Structural definition of a conserved neutralization epitope on HIV-1 gp120

The remarkable diversity, glycosylation and conformational flexibility of the human immunodeficiency virus type 1 (HIV-1) envelope (Env), including substantial rearrangement of the gp120 glycoprotein upon binding the CD4 receptor, allow it to evade antibody-mediated neutralization. Despite this complexity, the HIV-1 Env must retain conserved determinants that mediate CD4 binding. To evaluate how these determinants might provide opportunities for antibody recognition, we created variants of gp120 stabilized in the CD4-bound state, assessed binding of CD4 and of receptor-binding-site antibodies, and determined the structure at 2.3 Å resolution of the broadly neutralizing antibody b12 in complex with gp120. b12 binds to a conformationally invariant surface that overlaps a distinct subset of the CD4-binding site. This surface is involved in the metastable attachment of CD4, before the gp120 rearrangement required for stable engagement. A site of vulnerability, related to a functional requirement for efficient association with CD4, can therefore be targeted by antibody to neutralize HIV-1. Supplementary information The online version of this article (doi:10.1038/nature05580) contains supplementary material, which is available to authorized users.

We analyzed Cβ-Cβ distances within the context of the S375W, T257S structure to construct novel interdomain disulfides. Molecular modeling suggested four interdomain disulfides could be made to novel disulfides were constructed in the context of the S375W, T257S core. In addition, we tested a 123-431 disulfide that tied together the second and third strands of the bridging sheet as well as an A433M cavity-altering mutation. Structural analysis of these stabilized gp120 variants in complex with CD4 and Fab 17b at 2.0-2.2 Å resolution showed that four of five disulfides formed (except 231-268), and that M95W and A433M, in the context of the 96-275 disulfide, induced minimal structural perturbation.
To increase stabilization, we tested disulfide and cavity-altering combinations. Folding difficulties were encountered with combinations of three or four interdomain disulfides (supplementary Table 2). A four-disulfide core did not fold, and only two combination of three disulfides produced adequate levels of folded protein (supplementary Table 2). We analyzed one of these three-disulfide variants with linkages at 96-275, 109-428, and 123-431, bound to CD4 and 17b at 2.8 Å resolution. We also analyzed a two-disulfide variant with linkages at 96-275 and 109-428 at 2.5 Å resolution. In both structures, all potential disulfides formed.
a , Structural details of stabilization. The ribbon diagram depicts the gp120 inner domain (gray), outer domain (red), and bridging sheet (blue), along with stick models of introduced disulfides (with carbon atoms in green) and cavity-altering substitutions (with carbon atoms in magenta). Inset boxes show electron density (2Fo-Fc, contoured at 1σ) corresponding to each mutational substitution. b, Unliganded SIV gp120, with polypeptide in ribbon diagram colored the same as a. Positions corresponding to introduced cysteines are green, with disulfide connections that only form in the CD4-bound conformation highlighted with yellow dashes and with associated Cα distances. As can be seen, each of the four disulfides is structurally incompatible with the conformation of the unliganded SIV gp120. c, Antigenic analysis of conformational stabilization. The vertical axis corresponds to the change in entropy of each particular gp120 variant upon binding to CD4. In the left panel, the horizontal axis corresponds to the onrate for the CD4-induced (CD4i) antibodies, 17b and m6 (in blue and red, respectively). In the middle panel, the horizontal axis corresponds to the on-rate of CD4. In the right panel, the horizontal axis shows the aggregate change in ∆∆G for the 10 CD4-binding-site (CD4BS) antibodies listed in Table 1. In ∆∆G calculations, a maximal change in affinity of 10 5 was used. All three of the heavy chain complementarity-determining regions (CDRs) of b12 make extensive contact with gp120. The CDR H1 (197 Å 2 ) uses Arg 28 and Asn 31 to make polar interactions, and Phe 32 to pack hydrophobically over the N-terminal portion of the CD4-binding loop (β15-α3 of gp120). The CDR H2 (245 Å 2 ) inserts between the CD4-binding loop and β20/21, positioning Tyr 53 into a hydrophobic cleft at the C-terminus of the CD4-binding loop. The base of the CDR H3 (276 Å 2 ) also makes extensive contacts with the CD4-binding loop, with both polar (e.g. three hydrogen bonds) and hydrophobic (e.g. Tyr 98 stacks over gp120 proline residue 369) interactions. Meanwhile, the extended CDR H3 tip projects towards the glycosylated silent face, with Trp 100 at the CDR H3 tip sandwiched between gp120 Arg 419 and Asn 386, which is N-glycosylated (in a rare antibody:glycan interaction, the glycan attached to Asn 386 makes a hydrogen bond to the backbone amide of Trp 100 -"NAG" above).
