Suppression of mutant Kirsten-RAS (KRASG12D)-driven pancreatic carcinogenesis by dual-specificity MAP kinase phosphatases 5 and 6

The cytoplasmic phosphatase DUSP6 and its nuclear counterpart DUSP5 are negative regulators of RAS/ERK signalling. Here we use deletion of either Dusp5 or Dusp6 to explore the roles of these phosphatases in a murine model of KRASG12D-driven pancreatic cancer. By 56-days, loss of either DUSP5 or DUSP6 causes a significant increase in KRASG12D-driven pancreatic hyperplasia. This is accompanied by increased pancreatic acinar to ductal metaplasia (ADM) and the development of pre-neoplastic pancreatic intraepithelial neoplasia (PanINs). In contrast, by 100-days, pancreatic hyperplasia is reversed with significant atrophy of pancreatic tissue and weight loss observed in animals lacking either DUSP5 or DUSP6. On further ageing, Dusp6−/− mice display accelerated development of metastatic pancreatic ductal adenocarcinoma (PDAC), while in Dusp5−/− animals, although PDAC development is increased this process is attenuated by atrophy of pancreatic acinar tissue and severe weight loss in some animals before cancer could progress. Our data suggest that despite a common target in the ERK MAP kinase, DUSP5 and DUSP6 play partially non-redundant roles in suppressing oncogenic KRASG12D signalling, thus retarding both tumour initiation and progression. Our data suggest that loss of either DUSP5 or DUSP6, as observed in certain human tumours, including the pancreas, could promote carcinogenesis.

routinely tested for mycoplasma contamination (MycoAlert #LT07-118; Lonza, Basel, CH). Transfection of MEFs with either Ad5-CMV empty or Ad5-CMVCre vectors (Vector Core, University of Iowa, USA), assays of cell proliferation, clonogenicity and analysis of p-ERK staining and EdU incorporation using highcontent microscopy were as described previously (4). For all experiments using cultured cells technical replicates were performed in triplicate and, unless stated otherwise in the figure legends, experiments were repeated 3 times.
Tumour cell and acinar cell 3D culture. To isolate primary murine pancreatic adenocarcinoma cell lines, tumour tissue (300 mg) was diced in 0.5 ml ice cold PBS before addition of 5 ml 1 mg/mL collagenase V in DMEM. Following incubation at 37°C for 45 min, 5 ml of PDAC medium (DMEM supplemented with 20% FBS, 4.5 gl glucose, 1 mM L-glutamine, 0.11 g/l pyruvate, 50 mg/ml gentamycin, 100 U/ml penicillin and 100 mg/ml streptomycin) was added and centrifuged at 350 x g for 5 min. The partly digested tumour was then resuspended in 2 ml 0.05% trypsin/EDTA and incubated for 5 min at 37°C. The trypsin was then quenched with 2 ml PDAC medium, the solution centrifuged at 350g for 5 min and the cell pellet washed in PBS before being resuspended in 6 ml PDAC medium and seeded into a 25 cm 2 cell culture flask. After 7 days, cells were transferred into DMEM, supplemented with 10% FBS, 100 mM sodium pyruvate, 2.5 mM L-Glutamine and 100 U/ml each of Penicillin and Streptomycin and cultured for experimental use.
3D acinar cell culture was performed according to the method of Figura et al. (5) Briefly, pancreata from either wild type (WT); n = 7 or Dusp5M -/-; n = 6 mice were isolated and digested with collagenase P (0.4 mg/ml, Roche 11213857001) in HBSS for 10 min at 37 °C with gentle agitation. Digestion was then stopped by the addition of 5% FBS/HBSS. The cells were washed twice with 5% FBS/HBSS and filtered through a 100 µm cell strainer. The cell suspension was layered over a 30% FBS/HBSS solution to allow intact cell clusters to precipitate by low-speed centrifugation at 180 x g. The acinar cell clusters were resuspended in growth factor reduced Matrigel (BD 354230) with an equal volume of 3D culture media (Waymouth's media, 10% FBS, 100 U/ml penicillin, 100 µg/ml streptomycin, 2 mM L-glutamine, 100 µg/ml soybean trypsin inhibitor, 1 µg/ml dexamethasone) and seeded into rat-tail collagen-1 (BD 354236) coated wells of a 24-well plate. Cells were then covered with an additional layer of collagen-1 and 3D culture media was added. The transdifferentiation process was evaluated morphologically over time, based on the conversion of acinar cell clusters to ductal cyst structures that were comprised of a single layer of epithelial cells surrounding an empty luminal space. For each animal and time point, the mean rate of transdifferentiation in 4 optical fields was determined.
Tissue sectioning and histology. Organs were fixed, embedded, sectioned and stained with H & E or analysed by immunohistochemical (IHC) staining as previously described (1). Quantitative analysis of all H & E and IHC staining was performed blinded. Alcian blue staining was performed following rehydration for 30 min in Alcian blue solution (Sigma) before counterstaining with Nuclear Fast Red (Sigma) for 5 min. Slides were then dehydrated through increasing concentrations of ethanol (70%, 90%, 95% & 100% x 2 -2 min each) and cleared in xylenes (2 x 10 min) before mounting under coverslips using Omnimount (National Diagnostics, Atlanta, GA). All slides required for imaging or quantitative analysis were scanned on a Leica Biosystems Aperio XT slide scanner and quantitative analysis was performed using Aperio ImageScope software (Leica Biosystems, Wetzlar, Germany).

