Background:  We previously described envelope glycoproteins of an HIV-1 isolate adapted in vitro for growth in microglia that acquired a highly fusogenic phenotype, lower CD4 dependence, and higher CD4 affinity, as well as resistance to inhibition by anti-CD4 antibodies. Here, we investigated whether similar phenotypic changes are present in vivo.

Methods:  Envelope clones from the brain and spleen of an HIV-1-infected individual with neurological disease were amplified, cloned, and sequenced. Amino acid changes in V1/V2, V3, and V4 variable regions of gp120, and in the carboxy-terminal heptad repeat domain (HR2) of gp41 ectodomain, were found. Functional clones were then used in cell-to-cell fusion assays to test for CD4 and co-receptor utilization, and for sensitivity to various antibodies and inhibitors.

Results:  Phylogenetic analysis demonstrated clustering of sequences according to the tissue of origin, as expected. Both brain- and spleen-derived envelope clones mediated fusion in cells expressing both CD4 and CCR5, and brain envelopes also used CCR3 very efficiently as co-receptor. We found that brain envelopes had a lower CD4-dependence and higher CD4 affinity, since they were able to mediate fusion in the presence of low levels of CD4 on the target cell membrane, and they were significantly more resistant to blocking by anti-CD4 antibodies than spleen-derived envelopes. In contrast, we observed no difference in CCR5 utilization or in sensitivity to the CCR5 antagonist TAK-779. However, brain-derived envelopes were significantly more resistant than those from spleen to the fusion inhibitor T-1249, and concurrently showed slightly greater fusogenicity.

Conclusions:  Our results confirm an increased affinity for CD4 of brain-derived envelopes that may have originated from in vivo adaptation to replication in microglial cells. More interestingly, we note the presence of envelopes significantly less sensitive to a fusion inhibitor in the brain vs the spleen of an untreated, HIV-1-infected individual. Changes in the interaction with CD4 but not CCR5, or the specific polymorphism found in HR2 in these brain sequences, may be related with the observed phenotypic differences.