Intestinal Lipoprotein Overproduction, a Newly Recognized Component of Insulin Resistance, Is Ameliorated by the Insulin Sensitizer Rosiglit
http://www.100md.com
内分泌学杂志 2005年第1期
Departments of Medicine and Physiology (G.F.L., K.U., L.S.), Division of Endocrinology and Metabolism, and the Department of Laboratory Medicine and Pathobiology (M.N., M.H., K.A.), Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada M5G 2C4
Address all correspondence and requests for reprints to: Gary F. Lewis, Toronto General Hospital, 200 Elizabeth Street, EN11-229, Toronto, Ontario, Canada M5G 2C4. E-mail: gary.lewis@uhn.on.ca.
Abstract
We investigated whether intestinal lipoprotein overproduction in a fructose-fed, insulin-resistant hamster model is prevented with insulin sensitization. Syrian Golden hamsters were fed either chow, 60% fructose for 5 wk, chow for 5 wk with the insulin sensitizer rosiglitazone added for the last 3 wk, or 60% fructose plus rosiglitazone. In vivo Triton studies showed a 2- to 3-fold increase in the large (Svedberg unit > 400) and smaller (Sf 100–400) triglyceride-rich lipoprotein particle apolipoprotein B48 (apoB48) but not triglyceride secretion with fructose feeding in the fasted state (P < 0.01) and partial normalization with rosiglitazone in fructose-fed hamsters. Ex vivo pulse-chase labeling of enterocytes confirmed the oversecretion of apoB48 lipoproteins with fructose feeding. Intestinal lipoprotein oversecretion was associated with increased expression of microsomal triglyceride transfer protein expression. With rosiglitazone treatment of fructose-fed hamsters, there was approximately 50% reduction in apoB48 secretion from primary cultured enterocytes and amelioration of the elevated microsomal triglyceride transfer protein mass and activity in fructose-fed hamsters. In contrast, in the postprandial state, the major differences between nutritional and drug intervention protocols were evident in triglyceride-rich lipoprotein triglyceride and not apoB48 secretion rates. The data suggest that intestinal lipoprotein overproduction can be ameliorated with the insulin sensitizer rosiglitazone.
Introduction
THE HALLMARK DYSLIPIDEMIA of insulin-resistant states and type 2 diabetes consists grossly of hypertriglyceridemia, low high-density lipoprotein cholesterol and small dense, low-density lipoprotein particles (1). Postprandial elevation of triglyceride (TG)-rich lipoproteins (TRLs) is also a well-recognized feature of diabetic dyslipidemia and includes the accumulation of intestinally derived apolipoprotein B48 (apoB48)-containing lipoproteins (2, 3, 4). There is increasing evidence that TRLs, including the apoB48 intestinally derived lipoproteins, may be particularly atherogenic (5, 6, 7, 8). To date, studies in rats and humans have focused on the delayed clearance of TRLs as the dominant underlying mechanism for their postprandial accumulation in diabetes. The impaired removal of intestinally derived lipoproteins has been attributed to decreased activity of lipoprotein lipase (9, 10) and apolipoprotein composition of the carrier particles (11, 12). Despite ample evidence supporting the delayed postprandial clearance of intestinally derived lipoproteins, there is little information in the literature regarding the formation of apoB48-containing lipoproteins in the setting of insulin resistance and type 2 diabetes mellitus.
We have recently established the fructose-fed hamster as a nondiabetic model of whole-body and hepatic insulin resistance and metabolic dyslipidemia (13) and have shown that rosiglitazone treatment of the fructose-fed hamster ameliorates whole-body insulin resistance, improves hepatic insulin signaling, and reduces hepatic lipoprotein overproduction (14). The tissue-specific expression of apoB100 (only in the liver) and apoB48 (only in the intestine) is a distinct advantage of the hamster model over other rodent models in permitting the study of intestinal vs. hepatically derived lipoproteins in vivo (15, 16). ApoB48 is an essential structural component of intestinally derived chylomicrons and chylomicron remnants. Thus, the use of the hamster model makes it possible to investigate the mechanisms of chylomicron and chylomicron remnant metabolism in insulin-resistant states because apoB48-containing lipoproteins are almost exclusively secreted by the intestine.
We recently used the fructose-fed hamster model to investigate the role of intestinal lipoprotein production in the development of metabolic dyslipidemia (17). In this model, we found a 2- to 3-fold elevation in apoB48 particle production rate in vivo compared with chow-fed hamsters. The increased secretion of intestinal apoB48-containing particles was confirmed ex vivo in cultured enterocytes derived from fructose-fed hamsters. Similar to hepatic overproduction of apoB100-containing lipoproteins (13), the ex vivo experiments showed that chronic fructose feeding was associated with greater stability of intracellular apoB48, enhanced intestinal enterocyte de novo lipogenesis, and up-regulation of the key protein involved in intestinal lipoprotein assembly, microsomal TG transfer protein (MTP) (17).
In the present study we used the peroxisome proliferator-activated receptor agonist and insulin sensitizer rosiglitazone in the fructose-fed hamster to determine whether the administration of this insulin-sensitizing agent is associated with amelioration of intestinal lipoprotein overproduction in the fructose-fed hamster. We measured apoB48 production both in vivo and ex vivo in this animal model of insulin resistance and assessed the response to insulin sensitization with rosiglitazone. We also measured the effect of insulin sensitization on the expression of a key protein involved in intestinal lipoprotein assembly, MTP.
Materials and Methods
Animals and study protocols
Male Syrian Golden hamsters (Mesocricetus auratus) were purchased from Charles River (Quebec, Canada). All animals were housed in pairs and were given free access to food and water. Hamsters were fed a normal chow diet for 7 d to allow acclimatization to the new environment and recovery from the stress of shipping. Animals were then placed on one of four feeding protocols: 1) normal chow for 5 wk (CHOW); 2) high-fructose diet (hamster diet with 60% fructose, Dyets Inc., Bethlehem, PA) for 5 wk (FRUC); 3) chow diet for 5 wk with rosiglitazone (20 μmol/kg·d) (GlaxoSmithKline, Philadelphia, PA) diluted in water (100 μl) and given once daily by gavage without anesthetization for the last 3 wk of the chow feeding period (CHOW+RSG); and 4) high-fructose diet for 5 wk with rosiglitazone (20 μmol/kg·d) administered as above for the last 3 wk of the fructose feeding period (FRUC+RSG). Doses of rosiglitazone up to 50 μmol/kg per day have been administered to rodents for 9 months without adverse effects (18). The fat content of the chow diet was 4.0% (soybean oil) and that of the fructose diet was 6% (corn oil). Therefore, although we have called the high-fructose diet FRUC, we draw the reader’s attention to the fact that there were also small differences in type and quantity of the fat content of FRUC vs. CHOW. This is the same diet that was used in previously published studies in which we carefully characterized the insulin resistance and the intestinal and hepatic lipoprotein overproduction induced by FRUC vs. CHOW (13, 14, 17, 19). The hamsters’ weights were monitored every week. At the end of the 5 wk, the animals underwent either the in vivo protocol described below or were killed for isolation of enterocytes for the ex vivo protocols. All animal protocols were approved by the Animal Ethics Committee of the University Health Network.
In vivo determination of intestinal lipoprotein particle production rates
Methods for the in vivo determination of intestinal lipoprotein particle production have been published previously (17). Briefly, 1 d before these studies, femoral venous and arterial catheters were inserted as described previously (13). The animals were fasted overnight for 16 h. A baseline blood sample was drawn followed by an iv bolus of Triton-WR1339 (Sigma Chemical Co., St. Louis, MO). After Triton administration, blood samples were drawn at 10 and 20 min (total blood volume withdrawn for the entire study was 1.2 ml). The Svedberg unit (Sf) > 400 (large TRL) and Sf 100–400 (smaller TRL) fractions were isolated as previously described (17). ApoB48 was quantified using analytical SDS-PAGE as described previously (20), in which there is no contamination by apoB100. Large and small apoB48 and TG secretion rates were derived by multiplying the slope of the concentration increase of apoB48 and TG, respectively, over time by the intravascular distribution volume estimated as 3.8 ml/100 g body weight (14).
Studies performed in the postprandial state were as described above except that they were manually administered 400 μl of lard at time 0 and then every 20 min over 60 min, to achieve a steady fat-fed condition as previously described (17). The bolus of Triton-WR1339 was given 1 h after starting feeding, and blood samples were drawn as described above.
Triton method validation experiments
We performed additional experiments in three animals in which we compared the slope of apoB48 vs. time after Triton administration, calculated from three vs. five time points. We found no significant difference in the slopes calculated from either three or five time points [5.3 ± 0.45 μg/min vs. 5.1 ± 0.58 μg/min, respectively, for Sf > 400, P = not significant (NS); and 5.5 ± 0.85 μg/min vs. 5.3 ± 0.84 μg/min, respectively, for Sf 100–400, P = NS). To keep blood volumes drawn for each experiment to a minimum we have calculated the slope of the apoB48 vs. time curves using three rather than five time points.
Ex vivo protocols
Isolation of primary hamster enterocytes.
The isolation of viable adult villi from hamster small intestine was based on that described by Perreault and Beaulieu (21) with some modification as reported recently (17).
Characteristics of the isolated enterocytes and their functional viability have been documented previously (17). We have also assessed the stability of primary cultured enterocytes over the time course of the experiments reported in the manuscript. Primary enterocytes were radiolabeled for 20 min after 90, 120, and 150 min of incubation. No significant change was observed in the total amount of apoB48 secreted by cultured enterocytes over these time periods (data not shown), although there is loss of the cell’s ability to secrete apoB48 over time after 2.5 h after isolation. The experiments reported in this manuscript, however, always use cells within 2.5 h of isolation, at which time the apoB48 secretion is high and there is greater than 90% cell viability. Thus we have shown that there is no appreciable loss of cell viability or differentiation over the time course of our pulse-chase experiments reported in the manuscript. We have also tested the effect of rosiglitazone in vitro on primary enterocytes and have found no effect on cell viability as assessed by total protein synthesis.
Metabolic labeling of intact primary hamster enterocytes.
Primary hamster enterocytes were used for pulse-chase experiments as described (17). Briefly, cells were preincubated in methionine-free DMEM for 30 min and pulsed with 50 μCi/ml [35S]methionine (equals 1.85 MBq/ml) for 25 min. After the pulse, the cells were washed and chased in methionine-supplemented DMEM. At 90 min, triplicate dishes were harvested, and cells were lysed in solubilization buffer and the lysates were used for immunoprecipitation, as described (22). As a control, we have monitored changes in total protein synthesis by assessing both acid-precipitable radiolabeled protein (trichloroacetic acid counts) as well as total protein mass assays. No specific protein was used as apoB48 is itself a highly specific protein expressed only by intestinal enterocytes.
SDS-PAGE, fluorography, and immunoblotting of primary hamster enterocytes.
Immunoprecipitates were analyzed by SDS-PAGE and fluorography, essentially as described (23). ApoB48 bands, visualized by fluorography, were quantified by excision from the gel, digestion, and scintillation counting essentially as described (22). The 97-kDa subunit of MTP was measured by chemiluminescent immunoblotting as previously described (22), using a goat anti-bovine MTP antiserum provided by Dr. David Gordon (Bristol Myers Squibb, Princeton, NJ). For MTP blots, we measured total protein content and loaded equal amounts of total protein on each lane of the gel (typically 20 μg). Thus, the MTP data shown in Fig. 4 are normalized for protein content. In addition to the measurement of MTP mass in cultured hamster enterocytes, we also measured MTP mass in total intestinal tissue isolated from three hamsters from each treatment group (i.e. without primary culture of hamster enterocytes). MTP protein mass was quantified by densitometry using FluorChem Imaging software (Alpha Innotech Corp., San Leandro, CA). Lipoprotein fractionation was achieved by sequential ultracentrifugation of the culture medium to isolate large chylomicrons (17, 24).
FIG. 4. MTP mass (A and B) and transfer activity (C) in chow-fed hamsters (CHOW), chow-fed hamsters treated with rosiglitazone (CHOW+RSG), fructose-fed hamsters (FRUC), and fructose-fed hamsters treated with rosiglitazone (FRUC+RSG). Data are shown as percentage MTP mass (A and B) and transfer activity (C) compared with CHOW (set at 100%). A, Enterocytes from CHOW (n = 3 hamsters, black bars), FRUC (n = 3, white bars), and FRUC+RSG (n = 3, light gray bars) hamsters (after a 16-h fast) were solubilized, equal amounts of cell protein (20 μg) were subjected to SDS-PAGE, and proteins were transferred onto polyvinylidene difluoride membrane. B, Same analysis as above on enterocytes from CHOW (n = 3 hamsters, black bars) and CHOW+RSG (n = 3, dark gray bars) hamsters. Immunoblotting was performed to detect the 97-kDa MTP subunit with a rabbit anti-hamster MTP antiserum. Shown are representative immunoblots from three experiments expressed as a percentage of the MTP mass detected in chow-fed hamster intestinal cell lysates or intestinal tissue. For cell lysates, there was a significant increase in protein mass of MTP compared with CHOW controls (20% increase in MTP mass; *, P = 0.03 for FRUC vs. CHOW), which was improved with rosiglitazone treatment (30%; , P = 0.008 for FRUC+RSG vs. FRUC). No significant change was observed in MTP mass between CHOW and CHOW+RSG. For MTP transfer activity (C), there was a significant increase in MTP activity in FRUC compared with CHOW (11% increase in MTP activity; *, P < 0.05 for FRUC vs. CHOW), which tended to improve with rosiglitazone treatment (7% decrease; P = NS for FRUC+RSG vs. FRUC). There was no difference in the MTP activity between CHOW and CHOW+RSG.
MTP lipid transfer activity was measured in both primary cultured hamster enterocytes and intestinal tissue by fluorescence spectrophotometry using a kit from Roar Biomedical (kit RB-MTP, New York, NY). Briefly, intestinal pieces from CHOW, CHOW+RSG, FRUC, and FRUC+RSG were homogenized in a Tris buffer, and 100 μg of homogenate or 100 μg of cultured enterocyte lysate were incubated for 6 h at 37 C with donor and acceptor particles provided in the kit. After 6 h, the samples were read in the fluorescence spectrophotometer at an excitation wavelength of 465 nm and emission wavelength of 538 nm.
Other laboratory methods
Glucose was determined on whole blood using a glucometer (Sure Step, One Touch). Plasma insulin concentrations were determined by RIA using a rat insulin kit from Linco Research (St. Louis, MO). This assay has 100% cross-reactivity to hamster insulin, and the intra- and interassay coefficients of variation were 6.8 and 10.6%, respectively. Plasma free fatty acids (FFAs) were measured by a colorimetric method (kit supplied by Wako Industrials, Neuss, Germany) with a coefficient of variation of 4.7% for intra- and 8.2% for interassay variation. TG was measured using a colorimetric assay (Roche Mannheim GmbH Diagnostica, Laval, Quebec, Canada) with an intra- and interassay coefficient of variation of 3.0 and 4.9%, respectively.
Statistical analysis
All the values are reported as mean ± SEM. For comparison of TRL TG and apoB48 secretion rates between CHOW, CHOW+RSG, FRUC, and FRUC+RSG hamsters, ANOVA was used followed by post hoc analysis with Tukey’s test. A P value < 0.05 was considered to be significant.
Results
Effect of rosiglitazone treatment on body weight, plasma insulin, FFA, TG, and glucose (Table 1)
Because of blood volume limitations, fasting blood samples for the above analyses were not taken from animals undergoing in vivo protocols and therefore are not reported for the actual animals that underwent the in vivo studies reported in this manuscript. Data in Table 1 are cumulative data from our laboratory from animals that had undergone identical feeding and rosiglitazone treatment protocols, but these animals did not undergo in vivo studies for assessment of lipoprotein production. There was no significant difference in any of the variables between groups. There was a strong trend for FRUC to have higher TGs, insulin, and glucose, but perhaps because of high inter-animal variability, the differences were not significant.
TABLE 1. Characteristics of fructose-fed (FRUC), fructose fed plus rosiglitazone-treated (FRUC + RSG), chow-fed (CHOW), and chow-fed plus rosiglitazone-treated (CHOW + RSG) hamsters
Relative proportions of apoB48 and -B100 in the TRL fractions
As one might expect, the majority of apoB in the Sf > 100 fraction was apoB48, with very little apoB100 detected. In the Sf 60–100 fraction, 45% of the apoB was apoB48 in the CHOW group, 56% was apoB48 in the FRUC group, and 60% was apoB48 in the FRUC+RSG group. In the Sf 20–60 fraction, 5% of the apoB was apoB48 in the CHOW group, 24% in the FRUC group, and 13% in the FRUC+RSG group. This is in contrast to the proportion of apoB48 to apoB100 in these smaller TRL fractions in healthy humans previously reported from our laboratory (25). In the fasted human Sf 60–400 TRL fraction, apoB48 constituted only 11.5% of total apoB, and in the Sf 20–60 fraction, it constituted 7% of total apoB. In another published human study (26), only 5% of the Sf 60–400 TRL fraction apoB was apoB48, whereas only 6.7% of the Sf 20–60 fraction was apoB48. These data indicate that the CHOW-fed hamster Sf 60–100 fraction has a relatively far greater proportion of apoB48 to -B100 than does the human, whereas the Sf 20–60 fraction has a similar proportion of apoB48 to -B100 in hamsters and humans.
Triton WR-1339 studies for determination of in vivo intestinal lipoprotein production in the fasting state (Fig. 1)
There was an approximately 2.5-fold greater apoB48 secretion rate with chronic fructose feeding compared with chow in large TRL (Sf > 400 fraction; P < 0.01) and an approximately 2-fold higher apoB48 secretion rate in smaller TRL (Sf 100–400 fraction; P < 0.05) (Fig. 1, B and D). Treatment of fructose-fed hamsters with rosiglitazone resulted in an approximately 50% lower apoB48 secretion rate than fructose alone in both TRL fractions (for Sf > 400, P < 0.05 vs. FRUC and P = NS vs. other treatment groups; for the Sf 100–400 fraction, P < 0.05 vs. FRUC and P = NS vs. other treatment groups), but treatment of CHOW-fed hamsters with rosiglitazone resulted in no significant change in apoB48 secretion rate in either the large or smaller TRL fraction. There was no significant difference in TG secretion rates between any of the groups in either large or small TRL fractions (Fig. 1, A and C).
FIG. 1. In vivo production of Sf > 400 (large) (A and B) and Sf 100–400 (smaller) (C and D) TRL particle TG (A and C) and apoB48 (B and D) in the fasted state in control hamsters (CHOW), control hamsters treated with rosiglitazone (CHOW+RSG), fructose-fed hamsters (FRUC), and fructose-fed hamsters treated with rosiglitazone (FRUC+RSG). A, Sf > 400 TG secretion rate (in μmol/min) after iv administration of Triton WR-1339 in CHOW (black bars, n = 13), CHOW+RSG (dark gray bars, n = 6), FRUC (white bars, n = 19), and FRUC+RSG hamsters (gray bars, n = 18). TG secretion rates between the four groups were not statistically different. B, Sf > 400 apoB48 secretion rate was significantly higher in FRUC compared with CHOW (*, P < 0.01) and was reduced with rosiglitazone (+, P < 0.05 for FRUC+RSG vs. FRUC). C, Sf 100–400 TG secretion rates were not statistically different between the four groups. D, Sf 100–400 apoB48 secretion rate was higher in FRUC compared with CHOW (+, P < 0.01) and was significantly reduced with rosiglitazone (+, P < 0.05 for FRUC+RSG vs. FRUC).
Because we observed an increase in apoB48 but not TG production with FRUC vs. CHOW and intermediate apoB48 production rates with FRUC+RSG, we assessed the TRL TG/apoB48 ratio in the fasted state as a crude index of particle size determination. For the Sf > 400 TRL fraction, the ratio of TRL TG/apoB48 was lower in FRUC than CHOW (FRUC, 0.005 ± 0.001, vs. CHOW, 0.009 ± 0.002; P < 0.05), and the ratio was intermediate in FRUC+RSG (0.007 ± 0.001; P = NS compared with the other groups), indicating a tendency for particle size to shift from larger to smaller size with fructose feeding. The Sf 100–400 fraction tended to show a shift toward a smaller particle size in FRUC (CHOW, 0.012 ± 0.001; FRUC, 0.009 ± 0.001; and FRUC+RSG, 0.010 ± 0.001; P = NS). There was no significant difference in the TG/apoB48 between CHOW+RSG and the other groups (0.007 ± 001 for Sf > 400 and 0.010 ± 001 for Sf 100–400).
Triton WR-1339 studies for determination of in vivo intestinal lipoprotein production in the postprandial state (Fig. 2)
TG and apoB48 secretion rates were measured in the large (Sf > 400 TRL) and smaller (Sf 100–400) fractions in CHOW, CHOW+RSG, FRUC, and FRUC+RSG hamsters after 60 min of fat feeding. In the Sf > 400 fraction, TG secretion rate tended to be higher in FRUC compared with CHOW and was significantly decreased with rosiglitazone treatment (P = 0.08 for CHOW vs. FRUC; P < 0.05 for FRUC+RSG vs. FRUC and CHOW) (Fig. 2A). There was no significant change in TG production in the Sf > 400 fraction when CHOW-fed hamsters were treated with RSG. In the Sf 100–400 fraction, TG secretion was increased approximately 2-fold in FRUC and normalized with rosiglitazone (P < 0.05 for FRUC vs. CHOW; P < 0.01 for FRUC+RSG vs. FRUC) (Fig. 2C). There was no significant difference in Sf 100–400 TG secretion in the CHOW+RSG.
FIG. 2. In vivo production of Sf > 400 (large) (A and B) and Sf 100–400 (smaller) (C and D) TRL particle TG (A and C) and apoB48 (B and D) in the fed (postprandial) state in control hamsters (CHOW), control hamsters treated with rosiglitazone (CHOW+RSG), fructose-fed hamsters (FRUC), and fructose-fed hamsters treated with rosiglitazone (FRUC+RSG). TG and apoB48 secretion rates were measured in the Sf > 400 fraction (large TRL) and Sf 100–400 fraction (smaller TRL) in CHOW (black bars, n = 11), CHOW+RSG (dark gray bars, n = 6), FRUC (white bars, n = 11), and FRUC+RSG hamsters (gray bars, n = 8), A, In the Sf > 400 fraction, TG secretion rate tended to be increased in the FRUC hamsters compared with CHOW (P = 0.08) and was significantly decreased with rosiglitazone treatment (+, P < 0.05 for FRUC+RSG vs. FRUC). B, There were no differences in apoB48 secretion in the Sf > 400 fraction between the four groups of animals. C, In the Sf 100–400 fraction, TG secretion was increased approximately 2-fold in FRUC and normalized with rosiglitazone (+, P < 0.05 for FRUC vs. CHOW; *, P < 0.01 for FRUC+RSG vs. FRUC). D, ApoB48 secretion tended to increase in FRUC hamsters (P = NS for all).
