Poster # 1: GENE-EXPOSOME INTERACTIONS CRITICAL FOR DEMENTIA

Presenting author: Kyung-An Han, PhD (University of Texas at El Paso)

Background: The overarching goal of our research is to identify the mechanisms by which genetic and exposome risk factors interact for Alzheimer's disease and related dementias (ADRD) using the model organism Drosophila melanogaster. While memory loss has been the focus of most ADRD research, memory analyses in animal models are time-consuming and labor-intensive. Furthermore, executive deficits, which are often seen in the early stages of ADRD, have been largely overlooked in research. Our study focuses on inhibitory control deficits as an early and measurable dementia trait, providing an innovative approach to discovering dementia-related genes and their interactions with exposome factors. Method: Through an unbiased genetic screen, we have identified fifteen genetic loci that cause age-dependent dysfunctions in inhibitory control. The presentation will be on one of the loci named scully (scu) encoding a multifunctional mitochondria molecule that binds Amyloid beta. Result: We found that the scu-deficient flies exhibit inhibitory control deficit and memory loss in an aging-dependent manner and interacts with the exposomes social stress and sleep loss. We identified the mushroom body as the major neural site for the Scu's role in the aging-associated cognitive decline. Scu/HSD17b10 is known to be involved in steroid homeostasis in mammals and tRNA processing in Drosophila. We found no genetic interaction of scu with t-RNA processing enzyme genes but found strong interaction with ecdysone, a major steroid hormone in Drosophila. Notably, scu deficiency augments Ab and Tau pathologies. Conclusion: Our findings not only advance the knowledge of how scu contributes to dementia and ADRD pathogenesis but also provide a novel in vivo model system to validate GWAS and omics data obtained from human subjects.

Funding: This work was supported by the Orville Edward Egbert, M.D. Endowment fund and NIH NIGMS 1R16GM145548 and NIMHD 3U54MD007592-29S grants.; COI: None

Poster # 2: CIRCUIT DYNAMICS IN ALZHEIMER'S DISEASE: SCOPIOUS AND PARTIAL OSCILLONS

Presenting author: Nastaran Lotfi, PhD (UTHealth Houston)

Alzheimer's disease (AD) causes progressive degeneration of neuronal populations, leading to profound cognitive impairment, irreversible memory loss, and other debilitating symptoms. To date, most research has emphasized cellular and cognitive aspects of AD, while circuit-level mechanisms have received comparatively less attention due to their inherent complexity. Many properties of network dynamics are emergent, i.e., arising from interactions among neurons and synapses that individually do not exhibit such behavior, so conceptual dissection with advanced analytical tools is required. Crucial information about network dynamics, including desynchronization and hypersynchrony, is contained in electrophysiological signals such as subcortical local field potentials (LFPs) and scalp electroencephalography (EEG). These signals have intricate structure that is often obscured by common analysis methods. We have developed a powerful signal-processing technique, the Discrete Padé Transform (DPT), which reveals a new level of organization in LFPs and resolves fine, transient features that traditional approaches can miss. DPT shows that LFPs comprise a small number of discrete, frequency-modulated waveforms-oscillons-embedded in weak noise. We hypothesize that oscillons are the true physical oscillatory motifs of synchronized neural activity, with familiar Fourier-based θ- and γ"bands" rerpesenting coarse approximations to these underlying components. Recent extensions of DPT enable simultaneous analysis of many channels, allowing identification of spatially extended, scopious oscillons in human and rodent circuits. Together, these tools open new avenues for studying multifaceted alterations in circuit dynamics caused by AD and related dementias (ADRD), enabling robust electrophysiological biomarkers in humans and animal models and potentially revealing early indicators of conversion from mild cognitive impairment (MCI) to AD.

Funding: None reported.; COI: None reported.

Poster # 3: CIRCADIAN RHYTHM DYSREGULATION AS A MODIFIABLE RISK FACTOR FOR AGE-RELATED COGNITIVE IMPAIRMENT

Presenting author: Hannah Smith, PhD (Texas A&M UHSC)

Fragmentation of the sleep-wake cycle and reduced amplitude of circadian rhythms precede cognitive impairment during aging and in Alzheimer's disease and related dementias
(AD/ADRDs), suggesting that alterations in circadian rhythmicity may contribute to age-related declines in learning and memory. However, the extent to which-and how-early circadian rhythm dysregulation affects cognitive aging remain unclear. Increasing evidence indicates that sleep and circadian disturbances may precede cognitive decline, underscoring the need to elucidate the mechanisms linking circadian dysregulation to cognitive function. Building on our previous findings of long-term B cell expansion following an established light-dark (LD) cycle-shifting paradigm, we examined whether controlled circadian disruption (CCD) acutely affects both the adaptive immune system and cognitive performance. Young C57BL/6J mice (3-6 months old) were subjected to an 80-day CCD paradigm consisting of repeated 12-hour advances of the LD cycle every 5 days. One week after the final shift, cognitive performance was evaluated using the Barnes maze, along with additional behavioral assessments including the three-chamber sociability, open field, and novel object/location recognition tests. Immune cell activation was also analyzed in peripheral tissues and the brain. Our results show that CCD induces significant acute alterations in the adaptive immune system of young mice. While cognitive and social behavioral changes were present, they were relatively mild. These findings suggest that circadian desynchronization rapidly impacts immune function, potentially setting the stage for long-term immune and cognitive consequences associated with sleep and circadian rhythm disturbances.

Funding: Janell and Joe Marek '57 Alzheimer 's disease Research Fund, WoodNext Foundation, TARCC, and DOD; COI: None

Poster # 4: APOE4-INDUCED PATTERNED BEHAVIORAL DECLINE AND NEURODEGENERATION REQUIRES ENDOGENOUS TAU IN A C. ELEGANS MODEL OF ALZHEIMER'S DISEASE

Presenting author: Jon Pierce, PhD (UT Austin)

Background: Alzheimer's disease (AD) causes a characteristic spatiotemporal pattern of neurodegeneration, resulting in the loss of associated faculties such as cognition. The factors which account for this pattern of degeneration are unclear, as AD risk genes are numerous and broadly expressed. Previously, we generated a model of AD using the nematode Caenorhabditis elegans in which the AD risk variant of apolipoprotein E, APOE4, is pan-neuronally expressed. We showed that HSN class motor neurons degenerate in early adult. Method: Here, we expand on our past work by performing behavioral analyses to deduce the effect of APOE4 on the function of distinct neuronal circuits. Results: We found evidence that APOE4 induces dysfunction of other neurons, suggesting a spatiotemporal pattern of degeneration that roughly correlates with endogenous levels of ptl-1, the C. elegans homolog of human MAPT. Moreover, deletion of ptl-1 suppressed dysfunction of multiple behaviors, suggesting broad protective effects across the nervous system including the HSN neurons. Lastly, we show that PTL-1 in the touch receptor neurons, where PTL-1 is most abundant, is required for degeneration of the HSN neurons. Conclusion: Our results suggest that C. elegans may provide a useful in vivo system to study how endogenous Tau acts downstream of APOE4 to cause age-related patterned neurodegeneration.

Funding: NIH (R01GM122463, RF1AG057355, R21OD032463 - J.T.P.) Waggoner Fellowship for Alcohol Research (J.T.P.), David & Ellen Berman Fellowships for Huntington's Research (J.T.P.), George and Karen Casey (J.T.P.), Pine Family Foundation (J.T.P.), CNS Catalyst Grant (J.T.P. & E.S.C.), National Institutes of Health (NIH) NIGMS (R35GM138340) (E.S.C.), Welch Foundation (F-2133-20230405) (E.S.C.), and F.M. Jones and H.L. Bruce Graduate Fellowship (Z.W.).; COI: None

Poster # 5: DEVELOPING A METHOD FOR MULTI-OMIC IMAGING OF ALZHEIMER'S MOUSE MODELS AND HUMAN BRAIN TISSUE USING MALDI AND DESI MSI

Presenting author: Bandana Bera, MS (UTMB)

Background: Alzheimer's disease (AD) is increasingly recognized as a neurodegenerative disorder characterized by profound lipid and metabolic dysregulation, which plays a critical role in its pathogenesis. Mass spectrometry imaging (MSI) provides a powerful approach for mapping the spatial distribution of biomolecules in tissues, enabling the study of region-specific molecular alterations. However, conventional imaging methods often lack the spatial and molecular resolution required to capture the full spectrum of these intricate biochemical changes within the brain. Methods: To overcome these limitations, we developed an integrated multi-omic imaging workflow that combines Matrix-Assisted Laser Desorption/Ionization (MALDI) and Desorption Electrospray Ionization (DESI) MSI. Frozen brain tissue sections from transgenic AD mouse models and post-mortem human samples were analyzed using a serial MALDI-DESI workflow. Optimized matrix deposition and solvent systems were employed to enhance ionization efficiency across molecular classes. MALDI-MSI was used for spatial profiling of complex lipid species, while DESI-MSI, optimized for 20 µm near-single-cell resolution, enabled detailed mapping of small polar metabolites such as γ-aminobutyric acid (GABA), glutamine, and glutamate under ambient conditions. Results: The combined MALDI and DESI imaging approach enabled visualization of spatial lipidomic alterations and corresponding metabolic changes during AD progression, in both mouse models and human brain tissue. This dual-mode strategy provided comprehensive coverage of region-specific molecular perturbations associated with AD pathology. The MSI data was complemented by immunofluorescence (IF) and hematoxylin and eosin (H&E) staining of adjacent tissue sections, providing essential histological and cellular context for molecular distributions. Conclusion: This study establishes MALDI-DESI MSI as a robust, multimodal platform for the spatial mapping of targeted and untargeted lipids, metabolites, and neurotransmitters in brain tissue. The high spatial resolution and molecular specificity of this workflow offer new insights into the biochemical heterogeneity of AD, advancing our understanding of disease-associated metabolic dysregulation and supporting the identification of potential biomarkers for diagnosis and therapeutic targeting.

Funding: National Institutes of Health (NIH) grand (1S10OD038289-01) for DESI MIS Instrument; COI: No conflict of interest

Poster # 6: CHARACTERIZATION OF TAU OLIGOMER-INDUCED DYSFUNCTION IN A HUMAN IPSC-DERIVED CEREBRAL ORGANOID MODEL

Presenting author: Kevin Johnson, PhD (UTMB)

Background: Cerebral organoids replicate much of the complexity, organization, and connectivity of the human brain, and can be reliably generated and reproduced from induced pluripotent stem cells (iPSCs) derived from a single patient. This technology enables high throughput generation of real patient data under highly controlled, non-invasive conditions. Until recently, cerebral organoids were primarily used to study developmental neurogenesis and specific genetic conditions. However, as the field of AD and dementia related research begins to shift away from animal studies, cerebral organoids will serve a vital function in disease modeling and therapeutic testing for AD. Here we use a cerebral organoid model derived from a single control iPSC line to characterize histopathological and electrophysiological effects of tau oligomers (TauO) to create a template for the evaluation of patient-derived cerebral organoids and their capacity for innate or induced resilience against AD neuropathology. Methods: 100+ days in vitro cerebral organoids were exposed to recombinant TauO at varying concentrations. Exposed organoids were fixed, sectioned, and immunolabeled to assess the propagation of tau pathology. Additional organoids were mounted on multielectrode arrays and spontaneous electrical activity was recorded over a period of one week. Results: Our findings indicate that addition of exogenous oligomeric tau to cerebral organoids induces dissociation of microtubule-associated tau leading to peri-nuclear aggregation and hyperphosphorylation of endogenous tau. We also show that 50nM TauO increases electrical activity within 24 hours compared to baseline, while 250nM TauO silences activity. Conclusions: Our findings suggest that human iPSC-derived cerebral organoids effectively recapitulate the propagation of tau pathology found in the progression of AD. These models can be a valuable tool for elucidating mechanisms of neuronal resilience against AD, and allowing high throughput drug screening on an individual patient basis.

Funding: The Moody Brain Health Institute at UTMB; COI: The authors declare no conflicts of interest

Poster # 7: INTERPRETABLE MACHINE LEARNING FOR COGNITIVE AGING: HANDLING MISSING DATA AND UNCOVERING SOCIAL DETERMINANT

Presenting author: Xi Mao (UTRGV)

Early detection of Alzheimer's disease (AD) is critical because its neurodegenerative effects are irreversible, and neuropathological change as well as modifiable social-behavioral risk factors accumulate years before clinical diagnosis. Identifying higher-risk individuals earlier enables prevention, timely care, and more equitable resource allocation. We study prediction of cognitive performance from social determinants of health (SDOH) using the NIH NIA supported PREPARE Challenge Phase 2 dataset derived from the nationally representative Mex-Cog cohort from the 2003 and 2012 Mexican Health and Aging Study (MHAS). Data: The target is a validated composite cognitive score in seven domains: orientation, immediate and delayed memory, attention, language, constructional praxis, and executive function, which derived from 2021 and 2016 MHAS. We curated features across demographic, socioeconomic, health status, lifestyle, psychosocial, and healthcare access. Methodology: These feature domains were selected to capture the multidimensional nature of social and behavioral influences on cognitive aging. Substantial missingness was addressed with a singular value decomposition (SVD)-based data imputation pipeline that treats continuous and categorical variables separately. This approach leverages latent correlations among features to recover missing values and offers a strong balance between statistical reliability and scalability. Through comparative evaluation of multiple methods, XGBoost was selected as the default prediction model due to its superior performance. Results: Our results show that the proposed framework outperforms existing methods and achieves better predictive accuracy than the top results reported on the data challenge leaderboard. To better understand the relationship between input features and the composite cognitive score, we conducted a thorough post hoc analysis of the top contributing features, examining the mechanism by which these features are associated with cognitive scores. The study further stratified the analysis by age group based on SHAP analysis to explore whether the most predictive features differ across life stages. The proposed pipeline outperformed the state-of-theart methods, demonstrating robustness, interpretability, and computational efficiency, and underscoring its potential as a practical modeling strategy for datasets with substantial missingness in both continuous and categorical features. [Manuscript:https://arxiv.org/abs/2510.10952]

Funding: UTRGV Travel Resources and Investments Program (TRIP) UT Health San Antonio

CAPAS center; COI: NA

Poster # 8: SKULL BONE MARROW IS ACTIVATED TO PRODUCE A PROINFLAMMATORY RESPONSE AND SOURCE OF NEUROINFLAMMATION IN ALZHEIMER'S AFFECTED BRAINS

Presenting author: Iva Lazarova, BS (UTSW)

Background: Alzheimer's disease (AD) is known to cause neurodegeneration through aggregation of proteins such as Aβ. Research has also revealed neuroinflammation, indicating that despite the brain's immune privileged status, it communicates inflammatory signals to the periphery. The glymphatic system has been found to have direct CSF channels into the subarachnoid space and skull bone marrow niches, bypassing peripheral vasculature. We propose that an AD brain can induce skull bone marrow activation to secrete Reelin and other peripheral immune factors. Method To determine differences in gene expression and cell populations 5xFAD mice and wild type littermates (WT), were used in this study. For qRT-PCR and bulk RNA sequencing RNA was extracted from 4-month-old mouse skulls and libraries were prepared and run at the University of Texas Southwestern Genomics and Microarray Core. Analysis was performed with Galaxy online resource. Single-cell RNA sequencing (scRNA-seq) was performed on cells isolated from the skull and femur of 3 and 6-month-old 5xFAD Mice and their WT littermates. Barcoding and library creation was done by Parse Evercode WT kit v3.

Sequencing was performed by Novogene Co, Ltd. Trailmaker (Parse Biosciences) and PangaloDB were used for data analysis. Result Bulk RNA sequencing of 5xFAD and WT mice revealed increases in Reelin and multiple immune factors in skull bone marrow, including pro-inflammatory cytokines. The scRNA-seq on skull bone marrow from 5xFAD mice showed populations changed from a neutrophil-oriented population of acute response to an adaptive one containing pre-B cells. A currently unknown population was present in 3-month-old 5xFAD mice, which we hypothesize to be a unique endothelial cell group. The scRNA-seq also confirmed higher Reelin and immune factor expression in 5xFAD skull bone marrow at 3 months. Conclusion Data shows that there is increased inflammatory activity driven by the skull bone marrow in AD mouse models, which can be seen as an increase in Reelin, a proinflammatory phenotype, and immune cell population changes in the skull bone marrow niche. Therefore, skull bone marrow may be a component of neuroinflammatory AD pathophysiology and a target for therapeutics.

Funding: TARCC Deutsche Forschungsgemeinschaft National Heart, Lung, and Blood Institute National Institute on Aging National Institute of Neurological Disorders and Stroke BrightFocus Bluefield Project to Cure FTD Harrington Discovery Institute Circle of Friends Pilot Synergy Grant, Alzheimer's Association Presbyterian Village North Foundation Robert J. Kleberg Jr. & Helen C. Kleberg Foundation. NIH/NIAMS SBIR to Reelin Therapeutics, Inc. ; COI: L. Calvier, M.Z. Kounnas, and J. Herz are shareholders of Reelin Therapeutics, Inc. L.C. and J.H. are co-inventors of a patent related to anti-Reelin strategies (application number 15/763,047 and publication number 20180273637).

Poster # 9: INVESTIGATING CONFORMATIONAL DIVERSITY OF TAU INSIDE MAMMALIAN CELLS

Presenting author: Rupam Ghosh, PhD (UTSW)

Background: The conformational landscape of metastable proteins such as tau is highly sensitive to the cellular milieu, including molecular interactors, macromolecular crowding, and posttranslational modifications. These proteins exhibit complex folding landscapes characterized by multiple local energy minima. In neurodegenerative diseases such as Alzheimer's disease (AD), the aggregating forms of tau often share identical primary sequences with those present in healthy individuals. Despite the importance of the cellular environment, most high-resolution structural studies have been conducted either in vitro using purified proteins or ex vivo using patientderived fibrils. Consequently, these approaches fail to replicate the dynamic and heterogeneous conditions of the intracellular environment. Moreover, structural insights obtained from such studies are typically biased toward thermodynamically stable, sarkosyl-resistant species, thereby neglecting the broader ensemble of conformations that exist inside cells. Our research aims to characterize the conformational ensemble of tau within its native biological context to elucidate the structural basis of conformation-dependent pathologies such as AD. This understanding will identify biologically relevant conformations with potential diagnostic and therapeutic significance. Methods: Dynamic Nuclear Polarization (DNP)-enhanced solid-state NMR spectroscopy provides exceptional sensitivity for detecting proteins at endogenous concentrations within their native cellular environment. Building on our recently developed in-cell DNP NMR methodology, we have previously characterized the conformational ensemble of the intrinsically disordered protein α-synuclein inside living cells, revealing how specific intracellular conditions, locations and seeding with pre-formed fibrils modulate its conformational populations. Extending this approach, we are now leveraging in-cell DNP NMR to investigate the conformational ensemble of tau within mammalian cells. Results: Our findings demonstrate that monomeric wild-type tau adopts at least two distinct conformations in mammalian cells. These conformations are strongly influenced by stoichiometric interactions and macromolecular crowding. Ongoing studies aim to delineate the molecular determinants underlying this conformational diversity and to assess how the population distribution changes upon seeding with ex vivo patient-derived fibrils.

Funding: NIH, O'Donnell Brain Institute, Welch Foundation; COI: None

Poster # 10: DNP NMR REVEALS CELLS SELECTIVELY AMPLIFY A DISTINCT AMYLOID CONFORMATION

Presenting author: Shoyab Ansari, PhD (UTSW)

Neurodegeneration in Parkinson's disease (PD) is correlated with Lewy bodies, intracellular inclusions containing fibrillar aggregates of α-synuclein (α-syn). While α-syn assembles into diverse fibrillar forms in vitro, its relevance to disease remains unclear. Additionally, the intrinsically disordered regions that flank the amyloid core are challenging to resolve with diffraction-based methods. We used DNP NMR to investigate the amyloid fibrils resulting from seeded propagation in vitro and inside cells. We found that the amyloid fold and the conformational preferences of the flexible regions are faithfully propagated in vitro and in cellular lysates. However, during the seeded propagation of amyloids inside live cells, the minor conformation in the seeded population became predominant and more ordered. Additionally, the conformational preferences of the flexible regions were also altered inside the living cells. This suggests that fibril conformation changes as it propagates inside living cells and differs from the in vitro conformation.

Funding: RF1 NS143938 R01 NS134921 R21 NS136951 ; COI: NONE

Poster # 11: THE IMPACT AND PERCEPTIONS OF POLYGENIC RISK SCORE DISCLOSURE IN ALZHEIMER'S DISEASE

Presenting author: Fatima Chavez, MPH (BCM)

Background: Polygenic risk scores (PRS) quantify cumulative genetic risk for Alzheimer's disease (AD). We aimed to evaluate participant responses to PRS disclosure and identify communication lessons for clinical use. Methods: Sixty participants underwent whole genome sequencing and APOE genotyping. PRSs for AD and coronary artery disease (CAD) were generated using Allelica DISCOVER and PREDICT v3.1.6 platforms, excluding APOE for AD. Results were adjusted to ancestry-specific thresholds and reported as average or high risk (≥ 2X relative risk). One year after disclosure of other genetic test results, participants received structured counseling regarding their PRS. Post-disclosure surveys assessed understanding, emotional responses, perceived utility, and preferences regarding genetic risk information. Results: We disclosed AD and CAD PRS to 60 adults enrolled in as dementia risk assessment study (cognitively normal, n = 29; mild cognitive impairment, n = 21; AD dementia, n = 7; others, n = 3). Ten (20%) and 21 (35%) had high/elevated AD and CAD risks, respectively. Interestingly, nearly all participants with high AD PRS did not carry an APOE ε4 allele (n = 9, 90%), underscoring added value beyond traditional risk markers. Most participants reported understanding their PRS (95%) and found it useful for planning (90%); 83% endorsed broader PRS use in AD risk assessment. Emotional responses were largely positive (68%), with participants feeling prepared and glad to receive results, though 5% desired more preparation. Disclosure was feasible across cognitive groups without major adverse reactions. Conclusion: PRS can be integrated into AD genetic testing, enhancing risk communication and patient engagement. Larger-scale disclosure studies will be needed to confirm these findings and refine best practices.

Funding: 1. Baylor College of Medicine Learning Health System Scholarly Time Grant; COI: N/A

Poster # 12: CELL TYPE-SPECIFIC ALTERNATIVE POLYADENYLATION IN ALZHEIMER'S DISEASE

Presenting author: Yordi S. Tamayo-Molina, BS (Texas Tech HSC)

Background: Alzheimer's disease (AD) is an incurable human neurodegenerative disorder with a continuously increasing prevalence, and that is characterized by progressive cognitive and functional decline. Most gene expression studies on post-mortem AD brains have focused on differentially expressed genes as contributors to disease onset and progression, often assuming a linear correlation between mRNA and protein levels. Here, we investigated the mechanism of alternative polyadenylation (APA) as a possible contributor to gene expression regulation in AD. APA generates mRNA isoforms with distinct 3′ untranslated regions (3′UTRs). By altering the length and composition of 3′UTRs, APA can influence mRNA stability, localization, and translational efficiency. In addition to 3′UTRs, APA can also occur within intronic sequences, producing mRNA splice variants with truncated coding regions or altered regulatory potential, thereby modulating transcriptomic landscape and regulating protein expression. Importantly, dysregulation of APA has been implicated in various physiological processes and diseases, including cancer and neurological disorders, e.g., Parkinson's disease, but very little is known about this mechanism in the regulation of brain gene expression in the AD pathophysiology. Method: We have performed bioinformatic re-analysis of existing snRNA-Seq datasets from ageand gender-matched AD and cognitively unimpaired donors using existing bioinformatic tools to determine the frequency and type of APA changes in a cell type-specific manner. Result: We found that APA occurs at AD in a cell-type specific manner within the prefrontal cortex of AD patients. Our results suggest that APA is more prevalent in female than in male donors. Moreover, APA appears to affect different cell types depending on sex. For example, APA is most commonly observed in neurons in males vs. astrocytes in females. The identified APAassociated genes can be categorized as follows: 1) genes already linked to the onset, development, and progression of AD, and 2) genes not previously associated directly with AD but implicated in neurodevelopment. Conclusion: We propose that alternative polyadenylation acts as an additional modulator of gene expression in a cell type-specific manner in AD. We are currently refining advanced semi-automated methods to better analyze APA in existing RNA-Seq datasets and in AD samples and models.

Funding: NIH, National Institute on Aging R03AG087283 and startup funds from TTUHSC to PNG.; COI: None

Poster # 13: RAAS-APOE EPISTASIS AS A POPULATION SPECIFIC RISK FACTOR FOR COGNITIVE IMPAIRMENT

Presenting author: Amanda Tucker, BS (UNTHSC)

Background: The strongest genetic risk factor for development of Alzheimer's disease (AD) and/or AD-related dementias (ADRD) is the ε4 allele in the apolipoprotein E (APOE) gene, though the frequency and impact of this variant on cognitive decline can vary across populations. Vascular disorders (e.g. hypertension or cardiovascular disease) are also strong risk factors for AD/ADRD. These are particularly impactful in non-White populations and may play a role in differential AD outcomes. A multiethnic cohort study found that controlling for vascular disorders mediated the AD health disparity observed across populations, suggesting vascular health has a significant influence on cognitive functions across these groups. Genetic variants in angiotensin converting enzyme 1 (ACE1), ACE2, and angiotensinogen (AGT) are risk factors for vascular disorders due to the prominent role they play in the renin-angiotensin aldosterone systems (RAAS) regulation of vasculature tone and function. Recent studies suggest protein interactions occur between ACE2 and APOE proteins, making ACE2 and upstream RAAS mediating genes plausible candidates for epistatic studies connecting vascular disorders and AD/ADRD risks. This study aims to elucidate single nucleotide polymorphisms (SNPs) in genes of the RAAS pathway to identify candidate genes for epistasis studies with APOE. We hypothesize that RAAS genetic loci have population specific SNPs associated with hypertension and AD/ADRD when found in epistasis with APOE. Methods: Formal epistasis analysis was conducted on the Texas Alzheimer's Disease Research Care and Consortium dataset (TARCC) using PLINK, by Mexican-Americans or Non-Hispanic White populations and controlling for APOE ε4 status. Population specific epistasis profiles were generated to group epistatic interactions by their effects with AD status, MCI status, or the combined effect connecting AD or MCI status with a clinical phenotype of hypertension. Conclusion: We identified population specific epistatic interactions between the APOE and the AGT or ACE1 regions that confer genetic risk of AD or MCI in Mexican Americans and non-Hispanic Whites. We also identified population specific interactions that connect vascular functions with AD or MCI in non-Hispanic Whites. This study utilizes epistasis to define vascular co-morbidities with the APOE ε4 AD risk factor to improve current understanding of nuanced, population specific risk for AD.

Funding: This study is funded by the National Institute on Aging training fellowship (T32 AG020494), & IMSD Fellowship, Grant # 5 R25 GM125587-05 from the National Institutes of General Medical Sciences (NIGMS).; COI: This research is protected under provisional patent patent UNTX.P0013US.P1.

Poster # 14: INVESTIGATING GENETIC BASIS OF THE INVERSE ASSOCIATION BETWEEN CANCER AND ALZHEIMER'S DISEASE

Presenting author: Dipti Debnath, M.Pharm (UNTHSC)

Background: Epidemiological studies have reported that patients with cancer show lower Alzheimer's Disease (AD) likelihood, while AD patients exhibit reduced cancer prevalence. To identify mechanisms underlying this inverse association, we investigated genetically regulated gene expression to identify loci and genes that were genetically correlated with inverse risk. Methods: We analyzed GWAS data for eight cancer (N= 85,716- 448,150) types (lung, breast, prostate, colorectal, bladder, endometrial, melanoma, and basal cell carcinoma) and late-onset clinical AD (N= 290, 088). Global and local genetic correlations (rg) were assessed using LD score regression (LDSC) and LAVA across 2,495 LD-independent regions. We performed a transcriptome-wide association study (TWAS) integrating tissue specific gene expression weights using GTEx (Genotype Tissue Expression) and TCGA (The Cancer Genome Atlas) by FUSION. We used the STRING database to construct protein-protein interaction (PPI) networks and then conducted pathway enrichment analysis on the biggest protein cluster using g:Profiler. Results: Local genetic correlation analysis identified key genomic regions: the APOE locus on chromosome 19 at 19q13.31-32 region showed consistent negative correlations with five cancer types (breast: rg=-0.31, p=0.036; prostate: rg=-0.25, p=0.029; lung: rg=-0.38, p=0.002; melanoma: rg=-0.33, p=0.021; endometrial: rg=-0.31, p=0.006). Integrating gene expression weights from TCGA's Lung adenocarcinoma tissue, we additionally identified genes inversely regulated between AD and lung cancer (CHRNA5 LC.Z=-15.7162, AD.Z= 2.36; CKM LC.Z=2,77, AD.Z= 8.65). Using lung tissue gene expression, we also observed opposing Z scores of CHRNA5 between AD and cancer (LC.Z=-7.63, AD.Z=2.51). We identified 15 pathways related to lung cancer and AD from pathway enrichment analysis. Conclusions: The local rg identified consistent inverse effect sizes between AD and multiple cancer types at the APOE region, suggesting shared biological pathways with opposing effects. Similarly inverse effects at CKM, known to encode creatinine kinase is near the APOE-MARK4 region. CHRNA5 in a nicotinic receptor binding protein wherein variants have opposing effect on gene expression. These inverse effect estimates provide novel insights into the differences of the genetic architecture between AD and cancers.

Funding: Research reported in this abstract was supported by the Alzheimer's Association and the National Institute on Aging of the National Institutes of Health under Award Numbers; AARF-22-967171 (GAP); R00AG078503 (GAP), R01AG070862 (RCB), (NRP), HABLE; R01AG058533 and HABS-HD; U19AG078109 (Dr. Sid O'Bryant). ; COI: NA

Poster # 15: MITONUCLEAR GENE ANCESTRY AND ASSOCIATIONS WITH ALZHEIMER'S DISEASE RISK AND BIOMARKERS

Presenting author: Jiya Ghei (UNTHSC)

Background: Alzheimer's disease (AD) is a multifactorial neurodegenerative disorder influenced by genetic, vascular, and metabolic factors which may differ across ancestral backgrounds. Previous studies suggest that nuclear-encoded mitochondrial (mitonuclear) genes are crucial mediators of metabolic pathways involved in AD and may contribute to variation in risk factors and blood biomarkers across populations. This study examines the relationship between local ancestry at mitonuclear genes and AD risk factors and biomarkers in a cohort of Mexican Americans, African Americans, and non-Hispanic Whites. Methods. Genetic and phenotypic data were obtained from participants in the Health and Aging Brain Study - Health Disparities (HABS-HD). Genome-wide single nucleotide polymorphism (SNP) data were used to determine local ancestry segments in mitonuclear gene regions. These local ancestry segments were summarized into regional ancestry dosage matrices for each individual. Admixture mapping was performed, using linear mixed models to analyze the associations between local ancestry at mitonuclear gene regions and AD biomarkers and vascular risk factors. Models were adjusted for relevant demographic and genetic covariates and multiple testing correction was conducted with significance thresholds commonly used in admixture mapping studies. Results. Our preliminary analyses indicated significant admixture associations between local ancestry at mitonuclear gene regions and both hypertension and plasma phosphorylated tau (ptau181) levels. These findings suggest that genetic ancestry may influence traditional vascular risk factors for AD, such as hypertension, as well as molecular AD biomarkers like ptau181. Conclusion. These results highlight the importance of investigating the relationship between genetic ancestry and AD risk and progression. As additional data from the HABS-HD cohort becomes available, our analysis will expand to include additional clinical and genetic data for larger and more diverse populations. Additionally, future analyses will assess the impact of mitonuclear discordance on metabolic phenotypes and AD outcomes.

Funding: NIH/NIA 1U19AG078109-01, Contact PI- Sid O'Bryant; COI: N/A

Poster # 16: COMPARISON OF EPIGENETIC CLOCK PERFORMANCE AND DRIVERS OF AGE ACCELERATION IN MEXICAN AMERICANS AND NON-HISPANIC WHITES

Presenting author: Kahylen Minniefield, BS (UNTHSC)

Background: Epigenetic clocks, which use DNA methylation patterns to predict a person's biological age, have become widely used biomarkers for age-related disease risk. Previous research has shown significant associations between accelerated epigenetic aging and Alzheimer's Disease (AD). These epigenetic clocks were trained primarily on non-Hispanic white (NHW) cohorts. However, both AD-related pathologies and DNA methylation patterns vary across ethnic groups. This could lead to inaccurate measurements of epigenetic age acceleration and biased associations with AD-related biomarkers in other populations. Here we aim to 1) test the accuracy of several population epigenetic clocks between a NHW and Mexican American (MA) cohort and 2) investigate the associations between epigenetic age acceleration and AD-related biomarkers in these two groups. Methods: Previously collected genome-wide DNA methylation data obtained from peripheral blood of 551 participants (N = 299 MAs, N = 252 NHWs) from the Texas Alzheimer's Research and Care Consortium was used for this study. Epigenetic age estimates were calculated using five commonly used epigenetic clocks (Horvath, skinHorvath, Hannum, PhenoAge, GrimAge) and compared between groups. We then used generalized linear models to investigate associations between estimates of age acceleration and AD-related biomarkers. Results: We found no significant differences in epigenetic age estimates between NHW and MA groups for any of the clocks tested. However, we do find different biomarkers driving variation in age acceleration between the two groups. Higher levels of thymus and activation-regulated chemokine (TARC) are associated with increased age acceleration across 4 out of 5 clocks (excluding GrimAge; all p < 0.043), only for the NHW group. Likewise, higher levels of coagulation factor VII were associated with increased age acceleration across all 5 clock estimates (all p < 0.038) only in the MA group. Conclusion: Our results suggest that there is no bias in epigenetic age estimates between NHWs and MAs for the epigenetic clocks tested. However, different biomarkers are associated with variation in accelerated aging estimates for each group. Coagulation factor VII is of particular interest, as genetic variation in this gene has previously been associated with AD risk.

Funding: Texas Alzheimer's Research and Care Consortium; COI: COI: None

Poster # 17: IDENTIFICATION OF ALTERED DNA METHYLATION MARKERS IN ALZHEIMER'S DISEASE THROUGH GENOME-WIDE ARRAY-BASED ANALYSIS

Presenting author: Kumudu Himali Subasinghe, PhD (UNTHSC)

Background: Alzheimer's disease (AD) is the most prevalent form of dementia, accounting for the majority of dementia cases worldwide. It is a progressive neurodegenerative disease that disproportionately affects certain racial/ethnic groups, including Mexican Americans (MAs) compared to their non-Hispanic white (NHW) counterparts. Despite recent advances in therapeutic approaches, complex molecular mechanisms underlying the disease progression and its population-specific pathology are still understudied. Epigenetic mechanisms have recently emerged as a promising tools for elucidating the biological basis of complex diseases such as AD. Among these, DNA methylation (DNAm) is one of the most extensively studied epigenetic modifications. DNAm modifies gene expression by adding a methyl group to cytosines adjacent to guanines (CpG sites). This process can regulate gene activity through gene silencing and has been implicated in various biological processes and disease pathologies, including AD. Identifying CpG sites with differential methylation associated with cognitive imparment (CI) and population specificity may help uncover the molecular basis of observed disparities in AD risk and progression. Objective. This study aimed to identify CI-relevant CpG sites with population specific methylation petterns in MAs and NHWs using genome-wide DNA methylation microarrays. Methods. DNA was extracted from the buffy coat samples of the participants using the Hamilton robotic system with Omega Mag-Bind Blood & Tissue DNA HDQ 96 Kit. Between 250-500 ng of DNA was bisulfite-converted using the Zymo EZ DNA Methylation kit. Genomewide DNAm profiling was performed using the Illumina MethylationEPIC BeadChip (v1 and v2). Raw IDAT files were processed in R (v4.5.1) using minfi with preprocessNoob normalization to generate β-values. Post-QCed and ComBat adjusted for batch, array, and array version were analyzed using limma, controlling for age, APOE, education and sex. Analyses were conducted separately for each population and visit. CpG sites with nominal significance (p<0.05) and consistent direction across visits were mapped to genes and analyzed in IPA. Results. The top enriched protein network in both populations was anchored on AKT dysregulation, showing opposite predicted effects between groups. Canonical pathway enrichment analyses revealed distinct population-specific pathways. Conclusion: These findings highlight epigenetic differences underlying CI and emphasize population-specific mechanisms in AD pathogenesis.

Funding: None reported.; COI: None reported.

Poster # 18: IDENTIFYING APOLIPOPROTEIN E-GLYCOSAMINOGLYCANS INTERACTIONS FOR DISEASE-MODIFYING THERAPIES: RELEVANCE TO ALZHEIMER'S DISEASE

Presenting author: Ashlyn Mattern, BS (UTMB)

Background: Alzheimer's Disease (AD) is a prevalent neurodegenerative disorder affecting about 7 million individuals in the US, with the number of diagnoses projected to double in the next couple of decades. Disease-modifying therapies remain limited, and often with severe sideeffects. The link between Apolipoprotein E (APOE), a protein that supports lipid transport and brain repair, and AD has been widely studied. There are three known naturally occurring variants: ApoE3, or physiologically neutral, ApoE2, proposed to be protective, and ApoE4, the known strongest risk factor for AD. It has been recently discovered a novel protective genetic variant of APOE, APOE3Christchurch, against an autosomal dominant form of AD. Heparan sulfate proteoglycans (HSPG) have been shown to bind to APOE in a domain impacted by the Christchurch mutation. This study has allowed us to identify novel ApoE- protein/glycosaminoglycans interactions essential in trying to mimic the protective mechanisms, thus leading to disease-modifying therapy for AD. Methods: We tested three HSPG interactions including Glypican 2, Glypican 4 and Glypican 5 with a alphaFold 3, pull-down assay, Western blotting, and enzyme linked immunosorbent assay (ELISA) to detect the interactions. The pull-down assay includes 2 conformations with beads used to detect the ApoE variants, or beads used to detect the Glypicans. We used western blotting to detect ApoE variants (anti-ApoE clone D6E10), or the Glypicans (anti-GPC2, anti-GPC4 and anti-GPC5). Using ELISA, we tested the binding of serial dilutions of the Glypicans to a constant concentration of ApoE variants. Results: Computational predictions and experimental validation using pulldown and ELISA assays revealed that Glypican 4 binds more to prokaryotic ApoE2, and eukaryotic ApoE4 compared to Glypican 2, whereas Glypican 5 has no interaction with ApoE variants. Conclusion: Our data suggests selective ApoE-interactions that support future in vitro validation using cellular models of protein expression, and in vivo evaluation of pharmacokinetics and pharmacodynamics profiles.

Funding: Start-up fund from Sealy & Smith Foundation to CM and the Marvin Legator Scholarship Award to AM.; COI: No conflict of interests.

Poster # 19: SEX-DEPENDENT ALTERATIONS IN CELLULAR EXCITATORYINHIBITORY BALANCE AND SYNAPTIC PATHWAYS DURING ALZHEIMER'S DISEASE PROGRESSION: A SINGLE CELL ANALYSIS

Presenting author: Laura Sanchez-Sanchez, PhD (UTMB)

Alzheimer's disease (AD) disproportionately affects women. Emerging evidence indicates that neuronal hyperexcitability, frequently observed during the early stages of AD and more prominent in women, is associated with progressive disruption of both synaptic and cellular excitatory-inhibitory (sE/I and cE/I) balance. However, it remains unclear whether alterations in the sE/I balance contribute directly to cell-type-specific vulnerability and subsequent cE/I imbalance in women or resilience in men. Because the sE/I equilibrium within excitatory and inhibitory neurons is essential for maintaining neuronal integrity, circuit stability, and overall brain function, gradual shifts in the gene expression of synaptic pathways may underlie the selective vulnerability of particular neuronal subtypes observed in AD. Using single cell RNAseq data from human medial temporal gyrus in the Seattle Alzheimer's Disease Brain atlas (SEA-AD), we aimed to investigate cell-specific differences in synaptic gene expression between women and men across disease stages of AD. To elucidate pathways that may contribute to the higher susceptibility of synaptic alterations in women, gene ontology (GO) enrichment analysis was performed on differentially expressed genes (DEGs). The proportions of excitatory and inhibitory neurons exhibited sex-dependent patterns along disease progression, with an early decline of Sst-positive neurons in men (ordinary least squares (OLS) -866.81 corr.coefeicient, pvalue <0.001 ****) and a late decline of Pvalb-positive neurons in women (OLS -470.50 corr.coefficient, p-value 0.076 ns). Men showed a higher number of DEGs in inhibitory neuronal populations (35854 DEGs) than women (16127 DEGs), along with a predominance of downregulated genes (90%). In contrast, Women exhibited fewer DEGs, and no significant predominance of up- or down-regulation (66% of Down-regulated DEGs). GO enrichment analysis showed that women presented upregulated synaptic pathways – particularly Glutamatergic synapse - within vulnerable neuronal populations in early stages, consistent with hyperexcitability previously described. These findings suggested a link between sE/I imbalance and sex-specific cell vulnerability and point toward potential protective mechanisms presented in males that may mitigate AD progression.

Funding: AG070255 AG073133 ; COI: No conflict of interest Nothing to disclose

Poster # 20: GENETIC ASSOCIATIONS WITH ALZHEIMER'S DISEASE USING THE ALL OF US RESEARCH PROGRAM DATABASE

Presenting author: Chun Xu, MD, MSc, PhD (UTRGV)
Alzheimer's disease (AD) is a progressive neurodegenerative disorder characterized by memory loss and cognitive decline, representing a major public health challenge worldwide. While aging remains the most significant risk factor, numerous studies have shown that genetic, environmental, and lifestyle factors jointly influence disease onset and progression. Identifying genetic variants associated with AD can provide critical insights into its underlying mechanisms and potential therapeutic targets. With the advent of large-scale genomic databases, such as the National Institutes of Health (NIH) All of Us Research Program, researchers now have access to diverse datasets that enable the discovery of genetic risk factors across populations that have been historically underrepresented in genetic studies. In this study, we utilized version 7 of the All of Us Research Program database through the secure Researcher Workbench. A total of 498 participants diagnosed with AD and 33,215 matched controls without AD were included. We focused on three chromosomal regions-1, 17, and 19-previously implicated in AD risk, analyzing 19 candidate genes and their single nucleotide polymorphisms (SNPs) reported in the literature. The curated SNP panel was derived from publicly available genomic references (www.allofus.nih.gov). Our results identified three SNPs showing the strongest associations with AD, all located within the APOE and APOC1 genes on chromosome 19, consistent with the well-established role of APOE ε4 in AD susceptibility.

Importantly, we detected a stop-gain mutation in the CD46 gene (NM_172351.3(CD46): c.719C>G, p.Ser240*), which may represent a previously unreported variant relevant to AD pathogenesis. CD46 is involved in immune regulation and complement activation, suggesting a potential link between immune dysregulation and neurodegeneration in AD. In conclusion, our findings confirm known genetic risk factors while identifying new potential candidates that merit further investigation. Functional and replication studies are needed to determine the biological significance of CD46 and its interaction with other AD-associated genes. This study demonstrates the value of large, diverse datasets like All of Us in advancing precision medicine and improving our understanding of AD genetics across populations.

Funding: NIH -Healthy America Foundation (HAF) Research Grant using All of US data; COI: The authors declare no real or perceived conflicts of interest related to this study.

Poster # 21: EXPLORING GLYCAN SULFATION AT THE BLOOD-BRAIN BARRIER IN ALZHEIMER'S DISEASE

Presenting author: Aric F. Logsdon, PhD (Texas Tech HSC)

Background: Alzheimer's disease (AD) is a major health problem affecting approximately six million people in the United States. Multiple lines of evidence suggest poor perivascular clearance of amyloid beta (Aβ) as a major contributing factor for the "vascular hypothesis" of AD. An important, yet understudied facet of Aβ clearance in AD is the change in sugar composition on the surface of the brain endothelial cell (BEC). Chondroitin sulfate (CS) glycosaminoglycans coat the BEC matrix with long, anionic chains of repeating disaccharide units that can be differentially sulfated to modulate interactions with proteins, such as Aβ. While altered brain CS sulfation has been strongly associated with AD, the impact of altered BEC CS sulfation on perivascular Aβ clearance in AD has yet to be evaluated. We hypothesize that CS hypersulfation augments perivascular Aβ accumulation and contributes, in part, to the onset and progression of AD. Methods: CS hypersulfation has been reported in human AD brain tissue, as measured by liquid chromatography tandem mass spectrometry (LC-MSMS). Detailed description of CS composition at the cellular level has thus far been limited by the non-specificity and constrained sensitivity of traditional histochemical approaches. Leveraging highly specific and sensitive LC-MSMS methods that enable robust quantification of CS sulfation we will comprehensively profile CS glycans on BECs isolated from human brain tissue with confirmed AD pathology. Furthermore, this project explores CS sulfation as a mechanism of Aβ clearance binding and transport kinetics using an in vitro model of the human blood-brain barrier (BBB). Results: Isolated BECs demonstrated a unique CS sulfation pattern in relationship to their corresponding brain parenchyma partition. BECs isolated from human AD brain tissue exhibited CS hypersulfation, compared to non-AD controls. Enzymatic inhibition of CS sulfation reduced A binding to human BECs and altered human BBB function in vitro. Conclusions: Successful results will establish innovative LC-MSMS methods to identify CS sulfation patterns on human BECs to be strongly associated with AD-related pathology. Collectively, these outcomes will showcase an underappreciated pathway in the AD pathogenesis and identify CS sulfation as a novel therapeutic target to enhance perivascular Aβ clearance from the AD brain.

Funding: NIH-NIA K22AG081264 TTUHSC ORI (Startup Funding) TTUHSC PN (Startup Funding) TARCC Junior Investigator Research Grant; COI: N/A

Poster # 22: A FRAMEWORK FOR QUANTIFYING CEREBROVASCULAR PATHOLOGY USING NACC NP11 DATA

Presenting author: Angelique D. Gonzalez, BS (UTH San Antonio)

Background: Cerebral small vessel disease (SVD) is one of the most common contributors to vascular cognitive impairment and dementia (VCID) and often coexists with Alzheimer's disease neuropathologic change (ADNC). Yet, despite its prevalence, researchers still lack standardized, quantitative ways to measure cerebrovascular pathology across large, multicenter autopsy datasets. The National Alzheimer's Coordinating Center (NACC) Neuropathology Form 11 (NP11) captures valuable cerebrovascular data but omits the region-specific features required for frameworks like the Vascular Cognitive Impairment Neuropathology Guidelines (VCING). To bridge this gap, we developed two harmonized measures-the SVD burden index and a simplified
"VCING-Lite" proxy-to quantify vascular injury and estimate VCID likelihood directly from NP11 data. Method: We analyzed NACC neuropathologic data from 1,658 autopsied participants across 32 Alzheimer's Disease Centers. Nineteen NP11 variables were harmonized on a 0-3 scale and combined into a six-domain SVD burden index (range 0-18). The VCING-Lite proxy categorized cases as low, intermediate, or high VCID likelihood based on the presence of large infarcts (≥ 15 mm), and moderate-to-severe global arteriolosclerosis or cerebral amyloid angiopathy. Statistical analyses were performed in Python 3.11, using ANOVA, Kruskal-Wallis tests, and Spearman's correlations with adjustments for age at death, sex, APOE genotype, and ADNC score. Results: Among 1,658 cases, 1,604 had sufficient data for SVD scoring. Scaled SVD burden scores ranged from 0-14 (mean = 4.12 ± 2.60). VCING-Lite categories showed clear separation in SVD burden (p < 0.001), and correlations between the two measures were strong and consistent (r = 0.74-0.75). Even after adjusting for covariates and excluding severe ADNC cases, the association remained robust (partial r = 0.72, p < 0.001), confirming that both indices capture stable, biologically meaningful variation in cerebrovascular pathology. Conclusion: Our SVD burden index and VCING-Lite proxy provide scalable, reproducible tools for quantifying vascular pathology across NACC cohorts. By transforming existing NP11 data into standardized metrics, these measures support cross-cohort harmonization and enable more precise modeling of vascular contributions to dementia. Future work will link these indices to imaging and molecular data to better define how cerebrovascular disease influences cognitive decline in aging and Alzheimer's disease.

Funding: This research was supported by grants U24NS133945 (M.E.F. as PI), R01AG072080 (M.E.F. as PI), RF1AG072080 (M.E.F. as PI), R01AG082118 (M.E.F. as PI), and P30AG066546 (M.E.F. as Neuropathology Core Leader) from the National Institutes of Health. Additional support was provided by the Owen's Foundation Grant (M.E.F. as PI) and the Baptist Foundation of San Antonio Endowment (M.E.F.). The NACC database is funded by NIA/NIH Grant U24 AG072122. NACC data are contributed by the NIA-funded ADRCs: P30 AG062429 (PI James Brewer, MD, PhD), P30 AG066468 (PI Oscar Lopez, MD), P30 AG062421 (PI Bradley Hyman, MD, PhD), P30 AG066509 (PI Thomas Grabowski, MD), P30 AG066514 (PI Mary Sano, PhD), P30 AG066530 (PI Helena Chui, MD), P30 AG066507 (PI Marilyn Albert, PhD), P30 AG066444 (PI David Holtzman, MD), P30 AG066518 (PI Lisa Silbert, MD, MCR), P30 AG066512 (PI Thomas Wisniewski, MD), P30 AG066462 (PI Scott Small, MD), P30 AG072979 (PI David Wolk, MD), P30 AG072972 (PI Charles DeCarli, MD), P30 AG072976 (PI Andrew Saykin, PsyD), P30 AG072975 (PI Julie A. Schneider, MD, MS), P30 AG072978 (PI Ann McKee, MD), P30 AG072977 (PI Robert Vassar, PhD), P30 AG066519 (PI Frank LaFerla, PhD), P30 AG062677 (PI Ronald Petersen, MD, PhD), P30 AG079280 (PI Jessica Langbaum, PhD), P30 AG062422 (PI Gil Rabinovici, MD), P30 AG066511 (PI Allan Levey, MD, PhD), P30 AG072946 (PI Linda Van Eldik, PhD), P30 AG062715 (PI Sanjay Asthana, MD, FRCP), P30 AG072973 (PI Russell Swerdlow, MD), P30 AG066506 (PI Glenn Smith, PhD, ABPP), P30 AG066508 (PI Stephen Strittmatter, MD, PhD), P30 AG066515 (PI Victor Henderson, MD, MS), P30 AG072947 (PI Suzanne Craft, PhD), P30 ; COI: None.

Poster # 23: GENE EXPRESSION SIGNATURES OF GLIOVASCULAR CELLS ASSOCIATED WITH MICROINFARCT BURDEN

Presenting author: Kyra M. Clarke, BS (UTH San Antonio)

BACKGROUND: Cerebral small vessel disease (cSVD) is a major contributor to vascular cognitive impairment and is increasingly recognized as a significant co-pathology in Alzheimer's disease and related dementias (ADRD). Microinfarcts- microscopic ischemic lesions that are frequently missed on standard neuroimaging- are lesions indicative of underlying cSVD and prevalent among patients with dementia. Understanding the molecular and cellular consequences of microinfarcts is essential to elucidating how vascular injury interacts with neurodegenerative processes in ADRD. METHOD: We utilized the NanoString GeoMx DSP platform to profile gene expression in postmortem human brain tissue from four subjects, stratified by microinfarct burden and matched for arteriosclerosis and AD Neuropathologic Change severity. These samples were confirmed not to have microinfarcts in the regions examined to investigate diffuse changes due to global cerebral hypoxia vs. lesion specific alterations. Areas of interest (AOIs) were selected from the middle frontal gyrus and basal ganglia for a total of 190 AOIs (54 - 929 cells per AOI; median = 287). Whole transcriptome sequencing was performed on spatially segmented astrocytes, microglia, and endothelial cells. RESULT: Differential gene expression and gene set variation analyses revealed cell- and region-specific transcriptional changes with increasing microinfarct burden, identifying 5,197 unique DEGs across cell types and regions. Notably, the greatest transcriptional shifts in gliovascular cells were observed at early stages of pathology in a non-linear pattern. Functional pathway modules encompassed inflammatory pathways, cell signaling, hypoxia/metabolic responses to hypoperfusion, gliovascular aging, and protein quality control. These findings suggest that early gliovascular changes may represent compensatory responses or the onset of pathological dysfunction driven by cSVD. CONCLUSION: Increasing microinfarct burden differentially impacts gene expression across brain regions and cell types. By applying spatial transcriptomic analysis with pathway profiling, this study characterizes the heterogeneous molecular signatures of comorbid vascular injury and AD, advancing mechanistic insight into mixed neuropathologies in dementia.

Funding: National Institute on Aging, Owen's Foundation; COI: N/A

Poster # 24: CEREBROVASCULAR DISEASE MODIFIES TMEM106B FIBRIL ACCUMULATION IN NEURODEGENERATION

Presenting author: Matthew Dopler, PhD (UTH San Antonio)

Background: TMEM106B forms age-dependent amyloid fibrils that are elevated in dementia patients and correlated with increased TDP-43 pathology. Cerebrovascular disease (CVD), a major dementia co-pathology, may contribute to TDP-43 dysfunction. Understanding the relationships among CVD, TMEM106B, and TDP-43 is crucial for developing targeted therapies Methods: We compared autopsy brains across TDP-43 pathologies and CVD to investigate potential differences in TMEM106B levels. Autopsy brain samples (N=58) were selected from the Biggs Brain Bank to include frontotemporal lobar degeneration with TDP pathology (FTLDTDP), limbic-predominant age-related TDP-43 encephalopathy (LATE), TDP-43 negative Alzheimer's diseases neuropathologic change (ADNC), and controls. Immunofluorescence staining using anti-TMEM106B (239-250aa) was performed on formalin-fixed Posterior hippocampal sections to assess whether CVD burden had an impact on TMEM staining patterns. Quantitative digital pathology analyses were conducted on whole slide images (WSI) to quantify TMEM106B fibril accumulation. Results: Age was the strongest predictor of TMEM106B fibril accumulation (p < 0.001), and females showed higher fibril burden than males (p = 0.035). TDP-43 pathology was associated with increased TMEM106B accumulation (p = 0.03). Greater co-existing cardiovascular disease (CVD) pathology burden showed a trend toward reduced TMEM106B fibril accumulation, suggesting an attenuating, and possibly dosedependent effect of CVD pathology on the observed TDP-43 and TMEM106B association (p = 0.058). Across ADNC (Braak tangle stage, Thal amyloid plaque phase, CERAD neuritic plaque density score), increasing CVD pathology burden did not modify TMEM106B fibril levels with increasing Braak stage (p = 0.64), Thal phase (p = 0.39), or CERAD score (p = 0.56), and TDP43 did not increase TMEM106B in the presence of these AD lesions. Conclusion: Age, sex, and TDP-43 are determinants of TMEM106B fibril. Increasing CVD pathology appears to blunt the TDP-43 and TMEM106B association, possibly due to hypoxia driven lysosomal dysfunction. Additional co-pathologies do not appear to drive TMEM106B fibril accumulation. Future work will assess if lower TMEM106B may reflect compensatory change.

Funding: This research was supported by K08AG065463, R01AG072080, RF1AG072080, R01AG082118, T32GM113896, T32GM14532, T32AG082661, T32TR004544, and P30AG066546 from the National Institutes of Health. Additional support provided by Owen's Foundation Grant, and the Baptist Foundation of San Antonio Endowment. ; COI: The authors declare no conflicts of interests.

Poster # 25: MICROGLIAL SENSOME ALTERATIONS IN NEURODEGENERATION

Presenting author: Teniade Adetona, BS (UTH San Antonio)

Background: Microglia play a critical role in detecting and responding to chemotactic cues in neurodegeneration. Sensing environmental changes is mediated by the microglial sensome, a defined set of genes and their products that enable microglia to sense changes in the brain parenchyma. Current literature points to a distinct microglial sensome signature in Alzheimer's disease (AD), where homeostatic genes are downregulated and genes associated with microglial response are upregulated. However, it remains unclear how the microglial sensome is altered in tauopathies including chronic traumatic encephalopathy (CTE), corticobasal degeneration (CBD), and progressive supranuclear palsy (PSP). Taken together, we hypothesize that the microglial sensome protein landscape in CTE, CBD, and PSP resembles that observed in AD. Method: To explore the role of the microglial sensome in neurodegeneration, formalin-fixed, paraffinembedded postmortem brain tissue (n=47) from decedents with pathologically confirmed AD, CTE, CBD, and PSP were cut into serial sections. Slides were labeled with known microglial function and sensome markers: P2RY12, IBA1, TREM2, CD68, CD74, pSMAD2/3, Clec7a. Additionally, to characterize disease pathology, slides were labeled for hyperphosphorylated tau (AT8) and dystrophic neurons (SMI-31). Labelled tissue slides were visualized, imaged, and digitized on a Leica Aperio AT2 Slide Scanner, and analyzed using Aperio ImageScope pathology software. Result: Preliminary findings indicate that CTE exhibits higher protein levels of the homeostatic receptor P2RY12 relative to AD, PSP, and CBD. Notably, this increase is evident in both gray matter where pathology is most prevalent and white matter regions where pathology is atypical. Consistent with the literature, there were no significant differences across diseases in IBA1 and AT8 protein levels. Conclusion: These findings suggest that other factors that contribute to a CTE prognosis, mainly traumatic brain injuries, may elicit unique microglial adaptations, contributing to divergent neurodegenerative pathways. Our study highlights the therapeutic potential of strategic targeting of the microglial sensing apparatus to prevent the development of neurodegeneration.

