GPR109A is a G-protein-coupled receptor for the bacterial fermentation product butyrate and functions as a tumor suppressor in colon

Abstract

Short-chain fatty acids, generated in colon by bacterial fermentation of dietary fiber, protect against colorectal cancer and inflammatory bowel disease. Among these bacterial metabolites, butyrate is biologically most relevant. GPR109A is a G-protein-coupled receptor for nicotinate but recognizes butyrate with low affinity. Millimolar concentrations of butyrate are needed to activate the receptor. Although concentrations of butyrate in colonic lumen are sufficient to activate the receptor maximally, there have been no reports on the expression/function of GPR109A in this tissue. Here we show that GPR109A is expressed in the lumen-facing apical membrane of colonic and intestinal epithelial cells and that the receptor recognizes butyrate as a ligand. The expression of GPR109A is silenced in colon cancer in humans, in a mouse model of intestinal/colon cancer, and in colon cancer cell lines. The tumor-associated silencing of GPR109A involves DNA methylation directly or indirectly. Reexpression of GPR109A in colon cancer cells induces apoptosis, but only in the presence of its ligands butyrate and nicotinate. Butyrate is an inhibitor of histone deacetylases, but apoptosis induced by activation of GPR109A with its ligands in colon cancer cells does not involve inhibition of histone deacetylation. The primary changes in this apoptotic process include down-regulation of Bcl-2, Bcl-xL, and cyclin D1 and up-regulation of death receptor pathway. In addition, GPR109A/butyrate suppresses nuclear factor-kappaB activation in normal and cancer colon cell lines as well as in normal mouse colon. These studies show that GPR109A mediates the tumor-suppressive effects of the bacterial fermentation product butyrate in colon.

Figure 1

Expression of GPR109A in colon. A, Expression of GPR109A mRNA in mouse intestinal tract. B, Expression of GPR109A and GPR109B mRNA in human colon and in human colon cell lines. C, Immunolocalization of GPR109A protein in mouse intestinal tract (GPR109A, red; nuclei, blue). D, Immunolocalization of GPR109A and GPR109B in human colonic biopsies (GPR109A & GPR109B, red; nuclei, blue).

Figure 2

Silencing of GPR109A in colon cancer. A, Expression of GPR109A and GPR109B in normal colon and paired colon cancer tissues from 18 patients as assessed by RT-PCR. B, Expression of GPR109A in the colon and intestine of wild type and Apc^Min/+ mice. In Apc^Min/+ mice, the tissues were collected from tumor sites and also from areas with no tumor. C, Comparison of expression of GPR109A and GPR109B by RT-PCR between normal colon cell lines (NCM460 and CCD841) and cancer colon cell lines.

Figure 3

Involvement of DNMT1 in the silencing of GPR109A. A, Analysis of expression of GPR109A and GPR109B by RT-PCR in the human colon cancer cell line HCT116 which expresses all three isoforms of DNMT (WT), and in isogenic HCT116 cell lines with targeted deletion of DNMT1 (DNMT1^−/−), DNMT3b (DNMT3b^−/−) or both (DKO). B, Immunocytochemistry for GPR109A protein in WT, DNMT1^−/−, DNMT3b^−/−, and DKO cells. C, Effect of procainamide, a specific inhibitor of DNMT1, on GPR109A expression in colon cancer cell lines.

Figure 4

Induction of apoptosis in colon cancer cells by GPR109A ligands. A, The normal colon cell line CCD841 was transfected with vector or human GPR109A cDNA, and then treated with or without nicotinate (1 mM) or butyrate (1 mM) for 48h. Cells were then used for analysis of apoptosis by FACS. B, The colon cancer cell line KM12L4 was transfected with vector or human GPR109A cDNA, and then treated with or without nicotinate (1 mM) or butyrate (1 mM) for 48h. Cells were then used for analysis of apoptosis by FACS. C, The cell lysates from the experiments described in A and B were used to monitor caspase activation by Western blot using antibodies specific for cleaved fragments of caspases.

Figure 5

Blockade of LPS-induced NF-κB activation by GPR109A in the normal colon cell line CCD841 and in the colon cancer cell lines KM12L4 and HCT116. A, CCD841 cells were first transfected with a NF-κB-luciferase reporter construct. 24h later, cells were treated with LPS (100 ng/mL) for 4h with or without pretreatment with butyrate (But, 1 mM), nicotinate (Nic, 1 mM), or acifran (Aci, 0.25 mM) for 4h. The ligands were present for an additional 4h during treatment with LPS. LPS-induced activation of NF-κB was monitored by measuring the activity of luciferase as a reporter. UT, no treatment with GPR109A ligands. B, KM12L4 cells were transfected with a NF-κB-luciferase reporter together with either vector or GPR109A cDNA. 24h later, cells were treated with LPS (100 ng/mL) for 4h with or without pretreatment with butyrate (But, mM), nicotinate (Nic, 1 mM), and acifran (Aci, 0.25 mM) for 4h. In addition to the pretreatment, the ligands were present also during LPS treatment. LPS-induced activation of NF-κB was monitored by measuring the activity of luciferase as a reporter. C, HCT116 cells (DNMT^+/+ and DNMT1^−/−) were transfected with a NF-κB-luciferase reporter construct. 24h later, cells were treated with LPS (100 ng/mL) for 4h with or without pretreatment with butyrate (1 mM), nicotinate (1 mM), or acifran (0.25 mM) for 4h. The ligands were present also during LPS treatment. LPS-induced activation of NF-κB was monitored by measuring the activity of luciferase. UT, no treatment with GPR109A ligands.