Vascular Biology |
From the Departments of Endocrinology and Diabetes (M.A.M.d.B., P.J.V., J.A.R.), General Internal Medicine (J.F.P.B., L.M.H., P.C.N.R.), and Cardiology (L.M.H.), Leiden University Medical Center, Leiden; Netherlands Organization for Applied Scientific Research-Quality of Life (M.A.M.d.B., J.F.P.B., P.J.V., L.M.H., P.C.N.R.), Leiden; Department of Infectious Diseases, Tropical Medicine, and AIDS (P.R., M.v.d.V.), Amsterdam Medical Center, Amsterdam; and Laboratory of Pediatrics (F.K.), Center for Liver, Digestive, and Metabolic Diseases, University Hospital Groningen, Groningen, the Netherlands.
Correspondence to Marion A.M. den Boer, Department of Endocrinology and Diabetes, LUMC, PO Box 9600, 2300 RC Leiden, the Netherlands. E-mail A.M.den_Boer{at}lumc.nl
| Abstract |
|---|
|
|
|---|
Methods and Results Feeding female APOE*3-Leiden transgenic mice a Western-type diet supplemented with RTV (35 mg/kg per day) for 2 weeks resulted in a 2-fold increase in fasting plasma triglyceride (TG) levels, which was specific for very lowdensity lipoprotein (VLDL). RTV did not change the hepatic VLDL-TG production. Instead, RTV did increase the postprandial TG response to an oral fat load (area under the curve, 25.5±12.1 versus 13.8±6.8 mmol/L per hour in controls; P<0.05). Likewise, RTV hampered the plasma clearance of intravenously injected glycerol tri[3H]oleate-labeled VLDLlike emulsion particles (half time, 19.3±10.5 versus 5.0±1.3 minutes in controls; P<0.05) associated with a decrease of 44% in plasma lipoprotein lipase activity. Accordingly, RTV decreased the uptake of TG-derived fatty acids (FAs) into adipose tissue, as well as the uptake of albumin-bound FA.
Conclusions We conclude that RTV causes hypertriglyceridemia via decreased lipoprotein lipasemediated clearance of VLDL-TG. In addition, RTV specifically impairs the uptake of FA in adipose tissue, which may contribute to the lipodystrophy that is frequently observed in HIV-infected subjects on antiretroviral therapy.
We investigated the mechanism underlying the hypertriglyceridemia associated with ritonavir treatment in APOE*3-Leiden mice. Ritonavir increased the postprandial triglyceride response and decreased the clearance of glycerol tri[3H]oleatelabeled very low density lipoprotein-like emulsion particles, as related to decreased plasma lipoprotein lipase activity. These effects were accompanied by decreased fatty acid flux into adipose tissue.
Key Words: dyslipidemia protease inhibitor transgenic mice FA metabolism TG metabolism
| Introduction |
|---|
|
|
|---|
We used the APOE*3-Leiden transgenic mouse as an experimental model, because these mice have a humanized lipoprotein profile and are susceptible to diet- and drug-induced hyperlipidemia, obesity, and atherosclerosis.1315 In contrast to wild-type mice, APOE*3-Leiden transgenic mice are highly sensitive to treatment with hypolipidemic drugs, such as statins, fibrates, and peroxisome proliferator-activated receptor
(PPAR
) and PPAR
-agonists.16 Similar to humans, APOE*3-Leiden transgenic mice have a much lower clearance rate of VLDL-TG than wild-type mice. As a consequence, APOE*3-Leiden mice represent a suitable animal model for RTV-associated hyperlipidemia.
See page 7
The first aim of the present study was to assess the effects of RTV on both VLDL-TG production and clearance rates. We used a low dosage of RTV that induced hypertriglyceridemia without causing toxicity, as measured by plasma alanine amino transferase levels. The second aim was to evaluate the effects of RTV on tissue-specific uptake of fatty acids (FAs) derived from VLDL-TG and from the plasma-free FA pool by applying our recently described method using differentially labeled FA to quantify tissue-specific uptake of FA derived from VLDL-TG and from plasma-free FA.17 We found that RTV decreased the clearance of VLDL-TG from plasma by decreasing lipoprotein lipase (LPL) activity and decreased the uptake of FA derived from VLDL-TG and of albumin-bound FA in adipose tissue but not in other organs.