Funding: T32 NS082145 P30 AG066546 ; COI: N/A

Poster # 26: MUTANT KRAS-INDUCED ENDOTHELIAL CELLS TRIGGER AMYLOID-Β PRODUCTION IN A FOCAL AREA OF BRAIN ARTERIOVENOUS MALFORMATION WITH INTRACEREBRAL HEMORRHAGE

Presenting author: Eunsu Park, PhD (UTHealth Houston)

Background: Brain arteriovenous malformation (AVM) is a leading cause of hemorrhagic stroke. The brain AVM is characterized by abnormal connections between arteries and veins that often result in intracerebral hemorrhage (ICH, bleeding) and neurological deficits. Notably, 50-80% of brain AVM patients exhibit cognitive impairment, particularly memory deficits. Although brain AVM-associated bleeding can cause cognitive dysfunction, the underlying mechanisms remain unclear. Amyloid-β (Aβ) compromises blood-brain barrier (BBB) integrity and can trigger bleeding in Alzheimer's disease (AD). A clinical report shows an accumulation of Aβ-like proteins in vessel walls in ruptured human brain AVM tissue compared with unruptured AVMs. Based on this clinical evidence, we hypothesized that Aβ accumulation in the brain AVM area disrupts vascular integrity, thereby exacerbating hemorrhage and cognitive impairment. Method: Ruptured human brain AVM tissues are compared with the superficial temporal artery (STA). Blood was collected from ruptured human brain AVM patients and compared with blood from unruptured aneurysm patients. A mouse model of brain AVM was established by injecting brain endothelial cell-specific AAV-BR1-KRASG12V. Cognitive behavior testing was performed. Immunofluorescence staining was used to detect Aβ with Methoxy-X04 and antibodies MOAB-2, 4G8, and oligomer A11. Intracerebral hemorrhage (bleeding) was identified by Ter-119 (a marker of red blood cells). The FDA-approved donanemab was administered to the brain AVM mice. Result: We detected Aβ in ruptured human AVM tissues compared to STA. We also observed higher plasma Aβ(1-42) levels in ruptured AVM patients than in blood from unruptured aneurysm patients. Mutant KRAS-induced AVM mice display cognitive dysfunctions. In brain AVM mice, Aβ deposition aligns with vascular leakage. In contrast, non-ruptured AVM areas (without red blood cell detection) rarely exhibit Aβ. Finally, donanemab treatment mitigates Aβ detection and reduces bleeding in the AVM area in mice. Conclusion: Our study reveals, for the first time, that Aβ can drive vascular leakage and cognitive dysfunction in brain AVM, even in the absence of AD-related genetic mutations. These findings strongly indicate a causative role for Aβ in the pathology of brain AVM.

Funding: NINDS (R01NS126415), The Aneurysm and AVM Foundation (Cerebrovascular Research Grant); COI: None

Poster # 27: REGIONAL AND PATHOLOGICAL DIFFERENCES IN TAU SEEDING ACTIVITY REVEALED BY TAUHEK293 BIOSENSOR CELLS SEEDED WITH HUMAN BRAIN HOMOGENATES

Presenting author: Wileidy Gomez, PhD (UTHealth Houston)

Alzheimer's disease (AD) is characterized by the progressive accumulation of amyloid-β (Aβ) and tau aggregates in the brain, yet the molecular features that distinguish clinically silent from symptomatic stages remain poorly understood. We focused on characterizing tau seeding activities and aggregate morphologies induced by human brain homogenates derived from three conditions: non-demented without pathology (NDNP), non-demented with AD neuropathology (NDAN), and clinically diagnosed AD. We tested the whole homogenate, as well as pellets and supernatants, prepared from temporal cortices, hippocampi, and visual cortices. These different samples were tested for tau seeding activities in TauHEK293 biosensor cells. We observed that seeding activities were markedly enriched in the pellet fractions across regions, consistent with previous evidence that the majority of seeding competent aggregates are sedimented. Importantly, NDAN and AD brains displayed robust seeding activity, whereas NDNP brains showed minimal activity. The hippocampus exhibited the strongest seeding responses, underscoring its vulnerability at early stages of tau pathology. We quantified the number of puncta and found significant differences between groups. NDAN and AD brains induced higher numbers of tau inclusions compared to NDNP. Furthermore, the supernatant fractions showed limited seeding capacity, reinforcing the notion that most of the seeding-competent aggregates are present in the pellet. Beyond quantitative differences, we analyzed the morphological diversity of tau inclusions. AD-derived samples induced a wide range of structures, including threads, ordered aggregates, and large speckles, whereas NDAN brains predominantly triggered small speckles and diffuse cloud-like inclusions. NDNP samples generated only scarce inclusions, lacking complex morphologies. These findings support the concept of tau conformational strain heterogeneity and reveal that distinct pathological backgrounds shape both the extent and the structural outcomes of tau seeding. Our results provide novel evidence that clinically silent AD pathology (NDAN) is sufficient to drive significant tau seeding and distinct aggregate morphologies. This work expands on prior studies by directly comparing NDAN and AD cases, highlighting the importance of early, preclinical tau seeding events. We propose that morphological profiling of tau inclusions may serve as a powerful tool to differentiate tau strains across disease stages, paving the way for precision diagnostics and therapeutic strategies in AD.

Funding: AARGD-22-972125 (Alzheimer's Association) and 2R01AI132695 (NIH); COI: The authors declare no conflicts of interest

Poster # 28: THE EMERGENCE OF TAU-MUSASHI1 HETEROAMYLOIDS AS DISTINCTIVE CONTRIBUTORS TO ALZHEIMER'S PATHOLOGY

Presenting author: Abbigael Aday, BS (UTMB)

Alzheimer's disease (AD) is the most common age-related neurodegenerative disorder, histopathologically characterized by the accumulation of amyloid-β plaques and neurofibrillary tangles comprised of hyperphosphorylated Tau. Among Tau species, the smaller, soluble oligomers are now recognized as the most neurotoxic, with the capacity to propagate in a prionlike manner across cells and brain regions. Beyond homotypic aggregation, Tau interacts with other proteins to form aggregates, referred to as heteroamyloids, which are increasingly implicated in disease pathogenesis due to their enhanced toxicity and complex structural diversity. This study explores the formation of heteroamyloids between Tau and the RNA binding protein Musashi1 (MSI1), a regulator of mRNA translation and a protein associated with multiple neurological conditions. MSI1 has been observed, by our group, to aggregate and mis-localize in AD, exhibiting functional impairment in the presence of pathological Tau species. Here, the presence of MSI1:Tau heteroamyloids is examined in human brain tissue, while recombinant approaches are used to reconstitute and biochemically characterize these aggregates as distinct molecular species. We performed immunofluorescence, 3D reconstructive analysis, and sarkosyl fractionation to investigate the presence of heteroamyloids in AD brain tissue. We then produced heteroamyloids in vitro, followed by highly validated biochemical and biophysical characterizations including AFM, CD, PK-digestion, and EM to distinguish these species as unique heteroamyloids. The detection of MSI1:Tau heteroamyloids in Alzheimer's disease brain tissue introduces a novel and promising area for therapeutic and diagnostic development. Given the reported enhanced toxicity of other heteroamyloids in neurodegenerative disease, further investigation into their structure and function may uncover critical mechanisms driving disease progression.

Funding: Funding sources: Alzheimer's Association Research Fellowship AARF-21-720991, NIH 1R03AG088929-01, Moody Brain Health Seed Grant, Mitchell Center, Sealy Aging Center, and Drug Discovery Institute (Mauro Montalbano); NIH RF1AG077484, RO1AG054025, AG058522, and R01G077253 (Rakez Kayed).; COI: None

Poster # 29: DIFFERENTIAL EFFECTS OF BRAIN-DERIVED TAU OLIGOMERS ON NEUROGENESIS IN ALZHEIMER'S DISEASE AND COGNITIVELY RESILIENT INDIVIDUALS

Presenting author: Adam Trupp, BS, BM (UTMB)

Impairment of neurogenesis, a critical process for maintaining cognitive function, accelerates cognitive decline in Alzheimer's disease (AD). A subset of individuals however, termed nondemented with Alzheimer's disease neuropathology (NDAN), exhibit resilience, retaining cognitive function despite the presence of amyloid beta plaques and neurofibrillary tangles of tau. This cognitive preservation has been linked to the maintenance of hippocampal neural stem cells (NSCs) and enhanced neurogenesis in NDAN individuals. Here, we examine how brain-derived tau oligomers (BDTOs) isolated from AD or NDAN patients differentially regulate neurogenesis and cognition. Using both in vitro NSC cultures and in vivo models, we assessed neurogenesis via immunofluorescence of canonical markers. Cognition was assessed via novel object recognition (NOR) test. Our preliminary findings uncover a key distinction: BDTOs isolated from AD patients may accelerate NSC differentiation in vitro and in vivo, disrupting normal developmental trajectories, while NDAN BDTO groups undergo proper differentiation timelines, comparable to control groups. Behavioral data supports that long-term memory is more susceptible to AD BDTOs than that of NDAN BDTOs. These findings suggest that AD and NDAN BDTOs exert fundamentally different biological effects on neurogenesis and cognition.

Funding: R01-AG069422 (GT, MAM); COI: The author declares no conflicts of interest.

Poster # 30: TOXIC TAU ANTIBODY-GUIDED MALDI-MSI COMBINED WITH DESI-MSI REVEALS REGION-SPECIFIC TAU PATHOLOGY AND METABOLIC ALTERATIONS IN HTAU MICE

Presenting author: Dhvani Pansuriya, MS (UTMB)

Background: Alzheimer's disease (AD) is a progressive neurodegenerative disorder characterized by lipid dysregulation and pathological aggregation of tau protein. While neurofibrillary tangles are a hallmark of AD, increasing evidence points to soluble tau oligomers (TauO) as the key toxic species driving disease progression and spread. Mass spectrometry imaging (MSI) enables spatial mapping of biomolecules in tissues, offering insight into region-specific molecular changes. However, conventional methods often lack the spatial and molecular resolution needed to capture these complex biochemical alterations in the brain. Understanding how tau pathology evolves across different brain regions, ages, and sexes is essential for uncovering mechanisms underlying AD. Methods: We developed an integrated multi-omic imaging workflow combining MatrixAssisted Laser Desorption/Ionization (MALDI) and Desorption Electrospray Ionization (DESI) MSI to investigate tau pathology and related molecular changes. Tau-oligomer-specific antibodies (TOMA2 and TTCM2) were conjugated with photo-cleavable mass tags (PCMTs), enabling MSI-based visualization of antibody binding across brain regions. Frozen brain tissue sections from transgenic hTau mice were analyzed using a serial MALDI-DESI workflow and compared with their wild-type (WT) littermates in an age-dependent and sex-specific manner. MALDI-MSI profiled complex lipid species, while DESI-MSI-optimized for 20 µm near-singlecell resolution-mapped small polar metabolites such as γ-aminobutyric acid (GABA), glutamine, and glutamate. Adjacent sections were examined with immunofluorescence (IF) and hematoxylin and eosin (H&E) staining for histological validation. Results: The combined MALDI-DESI approach enabled spatial visualization of lipidomic and metabolic alterations in hTau mouse brains. Distinct age-dependent and region-specific differences in tau pathology were observed in hTau mice compared to WT controls. Antibody-based MSI revealed heterogeneous binding of TOMA2 and TTCM2 across cortical and hippocampal regions, correlating with tau oligomer accumulation confirmed by IF analysis. In parallel, DESI-MSI identified age-related metabolic shifts in neurotransmitter-related metabolites, including GABA, glutamine, and glutamate, highlighting the metabolic consequences of tau aggregation. Conclusion: This study demonstrates the utility of MALDI-DESI MSI coupled with PCMT-conjugated antibodies for high-resolution spatial mapping of tau pathology and related metabolic changes. By linking tau oligomer distribution with metabolic dysregulation, these findings provide new insight into the molecular mechanisms underlying Alzheimer's disease progression and identify potential targets for therapeutic intervention.

Funding: This research work was supported by Mitchell Center for Neurodegenerative Diseases, and National Institute of Health grants: (R01AG077253, R01AG054025 and U24AG072458 to RK).; COI: N/A

Poster # 31: NEUROPLASTIN TRANSCRIPTIONAL DOWNREGULATION REVEALS A GENE-PROTEIN DISCREPANCY IN ALZHEIMER'S DISEASE

Presenting author: Eileen Gomez, BS (UTMB)

Neuroplastin (NPTN) is a cell-adhesion glycoprotein that plays a key role in regulating synaptic calcium homeostasis. Its isoforms form a functional complex with plasma-membrane Ca²⁺ATPase 2 (PMCA2), facilitating calcium extrusion and restoring basal intracellular Ca²⁺ levels. Given its role in calcium handling, alterations in neuroplastin could have important implications for Alzheimer's disease (AD) pathophysiology. While protein-level studies have reported increased neuroplastin protein levels in the hippocampus of AD brains, its transcriptional regulation has not been examined. To address this gap, we analyzed publicly available human hippocampal microarray and RNA-seq datasets downloaded from the Gene Expression Omnibus (GEO2R) and the Allen Brain Atlas: Aging, Dementia and TBI Study (ADTBI), to characterize expression and co-regulation among NPTN, ATP2B2 (PMCA2), and MAPT (Tau), across diagnostic groups and disease progression (Control, n=116; AD, n=73). We tested the hypothesis that in Alzheimer's Disease, NPTN expression is reduced and that this decrease disrupts coregulation with ATP2B2 and MAPT, reflecting disturbance of the Np-PMCA complex and impaired calcium regulation at the synapse. We found that across independent cohorts and meta-analysis, NPTN transcript levels were consistently reduced in AD (g = -0.62, 95% CI [1.01, -0.22], p = 0.0022; I² = 34.8%), whereas ATP2B2 and MAPT showed greater variability. Despite this downregulation, NPTN and ATP2B2 expression remained positively correlated across diagnostic groups (r ≈ 0.5-0.6, p < 0.001), whereas the MAPT-NPTN relationship was also positive but weaker and more variable (r ≈ 0.5, p ≤ 0.01). Importantly, these transcriptional findings contrast with reports of elevated neuroplastin, implying possible post-transcriptional or compensatory mechanisms. Together, these findings highlight and reinforce the role of neuroplastin as a regulator of synaptic calcium clearance and suggest that the disruption of the Np-PMCA2 complex occurs post transcriptionally under tau pathology and may underlie the calcium dysregulation that contributes to the multifaceted synaptic vulnerability characteristic of Alzheimer's disease.

Funding: AG070255 AG073133; COI: N/A

Poster # 32: HETEROTYPIC SEEDING OF TAU BY AMYLOID OLIGOMERS GENERATES CONFORMATIONALLY AND FUNCTIONALLY DISTINCT TAU SPECIES

Presenting author: Nemil Bhatt, PhD (UTMB)

Background: Tau aggregation is a classic hallmark of tauopathies, yet how heterotypic interactions with other amyloid proteins influence tau conformation and pathogenicity remains unclear. We investigated whether heterotypic-seeding by amyloid-b, a-synuclein or TDP-43 aggregates drives the formation of distinct, disease relevant tau conformers. Methods Recombinant 4R, 3R, and 3R:4R tau monomers were cross-seeded with amyloid-β oligomers (AβO), α-synuclein oligomers (αSynO), or TDP-43 oligomers (TDP-43O). The resulting assemblies were analyzed by Western blotting, filter-trap, Bis-ANS/ThT fluorescence, and atomic-force microscopy (AFM). Functional activity was tested in tau RD-P301S biosensor cells, acute hippocampal slices, and primary neuron-glia co-cultures. Results: Heterotypic seeding promoted the formation of non-fibrillar, conformationally diverse tau oligomers that retained isoform-specific (3R vs 4R) features while acquiring seed-dependent epitopes. AFM revealed distinct morphologies-globular in homotypic seeded, heterogeneous in AβO/αSynO and denseclustered in TDP-43O conditions. Peptide profiling confirmed unique proteolytic and structural signatures. In tau biosensor assays, seed-dependent tau propagation was observed and electrophysiological analysis demonstrated broad synaptic impairment across tau isoforms in a seed dependent manner. In mixed neuron-glia cultures, heterotypic seeded tau displayed celltype-specific uptake, accumulating primarily in astrocytes and microglia. Cytokine profiling revealed that seed-dependent distinct inflammatory signatures, with αSynO- and TDP-43Oseeded tau eliciting the strongest glial cytokine responses. Conclusions: Amyloid heterotypic seeding triggers folding of tau into structurally and functionally distinct aggregates that differentially engage neuronal and glial pathways. These findings highlight importance and establish a mechanistic framework linking heterotypic seeding to strain-specific tau toxicity and glial-driven neuroinflammation, offering new insight into the molecular heterogeneity underlying Alzheimer's disease and related tauopathies.

Funding: This research work was supported by National Institute of Health grants: (R01AG077253 to RK and YZ, R01AG054025 and U24AG072458 to RK,) ; COI: No conflict

Poster # 33: APOE ISOFORM-ASSOCIATED TAU OLIGOMER POLYMORPHS DIFFER IN SYNAPTOTOXICITY AND SEEDING POTENCY

Presenting author: Nikita Shchankin, BS (UTMB)

BACKGROUND: Pathological tau aggregates form distinct polymorphic species across diseases and even across Alzheimer's disease (AD) patients. However, tau aggregate polymorphism across the apolipoprotein E isoforms (APOE ε2, ε3, ε4), the strongest predictors of late-onset AD development, is unknown. METHODS: This study assessed conformational and bioactivity properties of tau oligomers from 14 patients with varying APOE genotypes. RESULTS: Tau oligomers differ in proteolytic stability and cleavage site profiles across the APOE isoforms, indicating conformationally-distinct polymorphs. APOE isoform-associated tau oligomers affect synaptic plasticity differently, with ε4-associated oligomers having the highest potency and strongest impact on synaptic functioning. Bioactivity assays reveal that ε4-associated oligomers demonstrate particularly high seeding activity. Interestingly, tau oligomer synaptotoxicity and seeding activity are independent characteristics. CONCLUSIONS: The APOE isoforms are associated with distinct tau oligomer polymorphs with varying bioactivity, underscoring the importance of considering APOE status when generating AD therapies. Polymorph-specific targeting of pathological tau species could provide a novel method of combating AD.

Funding: This work was funded by the National Institutes of Health (grant numbers AG06071801, AG07245801, AG07725301, AG05402506, AG077484 to R.K., AG070255 to AL, and AG067952 to N.M.).; COI: N/A

Poster # 34: FUNCTIONAL CHANGES UNDERLYING A SMALL- MOLECULE INHIBITOR CONTRIBUTING TO IMPROVED COGNITIVE RESERVE AND SYNAPTIC RESILIENCE.

Presenting author: Sanjana Mohanty, MS (UTMB)

Background: Dendritic spines, crucial for synaptic plasticity, learning, and memory, are known to undergo significant loss in Alzheimer's disease (AD), contributing to cognitive impairment. Phospholipase D1 (PLD1), an enzyme implicated in diverse cellular processes including cytoskeletal modulation and vesicle trafficking, exhibits elevated expression in AD pathology. Previous work demonstrated that pharmacological inhibition of PLD1 using VU0155069 (VU01) ameliorated dendritic spine deficits specifically in the CA1 hippocampal region of 6- and 12month-old 3xTgAD mice, models displaying progressive Aβ and tau pathology. This study expands upon those findings to provide a comprehensive assessment of PLD1 inhibition across multiple brain regions. Methods: Six- and twelve-month-old 3xTgAD mice received intraperitoneal injections of VU01 (1 mg/kg) or saline control every other day for one month. Brain tissues were subsequently processed using the FD Rapid GolgiStain kit, and dendritic spines were imaged via high-resolution microscopy (100X) and quantified using ImageJ across hippocampal (CA2, CA3, DG), cortical (V1, V2, A1, A2), and amygdala (LA, BLA, CeA) subregions. Results: Results revealed a consistent trend of increased dendritic spine density in VU01-treated mice compared to saline controls across the diverse brain areas examined, apparent at both early and advanced stages of AD-like pathology. Conclusion: These findings suggest that inhibiting PLD1 provides widespread neuroprotection against spine loss, potentially conserving critical synaptic structures like memory-associated mushroom spines, thereby supporting cognitive reserve in AD.

Funding: None reported.; COI: None reported.

Poster # 35: INVESTIGATING THE ROLE OF CONNEXIN-50 IN THE PROPAGATION OF POLYMORPHIC Α-SYNUCLEIN IN SYNUCLEINOPATHIES

Presenting author: Twisha Chatterjee, MS (UTMB)

Background: Connexins, gap junction components, have been implicated in intercellular connectivity under physiological and pathophysiological conditions, including neurodegeneration. Synucleinopathies comprise a diverse group of neurodegenerative disorders pathologically characterized by α-synuclein (α-Syn) aggregates. However, little is known about connexin-associated α-Syn pathological spread in synucleinopathies. Method Here, we present evidence of connexin-50 (Cx50) directly interacting with α-Syn aggregates in human brains affected by synucleinopathies, including Alzheimer's disease (AD), dementia with Lewy bodies (DLB), and Parkinson's disease (PD). We also observed this interaction in pre-clinical αSyn mouse models that exhibit Parkinson's disease-related phenotypes. To achieve this, we utilized immunofluorescence using markers for astrocyte and neuron as well as antibodies specific to α-Syn, pharmacological, and genetic manipulation techniques such as using gap junction inhibitors and performing gene knockdown with the help of siRNA respectively in connexin cell models and primary co-cultures. Result Utilizing well-characterized α-Syn oligomers (BDSOs) isolated from brains affected by Alzheimer's disease (AD), Lewy body dementia (DLB), and Parkinson's disease (PD), we demonstrate that BDSOs preferentially enter cells expressing connexin 50 (Cx50), a protein found in gap junctions, as confirmed by pharmacological inhibition. In live-cell imaging experiments, we observed a significant reduction in BDSO uptake in primary neurons-astrocytes co-cultured from human wild type α-Syn transgenic mice expressing genetically modified Cx50. This reduction was accompanied by a decrease in α-Syn aggregates. Moreover, the downregulation of Cx50 was associated with a reduction in the activity of astrocytes, as evidenced by a decrease in pro-inflammatory and an increase in anti-inflammatory cytokines. These findings suggest that Cx50 plays a crucial role in the interplay between neurons and astrocytes. Conclusion This study presents compelling evidence of the connection between disease-relevant α-Syn aggregates and Cx50 in neurons that regulate astrocyte activity. This insight sheds light on the intricate process of pathogenic α-Syn spread in synucleiopathies.

Funding: Mitchell Center for Neurodegenerative Diseases, and National Institute of Health grants: (R01AG077253, R01AG054025 and U24AG072458 to RK); COI: NA

Poster # 36: IDENTIFICATION OF RESILIENCE AND SUSCEPTIBILITY FACTORS IN A MOUSE MODEL OF ALZHEIMER'S DISEASE NEUROPATHOLOGY

Presenting author: Chia-Hsuan Fu, PhD (BCM)

Background: Alzheimer's disease (AD) is the most common form of dementia and is characterized by amyloid plaques and neurofibrillary tangles. However, some individuals with these neuropathological hallmarks remain cognitively intact. The mechanism underlying differential resilience and susceptibility to cognitive impairment in disease is unclear. Method: To identify pathways that might confer resilience to AD neuropathology, we used a mouse model expressing a mutant human amyloid precursor protein (APP) transgene. We previously found that, similar to human individuals with AD neuropathology, some APP mice develop resilience to memory deficits and continued seizure activity, despite being genetically identical to their littermates who remain susceptible. We performed bulk RNA-sequencing on microdissected dentate gyrus (DG) tissue from a large group of APP mice and nontransgenic (NTG) littermates of both sexes. Result: Using principle component analysis, we found that whereas NTG mice clustered closely together in one quadrant of the PCA plot, APP mice were spread apart in all four quadrants. One subset of APP mice clustered closely together with NTG mice. Expression of ΔFosB, a seizure-induced transcription factor, and cognition-associated genes in these APP mice were similar to that in NTG mice, suggesting that these mice were resilient to disease progression. The majority of the remaining APP mice clustered in the quadrant directly opposite to that of the NTG mice. These mice had increased ΔFosB expression and decreased expression of cognition-associated genes, suggesting that these mice were susceptible to disease progression. Using this grouping, we identified a small set of genes that were differentially expressed in resilient APP mice compared with either NTG or susceptible APP mice, which likely represent resilience factors, while a larger set of genes was differentially expressed in susceptible APP mice compared with either NTG or resilient APP mice, which likely represents susceptibility factors. Conclusion: APP mice can be sorted into resilient and susceptible groups based on the transcriptomic profile of their DG, which allowed us to identify factors that might confer resilience and susceptibility to seizures and cognitive deficits in AD. Understanding the functions of these factors could help identify new therapeutic strategies to combat AD.

Funding: This work was funded by NIH NS086965 and NS085171 (JC).; COI: No conflict of interest.

Poster # 37: ASTROCYTES-SYNAPSE INTERACTIONS ARE REDUCED IN THE THALAMIC RETICULAR NUCLEUS OF APP MICE

Presenting author: Joseph Campbell, BS, MS (BCM)

Background: Deficits in sleep maintenance and slow wave sleep (SWS) during early stages of Alzheimer's disease (AD) can impact production and clearance of amyloid beta (Aβ), likely contributing to disease progression. Thus, understanding the mechanisms of AD-related sleep disruptions is of vital importance. Previously, our lab demonstrated mice expressing mutant human amyloid precursor protein (APP mice), exhibit hypofunctioning of the thalamic reticular nucleus (TRN), which is associated with sleep deficits early in disease progression prior to formation of Aβ plaques. The mechanisms underlying TRN dysfunction remain unclear. However, AD-related alterations in astrocyte function, such as the ability to stabilize or modulate synapses, have been identified as contributors to pathophysiology. Methods: Using bulk RNA sequencing, fluorescent immunohistochemistry, and an astrocyte specific AAV reporter, we investigated whether state-dependent changes occur in TRN astrocytes that might impact astrocyte-synapse interactions and alter glutamatergic TRN synapses numbers in APP mice. Results: Here we report that early in disease progression prior to plaque deposition, APP mice exhibit: 1) TRN-specific differentially expressed gene profiles related to astrocyte territory size. 2) TRN astrocyte population remains stable over time; however, GFAP-expressing astrocytes are increased in number at very early stages. 3) TRN astrocytes have more complex GFAPexpressing main processes while their territory and finer morphological features are reduced. 4) TRN astrocyte-synapse interactions are decreased, concomitant with a decrease in glutamatergic synapses. Conclusion: As reciprocal interactions between astrocytes and neurons play key roles in shaping network activity, AD-related morphological and functional changes in astrocytes could negatively impact TRN function, with consequences on sleep. Therefore, astrocytes may represent promising targets for innovative therapies to improve sleep and slow or stop the progression of AD. Further studies will be needed to 1) determine the mechanisms and precise role astrocytes play in loss of TRN synapses, 2) the impact that reduced astrocyte-synapse interactions have on TRN synaptic activity, and 3) whether restoring astrocyte morphology can attenuate TRN synaptic deficits and AD related pathology.

Funding: NIH grants NS085171 (JC) and AG065290 (JC and MB), and by the Belfer Neurodegeneration Consortium at MD Anderson (JC and MB).; COI: None

Poster # 38: CAUDAL NUCLEUS OF THE SOLITARY TRACT: A NOVEL INTEGRATION HUB FOR RESILIENCE AND SUSCEPTIBILITY IN ALZHEIMER'S DISEASE

Presenting author: Marina Rodriguez Alonso, BS (BCM)

Background: Alzheimer's disease (AD) is a neurodegenerative disorder characterized by amyloid-β (Aβ) accumulation and progressive cognitive decline. Although Aβ burden correlates with disease pathology, it does not fully explain clinical variability. Some individuals and ADmodel mice maintain normal cognition despite comparable Aβ levels, representing a "resilient" phenotype that offers insight into mechanisms preserving cognitive function. The caudal nucleus of the solitary tract (cNST), a brainstem region integrating vagal and peripheral inflammatory signals, has been implicated in physiological resilience. However, its role in neuronal and cognitive resilience during AD progression remains unexplored. Given its integrative function, the cNST may serve as a hub influencing whether individuals with AD pathology exhibit resilience or susceptibility. Method: In amyloid precursor protein (APP) transgenic mice, activity-dependent biomarker expression patterns distinguish resilient from susceptible phenotypes based on the degree of hippocampal dysfunction. Building on this framework, I examined activity of the cNST of non-transgenic (NTG), resilient, and susceptible APP mice to determine whether activity of cNST corresponds to resilience or susceptibility in APP mice. Result: Preliminary data reveal increased activity marker expression in the cNST of resilient APP mice relative to susceptible APP mice, suggesting heightened or more adaptive neuronal activation in this region that could contribute to the development of resilience. Conclusion: cNST activity differs between resilient and susceptible mice, suggesting differential neural regulation that may mediate communication between peripheral and central signals. Therefore, the cNST may function as an integrative hub where peripheral and central pathways converge to promote cognitive resilience. Defining how the cNST might contribute to resilience may uncover novel therapeutic targets that preserve cognitive function and mitigate susceptibility in AD, ultimately informing strategies to slow or prevent disease progression.

Funding: Belfer Neurodegeneration Consortium at MD Anderson.; COI: N/A

Poster # 39: ABLATION OF BACH1 IMPROVES MEMORY AND REDUCES PATHOLOGY IN HUMAN AMYLOID PRECURSOR PROTEIN TRANSGENIC MICE

Presenting author: Pei-Yun Chuang, MS (BCM)

Background: Bach1 is a transcriptional repressor that downregulates antioxidant defense genes and has been linked to several neurological conditions, including Alzheimer's disease (AD), Parkinson's disease, and Down syndrome. Transcriptomic profiling of transgenic mice expressing mutant human amyloid precursor protein (APP)-a well-established model of AD pathologyshowed elevated Bach1 expression in animals displaying greater vulnerability to cognitive deficits. Additionally, pharmacologically-induced seizures increased Bach1 levels even in wildtype mice, suggesting that Bach1 may generally modulate oxidative stress pathways contributing to neuronal susceptibility. Method: To investigate whether Bach1 directly influences neurological outcomes, we generated Bach1 knockout (Bach1-KO) mice and performed an extensive series of behavioral assessments. We also examined molecular responses following chemoconvulsant-induced seizures in Bach1-KO mice to evaluate antioxidant capacity and astrocytic activation. To assess the role of Bach1 in AD-related pathology, we crossed Bach1-KO mice with APP transgenic mice, and the resulting APP/Bach1-KO offspring were analyzed for cognitive performance and amyloid plaque accumulation. Result: Across behavioral assays, Bach1-KO mice exhibited normal performance, indicating that loss of Bach1 does not result in apparent behavioral abnormalities. Following seizure induction, Bach1-KO mice demonstrated elevated antioxidant responses and reduced astrocyte reactivity, suggesting improved resistance to neuronal stress. In the AD model, APP/Bach1-KO mice performed better on cognitive tasks and displayed decreased amyloid plaque burden compared with APP controls, indicating a protective effect of Bach1 deletion. Conclusion: Elimination of Bach1 enhances intrinsic neuroprotective and antioxidant mechanisms without causing behavioral deficits. Bach1 deficiency reduces amyloid pathology and ameliorates cognitive decline in APP mice, supporting its role as a key modulator of oxidative stress and neurodegeneration. These findings highlight Bach1 as a potential therapeutic target for preventing or slowing disease progression in Alzheimer's and seizure-related neurological disorders.

Funding: This work was funded by NIH grants NS085171, AG065290 (JC), and support from the Belfer Neurodegeneration Consortium at MDACC.; COI: N/A

Poster # 40: TITLE: REDUCED PROLIFERATIVE AND DIFFERENTIATION POTENTIAL OF SUBVENTRICULAR ZONE NEURAL STEM CELLS IN AN ALZHEIMER'S MOUSE MODEL

Presenting author: Abena Asua Dwamena, PharmD (Texas Tech HSC)

Background: Alzheimer's disease (AD) is a leading cause of severe dementia. It is characterized by the accumulation of amyloid plaques and hyperphosphorylated neurofibrillary tangles. This leads to defective synaptic transmission between neurons, resulting in the eventual death of the cells. It has been shown that the brain has an intrinsic regenerative capacity to compensate for lost neurons through a process called neurogenesis. Recent research points to a decline in the brain's regenerative capabilities in neurodegenerative diseases. Hence, our goal is that if this condition can be remedied, this could improve the body's repair mechanism to effectively address AD. Neural stem cells (NSCs) help in the regeneration of neurons; however, their specific role in AD recovery is not well-defined. This study aims to develop and evaluate NSC cultures derived from the subventricular zone (SVZ) of adult transgenic AD model mice and age-matched wild-type animals. Method: NSCs were maintained as neurospheres under serum-free conditions supplemented with growth factors and analyzed for stemness markers (Nestin and Sox2 expression), proliferation markers (analysis of neurosphere size and sphere-forming ability), as well as early markers of differentiation (expression of doublecortin (DCX) and GFAP in astrocytes). Result: Our findings suggest that NSCs in the AD model exhibit a decline in sphere formation and reduced expression of neuronal lineage marker doublecortin compared with wildtype mice, indicative of compromised proliferative and differentiation capacity. Conclusion: These findings are in line with our hypothesis that intrinsic NSC dysfunction contributes to defects in brain regeneration in AD patients. Ongoing studies will further explore molecular mechanisms linking proteostatic stress and NSC fate, opening a new path towards understanding ways that could help the brain regain its cognitive abilities.

Funding: Texas Tech University Health Sciences Center; COI: No conflict of interest

Poster # 41: UBE2O DEFICIENCY IN MICE AND PRIMARY CORTICAL NEURONAL CELL CULTURE DISRUPTS PROTEOSTASIS AND INCREASES AD-RELATED PATHOLOGY AND NEUROINFLAMMATION.

Presenting author: Augustina Potokiri, MSc (Texas Tech HSC)

Background: The rise in average U.S. life expectancy from 70 years in the 1980s to 80 years in 2025 has paralleled a threefold increase in Alzheimer's disease (AD) prevalence. AD is characterized by protein aggregation and neuroinflammation, processes tightly regulated by the ubiquitin-proteasome system (UPS). Ubiquitin-conjugating enzyme E2O (UBE2O), an E2-E3 hybrid enzyme within the UPS, plays a key role in substrate ubiquitination and degradation.

However, its precise role in AD remains unclear. Methods: Using UBE2O knockout (KO) mice and primary cortical neuronal cultures, we investigated the effects of UBE2O loss on cognition, protein homeostasis, and inflammation. Behavioral tests assessed learning and memory performance. Western blot analysis was performed to detect total ubiquitin, amyloid precursor protein (APP), and tau. Quantitative PCR measured Il 1β, Il 6, and Tnf-α expression. RNA sequencing, synapse isolation, and immunohistochemistry were performed to evaluate synaptic integrity and neuronal subtype vulnerability. Results: UBE2O KO mice exhibited significant learning and memory deficits compared to wild-type (WT) controls. Western blot analysis revealed increased accumulation of total ubiquitin, APP, and tau, indicating impaired proteostasis. Pro-inflammatory cytokines were markedly upregulated in KO brains, and RNA-seq data showed dysregulation of synaptic genes. Immunostaining identified selective loss of specific excitatory and inhibitory neuronal populations. Conclusion: Collectively, these findings demonstrate that loss of UBE2O disrupts proteostasis, enhances neuroinflammation, and impairs synaptic signaling, ultimately contributing to the cognitive decline characteristic of AD. Our results identify UBE2O as a critical regulator of neuronal protein homeostasis and suggest that enhancing UBE2O expression or activity may represent a promising therapeutic strategy for Alzheimer's disease.

Funding: NIH; COI: Non

Poster # 42: DYSFUNCTION OF SYNAPTIC ORGANIZER CEREBELLIN1-GLUD IMPAIRS AUTOPHAGY AND NEUROPLASTICITY IN ALZHEIMER'S DISEASE - EVIDENCE FROM MOUSE AND HUMAN BRAINS

Presenting author: Yong Chen, PhD (Texas Tech HSC)

Our recent study found that the expression of a synaptic organizing complex consisting of cerebellin1 (Cbln1) and glutamate delta receptors (GluD), was significantly decreased in postmortem brains of patients with Alzheimer's disease (AD) and in J20 AD mouse brains. Intracerebroventricular injection (ICV) of Cbln1 mitigated cognitive behavioral deficits and restored neuroplasticity (LTP) in J20 mice. However, the underlying mechanisms remain unclear. Accumulating evidence has shown that GluD regulates autophagy. Moreover, autophagy is impaired in AD patients and mouse models. Here we hypothesized that impaired Cbln1-GluD signaling in AD is causally linked to impaired autophagy and neuroplasticity but that these deficits can be restored by Cbln1 administration in J20 mice. We measured the expression of BDNF, a critical player in neuroplasticity and modulator of autophagy, beclin 1, an autophagy marker, and MnSOD2, a mitochondrial superoxide dismutase that can modulate autophagy, in the hippocampus of wild type (WT) and J20 mice using western blotting. Expression of BDNF, beclin 1, and MnSOD2 proteins was decreased significantly in J20 mice compared with WT mice. A single ICV injection of recombinant human Cbln1 blocked these changes. Knockdown of GluD1 or GluD2 by siRNAs blocked the effects of Cbln1 on hippocampal BDNF and beclin 1 expression, whereas only GluD1, but not GluD2 siRNA, blocked the Cbln1 effects on MnSOD2 expression, suggesting that GluD receptors are involved in the beneficial effects of Cbln1 on neuroplasticity and autophagy molecules in the J20 AD model. Finally, to explore the relevance of these findings for the human condition, we performed western blotting and coimmunoprecipitation (co-IP) experiments in post-mortem brain tissues from AD patients. The data showed that the expression of beclin 1 also decreased in the hippocampus from AD patients and is directly linked to GluD1, based on co-IP data showing the interaction between an autophagy regulator (PIST), GluD1 and beclin 1, confirming our findings in J20 mice. Together with our previous electrophysiological and behavioral data, these results suggest that dysfunction of the Cbln1-GluD trans-synaptic complex impairs autophagy and neuroplasticity, and that modulating the autophagy pathway by Cbln1 can restore neuroplasticity and improve cognitive function in AD.

Funding: The Garrison Family Foundation and Center of Excellence for Translational Neuroscience and Therapeutics, TTUHSC; COI: No

Poster # 43: DIURNAL DIFFERENCES IN IMMUNE CELL TRAFFICKING THROUGH THE CENTRAL NERVOUS SYSTEM IN AGING

Presenting author: Andrew Hynes, BS (UT Austin)

Background: Neuroinflammation is a critical feature of age-associated neurodegenerative disorders. Additionally, aging results in disruptions to circadian/sleep-wake cycles, which can regulate immune activity. Thus, understanding how circadian rhythms influence immune function may lead to greater understanding of the pathophysiology of dementia. In the periphery, the circadian clock regulates daily rhythms in immune cell migration through the blood and into tissues. Disruption of these rhythms is associated with worsened pathology in models of inflammatory disorders, such as inflammatory bowel disease, asthma, and arthritis. Therefore, we hypothesize disruption of immune cell migration rhythms through the brain and meninges as a potential mediator of neuropathology in aging. Methods: Immune cell subpopulations were assessed in meninges and brains using flow cytometry in male and female adult (6 mos) and aged (18 mos) C57BL/6J mice at 4 times of day: lights on (zeitgeber time [ZT]1), mid light phase (ZT6), lights off (ZT13), and mid dark phase (ZT18). Immune cell distribution in the meningeal vasculature was further investigated using immunohistochemistry at ZT1 and ZT13. Results: Aging was associated with elevated T cell populations, primarily driven by an increase in CD8+ T cell populations in the brain and meninges during the light phase, while adult female mice displayed diurnal differences in T cell migration through the meninges, with lower counts during the light phase. This diurnal difference was disrupted in aging with a light phase increase in CD8+ T cells. These results mirror previous analyses from our lab showing a light phase specific disruption in migratory factor expression in the meninges. Conclusions: These results show a time-of-day dependent effect of aging on T cell migration through the brain, driven by a light phase specific increase in CD8+ T cells. Gaining a better understanding of the mechanism driving this increase in CD8+ T cells could yield novel strategies to combat chronic neuroinflammation with age.

Funding: NIA Grant - R01AG078758; COI: N/A

Poster # 44: INFLUENCE OF MICROGLIAL DEPLETION ON IMMUNE CELL TRAFFICKING THROUGH THE AGING BRAIN.

Presenting author: John Chaney, BS (UT Austin)

Background: Humans over the age of 55 have an estimated 43% increased risk of developing dementia relative to adults under 55. A key contributor to dementia risk is age-related neuroinflammation, driven by a shift toward chronic microglia reactivity. Aging is also associated with changes in immune cells migrating through the brain, including an increase in CD8+ T cells . Recent work suggests that these two age-related changes may reinforce each other through a positive feedback loop: chronically reactive microglia release chemokines that recruit CD8+ T cells to the brain, and, in turn, these cells release proinflammatory cytokines to keep microglia in the reactive state. PLX-5622 is a CSF1 receptor inhibitor that can deplete microglia in the brain; once treatment stops, new microglia repopulate the brain. In neurodegenerative conditions, these repopulated microglia can express a less reactive phenotype. We hypothesize that microglial depletion by PLX-5622 will result in fewer peripheral immune cells migrating through the aged brain. Further, upon cessation of PLX-5622 treatment, we predict decreases in CD8+ T cells will persist, while other peripheral immune populations will return to baseline. Methods: To test this, we performed flow cytometry to quantify peripheral immune cell subpopulations, including CD8+ T cells, within the meninges of aged (24mos) Fisher 344 x Brown Norway rats. Meninges were chosen as they serve as the primary route by which peripheral immune cells enter and exit the brain. Peripheral immune cells were quantified in aged controls, aged rats treated with PLX5622, and a "recovery" group where PLX 5622 treatment was stopped to allow microglia to repopulate. Results: Preliminary data shows an association between microglial depletion and a reduction in several peripheral immune cell populations. After microglial repopulation, most populations return to their previous levels, aside from CD8+ T cells. Conclusion: These results suggest that microglia broadly drive peripheral immune cell migration to the brain; while specifically reactive microglia drive CD8+ T cell migration to the brain. Understanding this interaction could help guide strategies to disrupt age-related neuroinflammation and cognitive decline in the future.

Funding: R01AG062716; COI: No competing interests.

Poster # 45: ROLE OF ALTERNATIVE POLYADENYLATION IN MODULATING NEURONAL SENESCENCE AND INFLAMMATION IN AN IN VITRO AGING-LIKE MODEL

Presenting author: Katheeja Muhseena Neeraje, PhD (UT Houston)

Alternative polyadenylation (APA) is a key post-transcriptional mechanism that produces mRNA isoforms with distinct 3′ untranslated regions (3'UTR), thereby influencing transcript stability, localization, and translation efficiency. In neurons, precise APA regulation is critical for coordinating stress responses, axonal transport, and synaptic plasticity. NUDT21 (Nudix Hydrolase 21) is a core component of the APA machinery that modulates gene expression involved in cell cycle regulation, stress adaptation, and inflammatory signaling. Loss or deficiency of NUDT21 leads to global 3′UTR shortening, resulting in widespread dysregulation of genes controlling inflammation, senescence, and cellular homeostasis. Despite growing evidence linking APA dysregulation to neurodegeneration, the role of NUDT21 in mediating in vitro aging-like phenotypes remains largely undefined. This study investigates the role of NUDT21 in differentiated SH-SY5Y neurons, using three-week-old cultures as an in vitro model of aging-like phenotype. One- and three-week differentiated cells were compared to assess cell cycle status and the mRNA expression of senescence- and inflammation-related markers. Additionally, the impact of siRNA-mediated NUDT21 knockdown on the expression of these markers was evaluated to determine how NUDT21 deficiency influences cellular aging and inflammatory pathways. This study confirmed the neuronal differentiation of SH-SY5Y cells by βIII-tubulin (Tuj-1) staining. In flow cytometry analysis, three-week differentiated neurons demonstrated aberrant cell cycle re-entry, accompanied by increased p16 and NF-κB levels with simultaneous reduced NUDT21 expression. This is accompanied by increased expression of senescence-associated secretory phenotype (SASP) -associated factors, including GRO, IL-1β, and CCL5, collectively promoting inflammation and cellular senescence. Efficient knockdown of NUDT21 further upregulated NF-κB, IL-1β, CCL5, and p16, reinforcing the link between NUDT21 deficiency, pro-inflammatory signaling, and senescence activation. These findings demonstrate that deficiency of NUDT21, a key regulator of APA, enhances pro-inflammatory signaling and activates senescence pathways under in vitro aging-like conditions. Overall, our results highlight NUDT21 as an essential regulator of neuronal aging and SASP-like responses.

Funding: This work was supported by the University of Texas Health Science Centre at Houston. The authors acknowledge American Heart Association (AHA) - 23SCISA1143588, 23TPA1140944, and 24SCEFIA1262305 to VVG. We acknowledge Texas Alzheimer's Research and Care Consortium (TARCC) 1269191 funding and UTHealth seed funding to VVG.; COI: I declare that I have no known financial or personal conflicts of interest that could have influenced the work reported in this study.

Poster # 46: PLCG2 DEFICIENCY ALTERS CNS INFLAMMATORY AND LIPID RESPONSES TO LPS

Presenting author: Eduardo Gutierrez Kuri, MD, MS (UTH San Antonio)

Background: PLCG2, a critical immune regulator, has emerged as a key factor in Alzheimer's disease, with genetic variants conferring both risk and protection. In the brain, PLCG2 is predominantly expressed in microglia, influencing cellular functions including calcium signaling and inflammatory responses. Method: We investigated PLCG2's role in neuroinflammation using constitutive knockout mouse models subjected to lipopolysaccharide (LPS)-induced acute inflammation. Calcium imaging, transcriptomic profiling, and protein analyses were performed, followed by behavioral assessment. Result: In primary microglia, PLCG2 loss blunted LPSinduced calcium responses and reduced calcium release even after ionomycin stimulation, indicating impaired intracellular Ca^²⁺ handling or signaling capacity. In vivo, systemic LPS treatment induced over 200 differentially expressed genes in the brain, altering pathways related to innate and adaptive immune responses, inflammatory signaling, and microglial function. Knockout mice exhibited a distinct gene signature characterized by upregulated interleukin, prostaglandin, and interferon pathways. Consistent with the transcriptomic data, protein analysis confirmed elevated brain levels of the chemokine CCL2 (MCP-1), a key mediator of monocyte recruitment during inflammation, in LPS-treated knockout mice. Preliminary plasma measurements did not reveal clear genotype-dependent differences in CCL2, suggesting that the effect may be confined primarily to the brain, though additional samples are needed to confirm this. Targeted lipidomic profiling revealed significant increases in lysophospholipids in a genotype and treatment dependent manner. LPS exposure further accentuated these genotype dependent differences, indicating that inflammatory stimulation amplifies lipid alterations driven by PLCG2 deficiency. Elevated lysophospholipids have been associated with neuroinflammation and neurodegeneration. Interferon-responsive genes were markedly upregulated in homozygote knockout mice post-LPS, a phenotype associated with white matter pathology and inflammaging. Principal component analysis of myelin-enriched genes revealed LPS-treated homozygote knockout samples clustering separately from other genotypes. Behaviorally, LPS-treated knockout mice displayed reduced locomotion in open-field tests, correlating with expression of sickness behavior-related genes. Conclusion: These findings suggest that PLCG2 deficiency enhances interferon responses, potentially driving behavioral changes and myelin-related pathology. We are currently performing immunofluorescence staining of these brains to validate transcriptomic findings and to assess cell-type-specific and regional patterns of inflammation and myelin integrity.

Funding: This research was supported by the Alzheimer's Association (AARG-21-846012, SH and JPP); the William & Ella Owens Medical Research Foundation (JPP); and intramural funding (SH and JPP).; COI: None reported.

Poster # 47: EFFECTS OF CHRONIC JET LAG ON METABOLISM, BEHAVIOR, HEALTHSPAN, AND TAU PATHOLOGY IN A NOVEL MAPT GENE REPLACEMENT MOUSE MODEL

Presenting author: Andrea Gonzalez, Student Associate II (UTH San Antonio)

Background: Disturbed circadian rhythms from shift work, sleep disorders, or aging are linked to higher risk for metabolic syndrome and Alzheimer's disease (AD). In AD, circadian disruption manifests as fragmented sleep and behavioral symptoms, including sundowning. Emerging evidence suggests circadian dysfunction may contribute to disease onset and progression rather than simply reflect neurodegeneration. Methods: MAPT(H1.0*N279K) gene replacement (GR) mice harbor the human MAPT gene (H1 haplotype) carrying the pathogenic N279K mutationcausal for dominantly inherited frontotemporal dementia (FTDP-17)-which replaces the native mouse MAPT gene. At 4 months, these mice were housed in adjacent circadian cabinets: one under a normal light/dark cycle (NLC), the other under chronic jet lag (CJL; 8-hour phase advances twice weekly). Behavioral, metabolic, and healthspan assessments-including locomotor activity, spatial memory, aggression, repetitive behaviors, quantitative MRI, metabolic cage data, blood glucose and ketone levels, and frailty index-were performed longitudinally at multiple timepoints. FDG PET imaging was performed at 23 months, and brains were harvested at 24-25 months for tau pathology analysis. A separate 33-month-old cohort maintained under standard light cycles was also examined by immunofluorescence. Results: CJL mice exhibited lasting reductions in body and fat mass and sustained lower ketone production after fasting. CJL animals also showed increased frailty scores, spatial memory deficits, neuropsychiatric-like behaviors, and impaired glucose homeostasis appeared at initial timepoints and later resolved. FDG PET revealed no differences in brain glucose utilization between groups. At 24 months, histology revealed no AT8-positive neurons ("(pre)tangles") in either NLC or CJL mice. By 33 months, standard-housed mice exhibited numerous AT8-positive pyramidal neurons in hippocampal CA3, indicating age-dependent tau pathology in baseline conditions. Conclusion: Chronic circadian disruption in MAPT(H1.0*N279K)-GR causes metabolic and healthspan deficits without altering brain glucose uptake or accelerating tau accumulation in the conditions and timeframe tested. While some behavioral and metabolic effects are transient, lasting losses in body and fat mass and impaired ketogenesis remain. Tau pathology arises only at very advanced age, suggesting circadian misalignment impairs physiology and resilience but was not sufficient to trigger early tauopathy. Circadian-based interventions may extend healthspan even if they do not prevent agerelated tau accumulation.

Funding: 2024 William and Ella Owens Medical Research Foundation Award UT Health San Antonio Startup funds ; COI: None

Poster # 48: MODULATION OF MICROGLIA BY L-TYPE CALCIUM CHANNELS CAV1.2 AND CAV1.3

Presenting author: Brinda Palliyana, MS (UTH San Antonio)

Background: Microglia, the resident macrophages of the brain, play a major role in immune surveillance and maintaining homeostasis. One of the key pathways that regulates the activity of microglia, is intracellular calcium signalling. Triggered by various extracellular stimuli, intracellular calcium signals modulate microglia responses and can also influence lysosomal and autophagy functions, including phagocytosis, acidification, trafficking and debris clearance. Microglia dysregulation, dysfunction of lysosomes and impaired autophagy are notable features in pathogenesis of neurodegenerative diseases including Alzheimer's and Parkinson's diseases. Among the various potential sources of calcium influx, the L-type calcium channels (LTCCs) hold particular importance, as work from our group suggests that certain LTCC antagonists reduce microglia pro-inflammatory cytokine production, and can also reduce dystrophic neurite pathology in 5XFAD mouse model of Alzheimer's disease. However, studies elucidating the functional role of LTCC in microglia are limited. We hypothesize that microglial phenotype is regulated by LTCCs, particularly the subunits Cav1.2 and Cav1.3. Methods: To study this, we utilized in vivo microglia-specific conditional knockout (CKO) models of Cav1.2 and Cav1.3 to assess behaviour tasks like open field, novel object recognition and Barnes maze. In vitro, CKO of Cav1.2 and Cav1.3 in neonatal microglia were treated with lipopolysaccharide (LPS), and quantitative PCR was performed to check inflammatory and lysosomal gene expression. Results: In vivo, we found that microglial CKO of Cav1.2 and Cav1.3 in adult mice did not significantly alter behaviour in open field, novel object recognition, and Barnes maze tasks in the absence of immune stimulation. Future studies will examine how stimulation with LPS alters sickness behaviour in Cav1.2 and Cav1.3 CKO mice. In vitro, we found that CKO of Cav1.2 and Cav1.3 in neonatal microglia altered some responses to LPS. Quantitative PCR revealed that expression of lysosomal cysteine protease, cathepsin B (Ctsb) was suppressed by LPS, and 3-way ANOVA revealed an interaction between LPS and Cre treatment. Additionally, LPS increased expression of pro-inflammatory cytokine gene, Il1b. Conclusions: Overall, current results suggest that microglial Cav1.2 and Cav1.3 have limited role in regulating microglia phenotype, although additional work is required to understand how LTCC antagonists modulate microglia phenotype and function.

Funding:

Funding: R01AG085531; COI: None

Poster # 49: DELINEATING MICROGLIAL UPTAKE OF TAU FIBRILS AS A FUNCTION OF CELL-EXTRINSIC CUES

Presenting author: Celso S. G. Catumbela, PhD (UTH San Antonio)

BACKGROUND: Due to factors still unclear, microglia can either exacerbate or ameliorate the spread of tau pathology to influence the progression of Alzheimer's disease (AD). Perhaps, these opposing effects of microglia on tau accumulation contribute to sex and regional differences in vulnerability to tau pathology-two components of AD that remain poorly explained. Intriguingly, a recent report revealed that astrocytes modulate microglial uptake of amyloid-β (Aβ) aggregates to mitigate or worsen disease in vitro and in vivo. Perhaps, non-microglia cell types in brain also influence microglial uptake of tau fibrils to influence sex and regional differences in vulnerability to tau pathology. Further delineation of this topic may allow for development of therapies that mitigate sex- and region-specific vulnerability to tau pathology. METHODS: We isolated and cultured adult mouse non-microglia cells of both sexes from whole brains and separate brain regions (cortex+hippocampus, and cerebellum). Conditioned media were collected from mixed cultures of non-microglia cells and then applied to BV2 immortalized microglia for a period of 24h. Next, these pre-treated microglia were exposed to human recombinant P301S mutant tau preformed fibrils (PFFs) for 1h and, subsequently, submitted to confocal microscopy for quantification of microglial tau uptake propensity. RESULTS: So far, our data indicate that, relative to untreated controls, BV2 microglia pre-treated with conditioned media from nonmicroglia cells displayed significantly enhanced tau uptake propensity, albeit not in brain regionspecific manner (p < 0.05; n = 3 biological replicates). Furthermore, we pre-treated BV2 microglia with 24h- and 48h-conditioned media from mixed cultures of either male or female whole-brain non-microglial cells, which revealed trending effects of sex on microglial tau uptake propensity. Specifically, to enhance microglial tau uptake, the conditioned media required a longer period of incubation with non-microglia cells from males (48h), relative to incubation period with female counterparts (24h) (n = 3 biological replicates). CONCLUSION: Present data support a key role for non-microglia cells in modulation of microglial tau uptake, which may be influenced by sex. Future efforts will interrogate such cell-extrinsic effects on primary microglia as well as other tau-relevant processes including immune response.

Funding: This work is made possible by kind funding from the National Institute on Aging to Dr. Sarah C. Hopp (R01AG08553, K01AG066747). ; COI: We declare no conflict of interest.

Poster # 50: CDC25A EXPRESSION IS DIFFERENTIALLY MODIFIED BY HISPANIC ANCESTRY IN SINGLE & MIXED ETIOLOGY DEMENTIA

Presenting author: Elizabeth Ochoa PhD (UTH San Antonio)

Background: Hispanic individuals are at a 1.5-times higher risk for Alzheimer's disease and related dementias yet remain underrepresented in research. Longitudinal Hispanic cohorts provide antemortem biospecimen, cognitive, and epidemiological data, and GWAS/admixture mapping analyses identify Hispanic genetic ancestry as a modifier of Alzheimer's disease risk-associated gene expression. Additionally, neuropathological reports reveal a disproportional anatomydependent burden of Alzheimer's pathologies in Hispanic decedents. Despite these advances, few analyses investigate differential gene expression across dementia-associated pathologies from postmortem brain of Hispanic decedents. To address this gap, we investigated gene expression in Hispanic decedents with single or mixed etiology dementia. Method RNA extracted from postmortem frontal cortex brain tissue of Hispanic and non-Hispanic decedents with Alzheimer's disease (AD, n=6), cerebrovascular disease (CVD, n=6), or mixed etiology dementia (ADCVD, n=6) was hybridized to the NanoString Human Neuroinflammation panel, with gene expression quantified via SPRINT nCounter. Analysis completed via ROSALIND® using NanoString Advanced Analysis criteria. Low counts and low background genes were omitted from analysis. Targets were evaluated for Drosophila melanogaster orthologs, and causality tested using transgenic fly models of human amyloid beta (AB) and tau. Neurodegeneration quantified via locomotor and immunohistochemical methods. Result Comparison of Hispanic decedents with AD or ADCVD to non-Hispanic decedents reveals shared pathway enrichment of the Astrocyte Function gene set. Respective to non-Hispanic decedents, two genes in Hispanic AD and seven genes in Hispanic ADCVD are differentially expressed. Analysis indicates CDC25A is decreased, while eight other genes are increased at the transcript level in the frontal cortex of Hispanic decedents with CVD when compared to all decedent samples analyzed. Interestingly, panneuronal overexpression of string (stg), a CDC25A ortholog, is sufficient to reduce Drosophila locomotion, a trend consistent in the context of human AB42. Conclusion These preliminary findings suggest an ancestry- and etiology-specific transcriptional profile in CVD and point to genes modified by ancestry in AD and ADCVD. Ongoing experiments aim to determine differential gene expression relative to pathology in both postmortem human brain and transgenic Drosophila. Overall, we seek to identify Hispanic ancestry- and disease-specific genes among dementia etiologies as an avenue for precision medicine care among an at-risk, underrepresented population.

