Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Interaction of acetylcholine with members of the muscarinic receptor family (M1-M5) is important for the regulation of a host of physiological actions. Hence, the discovery of other endogenous ligands for these receptors would extend our understanding of mammalian physiology and pathophysiology. In addition to their direct biochemical roles, metabolites often have signaling functions. Classical feedback loops are an example. Although feedback regulation is primarily involved in maintaining controlled levels of a given metabolite, such molecules may also assume functions more distant or even unrelated to regulation of their own synthesis.

In the course of examining the actions of human bile acids on guinea pig gastric chief cells, we discovered that lithocholyltaurine, the taurine conjugate of lithocholic acid, is a partial muscarinic receptor agonist (Raufman et al., 1998). Chief cells are reported to express the M3 subtype muscarinic receptor (Sutliff et al., 1989; Kajimura et al., 1992). Taurine alone has no muscarinic ligand properties (Raufman et al., 1998). These observations may be explained by the structural similarity of lithocholyltaurine and acetylcholine (Fig. 1A). Both molecules contain a bulky tetrahedral atom, the quaternary ammonium group in acetylcholine and the sulfonic acid moiety in lithocholyltaurine. The tetrahedral atom is linked via two methylene groups to a polar center (ester oxygen or amide nitrogen) and further to a carbonyl group that is identically positioned in both molecules. The overall geometry of the molecules and specific distances between the positively charged center and the two additional polar atoms are very similar (Chothia and Pauling, 1968; Beers and Reich, 1970; Fraenkel et al., 1994; Fraenkel et al., 1996) (Fig. 1, A and B).

View larger version (23K): [in this window] [in a new window]   Fig. 1.   Structures of acetylcholine and lithocholyltaurine, and mass spectrometry of lithocholylcholine. A, structures of acetylcholine and lithocholyltaurine. B, three-dimensional models of acetylcholine and lithocholyltaurine showing predicted free-solution structures and distances between polar centers in angstroms (Vlahcevic et al., 1972; Makishima et al., 1999; Parks et al., 1999; Slack, 2000). The model of lithocholyltaurine was built using the program WebLab ViewerPro (Molecular Simulations, Inc.) on the basis of a cholic acid structure obtained from the Okanagan University server (www.sci.ouc.bc.ca/chem). The structure was visualized using WebLab ViewerPro or Rasmol (Sayle and Milner-White, 1995) on a Silicon Graphics O2 workstation. The structure of lithocholylcholine combines the circled portions of acetylcholine and lithocholyltaurine. C, structure and purity of lithocholylcholine was confirmed by mass spectrometry.

We formulated the hypothesis that the structural similarity between the taurine and choline moieties of lithocholyltaurine and acetylcholine, respectively, in conjunction with the presence of the steroidal nucleus in lithocholyltaurine, explains interaction with muscarinic receptors. Because the bulky steroid nucleus might provide an anchor to help position the choline moiety in the receptor binding site, we considered that acylating choline with lithocholate rather than acetate might produce a potent cholinergic ligand. To test this hypothesis, we synthesized lithocholylcholine, a molecule that combines the steroid nucleus of lithocholic acid with a choline moiety as is present in acetylcholine (Fig. 1B), and examined its actions on Chinese hamster ovary (CHO) cells expressing M1 to M5 muscarinic receptor subtypes. We also examined functional actions of lithocholylcholine on postreceptor signaling and on acetylcholine-induced relaxation of rat aorta.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Materials. Dulbecco's modified Eagle's medium, minimal essential medium nonessential amino acid, penicillin, streptomycin, and G418 were obtained from Invitrogen (Carlsbad, CA). N-Methyl-[³H]scopolamine ([³H]NMS) and [³H]myoinositol were supplied by PerkinElmer Life Sciences (Boston, MA). Cell Signaling (Beverly, MA) provided rabbit polyclonal anti-MAPK and mouse monoclonal anti-phospho-MAPK. Carbachol was purchased from CalBioChem (San Diego, CA). Lithocholylglycine was from Steraloids, Inc (Newport, RI). All other chemicals were obtained from Sigma-Aldrich (St. Louis, MO) or Fisher Scientific Co. (Pittsburgh, PA).

