|
|
||||||||
Journal of Lipid Research, Vol. 43, 706-712, May 2002
Copyright © 2002 by Lipid Research, Inc.




* Departments of Medicine, University of Western Australia
Physiology, University of Western Australia
Department of Nutrition, Dietetics and Food Sciences, Curtin University of Technology, Western Australia
1 To whom correspondence should be addressed at the University Department of Medicine, Royal Perth Hospital, GPO Box X2213, Perth, WA 6847, Australia. e-mail: gfwatts{at}cyllene.uwa.edu.au
| ABSTRACT |
|---|
|
|
|---|
Abbreviations: apo, apolipoprotein; FCR, fractional catabolic rate; HOMA, homeostasis model assessment; RLP, remnant-like particle
Supplementary key words HMG-CoA reductase inhibitor dyslipidaemia triglyceride-rich lipoproteins
| INTRODUCTION |
|---|
|
|
|---|
While a hypocaloric diet and other lifestyle modifications in obese subjects may achieve significant weight loss and correct lipoprotein metabolism (5), including postprandial dyslipidaemia, adherence to such programs is generally unsuccessful (6). Effective management of dyslipidaemia in obesity may therefore often require the use of lipid-regulating pharmacotherapy (7, 8). HMG-CoA reductase inhibitors have been shown to be effective in treating dyslipidaemia (9), but their effect and mechanism of action on chylomicron remnant metabolism in subjects with visceral obesity and insulin resistance is not fully established. Because de novo cholesterol synthesis is co-regulated with the expression of the LDL receptor (10), which may in turn clear chylomicron remnants, statins, by inhibiting cholesterol synthesis, may decrease plasma chylomicron remnant concentrations by increasing their fractional catabolism.
Standard methods for assessing the kinetics of chylomicron remnant metabolism involve measurement of the response of plasma triglycerides, retinyl esters, or apoB-48 to an oral fat load (2, 11). ApoB-48 and remnant-like particle (RLP) cholesterol are static markers for the metabolism of chylomicron and their remnants (12, 13). A single fasting plasma level of apoB-48 may also predict the kinetic response of retinyl esters and apoB-48 to a fat load (14). However, none of these tests provides specific information relating to chylomicron remnant kinetics or metabolism. We have previously described and validated a novel breath test based on the injection of a remnant-like emulsion labeled with cholesteryl [13C]oleate and subsequent measurement of 13CO2 enrichment in breath (15) that provides a specific functional assessment of chylomicron remnant metabolism in human (16, 17) and animal studies (18, 19).
In this study, we undertook a controlled trial to examine the effect of an HMG-CoA reductase inhibitor (atorvastatin) on chylomicron remnant metabolism in viscerally obese men with dyslipidaemia based on the the kinetics of chylomicron remnant-like emulsions measured with the breath test, and measurements of fasting apoB-48 and RLP cholesterol.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Study design
The study reported herein represents a substudy of a larger randomized, doubled-blind, placebo-controlled intervention trial examining the effects of atorvastatin and other interventions on aspects of lipoprotein metabolism in obesity, which will be reported separately. Eligible patients entered a 3 week run-in diet-stabilizing period, at the end of which they were randomized to a 6 week treatment period of either atorvastatin (40 mg orally at night) or matching placebo. All subjects were advised to continue their habitual isocaloric diet, and to keep physical exercise constant. Compliance with atorvastatin and placebo was checked by tablet count at week 3 and 6 treatment periods.
Clinical protocol
All subjects were admitted to the metabolic ward in the morning after a 14 h fast. They were studied in a semi-recumbent position and allowed water only. Venous blood was collected for measurements of biochemical analyses. Arterial blood pressure was recorded after 3 min in the supine position using a Dinamap 1846 SX/P monitor (Critikon Inc, Tampa, FL).
