Andrew McKenzie

A common haplotype lowers PU.1 expression in myeloid cells and delays onset of Alzheimer's disease

A genome-wide survival analysis of 14,406 Alzheimers disease (AD) cases and 25,849 controls identified eight previously reported AD risk loci and 14 novel loci associated with age at onset. Linkage disequilibrium score regression of 220 cell types implicated the regulation of myeloid gene expression in AD risk. The minor allele of rs1057233 (G), within the previously reported CELF1 AD risk locus, showed association with delayed AD onset and lower expression of SPI1 in monocytes and macrophages. SPI1 encodes PU.1, a transcription factor critical for myeloid cell development and function. AD heritability was enriched within the PU.1 cistrome, implicating a myeloid PU.1 target gene network in AD. Finally, experimentally altered PU.1 levels affected the expression of mouse orthologs of many AD risk genes and the phagocytic activity of mouse microglial cells. Our results suggest that lower SPI1 expression reduces AD risk by regulating myeloid gene expression and cell function.

Leveraging Social Networks for Toxicovigilance

The landscape of drug abuse is shifting. Traditional means of characterizing these changes, such as national surveys or voluntary reporting by frontline clinicians, can miss changes in usage the emergence of novel drugs. Delays in detecting novel drug usage patterns make it difficult to evaluate public policy aimed at altering drug abuse. Increasingly, newer methods to inform frontline providers to recognize symptoms associated with novel drugs or methods of administration are needed. The growth of social networks may address this need. The objective of this manuscript is to introduce tools for using data from social networks to characterize drug abuse. We outline a structured approach to analyze social media in order to capture emerging trends in drug abuse by applying powerful methods from artificial intelligence, computational linguistics, graph theory, and agent-based modeling. First, we describe how to obtain data from social networks such as Twitter using publicly available automated programmatic interfaces. Then, we discuss how to use artificial intelligence techniques to extract content useful for purposes of toxicovigilance. This filtered content can be employed to generate real-time maps of drug usage across geographical regions. Beyond describing the real-time epidemiology of drug abuse, techniques from computational linguistics can uncover ways that drug discussions differ from other online conversations. Next, graph theory can elucidate the structure of networks discussing drug abuse, helping us learn what online interactions promote drug abuse and whether these interactions differ among drugs. Finally, agent-based modeling relates online interactions to psychological archetypes, providing a link between epidemiology and behavior. An analysis of social media discussions about drug abuse patterns with computational linguistics, graph theory, and agent-based modeling permits the real-time monitoring and characterization of trends of drugs of abuse. These tools provide a powerful complement to existing methods of toxicovigilance.

Integrative network analysis of nineteen brain regions identifies molecular signatures and networks underlying selective regional vulnerability to Alzheimer's disease.

BackgroundAlzheimer’s disease (AD) is the most common form of dementia, characterized by progressive cognitive impairment and neurodegeneration. However, despite extensive clinical and genomic studies, the molecular basis of AD development and progression remains elusive.MethodsTo elucidate molecular systems associated with AD, we developed a large scale gene expression dataset from 1053 postmortem brain samples across 19 cortical regions of 125 individuals with a severity spectrum of dementia and neuropathology of AD. We excluded brain specimens that evidenced neuropathology other than that characteristic of AD. For the first time, we performed a pan-cortical brain region genomic analysis, characterizing the gene expression changes associated with a measure of dementia severity and multiple measures of the severity of neuropathological lesions associated with AD (neuritic plaques and neurofibrillary tangles) and constructing region-specific co-expression networks. We rank-ordered 44,692 gene probesets, 1558 co-expressed gene modules and 19 brain regions based upon their association with the disease traits.ResultsThe neurobiological pathways identified through these analyses included actin cytoskeleton, axon guidance, and nervous system development. Using public human brain single-cell RNA-sequencing data, we computed brain cell type-specific marker genes for human and determined that many of the abnormally expressed gene signatures and network modules were specific to oligodendrocytes, astrocytes, and neurons. Analysis based on disease severity suggested that: many of the gene expression changes, including those of oligodendrocytes, occurred early in the progression of disease, making them potential translational/treatment development targets and unlikely to be mere bystander result of degeneration; several modules were closely linked to cognitive compromise with lesser association with traditional measures of neuropathology. The brain regional analyses identified temporal lobe gyri as sites associated with the greatest and earliest gene expression abnormalities.ConclusionsThis transcriptomic network analysis of 19 brain regions provides a comprehensive assessment of the critical molecular pathways associated with AD pathology and offers new insights into molecular mechanisms underlying selective regional vulnerability to AD at different stages of the progression of cognitive compromise and development of the canonical neuropathological lesions of AD.

