Integrated analysis of robust sex-biased gene signatures in human brain

Relationship between sex biases in gene expression and sex biases in autism and Alzheimer's disease

BACKGROUND

Sex differences in the brain may play an important role in sex-differential prevalence of neuropsychiatric conditions.

METHODS

In order to understand the transcriptional basis of sex differences, we analyzed multiple, large-scale, human postmortem brain RNA-Seq datasets using both within-region and pan-regional frameworks.

RESULTS

We find evidence of sex-biased transcription in many autosomal genes, some of which provide evidence for pathways and cell population differences between chromosomally male and female individuals. These analyses also highlight regional differences in the extent of sex-differential gene expression. We observe an increase in specific neuronal transcripts in male brains and an increase in immune and glial function-related transcripts in female brains. Integration with single-nucleus data suggests this corresponds to sex differences in cellular states rather than cell abundance. Integration with case-control gene expression studies suggests a female molecular predisposition towards Alzheimer's disease, a female-biased disease. Autism, a male-biased diagnosis, does not exhibit a male predisposition pattern in our analysis.

CONCLUSION

Overall, these analyses highlight mechanisms by which sex differences may interact with sex-biased conditions in the brain. Furthermore, we provide region-specific analyses of sex differences in brain gene expression to enable additional studies at the interface of gene expression and diagnostic differences.

Novel missense variants in brain morphogenic genes associated with depression and schizophrenia

Introduction

Impaired function of brain morphogenic genes is considered one of the predisposing factors for the manifestation of psychiatric and cognitive disorders, such as paranoid schizophrenia (SCZ) and major depressive disorder (MDD). Identification of such genes (genes of neurotrophic factors and guidance molecules among them) and their deleterious genetic variants serves as a key to diagnosis, prevention, and possibly treatment of such disorders. In this study, we have examined the prevalence of genomic variants in brain morphogenic genes in individuals with SCZ and MDD within a Russian population.

Methods

We have performed whole-exome sequencing of 21 DNA samples: 11 from individuals with SCZ and 10 with MDD, followed by ARMS (Amplification-Refractory Mutation System) based screening of detected single nucleotide variants (SNVs) in larger groups: 102 for individuals with SCZ, 79 for those with MDD and 103 for healthy donors.

Results

Whole-exome sequencing has revealed 226 missense mutations in 79 genes (out of 140 studied), some of which occur in patients with psychiatric disorders significantly more frequently than in healthy donors. We have identified previously undescribed genomic variants in brain morphogenic genes: CDH2 (rs1944294-T and rs17445840-T), DCHS2 (rs11935573-G and rs12500437-G/T) and CDH23 (rs1227051-G/A), significantly associated with the incidence of SCZ and MDD in the Russian population. For some SNVs (rs6265-T, rs1944294-T, rs11935573-G, rs4760-G) sex-biased differences in their prevalence between SCZ/MDD patients and healthy donors was detected.

Discussion

However, the functional significance of the SNVs identified has still to be confirmed in cellular and animal models. Once it is fulfilled, these SNVs have the potential to complement the diagnostic toolbox for assessing susceptibility to mental disorders. The data obtained indirectly confirm the importance of adequate brain structure formation for its correct functioning and preservation of mental health.

Blood extracellular vesicles carrying brain-specific mRNAs are potential biomarkers for detecting gene expression changes in the female brain

The absence of non-invasive tests that can monitor the status of the brain is a major obstacle for psychiatric care. In order to address this need, we assessed the feasibility of using tissue-specific gene expression to determine the origin of extracellular vesicle (EV) mRNAs in peripheral blood. Using the placenta as a model, we discovered that 26 messenger RNAs that are specifically expressed in the placenta are present in EVs circulating in maternal blood. Twenty-three of these transcripts were either exclusively or highly expressed in maternal blood during pregnancy only and not in the postpartum period, verifying the feasibility of using tissue-specific gene expression to infer the tissue of origin for EV mRNAs. Using the same bioinformatic approach, which provides better specificity than isolating L1 cell-adhesion molecule containing EVs, we discovered that 181 mRNAs that are specifically expressed in the female brain are also present in EVs circulating in maternal blood. Gene set enrichment analysis revealed that these transcripts, which are involved in synaptic functions and myelination, are enriched for genes implicated in mood disorders, schizophrenia, and substance use disorders. The EV mRNA levels of 13 of these female brain-specific transcripts are associated with postpartum depression (adjusted p-vals = 3 × 10-5 to 0.08), raising the possibility that they can be used to infer the state of the brain. In order to determine the extent to which EV mRNAs reflect transcription in the brain, we compared mRNAs isolated from cells and EVs in an iPSC-derived brain microphysiological system differentiated for 3 and 9 weeks. We discovered that, although cellular and extracellular mRNA levels are not identical, they do correlate, and it is possible to extrapolate cellular RNA expression changes in the brain via EV mRNA levels. Our findings bring EV mRNAs to the forefront of peripheral biomarker development efforts in psychiatric diseases by demonstrating the feasibility of inferring transcriptional changes in the brain via blood EV mRNA levels.

