Research Article

A Potent and Broad Neutralizing Antibody Recognizes and Penetrates the HIV Glycan Shield

See allHide authors and affiliations

Science  25 Nov 2011:
Vol. 334, Issue 6059, pp. 1097-1103
DOI: 10.1126/science.1213256


The HIV envelope (Env) protein gp120 is protected from antibody recognition by a dense glycan shield. However, several of the recently identified PGT broadly neutralizing antibodies appear to interact directly with the HIV glycan coat. Crystal structures of antigen-binding fragments (Fabs) PGT 127 and 128 with Man9 at 1.65 and 1.29 angstrom resolution, respectively, and glycan binding data delineate a specific high mannose-binding site. Fab PGT 128 complexed with a fully glycosylated gp120 outer domain at 3.25 angstroms reveals that the antibody penetrates the glycan shield and recognizes two conserved glycans as well as a short β-strand segment of the gp120 V3 loop, accounting for its high binding affinity and broad specificify. Furthermore, our data suggest that the high neutralization potency of PGT 127 and 128 immunoglobulin Gs may be mediated by cross-linking Env trimers on the viral surface.

Viruses have evolved a variety of mechanisms to escape antibody recognition, many of which involve features of the viral surface proteins, such as high variability, steric occlusion, and glycan coating. For HIV, the dense shield of glycans (1, 2) that decorate the viral Env protein was once believed to be refractory to antibody recognition, masking conserved functionally significant protein epitopes for which greater exposure would result in increased susceptibility to antibody neutralization. However, the broadly neutralizing monoclonal antibody (bnmAb) 2G12 and several of the recently described PGT antibodies appear to bind directly to the HIV glycan coat. Although carbohydrate-protein interactions are typically weak (3), 2G12 recognizes terminal dimannose (Manα1,2Man) moieties on oligomannose glycans, using an unusual domain-exchanged antibody structure that creates a multivalent binding surface that enhances the affinity of the interaction through avidity effects (4). However, although 2G12 neutralizes clade B isolates broadly, it is less effective against other clades, particularly clade C viruses, which have a somewhat different oligomannose glycan arrangement than clade B viruses. In contrast, we have recently isolated six bnmAbs (PGTs 125 to 128, 130, and 131) that bind specifically to the Man8/9 glycans on gp120 and potently neutralize across clades (5). PGT 128, the broadest of these antibodies, neutralizes over 70% of globally circulating viruses and is, on average, an order of magnitude more potent than the recently described PG9, PG16, VRC01, and PGV04 (also known as VRC-PG04) bnmAbs (58) and is two orders of magnitude more potent than prototype bnmAbs described earlier (6, 9).

The neutralization potency exhibited by the PGT class of antibodies suggests that they may provide protection at relatively low serum concentrations. Hence, the epitopes recognized by these antibodies may be good vaccine targets if appropriate immunogens can be designed.

Crystal structures of PGTs 127 and 128 bound to Man9. To gain a structural understanding of the specificity for Man8/9 glycans by PGTs 127 and 128, we first determined crystal structures of the Fabs of PGTs 127 and 128 with a synthetic Man9 glycan lacking the core N-acetylglucosamine (GlcNAc) moieties at 1.65 and 1.29 Å resolution, respectively (table S1). The bound glycan is well ordered, except for the terminal mannose residue of the D2 arm (Fig. 1 and figs. S1 and S2A). The 127/Man9 and 128/Man9 structures show a similar conformation for the glycan (fig. S1), demonstrating a conserved mode of recognition by these clonally related antibodies.

