Several members of the chemokine receptor family are used together with CD4 for HIV-1 entry into target cells. T cell line-tropic (T-tropic) HIV-1 viruses use the chemokine receptor CXCR4 as a co-receptor, whereas macrophage-tropic (M-tropic) primary viruses use CCR5 (refs 2-6). Individuals with defective CCR5 alleles exhibit resistance to HIV-1 infection, suggesting that CCR5 has an important role in vivo in HIV-1 replication. A subset of primary viruses can use CCR3 as well as CCR5 as a co-receptor, but the in vivo contribution of CCR3 to HIV-1 infection and pathogenesis is unknown. HIV-1 infects the central nervous system (CNS) and causes the dementia associated with AIDS. Here we report that the major target cells for HIV-1 infection in the CNS, the microglia, express both CCR3 and CCR5. The CCR3 ligand, eotaxin, and an anti-CCR3 antibody inhibited HIV-1 infection of microglia, as did MIP-1beta, which is a CCR5 ligand. Our results suggest that both CCR3 and CCR5 promote efficient infection of the CNS by HIV-1.

The genetic evolution of human immunodeficiency virus type 1 (HIV-1) in the brain is distinct from that in lymphoid tissues, indicating tissue-specific compartmentalization of the virus. Few primary HIV-1 envelope glycoproteins (Envs) from uncultured brain tissues have been biologically well characterized. In this study, we analyzed 37 full-length env genes from uncultured brain biopsy and blood samples from four patients with AIDS. Phylogenetic analysis of intrapatient sequence sets showed distinct clustering of brain relative to blood env sequences. However, no brain-specific signature sequence was identified. Furthermore, there was no significant difference in the number or positions of N-linked glycosylation sites between brain and blood env sequences. The patterns of coreceptor usage were heterogeneous, with no clear distinction between brain and blood env clones. Nine Envs used CCR5 as a coreceptor, one used CXCR4, and two used both CCR5 and CXCR4 in cell-to-cell fusion assays. Eight Envs could also use CCR3, CCR8, GPR15, STRL33, Apj, and/or GPR1, but these coreceptors did not play a major role in virus entry into microglia. Recognition of epitopes by the 2F5, T30, AG10H9, F105, 17b, and C11 monoclonal antibodies varied among env clones, reflecting genetic and conformational heterogeneity. Envs from two patients contained 28 to 32 N-glycosylation sites in gp120, compared to around 25 in lab strains and well-characterized primary isolates. These results suggest that HIV-1 Envs in brain cannot be distinguished from those in blood on the basis of coreceptor usage or the number or positions of N-glycosylation sites, indicating that other properties underlie neurotropism. The study also demonstrates characteristics of primary HIV-1 Envs from uncultured tissues and implies that Env variants that are glycosylated more extensively than lab strains and well-characterized primary isolates should be considered during development of vaccines and neutralizing antibodies.

Retroviral capsid (CA) proteins contain a uniquely conserved stretch of 20 amino acids which has been named the major homology region (MHR). To examine the role of this region in human immunodeficiency virus type 1 morphogenesis and replication, four highly conserved positions in the MHR were individually altered by site-directed mutagenesis. Conservative substitution of two invariant residues (glutamine 155 and glutamic acid 159) abolished viral replication and significantly reduced the particle-forming ability of the mutant gag gene products. Conservative substitution of the third invariant residue in the MHR (arginine 167) or of an invariably aromatic residue (tyrosine 164) had only a moderate effect. However, removal of the extended side chains of these amino acids by substitution with alanine prevented viral replication and affected virion morphogenesis. The replacement of tyrosine 164 with alanine substantially impaired viral particle production. By contrast, the substitution of arginine 167 with alanine had only a two-to threefold effect on particle yield but led to the formation of aberrant core structures. The MHR mutants which were severely defective for particle production had a dominant negative effect on particle formation by the wild-type Gag product. The role of the MHR in the incorporation of the Gag-Pol precursor was examined by expressing the Gag and Gag-Pol polyproteins individually from separate plasmids. Only when the two precursor polyproteins were coexpressed did processed Gag and Pol products appear in the external medium. The appearance of these products was unaffected or only moderately affected by substitutions in the MHR of the Gag-Pol precursor, suggesting that the mutant Gag-Pol precursors were efficiently incorporated into viral particles. The results of this study indicate that specific residues within the MHR are required both for human immunodeficiency virus type 1 particle assembly and for the correct assembly of the viral core. However, mutant Gag and Gag-Pol polyproteins with substitutions in the MHR retained the ability to interact with wild-type Gag protein. Retroviral particles are released by budding from the plasma membrane of the host cell, thereby acquiring a plasma membrane-derived lipid envelope (40). The internal structural proteins of human immunodeficiency virus type 1 (HIV-1), like those of other retroviruses, are encoded by the viral gag gene and are translated as a polyprotein precursor (Pr55Mas) (47). During translation of gag, ribosomal frameshifting into the pol reading frame occurs at a frequency of 5 to 10% and leads to the synthesis of a Gag-Pol fusion protein (PrlW6QagsPo) (26, 49). HIV-1 particle morphogenesis begins with the assembly of Pr55rag and Prl60WagaPo precursors underneath the cell membrane (11). An increasingly spherical structure is formed that eventually pinches off, resulting in virus particle release (11). The cotranslational attachment of myristic acid at the N terminus of Pr55rag is required for particle assembly (3, 18). Expression of P...

No abstract