Functional Links between the Fusion Peptide-proximal PolarSegment and Membrane-proximal Region of HumanImmunodeficiency Virus gp41 in Distinct Phases ofMembrane Fusion*□S
Received for publication, April 26, 2007, and in revised form, May 24, 2007 Published, JBC Papers in Press, May 25, 2007, DOI 10.1074/jbc.M703485200
Anna K. Bellamy-McIntyre‡§1, Chan-Sien Lay‡¶1, Severine Bar�2, Anne L. Maerz‡, Gert H. Talbo**,Heidi E. Drummer‡§¶, and Pantelis Poumbourios‡¶3
From the ‡Macfarlane Burnet Institute for Medical Research and Public Health, Prahran, Victoria 3004, Australia, the §Departmentof Microbiology, Monash University, Clayton, Victoria 3070, Australia, the ¶Department of Microbiology and Immunology, theUniversity of Melbourne, Parkville, Victoria 3058, Australia, the �Department of Cell Biology, Institut Cochin, Paris, France 75014,and **Proteomics Centre, Baker Heart Research Institute, Melbourne, Victoria 3004, Australia
The binding of CD4 and chemokine receptors to the gp120attachment glycoprotein of human immunodeficiency virustriggers refolding of the associated gp41 fusion glycoproteininto a trimer of hairpinswith a 6-helix bundle (6HB) core. Theseevents lead tomembrane fusion and viral entry. Here, we exam-ined the functions of the fusion peptide-proximal polar segmentand membrane-proximal Trp-rich region (MPR), which areexterior to the 6HB. Alanine substitution of Trp666, Trp672,Phe673, and Ile675 in the MPR reduced entry by up to 120-foldwithout affecting gp120-gp41 association or cell-cell fusion.TheL537Apolar segmentmutation led to the loss of gp120 fromthe gp120-gp41 complex, reduced entry by�10-fold, but did notaffect cell-cell fusion. SimultaneousAla substitutionofLeu537withTrp666, Trp672, Phe673, or Ile675 abolished entry with 50–80%reductions in cell-cell fusion. gp120-gp41 complexes of fusion-de-fective double mutants were resistant to soluble CD4-inducedshedding of gp120, suggesting that their ability to undergo recep-tor-induced conformational changes was compromised. Consist-ent with this idea, a representative mutation, L537A/W666A, ledto an �80% reduction in lipophilic fluorescent dye transferbetween gp120-gp41-expressing cells and receptor-expressingtargets, indicating a block prior to the lipid-mixing phase. TheL537A/W666A double mutation increased the chymotrypsinsensitivity of the polar segment in a trimer of hairpins model,comprising the 6HB core, the polar segment, and MPR linkedN-terminally to maltose-binding protein. The data indicate thatthe polar segment and MPR of gp41 act synergistically in form-ing a fusion-competent gp120-gp41 complex and in stabilizingthe membrane-interactive end of the trimer of hairpins.
The envelope glycoprotein complex (Env)4 of human immu-nodeficiency virus type 1 (HIV-1) comprises a trimer of recep-tor binding gp120 subunits in non-covalent association with atrimer of transmembrane gp41 subunits on the surface ofinfected cells and virions. Viral entry is initiated when gp120binds to cell-surface CD4 molecules, inducing structuralchanges within the gp120 core domain, and leading to the for-mation of the binding site for the chemokine co-receptors,CCR5 and/or CXCR4 (1–4). gp120 comprises 5 variable loops(V1–V5) that exist outside the gp120 core; V3, and to a lesserdegree V1 and V2, contribute to chemokine receptor bindingspecificity (5–7). The sequential binding of gp120 to CD4 andchemokine receptor triggers the refolding of gp41 into a tri-mer of hairpins, which mediates membrane fusion (8, 9).gp41 is a class I fusion glycoprotein, being structurallyhomologous to the fusion glycoproteins of other retrovi-ruses, orthomyxoviruses, paramyxoviruses, filoviruses, andcoronaviruses. The gp41 ectodomain is comprised of anN-terminal fusion peptide, connected through a flexiblepolar segment to a coiled coil-forming amphipathic �-helix(N-helix), a centrally located disulfide-bonded loop, a C-ter-minal amphipathic �-helix (C-helix), and a membrane-prox-imal tryptophan-rich region (MPR) (10–16). The ectodo-main is anchored to the viral envelope by a C-terminallylocated transmembrane domain (TMD), which precedes an�150-residue cytoplasmic domain.Themajority of the gp41 ectodomain appears to be buried by
the gp120 trimer. This model for gp41 in the context of prefu-sion Env is based on the findings that the epitopes of mono-clonal antibodies (mAbs) encompassing the fusion peptide andpolar segment (residues 521–538), disulfide bonded region(579–613), andC-helix (644–663), are largely occluded in pre-
* This work was supported by the National Health and Medical ResearchCouncil of Australia Grants 296200 and 345413, American Foundation forAIDS Research Grant 106610-36-RGNT, and Sidaction. The costs of publi-cation of this article were defrayed in part by the payment of page charges.This article must therefore be hereby marked “advertisement” in accord-ance with 18 U.S.C. Section 1734 solely to indicate this fact.
□S The on-line version of this article (available at http://www.jbc.org) containssupplemental Fig. S1.
1 Both authors contributed equally to this work.2 Present address: Program Infection and Cancer, Abt. F010 and INSERM
U701, Deutsches 10 Krebsforschungszentrum, Heidelberg, Germany.3 To whom correspondence should be addressed: 85 Commercial Rd.,
Melbourne, Victoria 3004, Australia. Fax: 613-9282-2100; E-mail:apoumbourios@ burnet.edu.au.
fusion Env but become exposed following interaction betweengp120 and recombinant soluble CD4 (sCD4) or receptorexpressing cells (17–20). By contrast, the epitopes of mAbs 2F5and 4E10 within the MPR (662–667) are exposed in gp120-gp41 prior to CD4 binding indicating an external location forthis region (17, 20, 21). These data are largely reflected in astructural model of simian immunodeficiency virus (SIV) Envderived by cryoelectron tomography (22). In this model, a largeglobular domain comprises a gp120 trimer surrounding thegp41 ectodomain and is anchored to the plasmamembrane by asplayed tripod, which corresponds to the MPR in an extensiveinteraction with the viral envelope. However, an alternativecryoelectron tomographic model of SIV gp120-gp41 revealed amore compact head domain linked to the membrane via a stalkwith no evidence of a tripod-like structure (23). The structureof the MPR in situ therefore remains an open question.gp120-receptor interactions trigger the membrane fusion
cascade, beginning with insertion of the fusion peptide into theouter leaflet of the target membrane and gp41 adopting a pre-hairpin intermediate conformation that bridges the viral andcellular membranes (24–26). In the prehairpin intermediate,the trimeric coiled coil of N-helices is exposed and available forinteractionwith peptides derived from theC-helix that can pre-vent subsequent stages of the fusion cascade. Likewise, exposedC-helices are available for interaction with synthetic analoguesof the N-helix, which also inhibit fusion (18, 19, 27–29). Theseinteractions mimic the antiparallel packing of native C-helicesinto hydrophobic grooves on the exterior of the coiled-coil thatforms the 6-helix bundle (6HB) core, bringing together the N-and C-terminal membrane inserted ends of gp41 (the fusionpeptide and TMD), and the associated viral and cellular mem-branes for merger (30–34). Three phases of cell-cell and virus-cell membrane fusion have been discerned experimentally forclass I fusion proteins: lipid mixing or hemifusion; the openingof a small pore through which small solutes can pass; and poreexpansion (35–39). These events precede entry of the viralnucleocapsid into the cytosol. Six-helix bundle formation andhemifusion appear to be co-dependent processes as lipids thatpromote positive spontaneous membrane curvature and blockhemifusion also prevent completion of the trimer of hairpinsfold (37).How conserved sequences located outside the 6HB core
domain of gp41, such as the N-terminally located polar seg-ment adjacent to the fusion peptide, and the C-terminal Trp-richMPR linking theC-helix to theTMD (40), contribute to themembrane fusion mechanism is not clearly understood. TheMPR is of particular interest as it is a highly conserved regionthat overlaps with the C terminus of the Fuzeon inhibitor pep-tide and encompasses the epitopes of the two most broadlyneutralizing mAbs available, 2F5 and 4E10 (41–43). The MPRis functionally relevant as simultaneousAla replacements of theconserved Trp666-Trp670-Trp672-Trp678-Trp680 cluster blockfusion pore opening (44), an effect that may be related to thelipid destabilizing properties of this sequence (45–48).Like most sequences within the gp41 ectodomain, the MPR
undergoes structural transitions in the fusion cascade. Forexample, the 2F5 epitope (Glu657-Trp670) is available for mAbbinding in prefusion and prehairpin intermediate forms of Env
but becomes occluded when the 6HB forms (21, 49–51). Thischange in antigenicity corresponds to a structural transition forthe N-terminal portion (Glu657-Asp664) from an extended con-formation to�-helix (30–34, 51).Helical conformation has alsobeen observed for theMPRwhen bound to lipid (52). Antibodybinding studies suggest that the polar segment and MPRoccupy internal and external locations, respectively, in prefu-sion gp120-gp41 complexes (17, 20, 21). By contrast, thesesequenceswould become apposed at themembrane-interactiveend of the fusion-activated trimer of hairpins by formation ofthe 6HB. Consistent with this idea, extension of the 6HB core toinclude residues 528–535 of the polar segment and residues669–677 of the MPR confers stability to a trimer of hairpinsmodel protein (53).In this study we assessed the functional role of the polar seg-
ment and MPR. Alanine replacement of conserved residueswithin the MPR did not affect cell-cell fusion but resulted in8–120-fold reductions in viral entry. Simultaneous Ala replace-ment ofMPR residues with Leu537 of the polar segment inhibitedcell-cell fusion, viral entry, and sCD4-induced dissociation ofgp120 from the Env complex. By using a trimer of hairpins modelof gp41, we found that simultaneous mutations in the polar seg-ment andMPR destabilized themembrane interactive end of thisfusion-activated structure.We propose that theMPR has distinctroles in different stages of the fusion cascade.
