GASTROENTEROLOGY 1998;114:1024–1034
Reduced Plasma Cholesterol and Increased Fecal Sterol Loss in Multidrug Resistance Gene 2 P-Glycoprotein–Deficient Mice PETER J. VOSHOL,* RICK HAVINGA,* HENK WOLTERS,* ROEL OTTENHOFF,‡ HANS M. G. PRINCEN,§ RONALD P. J. OUDE ELFERINK,‡ ALBERT K. GROEN,‡ and FOLKERT KUIPERS* *Groningen Institute for Drug Studies, Laboratory of Nutrition and Metabolism, University Hospital Groningen, Groningen; ‡Department of Gastrointestinal and Liver Diseases, Academic Medical Center, Amsterdam; and §Gaubius Laboratory TNO Prevention and Health, Leiden, The Netherlands
Background & Aims: mdr2 P-glycoprotein (Pgp) deficiency in mice leads to the absence of biliary phospholipids and cholesterol in the presence of normal bile salt secretion. The aim of this study was to evaluate the importance of the biliary pathway in cholesterol homeostasis by determining the effects of mdr2 Pgp deficiency on hepatic and plasma lipid levels and cholesterol kinetics in chow-fed mice. Methods: Hepatic lipid content, enzyme activities, plasma lipoprotein levels, and fecal sterol excretion were measured in wild-type (1/1) and mdr2 Pgp–deficient (2/2) mice. Cholesterol kinetics were determined using radiotracer techniques. Results: No differences in hepatic lipid content were observed between (2/2) and (1/1) mice. Plasma high-density lipoprotein cholesterol and apolipoprotein A-I levels were strongly reduced in (2/2) mice compared with controls, whereas the apolipoprotein B contents of very-low-density lipoprotein and low-density lipoprotein were increased. Hepatic activity of 3-hydroxy-3methylglutaryl–coenzyme A reductase was threefold greater in (2/2) mice than in controls; however, compartmental analysis of plasma cholesterol decay showed no differences in cholesterol synthesis between (2/2) and (1/1) mice. A dual isotope approach for estimating cholesterol absorption yielded ,50% lower values in (2/2) mice than in controls. Surprisingly, (2/2) mice showed a fourfold increase in fecal neutral sterol secretion. Conclusions: This study unequivocally establishes the important direct role of biliary lipids in the regulation of plasma lipid levels in mice.
he hepatobiliary axis appears to play a key role in the maintenance of cholesterol homeostasis.1,2 Concerning the regulatory aspects of this axis, interest has mainly been focused on the bile salts. Bile salts represent the major catabolic products of cholesterol with a fecal disposal equivalent to 400–600 mg cholesterol per day in humans.1,2 In addition, bile salts fulfill important physiological functions in the intestinal absorption of lipids3,4 and, after their reabsorption from the intestine, in the regulation of hepatic cholesterol and bile salt synthesis,1,5
T
low-density lipoprotein (LDL) receptor activity,6 and very-low-density lipoprotein (VLDL) production.7,8 Finally, bile salts control secretion of cholesterol and phospholipids into the bile.9 Secretion of cholesterol into bile amounts up to 1200 mg/day in humans.1 A considerable fraction, approximately 60%, is reabsorbed from the intestine.10 In view of the fact that the amount of biliary cholesterol entering the intestine exceeds that of dietary cholesterol by a factor of at least two, intestinal reabsorption of biliary cholesterol may represent an important site of control in cholesterol homeostasis. Biliary cholesterol secretion is coupled to a fourfold greater secretion of bile-specific phospholipids, consisting mainly (95%) of phosphatidylcholine with 16:0–18:1, 16:0–18:2, or 16:0– 20:4 acyl chain configuration.11 Phospholipids are essential for efficient solubilization of biliary cholesterol12,13 and for protection of cells lining the biliary tree against the detergent action of bile salts.14–16 Virtually all biliary phospholipids are reabsorbed by the intestine.17 It has been claimed that these phospholipids are needed for adequate formation and secretion of the apolipoprotein (apo) B48–containing chylomicrons in the enterocytes4,18 and, therefore, for lipid absorption. Thus, both biliary cholesterol and phospholipids are conserved in the enterohepatic circulation. The impact of the circulation of biliary lipids per se on cholesterol homeostasis, however, has not been defined, mainly because it has not been possible to vary output rates of the individual biliary lipids independently from that of bile salts under in vivo conditions. Recently generated mice in which the mdr2 gene has been disrupted by homologous recombination (mdr2 [2/2]) provide a new, powerful tool to address this Abbreviations used in this paper: ACAT, acyl–coenzyme A: cholesterol acyltransferase; apo, apolipoprotein; FPLC, fast protein liquid chromatography; HDL, high-density lipoprotein; HMG-CoA, 3-hydroxy-3-methylglutaryl–coenzyme A; LDL, low-density lipoprotein; Pgp, P-glycoprotein; VLDL, very-low-density lipoprotein. r 1998 by the American Gastroenterological Association 0016-5085/98/$3.00
May 1998
issue. The (2/2) mice show a virtual absence of phospholipids in their bile and a strongly impaired cholesterol secretion,12–14 whereas bile salt output is not affected. Based on these data, it was proposed that mdr2 Pglycoprotein (Pgp) functions as a flippase that is responsible for translocation of phospholipids from the inner to the outer leaflet of the bile canalicular membrane before their secretion into bile.12,14 This view is supported by results from a number of recently published studies.19–22 In the present study, we evaluated the effects of mdr2 Pgp deficiency on plasma lipid levels and on cholesterol kinetics in mice fed a standard, low-cholesterol laboratory chow. The results of this study unequivocally establish a key role of biliary lipid secretion in the regulation of plasma lipoprotein levels. In addition, absence of biliary lipids surprisingly leads to enhanced fecal sterol loss that is not reflected in plasma cholesterol kinetics, indicating disequilibrium between body pools of cholesterol in mdr2 Pgp–deficient mice.
