Metabolic alterations such as dyslipidemia, impaired insulin sensitivity, and, in some cases, overt lipodystrophy with abnormal distribution of body fat are common in HIV-1–infected patients undergoing highly active antiretroviral treatment (HAART).1 Despite the presence of overt alterations is magnified in HIV-1–infected patients undergoing HAART, initial signs of altered metabolism, such as mild dyslipidemia or evidence of local inflammation in adipose tissue, have been reported in HIV-1–infected patients before starting HAART.2 The relative involvement of HIV-1 infection–related events and HAART-specific patterns, and the molecular mechanisms ultimately responsible for these metabolic alterations are poorly understood.
Fibroblast growth factors (FGFs) constitute a family of regulatory proteins, most of which act as autocrine/paracrine factors and cause their biological effects in the same cells that secrete them or in closely surrounding cells. However, members of a subfamily of FGF proteins composed of FGF19, FGF21, and FGF23—endocrine FGFs—are known to function as hormonal regulators, acting at a distance from secreting cells.3 FGF23 is produced by bone and regulates phosphate and vitamin D homeostasis, whereas FGF19 and FGF21 are involved in multiple aspects of metabolic regulation. FGF21 is produced mainly by the liver and has been shown to correct glucose intolerance and hyperlipidemia in rodent models of obesity and type 2 diabetes.4 FGF19 is secreted by the small intestine—mainly the ileum region—and acts on the liver to control the biosynthesis of bile acids. In addition to these effects, experimental data in rodents indicate that FGF19 has important effects that favor a healthy metabolic profile, enhancing energy expenditure in general and specifically promoting glucose disposal.5 Some studies have reported that FGF19 impairs insulin-induced fatty acid synthesis in the liver,6 whereas other data suggested that most of the effects of FGF19 on metabolism occur through tissues other than liver, probably adipose tissue.7 Recently, FGF19 has been reported to act as a postprandial insulin-independent activator of hepatic protein and glycogen synthesis, being able to restore the loss of hepatic glycogen in diabetic animals lacking insulin.8 The metabolic effects of FGF19 and FGF21 are mediated through fibroblast growth factor receptors (FGFRs), predominantly FGFR4, which is expressed in liver, and FGFR1, which is mostly expressed in white adipose tissue.9,10 Both receptors must interact with the membrane protein β-Klotho to form an FGF19/21-responsive receptor complex. Thus, β-Klotho seems to be an obligatory coreceptor for FGF19 and FGF21 responsiveness.3,9
In a recent study, we reported that the levels of FGF21 in serum were abnormally elevated in HIV-1–infected patients, especially those that developed lipodystrophy and metabolic syndrome after HAART.11 This paradoxical observation is similar to findings in obese/diabetic patients, in which FGF21 levels are also increased.12,13 It has been proposed that obesity constitute a “FGF21-resistant” state14; and this may also occur in HIV-1–infected patients with lipodystrophy. However, it is not known whether FGF19 is altered in HIV-1–infected patients in association with their metabolic disturbances, and, hence, whether the FGF19 may be considered a potential target of treatment strategies for the metabolic syndrome in these patients is not known. In the present study, we compared FGF19 levels in HIV-1–infected patients under distinct conditions of HAART treatment and lipodystrophy with those in healthy individuals and made a parallel assessment of FGF21 levels. We also determined the levels of FGFR1 and β-Klotho coreceptor gene expression in patient adipose tissue biopsies to approach indications of tissue sensitivity to endocrine FGFs in HIV-1–infected patients.
All patients and controls provided informed written consent to participate in the study. The study was approved by the ethics committee of Hospital de la Santa Creu i Sant Pau, Barcelona, Spain. Patients with opportunistic infections, neoplasms, or fever of undetermined origin were excluded from the study. At the time of the study entry, no patient or control used any other drug known to influence glucose metabolism or fat distribution, such as anabolic hormones or systemic corticosteroids, uridine, recombinant human growth hormone, or appetite stimulants or suppressors. Patients with serologic evidence of infection by hepatitis viruses (B and C) were excluded from the study. For hepatitis B virus infection exclusion, a negative serum hepatitis B virus DNA was also required. No control or patient had undergone previous cholecystectomy. Patient demographics, anthropometric and metabolic data, HIV-1 infection parameters, and antiretroviral treatment data are shown in Table 1. Body mass index (BMI) was calculated, and waist circumference was measured to the nearest millimeter using anatomical landmarks, as defined by the Third National Health and Nutrition Evaluation Survey.15 Whole-body dual-energy X-ray absorptiometry scans (Hologic QDR-4500A Hologic, INc, Waltham, MA) were conducted by a single operator. The percentage of fat in the arms, legs, and central abdomen (calculated from the mass of fat vs lean and bone mass) and total lean body mass (in kilograms) were recorded.
