CD4-Induced Activation in a Soluble HIV-1 Env Trimer

Transcription

1 Artile CD4-Indued Ativation in a Solule HIV-1 Env Trimer Miklos Guttman, 1 Natalie K. Garia, 1 Alert Cupo, 2 Tsutomu Matsui, 3 Jean-Philippe Julien, 4 Rogier W. Sanders, 2,5 Ian A. Wilson, 4 John P. Moore, 2 and Kelly K. Lee 1, * 1 Department of Mediinal Chemistry, University of Washington, Seattle, WA 98195, USA 2 Weill Medial College of Cornell University, New York, NY 10021, USA 3 Stanford Synhrotron Radiation Lightsoure, SLAC National Aelerator Laoratory, Menlo Park, CA 94025, USA 4 Department of Integrative Strutural and Computational Biology, International AIDS Vaine Initiative Neutralizing Antiody Center, Center for HIV/AIDS Vaine Immunology and Immunogen Disovery, and Skaggs Institute for Chemial Biology, The Sripps Researh Institute, La Jolla, CA 92037, USA 5 Department of Medial Miroiology, Aademi Medial Center, 1105 AZ Amsterdam, the Netherlands *Correspondene: kklee@uw.edu SUMMARY The HIV envelope glyoprotein (Env) trimer undergoes reeptor-indued onformational hanges that drive fusion of the viral and ellular memranes. Env onformational hanges have een oserved using low-resolution eletron mirosopy, ut only large-sale rearrangements have een visile. Here, we use hydrogen-deuterium exhange and oxidative laeling to gain a more preise understanding of the unliganded and CD4-ound forms of solule Env trimers (SOSIP.664), inluding their glyan omposition. CD4 ativation indues the reorganization of ridging sheet elements, V1/V2 and V3, muh of the gp120 inner domain, and the gp41 fusion suunit. Two CD4 inding site-targeted inhiitors have sustantially different effets: NBD-556 partially mimis CD4-indued destailization of the V1/V2 and V3 rown, whereas BMS-806 only affets regions around the gp120/gp41 interfae. The strutural information presented here inreases our knowledge of CD4- and small moleule-indued onformational hanges in Env and the allosteri pathways that lead to memrane fusion. INTRODUCTION The trimeri envelope glyoprotein (Env) omplex on the surfae of virions mediates HIV-1 entry and is the sole target for neutralizing antiodies (NAs) that are indued during natural infetion. A detailed understanding of Env s struture, its onformational rearrangements, and how it presents NA epitopes is ritial for guiding the design of oth protein-ased vaines and Env-targeting entry inhiitors (Jardine et al., 2013; Walker and Burton, 2010). The 4.7 Å rystal and 5.8 Å ryo-eletron mirosopy (ryo-em) strutures of a solule, leaved form of the Env trimer (BG505 SOSIP.664) have unveiled novel aspets of its antigeniity and arhiteture (Julien et al., 2013a; Lyumkis et al., 2013). However, the omplete organization of gp41 and the mehanism underlying CD4-indued strutural reorganizations within the trimer remain unresolved. It also is neessary to understand how flexile elements, inluding gp120 glyans and surfae loops, and the reathing of the overall Env omplex modulate its reognition y antiodies and host fators (Davenport et al., 2013; Kwong et al., 2002; Myszka et al., 2000; Sanlan et al., 2007). Charaterizing onformational hanges in multisuunit glyoprotein omplexes via lassial strutural methods an e very hallenging. Here, we proe the strutural dynamis of the Env trimer using hydrogen-deuterium exhange (HDX) to measure the rates of deuterium inorporation into akone amides under solution onditions. Unstrutured or flexile regions undergo deuterium exhange rapidly, in marked ontrast to ones involved in stale hydrogen onding networks as part of the protein seondary struture or those that are highly oluded from solvent. Coupled with mass spetrometry (MS), HDX information an e used to measure onformation dynamis with sequene-speifi detail, loalize protein-protein interations, and detet ligand-indued onformational hanges (Marsisin and Engen, 2010). As a omplementary approah, we performed oxidative laeling using short exposures to synhrotron radiation to generate hydroxyl radials that modify protein side hains. Beause solvent-oluded side hains reat less readily than exposed ones, this method allows solvent aessiility to e estimated, in some ases with single amino aid resolution (Tong et al., 2008; Wang and Chane, 2011). We have used these tehniques to study solule, leaved SOSIP.664 trimers ased on the sutype A sequenes KNH1144 and BG505. The sequene modifiations used to stailize these trimers are desried elsewhere and summarized in Figure 1A (Binley et al., 2000, 2002; Sanders et al., 2002, 2013). The resulting homogeneous trimers losely resemle native Env on virions, oth antigenially and struturally (Harris et al., 2011; Khayat et al., 2013; Sanders et al., 2013). Here we report a strutural analysis of the BG505 SOSIP.664 trimer in its unliganded prefusion and CD4-ound states. The HDX results reveal the nature of rearrangements in the gp120 reeptor inding suunit and the allosteri hanges in the gp41 fusion suunit. The sequene-speifi details omplement eletron mirosopy strutures of prefusion and CD4-ound trimers (Harris et al., 2011; Liu et al., 2008; Tran et al., 2012). 974 Struture 22, , July 8, 2014 ª2014 Elsevier Ltd All rights reserved

2 Figure 1. HDX Profile of SOSIP.664 Trimers (A) The sequene of mature, full-length Env is shown with the SOSIP modifiations indiated. To the right, strutural elements inluding variale loops 1 to 5, N/C termini, and heptad repeats of gp41 (HR1 and HR2) are mapped onto the rion diagram for one protomer from the BG505 SOSIP.664 trimer rystal struture (Protein Data Bank [PDB] aession numer 4NCO) (Julien et al., 2013a). (B and C) The HDX profiles are shown for unliganded BG505 and KNH1144 SOSIP.664 trimers. Perentage exhange is shown after 3 s, 1 min, 30 min, and 20 hr for all oservale pepti fragments at the midpoint of their primary sequene. For example, the exhange profile of the peptide from residue 105 to 111 is plotted at position 108. Individual exhange plots with errors are shown in Figures S2 and S3. We also use HDX to ompare the ations of two entry inhiitors that target the CD4 inding site: BMS-806 and NBD-556. The inding of NBD-556 to a monomeri gp120 ore fragment has een studied via thermodynami measurements and X-ray rystallography (Kwon et al., 2012; Shön et al., 2006; Zhao et al., 2005). NBD-556 inding has a major effet at the apex of the SOSIP.664 trimer, whih is sustantially