Although liver fatty acid binding protein (L-FABP) is postulated to influence cholesterol homeostasis, the physiological significance of this hypothesis remains to be resolved. This issue was addressed by examining the response of young (7 wk) female mice to L-FABP gene ablation and a cholesterol-rich diet. In control-fed mice, L-FABP gene ablation alone induced hepatic cholesterol accumulation (2.6-fold), increased bile acid levels, and increased body weight gain (primarily as fat tissue mass). In cholesterol-fed mice, L-FABP gene ablation further enhanced the hepatic accumulation of cholesterol (especially cholesterol ester, 12-fold) and potentiated the effects of dietary cholesterol on increased body weight gain, again mainly as fat tissue mass. However, in contrast to the effects of L-FABP gene ablation in control-fed mice, biliary levels of bile acids (as well as cholesterol and phospholipids) were reduced. These phenotypic alterations were not associated with differences in food intake. In conclusion, it was shown for the first time that L-FABP altered cholesterol metabolism and the response of female mice to dietary cholesterol. While the biliary and lipid phenotype of female wild-type L-FABP+/+ mice was sensitive to dietary cholesterol, L-FABP gene ablation dramatically enhanced many of the effects of dietary cholesterol to greatly induce hepatic cholesterol (primarily cholesterol ester) and triacylglycerol accumulation as well as to potentiate body weight gain (primarily as fat tissue mass). Taken together, these data support the hypothesis that L-FABP is involved in the physiological regulation of cholesterol metabolism, body weight gain, and obesity.
- cholesterol ester
liver fatty acid binding protein (L-FABP) is one of a large family of nonenzymatic, lipid-binding proteins present in high amounts (3–5% of liver cytosol protein, 100–400 μM) in liver cytosol (for a review, see Ref. 18). Because of its high affinity for fatty acids, much interest has focused on the physiological role of L-FABP in fatty acid metabolism (for a review, see Ref. 18). However, increasing data suggest that L-FABP may also play a role in intracellular cholesterol dynamics.
First, structural studies show that compared with other members of this protein family, L-FABP has a ligand-binding site at least twice as large, sufficient to accommodate molecules the size of cholesterol and bile acids (for a review, see Ref. 27). Indeed, in vitro studies show that L-FABP binds cholesterol with a dissociation constant as low as 0.3 μM (for a review, see Ref. 22), and chemically blocking the L-FABP ligand-binding site inhibits sterol binding (31). A variety of in vitro radioligand and fluorescence displacement studies show that L-FABP also binds bile acids (for a review, see Ref. 28). On the basis of cross-linking studies of photoreactive bile acids, it has been suggested that L-FABP is a major cytoplasmic bile acid binding protein (7).
Second, L-FABP selectively enhances intermembrane cholesterol transfer from isolated plasma membranes (24) and from isolated plasma membranes to isolated mitochondria in vitro (for a review, see Ref. 9). Chemically blocking the L-FABP ligand-binding site inhibits intermembrane sterol transfer activity (31). Furthermore, L-FABP enhances microsomal conversion of exogenous cholesterol to cholesterol esters by acyl CoA:cholesterol acyl transferase (ACAT) in vitro (6).
Finally, L-FABP overexpression in cultured cells increases cholesterol uptake, intermembrane transfer, and intracellular cholesterol ester mass (for a review, see Ref. 14). Inhibiting cholesterol binding abolishes L-FABP-mediated enhancement of cellular cholesterol uptake in transfected cells and inhibits L-FABP-mediated cholesterol transfer from the plasma membrane to the endoplasmic reticulum for cholesterol esterification in transfected cells (14).
Despite the above studies, relatively little is known regarding the physiological relevance of L-FABP to cholesterol dynamics. It was recently noted that L-FABP is upregulated as much as four- to fivefold in male sterol carrier protein (SCP)-x/SCP-2 gene-ablated mice, concomitant with decreased liver cholesterol ester mass (26) and hypersecretion of cholesterol in bile (10). While the latter finding suggests that L-FABP may be the cholesterol transporter responsible for the biliary cholesterol hypersecretion in SCP-2/SCP-x gene-ablated mice, the complexity of the SCP-2/SCP-x gene-ablated mouse precludes discrimination of the relative contributions of L-FABP upregulation, SCP-2 ablation, and SCP-x ablation to the cholesterol phenotype. Both SCP-2 and SCP-x are soluble proteins that also bind, transfer, or utilize cholesterol (for a review, see Ref. 23).
To begin to resolve these issues, the present investigation utilized sexually mature, female L-FABP gene-ablated mice to test the hypothesis that L-FABP may have a physiological role in cholesterol metabolism in response to high dietary cholesterol. It was shown that 1) L-FABP gene ablation increased hepatic cholesterol and bile acids as well as biliary lipid levels (bile acids, cholesterol, and phospholipids); 2) L-FABP gene ablation potentiated the effects of dietary cholesterol to synergistically redirect cholesterol and bile acid from the biliary lipid pool to hepatic accumulation of free and even more so cholesterol esters as well as bile acids; and 3) L-FABP gene ablation potentiated the effect of cholesterol on increased body weight gain and fat tissue mass (FTM).
MATERIALS AND METHODS
Protease inhibitor cocktail for mammalian tissues was purchased from Sigma-Aldrich (St. Louis, MO). Protein Assay Dye Reagent Concentrate was obtained from Bio-Rad Laboratories (Richmond, CA). Silica gel G thin-layer chromatography plates were purchased from Analtech (Newark, DE). Reference lipids were obtained from Nu-Chek-Prep (Elysian, MN). Biliary lipids were determined as follows: bile acids (Bile Acids-L3K Assay kit, Diagnostic Chemicals; Oxford, CT), free cholesterol (Wako kit no. 274-47109, Wako Diagnostics; Richmond, VA), and phospholipids (Wako kit no. 990-54009, Wako Diagnostics). All reagents and solvents used were of the highest available grade and were cell culture tested.
