Polymorphisms in Fas Gene Is Associated with HIV-Related Lipoatrophy in Thai Patients

Abstract

The present study aimed to evaluate the role of genetic polymorphisms in the emergence of lipoatrophy or lipodystrophy in HIV-infected patients with antiretroviral therapy (ART) in Thailand. Position 455 upstream of the Apolipoprotein C3 gene (ApoC3 T–455C, rs2854116), codon 64 of the Beta3 adrenergic receptor gene (ARβ3 Tcod64C, rs4994), and position 670 upstream of the Fas gene (Fas A–670G, rs1800682) were genotyped in 829 HIV-infected Thai patients who had started ART. Crude and adjusted incidence rate ratios (IRR) were calculated using Poisson regression. The serum levels of cholesterol, triglycerides, high-density lipoprotein (HDL), and low-density lipoprotein (LDL) were also analyzed. Multivariate analysis revealed an association between the Fas −670AA genotype, but not the ApoC3 −455 or ARβ3 cod64 genotypes, with the incidence of lipoatrophy after adjusting for gender and stavudine (d4T)-containing regimens (IRR=1.72, 95% CI=1.20–2.45, p=0.003). However, ApoC3 −455C homozygous patients showed elevated serum levels of triglycerides, while this genotype did not affect serum total cholesterol, HDL, or LDL levels in patients with lipoatrophy or lipodystrophy. In contrast, the ARβ3 cod64 genotype did not show any significant association with the serum levels of cholesterol, triglycerides, HDL, or LDL. In conclusion, Fas −670AA affected the incidence of lipoatrophy in HIV-1-infected Thai patients, while the ApoC3 −455C allele affected the serum levels of triglycerides. These results confirmed the role of genetics in the development of ART-related metabolic disorders.

Introduction

M etabolic and morphological alterations observed in human immunodeficiency virus (HIV)-infected patients represent a significant health problem related both to the possible long-term consequences and to the behavioral and psychological well-being of patients. Since the introduction of highly active antiretroviral therapy (HAART), the life expectancy of patients infected with HIV type 1 (HIV-1) has increased remarkably. Despite the clinical benefits, long-term HAART is associated with a complex spectrum of morphological alterations, characterized by body changes that may stigmatize patients and compromise their compliance with therapy, as well as metabolic effects, including dyslipidemia and insulin resistance, which lead to an elevated risk of developing diabetes mellitus and myocardial infarction.^1
–3

Morphological alterations are collectively named lipodystrophy (LD); however, in recent years it has been recognized that fat loss and fat gain represent distinct entities. Clinical features of fat loss include loss of subcutaneous fat in the face, arms, legs, and buttocks, while fat accumulation is characterized by excess fat deposition in visceral adipose tissue in the abdomen, breasts, dorsocervical region, and trunk.⁴

The underlying mechanism of antiretroviral therapy (ART)-related lipoatrophy (LA) has not been fully elucidated. Many mechanisms have been suggested, including mitochondrial toxicity by nucleoside reverse transcriptase inhibitors (NRTIs), inhibition of adipocyte differentiation by protease inhibitors (PIs), increased levels of inflammatory cytokines, and HIV itself.^5
–7 Mitochondrial toxicity has been clearly shown to play a pivotal role; in particular, thymidine analogs are strong inhibitors of DNA polymerase-γ and lead to mitochondrial depletion and dysfunction⁸; mainly stavudine (d4T), but also zidovudine (AZT) and didanosine (ddI) are strongly related to LA.^9,10 In previous in vitro and clinical studies, newer nucleoside and nucleotide agents, such as lamivudine (3TC), emtricitabine, abacavir, and tenofovir, appear to be much weaker inhibitors of mitochondrial DNA polymerase-γ or other mitochondrial functions, and appear to be associated with a lower risk of events thought to be related to mitochondrial toxicity.¹¹ However, a certain degree of mitochondrial toxicity is still present and will probably still affect HIV patients under ART, although to a lesser extent.^12
–14 In fact, switch studies from d4T to other NRTIs and complete switch off of thymidine analogs showed modest even if consistent increases in limb fat.^15
–17

The pathogenesis of lipohypertrophy appears to be multifactorial, with age, sex, HIV itself, immune depression, and duration and type of ART related to its appearance.^18
–20 PIs were suggested to be involved in lipohypertrophy; however, visceral fat accumulation also occurs in the absence of PIs.^21,22 A compartment-specific effect of mitochondrial toxicity within the adipose tissue may be related to lipohypertrophy²³ as well as dysregulation of fatty acid metabolism and altered expression of adipokines.^24
–26

However, LA and lipohypertrophy do not occur in all treated patients, and there is a very large degree of interindividual variability in the timing of emergence and severity of symptoms. A study of identical twins showed that genetic factors are involved in the accumulation of visceral adipose tissue.²⁷ These data suggest that host genetic factors may play a role, and inherited predispositions may have a significant influence on the appearance of LD and metabolic alterations as well as on the viroimmunological response to the drugs.