Materials and Methods Materials Tween 80, Triton WR-1339, glucose-6-phosphate, fatty acid–free bovine serum albumin, oleoyl coenzyme A, 3-hydroxy-3-methylglutaryl–coenzyme A (HMG-CoA), and mevalonic acid lactone were purchased from Sigma Chemical Co. (St. Louis, MO). Glucose-6-phosphate dehydrogenase, dithiothreitol, nicotinamide adenine dinucleotide phosphate, adenosine triphosphate, triglyceride, and cholesterol (ester) kits were obtained from Boehringer Mannheim GmbH (Mannheim, Germany). [14C]HMG-CoA (55 mCi/mmol), [1, 2[n]3H]cholesteryl oleate (45.1 Ci/mmol), and [7(n)-3H]cholesterol (3.5 Ci/mmol) were obtained from the Radiochemical Center, Ltd. (Amersham, Buckinghamshire, England). [5-3H]Mevalonic acid lactone, [4-14C]cholesterol (5.8 Ci/mmol), and [1-14C]oleoyl coenzyme A (59.35 mCi/mmol) were obtained from New England Nuclear Corp. (Boston, MA). All reagents used were of analytical grade.
Animals We used 2–4-month-old male and female mdr2 (2/2) and (1/1) mice with an FVB background.23 Animals were obtained from the colony at the Central Animal Facility, Academic Medical Center, Amsterdam, The Netherlands. Experimental protocols were approved by the Ethics Committee for Animal Experiments, Faculty of Medical Sciences, University of Groningen, Groningen, The Netherlands. The animals were housed in a light-controlled (lights on from 6 AM to 6 PM) and temperature-controlled (21°C) environment. After weaning, they were fed a commercial laboratory chow (RMH-B; Hope Farms BV, Woerden, The Netherlands) that contained 6.2% (wt/wt) fat and approximately 0.01% cholesterol (wt/wt). Food and tap water was available ad libitum.
PLASMA LIPIDS IN mdr2 KNOCKOUT MICE
1025
Experimental Procedures Male and female mice of both genotypes, 4–5 per group, were anesthetized with halothane. A large blood sample (0.6–1.0 mL) was obtained by heart puncture, transferred to ethylenediaminetetraacetic acid–containing tubes, and centrifuged to obtain plasma. A portion of the plasma samples was kept at 4°C and used for fast protein liquid chromatography (FPLC) size fractionation of lipoproteins as detailed below within 48 hours. The remainder of the samples was stored at 220°C until analysis. Immediately after blood collection, the liver was excised, weighed, and frozen in liquid nitrogen for subsequent determination of hepatic lipid content and of messenger RNA (mRNA) levels (see below). Additional small blood samples (0.1 mL) for plasma lipid analysis only were obtained from 7–9 animals per group by tail bleeding under light halothane anesthesia. Separate groups of male (2/2) and (1/1) mice kept on the same diet were used for determination of the hepatic activities of HMG-CoA reductase and acyl– coenzyme A:cholesterol acyltransferase (ACAT). Cholesterol kinetics were determined after an intravenous injection of 3H-labeled cholesterol (0.33 mg, 3 µCi) dissolved in a 20% Lipofundin S (Braun Medical BV, Oss, The Netherlands) solution in phosphate-buffered saline via a tail vein. Blood samples were drawn at 30 minutes, 4 hours, and 1, 2, 3, 4, 7, 10, 14, and 17 days after injection by tail bleeding. Blood samples (0.06 mL) were extracted according to the method of Quarfordt et al.,24 and 3H activity in the extracts was measured in a scintillation counter. Total cholesterol in the extracts was determined enzymatically (see below). Analysis of the data was performed by two-compartment analysis as proposed by Quarfordt et al.24 with MultiFit, a computer program developed by Dr. J. H. Proost, University Centre for Pharmacy, Groningen, The Netherlands. Feces were collected from individual mice in timed intervals. Freeze-dried samples were extracted, and radioactivity in the chloroform phase and the water phase, the latter after concentration on SepPak C18 columns (Waters Corp., Milford, MA), was determined. Cholesterol absorption was assessed by the dual-isotope method described by Zilversmit and Hughes25 as recently modified for use in rodents by Turley et al.26 In short, animals were given an intravenous dose of 3H-labeled cholesterol (0.27 mg, 2.5 µCi) dissolved in Lipofundin S and, at the same time, an oral dose of 14C-labeled cholesterol (0.07 mg, 1.0 µCi) dissolved in medium-chain triglyceride oil. After 48 hours, animals were killed, and the ratio between 14C-labeled and 3H-labeled cholesterol was determined in plasma by scintillation counting. The formula used to calculate the cholesterol absorption was: % Cholesterol absorption 5
1
2
% Intragastric 14C Dose per mL Plasma Intravenous H3 Dose per mL Plasma
3 100.
1026
VOSHOL ET AL.
The 14C/3H ratio of hepatic and intestinal cholesterol was also measured to check equilibration of cholesterol administered via the different routes. All experiments were performed between noon and 4 PM to exclude potential effects of circadian variations; animals were not fasted before experiments unless stated otherwise.
Lipid Analyses Plasma concentrations of cholesterol, cholesterol ester, and triglycerides (without measuring free glycerol) were measured using commercially available kits (Boehringer Mannheim) according to the instructions provided. Contents of cholesterol, cholesterol ester, triglycerides, and phospholipids in liver tissue were determined after lipid extraction27 as described previously.28
Lipoprotein Size Fractionation For FPLC size fractionation of plasma lipoproteins, 0.2 mL of pooled plasma from at least 3 animals per group was injected onto a 25-mL Superose 6 prep grade column (Pharmacia, Uppsala, Sweden) and was eluted at a constant flow of 0.5 mL/min with phosphate-buffered saline (pH 7.4). The effluent was collected in 0.5-mL fractions, and cholesterol and triglyceride concentrations were measured as described above. The lipoproteins were identified on the basis of the elution profile of human plasma lipoproteins.
