M2 Macrophages promote IL-33 expression, ILC2 expansion and mucous metaplasia in response to early life rhinovirus infections

Abstract

Wheezing-associated rhinovirus (RV) infections are associated with asthma development. We have shown that infection of immature mice with RV induces type 2 cytokine production and mucous metaplasia which is dependent on IL-33 and type 2 innate lymphoid cells (ILC2s) and intensified by a second heterologous RV infection. We hypothesize that M2a macrophages are required for the exaggerated inflammation and mucous metaplasia in response to heterologous RV infection. Wild-type C57Bl/6J mice and LysM^Cre IL4Rα KO mice lacking M2a macrophages were treated as follows: (1) sham infection on day 6 of life plus sham on day 13 of life, (2) RV-A1B on day 6 plus sham on day 13, (3) sham on day 6 and RV-A2 on day 13, or (4) RV-A1B on day 6 and RV-A2 on day 13. Lungs were harvested one or seven days after the second infection. Wild-type mice infected with RV-A1B at day 6 showed an increased number of Arg1- and Retnla-expressing lung macrophages, indicative of M2a polarization. Compared to wild-type mice infected with RV on day 6 and 13 of life, the lungs of LysM^Cre IL4Rα KO mice undergoing heterologous RV infection showed decreased protein abundance of the epithelial-derived innate cytokines IL-33, IL-25 and TSLP, decreased ILC2s, decreased mRNA expression of IL-13 and IL-5, and decreased PAS staining. Finally, mRNA analysis and immunofluorescence microscopy of double-infected LysM^Cre IL4Rα KO mice showed reduced airway epithelial cell IL-33 expression, and treatment with IL-33 restored the exaggerated muco-inflammatory phenotype.

Conclusion: Early-life RV infection alters the macrophage response to subsequent heterologous infection, permitting enhanced IL-33 expression, ILC2 expansion and intensified airway inflammation and mucous metaplasia.

Keywords: IL-33; ILC2; M2 macrophage; Rhinovirus; neonate.

Figure 1

Differential expansion of F4/80+CD11c+arginase-1+ M2a-polarized macrophages in wild-type immature mice following heterologous RV infection. Six-day old wild-type mice were inoculated with sham or RV-A1B on day 6 of life and sham or RV-A2 on day 13 of life. (A) On day 14, lungs were harvested, digested with liberase TM, collagenase XI, hyaluronidase 1a, and DNase I, and stained with anti-F4/80, anti-CD11c, anti-arginase-1, and PacBlue (for dead cells). Cells were washed, fixed, and processed for flow cytometry. The panel shows flow cytometry analysis of live F4/80+ CD11c+ arginase-1+ macrophages from the four groups. FMO controls for F4/80, CD11c and arginase are also shown. Data shown are mean ± SD; n=6 per group from two different experiments; *different from sham + sham, P < 0.05 by one-way ANOVA; †different from RV-A1B + sham, P < 0.05 by one-way ANOVA. (B) Lungs were collected on day 14 of life (1 day post RV-A2 or sham infection) from wild-type immature mice inoculated with sham or RV-A1B on day 6 of life and sham or RV-A2 on day 13 of life. Cell suspensions were sorted for CD45+ F4/80+ macrophages. The cell pellet was collected for mRNA expression by quantitative PCR (N=6/group, lungs from two mice were combined for each measurement). Data shown are mean ± SD; *different from sham+sham, P < 0.05 by one-way ANOVA; †different from RV-A1B+sham, P < 0.05 by one-way ANOVA.

Figure 2

LysM^Cre IL-4Rα KO attenuates heterologous RV infection-induced type 2 cytokine expression. Six-day old immature wild-type and LysM^Cre IL-4Rα KO mice were inoculated with sham or RV-A1B on day 6 and sham or RV-A2 on day 13 of life. (A) mRNA and protein expression of the type 2 cytokines IL-5 and IL-13. Lungs were harvested for qPCR or ELISA seven days after secondary sham or RV-A2 infection. (N =6 from two different experiments, mean ± SD; *different from wildtype sham + RV-A1B, P < 0.05 by one-way ANOVA; †different from wild-type RV-A1B + RV-A2, P < 0.05 by one-way ANOVA.) (B) mRNA and protein expression of the pro-inflammatory cytokines IL-1β, TNF-α, CXCL-1, and IFN-γ. Lungs were harvested for qPCR or ELISA one day after secondary sham or RV-A2 infection. (N =6 from two different experiments, mean ± SD; *different from wild-type sham + RV-A2, P < 0.05 by one-way ANOVA; †different from wild-type RV-A1B + RV-A2, P < 0.05 by one-way ANOVA. (C) Whole lung RV positive-strand RNA one day after secondary infection with sham or RV-A2. (N=6 from two different experiments, mean ± SD; different from sham + RV-A2, P < 0.05 by one-way ANOVA.

Figure 3

LysM^Cre IL-4Rα KO attenuated heterologous RV infection-enhanced innate cytokine expression. (A, B) Six-day old immature wild-type and LysM^Cre IL-4Rα KO mice were inoculated with sham or RV-A1B on day 6 and sham or RV-A2 on day 13 of life. Lungs were harvested for qPCR (A) or ELISA (B) one day post sham or RV-A2 secondary infection for IL-33 and seven days after sham or RV-A2 infection for IL-25 and TSLP expression. (N=6-10 from two different experiments, mean ± SD; *different from RV-A1B + sham (IL-25 and TSLP) or wild-type sham + RV-A2 (IL-33), P < 0.05 by one-way ANOVA; †different from wild-type RV-A1B+RV-A2, P < 0.05 by one-way ANOVA.

