Establishment of a pancreatic cancer animal model using the pancreas-targeted hydrodynamic gene delivery method.

This research developed an easy-to-use, reproducible pancreatic cancer animal model utilizing pancreas-targeted hydrodynamic gene delivery to deliver human pancreatic cancer-related genes to the pancreas of wild-type rats. KRAS G12D -induced pancreatic intraepithelial neoplasia lesions showed malignant transformation in the main pancreatic duct at 4 weeks and developed acinar-to-ductal metaplasia, which led to pancreatic ductal adenocarcinoma within 5 weeks, and the gene combination of KRAS G12D and YAP enhanced these effects. The repeat hydrodynamic gene delivery of KRAS G12D  + YAP combination at 4 weeks showed acinar-to-ductal metaplasia in all rats and pancreatic ductal adenocarcinoma in 80% of rats 1 week later. Metastatic tumors in the liver, lymph nodes, and subcutaneous lesions and nervous invasion were confirmed. KRAS G12D and YAP combined transfer contributes to the E- to N-cadherin switch in pancreatic ductal adenocarcinoma cells and to tumor metastases. This pancreatic cancer model will speed up pancreatic cancer research for novel treatments and biomarkers for early diagnosis.

Introduction

Pancreatic ductal adenocarcinoma (PDAC) is a leading cause of cancer-related deaths worldwide,, and the development of effective therapies is an unmet clinical need. The lack of an appropriate pancreatic cancer model is the major drawback. The model must mimic human pancreatic cancer in molecular pathogenesis, histological features, and multi-step malignant transformation for tumor marker and therapy development. Currently available models include chemical carcinogen-induced animal models,3, 4, 5, 6, 7, 8 cancer cell line-based xenografts, patient-derived xenografts, and organoid models. In addition, injection of adenovirus-expressing Cre recombinase into the pancreatic duct of Hras and Kras transgenic rats showed pancreatic neoplasia, which is considered to be a pancreatic cancer model. However, these models fail to achieve the reproducibility, cost benefit, efficacy, and similarity with human PDAC necessary for the development of novel therapies and biomarkers. Therefore, novel pancreatic cancer animal models with simple, easy, and reproducible methods are essential.

With this aim, we applied the hydrodynamic gene delivery (HGD) method, which has been used successfully to develop mice liver cancer in vivo by delivering oncogenes, including Yap, Ras, and Myc, to the liver11, 12, 13. To utilize this method, we established pancreas-targeted HGD to wild-type rats. The major advantages of this procedure include safe and effective gene delivery in a pancreas-specific manner, and optimization of injection parameters achieved effective transgene expression in vivo with no leakage of the transgene to other organs. For the transgenes, the KRAS variant is related to pancreatic intraepithelial neoplasia (PanIN), and this dosage activated PDAC occurrence in almost all pancreatic cancers.,, Sequentially activated genes, including CDKN2A, TP53, SMAD4, and YAP genes, increase the malignant potential of the tumors and contribute to the development of later-stage pancreatic intraepithelial neoplasia (PanIN) and acinar-to-ductal metaplasia (ADM).17, 18, 19 ADM is the initial stage of PDAC,20, 21, 22 and YAP showed high expression in PDAC contributing to malignant transformation and metastasis through epithelial-mesenchymal transition (EMT).23, 24, 25, 26, 27 Interestingly, it has been reported that YAP is a critical oncogenic effector of KRAS for the development of PDAC and can be a promising therapeutic target for PDAC., Therefore, in this study, we utilized the pancreas-targeted HGD to examine the development of in vivo pancreatic cancer models by delivering oncogenes, including KRAS and YAP.

