Sex-Biased Control of Inflammation and Metabolism by a Mitochondrial Nod-Like Receptor

Abstract

Mitochondria regulate steroid hormone synthesis, and in turn sex hormones regulate mitochondrial function for maintaining cellular homeostasis and controlling inflammation. This crosstalk can explain sex differences observed in several pathologies such as in metabolic or inflammatory disorders. Nod-like receptor X1 (NLRX1) is a mitochondria-associated innate receptor that could modulate metabolic functions and attenuates inflammatory responses. Here, we showed that in an infectious model with the human protozoan parasite, Leishmania guyanensis, NLRX1 attenuated inflammation in females but not in male mice. Analysis of infected female and male bone marrow derived macrophages showed both sex- and genotype-specific differences in both inflammatory and metabolic profiles with increased type I interferon production, mitochondrial respiration, and glycolytic rate in Nlrx1-deficient female BMDMs in comparison to wild-type cells, while no differences were observed between males. Transcriptomics of female and male BMDMs revealed an altered steroid hormone signaling in Nlrx1-deficient cells, and a "masculinization" of Nlrx1-deficient female BMDMs. Thus, our findings suggest that NLRX1 prevents uncontrolled inflammation and metabolism in females and therefore may contribute to the sex differences observed in infectious and inflammatory diseases.

Keywords: inflammation; innate immunity; metabolism; nod-like receptor X1; sex.

Figure 1

NLRX1 attenuated inflammation and tissue damage in infected female mice. Wild-type (WT) and Nlrx1-deficient (Nlrx1^-/- ) C57BL/6 female mice (n=5 mice per group) were infected in both hind footpads with 3x10⁶ stationary phase LgyLRV1+ parasites containing a luciferase gene. (A) Footpad swelling was measured weekly as a proxy of disease progression. At the peak of infection [4 weeks post-infection (p.i.)] (B) in vivo inflammation and (C) parasite burden was visualized and quantified by bioluminescence imaging after luminol and luciferin injection, respectively. Graphs are presented as mean +- SEM and are representative of three independent experiments. (D) Representative images of hematoxylin and eosin (H&E) staining of footpad sections. Upper panel show normal histological appearance of the epidermis and dermis of female WT and Nlrx1^-/- mouse footpads. Bottom panel: lesions from hind footpads of LgyLRV1+ infected female mice were dissected at 4 weeks p.i. and cell recruitment to lesion site was visualized. Magnification: 40x, scale bar: 50 μm. Sq, squames. E, epidermis. D, dermis. M, muscle. (n=5 mice per group). Relative mRNA levels of pro-inflammatory genes (E) Ifnb, (F) Il6 and (G) Tnfa were quantified in the lesions of infected WT and Nlrx1^-/- female mice at 4 weeks p.i. using RT-qPCR. Graphs are presented as mean +- SEM (n=8 mice per group). Statistical significance was assessed by two-way ANOVA with multiple comparisons (A) or unpaired, parametric t-test (B-C, E-G). ns = non-significant, *p ≤ 0.05 **p ≤ 0.01, ***p ≤ 0.001.

Figure 2

Inflammatory profile and transcriptomics analysis of female BMDMs. Bone marrow derived macrophages (BMDMs) from WT and Nlrx1^-/- female mice were isolated and infected with stationary phase LgyLRV1+ parasites or stimulated with the TLR3 agonist poly I:C (2 μg/ml). Nlrx1 mRNA levels were quantified by qRT-PCR (A) at 8 hours p.i. (n=3 independent experiments). (B–D) After 24 hours, supernatants were collected and IFNβ, IL-6 and TNFα secretion was quantified by ELISA in LgyLRV1+ or poly I:C stimulated BMDMs. (n=3 independent experiments). (E) At 24 hours p.i., BMDMs were fixed with 4% PFA and stained with DAPI and phalloidin. Cells were visualized with a high content microscope (40x) and intracellular parasite load was quantified using a MetaXpress software (n=2 independent experiments). (F) Transcriptomics analysis of WT and Nlrx1^-/- female BMDMs (n=3 mice per group) infected with LgyLRV1+ parasites or stimulated with poly I:C (2 μg/ml) for 24 hours. The heatmap represents the global weighted correlation network analysis (WGCNA) and module names are represented by a color. A gene ontology (GO) enrichment analysis for each module was performed to identify the biological processes associated to each module. represents modules enriched in GO terms associated with sex hormone signaling. Graphs are presented as mean +- SEM and significance was tested by two-way ANOVA with multiple comparisons (A-F). ns = non-significant, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001.

