Coreceptor Usage Of HIV-1 In Anti-retroviral Therapy Patients And Prediction Of Viral Coreceptor Usage By V3 Region

BackgroundIn 2007,the world's first co-receptor antagonist Maraviroc against HIV-1 was approved in the United States.Maraviroc,blocking HIV-1gp120 combination with CCR5, thereby prevents HIV-1 into the host cell.So,the drug will play an important role for occupational exposure to emergency prevention,and treatment failure.Meanwhile,there will be more co-receptor antagonist for clinical trial.In order to utilize the drugs rationally,it is necessary to know HIV-1 co-receptor usage before clinical application. However,we do not know what co-receptor HIV-1 strains use after anti-retroviral therapy. In this paper,in order to explore factors that influence HIV-1 co-receptor usage and analyze the feasibility that V3 region predicts HIV-1 co-receptor,HIV-1 co-receptor usage and V3 sequence were analyzed among anti-retroviral therapy and treatment-naive HIV-1 infection patients.And the study may provide information to help to design rational and effective treatment regimen for ART.MethodsHIV-1 infected individuals who has experienced the anti-retroviral therapy were selected from Henan,Anhui province as treatment group,and unexperienced ART patients were from Anhui province taken as treatment-na(i|¨)ve group.HIV-1 strains were isolated form PBMCs of whole blood while HIV-1 p24 was quantified using a commercial enzyme-linked immunosorbent assay(ELISA) kit.Then HIV-1 co-receptor was identified using Ghost cell lines expressing CD4 and the chemokine receptor CCR5 or CXCR4.HIV-1 V3 region was amplified from HIV-1 strains and further was analyzed the relationship between the characteristics of V3 sequence and HIV co-receptor usage. Then V3 sequence was submitted into web-server(WebPSSM and geno2pheno) to predict HIV-1 co-receptor usage,and predicted result was compared with HIV phenotype based Ghost cell line.Furthmore,the feasibility of Web-server assay was analyzed.Results45 HIV-1 B' strains were isolated from antiretroviral-treatment patients in Henan or Anhui province,of which 23(51%) were from Henan,therapeutic regimen for the AZT+DDI+NVP,and of which 22(49%) were from Anhui,therapeutic regimen for D4T+DDI+NVP.And the average treatment time was 26(6-48) months.109 HIV-1 B' strains were isolated from treatment-na(i|¨)ve patients.All the strain was identified as HIV-1 B' subtype after pol sequence analysis.Among 45 HIV strains from the treatment group,22(48.89%) strains used CCR5 as a co-receptor(R5 strain),21(46.67%) strains used CXCR4/CCR5 as a co-receptor(X4/R5 strain),and two(4.44%) used only CXCR4 as a co-receptor(X4 strain).In 109 strains from treatment-na(i|¨)ve group,96(88.07%) strains used CCR5 as a co-receptor(R5 strain), 13(11.93%) strains used CCR5/CXCR4 as a co-receptor use(X4/R5 strain).In order to analyze relationship between disease progression and HIV co-receptor usage,first of all,CD4+T count was classed three groups as follow:CD4＜100, 100=＜CD4＜200,CD4＞200.Statistical analysis showed that HIV-1 X4/R5 strains gradually increased as CD4~+T count decreasing in both treatment and treatment-na(i|¨)ve group.Even in the same CD4~+T count,the HIV-1 X4/R5 utilization in treatment groups was remarkable higher than in treatment-na(i|¨)ve group.The resullt showed some correlation between HIV-1 X4/R5 utilization rate and CD4~+T count after antiretroviral-treatment.Logistic regression analysis showed that the OR was 7.72(3.39-17.58)(p＜0.0001) between treatment and not,however treatment time,regimen,age and sex do not influence viral co-receptor usage,which statistical analysis was not significant difference(P＜0.05).Infection possibility based on Ghost-CCR5 cells line assay showed:HIV-1 CCR5 strain isolated from treatment group was 6.63 times of infection possibility than SF33 strain;while HIV-1 CCR5 strain isolated from treatment-na(i|¨)ve was 4.68 times than SF33 strain.For the dual-tropic strains isolated from treatment group,the infection possibility on Ghost-CXCR4 cell was 2.09 times than SF33,while 4.50 times than SF33 on Ghost-CCR5.For the dual-tropic strains isolated from treatment-naive group,the infection possibility was 0.69 times than SF33 on Ghost-CXCR4 cells,while 1.90 times than SF33 on Ghost-CCR5 cells.Among anti-retroviral therapy patients,V3 net charge was distributed from 2 to 7(4.20±1.63),while 3-6(4.26±0.78) in treatment-na(i|¨)ve population.There was no significant difference(p=0.38).The 11/25 rulle predicted 24(33.3%) X4 strains,16 strains form treatment group and 8 strains from treatment-na(i|¨)ve group.The fourfold table method was used to evaluate the agreement between phenotype based Ghost cell lines and predicted resullt based on 11/25 rulle.The result showed there was no relevance (X2=0.69,p=0.41) in the naive-treatment group.But in the treatment group,there was relevance(X2=5.33,p=0.02),suggesting 11/25 rule was adequate for predicting HIV-1 co-receptor usage after ART.As for Web-Server(WebPSSM and geno2pheno) for HIV-1 co-receptor prediction.WebPSSM predicted 53(72.6%) HIV-1 strain as CCRS-tropic, while geno2pheno predicted 34(46.6%) HIV-1 strain as CCR5-tropic.The concordance was 50%.Compared with viral phenotype based on Ghost cell assay,the concordance of WebPSSM was significantly better than geno2pheno.For predicting CCR5-tropic strains, there was 89.5%concordance between WebPSSM and phenotype,while 50% concordance between geno2pheno and phenotype.But for predicting CXCR4 tropic strains,the concordance was 44.1%between WebPSSM and Ghost cell lines,while 55.9%between geno2pheno and phenotype.Conclusion1.The study found that the HIV-1 CXCR4 utilization among anti-retroviral therapy HIV-1 infected patients was higher than the treatment-na(i|¨)ve population,whatever CD4~+T was in differ level,implying that it should pay attention to the choice of co-receptor antagonists after the failure in the Chinese central region,Furthermore,suggesting X4 strains could spread in population. 2.V3 net charge was higher among anti-retroviral therapy patients than the treatment-na(i|¨)ve population,and there was still correlation between HIV-1 X4/R5 utilization and V3 net charge after antiretroviral-treatment,and 11/25 rule was adequate for predicting HIV-1 co-receptor usage after antiviral therapy.3.For the treatment-na(i|¨)ve individual,if co-receptor antagonist is considered,WebPSSM is recommened to predict HIV-1 co-receptor usage.For the patients with onging treatment failure,if co-receptor antagonist is considered for replacement therapy,we also recommend WebPSSM.