Highly active antiretroviral therapy (HAART) has proven to be effective at reducing morbidity and mortality in HIV-infected individuals. However, it is recognized that up to 80% of patients receiving this treatment develop metabolic alterations, which include dyslipidaemia, insulin resistance, central adiposity and peripheral lipoatrophy . Emergence of lipodystrophy has been correlated temporally with the widespread use of protease inhibitors (PI). The deleterious effects of PI are difficult to investigate in patients receiving several classes of antiviral drug; therefore, in vitro models are required. Studies with these models have suggested that PI have a direct effect on the formation of adipocytes. However, conflicting reports on adipogenesis have been reported for a given drug [2,3]. These discrepancies underline the importance of the cellular model that is used. The molecular mechanisms of adipocyte differentiation that are inhibited by PI are unclear. Both extracellular and intracellular targets of PI have been proposed. It has been reported that indinavir modified the extracellular matrix integrity of 3T3-L1 preadipocytes by disrupting the activity of the matrix metalloproteases MMP-2 and MMP-9 . Modulation of the expression of SREBP1, a key transcription factor for adipocyte differentiation, has been observed in mouse adipose tissue after ritonavir (RTV) treatment , but evidence that PI enter adipocytes is still lacking.
This study investigates the effects of six PI drugs on adipocyte differentiation in several cell lines. A novel competitive enzyme-linked immunosorbent assay (ELISA)  was used to analyse uptake and efflux of RTV during the differentiation of drug-sensitive and drug-resistant preadipocytes.
Materials and methods
Nelfinavir, indinavir (IDV), RTV and saquinavir were obtained through the AIDS Research and Reference Reagent Program, National Institutes of Health (NIH). Amprenavir was kindly provided by GlaxoSmithKline Research and Development (Ware, UK). Large amounts of RTV, lopinavir, IDV and saquinavir were obtained by extraction from commercially available tablets and capsules. The purity was assessed by 1H and 13C nuclear magnetic resonance and mass spectroscopy. The PI drugs obtained from the NIH were used as reference products for purification and to validate the results obtained with drugs purified from tablets. Each PI was dissolved in dimethyl sulfoxide and diluted 1:1000 in culture media.
Cell lines and adipocyte differentiation
3T3-L1  and Ob1771  preadipocytes were maintained in DMEM (Dulbecco's minimum Eagle's medium; Gibco BRL, Life Technologies, Gaithersburg, Maryland, USA) containing 8% fetal calf serum (growth medium). Differentiation was induced by incubating postconfluent cells (day 0) for 48 h in growth medium supplemented with 0.1 μmol/l dexamethasone, 0.1 mmol/l 3-isobutyl-1-methylxanthine and 0.17 μmol/l insulin. Cells were then incubated in growth medium supplemented with insulin (1.7 μmol/l for 3T3-L1 or 0.17 μmol/l for Ob1771) and 1 μmol/l ciglitazone (BioMol, Plymouth Meeting, Pennsylvania, USA). PI were added at day 0 and maintained thereafter by changing the media every other day. 3T3-F442  preadipocytes were differentiated by incubating postconfluent cells (day 0) in growth medium supplemented with 2 nmol/l triiodothyronine, 0.17 μmol/l insulin, 1 μmol/l ciglitazone and treated with PI as indicated above.
The experimental protocol used for the differentiation of embryonic stem cells into adipocytes has been described . PI were added at day 7 and maintained thereafter.
RNA was prepared 10 days after addition of PI and 20 μg total RNA was subjected to Northern blot analysis as previously described . The blot was hybridized with a probe revealing the expression of the gene encoding the adipocyte fatty acid-binding protein (a-FABP) as an adipocyte-specific gene . Quantification of the hybridization signal was performed using a PhosphorImager (Fujix-Bas1000, Tokyo, Japan) coupled to the MacBas ver2.x bio-imaging analyser (Fujix-Bas).
