Regulation of Bile Acid Synthesis in Humans: Effect of Treatment with Bile Acids, Cholestyramine or Simvastatin on Cholesterol 7a-Hydroxylation Rates In Viuo MARC0

BERTOLOTTI,~ NICOLA.bATE,l PAOLA LO RIA,^ MICHELEDILENGITE,~ FRANCESCA CARUBBI,' ADRIANO PINETTI,2 ANTONIA DIGRISOL03 AND NICOLACARULLI'

'Zstituto di Patologia Medica, 2Zstituto di Chimica Organica and 3Centro Antidiabetico, Universita di Modena, 1-41100 Modena, Italy

The rates of cholesterol 7a-hydroxylation (the first and rate-limiting step of bile acid synthesis from cholesterol)were evaluated in uivo in patients administered bile acids with different structural properties, cholestyramineor simvastatin, a competitiveinhibitor of 3-hydroxy-3-methylglutaryl-coenzyme A reductase. Twenty-three subjects, with normal hepatic and intestinal functions, were studied in basal conditions and after one of the following treatment schedules,lasting 4 to 6 weeks: cholestyramine, 4 and 12 w d a y (four patients); ursodeoxycholicacid, 9 to 11mg/kg/day(four patients); chenodeoxycholic acid, 12 to 15 mg/kg/day (fivepatients);deoxycholic acid, 8 to 10mg/kg/day(four patients); and simvastatin, 40 mg/day (six patients). 7a-Hydroxylation of cholesterol was assayed by measuring the increase in body water tritium after intravenous bolus of cholesterol tritiated at the 7a position. Plasma bile acid composition, evaluated by gas-liquid chromatography,revealed a substantial enrichment of the recirculating pool by the administered bile acid, whereas treatment with cholestyramine decreased the content of dihydroxylated bile acids. Cholesterol 7ahydroxylation increased in a dose-related manner after cholestyramine, in parallel with a decrease of cholesterol in total plasma and low-density lipoproteins (1.006 to 1.063 gndml). Hydroxylation rates decreased by an average of 47% with chenodeoxycholic acid and by an average of 78% with deoxycholic acid; ursodeoxycholic acid treatment did not affect 7ahydroxylation significantly. Simvastatin markedly reduced plasma total and low-density lipoproteincholesterol but exerted no change on 7a-hydroxylation rates. Our results support the existence of a feedback inhibition exerted on cholesterol 7a-hydroxylation (and consequently on bile acid synthesis) by hydrophobic bile acids returning to the liver through the enterohepatic circulation. The finding emphasizes the importance of the physicochemical properties of

Received February 11, 1991; accepted July 2, 1991. This work was supported in part by the Research Funds (60%) of the University of Modena. Address reprint requests to: Marco Bertolotti, M.D., Istituto di Patologia Medica-Clinica Medica I, Policlinico, Via del Pozzo, 71,I-41100 Modena, Italy. 31/1/32218

bile acids in the regulation of hepatic cholesterol balance. Under these experimental conditions, inhibition of 3-hydroxy-3-methylglutaryl-coenzyme A reductase and presumably reduced availability of newly synthesized cholesterol are not critical for bile acid 1991;14:830-837.) synthesis. (HEPATOLOGY

Hepatic biotransformation of cholesterol to bile acids is a major mechanism whereby cholesterol is eliminated from the organism (1).The fist and rate-limiting step of this process is hydroxylation at the 7a position of the sterol nucleus, catalyzed by a microsomal P450dependent enzyme, cholesterol 7a-hydroxylase (2). The enzyme has recently been purified from rat liver and cloned (3-5). The intimate regulation of the overall synthetic process and its rate-limiting event is incompletely understood. Classically, bile acids returning to the liver through the enterohepatic circulation were considered as the effectors of a feedback control on the biosynthetic pathway (2,6).The alternate hypothesis was later raised that the availability of newly synthesized microsomal cholesterol could represent the ultimate regulator of bile acid synthesis (7). The hypothesis was supported by the lack of inhibitory effect exerted by bile acids on cultured liver cells (7, 8). More recent reports in the animal in vivo also yielded conflicting results. As a general rule, hydrophobic bile acids appeared to inhibit bile acid synthesis and 7a-hydroxylation rates more than hydrophilic ones (9-11). Even if some data in humans seemed to support the feedback hypothesis (121, research in man has been limited by methodological problems: exogenous administration of bile acids can, in fact, interfere with evaluation of bile acid kinetics according to isotope dilution techniques (13). On the other hand, sterol balance analysis is cumbersome and imprecise, requiring complex laboratory workup and control of alimentary sterol intake (14). We recently validated a methodological approach for the in uivo evaluation of cholesterol 7a-hydroxylation rates in humans. This technique is based on the tritium

830

83 1

CHOLESTEROL 701-HYDROXYLATION IN HUMANS

Vol. 14, No. 5, 1991

TABLE 1. Relevant clinical data of the patients examined Patient

Weight

(Yr)

43

% Ideal"

Cholesterol (mg/dl)

Triglyceride (mg/dl)

46 55 45 55 59 53 35 41 46 47 60 41 37 39 32 42 34 60 48 68 61 36 47

68 57 73 61 72 93 69 70 86 71 87 70 56 85 100 60 74 78 67 62 63 75 64

97 100 129 113 105 122 99 107 130 113 116 101 102 118 125 103 106 92 93 91 109 103 91