EXPERIMENTAL PROCEDURES
HIV-1 Env Expression Vectors—Preparation of the cytomeg-alovirus promoter-driven HIV-1AD8 Env expression vector,pCDNA3.1-AD8env, is described elsewhere (54). In vitromutagenesis of the gp41 region was carried out by overlapextension PCR techniques. Bacteriophage T7 promoter-drivenEnv expression vectors based on pTM.1 (55) were generated bytransferring the HIV-1AD8 env fragment bounded by NdeI andXhoI, from pCDNA3.1-AD8env to pTMenv.2, giving pTM-AD8env (56). DNA sequences were confirmed using the ABIBigDye terminator reagent. Env was either expressed in a Rev-dependent manner from the cytomegalovirus promoter ofpCDNA3.1-AD8env, or in a Rev-independentmanner from thebacteriophage T7 promoter in pTM-AD8env (55, 56). For Rev-independent expression, the cells were cotransfected withpTM-AD8env and pCAG-T7, the latter directing expression ofbacteriophage T7 RNA polymerase from a cytomegalovirusimmediate-early enhancer and chicken �-actin promoter (57).Alternatively, the pTM-AD8env-transfected cells were infectedwith the recombinant vaccinia virus vTF7.3, which directsexpression of bacteriophage T7 RNA polymerase (55).Cells—293T, BHK21, HeLa, and U373MG-CCR5 cells were
maintained in Dulbecco’s modification of minimal essentialmedium, 10% fetal calf serum (complete medium) and trans-fected with expression vectors using FuGENE 6 (Roche).U87.CD4.CCR5 cells were maintained in Dulbecco’s modifiedminimal essential medium, 15% fetal calf serum supplementedwith puromycin (0.1 mg/ml) and G418 (0.3 mg/ml).Antibodies—Immunoglobulin G number 14 (IgG 14) was
purified from the plasma of a HIV-1 positive individual usingprotein A-Sepharose (Amersham Biosciences). Anti-gp41mAbswere obtained through theAIDSResearch andReference
Functions of the HIV gp41 Polar and Membrane Proximal Regions
AUGUST 10, 2007 • VOLUME 282 • NUMBER 32 JOURNAL OF BIOLOGICAL CHEMISTRY 23105
Reagent Program, NIAID, National Institutes of Health, fromG. Lewis (C8 (58)), S. Zolla-Pazner (126-6, 98-6 (59)), and R.Myers (Md.1 (60)). Goat anti-Env 2–3 (nonglycosylated HIV-1SF2 gp120) antibody was obtained through the AIDS Researchand Reference Reagent Program, NIAID, National Institutes ofHealth, from K. Steimer (61).Processing of Env Glycoproteins—At 24-h post-transfection,
293T cells were lysed for 10 min on ice in phosphate-bufferedsaline (PBS) containing 1%TritonX-100, 0.02% sodium azide, 1mM EDTA. Lysates were clarified by centrifugation for 10 minat 10,000 � g at 4 °C prior to SDS-PAGE under reducing con-ditions. Proteins were transferred to nitrocellulose and probedwith mAb C8. The immunoblots were developed with horse-radish peroxidase-conjugated rabbit IgG to mouse Ig andenhanced chemiluminescence or with Alexa Fluor 680-con-jugated goat anti-mouse Ig (Invitrogen) and scanning in aLI-COR Odyssey infrared imager.Biosynthetic Labeling and Immunoprecipitation—293T cells
were cotransfected with pTM-AD8env and pCAG-T7 plas-mids. At 48-h post-transfection, the cells were incubated for 30min in cysteine and methionine-deficient medium (MP Bio-medicals, Seven Hills, New South Wales, Australia), and thenlabeled for either 20 or 45 min with 150 �Ci of Tran35S-label(MP Biomedicals). The cells were then washed and chased incomplete medium for 6 h prior to lysis. All incubations wereperformed at 37 °C in 5% CO2. Cell lysates and clarified cul-ture supernatants were immunoprecipitated with variousantibodies and protein G-Sepharose and subjected to SDS-PAGE under reducing conditions, as described previously(54). The labeled proteins were visualized by scanning in aFuji phosphorimager.Luciferase Reporter Assay of Cell to Cell Fusion—293T effec-
tor cells (250,000 cells per well of a 12-well culture plate) werecotransfected with pCDNA3.1-AD8env or pTM-AD8env andpCAG-T7 plasmids. BHK21 target cells (250,000 cells per wellof a 12-well culture plate) were cotransfected with pT4luc (62)and pc.CCR5 (obtained from the AIDS Research and ReferenceReagent Program from N. Landau (63)). At 24 h post-transfec-tion, targets and effectors were each resuspended in 1 ml ofcompletemedium; targets (100�l) were co-culturedwith effec-tors (100 �l) in a 96-well plate for a further 18 h at 37 °C. Thecells were assayed for luciferase activity using the PromegaSteadyGlo reagent (Promega, Madison, WI).LipidMixing Assays—Lipidmixing assays were performed as
described previously (64). HeLa effector cells (200,000 cells perwell of a 6-well plate) were transfectedwith wild type ormutantEnv-expression vectors. At 5 h post-transfection, the cells wereinfected with vTF7.3. At 18 h post-transfection, Env expressingcells were labeled with 2 �M DiO (3,3�-dioctadecyloxacarbo-cyanine perchlorate; green fluorescence, excitation at 484 nm,emission at 501 nm (Molecular Probes, Eugene, OR)), whereastarget U373MG-CCR5 cells were labeled with 2 �M DiI (1,1�-dioctadecyl-3,3,3�,3�- tetramethylindocarbocyanine perchlo-rate; red fluorescence, excitation at 549 nm, emission at 565nm) for 2 min at room temperature. After washing twice withPBS, target cells were detached with PBS, 1 mM EDTA andadded to an equivalent number of adherent effector cells. Fol-lowing 2 h of co-culture at 37 °C, the cells were detached with
trypsin, and fixed with 4% formaldehyde in PBS. Two-colorfluorescence analysis was performed on an Epics XL flowcytometer. Cells displaying �4 and �70 arbitrary log units inthe green and red wavelengths, respectively, were scored asdouble-fluorescent cells.Single Cycle Infectivity Assays—Env-pseudotyped luciferase
reporter viruses were produced by cotransfecting 293T cells(350,000 cells per well of a 6-well Linbro culture plate) witheither 1 �g of pCDNA3.1-AD8env plus 1 �g of luciferasereporter virus vector, pNL4.3.Luc.R�E� (obtained from theAIDS Research and Reference Reagent Program from N.Landau (65)), or 1�g of pTM-AD8env, 1�g of pCAG-T7 plus 1�g of pNL4.3.Luc.R�E� using FuGENE 6. Supernatants con-taining pseudotyped virions were harvested at 72 h post-trans-fection and filtered through 0.45-�m filters. The infectivity ofpseudotyped viruses was determined in U87.CD4.CCR5 cells(obtained from the AIDS Research and Reference Reagent Pro-gram fromH. Deng and D. Littman (66)) as described (54). Theprotein composition of pseudotyped virions was assessed byfirst pelleting the virions from 9 ml of transfection supernatantthrough a 1.5-ml 25% (w/v) sucrose/PBS cushion (BeckmanSW41 Ti rotor, 25,000 rpm, 2.5 h, 4 °C) followed by reducingSDS-PAGE in 7.5–15% polyacrylamide gradient gels andWest-ern blotting with IgG 14 and goat anti-Env 2–3. The immuno-blots were developed with IRDye 800CW-conjugated rabbitanti-human Ig (Rockland) or Alexa Fluor 680-conjugated don-key anti-goat Ig (Invitrogen), respectively, and scanning in aLI-COR Odyssey infrared imager.Soluble CD4-induced Shedding of gp120—293T cells were
transfected with Env expression plasmids as described above.At 24 h post-transfection, the cells were starved with methio-nine and cysteine-free medium for 30 min and then pulselabeled with 150 �Ci of Tran35S-label for 45 min. The labeledcells were washed and then chased for 5 h with completemedium in the presence or absence of sCD4 (15�g of sCD4 per0.6 ml of medium). All incubations were at 37 °C in 5% CO2.The cells and culture supernatants were processed for immu-noprecipitation with IgG 14 as described above.Expression and Purification of MBP/gp41 Chimeras—The
MBP/gp41(528-L-677) chimeras used in this study are derivedfromHIV-2ST and comprise the N-terminal polar segment andcoiled coil (Ala528-Trp596), linked through Ser-Gly-Gly-Arg-Gly-Gly, to the C-terminal helix and membrane proximal seg-ment (Trp610-Ser677) (32, 53). Residuenumbering is according tothe HXB2R convention. Ala528 of HIV-2ST gp41 was fused to theC terminus of MBP through an Asn-Ala linker incorporating auniqueNotI site. Overlap extension PCRwas used to generate themodifiedHIV-2ST gp41ectodomain fragment:Ala528-Trp596-Ser-Gly-Gly-Arg: forward primer, 5�-ATAAGAATGCGGCCGCG-ATGGGCGCGGCGTCCTTGACG, reverse primer, 5�-ACC-CCCACGACCCCCGGACCATGAATTTAGTTGCGCCTG-GTC; Arg-Gly-Gly-Trp610-Ser677: forward primer, 5�-TCCG-GGGGTCGTGGGGGTTGGGTAAATGACACCTTAACG-CCTG, reverse primer, 5�-CTGAATATAGTCGACTTAGG-AGGTTAAATCAAACCA. The PCR products were ligatedinto theMBP expression vector throughNotI-SalI. The L537A,W666A, and L537A/W666A mutations were introduced toMBP/gp41(528-L-677) by overlap extension PCR. DNA
Functions of the HIV gp41 Polar and Membrane Proximal Regions
23106 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 32 • AUGUST 10, 2007
sequences were confirmed using ABI prism BigDye terminator(Applied Biosystems, Foster City, CA). MBP/gp41 chimeraswere induced in Escherichia coli strain BL21(DE3) and purifiedby amylose-agarose affinity chromatography (New EnglandBiolabs) and gel filtration as described (53). MBP/gp41 trimerswere proteolyzedwith sequencing grade chymotrypsin (Roche)and analyzed by SDS-PAGE in 12–17% polyacrylamide gradi-ent gels as described (53).Mass Spectrometry—The digested proteins were mass ana-
lyzed by linearMALDI-MS (Bruker Daltonics, Germany) using2,5-dihydroxybenzoic acid/5-methylsalicylic acid (9:1) (superDHB, Bruker Daltonics) as a matrix. The locations of chymo-tryptic cleavage sites were verified by in-gel tryptic digestion ofthe protein band in question. The excised bandswerewashed in50 mM NH4HCO3/acetonitrile (1:1) (v/v) for 20 min and dried.Then, 0.3�g of trypsin in 25mMNH4HCO3were added and theproteins digested for 2 h at 37 °C. The resulting peptides wereextracted in 5% formic acid/acetonitrile (1:1) (v/v) and dried. Theextracts were dissolved in acetonitrile/water/formic acid (30:69:1)and applied to C18 ziptips. The peptides were eluted with DHBmatrix in acetonitrile, 0.1% trifluoroacetic acid (aq) (6:4) and theeluates were analyzed byMALDI-MS in reflector mode.
