JBC Anatrace, Inc.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M304950200 on February 23, 2004

J. Biol. Chem., Vol. 279, Issue 18, 18526-18534, April 30, 2004
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
279/18/18526    most recent
M304950200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Shnaper, S.
Right arrow Articles by Shai, Y.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Shnaper, S.
Right arrow Articles by Shai, Y.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

The C- and the N-terminal Regions of Glycoprotein 41 Ectodomain Fuse Membranes Enriched and Not Enriched with Cholesterol, Respectively*

Sophie Shnaper{ddagger}, Kelly Sackett{ddagger}, Stephen A. Gallo§, Robert Blumenthal§, and Yechiel Shai{ddagger}

From the {ddagger}Department of Biological Chemistry, The Weizmann Institute of Science, Rehovot 76100, Israel and §Laboratory of Experimental and Computational Biology, National Cancer Institute, National Institutes of Health, Frederick, Maryland 21702-1201

Received for publication, May 12, 2003 , and in revised form, February 22, 2004.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS AND DISCUSSION
 REFERENCES
 
To infect target cells, HIV-1 employs a virally encoded transmembrane protein (gp41) to fuse its viral envelope with the target cell plasma membrane. We describe the gp41 ectodomain as comprised of N- and C-terminal subdomains, each containing a heptad repeat as well as a fusogenic region, whose organization is mirrored by the intervening loop region. Recent evidence indicates that the gp41 directed fusion reaction proceeds to initial pore formation prior to gp41 folding into its low energy hairpin conformation. This implies that exposed regions of the gp41 ectodomain are responsible for the bulk of the fusion work, probably through direct protein-membrane interactions. Prevalent fusion models contend that the gp41 ectodomain initially interacts with the target cell surface through its highly hydrophobic N terminus, which is believed to insert into the target membrane, thereby linking the virus to the target cell. This arrangement allows the N-terminal subdomain to interact with the target cell surface, whereas the C-terminal subdomain remains proximal to the virion, allowing interaction with the viral envelope. The composition of the viral envelope and the target cell surface differ due to the virus budding from raft microdomains. We show here that constructs corresponding to the C-terminal subdomain specifically destabilize ordered and cholesterol rich membranes (33 molar %), whereas the N-terminal subdomain is more effective in fusing both unordered cholesterol-free membranes and those containing lower amounts of cholesterol (10 molar %). Moreover we show that, in the context of the C-terminal subdomain, the heptad repeat contributes helical structure, which may describe the enhanced inhibitory effect of the C-terminal subdomain relative to the C-terminal heptad repeat (C34) alone. Our results are discussed in light of recent findings that showcase the role of exposed gp41 regions in effecting membrane fusion.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS AND DISCUSSION
 REFERENCES
 
Infection of human cells by HIV-11 requires fusion between the viral envelope and plasma membrane whereby the virion transfers its nucleocapsid into the cytoplasm of the host cell. Virally encoded envelope glycoproteins play a major role in driving the membrane fusion reaction by helping to overcome the free energy barriers associated with this process (1). The envelope glycoprotein complex of HIV-1 is initially synthesized as a gp160 precursor molecule that is subsequently processed by a cellular protease into a mature surface, gp120 subunit noncovalently attached to a transmembrane gp41 subunit (2–4). The subunits are assembled as trimers (5), which are directed to the plasma membrane and then incorporated into budding virions. Recent studies (6, 7) show preferential budding of the virions through the so-called raft microdomains of the plasma membrane, which provides an explanation for the unusually high content of cholesterol and sphingomyelin in the viral envelope (8). Binding of gp120 with CD4 and a co-receptor (reviewed in Ref. 9) cooperatively initiates a series of conformational changes in the gp120·gp41 complex (10), which eventually promotes the actual merging of the viral and target cell membranes via structural rearrangements of the gp41 subunits.

Two main functional regions separated by a loop are identified within the extraviral 175-residue ectodomain of the gp41 subunit: N-terminal and C-terminal subdomains (see Fig. 1). The N-terminal subdomain is made of the N-terminal hydrophobic fusion peptide (FP) region and an adjacent {alpha}-helical leucine/isoleucine zipper termed the N-terminal heptad repeat (NHR). The FP is believed to play a pivotal role in the fusion event by inserting into the target membrane and directly effecting the fusion of apposing bilayers (reviewed in Refs. 11 and 12), although the mode of action remains illusory. The adjacent NHR is believed to be involved in structural re-arrangements that draw together opposing viral and target cell membranes (13). Recent studies show that the NHR acts synergistically with the FP to fuse vesicles, implying that these regions interact directly with membranes (14). We describe the C-terminal subdomain as reflecting its N-terminal counterpart. Specifically, it contains a heptad repeat termed the C-terminal heptad repeat (CHR) followed by a hydrophobic region called the tryptophan rich pre-transmembrane region, which was proposed to serve as an internal fusion peptide (IFP) (15). Peptides corresponding to the IFP are effective in inducing lipid mixing of model vesicles (15–17). The CHR, similar to the NHR, is believed to be involved in structural re-arrangements of gp41 that results in viral-target membrane apposition (13).



View larger version (28K):
[in this window]
[in a new window]
  		FIG. 1.
Scheme of the HIV-1 gp41 primary structure. Primary functional regions are dotted and outlined with additionally characterized regions specified above by a brace. Above and in the brackets are peptide segments that have been extensively studied and will be referred to throughout the text. Below are the peptides represented in this study, which are derived from the indicated N-and C-terminal subdomains. The residues are numbered according to their position in gp160. The C34 sequence is highlighted within C56.