Lipodystrophy was assessed using a lipodystrophy severity grading scale based on that reported by Lichtenstein,16 and a clinical diagnosis of lipodystrophy was assigned to patients with overall scores >7.
Sampling and Analytical Methodology
Plasma and serum were obtained from blood drawn from seated patients after a 12-hour overnight fast and at least 15 minutes after the placement of a peripheral intravenous catheter. All lipid measurements were performed using a Hitachi 911 system (Roche Diagnostic Systems, Basel, Switzerland). Serum triglycerides were measured by a fully enzymatic standard method; high-density lipoprotein (HDL) cholesterol was measured by a direct method using polyethylene glycol–modified enzymes (PEGME).17 Low-density lipoprotein cholesterol was measured after ultracentrifugation according to the method recommended by the Lipid Research Clinic. Insulin resistance was estimated by the homeostasis model assessment of insulin resistance (HOMA-IR) method.18
Plasma HIV-1 viral load was measured using the Amplicor HIV-1 Monitor assay (Roche Diagnostic Systems), which has a lower detection limit of 20 copies per milliliter.
Serum FGF19 and FGF21 levels were determined using non–cross-reactive enzyme-linked immunosorbent assays specific for the corresponding human proteins (Biovendor, Heidelberg, Germany). Serum FGF19 data exhibited a normal distribution in patients and control population, whereas the distribution of serum FGF21 data distribution was skewed and was thus log transformed before analysis, as described previously.11
Biopsy samples of subcutaneous fat from 10 patients from each patient group and 10 healthy controls were taken from the abdominal area. After homogenization in RNeasy lysis buffer (Qiagen, Hilden, Germany), RNA was isolated using a column affinity-based methodology that included on-column DNA digestion (RNeasy; Qiagen). One microgram of RNA was transcribed into cDNA using MultiScribe reverse transcriptase (RT) and random hexamer primers (TaqMan Reverse Transcription Reagents; Applied Biosystems, Foster City, CA). For quantitative mRNA expression analysis, TaqMan RT-polymerase chain reaction (PCR) was performed on the ABI PRISM 7700HT sequence detection system (Applied Biosystems). The TaqMan RT-PCR reactions were performed in a final volume of 25 μL using TaqMan Universal PCR Master Mix, No AmpErase UNG reagent, and primer pair probes specific for FGF21 (Hs00173927_m1), FGF19 (Hs00192780_m1), FGFR1 (Hs00222484_m1), β-Klotho (Hs00545621_m1), lactase-like protein (Lctl, Hs01385107_m1), and 18S rRNA (Hs99999901). Controls with no RNA, primers, or RT were included in each set of experiments. Each sample was run in duplicate, and the mean value of the duplicate was used to calculate the mRNA levels for the genes of interest. Values were normalized to that of the reference control (18S rRNA) using the comparative 2−ΔCT method, following the manufacturer instructions. Parallel calculations performed using the reference gene PPIA (Hs99999904) yielded essentially the same results.
Data were expressed as means ± standard error of the mean, frequencies, or percentages relative to healthy controls (defined as 100%). The normality of parameter distributions was determined using a Kolmogorov–Smirnov analysis. One-way analysis of variance and Tukey post hoc tests were performed for comparisons of parametric data. Regression analysis was used to determine the linear relationships of anthropometric, metabolic, and antiretroviral treatment variables with serum FGF19 quantitative parameters.
Statistical analyses were performed using the Statistical Package for Social Sciences version 17.0 (SPSS, Chicago, IL) and the SAS version 9.1.3 software (SAS Institute Inc, Cary, NC), The P values <0.05 (determined by 2-sided tests) were considered significant.
A total of 106 HIV-1–infected patients and 31 healthy controls were studied. Thirty-four of the HIV-1–infected patients had not received any antiretroviral treatment (naive), whereas 72 were receiving HAART. Thirty-eight of the HIV-1–infected patients undergoing HAART had been diagnosed as having lipodystrophy. There were no statistically significant differences between patients and controls with respect to sex distribution. Cumulative exposure to nucleoside reverse transcriptase inhibitors (NRTIs) and nonnucleoside reverse transcriptase inhibitors (NNRTIs) was significantly greater in patients with lipodystrophy than in those without lipodystrophy, but there were no significant differences with respect to protease inhibitors (PIs) exposure time (Table 1). Data on individual drug exposure profiles of HAART are shown in Table 2 and indicated that the only significant difference between lipodystrophy-positive and lipodystrophy-negative patients was a significantly higher duration of exposure to stavudine and didanosine in the lipodystrophy-positive group.