| Methods |
|---|
|
|
|---|
Animals and Experiments
Female APOE*3-Leiden transgenic mice, housed under standard conditions with free access to water and food, were fed a semisynthetic Western type diet (Hope Farms) with or without the addition of RTV (Norvir, Abbott) at a concentration of 35 mg/kg of body weight per day for 2 weeks. Plasma levels of TG, total cholesterol, and free FA were determined using commercially available enzymatic assays. FPLC analysis was performed on pooled plasma to determine the distribution of TG and cholesterol over the lipoprotein fractions using the AKTA purifier supplied with a Superose-6 column (Amersham Pharmacia Biotech). Hepatic VLDL-TG production was determined after injection of Triton WR1339, which completely blocks plasma VLDL-TG clearance.18 The postprandial TG response was determined after the administration of an intragastric olive oil bolus. Glycerol tri[3H]oleate-labeled, 80-nm-sized, protein-free VLDLlike emulsion particles, which are known to mimic endogenous VLDL-TG particles,19 were intravenously injected into mice to study the in vivo serum clearance. Plasma from heparin-injected mice was used to determine plasma total LPL levels in vitro as modified from Zechner.20 Modulated lipolytic activity was determined by incubation of postheparin plasma with [3H]triolein-labeled 80-nmsized VLDLlike emulsion particles in vitro. To determine the effect of RTV on the uptake of FA from VLDL-TG by peripheral tissues we used a steady-state approach by continuously infusing glycerol tri[3H]oleatelabeled emulsion particles and [14C]oleate bound to albumin.17
Statistical Analysis
The differences between experimental groups were determined by the MannWhitney U test for 2 independent samples. The level of statistical significance of the differences was set at P<0.05. Analyses were performed using SPSS 12.0 for Windows software (SPSS, Inc).
| Results |
|---|
|
|
|---|
|
RTV Does Not Change In Vivo VLDL-TG Production
To investigate whether the increase in plasma TG levels was due to increased hepatic VLDL-TG production, we injected fasted mice with Triton WR 1339, which completely inhibits lipolysis of VLDL-TG. However, as is shown in Figure IIA (available online at http://atvb.ahajournals.org), after 2 weeks of dietary RTV administration, no significant difference was observed in the rate of VLDL-TG production when the RTV-treated mice were compared with the controls (139±41 versus 177±60 µmol TG/kg per hour).
RTV Increases Postprandial TG Response
Subsequently, we investigated whether the increase in postprandial plasma TG levels was caused by impaired postprandial clearance of TG. For this purpose, an intragastric bolus of olive oil was administered, and, subsequently, plasma TG levels were determined. Figure IIB shows that RTV treatment caused a 2-fold increment in the postprandial TG response on an intragastric olive oil administration (area under the curve, 25.5±12.1 versus 13.8±6.8 mmol/L per hour; P<0.05), which, indeed, suggests impaired TG clearance.
RTV Increases Plasma Half-Life of TG-Rich VLDLLike Emulsion Particles
To investigate whether the decreased clearance of TG indeed contributes to the hypertriglyceridemia observed in RTV-treated mice, mice were intravenously injected with glycerol tri[3H]oleate-labeled protein-free VLDLlike emulsion particles. These particles mimic the metabolic behavior of TG-rich lipoproteins.19,21 Because LPL is more abundantly expressed on the adipose tissue in the postprandial state compared with the fasted state,22 we used fed mice for this study. As is shown in Figure 2, the clearance of glycerol tri[3H]oleate was markedly decreased in RTV-treated mice when compared with the control group, which is evident from an &4-fold increase in serum half-life of glycerol tri[3H]oleate (half time, 19.3±10.5 versus 5.0±1.3 minutes; P<0.05).
|
RTV Decreases Total LPL Activity in Postheparin Plasma
Impaired LPL-mediated TG hydrolysis can be due to decreased expression of LPL and/or by a direct effect of RTV on LPL activity. Therefore, we determined the effect of RTV on the total lipolytic activity in postheparin plasma by incubation with a glycerol tri[3H]oleatecontaining substrate mixture. As shown in Figure 3A, the postheparin hepatic lipase (HL) activity in RTV-treated mice did not differ significantly from that of control mice (15.1±3.7 versus 12.5±3.7 µmol FA/hr per milliliter). The postheparin LPL activity, however, was significantly decreased by 44% in RTV-treated mice versus control mice (11.2±3.3 versus 19.9±11.1 µmol FA/hr per milliliter; P<0.05). This observation shows that RTV impairs LPL-mediated TG lipolysis by lowering the total LPL activity present in plasma.
|
RTV Decreases the Modulated Lipolytic Activity in Postheparin Plasma
To study the modulated lipolytic activity in plasma by allowing interference of the endogenous activators [eg, apolipoprotein (apo)CII] and inhibitors (eg, apoCI and apoCIII) with the activity of LPL, we performed an additional assay in which the lipolytic activity of plasma was determined toward a relatively low amount of well-defined emulsion particles instead of an excess of solubilized TG. As is shown in Figure 3B, the postheparin modulated lipolytic activity is decreased significantly by 55% in the plasma of RTV-treated mice as compared with control mice (19.0±3.7 versus 42.8±12.7 nmol of free FA/hr per milliliter; P<0.05).