different than what was seen using monomeri gp120. In ontrast, BMS-806 does not affet the apex of the trimer, onsistent with its inhiition of CD4-indued onformational hanges (Guo et al., 2003; Lin et al., 2003; Madani et al., 2004). Overall, our studies of how the HIV-1 Env trimer interats with CD4, NBD-556, and BMS-806 provide insights into the allosteri networks that are involved in the reeptor-mediated ativation of this type 1 fusion protein. RESULTS HDX-MS Profile of SOSIP.664 Trimers The KNH1144 and BG505 SOSIP.664 trimers were analyzed y HDX-MS. After optimizing pepsin digestion onditions, sequene overage of 92% (89% of gp120 and 100% of gp41) from 103 unique KNH1144 peptides and 82% (76% of gp120 and 100% of gp41) from 120 unique BG505 peptides was ahieved y MS (overage maps are shown in Figures S2 and S3 availale online). Expression of oth trimers in N-aetylgluosaminyl transferase I defiient (GnTI / ) 293S ells aided the detetion of glyosylated peptides during MS analysis, as the presene of only high-mannose type glyoforms redued glyan heterogeneity (Tale 1). Pepti fragments from V4 or the N-terminal half of V1 ould not e monitored y HDX-MS. Their asene was presumaly due to dense glyosylation, as V1 and V4 fragments were oservale after the pepti fragments were deglyosylated with endoglyosidase H (EndoH) or peptide-n-glyosidase F (PNGaseF). The exhange profiles for the KNH1144 and BG505 SOSIP.664 trimers are shown in Figure 1; the exhange data for eah peptide are plotted along the primary protein sequene (all individual exhange plots are shown in Figures S2 and S3). The patterns of protetion were very similar for oth trimers. Regions of high protetion were seen throughout the gp120 suunits, as in previous studies of gp120 monomers (Davenport et al., 2013; Guttman et al., 2012) and gp140 monomers (Guttman and Lee, 2013). However, we now also oserved extensive protetion within the variale loops of the gp120 suunits in the trimers. For example, a highly proteted region spanned the N-terminal half of V2 (residues ). Although the region was not overed in the BG505 data set, there was strong protetion etween V1 and V2 up to residue 159 in the KNH1144 data set. Hene, the region around N160, a entral feature of the roadly neutralizing antiody (NA) PG9/16 epitope (MLellan et al., 2011; Walker et al., 2009), has a stale seondary struture in the unliganded trimer. In ontrast, the rest of V2 (residues ) undergoes rapid exhange, implying that this relatively short segment may e the only true flexile loop portion of the V2 region. The peptides spanning V3 were also signifiantly proteted, indiating that in the unliganded trimer, the V3 loop either has a onstrained seondary struture or is solvent inaessile. A omplex profile of protetion throughout the gp41 suunit provides detailed information aout its organization (Figure 1; Figures S2 and S3). The seond half of HR1 (residues ) is sustantially proteted, partiularly so for residues 579 to 592. In ontrast, the N-terminal half of HR1 (residues ), whih ontains the I559P sustitution integral to SOSIP.664 trimers, was not proteted, indiating that this region is not involved in a stale seondary struture. A moderate level of exhange protetion was seen within the fusion peptide proximal region (FPPR; residues ), whereas the fusion peptide (FP) itself (as proed via residues ) was fully exhanged within seonds. The sequene following HR1 (residues ), inluding the disulfide loop and the immunodominant epitope luster I, was also sustantially proteted. Immediately C-terminal to the disulfide loop, there was only moderate protetion for residues 603 to 622 spanning the N611 and N616/ 618 (KNH1144/BG505) glyosylation sites. The rapid exhange oserved in a smaller, overlapping BG505 peptide indiates that residues 615 to 622 do not have a stale seondary struture. In HR2, segments 623 to 634 and 641 to 646 are very strongly proteted in the prefusion form of the trimer, while the intervening segments 635 to 640 (inluding glyosylation site N637) are muh less well proteted. The seond half of HR2 Struture 22, , July 8, 2014 ª2014 Elsevier Ltd All rights reserved 975

3 Tale 1. Glyoforms Oserved in SOSIP.664 Trimers Glyosylation Sites a KNH1144 BG505 (293S) BG505 (293F) N88 Man 5 Man 5 Man 5 and omplex N156 Man 8,9 N188/190 and N190 Man 5 Man 5, omplex N197 Man 5 9 Man 5 9 Man 5 9 N234 N241 Man 6 9 Man 6 9 Man 8,9 Man 8,9 N262 Man 8,9 Man 8,9 Man 8,9 N276 Man 7,8 Man 5 8 Man 5-8 N295 Man 8 Man 8 N301 Man 7,8 Man 8 Man 8 N332 and N339 N356 and N363 Man 8,9 Man 7 N448 Man 8,9 Man 8,9 Man 8,9 N463/462 Man 5 Man 5 omplex N611 and N616/618, Man 5 Man 5 omplex N625 Man 5 Man 5 Man 5 N637 Man 5,6 Man 5 8 Man 5 8 a N-linked glyosylation site(s) where some peptides ontained more than one glyan; KNH1144/BG505 position. Data were otained from peptides earing two glyans, so only the average for the total omined glyosylation is reported. Glyan oupany was oserved at less than 90%, as assessed y the relative aundane of nonglyosylated and deglyosylated signal after PNGaseF digestion (see Tales S1 and S2). (residues ) undergoes rapid exhange and appears to e highly dynami in solution. Almost all of the SOSIP.664 peptides exhiited unimodal exhange profiles signifying onformational homogeneity within the samples, without any evidene of slow, large-sale, orrelated protein motions (EX1 exhange kinetis). The only exeption was fragment 538 to 547, whih exhiited imodal ehavior at early time points in oth the KNH1144 and BG505 data sets (see Figure S4). Effet of Glyosylation on Underlying Protein Strutural Dynamis We addressed whether differenes in trimer glyosylation affet the ehavior of the underlying protein struture. To do so, we ompared HDX profiles for BG505 SOSIP.664 trimers expressed in 293F ells (apale of making omplex glyans) and 293S GnTI / ells (only high-mannose glyans present). The differenes in glyosylation etween expression systems were apparent y SDS-PAGE and lue-native PAGE (BN-PAGE) (Figure S1). The pepti digest showed that the majority of oservale glyopeptides from 293F-trimers were still in the high-mannose form, and only glyans at sites N88 and N462 had een proessed to omplex type (Tale 1). Beause omplex glyoforms are generally harder to detet eause of their lower ionization effiieny and miroheterogeneity (Stavenhagen et al., 2013), we also examined the relative