RESULTS
Alanine Scan of the FusionPeptide-proximal andMembrane-proximal Regions of HIV-1AD8 gp41—Previously, we found thatextension of the 6-helix bundle core of gp41 to include Ala528-Leu535, of the N-terminal polar segment, and Phe669-Ser677,within the MPR, conferred stability to a model of the HIV-2STgp41 trimer of hairpins, MBP/gp41 (53). In this paper, weexamined the functional relevance of these sequences followingalanine replacement of component amino acids that are con-served in HIV-1, HIV-2, and SIV (Fig. 1). Met530 and Leu537 inthe polar segment andTrp672 in theMPRare found in allHIV-1,
HIV-2, and simian immunodeficiency virus isolates. Also withinthe MPR, Trp666, Phe673, and Ile675 are conserved in the majorityof isolates. Notably, three conserved aromaticMPR residues con-tribute the majority of contacts with neutralizing mAb: Trp666with 2F5 (51), Trp672 and Phe673 with 4E10 (67).
We first examined the synthesis of W666A, W672A, F673A,D674A, and I675A MPR mutants by Western blotting withmAb C8 (directed to the gp41 cytoplasmic domain) and foundthat the mutated gp160 precursors were expressed and pro-cessed to gp41 in a similar manner to wild type (Fig. 2A). Thefusion activities of the MPR mutants were assessed using aluciferase reporter assay of cell-cell fusion. 293T effector cellswere cotransfected with pCDNA3.1-AD8env and pCAG-T7,whichdirects expressionof bacteriophageT7RNApolymerase. Inthis assay, luciferase is induced after fusion of Env-293T effectorswith CD4-CCR5-expressing BHK21 cell targets that contain theluciferase open reading frame under the control of the bacterio-phage T7 promoter. Consistent with the observation that themutants were expressed and cleaved normally, the MPRmutantswere also able tomediate cell-cell fusion to a similar extent aswildtype across a range of Env expression levels (Fig. 2B).We next investigated whether luciferase reporter viruses
pseudotyped with the MPR mutants could establish a singlecycle of replication in U87.CD4.CCR5 target cells. Fig. 2C indi-cates that viral entry was reduced by�120-fold forW666A andI675A, by 35-fold forW672A, and by 8-fold for F673A. By con-trast, Ala replacement of the acidic side chain of Asp674 wastolerated (Fig. 2C). Western blotting of Env-pseudotypedHIV-1 particles that had been pelleted through 25% (w/v)sucrose indicated that the pseudovirions had retained gp120 inall cases (Fig. 2D). These data suggest that bulky hydrophobicside chains of the MPR are important for viral fusion but notcell-cell fusion.
FIGURE 1. Alignment of gp41 amino acid sequences. Secondary structures observed by NMR for the micelle-inserted fusion peptide and adjacent polarsegment (24), are indicated by a line for random coil and cylinder for �-helix. The N- and C-terminal portions, respectively, of the coiled coil-forming N-helix andantiparallel C-helix of the 6-helix bundle core domain, as determined by x-ray crystallography (31, 33), are shown as truncated cylinders. The residues of theN-terminal polar segment and C-terminal MPR that were replaced with alanine in HIV-1AD8 and HIV-2ST gp41 in this study are shown and in bold type.
Functions of the HIV gp41 Polar and Membrane Proximal Regions
AUGUST 10, 2007 • VOLUME 282 • NUMBER 32 JOURNAL OF BIOLOGICAL CHEMISTRY 23107
Effect of Simultaneous Mutations in the N-terminal PolarSegment and MPR—Our previous study with the MBP/gp41chimeric model of the HIV-2 gp41 trimer of hairpins suggestedthat the polar segment and MPR confer trimer stability in asynergistic manner (53).Whereas these terminal sequences arelikely to be brought together in late fusion forms of gp41 by
6-helix bundle formation, theyappear to be spatially separate in theprefusion gp120-gp41 complexwiththe polar segment located in theinterior of the trimer, whereas theMPR is external (17, 20, 21). Wetherefore examined the effects ofsimultaneous Ala substitutions inthe polar segment andMPR at var-ious stages of the fusion cascade.For these experiments, we usedthe bacteriophage T7 promoter-driven vector pTM.1 to overcomean Env expression defect resultingfrom disruption of stem-loop 2Aof the Rev responsive element bythe L537A mutation (data notshown). Western blotting of theL537A/W666A, L537A/W672A,L537A/F673A, L537A/ D674A,and L537A/I675A double mu-tants, and L537A andW666A singlemutants, revealed gp160 and gp41at levels that were comparable withwild type (Fig. 3A). The gp120-an-choring abilities of gp41 L537A andthe double mutants were confirmedby immunoprecipitation of pulse-chase biosynthetically labeled Env-transfected 293T cells. Whereassimilar amounts of gp160 were im-munoprecipitated from the lysates ofwild typeandmutantEnv-transfectedcells, slightly more gp120 was shedinto culture supernatants for theL537A-containing mutants (Fig. 3B),suggesting that L537A is associatedwith a subtle gp120-shedding pheno-type. TheM530Amutant was largelydevoid of gp120, consistent with de-fective biosynthesis, andwas not ana-lyzed further.The fusogenic abilities of L537A
and L537A-containing double mu-tants were assessed in the luciferasereporter assay of cell-cell fusion. Fig.3C shows that the L537A mutantwas able to mediate fusion at wildtype levels indicating that the shed-ding phenotype of this mutantobserved in Fig. 3B was not associ-ated with loss of cell-cell fusion
function. However, the introduction of W666A, W672A,F673A, D674A, and I675AMPRmutations to the L537A back-ground resulted in decreases in fusion activity of 80% forL537A/W666A, 69% for L537A/W672A, 61% for L537A/F673A, 56% for L537A/D674A, and 63% for L537A/I675A (Fig.3C). The fusion activities of double mutants were significantly
FIGURE 2. A, synthesis and processing of HIV-1AD8 Env glycoprotein mutants in 293T cells. 293T cells weretransfected with the indicated amounts of wild type (WT) and mutated pCDNA3.1-AD8env expression plas-mids. At 36 h post-transfection, the cells were lysed and subjected to reducing SDS-PAGE in 10% polyacrylam-ide gels. Proteins were transferred to nitrocellulose prior to Western blotting with mAb C8 and Alexa Fluor680-conjugated goat immunoglobulins to mouse immunoglobulins and scanning in an LI-COR Odyssey infra-red imager. Asterisk, �80-kDa species corresponding in molecular mass to a biosynthetic intermediate ordegradation product of gp160 (96). Representative of three independent experiments. B, cell-cell fusion activ-ities of Env glycoprotein mutants using a luciferase reporter assay. 293T effector cells were cotransfected withthe T7 polymerase expression vector, pCAG-T7, together with the indicated amounts of wild type (WT) ormutated pCDNA3.1-AD8env vectors. BHK21 target cells were cotransfected with pT4luc and pc.CCR5. At 18 h post-transfection, the effectors and targets were mixed and co-cultured for 24 h prior to lysis and assay for luciferaseactivity. Mean relative light units (RLU) � standard error (error bars) is shown from at least 4 independent experi-ments. C, ability of Env mutants to mediate entry of Env-pseudotyped luciferase reporter viruses. Env-pseudotyped luciferase reporter viruses were prepared by cotransfecting 293T cells with 1 �g ofpCDNA3.1-AD8env plus 1 �g of pNL4.3.Luc.R�.E�. At 72 h post-transfection, the virus-containing supernatantswere filtered and used to infect U87.CD4.CCR5 monolayers for 18 h. The inoculum was then replaced with freshmedium and the cells assayed for luciferase activity at 52 h postinfection. Luciferase activity was normalizedagainst the reverse transcriptase activity present in each virus inoculum. Mean RLU � standard error from atleast 3 independent experiments are shown. *, p � 0.05; **, p � 0.01 relative to wild type (two sample t testassuming unequal variances). D, protein content of virions. Env-pseudotyped virions were pelleted through a25% sucrose cushion and subjected to SDS-PAGE under reducing conditions in 7.5–15% polyacrylamide gra-dient gels. Viral proteins were visualized by Western blotting with anti-Env 2–3 and Alexa Fluor 680-conjugateddonkey anti-goat Ig (upper panel) or IgG 14 and IRDye 800CW-conjugated rabbit anti-human Ig (lower panel).