Therefore, host genetic polymorphisms may have a significant effect on the response to ART in terms of metabolism and immune response.²⁸ A study of Italian HIV patients suggested that single nucleotide polymorphisms (SNPs) in genes involved in apoptosis and lipid metabolism mediate the development of LA.²⁹ In particular, the TT and CT genotypes at position −455 of the Apolipoprotein C3 (ApoC3) gene (ApoC3 T–455C, rs2854116), nonsynonymous (W to R) CC and CT genotypes at codon 64 of the Beta3 adrenergic receptor (ARβ3) gene (ARβ3 Tcod64C, rs4994), and the AA genotype at position −670 of the Fas gene (Fas A–670G, rs1800682) were associated with increased incidence rates of LA.²⁹ ApoC3 protein, localized mainly in very low-density lipoprotein and high-density lipoprotein (HDL), is involved in fat metabolism and may delay the catabolism of triglyceride-rich particles. ApoC3 inhibits lipoprotein lipase and hepatic lipase and decreases the chylomicrons in hepatic cells.^30,31 ARβ3 protein, expressed mainly in visceral adipose tissue,^32,33 contributes to both lipolysis and the delivery of free fatty acids. Fas is the main gene that controls cell death by inducing apoptosis. A previous study indicated that Fas and Fas ligand influenced the immune reconstitution induced by ART.³⁴

In Thailand, the National Access to Antiretroviral Care program started in 2002 using standard regimens with GPO-VIR (a combination of d4T, 3TC, and nevirapine) as the first line of ART. The present study was performed to determine whether these genetic polymorphisms found in the Italian cohort also mediate the emergence of LA or LD in HIV-infected patients starting ART in Thailand in consideration of the relevance of these side effects in the currently used regimens and of the differences in ethnicity, alimentation, and way of life of the Thai population.

Materials and Methods

Study design

HIV-1-infected individuals receiving ART at Bamrasnaradura Infectious Diseases Institute, Nonthaburi, Thailand were enrolled in this study. LA was defined as the presence of fat loss at any site of the body (face, arms, legs, or buttocks). LD was defined as LA combined with fat accumulation diagnosed at the same time. Fat accumulation was defined as the presence of at least one alteration, such as buffalo hump, or fat accumulation in the breast, abdomen, neck, or lipomas. To qualify, the alteration had to be recognized by both the patient and the physician.³⁵ Most patients who developed LA or LD changed regimen for those without d4T. A 200 μl sample of whole blood was collected from each patient and stored at −20°C until DNA extraction. The genotypes of ApoC3 −455, ARβ3 cod64, and Fas −670 were compared between patients who developed LA or LD and those who did not. Patients' age, gender, transmission route, CD4 counts at pre-ART and at 1 year post-ART (cells/μl), drug regimens, and duration of therapy were obtained from their medical records. Lipid profile data were obtained from patients' records of serum biochemical tests. Blood for this test was collected when patients were in the fasting state; however, not all patients had serum biochemical test data as the test incurs extra costs for patients. All participants signed an informed consent form. The study was approved by the institutional ethical committees at the Bamrasnaradura Infectious Diseases Institute and Department of Disease Control, Ministry of Public Health, Thailand.

DNA extraction

Genomic DNA was extracted from 200 μl of each whole blood sample using a QIAamp DNA Blood Mini Kit (QIAGEN, Hilden, Germany) according to the manufacturer's instructions.³⁶

Genotyping

Genotyping of Fas gene promoter polymorphism, A–670G (rs1800682)

Polymerase chain reaction (PCR) and restriction fragment length polymorphism analysis for Fas A–670G, modified from a previously published technique,³⁷ was conducted with the primers 5′-CTACCTAAGAGCTATCTACCGTTC-3′ and 5′-GGCTGTCCATGTTGTGGCTGC-3′. The PCR product was digested with the restriction enzyme MvaI (Roche Biochemicals, Mannheim, Germany) at 37°C overnight. Two polymorphic alleles could be distinguished on the basis of restriction length polymorphisms (188 bp for G and 232 bp for A).

Genotyping of ApoC3 gene promoter polymorphism, T–455C (rs2854116)

TaqMan SNP assay was performed using the StepOne real-time PCR system (Applied Biosystems, Foster City, CA) with primers (5′-GAGCTCAGCCCCTGTAACCAG-3′ and 5′-ACACAGCCTGGAGTAGAGGG-3′) and TaqMan MGB probes (5′-VIC-CTCCAAACACCCC-MGB-3′ for the C allele and 5′-FAM-TTTACTCCAAACATCC-MGB-3′ for the T allele).

Genotyping of ARβ3 gene coding polymorphism, Tcod64C (rs4994)

TaqMan SNP assay was performed with a TaqMan SNP assay kit (C___2215549_20; Applied Biosystems).

Statistical analysis

Person-year analysis was conducted to investigate a single endpoint: the incidence of LA or LD. If the patient was diagnosed as both LA and LD, the earlier date of the first diagnosis was used as the endpoint. We focused on the time to the occurrence of the first event. Person-years at risk were calculated from the start date of ART until the last available follow-up or development of the event, whichever occurred first. The Poisson regression univariate model was used to determine whether the genetic polymorphisms as well as patients' age, gender, transmission route, CD4 counts at pre-ART and at 1 year post-ART, and starting drug regimes were predictors of LA or LD. Associations were expressed as incidence rate ratio (IRR) and corresponding 95% confidence interval (CI). Factors showing significant association with LA or LD were then included as covariates in the Poisson regression multivariate model to adjust for their possible confounding effects. Starting drug regimens were grouped into either d4T-containing regimens or other backbone regimens. Continuous variables were compared using the Wilcoxon signed rank test for paired samples and Kruskal–Wallis test for unpaired samples. A variable was considered significant at p<0.05.

Results

Incidences of LA and/or LD in HIV-1-infected Thai patients undergoing ART

A total of 829 HIV-1-infected individuals who had started ART were enrolled in this study at the Bamrasnaradura Infectious Diseases Institute, Nonthaburi, Thailand, from April 2008 to August 2009. Patient demographic data are shown in Table 1. The study population consisted of 478 males (57.7%) and 351 females (42.3%). Most of the patients contracted infection through sexual contact. Almost half of the patients started ART immediately after being diagnosed with HIV infection. Median CD4 count per microliter pre-ART was 45 and that post-ART (after 1 year) was 210.5. The mean follow-up period was 4.85 years (SD=2.42).

Table 1.