ApoB48 secretion was measured in both large and small TRL fractions in the postprandial state. In the Sf > 400 fraction, there were no differences in apoB48 secretion (P = NS for all) (Fig. 2B). In the Sf 100–400 fraction, apoB48 secretion tended to increase (by 2-fold) in FRUC (P = NS for all) (Fig. 2D). There was no difference in apoB48 production rate in CHOW+RSG compared with the other treatment groups.
TG/apoB48 was assessed in the postprandial state. In this case, the TG secretion rate was increased in FRUC, whereas the apoB48 secretion rate only tended to be increased. For the Sf > 400 fraction, the TG/apoB48 was increased in FRUC vs. CHOW and decreased in FRUC+RSG vs. FRUC (0.005 ± 001 for CHOW, 0.005 ± 001 for CHOW+RSG, 0.009 ± 002 for FRUC, and 0.003 ± 001 for FRUC+RSG; P < 0.05 for FRUC vs. CHOW; P < 0.01 for FRUC vs. FRUC+RSG). In the Sf 100–400 fraction, the TG/apoB48 was also increased in the FRUC group and decreased to normal in the FRUC+RSG group (0.005 ± 001for CHOW, 0.005 ± 001 for CHOW+RSG, 0.008 ± 001 for FRUC, and 0.005 ± 001 for FRUC+RSG; P < 0.001 for FRUC vs. CHOW; P < 0.01 for FRUC vs. FRUC+RSG; and P < 0.01 FRUC vs. CHOW+RSG).
Studies of apoB48 secretion in primary cultured enterocytes (Fig. 3)
Pulse-chase-labeling experiments were used to assess the secretion of apoB48 in villus enterocytes isolated from CHOW, FRUC, and FRUC+RSG hamsters. Figure 3, A and B, shows the extracellular secretion of chylomicron-specific and total apoB48 in enterocytes isolated from CHOW, FRUC, and FRUC+RSG hamsters. Enterocytes from FRUC hamsters secreted approximately 4-fold more newly synthesized chylomicron-specific and total apoB48 over a 90-min chase compared with that in CHOW animals (P = 0.01 for chylomicron-specific and total apoB48 in FRUC vs. FRUC+RSG). In fructose-fed animals, there was a 2-fold reduction in chylomicron and total apoB48 secretion with rosiglitazone treatment (P = 0.02 for chylomicron apoB48, and P = 0.01 for total apoB48 in FRUC vs. FRUC+RSG). Total apoB48 secretion was also assessed in a separate group of hamsters fed either CHOW or CHOW+RSG (Fig. 3C). No significant change in apoB48 was observed in CHOW hamsters treated with RSG. It should be noted that, in Fig. 3, apoB48 synthesis and secretion were normalized for both total cellular protein mass and the incorporation of [35S]methionine into total trichloroacetic acid-insoluble cellular and secreted proteins. Therefore, the stimulation of apoB48 synthesis and secretion was not a consequence of global effects of fructose-feeding on protein synthesis and secretion.
FIG. 3. Ex vivo secretion of apoB48 by primary cultured hamster enterocytes in chow-fed hamsters (CHOW), chow-fed hamsters treated with rosiglitazone (CHOW+RSG), fructose-fed hamsters (FRUC), and fructose-fed hamsters treated with rosiglitazone (FRUC+RSG). Primary enterocytes, isolated from the treated hamsters, were pulsed with [35S]methionine and chased for 0, 45, and 90 min. The media samples and cell lysates collected at each chase time point were subjected to immunoprecipitation and then analyzed by SDS-PAGE and fluorography. A and B, Chylomicron and total apoB48 secreted in CHOW, FRUC, and FRUC+RSG, respectively. C, Total apoB48 secreted in CHOW and CHOW+RSG. There was a 4-fold increase in chylomicron and total apoB48 secretion with fructose-feeding, which was reduced 2-fold with rosiglitazone treatment (A: *, P = 0.01 for FRUC vs. CHOW; , P = 0.02 for FRUC+RSG vs. FRUC; B: **, P = 0.01 for FRUC vs. CHOW; , P = 0.01 for FRUC+RSG vs. FRUC). No significant difference was observed in apoB48 secretion in CHOW vs. CHOW+RSG.
MTP mass (Fig 4, A and B) and transfer activity (Fig. 4C) in primary cultured enterocytes and MTP mass and transfer activity measured in intestinal tissue isolated from treated hamsters (not illustrated)
Facilitated secretion of apoB48 and core lipoprotein lipids in FRUC hamster enterocytes could be related to an increase in MTP mass, the key factor involved in the lipoprotein assembly process. To test this hypothesis, equal quantities of primary cultured intestinal cell lysate (20 μg) were analyzed by immunoblotting. As shown in Fig. 4A, in FRUC hamster enterocyte cell lysates, there was a significant increase in protein mass of MTP compared with CHOW controls (20% increase in MTP mass; P = 0.03), which was ameliorated by rosiglitazone treatment (30%; P = 0.008). Interestingly, no significant change was observed in MTP mass between CHOW and CHOW+RSG (Fig. 4B). The MTP transfer activity measured in enterocyte cell lysates mirrored the MTP protein mass data (Fig. 4C), with a significant increase in MTP transfer activity in FRUC vs. CHOW (110.5 ± 2.57 vs. 100.0 ± 2.58%; P < 0.05) and partial amelioration of this FRUC-induced increase with the addition of rosiglitazone treatment (103.2 ± 3.63%; P = NS for FRUC+RSG vs. FRUC). No difference was observed between CHOW+RSG (101.0 ± 1.65%) and CHOW.
MTP mass and transfer activity were also measured directly in intestinal tissue (30 μg total protein per lane), but unlike the cell lysates, there was no significant difference in tissue MTP mass between the treatment groups, although there was a tendency for the mass to be higher in FRUC and FRUC+RSG vs. CHOW (MTP mass in CHOW = 100.0 ± 3.0%, FRUC = 110.4 ± 4.3%, and FRUC+RSG =118.7 ± 3.3%; P = NS between groups) (not illustrated). MTP activity measured directly in intestinal tissue was significantly increased by 13% in FRUC vs. CHOW (MTP activity in CHOW = 100 ± 2.94% and FRUC = 112.9 ± 4.83%; P < 0.01). Rosiglitazone did not, however, ameliorate this increase in MTP activity (FRUC+RSG=115.9 ± 4.72%, p=ns vs. FRUC, P < 0.001 vs. CHOW). CHOW+RSG (103.1 ± 1.2%) was not different from CHOW.
Discussion
We have demonstrated previously that treatment of fructose-fed, nondiabetic, insulin-resistant hamsters with rosiglitazone, a member of the thiazolidinedione class of insulin sensitizers with specific peroxisome proliferator-activated receptor- agonist activity, improved whole-body and hepatic insulin sensitivity and also decreased the overproduction of very-low-denisty lipoprotein apoB by hepatocytes ex vivo (14). In the present study, we have also shown that rosiglitazone treatment reduced the secretion of intestinally derived apoB48-containing lipoproteins in the fasting state and TRL TG secretion in the postprandial state in fructose-fed but not in chow-fed hamsters. The secretion of intestinal apoB48-containing particles was examined both in vivo and ex vivo, and similar results were obtained. Furthermore, rosiglitazone treatment was associated with a reversal of the increased intestinal expression of MTP seen with fructose feeding when measured in intestinal cell lysates, indicating one potential molecular mechanism by which rosiglitazone led to reduction of intestinal particle secretion in this insulin-resistant animal model. Whether this improvement in intestinal lipoprotein particle secretion resulted from the insulin-sensitizing effects of rosiglitazone or from another unrelated action of the drug will require additional study perhaps using relevant knockout and transgenic animal models.
In the present study, apoB48 secretion in the fasting state was increased in large (Sf > 400) and smaller (Sf 100–400) TRL fractions, whereas there was no significant difference in TG secretion. The absence of a significant difference in TG secretion between control and insulin-resistant fructose-fed animals in the fasted state suggests that there is increased production of small, lipid-poor, apoB48-containing lipoproteins in these size ranges in the fasting state with fructose feeding. Fructose feeding was indeed associated with the production of smaller particles in the larger TRL Sf > 400 fraction, as evidenced by a reduction in the TG/apoB48 ratio, and a tendency toward smaller particle production in the smaller Sf 100–400 fraction. We speculate that the intestine constitutively secretes small apoB48-containing lipoprotein particles (and a greater number in insulin-resistant states) to be primed and ready for the ingestion of fat. In the fed (postprandial) state, fructose feeding and rosiglitazone treatment manifested in an increase and correction, respectively, of TRL TG, a feature that was not evident in the fasted state, in which predominantly lipid-poor TRLs are secreted. Both chronic fructose feeding and acute fat ingestion also stimulate hepatic very-low-density lipoprotein secretion, as we have shown previously for the former in the fructose-fed hamster (13) and others have shown in humans for the latter after a high-fat meal (27).
The relative proportion of apoB48 as a fraction of total apoB in the Sf 60–100 but not the Sf 20–60 fraction in fasted, chow-fed hamsters was greater than that found in humans and increased proportionately with fructose feeding. Unlike most other rodents, in the hamster liver there is negligible editing of apoB, and we have never been able to detect apoB48 secreted by cultured hamster hepatocytes. The intestine of the hamster, therefore, produces far more TRL apoB48 relative to hepatic apoB100 in this TRL fraction than does the human. Perhaps the hamster, because it nibbles rather than gorges like the human and stores food in its cheek, has evolved to secrete relatively more apoB48 from its intestine relative to hepatic apoB100, compared with the human, to accommodate the transport of fairly persistent food ingestion. These differences between hamster and human caution us not to generalize the present findings in the hamster model to other species such as humans. Future studies will need to examine this phenomenon directly in other species, including the human.
Our ex vivo studies of enterocyte cell lysates showed that treatment with rosiglitazone was associated with a reversal of the increased expression of MTP seen with fructose feeding. Measurement of MTP mass and lipid transfer activity directly in intestinal tissue, however, showed less consistent changes, possibly because of the presence of other cell types masking enterocyte-specific changes in MTP mass and transfer activity. Although these changes in MTP expression are small, there is ample evidence from the literature to support the finding that relatively small changes in MTP levels can result in significant changes in lipoprotein production rates (28, 29, 30, 31). The reduction in MTP levels with rosiglitazone treatment may have been implicated in the reduction of intestinal lipoprotein secretion in the present study. MTP is an important factor in intestinal lipoprotein assembly. In Caco-2 cells, MTP has been shown not only to be involved in the first step of lipoprotein synthesis, i.e. the rescue of apoB from intracellular degradation through early lipidation of the protein, but also to be involved in further steps involving association of lipoprotein particles with TG droplets (32). The promoter region of the MTP gene contains a negative insulin-response element (33), and intestinal MTP mRNA has been shown to be raised in diabetic and insulin-resistant rats (34, 35). It is possible, therefore, that the reduction in MTP levels induced by rosiglitazone treatment was the result of improved insulin signaling at the level of the enterocyte. The precise molecular signaling pathway involved in insulin-mediated modulation of MTP expression is currently unclear. Additional studies are needed to elucidate the role of insulin in modulating MTP expression and in regulating intestinal particle secretion.