Fig. 1

Binding mode of Man9 by antibody PGT 128 revealed by the high-resolution crystal structure of the complex. (A) Front (top) and side (bottom) views of PGT 128 Fab with bound Man9 glycan. The light and heavy chains are depicted as gray and magenta ribbons, respectively, and the glycan as yellow (carbons) and red (oxygen) ball-and-stick models. (B) Close-up view of glycan binding site of PGT 128 showing electron density (2Fo-Fc) at 1.0σ for glycan and associated water molecules. Water molecules are shown as red spheres with the electron density colored red for waters that bridge mannose residues and green for waters in the glycan-antibody interface. (C) Detailed view of the interactions in the Man9 glycan-binding site at the interface of CDRs H2, H3, L3, and FR2. Tryptophan (VH W52f, W56, and W100e and VL W95) and Asn/Asp (VH N53 and VL N94 and D95a) residues from the Fab are enriched at the interface and dominate the interactions with the glycan. The D1 arm is bound by residues in the six–amino acid CDR H2 insert and VH FR2. The D3 arm is bound by residues within CDR L3. Hydrogen bonds are shown as green dashes.

Analysis of these crystal structures reveals the origin of their specificity for Man8/9 glycans. The terminal mannose residues of both the D1 and D3 arms, which are only present on Man8/9 glycans (Fig. 1B, Fig. 2A, and fig. S2A), are heavily contacted, forming 11 of the 16 total hydrogen-bonding interactions with the antibody (table S2). This specificity for glycans is consistent with glycan array data showing binding of PGT 127 and 128 to Man8 and Man9, but not to monoglucosylated Man9 N-glycans (fig. S3A), and with glycosidase inhibitor specificity profiling (fig. S3B). The D3 arm of Man8/9 is bound by complementarity-determining region (CDR) L3 residues Asn94, Trp95, and Asp95a (Fig. 1C and table S2). Several ordered water molecules are present in the glycan-antibody interface and also bridge the mannose residues (Fig. 1C), as previously noted as key features of other antibody-carbohydrate interfaces (10). In addition, two hydrogen bonds are observed between mannose residues that reside on different arms. The individual dihedrals of the glycan are in stable low-energy conformations (fig. S2), which are consistent with a high-affinity interaction. PGTs 125 to 128 contain a six-residue insertion in CDR H2 (5), which was probably introduced somatically during affinity maturation (11). This insertion mediates an outward displacement of the C′′ β-strand of VH (fig. S4) and promotes contact with the Man9 D1 arm (Fig. 1 and table S2). Deletion of the insert resulted in diminished gp120 binding and neutralization potency for PGTs 127 and 128 (Fig. 3C). However, a reciprocal swap of the PGT 127 and 128 insert residues did not result in a complete interchange of their binding to gp120 or their neutralization profiles (fig. S5), indicating that the insert does not solely account for their differences in breadth and potency (12, 13). The high affinity for Man9 is explained by its extensive buried surface area (394 Å2 for PGT 128 and 352 Å2 for PGT 127) (table S2) in a binding mode that differs from other carbohydrate-binding antibodies or lectins and notably from 2G12, which contacts only the terminal Manα1,2Man moieties of Man9, particularly at the tip of the D1 arm (4).

Fig. 2

Crystal structure of PGT 128 Fab in complex with an engineered glycosylated gp120 outer domain (eODmV3). (A) Overall view of PGT 128/eODmV3. PGT 128 Fab heavy and light chains are depicted as in Fig. 1. The eODmV3 is shown in a green cartoon ribbon representation. Glycans are depicted in a ball-and-stick representation with carbons in yellow, oxygens in red and, nitrogens in blue. PGT 128 binds the N332 glycan in the primary glycan-binding site by interactions with the terminal mannose residues of the D1 and D3 arms. The mode of interaction and site of recognition are identical to that visualized in the high-resolution Man9 complex. The secondary glycan-binding site recognizes the N301 glycan. (B) Close-up view of the secondary glycan interaction site and contacts made with N301 glycan. The mannose residues of the N301 glycan splay out around FR3 residues VH D72, T73, P74, and K75. The terminal mannose residues are not ordered in the electron density. (C) Close-up view of V3 interactions with CDR H3. The C terminus of V3, residues Ile323 to Asp325, makes van der Waals and hydrogen-bonding contacts to one side of an extended β-strand region of PGT 128 CDR H3, which includes L100-D100d. The V3 base is intercalated between the apex of the CDR H2 insert (Y52e and W52f) and CDR H3.