Experimental protocols for the use of laboratory animals were approved by the University Lab Animal Care Committee and met American Association for Accreditation of Laboratory Animal Care guidelines. L-FABP null (L-FABP−/−) C57BL/6 mice were obtained as previously documented (17). Mice were maintained on a standard low-fat (5% of calories from fat) pelleted rodent chow (Teklad Rodent Diet W8604, Harlan Teklad; Madison, WI), housed in a temperature-controlled (25°C) facility on a 12:12-h light-dark cycle, and allowed free access to food and water.
Because of the gender-dependent differences in expression of enzymes involved in bile acid metabolism (for a review, see Ref. 25) and the greater sensitivity of female mice to dietary branched-chain lipids such as cholesterol (for a review, see Ref. 25), the dietary study was performed with 7-wk-old female L-FABP gene-ablated (L-FABP−/−) mice along with age- and sex-matched L-FABP wild-type (L-FABP+/+) littermates as controls. Mice were housed individually and acclimated for 1 wk to water and a control-defined pelleted diet fed ad libitum. The control diet was a modified AIN-76A phytol-free and phytoestrogen-free pelleted rodent diet (5% of calories from fat, no. D11243, Research Diets; New Brunswick, NJ) (2). Once acclimated to the modified control diet (Research Diets no. D11243), mice were then either continued for 5 wk on the control pelleted diet (Research Diets no. D11243) or placed on an isocaloric cholesterol-rich pelleted diet. The latter was comprised of the same control pelleted diet (5% of calories from fat, no. D01091702, Research Diets) supplemented with 1.25% cholesterol as formulated by the manufacturer. Of the total of 32 mice used, 16 animals were maintained on the control diet and 16 animals were exposed to the 1.25% cholesterol diet. Each diet consisted of two separate groups of animals: eight L-FABP−/− females and eight L-FABP+/+ females. Throughout the course of the study, at the same time of day, the mice were weighed every other day. At the same time, all pellets and fragments of rodent food were removed from each cage and weighed, and this weight was subtracted from the weight of pelleted food initially placed in the feeding bin to determine the amount of food consumed by the animal during each 2-day period. These measurements were taken at exactly the same time of day every 2 days throughout the course of the dietary study to minimize any differences in amounts of food spilled or overall pattern of food consumption. However, because mice spill food, it is difficult to precisely measure food intake by this method. Furthermore, any differences in diurnal or other pattern of food intake throughout each day/night were not measured by this method. There were no significant differences in the amount of food consumed by the animals on either diet, as described below in results. This indicated that the mice did not exhibit any significant dietary preferences. After being weighed, animals were returned to their respective cages, and preweighed amounts of the appropriate pelleted diet were placed in the feeding bins.
Animal euthanization, tissue collection, and morphometric analysis.
At the beginning and conclusion of the dietary study, mice were examined by dual-energy X-ray absorptiometry (DEXA) to quantify the amount of body FTM and bone-free lean tissue mass (LTM) as described previously (2). Before death, each animal was fasted for 12 h, weighed, anesthetized, and again examined by DEXA. Blood was then collected from the mouse via cardiac puncture, and the blood was immediately processed to serum and stored at −80°C. The animal was euthanized by cervical dislocation; the tissues of interest were removed, flash frozen with dry ice, and stored at −80°C for subsequent analysis. The liver was removed, and a small piece of the liver was used immediately for histological analysis. The remainder of the liver was divided into small portions, flash frozen with dry ice, and stored at −80°C for subsequent analysis. The gall bladder was also removed from each animal; the bile was collected, flash frozen with dry ice, and stored at −80°C for subsequent biliary lipid analysis. Histological analysis was performed as described previously (2). The severity of fatty vacuolation in hepatocytes was scored as follows: 0, normal; 1, minimal fatty change; 2+, mild fatty change; 3+, moderate fatty change; and 4+, severe fatty change.
Western blot analysis.
Primary antibodies against the following proteins involved in cholesterol metabolism were obtained as follows: 1) rabbit polyclonal anti-mouse scavenger receptor class B type I (SR-BI) from Novus Biologicals (Littleton, CO); 2) goat polyclonal anti-mouse low-density lipoprotein (LDL) receptor, anti-human acyl-CoA:cholesterol acyltransferase 1 (ACAT-1), goat polyclonal anti-human cholesterol 7α-hydroxylase (CYP7A1), goat polyclonal anti-human sterol 27-hydroxylase (CYP27A1), rabbit polyclonal anti-human peroxisome proliferator-activated receptor-α (PPAR-α), rabbit polyclonal anti-human sterol regulatory element binding protein 1 (SREBP-1), rabbit polyclonal anti-human farnesoid X receptor (FXR), goat polyclonal anti-human liver X receptor-α (LXR-α), goat polyclonal anti-mouse short heterodimer partner protein (SHP), goat polyclonal anti-mouse bile salt export protein (BSEP), goat polyclonal anti-human multidrug resistance protein 2 (MRP2), and rabbit polyclonal anti-mouse fatty acid transport protein 1 (FATP-1) from Santa Cruz Biotechnology (Santa Cruz, CA); 3) rabbit polyclonal antisera to recombinant rat L-FABP, rabbit anti-recombinant mouse SCP-2, rabbit anti-recombinant mouse acyl-CoA binding protein (ACBP), rabbit anti-recombinant mouse SCP-x, and rabbit anti-porcine aspartate aminotransferase (AAT, i.e., GOT) were obtained as described previously (3); 4) anti-mouse caveolin-1 from Affinity Bioreagents (Golden, CO); 5) anti-human 3-hydroxy-3-methylglutaryl CoA (HMG-CoA) reductase from Upstate Cell Signaling Solutions (Lake Placid, NY); 6) anti-human ACAT-2 from Cayman Chemical (Ann Arbor, MI); 7) rabbit polyclonal anti-mouse glutathione S-transferase (GST) and rabbit polyclonal anti-Pseudomonas 3α-hydroxysteroid dehydrogenase (3α-HSD) from USBiological (Swampscott, MA); 8) rabbit polyclonal anti-rat organic anion transport polypeptide 1 (OATP1) from Alpha Diagnostic (San Antonio, TX); and 9) goat polyclonal anti-mouse fatty acid translocase (FAT; CD36) from Research Diagnostics (Flanders, NJ). The above primary antibodies were detected with secondary antibodies as follows: alkaline phosphatase-conjugate goat anti-rabbit IgG and alkaline phosphatase-conjugate rabbit anti-goat IgG were purchased from Sigma-Aldrich. For Western blot analysis, liver samples from female L-FABP−/− and L-FABP+/+ mice on control and cholesterol-rich diets were homogenized followed by centrifugation at 600 g for 10 min to remove insoluble debris (21). Western blot analysis was then performed to determine protein expression levels using the above antisera basically as described previously (3).