Competitive enzyme immunoassay
Cells were washed three times with cold phosphate buffer saline, which has been shown to be efficient in removing extracellular PI . Cells were then scraped with 100% methanol. After 1 h at −20°C, the insoluble material was removed by centrifugation and the liquid phase was dried. Dried residues were dissolved in the assay buffer for 1 h and RTV was quantified by competitive ELISA as previously described . This technique was approximately 15 times more sensitive than the high performance liquid chromatography–ultraviolet method . Results were expressed as nanomoles RTV/106 cells.
The effects of pharmacological concentrations of IDV, RTV, amprenavir, saquinavir and lopinavir on the differentiation of 3T3-L1, 3T3-F442A, Ob1771 and embryonic stem cells was investigated. Adipocyte differentiation was analysed by quantification of the expression of a-FABP. Lopinavir and nelfinavir inhibited adipocyte differentiation of Ob1771 and 3T3-L1 preadipocytes whereas treatment with amprenavir did not produce any significant effect (Table 1). Interestingly, treatment with saquinavir, IDV or RTV led to different effects depending on the cell lines. Adipocyte differentiation of 3T3-L1 and 3T3-F442A cells was inhibited by RTV and IDV as previously described [2,13,14]. RTV did not have a cytotoxic effect and cells appeared to be arrested at the preadipocyte stage (data not shown). In contrast, Ob1771 and embryonic stem cells were unaffected by IDV and RTV, as indicated by a significant lipid droplets formation and the expression of the gene for a-FABP. Altogether, these results demonstrated that responsiveness of preadipocytes to PI differed from cell model to cell model. As the molecular mechanisms leading to inhibition or maintenance of adipocyte differentiation from PI remained unknown, the accumulation of PI by the cells was investigated. The ELISA protocol was adapted to determine the level of cell-associated RTV. Treatment of 3T3-L1 preadipocytes with RTV for 8 h led to dose-dependent accumulation of the drug (Fig. 1a). Interestingly, RTV was also able to accumulate into adipocytes with a similar efficiency.
Cell-associated RTV was examined during the differentiation of the RTV-resistant Ob1771 and the RTV-sensitive 3T3-L1 preadipocytes. After 20 min incubation, the level of cell-associated RTV was 9 nmoles/106 cells for Ob1771 and 5 nmoles/106 cells for 3T3-L1. After 20 h, similar levels of RTV were associated with 3T3-L1 cells and Ob1771 cells (Fig. 1b). These results indicate that the resistance of Ob1771 cells to RTV was not a consequence of their inability to take up the drug.
RTV efflux was also investigated from preadipocytes; this occurred rapidly, with only 20–30% of PI remaining associated with Ob1771 and 3T3-L1 cells after 20 min (Fig. 1c). Uptake, cell-association and efflux of RTV were similar in preadipocytes regardless of their sensitivity to the drug.
The anti-adipogenic effect of different PI was investigated on the most commonly used murine preadipocyte clonal lines. Amprenavir, nelfinavir and lopinavir had the same effect with cells from different origin. Moreover IDV, RTV and saquinavir prevented the appearance of lipid droplets and inhibited the differentiation of 3T3-L1 and 3T3-F442A cells but not that of Ob1771 and embryonic stem cells. Several hypotheses could explain the difference in sensitivity to RTV between 3T3-L1 and Ob1771 cells. The ability of cultured adipocytes to accumulate PI was also examined. To our knowledge, this is the first evidence showing that cultured adipocytes are rapidly able to accumulate and subsequently release a PI. In the experimental protocol used, 1–2% of PI in the external medium was found to be associated with the cell compartment ((Fig. 1a). It has been reported that 99% of RTV in plasma is bound to proteins . The percentage of unbound and bound RTV in our culture conditions remains to be determined, but these data strongly suggest that a high percentage of free RTV accumulated into preadipocytes. Like blood cells, adipocytes are able to accumulate RTV but the clinical relevance of this observation (i.e., that adipose tissue can indeed accumulate PI) has still to be determined. The molecular mechanisms of uptake and efflux of RTV in preadipocytes and adipocytes are also unknown but we cannot exclude the participation of the P-glycoprotein transporters and multidrug resistant-associated proteins.