Figure 3

In male mice NLRX1 did not regulate inflammation or tissue damage. Wild-type (WT) and Nlrx1-deficient (Nlrx1^-/- ) C57BL/6 male mice (n=4 mice per group) were infected in both hind footpads with 3x10⁶ stationary phase LgyLRV1+ parasites (A) Footpad swelling was measured weekly as a proxy of disease progression. At the peak of infection (4 weeks p.i.), (B) in vivo inflammation and (C) parasite burden was visualized and quantified by bioluminescence imaging after luminol and luciferin injection, respectively. Graphs are presented as mean +- SEM and are representative of three independent experiments. (D) Representative images of hematoxylin and eosin (H&E) staining of footpad sections. Upper panel show normal histological appearance of the epidermis and dermis of male mouse footpads. Bottom panel: lesions from hind footpads of LgyLRV1+ infected male mice were dissected at 4 weeks p.i. and cell recruitment to lesion site was visualized. Magnification: 40x, scale bar: 50 μm. Sq, squames. E, epidermis. D, dermis. M, muscle. (n=5 mice per group). Relative mRNA levels of pro-inflammatory genes (E) Ifnb, (F) Il6 and (G) Tnfa were quantified in the lesions of infected WT and Nlrx1^-/- male mice at 4 weeks p.i. using RT-qPCR. Graphs are presented as mean +- SEM. (n=9-10 mice per group). Statistical significance was assessed by two-way ANOVA with multiple comparisons (A) or unpaired, parametric t-test (B, C, E–G). ns = non-significant, *p ≤ 0.05.

Figure 4

Sex bias in inflammation and infection in absence of NLRX1. BMDMs from female and male WT and Nlrx1^-/- mice were isolated simultaneously and infected with LgyLRV1+ parasites or stimulated with TLR3 agonist poly I:C (2 μg/ml). After 8 hours (A) and 24 hours p.i. (B), Nlrx1 mRNA levels were quantified by qRT-PCR (n=3 independent experiments). (C–E) After 24 hours, proinflammatory cytokines IFNβ, IL-6 and TNFα were quantified in cell-free supernatants by ELISA in LgyLRV1+ infected or poly I:C stimulated BMDMs. (n=3-4 independent experiments). (F) At 24 hours p.i., BMDMs were fixed with 4% PFA and stained with DAPI and phalloidin. Cells were visualized with a high content microscope (40x) and intracellular parasite load was quantified using a MetaXpress software. (n=3 independent experiments). Graphs are presented as mean +- SEM and significance was tested by two-way ANOVA with multiple comparisons (A–F). ns = non-significant, *p ≤ 0.05, ***p ≤ 0.001.

Figure 5

In absence of NLRX1 female BMDMs had a male-like metabolic response. BMDMs from female and male WT and Nlrx1^-/- mice were isolated simultaneously and infected with LgyLRV1+ parasites for 8 hours. After 8 hours, (A) basal mitochondrial respiration and (B) basal glycolytic rate were assessed by Seahorse XFe96 analyzer and adjusted to protein concentration per well (n=3-4 independent experiments). (C) Mitochondrial ROS (mtROS) production was quantified in female and male BMDMs infected with LgyLRV1+ parasites or treated with poly I:C (2 μg/ml) for 8 hours. Cells were stained with MitoSOX Red (5 μM) for 20 min at 37°C. Fluorescence was measured using a Spectramax i3 plate reader and adjusted to protein concentration per well. (n=3 independent experiments). (D) Mitochondria structure and (E) the percentage of mitochondria volume per cell of female and male BMDMs infected with LgyLRV1+ parasites for 8 hours were analyzed by transmission electron microscopy. Representative images are shown. Red arrows show examples of normal mitochondrial structure. LgyLRV1+ parasites are contoured in red (n=2 independent experiments, total of minimum 60 cells analyzed per group). Magnification: 4800x. Scale bar: 1 μm. (F) BMDMs from female and male WT and Nlrx1^-/- mice were pre-treated with 17β-estradiol (200 pg/ml) for 2 hours and estradiol was kept in the assay medium for the duration of the assay. Basal mitochondrial respiration was assessed by Seahorse XFe96 analyzer and adjusted to protein concentration per well (n=4 independent experiments). Graphs are presented as mean +- SEM and significance was assessed by two-way ANOVA with multiple comparisons (A–C) or unpaired, parametric t-test (E, F). ns = non-significant, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001.

Figure 6

Transcriptomics analysis of female and male BMDMs revealed a male-like phenotype of Nlrx1^-/- female BMDMs. (A) Transcriptomics analysis of both female and male WT and Nlrx1^-/- BMDMs (n=3 mice per group) infected with LgyLRV1+ and LgyLRV1- parasites or stimulated with poly I:C (2 μg/ml) for 24 hours. The heatmap represents the global weighted correlation network analysis (WGCNA) and module names are represented by a color. A gene ontology (GO) enrichment analysis for each module was performed to identify the biological processes associated to each module. represents modules enriched in GO terms associated with sex hormone signaling. At 24 hours, the number of differentially expressed genes between groups in (B) non-infected, (C) LgyLRV1+ infected and (D) poly I:C treated conditions is plotted. On each barplot the number represents the total number of differentially upregulated (in red) and downregulated (in blue) genes. Pairwise comparisons are done by “sex” (same genotype, different sex, “WT.F._WT.M” and “Nlrx1.F_Nlrx1.M”), by “genotype” (same sex, different genotype (“Nlrx1.F_WT.F” and “Nlrx1.M_WT.M”), or by combining both “sex and genotype” (“Nlrx1.M_WT.F” and “Nlrx1.F_WT.M”). See Supplementary Tables 9–11 for the list of significantly differentially expressed genes. (E) Identification of WT female specific genes. Genes that were differentially expressed in only in WT female BMDMs but not between the other three groups (Nlrx1-/- female, Nlrx1-/- male and WT male BMDMs) were identified in non-infected samples at 8 hours.