337 276 249 194 25 1 179 171 304 235 405 214 230 162 273 231 343 247 396 356 265 407 322 370

101 61 151 117 145 122 62 620 725 585 238 382 105 240 314 54 244 57 67 247 81 163 111

Age

Sex

Diagnosis

~

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

M F F F M F M M M M M M M M M F M M M F F M F

Hypercholesterolemia Hypercholesterolemia Unexplained pruritus Unexplained pruritus Gastritis Hepatic fibrosis Diabetes Combined hyperlipidemia Hypertriglyceridemia Combined hyperlipidemia Hypertrigly ceridemia Hypertriglyceridemia Diabetes Combined hyperlipidemia Hypertriglyceridemia Hypercholesterolemia Hypertriglyceridemia Hypercholesterolemia Hypercholesterolemia Combined hyperlipidemia Hypercholesterolemia Hypercholesterolemia Hypercholesterolemia

"Calculated as: weight (kg)/[height (cm) - 1001 x 100.

release method first described by Van Cantfort et al. (15) and later used in this laboratory under in uitro (16) and in uiuo (17, 18) conditions. The technique does not necessitate duodenal intubation or exogenous administration of bile acids. We used this procedure to test the effect of bile acids with different degrees of hydrophobicity, of the bile acid-binding resin cholestyramine and simvastatin, a competitive inhibitor of 3-hydroxy-3methylglutaryl CoA (HMG-CoA) reductase, on cholesterol 7a-hydroxylation. Inhibitors of HMG-CoA redudase, the limiting step of cholesterol synthesis, are increasingly used as hypocholesterolemic agents (191, and their pharmacological effect is likely related to reduced cholesterol synthesis (particularly in the liver). Simvastatin is a derivative of lovastatin, the bestcharacterized drug of this class, with analogous properties (20). Our results support the occurrence of feedback inhibition on cholesterol 7a-hydroxylation in man, exerted by more hydrophobic bile acids. Long-term inhibition of HMG-CoA reductase, on the contrary, does not affect this metabolic step. PATIENTS AND METHODS Putients. A total of 23 subjects (aged 32 to 68 yr; 8 women and 15 men) admitted to the Istituto di Patologia Medica of the University of Modena as inpatients or as day-hospital patients were studied. Table 1 shows the relevant clinical data of the patients. Physical examination and laboratory evaluation showed normal hepatic, intestinal and thyroid functions. The patients were instructed to follow individual diets containing

approximately 400 mg cholesteroUday.When a hypoglycidic or a hypolipidic regimen was administered, such as in diabetic or in hyperlipidemic patients, investigation was started after at least 1 mo, when a stable metabolic situation was achieved. Diabetic patients were not receiving insulin. Body weight remained constant throughout different evaluations in the same subject. Patients gave their consent to the study protocol, which was approved by the Ethical Committee of the University of Modena. AU patients were studied in pretreatment conditions. Then, four subjects (Nos. 1 to 4) had the study repeated after treatment with both low dosage (4 gm/day as a single dose) and medium-high dosage (12 &day in three doses) cholestyramine, over 4 to 6 wk, in random order. The drug was administered as 4 gm packages 15 min before meals. Four patients (Nos. 5 to 8)were reevaluated after treatment with ursodeoxycholicacid (9 to 11 mg/kg/day), 5 patients (Nos. 9 to 13) were reevaluated after treatment with chenodeoxycholic acid (12 to 15 m&g/day) and 4 patients (Nos. 14 to 17) were reevaluated after treatment with deoxycholic acid (8 to 10 mg/kg/day). Bile acids were administered in two or three daily doses. Treatment regimens with bile acids lasted 4 to 6 wk. Finally, six patients (Nos. 18 to 23) were studied before and after treatment with simvastatin. This was started at a lower dose (10 to 20 mg/day) and gradually increased over a 2- to 4-wk period to the final dosage of 40 mg/day (as a single bedtime dose), which was maintained through 6 wk before repeating the evaluation. At least 2 mo elapsed between different evaluations in the same subject. Patients underwent clinical and routine laboratory controls at the end of each treatment period. Matericrh. Synthesis of [7a-3H]cholesterol(specificactivity, 3 to 10 mCi/mmol) and assay of label stereospecificityat the 7a

832

HEPATOLOGY

BERTOLOTTI ET AL.

TABLE 2. Effect of treatment with different bile acids on the

TABLE 3. Effect of treatment with cholestyramine on plasma

rates of cholesterol 7a-hydroxylation

bile acid composition

Patient no.

