Paraoxonase‐1 Gene Haplotypes Are Associated with Metabolic Disturbances, Atherosclerosis, and Immunologic Outcome in HIV‐Infected Patients

Paraoxonase‐1 Gene Haplotypes Are Associated with Metabolic Disturbances, Atherosclerosis, and Immunologic Outcome in HIV‐Infected Patients

Please download to get full document.

View again

of 8
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.


Publish on:

Views: 3 | Pages: 8

Extension: PDF | Download: 0

  Paraoxonase-1 Haplotypes in HIV Infection  •  JID 2010:201 (15 February)  •  627 M A J O R A R T I C L E Paraoxonase-1 Gene Haplotypes Are Associatedwith Metabolic Disturbances, Atherosclerosis,and Immunologic Outcome in HIV-Infected Patients Sandra Parra, 1,2 Judit Marsillach, 1 Gerard Aragone´s, 1 Rau´l Beltra´n, 1 Manuel Montero, 3 Blai Coll, 4 Bharti Mackness, 1 Michael Mackness, 1 Carlos Alonso-Villaverde, 1,2 Jorge Joven, 1 and Jordi Camps 1 1 Centre de Recerca Biome`dica, Serveis de  2 Medicina Interna and  3 Radiologia, Hospital Universitari de Sant Joan, Institut d’Investigacio´ Sanita`riaPere Virgili, Universitat Rovira i Virgili, Reus, and  4 Institut de Recerca Biome`dica, Hospital Arnau Vilanova, Lleida, Spain Background.  Oxidative stress is associated with human immunodeficiency virus (HIV) infection. Paraoxonase-1 (PON1) is an antioxidant enzyme that is bound to high-density lipoproteins (HDLs). We evaluated whether PON1  gene haplotypes influence the metabolic disturbances, presence of subclinical atherosclerosis, and virologicoutcome associated with the infection. Methods.  DNA from blood samples collected from 234 HIV-infected patients and 633 healthy control sub- jects had single-nucleotide polymorphisms of   PON1 192  , PON1 55  , PON1  162  , PON1  832  , PON1  909  , PON1  1076  ,  and PON1  1741  analyzed using the Iplex Gold MassArray method. Subsequently, the influence of these single-nucleo-tide polymorphisms on measured biochemical and clinical variables was assessed. Results.  We observed significant differences in the haplotype distribution between the control subjects and theHIV-infected patients. Haplotype H10 (GTCCGTC) was more prevalent in the HIV-infected patients (6.41% vs0.64%; ), and haplotype H5 (GACCGTC) was less prevalent in HIV-infected patients (27.7% vs 42.9%; P  ! .001). In HIV-infected patients, haplotype H7 (AATTCCT) was associated with better CD4 + cellcountrecovery, P  p .001higher levels of HDL cholesterol ( ) and apolipoprotein A-I ( ), lower levels of triglycerides P  p .048  P  p .019( ), and lower rates of subclinical arteriosclerosis ( ). P  p .004  P  ! .001 Conclusions.  PON1  haplotypes segregate with HIV infection, HDL metabolism, the presence of subclinicalatherosclerosis, and CD4 + cell recovery after treatment.The spread of human immunodeficiency virus type 1(HIV-1) infection continues to increase, whereas thesurvival of HIV-1–infected patients is considerably ex-tended by new and more-effective antiretroviral ther-apy. Hence, it is likely that long-term consequences of treatment and infection will become increasingly com-mon. These consequences involve not only the im- Received 14 April 2009; accepted 25 September 2009; electronically published14 January 2010.Potential conflicts of interest: none reported.Financial support: Red de Centros de Metabolismo y Nutricio´n (grant C03/08)and Fondo de Investigacio´n Sanitaria (grants 04/1752, 05/1607, and 08/1175) fromthe Instituto de Salud Carlos III, Madrid, Spain. S.P. is the recipient of a careerdevelopment award from the Instituto de Salud Carlos III (grant CM06/00246).J.M. is the recipient of a postgraduate research fellowship from the Generalitatde Catalunya (grant FI 05/00068).Reprints or correspondence: Dr Jordi Camps, Centre de Recerca Biome`dica, Hos-pital Universitari de Sant Joan, C. Sant Joan s/n, 