RTV Decreases FA Uptake in Adipose Tissue
The effect of RTV on the uptake of FA from VLDL-TG and albumin-bound FA by various tissues was studied during steady state infusion of glycerol tri[3H]oleate TGrich, VLDL-like emulsion particles. RTV treatment did not affect VLDL-TGderived FA uptake by the liver, skeletal muscle, and the heart (Figure 4A). In adipose tissue, however, the uptake of VLDL-TGderived FA was significantly decreased (639±220 versus 986±80 nmol FA/mg tissue protein; P<0.05). The uptake of FA bound to albumin was also decreased in adipose tissue of RTV-treated mice (514±176 versus 1078±194 nmol FA/mg tissue protein; P<0.05) and not in the liver, skeletal muscle, and the heart when compared with control mice (Figure 4B).
|
| Discussion |
|---|
|
|
|---|
Human studies remain inconclusive with respect to the underlying mechanism of RTV-induced hypertriglyceridemia.311 Purnell et al23 showed that RTV decreased hepatic lipase activity, although there was no difference in postheparin LPL levels between RTV- and placebo-treated healthy subjects. In contrast, a study by Baril et al3 showed that RTV caused decreased LPL activity, whereas no differences in the amount of apoCII (cofactor for LPL) or apoCIII (inhibitor of LPL) were found, indicating a direct effect of RTV on the LPL enzyme as we now conclusively show in our study. Shahmanesh et al10 showed a significant decrease in the fractional catabolic rate of VLDL-TG in individuals treated with RTV either alone or in combination with other antiretroviral drugs because of a decreased activity of LPL even in the postabsorptive state. Another study in HIV-negative subjects treated with RTV showed a trend toward decreased fat clearance as measured by an intravenous fat tolerance test after a 10-hour fast.5 A recent study by Sekhar et al9 revealed marked abnormalities in the ability of HIV lipodystrophy patients to metabolize dietary TG, suggesting an impairment of the function of LPL. In humans, it is impossible to conclusively show the direct effects of the individual drugs on the lipid metabolism, because HAART-treated patients are usually on a therapy regimen of
3 drugs. Moreover, in humans there is considerable heterogeneity in both environmental and genetic background.
To conclusively determine the mechanism underlying RTV-induced hypertriglyceridemia, we used the APOE*3-Leiden transgenic mouse as our model. Studies in AKR/J mice24 and in C57BL/6 wild-type mice25 showed an effect of RTV only on the hepatic VLDL-TG production rate. In contrast to AKR/J and wild-type mice, the APOE*3-Leiden transgenic mouse has a lipoprotein profile with close resemblance to the human profile.1315 In these mice, plasma cholesterol levels can be titrated to any desired level by varying the amount of cholesterol in the diet. In contrast to wild-type mice, APOE*3-Leiden transgenic mice are highly sensitive to treatment with hypolipidemic drugs, such as statins, fibrates, and PPAR-
and PPAR-
agonists.16 These observations imply that the APOE*3-Leiden transgenic mice on a Western type diet represent a suitable animal model for hyperlipidemia.
An in vitro study in human and rat hepatoma cells and primary hepatocytes from mice showed that protease inhibitor treatment inhibits proteasomal degradation of nascent apoB.26 However, protease inhibitors also inhibited secretion of apoB. The concentrations of drugs used in these in vitro studies are much higher than the maximal plasma concentrations in subjects taking these drugs.27 RTV may affect different components of the lipid metabolism depending on the dosage used. The dosage we used in our mice was 2 times higher than what an average adult would receive per kilogram per day. Taking into account the much faster metabolic rate in mice, it is clear that we used a low-physiological dosage in our mice. Unfortunately, we did not have the opportunity to assess plasma RTV concentrations. It may be that at superphysiological concentrations RTV affects the VLDL-TG production rate as well.
In the present study, RTV impaired FA uptake in adipose tissue under steady-state conditions while infusing glycerol tri[3H]oleate-labeled VLDLlike particles together with albumin-bound 14Clabeled FA. Before tissues can take up FA derived from VLDL-TG, these TGs have to be lipolyzed by LPL. In the current study, we show that RTV decreased plasma LPL activity by 44%. As expected, because of decreased LPL activity, the adipose tissue of RTV-treated mice took up significantly less FA derived from VLDL-TG compared with control mice under fed conditions. In the fed state, LPL is more abundant in adipose tissue than in muscle,17,22 explaining why no change is seen in the uptake of VLDL-TGderived FA in muscle. In addition to decreased uptake of FA derived from VLDL-TG, the adipose tissue of RTV-treated mice also took up less albumin-bound FA, a process independent of LPL. The active transport of FA into tissues occurs mainly via CD36. CD36 functions as a high-affinity transporter of long-chain FA in adipose tissue and the muscle.28,29 Serghides et al30 have shown that CD36 deficiency was induced by antiretroviral therapy both in healthy humans and in HIV-infected subjects. They also showed that RTV significantly decreased CD36 levels in THP1 and C32 cells. The observed decrease in the uptake of albumin-bound FA in adipose tissue as we observed is in accordance with a decrease in CD36 levels. Another study showed that in murine peritoneal macrophages CD36 can be upregulated by protease inhibitor therapy leading to increased uptake of cholesterol and cholesteryl esters.31 The difference in outcome of these studies may be a matter of different concentrations that are used in the in vitro studies. Many protease inhibitors, especially RTV, are very poorly soluble and difficult to handle in an in vitro assay.32 Alternatively, it may be that the same drug exerts different effects in different types of cells.
