Oncogenic KRAS/ERK/JUNB signaling suppresses differentiation regulator GATA6 in pancreatic cancer
Zheng Zhong, Xinang Cao, Pei-Ju Liao, Raman Sethi, Jeffrey A. Klomp, Clint A. Stalnecker, Jinmiao Chen, Yue Wan, Channing J. Der, David M. Virshup
Zheng Zhong, Xinang Cao, Pei-Ju Liao, Raman Sethi, Jeffrey A. Klomp, Clint A. Stalnecker, Jinmiao Chen, Yue Wan, Channing J. Der, David M. Virshup
View: Text | PDF
Research Article Cell biology Oncology

Oncogenic KRAS/ERK/JUNB signaling suppresses differentiation regulator GATA6 in pancreatic cancer

Abstract

GATA6 is a master regulator of differentiation in the pancreas, and its expression levels determine the 2 main molecular subtypes of pancreatic cancer. High GATA6 levels contribute to the classical pancreatic cancer subtype, which is associated with a higher degree of tumor differentiation and better disease prognosis. However, why GATA6 expression varies across pancreatic cancers and what regulates GATA6 expression remain elusive. Here, we report that oncogenic KRAS-activated ERK signaling suppresses GATA6 transcription in pancreatic cancers. GATA6 mRNA levels inversely correlated with KRAS/ERK activity in pancreatic tumors. A genome-wide CRISPR screen in a GATA6-EGFP reporter knockin cell line identified JUNB as the ERK-regulated transcriptional repressor for GATA6. Active ERK stabilized JUNB protein, while KRAS/ERK inhibition led to ubiquitin-independent proteasomal degradation of JUNB and increased transcription of GATA6. Upregulation of GATA6 enhanced chemosensitivity of pancreatic cancer cells, and KRAS/ERK inhibitors synergized with chemotherapy in a GATA6-dependent manner. Our study identifies how oncogenic KRAS/ERK signaling suppresses GATA6 to cause dedifferentiation in pancreatic cancer. Combining KRAS/ERK inhibitors with standard-of-care chemotherapies could be a promising therapeutic strategy for treating pancreatic cancers.

Authors

Zheng Zhong, Xinang Cao, Pei-Ju Liao, Raman Sethi, Jeffrey A. Klomp, Clint A. Stalnecker, Jinmiao Chen, Yue Wan, Channing J. Der, David M. Virshup

×

Figure 2

Oncogenic KRAS-activated ERK signaling inhibits GATA6 transcription.

Options: View larger image (or click on image) Download as PowerPoint
Oncogenic KRAS-activated ERK signaling inhibits GATA6 transcription. (A)...
(A) MEK inhibition did not stabilize GATA6 protein. HPAF-II cells were treated with 40 μg/mL CHX in the presence or absence of 100 nM trametinib for the indicated time. Quantification (mean ± SD) of GATA6 bands is shown. n = 2 biological replicates/condition. (B) MEK inhibition upregulated GATA6 transcript abundance. HPAF-II cells were treated with 100 nM trametinib for the indicated time. DUSP6 is an ERK target gene, while DAB2 is a GATA6 target gene. n = 2 biological replicates × 2 technical replicates/condition. Data are shown as the mean ± SD. (C) Oncogenic activation of KRAS/ERK signaling downregulated GATA6. Analysis of data from GSE58055 (27). Human pancreatic duct epithelial (HPDE) cells expressing dox-inducible GFP, WT KRAS, or KRASG12D mutant were profiled by microarray. (D and E) KRAS/ERK inhibition upregulated GATA6 in vivo. (D) Mice bearing HPAF-II xenografts were treated with vehicle or trametinib (3 mg/kg) once daily for 3 days, and tumors were harvested 8 h after the last dose for analysis. n = 4–6 tumors per condition each analyzed with technical replicates. P values of 2-tailed, unpaired t test are shown. (E) Analysis of data from GSE201412 (28). Mice bearing HPAC xenografts were treated with vehicle or KRASG12D inhibitor and harvested at the indicated time points. (F and G) KRAS/ERK activity is inversely correlated with GATA6 transcript abundance. Analyses of data from references 31 and 32. Each dot represents a human primary pancreatic cancer (F) or a PDAC PDX (G). Samples were colored based on the Bailey molecular subtype in F or histological differentiation status in G. The R and P values of simple linear regression are shown. (H) Pancreatic cancer cells with high KRAS/ERK activity showed lower GATA6 expression (CRA001160) (34). The P values of Dunn’s multiple-comparison test are shown.