With respect to anthropometric parameters, there were no significant differences in BMI between either of the 2 groups of HAART, patients and controls; however, the waist-to-hip ratio was significantly higher in HAART patients. There was a small but significant decrease in BMI in “naive” patients. Among parameters indicative of circulating lipid homeostasis, triglyceridemia was markedly increased in HIV-1–infected HAART patients, whereas total cholesterol, HDL cholesterol, and low-density lipoprotein cholesterol were significantly lower in all the HIV-1–infected patient groups compared with healthy controls. Blood glucose levels were increased in both HIV-1–infected subsets, HAART patients compared with controls, and the increase in glycemia was significantly higher in those with lipodystrophy than in those without. The same pattern was observed for plasma insulin as follows: its levels were increased in HIV-1–infected individuals compared with HIV-1–uninfected individuals, the greatest increase occurring in patients with lipodystrophy. The HOMA-IR index was increased in both groups of HIV-1–infected HAART patients, with the greatest increase occurring in those with lipodystrophy.
Serum FGF19 levels in patients and healthy controls ranged from 26.1 pg/mL to 64.1 pg/mL. There were no significant differences (P = 0.89) in serum FGF19 levels between men (217.9 ± 12.1 pg/mL, n = 104) and women (214.8 ± 20.8 pg/mL, n = 33). Serum levels of FGF19 in all HIV-1–infected patient groups were significantly reduced relative to those in controls (Fig. 1). Among HIV-1–infected patients, those undergoing HAART and presenting lipodystrophy had the lowest FGF19 levels, nontreated patients had the highest levels, and HAART patients without lipodystrophy had intermediate levels; however, these differences among HIV-1–infected patient groups did not attain statistical significance. Serum FGF21 levels were significantly higher in all HIV-1–infected patient groups than in healthy controls (Fig. 1) in agreement with a previous report.11
We investigated the relationship between serum FGF19 levels and anthropometric parameters, metabolic data (including FGF21 levels), and parameters related to HIV-1 infection and HAART patterns (Table 2). Serum FGF19 levels correlated positively with HDL cholesterol levels and negatively with insulin levels and HOMA-IR. We found no significant correlation with parameters indicative of overall fat mass, such as BMI or fat percentage, or in relation to fat distribution, such as waist-to-hip ratio. In fact, when the overall patient population and controls were considered together, there were no statistically significant differences in serum FGF19 levels between lean (216.6 ± 12.3 pg/mL, BMI < 25 kg/m2, n = 93) and overweight (218.4 ± 19.6 pg/mL, BMI > 25 kg/m2, n = 44) subjects. Serum FGF19 levels correlated negatively with cumulative months of treatment with NRTI and NNRTI drugs but not with cumulative duration of PI exposure. When these correlations were investigated in relation to individual drug treatment patterns for NRTIs (Table 3), a statistically significant negative correlation was found for the extent of treatment with stavudine (R = −0.21, P = 0.015), lamivudine (R = −0.18, P = 0.037) and tenof,ovir (R = −0.20, P = 0.021). No significant correlation was found for cumulative treatment with other NRTI drugs present in the HAART regimes of the patients such as zidovudine (R = −0.02, P = 0.78), abacavir (R = −0.08, P = 0.36), or didanosine (R = −0.07; P = 0.39). For NNRTIs, only cumulative months of treatment with efavirenz showed a significant negative correlation with serum FGF19 levels (R = −0.19, P = 0.029). No significant correlation between serum FGF21 and serum FGF19 levels was found.