The sterile isotopically labeled chylomicron remnant-like emulsion (14 ml) was injected intravenously into an antecubital vein via a 21G butterfly needle. End-expiratory breath samples were collected into a vacutainer (Labco Ltd) at baseline and post-injection every 10 min for the first hour, every 20 min for the second hour, every 30 min for the next 5 h and hourly for another 3 h. Subjects were then given a snack and allowed to go home. Two additional breath samples were collected on the following days (24 h). During the first 10 h of collection of breath samples subjects sat quietly in a chair and were allowed to drink only water. After leaving the metabolic ward, they were requested to refrain from vigorous activity and to eat a light supper before retiring.
Dietary analysis
Dietary intake was assessed for energy and major nutrients using at least two 24 h dietary diaries. Diets were subsequently analyzed using DIET 4 Nutrient Calculation Software (Xyris Software, Queensland, Australia). Dietary intake, alcohol, and exercise diaries were also completed at baseline and at the end of the study.
Preparation of stable isotope-labeled remnant-like emulsions
This was carried out as described previously (16). Briefly, pure lipid mixtures containing triolein (135 mg), phosphatidylcholine (75 mg), cholesteryl [13C]oleate (70 mg), and cholesterol (24 mg) were emulsified by sonication for 1 h in 2.2% glycerol in water. After sonication the mixture was centrifuged for 10 min to remove titanium fragments and then filtered into sterile vessels. All emulsion preparations were confirmed to be sterile and pyrogen-free (Pharmacy Dept., Royal Perth Hospital). Uniformly labeled [13C]oleate was purchased from Novachem Pty. Ltd., Victoria, Australia, and cholesteryl [13C]oleate was synthesized from cholesterol and [13C]oleic acids as described previously (15).
Biochemical analyses
Plasma triglyceride and cholesterol concentrations were determined by standard enzymatic methods using a Hitachi 917 Biochemical Analyser (Hitachi Ltd, Tokyo, Japan). HDL cholesterol (HDL-C) was measured by an enzymatic colorimetric method using a commercial kit (Boehringer Mannheim, Mannheim, Germany). LDL-C was calculated using the Friedewald equation. ApoB and apoA-I were determined by immuno-nephelometry. ApoB-48 was measured as described previously (14) with the exception that apoB-48 was assayed directly from plasma rather than 1.063 g/ml lipoprotein fraction. Values obtained using either method are identical (J. C. L. Mamo and A. P. James, unpublished observation). Plasma RLP-C was determined with a JIMRO-II (Japan Immunoresearch Laboratories, Takasaki, Japan) assay kit using an immunoseparation method described by Nakajima et al. (20). Plasma non-esterified fatty acids (NEFAs) were measured by an enzymatic, colorimetric method using a commercial kit (Randox, Co., Antrim, UK). Plasma insulin was measured by radioimmunoassay (DiaSorini s.r.l., Saluggia, Italy) and glucose concentration by a hexokinase method on a Hitachi 917 analyser. Insulin resistance was estimated by the homeostasis model assessment (HOMA) that employs the formula: fasting insulin (mU/l) x fasting plasma glucose (mmol/l)/22.5, as described by Matthews et al. (21). Plasma lathosterol concentration and apoE genotype were determined as described elsewhere (22). CETP activity was determined using a CETP Activity Kit (Roar Biomedical Inc.): briefly, 10 µl of plasma was diluted (1:10) in 90 µl of sample buffer; 20 µl of the diluted plasma was combined with 4 µl of donor (synthetic phospholipid and cholesteryl ester) and 4 µl of VLDL acceptor particle to a total volume of 200 µl and incubated for 3 h at 37°C in a fluorescence-compatible microtiter plate; the fluorescence intensity was then read at excitation wavelength of 465 nm and emission wavelength of 535 nm in a fluorescence spectrometer. Plasma liver (alanine transferase, asparate transferase, alkaline phosphatase) and muscle (creatinine kinase) enzymes were measured at the beginning and end of the study. Interassay CV of all measurements were less than 6%. The expired CO2 in the exhaled breath samples was analyzed at each time point by isotope ratio mass spectroscopy using a Finnigan BreathPlus machine (Thermoquest Systems Pty. Ltd, Sydney, Australia). The ratio of 13CO2/12CO2 was referenced to Peedeebelimnite standard values and the delta unit value was calculated using the Breathmat software. The delta units reference a sample of limestone, a standard in the 13-carbon isotope ratio field, and basal (non-enriched) values correspond approximately to 1% 13 carbon (16).
