Ravi Shanker Pandey

Background: Late-onset Alzheimer’s disease (LOAD) comprises more than 95% of all AD cases. Transgenic, overexpression animal models have off target side effects, do not effectively produce the heterogeneity observed clinically in LOAD patients, and are therefore not best suited for preclinical therapeutic development. The Model Organism Development and Evaluation for Late-onset Alzheimer’s Disease (MODEL-AD) Consortium was established to develop novel mouse strains to model human-relevant genetic and environmental risk factors for LOAD. Methylenetetrahydrofolate reductase (MTHFR) is an enzyme in the folate/methionine pathway. Variants in the MTHFR gene, notably 677C>T, are associated with ADRD, and we have previously shown the Mthfr ^677C>T mouse model phenocopies humans carrying the variant and develop cerebrovascular deficits. Methods: To examine the contributions of Mthfr ^677C>T in the context of late-onset Alzheimer’s disease (LOAD), MODEL-AD created a novel mouse strain on the C57BL/6J (B6) background that was homozygous for Mthfr ^677C>T , in combination with humanized Aβ , APOEe4 , and Trem2*R47H ( referred to as LOAD2.Mthfr ^677C>T ) . Mice were assessed over multiple ages for disease-relevant phenotypes. Regular behavior measurements and biometric samples were collected longitudinally to 24 months of age. Blood and brain tissue were collected for transcriptomics, proteomics, human disease correlation, and neuropathology. Results: Despite lacking hallmark pathologies such as amyloid deposition and significant neuroinflammation, compared to LOAD2 controls, LOAD2.Mthfr ^677C>T mice showed transcriptional and proteomic signatures in the brain that relate to the cerebrovasculature, myelination, and synaptic biology, similar to those seen in human LOAD patients. Conclusions: These data further support the use of the LOAD2.Mthfr ^677C>T mouse model to study aspects of ADRD such as cerebrovascular compromise.

Background: Alzheimer's disease (AD) is a complex, multifactorial pathology characterized by high heterogeneity in biological alterations. New genetic and genomic resources are identifying multiple genetic risk factors for late‐onset Alzheimer's disease (LOAD). However, our understanding of the cellular and molecular mechanisms linking disease risk variants to various phenotypes remains limited. Therefore, it is essential to integrate information from multiple data modalities to thoroughly explore endophenotype networks and biological interactions related to the disease, thereby accelerating our understanding of heterogeneity in Alzheimer's disease. Method: We obtained transcriptomics and proteomics data from whole hemibrain samples of mouse models harboring the genetic risk variants Abca7 ^A1527G , Mthfr ^677C>T , and Plcg2 ^M28L . These mouse model already carrying humanized amyloid‐beta , APOE4, and Trem2 ^R47H alleles, all knocked into a C57BL/6J background. We included mouse models of multiple ages for both sexes. Using state‐of‐the‐art bioinformatics tools, we conducted multi‐omics analysis to identify molecular alterations in these mouse models. Furthermore, we systematically aligned the multimodal mouse data with relevant human study cohorts to determine the AD relevance of risk genes. Result: The effects of these genetic variants recapitulated a variety of human gene and protein expression patterns observed in the LOAD study cohort. The Abca7 variant exhibited extracellular matrix, neuroimmune, and oligodendrocyte‐related gene signatures at an early age, correlating with postmortem LOAD cases compared to controls. By 18 months of age, the Mthfr variant exhibited vasculature, myelination, and synapse‐related gene and protein signatures, also correlating with postmortem LOAD cases relative to controls. The Plcg2 variant exhibited neuroimmune, endolysosome, and synapse‐related gene signatures and altered cell‐ECM interaction processes at the protein level, correlating with postmortem LOAD cases compared to controls. Conclusion: We have characterized in vivo signatures of three genetic candidates for late‐onset Alzheimer's disease (LOAD), identifying alterations in specific LOAD‐related pathways for each variant. Our study highlights that assembling multi‐omics measurements reveals interrelated pathway alterations in Alzheimer's Disease (AD) and enables the identification of biomarker combinations that may inform clinical practice. Our approach provides a platform for further exploration into the causes and progression of AD by assessing animal models at different ages and/or with different combinations of LOAD risk variants.

