|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Lipids, Lipoproteins, and Cardiovascular Risk Factors |
C Gene Variation
Institut de Recerca en Ciències de la Salut, Hospital Universitari de Sant Joan de Reus, Universitat Rovira i Virgili, Reus, Spain
aAddress correspondence to this author at: Unitat de Recerca de Lípids i Arteriosclerosi, Facultat de Medicina, Universitat Rovira i Virgili, Sant Llorenç, 21, 43201 Reus, Spain. Fax 34-977-75-9322; e-mail josep.ribalta{at}urv.cat.
| Abstract |
|---|
|
|
|---|
C genotype.
Methods: We measured lipoprotein changes in HIV-positive patients (n = 229) who had been followed for 5 years. For statistical analyses, we segregated the patients with respect to PI treatment and APOA51131T
C genotype.
Results: The frequency of the C allele was 0.08, similar to that in the general population. We found a strong effect of the APOA51131T
C genotype among patients receiving PIs. Carriers of the C allele had consistently increased mean (SD) triglyceride concentrations compared with noncarriers after 1 year [2.11 (1.62) vs 3.71 (4.27) mmol/L; P = 0.009], 2 years [2.48 (2.09) vs 4.02 (4.05) mmol/L, P = 0.050], 3 years [2.32 (1.71) vs 4.13 (4.26) mmol/L; P = 0.013], 4 years [2.90 (2.95) vs 5.35 (7.12) mmol/L; P was not significant], and 5 years [4.25 (5.58) vs 9.23 (9.63) mmol/L; P was not significant]. We observed the same effect on total cholesterol concentrations: after 1 year [4.93 (1.31) vs 5.87 (1.66) mmol/L; P = 0.006], 2 years [5.03 (1.12) vs 6.42 (2.48) mmol/L; P = 0.001], 3 years [5.11 (1.17) vs 6.38 (2.43) mmol/L; P = 0.009], 4 years [5.49 (1.71) vs 6.78 (3.03) mmol/L; P was not significant], and 5 years [5.56 (1.75) vs 7.90 (3.60) mmol/L; P was not significant]. HDL cholesterol showed a progressive reduction, leading to a considerably higher cholesterol/HDL cholesterol ratio after 3 years.
Conclusion: Variability in the APOA5 gene predisposes patients with HIV, particularly those treated with PI, to severe hyperlipidemia.
| Introduction |
|---|
|
|
|---|
C is the commonly used variant. APOA5 is expressed mainly in the liver and distributed predominantly on TG-rich lipoproteins such as chylomicrons and VLDL, and also on HDL. In human plasma, apo A-V is found at lower concentrations than other apolipoproteins. Its exact function has not been completely elucidated, but the available data indicate that it modulates TG metabolism by controlling production of VLDL and catabolism of the lipolysis of TG-rich lipoproteins (10). The latter role has been confirmed not only in vitro but also in patients who develop severe hypertriglyceridemia attributable to apo A-V deficiency (11). Because protease inhibitor (PI)-associated dyslipidemia is caused by increases in VLDL production (12) and by impaired lipolysis (13), we hypothesized that, in HIV-infected individuals, the APOA5 gene could be an important indicator of predisposition to PI-related deterioration of the lipid profile. Hence, we analyzed lipid changes in HIV patients segregated with respect to treatment strategy and APOA5 genotype.
| Materials and Methods |
|---|
|
|
|---|
C genotype have been measured for all participants. Exclusion criteria were age <18 years and AIDS-related opportunistic disease when the study began. The Ethics and Investigation Committee of our Hospital approved the study.
lipid profile analyses
We used standard laboratory methods to measure total cholesterol, HDL cholesterol, and TGs.
high-sensitive c-reactive protein analyses
We measured serum high-sensitive C-reactive protein (hsCRP) concentrations with a turbidimetric immunoassay (Biokit), according to the manufacturers instructions.
apoa5 1131t
c genotype analyses
Genomic DNA was obtained from leukocytes and extracted with calibrated methods. The 1131T
C variation in the APOA5 gene was detected with the oligonucleotide primers AV-15'-GAT TGA TTC AAG ATG CAT TTA GGA C-3' and AV-25'CCC CAG GAA CTG GAG CGA AAT T-3' to amplify a 187bp segment, and AV-2 primer forced a site for Mse I (New England Biolabs) enzymatic restriction. PCR was performed as described previously (8).
statistical analyses
We used a multivariate analysis on patients with complete data available [age, body mass index (BMI), sex, and lipid data at all time points] being entered into the analysis. To improve the statistical power of our dataset, we performed ANOVA analyses of data from every single time-point, used lipid values as a dependent variable, and normalized for confounding factors such as sex, age, BMI, and lipodystrophy. We also performed multivariate analyses of repeated measures, which confirmed the trends but did not reach statistical significance. For TG and hsCRP, calculations were performed on log-transformed values, although nontransformed concentrations are shown in the Tables and Figs.