5 6 7 8 Mean k S.D. treated with UDCA (n = 4) 9 10 11 12 13 Mean ? S.D. treated with CDCA (n = 5) 14 15 16 17 Mean t S.D. treated with DCA (n = 4)

Bile acid administered

Rates of cholesterol 7cr-hvdroxvlation imgld& Before

After

UDCA UDCA UDCA UDCA

295 424 458 323 375 t 78

318 530 440 292 395 t 111

CDCA CDCA CDCA CDCA CDCA

465 554 785 529 292 525 2 178

291 248 264 294 199 259 +- 39"

Pretreatment Cholestyramine (4 gm/day) Cholestyramine (12 gm/day)

CA (%)

CDCA (%)

DCA (%)

35 t 8 47 k 8"

43 t 8 35 t 8b

22 2 4 18 t 6

61

27 t l l b

12 t 5"

5

12"

Four patients (Nos. 1 to 4) were treated for 4 to 6 wk with both dosage regimens of cholestyramine. Data indicate the mean 2 S.D. CA = cholic acid; CDCA = chenodeoxycholic acid; DCA = deoxycholic acid. "p < 0.05 vs. pretreatment, Student's t test for paired data. bp < 0.01 vs. pretreatment, Student's t test for paired data.

hr thereafter in all patients. Blood or urine samples were collected every 12 hr from tracer administration for 5 to 6 days. DCA 405 99 Plasma samples were used for evaluation of cholesterol DCA 472 56 specific activity. After alkaline extraction, cholesterol radioacDCA 328 95 tivity was measured by liquid scintillation, and cholesterol DCA 462 111 mass was assayed according to Abell et al. (26). The specific 417 ? 66 90 & 24' activity could be determined as the radioactivity/mass ratio. The radioactivity time course of body water was evaluated using 8 to 10 samples of erythrocytes or urine from each study. Patients were treated for 4 to 6 wk with either 9 to 11 mgiday The samples were distilled with a device analogous to that ursodeoxycholic acid (UDCA), 12 to 15 mgiday chenodeoxycholic acid described by Hutton et al. (271, and the distillate was assayed (CDCA)or 8 to 10 mg/day deoxycholic acid (DCA).Results obtained in for radioactivity. Patients 5 , 8 , 9 , 1 0 and 11 were already presented in a previous report In an initial subgroup of six patients, body water volume was from this laboratory (17). assayed directly by isotope dilution, administering a bolus of "p < 0.05 vs. pretreatment, Student's t test for paired data. tritiated water after completion of the study (17).The volumes bp < 0.01 vs. pretreatment, Student's t test for paired data. were consistently close to the 60% of ideal body weight; therefore the latter volumes were used in the remaining patients. Water turnover was investigated in two patients by measuring tritium decay after oral administration; body water position were performed as described (21,22). Cholestyramine half-life equaled 8 and 10 days. We assumed this turnover rate was administered as 4 gm packages (Questran, Bristol Italiana, to be too slow to interfere with the evaluation of body tritium Rome, Italy). Bile acids were administered as 150, 250 or 300 increment, which was calculated over a 12-hr interval. Body water tritium was plotted vs. time, and tritium mg gelatine capsules. Chenodeoxycholic and ursodeoxycholic acids were supplied by Gipharmex (Milan, Italy) and increment was determined by interpolation using nonlinear Sanofi-Midy(Milan,Italy). Deoxycholicacid was obtained from regression analysis. Usually, the interval chosen was between Sigma Chemical Co. (St.Louis, MO). Simvastatin was provided 60 and 72 hr after infusion (17). Cholesterol 7a-hydroxylation by Merck, Sharp & Dohme Italiana (Rome, Italy) as 10 or 20 was calculated as the ratio between body water tritium mg tablets. All reagents were analytical grade and were enrichment and plasma cholesterol specific activity in the same interval: purchased from Car10 Erba (Milan, Italy). Increase of body water 3H(cpm/l2 hr) x 2 Evaluation of Cholesterol 7wHydroxylation. The rates of Plasma cholesterol specific activity (cpm/mg) cholesterol 7a-hydroxylationwere evaluated by in uiuo tritium Hydroxylation rates were expressed as the amount of release assay (17). Cholesterol 7a-hydroxylationis a stereospecific reaction by which a hydrogen at the 7a position of the cholesterol undergoing 7a-hydroxylation per day. A correction sterol nucleus is replaced by a hydroxyl group (23). Therefore, was made to take into account the degree of radiolabel after administration of [7a-3Hlcholesterol, tritium release is stereospecificity,which ranged from 70% to 86% in this study. proportional to the activity of 7a-hydroxylation, which can be When a subject was reevaluated after treatment, the same quantitated by measuring the increment of body water batch of labeled cholesterol and the same protocol of sampling radioactivity. This approach is similar to that used by collection were used. Body water volume was assumed to be Rosenfeld et al. (24), who used [24,25-3H]cholestero1to assay constant throughout the study. Plasma Lipids, Lipoproteins and Bile Acids. Blood samples bile acid synthesis from the cleavage of the cholesterol side were obtained in fasting conditions in heparinized tubes. chain. After an overnight fast, patients were infused intravenously Plasma cholesterol and triglyceride concentrations were defor 10 to 15 min with trace amounts of [7a-3Hlcholesterol(200 termined by standard enzymatic technique. Lipoprotein choto 400 pCi) dissolved in 50 ml of homologous plasma or human lesterol was assayed on the different density fractions albumin. Whole body radiation exposure caused by tritium ( < 1.006, 1.006 to 1.063 and > 1.063 gm/ml) obtained after administration (25) was 25 to 50 mrad. Blood samples were sequential ultracentrifugation of plasma samples. The 1.006to collected in heparinized tubes before the infusion and 60 to 72 1.063 gm/ml fraction is largely made of low-density lipoprotein

833

CHOLESTEROL 7n-HYDROXYLATION IN HUMANS

Vol. 14, No. 5, 1991

ILL U DCA

COCA

800 ..