Synthesis of Lithocholylcholine. Lithocholic acid was formylated to 3-formyloxylithocholic acid by methods described previously (Dayal et al., 1991). In the presence of 62 mg of N,N'-dicyclohexylcarbodiimide and 4.4 mg of 4-dimethylaminopyridine, 3-formyloxylithocholic acid (180 mg in N,N-dimethylformamide) was conjugated with 150 mg of choline overnight at room temperature. The resulting formylated lithocholylcholine was precipitated with water, washed with methanol, and dried. Lithocholylcholine was obtained by deformylation in the presence of 5% methanesulfonic acid/methanol (Dayal et al., 1995). The structures of formylated lithocholylcholine and lithocholylcholine were confirmed by mass spectrometry performed by the High Resolution Mass Spectrometry Facility at the University of Iowa (Iowa City, IA).

Acetylcholinesterase Assay. Recombinant acetylcholinesterase (EC3.1.1.7) (Sigma-Aldrich; 3200 units/mg of protein; 0.14 µg of protein/ml of reaction mixture) was incubated with 0.06 µmol of lithocholylcholine, lithocholyltaurine (in 6 ml), or lithocholylglycine (in 0.6 ml) for 1 h at 37°C. As a control, acetylcholinesterase was denatured by boiling for 5 min (heated enzyme). The reaction mixture was dried, extracted into 30 µl of MeOH, and spotted onto Silica Gel G plates (Analtech, Newark, DE). After separation in a chloroform/methanol/acetic acid (21:15:0.75, v/v/v) system, spots were visualized by spraying with half-saturated (NH₄)₂SO₄ solution and heating in an oven at 120°C for 45 min.

Cell Lines. CHO-K1 and CHO cells expressing rat M1 and M3 (rM1 and rM3) subtype muscarinic receptors were obtained from American Type Culture Collection (Manassas, VA). CHO cells expressing human M2 to M5 (hM2-hM5) subtype muscarinic receptors were provided by Drs. Tom Bonner and Mark Brann [National Institute of Mental Health (Bethesda, MD) and Acadia Pharmaceuticals (San Diego, CA), respectively] (Buckley et al., 1989). Cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (BioWhittaker, Walkersville, MD), 1× minimal essential medium nonessential amino acids, penicillin (50 units/ml), and streptomycin (50 µg/ml). The media for M1 to M5 cells also contained 0.1 mg/ml G418.

Cell Toxicity Assays. Before proceeding with these studies, we evaluated potential toxic effects of bile acid derivatives, DMSO used as the solvent, and other test agents using conditions similar to those for the other experiments described in this article by examining their actions on the release of lactate dehydrogenase and exclusion of trypan blue from CHO cells (assay kits; Sigma-Aldrich). At concentrations used in the following experiments, none of these agents altered these two measures of cell damage (data not shown).

Radioligand Binding. Binding of [³H]NMS to CHO cells was determined as described previously (Raufman et al., 1998). Briefly, adherent cells were harvested with 0.05% trypsin and suspended in Krebs-Henseleit buffer (pH 7.4) composed of 1.2 mM MgSO₄, 118 mM NaCl, 4.7 mM KCl, 1.3 mM CaCl₂, 25 mM NaHCO₃, and 1.2 mM glucose. Cells (5 × 10⁵ cells/0.5 ml) were incubated with [³H]NMS (82 Ci/mmol stock solution; final concentration 0.6 nM) alone or with test agents for 45 min at 37°C. Bound and free radioligand were separated by centrifugation. The cell pellet was dissolved in 100 µl of Soluene 350 and counted in a liquid scintillation counter (1214 Rackbeta; PerkinElmer Wallac, Gaithersburg, MD). Nonspecific binding was determined in the presence of 10 µM unlabeled NMS and, in all experiments, was <10% of total binding. Values shown represent binding with radioligand alone (total binding) minus nonspecific binding.

Measurement of Inositol Phosphates. Total inositol phosphates in CHO cells were determined as described previously (Ryan et al., 1998). Briefly, cells were seeded onto 24-well plates (5 × 10⁴ cells/well), preincubated for 24 h at 37°C, and incubated for an additional 24 h with [³H]myoinositol (1 µCi/ml) in growth medium supplemented with 2% fetal bovine serum. Cells were treated with 20 mM LiCl for 30 min, followed by a 30-min incubation at 37°C with the test agents. The reaction was terminated with 1 ml of HCl/methanol (0.14%, v/v), and labeled inositol phosphates were purified by Dowex resin chromatography and measured in a liquid scintillation counter.

