Digital quantification of the MMSE interlocking pentagon areas: a three-stage algorithm.

The Mini-Mental State Examination (MMSE) is a widely employed screening tool for the severity of cognitive impairment. Among the MMSE items, the pentagon copying test (PCT) requires participants to accurately replicate a sample of two interlocking pentagons. While the PCT is traditionally scored on a binary scale, there have been limited developments of granular scoring scale to assess task performance. In this paper, we present a novel three-stage algorithm, called Quantification of Interlocking Pentagons (QIP) which quantifies PCT performance by computing the areas of individual pentagons and their intersection areas, and a balance ratio between the areas of the two individual pentagons. The three stages of the QIP algorithm include: (1) detection of line segments, (2) unraveling of the interlocking pentagons, and (3) quantification of areas. A set of 497 PCTs from 84 participants including their baseline and follow-up PCTs from the Rush Memory and Aging Project was selected blinded about their cognitive and clinical status. Analysis of the quantified data revealed a significant inverse relationship between age and balance ratio (beta = - 0.49, p = 0.0033), indicating that older age was associated with a smaller balance ratio. In addition, balance ratio was associated with perceptual speed (r = 0.71, p = 0.0135), vascular risk factors (beta = - 3.96, p = 0.0269), and medical conditions (beta = - 2.78, p = 0.0389). The QIP algorithm can serve as a useful tool for enhancing the scoring of performance in the PCT.

Polygenic risk associated with Alzheimer's disease and other traits influences genes involved in T cell signaling and activation.

Introduction: T cells, known for their ability to respond to an enormous variety of pathogens and other insults, are increasingly recognized as important mediators of pathology in neurodegeneration and other diseases. T cell gene expression phenotypes can be regulated by disease-associated genetic variants. Many complex diseases are better represented by polygenic risk than by individual variants. Methods: We first compute a polygenic risk score (PRS) for Alzheimer's disease (AD) using genomic sequencing data from a cohort of Alzheimer's disease (AD) patients and age-matched controls, and validate the AD PRS against clinical metrics in our cohort. We then calculate the PRS for several autoimmune disease, neurological disorder, and immune function traits, and correlate these PRSs with T cell gene expression data from our cohort. We compare PRS-associated genes across traits and four T cell subtypes. Results: Several genes and biological pathways associated with the PRS for these traits relate to key T cell functions. The PRS-associated gene signature generally correlates positively for traits within a particular category (autoimmune disease, neurological disease, immune function) with the exception of stroke. The trait-associated gene expression signature for autoimmune disease traits was polarized towards CD4+ T cell subtypes. Discussion: Our findings show that polygenic risk for complex disease and immune function traits can have varying effects on T cell gene expression trends. Several PRS-associated genes are potential candidates for therapeutic modulation in T cells, and could be tested in in vitro applications using cells from patients bearing high or low polygenic risk for AD or other conditions.

SR-TWAS: Leveraging Multiple Reference Panels to Improve TWAS Power by Ensemble Machine Learning.

Multiple reference panels of a given tissue or multiple tissues often exist, and multiple regression methods could be used for training gene expression imputation models for TWAS. To leverage expression imputation models (i.e., base models) trained with multiple reference panels, regression methods, and tissues, we develop a Stacked Regression based TWAS (SR-TWAS) tool which can obtain optimal linear combinations of base models for a given validation transcriptomic dataset. Both simulation and real studies showed that SR-TWAS improved power, due to increased effective training sample sizes and borrowed strength across multiple regression methods and tissues. Leveraging base models across multiple reference panels, tissues, and regression methods, our real application studies identified 6 independent significant risk genes for Alzheimer's disease (AD) dementia for supplementary motor area tissue and 9 independent significant risk genes for Parkinson's disease (PD) for substantia nigra tissue. Relevant biological interpretations were found for these significant risk genes.

Associations of cortical SPP1 and ITGAX with cognition and common neuropathologies in older adults.