Interrogation of the human cortical peptidome uncovers cell-type specific signatures of cognitive resilience against Alzheimer's disease

Alzheimer's disease (AD) is characterised by age-related cognitive decline. Brain accumulation of amyloid-β plaques and tau tangles is required for a neuropathological AD diagnosis, yet up to one-third of AD-pathology positive community-dwelling elderly adults experience no symptoms of cognitive decline during life. Conversely, some exhibit chronic cognitive impairment in absence of measurable neuropathology, prompting interest into cognitive resilience-retained cognition despite significant neuropathology-and cognitive frailty-impaired cognition despite low neuropathology. Synapse loss is widespread within the AD-dementia, but not AD-resilient, brain. Recent evidence points towards critical roles for synaptic proteins, such as neurosecretory VGF, in cognitive resilience. However, VGF and related proteins often signal as peptide derivatives. Here, nontryptic peptidomic mass spectrometry was performed on 102 post-mortem cortical samples from individuals across cognitive and neuropathological spectra. Neuropeptide signalling proteoforms derived from VGF, somatostatin (SST) and protachykinin-1 (TAC1) showed higher abundance in AD-resilient than AD-dementia brain, whereas signalling proteoforms of cholecystokinin (CCK) and chromogranin (CHG) A/B and multiple cytoskeletal molecules were enriched in frail vs control brain. Integrating our data with publicly available single nuclear RNA sequencing (snRNA-seq) showed enrichment of cognition-related genes in defined cell-types with established links to cognitive resilience, including SST interneurons and excitatory intratelencephalic cells.

Relationship between sex biases in gene expression and sex biases in autism and Alzheimer's disease

Sex differences in the brain may play an important role in sex-differential prevalence of neuropsychiatric conditions. In order to understand the transcriptional basis of sex differences, we analyzed multiple, large-scale, human postmortem brain RNA-seq datasets using both within-region and pan-regional frameworks. We find evidence of sex-biased transcription in many autosomal genes, some of which provide evidence for pathways and cell population differences between chromosomally male and female individuals. These analyses also highlight regional differences in the extent of sex-differential gene expression. We observe an increase in specific neuronal transcripts in male brains and an increase in immune and glial function-related transcripts in female brains. Integration with single-cell data suggests this corresponds to sex differences in cellular states rather than cell abundance. Integration with case-control gene expression studies suggests a female molecular predisposition towards Alzheimer's disease, a female-biased disease. Autism, a male-biased diagnosis, does not exhibit a male predisposition pattern in our analysis. Finally, we provide region specific analyses of sex differences in brain gene expression to enable additional studies at the interface of gene expression and diagnostic differences.

Insights into Sex and Gender Differences in Brain and Psychopathologies Using Big Data

The societal implication of sex and gender (SG) differences in brain are profound, as they influence brain development, behavior, and importantly, the presentation, prevalence, and therapeutic response to diseases. Technological advances have enabled speed up identification and characterization of SG differences during development and in psychopathologies. The main aim of this review is to elaborate on new technological advancements, such as genomics, imaging, and emerging biobanks, coupled with bioinformatics analyses of data generated from these technologies have facilitated the identification and characterization of SG differences in the human brain through development and psychopathologies. First, a brief explanation of SG concepts is provided, along with a developmental and evolutionary context. We then describe physiological SG differences in brain activity and function, and in psychopathologies identified through imaging techniques. We further provide an overview of insights into SG differences using genomics, specifically taking advantage of large cohorts and biobanks. We finally emphasize how bioinformatics analyses of big data generated by emerging technologies provides new opportunities to reduce SG disparities in health outcomes, including major challenges.