Western Blot Analysis For Western blot analysis, proteins of FPLC samples containing VLDL/LDL and high-density lipoprotein (HDL) fractions were size-separated by sodium dodecyl sulfate– polyacrylamide gel electrophoresis using 4%–15% gradient gels (Tris-glycerine, Ready Gels; Bio-Rad, Hercules, CA). Proteins were transferred to nitrocellulose membranes (enhanced chemiluminescence–hyperbound; Amersham) by tankblotting (Bio-Rad) followed by incubation with polyclonal rabbit anti-rat apoB (provided by Dr. R. A. Davis, San Diego, CA) and with rabbit anti-human apoA-I (immunoglobulin G fraction; Calbiochem, San Diego, CA). As a second antibody, a goat anti-rabbit immunoglobulin G horseradish peroxidase (Amersham) was used. Detection was performed by Western blotting detection reagents (Amersham) according to the instructions provided.
HMG-CoA Reductase and ACAT Assays Microsomes were isolated29 from livers of (2/2) and (1/1) male mice for determination of HMG-CoA reductase and ACAT activities, as described by Philipp and Shapiro30 and Billheimer et al.,31 respectively.
Determination of mRNA Levels mRNA was isolated from livers of (2/2) and (1/1) male mice and analyzed for mRNA levels of HMG-CoA reductase, cholesterol 7a-hydroxylase, cholesterol 27-hydroxy-
GASTROENTEROLOGY Vol. 114, No. 5
lase, LDL receptor, apoB, apoA-I, albumin, and fibrinogen as described previously.32
Fecal Sterol Analysis Seven to nine male (2/2) and (1/1) mice were housed individually, and total fecal production during a 1-week period was separated from the wood shavings. Fecal samples were lyophilized and weighed. Aliquots hereof were used for determination of neutral and acidic sterol content by gas-liquid chromatographic procedures described previously.33,34
Miscellaneous Methods Protein was determined according to Lowry et al.,35 with bovine serum albumin as the standard.
Statistics All values represent means 6 SD for the indicated number of animals per group. Differences between the two groups of mice were analyzed by means of a Mann–Whitney U test.36 A P value of ,0.05 was considered significant.
Results Hepatic Lipid Content Figure 1 shows the hepatic content of free cholesterol, cholesterol esters, triglycerides, and phospholipids, expressed as nanomoles per milligram of protein, in male (Figure 1A) and female (Figure 1B) mice of the (1/1) and (2/2) genotype. Free cholesterol in the livers of both male and female (2/2) mice tended to be increased compared with the (1/1) mice, but these differences did not reach statistical significance. No significant differences were observed in cholesterol ester and phospholipid content between (1/1) and (2/2) male and female mice, but triglyceride levels in the liver were significantly lower in female (2/2) mice than in female (1/1) mice. mdr2 Pgp deficiency resulted in increased liver weight, particularly in females, as described previously.12 Plasma Lipid Levels and Lipoprotein Profiles Figure 2 shows plasma cholesterol (Figure 2A) and triglyceride (Figure 2B) concentrations in nonfasted male and female (1/1) and (2/2) mice. Plasma cholesterol levels of both male and female (2/2) mice were reduced by about 65% compared with (1/1) controls. Cholesterol esters amounted up to about 70% of total plasma cholesterol and was similar in all groups. Heterozygous (1/2) mice showed a 25% reduction in plasma cholesterol (data not shown). Plasma triglyceride levels were decreased by about 55% in male (2/2) mice compared with male controls, whereas female (2/2) mice showed no significant differences in comparison to (1/1) mice in
May 1998
this respect. Fasted values for plasma cholesterol and triglycerides were similar to the nonfasted ones (data not shown). Figure 3 shows the cholesterol and triglyceride contents in the different plasma lipoprotein fractions of male mice after FPLC size chromatography. The reduction in plasma cholesterol in the (2/2) mice was caused by a strong reduction of the HDL size fractions (Figure 3A). The decrease of plasma triglyceride levels in the male (2/2) mice was caused by a strong reduction in the VLDL size fractions (Figure 3B), whereas the triglyceride content of the intermediate-density lipoprotein/LDL size fractions appeared to be slightly, but consistently, increased. Western blot analysis of apoB and apoA-I contents in the different FPLC fractions showed a markedly higher content of apoB48 and apoB100 (Figure 4, upper lanes) in both the VLDL and LDL fractions of knockout mice. As expected, apoA-I content was decreased (Figure 4, lower lanes) in the HDL fractions of the (2/2) mice compared with their controls. Similar patterns were observed for female mice.
Figure 1. Hepatic lipid content in (A ) male and (B ) female (1/1) (h) and (2/2) (j) mice on a regular chow diet. FC, free cholesterol; CE, cholesterol ester; TG, triglycerides; PL, phospholipids (value divided by 10). All values represent nanomoles per milligram of protein (n 5 4; mean 6 SD). *P 5 0.02, Mann–Whitney U test.
PLASMA LIPIDS IN mdr2 KNOCKOUT MICE
1027
Hepatic Enzyme Activity and mRNA Levels Table 1 shows the activities of HMG-CoA reductase and ACAT in the livers of male (1/1) and (2/2) mice. Enzyme activities were increased threefold and twofold, respectively, in (2/2) mice compared with (1/1) mice. Figure 5 shows the relative mRNA levels of HMG-CoA reductase, cholesterol 7a-hydroxylase, cholesterol 27-hydroxylase, LDL receptor, apoB, apoA-I, albumin, and fibrinogen in male (2/2) and (1/1) mouse livers. No significant differences were observed for these specific mRNA levels between (1/1) and (2/2) mice. Cholesterol Kinetics The plasma decay of radiolabeled cholesterol showed a rapid initial phase followed by a distinctly slower terminal phase (Figure 6A). Kinetic analysis of the curves showed increased distribution volumes, i.e., 12.2 6 3.3 vs. 6.3 6 0.2 mL (P , 0.05) for the first (‘‘rapid’’) compartment and 12.7 6 3.7 vs. 6.1 6 0.5 mL (P , 0.05) for the second compartment in (2/2) mice compared with (1/1) controls. No difference in choles-
Figure 2. (A ) Plasma cholesterol and (B ) triglyceride levels in chow-fed male and female (1/1) (h) and (2/2) (j) mice. Values represent millimoles per liter; n 5 10 male (1/1), n 5 8 male (2/2), n 5 8 female (1/1), and n 5 8 female (2/2). *P , 0.05, Mann–Whitney U test.