INTRODUCTION: The secreted phosphoprotein 1 (SPP1) gene expressed by CD11c+ cells is known to be associated with microglia activation and neuroinflammatory diseases. As most studies rely on mouse models, we investigated these genes and proteins in the cortical brain tissue of older adults and their role in Alzheimer's disease (AD) and related disorders. METHODS: We leveraged protein measurements, single-nuclei, and RNASeq data from the Religious Orders Study and Rush Memory and Aging Project (ROSMAP) of over 1200 samples for association analysis. RESULTS: Expression of SPP1 and its encoded protein osteopontin were associated with faster cognitive decline and greater odds of common neuropathologies. At single-cell resolution,  integrin subunit alpha X (ITGAX) was highly expressed in microglia, where specific subpopulations were associated with AD and cerebral amyloid angiopathy. DISCUSSION: The study provides evidence of SPP1 and ITGAX association with cognitive decline and common neuropathologies identifying a microglial subset associated with disease.

The molecular genetic landscape of human brain size variation.

Human brain size changes dynamically through early development, peaks in adolescence, and varies up to 2-fold among adults. However, the molecular genetic underpinnings of interindividual variation in brain size remain unknown. Here, we leveraged postmortem brain RNA sequencing and measurements of brain weight (BW) in 2,531 individuals across three independent datasets to identify 928 genome-wide significant associations with BW. Genes associated with higher or lower BW showed distinct neurodevelopmental trajectories and spatial patterns that mapped onto functional and cellular axes of brain organization. Expression of BW genes was predictive of interspecies differences in brain size, and bioinformatic annotation revealed enrichment for neurogenesis and cell-cell communication. Genome-wide, transcriptome-wide, and phenome-wide association analyses linked BW gene sets to neuroimaging measurements of brain size and brain-related clinical traits. Cumulatively, these results represent a major step toward delineating the molecular pathways underlying human brain size variation in health and disease.

Benchmarking of deep neural networks for predicting personal gene expression from DNA sequence highlights shortcomings.

Deep learning methods have recently become the state of the art in a variety of regulatory genomic tasks1-6, including the prediction of gene expression from genomic DNA. As such, these methods promise to serve as important tools in interpreting the full spectrum of genetic variation observed in personal genomes. Previous evaluation strategies have assessed their predictions of gene expression across genomic regions; however, systematic benchmarking is lacking to assess their predictions across individuals, which would directly evaluate their utility as personal DNA interpreters. We used paired whole genome sequencing and gene expression from 839 individuals in the ROSMAP study7 to evaluate the ability of current methods to predict gene expression variation across individuals at varied loci. Our approach identifies a limitation of current methods to correctly predict the direction of variant effects. We show that this limitation stems from insufficiently learned sequence motif grammar and suggest new model training strategies to improve performance.

Differential gene expression analysis based on linear mixed model corrects false positive inflation for studying quantitative traits.

Differential gene expression (DGE) analysis has been widely employed to identify genes expressed differentially with respect to a trait of interest using RNA sequencing (RNA-Seq) data. Recent RNA-Seq data with large samples pose challenges to existing DGE methods, which were mainly developed for dichotomous traits and small sample sizes. Especially, existing DGE methods are likely to result in inflated false positive rates. To address this gap, we employed a linear mixed model (LMM) that has been widely used in genetic association studies for DGE analysis of quantitative traits. We first applied the LMM method to the discovery RNA-Seq data of dorsolateral prefrontal cortex (DLPFC) tissue (n = 632) with four continuous measures of Alzheimer's Disease (AD) cognitive and neuropathologic traits. The quantile-quantile plots of p-values showed that false positive rates were well calibrated by LMM, whereas other methods not accounting for sample-specific mixed effects led to serious inflation. LMM identified 37 potentially significant genes with differential expression in DLPFC for at least one of the AD traits, 17 of which were replicated in the additional RNA-Seq data of DLPFC, supplemental motor area, spinal cord, and muscle tissues. This application study showed not only well calibrated DGE results by LMM, but also possibly shared gene regulatory mechanisms of AD traits across different relevant tissues.

Robust, scalable, and informative clustering for diverse biological networks.

