BET Inhibition Rescues Acinar-Ductal-Metaplasia and Ciliogenesis and Ameliorates Chronic Pancreatitis-Driven Changes in Mice With Loss of the Polarity Protein Par3

Abstract

Background & aims: The apical-basal polarity of pancreatic acinar cells is essential for maintaining tissue architecture. However, the mechanisms by which polarity proteins regulate acinar pancreas injury and regeneration are poorly understood.

Methods: Cerulein-induced pancreatitis was induced in mice with conditional deletion of the polarity protein Par3 in the pancreas. The impact of Par3 loss on pancreas injury and regeneration was assessed by histologic analyses and transcriptional profiling by RNA sequencing. Mice were pretreated with the bromodomain and extraterminal domain (BET) inhibitor JQ1 before cotreatment with cerulein to determine the effect of BET inhibition on pancreas injury and regeneration.

Results: Initially, we show that Par3 is increased in acinar-ductal metaplasia (ADM) lesions present in human and mouse chronic pancreatitis specimens. Although Par3 loss disrupts tight junctions, Par3 is dispensable for pancreatogenesis. However, with aging, Par3 loss results in low-grade inflammation, acinar degeneration, and pancreatic lipomatosis. Par3 loss exacerbates acute pancreatitis-induced injury and chronic pancreatitis-induced acinar cell loss, promotes pancreatic lipomatosis, and prevents regeneration. Par3 loss also results in suppression of chronic pancreatitis-induced ADM and primary ciliogenesis. Notably, targeting BET proteins attenuates chronic pancreatitis-induced loss of primary cilia and promotes ADM in mice lacking pancreatic Par3. Targeting BET proteins also attenuates cerulein-induced acinar cell loss and enhances recovery of acinar cell mass and body weight of mice lacking pancreatic Par3.

Conclusions: Combined, this study demonstrates how Par3 restrains chronic pancreatitis-induced changes in the pancreas and identifies a potential role for BET inhibitors to attenuate pancreas injury and facilitate regeneration.

Keywords: ADM; BET Inhibitors; Inflammation; Par3; Primary Cilia.

Figure 1

Increased Par3 expression in human and mouse chronic pancreatitis specimens. (A) Human pancreatitis tissue array was stained for Par3 by immunohistochemistry. Par3 expression in the mild/acute pancreatitis samples (n = 11) was compared with Par3 expression in the chronic pancreatitis samples (n = 15). t test, mean ± standard deviation (SD); ∗∗P ≤ .01. (B) In chronic pancreatitis specimens (n = 15), the relative expression of Par3 in ADM was compared with Par3 expression in the adjacent normal pancreas. t test, mean ± SD; ∗∗P ≤ .01. (C and D) Wild-type B6 mice were treated with 8 hourly injections of cerulein to induce acute pancreatitis (n = 4, 5) or treated with cerulein twice daily for 5 days to induce chronic pancreatitis (n = 4, 4), and the pancreas immunofluorescence stained for Par3 and DAPI. t test, mean ± SD; ∗∗P ≤ .01. (E) The chronic pancreatitis mouse samples (n = 3) were costained with Par3, CK19, amylase, and DAPI to evaluate Par3 expression in mouse ADM lesions. Par3 expression in the ADM was compared with Par3 expression in the adjacent normal pancreas. In each mouse, at least 10 adjacent regions and ADM lesions were analyzed for Par3 expression. t test, mean ± SD; ∗∗P ≤ .01, ∗∗∗∗P ≤ .0001. Scale bars in A–E = 100 mm. IHC, immunohistochemistry; ns, not significant; Ceru, cerulein.

Figure 2

Mice with pancreatic Par3 loss exhibit low-grade inflammation and acinar cell loss with aging. (A) Alleles of Pdx1-Cre (C) and Par3fl mice. Par3 expression in the pancreas of mice of indicated genotypes was determined by Western blotting. (B) Immunofluorescence stains for Par3, Zo-1, and E-cadherin of the pancreas collected from CPar3+/+ and CPar3fl/fl mice at 13 weeks. Scale bar = 50 μm. Quantification of immunofluorescence stains for Par3 (n = 3, 3), Zo-1 (n = 6, 6), and E-cadherin (n = 6, 6). t test, mean ± standard deviation (SD); ∗P ≤ .05, ∗∗P ≤ .01. (C) Body weights and pancreas weights of CPar3+/+ and CPar3fl/fl mice at 3 (n = 9, 10) and 6 (n = 9, 10) months of age. t test, mean ± SD; ∗P ≤ .05. (D and E) H&E stains of the pancreas from CPar3+/+ and CPar3fl/fl mice at 3, 6, and 12 months of age. The extent of fat accumulation was assessed as <50% or >50% at 12 months of age (n = 5, 9). Fisher exact test; P = .09. Scale bars = 1 mm. (F) F4/80 IHC stains of the pancreas from CPar3+/+ and CPar3fl/fl mice at 3 (n = 8, 9) and 6 months (n = 8, 7) of age. t test, mean ± SD; ∗P ≤ .05. Scale bar = 100 μm. H&E, hematoxylin and eosin; IHC, immunohistochemistry; ns, not significant.

