Preclinical mouse model of a misfolded PNLIP variant develops chronic pancreatitis

Abstract

Objective: Increasing evidence implicates mutation-induced protein misfolding and endoplasm reticulum (ER) stress in the pathophysiology of chronic pancreatitis (CP). The paucity of animal models harbouring genetic risk variants has hampered our understanding of how misfolded proteins trigger CP. We previously showed that pancreatic triglyceride lipase (PNLIP) p.T221M, a variant associated with steatorrhoea and possibly CP in humans, misfolds and elicits ER stress in vitro suggesting proteotoxicity as a potential disease mechanism. Our objective was to create a mouse model to determine if PNLIP p.T221M causes CP and to define the mechanism.

Design: We created a mouse model of Pnlip p.T221M and characterised the structural and biochemical changes in the pancreas aged 1-12 months. We used multiple methods including histochemistry, immunostaining, transmission electron microscopy, biochemical assays, immunoblotting and qPCR.

Results: We demonstrated the hallmarks of human CP in Pnlip p.T221M homozygous mice including progressive pancreatic atrophy, acinar cell loss, fibrosis, fatty change, immune cell infiltration and reduced exocrine function. Heterozygotes also developed CP although at a slower rate. Immunoblot showed that pancreatic PNLIP T221M misfolded as insoluble aggregates. The level of aggregates in homozygotes declined with age and was much lower in heterozygotes at all ages. The Pnlip p.T221M pancreas had increased ER stress evidenced by dilated ER, increased Hspa5 (BiP) mRNA abundance and a maladaptive unfolded protein response leading to upregulation of Ddit3 (CHOP), nuclear factor-κB and cell death.

Conclusion: Expression of PNLIP p.T221M in a preclinical mouse model results in CP caused by ER stress and proteotoxicity of misfolded mutant PNLIP.

Keywords: CELL DEATH; CHRONIC PANCREATITIS; PANCREATITIS.

Fig 1.

The changes of the pancreas from Pnlip p.T221M homozygous (HOM), heterozygous (HET), and wild type (WT) male mice with age from 1-12 months. The quantification graphs are shown as means of individual values with standard deviation at each indicated age. A. Body weight. B. Pancreas weight. By 2-way ANOVA, there was a statistically significant interaction between genotype and age (p≤0.001). The pancreas from both HET and HOM mice was smaller compared to the pancreas from WT mice at all ages except for 1 month (p≤0.001). The pancreas from HOM mice were statistically smaller compared to the pancreas from HET mice at all ages except 1 month (p≤0.001). C. Serum amylase. Multiple t-tests were conducted to compare serum amylase in HOM vs WT mice and HET mice vs WT mice at each matched age. D. H&E stained mouse pancreatic sections. n > 10 per genotype at each age point except for n ≥ 5 in D.

Fig 2.

Fibrogenesis in the pancreas from Pnlip p.T221M homozygous (HOM) and wild type (WT) mice at age 6 months. A. Masson’s trichrome staining shows widespread fibrotic change (blue stain) in the pancreas from HOM mice. B. The quantification of fibrosis as assessed by trichrome staining shown in A. C. Hydroxyproline content is increased in the pancreas from HOM mice. D. Representative of immunoblots of soluble protein extraction from the mouse pancreas at ages of 2 and 6 months against α-smooth muscle actin (α-SMA), and α-tubulin served as an endogenous control (top panel). The quantification graph of immunoblots (bottom panel). E. Transmission electron micrograph (TEM) of the pancreatic sections from HOM mice shows increased collagen (Col), with ER, Mt, and ZG labeled for endoplasmic reticulum, mitochondria, and zymogen granules, respectively. n ≥ 3 in A, B, and E; n ≥ 6 in C; n ≥ 4 in D.

Fig 3.

Inflammation in the pancreas of Pnlip p.T221M homozygous (HOM) and wild type (WT) mice at age 3.5 months. A. Immunohistochemistry stained pancreatic sections from HOM mice show increased immune cell infiltration (brown stain). F4-80 for macrophages, myeloperoxidase for neutrophils, B220/CD45R for B cells, and CD3 for T cells. B. The quantification of cells stained positive for each marker shown in A. n ≥ 10.

Fig 4.