Funding:

Funding: TARCC Postdoctoral Fellowship, NIGMS IRACDA K12 GM111726, NIA P30AG066546, Bartell and Mollie Zachry Endowment, Reed Precision Medicine Center; COI: N/A

Poster # 51: THE ROLE OF THE INTEGRATED STRESS RESPONSE IN ALZHEIMER'S DISEASE

Presenting author: Kyle Tengler, MS (UTH San Antonio)

Background: Alzheimer's disease (AD) is a leading cause of dementia and poses a massive and growing public health burden. Despite enormous research efforts, the mechanisms underlying AD pathogenesis remain incompletely understood, limiting effective clinical interventions. Emerging evidence suggests that the chronic activation of Integrated Stress Response (ISR), a conserved signaling pathway regulating cellular responses to viral, metabolic, oxidative, and ER stress may contribute to neurodegenerative processes in AD. Aberrant ISR activation has been implicated in synaptic dysfunction, neuronal loss, and memory impairment. To better understand the role of the ISR in AD, we analyze transcriptional changes of ISR target genes in established mouse models of the disease. Methods We analyzed publicly available microarray data from 192 5xFAD and wildtype (WT) mice, including both sexes and spanning ages 4 to 18 months. Additionally, we generated and examined RNA-seq data from 10-month-old PS19 and WT mice. Differential gene expression analysis focused on canonical ISR target genes, comparing expression profiles between AD model and WT mice to identify disease-associated transcriptional changes. Results A subset of ISR target genes was found to be differentially expressed in both 5xFAD and PS19 AD mouse models compared to their respective WT controls. These changes were observed across multiple ages and brain regions, with several ISR-related genes showing consistent up- or down-regulation in AD models, suggesting activation of stress response pathways during disease progression. The identified gene expression patterns overlapped with known markers of neuroinflammation and proteostasis disruption, key features of AD pathology. Conclusion Our findings indicate that ISR-related transcriptional changes occur in multiple mouse models of Alzheimer's disease. The consistent differential expression of ISR target genes across distinct pathological contexts supports a potential role for the ISR in AD pathogenesis. These results provide a basis for further studies investigating whether modulation of the ISR could represent a therapeutic strategy to mitigate neurodegeneration in Alzheimer's disease.

Funding: NIA awarded funds; COI: None

Poster # 52: ASTROCYTE L-TYPE CALCIUM CHANNELS REGULATE TAU HANDLING AND STRESS-RESPONSE

Presenting author: Lubov Ezerskiy PhD (UTH San Antonio)

Alzheimer's disease and related dementias (ADRD) affect over 6 million Americans and almost 460,000 Texans annually, yet disease-modifying therapies remain elusive. Tauopathies, a class of neurodegenerative disease, are characterized by the abnormal accumulation of hyperphosphorylated tau. Pathological tau inclusions have previously been associated solely with neurons where tau is predominantly expressed. However, many tauopathies also exhibit tau deposits in astrocytes and microglia. These deposits contribute to disease progression through synaptic dysfunction, neuronal death, and severe gliosis. Preliminary patient autopsy data showed that AD patients who happened to receive L-type calcium channel (LTCC) blockers displayed lower Braak stage and improved cognition. These drugs act on many cell types, including microglia and astrocytes, but their neuroprotective mechanism is unknown. We hypothesized that nimodipine, an FDA-approved LTCC blocker, would improve astrocytic responses to tau exposure, thereby slowing disease progression. Primary cortical astrocytes isolated from C57BL/6J mice were purified, cultured, and pretreated with 5µM nimodipine before a 24-hour incubation with 50nM wild-type 2N4R tau pre-formed fibrils (PFFs). Nimodipine-treated astrocytes exhibited reduced tau uptake and transcriptomic changes that shifted toward untreated controls. While reduced tau uptake was surprising, it is possible that dampening LTCC activity curtails calcium-dependent internalization and/or limits intracellular storage of pathogenic tau, thereby preserving homeostatic functions and decreasing the potential for subsequent seeding to neighboring cells. Alternatively, LTCC blockade may facilitate more efficient degradation of internalized tau. These mechanistic possibilities will be tested in ongoing work. Together, our findings support a model in which restraining LTCC overactivity in astrocytes normalizes stress responses to extracellular tau and reduces cellular tau burden. Given the clinical availability of nimodipine, targeting glial LTCCs represents a potential therapeutic strategy to slow or prevent tau-driven disease progression.

Funding: None reported.; COI: None reported.

Poster # 53: CHARACTERIZATION OF NOVEL MOUSE MODELS OF MICROGLIAL CONDITIONAL KNOCKOUT OF L-TYPE CALCIUM CHANNEL CAV1.2 AND CAV1.3 SUBTYPES

Presenting author: Rachael Cundey, BS (UTH San Antonio)

Background: L-type calcium channel (LTCC) antagonists act at Cav1.2 and Cav1.3 and have been associated with reduced risk of dementia in humans, but the mechanism of this effect remains unclear. We have previously observed that LTCC antagonists reduce microglial transition to a pro-inflammatory state. We posit that LTCC antagonists act at microglial Cav1.2 or Cav1.3 to exert their neuroprotective effect by blocking the neurotoxic effects of proinflammatory microglia. Method: To test the hypothesis that microglial Cav1.2 or Cav1.3 controls microglia phenotype, we generated Cav1.2 and Cav1.3 microglia-specific conditional knockout (CKO) mice. We crossed CX3CR1-CreERT2 mice with Cav1.2fl/fl or Cav1.3flexGFP mice to generate Cav1.2 CKO (CKO2) and Cav1.3 CKO (CKO3) mice, respectively. We assessed the effects of each CKO on mouse behavior in tests for locomotion (open field), recognition memory (novel object recognition), and spatial learning and memory (Barnes maze). We also measured inflammatory gene expression using qPCR. For each test, we compared each CKO to the corresponding floxed control mouse and all groups to CX3CR1-CreERT2 controls. Result: In the Barnes maze, Cav1.2fl/fl and CKO2 mice had a significantly increased latency to target during training compared to Cre-only mice, suggesting impaired spatial learning in the former genotypes. There were significant changes in inflammatory gene expression in the brains of Cav1.3flexGFP and CKO3 mice including an increase in TGF-β expression in CKO3, suggesting increased anti-inflammatory signaling in these mice. Conclusion: Preliminary results reveal an anti-inflammatory effect of microglial Cav1.3 knockout and spatial learning impairment in microglial Cav1.2 knockout. These findings indicate that LTCC antagonist action at microglial Cav1.3 may be the source of the neuroprotective effect of these drugs. Further work will include immunohistochemical staining to determine the efficacy of each microglial conditional knockout.

Funding: National Institutes of Health [R01AG085531]; COI: n/a

Poster # 54: HNRNPK REGULATES ACSL4 TO MODULATE SEX-SPECIFIC LIPID PATHWAYS

Presenting author: Amay Parmar, BS (UTHealth Houston)

Background: Heterogeneous nuclear ribonucleoproteins (hnRNPs) are a diverse family of nucleic acid-binding proteins that regulate nearly all DNA- and RNA-dependent processes. HNRNPK, a member of this family, is a multifunctional DNA- and RNA-binding protein involved in transcriptional regulation, RNA splicing, mRNA stability, and translation. Mutations in HNRNPK cause Kabuki-like and Au-Kline syndromes, both characterized by progressive neurological decline. Hnrnpk knockout in mice is embryonically lethal, and hnRNPK reduction impairs synaptic plasticity, indicating the significance of hnRNPK. Nevertheless, the molecular mechanisms by which HNRNPK functions in neurons remain poorly understood. Methods We performed a genome-wide gene expression analysis (RNA-seq) to identify genes modulated by primary neurons expressing hnRNPK, thus allowing us to define genes responsive to hnRNPK regulation. Differential expression and pathway enrichment analyses were conducted using Advaita iPathwayGuide to define molecular networks responsive to hnRNPK. Among the highlighted genes, the X-linked ACSL4 (Acyl-CoA synthetase long-chain family member 4) showed consistent enrichment in multiple lipid metabolism pathways modulated by hnRNPK and was selected for validation by qPCR, protein expression, and functional studies. Results We found that hnRNPK significantly changed the expression of 1,956 of 23,721 genes. hnRNPK modulated neuronal networks involved in differentiation, migration, remodeling, and synaptic plasticity, and it also affected pathways linked to cell death, cell cycle, cognition, learning and memory, autophagy, sex differentiation, and amyloid formation. Many of the affected pathways were enriched for lipid regulation and metabolism, with Acsl4 prominently represented in multiple lipid networks. We further showed that ACSL4 expression is sex dependent in primary neurons, with higher levels in females, and that ACSL4 modulates autophagy in both male and female neurons. Finally, female human brains expressed higher levels of ACSL4 than male brains, with even higher levels in female brains with Alzheimer's disease. Conclusions Our study showed that hnRNPK regulates neuronal transcription, with a strong influence on lipid metabolic pathways through the X-linked Acsl4 gene. These findings provide insight into how hnRNPK dysfunction may contribute to age-related neurodegenerative diseases.

Funding: TARCC grant 2024-2026 (Andrey Tsvetkov); COI: None

Poster # 55: LATERAL BINDING OF OLIGOMERIC AΒ42 REMODELS THE PROTEASOME TO DRIVE COGNITIVE DECLINE IN ALZHEIMER'S DISEASE, REVERSED BY PROTEASOME ACTIVATION

Presenting author: Andrew Pickering, PhD (UTHealth Houston)

Background: Impaired proteasome activity is a robust and reproducible feature of Alzheimer's disease (AD) in both patients and animals models. Prior work suggest that β-amyloid (Aβ) aggregates may inhibit proteasomes, but it is unclear if this reflects direct structural interference or indirect toxicity. The proteasome controls a myriad of critical neuronal processes with its disruption potentially driving downstream AD pathology. Methods and Results We systematically examined how distinct Aβ42 and Aβ40 forms: monomeric, oligomeric, and fibrillar, impact proteasome structure, dynamics, and function. Consistently with acknowledged intracellular toxicity of soluble Aβ42 oligomers, oligomeric Aβ42 selectively impairs 20S and 26S proteasome activity, while monomeric and fibrillar forms as well as Aβ40 had minimal inhibitory effects. Morphometric analysis of AFM images of proteasome particles suggests that oligomeric Aβ42 binds laterally to the 20S proteasomes, and impairs proteasome dynamics by disabling conformational shifts toward an open gate conformation, critical for catalytic activity. Oligomers-decorated proteasome particles are locked in catalytically inert cycling between closed-gate and intermediate conformations. Oligomeric Aβ42 also induce disassembly of the 26S complex, increasing participation of free 20S cores, as confirmed by native gel and AFM imaging. Importantly, we show that proteasome dysfunction induced by the oligomeric Aβ42 is both preventable and reversible by pharmacological intervention. Employing a proteasome activator TAT-DEN which promotes opening of the proteasome gate we restored proteasome activity, and reduced the conformational remodeling induced by oligomeric Aβ42. In cellular and animal models, proteasome activation mitigated Aβ-induced cell death, memory deficits, and survival impairments, even under conditions of pre-established Aβ deficits. Oligomeric Aβ42 but not other Aβ42 species, can decorate the lateral sides of the 20S proteasome restricting cycling between conformations and in particular restricting gate opening. Oligomeric Aβ42 also reduces the pool of 26S proteasomes impairing the ubiquitin proteasome system. Conclusions We demonstrate proteasome dysfunction can be both prevented and reversed by proteasome agonists. We show that proteasome activation with agonists is sufficient to prevent cognitive deficits in AD animal models. This suggests proteasome dysfunction is a key downstream target in AD pathology and supports proteasome activation as a therapeutic strategy to restore neuronal function in AD.

Funding: This work was supported by the National Institute of Aging R01 AG065301 (to A.M.P.), National Institute of General Medical Science R01 GM069819 (to M.G.), William and Ella Owens Medical Research Foundation (to M.G. and P.A.O.) and Fulbright Scholarship program (MB). ; COI: A.M.P., M.G., and P.A.O are inventors on a patent application related to this work filed by The University of Texas Health Science Center at San Antonio (HSC1567). The authors are also co-founders of ProAllostera Pharmaceuticals.

Poster # 56: STUDYING AMYLOID-INDUCED MTORC1 ACTIVATION IN DROSOPHILA

Presenting author: Andy Delgado, BA (UTHealth Houston)

Studying Amyloid-induced mTORC1 Activation in Drosophila Andy Delgado, Andrew M. Pickering

Background: Disturbances in the mammalian target of rapamycin complex 1 (mTORC1) is frequently reported in Alzheimer's disease (AD) patients and various mammalian model organisms and cell systems. As both AD and mTORC1 are strongly associated with aging, the fruit fly Drosophila Melanogaster could be used for its advantages of short lifespan and control of gene expression. The S6 kinase (S6K) and 4E binding protein 1 (4E-BP1) are the best characterized downstream effectors of mTORC1 and key regulators of translation.

Phosphorylation of S6K and 4E-BP1 are routinely used as a readout of mTORC1 activity and is important for studying mTORC1-mediated translation. In this Poster, amyloid-beta's effect on mTORC1 in the fruit fly model is investigated at various ages. Methods: Flies panneuronally expressing amyloid-beta were aged to 20 and 40 days of age and tested for S6K and 4E-BP1 phosphorylation via western blot. Results: Flies at 20 days of age show substantial increase in S6K phosphorylation, but no increase in 4E-BP1 phosphorylation. They also see a dramatic downregulation of the total S6K and 4E-BP1 proteins. Flies at day 40 see no changes in S6K and 4E-BP1 abundance or phosphorylation. Conclusion: The fruit fly recapitulates the well-established mechanism of amyloid stimulation of mTORC1. However, at least when using a constitutive expression system, this activation seems to happen strongly at early adult stages but is not observed at later ages. Future studies should test flies in early adulthood, to see if younger adults yield phosphorylation phenotypes without changes in total abundance. The mTORC1 activation of various flies could also be explored throughout the lifespan as a possible mechanism of resilience. Studies focused on mTORC1-mediated translation in AD can be done in flies.

Funding: N/A; COI: N/A

Poster # 57: BRAIN IRON METABOLISM IS REGULATED IN A SEX-DEPENDENT MANNER IN AN AGED MOUSE MODEL OF VASCULAR DEMENTIA

Presenting author: Areej Ashraf (UTHealth Houston)

Background: Impairment of the iron metabolism in the brain is associated with multiple brain pathologies, including vascular contributions to cognitive impairment and dementias (VCID). However, the underlying mechanisms are not well understood yet. Method: We utilize an aged mouse model of VCID, where micro-coils (0.18 mm diameter) are surgically implanted around the two bilateral carotid arteries (BCAS, bilateral carotid artery stenosis). Mice were maintained for two months. Mice were tested for behavior (open field, elevated plus maze, novel object recognition, and fear conditioning), and their brains were analyzed by IHC, Western blotting, and qPCR. Results: We found that female BCAS mice exhibited cognitive deficits, neuroinflammation, and demyelination, compared with female sham mice. We also found that the brains of aged female BCAS mice had enhanced levels of ferrous iron and increased iron deposition, compared with female sham mice. Interestingly, these differences were not observed between male sham and male BCAS mice. We found that the expression of genes involved in iron transport and iron storage was significantly increased in the brains of male BCAS mice, but not in female BCAS, compared with sham mice. Western blot data revealed that ferritinophagy was more efficiently inhibited in the brains of male BCAS mice, compared with male shams, but not in female BCAS mice. Conclusion: Our study establishes a foundation for sex differences in brain iron metabolism in a mouse model of VCID in aged mice, which may help to identify different approaches to treat patients of both sexes with vascular dementia.

Funding: NIA (R21AG075750) Texas Alzheimer's Research and Care Consortium (#957578) American Heart Association (#856061); COI: The authors have no relevant financial or nonfinancial interests to disclose.

Poster # 58: EFFECT OF ANTI-MICROBIAL IMMUNIZATIONS IN THE PROGRESSION OF AMYLOID PATHOLOGY IN A MOUSE MODEL OF ALZHEIMER'S DISEASE

Presenting author: Catalina Valdes, BS (UTHealth Houston)

Background: The role of the immune system as a biological contributor of Aβ and tau deposition has been extensively explored. Most of this research has focused in describing the detrimental effects of neuroinflammation. In turn, the role of the peripheral immune system in brain's health has been reported as critical in preventing microbial invasion and AD pathology. Along this line, studying the role of antimicrobial vaccinations are relevant in the context of brain health. Antimicrobial vaccinations have been associated with the preservation of brain health in aging and disease contexts. The support for this hypothesis comes from epidemiological studies indicating a reduction in the incidence of AD in elderly populations that get periodically vaccinated. Method: Whether vaccinations can prevent the progression of the pathogenic markers of AD is highly relevant. We hypothesize that Alzheimer's amyloid and inflammatory pathologies can be reduced in a vaccine-technology dependent manner. To asses this question, a mouse model which produce Aβ deposition in the brain was chosen to study the role of immunogenic treatments in AD pathology. Three highly recommended vaccines were tested; COVID-19 mRNA, Inactivated influenza virus, and a polysaccharide-based vaccine to prevent pneumonia. Result: The activation of the peripheral immune system was confirmed in all vaccinated mice. Regardless, rescue of short-term memory loss was observed only in animals injected with the COVID-19 and influenza vaccine. This was correlated with unique cytokine signatures in the brain. We also evaluated the amyloid deposition in the brain of treated mice and found that only the of COVID-19 mRNA injected animals showed a decreased in total burden. When checking the total number of plaques in each group, we found a reduction in those treated with the COVID-19 and influenza vaccines. Finally, when we quantified the specific size of plaques within the hippocampus, we observed fewer small, medium and large size plaques in the same treated groups. Conclusion: Overall, our findings support the idea that antimicrobial vaccines not only prevent severe infections but also protect the brain from neurodegenerative insults.

Funding: Nothing to disclose; COI: Nothing to disclose

Poster # 59: INTERACTIONS BETWEEN BACTERIAL AND MAMMALIAN AMYLOIDS IN DIFFERENT MODEL SYSTEMS.

Presenting author: Francisca Bravo-Risi, PhD (UTHealth Houston)

Background: Protein misfolded diseases (PMDs) are characterized by the accumulation of misfolded protein (MP) aggregates generated through nucleation-polymerization processes. These can be accelerated by adding MP in homologous or heterologous (cross-seeding) fashions. Notably, these mechanisms are not limited to mammalian amyloids (MAs), but also present for bacterial amyloids (BAs). Moreover, the presence of pathogens and their molecules has been reported in the periphery and CNS of Alzheimer's disease (AD) patients, raising the question whether BAs trigger the misfolding of mammalian amyloids involved in AD. Methods: We studied the seeding activity of BAs (CsgA, Mcc, and TasA) over Aβ40 by PMCA. To further explore interactions in a cellular context, we exposed the biosensor cells HEK293T-FRET-tauP301S and HEK293T-FRET-α-Syn-A53T, expressing mutated tau and α-Syn, with Mcc and TasA. Bioassays in animal models of Aβ, tau, and α-Syn pathologies were also included. Results: We found that: i) the three BAs promoted Aβ aggregation in vitro, ii) we observed specificity in brain tropisms in CsgA-treated mice, and iii) Mcc promoted specifically tau aggregation in the biosensor cells. Conclusion: Overall, our findings highlight both the broad and specific cross-interaction between amyloids and its extent to functional ones. This data is relevant for understanding the molecular mechanisms behind the progression of PMDs.

Funding: Alzheimer's Association (AARGD-22-972125) and UTHealth-Houston funds.; COI:
The authors have no conflict of interest.

Poster # 60: SPATIAL TRANSCRIPTOMICS AND SNRNA-SEQ UNRAVEL CELL-CELL INTERACTIONS AND SIGNALING PATHWAYS IN THE ENTORHINAL CORTEX DRIVING ALZHEIMER'S DISEASE INITIATION AND PROGRESSION

Presenting author: Haichao Wei, PhD (UTHealth Houston)

Background: Alzheimer's disease (AD) is a progressive neurodegenerative disorder characterized by complex molecular and cellular changes that lead to cognitive decline and dementia. The entorhinal cortex (EC) is one of the earliest cortical regions affected in AD, and the propagation of amyloid-β and tau pathology from the EC is thought to drive disease progression. However, the cellular and molecular mechanisms underlying the selective vulnerability of EC neurons remain poorly understood. Emerging evidence indicates that astrocytes play critical roles in modulating AD pathology and influencing neuronal health, suggesting that glial-neuronal interactions may contribute to early disease onset. Methods: To elucidate the cellular and molecular landscape of EC degeneration, we applied CosMx spatial transcriptomics at true single-cell level across healthy, early, mid, and late AD stages. We further integrated these data with single-nucleus RNA sequencing (snRNA-seq) to validate and characterize the identified transcriptional changes across distinct glial and neuronal populations in the EC. This multi-omics approach enabled high-resolution mapping of spatial gene expression and intercellular communication dynamics during AD progression. Results: Our analysis uncovered distinct transcriptional programs and spatially resolved cell-cell interactions that varied with disease stage. We identified pronounced astrocyte-neuron signaling networks in early AD, suggesting that astrocyte-derived molecular cues may initiate or exacerbate neuronal dysfunction. Multiomics integration revealed stage-dependent changes in ligand-receptor signaling, highlighting astrocytic pathways that modulate neuronal homeostasis and synaptic integrity. Furthermore, we identified a set of astrocyte-enriched, early stage-specific genes potentially involved in the initiation of AD-related neurodegeneration. These findings suggest that astrocytes may play pivotal roles in shaping the earliest pathological landscape of the EC. Conclusion: Our findings provide new insights into the molecular mechanisms underlying EC vulnerability in AD. By linking spatially resolved gene expressions to cellular communication networks, our study highlights potential astrocyte-mediated mechanisms driving early neurodegenerative changes and identifies candidate targets for early intervention aimed at slowing or preventing AD progression.

Funding: R01AG078728-01, R21 NS113068, the Amy and Edward Knight Fund, and the UTHSC Senator Lloyd Bentsen Center for Stroke Research; ; COI: No

Poster # 61: CHARACTERIZING THE BASIC MECHANISMS UNDERLYING THE PATHOGENIC SPREAD OF SOLUBLE TAU.

Presenting author: Madeline Marques, PhD (UTHealth Houston)

Background: Tauopathies are highly heterogeneous diseases characterized by the Tau isoforms (3R vs. 4R) found in the pathological inclusions, as well as the cell types and brain regions affected. Previous work has found that not all Tau species are equally capable of pathogenic seeding and propagation, suggesting that important differences in structural and/or biochemical properties of Tau isoforms could be playing an important role in disease-specific trajectories.

However, the mechanisms underlying how soluble Tau spreads in the brain remain incompletely understood. Method Although Tau neuropathology has been studied extensively in Drosophila, there are no reports thoroughly assessing Tau spread in the fly brain. To rectify this gap in knowledge, we are characterizing which of the 6 major human Tau (hTau) isoforms spread in the fly brain, co-expressing Tau isoforms with GFP in glutamatergic neurons and examining the localization of Tau. To identify which cell populations are receiving the released Tau, we are using single-cell RNA sequencing. To identify cognate receptors needed for internalization of Tau into recipient cells we are using in-situ cell surface proteomics. Our study will be the first to use a fully integrated approach to examine the mechanisms underlying the release of Tau isoforms from glutamatergic neurons, the identities of cell-types that are competent to accept released Tau, and the cell-surface proteins needed for internalization of Tau by recipient cells. Result Our preliminary findings indicate that the 6 major Tau isoforms spread to varying extents in both larval and adult fly brain. Next, we will identify the cell populations accepting each of the Tau isoforms that spread from glutamatergic neurons, using single-cell RNA sequencing. The use of Drosophila for these studies allows for rapid, affordable, and exhaustive investigations that may have otherwise been prohibitively expensive or time consuming in vertebrate models. Conclusion Our findings suggest that key differences in the structural and/or biochemical properties of these Tau isoforms may underly their varying propensities to spread. We predict our integrative approach will allow us to identify new mechanisms of Tau spread, as well as cellsurface receptors that may be targeted to mitigate soluble Tau spread.

Funding: This project will be supported by a TARCC Postdoctoral Fellowship Grant (Award ID: 1539058; PI: Dr. Madeline Marques). This work project is also supported by an NIH R01 grant (Project Number: 1RF1AG072176-01; PI: Kartik Venkatachalam).; COI: No conflicts of interest to declare.

Poster # 62: MODULATION OF ASTROCYTE REACTIVITY BY LRP1 IN NEUROINFLAMMATION

Presenting author: Meng Wang, BS (UTHealth Houston)

Chronic neuroinflammation significantly increases the risks of neurodegenerative diseases, particularly in aging and in pathological conditions such as brain injury and infection. A hallmark of chronic neuroinflammation is the shift of astrocytes into reactive states, including subtypes that disrupt neuronal function and actively cause neural cell death. Markers associated with these neurotoxic reactive astrocytes are substantially elevated in brain tissue from individuals with major neurodegenerative diseases such as Alzheimer's and Parkinson's, implicating them in disease pathogenesis. Targeting mechanisms that modulate astrocyte reactivity may therefore help preserve neuronal integrity in these conditions. This project investigates the role of astrocytic low-density lipoprotein receptor-related protein 1 (LRP1) in regulating inflammatory responses. We hypothesize that astrocyte-LRP1 dampens astrocyte reactivity to pro-inflammatory signals. To test this, we are assessing how astrocyte-specific LRP1 knockout (KO) alters astrocyte responses to cytokine exposure. Our current findings indicate that astrocyte-specific LRP1KO enhances cytokine-induced reactivity in vitro, promoting astrocytic transition into reactive states with elevated expression of genes associated with neurotoxic astrocyte phenotypes.

Ongoing experiments are evaluating the functional impact of astrocyte-LRP1KO on neurons in co-culture, characterizing the effects of astrocyte-specific LRP1KO in vivo, testing whether an LRP1 agonist can reduce astrocyte reactivity after inflammatory stimulation, and profiling wholebrain transcriptomic changes in response to astrocyte-LRP1KO via single-cell RNA sequencing.

Funding: NIH T32AG082661 to MW, VA Career Development Award IK2BX003240 to NLS, NIH 1R01NS132778 to NLS, NIH T32GM113896 and T32GM145432 to STX-MSTP.; COI: None reported.

Poster # 63: SEPSIS AS A RISK FACTOR FOR ALZHEIMER'S DISEASE: TRANSLATIONAL INSIGHTS FROM MOUSE MODELS TO ICU PATIENTS

Presenting author: Mohd Tayyab, PhD (UTHealth Houston)

Background: Alzheimer's disease (AD), the predominant cause of dementia, has been epidemiologically linked to hospital-diagnosed infection such as sepsis, which increase dementia risk by 1.2-5-fold. Sepsis induces widespread immune dysregulation, systemic inflammation, and blood-brain barrier disruption, leading to persistent neuroimmune activation and cognitive decline. The mechanistic pathways connecting sepsis to AD pathogenesis remain poorly defined. Emerging evidence suggests that sex-specific immune and neuroinflammatory responses modulate these outcomes. We hypothesize that systemic infection accelerates AD-related neurodegeneration through inflammatory and neuroimmune cascades in a sex-dependent manner. Methods: In preclinical study, 50-day old male and female APP/PS1 mice (n=6/group) underwent cecal ligation and perforation (CLP) or sham surgery. At 130 days post-CLP, immunofluorescence quantified IBA1, GFAP, and amyloid-β (Aβ) in the prefrontal cortex (PFC) and hippocampus, and inflammasome markers in female PFC. Clinically, plasma from 263 ICU patients with bacterial (n=122) or viral sepsis (COVID-19; n=101), 20 AD patients, 37 sepsis patients with delirium, and 20 controls was analyzed for neuroinflammatory and AD-related biomarkers using the NULISA proteomic assay. Results: Our preclinical study demonstrated that APP/PS1 mice subjected to polymicrobial sepsis showed increased GFAP⁺ astrocytes and IBA1⁺ microglia in the hippocampus and PFC of both sexes, with females exhibiting greater astrocyte reactivity, microglial activation, and inflammasome signaling (ASC, NLRP3). Sepsis also increased Aβ deposition in female PFC, indicating accelerated AD pathology. Clinically, plasma proteomics revealed distinct signatures, with bacterial sepsis sharing 50 differentially expressed proteins with AD versus 33 for viral sepsis. Overlapping biomarkers (GFAP, S100B, SNCA, UCHL1, pTau217) reflected neuroinflammation and tau pathology. Bacterial sepsis shared canonical AD markers (NEFL, NRGN, Aβ1-42, pTDP43.409), while viral sepsis showed unique inflammatory pathways. Delirium correlated with elevated IL-17, NEFL, TARDBP, Aβ1-42, and pTau217, linking sepsis to AD-related neurodegeneration. Conclusion: Sepsis triggers pronounced neuroinflammation and accelerates AD-related pathology in APP/PS1 mice, with females showing greater vulnerability. Clinically, bacterial sepsis mirrors AD-associated biomarker profiles, while viral sepsis exhibits distinct inflammatory patterns. These results suggest systemic infections may initiate or amplify neurodegenerative cascades, highlighting the importance of early biomarker-based monitoring and interventions to prevent long-term cognitive decline in sepsis survivors.

Funding: This work was supported by UT Health, Houston granted TARCC 2022-26, NIH/NIA grant RF1AG072491/R01 AG072491 to RM and TB.; COI: No

Poster # 64: UNRAVEL THE ROLE OF PEROXISOME PROLIFERATOR-ACTIVATED RECEPTOR DELTA (PPARΔ/Β) IN MICROGLIAL PHAGOCYTOSIS ASSOCIATED WITH ALZHEIMER'S DISEASE

Presenting author: Abhijeet Arun Patil, MS (UTHealth Houston)

Alzheimer's disease (AD) is a progressive neurodegenerative disorder characterized by amyloid-β (Aβ) plaques, tau tangles, and chronic neuroinflammation, leading to synaptic dysfunction and cognitive decline. Microglia, the brain's resident immune cells, are essential for Aβ clearance and maintenance of homeostasis, but their function becomes dysregulated during disease progression. Peroxisome proliferator-activated receptor beta/delta (PPARβ/δ), a nuclear receptor that regulates lipid metabolism and inflammation, is highly expressed in microglia and serves as a key transcriptional regulator of their metabolic state. In vitro activation of PPARβ/δ using the selective agonist KD3010 significantly enhanced microglial phagocytosis and upregulated target genes involved in lipid metabolism (ANGPTL4, CD36) and RNA regulation (Qki), indicating a shift toward a metabolically adaptive, homeostatic phenotype. Supporting this, single-nucleus transcriptomic data from human AD brains demonstrate that multiple microglial subclusters exhibit disease-associated gene-expression changes, including dysregulation of lipid-handling, phagocytic, and inflammatory pathways, linking our mechanistic findings to human pathology. Future studies will use the 5xFAD model with microglia-specific PPARβ deletion (PPARβL/LCx3cr1-Cre) to determine its impact on amyloid pathology, neuroinflammation, and cognition. Collectively, these studies will elucidate how PPARβ/δ coordinates microglial lipid metabolism and immune function and establish it as a potential therapeutic target to mitigate Alzheimer's disease progression.

Funding: None reported.; COI: None reported.

Poster # 65: G4-RNA DESTABILIZATION REVEALS SEX- AND AGE-LINKED MECHANISMS IN BRAIN PERICYTES

Presenting author: Nicholas Turner, BS (UTHealth Houston)

Background: Pericytes are critical for maintaining cerebrovascular stability and blood-brain barrier integrity. Their dysfunction contributes to vascular leakage, neuroinflammation, and neurodegeneration during aging. Although sex and age are key biological variables influencing cerebrovascular health, the molecular basis of pericyte aging and its sex dependence remains poorly understood. G-quadruplex (G4) structures are non-canonical four-stranded nucleic acid elements that regulate transcription and translation. Dysregulation of G4 structures has been linked to aging, inflammation, and neurodegenerative disease, but their role in pericyte biology is unknown. Phenylpyrrolocytosine (PhpC) is a small molecule that selectively destabilizes G4RNA structures, providing a tool to probe how RNA structural dynamics influence cellular proteomes. We hypothesize that G4-RNA remodeling contributes to sex- and age-specific molecular changes. Methods Primary brain pericytes were isolated from young (4 months) and aged (22 months) wild-type male and female mice and cultured under basal conditions or treated with PhpC. G4s were stained using the G4-specific fluorescent probe N-TASQ to assess age- and sex-dependent differences. Quantitative proteomics followed by mass spectrometry was performed to assess global protein expression. Differentially expressed proteins and pathways were analyzed using Advaita iPathwayGuide to identify molecular processes affected by age, sex, and G4-RNA destabilization. Results N-TASQ staining revealed a marked accumulation of G4s in aged pericytes compared to young cells, indicating increased G4 stress with age. Proteomic analysis revealed distinct sex-dependent protein signatures in pericytes, with female cells enriched for cytoskeletal and junctional components and male cells enriched for translational and chromatin-associated factors. Aging amplified these differences, highlighting alterations in actinregulatory and RNA-processing pathways. PhpC treatment minimally affected young cells but enhanced sex-specific divergence in aged pericytes, suggesting increased sensitivity of translational control to G4-RNA modulation with age. Among the most responsive targets were Shroom family proteins involved in actin organization and cell polarity, including the X-linked Shroom4, whose sex-linked expression may underlie differences in cytoskeletal regulation and contribute to sex-specific responses during pericyte aging. Conclusion Our findings suggest that age and sex jointly shape the pericyte proteome through G4-RNA-dependent mechanisms. PhpCsensitive pathways reveal a potential link between RNA structural dynamics, cytoskeletal remodeling, and sex-specific aspects of cerebrovascular aging.

Funding: American Federation for Aging Research and Glenn Foundation for Medical Research Breakthroughs in Gerontology (BIG) Award, #BIG21042 (Andrey Tsvetkov), and TARCC postdoctoral research fellowship 2024-2026 (Vijay Kumar M J); COI: None

Poster # 66: THERAPEUTIC TARGETING OF ASTROCYTIC CHI3L1 WITH A BRAINPENETRANT BISPECIFIC ANTIBODY IN A MOUSE MODEL OF ALZHEIMER'S DISEASE

Presenting author: Qiang Wang, PhD (UTHealth Houston)

Background: Neuroinflammation plays a central role in Alzheimer's disease (AD) progression, contributing to synapse loss and cognitive impairment. Chitinase-3-like protein 1 (Chi3l1/YKL40), a highly conserved secreted glycoprotein, promotes tissue inflammation across multiple human diseases. Elevated Chi3l1 levels are consistently detected in AD biofluids and serve as biomarkers of ongoing inflammatory activity. Despite these associations, the therapeutic value of selectively targeting Chi3l1 in AD remains unknown. Method Using immunostaining, singlecell RNA sequencing, CSF analyses, and primary glial cultures, we profiled Chi3l1 expression in mouse brain at baseline and in the presence of AD-related pathology. We generated a Chi3l1blocking monoclonal antibody and engineered a transferrin receptor (TfR)-mediated, blood-brain barrier-penetrant bispecific. 5xFAD mice received peripheral (i.p.) administration of the bispecific; behavioral performance was assessed, and brains underwent histopathological evaluation. Result Astrocytes were the exclusive brain source of Chi3l1 mRNA and protein, and AD-related pathology increased Chi3l1 production in vitro and in vivo. A cross-reactive antiChi3l1 monoclonal antibody fused to an anti-TfR single-chain variable fragment (C59-αTfR) exhibited enhanced brain exposure. Treatment with C59-αTfR significantly restored exploratory activity in 5xFAD mice. Although effects on amyloid plaque burden were minimal, C59-αTfR robustly reduced plaque-associated gliosis, including microglial and astrocytic activation (notably DAM populations). In parallel, treatment lowered established neurodegeneration biomarkers, indicating attenuation of downstream neuronal injury. Conclusion Our findings reveal that Chi3l1 functions as a critical astrocyte-derived mediator driving neuroinflammation and neurodegeneration in AD. Therapeutic targeting of Chi3l1 with a selective brain-penetrant biologic achieves both functional and pathological improvements in a mouse AD model, supporting Chi3l1 as a promising target for precision immunotherapy in AD intervention.

Funding: This program and facilities are supported by NIH grants AG057587, AG074283, DK122708-03S1, BrightFocus ADR A20183775, and Brown Foundation 2020 Healthy Aging Initiative (W.C.).; COI: no

Poster # 67: WARBURG-LIKE GLYCOLYTIC REPROGRAMMING UNDERLIES THE PATHOLOGICAL EFFECTS OF TAU

Presenting author: Richa Gupta, PhD (UTHealth Houston)

Background: Degenerating neurons in Alzheimer's disease (AD) and tauopathies display bioenergetic failure caused by the accumulation of abnormal tau aggregates. Early changes in neuronal metabolism that lead to these persistent issues, occurring before cell degeneration, are not fully understood. To fill this critical knowledge gap, we examined bioenergetic changes in Drosophila glutamatergic neurons expressing the 0N3R isoform of human tau (tau0N3R). Methods Whole-brain RNA sequencing of larval brains was performed to assess gene expression changes. Protein analysis verified enzymatic levels. To directly evaluate bioenergetic function, we analyzed glutamatergic neurons expressing both tau0N3R and a sensor of cellular ATP/ADP ratio (PercevalHR), followed by metabolic assessment using Seahorse assay. The pathological nature of these metabolic changes was confirmed through RNAi-mediated knockdown of upregulated glycolytic genes. Results RNA sequencing revealed a significant increase in genes encoding sugar transporters and all glycolytic enzymes. In contrast, genes encoding proteins of the mitochondrial oxidative phosphorylation (OXPHOS) pathway mostly remained unchanged, supporting the idea of selective upregulation of glycolysis rather than a general increase in metabolic genes. Protein analysis confirmed higher levels of key glycolytic enzymes in tau0N3R brains. Bioenergetic analysis indicated a shift toward utilizing glycolysis instead of mitochondria for energy, with no change in overall ATP production or demand. This switch from OXPHOS to aerobic glycolysis suggests a neuronal Warburg-like metabolic reprogramming, supported by the glycolytic rate assay showing that tau expression in neurons causes a shift toward glycolysis for energy despite maintaining basal mitochondrial oxidative phosphorylation (OXPHOS)-this reliance on glycolysis results in neurons functioning in a metabolically compromised state. Additionally, RNAi-mediated knockdown of the upregulated glycolytic genes significantly reduced the lethality caused by tau0N3R expression in neurons. Conclusion These findings suggest that abnormal glycolytic reprogramming, similar to the Warburg effect, in neurons expressing tau0N3R contributes to neurodegeneration in a Drosophila model. Human tau expression induces a metabolic shift toward increased glycolysis for ATP production, despite preserved mitochondrial function. This metabolic change diminishes cellular energy reserves and actively promotes neurodegeneration, premature death, and accelerated biological aging.

Funding: R01AG069076, R01AG072176, R21AG087381; COI: No conflict of interest

Poster # 68: DUAL ROLE OF HELICASES DDX3X AND DDX3Y IN NEURONS

Presenting author: Rocio Diaz Escarcega, PhD (UTHealth Houston)

Background: Non-canonical base pairing between four guanines (G) within single-stranded Grich sequences leads to the formation of а G-quartet. Self-stacking of G-quartets results in the formation of a columnar four-stranded DNA structure known as the G-quadruplex (G4 or G4DNA). In cancer cells, G4-DNA regulates a broad variety of DNA-dependent processes, including transcription, replication, and telomere function. How G4s function in neurons is poorly understood. Methods We performed proteomics analysis to identify the protein networks modulated by a G4S ligand, quarfloxin (CX-3543), in primary cultured neurons. Quarfloxin promotes the stabilization of G4 structures, thus allowing us to gain insight into physiologically relevant G4-associated pathways. We performed a genome-wide gene expression analysis (RNAseq) to identify DDX3X- and DDX3Y-regulated genes in primary cultured male and female neurons Results We found that quarfloxin treatment induces DNA damage, with double-strand breaks enriched in the nucleolus. Proteomic analysis of treated neurons revealed widespread changes, impacting protein networks associated with Alzheimer's, Parkinson's and Huntington's diseases, and amyotrophic lateral sclerosis. The G4 helicase DDX3X, encoded on the X chromosome, was among the upregulated proteins, prompting further investigation of DDX3X and its Y-linked homolog DDX3Y in male and female neurons. RNA sequencing identified DDX3X- and DDX3Y-regulated genes involved in DNA damage responses, inflammation, cell cycle, and senescence, and revealed sex-specific expression profiles. We also discovered that in human brains, DDX3X levels were higher and more enriched in neuronal nuclei of older females compared to younger subjects, with a further increase observed in Alzheimer's disease. Conclusions Our study showed that DDX3X and DDX3Y may serve dual roles in neurons. When activated acutely, they promote beneficial DNA damage response pathways, but when chronically upregulated, they drive neurons toward senescent states. Taken together, our data suggest that DDX3X and DDX3Y may contribute to neuronal stress responses and neurodegeneration with sex differences. Our study reveals novel G4-dependent mechanisms and highlights the potential of DDX3X and DDX3Y as regulators of transcriptional responses in neurons, offering new molecular insights into neuronal sex-specific pathways and neurodegenerative diseases.

Funding: National Institute on Aging [1RF1AG068292] (A.S.T., S.P.M., A.U.), by the Glenn Foundation and the American Federation for Aging Research [AFAR BIG21042] (A.S.T), and by postdoctoral fellowship grant from Texas Alzheimer's Research and Care Consortium [TARCC grant ID: 1285130] (V. K. M. J.). The Biorepository of Neurological Disorders is supported by the Huffington Foundation (L.D.M.).; COI: None

Poster # 69: AXL BLOCKADE BLUNTS DISEASE PROGRESSION IN A MOUSE MODEL OF AD

Presenting author: Sanming Li, PhD (UTHealth Houston)

Background: Axl is a phagocytic receptor belonging to the TAM family of receptor tyrosine kinases. Unlike other TAM members, Axl expression on microglia is induced by inflammation in the brain. We previously reported that type I interferon (IFN-I) signaling upregulates Axl expression by plaque-associated microglia and promotes microglia-mediated synaptic loss under Alzheimer's disease (AD) conditions. However, the specific roles and therapeutic value of Axl receptor per se in AD pathogenesis remain not specified. Methodology: To reveal the molecular function of Axl in microglia, we performed RNA sequencing (RNA-seq) comparing primary microglia derived from Axl knockout (KO) and wild-type (WT) mice. Synaptosomes isolated from adult mouse brains were used in an uptake assay to assess microglial phagocytosis in vitro. To explore the therapeutic effects of Axl blockade, Bemcentinib (R428), a brain-penetrant selective Axl inhibitor, was administered orally to adult 5xFAD mice for 2 weeks before the animals were analyzed. Results: RNA-seq analysis revealed that Axl deletion downregulated genes associated with microglial proliferation and dysregulated genes related to the lysosomal pathway. Activation of Axl signaling by its ligand Growth arrest-specific 6 enhanced microglial phagocytosis of synaptosomes, whereas inhibition of this pathway-either pharmacologically or via Axl deletion-reduced microglial proliferation and synaptosome engulfment. Importantly, treatment with R428 in 5xFAD mice significantly alleviated synapse loss and neuronal dystrophy, accompanied by diminishment of IFN-responsive microglia population. Conclusions: Our findings demonstrate that Axl signaling promotes microglial proliferation and synaptic engulfment. Targeting Axl receptor may therefore represent a promising therapeutic strategy for mitigating microglia-mediated pathologies in Alzheimer's disease.

Funding: NIH grants AG057587, AG074283, DK122708-03S1, BrightFocus ADR A20183775, and Brown Foundation 2020 Healthy Aging Initiative (W.C.).; COI: No

Poster # 70: REBUILDING THE ALZHEIMER'S BRAIN FROM THE VASCULATURE

Presenting author: Sithara Thomas, PhD (UTHealth Houston)

Alzheimer's disease (AD) is characterized by profound cerebrovascular dysfunction, including alterations in vascular density, microstructure, integrity, and cerebral blood flow. Therapeutic angiogenesis may can enhance brain function and potentially ameliorate cognitive decline. Despite this, adult brain angiogenesis is not yet achieved and thus its effect in AD brain is unknown. Our lab demonstrated that histone deacetylase 2 (HDAC2) in brain endothelial cells (ECs) epigenetically represses pro-angiogenic genes during development, to ensure vascular maturation. Interestingly, EC Hdac2 deletion from the adult brain reactivated these genes and induced angiogenesis; furthermore, the angiogenic vessels promoted neurogenesis in the adult brain. Here, we hypothesized that EC Hdac2 deletion can induce angiogenesis in the 5xFAD mouse model of Alzheimer's disease, thereby enhancing Aβ clearance and promoting neurogenesis. Methods: Endothelial Hdac2 was deleted from one-month-old
5xFAD:Hdac2:Cdh5CreERT2 mice, and four months later, we evaluated the change in the vascular density using laminin and vascular proliferation by EdU staining. AD pathology was assessed by thioflavin S (ThS) staining and immunostaining with Aβ-6E10 and Aβ-4G8 antibodies. Neurogenesis was studied using EdU+/NeuN (newly generated neurons), EdU+/TBR1 (excitatory neurons) and EdU+/calbindin (interneurons) co-staining. Results: Hdac2 deleted 5xFAD (5xFAD Hdac2 ECKO) mice showed increased vessel density as measured by laminin staining. The presence EdU+ vessels further confirmed the increased angiogenesis. Quantification of Aβ load using ThS, Aβ-4G8, and Aβ-6E10 staining revealed a marked reduction in amyloid deposition in 5xFAD Hdac2 ECKO mice. Importantly, enhanced neurogenesis was evident from the increased population of EdU+/NeuN, EdU+/TBR1 and EdU+/calbindin cells (all p≤ 0.05). Discussion: Our findings reveal a transformative breakthrough: EC Hdac2 deletion triggers potent therapeutic angiogenesis in the AD brain. This robust neovascularization not only slashes Aβ plaque burden but robustly drives neurogenesis. Conclusion: Brain EC-specific Hdac2 is a promising therapeutic target for inducing adult cerebral angiogenesis, attenuating Alzheimer's disease pathology, and enhancing neurogenesis.

Funding: TARCC grant 2024-2026 ; COI: No

Poster # 71: RNA BINDING PROTEIN QUAKING IS A NOVEL REGULATOR OF MICROGLIAL FUNCTIONAL STATE IN ALZHEIMER'S DISEASE

Presenting author: Takese McKenzie, BS (UTHealth Houston)

Alzheimer's Disease (AD) is one of the most prevalent neurodegenerative diseases affecting elderly adults. The dysfunction of microglia, the resident phagocytes of the brain, is a defining hallmark of AD. As AD progresses, phagocytosis is gradually impaired, which promotes the accumulation of amyloid-beta plaques (Aβ) and neurodegeneration. However, the mechanism behind this impairment is not fully understood. We identified the RNA binding protein Quaking (Qki) as a key protein downregulated in microglia localized in dense Aβ regions in human AD brains. We also showed that Qki promotes microglial phagocytosis during demyelination. Currently, the role of microglial Qki in AD pathogenesis is unknown. Given the downregulation of microglial Qki in dense Aβ brain regions and Qki's regulation of phagocytosis, we hypothesize that Qki promotes microglial phagocytosis to attenuate AD progression. We tested this hypothesis by conditionally deleting Qki in microglia using the CX3CR1-CRE-ERT2 system. These mice were then crossed with the 5xFAD transgenic AD mouse model. Tamoxifen injections were given to mice on post-natal day 15 to induce microglial Qki deletion postdevelopment. Analysis of AD pathology was performed at 6.5 months old using immunofluorescence staining for Aβ burden, neuronal density, astrogliosis, and microgliosis. The hindlimb clasping test was performed to characterize neurodegeneration, and survival analysis was done to determine the effect of microglial Qki loss on the survival of 5xFAD mice. Our findings revealed that Qki loss increased Aβ burden in both the cortex and hippocampus of 5xFAD mice, indicating that microglial Qki facilitates Aβ clearance. These mice also displayed increased microgliosis, astrogliosis, and hindlimb clasping, accompanied by reduced survival and neuronal density. These findings suggest that microglial Qki loss accelerates AD progression, whereby its loss aggravated neurodegeneration, neuroinflammation, and survival. Future studies will focus on deciphering the effect of microglial Qki loss on cognitive function and the Qkispecific mechanism in regulating microglial phagocytosis. In summary, we have identified a novel role of Qki in promoting microglial phagocytosis in AD, thereby highlighting Qki upregulation as a potential therapeutic strategy and using Qki expression as a new marker of microglial phagocytic state in AD.

Funding: Pre-doc to Post-doc Transition Award National Institute On Aging of the National Institutes of Health (F99AG088439); COI: None reported.

Poster # 72: STRESS-INDUCED G-QUADRUPLEX HELICASE DHX36 REGULATES NEUROINFLAMMATION

Presenting author: Vijay Kumar M J, MSc, PhD (UTHealth Houston)

Background: G-quadruplexes (G4s) are non-canonical four-stranded DNA and RNA structures that regulate replication, transcription, and translation. Stabilization and accumulation of G4s have been linked to aging and neurodegeneration, where unresolved G4s can stall transcription and translation, leading to stress and genome instability. G4-resolving helicases such as DHX36 maintain G4 homeostasis by unwinding these structures, regulating G4-associated processes. During brain aging and chronic stress, impaired G4 resolution may disrupt gene expression programs that promote neuroinflammation, a key driver of neuronal dysfunction and degeneration. We hypothesize that DHX36 responds to stress and regulates G4-dependent transcriptional networks that contribute to neuronal inflammaging. Methods By using a G4DNA probe and mass spectrometry proteomic analysis, we identified that DHX36 is enriched in the aged brain. We analyzed DHX36 expression in mouse and human brains across age and Alzheimer's disease (AD) stages using immunohistochemistry. Cultured primary neurons were treated with the G4 stabilizer pyridostatin and stress-inducing agents, including tunicamycin, to identify stimuli regulating DHX36 expression. Lentiviral overexpression of DHX36 was performed to assess its molecular function, followed by RNA sequencing and pathway analysis to identify downstream targets linked to inflammation, aging, and chromatin regulation. Results DHX36 expression was significantly elevated in neurons of aged mouse and human brains, with further upregulation in AD, indicating a stress-responsive pattern. G4 stabilization by pyridostatin did not upregulate Dhx36 mRNA and protein, suggesting that G4 accumulation alone does not induce DHX36. In contrast, ER stress induced by tunicamycin upregulated DHX36 expression, implicating the unfolded protein response. In neurons in the aged mouse brain, DHX36 localized predominantly to the nucleus, consistent with a transcriptional regulatory function.
Transcriptomic profiling revealed that DHX36 repressed inflammatory and senescence-related pathways, including cytokines, interferon-stimulated genes, NF-κB regulators, and chromatin modifiers such as Hmga1 and Hmga2. Conclusion Our study identifies DHX36 as a stressinduced helicase that links ER stress signaling to the transcriptional repression of neuroinflammatory and senescence programs. By maintaining G4 homeostasis and suppressing maladaptive inflammatory responses, DHX36 emerges as a potential therapeutic target to mitigate neuronal inflammaging and age-associated neurodegeneration.

Funding: TARCC postdoctoral research fellowship 2024-2026 (Vijay Kumar M J), American Federation for Aging Research and Glenn Foundation for Medical Research Breakthroughs in Gerontology (BIG) Award, #BIG21042 (Andrey Tsvetkov); COI: None

Poster # 73: ROLE OF TYPE 2 INNATE LYMPHOID CELLS ON ALZHEIMER'S PATHOLOGY

Presenting author: Vijayasree V Giridharan, PhD (UTHealth Houston)

Background: Innate lymphoid cells type 2 (ILC2s) are pivotal immune regulators implicated in tissue homeostasis and repair, yet their roles in neurodegeneration and modulation of the gutbrain axis in Alzheimer's disease (AD) are largely uncharacterized. Elucidating mechanisms by which ILC2s influence amyloid pathology and gut microbial networks may uncover promising therapeutic strategies for AD. Methods: Male APP/PS1, beginning at 4 months of age, underwent weekly anti-CD90.2 monoclonal antibody administration until 6 months to induce ILC2 depletion. Cognition was evaluated via novel object recognition, while fecal samples were analyzed for microbiome composition, alpha and beta diversity, taxonomic shifts, and pathway abundance via functional metagenomics. Flow cytometry was utilized to isolate and sort ILC2s from the choroid plexus (CP), meninges, and lungs of wild-type mice. Conditioned media derived from each ILC2 population were applied to differentiated SH-SY5Y neurons challenged with Aβ fibrils to quantify neuroprotective effects. Results: In vivo, APP/PS1 mice subjected to ILC2 depletion exhibited pronounced memory deficits and reduced cortical levels of IL-4, IL-5, and IL13 relative to wild-type controls. Flow cytometric analysis revealed a marked reduction in ILC2s in the CP of APP/PS1 mice, suggesting an inherent ILC2 deficiency in the amyloidosis model. Microbiome analyses revealed substantial drops in alpha diversity (Observed OTUs) and pronounced beta diversity shifts, with ILC2-deficient mice exhibiting reduced proportions of short-chain fatty acid-producing taxa, most notably Lachnospiraceae and Clostridia, as well as decreased Firmicutes abundance. Taxonomic profiling further demonstrated altered relative abundance of key microbial families consistent with disruption of gut microbial structure. Conditioned media from CP- and lung-derived activated ILC2s significantly enhanced neuronal viability in Aβ-exposed differentiated SH-SY5Y neuronal cells, with robust secretion of IL-5 and IL-13 by activated ILC2s. Conclusion: These findings demonstrate that ILC2s confer neuroprotective actions against Aβ toxicity, and their deficiency modulates gut microbial diversity. ILC2 depletion accelerates cognitive decline, disrupts central cytokine balance, and reshapes the intestinal microbiome in APP/PS1 mice, underscoring the existence of a brain-gutimmune axis. Enhancing ILC2 function may offer a novel and integrative approach to curtailing AD progression by simultaneously targeting immune, neural, and microbial systems.

Funding: This work was supported by The University of Texas Health Science Center at Houston; Texas Alzheimer's Research and Care Consortium (TARCC) 1269191 to VVG, TB, and RM; and American Heart Association (AHA) 23SCISA1143588, 23TPA1140944, and 24SCEFIA1262305 to VVG. ; COI: None

Poster # 74: DISEASE-SPECIFIC TAU POLYMORPHS DEFINE UNIQUE PROTEIN INTERACTION NETWORKS ACROSS PROTEINOPATHIES

Presenting author: Ahmet Ege Balci (UTMB)

Tau protein aggregates exhibit distinct conformations across tauopathies, but their diseasespecific protein interactions remain poorly understood. Here, we demonstrate that disease-specific tau conformations determine unique protein interaction landscapes across Alzheimer's disease (AD), progressive supranuclear palsy (PSP), and dementia with Lewy bodies (DLB). Through comprehensive interactome profiling of misfolded tau aggregates from PBS- and sarkosyl-soluble fractions. We identified 493 high-confidence proteins with remarkable disease specificitynotably, no common interactors overlapping across all three tauopathies. Machine learning classification achieved compelling discrimination between diseases using as few as 4-6 proteins features, demonstrating robust molecular signatures underlying clinical heterogeneity. AD derived tau aggregates uniquely engaged cellular metabolism machinery, including key glycolytic enzymes and TCA cycle proteins, alongside glutamate/GABA neurotransmitter cycling components, with the astrocytic glutamate transporter SLC1A2 showing 27-fold enrichment over other tauopathies. In contrast, PSP tau displayed the most distinctive profile, with extensive protein depletion (52/57 significant proteins) and selective enrichment of proteasome components, particularly PSMB7 showing >3000-fold abundance. DLB tau is associated with neurogenesis modulators while depleting neuroinflammatory mediators. These interaction patterns were validated through proximity ligation assays and correlated with distinct posttranslational modification profiles, with PSP tau exhibiting globally elevated ubiquitination, AD showing mixed modification patterns, and DLB displaying minimal ubiquitination. Critically, sarkosyl-soluble fractions revealed reduced interactome complexity across diseases, except for PSP tau which maintained robust interactions with GPCR-ERK signaling and kinetochore proteins, suggesting unique aggregation mechanisms. Our findings establish that conformationally distinct tau strains dictate disease-specific protein interaction networks, providing molecular insight into tauopathy diversity and identifying novel therapeutic targets for precision medicine approaches in neurodegeneration.

Funding: This research work was supported by Mitchell Center for Neurodegenerative Diseases, the Gillson Longenbaugh Foundation and National Institute of Health grants: (R01AG077253 to RK and YZ, R01AG054025 and U24AG072458 to RK, R03AG088929-01 to MM).; COI: The authors declare no competing financial interests.