Kinetic analysis
A compartment model describing the appearance of labeled CO2 in breath was developed using the SAAM II program (SAAM Institute, Seattle, WA, USA). The model was developed assuming that the fractional rate constants [k(i,j)] were time invariant and first order. The model used to fit the tracer data is shown in Fig. 1
. The differential equations describing this model are shown below:
|
Compartment 1 of the model represents the plasma compartment into which the labeled remnant-like emulsion is injected. Material in compartment 1 can be lost from plasma to compartments 2 or 3. Compartment 2 represents labeled CO2 in breath and is sampled during the course of the study. The primary pathway for clearance of emulsion and appearance of labeled CO2 is via compartment 3. This compartment may include the hepatic and extrahepatic processes associated with uptake, hydrolysis, and oxidation of the labeled oleate.
The compartment model was fitted to the observed 13CO2 breath data to provide estimates of the model rate constants and FCR. By using measures of model order, such as the Akaike information criterion (23), the model structure shown in Fig. 1 was the simplest model that was consistent with the tracer data. In the absence of additional data, a more complex model, including hepatic compartmentalization, could not be included in the present model structure.
Statistical analysis
All analyses were carried out using SPSS 10.1 (SPSS, Inc., Chicago, IL). Skewed variables were examined after logarithmic transformation. Differences were tested using one-sample and two-sample t-tests. Associations were examined by linear regression methods. Statistical significance was defined at the 5% level using a two-tailed test.
| RESULTS |
|---|
|
|
|---|
|
|
Figure 2 shows the enrichment of 13CO2 in the breath before and during treatment with atorvastatin over time following injection of the emulsion. In both groups there was a rapid increase in enrichment of 13CO2 with a peak at approximately 6 h, reflecting the catabolism of remnant-like particles via lipolysis and hepatic removal, followed by a slow decline period over the following 19 h. Although enrichment curves were of similar contour, peak enrichment in the atorvastatin group was higher and maximum enrichment was attained earlier. Table 3 shows the model rate constants and FCR for the labeled chylomicron remnants in the two groups. Relative to placebo, there were no significant changes in k(0,2) and k(2,3), but k(2,1), k(3,1), and FCR increased significantly after atorvastatin treatment. The FCR during treatment with atorvastatin was also lower compared with a reference, normolipidaemic population (16, 18).
|
|
| DISCUSSION |
|---|
|
|
|---|
The effects of HMG-CoA reductase inhibitors on intestinally-derived triglyceride-rich lipoprotein metabolism have been examined previously (2529). In two uncontrolled trials, retinyl palmitate was used as a postprandial marker. Cabezas et al. (25) reported in patients with familial combined hyperlipidaemia that simvastatin improved postprandial chylomicron remnant clearance. Cianflone et al. (26) also reported increased remnant clearance with lovastatin in normotriglyceridaemic subjects. However, the reliability of retinyl esters as a postprandial marker is questionable (2, 30). The rationale for its use as a postprandial marker is based on the assumption that the retinyl esters remain associated with the chylomicron remnant. However, Krasinski et al. (30) found that about one third of the postprandial retinyl esters associated with HDL and LDL. In another two uncontrolled trials using apoB-48 as a postprandial maker, Parhofer et al. (27) found that chylomicron remnant-apoB-48 clearance was increased with atorvastatin in normolipidaemic subjects. Battula et al. (28) found in Type 2 diabetes that cerivastatin improved postprandial response of chylomicron apoB-48. However, in these two trials they failed to detect a decrease in fasting plasma apoB-48 level with statin treatment. Using a radioactive labeled chylomicron-like emulsion, Santos et al. (29) recently reported in patients with coronary artery diseases pravastatin increased chylomicron remnant clearance. Not all studies suggest that statins universally improve postprandial lipidaemia (31), although the clearance of chylomicron remnants has not been specifically examined in studies reporting no treatment effects. We have extended the aforementioned studies by investigating the effect of atorvastatin on chylomicron remnant metabolism in obesity employing a stable isotope remnant-like emulsion and measurements of fasting plasma apoB-48 and RLP-C. Our data are consistent with the findings that atorvastatin increases catabolism of chylomicron remnants in obese patients as reflected by the increase in FCR of remnant-like emulsions. Although we did not examine postprandial response of apoB-48, we would expect atorvastatin would have attenuated postprandial apoB-48 response based on our previous findings that the fasting apoB-48 predicts the postprandial response (14).