Figure 3

Pancreatic Par3 loss results in exacerbation of cerulein-induced acute pancreatitis. (A) CPar3+/+ and CPar3fl/fl mice were treated with 8 hourly intraperitoneal injections of cerulein (50 μg/kg) and the pancreas collected and stained with hematoxylin and eosin 1 hour (n = 5, 5), 1 day (n = 5, 5), or 2 days (n = 5, 5) after the last cerulein injection. (B and C) The acute pancreatitis specimens collected at 1 day after the last cerulein injection of CPar3+/+ and CPar3fl/fl mice were stained for F4/80 (n = 4, 5), Ly-6B.2, clone 7/4 (n = 5, 5), Ki67 (n = 5, 5), and cC3 (n = 5, 6), and the relative staining was quantified. t test, mean ± standard deviation; ∗P ≤ .05, ∗∗∗P ≤ .001. (D) The acute pancreatitis specimens collected 14 days after the last cerulein injection of CPar3+/+ and CPar3fl/fl mice were stained with hematoxylin and eosin (n = 4, 4). Scale bars in A–D = 100 μm. KO, knock out; ns, not significant; WT, wild-type.

Figure 4

Pancreatic Par3 loss exacerbates acinar cell loss in cerulein-induced chronic pancreatitis. (A) CPar3+/+ and CPar3fl/fl mice were treated with intraperitoneal injections of saline or cerulein (250 μg/kg) twice daily for 14 days. (B) Weights of mice treated with saline (n = 8) or cerulein (CPar3+/+, n = 12; CPar3fl/fl, n = 11) during treatment. One-way analysis of variance at end of treatment, mean ± standard deviation (SD); ∗∗P ≤ .01, ∗∗∗∗P ≤ .0001. (C) Pancreas weights as a percentage of body weight were determined for mice treated with saline (n = 5) or cerulein (CPar3+/+, n = 8; CPar3fl/fl, n = 8). One-way analysis of variance, mean ± min to max; ∗P ≤ .05, ∗∗∗P ≤ .001, ∗∗∗∗P ≤ .0001. (D) CPar3+/+ and CPar3fl/fl mice were treated with cerulein (250 μg/kg) twice daily for 3 days (n = 6, 5), 5 days (n = 7, 7), 10 days (n = 3, 6), and 14 (n = 4, 5) days and the pancreas hematoxylin and eosin stained. (E) CPar3+/+ and CPar3fl/fl mice were treated with saline or cerulein (250 μg/kg) twice daily for 14 days. Weights of mice treated with saline (n = 4) or cerulein (CPar3+/+, n = 7; CPar3fl/fl, n = 9) during and after treatment. t test, mean ± SD; ∗P ≤ .05. (F) Amylase staining of the pancreas 2 weeks (saline, n = 3; cerulein, CPar3+/+, n = 4; cerulein, CPar3fl/fl, n = 3) and 2 months (cerulein, CPar3fl/fl, n = 5) postcerulein treatment, and the relative amylase staining was quantified. One-way analysis of variance, mean ± SD; ∗P ≤ .05, ∗∗P ≤ .01. Scale bars for low magnification in D and F = 1 mm. Scale bars for high magnification in D = 100 μm. H&E, hematoxylin and eosin; IHC, immunohistochemistry; ns, not significant.

Figure 5

Pancreatitis exacerbates loss of acinar compartment in mice with pancreatic Par3 loss. (A) Pancreas from CPar3+/+ and CPar3fl/fl mice treated with cerulein for 5 days were stained for Ly-6B.2, clone 7/4 (n = 4, 4), F4/80 (n = 6, 7), and CD206 (n = 6, 5), and the relative staining was quantified. t test, mean ± standard deviation; ∗P ≤ .05. Scale bar = 100 μm. (B) Pancreas from CPar3+/+ (n = 5) and CPar3fl/fl (n = 5) mice treated with cerulein twice daily for 14 days were trichrome stained or stained for α-SMA. Scale bars for low magnification = 1 mm and scale bars for high magnification = 100 μm. IHC, immunohistochemistry; ns, not significant.

Figure 6

Par3 loss results in reduced chronic pancreatitis-induced ADM formation and proliferation. CPar3+/+ and CPar3fl/fl mice were treated with intraperitoneal injections of cerulein (250 μg/kg) twice daily for 5 or 10 days. (A and C) The chronic pancreatitis mouse samples were costained with amylase, CK19, and DAPI to evaluate mouse ADM lesions, and the relative staining was quantified. (n = 4, 4 for samples in A; n = 6, 4 for samples in C). t test, mean ± standard deviation; ∗P ≤ .05, ∗∗P ≤ .01. Scale bar = 100 μm. (B and D) The specimens were stained for Ki67 and cC3, and the relative staining was quantified. (n = 4, 4 for samples in B; n = 6, 7 for samples in D). t test, mean ± standard deviation; ∗P ≤ .05; ∗∗P ≤ .01. Scale bar = 100 μm. KO, knock out; ns, not significant; WT, wild-type.