Results

Effect of the in vivo oncogene HGD on the oncogenic signaling pathway

To examine whether in vivo oncogene HGD could activate the following carcinogenic signaling pathway, the time-dependent gene expression of the KRAS-related signaling pathway was examined after pancreas-targeted HGD. The expressions of KRASG12D, Akt, Erk, Yap, Tgf-β1, and CTGF and the phosphorylation of Akt and Erk were assessed (Figure 2). Western blotting showed effective expression of KRASG12D, an increase of Akt, and phosphorylation of Akt 1 day after HGD, which was sustained for 5 days and decreased or was inactivated within 4 weeks. Erk showed a longer increase and phosphorylation duration, which was sustained for 4 weeks. Increased Yap, Tgf-β1, and Ctgf expressions were confirmed following the HGD of KRASG12D. HGD of the YAP gene showed increased Yap, Tgf-β1, and Ctgf expressions more clearly than KRAS HGD, while no increase of Kras, Akt, and Erk expression was seen (Figure 2A). Protein expression analyses in vitro using BxPC-3 and Panc-10.05, derived from pancreatic adenocarcinoma and hTERT-HPNE, with hTERT-immortalized pancreatic duct-derived epithelial cell used as a control, showed activation of Akt or Erk followed by the expressions of Yap, Tgf-β1, and Ctgf (Figure 2B). These results indicate that in vivo gene transfer of oncogenes to the pancreas achieved activation of the oncogene-related signaling pathway, which is activated in pancreatic cancer.

Figure 2

Expression of carcinogenic signaling pathway proteins

(A) Expressions of KRASG12D, Akt, pAKT, Erk, pErk, Yap, Tgf-β1, Ctgf, and β-actin in the pancreas of KRAS-HGD and YAP-HGD rats. (B) Expression of the proteins in the cell lines of hTERT-HPNE and hTERT-immortalized pancreatic duct-derived epithelial cell, and BxPC-3 and Panc-10.05, derived from pancreatic adenocarcinoma. HGD, hydrodynamic gene delivery.

The effect of pancreas-targeted HGD of oncogene on pancreatic tumor development

To examine the effect of pancreas-targeted HGD of oncogenes on pancreatic tumor development, pancreatic carcinogenesis-related oncogenes of KRAS, KRAS,,,, MYC, and YAP, were hydrodynamically transferred to the wild-type rat pancreas, following a previously reported method, which is briefly described in the materials and methods section (Figure 1). The plasmids were injected either individually or in combination (Figure 3A). KRAS single gene delivery showed the development of some but not all macroscopic nodular lesions (Figure 3B), and all pancreatic tissues were harvested regardless of the mass development 4 weeks after HGD during the active period of transferred gene expression, as shown in Figure 2. Main pancreatic ducts were stained with hematoxylin and eosin (H&E) and MUC5AC, the indicator for the PanIN-1B, which is the preliminary stage of pancreatic cancer, and Ki-67 as a marker of malignant potential, which is thought to be the marker for PanIN-3 (Figure 3C). The results showed that genes including KRAS plasmid-transferred rat groups (KRAS, KRAS + MYC, KRAS + MYC + YAP, KRAS + YAP) showed a predominantly papillary or micropapillary structure and variable degrees of cytologic and architectural atypia compared to the flat epithelium composed of tall columnar mucin-producing cells in the main pancreatic ducts of rats in groups transferring plasmids, including KRAS genes (KRAS, KRAS + MYC, KRAS + MYC + YAP) (Figure 3C). Immunohistochemical analyses revealed a significantly higher level of MUC5AC and Ki-67 positively stained cells in groups transferred with KRAS-expressing plasmids compared to those of the control groups (WT, NS), YAP-transferred group, and plasmid-transferred groups, including KRAS genes (KRAS, KRAS + MYC, KRAS + MYC + YAP) (Figures 3C and 3D). These results showed that pancreas-targeted HGD of KRAS gene achieved a microscopic neoplastic lesion of the pancreas with variable mucin production, which is a feature of PanIN, the precursor lesion of invasive PDAC. In addition, while single transfer of the YAP gene showed no PanIN development, combinatory transfer with the KRAS gene and YAP gene showed statistically higher MUC5AC expression than that of KRAS and KRAS + MYC, which indicated that KRAS + YAP HGD to the pancreas causes more malignant tumors in wild-type rats (Figures 3C and 3D).

