Inhibition of calcium phosphate precipitation by bile salts: a test of the Can+-buffering hypothesis

The ability of bile salts to inhibit the precipitation of either calcium hydroxyapatite or its precursor, amorphous calcium phosphate, by reducing CaZ+ activity or poisoning nascent crystals was determined. When apatite precipitated rapidly (1-4 h), glycocholate and taurine-conjugated bile salts (up to 100 mM) had little effect on apatite formation, but prevented amorphous calcium phosphate precipitation by lowering Ca2+ activity. In contrast, glycodeoxycholate and glycochenodeoxycholate (2-3 mM) inhibited apatite formation for at least 24 h by poisoning embryonic apatite. When apatite precipitated slowly (> 24 h), all the dihydmxy bile salts prevented apatite formation for at least 4 days. At constant initial supersaturation, the phosphate concentration determined the degree of inhibition caused by the six bile salts mixed together in physiologic proportion. At low phosphate concentrations (1.2 mM) total inhibition was achieved by poisoning embryos ( 5 mM total bile salt), but with 4.0 mM phosphate only -60% inhibition was attained (150 mM bile salt) by a combination of poisoning and Caz+-buffering. B Thus, at low supersaturation all dihydroxy bile salts can prevent apatite formation by reducing free Ca2+ (taurine and glycine conjugates) or poisoning embryos (glycine conjugates). With mixtures of bile salts at higher supersaturation, inhibition of apatite depends on a combination of poisoning and reduction of free Ca2*, mainly caused by glycodeoxycholate and glycochenodeoxycho1ate.Crowther, R. S., and M. Okido. Inhibition of calcium phosphate precipitation by bile salts: a test of the Ca2+-buffering hypothesis. J Lipid Res. 1994. 35: 279-290. Supplementary key words calcium hydroxyapatite gallstones calcification biomineralization Calcium salts of bilirubin, carbonate, phosphate, and fatty acids are common constituents of all types of gallstones (1, 2) and understanding the factors that promote or inhibit precipitation of these salts is necessary for the development of strategies to prevent gallstone formation. Precipitation of ionic salts is governed by well-established physical laws, and it therefore appears a trivial undertaking to try to understand the formation of these salts in bile. However, bile is a complex fluid that undergoes both physical and chemical compositional changes, any one of which may influence the precipitation of Ca2+-sensitive anions. Because bile salts are the major biliary solutes, initial attempts to understand the regulation of Ca2+ salt formation in bile have focused on the role of these anions. Seminal work by Williamson and Percy-Robb (3, 4), subsequently confirmed by others (5-13), showed that bile salts bind Ca2+ ions and lower Ca2+ activity. This observation led Moore, Celic, and Ostrow (6 ) to propose the "Ca2+-buffering" hypothesis: inhibiting Ca2+-sensitive anion precipitation in bile by reducing Ca2+ activity is a major physiological function of bile salts. However, Caz+-binding per se will not necessarily cause physiologically significant inhibition of Ca2+ salt formation. Any reduction of Ca2+ activity may alter the kinetics of salt formation (Le., alter the onset of precipitation), but longterm, thermodynamic prevention of precipitation requires that the bile salt concentration and its Ca2+ binding affinity must together be sufficient to make bile unsaturated with respect to the target Ca2+ salt. Unfortunately, although many groups have studied bile salt-Ca2+ binding, relatively few investigators have examined the consequences of this binding for Ca2+ salt precipitation. Thus the ea2+-buffering hypothesis has not been adequately tested. Angelic0 et al. (14) demonstrated in the bile-fistula rat model that the precipitation of mixed Ca2+ salts of monoand unconjugated bilirubin and palmitate occurred under conditions of bile salt depletion. Precipitation was promoted when Ca2+ was added to bile salt-depleted bile, but not when added to bile salt-rich bile. Conversely, precipitation was inhibited by the addition of micellar concentrations of bile salts, but it was uncertain whether this effect resulted from Ca2+ buffering or from the solubilization of excess unconjugated bilirubin. In vitro Abbreviations: ACP, amorphous calcium phosphate; CMC, critical micellar concentration; DSHAP, degree of saturation with respect to hydroxyapatite; GC, glycocholate; GCDC, glycochenodeoxycholate; GDC, glycodeoxycholate; HAP, calcium hydroxyapatite; Pi, inorganic phosphate; TC, taurocholate; TCDC, taurochenodeoxycholate; TDC, taurodeoxycholate. 'To whom correspondence should be addressed. Visiting scholar from the Department of Surgery, Kyushu University, Fukuoka, Japan. Journal of Lipid Research Volume 35, 1994 279 by gest, on O cber 8, 2017 w w w .j.org D ow nladed fom studies of fatty acid precipitation from micellar bile salt solutions showed that calcium palmitate precipitation was promoted when Ca2+ was bound to the bile saltlpalmitate mixed micelles (15). Glycine-conjugated bile salts were shown to inhibit calcium carbonate formation in vitro, but it was concluded that an unknown mechanism, unrelated to the reduction of Ca2+ activity, was responsible (16). Thus, of the few studies performed, none have verified the Ca2+-buffering hypothesis, and some have suggested that other mechanisms may be involved in inhibiting CaZ+-sensitive anion precipitation. Sutor and Percival (17) showed that small volumes of bile inhibited calcium phosphate formation in vitro, and this effect was reproduced by low concentrations of mixed taurocholatelphosphatidylcholine micelles. We recently showed that glycochenodeoxycholate (GCDC) inhibited calcium phosphate precipitation by “poisoning” calcium hydroxyapatite (HAP) embryos, which prevented the transformation of amorphous calcium phosphate (ACP) into HAP (18). Moore et al. (19) calculated that bile was normally undersaturated with respect to CaHP04, and they suggested that this mineral phase could precipitate in bile only when excess inorganic phosphate was produced by phospholipid hydrolysis. CaHP04 is one of the possible forms of ACP, and although ACP was once believed to be an obligate precursor of HAP, this is now known to be false (20). Because HAP is much more insoluble than CaHP04 (21), it may precipitate from solutions that are undersaturated with respect to CaHP04 and therefore phospholipid hydrolysis may not be necessary for HAP to precipitate in bile. The major forms of calcium phosphate in gallstones are HAP and whitlockite (l), and the object of the current work was to study the effects of the major human bile salts on the formation of HAP and its sometime precursor ACP, and to deduce the mechanism of inhibition in each case. MATERIALS AND METHODS CaCl,, Na2HP04, and NaCl were obtained from J. T. Baker Chemical Co. (Phillipsburg, NJ). Sodium salts of GCDC, glycocholic (GC), taurocholic (E), glycodeoxycholic (GDC), taurodeoxycholic (TDC), and taurochenodeoxycholic ( E D C ) acids were obtained from Sigma Chemical Co. (St. Louis, MO), Steraloids Inc. (Wilton, NH), and Calbiochem (La Jolla, CA). Except where noted, experiments were performed with bile salts obtained from Sigma. Tetrahydrofuran and 1,6-dipheny11,3,5-hexatriene were obtained from Aldrich Chemical Co. (Milwaukee, WI). HAP (type VI) and all other reagents were from Sigma Chemical Co. HAP was analyzed for calcium and phosphate content and the C d P ratio was 1.69 + 0.01, which is consistent with HAP, Ca5(P04)30H, (1.67). Calcium phosphate precipitation Stock solutions of NaCl, Na2HP04, and bile salts, all dissolved either in 50 mM Tris or 10 mM HEPES buffer, pH 7.5, were mixed together in polypropylene microcentrifuge tubes. The NaCl concentration was adjusted in inverse proportion to the bile salt concentration to keep the ionic strength approximately constant. Solutions were not purged of COz because previous experience had shown that precipitation of CaC03 did not occur under these experimental conditions (18). To initiate the reaction, CaClZ in buffer was added to give a final volume of 1.0 ml and the tubes were vortexed. In different series of experiments the initial CaClZ and Na2HP04 concentrations were adjusted so that HAP would precipitate either directly or through its ACP precursor. For each experiment the initial concentrations of reactants after mixing are given in the figure legends. After mixing, 200 p1 of solution was removed and the absorbance at 405 nm was measured (Titertek Multiskan Plus plate reader, Flow Labs, McLean, VA). The samples were returned to their respective tubes, which were capped and incubated at 37°C. During temperature equilibration the pH of the solutions fell to 7.3 when Tris buffer was used and to 7.4 when HEPES buffer was used. These pH values were used when initial saturations of solution with respect to HAP were calculated (see Calculation of solution saturation). At intervals the tubes were vortexed and the absorbance of a 200-4 aliquot of solution was measured as before. After 24 h incubation the solutions were centrifuged at 11,000 g for 5 min and the supernatants were assayed for inorganic phosphate (Pi) and, in some cases, for total calcium and bile salt (see Analytical methods). In some experiments the pellets were washed by resuspending them twice in 1 ml of deionized water, and the washed pellets were placed on glass slides and dried at 100OC. A sample of the dried pellet was transferred to a BaF, window, and the infrared spectrum of the precipitate was obtained at a resolution of 2 cm-1 using a UMA300A infrared microscope attached to an FTS6O FT-IR spectrometer (Bio-Rad, Digdab Division, Cambridge, MA).