Fig. 3

Effect of PGT 128 paratope mutations in the individual glycan subsites on neutralization of HIV-1JR-FL and glycan binding. Binding of PGT 128 mutants to gp120 was tested by enzyme-linked immunosorbent assay (left panel) or to glycans on the high-mannose glycan microarray (right panel). (A) Mutation of select residues in the primary glycan-binding site (Man8/9) that recognizes the N332 glycan. Residues (HC, heavy chain; LC, light chain) that disrupt the formation of the hydrophobic core of the binding site (VH K100gA and W100eA, and VL W95A) or disrupt hydrogen bonding to terminal mannose residues (VH H59A and VL D95aA) compromise neutralization (middle panel), as well as gp120 and glycan binding. (B) Mutation of select residues interacting with the secondary glycan-binding site that recognizes the N301 glycan. Mutation of VH H52aA results in a decrease in gp120 binding and neutralization, whereas disruption of the CDR H1-H2 disulfide (VH C32A or C52bA, or a double mutant) greatly compromises both gp120 binding and neutralization. There is much less effect on the glycan array, which primarily reflects binding to the primary glycan-binding site. A complete list of paratope mapping, as well as the effect on gp120 binding, is provided in table S3. (C) Contribution of the six-residue CDR H2 insert to neutralization and glycan binding. PGT 128 retains the ability to bind Man8/9 and to neutralize to a lesser extent on deletion of the insert, whereas PGT 127 no longer neutralizes, although still has some ability to bind Man8/9. Swapping of the insert between 127 and 128 allows 128 to retain some binding and neutralization, but substantially reduces binding and abrogates neutralization when the PGT 128 H2 insert is transplanted onto PGT 127 (fig. S5).

Crystal structure of PGT 128 bound to a glycosylated gp120 outer domain. To structurally define the epitope recognized by PGT 128 in the context of gp120, we cocrystallized Fab PGT 128 with a glycosylated gp120 outer domain construct containing a truncated V3 loop [engineered outer domain mini-V3 (eODmV3) (14)] (fig. S6). PGT 128 binds to eODmV3 with an apparent affinity of 46 nM, which is about eightfold less than its interaction with the HIV-1JR-FL gp120 core with a full-length V3 (fig. S7). The V3 loop truncation did not affect PGT 128 binding (fig. S7). The purified complex was homogenous as assessed by size exclusion chromatography multiangle light scattering (fig. S8), and the crystal structure was solved by molecular replacement and refined at 3.25 Å resolution to an Rcryst of 0.21 and Rfree of 0.26 (table S1).

The crystal structure unexpectedly revealed that PGT 128 engages two different glycans, as well as the C-terminal end of the V3 loop, within the binding site (Fig. 2A). The primary glycan-binding site is occupied by the high-mannose glycan attached to N332 (Man8/9GlcNAc2), whereas a secondary glycan-binding site is occupied by electron density associated with N301 (Fig. 2B). The secondary glycan-binding site is focused on the core pentasaccharide attached to N301, as only the Man5GlcNAc2 portion of the glycan is visible in the density map. The two GlcNAc residues bind atop the CDR H1-H2 disulfide in a favorable hydrophobic interaction; hydrogen bonds are formed between the backbone amide and carbonyl of Ala52c and the N-acetyl and O3 hydroxyl of the first Asn-linked GlcNAc. FR3 and CDR H1 residues form the contact site for the mannose sugars (Fig. 2B and table S4).

The CDR H3 apex contacts the V3 loop on the gp120 outer domain (Fig. 2C). The C-terminal residues of V3, Ile323-Asp325, are bound in a groove between CDRs H2 and H3. Leu100-Asp100d in CDR H3 adopt a β-strand conformation that is primed for β-sheet–type interactions with the gp120 V3 loop (15).