Activation of NF-kB pathway in the pancreas of Pnlip p.T221M homozygous (HOM) and wild type (WT) mice at age 3 months. A. Representative immunoblot image for NF-kB subunit p65 and its phosphorylated form, p-p65, in the soluble pancreatic protein extracts, with α-tubulin serving as an endogenous control. B. The quantification graph for the protein abundance of p65 and p-p65 shown in A. C. Representative image of immunohistochemistry of mouse pancreatic sections for p65 and p-p65. D. The quantification of nuclear positive stains for p65 and p-p65. n ≥ 10 for A and B, and n ≥ 14 for C and D.

Fig 5.

Pancreatic zymogen content from Pnlip p.T221M homozygous (HOM) and wild type (WT) mice with age from 1-12 months. A. The activity of pancreatic lipase. B. The activity of amylase. C. The activity of trypsin. The activity of each enzyme was normalized by protein content. Pancreatic total activity for each enzyme was calculated for the entire pancreas. The data are shown as Whiskers plots with individual values and the median box for the 25th to 75th percentiles. n ≥ 3 per genotype per age point.

Fig 6.

PNLIP trafficking and localization in the pancreatic acinar cells from Pnlip p.T221M homozygous (HOM) and wild type (WT) mice at 3 months. Co-immunofluorescence staining was performed on pancreatic sections, and Hoechst stained for nuclear. A. Co-immunostaining of PNLIP and amylase (another digestive enzyme); B. Co-immunostaining of PNLIP and calreticulin (an ER marker). Bar, 10 μm. n = 3 per genotype.

Fig 7.

PNLIP misfolding and ER stress in the pancreas of Pnlip p.T221M homozygous (HOM) and wild type (WT) mice with age. A. Representative immunoblots of PNLIP and BiP in the detergent-soluble and - insoluble fractions of pancreatic protein extracts, and α-tubulin served as an endogenous control for the soluble fractions. B. The quantification graphs of immunoblot analysis of PNLIP and BiP in A. C. qPCR analysis of Hsp5a (BiP) mRNA abundance in the mouse pancreatic RNA extracts, and Rpl13a served as a control. The results are expressed as fold change relative to the mean of the results of age-matched WT mice. D. Representative of transmission electron micrographs (TEM) of the mouse pancreatic sections at 6 months shows a profound dilation of the ER in the HOM pancreas (8000x and 25000x). ER: endoplasmic reticulum; ZG: zymogen granule. n ≥ 6 per genotype per age point in A, B, C; n = 3 in D.

Fig 8.

Age-dependent activation of the unfolded protein response (UPR) in the pancreas of Pnlip T221M homozygous (HOM) and wild type (WT) mice. A. qPCR analysis of mRNA abundance of various markers of the UPR in the mouse pancreatic RNA extracts, and Rpl13a served as controls. The makers include Xbp1 mRNA splicing, Atf6, Atf4, and Atf3. B. Representative immunoblots of phosphorylated eIF2α (p-eIF2α) in the detergent-soluble pancreatic protein extracts, and α-tubulin served as an endogenous control (top panel); the quantification graph of immunoblot analysis of p-eIF2α (bottom panel). C. Immunohistochemistry of the mouse pancreatic sections for p-eIF2α. D. qPCR analysis of mRNA abundance of Ddit3 (CHOP) in the mouse pancreatic RNA extracts, and Rpl13a served as controls. n ≥ 6 per genotype per age point except for n = 3 in C.

Fig 9.

Increased cell death in the pancreas of Pnlip p.T221M homozygous (HOM) and wild type (WT) mice at age 3.5 months. A. H&E staining (40 x magnification) of the pancreatic sections showing both necrotic (arrow head) and apoptotic (long arrow) cells and larger acinar cells. B. Immunohistochemistry staining of cleaved-caspase 3 in the pancreatic sections (brown), and the staining on the pancreatic sections at 1 month also included. C. TUNEL staining of the pancreatic sections (brown). D. Immunohistochemistry staining of high mobility group box 1 (HMGB1) in the pancreatic section showing increased nuclear to cytoplasmic HMGB1 transition (arrowhead). In B-D, the corresponding quantification graph is listed on the right. n=3 in A and n ≥ 10 in B-D.