Intestinal lipoprotein production has been felt to be regulated predominantly by the amount of fat ingested because of the efficiency of fat absorption and rapid turnover of enterocytes. The mucosal surface of the gastrointestinal tract is remarkable for the very rapid turnover of the epithelial cell population. Enterocytes of the small intestine are replaced rapidly by the division of stem cells at the base of the crypts of Lieberkuhn and movement of cells up the crypts to the villi as they differentiate into mature enterocytes. It is likely that the population of absorptive cells is replaced after 24–72 h. The rapid turnover of intestinal cells, therefore, would seem to preclude a chronic state such as insulin resistance from affecting the intestine’s capacity to produce lipoprotein particles. We have previously performed short-term (2 d) fructose feeding studies in hamsters to investigate the effect on apoB48 lipoprotein formation and found no significant effect on the secreted level of intestinal chylomicron-apoB48, showing that the effect of fructose feeding on intestinal lipoprotein production is a chronic process (17). It has been suggested that consumption of high-fat diets during maturation causes adaptation of the enterocytes so that they have an increased capacity to absorb lipid and secrete chylomicrons (36, 37). Such intestinal adaptation may also underlie the effect of an insulin-resistant state on the intestine’s capacity to produce lipoprotein particles.
In conclusion, we have shown that whole-body insulin sensitization with rosiglitazone treatment is associated with a reduction in intestinal MTP overexpression and apoB48-containing particle hypersecretion in the fructose-fed insulin-resistant hamster. Although we cannot be sure that the beneficial effect of rosiglitazone was mediated by insulin sensitization, our findings suggest that therapeutic measures that effectively ameliorate insulin resistance, or that reduce MTP overexpression in insulin-resistant states, could be part of the strategy to correct the intestinal particle oversecretion associated with insulin resistance. This could have significant implications in the management of the dyslipidemia and possibly atherosclerosis of type 2 diabetes mellitus and insulin resistance. Additional studies examining the link between insulin resistance and intestinal lipoprotein overproduction need to be conducted in humans to determine whether this process is relevant to human disease states.
References
Lewis GF, Steiner G 1996 Hypertriglyceridemia and its metabolic consequences as a risk factor for atherosclerotic cardiovascular disease in non-insulin-dependent diabetes mellitus. Diabetes Metab Rev 12:37–56
Lewis GF, O’Meara NM, Soltys PA, Blackman JD, Iverius PH, Pugh WL, Getz GS, Polonsky KS 1991 Fasting hypertriglyceridemia in non-insulin-dependent diabetes mellitus is an important predictor of postprandial lipid and lipoprotein abnormalities. J Clin Endocrinol Metab 72:934–944
Curtin A, Deegan P, Owens D, Collins P, Johnson A, Tomkin GH 1996 Elevated triglyceride-rich lipoproteins in diabetes: a study of apolipoprotein B-48. Acta Diabetol 33:205–210
Mero N, Syvanne M, Taskinen MR 1998 Postprandial lipid metabolism in diabetes. Atherosclerosis 141(Suppl 1):S53–S55
Doi H, Kugiyama K, Ohgushi M, Sugiyama S, Matsumura T, Ohta Y, Nakano T, Nakajima K, Yasue H 1998 Remnants of chylomicron and very low density lipoprotein impair endothelium-dependent vasorelaxation. Atherosclerosis 137:341–349
Mero N, Malmstrom R, Steiner G, Taskinen MR, Syvanne M 2000 Postprandial metabolism of apolipoprotein B-48- and B-100-containing particles in type 2 diabetes mellitus: relations to angiographically verified severity of coronary artery disease. Atherosclerosis 150:167–177
McNamara JR, Shah PK, Nakajima K, Cupples LA, Wilson PW, Ordovas JM, Schaefer EJ 2001 Remnant-like particle (RLP) cholesterol is an independent cardiovascular disease risk factor in women: results from the Framingham Heart Study. Atherosclerosis 154:229–236
Karpe F, Boquist S, Tang R, Bond GM, de Faire U, Hamsten A 2001 Remnant lipoproteins are related to intima-media thickness of the carotid artery independently of LDL cholesterol and plasma triglycerides. J Lipid Res 42:17–21
Haffner SM, Foster DM, Kushwaha RS, Hazzard WR 1984 Retarded chylomicron apolipoprotein-B catabolism in type 2 (non-insulin-dependent) diabetic subjects with lipaemia. Diabetologia 26:349–354
Mamo JC, Hirano T, Sainsbury A, Fitzgerald AK, Redgrave TG 1992 Hypertriglyceridemia is exacerbated by slow lipolysis of triacylglycerol-rich lipoproteins in fed but not fasted streptozotocin diabetic rats. Biochim Biophys Acta 1128:132–138
Bar-On H, Levy E, Oschry Y, Ziv E, Shafrir E 1984 Removal defect of very-low-density lipoproteins from diabetic rats. Biochim Biophys Acta 793:115–118
Steinbrecher UP, Witztum JL 1984 Glucosylation of low-density lipoproteins to an extent comparable to that seen in diabetes slows their catabolism. Diabetes 33:130–134
Taghibiglou C, Carpentier A, Van Iderstine SC, Chen B, Rudy D, Aiton A, Lewis GF, Adeli K 2000 Mechanisms of hepatic very low density lipoprotein overproduction in insulin resistance. Evidence for enhanced lipoprotein assembly, reduced intracellular ApoB degradation, and increased microsomal triglyceride transfer protein in a fructose-fed hamster model. J Biol Chem 275:8416–8425
Carpentier A, Taghibiglou C, Leung N, Szeto L, Van Iderstine SC, Uffelman KD, Buckingham R, Adeli K, Lewis GF 2002 Ameliorated hepatic insulin resistance is associated with normalization of microsomal triglyceride transfer protein expression and reduction in very low density lipoprotein assembly and secretion in the fructose-fed hamster. J Biol Chem 277:28795–28802
Arbeeny CM, Meyers DS, Bergquist KE, Gregg RE 1992 Inhibition of fatty acid synthesis decreases very low density lipoprotein secretion in the hamster. J Lipid Res 33:843–851
Liu GL, Fan LM, Redinger RN 1991 The association of hepatic apoprotein and lipid metabolism in hamsters and rats. Comp Biochem Physiol A 99:223–228
Haidari M, Leung N, Mahbub F, Uffelman KD, Kohen-Avramoglu R, Lewis GF, Adeli K 2002 Fasting and postprandial overproduction of intestinally derived lipoproteins in an animal model of insulin resistance. Evidence that chronic fructose feeding in the hamster is accompanied by enhanced intestinal de novo lipogenesis and ApoB48-containing lipoprotein overproduction. J Biol Chem 277:31646–31655
Buckingham RE, Al Barazanji KA, Toseland CD, Slaughter M, Connor SC, West A, Bond B, Turner NC, Clapham JC 1998 Peroxisome proliferator-activated receptor- agonist, rosiglitazone, protects against nephropathy and pancreatic islet abnormalities in Zucker fatty rats. Diabetes 47:1326–1334
Taghibiglou C, Rashid-Kolvear F, Van Iderstine SC, Le Tien H, Fantus IG, Lewis GF, Adeli K 2002 Hepatic very low density lipoprotein-ApoB overproduction is associated with attenuated hepatic insulin signaling and overexpression of protein-tyrosine phosphatase 1B in a fructose-fed hamster model of insulin resistance. J Biol Chem 277:793–803
Karpe F, Hamsten A, Uffelman K, Steiner G 1996 Apolipoprotein B-48. Methods Enzymol 263:95–104
Perreault N, Beaulieu JF 1998 Primary cultures of fully differentiated and pure human intestinal epithelial cells. Exp Cell Res 245:34–42
Taghibiglou C, Rudy D, Van Iderstine SC, Aiton A, Cavallo D, Cheung R, Adeli K 2000 Intracellular mechanisms regulating apoB-containing lipoprotein assembly and secretion in primary hamster hepatocytes. J Lipid Res 41:499–513
Laemmli UK 1970 Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680–685
Luchoomun J, Zhou Z, Bakillah A, Jamil H, Hussain MM 1997 Assembly and secretion of VLDL in nondifferentiated Caco-2 cells stably transfected with human recombinant ApoB48 cDNA. Arterioscler Thromb Vasc Biol 17:2955–2963
Lemieux S, Fontani R, Uffelman KD, Lewis GF, Steiner G 1998 Apolipoprotein B-48 and retinyl palmitate are not equivalent markers of postprandial intestinal lipoproteins. J Lipid Res 39:1964–1971
Karpe F, Steiner G, Olivecrona T, Carlson LA, Hamsten A 1993 Metabolism of triglyceride-rich lipoproteins during alimentary lipemia. J Clin Invest 91:748–758
Cohn JS, Wagner DA, Cohn SD, Millar JS, Schaefer EJ 1990 Measurement of very low density and low density lipoprotein apolipoprotein (Apo) B-100 and high density lipoprotein Apo A-I production in human subjects using deuterated leucine. Effect of fasting and feeding. J Clin Invest 85:804–811
Jamil H, Gordon DA, Eustice DC, Brooks CM, Dickson Jr JK, Chen Y, Ricci B, Chu CH, Harrity TW, Ciosek Jr CP, Biller SA, Gregg RE, Wetterau JR 1996 An inhibitor of the microsomal triglyceride transfer protein inhibits apoB secretion from HepG2 cells. Proc Natl Acad Sci USA 93:11991–11995
Jamil H, Chu CH, Dickson Jr JK, Chen Y, Yan M, Biller SA, Gregg RE, Wetterau JR, Gordon DA 1998 Evidence that microsomal triglyceride transfer protein is limiting in the production of apolipoprotein B-containing lipoproteins in hepatic cells. J Lipid Res 39:1448–1454
Ledmyr H, Karpe F, Lundahl B, McKinnon M, Skoglund-Andersson C, Ehrenborg E 2002 Variants of the microsomal triglyceride transfer protein gene are associated with plasma cholesterol levels and body mass index. J Lipid Res 43:51–58
Liao W, Kobayashi K, Chan L 1999 Adenovirus-mediated overexpression of microsomal triglyceride transfer protein (MTP): mechanistic studies on the role of MTP in apolipoprotein B-100 biogenesis. Biochemistry 38:10215
van Greevenbroek MM, Robertus-Teunissen MG, Erkelens DW, de Bruin TW 1998 Participation of the microsomal triglyceride transfer protein in lipoprotein assembly in Caco-2 cells: interaction with saturated and unsaturated dietary fatty acids. J Lipid Res 39:173–185
Hagan DL, Kienzle B, Jamil H, Hariharan N 1994 Transcriptional regulation of human and hamster microsomal triglyceride transfer protein genes. Cell type-specific expression and response to metabolic regulators. J Biol Chem 269:28737–28744
Gleeson A, Anderton K, Owens D, Bennett A, Collins P, Johnson A, White D, Tomkin GH 1999 The role of microsomal triglyceride transfer protein and dietary cholesterol in chylomicron production in diabetes. Diabetologia 42:944–948
Phillips C, Owens D, Collins P, Tomkin GH 2002 Microsomal triglyceride transfer protein: does insulin resistance play a role in the regulation of chylomicron assembly? Atherosclerosis 160:355–360
Thomson AB, Cheeseman CI, Keelan M, Fedorak R, Clandinin MT 1994 Crypt cell production rate, enterocyte turnover time and appearance of transport along the jejunal villus of the rat. Biochim Biophys Acta 1191:197–204
Thomson AB, Keelan M, Wild GE 1996 Nutrients and intestinal adaptation. Clin Invest Med 19:331–345(Gary F. Lewis, Kristine U)
Address all correspondence and requests for reprints to: Gary F. Lewis, Toronto General Hospital, 200 Elizabeth Street, EN11-229, Toronto, Ontario, Canada M5G 2C4. E-mail: gary.lewis@uhn.on.ca.