intensities of mathed glyopeptides derived under idential onditions from BG505 SOSIP.664 trimers that had een produed in the two different ell types for a more diret assessment of glyosylation differenes (Figure S5). The signals for the high-mannose glyopeptides at N276, N197, Figure 2. HDX-MS Comparisons of 293F versus 293S GnTI / Expression System (A) The utterfly plot shows the exhange profiles of BG505 SOSIP.664 trimers from 293S GnTI / ells (all high-mannose glyoforms, top) versus 293F ells (ontains omplex glyosylation, ottom). (B) The differenes at eah time point are plotted in the differene plot underneath, revealing no major hanges attriuted to differenes in glyosylation types. N262, N234, N448, N625, and N637 were relatively onsistent etween 293F and 293S data sets, suggesting that high mannose glyans are indeed present at all these sites, irrespetive of the ell type used to make the trimers. In ontrast, the signal loss for the high-mannose glyoforms at N88, N190, N462, and N611/N618 in the 293F data set indiates that the glyans at these sites had proaly een proessed to omplex types. Despite the glyosylation differenes, the HDX profiles for the 293F and 293S material were very similar (Figure 2), suggesting that the onformational dynamis of the underlying protein struture is largely insensitive to any differenes in glyan proessing. CD4-Indued Changes within the Trimer The inding of solule, two-domain CD4 leads to dramati onformational hanges that are virtually idential for BG505 and KNH1144 SOSIP.664 trimers, as monitored y HDX-MS (Figure 3). The most striking hanges ourred within V1/V2 and V3, whih lost all protetion upon CD4 inding. As expeted, the CD4 inding region itself was more proteted in the ound state. Major hanges also ourred within regions that omprise the ridging sheet; strand 21 eame more proteted, whereas the other three strands (2, 3, and 20) and the segment diretly C-terminal to 21 were less proteted. There was also an inrease in protetion within all three layers of the gp120 inner domain. Within gp41, the FPPR eame less proteted, whereas the onverse was seen for the C-terminal half of HR1. The slight protetion of the seond half of HR2 in the unliganded BG505 trimer was lost upon CD4 inding. A minor fration the trimer- CD4 omplex showed dimerization when analyzed on BN- PAGE gels for the BG505, ut not the KNH1144 onstrut (Figure S1). However, the HDX data yielded no indiations of the presene of multiple Env speies, as assessed y the lak of imodal distriutions in the mass envelopes (Guttman et al., 976 Struture 22, , July 8, 2014 ª2014 Elsevier Ltd All rights reserved

4 2013; Weis et al., 2006). Thus, if some degree of oligomerization did our in solution, it did not affet the HDX profile of the BG505 trimer-cd4 omplex. Oxidative laeling with MS was used as a omplementary approah to trak hanges in the side-hain solvent aessiility of BG505 SOSIP.664 trimers upon CD4 inding. Although oxidative laeling was highly reproduile (Figure 4A), the sample-tosample variaility in digestion effiieny signifiantly affeted the signal intensities of individual peptides, suh that only residues highly suseptile to oxidation ould e monitored relialy (see Experimental Proedures). We speifially examined residues M95, M104, M150, M161, F316, M475, and M626, as well as leuine residues L660, L661, and L663 near the C terminus of gp41 in this onstrut (Figure 3C). Whereas M104, M95, and the triad of gp41 leuines exhiited little response to CD4 inding, signifiant hanges were oserved for M150 and M161 in V1/V2 and F316 in V3. These residues were all oxidized muh more rapidly after CD4 ound, indiative of a marked inrease in their solvent aessiility (Figure 4B). There was also a slight inrease in the oxidation rate for M626 in gp41 HR2. In ontrast, the oxidation rate at M475 in layer 3 of the gp120 inner domain was dereased, suggesting that this residue eomes more uried following CD4 inding. Figure 3. Changes upon CD4 Binding y HDX (A and B) Butterfly plots omparing the exhange profiles for SOSIP.664 trimers unliganded (top) and in omplex with scd4 (ottom) for BG505 (A) and KNH1144 (B). The orresponding differene plots elow highlight the regions that gain and lose protetion upon scd4 inding, plotted aove and elow the axis, respetively. (C) Regions that are more proteted (lue) or less proteted (red) upon CD4 inding are mapped on the BG505 SOSIP.664 trimer rystal struture (PDB Effets of Env Inhiitors BMS-806 and NBD-556 To further investigate ligand-indued allosteri hanges in the struture of SOSIP.664 trimers, we used HDX to examine the effets of two small-moleule ompounds that target the CD4 inding site and inhiit a suset of HIV-1 isolates (Zhao et al., 2005). NBD-556 inds to the Phe43 avity in gp120 (Kwon et al., 2012) and indues onformational hanges similar to CD4, at least in the ontext of monomeri gp120 (Shön et al., 2006). BMS-806 is also thought to ind at the CD4 inding site ut has een shown to lok these same onformational hanges (Guo et al., 2003; Madani et al., 2004). Reported half maximal inhiitory onentration values for NBD-556 are generally in the miromolar range (Shön et al., 2006; Zhao et al., 2005), whereas for BMS-806 they are in the nanomolar range (Madani et al., 2004; Si et al., 2004). Thus, HDX experiments with the KNH1144 SOSIP.664 trimers were performed in the presene of 100 mm of eah ompound to ensure saturating oupany, and all uffers ontained 2% DMSO to maintain the ompounds in solution. Under these onditions, the inding of antiody 17, whih reognizes a CD4-indued epitope overlapping the oreeptor inding site (Thali et al., 1993), to the trimers was enhaned y NBD-556 ut hindered y BMS-806 (Figure S7A). An additional unliganded HDX data set derived using the same uffer showed that the inlusion of 2% DMSO had minimal effets (Figure S7B). The onformational hanges indued in the KNH1144 SOSIP.664 trimer y NBD-556 inding were in part similar to those seen upon CD4 inding (Figure 5A; f. Figure 3A). For example, destailization ourred within V1/V2 and V3, aession numer 4NCO) (Julien et al., 2013a). KNH1144 sequene data were used for these heatmaps eause of the greater overage, although BG505 shows nearly idential trends for the regions overed. Individual exhange plots with errors are shown in Figures S2 and S3. Laeled spheres highlight the speifi residues monitored y oxidative laeling. Struture 22, , July 8, 2014 ª2014 Elsevier Ltd All rights reserved 977