43201-Reus, Catalunya, Spain( The Journal of Infectious Diseases 2010;201:627–34   2010 by the Infectious Diseases Society of America. All rights reserved.0022-1899/2010/20104-0020$15.00DOI: 10.1086/650312 munocompromised status of the patient but also met-abolic derangements, including lipoprotein disordersthat may lead to cardiovascular disease[1].Oneoftheseproblems is increased oxidative stress resulting fromeither the infection itself or the secondary effects of treatments [2]. Paraoxonase-1 (PON1) is an enzymewith antioxidant properties. PON1 is an esterase andlactonase that catalyzes the hydrolysis of oxidized phos-pholipids and lipophylic lactones [3]. In the generalpopulation and in patients with diabetes, PON1 pre-serves high-density lipoproteins (HDLs) and low-den-sity lipoproteins (LDLs) from peroxidation and, assuch, has been associated with a protective role againstthe development of atherosclerosis [4–8]. PON1 knock-out mice have increased macrophage oxidative stressand are more susceptible to atherosclerosis than areother mice [9], and human  PON1  transgenic mice havedecreased atherosclerosis formation and lipoproteinoxidation [10]. PON1 attenuates in vitro productionof monocyte chemoattractant protein–1 (MCP-1) by    a  t  Bi   b l  i   o t   e  c  a  d  e l   a  Uni   v e r  s i   t   a  t   d  e B a r  c  e l   on a  on J   a n u a r  y1  8  ,2  0 1  3 h  t   t   p :  /   /   j  i   d  . oxf   or  d  j   o ur n a l   s  . or  g /  D o wnl   o a  d  e  d f  r  om   628  •  JID 2010:201 (15 February)  •  Parra et al Table 1. General Characteristics of 234 Human Immunodeficiency Virus (HIV)–In-fected Patients Characteristic (no. of patients) a ValueAge, mean    SD, years ( n p 234) 38.7    6.8Sex, male ( n p 234) 162 (69.2)Conventional risk factor for cardiovascular disease ( n p 234)Current smoker 184 (78.6)Hypertension 21 (9.0)Abnormal fasting glucose level 20 (8.5)Body mass index, mean    SD, kg/m 2 23.1    3.2Dyslipidemia 85 (36.3)Risk factor for HIV infection ( n p 234)Intravenous drug use 133 (56.8)Male homosexual contact 29 (12.3)Heterosexual contact 68 (29.1)Time since HIV diagnosis, mean    SD, years ( n p 234) 5.4    3.3Baseline CD4 + cell count, mean    SD, U/mm 3 ( n p 185) 359.9    297.3Viral load  ! 200 copies/mL ( n p 234) 92 (39.3)AIDS-related disease ( n p 234) 77 (32.9)HCV coinfection ( n p 234) 142 (60.6)Lipodystrophy ( n p 234) 51 (21.8)Duration of previous ART received, mean    SD, months ( n p 234)Nucleoside analogues 103.3    62.7Protease inhibitor 30.2    27.3Nonnucleoside analogues 8.1    10.4Treated with statins ( n p 234) 6 (2.6)Treated with fibrates ( n p 234) 20 (8.5)Lipid profileCholesterol, mean    SD, mmol/L ( n p 186) 4.9    1.3HDL cholesterol, mean    SD, mmol/L ( n p 202) 1.2    .5LDL cholesterol, mean    SD, mmol/L ( n p 171) 2.8    1.0Triglycerides, mean    SD, mmol/L ( n p 184) 2.3    2.1Apolipoprotein A-I, mean    SD, g/L ( n p 146) 1.4    .3Apolipoprotein A-II, mean    SD, g/L ( n p 196) .33    .06 NOTE.  Quantitative variables are expressed as the mean value    standarddeviation.Qualitativevariables are expressed as the no. (%) of patients. ART, antiretroviral therapy; HCV, hepatitis C virus;HDL, high-density lipoprotein; HIV, human immunodeficiency virus;LDL,low-densitylipoprotein;SD,standard deviation. a No. of patients for whom information was available. monocytes. MCP-1 is a proinflammatory chemokine that isinvolved in the initial step of formation of the atheromatousplaque [11]. Previous studies from our group demonstrated anincreased plasma MCP-1 concentration in HIV-infected pa-tients and an association between the polymorphisms in the MCP-1  2518   allele and the presence of subclinical atherosclerosis[12]. We have also reported elsewhere [13] that PON1 statusis influenced by the course of HIV infection and results in adecrease in PON1 activity.The hypothesis of the present study is that the relationshipbetween oxidative stress, HIV infection, and atherosclerosis issuch that  PON1  gene polymorphisms could be associated withthe metabolic disturbances associated with the infection, as wellas with the immunologic, virologic, and clinical course of thisdisease. METHODS Study population and design.  