1028
VOSHOL ET AL.
GASTROENTEROLOGY Vol. 114, No. 5
measurements by the dual-isotope method.25,26 Apparent cholesterol absorption was 70% and 42% for (1/1) and (2/2) mice, respectively. As expected, the normalized ratios between 14C- and 3H-labeled cholesterol in plasma, liver, and intestine were similar in the (1/1) mice at 48 hours after injection, which was indicative for equilibration between orally and intravenously administered cholesterol across these organs. In (2/2) mice, in contrast, liver and intestine were clearly not (yet) in equilibrium with plasma, thus prohibiting any conclusion about actual cholesterol absorption efficiency by dual-isotope methodology in these animals. Fecal Neutral and Acidic Sterol Content and Secretion Figure 7 shows the fecal neutral and acidic sterol output in male (1/1) and (2/2) mice, expressed as micromoles per day. Fecal total neutral sterol (Figure 7A) and cholesterol (data not shown) excretion was increased fourfold in the (2/2) mice compared with the control (1/1) mice. Total fecal acidic sterol excretion was slightly reduced in male (2/2) mice (Figure 7B). Analysis of fecal bile salt composition showed a decrease in the relative amounts of cholate and deoxycholate in (2/2) mice, whereas the relative contribution of muricholate showed a twofold increase. Food intake and feces mass production did not differ between the groups (data not shown).
Discussion Figure 3. (A ) Cholesterol and (B ) triglyceride profiles after FPLC size chromatography of plasma from male (1/1) (dotted line) and (2/2) (bold line) mice. Values represent millimoles per liter in each fraction. The mean of three samples is shown.
terol transport (equivalent to synthesis) was noted, i.e., 7.13 6 1.36 and 7.09 6 0.98 µmol/day in (2/2) and (1/1) mice, respectively. There were no significant differences in other kinetic parameters (see legend to Figure 6). Analysis of fecal radioactivity (Figure 6B) showed that the fraction of 3H-cholesterol recovered in the neutral sterol fraction was considerably smaller in (2/2) mice than in controls but was still appreciable, i.e., ,4% dose in 17 days. Conversion of intravenously administered 3H-cholesterol to bile salts disposed into feces during the course of the experiment was somewhat slower in (2/2) than in (1/1) mice. Table 2 shows the results of cholesterol absorption
In this study, we used the mdr2 Pgp–deficient mouse model to evaluate the role of biliary lipids in the maintenance of cholesterol homeostasis and in the regulation of plasma lipid levels. Originally, mdr2 Pgp deficiency was induced in mice of the 129/Ola strain,14 but because of poor breeding of this strain, the deficiency has been bred into the FVB background.12 In the FVB strain, mdr2 Pgp deficiency leads to similar changes in bile composition as observed in the 129/Ola strain, i.e., no phospholipid secretion and a strongly reduced cholesterol secretion (297%) in the presence of a normal bile salt secretion, as described in detail by Oude Elferink et al.12 In the present study, we analyzed hepatic and plasma lipid levels, cholesterol kinetics, cholesterol absorption, and fecal sterol output in control (1/1) and mdr2 gene knockout (2/2) FVB mice that were kept on a standard, low-cholesterol/low-fat chow diet. No significant differences were observed in hepatic cholesterol and cholesterol ester content of the (2/2)
May 1998
PLASMA LIPIDS IN mdr2 KNOCKOUT MICE
1029
Figure 4. Western blot analysis of apoB and apoA-I in FPLC fractions of plasma from male (1/1) and (2/2) mice as shown in Figure 3. Protein present in the various fractions was separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis, blotted onto nitrocellulose membranes, and visualized as described in detail in Materials and Methods.
mice compared with the (1/1) mice when expressed as nanomoles per milligram of protein, the latter despite increased ACAT activity. The phospholipid content of the liver showed no significant differences across all groups, presumably indicating that hepatic phosphatidylcholine synthesis has adapted to the absence of biliary secretion. Plasma lipid analysis showed an unexpected reduction in cholesterol concentration in (2/2) mice compared with (1/1) mice in both sexes, whereas triglycerides were reduced in males only. Surprisingly, the apoB contents of VLDL and LDL fractions were increased in (2/2) animals. These effects are unlikely because of liver pathology associated with this deficiency: (1) the effects are larger in the FVB strain than previously found in the 129/Ola strain (plasma cholesterol reduction of approximately 50%), which showed more severe liver pathology than the FVB (2/2) mice12; (2) effects of mdr2 Pgp deficiency on plasma cholesterol were similar in male and female mice, whereas the latter show markedly more liver pathology15,16; (3) in mdr2 (2/2) mice, no signs of hepatocytic damage are found, and mRNA levels of acute Table 1. Hepatic Activities of HMG-CoA Reductase and ACAT in Chow-Fed Male (1/1) and (2/2) Mice
HMG-CoA reductase ACAT
(1/1) Mice
(2/2) Mice
0.06 6 0.01 0.53 6 0.01
0.21 6 0.04 a 0.95 6 0.08 a
NOTE. Values are expressed as means 6 SD (n 5 2). aP , 0.05.
phase markers albumin and fibrinogen were similar in (2/2) and (1/1) mice; (4) liver disease associated with impaired lipid secretion into bile, such as cholestasis, is generally associated with increased plasma lipid levels
Figure 5. Hepatic mRNA levels in livers of chow-fed male (1/1) (h) and (2/2) (j) mice. Values represent the ratio of specific mRNA to RIBO mRNA as detailed in Materials and Methods. Values for (1/1) mice were taken as 100%. 7a, cholesterol 7a-hydroxylase; 270H, cholesterol 27-hydroxylase; LDLr, LDL receptor; HMG, HMG-CoA reductase; ApoA-I, apolipoprotein A-I; ApoB, apolipoprotein B; alb, albumin; FNA, a-fibrinogen (n 5 3; mean 6 SD). No significant differences were observed between groups, as assessed by Mann–Whitney U test.
1030
VOSHOL ET AL.
