- Email: [email protected]
Fatty acid composition of phospholipids in bile in man: Promoting effect of deoxycholate on arachidonate
Clinica Chimica Acta, 165 (1987) 21-37 Elsevier
21
CCA 03755
Fatty acid composition of phospholipids in bile in man: promoting effect of deoxycholate on arachidonate G.P. van Berge Henegouwen,S.D.J. van der Werf and A.T. Ruben Department ofJnternaf Medicine and G~tr~~tero~o~, Ma&erg-GZ, Arnhem {The ~ether~~~) (Received 8 May 1986; revision received 8 December 1986; accepted after revision 9 December 1986) Key work
Biliary lipids; Phospholipids; Fatty acid composition; Bile acids
Ninety-five percent of phospholipids (PLs) in bile is secreted as phosphatidylcholine or lecithin. The study of fatty acid patterns of phospholipids present in gallbladder bile could help clarify whether a preponderance of certain fatty acids could play a role in cholesterol gallstone fo~ation in man. In acute bile acid-exchange experiments, it was found that more hydrophobic bile acids did promote the excretion in bile of PL rich in arachidonic acid (a prostaglandin precursor) and steak acid. We studied, therefore, bile acid, cholesterol and phospholipid fatty acid patterns (measured by gas chromatography) in gallbladder bile, obtained by duodenal intubation and ~hol~ysto~-s~ulation of 24 healthy volunteers with normal hver/g~bladder fiction (ultraso~d~. PL-fatty acid composition (mean % f SD) was 41.40 ( f 1.41) for palmitic acid, 2.68 (f 0.82) for palmitoleic acid, 5.50 (f 1.55) for steak acid, 12.09 (rf: 0.98) for oleic acid, 32.83 ( f 3.04) for linoleic acid and 5.64 ( f 1.59) for arachidonic acid. The proportion of biliary deoxycholate was positively correlated with arachidonic acid (r = 0.71; p < 0.011, whereas ~hen~eoxycholate was inversely correlated with arachidonic acid (r = -0.53; p -z 0.02). There was a positive correlation between biliary chenodeoxycholate and linoleic acid (r = 0.48; p < 0.05) and a negative correlation between biliary deoxycholate and linoleic acid composition (r = 0.68; p < 0.01). Also a correlation was found between palmitic acid and cholesterol saturation index (r = 0.49; p -C0.05). We conclude that the hydrophobic bile acid deoxycholate, which does not desaturate cholesterol in bile, promotes the biliary excretion of arachidonic acid. Since
Correspondence Netherlands.
to: G.P. van Berge Henegouwen, Malberg-GZ, Wagnerlaan 55, 6815 AD Ar&em, ‘&e
~9-8981/87/$03.50
@ 1987 Efsevier Science Publishers B.V. (BiomedicaI Division)
28
arachidonic acid could induce the gallbladder mucosa to produce prostaglandins and mucus, increased biliary PL-arachidonic acid composition might be a factor in cholesterol gallstone disease. Introduction
Phospholipids (PLs) are major constituents of mammalian cell membranes. Especially PLs rich in arachidonic acid and stearic acid are strongly represented in intracellular membranes in the liver [l] and 95% of the PLs in bile are secreted as phosphatidylcholine (PC) or lecithin, in particular as the two molecular species 1-palmitoyl, 2-linoleyl (16 : O-18 : 2) PC and 1-palmitoyl, 2-oleoyl (16 : O-18 : 1) PC [l-3]. It has been shown that diets rich in n-6 or n-3 fatty acids induce a lowering in plasma cholesterol, especially LDL-cholesterol [4], but also increase cholesterol secretion in bile, suggesting that these fatty acids increase the transfer of cholesterol into bile [5-81. This could indicate that diets rich in n-3 or n-6 fatty acids might induce the hthogenic properties of bile. The study of fatty acid patterns of PLs present in gallbladder bile could help clarify whether a preponderance of certain fatty acids could play a role in the pathogenesis of cholesterol gallstone disease in man. Cantafora et al [9] and AhIberg et al [2] showed that patients with gallstones had an increased proportion of deoxycholate and also an increased proportion of arachidonate-containing PLs in bile. A negative correlation between the proportion of arachidonate in fatty acid composition and that of chenodeoxycholate was present in gallbladder bile of gallstone patients [9], but this has never been shown in normal healthy subjects. Acute bile acid exchange experiments in gallstone patients [lo] with T-drams in situ showed that the more hydrophobic bile acids like deoxycholate (and chenodeoxycholate), would seem to promote the preferential secretion into bile of PL-molecular species from liver cell plasma membranes rich in arachidonic acid and stearic acid. The objective of our study was to investigate the fatty acid composition in PLs in bile of a large group of normal healthy volunteers and to look at possible relations between biliary bile acid composition and the fatty acid composition present in the PLs in bile. Materials and methods Subjects
Twenty-four individuals (14 females and 10 males) participated in the study after informed consent. All had normal gallbladder and liver structure on ultrasonography. Eligibility required a non-compromised renal and hepatic function according to routine laboratory tests. Patients with hyperlipidaemia or obesity (actual weight more than 110% of ideal body wt.) were not included. The mean age was 49 yr (SD Ifr 16).
29
Microscopic examination of the deposit of the duodenal bile after centrifugation revealed no cholesterol crystals in all subjects. Bile sampling Dark brown bile was aspirated for the duodenum at 09.00 a.m. through a polyvinyl tube after 12-h overnight fast. Gallbladder contraction was induced by slow i.v. administration of cholecystokinin-octapeptide (CCK-OP, Sincalide@, E. Squibb, Prince-town, NY, USA). In the 24 healthy individuals, 4 bile samples were taken on 4 consecutive days. A 2-ml aliquot of the most concentrated bile was taken. The remainder of the aspirate was returned to avoid depletion of the bile acid pool. A 0.5~ml portion was diluted immediately after sampling with 4.5 ml isopropanol (1 : 10, v/v) to inactivate possible phospholipase activity. 1.5 ml was frozen to -20°C and stored at that temperature together with the isopropanolic solution for further processing. Analyses Biliary lipid composition The concentration of biliary lipids was determined in the isopropanolic solution of duodenal bile. Cholesterol and bile acids were measured simultaneously by gas-liquid chromatography in a l-ml aliquot [ll]. After evaporation enzymatic hydrolysis was performed with cholylglycine hydrolase (EC 1.51.2.4, Sigma Chemical C., St. Louis, MO, USA) as previously described [12,13]. The mixture was acidified with 0.5 ml concentrated hydrochloric acid. Biliary lipids were extracted thrice with 10 ml diethylether (recovery 97-1018). The evaporated extract was methylated with an acidified dimethoxypropane-methanol mixture (1 : 1, v/v) [14]. The cholesterol and bile acid methylesters were quantified after conversion into trifluoroacetates against the 7-ketodeoxycholic acid internal standard using a gas-chromatograph type Becker 420 (Packard-Becker, Delft, The Netherlands) with a U-shaped all glass column, 1 m X 4 mm i.d. packed with 1% OV 210 on gaschrom (Y,100-120 mesh (Chrompack, Middelburg, The Netherlands) [14,15]. Total bile acid concentrations were calculated as the sum of cholic acid, deoxycholic acid, lithocholic acid and chenodeoxycholic acid. The analyses of cholesterol concentration by gas-liquid chromatography were in close agreement with those performed by the enzymatic calorimetric method using the two Boehringer Mannheim test combinations [15,16] (n = 15; r = 0.9930, y = 1.0170x - 0.5724, p < 0.0001) as has been reported by Bolton et al [16]. Total PL content of the isopropanolic bile solutions was measured in 0.1 ml as lipid soluble phosphorus [17]. Only bile samples containing total bile acid concentrations of > 30 mm01 were included in this study. Total bile acid concentrations varied from 30-80 mmol. Lipid composition of bile was expressed as mole fraction of total bile acids, PLs and cholesterol. The lithogenic index was derived from the polynomial equations developed by Thomas and Hofmann [18] to describe the cholesterol solubility line as proposed by Hegardt and Dam [19], and Holzbach et al [20].
