Catherine C. Kaczorowski

Summary: Single-cell patch-proteomics is a novel and powerful technique combining electrophysiological and quantitative proteomic characterization from individually sampled neurons. Following whole-cell patch clamp characterization of neuronal excitability, we collect intracellular contents and, using the latest advancements in Liquid Chromatography and Mass Spectrometry, identify and quantify peptides and their corresponding proteins. By cell-matched one-to-one associations between electrophysiological properties and protein abundance, this approach provides a means to elucidate molecular mechanisms of neuronal function on individual neuron basis. Graphical

Single-cell patch-proteomics is a novel and powerful technique combining electrophysiological and quantitative proteomic characterization from individually sampled neurons. Following whole-cell patch clamp characterization of neuronal excitability, we collect intracellular contents and, using the latest advancements in Liquid Chromatography and Mass Spectrometry, identify and quantify peptides and their corresponding proteins. By cell-matched one-to-one associations between electrophysiological properties and protein abundance, this approach provides a means to elucidate molecular mechanisms of neuronal function on individual neuron basis.

Temporal dynamics in functional connectomes provide a physiologically grounded signature of ′hidden′ pathologies during preclinical stages of Alzheimer′s disease (AD). We evaluated the effect of beta-amyloid (Aβ) on dynamic functional connectomes in transgenic mice and human subjects. Functional magnetic resonance images (fMRI) were collected in two strains of Aβ mice. fMRI-derived connectomes were segmented into discrete states using a hidden Markov model, and network strength, efficiency, and transitivity were analyzed per state. Human fMRI-derived connectome measures were analyzed across 3 states. Static network measures were significantly different between Aβ mice and controls, the former having high values for strength, efficiency and clustering coefficient in anterior cingulate, hippocampus, and retrosplenium. Dynamic network measures were stable within-states in Aβ mice. Similarly, human subjects with high Aβ had high node strength in precuneus and temporoparietal areas compared to low Aβ. Conversely, high Aβ was associated with high switch rates, high fractional occupancy, and state dwell times. Also, global strength, efficiency, and transitivity were less stable within states in the high Aβ group. Our results indicate that static, but not dynamic, connectome strength, efficiency, and network integration are increased in Aβ mice, while dynamic network states appear less stable in human functional connectomes. This data supports a dissociable, species-specific impact of Aβ, with dynamic network alterations present in humans but not in Aβ mouse models, suggesting additional non-Aβ-driven influences on dynamic functional connectivity in preclinical AD.

Alzheimer’s disease has widespread effects on brain structure, function and behavior, but we lack a systematic dissection of its impact across hundreds of forebrain and brainstem regions. Here, using diffusion tensor MRI at 25 µm, we mapped the global consequences of mutations in APP and PSEN1 across 231 regions of interest (ROIs) in male and female 5×FAD BXD hybrid mice at 14 months. Over half of the ROIs change in volume along rostrocaudal and mediolateral axes of the CNS, with unexpected swelling in the neocortex, hippocampus and amygdala of up to 10%, counterbalanced by shrinkage in the thalamus, brainstem and most white matter tracts. Yet, total brain volume is unaltered. Variation in individual ROI volumes is highest in females. Differences in fear acquisition and contextual memory performance covary with volumes of several regions and can have opposite polarities between cases and controls. These structural benchmarks establish a foundation for testing therapeutic interventions in preclinical trials.

Understanding the molecular mechanisms underlying cognitive resilience in Alzheimer’s disease (AD) is essential for identifying novel drivers of preserved cognitive function despite neuropathology. Rather than directly searching for individual genetic factors, we focus on latent factors and deep learning modeling as a systems-level approach to capture coordinated transcriptomic patterns and address the problem of missing heritability. We developed a conditional-gaussian mixture variational autoencoder (C-GMVAE) that integrates single-cell transcriptomic data with behavioral phenotypes from a genetically diverse BXD mouse population carrying 5XFAD mutations. This framework learns a structured latent space that captures biologically meaningful variation linked to cognitive resilience. The resulting latent variables are highly heritable and reflect genetically regulated molecular programs. By projecting samples along phenotype-aligned axes in the latent space, we obtain continuous gradients of genetic features associated with AD cognitive resilience. These findings highlight the potential of latent variable approaches not only to model high-dimensional biological data but also to reveal hidden factors driving phenotypic variability in neurodegeneration.