To assess the importance of the individual glycan binding sites for epitope recognition, we tested a series of antibody variants containing single amino acid substitutions in each subsite. Mutations in the primary glycan binding site (N332) compromised neutralization, gp120 binding, and binding to Man8/9 on the glycan array (Fig. 3A and table S3). Although numerous interactions are made with the glycan, including a total of 16 hydrogen bonds, disruption of the bidentate interaction between Man D3 and CDR L3 Asp95a resulted in a loss of gp120 and glycan binding and neutralizing activity (Fig. 3A). The mutation of residues involved in the secondary site (N301), particularly the H1-H2 disulfide, also resulted in a loss of gp120 binding and virus neutralization (Fig. 3B and table S3). Nevertheless, the affinity of this secondary site (N301) was too low to detect directly by glycan array, as evidenced by the lack of glycan binding capacity by a primary glycan-binding site loss-of-function variant (VL Asp95a→Ala). Also, the mutation of FR3 and CDR H1 residues that interact with the mannose residues in the secondary binding site generally had little to no effect on neutralization by PGT 128, suggesting that these interactions are not as crucial as the GlcNAc interactions (table S3). Notwithstanding, the N301 glycan is required for high-affinity binding to gp120 and neutralization. The importance of the N332 and N301 glycans in forming the PGT 128 epitope was confirmed by alanine scanning mutagenesis, where substitutions at positions 332 and 301 resulted in the loss of neutralizing activity against most isolates tested (table S5). PGT 127 displayed a similar glycan reactivity profile as PGT 128 against most isolates, suggesting that the two antibodies share a similar conserved mode of epitope recognition.

The N301 and N332 glycans are 93 and 73% conserved among HIV isolates (fig. S9), respectively, which accounts for the ability of PGT 128 to neutralize 72% of circulating viruses. In the HIV-1JR-CSF strain, individual alanine mutations at positions 332 and 301 had little to no effect on neutralization by PGT 128 (5), but various combinations of double glycan substitutions, which included the nearby N295 as well as N332 and N301, completely abolished neutralizing activity (fig. S10). These results suggest that, for isolate JR-CSF, the epitope may be more promiscuous and accommodate antibody binding to two out of three glycans. The PGT 128 requirement for two closely spaced N-linked glycans (table S5 and fig. S10) probably accounts for its lack of reactivity with self-glycoproteins displaying single Man8/9GlcNAc2 sugars (fig. S11) and for the resistance of HIV-2 and SIV viruses to neutralization (fig. S12). Specific interactions with V3 were more difficult to investigate, because the V3 contacts with PGT 128 CDR H3 are primarily mediated through backbone hydrogen bonding and van der Waals interactions that are tolerant of side-chain variation, as seen for the V3 crown-specific antibody 447-52D (16). Thus, three discontinuous sites on the gp120 outer domain (449 Å2 from N332, 328 Å2 from N301, and 305 Å2 from V3) combine to form 1081 Å2 of buried surface area (table S4), which is similar in overall size to other anti-HIV bnmAbs VRC01 and PGV04/VRC-PG04, which bury 1229 Å2 and 1080 Å2 on the CD4 binding site of core gp120, respectively (8, 17).

The PGT 128 epitope is accessible on the HIV trimer. To gain a structural understanding of the epitope recognized by PGT 128 in the context of the HIV trimer, we generated a negative stain reconstruction of a soluble, partially deglycosylated 664G trimer in complex with PGT 128 Fab. This engineered Env trimer incorporates stabilizing mutations that allow it to maintain integrity upon deglycosylation (1823). Three Fabs bind to the trimer with no close contacts to neighboring gp120 protomers, indicating that the outer domain epitope is accessible and exposed (Fig. 4 and fig. S13). Fitting of the crystal structure of the PGT 128/eODmV3 complex into the reconstruction also revealed that the V3 base (Fig. 4A and fig. S13D) is surface-exposed, but below and adjacent to the density corresponding to the V1/V2 loops. No large-scale conformational changes in the trimer appear to take place upon Fab binding (Fig. 4B). Thus, the elements that form the PGT 128 epitope are almost directly opposite the CD4bs on gp120 and appear to be accessible and not subject to steric occlusion in the trimer.