30
Fatty acid composition of Mary PLS
Methylesters of fatty acids were analysed by gas-liquid chromatography. A mixture of normal human fatty acids was used as standards (Sigma Chemical Corp., Poole, Dorset, UK; high grade purity > 99%). First 1 ml of diluted bile (1: 10; v/v) isopropanolic solution) was dried under a nitrogen stream at 40 “C. Then 0.5 ml CH,OH was added, mixed and then 0.2 ml of internal standard solution (2 mg of heptadecanoid acid in 2.5 ml benzene). Subsequently, 0.3 ml of benzene was added together with 0.5 ml of 2.2-dimethoxypropane (DMP) and ca 0.03 ml of concentrated HCl (38%) was pipetted into this solution. The reaction mixture was left overnight or for at least 6 h at room temperature to ensure complete hydrolysis and subsequent esterification. The extraction of fatty acid-methylesters was performed with 3 ml hexane (2 times) after addition of 2.5 ml of H,O. The hexane solution was dried under an nitrogen stream at 40 o C. After addition of 0.1 ml hexane a sample of 0.6 ~1 was injected into the gas-chromatograph. With the DMP-method both esterified (bound in PLs) and free fatty acids in bile are analysed [21]. Selected samples were also analysed by transesterification with methanolic base (0.5 mol/l) made from metallic sodium and dry methanol (Supelchem BV, Hilversum, The Netherlands) [22]. Only esterified fatty acids in bile react with methanolic base, but not the free fatty acids. By combining the results of the DMP-method and the methanohc base-method the free fatty acid fraction potentially present in bile could be estimated. The methanolic base-method was done as follows: 0.2 ml of internal standard solution was added to 0.3 ml of benzene and 0.5 ml of CH,OH together with 0.5 ml of DMP and 0.03 ml of concentrated HCl. This reaction mixture was left overnight and then evaporated. Subsequently, 1 ml of the isopropanolic bile solution was added. This again was evaporated and then dissolved in 1 ml of methanolic base and 0.5 ml of benzene. This solution was kept for 30 mm at 60 ‘C; after cooling to room temperature 2.5 ml of H,O was added and fatty acid methylesters were extracted twice with 3 ml of hexane. The hexane solution was dried under a nitrogen stream and further gas-chromatographic analysis was performed as described with the DMP-method. When we analysed the undiluted bile stored at - 20 o C with the methanolic baseand the DMP-method a considerable amount of free fatty acids was present. However, if we analysed the isopropanolic solution of bile stored at - 20 o C with the two methods no free fatty acids were present. These results indicate that in the undiluted bile samples after thawing some hydrolysis of PLs to free fatty acids did occur. The methylesters of fatty acids were separated by using a gas-chromatograph type Becker 420 (Packard-Becker) filled with a coiled all-glass-column (6 ft X 2 mm i.d. and 6 mm o.d.) packed with 10% SP 2310 on 100/120 chromosorb W-AW at 195 o C. The flow rate was 20 ml/mm. N,-gas was used as a carrier-gas. Percentages were corrected for detector response by using a known amount of internal standard in each sample.
31
Statistical methods
The data on biliary lipids and fatty acid composition (obtained by averaging the results of analyses of the four bile samples taken on four consecutive days in one person) were normally distributed and are presented as individual values with their means. Correlations of individual biliary bile acids (x) with PL-fatty acid composition (y), were calculated by the method of least-squares with associated confidence limits of 95% for the regression of y on X. Their significance was derived from Pearson’s correlation coefficient. Results Biliary fatty acid composition of PLs in hearthy individuals
The composition present in the 24 healthy individuals is shown in Table I. As should be expected palrnitic acid and linoleic acid were the predominant fatty acids present in PLs in bile. They accounted for approximately 75% of total fatty acids. Free fatty acids could not be detected. There was no significant difference in fatty acid-PL composition between males and females in our study group. Relation biliary bile acids and PL-fatty acid composition
In the 24 healthy individuals, we looked at possible correlations between individual biliary bile acid composition and PL-fatty acid composition. A negative correlation was found between the proportion of biliary cholic acid and that of biliary arachidonic acid (Fig. 1; r = - 0.52; p < 0.01) and also between the other primary bile acid chenodeoxycholic acid and arachidonic acid (Fig. 2; r = - 0.53; p < 0.01). On the other hand a significantly positive correlation was found between the secondary bile acid deoxycholic acid and arachidonic acid (Fig. 3; r = 0.71; p < 0.01). Opposite correlations as for arachidonic acid were found for linoleic acid. There was a positive correlation between the proportion of biliary chenodeoxycholic acid and that of linoleic acid (r = 0.48; p < 0.05) and a negative correlation between biliary deoxycholic acid and linoleic acid (Fig. 4; r = -0.68; p < 0.01). A positive
TABLE Biliary
I fatty acid composition
Biliary PL-fatty Type fatty acid Palmitic acid Pahnitoleic acid Steak acid Oleic acid Linoleic acid Linolenic acid Dihomo-y-linoleic Arachidonic acid
in PLs
acids
acid
Healthy individuals Mean % (*SD) 16:O 16:l 18:0 18:l 18:2 18:3 20:3 2014
41.40 2.68 5.50 12.09 32.83 Trace Trace 5.64
(1.41) (0.82) (1.55) (0.98) (3.06)
(1.59)
(n = 24)
32 Biliary PL-arachidonic acid (mol %I 10 _
?-
.
