XX sex chromosome complement modulates immune responses to heat-killed Streptococcus pneumoniae immunization in a microbiome-dependent manner

Abstract

Background: Differences in male vs. female immune responses are well-documented and have significant clinical implications. While the immunomodulatory effects of sex hormones are well established, the contributions of sex chromosome complement (XX vs. XY) and gut microbiome diversity on immune sexual dimorphisms have only recently become appreciated. Here we investigate the individual and collaborative influences of sex chromosome complements and gut microbiota on humoral immune activation.

Methods: Male and female Four Core Genotype (FCG) mice were immunized with heat-killed Streptococcus pneumoniae (HKSP). Humoral immune responses were assessed, and X-linked immune-related gene expression was evaluated to explain the identified XX-dependent phenotype. The functional role of Kdm6a, an X-linked epigenetic regulatory gene of interest, was evaluated ex vivo using mitogen stimulation of B cells. Additional influences of the gut microbiome on sex chromosome-dependent B cell activation was also evaluated by antibiotically depleting gut microbiota prior to HKSP immunization. Reconstitution of the depleted microbiome with short-chain fatty acid (SCFA)-producing bacteria tested the impact of SCFAs on XX-dependent immune activation.

Results: XX mice exhibited higher HKSP-specific IgM-secreting B cells and plasma cell frequencies than XY mice, regardless of gonadal sex. Although Kdm6a was identified as an X-linked gene overexpressed in XX B cells, inhibition of its enzymatic activity did not affect mitogen-induced plasma cell differentiation or antibody production in a sex chromosome-dependent manner ex vivo. Enhanced humoral responses in XX vs. XY immunized FCG mice were eliminated after microbiome depletion, indicating that the microbiome contributes to the identified XX-dependent immune enhancement. Reconstituting microbiota-depleted mice with select SCFA-producing bacteria enhanced fecal SCFA concentrations and increased humoral responses in XX, but not XY, FCG mice. However, exposure to the SCFA propionate alone did not enhance mitogenic B cell stimulation in ex vivo studies.

Conclusions: FCG mice have been used to assess sex hormone and sex chromosome complement influences on various sexually dimorphic traits. The current study indicates that the gut microbiome impacts humoral responses in an XX-dependent manner, suggesting that the collaborative influence of gut bacteria and other sex-specific factors should be considered when interpreting data aimed at delineating the mechanisms that promote sexual dimorphism.

Keywords: Four Core Genotype; Gut microbiome; HKSP; IgM; Kdm6a; Plasma cells; SCFA; Sex differences; X chromosome.

Fig. 1

Possession of an XX vs. XY sex chromosome complement influences humoral responses to HKSP immunization. Immune responses against HKSP were assessed in FCG females and males one-week post-HKSP immunization. Numbers of HKSP-specific IgM-secreting B cells (A) and percentages of CD138 + plasma cells (B) were measured using ELISpot and flow cytometry, respectively. To assess the impact of sex hormones on sex chromosome-dependent phenotypes, percentages of CD138 + plasma cells were assessed in sham-operated (Sham) vs. gonadectomized (Gdx) female and male FCG mice following the same immunization protocol (C). Main effects of chromosomes and gonadal sex on variables assessed in A and B are presented by graphing the predicted (LS) means of each variable (D). Data in A-C are represented as the mean ± SEM with each data point representing one mouse. Statistical analyses by two-way ANOVA followed by Sidak’s multiple comparisons tests (A, B) or three-way ANOVA followed by Tukey’s multiple comparisons tests (C). Comparisons in graphs are representative of multiple comparisons tests; *p < 0.05; **p < 0.01; ***p < 0.001. ANOVA main and interactive effects are provided in Additional file 9: Tables S1-S2