Fig. 4

Negative stain EM reconstruction of partially deglycosylated soluble 664G Env trimer in complex with PGT 128 Fab. Soluble (664G) Env trimer was complexed with Fab PGT 128 and treated with Endo H to remove nonprotected glycans. (A) Coordinates of the 128/eODmV3 complex structure fitted into the reconstruction density (blue). Overhead (top) and side (bottom) views show the fit of the crystal structure to the EM density (supporting online material). Fab 128, depicted as blue (heavy) and white (light), and eODmV3 (red) are depicted in schematic backbone representation with glycans shown as yellow sticks. (B) Reconstruction density overlayed with cryoelectron tomographic reconstruction of native, unliganded trimer (yellow) (30). The putative location of V1/V2 is indicated. The V3 base, N301, and N332 are exposed on the surface of the outer domain and slightly below the trimer apex, which corresponds to the location of the V1/V2 loops. The PGT 128 epitope is located approximately on the opposite side of gp120 from the CD4bs (fig. S13C).

Mechanism of exceptional neutralization potency by PGTs 127 and 128. Because a strong correlation has been described in other systems between antibody apparent binding affinity for native Env trimers expressed on the cell surface and neutralization potency (2427), we first compared the neutralization potency of PGTs 127 and 128 to their binding affinity for cell-surface–expressed HIV-1JR-FL Env trimers (28, 29). The neutralization median inhibitory concentration (IC50) values of PGT 127 and 128 immunoglobulin Gs (IgGs) against HIV-1JR-FL were ~17- and 31- fold lower (i.e., more potent neutralization) than their cell-surface trimer-binding median effective concentration (EC50) values (Fig. 5), whereas the neutralization potency of Fabs PGT 127 and 128 correlated strongly with their binding affinity for cell-surface HIV-1JR-FL Env trimers (Fig. 5). Furthermore, although PGT 127 and 128 IgGs bound with similar apparent affinity to cell-surface Env trimers as their Fab counterparts, they neutralized approximately 81- and 70-fold more potently, respectively, than their corresponding Fab fragments (Fig. 5). Similar results were also obtained with HIV-1YU2 (fig. S14). Collectively, these results suggest that PGT 127 and 128 IgGs may cross-link spikes on the surface of the virus, giving an increase in affinity through avidity effects, but not on the surface of Env-expressing cells. The comparable binding affinity of PGT 127 and 128 IgGs and Fabs for cell-surface Env is consistent with IgG cross-linking of trimers on the viral surface occurring between spikes rather than within a single spike. In addition, intraspike cross-linking by PGT 128 IgG appears unlikely based on the inter-Fab distances observed for PGT 128 Fab-trimer complexes by electron microscopy (EM) (Fig. 4). Considering the scarcity of native Env trimers on the viral surface (30), a possible explanation for this observation is that two or more viral spikes are clustered to form an infectious unit, as proposed previously (31). In this scenario, neutralization measures binding to infectious Env units but not to single spikes. This interpretation also requires few infectious units on transfected cells as compared to single spikes, and that the single spikes are not in close enough proximity for cross-linking to occur. Previous studies have reported that avidity generally plays a limited role in antibody neutralization of HIV, as suggested by the relatively modest increases in neutralization potencies of IgGs as compared to their Fab counterparts (32, 33). However, in contrast to other broadly neutralizing epitopes on HIV Env, the epitopes recognized by PGTs 127 and 128 appear to be highly accessible (Fig. 4), which may promote interspike cross-linking.

Fig. 5

Cell-surface binding and neutralization properties of PGT 127 and PGT 128 IgGs and Fabs. (A) (Left) Binding of PGT 127 and PGT 128 Fabs and IgGs to HIV-1JR-FL trimers expressed on the surface of transfected 293T cells as determined by flow cytometry. (Right) Neutralizing activity of PGT 127 and PGT 128 IgGs and Fabs against HIV-1JR-FL . 2G12 is included for comparison. Experiments were performed in duplicate, and data are representative of at least two independent experiments. MFI, mean fluorescence intensity. (B) (Top) Comparison of binding (EC50) and neutralization (IC50) for PGT 127 and PGT 128 Fabs and IgGs against HIV-1JR-FL . 2G12 is included for comparison. (Bottom) Bar graph representation of Fab (IC50):IgG (IC50) ratios for PGT 127, PGT 128, b12, PG16, PGT 121, 2F5, and 4E10. 2G12 is not included because its two Fabs form a domain-swapped dimer (4). Ratios were calculated as IC50 of the Fab/IC50 of IgG.