cAMP Assay. cAMP accumulation in CHO-hM2 cells was determined as described previously (Birdsall et al., 1999). Briefly, adherent cells were harvested with 0.05% trypsin and suspended in a solution containing 118 mM NaCl, 1.8 mM CaCl₂, 2.7 mM KCl, 0.81 mM MgSO₄, 1.0 mM Na₂HPO₄, 5.6 mM glucose, and 10 mM HEPES (pH 7.4) at a density of 10⁶ cells/ml. Aliquots of the cell suspension (0.5 ml) were incubated with 0.5 mM 3-isobutyl-1-methylxanthine for 5 min at 37°C. The experiment was initiated by the addition of agent alone for 5 min followed by the addition of 5 µM forskolin, with the incubation allowed to proceed for an additional 2.5 min. The reaction was stopped by adding 5 M HCl (final concentration 0.1 M), and the cells were lysed by boiling for 10 min. cAMP levels in the supernatant (0.1 ml) were measured with a direct cAMP enzyme immunoassay kit (Sigma-Aldrich).

Determination of MAP Kinase Phosphorylation. Phosphorylation of p44/p42 MAP kinase was determined by methods described previously (Schmidt et al., 1994). Briefly, cells were subcultured in six-well plates (2 × 10⁵ cells/well). After a 24-h incubation at 37°C, the cells were serum-starved for an additional 24 h, washed with phosphate-buffered saline, and allowed to recover for 1 h at 37°C before adding test agents. After a 10-min incubation with test agents, the reaction was terminated by adding lysis buffer [150 mM NaCl, 10 mM Tris/HCl, 1% (mass/volume) deoxycholic acid, 1% (volume) Nonidet P-40, 0.1% (mass/volume) SDS, 4 mM EDTA, 1 mM Na₃VO₄, 20 µg/ml leupeptin, 20 µg/ml aprotinin, 250 µg/ml p-nitrophenylphosphate, and 1 mM phenylmethylsulphonyl fluoride, pH 8]. Cell lysates were subjected to SDS-polyacrylamide gel electrophoresis (10% gel; Invitrogen). Resolved proteins were transferred electrophoretically to nitrocellulose membranes (Osmonics, Westborough, MA) and blotted with anti-phospho-p44/42 (ERK1/2) MAP kinase. Bound antibody was detected by super-signal chemiluminescence detection (Pierce Chemical, Rockford, IL), and phosphorylation was quantified by densitometry of the resulting lumigrams from the combined phospho-p44/p42 bands. To verify that equal amounts of protein were added to each lane, the blots used with anti-phospho-p44/42 (ERK1/2) MAP kinase were stripped and reprobed with anti-p44/p42 (rows labeled "b").

Preparation of Rat Aortic Rings. Rat thoracic aortic rings were prepared by methods described previously (Stewart and Kennedy, 1999). Briefly, after being anesthetized with a volatile anesthetic, male Sprague-Dawley rats (300-400g) were sacrificed by exsanguination. The abdomen was opened via a midline incision, 100 units of heparin was injected into the inferior vena cava, and the thoracic aorta was dissected and placed in oxygenated Krebs-Henseleit solution. After removal of connective tissue, the aorta was cut into two rings of 3 to 4 mm each. The rings were mounted on two stainless steel hooks and bathed in the oxygenated (saturated with 95% O₂/5% CO₂ gas) Krebs-Henseleit solution (37°C). The tension on each ring was monitored via a strain gauge and recorded. Resting tension was set at 2.0 g, and the rings were exposed repeatedly to 80 mM KCl until a stabile contractile response was obtained. The preparations were then constricted by addition of 10⁵ M phenylephrine (an -adrenoceptor agonist), and the vasodilatory response to 10⁵ M acetylcholine was examined as a measure of endothelial integrity (all preparations responded with at least a 60% decrease in phenylephrine-induced tension). The effect of lithocholylcholine on acetylcholine-induced relaxation was then monitored. One ring from each aorta was incubated with 0.1 mM lithocholylcholine (the other being exposed to an equal volume of diluent) for 15 to 20 min before being treated with 10⁵ M phenylephrine. After the contractile response to phenylephrine reached steady state, concentration response curves for acetylcholine were obtained by cumulative addition (n = 4).

Statistical Analysis. Results are expressed as mean ± S.E.M. n indicates the number of independent experiments. For experiments using CHO cells, significant difference between values were determined using the unpaired Student's t test. For experiments using aortic ring tension, data are expressed as a percentage of the maximal contraction caused by phenylephrine (10⁵ M). In these experiments, n indicates the number of rats from which blood vessels were obtained. Statistical analysis was performed using two-way repeated analysis of variance. Significance (P) was determined at the 0.05 level.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Verification of Synthesis of Lithocholylcholine. The structure of synthetic lithocholylcholine was confirmed by mass spectrometry (Fig. 1C). For the lithocholylcholine cation, the measured and calculated masses were 462.3951 and 462.3947, respectively. Negative ion electrospray interface-mass spectrometry revealed that lithocholylcholine was obtained as the chloride salt.