Abstract
We investigated whether intestinal lipoprotein overproduction in a fructose-fed, insulin-resistant hamster model is prevented with insulin sensitization. Syrian Golden hamsters were fed either chow, 60% fructose for 5 wk, chow for 5 wk with the insulin sensitizer rosiglitazone added for the last 3 wk, or 60% fructose plus rosiglitazone. In vivo Triton studies showed a 2- to 3-fold increase in the large (Svedberg unit > 400) and smaller (Sf 100–400) triglyceride-rich lipoprotein particle apolipoprotein B48 (apoB48) but not triglyceride secretion with fructose feeding in the fasted state (P < 0.01) and partial normalization with rosiglitazone in fructose-fed hamsters. Ex vivo pulse-chase labeling of enterocytes confirmed the oversecretion of apoB48 lipoproteins with fructose feeding. Intestinal lipoprotein oversecretion was associated with increased expression of microsomal triglyceride transfer protein expression. With rosiglitazone treatment of fructose-fed hamsters, there was approximately 50% reduction in apoB48 secretion from primary cultured enterocytes and amelioration of the elevated microsomal triglyceride transfer protein mass and activity in fructose-fed hamsters. In contrast, in the postprandial state, the major differences between nutritional and drug intervention protocols were evident in triglyceride-rich lipoprotein triglyceride and not apoB48 secretion rates. The data suggest that intestinal lipoprotein overproduction can be ameliorated with the insulin sensitizer rosiglitazone.
Introduction
THE HALLMARK DYSLIPIDEMIA of insulin-resistant states and type 2 diabetes consists grossly of hypertriglyceridemia, low high-density lipoprotein cholesterol and small dense, low-density lipoprotein particles (1). Postprandial elevation of triglyceride (TG)-rich lipoproteins (TRLs) is also a well-recognized feature of diabetic dyslipidemia and includes the accumulation of intestinally derived apolipoprotein B48 (apoB48)-containing lipoproteins (2, 3, 4). There is increasing evidence that TRLs, including the apoB48 intestinally derived lipoproteins, may be particularly atherogenic (5, 6, 7, 8). To date, studies in rats and humans have focused on the delayed clearance of TRLs as the dominant underlying mechanism for their postprandial accumulation in diabetes. The impaired removal of intestinally derived lipoproteins has been attributed to decreased activity of lipoprotein lipase (9, 10) and apolipoprotein composition of the carrier particles (11, 12). Despite ample evidence supporting the delayed postprandial clearance of intestinally derived lipoproteins, there is little information in the literature regarding the formation of apoB48-containing lipoproteins in the setting of insulin resistance and type 2 diabetes mellitus.
We have recently established the fructose-fed hamster as a nondiabetic model of whole-body and hepatic insulin resistance and metabolic dyslipidemia (13) and have shown that rosiglitazone treatment of the fructose-fed hamster ameliorates whole-body insulin resistance, improves hepatic insulin signaling, and reduces hepatic lipoprotein overproduction (14). The tissue-specific expression of apoB100 (only in the liver) and apoB48 (only in the intestine) is a distinct advantage of the hamster model over other rodent models in permitting the study of intestinal vs. hepatically derived lipoproteins in vivo (15, 16). ApoB48 is an essential structural component of intestinally derived chylomicrons and chylomicron remnants. Thus, the use of the hamster model makes it possible to investigate the mechanisms of chylomicron and chylomicron remnant metabolism in insulin-resistant states because apoB48-containing lipoproteins are almost exclusively secreted by the intestine.
We recently used the fructose-fed hamster model to investigate the role of intestinal lipoprotein production in the development of metabolic dyslipidemia (17). In this model, we found a 2- to 3-fold elevation in apoB48 particle production rate in vivo compared with chow-fed hamsters. The increased secretion of intestinal apoB48-containing particles was confirmed ex vivo in cultured enterocytes derived from fructose-fed hamsters. Similar to hepatic overproduction of apoB100-containing lipoproteins (13), the ex vivo experiments showed that chronic fructose feeding was associated with greater stability of intracellular apoB48, enhanced intestinal enterocyte de novo lipogenesis, and up-regulation of the key protein involved in intestinal lipoprotein assembly, microsomal TG transfer protein (MTP) (17).
In the present study we used the peroxisome proliferator-activated receptor agonist and insulin sensitizer rosiglitazone in the fructose-fed hamster to determine whether the administration of this insulin-sensitizing agent is associated with amelioration of intestinal lipoprotein overproduction in the fructose-fed hamster. We measured apoB48 production both in vivo and ex vivo in this animal model of insulin resistance and assessed the response to insulin sensitization with rosiglitazone. We also measured the effect of insulin sensitization on the expression of a key protein involved in intestinal lipoprotein assembly, MTP.
Materials and Methods
Animals and study protocols
Male Syrian Golden hamsters (Mesocricetus auratus) were purchased from Charles River (Quebec, Canada). All animals were housed in pairs and were given free access to food and water. Hamsters were fed a normal chow diet for 7 d to allow acclimatization to the new environment and recovery from the stress of shipping. Animals were then placed on one of four feeding protocols: 1) normal chow for 5 wk (CHOW); 2) high-fructose diet (hamster diet with 60% fructose, Dyets Inc., Bethlehem, PA) for 5 wk (FRUC); 3) chow diet for 5 wk with rosiglitazone (20 μmol/kg·d) (GlaxoSmithKline, Philadelphia, PA) diluted in water (100 μl) and given once daily by gavage without anesthetization for the last 3 wk of the chow feeding period (CHOW+RSG); and 4) high-fructose diet for 5 wk with rosiglitazone (20 μmol/kg·d) administered as above for the last 3 wk of the fructose feeding period (FRUC+RSG). Doses of rosiglitazone up to 50 μmol/kg per day have been administered to rodents for 9 months without adverse effects (18). The fat content of the chow diet was 4.0% (soybean oil) and that of the fructose diet was 6% (corn oil). Therefore, although we have called the high-fructose diet FRUC, we draw the reader’s attention to the fact that there were also small differences in type and quantity of the fat content of FRUC vs. CHOW. This is the same diet that was used in previously published studies in which we carefully characterized the insulin resistance and the intestinal and hepatic lipoprotein overproduction induced by FRUC vs. CHOW (13, 14, 17, 19). The hamsters’ weights were monitored every week. At the end of the 5 wk, the animals underwent either the in vivo protocol described below or were killed for isolation of enterocytes for the ex vivo protocols. All animal protocols were approved by the Animal Ethics Committee of the University Health Network.
In vivo determination of intestinal lipoprotein particle production rates
Methods for the in vivo determination of intestinal lipoprotein particle production have been published previously (17). Briefly, 1 d before these studies, femoral venous and arterial catheters were inserted as described previously (13). The animals were fasted overnight for 16 h. A baseline blood sample was drawn followed by an iv bolus of Triton-WR1339 (Sigma Chemical Co., St. Louis, MO). After Triton administration, blood samples were drawn at 10 and 20 min (total blood volume withdrawn for the entire study was 1.2 ml). The Svedberg unit (Sf) > 400 (large TRL) and Sf 100–400 (smaller TRL) fractions were isolated as previously described (17). ApoB48 was quantified using analytical SDS-PAGE as described previously (20), in which there is no contamination by apoB100. Large and small apoB48 and TG secretion rates were derived by multiplying the slope of the concentration increase of apoB48 and TG, respectively, over time by the intravascular distribution volume estimated as 3.8 ml/100 g body weight (14).
Studies performed in the postprandial state were as described above except that they were manually administered 400 μl of lard at time 0 and then every 20 min over 60 min, to achieve a steady fat-fed condition as previously described (17). The bolus of Triton-WR1339 was given 1 h after starting feeding, and blood samples were drawn as described above.
Triton method validation experiments
We performed additional experiments in three animals in which we compared the slope of apoB48 vs. time after Triton administration, calculated from three vs. five time points. We found no significant difference in the slopes calculated from either three or five time points [5.3 ± 0.45 μg/min vs. 5.1 ± 0.58 μg/min, respectively, for Sf > 400, P = not significant (NS); and 5.5 ± 0.85 μg/min vs. 5.3 ± 0.84 μg/min, respectively, for Sf 100–400, P = NS). To keep blood volumes drawn for each experiment to a minimum we have calculated the slope of the apoB48 vs. time curves using three rather than five time points.
Ex vivo protocols
Isolation of primary hamster enterocytes.
The isolation of viable adult villi from hamster small intestine was based on that described by Perreault and Beaulieu (21) with some modification as reported recently (17).
Characteristics of the isolated enterocytes and their functional viability have been documented previously (17). We have also assessed the stability of primary cultured enterocytes over the time course of the experiments reported in the manuscript. Primary enterocytes were radiolabeled for 20 min after 90, 120, and 150 min of incubation. No significant change was observed in the total amount of apoB48 secreted by cultured enterocytes over these time periods (data not shown), although there is loss of the cell’s ability to secrete apoB48 over time after 2.5 h after isolation. The experiments reported in this manuscript, however, always use cells within 2.5 h of isolation, at which time the apoB48 secretion is high and there is greater than 90% cell viability. Thus we have shown that there is no appreciable loss of cell viability or differentiation over the time course of our pulse-chase experiments reported in the manuscript. We have also tested the effect of rosiglitazone in vitro on primary enterocytes and have found no effect on cell viability as assessed by total protein synthesis.
Metabolic labeling of intact primary hamster enterocytes.
Primary hamster enterocytes were used for pulse-chase experiments as described (17). Briefly, cells were preincubated in methionine-free DMEM for 30 min and pulsed with 50 μCi/ml [35S]methionine (equals 1.85 MBq/ml) for 25 min. After the pulse, the cells were washed and chased in methionine-supplemented DMEM. At 90 min, triplicate dishes were harvested, and cells were lysed in solubilization buffer and the lysates were used for immunoprecipitation, as described (22). As a control, we have monitored changes in total protein synthesis by assessing both acid-precipitable radiolabeled protein (trichloroacetic acid counts) as well as total protein mass assays. No specific protein was used as apoB48 is itself a highly specific protein expressed only by intestinal enterocytes.
SDS-PAGE, fluorography, and immunoblotting of primary hamster enterocytes.
Immunoprecipitates were analyzed by SDS-PAGE and fluorography, essentially as described (23). ApoB48 bands, visualized by fluorography, were quantified by excision from the gel, digestion, and scintillation counting essentially as described (22). The 97-kDa subunit of MTP was measured by chemiluminescent immunoblotting as previously described (22), using a goat anti-bovine MTP antiserum provided by Dr. David Gordon (Bristol Myers Squibb, Princeton, NJ). For MTP blots, we measured total protein content and loaded equal amounts of total protein on each lane of the gel (typically 20 μg). Thus, the MTP data shown in Fig. 4 are normalized for protein content. In addition to the measurement of MTP mass in cultured hamster enterocytes, we also measured MTP mass in total intestinal tissue isolated from three hamsters from each treatment group (i.e. without primary culture of hamster enterocytes). MTP protein mass was quantified by densitometry using FluorChem Imaging software (Alpha Innotech Corp., San Leandro, CA). Lipoprotein fractionation was achieved by sequential ultracentrifugation of the culture medium to isolate large chylomicrons (17, 24).
