First discovered as naturally occurring by Sir Henry Dale and his colleague Arthur Erwins in 1913 and subsequently characterized as a neurotransmitter by Otto Loewi in the 1920s [37], ACh is the principal muscarinic receptor ligand. Neuronal production of ACh is catalyzed principally by the enzyme cytoplasmic choline acetyltransferase (ChAT) [32]. Due to the efficiency of acetylcholinesterases (AChE) at synaptic junctions, neuronal ACh is quickly hydrolyzed into choline, acetate and water, and, therefore, its effects are likely limited to synaptic transmission. However, the presence of ChAT and production of ACh by non-neuronal tissue was first reported in the 1930s [38]. Subsequently, ChAT expression has been reported in multiple cell types [25]. An additional mechanism of ACh production utilizes the widely-expressed enzyme carnitine acyltransferase (CrAT).

Using quantitative real-time PCR, our group demonstrated that H508, WiDr, and Caco-2 human colon cancer cell lines all express ChAT, that H508 and Caco-2 cells release ACh into cell culture media, and that muscarinic receptor antagonists inhibit proliferation of unstimulated H508 cells by about 40% [39]. Immunohistochemical staining of surgical specimens revealed that normal colonic epithelium has weak or no staining for ChAT, whereas half of the colon cancer specimens examined exhibited moderate to strong staining [39]. Collectively, these findings provide strong evidence that ACh is an autocrine growth factor in colon cancer.

Bile acids are produced by hepatocytes and secreted into bile, predominantly after conjugation (amidation) with taurine or glycine [40]. As illustrated in Figure 2, primary human bile acids derive from cholesterol metabolism following a series of steroid nucleus hydroxylations. These primary bile acids, cholic and chenodeoxycholic acids, are secreted into the small intestine where bacterial dehydroxylases, primarily from Clostridia species, modify these molecules to form the secondary bile acids, deoxycholic and lithocholic acids, respectively. Both primary and secondary bile acids are absorbed in the distal ileum by a highly efficient transporter on the apical surface of enterocytes (apical sodium-dependent bile acid transporter, commonly referred to as ASBT). Via the portal circulation, bile acids are transported back to the liver where secondary bile acids undergo conjugation (amidation) with either taurine or glycine, and are secreted into the biliary tree and intestines. Reduced expression or inactivation (mutation) of ASBT results in reduced bile acid up-take from the distal small intestine and increased fecal bile acids. The bile acid nuclear receptor Farnesoid X Receptor (FXR), a key regulator of bile acid homeostasis, is expressed widely in the GI tract. The actions of FXR and its possible role in colon cancer have recently been reviewed [41].

Novel observations in our lab revealed the interaction of bile acids with muscarinic receptors. In the course of examining actions of bile acids on pepsinogen secretion by gastric chief cells we found that taurine conjugates of lithocholic acid bind to muscarinic receptors, increase inositol phosphates (IP), and stimulate secretion by a cholinergic mechanism [8]. In collaboration with the Frucht lab, we revealed the functional interaction of lithocholic acid conjugates with CHRM3 receptors on a human colon cancer cell line [34]. Specifically, in H508 colon cancer cells which express CHRM3, lithocholytaurine (LCT) was shown to stimulate dose-dependent increases in cell proliferation up to 200% compared to control, dose-dependent inhibition of radioligand binding, and increased IP formation. No such changes were observed in a colon cancer cell line (SNU-C4) that does not express CHRM3 [34]. Molecular modeling suggests that the functional interaction of bile acids with muscarinic receptors is due to shape and surface charge similarities to ACh [36].