Recombinant Acetylcholinesterase Hydrolyzes Lithocholylcholine. To determine whether lithocholylcholine is subject to the same in vivo metabolism as acetylcholine and to confirm its structure, we examined the actions of recombinant acetylcholinesterase on the newly synthesized molecule. As shown in Fig. 2, incubation of lithocholylcholine with recombinant acetylcholinesterase, but not denatured enzyme, resulted in hydrolysis and formation of lithocholic acid. Naturally occurring lithocholic acid conjugates (lithocholyltaurine and lithocholylglycine) were not affected by the acetylcholinesterase.

View larger version (56K): [in this window] [in a new window]   Fig. 2.   Hydrolysis of lithocholylcholine by recombinant acetylcholinesterase. Lithocholylcholine (LCC), lithocholylglycine (LCG) or lithocholyltaurine (LCT) (0.06 µmol) was incubated with or without recombinant acetylcholinesterase (0.14 µg of protein/ml of reaction mixture) for 1 h at 37°C. The reaction mixture was dried, extracted into MeOH and spotted onto a Silica Gel G plate. As a control, lithocholylcholine was also treated with heat-denatured enzyme. After separation by silica gel TLC in a chloroform/methanol/acetic acid (21:15:0.75, v/v/v) system, spots were visualized by charring. The data shown are representative of two separate experiments. LA, lithocholic acid.

Inhibition of Muscarinic Radioligand Binding by Lithocholylcholine and Other Muscarinic Ligands. Interaction of lithocholylcholine and other cholinergic ligands with CHO cells expressing M1 to M5 muscarinic receptors was determined by examining their effects on binding of a well characterized muscarinic radioligand, [³H]NMS. Dose-response analysis revealed different patterns of interactions with the receptors (Fig. 3; Tables 1 and 2). With M3 receptors, the subtypes most commonly expressed in gastrointestinal epithelium, lithocholyltaurine and lithocholylglycine, the naturally occurring bile acids, were able to displace [³H]NMS at concentrations that are well within those reported for these bile acids in the human biliary tree and intestines (Table 1) (Vlahcevic et al., 1971, 1972; Mallory et al., 1973). Lithocholylcholine had the greatest potency for M3 compared with the other receptors (K_i = 1 µM) (Table 2). The binding affinity (K_i) of lithocholylcholine was the same for human and rat M3 receptors (Table 2). Moreover, at the M3 receptor, lithocholylcholine was approximately 2-, 25-, and 35-fold more potent than acetylcholine, lithocholyltaurine, and lithocholylglycine, respectively. With hM2 receptors, the subtype most commonly expressed on intestinal smooth muscle cells, lithocholylcholine, was 12-fold less potent than acetylcholine, but much more potent (>60-fold) than lithocholyltaurine and lithocholylglycine (Table 2).

View larger version (30K): [in this window] [in a new window]   Fig. 3.   Dose response for lithocholylcholine (LCC), lithocholyltaurine (LCT), and lithocholylglycine (LCG) on binding of a muscarinic radioligand to rM1, hM2, rM3, hM3, hM4, and hM5 receptors. Increasing concentrations of atropine (closed circles), acetylcholine (ACh, closed squares), carbachol (Carb, open triangles), lithocholylcholine (open circles), and lithocholylglycine (closed triangles) inhibit binding of [3H]NMS to CHO cells expressing M1 to M5 muscarinic receptors. Increasing concentrations of lithocholyltaurine (open squares) inhibit [3H]NMS binding to CHO cells expressing rM1, rM3, and hM3, but not hM2, hM4, and hM5 muscarinic receptors. Data points represent mean ± S.E.M. of at least three independent experiments. Error bars not shown when smaller than the symbol. , indicates values that are significantly less than those observed with the same concentration of acetylcholine (P < 0.005). Curves were obtained by computer analysis using a nonlinear, least-squares curve-fitting program, LIGAND (Munson and Rodbard, 1980).

View this table: [in this window] [in a new window]   TABLE 1 Potency of cholinergic agents and bile acids for inhibition of [3H]NMS binding to CHO cells expressing M1 to M5 muscarinic receptors IC50 values (micromolar) were obtained by computer analysis of dose-inhibition curves of [3H]NMS binding shown in Fig. 3 using a nonlinear, least-squares curve-fitting program, LIGAND (Munson and Rodbard, 1980). Values are mean ± S.E.M. of at least three independent experiments.

