Stromal-derived factor-1 chemokine gene variant is associated with the delay of HIV-1 disease progression in two longitudinal cohorts

HIV-1 disease is multifactorial and is associated with a spectrum of clinical outcomes among infected persons [1]. Viral and host factors influence HIV-1 disease progression. Chemokine receptors function as co-receptors for viral entry, and studies have shown that viral isolates from long-term non-progressors (LTNP) maintain exclusive usage of the CCR5 co-receptor, whereas isolates from progressors generally broaden their co-receptor usage to include CCR2b, CCR3, and CXCR4 [2,3]. Host genetic factors, such as polymorphisms in the CCR2b receptor and a deletion in the CCR5 (CCR5Δ32) receptor, delay disease progression or prevent infection [4–6]. The later studies emphasize the role of host genetic factors in the natural history of HIV-1. No polymorphisms have been found in the CXCR4 gene that would explain why HIV isolates from LTNP maintain CCR5 as their predominant co-receptor [7]. Subsequent study of the gene encoding the CXCR4 ligand, stromal-derived factor (SDF-1) has shown a G to A polymorphism in the conserved 3′ untranslated region (SDF1-3′A), which was implicated in delaying HIV-1 disease progression [8,9]. Other studies have found an accelerated rate of HIV-1 disease progression [10] or death [11] in individuals with this polymorphism. To clarify this issue, we analysed for the presence of the SDF1-3′A polymorphism in two longitudinal, HIV-1-infected cohorts. The study participants included 90 HIV-seronegative individuals who comprised 52 normal blood donors and 38 high-risk, homosexual men, as well as HIV-1-infected individuals followed for more than 12 years [12,13]. The latter group was subdivided into 61 progressors and 31 LTNP. The progressors were characterized by AIDS-defining symptoms and a decline in CD4 T cell count to less than 200 in less than 8 years after infection, whereas the LTNP maintained CD4 T cell levels of over 400 for more than 8 years after infection and had no clinical symptoms. None of the LTNP were on therapeutic interventions while enrolled in these studies. DNA was purified from peripheral blood mononuclear cells and analysed for the presence of the SDF1-3′A polymorphism [8] and the CCR5Δ32 base pair deletion [5] using polymerase chain reaction-based assays. Among the HIV-seronegative individuals studied, the frequency of the SDF1-3′A allele was 0.178 (Table 1). Within this population, analysis of the blood donors and the homosexual men showed that the frequencies for SDF1-3′A were 0.221 (n = 52) and 0.118 (n = 38), respectively. The progressors had a lower SDF1-3'A allele frequency of 0.148 when compared with the seronegative individuals. Of note is the fact that the allele frequency for the LTNP was 0.306, which was significantly higher when compared with the frequencies for seronegative individuals (P = 0.046) and progressors (P = 0.019) (Table 1). Interestingly, the CCR5Δ32 allele frequency for the LTNP was not significantly higher than that of the seronegative individuals, but was significantly higher when compared with the progressors (P = 0.033) (Table 1).Table 1: Frequency of the SDF1-3′A polymorphism and the CCR5Δ32 deletion in two longitudinal cohorts of HIV-1-infected persons and HIV-seronegative individuals. Although these data are in agreement with the findings from two previous studies [8,9], they differ from the results of two others [10,11]. Reasons for the discrepancies between these studies may be the use of antiretroviral therapies [11], the inability to determine accurately the time of seroconversion [10], or the limited median follow-up time [11], all of which would impact the interpretation of these genetic polymorphisms on the natural history of HIV-1. The mechanism underlying the observed SDF1-3′A protective effect is unknown. This polymorphism occurs in the 3′ untranslated region of the gene [8], which may stabilize the RNA, thus affecting protein expression. SDF-1 binding of CXCR4 has been shown to result in rapid internalization of this receptor from the cell surface [14,15]. If increased protein expression occurs in vivo, it may explain the exclusive usage of CCR5 by HIV-1 from the LTNP [3]. The lack of a strong association of the CCR5Δ32 bp deletion with disease progression in this study is intriguing. Previous work has shown that LTNP who are wild type for CCR5 are virologically and immunologically similar to persons who are heterozygous for the 32 bp deletion [16]. Although the virological and immunological responses of the HIV-1-infected individuals have not been studied, HIV-1-seronegative persons showed a similar spontaneous and induced production of β-chemokines, macrophage-inflammatory protein types 1α and 1β, or regulated upon activation, normal T cell expressed and secreted (data not shown), regardless of their SDF-1 genotypes. This implies that the SDF1-3′A polymorphism does not affect the production of β-chemokines, which could influence the utilization of the CCR5 co-receptor by HIV-1. In summary, it was demonstrated that LTNP in two HIV-1-infected cohorts have a higher frequency of the SDF1-3′A polymorphism compared with the frequencies for HIV-seronegative individuals and progressors. These data suggest that this polymorphism is associated with delayed HIV-1 disease progression. Charlene S. Dezzuttia Patricia C. Guenthnera Timothy A. Greenb Oren J. Cohend Thomas J. Spirac Renu B. Lala