GASTROENTEROLOGY Vol. 114, No. 5
Table 2. Apparent Cholesterol Absorption and Normalized 14C/3H Ratios in Plasma, Liver, and Intestine in (1/1) and (2/2) Mice at 48 Hours After an Oral Dose of 14C-Labeled Cholesterol and an Intravenous Dose of 3H-Labeled Cholesterol
(1/1) (2/2)
Apparent absorption
14C/3H plasma
14C/3H
liver
14C/3H intestine
70 6 13 42 6 8 a
0.7 6 0.1 0.4 6 0.1 a
0.9 6 0.2 0.7 6 0.1 a,b
0.9 6 0.3 1.3 6 0.3 b
NOTE. The results are expressed as means 6 SD. aP , 0.05, significant difference between (2/2) and (1/1) mice. bP , 0.05, significant difference ratio liver and intestine from ratio plasma.
Figure 6. (A ) Plasma cholesterol decay and (B ) recovered radioactivity in neutral and acidic fractions of feces after intravenous injection of 3H-cholesterol in chow-fed control (1/1) and (2/2) mice. Data for (1/1) mice are indicated by open symbols and those for (2/2) mice by closed symbols. (A ) Isotopic data are expressed as the fraction of injected radioactivity per micromoles of cholesterol. The plasma decay data were analyzed by a two-compartment model, showing the following kinetic parameters:
V1 (mL)
Pool size1 (mmol)
V2 (mL)
k10
k12
k21
(1/1) 6.3 6 0.2 34,4 6 4.7 6.1 6 0.5 0.24 6 0.02 0.39 6 0.18 0.39 6 0.15 (2/2) 12.2 6 3.3* 40.7 6 11.4 12.7 6 3.7* 0.24 6 0.03 0.32 6 0.11 0.32 6 0.14
V1 and V2, distribution volume of first and second compartment, respectively; k10, k12, k21, elimination rate constants describing removal and transfer between pools. (B ) Cumulative excretion of radioactivity in neutral sterol fraction (diamonds) and acidic sterol fraction (squares) after intravenous injection of 3H-cholesterol in (1/1) and (2/2) mice. Values represent the percentage of injected dose (mean 6 SD; n 5 5 per group). *Significant difference between (1/1) and (2/2) shown for the 17-day point only for reasons of clarity.
rather than reduced levels37–40; and finally, (5) heterozygous (1/2) mice show no signs of liver pathology14 but do show a 25% reduction in plasma cholesterol. At the moment, we can only speculate about the mechanisms underlying these changes in (apo)lipoprotein levels. First, a reduced absorption of dietary lipids and the absence of biliary lipids for reabsorption from the intestine may contribute. Biliary phospholipids, cholesterol, and bile salts appear to be required for efficient absorption of lipids from the intestine.3,4 Absence of biliary lipid components impairs apoB48 synthesis and secretion of apoB48-containing chylomicrons by the enterocytes.41 Recently, it was shown that specifically PC, the major phospholipid species in bile, stimulates secretion of triglyceride-rich lipoproteins by Caco-2 cells.42 Therefore, one may expect lipid malabsorption and impaired chylomicron formation in the intestine of the (2/2) mice. Because HDL particles are, in part, derived from chylomicron surface material, reduced chylomicron formation may contribute to low HDL levels in (2/2) mice. Impaired intestinal HDL formation may also play a role: interruption of the enterohepatic circulation by cholestyramine feeding43 and bile diversion44 has been shown to reduce intestinal apoA-I mRNA levels. Alternatively, HDL cholesterol45 and phospholipids46 are used preferentially for bile secretion: the reduction in HDL could be a physiological reaction of the liver to compensate for the absence of biliary lipid secretion. However, hepatic apoA-I mRNA levels were equal in both groups. Elevated apoB levels in the plasma of mdr2 (2/2) mice could be a consequence of the reciprocal relationship that has been described between biliary cholesterol secretion and VLDL secretion.47–49 Thus, maneuvers leading to increased disposition of cholesterol into bile result in decreased VLDL secretion and vice versa. This scenario would, in part, explain elevated levels of liver-derived apoB100 and apoB48 in VLDL and LDL fractions in plasma of (2/2) mice. The resulting plasma lipid profiles obviously
May 1998
Figure 7. (A ) Fecal neutral sterol and (B ) bile salt output in male (1/1) (h) and (2/2) (j) mice. Values represent the total output in micromoles per day (n 5 10 [1/1] mice and n 5 8 [2/2] mice; mean 6 SD). *Significant difference between groups, as assessed by Mann–Whitney U test.