View this table: [in this window] [in a new window]   TABLE 2 Calculated ligand binding affinities [Ki (micromolar)] for interaction of agents with M1 to M5 muscarinic receptors expressed in CHO cells Mean Ki values were calculated from the IC50 values shown in Table 1 by the method of Cheng and Prusoff (1973). KD and Bmax values necessary for the calculation were obtained from Wang el-Fakahany (1993) and Buckley et al. (1989). Due to the nature of the calculation, the error associated with the Ki values cannot be reliably estimated.

Actions of Lithocholylcholine on Downstream Signaling Resulting from Activation of Muscarinic Receptors. Lithocholylcholine was characterized functionally by its ability to alter CHO cell inositol phosphate formation, cAMP accumulation, and phosphorylation of MAP kinase (Figs. 4-6). Acetylcholine and carbachol caused a dose-dependent increase in inositol phosphate formation in CHO-rM3, but not in CHO-hM2 cells (Fig. 4, A and B). Lithocholylcholine alone did not alter inositol phosphate formation in cells expressing either muscarinic receptor subtype. In CHO-rM3 cells, increasing concentrations of lithocholylcholine, lithocholyltaurine, and lithocholylglycine caused a progressive decrease in acetylcholine-induced inositol phosphate formation (Fig. 4C). For this action, lithocholylcholine was almost 1 log more potent than lithocholyltaurine and lithocholylglycine (Fig. 4C). These observations indicate that lithocholylcholine is an M3-receptor antagonist.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Cholinergic agents and bile acids alter formation of inositol phosphates. A, dose response for actions of cholinergic agonists and bile acids on stimulation of inositol phosphate formation in CHO cells expressing rat M3 muscarinic receptors. Acetylcholine (ACh, closed squares), carbachol (Carb, open triangles), lithocholylcholine (LCC, open circles), lithocholyltaurine (LCT, open squares), and lithocholylglycine (LCG, closed triangles). B, cholinergic agonists and bile acids do not alter inositol phosphate formation in CHO cells expressing human M2 receptors. C, in CHO cells expressing rat M3 receptors, increasing concentrations of lithocholylcholine, lithocholyltaurine, and lithocholylglycine reduce acetylcholine (10 µM)-induced inositol phosphate formation. Data points represent mean ± S.E.M. of at least three independent experiments.  and , indicate values that are significantly less than those observed with 10 µM acetylcholine alone (P < 0.05 and 0.001, respectively).

View larger version (36K): [in this window] [in a new window]   Fig. 5.   Effect of lithocholylcholine on cAMP formation in CHO cells expressing human M2 muscarinic receptors. cAMP accumulation in the presence of 3-isobutyl-1-methylxanthine and 1% DMSO was measured. Acetylcholine (ACh, 0.1 µM) inhibited forskolin-stimulated cAMP accumulation. Lithocholylcholine (LCC, 25 µM) and atropine (10 µM) blocked the actions of acetylcholine. Data points represent mean ± S.E.M. of at least three independent experiments.  and , indicate values that are significantly greater than those observed with forskolin plus acetylcholine (P < 0.01 and 0.001, respectively).

View larger version (33K): [in this window] [in a new window]   Fig. 6.   Cholinergic agents and bile acids alter MAP kinase phosphorylation. A, in CHO cells expressing M1 to M5 muscarinic receptors, acetylcholine (ACh), and carbachol (Carb), but not lithocholylcholine (LCC), lithocholyltaurine (LCT), and lithocholylglycine (LCG) (all 100 µM), stimulate phosphorylation of p44 (ERK1) and p42 (ERK2). In CHO-K1 cells that do not express muscarinic receptors, none of these agents stimulate p44 or p42 phosphorylation. H2O is the solvent for acetylcholine, carbachol, and NMS, whereas 0.5% DMSO is the solvent for the bile acids. For all immunoblots in Fig. 6, proteins in row a were immunoblotted with anti-phospho-p44/42. To show that equal amounts of protein were added to each lane; those in row b were immunoblotted with anti-p44/42. B, N-methylscopolamine (NMS, 10 µM), atropine (10 µM), lithocholylcholine (50 µM), lithocholyltaurine (1 mM), and lithocholylglycine (0.1 mM) inhibit acetylcholine (10 µM)-induced phosphorylation of MAP kinase in M1 and M3 CHO cells, but not in M2, M4, and M5 CHO cells. C, N-methylscopolamine (NMS, 10 µM), lithocholylcholine (0.3 mM), lithocholyltaurine (1 mM), and lithocholylglycine (0.1 mM) inhibit acetylcholine (0.1 µM)-induced phosphorylation of MAP kinase in M2 CHO cells. D, dose response for effects of lithocholylcholine (circles) and lithocholyltaurine (squares) on acetylcholine (10 µM)-induced MAP kinase phosphorylation in rat M3 cells. The graph in the lower panel represents the mean ± S.E.M. of at least three independent experiments.  and , indicate values that are significantly less than those observed with 10 µM acetylcholine alone (C) (P < 0.05 and 0.001, respectively).