INTRODUCTION: Our previous work established the AD‐BXD mouse panel as an innovative model for studying the genetic complexity and heterogeneity underlying Alzheimer's disease (AD). In this study, we leveraged this model and proteomics approach to identify protein signatures linked to cognitive resilience in AD. METHODS: We assessed cognitive performance in 6‐month‐old AD‐BXD mice using contextual fear conditioning and calculated a quantitative resilience score. Frontal cortex proteomes were analyzed using data‐independent acquisition mass spectrometry. Protein quantitative trait loci (pQTL) mapping and transcription factor motif analysis were performed. RESULTS: Cognitive resilience was highly heritable. Of nine pQTL proteins associated with resilience, eight mapped to a shared locus on chromosome 1, forming a genetically regulated module. This module mediates the link between specific SNPs and cognitive outcomes in AD. DISCUSSION: These findings reveal a protein network underlying resilience to early AD pathology, with Nr1d1 emerging as a key transcriptional regulator. Highlights: Quantitative proteomics analysis using the genetically diverse AD‐BXD mouse panel identified nine proteins whose abundances are under genetic control and associated with differences in cognitive response to early‐onset AD mutations. Genetic mapping revealed that the abundances of eight of these nine proteins are regulated by the same genomic region on chromosome 1, forming a module that mediates the relationship between specific SNPs and cognitive outcomes in response to AD mutations. Integrating proteomic and genomic data suggests that SNP rs46128598, located in the transcription factor binding site of nuclear receptor subfamily 1 group D member 1 (Nr1d1), is a novel effector of metabolic pathways involved in cognitive performance differences in AD mutant carriers. This study indicates that targeting the molecular drivers of genetic resilience to AD mutations, including Nr1d1‐mediated proteins (Ak1a1, Gars1, Nudt3, Ogdh, Ptpn11, Iars2, Uba1, Ppt1, and Tmem223), could lead to new therapeutic approaches to delay the onset and/or progression of AD.

INTRODUCTION: Developing effective therapeutics for Alzheimer's disease (AD) requires a better understanding of the molecular drivers of the disease. Our previous work nominated DLGAP2 as a modifier of age‐related cognitive decline and risk for AD. We tested the hypothesis that overexpression of DLGAP2 in the hippocampus would protect against cognitive and synaptic deficits in a susceptible F1 5XFAD model. METHODS: DLGAP2 was overexpressed in the hippocampus of F1 hybrid 5XFAD and non‐transgenic littermates using a viral approach. Cognitive function, electrophysiological properties, and dendritic spine morphology were assessed at 6 and 14 months of age. RESULTS: DLGAP2 overexpression impaired synaptic plasticity and exacerbated AD‐related memory deficits but had minimal effect on spine structure or intrinsic neuronal properties. DISCUSSION: We highlight the complex role of DLGAP2 in AD pathology. Targeted interventions involving postsynaptic proteins must consider potential adverse effects on synaptic integrity and cognitive performance, particularly in the context of AD. Highlights: DLGAP2 overexpression accelerates AD‐related impairment of contextual fear acquisition and memory. DLGAP2 overexpression impairs synaptic plasticity prior to AD‐related memory impairment, but not intrinsic excitability. Effect of DLGAP2 overexpression on thin spine density was blunted in AD mice from in vivo dendritic spine results that were replicated in cultured rodent neurons.

INTRODUCTION: Developing effective therapeutics for Alzheimer’s Disease (AD) requires a better understanding of the molecular drivers of the disease. Our previous work nominated DLGAP2 as a modifier of age-related cognitive decline and risk for AD. We tested the hypothesis that overexpression of DLGAP2 in the hippocampus would protect against cognitive and synaptic deficits in a susceptible F1 5XFAD model. METHODS: DLGAP2 was overexpressed in the hippocampus of F1 hybrid 5XFAD and nontransgenic littermates using a viral approach. Cognitive function, electrophysiological properties, and dendritic spine morphology were assessed at 6 and 14 months of age. RESULTS: DLGAP2 overexpression impaired synaptic plasticity and exacerbated AD-related memory deficits but had minimal effect on spine structure or intrinsic neuronal properties. DISCUSSION: We highlight the complex role of DLGAP2 in AD pathology. Targeted interventions involving postsynaptic proteins must consider potential adverse effects on synaptic integrity and cognitive performance, particularly in the context of AD.