That visceral obesity is associated with disturbances in lipoprotein metabolism is well recognized (1, 32, 33). As seen in our patients, the principal abnormalities include elevation in plasma TRLs and to a lesser extent LDL-C, with a reciprocal decrease in HDL-C. Elevated fasting levels and postprandial responses of apoB-48 and apoB-100 containing TRLs have been shown in viscerally obese subjects, and these may be partially related to insulin resistance (34). Insulin resistance has two potential effects on chylomicron remnant metabolism. First, it down-regulates LDL receptor expression and activity (35). Second, it increases hepatic de novo cholesterol synthesis and subsequently VLDL apoB-100 secretion (5, 36). These effects would result in increased competition for hepatic receptors between chylomicron and VLDL remnants, thereby disturbing or delaying the uptake of chylomicron remnants by this pathway. As recently emphasized, the binding of apoE on the remnant particle to the hepatic LDL/E receptor is central to the disposal of chylomicron remnants (37). Compared with a reference normolipidaemic population (17, 24), plasma apoB-48 and RLP-C levels were markedly elevated and the FCR of remnant-like emulsion was lower in our obese patients. This reflects delayed chylomicron remnant clearance in these obese subjects with dyslipidaemia and insulin resistance.
HMG-CoA reductase inhibitors have been shown to decrease de novo cholesterol synthesis, thereby upregulating LDL receptor activity and decreasing hepatic secretion of apoB containing lipoproteins. If there was an increased LDL receptor activity and decreased competition with apoB-100 containing lipoproteins for receptor-mediated clearance pathway, the clearance of chylomicron remnants following atorvastatin would be subsequently enhanced. Our findings show that in the setting of insulin resistance, inhibition of cholesterol synthesis with atorvastatin, as reflected by a significant decrease in plasma lathosterol, does improve these abnormalities.
Our findings demonstrated there was a greater reduction on atorvastatin in plasma RLP-C (-33%) than in apoB-48 (-23%) levels, and the decrease in plasma triglyceride in our study was associated with RLP-C but not with apoB-48. This may be due to several reasons. The decrease in plasma apoB-48 and RLP-C levels may reflect a reduction in chylomicron remnant number and size. Plasma apoB-48 level exclusively reflects intestinal-derived chylomicrons and their remnants, whereas RLP-C constitutes a spectrum of intestinal and hepatic derived triglyceride-rich lipoprotein remnants (20). The action of atorvastatin may be to primarily decrease triglyceride-rich lipoproteins of hepatic origin. Given these factors, we might expect that RLP-C would be more sensitive to the change in TRL metabolism on atorvastatin treatment than apoB-48. However, apoB-48 and RLP-C do not specifically reflect chylomicron remnants. These plasma markers of chylomicron remnant metabolism have considerable kinetic heterogeneity. Hence, the lack of correlation between apoB-48, RLP-C, and the FCR of the emulsion might have been anticipated. The decrease in apoB-48 and RLP-C might also possibly have been due to a decrease in apoB-48 and hepatic VLDL apoB-100 secretions (38). That plasma apoB-48 and RLP-C levels and FCR of the remnant-like particle emulsion did not return to normal following treatment reflects the effects of continuing insulin resistance of TRLs, including chylomicron remnants, as was reviewed previously (34).