Demographic Data of All the Patients

Demographic data	n=829
Mean age (SD) years	41.12 (7.77)
Male (%)	478 (57.66)
Transmission
Sexual (%)	806 (97.23)
Intravenous drug user (%)	8 (0.97)
Unknown (%)	15 (1.81)
Median baseline CD4 (cells/μl) (IQR)	45 (15, 127.5)
Median CD4 (cells/μl) after treatment (IQR)	210.5 (142, 305)
Median duration of therapy before onset of lipoatrophy or lipodystrophy (years, IQR) (n=270)	2.63 (1.67, 4.00)
Median duration of therapy of nonlipoatrophy or nonlipodystrophy (years, IQR) (n=559)	5.53 (4.10, 7.18)

SD, standard deviation; IQR, interquartile range.

We first combined patients with LA and those with LD for analysis, as fat loss was commonly observed in both LA and LD cases. Among the 829 patients, 270 (32.6%) were diagnosed with LA or LD within the observation period. We kept on monitoring the other 559 (67.4%) until the end of the observation period and found that they developed neither LA nor LD (non-LA/LD cases). The median duration of therapy until diagnosis was 2.63 years for the LA or LD cases and 5.53 years for the non-LA/LD cases. In a follow-up of 4019.8 person-years, the incidence rate of LA or LD was 6.70 per 100 person-years. LA or LD cases were found more frequently in females (40.7%) than in males (26.6%), and the incidence rate per 100 person-years for females (9.11) was significantly higher than that for males (5.18) (IRR=1.76, 95% CI=1.37–2.25, p<0.001) (Table 2). Patients using d4T/3TC backbone regimens as starting drugs showed a significantly higher incidence rate per 100 person-years (7.71) than those with other non-d4T-containing (mostly AZT/3TC) regimens (3.49) (IRR=2.21, 95% CI=1.53–3.28, p<0.001) (Table 2). These findings were consistent with those of previous studies in other ethnic groups.^{9,20,29,38,39} In contrast, age, transmission route, and CD4 counts pre-ART and at 1 year post-ART did not significantly affect the incidence of LA or LD (Table 2).

Table 2.

Univariate and Multivariate Analysis of All Patients (n=829)

				Crude		Adjusted ² (n=829)
Factors	Events ¹	Person-years	Events/100 person-years	IRR (95% CI)	p-value	IRR (95% CI)	p-value
Gender
Male	127	2449.9	5.18	1
Female	143	1569.9	9.11	1.76 (1.37–2.25)	<0.001	1.70 (1.33–2.16)	<0.001
Transmission
Unknown³	7	95.9	7.3	1
Sexual	263	3924.0	6.7	0.92 (0.44–2.31)	0.782
NRTIs
No-d4T	33⁴	945.8	3.49	1
d4T	237⁵	3074.0	7.71	2.21 (1.53–3.28)	<0.001	2.05 (1.42–2.96)	<0.001
Age×10 (years)				0.87 (0.74–1.02)	0.090
CD4 at baseline⁵				0.9988 (0.9973–1.0003)	0.107
CD4 after therapy				0.9497 (0.9987–1.0005)	0.458
Fas−670
AG/GG	200	3221.6	6.21	1
AA	70	798.2	8.77	1.41 (1.06–1.86)	0.015	1.47 (1.12–1.94)	0.005
ApoC3−455
CT/TT	206	3139.0	6.56	1
CC	64	880.9	7.27	1.11 (0.82–1.47)	0.474
ARβ3 cod64
CC/CT	58	815.8	7.11	1
TT	212	3204.0	6.62	0.93 (0.69–1.27)	0.621

Diagnosis of lipoatrophy or lipodystrophy.

Gender, NRTIs, and Fas−670 were included as covariates.

Including eight intravenous drug users.

No-d4T group mainly used AZT. One case had received ddI instead of AZT as an initial regimen.

CD4 cells/μl.

IRR, incidence rate ratios; NRTIs, nucleoside reverse transcriptase inhibitors; d4T, stavudine.

Fas −670AA, but not ApoC3 −455 or ARβ3 cod64 genotype, affects incidence of LA or LD in HIV-1-infected Thai patients

To elucidate the possible effects of genetic polymorphisms on the development of LA or LD, we determined the genotypes of Fas −670, ApoC3 −455, and ARβ3 cod64 of 829 HIV-1-infected Thai patients. LA or LD cases were found significantly more frequently among patients with the Fas −670AA genotype than in those with other Fas −670 genotypes (IRR=1.41, 95% CI=1.06–1.86, p=0.015, Table 2). The Fas −670AA genotype was found in 25.9% of the LA or LD cases and in 20.2% of the non-LA/LD cases during the observation period. Increased frequency of the Fas −670AA genotype in LA or LD cases was observed in both males and females (data not shown). In cases of ApoC3 −455 and ARβ3 cod64 genotypes, however, there were no statistically significant differences in the frequency of LA or LD between groups (Table 2). Adjustment for gender and starting drug regimens in multivariate Poisson regression analysis resulted in a slightly higher IRR for Fas −670AA (1.47, 95% CI=1.12-194, p=0.005, Table 2). This difference was statistically significant after correction for multiple testing of three SNPs (p=0.015).