 
Current models describe the fusion process through a progression of conformational changes in the gp41 ectodomain. In the native state, gp41 is held in a metastable state by interaction with gp120 (13, 18). Binding of gp120 to receptors releases gp41 from metastable constraints, allowing it to transform (19) to an extended conformation termed the pre-hairpin intermediate (PHI). Although not observed, the PHI is characterized by exposed NHR and CHR regions (20–22), and exists transiently (20, 22), and the N-terminal fusion peptide region is believed to be inserted into the target membrane (11) thereby bridging the gap between the two membranes. Furthermore, the PHI is believed to be the target of peptide inhibitors derived from the gp41 ectodomain that presumably act by binding to regions in the PHI and competitively inhibit further conformational changes (21, 23–26). One example is C34, which is derived from the CHR of gp41 and inhibits cell-cell fusion at nanomolar concentrations by binding the NHR coiled-coil (25). Folding of the PHI into a stable hairpin conformation is believed to coincide with fusion. Contrary to the widely accepted belief that hairpin folding drives the fusion reaction by initiating pore formation, it was recently proposed that fusion pore formation occurs prior to hairpin formation and that hairpins are required for pore stabilization (27). The hairpin conformation was contrived based on studies of soluble fragments of the ectodomain, which revealed a thermostable core made of a parallel trimeric coiled-coil of NHR peptides with {alpha}-helical CHR peptides packed in an antiparallel manner into the hydrophobic grooves of the coiled-coil (28–30). It is believed that folding to a hairpin helps draw the apposing membranes together and aligns the target membrane inserted FPs with the transmembrane domains in the viral envelope (13, 18). This model implies direct interaction of the two membrane proximal regions with their respective membrane targets. Specifically, the FP region interacts with the target membrane, whereas the pre-transmembrane region interacts with the viral membrane. The high cholesterol content of the viral membrane was recently shown to be indispensable for HIV-1 entry at a post receptors-binding stage, implying that the highly ordered viral membrane is required for fusion (31, 32). On the other hand, Blumenthal and co-workers (33) have shown that depletion of cholesterol from receptors-bearing target cell membranes affects the lateral association of fusion protein-receptors complexes believed to be important for activating fusion but does not inhibit the fusion event itself given adequate expression of receptors. Together, these findings delineate the basic membrane demands that the HIV-1 fusion reaction must meet: a highly ordered, cholesterol-rich viral envelope, and a less ordered target membrane that does not require cholesterol. Recent findings, identifying exposed regions of the gp41 ectodomain as being responsible for the process of membrane merger up to and including initial pore formation (27) prompt specific evaluation of how the ectodomain interacts with and perturbs membranes.

This study addresses the specificity to which the C- and N-terminal subdomains destabilize membranes either enriched in or devoid of cholesterol. We mimic the structural organization of the PHI through two separate fragments: the C-terminal subdomain (C56) is composed of both the CHR and the IFP, whereas the N-terminal subdomain (N70) includes both the NHR and FP. Our results suggest that C- and N-terminal subdomains are specific in their mode of action with the N-subdomain as a potent fusogen toward cholesterol-free membranes and the C-subdomain as superior in perturbing cholesterol-rich membranes. For the membrane bound C56, the CHR contributes {alpha}-helical structure, which may explain both the enhancement in lipid mixing of C56 relative to C20 (IFP) as well as the enhanced inhibitory effect of C56 compared with C34.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS AND DISCUSSION
 REFERENCES
 
Materials—Rink-amide methylbenzhydrylamine resin and 9-Fmoc amino acids were purchased from Calbiochem-Nova-Biochem AG (Laufelfingen, Switzerland). Other reagents used for peptide synthesis included trifluoroacetic acid (Sigma), piperidine (Merck), N,N-diisopropylethylamine (Sigma), N-hydroxybenzotriazole hydrate (Aldrich), 2-(1H-benzotriazole-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate, and dimethylformamide (peptide synthesis grade, BioLab) were purchased from Sigma. Egg phosphatidylcholine (PC), phosphatidylserine, egg sphingomyelin (Spm), cholesterol (extra pure), and lysophosphatidylcholine (LPC) were purchased from Sigma as well. 5- and 6-([(4-chloromethyl)benzoyl]-amino)tetramethylrhodamine (CMTMR) and calcein-AM were purchased from Molecular Probes (Eugene, OR). All of the additional reagents were of analytical grade. Buffers were prepared using double glass-distilled water.

Peptide Synthesis—The peptides were synthesized by the Fmoc solid phase method on Rink-amide methylbenzhydrylamine resin using an ABI 433A automatic peptide synthesizer and then cleaved from the resin using trifluoroacetic acid:double-distilled water:TES at 18.5:1:0.5 (v/v). N70 was prepared as described previously (14). The peptides were purified by reverse-phase high pressure liquid chromatography on C4 reverse-phase Vydac semi-preparative and analytic columns to >98% homogeneity. The purified peptides were homogenous (>98%) by analytical high pressure liquid chromatography. The peptides were further subjected to amino acid analysis and to Platform LCZ electrospray mass spectroscopy to confirm their composition and molecular weight.

Preparation of Large Unilamellar Vesicles (LUV)—The lipids were dissolved in a 2:1 (v/v) mixture of chloroform:methanol and then dried under a stream of nitrogen while rotating, thereby depositing a thin film of lipids on the wall of a glass vial. Excess solvent was removed by overnight lyophilization, and the lipid films were stored under argon to prevent oxidation. Dry lipid films were suspended in PBS buffer by vortexing to produce large multilamellar vesicles. The lipids suspension was further processed with five cycles of freezing and thawing followed by extrusion through polycarbonate membranes initially with 1-µm diameter pores and then with 0.2-µm diameter pores to generate LUV. LUV were used within 24 h following preparation.