Poster # 75: THE EFFECTS OF THE RESTORATION AND ENHANCEMENT OF THE DNA DAMAGE REPAIR PATHWAY IN CELL MODELS

Presenting author: David Lynch, PhD (UTMB)

Background: The reduction of the DNA damage repair (DDR) pathway is associated with the development of Alzheimer's disease (AD). The DDR pathway is a complex system, but is an attractive target for the reduction of aging-related symptoms in AD. There are two ways that this could be carried out: firstly, the protein p53, a vital part of the DDR pathway, has been shown to be aggregated in AD patients, inhibiting its function. By restoring this function, the DDR may be restored to functionality. Secondly, it may be possible to enhance the DDR through the use of small molecules such as remodelin, a NAT10 inhibitor. Our hypothesis is that by combining both methods we will be able to improve the lifespan of cells that model AD. Methods: SH-SY5Y cells were used to model uptake of anti-p53 aggregate antibody into cells. Primary neurons were then used to determine the survival of the cells when exposed to both the antibody and remodelin. This model was then used to detemrine the effect that anti-p53 aggregate antibodies and remodelin have on the cell's transcriptome and proteome following DNA damage. Results: Our preliminary results have shown the uptake of antibody by the cells, indicating that the antibody will be able to remove the intracellular p53 aggregates that develop in AD. Moreover, the mouse primary neurons showed a high tolerance for both the antibody and remodelin individually and together. Fnally, transcriptomics and proteomics revealed that there are large changes in the expression of RNA and proteins following treatment with the antibody and the remodelin, including after UV exposure, suggesting that the treatments are able to modulate the response to DNA damage. Conclusions: In combination, the treatments with antibodies and remodelin are able to significantly alter cellular gene and protein expression, suggesting that they could be useful in restoring DDR pathway function in AD.

Funding: Funding for this project was received from the Pepper centre and the National Institute for Health. ; COI: None reported.

Poster # 76: HMGB1-MEDIATED ASTROCYTE SENESCENCE INDUCED BY DISEASE-SPECIFIC BDTO POLYMORPHS FROM AD, DLB, AND PSP

Presenting author: Fadhl Alshaebi, PhD (UTMB)

Background: Neurodegenerative diseases, including Alzheimer's disease (AD), dementia with Lewy bodies (DLB), and progressive supranuclear palsy (PSP), exhibit marked clinical and pathological heterogeneity. The molecular mechanisms underlying this variability remain poorly understood. Soluble brain-derived tau oligomer (BDTO) polymorphs differ in conformation and bioactivity across diseases and may drive disease-specific cellular responses. Astrocytes, which play critical roles in neuronal support and homeostasis, are increasingly recognized as active mediators of neuroinflammation through HMGB1 release and activation of the senescenceassociated secretory phenotype (SASP). We hypothesized that BDTO polymorphs from AD, DLB, and PSP induce HMGB1-mediated astrocyte senescence profiles contributing to diseasespecific neuroinflammatory signatures. Methods: Primary astrocytes were isolated from postnatal day 0-3 (P0-P3) wild-type C57BL/6 mice. Cells were treated with BDTO polymorphs isolated from AD, DLB, and PSP brain tissue; untreated astrocytes served as controls. Senescence was assessed by immunofluorescence staining for established markers, including HMGB1, p16, γH2AX, and senescence-associated β-galactosidase (SA-β-Gal) activity. To evaluate senescenceassociated secretory phenotype (SASP) activation, Western blotting and ELISA were performed to quantify secreted cytokines, specifically IL-1β, IL-6, and TNF-α. Planned RNA sequencing (RNA-seq) will define global transcriptional changes and identify HMGB1-dependent signaling pathways associated with BDTO-induced astrocyte senescence. Results: Exposure to BDTO polymorphs significantly induced astrocyte senescence, with distinct variations in the extent of the effect across groups. Astrocytes treated with AD-derived BDTO polymorphs exhibited the most pronounced senescent phenotype, characterized by marked increases in HMGB1 translocation, p16, and γ-H2AX expression, enhanced SA-β-Gal activity, and strong upregulation of SASP cytokines (IL-1β, IL-6, TNF-α). DLB- and PSP-derived BDTO polymorphs also induced astrocyte senescence, though to a slightly lesser degree than AD, indicating a lower but still significant pathogenic potential. These findings suggest differential HMGB1 activation and inflammatory potential among disease-specific BDTO polymorphs. Conclusion: These results demonstrate that BDTO polymorphs from AD, DLB, and PSP induce HMGB1-mediated astrocyte senescence phenotypes, which may represent a critical mechanism underlying the molecular and pathological heterogeneity of tauopathies. HMGB1 activation appears to link tau oligomer polymorphism to astrocyte dysfunction and neuroinflammatory signaling. Targeting HMGB1-dependent senescence pathways could provide a novel therapeutic strategy to attenuate neuroinflammation and modify disease progression.

Funding: NIH; COI: No Conflict of Interest

Poster # 77: MUSASHI-1 ISOFORMS IN AGED AND ALZHEIMER'S DISEASE BRAINS

Presenting author: Jack Tiller, MS (UTMB)

Background: Musashi-1 (MSI1) is an RNA-binding protein implicated in neurodegenerative processes, but the role of its isoforms in Alzheimer's disease (AD) remains unclear. Method: To explore this, we analyzed RT-qPCR sequencing data for MSI1 from human brain tissue (AD, early AD, and aged controls) and constructed a transcript map of all known MSI1 isoforms present in our samples. Result: Expression analysis showed that multiple MSI1 isoforms are expressed across groups, with no clear disease-specific pattern. Transcriptome-based mutation calling identified numerous nucleotide changes across MSI1 isoforms; however, we revealed that most mutations were shared across nearly all samples, and no unique mutations distinguished AD from controls. We also examined protein-level data from western blots of the Sarkosyl-soluble fractions, and consistent bands were observed at ~25, ~31, and ~50-56 kDa, while the expected ~39 kDa (monomeric MSI1) wildtype band was absent. PBS-soluble fractions occasionally showed a 39 kDa band. Banding patterns did not reliably distinguish AD from controls, suggesting that MSI1 alterations may reflect general features of brain aging rather than diseasespecific pathology. Conclusion: Together, these findings indicate that coding mutations are unlikely to drive MSI1 isoform differences in AD and that the observed bands on western blot may reflect noncanonical isoforms or aggregation states. Ongoing work includes expanding protein/transcript isoforms characterization across all sequenced samples and exploring additional pathways of interest, such as nuclear dysfunction linked to lamin proteins. This exploratory study establishes a framework for clarifying the potential role of specific MSI1 isoforms in AD pathology.

Funding: Alzheimer's Association (AARFD-NTF-22-968603, AARF-21-720991) NIH (AG054025, AG055771, 1R03AG088929-01); COI: None.

Poster # 78: DOES ZIKA INFECTION LEAD TO ALZHEIMER'S DISEASE? EXPLORING COMMON MOLECULAR PATHWAYS AND GENES

Presenting author: Kashish Khatkar, MS (UTMB)

Background: Recent evidence suggests that viral infections may contribute to the development of neurodegenerative diseases. Zika virus (ZIKV), a neurotropic flavivirus associated with microcephaly and neuronal apoptosis, induces neuroinflammation, oxidative stress, and endoplasmic reticulum (ER) stress, key processes also linked to the pathology of Alzheimer's disease (AD). This study investigates the potential connection between ZIKV infection and the onset or progression of AD, emphasizing shared molecular pathways and genetic factors. Methods: Human iPSC-derived neurons from healthy controls (CN) and AD patients were used to compare baseline gene expression differences. In parallel, CN cells were infected with ZIKV (MOI 1) for 48 hours, followed by total RNA extraction for transcriptomic profiling and small non-coding RNA (sncRNA) sequencing (s. Genes commonly affected by ZIKV and AD were identified and then projected to Gene Ontology (GO) enrichment analysis, which revealed them to be associated with signaling involved in neuroinflammation, oxidative stress, and cell death. Differential expression analysis was also done for sncRNA seq data and real-time PCR was performed to confirm the sequencing results. Results: Both ZIKV infection and AD shared molecular disease mechanisms involved in neuroinflammation and immune activation, oxidative stress with mitochondrial dysfunction, and autophagy and ER stress impairment. Key genes identified include MX1, an antiviral response gene indicating chronic interferon activation; CHAC1, a stress-inducible gene promoting glutathione degradation and neuronal apoptosis; RCAN1, a regulator of calcineurin signaling involved in synaptic dysfunction; and TRIM22, an innate immune modulator sustaining inflammation. ZIKV-affected tRNA-derived fragment tRF5Ala-AGC-4-1 negatively correlated with MX1 and CHAC1, suggesting that increased tRF expression may suppress antiviral and oxidative stress pathways, disrupt neuronal homeostasis and amplifying neurodegenerative signaling. Conclusion: ZIKV infection may initiate or accelerate Alzheimer-like pathology through shared molecular and genetic mechanisms. The convergence of immune activation, oxidative stress, and RNA regulatory pathways supports a potential virus-neurodegeneration link. Further gain/loss-of-function studies are needed to clarify the causal relationship between ZIKV neurotropism and AD progression.

Funding: This work was supported by grants from the U.S. National Institutes of Health (NIH) (R21 AI166543, R21 AG069226, R61 AG075725), the American Lung Association (ERP1252718), the Department of Defense (DoD) (MS240193), and the TARRC Investigator-Initiated Research Award to X.B.; COI: The authors declare no conflicts of interest relevant to the content of this study.

Poster # 79: EXPLORING THE ROLE OF NEURAL STEM CELL SMALL EXTRACELLULAR VESICLES IN MODULATING MICROGLIAL ACTIVATION IN ALZHEIMER'S DISEASE

Presenting author: Madison Khatkar, BS (UTMB)

Background: Adult neurogenesis, the generation of new neurons from neural stem cells (NSCs) in the hippocampus, is impaired in Alzheimer's disease (AD). We previously showed that small extracellular vesicles secreted by hippocampal NSCs (NSCsEv) provide synaptic protection against amyloid oligomers and significantly modulate microglial activation, which could underlie their protective effect against AD pathology. This study aims to investigate the mechanisms by which NSCsEv activates microglia. Methods Small extracellular vesicles were isolated by serial ultracentrifugation from the conditioned media of neural stem cells (NSC) differentiated from human induced pluripotent stem cells (NSCsEv) or mature neurons differentiated from NSCs (MNsEv) and labeled with PKH26 red fluorescent dye. Human microglia (HMC3) were cultured in a complete growth medium and treated with labeled small extracellular vesicles (NSCsEv; 3.6 x 10³ exosomes/μl) PBS, or LPS. Twenty-four hours and seventy-two hours after the NSCsEv and PBS treatments, the cells were washed and either fixed in 4% paraformaldehyde for immunofluorescence, lysed for western blot analysis or underwent RNA extraction for PCR. The culture medium for the LPS and 72-hour treatment group was changed every other day for 3 days. The cells were washed and fixed in 4% paraformaldehyde for immunofluorescence. Results Treatment of HMC3 human microglial cells with NSCsEv for 24 hours resulted in a marked upregulation of TREM2 protein levels compared to LPS and PBS, as determined by western blot. Western blotting also revealed an increase in CD68 protein levels in NSCsEvtreated cells compared to PBS and LPS controls, although this increase did not reach statistical significance. In parallel, qPCR analysis confirmed a significant upregulation of TREM2 in HMC3 cells following 24 hours of treatment with NSCsEv compared with PBS, supporting the proteinlevel findings. Conclusion: Treatment of HMC3 human microglial cells with neural stem cellderived small extracellular vesicles (NSCsEv) for 24 hours induces a distinct activation profile characterized by enhanced expression of TREM2. The consistent upregulation of TREM2 at both the mRNA and protein levels, together with a trend toward increased CD68 expression, suggests that NSCsEv modulates microglial activation, highlighting their potential role in supporting neuroprotective or regenerative processes within the central nervous system.

Funding: R01-AG069433 (GT, MAM); COI: None reported.

Poster # 80: IDENTIFICATION AND CHARACTERIZATION OF A TRNA-DERIVED FRAGMENT WITH AN IMPACT DURING NEURONAL INFECTION OF SARS-COV-2

Presenting author: Meagan Rippee-Brooks, MS (UTMB)

Background: SARS-CoV-2, with its neurotropism, has been of interest since its emergence because of associated neurological disease sequelae, which include brain fog and cognitive impairment. Increasing evidence suggests there is a close relationship between SARS-CoV-2 infection and altered accumulation of a specific family of endogenous small non-coding RNAs (sncRNAs) known as tRNA-derived RNA fragments (tRFs). Previously, our lab has published several studies highlighting the presence and regulatory role tRFs during neurodegenerative disorders (Alzheimer's Disease (AD) and during SARS-CoV-2 infections in airway epithelial cells. We investigated the role of enhanced tRF5-GlnCTG during SARS-CoV-2 infection within iPSC-neurons to model its regulatory role amongst AD-related gene targets. Method: We infected iPSC-neurons with the omicron variant of SARS-CoV-2 to identify impacted tRFs. We then pre-treated iPSC-neurons with antisense oligos (ASOs) to tRF5-GlnCTG and infected these cells. qRT-PCR determined the impact this tRF5 has upon SARS-CoV-2 viral replication, and the computationally predicted gene targets (n=11) identified to interact with tRF5-GlnCTG, with a hybridization free energy (ΔG) less than or equal to -30 kJ/mol, shared commonly with iPSCneurons infected with omicron. Result: T4PNK-RNA-Seq revealed that tRF5-GlnCTG is a significant sncRNA signature of omicron-infected iPSC-neurons. We found that ASO treatment targeting tRF5-GlnCTG suppresses viral replication. To understand the associated molecular consequence of enhanced omicron replication by tRF5-GlnCTG in neurons, RNA hybrid energybased prediction and the omicron variant changed mRNA were performed. qRT-PCR revealed that CACNA2D2, one of 11 predicted targets, was suppressed by omicron-induced tRF5GlnCTG. Given the reported role of CACNA2D2 in central nervous system (CNS) function, including its involvement in AD-related signaling, we consider that omicron may affect neuronal function via tRF5-GlnCTG. Conclusion: This study highlights the implication of sncRNAs as regulators of molecular signatures during neurotropic viral infections. Our validation of computationally predicted gene target CACNA2D2 supports the importance of SARS-CoV-2induced tRF5-GlnCTG in CNS function, which has since become of interest to the scientific community in relation to neurodegenerative phenotypes. Here, we show that viral infections, particularly neurotropic viruses such as SARS-CoV-2, have the propensity to act as molecular switches upon infection, accelerating associated neurological disease.

Funding: This work was supported by grants from the US National Institutes of Health (NIH) R21 AI166543, R21 AG069226, R61 AG075725, ERP-1252718 from the American Lung Association, MS240193 from the Department of Defense (DoD), and TARRC InvestigatorInitiated Research Award to X.B.; COI: The authors have no conflicts of interest to declare that are relevant to the content of this study.

Poster # 81: SEX DIFFERENCES IN TAU PATHOGENESIS: X-LINKED UBL4A DYSREGULATION IN ALZHEIMER'S DISEASE

Presenting author: Nicha Puangmalai, PhD (UTMB)
Tau aggregation is one of the major hallmarks of Alzheimer's disease (AD). Sex differences also play an important role in AD pathology; however, the influence of sex differences on tau pathology in AD is still not fully understood. X-chromosome-wide association studies (XWAS) identified Ubl4a as a locus linked to neurofibrillary tangle formation specifically in male AD patients. Ubl4a forms the BAG6 complex with BAG6 and GET4, functioning in protein quality control and ubiquitin-dependent degradation. We hypothesize that Ubl4a and BAG6 complex dysregulation underlies sex-differential tau ubiquitination and aggregation in AD. Western blot analysis of postmortem AD and age-matched control brain samples revealed pronounced reductions in Ubl4a, BAG6, and GET4 in AD cases, strongly correlating with increased hyperphosphorylated tau levels. Since ubiquitination is fundamental to tau turnover, reduced Ubl4a activity may impair tau processing and contribute to NFT accumulation in males. Immunofluorescence analysis in hTau mice expressing six human tau isoforms demonstrated agedependent BAG6 complex alterations paralleling progressive tau pathology. Tau biosensor cell assays showed that BAG6 complex knockdown heightened AD brain-derived oligomeric tau seeding activity, while Ubl4a knockdown produced synergistic effects, further intensifying tau aggregation. Mechanistically, Ubl4a facilitates tau ubiquitination via BAG6-GET4 complex engagement; Ubl4a reduction destabilizes this machinery, impairs tau recognition and clearance, and exacerbates disease progression. These findings link X-linked Ubl4a and the BAG6-GET4 complex to sex-specific tau ubiquitination and propagation in AD, identifying the Ubl4a-BAG6GET4 axis as a promising precision therapeutic target for addressing sex-dependent vulnerabilities in AD.

Funding: This research work was supported by Mitchell Center for Neurodegenerative Diseases, the Gillson Longenbaugh Foundation and National Institute of Health grants: (R01AG077253 to RK and YZ, R01AG054025 and U24AG072458 to RK). ; COI: no conflicts of interest.

Poster # 82: PATHOLOGICAL EFFECTS OF AMYLOID POLYMORPHS ON MICROGLIA DERIVED FROM APOLIPOPROTEIN E3/3 AND E4/4 INDUCED PLURIPOTENT STEM CELLS

Presenting author: Nopparat Suthprasertporn, PhD (UTMB)

Alzheimer's disease (AD) is driven by pathological tau aggregation leading to neurofibrillary tangle formation, alongside the accumulation of amyloid-β plaques, with emerging evidence implicating toxic alpha-synuclein species in disease progression. The apolipoprotein E4 (ApoE4) allele is the strongest genetic risk factor for late-onset AD and is linked to impaired lipid metabolism and altered microglial function, yet how ApoE4 influences microglial responses to tau and alpha-synuclein remains unclear. Using induced pluripotent stem cell (iPSC)-derived microglia from ApoE3/3 and ApoE4/4 backgrounds, we quantified lipid droplet dynamics using flow cytometry after exposure to recombinant tau oligomers, brain-derived tau oligomers, and alpha-synuclein oligomers. ApoE4 microglia exhibited a greater proportion of high lipid droplet cells compared to ApoE3, especially following tau treatment, indicating enhanced lipid metabolic dysregulation. These findings reveal that ApoE4 drives microglial lipid droplet accumulation in response to amyloid polymorphs implicated in AD. Transcriptomic and proteomic approaches will elucidate how genetic background broadly shapes microglial responses to amyloid polymorphs, providing deeper insights into mechanisms behind increased disease susceptibility in ApoE4 carriers.

Funding: This research work was supported by Mitchell Center for Neurodegenerative Diseases, the Gillson Longenbaugh Foundation and National Institute of Health grants: (R01AG077253 to RK and YZ, R01AG054025 and U24AG072458 to RK); COI: I have no conflict of interest.

Poster # 83: PATHOLOGICAL EFFECTS OF AMYLOID POLYMORPHS ON MICROGLIA DERIVED FROM APOLIPOPROTEIN E3/3 AND E4/4 INDUCED PLURIPOTENT STEM CELLS

Presenting author: Shrinath Kadamangudi, MPhil (UTMB)

Background: Synaptic dysfunction is a hallmark of tauopathies and a key driver of cognitive decline. Tau pathology progression correlates with cognitive deterioration, underscoring the therapeutic potential of targeting tau at the synapse. However, cellular and regional susceptibilities to tau pathology complicate therapeutic efforts, and the mechanisms underlying synaptic vulnerability in human tauopathies remain poorly understood. Methods: Post-mortem brain tissue from control and primary age-related tauopathy (PART) cases was used to investigate synaptic vulnerability to soluble tau oligomers (tauO), highly toxic pathological conformers. Definite PART samples, characterized by selective tau accumulation without co-pathologies (e.g., Aβ, α-syn, TDP-43), were studied. Synaptosomes from hippocampus and temporal cortex were: (1) characterized for tau pathology (p-tau, tauO, total tau) via western blotting, (2) treated with recombinant tau oligomers (r-tauO) and analyzed by flow cytometry (FC) immunophenotyping to examine pre-/post-synaptic and GABAergic/glutamatergic tauO interactions, (3) microtransplanted into frog oocytes for two-electrode voltage clamp (TEVC) recordings of GABA and AMPA currents, and (4) examined via EM immunogold to confirm tauO-synapse interactions. Additionally, brain-derived tau oligomers (BDTO) from PBS-soluble PART hippocampal lysates were isolated via co-immunoprecipitation and analyzed by LC-MS/MS to identify synaptic BDTO interactomes. Results: Outcome-1 | In control synaptosomes treated with r-tauO, FC and EM revealed that tauO preferentially engage pre-synaptic terminals and associate with synaptic vesicles. TauO showed higher affinity for GABAergic synapses (FC) and selectively potentiated GABAergic currents (TEVC). Outcome-2 | In PART synaptosomes, tau aggregates (tauO and PHF-tau) were significantly elevated in the hippocampus compared to the temporal cortex (western blot), correlating with a decreased excitatory/inhibitory (E/I) ratio (TEVC), suggesting a pro-inhibitory shift in regions with elevated tau pathology. Outcome-3 | PART BDTO interactomes revealed significant enrichment of pre-synaptic proteins involved in synaptic vesicle cycling, aligning with Outcome-1. Conclusions: This study provides the first direct evidence of selective synaptic vulnerability to tauO in human brain tissue, revealing a preference for pre-synaptic and inhibitory synapses. These findings challenge post-synapticfocused therapeutic strategies, such as targeting NMDA and acetylcholine pathways. By identifying pre-synaptic vesicle cycling as a tauO target, our study underscores the need for tau therapeutics tailored to specific synaptic populations and circuits.

Funding: This work was supported by the UTMB Kempner Fellowship and NIA/NIH grants F30AG085974 (SK) and R01AG060718 (GT).; COI: None.

Poster # 84: PHOSPHOLIPASE D1 ATTENUATION MITIGATES SYNAPTIC AND COGNITIVE DYSFUNCTION IN ALZHEIMER'S DISEASE AND RELATED DEMENTIA BY USING AN ADENO-ASSOCIATED VIRAL APPROACH

Presenting author: Sravan Gopalkrishna Shetty Sreenivasa Murthy, PhD (UTMB)

Background: Synaptic dysfunction is among the earliest pathological events in Alzheimer's Disease (AD), preceding neuronal loss and closely correlating with cognitive decline. Identifying molecular drivers of synaptic impairment is critical for developing effective therapies. Our lab has observed a pathological increase in phospholipase D isoform 1 (PLD1) within the synaptosomal fraction of AD mouse models and postmortem human brain samples. Earlier studies have shown that amyloid-β and tau deposits elevate PLD1 expression in wild-type and 6month-old 3xTg-AD mice. This study investigates the role of PLD1 in synaptic dysregulation and cognitive outcomes in aged AD models. Methods: Twelve-month-old 3xTg-AD mice received intracerebroventricular injections of adeno-associated viral vector serotype 2 (AAV2) to induce PLD1 overexpression (PLD1 OXP) or attenuation (PLD1 ATT). Behavioral assays assessing learning and memory included the Y-maze, open field test (OFT), novel object recognition (NOR), and elevated plus maze (EPM). Immunofluorescence was used to examine PLD1 interaction with pathological tau. Synaptic function was evaluated via high-frequency stimulation long-term potentiation (HFS-LTP) using conventional electrophysiology and multi-electrode array (MEA) recordings. Result: AAV2-mediated modulation of PLD1 revealed distinct outcomes in aged 3xTg-AD mice. PLD1 attenuation (PLD1 ATT) significantly improved cognitive performance. Furthermore, PLD1 modulation altered synaptic potentiation profiles, particularly in the Schaffer collateral-CA1 hippocampal pathway, a region highly susceptible to AD pathology. Conclusion: PLD1 plays a critical role in maintaining synaptic integrity during healthy aging. PLD1 dysregulation-driven by AD-related neuropathological proteinopathies, contributed to synaptic vulnerability and cognitive decline. Targeting PLD1 may offer a promising therapeutic strategy for mitigating synaptic dysfunction in Alzheimer's Disease and related dementia.

Funding: Jeanne B. Kempner Scholarship (2022-2023, C.N.), AARG-17-533363 (B.K.), NIA R21 - AG059223 (B.K.), NIA R01 - AG063945 (B.K.), The Don and Nancy Mafrige Professor in Neurodegenerative Disease Endowment (B.K.), Mitchell Center for Neurodegenerative Diseases ; COI: No conflict of interest

Poster # 85: PROGRESSIVE TAU CO-PATHOLOGY IN A HUMAN Α-SYNUCLEIN MOUSE MODEL OF SYNUCLEINOPATHY AND ITS REVERSAL BY TAU OLIGOMERSELECTIVE IMMUNOTHERAPY

Presenting author: Sudipta Senapati, PhD (UTMB)

Background: Pathological aggregation of α-synuclein (α-Syn) is a defining hallmark of Parkinson's disease (PD); however, increasing evidence implicates tau co-pathology as a critical contributor to disease progression and neurodegeneration. The mechanistic interplay between αSyn and tau remains incompletely understood, particularly regarding how tau pathology evolves over time and influences α-Syn-mediated toxicity. Methods: We utilized the human wild-type α-synuclein overexpressing mouse model (hSyn) to examine the temporal progression of behavioral impairments and associated pathological changes. Motor and cognitive performances were longitudinally assessed, and brain tissues were analyzed for α-Syn and tau aggregation, hyperphosphorylation, and synaptic integrity. To determine the therapeutic potential of tautargeted immunotherapy, hSyn mice received TTCM2-a conformation-selective IgG2b antibody that recognizes discontinuous β-strand-rich epitopes within the tau microtubule-binding region, exhibiting high specificity for aggregation-prone tau oligomers. Results: Motor deficits emerged around 9 months of age and progressively worsened, followed by cognitive impairments at later stages. These behavioral declines correlated with marked accumulation of α-Syn aggregates and hyperphosphorylated tau species in the cortex, midbrain, and hippocampus. Treatment with TTCM2 significantly reduced pathogenic tau oligomers, attenuated α-Syn aggregation, and restored synaptic marker expression. These molecular improvements were accompanied by robust recovery of motor coordination and cognitive performance. Conclusion: Our findings identify tau oligomers as key amplifiers of α-synuclein-induced neurotoxicity, driving both pathological propagation and behavioral dysfunction. Selective targeting of tau oligomers with TTCM2 effectively disrupted the synergistic tau-α-Syn interplay, resulting in substantial functional and molecular rescue. These results support tau oligomer-selective immunotherapy as a promising disease-modifying strategy for PD and related synucleinopathies characterized by tau co-pathology.

Funding: This research work was supported by Mitchell Center for Neurodegenerative Diseases, the Gillson Longenbaugh Foundation and National Institute of Health grants: (R01AG077253 to RK , R01AG054025 and U24AG072458 to RK).; COI: The authors declare no conflict of interest.

Poster # 86: NON-TOXIC TAU OLIGOMER POLYMORPHS ISOLATED FROM NON-DEMENTED INDIVIDUALS WITH ALZHEIMER'S NEUROPATHOLOGY

Presenting author: Thai T. Nguyen, BS (UTMB)

Background: Alzheimer's disease (AD) is a neurodegenerative disorder that is characterized by the presence of extracellular Aβ plaques and intracellular tau neurofibrillary tangles. A subset of individuals classified as A+T+N-, termed Non-Demented with Alzheimer's Neuropathology (NDAN), maintain normal cognition despite harboring full AD pathology. This suggests that tau oligomers (TauO) may exist as distinct polymorphs with divergent functional consequences as well as offer innate protective mechanisms opposing the events that normally lead to cognitive impairment. Methods: We investigated whether NDAN brain-derived tau oligomers (NDANBDTOs) differ from AD-BDTOs in toxicity and microglial responses. Wild-type mice received intracerebroventricular injections of AD- or NDAN-BDTOs and molecular outcomes were analyzed at 24h, 48h. Brain tissue was analyzed by immunofluorescence and western blotting to characterize microglial phenotypes (Iba1, CD68, TREM2) and assess pathways associated with phagocytosis, autophagy (Atg5, Atg7, LC3A/B, p62), and oxidative stress (SOD1/2, catalase, 8oxo-dG). Results: AD- and NDAN-BDTOs were identified as structurally distinct polymorphs with divergent biological effects. AD-BDTOs induced neuronal apoptosis and proinflammatory microglial activation, while NDAN-BDTOs enhanced autophagy, preserved antioxidant defenses, and reduced oxidative DNA damage. Furthermore, NDAN-BDTOs elicited a specialized microglial phenotype distinct from that induced by AD-BDTOs, suggesting differential engagement of neuroinflammatory pathways. Conclusion: Our findings support the existence of a non-toxic, propagating tau oligomer in NDAN brains that promotes resilience by avoiding neurodegeneration. These results highlight the therapeutic potential of strategies that stabilize or redirect tau oligomers into non-toxic conformations, offering a novel avenue for disease-modifying interventions in AD.

Funding: UT-MSTAR (NIH 1T35AG092747-01) R21AG089708 to GT R01AG073133 to GT R21AG082230 to AF AARF22973974 to AF (Alzheimer's Association Research Fellowship) ; COI: No conflicts of interest.

Poster # 87: REELIN DYSREGULATION AND COGNITIVE IMPAIRMENT IN DP16 MICE: INSIGHTS INTO NEUROINFLAMMATION IN DOWN SYNDROME

Presenting author: Costanza Calamai, PhD (UTSW)

Background: The immune system is essential for tissue integrity, but its excessive activation can lead to disease. In the central nervous system, neuroinflammation plays a key role in the progression of neurodegenerative disorders. Reelin, an extracellular glycoprotein involved in neuronal development, synaptic function, and immune regulation, has been associated with increased neuroinflammation in neurodegenerative diseases such as Alzheimer's. Down syndrome (DS) is characterized by altered neuronal development and chronic neuroinflammation, contributing to cognitive decline and neurodegeneration. Understanding the molecular mechanisms driving neuroinflammation in DS is crucial for identifying potential biomarkers and therapeutic targets. Method. Dp(16)1Yey mice (hereafter referred to as DP16 mice) were used to investigate cognitive function. Spontaneous alternation was assessed using the Y-maze test, while associative learning and contextual memory were evaluated with the Contextual Fear Conditioning test. Reelin levels were measured by Western blot analysis on plasma samples obtained from both wild-type (WT) and DP16 mice, as well as from human plasma samples collected from individuals with and without DS across different age groups. Intravital microscopy (IVM) experiments were also performed in the brains of DP16 and WT mice using two-photon microscopy to assess leukocyte rolling. Result. Behavioral tests revealed reduced spontaneous alternation, associative learning, and contextual memory in DP16 mice compared to WT controls. Regarding Reelin, plasma levels were found to be increased in DP16 mice compared to WT, in both males and females. Similarly, in human samples, Reelin levels appeared higher in individuals with Down syndrome compared to euploid controls, reaching statistical significance in the subgroup of patients aged 43-56 years. Based on these findings, we also expect to observe an increased number of rolling leukocytes in IVM experiments performed in the brains of DP16 mice compared to WT controls. Conclusion. These findings indicate that DP16 mice display cognitive impairment, and that elevated Reelin levels may contribute to this phenotype, suggesting a potential link between Reelin dysregulation and cognitive decline in Down syndrome.

Funding: C. Calamai was supported by a postdoctoral fellowship grant from TARCC; COI: L. Calvier, M.Z. Kounnas, and J. Herz are shareholders of Reelin Therapeutics, Inc. L.C. and J.H. are co-inventors of a patent related to anti-Reelin strategies (application number 15/763,047 and publication number 20180273637).

Poster # 88: INVESTIGATING ANTI-REELIN THERAPY ACROSS ALZHEIMER'S DISEASE MOUSE MODELS

Presenting author: Miranda E. Leal-Garcia, PhD (UTSW)

Background: Alzheimer's disease (AD) is characterized by cognitive decline, amyloid beta plaques and neurofibrillary tau tangles. Despite efforts to target these proteins, current therapies have demonstrated limited success in preventing cognitive decline. In AD, blood brain barrier dysfunction permits immune cells infiltration which exacerbates inflammation, secondary tissue damage and the amyloid and tau pathology. We previously found that Reelin, an extracellular glycoprotein, has been shown to promote endothelial activation and inflammation in chronic inflammatory diseases such as atherosclerosis and multiple sclerosis. However, the role of circulating Reelin has yet to be investigated in AD. Method: To investigate the role of circulating Reelin in AD pathology, we utilized three transgenic mouse models, 5xFAD and APPNL-F which exhibit amyloid plaque formation and PS19 which develop tau pathology. Wildtype (WT) littermates were used as controls. To reduce circulating Reelin, male and female mice received intraperitoneal injections of either control IgG or anti-Reelin antibody (CR-50) twice weekly. To investigate memory and cognitive function, spontaneous alternative behavior (SAB) was assessed using an open Y-maze. To measure circulating Reelin, a western blot was performed on the plasma. To measure total and phosphorylated tau, a western blot was performed on cortex protein lysate. Results: Circulating Reelin in the blood of 5xFAD and APPNL-F mice was elevated throughout the progression of the disease. Following CR-50 treatment, 5xFAD and APPNL-F mice had decreased circulating Reelin levels. 5xFAD and APPNL-F exhibited improved cognition when compared to their IgG counterparts. We observed a reduction in phosphorylated Tau in our 5xFAD CR-50 treated group. However, there were no observable differences amongst our APPNL-F groups. PS19 mice have elevated Reelin when compared to WT controls but is reduced following CR-50 treatment. CR-50 treatment appeared to have no impact on Tau phosphorylation in PS19 mice. Conclusion: Our data shows that depleting Reelin improves cognition in 5xFAD and APPNL-F mice. Overall, our data suggests that Reelin may play a role in disease progression. However, additional experiments are necessary to evaluate the inflammatory response and the role of Reelin in AD pathology in PS19 mice

Funding: J. Herz was supported by grants from the NHLBI (R37 HL063762), NIA (RF1/RF3 AG053391), the NINDS and NIA (R01 NS093382 and R01 NS108115), BrightFocus (A2013524S & A2016396S), the Bluefield Project to Cure FTD, Harrington Discovery Institute (HDI2019-SI-4479), Circle of Friends Pilot Synergy Grant, Alzheimer's Association (ABA-22970304), the Presbyterian Village North Foundation, and the Robert J. Kleberg Jr. & Helen C. Kleberg Foundation. M.Z. Kounnas was supported by a NIH/NIAMS SBIR to Reelin Therapeutics, Inc. (1R43AR081762) J. Herz and L. Calvier (Co-Investigators Subcontract), and NIH/NIA SBIR grant to Reelin Therapeutics, Inc. (1R43AG084450) with L. Calvier and J. Herz (PIs Subcontract).; COI: L. Calvier, M.Z. Kounnas, and J. Herz are shareholders of Reelin Therapeutics, Inc. L.C. and J.H. are co-inventors of a patent related to anti-Reelin strategies (application number 15/763,047 and publication number 20180273637).

Poster # 89: PERIPHERAL ANTI-REELIN THERAPY RESTORES NEUROVASCULAR INTEGRITY AND PROMOTES MICROGLIAL PLAQUE CLEARANCE IN ALZHEIMER'S DISEASE

Presenting author: Sharon Wu, BS (UTSW)

In Alzheimer's disease (AD), beyond the hallmark amyloid-β plaques and tau tangles, increasing evidence highlights the critical role of chronic neuroinflammation in disease progression. This inflammation is driven in part by circulating leukocytes that infiltrate the brain and release cytokines and chemokines. In our lab, we have identified Reelin, a secreted endothelial factor present in the circulation, as a key mediator that exacerbates neuroinflammation by promoting leukocyte transmigration across the compromised blood-brain barrier (BBB). Our central hypothesis posits that peripheral inhibition of Reelin-without the need to traverse the BBB-can effectively attenuate neuroinflammation and restore cognitive function by preserving BBB endothelial integrity. To test this hypothesis, we administered the anti-Reelin antibody CR-50 for one month in the APPNL-F mouse model of late-stage Alzheimer's disease. Our studies demonstrate that peripheral anti-Reelin CR-50 treatment not only reduces plasma Reelin levels but also repairs BBB integrity, thereby limiting leukocyte infiltration and reestablishing a protective neurovascular environment. In addition, anti-Reelin therapy effectively reduces amyloid plaque load in the brain, which we believe is driven by activated microglial phagocytosis. Biochemical and histological analyses further support these findings, revealing decreased Reelin accumulation in plasma and improved BBB tight junction integrity. This restoration of BBB function, combined with enhanced amyloid clearance, is directly associated with improved cognitive outcomes in treated AD mice. Our results underscore the mechanistic link between peripheral Reelin inhibition and central neuroinflammatory resolution. By targeting Reelin in the circulation, this therapeutic approach bypasses the limitations of BBB permeability while simultaneously restoring vascular and immune homeostasis. Collectively, these findings highlight peripheral anti-Reelin therapy as a promising and translationally relevant strategy to mitigate neuroinflammation, promote amyloid clearance, and improve cognitive function in Alzheimer's disease.

Funding: L. Calvier was supported by a postdoctoral fellowship grant from the DFG (CA 1303/11) and by the Texas Alzheimer's Research and Care Consortium (TARCC). J. Herz was supported by grants from the NHLBI (R37 HL063762), NIA (RF1/RF3 AG053391), the NINDS and NIA (R01 NS093382 and R01 NS108115), BrightFocus (A2013524S & A2016396S), the Bluefield Project to Cure FTD, Harrington Discovery Institute (HDI2019-SI-4479), Circle of Friends Pilot Synergy Grant, Alzheimer's Association (ABA-22-970304), the Presbyterian Village North Foundation, and the Robert J. Kleberg Jr. & Helen C. Kleberg Foundation. M.Z. Kounnas was supported by a NIH/NIAMS SBIR to Reelin Therapeutics, Inc. (1R43AR081762) J. Herz and L. Calvier (Co-Investigators Subcontract), and NIH/NIA SBIR grant to Reelin Therapeutics, Inc. (1R43AG084450) with L. Calvier and J. Herz (PIs Subcontract).; COI: L. Calvier, M.Z. Kounnas, and J. Herz are shareholders of Reelin Therapeutics, Inc. L.C. and J.H. are co-inventors of a patent related to anti-Reelin strategies (application number 15/763,047 and publication number 20180273637).

Poster # 90: SKULL BONE MARROW NICHES MEDIATE CNS-PERIPHERAL IMMUNE CROSSTALK IN ALZHEIMER'S DISEASE

Presenting author: Thandiwe Chavula, MBBS, PhD (UTSW)

Background: Amyloid beta (Aβ) plaque formation is a hallmark of Alzheimer's disease (AD), yet Aβ-targeted therapies have limited clinical benefit, highlighting the need for new therapeutic strategies. Neuroinflammation has emerged as a third core pathology in AD, driven by resident microglia and infiltrating immune cells. However, mechanisms for recruitment of peripheral immune cells remain poorly understood. Recent evidence identifies the skull bone marrow as a unique immune niche directly communicating between the central nervous system (CNS) and peripheral immune tissue, though its role in AD is unclear. Our preliminary data suggest that Reelin, a proinflammatory endothelial factor, is upregulated in AD and may serve as a marker of skull bone marrow activation. Here, we investigated whether skull bone marrow activation represents a novel pathway linking CNS and peripheral inflammation in AD. Methods: Plasma reelin was analyzed in AD patients and mouse models (5xFAD and APPNL-F) via western blotting. To assess skull bone marrow activation, WT mice were injected with PBS or Aβ1-42 fibrils and Reelin expression quantified by PCR. Effects of Reelin inhibition with CR-50 antibody on skull bone marrow niche size and volume were examined in 5xFAD and APPNL-F mice. Longitudinal 9.4T T₂-weighted MRI with gadolinium contrast was used to validate in vivo skull bone marrow changes following CR-50 treatment. Results: Plasma Reelin levels increased with disease severity in both AD patients and mouse models. Reelin mRNA was upregulated two-fold in skull bone marrow, decreased in femur, and unchanged in liver, correlating with plasma Reelin levels. Aβ1-42 injection elevated plasma Reelin and selectively increased skull bone marrow mRNA expression. Strikingly, 5xFAD mice exhibited enlarged skull bone marrow niches, with area and intensity correlating with Reelin mRNA levels. CR-50 treatment significantly reduced niche area and integrated density in both 5xFAD and APPNL-F mice. Finally, MRI confirmed reduced skull bone marrow pocket volume one-month post-treatment in APPNL-F mice. Conclusion: These findings identify skull bone marrow niches as mediators of CNS-peripheral immune crosstalk in AD. Reelin reflects this activation, and its inhibition mitigates skull bone marrow expansion, supporting Reelin as a potential target in AD-related neuroinflammation.

Funding: TARCC DFG NIA NINDS BrightFocus; COI: Laurent Calvier and Joachim Herz are shareholders of Reelin Therapeutics, Inc. Laurent Calvier and Joachim Herz are co-inventors of a patent related to anti-Reelin strategies (application number 15/763,047 and publication number 20180273637).

Poster # 91: EPILEPTIFORM SPIKES SHIFT FROM REM TO NREM SLEEP ACROSS AGE AND DISEASE PROGRESSION IN A MOUSE MODEL OF ALZHEIMER'S NEUROPATHOLOGY

Presenting author: Manuel Silva-Pérez, PhD (BCM)

Background: Individuals with Alzheimer's disease (AD) have an increased incidence of epileptiform activity, which may contribute to and accelerate cognitive decline, and occur primarily during NREM sleep. Several transgenic mouse lines used to study AD-related neuropathology also exhibit epileptiform activity predominantly during sleep. However, these mice tend to be studied quite early in disease progression and exhibit epileptiform spikes primarily during REM rather than NREM sleep. In APP mice from line Tg2576, epileptiform spikes first emerge in REM sleep at very young ages and then propagate to NREM sleep at later timepoints, suggesting that the distribution of epileptiform activity across the sleep-wake states may shift with age and disease progression. Method: We performed continuous video-EEG recordings in female and male APP mice from line J20 at 2, 13, 20, and 30 months of age. We quantified epileptiform spikes, time spent in wakefulness, NREM sleep, and REM sleep, and sleep fragmentation, a prominent feature of AD. Result: In young mice in early stages of disease, epileptiform spikes were abundant during REM sleep, less frequent during NREM sleep, and almost absent during wakefulness. Over half of the total daily epileptiform spikes occurred during REM sleep. The proportion of epileptiform spikes during NREM sleep increased at later timepoints, comprising ~80% of total epileptiform spikes at 20 and 30 months of age. This shift was due to gradual age-related changes in spike rates, which decreased during REM and increased during NREM. Sleep also became increasingly fragmented. Conclusion: Epileptiform spikes shift from occurring primarily during REM to occurring predominantly during NREM as APP mice age and disease progresses. The prevalence of epileptiform spikes in NREM sleep in aged APP mice better reflects what is observed in individuals with AD. Therefore, identifying the mechanisms that give rise to epileptiform spikes in APP mice, and underlie the shift from REM to NREM sleep, may provide novel insights into the basis of epileptiform activity in AD. These mice provide an opportunity to study the origin and progression of epileptiform activity in AD from its earliest asymptomatic stages, when diagnosis and evaluation of human patients is not yet possible.

Funding: Supported by a grant from the American Epilepsy Society (MSP) and grants NS085171 and AG065290 from the National Institutes of Health (JC).; COI: The authors report no conflicts of interest.

Poster # 92: MIRNA-501-3P AND MIRNA-502-3P: A PROMISING BIOMARKER PANEL FOR ALZHEIMER'S DISEASE

Presenting author: Subodh Kumar, PhD (Texas Tech HSC - El Paso)

Background: Alzheimer's disease (AD) lacks a less invasive and early detectable biomarker. Here, we investigated the biomarker potential of miR-501-3p and miR-502-3p using different AD sources. Method: MiR-501-3p and miR-502-3p expressions were evaluated in AD cerebrospinal fluid (CSF) exosomes, serum exosomes, familial and sporadic AD fibroblasts and Blymphocytes by qRT-PCR analysis. Further, miR-501-3p and miR-502-3p expressions were analysed in APP, Tau plasmid transfected cells media exosomes and in different brain cell types. Result: MiR-501-3p and miR-502-3p expressions were significantly upregulated in AD CSF exosomes relative to controls. MiRNA levels were high in accordance with amyloid plaque and NFT density in multiple brain regions. Similarly, both miRNAs were elevated in AD and MCI serum exosomes compared with controls. MiR-502-3p expression was high in familial AD and sporadic AD B-lymphocytes. MiR-501-3p and miR-502-3p expression were elevated intracellularly and secreted extracellularly in response to APP and Tau pathology. Finally, neurons and astrocytes displayed high expression of these miRNAs. Conclusion: These results suggest that miR-501-3p and miR-502-3p could be promising biomarkers for AD.

Funding: K99AG065645, R00AG065645, The Edward N. & Margaret G. Marsh Foundation, ; COI: The author would like to inform that he filed a patent on 'Synaptosomal miRNAs and Synapse Functions in Alzheimer's Disease' TTU Ref. No. 2022-016, U.S. Patent App.No. PCT/US2023/019298 on 26 October 2023 related to the contents of this manuscript.

Poster # 93: LONGITUDINAL CHANGES IN WALKING SPEED RESERVE MAY BE AN INDICATOR OF PTAU AND AMYLOID BETA CHANGES

Presenting author: John D Tebbetts, MS (TAMU)

Background: Functional reserve is related to resilience. Resilience is defined as the ability to recover from a stressor while reserve is the capacity. An individual's functional reserve is the difference between baseline and maximal function, representative of their capacity to respond to perturbations. Reserve narrows with age through decrease in maximal capacity and/or increase in baseline requirements to maintain function. If reserve decreases over time, this may be an indicator of accelerated cognitive decline (measured by Plasma phosphorylated Tau (pTau) and Amyloid-Beta (AB)) and increased vulnerability to disease and injury. It is currently unknown if walking speed reserve may be indicative of a change in these markers of dementia risk. Methods: Five middle aged (60.4 3.2yrs) and 4 older (67.8 2.6yrs) individuals underwent gait assessments and a blood draw twice, one year apart. Walking speed reserve (WSR) was quantified by subtracting preferred walking speed from the participants preferred walking speed. Levels of pTau181, pTau217, total pTau, AB 40, and AB 42 were determined. pTau ratios to total Tau and AB42/40 were calculated for both years. Longitudinal changes were calculated by subtracting year one from year two. Spearman correlations between WSR and markers of dementia were run. Results: Longitudinal change in WSR was moderately correlated with pTau217 (r=0.42, p=0.27), although not significant. WSR had non-significant weak relationships with pTau181 (r=-0.13, p=0.74) and AB (r=0.28, p=0.46). Conclusion: At this sample size, longitudinal change in WSR was associated with pTau217, and weakly with AB. Indicating that greater changes in WSR were associated with greater changes in blood biomarkers of dementia. Changes in walking speed are indicative of mobility and mortality. These results provide some evidence that WSR may be an early indicator of cognitive decline and should be investigated in studies with larger samples.

Funding: Funding for this project was provided by a Texas A&M University College of Education and Human Development R3 award. ; COI: No conflicts of interest

Poster # 94: SYSTEMIC METABOLOMICS AND JAK3-SIGNALING IN ADRD

Presenting author: Narendra Kumar PhD (TAMU)

Background: Alzheimer's Disease and Related Dementias (ADRD) remain a major health challenge, with rising prevalence in individuals with Type-2 Diabetes (T2D). T2D and its upstream metabolic syndrome/obesity is a substantive risk factor for ADRD where roughly a 1.52-fold increased risk has been observed in many cohorts, and it accelerates cognitive decline. Multiple, interacting mechanisms link T2D to ADRD risk including brain (central) insulin resistance, chronic systemic and neuro-inflammation, vascular injury (small-vessel disease), glycation/oxidative stress, mitochondrial dysfunction, and disturbed lipid metabolism. These act together to promote amyloid and tau pathology and synaptic failure. Since T2D is vastly heterogenous, multimodal approaches are required to establish causal chains and treatment readouts. Emerging evidence suggests that gut-brain communication - involving the intestinal epithelium, immune system, and microglia - plays a pivotal role in accelerating neuroinflammation, amyloid/tau pathology, and cognitive decline. However, the precise mechanisms that link gut dysbiosis and metabolic disease to brain pathology remain poorly understood. Janus kinase 3 (Jak3) is a signaling node that integrates immune, epithelial, and neuroglial responses, making it an ideal molecular handle to dissect gut-brain mechanisms in ADRD. Previously we reported that obese humans have reduced expression of intestinal Janus kinase-3 (Jak3), a non-receptor tyrosine kinase, and reconstituting these conditions using global and intestinal mucosal-specific Jak3-deficient mouse demonstrate predisposition to obesity associated T2D, gut-dysbiosis, and ADRD-like brain phenotype. Methodology: Towards characterizing these phenotypes in a longitudinal multi-omic cohorts specifically enriched for T2D to ADRD trajectories, in this study we used serum metabolomic profiling through metabolic pathways enrichment approach to validate the Jak3-deficient mouse model with corresponding human sample. Results: We reveal that Jak3-deficiency leads to intestinal inflammation and gutdysbiosis which was associated with increased microglial activation, increased TLR-4 and HIF1 expression in brain, and compromised brain homeostasis. In addition to these phenotypes, there was 80-95% similarities in top 25 enriched metabolic pathways impacted. Conclusion: Thus, we not only demonstrate the metabolic pathways impacted systemically during T2D+ADRD but specific pathophysiology shown by the genetically modified mouse that closely and metabolomically recapitulates corresponding human conditions.

Funding: TAM-seedling grant.; COI: None reported.

Poster # 95: PRELIMINARY EFFECTS OF CEREBELLAR ITBS ON COGNITION IN COGNITIVE DECLINE: A MULTIMODAL TMS-FMRI INVESTIGATION

Presenting author: Rushali Patel, MA (TAMU)

Background: Prior to the diagnosis of Alzheimer's disease (AD) individuals may experience cognitive decline (CD), and early disruption of frontoparietal-cerebellar networks underlying cognitive control and working memory. Non-invasive brain stimulation targeting these networks may reveal or enhance compensatory neural mechanisms relevant to AD. This study investigated the effects of intermittent theta burst stimulation (iTBS) applied to the right Crus I region of the cerebellum, followed by neuroimaging, to evaluate stimulation-related changes in cognition and brain activity in CD and young adults (YA). Methods Forty participants will be recruited per group (CD and YA) as part of an ongoing larger study. Here we present preliminary behavioral results from a subset of participants (N-back: 17 YA, 4 CD; Stroop: 12 YA, 5 CD; Sequence Learning: 27 YA, 6 CD). Following either Active or Sham iTBS, participants underwent MRI scanning beginning with resting-state acquisition and followed by task-based fMRI during Nback, Stroop, and Sequence Learning paradigms. Behavioral outcomes included accuracy and reaction time (RT); imaging analyses (in progress) will examine cerebello-cortical connectivity and task-related activation. Results YA participants showed minimal stimulation effects. In the N-back task, CD participants demonstrated higher accuracy and faster RTs under Active iTBS, particularly in the 2-back condition, suggesting improved working-memory efficiency. In the Stroop task, CD participants exhibited higher accuracy and slightly reduced RTs under Active stimulation, consistent with partial mitigation of interference effects. YA accuracy remained at ceiling across conditions. During the Sequence Learning task, both groups showed improvements in performance. However, CD participants under the Active condition consistently showed lower accuracy and slower RTs compared to Sham across phases, indicating that cerebellar iTBS did not enhance sequence learning and may have altered response dynamics without improving efficiency. Conclusions These preliminary findings demonstrate the feasibility of integrating cerebellar iTBS with fMRI to probe neural and behavioral plasticity in CD. Stimulation effects were task-specific, facilitating working memory and cognitive control but not sequence learning. This pattern suggests differential engagement of cerebello-cortical circuits across task domains. Given small samples and ongoing recruitment, interpretations remain preliminary; forthcoming neuroimaging analyses will clarify network-level mechanisms and AD-relevant implications of cerebellar modulation.

Funding: None reported.; COI: None reported.

Poster # 96: SEX DIFFERENCES IN THE EFFECTS OF CONTROLLED CIRCADIAN DYSREGULATION ON GUT MICROBIOME AND BARRIER ARCHITECTURE

Presenting author: Marivelle Levett, BS (Texas A&M UHSC)

Background: Circadian rhythm dysregulation, a common outcome of modern lifestyles, contributes to chronic inflammation and heightened risk for neurologic and metabolic diseases. We investigated sex differences and whether chronic circadian misalignment induces gut dysbiosis that worsens stroke severity and mortality. Methods: Adult Sprague Dawley rats were maintained under either a standard 12:12 light-dark (LD) cycle (fixed LD) or a chronic shift-work paradigm in which the LD cycle was phase-advanced by 12 hours every 5 days (shifted LD) for 50 days. Fecal and serum samples collected before and after treatment underwent 16S rRNA gene sequencing and short-chain fatty acid (SCFA) analysis. Additional groups were used for gut histology, tight junction protein immunostaining, and measurements of circulating inflammatory mediators, including the bacterial endotoxin LPS and cytokine IL-17A. Results: Shifted LD cycles produced sex-dependent alterations in gut microbiota. Male rats exhibited decreased microbial richness and evenness, loss of beneficial SCFA-producing taxa (Turicibacter, Lachnospiraceae), and marked beta diversity shifts, whereas females showed minimal disruption. The abundance of Akkermansia muciniphila, Butyricicoccus, and Prevotella (taxa linked to gut integrity and neuroprotection) correlated with improved stroke survival in males but not females. Histological analysis revealed significant epithelial injury in shifted LD males, including villus shortening, crypt hyperplasia, and loss of the tight junction protein ZO-1, indicating impaired barrier function. These pathological changes were absent in females. Correspondingly, serum LPS and IL-17A levels were significantly elevated in shifted LD males, while inflammatory responses in females were more variable and attenuated. Although total SCFA levels remained unchanged, butyrate, a key SCFA with anti-inflammatory and neuroprotective effects, was selectively decreased in shifted LD males, linking circadian disruption to gut-derived systemic inflammation. Conclusion: Chronic circadian misalignment compromises gut integrity and promotes systemic inflammation in a sex-specific manner, with males displaying greater vulnerability. These findings highlight the gut microbiome as a potential mediator of circadian rhythm-associated disease risk and suggest that targeting microbiota-host interactions may mitigate inflammation and improve outcomes in circadian disruption-related disorders.

Funding: Woodnext Foundation; COI: The authors have nothing to disclose.

Poster # 97: DIGITAL BIOMAKERS OF EXTREME HEAT AND COGNITION

Presenting author: So-Min Cheong, PhD (Texas A&M UHSC)

Background: Extreme heat exposure may pose a serious dementia risk but limited research exists especially reflecting one's daily living environment. Individually experienced temperatures vary widely within and between individuals as they move in and out of locations that are equipped with and without air conditioning (AC) and ventilation. While these small variations in one's microenvironment are important to ascertain the overall exposure level and its impact on health, they are not well established. We demonstrate the effects of changing person-specific ambient temperature on physical activity and sleep measured with the wrist-worn accelerometer. Physical activity and sleep are major modifiable risk factors for dementia. We also present a pilot study showing the effect of person-specific ambient temperature on momentary cognition. Methods: 1) We modelled the effects of person-specific ambient temperature on digital biomarkers of physical activity and sleep using linear mixed effects model (Cheong and Gaynanov 2024). There were 30 participants with a mean age of 58 living in low-income and predominantly Black neighborhood in Houston, TX. 2) We measured the effects of person-specific ambient temperature on daily cognition. 9 participants, aged 70-92 (mean age=78) living in College Station, TX, completed a suite of nine NeuroUX cognitive ecological momentary assessments on their smartphones daily for 14 days in September of 2024 (Moore et al. 2025). These assessments measured recognition memory, working memory, processing speed, and executive functions. A random-intercept multilevel model was used to examine the dynamic relationship between temperature and daily cognitive performance. Results were significant for 1) both physical activity and sleep. A 1° increase in ambient temperature led to around a 2% increase in sedentary time and sleep efficiency. 2) Although not statistically significant due to the small sample size, small to medium effects were observed in working memory, executive functions, and recognition memory. Conclusion: Digital biomarkers provide data-driven insights into the effects of extreme heat on cognition.