Fas–670AA genotype affects incidence of LA but not that of LD in HIV-1-infected Thai patients

Among 270 LA and/or LD cases, 148 were diagnosed as LA singly and 105 were diagnosed as LD singly. There were 17 cases diagnosed as LA first who then subsequently developed fat accumulation. These 17 cases were grouped as LD cases. The median durations of therapy until diagnosis were 2.58 years for LA and 2.83 years for LD cases. When we analyzed LA and LD separately, females showed a higher IRR in both LA (1.60, 95% CI=1.14–2.24, p=0.005) and LD (2.31, 95% CI=1.59–3.38, p<0.001) (Tables 3 and 4)., Similarly, d4T-containing regimens showed significantly higher IRR in both LA (2.46, 95% CI=1.48–4.36, p<0.001) and LD (2.10, 95%CI=1.23–3.80, p=0.003) (Tables 3 and 4). In the case of the Fas −670AA genotype, however, significantly higher IRR was observed only in LA (1.68, 95% CI=1.15–2.41, p=0.006) but not in LD (1.19, 95%CI=0.74–1.83, p=0.433) (Tables 3 and 4). The statistical significance of higher IRR of LA for Fas −670AA did not change after adjustment for gender and starting drug regimens in multivariate Poisson regression analysis (1.72, 95% CI=1.20–2.45, p=0.003, Table 3). The difference was statistically significant even after correction for multiple testing of three SNPs in two groups (LA or LD) (p=0.018). These results suggested that the Fas −670 genotype affects LA without fat accumulation.

Table 3.

Lipoatrophy: Rate of Incidence and Multivariate Analysis (n=707)

				Crude		Adjusted ² (n=707)
Factors	Events ¹	Person-years	Events/100 person-years	IRR (95% CI)	p-value	IRR (95% CI)	p-value
Gender
Male	76	2282.1	3.33	1
Female	72	1350.9	5.33	1.60 (1.14–2.24)	0.005	1.56 (1.13–2.16)	0.007
Transmission
Unknown	3	85.2	3.52	1
Sexual	145	3547.8	4.09	1.16 (0.39–5.69)	0.865
NRTIs
No-d4T	17	880.5	1.93	1
d4T	131	2752.5	4.76	2.46 (1.48–4.36)	<0.001	2.32 (1.40–3.85)	0.001
Age×10 (years)				0.88 (0.71–1.08)	0.228
CD4 at baseline³				0.9985 (0.9964–1.0006)	0.151
CD4 after therapy³				0.9991 (0.9979–1.0003)	0.151
Fas−670
AG/GG	105	2920.4	3.59	1
AA	43	712.6	6.03	1.68 (1.15–2.41)	0.006	1.72 (1.20–2.45)	0.003
ApoC3−455
CT/TT	118	2866.1	4.12	1
CC	30	766.9	3.91	0.95 (0.61–1.43)	0.818
ARβ3 cod64
CC/CT	38	748.0	5.08	1
TT	110	2885.0	3.81	0.75 (0.51–1.12)	0.134

Diagnosis of lipoatrophy.

Gender, NRTIs, and Fas−670 were included as covariates.

CD4 cells/μl.

IRR, incidence rate ratios; NRTIs, nucleoside reverse transcriptase inhibitors; d4T, stavudine.

Table 4.

Lipodystrophy: Rate of Incidence and Multivariate Analysis (n=681)

				Crude		Adjusted ² (n=681)
Factors	Events ¹	Person-years	Events/100 person-years	IRR (95% CI)	p-value	IRR (95% CI)	p-value
Gender
Male	51	2234.3	2.28	1
Female	71	1346.3	5.27	2.31 (1.59–3.38)	<0.001	2.19 (1.53–3.15)	<0.001
Transmission
Unknown	4	86.0	4.65	1
Sexual	118	3494.5	3.38	0.73 (0.28–2.71)	0.508
NRTIs
No-d4T	16	860.3	1.86	1
d4T	106	2720.2	3.90	2.10 (1.23–3.80)	0.003	1.82 (1.07–3.09)	0.027
Age×10 (years)				0.85 (0.67–1.07)	0.171
CD4 at baseline³				0.9988 (0.9966–1.0010)	0.278
CD4 after therapy³				1.0000 (0.9990–1.0014)	0.696
Fas−670
AG/GG	95	2888.0	3.29	1
AA	27	692.6	3.90	1.19 (0.74–1.83)	0.433	1.27 (0.83–1.95)	0.278
ApoC3−455
CT/TT	88	2781.9	3.16	1
CC	34	798.7	4.26	1.34 (0.88–2.02)	0.148
ARβ3 cod64
CC/CT	20	714.5	2.80	1
TT	102	2866.1	3.56	1.27 (0.78–2.17)	0.328

Diagnosis of lipoatrophy.

Gender, NRTIs, and Fas−670 were included as covariates.

CD4 cells/μl.

Lipodystrophy cases included 17 cases diagnosed with both lipoatrophy and lipodystrophy.

IRR, incidence rate ratios; NRTIs, nucleoside reverse transcriptase inhibitors; d4T, stavudine.

Lipid profiles of HIV-1-infected Thai patients

We analyzed the serum lipid profiles of the patients. From the records of 829 patients, serum cholesterol levels were obtained in 610 and 628 cases at the first and last measurement of the observation period, respectively. Serum triglyceride (TG) levels were obtained in 646 and 716 cases at the first and last measurement, respectively. Serum HDL levels were obtained in 393 and 367 cases at the first and last measurement, respectively. Serum low-density lipoprotein (LDL) levels were obtained in 382 and 438 cases at the first and last measurement, respectively. Serum cholesterol levels showed a normal distribution, while serum TG, HDL, and LDL levels did not. Therefore, we used a nonparametric method to evaluate the statistical significance of differences in patient lipid profiles. Slight increases in serum cholesterol levels were observed in the LA (from median 194.5 to 206 mg/dl, p=0.016), LD (from median 194 to 205 mg/dl, p=0.001), and non-LA/LD cases (from median 195 to 200 mg/dl, p=0.033) during ART (Table 5). Similarly, significant increases in TG were observed in the LA (from median 130 to 183.5 mg/dl, p=0.019) and LD (from median 122 to 165.5 mg/dl, p<0.001), but not in non-LA/LD cases (from median 138 to 148.5 mg/dl, p=0.359; Table 5). Accordingly, the LA and LD cases showed significantly higher TG levels than the non-LA/LD cases at the last measurement (p=0.007; Table 5). These results indicated that lipid metabolism in LA and LD cases was more severely damaged by ART than that in non-LA/LD cases. It should also be noted here that the LA and LD cases showed weak but clear trends toward lower TG levels than non-LA/LD cases at the first measurement (p=0.091; Table 5). In fact, when combined, the LA and LD cases showed significantly lower TG levels (median 122 mg/dl) than the non-LA/LD cases (median 138 mg/dl) at the first measurement (p=0.049). On the other hand, HDL and LDL levels remained almost unchanged during the treatment period in the LA and LD, but non-LA/LD cases showed a slight decrease in LDL (from median 137 to 132.5 mg/dl, p=0.026; Table 5).

