Interplay between murine sex-biased gene expression and hepatic zonation revealed by single nucleus RNA sequencing

Abstract

The zonation of liver metabolic processes is well-characterized; however, little is known about the cell type-specificity and zonation of sexually dimorphic gene expression or its growth hormone (GH)-dependent transcriptional regulators. We address these issues using single nucleus RNA sequencing of 32,000 nuclei representing nine major liver cell types. Nuclei were extracted from livers from young adult male and female mice, from male mice infused with GH continuously to mimic the female plasma GH pattern, and from mice treated with TCPOBOP, a xenobiotic agonist ligand of the liver nuclear receptor CAR (Nr1i3). Analysis of these rich transcriptomic datasets revealed: 1) expression of sex-biased genes and their key GH-dependent transcriptional regulators is primarily restricted to hepatocytes and is not a feature of liver non-parenchymal cells; 2) many sex-biased transcripts show sex-dependent zonation within the liver lobule; 3) gene expression is substantially feminized in both periportal and pericentral hepatocytes when male mice are infused with GH continuously; 4) sequencing nuclei increases the sensitivity for detecting thousands of nuclear-enriched lncRNAs and enables determination of their liver cell type-specificity, sex bias and hepatocyte zonation profiles; 5) the periportal to pericentral hepatocyte cell ratio is significantly higher in male than female liver; and 6) TCPOBOP exposure disrupts sex-specific gene expression and hepatocyte zonation within the liver lobule. These findings highlight the complex interconnections between hepatic sexual dimorphism and zonation at the single cell level and reveal how endogenous hormones and foreign chemical exposure can alter these interactions across the liver lobule with large effects on both protein-coding genes and lncRNAs.

Constitutively active STAT5b feminizes mouse liver gene expression

“…RNA-seq data was analyzed using a modified version of the custom pipeline described earlier (15). Briefly, sequence reads were aligned to mouse genome build mm9 (NCBI 37) using STAR aligner (38) and FeatureCounts (39) was used to count sequence reads mapping to the union of the exonic regions in all isoforms of a given gene (collapsed exon counting) based on an mm9 Gene Transfer Format file comprised of 75,798 mouse genes: n = 20,884 RefSeq protein coding genes, n=48,360 mouse liver-expressed lncRNAs genes (a much more complete listing than the set described previously in (40)), n=2,061 RefSeq non-coding genes (NR accession numbers) that do not overlap the set of 48,360 lncRNAs, n =4,490 Ensembl non-coding lncRNAs that do not overlap either the RefSeq NR gene set or the 48,360 lncRNA gene set, and n=3 other lncRNAs with interesting liver functions (lnc-LFAR1, LeXis, Lnclgr) (41) and in the gene annotations provided in Table S1 . Raw sequencing files and processed data files are available at GEO (https://www.ncbi.nlm.nih.gov/geo/) under accession numbers GSE196014 and GSE196015.…”