RNAscope analysis.
In situ hybridisation detection of Dusp6 (Advanced Cell Diagnostics, Hayward, CA; 429328) mRNA was performed using RNAscope 2.5 LS (Brown) detection kit (Advanced Cell Diagnostics, Hayward, CA; 322100) on a Leica Bond Rx autostainer according to the manufacturer's instructions Leica Biosystems, Wetzlar, Germany). Staining was performed on 4um formalin-fixed paraffin-embedded sections, which were cut and placed in a 60°C oven 2 h prior to staining. In order to verify RNA integrity in the tissue, a positive control probe (mm-Ppib, 313918) was used. The tubulin blot is shown as a representative loading control (p-ERK1/2/ERK1/2). For all quantitative data individual experimental points and mean values from 3 independent experiments (n = 3) ± SEM are shown, ns = not significant, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 using one-way ANOVA and Bonferroni post hoc test.  Quantification of the percentage acinar tissue remaining in either 56 (C) or 100day (F) age-matched pancreata of the indicated cohorts following KRAS G12Ddriven ADM and PanIN initiation. Quantification was performed on one representative section per mouse, following serial sectioning of the pancreas. Mean values are shown, n = 7 mice per cohort, ns = not significant, **P < 0.01, ****P < 0.0001, using one-way ANOVA and Bonferroni post hoc test. Mean values ± SEM are shown, n = 7 mice per cohort, ns = not significant, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 using one-way ANOVA and Bonferroni post hoc test. Please note that due to concomitant recruitment of both the KCSB2 -/and KCDKO mice into the aged cohorts the quantitative data pertaining to outcomes in KC, KCD5 +/and KCD5 -/mice shown in D, E and F is the same as that presented in Figure 2 (E-G).   Logrank test, ns = not significant, **P < 0.01, ***P < 0.001. , ns = not significant, **P < 0.01,****P < 0.0001 using a 2 x 2 contingency table analysed by Fisher's exact test with a two-tailed P value. (B) Images of isolated pancreata from 2 mice of the indicated genotype demonstrating the reduction in size of KCD5 -/pancreata, relative to KC. (C) Quantification of the pancreas to body weight ratio of the indicated cohorts. The KCD5 -/and KCD6 -/-WL cohorts represent all the mice of these genotypes that were included in the survival study, but which had to be removed following dramatic weight-loss, and did not present with PDAC. KCD5 -/-WL cohort median age, 94 days; KCD6 -/-WL cohort median age, 113 days. Mean shown, n = 5, 4, 6, 9, 6, 7, 10, 7, 7, 9. (D) A representative image of an H&E stained pancreatic tissue section from a KCD5 -/-WL cohort animal. (Scale bars, 2mm.) (E) Quantification of the percentage acinar tissue remaining in the pancreata of the indicated 100-day age-matched cohorts alongside the KCD5 -/-WL and KCD6 -/-WL cohorts. Mean shown, n = 9, 7, 7, 12, 7, 8, 7, ns = not significant, **P < 0.01, ***P < 0.001,****P < 0.0001, using oneway ANOVA and Bonferroni post hoc test.