Table 5.

Lipid Profile of Patients

	LA cases ¹			LD cases ²			Non-LA/LD cases ³
Lipid profile	n	Median	IQR	n	Median	IQR	n	Median	IQR	p-value ⁴
Cholesterol (mg/dl)
First test⁵	96	194.5	165.25–231.25	77	194	159.00–216.50	437	195	160.00–224.00	0.507
Last test⁶	121	206	177.50–236.00	95	205	180.00–239.00	412	200	176.500–227.75	0.106
p-value⁷		0.016			0.001			0.033
TG (mg/dl)
First test⁵	98	130	92.25–190.25	89	122	88.00–164.50	459	138.0	95.00–201.00	0.091
Last test⁶	140	183.5	109.50–285.00	116	165.5	103.75–256.25	460	148.5	97.00–224.75	0.007
p-value⁷		0.019			<0.001			0.359
HDL (mg/dl)
First test⁵	43	50	40.00–67.00	45	51	44.00–70.00	305	50	40.00–62.00	0.648
Last test⁶	63	52	42.00–66.00	51	52	42.00–63.00	253	50	40.00–62.00	0.419
p-value⁷		0.933			0.885			0.392
LDL (mg/dl)
First test⁵	41	136	117.00–174.00	51	133	116.00–168.00	290	137	108.00–167.00	0.731
Last test⁶	70	135.5	107.75–166.25	88	136.5	114.25–165.75	280	132.5	111.00–157.00	0.295
p-value⁷		0.777			0.735			0.026

Lipoatrophy cases.

Lipodystrophy cases.

Patients without lipoatrophy or lipodystrophy.

Kruskal–Wallis test.

The first measurement after patients started antiretroviral therapy.

The last measurement of the observation period. For LA and LD cases, this corresponded to the diagnosis of LA and LD, respectively.

Wilcoxon signed rank test.

LA, lipoatrophy; LD, lipodystrophy; IQR, interquartile range; TG, triglyceride; HDL, high-density lipoprotein; LDL, low-density lipoprotein.

As the LA and LD cases showed similar lipid profiles, we combined the LA and LD cases again and analyzed the effects of patients' genotypes on the lipid profile together. The results indicated that the ApoC3 −455CC genotype was associated with decreased serum cholesterol levels in non-LA/LD cases at the last measurement (p=0.035; Table 6). However, this association was not observed in LA and LD cases (p=0.393; Table 6) or at the first measurement in both LA and LD cases and non-LA/LD cases (p=0.254 and p=0.583, respectively). The ApoC3 −455CC genotype was also found to be associated with elevated serum TG levels in the LA and LD cases at the time of LA or LD diagnosis (p=0.021; Table 6). In the non-LA/LD cases, the ApoC3 −455CC genotype had virtually no effect on serum TG levels. These results suggested that the ApoC3 −455CC genotype promotes dysregulation of lipid metabolism in LA or LD cases, even though this allele had no deleterious effect on the onset of LA or LD (Table 2). On the other hand, the ApoC3 −455CC genotype affected neither HDL nor LDL serum levels (data not shown). The ARβ3 cod64 genotype had no effect on the serum lipid data of any patient group (Table 6). Similarly, Fas −670AA, which was found to be significantly associated with LA, did not show any effects on serum lipid levels (Table 6).

Table 6.

Patient Genotype and Lipid Profile

	Cholesterol (mg/dl)
	LA and LD cases ¹						Non-LA/LD cases ²
	First test			Last test			First test			Last test
Genotype	n	Median	IQR	n	Median	IQR	n	Median	IQR	n	Median	IQR
ApoC3−455 CC	42	202	172.50–225.00	51	214	183.00–239.00	86	196	154.75–220.25	86	193.5	164.50–220.50
ApoC3−455 CT/TT	131	191	159.00–218.00	165	204	179.50–240.00	351	195	161.00–224.00	326	202.5	177.75–229.00
p-value³		0.254			0.393			0.583			0.035
ARβ3 cod64 TT	138	194	159.75–224.25	169	206	177.50–216.50	347	198	162.00–224.00	332	202	175.25–228.00
ARβ3 cod64 CT/CC	35	194	177.00–207.00	47	202	178.00–246.00	90	189	156.00–223.50	80	194	174.25–221.00
p-value³		0.806			0.658			0.360			0.328
Fas−670 AA	39	192	158.00–217.00	53	211	182.50–240.00	87	197	158.00–223.00	85	201	176.50–232.50
Fas−670 AG/GG	134	195	165.75–223.50	163	205	173.50–236.75	350	195	161.00–224.00	327	199	175.50–226.00
p-value³		0.259			0.602			0.820			0.493

