GASTROENTEROLOGY 1983;84:253-64

Transcription

1 GASTROENTEROLOGY 1983;84: Alteration of the Degree of Biliary Cholesterol Saturation in the Hamster and Rat by Manipulation of the Pools of Preformed and Newly Synthesized Cholesterol STEPHEN D. TURLEY, DAVID K. SPADY, and JOHN M. DIETSCHY Department of Internal Medicine, University of Texas Health Science Center at Dallas, Dallas, Texas Our studies compared the effects of changing the availability of newly synthesized and preformed cholesterol by various dietary manipulations on biliary cholesterol secretion in the hamster and rat. In hamsters fed a plain pelleted diet, only 2%-5% of biliary cholesterol was derived directly from newly synthesized stero1. Cholestyramine feeding, through a stimulation of hepatic sterol synthesis, increased this fraction fivefold but did not change total biliary cholesterol output. The relative cholesterol content increased significantly due to a reduction in bile acid and phospholipid output. In contrast, biliary cholesterol output was increased several-fold in hamsters fed a fat-free diet. These animals also manifested a pronounced increase in whole-body sterol synthesis, this being due principally to an increase in hepatic sterol synthesis. Although this resulted in the transport of much more newly synthesized cholesterol directly into bile, this did not account for the disproportionately high rate of biliary cholesterol output. Such excess sterol was derived predominantly from a preformed source. Unlike hamsters, rats fed the fat-free diet manifested a marked reduction in hepatic and whole-body sterol Received May 21, Accepted September 14, Address requests for reprints to: Dr. John M. Dietschy, University of Texas Health Science Center at Dallas, 5323 Harry Hines Boulevard, Dallas, Texas This work was supported by U.S. Public Health Service Research Grants HL and AM and by the Moss Heart Fund. The authors thank Joan Thorson, Debra Cohen, Karen Schaid, Edith Cole, Nancy Hammack, and Nancy Tomlinson for their excellent technical assistance and Dorothy Dunham for preparation of the manuscript by the American Gastroenterological Association /83/ $03.00 synthesis, bile acid pool size, and bile acid and cholesterol output in bile. These studies demonstrate that when hepatic cholesterol synthesis increases in response to a need for more sterol in the body, a greater proportion of biliary cholesterol is derived directly from newly synthesized sterol, but total biliary cholesterol output is unchanged. In contrast, when more cholesterol is synthesized than is needed to maintain cholesterol balance, biliary cholesterol output may increase. Such excess biliary sterol is derived predominantly from a preformed source rather than from the transport of newly synthesized sterol directly across the canalicular membrane. The finding more than a decade ago that the primary event in the pathogenesis of cholesterol gallstone disease is the secretion by the liver of saturated or lithogenic bile has led to an extensive study of the factors that regulate biliary cholesterol secretion (1-3). Although there have been significant advances in the study of the physicochemical factors that determine the solubility of cholesterol in bile, the mechanism(s) that ultimately controls the rate of biliary cholesterol secretion has not been elucidated (2-8). The ultimate regulatory step could occur either at the level of the transport process that facilitates the movement of cholesterol across the canalicular membrane, or alternatively, at the level of events within the hepatocyte related to the synthesis of cholesterol and the uptake of lipoproteins from the plasma. Recent studies in this laboratory have examined in detail the relationship between the rate of hepatic cholesterol synthesis and biliary cholesterol secretion using the rat as an experimental model (7,9). In

2 254 TURLEY ET AL. GASTROENTEROLOGY Vol. 84, No.2 those experiments, the contribution of newly synthesized cholesterol to biliary cholesterol was quantitated using 3H 2 0 under conditions where the rate of cholesterol synthesis in the liver had been made to vary over a wide range by subjecting the rats to various manipulations including the feeding of cholesterol and cholestyramine and fasting. It was found that while the proportion of biliary cholesterol derived from newly synthesized sterol was directly related to the rate of hepatic cholesterol synthesis, total biliary cholesterol output was independent of the rate of synthesis in the liver, and, instead was governed by the rate of bile acid and phospholipid output (7). The extent to which this finding can be applied to humans and other animals remains uncertain because there are several aspects of cholesterol metabolism in the rat that set it apart from the other species. For example, in the normal rat the rate of hepatic cholesterol synthesis is much higher than in other species such as the rabbit, guinea pig, hamster, and probably humans (8,10). Furthermore, the rat possesses a highly efficient mechanism for converting excess cholesterol into bile acid (11). More importantly, in the rat biliary cholesterol secretion is tightly coupled to bile acid and phospholipid output and many dietary and pharmacologic treatments that disturb this coupling process in humans and other animals are without effect in the rat (8.12). In a number of ways, sterol metabolism in the hamster differs from that in the rat and more closely approximates that seen in humans. The hamster, for example, carries more of its plasma cholesterol in the low density lipoprotein fraction (10) and manifests much lower rates of whole-body and hepatic cholesterol synthesis than the rat (10). Furthermore, this species has the propensity to produce lithogenic bile in response to a variety of dietary manipulations. One of these involves the feeding of a diet deficient in essential fatty acids. It is well documented that in hamsters (particularly males) fed a fat-free diet, there is a dramatic increase in both hepatic cholesterol synthesis and biliary cholesterol output (13-19). However, it is unclear whether the secretory defect simply involves the transport of disproportionately large quantities of newly synthesized sterol directly across the canalicular membrane, or whether it is a consequence of changes in hepatic lipoprotein secretion, bile acid production, or some other aspect of sterol metabolism that are somehow associated with the marked stimulation of cholesterol synthesis in the liver. In rats fed a fat-free diet, hepatic sterol synthesis is suppressed, but it is not known whether or not biliary lipid secretion is affected (20). The purpose of these studies was to investigate the effects of modifying the availability of newly synthesized and preformed cholesterol on biliary cholesterol output in