5 throughout the ridging sheet, and even at the FPPR within gp41. However, in ontrast to CD4, NBD-556 inding did not lead to oservale hanges in the gp120 inner domain or in HR1 of gp41 (Figure 5B). In ontrast, the inding of BMS-806 altered the trimer in a way that is sustantially different to the effet of CD4, or of NBD-556 (Figures 5C and 5D). When BMS- 806 ound, layers 2 and 3 of the gp120 inner domain eame more proteted, with only modest effets on gp41 FPPR and HR1 and, notaly, no major hanges in V1/V2 or V3. DISCUSSION Figure 4. Solvent Aessiility Changes upon CD4 Binding Monitored y Oxidative Laeling (A) Modifiation rates of peptides used as internal dosimeters for ensuring that oth the unliganded and CD4-ound data sets were exposed to similar amounts of oxidative laeling. The signal for the unmodified form (normalized to a nonirradiated sample) of leuine enkephalin (left) and Sustane P (right) are shown as a funtion of exposure to synhrotron radiation. (B) The degree of oxidation is shown as a funtion of radiation exposure for various regions of SOSIP.664 (BG505) in the unliganded (gray irles) or CD4- ound (squares) state. Perentage modified refers to the signal intensity of Strutural Organization of Env Trimers in Solution Cleaved, stailized SOSIP gp140s are presently the only form of reominant solule trimers with a native-like struture, as exemplified y the KNH1144 (Bartesaghi et al., 2013; Harris et al., 2011) and BG505 SOSIP trimers (Julien et al., 2013a, 2013; Khayat et al., 2013; Lyumkis et al., 2013; Sanders et al., 2013). The data presented here provide additional evidene that these SOSIP trimers have onformations onsistent with the funtional, prefusion Env struture. In ontrast, the alternative way to make solule gp140s, y knoking out the intersuunit leavage site, yields nonnative proteins with heterogeneous, splayed-out onfigurations of the gp120 heads held together y the gp41 stem, whih has proaly adopted a postfusion (i.e., six-helix undle) state (Guttman and Lee, 2013; Ringe et al., 2013). The present analysis gives insight into the underlying strutural dynamis in the unliganded trimer while providing sequenespeifi information aout elements that ould not e resolved in the reent 5 Åstrutures. Both the ryo-em trimer struture in omplex with antiody PGV04 and the rystal struture in omplex with antiody PGT122 were otained using the same BG505 SOSIP.664 onstrut studied here (Julien et al., 2013a; Lyumkis et al., 2013). These strutures revealed how the V1/V2 and V3 variale regions form intimate interations at the trimer rown. We find that several regions of V1/V2 do indeed have a high degree of HDX protetion, whih is onsistent with the extensive sheet seondary struture present in the Greek key motif seen in the PG9-ound saffolded V1/V2 struture (MLellan et al., 2011) and in the SOSIP.664 trimer strutures (Julien et al., 2013a; Lyumkis et al., 2013). Similarly, the V3 region of the trimer is extensively proteted efore, ut not after, CD4 inding (Figures 3A and 3B). No suh protetion of the V1/V2 or V3 regions was seen in our previous studies of monomeri gp120 (Davenport et al., 2013; Guttman et al., 2012) or monomeri, unleaved gp140 (Guttman and Lee, 2013) implying that these regions of Env are well ordered only in prefusion trimers. The slow oxidative laeling kinetis seen for residues M161 and F316 reinfore the assessment that V1/V2 and V3 segments are largely inaessile to the solvent in the prefusion form of the the oxidized from (+16 Da) relative to the sum of signals for the modified and unmodified peptide, with the sites of oxidation shown in oldfae type. Error ars represent the SD etween dupliate measurements. Examples of mass spetra revealing hanges in oxidation rates are shown in Figure S6. The oxidation sites are illustrated on the trimer in Figure 3C. *Data for peptide 656 to 664 enompass oxidation at residues L660, L661, and L Struture 22, , July 8, 2014 ª2014 Elsevier Ltd All rights reserved

6 Figure 5. Changes upon NBD-556 and BMS-806 Binding y HDX (A and B) Butterfly and differene plots showing the effets of NBD-556 (A) and BMS-806 (B) on KNH1144 SOSIP.664 trimers. (C and D) Differenes upon inding NBD-556 (C) and BMS-806 (D) are plotted on the trimer rystal struture (PDB aession numer 4NCO), revealing regions that are more proteted (lue) or less proteted (red) upon CD4 inding. All of the major hanges in the highlighted regions are signifiant as assessed y the experimental error (individual plots shown in Figure 6). BG505 SOSIP.664 trimer (Figure 4B). Thus, although V3 is well exposed on gp120 monomers, our results reinfore its sequestration within the trimer struture (Julien et al., 2013a; Lyumkis et al., 2013). Although the availale strutures reveal the overall organization of the prefusion form of gp41, for example highlighting a entral helial undle, muh of this suunit has not yet een interpreted in atomi detail (Julien et al., 2013a; Lyumkis et al., 2013). The HDX-MS data provide additional sequene-speifi information for gp41 strutural order. The N-terminal half of HR1 (residues ), whih ontains the I559P gp41-stailizing modifiation, is in rapid exhange and, hene, is not involved in any stale seondary struture. In ontrast, the rest of HR1 (TVWGIKQLQARVLAVERYLRDQQL; residues ) is extensively proteted (Figures 1B and 1C). We infer that the long entral helies oserved in the trimer strutures (Bartesaghi et al., 2013; Julien et al., 2013a; Lyumkis et al., 2013) orrespond to residues 569 to 592, rather than the entirety of HR1 (residues ) that is oserved as a helix in the postfusion gp41 struture (Chan et al., 1997). In oth the rystal and EM strutures of the BG505 SOSIP.664 trimer, HR2 is proposed to form a long helix. We now find, via HDX, that the C-terminal helial portion of HR2 (residues ) is only modestly proteted. This region thus appears to e more dynami in solution than is implied y the stati strutures in the rystal and in vitreous ie. In ontrast, signifiant segments in the N-terminal portion of HR2 and the preeding gp41 loop region are highly proteted and, hene, likely to e engaged in a stale seondary struture. These elements of the trimer are situated lose to the gp120/gp41 interfae as they are linked, via the engineered SOS disulfide ond (C501 C605), to the