6-
4-
0 25
M
35
40
1 50
45
Biliary CA tmol %b)
Fig. 1. Relation between biliary cholic acid and arachidonic acid in healthy subjects (n = 24). Arachidonic acid composition (y, mol W) is expressed as function of biliary cholic acid composition (x, mol W). The solid line depicts the regression function generated by the least squares method. The correlation was significant ( r = - 0.52; p < 0.01). The curves depict the 95% confidence limits for the regression of y on X.
Biliary P L-arachidonic acid I mol k b
‘EI l-
.
:.:
.
. .
r
24 =-0.53
6-
Biliary COCA tmol %I Fig. 2. Relation between biliary chenodeoxycholic acid and arachidonic acid in healthy subjects (n = 24). Arachidonic acid composition (y, mol W) is expressed as function of biiary chenodeoxycholic acid composition (x, mol %). The solid line depicts the regression function generated by the least-squares method. Correlation was significant (r = -0.53; p -c 0.01). The curves depict the 95% confidence limits for the regression of y on x.
33 Biliary PL-arachidonic acid (mol %
1
Biliary
DCA
lmol
%)
Fig. 3. Relation between biliary deoxycholic acid and arachidonic acid in healthy subjects (n = Arachidonic acid composition (y, mol %) is expressed as function of deoxycholic acid composition mol W). The solid line depicts the regression function generated by the least-squares method. correlation was significant (r = + 0.71; p < 0.01). The curves depict 95% confidence limits for regression of y on X. Biliary acid
24). (x, The the
P L - linoleic
1mol
%)
.
“’
24
r = -0.68 .
\
251 0
5
10
15
20
25
I
35
30 Biliary
DCA
(mol%)
Fig. 4. Relation between biliary deoxycholic acid and linoleic acid in healthy subjects (n = 24). Linoleic acid composition (v, mol W) is expressed as function of deoxycholic acid composition (x, mol W). The solid line depicts the regression function generated by the least-squares method. The correlation was significant (r = -0.68; p < 0.01). The curves indicate the 95% confidence limits for the regression of y on x.
34 Biliary CDCA (mol%)
“=
24
r = -0.87
Eiliary
DCA (mol81
Fig. 5. Relation between biliary chenodeoxycholic and deoxycholic acid in healthy subjects (n = 24). Chenodeoxycholic acid composition (y, mol %) is showed as function of deoxycholic acid (x, mol R;). The solid line depicts the regression function generated by the least-squares method. The correlation was significant (r = -0.87; p < 0.01). The curves depict the 95%confidence limits for the regression of y on X.
correlation was found between biliary cholic acid and steak acid (r = 0.45; p -C0.05). Positive correlations were also found between deoxycholic acid and oleic acid (r = 0.44; p < 0.05) and between biliary cholate and linoleic acid (r = 0.44; p -C0.05). The opposite relationships of deoxycholic acid and chenodeoxycholic acid with arachidonic acid and linoleic acid were not surprising because a strong reciprocal relationship existed between the proportion of biliary chenodeoxycholic acid and that of deoxycholic acid (Fig. 5; r = -0.87; p -c 0.01). Finally, also a positive correlation was found between the cholesterol saturation index and biliary palmitic acid (r = 0.49; p < 0.05) and between the cholesterol saturation index and linoleic acid (r = 0.45; p -c 0.05). Discussion
Fatty acid composition of lecithin in gallbladder bile of the healthy individuals showed that palmitic acid and linoleic acid were the predominant fatty acids. Similar results have been obtained by other studies [2,3,10]. In the present study, it was shown that an inverse relationship did exist between the proportion of both primary bile acids in bile, cholate and chenodeoxycholate, and that of arachidonic acid. There was also a positive relationship between both primary biliary acids and linoleic acid. On the other hand deoxycholate showed the
35
opposite effect. A strong positive relationship of deoxycholate with arachidonate acid was shown and a negative relation with linoleic acid. Previous studies have shown that gallbladder bile of cholesterol gallstone patients shows an increased lecithin arachidonate composition and also a higher deoxycholate proportion in biliary bile acids [2,9]. Furthermore, Cantafora et al [9] observed an inverse relationship between arachidonate and chenodeoxycholate composition. Their study, however, was confined to gallstone patients. Healthy controls were not included. Our study extends their findings of an inverse correlation between chenodeoxycholate and arachidonate and the positive correlation between deoxycholate and arachidonate to the normal population, too. In view of the strong inverse relationship between chenodeoxycholate and deoxycholate in bile, the opposite relation of chenodeoxycholate and deoxycholate with linoleate and arachidonate does not seem surprising. This strong inverse relationship between chenodeoxycholate and deoxycholate in bile is likely to be explained by competition of intestinal transport. In our study, the correlations detected between deoxycholate and arachidonate and between deoxycholate and linoleate in bile, were stronger than for chenodeoxycholate. For this reason we believe that more the relationship between the proportion of biliary deoxycholate and biliary PL-fatty acid composition than the relationship between chenodeoxycholate and PL-fatty acid composition, is likely to be explained by acute coupling effects in induced biliary lipid secretion, Ahlberg et al [2], however, showed that chenodeoxycholic acid treatment in gallstone patients did induce phospholipid species in gallbladder bile containing less arachidonate and Salvioli et al [lo] showed in acute bile exchange experiments that both deoxycholate and chenodeoxycholate increased the secretion of bilk-y lecithin, rich in arachidonate and stearate. The findings of Salvioli et al [lo] are not completely confirmed by our study. As no actual biliary secretion of lipids was measured in our study, we have no proof of the validity of our data. On the other hand our studies were performed in the natural situation without any possible unphysiological manipulation. Although quantification of biliary PL molecular species could provide more useful information, this seems unlikely in relation to the arachidonic acid composition, since Ahlberg et al [2] showed that arachidonic acid was present in PC in human bile only in combination with palmitic acid (1-palmitoyl-2-arachidonyl-snglycero-phosphocholine). Several studies have shown an increased proportion of deoxycholate - generally held to be a saturating or unequivocally non-desaturating bile acid - in gallbladder bile of cholesterol gallstone patients [23-251. Also, the National Cooperative Gallstone Study demonstrated a positive correlation between biliary deoxycholate percentage and biliary cholesterol saturation [26]. The finding in our study, employing only healthy subjects, of a strong positive correlation between biliary deoxycholate and PC-arachidonate could have some implication for the pathogenesis of cholesterol gallstone disease. It could mean that a bile acid pool size rich in deoxycholate could preferentially promote the biliary secretion of phospholipid species rich in arachidonate from the hepatocytic cell membrane. It has been suggested from animal experiments that