DGCA: A comprehensive R package for Differential Gene Correlation Analysis

BackgroundDissecting the regulatory relationships between genes is a critical step towards building accurate predictive models of biological systems. A powerful approach towards this end is to systematically study the differences in correlation between gene pairs in more than one distinct condition.ResultsIn this study we develop an R package, DGCA (for Differential Gene Correlation Analysis), which offers a suite of tools for computing and analyzing differential correlations between gene pairs across multiple conditions. To minimize parametric assumptions, DGCA computes empirical p-values via permutation testing. To understand differential correlations at a systems level, DGCA performs higher-order analyses such as measuring the average difference in correlation and multiscale clustering analysis of differential correlation networks. Through a simulation study, we show that the straightforward z-score based method that DGCA employs significantly outperforms the existing alternative methods for calculating differential correlation. Application of DGCA to the TCGA RNA-seq data in breast cancer not only identifies key changes in the regulatory relationships between TP53 and PTEN and their target genes in the presence of inactivating mutations, but also reveals an immune-related differential correlation module that is specific to triple negative breast cancer (TNBC).ConclusionsDGCA is an R package for systematically assessing the difference in gene-gene regulatory relationships under different conditions. This user-friendly, effective, and comprehensive software tool will greatly facilitate the application of differential correlation analysis in many biological studies and thus will help identification of novel signaling pathways, biomarkers, and targets in complex biological systems and diseases.

Multiscale network modeling of oligodendrocytes reveals molecular components of myelin dysregulation in Alzheimer's disease

BackgroundOligodendrocytes (OLs) and myelin are critical for normal brain function and have been implicated in neurodegeneration. Several lines of evidence including neuroimaging and neuropathological data suggest that Alzheimer’s disease (AD) may be associated with dysmyelination and a breakdown of OL-axon communication.MethodsIn order to understand this phenomenon on a molecular level, we systematically interrogated OL-enriched gene networks constructed from large-scale genomic, transcriptomic and proteomic data obtained from human AD postmortem brain samples. We then validated these networks using gene expression datasets generated from mice with ablation of major gene expression nodes identified in our AD-dysregulated networks.ResultsThe robust OL gene coexpression networks that we identified were highly enriched for genes associated with AD risk variants, such as BIN1 and demonstrated strong dysregulation in AD. We further corroborated the structure of the corresponding gene causal networks using datasets generated from the brain of mice with ablation of key network drivers, such as UGT8, CNP and PLP1, which were identified from human AD brain data. Further, we found that mice with genetic ablations of Cnp mimicked aspects of myelin and mitochondrial gene expression dysregulation seen in brain samples from patients with AD, including decreased protein expression of BIN1 and GOT2.ConclusionsThis study provides a molecular blueprint of the dysregulation of gene expression networks of OL in AD and identifies key OL- and myelination-related genes and networks that are highly associated with AD.

A COMMON ALLELE IN SPI1 LOWERS RISK AND DELAYS AGE AT ONSET FOR ALZHEIMER'S DISEASE

Kuan-Lin Huang, Sheng Chih Jin, Oscar Harari, Manav Kapoor, Sarah Bertelsen, Jake Czajkowski, jean-Charles Lambert, Vincent Chouraki, C eline Bellenguez, Benjamin Grenier-Boley, Yuetiva Deming, Andrew McKenzie, Alan E. Renton, John Budde, Jorge L. Del-Aguila, Maria Victoria Fernandez, Laura Ibanez, Denise Harold, Paul Hollingworth, Richard Mayeux, Jonathan L. Haines, Lindsay A. Farrer, Margaret A. Pericak-Vance, Sudha Seshadri, Julie Williams, Philippe Amouyel, Gerard D. Schellenberg, Bin Zhang, Ingrid Borecki, John Kauwe, Eduardo Marcora, Carlos Cruchaga, Alison M. Goate, The Alzheimer’s Disease Neuroimaging Initiative, Washington University in St. Louis, Saint Louis, MO, USA; 2 Yale University, New Haven, CT, USA; 3 Icahn School of Medicine at Mount Sinai, New York, NY, USA; Institut Pasteur de Lille, Lille, France; Boston University School of Medicine, Boston, MA, USA; 6 Washington University School of Medicine, Saint Louis, MO, USA; 7 Cardiff University, Cardiff, United Kingdom; 8 Columbia University, New York, NY, USA; Case Western Reserve University, Cleveland, OH, USA; Boston University, Boston, MA, USA; University of Miami, Miller School of Medicine, Miami, FL, USA; 12 MRC Centre for Neuropsychiatric Genetics & Genomics, Institute of Psychological Medicine and Clinical Neurosciences, Cardiff University, Cardiff, United Kingdom; University of Pennsylvania Perelman School of Medicine, Philadelphia, PA, USA; 14 Brigham Young University, Provo, UT, USA. Contact e-mail: kuan-lin. huang@wustl.edu

The human-specific CASP4 gene product contributes to Alzheimer-related synaptic and behavioural deficits