Using H508 human colon cancer cells with high-level expression of CHRM3 [42] we identified bile acids as CHRM3 agonists and revealed that, as with ACh [43], proliferative actions of bile acids require cross-talk between CHRM3 and epidermal growth factor receptors (EGFR) and post-receptor ERK1/2 activation [44,45]. Bile acids stimulated proliferation of H508 cells that co-express CHRM3 and EGFR, but did not alter proliferation of SNU-C4 cells that express EGFR but not CHRM3 [44]. The requirement for CHRM3 activation was confirmed using chemical inhibitors of muscarinic receptor activation [44,45]. Transactivation of EGFR by GPCR (e.g., CHRM3) is a common feature of mitogenic signaling [46]. Work detailed in [44] reveals that: (1) In H508 cells that co-express CHRM3 and EGFR, bile acids stimulate cell proliferation. (2) Bile acids do not alter proliferation of SNU-C4 cells that express EGFR but not CHRM3. (3) Proliferative actions of bile acids are inhibited by CHRM3 antagonists. (4) Rapid and reversible post-EGFR ERK activation is detected with micromolar levels of bile acids that stimulate cell proliferation [45] and were detected by us in the human cecum [47]. Collectively, these findings define efficacious bile acid concentrations, emphasize the importance of co-expression of CHRM3 and EGFR for bile acid-induced colon cancer cell proliferation, and identify the key role of post-EGFR ERK signaling.

EGFR actions on cell proliferation and survival are mediated by several signaling cascades. Bile acids protect colon cancer cells from apoptosis by EGFR-, PI3K/AKT-dependent mechanisms that involve activation of NF-κB [48]. Treatment with bile acids increases resistance of colon cancer cells to TNF-α- and UV-induced apoptosis, and stimulates nuclear translocation and transcriptional activity of NF-κB [49]. Reduced activation of NF-κB using an IκBα super-repressor (AdIκBSR) or chemical inhibitors attenuates anti-apoptotic actions of bile acids [49]. Bile acid-induced NF-κB activation and rescue from apoptosis is regulated by PI3K/AKT signaling downstream of EGFR [49]. Bile acid-induced resistance to TNF-α- and UV-stimulated apoptosis requires activation of AKT; both NF-κB activation and anti-apoptotic actions of bile acids were attenuated when AKT expression and activation were reduced by transfection with mutant AKT or treatment with an AKT inhibitor, respectively [49]. These observations in H508 cells were confirmed in another human colon cancer cell line (HT-29 cells) [49]. Collectively, these findings demonstrate that, downstream of EGFR activation, bile acid-induced PI3K/AKT and NF-κB activation regulate colon cancer cell apoptosis and survival. These actions of bile acids are likely to be important for colon cancer cell resistance to chemo- and radiation-therapy.

Matrix metalloproteinases (MMPs) catalyze EGFR ligand release. Work detailed in [45] reveals that MMP7 mediates bile acid-induced activation of EGFR by catalyzing release of HBEGF (Figure 3). Bile acid-induced H508 cell proliferation is blocked by anti-HBEGF antibody, by CRM197, an agent that blocks HBEGF release, and by GM6001, a non-selective MMP inhibitor [45]. Recombinant MMP7 and HBEGF both mimic proliferative actions of bile acids. Using q-PCR, immunohistochemistry and ELISA, we showed that H508 colon cancer cells express MMP7 and that MMP7 gene transcription is induced by bile acids. The half-life of MMP7 mRNA (∼16 h) was not altered indicating that MMP7 expression is stimulated at the transcriptional level [50]. Proliferative actions of bile acids are blocked by anti-MMP7 antibody and knockdown with MMP7 siRNA [45]. Collectively, these findings identify molecular mechanisms leading from bile acid interaction with CHRM3 to activation of EGFR (Figure 3). The role of muscarinic receptor ligand induction of MMP1 and 10 gene transcription remains to be determined.

To investigate the importance of CHRM3 expression in vivo we utilized the azoxymethane (AOM) model of colonic neoplasia in wild type (WT) and Chrm3-deficient (Chrm3^−/−) mice [31]. When compared to AOM-treated WT mice, AOM-treated Chrm3^−/− mice revealed 40% reduction in colon tumor number and 60% reduction in tumor volume. Additionally, cell proliferation (determined by BrdU staining) was reduced by 43% in Chrm3^−/− compared to WT mice. These findings confirmed the importance of CHRM3 receptor expression in colon neoplasia. To determine the therapeutic potential of muscarinic antagonists in human colon cancer, additional studies, including investigations of specific CHRM3 antagonists and human trials, are required.