Peptide-induced Lipid Mixing—Lipid mixing of LUV was measured using a fluorescence-probe dilution assay (34). LUV containing 0.6 molar % each of NBD-PE as the energy donor and Rho-PE as the energy acceptor were prepared in PBS as described above. A 1:4 mixture of labeled and unlabeled vesicles (110 µM total phospholipid concentration) was suspended in 400 µl of PBS, and a small volume of peptide from a stock solution was added. The increase in NBD fluorescence at 530 nm (8-nm slit) was monitored with the excitation set at 467 nm (8-nm slit). The fluorescence intensity before the addition of the peptide was referred to as 0% lipid mixing, and the fluorescence intensity upon addition of reduced Triton X-100 (0.05% (v/v)) was referred to as 100% lipid mixing. All of the fluorescence measurements were done on an SLM-AMINCO Bowman series 2 luminescence spectrometer.

ATR-FTIR Measurements—Spectra were obtained with a Bruker Equinox 55 FTIR spectrometer equipped with a deuterated triglyceride sulfate detector that was coupled to an ATR device. For each spectrum, 150 scans were collected with a resolution of 4 cm-1. Samples were prepared as described previously (35). A mixture of lipids (0.5 mg) alone or with peptide was deposited on a ZnSe horizontal ATR prism (80 x 7 mm). Prior to preparing the samples, the trifluoroacetate (CF3COO-) counterions, which strongly associate with the peptide, were replaced with chloride ions by washing in 0.1 M HCl and lyophilization. This eliminated the strong C = O stretching absorption band near 1673 cm-1 (36). Peptides were dissolved in MeOH, and lipids were dissolved in a 1:2 MeOH/CH2Cl2 mixture. Lipid-peptide mixtures or lipids alone were spread with a Teflon bar on the ZnSe prism and then dried under vacuum for 15 min. Polarized and non-polarized spectra were recorded. For analysis of the peptide amide I contribution alone, the non-polarized spectra of the pure phospholipids were subtracted from that of the peptide + phospholipid sample. For analysis of phospholipid acyl chain order, symmetric (~2853 cm-1) and antisymmetric (~2922 cm-1) stretching vibrations were analyzed of spectra following polarization (both parallel and perpendicular). The ratio between parallel and perpendicular polarized absorbance maxima for these two stretching vibrations was compared for samples of phospholipid alone and phospholipid + peptide. The background for each spectrum was a clean ZnSe prism. The multilayers formed, composed of PC:Spm:chol at 1:1:1 molar ratio, produce significant background signal in the amide I region due to absorbance from the amide bond in sphingomyelin. Procuring an accurate peptide amide I signal necessitated the use of relatively high peptide:lipid molar ratios (1:50). For measurements conducted of peptides in PC multilayers, a peptide:lipid molar ratio of 1:100 was used. Deuteration of the sample was achieved by introducing an excess of deuterium oxide (2H2O) into a chamber placed on top the ZnSe prism in the ATR casting and incubating for 5 min before acquisition of spectra. The H/D exchange was considered complete due to a complete shift of the amide II band. Any contribution of 2H2O vapor to the absorbance spectra near the amide I peak region was eliminated by subtraction of the spectra of pure lipids equilibrated with 2H2O under the same conditions.

ATR-FTIR Data Analysis—Prior to curve fitting, a straight base line passing through the ordinates at 1700 and 1600 cm-1 was subtracted. To resolve overlapping bands, the spectra were processed using PEAK-FITTM (Jandel Scientific, San Rafael, CA) software. Second- and fourth-derivative spectra were calculated to identify the positions of the component bands in the spectra. These wave numbers were used as initial parameters for curve fitting with Gaussian component peaks. Positions, bandwidths, and amplitudes of the peaks were varied until: (i) the resulting bands shifted by no more than 2 cm-1 from the initial parameters; (ii) all of the peaks had reasonable half-widths (<20–25 cm-1); and (iii) good agreement between the calculated sum of all of the components and the experimental spectra was achieved (r2 > 0.999). The relative contents of different secondary structure elements were estimated by dividing the areas of individual peaks, which were assigned to a particular secondary structure, by the whole area of the resulting amide I band. The results of two independent experiments were averaged.

Analysis of the Polarized ATR-FTIR Spectra—The ATR electric fields of incident light were calculated as shown in Equation 1 (37),

(Eq. 1)
where {theta} is the angle of a light beam to the prism normal at the point of reflection (45°) and n21 = n2/n1 (n1 and n2 are the refractive indices of ZnSe taken as 2.4 and of the membrane sample taken as 1.5, respectively). Under these conditions, Ex, Ey, and Ez are 1.09, 1.81, and 2.32, respectively. The electric field components together with the dichroic ratio (RATR, defined as the ratio between absorption of parallel to a membrane plane, Ap, and perpendicularly polarized incident light, As) are used to calculate the orientation order parameter, f, by the following formula shown in Equation 2 (38),

(Eq. 2)
where h is the fraction of transition dipoles in the molecule that belong to the ordered structure of the absorption band whose order parameter has been calculated. Lipid order parameters were obtained from the symmetric (~2853 cm-1) and antisymmetric (~2922 cm-1) lipid-stretching modes using the same equations, setting {alpha} = 90° (37). The average angle of orientation of the helical axis with respect to the bilayer normal, {gamma}, is calculated based on the order parameter with the formula shown in Equation 3 (38),

(Eq. 3)

Dye Transfer Fusion Assay—Peptide inhibition of cell-cell fusion was assayed by monitoring the redistribution of a water-soluble fluorescent probe from the effector cell to the target cell upon their co-incubation with each other. SupT target cells were labeled with calcein-AM at 10 µM for 60 min at 37 °C and then washed and resuspended in RPMI 1640 medium. The plated HeLa cells infected with the recombinant vaccinia construct VPE16, which expresses HIV-1 IIIB gp120–41 (4), were labeled with CMTMR at 20 µM for 1 h at 37 °C, washed several times, and combined with the target cells (1:3 effector-target cell ratio). Different concentrations of peptides, dissolved in PBS buffer, were then added. The cells were co-cultured for 2 h at 37 °C in 24-well plates (Costar, Cambridge, MA). C34 and C56 inhibition were tested in parallel on the same day. These experiments were repeated several times, and the general trend was found to be the same. Both phase and fluorescent images were collected using an Olympus IX70 with a x20 objective lens and a CCD camera (Princeton Instruments, Trenton, NJ). A 82,000 optical filter cube (Chroma Technology Corporation., Brattleboro, VT) was used for the excitation of calcein (494/517) and CMTMR (541/565). Three images per well were collected, and dye transfer from the donor to the acceptor cell was counted via Metamorph software (Universal Imaging, West Chester, PA). The scoring of fusion events was conducted as described previously (22).