Recent studies have indicated that innate immune signalling molecules are involved in late-onset Alzheimers disease (LOAD) risk. Amyloid beta (Aβ) accumulates in AD brain, and has been proposed to act as a trigger of innate immune responses. Caspase-4 is an important part of the innate immune response. We recently characterized transgenic mice carrying human CASP4, and observed that the mice manifested profound innate immune responses to lipopolysaccharide (LPS). Since these inflammatory processes are important in the aetiology of AD, we have now analysed the correlation of expression of caspase-4 in human brain with AD risk genes, and studied caspase-4 effects on AD-related phenotypes in APPswe/PS1deltaE9 (APP/PS1) mice. We observed that the expression of caspase-4 was strongly correlated with AD risk genes including TYROBP, TREM2, CR1, PSEN1, MS4A4A and MS4A6A in LOAD brains. Caspase-4 expression was upregulated in CASP4/APP/PS1 mice in a region-specific manner, including hippocampus and prefrontal cortex. In APP/PS1 mice, caspase-4 expression led to impairments in the reversal phase of a Barnes maze task and in hippocampal synaptic plasticity, without affecting soluble or aggregated Aβ levels. Caspase-4 was expressed predominantly in microglial cells, and in the presence of CASP4, more microglia were clustered around amyloid plaques. Furthermore, our data indicated that caspase-4 modulates microglial cells in a manner that increases proinflammatory processes. We propose that microglial caspase-4 expression contributes to the cognitive impairments in AD, and that further study of caspase-4 will enhance our understanding of AD pathogenesis and may lead to novel therapeutic targets in AD.

Zfp189 Mediates Stress Resilience Through a CREB-Regulated Transcriptional Network in Prefrontal Cortex

Stress resilience involves numerous brain-wide transcriptional changes. Determining the organization and orchestration of these transcriptional events may reveal novel antidepressant targets, but this remains unexplored. Here, we characterize the resilient transcriptome with co-expression analysis and identify a single transcriptionally-active uniquely-resilient gene network. Zfp189, a previously unstudied zinc finger protein, is the top network key driver and its overexpression in prefrontal cortical (PFC) neurons preferentially activates this network, alters neuronal activity and promotes behavioral resilience. CREB, which binds Zfp189, is the top upstream regulator of this network. To probe CREB-Zfp189 interactions as a network regulatory mechanism, we employ CRISPR-mediated locus-specific transcriptional reprogramming to direct CREB selectively to the Zfp189 promoter. This single molecular interaction in PFC neurons recapitulates the pro-resilient Zfp189-dependent downstream effects on gene network activity, electrophysiology and behavior. These findings reveal an essential role for Zfp189 and a CREB-Zfp189 regulatory axis in mediating a central transcriptional network of resilience.

Brain Cell Type Specific Gene Expression and Co-expression Network Architectures

Elucidating brain cell type specific gene expression patterns is critical towards a better understanding of how cell-cell communications may influence brain functions and dysfunctions. We set out to compare and contrast five human and murine cell type-specific transcriptome-wide RNA expression data sets that were generated within the past several years. We defined three measures of brain cell type-relative expression including specificity, enrichment, and absolute expression and identified corresponding consensus brain cell “signatures,” which were well conserved across data sets. We validated that the relative expression of top cell type markers are associated with proxies for cell type proportions in bulk RNA expression data from postmortem human brain samples. We further validated novel marker genes using an orthogonal ATAC-seq dataset. We performed multiscale coexpression network analysis of the single cell data sets and identified robust cell-specific gene modules. To facilitate the use of the cell type-specific genes for cell type proportion estimation and deconvolution from bulk brain gene expression data, we developed an R package, BRETIGEA. In summary, we identified a set of novel brain cell consensus signatures and robust networks from the integration of multiple datasets and therefore transcend limitations related to technical issues characteristic of each individual study.

A common haplotype lowers SPI1 (PU.1) expression in myeloid cells and delays age at onset for Alzheimer's disease

In this study we used age at onset of Alzheimer’s disease (AD), cerebrospinal fluid (CSF) biomarkers, and eQTL datasets to fine map AD-associated GWAS loci and investigate the underlying mechanisms. In a genome-wide survival analysis of 40,255 samples, eight of the previously reported AD risk loci are significantly (p < 5×10−8) or suggestively (p < 1×10−5) associated with age at onset-defined survival and a further fourteen novel loci reached suggestive significance. One third (8/22) of these SNPs are cis-eQTLs in monocytes and/or macrophages, including rs7930318 associated with expression of MS4A4A and MS4A6A. The minor allele of rs1057233 (G), within the previously reported CELF1 AD risk locus, shows association with higher age at onset of AD (p=8.40×10−6), higher CSF levels of Aβ42 (p=1.2×10−4), and lower expression of SPI1 in monocytes (p = 1.50×10−105) and macrophages (p = 6.41×10−87). SPI1 encodes PU.1, a transcription factor critical for myeloid cell development and function. AD heritability is enriched within the SPI1 cistromes of monocytes and macrophages, implicating a myeloid PU.1 target gene network in the etiology of AD. Finally, experimentally altered PU.1 levels are correlated with phagocytic activity of BV2 mouse microglial cells and specific changes in the expression of multiple myeloid-expressed genes, including the mouse orthologs of AD risk genes, MS4A4A and MS4A6A. Our results collectively suggest that lower SPI1 expression reduces AD risk by modulating myeloid cell gene expression and function.