Previous studies in various virus systems, including murine leukemia virus, dengue virus (DENV), West Nile virus (WNV), poliovirus, and HIV, have shown that viral infectivity decays exponentially with time (3436). Furthermore, certain neutralizing antibodies have been shown to accelerate the decay of viral infectivity (37, 38). For example, recent studies have demonstrated that the half-life of WNV and DENV decreases in the presence of virus-specific antibodies (37). To determine whether PGTs 127 and 128 affect the rate of viral infectivity decay, we measured the half-life of HIV-1JR-FL in the presence and absence of PGT 127 and 128 IgGs and Fabs. At antibody concentrations corresponding to 90% neutralization, PGTs 127 and 128 IgGs reduced the half-life of HIV-1JR-FL by approximately 9.7- and 11.2-fold, respectively, whereas the corresponding Fab fragments and 2G12 IgG had little to no effect on viral infectivity decay (Fig. 6, A and B). No evidence for antibody-induced gp120 shedding was observed (fig. S15). Collectively, these data suggest that interspike cross-linking by PGT 127 and 128 IgGs may accelerate the inactivation of HIV-1JR-FL Env spikes, perhaps by inducing conformational changes that perturb trimer functionality (39), resulting in enhanced neutralization potency.

Fig. 6

Impact of PGT 127 and PGT 128 on viral infectivity decay. (A) Viral infectivity decay of HIV-1JR-FL was measured in the presence of PGT 127 and PGT 128 IgGs and Fabs. 2G12 is included for comparison. Data were fitted to a single-phase exponential decay to obtain half-life. Individual experiments were performed in triplicate, and error bars represent the standard error of two independent experiments. (B) The reduction in the half-life of HIV-1JR-FL (expressed as an x-fold decrease) in the presence of antibodies at concentrations providing 90% neutralization, as compared to the absence of antibody. Error bars represent the standard error of two independent experiments.

Studies of protein-carbohydrate interactions have established various principles of molecular recognition. For example, spike glycan-protein interactions are weak because of unfavorable entropy contributions associated with glycan binding, multivalency is crucial to enhance binding affinity. Here we provide an example of multivalency achieved through the combination of glycan and protein; the three subsites for N332, N301, and the C-terminal V3 stem are essentially independent, but combine to mediate high-affinity recognition of a glycan-based epitope on HIV Env. Considering the highly exposed nature of this epitope and the high conservation of its two glycan and V3 loop backbone components, coupled with recent studies demonstrating that broad and potent serum neutralizing activity is frequently mediated by antibodies that target N332A-sensitive epitopes (5, 4043), it appears that this antigenic region may serve as an attractive vaccine target if appropriate immunogens can be designed.

Supporting Online Material

Materials and Methods

Figs. S1 to S15

Tables S1 to S6

References (4478)