Background: Alzheimer’s disease (AD) is a complex, multifactorial pathology with high heterogeneity in biological alterations. Our understanding of cellular and molecular mechanisms from disease risk variants to various phenotypes is still limited. Mouse models of AD serve as indispensable platforms for comprehensively characterizing AD pathology, disease progression, and biological mechanisms. However, selection of the right model in preclinical research and translation of findings to clinical populations are intricate processes that require identification of pathophysiological resemblance between model organisms and humans. Many existing clinical trials that showed promising efficacy in one particular mouse model later do not align with human trial results, assuming that study had consisted of a heterogeneous group of participants, and individual animal models may only recapitulate features of a subgroup of human cases. To improve interspecies translation, it is necessary to comprehensively compare molecular signatures in mouse models with subgroup of human AD cases with distinct molecular signatures.Method: We performed transcriptomic and proteomics analysis on whole brain samples from mouse models carrying LOAD risk variants. To assess the human disease relevance of LOAD risk variants in mice, we determined the extent to which changes due to genetic perturbations in mice matched those observed in human AD subtypes and disease stages of AD in the ROS/MAP, Mayo and Mount Sinai Brain Bank cohorts. Genesets within these disease subtypes are highly co‐expressed and represent specific molecular pathways.Result: We have identified that distinct mouse models match to distinct human AD subtypes in age‐dependent manner. Specifically, mouse models carrying human AD risk variants such as Abca7*A1527G showed strong correlation with inflammatory AD subtypes, while mouse models carrying risk variant such as Plcg2*M28L exhibited transcriptomics changes similar to non‐inflammatory AD subtypes.Conclusion: In this study, we highlighted that mouse model of AD may match to a particular subset of human AD subtypes but not all subtypes simultaneously, and that risk for these subtypes may be influenced by distinct AD genetic factors. Additional work toward validating and better understanding the role of each subtype key regulator in its matching mouse model will provide great value and have a great impact on future studies of AD.

Background: APOE^e4 is the greatest risk factor for late‐onset Alzheimer’s disease (AD), accounting for ∼30% of genetic risk and significantly decreasing age of onset. APOE is implicated to play central roles across multiple pathways such as neuroinflammation, lipid metabolism and vascular health. Current mechanistic work has introduced humanized APOE alleles in mouse models of AD on a single laboratory mouse strain, C57BL/6J (B6). Recent work from our lab has taken advantage of genetically diverse wild‐derived strains which were captured from different geographical regions and then inbred, thus, their genomes are closer to recapitulating combinations that would exist in the natural world. Here, we seek to identify and validate human‐relevant genetic interactors of APOE within these contexts and in the presence of amyloid.Method: Novel wild‐derived mice were developed by backcrossing the respective strain (PWK/PhJ, WSB/EiJ) to B6.APOE^e4 (N5). These were crossed with congenic APP/PS1 carriers, and offspring intercrossed to generate cohorts of 4‐ and 8‐month males and females homozygous for APOE^e4, with and without APP/PS1. A cohort of B6.APOE^e4 carrying APP/PS1 were also generated for comparison. Neuropathology was assessed along with blood chemistry, and transcriptional profiling was performed on brain hemispheres.Result: Brain transcriptional profiling revealed that APOE^e4 had the greatest impact on differential gene expression (dE) in the PWK genetic context at both timepoints, with the strongest effect in females. Comparison with AMP‐AD human‐derived consensus clusters also indicated the broadest overlap when variants were present in the PWK context. Furthermore, APOE^e4 genotype was sufficient to cause neuronal damage at 8 months, despite other studies indicating neuronal resilience to amyloid and aging in PWK. In both WSB and B6 carrying APP/PS1, APOE^e4 increased the incidence and severity of cerebral amyloid angiopathy (CAA). Further neuropathological assessments are ongoing, and will be correlated with dE.Conclusion: This work highlights the importance of assessing risk alleles for AD in a variety of genetic contexts. We have also created genetically diverse panels of mice that carry humanized amyloid beta and humanized tau in that hope that these strains will provide an improved preclinical platform for therapeutic discovery and testing.

