The Nef protein of primate immunodeficiency viruses is essential for establishing a highly productive pathogenic infection in vivo. In tissue culture, Nef is not required for infection but enhances viral infectivity. This effect is most pronounced in unstimulated primary lymphocytes and occurs in the early phase of infection prior to viral gene expression. Since Nef expression does not lead to obvious changes in virus composition, it was of interest to analyze whether Nef is incorporated into virus particles. Here, we show that Nef is specifically immunoprecipitated from radioactively labeled human immunodeficiency virus type 1 (HIV-1)-infected cells and virus particle preparations. Quantitative analysis revealed Nef to be incorporated on the order of 10% of reverse transcriptase incorporation, which corresponds to 5 to 10 molecules of Nef per virion. In infected cells, Nef was detected as a full-length 27-kDa protein. In contrast, approximately 50% of particle-associated Nef corresponded to an 18-kDa species which comigrated with the larger product after in vitro cleavage of purified HIV-1 Nef by the viral proteinase. Nef cleavage in particle preparations was completely abolished by a specific inhibitor of HIV-1 proteinase. Most likely, Nef is cleaved concomitantly with viral structural proteins on maturation of virus particles. This cleavage is likely to be functionally significant because it dissociates the conserved core domain from the N-terminal membrane attachment region. Our results suggest that the profound influence of Nef on establishing infection of unstimulated cells in tissue culture and in vivo is mediated by virion-associated Nef which functions in early infection before viral gene expression.

Retroviruses are produced as immature particles containing structural polyproteins, which are subsequently cleaved by the viral proteinase (PR). Extracellular maturation leads to condensation of the spherical core to a capsid shell formed by the capsid (CA) protein, which encases the genomic RNA complexed with nucleocapsid (NC) proteins. CA and NC are separated by a short spacer peptide (spacer peptide 1 [SP1]) on the human immunodeficiency virus type 1 (HIV-1) Gag polyprotein and released by sequential PR-mediated cleavages. To assess the role of individual cleavages in maturation, we constructed point mutations abolishing cleavage at these sites, either alone or in combination. When all three sites between CA and NC were mutated, immature particles containing stable CA-NC were observed, with no apparent effect on other cleavages. Delayed maturation with irregular morphology of the ribonucleoprotein core was observed when cleavage of SP1 from NC was prevented. Blocking the release of SP1 from CA, on the other hand, yielded normal condensation of the ribonucleoprotein core but prevented capsid condensation. A thin, electron-dense layer near the viral membrane was observed in this case, and mutant capsids were significantly less stable against detergent treatment than wild-type HIV-1. We suggest that HIV maturation is a sequential process controlled by the rate of cleavage at individual sites. Initial rapid cleavage at the C terminus of SP1 releases the RNA-binding NC protein and leads to condensation of the ribonucleoprotein core. Subsequently, CA is separated from the membrane by cleavage between the matrix protein and CA, and late release of SP1 from CA is required for capsid condensation.

Infectious retrovirus particles are derived from structural polyproteins which are cleaved by the viral proteinase (PR) during virion morphogenesis. Besides cleaving viral polyproteins, which is essential for infectivity, PR of human immunodeficiency virus (HIV) also cleaves cellular proteins and PR expression causes a pronounced cytotoxic effect. Retroviral PRs are aspartic proteases and contain two copies of the triplet Asp-Thr-Gly in the active center with the threonine adjacent to the catalytic aspartic acid presumed to have an important structural role. We have changed this threonine in HIV type 1 PR to a serine. The purified mutant enzyme had an approximately 5-to 10-fold lower activity against HIV type 1 polyprotein and peptide substrates compared with the wild-type enzyme. It did not induce toxicity on bacterial expression and yielded significantly reduced cleavage of cytoskeletal proteins in vitro. Cleavage of vimentin in mutant-infected T-cell lines was also markedly reduced. Mutant virus did, however, elicit productive infection of several T-cell lines and of primary human lymphocytes with no significant difference in polyprotein cleavage and with similar infection kinetics and titer compared with wild-type virus. The discrepancy between reduced processing in vitro and normal virion maturation can be explained by the observation that reduced activity was due to an increase in K m which may not be relevant at the high substrate concentration in the virus particle. This mutation enables us therefore to dissociate the essential function of PR in viral maturation from its cytotoxic effect.