The significant inverse association between changes in FCR- and LDL-C also supported the notion that there was a decreased competition for the hepatic LDL/E receptor. The importance of the LDL/E receptor in regulating chylomicron remnant metabolism has been shown in previous studies (18, 37). However, we recently reported that in patients with homozygous or heterozygous FH the FCR of the emulsion was not different from that in normolipidaemic controls (17) and was also not altered by simvastatin treatment (39). The discrepancies with the present report might be due to differences between FH and obese subjects in the expression and functionality of LDL and other receptors that clear chylomicron remnants, as well as to group differences in the response of these receptors to statin therapy. In our study, the significant increase in the rate constants k(3,1) and k(2,1) values reflect a parallel increase in both hepatic and extrahepatic removal of remnant-like emulsions on atorvastatin treatment. Total FCR of remnant-like emulsions, however, predominantly involved an increase in their removal by the liver, as reflected by the relative magnitude of these two rate constants.
Direct measurement of chylomicron remnant metabolism is technically difficult. The traditional methods for accessing chylomicron metabolism lack specificity for distinguishing defects in lipolysis from defects in remnant clearance. The breath test is relatively simple to administer in clinical studies. It provides a specific functional assessment of chylomicron remnant metabolism and has been previously validated in patients with familial dyslipidaemias (16, 17). We standardized conditions for carrying out the breath test by studying volunteers in the postabsorptive state, restricting physical activity, and maintaining habitual dietary intake constant.
The breath test depends not only on the plasma clearance of remnant-like emulsions, but also on variations in respiratory quotient, fatty acid uptake, and oxidation. This might explain why the FCR of the emulsion using this test was of smaller magnitude than that reported by Santos et al., who used direct measurements in plasma to estimate the FCR of chylomicron-like emulsion (29). We cannot exclude the possibility of increased insulin sensitivity and decreased cytosolic triglyceride stores in the liver by atorvastatin would enhance the FCR of remnant-like emulsions. However, there were no significant changes in body weight, HOMA score, or plasma NEFAs on atorvastatin treatment. The confounding influence of CETP also needs to be acknowledged. The absence of significant effect of atorvastatin on plasma CETP activity in this study suggests that the effect of atorvastatin on the FCR of the emulsion is not likely to be mediated by CETP. Also, CETP activity was not correlated with the FCR of the emulsion at baseline (data not shown). However, further in vivo studies of the effect of CETP activity on the breath test are required.
In conclusion, we suggest that in obese subjects inhibition of cholesterol synthesis with atorvastatin improves chylomicron remnant metabolism by increasing chylomicron remnant catabolism. Since elevation of triglyceride-rich remnants is a risk factor for atherosclerosis (40), our results suggest that statins may decrease coronary heart disease risk in obesity not only by improving the kinetics of apoB-100 containing lipoproteins but also by improving chylomicron remnant kinetics. Further studies should examine the additive effect of weight reduction or insulin sensitizers together with a statin in dyslipidaemic obese subjects. Finally, the kinetic bases for other lipoprotein changes also require investigation.
| ACKNOWLEDGMENTS |
|---|
The authors thank Dr. Tao Wang from Otsuka America Pharmaceutical Incorporation for supplying the RLP-C kits and Assistant Professor Frank van Bockxmeer for the apoE genotyping. We thank the nursing staff of the Clinical Research Studies Unit of the University Department of Medicine (Royal Perth Hospital, University of Western Australia) for providing expert clinical assistance.
Manuscript received July 3, 2001 and in revised form January 11, 2002.
| REFERENCES |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Journal of Biological Chemistry |
| Molecular and Cellular Proteomics | ASBMB Today |