FIG. 4. MTP mass (A and B) and transfer activity (C) in chow-fed hamsters (CHOW), chow-fed hamsters treated with rosiglitazone (CHOW+RSG), fructose-fed hamsters (FRUC), and fructose-fed hamsters treated with rosiglitazone (FRUC+RSG). Data are shown as percentage MTP mass (A and B) and transfer activity (C) compared with CHOW (set at 100%). A, Enterocytes from CHOW (n = 3 hamsters, black bars), FRUC (n = 3, white bars), and FRUC+RSG (n = 3, light gray bars) hamsters (after a 16-h fast) were solubilized, equal amounts of cell protein (20 μg) were subjected to SDS-PAGE, and proteins were transferred onto polyvinylidene difluoride membrane. B, Same analysis as above on enterocytes from CHOW (n = 3 hamsters, black bars) and CHOW+RSG (n = 3, dark gray bars) hamsters. Immunoblotting was performed to detect the 97-kDa MTP subunit with a rabbit anti-hamster MTP antiserum. Shown are representative immunoblots from three experiments expressed as a percentage of the MTP mass detected in chow-fed hamster intestinal cell lysates or intestinal tissue. For cell lysates, there was a significant increase in protein mass of MTP compared with CHOW controls (20% increase in MTP mass; *, P = 0.03 for FRUC vs. CHOW), which was improved with rosiglitazone treatment (30%; , P = 0.008 for FRUC+RSG vs. FRUC). No significant change was observed in MTP mass between CHOW and CHOW+RSG. For MTP transfer activity (C), there was a significant increase in MTP activity in FRUC compared with CHOW (11% increase in MTP activity; *, P < 0.05 for FRUC vs. CHOW), which tended to improve with rosiglitazone treatment (7% decrease; P = NS for FRUC+RSG vs. FRUC). There was no difference in the MTP activity between CHOW and CHOW+RSG.
MTP lipid transfer activity was measured in both primary cultured hamster enterocytes and intestinal tissue by fluorescence spectrophotometry using a kit from Roar Biomedical (kit RB-MTP, New York, NY). Briefly, intestinal pieces from CHOW, CHOW+RSG, FRUC, and FRUC+RSG were homogenized in a Tris buffer, and 100 μg of homogenate or 100 μg of cultured enterocyte lysate were incubated for 6 h at 37 C with donor and acceptor particles provided in the kit. After 6 h, the samples were read in the fluorescence spectrophotometer at an excitation wavelength of 465 nm and emission wavelength of 538 nm.
Other laboratory methods
Glucose was determined on whole blood using a glucometer (Sure Step, One Touch). Plasma insulin concentrations were determined by RIA using a rat insulin kit from Linco Research (St. Louis, MO). This assay has 100% cross-reactivity to hamster insulin, and the intra- and interassay coefficients of variation were 6.8 and 10.6%, respectively. Plasma free fatty acids (FFAs) were measured by a colorimetric method (kit supplied by Wako Industrials, Neuss, Germany) with a coefficient of variation of 4.7% for intra- and 8.2% for interassay variation. TG was measured using a colorimetric assay (Roche Mannheim GmbH Diagnostica, Laval, Quebec, Canada) with an intra- and interassay coefficient of variation of 3.0 and 4.9%, respectively.
Statistical analysis
All the values are reported as mean ± SEM. For comparison of TRL TG and apoB48 secretion rates between CHOW, CHOW+RSG, FRUC, and FRUC+RSG hamsters, ANOVA was used followed by post hoc analysis with Tukey’s test. A P value < 0.05 was considered to be significant.
Results
Effect of rosiglitazone treatment on body weight, plasma insulin, FFA, TG, and glucose (Table 1)
Because of blood volume limitations, fasting blood samples for the above analyses were not taken from animals undergoing in vivo protocols and therefore are not reported for the actual animals that underwent the in vivo studies reported in this manuscript. Data in Table 1 are cumulative data from our laboratory from animals that had undergone identical feeding and rosiglitazone treatment protocols, but these animals did not undergo in vivo studies for assessment of lipoprotein production. There was no significant difference in any of the variables between groups. There was a strong trend for FRUC to have higher TGs, insulin, and glucose, but perhaps because of high inter-animal variability, the differences were not significant.
TABLE 1. Characteristics of fructose-fed (FRUC), fructose fed plus rosiglitazone-treated (FRUC + RSG), chow-fed (CHOW), and chow-fed plus rosiglitazone-treated (CHOW + RSG) hamsters
Relative proportions of apoB48 and -B100 in the TRL fractions
As one might expect, the majority of apoB in the Sf > 100 fraction was apoB48, with very little apoB100 detected. In the Sf 60–100 fraction, 45% of the apoB was apoB48 in the CHOW group, 56% was apoB48 in the FRUC group, and 60% was apoB48 in the FRUC+RSG group. In the Sf 20–60 fraction, 5% of the apoB was apoB48 in the CHOW group, 24% in the FRUC group, and 13% in the FRUC+RSG group. This is in contrast to the proportion of apoB48 to apoB100 in these smaller TRL fractions in healthy humans previously reported from our laboratory (25). In the fasted human Sf 60–400 TRL fraction, apoB48 constituted only 11.5% of total apoB, and in the Sf 20–60 fraction, it constituted 7% of total apoB. In another published human study (26), only 5% of the Sf 60–400 TRL fraction apoB was apoB48, whereas only 6.7% of the Sf 20–60 fraction was apoB48. These data indicate that the CHOW-fed hamster Sf 60–100 fraction has a relatively far greater proportion of apoB48 to -B100 than does the human, whereas the Sf 20–60 fraction has a similar proportion of apoB48 to -B100 in hamsters and humans.
Triton WR-1339 studies for determination of in vivo intestinal lipoprotein production in the fasting state (Fig. 1)
There was an approximately 2.5-fold greater apoB48 secretion rate with chronic fructose feeding compared with chow in large TRL (Sf > 400 fraction; P < 0.01) and an approximately 2-fold higher apoB48 secretion rate in smaller TRL (Sf 100–400 fraction; P < 0.05) (Fig. 1, B and D). Treatment of fructose-fed hamsters with rosiglitazone resulted in an approximately 50% lower apoB48 secretion rate than fructose alone in both TRL fractions (for Sf > 400, P < 0.05 vs. FRUC and P = NS vs. other treatment groups; for the Sf 100–400 fraction, P < 0.05 vs. FRUC and P = NS vs. other treatment groups), but treatment of CHOW-fed hamsters with rosiglitazone resulted in no significant change in apoB48 secretion rate in either the large or smaller TRL fraction. There was no significant difference in TG secretion rates between any of the groups in either large or small TRL fractions (Fig. 1, A and C).
FIG. 1. In vivo production of Sf > 400 (large) (A and B) and Sf 100–400 (smaller) (C and D) TRL particle TG (A and C) and apoB48 (B and D) in the fasted state in control hamsters (CHOW), control hamsters treated with rosiglitazone (CHOW+RSG), fructose-fed hamsters (FRUC), and fructose-fed hamsters treated with rosiglitazone (FRUC+RSG). A, Sf > 400 TG secretion rate (in μmol/min) after iv administration of Triton WR-1339 in CHOW (black bars, n = 13), CHOW+RSG (dark gray bars, n = 6), FRUC (white bars, n = 19), and FRUC+RSG hamsters (gray bars, n = 18). TG secretion rates between the four groups were not statistically different. B, Sf > 400 apoB48 secretion rate was significantly higher in FRUC compared with CHOW (*, P < 0.01) and was reduced with rosiglitazone (+, P < 0.05 for FRUC+RSG vs. FRUC). C, Sf 100–400 TG secretion rates were not statistically different between the four groups. D, Sf 100–400 apoB48 secretion rate was higher in FRUC compared with CHOW (+, P < 0.01) and was significantly reduced with rosiglitazone (+, P < 0.05 for FRUC+RSG vs. FRUC).
Because we observed an increase in apoB48 but not TG production with FRUC vs. CHOW and intermediate apoB48 production rates with FRUC+RSG, we assessed the TRL TG/apoB48 ratio in the fasted state as a crude index of particle size determination. For the Sf > 400 TRL fraction, the ratio of TRL TG/apoB48 was lower in FRUC than CHOW (FRUC, 0.005 ± 0.001, vs. CHOW, 0.009 ± 0.002; P < 0.05), and the ratio was intermediate in FRUC+RSG (0.007 ± 0.001; P = NS compared with the other groups), indicating a tendency for particle size to shift from larger to smaller size with fructose feeding. The Sf 100–400 fraction tended to show a shift toward a smaller particle size in FRUC (CHOW, 0.012 ± 0.001; FRUC, 0.009 ± 0.001; and FRUC+RSG, 0.010 ± 0.001; P = NS). There was no significant difference in the TG/apoB48 between CHOW+RSG and the other groups (0.007 ± 001 for Sf > 400 and 0.010 ± 001 for Sf 100–400).
Triton WR-1339 studies for determination of in vivo intestinal lipoprotein production in the postprandial state (Fig. 2)
TG and apoB48 secretion rates were measured in the large (Sf > 400 TRL) and smaller (Sf 100–400) fractions in CHOW, CHOW+RSG, FRUC, and FRUC+RSG hamsters after 60 min of fat feeding. In the Sf > 400 fraction, TG secretion rate tended to be higher in FRUC compared with CHOW and was significantly decreased with rosiglitazone treatment (P = 0.08 for CHOW vs. FRUC; P < 0.05 for FRUC+RSG vs. FRUC and CHOW) (Fig. 2A). There was no significant change in TG production in the Sf > 400 fraction when CHOW-fed hamsters were treated with RSG. In the Sf 100–400 fraction, TG secretion was increased approximately 2-fold in FRUC and normalized with rosiglitazone (P < 0.05 for FRUC vs. CHOW; P < 0.01 for FRUC+RSG vs. FRUC) (Fig. 2C). There was no significant difference in Sf 100–400 TG secretion in the CHOW+RSG.
FIG. 2. In vivo production of Sf > 400 (large) (A and B) and Sf 100–400 (smaller) (C and D) TRL particle TG (A and C) and apoB48 (B and D) in the fed (postprandial) state in control hamsters (CHOW), control hamsters treated with rosiglitazone (CHOW+RSG), fructose-fed hamsters (FRUC), and fructose-fed hamsters treated with rosiglitazone (FRUC+RSG). TG and apoB48 secretion rates were measured in the Sf > 400 fraction (large TRL) and Sf 100–400 fraction (smaller TRL) in CHOW (black bars, n = 11), CHOW+RSG (dark gray bars, n = 6), FRUC (white bars, n = 11), and FRUC+RSG hamsters (gray bars, n = 8), A, In the Sf > 400 fraction, TG secretion rate tended to be increased in the FRUC hamsters compared with CHOW (P = 0.08) and was significantly decreased with rosiglitazone treatment (+, P < 0.05 for FRUC+RSG vs. FRUC). B, There were no differences in apoB48 secretion in the Sf > 400 fraction between the four groups of animals. C, In the Sf 100–400 fraction, TG secretion was increased approximately 2-fold in FRUC and normalized with rosiglitazone (+, P < 0.05 for FRUC vs. CHOW; *, P < 0.01 for FRUC+RSG vs. FRUC). D, ApoB48 secretion tended to increase in FRUC hamsters (P = NS for all).