the hamster and rat with particular emphasis on establishing whether or not the mass of newly synthesized sterol being transported directly into bile was, in any case, a determinant of the overall rate of biliary cholesterol secretion. Two series of studies were carried out. In the first of these, female hamsters were fed diets containing either cholestyramine or cholesterol, while the second group of experiments specifically compared the effects of a fat-free diet in male hamsters and rats. Materials and Methods Animal Preparations and Diets Female and male Golden Syrian hamsters were obtained from Charles River Lakeview, Newfield, N.J. The females were originally purchased in the weight range g, whereas the males were weanlings with a weight range g. Weanling male Sprague-Dawley-derived rats (weight range, g) were obtained from Charles River Breeding Laboratories, Inc., Wilmington, Mass. The hamsters were housed in individual cages and the rats in colony cages. The hamsters and rats were maintained in separate rooms, each with alternating 12-h periods of light (3:00 PM-3:00 AM) and darkness (3:00 AM-3:00 PM). After an adjustment period of 5 days, during which all animals had free access to a plain pelleted diet and water, the hamsters and rats were either started on their respective experimental diet or continued to receive a plain diet. In all studies, the control hamsters and rats were fed the same plain pelleted diet (Wayne Laboratory Animal Diets, Allied Mills, Inc., Chicago, Ill.). The digestible energy content of the diet was 3.23 kcal/g, and the cholesterol content, by direct analysis. was 0.26 mg/g (0.026% wt/wt). The other diets used in the studies with female hamsters consisted of a ground plain diet containing either cholestyramine (3% wt/wt) (Mead Johnson Research Center, Evansville, Ind.) or cholesterol (0.12% wt/wt) (Byron Chemical Company, Inc., Long Island City, N.Y.). The additional cholesterol was incorporated into the diet by first dissolving it in warm ethanol and then after mixing with the pelleted diet, allowing the ethanol to evaporate. The cholestyramine and cholesterol diets were fed for 2 and 4 wk, respectively. In preliminary studies, it was found that the hamsters fed the cholestyramine diet developed symptoms of vitamin K deficiency. Therefore, all cholestyramine-fed animals, as well as the cholesterol-fed and corresponding control groups, were given two subcutaneous injections of vitamin K during the last 10 days of their respective feeding periods. Aquamephyton (Merck Sharp & Dohme, West Point, Pa.) was diluted in 0.9% NaCl solution, and a volume containing 8,..,.,g of vitamin K was injected into each animal. The fat-free diet (No ) was obtained in pelleted form from ICN Nutritional Biochemicals, Cleveland, Ohio and stored at 4 C. It consisted of vitamin-free casein (21.10%), alphacel (16.45%), sucrose (57.85%), salt mix-

3 February 1983 BILIARY LIPID SECRETION IN HAMSTERS AND RATS 255 ture USP XIV (4.00%). choline chloride (0.60%), plus ICN vitamin diet fortification mixture. The digestible energy content of the diet was 3.18 kcallg. The experiments with the fat-free diet were carried out using the weanling male hamsters and rats and involved a feeding period of 5 wk. In all studies the diets were fed ad libitum, and body weights were measured at the commencement of the feeding of the experimental diets and again on the day of the experiment. The hamsters fed the cholestyramine and cholesterol diets gained weight at the same rate as those fed a plain diet. However, both the hamsters and rats fed the fat-free diet had significantly lower body weights at the end of the 5-wk feeding period than the respective control animals. In the studies in which rates of biliary lipid secretion and the contribution of newly synthesized cholesterol to biliary cholesterol were measured, the same protocol was used for hamsters and rats. The animals were anesthetized with diethyl ether, and then 150 mci of 3HzO (New England Nuclear, Boston, Mass.) contained in 1 ml isotonic saline was injected directly into a femoral vein. The animals were then placed in restraining cages and left for 6 h. They were then anesthetized again and, in the studies with hamsters, the cystic duct was ligated and a catheter placed in the common bile duct using PE-lO polyethylene tubing. In the studies with rats, the common bile duct was cannulated with PE-50 tubing. The animals were again restrained, and bile was collected for the following 4 h, either as a single sample or at hourly intervals, and the volume was recorded. The animals were then bled from the abdominal aorta and aliquots of plasma were taken for the measurement of plasma water specific activity (SA). In one study involving the measurement of the SA of cholesterol in bile, liver, and plasma, hamsters that had been fed the fat-free diet were administered 250 mci of 3H z O and then subjected to the above protocol. For each hourly collection period, equal aliquots of bile from all animals were pooled and these, as well as aliquots of liver and plasma from each animal, were extracted, and the SA of free and esterified cholesterol was measured. In all studies, the 3H z O was administered at about the mid-dark phase of the lighting cycle, and thereafter the animals were not given any food or drinking water. The administration of the 3H z O and all subsequent procedures were carried out under well-ventilated fume h~ods. Assay of Sterol Synthesis In Vivo The hamsters and rats were anesthetized with diethyl ether, a 50-mCi bolus of 3H z O was administered directly into a femoral vein, and they were then left for 1 h, during which time no food or water was given. The animals were then anesthetized again, a fixed volume of blood was taken from the abdominal aorta, and various tissues were removed. The SA of plasma water and the 3H_ digitonin precipitable sterols (DPS) content of the tissues were measured as described (21,22). Biliary Lipid Assays In the studies with hamsters, biliary cholesterol levels and the cholesterol content of gallstones were mea- sured with a gas chromatograph using stigmasterol (Applied Science Laboratories, Inc., State College, Pa.) as an internal standard (23). In the studies with rats, biliary cholesterol levels were assayed colorimetrically (24). The concentrations of total bile acids and phospholipid in bile were determined as described (25,26). Assay of Plasma, Liver, and Dietary Cholesterol Levels and of Bile Acid Pool Size The level of cholesterol in the plasma, liver, and diet, and bile acid pool size were measured as described (7,23,27,28). Because the liver was used for the assay of hepatic cholesterol levels, the bile acid contained in the liver was not included in the pool size measurements. However, the error incurred in these measurements was negligible