Figure 7

Mice with pancreatic Par3 loss demonstrate reduced chronic pancreatitis-induced primary cilia. (A and B) CPar3+/+ and CPar3fl/fl mice were treated with intraperitoneal injections of cerulein (250 μg/kg) twice daily for 5 days. Gene set enrichment analysis and enrichment plot for primary cilium using Day 5 pancreatitis RNA samples from CPar3+/+ and CPar3fl/fl mice (n = 3, 3). Heatmap for the top genes associated with the primary cilium. (C–E) Pancreas from control (UT) CPar3+/+ and CPar3fl/fl mice or from mice treated with cerulein for 5 days were stained for acetylated tubulin (n = 5, 4, 4, 4), Arl13b (n = 4, 3, 4, 4), and FoxJ1 (n = 5, 5, 5, 5), and the relative staining was quantified. One-way analysis of variance, mean ± standard deviation; ∗∗P ≤ .01, ∗∗∗∗P ≤ .0001. Scale bar in C and D = 25 μm and in E =100 μm. ns, not significant; UT, untreated.

Figure 8

BET inhibition attenuates pancreatitis-induced loss of primary cilia in mice lacking pancreatic Par3. (A) CPar3+/+ and CPar3fl/fl mice were treated with DMSO or JQ1 (50 mg/kg) for 7 days and then concurrently treated with DMSO or JQ1 (50 mg/kg) and cerulein (250 μg/kg) for additional 5 days. Gene set enrichment analysis and enrichment plot for primary cilium using RNA samples from CPar3fl/fl mice treated with cerulein and cotreated with DMSO or JQ1 (n = 3, 3). Heatmap for the top genes associated with the primary cilium. (B) CPar3+/+ and CPar3fl/fl mice were treated with DMSO or JQ1 (50 mg/kg) for 7 days and then concurrently treated with DMSO or JQ1 (50 mg/kg) and cerulein (250 μg/kg) for additional 5 days. Pancreas were stained for F4/80 (n = 5, 5, 6), and the relative staining was quantified. One-way analysis of variance, mean ± standard deviation. Scale bar = 100 μm. (C–E) Pancreatic sections from mice treated with cerulein for 5 days and cotreated with DMSO or JQ1 were stained for FoxJ1 (n = 4, 4, 5), acetylated tubulin (n = 5, 4, 4), and Arl13b (n = 4, 4, 4) and the relative staining was quantified. One-way analysis of variance, mean ± standard deviation; ∗P ≤ .05, ∗∗∗∗P ≤ .0001, Scale bar in C = 100 μm and in D and E = 25 μm. DMSO, dimethylsulfoxide; ns, not significant.

Figure 9

BET inhibition promotes ADM in mice lacking pancreatic Par3. CPar3+/+ and CPar3fl/fl mice were treated with DMSO or JQ1 (50 mg/kg) for 7 days and then concurrently treated with DMSO or JQ1 (50 mg/kg) and cerulein (250 μg/kg) for additional 5 days. (A) The chronic pancreatitis mouse samples were costained with amylase, CK19, and DAPI to evaluate mouse ADM lesions, and the relative staining was quantified. (n = 4, 4, 4). One-way analysis of variance, mean ± standard deviation; ∗∗P ≤ .01, ∗∗∗P ≤ .001, ∗∗∗∗P ≤ .0001. Scale bar = 100 μm. (B) The specimens were stained for Ki67 and cC3, and the relative staining was quantified. (n = 5, 4, 4). t test, mean ± standard deviation; ∗∗P ≤ .01, ∗∗∗P ≤ .001. Scale bar = 100 μm. DMSO, dimethylsulfoxide; ns, not significant.

Figure 10

BET inhibitors attenuate cerulein-induced acinar cell loss in mice lacking pancreatic Par3 and enhance recovery of acinar cell mass and body weight. (A and B) CPar3+/+ and CPar3fl/fl mice were treated with DMSO or JQ1 (50 mg/kg) for 7 days and then concurrently treated with DMSO or JQ1 (50 mg/kg) and cerulein (250 μg/kg) for additional 2 weeks. Mice were monitored for weight loss during treatment. Acinar cell loss at the end of treatment was analyzed by amylase staining (n = 5, 5, 5, 5), and the relative amylase staining was quantified. One-way analysis of variance, mean ± standard deviation. ∗P ≤ .05. Scale bar = 1 mm. (C and D) CPar3+/+ and CPar3fl/fl were pretreated with DMSO or JQ1 (50 mg/kg) for 7 days and then concurrently treated with DMSO or JQ1 (50 mg/kg) and cerulein (250 μg/kg) for 2 weeks. Mice were then monitored for additional 25 days. Recovery of body weights monitored over 25 days. Acinar cell recovery 25 days posttreatment was analyzed by amylase staining (n = 4, 5, 4, 5), and the relative amylase staining was quantified. One-way analysis of variance, mean ± standard deviation; ∗P ≤ .05, ∗∗∗P ≤ .001. Scale bar = 1 mm. DMSO, dimethylsulfoxide; IHC, immunohistochemistry.