CD11c identifies microbiota and EGR2-dependent MHCII+ serous cavity macrophages with sexually dimorphic fate in mice

Abstract

The murine serous cavities contain a rare and enigmatic population of short-lived F4/80^lo MHCII⁺ macrophages but what regulates their development, survival, and fate is unclear. Here, we show that mature F4/80^lo MHCII⁺ peritoneal macrophages arise after birth, but that this occurs largely independently of colonization by microbiota. Rather, microbiota specifically regulate development of a subpopulation of CD11c⁺ cells that express the immunoregulatory cytokine RELM-α, are reliant on the transcription factor EGR2, and develop independently of the growth factor CSF1. Furthermore, we demonstrate that intrinsic expression of RELM-α, a signature marker shared by CD11c⁺ and CD11c^- F4/80^lo MHCII⁺ cavity macrophages, regulates survival and differentiation of these cells in the peritoneal cavity in a sex-specific manner. Thus, we identify a previously unappreciated diversity in serous cavity F4/80^lo MHCII⁺ macrophages that is regulated by microbiota, and describe a novel sex and site-specific function for RELM-α in regulating macrophage endurance that reveals the unique survival challenge presented to monocyte-derived macrophages by the female peritoneal environment.

Keywords: macrophage ⋅ peritoneal cavity ⋅ regulation.

Figure 1

Development of peritoneal macrophages after birth. Peritoneal cells were isolated from mice of mixed sex at various ages between 4 days and 17 weeks and analyzed by flow cytometry (for full gating strategy, see Supporting information Fig. S1). (A) Representative expression of F4/80 and MHCII by live CD45⁺ Lin^– CD11b⁺ peritoneal cells. Data are from one experiment representative of three performed. (B) Number of F4/80^hi and F4/80^loMHCII⁺ peritoneal MNPs obtained at the indicated ages. Symbols represent individual animals with line at mean of 9–13 mice per group pooled from three independent experiments. (C) Ratio of F4/80^hi to F4/80^lo MHCII⁺ peritoneal macrophages from data in (B). (D) Representative expression of intracellular RELM‐α by total F4/80^loMHCII⁺ MNPs (left) and proportion of RELM‐α expressing F4/80^lo MHCII⁺ MNPs and F4/80^hi resident macrophages at the indicated ages. Symbols represent individual animals with line at mean of three to four mice per time point from one experiment. (E) Representative expression of Ly6C and MHCII by CD45⁺ Lin^– CD11b⁺ F4/80^lo cells (left) and proportion of Ly6C⁺ F4/80^lo cells that express MHCII (right) at the indicated ages. Symbols represent individual animals with line at mean of six to eight mice per group pooled from two independent experiments. (F) Representative expression of CD115 and CD11c by F4/80^lo MHCII⁺ MNPs (upper) and CD11c expression by CD115⁺ F4/80^lo MHCII⁺ MNPs (lower) at the indicated ages. Right, proportion of CD11c expressing cells among CD115⁺ F4/80^lo MHCII⁺ MNPs. Symbols represent individual animals with line at mean of six to eight mice per group pooled from two independent experiments. Symbols on graphs represent individual mice with an n of 9–13 (B), 3–4 (D), or 6–8 (E‐F) mice per time point pooled from 1 (D), 2 (E‐F), or 3 (B) experiments. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 (C‐F; one‐way ANOVA).