Alzheimer’s disease (AD) has a complex etiology arising from largely unknown interactions between genetic and environmental (GxE) factors. Even in populations with causal familial Alzheimer’s disease mutations, there is variation in disease onset and progression, suggesting that clinical symptoms are modified by genetics and environment. Identification of such modifiers is critical, as mechanisms that promote resilience to high-risk AD mutations, unhealthy diet, or aging represent promising therapeutic targets for AD; global resilience factors that protect against multiple “hits” are among the highest priority for discovery. Both genetic and environmental protective factors in AD have been identified; however, GxE factors are incredibly difficult to study in human populations given complex genomes, poor self-reporting, limited data from underrepresented groups, and incompletely documented exposomes. Here, we (1) validate novel GxE tools using population of mouse strains that model the polygenic nature of human AD, (2) characterize individuals with cognitive resilience to high-risk genetic and dietary perturbations, (3) define and quantitate roles for genetics, sex, age, and diet, and (4) present data for the discovery of complex interactions that are nearly impossible to elucidate from humans or inbred mice. We found that a high-fat high-sugar (HFHS) diet is not universally damaging, as some strains showed an improved AD-related cognitive outcome when fed a HFHS diet, suggesting the need for personalized recommendations for dietary interventions in AD. Cognitive resilience to AD is polygenic; however, we found a locus on Chr 10 that was modestly associated with cognitive resilience to AD in females, and this association was strengthened by HFHS diet, pointing to an unexpected interaction between specific genetic loci and ‘unhealthy’ diet in AD risk and resilience. This study is the first of its kind to explore characteristics of AD resilience and GxE interactions in a genetically diverse mouse model. We present a subset of strains that exemplify global cognitive resilience to be leveraged for deep mechanistic studies aimed toward development of resilience-based, personalized therapeutic interventions. Highlights: AD-BXD mouse models and data repository provide unprecedented tools for understanding individual differences arising from Gene x Environment factors in aging and AD. Evidence of global cognitive resilience to AD-relevant outcomes that is partially mediated by a resilience to aging and dietary risk factors. Cognitive resilience in AD-BXDs its polygenic nature, with a modest locus on Chr 10 whose association with resilience was amplified by a high-fat/high-sugar diet. The genetic factors mediating resilience to AD in humans also predict cognitive resilience in AD-BXDs, suggesting robust, translatable global resilience factors.

Alzheimer’s disease (AD) affects brain regions with remarkable heterogeneity, but the precise impact of this disease on hundreds of small cortical, subcortical, and brainstem regions remains poorly defined. Here, as a prelude to testing preclinical models to prevent AD, we systematically quantified effects of human AD mutations in APP and PSEN1 on 231 regions and comprehensively evaluated changes in volume with unprecedented resolution in genetically diverse mice as a function of sex and genetic background. We studied 34 5XFAD F1 hybrids and 23 sibling controls at 14 months, evaluating learning and memory behaviors, followed by ex vivo diffusion tensor images (DTI) at 25 µm resolution. We delineated 231 regions of interest (ROIs) bilaterally with high precision. Remarkably, we found bidirectional changes: marked volume increases (up to 10%) in neocortex, hippocampus, amygdala, and sensory nuclei, contrasted with decreases in striatum, pallidum, thalamus, hypothalamus, and most fiber tracts. These opposing effects are unrelated to amyloid load and are likely to reflect temporal gradients in susceptibility of ROIs. Effects are similar in both sexes but far more prominent in females. Genetic background strongly modulates penetrance of the human mutations with the AD-BXD77 F1 type having the greatest sensitivity. Light sheet microscopy and stereological analysis of NeuN+ neurons and amyloid β aggregates in 22 regions revealed up to 20% loss of cells in CA3 and in anterior and intralaminar parts of the thalamus. Volumetric changes correlated with impaired fear acquisition and memory, with cases and controls often showing opposite relations between performance and regional volumes. These findings reveal unprecedented regional heterogeneity in AD progression and suggest therapeutic efficacy may vary substantially across genetic backgrounds and between sexes.