Inhibitory Action of Lithocholylcholine on Acetylcholine-Induced Rat Aortic Ring Relaxation. To determine whether the inhibitory effects of lithocholylcholine could be demonstrated in intact mammalian tissue, we examined its effects on acetylcholine-induced relaxation of phenylephrine-constricted rat aortic rings, an established model of muscarinic cholinergic action (Stewart and Kennedy, 1999). As shown in Fig. 7, increasing concentrations of acetylcholine caused progressive relaxation of phenylephrine-constricted rings. The addition of 100 µM lithocholylcholine inhibited the actions of acetylcholine. These results provide evidence that lithocholylcholine can inhibit cholinergic actions at the organ level.

View larger version (19K): [in this window] [in a new window]   Fig. 7.   Actions of lithocholylcholine on acetylcholine-induced relaxation of 105 M phenylephrine-treated rat aortic rings. Increasing concentrations of acetylcholine (ACh) were added cumulatively to aortic rings incubated with lithocholylcholine (LCC, 100 µM) or diluent. The effect of acetylcholine is expressed as a percentage of phenylephrine-induced tension. , indicates P < 0.05 compared with the values obtained in the presence of lithocholylcholine (n = 4 per group).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Lithocholylcholine, a hybrid molecule synthesized by joining the steroid nucleus of lithocholic acid with the choline moiety of acetylcholine, acts as a muscarinic receptor antagonist. In CHO cells expressing muscarinic receptor subtypes, this is evidenced by the observations presented above. Lithocholylcholine inhibits muscarinic radioligand binding, postreceptor signal transduction, and, in rat aortic rings, acetylcholine-induced relaxation. Moreover, the findings presented here confirm our previous observation (Raufman et al., 1998) that, at physiological concentrations present in the human biliary tree and intestines (Vlahcevic et al., 1971; Vlahcevic et al., 1972; Mallory et al., 1973), some bile acids interact with muscarinic receptors and alter postreceptor signaling events. Interestingly, other investigators have reported that, over the same range of concentrations, bile acids interact with nuclear receptors to alter gene transcription, thereby affecting bile acid synthesis and transport (Makishima et al., 1999; Parks et al., 1999). Our findings provide evidence that bile acids may have signaling functions that extend beyond their effects on sterol metabolism, lipid absorption, and cholesterol elimination. Moreover, the present studies indicate that the approach of creating hybrid molecules from bile acids and acetylcholine may provide new selective muscarinic receptor ligands.

A brief discussion of the potential clinical utility of lithocholylcholine is warranted. Based on the data shown in Fig. 3 and Tables 1 and 2, it is apparent that for each of the muscarinic receptor subtypes tested, lithocholylcholine is approximately 1000-fold less potent than atropine. The efficacy of lithocholylcholine is similar to that of atropine. The ability of tissue acetylcholinesterases to hydrolyze lithocholylcholine suggests that the in vivo half-life of the agent may be limited. This might be advantageous if one is seeking a reversible agent. However, the actual serum and tissue half-life of lithocholylcholine can only be determined using in vivo studies that are beyond the scope of the present work. Furthermore, the lipophilicity of lithocholylcholine compared with currently available water-soluble anticholinergics may provide benefit in terms of access to receptors in the cell membrane lipid bilayer, particularly in the central nervous system. Again, this is better addressed using in vivo models. Nevertheless, the data in Fig. 7 indicating the ability of lithocholylcholine to inhibit the actions of acetylcholine on vasodilatation of rat aorta, demonstrate the potential functional utility of lithocholylcholine in an intact tissue model. The synthesis of lithocholylcholine and its demonstrated actions as a muscarinic receptor antagonist opens the door to the creation of other potentially more potent and nonhydrolyzable hybrids of bile acids and acetylcholine.