The mouse liver was homogenized and fractionated, and lipids were analyzed as described previously (2, 17). Total fatty acid was determined utilizing the following relationship: moles of fatty acid = [moles of cholesterol ester + moles of nonesterified fatty acid + (2 × moles of phospholipid) + (3 × moles of triacylglycerol)]. Quantification of total bile acid content from the mouse liver 105k supernatant, serum, and bile was achieved utilizing the commercially available Bile Acids-L3K Assay kit (Diagnostic Chemicals) used according to the manufacturer's directions. Biliary cholesterol and phospholipids were determined using the following commercially available kits (Wako Diagnostics): cholesterol, Wako no. 274-47109; and phospholipid, Wako no. 990-54009. Bile acid, cholesterol, and phospholipid levels were expressed as concentration (mM) and as total mass (nmol) calculated from the biliary lipid concentration and volume of bile. Protein was quantified by the Bradford protein assay (Bio-Rad) (5).
Unless otherwise stated, data are presented as means ± SD with n = 7–8 for all analyses and P indicated in the results. Graphical analysis was performed utilizing SigmaPlot 2000 for Windows version 6.10 (SPSS; Chicago, IL). Statistical analysis was performed using the one-way ANOVA with the Newman-Keuls posttest utilizing the GraphPad Prism data analysis package version 3.02 for Windows (GraphPad Software; San Diego, CA).
Effect of L-FABP gene ablation on initial body weight and food consumption.
The initial body weight of female L-FABP−/− mice (20 ± 2 g) did not significantly differ from that of wild-type L-FABP+/+ female littermates (20 ± 1 g). Control-fed, female L-FABP−/− mice consumed 12 ± 1 kcal/day, which was not significantly different from that of female wild-type L-FABP+/+ littermates fed the control diet (12 ± 1 kcal/day) regardless of whether calculated on a daily basis or at the end of the dietary study. Dietary cholesterol did not significantly affect food consumption in any of the groups examined. Qualitative visual observation of the mice every 2 days over the 5-wk dietary study did not reveal any obvious differences in activity level between control and mutant mice fed either the control or cholesterol-rich diet. These data indicate that any differences in body weight gain or lipid phenotype noted by the end of the 5-wk 1.25% cholesterol dietary study (see below) were not due to differences in 1) initial body weight, 2) visually obvious differences in activity, or 3) food intake.
Effect of L-FABP gene ablation on whole body phenotype.
Control-fed female wild-type L-FABP+/+ mice gained 0.003 ± 0.001 g body wt·day−1·kcal food consumed−1 (Fig. 1A). L-FABP gene ablation increased 3.4-fold the weight gain per day per kilocalorie of food consumed in control-fed mice (Fig. 1A). Female L-FABP−/− mice on the cholesterol-rich diet also gained 2.3-fold more weight per day per kilocalorie of cholesterol-rich food consumed than their cholesterol-fed wild-type L-FABP+/+ littermates (Fig. 1A), which was 1.6-fold (P < 0.001) more than their control-fed L-FABP−/− littermates (0.012 ± 0.001 g·day−1·kcal−1; Fig. 1A). Thus L-FABP gene ablation increased body weight gain per kilocalorie of food consumed, an effect exacerbated by cholesterol-rich diet. Basically, similar findings were obtained when weight gain was expressed as grams per day (not shown). Thus, although L-FABP gene ablation did not affect the initial body weight of female mice aged 8 wk, the L-FABP gene-ablated mice exhibited a specific alteration in body weight by 13 wk of age, as evidenced by higher body weight per kilocalorie of food consumed, higher rate of body weight gain, and higher final body weight. These effects were exacerbated by the cholesterol-rich diet.
To resolve whether increased body weight gain was due to increased FTM or LTM, mice were anesthetized and examined by DEXA at the beginning and conclusion of the dietary study. Although L-FABP gene ablation did not significantly alter the overall appearance of initial DEXA images of female mice taken at 8 wk of age (not shown), L-FABP gene ablation significantly influenced the size of female mice visualized by DEXA by 13 wk of age (Fig. 1, D vs. C). This was consistent with the 3.4-fold greater percent increase in body weight gain per day per kilocalorie of food consumed (Fig. 1B). Analysis of DEXA images revealed that L-FABP gene ablation increased the body weight of the female mice significantly, primarily by inducing a fourfold increased body FTM from 50 ± 10% to 210 ± 10% (Fig. 1G). In contrast, bone-free LTM increased threefold from 4 ± 2% to 13 ± 1% (Fig. 1H). Thus L-FABP gene ablation significantly increased weight gain, primarily as FTM, in female mice. While control-fed wild-type L-FABP+/+ female mice gained nearly 10-fold more weight as FTM versus LTM, L-FABP gene ablation increased the proportion gained as FTM versus LTM to >20-fold. Representative DEXA images of cholesterol-fed L-FABP+/+ mice (Fig. 1E) were visually slightly larger compared with control-fed L-FABP+/+ littermates (Fig. 1C). The cholesterol-rich diet increased the percent FTM by 2.4-fold from 50 ± 10% to 120 ± 10% (Fig. 1G) and increased by 2-fold the percent LTM from 4 ± 1% to 8 ± 2% (Fig. 1H) in wild-type L-FABP+/+ female mice. Thus cholesterol-fed wild-type L-FABP+/+ female mice gained more weight as FTM versus LTM. L-FABP gene ablation further exacerbated the effects of cholesterol, such that the overall appearance of representative cholesterol-fed L-FABP−/− mice was obese (Fig. 1F) compared with either cholesterol-fed L-FABP+/+ littermates (Fig. 1E) or control-fed L-FABP+/+ littermates (Fig. 1C). Cholesterol-fed L-FABP−/− mice exhibited a 2.7-fold increased percent FTM from 120 ± 10% to 320 ± 40% (Fig. 1G). Concomitantly, the percent LTM was increased twofold from 8 ± 2% to 16 ± 1% (Fig. 1H). Thus cholesterol-fed L-FABP−/− female mice gained even more weight as FTM versus LTM than did cholesterol-fed wild-type L-FABP+/+ littermates, indicating that cholesterol feeding exacerbated the effects of L-FABP gene ablation.