a , Details from a b12 perspective. The gp120 molecular surface is depicted along with specific residues of b12 and gp120 as stick and line models, respectively. The gp120 molecular surface is colored purple for the surface associated with the CD4-binding loop, red for the surface associated with the rest of green (CDR H3) for b12 and red (outer domain) and purple (CD4-binding loop) for gp120, with nitrogen atoms in blue, oxygen atoms in red, and select water molecules in blue. Atoms involved in hydrogen bonds are connected by dashed lines. Labels specifying gp120 residues are italicized, with those specifying b12 residues in normal script. Arg 28 in the b12 CDR H1 assumes two conformations, both of which are shown.
Inset of the same orientation shows the CDR H1, H2 and H3 loops (depicted in ribbon diagram) arrayed around the CD4-binding loop (purple surface). b, Details from a gp120 perspective. The orientation has been rotated 180º about a vertical axis. The b12 molecular surface is depicted along with highlighted residues of gp120 and b12 as stick and line models, respectively. The b12 molecular surface is colored orange-yellow for the surface associated with the CDR H1, cyan for the surface associated with the CDR H2, green for the surface associated with the CDR H3, and gray otherwise. Stick models are colored as specified in a. Labels specifying gp120 residues are italicized, with those specifying b12 residues in normal script. Inset of the same orientation shows the CDR H1, H2 and H3 surfaces grasping onto the CD4-binding loop, which is displayed as a purple ribbon diagram. to the "CD4-binding loop" (containing secondary structural elements, β15 and α3) and the "outer domainexit loop" (the loop between β24 and α5) are denoted and labeled in purple and red, respectively.
As expected from the use of the conformationally constrained gp120 in crystallization, the overall b12-bound conformation was similar to that induced by CD4, though with important differences. These differences relate primarily to the regions, which in the CD4-bound state form the bridging sheet, and to proximal portions of the inner domain.
In the bridging sheet (β2, β3, β20 and β21), significant differences were observed between the b12-bound conformation and the unliganded and CD4-bound states. β2 and β3 are not ordered sufficiently in the b12-bound conformation to be resolved. Meanwhile, β20 proceeds in the same direction as the preceding β19, so that in the b12-conformation the β20-β21 ribbon is turned 90º with respect to unliganded and CD4-bound conformations (Fig. 2).
In the inner domain around the core N-and C-termini, the b12 conformation is different from the The conformation of the outer domain, which serves as the central binding site for b12 and CD4, is remarkably well preserved between unliganded, b12-and CD4-bound states (Fig. 2). This similarity is reflected in a Cα RMSD of only 1.3 Å between outer domains of b12-and CD4-bound states. The differences in outer domain are greater between unliganded SIV gp120 and liganded HIV-1 gp120 (average SIV difference is 4.36 Å). The CD4-binding loop, for example, is significantly different in the unliganded SIV structure and much more similar in b12-and CD4-bound states. Interestingly, the Cα RMSD for residues of the outer domain in contact with b12 and CD4 is 2.0 Å, significantly higher than non-contact residues (Cα RMSD of 1.1 Å) (significance of contact versus non-contact RMSD, P < 0.0019), suggesting that the few differences observed in the HIV-1 outer domain are induced by b12 or CD4 binding.
Overall, these results confirm the conformational flexibility in liganded versus unliganded gp120 The exploitation of D1D2-Igαtp avidity can only occur if it is the analyte. Wild-type core and OD1 variants of gp120 were immobolized onto CM5 sensor chips at densities of 430 and 385 RU, respectively. To minimize nonspecific binding of D1D2-Igαtp, alcohol dehydrogenase (S. cerevisiae, Sigma) was immobilized onto the reference surface and 0.1% carboxymethyl dextran was added to the buffer. Increasing concentrations of D1D2-Igαtp were allowed to bind for 2 min and then to dissociate for 2 min, with a flow rate of 30uL/min. Concentrations of D1D2-Igαtp (right panels) were 0.62 nM (cyan), 1.25 nM (red), 2.50 nM (green) and 5.00 nM (blue) for wild-type core (top row), and 1.25 nM (cyan), 2.50 nM (red), 5.00 nM (green) and 10.0 nM (blue) for OD1 (bottom row).