because only a small fraction of the bile acid pool is present in the liver (29). Assay of Specific Activity of Cholesterol in Bile, Liver, and Plasma Aliquots of bile, liver, and plasma were extracted in chloroform/methanol (2:1 vollvol) on a steambath. The free and esterified cholesterol fractions were then separated on silicic acid/celite columns. After saponification, the sterols were precipitated as the digitonides and aliquots were taken for the determination of 3H content and of cholesterol content. Calculations The equations used to calculate the SA of plasma water, the rate of incorporation of 3H z O into DPS by the liver and extrahepatic tissues, and the mass of newly synthesized cholesterol secreted in bile have been described in detail (9). In the studies with rats, the 3H-DPS content of bile was converted to a mass of newly synthesized sterol using a value of 1.45 for the number of nanomoles of acetyl coenzyme A (Co A) incorporated into DPS for each nanomole of 3H z O. In the studies with hamsters, the calculations were made using a value of 1.59 for this ratio (10). Where appropriate, mean values ± 1 SEM for groups of data are given. Differences in mean values were tested for significance using the unpaired Student's t-test. Results In previous studies with rats it was established that 6 h after the animals had been administered 3H z O, the amount of labeled sterol secreted in bile relative to total cholesterol output reached a plateau and remained essentially constant over a further 4 h (9). This was taken as a measure of the fraction of biliary cholesterol that was derived directly from newly synthesized sterol. In female rats fed a plain diet, this fraction was equal to 15%-20%

4 256 TURLEY ET AL. GASTROENTEROLOGY Vol. 84, No. 2 Table 1. Effect of Feeding Diets Containing Cholestyramine or Cholesterol on Rates of Sterol Synthesis In Vivo in Various Tissues of Female Hamsters a Final Content of 3H-DPS per gram of tissue (nmollg-h) Synthesis of 3H-DPS body Small per whole animal Experimental diet weight (g) Liver intestine Blood Carcass (f-lmol/100 g h) Control (6) 136 ± ± ± ± ± ± 0.51 Cholestyramine (5) 129 ± ± 280 b 907 ± 73 b 299 ± 25 b 40 ± 4 b ± 1.41 b Cholesterol (5) 141 ± 2 10 ± Ib 184 ± ± ± Ib 1.58 ± 0.04b a Female hamsters that had been fed a diet containing either cholestyramine (3%) for 2 wk or cholesterol (0.12%) for 4 wk, as wejl as control animals that had been fed a plain pelleted diet only, were administered intravenously 50 mci of 3H 2 0 and killecll h later. The SA of plasma water and the content of 3H-DPS in the liver, small intestine, blood, and carcass were determined as described in Materials and Methods. These contents correspond to the nanomoles of 3H 2 0 incorporated into DPS per gram of tissue per hour, except for the whole animal where the units are micromoles per 100 grams body weight per hour. The values represent the mean ± SEM for the number of animals shown in parentheses. b Significantly different from the corresponding control value at the p<0.05 level. of total cholesterol output. A preliminary experiment was therefore carried out to determine whether the same protocol could be applied to accurately quantitate the contribution of newly synthesized sterol to biliary cholesterol in the hamster. Female hamsters that had been fed plain diets were administered 3HzO and after 6 h were subjected to total biliary diversion for 4 h. Although the data are not shown, the output of bile acid, phospholipid, and total cholesterol decreased markedly over the 4 h. The output of each of the lipids during the fourth hour was only ~30% of that during the first hour. The quantity of labeled sterol secreted in the bile also decreased by a similar extent. Therefore, the contribution of newly synthesized sterol to total cholesterol output remained essentially constant over this time. Although this fraction was only 2%- 3% of the total, the fact that it did not increase over the 4 h indicated that the pool of newly synthesized sterol being used for biliary secretion had fully equilibrated with the labeled sterol being generated within the hepatocyte during the 6 h after the administration of 3H z O. The first series of studies examined the effect of cholestyramine and cholesterol feeding on biliary lipid secretion, the contribution of I1ewly synthesized sterol to biliary cholesterol, and various other aspects of sterol metabolism in female hamsters. The amounts of 3HzO incorporated into DPS by the liver and extrahepatic tissues of these animals are given in Table 1. Because these measurements were carried out in vivo, the data actually represent the content of 3H-DPS per gram of tissue found 1 h after the administration of 3H z O. In the cholestyramine-fed group, hepatic sterol synthesis was increased 14- fold, and this resulted in more labeled sterol appearing in the blood (10,22). The intestine and the tissues present within the carcass also manifested a higher 3H-DPS content, but in the latter case this was probably due mainly to uptake of labeled sterol from the blood (22). Total body sterol synthesis was increased fivefold. In the cholesterol-fed group, sterol synthesis in the liver was markedly suppressed. There was also a significant decrease in the 3H-DPS content of the various extrahepatic tissues, but, again, this would have resulted partly from the uptake of less labeled sterol from the blood. Cholesterol feeding reduced total body sterol synthesis by 50%. The data given in Table 2 show that cholestyramine feeding lowered plasma cholesterol levels but did not affect the level of free and esterified choles- Table 2. Effect of Feeding Diets Containing Cholestyramine or Cholesterol on Plasma and Liver Cholesterol Levels and Bile Acid Pool Size in Female Hamsters a Final Liver plasma Ljver cholesterol content (mg/g) body weight cholesterol Bile acid pool size Experimental diet weight (g) (g) (mg/dl) free Esterified (f-lmollanimal) (f-lmollkg) Control (6) 138 ± ± ± ± ± ± ± 9 Cholestyramine (6) 136 ± ± ± ± :) ± ± 0.7 b 76 ± 5 b Cholesterol (6) 136 ± ± 0.5 H14 ± 6 b 2.21 ± 0.09 b 5.95 ± 2.55 b 16.7 ± ± 11 a Female hamsters that had been fed the same diets as described in the legend to Table 1 were anesthetized, and blood was collected from the abdominal aorta i'nto a syringe containing ethylenediaminetetraacetate. The liver was removed, and aliquots were placed in chloroform/methanol (2:1). The gallbladder and the "mtire small intestine, along with its contents, were added to a beaker containing ethanol. The level of cholesterol in the plasma and liver, and the bile acid pool size were measured as described in Materials and Methods. The values represent the mean ± SEM for the number of animals shown in parentheses. b Significantly different from the corresponding control value at the p<0.05 level.

5 February 1983 BILIARY LIPID SECRETION IN HAMSTERS AND RATS 257 Table 3. Effect of Feeding Diets Containing Cholestyramine or Cholesterol on Biliary Lipid Concentration and Output in Female Hamsters a Bile se- Biliary lipid concentration Biliary lipid output (J..tmollml) Final cretion (J..tmollkg h) Molar percentage in bile body rate (mll Bile Phospho- Bile Phospho- Bile Phospho- Experimental diet weight (g) kg h) acid lipid Cholesterol acid lipid Cholesterol acid lipid Cholesterol Control (12) ±3 ±0.2 ±0.6 ±0.13 ±0.017 ±2.3 ±0.6 ±0.05 ±1.0 ±1.0 ±0.1 Cholestyramine (10) ±2 ±O.lb ±0.5 ±0.07 b ±0.022 ±0.7 b ±0.2b ±0.05 ±0.8 ±0.9 b ±0.2b Cholesterol (5) ±3 ±0.3 b ±0.5 ±0.11 b ±0.013 ±2.8 ±0.8 ±0.06 ±1.4 ±1.5 ±0.1 a Female hamsters that had been fed the same diets as described in the legend to Table 1 were administered intravenously 150 mci of 3H 2 0, placed in a restraining cage, and left for 6 h. They were then anesthetized again, the cystic duct was ligated, and a catheter was placed in the common bile duct. The hamsters were again restrained and bile was collected for the following 4 h. The content of total and newly synthesized cholesterol and of bile acid and phospholipid in bile was measured as described in Materials and Methods. The data on newly synthesized cholesterol are shown in Figure 1. The values represent the mean ± SEM for the number of animals shown in parentheses. b Significantly different from the corresponding control value at the p<0.05 level. terol in the liver. The cholesterol-rich diet, which contained fo~r to five times more cholesterol than the plain diet, produced modest hypercholesterolemia and a marked accumulation of cholesteryl ester in the liver. The results of the studies on biliary lipid secretion in the cholestyramine- and cholesterol-fed hamsters are shown in Table 3. In the group fed cholestyramine, the rate of bile secretion and the output of bile acid and phospholipid were significantly lower than in the control group. The lower rate of bile acid secretion was consistent with the significant reduction of bile acid pool size found in a separate group of animals (Table 2). Although bile acid and phospholipid output decreased, total cholesterol output was unchanged. Thus, there was a slight, but significant increase in the relative cholesterol content of the bile. In the cholesterol-fed group, there was a 30% increase in liver size and this contributed to a significantly higher rate of bile secretion. However, overall there was no change in either the output or relative content of any of the three lipids. There was no evidence of gallstone formation in any of the groups. In the studies described in Table 3, the fraction of biliary cholesterol derived from newly synthesized sterol was also measured. Figure 1 shows the average hourly output of both total and newly synthesized cholesterol in bile during the 4 h of diversion. In the control group, newly synthesized sterol represented only 5% of total output. This fraction was increased to 21% in the cholestyramine-fed animals but was reduced to only 0.2% in the hamsters fed cholesterol. Despite the variable contribution of newly synthesized cholesterol, total cholesterol output was similar in the three groups. The second major series of studies involved a comparison of the effect of a fat-free diet on biliary lipid secretion and other aspects of sterol metabolism in male hamsters and rats. The data given in Table 4 show the effect of the fat-free diet on the content of newly synthesized sterols in the major tissues of both species. In the hamsters fed the fatfree diet, sterol synthesis in the liver was increased >15-fold, with little change in most of the extrahepatic tissues except the small intestine in which the content was increased about twofold. The increase 0.1 ~ NEWLY SYNTHESIZED CHOLESTEROL CHOLESTEROL CONTROL CHOLESTYRAMINE DIET DIET DIET Figure 1. Effect of feeding cholestyramine and cholesterol on biliary cholesterol output and the contribution of newly synthesized cholesterol to biliary cholesterol in female hamsters. These data are derived from the studies described in Table 3. The values represent the average hourly output of both total and newly synthesized cholesterol in bile over 4 h.