C-terminal region of gp120 (Binley et al., 2000). Of note is that the N- and C-terminal extensions of gp120 are also highly proteted in the trimer ut not in gp120 (Davenport et al., 2013; Guttman et al., 2012), again onsistent with their involvement in gp41 assoiation (Helseth et al., 1991). The N-terminal portion of the FP was proteted negligily, whih is surprising eause this hydrophoi region might e antiipated to e solvent oluded in the prefusion state. In a similar study with influenza hemagglutinin, also a type I fusion protein, the FP was indeed moderately proteted in a way attriutale to either solvent olusion or hydrogen onding with residues lining the FP poket (N.K.G., M.G., J.L. Ener, and K.K.L., unpulished data; Wilson et al., 1981). In Env trimers, the FP may e only loosely uried, without any partiipation of its akone in hydrogen onds. It is also possile that the SOSIP modifiations partially disrupt the paking of the FP. Alternatively, the memrane-proximal external region, whih is asent from the Struture 22, , July 8, 2014 ª2014 Elsevier Ltd All rights reserved 979

7 Figure 6. Key CD4-Indued Changes in Trimeri SOSIP.664 Monitored y HDX-MS (A) The SOSIP.664 trimer rystal struture (PDB aession numer 4NCO) (Julien et al., 2013a) was modeled into the eletron mirosopi density of unliganded Env on virions (Eletron Mirosopy Dataase [EMD] 5019) (Liu et al., 2008): V1/V2 (yellow) and V3 (green), the ridging sheet (red), HR1 (purple), and gp120 inner domain (lak). (B) The scd4-ound gp120 ore struture (PDB aession numer 3JWD) (Panera et al., 2010) was modeled into the scd4-ound eletron density map (EMD 5455) (Tran et al., 2012) with CD4 shown in lue. In the CD4-ound state, the ridging sheet is reorganized, positioning V1/V2 away from the trimeri interfae and exposing the V3 loop. Individual exhange profiles of the key regions of interest are shown for unliganded trimer (lue), scd4 ound (red), NBD-556 ound (orange) and BMS-806 ound (yan). Error ars represent SDs from dupliate measurements. Eletron mirosopi renderings were made with Chimera (Pettersen et al., 2004). SOSIP.664 onstrut, may play a role in oluding the FP in fulllength, funtional trimers. SOSIP.664 trimers have a muh higher ontent of high mannose glyans than was found in a previous study of monomeri gp120 (Leonard et al., 1990). This finding is onsistent with previous reports that quaternary interations within Env trimers olude several glyan hains from glyosidases and therefore maintain their high-mannose form (Doores et al., 2010). Positions N276, N197, and N301, whih ear omplex glyans in the ontext of gp120 monomers and unleaved gp140 onstruts (Go et al., 2008), show little mannose trimming and are present as predominantly Man 9 and Man 8 glyoforms in SOSIP.664 trimers, even from 293F ells. Beause none of these glyans are part of the 2G12 epitope, the use of a 2G12 affinity olumn to purify the trimers is unlikely to skew the oserved glyosylation profiles (Sanders et al., 2002a). Colletively, these oservations may e relevant to immunogen design, eause several of these glyans are involved in NA epitopes (Jardine et al., 2013; Panera et al., 2013; Pejhal et al., 2011; Walker et al., 2011). CD4-Indued Transitions in Env Trimers These data provide a detailed glimpse into the CD4-indued reorganization of the trimer from the losed to the open form, a proess that has een imaged at low-resolution y ryo-em (Harris et al., 2011; Liu et al., 2008; Tran et al., 2012). In the losed, prefusion form, the V1/V2 region is involved in interand intraprotomer interations that form a ap at the trimer apex (Figure 6A). Likewise, the V3 and the ridging sheet segments, whih form the oreeptor inding site, are also onstrained y their involvement in the quaternary struture within the prefusion trimer. Upon CD4 inding, the struture at the rown is disrupted, and the V1/V2 and V3 regions eome disordered (Figure 3C). The oxidation rates for V2 residues M150 and M161 and V3 residue F316 are also drastially inreased when CD4 inds (Figure 4B). This inrease in the loal solvent 980 Struture 22, , July 8, 2014 ª2014 Elsevier Ltd All rights reserved

8 aessiility is again onsistent with an unraveling of the V1/V2 and V3 interations at the trimer apex as the entire sudomain struture opens up. Overall, our data illustrate how CD4 inding ats to disrupt key quaternary interations and therey expose the elements neessary for oreeptor inding and the susequent onformational hanges that drive fusion. Moreover, these same onformational hanges unover various NA epitopes that are effiiently shielded on the prefusion form of the trimer, suh as those assoiated with the oreeptor inding site, inluding V3 elements. The effets of CD4 inding extend eyond the immediate ontat zone to the gp120 inner domain as well as to previously unresolved elements in gp41. The inreased protetion of the inner domain, along with a derease in solvent aessiility (as proed y the oxidation rate of M475), is not surprising given that these regions reside at the interfae with CD4. More unexpeted was the large allosteri effet oserved in gp41 FPPR and HR1. Upon CD4 inding, the FPPR eomes less proteted, alleviating onstraints around the FP, whereas the enter of HR1 eomes more proteted, possily a refletion of enhaned paking of the entral helial ore. For the BG505 trimers, a slight derease in protetion at the C-terminal half of HR2 was also evident (Figure 3B). These alterations to gp41 might serve as an initial priming event prior to the full ativation of Env y oreeptor inding. Figure 7. Allosteri Networks in Env Trimers CD4 inding leads to two allosteri effets within Env trimers (red to yellow pathways). Pathways are highlighted on a single protomer of the trimer with neighoring protomers rendered transparent. (A) Repositioning of the elements of the ridging sheet leads to the disruption of the trimeri interations within V1/V2 and V3, while somehow affeting the FPPR, whose loation and relation to the rown region have yet to e interpreted in high-resolution strutures. (B) The onformational hanges indued within three layers of the inner domain also influene HR1 within gp41. CD4 Binding Site-Targeted Ligands Reveal Distint Allosteri Networks in Env A omparison of the effets imparted on Env trimers y the ligands examined here suggests that two allosteri networks are engaged when CD4 ativates the