Effect of L-FABP gene ablation on liver morphology and total lipids.
Gross liver weight of control-fed female L-FABP−/− mice (3.6 ± 0.3 g/100 g body wt) did not significantly differ from that of wild-type L-FABP+/+ female littermate mice (4.2 ± 0.8 g/100 g body wt). Cholesterol feeding also did not significantly alter liver weight per 100 g body wt of either L-FABP−/− or wild-type L-FABP+/+ littermates (data not shown). Neither L-FABP gene ablation nor cholesterol feeding had a significant effect on the amount of total protein isolated from either liver homogenate or the 105k supernatant (data not shown). On histological examination, livers of control-fed L-FABP−/− mice appeared similar and exhibited small numbers of fatty vacuoles, distributed midzonal to periportal, and these parameters were similar to those of control-fed wild-type L-FABP+/+ littermates (not shown). In contrast, in livers of cholesterol-fed wild-type L-FABP+/+ mice, fatty vacuolation was apparent and increased fourfold (P < 0.01, not shown). Although the extent of fatty vacuolation in cholesterol-fed L-FABP−/− mice distributed more diffusely with a tendency toward microvesicular vacuolation in centrilobular hepatocytes and macrovesicular vacuolation in midzonal to periportal hepatocytes and appeared greater than in the cholesterol-fed L-FABP+/+ littermates, this morphological change did not achieve statistical significance (not shown). To confirm that the morphological increased vacuolation in livers of cholesterol-fed mice, especially L-FABP−/− mice, was due to increased lipids, liver lipids were extracted and quantified as described in materials and methods. Livers of control-fed wild-type L-FABP+/+ mice contained 250 ± 20 nmol/mg protein of total lipid (Fig. 2A). The cholesterol-rich diet significantly increased liver total lipid content in wild-type L-FABP+/+ mice by 1.9-fold (P < 0.001; Fig. 2A). L-FABP gene ablation alone increased these levels much less, as evidenced by only 30% higher total lipid (Fig. 2A). The cholesterol-rich diet did not further exacerbate the effects of L-FABP gene ablation on total lipid content (Fig. 2A).
Effect of L-FABP gene ablation on liver lipid distribution.
Livers of control-fed wild-type L-FABP+/+ mice contained only a small amount (12 ± 3 nmol/mg liver homogenate protein) of total cholesterol (71% unesterified + 29% esterified cholesterol; Fig. 2, H and G), representing only 4.6% of total liver lipid (Fig. 2C). L-FABP gene ablation alone increased the total cholesterol level by 2.8-fold (P < 0.05; Fig. 2C), with unesterified cholesterol increased by 2.2-fold (P < 0.01; Fig. 2H) and esterified cholesterol increased by 3.6-fold (P < 0.05; Fig. 2G). The cholesterol-rich diet alone significantly increased liver total cholesterol content in wild-type L-FABP+/+ mice even more, by 6.3-fold (P < 0.01; Fig. 2C), primarily as esterified cholesterol by 14.5-fold (P < 0.001; Fig. 2G) and less so as unesterified cholesterol by 2.4-fold (P < 0.01; Fig. 2H). Finally, L-FABP gene ablation exacerbated this effect of the cholesterol-rich diet by 2.2-fold (P < 0.01) compared with cholesterol-fed wild-type L-FABP+/+ mice (Fig. 2C), primarily as cholesterol ester such that livers of cholesterol-fed L-FABP−/− mice exhibited the highest levels of cholesterol ester (12-fold, P < 0.01; Fig. 2G) of any group examined. The dramatic hepatic cholesterol accumulation was not reflected in serum cholesterol, which increased only slightly (20%, P < 0.05, not shown) in cholesterol-fed, but not control fed, L-FABP−/− mice. Taken together, these data indicated that the most prominent effect of dietary cholesterol on liver cholesterol distribution was to reverse the distribution of unesterified (highest in control-fed L-FABP+/+ mice) to primarily esterified cholesterol (highest in cholesterol-fed gene-ablated L-FABP−/− mice). While L-FABP gene ablation alone increased the total cholesterol, especially esterified and less so unesterified cholesterol in liver lipids, L-FABP gene ablation significantly enhanced the responsiveness of the liver to dietary cholesterol such that unesterified and esterified cholesterol represented 17% and 83%, respectively.
Examination of the fatty acid-containing liver lipid species in control-fed female wild-type L-FABP+/+ mice revealed the following quantitative rank order: triacylglycerol > phospholipids > nonesterified fatty acid > cholesterol ester (Fig. 2). While L-FABP gene ablation alone did not alter the total fatty acid content of the liver (Fig. 2B), the pattern of fatty acid distribution was significantly modified, such that the rank order of these lipid classes shifted to phospholipids and triacylglycerol > nonesterified fatty acid > cholesterol ester. This effect differed markedly from that of the cholesterol-rich diet alone, which altered the rank quantitative order of fatty acid-containing species to phospholipids > triacylglycerol > cholesterol ester > nonesterified fatty acid (Fig. 2). L-FABP gene ablation further exacerbated these effects of the cholesterol-rich diet such that the level of total fatty acid was increased to the highest level of any group examined (Fig. 2B). Specifically, nonesterified fatty acid was increased 2.5-fold (P < 0.01; Fig. 2D), cholesterol ester was increased 12-fold (Fig. 2G), and triacylglycerol was increased 1.5-fold (Fig. 2E), whereas phospholipids were unchanged compared with control-fed L-FABP−/− mice. As a result, cholesterol-fed L-FABP−/− mouse liver lipids exhibited a quantitative rank order of triacylglycerol, cholesterol ester, and phospholipid > nonesterified fatty acid (Fig. 2). Thus cholesterol-fed female L-FABP−/− mice exhibited a significantly altered pattern of esterified fatty acids in favor of lipids typically found in lipid droplets, i.e., triacylglycerols and cholesterol esters, consistent with the morphological observations.