Given the profound—and opposite—alterations in the endocrine FGF system in HIV-1–infected patients, with the abnormally low serum levels of FGF19 found here and the high levels of FGF21 reported previously and confirmed here, we analyzed the expression of mRNA for receptors and coreceptors for these endocrine FGFs in adipose tissue from patients and controls, and the expression of the FGF19 and FGF21 mRNA. The FGF19 mRNA expression was undetectable in adipose tissue from all patient groups, as expected.19 In fact, FGF19 is known to be expressed mainly in the intestine, although under pathological conditions it may be expressed in ectopic sites such as liver.20 Similarly, little or no FGF21 mRNA expression was detected in subcutaneous adipose tissue from HIV-1–infected patients; FGF21 expression was also virtually undetectable in adipose tissue from healthy controls, in agreement with other previous reports.21 Expression of the mRNA of FGFR1, the main FGF receptor expressed in adipose tissue known to mediate FGF19 and FGF21 effects, was significantly reduced in untreated HIV-1–infected patients and in HIV-1–infected HAART patients with no lipodystrophy compared with controls (Fig. 2). There was no significant difference in FGFR1 gene expression in adipose tissue between HIV-1–infected HAART patients with lipodystrophy and controls. Next, we determined the mRNA expression levels for β-Klotho, the coreceptor of FGFR1 that determines the specificity of cell responsiveness to the FGF19 and FGF21 members of the FGF family. The results indicated a dramatic decrease in β-Klotho transcript expression in adipose tissue from all patient groups, a decrease that was specially marked in naive patients (Fig. 2). The mRNA levels of Lctl, a recently described potential coreceptor for FGF19 signaling,22 were negligible in human adipose tissue from both controls and patients. In light of the reported role of peroxisome proliferator–activated receptor-γ (PPAR-γ), the master regulator of adipogenesis, in β-Klotho gene regulation,23 and the known alterations in PPAR-γ gene expression in adipose tissue from HIV-1–infected patients,2 PPAR-γ mRNA levels were determined in the patient groups. Results indicated that PPAR-γ expression was significantly lowered in all patient groups (percent values respect to controls: 16% ± 2% in naive patients, P < 0.001; 36% ± 3% in lipodystrophy-negative patients, P = 0002; 14% ± 3% in lipodystrophy-positive patients, P < 0.0001), in agreement with previous reports.
The present study demonstrates that serum FGF19 levels are reduced in HIV-1–infected patients. The fact that FGF19 levels were negatively correlated with parameters indicative of insulin resistance, such as HOMA-IR, is consistent with the proposed role of FGF19 in promoting insulin sensitivity and glucose uptake in target tissues based on experimental animal studies.5,20 Although the cross-sectional nature of the present study does not allow cause-to-effect relationships, it may be speculated that the abnormally low levels of FGF19 in HIV-infected patients contribute to metabolic disturbances such as hyperglycemia and insulin resistance that occur frequently in these patients. The recent proposal indicating that pharmacological treatment of rodent models with FGF19 may boost certain effects of insulin on liver is also consistent with this scenario.8 On the other hand, the lack of association of FGF19 levels and BMI in the HIV-1–infected population is in contrast with a recently reported inverse association in obese population.24 It is likely that the narrow range of BMI in our HIV-1–infected patient population precluded detection of such association in contrast with the extremely wide range of BMI values in obesity studies.
Given the very limited knowledge about the regulation of FGF19 production and release, the mechanisms by which FGF19 levels are reduced in patients is a matter of speculation. The ileum is a main site of FGF19 production; in this part of the intestine, FGF19 production is induced by bile acids.20 There is evidence that chenodeoxycholic acid increases FGF19 levels in volunteers, whereas treatment with the bile acid–binding resin cholestyramine decreases serum FGF19.25 To our knowledge, only 1 published study has assessed bile acid levels in HIV-infected patients.26 This report describes distinct alterations in individual bile acid concentrations in HIV-1–infected patients, from a rise in the relative concentrations of lithocholic and taurocholic acids to a relative decrease in chenodeoxycholic acid. Further studies will be required to establish whether the reduction in the serum FGF19 levels in patients is a consequence of altered bile acid levels.