Functions of the HIV gp41 Polar and Membrane Proximal Regions
23108 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 32 • AUGUST 10, 2007
higher than the “No Env” control (p � 0.04). The M530Amutant lacked cell-cell fusion activity, consistent with its proc-essing defect.
We next examined the ability of the combined L537A/MPRmutants to mediate viral entry. Wild type and L537A-contain-ing mutant pseudovirions were produced for functional
FIGURE 3. Synthesis, processing, and cell-cell fusion activities of Env mutants expressed from the bacteriophage T7 promoter. A, 293T cells were cotransfectedwith pTM-Env and pCAG-T7 expression vectors. At 24 h post-transfection, the cells were lysed and subjected to reducing SDS-PAGE in 7.5–15% gradient polyacryl-amide gels. The proteins were visualized by Western blotting with mAb C8 and Alexa Fluor 680-conjugated goat anti-mouse Ig. Representative of 3 independentexperiments. B, gp120-anchoring ability of gp41 mutants. 293T cells were cotransfected with pTM-Env and pCAG-T7 vectors. At 48 h post-transfection, cells werelabeled with Tran35S-label for 45 min and chased in complete medium for 6 h before lysis. Cell lysates (C) and clarified culture supernatants (S) were immunoprecipi-tated with IgG 14 and protein G-Sepharose. gp160 and gp120 bands were visualized following SDS-PAGE in 8–12% polyacrylamide gradient gels under reducingconditions and scanning in a PhosphorImager. The image was prepared from samples run on 2 gels in a single experiment; representative of three independentexperiments. C, cell-cell fusion activity of gp41 mutants. 293T effector cells were cotransfected with pTM.1-Env and pCAG-T7 expression vectors. BHK21 target cellswere cotransfected with pT4luc and pc.CCR5. At 24 h post-transfection, the effectors and targets were co-cultured for 18 h prior to assay for luciferase activity. MeanRLU � standard error (error bars) from three independent experiments are shown. *, p � 0.03, relative to wild type (two sample t test assuming unequal variances). D,single cycle entry of Env-pseudotyped luciferase reporter viruses. Env-pseudotyped luciferase reporter viruses were prepared by cotransfecting 293T cells with 1 �gof pTM-Env, 1 �g of pCAG-T7, and 1 �g of pNL4.3.Luc.R�.E� and infectivity determined as described for Fig. 2D. �, p � 0.005 relative to wild type; †, p � 0.004 relativeto L537A (two sample t test assuming unequal variances). E, protein content of virions. Env-pseudotyped virions were pelleted through a 25% sucrose cushion andsubjected to SDS-PAGE under reducing conditions in 7.5–15% polyacrylamide gradient gels. Viral proteins were visualized by Western blotting with anti-Env 2–3 andAlexa Fluor 680-conjugated donkey anti-goat Ig (upper panel) or IgG 14 and IRDye 800CW-conjugated rabbit anti human Ig (lower panel). F, conformation of the gp41domain of the L537A/W666A mutant. Pulse-chase biosynthetically labeled wild type (w) and L537A/W666A-mutated (m) glycoproteins were immunoprecipitatedwith the indicated mAbs and visualized with a PhosphorImager following reducing SDS-PAGE on 8–12% gradient polyacrylamide gels. The image was prepared fromsamples run on 2 gels in a single experiment.
Functions of the HIV gp41 Polar and Membrane Proximal Regions
AUGUST 10, 2007 • VOLUME 282 • NUMBER 32 JOURNAL OF BIOLOGICAL CHEMISTRY 23109
analysis by cotransfecting pTM-AD8env, pCAG-T7, andpNL4.3LucR�E� plasmids into 293T cells. The L537A muta-tion caused an �10-fold decrease in viral entry competence,consistentwith a lower level of gp120 retained on virions (Fig. 3,D and E). The introduction of W666A, W672A, F673A, andI675A MPR mutations to the L537A background ablated viralentry altogether even though the levels of virion-associatedgp120 for the double mutants were similar to that observed forL537A. Of these double mutants, only L537A/D674A retainedsubstantial entry activity, however, this function was decreased�5-fold with respect to L537A (p � 0.0022). These data indi-cate that Leu537 in the polar segment is important for gp120-gp41 stability in both cellular and viral contexts. Furthermore,the data suggest that hydrophobic residues within the polarsegment and MPR act together in membrane fusion in bothcellular and viral contexts.The folding of the gp41 domain of a representative fusion-
defective double mutant, L537A/W666A, was assessed byimmunoprecipitation with the human mAbs, Md.1, 126.6, and98.6, which bind to conformational epitopes in the C-terminalportion of the gp41 ectodomain (59, 60, 68). These mAbs areable to detect structural disturbances in gp41, as mutations inthe gp41 coiled-coil sequence that block gp160 oligomerizationalso block epitope formation (68). Furthermore, mAbs 126.6and 98.6 can recognize the hairpin form of gp41 (50). Mono-clonal antibody C8 binds to a linear epitope in the cytoplasmicdomain of gp41 (58) andwas used as a positive control, whereas12CA5 specific to the influenza hemagglutinin was included asan irrelevant antibody control. Fig. 3F shows that the L537A/W666A double mutant was efficiently immunoprecipitated bythe mAbs indicating that the gp41 domain of this mutantacquires a global conformation that is similar to the wild type.sCD4-induced Shedding of gp120—Soluble CD4 triggers the
formation of pre-hairpin intermediate conformations of gp41(18, 19) and can activate Env-mediated fusion with target cellsexpressing only CCR5 or CXCR4 co-receptors (9). Functionalgp120-gp41 complexes can also be induced to shed gp120 bysCD4 (69, 70).We examinedwhether sCD4 could induce gp120shedding from cell surface-expressed [35S]Met/[35S]Cys-la-beled gp120-gp41 to monitor the ability of the mutants toundergo conformational changes in response to receptor. Thetreatment of wild type and single MPR fusogenic mutants withsCD4 led to increased shedding of gp120 into culture superna-tants when compared with untreated controls in all cases (Fig.
4A). These data indicate that gp120-gp41 complexeswith singleMPRmutations have similar stabilities to the wild type and thatsCD4 can induce conformational changes in these mutants.Similar trends were observed for the fusogenic L537A andW666A mutants that had been expressed from the T7 pro-moter (Fig. 4B). However, relatively small or non-existentincreases in sCD4-induced gp120 shedding were observed forthe L537A/W666A, L537A/W672A, L537A/F673A, andL537A/I675A double mutants, which were defective for fusionand viral entry functions. The sCD4 shedding defect was not asapparent for L537A/D674A, consistent with the intermediatelevel of entry competence retained by this double mutant.These data suggest that gp120-gp41 complexes bearing L537A/W666A, L537A/W672A, L537A/F673A, and L537A/I675Adouble mutations cannot efficiently undergo CD4-inducedconformational changes, perhaps contributing to their defec-tive fusion and entry functions.Effects of Mutations on Lipid Mixing—If fusion defective
mutants lack the ability to undergo receptor-induced confor-mational changes in an efficient manner, then it follows thatthey will also be defective in mediating the lipid mixing stage offusion. This stage is considered to precede pore expansion,which is measured in the luciferase assay. We tested this ideawith a representative fusion defective mutant, L537A/W666A,by using flow cytometry to quantify lipophilic dye exchangebetween Env-expressingHeLa cells, labeled with the DiO greenfluorescent probe, andU373MG-CCR5 targets labeled with theDiI red fluorescent probe (64). Fig. 5 indicates that Env glyco-proteins with single L537A and W666A mutations retainedwild type levels of lipid mixing activity, consistent with theircell-cell fusion abilities. The control X4 Env of HIV-1BH8 wasunable to mediate lipid mixing with the U373MG-CCR5 tar-gets, confirming the specificity of the assay. The L537A/W666A double mutation inhibited lipid mixing function by�80%, indicating that these hydrophobic residues contributejointly to the early stages of fusion.Effects of Polar Linker andMPRMutations on gp41 Trimer of
Hairpins Stability—The fusion-activated trimer of hairpins/6HB conformation of gp41 is thought to facilitate virus-cellmembrane fusion by bringing membrane-inserted fusion pep-tides and TMDs into close proximity. The polar and MPR seg-ments link the 6HB to the terminal membrane-interactivesequences, and are therefore also likely to be juxtaposed in thislate fusion form of gp41 (Fig. 6A). We asked whether these
FIGURE 4. Soluble CD4-induced shedding of gp120. 293T cells were transfected with pCDNA3.1-AD8env vectors (A) or cotransfected with pTM-Env pluspCAG-T7 vectors (B). At 24 h post-transfection, the cells were labeled with 150 �Ci of Tran35S-label and then chased for 5 h in the absence (�) or presence (�)of 15 �g of sCD4. The culture supernatants (upper panels) and cell lysates (lower panels) were immunoprecipitated with IgG 14 and protein G-Sepharose andsubjected to SDS-PAGE in 8 –12% gradient gels followed by scanning in a PhosphorImager. The images were prepared from samples run on 4 gels in a singleexperiment.