36
arachidonate in gallbladder bile could stimulate prostaglandin synthesis in the gallbladder wall, which in turn could induce glycoprotein or mucin secretion by the mucosa [26,27]. Glycoproteins are incriminated in cholesterol gallstone formation because they can act as nucleating factors [28]. The fact that we did not find detectable amounts of free fatty acids in bile does not necessarily contradict this hypothesis, because only minute amounts of free arachidonic acid, below the detection limit of our method, will be necessary for an appreciable prostaglandin effect. Thus both the positive correlation between biliary deoxycholate and biliary cholesterol saturation and between biliary deoxycholate and arachidonate could have some relevance with respect to the pathogenesis of cholesterol gallstone disease. Our study also provides additional arguments for the hypothesis that bile acid pool size composition is important in the biliary secretion of PL species. Although it is well known that the biliary secretion of PLs quantitatively is dependent of total bile acid secretion, our study indicates also that especially the biliary secretion of PL species rich in arachidonate and linoleate is dependent of the secretion of the secondary bile acid deoxycholate. Acknowledgements
The authors are grateful to Mr. B.H. Jansen, Department of Nuclear Chemistry, Malberg-GZ Hospital, for the statistical analysis and to Mrs. Ineke Janssen-Volman for expert secretarial assistance. This work was supported in part by a ZWO/ FUNGO-grant and a grant from the Dutch Liver and Digestive Foundation (De Nederlandse Lever-Darn-i Stichting). References 1 Kwamoto T, Okana G, Akin0 T. Biosynthesis and turn-over of individual molecular species of phosphatidyl-choline in liver and bile. Biochim Biophys Acta 1980;619:20-34. 2 Ahlberg J, Curstedt T, Einarsson H, Sjiivall J. Molecular species of biliary phosphatidyl-choline in gallstone patients. J Lipid Res 1981;22:404-409. 3 Cantafora A, DiBiase A, Alvaro D, Angelic0 M, Marin M, Attili AF. High performance liquid chromatographic analysis of molecular species of phosphatidylcholine-development of quantitative assay and its application to human bile. Clin Chim Acta 1983;134:281-295. 4 Goodnight SH, Harris WS, Connor WE, Illingworth DR. Polyunsaturated fatty acids, hyperlipidemia and thrombosis. Arteriosclerosis 1982;2:87-113. 5 Rameska CS, Paul R, Ganerly J. Effect of dietary unsaturated oils on the biosynthesis of cholesterol and on biliary and fecal excretion of cholesterol and bile acids in rats. J Nutr 1980;110:2149-2158. 6 Balasabramaniam S, Simons LA, Chang S, Hi&e JB. Reduction in plasma cholesterol and increase in biliary cholesterol by a diet rich in n-3 fatty acids in the rat. J Lipid Res 1985;26:684-689. 7 Grundy SM, Abrens Jr. EH. The effects of unsaturated dietary fats on absorption excretion, synthesis and distribution of cholesterol in man. J Clin Invest 1970;49:1135-1152. 8 Grundy SM. Effects of polyunsaturated fats on lipid metabolism in patients with hypertriglyceridemia. J Clin Invest 1975;55:269-282. 9 Cantafora A, Angel& M, DiBiase A, Pieche U, Bracci F, Attili AF, Capocaccio L. Structure of biliary phosphatidyl-choline in cholesterol gallstone patients. Lipids 1981;16:589-592. 10 Salvioli G, Romani M, Loria P, Carulli M, Pradelli JM. Effect of acute administration of bile acids on fatty acid composition of biiary phosphatidylcholine in man. J Hepatol 1985;1:291-300.
31 11 Junker RL, Hassan AS, Subbiah MTR. Simultaneous quantitation of biliary cholesterol, bile acids and phospholipids and fatty acids, by gas-liquid chromatography. Clin Chem 1981;27:1779-1780. 12 Van Berge Henegouwen GP, Ruben AT, Brandt K-H. Quantitative analysis of bile acids in serum and bile, using gas-liquid chromatography. Clin Chim Acta 1974;54:249-261. 13 Huijbregts AWM, Van Schaik A, Van Berge Henegouwen GP, Van der Werf SDJ. Serum lipids, biliary lipid composition and bile acid metabolism in vegetarians as compared to normal controls. Eur J Clin Invest 1980;10:443-449. 14 Shaw R, Elliot WH. Bile acids. LV, 2.2-Dimethoxypropane: an esterifying agent preferred to diazomethane for chenodeoxycholic acid. J Lipid Res 1978;19:783-787. 15 Fromm H, Amin P, Klein M, Kupke I. Use of a simple enzymatic assay for cholesterol analysis in human bile. J Lipid Res 1980;65:259-261. 16 Bolton CH, Nicholls JS, Heaton KW. Estimation of cholesterol in bile. Clin Chem Acta 1980;105:225-230. 17 La Russo NF, Hoffman NE, Hofmann Northfield TCN, Thistle JL. Effect of primary bile acid ingestion on bile acid metabolism and biliary lipid secretion in gallstone patients. Gastroenterology 1975;69:1301-1314. 18 Thomas PJ, Hofmann AF. A simple calculation of the lithogenic index of bile: expressing biliary composition on rectangular coordinates. Gastroenterology 1973;65:698-700. 19 Hegardt FG, Dam H. The solubility of cholesterol in aqueous solution of bile salts and lecithin. Z Emahnmgswiss 1971;10:223-233. 20 Holzbach RT, Marsh M, Olszewski M, Holan K. Cholesterol solubility in bile: evidence that supersaturated bile is frequent in healthy man. J Clin Invest 1973;52:1467-1479. 21 Mason ME, Waller GK. Dimethoxypropane induced transesterification of fats and oils in preparation of methyl esters for gaschromatographic analysis. Anal Chem 1964;36:583-586. 22 Luddy EF, Barford BA, Herb SF, Magidman P. A rapid and quantitative procedure for the preparation of methyl esters of butter oil and other fats. J Am Oil Chem Sot 1968;45:549-552. 23 Ahlberg J, Angelin B, Einarsson K. Influence of deoxycholic acid on biliary lipids in man. Clin Sci 1977;53:249-256. 24 Pomare EW, Heaton KW. Bile salt metabolism in patients with gallstones in functioning gallbladders. Gut 1973;14:885-890. 25 Hofmann AF, Grundy SM, La&in JM, et al and the NCGS-group. Pretreatment biliary lipid composition in white patients with radiolucent gallstones in the National Cooperative Gallstone Study. Gastroenterology 1982;83:738-752. 26 LaMont JT, Turner BS, Dibenedetto D, Handin R, Schafer AI. Aracbidonic acid stimulates mucin secretion in prairie dog gallbladder. Am J Physiol 1983;245:992-998. 27 Lee SP, LaMont SP, Carey MC. Role of the gallbladder mucus hypersecretion in the evolution of cholesterol gallstones. J Clin Invest 1981;67:1712-1723. 28 Smith BF, LaMont JT. Identification of gallbladder mu& bilirubin complex in human cholesterol gallstone matrix. J Clin Invest 1985;76:439-445.