. .

fi_ DCA

FIG.1. Percent changes in cholesterol 7a-hydroxylation rates, compared with pretreatment, after ursodeoxycholic acid IUDCA), chenodeoxycholicacid (CDCAI and deoxycholic acid (DCA). (SeeTable 2 for single values and for explanations.) 0 = single data points: burs mean values.

2oo 100 0

E 0

y

(LDL) and to a lesser extent of intermediate density lipoprotein. For simplicity, it is referred to here as LDL. Recirculating bile acid pool composition was determined on plasma samples by gas-liquid chromatography (28) on a Fractovap 4200 instrument (Carlo Erba) using 3a-7a-12-ketocholanoic acid as an internal standard. A composite hydrophobicity index could be calculated for each patient before and after bile acid treatment, as the weighted sum of the indexes of the different bile acids in the pool (29).Because the proportion of glycine and taurine conjugated was not determined by our method, the hydrophobicity index of the unconjugated salt was used. Statistical Evaluation. Data were expressed as the mean t S.D. Statistical analysis was made according to Student's t test for paired data. Linear regression analysis was performed by the least-squares method.

RESULTS Table 2 shows the values of cholesterol 7ahydroxylation in the subjects investigated before and after treatment with different bile acids. Ursodeoxycholic acid did not affect hydroxylation rates. Cholesterol 7a-hydroxylation was reduced after chenodeoxycholic acid and, even more markedly, after deoxycholic acid. These alterations are illustrated in Figure 1, which shows percent changes of 7a-hydroxylation rates compared with pretreatment. After treatment, the administered bile acid became predominant in the recirculating pool, as estimated by gas-liquid chromatography of plasma samples. Deoxycholic and chenodeoxycholic acids represented 70% to 85%of total bile acids, whereas percent concentrations of ursodeoxycholic acid ranged from 48% to 60%'.As shown in Figure 2, a significant inverse correlation was present between cholesterol 7a-hydroxylation rates and the composite hydrophobicity index of the bile acid pool (29). No significant changes of cholesterol concentration in total plasma and the LDL (1.006 to 1.063 gm/ml) fraction were detected after treatment with ursodeoxycholic and chenodeoxycholic acids, whereas cholesterol levels decreased after deoxycholic acid (total

0

0

a

.

0.40 0.60 Hydrophobicity index

0.20

0.80

FIG.2. Correlation between cholesterol 7u-hydroxylation rate and the hydrophobicity index of the bile acid pool in 13 patients (Nos. 5 to 17) treated with different bile acids. 0 = pretreatment (all values); = after ursodeoxycholic acid; = after chenodeoxycholic acid; A = after deoxycholic acid. A composite hydrophobicity index was calculated as described (29) using individual indexes of unconjugated bile acids. Equation of the regression line (n = 26): y = 509 - 389x; r = -0.44:p < 0.05

2500

I .

22so:

/

BEFORE

4

9/d

1 2 g/d

FIG. 3. Effect of cholestyramine treatment on cholesterol 7ahydroxylation rates. Cholestyramine was administered at the daily doses of 4 and 12 &day for 4 to 6 wk to four patients (Nos. 1to 4). Pretreatment and high-dose data points for Patient 3 were already presented (17). = single data points; bars mean values. *p c 0.05 vs. low dose, Student's t test for paired data; **p i0.01 vs. pretreatment, Student's t test for paired data. y

cholesterol- before, 263 ? 80 mg/dl; deoxycholic acid, 228 & 40 mg/dl; LDL-cholesterol-before, 185 t 30 mg/dl; deoxycholic acid, 157 t 25 mg/dl). The effect of different doses of cholestyramine on cholesterol 7a-hydroxylation is shown in Figure 3. Cholestyramine induced a dose-related increase of hydroxylation rates (values before treatment: 454 t 173 mg/day; low dose, 1,151 t 167 mg/day; high dose, 1,670 ? 304 mg/day). This effect was paralleled by a decrease of circulating total and LDL-cholesterol (Fig. 4).Plasma total cholesterol decreased from 269 2 59 to

834

BERTOLOTTI ET AL.

HEPATOLOGY

0

I BEFORE

SIMVASTATIN

FIG.6. Effect of simvastatin on cholesterol 7a-hydroxylationrates (see Fig. 5 for details). 0 = single data points; bars = mean values. p = not significant, Student’s t test for paired data. 0 1

I

I

BEFORE

I 4

1

o/d

0

.

1

12 d d

FIG.4. Effect of cholestyramine treatment on cholesterol concentration in total plasma (upperpanel) and in LDL flowerpanel) (see Fig. 3 for details).LDL represents the 1.006 to 1.063 gm/ml density fraction after ultracentrifugation.0 = single data points; burs = mean values. *p < 0.05 vs. pretreatment and vs. low dose, Student’s t test for paired data; **p < 0.01 vs. pretreatment and between the two doses, Student’s t test for paired data. 500

TOTAL 450 400

-c

._

350

0

c

300

L B cv = Y

g>

p r c -

250

200

r 0

150 100 50

0

k

BEFORE

AFTER

LDL

.

BEFORE

AFTER

FIG.5. Effect of simvastatin on plasma total and LDL-cholesterol concentration.Six patients (Nos. 18 to 23)were treated with 40 mg/day simvastatinfor 6 wk. LDL represents the 1.006 to 1.063 gm/ml density fraction after ultracentrifugation. 0 = single data points; bars = mean values. *p < 0.01 vs. pretreatment, Student’s t test for paired data.

237 & 40 mg/dl (low dose) and to 208 t 49 mg/dl (high dose), whereas the values for LDL-cholesterol were, respectively, 178 & 44, 157 & 44 and 134 ? 48 mg/dl. Table 3 shows the plasma bile acid patterns before and after cholestyramine. Plasma concentration of dihydroxylated bile acids, chenodeoxycholic and deoxycholic acids tended to be lower during treatment with the lower dose, and even more markedly after the higher dosage. Percent cholic acid concentration increased after cholestyramine.

Simvastatin treatment significantly reduced plasma cholesterol levels as shown in Figure 5. Total cholesterol decreased from 353 f 52 to 247 & 77 mg/dl, and LDLcholesterol fell from 247 f 77 to 140 & 48 mg/dl. Cholesterol 7a-hydroxylation rates were unaffected by treatment with simvastatin as illustrated in Figure 6 (before treatment = 380 & 99 mg/day; after simvastatin = 390 f 124 mg/day). No correlation was detected between changes in plasma total or LDL-cholesterol and changes in 7a-hydroxylation rates. Circulating bile acid composition was not affected by simvastatin treatment. Cholesterol levels in the remaining lipoprotein fractions and plasma triglycerides were unaffected by either treatment. No major side effects were noticed with any drug. At the end of treatment, among the four patients who received deoxycholic acid, one patient complained of mild diarrhea and two patients had slight hypertransaminasemia (1.5 to 2 times the upper normal values). In no case was modification of the treatment schedule necessary, and these alterations quickly disappeared after suspending treatment. DISCUSSION In this study, a new in uiuo procedure, which proved

to be relatively simple, reproducible and well tolerated, was used to investigate the importance of bile acid feedback and HMG-CoA reductase activity on the regulation of bile acid synthesis in humans. In particular, the rate of its limiting step, 7a-hydroxylation, was related to the physicochemical property of hydrophobicity (30) of the bile acid pool. We first considered possible sources of errors with this technique. The method assumes that labeled cholesterol mixes homogeneously with the hepatic microsomal pool accessible to 7a-hydroxylase. Indeed exogenous cholesterol was shown to mix very quickly into a “rapidly exchangeable” pool, including plasma and liver cholesterol (31).Furthermore, our technique has been able to detect changes in 7a-hydroxylation rates even in condi-

Vol. 14,No.5, 1991

CHOLESTEROL 7a-HYDROXYLATIONIN HUMANS

tions such as cholestyramine treatment, when the contribution of newly synthesized cholesterol to the substrate pool for bile acid production should increase and the risk of underestimating bile acid synthesis should be greater (32). We believe that important errors resulting from incomplete mixing of preformed and newly synthesized hepatic cholesterol can be reasonably ruled out in this study. We also considered the problem of partial labeling of [3H]cholesterol at the 7p position. In a previous paper, we estimated that metabolism of [7p-3Hlbile acids could release an amount of tritiated water corresponding to less than 10% of the increment in body water radioactivity in the 12-hr interval considered (17). Therefore it should not significantly interfere with the calculated 7a-hydroxylation rates, provided a correction is made for the degree of stereospecificity. Finally, the contribution of tritiated water to cholesterol synthesis can be considered negligible for the purposes of this study (1,17, 24). This technique could previously detect changes in cholesterol 7a-hydroxylation in conditions known to affect bile acid synthesis. In particular, hydroxylation rates were decreased in patients with advanced cirrhosis (17), and preliminary results showed an inverse correlation with age (18).Also, unpublished reports from this laboratory have shown a good correspondence between this technique and isotope dilution analysis (less than 15% difference in four control subjects). The present findings show a reduction of bile acid synthesis when the recirculating pool is mainly composed of hydrophobic bile acids (Table 2, Figs. 1 and 2). Our results are in agreement with previous data in humans, both in uiuo (33-35)and in uitro (36, 37), even if the effect of all three bile acids was never investigated in a single study. This finding also substantiates recent data in the animal (9-11, 29). It cannot be excluded that some of our patients were affected by familial hypertriglyceridemia, a condition that was reported by Angelin et al. (38)to be associated with increased bile acid synthesis. In those patients, cholesterol 7a-hydroxylation rates were consistently higher than in our subjects (see Table 2). Moreover, response to treatment was rather uniform in each treatment group. Therefore we assume that this possibility should bring no relevance to the interpretation of our data. Treatment with cholestyramine significantly increased the rates of cholesterol 7a-hydroxylation (see Fig. 3) in agreement with previous studies in uiuo and in uitro (12, 37). Our results indicate a dose-related effect accompanied by a parallel decrease in plasma total and LDL-cholesterol (as illustrated in Fig. 4) that supports a relationship between increased bile acid production and enhanced expression of hepatic LDL-receptors (39). The effect of cholestyramine on bile acid synthesis and cholesterol 7a-hydroxylation is probably due to reduction of the amount of bile acids recirculating to the liver. Postprandial serum bile acids were shown to be reduced after cholestyramine treatment (40). Our re-