Functions of the HIV gp41 Polar and Membrane Proximal Regions
23110 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 32 • AUGUST 10, 2007
linking sequences play a role in the formation and/or stability ofthe trimer of hairpins. For these studies, we used anMBP/gp41chimeric trimer of hairpins model, MBP/gp41(528-L-677),which comprises the HIV-2ST gp41 ectodomain fragmentAla528-Ser677 (53) with the interhelical disulfide-bonded region(Gly597-Pro609) replaced by Ser-Gly-Gly-Arg-Gly-Gly (32) (Fig.6A). This modification enabled the purification of MBP/gp41trimers by gel filtration (supplementarymaterials Fig. S1) at�2mg/liter of culture, corresponding to�7-fold higher yields thanversions containing the disulfide-bonded region (53). Wefound that chimeras comprised of corresponding HIV-1sequenceswere either insoluble or presented as high-molecularweight aggregates that could not be analyzed further (data notshown). We used limited chymotrypsin proteolysis, whichcleaves on the C-terminal side of Tyr, Phe, and Trp and to alesser extent Leu, Met, Ala, Asp, and Glu, to examine howL537A,W666A, andL537A/W666Amutations affect the integ-rity of the MBP/gp41(528-L-677) trimer of hairpins.Chymotrypsin released �49, �41, and �14.4-kDa products
from wild type MBP/gp41(528-L-677) trimer (Fig. 6B, bands1–3, respectively). When the protease:protein ratio wasincreased, an inverse correlation between the intensity of band1 versus that of bands 2 and 3 was observed (Fig. 6, B–D); bands2 and 3 are therefore chymotryptic peptides of band 1. Massspectrometric analysis of the 1:40 and 1:5 chymotryptic digestsof wild type MBP/gp41 indicated that the N- and C-helicalsequences of the gp41 core remained largely intact with a num-ber of their subfragments also released (Fig. 7). In-gel trypsindigestion and mass spectrometry was used to infer the C-ter-
minal boundaries of bands 1 and 2. Trypsin cleaves on theC-terminal side of Lys and Arg; therefore peptides terminatingwith a residue preferred by chymotrypsin indicate the locationof the chymotrypsin site(s). Trypsin treatment of band 1released the chimeric peptide, MBP(Gln356-Thr367)-Asn-Ala-gp41(Ala528-Arg542) (2,760.4 Da), and the gp41 peptides,Thr543-Lys560 (1,994.2 Da) and Gln562-Lys585 (2,813.6 Da); wedid not detect peptide(s) that were C-terminal to thesesequences. Consistent with the lowermolecular weight of band2, MBP(Gln356-Thr367)-Asn-Ala-gp41(Ala528-Leu535) (2,016.9Da) was identified as the C-terminal peptide. These data sug-gest that chymotrypsin cleaves initially on the C-terminal sideof the N-helix to give band 1, followed by cleavage at Leu537and/or Leu535 in the polar segment to give band 2; the diffuseband “3” is thus likely to comprise gp41 core domain peptides.We compared the amounts of bands 1 and 2 released by chy-
motrypsin from wild type, L537A-, W666A-, and L537A/W666A-mutated MBP/gp41(528-L-677) to gauge how themutations affected the stability of the polar segment (the polarsegment contains the Leu535 and Leu537 chymotrypsin sitesthat give rise to band 2). The chymotrypsin sensitivity ofW666Awas almost identical to that of wild type, indicating thatthis mutation did not detectably affect the exposure of the pre-ferred chymotrypsin site(s) (Fig. 6,B–D).Whereas treatment ofthe L537A mutant at a chymotrypsin:protein ratio of 1:40released approximately wild type levels of band 1, the mutantresisted further cleavage to band 2 at higher protease:proteinratios (Figs. 6, B–D). This finding is consistent with removal ofa favored chymotrypsin site (i.e. Leu537) and the alternate site,Leu535, remaining protected. Simultaneous L537A/W666Amutations led to increased amounts of band 2 being released atall protease:protein ratios when compared with the other con-structs, and in particular with respect to L537A, which alsolacks the Leu537 chymotrypsin site (Fig. 6, B–D). The N- andC-helical sequences of gp41 were detected by mass spectrome-try at chymotrypsin:MBP/gp41 ratios of both 1:40 and 1:5 forL537A, W666A, and L537A/W666A-mutated MBP/gp41 sug-gesting that the structural integrity of the gp41 core is not sub-stantially affected by mutations within the polar segment orMPR (Fig. 7). Thus, simultaneous L537A/W666A mutationsincreased the sensitivity of the polar segment to chymotrypsincleavage at Leu535, suggesting that Leu537 and Trp666 cooperatein stabilizing the membrane-interactive end of the gp41 trimerof hairpins.
DISCUSSION
In this study, we found that simultaneous Ala mutations inthe polar segment and MPR inhibited Env fusogenicity andabolished viral entry competence. These functional defects cor-related with reduced or absent sCD4-induced dissociation ofgp120 from the gp120-gp41 complex and inhibition of themembrane fusion cascade prior to hemifusion. We also foundthat simultaneous L537A/W666A mutations, but not the indi-vidual L537A andW666Amutations, led to an increase in chy-motrypsin sensitivity of the polar segment within the MBP/gp41(528-L-677) chimera, indicating that the stability of thisregion was affected by the double mutation in a trimer of hair-pins structure. We infer a functional association between
FIGURE 5. Lipid mixing activities of Env mutants. HeLa effector cells weretransfected with wild type (WT) or mutated Env expression vectors and at 5 hpost-transfection, infected with the recombinant vaccinia virus vTF7.3. At18 h post-transfection, effector and U373MG-CCR5 target cells were labeledwith green (DiO) and red (DiI) fluorescent membrane probes, respectively,and co-cultured for 2 h at 37 °C prior to detachment, fixation with 4% formal-dehyde in PBS, and analysis of the green and red fluorescence in flow cytom-etry. The percentage lipid mixing activities were determined following thesubtraction of background dye redistribution between empty vector-trans-fected/vTF7.3-infected HeLa cells labeled with DiO and U373MG-CCR5 tar-gets labeled with DiI. The mean � standard error (error bars) from six inde-pendent experiments are shown. *, p � 0.0001 relative to L537A and W666A(two sample t test assuming unequal variances).