In an initial study, we per-formed case-control comparisons to assess the differences ingenotype distributions of the  PON1  genes in HIV-infected pa-tients, compared with differences in the  PON1  gene distribu-tions noted in the general population. The study participantswho were used as the control group enrolled in a population-based study conducted in our area. Details of this study havebeen published elsewhere [14]. In brief, the 633 participants   a  t  Bi   b l  i   o t   e  c  a  d  e l   a  Uni   v e r  s i   t   a  t   d  e B a r  c  e l   on a  on J   a n u a r  y1  8  ,2  0 1  3 h  t   t   p :  /   /   j  i   d  . oxf   or  d  j   o ur n a l   s  . or  g /  D o wnl   o a  d  e  d f  r  om   Paraoxonase-1 Haplotypes in HIV Infection  •  JID 2010:201 (15 February)  •  629 Table 2. Differences between Haplotype Distribution in theControl Subjects and the Human Immunodeficiency Virus (HIV)–Infected Patients Haplotype (sequence)Controlsubjects,no. (%)( n p 633)HIV-infectedpatients,no. (%)( n p 234)  P  H1 a (AACCGTC) 111 (17.5) 55 (23.5) .054H2 a (ATCCGTC) 563 (88.9) 199 (85.04) .760H3 a (GACTCTT) 57 (9.0) 16 (6.8) .419H4 a (AACCCTT) 96 (15.1) 41 (17.5) .368H5 a (GACCGTC) 272 (42.9) 65 (27.7) .001H6 (ATTTCCT) 11 (1.7) 3 (1.3) 1.000H7* (AATTCCT) 240 (37.9) 81 (34.61) .631H8* (GACCCTT) 80 (12.6) 31 (13.2) .740H9 (ATCCGTT) 1 (.15) 0 (0) 1.000H10* (GTCCGTC) 4 (.63) 15 (6.41)  ! .001H11* (AATTCCC) 47 (7.42) 15 (6.41) .771H12* (GATTCCT) 48 (7.58) 21 (8.97) .488H13 (GACCTGG) 4 (.63) 0 (0) .578H14 (ATCCCTT) 32 (5.05) 8 (3.41) .469H15 (GTTCCTT) 1 (.15) 0 (0) 1.000H16 (AACTCTT) 9 (1.42) 9 (3.84) .031H17 (GTCCCTT) 12 (1.89) 9 (3.84) .131H18 (AACCCTC) 4 (.63) 3 (1.28) .390H19 (AACTCCT) 1 (.15) 0 (0) 1.000H20 (ATTCCTT) 1 (.15) 0 (0) 1.000H21 (ATTTCCC) 1 (.15) 0 (0) 1.000H22 (GTTTCCT) 1 (.15) 0 (0) .265H23 (GATCCTT) 1 (.15) 0 (0) .265H24 (AATCCTT) 1 (.15) 0 (0) .265H25 (GACCCCT) 1 (.15) 0 (0) .265 a Haplotype with a frequency  1 5%. were ostensibly healthy individuals (339 women and 294 men;mean age, 45 years [range, 18–81 years]) of white ethnic srcinwho were from the Mediterranean region of Catalonia. The234 HIV-infected patients who were studied (72 women and162 men; mean age, 39 years [range, 20–66 years]) were amongthose attending our outpatient AIDS clinic and were of thesame ethnic srcin as the control participants in the study. Theonly exclusion criterion was age  ! 18 years.Expanding on the initial study, we performed a case-controlassessment of the HIV-infected patients receiving antiretroviraltreatment. The purpose of the assessment was to evaluate any associations of   PON1  gene haplotypes with the metabolic, im-munologic, and virologic variables measured. In this secondstudy, we defined as “cases” those HIV-infected patients whohad lipodystrophy, metabolic disturbances,dyslipidemia(asde-fined by the National Cholesterol Education Program AdultTreatment Panel III as a total cholesterol level  1 5.0 mmol/L oran LDL cholesterol level 1 3.0 mmol/L; an HDL cholesterol ! 1.0mmol/L in men and  ! 