trimer: one involves opening of the Env rown, and the other leads to priming of gp41. The network attriutale to opening involves rearrangements of the ridging sheet elements and the disruption of interprotomer interations mediated y V1/V2 and V3 and ulminates in the formation and exposure of the oreeptor inding site (Figure 7A). The priming network links the CD4 inding site to HR1 through layers 1 to 3 of the gp120 inner domain (Figure 7B). Whereas CD4 inding triggers oth networks, NBD-556 leads only to the opening of Env and does not appear to prime gp41 (Figures 5 and 6). This mode of ation was not evident from earlier strutural studies that used a trunated gp120 ore, whih adopts a reeptor-ound onformation with an ordered inner domain that is nearly idential in struture to the CD4-ound ore (Kwon et al., 2012). Conversely, BMS-806 appears to influene the priming network, without affeting the opening network. On the asis of our interpretation that residues 569 to 592 orrespond to the entral gp41 helix, the BMS-806-indued hanges in HR1 appear to e driven through the a1 helix in gp120 layer 2, whih like layer 3 also eomes sustantially more ordered (Figures 5D and 7B). The gp120 inner domain hanges that we impliate as part of the priming pathway are onsistent with mutagenesis studies that proe the transdution of CD4-indued onformational hanges through the inner domain to gp41 (Désormeaux et al., 2013; Finzi et al., 2010). It is surprising perhaps that the FPPR in gp41 responds to oth CD4 and NBD-556, whereas the rest of gp41 appears unhanged y NBD-556 inding. The position of the FPPR was not interpreted in the existing SOSIP.664 trimer strutures, so its linkage to the opening network is not yet lear. Overall, our present HDX-MS analysis reveals how CD4 and CD4 mimetis indue profound onformational hanges in the trimer y ativating distint networks that open the trimer rown, unmask key elements of the oreeptor inding site, and prime the fusogeni properties of the gp41 suunit. Further studies will e required to address how the remaining unmapped elements of the trimer, and the oreeptor inding event, are orhestrated to drive the memrane fusion stages of virus entry. EXPERIMENTAL PROCEDURES Sample Preparation KNH1144 and BG505 SOSIP.664 were expressed and purified y 2G12 affinity hromatography, as desried previously (Iyer et al., 2007). A final round of size exlusion hromatography over a Superdex 200 olumn (GE Healthare) in PBS (20 mm sodium phosphate [ph 7.4], 150 mm NaCl, 1 mm EDTA, 0.02% sodium azide) was performed prior to HDX to remove Env aggregates, dimers, and monomers. Proteins were onentrated with VivaSpin spin filters (GE Healthare). Sample purity was assessed with reduing and nonreduing SDS-PAGE and BN-PAGE (Figure S1). The omplex with solule CD4 Struture 22, , July 8, 2014 ª2014 Elsevier Ltd All rights reserved 981

9 (scd4) was formed y overnight inuation with a 3-fold molar exess of twodomain human scd4 at 4 C. scd4 was otained from the NIH aquired immune defiieny syndrome (AIDS) reagents program (Garlik et al., 1990). BMS-806 was purhased from SellekChem. NBD-556 was synthesized and purified as desried previously (Shön et al., 2006). HDX-MS A 10 mg (80 pmol) aliquot of SOSIP.664 (per time point) was inuated in deuterated uffer (85% D 2 O) at 22 C for 3 s, 1 min, 30 min, and 20 hr. Samples were added to an equal volume (100 ml) of ie-old quenh solution (200 mm tris(2-aroxyethyl)phosphine, 0.02% formi aid) with 10 ml of a 3 mg/ml pepsin solution in 100 mm Na 3 PO 4 (ph 4.0), to a final ph of 2.5. After 5 min of pepsin digestion on ie, samples were flash frozen in liquid nitrogen and stored at 80 C until analysis. Deglyosylated pepti digests were prepared y treatment with either 0.5 mu of N-glyanase (ProZyme) (ph 7.0) or 0.1 mu EndoH (ProZyme) (ph 5.5) for 2 hr at 37 C. Compounds NBD-556 and BMS-806 were resuspended in DMSO (>99.9%; Sigma-Aldrih) and preinuated with SOSIP.664 at a final onentration of 100 mm in a uffer ontaining 2% DMSO. Idential ompound and DMSO onentrations were inluded in the deuterium inuations. Deuterated samples were analyzed y liquid hromatography (LC)-MS as previously desried (Guttman and Lee, 2013). Peptides were identified y exat mass and MS/MS spetra with the aid of protein prospetor (Baker et al., 2011). Identifiation of glyopeptides was aided y MS/MS spetra of the enzymatially deglyosylated pepsin digests. Deuterium shifts were alulated with using HX-Express version 2 (Guttman et al., 2013; Weis et al., 2006) to monitor the mass envelope widths for possile imodal ehavior. The perentage exhange for eah fragment was alulated relative to zero and fully deuterated standards, as desried previously (Guttman et al., 2012). Heatmaps were made using PyMOL (DeLano, 2002). Beause additional GlNA groups in omplex type glyan hains an ontriute to the oserved deuterium uptake kinetis, omparisons of the exhange kinetis of glyopeptides with different glyan ompositions may e misleading (Guttman et al., 2011). For this reason, the HDX omparisons etween 293S GnTI / and 293F expressed material were made only with data from the high mannose glyoforms within the 293F data set, whih were a minor speies for sites showing omplex type glyosylation. Although this partially limits the sope of the omparison, if altered glyosylation does have strutural or dynami effets eyond the glyosylation site, it is likely that differenes in the HDX profiles of peripheral regions will e seen. Oxidative Laeling X-ray irradiation experiments were performed at Beamline 4-2 of the Stanford Synhrotron Radiation Lightsoure (SSRL). The X-ray eam (11 kev with a ring urrent of ma) was attenuated to give dose responses in the appropriate range for oxidative laeling as assessed on site using samples ontaining 80 mm Alexafluor-488 dye in an idential phosphate uffer and rionulease A at 1 mg/ml, to mimi sample onditions (Xu and Chane, 2007). Samples of unliganded and CD4-ound trimer (with a 3-fold molar exess of scd4 relative to eah SOSIP.664 protomer) in PBS were all prepared at a total protein onentration of 1 mg/ml to avoid differenes in savenging due to protein onentration effets (Tong et al., 2008). Leuine enkephalin (330 ng) and Sustane P (580 ng; Sigma-Aldrih) were inluded in all samples as internal peptide dosimeters to ensure onsisteny of laeling etween sample sets (Tong et al., 2008). Samples (20 ml) were loaded with an autosampler using a Mirola 560 syringe pump (Hamilton) and irradiated as they passed through a quartz apillary. The flow rates during irradiation were adjusted to ahieve exposure times ranging from 15 to 240 mse. Immediately after exposure, the samples were dispensed and mixed into a tue ontaining 2.5 ml of 200 mm methionine to prevent seondary oxidation (Xu et al., 2005) and stored at 20 C. Samples with no irradiation were olleted under idential onditions to serve as the nonirradiated standards. All experiments were performed in dupliate. Irradiated samples were mixed 1:1 with a solution of 6 M guanidine HCl, 20 mm dithiothreitol (DTT) and 50 mm Tris (ph 8.0) and denatured at 85 C for 30 min. Cysteines were then alkylated y inuation in the dark for 1 hr with 20 mm iodoaetamide, followed y an addition of 10 mm DTT to quenh the reation. After dilution with 20 mm Tris (ph 8.0) to ahieve a final Gnd-HCl onentration of 0.5 M, samples were treated with 0.5 mu of N-glyanase (ProZyme) at 37 C for 90 min to ahieve full enzymati deglyosylation. LysC and GluC (Promega) were then added at ratios of 6:1 and 3:1 (sustrate to enzyme) and left at 37 C for 18 hr. Samples were analyzed y LC-MS on a Synapt Q-TOF mass spetrometer oupled to a Aquity UPLC system (Waters). Peptides were loaded onto a Hypersil mm 2.1 mm C18 olumn (Thermo Sientifi) and resolved with a gradient of 0% to 32% B over 15 min (A: 3% aetonitrile, 0.1% formi aid; B: 100% aetonitrile, 0.1% formi aid). The GluC and LysC digests were run in alternating order to minimize any effets of sample arryover. Chromatographi peaks for the most aundant and nonoverlapped isotopi peaks were integrated with MassLynx (Waters). Oxidation rates were initially alulated y monitoring the net signal intensity of the unmodified peptide as a funtion of irradiation dose. Although this type of analysis was useful for monitoring the oxidation rates of the internal dosimeter peptides, it was not suitale for the analysis of the Env-derived proteolyti fragments. The high variaility in digestion effiieny during sample proessing led to signifiant variaility in the resulting peptide signal intensities. Instead, the intensity of eah modified form of a peptide (+16 Da) was measured relative to the sum of the modified and unmodified signal intensities (Chen et al., 2012), a metri that should e independent of digestion effiieny. This approah provided preise, residue-speifi information ut limited the sope of the analysis eause only peptides with suffiient degrees of modifiation ould e analyzed. Peptides and oxidation sites were identified y manual inspetion of MS/MS spetra aided y Protein Prospetor (Baker et al., 2011). SUPPLEMENTAL INFORMATION Supplemental Information inludes seven figures and two tales and an e found with this artile online at AUTHOR CONTRIBUTIONS M.G., N.K.G., and T.M. onduted the experiments. A.C., J.-P.J. and R.W.S. provided essential reagents. M.G. and K.K.L. analyzed the data and made the figures. M.G., N.K.G., J.-P.J., J.P.M., R.W.S., I.A.W., and K.K.L. wrote the manusript. ACKNOWLEDGMENTS We thank Matthew MDonald for assistane with the synthesis of NBD-556 and Asim Denath, Per Johan Klasse, Max Crispin, and Andrew B. Ward for valuale disussions. Two-domain scd4-183 (Pharmaia) was otained through the AIDS Researh and Referene Reagent Program, Division of AIDS, National Institute of Allergy and Infetious Diseases, NIH. This work was supported y NIH grants F32-GM (M.G.), T32-GM (N.K.G.), R00-GM (K.K.L.), R01-GM (K.K.L.), P01-AI82362 (J.P.M., I.A.W.), R01-AI (I.A.W.), R37-AI36082 (J.P.M.), and R01- AI41420 (J.P.M.); grant OPP from the Bill and Melinda Gates Foundation Collaoration for AIDS Vaine Disovery (M.G., K.K.L.); grant P30-AI from the University of Alaama at Birmingham s Center for AIDS Researh through the Creative and Novel Ideas in HIV Researh program for new HIV investigators (K.K.L.); and the International AIDS Vaine Initiative Neutralizing Antiody Consortium and Center (I.A.W., J.P.M.), Center for HIV/ AIDS Vaine Immunology and Immunogen Disovery grant UM1 AI (I.A.W.), the Hope Barns Fellowship (N.K.G.), a Vidi grant from the Netherlands Organization for Sientifi Researh (R.W.S.), Starting Investigator Grant ERC- StG SHEV from the European Researh Counil (R.W.S.), and a Canadian Institutes of Health Researh fellowship (J.-P.J.). Portions of this researh were arried out at the SSRL. The SSRL Strutural Moleular Biology Program is supported y the Department of Energy s Offie of Biologial and Environmental Researh and y the NIH s National Institute of General Medial Sienes (grant P41-GM103393) and National Center for Researh Resoures (grant P41-RR001209). Reeived: Marh 11, 2014 Revised: April 28, 2014 Aepted: May 1, 2014 Pulished: June 12, Struture 22, , July 8, 2014 ª2014 Elsevier Ltd All rights reserved

10 REFERENCES Baker, P.R., Trinidad, J.C., and Chalkley, R.J. (2011). Modifiation site loalization soring integrated into a searh engine. Mol. Cell. Proteomis 10, M Bartesaghi, A., Merk, A., Borgnia, M.J., Milne, J.L., and Suramaniam, S. (2013). Prefusion struture of trimeri HIV-1 envelope glyoprotein determined y ryo-eletron mirosopy. Nat. Strut. Mol. Biol. 20, Binley, J.M., Sanders, R.W., Clas, B., Shuelke, N., Master, A., Guo, Y., Kajumo, F., Anselma, D.J., Maddon, P.J., Olson, W.C., and Moore, J.P. (2000). A reominant human immunodefiieny virus type 1 envelope glyoprotein omplex stailized y an intermoleular disulfide ond etween the gp120 and gp41 suunits is an antigeni mimi of the trimeri virion-assoiated struture. J. Virol. 74, Binley, J.M., Sanders, R.W., Master, A., Cayanan, C.S., Wiley, C.L., Shiffner, L., Travis, B., Kuhmann, S., Burton, D.R., Hu, S.L., et al. (2002). Enhaning the proteolyti maturation of human immunodefiieny virus type 1 envelope glyoproteins. J. Virol. 76, Chan, D.C., Fass, D., Berger, J.M., and Kim, P.S. (1997). Core struture of gp41 from the HIV envelope glyoprotein. Cell 89, Chen, J., Rempel, D.L., Gau, B.C., and Gross, M.L. (2012). Fast photohemial oxidation of proteins and mass spetrometry follow sumilliseond protein folding at the amino-aid level. J. Am. Chem. So. 134, Davenport, T.M., Guttman, M., Guo, W., Cleveland, B., Kahn, M., Hu, S.L., and Lee, K.K. (2013). Isolate-speifi differenes in the onformational dynamis and antigeniity of HIV-1 gp120. J. Virol. 