21
CCA 03755
Fatty acid composition of phospholipids in bile in man: promoting effect of deoxycholate on arachidonate G.P. van Berge Henegouwen,S.D.J. van der Werf and A.T. Ruben Department ofJnternaf Medicine and G~tr~~tero~o~, Ma&erg-GZ, Arnhem {The ~ether~~~) (Received 8 May 1986; revision received 8 December 1986; accepted after revision 9 December 1986) Key work
Biliary lipids; Phospholipids; Fatty acid composition; Bile acids
Ninety-five percent of phospholipids (PLs) in bile is secreted as phosphatidylcholine or lecithin. The study of fatty acid patterns of phospholipids present in gallbladder bile could help clarify whether a preponderance of certain fatty acids could play a role in cholesterol gallstone fo~ation in man. In acute bile acid-exchange experiments, it was found that more hydrophobic bile acids did promote the excretion in bile of PL rich in arachidonic acid (a prostaglandin precursor) and steak acid. We studied, therefore, bile acid, cholesterol and phospholipid fatty acid patterns (measured by gas chromatography) in gallbladder bile, obtained by duodenal intubation and ~hol~ysto~-s~ulation of 24 healthy volunteers with normal hver/g~bladder fiction (ultraso~d~. PL-fatty acid composition (mean % f SD) was 41.40 ( f 1.41) for palmitic acid, 2.68 (f 0.82) for palmitoleic acid, 5.50 (f 1.55) for steak acid, 12.09 (rf: 0.98) for oleic acid, 32.83 ( f 3.04) for linoleic acid and 5.64 ( f 1.59) for arachidonic acid. The proportion of biliary deoxycholate was positively correlated with arachidonic acid (r = 0.71; p < 0.011, whereas ~hen~eoxycholate was inversely correlated with arachidonic acid (r = -0.53; p -z 0.02). There was a positive correlation between biliary chenodeoxycholate and linoleic acid (r = 0.48; p < 0.05) and a negative correlation between biliary deoxycholate and linoleic acid composition (r = 0.68; p < 0.01). Also a correlation was found between palmitic acid and cholesterol saturation index (r = 0.49; p -C0.05). We conclude that the hydrophobic bile acid deoxycholate, which does not desaturate cholesterol in bile, promotes the biliary excretion of arachidonic acid. Since
Correspondence Netherlands.
to: G.P. van Berge Henegouwen, Malberg-GZ, Wagnerlaan 55, 6815 AD Ar&em, ‘&e
~9-8981/87/$03.50
@ 1987 Efsevier Science Publishers B.V. (BiomedicaI Division)
28
arachidonic acid could induce the gallbladder mucosa to produce prostaglandins and mucus, increased biliary PL-arachidonic acid composition might be a factor in cholesterol gallstone disease. Introduction
Phospholipids (PLs) are major constituents of mammalian cell membranes. Especially PLs rich in arachidonic acid and stearic acid are strongly represented in intracellular membranes in the liver [l] and 95% of the PLs in bile are secreted as phosphatidylcholine (PC) or lecithin, in particular as the two molecular species 1-palmitoyl, 2-linoleyl (16 : O-18 : 2) PC and 1-palmitoyl, 2-oleoyl (16 : O-18 : 1) PC [l-3]. It has been shown that diets rich in n-6 or n-3 fatty acids induce a lowering in plasma cholesterol, especially LDL-cholesterol [4], but also increase cholesterol secretion in bile, suggesting that these fatty acids increase the transfer of cholesterol into bile [5-81. This could indicate that diets rich in n-3 or n-6 fatty acids might induce the hthogenic properties of bile. The study of fatty acid patterns of PLs present in gallbladder bile could help clarify whether a preponderance of certain fatty acids could play a role in the pathogenesis of cholesterol gallstone disease in man. Cantafora et al [9] and AhIberg et al [2] showed that patients with gallstones had an increased proportion of deoxycholate and also an increased proportion of arachidonate-containing PLs in bile. A negative correlation between the proportion of arachidonate in fatty acid composition and that of chenodeoxycholate was present in gallbladder bile of gallstone patients [9], but this has never been shown in normal healthy subjects. Acute bile acid exchange experiments in gallstone patients [lo] with T-drams in situ showed that the more hydrophobic bile acids like deoxycholate (and chenodeoxycholate), would seem to promote the preferential secretion into bile of PL-molecular species from liver cell plasma membranes rich in arachidonic acid and stearic acid. The objective of our study was to investigate the fatty acid composition in PLs in bile of a large group of normal healthy volunteers and to look at possible relations between biliary bile acid composition and the fatty acid composition present in the PLs in bile. Materials and methods Subjects
Twenty-four individuals (14 females and 10 males) participated in the study after informed consent. All had normal gallbladder and liver structure on ultrasonography. Eligibility required a non-compromised renal and hepatic function according to routine laboratory tests. Patients with hyperlipidaemia or obesity (actual weight more than 110% of ideal body wt.) were not included. The mean age was 49 yr (SD Ifr 16).
29
Microscopic examination of the deposit of the duodenal bile after centrifugation revealed no cholesterol crystals in all subjects. Bile sampling Dark brown bile was aspirated for the duodenum at 09.00 a.m. through a polyvinyl tube after 12-h overnight fast. Gallbladder contraction was induced by slow i.v. administration of cholecystokinin-octapeptide (CCK-OP, Sincalide@, E. Squibb, Prince-town, NY, USA). In the 24 healthy individuals, 4 bile samples were taken on 4 consecutive days. A 2-ml aliquot of the most concentrated bile was taken. The remainder of the aspirate was returned to avoid depletion of the bile acid pool. A 0.5~ml portion was diluted immediately after sampling with 4.5 ml isopropanol (1 : 10, v/v) to inactivate possible phospholipase activity. 1.5 ml was frozen to -20°C and stored at that temperature together with the isopropanolic solution for further processing. Analyses Biliary lipid composition The concentration of biliary lipids was determined in the isopropanolic solution of duodenal bile. Cholesterol and bile acids were measured simultaneously by gas-liquid chromatography in a l-ml aliquot [ll]. After evaporation enzymatic hydrolysis was performed with cholylglycine hydrolase (EC 1.51.2.4, Sigma Chemical C., St. Louis, MO, USA) as previously described [12,13]. The mixture was acidified with 0.5 ml concentrated hydrochloric acid. Biliary lipids were extracted thrice with 10 ml diethylether (recovery 97-1018). The evaporated extract was methylated with an acidified dimethoxypropane-methanol mixture (1 : 1, v/v) [14]. The cholesterol and bile acid methylesters were quantified after conversion into trifluoroacetates against the 7-ketodeoxycholic acid internal standard using a gas-chromatograph type Becker 420 (Packard-Becker, Delft, The Netherlands) with a U-shaped all glass column, 1 m X 4 mm i.d. packed with 1% OV 210 on gaschrom (Y,100-120 mesh (Chrompack, Middelburg, The Netherlands) [14,15]. Total bile acid concentrations were calculated as the sum of cholic acid, deoxycholic acid, lithocholic acid and chenodeoxycholic acid. The analyses of cholesterol concentration by gas-liquid chromatography were in close agreement with those performed by the enzymatic calorimetric method using the two Boehringer Mannheim test combinations [15,16] (n = 15; r = 0.9930, y = 1.0170x - 0.5724, p < 0.0001) as has been reported by Bolton et al [16]. Total PL content of the isopropanolic bile solutions was measured in 0.1 ml as lipid soluble phosphorus [17]. Only bile samples containing total bile acid concentrations of > 30 mm01 were included in this study. Total bile acid concentrations varied from 30-80 mmol. Lipid composition of bile was expressed as mole fraction of total bile acids, PLs and cholesterol. The lithogenic index was derived from the polynomial equations developed by Thomas and Hofmann [18] to describe the cholesterol solubility line as proposed by Hegardt and Dam [19], and Holzbach et al [20].
30
Fatty acid composition of Mary PLS
Methylesters of fatty acids were analysed by gas-liquid chromatography. A mixture of normal human fatty acids was used as standards (Sigma Chemical Corp., Poole, Dorset, UK; high grade purity > 99%). First 1 ml of diluted bile (1: 10; v/v) isopropanolic solution) was dried under a nitrogen stream at 40 “C. Then 0.5 ml CH,OH was added, mixed and then 0.2 ml of internal standard solution (2 mg of heptadecanoid acid in 2.5 ml benzene). Subsequently, 0.3 ml of benzene was added together with 0.5 ml of 2.2-dimethoxypropane (DMP) and ca 0.03 ml of concentrated HCl (38%) was pipetted into this solution. The reaction mixture was left overnight or for at least 6 h at room temperature to ensure complete hydrolysis and subsequent esterification. The extraction of fatty acid-methylesters was performed with 3 ml hexane (2 times) after addition of 2.5 ml of H,O. The hexane solution was dried under an nitrogen stream at 40 o C. After addition of 0.1 ml hexane a sample of 0.6 ~1 was injected into the gas-chromatograph. With the DMP-method both esterified (bound in PLs) and free fatty acids in bile are analysed [21]. Selected samples were also analysed by transesterification with methanolic base (0.5 mol/l) made from metallic sodium and dry methanol (Supelchem BV, Hilversum, The Netherlands) [22]. Only esterified fatty acids in bile react with methanolic base, but not the free fatty acids. By combining the results of the DMP-method and the methanohc base-method the free fatty acid fraction potentially present in bile could be estimated. The methanolic base-method was done as follows: 0.2 ml of internal standard solution was added to 0.3 ml of benzene and 0.5 ml of CH,OH together with 0.5 ml of DMP and 0.03 ml of concentrated HCl. This reaction mixture was left overnight and then evaporated. Subsequently, 1 ml of the isopropanolic bile solution was added. This again was evaporated and then dissolved in 1 ml of methanolic base and 0.5 ml of benzene. This solution was kept for 30 mm at 60 ‘C; after cooling to room temperature 2.5 ml of H,O was added and fatty acid methylesters were extracted twice with 3 ml of hexane. The hexane solution was dried under a nitrogen stream and further gas-chromatographic analysis was performed as described with the DMP-method. When we analysed the undiluted bile stored at - 20 o C with the methanolic baseand the DMP-method a considerable amount of free fatty acids was present. However, if we analysed the isopropanolic solution of bile stored at - 20 o C with the two methods no free fatty acids were present. These results indicate that in the undiluted bile samples after thawing some hydrolysis of PLs to free fatty acids did occur. The methylesters of fatty acids were separated by using a gas-chromatograph type Becker 420 (Packard-Becker) filled with a coiled all-glass-column (6 ft X 2 mm i.d. and 6 mm o.d.) packed with 10% SP 2310 on 100/120 chromosorb W-AW at 195 o C. The flow rate was 20 ml/mm. N,-gas was used as a carrier-gas. Percentages were corrected for detector response by using a known amount of internal standard in each sample.
31
Statistical methods
The data on biliary lipids and fatty acid composition (obtained by averaging the results of analyses of the four bile samples taken on four consecutive days in one person) were normally distributed and are presented as individual values with their means. Correlations of individual biliary bile acids (x) with PL-fatty acid composition (y), were calculated by the method of least-squares with associated confidence limits of 95% for the regression of y on X. Their significance was derived from Pearson’s correlation coefficient. Results Biliary fatty acid composition of PLs in hearthy individuals
The composition present in the 24 healthy individuals is shown in Table I. As should be expected palrnitic acid and linoleic acid were the predominant fatty acids present in PLs in bile. They accounted for approximately 75% of total fatty acids. Free fatty acids could not be detected. There was no significant difference in fatty acid-PL composition between males and females in our study group. Relation biliary bile acids and PL-fatty acid composition
In the 24 healthy individuals, we looked at possible correlations between individual biliary bile acid composition and PL-fatty acid composition. A negative correlation was found between the proportion of biliary cholic acid and that of biliary arachidonic acid (Fig. 1; r = - 0.52; p < 0.01) and also between the other primary bile acid chenodeoxycholic acid and arachidonic acid (Fig. 2; r = - 0.53; p < 0.01). On the other hand a significantly positive correlation was found between the secondary bile acid deoxycholic acid and arachidonic acid (Fig. 3; r = 0.71; p < 0.01). Opposite correlations as for arachidonic acid were found for linoleic acid. There was a positive correlation between the proportion of biliary chenodeoxycholic acid and that of linoleic acid (r = 0.48; p < 0.05) and a negative correlation between biliary deoxycholic acid and linoleic acid (Fig. 4; r = -0.68; p < 0.01). A positive
TABLE Biliary
I fatty acid composition
Biliary PL-fatty Type fatty acid Palmitic acid Pahnitoleic acid Steak acid Oleic acid Linoleic acid Linolenic acid Dihomo-y-linoleic Arachidonic acid
in PLs
acids
acid
Healthy individuals Mean % (*SD) 16:O 16:l 18:0 18:l 18:2 18:3 20:3 2014
41.40 2.68 5.50 12.09 32.83 Trace Trace 5.64
(1.41) (0.82) (1.55) (0.98) (3.06)
(1.59)
(n = 24)
32 Biliary PL-arachidonic acid (mol %I 10 _
?-
.