Several iPSC-derived three-dimensional (3D) cultures have been generated to model Alzheimer’s disease (AD). While some AD-related phenotypes have been identified across these cultures, none of them could recapitulate multiple AD-related hallmarks in one model. To date, the transcriptomic features of these 3D models have not been compared with those of human AD brains. However, these data are crucial to understanding the pertinency of these models for studying AD-related pathomechanisms over time. We developed a 3D bioengineered model of iPSC-derived neural tissue that combines a porous scaffold composed of silk fibroin protein with an intercalated collagen hydrogel to support the growth of neurons and glial cells into complex and functional networks for an extended time, a fundamental requisite for aging studies. Cultures were generated from iPSC lines obtained from two subjects carrying the familial AD (FAD) APP London mutation, two well-studied control lines, and an isogenic control. Cultures were analyzed at 2 and 4.5 months. At both time points, an elevated Aβ42/40 ratio was detected in conditioned media from FAD cultures. However, extracellular Aβ42 deposition and enhanced neuronal excitability were observed in FAD culture only at 4.5 months, suggesting that extracellular Aβ deposition may trigger enhanced network activity. Remarkably, neuronal hyperexcitability has been described in AD patients early in the disease. Transcriptomic analysis revealed the deregulation of multiple gene sets in FAD samples. Such alterations were strikingly similar to those observed in human AD brains. These data provide evidence that our patient-derived FAD model develops time-dependent AD-related phenotypes and establishes a temporal relation among them. Furthermore, FAD iPSC-derived cultures recapitulate transcriptomic features of AD patients. Thus, our bioengineered neural tissue represents a unique tool to model AD in vitro over time.

Transcriptomic analysis of the mammalian retinal pigment epithelium (RPE) aims to identify cellular networks that influence ocular development, maintenance, function, and disease. However, available evidence points to RPE cell heterogeneity within native tissue, which adds complexity to global transcriptomic analysis. Here, to assess cell heterogeneity, we performed single-cell RNA sequencing of RPE cells from two young adult male C57BL/6J mice. Following quality control to ensure robust transcript identification limited to cell singlets, we detected 13,858 transcripts among 2667 and 2846 RPE cells. Dimensional reduction by principal component analysis and uniform manifold approximation and projection revealed six distinct cell populations. All clusters expressed transcripts typical of RPE cells; the smallest (C1, containing 1–2% of total cells) exhibited the hallmarks of stem and/or progenitor (SP) cells. Placing C1–6 along a pseudotime axis suggested a relative decrease in melanogenesis and SP gene expression and a corresponding increase in visual cycle gene expression upon RPE maturation. K-means clustering of all detected transcripts identified additional expression patterns that may advance the understanding of RPE SP cell maintenance and the evolution of cellular metabolic networks during development. This work provides new insights into the transcriptome of the mouse RPE and a baseline for identifying experimentally induced transcriptional changes in future studies of this tissue.

Background: Late‐onset Alzheimer’s disease (LOAD) is the leading cause of dementia in the elderly. However, most transgenic animal models are based on rare, early‐onset AD genes which have not been great preclinical models as they may not sufficiently capture the full transcriptomic signatures and complete neuropathology of disease. Animal models based on LOAD‐associated genes are necessary to connect common genetic variation with LOAD transcriptomes.Method: We performed transcriptomic analysis at 4, 8, 12 and 24 months of age on whole brain samples from APOE4KI mouse, carrying a humanized version of the prominent APOE4genetic risk factor for LOAD, and the Trem2*R47Hmouse, carrying a rare deleterious variant R47H allele of Trem2. In addition, a mouse model expressing both human APOE4and the Trem2*R47Hmutation was used to compare the transcriptional effects in mice carrying both variants to mice carrying only a single risk allele and B6 controls.We compared mouse modules with human postmortem brain modules from the Accelerating Medicine’s Partnership for AD (AMP‐AD) to determine the AD relevance of risk genesResult: Differentially expressed genes in these mice were significantly enriched in multiple AD‐related pathways, including immune response, osteoclast differentiation, metabolism, and mRNA/protein processing. We also observed age related changes, specifically in Trem2*R47H mice, such as more differentially expressed genes and significant differential splicing events compare to the APOE4mouse models.Further, these mouse models overlapped with immune, myelination, neuronal, and DNA repair related AMP‐AD modules.Conclusion: We have characterized three novel mouse models of LOAD at young and advanced ages and observed sex as well as age‐specific transcriptional changes. These mouse models exhibited similar transcriptional changes as seen in clinical samples. Hence, this double mutant mouse model will be used as a sensitized genetic background for further addition of human AD risk variants to model late‐onset Alzheimer’s disease in mice.

Background: Late-onset Alzheimer’s disease (LOAD) is the most common form of dementia worldwide. To date, animal models of Alzheimer’s have focused on rare familial mutations, due to a lack of frank neuropathology from models based on common disease genes. Recent multi-cohort studies of postmortem human brain transcriptomes have identified a set of 30 gene co-expression modules associated with LOAD, providing a molecular catalog of relevant endophenotypes. Results: This resource enables precise gene-based alignment between new animal models and human molecular signatures of disease. Here, we describe a new resource to efficiently screen mouse models for LOAD relevance. A new NanoString nCounter® Mouse AD panel was designed to correlate key human disease processes and pathways with mRNA from mouse brains. Analysis of the 5xFAD mouse, a widely used amyloid pathology model, and three mouse models based on LOAD genetics carrying APOE4 and TREM2*R47H alleles demonstrated overlaps with distinct human AD modules that, in turn, were functionally enriched in key disease-associated pathways. Comprehensive comparison with full transcriptome data from same-sample RNA-Seq showed strong correlation between gene expression changes independent of experimental platform. Conclusions: Taken together, we show that the nCounter Mouse AD panel offers a rapid, cost-effective and highly reproducible approach to assess disease relevance of potential LOAD mouse models.