Fig. 2

Kdm6a is differentially expressed in XX vs. XY splenocytes and B cells isolated from HKSP-immunized FCG mice. One-week post-HKSP immunization, RNA-Sequencing was performed on splenocytes isolated from male and female FCG mice. Levels of Kdm6a expression are represented as relative expression using EdgeR values (A). Differential expression of Kdm6a was confirmed by qRT-PCR using both splenocytes and B cells isolated from HKSP-immunized females (B) and males (C). KDM6a protein levels were assessed by western blot using lysates generated from HKSP-immunized FCG splenocytes (D, E). Relative levels of protein expression were quantified as KDM6a:beta-tubulin ratios (E). Kdm6a expression was also assessed in B cells from HKSP-immunized gonadectomized (Gdx) and sham-operated (Sham) females (F) and males (G, Additional file 9: Table S4). RNA-FISH was utilized to determine if Kdm6a was expressed from the inactivate X chromosome in B cells isolated from female XX FCG mice previously immunized with HKSP. Representative images of individual DAPI, Xist, and Kdm6a, as well as merged images, are presented (H). Colocalization of Kdm6a and Xist signals was considered indicative Kdm6a expression from the inactive X chromosome. Data are represented as the mean ± SEM with each data point representing one mouse. Statistical analyses by two-way ANOVA followed by Sidak’s multiple comparisons test (A, E), unpaired t-tests (B, C), or by three-way ANOVA followed by Tukey’s multiple comparisons test (F, G). Comparisons in graphs are representative of multiple comparisons tests; *p < 0.05; **p < 0.01; ****p < 0.0001. ANOVA main and interactive effects for A & E are provided in Additional file 9: Table S3

Fig. 3

KDM6a inhibition enhances plasma cell differentiation, but not IgM secretion, similarly in all four genotypes. Splenocytes isolated from naïve FCG mice were stimulated ex vivo with IL-4 (0.01 µg/mL) and LPS (5 µg/mL) in the presence or absence of 2 µM GSK J4 or 2 µM GSK J5. Representative density plots of flow cytometric data from naïve (day 0) or stimulated splenocytes (day 3) ± GSK J4 exposure are depicted in (A). Percentages of CD138 + plasma cells in FCG mice were quantified by flow cytometry (B). Supernatants were collected and total IgM concentrations were assessed by ELISA for each stimulation condition (C). Data are represented as the mean ± SEM with each data point representing one mouse. Three-way ANOVA with Tukey’s multiple comparisons tests. Comparisons in graphs are representative of multiple comparisons tests; **p < 0.01; ****p < 0.0001. ANOVA main and interactive effects are provided in Additional file 9: Table S6. Results from additional concentrations of GSK J4 and GSK J5 are provided in Additional file 5: Fig. S5 and Additional file 9: Table S7

Fig. 4

XX-specific enhancement in response to HKSP is dependent on the gut microbiome. The number of IgM-secreting B cells produced in response to HKSP immunization was assessed by ELISpot in male and female FCG mice possessing intact or antibiotically depleted gut microbiomes. Data are presented collectively for three-way ANOVA analyses (A) and separated by sex [females (B) and males (C)] for two-way ANOVA analyses to uncover chromosome roles independent of circulating sex hormones. Data are represented as the mean ± SEM with each point representing one mouse. Statistics by three-way ANOVA with Tukey’s multiple comparisons (A) or two-way ANOVA with Tukey’s multiple comparisons test (B, C). Significance indicated are representative of Tukey’s multiple comparisons tests. Comparisons in graphs are representative of multiple comparisons tests; **p < 0.01; ***p < 0.001; ****p < 0.0001. Abx = antibiotics. ANOVA main and interactive effects are provided in Additional file 9: Tables S8-S9

Fig. 5

Assessment of gut microbiota diversity in FCG mice. 16s rRNA gene sequencing and metagenomic analyses were performed to assess microbiota diversity in FCG mice. The number of different OTUs as a function of the number of sequence reads (A) and Shannon Diversity Indexes (B) were determined for gonadally intact and gonadectomized (Gdx) animals. Representative PCoA plots of Bray–Curtis pair-wise comparison distances demonstrate clustering differences between males and females in both intact (C) and gonadectomized (D) animals. For each distinct OTU identified, percent abundancies were calculated for intact animals (E, Additional file 9: Table S12). The total height of y-axis represents 100% of the assigned sequences after quality filtering, and the size of the colored regions represents proportional contributions of each phylotype shown with the top 11 families being visualized. Abundances greater than 5% are labeled with percentages. Data in A and B are represented as the mean ± SEM with each point representing one mouse. Statistics by Kruskal–Wallis test for A and B, followed by Dunn’s multiple comparisons tests provided in Additional file 9: Table S10. Bray–Curtis comparisons between each group are provided in Additional file 9: Table S11. A full list of taxon abundancies in E are provided in Additional file 9: Table S12. Gdx = gonadectomized; OTUs = observable taxonomic units