CD Spectroscopy—The CD spectra of the different peptides (10 µM)in 1% LPC were determined in an Aviv 202 spectropolarimeter in a quartz optical cell with a 1-mm path length at 20 °C in the range of 193–260 nm (1-nm steps, 10-s averaging). The spectrum of 1% LPC was subtracted from the spectra of the peptides in LPC.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS AND DISCUSSION
REFERENCES
 
The C- and N-terminal Subdomains Potently Induce Lipid Mixing of LUV either Enriched in or Devoid of Cholesterol, Respectively—To compare the specificity with which peptides from the C- and N-terminal subdomains perturb membranes, we tested the ability of C20, C56, FP33, and N70 (refer to Fig. 1) to induce lipid mixing of LUV designed to reproduce the disparate membrane order conditions on the virion and target cell surface (following cholesterol depletion) (Figs. 2 and 3). PC LUV were chosen to mimic the cholesterol-depleted target cell surface because of their zwitterionic nature and the fact that the outer leaflet of blood cell membranes is enriched in PC and sphingomyelin (reviewed in Ref. 39), both of which share the same head group. PC alone does not self-organize into ordered versus unordered regions and therefore represents a homogenous membrane surface (40). To reproduce the well ordered viral membrane surface, we chose a lipid composition based on the work of Aloia et al. (8) who quantified the lipid composition of virions and target cells. For the viral envelope, they identify the dominant outer leaflet choline lipids PC and Spm to be in approximate equal abundance and show an approximate 1:1 ratio of cholesterol to phospholipid. Accordingly, we model the virion surface with a PC:Spm:chol (1:1:1) lipid composition. It has been shown that cholesterol promotes lateral organization in membranes at physiologic temperatures when combined with Spm (40). Organization of raft structures follows the association of Spm and cholesterol. In in vitro vesicle preparations of Spm and cholesterol, optimal inhomogeneity in lipid dispersions occurs with 25–40 molar % cholesterol (40). Therefore, in addition to modeling the virion lipid surface, PC:Spm: chol (1:1:1) vesicles have a strong propensity to spontaneously form ordered regions.



View larger version (19K):
[in this window]
[in a new window]
  		FIG. 2.
Dose dependence of lipid mixing of PC LUV induced by the different peptides. Peptide aliquots were added to mixtures of LUV (22 µM phospholipids concentration) containing 0.6 mol % each of NBD-PE and Rho-PE and unlabeled LUV (88 µM phospholipids concentration) in PBS. The increase of the fluorescence intensity of NBD-PE was measured 5 min after the addition of the peptide, and the percentage from the maximum is plotted versus the peptide/lipid molar ratio. The fluorescence intensity upon addition of Triton X-100 (0.05% v/v) is referred to as 100% lipid mixing. The experiments were repeated at least twice with the error bars for each point reflecting the variation between different sample measurements. Symbols: ({blacksquare}), C56; ({blacktriangleup}), C20; ({diamond}), N70; and ({circ}), fusion peptide. As a control, the dose-dependent lipid mixing of Spm:PC:PE:chol:phosphatidylserine (11:10:4:3:1) caused by N70 is indicated by filled circles connected by a dashed line.

 



View larger version (17K):
[in this window]
[in a new window]
  		FIG. 3.
Dose dependence of lipid mixing of PC:Spm:Chol (1:1:1) LUV induced by the different peptides. In separate experiments, increasing amounts of peptide were added to a fixed amount of vesicles as described in Fig. 2. Symbols: ({blacksquare}), C56; ({blacktriangleup}), C20; ({diamond}), N70; and ({circ}), fusion peptide.

 
Comparing Fig. 2 with Fig. 3, we show a clear correlation of the ability of C- and N-terminal constructs to perturb cholesterol-rich and cholesterol-free vesicles, respectively. Fig. 2 confirms our previous finding that N70 is a potent fusogen toward PC LUV, reiterating the important role that the NHR plays in membrane destabilization (14). In addition, we checked the ability of the N-terminal subdomain to disrupt vesicles whose lipid composition models that found in the outer leaflet of blood cells (Spm:PC:PE:chol:phosphatidylserine at a 11:10:4:3:1 ratio (reviewed in Refs. 39 and 41). This mixture of lipids includes the microdomain inducing sphingomyelin and cholesterol, although the cholesterol content at ~10% is much less than that in Spm:PC:chol (1:1:1). Evidently, the absence or presence of ~10% cholesterol does not affect the fusogenic function of the N-terminal subdomain. Comparable to FP33, C20 has relatively low activity toward PC LUV, whereas inclusion of the CHR (C56) enhances lipid-mixing activity (Fig. 2). In addition, C34 is ineffective at inducing lipid mixing of PC LUV (data not shown). Our results here confirm the lipid-mixing ability of C20 (15–17) alone. More importantly we show that, in the context of the C-terminal subdomain, the function of C20 is preserved with the continuous CHR now contributing to membrane destabilization.