Overall, our data indicate that the lack of effect of RTV on adipocyte differentiation cannot be explained by its inability to enter the cells. The differential effect of RTV in drug-sensitive and drug-resistant cells could be a result of differential expression of intracellular binding proteins, leading to variation in subcellular localization and access or not to a target protein. Detailed characterization of drug-sensitive and drug-resistant cells will be important for the identification of molecular events required to mediate the inhibitory effect of RTV on adipogenesis.
We are grateful to Drs C. Frelin, E. Z. Amri, P. Peraldi and B. Phillips for critical reading of the manuscript. We are also indebted to M. C. Nevers for the preparation of RTV-tracer.
Sponsorship: This work was supported in part by grants from the French National Agency for AIDS research (grant ANRS to CD) and ‘Ensemble contre le Sida'. C. Vernochet is a recipient of an ANRS fellowship.
1. Saves MF, Raffi J Capeau W, Rozenbaum JM, Ragnaud C, Perronne A, et al. Factors related to lipodystrophy and metabolic alterations in patients with human immunodeficiency virus infection receiving highly active antiretroviral therapy. Clin Infect Dis
2. Zhang B, MacNaul K, Szalkowski D, Li Z, Berger J, Moller DE. Inhibition of adipocyte differentiation by HIV protease inhibitors. J Clin Endocrinol Metab
3. Nguyen AT, Gagnon A, Angel JB, Sorisky A. Ritonavir increases the level of active ADD-1/SREBP-1 protein during adipogenesis. AIDS
4. Mondal D, Larussa VF, Agrawal KC. Synergistic antiadipogenic effects of HIV type 1 protease inhibitors with tumor necrosis factor alpha: suppression of extracellular insulin action mediated by extracellular matrix-degrading proteases. AIDS Res Hum Retroviruses
5. Riddle TM, Kuhel DG, Woollett LA, Fichtenbaum CJ, Hui DY. HIV protease inhibitor induces fatty acid and sterol biosynthesis in liver and adipose tissues due to the accumulation of activated sterol regulatory element-binding proteins in the nucleus. J Biol Chem
6. Akeb F, Ferrua B, Creminon C, Roptin C, Grassi J, Nevers MC, et al. Quantification of plasma and intracellular levels of the HIV protease inhibitor ritonavir by competitive ELISA. J Immunol Meth
7. Green H, Kehinde O, Spontaneous heritable changes leading to increased adipose conversion in 3T3 cells. Cell
8. Doglio A, Dani C, Grimaldi P, Ailhaud G. Growth hormone regulation of the expression of differentiation-dependent genes in preadipocyte Ob1771 cells. Biochem J
9. Antras-Ferry J, Franckhauser S, Robin D, Robin P, Granner DK, Forest C. Expression of the phosphoenolpyruvate carboxykinase gene in 3T3-F442A adipose cells: effects of retinoic acid and differentiation. Biochem J
10. Dani C, Smith AG, Dessolin S, Leroy P, Staccini L, Villageois P, et al. Differentiation of embryonic stem cells into adipocytes in vitro. J Cell Sci
11. Reese-Wagoner A, Thompson J, Banaszak L. Structural properties of the adipocyte lipid binding protein. Biochim Biophys Acta
12. Khoo SH, Hoggard PG, Williams I, Meaden ER, Newton P, Wilkins EG, et al. Intracellular accumulation of human immunodeficiency virus protease inhibitors. Antimicrob Agents Chemother
13. Roche R, Poizot-Martin I, Yazidi CM, Compe E, Gastaut JA, Torresani J, Planells R. Effects of antiretroviral drug combinations on the differentiation of adipocytes. AIDS
14. Caron M, Auclair M, Vigouroux C, Glorian M, Forest C, Capeau J. The HIV protease inhibitor indinavir impairs sterol regulatory element-binding protein-1 intranuclear localization, inhibits preadipocyte differentiation, and induces insulin resistance. Diabetes
15. Acosta EP, Kakuda TN, Brundage RC, Anderson PL, Fletcher CV. Pharmacodynamics of human immunodeficiency virus type 1 protease inhibitors. Clin Infect Dis
(suppl 2): S151–S159.