Funding: texas A&M startup fund; COI: n/a

Poster # 98: INFLAMMATION AND COGNITIVE IMPAIRMENT: EVIDENCE FOR NEURODEGENERATION AS A POTENTIAL MEDIATOR

Presenting author: Edna Patricia P. Mendoza, PhD (UNTHSC)

Background: Systemic inflammation is associated with cognitive decline and Alzheimer's Disease and Related Dementias (ADRD), but the mediation role of plasma biomarkers amyloid, tau, and neurodegeneration (ATN) remain unclear. The objective of this study is to identify inflammatory biomarkers for predicting cognition across diverse populations. Methods: Cross-sectional analysis of 4,134 adults aged ≥ 50 years from the Health and Aging Brain Study-Health Disparities measured plasma biomarkers using Quanterix Simoa. The outcome variable was cognitive diagnosis (normal 73% vs. impaired 27%, MCI/dementia). The exposure was systemic inflammation, tested using (1) grouped anti-, pro-, and acute-phase inflammatory biomarkers (IL10, IL-6, TNF-α, CRP) and (2) pro-inflammatory cytokines only (IL-6, TNF-α). This approach was used to assess whether varied inflammatory biomarkers or cytokines-only can capture inflammation's impact on cognition. The mediators were three ATN categories, based on corresponding plasma biomarkers: amyloid pathology (Aβ42 and Aβ40), tau pathology (p-tau181 and total tau), and neurodegeneration (neurofilament light [NfL]). Confirmatory factor analysis was used to create latent constructs, and structural equation modeling tested ATN mediation of inflammation-cognition relationships. Models were stratified by race/ethnicity and sex, adjusting for demographics, APOE4, and comorbidities. Results: Both inflammatory measures showed adequate loadings (grouped: 0.39-0.68; cytokines-only: 0.44-0.77) and predicted ATN biomarkers. The grouped model (tested in overall population, Hispanics, Non-Hispanic Whites [NHW], by sex) demonstrated: (1) inflammation increased amyloid burden, but amyloid did not mediate cognitive effects; (2) sex-specific tau mediation only in males (inflammation predicted tau accumulation β=0.394, p=0.002, which predicted cognitive impairment, indirect effect β=8.12, p=0.001); (3) for neurodegeneration as primary mediator, inflammation predicted NfL elevation, which predicted cognitive impairment in overall population (indirect effect β=0.0094, p=0.013) and NHW (β=0.024, p=0.006), with non-significant direct effects indicating complete mediation. The cytokines-only model (tested in overall population, Non-Hispanic Blacks, Hispanics, NHW, by sex) showed weaker results: poor tau model fit (CFI=0.59-0.76, TLI=0.260.56), loss of male tau mediation, and attenuated NfL mediation limited to NHW only (p=0.027). CRP and IL-10, despite modest loadings, contributed essential inflammatory variance for detecting mediation. Conclusion: Our study indicates neurodegeneration as the primary mediator linking inflammation to cognitive impairment, with pathways that differ by race/ethnicity and by sex.

Funding: The HABS-HD Study was supported by the National Institute on Aging of the National Institutes of Health under Award Numbers R01AG054073, R01AG058533, R01AG070862, and U19AG078109. ; COI: There are no conflicts of interest to disclose.

Poster # 99: AGE-RELATED ALPHA CONNECTIVITY USING VOLUME-CONDUCTIONCORRECTED WPLI² IN THE LEMON DATASET

Presenting author: Gabriela A. Renta-Lopez, MS (UT Austin)

Background: Recent work in younger-old adults (ages 45-64) demonstrated age-related changes in resting-state EEG coherence, showing that alpha-band connectivity increased during eyesclosed conditions, with greater coherence observed in older individuals with high cognitive reserve (CR). This suggests preserved connectivity may be essential for CR. Traditional coherence measures are confounded by volume-conduction artifacts, leading to spurious zero-lag synchronization caused by instantaneous electrical-field spread through the skull. We extended these findings to an older cohort using debiased weighted phase lag index (wPLI²), which removes zero-lag connections to isolate genuine phase-based connectivity. Method: We analyzed 60 older adults (M=67.6±4.9) from the LEMON (Leipzig Study for Mind-BodyEmotion Interactions) dataset, an open-access multimodal neuroimaging repository. Participants completed a 16-minute resting-state EEG with alternating eyes-closed and eyes-open conditions. Following ICA-based artifact rejection, we computed wPLI² connectivity across six frequency bands using multitaper spectral estimation. CR was operationalized as a composite of years of education and vocabulary IQ. We assessed relationships with age and episodic memory (CVLT). Result: After Benjamini-Hochberg FDR correction (q<0.05), low-alpha connectivity showed robust eyes-closed elevation (50% increase [95% CI: 20%, 81%], t(59)=3.29, p_FDR=0.010, d=0.42 [95% CI: 0.15, 0.69]), replicating classic alpha blocking with medium effect size. Extending prior findings to an older cohort, we observed differential age-related patterns by CR level. High-reserve individuals showed age-related decline in low-alpha (r=−0.38 [95% CI: −0.65, −0.03], p=0.037) and high-alpha (r=−0.39 [95% CI: −0.66, −0.03], p=0.034), while lowreserve individuals showed stable connectivity (both r=+0.07, p>0.69). Formal interaction analyses revealed that CR significantly moderated age-related connectivity trajectories in theta frequencies, with similar patterns observed in alpha bands. These differential trajectories were detected after removing volume-conduction confounds. Conclusion: Strong alpha-blocking replication validates wPLI² as a robust method for measuring EEG connectivity free of volumeconduction artifacts. The differential age-trajectories in high-CR versus low-CR groups extend traditional models of CR to older populations. This preliminary work establishes a foundation for investigating whether preserved connectivity serves as a mechanism underlying CR effects on cognitive function in aging, with future analyses planned to explore these relationships across multiple frequency bands and network configurations.

Funding: Coleman Fung Foundation; COI: No conflict of interest

Poster # 100: RELATIONSHIP OF GLOBAL COGNITIVE SCALES AND NEUROPSYCHOLOGICAL PROFILE WITH AMYLOID BURDEN

Presenting author: Khushboo Verma, MD (UT Tyler)

Background: Global cognitive scales such as the Mini-Mental State Examination (MMSE) and Clinical Dementia Rating (CDR) guide diagnosis and staging, but it is unclear how well highdimensional biomarkers derived from β-amyloid (Aβ) PET and neuropsychological profiles map onto these discrete clinical grades. Methods: We analyzed OASIS-3, including cognitively normal participants (CDR = 0; n = 755) and individuals with impairment (CDR = 0.5-1; n = 622). Aβ deposition was measured with Pittsburgh compound-B PET (PIB-PET). Images were registered to structural MRI, segmented volumetrically, and processed to yield partial-volumecorrected standardized uptake value ratios (SUVRs) in anatomically defined regions.
Associations between regional SUVRs and neuropsychological test scores were evaluated using correlation analyses with false-discovery-rate (FDR) adjustment (q < 0.001). To visualize joint structure in the psychometric and Aβ spaces, we applied t-distributed stochastic neighbor embedding (t-SNE) for dimensionality reduction and examined whether the resulting lowdimensional embeddings tracked CDR strata. Results: Greater clinical severity (higher CDR, lower MMSE) was associated with elevated Aβ burden across widespread cortical and select subcortical regions (FDR-corrected q < 0.001). In low-dimensional embeddings, psychometric scores and regional Aβ measures were arranged along a graded trajectory that corresponded to global dementia status, supporting a continuum rather than discrete clustering of profiles (global alignment tests, q < 0.001). Conclusions: MMSE/CDR staging is mirrored by a brain-wide pattern of Aβ accumulation and coordinated shifts in neuropsychological performance. Dimensionality reduction reveals a continuous manifold linking psychometrics and Aβ-PET, indicating that clinical severity is embedded within high-dimensional cognitive and pathological feature spaces. These findings support a parsimonious, interpretable computational representation of dementia severity that explains substantial variance in PIB-PET signal while maintaining clinical relevance.

Funding: No funding was received for this study; COI: None

Poster # 101: DIGITAL HEALTH AND SURVEY RESPONSE PATTERNS AMONG ADULTS ACROSS THE COGNITIVE SPECTRUM: COMPADRE-CART AND DETECT-AD STUDIES

Presenting author: Marialy Salinas Valdez, BA (UTH San Antonio)

Introduction: Early detection of cognitive decline remains challenging, particularly for identifying subtle functional changes. Passive digital health monitoring offers continuous, unobtrusive assessment of real-world behaviors that may signal emerging cognitive changes. However, the sensitivity of digital health metrics for differentiating stages across the cognitive continuum (i.e., cognitively unimpaired [CU], subjective cognitive impairment [SCI], mild cognitive impairment [MCI]) remains unclear. We examined whether home-based digital health metrics and weekly survey patterns differ across three cognitive status groups in two cohort studies. Methods: A total of 219 participants (mean age: 72.8±6.2 years; 68% female) from the COMPADRE-CART (South Texas) and DETECT-AD (Portland Metro) studies were included. Participants were classified as CU (Clinical Dementia Rating [CDR]=0 without subjective cognitive complaints; n=76), SCI (CDR]=0 with subjective cognitive complaints, n=83), or MCI (CDR=0.5; n=60). Using Kruskal-Wallis tests for continuous outcomes and Chi-square test for categorical outcomes (α=0.05), we examined associations between cognitive status and digital metrics derived from wearable activity trackers (daily steps, sleep duration), in-home motion sensors (room transitions, time outside home, time in independent life-space activities), and weekly online surveys (survey completion time, self-reported pain (0-10), blue mood (≥ 3 days), and loneliness. We analyzed inhome sensor data separately for single and double resident households. Results: Weekly survey completion time in seconds (p=0.017) and self-reported pain levels (p=0.046) differed significantly across groups. Post-hoc testing indicated the MCI group had longer survey completion time (330.47±241.47) compared to CU (243.30±168.63; p=0.017) and SCI (237.14±141.99; p=0.007) groups. Pain levels were higher in the MCI group (2.31±1.96) relative to the CU group (1.48±1.57; p =0.0125). Metrics from the wearable activity trackers and in-home sensors did not differ between groups (all p>0.05) Conclusion: Individuals with MCI displayed longer survey completion times and endorsed higher pain levels. Digital metrics did not distinguish between the SCI and CU groups, which could be attributable to the clinical heterogeneity associated with SCI. Future studies incorporating Alzheimer's disease biomarkers will be important for determining whether digital metrics can detect subtle changes in the prodromal stage. Longitudinal monitoring may also improve digital device sensitivity for identifying mild functional changes relative to an individual's baseline.

Funding: R01 AG077472 P30 AG066546 U2C AG054397; COI: Zachary Beattie and Jeffrey Kaye have a financial interest in Life Analytics, Inc., a company that may have a commercial interest in the results of this research and technology. This potential conflict of interest has been reviewed and managed by OHSU. Vanessa Young is an executive committee member of the technology & dementia PIA of the Alzheimer's Association.

Poster # 102: DEPRESSION IS ASSOCIATED WITH HIGHER T-TAU LEVEL IN PLASMA: A REPORT FROM THE FRAMINGHAM HEART STUDY.

Presenting author: Carlos Alberto Gaona Ortiz Gil, MD (UTH San Antonio)

Background: Depression is an established risk factor for neurodegenerative diseases, including Alzheimer's disease (AD). While blood-based biomarkers (BBMs) have been promising for early detection of neurodegeneration, their relationship with depressive symptoms remains controversial. We explored the cross-sectional relationship between depression and BBMs of neurodegeneration and AD pathology (t-tau, NfL, GFAP, and p-tau181) using ultrasensitive immunoassay. Methods: The current study comprises the Offspring (exam 9) and Omni 1 (exam 4) cohorts of the Framingham Heart study (FHS) who completed their visits between 2011 and 2014. Depression was considered a binary variable (yes/no) and defined as having a score ≥ 16 on the Center for Epidemiologic Studies Depression Scale or reporting antidepressant use. Single Molecule Array (SIMOA) assays were used to assess plasma levels of t-tau, NfL, GFAP, and p-tau181. All BBM values were log-transformed and standardized. Multivariable regression models were adjusted for (1) sex, age, cohort, and additionally (2) education, APOE4 status, FHS cardiovascular risk score, estimated glomerular filtration rate, and C-reactive protein.  Results: Among 2,466 participants (mean age 69.8 years, 44.9 % male), 19.8% met criteria for depression (see Table 1 for clinical characteristics). Depression was significantly associated with higher t-tau levels in both model 1 (β±SE =0.24±0.05, p<0.001) and model 2 (0.21±0.05, p<0.001). No associations were found between depression and p-tau 181, NfL, or GFAP, as shown in Table 2.  Conclusion: Depression was associated with higher t-tau levels in these cohorts. While current findings corroborate previous studies showing that depressive symptoms were not associated with plasma biomarkers of AD pathology, the positive association with t-tau level suggests the involvement of neuronal injury.

Funding: P30AG066546, R01AG062531, AG059421, HL105756, N01-HC-25195, HHSN268201500001I,75N92019D00031. ; COI: No

Poster # 103: MULTI-SITE VALIDATION OF THE FUJIREBIO LUMIPULSE NFL KIT FOR THE MARKVCID CONSORTIUM

Presenting author: Sarah Dean, BS (UTH San Antonio)

Background: The MarkVCID consortium was established in 2016 to identify biomarkers for vascular contributions to cognitive impairment and dementia (VCID). Plasma neurofilament light (NfL) is released upon neuroaxonal injury and is commonly elevated in neurodegenerative diseases. We previously demonstrated that plasma NfL measured using the Quanterix Simoa HDX could be a reliable biomarker for risk stratification in VCID clinical trials, but this kit was discontinued. Building on prior validation using the Quanterix Simoa platform, we evaluated the precision, repeatability, and reproducibility of the Fujirebio Lumipulse system for quantifying plasma NfL. Methods: Plasma samples were sourced from the MarkVCID consortium and NfL was measured at four sites: UT Health San Antonio, University of Kentucky, John Hopkins University, and the Mayo Clinic. We assessed intra- and inter-site repeatability (n=35 participant samples measured in duplicate at all sites), test-retest repeatability (n=30 participant samples collected three times over one month), and inter-run reproducibility (n=10 participant samples measured over 3 runs) using the Lumipulse G1200 instrument, a fully automated immunoassay system. The results were compared to previous results generated using the Quanterix Simoa Neuro 4-plex A kit. Results: During a single run (intra-run), both platforms showed minimal variation (Lumipulse mean site coefficient of variation [CV] = 2.0-5.6%; Quanterix CV = 3.65.1%). Across runs (inter-run), the Lumipulse demonstrated greater consistency (CV = 2.3-6.3%) compared to the Quanterix (CV = 6.0-11.5%). Test-retest reliability was high and identical for both platforms (intraclass correlation coefficient [ICC] = 0.97). Between sites, results were slightly more consistent for the Lumipulse (ICC = 0.98-0.998) than the Quanterix (ICC = 0.960.98). However, NfL concentrations were found to be significantly different at each site (p<0.0001), whereas there was no difference when measured by Lumipulse (p=0.44). Overall agreement between platforms was high, with a mean ICC of 0.884. Conclusion: The Fujirebio Lumipulse NfL assay demonstrated strong reliability, with higher precision and greater consistency across runs and sites than the Quanteix Simoa assay. These results support the use of the Fujirebio Lumipulse platform to quantify plasma NfL for future VCID research and multi-site clinical trials.

Funding: Data collection and sharing for this project was funded by NINDS/NIA as part of the MarkVCID Consortium (U24NS100591, UF1NS100599, UF1NS125513, UF1NS100588, UF1NS100608, UF1NS125488, UF1NS100598, UF1NS100614, UF1NS125417, UF1NS125512).; COI: Fujireibo provided the kits for the Lumipulse for free.

Poster # 104: PILOT STUDY TO EVALUATE SALIVARY LACTOFERRIN LEVELS IN A SOUTH TEXAS ALZHEIMER'S DISEASE AND CAREGIVER POPULATION

Presenting author: Valentina R. Garbarino, PhD (UTH San Antonio)

Background: Lactoferrin is an endogenous iron-binding glycoprotein with important functions in innate immunity. Acute stress has been associated with increases in salivary lactoferrin levels, and since 2017, conflicting results have suggested that salivary lactoferrin levels could serve as a sensitive and non-invasive biomarker for the diagnosis Alzheimer's disease (AD). The conflict in the field could be due to a number of factors including medication effects, heterogeneity in ethnic backgrounds and diagnosis, sample collection methodology and oral health variability, etc. In an attempt to understand the controversy in the literature we have designed a study to evaluate salivary lactoferrin levels in a diverse South Texas population. Method Lactoferrin and cortisol levels were measured using an Abcam Salivary Lactoferrin Kit in samples collected from two separate studies including individuals who have a diagnosis of AD or serve as family caregivers for a loved-one with AD. The first study included adult white Hispanic (N=9) and non-Hispanic (N=9) female caregivers. The second pilot study included individuals with a diagnosis of AD (N=3) and their associated spousal caregivers (N=3) stored with and without protease inhibitor (PI). Result Un-normalized, lactoferrin levels without the addition of PI, were not different from each other across the studies caregiver populations (144 ± 134 ug/ml and 74.7 ± 48.6 ug/ml), but salivary lactoferrin levels in the AD group were significantly higher (431 ± 315 ug/ml) than the control groups (P< 0.036), which were approximately 90 times higher than previous reports. This difference disappeared (P= 0.586) with the addition of PI indicating a potential masking of lactoferrin by the contents of the PI. Cortisol levels were not different between caregivers and AD participants in this small cohort (P= 0.534). There were no differences in salivary lactoferrin levels between Hispanic and non-Hispanic individuals. Conclusion Our results conflict with previous reports of a negative association between salivary lactoferrin levels and cognitive impairment. Multiple factors will need to be considered and addressed in future studies to standardize the collection, processing, storage, and data stratification and normalization methods for determining salivary lactoferrin levels.

Funding: VRG: 5T-P30-AG066546-03, Subproject 6187; NIAP25-1446397; TR004529; Institute for Integration of Medicine and Science (IIMS) Clinical and Translational Science Award Pilot Program; C-KY:GRECC, STVHC. ; COI: None.

Poster # 105: LONG SLEEP DURATION IS ASSOCIATED WITH ELEVATED PLASMA P-TAU181 AND TOTAL TAU IN THE FRAMINGHAM HEART STUDY

Presenting author: Vanessa M. Young, MS (UTH San Antonio)

BACKGROUND: Short and long sleep durations have been associated with increased dementia risk. However, the biological pathways linking sleep to dementia are poorly understood. Prior studies have predominantly used linear or categorical approaches that might have missed nonlinear relationships, and few have systematically accounted for kidney function, which declines with age and might confound biomarker interpretation. Here, we explored whether associations between sleep duration and plasma biomarkers of Alzheimer's disease (AD) and neurodegeneration could reveal mechanisms linking sleep duration to dementia, thereby, informing preventive intervention strategies. METHODS: The sample included Framingham Heart Study Generation 2 (exam 9) and Omni 1 (exam 4) cohorts with visit cycles conducted between 2011-2014 (n= 2,410; mean age: 70.0±8.5 years; 55.2% female). We examined crosssectional associations between self-reported sleep duration (hours/night) and plasma biomarkers using linear models and restricted cubic splines. Biomarkers included p-tau181, t-tau, NfL, and GFAP. Model 1 was adjusted for age, sex, cohort, APOEε4; model 2 was further adjusted for obstructive sleep apnea, depression, and cystatin C-based estimated glomerular filtration rate (eGFR). RESULTS: The sleep-biomarker associations were best explained by non-linear terms.
Sleep duration showed significant nonlinear associations with p-tau181 (J-shaped pattern), t-tau (U-shaped pattern), and NfL (positive linear trend) in Model 1 (p<0.01). In Model 2, sleep was only associated with p-tau181 (J-shaped) (p=0.005), with lowest levels at 7-8h and elevations at ≥ 10h. T-tau showed a significant association with sleep duration (p=0.017) but with a less pronounced nonlinear pattern. Kidney function fully explained NfL associations. CONCLUSIONS: Sleep duration exhibited robust nonlinear associations with p-tau181 independent of kidney function, while t-tau showed weaker, kidney-sensitive patterns. Extended sleep (≥ 10h) might serve as a clinical alert for AD pathology, though longitudinal studies are needed to establish causality and intervention potential. These findings highlight the importance of kidney function adjustment when interpreting plasma biomarkers in aging research.

Funding: P30AG066546, R01AG062531, AG059421, HL105756, N01-HC-25195, HHSN268201500001I,75N92019D00031. This work is part of the Translational Science program at UTHSCSA supported by NCATS UM1 TR004538.; COI: Mrs. Young is a PIA executive committee member of the Alzheimer's Association. Dr. Seshadri reports consulting for Eisai and Biogen. Dr. Baril reports speaking fees from Eisai. Dr. Salardini reports speaking fees from Lilly.

Poster # 106: BLOOD-BASED RNA PROFILING IDENTIFIES DISEASE-COMMON AND DISEASE-SPECIFIC MOLECULES IN ALZHEIMER'S DISEASE AND MULTIPLE SCLEROSIS

Presenting author: Eun Seok Choi, PhD (UTMB)

Background: Serum biomarkers, including Tau and NfL, show promise for Alzheimer's disease (AD) diagnosis and prognosis but are not yet standalone tools. Using a buffy coat for biomarker identification can reduce donor variability by removing donor-specific soluble serum factors. We profiled mRNA and small RNA expression in buffy coat samples from healthy individuals and patients with AD and multiple sclerosis (MS). Methods: Buffy coat samples were obtained from individuals with AD (n=4), MS (n=4), and healthy controls (n=5). Bulk mRNA and small RNA sequencing were performed, followed by differential expression and Gene Ontology (GO) enrichment analysis. Candidate mRNAs and small RNAs were further validated in an expanded cohort (AD (n=24), MS (n=20), control (n=23)). Results: mRNA profiling revealed diseasedistinct cellular signatures: AD showed significant reductions in resting NK cells and monocytes, whereas MS exhibited elevated NK and T cell populations, including naive CD4⁺ and memory T cells. There were also trends for MS patients having fewer CD4⁺ memory T cells and monocytes. GO analysis identified shared molecular signatures of enhanced antiviral responses, including five common genes, and impaired immune regulation. Disease-specific pathways included kinase activity and single-stranded RNA binding (AD), and IgA immunoglobulin complex formation (MS). Furthermore, small RNA sequencing detected 296 dysregulated sncRNAs in AD and 42 in MS. Analysis prioritized disease-specific tRNA-derived fragments (tRF5-ValCAC in AD, tRF5HisGTG in MS) and one tRF (tRF5-GlyCCC-2), which was commonly downregulated in both. Conclusion: This study highlights shared molecular signatures between AD and MS, including activation of the antiviral pathway, and identifies disease-specific tRFs. Although our sample size is limited, these findings support the potential of buffy coat-derived small RNA, particularly tRFs, as minimally invasive biomarkers. Larger cohorts and mechanistic studies are warranted to validate their diagnostic and predictive utility.

Funding: This work was supported by grants from the US National Institutes of Health (NIH) R21 AI166543, R21 AG069226, R61 AG075725, ERP-1252718 from the American Lung Association, MS240193 from the Department of Defense (DoD), and TARRC InvestigatorInitiated Research Award to Xiaoyong Bao.; COI: The authors have no conflicts of interest to declare that are relevant to the content of this article.

Poster # 107: TAU OLIGOMERS IN CIRCULATING EXTRACELLULAR VESICLES AS PREDICTIVE BIOMARKERS FOR VULNERABILITY AND RESILIENCE TO ALZHEIMER'S DISEASE

Presenting author: Michela Marcatti, PhD (UTMB)

Alzheimer's disease (AD) is a progressive neurodegenerative disorder that often begins with mild cognitive impairment (MCI), a prodromal stage where symptoms are subtle and progression uncertain. Not all individuals with MCI develop AD, underscoring the need to understand the molecular features distinguishing resilience from vulnerability to cognitive decline. Among AD's hallmarks, misfolded tau assembles into soluble, toxic oligomers (TauOs) that are structurally heterogeneous and can form distinct polymorphs, some behaving as strains with unique biochemical and pathological properties. Yet, how these TauO polymorphs shape disease trajectories remains largely unexplored. As one of the most prevalent dementias worldwide, AD demands predictive biomarkers for early detection and intervention. Recent efforts have focused on blood-based markers, particularly plasma brain-derived extracellular vesicles (pl-BDEVs), which reflect central nervous system alterations. Unlike total tau, often derived from peripheral sources, tau species within pl-BDEVs may provide a more specific and accessible window into early AD-related changes, including those driven by TauO polymorphs. In this study, we analyzed plasma extracellular vesicles (pl-EVs) from longitudinal samples of participants in the Texas Alzheimer's Research and Care Consortium (TARCC), who were initially diagnosed with MCI and later either progressed to AD (vulnerable) or remained stable (resilient). We characterized pl-EVs by nanoparticle tracking analysis, western blot (CD63, CD9, CD81), and electron microscopy. TauOs were immunoprecipitated and characterized by proteinase K (PK) digestion, seeding assay, atomic force microscopy (AFM), and western blot. We further assessed their cytotoxicity in SH-SY5Y cells and synaptotoxicity in human cortical synaptosomes. Moreover, we isolated CNS cell-type-specific pl-BDEVs using immunomagnetic beads conjugated with cell-type markers (L1CAM, TMEM119, GLAST, and MOG) and immunoprecipitated brain-derived TauOs (BDTOs). We identified distinct PK digestion patterns, morphology, and toxicity between TauOs from vulnerable and resilient MCI individuals, as well as different distributions across CNS-derived EVs. These findings reveal conformationally distinct TauO polymorphs linked to cognitive vulnerability and resilience, addressing the need for predictive AD biomarkers. Detecting BDTOs in accessible, brain-derived pl-EVs could enable preclinical forecasting and support early-stage therapeutic interventions.

Funding: NHI/NIA R01AG07313302 to GT; COI: NA

Poster # 108: COMPARING FUNCTIONAL CONNECTIVITY CHANGES IN TWO TAUOPATHY MOUSE MODELS

Presenting author: Kevin Nguyen, BS (BCM)

Background: Tau accumulation plays a central role in the progression of Alzheimer's Disease (AD). MRI is a non-invasive imaging technique to measure changes in brain structure and function. Over the course of AD, changes in the relationship between the activity of brain regions, described as functional connectivity (FC), have been observed. PET studies in AD patients have proposed a link between tau tangles and functional hypoconnectivity. We used two tauopathy mouse models to evaluate the contribution of tau to FC changes across disease progression. Methods: We used the rTg4510 and PS19 mouse models of tauopathy. rTg4510 mice were assessed at 3 months (prior to tau tangle formation), 6-months (beginning of tangle deposition), and 10-months (tangle accumulation, some neurodegeneration). PS19 mice were assessed at 6months (early tangle accumulation). Spatial learning and memory was evaluated using Morris Water Maze (MWM). FC was measured between 30 brain regions using resting-state functional MRI. Imaging was done using awake mice to avoid the potentially confounding effects of anesthesia. Results: Although tangle formation was absent from the rTg4510 models at the 3month time point, there were significant deficits in spatial learning and memory. These deficits worsen by the 10-month mark. At 6-months, PS19 mice also exhibited deficits in spatial learning and memory. Functional hyperconnectivity was observed in the rTg4510 mice at 3-months, and a mix of hyper- and hypo-connectivity at 6- and 10-months. PS19 mice displayed mixed changes in FC at 6-months. Conclusion: In both tauopathy mouse models, tau pathology is associated with spatial learning and memory deficits. In line with these cognitive deficits, FC changes were observed in memory-related brain regions at all time points. While previous PET studies detected a relationship between tau tangles and hypoconnectivity, the observed early hyperconnectivity suggests a more complex relationship between tau pathology and FC.

Funding: NIH 1R56AG071152 NIH 1R01AG081192; COI: The authors have no conflicts of interest to declare

Poster # 109: A LONGITUDINAL INVESTIGATION OF THE CEREBELLUM IN ADULTHOOD: ANATOMICAL AND NETWORK CHANGES, MOTOR FUNCTION, AND COGNITION

Presenting author: BriAnna Lowe, MS (TAMU)

Abstract
Background: The brain is subject to changes in both function and structure with aging, which can result in deficits in function. Past literature has emphasized the cerebral cortex in late adulthood, the cerebellum if often excluded, despite undergoing changes that impact motor and cognitive functioning in later life. The inclusion of these factors from a longitudinal perspective is essential to form a comprehensive picture of how the cerebellum contributes to performance over time and may relate to risk and resilience for symptoms associated with age-related neurodegenerative disease. Method: 175 participants were recruited for a 2-year longitudinal study that used multi-modal neuroimaging, motor, and cognitive testing. Participants wore a wearable activity monitor, completed cognitive activities, and a questionnaire battery during this time. The second session consisted of multi-modal neuroimaging. Participants were assessed at three time points (baseline, 12-months, 24-months). Here we assessed cerebellar resting state connectivity changes from baseline to 12-months in 86 participants. We also considered relationships between baseline connectivity and behavioral performance 12 months later. All analyses were completed using the CONN toolbox with methods and thresholding consistent with recent work from our group. Results: Preliminary analyses indicate that connectivity between cerebellar lobule VI and the prefrontal cortex decrease over 12 months. Further, Crus I connectivity with the prefrontal cortex at baseline is predictive of cognitive performance at time 2, and in those with better cognitive function at baseline, they see more robust connectivity between Crus I and the parietal cortex over time. Lobule V will also be further examined in the context of motor behavior. Conclusion: Understanding the role of the cerebellum in motor and non-motor performance is essential for identifying changes that occur earlier in adulthood as the cerebellum may be a key metric for risk and resilience to cortical pathology, such as that seen in Alzheimer's Disease. Our preliminary analyses in particular point to a potential role of cerebellofrontal networks in cognitive trajectories over 12 months.

Funding: 1 R01 AG064010-01; COI: No conflicts of interest

Poster # 110: CEREBELLO-BASAL GANGLIA NETWORKS AND BEHAVIOR ACROSS ADULTHOOD AND AGING

Presenting author: Casey Delaney, BA, BS (TAMU)

Background: The cerebellum has been under-studied in the context of aging, cognition, and reward related processing, despite research emphasizing its involvement in these functions. The cerebello-basal ganglia (CB-BG) circuit may be a key point to better understand how the brain ages. As our population continues to age, it is vital to better understand the relationship between cerebellar changes and age-related cognitive and motor decline. Method: We investigated the cerebello-basal ganglia network, which has been implicated in modulation of cognitive, motor, and reward-related functions. This study specifically investigated CB-BG resting state networks associated with cognitive and motor circuits in 138 participants between the ages of 35 and 86, grouped by early middle age, late middle age, and older adulthood. Participants completed motor and cognitive batteries then underwent 26 minutes of resting state functional connectivity magnetic resonance imaging (fcMRI). This data was analyzed using the CONN toolbox with an ROI-ROI approach targeting the CB-BG circuit. Results: We replicated our previously mapped cognitive subcircuits, with Crus I and II of the cerebellum connected to VSs and DC of the basal ganglia. Similarly we have replicated the motor subcircuits, showing Lobules V and VI of the cerebellum functionally connected with the dorsal rostral and dorsal caudal putamen in the basal ganglia. Additionally, we found higher functional connectivity in early middle age compared to older adulthood, with no differences in late middle age. Analysis of connectivity associations with reaction times in the Stroop task showed differential relationships in early middle aged adults compared to older adults. When investigating sex, higher functional connectivity was seen in females compared to males. Further analysis into hormone influences revealed higher connectivity associated with progesterone levels for females compared to males. Conclusion: Here, we have shown that the CB-BG resting state network can be parsed into motor and cognitive subcircuits. This network seems to be sensitive to progesterone levels and is associated with cognitive performance. This expanded investigation of CB-BG networks represents an essential new step in increasing our knowledge of age-related cognitive and motor declines, and further investigation of this circuit in Alzheimer's Disease or other dementias is warranted.

Funding: National Institute on Aging Grants R01AG064010 and R01AG064010-S1; COI: None reported.

Poster # 111: EVALUATING THE RELATIONSHIP BETWEEN PRE-DEATH GRIEF AND NUCLEUS ACCUMBENS ACTIVATION FOR CARE PARTNERS OF PERSONS WITH DEMENTIA: AN FMRI STUDY

Presenting author: Anna McClendon (Texas A&M UHSC)

Background: By 2050, an estimated 139 million people worldwide will be living with Alzheimer's Disease Related Dementias (AD/ADRD). With approximately six care partners per person, this translates to nearly 800 million individuals providing care. Many of these care partners experience pre-death grief, which is grief that occurs while the person with AD/ADRD is still alive. However, the nature and neural mechanisms underlying pre-death grief remain poorly understood. Previous studies of other grief experiences, such as prolonged grief disorder, have shown that the nucleus accumbens (NA) is activated in individuals with higher grief intensity. This study investigates whether a similar relationship exists between pre-death grief and NA activation. Method: 11 care partners of persons with AD/ADRD and 11 control participants engaged in a 35-minute fMRI study. Participants provided six photographs of the person with AD/ADRD and another family member. A database was used to select three more images of a stranger that were matched to the person with AD/ADRD for sex, age, and race. Ten grief-related words primarily derived from previous interviews with bereaved individuals were selected. These photos and words were made into 60 composites, each consisting of one picture (care recipient, family member, or stranger) and one word (grief-related or neutral) for a 3 (picture groups) x 2 (grief vs neutral words) factorial design. The 60 composites were presented in a fMRI and ROI analysis was conducted on the NA of the brain. Result: Utilizing maximum voxel neighborhood analysis bilateral significant differences were found, with care partners evidencing greater NA activation compared to the control group (t(21) = 3.018; p =.006). Also, care partners experienced significant more activation in the NA when shown a picture of the person with AD/ADRD than a picture of the family member (t(10) = 4.07; p<.001). Conclusion: These findings offer preliminary evidence of distinct neural patterns among care partners during a grief-elicitation task. Care partners experiencing pre-death grief demonstrate neurobiological similarities to those with prolonged grief disorder. Future research should examine how pre-death grief evolves over time to better understand this relationship.

Funding: Center for Excellence for Translational Neuroscience and Therapeutics ; COI: None

Poster # 112: THE CEREBELLUM AS A SOURCE OF SCAFFOLDING AND RESILIENCE IN AGING AND ALZHEIMER'S DISEASE: MULTI-MODAL NEUROIMAGING EVIDENCE

Presenting author: Jessica Bernard, PhD (Texas A&M UHSC)

Background: As the population ages, so to does the number of individuals with Alzheimer's Disease (AD) and related dementias. Few states are as impacted as Texas, which is number three in the nation for number of cases and number two for the number of deaths. Though we have made a great deal of advances in our understanding of AD, we still lack effective treatments and need novel targets of intervention and remediation. The cerebellum is one such target. The human cerebellum is a processing powerhouse, and while it is not a primary site of pathology in AD, it has been conceptualized as a key source of scaffolding and resilience for function in AD. Methods: Here, I present an overview of key data highlighting cerebellar scaffolding in cognitively normal older adults, but also in AD and mild cognitive impairment. Data from a large longitudinal study of adults (n=136 at baseline with complete data, ages 35-86 years at baseline) encompassing multimodal neuroimaging and an intensive behavioral phenotyping will be leveraged along with secondary analyses from the Cambridge Center for Ageing and Neuroscience and the Alzheimer's Disease Neuroimaging Initiative. Results: Across samples we demonstrate that regional cerebellar structure and resting state connectivity are associated with age, such that older age is typically related to poorer neuroimaging metrics. Further, both resting state connectivity and lobular cerebellar volume are associated with cognitive and motor behavior in patterns that are specific and consistent with the known cerebellar functional topography. We also have evidence of longitudinal prediction wherein baseline cerebellar measures predict function over time. Brain-behavior associations in AD suggest differential engagement of cerebello-cortical resources. Conclusion: Convergent results across datasets suggest that maintaining cerebellar resources is associated with better behavior, though resources are differentially engaged in AD. Together, this points to the cerebellum as a source of resilience such that those with better cerebellar metrics maintain function. A lack of cerebellar resources however may also mark vulnerability to decline in AD, as individuals are less able to engage these resources which may compensate for the cortex in the face of disease pathology.

Funding: R01AG064010 WoodNext Foundation; COI: None

Poster # 113: EVALUATING PRE-DEATH GRIEF AND NEURAL SIGNATURES OF PAIN FOR CARE PARTNERS OF PERSONS LIVING WITH DEMENTIA: AN FMRI STUDY

Presenting author: Zoe Friedman, BS (Texas Tech HSC)

Background: More than 57 million people worldwide have Alzheimer's Disease and Alzheimer's Disease Related Dementias (AD/ADRD). Individuals with AD/ADRD usually live with the disease for years, placing tremendous burden on care partners. A large body of literature has examined caregiver experiences of burden and pre-death grief. However, there is a lack of understanding of the neurobiological mechanism of pre-death grief. The current study investigates the neurobiological underpinnings of pre-death grief in care partners, examining the relationship between pre-death grief and activation in pain regions in the brain during an fMRI grief elicitation task. Method: Participants provided three photographs of the person with AD/ADRD and three of another family member. Three images of a stranger, matched to the person with AD/ADRD for sex, age, and race, were selected. Ten grief-related words from interviews with bereaved individuals were selected. 60 composites with one picture (care recipient, family member, or stranger) and one word (grief-related or neutral) were created for a 3 (picture groups) x 2 (grief vs neutral words) factorial design. The 60 composites were presented in an fMRI and ROI analysis was conducted. This study enrolled 11 care partners of persons with AD/ADRD and 11 control participants. Result: Using maximum voxel neighborhood analysis, care partners showed bilateral, significant differences and increased amygdala activation compared to controls (t(21) = 3.91; p <.001). When viewing the same contrast that included all words (neutral; grief), similarly significant patterns were found, with care partners evidencing greater amygdala activation compared to controls (t(21)= 2.24; p = .04). Also, a significant difference was found in the left dACC, with care partners evidencing greater activation compared to controls (t(21) = 4.80; p <.001). Lastly, there were significant negative correlations in the left (rho=-.60; p = .05) and right (rho=-.60; p = .05) dACC, indicating those with higher pre-death grief showed lower activation in these regions while viewing images of their care recipient compared to strangers. Conclusion: This is the first study to identify the relationship between pre-death grief and pain-related neural networks. These preliminary findings suggest regions of the brain implicated in the grief process.

Funding: Center for Excellence for Translational Neuroscience and Therapeutics; Medical Student Research Summer Program. ; COI: None

Poster # 114: BIOENERGETIC VULNERABILITY IN HIPPOCAMPAL MITOCHONDRIAL METABOLIC ALTERATIONS OF A VCD-INDUCED MOUSE MODEL OF MENOPAUSE.

Presenting author: Eul Hyun Suh (UNTHSC)

Background: The menopausal transition involves a decline in ovarian hormones, increasing women's vulnerability to cognitive decline and neurodegenerative diseases such as Alzheimer's. Estrogen is essential for maintaining neuronal energy balance and its loss disrupts brain glucose metabolism, causing bioenergetic instability and metabolic stress in the hippocampus of a key region for learning and memory. Identifying these early metabolic changes is critical to understanding menopause-related brain aging. This objective of this study to investigate menopause-associated alterations in brain metabolism, structure, and mitochondrial function in a VCD-induced mouse model. Methods Menopause was induced in female C57BL/6 mice by intraperitoneal injection of VCD to model gradual ovarian follicle loss. Menopausal status was verified by vaginal cytology and serum levels of estrogen and FSH. Brain structure was evaluated using 7T T₂-weighted MRI, and hippocampal subfield volumes were quantified by multi-atlas segmentation. Mitochondrial metabolism in the hippocampal CA3 region was analyzed using the Seahorse XFe24 Cell Mito Stress Test to measure oxygen consumption, basal, ATP-linked, and maximal respiration. In addition, ¹³C NMR isotopomer analysis following [U-¹³C]glucose administration assessed oxidative versus glycolytic carbon flux, revealing early metabolic changes linked to ovarian hormone depletion. Statistical analysis was performed using one-way ANOVA (p < 0.05). Results Vaginal smear analysis confirmed the reproductive stages across groups and the menopause group showed significantly lower estrogen and higher FSH levels than the young group (p < 0.001). Hippocampal mitochondrial analysis revealed a marked reduction in basal respiration in menopausal mice (60.0 ± 37.37 pmol/min) compared to perimenopausal mice (108.62 ± 54.17 pmol/min; p = 0.039), indicating impaired oxidative metabolism. ATP-linked respiration was also significantly reduced (6.53 ± 6.97 vs. 19.85 ± 15.22 pmol/min; p = 0.02). Additionally, significant decreases in dorsal, ventral, and total hippocampal volumes were observed in the menopause group compared with perimenopausal controls. Conclusion This study shows that menopause related hormone changes leads to metabolic alternation in the hippocampal CA3 region, accompanied by structural atrophy. The resulting bioenergetic decline reveals a connection between menopause and neurodegenerative risk. In vivo assessment of hippocampal metabolism may serve as a noninvasive biomarker for early detection and intervention in menopause-related brain dysfunction.

Funding: Pharmacology and Neuroscience Pilot Grant from UNT Health at Fort Worth. ; COI: NA

Poster # 115: GLOBAL SCALE-FREE BRAIN ACTIVITY IN EYES-OPEN AND EYESCLOSED RESTING STATES AND ITS ASSOCIATION WITH COGNITIVE PERFORMANCE IN AGING

Presenting author: Frigyes Samuel Racz, MD, PhD (UT Austin)

Background: Converging evidence indicates that the spectral slope of neural recordings such as electroencephalography (EEG) - reduces with aging. This phenomenon is hypothesized to reflect an increase in the ratio of excitatory and inhibitory (E/I) activity and has been proposed as a neural mechanism associated with age-related cognitive decline. Therefore, it is of interest to investigate if the spectral slope is associated with cognitive performance and thus could be utilized as a potential neural signature of cognitive aging. To date, there is a lack of studies addressing how spectral slope estimates obtained from both eyes-closed (EC) and eyes-open (EO) resting states might relate to cognitive performance in the aging population. METHODS: We analyzed resting-state EEG recordings of 19 older (aged 60+, HE) and 24 younger (aged 18-40, HY) healthy volunteers collected in EC and EO state. Global spectral slopes (β) were estimated from the 2-22.5 Hz regime after separating scale-free from oscillatory components in the power spectrum to avoid the biasing effect of oscillatory peaks. Cognitive performance was assessed using seven tasks of the Cambridge Neuropsychological Test Automated Battery, and the association between β and cognitive performance scores was assessed via regression analysis. RESULTS: Reduced β was observed in HE in contrast to HY in both states. While the HY group exhibited an increase in β when transitioning to EO from EC, this could not be observed in HE. A strong, anticorrelated relationship between β in EO and performance in a rapid visual information processing task was observed only in the HE but not in the HY group. CONCLUSION: Both between- (HE vs. HY) and within-group differences (EO vs. EC) in β indicate a change in E/I ratio with aging, although age as a continuous variable was not predictive of β. The inverse relationship of β with task performance suggests that this pattern might reflect a compensatory mechanism of the brain to maintain cognition with age. Directions for future research are discussed.

Funding: This study was partially funded by the Coleman Fung Foundation and the Charlie Sinclair Foundation.; COI: The authors declare no real or perceived conflicts of interest.

Poster # 116: ENHANCING COGNITION IN OLDER ADULTS VIA ACCELERATED PREFRONTAL THETA-BURST STIMULATION

Presenting author: Minsu Zhang, MS (UT Austin)

Background: Intermittent theta-burst stimulation (iTBS) over the left dorsolateral prefrontal cortex (DLPFC) has shown promise in improving cognitive functions in older adults, both with and without cognitive impairment. Accelerated iTBS (aiTBS), a recent variant, has demonstrated efficacy in treating depression. This pilot study investigates whether (1) aiTBS applied to the DLPFC can improve cognitive performance in older adults with cognitive deficits, and (2) whether contingent negative variation (CNV) - an EEG biomarker of cognitive control - changes over the course of the intervention. Method: Two older adults (ages: 69, 70; one female) completed eight visits over two weeks. During visits 2-7, participants received two daily sessions of 1800-pulse iTBS. Cognitive control was assessed using a custom designed go/no-go task with EEG recordings at the start and the end of each visit. The task incorporated visual distractions and required delayed responses to maximize CNV amplitude. Neuropsychological assessments were conducted at baseline (PRE, visit 1) and post-intervention (POST, visit 8), including:
Dimensional Change Card Sort (DCCS), Face-Name Association Memory Exam (FNAME), Flanker Inhibitory Control and Attention, Picture Sequence Memory (PSM), Rey Auditory Verbal Learning Test (RAVLT), RAVLT Delay, Hamilton Depression Rating Scale (HAMD), and Perseverative Thinking Questionnaire (PTQ). Result: Average age-adjusted score improvements were observed in DCCS (+6.3%), FNAME Delay (+24.6%), Flanker (+1.6%), PSM (+18.5%), RAVLT (+8.2%), RAVLT Delay (+14.8%). HAMD scores decreased from 2 to 1.5, and PTQ scores from 14 to 12, indicating slight reductions in depressive symptoms and repetitive negative thinking. Event-related potential and multivariate pattern analyses revealed a 13.7% increase in CNV amplitude and a 6.3% improvement in classification accuracy. These findings should be interpreted cautiously due to the small sample size and potential retest effects, particularly in tests lacking alternate forms (e.g., FNAME Delay, RAVLT). Conclusion: Preliminary results suggest that aiTBS targeting the DLPFC may (1) enhance cognitive function and mood in older adults with cognitive impairment, and (2) and 2) produce a more discriminant CNV biomarker potentially associated with behavioral improvements. This pilot study provides foundational evidence for future investigations into non-invasive neuromodulation techniques for cognitive enhancement.

Funding: TARCC Investigator Initiated Research Grant; COI: None.

Poster # 117: SPATIOTEMPORAL HETEROGENEITY OF ALZHEIMER'S DISEASE-RELATED FUNCTIONAL BRAIN NETWORK REORGANIZATION

Presenting author: Roy Massett, BA (University of Texas at Dallas)

Background: Alzheimer's disease (AD) is a highly heterogeneous condition both in terms of the distribution of pathology deposition and clinical manifestation. The organization of functional brain networks is related to AD pathology and cognitive impairment, with recent work showing that higher dementia severity is related to the desegregation of large-scale functional brain networks (Zhang et al., 2023). However, the spatiotemporal heterogeneity of these changes in network segregation and how they contribute to different cognitive deficit profiles remains uncertain. To contribute to this question, we applied a disease progression model to system-level measures of network organization. Method: We included 345 resting-state functional magnetic resonance imaging (fMRI) scans from amyloid positive individuals enrolled in the Alzheimer's Disease Neuroimaging Initiative. The correlations of fMRI resting-state time series between nodes were used to construct brain networks (Chan et al., 2014). Nodes were assigned to a functional system based on a previously defined functional atlas (Power et al., 2011). Network segregation was calculated for each brain system as measures of system-level organization. The SuStaIn algorithm was used to simultaneously stage and subtype subjects according to their patterns of network segregation (Young et al., 2018). Predicted subtypes and stages of subjects were further analyzed using linear regression models. Result: SuStaIn identified two subtypes of alterations in system-level network segregation, which were upheld in k-fold cross validation. One subtype showed lower segregation in sensory-motor systems, including the visual and auditory systems, while the other showed decreases primarily in segregation of association systems including the default mode and task control systems. Further, we observed worse pathological and cognitive measures in the association system subtype. The only measure analyzed that contradicted this pattern was cortical thickness, which exhibited a significant relationship with stage. Conclusion: Resting-state fMRI signals can be used to identify dissociable patterns of AD-related brain network desegregation relating to different profiles of brain network dysfunction and cognitive impairment. These findings begin to describe the spatiotemporal heterogeneity in brain network decline, and their relation to cognitive trajectories, which is relevant not only to AD prognostics but also evaluating clinical trial outcomes in cognitively-diverse patient populations.

Funding: James S. McDonnell Foundation National Institutes of Health AG-063930 ; COI: None.

Poster # 118: CURRENT AND EMERGING TARGETS FOR GLYMPHATIC DRAINAGE AS IT RELATES TO NEURODEGENERATION: A COMPREHENSIVE REVIEW

Presenting author: Camry A'Keen, MS (UTH San Antonio)

Introduction Neurodegenerative diseases such as Alzheimer's (AD), Parkinson's (PD), Huntington's (HD), and amyotrophic lateral sclerosis (ALS) represent a major global health burden. Although progress has been made in diagnostics and symptomatic treatment, curative approaches remain elusive. Emerging evidence implicates the glymphatic system a glialdependent perivascular network for cerebrospinal and interstitial fluid exchange as a key in the clearance of neurotoxic proteins such as amyloid-beta and tau. Impairments in this system are linked to waste accumulation and neuroinflammation. Imaging and translational applications remain underdeveloped. This systematic review examines molecular targets and protein regulators relevant to glymphatic and meningeal lymphatic dysfunction, emphasizing PET-tracer potential. Method We assessed 39 peer-reviewed articles published 2014-2025, including molecular, murine, and human analysis. Articles were sourced from PubMed database using search terms "glymphatic drainage neurodegeneration". All 36 were included, while an additional 3 articles were included addressing molecular regulators, imaging modalities, or lymphatic dysfunction relevant to neurodegeneration mechanisms. Results The most frequently cited molecular marker was AQP4, used for CSF exchange, in which mislocalization was associated with impaired glymphatic flow. Upstream regulation by STAT3, acts via the JAK/STAT3 pathway, modulating AQP4 polarization. IL-33 is a crucial regulator to maintain healthy AQP4. VEGF-C and its receptor VEGFR3 were shown to regulate meningeal lymphangiogenesis and CSF outflow, especially via cervical lymph nodes. α-synuclein was exacerbated by meningeal lymphatic blockage and associated with impaired glymphatic-flow in Parkinsonian models. Immune and glial-related markers were identified -C1q, CD68, MMP-9, GFAP, and TREM2.
C1q was implicated in classical complement activation and synaptic pruning. CD68 co-localizes TSPO PET signal, suggesting TSPO PET captures broader activation. CD68 provides greater cellular specificity for activated microglia. MMP-9 contributed to BBB and perivascular disruption. GFAP, a hallmark of reactive astrocytes, was identified as a biomarker in CSF and peripheral blood relevant to traumatic-brain-injury and AD. TREM2, a microglial receptor in immune surveillance and amyloid clearance in Alzheimer's and frontotemporal-dementia. Conclusion These articles present no tracers that directly target the glymphatic system. Molecular regulators- AQP4, STAT3, IL-33, and VEGF-C give us insight into future developments. 21 targets were identified across the dataset, many of which present potential for PET imaging.

Funding: None reported.; COI: None reported.

Poster # 119: MACHINE-LEARNING DERIVED SPARE-TAU, A TAU-PET INDEX, DEMONSTRATES CORRELATION WITH BIOMARKERS OF ALZHEIMER'S DISEASE AND RELATED DEMENTIAS

Presenting author: David Haoyuan Wang, BS (UTH San Antonio)

Background: Tau pathology drives Alzheimer's disease (AD) clinical progression, but global tauPET indices like composite SUVR lack regional specificity. We developed SPARE-Tau (Spatial Pattern of Abnormality for Recognition of Tau), a machine-learning index, to summarize weighted regional tau deposition, enhancing sensitivity and precision. We then validated SPARETau with established AD CSF markers, and investigated SPARE-Tau's association with established AD plasma markers. Method: Using 1,034 ADNI participants with 18F-Flortaucipir tau-PET, SPARE-Tau was trained on 625 control and pathological participants (269 Aβ+ MCI/dementia, 356 Aβ- CN) via support vector machines and validated through leave-one-out cross-validation. Linear regression models examined SPARE-Tau associations with CSF and plasma AD biomarkers, adjusting for age, sex, and education. Analyses were stratified by βamyloid status with interaction terms testing differential associations. Spearman's rank correlation compared biomarker associations. We controlled for multiple comparisons using the BenjaminiHochberg procedure with a false discovery rate (FDR) threshold of 5%. Result: SPARE-Tau achieved AUC=0.912 [0.895,0.942] (sensitivity=0.781 [0.730,0.829], specificity=0.946 [0.922,0.969]) in classifying Aβ+ MCI/dementia. Top regions contributing to SPARE-Tau followed the Braak staging, including medial temporal lobes and limbic structures. In β-amyloidpositive participants, SPARE-Tau showed strong associations with AD biomarkers, particularly tau-related markers: CSF pTau181/Aβ42 (β=0.513 [0.414,0.612]; ρ=0.577 [0.487,0.650]) and plasma pTau217/Aβ42 (β=0.766 [0.686,0.847]; ρ=0.752 [0.697,0.800]). CSF pTau181/Aβ42 showed stronger associations than CSF pTau181 (β=0.261 [0.191,0.331]; ρ=0.439 [0.327,0.543]), but plasma pTau217/Aβ42 and pTau217 (β=0.729 [0.647,0.811]; ρ=0.735 [0.677,0.784]) were comparable to each other. Associations were consistently stronger in β-amyloid-positive participants. SPARE-Tau associated more strongly with GFAP (β=0.184 [0.115,0.252]; ρ=0.320 [0.205,0.427]) than NfL (β=0.086 [0.037,0.134]; ρ=0.221 [0.111,0.324]), p<0.001. Conclusion: SPARE-Tau demonstrates robust alignment with CSF tau while enabling plasma tau evaluation relative to PET, advancing noninvasive tau pathology monitoring.

Funding: This study was supported in part by the National Institutes of Health (NIH) grant P30AG066546 (South Texas Alzheimer's Disease Research Center) and grant numbers 5R01HL127659, 1U24AG074855, 5R01AG080821, 1R01AG085571, 5R01AG083865, and the TARCC Grants Program.; COI: J.T. has served as a speaker for EISAI and a consultant for GE Healthcare. Other authors have declared that no competing interests exist.

Poster # 120: SYNTHESIS AND CGMP VALIDATION OF [11C]ER176 FOR PET IMAGING OF NEUROINFLAMMATION

Presenting author: Devin Gondowardojo, BS (UTH San Antonio)

Background: Chronic neuroinflammation has been implicated in various neurodegenerative diseases, such as Alzheimer's disease (AD), Parkinson's disease (PD), and Traumatic brain injury (TBI), manifesting as inflammatory cell infiltration and blood-brain barrier dysfunction, causing chronic microglial activation, resulting in the release of pro-inflammatory cytokines, leading to neuronal damage and cognitive impairment. TSPO PET is a validated imaging method in detecting neuroinflammation. However, there are concerns over the use of TSPO PET with previous generation tracers due to TSPO polymorphism in humans. As a next generation TSPO PET tracer, [11C]ER176 has been recognized as a high affinity PET tracer with excellent kinetics, and minimum sensitivity towards genetic polymorphism. Herein, we present the radiosynthesis and cGMP validation of [11C]ER176 production for human use in a clinical trial evaluating stem-cell-based immune modulation (NCT05951777). Method: The radiosynthesis of [11C]ER176 was automated using a GE Healthcare TRACERlab FX2C module, with an inhouse designed reaction sequence. Cyclotron produced [11C]carbon dioxide was first catalytically converted into [11C]methane (Ni catalyst, 360º C) and reacted with iodine to obtain [11C]iodomethane . Next, N-desmethyl-ER176 precursor was methylated with [11C]iodomethane in the presence of a base (tBuOK) using a custom prepared stainless-steel loop (room temp., 5 min). Crude [11C]ER176 was then purified using high-performance liquid chromatography (HPLC) under isocratic conditions. The separated clean product was formulated into an injectable dose using sterile 10% ethanol-saline solution and sterile filtered for quality control procedures. Results: With a precursor load of 1 mg, 2.6 - 4.2 GBq of [11C]ER176 was obtained with an radiochemical yield of 9% and a molar activity of 59.9 ± 14.5 GBq/μmol at the end of synthesis (~20 minutes; n=3). The final product formulation was tested on a strict quality control criteria for human use including, pH , chemical concentration, impurity content, endotoxins, residual solvents, radionuclidic purity, half life, and sterility. Conclusion: [11C]ER176 was successfully synthesized under 21 CFR Part 212 and validated for human PET studies meeting all Food and Drug Administration and United States Pharmacopeia requirements for a PET radiopharmaceutical production and can be used in any neuroimaging study that investigate neuroinflammation including TBI, AD, and PD.

Funding: Congressionally Directed Medical Research Programs (CDMRP) Award Number: HT9425-23-1-0991 PI: Charles S Cox ; COI: None reported.

Poster # 121: QUANTIFYING VASCULAR BURDEN IN ALZHEIMER'S DISEASE: INTEGRATING MULTIPLE SVD MEASURES TO QUANTIFY VASCULAR BURDEN IN ALZHEIMER'S DISEASE: THE SPARE-SVD SCORE

Presenting author: Mengjin Dong, PhD (UTH San Antonio)

Background: Cerebral small vessel disease (SVD) is a critical contributor to cognitive impairment in Alzheimer's disease and related dementia (ADRD). Although the revised AD framework recognizes vascular (V) pathology as an essential copathology alongside β-amyloid, tau, and neurodegeneration (ATN), prior research has mainly focused on individual SVD markers in isolation, limiting understanding of their collective impact. To address this gap, we developed the Spatial Pattern of Abnormality for Recognition of SVD (SPARE-SVD) score, a machine learning-based "V" biomarker that integrates multiple SVD features to provide a comprehensive measure of vascular burden. Method: SPARE-SVD was trained using a leave-one-out strategy on DTI features and white matter lesion (WML) burden from 362 ADNI participants using a Support Vector Machine to differentiate participants with Alzheimer's disease dementia (ADD) from CN. Its effectiveness was compared with individual SVD biomarkers. Associations between SPARE-SVD and cognitive scores were examined using linear regression models, adjusting for age, sex, SPARE-AD score, CSF β-amyloid, and CSF ptau positivity. Result: SPARE-SVD achieved a balanced accuracy of 0.82 and ROC AUC of 0.85 for predicting CN/ADD, outperforming total WML or free-water-corrected FA alone. SPARE-SVD was significantly associated with poorer cognitive performance in memory, executive function, and language performance. After stratifying by β-amyloid and tau positivity, these associations remain significant only in the SPARE-SVD positive (V+, SPARE-SVD>0) group, but not in the Vgroup. Conclusion: The SPARE-SVD score captures the aggregate impact of SVD on ADRD. By integrating various MRI features into a single interpretable index, SPARE-SVD provides a scalable approach to quantify vascular burden and its interaction with neurodegenerative processes for cognitive decline.