	TG (mg/dL)
	LA and LD cases ¹						Non-LA/LD cases ²
	First test			Last test			First test			Last test
Genotype	n	Median	IQR	n	Median	IQR	n	Median	IQR	n	Median	IQR
ApoC3−455 CC	44	132	102.25–209.75	61	222	129.00–327.50	91	145	99.00–204.00	94	158	99.25–210.25
ApoC3−455 CT/TT	143	122	87.00–179.00	195	166	101.00–248.25	368	136	93.25–201.00	366	144	97.00–226.00
p-value³		0.277			0.021			0.771			0.945
ARβ3 cod64 TT	150	126	93.00–185.75	198	175	77.50–281.00	363	140	97.00–199.00	360	144.5	93.25–218.75
ARβ3 cod64 CT/CC	37	108	82.50–160.00	58	167.50	109.00–274.50	93	124.5	77.75–206.75	100	154	107.00–232.50
p-value³		0.174			0.973			0.368			0.500
Fas−670 AA	42	120	89.50–162.75	64	167	99.25–270.75	92	128	94.00–202.50	96	123.5	96.50–223.75
Fas−670 AG/GG	145	123	89.00–184.00	192	175	103.00–260.00	367	140	95.00–201.00	364	152	97.00–224.75
p-value³		0.580			0.530			0.655			0.498

Lipoatrophy and lipodystrophy cases.

Patients without lipoatrophy or lipodystrophy.

Kruskal–Wallis test.

LA, lipoatrophy; LD, lipodystrophy; IQR, interquartile range.

Discussion

In the present study, we evaluated the influence of SNPs involved in apoptosis and lipid metabolism on developing LA or LD in Thai HIV-1 patients. Specifically, we analyzed the SNPs of the Fas A–670G, ApoC3 T–455C, and ARβ3 Tcod64C in patients undergoing ART.

Of the SNPs investigated, our findings indicated that the AA genotype of Fas −670 was significantly associated with the risk of LA but not that of LD. This result is consistent with that of a previous study in Italian HIV patients by Zanone Poma et al.²⁹ It was suggested that adipocyte apoptosis induced by mitochondrial toxicity of NRTIs was the primary mechanism involved in LA.¹⁵ Consistent with this suggestion, increased Fas gene expression levels were observed in adipose tissue of HIV-1 patients with NRTI-associated LA.⁴⁰ The Fas A–670G polymorphism is located in the γ-interferon activation site, to which transcription factors such as signal transducer and activator of transcription (STAT) bind.⁴¹ Therefore, this SNP may affect the level of Fas gene transcription and may induce apoptosis, consequently increasing the risk of LA. The Fas A–670G polymorphism has been reported to be associated with several autoimmune diseases and inflammatory disorders,^41
–43 which may involve a mechanism similar to LA. The frequency of the AA genotype of Fas −670 in 829 HIV-1-infected Thais was 22.1%, which is similar to the percentages in Europeans and Japanese but is markedly higher than that in Africans. As Fas −670AA caused a 70% increase in risk of LA, this genotype is a candidate for future pretherapy genetic screening, although it is still necessary to find other SNPs that increase the risk of LA for accurate screening. It is also necessary to confirm our findings on Fas −670AA in a larger patient cohort.

On the other hand, the frequency distribution of the ARβ3 cod64 and ApoC3 −455 genotypes was not significantly different between LA or LD cases and non-LA/LD cases. The ARβ3 Tcod64C polymorphism failed to show any effect on lipid metabolism among Thai patients in this study. The results for ARβ3 cod64 were consistent with a previous report that showed no association of this polymorphism with lipid metabolism during weight loss in obese subjects.⁴⁴ However, in the Italian cohort, this genotype was associated with the risk of developing LA.²⁹ Several studies have reported that people with this SNP showed an increased risk of developing obesity and glucose intolerance,^32,33 whereas other studies failed to find any correlation between this mutation and body mass index.^45,46

The results for ApoC3 −455 were also not consistent with those of the Italian cohort, which showed an association of the ApoC3 −455T allele with LA. However, it was also reported that the ApoC3 −455C but not the ApoC3 −455T allele was associated with metabolic syndrome.⁴⁷ In fact, a recent study in French HIV patients showed that the ApoC3 −455C but not the ApoC3 −455T allele is associated with the severity of ART-induced dyslipidemia and occurrence of LD.^48,49 Therefore, the precise role of the ApoC3 T–455C polymorphism in LA or LD during ART should be carefully reevaluated in a larger cohort of HIV patients.

Nevertheless, our study also showed that the ApoC3 −455CC genotype was associated with elevated serum levels of TG, which was consistent with the findings of a previous report showing that the ApoC3 −455C allele was associated with metabolic syndrome.⁴⁷ These data suggest that the mechanism underlying the promotion of LA and LD in HIV-infected individuals receiving ART may be different from that promoting metabolic syndrome. However, the differences in results may be due to ethnic differences in the cohorts. It was reported that genetic polymorphisms in hepatic lipase showed different effects on plasma lipid levels in different ethnic groups.⁵⁰ Foulkes et al. found race- and ethnic-specific differences in plasma lipid levels under ART, as well as differences in the influence of ApoC3 T–455C and the other SNP in ApoC3 gene on the development of protease inhibitor-related hypertriglyceridemia.⁵¹ Further studies including reevaluation of these SNPs in larger patient cohorts are therefore necessary to elucidate the precise mechanisms underlying LA and LD in HIV-infected individuals receiving ART.

In conclusion, the Fas −670AA genotype, but not the ApoC3 −455 or ARβ3 cod64 genotypes, affected the incidence of LA in HIV-1-infected Thai patients. None of these alleles affected the incidence of LD. On the other hand, the ApoC3 −455CC, but not the ARβ3 cod64 genotype, affected the serum levels of TG.

Footnotes

Acknowledgments

We thank Ms. Taksin Pimpak for her help in statistical analysis, Professor Massimo Galli at the University of Milan for his support, and all the participants in this study. This work was supported by grants from the Health Science Foundation, the Ministry of Education, Culture, Sports, Science, and Technology, and the Ministry of Health, Labour, and Welfare, Japan.