Effect of L-FABP gene ablation on the major liver oxidation products of cholesterol (bile acids) and biliary lipids (bile acids, cholesterol, and phospholipids).
L-FABP gene ablation alone increased liver bile acid concentration (nmol/mg protein) 1.5-fold (P < 0.05; Fig. 3A), whereas dietary cholesterol alone increased the bile acid concentration in the soluble fraction of liver homogenate 2.5-fold (Fig. 3A), of wild-type L-FABP+/+ mice. Surprisingly, L-FABP gene ablation did not synergistically increase the effect of dietary cholesterol on liver bile acid level but instead decreased the concentration of bile acids in the soluble fraction of liver homogenate 2.4-fold (Fig. 3A), which was 4-fold lower than that in cholesterol-fed wild-type L-FABP+/+ mice (Fig. 3A). As expected, serum bile acid concentration (not shown) and total mass in nanomoles (Fig. 4A) were very low (<0.1% of total), likely due to the rapid reabsorption of bile acids from serum back into the liver, but overall reflected the pattern observed in the liver (not shown). The tissue accounting for the highest bile acid concentration in millimolars (Fig. 3B) and total mass in nanomoles (Fig. 4C) was gall bladder bile. L-FABP gene ablation alone also increased the biliary concentration (Fig. 3B) and total mass (P < 0.01; Fig. 4C) of bile acid by 1.22-fold (P < 0.01) in the gall bladder. In contrast, dietary cholesterol alone did not increase but rather decreased slightly the bile acid concentration (Fig. 3B) and total mass (Fig. 4C) in bile isolated from the gall bladder of wild-type L-FABP+/+ mice, an effect exacerbated by L-FABP gene ablation (Figs. 3B and 4C). Thus L-FABP gene ablation alone increased bile acid concentration and mass in serum, the gall bladder, and, to a lesser extent, the liver. Dietary cholesterol lowered biliary bile acid concentration and mass (but not liver or serum), an effect exacerbated by L-FABP gene ablation.
Biliary cholesterol concentration in millimolars (Fig. 3C) and total mass in nanomoles (Fig. 4D) for control-fed wild-type L-FABP+/+ mice were increased 1.7-fold (P < 0.05) by L-FABP gene ablation alone but not by dietary cholesterol alone (Figs. 3C and 4D). In contrast, cholesterol-fed L-FABP−/− mice exhibited decreased biliary cholesterol concentration (Fig. 3C) and total mass (Fig. 4D). The other major lipid present in gall bladder bile is phospholipid. Biliary concentration (mM) and mass (nmol) of phospholipids were two- to threefold greater than those of cholesterol in control-fed wild-type L-FABP+/+ mice (Figs. 3, C and D, and 4, D and E). Dietary cholesterol alone significantly reduced the biliary phospholipid concentration (Fig. 3D) and total mass (Fig. 4E) 1.7-fold in the gall bladder of wild-type L-FABP+/+ mice. In contrast, L-FABP gene ablation alone increased the biliary concentration (Fig. 3D) and mass (Fig. 4E) of phospholipids 1.3-fold (P < 0.05). L-FABP gene ablation exacerbated the effect of dietary cholesterol to decrease the concentration (Fig. 3D) and mass (Fig. 4E) of phospholipids in gall bladder bile. These effects were reflected in sum total biliary lipid concentration (Fig. 3E) and mass (Fig. 4F). However, neither L-FABP gene ablation alone, cholesterol-rich diet alone, nor both together significantly altered the relative proportion of biliary cholesterol, which remained near 4% (Fig. 4G).
Effect of cholesterol and L-FABP gene ablation on key proteins in endogenous cholesterol synthesis (HMG-CoA reductase), storage (ACAT), and uptake (LDL receptor and SR-BI).