Funding: This study was supported in part by the National Institutes of Health (NIH) grant P30AG066546 (South Texas Alzheimer's Disease Research Center) and grant numbers 5R01HL127659, 1U24AG074855, 5R01AG080821, 1R01AG085571, 5R01AG083865, and the TARCC Grants Program.; COI: J.T. has served as a speaker for EISAI and a consultant for GE Healthcare. Other authors have declared that no competing interests exist.

Poster # 122: DECIPHERING THE ROLE OF TLR4 IN ALZHEIMER'S DISEASE THROUGH NOVEL TLR4-TARGETED IMAGING PROBES

Presenting author: Sandun Perera, PhD, Andre Aguirre (UTH San Antonio)

Background: Neuroinflammation plays a critical role in the initiation and progression of various central nervous system (CNS) disorders, including Alzheimer's disease. PET imaging has become indispensable for assessing inflammatory processes although lack specific tracers towards neuroimmune targets. Toll-like receptor 4 (TLR4) is a key mediator of neuroinflammatory signaling and represents a promising molecular target for imaging CNS inflammation. Herein, we report the development of a fluorescently labeled TLR4-specific DNA aptamer, ApTOLL as an imaging probe to interrogate TLR4 surface expression and activation without perturbing ongoing signaling. Methods: BV2 microglia and HEK-Blue-hTLR4 cells were were fixed with 4% paraformaldehyde in PBS prior to blocking with 1% BSA-PBS and treatment with imaging probes: anti-TLR4-AF594 mAb or FAM-labeled ApTOLL at 0.2-200 nM. Incubations were carried out at 4 °C or 25 °C for 24 h or 1 h, respectively. Nuclei were counterstained with DAPI, and coverslips were mounted in ProLong Gold antifade medium and cured overnight in the dark. Imaging was performed on a Zeiss LSM 980 confocal microscope equipped with Airyscan 2 super-resolution and a 63×/1.4 NA oil-immersion objective. Image analysis was conducted using Fiji. In parallel, Flow cytometry was used for quantitative validation of surface labeling. Results: In mouse BV2 microglia showed only a weak FAM-ApTOLL signal while HEK-Blue-hTLR4 cells exhibited a clear concentration-dependent increase in fluorescence, with the 200 nM FAMApTOLL sample showing approximately fourfold higher intensity relative to controls. The antiTLR4-AF594 antibody yielded a twofold increase under comparable conditions. Conclusions: The FAM-labeled ApTOLL effectively detected surface TLR4 expression in HEK-Blue-hTLR4 cells at high nanomolar concentrations, supporting its utility as a potential TLR4-targeted imaging probe. The comparable labeling efficiency to antibody-based detection further underscored its specificity and translational promise. However, the absence of detectable binding in BV2 cells and the requirement for relatively high probe concentrations indicated that additional optimization is needed to enhance sensitivity across cell lines. Future work will focus on refining aptamer secondary structure, incubation parameters, and ionic conditions (e.g., Mg²⁺ concentration) to improve binding affinity and enable reliable detection of TLR4 in lowexpressing cells, ultimately advancing novel imaging probes toward in vivo imaging applications.

Funding: UTH San Antonio; COI: None

Poster # 123: NOVEL BLOOD POOL IMAGING ANALYSIS DEMONSTRATES SIGNIFICANT CORRELATION OF NASAL BLOOD POOL WITH RISK FACTORS FOR ALZHEIMER'S DISEASE

Presenting author: William Phillips, MD (UTH San Antonio)

Background: A novel imaging analysis methodology was developed to assess nasal blood pool due to nasal turbinate vasodilation in patients with known risk factors for Alzheimer's disease composed of metabolic syndrome entities i.e. elevated BMI, diabetes, hypertension and sleep apnea. Nasal turbinates are important due to their transmission of CSF lymphatic drainage from the brain. Nasal lymphatics clear toxic brain proteins such as amyloid that are drained from the recently discovered glymphatic/meningeal brain lymphatic system. Method: A retrospective study was performed on 200 patients having whole-body blood pool scans referred to the Nuclear Medicine department over a 3-year period. Relative nasal vasodilation was assessed with max pixel ratios of the nasal region/ heart ROI maximum (max) pixel ratios. These ratios were correlated with clinical conditions, including obesity, diabetes, hypertension, and sleep apnea. Results: A direct significant association of nose to heart max ratio (NHMR) was found with diabetes, sleep apnea, and hypertension with an increase in the ratio of +0.10 (P= 0.002), +0.13 (P= 0.0002), +0.08 (P= 0.0123), respectively. Examples of typical whole body blood pool images will be shown in typical patients with no metabolic syndrome Alzheimer's risk factors as compared to patients with only one and those with multiple Alzheimer's metabolic syndrome risk factors which include elevated BMI, hypertension, diabetes and sleep apnea. Conclusion: Nasal vasodilation is significantly increased in patients with risk factors for Alzheimer's disease. The authors propose that nasal turbinate vasodilation, due to increase parasympathetic activity of the nasal turbinates, contrasting with known increased sympathetic activity of the cardiovascular system in these same patients with Alzheimer's risk factors. The authors propose that nasal vasodilation interferes with CSF lymphatic brain waste clearance. Nasal lymphatic obstruction provides an explanation for other previously reported MRI brain imaging abnormalities in Alzheimer's disease of enlarged perivascular spaces and lower indices of diffusion tensor imaging-analysis along the perivascular space (DTI-ALPS) observed in Alzheimer's disease. These MRI findings suggest impaired CSF movement through the brain. If further studies support these findings, the nasal turbinates and the increased vasodilating parasympathetic activity of the nasal turbinates could lead to novel therapeutic approaches for Alzheimer's.

Funding: None; COI: None

Poster # 124: AGE-DEPENDENT VASCULAR AMYLOID PATHOLOGY INITIATES TAUOPATHY AND NEUROINFLAMMATION IN THE 3XTG-AD MOUSE MODEL

Presenting author: Pedro Pablo Martinez Cuevas, MSc (UTMB)

Background: Cerebrovascular dysfunction, particularly the accumulation of amyloid-beta (Aβ) protein in brain capillaries, known as cerebral amyloid angiopathy (CAA), is increasingly recognized as a significant factor in Alzheimer's disease (AD) progression. While the 3xTg-AD mouse model effectively mimics many AD features, the full extent of vascular alterations and their relevance to human AD remains to be thoroughly investigated. Method: Our study aimed to evaluate the age-related progression of Aβ vascular deposition, tau pathology, and neuroinflammation in the dentate gyrus of the hippocampus in 3xTg-AD mice, comparing them to wild-type controls at 3, 6, 9, and 12 months of age. Result: We observed that vascular Aβ deposition began as early as 6 months and persisted through 12 months. A trend towards increased total tau pathology was noted by 12 months, and significant neuroinflammation became evident after 9 months. These findings suggest that early Aβ vascular pathology likely contributes to neurovascular dysfunction. This dysfunction appears to initiate a detrimental cycle, leading to further Aβ accumulation, increased phosphorylated tau, and heightened neuroinflammation, thereby exacerbating overall AD pathology. Conclusion: Our data underscore vascular dysfunction as a critical contributor to AD onset and progression. Further evaluation, especially in older mouse models (16-20 months), is essential to fully comprehend AD pathology and identify crucial therapeutic windows for intervention.

Funding: Startup package of NVRCL and EC.; COI: NA

Poster # 125: FOCUSED ULTRASOUND MEDIATED DELIVERY OF NANOBODIES ENHANCES HUMAN TAU DETECTION IN-VIVO

Presenting author: Deepshikha Bhardwaj, PhD (UTSW)

Background: Limited permeability of the blood-brain barrier (BBB) remains a major obstacle in the diagnosis and treatment of Alzheimer's disease (AD). More than 98% of small molecules and nearly all large therapeutics are unable to cross the BBB in effective concentrations. As a result, noninvasive and targeted techniques such as Focused Ultrasound (FUS) are being actively explored to transiently and reversibly open the BBB. FUS is an incisionless, image-guided procedure that can modulate BBB permeability in localized brain regions with high spatial precision. The inability to detect AD during its pre-clinical phase, before widespread neuronal loss-severely limits the efficacy of emerging diagnostic and therapeutic strategies. Early detection of tau pathology, a hallmark of AD and related tauopathies, is therefore critical. This study investigates the feasibility of using FUS to deliver Tau-specific nanobodies (Tau-VHH) across the BBB in a preclinical mouse model, enabling early detection of tau aggregates that would otherwise remain inaccessible due to the intact BBB. Methods: Tau-VHH nanobodies were fluorescently labeled with Alexa Fluor 488 (AF488) Maleimide, and successful conjugation was confirmed using Matrix-Assisted Laser Desorption/Ionization Time-of-Flight (MALDI-TOF) mass spectrometry. P301S transgenic mice, which exhibit tau aggregation, and their wild-type (WT) littermates were subjected to FUS targeting at right hippocampi to induce localized BBB opening. The extent of BBB disruption was confirmed by magnetic resonance imaging (MRI). Subsequently, 1 mg of AF488-labeled Tau-VHH (510 nM) was administered via intravenous injection, and mice were sacrificed at varied post-injection time points. After cardiac perfusion, brains were extracted, sectioned, and imaged using tissue cytometry to assess nanobody localization. Results: Compared to WT littermates, P301S mice showed markedly higher hippocampal Tau-VHH fluorescence. In P301S animals, FUS-targeted hippocampi exhibited a significantly greater fluorescent signal than the contralateral, non-sonicated side, 90 minutes postinjection (p < 0.05). This indicates successful BBB opening and selective binding of Tau-VHH to pathological tau aggregates. Conclusion: FUS-mediated BBB opening enables targeted delivery of Tau-specific nanobodies into the hippocampus, facilitating in vivo detection of tau pathology in P301S mice. This approach holds promise for early, noninvasive identification of tauopathies and the future development of image-guided theranostic strategies for Alzheimer's disease.

Funding: None reported.; COI: None

Poster # 126: THE HUMAN CEREBELLUM: A NOVEL TARGET FOR COGNITIVE & MOTOR INTERVENTIONS IN AGING & ALZHEIMER'S DISEASE

Presenting author: Jerilyn James, MS (TAMU)

Background: Over 7 million Americans are living with Alzheimer's. This number is expected to increase to 13 million by 2050, as the aging population continues to grow. Emerging literature suggests that the cerebellum is impacted by Alzheimer's Disease (AD). There is a decrease in functional connectivity and changes to cerebellar volume in regions associated with the cortical areas that are most significantly affected by AD. These changes position the cerebellum as a potential target for intervention in age-related decline. Non-invasive brain stimulation is an especially promising methodology. Methods: In this study, intermittent theta burst stimulation (iTBS), a form of non-invasive brain stimulation that excites the underlying brain tissue, was combined with task-based fMRI in cognitively normal OA (adults aged 60 and older with a MOCA score of at least 27). Forty older adults will be recruited, though our preliminary analyses here represent only the subset of participants who have completed the study and been unblinded. All participants underwent active and sham cerebellar stimulation in random order, using a double-blind crossover design. Following stimulation, participants completed a 90-minute scan that involved a resting-state functional scan and several behavioral tasks. Participants completed an n-back working memory task, an explicit sequence learning task, and the Stroop task. Results: A within-subjects comparison of sham stimulation relative to active iTBS was conducted to analyze the effects of the stimulation. A block-by-block analysis of the sequence learning task reveals that accuracy improved over the course of the task following active stimulation. Specifically, accuracy was noticeably higher in the last three blocks compared to the first three blocks. Alternatively, there was little to no improvement in accuracy following sham stimulation, suggesting that iTBS facilitated learning. Conclusion: These results suggest that non-invasive cerebellar stimulation may improve visuomotor task learning and working memory in OA. These promising preliminary findings highlight the potential of cerebellar stimulation for ameliorating functional declines associated with aging, and future work will target Alzheimer's disease and atrisk groups.

Funding: WoodNext Foundation; COI: None reported.

Poster # 127: MILD BEHAVIORAL IMPAIRMENT AS AN EARLY STAGE OF ALZHEIMER'S DISEASE: EVIDENCE FROM BIOMARKER PROFILES IN THE TARCC COHORT

Presenting author: Natalia Pessoa Rocha, PhD (UTHealth Houston)

Background: Mild behavioral impairment (MBI) describes the later-life onset of persistent neuropsychiatric symptoms (NPS) that may precede or accompany cognitive decline. MBI is hypothesized to represent a prodromal stage of Alzheimer's disease (AD), reflecting underlying neurodegenerative changes before overt cognitive symptoms emerge. This study aimed to validate MBI as an early stage within the AD continuum by determining whether individuals with MBI exhibit intermediate biomarker profiles compared with cognitively normal and AD groups. Methods: Data from the Texas Alzheimer's Research and Care Consortium (TARCC) were analyzed to categorize participants into four groups: MBI, mild cognitive impairment (MCI) without MBI, AD, and controls. MBI classification followed published algorithms aligned with ISTAART-AA criteria, using ten domains from the Neuropsychiatric Inventory Questionnaire (NPI-Q) to calculate a total MBI symptom severity score (range 0-30). Participants were classified as MBI-positive if their total symptom severity score was ≥ 1 across two consecutive visits prior to a dementia diagnosis. Group comparisons were conducted for demographics, APOE4 carrier status, and plasma biomarkers-YKL-40, glial fibrillary acidic protein (GFAP), neurofilament light chain (NfL), and CD14-measured using Quanterix assays. Proteomic analyses were performed with Meso Scale Discovery (MSD), an electrochemiluminescence-based ELISA platform. Results: Among 3,666 participants, 298 were classified as MBI, 606 as MCI (without MBI), 1,442 as controls, and 1,320 as AD. AD participants were older than other groups, and APOE4 prevalence followed a graded pattern (AD > MCI ≈ MBI > controls). Plasma biomarker analyses revealed significant group differences in YKL-40, GFAP, NfL, and CD14. The MBI group exhibited intermediate YKL-40 and GFAP levels between controls and MCI, while NfL and CD14 levels were similar to controls but lower than AD. Proteomic analyses showed that MBI participants occupied an intermediate biological position between controls and AD, closely resembling MCI profiles. Conclusion: Similar to MCI, MBI is associated with intermediate biomarker alterations between cognitively normal and AD groups. These findings provide biological evidence supporting MBI as a prodromal stage of AD and underscore its potential utility for early detection and intervention in AD.

Funding: TARCC Alzheimer's Association; COI: The authors declare no conflicts of interest.

Poster # 128: VARIANCE PARTITIONING AND OVERLAP CORRECTION IN PSYCHEDELIC THERAPY RESEARCH: AN UMBRELLA META-ANALYTIC STUDY

Presenting author: Alissa A. Lopez, MSc (UTRGV-SOM)

Background: Psychedelics such as psilocybin, MDMA, LSD, and ayahuasca are gaining renewed interest for the treatment of psychiatric disorders including depression, anxiety, post-traumatic stress disorder (PTSD), and substance-use disorders. Several meta-analyses suggest meaningful symptom improvement, but the evidence base is small and overlapping, potentially inflating precision and limiting generalizability. A comprehensive umbrella review is needed to compare therapeutic effects across compounds and conditions while correcting for redundancy and smallsample bias. Methods: We conducted an umbrella review of meta-analyses assessing psychedelic-assisted psychotherapy for psychiatric disorders. Eligible reviews reported quantitative effect sizes or sufficient data for conversion to standardized mean differences. To address small-sample bias, standardized mean differences were adjusted to Hedges' g using the correction factor J. Overlap among meta-analyses was quantified using the Jaccard index and redundancy ratio, visualized via a heatmap. A two-level mixed-effects model (effect sizes nested within original studies and meta-analyses) estimated pooled effects and variance partitioning. Quality and heterogeneity were assessed using AMSTAR-2, and I² statistics. Results: Nineteen meta-analyses covering four psychedelic compounds and five psychiatric disorders were identified. Substantial overlap existed-each unique trial appeared in nearly two meta-analyses on average. The pooled effect size for psilocybin trials was Hedges' g = −1.34 (95% CI −1.80 to −0.89, p < 0.001), consistent with significant symptom reduction versus control. Variance partitioning indicated that heterogeneity was largely attributable to differences among metaanalyses, with minimal additional variance from original studies. Conclusion: Psychedelicassisted therapy demonstrates robust efficacy across psychiatric disorders, particularly psilocybin in depression and MDMA in PTSD. However, evidence overlap and small sample sizes limit interpretability. Correcting for redundancy through Jaccard analysis and hierarchical modeling improves precision and transparency. Future syntheses should preregister overlap strategies, deduplicate studies, and pursue individual-participant meta-analyses to enhance reliability.

Funding: N/A - This is a secondary analysis of literature; COI: No conflict of interest

Poster # 129: NEUROPSYCHOLOGICAL IMPROVEMENTS IN MILD COGNITIVE IMPAIRMENT FOLLOWING NEUROSTIMULATION

Presenting author: Hannah Cabrera, MS (UTSW)

Background: Current treatments for neurodegenerative disorders have not shown efficacy in improving cognition. Thus, new treatments are needed that can alleviate the cognitive deficits that occur. We aimed to examine if high-definition transcranial direct current stimulation (HD-tDCS) improves neuropsychological functioning in patients with amnestic Mild Cognitive Impairment (aMCI). Method: 22 participants underwent 10 daily sessions of HD-tDCS over the medial prefrontal cortex (mPFC) for 20 minutes. Participants were randomized into 1mA (n=10) or sham (n=12) HD-tDCS. Neuropsychological assessments were administered at baseline and immediately following the last session. An Analysis of Covariance evaluated for post-treatment differences between groups, controlling for baseline scores. Result: Statistically significant improvements were observed following active HD-tDCS compared to sham (Adj MDiff=6.868, SE=2.359) for verbal memory (p=0.009, d=1.254). No other statistically significant differences were observed. Examination of effect sizes showed modest differences suggestive of executive function enhancement for active HD-tDCS relative to sham (Adj MDiff=2.301, d=0.489), but improvements in visual memory (Adj MDiff=4.878, d=0.788) and confrontation naming (Adj MDiff=7.657, d=0.897) favored sham over active HD-tDCS. Conclusion: HD-tDCS over the mPFC was associated with improved verbal memory, an ability commonly impaired in aMCI with minimal test-retest effects. Future research is needed to delineate if effects from test-retest, the low electric current inherent in sham protocols, or certain clinical features might underlie the modest gains in the sham group, obscuring interpretation about other cognitive skills. Also, a treatment mechanism remains unclear and needs further investigation to assess for HD-tDCS related neural changes to understand efficacy.

Funding: TARCC - Public; COI: None to report

Poster # 130: EVALUATING BIOPSYCHOSOCIAL FACTORS TO UNDERSTAND COGNITIVE FUNCTIONING IN FAMILY CAREGIVERS OF PERSONS WITH ALZHEIMER'S DISEASE AND RELATED DEMENTIAS

Presenting author: Emily Post, BS (Texas Tech HSC)

Background: The number of older adults and the prevalence of Alzheimer's Disease and Alzheimer's Disease-related dementias (AD/ADRD) are increasing, as is the number of care partners (i.e., unpaid family members) that are predominantly older adults. The demands of responsibilities on care partners can negatively affect their own physical, psychological, and social functioning, increasing their risk for cognitive decline. Yet, these risk and protective factors have not been well investigated in this population. This study investigated the effects of physiological and social factors on care partners' neurocognitive functioning. Method: Forty-two unpaid care partners (Mage = 69.40, SDage = 11.58) providing care to a relative with AD/ADRD completed psychological measures (e.g., Oslo Social Support, Social Support Inventory), neuropsychological tests (i.e., Brief Visuospatial Memory Test- Revised (BVMT-R); California Verbal Learning Test-Third Edition (CVLT-3); Trail Making Test (TMT-Delta); Wechsler Abbreviated Scale of Intelligence, Second Edition (WASI-II); Wisconsin Card Sorting Test (WCST)) and wore a Fitbit watch for two weeks measuring daily step count, sleep, and heart rate variability. Result: A Kendall's tau correlation analysis revealed a significant negative correlation between sleep and TMT (rτ = -.46, p = .04) and a trending positive correlation between daily step count and BVMT-R (rτ = .38, p = .075). A multiple regression found social support (β = .41, p = .049) as a significant predictor of BVMT-R performance (F(2,27) = 2.85, p = .075, R2 =.17). As social support increased, visuospatial memory ability increased while controlling for social network size (β = -.41, p = .05). Social support (β = -.43, p = .04) was also a significant predictor of TMT-Delta (F(2,29) = 2.87, p = .07, R2 =.17). As social support increased, processing speed improved while controlling for social network size (β = .05, p = .81). Conclusion: These preliminary results provide novel findings regarding the impact of sleep and social support on care partners' cognitive functioning. Interventions targeting improvements in sleep and functional social support for care partners of persons with AD/ADRD may be particularly beneficial in protecting against declines in processing speed and visuospatial memory.

Funding: None; COI: None

Poster # 131: RELATIONSHIPS BETWEEN COGNITION AND NUTRITIONAL INDICES AMONG AGING RURAL WEST TEXANS: A PROJECT FRONTIER STUDY

Presenting author: Shivani Challakonda, BS (Texas Tech HSC)

Background: Suboptimal nutrition is a recognized risk factor for Alzheimer's disease (AD). The Nutritional Risk Index (NRI) incorporates serum albumin and body weight, making it more sensitive to obesity or undernutrition, while Prognostic Nutritional Index (PNI) incorporates serum albumin and lymphocyte count, reflecting immune and inflammatory status. Previous studies suggested that lower NRI and PNI values are associated with increased risk of cognitive decline. This study examined relationships between these indices and cognitive performance measures in a rural cohort from Project FRONTIER (PF) in West Texas. We hypothesized that lower NRI and PNI values correlated with poorer cognitive outcomes, identifying individuals at greater risk of cognitive impairment and enabling future nutrition-based interventions. Methods Data from 1,287 PF participants were analyzed (age 58.8±12.2 years, 47% female, 58% Hispanic). Spearman correlation analyses were conducted using IBM SPSS to examine associations among NRI, PNI, and performance on the Mini-Mental State Examination (MMSE; 7 domains), the Executive Interview (EXIT25; 7 domains), and the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS, 5 domains). Results NRI and PNI were negatively correlated with each other. Only one MMSE domain, recall, correlated positively with NRI (ρ=0.087, p=0.027) and PNI (ρ=0.086, p=0.027). For the EXIT25, PNI was negatively correlated with Initiation/Generativity (Exit_C; ρ=-0.075, p=0.008) and positively correlated with Inhibition/Interference Control (Exit_G; ρ=0.057, p=0.046) and Motor Programming/Sequencing (Exit_R; ρ=0.059, p=0.036). NRI was negatively correlated with Automaticity/Utilization Behavior (Exit_H; ρ=-0.066, p=0.020), Inhibition/Motor Control (Exit_S; ρ =-0.068, p=0.017), and Abstract Reasoning/Complex Command (Exit_U; ρ=-0.060, p=0.034). No significant associations were found with total RBANS scores. However, NRI showed a negative correlation with the Visuospatial/Constructional domain (ρ=-0.109, p<0.001) and PNI showed a positive correlation with the same domain (ρ=0.062, p=0.029). NRI was also negatively correlated with the Attention domain (ρ=-0.113, p <0.001). Conclusion NRI and PNI demonstrated significant but distinct relationships with specific cognitive domains, suggesting that nutritional status influences cognition through multiple neural pathways. Their inverse relationship may reflect how these indices capture different nutritional risk profiles associated with cognitive decline, overnutrition and metabolic risk compared with inflammatory or immune dysregulation, particularly within rural populations.

Funding: Medical Student Summer Research Program funded by the TTUHSC SOM; COI: None

Poster # 132: THE ROLE OF CARDIOVASCULAR RISK FACTORS AND APATHY ON EXECUTIVE FUNCTIONING IN A MULTI-ETHNIC COMMUNITY-DWELLING SAMPLE

Presenting author: Santiago Sanchez Cabarcas (University of Houston)

Background: Cardiovascular risk factors (CVRFs) contribute to biological changes that are associated with mood symptoms and cognitive dysfunction. Apathy frequently co-occurs with CVRFs and may reflect underlying neurovascular changes. We aimed to evaluate the relative contribution of CVRFs and apathy on executive function in a community-dwelling sample of adults in Texas. Method We used English-speaking baseline data of participants without history of stroke from the Health and Aging Brain Study-Health Disparities (HABS-HD; N=3,015; Mage=66±9; 19% Hispanic/Latino [H/L], 37% Non-Hispanic Black [NHB], 45% NonHispanic White [NHW]). An executive function (EF) composite score was calculated from scores on Digit Span Backward, Verbal Fluency, Trail-Making Test Part B, and Digit Symbol Substitution using principal components analysis. Other variables of interest included depression and apathy symptoms (as measured by subscales of the Geriatric Depression Scale) and selfreported CVRFs (i.e., diabetes, hypertension, dyslipidemia, obesity, smoking status). We used linear regression to evaluate variance of EF explained by CVRFs, depression, and apathy after controlling for sociodemographic covariates (age, education, sex). Relative importance (LMG) was the primary effect size examined. Result In unadjusted analyses, apathy, depression, and CVRFs were negatively associated with EF. Of the depression subscales, dysphoria yielded the largest effect size (r=-0.20) and apathy the smallest (r=-0.09). After adjusting for covariates, CVRFs explained 3.8% with strongest contributions from smoking status (LMG=1.4%), diabetes (LMG=1.2%), and hypertension (LMG=1.1%). Depression explained 2.7%, and apathy a marginal 0.4% of variance in EF. In models stratified by race/ethnicity, CVRFs and depression had differential contributions; depression explained more variance than CVRFs for NHW and H/L participants while CVRFs explained more variance than depression for NHB participants. Conclusion In this multi-ethnic sample, CVRFs were associated with EF at small-to-medium effect size while depression and apathy accounted for additional variance. The impact of depression on EF was stronger than apathy, although this may be partially explained by the GDS subscale inadequately capturing the construct of apathy. Future work involving an independent measure of apathy and longitudinal analyses of cognitive and mood trajectories may better elucidate these associations.

Funding: HABS-HD was supported by the National Institute on Aging of the National Institutes of Health under Award Numbers R01AG054073, R01AG058533, R01AG070862, P41EB015922 and U19AG078109; LDM is supported by NIH Award UH2/UH3 AG083264. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Additional support was received by the Provost's Undergraduate Research Scholarship of the University of Houston.; COI: The authors declare no conflicts of interest.

Poster # 133: REVISITING THE LAWTON AND BRODY: A PSYCHOMETRIC AND PRAGMATIC ANALYSIS

Presenting author: Jared F. Benge, PhD, ABPP (UT Austin)

Background: Instrumental activities of daily living (IADLs) are critical for assessing functional independence and disease progression in Alzheimer's disease and related disorders (ADRD). The Lawton and Brody IADL Scale, developed over half a century ago, remains widely used. However, there is little information available on its psychometric applicability and practical use in the modern ADRD clinical research space. With the advent of disease-modifying treatments and an emphasis on detecting early functional decline, reevaluating this scale is essential. Methods The study analyzed data from 3,552 individuals and their co-participants in the Texas Alzheimer's Research and Care Consortium (TARCC) database, spanning 2007-2018.

Participants were classified into three cognitive groups: no cognitive impairment, mild cognitive impairment, and dementia based on their Clinical Dementia Rating score. Using graded response models and differential item functioning (DIF) analyses, we assessed item discrimination and difficulty across demographic groups. Additionally, expert review was conducted to evaluate item content, cultural appropriateness, and technological relevance in modern contexts. Results Most items clustered around the same overall functional ability level, with items assessing shopping, food preparation, and transportation the most sensitive to mild functional impairment. However, the scale exhibited limited sensitivity to subtler functional declines, raising concerns about its effectiveness in early detection or tracking disease state at milder stages. Statistically significant DIF was observed between demographic groups across items. Expert reviewers identified outdated item content particularly for questions assessing telephone use and finances, given technological advancements. More broadly, shifts in normative behaviors and clinical trial composition could impact usability of the scale. Conclusion While the Lawton and Brody remains foundational for assessing functional independence, its utility in modern clinical and research contexts is challenged by item-level limitations and the potential for differential item functioning across groups. Findings underscore the need for refining assessment tools for the modern ADRD clinical research space, while preserving continuity for comparative and longitudinal research. In particular, refining the instrument to address the changing technological and socioeconomic demands of modern older adults holds potential to strengthen functional evaluations and support the field's focus on early disease detection.

Funding: TARCC Collaborative Grant 1280666; COI: None reported.

Poster # 134: COGNITIVE INTRAINDIVIDUAL VARIABILITY: PRECLINICAL HARBINGER OF DEMENTIA OR BEHAVIORAL MARKER OF COGNITIVE RESERVE, INTRINSIC CAPACITY, AND RESILIENCE?

Presenting author: Anika Bhatia (UT Dell)

Background: Prior work suggests that intraindividual variability across tests (IIV-Across) and between cognitive domains (IIV-Between) may reflect distinct neurocognitive mechanisms that are context dependent. To help clarify these mechanisms in aging adults, we tested the hypothesis that different demographic and clinical characteristics would predict these IIV metrics in cognitively normal adults with and without subjective cognitive impairment (SCI) versus mild cognitive impairment (MCI). Method: Baseline data from 1790 non-demented adults (1002 older adults controls: OAC; 287 SCI, 501 MCI) enrolled in the Texas Alzheimer's Research and Care Consortium Longitudinal Hispanic Cohort study were included for analysis. Cognitive IIV metrics were calculated as intraindividual standard deviations and hierarchical linear regression analyses were conducted separately for each IIV metric in each diagnostic group with 3 blocks: Block 1 = Sociodemographics (age and premorbid IQ), Block 2 = Psychiatric comorbidities (Anxiety, Apathy, and Dysphoria severity ratings), and Block 3 = Mean Performance Level (Language, Attention, Memory, and Executive Function). Results: Higher estimated premorbid IQ was predictive of higher IIV-Across tests in the OAC group only, while older age was predictive of greater IIV-Across tests in the SCI group, and neither age nor premorbid IQ were predictive of IIV- Across in the MCI group nor IIV-Between domains in any study group. All mean domain scores were predictive of both IIV metrics in the OAC group with the exception of attention, which was identified as a significant predictor of both IIV metrics in the SCI and MCI groups. The direction of each relationship between mean domain scores with IIV metrics in the SCI and MCI groups were consistently opposite to that of the OAC group with the exception of language which showed an inverse relationship with IIV-Between scores in both OAC and MCI groups and was unrelated to either IIV metric in the SCI group. Conclusions: Obtained findings are consistent with the view that IIV metrics are context-dependent and align with the World Health Organization's concept of intrinsic capacity, which emphasizes the interdependence of physical and mental abilities that all contribute to healthy aging and strengths in one domain may provide the capacity to compensate for declines in another.

Funding: None reported.; COI: None reported.

Poster # 135: VALIDATION OF SOUTHWESTERN ASSESSMENT OF PROCESSING SPEED IN A MIXED CLINICAL SAMPLE

Presenting author: Nusha Kheradbin, BS (UT Dell)

Background: Digit Symbol Substitution (DSS) tasks are among the most widely used measures of processing speed due to their reliability and clinical relevance across a broad spectrum of neurocognitive disorders. In 2021, researchers at UT Southwestern developed the Southwestern Assessment of Processing Speed (SWAPS), a 60-second digit-symbol transcription task. While similar to other commercially available DSS tests, this stand-alone measure of processing speed offers a more cost-effective alternative and incorporates modern regression-based normative data derived from a racially, ethnically, and educationally diverse population. This study aims to validate SWAPS in a mixed clinical sample and assess its convergent and divergent validity. Method: Participants included 117 older adults (51% males; 85% non-Hispanic white) from two neuropsychology clinics (Age: X̄= 73.97, SD= 7.78; Education: X̄= 15.81, SD= 2.84).

Participants completed a comprehensive evaluation including measures of memory, language, and executive functioning, and were categorized as cognitively normal (CN), mild cognitive impairment (MCI), or dementia (CI). Correlations were used to assess bivariate relationships and ANOVAs were used to examine differences between groups. Results: Older age was correlated with lower SWAPS performance (r = -0.34, p < .001), while no significant relationships were found with other sociodemographic characteristics: (Sex: ρ = -0.67, p = 0.48; Education: r = 0.04, p = 0.66; Race/Ethnicity: ρ = 0.8, p = 0.4). Correlations with neuropsychological measures ranged from 0.01 to 0.64, with the weakest association observed with Multilingual Naming Test raw score. The strongest associations were found with raw scores on Stroop Color (r = .64), Trail Making Part A (r = -0.62), and Immediate Recall on Brief Visuospatial Memory Test (r = 0.60).

SWAPS scores significantly differed across diagnostic groups (CN>MCI>CI). Conclusions: SWAPS demonstrated strong convergent validity with measures of working memory, processing speed, and visuospatial skills, and weak associations with language and self-reported symptoms of depression. It effectively differentiated between diagnostic groups and was not associated significantly with gender, education, or race/ethnicity, although our sample was generally well educated and primarily non-Hispanic white. Findings suggest SWAPS is a valid measure of working memory and processing speed but additional research with heterogeneous clinical populations is needed.

Funding: NIH/NIA AG069780 and P30AG066546, TARCC 1280666, ALZpath Bio; COI: none

Poster # 136: A PRINCIPAL COMPONENT ANALYSIS OF PASSIVE DIGITAL BIOMARKERS AMONG OLDER ADULTS

Presenting author: Sidonia E. Compton, PhD (UT Dell)

Background: Passive sensing biomarkers hold promise as tools to identify functional and cognitive decline in older adults, yet the plethora of data types available from such approaches can make analytic strategies challenging. Principal component analysis (PCA) offers a datadriven approach to identify dimensionality and latent structures across heterogeneous variables in these settings. Thus, we characterized facets of passive digital biomarkers aggregated across 30 days in a sample of older adults to identify key underlying dimensions for clinical research. Method: Participants (N = 21; M age = 74.87, SD = 5.03; 66.7% female) were drawn from the ongoing, longitudinal Technology for Smartphone Assessment of Neurocognitive Symptoms (TechSANS) study. At baseline, participants installed the TechSANS application on their smartphones to capture passive sensing metrics. A PCA with varimax rotation was performed to explore the underlying structure across the following variables: use-related (i.e., unlock duration), location-related (i.e., distance traveled from home, number of locations visited), gait (i.e., step count, average walking speed), keyboard metrics (i.e., total words typed, word deletions relative to total words typed), and social engagement (i.e., unique call contacts, outgoing text messages). Result: Four principal components with eigenvalues >1 were identified, accounting for 75.87% of the total variance. Component 1, explaining 28.50% of the total variance, was defined by faster average walking speed, greater distance traveled, and higher step count, a component marked by movement. Component 2 reflected greater unique phone contacts and visiting more unique locations; this novelty factor explained 18.96% of the total variance. Component 3, which explained 14.37% of the total variance, represented longer unlock duration and more words typed, suggestive of a general use factor. Component 4 was marked by sending more text messages and deleting fewer words, explaining 14.03% of the total variance. Conclusion: Initial findings suggest that movement, novelty of use, general use, and messaging represent four broad components of multi-modal passive smartphone sensing markers in older adults. Future work is needed to clarify which facets of use optimally serve as early indicators of functional and cognitive decline to help guide individualized interventions, and to test the stability of these factors over time in larger samples.

Funding: National Institutes of Health (R01AG082783), Texas Alzheimer's Research and Care Consortium (1280720), and National Academy of Neuropsychology (NAN) Clinical Research Grant.; COI: None.

Poster # 137: COMPARING VISUOSPATIAL TASKS IN COGNITIVE EVALUATION

Presenting author: Hanna Hausman, PhD (UTH San Antonio)

Background: Construction tasks such as cube copy (CC) and figure copy (FC) are often administered with perceptual estimation tasks like Number Location (NL) in the cognitive evaluation of dementia. Similar tasks are included in common screening measures (e.g., MoCA) and test batteries (e.g., RBANS). Questions remain about the relative contributions of these measures to cognitive diagnosis. This study examined classification accuracy of CC, FC, and NL singly and in combination in outpatients referred for neuropsychological evaluation. Method: The sample (N=122) included valid cases from a local clinical dataset with CC, FC, and NL data and clinical diagnosis: normal (n=46), mild cognitive impairment (MCI; n=58), or dementia (DEM; n=18). Participants averaged 74.9 (SD=8.1) years old with average education of 14.3 (SD=2.9) years; 57% of participants were female. Race/ethnicity percentages were 3.6% African American, 2.7% Asian, 30.6% Hispanic/Latino, and 63.1% White non-Hispanic/Latino. CC qualitative scoring was converted to a three-level score (0, 0.5 and 1) with higher score indicating better performance. FC and NL were scored in standard fashion. Classification accuracy was examined using receiver operating characteristic curve (ROC) analysis with these participant groups: MCI-DEM/normal, MCI/normal, and DEM/normal. Logistic regression models were compared using CC, FC, and NL as predictors. Results: Area under the ROC curve (AUC) values were modest for predictors individually in differentiating MCI-DEM/normal (AUC=.64.67) and MCI/normal (AUC=.59-.62) cases. AUC values were acceptable to excellent (AUC=.78.82) differentiating DEM/normal cases. CC was the only significant predictor in an LR model predicting MCI/normal cases with modest accuracy (AUC=.66). In the model predicting DEM/normal cases, NL and FC were significant predictors with excellent accuracy (AUC=.88). Conclusion: Classification accuracy of visuospatial tasks remained modest in differentiating MCI from normal cases with improved accuracy distinguishing DEM from normal cases. LR results indicate evidence of differential roles for the three measures with CC being more useful in MCI/normal distinction and FC and NL showing utility in the DEM/normal distinction. Further research examining visuospatial tasks in a larger sample is warranted.

Funding: Not applicable.; COI: None.

Poster # 138: ASSOCIATIONS BETWEEN ALZHEIMER'S BIOMARKER STATUS, NEUROPSYCHIATRIC SYMPTOMS AND PRECLINICAL ALZHEIMER'S COGNITIVE COMPOSITE SUBSET PERFORMANCE

Presenting author: Katherine Goulden, MA (UTRGV)

Among the earliest domains of cognitive functioning to decline in Amnestic and Non-Amnestic MCI are episodic memory and executive function, domains of functioning measured by the Trail making Test and Digit Span test within the Preclinical Alzheimer's Cognitive Composite Performance (PACC) measure. The PACC is designed to identify subtle cognitive changes in individuals who are cognitively unimpaired but may be at risk for developing Alzheimer's Disease.PET imaging of 247 participants from the Alzheimer's Disease Neuroimaging Initiative was classified by Braak stage based on regional tau deposition. We examined the relationship of CSF Tau and β-amyloid (Aβ)-PET status with reported neuropsychiatric symptoms on the Neuropsychiatric Score Inventory (NPI-Q) and neuropsychological performance on the PACC subtests of Trail making Test and Digit Span test.Accounting for age, education, sex and apoe4 status, a higher CSF tau burden was associated with lower PACC trail making performance. Accounting for age, education, sex and apoe4 status, a higher CSF tau burden was associated with lower PACC Digit Span performance. APOE4 Allele and education significantly moderated performance on both measures, with a significant interaction of gender. Accounting for age, education, sex and apoe4 status, a higher amyloid burden was associated with lower PACC performance on both tests, with age significantly moderating the relationship. Aβ-PET presence was found to be associated with increased frequency and severity of neuropsychiatric symptoms, especially affective symptoms like anxiety and apathy. In Tau-positive individuals mood and anxiety symptoms show an inconsistent relationship with objective cognition. Lower PACC scores are associated with more neuropsychiatric symptoms, especially in domains like motivation, drive, and mood. The direction of association between lower PACC scores and more neuropsychiatric symptoms is consistent with the literature. The results suggest that neuropsychiatric symptoms may be an early marker or risk factor for cognitive decline.

Funding: N/A; COI: N/A

Poster # 139: EXPLORING THE ASSOCIATION OF BILINGUALISM IN HISPANIC/LATINX ADULTS WITH AGE OF COGNITIVE IMPAIRMENT

Presenting author: Diego Jofre-Zarate, MS (UTSW)

Background: Hispanic/Latinx adults (HLAs) in the USA face higher rates and have an earlier onset of Alzheimer's disease compared to European Americans. One factor of growing interest is bilingualism, which has been proposed as a contributor to cognitive reserve and a buffer against cognitive impairment (CI). Thus, we examined whether bilingualism might be linked to a later age of CI within an HLA population in the USA. Method: Data were obtained from the Health and Aging Brain Study-Health Disparities (HABS-HD). Participants were selected for the current study if they self-reported as Hispanic, were aged ≥ 50, and had CI using the Clinical Dementia Rating (CDR) Global Score ≥ 0.5. The sample included Alzheimer's dementia cases and excluded those with other dementias, a lifetime history of stroke, or alcohol use disorder. Language status was classified according to participants' self-reported English/Spanish monolingual or bilingual capability. To investigate if there was a difference in the age of CI between monolingual (n=174) and bilingual (n=173) participants, we conducted an ANCOVA, with education and apolipoprotein E ε4 allele (APOE4) status entered as covariates. Age of CI was operationalized as being each participants' age at study entry into HABS-HD. Results: Monolingual and bilingual groups were similar in sex and CDR-Global Scores, but bilingual adults had more years of education (M = 11.88 vs. 7.44 years; p<.001.) and fewer APOE4 carriers (31% vs. 47%; p=.04). There was no significant difference in the age of CI between the groups, while controlling for education and APOE4 status, F(1, 343)=.13, p=.72, partial η2 <.001. Conclusions: Among HLAs, bilingualism was not associated with a later age of CI when education and APOE4 status were considered. While this suggests that language status alone may not delay cognitive decline, the true age of onset for CI may occur earlier than could be detected within the current study. Thus, interpretation of CI timing is limited and unknown within HABS-HD. Future studies exploring the estimated or real timing of CI are needed, as are investigations with longitudinal designs and objective measures of bilingual proficiency to clarify the role of bilingualism in cognitive reserve and dementia risk.

Funding: This work is not supported by any grant funding.; COI: Vishal Thakkar, PhD, receives partial salary support and is actively affiliated with TARCC.

Poster # 140: STRUCTURED PROMPTING IMPROVES LARGE LANGUAGE MODELBASED IDENTIFICATION OF ALZHEIMER'S DEMENTIA: PROOF OF CONCEPT USING SYNTHETIC TRANSCRIPTS

Presenting author: Oscar Kronenberger, BA (UTSW)

Objective: Lexical, syntactic, and semantic domains of language are often impaired in Alzheimer's dementia (AD), but efficient quantification of unstructured language data poses challenges to large-scale clinical evaluation. Large language models (LLMs) may automate quantification of linguistic features associated with AD and ultimately support detection of cognitive impairment, but optimal methods remain unclear. In this proof-of-concept study, we highlight the impact of LLM prompting methods on AD classification accuracy using synthetically generated transcripts. Methods: We reviewed literature to identify linguistic features that differentiate AD from cognitively unimpaired (CU) adults. We then generated 100 (50 AD, 50 CU) transcripts for the Boston Cookie Theft Picture description task, a common clinical test of language production and organization, based upon empirical features using Gemini 2.5 Pro. Next, the ChatGPT-5 chain-of-thought "thinking" model was asked to predict group membership of each transcript using two prompts: zero-shot versus structured. The zero-shot method stated, "Based on the transcript below, assess whether the speaker is cognitively unimpaired or shows evidence of Alzheimer's dementia." In comparison, the structured prompt provided a linguistic deficits rating scale with objective coding guidelines (1=very impaired to 5=normal) indexing lexical, syntax/grammar, semantics, and content features, concluding with a prediction of group membership. Classification accuracy (ACC), sensitivity, and specificity were assessed for each method. Index scores were assessed using t-tests, Cohen's d effect sizes, and logistic regression. Results: Zero-shot prompting had an ACC of 70% with poor sensitivity (40%) but strong specificity (100%) in identifying the AD and CU transcripts. Structured prompting had 100% total ACC and exhibited very large group differences across lexical (d=11.4; ACC=100%), syntax/grammar (d=15.2; ACC=100%), semantic (d=6.6; ACC=100%), content (d=2.8; ACC=91%), and composite scores (AD M=2.5, SD=0.2; CU M=4.7, SD=0.1; d=15.5; ACC=100%). Conclusions: These findings provide preliminary proof of concept for the use of structured prompting methods in LLM-based AD linguistic feature extraction and identification. Critically, the accuracy reported here is likely inflated by the use of synthetic transcripts and does not provide real-world validation; thus, further research in real-world natural communication samples is needed.

Funding: N/A; COI: N/A

Poster # 141: NONINVASIVE BRAIN STIMULATION ON CATEGORY FLUENCY PERFORMANCE IN MILD COGNITIVE IMPAIRMENT AND ALZHEIMER'S DEMENTIA

Presenting author: Tara Driskill, BS (UTSW)

Introduction In prior clinical trials, High-Definition Transcranial Direct Current Stimulation (HD-tDCS) was associated with improvement in verbal memory and phonemic fluency in cognitively impaired older adults, but no effect on semantic fluency when using the Delis-Kaplan Executive Function System (DKEFS), which includes common noun (animals) and proper noun (boy's names) tasks. These semantic fluency categories may differ in underlying cognitive processes, with common nouns relying more on semantic networks to generate words. Two components of semantic fluency, clustering (generating words within subcategories) and switching (shifting between subcategories) may represent different retrieval strategies within the semantic network. This secondary analysis of trial data evaluated potential HD-tDCS effects on total scores, clustering, and switching performance of animal fluency among older adults with amnestic Mild Cognitive Impairment (aMCI) and mild Alzheimer's Dementia (AD). Methods 44 participants (M age=71.75) underwent ten 20-minute sessions of HD-tDCS over the medial prefrontal cortex and were randomized to receive either active treatment (n=28) or sham treatment (n=16). Individuals who received either 1 mA (n=19) or 2mA (n=9) active HD-tDCS were combined due to similar effects seen on other cognitive outcomes regardless of current intensity in our initial work. T-scores for animal fluency, total cluster size (number of words within a subcategory), and total switches between clusters were calculated. Repeated measures analyses of variance and effect size calculations were performed to evaluate changes from baseline to immediate post-treatment between conditions. Results No statistically significant changes were demonstrated on cluster size, total switches, or total Animal performance between HD-tDCS conditions (p>.05). Effect sizes were negligible for cluster size (d=0.25) and switching (d=0.15), but moderate for timepoint x condition for animal fluency performance (d=0.51), with the active group showing improvement (M difference = 5.42 T-Scores). Conclusions HDtDCS over the medial prefrontal cortex did not produce robust improvement on semantic fluency in this small sample when measured through animal fluency, mean cluster size, or total switching, but did produce a moderate effect. Further research examining neuromodulation effects on semantic fluency measures between individuals with AD versus aMCI may be necessary to explore if there could be differential effects based on cognitive stage.

Funding: This study received funding from AWARE North Texas and the Texas Alzheimer's Research Care Consortium.; COI: No conflicts of interest to disclose.

Poster # 142: LANGUAGE INFLUENCES ON VERBAL MEMORY ORGANIZATION: EVIDENCE FROM NETWORK ANALYSIS OF WMS LOGICAL MEMORY

Presenting author: William Goette, PhD (UTSW)

Background: Translating cognitive measures into other languages is essential for equitable neuropsychological assessment. Literal translations of tests may preserve content, but linguistic differences in vocabulary, syntax, and grammar can preclude true psychometric equivalence. This issue is especially relevant for verbal memory tests where semantic associations may influence encoding and retrieval. Method: Data were drawn from the Health and Aging Brain Study: Health Disparities (HABS-HD). Item scores from the immediate recall trial of Wechsler Memory Scale-III Logical Memory (LM) were analyzed for N=3004 cognitively unimpaired participants (English: n=2369; Spanish: n=635). Bayesian conditional latent network analysis compared LM item-response networks between language groups, adjusting for age, education, and race. Posterior predictive p-values (ppp) were used to test differences in network structure using correlation of the network weights, Hamming distance, and expected influence. Bayesian betabinomial regression tested differences in expected raw scores controlling for the same covariates.

Result: LM item networks differed significantly by language. Correlation of network associations (ppp<.001), Hamming distance (ppp=.001), and expected influence (ppp=.04) all indicated structural divergence. English speakers exhibited 9 item clusters while Spanish speakers showed 11, reflecting distinct associative patterns. Despite these network-level differences, expected total recall did not differ by language (β=-0.01), controlling for covariates. Conclusion: Different network structures for LM learning demonstrates that translation of the story did not result in equivalent associational relationships between details. Thus, while the expected total scores on the test are the same, the process of associating the story's details are seemingly not equivalent between English and Spanish. Findings like these are cautionary reminders that critical-item analyses or validity scales based on rarely missed items require direct validation even when all other indicators of translation equivalence are seemingly acceptable. These results underscore the need for process-level evaluation of test adaptation because reliance on aggregate scores may obscure meaningful differences in cognitive organizational and utilization of semantic association networks.

Funding: Research reported in this publication was supported by the National Institute on Aging of the National Institutes of Health under Award Numbers R01AG054073 and R01AG058533, R01AG070862, P41EB015922 and U19AG078109. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.; COI: None reported.

Poster # 143: PAID AND FAMILY CAREGIVING IN ALZHEIMER'S DISEASE: A SCOPING REVIEW OF OUTCOMES AND EQUITY IN COMMUNITY-DWELLING POPULATIONS

Presenting author: Morad Marikh, BS (Texas Tech University Foster School of Medicine )

Background: Alzheimer's disease (AD) is a leading cause of dependence in older adults, who primarily receive care at home. Family members provide the majority of unpaid support, yet the role and impact of paid caregiving on patient outcomes and equity remain insufficiently understood. Clarifying how caregiving type and socioeconomic context influence functional decline, behavioral and psychological symptoms of dementia (BPSD), and quality of life is essential for designing equitable, sustainable care models. Method: Following PRISMA-ScR guidelines, a scoping review of Ovid MEDLINE (January 2020-October 2025) was conducted to identify peer-reviewed studies comparing outcomes among community-dwelling adults with AD or related dementias receiving paid versus family (unpaid) caregiving. Eligible studies reported patient-level outcomes, including functional status, activities of daily living (ADLs), BPSD, or quality of life. Studies focusing solely on caregiver burden, institutionalized patients, or nonEnglish publications were excluded. Data were extracted on study design, sample characteristics, caregiving context. Results: Sixteen studies encompassing over 58,000 dementia-care dyads were included. Paid caregiving improved symptom management, reduced unmet needs, and lessened caregiver strain, particularly within higher-resource settings. Family caregiving provided stronger emotional support but was associated with greater burden and less consistent follow-up. Socioeconomic and geographic disparities significantly influenced outcomes. Hybrid models integrating professional and family caregivers yielded the most favorable outcomes in maintaining independence, reducing symptoms, and improving quality of life. Conclusion: Integrating professional and family caregiving may optimize outcomes while reducing disparities. Findings should be interpreted cautiously given the scoping nature of this review and heterogeneity across included studies. Future work should evaluate these hybrid models across diverse U.S. populations.

Funding: None reported.; COI: None reported.

Poster # 144: PHYSICAL AND BEHAVIORAL TRAJECTORIES AND COGNITIVE OUTCOMES: A LONGITUDINAL ANALYSIS OF THE HEALTH AND RETIREMENT STUDY

Presenting author: Breanna Morales, MS (TAMU)

Background: Alzheimer's disease and related dementias (ADRD) are preceded by subtle physiological changes that may appear years before clinical diagnosis. Declines in gait speed and daily activity have been proposed as early indicators of neurodegenerative vulnerability, yet longitudinal evidence remains limited. The Health and Retirement Study (HRS) is a nationally representative cohort of U.S. older adults, aged 51 years and older, re-interviewed biennially. This dataset provides longitudinal data to examine how trajectories of physical function and lifestyle behaviors predict cognitive outcomes. Objective: This study evaluated whether longitudinal changes in gait speed and physical activity serve as early, modifiable indicators of Alzheimer's disease and dementia in a nationally representative cohort of older adults. Methods: Data were obtained from the Health and Retirement Study (HRS). Participants with at least two assessments of gait speed or physical activity and no diagnosis of Alzheimer's disease or dementia at baseline were included. Alzheimer's and dementia status were determined through self- or proxy-reported physician diagnosis. Annual rates of change for each variable were calculated as the slope between available waves (2004-2022), with missing intervals estimated under the assumption of linear change. These slopes represented the yearly trajectory of change in gait speed and physical activity. Generalized estimating equation logistic regressions and Cox proportional hazards models were used to test associations with subsequent Alzheimer's or dementia diagnosis, adjusting for baseline age. Results & Discussion: Gait speed was a significant predictor of Alzheimer's and dementia diagnosis. Each one-second per year slowing in gait speed increased the odds of Alzheimer's by 3.7 times and dementia by 2.8 times (p < .001). Annual declines in physical activity were also associated with a greater risk, while maintaining higher activity levels was protective (OR = 0.95, p < .001). Findings were consistent across Cox models, indicating that steeper yearly declines were linked to earlier disease onset. Even after adjusting for age, gait speed remained the strongest predictor. These results align with prior evidence linking slower gait to cognitive decline and suggest that subtle, measurable decreases in gait and physical activity may serve as early, modifiable indicators of neurodegenerative risk.

Funding: N/A; COI: n/a

Poster # 145: A STATEWIDE NEEDS ASSESSMENT OF SERVICES AND SUPPORTS FOR UNPAID CAREGIVERS OF PEOPLE LIVING WITH ALZHEIMER'S DISEASE AND RELATED DEMENTIAS

Presenting author: Matthew Lee Smith, PhD, MPH, CHES, CHW-I (TAMU)

Background: Texas ranks in the top five states for individuals living with Alzheimer's Disease and related dementias (ADRD), highlighting the growing importance of unpaid caregiving by family members and friends. A statewide needs assessment was conducted to better understand the availability, accessibility, and use of services for ADRD caregivers in Texas. It also sought to identify the unmet needs and barriers to service access and utilization from the perspective of unpaid caregivers and the organizations that serve and support them. Method: Two interrelated statewide surveys and a series of four virtually-hosted Community Discussion Groups (CDG) were conducted. Both surveys were designed to complement and build upon one another to receive comparable information from the vantage point of: (1) unpaid ADRD caregivers; and (2) organizations serving unpaid ADRD caregivers. The CDG gathered more in-depth qualitative perspectives about the needs of unpaid ADRD caregivers, service adequacy and deficits, barriers to service utilization, and needs to improve service delivery. Result: About 580 caregivers and 50 organizations across Texas completed the surveys, with responses coming from both urban and rural regions. Nearly 30 representatives of community organizations participated in CDG to offer additional in-depth insights and confirm survey findings. Awareness and access were central issues for caregivers and organizations. Caregivers stressed day-to-day needs like health and wellness and respite care, while organizations highlighted gaps in structural supports like financial and legal planning. Caregivers most often reported needing, but not using, services such as caregiving technologies, dementia research updates, health and wellness programs, respite care, and transportation. These needs were especially pronounced in rural areas, where caregivers reported more limited availability and accessibility of services. Organizations confirmed many of these gaps reported by caregivers. Despite the range of challenges reported, caregivers and organizations had strong agreement about what should be prioritized to support ADRD care in their areas. Conclusion: This initiative provided a glimpse into the service gaps, access obstacles, and opportunities for improving ADRD caregiving support in Texas. Practical opportunities to improve ADRD-related service availability, access, and use are provided.

Funding: This project was supported by the Centers for Disease Control and Prevention of the U.S. Department of Health and Human Services (HHS) as part of a financial assistance award totaling $224,000 with 100 percent funded by CDC/HHS. The contents are those of the author(s) and do not necessarily represent the official views of, nor an endorsement, by CDC/HHS, or the U.S. Government.; COI: N/A

Poster # 146: EVALUATION OF THE TECHNOLOGY ACCEPTANCE OF A DIGITAL CAREGIVING ASSISTANCE PLATFORM: A MIXED-METHODS RESEARCH

Presenting author: Minh-Nguyet Hoang, MBA (Texas A&M UHSC)

Background: The Olera.care digital platform was developed informed by needs assessment research with informal caregivers. The objective of this study was to evaluate the technology acceptance of the Olera.care platform among family caregivers in the United States using validated surveys, and to characterize caregiving challenges, user experience with the platform, and attitudes towards integrating artificial intelligence (AI) in future iterations of the tool using open-ended qualitative questions. Methods: Participants were recruited from online and asked to engage with the Olera.care platform for four weeks. They completed an initial eligibility survey and a technology acceptance survey (TAS) at the end of four weeks. Surveys included socioeconomic characteristics, caregiving experiences, technology engagement and literacy, and perceived usefulness and ease-of-use of the platform. Qualitative feedback regarding the usefulness of the platform was analyzed via a thematic analysis framework approach. Descriptive statistics were used to summarize TAS responses. Analyses of variance, t-tests, and linear regressions were used to compare the differences in the overall TAS scores by caregiver characteristics. Results: A total of 65 caregivers completed the study. Participants had an average age of 59.9 (SD 9.8 years), were female (61/64, 95.3%), White (45/65, 68.2%), and the adult child of persons living with dementia (PLwD) (52/65, 64.6%). The Olera.care platform evaluation showed a high acceptance rate, with an overall mean TAS score of 5.83/7 (SD 0.85) and all individual TAS items scored above 5.0. Technology acceptance significantly differed by platform use frequency (F(3,61)= 7.88, p< .001), with daily users reporting the highest scores (mean 6.55, SD 0.44), followed by those using the platform 4-6 times weekly (mean 6.05, SD 0.70), 2-3 times weekly (mean 5.82, SD 0.62), and once weekly (mean 5.20, SD 0.99). Thematic analyses elicited the caregiving challenges, experiences of the Olera.care platform, and receptiveness toward AI-assisted support. Conclusion: Findings reinforce evidence that informal caregivers of PLwD face substantial challenges coordinating care services, managing costs, and balancing daily responsibilities. The Olera.care platform demonstrated strong acceptance and perceived utility, though caregiver feedback emphasized the need for expanded care-service directories, assistance with financial navigation, and integrated AI-driven decision support to streamline future caregiving demands.