Despite the fact that levels of FGF19 were significantly reduced in patients before HAART, the present study demonstrated a negative correlation between FGF19 levels and the extent of the treatment of patients with NRTIs, particularly with stavudine, and with the NNRTI, efavirenz. Stavudine is one of the components of antiretroviral regimes that is more closely associated with adipose tissue alterations, including lipodystrophy.27 Although we found no correlation between FGF19 levels and specific indicators of lipodystrophy, the lowest levels of FGF19 in our patient groups were observed in patients with lipodystrophy. On the other hand, efavirenz has not traditionally been associated with the appearance of signs of metabolic syndrome or lipodystrophy symptoms. However, an extensive randomized study (ACTG A5142) recently reported that lipoatrophy was more common with efavirenz than with PIs when combined with stavudine or zidovudine.28
A major finding of the present study is the reduction in the expression of FGFR1 and β-Klotho—main mediators of FGF19 and FGF21 responsiveness in adipose tissue—in HIV-infected patients. This result—to our knowledge the first reported finding of altered gene expression for components of the endocrine FGF system responsiveness in human tissues—is strongly supportive of impaired responsiveness to endocrine FGFs. It is also consistent with the proposal that resistance to FGF21 accounts for the paradoxical increase of FGF21 in these patients in the absence of an improvement in glucose utilization and lipid handling that would normally be expected from exposure of tissues to high levels of FGF21. A similar scenario has been recently reported in rodent models of obesity, in which resistance to the effects of FGF21 is associated with downregulation of FGF receptors and β-klotho in white fat.14 For FGF19, the situation is such that the beneficial metabolic effects of FGF19 would be strongly blunted in adipose tissue from patients by the combination of lowered levels of the hormone and reduced expression of its receptor system components. In fact, it has been recently proposed that beneficial metabolic effects of FGF19 apply mainly to its action in adipose tissue.7
In light of the minimal available information about the regulation of FGF receptors and β-klotho coreceptor proteins, any proposed mechanisms to account for the reduction of their expression can only be speculative. However, because the gene expression for these proteins is markedly reduced in HIV-1–infected patients even without HAART, it is likely that alterations caused by HIV-1 infection itself, ranging from inflammation to initial alterations in lipid metabolism, may contribute to abnormally low expression. Among the initial alterations reported in fat from HIV-1–infected patients before treatment, and as confirmed here, is the abnormal downregulation of the transcription factor PPAR-γ.2,29 This alteration seems to be the consequence of pro-inflammatory signals (eg, enhanced tumor necrosis factor-α levels) and the direct action of HIV-1–encoded proteins such as Vpr.30 It has been reported that β-Klotho is strongly regulated by the PPAR-γ activator rosiglitazone in cultured adipocytes,23 suggesting that abnormal PPAR-γ–dependent regulation of gene expression may account for the strong reduction in β-Klotho in adipose tissue from HIV-1–infected patients.
In summary, we have found that the levels of FGF19 are decreased and most probably the sensitivity to FGF19 and FGF21 is reduced in adipose tissue in HIV-1–infected patients. These alterations may attenuate the beneficial effects of endocrine FGFs in HIV-infected patients and contribute to their metabolic alterations. However, considering that, in experimental models of reduced expression of endocrine FGFs receptors, pharmacological doses of FGF21 still produce beneficial effects,4,13 the potential of FGF19 and/or FGF21 treatment to ameliorate metabolic alterations in HIV-1–infected patients undergoing HAART warrants future attention.
1. Pullinger CR, Aouizerat BE, Gay C, et al.. Metabolic abnormalities and coronary heart disease risk in human immunodeficiency virus-infected adults. Metab Syndr Relat Disord. 2010;8:279–286.
2. Giralt M, Domingo P, Villarroya F. Adipose tissue biology and HIV-infection. Best Pract Res Clin Endocrinol Metab. 2011;25:487–499.
3. Itoh N. Hormone-like (endocrine) Fgfs: their evolutionary history and roles in development, metabolism, and disease. Cell Tissue Res. 2010;342:1–11.
4. Kharitonenkov A, Shiyanova TL, Koester A, et al.. FGF-21 as a novel metabolic regulator. J Clin Invest. 2005;115:1627–1635.
5. Tomlinson E, Fu L, John L, et al.. Transgenic mice expressing human fibroblast growth factor-19 display increased metabolic rate and decreased adiposity. Endocrinology. 2002;143:1741–1747.
6. Bhatnagar S, Damron HA, Hillgartner FB. Fibroblast growth factor-19, a novel factor that inhibits hepatic fatty acid synthesis. J Biol Chem. 2009;284:10023–10033.
7. Wu X, Ge H, Lemon B, et al.. Selective activation of FGFR4 by an FGF19 variant does not improve glucose metabolism in ob/ob mice. Proc Natl Acad Sci U S A. 2009;106:14379–14384.
8. Kir S, Beddow SA, Samuel VT, et al.. FGF19 as a postprandial, insulin-independent activator of hepatic protein and glycogen synthesis. Science. 2011;331:1621–1624.
9. Kurosu H, Choi M, Ogawa Y, et al.. Tissue-specific expression of betaKlotho and fibroblast growth factor (FGF) receptor isoforms determines metabolic activity of FGF19 and FGF21. J Biol Chem. 2007;282:26687–26695.
