HIV-1 variability in relation to host defence mechanisms and disease outcome

Abstract

Genetic variability is a major characteristic of HIV-1. Virus variants evolve rapidly, not only globally, but also within single individuals as a result of an error-prone reverse transcriptase, a rapid virus population turnover and selective pressure from the host. Genetic variation translates into biological variation, such as cell tropism, virulence and sensitivity to host immune response. The aim of the present study was to examine HIV-1 variability at a genetic, antigenic and biological level in relation to host defence mechanisms and disease outcome of infected individuals.

The relationship between envelope gp120 V3 sequence variation and seroreactivity was analysed in a cohort of infected Ugandans. Peptides were synthesised on the basis of the patients' own autologous V3 sequence, consensus sequences within the cohort and previously published sequences. A majority of the sera reacted strongest with the autologous V3 peptides. It was also possible to differentiate antibody populations reactive with distinct variants within the virus population in single individuals.

Next, kinetics of development of HIV-1 specific IgG subclasses in sera of infants born to seropositive mothers was evaluated as a diagnostic and prognostic tool. In many newborns infection could be diagnosed between three to six months of age by de novo synthesis of either IgG1 or IgG3 reactive with HIV-1 specific peptides. Interestingly, it was also found that children with rapid disease progression failed to produce antibodies reactive with V3 peptides.

The correlation between disease progression, HIV-1 sequence evolution and specific immune response in infected children was also studied. Children with slow disease progression were found to accumulate non-synonymous substitutions faster than children with rapid disease progression. Furthermore, the ratio of synonymous/non-synonymous substitutions showed an inverse correlation with the seroreactivity against peptides representing autologous V3 sequences. These data indicate that V3 antibodies exert a positive selective pressure for amino acid change in the V3 loop and that a stronger selective pressure is associated with slower rate of disease progression. Certain chemokine receptors were recently identified as coreceptors for HIV-1 cell entry. For this reason we examined B-chemokine sensitivity and V3 sequence variation of primary isolates with defined biological phenotype, i.e. slow/low, nonsyncytium-inducing (NSI) or rapid/high, syncytium-inducing (SI), sequentially obtained from patients with progressive disease. A switch in virus phenotype from NSI to SI was paralleled by a loss of sensitivity to inhibition by B chemokines. Furthermore, B-chemokine sensitivity correlated with V3 amino acid changes, which have previously been shown to be determinants for biological phenotype. In a separate study we found that loss of sensitivity to inhibition by B-chemokines correlated with a shift in virus coreceptor usage. Isolates with NSI phenotype used CCR5 as coreceptor, while SI isolates used CXCR4. Furthermore, many SI isolates displayed a broadened coreceptor usage and could use CCR3 and CCR5 in addition to CXCR4.

A NSI to SI switch is a sign of poor prognosis, but still many HIV-1 infected individuals develop AIDS while carrying NSI virus. To study virus coreceptor usage and B-chemokine sensitivity in relation to disease progression in such patients we analysed sequential NSI isolates obtained before and after development of AIDS. NSI isolates obtained from AIDS patients were found to remain CCR5 dependent and relatively sensitive to B-chemokine inhibition. However, NSI isolates from patients with very late stage disease tended to show reduced B-chemokine sensitivity. The length of the gpl20 V2 domain has, in addition to V3 amino acid sequence, been suggested to influence biological phenotype. Therefore, we analysed the V2 length of sequential HIV-l isolates from patients who did, or did not, display a NSI to SI switch during disease progression. Isolates from patients with stable NSI virus phenotype had shorter V2 regions than isolates obtained from switch patients. Similarly, CCR5 dependent isolates had shorter V2 regions than isolates with the ability to utilize CXCR4, but there was no difference in V2 length between SI isolates that were monotropic for CXCR4 or multitropic for several coreceptors.