We analyzed deviation from Hardy Weinberg equilibrium with the
2 goodness-of-fit test. Results are conveyed as mean (SD). Statistical significance was accepted at a value of P <0.05.
| Results |
|---|
|
|
|---|
C genotype and treatment scheme. For statistical purposes, the single patient homozygous for the C allele was pooled with those patients who were T/C heterozygote. For lipid analyses, patients undergoing the PI regimen (n = 148) and those not receiving PIs (n = 81) were studied separately. The frequency of the C allele was 0.08, which is similar to that found in the Spanish general population (0.07) (16). Allelic distribution was in Hardy-Weinberg equilibrium.
apoa5-associated changes in lipids and lipoproteins
The group of patients with the wild-type genotype (TT) and carriers of the rare variant (TC and CC) were comparable at baseline with respect to age, sex, immunologic status, hsCRP, total cholesterol, HDL cholesterol, and TGs (Table 1
). Only BMI was considerably higher in the carriers of the wild-type allele.
|
patients receiving pi therapy
Because hyperlipidemia is strongly associated with the PI regimen, we focused on the subgroup of 148 patients receiving PI as a component of their antiretroviral therapy, and we analyzed their lipid profile changes over the 5-year follow-up period. The 2 genotype groups were also comparable at baseline (pretreatment), including the percentage of patients receiving ritonavir (Table 1
).
Carriers of the C allele had consistently higher TG concentrations than noncarriers at 1 year [2.11 (1.62) vs 3.71 (4.27) mmol/L; P = 0.009], 2 years [2.48 (2.09) vs 4.02 (4.05) mmol/L; P = 0.050], 3 years [2.32 (1.71) vs 4.13 (4.26) mmol/L, P = 0.013], 4 years [2.90 (2.95) vs 5.35 (7.12) mmol/L; P not significant], and 5 years [4.25 (5.58) vs 9.23 (9.63) mmol/L; P not significant], after adjustment of the data for age, sex, BMI, and the presence of lipodystrophy (Fig. 1
).
|
Results were similar for total cholesterol. Carriers of the C allele had higher plasma cholesterol concentrations at 1 year [4.93 (1.31) vs 5.87 (1.66) mmol/L; P = 0.006], 2 years [5.03 (1.12) vs 6.42 (2.48) mmol/L; P = 0.001], 3 years [5.11 (1.17) vs 6.38 (2.43) mmol/L; P = 0.009], 4 years [5.49 (1.71) vs 6.78 (3.03) mmol/L; P = not significant], and 5 years [5.56 (1.75) vs 7.90 (3.60) mmol/L; P = not significant] (Fig. 2
). HDL cholesterol concentrations showed a tendency toward decrease in carriers of the C allele and increase in patients with the wild-type alleles, but these differences did not reach statistical significance (data not shown).
|
The total cholesterol/HDL cholesterol ratio, which was 78% higher in carriers of the C allele than in carriers of the wild-type allele (Fig. 3
), indicated that these lipid changes increased the risk of atherogenesis.
|
patients not receiving treatment with pis
To investigate whether the effect of APOA5 on the lipid profile in HIV patients was influenced by treatment with PIs or whether the effect was more generalized, we separately evaluated the 81 patients who were not receiving treatment with PIs. At baseline, the 2 genotype groups were comparable (Table 1
). There were no differences between genotypes with respect to total cholesterol, TGs, HDL cholesterol, or the total cholesterol/HDL cholesterol ratio over the 5-year follow-up period (Figs. 13
).
| Discussion |
|---|
|
|
|---|
apoa5 enhances pi-associated hyperlipidemia
The PIs used in combined therapies can produce major fat redistribution, hyperlipidemia, and insulin resistance. These effects can be mitigated by replacing PIs with other antiretroviral drugs (17). That these abnormalities do not develop in all patients on PI regimens suggests the involvement of genetic or environmental predisposing factors. We focused on the APOA5 gene because it is probably the strongest genetic determinant of plasma TG (9) identified to date, and few data on the influence of APOA5 on PI-induced hyperlipidemia are available.