6 258 TURLEY ET AL. GASTROENTEROLOGY Vol. 84, No.2 Table 4. f;ffect of Feeding a Fat-Free Diet on Rates of Sterol Synthesis In Vivo in Various Tissues of Male Hamsters and Rats a Final Content of 3H-DPS per gram of tissue (nmollg'h) Synthesis of 3H-DPS per Experimental body Small whole ani mill diet weight (g) Liver intestine Adrenal Testis Blood Carcass (JLmoll100 g h) Hamster Control (4) 111 ± 7 74 ± ± ± ± 2 22 ± 3 41 ± ± 0.88 Fat -free (7) 82 ± 3 b 1144 ± 122b 452 ± 44 b 1593 ± ± ± 13 b 55 ± ± 0.81 b Rat Control (5) 320 ± ± ± ± ± 2 90 ± 6 17 ± ± 0.59 Fat-free (5) 279 ± 5 b 439 ± 43 b 365 ± ± ± 2b 31 ± 6 b 30 ± Ib 5.64 ± 0.16 b a Male hamsters and rats that had been fed a fat-free diet for 5 wk, as well as control animals that had been fed a plain pelleted diet only, were administered intravenously 50 mci of 3H 2 0 and killed 1 h later. The SA of plasma water and the content of 3H-DPS in the liver and various extrahepatic tissues were determined as described in Materials and Methods. These contents correspond to the nanomoles of 3H 20 incorporated into DPS per gram of tissue per hour, except for the whole animal where the units are micromoles per 100 grams body weight per hour. The values represent the mean ± SEM for the number of animals shown in parentheses. b Significantly different from the corresponding control value at the p<0.05 level. in the 3H-DPS content of the blood reflected the enhanced rate of sterol synthesis in the liver. Overall, total body sterol synthesis was elevated 2.5-fold, this being due principally to the increase in hepatic sterol synthesis. In contrast to the hamsters, the rats fed the fat-free diet showed marked inhibition of sterol synthesis in the liver and a corresponding decrease in the 3H-DPS content of the blood. Although the 3H-DPS content of the small intestine, adrenal glands, and testes was also slightly decreased, one or more tissues contained within the residual carcass must have had an enhanced rate of sterol synthesis because the 3H-DPS content of the carcass in the rats fed the fat-free diet was 76% higher than that in the control animals. Despite this, total body sterol synthesis in the rats fed the fat-free diet was significantly less than in those fed a plain diet. The effect of the fat-free diet on cholesterol levels in the liver is shown in Table 5. In contrast to the markedly different effect on hepatic sterol synthesis in the two species, there was a four- to fivefold increase in hepatic cholesteryl ester levels and no significant change in free cholesterol levels in both the hamsters and rats fed the fat-free diet. In these same groups of animals, the distribution of cholesterol in the various plasma lipoprotein fractions was measured. Although in neither species was there a significant change in the total plasma cholesterol levels, there were clearly different effects on the distribution of cholesterol carried in the various lipoprotein fractions. Thus, compared with the hamsters fed a plain diet, those fed the fat-free diet had about twice as much cholesterol in the fractions of p < g/ml (21.5 mg/dl compared with 10.2 mg/dl) and p = g/ml (5.0 mg/dl compared with 2.6 mg/dl). In contrast, the rats fed the fat-free diet generally had less cholesterol in the various fractions of p < but had substantially more cholesterol in the high density lipoprotein fraction (p = Table 5. Effect of Feeding a Fat-Free Diet on Liver Cholesterol Levels in Male Hamsters and Rats a Final Experimental body Liver Liver cholesterol content (mg/g) diet weight (g) weight (g) Free Esterified Hamster Control (8) 99 ± ± ± ± 0.13 Fat-free (9) 77 ± 3 b 3.4 ± 0.2b 2.23 ± ± 0.73 Q Rat COlltrol (5) 288 ± ± ± ± 0.03 Fat-free (5) 253 ± 4 b 11.1 ± ± ± 0.13 b a Male hamsters and rats that had been fed a fat-free diet for 5 wk, as well as control animals that had been fed a plain pelleted diet only, were anesthetized and bled from the abdominal aorta. The liver was removed, weighed, and aliquots of tissue were placed in chloroform/methanol (2:1). The 1t;lVel of free and esterified cholesterol was measured as described in Materials and Methods. The values represent the mean ± SEM for the number of animals shown in parentheses. b Significantly different from the respective control value at the p<0.05 level.