represent the consequence of complex interactions between secretory processes and removal of lipids from plasma, interactions that cannot be fully appreciated on the basis of our results so far. To gain more insight herein, we investigated cholesterol synthesis and kinetics in detail in these animals by standard techniques. Activity of HMG-CoA reductase was increased in livers of male (2/2) mice, indicative for increased hepatic cholesterol synthesis. The mRNA levels of HMGCoA reductase were similar in the livers of male mdr2 (2/2) mice and (1/1) mice, suggesting posttranscriptional modulation of enzyme activity. Cooper et al.37,38 and others39,50 have reported increased cholesterol synthesis as well as HMG-CoA reductase activity and mRNA levels in the liver of rats with biliary obstruction, another experimental model in which the flow of biliary cholesterol, phospholipids, as well as bile salts to the intestine is absent. Reduced delivery of chylomicron-remnant cholesterol for feedback inhibition because of impaired cholesterol absorption51,52 and defective control of hepatic
PLASMA LIPIDS IN mdr2 KNOCKOUT MICE
1031
cholesterogenesis because of reduced formation of regulatory sterols37,39 have been proposed to underlie this apparent paradoxical response to bile duct ligation. It should be noted that bile duct ligation results in an increased hepatic cholesterol content when expressed per milligram of protein and in elevated plasma cholesterol concentrations in rats.39 Because these phenomena do not occur in (2/2) mice, the mechanism(s) involved in the up-regulation of HMG-CoA reductase may be different in both situations. It is clear, however, that increased bile salt synthesis, in most experimental models associated with increased hepatic cholesterol synthesis, is not the cause. We found no effects of mdr2 Pgp deficiency on hepatic mRNA levels of cholesterol 7a-hydroxylase and cholesterol 27-hydroxylase, enzymes catalyzing the first steps of the neutral and acidic pathways in bile salt synthesis, respectively.53 Furthermore, fecal bile salt output, under steady-state conditions identical to hepatic production rate, was even slightly decreased in the male knockout mice. The differences in fecal bile salt composition between (1/1) and (2/2) mice are in accordance with the differences in biliary bile salts noted earlier.13 In contrast to the measurement of hepatic HMG-CoA reductase activity, kinetic analysis of plasma cholesterol decay did not show differences in cholesterol synthesis in the central compartment, including the liver, between (2/2) and (1/1) mice, i.e., about 7–8 µmol/day in both strains. Values obtained by this approach were in excellent agreement with those of Quarfordt et al.,24 using similar methodology in mice. In fact, estimates of total body cholesterol synthesis based on the difference between intake (about 0.9 µmol/day) and output (sum of fecal neutral and acidic sterols) yielded values in the same order of magnitude for control mice, i.e., about 11 µmol/day. In marked contrast, total body cholesterol synthesis based on fecal analysis in (2/2) mice, i.e., about 30 µmol/day, was much greater than the value derived from plasma cholesterol kinetics. This discrepancy can, in our opinion, only be explained by assuming increased cholesterol synthesis in the intestine of (2/2) mice that does not contribute to plasma cholesterol. The several-fold increase in ‘‘endogenous cholesterol secretion’’ in (2/2) mice may therefore be caused by an accelerated desquamation of enterocytes, possibly caused by cytotoxic effects of lipid-free bile in conjunction with a derepressed intestinal cholesterol synthesis as a result of the absence of biliary cholesterol and disturbed cholesterol absorption. The relevance of bile in the regulation of fecal sterol secretion was also evident in a recent study performed in our laboratory,54 showing a twofold increase in neutral sterol secretion in long-term bile-diverted rats.
1032
VOSHOL ET AL.
In this condition, we noted a marked proliferation of intestinal mucosa (Minich et al., unpublished results, June 1997). Attempts to quantify the efficiency of intestinal cholesterol absorption in (2/2) mice using the dualisotope approach optimized for use in rodents by Turley et al.26 were hampered by incomplete equilibration of intravenously and intragastrically administered cholesterol in (2/2) mice. In particular, the high 14C/3H ratio in the intestine indicates that orally administered cholesterol is sequestered in the intestinal wall. Therefore, the estimates of cholesterol absorption obtained by this method, i.e., about 70% in controls and 40% in the knockouts, are not reliable for (2/2) mice. It is very likely, however, that cholesterol absorption is impaired in (2/2) mice. For instance, a recent study by Mackay et al.55 shows that pancreatic phospholipase A2–mediated phosphatidylcholine hydrolysis is required for efficient uptake of cholesterol by Caco2 cells. In any case, the data summarized in Table 2 support our suggestion that, in the absence of biliary cholesterol, cholesterol of intestinal origin does not or only slowly reaches the plasma compartment, possibly because of impaired chylomicron formation. Another point worth noting relates to the contribution of biliary cholesterol to fecal neutral sterol secretion in mice. In (1/1) mice, total neutral sterol excretion was about three times larger than the sum of the calculated daily input of biliary cholesterol into the intestine (0.96 µmol/day13) and dietary cholesterol intake (0.9 µmol/ day). Assuming that at least 70% of biliary and dietary cholesterol will be (re)absorbed, this means that the majority of fecal sterols under the dietary conditions used in this study must originate from endogenous secretion, i.e., most likely from the turnover of intestinal cells. Data shown in Figure 6B show that, in the absence of biliary cholesterol secretion, a relatively small part (4% dose per 17 days) of circulating plasma cholesterol is secreted into feces in the form of neutral sterol, a fraction that is doubled in animals with normal bile composition. Thus, endogenous secretion mainly involves cholesterol locally synthesized in the intestine. This simple calculation also supports our hypothesis that increased fecal sterol secretion in (2/2) mice originates from accelerated turnover of intestinal cells in these animals and, most importantly, underlines the role of biliary lipids in the regulation of this quantitatively very important pathway in cholesterol homeostasis. In conclusion, this study delineates the important role of biliary phospholipid and cholesterol secretion in the regulation of plasma lipid levels and maintenance of a
GASTROENTEROLOGY Vol. 114, No. 5
normal sterol balance in mice. Our results suggest that, in the absence of biliary lipid secretion, intestinal cholesterol absorption is impaired and intestinal cell turnover is increased, leading to a strongly increased fecal excretion of the sterol. This combination of effects more than compensates for the lack in biliary cholesterol secretion. Interestingly, the increase in fecal sterol excretion is accompanied by a decrease in serum HDL concentration. The molecular mechanism underlying this interaction is presently under investigation.