Figure 2

Effects of microbial colonization on peritoneal macrophages. (A) Representative Ly6C and MHCII expression by live Lin^– CD11b⁺ CD115⁺ F4/80^lo peritoneal leukocytes obtained from 6‐ to 9‐week‐old SPF or GF male mice. (B) Absolute number of Ly6C⁺ MHCII^– and Ly6C⁺ MHCII⁺ cells in the peritoneal cavity of SPF and GF mice detailed in (A). (C) Absolute number of CD11c^– and CD11c⁺ fractions of Ly6C^– MHCII⁺ CD115⁺ F4/80^lo macrophages in the peritoneal cavity of SPF and GF mice detailed in (A). (D)Frequency of RELM‐α⁺ cells within CD11c^– and CD11c⁺ fractions of Ly6C^–MHCII⁺CD115⁺F4/80^lo cells from the peritoneal cavity of SPF of GF mice detailed in (A). (E) Absolute number of F4/80^hi macrophages in the peritoneal cavity of SPF or GF mice detailed in (A). (F) Frequency of RELM‐α⁺ cells within F4/80^hi macrophages from the peritoneal cavity of SPF or GF mice detailed in (A). (G) Relative number of F4/80^hi macrophages and CD11c^– and CD11c⁺ fractions of Ly6C^– MHCII⁺ CD115⁺ F4/80^lo macrophages of male mice treated with an antibiotic “cocktail “of vancomycin, neomycin, ampicillin, and metronidazole (ABX) or without (control = Con) antibiotics normalized to numbers in controls. H. Frequency of RELM‐α⁺ (left) and CD226⁺ (right) cells within CD11c^– and CD11c⁺ fractions of Ly6C^–MHCII⁺CD115⁺F4/80^lo cells from the peritoneal cavity of mice detailed in (G). (I) Frequency of RELM‐α⁺ cells within F4/80^hi macrophages from the peritoneal cavity of mice detailed in (G). Symbols on graphs represent individual mice with an n = 8 (ABX study) or 10 (GF study) animals per group pooled from two independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001 (C, D, G, H, Student's t‐test with Holm–Sidak correction; E, F, I, Student's t‐test).

Figure 3

EGR2 controls development of CD11c⁺ peritoneal F4/80^loMHCII⁺ macrophages. (A) Heat map showing expression of Egr2 and Irf4 in the indicated peritoneal and blood populations from the ImmGen Consortium. The heat map displays the relative log2‐normalized expression per gene, calculated as the gene expression value of each cell type minus the mean expression value per gene over all cells in the heatmap. B. Representative expression of EGR2 by Ly6C⁺ MHCII^– and Ly6C⁺ MHCII⁺ monocytes, CD11c^– and CD11c⁺ fractions of Ly6C^– MHCII⁺ CD115⁺ F4/80^lo macrophages and F4/80^hi resident macrophages obtained from the peritoneal cavity of Egr2 ^fl/fl mice (left) and mean fluorescence intensity of EGR2 expression in these populations from six male (square symbols) and five female (circles) Egr2 ^fl/fl mice from one of three representative experiments. ****p < 0.0001 (one‐way ANOVA). (C) Absolute number of Ly6C⁺ MHCII^– and Ly6C⁺ MHCII⁺ monocytes, CD11c^– and CD11c⁺ fractions of Ly6C^– MHCII⁺ CD115⁺ F4/80^lo macrophages, F4/80^hi resident macrophages and CD11b⁺ DC in the peritoneal cavity of male and female Lyz2 ^Cre .Egr2 ^fl/fl (Cre⁺) and Egr2 ^fl/fl (Cre^–) mice. Symbols represent individual animals with 10 (male Cre^–), 11 (male Cre⁺), 12 (female Cre^–), and 13 (female Cre⁺) mice per group pooled from two to three independent experiments. **p < 0.01, ***p < 0.001 (Student's t‐test with Holm–Sidak correction). (D) Frequency of RELM‐α⁺ cells within CD11c^– and CD11c⁺ fractions of Ly6C^–MHCII⁺CD115⁺F4/80^lo cells from the peritoneal cavity of Cre⁺ and Cre^– mice in (C). Symbols represent individual animals, with cells 10 (male Cre^–), 11 (male Cre⁺), and 8 (female Cre⁺ and Cre⁺) mice per group pooled from two independent experiments. ***p < 0.001, ****p < 0.0001 (Student's t‐test with Holm–Sidak correction). (E) Frequency of RELM‐α⁺ cells within F4/80^hi peritoneal macrophages from Cre⁺ and Cre^– mice in (C). Symbols represent individual animals, with 13 (male and female Cre⁺), 15 (female Cre^–), and 17 (male Cre^–) mice per group pooled from four independent experiments. **p < 0.01 (Student's t‐test with Holm–Sidak correction).

