Dendritic cells, exposed to primary, mixed phenotype HIV-1 isolates preferentially, but not exclusively, replicate CCR5-using clones

HIV isolates can be phenotypically distinguished by their requirement for either CCR5 (R5 virus) or CXCR4 (X4 virus) co-receptors to infect CD4 leukocytes [1]. During the early phase of HIV infection, the entire ‘quasispecies’ usually displays the R5 phenotype, even if a ‘mixed phenotype’ virus (containing both R5 and X4 clones) was originally transmitted [2]. This observation could be explained by the positive selection of R5 clones in the primary target cells: mucosal dendritic cells and interacting CD4 T cells [3,4]. Alternatively, X4 clones may be eliminated by the immune system that is still intact early after infection. Using R5 and X4 HIV strains/clones separately, isolated dendritic cells seem selectively susceptible to R5 variants, but X4 viruses can be rescued by co-culture with CD4 T cells [5,6] (G. Vanham, L. Penne, H. Allemeersch, et al., 2000, in preparation). In order to simulate in-vivo conditions, we used primary mixed phenotype isolates and compared their capacity to infect peripheral blood mononuclear cells (PBMC) (as a ‘universally permissive’ system), monocyte-derived dendritic cells, which closely resemble mucosal intertstitial dendritic cells [7], and co-cultures of monocyte-derived dendritic cells with autologous resting T cells. Compared with activated PBMC, dendritic cells were consistently less efficient in supporting the replication of HIV strains/clones (lower overall titres in dendritic cells), but the addition of resting T cells boosted viral production (see Table 1).Table 1: Differential selection of R5 or X4 phenotype from primary HIV-1 isolates and clones in three types of primary cell cultures. As expected, both R5 and X4 biological clones of the mixed phenotype isolate VI 820 readily replicated in PBMC; dendritic cells alone were selectively susceptible to the R5 clone but the addition of CD4 T cells rescued the X4 clone. In the four different isolates used, the proportion of X4 clones varied between 27 and 81% (see the R5/X4 ratios obtained after culture in activated PBMC). We assumed that activated PBMC are able to replicate all infectious virus, and we observed that PBMC and the U87 are roughly equally sensitive to R5 respectively X4 viruses. On the basis of the observed infectious titres and R5 : X4 ratios, we calculated the 50% PBMC infectious doses (PBMC ID50) required to infect dendritic cells alone or dendritic cells in co-culture with CD4 T cells (see Table 1, footnote f). Both R5 and X4 varied 1000-fold in their virulence for dendritic cells (R5: 1.36–2280 and X4 1.96–1650 ‘PBMC ID50’ required for infection). Nevertheless, R5 viruses were more virulent (required much less infectious doses) to infect dendritic cells in most virus–donor combinations. In dendritic cells from donor 3, the virulence of R5 and X4 viruses from VI 820 and VI 1363 was fairly comparable. A mixture of highly virulent X4 and avirulent R5 viruses was not observed. Differences in virulence were reduced after the addition of CD4 T cells. In conclusion, we used a model for interstitial dendritic cells with or without autologous resting T cells to simulate the primary targets of sexual HIV transmission. For the first time, we examined whether these cells were selectively permissive to R5 viruses, if they were in competition with X4 viruses in the context of a primary isolate. Our results suggest that, compared with PBMC, dendritic cells are more susceptible to R5 viruses. The simultaneous presence of CD4 T cells reduces this selection, but fails to reverse it. The extent of R5 preference, however, was both isolate- and donor-dependent. Clearly, this positive selection mechanism alone cannot explain the strong predominance of R5 over X4 clones, observed in vivo, early after transmission. Therefore, an additional (immune-mediated?) negative selection against the X4 phenotype may also play an important part in early R5 dominance. Acknowledgement The authors would like to thank Dr G. Mertens from the Antwerp Red Cross Blood Transfusion Centre for allowing access to buffy coats from healthy blood donors. Guido Vanhama David Davisb Betty Willemsb Lieve Pennea Luc Kestensa Wouter Janssensb Guido van der Groenb