References 1. Dietschy JM, Turley SD, Spady DK. Role of liver in the maintenance of cholesterol and low density lipoprotein homeostasis in different animal species, including humans. J Lipid Res 1993;34: 1637–1659. 2. Packard CJ, Shepherd J. The hepatobiliary axis and lipoprotein metabolism: effects of bile acid sequestrants and ileal bypass surgery. J Lipid Res 1982;23:1081–1098. 3. Carey MC, Hernell O. Digestion and absorption of fat. Semin Gastrointest Dis 1992;3:189–208. 4. Tso P. Intestinal lipid absorption, In: Johnson LR, ed. Physiology of the gastrointestinal tract. New York: Raven, 1994:1867–1907. 5. Duckworth PF, Vlahcevic ZR, Studer EJ, Gurley EC, Heuman DM, Beg ZH, Hylemon PB. Effect of hydrophobic bile acids on 3-hydroxy3-methylglutaryl–coenzyme A reductase activity and mRNA levels in the rat. J Biol Chem 1991;266:9413–9418. 6. Spady DK, Stange EF, Bilhartz LE, Dietschy JM. Bile acids regulate hepatic low density lipoprotein receptor activity in the hamster by altering cholesterol flux across the liver. Proc Natl Acad Sci USA 1986;83:1916–1920. 7. Lin Y, Havinga R, Schippers IJ, Verkade HJ, Vonk RJ, Kuipers F. Characterization of the inhibitory effects of bile acids on very-lowdensity lipoprotein secretion by rat hepatocytes in primary culture. Biochem J 1996;316:531–538. 8. Lin Y, Havinga R, Verkade HJ, Moshage H, Slooff MJ, Vonk RJ, Kuipers F. Bile acids suppress the secretion of very-low-density lipoprotein by human hepatocytes in primary culture. Hepatology 1996;23:218–228. 9. Verkade HJ, Vonk RJ, Kuipers F. New insights into the mechanism of bile acid–induced biliary lipid secretion. Hepatology 1995;21: 1174–1189. 10. Wilson MD, Rudel LL. Review of cholesterol absorption with emphasis on dietary and biliary cholesterol. J Lipid Res 1994;35: 943–955. 11. Hay DW, Cahalane MJ, Timofeyeva N, Carey MC. Molecular species of lecithins in human gallbladder bile. J Lipid Res 1993;34:759–768. 12. Oude Elferink RP, Ottenhoff R, van Wijland M, Smit JJ, Schinkel AH, Groen AK. Regulation of biliary lipid secretion by mdr2 P-glycoprotein in the mouse. J Clin Invest 1995;95:31–38. 13. Oude Elferink RP, Ottenhoff R, van Wijland M, Frijters CM, van Nieuwkerk C, Groen AK. Uncoupling of biliary phospholipid and cholesterol secretion in mice with reduced expression of mdr2 P-glycoprotein. J Lipid Res 1996;37:1065–1075. 14. Smit JJ, Schinkel AH, Oude Elferink RP, Groen AK, Wagenaar E, van Deemter L, Mol CA, Ottenhoff R, van der Lugt NM, van Roon MA, van der Valk MA, Offerhaus GJA, Berns AJM, Borst P. Homozygous disruption of the murine mdr2 P-glycoprotein gene leads to a complete absence of phospholipid from bile and to liver disease. Cell 1993;75:451–462. 15. Nieuwkerk van CMJ, Oude Elferink RPJ, Groen AK, Ottenhoff R, Tytgat GNJ, Dingemans KP, Van den Bergh Weerman MA, Offer-
May 1998
16.
17. 18.
19.
20.
21. 22.
23.
24.
25.
26.
27. 28.
29.
30.
31.
32.
33.
34.
35.
haus GJA. Effects of ursodeoxycholate and cholate feeding on liver disease in FVB mice with a disrupted mdr2 P-glycoprotein gene. Gastroenterology 1996;111:165–171. Nieuwkerk van CMJ, Groen AK, Ottenhoff R, et al. The role of bile salt composition in liver pathology of mdr2 (2/2) mice: differences between males and females. J Hepatol 1997;26:138– 145. Robins SJ. Recirculation and reutilization of micellar bile lecithin. Am J Physiol 1975;229:598–602. Davidson NO, Kollmer ME, Glickman RM. Apolipoprotein B synthesis in rat small intestine: regulation by dietary triglyceride and biliary lipid. J Lipid Res 1986;27:30–39. Nies AT, Gatmaitan Z, Arias IM. ATP-dependent phosphatidylcholine translocation in rat liver canalicular plasma membrane vesicles. J Lipid Res 1996;37:1125–1136. Ruetz S, Gros P. Enhancement of mdr2-mediated phosphatidylcholine translocation by the bile salt taurocholate. Implications for hepatic bile formation. J Biol Chem 1995;270:25388–25395. Ruetz S, Gros P. Phosphatidylcholine translocase: a physiological role for the mdr2 gene. Cell 1994;77:1071–1081. Ruetz S, Raymond M, Gros P. Functional expression of Pglycoprotein encoded by the mouse mdr3 gene in yeast cells. Proc Natl Acad Sci USA 1993;90:11588–11592. Groen AK, Van Wijland MJ, Frederiks WM, Smit JJ, Schinkel AH, Oude Elferink RP. Regulation of protein secretion into bile: studies in mice with a disrupted mdr2 p-glycoprotein gene. Gastroenterology 1995;109:1997–2006. Quarfordt SH, Oswald B, Landis B, Xu HS, Zhang SH, Maeda N. In vivo cholesterol kinetics in apolipoprotein E–deficient and control mice. J Lipid Res 1995;36:1227–1235. Zilversmit DB, Hughes LB. Validation of a dual-isotope plasma ratio method for measurement of cholesterol absorption in rats. J Lipid Res 1974;15:465–473. Turley SD, Herndon MW, Dietschy JM. Reevaluation and application of the dual-plasma isotope plasma ratio method for the measurement of intestinal cholesterol absorption in the hamster. J Lipid Res 1994;35:328–339. Bligh EG, Dyer WJ. A rapid method of total lipid extraction and purification. Can J Biochem Biophys 1959;37:911–917. Kuipers F, Van Ree JM, Hofker HM, et al. Altered lipid metabolism in apolipoprotein E–deficient mice does not affect cholesterol balance across the liver. Hepatology 1996;24:241–247. Smit HJ, Temmerman AM, Wolthers H, Kuipers F, Beyen AC, Vonk RJ. Dietary fish-oil induced changes in intrahepatic cholesterol transport and bile acid synthesis in rats. J Clin Invest 1991;88: 943–951. Philipp BW, Shapiro DJ. Improved methods for the assay and activation of 3-hydroxy-3-methylglutaryl coenzyme A reductase. J Lipid Res 1979;20:588–593. Billheimer JT, Tarani D, Nes WR. Effect of a dispersion of cholesterol in Triton WR 1339 on acylCoA:cholesterol acyltransferase in rat liver microsomes. Anal Biochem 1981;111:331–335. Twisk J, Lehmann EM, Princen HM. Differential feedback regulation of cholesterol 7 alpha-hydroxylase mRNA and transcriptional activity by rat bile acids in primary monolayer cultures of rat hepatocytes. Biochem J 1993;290:685–691. Arca M, Montali A, Ciocca S, Angelico F, Cantafora A. An improved gas-liquid chromatographic method for the determination of fecal neutral sterols. J Lipid Res 1983;24:332–335. Setchell KD, Lawson AM, Tanida N, Sjovall J. General methods for the analysis of metabolic profiles of bile acids and related compounds in feces. J Lipid Res 1983;24:1085–1100. Lowry OH, Rosebrough NJ, Farr AL, Randall RL. Protein measurement with the folin reagens. J Biol Chem 1951;193:265–275.
