Arginase-1-expressing macrophages suppress Th2 cytokine-driven inflammation and fibrosis

Abstract

Macrophage-specific expression of Arginase-1 is commonly believed to promote inflammation, fibrosis, and wound healing by enhancing L-proline, polyamine, and Th2 cytokine production. Here, however, we show that macrophage-specific Arg1 functions as an inhibitor of inflammation and fibrosis following infection with the Th2-inducing pathogen Schistosoma mansoni. Although susceptibility to infection was not affected by the conditional deletion of Arg1 in macrophages, Arg1(-/flox);LysMcre mice died at an accelerated rate. The mortality was not due to acute Th1/NOS2-mediated hepatotoxicity or endotoxemia. Instead, granulomatous inflammation, liver fibrosis, and portal hypertension increased in infected Arg1(-/flox);LysMcre mice. Similar findings were obtained with Arg1(flox/flox);Tie2cre mice, which delete Arg1 in all macrophage populations. Production of Th2 cytokines increased in the infected Arg1(-/flox);LysMcre mice, and unlike alternatively activated wild-type macrophages, Arg1(-/flox);LysMcre macrophages failed to inhibit T cell proliferation in vitro, providing an underlying mechanism for the exacerbated Th2 pathology. The suppressive activity of Arg1-expressing macrophages was independent of IL-10 and TGF-beta1. However, when exogenous L-arginine was provided, T cell proliferation was restored, suggesting that Arg1-expressing macrophages deplete arginine, which is required to sustain CD4(+) T cell responses. These data identify Arg1 as the essential suppressive mediator of alternatively activated macrophages (AAM) and demonstrate that Arg1-expressing macrophages function as suppressors rather than inducers of Th2-dependent inflammation and fibrosis.

Figure 1. Arg1 ^−/flox ;LysMcre mice exhibit reduced Arg1 activity in macrophages.

(A) Macrophages were isolated from control (filled bars) and Arg1 ^−/flox ;LysMcre (open bars) mice and analyzed for arginase activity by measuring urea. Macrophages were isolated and assayed with either medium alone or treated with a combination of IL-4, IL-13, and IL-21 overnight. (B) Macrophages were assayed for NO by Griess reaction in the presence of medium, IL-4, IL-13, and IL-21, or IFN-γ and LPS (Mean±SEM). (C) RNA from thioglycollate elicited macrophages stimulated with IL-4, IL-13, and IL-21 was isolated, purified, and assessed for the expression of Mrc1 (mannose receptor) and Chi3l3 (Ym1) (Mean±SEM). All assays were repeated with similar results and performed in triplicate.

Figure 2. Arg1 expression is required to control fibrosis.

Control Arg1 ^flox/flox (filled bars) and Arg1 ^−/flox ;LysMcre (open bars) mice were exposed to 35 S. mansoni cercariae percutaneously. Mice (n = control/Arg1 ^−/flox ;LysMcre ) were sacrificed at weeks 9 (n = 14/14), 12 (n = 21/10), and 22 (n = 12/10) post-infection and analyzed for the development of fibrosis and portal hypertension. (A) Collagen was assessed by measuring hydroxyproline and normalizing to infectious worm pairs per mouse (Mean±SEM). (B) Total animal weight was compared with the weight of total excised liver to determine liver as a percent of body weight (Mean±SEM). (C) S. mansoni eggs within the lungs of control and Arg1 ^−/flox ;LysMcre mice as an measure of collateral vessel development and portal hypertension; eggs were enumerated by digesting lungs in 4% KOH at 37°C for 12 hours and 1ml of the suspension was counted in a Sedgwick-Rafter chamber (Mean±SEM). (D) Representative gross pathology of 22-week infected control and Arg1 ^−/flox ;LysMcre. Upper arrow indicates presence of S. mansoni eggs which were shunted into the lungs of Arg1 ^−/flox ;LysMcre mice. (E) Individual liver sections from Control Arg1 ^flox/flox (top) and Arg1 ^−/flox ;LysMcre (bottom) mice were analyzed for collagen via second harmonic emission (red). Representative granulomas from 9, 12, and 22 weeks post-infection with S. mansoni. All images were taken at 20× magnification. Mean fluorescence intensities for individual control (filled bars) and Arg1 ^−/flox ;LysMcre (open bars) mice at week 9 (n = 14/14), 12 (n = 21/10), and 22 (n = 12/9) weeks post-infection with S. mansoni. All assays were repeated three times with similar results.

Figure 3. Macrophage-associated Arg1 inhibits granulomatous inflammation.

Week 9 (n = 14/14), 12 (n = 21/10), and 22 (n = 12/9) S. mansoni–infected control and Arg1 ^−/flox ;LysMcre mice were individually assessed for granuloma volume (Mean±SEM) (A), and individual populations of small/large lymphocytes, macrophages, fibroblasts, and eosinophils were enumerated (B). Individual granulomas from control (filled bars) and Arg1 ^−/flox ;LysMcre mice (open bars) were scored for granuloma-associated necrosis on a scale of 1–4 at weeks 9 (n = 14/14), 12 (n = 21/10) and 22 (n = 12/9) post-infection with S. mansoni (mean±SEM) (C). Representative granulomas from 12-week S. mansoni–infected control and Arg1 ^−/flox ;LysMcre mice were stained with Picrosirius Red 20× or DAPI (blue) and F4/80⁺ (green) and photographed at 40×. Arrows point to macrophage-rich regions. Sm = S. mansoni egg (D). All assays were repeated three times with similar results.