Interestingly, we find that C-subdomain peptides are superior to N-subdomain peptides in perturbing cholesterol-rich vesicles (Fig. 3). C56 and C20 vigorously perturb PC:Spm:chol LUV with C56 showing enhanced activity compared with C20 (as in PC LUV). It has been previously reported that C20 effectively induces lipid mixing of PC:Spm:chol (1:1:1) LUV (17). We confirm this finding and show that, in identical vesicles, the CHR contributes to the lipid-mixing function of the C-terminal subdomain. On the other hand, FP33 is ineffective at inducing lipid mixing of PC:Spm:chol LUV over the concentration range tested, whereas the activity of N70 is significantly reduced compared with PC LUV. Collectively, these results delineate the membrane-destabilizing potency of gp41 ectodomain fragments. C-terminal subdomain fragments (C20 and C56) destabilize cholesterol-rich LUV preferentially, whereas N-terminal subdomain fragments (FP33 and N70) preferentially fuse vesicles that are either cholesterol-free or contain ~10% cholesterol. The finding that FP33 is ineffective in perturbing PC:Spm:chol LUV may represent a safety mechanism that prevents the FP from destabilizing the viral envelope shortly after metastable release of gp41 from gp120 caused by receptors binding or highly labile gp120-gp41 contacts due to the hypermutability of HIV in the absence of receptor binding

For the two subdomains, the heptad repeats differ in the relative degree to which they contribute to lipid mixing of both cholesterol-rich and cholesterol-free vesicles. The NHR contributes dramatically in the N-terminal subdomain, whereas the CHR contributes to a lesser degree in the C-terminal subdomain. This implies that the IFP is primarily responsible for membrane destabilization induced by the C-terminal subdomain and correlates with the findings that the IFP effectively partitions into and self-assembles in membranes (17), whereas the CHR (C34) binds poorly to membranes (42). Further support for the dominant role of IFP in peptide-membrane interactions in context of the C-terminal subdomain can be gleaned from the evidence that DP-178, which represents C34 frameshifted 10 residues into the IFP, binds effectively to membranes and self-assembles therein (42) in direct correlation to IFP actions (17). In their pioneering in vivo work on the role of the IFP during fusion, Hunter et al. (43) suggested that the IFP is involved in destabilization of the viral membrane during fusion. Our findings align with this suggestion and identify the C-terminal subdomain as fusogenic with both IFP and CHR regions contributing.

C56 Is Capable of Inducing Lipid Mixing of PC LUV Alone and in the Presence of N70 —Acting alone, both C56 and N70 perturb PC LUV. To understand the interactive effect of these peptides, we compared the lipid-mixing ability of N70 and C56 alone and pre-mixed (Fig. 4). The data indicate that C56 and N70 act additively to induce lipid mixing of PC LUV, meaning that there is neither synergism nor interference between free peptides within the context of the assay. Formation of the N70-C56 core would sequester the membrane-interacting NHR and possibly also the IFP regions, probably resulting in a decreased ability of this complex to induce lipid mixing relative to the combined effects of the free N70 and C56 peptides. Our finding that the lipid-mixing function of the pre-mixed N70 and C56 parallels the additive effect of the peptides alone can be interpreted in different ways. (i) The non-covalent N70-C56 core is formed when the peptides are pre-mixed in Me2SO and then disassociates upon interaction with membranes as shown for the analogous N36-C34 core (44). (ii) The N70-C56 core is unable to form in Me2SO, with the kinetics of lipid mixing of free peptides, much faster than core formation in LUV. (iii) The N70-C56 core is not formed in either Me2SO or upon addition to membranes. It should be noted that we know of no reports showing HIV-1 non-covalent core formation using constructs that include the FP. It could be that for the N70 construct in solution, the hydrophobic FP folds back to interact with the hydrophobic groove of the NHR, thereby inhibiting CHR binding.



View larger version (25K):
[in this window]
[in a new window]
  		FIG. 4.
Dose dependence of lipid mixing of PC LUV induced by the peptides, alone or pre-mixed. In separate experiments, increasing amounts of peptide alone or pre-mixed were added to a fixed amount of vesicles as described in Fig. 2. Under conditions of pre-mixing, concentrated peptides dissolved in Me2SO were combined and allowed to incubate for 10 min prior to introducing to LUV. Symbols: C56, black bars; N70, gray bars; and C56 and N70 premixed together, white bars.

 
The ability of free peptides to induce lipid mixing and fusion of model vesicles follows the ability of the peptides to bind to and induce aggregation and fusion of vesicles. In the case of N70, we have shown that the peptides bind to vesicles and induce aggregation and subsequent fusion of the vesicles (14). Elucidation of the details by which membrane-associated peptides conduct the final fusion step promises to resolve central questions in protein-induced membrane fusion.

Circular Dichroism Spectroscopy Indicates That C56 Has Higher {alpha}-Helical Content Than C20 in a Membrane-mimetic Environment—The structure of fusogenic peptides has been shown to be important for their ability to promote fusion (45). Because C56 is more effective than C20 at inducing lipid mixing of both PC and PC:Spm:chol (1:1:1) LUV, we compared their secondary structure in membrane-mimetic LPC micelles using CD. Light scattering caused by the liposomes in suspension in the range of wavelengths essential for secondary structure determination prevented the usage of any type of vesicles in the CD measurements. Therefore, LPC at micellar concentrations (1% w/v) was chosen as an applicable membrane mimetic environment with low light scattering effect. Both peptides displayed spectra with minima at 208 and 222 nm (Fig. 5), typical of {alpha}-helix structure. Under these conditions, it has been reported that {alpha}-helical contents are underestimated (46). The fractional helicity of a peptide is proportional to its mean residue ellipticity at 222 nm. Given the 1.5-fold increase in absorbance of C56 over C20 at 222 nm along with the fact that C56 is far longer than C20, the CD data clearly demonstrate that the {alpha}-helical content of C56 is significantly higher than that of C20.



View larger version (18K):
[in this window]
[in a new window]
  		FIG. 5.
CD spectra of the peptides in 1% LPC. Lyophilized samples of peptide were dissolved directly in double-distilled water containing 1% w/v LPC. The spectra were taken at peptide concentrations of 10 µM. The upper and lower spectra correspond to the C20 and C56, respectively.