PLASMA LIPIDS IN mdr2 KNOCKOUT MICE
1033
36. Dawson-Saunders B, Trapp RG. Basic and clinical biostatistics. East Norwalk, CT: Lange, 1990. 37. Cooper AD, Ockner RK. Studies of hepatic cholesterol synthesis in experimental acute biliary obstruction. Gastroenterology 1974; 66:586–595. 38. Cooper AD, Jones AL, Koldinger RE, Ockner RK. Selective biliary obstruction: a model for the study of lipid metabolism in cholestasis. Gastroenterology 1974;66:574–585. 39. Dueland S, Reichen J, Everson GT, Davis RA. Regulation of cholesterol and bile acid homoeostasis in bile-obstructed rats. Biochem J 1991;280:373–377. 40. Long TT, Jakoi L, Stevens R, Quarfordt S. The sources of rat biliary cholesterol and bile acid. J Lipid Res 1978;19:872–878. 41. Davidson NO, Drewek MJ, Gordon JI, Elovson J. Rat intestinal apolipoprotein B gene expression. Evidence for integrated regulation by bile salt, fatty acid, and phospholipid flux. J Clin Invest 1988;82:300–308. 42. Field FJ, Born E, Chen H, Murthy S, Mathur SN. Regulation of apolipoprotein B secretion by biliary lipids in CaCo-2 cells. J Lipid Res 1994;35:749–762. 43. Felgines C, Mazur A, Rayssiguier Y. Effect of the interruption of enterohepatic circulation of bile acids by cholestyramine on apolipoprotein gene expression in the rat. Life Sci 1994;55:1053– 1060. 44. Sonoyama K, Nishikawa H, Kiriyama S, Niki R. Bile diversion lowers apolipoprotein A-I and A-IV mRNA levels in rat ileum. J Nutr Sci Vitaminol Tokyo 1994;40:343–352. 45. Robins SJ, Fasulo JM. High density lipoprotein, but not other lipoproteins, provide a vehicle for sterol transport in bile. J Clin Invest 1997;99:380–384. 46. Portal I, Clerc T, Sbarra V, Portugal H, Pauli AM, Lafont H, Tuchweber B, Yousef I, Chanussot F. Importance of high-density lipoprotein-phosphatidylcholine in secretion of phospholipid and cholesterol in bile. Am J Physiol 1993;264:G1052–G1056. 47. Stone BG, Erickson SK, Craig WY, Cooper AD. Regulation of rat biliary cholesterol secretion by agents that alter intrahepatic cholesterol metabolism. Evidence for a distinct biliary precursor pool. J Clin Invest 1985;76:1773–1781. 48. Nervi F, Marinovic I, Rigotti A, Ulloa N. Regulation of biliary cholesterol secretion. Functional relationship between the canalicular and sinusoidal cholesterol secretory pathways in the rat. J Clin Invest 1988;82:1818–1825. 49. Nervi F, Bronfman M, Allalon W, Depiereux E, Del Pozo R. Regulation of biliary cholesterol secretion in the rat. Role of hepatic cholesterol esterification. J Clin Invest 1984;74:2226– 2237. 50. Weis HJ, Dietschy JM. Failure of bile acids to control hepatic cholesterogenesis: evidence for endogenous cholesterol feedback. J Clin Invest 1969;48:2398–2408. 51. Nervi FO, Dietschy JM. The mechanisms of and the interrelationship between bile acid and chylomicron-mediated regulation of hepatic cholesterol synthesis in the liver of the rat. J Clin Invest 1978;61:895–909. 52. Weis HJ, Dietschy JM. Presence of an intact cholesterol feedback mechanism in the liver in biliary stasis. Gastroenterology 1971; 61:77–84. 53. Russell DW, Setchell KD. Bile acid biosynthesis. Biochemistry 1992;31:4737–4749. 54. Bandsma RHJ, Stellaard F, Vonk RJ, Nagel GT, Neese RA, Hellerstein MK, Kuipers F. The contribution of newly synthesised cholesterol to rat plasma and bile determined by mass isotopomer distribution analysis: bile salt flux promotes secretion of newly synthesised cholesterol into bile. Biochem J 1998;329:699– 703.
1034
VOSHOL ET AL.
55. Mackay K, Starr JR, Lawn RM, Ellsworth RL. Phosphatidylcholine hydrolysis is required for pancreatic cholesterol esterase– and phospholipase A(2)-facilitated cholesterol uptake into intestinal Caco-2 cells. J Biol Chem 1997;272:13380–13389.
Received October 17, 1997. Accepted January 15, 1998. Address requests for reprints to: Peter J. Voshol, M.Sc., B.Sc., Groningen Institute for Drug Studies, Laboratory of Nutrition and Metabolism, University Hospital Groningen, CMC IV, Hanzeplein 1, 9713 GZ, Groningen, The Netherlands. e-mail:
[email protected]. Supported by grant 902-23-097 from The Netherlands
GASTROENTEROLOGY Vol. 114, No. 5
Organization for Scientific Research (NWO) and by a grant from Nutricia B.V., Zoetermeer, The Netherlands. Presented in part at the 1995 Scientific Sessions of the American Heart Association in Anaheim, California, and published in abstract form (Circulation 1995;98:I-105) and at the 1997 meeting of the European Association for Study of the Liver, London, England, and published in abstract form (J Hepatol 1997;26:123). The authors thank Elly de Wit, Renze Boverhof, and Hans Bartels for skillful technical assistance and Dr. J. H. Proost (University Centre for Pharmacy, Department of Pharmacokinetics and Drug Delivery, University of Groningen, The Netherlands) for providing the MultiFit computer software used for kinetic studies.