1.2 in women; or a triglyceride level 1 1.7mmol/L), a positive cardiovascular disease risk (as assessed by the Framingham risk score), and thepresenceofatherosclerosis,as determined by measurement of the intima-media thickness(IMT) in the carotid artery. Although IMT is a continuousvariable, and although we have reported the data in this form,we also defined subclinical atherosclerosis as (1) a categorical(dichotomized) variable when the IMT was  1 0.8 mm or (2)the presence of an atheromatous plaque in the analyzed zonesof the arteries.We used previously established definitions of cases and con-trols in association with the magnitude of the increase in theCD4 + cell count—that is, we reassigned the HIV-infected pa-tients as cases or controls on the basis of their response or lack of response, respectively, to anti-HIV treatment. Patients whodid not have a CD4 + cell count increase of   1 50 cells/mm 3 after12 months of treatment follow-up were considered to be “non-responders” [15]. In terms of virologic variables, we consideredcases to be those patients who experienced a rebound in theviral load during the course of the 12-month observationalperiod of the study. Patients who abandoned treatment andthose who completed the 12-month observational period witha negative viral load were censored from the present statisticalanalyses.All the participants provided fully informed consent. Thedata were coded to ensure anonymity. The study was approvedby the ethics committee of the Hospital Universitari de SantJoan de Reus. All procedures were performed according to theprinciples of the Declaration of Helsinki and Good ClinicalPractice. Biochemical measurements.  A sample of fasting venousblood was obtained duringtheclinicalexamination.Theplasmaviral load was measured using the Cobas TaqMan HIV-1 assay (Roche), and the CD4 + T cell count was determined using flow cytometry (Beckman-Coulter). HDL cholesterol levels were an-alyzed by a homogeneous method (Beckman-Coulter). LDLcholesterol levels were calculated using the Friedewald formula[16]. Serum total cholesterol and triglyceride concentrationswere measured by standard methods (Beckman-Coulter). Se-rum PON1 esterase activity was measured as the rate of hy-drolysis of paraoxon at 410 nm and 37  C in 0.05 mmol/Lglycine buffer (pH 10.5) with 1 mmol/L calcium chloride [17].Activities were expressed as the number of units per liter, where1 unit equals 1 micromole of paraoxon hydrolyzed per minute.Serum PON1 lactonase activity was measured in an assay re-agent containing 1 mmol/L calcium chloride, 0.25 mmol/L 5-(thiobutyl)-butyrolactone (TBBL), and 0.5 mmol/L 5,5 ′ -dithio-bis-2-nitrobenzoic acid (DTNB) in 0.05 mmol/L Tris-HCLbuffer (pH 8.0). The increase in absorbance was monitored at412 nm [18, 19]. Activities were expressed as the number of units per liter, where 1 unit equals 1 millimole of TBBL hy-drolyzed per minute. The serum PON1 concentration was de-   a  t  Bi   b l  i   o t   e  c  a  d  e l   a  Uni   v e r  s i   t   a  t   d  e B a r  c  e l   on a  on J   a n u a r  y1  8  ,2  0 1  3 h  t   t   p :  /   /   j  i   d  . oxf   or  d  j   o ur n a l   s  . or  g /  D o wnl   o a  d  e  d f  r  om   630  •  JID 2010:201 (15 February)  •  Parra et al Figure 1.  Linkage disequilibrium structure of the  PON1  gene in control subjects  (A)   and human immunodeficiency virus (HIV)–infected patients  (B). Linkage disequilibrium structure of 7 common single-nucleotide polymorphisms (SNPs) labeled by their RS numbers vertically. Note that SNP number1 is  PON1 192  , followed by  PON1 55  , PON1  162  , PON1  832  , PON1  909  , PON1  1076  ,  and  PON1  1741 .  Each square denotes the strength and significance oflinkage disequilibrium between pairs of markers in the region. Black denotes no (or minimal) evidence of historical recombination. The nos. of squaresdenote 100    D  ′ (a statistical measure of linkage disequilibrium), with missing values denoting a result of 100. termined by enzyme-linked immunosorbentassay(ELISA)per-formed using an antibody raised against a peptide derived fromthe sequence of mature PON1 [6]. The serum concentrationof oxidized LDL was measured by ELISA (Mercodia). Serumapolipoprotein (apo) A-I and A-II concentrations were deter-mined by immunoturbidimetry (Beckman-Coulter and DialabGmbh, respectively). The serum concentration of C-reactiveprotein (CRP) was measured using a high-sensitivity method(Beckman-Coulter). The plasma concentration of MCP-1 wasmeasured by ELISA (Human MCP-1 ELISA Development Kit;Prepotech). Genotyping.  