In accordance with decreased FA uptake by peripheral tissues, we found an increase of &16% in plasma FA levels in RTV-treated mice. As we have shown recently,33 increased plasma FA levels can directly impair LPL activity most probably via product inhibition, because free FA can bind to the active site of LPL. In the present study, plasma free FA levels are slightly but significantly increased; therefore, in addition to direct impairment of LPL activity; RTV may also be contributing indirectly to decreased LPL-mediated lipolysis via increased plasma FA.
Lipodystrophic HAART-treated HIVinfected patients showed an increased postprandial TG and FA response compared with nonlipodystrophic HIVinfected patients and healthy controls most likely caused by inadequate trapping of FA into adipose tissue.34 Decreased postprandial adipose tissue FA uptake was already observed in our study after 2 weeks of drug administration, although no obvious lipodystrophy as measured by weighing fat pads was observed yet. The flux of FA to adipose tissue mediated by LPL is an important determinant of adipogenesis. Deletion of LPL in adipose tissue in leptin-deficient ob/ob mice has been shown to prevent excessive storage of TG in the adipose tissue.35 In contrast, the absence of apoCIII, the natural LPL inhibitor, enhances FA uptake from plasma TGs in adipose tissue, which leads to higher susceptibility to diet-induced obesity.36 In mice that were administered RTV for a much longer period, generalized lipoatrophy was shown in male mice, whereas this lipodystrophy was restricted to the gonadal depot in female mice.37 The investigators proposed that the lipodystrophy in these mice is caused, at least in part, by reduced PPAR-
function. PPAR-
transcriptionally activates a number of genes that are essential for adipogenesis, lipid storage, and metabolism, including CD36.
The cause of the HAART-associated hypertriglyceridemia, as observed in humans, may be multifactorial in nature because of the use of different protease inhibitors simultaneously in combination with antiretroviral drugs of other classes. We propose that the main mechanism by which RTV increases plasma TG is by decreasing the LPL-mediated clearance of TG-rich lipoproteins. In the present study, we directly show that RTV decreases the uptake of VLDL-TGderived FA and albumin-bound FA specifically in adipose tissue, an effect that may well contribute to HAART-associated lipodystrophy.
| Acknowledgments |
|---|
Received August 23, 2005; accepted October 20, 2005.
| References |
|---|
|
|
|---|
Related Article:
This article has been cited by other articles:
![]() |
M. J. Carper, W. T. Cade, M. Cam, S. Zhang, A. Shalev, K. E. Yarasheski, and S. Ramanadham HIV-protease inhibitors induce expression of suppressor of cytokine signaling-1 in insulin-sensitive tissues and promote insulin resistance and type 2 diabetes mellitus Am J Physiol Endocrinol Metab, March 1, 2008; 294(3): E558 - E567. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. T. Rodgers and P. Puigserver Fasting-dependent glucose and lipid metabolic response through hepatic sirtuin 1 PNAS, July 31, 2007; 104(31): 12861 - 12866. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Zhou, E. C. Gurley, S. Jarujaron, H. Ding, Y. Fang, Z. Xu, W. M. Pandak Jr., and P. B. Hylemon HIV protease inhibitors activate the unfolded protein response and disrupt lipid metabolism in primary hepatocytes Am J Physiol Gastrointest Liver Physiol, December 1, 2006; 291(6): G1071 - G1080. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Guardiola, R. Ferre, J. Salazar, C. Alonso-Villaverde, B. Coll, S. Parra, L. Masana, and J. Ribalta Protease Inhibitor-Associated Dyslipidemia in HIV-Infected Patients Is Strongly Influenced by the APOA5-1131T->C Gene Variation Clin. Chem., October 1, 2006; 52(10): 1914 - 1919. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. G. Morse and J. A. Kovacs Metabolic and skeletal complications of HIV infection: the price of success. JAMA, August 16, 2006; 296(7): 844 - 854. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. A. Spector HIV Protease Inhibitors and Hyperlipidemia: A Fatty Acid Connection Arterioscler. Thromb. Vasc. Biol., January 1, 2006; 26(1): 7 - 9. [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
ATVB Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2006 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |