Abstract
GPRC5A is an orphan G-protein coupled receptor with an intriguing dual behavior, acting as an oncogene in some cancers and as a tumor suppressor in other cancers. In the pancreatic cancer context, very little is known about GPRC5A. By analyzing messenger RNA (mRNA) expression data from 675 human cancer cell lines and 10 609 samples from The Cancer Genome Atlas (TCGA) we found that GPRC5A's abundance in pancreatic cancer is highest (cell lines) or second highest (TCGA) among all tissues and cancer types. Further analyses of an independent set of 252 pancreatic normal and cancer samples showed GPRC5A mRNA to be more than twofold upregulated in primary tumor samples compared with normal pancreas (P-value<10(-5)), and even further upregulated in pancreatic cancer metastases to various organs (P-value=0.0021). Immunostaining of 208 cores (103 samples) of a tissue microarray showed generally low expression of GPRC5A protein in normal pancreatic ductal cells; on the other hand, in primary and metastatic samples, GPRC5A protein levels were dramatically increased in pancreatic ductal cells. In vitro studies of multiple pancreatic cancer cell lines showed that an increase in GPRC5A protein levels promoted pancreatic cancer cell growth and migration. Unexpectedly, when we treated pancreatic cancer cell lines with gemcitabine (2',2'-difluorodeoxycytidine), we observed an increase in GPRC5A protein abundance. On the other hand, when we knocked down GPRC5A we sensitized pancreatic cancer cells to gemcitabine. Through further experimentation we showed that the monotonic increase in GPRC5A protein levels that we observe for the first 18 h following gemcitabine treatment results from interactions between GPRC5A's mRNA and the RNA-binding protein HuR, which is an established key mediator of gemcitabine's efficacy in cancer cells. As we discovered, the interaction between GPRC5A and HuR is mediated by at least one HuR-binding site in GPRC5A's mRNA. Our findings indicate that GPRC5A is part of a complex molecular axis that involves gemcitabine and HuR, and, possibly, other genes. Further work is warranted before it can be established unequivocally that GPRC5A is an oncogene in the pancreatic cancer context.