Funding: National Institute on Aging Small Business Innovative Research 1R44AG074116-01; solicitation AG21-025; COI: None declared

Poster # 147: IMPROVING VIETNAMESE REFUGEES AND IMMIGRANTS' COGNITIVE HEALTH LITERACY THROUGH UNIVERSITY-COMMUNITY PARTNERSHIPS

Presenting author: Nguyen T.K. Nguyen, PhD, MSW, MA (University of Houston)

Background: Since the fall of Saigon in 1975, waves of Vietnamese refugees/immigrants have migrated to the United States (U.S.). Currently, 2.3 million Vietnamese live in the U.S. Vietnamese arrived with adverse conditions, including wartime traumas, life-threatening journeys, low socioeconomic status, and linguistic barriers, with much of their health data unknown. To fill this gap, we developed the Vietnamese Aging and Care Survey and collected their health data in Houston, the 2nd largest Vietnamese-populated metropolitan area in the U.S. It showed high rates of physical, mental, and cognitive disabilities. Using the UniversityVietnamese community partnerships, we further developed a linguistically and culturally tailored dementia one-pager to improve the community's health/dementia literacy, formed the Cognitive Health Initiative (CHAIN), and offered complimentary memory tests to monolingual Vietnamese refugees/immigrants. This study reports how CHAIN trains university students and fosters multigenerational collaborations among the Houston Vietnamese-American community. Method: Following the Cultural Exchange Model, we trained bilingual/bicultural Vietnamese students (Cohort 1) in the Vietnamese Montreal Cognitive Assessment (MoCA-V). When Cohort 2 students joined, we formed a mentor (Cohort 1)-mentee (Cohort 2) relationship, and Cohort 1 demonstrated the MoCA-V while Cohort 2 observed their mentors perform. After several demonstrations, Cohort 2 tried the MoCA-V supervised by Cohort 1. They repeated this sequence until Cohort 2 felt comfortable conducting the MoCA-V independently. Result: During 20232024, we trained 52 students, attended 16 Vietnamese health fairs, and offered 406 MoCA-V assessments. Participants were on average 73±7.1 years old, female, married, retired, highschool-educated, in good health, and scored 22/30 MoCA-V. Cohort 1 successfully trained Cohort 2, providing them with tips for conducting the MoCA-V with older Vietnamese. Conclusion: We trained Vietnamese university students and conducted culturally and linguistically tailored cognitive assessments to monolingual older Vietnamese with low health literacy. The mentor-mentee relationships promoted teamwork, accountability, ethics, and responsibility to the community among CHAIN students, and CHAIN demonstrated successful University-Community partnerships. Improving the community's health literacy requires longterm commitment and the efforts of all parties involved. Researchers, university students, and the Vietnamese community need to continue collaborating and achieve a common goal of improving the health literacy of older Vietnamese refugees/immigrants.

Funding: University of Houston Cougar Initiative to Engage (CITE) Grant; University of Houston Small Grants Program; National Institutes of Health (NIH)/National Heart, Lung, and Blood Institute (NHLBI) OT2HL158287 (subaward); COI: Both authors have no conflicts of interest.

Poster # 148: THE WHO, WHAT, AND HOW OF REAL-WORLD TECHNOLOGY USE BY CAREGIVERS OF OLDER ADULTS WITH DEMENTIA

Presenting author: Dariella Fernandez, PhD (UT Dell)

Background: Caregivers of people living with dementia (PLWD) experience elevated rates of burden, depression, and anxiety. Increasingly, caregivers turn to technology-based strategies to improve care and mitigate these outcomes, however, little is known about the spontaneous use of caregiving technology in real world settings. Kiselica et al. (2024) recently proposed the technology that CARES framework to characterize how caregivers use technologies for cognitive offloading (C) of caregiving demands, automated task management (A), remote monitoring and intervention (R), emotional and social support (E), and symptom treatment (S) of PLWD. This study applied the CARES framework to describe spontaneous technology use among caregivers, focusing on who uses technology most, what tasks they use it for, and how use relates to caregiver outcomes. Method: 100 caregivers (75% female; 60% spouses) of people with mild cognitive impairment to moderate dementia completed the Technology in Caregiving Questionnaire (TCQ) and measures of caregiver burden, depression, anxiety, and dementia severity. TCQ items were classified into CARES categories through expert consensus, with mean scores calculated for each domain. Correlations examined associations between technology use, caregiver demographics, and psychosocial outcomes. Results: Caregivers reported most frequently using technology for cognitive offloading (M=2.1, SD=.91) and least use for remote monitoring (M=.97, SD=.66). Commonly reported technologies (rated "very often") included digital calendars (54%), patient healthcare portals (34%), and automated banking (48%). Younger age correlated with greater technology use for cognitive offloading (r(97)=-.32, p<.01), remote monitoring (r(97)=-.41, p<.001), emotional support (r(97)=-.31, p<.01), and symptom management (r(97)=-.22, p<.05). Higher caregiver burden, depression, and anxiety were associated with greater overall technology use (p < .05); greater dementia severity correlated with more frequent use for emotional support (r(97)=.33, p < .001). Conclusion: Results indicate that caregivers of PLWD employ many technology-based strategies, particularly those that reduce cognitive load or automate complex tasks. Age emerged as a key correlate, suggesting a continued age-related digital divide among caregivers. Moreover, greater caregiving burden and symptom severity were linked to higher technology use, implying worsening care situations prompt caregivers to adopt digital solutions. Thus, there is likely a need to direct optimal digital resources to caregivers at times of need.

Funding: This work is supported in part by the National Institutes of Health (R01AG082783) and Texas Alzheimer's Research and Care Consortium (1280720). ; COI: None

Poster # 149: CATARACT SURGERY REDUCES DEMENTIA RISK IN OLDER ADULTS:
AN ANALYSIS OF 15-YEAR RETROSPECTIVE STUDIES INCLUDING MORE THAN 100,000 PATIENTS

Presenting author: Salvatore Saieva, PhD (UT Houston)

Background: Dementia affects more than seven million Americans. Recent evidence identified vision loss as a modifiable risk factor for dementia. One of the leading causes of visual impairment is cataract, a pathology characterized by the aggregation of the crystallin protein in the lens. Compelling evidence shows increased risk of dementia in older adults with cataract. Moreover, recent evidence suggests that cataract extraction reduces the risk of dementia. Although most evidence relates eye impairments and dementia risk, other reports suggest limited interactions. Notably, these reports were conducted in limited cohorts. Methods: This retrospective study was conducted on the TriNetX US Collaborative Network. Patients aged ≥ 65 years with a diagnosis of cataract (ICD-10: H25, H26, H28) were included. Two cohorts were defined: those who underwent cataract surgery (CPT: 66982, 66983, 66984, n= 122,705) and those who did not (n= 122,705). The index event was the first cataract diagnosis, with outcomes evaluated from 1-day post-index over 1-, 5-, 10-, and 15-year periods. Primary outcomes were incident diagnoses of dementia, including Alzheimer's disease, vascular dementia, frontotemporal dementia, and Lewy body dementia, identified by ICD-10 codes. Patients with prior dementia diagnoses were excluded. Analyses were performed on the TriNetX platform. Results: The incidence of all-cause dementia was consistently lower in the surgery cohort across all follow-up periods. At 1 year, cataract surgery was associated with a 42% lower risk of dementia (hazard ratio [HR] 0.62, 95% CI 0.53-0.72). This protective association persisted over time: HR 0.88 (95% CI 0.82-0.95) at 5 years, HR 0.86 (95% CI 0.81-0.91) at 10 years, and HR 0.86 (95% CI 0.81-0.91) at 15 years. Risk ratios for Alzheimer's disease were similarly reduced (RR 0.71-0.78 across analyses). Kaplan-Meier survival curves demonstrated significantly delayed onset of dementia among surgical patients (log-rank p < 0.01 at all intervals). Conclusions: Cataract surgery in older adults is associated with a significantly lower long-term risk of developing dementia, including Alzheimer's and vascular subtypes. Further research is needed to elucidate the biological mechanisms underlying such associations. Our work emphasizes the need for integrating ophthalmological health as a strategy to prevent dementia.

Funding: TARCC Postdoctoral Fellowship to SS; COI: None reported.

Poster # 150: RELATIONSHIP OF HEARING LOSS LATERALITY AND DEMENTIA DIAGNOSES

Presenting author: Satwant Kumar, MD, PhD (UT Tyler)

Background: Hearing loss is linked to cognitive decline, but clinic-ready, ear-level measures (e.g., left-ear word-recognition in quiet/noise) are seldom integrated with neuroimaging patterns. Establishing descriptive signals can inform screening and referral. Methods: A single-center, retrospective pilot study was conducted at a rural memory clinic. Inclusion criterion: patients referred from the memory clinic to the audiology assessment for subjective hearing issues (N=22). We reviewed age; diagnosis; hemispheric atrophy on MRI; which ear had greater hearing loss; and whether word-recognition scores (WRS) were worse in one ear. We summarized proportions with 95% Wilson confidence intervals and ran exploratory contingency tables. No adjusted modeling or p-value inference was planned a priori, given the sample size; analyses are descriptive and hypothesis-generating. Results: Left-ear word recognition was worse in 9/22 (40.9%, 95% CI 23.3-61.3%). Audiologic asymmetry was noted with left-sided hearing issues in 7/22 (31.8%, 16.4-52.7%), and right-sided in 4/22 (18.2%, 7.3-38.5%). Of the patients, 14/22 (63.6%, 43.0-80.3%) had Alzheimer's disease, 8/22 (36.4%, 19.7-57.0%) had Lewy Body Disease (LBD), and 9/22 (40.9%, 23.3-61.3%) had vascular cognitive impairment. MRI laterality was right>left for 8/22 (36.4%, 19.7-57.0%), and left>right for 3/22 (13.6%, 4.7-33.3%), whereas it was bilateral for 5/22 (22.7%, 10.1-43.4%). Conclusions: In this pilot, ear-specific wordrecognition deficits and MRI laterality were measured within routine care. No significant correlation emerged between laterality of hearing loss and hemispheric atrophy, suggesting that any linkage between ear-level speech clarity and hemispheric atrophy may necessitates a larger sample size.

Funding: None; COI: None

Poster # 151: CHARACTERIZING SUICIDE IDEATION AMONG DEMENTIA FAMILY CAREGIVERS

Presenting author: Wesley R. Browning, PhD (UTHealth Houston)

Background: Family caregivers of people living with Alzheimer's disease and related dementias (ADRD) face persistent emotional, financial, and physical challenges. Despite well-documented distress in this population, suicide ideation (SI) remains largely overlooked. Caregiving is often assumed to be protective against suicide, yet emerging evidence challenges this view. Little is known about SI rates or the daily factors that influence risk among caregivers. This study addresses this gap by examining the prevalence and daily correlates of SI in ADRD. Method Participants were 453 ADRD family caregivers who completed 9,513 daily diary entries about mood, stress, social interactions, and caregiving experiences over 21 days. SI was measured via a single adapted item from the Patient Health Questionnaire. We calculated SI incidence and frequency, and used generalized linear mixed modeling (GLMM) to assess person-level and daily predictors of SI. Result Across 21 days, 273 (60.3%) caregivers reported SI at least once. These caregivers reported SI on an average of 3.9 days. In the final GLMM, Non-Hispanic White (NHW) (OR=0.61, 95%CI=0.41-0.91) and female caregivers (OR=0.33, 95%CI=0.19-0.58) were less likely to report SI on a given day. Caregivers who had more depressive symptoms (OR=1.43, 95%CI=1.29-1.59), anger (OR=1.20, 95%CI=1.07-1.35), anxiety (OR=1.11, 95%CI=1.00-1.23), higher emotional impulsivity (OR=1.59, 95%CI=1.41-1.80), or managed more behavioral symptoms on a given day were more likely to experience SI on the same day. Hypothesized protective factors, such as social support, did not reduce likelihood of SI. Conclusion Our results show that SI is not only prevalent but also highly sensitive to daily changes in mood, emotion regulation, and caregiving demands among ADRD caregivers. Contrary to expectations, protective factors did not reduce daily SI risk, suggesting that traditional sources of resilience may do little to buffer against SI under caregiving stress. This pattern underscores that prevention efforts must be context-specific and responsive to caregivers' daily experiences. Male and nonNHW caregivers appear particularly vulnerable, highlighting the importance of person-centered approaches. Our results underscore the need of integrating routine SI screening into caregiver support programs and point to the value of timely interventions targeting emotion regulation and behavioral symptom management to prevent escalation of suicide risk.

Funding: National Institute on Aging (NIA) - R01AG072422; COI: None to report.

Poster # 152: BEYOND THE SCAN: VISUAL AI DECISION SUPPORT FOR EARLY ADRD DETECTION AND PROGNOSIS

Presenting author: Xuan Wang, PhD (UTRGV)

Alzheimer's Disease and Related Dementias (AD/ADRD) are among the most prevalent, particularly in individuals aged 65 and older, making them urgent public health concerns. In 2023, AD/ADRD accounted for $345 billion in healthcare costs in the U.S.-the highest of any disease, and healthcare costs for AD/ADRD are projected to rise to $1 trillion by 2050. As no clinical cure currently exists, early identification during the prodromal stages is critical to slowing disease progression. Our project aims to build an AI-powered visual decision platform that supports earlier detection of Alzheimer's disease in Hispanic communities, with a design that can readily translate to broader populations. The system is tailored to real-world barriers uneven access to care and the complexity of managing chronic conditions that elevate neurodegenerative risk. By fusing machine learning with intuitive visual analytics, the platform learns patterns linked to prodromal disease, raises timely alerts, and enables administrators and clinicians to filter and explore data by diagnosis, risk factors, or workflow needs. The result is decision support that strengthens evidence-based practice surfacing intervention impact, informing resource allocation, and ultimately improving outcomes and care delivery. In a recent usability and acceptance study with 188 participants aged between 50 and 75, a clear majority reported positive feedback: they indicated strong willingness to adopt lifestyle changes prompted by the tool's recommendations and rated the prototype as easy to use and the insights as practically useful. These findings reinforce the platform's potential to motivate preventive behaviors while fitting smoothly into everyday use.

Funding: NA; COI: The authors do not have a conflict of interest

Poster # 153: EXPLORING THE IMPACT OF FALL DETECTION TECHNOLOGY ON DEMENTIA CAREGIVERS AND CARE RECIPIENTS: PRELIMINARY RESULTS

Presenting author: Marcia Ory, PhD (Texas A&M UHSC)

Background: Falls are common among people living with dementia (PLWD) and can significantly impact caregiver burden and quality of life. This study presents preliminary findings from an ongoing evaluation of a fall detection and prediction system implemented in households of PLWD and their caregivers. This system uniquely uses Ultra-Wideband technology, AI/Neural Network and smartwatches for seamless indoor/outdoor coverage. Falls can happen anywhere and fast; accurate fall detection that provides immediate communication with the caregiver minimizes the cost of falls and maximizes the quality of life of PLWD and caregivers. Methods: Participants of this study include dyads of PLWD (ages 65 and experiencing mild to moderate cognitive impairment and at risk of falls) and caregivers, which are continuing to be enrolled on a rolling basis with a goal of reaching 60 dyads. Survey-based data collection with caregivers begins at baseline (when the fall detection system is installed) and continues monthly for one year to assess psychosocial and usability outcomes, including but not limited to technology usability, fear of falling, anxiety and depression, loneliness, and caregiver burden. Data will be analyzed using descriptive statistics and Wilcoxon signed-rank tests. This NIH-funded SBIR grant is a unique industry-academic-foundation partnership among Clairvoyant Networks Inc., Texas A&M University, and the UK Longitude Prize. Results: Initial data analyzed from ten dyads indicate participants' mental health, social connectedness, and burden scores remained stable across the first two months. PLWD were at fall risk and representative of those with mild cognitive impairment and mild dementia. Caregivers' mean system usability scale scores showed a nonsignificant increase from 65.5 to 69.8, indicating moderate usability. No statistically significant pre-post changes were observed in caregiver anxiety/depression, loneliness, or burden, nor in PLWD anxiety/depression, loneliness, or burden. Regarding technology usability, participants appear to find the system easy to use and would like to use the system frequently. Conclusion: Early results suggest that the fall detection system is usable and acceptable among PLWD-caregiver dyads, with stable psychosocial outcomes over six weeks. Continued data collection and longer follow-up will allow evaluation of the system's impact on falls, caregiver stress, and overall wellbeing.

Funding: 1) The project described was supported by Grant Number RAG076218A from the National Institutes of Health (NIH) awarded to Clairvoyant Networks, Inc., with Texas A&M being a subcontracted academic partner. 2) Clairvoyant Networks Inc received funding from the Longitude Prize on Dementia as a Semi-Finalist and Finalist Award winner. ; COI: Clairvoyant Networks Inc is a commercial company offering novel fall detection services built on its AI/neural network system. However, note that all data collection/analyses are independently conducted by its academic partner, Texas A&M University.

Poster # 154: DEMENTIA PREVALENCE AND URBAN CONFIGURATION: A SPACE SYNTAX ANALYSIS.

Presenting author: Cristian A. Maestre BSc (UTRGV)

Background: While advancing age is considered the greatest risk factor for dementia, an expanding body of research highlights the vital role social and environmental factors play. The current study looked at urban design characteristics in which population-based studies of dementia were conducted. Our main objective was to assess whether the reported prevalence of dementia was associated with the layout of spaces in the catchment areas where the studies took place. Methods A systematic review was conducted across the following databases: PubMed, EBSCO, Web of Science, and Cochrane to collect population-based studies related to the prevalence of dementia. The search yielded 69 articles and 123 distinct catchment areas spanning five continents. To analyze the spatial layout of these areas, the DepthmapX software was used to characterize the catchment area data. Two key space syntax measures were selected for the analysis: Angular Connectivity and Angular Depth, as they capture the principal characteristics of wayfinding and legibility within an urban environment. Results The analysis revealed a significant negative association between the prevalence of dementia and angular connectivity (β = -0.739, 95% CI: -1.389 to -0.090, p = 0.027). This statistical relationship indicated that catchment areas featuring higher angular connectivity tend to have a lower dementia prevalence. Furthermore, the prevalence of dementia showed a significant positive correlation with Angular Depth (β = 0.599, 95% CI: 0.113 to 1.086, p = 0.017). After adjustment for mean age and sex distribution, the association between dementia prevalence and Angular Connectivity remained consistent in its direction and statistically significant (β = -0.513, 95% CI: -0.962 to -0.063, p = 0.026). In contrast, the relationship between dementia prevalence and Angular Depth showed only a non-significant tendency after the adjustment. Conclusion These findings suggest that the specific urban layout of an urban environment may influence the prevalence of dementia at the population level. Areas characterized by high connectivity, and to a lesser extent, lower Angular Depth were associated with a lower dementia prevalence. This study highlights the role of urban configuration as a potential environmental determinant of cognitive health.

Funding: None reported.; COI: None reported.

Poster # 155: DESIGN THINKING AND INNER DEVELOPMENT SKILLS FOR DEMENTIA CAREGIVERS: A SUSTAINABLE FRAMEWORK TO REDUCE BURNOUT AND FOSTER INNOVATION IN CAREGIVING

Presenting author: Sylvia Azucena Robles, MBA, PhD (UTRGV)

Background: Alzheimer's disease and related dementias (ADRD) place increasing emotional and cognitive demands on caregivers, leading to high levels of stress and burnout. Despite the growing prevalence of ADRD, psychosocial support interventions for caregivers remain fragmented and limited in scalability. Building on evidence from the Vaqueros Innovation Platform (VIP) Idea Lab Design Thinking Workshop, which demonstrated how Design Thinking (DT), Inner Development Goals (IDGs), and Sustainable Development Goals (SDGs) frameworks cultivate empathy, creativity, problem-solving, and resilience among participants, this study proposes applying these human-centered innovation methods to strengthen the caregiving ecosystem for ADRD. Grounded in the Transactional Model of Stress and Coping and Resilience Theory, this research envisions a pathway to empower caregivers through co-created, technologyenabled interventions. Method: Drawing on mixed-methods data collected from over 700 participants of previous VIP Idea Lab workshops, key inner development competencies-empathy, mindfulness, collaboration, and reflective awareness-were mapped to the stages of the design thinking process. Qualitative insights revealed significant growth in participants' emotional intelligence, creative confidence, and problem-solving skills. These findings informed the conceptual development of a new caregiver-focused DT/IDGs framework designed to address psychosocial stressors and co-develop sustainable, AI-assisted solutions to support caregiver well-being and dementia care innovation. Results: Preliminary analysis suggests that integrating IDGs within design thinking promotes the development of cognitive and emotional capacities essential for resilience and adaptive coping. Participants reported higher levels of empathy, collaboration, and perceived purpose, suggesting strong transferability to caregiver training contexts. Conclusion: This conceptual framework lays the foundation for future empirical testing through an AD-RCMAR pilot intervention designed to enhance caregiver wellbeing and innovation capacity. By merging sustainability, inner development, and humancentered design, the study offers a transformative model for addressing ADRD caregiver burnout and improving the psychosocial environment of dementia care.

Funding: The VIP Idea Lab Design Thinking workshop is funded by the UTRGV Robert C. Vackar College of Business and Entrepreneurship Deanship. ; COI: The proposed study is novel, does not address a specific aim of another ongoing study, does not extend the scope of another study, and is not funded by another source.

Poster # 156: ORAL HEALTH, VITAMIN D, AND COGNITIVE FUNCTION IN LATE LIFE

Presenting author: Juan Carlos Lopez-Alvarenga (UTRGV-SOM)

Objective. To evaluate the association between serum 25-hydroxyvitamin D₃ [25(OH)D₃] levels, periodontal health, and cognitive performance in older U.S. adults, accounting for diabetes and systemic conditions. Methods. Cross-sectional data from 905 participants aged ≥ 60 years in NHANES 2011-2012 were analyzed. Periodontal indicators included clinical attachment loss (CAL), probing pocket depth (PPD), and cement-junction margin (CJM). Cognitive function was assessed with the Consortium to Establish a Registry for Alzheimer's Disease (CERAD) Word Recall, Animal Fluency, and Digit Symbol Substitution tests, summarized through principal component analysis (F1 = memory recall). Multivariable linear regression models were adjusted for age, sex, diabetes, BMI, serum creatinine, and education. Results. Diabetes prevalence was 23%. Higher CAL and PPD were independently associated with poorer memory recall (β ≈ -0.37; p < 0.001), whereas greater CJM was associated with better recall (β = 0.24; p < 0.001). Each 10 ng/mL increase in 25(OH)D₃ was linked to ~10% lower CAL and 8% lower PPD. Vitamin D's effect on cognition was attenuated after adjusting for education. Conclusions. Periodontal disease and low vitamin D levels were associated with reduced memory performance, particularly among individuals with diabetes. Findings support an oral-systemic-cognitive axis relevant to metabolic aging and dementia prevention.

Funding: NA; COI: None of the authors has a conflict of interest

Poster # 157: HOW LONELINESS AND SOCIAL ISOLATION ARE LINKED TO COGNITIVE DECLINE AMONG OLDER ADULTS? A SYSTEMATIC REVIEW OF UNDERLYING PHYSIOLOGICAL AND PSYCHOBEHAVIORAL MECHANISMS

Presenting author: Kexin Yu, PhD, LMSW (UTSW)

Background: Despite extensive evidence linking loneliness and social isolation to cognitive decline in later life, the underlying pathophysiological and psychosocial mechanisms remain unclear. This systematic review synthesizes existing research to clarify these pathways. Methods: A comprehensive search was conducted across PubMed, Web of Science, EMBASE, and PsycINFO until October 2024. Using Covidence for literature management, we included peer-reviewed studies in English involving human participants aged 50+. Studies examining psychosocial pathways required mediation analyses, whereas studies on physiological pathways were included if outcomes were related to dementia pathology. Two reviewers independently screened titles, abstracts, and full texts, extracted study characteristics, and assessed methodological quality across key domains (sample selection, measurement validity, confounding control, and analysis). Results: Thirty-eight empirical articles met inclusion criteria. Across loneliness studies, seven domains of physiological pathways (including Alzheimer's disease pathology, neurotrophic factors, brain structural changes, cardiovascular risks, inflammation, HPA-axis functioning, and general health) and eight psychobehavioral domains (depressive symptoms, psychological well-being, personality traits, cognitive and social engagement, selfperception, etc.) were identified. Social isolation studies examined six physiological and five psychobehavioral domains with overlapping but fewer mechanistic pathways. Depressive symptoms and lower gray matter volume most consistently mediated the association between loneliness and cognitive decline. Evidence also suggested moderating effects of brain-derived neurotrophic factor (BDNF) and inflammation (IL-6) on loneliness-related decline in specific cognitive domains. Although cross-sectional findings predominated, longitudinal studies supported that loneliness and isolation may precede neurobiological and behavioral changes contributing to cognitive impairment, rather than merely reflecting prodromal symptoms. Conclusion: This systematic review highlights converging physiological and psychobehavioral pathways linking loneliness and social isolation to cognitive decline, underscoring the multifactorial nature of these associations. Future studies should prioritize longitudinal and mechanistic designs integrating biomarkers, neuroimaging, and behavioral data to establish temporal sequences and causality. Findings inform intervention development targeting both social and neurobiological resilience pathways to mitigate cognitive decline in at-risk older adults.

Funding: This work was supported by the National Institutes of Health R01AG051628, R01AG056102, R01AG056712, R03AG081719, P30AG066518, T32AG049676, and K00AG068492. ; COI: None

Poster # 158: SOCIAL AND PSYCHOLOGICAL FACTORS OF AI ACCEPTANCE AMONG INFORMAL CAREGIVERS OF PEOPLE LIVING WITH DEMENTIA: A PILOT SURVEY STUDY

Presenting author: Louis Fisher, BS (Clemson University)

Background: Informal caregivers provide a substantial proportion of care for people living with dementia (PLwD) and frequently experience significant emotional, physical, and financial strain. These stressors place caregivers at elevated risk for anxiety, depression, and caregiver burden. Technology-based interventions, including mobile and web-based applications, have emerged as an accessible approach to support caregivers by offering informational resources, skills training, and mental or social support. As caregiving experiences are highly heterogeneous, artificial intelligence (AI)-enabled functions are increasingly integrated into digital platforms to deliver personalized recommendations and tailored assistance. However, hesitancy toward AI may serve as a barrier to adoption and sustained engagement. Understanding factors associated with AI acceptance is critical for informing the design of equitable and user-centered caregiver technologies. This study aims to assess levels of AI acceptance among adult informal caregivers of PLwD in the United States, and to examine psychological flexibility, social need, and selected demographic characteristics as potential determinants. Methods: Eligible Informal caregivers were recruited through online social media outlets. Participants completed validated measures assessing AI acceptance, psychological flexibility (PF), social needs, and demographic characteristics. Descriptive statistics were used to summarize participant characteristics. Correlational analyses and linear regressions were conducted to examine associations between AI acceptance and key determinants. Results: A total of 31 informal caregivers completed the survey. Participants had an average age of 60 years, were predominantly female (29/31, 93.5%), white (25/31, 80.6%), and educated (24/31, 77.4% completed some form of higher education). Most participants were actively providing care at the time of the survey (21/31, 67.7%). Mean PF and AI acceptance scores were moderate (66.4 and 3.01, respectively). Regression analyses indicated that perceived illness severity of the care recipient (p < 0.001), perceived financial status (p = 0.055), and the harnessing and acceptance subscales of the Personalized Psychological Flexibility Index were associated with AI acceptance. Conclusion: AI acceptance among informal caregivers of PLwD appears to be influenced by psychological flexibility, perceived illness severity, and financial strain. Digital platforms incorporating AI should consider these contextual factors and provide adaptable, responsive features that align with caregiver needs, resources, and readiness.

Funding: National Institute on Aging Small Business Innovative Research No. 2R44AG07411604 ; COI: None Declared

Poster # 159: A PUBLIC HEALTH APPROACH TO ALZHEIMER'S DISEASE AND RELATED DEMENTIA: TEXAS DSHS ALZHEIMER'S PROGRAM INITIATIVES

Presenting author: Megan Rowe, MPS (Texas Department of State Health Services)

Background: In Texas, nearly 12% of individuals aged 65 and older are living with Alzheimer's disease and an estimated 1.1 million family caregivers provide their essential care. As the number of Texans experiencing cognitive decline increases, so does the need for coordinated strategies to address the growing impact of Alzheimer's disease and related dementias (ADRD). Method: An implementation approach was used to design and launch the ADRD program initiatives within the Alzheimer's Disease Program at the Texas Department of State Health Services (DSHS). Program implementation was guided by the Texas State Plan for Alzheimer's Disease and the CDC's National Healthy Brain Road Map Initiative using a public health approach to address dementia. Result: Implementation efforts led to the development and establishment of a variety of strategic public health initiatives. DSHS launched Interprofessional Education Curricula to address education gaps for future healthcare providers. We partnered with Texas A&M AgriLife to increase public understanding of dementia and risk reduction strategies. The media awareness campaign continued and was expanded to include provider education and resources. DSHS revamped the statewide Alzheimer's Disease Partnership to increase stakeholder engagement and implemented a quarterly newsletter to improve collaboration. We launched a competitive grants program, the Grants to Increase Local Dementia Support (GILDS), to expand statewide capacity to provide coordinated outreach to families impacted by dementia. These activities strengthened statewide partnerships to advance early detection, diagnosis, risk reduction awareness, provider education, and community-based support. Statewide capacity was expanded to identify and connect individuals impacted by dementia and their caregivers through evidence-informed approaches and coordinated support. Together, these efforts enhanced Texas's foundation for sustainable, dementia-focused public health action. Conclusion: These efforts demonstrate how a coordinated public health approach can strengthen dementia infrastructure, partnerships, and community capacity across Texas. By aligning statewide activities with the Texas State Plan for Alzheimer's Disease 2024-2028 and the CDC's Healthy Brain Initiative Road Map, the program has advanced sustainable systems to address dementia through awareness, early detection and diagnosis, and community-based support. Continued collaboration and investment are essential to maintain this momentum, reduce dementia risk, and improve quality care for Texans impacted by dementia.

Funding: State Funds and the CDC BOLD Grant ; COI: N/A

Poster # 160: STRENGTHENING DEMENTIA COLLABORATION ACROSS TEXAS: TWO YEARS OF PROGRESS IN THE TEXAS ALZHEIMER'S DISEASE PARTNERSHIP

Presenting author: Vanessa Pierson and Stephanie Daly, LMSW (Texas Department of State Health Services)

Background: According to the CDC, one of the 10 Essential Public Health Services to improve health outcomes is strengthening, supporting, and mobilizing communities and partnerships. The volunteer-based Texas Alzheimer's Disease Partnership (Partnership) was established by the Department of State Health Services (DSHS) in 2009. Through CDC's Building our Largest Dementia Infrastructure (BOLD) Grant, DSHS implemented activities to expand Partnership membership and strengthen stakeholder engagement during grant years 1 (October 2023September 2024) and 2 (October 2024-September 2025). Methods: To increase program capacity and strengthen statewide coordination, DSHS designated dedicated staffing to implement BOLD grant strategies and facilitate Partnership activities. Outreach initiatives were developed to increase attendance at quarterly meetings and expand membership within priority Public Health Regions (PHRs)(South Texas, the Panhandle, and Northeast Texas). To enhance engagement, DSHS established topic-specific workgroups, introduced a quarterly newsletter, and incorporated educational presentations on topics of interest into meetings. Results: The Partnership grew by 67 members during grant years 1 and 2, including 31 new members from the three priority PHRs. Average quarterly meeting attendance increased from 77 to 93 participants, reflecting stronger engagement across the coalition. Six educational presentations were completed over grant years 1 and 2. Newsletter reach expanded substantially, from 441 recipients for the inaugural issue in November 2024 to 1,986 in August 2025, representing a 450% increase in one grant year. Conclusion: In 2024 - 2025, the Alzheimer's Disease Program expanded the Partnership by 67 members and increased attendance by an average of 16 members through targeted outreach and engagement activities. BOLD grant funding increased program capacity and provided the foundation needed to strengthen statewide infrastructure, support multisector collaboration, and enhance information sharing across Texas. These efforts demonstrate the impact of sustained investment and coordination in advancing public health response to Alzheimer's disease and related dementias.

Funding: CDC BOLD Grant Funding. ; COI: No conflicts of interest to disclose.

Poster # 161: INTEGRATION OF THE CMS GUIDE MODEL AND CLINICAL TRIALS WITHIN A NEW COMPREHENSIVE MEMORY CLINIC SERVING RURAL AREAS IN TEXAS

Presenting author: Alayna Jump, BS (Texas Tech HSC)

Background: Alzheimer's disease and Alzheimer's disease-related dementias (AD/ADRD) continue to rise across Texas. The longest and most complex disease trajectories occur among rural older adults, who face dual inequities: higher rates of AD/ADRD and lower access to timely diagnosis or coordinated care. Rural residents obtain diagnoses at later stages, when treatment options are limited because specialized clinicians (e.g., neurologists, neuropsychologists) are less accessible. These disparities undermine older adults' ability to live fully and independently, as well as heighten caregiver stress and economic insecurity. Method: Our team at TTUHSC Lubbock built a comprehensive, interdisciplinary memory clinic model including a behavioral neurologist, geriatric psychiatrist, neuropsychologist, social worker, speech-language pathologist, and other support staff. This model aimed to provide dementia screening, diagnoses, care planning, advanced care planning, patient and caregiver resources following diagnosis (including the CMS GUIDE program), and access to clinical trials. Result: Over the last year, our team has served over 210 families across 33 counties, 69% of which are rural. Every family received access to a comprehensive examination, care planning for the patient and care partner, advanced care planning, and was provided appropriate resources. Starting July 1, 2025, we enhanced our clinic by integrating the CMS GUIDE Model which provides free respite, an in-home safety assessment, medication management, and other services to persons with AD/ADRD and their caregivers. 68% of dyads have utilized in-home safety assessments and 100% of moderate-high complexity dyads have received access to in-home respite. Lastly, we currently have 4 nonpharmacological clinical trials that target suicidal ideation, pre-death grief, and caregiver burden that have reached over 80 participants and have been implemented in-person and via telehealth. Conclusion: Over the first year, our clinic demonstrated feasibility and acceptability in providing comprehensive care to persons with AD/ADRD and care partners, including implementation of the CMS GUIDE program. Our clinic has also been used as a model for other clinics around the country (e.g., OHSU; Stanford; TTUHSC El Paso) of how to implement the GUIDE program into a comprehensive memory clinic. Further, we will continue to implement clinical trials, including pharmacological trials.

Funding: Garrison Family Foundation; Pickering Family; COI: None

Poster # 162: LIFE REVIEW NONPHARMACOLOGICAL INTERVENTION DELIVERED BY CAREGIVERS OF PEOPLE LIVING WITH DEMENTIA IMPROVES DEPRESSION: A MIXED-METHODS STUDY

Presenting author: Christina E Miyawaki, PhD, MSW, MA (University of Houston)

Background: Depression rose from 6% to 23% in older Americans post-COVID-19 pandemic. Furthermore, 40% of them experience depression and dementia simultaneously, as they are the two most common neuropsychiatric disorders in older adults. Previous studies have found that nonpharmacological interventions work better for older adults. Life review therapy is a nonpharmacological, evidence-based depression intervention primarily for older adults, wherein participants discuss their lives in chronological order with a therapist. Seeing a therapist, however, is costly, stigmatizing, and subject to therapists' availability. To alleviate these barriers, we developed an innovative depression intervention, Caregiver-Provided Life Review (C-PLR) for people with depression and dementia (PwDD) and their family caregivers. We trained caregivers in life review skills, and caregivers conducted life reviews with their PwDD at home. This study reports the results and their implications. Method: This six-week intervention used a mixed-methods design with 45 PwDD-caregiver dyads (N=90). Eligible PwDD were 65+ years old and lived with mild depressive symptoms (≥ 5 Geriatric Depression Scale) and early-stage dementia (≥ 13 Telephone-Montreal Cognitive Assessment). Caregivers were PwDD's family members (≥ 18 years), cared for PwDD for ≥ 8 hours/week for ≥ 1 year with no depression and no severe caregiver burden. We measured PwDD's pre- and post-intervention depression (primary outcome), life satisfaction, caregivers' caregiving burden, rewards of caregiving, and dyads' relationship quality (secondary outcomes) levels using paired t-tests. Result: PwDD were on average 81 years old, widowed, retired, female, and in poor/fair health, while caregivers were 58 years old, married, college-educated, working, female, and in good/excellent health. PwDD's depression significantly improved (p<0.001), as did the rewards of caregiving (p=0.029), and relationship quality (p=0.041) without increasing caregiver burden (p=0.519). Caregivers' postintervention interviews supported the quantitative results. All caregivers confirmed the feasibility of training and confidence in conducting life reviews. Conclusion: Training family caregivers to conduct life reviews may be a way to improve PwDD's depression. C-PLR allows busy caregivers to complete training and life reviews with PwDD at home on their own schedule, and it extends nonpharmacological care for PwDD. Leveraging C-PLR's flexible schedule, non-stigmatizing setting, and cost-effective approach, C-PLR can potentially reach a wide range of PwDD caregiver dyads.

Funding: Alzheimer's Association Research Grant AARG-20-685255 & Alzheimer's Association Research Grant Rapid Program in Dementia Funding Grant AARG-20-685255-RAPID.; COI: All authors have no conflicts of interest.

Poster # 163: PREDICTORS OF UNSTEADINESS AND FALLS IN OLDER ADULTS WITH COGNITIVE IMPAIRMENT: INSIGHTS FROM A LARGE CLINICAL DATASET

Presenting author: Shaoqing Ge, PhD, MPH, RN (UT Austin)

Background: Falls and unsteadiness are critical safety concerns for older adults with diminished cognitive function, often leading to severe adverse outcomes. This study aims to identify factors contributing to unsteadiness or sustained falls among community-dwelling older adults. Methods: Data were analyzed from the BrainCheck Plan (n=18,837), administered by primary care and neurology providers during clinical visits to individuals and their caregivers upon identification of cognitive impairment. Impairment was determined by failure on both the BrainCheck Screen-a brief computerized screener-and BrainCheck Assess, which includes validated cognitive batteries (immediate and delayed recognition, Trail Making Test A and B, Stroop Color and Word Test, Digit Symbol Substitution). This tiered protocol ensures inclusion of participants with clinically significant cognitive concerns, making the dataset appropriate for research on older adults at elevated risk for dementia. Since 2017, the BrainCheck Plan workflow has been reimbursable under Medicare, supporting its sustained use in clinical settings. Of 25,462 available records, the analytic sample was restricted to individuals aged ≥ 40 years with complete data (n=18,837) to enhance data quality and relevance. Logistic regression was performed to identify factors associated with increased risk of unsteadiness or falls. The outcome variable was defined as older adults or their caregivers answering "yes" to the Alzheimer's Association safety assessment checklist item: "Has the patient experienced unsteadiness or sustained falls?" Predictor variables included demographic characteristics, caregiver relationship, living arrangement, anxiety (GAD score), activities of daily living (ADL), and dementia severity. Results: Significant predictors of increased risk included male gender (β=0.16, p<0.001), higher age (β=0.02, p<0.001), and elevated anxiety (GAD score; β=0.06, p<0.001). Protective factors included spousal relationship (β=-0.23, p<0.001), family/friend caregiver (β=-0.71, p<0.001), living with the patient (β=-0.23, p<0.001), greater ADL independence (β=-0.08, p<0.001), and confirmed dementia diagnosis (β=-0.45, p<0.001). Race (white or other/mixed) was not a significant predictor. Conclusion: These findings provide actionable guidance for clinical practice. Identifying male gender, older age, and anxiety as risk factors can help clinicians prioritize fall risk screening and prevention. Recognizing the protective effects of caregiver support and cohabitation suggests that strengthening social and living arrangements may reduce fall risk.

Funding: Author Shaoqing Ge is supported by the UT Austin School of Nursing Start Up Grant. Author [Bin Huang] is supported by the National Institute on Aging of the National Institutes of Health (award R44AG078006).; COI: Author B.H. received salary and stock options from BrainCheck, Inc.

Poster # 164: TRAINING TOMORROW'S DEMENTIA CARE TEAMS: OUTCOMES OF AN ONLINE INTERPROFESSIONAL TRAINING INITIATIVE

Presenting author: Alyssa Aguirre, LCSW (UT Dell)

Background: Health professional students require enhanced training in Alzheimer's Disease and Related Dementias (ADRD) and interprofessional collaboration to address workforce gaps in dementia-capable care. The goals of the study were to improve students' knowledge, attitudes, and collaborative skills in ADRD and interprofessional care, while centering the lived experiences of individuals, families and caregivers affected by dementia. This study developed and evaluated a scalable online interprofessional education (IPE) curriculum for students in nursing, medicine, pharmacy, social work, and communication at The University of Texas in Austin. Method: The curriculum included two asynchronous modules covering IPE principles, brain health, and early detection, followed by a virtual IPE Day featuring case-based, teamoriented learning. Pre- and post-surveys assessed ADRD knowledge, attitudes, and IPE competencies using validated scales. Qualitative feedback was collected through open-ended responses and analyzed thematically. Result: Of 58 students enrolled, 42 completed both modules and 30 participated in the IPE Day. Statistically significant improvements were observed in ADRD knowledge (mean increase = 1.43, p < .001) and preparedness to support families (1.49, p < .001) and all five IPE competency domains - communication (mean increase = 0.48, p = .003), collaboration (0.53, p < .001), roles and responsibilities (0.43, p = .005), patient-centered care (0.50, p = .002), and team functioning and conflict management (0.39, p = .001). Over 90 percent of participants agreed the curriculum supported their learning and professional development. Qualitative analysis revealed five themes: students valued active learning strategies, clinical relevance, and self-paced flexibility, while identifying opportunities for greater interprofessional integration and extended simulation time. Conclusion: This brief, scalable virtual IPE curriculum demonstrated feasibility and impact on enhancing dementia care competencies among health professions students. The significant improvements in both ADRD knowledge and interprofessional competencies, combined with high student satisfaction, support the effectiveness of this educational approach. These findings provide important implications for interprofessional curriculum design in ADRD education and training, demonstrating that online formats can prepare students for dementia-capable, collaborative practice.

Funding: Texas Department of State Health Services; COI: None

Poster # 165: A RURAL PERSPECTIVE: ANTI-AMYLOID SAFETY AND PRACTICAL CONSIDERATIONS

Presenting author: Elisabeth Burnett, MSN, FNP (UT Tyler)

Background: Anti-amyloid monoclonal antibodies such as lecanemab and donanemab mark a new era of disease-modifying therapy in Alzheimer's disease. However, most clinical guidelines and appropriate-use recommendations are developed in larger academic or urban centers, which may not be fully aligned with rural practice. A lack of educational resources and implementation support for community neurologists contributes to a lack of diagnostic accuracy and treatment safety. An understanding of these disparities is essential to ensuring equitable access to and outcomes in rural memory care. Methods: This is a retrospective analysis of patient charts (n = 6, April 2024 - October 2025) treated with either Lecanemab or Donanemab. All the patients were initiated on anti-amyloid therapies by general neurologists and transferred for care at the memory disorders clinic at the tertiary care center. Results: Four of six (67%) had mixed etiologies of vascular cognitive impairment (Fazekas 3) and/or Lewy body disease, despite antiamyloid initiation. One patient received Lecanemab while on Apixaban; therapy was suspended pending Watchman placement. Two of six (33%) developed symptomatic ARIA-E: one was discharged from the ED without MRI (diagnosed as sinusitis); the other had severe ARIAE/ARIA-H on donanemab without prior APOE genotyping. Conclusions: In this early rural implementation cohort, diagnostic overlap, limited biomarker access, and insufficient local education on anti-amyloid prescribing contributed to non-guideline-concordant therapy and preventable ARIA events. Targeted education for community neurologists, structured diagnostic pathways, and improved access to MRI and APOE testing are needed to promote safer, more equitable deployment of anti-amyloid therapies across rural populations.

Funding: None; COI: None

Poster # 166: DEPRESSION AND CAREGIVER BURDEN: THE MEDIATING ROLE OF HEALTH BELIEFS IN A MULTIETHNIC COHORT STUDY

Presenting author: Diamond Lee, MS (UTSW)

Background: Existing research indicates that caregivers of individuals with severe health conditions exhibit elevated levels of depression, adverse health beliefs, and stress associated with their caregiving responsibilities. While these factors have been extensively studied in isolation, limited attention has been given to the impact of caregivers' health beliefs and stress in the context of depressive symptoms. The present study investigates the influence of caregivers' health beliefs on perceived caregiving burden, with a focus on the role of experienced depression. Methods: Participants (N = 85; Mage = 61.2 years; Medu = 15.89 years; 83% male; 62.4% White, 16.5% Hispanic,13.3% Black) were drawn from a community-based, multiethnic cohort. All participants completed the Multidimensional Health Perceptions Questionnaire (MHPQ), the Zarit Burden Interview (ZBI), and the Patient Health Questionnaire (PHQ-9). A correlation analysis was conducted to assess the strength of associations between depression, healthcare beliefs, and perceived caregiver burden. Additionally, parallel meditation regression analyses examined how specific health beliefs influenced caregiver burden and depressive symptoms. (89) Results: Depression was found to have a small effect on various health beliefs, suggesting weak relationships between social-emotional healthcare beliefs (r = .27, p = .01, R2=.07), trust in medicine (r =.29, p < .01, R2 = .08), and health literacy (r = .27, p = .04, R2 =.07), in contrast to a large effect for caregiver burden (r = .58, p <.01, R2= .33). When controlling for one another, anticipated discrimination, spiritual beliefs, socio-emotional health care beliefs, trust in medicine, self-efficacy, health literacy, and trust in social health advice did not mediate the relationship between depression and caregiver burden. Conclusion: Findings revealed distinct relationships between depression and healthcare beliefs, while healthcare beliefs had minimal effects on caregiver burden in this sample. Although these beliefs did not mediate the link between depression and caregiver burden, future research should explore potential pathways through which psychological and belief-based factors impact caregiver outcomes.

Funding: None; COI: None

Poster # 167: STEROID RECEPTOR CO-ACTIVATOR STIMULATOR PROMOTES NEUROPROTECTION AND PRESYNAPTIC INTEGRITY IN TAUOPATHY MICE

Presenting author: Rohit K. Srivastava, PhD (BCM)

Introduction: Alzheimer's disease (AD) is the most common cause of dementia worldwide and represents a major unmet clinical challenge, with its incidence projected to rise sharply in the coming decades. The disease is characterized by the accumulation of extracellular amyloid-β (Aβ) plaques and intracellular neurofibrillary tangles composed of hyperphosphorylated tau (pTau), accompanied by progressive neuronal loss and cognitive decline. Mounting evidence indicates that tau hyperphosphorylation and aggregation are central drivers of neurodegeneration and disease progression, underscoring the urgent need for therapies that specifically target tau pathology. In this study, we evaluated the therapeutic efficacy of MCB-613-10-1 (10-1), a derivative of MCB-613 and a potent small-molecule stimulator of steroid receptor coactivators (SRCs), using the ps19 transgenic mouse model expressing the human P301S tau mutation. We investigated whether 10-1 treatment could attenuate tau-associated neurodegeneration, preserve synaptic integrity, and improve behavioral outcomes, providing translational insight into its potential as a disease-modifying therapy for Alzheimer's disease and related tauopathies. Method: Transgenic ps19 mice were administered 10-1 (20 mg/kg, five times per week) or saline for five months, beginning at four months of age, with age-matched wild-type (WT) mice serving as controls. At nine months, cortical brain regions were used for molecular and cellular analyses, including immunofluorescence, ELISA, western blotting, qPCR, and TUNEL assay to evaluate neurodegenerative and molecular changes. Behavioral assessments were performed to evaluate cognitive performance, exploratory activity, and anxiety-like behaviors. Results: Prolonged 101 administration significantly reduced cortical pTau accumulation in ps19 mice without affecting total tau levels. Treatment with 10-1 exerted pronounced neuroprotective effects, as evidenced by a marked decrease in TUNEL-positive neurons, indicating attenuation of apoptosis. Furthermore, 10-1 preserved synaptic integrity, reflected by sustained synaptophysin (SYP) expression, and mitigated neuroinflammation by reducing glial fibrillary acidic protein (GFAP) levels, indicative of decreased astrocytic activation. Behaviorally, 10-1-treated ps19 mice exhibited improved hippocampus-dependent learning and memory. Conclusion: These findings show that 10-1 suppresses several pathogenic pathways, including tau phosphorylation, apoptosis, synaptic dysfunction, and neuroinflammation in the brain. Collectively, these results identify 10-1 as a promising disease-modifying therapeutic candidate with translational potential for the treatment of tau-associated neurodegenerative disorders, including Alzheimer's disease and related tauopathies.

Funding: Funding has been provided by CoRegen, Inc. to develop SRC small molecule stimulators for clinical use.; COI: David M. Lonard and Bert W. O'Malley are founding members of a nonpublic company, called CoRegen Inc. David M. Lonard, Lisa K. McClendon, Rohit K. Srivastava, Adam M. Dean, Clifford C. Dacso, and Bert W. O'Malley have stock ownership in CoRegen.

Poster # 168: A NEW APPROACH FOR PARKINSON'S TREATMENT: NEUROPROTECTIVE NANOMATERIALS ALLEVIATE AMYLOID BURDEN, REDUCE CELLULAR STRESS, AND EXTEND LIFESPAN

Presenting author: Harris C. Brown (TAMU)

Background: Parkinson's Disease (PD) is the fastest growing neurological disorder. PD is characterized by the neurodegeneration of dopaminergic neurons in the substantia nigra of the midbrain. By 2040, an estimated 12 million patients will have PD. Current PD treatments only counteract classical symptoms but are not neuroprotective. The exact etiology of disease is not understood, but neurotoxic aggregation of alpha synuclein and mitochondrial dysfunction serve a major role. Previous studies have shown beneficial neuronal cellular responses to the administration of transition metal dichalcogenide (TMD) nanoflowers (NFs). These benefits include increased healthy neuronal cell proliferation, reduced mitochondrial damage and cellular ROS levels, induced mitochondrial biogenesis, and extended the lifespan of healthy C. elegans nematodes. Method- Neuronal cells were impregnated with preformed alpha synuclein aggregates and subsequently treated with a 10% total solution of TMD NFs. Experiments were conducted to examine the cellular health and therapeutic response. Three cell types were utilized: N27 rat neurons, DI TNC1 rat astrocytes, and CTX TNA2 rat astrocytes. The cellular health indicators used include ROS stress granule immunofluorescence, JC-1 mitochondrial biogenesis ELISA for mitochondrial health, and qPCR with custom rat primers to study genetic consequences of the therapeutic. Results- Mitochondrial impairment and biogenesis significantly improved in a dose-dependent manner. TMD NF treatments alleviated pathological cellular stress with dose-dependent reduction in reactive oxygen species. Reduction in cytosolic TIA-R stress granule protein. Dynamic transcriptional response over time to mitigate the effects of aggregated protein. MoSe2 administration resulted in a significant increase in C. Elegans lifespan, while MoS2 had minimal effect. Both MoS2 and MoSe2 NFs offered significant reduction in amyloid burden. Conclusion- Administration of TMD NFs in an induced PD model resulted in significant beneficial mitochondrial response in three neuronal cell types. MoS2 and MoSe2 NF treatments resulted in transcriptional responses contrary to incubation with amyloid aggregates. TMD NFs are a biocompatible nanotherapeutic that resulted in reduced amyloid burden and extension of lifespan in C. elegans. Overall, MoSe2 NFs have shown to be a more promising neurotherapeutic candidate in the treatment of PD.

Funding: NIGMS: R35GM142869; COI: No conflict of interest.

Poster # 169: THE UNIQUE MECHANISM OF PROTEASOME ACTIVATION WITH TAT PEPTIDES PROTECTS FROM AD RELATED COGNITIVE DECLINE.

Presenting author: Pawel A. Osmulski, PhD (UTH San Antonio)

Background: The proteasome as an essential protease of ubiquitin-proteasome pathway is the critical keeper of intracellular proteostasis. The pathway plays fundamental roles in key neuronal functions impacted in AD. We show that proteasome impairment in AD patients precedes other biomarkers' changes such as accumulation of amyloid β and phospho-tau. The compromised proteasome activity comes as a victim of AD-related intracellular alterations, but it can also act as a perpetrator of such changes in a vicious positive feedback loop. Our innovative approach calls for holistic improvement of proteostasis. Methods To achieve this goal, we developed peptidomimetic (peptide-like) leads, TAT-TOD and TAT-DEN, based on a viral proteasomeinteracting protein HIV-1 TAT modified with synthetic structural turn stabilizers. We tested the actions of our leads in silico, in vitro, in cellulo and in vivo. Results The leads bind to the catalytic core of the proteasome to the pocket used by natural regulatory proteins and activate both the core (20S) and the most physiologically relevant proteasome holoenzyme (26S) in vitro, in cultured cells, and in animal models (flies and mice). Moreover, in our in vitro studies TATDEN not only boosts catalytic activity: it also protects the core proteasome from binding the toxic oligomers of Aβ42; a capability not reported for other proteostasis-targeting drug candidates. Our structural studies suggest a unique mechanism of the proteasome - oligomeric Aβ42 - TAT-DEN interactions that breaks the vicious cycle of amyloid-dependent proteasome inhibition and proteostatic decline. Importantly, the non-toxic and non-immunogenic peptidomimetics pass the brain-blood barrier (BBB), are detectable in brains of AD model mice and protect against a cognitive decline. Conclusions We developed small peptidomimetics that (i) activate the proteasome in vitro and in vivo, (ii) protect the proteasome from the inhibition by Aβ42 oligomers with a unique molecular mechanism, (iii) pass the BBB and alleviate AD-like cognitive deficits in animal models with no detectable toxicity, (iv) mitigate survival impairments in AD models. The candidates are ready for further drug-relevant optimization and testing.

Funding: This work was supported by the National Institute of General Medical Science R01 GM069819 (to M.G.), National Institute of Aging R01 AG065301 (to A.M.P.), William and Ella Owens Medical Research Foundation (to M.G. and P.A.O.). ; COI: PAO, MG and AMP are inventors on a patent application related to this work filed by The University of Texas Health Science Center at San Antonio (HSC1567). The authors are also co-founders of ProAllostera Pharmaceuticals.

Poster # 170: THE UNIQUE MECHANISM OF PROTEASOME ACTIVATION WITH TAT PEPTIDES PROTECTS FROM AD RELATED COGNITIVE DECLINE.

Presenting author: Romeesa Khan, BS (UTHealth Houston)

Background: Stroke is the leading cause of long-term disability in the United States and a leading cause of death. Although advancements in acute treatment have increased stroke survival rates, more than 60% stroke survivors live with some degree of cognitive impairment. Secondary injury cascades lead to chronic increases in oxidative stress, non-resolving inflammation, and cellular dysfunction, which implicate cellular senescence as a key driver of chronic post-stroke neurodegeneration. We tested the hypothesis that targeting senescent cells via administration of the senolytic cocktail, dasatinib and quercetin (D+Q), will improve long-term recovery after stroke. Method: Middle-aged male and female C57BL/6 mice (14-15 mos) underwent transient (60-min) middle cerebral artery occlusion (tMCAO) or sham surgery. Starting at one-month post-tMCAO, mice were treated with D+Q (or vehicle) via oral gavage on a monthly basis for 3 months. Functional outcomes were assessed at multiple timepoints using a battery of behavioral tests. Inflammation and cellular senescence were assessed using flow cytometry, qPCR, and immunohistochemistry. Result: D+Q treatment significantly improved recognition memory in stroke mice at 2-months post-stroke, as assessed through the Novel Object Recognition Test (p=.036). D+Q-treated stroke mice also displayed significant improvement in spatial memory via the Radial Arm Water Maze test at 3-months post-stroke (p=.003). Enhanced associative learning was seen in the fear conditioning test, indicating D+Q has potent therapeutic effects on poststroke cognition. Additionally, we found that D+Q-treated mice had improved grip strength when compared to vehicle groups at both 2- and 3-months post-stroke. Flow cytometric analysis of blood showed decreased bone marrow myeloid skewing, blood monocytosis/neutrophilia, and inflammatory activation in D+Q-treated female stroke mice, compared to vehicle-treated females. After further examination, senescence-associated beta-galactosidase activity and age/injuryrelated expansion of HSCs were significantly attenuated in D+Q treated mice compared to vehicle controls. Conclusion: We report significant improvements in physiological, neuropathology, and cognitive function in stroke mice, when treated long-term with the combinatorial senolytic therapy D+Q, chronically after stroke. Our results indicate injury-induced senescence may be responsible for accelerated brain aging after stroke, and senolytic treatment may be a first to expand the treatment window to millions of stroke survivors, reducing risk of post-stroke dementia.

