Cerulein-induced chronic pancreatitis does not require intra-acinar activation of trypsinogen in mice

Abstract

Background & aims: Premature activation of trypsinogen activation can cause pancreatic injury and has been associated with chronic pancreatitis (CP). Mice that lack intra-acinar activation of trypsinogen, such as trypsinogen-7-null (T(-/-)) and cathepsin B-null (CB(-/-)) mice, have been used to study trypsin-independent processes of CP development. We compared histologic features and inflammatory responses of pancreatic tissues from these mice with those from wild-type mice after the development of CP.

Methods: CP was induced in wild-type, T(-/-), and CB(-/-) mice by twice-weekly induction of acute pancreatitis for 10 weeks; acute pancreatitis was induced by hourly intraperitoneal injections of cerulein (50 μg/kg × 6). Pancreatic samples were collected and evaluated by histologic and immunohistochemical analyses. Normal human pancreas samples, obtained from the islet transplant program at the University of Minnesota, were used as controls and CP samples were obtained from surgical resections.

Results: Compared with pancreatic tissues from wild-type mice, those from T(-/-) and CB(-/-) mice had similar levels of atrophy, histomorphologic features of CP, and chronic inflammation. All samples had comparable intra-acinar activation of nuclear factor (NF)-κB, a transcription factor that regulates the inflammatory response, immediately after injection of cerulein. Pancreatic tissue samples from patients with CP had increased activation of NF-κB (based on nuclear translocation of p65 in acinar cells) compared with controls.

Conclusions: Induction of CP in mice by cerulein injection does not require intra-acinar activation of trypsinogen. Pancreatic acinar cells of patients with CP have increased levels of NF-κB activation compared with controls; regulation of the inflammatory response by this transcription factor might be involved in the pathogenesis of CP.

Figure 1. (A) Characterization of Trypsinogen-7 knock-out (T−/−) mice

FPLC analysis of pancreatic homogenates showed absence of trypsinogen-7 peak in T−/− mice. The specificity of this peak was confirmed by N-terminal sequencing. (B) T−/− mice lack pathologic trypsinogen activation. Trypsin activity was measured at indicated time-points in caerulein treated groups (compared to respective saline treated controls) as described in methods. N =10-16/group, pooled from 3 independent experiments. (C, D) Caerulein-induced activation of chymotrypsinogen (C) and elastase (D) in WT and T−/− mice. Significant activation of chymotrypsin and elastase was seen in WT mice at 30 minute time-point which was not observed in T−/− mice. P>.9 for chymotrypsin activity in control vs 8-hr or 12-hr time-points both for WT and T−/− mice. N=10-15/group for chymotrypsin activity. N=4-6 for controls and 30 minute groups, 10-12 for 8 hr and 12 hr groups for elastase activity. WT and T−/− control mice showed similar elastase activities (11±2 vs 12±1 slope/mg protein, P>.5) suggesting lack of compensatory changes in elastase expression in T−/− mice. (E) Cathepsin B knock-out (CB−/−) mice as a parallel tool. These mice have been previously characterized to lack significant pathologic trypsinogen activation which was confirmed in our laboratory. Trypsin activity was measured 30 minutes after caerulein injection (50μg/kg i.p.). N≥6 per group. *=statistically significant. Enzyme activities reported in B-E were measured in pancreatic homogenates, are normalized to respective controls and expressed as percent of maximal activity which was 90.4±12 for trypsin, 116.4±51 for chymotrypsin and 16.4±3 for elastase activities (slope/mg protein).

Figure 2. Atrophy of pancreas in WT, T−/− and CB−/− CP groups was comparable

(A)Representative pictures demonstrating pancreas atrophy. (B) Comparison of whole pancreas weight. N= 16-20/group, P<.0001 pair wise for each CP group vs control, not significant pair wise among CP groups.