7 February 1983 BILIARY LIPID SECRETION IN HAMSTERS AND RATS 259 Table 6. Effect of Feeding a Fat-Free Diet on Biliary Lipid Concentration and Output in Male Hamsters and Rats a Bile se- Biliary lipid concentration Biliary lipid output Final (p,mollml) (p,mollkg h) cretion Molar percentage in bile Experimental body rate (mll Bile Phospho- Bile Phospho- Bile Phosphodiet weight (g) kg h) acid lipid Cholesterol acid lipid Cholesterol acid lipid Cholesterol Hamster Control (5) ±:4 ±:0.2 ±:0.5 ±:0.13 ±:0.02 ±: 1.1 ±:0.5 ±:0.06 ±:1.5 ±:1.3 ±:0.3 Fat-free (8) ±:3 ±:0.1 b ±:0.6 b ±:0.16 b ±:0.14b ±:2.6 ±:0.6 b ±:0.42 b ±: 1.2 b ±:0.6 b ±:0.8 b Rat Control (5) ±:6 ±:0.1 ±:1.9 ±:0.2 ±:0.02 ±:9 ±:0.8 ±:0.10 ±:1.0 ±:0.9 ±:0.1 Fat-free (4) ±:6 ±O.lb ±1.5 b ±:0.6 ±0.04 b ±4b ±1.7 ±0.10b ±1.2 b ±:l.l b ±:0.1 a Male hamsters and rats that had been fed a fat-free diet for 5 wk, as well as control animals that had been fed a plain pelleted diet only, were administered intravenously 150 mci of 3H 2 0 and subjected to exactly the same protocol that is described in the legend to Table 3. The values represent the mean ±: SEM for the number of animals shown in parentheses. b Significantly different from the corresponding control value at the p<0.05 level g/ml) compared with the control group. The next group of experiments examined the effect of the fat-free diet on biliary lipid secretion, bile acid pool size, and the contribution of newly synthesized cholesterol to biliary cholesterol. The results of the studies on biliary lipid secretion in the hamster and rat are shown in Table 6. In the hamsters fed the fatfree diet, the concentration of all three biliary lipids was increased significantly, particularly that of cholesterol, which was elevated ninefold. Although the rate of bile secretion was significantly lower, the overall secretory rate of cholesterol was 7.4-fold higher in the fat-free animals. Cholesterol was the Table 7. Effect of Feeding a Fat-Free Diet on Bile Acid Pool Size in Male Hamsters and Rats a Final Experimental body Bile acid pool size diet weight (g) (p,mollanimal) (p,mollkg) Hamster Control (11) 108 ± ±: ±: 7 Fat-free (20) 85 ±: 2b 13.4 ±: ±: 16 Rat Control (4) 334 ±: ±: ±: 29 Fat-free (4) 228 ±: 3 b 57 ±: 3 b 203 ±: llb a Male hamsters and rats that had been fed a fat-free diet for 5 wk, as well as control animals that had been fed a plain pelleted diet only, were anesthetized and bled from the abdominal aorta. The entire small intestine and its contents were removed and placed in a beaker containing ethanol. In the studies with hamsters, the gallbladder and its contents were combined with the intestine. Bile acid pool size was measured as described in Materials and Methods. The values represent the mean ±: SEM for the number of animals shown in parentheses. b Significantly different from the respective control value at the p<0.05 level. only sterol detected in the bile. The output of bile acid and phospholipid was also increased but to a much lesser extent than cholesterol. Thus, the molar percentage of cholesterol was fivefold higher in the group fed the fat-free diet. All animals in this particular group showed some degree of gallstone formation, and by direct analysis, cholesterol was found to represent at least 75% of the mass of these stones. None of the control animals developed gallstones. In contrast to the hamsters, the rats fed the fat-free diet showed significantly lower biliary concentrations and outputs of both cholesterol and bile acid than the control animals. The concentration and output of phospholipid in bile were unchanged. Because both bile acid and cholesterol output decreased markedly, overall there was no significant change in the molar percentage of cholesterol. The proportion of phospholipid in bile increased significantly while there was a corresponding decrease in the relative content of bile acid. The effect of the fat-free diet on the rates of biliary bile acid output was consistent with the changes found in bile acid pool size in a separate group of animals (Table 7). Thus, in the hamsters fed the fatfree diet, both bile acid output and pool size were marginally elevated, whereas in the rats fed the fatfree diet, the secretion rate and pool size of bile acid were both - 55% less than in the control animals. The studies described in Table 6 involved not only the measurement of the absolute rates of biliary lipid secretion but also the quantitation of the fraction of biliary cholesterol that was derived directly from newly synthesized cholesterol. The results of this experiment with the hamsters are shown in panel A of Figure 2. In the hamsters fed a plain diet, very little labeled sterol was detected in the bile even

8 260 TURLEY ET AL. GASTROENTEROLOGY Vol. 84, No.2 I ~ ~ 2.5 ~ o c5 _ 2.0 0::-<= W' I-~ ~::::: E :r:::t, u >- 0:: <I: ~ 0.5 Figure A HAMSTER ~ NEWLY SYNTHESIZED ~CHOLESTEROL I B RAT ~ NEWLY SYNTHESIZED CHO LESTEROL,-1-,-1- ~.,... CONTROL FAT-FREE DIET DIET 2. Effect of feeding a fat-free diet on biliary cholesterol output and the contribution of newly synthesized cholesterol to biliary cholesterol in male hamsters and rats. These data are derived from the studies described in Table 6. The values represent the average hourly output of both total and newly synthesized cholesterol in bile over 4 h. bile remained essentially constant over this time. The SA of free cholesterol in the liver and plasma of the same animals at the end of the fourth hour of diversion was lower than the SA of biliary free cholesterol during the fourth hour. Although the data are not shown, the SA of the cholesteryl ester fraction in both the liver and plasma was in turn only about half that of the respective free cholesterol fraction. There was no detectable mass of cholesteryl ester in the bile. Using the SA data for biliary cholesterol, it was calculated that only 5%-6% of all cholesterol secreted in the bile over the 4-h period was newly synthesized (Figure 3, panel B). These various findings thus demonstrated that the pool of newly synthesized sterol being used for biliary secretion had become fully equilibrated with labeled sterol by the time biliary diversion had commenced, and, therefore, that much of the excess cholesterol transported into bile under these conditions was not derived directly from the additional cholesterol being synthesized within the hepatocyte. though the SA of plasma water had been raised to very high levels. Overall, only 1.5% of biliary cholesterol was derived from newly synthesized sterol in the control group. In the hamsters given the fatfree diet, the mass of newly synthesized cholesterol secreted in the bile was increased 29-fold, but because overall biliary cholesterol output had increased more than sevenfold, newly