Pirfenidone increases IL-10 and improves acute pancreatitis in multiple clinically relevant murine models

Abstract

Despite decades of research, there is no specific therapy for acute pancreatitis (AP). In the current study, we have evaluated the efficacy of pirfenidone, an antiinflammatory and antifibrotic agent that is approved by the FDA for treatment of idiopathic pulmonary fibrosis (IPF), in ameliorating local and systemic injury in AP. Our results suggest that treatment with pirfenidone in therapeutic settings (e.g., after initiation of injury), even when administered at the peak of injury, reduces severity of local and systemic injury and inflammation in multiple models of AP. In vitro evaluation suggests that pirfenidone decreases cytokine release from acini and macrophages and disrupts acinar-macrophage crosstalk. Therapeutic pirfenidone treatment increases IL-10 secretion from macrophages preceding changes in histology and modulates the immune phenotype of inflammatory cells with decreased levels of inflammatory cytokines. Antibody-mediated IL-10 depletion, use of IL-10-KO mice, and macrophage depletion experiments confirmed the role of IL-10 and macrophages in its mechanism of action, as pirfenidone was unable to reduce severity of AP in these scenarios. Since pirfenidone is FDA approved for IPF, a trial evaluating the efficacy of pirfenidone in patients with moderate to severe AP can be initiated expeditiously.

Keywords: Chemokines; Cytokines; Gastroenterology; Inflammation; Macrophages.

Conflict of interest statement

Conflict of interest: AKS is one of the inventors of Minnelide (patented by the University of Minnesota; patent no. 8507552), which has been licensed to Minneamrita Therapeutics. He has ownership interest in this company and is its Chief Scientific Officer and Cofounder. This relationship is managed by the Office of the Vice Provost for Research at the University of Miami.

Figures

Figure 1. Therapeutic pirfenidone administration reduces local pancreatic injury and lung injury in caerulein 2-day model of acute pancreatitis.

(A) Schematic of therapeutic administration of pirfenidone in a caerulein 2-day model of acute pancreatitis. (B) Representative histology (H&E, 100×) and histological analysis of pancreatic histology from AP-only group and pirfenidone-treated group. Therapeutic pirfenidone treatment leads to a decrease in pancreatic edema, necrosis, and inflammatory infiltrates. Histologic quantification of edema, necrosis, and inflammation is also shown. (C) Serum amylase levels were significantly decreased in the pirfenidone treatment group. (D) IHC for coronin 1A (200×), which stains leukocytes, shows a decrease in immune cell infiltration with pirfenidone treatment. Pancreatic MPO, which is a marker of neutrophilic infiltration, also shows a significant decrease with treatment. (E) Pancreas wet/dry weight ratio, a measure of pancreatic edema, shows a significant decrease in the pirfenidone-treatment group. (F and G) Serum HMGB1 and serum CRP, biomarkers that correlate with severity of AP, were significantly reduced with therapeutic pirfenidone treatment. (H) Lung H&E (100×) shows a reduction in injury with treatment. (I) IHC of lung sections for coronin 1A (200×) shows a decrease in immune infiltration with pirfenidone treatment. Lung MPO (myeloperoxidase) also shows a significant decrease with treatment. (J) Lung wet/dry weight ratio (a measure of pulmonary edema) shows a significant decrease in the treatment group. Pirf., pirfenidone. n = 8 each in AP-only and AP + pirf group. n = 5 each in control groups in F and G. Data represent mean ± SEM. *P < 0.05 by Mann-Whitney U test for B–E, I, and J and Kruskal-Wallis test (Dunn’s multiple-comparison test) for F and G.

Figure 2. Evaluation of the effect of pirfenidone on early events in acute pancreatitis.

Pancreatic acini were isolated from healthy C57BL/6J mice and treated with supramaximal carbachol, with or without pirfenidone pretreatment (0.5 mg/mL for 30 minutes before supramaximal stimulation) in vitro. (A) Treatment of pancreatic acini with supramaximal carbachol (1 mM) leads to trypsin activation. Pirfenidone pretreatment is unable to inhibit supramaximal carbachol–induced trypsin activation (n = 3). (B) Treatment with caerulein in vivo (50 μg/kg, i.p) results in IκBα degradation at 1 hour, as shown by Western blotting in pancreatic tissue. Pretreatment with pirfenidone gavage (400 mg/kg, 30 minutes prior to caerulein injection) does not affect IκBα degradation (n = 3). (C and D) Pretreatment with pirfenidone resulted in decreased secretion of TNF-α (C) and IL-6 (D) from acini treated with supramaximal caerulein (100 nM) for 4 hours (n = 3). (E and F) Pirfenidone disrupts acinar cell–macrophage crosstalk. Pretreatment with pirfenidone resulted in attenuation of the acinar cell mediated activation of macrophages as shown by decreased secretion of TNF-α (E) and IL-6 (F) (n = 3). (G–J) Splenocytes isolated from C57BL/6J mice with L-arginine AP (G and H) (n = 6) or caerulein 2-day model AP (I and J) (n = 3) at the peak of injury were cultured in duplicate, with or without pirfenidone (0.5 mg/mL) for 12 hours, and cells were analyzed using flow cytometry. Pirfenidone treatment led to a significant decrease in TNF-α+Ly6C+ cells (G and I) and MHCII+ Ly6C+ cells (H and J) in the L-arginine model, as well as caerulein 2-day model. Pirf., pirfenidone. Data represent mean ± SEM. *P <.05 by ordinary 1-way ANOVA for (C and D); 2-way ANOVA for (E and F), and Mann-Whitney U test for (G–J).

Figure 3. Therapeutic pirfenidone, when administered at the peak of injury in a well-established L-arginine model of acute pancreatitis, starts reducing local pancreatic and lung injury at 120 hours.

(A) Schematic of pirfenidone administration in a therapeutic manner in well-established L-arginine model of acute pancreatitis. (B) Histological analysis of representative H&E sections (100×) of pancreata from AP-only group and pirfenidone-treated AP group. A decrease in pancreatic edema, necrosis, and inflammatory infiltrate can be seen. Histologic quantification of edema, inflammation, and necrosis is shown. (C) Lung H&E (100×) at 120 hours shows a reduction in injury with pirfenidone treatment. Pirf., pirfenidone. Data represent mean ± SEM and n = 5 in each group. *P < 0.05 by Mann-Whitney U test for each time point (as we are not comparing changes between time points).

Figure 4. Therapeutic pirfenidone modulates immune and cytokine microenvironment at 120 hours, when histological changes start to appear in L-arginine model.

(A–K) mRNA levels of proinflammatory cytokines TNF-α, IL-6, IL-12α, iNOS, IL-23, IFN-γ, TGF-β2, MMP-9, COX-2, and NLRP3 showed a significant decrease with pirfenidone treatment, while the antiinflammatory cytokine IL-10 was significantly increased with pirfenidone treatment. (L–N) Flow cytometry analysis of pancreatic immune cells at 120 hours demonstrates that pirfenidone treatment leads to a significant decrease in percentage of TNF-α secreting macrophages and a significant increase in percentage of IL-10 secreting Th cells and macrophages. (O) Pirfenidone treatment also led to statistically significant increase in Tregs. (P) Pirfenidone treatment significantly increased intrapancreatic IL-10 levels at 120 hours, as measured by ELISA. (Q and R) Multi-Analyte Flow Assay for cytokines TNF-α and IL-17 in serum demonstrates that pirfenidone is able to suppress the level of these cytokines. Data represent mean ± SEM and n = 5 in each group. *P < 0.05 by Kruskal-Wallis Test (Dunn’s multiple-comparison test) for A–K and P and Mann-Whitney U test for L–O, Q, and R.

Figure 5. Therapeutic administration of pirfenidone modulates immune infiltration as well as immune cell phenotype in the well-established L-arginine model of acute pancreatitis at 144 hours.

(A–G) Flow cytometry analysis of pancreatic immune cells at the 144-hour time point shows a reduction in infiltration of neutrophils (A), Gr1+ cells (B), and macrophages (C); pirfenidone also modulated the phenotype of immune cells with decreased intrapancreatic levels of activated macrophages (D) but increased levels of IL-10 secreting macrophages (E), and pirfenidone treatment led to decreased intrapancreatic abundance of IL-17 secreting Th (F) and cytotoxic T cells (G). Pirf., pirfenidone. n = 8–9 each in macrophage/neutrophil panel (n = 8 in AP-only group; n = 9 in AP + pirf group). n = 4–5 each in T cell panel (n = 4 each in AP-only group; n = 5 each in AP + pirf group). Data represent mean ± SEM. *P < 0.05 by Mann-Whitney U test.

Figure 6. Therapeutic administration of pirfenidone increases IL-10 levels before the histology changes in L-arginine model.

(A) Schematic of pirfenidone administration in a therapeutic manner in well-established L-arginine model of acute pancreatitis. (B) Histology of representative H&E sections (50× and 200×) of pancreas from AP-only group and pirfenidone-treated AP group at 110 hours after L-arginine injections is shown. (C) Histologic quantification of edema, necrosis, and inflammation is shown. Pirfenidone treatment does not change pancreatic edema, necrosis, and inflammatory infiltrate at 110 hours. (D) Lung H&E staining (50× and 200×) shows no reduction in injury with pirfenidone treatment at 110-hour time point. (E–L) mRNA levels of various cytokines is shown. mRNA levels of antiinflammatory cytokine IL-10 (E) was significantly increased with pirfenidone treatment at 110 h time point. (F–L) mRNA levels of TNF-α, IL-6, i NOS, IFN-γ, IL-4, Arginase-1, and CD206 did not change at 110 hours. (M) Serum CRP levels do not change with pirfenidone treatment at 110 hours. Pirf., pirfenidone. Data represent mean ± SEM (n = 13 in AP-only group; n = 14 in pirfenidone-treatment group). *P < 0.05 by Mann-Whitney U test for C and E–L and Kruskal-Wallis test (Dunn’s multiple-comparison test) for M.