Fig. 6

Concentrations of SCFA in the feces of male and female FCG mice. Fecal pellets from naïve FCG mice were collected and analyzed for a panel of short-chain fatty acids: acetate (A), butyrate (B), propionate (C), 2-methylbutyric acid (D), hexanoic acid (caproic acid, E), isobutyric acid (F), and isovaleric acid (G) by LC–MS/MS (Metabolon). Data are represented as the mean ± SEM with each point representing one mouse. Statistics by two-way ANOVA with Tukey’s multiple comparisons test. Comparisons in graphs are representative of multiple comparisons tests; *p < 0.05; **p < 0.01. ANOVA main and interactive effects are provided in Additional file 9: Table S13

Fig. 7

Reconstitution of gut microbiota with SCFA-producing bacteria increased humoral responses in an XX-dependent manner. Overview of experimental design (A). Briefly, gut microbiota of FCG mice were depleted using antibiotic oral gavage (3 days). On Day 4, mice were administered one of the following via oral gavage: inulin alone (Group 1); SCFA-producing bacteria alone (Group 2); or Inulin + SCFA-producing bacteria (Group 3). On Day 6, all mice were immunized with 2 × 10⁸ CFU heat-killed Streptococcus pneumoniae. Mice in the inulin alone exposed group received antibiotics by oral gavage every other day through Day 12, while mice receiving SCFA-producing bacteria ± inulin received water. Mice were euthanized on Day 13 and samples collected for immune response evaluations. Fecal pellets were collected on Day 0, Day 4 (prior to gavage treatments), and Day 13. The number of living bacteria per pellet for females (B) and males (C) was assessed by flow cytometry to assess successful gut microbiota depletion and reconstitution. Immune responses were evaluated as numbers of IgM-secreting B cells and analyzed as Inulin alone vs. SCFA-producing bacteria (D) or as Inulin alone vs. Inulin + SCFA-producing bacteria (E). Concentrations of acetate (F), butyrate (G), and propionate (H), three well-studied SCFAs with immunomodulatory function, were assessed by LC–MS/MS (Metabolon) to confirm effective SCFA-producing bacteria recolonization and subsequent SCFA production. The concentrations of additional SCFAs are provided in Additional file 7: Fig. S7. Data point labels in B and C indicate the sex chromosome complement (XX or XY) and the group number as indicated in A. Data are represented as the mean ± SEM with each point representing one mouse. Statistics by three-way ANOVA with Tukey’s multiple comparisons test. Comparisons in graphs are representative of multiple comparisons tests; *p < 0.05. ANOVA main and interactive effects are provided in Additional file 9: Tables S14 and S15

Fig. 8

Propionate decreases plasma cell frequencies and IgM secretion similarly in all four genotypes ex vivo. Splenocytes isolated from naïve FCG mice were stimulated ex vivo with IL-4 (0.01 µg/mL) and LPS (5 µg/mL) in the presence or absence of 1 mM C3 (propionate). Flow cytometric analyses evaluated cell viability (A) and the number of CD138 + plasma cells (B) 4 days post-stimulation. Supernatants were collected and total IgM concentrations were assessed by ELISA (C). Data are represented as the mean ± SEM with each point representing one mouse. Statistics by three-way ANOVA with Tukey’s multiple comparisons test. ANOVA results provided in Additional file 9: Table S16. Comparisons in graphs are representative of multiple comparisons tests; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. C3 = propionate. Additional concentrations of C3 are presented in Additional file 8: Fig. S8 with accompanying ANOVA table in Additional file 9: Table S17