87, DeLano, W.L. (2002). The PyMOL Moleular Graphis System. (San Carlos, CA: DeLano Sientifi). Désormeaux, A., Coutu, M., Medjahed, H., Paheo, B., Hershhorn, A., Gu, C., Xiang, S.H., Mao, Y., Sodroski, J., and Finzi, A. (2013). The highly onserved layer-3 omponent of the HIV-1 gp120 inner domain is ritial for CD4-required onformational transitions. J. Virol. 87, Doores, K.J., Bonomelli, C., Harvey, D.J., Vasiljevi, S., Dwek, R.A., Burton, D.R., Crispin, M., and Sanlan, C.N. (2010). Envelope glyans of immunodefiieny virions are almost entirely oligomannose antigens. Pro. Natl. Aad. Si. U S A 107, Finzi, A., Xiang, S.H., Paheo, B., Wang, L., Haight, J., Kassa, A., Danek, B., Panera, M., Kwong, P.D., and Sodroski, J. (2010). Topologial layers in the HIV-1 gp120 inner domain regulate gp41 interation and CD4-triggered onformational transitions. Mol. Cell 37, Garlik, R.L., Kirshner, R.J., Ekenrode, F.M., Tarpley, W.G., and Tomih, C.S. (1990). Esherihia oli expression, purifiation, and iologial ativity of a trunated solule CD4. AIDS Res. Hum. Retroviruses 6, Go, E.P., Irungu, J., Zhang, Y., Dalpathado, D.S., Liao, H.X., Sutherland, L.L., Alam, S.M., Haynes, B.F., and Desaire, H. (2008). Glyosylation site-speifi analysis of HIV envelope proteins (JR-FL and CON-S) reveals major differenes in glyosylation site oupany, glyoform profiles, and antigeni epitopes aessiility. J. Proteome Res. 7, Guo, Q., Ho, H.T., Diker, I., Fan, L., Zhou, N., Friorg, J., Wang, T., MAuliffe, B.V., Wang, H.G., Rose, R.E., et al. (2003). Biohemial and geneti haraterizations of a novel human immunodefiieny virus type 1 inhiitor that loks gp120-cd4 interations. J. Virol. 77, Guttman, M., and Lee, K.K. (2013). A funtional interation etween gp41 and gp120 is oserved for monomeri ut not oligomeri, unleaved HIV-1 Env gp140. J. Virol. 87, Guttman, M., Sian, M., and Lee, K.K. (2011). Traking hydrogen/deuterium exhange at glyan sites in glyoproteins y mass spetrometry. Anal. Chem. 83, Guttman, M., Kahn, M., Garia, N.K., Hu, S.L., and Lee, K.K. (2012). Solution struture, onformational dynamis, and CD4-indued ativation in full-length, glyosylated, monomeri HIV gp120. J. Virol. 86, Guttman, M., Weis, D.D., Engen, J.R., and Lee, K.K. (2013). Analysis of overlapped and noisy hydrogen/deuterium exhange mass spetra. J. Am. So. Mass Spetrom. 24, Harris, A., Borgnia, M.J., Shi, D., Bartesaghi, A., He, H., Pejhal, R., Kang, Y.K., Depetris, R., Marozsan, A.J., Sanders, R.W., et al. (2011). Trimeri HIV-1 glyoprotein gp140 immunogens and native HIV-1 envelope glyoproteins display the same losed and open quaternary moleular arhitetures. Pro. Natl. Aad. Si. U S A 108, Helseth, E., Olshevsky, U., Furman, C., and Sodroski, J. (1991). Human immunodefiieny virus type 1 gp120 envelope glyoprotein regions important for assoiation with the gp41 transmemrane glyoprotein. J. Virol. 65, Iyer, S.P., Franti, M., Krauhuk, A.A., Fish, D.N., Ouattara, A.A., Roux, K.H., Krawie, L., Dey, A.K., Beddows, S., Maddon, P.J., et al. (2007). Purified, proteolytially mature HIV type 1 SOSIP gp140 envelope trimers. AIDS Res. Hum. Retroviruses 23, Jardine, J., Julien, J.P., Menis, S., Ota, T., Kalyuzhniy, O., MGuire, A., Sok, D., Huang, P.S., MaPherson, S., Jones, M., et al. (2013). Rational HIV immunogen design to target speifi germline B ell reeptors. Siene 340, Julien, J.P., Cupo, A., Sok, D., Stanfield, R.L., Lyumkis, D., Deller, M.C., Klasse, P.J., Burton, D.R., Sanders, R.W., Moore, J.P., et al. (2013a). Crystal struture of a solule leaved HIV-1 envelope trimer. Siene 342, Julien, J.P., Lee, J.H., Cupo, A., Murin, C.D., Derking, R., Hoffenerg, S., Caulfield, M.J., King, C.R., Marozsan, A.J., Klasse, P.J., et al. (2013). Asymmetri reognition of the HIV-1 trimer y roadly neutralizing antiody PG9. Pro. Natl. Aad. Si. U S A 110, Khayat, R., Lee, J.H., Julien, J.P., Cupo, A., Klasse, P.J., Sanders, R.W., Moore, J.P., Wilson, I.A., and Ward, A.B. (2013). Strutural haraterization of leaved, solule HIV-1 envelope glyoprotein trimers. J. Virol. 87, Kwon, Y.D., Finzi, A., Wu, X., Dogo-Isonagie, C., Lee, L.K., Moore, L.R., Shmidt, S.D., Stukey, J., Yang, Y., Zhou, T., et al. (2012). Unliganded HIV-1 gp120 ore strutures assume the CD4-ound onformation with regulation y quaternary interations and variale loops. Pro. Natl. Aad. Si. U S A 109, Kwong, P.D., Doyle, M.L., Casper, D.J., Ciala, C., Leavitt, S.A., Majeed, S., Steeneke, T.D., Venturi, M., Chaiken, I., Fung, M., et al. (2002). HIV-1 evades antiody-mediated neutralization through onformational masking of reeptor-inding sites. Nature 420, Leonard, C.K., Spellman, M.W., Riddle, L., Harris, R.J., Thomas, J.N., and Gregory, T.J. (1990). Assignment of intrahain disulfide onds and haraterization of potential glyosylation sites of the type 1 reominant human immunodefiieny virus envelope glyoprotein (gp120) expressed in Chinese hamster ovary ells. J. Biol. Chem. 265, Lin, P.F., Blair, W., Wang, T., Spier, T., Guo, Q., Zhou, N., Gong, Y.F., Wang, H.G., Rose, R., Yamanaka, G., et al. (2003). A small moleule HIV-1 inhiitor that targets the HIV-1 envelope and inhiits CD4 reeptor inding. Pro. Natl. Aad. Si. U S A 100, Liu, J., Bartesaghi, A., Borgnia, M.J., Sapiro, G., and Suramaniam, S. (2008). Moleular arhiteture of native HIV-1 gp120 trimers. Nature 455, Lyumkis, D., Julien, J.P., de Val, N., Cupo, A., Potter, C.S., Klasse, P.J., Burton, D.R., Sanders, R.W., Moore, J.P., Carragher, B., et al. (2013). Cryo- EM struture of a fully glyosylated solule leaved HIV-1 envelope trimer. Siene 342, Madani, N., Perdigoto, A.L., Srinivasan, K., Cox, J.M., Chruma, J.J., LaLonde, J., Head, M., Smith, A.B., 3rd, and Sodroski, J.G. (2004). Loalized hanges in the gp120 envelope glyoprotein onfer resistane to human immunodefiieny virus entry inhiitors BMS-806 and #155. J. Virol. 78, Marsisin, S.R., and Engen, J.R. (2010). Hydrogen exhange mass spetrometry: what is it and what an it tell us? Anal. Bioanal. Chem. 397, MLellan, J.S., Panera, M., Carrio, C., Gorman, J., Julien, J.P., Khayat, R., Louder, R., Pejhal, R., Sastry, M., Dai, K., et al. (2011). Struture of HIV-1 gp120 V1/V2 domain with roadly neutralizing antiody PG9. Nature 480, Myszka, D.G., Sweet, R.W., Hensley, P., Brigham-Burke, M., Kwong, P.D., Hendrikson, W.A., Wyatt, R., Sodroski, J., and Doyle, M.L. (2000). Struture 22, , July 8, 2014 ª2014 Elsevier Ltd All rights reserved 983