L-FABP gene ablation alone did not significantly alter the HMG-CoA reductase protein level in control-fed mice, and, as expected, the cholesterol-rich diet alone decreased HMG-CoA reductase 40–50% (P < 0.001; Fig. 5A). Cholesterol-fed L-FABP−/− mice had essentially the same level of HMG-CoA reductase protein as exhibited by cholesterol-fed wild-type L-FABP+/+ mice (Fig. 5A). With regard to the two isoforms of ACAT (ACAT2 and ACAT1) in the mouse liver, L-FABP gene ablation alone increased the level of ACAT2 modestly (30%, P < 0.001; Fig. 5C) and ACAT1 only slightly (10%; Fig. 5B) in control-fed female L-FABP−/− mice, again consistent with a modest increase in the liver cholesterol ester level (Fig. 2G). The high-cholesterol diet alone increased the level of ACAT2 protein much more, by 1.7-fold (P < 0.001; Fig. 5C), and ACAT-1 less so (33%, P < 0.01; Fig. 5B) in wild-type L-FABP+/+ mice, consistent with the increased concentration of cholesterol ester in the liver (Fig. 2G). The high-cholesterol diet also increased the level of ACAT2 protein by 1.7-fold (P < 0.001; Fig. 5C) and less so ACAT-1 (Fig. 5B) in L-FABP−/− female mice compared with control-fed female L-FABP−/− mice, and the increase in ACAT-2 was 1.3-fold greater than that elicited by cholesterol in female wild-type L-FABP+/+ mice (P < 0.001; Fig. 5C). Thus ACAT2 protein in cholesterol-fed female L-FABP−/− mice was the highest of any group examined (Fig. 5C), consistent with the highest level in cholesterol ester (Fig. 2G). Thus L-FABP gene ablation did not obscure the normal effects of dietary cholesterol on regulating the levels of liver HMG-CoA reductase (decrease) and ACAT2 (increase). However, the fact that the level of HMG-CoA reductase was lowered indicated that this did not account for the observed hepatic cholesterol accumulation and hypercholesterolemia observed in L-FABP−/− mice and even more so in cholesterol-fed L-FABP−/− mice. In contrast, L-FABP gene ablation significantly potentiated the effects of dietary cholesterol on ACAT2 but not ACAT1 protein expression compared with control-fed L-FABP−/− mice and cholesterol-fed wild-type L-FABP+/+ female mice. With regard to the expression of proteins involved in cholesterol uptake, LDL receptor protein was increased by 70% (P < 0.05) in L-FABP−/− mice compared with that observed in wild-type L-FABP+/+ animals (Fig. 5D). The cholesterol-rich diet alone increased (40%) the level of LDL receptor in wild-type L-FABP+/+ mice. Likewise, the cholesterol-rich diet tended to increase (20%) LDL receptor levels in liver homogenates from L-FABP−/− mice compared with control-fed L-FABP−/− littermates (Fig. 5D). With regard to SR-BI, the high-density lipoprotein receptor, L-FABP gene ablation alone increased SR-BI expression slightly (Fig. 5E). The cholesterol-rich diet alone increased SR-BI by ∼40% (P < 0.01) in L-FABP+/+ mice (Fig. 5E). L-FABP gene ablation did not further exacerbate the effect of the cholesterol-rich diet on increasing SR-BI (Fig. 5E). Thus upregulation of ACAT-2, the major ACAT of importance in the liver hepatocyte and the one most closely associated with cholesterol ester synthesis for secretion (16), likely contributed to the marked elevation of hepatic esterified cholesterol in cholesterol-fed L-FABP −/− mice. Furthermore, the marked hepatic cholesterol accumulation induced by the cholesterol-rich diet in the liver of L-FABP−/− mice was not associated with marked upregulation of LDL receptor or SR-BI.
Response of liver intracellular (L-FABP, SCP-2, and caveolin-1) and plasma membrane (FATP-1, FAT/CD36, and AAT) proteins that bind/transport cholesterol and/or fatty acids.
Because cholesterol and fatty acid metabolism are coregulated (for a review, see Ref. 32), it was important to investigate whether concomitant upregulation of other cholesterol or fatty acid binding/transport proteins present in the mouse liver might contribute to the effects of L-FABP gene ablation and the cholesterol-rich diet on the cholesterol and lipid phenotype. With regard to the major intracellular cholesterol and fatty acid binding proteins, L-FABP gene ablation alone increased the level of SCP-2 by 1.9-fold (P < 0.001; Fig. 6B) while concomitantly decreasing expression of caveolin-1 in liver homogenates of control-fed mice (Fig. 6C). Although the cholesterol-rich diet elicited small decreases (14% and 18%) in the amounts of L-FABP (Fig. 6A) and caveolin-1 (Fig. 6C), the level of SCP-2 was increased 1.5-fold (P < 0.01; Fig. 6B) in wild-type L-FABP+/+ mice. However, L-FABP gene ablation only modestly potentiated this cholesterol-induced increase by 10% compared with control-fed L-FABP−/− mice (Fig. 6B). With regard to the plasma membrane fatty acid binding/translocase proteins, L-FABP gene ablation elicited no or only very small changes in the level of these proteins (data not shown), consistent with earlier findings from this laboratory (17). Likewise, cholesterol-rich diets elicited no or only minor alterations in these proteins regardless of L-FABP gene ablation (not shown). Neither wild-type L-FABP+/+ nor L-FABP−/− female mice fed the cholesterol-rich diet exhibited significantly altered levels of the fatty acyl CoA binding protein ACBP (data not shown). Overall, these changes in cholesterol and fatty acid binding proteins were unlikely to account for the exacerbation of hepatic cholesterol ester accumulation in cholesterol-fed L-FABP−/− mice.
Effect of L-FABP gene ablation and cholesterol on nuclear receptors involved in cholesterol and fatty acid metabolism (SREBP-1 and PPAR-α).
It is known that SREBP-1 and PPAR-α are primarily involved in cholesterol and fatty acid metabolism, respectively. However, both proteins are coregulated by dietary fat (20). SREBP-1 exists in two forms. Neither cholesterol-rich diet nor L-FABP gene ablation significantly altered the level of the inactive P form of SREBP-1 in female mice (data not shown). Western blot analysis showed that L-FABP gene ablation alone had no effect on the level of the active N form of SREBP-1 in livers of control-fed L-FABP−/− mice (Fig. 6D). In contrast, as expected, the cholesterol-rich diet decreased the level of the active N form of SREBP-1 by 44% (P < 0.001) in female L-FABP+/+ mice (Fig. 6D). L-FABP gene ablation reversed the effect of the cholesterol-rich diet in L-FABP−/− mice but did not increase active N form of SREBP-1 above the level observed in control-fed L-FABP−/− mice (Fig. 6D). Finally, Western blot analysis of PPAR-α in female mouse liver homogenates showed no significant effects of L-FABP gene ablation, cholesterol-rich diet, or both together (Fig. 6E). Taken together, these data were consistent with the hepatic accumulation of cholesterol in the liver of cholesterol-fed L-FABP−/− mice not being due to dramatically increased levels of the nuclear receptors SREBP-1 and PPAR-α.
L-FABP gene ablation and cholesterol alter expression of liver enzymes, nuclear receptors, and transporters involved in bile acid metabolism.
