No abstract

Two distinct regions or epitope clusters of human immunodeficiency virus type 1 (HIV-1) gp120 have been shown to elicit neutralizing antibodies: the V3 loop and the CD4-binding site. We have isolated neutralizing human monoclonal antibodies (HuMAbs) against conserved epitopes in both of these regions. In this study, we demonstrate that an equimolar mixture of two of these HuMAbs, one directed against the V3 loop and the other against the CD4-binding site, neutralizes HIV-1 at much lower concentrations than does either of the individual HuMAbs. Mathematical analysis of this effect suggests cooperative neutralization of HIV-1 by the two HuMAbs and demonstrates a high level of synergy, with combination indices (CIs) of 0.07 and 0.16 for 90% neutralization of the MN and SF-2 strains, respectively. The dose reduction indices (DRIs) for each of the two HuMAbs at 99% neutralization range approximately from 10 to 150. A possible mechanism for this synergism is suggested by binding studies with recombinant gp160 of the MN strain; these show enhanced binding of the anti-CD4 binding site HuMAb in the presence of the anti-V3 loop HuMAb. These results demonstrate the advantage of including both V3 loop and CD4-binding site epitopes in a vaccine against HIV-1 and indicate that combinations of HuMAbs against these two sites may be particularly effective in passive immunotherapy against the virus.

A human monoclonal antibody (HuMAb), 5145A, against HIV-1 gp120 was isolated from an asymptomatic, seropositive hemophiliac. The epitope of this HuMAb was destroyed by reduction of gp120 disulfide bonds, but not by removal of N-linked carbohydrates. This epitope overlaps the CD4-binding site of gp120, because binding of 5145A to gp120 is inhibited by soluble CD4 and by 1125H, a previously described HuMAb directed toward the CD4-binding site. However, the 5145A epitope differs from those of 1125H and other anti-CD4-binding site HuMAbs previously described, as documented by the viral strain specificity of 5145A and its reactivity with a panel of gp120 mutants. Specifically, 5145A reacted with 14 of 15 HIV-1 isolates tested, including 9 isolates from the Central African Republic, 6 of which were not recognized by 1125H. Partial epitope mapping of 5145A, using a series of gp120 mutants, demonstrated its lack of sensitivity to mutations in residues 257 and 427, contrasting with a marked sensitivity to mutations in residues 368 and 370. This pattern of reactivity distinguishes its epitope from that of any HuMAb against the CD4-binding site region described to date. In addition, 5145A exhibited potent and essentially equivalent neutralization of the MN, SF-2, IIIB, and RF strains and possessed significant neutralizing activity against three of three African strains tested. Finally, 5145A synergistically neutralized the MN and SF-2 strains of HIV-1 when combined with 4117C, a HuMAb against the V3 loop. The broad strain specificity and potent neutralizing activity of 5145A, together with its ability to synergize with an anti-V3 loop HuMAb in neutralizing HIV-1, indicate that 5145A has excellent potential as a passive immunotherapeutic agent against HIV-1.