Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Sep;243(1):65-77.
doi: 10.1002/path.4928. Epub 2017 Jul 27.

Glycogen Synthase kinase-3β Ablation Limits Pancreatitis-Induced Acinar-To-Ductal Metaplasia

Affiliations
Free PMC article

Glycogen Synthase kinase-3β Ablation Limits Pancreatitis-Induced Acinar-To-Ductal Metaplasia

Li Ding et al. J Pathol. .
Free PMC article

Abstract

Acinar-to-ductal metaplasia (ADM) is a reversible epithelial transdifferentiation process that occurs in the pancreas in response to acute inflammation. ADM can rapidly progress towards pre-malignant pancreatic intraepithelial neoplasia (PanIN) lesions in the presence of mutant KRas and ultimately pancreatic adenocarcinoma (PDAC). In the present work, we elucidate the role and related mechanism of glycogen synthase kinase-3beta (GSK-3β) in ADM development using in vitro 3D cultures and genetically engineered mouse models. We show that GSK-3β promotes TGF-α-induced ADM in 3D cultured primary acinar cells, whereas deletion of GSK-3β attenuates caerulein-induced ADM formation and PanIN progression in KrasG12D transgenic mice. Furthermore, we demonstrate that GSK-3β ablation influences ADM formation and PanIN progression by suppressing oncogenic KRas-driven cell proliferation. Mechanistically, we show that GSK-3β regulates proliferation by increasing the activation of S6 kinase. Taken together, these results indicate that GSK-3β participates in early pancreatitis-induced ADM and thus could be a target for the treatment of chronic pancreatitis and the prevention of PDAC progression. Copyright © 2017 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Keywords: KRas; S6 K; acinar-to-ductal metaplasia; glycogen synthase kinase-3β; pancreatic cancer; pancreatitis.

Conflict of interest statement

The other authors disclosed no potential conflicts of interest.