Figure 1

Schematic presentation of pancreas-targeted hydrodynamic gene delivery induced pancreatic cancer model

Schema of pancreas-targeted HGD of oncogenes from the SMV with a temporary vascular blockade at the PV. HGD, hydrodynamic gene delivery; HV, hepatic vein; PV, portal vein; SV, splenic vein; IVC, inferior vena cava; SMV, superior mesenteric vein.

Figure 3

Effect of pancreas-targeted HGD of oncogene on pancreatic tumor development

(A) The list of animal group and genes injected (n = 6 for each group). WT, wild-type; NS, normal saline. (B) Macroscopic nodular lesions in the pancreas of the rats injected with KRAS. Black arrows represent nodular lesions. (C) Pancreatic tissues from rats were harvested 4 weeks after the HGD and main pancreatic ducts were stained with HE, MUC5AC, and Ki-67. Representative images of staining are shown for each group. Scale bar represents 100 μm. (D) Five different sections from each of the six rats in all groups (n = 30) were quantitatively analyzed for positively stained cells for MUC5AC and Ki-67 using ImageJ software. The values represent mean ± SEM, ∗∗p < 0.01, and ∗∗∗p < 0.001, and one-way ANOVA with post hoc Tukey’s test. (E) Effect of the repeat injection of KRAS on atypical changes in PanIN. Representative images of HE, MUC5AC, and Ki-67 stainings are shown for each group. Scale bar represents 100 μm. (F) Quantitative analyses of MUC5AC and Ki67 positively stained cells in rats with repeat injection conducted analyzed using ImageJ software. The values represent mean ± SEM, ∗p < 0.05, and ∗∗∗∗p < 0.0001, one-way ANOVA with post hoc Tukey’s test.

The effect of repeat injection of KRAS on the PanIN progression

Based on the results showing that KRAS increases the atypia of PanIN, oncogenic dosage gain affects the progression of PDAC, and hydrodynamic injection reactivates silenced transgene expression, we examined the effect of repeat HGD of KRAS on malignant transformation and carcinogenesis in the pancreas. KRAS expression peaked within a week after pancreas-targeted HGD, and its signal transduction to Akt and Erk was sustained for 4 weeks. The second administration of KRAS was performed 4 weeks after the first administration, and pancreatic tissues were harvested a week after second administration. The pancreatic tissue that received repeat injections of KRAS (KRAS/KRAS) showed high-grade dysplastic changes in the intraepithelial lesion and atypia (Figure 3E), as evidenced by the immunohistochemical staining of Ki-67 showing a larger number of positively stained cells, while no further increase of MUC5AC stainability was observed (Figure 3F). These results suggest that KRAS oncogenic dosage gain increased the malignant potential of PanIN and induced atypical lesions (Figure 3C).

The effect of combination and repeat injection of the oncogene to the ADM and PDAC development