Functions of the HIV gp41 Polar and Membrane Proximal Regions
AUGUST 10, 2007 • VOLUME 282 • NUMBER 32 JOURNAL OF BIOLOGICAL CHEMISTRY 23111
Leu537 in the polar segment and hydrophobic residues of theMPR in distinct stages of the fusion cascade: in the formation ofa stable gp120-gp41 prefusion complex that is responsive toreceptor engagement, and in stabilization of the membrane-interactive end of the late fusion trimer of the hairpins form ofgp41.Alanine replacement of individual MPR residues, Trp666,
Trp672, Phe673, and Ile675, did not detectably affect cell-cellfusion, however, �8–120-fold reductions in viral entry wereobserved. The fusion activities of theMPRmutants were indis-tinguishable over a range of Env concentrations, suggesting thatsmall numbers of Env spikes and limited potential for cooper-ativity to overcome a minor functional defect is unlikely toaccount for the effects of the mutations on viral entry (22, 39,71–75). The MPR has been shown to bind to and destabilize
model membranes with cholesterolmodulating this function (45–48,76), therefore the functional differ-ences observed for viral and cell sur-face Env forms may be attributed tothe relative amounts of cholesterolin these membranes. The virionenvelope is enriched in cholesteroland is less fluid than the plasmamembrane with both attributescontributing to HIV-1 infectivity(77–80). The conformation andtopology of the MPR in the lipid-bound state has been suggested tobe analogous to that ofTrp-contain-ing interfacial helical peptides,which destabilize membranes byintroducing positive curvaturestrain to the outer leaflet of thebilayer (47, 48, 52). In the presenceof cholesterol, such peptides pene-trate the bilayer less deeply andexhibit decreased destabilizingactivity than for membranes thatlack cholesterol (81, 82). Viral entrydefects are associated with Alareplacement of Trp666, Trp672,Phe673, and Ile675 (this study) as wellas Leu669 and Leu679 (83), suggest-ing that a full complement of aro-matic and hydrophobic residues isrequired for efficient MPR-medi-ated destabilization of the viralenvelope. By contrast, at least 3 aro-matic residues must be mutatedsimultaneously for cell-cell fusionfunction to be blocked (40).The L537A polar segment
mutant retained wild type levels ofcell-cell fusion function despite hav-ing a subtle gp120 shedding pheno-type for cellular Env, whereas amore pronounced loss of gp120
from the viral Env complex was linked to an �10-fold decreasein viral entry competence. The L537A/MPR double mutantsexhibited more severe cell-cell fusion defects and lacked viralentry function even though no further increase in gp120 shed-ding was observed. The results of mAb binding studies,together with the cryoelectron tomographic structure of theprefusion trimer solved by Zhu et al. (22) indicate that the gp41ectodomain (excluding the MPR) forms the interior of thegp120-gp41 complex (17–20, 22). The L537A shedding pheno-typemay be due to the introduction of a cavity into this interior,which decreases the stability of the gp120-gp41 complex. Themore pronounced shedding phenotype in virus suggests thatthe structure of the prefusion complex may be subtly alteredfollowing incorporation into virions, perhaps due to interac-tions with a more ordered cholesterol-enriched lipid bilayer
FIGURE 6. Effects of mutations on the chymotrypsin sensitivity of a trimer of hairpins model protein,MBP/gp41(528-L-677). A, schematic representation of MBP/gp41(528-L-677). Maltose-binding protein waslinked through Asn-Ala to the HIV-2ST gp41 N-terminal sequence Ala528-Trp596, linked through Ser-Gly-Gly-Arg-Gly-Gly to the gp41 C-terminal sequence Trp610-Ser677. The upper sequence is that of HIV-1AD8, the lowersequence is that of HIV-2ST. The HXB2R numbering convention is used. The N-helix and C-helix are from thecrystal structure of the HIV-1 gp41 6HB core (31). A monomer is depicted for clarity. B, chymotrypsin cleavageof wild type and mutated MBP/gp41(528-L-677). Purified MBP/gp41(528-L-677) trimers were treated withchymotrypsin for 10 min at 37 °C at the indicated protease:protein ratios and then subjected to SDS-PAGEunder reducing conditions in 12–17% polyacrylamide gradient gels. The protein bands were visualized follow-ing staining with Coomassie Brilliant Blue and scanning in a LI-COR Odyssey infrared imager. C and D, quanti-tation of chymotryptic peptides. The intensities of bands 1 (C) and 2 (D) across the various protease:proteinratios were quantified using Odyssey version 1.2 software and expressed as a fraction of the correspondingmock treated protein. Representative results of three independent experiments is shown. WT, wild type.
Functions of the HIV gp41 Polar and Membrane Proximal Regions
23112 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 32 • AUGUST 10, 2007
and/or with the underlying matrix shell via the gp41 cytoplas-mic tail (77, 84–86). The finding that a L49D mutation inmatrix reduces virion-associated gp120 without affecting thelevels of virion-associated gp41 indicates that gp41-matrixinteractions can affect gp120-gp41 complex stability (87).The simultaneous removal of hydrophobic side chains from
the polar segment andMPR in L537A/W666A, L537A/W672A,L537A/F673A, and L537A/I675A mutants led to markeddecreases in cell-cell fusion and the abolition of viral entry. Thespontaneous gp120 shedding phenotypes of the doublemutantswere similar to that of L537A for both cellular and viralforms of Env. Therefore, the more severe fusion and entrydefects of the double mutants relative to the single mutantscould not be simply attributed to further loss of gp120 from theEnv complex. However, these functional defects correlatedwith decreased responsiveness of themutant gp120-gp41 com-plexes to sCD4, suggesting that the mutants are activated lessefficiently by receptor than is wild type. This idea is consistentwith the observation that L537A/W666A (but not individual
L537A and W666A mutants) was inhibited prior to the hemi-fusion stage of the fusion cascade. Leu537 therefore appears toact together with the MPR in the formation of a fusion compe-tent gp120-gp41 complex. Lorizate and co-workers (88, 89) putforward a model whereby the N-terminal fusion peptide andMPR form a complex within prefusion Env, acting as a kinetictrap to halt fusion. This idea is based on the findings with syn-thetic peptides that fusion peptide-MPR interactions enhancemAb 2F5 binding (mAb 2F5 binds to the prefusion Env complex),and inhibit the membrane destabilizing properties of the compo-nent sequences. The synthetic fusion peptide-MPR complex islargely non-helical with a predominance of �-turns, consistentwith the structure of the Glu657-Asp664 sequence when bound to2F5 (51). The fusion peptide-MPR prefusion claspmodel theoret-ically brings the polar segment into proximity with theMPR. Theloss of hydrophobicity in these 2 sequences due to L537A/MPRmutationsmay affect the fusion peptide-MPR clasp and its releasefollowing gp120-receptor interactions.Previously, we reported that N- and C-terminal extension of
FIGURE 7. Chymotryptic peptides of wild type and L537A-, W666A- and L537A/W66A-mutated MBP/gp41(528-L-677). Wild type (WT) and mutatedMBP/gp41(528-L-677) derived from HIV-2ST was digested with chymotrypsin at protease:protein ratios (w/w) of 1:40 and 1:5 at 37 °C for 10 min. The digestedproteins were mass analyzed by linear MALDI-MS using 2,5-dihydroxybenzoic acid:5-methylsalicylic acid (9:1) as a matrix. The identity of gp41 peptides wereinferred from the observed molecular masses. Black, N-helix; gray, C-helix; dashed line, interhelical region.
Functions of the HIV gp41 Polar and Membrane Proximal Regions
AUGUST 10, 2007 • VOLUME 282 • NUMBER 32 JOURNAL OF BIOLOGICAL CHEMISTRY 23113
the 6-helix bundle core to include the polar segment residues,Ala528-Leu535, and MPR residues, Phe669-Ser677, respectively,conferred thermal stability to a MBP/gp41 trimer of hairpinsmodel protein (53). In the current study, we found that theL537A/W666A double mutation increased chymotrypsin sen-sitivity of the polar segment. The L537Amutation alone inhib-ited proteolysis of the polar segment, indicating that theupstream Leu535 site (Ile535 in AD8) is not readily accessed bychymotrypsin. By contrast, mutating L537A together withW666A led to enhanced proteolysis at Leu535. Proteases at lim-iting concentrations preferentially cleave protein substrates inmobile loops rather than within folded domains, suggestingthat L537A/W666A disrupts the structural order at the end ofthe hairpin, enabling the protease to access Leu535. These
observations indicate a structural link between Leu537 andTrp666 in stabilizing themembrane proximal end of the hairpin.However, because the individual L537A andW666Amutationsdid not increase the protease sensitivity of the polar segment,they are either unlikely to interact directly, or the effects of themutations are mitigated by other contacts in this region.A potential mechanism whereby these terminal sequences
could confer stability to gp41 is illustrated by HA2 of influenzavirus. In this case, low pH activates a loop-to-helix transitionthat extends the N terminus of the central coiled coil by some100Å that is terminated by anN-cap.The surface grooves of thecoiled coil terminus close to the N-cap act as a hydrophobicdocking site for the C-terminal MPR (90). Mutations designedto destabilize this terminal interaction have been reported to
FIGURE 8. Model of the HIV-1 fusion cascade. The metastable state of the gp120-gp41 complex is released by binding of gp120 to CD4 and chemokinereceptor (CKR). The majority of gp41 is occluded by the gp120 trimer. The MPR, which encompasses the mAb 2F5 and 4E10 epitopes, is exposed and adjacentto the viral envelope (A). Formation of the prehairpin intermediate of gp41 (B) precedes refolding of the core into a 6-helix bundle and hemifusion (C and D).Coil to helix transition in the polar segment creates a hydrophobic binding site for the MPR (E) completing the trimer of hairpins fold (F). Fusion peptide, orangecylinder; polar segment, purple tube (random coil)/purple cylinder (�-helix); coiled coil, pink cylinder; C-helix, light blue tube and cylinder; MPR, dark blue cylinder;TMD, hatched oblong. The cytoplasmic domain is not shown. Enlarged inset, theoretical model of the 6-helix bundle core (33) with an N-terminal helicalextension (purple ribbon) to Ser528 at the fusion peptide-polar segment boundary (94). The extension creates a hydrophobic binding site (Met530, Ile535, Leu537)for the MPR (the 2F5- and 4E10-bound structures are shown (51, 67)). The model was prepared with Swiss PDB Viewer (version 3.7) and POV Ray (version 3.5.1).