Figure 7. Therapeutic administration of pirfenidone increases IL-10–secreting macrophages in IL-10 reporter mice preceding changes in histology or reduction in TNF-α+ macrophages at 110 hours in the L-arginine model.

(A) Flow cytometry analysis of pancreatic immune cells at 110-hour time point shows a significant increase in the levels of IL-10–secreting macrophages in IL-10 reporter mice (C57BL/6 background). (B and C) TNF-α+ and MHCII+ M1 macrophages did not show any change with treatment at this time point. (D and E) IL-4+ and CD206+ M2 macrophages showed an increasing trend with treatment, but it was not statistically significant. Pirf., pirfenidone. Data represent mean ± SEM and n = 4 in each group. *P < 0.05 by Mann-Whitney U test.

Figure 8. Therapeutic pirfenidone improves pancreatic and lung injury in the L-arginine model of acute pancreatitis in an IL-10–dependent fashion.

(A) Schematic of pirfenidone administration in a therapeutic manner in L-arginine model of acute pancreatitis with well-established disease, with or without IL-10 neutralizing antibody, is shown. (B) Histology of representative H&E sections (100×) of pancreata from AP-only group, as well as pirfenidone-treated AP group, with or without IL-10 neutralization, is shown. (C) Pirfenidone starts improving L-arginine–induced acute pancreatitis at 120 hours. (D) Pirfenidone is not able to improve L-arginine induced AP when IL-10 is neutralized with antibodies, as evident by unchanged edema, necrosis, and inflammation scores. (E) Lung H&E (100×) shows a reduction in injury with pirfenidone treatment but not in the presence of IL-10 neutralization. Pirf., pirfenidone; IL-10 Antibody, IL-10 neutralizing antibody. Data represent mean ± SEM and n = 5 in each group. *P < 0.05 by Mann-Whitney U test.

Figure 9. Therapeutic pirfenidone does not reduce proinflammatory cytokines in the absence of IL-10 in the L-arginine model.

(A–J) Quantification of the mRNA levels of TNF-α, IL-6, IL-12α, iNOS, IL-23, IFN-γ, TGF-β2, MMP-9, COX-2, and NLRP3 demonstrates that pirfenidone is unable to reduce their levels in the presence of IL-10 neutralization. (K and L) Multi-Analyte Flow Assay for cytokines TNF-α and IL-17 in serum is in agreement with the notion that pirfenidone is not able to reduce their levels in the presence of IL-10 neutralization. Pirf., pirfenidone; IL-10 Ab, IL-10 neutralizing antibody. Data represent mean ± SEM and n =5 in each group. *P < 0.05 by Kruskal-Wallis test (Dunn’s multiple-comparison test).

Figure 10. Pirfenidone treatment of macrophages in M1-polarizing conditions increases IL-10 and other M2 markers.

BMDMs were isolated from healthy C57BL/6J mice and treated with LPS 100 ng/mL (M1 polarizing conditions), with or without pirfenidone treatment (0.5 mg/mL) and incubated for 24 hours. (A–E) Pirfenidone significantly reduces M1 markers TNF-α and IL-6 secretion from macrophages at protein levels (A and B) but not at mRNA levels (C and D); iNOS (M1) (E) is also not affected at mRNA levels with pirfenidone. (F–H) Pirfenidone significantly increases M2 markers, viz IL-10 (F), IL-4 (G), and Arginase-1 (H) in M1 polarizing conditions. Pirf., pirfenidone. For A and B, n =6 in each group. For C–H, n = 6 in each group treated with LPS, and n = 4 in macrophage-only and macrophage + pirf. groups. Data represent mean ± SEM *P < 0.05 by ordinary 1-way ANOVA for A and B and Kruskal-Wallis test (Dunn’s multiple-comparison test) for C–H.

Figure 11. Depletion of macrophages with clodronate liposomes abrogates the beneficial effects of therapeutic pirfenidone in L-arginine AP.

(A) Schematic of pirfenidone administration in a therapeutic manner in well-established L-arginine model of acute pancreatitis with clodronate liposomes (i.p., 200 μL) administration at 96 and 120 hours. (B and C) Confirmation of macrophage depletion using clodronate in pancreas (B) and spleen (C). (D) Histological analysis of representative H&E sections (200×) of pancreas from AP-only group and pirfenidone-treated AP group. There is no decrease in pancreatic edema, necrosis, and inflammatory infiltrate with pirfenidone treatment with depletion of macrophages. (E) Lung H&E (200×) does not show a reduction in injury with pirfenidone treatment with depletion of macrophages. (F–I) Histologic quantification of edema, inflammation, and necrosis in pancreas is shown, which concurs with our findings. (J–O) mRNA levels of proinflammatory markers TNF-α, IL-6, IFN-γ, IL-23, IL-12α,and NLRP3 did not show a significant change with pirfenidone treatment with depletion of macrophages. (P) Serum HMGB1 levels are not significantly different at 144 hours with macrophage depletion. Pirf., pirfenidone; Clod., liposomal clodronate. n =7 in each group. Data represent mean ± SEM. *P < 0.05 by Mann-Whitney U test.