As ADM of the pancreas is the origin of PDAC and related to KRAS tumorigenesis in pancreas, the effect of repeat injections of the oncogenes is examined on ADM formation. The rat groups of WT, NS/NS, KRAS, KRAS + YAP, KRAS/NS, KRAS/KRAS, KRAS/YAP, and KRAS + YAP/KRAS + YAP (Figure 4A) were histologically examined for ADM development (Figures 4B and 4C). H&E staining showed ADM formation after single KRAS administration; its dysplastic change was observed when combined with YAP (KRAS + YAP group), and MUC5AC and Ki-67 positively stained cells showed its increase. This tendency was enhanced by a second injection of genes, and KRAS/KRAS, KRAS/YAP, and KRAS + YAP/KRAS + YAP showed clear increases in Ki-67 stainability and MUC5AC-positive cells (Figure 4C). Among these groups, KRAS + YAP/KRAS + YAP showed severe dysplasia and macroscopic tumors (Figure 4D). We summarized the frequency of macroscopic tumor development in these repeat injection groups in Figure 4E. H&E images of the representative pancreatic tumor including ADM and PDAC lesions in the rat injected with KRAS + YAP/KRAS + YAP genes are shown in Figure 4F. ADM lesions are seen in acinar cells and PDAC lesions showing severe dysplasia in ADM lesions with fibrotic changes. While single KRAS administration showed ADM in 33% of rats, repeat KRAS injection (KRAS/KRAS) showed significantly increased ADM formation to 66% and PDAC in 33% of rats, thus indicating that ADM formation is related to the dosage of KRAS proteins and activation of the signaling pathway (Figure 4E). KRAS/YAP showed tumorous lesions macroscopically (Figure 4D); among rats, all showed ADM, 40% showed PDAC, and 20% showed metastatic tumors in lymph nodes and the liver. This indicates that induction of ADM by KRAS followed by YAP gene transfer activates malignant transformation, which was confirmed by KRAS + YAP/KRAS + YAP showing several macroscopically defined tumors in the pancreas (Figure 4D) that were histologically 100% ADM and 80% PDAC. In addition, 30% of rats showed metastatic lesions in the lymph node and liver, subcutaneous lesions, and nervous invasion. Immunohistochemical staining of Ki-67 on the pancreatic tissues of KRAS + YAP/KRAS + YAP showed the highest number of positively stained cells among the groups (Figure 4C). In addition, PDAC tumors showed activation of Akt, Erk, Yap, Tgf-β1, and Ctgf, which was consistent with in vivo and in vitro gene assessment (Figure 4G). Serum levels of CA19-9 were statistically higher in KRAS-transferred group and further in KRAS and YAP-transferred rats (Figure 4H). These results suggest that pancreas-targeted HGD of KRAS and YAP genes develop pancreatic cancer in wild-type rats in an efficient and timely manner.

Figure 4

Effect of combination and repeat injection of the oncogene to the ADM and PDAC development

(A) The list of animal group and genes injected (n = 6 for each group). NS, normal saline. (B) Pancreatic tissues from the rat group of WT, KRAS, and KRAS + YAP were harvested 4 weeks after the HGD and NS/NS, KRAS/NS, KRAS/KRAS, KRAS/YAP, KRAS + YAP/KRAS + YAP were harvested a week after the second injections (5 weeks after the first injection). Representative images of HE, MUC5AC, and Ki-67 stainings of ADM lesions of each rat group. Scale bar represents 100 μm. (C) Five different sections from each of the six rats in all groups (n = 30) were quantitatively analyzed for positively stained cells for MUC5AC and Ki-67 using ImageJ software. The values represent mean ± SEM, ∗∗p < 0.01, and ∗∗∗∗p < 0.0001, one-way ANOVA with post hoc Tukey’s test. (D) Macroscopic nodular lesions in the pancreas of the rats injected with KRAS/YAP and KRAS + YAP/KRAS + YAP. Black dotted circles and arrows represent the tumor lesions. (E) The summary of the tumorous lesions found in the repeatedly injected rats. N/A, not applicable. (F) Representative HE stainings of ADM and PDAC tumor lesions in the rat injected with KRAS + YAP/KRAS + YAP genes. (G) Representative immunohistochemical staining of Akt, Erk, Yap, TGF-β1, and CTGF in PDAC tumor in KRAS + YAP/KRAS + YAP genes injected rats. (H) Serum level of CA19-9 in rat groups (n = 6 from each group). The values represent mean ± SEM, ∗p < 0.05, and ∗∗∗∗p < 0.0001, one-way ANOVA with post hoc Tukey’s test.