Author Disclosure Statement

No competing financial interests exist.

References

Behrens

, Dejam

, Schmidt

et al. Impaired glucose tolerance, beta cell function and lipid metabolism in HIV patients under treatment with protease inhibitors. AIDS, 1999; 13:F63–70.

Tsiodras

, Mantzoros

, Hammer

, Samore

. Effects of protease inhibitors on hyperglycemia, hyperlipidemia, and lipodystrophy: A 5-year cohort study. Arch Intern Med, 2000; 160:2050–2056.

Vigouroux

, Gharakhanian

, Salhi

et al.

Diabetes, insulin resistance and dyslipidaemia in lipodystrophic HIV-infected patients on highly active antiretroviral therapy (HAART)

Diabetes Metab, 1999; 25:225–232.

Carr

. HIV lipodystrophy: Risk factors, pathogenesis, diagnosis and management. AIDS, 2003; 17,Suppl 1:S141–148.

Birkus

, Hitchcock

, Cihlar

. Assessment of mitochondrial toxicity in human cells treated with tenofovir: Comparison with other nucleoside reverse transcriptase inhibitors. Antimicrob Agents Chemother, 2002; 46:716–723.

Maggiolo

, Roat

, Pinti

et al. Mitochondrial changes during D-drug-containing once-daily therapy in HIV-positive treatment-naive patients. Antivir Ther, 2010; 15:51–59.

Giralt

, Domingo

, Guallar

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.

Collins

, Sondel

, Cesar

, Hellerstein

. Effect of nucleoside reverse transcriptase inhibitors on mitochondrial DNA synthesis in rats and humans. J Acquir Immune Defic Syndr, 2004; 37:1132–1139.

Mallal

, John

, Moore

, James

, McKinnon

. Contribution of nucleoside analogue reverse transcriptase inhibitors to subcutaneous fat wasting in patients with HIV infection. AIDS, 2000; 14:1309–1316.

10.

van Vonderen

, van Agtmael

, Hassink

et al. Zidovudine/lamivudine for HIV-1 infection contributes to limb fat loss. PLoS One, 2009; 4:e5647.

11.

Moyle

. Mechanisms of HIV and nucleoside reverse transcriptase inhibitor injury to mitochondria. Antivir Ther, 2005; 10,Suppl 2:M47–52.

12.

Minami

, Yamamoto

, Takahama

, Ando

, Miyamura

, Suematsu

. Comparison of the influence of four classes of HIV antiretrovirals on adipogenic differentiation: The minimal effect of raltegravir and atazanavir. J Infect Chemother, 2011; 17:183–188.

13.

Lebrecht

, Venhoff

, Kirschner

, Wiech

, Venhoff

, Walker

. Mitochondrial tubulopathy in tenofovir disoproxil fumarate-treated rats. J Acquir Immune Defic Syndr, 2009; 51:258–263.

14.

Perazella

. Tenofovir-induced kidney disease: An acquired renal tubular mitochondriopathy. Kidney Int, 2010; 78:1060–1063.

15.

McComsey

, Paulsen

, Lonergan

et al. Improvements in lipoatrophy, mitochondrial DNA levels and fat apoptosis after replacing stavudine with abacavir or zidovudine. AIDS, 2005; 19:15–23.

16.

Gallant

, DeJesus

, Arribas

et al. Tenofovir DF, emtricitabine, and efavirenz vs. zidovudine, lamivudine, and efavirenz for HIV. N Engl J Med, 2006; 354:251–260.

17.

Benn

, Sauret-Jackson

, Cartledge

et al. Improvements in cheek volume in lipoatrophic individuals switching away from thymidine nucleoside reverse transcriptase inhibitors. HIV Med, 2009; 10:351–355.

18.

Lichtenstein

. Redefining lipodystrophy syndrome: Risks and impact on clinical decision making. J Acquir Immune Defic Syndr, 2005; 39:395–400.

19.

Leung

, Glesby

. Pathogenesis and treatment of HIV lipohypertrophy. Curr Opin Infect Dis, 2011; 24:43–49.

20.

Galli

, Veglia

, Angarano

et al. Gender differences in antiretroviral drug-related adipose tissue alterations. Women are at higher risk than men and develop particular lipodystrophy patterns. J Acquir Immune Defic Syndr, 2003; 34:58–61.

21.

Dube

, Komarow

, Mulligan

et al. Long-term body fat outcomes in antiretroviral-naive participants randomized to nelfinavir or efavirenz or both plus dual nucleosides. Dual X-ray absorptiometry results from A5005s, a substudy of Adult Clinical Trials Group 384. J Acquir Immune Defic Syndr, 2007; 45:508–514.

22.

Shikuma

, Yang

, Glesby

et al.

Metabolic effects of protease inhibitor-sparing antiretroviral regimens given as initial treatment of HIV-1 Infection (AIDS Clinical Trials Group Study A5095)

J Acquir Immune Defic Syndr, 2007; 44:540–550.

23.

Sekhar

, Jahoor

, Pownall

et al. Severely dysregulated disposal of postprandial triacylglycerols exacerbates hypertriacylglycerolemia in HIV lipodystrophy syndrome. Am J Clin Nutr, 2005; 81:1405–1410.

24.

Kratz

, Purnell

, Breen

et al. Reduced adipogenic gene expression in thigh adipose tissue precedes human immunodeficiency virus-associated lipoatrophy. J Clin Endocrinol Metab, 2008; 93:959–966.

25.

Addy

, Gavrila

, Tsiodras

, Brodovicz

, Karchmer

, Mantzoros

. Hypoadiponectinemia is associated with insulin resistance, hypertriglyceridemia, and fat redistribution in human immunodeficiency virus-infected patients treated with highly active antiretroviral therapy. J Clin Endocrinol Metab, 2003; 88:627–636.

26.