L-FABP gene ablation alone slightly increased the expression of CYP7A1 (the rate-limiting enzyme of bile acid synthesis) and CYP27A1 without altering SCP-x in control-fed L-FABP−/− mice (Table 1). The cholesterol-rich diet alone increased the expression of both CYP7A1 and CYP27A1 by 1.4-fold without significantly altering SCP-x in wild-type L-FABP+/+ mice (Table 1). In contrast, L-FABP gene ablation reduced the expression of CYP7A1 and CYP27A1 by 20–28% without significantly altering the level of SCP-x in cholesterol-fed L-FABP−/− mice (Table 1). With regard to nuclear receptors involved in bile acid metabolism, L-FABP gene ablation alone modestly increased positive nuclear hormone receptors (1.3-fold LXR-α) and decreased negative nuclear hormone receptors (30% FXR and 24% SHP; Table 1), consistent with the increased expression of CYP7A1 and increased liver bile acid. In cholesterol-fed wild-type L-FABP+/+ mice, the level of the positive nuclear hormone receptor LXR-α was increased 1.6-fold, whereas the levels of the negative nuclear hormone receptors FXR and SHP decreased 2-fold (Table 1). The net effect of these cholesterol-induced changes was consistent with the observed increase in CYP7A1 and liver bile acid level in cholesterol-fed wild-type L-FABP+/+ mice. In contrast, L-FABP gene ablation oppositely altered the expression of these nuclear receptors in cholesterol-fed L-FABP−/− mice (Table 1): the positive regulator (LXR-α) was decreased 1.9-fold and the negative regulators FXR and SHP were increased 2- and 1.6-fold, respectively. The predominance of negative regulatory effects of the cholesterol-rich diet in female L-FABP−/− mice was consistent with the observed reduced expression of CYP7A1 and reduced liver level of bile acid.
At least two classes of proteins involved in bile acid binding/transport through the cytoplasm (L-FABP, GST, and 3α-HSD) or across the plasma membrane (BSEP, MRP2, and OATP-1) were examined by Western blot analysis of liver homogenates. In response to loss of L-FABP alone, the levels of GST and 3α-HSD were upregulated 1.5- and 1.4-fold, respectively, in livers of L-FABP−/− mice (Table 2). In response to the cholesterol-rich diet alone, the expression of some cytosolic bile acid binding/transport proteins was decreased slightly, e.g., L-FABP and GST, whereas that of 3α-HSD was increased significantly by 1.7-fold (Table 2). However, L-FABP gene ablation markedly altered the response to dietary cholesterol such that not only GST but also 3α-HSD was significantly decreased up to twofold (Table 2). L-FABP gene ablation significantly increased the expression of canalicular bile acid transporters (BSEP and MRP2) but not that of the basolateral/serosal bile acid transporter OATP1 (Table 2). The cholesterol-rich diet alone did not affect (BSEP and OATP1) or only slightly affected (MRP2) the expression of plasma membrane bile acid transporters (Table 2). In contrast, cholesterol-fed L-FABP−/− littermates exhibited dramatically decreased levels of canalicular bile acid transporters (BSEP and MRP2) but not that of the basolateral/serosal bile acid transporter OATP1 (Table 2). Thus decreased bile acid level in serum, the liver, and bile of cholesterol-fed L-FABP−/− mice compared with their control-fed wild-type L-FABP+/+ littermates was associated with 1) reduced expression of two key enzymes in bile acid synthesis (CYP7A1 and CYP27A1), which in turn was consistent with decreased nuclear receptors positively regulating transcription and increased nuclear receptors negatively regulating transcription of such enzymes; and 2) decreased levels of nearly all the proteins involved in transporting bile acids to the bile canaliculus and across the bile canalicular membrane of the liver hepatocyte.
Since its discovery over two decades ago, studies in vitro and with cultured cells have begun to resolve direct roles of L-FABP in lipid metabolism. The available evidence suggests that L-FABP may function not only in fatty acid but also cholesterol metabolism (for reviews, see Refs. 18 and 22). Unlike other members of the fatty acid binding protein family, the ligand-binding cavity of L-FABP is severalfold larger and can accommodate at least two (rather than one) fatty acids or even larger ligands (27) such as cholesterol (for a review, see Ref. 22). Chemically blocking the sterol binding site of L-FABP inhibits 1) L-FABP-mediated cholesterol transfer in vitro (31), 2) L-FABP-mediated cholesterol transfer in cultured cells (14), and 3) ACAT-mediated esterification of plasma membrane-derived cholesterol in the endoplasmic reticulum of cultured cells (14). Although the affinity of L-FABP for cholesterol is at least 10-fold weaker than that of other intracellular cholesterol binding proteins (for reviews, see Refs. 11 and 22), the cytosolic concentration of L-FABP is two to three orders of magnitude higher than any of the other intracellular cholesterol binding proteins (for reviews, see Refs. 11 and 18). Despite these and other studies performed in vitro and with cultured cells, however, the physiological relevance of L-FABP in cholesterol metabolism remains to be resolved. The results presented herein with L-FABP gene-ablated mice support the hypothesis that L-FABP has a physiological role in cholesterol metabolism.
The data suggest that L-FABP is necessary to support the normal increase of hepatic conversion of cholesterol to bile acids during cholesterol feeding. Several factors may contribute to this observation. First, because L-FABP binds and is directly involved in the transfer of cholesterol (contains a branched side chain) within the cell (for reviews, see Refs. 14 and 22), it may target cholesterol to the peroxisome for oxidation. Analogously, L-FABP not only binds another branched-chain lipid, phytanic acid, but also mediates its transfer to peroxisomes for oxidation (for a review, see Ref. 1). Second, upregulation of L-FABP correlates with hypersecretion of cholesterol in bile (10, 15). Conversely, as shown herein, cholesterol-fed L-FABP−/− mice exhibited significantly reduced mass of cholesterol as well as bile acid and phospholipids in gall bladder bile. Third, altered cholesterol metabolism may be an indirect effect of altered fatty acid metabolism. A more rapid secretion of very-low-density lipoproteins in cholesterol-fed L-FABP−/− mice could bind to and thus “wash out” cholesterol and cholesterol ester from the cytoplasm before it has a chance to reach peroxisomes. However, serum levels of cholesterol ester were only modestly or slightly altered, whereas serum cholesterol was slightly reduced (not shown). Finally, cholesterol-fed L-FABP−/− mice accumulated higher levels of total cholesterol, not only as unesterified cholesterol but even more so as esterified cholesterol (consistent with significant upregulation of ACAT2), in their livers. This was consistent with the interpretation that lipid-soluble cholesterol ester accumulates by partitioning into the triglyceride droplets, yet another mechanism whereby cholesterol could be diverted from processing into bile acids.