Assessment of metastatic tumors

To confirm the histological diagnosis of the metastatic tumors found in rats transferred with KRAS and YAP, tumors in the liver and lymph node, and subcutaneous lesions and nervous invasion that were diagnosed in the KRAS + YAP/KRAS + YAP group rats were assessed on the basis of histological characters determined by H&E, CK7, CK20, Ki-67, and E− and N-cadherin expressions that were related to cancer metastasis (Figure 5). For liver metastasis, the tumor cells surrounded by normal liver cells were stained with an anti-hepatocyte-specific antigen, anti-Hep-Par1. H&E staining showed tubular, small-sized duct-like irregular gland structures with marked nuclear atypia close to the bile duct in the liver. These cells were positively stained for CK7 and CK20 by a marked number of Ki-67-positive cells. Both E- and N-cadherin stained positive (Figure 5A). Subcutaneous metastatic tumors showed a solid white-yellowish mass with small to middle-sized irregular duct-like structures significantly stained positive with Ki-67, CK7, CK20, and N-cadherin (Figure 5B). Lymph node metastasis showed tubular, middle-sized duct-like structures with papillary formations surrounded by lymphoid tissue. The tumor showed marked nuclear atypia stained significantly positive for Ki-67, and the cells were stained positive for CK7, 20, and N-cadherin (Figure 5C). Direct invasive growth toward the nervous plexus surrounding the pancreas showed papillary formation with severe atypical nuclear formation, which stained positive for CK7, CK20, and both E- and N-cadherin. Tumor cells closely attached to and surrounded by nervous tissues stained positive with S-100 (Figure 5D). These results suggested that PDAC developed from repeat HGD of KRAS + YAP transfer achieved metastatic tumor development in the pancreas following malignant transformation by N-cadherin upregulation due to the gene combination. Subcutaneous and lymph node metastases showed a weakly stained E-cadherin, which might be related to the epithelial-mesenchymal transition effect (Figures 5B and 5C).

Figure 5

Metastatic and invasive lesions developed in the rat pancreatic cancer model

Representative macroscopic and microscopic images of metastatic tumors found in KRAS + YAP/KRAS + YAP gene-delivered rats. HE, Ki-67, CK7, CK20, E-cadherin, N-cadherin, HSA, and S100 staining. (A) Liver metastasis. (B) Subcutaneous metastasis. (C) Lymph node metastasis. (D) Direct invasion of the pancreatic tumor to the nervous plexus. HE, hematoxylin and eosin, CK7, Cytokeratin 7, CK20, Cytokeratin 20, HSA, hepatocyte specific antigen. Scale bar represents 100 μm.

Effect of YAP on KRASG12D-transferred pancreas

To determine the effect of YAP on malignant transformation in the KRAS-transferred pancreas, signal transduction was examined in vitro. Panc 10.05 was transfected with either a mock or YAP-expressing plasmid. YAP expression induced TGF-β1 and N-cadherin expressions and decreased E-cadherin expression (Figure 6A). The rats that received single KRAS gene delivery followed by the second NS injection (KRAS/NS) showed a higher increase of E-cadherin than that of the control (NS/NS) or KRAS/KRAS, KRAS/YAP, or KRAS + YAP/KRAS + YAP groups. However, the N-cadherin showed an increase in KRAS/YAP and KRAS + YAP/KRAS + YAP rat groups. These results suggest that KRAS gene transfer induces E-cadherin expression, and additional YAP expression could cause a cadherin switch to N-cadherin and promote epithelial-mesenchymal transition leading to increased malignant potential and frequency of metastatic lesions (Figures 6B and 6C). No significant increase or difference in serum aspartate transaminase (AST), alkaline phosphatase (ALP), amylase (AMY), or lactate dehydrogenase (LDH) was observed upon euthanasia (Figure 7).

Figure 6

Effect of YAP on Cadherin protein expression

(A) Expressions of YAP, TGF-β1, E-cadherin, N-cadherin, and β-actin in mock or YAP-expressing plasmid transfected Panc-10.05 cells. (B) Representative histological images of E- and N-cadherin staining in rat groups. (C) Five different sections from each of the six rats in all groups (n = 30) were quantitatively analyzed for positively stained cells for E- and N-cadherin using ImageJ software. The values represent mean ± SEM, ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001, one-way ANOVA with post hoc Tukey’s test.