Functions of the HIV gp41 Polar and Membrane Proximal Regions
23114 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 32 • AUGUST 10, 2007
block hemifusion and/or pore opening, depending on themutation and assay system used (91, 92). Random coil and�-helical conformations have been observed for the polar seg-ment in synthetic peptide models of the N-terminal portion ofthe gp41 prehairpin intermediate (24, 26, 93) with propagationof helical structure beyond the core to Ser528 at the fusion pep-tide-polar segment boundary improving the thermal stability ofthe gp41 core domain (94).Helical extension of the coiled coilNterminus indicates that Met530, Ile535, and Leu537 form thegrooves of the coiled coil (Met530 and Leu537 occupy g positionsof the hydrophobic heptad repeat, whereas Ile535 occupies e)potentially creating a hydrophobic docking site for the MPRwhen brought into proximity by 6HB formation. Such interac-tions would “zip up” the membrane-interactive end of the gp41hairpin, aiding in the apposition of virus and target cell mem-branes and contributing additional free energy to membranefusion. The idea that the MPR acquires structural order in thecontext of the trimer of hairpins is supported by our earlieranalysis of a MBP/gp41 chimera comprising the entire gp41ectodomain where the Phe669-Ser677 MPR sequence resistedchymotrypsin despite the presence of 7 potential protease sites(53).These scenarios are illustrated in Fig. 8, which summarizes
the HIV-1 fusion cascade. The mAb 2F5 and 4E10 epitopes areavailable for antibody binding in prefusion and prehairpinintermediate forms of gp41 (Fig. 8, A and B, respectively) butare lost or occluded by 6-helix bundle formation (21, 49–51)(Fig. 8,D–F). According to themodel put forward by Lorizate etal. (88, 89), the fusion peptide and MPR interact to form a pre-fusion clasp that is released by gp120-receptor interactions (Fig.8A). Membrane-inserted fusion peptides are depicted as �-hel-ices with the adjacent polar segment as a random coil, based onNMR and Fourier transform infrared spectroscopy findings(24, 25, 93, 95) (Fig. 8, B–D). However, it should be noted that�-strand conformations have also been documented for thissequence (26). Formation of the 6-helix bundle core leads tohemifusion (37) (Fig. 8D), and a coil-to-helix transition in thepolar segment (94) (Fig. 8,D andE) creates a hydrophobic bind-ing site (Met530, Ile535, and Leu537) for the MPR (Fig. 8E, inset),completing the trimer of hairpins fold (Fig. 8F). A possiblemechanism for mAb 2F5 neutralization, which binds to theMPR in the prefusion Env structure, is the blockade of terminalinteractions between the polar segment and MPR that arerequired to complete the trimer of hairpins fold. In summary,our data indicate that the N-terminal polar segment and C-ter-minal MPR of gp41 are functionally linked in distinct phases ofthe fusion cascade, acting synergistically to form a fusion-com-petent prefusion gp120-gp41 complex and in stabilizing themembrane-interactive end of the trimer of hairpins.
REFERENCES1. Basmaciogullari, S., Babcock, G. J., Van Ryk, D., Wojtowicz, W., and So-
droski, J. (2002) J. Virol. 76, 10791–108002. Chen, B., Vogan, E. M., Gong, H., Skehel, J. J., Wiley, D. C., and Harrison,
S. C. (2005) Nature 433, 834–8413. Kwong, P. D., Wyatt, R., Robinson, J., Sweet, R. W., Sodroski, J., and Hen-
drickson, W. A. (1998) Nature 393, 648–6594. Rizzuto, C. D., Wyatt, R., Hernandez-Ramos, N., Sun, Y., Kwong, P. D.,
Hendrickson, W. A., and Sodroski, J. (1998) Science 280, 1949–1953
5. Hoffman,N.G., Seillier-Moiseiwitsch, F., Ahn, J.,Walker, J.M., and Swan-strom, R. (2002) J. Virol. 76, 3852–3864
6. Pastore, C., Nedellec, R., Ramos, A., Pontow, S., Ratner, L., and Mosier,D. E. (2006) J. Virol. 80, 750–758
7. Rosen, O., Sharon, M., Quadt-Akabayov, S. R., and Anglister, J. (2006)Proc. Natl. Acad. Sci. U. S. A. 103, 13950–13955
8. Pierson, T. C., and Doms, R. W. (2003) Curr. Top. Microbiol. Immunol.281, 1–27
9. Salzwedel, K., Smith, E. D., Dey, B., and Berger, E. A. (2000) J. Virol. 74,326–333
10. Colman, P. M., and Lawrence, M. C. (2003) Nat. Rev. Mol. Cell. Biol. 4,309–319
11. Eckert, D. M., and Kim, P. S. (2001) Annu. Rev. Biochem. 70, 777–81012. Gallaher,W. R., Ball, J.M., Garry, R. F., Griffin,M. C., andMontelaro, R. C.
(1989) AIDS Res. Hum. Retroviruses 5, 431–44013. Kielian, M., and Rey, F. A. (2006) Nat. Rev. Microbiol. 4, 67–7614. Lamb, R. A., Paterson, R. G., and Jardetzky, T. S. (2006) Virology 344,
30–3715. Skehel, J. J., and Wiley, D. C. (2000) Annu. Rev. Biochem. 69, 531–56916. Xu, Y., Lou, Z., Liu, Y., Pang, H., Tien, P., Gao, G. F., and Rao, Z. (2004)
J. Biol. Chem. 279, 49414–4941917. Finnegan, C. M., Berg, W., Lewis, G. K., and DeVico, A. L. (2002) J. Virol.
76, 12123–1213418. Furuta, R. A., Wild, C. T., Weng, Y., and Weiss, C. D. (1998) Nat. Struct.
C. D. (2003) J. Virol. 77, 1666–167120. Sattentau, Q. J., Zolla-Pazner, S., and Poignard, P. (1995) Virology 206,
713–71721. Dimitrov, A. S., Jacobs, A., Finnegan, C. M., Stiegler, G., Katinger, H., and
Blumenthal, R. (2007) Biochemistry 46, 1398–140122. Zhu, P., Liu, J., Bess, J., Jr., Chertova, E., Lifson, J. D., Grise, H., Ofek, G. A.,
Taylor, K. A., and Roux, K. H. (2006) Nature 441, 847–85223. Zanetti, G., Briggs, J. A., Grunewald, K., Sattentau, Q. J., and Fuller, S. D.
(2006) PLoS Pathog. 2, e8324. Chang, D. K., Cheng, S. F., and Trivedi, V. D. (1999) J. Biol. Chem. 274,
5299–530925. Gordon, L. M., Mobley, P.W., Pilpa, R., Sherman,M. A., andWaring, A. J.
(2002) Biochim. Biophys. Acta 1559, 96–12026. Sackett, K., and Shai, Y. (2005) J. Mol. Biol. 350, 790–80527. Root, M. J., Kay, M. S., and Kim, P. S. (2001) Science 291, 884–88828. Wild, C., Oas, T., McDanal, C., Bolognesi, D., and Matthews, T. (1992)
Proc. Natl. Acad. Sci. U. S. A. 89, 10537–1054129. Wild, C. T., Shugars, D. C., Greenwell, T. K., McDanal, C. B., and Mat-
thews, T. J. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 9770–977430. Caffrey,M., Cai,M., Kaufman, J., Stahl, S. J.,Wingfield, P. T., Covell, D. G.,
Gronenborn, A. M., and Clore, G. M. (1998) EMBO J. 17, 4572–458431. Chan, D. C., Fass, D., Berger, J. M., and Kim, P. S. (1997) Cell 89, 263–27332. Tan, K., Liu, J., Wang, J., Shen, S., and Lu, M. (1997) Proc. Natl. Acad. Sci.
U. S. A. 94, 12303–1230833. Weissenhorn,W., Dessen, A., Harrison, S. C., Skehel, J. J., andWiley, D. C.
(1997) Nature 387, 426–43034. Yang, Z. N., Mueser, T. C., Kaufman, J., Stahl, S. J., Wingfield, P. T., and
Hyde, C. C. (1999) J. Struct. Biol. 126, 131–14435. Blumenthal, R., Sarkar, D. P., Durell, S., Howard, D. E., and Morris, S. J.
(1996) J. Cell Biol. 135, 63–7136. Markosyan, R. M., Cohen, F. S., and Melikyan, G. B. (2005)Mol. Biol. Cell
16, 5502–551337. Melikyan, G. B., Markosyan, R. M., Hemmati, H., Delmedico, M. K., Lam-
bert, D. M., and Cohen, F. S. (2000) J. Cell Biol. 151, 413–42338. Melikyan, G. B., Barnard, R. J., Abrahamyan, L. G., Mothes, W., and
Young, J. A. (2005) Proc. Natl. Acad. Sci. U. S. A. 102, 8728–873339. Munoz-Barroso, I., Durell, S., Sakaguchi, K., Appella, E., and Blumenthal,
R. (1998) J. Cell Biol. 140, 315–32340. Salzwedel, K., West, J. T., and Hunter, E. (1999) J. Virol. 73, 2469–248041. Binley, J. M., Wrin, T., Korber, B., Zwick, M. B., Wang, M., Chappey, C.,
Stiegler, G., Kunert, R., Zolla-Pazner, S., Katinger, H., Petropoulos, C. J.,and Burton, D. R. (2004) J. Virol. 78, 13232–13252
Functions of the HIV gp41 Polar and Membrane Proximal Regions
AUGUST 10, 2007 • VOLUME 282 • NUMBER 32 JOURNAL OF BIOLOGICAL CHEMISTRY 23115
42. Muster, T., Steindl, F., Purtscher, M., Trkola, A., Klima, A., Himmler, G.,Ruker, F., and Katinger, H. (1993) J. Virol. 67, 6642–6647
43. Zwick, M. B., Labrijn, A. F., Wang, M., Spenlehauer, C., Saphire, E. O.,Binley, J. M., Moore, J. P., Stiegler, G., Katinger, H., Burton, D. R., andParren, P. W. (2001) J. Virol. 75, 10892–10905
44. Munoz-Barroso, I., Salzwedel, K., Hunter, E., and Blumenthal, R. (1999)J. Virol. 73, 6089–6092
45. Saez-Cirion, A., Nir, S., Lorizate,M., Agirre, A., Cruz, A., Perez-Gil, J., andNieva, J. L. (2002) J. Biol. Chem. 277, 21776–21785
46. Shnaper, S., Sackett, K., Gallo, S. A., Blumenthal, R., and Shai, Y. (2004)J. Biol. Chem. 279, 18526–18534
47. Suarez, T., Nir, S., Goni, F.M., Saez-Cirion, A., andNieva, J. L. (2000) FEBSLett. 477, 145–149
48. Suarez, T., Gallaher, W. R., Agirre, A., Goni, F. M., and Nieva, J. L. (2000)J. Virol. 74, 8038–8047
49. de Rosny, E., Vassell, R., Jiang, S., Kunert, R., and Weiss, C. D. (2004)J. Virol. 78, 2627–2631
50. Gorny, M. K., and Zolla-Pazner, S. (2000) J. Virol. 74, 6186–619251. Ofek, G., Tang,M., Sambor, A., Katinger, H., Mascola, J. R.,Wyatt, R., and
Kwong, P. D. (2004) J. Virol. 78, 10724–1073752. Schibli, D. J., Montelaro, R. C., and Vogel, H. J. (2001) Biochemistry 40,
9570–957853. Lay, C. S., Wilson, K. A., Kobe, B., Kemp, B. E., Drummer, H. E., and
Poumbourios, P. (2004) FEBS Lett. 567, 183–18854. Poumbourios, P., Maerz, A. L., and Drummer, H. E. (2003) J. Biol. Chem.