ApoB48 secretion was measured in both large and small TRL fractions in the postprandial state. In the Sf > 400 fraction, there were no differences in apoB48 secretion (P = NS for all) (Fig. 2B). In the Sf 100–400 fraction, apoB48 secretion tended to increase (by 2-fold) in FRUC (P = NS for all) (Fig. 2D). There was no difference in apoB48 production rate in CHOW+RSG compared with the other treatment groups.
TG/apoB48 was assessed in the postprandial state. In this case, the TG secretion rate was increased in FRUC, whereas the apoB48 secretion rate only tended to be increased. For the Sf > 400 fraction, the TG/apoB48 was increased in FRUC vs. CHOW and decreased in FRUC+RSG vs. FRUC (0.005 ± 001 for CHOW, 0.005 ± 001 for CHOW+RSG, 0.009 ± 002 for FRUC, and 0.003 ± 001 for FRUC+RSG; P < 0.05 for FRUC vs. CHOW; P < 0.01 for FRUC vs. FRUC+RSG). In the Sf 100–400 fraction, the TG/apoB48 was also increased in the FRUC group and decreased to normal in the FRUC+RSG group (0.005 ± 001for CHOW, 0.005 ± 001 for CHOW+RSG, 0.008 ± 001 for FRUC, and 0.005 ± 001 for FRUC+RSG; P < 0.001 for FRUC vs. CHOW; P < 0.01 for FRUC vs. FRUC+RSG; and P < 0.01 FRUC vs. CHOW+RSG).
Studies of apoB48 secretion in primary cultured enterocytes (Fig. 3)
Pulse-chase-labeling experiments were used to assess the secretion of apoB48 in villus enterocytes isolated from CHOW, FRUC, and FRUC+RSG hamsters. Figure 3, A and B, shows the extracellular secretion of chylomicron-specific and total apoB48 in enterocytes isolated from CHOW, FRUC, and FRUC+RSG hamsters. Enterocytes from FRUC hamsters secreted approximately 4-fold more newly synthesized chylomicron-specific and total apoB48 over a 90-min chase compared with that in CHOW animals (P = 0.01 for chylomicron-specific and total apoB48 in FRUC vs. FRUC+RSG). In fructose-fed animals, there was a 2-fold reduction in chylomicron and total apoB48 secretion with rosiglitazone treatment (P = 0.02 for chylomicron apoB48, and P = 0.01 for total apoB48 in FRUC vs. FRUC+RSG). Total apoB48 secretion was also assessed in a separate group of hamsters fed either CHOW or CHOW+RSG (Fig. 3C). No significant change in apoB48 was observed in CHOW hamsters treated with RSG. It should be noted that, in Fig. 3, apoB48 synthesis and secretion were normalized for both total cellular protein mass and the incorporation of [35S]methionine into total trichloroacetic acid-insoluble cellular and secreted proteins. Therefore, the stimulation of apoB48 synthesis and secretion was not a consequence of global effects of fructose-feeding on protein synthesis and secretion.
FIG. 3. Ex vivo secretion of apoB48 by primary cultured hamster enterocytes in chow-fed hamsters (CHOW), chow-fed hamsters treated with rosiglitazone (CHOW+RSG), fructose-fed hamsters (FRUC), and fructose-fed hamsters treated with rosiglitazone (FRUC+RSG). Primary enterocytes, isolated from the treated hamsters, were pulsed with [35S]methionine and chased for 0, 45, and 90 min. The media samples and cell lysates collected at each chase time point were subjected to immunoprecipitation and then analyzed by SDS-PAGE and fluorography. A and B, Chylomicron and total apoB48 secreted in CHOW, FRUC, and FRUC+RSG, respectively. C, Total apoB48 secreted in CHOW and CHOW+RSG. There was a 4-fold increase in chylomicron and total apoB48 secretion with fructose-feeding, which was reduced 2-fold with rosiglitazone treatment (A: *, P = 0.01 for FRUC vs. CHOW; , P = 0.02 for FRUC+RSG vs. FRUC; B: **, P = 0.01 for FRUC vs. CHOW; , P = 0.01 for FRUC+RSG vs. FRUC). No significant difference was observed in apoB48 secretion in CHOW vs. CHOW+RSG.
MTP mass (Fig 4, A and B) and transfer activity (Fig. 4C) in primary cultured enterocytes and MTP mass and transfer activity measured in intestinal tissue isolated from treated hamsters (not illustrated)
Facilitated secretion of apoB48 and core lipoprotein lipids in FRUC hamster enterocytes could be related to an increase in MTP mass, the key factor involved in the lipoprotein assembly process. To test this hypothesis, equal quantities of primary cultured intestinal cell lysate (20 μg) were analyzed by immunoblotting. As shown in Fig. 4A, in FRUC hamster enterocyte cell lysates, there was a significant increase in protein mass of MTP compared with CHOW controls (20% increase in MTP mass; P = 0.03), which was ameliorated by rosiglitazone treatment (30%; P = 0.008). Interestingly, no significant change was observed in MTP mass between CHOW and CHOW+RSG (Fig. 4B). The MTP transfer activity measured in enterocyte cell lysates mirrored the MTP protein mass data (Fig. 4C), with a significant increase in MTP transfer activity in FRUC vs. CHOW (110.5 ± 2.57 vs. 100.0 ± 2.58%; P < 0.05) and partial amelioration of this FRUC-induced increase with the addition of rosiglitazone treatment (103.2 ± 3.63%; P = NS for FRUC+RSG vs. FRUC). No difference was observed between CHOW+RSG (101.0 ± 1.65%) and CHOW.
MTP mass and transfer activity were also measured directly in intestinal tissue (30 μg total protein per lane), but unlike the cell lysates, there was no significant difference in tissue MTP mass between the treatment groups, although there was a tendency for the mass to be higher in FRUC and FRUC+RSG vs. CHOW (MTP mass in CHOW = 100.0 ± 3.0%, FRUC = 110.4 ± 4.3%, and FRUC+RSG =118.7 ± 3.3%; P = NS between groups) (not illustrated). MTP activity measured directly in intestinal tissue was significantly increased by 13% in FRUC vs. CHOW (MTP activity in CHOW = 100 ± 2.94% and FRUC = 112.9 ± 4.83%; P < 0.01). Rosiglitazone did not, however, ameliorate this increase in MTP activity (FRUC+RSG=115.9 ± 4.72%, p=ns vs. FRUC, P < 0.001 vs. CHOW). CHOW+RSG (103.1 ± 1.2%) was not different from CHOW.
Discussion
We have demonstrated previously that treatment of fructose-fed, nondiabetic, insulin-resistant hamsters with rosiglitazone, a member of the thiazolidinedione class of insulin sensitizers with specific peroxisome proliferator-activated receptor- agonist activity, improved whole-body and hepatic insulin sensitivity and also decreased the overproduction of very-low-denisty lipoprotein apoB by hepatocytes ex vivo (14). In the present study, we have also shown that rosiglitazone treatment reduced the secretion of intestinally derived apoB48-containing lipoproteins in the fasting state and TRL TG secretion in the postprandial state in fructose-fed but not in chow-fed hamsters. The secretion of intestinal apoB48-containing particles was examined both in vivo and ex vivo, and similar results were obtained. Furthermore, rosiglitazone treatment was associated with a reversal of the increased intestinal expression of MTP seen with fructose feeding when measured in intestinal cell lysates, indicating one potential molecular mechanism by which rosiglitazone led to reduction of intestinal particle secretion in this insulin-resistant animal model. Whether this improvement in intestinal lipoprotein particle secretion resulted from the insulin-sensitizing effects of rosiglitazone or from another unrelated action of the drug will require additional study perhaps using relevant knockout and transgenic animal models.
In the present study, apoB48 secretion in the fasting state was increased in large (Sf > 400) and smaller (Sf 100–400) TRL fractions, whereas there was no significant difference in TG secretion. The absence of a significant difference in TG secretion between control and insulin-resistant fructose-fed animals in the fasted state suggests that there is increased production of small, lipid-poor, apoB48-containing lipoproteins in these size ranges in the fasting state with fructose feeding. Fructose feeding was indeed associated with the production of smaller particles in the larger TRL Sf > 400 fraction, as evidenced by a reduction in the TG/apoB48 ratio, and a tendency toward smaller particle production in the smaller Sf 100–400 fraction. We speculate that the intestine constitutively secretes small apoB48-containing lipoprotein particles (and a greater number in insulin-resistant states) to be primed and ready for the ingestion of fat. In the fed (postprandial) state, fructose feeding and rosiglitazone treatment manifested in an increase and correction, respectively, of TRL TG, a feature that was not evident in the fasted state, in which predominantly lipid-poor TRLs are secreted. Both chronic fructose feeding and acute fat ingestion also stimulate hepatic very-low-density lipoprotein secretion, as we have shown previously for the former in the fructose-fed hamster (13) and others have shown in humans for the latter after a high-fat meal (27).
The relative proportion of apoB48 as a fraction of total apoB in the Sf 60–100 but not the Sf 20–60 fraction in fasted, chow-fed hamsters was greater than that found in humans and increased proportionately with fructose feeding. Unlike most other rodents, in the hamster liver there is negligible editing of apoB, and we have never been able to detect apoB48 secreted by cultured hamster hepatocytes. The intestine of the hamster, therefore, produces far more TRL apoB48 relative to hepatic apoB100 in this TRL fraction than does the human. Perhaps the hamster, because it nibbles rather than gorges like the human and stores food in its cheek, has evolved to secrete relatively more apoB48 from its intestine relative to hepatic apoB100, compared with the human, to accommodate the transport of fairly persistent food ingestion. These differences between hamster and human caution us not to generalize the present findings in the hamster model to other species such as humans. Future studies will need to examine this phenomenon directly in other species, including the human.
Our ex vivo studies of enterocyte cell lysates showed that treatment with rosiglitazone was associated with a reversal of the increased expression of MTP seen with fructose feeding. Measurement of MTP mass and lipid transfer activity directly in intestinal tissue, however, showed less consistent changes, possibly because of the presence of other cell types masking enterocyte-specific changes in MTP mass and transfer activity. Although these changes in MTP expression are small, there is ample evidence from the literature to support the finding that relatively small changes in MTP levels can result in significant changes in lipoprotein production rates (28, 29, 30, 31). The reduction in MTP levels with rosiglitazone treatment may have been implicated in the reduction of intestinal lipoprotein secretion in the present study. MTP is an important factor in intestinal lipoprotein assembly. In Caco-2 cells, MTP has been shown not only to be involved in the first step of lipoprotein synthesis, i.e. the rescue of apoB from intracellular degradation through early lipidation of the protein, but also to be involved in further steps involving association of lipoprotein particles with TG droplets (32). The promoter region of the MTP gene contains a negative insulin-response element (33), and intestinal MTP mRNA has been shown to be raised in diabetic and insulin-resistant rats (34, 35). It is possible, therefore, that the reduction in MTP levels induced by rosiglitazone treatment was the result of improved insulin signaling at the level of the enterocyte. The precise molecular signaling pathway involved in insulin-mediated modulation of MTP expression is currently unclear. Additional studies are needed to elucidate the role of insulin in modulating MTP expression and in regulating intestinal particle secretion.