Figure 4

LysM^Cre IL-4Rα KO blocks lung ILC2 expansion in 6-day old mice following heterologous RV infection. Wild-type and LysM^Cre IL-4Rα KO mice were inoculated with sham or RV-A1B on day 6 of life and sham or RV-A2 on day 13 of life. On day 20, lungs were harvested, digested with liberase TM, collagenase XI, hyaluronidase 1a, and DNase I, and stained with lineage antibody cocktail, anti-CD127, anti-ST2 and Pacific blue (for dead cells). Cells were washed, fixed, and processed for flow cytometry. (A) Figure showing flow cytometry analysis of live lineage-negative, CD127+ ST2+ ILC2s from sham + sham, RV-A1B + sham, sham + RV-A2 and RV-A1B + RV-A2 groups. (B) Graph showing group mean data for ILC2s as a percentage of lineage-negative cells (left panel), as a percentage of live lung cells (middle panel) and as a total ILC2s number per lung (right panel) (N=6 from two different experiments, mean± SD; *different from wild-type RV-A1B+sham, P < 0.05 by one-way ANOVA; †different from wild-type RV-A1B + sham or RV-A1B + RV-A2, P < 0.05 by one-way ANOVA.

Figure 5

LysM^Cre IL-4Rα KO attenuated heterologous RV infection-exaggerated mucus metaplasia and type 2 immune responses. Six-day old immature wild-type and LysM^Cre IL-4Rα KO mice were inoculated with sham or RV-A1B on day 6 and sham or RV-A2 on day 13 of life. Lungs were harvested on seven days after the secondary infection (sham or RV-A2) and processed for qPCR and histology. Lungs sections were stained for PAS and quantified using NIH ImageJ. (A) mRNA expression of Muc5ac, and Gob5 (Clca1) in sham + sham, sham + RV-A1B, sham + RV-A2, and RV-A1B + RV-A2-infected mice. (N =6 from two different experiments, mean± SD; *different from wild-type RV-A1B + sham, P < 0.05 by one-way ANOVA; †different from wild-type RV-A1B + sham or RV-A1B + RV-A2, P < 0.05 by one-way ANOVA. (B) PAS staining in sham, RV-A1B, RV-A2, RV-A1B+RV-A2-infected wild-type and LysM^Cre IL-4Rα KO mice. The black bar is 50 microns (μ). (C) Quantification of PAS staining in the airways. Data are represented are PAS-positive cells per micron of basement membrane length. Data shown are mean ± SD; n=2-3 airways/mouse, 6 mice per group from two different experiments; *different from wild-type RV-A1B + sham, P < 0.05 by one-way ANOVA; †different from wild-type RV-A1B + RV-A2, P < 0.05 by one-way ANOVA.

Figure 6

Recombinant IL-33 restores type 2 immune responses and mucus metaplasia in LysM^Cre IL-4Rα KO mice with heterologous RV infection. Six-day old immature LysM^Cre IL-4Rα KO mice were inoculated with either sham on day 6 and 13 or RV-A1B on day 6 and RV-A2 on day 13 of life. Selected LysM^Cre IL-4Rα KO mice were also treated with 0.1 µg of mouse recombinant IL-33 intranasally on day 13 of life. Lungs were harvested on day 20 (7 days post secondary sham or RV-A2 infection) and processed for qPCR (A) and histology (C). (A) mRNA expression of IL-5, IL-13, IL-25, Muc5ac, and Gob5 (Clca1); N =6 from two different experiments, mean± SD; *different from LysM^Cre IL-4Rα KO sham + sham, P < 0.05 by one-way ANOVA; †different from LysM^Cre IL-4Rα KO RV-A1B+RV-A2, P < 0.05 by one-way ANOVA. (B) PAS staining in LysM^Cre IL-4Rα KO mice. The black bar is 50 microns. (C) Quantification of PAS staining in the airways. Data are represented are PAS-positive cells per micron of basement membrane length. Data shown are mean ± SD; n=2-3 airways/mouse, 6 mice per group from two different experiments; *different from LysM^Cre IL-4Rα KO sham+sham, P < 0.05 by one-way ANOVA; †different from LysM^Cre IL-4Rα KO RV-A1B+RV-A2, P < 0.05 by one-way ANOVA.

Figure 7

LysM^Cre IL-4Rα KO attenuated epithelial IL-33 expression after heterologous RV infection. Six-day old immature wild-type and LysM^Cre IL-4Rα KO mice were inoculated with sham or RV-A1B on day 6 and sham or RV-A2 on day 13 of life. Lungs were harvested on day 14 (1 day post RV-A2 infection) and digested with liberase TM, collagenase XI, hyaluronidase 1a, and DNase I, and stained with anti-F4/80, anti-CD45, anti-EpCAM, and DAPI (for dead cells). (A) Cell suspensions were sorted for CD45+F4/80+ macrophages, CD45-EpCAM+ epithelial cells, and CD45-EpCAM- cells. (B) The cell pellet was collected for IL-33 or IL-25 mRNA expression by quantitative PCR. IL-33 and IL-25 mRNA expression of CD45+F4/80+ macrophages, CD45-EpCAM+ epithelial cells, and CD45-EpCAM- cells in wild-type mice. (N=6/group, lungs from two mice were combined for each measurement). Data shown are mean ± SD. (C) Epithelial derived IL-33 mRNA expression in CD45- EpCAM+ epithelial cells in wild-type and LysM^Cre IL-4Rα KO mice. Data shown are mean ± SD, *different from wild-type RV-A1B + sham, P < 0.05 by one-way ANOVA; †different from wild-type RV-A1B + + RV-A2, P < 0.05 by one-way ANOVA. (D) Lung IL-33 in wild-type and LysM^Cre IL-4Rα KO mice. .