Figure 4

Inter‐relationship of peritoneal macrophage subsets. (A) Absolute number of peritoneal F4/80^hi and F4/80^loMHCII⁺ macrophages (left) and frequency of CD11c^– RELM‐α⁺ and CD11c⁺ RELM‐α⁺ cells within F4/80^loMHCII⁺ peritoneal macrophages (right) from Csf1r^+/+ and Csf1r ^{ΔFIRE/ΔFIRE} mice. Symbols represent individual male (blue) or female (red) animals, with 4 Csf1r^+/+ and 3 Csf1r ^{ΔFIRE/ΔFIRE} mice pooled from two independent experiments. **p < 0.01 (Student's t‐test with Holm–Sidak correction). (B) Frequency of RFP⁺ cells within the indicated cell populations from the peritoneal cavity of CD64^iCre/+.Rosa26 ^LSL‐RFP/+ mice (both sexes). Symbols represent seven individual animals pooled from three independent experiments. p > 0.05 for all comparisons, except Ly6C⁺ MHCII^– versus CD11b⁺ DC that was not significant (one‐way ANOVA with Tukey's multiple comparisons test). (C) Experimental schematic for construction of mixed bone marrow chimeric mice. (D) Relative frequency of CD45.1^– CD45.2⁺ cells to CD45.1⁺ CD45.2^– cells within blood Ly6C^hi monocytes from male mixed bone marrow chimeric given bone marrow from WT CD45.1^– CD45.2⁺ and CD45.1^–CD45.2⁺ bone marrow from Lyz2 ^Cre.Egr2 ^fl/fl (Cre⁺) or Egr2 ^fl/fl (Cre^–) mice. Symbols represent individual mice. Data represent eight mice per group pooled from two independent experiments. (E) As in (D), but for female bone marrow chimeric mice. Data represent eight mice per group pooled from two independent experiments. (F) Contribution of CD45.1^– CD45.2⁺ Egr2 ^fl/fl bone marrow to the indicated peritoneal and pleural cavity populations in male mixed chimeras given Cre⁺ or Cre^– bone marrow. Chimerism was normalized to Ly6C^hi blood monocytes before normalization of Cre⁺ to Cre^–. Data represent eight mice per group pooled from two independent experiments. **p < 0.01, ***p < 0.001 (Student's t‐test with Holm–Sidak correction). (G) As in (F), but in female bone marrow chimeric mice. Data represent eight mice per group pooled from two independent experiments. **p < 0.01, ***p < 0.001 (Student's t‐test with Holm–Sidak correction).

Figure 5

RELM‐α promotes survival or differentiation of monocyte‐derived cells in the female peritoneal cavity. (A) Absolute number of Ly6C⁺ MHCII^– and Ly6C⁺ MHCII⁺ monocytes, CD11c^– and CD11c⁺ fractions of Ly6C^– MHCII⁺ CD115⁺ F4/80^lo macrophages, F4/80^hi resident macrophages, and CD11b⁺ DC in the peritoneal cavity of female Retnla ^+/+ and Retnla ^–/– mice. Symbols represent individual animals, with eight (Retnla ^+/+) and six (Retnla ^–/–) mice per group, pooled from three independent experiments. (B) As for (A) but from male Retnla ^+/+ and Retnla ^–/– mice. Symbols represent individual mice, with eight (Retnla ^+/+) or ten (Retnla ^–/–) mice per group pooled from three experiments. (C) Frequency of TIM4^– cells within peritoneal F4/80^hi macrophages from mice in (A). Symbols represent individual mice, with ten (Retnla ^+/+) or eight (Retnla ^–/–) mice per group pooled from four experiments. (D) Frequency of TIM4^– cells within peritoneal F4/80^hi macrophages from male Retnla ^+/+ and Retnla ^–/– mice in (B). (E) Experimental schematic for construction of tissue‐protected single bone marrow‐chimeric mice. (F) Contribution of CD45.1^– CD45.2⁺ bone marrow to the indicated peritoneal populations in female tissue‐protected chimeras given Retnla ^+/+ or Retnla ^–/– bone marrow. Chimerism was normalized to Ly6C^hi blood monocytes before normalization to chimerism in mice receiving Retnla ^+/+ bone marrow. Data represent nine (Retnla ^+/+) or ten (Retnla ^–/–) mice per group pooled from two independent experiments, except for Ly6C^hi monocytes in Retnla ^–/– group, where n = 9 as too few of these cells in one animal. ***p < 0.001 (Student's t‐test with Holm–Sidak correction). (G) As for (F), but in male tissue‐protected bone marrow‐chimeric mice. Data represent eight (Retnla ^+/+) or nine (Retnla ^–/–) mice per group pooled from two independent experiments.