ImmGen report: sexual dimorphism in the immune system transcriptome

Abstract

Sexual dimorphism in the mammalian immune system is manifested as more frequent and severe infectious diseases in males and, on the other hand, higher rates of autoimmune disease in females, yet insights underlying those differences are still lacking. Here we characterize sex differences in the immune system by RNA and ATAC sequence profiling of untreated and interferon-induced immune cell types in male and female mice. We detect very few differentially expressed genes between male and female immune cells except in macrophages from three different tissues. Accordingly, very few genomic regions display differences in accessibility between sexes. Transcriptional sexual dimorphism in macrophages is mediated by genes of innate immune pathways, and increases after interferon stimulation. Thus, the stronger immune response of females may be due to more activated innate immune pathways prior to pathogen invasion.

Fig. 1

Cell-type transcriptional signature masks the sex effect. a Overview of the datasets used in this study. All four datasets include samples from male and female mice, from some or all of ImmGen 11 cell set, whose cell types are shown in the tree on the top. On the bottom, details of the four datasets used. PC peritoneal cavity, Sp spleen, CNS central nervous system, ATAC-seq Assay for Transposase-accessible Chromatin, GN granulocytes, DC dendritic cells, MF macrophages, T4 CD4⁺ T cells, T8 CD8⁺ T cells, Treg regulatory T cells, NK natural killer cells, NKT natural killer T cells, Tgd gamma delta T cells. b Pearson correlation matrix of the samples comprising datasets A and B (including the entire 11 cell set) for the 2904 genes that are expressed in both datasets, sorted by cell type and sex. Black rectangles mark samples of the same cell type. Source data for b are provided as a Source Data file

Fig. 2

Pan-immune transcriptional sexual dimorphism. a Heatmap of log₂ fold change (female–male expression) of pan-immune female and male sexually differentially expressed genes (SDEGs) (datasets A and B, paired t test pFDR < 0.2, female–male fold change >1.5). For genes with multiple names, only the first is shown and then period. Source data included in Supplementary Table 2. b Female enriched pathways identified by pre-ranked gene set enrichment analysis [GSEAPreranked (FDR < 25%) of the t statistic, sorted by normalized enrichment scores]. FDR corrected p value is displayed to the right of each bar. Source data included in Supplementary Table 3

Fig. 3

Cell-type-specific sex signature. a Number of male (light blue) and female (pink) sexually differentially expressed genes (SDEGs) in each cell type (datasets A and B, paired t test pFDR < 0.2, female–male fold change >1.5). b Heatmap of relative expression levels of male and female SDEGs in MFs from datasets A and B. Genes are sorted by female–male fold change. Expression values are trimmed to range [−2, 2]. For genes with multiple names, only the first is shown and then period. c Heatmap of relative expression levels of 71 previously identified MF IFN-stimulated genes (MF-ISGs) in datasets A and B. Gene symbols are indicated for MF-ISGs that are also SDEGs or known IFNα-response genes (according to MsigDB Hallmark gene sets). Expression values are trimmed to range [−1.5, 1.5]. In b, c, dotted horizontal white lines separate female from male upregulated genes. Dotted vertical white lines separate female and male samples. d Female–male fold change (log₂) distribution of MF-ISGs (71 genes, green) and all genes (1568 genes, excluding MF-ISGs, orange). The t test p value between distributions is 7.6 × 10⁻¹², with permutation p value <0.001. e Box plot presenting female–male log₂ fold change (FC) distribution of the genes in modules C1–C5 defined by Mostafavi et al. from unstimulated macrophages. On each box, the central mark indicates the median, and the bottom and top edges of the box indicate the 25th and 75th percentiles, respectively. The whiskers extend to the most extreme data points not considered outliers, and the outliers are plotted individually in red. Modules whose distributions are significantly different from zero (one-sample two-sided t test pFDR < 0.05) are marked by asterisk. Source data for b–e are provided as a Source Data file