References and Notes

  1. It has been shown that a three–amino acid insertion in the heavy-chain CDR H2 of the influenza virus-specific mAb 2D1 rearranges the antibody-combining site and enhances neutralization potency (12), demonstrating that somatically introduced amino acid insertions may be a conserved molecular mechanism for increasing antibody potency against viral pathogens.
  2. The eODmV3 was expressed in GnTI–/– HEK 293S cells to mimic the oligomannose-type glycosylation of that domain within intact gp120.
  3. Three canonical strand-pairing hydrogen bonds are formed as well as a hydrogen bond between V3 Asp325 and the backbone amide of AspH100d (Fig. 2C). Ile323 also interacts with the CDR H1-H2 disulfide and with LeuH100 in CDR H3, and Arg327 is located in close proximity to AspH100d. TyrH100b makes aromatic interaction with the Gly324-Asp325 peptide bond. Also, similar to many other bnmAbs to HIV, the PGT 128 CDR H3 loop is relatively long (19 amino acids), although not the longest seen to date for human Abs [31 or 32 residues for the PGT 140 series (5)].
  4. The HIV-1 664G trimer is based on the clade A strain KNH1144 and incorporates stabilizing mutations A501C, T605C, and I559P. We have previously shown that trimers incorporating these stabilizing mutations are competent to undergo CD4-induced conformational changes akin to those observed in the native trimer (22).
  5. HIV-1JR-FL is the only HIV isolate that has been shown to express a high proportion of fully cleaved Env trimers on the surface of transfected cells and was, therefore, selected for binding studies.
  6. Acknowledgments: We thank Y. Hua, K. Le, and V. Thaney for technical assistance; D. Ekiert for help with initial cloning of PGTs 125 and 126; X. Dai and X. Zhu for help with x-ray data collection and analysis; C. Corbaci and C. Williams for help with figure preparation; B. Walker for helpful comments on the manuscript; and members of The Glycosciences Laboratory for their collaboration in the establishment of the neoglycolipid-based microarray system. This work was supported by the International AIDS Vaccine Initiative Neutralizing Antibody Center, NIH grant AI84817 (I.A.W.), National Institute of Allergy and Infectious Diseases grant AI33292 (D.R.B.), NIH/National Research Service Award fellowship AI74372 (R.P.), Canadian Institutes of Health Research fellowship FRN HFE-224662 (J.-P.J.), HIV Vaccine Research and Design grant AI082362 (W.C.O., J.P.M., and I.A.W.), UK Research Councils’ Basic Technology Initiative “Glycoarrays” grant GRS/79268, Engineering and Physical Sciences Research Council Translational Grant EP/G037604/1, National Cancer Institute (NCI) Alliance of Glycobiologists for Detection of Cancer and Cancer Risk grant U01 CA128416, and the Ragon Institute. The three-dimensional reconstructions were conducted at the National Resource for Automated Molecular Microscopy, which is supported by NIH through the National Center for Research Resources' P41 program (RR017573). Portions of this research were carried out at the Stanford Synchrotron Radiation Lightsource (SSRL), a Directorate of the SLAC National Accelerator Laboratory and an Office of Science User Facility operated for the U.S. Department of Energy (DOE) Office of Science by Stanford University. The SSRL Structural Molecular Biology Program is supported by the DOE Office of Biological and Environmental Research; NIH’s National Center for Research Resources, Biomedical Technology Program (P41RR001209); and the National Institute of General Medical Sciences (NIGMS). Use of the Advanced Photon Source was supported by the DOE, Basic Energy Sciences, Office of Science, under contract no. DE-AC02-06CH11357. GM/CA CAT has been funded in whole or in part with federal funds from NCI (grant Y1-CO-1020) and NIGMS (grant Y1-GM-1104). Coordinates and structure factors for the Fab PGT 128/Man9, Fab PGT 127/Man9, and Fab PGT 128/eODmV3 structures have been deposited with the Protein Data Bank under accession codes 3TV3, 3TWC, and 3TYG. The Fab PGT 128/d664G trimer EM reconstruction density has been deposited with the Electron Microscopy Data Bank under accession code EMD-1970. Progenics Pharmaceuticals and the Cornell Research Foundation, on behalf of Cornell University, are the joint owners of U.S. Patent 7,939,083 and have additional patent applications on the stabilized Env trimer used in this manuscript. The IAVI and Theraclone hold U.S. patent 61/515,528 on the PGT antibodies. The Scripps Research Institute and IAVI have applied for a patent relating to the eODmV3 construct. Materials will be made available for noncommercial use under material transfer agreements with Progenics or Cornell (stabilized Env 664G trimer), IAVI (PGT antibodies), and TSRI (eODmV3). This is manuscript 21407-MB from The Scripps Research Institute.

Stay Connected to Science

Navigate This Article