10. Wu X, Li Y. Role of FGF19 induced FGFR4 activation in the regulation of glucose homeostasis. Aging (Albany NY). 2009;1:1023–1027.
11. Domingo P, Gallego-Escuredo JM, Domingo JC, et al.. Serum FGF21 levels are elevated in association with lipodystrophy, insulin resistance and biomarkers of liver injury in HIV-1-infected patients. AIDS. 2010;24:2629–2637.
12. Chen WW, Li L, Yang GY, et al.. Circulating FGF-21 levels in normal subjects and in newly diagnose patients with Type 2 diabetes mellitus. Exp Clin Endocrinol Diabetes. 2008;116:65–68.
13. Zhang X, Yeung DC, Karpisek M, et al.. Serum FGF21 levels are increased in obesity and are independently associated with the metabolic syndrome in humans. Diabetes. 2008;57:1246–1253.
14. Fisher FM, Chui PC, Antonellis PJ, et al.. Obesity is a fibroblast growth factor 21 (FGF21)-resistant state. Diabetes. 2010;59:2781–2789.
15. Lemieux S, Prud'homme D, Bouchard C, et al.. A single threshold value of waist girth identifies normal-weight and overweight subjects with excess visceral adipose tissue. Am J Clin Nutr. 1996;64:685–693.
16. Lichtenstein KA. Redefining lipodystrophy syndrome: risks and impact on clinical decision making. J Acquir Immune Defic Syndr. 2005;39:395–400.
17. Sugiuchi H, Uji Y, Okabe H, et al.. Direct measurement of high-density lipoprotein cholesterol in serum with polyethylene glycol-modified enzymes and sulfated alpha-cyclodextrin. Clin Chem. 1995;41:717–723.
18. Matthews DR, Hosker JP, Rudenski AS, et al.. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985;28:412–419.
19. Gabrielsson BG, Johansson JM, Jennische E, et al.. Depot-specific expression of fibroblast growth factors in human adipose tissue. Obes Res. 2002;10:608–616.
20. Jones S. Mini-review: endocrine actions of fibroblast growth factor 19. Mol Pharm. 2008;5:42–48.
21. Dushay J, Chui PC, Gopalakrishnan GS, et al.. Increased fibroblast growth factor 21 in obesity and nonalcoholic fatty liver disease. Gastroenterology. 2010;139:456–463.
22. Fon TK, Bookout AL, Ding X, et al.. Research resource: comprehensive expression atlas of the fibroblast growth factor system in adult mouse. Mol Endocrinol. 2010;24:2050–2064.
23. Moyers JS, Shiyanova TL, Mehrbod F, et al.. Molecular determinants of FGF-21 activity-synergy and cross-talk with PPARgamma signaling. J Cell Physiol. 2007;210:1–6.
24. Mráz M, Lacinová Z, Kaválková P, et al.. Serum concentrations of fibroblast growth factor 19 in patients with obesity and type 2 diabetes mellitus: the influence of acute hyperinsulinemia, very-low calorie diet and PPAR-α agonist treatment. Physiol Res. 2011;60:627–636.
25. Lundåsen T, Gälman C, Angelin B, et al.. Circulating intestinal fibroblast growth factor 19 has a pronounced diurnal variation and modulates hepatic bile acid synthesis in man. J Intern Med. 2006;260:530–536.
26. McRae M, Rezk NL, Bridges AS, et al.. Plasma bile acid concentrations in patients with human immunodeficiency virus infection receiving protease inhibitor therapy: possible implications for hepatotoxicity. Pharmacotherapy. 2010;30:17–24.
27. Nolan D, Hammond E, James I, et al.. Contribution of nucleoside-analogue reverse transcriptase inhibitor therapy to lipoatrophy from the population to the cellular level. Antivir Ther. 2003;8:617–626.
28. Haubrich RH, Riddler SA, DiRienzo AG, et al.. Metabolic outcomes in a randomized trial of nucleoside, nonnucleoside and protease inhibitor-sparing regimens for initial HIV treatment. AIDS. 2009;23:1109–1118.
29. Giralt M, Domingo P, Guallar JP, et al.. HIV-1 infection alters gene expression in adipose tissue, which contributes to HIV-1/HAART-associated lipodystrophy. Antivir Ther. 2006;11:729–740.
30. Giralt M, Domingo P, Villarroya F. HIV-1 infection and the PPARgamma-dependent control of adipose tissue physiology. PPAR Res. 2009;2009:607902.