6-
4-
0 25
M
35
40
1 50
45
Biliary CA tmol %b)
Fig. 1. Relation between biliary cholic acid and arachidonic acid in healthy subjects (n = 24). Arachidonic acid composition (y, mol W) is expressed as function of biliary cholic acid composition (x, mol W). The solid line depicts the regression function generated by the least squares method. The correlation was significant ( r = - 0.52; p < 0.01). The curves depict the 95% confidence limits for the regression of y on X.
Biliary P L-arachidonic acid I mol k b
‘EI l-
.
:.:
.
. .
r
24 =-0.53
6-
Biliary COCA tmol %I Fig. 2. Relation between biliary chenodeoxycholic acid and arachidonic acid in healthy subjects (n = 24). Arachidonic acid composition (y, mol W) is expressed as function of biiary chenodeoxycholic acid composition (x, mol %). The solid line depicts the regression function generated by the least-squares method. Correlation was significant (r = -0.53; p -c 0.01). The curves depict the 95% confidence limits for the regression of y on x.
33 Biliary PL-arachidonic acid (mol %
1
Biliary
DCA
lmol
%)
Fig. 3. Relation between biliary deoxycholic acid and arachidonic acid in healthy subjects (n = Arachidonic acid composition (y, mol %) is expressed as function of deoxycholic acid composition mol W). The solid line depicts the regression function generated by the least-squares method. correlation was significant (r = + 0.71; p < 0.01). The curves depict 95% confidence limits for regression of y on X. Biliary acid
24). (x, The the
P L - linoleic
1mol
%)
.
“’
24
r = -0.68 .
\
251 0
5
10
15
20
25
I
35
30 Biliary
DCA
(mol%)
Fig. 4. Relation between biliary deoxycholic acid and linoleic acid in healthy subjects (n = 24). Linoleic acid composition (v, mol W) is expressed as function of deoxycholic acid composition (x, mol W). The solid line depicts the regression function generated by the least-squares method. The correlation was significant (r = -0.68; p < 0.01). The curves indicate the 95% confidence limits for the regression of y on x.
34 Biliary CDCA (mol%)
“=
24
r = -0.87
Eiliary
DCA (mol81
Fig. 5. Relation between biliary chenodeoxycholic and deoxycholic acid in healthy subjects (n = 24). Chenodeoxycholic acid composition (y, mol %) is showed as function of deoxycholic acid (x, mol R;). The solid line depicts the regression function generated by the least-squares method. The correlation was significant (r = -0.87; p < 0.01). The curves depict the 95%confidence limits for the regression of y on X.
correlation was found between biliary cholic acid and steak acid (r = 0.45; p -C0.05). Positive correlations were also found between deoxycholic acid and oleic acid (r = 0.44; p < 0.05) and between biliary cholate and linoleic acid (r = 0.44; p -C0.05). The opposite relationships of deoxycholic acid and chenodeoxycholic acid with arachidonic acid and linoleic acid were not surprising because a strong reciprocal relationship existed between the proportion of biliary chenodeoxycholic acid and that of deoxycholic acid (Fig. 5; r = -0.87; p -c 0.01). Finally, also a positive correlation was found between the cholesterol saturation index and biliary palmitic acid (r = 0.49; p < 0.05) and between the cholesterol saturation index and linoleic acid (r = 0.45; p -c 0.05). Discussion
Fatty acid composition of lecithin in gallbladder bile of the healthy individuals showed that palmitic acid and linoleic acid were the predominant fatty acids. Similar results have been obtained by other studies [2,3,10]. In the present study, it was shown that an inverse relationship did exist between the proportion of both primary bile acids in bile, cholate and chenodeoxycholate, and that of arachidonic acid. There was also a positive relationship between both primary biliary acids and linoleic acid. On the other hand deoxycholate showed the
35
opposite effect. A strong positive relationship of deoxycholate with arachidonate acid was shown and a negative relation with linoleic acid. Previous studies have shown that gallbladder bile of cholesterol gallstone patients shows an increased lecithin arachidonate composition and also a higher deoxycholate proportion in biliary bile acids [2,9]. Furthermore, Cantafora et al [9] observed an inverse relationship between arachidonate and chenodeoxycholate composition. Their study, however, was confined to gallstone patients. Healthy controls were not included. Our study extends their findings of an inverse correlation between chenodeoxycholate and arachidonate and the positive correlation between deoxycholate and arachidonate to the normal population, too. In view of the strong inverse relationship between chenodeoxycholate and deoxycholate in bile, the opposite relation of chenodeoxycholate and deoxycholate with linoleate and arachidonate does not seem surprising. This strong inverse relationship between chenodeoxycholate and deoxycholate in bile is likely to be explained by competition of intestinal transport. In our study, the correlations detected between deoxycholate and arachidonate and between deoxycholate and linoleate in bile, were stronger than for chenodeoxycholate. For this reason we believe that more the relationship between the proportion of biliary deoxycholate and biliary PL-fatty acid composition than the relationship between chenodeoxycholate and PL-fatty acid composition, is likely to be explained by acute coupling effects in induced biliary lipid secretion, Ahlberg et al [2], however, showed that chenodeoxycholic acid treatment in gallstone patients did induce phospholipid species in gallbladder bile containing less arachidonate and Salvioli et al [lo] showed in acute bile exchange experiments that both deoxycholate and chenodeoxycholate increased the secretion of bilk-y lecithin, rich in arachidonate and stearate. The findings of Salvioli et al [lo] are not completely confirmed by our study. As no actual biliary secretion of lipids was measured in our study, we have no proof of the validity of our data. On the other hand our studies were performed in the natural situation without any possible unphysiological manipulation. Although quantification of biliary PL molecular species could provide more useful information, this seems unlikely in relation to the arachidonic acid composition, since Ahlberg et al [2] showed that arachidonic acid was present in PC in human bile only in combination with palmitic acid (1-palmitoyl-2-arachidonyl-snglycero-phosphocholine). Several studies have shown an increased proportion of deoxycholate - generally held to be a saturating or unequivocally non-desaturating bile acid - in gallbladder bile of cholesterol gallstone patients [23-251. Also, the National Cooperative Gallstone Study demonstrated a positive correlation between biliary deoxycholate percentage and biliary cholesterol saturation [26]. The finding in our study, employing only healthy subjects, of a strong positive correlation between biliary deoxycholate and PC-arachidonate could have some implication for the pathogenesis of cholesterol gallstone disease. It could mean that a bile acid pool size rich in deoxycholate could preferentially promote the biliary secretion of phospholipid species rich in arachidonate from the hepatocytic cell membrane. It has been suggested from animal experiments that