synthesized sterol accounted for only 4.8% of total cholesterol output. The data for the corresponding experiment in rats are shown in panel B of Figure 2. In the control rats, 9.4% of biliary cholesterol was newly synthesized. However, in the rats fed the fat-free diet, not only was there a significant decrease in total biliary cholesterol secretion, but also the contribution of newly synthesized sterol was reduced to only 2.1% of total output. Although the hamsters fed the fat-free diet showed a dramatic increase in the mass of newly synthesized sterol transported into bile, the finding that its contribution to total cholesterol output remained very small raised the possibility that, under these particular dietary conditions, the pool of newly synthesized sterol being used for biliary secretion had been expanded greatly and thus had not fully equilibrated with labeled sterol being generated within the hepatocyte during the 6-h interval after the administration of the 3H z O. A further study was therefore carried out using hamsters fed the fat-free diet to measure the SA of biliary free cholesterol during each hour of the 4-h diversion period. As shown in panel A of Figure 3, the SA of free cholesterol in the A INJECT ['H]WATER ~:L' 10 5 / /' , COMMENCE BILIARY DIVERSION /Bile /.-._7_._. Liver-...~ Plasma..,/' / 1~0~--71--~2~--3~--~4~ TIME (h) Figure 3. Specific activity of free cholesterol in bile and the proportion of biliary cholesterol newly synthesized in male hamsters subjected to total biliary diversion for 4 h. Male hamsters that had been fed a fat-free diet for 5 wk were administered intravenously 250 mci of 3H 2 0 and subjected to the same protocol as described in the legend to Table 3. The animals were then killed and samples of plasma and liver obtained. For each hourly collection period, equal aliquots of bile from all animals were pooled and these, as well as aliquots of plasma and liver from each animal were added to chloroform methanol (2: 1). The SA of plasma water, the SA of free and esterified cholesterol in the bile, liver, and plasma, and the content of bile acid and phospholipid in bile were measured as described in Materials and Methods. The data represent determinations in a total of 10 animals.

9 February 1983 BILIARY LIPID SECRETION IN HAMSTERS AND RATS 261 Discussion Despite extensive use of the hamster as a model for studies on the pathogenesis and treatment of cholesterol gallstone disease (30-32)' there are few published data on the normal characteristics of biliary lipid secretion and related aspects of sterol metabolism in this species. In this regard, our studies raise three points that warrant emphasis. The first of these concerns the dissociation of biliary cholesterol content from the rate of hepatic cholesterol synthesis. In other studies using age-matched hamsters, it was shown that in females the rate of hepatic sterol synthesis was several-fold higher than in males (33). Although in our studies the animals were not of identical age, the females still had a rate of hepatic synthesis that was about twice that in the males (Tables 1 and 4). However, the absolute and relative concentrations of cholesterol in the bile did not show a corresponding sex difference (Tables 3 and 6). This finding is consistent with that of other investigators (30). A dissociation of these two variables is also evident from a comparison of the data for male rats and hamsters. Thus, in the control male hamsters hepatic sterol synthesis was only a fraction of that in male rats fed the same plain pelleted diet, but the relative cholesterol content of the bile was essentially the same in both groups (Tables 4 and 6). A direct relationship between the level of cholesterol in bile and the rate at which cholesterol is synthesized in the liver might be expected only if a major portion of biliary cholesterol were derived directly from newly synthesized sterol. The present studies using 3H z O demonstrate that in the hamster newly synthesized sterol makes an even smaller contribution to biliary cholesterol than it does in the rat (9). In both the female and male hamsters fed the plain pelleted diet, only trace amounts of labeled sterol were present in the bile, even up to 10 h after the administration of a bolus of 3H z O that was sufficient to raise the SA of plasma water to very high levels. Overall, in both the females and males, only ~2%-5% of biliary cholesterol was newly synthesized. The experiment with cholestyramine-fed hamsters demonstrated that in this species, as in the case of the rat, the amount of newly synthesized cholesterol secreted directly into bile can be increased severalfold without changing total cholesterol output. It must be emphasized, however, that in this case the enhanced rate of hepatic synthesis is a compensatory response to the accelerated loss of sterol from the body. This does not necessarily mean that biliary cholesterol output would remain constant under conditions where cholesterol production by the liver, or by some other extrahepatic compartment, is elevated to a rate that exceeds that which is required to maintain normal cholesterol balance. The slight increase in biliary cholesterol saturation that occurred in the cholestyramine-fed hamsters can probably be attributed to a reduction in bile acid pool size. A similar, but more dramatic effect has been demonstrated in baboons treated with cholestyramine (34). Although biliary cholesterol secretion remained unchanged in the hamsters fed the diet containing additional cholesterol (0.12%) for 4 wk, it is possible that it would have increased had the diet been fed for a much longer period. It was recently reported that in female hamsters fed a standard diet containing cholesterol at a level of 1 % for 30 days, there was a slight increase in biliary cholesterol output, but cholesterol gallstones were not formed (35). Studies from another laboratory showed that the degree of cholesterol saturation of gallbladder bile increased and cholesterol gallstones were produced in female hamsters fed a diet with a cholesterol level of 0.24% for 12 wk (31). Thus, these studies have shown that biliary cholesterol output is relatively constant even though there are marked differences between the rates of cholesterol synthesis in the livers of male and female hamsters and in the livers of hamsters and rats. Furthermore, cholesterol and cholestyramine feeding can induce marked changes in rates of hepatic sterol synthesis in both species without changing the rates of biliary cholesterol secretion. Hence, biliary cholesterol secretion rates do not correlate with differences in rates of hepatic sterol synthesis regardless of whether these differences are due to the use of experimental animals of a different sex