Evolutionary conservation and divergence of the human brain transcriptome

Abstract

Background Mouse models have allowed for the direct interrogation of genetic effects on molecular, physiological, and behavioral brain phenotypes. However, it is unknown to what extent neurological or psychiatric traits may be human- or primate-specific and therefore which components can be faithfully recapitulated in mouse models. Results We compare conservation of co-expression in 116 independent data sets derived from human, mouse, and non-human primate representing more than 15,000 total samples. We observe greater changes occurring on the human lineage than mouse, and substantial regional variation that highlights cerebral cortex as the most diverged region. Glia, notably microglia, astrocytes, and oligodendrocytes are the most divergent cell type, three times more on average than neurons. We show that cis-regulatory sequence divergence explains a significant fraction of co-expression divergence. Moreover, protein coding sequence constraint parallels co-expression conservation, such that genes with loss of function intolerance are enriched in neuronal, rather than glial modules. We identify dozens of human neuropsychiatric and neurodegenerative disease risk genes, such as COMT, PSEN-1, LRRK2, SHANK3, and SNCA, with highly divergent co-expression between mouse and human and show that 3D human brain organoids recapitulate in vivo co-expression modules representing several human cell types. Conclusions We identify robust co-expression modules reflecting whole-brain and regional patterns of gene expression. Compared with those that represent basic metabolic processes, cell-type-specific modules, most prominently glial modules, are the most divergent between species. These data and analyses serve as a foundational resource to guide human disease modeling and its interpretation.

In search of a core cellular network with single cell transcriptome data in the fruit fly

“…Our observation of core modules with genes of different ages is in line with previous comparative studies based on bulk transcriptomics, which suggest that core modules may have evolved at different times (Pembroke, Hartl, and Geschwind 2021; Stuart et al 2003). Our work however points toward a more limited core network, both in size and in function, than these studies.…”

Functional and molecular characterization of a non-human primate model of autism spectrum disorder shows similarity with the human disease

“…For example, the Fmr1 knockout mouse had an astrocyte-related module that tended to be modulated in the same direction as human ASD, and the Mecp2 heterozygous mouse had a neuron/oligodendrocyte-related module modulated in the same direction with human ASD, but none of these modules were significantly concordant with human ASD. The lower concordance between rodents and humans is consistent with a greater divergence of gene coexpression networks, especially in glial-cell-related modules, between rodents and humans than between non-human primates and humans 13 . The replication of gene expression modulations in glia-related modules in the marmoset model seems particularly important, since ASD is thought to be caused by a mechanism involving glial regulation of synaptic formation and function 59 .…”

“…ASD affects primate-specific brain areas involved in social functions unique to primates (e.g., the medial prefrontal cortex) 10 , 11 . Moreover, the developmental expression pattern of some genes is primate specific 12 , 13 , and proteome data suggest primate-specific synaptic functions 14 . Different molecular networks between humans and rodents may limit the utility of rodent models for human diseases, and therefore, ASD primate models have advantages over rodent models (also see ref.…”

Comparative Analyses of Gene Co-expression Networks: Implementations and Applications in the Study of Evolution

“…Next, WGCNA uses a topological overlap measure (TOM), which is a combination of the adjacency value between a pair of genes as well as the adjacency values these genes have with other genes to which they are connected. WGCNAis also utilized for its module conservation statistics to make comparisons across modules of different clusterings and species (Du et al, 2020;Pembroke et al, 2021). In order to measure the preservation of a module, WGCNA can be used to determine if it is reproducible (or preserved) in an independent test network.…”