Background: New genetic and genomic resources have identified multiple genetic risk factors for late-onset Alzheimer’s disease (LOAD) and characterized this common dementia at the molecular level. Experimental studies in model organisms can validate these associations and elucidate the links between specific genetic factors and transcriptomic signatures. Animal models based on LOAD-associated genes can potentially connect common genetic variation with LOAD transcriptomes, thereby providing novel insights into basic biological mechanisms underlying the disease. Methods: We performed RNA-Seq on whole brain samples from a panel of six-month-old female mice, each carrying one of the following mutations: homozygous deletions of Apoe and Clu; hemizygous deletions of Bin1 and Cd2ap; and a transgenic APOEε4. Similar data from a transgenic APP/PS1 model was included for comparison to early-onset variant effects. Weighted gene co-expression network analysis (WGCNA) was used to identify modules of correlated genes and each module was tested for differential expression by strain. We then compared mouse modules with human postmortem brain modules from the Accelerating Medicine’s Partnership for AD (AMP-AD) to determine the LOAD-related processes affected by each genetic risk factor. Results: Mouse modules were significantly enriched in multiple AD-related processes, including immune response, inflammation, lipid processing, endocytosis, and synaptic cell function. WGCNA modules were significantly associated with Apoe^−/−, APOEε4, Clu^−/−, and APP/PS1 mouse models. Apoe^−/−, GFAP-driven APOEε4, and APP/PS1 driven modules overlapped with AMP-AD inflammation and microglial modules; Clu^−/− driven modules overlapped with synaptic modules; and APP/PS1 modules separately overlapped with lipid-processing and metabolism modules. Conclusions: This study of genetic mouse models provides a basis to dissect the role of AD risk genes in relevant AD pathologies. We determined that different genetic perturbations affect different molecular mechanisms comprising AD, and mapped specific effects to each risk gene. Our approach provides a platform for further exploration into the causes and progression of AD by assessing animal models at different ages and/or with different combinations of LOAD risk variants.

Background: New genetic and genomic resources have identified multiple genetic risk factors for late-onset Alzheimer’s disease (LOAD) and characterized this common dementia at the molecular level. Experimental studies in model organisms can validate these associations and elucidate the links between specific genetic factors and transcriptomic signatures. Animal models based on LOAD-associated genes can potentially connect common genetic variation with LOAD transcriptomes, thereby providing novel insights into basic biological mechanisms underlying the disease. Methods: We performed RNA-Seq on whole brain samples from a panel of six-month-old female mice, each carrying one of the following mutations: homozygous deletions of Apoe and Clu ; hemizygous deletions of Bin1 and Cd2ap ; and a transgenic APOEε4 . Similar data from a transgenic APP/PS1 model was included for comparison to early-onset variant effects. Weighted gene co-expression network analysis (WGCNA) was used to identify modules of correlated genes and each module was tested for differential expression by strain. We then compared mouse modules with human postmortem brain modules from the Accelerating Medicine’s Partnership for AD (AMP-AD) to determine the LOAD-related processes affected by each genetic risk factor. Results: Mouse modules were significantly enriched in multiple AD-related processes, including immune response, inflammation, lipid processing, endocytosis, and synaptic cell function. WGCNA modules were significantly associated with Apoe ^−/− , APOEε4, Clu ^−/− , and APP/PS1 mouse models. Apoe ^−/− , GFAP-driven APOEε4, and APP/PS1 driven modules overlapped with AMP-AD inflammation and microglial modules; Clu ^−/− driven modules overlapped with synaptic modules; and APP/PS1 modules separately overlapped with lipid-processing and metabolism modules. Conclusions: This study of genetic mouse models provides a basis to dissect the role of AD risk genes in relevant AD pathologies. We determined that different genetic perturbations affect different molecular mechanisms comprising AD, and mapped specific effects to each risk gene. Our approach provides a platform for further exploration into the causes and progression of AD by assessing animal models at different ages and/or with different combinations of LOAD risk variants.