Genomic DNA was obtained from leuko-cytes (Puregene DNA Isolation reagent set; Gentra Systems). PON1 192  , PON1 55  , PON1  162  , PON1  832  , PON1  909  , PON1  1076  ,PON1  1741 ,  and  MCP-1  2518   single-nucleotide polymorphisms(SNPs) were analyzed using the Iplex Gold MassArray method(Sequenom) at the Spanish National Genotyping Center (Cen-tro Nacional de Genotipado of the Universitat Pompeu Fabra,Barcelona, Spain). Arterial IMT measurement.  We performed carotid andfemoral ultrasound measurements for 183 HIV-infected pa-tients. By use of an identical protocol, these measurements wereconducted by the same investigators who performed a previ-ously published study [12], all of whom were blinded withrespect to the results of the other variables studied. We used aGE Logiq 700 with an ultrasound probe of 7–10 MHz. Weidentified 3 segments in the carotid arteries on which to con-duct the measurements: the common carotidartery(1cmprox-imal to the bifurcation), the carotid bulb (in the bifurcation),and the internal carotid artery (1 cm distal to the bifurcation).We evaluated the common femoral artery 1 cm proximal tothe bifurcation. The far-wall IMT images were obtained anddigitized for each participant. Statistical analysis.  We used the test to assess the degree 2 x of association between categorical variables. An analysis of var-iance (ANOVA) test or Student’s  t   test was used for continuousvariables that followed a normal distribution. The Mann-Whit-ney   U   test and the Wilcoxon rank-sum test were used for non-parametric variables. The results are presented as the mean val-ue    standard deviation (SD), for parametric variables, andthe median value (range), for nonparametric variables. SNPswere tested for Hardy-Weinberg equilibrium by use of Hap-loview software (version 4.0; Broad Institute) [20]. Estimatesof linkage disequilibrium between SNPs were calculated using D  ′ and . Haplotype estimations were performed usingPHASE 2 r  software (version 2; University of Chicago) [21] with defaultsettings and the SNPator package (CeGen) [22]. Linear or lo-gistic regression models were used to identify the haplotypesthat predicted the dependent variables after adjustment forpotential confounding factors, such as age, sex, dyslipidemia,smoking habit, hypertension, fasting glucose, body mass index,hepatitis C virus coinfection,lipid-loweringtreatment,durationof each antiretroviral treatment scheme, and basal CD4 + cellcount. The Kaplan-Meier hazard model was used to determinethe association between the haplotypes and the time to anundetectable viral load at a 95% confidence interval (CI). Allstatistical analyses were performed using the SPSS statisticalpackage (version 15.0; SPSS).   a  t  Bi   b l  i   o t   e  c  a  d  e l   a  Uni   v e r  s i   t   a  t   d  e B a r  c  e l   on a  on J   a n u a r  y1  8  ,2  0 1  3 h  t   t   p :  /   /   j  i   d  . oxf   or  d  j   o ur n a l   s  . or  g /  D o wnl   o a  d  e  d f  r  om   Paraoxonase-1 Haplotypes in HIV Infection  •  JID 2010:201 (15 February)  •  631 Figure 2.  A,  Differences between the haplotype distributions in controlsubjects and human immunodeficiency virus (HIV)–infected patients.  B, Differences in the proportion of HIV-infected patients with a CD4 + cellresponse in relation to the presence of absence of haplotype H7.  C, Differences in the proportion of HIV-infected patients with subclinicalatherosclerosis in relation to the presence or absence of haplotype H7. RESULTS Genotype distributions in control subjects and HIV-infected  patients.  