835

sults on plasma bile acids (Table 3) also confirm the existence of a different bile acid pool composition. The data support the hypothesis that both quantitative reduction and qualitative changes (decrease of dihydroxylated, more hydrophobic bile acids) of the recirculating pool can contribute to the effect of cholestyramine (37). The mechanism responsible for the inhibitory effect of bile acids on cholesterol 7a-hydroxylation is unknown. Because regulation of 7a-hydroxylase activity was shown to occur at the transcriptional level (4,5,41),we may hypothesize that hydrophobic bile acids, which cross lipid membranes more easily (42), can interact with regulatory regions in the cell nucleus. In addition, alterations in the turnover rate of 7a-hydroxylase mRNA (5) or enzyme protein (37) or posttranslational events caused by direct interaction with microsomal membranes could play a role as well. More direct studies are clearly needed in this respect. Treatment with chenodeoxycholic acid was previously shown to increase plasma total and LDL-cholesterol, possibly as a consequence of reduced bile acid synthesis (43). Because of the small number of patients, no changes could be detected in this study. A decrease in plasma total and LDL-cholesterol levels was observed with deoxycholic acid. Decreased intestinal absorption and hepatic synthesis of cholesterol may account for these changes, even if the conversion of cholesterol to bile acids is reduced (44). Finally, in our experimental conditions, inhibition of HMG-CoA reductase did not affect cholesterol 7ahydroxylation in uiuo, in agreement with recent data in uitro (45) and with the isotope dilution technique (46). We do not know whether treatment with HMG-CoA reductase inhibitors actually reduces cholesterol synthesis in man. After treatment with reductase inhibitors, compensatory increases of both HMG-CoA reductase and LDL-receptor mRNAs were described (47). Whereas activation of LDL-receptors clearly accounts for its hypocholesterolemic action, increased activity of HMG-CoA reductase (45, 47, 48) could attenuate the inhibitory effect on cholesterol synthesis. No effects on cholesterol turnover rate were reported after very long-term treatment with lovastatin (49). Instead, other data supported the hypothesis of a reduced cholesterol synthesis in uiuo in the body and presumably in the liver (45, 46, 50, 51). The discrepancy could be due to differences in the drug administration period or in the evaluation of cholesterol synthesis. It is indeed possible that subtle changes in the amount of hepatic newly synthesized cholesterol might go undetected with most techniques. In the assumption that these drugs ultimately inhibit hepatic cholesterol synthesis, from our data it can be derived that the availability of newly synthesized cholesterol is not critical for bile acid synthesis. This is consistent with the recent data showing high saturation of cholesterol 7a-hydroxylation in human microsomes (52). After treatment with cholestyramine, the degree of saturation decreases, and the amount of cholesterol

836

BERTOLOTTI ET AL.

available as a substrate may become determinant. This could explain why bile acid production and cholesterol 7a-hydroxylaseactivity were reduced after acute administration of HMG-CoA reductase inhibitors to bile fistula rats (53) or to T-tube patients (P. Loria et al., Unpublished observations from this laboratory, 1990); that is, in conditions of derepressed bile acid synthesis. After long-term treatment in patients with intact enterohepatic circulation, compensatory mechanisms can instead restore an intracellular cholesterol content compatible with normal biosynthetic functions. Research is presently ongoing in this laboratory to assay the effect of reductase inhibitors during cholestyramine treatment as a model of “chronically” derepressed bile acid synthesis. In conclusion, our data support the hypothesis of a feedback control of bile acid synthesis and emphasize the importance of the physicochemical properties of the bile acid pool in the regulation of hepatic cholesterol homeostasis. On the contrary, inhibition of HMG-CoA reductase does not appear to affect bile acid synthesis in physiological conditions. REFERENCES 1. Turley SD, Dietschy JM. The metabolism and excretion of cholesterol by the liver. In: Arias IM,Jakoby WB, Popper H, Schachter D, Shafritz DA, eds. The liver: biology and pathobiology. New York Raven Press, 1988:617-641. 2. Myant NB, Mitropoulos KA. Cholesterol 7a-hydroxylase. J Lipid R ~ 1977;18:135-153. s 3. Noshiro M, Nishimoto M, Morohashi K, Okuda K. Molecular cloning of cDNA for cholesterol 7a-hydroxylase from rat liver microsomes. FEBS Lett 1989;257:97-100. 4. Jelinek DF, Andersson S, Slaughter CA, Russell DW. Cloning and regulation of cholesterol 7a-hydroxylase, the rate-limiting enzyme in bile acid biosynthesis. J Biol Chem 1990;265:8190-8197. 5. Li YC, Wang DP, Chiang Jn.Regulation of cholesterol 7uhydroxylase in the liver: cloning, sequencing, and regulation of cholesterol 7a-hydroxylase mRNA. J Biol Chem 1990;265:1201212019. 6. Shefer S, Hauser S, Bekersky I, Mosbach EH. Biochemical site of regulation of bile acid synthesis in the rat. J Lipid Res 1970;ll: 404-411. 7. Davis RA, Highsmith WE, Malone-McNeal M, ArchambaultSchexnayder J , Kuan J-CW. Bile acid synthesis by cultured rat hepatocytes: inhibition by mevinolin but not by bile acids. J Biol Chem 1983;258:4079-4082. 8. Kubaska WM,Gurley EC, Hylemon PB, Guzelian PS, Vlahcevic ZR. Absence of negative feedback control of bile acid biosynthesis in cultured rat hepatocytes. J Biol Chem 1985;260:13459-13463. 9. Stange EF, Scheibner J , Ditschuneit H. Role of primary and secondary bile acids as feedback inhibitors of bile acid synthesis in the rat in vivo. J Clin Invest 1989;84:173-180. 10. Shefer S, Nguyen L, Salen G, Batta AK, Brooker D, Zaki FG, Rani I, et al. Feedback regulation of bile-acid synthesis in the rat: differing effects of taurocholate and tauroursocholate. J Clin Invest 1990;85:1191-1198. 11. Vlahcevic ZR, Heuman DM, Hylemon PB. Regulation of bile acid synthesis. HEPATOLOGY 1991;13:590-600. 12. Einarsson K, Hellstrom K, Kallner M. Feedback regulation of bile acid formation in man. Metabolism 1973;22:1477-1483. 13. Lindstedt S. The turnover of cholic acid in man. Acta Physiol S-d 1957;40:1-9. 14. Hardison WGM, Grundy SM. Effect of ursodeoxycholate and its taurine conjugate on bile acid synthesis and cholesterol absorption. Gastroenterology 1984;87:130-135. 15. Van Cantfort J, Renson J , Gielen J. Rat liver cholesterol

