HIV-1 Nef interacts with cellular adaptor protein (AP) complexes and their medium (mu) subunits. However, the role of the dileucine-based sorting motif within Nef in these interactions has been incompletely characterized. Here, yeast two-hybrid assays indicated that HIV-1 Nef interacted not only with the mu subunits of AP-1 and AP-2, but also with that of AP-3. The interactions with mu1 and mu3 were markedly stronger than the interaction with mu2. Leucine residues of the sorting motif were required for the interactions with mu3 and mu2 and contributed to the interaction with mu1. Confocal immunofluorescence microscopy indicated that Nef, AP-1, and AP-3 (but not AP-2) were concentrated in a juxtanuclear region near the cell center, potentially facilitating interaction between Nef and the mu1 and mu3 subunits. However, leucine residues of the sorting motif were not required for this subcellular localization of Nef. These data suggest that the dileucine motif, required for optimal viral replication, functions through interactions with a variety of AP complexes, including AP-3, potentially by recruiting adaptor complexes to subcellular locations specified by additional determinants in the Nef protein.

Potent antiretroviral therapy can reduce human immunodeficiency virus (HIV) in plasma to levels below the limit of detection for up to 2 years, but the extent to which viral replication is suppressed is unknown. To search for ongoing viral replication in 10 patients on combination antiretroviral therapy for up to 1 year, the emergence of genotypic drug resistance across different compartments was studied and correlated with plasma viral RNA levels. In addition, lymph node (LN) mononuclear cells were assayed for the presence of multiply spliced RNA. Population sequencing of HIV-1 pol was done on plasma RNA, peripheral blood mononuclear cell (PBMC) RNA, PBMC DNA, LN RNA, LN DNA, and RNA from virus isolated from PBMCs or LNs. A special effort was made to obtain sequences from patients with undetectable plasma RNA, emphasizing the rapidly emerging lamivudine-associated M184V mutation. Furthermore, concordance of drug resistance mutations across compartments was investigated. No evidence for viral replication was found in patients with plasma HIV RNA levels of <20 copies/ml. In contrast, evolving genotypic drug resistance or the presence of multiply spliced RNA provided evidence for low-level replication in subjects with plasma HIV RNA levels between 20 and 400 copies/ml. All patients failing therapy showed multiple drug resistance mutations in different compartments, and multiply spliced RNA was present upon examination. Concordance of nucleotide sequences from different tissue compartments obtained concurrently from individual patients was high: 98% in the protease and 94% in the reverse transcriptase regions. These findings argue that HIV replication differs significantly between patients on potent antiretroviral therapy with low but detectable viral loads and those with undetectable viral loads.

A dileucine-based protein sorting motif has recently been identified within the C-terminal, solvent-exposed loop of HIV-1 Nef and has been shown to be required for Nef-mediated down-regulation of CD4 and for optimal viral infectivity. Here, we report that mutation of the dileucine motif has no effect on Nef-mediated down-regulation of class I MHC heavy chain. Instead, deletion of an acidic domain just N-terminal of the polyproline helix of the SH3-binding domain significantly impairs this function. These data indicate that down-regulation of class I MHC and CD4 are mechanistically distinct processes. The data also suggest that protein interactions mediated by the acidic domain, rather than by the dileucine motif, may contribute to this function of Nef.

An SH3-binding domain within the Nef protein of primate lentiviruses has been reported to be important to viral replication and infectivity and dispensable for CD4 downregulation, but its precise role remains unclear. This study investigates the effects of mutations in both the polyproline helix and in the hydrophobic pocket that constitute the SH3-binding domain of Nef. The data demonstrate that the well-studied mutation of the central prolines is only partially disruptive to viral infectivity and replication. The central prolines also make a subtle contribution to the efficiency of CD4 downregulation, detectable only using low levels of Nef expression. Mutation of a conserved arginine in the polyproline helix abrogated more completely Nef-mediated enhancement of viral infectivity; this mutation also adversely affected CD4 downregulation at low levels of Nef expression. Only the R77A mutation substantially impaired downregulation of class I MHC. However, mutation of the central prolines and of R77 yielded proteins that were expressed less efficiently than wild-type Nef. The R77A mutant was expressed most poorly, compatible with its defective phenotypes in all assays. Mutations of the hydrophobic pocket were minimally detrimental to both the virologic and the receptor modulatory functions of Nef. Taken together, this analysis suggests that mutations in the SH3-binding domain do not abrogate fully any Nef-associated phenotype in the absence of detrimental effects on protein expression. We suggest that mutations in this domain can introduce incomplete effects caused by subtle impairments to protein expression; these effects may appear selective under certain experimental conditions due to different sensitivities of the assays to the level of Nef expression.

No abstract