or a different species, or to compensatory changes in hepatic sterol synthesis induced by cholesterol or cholestyramine feeding. The second point to be made is that in the hamster the level of cholesterol in the bile also is unrelated to hepatic cholesteryl ester content. Thus, in these and other studies it was found that male hamsters fed a plain diet manifested cholesteryl ester levels in the liver that were several-fold higher than those in females fed the same diet, but both groups secreted bile with a comparable cholesterol content (Tables 2, 3, 5, and 6, reference 33). Similarly, in the female hamsters fed the cholesterol diet, biliary cholesterol levels were unchanged even though hepatic cholesteryl ester content had increased dramatically. Although the male hamsters fed the fat-free diet showed a fivefold increase in the level of cholesteryl ester in the liver and a dramatic increase in biliary cholesterol levels, it is unlikely that expansion of the cholesteryl ester pool had any direct role in the formation of saturated bile because in the male rats

10 262 TURLEY ET AL. GASTROENTEROLOGY Vol. 84, No.2 fed the fat-free diet, hepatic cholesteryl ester levels also increased fivefold, but biliary cholesterol secretion actually declined. The third point to be made concerns the size of the bile acid pool in the hamster. Relative to body weight, bile acid pool size in the normal hamster was only ~30% of that in the rat and, over a 4-h period of diversion, became depleted to a much greater extent than normally occurs in the rat (Table 7. reference 9). Thus, since the rates of biliary lipid secretion given in the present studies generally represent the average over 4 h of diversion, the absolute rates found for the hamster and rat cannot be directly compared. Our studies with hamsters fed the diet deficient in essential fatty acids demonstrated, in agreement with the findings of several other laboratories, that the characteristic secretory change in these animals involves a dramatic increase in cholesterol secretion with either comparatively little or no increase in bile acid and phospholipid secretion. In theory, this excess cholesterol could come from preformed sterol within the hepatocyte or from the pool of newly synthesized cholesterol. Although the mass of newly synthesized sterol secreted directly into bile is increased in hamsters with essential fatty acid deficiency, this increase cannot account for the disproportionately large quantity of biliary cholesterol secreted in these animals. This conclusion is based on the finding that >90% of the sterol appearing in bile was unlabeled, and therefore not newly synthesized (Figure 2). Although several studies by other investigators involving the measurement of hepatic hydroxymethylglutaryl Co A reductase activity or the incorporation of [ 14 CJacetate into sterol by liver slices had clearly demonstrated a marked increase in sterol production by the liver in hamsters with essential fatty acid deficiency, it remained unclear as to what extent extrahepatic sterol synthesis was affected and whether or not there was actually any change in total body sterol synthesis under these conditions (14,17,18). The present experiments using 3HzO to measure sterol synthesis in vivo showed that in hamsters fed the fat-free diet there was a dramatic increase in hepatic cholesterol synthesis but relatively little change in the extrahepatic tissue compartments except the small bowel where an increase also occurred. Overall, total body sterol synthesis was elevated 2.5-fold, this being due principally to the increase in hepatic sterol synthesis. The signal that initiates the enhancement of cholesterol synthesis in the liver is unknown. It is not simply the lack of cholesterol in the diet as might be expected because the rate of synthesis can be returned toward normal levels merely by the addition of one of the essential fatty acids to the fat-free diet, or alternative- ly by substituting the sucrose component of the diet with a more complex carbohydrate such as starch (14,18). In hamsters fed cholestyramine, the compensatory increase in hepatic cholesterol synthesis, despite its magnitude, is apparently insufficient to balance the loss of sterol from the body because the level of cholesterol in plasma and bile acid pool size are both reduced. In the hamsters with essential fatty acid deficiency, hepatic sterol synthesis is increased by about the same magnitude as in the cholestyraminefed animals, but because the stimulation of hepatic synthesis produced by the fat-free diet is apparently unrelated to the lack of cholesterol in the diet or to an increased rate of bile acid degradation, the additional cholesterol that is synthesized in liver and other tissues probably exceeds the amount needed to actually compensate for lower dietary intake. This interpretation is consistent with the report that when hamsters are fed a fat-free diet containing cholestyramine there is a significant reduction in the incidence of cholesterol gallstone formation (36). Presumably, in this situation the enhanced degradation of cholesterol induced by the cholestyramine counteracts, to a major extent, the putative overproduction of cholesterol by these animals. If this interpretation is correct, then it implies that in the hamster fed the fatfree diet, the transport of excessive quantities of cholesterol into bile is related in some way, however indirectly, to the high rate of cholesterol synthesis in the liver of these animals. The present studies with rats fed the fat-free diet support this interpretation because in almost every respect, the response of the rat was the opposite of that found in the hamster. It is known that in rats with essential fatty acid deficiency, triglyceride and cholesteryl ester accumulate in the liver, apparently because of a diminished output of very low density lipoprotein (20,37). Furthermore, hepatic cholesterol synthesis is suppressed to less than normal levels even though cholesterol intake is significantly lower on this diet than on the plain pelleted diet (Table 4). This reduction in hepatic sterol synthesis, together with an insufficient compensatory increase in extrahepatic sterol synthesis, and the absence of cholesterol in the diet may critically limit the availability of cholesterol for bile acid synthesis. This, in turn, may account for the observed reduction in bile acid pool size and for the lower rates of bile acid and cholesterol output in bile in these animals (Tables 6 and 7). In summary, these studies appear to delineate two fundamentally different situations with respect to the relationship between cholesterol synthesis within the body and the rate at which cholesterol is secreted into bile. On the one hand, when a change