 
FTIR Spectroscopy Indicates That C56 Has Predominantly {alpha}-Helical Content in Cholesterol-enriched Membranes—The enhanced fusogenic function of C56 compared with C20 may be related to differences in structure and/or peptide-membrane interactions. We used FTIR spectroscopy to study peptide structure in association with PC:Spm:chol (1:1:1) membranes and the ability of the peptides to disrupt acyl chain order of the membranes. The experimental conditions of peptides interacting with deuterated phospholipids multilayers closely approximate biological conditions. For structural characterization, deconvolution of the amide bond absorbance in the amide I region (1600–1700 cm-1) allows identification of the main secondary structural components, which are characterized by regions of resolvable absorbance as described under the "Experimental Procedures." The spectra corresponding to C20, C56, and C34 are shown in Fig. 6. The assignments, wave numbers, and relative areas of the component peaks are listed in Table I. The different secondary structural elements were assigned according to Jackson and Mantsch (47). The spectrum between 1700 cm-1 and 1730 cm-1 in which proteins do not absorb was practically a straight line, confirming no contribution of residual water to the spectrum (data not shown).



View larger version (18K):
[in this window]
[in a new window]
  		FIG. 6.
FTIR spectra deconvolution of the amide I band (1600–1700 cm-1) of C20 (panel A1), C56 (panel B1), and C34 (panel C1) in deuterated PC:Spm:Chol (1:1:1) multilayers. The solid line represents the experimental FTIR spectra. The fitted spectra superimpose on the experimental spectra with good agreement (r2 >0.999) and are therefore not shown. The individual fitted components are represented below by dashed lines. The fourth derivatives, calculated to identify the positions of the components bands in the spectra, are shown in panels A2, B2, and C2 for C20, C56, and C34, respectively.

 


View this table:
[in this window]
[in a new window]
  		TABLE I
Secondary structure content of the peptides in deuterated PC:Spm:chol (1:1:1) multilayers determined by ATR-FTIR spectroscopy

 
Comparing C20 with C56 in Fig. 6, we see that their absorbance spectra in the amide I region are very similar. Deconvolution reveals the major component peak for C20 and C56 as being centered at 1656 and 1654, respectively, indicating predominant {alpha}-helical structure for both membrane-associated constructs. In this initial structural analysis of C20 and C56, we see only small differences. Specifically, a component peak at 1676 corresponding to 310-helical contribution is revealed in the fourth derivative of C56 and not C20 with a relative contribution of 11%. Otherwise, the two peptides share similar and dominant {alpha}-helical contributions (>55%) with lesser {beta}-sheet and {beta}-aggregate contributions. Under similar conditions, Nieva and co-workers (15) report that C20 adopts mainly {alpha}-helical structure. Because C56 is 36 residues longer than C20, it has many more residues contributing to {alpha}-helical structure. Examining the structure of the CHR (C34) alone in membranes, we see that its amide I absorbance spectra differs from that of C20 and C56 by being more broad with the maximal absorbance shifted to the left, away from the {alpha}-helical band. Deconvolution reveals that the major component peak is centered at 1643 in the random coil region, indicating that C34 is largely unstructured in deuterated membranes.

The findings that the C-terminal subdomain is more effective than the IFP at inducing lipid mixing of both cholesterol-rich and cholesterol-free LUV may be explained by cooperative stabilization of the CHR and IFP regions in the context of the C-terminal subdomain. As suggested by FTIR experiments in deuterated PC:Spm:chol multibilayers and supported by CD measurements in membrane-mimetic LPC micelles, the C-terminal subdomain has an {alpha}-helical fragment that is longer than that of the IFP. Because C56 is predominantly {alpha}-helical with no random coil contributions, helical structure is induced in the CHR when conjugated to the largely helical IFP since the free CHR peptide (C34) is largely unstructured in deuterated membranes as in solution (48), which may explain its lack of lipid-mixing function. Therefore, interaction with the deuterated membranes does not induce or stabilize {alpha}-helical structure in the CHR as does interaction with other regions in the gp41 ectodomain (IFP and NHR). It seems likely that the primary and/or secondary structural arrangement of the IFP as an adjacent region is required to stabilize the CHR secondary structure toward a predominantly helical state, similar to the one it assumes in the hairpin. As the CHR structure is stabilized by the IFP, alternatively, the CHR probably stabilizes the IFP to a degree, specifically at the N terminus of the IFP. The complementary ability of the two regions to stabilize one another will affect their functional roles as discrete regions in the context of the protein. Stabilization of the IFP by the CHR will enhance the membrane-perturbing ability of the IFP. Stabilization of the CHR by the IFP will enhance the ability of the CHR to interact with the coiled-coil NHR regions, facilitating hairpin formation.