Figure 3. (A) Histologic features characteristic of chronic pancreatitis were similar in all experimental CP groups

Representative hematoxylin and eosin-stained sections (100×) showing moderate acinar loss, fibrosis, microscopic duct dilatations, tubular complexes and inflammatory infiltrate WT, T−/− and CB−/− CP groups (N=16-20/group). (B) Fibrosis: Sirius red-stained sections (100×) under fluorescence showing marked fibrosis in CP groups. Quantification (C) confirmed comparable fibrosis in WT, T−/− and CB−/− CP groups (P<.0001 pair wise for each CP group vs control, not significant pair wise among CP groups, N=16-20/group). (D,E) Comparable overexpression of collagen-1, desmin and αSMA in CP groups. Representative western blots are shown in (D) and their quantification is shown as box-whisker plots with outliers in (E). N=5-8/group for each protein. Control vs WT, T−/−, CB−/− CP groups respectively: P= .0008, .001, .04 for collagen-1; .001, .001, .009 for desmin and .001, .001, .009 αSMA. P-values not significant for all pair wise comparisons among CP groups. Vimentin staining confirmed marked increase in stellate cell population in the CP groups (F). Representative sections (200×) stained with vimentin are shown.

Figure 4. (A) Ductular metaplasia in CP groups

Representative cytokeratin-19 stained sections (100×) from controls and CP groups showing increase in ductular/tubular complexes. These structures were counted in high power (100×) fields (B). Tubular complexes (per field) increased to 30±1, 34±1, 32±1 in WT, T−/−, CB−/− CP groups from 3±0.2 in controls (P<.0001 for each CP group vs controls, N=10/group). (C) Aquaporin-1 (AQP1) overexpression in CP: Dysfunction of ducts in CP is believed to cause compensatory overexpression of water channels Aquaporin-1 (see text). Comparable overexpression of AQP1 was seen in WT, T−/− and CB−/− groups with CP as compared to controls. Representative AQP-1 stained sections (200×) are shown (N=10/group).

Figure 5. (A, B) Inflammatory infiltrate in CP groups

Representative sections (200×) stained for CD3 (T-Cell marker) shown in (A), and quantified in (B). Comparable in WT, T−/−, CB−/− CP groups (P=.03, .01, .01 respectively vs control, not significant pair wise among CP groups, N=10/group). (C) Sustained NFκB activation in chronic pancreatitis: NFκB p65 immunostaining demonstrates its nuclear localization in acinar cells. Representative sections (200×) from controls and WT, T−/−, and CB−/− CP groups (insets are zoomed-in views showing closer details). (D) NFκB activation occurs independent of trypsinogen activation in immediate response to pathologic stimulus: NFκB activation in the pancreas 30 minutes after caerulein injection (50μg/kg) was assessed by EMSA and relative band intensities have been plotted. p<0.0001 for each group vs controls, not significant for other pairwise comparisons; N=3-5/group. Controls include N=3 each of WT, T−/− and CB−/− mice which were similar. Intra-acinar localization of NFκB was confirmed by in-vitro experiments on isolated acinar cells (see text). (E) Cyclooxygenase-2 (COX2): Overexpression of COX2, which is regulated by NFκB, was comparable in WT, T−/−, CB−/− CP groups (P= .0008, .001, .014 respectively vs controls, not significant pair wise among CP groups; N=5-8/group).

Figure 6. NFκB activation in human chronic pancreatitis

A) Representative hematoxylin and eosin-stained sections (100×) of pancreas from healthy controls and CP patients demonstrating histo-morphologic features of CP. Similar features were seen in the mouse CP groups. B and C) Persistent NFκB activation occurs in human chronic pancreatitis. NFκB component protein p65 immunostaining demonstrates its nuclear localization in acinar cells (B) in human chronic pancreatitis sections (200×) (insets are zoomed-in views showing closer details). Acinar cells whose nuclei stained positive for p65 were counted in each field (200×) and are shown in (C). Pancreas sections from seven CP patients and seven controls were analyzed (P<.001).

Figure 7. A simplistic schematic summarizing novel findings in the pathogenesis of pancreatic injury

The section on acute pancreatitis (in navy blue) is based on our previous study (9) and is being mentioned here for completeness. The left side of the schematic (dark red) describes the proposed paradigm of chronic pancreatitis stating that sustained activation of inflammatory pathways (NFκB pathway is thought to be the most important) in acinar cells results in chronic pancreatitis. Chronicity of pathogenic stimulus (which may be any of the recognized etiologic associations) may drive the sustained NFκB response. A complex interplay of genetic predispositions as well as environmental factors may influence the effect of pathogenic stimuli. In this schematic, solid lines depict links with experimental proof while hypothesized links are shown by dotted lines.