The emergence of SI T cell-line tropic variants in longterm infected individuals progressing to AIDS is well established. This overt replication of SI clones is thought to be due to an increasing qualitative immune deficiency with ongoing infection. Remarkably, in the three individuals with SI isolates described here, these isolates could be detected from the start ofinfection on. This finding is most easily explained by assuming that these three individuals were infected by carriers of SI viruses, whereas the other 16 individuals were infected by persons with NSI viruses. Indeed, comparison of the viral phenotype of patients 2 and 8 with the phenotypes of the isolates from their donors indicates that both NSI and SI viruses may be transmitted. Nevertheless, transmission of SI viruses may not invariably result in rapid CD4+ T cell depletion in the recipient. In clonal isolation studies we demonstrated that in carriers ofSI viruses HIV -1 clones with an NSI phenotype may coexist next to clones with the dominant SI phenotype. In the case of transmission by an SI virus carrier, therefore, probably both NSI and SI clones will be present in the inoculum. Incidental observations indicate that after infection, SI clones may be selectively cleared, presumably by the hostimmune response, leaving only NSI clones detectable. In subject 17 we found only NSI viruses, although his infecting partner had SI isolates before and after the putative moment of transmission.
Analysis of the phenotype of individual clones from the person who infected subject 17, obtained by isolation under limiting-dilution conditions, revealed a simultaneous presence ofNSI and SI clones in the PBMC of the donor at the putative moment of transmission (not shown). Therefore, selective transmission ofNSI clones cannot be totally excluded as an alternative explanation for the absence of SI viruses in subject 17. However, evidence for suppression of transmitter SI virus in the recipient has been observed in mother-child pairs. The relative inability ofSI viruses to infect monocytes may explain why these variants are preferentially cleared early in infection. Indeed, in the asymptomatic phase after seroconversion, monocytotropic NSI variants were found to be predominant, indicating that only viruses able to establish a low-grade persistent infection of monocytes can survive the vigorous anti-Hlv-I immune response at that time. The temporary disappearance ofSI isolates observed in patient 7 may reflect inadequate suppression of these variants by the immune system. Also in favor of this mechanism are two recent studies providing evidence for an important role ofthe immune system in the initial containment ofHIVI infection. Thus overt persistence ofSI isolates from the beginning of infection onward would be found only in individuals infected with SI viruses in whom the primary anti-HIV-l immune response is defective, as is suggested by the relatively low numbers of CD8+ T cells and the more transient rise in numbers of activated CD8+ T cells observed in individuals with SI viruses. Such an inefficient anti-HIV-1 immune response may be due to a compromised immune system at the time ofexposure, caused possibly by concurrent infections as noted in patient 7.