Clustering molecular data into informative groups is a primary step in extracting robust conclusions from big data. However, due to foundational issues in how they are defined and detected, such clusters are not always reliable, leading to unstable conclusions. We compare popular clustering algorithms across thousands of synthetic and real biological datasets, including a new consensus clustering algorithm-SpeakEasy2: Champagne. These tests identify trends in performance, show no single method is universally optimal, and allow us to examine factors behind variation in performance. Multiple metrics indicate SpeakEasy2 generally provides robust, scalable, and informative clusters for a range of applications.

Explainable deep learning approach for extracting cognitive features from hand-drawn images of intersecting pentagons.

Hand drawing, which requires multiple neural systems for planning and controlling sequential movements, is a useful cognitive test for older adults. However, the conventional visual assessment of these drawings only captures limited attributes and overlooks subtle details that could help track cognitive states. Here, we utilized a deep-learning model, PentaMind, to examine cognition-related features from hand-drawn images of intersecting pentagons. PentaMind, trained on 13,777 images from 3111 participants in three aging cohorts, explained 23.3% of the variance in the global cognitive scores, 1.92 times more than the conventional rating. This accuracy improvement was due to capturing additional drawing features associated with motor impairments and cerebrovascular pathologies. By systematically modifying the input images, we discovered several important drawing attributes for cognition, including line waviness. Our results demonstrate that deep learning models can extract novel drawing metrics to improve the assessment and monitoring of cognitive decline and dementia in older adults.

A Molecular Basis of Human Brain Connectivity.

Neuroimaging is commonly used to infer human brain connectivity, but those measurements are far-removed from the molecular underpinnings at synapses. To uncover the molecular basis of human brain connectivity, we analyzed a unique cohort of 98 individuals who provided neuroimaging and genetic data contemporaneous with dendritic spine morphometric, proteomic, and gene expression data from the superior frontal and inferior temporal gyri. Through cellular contextualization of the molecular data with dendritic spine morphology, we identified hundreds of proteins related to synapses, energy metabolism, and RNA processing that explain between-individual differences in functional connectivity and structural covariation. By integrating data at the genetic, molecular, subcellular, and tissue levels, we bridged the divergent fields of molecular biology and neuroimaging to identify a molecular basis of brain connectivity. One-Sentence Summary: Dendritic spine morphometry and synaptic proteins unite the divergent fields of molecular biology and neuroimaging.

The MIND diet, brain transcriptomic alterations, and dementia.

Identifying novel mechanisms underlying dementia is critical to improving prevention and treatment. As an approach to mechanistic discovery, we investigated whether MIND diet (Mediterranean-DASH Diet Intervention for Neurodegenerative Delay), a consistent risk factor for dementia, is correlated with a specific profile of cortical gene expression, and whether such a transcriptomic profile is associated with dementia, in the Religious Orders Study (ROS) and Rush Memory and Aging Project (MAP). RNA sequencing (RNA-Seq) was conducted in postmortem dorsolateral prefrontal cortex tissue from 1,204 deceased participants; neuropsychological assessments were performed annually prior to death. In a subset of 482 participants, diet was assessed ~6 years before death using a validated food-frequency questionnaire; in these participants, using elastic net regression, we identified a transcriptomic profile, consisting of 50 genes, significantly correlated with MIND diet score (P=0.001). In multivariable analysis of the remaining 722 individuals, higher transcriptomic score of MIND diet was associated with slower annual rate of decline in global cognition (ß=0.011 per standard deviation increment in transcriptomic profile score, P=0.003) and lower odds of dementia (odds ratio [OR] =0.76, P=0.0002). Cortical expression of several genes appeared to mediate the association between MIND diet and dementia, including TCIM, whose expression in inhibitory neurons and oligodendrocytes was associated with dementia in a subset of 424 individuals with single-nuclei RNA-seq data. In a secondary Mendelian randomization analysis, genetically predicted transcriptomic profile score was associated with dementia (OR=0.93, P=0.04). Our study suggests that associations between diet and cognitive health may involve brain molecular alterations at the transcriptomic level. Investigating brain molecular alterations related to diet may inform the identification of novel pathways underlying dementia.

Unconventional Source of Neurotoxic Protein Aggregation from Organelle Off-Target Bax∆2 in Alzheimer's Disease.

Protein aggregates are a hallmark of Alzheimer's disease (AD). Extensive studies have focused on ß-amyloid plaques and Tau tangles. Here, we illustrate a novel source of protein aggregates in AD neurons from organelle off-target proteins. Bax is a mitochondrial pore-forming pro-death protein. What happens to Bax if it fails to target mitochondria? We previously showed that a mitochondrial target-deficient alternatively spliced variant, Bax∆2, formed large cytosolic protein aggregates and triggered caspase 8-mediated cell death. Bax∆2 protein levels were low in most normal organs and the proteins were quickly degraded in cancer. Here, we found that 85% of AD patients had Bax∆2 required alternative splicing. Increased Bax∆2 proteins were mostly accumulated in neurons of AD-susceptible brain regions. Intracellularly, Bax∆2 aggregates distributed independently of Tau tangles. Interestingly, Bax∆2 aggregates triggered the formation of stress granules (SGs), a large protein-RNA complex involved in AD pathogenesis. Although the functional domains required for aggregation and cell death are the same as in cancer cells, Bax∆2 relied on SGs, not caspase 8, for neuronal cell death. These results imply that the aggregation of organelle off-target proteins, such as Bax∆2, broadens the scope of traditional AD pathogenic proteins that contribute to the neuronal stress responses and AD pathogenesis.

Interpretable deep learning approach for extracting cognitive features from hand-drawn images of intersecting pentagons in older adults.

Hand drawing involves multiple neural systems for planning and precise control of sequential movements, making it a valuable cognitive test for older adults. However, conventional visual assessment of drawings may not capture intricate nuances that could help track cognitive states. To address this issue, we utilized a deep-learning model, PentaMind, to examine cognition-related features from hand-drawn images of intersecting pentagons. PentaMind, trained on 13,777 images from 3,111 participants in three aging cohorts, explained 23.3% of the variance in global cognitive scores, a comprehensive hour-long cognitive battery. The model’s performance, which was 1.92 times more accurate than conventional visual assessment, significantly improved the detection of cognitive decline. The improvement in accuracy was due to capturing additional drawing features that we found to be associated with motor impairments and cerebrovascular pathologies. By systematically modifying the input images, we discovered several important drawing attributes for cognition, including line waviness. Our results demonstrate that hand-drawn images can provide rich cognitive information, enabling rapid assessment of cognitive decline and suggesting potential clinical implications in dementia.

Associations of VGF with Neuropathologies and Cognitive Health in Older Adults.

OBJECTIVE: VGF is proposed as a potential therapeutic target for Alzheimer's (AD) and other neurodegenerative conditions. The cell-type specific and, separately, peptide specific associations of VGF with pathologic and cognitive outcomes remain largely unknown. We leveraged gene expression and protein data from the human neocortex and investigated the VGF associations with common neuropathologies and late-life cognitive decline. METHODS: Community-dwelling older adults were followed every year, died, and underwent brain autopsy. Cognitive decline was captured via annual cognitive testing. Common neurodegenerative and cerebrovascular conditions were assessed during neuropathologic evaluations. Bulk brain RNASeq and targeted proteomics analyses were conducted using frozen tissues from dorsolateral prefrontal cortex of 1,020 individuals. Cell-type specific gene expressions were quantified in a subsample (N = 424) following single nuclei RNASeq analysis from the same cortex. RESULTS: The bulk brain VGF gene expression was primarily associated with AD and Lewy bodies. The VGF gene association with cognitive decline was in part accounted for by neuropathologies. Similar associations were observed for the VGF protein. Cell-type specific analyses revealed that, while VGF was differentially expressed in most major cell types in the cortex, its association with neuropathologies and cognitive decline was restricted to the neuronal cells. Further, the peptide fragments across the VGF polypeptide resembled each other in relation to neuropathologies and cognitive decline. INTERPRETATION: Multiple pathways link VGF to cognitive health in older age, including neurodegeneration. The VGF gene functions primarily in neuronal cells and its protein associations with pathologic and cognitive outcomes do not map to a specific peptide. ANN NEUROL 2023;94:232-244.

Benchmarking of deep neural networks for predicting personal gene expression from DNA sequence highlights shortcomings.

Deep learning methods have recently become the state-of-the-art in a variety of regulatory genomic tasks1-6 including the prediction of gene expression from genomic DNA. As such, these methods promise to serve as important tools in interpreting the full spectrum of genetic variation observed in personal genomes. Previous evaluation strategies have assessed their predictions of gene expression across genomic regions, however, systematic benchmarking is lacking to assess their predictions across individuals, which would directly evaluates their utility as personal DNA interpreters. We used paired Whole Genome Sequencing and gene expression from 839 individuals in the ROSMAP study7 to evaluate the ability of current methods to predict gene expression variation across individuals at varied loci. Our approach identifies a limitation of current methods to correctly predict the direction of variant effects. We show that this limitation stems from insufficiently learnt sequence motif grammar, and suggest new model training strategies to improve performance.

Estrogen Receptor Genes, Cognitive Decline, and Alzheimer Disease.

BACKGROUND AND OBJECTIVES: Lifetime risk of Alzheimer disease (AD) dementia is twofold higher in women compared with men, and low estrogen levels in postmenopause have been suggested as a possible contributor. We examined 3 ER (GPER1, ER2, and ER1) variants in association with AD traits as an indirect method to test the association between estrogen and AD in women. Although the study focus was on women, in a comparison, we separately examined ER molecular variants in men. METHODS: Participants were followed for an average of 10 years in one of the 2 longitudinal clinical pathologic studies of aging. Global cognition was assessed using a composite score derived from 19 neuropsychological tests' scores. Postmortem pathologic assessment included examination of 3 AD (amyloid-ß and tau tangles determined by immunohistochemistry, and a global AD pathology score derived from diffuse and neurotic plaques and neurofibrillary tangle count) and 8 non-AD pathology indices. ER molecular genomic variants included genotyping and examining ER DNA methylation and RNA expression in brain regions including the dorsolateral prefrontal cortex (DLPFC) that are major players in cognition and often have AD pathology. RESULTS: The mean age of women (N = 1711) at baseline was 78.0 (SD = 7.7) years. In women, GPER1 molecular variants had the most consistent associations with AD traits. GPER1 DNA methylation was associated with cognitive decline, tau tangle density, and global AD pathology score. GPER1 RNA expression in DLPFC was related to cognitive decline and tau tangle density. Other associations included associations of ER2 and ER1 sequence variants and DNA methylation with cognition. RNA expressions in DLPFC of genes involved in signaling mechanisms of activated ERs were also associated with cognitive decline and tau tangle density in women. In men (N = 651, average age at baseline: 77.4 [SD = 7.3]), there were less robust associations between ER molecular genomic variants and AD cognitive and pathologic traits. No consistent association was seen between ER molecular genomic variations and non-AD pathologies in either of the sexes. DISCUSSION: ER DNA methylation and RNA expression, and to some extent ER polymorphisms, were associated with AD cognitive and pathologic traits in women, and to a lesser extent in men.

Proteome-Wide Discovery of Cortical Proteins That May Provide Motor Resilience to Offset the Negative Effects of Pathologies in Older Adults.

BACKGROUND: Motor resilience proteins have not been identified. This proteome-wide discovery study sought to identify proteins that may provide motor resilience. METHODS: We studied the brains of older decedents with annual motor testing, postmortem brain pathologies, and proteome-wide data. Parkinsonism was assessed using 26 items of a modified United Parkinson Disease Rating Scale. We used linear mixed-effect models to isolate motor resilience, defined as the person-specific estimate of progressive parkinsonism after controlling for age, sex, and 10 brain pathologies. A total of 8 356 high-abundance proteins were quantified from dorsal lateral prefrontal cortex using tandem mass tag and liquid chromatography-mass spectrometry. RESULTS: There were 391 older adults (70% female), mean age 80 years at baseline and 89 years at death. Five proteins were associated with motor resilience: A higher level of AP1B1 (Estimate -0.504, SE 0.121, p = 3.12 × 10-5) and GNG3 (Estimate -0.276, SE 0.068, p = 4.82 × 10-5) was associated with slower progressive parkinsonism. By contrast, a higher level of TTC38 (Estimate 0.140, SE 0.029, p = 1.87 × 10-6), CARKD (Estimate 0.413, SE 0.100, p = 3.50 × 10-5), and ABHD14B (Estimate 0.175, SE 0.044, p = 6.48 × 10-5) was associated with faster progressive parkinsonism. Together, these 5 proteins accounted for almost 25% of the variance of progressive parkinsonism above the 17% accounted for by 10 indices of brain pathologies. DISCUSSION: Cortical proteins may provide more or less motor resilience in older adults. These proteins are high-value therapeutic targets for drug discovery that may lead to interventions that maintain motor function despite the accumulation of as yet untreatable brain pathologies.

Inferring protein expression changes from mRNA in Alzheimer's dementia using deep neural networks.

Identifying the molecular systems and proteins that modify the progression of Alzheimer's disease and related dementias (ADRD) is central to drug target selection. However, discordance between mRNA and protein abundance, and the scarcity of proteomic data, has limited our ability to advance candidate targets that are mainly based on gene expression. Therefore, by using a deep neural network that predicts protein abundance from mRNA expression, here we attempt to track the early protein drivers of ADRD. Specifically, by applying the clei2block deep learning model to 1192 brain RNA-seq samples, we identify protein modules and disease-associated expression changes that were not directly observed at the mRNA level. Moreover, pseudo-temporal trajectory inference based on the predicted proteome became more closely correlated with cognitive decline and hippocampal atrophy compared to RNA-based trajectories. This suggests that the predicted changes in protein expression could provide a better molecular representation of ADRD progression. Furthermore, overlaying clinical traits on protein pseudotime trajectory identifies protein modules altered before cognitive impairment. These results demonstrate how our method can be used to identify potential early protein drivers and possible drug targets for treating and/or preventing ADRD.

Cascading epigenomic analysis for identifying disease genes from the regulatory landscape of GWAS variants.

The majority of genetic variants detected in genome wide association studies (GWAS) exert their effects on phenotypes through gene regulation. Motivated by this observation, we propose a multi-omic integration method that models the cascading effects of genetic variants from epigenome to transcriptome and eventually to the phenome in identifying target genes influenced by risk alleles. This cascading epigenomic analysis for GWAS, which we refer to as CEWAS, comprises two types of models: one for linking cis genetic effects to epigenomic variation and another for linking cis epigenomic variation to gene expression. Applying these models in cascade to GWAS summary statistics generates gene level statistics that reflect genetically-driven epigenomic effects. We show on sixteen brain-related GWAS that CEWAS provides higher gene detection rate than related methods, and finds disease relevant genes and gene sets that point toward less explored biological processes. CEWAS thus presents a novel means for exploring the regulatory landscape of GWAS variants in uncovering disease mechanisms.

Stem cell-derived neurons reflect features of protein networks, neuropathology, and cognitive outcome of their aged human donors.

We have generated a controlled and manipulable resource that captures genetic risk for Alzheimer's disease: iPSC lines from 53 individuals coupled with RNA and proteomic profiling of both iPSC-derived neurons and brain tissue of the same individuals. Data collected for each person include genome sequencing, longitudinal cognitive scores, and quantitative neuropathology. The utility of this resource is exemplified here by analyses of neurons derived from these lines, revealing significant associations between specific Aß and tau species and the levels of plaque and tangle deposition in the brain and, more importantly, with the trajectory of cognitive decline. Proteins and networks are identified that are associated with AD phenotypes in iPSC neurons, and relevant associations are validated in brain. The data presented establish this iPSC collection as a resource for investigating person-specific processes in the brain that can aid in identifying and validating molecular pathways underlying AD.