Figure 7

Serum biochemical analyses

Serum biochemical levels of aspartate transaminase (AST), alkaline phosphatase (ALP), amylase (AMY), and lactate dehydrogenase (LDH) upon the euthanasia. The values represent mean ± SEM, ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001, one-way ANOVA with post hoc Tukey’s test.

Discussion

The pancreas consists of acinar, ductal, and pancreatic islet cells, and the ADM is considered to be a major precancerous lesion of PDAC.,20, 21, 22,34, 35, 36 KRAS activates YAP and TAZ,, which induces continuous ADM followed by PDAC development.,, Considering that overexpression of oncogenic genes in the pancreas is necessary to establish PDAC animal models, the complex vascular structures surrounding the pancreas were major concerns for pancreatic gene delivery methods in the past, compared to systemic HGD. Such a technical barrier can be overcome by pancreas-targeted HGD with gene injection into the superior mesenteric vein, a major vein linking to pancreatic veins that lead venous blood to the portal vein. This pancreas-targeted HGD procedure was previously shown by our group to be pancreas specific, safe, and efficient.

In this study, utilizing this procedure, we demonstrated that oncogene delivery to the pancreas led to the development of PDAC models in wild-type rats. The combination and the dosage of the oncogene were related to tumor development. Specifically, a high dosage of KRAS in the ductal structures showed PanIN transformation and highly atypical intraepithelial cells, which was supported by previous evidence. The acinar structure showed ADM via KRAS transfer, and when combined with YAP activation, further malignant transformation to PDAC was marked within 5 weeks of initiation of genes in wild-type rats. This was evidenced by the fact that acinar cells in the pancreas show transformation into ductular cells (ADM) with KRAS mutant activation., In addition, the combination of KRAS and YAP genes and the increased dosage by repeat injection in the KRAS + YAP/KRAS + YAP group showed transformation from ADM to a significantly malignant PDAC sequence with a high rate of metastatic lesions. Our protein expression assay evidenced activation of YAP expression following KRAS transfer. These results were supported by evidence showing that YAP contributes to the occurrence of ADM and PDAC, and contributes to the cancer metastasis.,

In addition, YAP induced EMT, via the cadherin switch, which refers to expression changes from E-cadherin to N-cadherin that occur during malignant transformation of cancer cells. Our results supported that an in vivo switch in KRAS- and YAP-transferred rats induced metastatic tumors. The fact that the YAP single transfer showed no significant tumor development or metastatic changes suggests that the malignant induction of KRAS and activation of the pathway through YAP addition is important in tumor development. These results supported that pancreatic HGD of the oncogene with appropriate dosage achieved pancreatic cancer model development in the wild-type animal. Further studies are needed to apply this method to other cancer types and animal models. For this purpose, the strategy we reported in various animal species and tissues38, 39, 40, 41, 42 could be considered. The limitations of our study involve the mechanisms of transformation of the gene delivery in normal acinar and pancreatic ductular cells; further studies should include single-cell assays to determine cell type-based phenotypic changes. Such studies will identify effective gene combinations for model development and strengthen the usefulness of this animal model for novel therapy development and biomarker establishment. Additionally, modified procedures might be applied to develop a mouse pancreatic cancer model to improve cost effectiveness.

In summary, pancreas-targeted oncogene HGD induced pancreatic cancer models within 5 weeks in wild-type rats. The tumor occurrence efficacy of this approach depended on the combination and dosage of genes. With molecular signaling activation, the malignant tumor potential increased and exhibited metastatic lesions partly through the cadherin switch. This animal model will speed up pancreatic cancer research for the establishment of the novel treatment strategies and markers for early diagnosis.