Figure 4. Arg1^flox/flox;Tie2cre mice develop exacerbated inflammation and fibrosis.

(A) BALB/c control (n = 5), BALB/c Arg1 ^flox/flox ;Tie2cre (n = 5), C57BL/6 control (n = 5), and C57BL/6 Arg1 ^flox/flox ;Tie2cre (n = 7) were infected with S. mansoni cercariae and euthanized on week 9 and assessed individually for granuloma volume and fibrosis. Additional control (n = 6) and C57BL/6 Arg1 ^flox/flox ;Tie2cre (n = 5) were examined on week 12. The data show the means±SEM; significant differences are noted in the figure. (B) Granuloma volumes were measured in the intestines of control and C57BL/6 Arg1 ^flox/flox ;Tie2cre on week 12 (Mean±SEM). A second experiment produced similar results.

Figure 5. Arg1-expressing AAMs regulate survival after S. mansoni infection.

Control Arg1 ^flox/flox and Arg1 ^−/flox ;LysMcre mice were infected with S. mansoni for 12 or 22 weeks, and survival was monitored. (A) Survival rates for week 12 (n = 21/20) and 22 (n = 12/15) control and Arg1 ^−/flox ;LysMcre mice infected with 35 cercariae or week 22 mice (n = 5/5) infected with 100 cercariae. (B) Survival of control (n = 8) and Arg1 ^−/flox ;LysMcre (n = 9) mice infected with 35 cercariae in the presence of the NO inhibitor aminoguanidine (2.5% in drinking water). Individual serum samples were isolated from control and Arg1 ^−/flox ;LysMcre mice infected with S. mansoni for 9, 12, and 22 weeks. Samples were analyzed for liver enzymes and LPS. (C) AST/GOT. (D) Alkaline Phosphatase. (E) ALT/GPT. (F) Endotoxin. All assays were repeated three times with similar results.

Figure 6. Arg1-expressing AAMs suppress Th2 cytokine production.

Liver leukocytes isolated from 9-week S. mansoni–infected mice were separated, counted, and cultured with PMA, Ionomycin, and Brefeldin A for 3 hours and analyzed for ex-vivo cytokine production capability. (A) Representative FACS plots of IL-13, IL-4, IL-5, IFN-γ, and IL-17-producing CD4⁺ T cells. (B) Total number of cytokine-producing CD4⁺ cells in the liver leukocyte preparation (n = 7/group). (C) Magnitude of cytokine production by CD4⁺ cells displayed as geometric mean of fluorescence intensity. Statistically significant differences are shown (n = 7/group). (D) 9-week S. mansoni–infected mice were treated with BrdU for 3 days by i.p. injection prior to sacrifice, and then mesenteric lymph nodes cells were isolated and stained for CD4⁺ and BrdU incorporation (n = 3/group). All experiments were repeated twice with similar results.

Figure 7. Arg1-expressing AAMs suppress T cell proliferation.

Day-4 thioglycollate-elicited peritoneal macrophages from WT or Arg1 ^−/flox ;LysMcre mice were treated with 1 ng/ml of cytokine cocktail (IL4+IL13+GM-CSF) for 36 hours before addition of CFSE labeled LN cells from OT-II mice and cognate OVA peptide. Proliferation of CD4⁺ cells was examined at 96 hours of co-culture in RPMI+10% FCS. The experiment was repeated twice with similar results.

Figure 8. The suppressive activity of AAMs is IL-10 and TGF-β independent.

(A) Day 4 peritoneal macrophages from thioglycollate-injected WT mice were treated with increasing doses of a cytokine cocktail of IL-4, IL-13, and GM-CSF for 36 hours before addition of CFSE-labeled LN cells from OT-II mice and cognate OVA peptide. Proliferation of CD4⁺ cells was examined at 96 hours of co-culture in RPMI+10% FCS. Some groups were also treated with anti-TGF-β (10 µg/ml), anti-IL-10 (10 µg/ml), or a combination of both mAbs when T cells were added to the macrophage APCs. (B) In a second series of experiments, the macrophage/T cell cultures were supplemented with L-arginine hydrochloride at the indicated concentrations, again at the time when T cells were added. All experiments were repeated three times with similar results.

Figure 9. Arg1 is required to suppress schistosome-specific T cell proliferation.

2×10⁵ thioglycollate-elicited macrophages from (WT) control Arg1^flox/flox (A,C) and (KO) Arg1 ^−/flox ;LysMcre mice (B,D) were pre-treated with a cocktail of IL-4/IL-13/GM-CSF at a concentration of 1 ng/ml of each cytokine or PBS. CD4⁺ cells were isolated from the livers of 9-week S. mansoni–infected WT (A,B) or Arg1 ^−/flox ;LysMcre mice (C,D), labeled with CFSE, and then added to the macrophage cultures at a concentration of 1×10⁵ cells per well in RPMI+10% FCS. Wells were either left unstimulated (Med) or stimulated with 20 µg/ml of SEA. After 84 hours, proliferation was analyzed by flow cytometry and represented as the percentage of CD4 cells that had divided in each of the conditions. The experiment was conducted twice with similar results.