HEPATOLOGY

7a-hydroxylase. I. Development of a new assay based on the enzymic exchange of the tritium located on the 7a position of the substrate. Eur J Biochem 1975;55:23-31. 16. Carulli N, Ponz de Leon M, Zironi F, Pinetti A, Smerieri A, Iori R, Loria P. Hepatic cholesterol and bile acid metabolism in subjects with gallstones: comparative effects of short-term feeding of chenodeoxycholic and ursodeoxycholic acid. J Lipid Res 1980;21: 35-43. 17. Bertolotti M, Carulli N, Menozzi D, Zironi F, Digrisolo A, Pinetti A, Baldini MG. In vivo evaluation of cholesterol 7a-hydroxylation in humans: effect of disease and drug treatment. J Lipid Res 1986;27:1278-1286. 18. Bertolotti M, Bertolotti S, Menozzi D, Iori E, Carulli N. Ageingand bile acid metabolism: studies on 7a-hydroxylation of cholesterol in humans. In: Paumgartner G, Stiehl A, Gerok W, eds. Trends in bile acid research. Dordrecht, The Netherlands: Kluwer Academic, 1989:75-78. 19. Grundy SM. HMG-CoA reductase inhibitors for treatment of hypercholesterolemia. N Engl J Med 1988;319:24-33. 20. Alberts AW. HMG-CoA inhibitors: the development. In: Stokes J 111, Mancini M, eds. Atherosclerosis reviews. Vol 18. New York: Raven Press, 1988:123-131. 21. Corey EJ, Gregoriou GA. Stereospecific synthesis of the 7-deuterio and 7-tritio cholesterols: the mechanism of enzyme-catalyzed hydroxylation at a saturated carbon atom. J Am Chem Soc 1959;81:3127-3133. 22. Bide AE, Henbest HB, Jones ERH, Wilkinson P. Studies in the sterol group. Part XLVIII. 7-Substituted cholesterol derivatives and their stereochemistry. Part I. 7-Halogeno-derivatives. J Chem SOC 1948:1788-1792. 23. Bergstrom S, Lindstedt S, Samuelsson B, Corey E, Gregoriou G. The stereochemistry of 7a-hydroxylation in the biosynthesis of cholic acid from cholesterol. J Am Chem Soc 1958:80:2337-2338. 24. Rosenfeld RS, Bradlow HL, Levin J , Zumoff B. Preparation of [24,25-3H]cholestero1.Oxidation in man as a measure of bile acid formation. J Lipid Res 1978;19:850-855. 25. Tubiana M, Dutreix J , Dutreix A, Jockey P. Bases physiques de la radiotherapie et de la radiologie. Paris: Masson, 1963:578-580. 26. Abell LL, Brodie BB, Levy BB, Kendall FL. A simplified method for estimation of total cholesterol in serum and demonstration of its specificity. J Biol Chem 1962;195:357-366. 27. Hutton JJ, Tappel AL, Udenfriend S. A rapid assay for collagen proline hydroxylase. Anal Biochem 1966;16:384-394. 28. Ali SS, Javitt NB. Quantitative estimation of bile salts in serum. Can J Biochem 1970;48:1054-1057. 29. Heuman DM, Hylemon PB, Vlahcevic ZR. Regulation of bile acid synthesis. 111. Correlation between biliary bile salt hydrophobicity index and the activities of enzymes regulating cholesterol and bile acid synthesis in the rat. J Lipid Res 1989;30:1161-1171. 30. Armstrong MJ, Carey MC. The hydrophobic-hydrophc balance of bile salts. Inverse correlation between reverse-phase high performance liquid chromatographic mobilities and micellar cholesterol-solubilizingcapacities. J Lipid Res 1982;23:70-80. 31. Wilson JD. The measurement of the exchangeable pools of cholesterol in the baboon. J Clin Invest 1970;49:655-665. 32. Redinger RN, Chow L, Grace DM. Cholesterol oxidation in primates by simultaneous sterol balance and breath analysis. Am J Physiol 1978;235:R55-R63. 33. LaRusso NF, Szczepanik PA, Hofmann AF. Effect of deoxycholic acid ingestion on bile acid metabolism and biliary lipid secretion in normal subjects. Gastroenterology 1977;72:132-140. 34. Tint GS, Salen G, Shefer S. Effect of ursodeoxycholic acid and chenodeoxycholic acid on cholesterol and bile acid metabolism. Gastroenterology 1986;91:1007-1018. 35. Pooler PA, Duane WC. Effects of bile acid administration on bile acid synthesis and its circadian rhythm in man. HEPATOLOGY 1988;8:1140-1146. 36. Coyne MJ, Bonorris GG, Goldstein LI, Schoenfield U.Effect of chenodeoxycholic acid and phenobarbital on the rate-limiting enzymes of hepatic cholesterol and bile acid synthesis in patients with gallstones. J Lab Clin Med 1976;87:281-291. 37. Reihnbr E, Bjorkhem I, Angelin B, Ewerth S, Einarsson K. Bile acid synthesis in humans: regulation of hepatic microsomal