278, 42149–4216055. Moss, B., Elroy-Stein, O., Mizukami, T., Alexander, W. A., and Fuerst,
T. R. (1990) Nature 348, 91–9256. Poumbourios, P., Ahmar, W. E., McPhee, D. A., and Kemp, B. E. (1995)
J. Virol. 69, 1209–121857. Takikawa, S., Ishii, K., Aizaki, H., Suzuki, T., Asakura, H., Matsuura, Y.,
and Miyamura, T. (2000) J. Virol. 74, 5066–507458. Abacioglu, Y. H., Fouts, T. R., Laman, J. D., Claassen, E., Pincus, S. H.,
Moore, J. P., Roby, C. A., Kamin-Lewis, R., and Lewis, G. K. (1994) AIDSRes. Hum. Retroviruses 10, 371–381
59. Xu, J.-Y., Gorny, M. K., Palker, T., Karwowska, S., and Zolla-Pazner, S.(1991) J. Virol. 65, 4832–4838
60. Myers, R., Meiller, T., Falkler, W., Jr., Patel, J., and Joseph, J. (1993) inAbstracts of the 93rd General Meeting of the American Society for Micro-biology, American Society for Microbiology, Washington, D. C.
61. Haigwood, N. L., Shuster, J. R., Moore, G. K., Lee, H., Skiles, P. V., Higgins,K. W., Barr, P. J., George-Nascimento, C., and Steimer, K. S. (1990) AIDSRes. Hum. Retroviruses 6, 855–869
62. Maerz, A. L., Drummer, H. E., Wilson, K. A., and Poumbourios, P. (2001)J. Virol. 75, 6635–6644
63. Deng, H., Liu, R., Ellmeier, W., Choe, S., Unutmaz, D., Burkhart, M., DiMarzio, P., Marmon, S., Sutton, R. E., Hill, C.M., Davis, C. B., Peiper, S. C.,Schall, T. J., Littman, D. R., and Landau, N. R. (1996)Nature 381, 661–666
64. Bar, S., and Alizon, M. (2004) J. Virol. 78, 811–82065. Connor, R. I., Chen, B. K., Choe, S., and Landau, N. R. (1995)Virology 206,
935–94466. Bjorndal, A., Deng, H., Jansson,M., Fiore, J. R., Colognesi, C., Karlsson, A.,
Albert, J., Scarlatti, G., Littman, D. R., and Fenyo, E. M. (1997) J. Virol. 71,7478–7487
67. Cardoso, R. M., Zwick, M. B., Stanfield, R. L., Kunert, R., Binley, J. M.,Katinger,H., Burton,D. R., andWilson, I. A. (2005) Immunity22, 163–173
68. Poumbourios, P., Wilson, K. A., Center, R. J., El Ahmar, W., and Kemp,B. E. (1997) J. Virol. 71, 2041–2049
69. Sattentau, Q. J., Moore, J. P., Vignaux, F., Traincard, F., and Poignard, P.(1993) J. Virol. 67, 7383–7393
70. Willey, R. L., Martin, M. A., and Peden, K. W. (1994) J. Virol. 68,1029–1039
71. Chertova, E., Bess, J. W., Jr., Crise, B. J., Sowder, I. R., Schaden, T. M.,Hilburn, J.M., Hoxie, J. A., Benveniste, R. E., Lifson, J. D., Henderson, L. E.,and Arthur, L. O. (2002) J. Virol. 76, 5315–5325
72. Kuhmann, S. E., Platt, E. J., Kozak, S. L., and Kabat, D. (2000) J. Virol. 74,7005–7015
73. Yang, X., Kurteva, S., Ren, X., Lee, S., and Sodroski, J. (2005) J. Virol. 79,12132–12147
74. Yang, X., Kurteva, S., Ren, X., Lee, S., and Sodroski, J. (2006) J. Virol. 80,4388–4395
75. Zhu, P., Chertova, E., Bess, J., Jr., Lifson, J. D., Arthur, L. O., Liu, J., Taylor,K. A., and Roux, K. H. (2003) Proc. Natl. Acad. Sci. U. S. A. 100,15812–15817
76. Lorizate, M., Cruz, A., Huarte, N., Kunert, R., Perez-Gil, J., and Nieva, J. L.(2006) J. Biol. Chem. 281, 39598–39606
77. Aloia, R. C., Tian, H., and Jensen, F. C. (1993) Proc. Natl. Acad. Sci. U. S. A.90, 5181–5185
78. Campbell, S. M., Crowe, S. M., and Mak, J. (2002) AIDS 16, 2253–226179. Guyader, M., Kiyokawa, E., Abrami, L., Turelli, P., and Trono, D. (2002)
J. Virol. 76, 10356–1036480. Liao, Z., Graham, D. R., and Hildreth, J. E. (2003) AIDS Res. Hum. Retro-
viruses 19, 675–68781. Egashira, M., Gorbenko, G., Tanaka, M., Saito, H., Molotkovsky, J., Na-
kano, M., and Handa, T. (2002) Biochemistry 41, 4165–417282. Matsuzaki, K., Sugishita, K., Ishibe, N., Ueha, M., Nakata, S., Miyajima, K.,
and Epand, R. M. (1998) Biochemistry 37, 11856–1186383. Zwick, M. B., Jensen, R., Church, S., Wang, M., Stiegler, G., Kunert, R.,
Katinger, H., Burton, D. R., Parker, C. E., Deterding, L. J., Hager-Braun, C.,Binley, J. M., Schulke, N., Moore, J. P., and Tomer, K. B. (2005) J. Virol. 79,1252–1261
84. Bhattacharya, J., Repik, A., and Clapham, P. R. (2006) J. Virol. 80,5292–5300
85. Murakami, T., and Freed, E. O. (2000) J. Virol. 74, 3548–355486. Wyma, D. J., Kotov, A., and Aiken, C. (2000) J. Virol. 74, 9381–938787. Davis, M. R., Jiang, J., Zhou, J., Freed, E. O., and Aiken, C. (2006) J. Virol.
80, 2405–241788. Lorizate, M., Gomara, M. J., de la Torre, B. G., Andreu, D., and Nieva, J. L.
(2006) J. Mol. Biol. 360, 45–5589. Lorizate, M., de la Arada, I., Huarte, N., Sanchez-Martinez, S., de la Torre,
B. G., Andreu, D., Arrondo, J. L., Nieva, J. L., Katinger, H., Kunert, R., Cruz,A., Perez-Gil, J., Gomara, M. J., Mingarro, I., Perez-Paya, E., Hermann, K.,Basanez, G., Saez-Cirion, A., Iloro, I., Melikyan, G., Nir, S., and Agirre, A.(2006) Biochemistry 45, 14337–14346
90. Chen, J., Skehel, J. J., andWiley, D. C. (1999) Proc. Natl. Acad. Sci. U. S. A.96, 8967–8972
91. Borrego-Diaz, E., Peeples, M. E., Markosyan, R. M., Melikyan, G. B., andCohen, F. S. (2003) Virology 316, 234–244
92. Park, H. E., Gruenke, J. A., and White, J. M. (2003) Nat. Struct. Biol. 10,1048–1053
93. Jaroniec, C. P., Kaufman, J. D., Stahl, S. J., Viard,M., Blumenthal, R.,Wing-field, P. T., and Bax, A. (2005) Biochemistry 44, 16167–16180
94. Sackett, K., Wexler-Cohen, Y., and Shai, Y. (2006) J. Biol. Chem. 281,21755–21762
95. Chang, D. K., Cheng, S. F., and Chien,W. J. (1997) J. Virol. 71, 6593–660296. Earl, P. L., Moss, B., and Doms, R. W. (1991) J. Virol. 65, 2047–2055
Functions of the HIV gp41 Polar and Membrane Proximal Regions
23116 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 32 • AUGUST 10, 2007