Materials and methods

Animals

Animal experiments were approved by and conducted in full compliance with the regulations of the Institutional Animal Care and Use Committee at Niigata University, Niigata, Japan (SA00423). Wistar rats (n = 100, female, 200–250 g) were purchased from Japan SLC (Hamamatsu, Shizuoka, Japan). Rats were housed under the standard conditions of a 12-h light/dark cycle, temperature of 20°C–23°C, and humidity of 45%–55% in pathogen-free facilities. Three rats were housed in a cage and provided with ad libitum access to food and water. Rats were given pancreas-targeted hydrodynamic injections as previously described. Under anesthesia using a combination anesthetic prepared with 0.3 mg/kg medetomidine, 2.0 mg/kg midazolam, and 2.5 mg/kg butorphanol, the portal vein in the hilus and superior mesenteric vein were dissected out and isolated. The catheter (SURFLO 22G, Terumo, Shibuya-ku, Tokyo, Japan) was inserted into the superior mesenteric vein with temporary occluding of the blood flow at the portal vein by vessel loops, and the plasmid DNA solution in a volume of 2% body weight (5 μg/mL in saline, 20-μg plasmid in 4 mL for a 200-g rat) was hydrodynamically injected at a flow rate of 1 mL/s (Figure 1). For the HGD of a combination of plasmids, equal amounts of individual plasmids were prepared in a volume of 2% body weight. Therefore, in a 200-g rat simultaneously receiving two plasmids, 20 μg of each plasmid was diluted in 4 mL of saline solution.

Cells

Human pancreatic adenocarcinoma BxPC-3 (CRL-1687) and Panc 10.05 (CRL-2547) cell lines and human pancreatic duct epithelial cell line hTERT-HPNE (CRL-4023) were purchased from American Type Culture Collection (ATCC, Manassas, VA, USA) and cultured in minimum essential medium containing 10% fetal bovine serum, 100 IU/mL of penicillin, and 100 μg/mL of streptomycin. Cells were incubated in a 5% CO2 humidified incubator at 37°C.

Plasmids

The YAP-expressing plasmid was constructed using full-length cDNA of human YAP ligated into XbaI restriction sites of the expression vector of the pLIVE vector (Mirus Bio, Madison, WI, USA). The KrasWT-expressing plasmid (pCMV6-Entry-KRAS, MR201779), Myc-expressing plasmid (pCMV6-Entry-KRAS G12D, MR201779), and KrasG12D-expressing plasmid (pCMV6-Entry-KRASG12D, RC400104) were purchased from OriGene (OriGene Technologies, Rockville, MD, USA). Either a mock or YAP-expressing plasmid were transfected into the Panc 10.05 cell using FuGENE HD Transfection Reagent (Promega, Madison, WI, USA) following the manufacturer’s instructions and were harvested 48 h after transfection.

Western blotting

For western blotting analysis, culture cells and a tissue lysate were suspended in phosphate-buffered saline and mixed with equal volumes of lysis buffer, 0.125 M tris-HCl (pH 6.8), 10% sucrose, 10% sodium dodecyl sulfate (SDS), 10% 2-mercaptoethanol, and 0.004% bromophenol blue. The extract was subjected to 12% SDS-polyacrylamide gel electrophoresis and blotted onto Hybond membranes (GE Healthcare Life Sciences, Pittsburgh, PA, USA). The membranes were blocked by EzBlockChemi (AE-1475; ATTO Corporation Taito-ku, Tokyo, Japan) for 1 h at room temperature and probed with primary antibodies overnight at 4°C. The antibodies used were mouse anti-CTGF antibody (sc-365970, Santa Cruz Biotechnology, Dallas, TX, USA) at 1:100 dilution; mouse anti-β-actin antibody (ab8227, Abcam, Cambridge, UK) at 1:2000 dilution; mouse anti-E-cadherin antibody (ab76055, Abcam) at 1:1,000 dilution; rabbit anti-Ras (mutated G12D) antibody (ab221163, Abcam) at 1:1,000 dilution; rabbit anti-Akt antibody (ab179463, Abcam) at 1:250 dilution; rabbit anti-pAkt antibody (ab192623, Abcam) at 1:500 dilution; rabbit anti-Erk antibody (No. 4695, Cell Signaling Technology, Danvers, MA, USA) at 1:250 dilution; rabbit anti-pErk antibody (No. 4370, Cell Signaling Technology) at 1:500 dilution; rabbit anti-YAP antibody (No. 4912, Cell Signaling Technology) at 1:100 dilution; rabbit anti-TGF-β1 antibody (ab92486, Abcam) at 1:500 dilution; and rabbit anti-N-cadherin antibody (ab18203, Abcam) at 1:1,000 dilution. Then, the membrane was incubated with secondary antibodies of anti-mouse (NA931-1ML; GE Healthcare Life Sciences) or anti-rabbit (NA934-1ML; GE Healthcare Life Sciences) antibody conjugated with horseradish peroxidase. Protein bands were visualized using the ECL plus Western Blotting Detection System (GE Healthcare Life Sciences).