Funding: AHA Sacco Award; COI: N/A

Poster # 171: THERAPEUTIC TARGETING OF PATHOGENIC TAU ATTENUATES COGNITIVE DECLINE AND INFLAMMAGING IN AN ALZHEIMER'S MOUSE MODEL

Presenting author: Jia Yi Liew (UTMB)

Background: Alzheimer's disease (AD) is characterized by the accumulation of tau protein in the brain, which forms neurofibrillary tangles and contributes to the gradual deterioration of brain function. Consequently, cellular senescence occurs, leading to cognitive impairment and hastening the aging process. Immunotherapies targeting Aβ and other protein aggregates are also being developed in the meantime. This study looks at immunotherapy in a Mapt (hTau) mouse model. This could become a useful immunological treatment in the future by lowering inflammation and cellular senescence that come with getting older. Methods: Mapt (hTau) mice were inoculated in the hippocampus with 1 μg of brain-derived Tau oligomers (BDTOs) from AD patients. After seven months, mice were injected in the tail vein with 120 μg of in-house mouse monoclonal anti-toxic tau antibodies (TTCM1-2 and TOMA1-4) or IgG isotype control. Mice were evaluated for cognitive and motor function before euthanasia, and brain neuropathology was investigated with immunohistochemistry, immunofluorescence, and western blot techniques. Results: Treatment with anti-toxic tau antibodies reduced total tau levels and tau phosphorylation in the brain. Furthermore, immunotherapy reduced inflammation and cellular senescence markers. Conclusions: The observed decrease in tau aggregates following immunotherapy indicates that the mechanisms responsible for tau phosphorylation and/or aggregation may have been altered. Additionally, the decrease in senescence and inflammation markers suggests that immunotherapeutic strategies targeting tau are a promising approach in the management of tauopathies.

Funding: This research work was supported by Mitchell Center for Neurodegenerative Diseases, the Gillson Longenbaugh Foundation and National Institute of Health grants: (R01AG077253 to RK , R01AG054025 and U24AG072458 to RK); COI: N/A

Poster # 172: INFLUENCE ON AGE AND BIOLOGICAL SEX ON SHORT-TERM SPATIAL MEMORY IN THE SYRIAN HAMSTER (MESOCRICETUS AURATUS).

Presenting author: Laura Emma Garcia, MA (UTRGV)

Background: Memory is a key factor in many diseases, and memory impairment is a hallmark of Alzheimer's disease. Although substantial progress has been made in the memory literature towards the understanding of memory impacts in Alzheimer's disease, further research is needed to develop new animal models for age-related disorders. Place preference behavioral models have been used effectively in the literature to understand how animals encode and recall environmental cues (Hong et al., 2025). For our present study, we aim to utilize the Syrian hamster (Mesocricetus auratus) as a model of spatial memory at different ages. The Syrian Hamster has been identified in the literature as being a good model of memory due to the brain's similarities, social exploration, ease of handling and low cost (Warner et al., 2017). We expect hamsters to prefer spending time in zones containing bedding of opposite-sex odors, suggesting the information of a short-term memory that may be influenced by age. Method: Using a repeated measures design, with trials as the within-subjects factor and age and sex as between-subjects factors, three cohorts of males (n = 28) and females (n = 36) of different ages (PND 161-602), were tested using a 16-hole board apparatus. Animals were placed in the hole-board for five 3minute trials: Preexposure, Treatment/Exposure, and Memory test 1-3. The inter-trial intervals for tests 2 and 3 were 15 and 30 minutes, respectfully. The preexposure trial was completed to determine preference and expose hamsters to the apparatus. Results: There was a significant interaction between control and treatment trials with the time spent in the stimulus zone (6) F (4,208) = 2.838, p = .025. There was also a significant interaction between sex and trials F (1,52) = 4.982, p = .030, and a significant age*trials interaction with differences between the youngest and middle cohorts F (2,52) = 2.875, p = .020. Conclusion: Our results provide evidence that the Syrian hamster is an excellent animal model for the study of sex differences/aging in learning and memory. Experimental subjects spent more time in the stimulus zone compared to the controls, confirming short-term memory formation that was influenced by age.

Funding: This project is funded by grant #1K01HL145339-01A1 (PI: M. Gil); COI: No conflicts of interest.

Poster # 173: DEVELOPING PEPTIDE INHIBITORS TO BLOCK TAU AGGREGATION IN ALZHEIMER'S AND RELATED DISEASES

Presenting author: Ayde Mendoza-Oliva, PhD (UTSW)

Background: Tauopathies such as Alzheimer's disease (AD) are characterized by the abnormal assembly of tau protein into amyloid fibrils. Currently, there are no approved therapies that block tau aggregation. Our work focuses on developing peptide inhibitors that bind to the ends of growing tau fibrils to prevent their extension. We hypothesized that the strategic incorporation of large hydrophobic residues into tau would disrupt protofilament elongation. Furthermore, we predicted that longer tau sequences would exhibit higher binding affinity and more effectively block fibril growth. Methods: Using Rosetta structural modeling, we identified positions where hydrophobic substitutions were predicted to meet these criteria. We tested a panel of candidate sequences. We evaluated their ability to inhibit seeded aggregation induced by AD and corticobasal degeneration (CBD) brain homogenates in wild-type (WT) tau HEK biosensor cells. These cells express the tau repeat domain fused to FRET-compatible fluorescent proteins, enabling quantitative detection of seeded aggregation. Results: We identified specific tau substitutions that blocked tau aggregation in cells. After selecting one of these inhibitors, we defined the minimal sequence required to retain its activity, revealing a dose-dependent inhibitory effect. Using a WT tau neuronal biosensor system derived from induced pluripotent stem cells (iPSCs), we validated its activity. To evaluate its therapeutic potential, we synthesized the peptide fused to a cell-penetrating sequence, observing efficient internalization in HEK and neuronal cells. We are currently conducting studies with the synthetic inhibitor to characterize its activity, perform toxicity assays, and analyze its interaction with tau fibrils to elucidate its mechanism of action. We will also evaluate the gene expression of the inhibitor and its efficacy in reducing pathology in an AD murine model. Conclusion: Through a combined strategy of structural modeling, molecular engineering, and cellular screening, we identified a tau-based peptide inhibitor that prevents seeding and disrupts the maintenance of pathological aggregates. We are now optimizing peptide production and delivery strategies. This approach may lead to a new therapeutic avenue for AD and related tauopathies.

Funding: Junior Investigator Research Grant; COI: None

Poster # 174: LINKING CHRONIC SOCIAL ADVERSITY TO BIOLOGICAL AGING AND ALZHEIMER'S RISK: A SYSTEMATIC REVIEW IN HISPANIC AND BORDER POPULATIONS

Presenting author: Moamin Elhadi, BS (Texas Tech University Foster School of Medicine)

Background: Hispanic and Latino adults are among the fastest-growing aging groups in the United States and face a disproportionately higher burden of Alzheimer's disease (AD) than nonHispanic Whites. Yet most biomarker and longitudinal studies focus on non-Hispanic populations, leaving major mechanistic gaps in understanding AD disparities. Residents of U.S.Mexico border regions experience persistent structural adversities. Poverty, discrimination, acculturative stress, limited healthcare access, and environmental disadvantage that may independently accelerate biological aging and worsen cognitive decline. Chronic adversity can trigger inflammatory, metabolic, and epigenetic processes linked to neurodegeneration; however, the intersecting roles of ethnicity and border residence remain largely unexplored. To our knowledge, this is the first systematic review synthesizing evidence on chronic social adversity, biological aging, and cognitive decline in Hispanic and border-region adults. Method: A structured search of Ovid MEDLINE identified peer-reviewed human studies examining associations among chronic social adversity, biological aging indicators, and cognitive outcomes, including dementia, mild cognitive impairment, or domain-specific decline. Studies were included if they involved Hispanic, Latino, Mexican American, or border-resident adults and quantitatively assessed at least one chronic adversity measure in relation to a biological aging or cognitive outcome. Observational or interventional human studies published in English were included; non-human, pediatric, or acute-stress studies were excluded. Results: Twelve records met inclusion criteria, representing five unique studies. Across studies, chronic stress, socioeconomic disadvantage, and racialization were associated with elevated inflammation and metabolic dysregulation in Hispanic adults, predicting poorer cognitive performance and increased dementia risk. Inflammatory pathways partially mediated racialized disparities in dementia incidence, while stress-related biological aging markers correlated with cognitive deficits in large Hispanic cohorts. No studies directly measured epigenetic aging or telomere attrition in border populations, underscoring a critical mechanistic and geographic gap. Conclusion: Emerging yet consistent evidence links chronic social adversity to accelerated biological aging and cognitive decline in Hispanic adults. Distinguishing border context from ethnicity is essential to clarify social and biological pathways underlying AD disparities. Future research should include borderregion participants, multidimensional adversity metrics, and integrative biomarker designs to advance equitable, mechanistically informed Alzheimer's prevention across the U.S.-Mexico border.

Funding: None reported.; COI: None reported.

Poster # 175: HIGH BLOOD PRESSURE AMONG ADULTS WITH COGNITIVE DECLINE:
FINDINGS FROM THE TEXAS BEHAVIORAL RISK FACTOR SURVEILLANCE SYSTEM, 2011-2023

Presenting author: Joseph O'Donnell, MPH (Texas Health and Human Services Commission)

Background: High blood pressure (HBP) significantly increases the risk of cognitive decline (CD). Understanding the interaction between HBP and CD will help identify at-risk populations and inform interventions. Our aim is to estimate the prevalence of HBP among those living with CD and identify its significant predictors. Methods: Odd years of data (2011-2023) from 4,459 respondents aged ≥ 18 who reported having CD, were analyzed from the Texas Behavioral Risk Factor Surveillance System. HBP prevalence was examined by sociodemographic and healthrelated characteristics, including age, sex, race/ethnicity, education, insurance, income, employment, obesity, high cholesterol, cardiovascular disease (CVD), diabetes, and arthritis. Multivariable logistic regression models were used to estimate the odds ratio (OR) and 95% confidence interval (CI) of HBP by the aforementioned sociodemographic and chronic-related conditions. Results: Among those with CD, 57.8% reported HBP. Older age was significantly associated with HBP prevalence with 62.3% and 69.9% among adults aged 55-64 and ≥ 65, respectively (p<0.0001 for both). Black Texans (76.5%) had significantly higher HBP prevalence compared to White (57.0%), Hispanic (51.8%), and Other Texans (57.4%; p<0.0001 for all).

While adjusting for the aforementioned sociodemographic and chronic conditions, Black Texans with CD had 3.4 times higher odds of HBP compared to White Texans (95% CI: 1.8- 6.4). Texans aged ≥ 65 years had 2.3 times higher odds of HBP compared to those aged 18-44 (95% CI: 1.1- 4.5). Texans with CD that also reported obesity (OR= 1.8, 95% CI: 1.3- 2.5), CVD (OR= 1.8, 95% CI: 1.2- 2.7), high cholesterol (OR= 1.9, 95% CI: 1.4- 2.5), arthritis (OR= 1.7, 95% CI: 1.22.3), or diabetes (OR= 1.8, 95% CI: 1.3- 2.6) were significantly more likely to have HBP compared to those who do not. Conclusion: Findings show that about 6 in 10 Texans living with CD have HBP. Black Texans, those aged ≥ 45, those who have obesity, CVD, high cholesterol, diabetes, or arthritis, have significantly higher HBP odds. With increasing CD rates among Texans, and a majority having HBP, effective HBP management may help increase the quality of life among Texans living with CD and related conditions such as dementia and Alzheimer's disease.

Funding: None; COI: None

Poster # 176: INCREASING CARDIOVASCULAR DISEASE RISK PREDICTS FASTER COGNITIVE DECLINE: INSIGHTS FROM AN ONGOING LONGITUDINAL COHORT OF ADULTS IN RURAL WEST TEXAS

Presenting author: Chathurika Samudani Dhanasekara, MD, PhD (Texas Tech HSC)

Background: While Alzheimer's disease remains the leading cause of dementia, vascular and mixed dementias account for approximately 25% of cases. Moreover, cardiovascular diseases and dementias share common metabolic risk factors. We aimed to examine the effects of cardiovascular disease risk on cognitive decline in a secondary analysis of a longitudinal cohort of middle-aged to older adults. We hypothesized faster cognitive decline with worsening cardiovascular disease risk. Methods: Participants with > 2 visits in Project FRONTIER (Facing Rural Obstacles to healthcare Now Through Intervention, Education & Research), an ongoing longitudinal cohort study, were included. Repeatable Battery for the Assessment of Neuropsychological Status (RBANSTotal) was the primary outcome; exposure was the 10-year cardiovascular disease risk (ASCVDR10) computed using PREVENT Equations. We fit linear mixed-effects models in R (lmerTest) with random intercepts and a Mundlak between-within decomposition of ASCVDR10; fixed effects were time (years since baseline), between-person mean risk, within-person deviation, and their interactions with time, using Satterthwaite degrees of freedom. A fully adjusted model additionally included age, sex, ethnicity, years of education, and visit. Results: The base model included 437 participants, while the full model included 429 participants. Base model revealed that RBANSTotal decreased with time (β = -3.48, p <0.001). Coefficient for between-person mean risk was not significant, while within-person deviation was associated with the same-visit RBANSTotal (β = 1.59, p <0.001). Time x between-person mean risk interaction (β = 0.10, p = 0.014), suggested a less steep decline for subjects with a higher typical risk. Importantly, time × within-subject risk interaction was significant, with a large effect size (β = -0.97 per 1% annual risk increase, p <0.001), suggesting a steeper decline when a subject exceeds their individual average risk. The full model reproduced these effects; additionally, visit, male sex, and Hispanic ethnicity significantly predicted lower RBANSTotal, whereas education was protective. Age was not significant. Conclusions: Persistent cognitive decline was observed over time in aging adults, and within-person increases in ASCVDR10 independently predicted steeper deterioration. These findings highlight the importance of monitoring and modifying cardiovascular risk trajectories, not just baseline risk, to help preserve cognitive functions in aging adults.

Funding: This research did not receive any specific grants from funding agencies in the public, commercial, or not-for-profit sectors.; COI: No conflicts of interest to declare.

Poster # 177: RELATIONSHIPS BETWEEN SLEEP QUALITY AND NUTRITIONAL INDICES AMONG AGING RURAL WEST TEXANS: A PROJECT FRONTIER STUDY

Presenting author: Luis Serrano-Rubio, BS (Texas Tech HSC)

Background: Sleep disturbances, including poor sleep quality, have been linked to increased Alzheimer's disease (AD) risk. Obstructive sleep apnea (OSA), a prevalent sleep disorder associated with obesity and anthropometric risk factors, disrupts sleep quality and may accelerate AD pathological changes. The Nutritional Risk Index (NRI) integrates protein status (albumin) and body mass, while the Prognostic Nutritional Index (PNI) combines protein status with immune function (lymphocyte count). Both markers have been associated with cognitive decline risk, yet their relationships with OSA-related variables remain unclear. In this study, we explored whether sleep-related variables (OSA risk factors and daytime sleepiness) and anthropometric measures correlated with nutritional status as assessed by NRI and PNI. Methods The sample included 470 participants from Project FRONTIER in rural West Texas (mean age 55±12 years; 66% female, 74% Hispanics). Primary outcomes were NRI and PNI. Predictors were item-level responses from the Berlin Questionnaire Sleep Evaluation (BQSE) and Epworth Sleepiness Scale (ESS), the Mallampati score (MPI), and anthropometrics [BMI (kg/m²) and abdominal circumference (cm)]. Data was analyzed using IBM SPSS and we used two-tailed Spearman correlations (α=0.05). Results Multiple sleep-related variables correlated with NRI.
Anthropometric measures showed the strongest associations: BMI (ρ=0.938, p<0.001), abdominal circumference (ρ=0.647, p<0.001), neck circumference (ρ=0.186, p<0.001), and Mallampati score (ρ=0.114, p<0.001). OSA-related variables also correlated with NRI, including Category 3 high-risk classification (ρ=0.374, p<0.001), Category 1 high-risk classification (ρ=0.150, p<0.001), high blood pressure (ρ=0.128, p=0.002), snoring frequency (ρ=-0.129, p=0.002), and witnessed apneas (ρ=-0.097, p=0.017). ESS items did not correlate with NRI. In contrast, few variables correlated with PNI. Anthropometric measures showed weak negative associations: BMI (ρ=-0.128, p<0.001) and abdominal circumference (ρ=-0.101, p<0.001). Only neck circumference (ρ=-0.089, p=0.032) and dozing while sitting and talking (ESS Q6; ρ=0.096, p=0.019) showed significant correlations. Conclusion Positive associations between obesityrelated measures and NRI, contrasted with negative associations with PNI, suggest these indices capture distinct aspects of nutrition. NRI's associations with anatomical OSA risk factors (neck circumference, Mallampati score) and sleep symptoms (snoring, witnessed apneas) suggest it could be relevant for examining obesity-mediated links between sleep quality and cognitive health in at-risk populations in West Texas.

Funding: R01 AG071859, R01 AG073826, P30AG066546, TTUHSC MSSRP; COI: no conflicts to disclose

Poster # 178: RETHINKING BILINGUAL COGNITIVE RESERVE: A MEMORY-DRIVEN PATHWAY TO PRESERVED COGNITION IN AGING

Presenting author: Yinan Xu, MA (University of Houston)

Bilingualism has been proposed as a potential cognitive reserve factor that mitigates age-related cognitive decline. Managing two active languages requires constant monitoring, selection, and inhibition, thereby engaging and strengthening executive control processes over time. The present study examined whether bilingualism supports cognitive flexibility and inhibitory control in aging. Data includes 50 middle-aged women (26 bilinguals, 24 monolinguals) whose ages did not significantly differ between groups. Participants completed the NIH Toolbox cognitive assessments, the Montreal Cognitive Assessment (MoCA), and a Shape-Color task measuring cognitive flexibility. Drift Diffusion Modeling (DDM) was used to calculate performance into drift rates representing switch cost. No main group differences were found in Shape-Color performance or flexibility measures. However, bilinguals showed higher MoCA delayed recall scores, and a stronger relationship between memory performance and reduced switch costs, suggesting tighter coupling between episodic memory and cognitive flexibility in bilinguals. These results imply that bilingualism may preserve cognitive efficiency in aging not just through cognitive flexibility but an enhanced integration across memory and control systems. More importantly, our results highlight that the protective effects of bilingualism in aging may stem more from memory-related mechanisms than from the traditionally emphasized inhibitory control.

Funding: NIH; COI: N/A

Poster # 179: HEARING LOSS IN THE YOUNG-OLD INCREASES RISK FOR ALZHEIMER'S DISEASE AND VASCULAR DEMENTIA

Presenting author: Lily Francis MD, PhD (UTH San Antonio)

Background: Hearing loss is a risk factor for dementia, but dementia subtypes underlying this association and effect modifiers are unclear. Here we examined the association of hearing loss with all cause dementia, Alzheimer's disease and vascular dementia, and effect modifiers of this association in the Framingham Study. Method: Framingham Study participants >60 years, who underwent Pure Tone Audiometry at the 15th biennial exam (1977-1979), and had subsequent dementia surveillance for up to 40 years were included. We defined hearing loss as pure-tone average (0.5-4.0 kHz) > 25 dB in the better ear and used National Health and Aging Trends Study (NHATS) categories: normal (range: 0 - 25dB), mild hearing loss (range: 26 - 40dB), moderate to severe hearing loss (>40dB). All cause dementia diagnosed using DSM IV criteria, Alzheimer's disease diagnosed using National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer's Disease and Related Disorders Association (NINCDS-ADRDA) criteria, vascular dementia diagnosed using the National Institute of Neurological Disorders and Stroke and the Association Internationale pour la Recherche et l'Enseignement en Neurosciences (NINDS-AIREN) NINDS- AIREN criteria. Result: Among 2000 participants (mean age 69.5 years (IQR 60.04-88.74), 60% women, followed for up to four decades, we identified 633 participants with new onset dementia. Moderate to severe hearing loss was associated with 30% increased risk of dementia (Alzheimer's disease HR 1.27 [95%CI 0.99-1.62] p=.06; vascular dementia: HR 1.7[CI 1.06-2.59] p=.027). Age stratified analysis revealed the effect was restricted to persons aged <70 years at hearing assessment (all cause dementia: HR 1.56[CI 1.19-2.05] p=.001; Alzheimer's disease: HR 1.46[CI 1.07-2.0] p=.017; vascular dementia: HR 2.08[CI 1.22-3.56] p=.007) and strongest in APOE-ε4 carriers (all cause dementia (ε4 positive: HR 2.99 [CI 1.77-5.06] p=<.0001, ε4 negative: HR 1.16 [CI 0.85-1.58] p=0.36); Alzheimer's disease (ε4 positive: HR 3.44 [CI 1.92-6.17] p=<.0001, ε4 negative: HR 1.09 [CI 0.78-1.54] p=.64)). Conclusion: Hearing loss is associated with increased risk of all cause dementia, Alzheimer's disease and vascular dementia among Framingham Study participants, with greatest risk in those aged <70 years and in APOE-ε4 carriers. Screening for and managing hearing loss starting in mid-life might reduce dementia risk.

Funding: This research was supported by the National Institute on Aging AG 054076(SS), AG0081225(SS), P30 AG066546(SS), 5R01 AG066524(SS), AG049607(SS), N01HC25195-41-01(SS), 268201500001I-4-26800001-1(SS), 75N92019D00031-0-759201900126-1(SS). ; COI: None

Poster # 180: SLEEP, COGNITION, AND DEPRESSION: SEX-SPECIFIC ANALYSES FROM THE FRAMINGHAM HEART STUDY

Presenting author: Samantha Gates, BA (UTH San Antonio)

Background: Sleep patterns exhibit marked sex differences, with females disproportionately reporting sleep disturbances and extremes in duration. Similarly, depression is more prevalent in females, frequently co-occurs with sleep disturbances, and may modify sleep-cognition relationships. Despite this evidence, sex-specific interactions among sleep duration, depression severity, and treatment remain underexplored. This analysis extends prior work (Young et al., 2025) by examining sex-specific associations between sleep duration and cognition, and the moderating role of depression. Method: The sample included middle-aged adults (mean age ≈ 50 years) in the Third-Generation, Omni 2, and New Offspring Framingham Heart Study cohorts. Participants were a priori sex-stratified and categorized into three sleep groups: short (≤ 6h), average (>6 to <9h; reference), and long (≥ 9h). Depression status was defined using CES-D scores (≥ 16) and antidepressant use, yielding four groups: control (no depression/no medication), medication only (CES-D <16/antidepressant use), depression only (CES-D ≥ 16/no medication), and both conditions (CES-D ≥ 16/antidepressant use). We conducted multivariable linear regressions adjusted for age, education, comorbidities, and lifestyle factors. Interactions between sleep duration and depression status were tested. Results: In males, long sleep duration was significantly associated with reduced performance in visual reproduction (β±SE= -1.64±0.75, p = 0.030) and logical memory (-2.05±1.04, p = 0.049). Notably, untreated depressive symptoms in males sleeping ≤ 6 hours were linked to better similarities test performance (2.53±1.03, p = 0.018), suggesting a complex interplay between sleep, cognition, and depression. In females, long sleep duration was associated with worse global cognition (-0.27±0.08, p = 0.001), slower trail B (-0.13±0.04, p = 0.002), and poorer visual reproduction (-1.80±0.51, p = <0.001). In stratification analyses, females with long sleep duration, antidepressant use, and depressive symptoms exhibited deficits in global cognition (-1.05±0.29, p = 0.001) and visual reproduction (-5.78 ±1.96, p = 0.006). Conclusion: Long sleep duration is a sex-differentiated marker of cognitive risk, with females exhibiting heightened vulnerability, particularly when depression and antidepressant use coincide. These findings emphasize the need for sex-specific research and targeted clinical interventions to optimize sleep duration and mitigate cognitive decline.

Funding: P30 AG066546, R01AG077472, R01 AG054076, R01 AG049607, U01 AG052409, R01 AG059421, RF1 AG063507, RF1 AG066524, U01 AG058589; COI: None

Poster # 181: RACIAL AND ETHNIC DIFFERENCES IN THE ASSOCIATION BETWEEN HEARING LOSS AND COGNITIVE OUTCOMES IN THE HABS-HD COHORT

Presenting author: Matthew Sun, BS (UTHealth Houston)

Background: As people age, their risk for cognitive impairment and sensory loss increases. Recent research has linked these conditions, identifying hearing loss as a modifiable risk factor for dementia in some populations. However, its contribution to dementia risk in diverse populations is unknown. Method: Using data from an ethnically diverse cohort (n = 1,349) in Dallas, Texas, enrolled in the Health and Aging Brain Study - Health Disparities (HABS-HD), we conducted a cross-sectional analysis into the association between hearing impairment and cognitive outcomes, examining potential ethnic differences. Participants were classified as follows. Those who self-reported being advised by a physician to use hearing aids but did not do so were classified as having uncorrected hearing loss (UC). Individuals who were not advised to use hearing aids or who adhered to physician recommendations were classified as having normal or corrected hearing (NC). Result: Hearing loss rates were similar across ethnic groups. For White participants (n = 451), 8.0% reported UC hearing loss; for Black participants (n = 532), 4.7%; and for Hispanic participants (n = 366), 6.0%. In logistic regression models adjusted for demographic, socioeconomic, and health covariates, White participants with UC hearing loss had significantly higher odds of cognitive impairment (OR = 2.72, 95% CI 1.16-6.13, p = 0.018) and depression/anxiety (OR = 2.25, 95% CI 1.09-4.70, p = 0.029) compared to those with NC hearing. In Black or Hispanic participants, there was no significant association between hearing status and either outcome measure. In adjusted linear regression models, there were no associations between hearing status and perceived social support across all ethnic groups; however, the association approached significance in the White group (β = -1.9, 95% CI -4.0-0.26, p = 0.085), suggesting a trend toward lower perceived social support among White participants with UC hearing loss. Conclusion: Among White participants, uncorrected hearing loss was associated with poorer cognitive and emotional outcomes. No similar associations were observed in Black or Hispanic participants, though limited sample sizes in these groups may have reduced power to detect these effects. These findings highlight the importance of including diverse populations in future dementia research.

Funding: Self; COI: None reported.

Poster # 182: SEPSIS ASSOCIATED DELIRIUM INCREASES RISK OF DEMENTIA: A
MULTICENTER RETROSPECTIVE COHORT STUDY USING A REAL-WORLD DATABASE TRINETX

Presenting author: Lucineia Danielski, PharmD (UTMB)

Background: Delirium is a frequent complication among critical patients, particularly in sepsis patients. Although traditionally considered a transient condition, delirium may signal acute brain dysfunction with long-term consequences. Whether delirium following sepsis is associated with an increased risk of cognitive decline and dementia remains insufficiently explored. Objective: We aimed to evaluate the association between post-sepsis delirium and the subsequent risk of dementia in a large real-world cohort (TriNetX). Methods: We performed a retrospective cohort study using TriNetX data from 67 U.S. healthcare organizations, including adults (≥ 18 years) hospitalized with sepsis between 2000 and 2025. Two cohorts were defined: sepsis with delirium and sepsis without delirium, excluding patients with prior cognitive impairment. After matching a 1:1 propensity score (72,454 per group), outcomes assessed were all-cause mortality and incident cognitive diagnoses (mild cognitive impairment, Alzheimer's disease, dementia). Results: Sepsis complicated by delirium was associated with substantially worse long-term outcomes. Patients experiencing delirium during sepsis had a 77% higher risk of death within 1 year and this excess risk, though attenuated, remained significant even after 15 years (48% higher). Beyond mortality, delirium during sepsis predicted persistent cognitive decline: the risk of early-onset Alzheimer's disease was more than doubled and remained elevated for up to 15 years. Similarly, the risk of late-onset Alzheimer's disease, mild cognitive impairment, and unspecified dementia was consistently 1.7-2.5 times higher compared with sepsis patients without delirium. Conclusions: Sepsis complicated by delirium was associated with a persistent increase in long-term mortality and dementia risk, nearly doubling the likelihood of cognitive decline up to 15 years after the acute event.

Funding: None; COI: None

Poster # 183: AGING ATHLETES' CONCERN FOR COGNITIVE DECLINE AND CHRONIC TRAUMATIC ENCEPHALOPATHY (CTE)

Presenting author: Robbie Magill, BA (UTSW)

Background: Traumatic brain injury is a risk factor for developing dementia, and repetitive head hits have also been putatively associated with the neuropathological condition chronic traumatic encephalopathy (CTE). Injury-related factors, such as loss of consciousness (LOC) and number of concussions, may heighten risk for cognitive decline later in life. Concerns for both cognitive decline and CTE are common in former professional American-style football players, but the prevalence of these concerns in other aging athlete (particularly female) populations is unknown. Method: 1,141 former collegiate athletes aged 41-87 (m=60.8) completed surveys about concussions and concerns about cognitive decline and developing CTE for the College Level Aging Athlete Study (CLEAATS). Concerns for cognitive decline and CTE were assessed via asking: "Currently, are you concerned that the concussions or hits to the head you sustained have had or will have a permanent effect on your memory and/or thinking skills as you get older?" and "Currently, are you concerned that the concussions you sustained will cause you to develop chronic traumatic encephalopathy (CTE)?" Participants were grouped by number of diagnosed concussions (≤ 2,>2), sex, and +/-LOC history. Chi-square tests of independence (χ2) with Cramér's V effect sizes were used to examine associations between these variables and cognitive concerns. Result: 494 (43%) participants reported concern about future cognitive decline, with males (54%) endorsing significantly more concern than females (33%; p<.001,V=.21). Those with more concussions (>2) expressed greater concern for cognitive decline than those with ≤ 2 (p<.001,V=.25), and those with +LOC history reported greater concern than those without (p<.001,V=.18). 395 (34.6%) participants reported specific concern about developing CTE, with males (49%) endorsing significantly more concern than females (21%;p<.001,V=.30). Those with more concussions (>2) also expressed greater concern for CTE than those with ≤ 2 (p<.001,V=.22), and those with +LOC history reported greater concern than those without (p<.001,V=.17). Conclusion: Concerns for cognitive decline and CTE were prevalent among this sample of aging athletes, especially for males with repeated concussions and LOC history. Since preserving brain health is a common worry among older adults, these concerns may manifest in psychological distress for aging athletes.

Funding: Funded by the Darrell K Royal Foundation; COI: None

Poster # 184: A NOVEL COMPREHENSIVE DEMENTIA CARE MODEL FOR RURAL EAST TEXAS

Presenting author: Khushboo Verma, MD (UT Tyler)

Background: Rural patients face prolonged waits for specialty dementia care and fragmented services, amplifying caregiver burden and delaying timely intervention. East Texas lacks an integrated pathway that links primary-care, rapid pre-visit workup, and definitive behavioral neurology assessment within a single, coordinated model. Methods: Prospective implementation at UT Health North Campus, Tyler. The pathway comprises: - Referral optimization: standardized completeness metrics, feedback to PCPs, and specialty-visit targets from referral (T0): ≤ 8 weeks (Year 1), ≤ 6 weeks (Year 2). - Integrated pre-visit (≤ 2-3 weeks from T0): demographics and SDOH; cognitive screens (MoCA/MMSE, SMCQ, FAQ); mood (PHQ-4 ± PHQ-9/GAD-7); sleep (ISI-7, STOP-Bang); therapy needs (PT/OT/SLP/audiology); caregiver burden (ZBI-8); goal setting/barriers; immediate lifestyle counseling and referrals. Behavioral neurology visit: complete diagnostic workup (serum labs, MRI, neuropsychology) and, when appropriate, biomarker-guided anti-amyloid eligibility counseling (PET amyloid/APOE). All data are captured in REDCap; PDSA microcycles occur every 6 months with run/control charts. Pre-specified success thresholds (Months 0-24): ≥ 90% pathway completion ≤ 8 weeks, mean referral completeness ≥ 4.5/5, ≥ 3-point ZBI-8 reduction, ≥ 85% SDOH linkage, and Net Promoter Score ≥ +50. Results: We will report feasibility, fidelity to time targets, changes in referral completeness, caregiver burden, SDOH linkage rates, patientreported outcomes, and stakeholder experience. Iterative PDSA changes (e.g., EMR hard-stops, dedicated MRI blocks, navigator scripting) will be described with pre/post process-control analyses. Conclusions: EDCC operationalizes a practical, exportable pathway for rural dementia care by compressing time-to-diagnosis, activating lifestyle and support services during the wait, and embedding continuous QI. If benchmarks are achieved, we will disseminate a living implementation toolkit to enable multisite replication across similar underserved regions.

Funding: None; COI: None

Poster # 185: EARLY UPTAKE OF BREXPIPRAZOLE FOR AGITATION IN ALZHEIMER'S DISEASE: MEDICARE CLAIMS ANALYSIS IN TEXAS, 2021-2023

Presenting author: Youngran Kim, PhD (UTHealth Houston)

Background: Antipsychotics have been widely used off-label to manage behavioral and psychological symptoms in individuals with Alzheimer's disease and Alzheimer's disease-related dementias (AD/ADRD), particularly in nursing homes, despite safety concerns and the lack of regulatory approval for this indication. In May 2023, brexpiprazole, which was originally approved in 2015 for schizophrenia and as adjunctive therapy for major depressive disorder, became the first medication approved by the U.S. Food and Drug Administration (FDA) for agitation associated with dementia due to AD. Understanding early adoption of brexpiprazole may provide insight into clinical uptake of this newly approved therapy and inform ongoing policy and safety monitoring efforts. This study examines changes in brexpiprazole prescribing among Medicare fee-for-service (FFS) beneficiaries with AD/ADRD in Texas from 2021-2023. Methods: We identified Medicare FFS beneficiaries in Texas with AD/ADRD using ICD-10 diagnosis codes who were enrolled in Medicare Part D during any year from 2021-2023. Using Part D pharmacy claims, we identified brexpiprazole prescriptions and compared annual claim counts across years to assess early prescribing trends surrounding the May 2023 FDA approval. Results: Brexpiprazole use among Medicare beneficiaries with AD/ADRD remained low overall. There were 874 claims in 2021 and 864 claims in 2022 (pre-indication), followed by a notable increase to 2,317 claims in 2023, representing nearly a three-fold rise after FDA approval. Conclusion: Although the absolute number of prescriptions remained low relative to the AD/ADRD population, brexpiprazole use increased substantially after FDA approval. Continued monitoring is warranted to assess prescribing patterns, clinical appropriateness, and patient safety as uptake expands, particularly given known risks associated with antipsychotic use in older adults with dementia.

Funding: TARCC Junior Investigator Research Grant and NIH R21AG091282; COI: None

Poster # 186: HOW DO MEDICARE ANNUAL WELLNESS VISITS CONTRIBUTE TO EARLY DIAGNOSES OF MILD COGNITIVE IMPAIRMENT AND ALZHEIMER'S DISEASE & RELATED DEMENTIA?

Presenting author: Mukaila Raji, MD, MS, F.A.C.P. (UTMB)

Background: Early recognition of mild cognitive impairment (MCI) and timely diagnoses of Alzheimer's Disease and Related Dementias (ADRD) are key to optimal dementia care. No data exist on how healthcare interventions/services (blood tests, specialist referrals, and brain imaging) conducted after annual wellness visits (AWVs) contribute to subsequent MCI/ADRD diagnosis. We assessed the role of post-AWV healthcare services on the association of incident AWVs with the first ADRD or MCI diagnosis among Texas Medicare enrollees. Methods: We used 2015 to 2019 Medicare administrative data for patients aged 68 or older in 2018. We excluded prior ADRD or MCI diagnoses or AWV service in 2015-2017. Beneficiaries with and without AWV (n=66,443 in each group) in 2018 were matched 1:1 by propensity scores based on demographics, comorbidities, and prior healthcare utilization. Healthcare services (specific blood tests: vitaminB12, thyroid stimulating hormone; specialist referrals: neurology, psychiatry; and brain images: CT, MRI) occurring after AWV and before the first MCI/ADRD diagnosis were examined. The outcomes, in the first year after the AWV index date, were MCI diagnosis, ADRD diagnosis, and MCI or ADRD diagnosis. Results: Both groups had similar proportion of patients receiving specific blood tests, specialist referrals, and brain images in the 12 months before AWV/index date. Among those with no prior specialist visits, 4.1% in AWV group and 3.5% in non-AWV group received specialist referrals in the follow-up period. Among those with no previous brain images, 12.4% in the AWV group and 11.7% in the non-AWV group received brain images in the follow-up period. AWV receipt was associated with a 41% increase in MCI diagnosis; mediation analyses indicated that about 20% of this increase was explained through post-AWV laboratory testing, specialist referrals, and brain imaging. AWV receipt was associated with an 8% increase in ADRD diagnosis, of which about 75% was explained through these post-AWV healthcare services. Conclusion: This study showed a modest increase in post-AWV healthcare service use. Medicare AWV had larger direct effect on MCI than on ADRD diagnosis. ADRD diagnosis largely reflected post-AWV tests/referrals. However, the cognitive assessment and clinical evaluations during AWV directly contributed to the early recognition of MCI.

Funding: This work was supported by a grant from the National Institute on Aging, National Institutes of Health (Grant No. 1R01AG083102-01; PIs: Tzeng, Kuo, & Raji); COI: None

Poster # 187: EXAMINING THE COGNITIVE AND PSYCHOLOGICAL BENEFITS OF A VIRTUAL REALITY-BASED COGNITIVE TRAINING PROGRAM FOR OLDER ADULTS

Presenting author: Yongseop Kim, PhD (University of Utah)

Virtual reality (VR)-based cognitive training has emerged as an innovative approach to promoting cognitive health and engagement among older adults. By providing immersive, interactive environments that stimulate visuospatial and attentional processes, VR training may enhance both objective cognitive performance and self-perceived cognitive confidence. This pilot study examined pre-post changes in visuospatial reasoning, memory confidence, and psychological well-being following participation in an 12-session VR-based cognitive training program. Methods: Twenty-seven community-dwelling older adults completed a 6-week (12-session) VRbased cognitive training program designed to strengthen visuospatial, attentional, and executive skills through embodied and interactive VR tasks. Measures included the Memory and Cognitive Confidence Scale (MACCS) for self-perceived memory and cognitive confidence, the Vandenberg Mental Rotation Test for visuospatial ability, the Montreal Cognitive Assessment (MoCA) for global and executive cognitive functioning, and the Patient Health Questionnaire-8 (PHQ-8) for depressive symptoms. We analyzed data using paired-sample t-tests, and effect sizes were calculated using Cohen's d. Results: Significant improvements were observed in visuospatial reasoning (t(26) = 2.83, p = .0089, d = 0.54) and memory confidence (t(26) = −2.80, p = .0096, d = 0.54), reflecting moderate effect sizes. Participants demonstrated modest but nonsignificant gains in MoCA total scores (M change = +0.63, p = .212) and visuospatial-executive subscale (ΔM = +0.11, p = .416), and a slight decrease in depressive symptoms (ΔM = −0.07, p = .626). These findings suggest that visuospatial performance and cognitive self-confidence were the most responsive domains to short-term VR-based cognitive training. Conclusion: This pilot study provides preliminary evidence that a 6-week, 12-session VR-based cognitive training program can enhance visuospatial ability and self-perceived cognitive confidence among older adults. Although global cognition and mood did not significantly change, the moderate effect sizes support the potential of VR as an engaging and feasible modality for cognitive enrichment. Future studies should include larger samples, extended durations, and follow-up assessments to examine the sustainability and mechanisms underlying VR-based cognitive improvements in aging populations.

Funding: None.; COI: No conflict of interest.

Poster # 188: HEALTHY BEHAVIOR PATTERNS AND DIFFERENTIAL DEMENTIA RISK AMONG OLDER ADULTS WITH MILD COGNITIVE IMPAIRMENT

Presenting author: Jungjoo Lee, PhD (Texas A&M UHSC)

Background: Existing research highlights that older adults with mild cognitive impairment (MCI) face an elevated risk of dementia, emphasizing the need for non-pharmacological prevention. This study investigated how three types of activity differentially affect Alzheimer's Disease and related dementia (AD/ADRD) risk. Method The Health and Retirement Study data (2012 to 2020) were used, and participants who met the MCI criteria based on the Telephone Interview for Cognitive Status 27 were extracted (n = 4,207). Participation in activities (i.e., cognitively stimulating activity (CSA), physical activity (PA), and community-based activity (CBA)) was assessed by a six-point Likert scale, and the average level between 2012 and 2020 was used as an input variable. The outcome variable was a self-reported physician diagnosis of AD or ADRD. Participants who reported a new diagnosis of Alzheimer's Disease or a related dementia between 2012 and 2020 were included as cases. (n = 130) A Cox proportional hazards regression was employed to examine the dose-response relationship between each type of activity participation and AD/ADRD risk. Regression models adjusted for residence type (community or nursing home), cognition, functionality (ADL/IADL), and chronic conditions (diabetes, heart attack, stroke, depression). Results CSA, PA, and CBA participation were each inversely associated with AD/ADRD risk, though with distinct patterns and magnitudes. CSA showed the most robust reduction effect, reaching about an 8% lower risk at the highest participation level (95% CI: 0.860.98). PA showed a linear and steady decline in risk of approximately 3% (95% CI: 0.94-0.99). In contrast, CBA demonstrated a rapid reduction even at minimal participation, once every other week, corresponding to about a 1.6% decrease in risk (95% CI: 0.96-0.99), suggesting an early and efficient protective effect. Conclusion The greatest risk reduction was observed for participation in CSA, followed by PA and CBA. These findings suggest that therapeutic or behavioral interventions can benefit from focusing on CSA while integrating other types of health behaviors to enhance protective effects on cognitive health. A combined approach, rather than reliance on a single activity type, can achieve optimal cognitive benefits by generating reciprocal and complementary influences that collectively mitigate dementia risk

Funding: Nothing to disclose.; COI: Nothing to disclose.

Poster # 189: VITAMIN A SUPPLEMENTATION REDUCES ANXIETY AND IMPROVES SPATIAL COGNITION IN THE J20 MOUSE MODEL OF ALZHEIMER'S DISEASE

Presenting author: Adam Baker, MS (Texas Tech HSC)

Background: Preclinical and human studies converge on the vitamin A (VA) metabolite all-trans retinoic acid (ATRA) promoting nonamyloidogenic processing of amyloid precursor protein (APP), potentially accounting for the neuroprotective effects of the Mediterranean diet in delaying Alzheimer's disease (AD) onset. In a previous clinical study, VA plasma levels in individuals diagnosed with mild cognitive impairment (MCI) were significantly lower compared to age-matched unimpaired controls. We hypothesized that long-term VA supplementation would protect against cognitive decline and AD-related pathology in an AD mouse model. Methods: We utilized the J20 AD mouse model, which overexpresses human APP with familial AD-linked mutations. In this model, amyloid beta (Aꞵ) plaque deposition is well-established in the hippocampus and increases with age. J20 AD mice were weaned onto either a VA-supplemented diet (VAS, 20 IU/g) or a control diet (VAN, 4 IU/g) and underwent a behavioral test battery at 14±1 months. Behavioral performance was evaluated using the Open Field Test (OFT), Novel Object Location/Recognition tasks (NOL/NOR), and Water T-maze (WTM). HPLC-MS was used to detect ATRA in serum. Results: In the OFT, J20 mice exhibited increased locomotion and decreased time spent in the center compared with wild-type (WT) controls (p<0.001, 3-way RM ANOVA), consistent with an anxiety-like phenotype. WT mice on the VAS diet spent more time in the center on the final trial compared to the VAN group (p=0.013), suggesting an anxiolytic effect of VAS. During the NOL/NOR trials, discrimination indices did not differ significantly between experimental groups; however, surprisingly, J20 VAN mice spent significantly more time in the center than J20 VAS mice, suggesting a normalization towards WT by VAS. In the WTM, J20 VAS mice showed a shorter path length than VAN J20 mice (p=0.003) during simple discrimination, indicating improved spatial learning. Confocal imaging revealed abundant plaques in J20 mice at 14 months, though further analyses are ongoing to determine how VAS effects Aꞵ plaque burden and neurodegeneration. Conclusion: Collectively, these findings support the potential of dietary VA sufficiency to mitigate cognitive deficits and reduce amyloid pathology in a transgenic AD model.

Funding: NIH R01 AG073826-01A1 NIH R01 AG071859-01A1; COI: NA

Poster # 190: SMARTPHONE BASED GAIT METRICS IN NATURALISTIC SETTINGS
DIFFERENTIATE OLDER ADULTS WITH AND WITHOUT FALL HISTORIES

Presenting author: Polina Lee (UT Austin)

Background: Fall-related injuries are a significant cause of morbidity and mortality among older adults worldwide, especially those with histories of cognitive impairment and dementia. Assessments of gait and balance in this population typically consist of qualitative analysis of provider-conducted observations of gait instability or self-reported history of falls. However, the predictive ability of these assessments has been inconsistent across studies, and no single assessment in isolation can predict fall risk with high certainty. Machine learning analysis of gait patterns has been a topic of interest due to its potential to identify subtle walking features predictive of fall risk. However, many of these studies are limited by in-clinic assessments and instrumentation. The purpose of this study was to identify metrics associated with a history of falls from passively collected, smartphone-based gait metrics. Methods: 43 older adults enrolled in a passive smartphone sensing study had the TECHSANS app installed on their phone, and were monitored for an average of 154 days. Features measuring location, physical activity, and device usage were recorded by the user's phone. The patients were divided into 3 groups: those who had never fallen, fallen once, and fallen twice in the past year. Results: The 10 pedometer and gyroscope features demonstrated pairwise differences between all three fall history groups, with maximum walking speed and walking duration displaying the strongest contrasts. Random Forest (RF) models were trained to predict patient fall history based on these metrics using k-fold cross-validation (k=10). At the 95% confidence level, the Area Under the Curve (AUC) was 0.94 ± 0.0005, the Area Under the Precision-Recall Curve (AUC-PR) was 0.88 ± 0.0008, and balanced accuracy was 0.69 ± 0.0010. Conclusions: Passive gait metrics collected using smartphones in uncontrolled settings hold the potential for differentiating older individuals with a history of falls from those without a history of falls. Further work is required to determine whether passive gait metrics can identify older individuals who are likely to fall in the future, especially those with cognitive impairments.

Funding: This work is supported by The National Institutes of Health (R01AG077017).; COI: None reported.

Poster # 191: ASSOCIATION BETWEEN THE GUT BACTERIAL METABOLOME AND COGNITIVE IMPAIRMENT IN TEXAS POPULATIONS

Presenting author: Yannick J Wadop Ngouongo, PhD (UTH San Antonio)

Background: Growing evidence suggests that the gut microbiome contributes to cognitive aging and Alzheimer's disease (AD) through complex gut-brain axis interactions. However, the biological mechanisms linking microbial activity to dementia remain poorly understood. Gutderived metabolites-products of bacterial metabolic activity-may influence systemic inflammation, vascular health, and neurodegeneration, providing a plausible mechanistic pathway. Yet, few studies have examined these relationships in diverse populations. This study aimed to identify fecal metabolites that differ between individuals with mild cognitive impairment (MCI) or dementia and cognitively normal controls (NC) within the Texas Alzheimer's Research and Care Consortium (TARCC). Methods: In this cross-sectional study, we included 152 participants (mean age 72.1 years; 63% female) from TARCC and community cohorts. Stool samples were collected at home using OmniMet tubes (DNA Genotek). Untargeted liquid chromatography-mass spectrometry (LC-MS) metabolomic profiling was conducted by Metabolon. Differential abundance analyses were performed using MaAsLin2, adjusting for age, sex, body mass index, and race/ethnicity. Results: Several fecal metabolites differed significantly between participants with MCI/dementia and NC (fold change > 1, adjusted p < 0.06). Long-chain polyunsaturated fatty acids including docosahexaenoate (DHA; 22:6n3) and docosadienoate (22:2n6) were reduced in MCI/dementia, as were medium-chain fatty acids such as caprylate (8:0) and laurate (12:0). Dicarboxylate compounds were also decreased in those with cognitive impairment. In contrast, endocannabinoid metabolites oleoyl ethanolamide (18:1) and stearoyl ethanolamide (18:0) were elevated in MCI/dementia. Additionally, metabolites from vitamin A metabolism were reduced, while those involved in histidine and lysine metabolism were increased in MCI/dementia. Conclusion: Our findings highlight several gut bacterial metabolic pathways associated with cognitive impairment, particularly those related to lipid and amino acid metabolism. These alterations suggest potential mechanisms linking gut metabolic dysregulation to neurodegeneration. Further replication and mechanistic studies are warranted to determine whether these pathways represent modifiable targets for dementia prevention.

Funding: None reported.; COI: No conflicts

Poster # 192: ALZHEIMER DEMENTIA AND EXERCISE: A REVIEW ON RECENT FINDINGS OF PROTECTIVE LIFESTYLE MODIFICATIONS

Presenting author: Pierce Bassett, BS (UTMB)

Background: Alzheimer's disease (AD) is a progressive neurodegenerative disorder which insidiously impacts the functionality, memory, and cognition of more than seven million Americans, including an estimated one-third of those over the age of 851. While current pharmacological therapies inhibit the enzymes responsible for the breakdown of acetylcholine, these treatments are limited in their efficacy and in preventing the progression of AD. Because of this, new interest has emerged which focuses on the impact of exercise and healthy lifestyle factors in reducing risk, delaying onset, and slowing progression of AD1. Goals: The goal of this analysis is to review recent literature regarding the role of physical activity and modifiable lifestyle factors in the prevention and progression of Alzheimer's dementia. Methods: A keyword search was performed of medical literature using the terms "Alzheimer's dementia" and "exercise". Relevant studies from peer-reviewed journals published in the last ten years were included. Results: Recent studies have shown that physical activity can reduce the risk of AD. One study found that participants who underwent strength training twice a week saw a reversal of mild cognitive impairment along with improved hippocampal volumes2. Additionally, those who underwent moderate aerobic exercise in one retrospective observational study not only saw a mitigated risk of AD, but also saw the risk reduced by 35% in persons with a genetic predisposition for AD3. Conclusions: The data support the critical role of physical activity and lifestyle interventions in reducing the risk and progression of Alzheimer's disease while emphasizing the potential for synergistic opportunities with both targeted pharmacological and non-pharmacologic strategies in AD prevention and management. Future research should focus on optimizing exercise routines and dietary management in combination with what we already know about AD to ensure more positive outcomes. References will be included on Poster.

Funding: None; COI: None

Poster # 193: TOWARDS DATA DRIVEN PHENOTYPIC CLUSTERING OF FRONTOTEMPORAL DEMENTIA SUBTYPES

Presenting author: Satwant Kumar (UT Tyler)

Background: Categorical Frontotemporal Dementia diagnoses (bvFTD, PPA variants, PSP/CBS, FTD-MND, rtFTD) overlap and evolve with stage. Unsupervised methods applied to structured, stage-labeled phenotype knowledge may expose latent structure that complements traditional labels. Methods: We assembled a stage-resolved phenotypic knowledge base
(Early/Middle/Late) comprising n=438 information units (146 per stage) linking Subtype × Symptom/Domain × Temporal Stage with associated region/pathology descriptors. For each stage, units were embedded and partitioned into k=4 clusters. Cluster composition by Subtype was summarized as proportions; binomial tests compared within-cluster subtype proportions to stage-level prevalence with Benjamini-Hochberg FDR correction. Primary endpoint: reproducible, FDR-significant subtype enrichment across stages; secondary endpoints: cluster sizes and leading subtype proportions. Results: Across all three stages, one cluster showed FDR-significant enrichment of PPA variants (svPPA, nfvPPA, lvPPA): 1. Early, Cluster 1
(n=37): svPPA 35.1%, nfvPPA 32.4%, lvPPA 32.4%; q=0.00080-0.00146. 2. Middle, Cluster 3
(n=35): svPPA 37.1%, nfvPPA 34.3%, lvPPA 28.6%; q=0.00038-0.010. 3. Late, Cluster 2 (n=32): svPPA 40.6%, lvPPA 31.3%, nfvPPA 28.1%; q=0.00012-0.039. Non-PPA subtypes predominated nominally in other clusters without FDR-significant enrichment, with stage-specific leaders including bvFTD/rtFTD (e.g., Middle-0; Late-1) and PSP/CBS or FTD-MND (e.g., Late0/Late-3). Cluster sizes (Early/Middle/Late, Clusters 0-3): [27,37,41,41] / [42,37,32,35] / [23,52,32,39]. Conclusions: Unsupervised analysis of stage-labeled phenotype information (not patient-level outcomes) consistently isolates a language-dominant PPA cluster with robust, FDRcorrected enrichment across stages, suggesting a stable latent phenotype axis that complements categorical diagnoses. Nominal, stage-dependent predominance of non-PPA subtypes in other clusters further supports heterogeneous trajectories. Findings generate hypotheses for trial stratification and longitudinal monitoring, but require validation on patient-level datasets with imaging/biomarkers to confirm generalizability.

Funding: None; COI: None

Poster # 194: VFS-NET: A VERTEX AND FACE-BASED DEEP LEARNING PIPELINE FOR 3D BRAIN MODEL SEGMENTATION AND ANALYSIS IN ALZHEIMER'S DISEASE

Presenting author: Morgan Mekale Smith, MS (UTH San Antonio)

Background: Neurodegenerative diseases, such as Alzheimer's disease (AD), are characterized by progressive regional loss. In neuropathology, three-dimensional (3D) model segmentation and analysis via magnetic resonance imaging (MRI) is the gold standard for quantifying cortical atrophy. MRI requires access to specialized facilities, advanced technical training, neuroimaging expertise, and is time-consuming. To address these challenges, we have developed a pipeline leveraging structured-light scanning and deep learning to investigate structural patterns of atrophy associated with AD. Method: 3D files of postmortem human brains were decimated and manually segmented using the Blender application. Models were segmented into several relevant neuroanatomical regions and exported as .csv files as input into the Python-based deep learning pipeline. A neural network model was trained on 3D models of brains with minimal cortical atrophy, tracking classification metrics of accuracy, precision, loss, recall, the number of trainable parameters, and F1-score. Segmentation performance was evaluated using mean intersection over union (mIoU), per class intersection over union (pIoU), boundary intersection over union (bIoU), dice coefficient, and chamfer distance (CD), allowing visualization and quantification of predicted regional atrophy. Surface area measurements from regions of interest were also obtained to assign quartiles of atrophy when compared to controls. Result: Our results demonstrate that VFS-Net outperformed traditional machine learning methods in both classification and segmentation metrics. This suggests that using this pipeline, 3D models obtained from structured-light scanning can be classified and segmented with favorable evaluation metrics (70-98%). Conclusion: Our findings support the utilization of this pipeline to generate quartiles of regional atrophy associated with complex disease processes, including AD. Future work will validate these findings with cortical thickness measurements obtained from MRI upon the integration of two-dimensional images within the 3D models as volumetric data.

Funding: This work is supported by funding from the National Institute on Aging (P30-AG066546; R01-AG070214), Texas Alzheimer's Research & Care Consortium, Zachry Endowment for Alzheimer Research, Reed Precision Medicine Center, and the J.M.R. Barker Foundation. ; COI: N/A

Poster # 195: GRPA-PRS: A RISK STRATIFICATION METHOD TO IDENTIFY GENETICALLY-REGULATED PATHWAYS IN POLYGENIC DISEASES

Presenting author: Yulin Dai, PhD (UTHealth Houston)

Background: Polygenic risk scores (PRS) estimate genetic liability to complex diseases by aggregating numerous common variants. However, some high-PRS individuals remain diseasefree (resilience), whereas some low-PRS individuals develop disease (extra-burden), indicating the existence of orthogonal effects that genetic factors are independent of PRS that counteract or amplify predicted risk. Characterizing these effects may uncover mechanisms that modulate disease penetrance, offering insights into disease mechanisms and early clinical interventions. Methods: We developed the genetically regulated pathways-polygenic risk score (GRPa-PRS) framework to identify genetically regulated pathway activities (GRPas) through PRS-based stratification, applied to Alzheimer's disease (AD; 2,722 participants in the discovery and 2,854 in the replication cohort) and schizophrenia (SCZ; 6,628 participants in the discovery and 5,334 in the replication cohort). We calculated PRS models, stratified individuals by risk and diagnosis, and analyzed differential GRPas within the biological meaningful subgroups. To verify the orthogonal effect for potential resilience and extra-burden, we reformulated the standard Pearson correlation test into a Two One-Sided Tests (TOST) equivalence procedure to determine whether the correlation between PRS and GRPa is statistically equivalent to zero within a margin delta. For AD, analyses were further conducted with and without APOE effects, and across APOE haplotype. We further benchmark our framework against another individual variant-based method PRSet for functional enrichment of genetic risks. Results: In AD, we identified several wellknown AD-related pathways, including amyloid-beta clearance, tau protein binding, and astrocyte response to oxidative stress. Additionally, we discovered resilience-related GRPs, such as the calcium signaling pathway and divalent inorganic cation homeostasis. Stratifying APOE haplotype does not show new differential pathways among AD and control groups, suggesting no APOE haplotype-specific effects. In SCZ, fewer significant GRPas were detected, with mitochondrial and muscle development pathways emerging. Compared to other approaches, GRPa-PRS consistently identified more robust and biologically meaningful pathway differences. Conclusions: Our GRPa-PRS framework offers enhanced flexibility for exploring GRPas among subgroups stratified by PRS with either summary statistics or individual-level data. Fewer GRPas identified in (no-APOE) AD model and SCZ suggest a more polygenic architecture without dominant effects (like APOE), requiring larger samples to detect significant GRPas.

Funding: R21AG087299; COI: None