36
arachidonate in gallbladder bile could stimulate prostaglandin synthesis in the gallbladder wall, which in turn could induce glycoprotein or mucin secretion by the mucosa [26,27]. Glycoproteins are incriminated in cholesterol gallstone formation because they can act as nucleating factors [28]. The fact that we did not find detectable amounts of free fatty acids in bile does not necessarily contradict this hypothesis, because only minute amounts of free arachidonic acid, below the detection limit of our method, will be necessary for an appreciable prostaglandin effect. Thus both the positive correlation between biliary deoxycholate and biliary cholesterol saturation and between biliary deoxycholate and arachidonate could have some relevance with respect to the pathogenesis of cholesterol gallstone disease. Our study also provides additional arguments for the hypothesis that bile acid pool size composition is important in the biliary secretion of PL species. Although it is well known that the biliary secretion of PLs quantitatively is dependent of total bile acid secretion, our study indicates also that especially the biliary secretion of PL species rich in arachidonate and linoleate is dependent of the secretion of the secondary bile acid deoxycholate. Acknowledgements
The authors are grateful to Mr. B.H. Jansen, Department of Nuclear Chemistry, Malberg-GZ Hospital, for the statistical analysis and to Mrs. Ineke Janssen-Volman for expert secretarial assistance. This work was supported in part by a ZWO/ FUNGO-grant and a grant from the Dutch Liver and Digestive Foundation (De Nederlandse Lever-Darn-i Stichting). References 1 Kwamoto T, Okana G, Akin0 T. Biosynthesis and turn-over of individual molecular species of phosphatidyl-choline in liver and bile. Biochim Biophys Acta 1980;619:20-34. 2 Ahlberg J, Curstedt T, Einarsson H, Sjiivall J. Molecular species of biliary phosphatidyl-choline in gallstone patients. J Lipid Res 1981;22:404-409. 3 Cantafora A, DiBiase A, Alvaro D, Angelic0 M, Marin M, Attili AF. High performance liquid chromatographic analysis of molecular species of phosphatidylcholine-development of quantitative assay and its application to human bile. Clin Chim Acta 1983;134:281-295. 4 Goodnight SH, Harris WS, Connor WE, Illingworth DR. Polyunsaturated fatty acids, hyperlipidemia and thrombosis. Arteriosclerosis 1982;2:87-113. 5 Rameska CS, Paul R, Ganerly J. Effect of dietary unsaturated oils on the biosynthesis of cholesterol and on biliary and fecal excretion of cholesterol and bile acids in rats. J Nutr 1980;110:2149-2158. 6 Balasabramaniam S, Simons LA, Chang S, Hi&e JB. Reduction in plasma cholesterol and increase in biliary cholesterol by a diet rich in n-3 fatty acids in the rat. J Lipid Res 1985;26:684-689. 7 Grundy SM, Abrens Jr. EH. The effects of unsaturated dietary fats on absorption excretion, synthesis and distribution of cholesterol in man. J Clin Invest 1970;49:1135-1152. 8 Grundy SM. Effects of polyunsaturated fats on lipid metabolism in patients with hypertriglyceridemia. J Clin Invest 1975;55:269-282. 9 Cantafora A, Angel& M, DiBiase A, Pieche U, Bracci F, Attili AF, Capocaccio L. Structure of biliary phosphatidyl-choline in cholesterol gallstone patients. Lipids 1981;16:589-592. 10 Salvioli G, Romani M, Loria P, Carulli M, Pradelli JM. Effect of acute administration of bile acids on fatty acid composition of biiary phosphatidylcholine in man. J Hepatol 1985;1:291-300.
31 11 Junker RL, Hassan AS, Subbiah MTR. Simultaneous quantitation of biliary cholesterol, bile acids and phospholipids and fatty acids, by gas-liquid chromatography. Clin Chem 1981;27:1779-1780. 12 Van Berge Henegouwen GP, Ruben AT, Brandt K-H. Quantitative analysis of bile acids in serum and bile, using gas-liquid chromatography. Clin Chim Acta 1974;54:249-261. 13 Huijbregts AWM, Van Schaik A, Van Berge Henegouwen GP, Van der Werf SDJ. Serum lipids, biliary lipid composition and bile acid metabolism in vegetarians as compared to normal controls. Eur J Clin Invest 1980;10:443-449. 14 Shaw R, Elliot WH. Bile acids. LV, 2.2-Dimethoxypropane: an esterifying agent preferred to diazomethane for chenodeoxycholic acid. J Lipid Res 1978;19:783-787. 15 Fromm H, Amin P, Klein M, Kupke I. Use of a simple enzymatic assay for cholesterol analysis in human bile. J Lipid Res 1980;65:259-261. 16 Bolton CH, Nicholls JS, Heaton KW. Estimation of cholesterol in bile. Clin Chem Acta 1980;105:225-230. 17 La Russo NF, Hoffman NE, Hofmann Northfield TCN, Thistle JL. Effect of primary bile acid ingestion on bile acid metabolism and biliary lipid secretion in gallstone patients. Gastroenterology 1975;69:1301-1314. 18 Thomas PJ, Hofmann AF. A simple calculation of the lithogenic index of bile: expressing biliary composition on rectangular coordinates. Gastroenterology 1973;65:698-700. 19 Hegardt FG, Dam H. The solubility of cholesterol in aqueous solution of bile salts and lecithin. Z Emahnmgswiss 1971;10:223-233. 20 Holzbach RT, Marsh M, Olszewski M, Holan K. Cholesterol solubility in bile: evidence that supersaturated bile is frequent in healthy man. J Clin Invest 1973;52:1467-1479. 21 Mason ME, Waller GK. Dimethoxypropane induced transesterification of fats and oils in preparation of methyl esters for gaschromatographic analysis. Anal Chem 1964;36:583-586. 22 Luddy EF, Barford BA, Herb SF, Magidman P. A rapid and quantitative procedure for the preparation of methyl esters of butter oil and other fats. J Am Oil Chem Sot 1968;45:549-552. 23 Ahlberg J, Angelin B, Einarsson K. Influence of deoxycholic acid on biliary lipids in man. Clin Sci 1977;53:249-256. 24 Pomare EW, Heaton KW. Bile salt metabolism in patients with gallstones in functioning gallbladders. Gut 1973;14:885-890. 25 Hofmann AF, Grundy SM, La&in JM, et al and the NCGS-group. Pretreatment biliary lipid composition in white patients with radiolucent gallstones in the National Cooperative Gallstone Study. Gastroenterology 1982;83:738-752. 26 LaMont JT, Turner BS, Dibenedetto D, Handin R, Schafer AI. Aracbidonic acid stimulates mucin secretion in prairie dog gallbladder. Am J Physiol 1983;245:992-998. 27 Lee SP, LaMont SP, Carey MC. Role of the gallbladder mucus hypersecretion in the evolution of cholesterol gallstones. J Clin Invest 1981;67:1712-1723. 28 Smith BF, LaMont JT. Identification of gallbladder mu& bilirubin complex in human cholesterol gallstone matrix. J Clin Invest 1985;76:439-445.