Histological analysis

Tissue samples for immunohistochemical staining were collected at the appropriate time points after the procedures and then fixed in 10% formalin upon tissue collection before embedding in paraffin. Five sections from the pancreas or tumors (10 μm) were collected from each of the 10 rats in the group, and standard H&E staining and immunohistochemistry were performed. For immunohistochemical staining, anti-MUC5AC antibody (ab3649, Abcam) at 1:200 dilution; anti-CTGF antibody (sc-365970, Santa Cruz Biotechnology) at 1:1,000 dilution; anti-CK7 antibody (ab9021, Abcam) at 1:250 dilution; anti-E-cadherin antibody (ab76055, Abcam) at 1:1,000 dilution; and anti-hepatocyte specific antigen antibody (sc-58693, Santa Cruz Biotechnology) at 1:50 dilution with Vectastain Elite ABC mouse IgG kit (PK-6102, Vector Laboratories, Burlingame, CA, USA) and DAB chromogen tablets (Muto Pure Chemicals, Tokyo, Japan); anti-Ki67 antibody (ab15580, Abcam) at 1:1,000 dilution; anti-Ras (mutated G12D) antibody (ab221163, Abcam) at 1:250 dilution; anti-pAkt antibody (ab179463, Abcam) at 1:100 dilution; anti-TGF-β1 antibody (ab92486, Abcam) at 1:100 dilution; anti-pErk antibody (No. 4370, Cell Signaling Technology) at 1:250 dilution; anti-YAP antibody (No. 4912, Cell Signaling Technology) at 1:250 dilution; anti-CK20 antibody (ab76126, Abcam) at 1:100 dilution; anti-N-cadherin antibody (ab18203, Abcam) at 1:1,000 dilution; and anti-S100 antibody (ab34686, Abcam) at 1:500 dilution with Vectastain Elite ABC rabbit IgG kit (PK-6101, Vector Laboratories) and DAB chromogen tablets (Muto Pure Chemicals, Tokyo, Japan) were employed. Then, images of each tissue section were captured randomly, and quantitative analysis was performed with ImageJ software (version 1.6.0_20; National Institutes of Health, Bethesda, MD, USA) with RGB-based protocol, as reported previously.

Biochemical analysis

Blood samples were collected from each rat before euthanasia, and serum biochemical analyses of AST, ALP, AMY, and LDH were performed (Oriental Yeast, Shiga, Japan). The serum level of carbohydrate antigen 19-9 (CA19-9) was analyzed by enzyme-linked immunosorbent assay using Rat Sialylated Lewis a/CA 19-9 ELISA Kit (LS-F27918, LSBio, Seattle, WA, USA).

Statistical analyses

Histological and serum biochemical factors were statistically evaluated by one-way analysis of variance (ANOVA) followed by Bonferroni’s multiple comparison test using GraphPad Prism7 software (version 7.04; MDF, Tokyo, Japan). p ≤ 0.05 denoted statistical significance.

Data availability statement

The authors confirm that the data supporting the findings of this study are available within the article.