Fig. 4

Interferon boosts the transcriptional sexually differential immune response. a Principal component analysis (PCA) of GN, MF, and B cells (circles, triangles, and diamonds, respectively) of females and males (pink and light blue) with and without IFN (filled or open shapes, respectively), based on 1000 autosomal genes with the highest standard deviation. Axes are the first three principal components (PCs). b The distribution of the log₂ fold change values between IFN-stimulated MF and unstimulated MF for the IFN upregulated genes in male and/or female MF. Distribution for female samples is shown in pink and for male in light blue. The t test p value between distributions is 2.7 × 10⁻¹³ with permutation p value <0.001. c Expression heatmap of 313 MF IFN-stimulated genes (MF-ISGs) expressed in unstimulated and IFN-stimulated male and female macrophage samples (dataset B). Expression values are trimmed to range [−2, 2]. Dotted vertical white line separate IFN-stimulated and unstimulated samples. d Female–male fold change distribution of MF-ISGs and all remaining genes from unstimulated samples (red and light blue, respectively) and MF-ISGs and all remainder genes from IFN-stimulated samples (green and yellow, respectively). e Box plot presenting female–male log₂ fold change (FC) distribution of the genes in modules C1–C5 defined by Mostafavi et al. from IFN-stimulated macrophages. On each box, the central mark indicates the median, and the bottom and top edges of the box indicate the 25th and 75th percentiles, respectively. The whiskers extend to the most extreme data points not considered outliers, and the outliers are plotted individually in red. Unstm unstimulated, IFN interferon. Modules whose distributions are significantly different from zero (one-sample two-sided t test pFDR < 0.05) are marked by asterisk. Source data for a–e are provided as a Source Data file

Fig. 5

ATAC-seq data of MFs display sexual dimorphism in chromatin accessibility. a Female–male fold change (log₂ gene body OCR value) distribution of MF-specific female (pink) and male (blue) SDEGs and all other genes (termed all genes, excluding SDEGs, black). b Sexual differential accessibility in macrophage TSS OCRs (adjusted two-sided t test pFDR < 0.05 and female–male fold change >2, shown in blue). c Sexual differential accessibility in macrophage distal enhancer OCRs (adjusted two-sided t test pFDR < 0.05 and female–male fold change >2, shown in blue). d The Firre locus chromatin accessibility data on the UCSC data browser (http://rstats.immgen.org/Chromatin/chromatin.html) in MF male (top) and female (bottom). Source data for a–c are provided as a Source Data file

Fig. 6

Sexual dimorphism in macrophages across tissues. a Pearson correlation matrix of MFs from the peritoneal cavity (PC-MF, green), spleen (Sp-MF, yellow), and microglia (labeled CNS, purple), based on 10,264 genes expressed above the noise threshold, sorted by cell type and sex. b Heatmap of the expression values of genes that are significantly contributing (two-way ANOVA, pFDR < 0.2, female–male fold change >log₂(1.5) in all three tissues) to sex effect across tissues (16 genes). c Heatmap of the expression values of genes that are significantly contributing to the sex–tissue interaction effect in specific tissue (two-way ANOVA pFDR < 0.2, male–female tissue-specific fold change >log₂(1.5)): left PC-MF, middle Sp-MF, 127 female and 101 male specific genes; right microglia (labeled CNS), 5 female and 43 male specific genes. In b, c, dotted horizontal white lines separate female from male upregulated genes. Dotted vertical white lines separate female and male samples. Values were normalized per tissue. Source data for a–c are provided as a Source Data file

Fig. 7

Comparison of human and mouse transcriptional immune sexual dimorphism. Female–male fold change (log₂) of ImmVar genes that are differentially expressed between female and male in a human CD14 and mouse MF cells from dataset B, and b human CD4 and mouse T4 cells from dataset B. The colored genes have >0.1 (log₂) fold change in both datasets in females (pink) or males (light blue). Selected gene symbols are shown. Source data for a, b are provided as a Source Data file