Both C20 and C56 Disrupt phospholipid Acyl Chain Order in Cholesterol-enriched Membranes—Polarized ATR-FTIR spectroscopy was used to analyze the effect of the peptides on membrane order of both PC and PC:Spm:chol (1:1:1) membranes. The orientation of each multilayer with and without peptide was determined. The symmetric ({nu}sym ~2853 cm-1) and antisymmetric ({nu}antisym ~2923 cm-1) vibrations of lipid methylene C–H bonds are perpendicular to the molecular axis of fully extended hydrocarbon chains in membranes. Measuring the dichroism of infrared light absorbance can reveal the order and orientation of the membrane sample relative to the prism surface. The R values corresponding to both the symmetric and anti-symmetric vibrations for the lipids in the absence of peptides were determined, and the order parameters f were calculated. The order parameters based on the symmetric and anti-symmetric vibrations are both similar in magnitude and follow an identical trend in variation between pure multilayers and peptides in multilayers. Therefore, we show only the order parameter derived from the symmetric vibrations. The effect of the peptides on the acyl chain order was estimated by comparing the CH2-stretching dichroic ratio of pure multilayers alone with that obtained for peptide bound to identical multilayers with membrane-bound peptide. The calculated ratios given as R values with corresponding order parameters f are shown in Table II. The data reveal that both C20 and C56 disrupt the acyl chain order of PC:Spm:chol membranes to a similar degree with C34 slightly more disruptive. These results suggest that the differential lipid-mixing ability of C56 and C20 is not a function of their ability to disrupt the order of the membranes with which they interact. Our findings that the CHR is largely unstructured in deuterated membranes but adopts helical structure when conjugated to C20 suggest that free CHR peptides are artifactual in structure. Therefore, the effect of C34 alone on membrane order should be considered with caution. In correlation to their ability to induce lipid mixing of PC:Spm: chol vesicles, N70 disrupts the same multilayers to a lesser degree relative to C20 and C56, whereas FP33 is ineffective in perturbing the membrane order.


View this table:
[in this window]
[in a new window]
  		TABLE II
ATR dichroic analysis of lipid order in deuterated multilayers

 
The ability of the peptides to affect membrane organization of PC multilayers runs parallel to their ability to induce lipid mixing of the same vesicles. FP33 and C20 are marginal in inducing lipid mixing of PC LUV, and each peptide in PC multilayers shows no measurable change in lipid order relative to pure PC multilayers. The same is true for C34, which displays no lipid mixing of PC LUV. For the constructs corresponding to the two subdomains, we see that both N70 and C56 significantly perturb PC multilayers.

When exposed and interacting with membranes during fusion in vivo, it is probable that the CHR region preferably occupies an {alpha}-helical conformation similar to its hairpin helix. This amphipathic {alpha}-helical arrangement provides a hydrophobic interface for associating with the membrane, with the potential to affect membrane destabilization. However, our results comparing peptide-membrane interactions show that both the IFP and C-terminal subdomain disrupt the lipid acyl chain order to the same degree for ordered multilayers. Although not conclusive, these results suggest that {alpha}-helical regions of the CHR are not contributing to membrane destabilization directly and that their primary role may be interdigitation with the coiled-coil to form the hairpin. Collectively, our findings point to the IFP, stabilized in context of the C-terminal subdomain, as being primarily responsible for the membrane fusion function of the C-terminal subdomain.

C56 Is an Inhibitor of HIV-1-induced Cell-Cell Fusion—Free peptides derived from the C-terminal subdomain, such as C34 and T20, have been shown to inhibit cell-cell fusion at nanomolar concentrations (23, 25). The peptides are believed to act through a dominant-negative mechanism to interfere with formation of the "hairpin"-like conformation of the ectodomain by specific binding to the hydrophobic groove on the surface of the N-helical coiled-coil (13, 18, 23, 44, 49). Peptide binding to the exposed coiled-coil presumably prevents gp41 folding into the hairpin, which is believed to correspond with fusion. Modifications to C34 and T20 that stabilize {alpha}-helical structure improve their inhibitory activity (50, 51), presumably by enhancing their interaction with the exposed viral coiled-coil in the PHI conformation. Because of the strong {alpha}-helical signal of C56 in a membrane-mimetic environment, we postulated that C56 will be an effective inhibitor of cell-cell fusion. We performed separate experiments with labeled cells testing the inhibition activity of C56 with C34 used as a control. The plotted data were fit to a hyperbolic decay function (Fig. 7). The concentration at which 50% inhibition is reached (IC50) can be extracted from each curve. The IC50 for C34 and C56 were estimated at 125 ± 34 nM (R2 = 0.98) and at 66 ± 38 nM (R2 = 0.97), respectively. The difference in IC50 between the two peptides is ~2-fold. Student's t tests showed that the difference in the patterns of inhibition shown in Fig. 7 is statistically significant (p = 4 x 10-7). It should be noted that the previously reported value of IC50 for C34 in a similar assay was lower than the one determined here (42). The main reason for the difference is that the vaccinia system utilized in our experiment expresses a higher amount of gp120/gp41 on the cell surface. Accordingly, more C34 is required to inhibit fusion. Our results show that the C-terminal subdomain preserves the highly potent inhibitory activity of C34. Moreover, the data suggest that the C-terminal subdomain is even a slightly better inhibitor than C34. This leads us to postulate that the C-terminal subdomain maintains the required structure to carry out its function throughout the fusion event.



View larger version (14K):
[in this window]
[in a new window]
  		FIG. 7.
Cell-cell content mixing in the presence of ({blacksquare}) C56 and ({blacktriangleup}) C34. HIV-1 envelope expressing HeLa cells were labeled with CMTMR, and the CD4+ CXCR4+ SupT1 cells were labeled with calcein. After 2 h of co-culture at 37 °C, the cells were examined for solute dye transfer. Three fields are taken of each sample and averaged. The solid and dashed lines represent hyperbolic decay fits to the data with correlation coefficients of 0.97–0.98. C34 and C56 curve fits yield an IC50 of 125 and 66 nM, respectively. These differences are statistically significant (p = 4 x 10-7).