Nagy

, Tsiodras

, Martin

et al. Human immunodeficiency virus type 1-related lipoatrophy and lipohypertrophy are associated with serum concentrations of leptin. Clin Infect Dis, 2003; 36:795–802.

27.

Bouchard

, Tremblay

, Despres

et al. The response to long-term overfeeding in identical twins. N Engl J Med, 1990; 322:1477–1482.

28.

Fox

, Boffito

, Winton

. The clinical implications of anti-retroviral pharmacogenomics. Pharmacogenomics, 2006; 7:587–596.

29.

Zanone Poma

, Riva

, Nasi

et al. Genetic polymorphisms differently influencing the emergence of atrophy and fat accumulation in HIV-related lipodystrophy. AIDS, 2008; 22:1769–1778.

30.

Hodis

, Mack

. Triglyceride-rich lipoproteins and the progression of coronary artery disease. Curr Opin Lipidol, 1995; 6:209–214.

31.

Wang

, McConathy

, Kloer

, Alaupovic

. Modulation of lipoprotein lipase activity by apolipoproteins. Effect of apolipoprotein C-III. J Clin Invest, 1985; 75:384–390.

32.

Widen

, Lehto

, Kanninen

, Walston

, Shuldiner

, Groop

. Association of a polymorphism in the beta 3-adrenergic-receptor gene with features of the insulin resistance syndrome in Finns. N Engl J Med, 1995; 333:348–351.

33.

Clement

, Vaisse

, Manning

et al. Genetic variation in the beta 3-adrenergic receptor and an increased capacity to gain weight in patients with morbid obesity. N Engl J Med, 1995; 333:352–354.

34.

Nasi

, Pinti

, Bugarini

et al. Genetic polymorphisms of Fas (CD95) and Fas ligand (CD178) influence the rise in CD4+ T cell count after antiretroviral therapy in drug-naive HIV-positive patients. Immunogenetics, 2005; 57:628–635.

35.

Blanco

, Carr

. Lipodystrophy syndrome: Diagnostic, clinic and therapeutic aspects. AIDS Rev, 2001; 3:98–105.

36.

Likanonsakul

, Rattanatham

, Feangvad

et al. HLA-Cw*04 allele associated with nevirapine-induced rash in HIV-infected Thai patients. AIDS Res Ther, 2009; 6:22.

37.

Ueda

, Hung

, Terai

et al. Fas gene promoter −670 polymorphism (A/G) is associated with cervical carcinogenesis. Gynecol Oncol, 2005; 98:129–133.

38.

Walker

, Bickel

, Lutke Volksbeck

et al. Evidence of nucleoside analogue reverse transcriptase inhibitor-associated genetic and structural defects of mitochondria in adipose tissue of HIV-infected patients. J Acquir Immune Defic Syndr, 2002; 29:117–121.

39.

Gallant

, Staszewski

, Pozniak

et al. Efficacy and safety of tenofovir DF vs stavudine in combination therapy in antiretroviral-naive patients: A 3-year randomized trial. JAMA, 2004; 292:191–201.

40.

Sievers

, Walker

, Sevastianova

et al. Gene expression and immunohistochemistry in adipose tissue of HIV type 1-infected patients with nucleoside analogue reverse-transcriptase inhibitor-associated lipoatrophy. J Infect Dis, 2009; 200:252–262.

41.

Kanemitsu

, Ihara

, Saifddin

et al. A functional polymorphism in fas (CD95/APO-1) gene promoter associated with systemic lupus erythematosus. J Rheumatol, 2002; 29:1183–1188.

42.

Huang

, Danis

, Lassere

, Edmonds

, Manolios

. Evaluation of a new Apo-1/Fas promoter polymorphism in rheumatoid arthritis and systemic lupus erythematosus patients. Rheumatology (Oxford), 1999; 38:645–651.

43.

van Veen

, Kalkers

, Crusius

et al. The FAS-670 polymorphism influences susceptibility to multiple sclerosis. J Neuroimmunol, 2002; 128:95–100.

44.

Oksanen

, Mustajoki

, Kaprio

et al. Polymorphism of the beta 3-adrenergic receptor gene in morbid obesity. Int J Obes Relat Metab Disord, 1996; 20:1055–1061.

45.

de Silva

, Walder

, Boyko

et al. Genetic variation and obesity in Australian women: A prospective study. Obes Res, 2001; 9:733–740.

46.

Garenc

, Perusse

, Rankinen

et al. The Trp64Arg polymorphism of the beta3-adrenergic receptor gene is not associated with training-induced changes in body composition: The HERITAGE Family Study. Obes Res, 2001; 9:337–341.

47.

Pollex

, Ban

, Young

et al. Association between the −455T>C promoter polymorphism of the APOC3 gene and the metabolic syndrome in a multi-ethnic sample. BMC Med Genet, 2007; 8:80.

48.

Bonnet

, Bernard

, Fauvel

, Massip

, Ruidavets

, Perret

. Association of APOC3 polymorphisms with both dyslipidemia and lipoatrophy in HAART-receiving patients. AIDS Res Hum Retroviruses, 2008; 24:169–171.

49.

Bonnet

, Genoux

, Bernard

, Fauvel

, Massip

, Perret

. Impact of genetic polymorphisms on the risk of lipid disorders in patients on anti-HIV therapy. Clin Chem Lab Med, 2007; 45:815–821.

50.

Shohet

, Vega

, Bersot

et al.

Sources of variability in genetic association studies: Insights from the analysis of hepatic lipase (LIPC)

Hum Mutat, 2002; 19:536–542.

51.

Foulkes

, Wohl

, Frank

et al. Associations among race/ethnicity, ApoC-III genotypes, and lipids in HIV-1-infected individuals on antiretroviral therapy. PLoS Med, 2006; 3:e52.