Among the 229 HIV-patients we followed for a period of 5 years, those receiving the PI regimen tended, as expected, to have higher concentrations of total cholesterol and TGs during the treatment period. Despite similar baseline lipid values among all patients, however, the individuals carrying the 1131C variant of the APOA5 gene and undergoing treatment with PIs had the highest values at all of the follow-up time-points. Conversely, carriers of the 1131C variant not receiving PI treatment did not have such strongly increased concentrations of these lipids. This observation is in accordance with our previous studies demonstrating that the magnitude of the effect of the APOA5 gene is more pronounced when conditions are more metabolically challenging (16).
Increased cholesterol concentrations in carriers of the 1131C allele could also be the result of increased cholesterol delivery by the TG-rich lipoproteins, a welldescribed secondary feature of VLDL overproduction. Although detailed lipoprotein subfractionation was not available in the present group of patients, it is clear that increased VLDL synthesis, decreased catabolism, or a combination of these processes can lead to PI-induced hyperlipidemia. It is also clear that these metabolic perturbations are characterized essentially by an increasing TG component with a concomitant increase in the cholesterol concentration in these lipoproteins.
In our study sample, carriers of the APOA5 variant allele had higher rates of lipodystrophy. The APOA5 gene may somehow predispose individuals to this lipid and fatty-tissue redistribution, but this hypothesis is not supported by our finding that the frequency of lipodystrophy among APOA5 genotypes in the non-PI group did not differ from the frequency in the overall patient group. Conversely, after adjustment for the presence of lipodystrophy, all of the observed differences with respect to lipid concentrations remained considerable, indicating that the hyperlipemic effect was truly associated with the APOA5 gene and was not influenced by the presence of lipodystrophy.
A mechanistic approach to this observation could be that, although the exact function in vivo of apo A-V is not known, in vitro studies suggest that it acts by decreasing the assembly of the VLDL particle (18) and its secretion and by stimulating lipolysis in the circulation (10). Conversely, PI-induced hyperlipidemia has been shown to be caused by an increase in VLDL production (12) as well as by impaired lipolysis (13) as a consequence, in part, of a direct interaction between the PI and the sterol regulatory element-binding proteins (SREBP)1 and 2, leading to an accumulation of SREBPs in the nucleus, which stimulates lipid synthesis (19). Because APOA5 expression is down-regulated by SREBP1c (20), we assume that apo A-V plays a determining role in PI-induced hyperlipidemia. Whether the 1131C allele is a functional variant or is acting as a marker of a functional variant elsewhere in the gene is not clear (21).
apoa5 and atherogenic lipid profile
Patients who are receiving PI-treatment and who carry the 1131C allele present with a lipoprotein profile that deteriorates rapidly because of increasing of total cholesterol and TGs and a decrease in the HDL fraction, with the atherogenic total/HDL cholesterol ratio reaching the highest quintile after 3 years in carriers of the mutant C allele (22). This scenario does not occur in patients who are receiving the same treatment but who carry the wild-type gene or who are not receiving PI therapy.
One of the limitations of our study is that there were fewer patients available to follow up at years 4 and 5 than for the first 3 years because 4 to 5 years ago there was less concern regarding lipid alterations in individuals with HIV-AIDS, so the percentage of individuals with HIV-AIDS being treated for hyperlipidemia was considerably lower in the first 2 years of recruitment into the present study. We suspect that the patients who were recruited were those with more evident dyslipidemia, which might explain the greater increase in plasma TG concentrations in year 5. It is important to note, however, that all the main conclusions of the study were drawn from the data obtained during 13 years of follow-up. Data from 45 years of follow-up were included for completeness, and to indicate that the trends toward increased lipids observed in the first 3 years continued in the same direction over the subsequent years of follow-up in those patients, for whom detailed lipid datasets were available.
In conclusion, our results indicate that variability at the APOA5 gene variation predisposes HIV patients, particularly those treated with PIs, to severe hyperlipidemia. Although these results must be confirmed in future studies, they suggest the possibility of using the APOA5 gene as a marker of predisposition to severe hyperlipidemia as a consequence of treatment with PIs. In HIV-positive individuals found to carry this variation, treatment alternatives that exclude PI may need to be considered.
| Acknowledgments |
|---|
| Footnotes |
|---|
1 Human gene: APOA5, apolipoprotein A-V. ![]()
2 Nonstandard abbreviations: TG, triglyceride; PI, protease inhibitor; HsCRP, high-sensitive C-reactive protein; SREBP, sterol regulatory element-binding proteins; BMI, body mass index. ![]()
| References |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |