Epitope mapping of broadly neutralizing HIV-2 human monoclonal antibodies

Abstract

Recent studies have shown that natural infection by HIV-2 leads to the elicitation of high titers of broadly neutralizing antibodies (NAbs) against primary HIV-2 strains (T. I. de Silva, et al., J. Virol. 86:930-946, 2012; R. Kong, et al., J. Virol. 86:947-960, 2012; G. Ozkaya Sahin, et al., J. Virol. 86:961-971, 2012). Here, we describe the envelope (Env) binding and neutralization properties of 15 anti-HIV-2 human monoclonal antibodies (MAbs), 14 of which were newly generated from 9 chronically infected subjects. All 15 MAbs bound specifically to HIV-2 gp120 monomers and neutralized heterologous primary virus strains HIV-2(7312A) and HIV-2(ST). Ten of 15 MAbs neutralized a third heterologous primary virus strain, HIV-2(UC1). The median 50% inhibitory concentrations (IC(50)s) for these MAbs were surprisingly low, ranging from 0.007 to 0.028 μg/ml. Competitive Env binding studies revealed three MAb competition groups: CG-I, CG-II, and CG-III. Using peptide scanning, site-directed mutagenesis, chimeric Env constructions, and single-cycle virus neutralization assays, we mapped the epitope of CG-I antibodies to a linear region in variable loop 3 (V3), the epitope of CG-II antibodies to a conformational region centered on the carboxy terminus of V4, and the epitope(s) of CG-III antibodies to conformational regions associated with CD4- and coreceptor-binding sites. HIV-2 Env is thus highly immunogenic in vivo and elicits antibodies having diverse epitope specificities, high potency, and wide breadth. In contrast to the HIV-1 Env trimer, which is generally well shielded from antibody binding and neutralization, HIV-2 is surprisingly vulnerable to broadly reactive NAbs. The availability of 15 human MAbs targeting diverse HIV-2 Env epitopes can facilitate comparative studies of HIV/SIV Env structure, function, antigenicity, and immunogenicity.

Fig 1

Pepscan analysis of HIV-2_7312A for 6.10F and 1.4B epitopes. The reactivity of each peptide in the linear (A and C) or cyclic (B and D) peptide library against 6.10F and 1.4B is represented in bar graphs. The peptide numbers are according to the HIV-2_7312A sequence. The sequences of peptides with significant binding are shown, and the most important region is highlighted in red. The peptide reactivity with 1.4B is shown in parentheses (C and D).

Fig 2

Envelope gp120 alignment for HIV-2_1871/5-5, HIV-2_1871/5-10, and HIV-2_7312A. The variable regions (V1/V2, V3, V4, and V5) are indicated by green bars as described previously (13). The cleavage site of V8 protease (10, 27) is indicated by the arrow. The amino acid substitutions between HIV-2_1871/5-5 and HIV-2_{1871/5_10} are highlighted in yellow. The residues (N409 and P411) contributing to the site mutagenesis and alanine scanning on HIV-2_7312A (K307 to I427) are also indicated. The numbering is according to the HIV-2_7312A sequence.

Fig 3

1.7A neutralization against HIV-2_1871/5 V4 and V5 chimeric viruses and site-directed mutants. (A) The V4 and V5 sequences of the parental strains (HIV-2_1871/5-5 and HIV-2_1871/5-10) and chimeras (HIV-2_{1871/5-5_C10V4} and HIV-2_{1871/5-5_C10V5}) are shown, and the regions contributing to the switch are highlighted in blue (original sequence from parental strains) or red (swapped sequence). The V4 sequences of the site-directed mutants are also shown, and the mutations are highlighted in red. The numbering is according to the HIV-2_7312A sequence. (B) Neutralization of the parental strains and chimeras by 1.7A. (C) Neutralization of parental strains and site-directed mutants by 1.7A. The error bars indicate standard error of the mean.

Fig 4

Blocking of sCD4-gp120 binding by 6.10B and 1.1F. Wells with HIV-2_ST gp120 immobilized in 2.6C-coated plates were preincubated with 10 μg/ml HIV-2 MAbs or buffer, washed, and then incubated with sCD4. The percent inhibition of sCD4 binding by MAbs is shown in the bar graph. More than 70% inhibition was considered blocking of sCD4-gp120 binding. Similar results were obtained in at least three independent experiments. The error bars indicate standard error of the mean.

Fig 5

sCD4-induced gp120 binding of 19.11F and 1.4H. Wells with HIV-2_ST gp120 immobilized in 2.6C-coated plates were preincubated with sCD4 (0.5 μg/ml) or buffer, washed, and then incubated with biotinylated MAbs. Similar results were obtained in three independent experiments. The error bars indicate standard error of the mean.

Fig 6

Sequence alignments of V3, the N terminus of C3, V4, and the N terminus of C4 for HIV-2 and SIV strains. The HIV-2 sequences were aligned with HIV-2_7312A, and the SIVsmm/SIVmac sequences were aligned with SIVmac239. The linear 6.10F and 1.4B epitopes on V3 and critical residues involved in 1.7A recognition are indicated in red. The numbering is according to the HIV-2_7312A sequence.

Fig 7

Epitope mapping of the CG-I, CG-II, and CG-III groups on HIV-2 structural models. (A) A model of the gp120 molecule for HIV-2 primary strain 7312A is shown with highlighted V3, V4, CD4bs, CD4i, and α2 regions. Potential N-linked glycans are modeled as pink spheres. (B) Schematic of approximate locations of V3, V4, CD4bs, CD4i, and α2 on the Env trimer.