Intestinal lipoprotein production has been felt to be regulated predominantly by the amount of fat ingested because of the efficiency of fat absorption and rapid turnover of enterocytes. The mucosal surface of the gastrointestinal tract is remarkable for the very rapid turnover of the epithelial cell population. Enterocytes of the small intestine are replaced rapidly by the division of stem cells at the base of the crypts of Lieberkuhn and movement of cells up the crypts to the villi as they differentiate into mature enterocytes. It is likely that the population of absorptive cells is replaced after 24–72 h. The rapid turnover of intestinal cells, therefore, would seem to preclude a chronic state such as insulin resistance from affecting the intestine’s capacity to produce lipoprotein particles. We have previously performed short-term (2 d) fructose feeding studies in hamsters to investigate the effect on apoB48 lipoprotein formation and found no significant effect on the secreted level of intestinal chylomicron-apoB48, showing that the effect of fructose feeding on intestinal lipoprotein production is a chronic process (17). It has been suggested that consumption of high-fat diets during maturation causes adaptation of the enterocytes so that they have an increased capacity to absorb lipid and secrete chylomicrons (36, 37). Such intestinal adaptation may also underlie the effect of an insulin-resistant state on the intestine’s capacity to produce lipoprotein particles.
In conclusion, we have shown that whole-body insulin sensitization with rosiglitazone treatment is associated with a reduction in intestinal MTP overexpression and apoB48-containing particle hypersecretion in the fructose-fed insulin-resistant hamster. Although we cannot be sure that the beneficial effect of rosiglitazone was mediated by insulin sensitization, our findings suggest that therapeutic measures that effectively ameliorate insulin resistance, or that reduce MTP overexpression in insulin-resistant states, could be part of the strategy to correct the intestinal particle oversecretion associated with insulin resistance. This could have significant implications in the management of the dyslipidemia and possibly atherosclerosis of type 2 diabetes mellitus and insulin resistance. Additional studies examining the link between insulin resistance and intestinal lipoprotein overproduction need to be conducted in humans to determine whether this process is relevant to human disease states.
References
Lewis GF, Steiner G 1996 Hypertriglyceridemia and its metabolic consequences as a risk factor for atherosclerotic cardiovascular disease in non-insulin-dependent diabetes mellitus. Diabetes Metab Rev 12:37–56
Lewis GF, O’Meara NM, Soltys PA, Blackman JD, Iverius PH, Pugh WL, Getz GS, Polonsky KS 1991 Fasting hypertriglyceridemia in non-insulin-dependent diabetes mellitus is an important predictor of postprandial lipid and lipoprotein abnormalities. J Clin Endocrinol Metab 72:934–944
Curtin A, Deegan P, Owens D, Collins P, Johnson A, Tomkin GH 1996 Elevated triglyceride-rich lipoproteins in diabetes: a study of apolipoprotein B-48. Acta Diabetol 33:205–210
Mero N, Syvanne M, Taskinen MR 1998 Postprandial lipid metabolism in diabetes. Atherosclerosis 141(Suppl 1):S53–S55
Doi H, Kugiyama K, Ohgushi M, Sugiyama S, Matsumura T, Ohta Y, Nakano T, Nakajima K, Yasue H 1998 Remnants of chylomicron and very low density lipoprotein impair endothelium-dependent vasorelaxation. Atherosclerosis 137:341–349
Mero N, Malmstrom R, Steiner G, Taskinen MR, Syvanne M 2000 Postprandial metabolism of apolipoprotein B-48- and B-100-containing particles in type 2 diabetes mellitus: relations to angiographically verified severity of coronary artery disease. Atherosclerosis 150:167–177
McNamara JR, Shah PK, Nakajima K, Cupples LA, Wilson PW, Ordovas JM, Schaefer EJ 2001 Remnant-like particle (RLP) cholesterol is an independent cardiovascular disease risk factor in women: results from the Framingham Heart Study. Atherosclerosis 154:229–236
Karpe F, Boquist S, Tang R, Bond GM, de Faire U, Hamsten A 2001 Remnant lipoproteins are related to intima-media thickness of the carotid artery independently of LDL cholesterol and plasma triglycerides. J Lipid Res 42:17–21
Haffner SM, Foster DM, Kushwaha RS, Hazzard WR 1984 Retarded chylomicron apolipoprotein-B catabolism in type 2 (non-insulin-dependent) diabetic subjects with lipaemia. Diabetologia 26:349–354
Mamo JC, Hirano T, Sainsbury A, Fitzgerald AK, Redgrave TG 1992 Hypertriglyceridemia is exacerbated by slow lipolysis of triacylglycerol-rich lipoproteins in fed but not fasted streptozotocin diabetic rats. Biochim Biophys Acta 1128:132–138
Bar-On H, Levy E, Oschry Y, Ziv E, Shafrir E 1984 Removal defect of very-low-density lipoproteins from diabetic rats. Biochim Biophys Acta 793:115–118
Steinbrecher UP, Witztum JL 1984 Glucosylation of low-density lipoproteins to an extent comparable to that seen in diabetes slows their catabolism. Diabetes 33:130–134
Taghibiglou C, Carpentier A, Van Iderstine SC, Chen B, Rudy D, Aiton A, Lewis GF, Adeli K 2000 Mechanisms of hepatic very low density lipoprotein overproduction in insulin resistance. Evidence for enhanced lipoprotein assembly, reduced intracellular ApoB degradation, and increased microsomal triglyceride transfer protein in a fructose-fed hamster model. J Biol Chem 275:8416–8425
Carpentier A, Taghibiglou C, Leung N, Szeto L, Van Iderstine SC, Uffelman KD, Buckingham R, Adeli K, Lewis GF 2002 Ameliorated hepatic insulin resistance is associated with normalization of microsomal triglyceride transfer protein expression and reduction in very low density lipoprotein assembly and secretion in the fructose-fed hamster. J Biol Chem 277:28795–28802
Arbeeny CM, Meyers DS, Bergquist KE, Gregg RE 1992 Inhibition of fatty acid synthesis decreases very low density lipoprotein secretion in the hamster. J Lipid Res 33:843–851
Liu GL, Fan LM, Redinger RN 1991 The association of hepatic apoprotein and lipid metabolism in hamsters and rats. Comp Biochem Physiol A 99:223–228
Haidari M, Leung N, Mahbub F, Uffelman KD, Kohen-Avramoglu R, Lewis GF, Adeli K 2002 Fasting and postprandial overproduction of intestinally derived lipoproteins in an animal model of insulin resistance. Evidence that chronic fructose feeding in the hamster is accompanied by enhanced intestinal de novo lipogenesis and ApoB48-containing lipoprotein overproduction. J Biol Chem 277:31646–31655
Buckingham RE, Al Barazanji KA, Toseland CD, Slaughter M, Connor SC, West A, Bond B, Turner NC, Clapham JC 1998 Peroxisome proliferator-activated receptor- agonist, rosiglitazone, protects against nephropathy and pancreatic islet abnormalities in Zucker fatty rats. Diabetes 47:1326–1334
Taghibiglou C, Rashid-Kolvear F, Van Iderstine SC, Le Tien H, Fantus IG, Lewis GF, Adeli K 2002 Hepatic very low density lipoprotein-ApoB overproduction is associated with attenuated hepatic insulin signaling and overexpression of protein-tyrosine phosphatase 1B in a fructose-fed hamster model of insulin resistance. J Biol Chem 277:793–803
Karpe F, Hamsten A, Uffelman K, Steiner G 1996 Apolipoprotein B-48. Methods Enzymol 263:95–104
Perreault N, Beaulieu JF 1998 Primary cultures of fully differentiated and pure human intestinal epithelial cells. Exp Cell Res 245:34–42
Taghibiglou C, Rudy D, Van Iderstine SC, Aiton A, Cavallo D, Cheung R, Adeli K 2000 Intracellular mechanisms regulating apoB-containing lipoprotein assembly and secretion in primary hamster hepatocytes. J Lipid Res 41:499–513
Laemmli UK 1970 Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680–685
Luchoomun J, Zhou Z, Bakillah A, Jamil H, Hussain MM 1997 Assembly and secretion of VLDL in nondifferentiated Caco-2 cells stably transfected with human recombinant ApoB48 cDNA. Arterioscler Thromb Vasc Biol 17:2955–2963
Lemieux S, Fontani R, Uffelman KD, Lewis GF, Steiner G 1998 Apolipoprotein B-48 and retinyl palmitate are not equivalent markers of postprandial intestinal lipoproteins. J Lipid Res 39:1964–1971
Karpe F, Steiner G, Olivecrona T, Carlson LA, Hamsten A 1993 Metabolism of triglyceride-rich lipoproteins during alimentary lipemia. J Clin Invest 91:748–758
Cohn JS, Wagner DA, Cohn SD, Millar JS, Schaefer EJ 1990 Measurement of very low density and low density lipoprotein apolipoprotein (Apo) B-100 and high density lipoprotein Apo A-I production in human subjects using deuterated leucine. Effect of fasting and feeding. J Clin Invest 85:804–811
Jamil H, Gordon DA, Eustice DC, Brooks CM, Dickson Jr JK, Chen Y, Ricci B, Chu CH, Harrity TW, Ciosek Jr CP, Biller SA, Gregg RE, Wetterau JR 1996 An inhibitor of the microsomal triglyceride transfer protein inhibits apoB secretion from HepG2 cells. Proc Natl Acad Sci USA 93:11991–11995
Jamil H, Chu CH, Dickson Jr JK, Chen Y, Yan M, Biller SA, Gregg RE, Wetterau JR, Gordon DA 1998 Evidence that microsomal triglyceride transfer protein is limiting in the production of apolipoprotein B-containing lipoproteins in hepatic cells. J Lipid Res 39:1448–1454
Ledmyr H, Karpe F, Lundahl B, McKinnon M, Skoglund-Andersson C, Ehrenborg E 2002 Variants of the microsomal triglyceride transfer protein gene are associated with plasma cholesterol levels and body mass index. J Lipid Res 43:51–58
Liao W, Kobayashi K, Chan L 1999 Adenovirus-mediated overexpression of microsomal triglyceride transfer protein (MTP): mechanistic studies on the role of MTP in apolipoprotein B-100 biogenesis. Biochemistry 38:10215
van Greevenbroek MM, Robertus-Teunissen MG, Erkelens DW, de Bruin TW 1998 Participation of the microsomal triglyceride transfer protein in lipoprotein assembly in Caco-2 cells: interaction with saturated and unsaturated dietary fatty acids. J Lipid Res 39:173–185
Hagan DL, Kienzle B, Jamil H, Hariharan N 1994 Transcriptional regulation of human and hamster microsomal triglyceride transfer protein genes. Cell type-specific expression and response to metabolic regulators. J Biol Chem 269:28737–28744
Gleeson A, Anderton K, Owens D, Bennett A, Collins P, Johnson A, White D, Tomkin GH 1999 The role of microsomal triglyceride transfer protein and dietary cholesterol in chylomicron production in diabetes. Diabetologia 42:944–948
Phillips C, Owens D, Collins P, Tomkin GH 2002 Microsomal triglyceride transfer protein: does insulin resistance play a role in the regulation of chylomicron assembly? Atherosclerosis 160:355–360
Thomson AB, Cheeseman CI, Keelan M, Fedorak R, Clandinin MT 1994 Crypt cell production rate, enterocyte turnover time and appearance of transport along the jejunal villus of the rat. Biochim Biophys Acta 1191:197–204
Thomson AB, Keelan M, Wild GE 1996 Nutrients and intestinal adaptation. Clin Invest Med 19:331–345(Gary F. Lewis, Kristine U)