L-FABP gene ablation inhibited the biliary response to dietary cholesterol. Wild-type cholesterol-fed rodents do not develop hypercholesterolemia because cholesterol catabolism to form bile salts is increased (for a review, see Ref. 29), a conclusion supported by the findings of 1) increased expression of positive (LXR) and decreased expression of negative (FXR and SHP) nuclear receptors regulating bile acid synthesis, 2) increased expression of the rate-limiting enzyme of bile acid synthesis (CYP7A1), and 3) increased expression of intracellular bile acid binding/transport proteins (3α-HSD and MRP2). In contrast, cholesterol-fed L-FABP−/− mice exhibited hepatic cholesterol accumulation concomitant with decreased concentrations of bile salts (serum, liver, and gall bladder bile), biliary cholesterol, and biliary phospholipids. Although it would be tempting to suggest that the reduced bile acid metabolic parameters resulted in decreased biliary secretion into gall bladder bile, such a conclusion would require the measurement of actual biliary secretion rates, an undertaking for future investigations. Consistent with such a suggestion, however, L-FABP upregulation (i.e., in SCP-2/SCP-x gene-ablated mice) results in increased expression of CYP7A1 and biliary hypersecretion (10, 15) along with reduced hepatic levels of neutral lipids (cholesterol ester and triacylglycerol) (26). Although the exact mechanism whereby L-FABP gene ablation elicits these effects is not well understood, L-FABP binds not only fatty acids but also bile acids and cholesterol (see the Introduction). This suggests that with excess exogenous cholesterol, the absence of L-FABP (together with lack of compensation by other proteins that bind these ligands) may prevent the normal targeting of these ligands to nuclear receptors responsible for expression of proteins involved in bile acid metabolism, transport, and secretion. Although such a possibility is based on the finding that L-FABP gene ablation inhibits the targeting of fatty acids and fatty acyl CoA to nuclei of living cells and directly interacts with another nuclear receptor PPAR-α (13), it remains to be shown whether this is the case for the other ligands (bile acids and cholesterol) and respective nuclear receptors.
L-FABP gene ablation induced an increase in body weight, weight gain per day per kilocalorie of food consumed, and body size in control-fed mice. This effect was specific for L-FABP−/− mice compared with ablation of other fatty acid binding proteins. Female intestinal (I-FABP) (30), adipocyte (A-FABP) (12), or heart (H-FABP) fatty acid binding protein (4) gene-ablated mice fed a control diet did not exhibit altered body weight or body weight gain. The observation of increased body weight gain in 13-wk-old female L-FABP−/− mice was apparently dependent not only on gender but also age and additional parameters (e.g., construct, generation number, diet, etc.) (8, 17, 19).
The cholesterol-rich diet significantly exacerbated the effect of L-FABP gene ablation on increased body weight gain and increased weight gain per kilocalorie of food consumed in female L-FABP−/− mice compared with their cholesterol-fed littermates. A-FABP gene-ablated mice also exhibit higher weight gain on a high-fat diet compared with their wild-type A-FABP+/+ counterparts (12). In contrast, when I-FABP gene-ablated female mice were fed a high-fat, high-cholesterol diet, weight gain was similar to that of their wild-type I-FABP+/+ counterparts (30). These data suggest that L-FABP, as well as some other intracellular fatty acid binding proteins (12), may function as lipid-sensing components of energy homeostasis.
L-FABP gene ablation induced obesity, primarily as increased FTM, in female mice, and this effect was exacerbated by the cholesterol-rich diet. This obesity-inducing effect of L-FABP gene ablation differed somewhat from that of A-FABP gene ablation, where increased adiposity was noted in the cholesterol-fed, but not control-fed, A-FABP−/− mice (12). Comparison of the current data with those of earlier studies from this and other laboratories suggests that the obesity observed in the present study with 13-wk-old sexually mature female control-fed L-FABP−/− mice may be age and endocrine related (17, 19). In future studies beyond the scope of the present investigation, it would be of interest to examine in more detail the role of these parameters in obesity in L-FABP−/− mice.
In summary, the present investigation contributed significantly to our understanding of the physiological significance of L-FABP in cholesterol metabolism in animals. These studies with sexually mature (13 wk) female L-FABP−/− mice showed for the first time that L-FABP gene ablation significantly altered the response to dietary cholesterol to 1) reduce bile acid levels, thereby reducing cholesterol elimination; 2) induce accumulation of neutral lipid species (triacylglycerol and cholesterol ester) in the liver; and 3) exacerbate body weight gain, weight gain per kilocalorie of food consumed, and obesity. It should be noted that these effects of dietary cholesterol and L-FABP gene ablation were not due to alterations in food consumption or grossly apparent differences in activity. In addition, L-FABP gene ablation had no effect on dietary fat absorption (19). Similarly, ablating other fatty acid binding proteins, i.e., I-FABP (30) or SCP-2/SCP-x (26), also did not alter food intake. In conclusion, it was shown for the first time that L-FABP altered cholesterol metabolism and the response of female mice to dietary cholesterol. Whereas female wild-type L-FABP+/+ mice were sensitive to dietary cholesterol, female L-FABP−/− littermates fed a cholesterol-rich diet exhibited increased hepatic cholesterol and neutral lipid (triacylglycerol and cholesterol ester) accumulation. Thus studies with gene-ablated mice demonstrate L-FABP influences not only fatty acid metabolism (1, 8, 17, 19) but, as shown herein, support the hypothesis that L-FABP also exhibits a physiological role in cholesterol metabolism.
This study was supported in part by National Institutes of Health Grants DK-41402 and GM-31651.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 the American Physiological Society