Val. 14, No. 5, 1991

38.

39.

40. 41.

42. 43. 44.

45.

CHOLESTEROL 7a-HYDR(3XYLATION IN HUMANS

cholesterol 7a-hydroxvlase activity. Gastroenterology -. 1989;97: 1498-1505. Aneelin B. Hershon K. Brunzell JD. Bile acid metabolism in hereditary forms of hypertriglyceridemia: evidence for an increased synthesis rate in monogenic familial hypertriglyceridemia. Proc Natl Acad Sci USA 1987;84:5434-5438. Rudling MJ, Reihn6r E, Einarsson K, Ewerth S, Angelin B. Low density lipoprotein receptor-binding activity in human tissues: quantitative importance of hepatic receptors and evidence for regulation of their expression in vivo. Proc Natl Acad Sci USA 1990;87:3469-3473. Angelin B, Bjorkhem I, Einarsson K, Ewerth S. Cholestyramine treatment reduces postprandial but not fasting serum bile acid levels in humans. Gastroenterology 1982;83:1097-1101. Pandak WM,Li YC, Chiang JYL,Studer EJ, Gurley EC, Heuman DM, vlahmvic ZR, et al. Regulation of cholesterol 7a-hydroxylase mRNA and transcriptional activity of taurocholate and cholesterol in the chronic biLiary diverted rat. J Biol Chem 1991;266:3416342 1. Cabral DJ, Small DM, Lilly HS, Hamilton JA. Transbilayer movement of bile acids in model membranes. Biochemistry 1987;26:1801-1804. Fromm H. Gallstone dissolution and the cholesterol-bile acidlipoprotein axis. Propitious effects of ursodeoxycholic acid. Gastroenterology 1984;87:229-233. Bertolotti M, Iori R, Zironi F, Montanari M, Tripodi A, Carulli N. Bile acids and cholesterol metabolism: effect of deoxycholic acid on plasma lipoprotein and biliary cholesterol concentration. Ital J Gastroenterol 1988;20:240-245. Reihn6r E, Rudling M, Stahlberg D, Berglund L, Ewerth S, Bjorkhem I, Einarsson K, et al. Influence of pravastatin, a specific

-

46. 47.

48.

49.

50.

51. 52. 53.

837

inhibitor of HMG-CoA reductase, on hepatic metabolism of cholesterol. N Engl J Med 1990;323:224-228. Mitchell JC, Logan GM, Stone BG, Duane WC. Effects of lovastatin on biliary lipid secretion and bile acid metabolism in humans. J Lipid Res 1991;32:71-78. Ma PTS, Gil G, Siidhof TC, Bilheimer DW, Goldstein JL,Brown MS. Mevinolin, an inhibitor of cholesterol synthesis, induces mRNA for low density lipoprotein receptor in livers of hamsters and rabbits. Proc Natl Acad Sci USA 1986;83:8370-8374. Bilhartz LE, Spady DK, Dietachy JM. Inappropriate hepatic cholesterol synthesis expands the cellular pool of sterol available for recruitment by bile acids in the rat. J Clin Invest 1989;84: 1181-1187. Goldberg IJ,Holleran S, Ramakrishnan R, Adams M, Palmer RH, Dell RB, Goodman DS. Lack of effect of lovastatin therapy on the parameters of whole-body cholesterol metabolism. J Clin Invest 1990;86:801-808. Grundy SM, Bilheimer DW. Inhibition of 3-hydroxy-3methylglutaryl-CoA reductase by mevinolin in familial hypercholesterolemia heterozygotes: effects on cholesterol balance. Proc Natl Acad Sci USA 1984;81:2538-2542. Parker TS, McNamara DJ, Brown CD, Kolb R, Ahrens EH Jr, Alberta AW, Tobert J , et al. Plasma mevalonate as a measure of cholesterol synthesis in man. J Clin Invest 1984;74:795-804. Einarsson K, Reihn6r E, Bjorkhem I. On the saturation of the cholesterol 7a-hydroxylase in human liver microsomes. J Lipid R ~ 1989;30:1477-1481. s Pandak WM, Heuman DM, Hylemon PB, Vlahcevic ZR. Regulation of bile acid synthesis. IV.Interrelationship between cholesterol and bile acid biosynthesis pathways. J Lipid Res 1990;31: 79-90.