Figures

Figure 1.
Figure 1.. GSK-3β protein expression is elevated following caerulein-induced pancreatitis in mice.
(A) Representative IHC images of GSK-3β in pancreatic sections from Pdx1-cre wild-type (WT) (upper panel) and Pdx1-cre;LSL-KRasG12D (KRas) mice (lower panel) treated with or without caerulein. Arrows point to the abnormal pancreas area. Bars, 200 μm. (B) Representative immunoblot analysis for phosphorylated Erk1/2, total Erk1/2, GSK-3α and GSK-3β in pancreatic tissue of WT and KRas mice before and 1, 2, 5 days post caerulein treatment. β-actin was used as loading controls. The black vertical line in (B) denotes where an additional time point was removed. Samples were developed from the same membrane. Shown are representative results from 3 experiments. (C) The average signal intensity of phospho-Erk1/2, GSK-3α, GSK-3β was quantified and expressed as mean ± SEM. n=6. *P<0.05 KRas versus WT mice.
Figure 2.
Figure 2.. GSK-3β promotes TGF-α-induced ADM.
(A) Freshly isolated primary mouse acinar cells were embedded in a 3D collagen culture. Images show acinar cell clusters or ducts on day 6 following treatment with DMSO, 50 ng/ml TGF-α, 1 μM GSK-3i or 50 ng/ml TGF-α combined with 1 μM GSK-3i. Bars, 250 μm. (B) Effect of GSK-3 inhibition on TGF-α-induced ADM in vitro. Number of ducts was quantified by counting. *P<0.05 TGF-α versus DMSO. #P<0.05 TGF-α combined with GSK-3i versus TGF-α only. (C) Primary mouse pancreatic acinar cells were isolated, transduced with GSK-3β kinase-dead (KD) lentiviral particles and seeded in collagen with or without TGF-α (50 ng/ml). The number of ducts formed was enumerated on day 6 post stimulation. (D) Primary mouse pancreatic acinar cells were isolated, transduced with constitutively active GSK-3β (S9A) lentiviral particles and seeded in collagen. The number of ducts formed was enumerated on day 6 post stimulation. *P<0.05 TGF-α or constitutively active GSK-3β (S9A) versus Control. #P<0.05 TGF-α combined with kinase-dead (KD) mutant versus TGF-α only. (E) Primary acinar cells from Pdx1-cre; GSK-3β (WT) and Pdx1-cre;GSK-3βF/F (KO) mice were isolated and cultured for 6 days in collagen in the presence of DMSO or TGF-α 50 ng/ml. Representative images of ductal-like structures are shown. Bars, 250 μm. (F) Quantification of ducts formed following 6-days of stimulation. *P<0.05 TGF-α treated versus untreated WT acinar cells. #P<0.05 TGF-α treated KO acinar cells versus WT acinar cells. All experiments were repeated using acinar cells derived from at least three different mice. Data were analyzed and expressed as mean ± SEM. n = 15.
Figure 3
Figure 3. Deletion of GSK-3β in the pancreas of genetically engineered mice.
(A) Pdx1-cre, LSL-KRasG12D, GSK-3βF/F mice were used to generate mice of desired genotypes as described in the Materials and Methods. The expression of phospho-GSK-3β (Ser9), GSK-3α/β and β-catenin in pancreatic tissues was examined by immnoblot. β-actin was used as a loading control. Shown are representative results from 6 experiments. (B) The average signal intensity of phospho-GSK-3β (Ser9), GSK-3α/β, β-catenin were quantified and expressed as mean ± SEM. n=6. *P<0.05 KO versus WT mice. #P<0.05 RKO versus KRas mice. (C) H&E and IHC staining for detection of GSK-3β expression in pancreas sections of WT, KRas, KO, and RKO mice. Representative images were taken under low and high magnification lens. Bars, 200 μm.
Figure 4.
Figure 4.. GSK-3β is necessary for KRas-initiated ADM in vivo.
(A) Scheme for caerulein-induced acute pancreatitis model and analysis. (B) Gross pathology of pancreas and adjacent tissues from transgenic mice of indicated genotypes treated with saline or 2 and 7 days post caerulein injection. (C) H&E stained pancreatic sections from WT, KRas or RKO mice (left panel) and immunofluorescence staining of amylase and CK19 (right panel) of mice treated with saline or 2 and 7 days post caerulein injection. Nuclei were counter-stained with Hoechst (blue). (D) H&E stained tissue samples from WT (black), KRas (red), KO (grey) and RKO (green) mice, respectively, were evaluated and quantitatively analyzed for ADM area as percentage to the total area. Data were analyzed and expressed as mean ± SEM. n = 25. *P<0.05 KRas versus WT mice. #P<0.05 RKO versus KRas mice. (E) Real-time PCR quantification of pancreatic gene expression for amylase (black), CK7 (red) and CK19 (blue) from WT, KRas or RKO mice treated with saline or 2, 7 days post caerulein injection. RPLP0 and GAPDH were used as internal housekeeping gene controls. Data were analyzed and expressed as mean ± SEM. n = 5. *P<0.05 KRas versus WT mice. #P<0.05 RKO versus KRas mice.
Figure 5.
Figure 5.. GSK-3β contributes to proliferation of ADM cells.
(A) IHC staining and analysis of WT, KRas and RKO mice for Ki-67 expression in the pancreas. Representative images were taken under 10× magnification lens. Bars, 200 μm. (B) Double-labeling of pancreatic sections from KRas and RKO mice was performed 2 or 7 days post caerulein injection using Ki-67 (red) and CK19 (green) antibodies. Nuclei were counter-stained with Hoechst (blue). (C) Quantification of the percentage of Ki-67 positive ADM cells in KRas and RKO mice 2 or 7 days post caerulein injection. Data were analyzed and expressed as mean ± SEM. n = 6. *P<0.05 RKO versus KRas mice.
Figure 6.
Figure 6.. GSK-3β deletion decreases the levels of phospho-S6.
(A) Double-labeling of the pancreatic sections for pS6 (red) and CK19 (green). The pancreatic sections were derived from KRas and RKO mice 2 or 7 days post caerulein injection. Nuclei were counter-stained with Hoechst (blue). (B) AR42J cells were pre-treated with diluent (DMSO) or GSK-3i for 2 hours. They were then treated with caerulein for an additional 2 or 6 hours. Cell lysates were prepared and probed with the indicated antibodies. (C) Stable control and shGSK-3β AR42J cells were treated with caerulein for 2, 6 and 12 hours. Cell lysates were prepared and probed with the indicated antibodies. Shown are representative results from 6 independent experiments.

Similar articles

See all similar articles

Cited by 4 articles

Publication types

MeSH terms

Feedback