The general characteristics of the patients are sum-marized in Table 1. We obtained 25 different haplotypes cor-responding to the SNP mutations  PON1 192  , PON1 55  , PON1  162  ,PON1  832  , PON1  909  , PON1  1076  ,  and  PON1  1741  (Table 2). Weobserved strong linkage disequilibrium between all the SNPsin HIV-infected patients ( D  ′ p 43), except for the SNPbetweenthe  PON1 192   and  PON1 55   polymorphisms (Figure 1). We ob-served significant differences in the distributions of haplotypeH10 (odds ratio [OR], 10.6 [95% CI, 3.52–32.2]; ) and P  ! .001haplotype H5 (OR, 0.61 [95% CI, 0.46–0.83]; ) be- P  p .001tween control subjects and HIV-infected patients (Figure 2 A  ). Association of    PON1  haplotypes with the immunologic and virologic outcomes in HIV-infected patients.  In a bivariateanalysis, we found significant differences in the CD4 + cellcounts of HIV-infected patients segregated according to hap-lotype. Patients carrying the H7 haplotype had higher basalCD4 + cell counts (Figure 2 B   and Table 3). After segregating pa-tients according to CD4 + cell count, we also observed a greaternumber of responders among patients carryingtheH7haplotypethan among patients who did not carry this haplotype (72.9%vs 52.7%; ). P  p .017The probability of maintaining an undetectable viral loadwhile receiving antiretroviral treatment was also significantly associated with  PON1  gene haplotypes. Carrying haplotypesH10 and H5 was associated with a higher probability of main-taining viral suppression ( and , respectively). P  p .047  P  p .022 Association of    PON1  haplotypes with the presence of lipo- dystrophy, metabolic disturbances,cardiovasculardiseaserisk,and dyslipidemia.  We did not observe (by use of bivariateanalysis) any association between  PON1  haplotypes and lipo-dystrophy, metabolic disturbances, or the 10-year risk of car-diovascular disease, as assessed by the Framingham risk score.However, we did find an association between dyslipidemia andthe H7 and H4 haplotypes ( and , respec- P  p .025  P  p .006tively). Patients carrying the H7 haplotype had lower serumtriglyceride concentrations, higher HDL cholesterol levels, andhigher apo A-I and apo A-II concentrations (Table 3).HIV-infected patients carrying the H4 haplotype also hadlower serum triglyceride levels (mean    SD, 2.16    2.05mmol/L vs 3.09    2.35 mmol/L; ) and higher HDL P  p .042cholesterol levels (1.23    0.53 mmol/L vs 0.97    0.25 mmol/L; ). In a linear logistic regression model, HDL cho- P  p .012lesterol levels (as a dependent variable) were predicted by tri-glycerides ( [95% CI,   .107 to   .044];  P   ! .001) b p  .075and the duration of nonnucleoside antiretroviral treatment( [95% CI, .004 to .015]; ). In a linear re- b p .009  P  !  .001gression model, serum apo A-I concentrations were predictedby the H7 haplotype ( [95% CI, .070–.283]; b p .176  P  p ), the duration of antiretroviral treatment ( [95%.008  b p .008CI, .003–.0012]; ), and hepatitis C virus coinfection P  ! .001( [95% CI,   .119 to .012]; ). b p  .106  P  p .028 Association of    PON1  haplotypes with the presence of sub- clinical atherosclerosis in HIV-infected patients.  We foundan association between the presence of subclinical atheroscle-rosis and the H7 haplotype (Figure 2 C  ). Patients who did notcarry this haplotype had higher rates of subclinical atheroscle-rosis (83.3% vs 16.7%; ). In the logisticregressionmodel, P  ! .001   a  t  Bi   b l  i   o t   e  c  a  d  e l   a  Uni   v e r  s i   t   a  t   d  e B a r  c  e l   on a  on J   a n u a r  y1  8  ,2  0 1  3 h  t   t   p :  /   /   j  i   d  . oxf   or  d  j   o ur n a l   s  . or  g /  D o wnl   o a  d  e  d f  r  om 
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks