Nucleotide sequences of the central portion of gp120, including the third hypervariable (V3) loop, were obtained from lymphocytes cocultivated with SupT1 cells from 29 AIDS patients in Bangui, Central African Republic. These sequences displayed significantly greater diversity (average distance, 23%) than has been previously observed in isolates from comparably restricted geographical areas. Isolates belonging to four major subtypes of HIV-1 were found; the only subtype not represented was the North American/European subtype B. Unlike the situation in Zaire and Uganda, where subtypes A and D account equally for virtually all isolates of HIV-1, the predominant subtypes in the Central African Republic, accounting for two-thirds of the isolates, were subtypes A (10 isolates) and E (9 isolates). Subtype E represents a group of variants that have previously been found only in Thailand. Only one isolate belonging to subtype D was found. Also recovered were two isolates of subtype C, a subtype associated with southern African and Indian isolates but not previously detected in central Africa. These isolates, although clearly clustering with subtype C, formed a distinct subset, differing from one another by 8.8% and from the Indian and South African subtype C isolates by an average of 22.5%. High interpatient, intrasubtype variation was also seen among the CAR subtype A (average pairwise difference, 19.3%) and subtype E (10.9%) isolates. The diversity of V3 sequences in this set has implications for immunization protocols that rely on the recognition of V3. This study underscores the necessity of basing intervention strategies on knowledge of the particular sequences present in the target population or geographical area.

No abstract

Two monoclonal antibodies (MAbs), 42F and 43F, were isolated some 14 months apart from a single long-term survivor of human immunodeficiency virus type 1 (HIV-1) infection. These MAbs were found to be indistinguishable in terms of their isotypes, specificities, affinities, and biological activities. Both 42F and 43F directed substantial antibody-dependent cellular cytotoxicity (ADCC) against cells infected with four divergent lab-adapted strains of HIV-1, but no neutralizing activity against these strains was detectable. The ability of MAbs 42F and 43F, as well as that of MAbs against two other gp120 epitopes, to direct ADCC against uninfected CD4 ؉ cells to which recombinant gp120 SF2 had been adsorbed (i.e., "innocent bystanders") was demonstrated to be less efficient by at least an order of magnitude than their ability to direct ADCC against HIV-1-infected cells. Flow cytometry analyses showed that 42F and 43F also bind to native primary isolate Envs from clades B and E expressed on cell surfaces. By direct binding and competition assays, it was demonstrated that the 42F/43F epitope lies in a domain of gp120 outside the previously described CD4-binding site and V3 loop ADCC epitope clusters. Immunoblot analysis revealed that the 42F/43F epitope is not dependent on disulfide bonds or N-linked glycans in gp120. Epitope mapping of 42F and 43F by binding to linear peptides demonstrated specificity of these MAbs for a sequence of 10 amino acids in the C5 domain comprising residues 491 to 500 (Los Alamos National Laboratory numbering for the HXB2 strain). Thus, 42F and 43F define a new ADCC epitope in gp120. Because of the relative conservation of this epitope and the fact that it appears to have been significantly immunogenic in the individual from which these MAbs were derived, it may prove to be a useful component of HIV vaccines. Furthermore, these MAbs may be used as tools to probe the potential importance of ADCC as an antiviral activity in HIV-1 infection.

No abstract