 
Concluding Remarks—Membrane fusion can be described through a succession of stages: (i) apposition of membranes; (ii) hemifusion of outer leaflets to form a stalk; (iii) fusion pore formation through fusion of both bilayers at the locus of the stalk; and (iv) fusion pore enlargement with full content mixing. Recently, Markosyan et al. (27) showed that fusion pore formation occurs prior to hairpin formation. These results suggest that exposed regions of the PHI are responsible for the bulk of the fusion work (up until fusion pore initiation) through protein-membrane interactions because incubation with free NHR and CHR peptides freezes fusion presumably by sequestering exposed viral NHR and CHR regions through peptide-protein interactions (27). Prior to hairpin folding, the two subdomains have the opportunity to interact with their adjacent membranes. Specifically, the C-terminal subdomain can interact with the adjacent viral envelope, whereas the N-terminal subdomain can interact with the target cell membrane. Membrane fusion requires membrane destabilization, and it has been shown that regions from both C- and N-terminal subdomains perturb membranes (14, 16, 17, 45, 52–54). Based on our results, we propose that the C- and N-terminal subdomains exert their effects following PHI formation and up until the hairpin formation through direct protein-membrane interactions (refer to Fig. 8). We believe that the role of the C-terminal subdomain, similar to the N-terminal subdomain, has the dual function of binding to and destabilizing membranes and folding into the hairpin conformation. These actions are necessarily sequential. For the C-terminal subdomain, the IFP binds to and destabilizes the viral membrane as an {alpha}-helix, which is stabilized by the CHR. The CHR is also helical, stabilized by the IFP, but seems to be less involved in membrane interaction and more reserved for interdigitation with the NHR coiled-coil to form the hairpin at later stages when these regions become juxtaposed through membrane apposition or changes in membrane architecture. For the N-terminal subdomain, the NHR binds to membranes, localizing inserted FPs, and both regions combine to destabilize the target cell membrane. The N-terminal subdomain is as effective in destabilizing both unordered cholesterol-free membranes and cholesterol-containing membranes whose lipid composition mimics that of the blood cell surface but has diminished effect on inhomogenous cholesterolrich membranes. Our observations imply that in addition to the known function of the CHR to interact with the NHR to form the hairpin, in the context of the C-terminal subdomain, the CHR can contribute to the HIV-1 fusion process by coordinating with the IFP to affect the organization of the highly ordered cholesterol-rich viral membrane.



View larger version (44K):
[in this window]
[in a new window]
  		FIG. 8.
Model of gp41-membranes interaction following metastable release of gp41 by gp120-receptors binding. A, in the PHI, the FP is inserted into the target cell membrane with exposed helical NHR, CHR, and IFP regions. B, affinity of the IFP and NHR regions for membranes will draw them to adjacent membrane surfaces (viral and target cell, respectively) allowing them to affect membrane organization. C, membrane destabilization to the point of hemifusion can juxtapose the membrane-associated NHR with the helical CHR to facilitate hairpin folding. D, much of the IFP is likely accommodated in the hairpin conformation as an extension of the CHR. In the hairpin, the hydrophobic surfaces of the NHR coiled-coil and the extended CHR are sequestered, thereby diminishing the likelihood of hairpin-membrane interactions from these regions. Note: in B and C, an additional dimension was included to illustrate the binding of NHR and IFP regions to membranes.

 

   FOOTNOTES
 
* This study was supported by the Minerva Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

To whom correspondence should be addressed: Dept. of Biological Chemistry, The Weizmann Institute of Science, Rehovot 76100, Israel, Tel.: 972-8-9342712; Fax: 972-9344112, E-mail: Yechiel.Shai{at}weizmann.ac.il.

1 The abbreviations used are: HIV-1, human immunodeficiency virus, type 1; ATR-FTIR, attenuated total reflection Fourier transform infrared spectroscopy; chol, cholesterol; FP, fusion peptide; IFP, internal fusion peptide; CD, circular dichroism; CMTMR-5 and -6, ((4-chloromethyl)benzoyl)amino))tetramythylrhodamine; Fmoc, N-(9-fluorenyl)-methoxycarbonyl; gp41, 41,000-Da transmembrane subunit glycoprotein of HIV-1; LUV, large unilamellar vesicles; LPC, lysophosphatidylcholine; NBD, 4-fluoro-7-nitrobenz-2-oxa-1,3-diazole; PC, egg phosphatidylcholine; PE, dioleoylphosphatidylethanolamine; PBS, phosphate-buffered saline; Spm, egg sphingomyelin; Rho, tetramethylrhodamine; TES, triethylsilane; NHR, N-terminal heptad repeat; CHR, C-terminal heptad repeat; PHI, pre-hairpin intermediate; ZnSe, zinc-selenium. Back



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS AND DISCUSSION
 REFERENCES
 

  1. Stegmann, T., Doms, R. W., and Helenius, A. (1989) Annu. Rev. Biophys. Biophys. Chem. 18, 187-211[CrossRef][Medline] [Order article via Infotrieve]
  2. Stein, B. S., and Engleman, E. G. (1990) J. Biol. Chem. 265, 2640-2649[Abstract/Free Full Text]
  3. Hallenberger, S., Bosch, V., Angliker, H., Shaw, E., Klenk, H. D., and Garten, W. (1992) Nature 360, 358-361[CrossRef][Medline] [Order article via Infotrieve]
  4. Earl, P. L., Moss, B., and Doms, R. W. (1991) J. Virol. 65, 2047-2055[Abstract/Free Full Text]
  5. Center, R. J., Leapman, R. D., Lebowitz, J., Arthur, L. O., Earl, P. L., and Moss, B. (2002) J. Virol. 76, 7863-7867[Abstract/Free Full Text]
  6. Nguyen, D. H., and Hildreth, J. E. (2000) J. Virol. 74, 3264-3272[Abstract/Free Full Text]
  7. Ono, A., and Freed, E. O. (2001) Proc. Natl. Acad. Sci. U. S. A. 98, 13925-13930[Abstract/Free Full Text]
  8. Aloia, R. C., Tian, H., and Jensen, F. C. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 5181-5185[Abstract/Free Full Text]
  9. Clapham, P. R., and McKnight, A. (2002) J. Gen. Virol. 83, 1809-1829[Abstract/Free Full Text]
  10. Salzwedel, K., and Berger, E. A. (2000) Proc. Natl. Acad. Sci. U. S. A. 97, 12794-12799[Abstract/Free Full Text]
  11. White, J. M. (1992) Science 258, 917-924[Abstract/Free Full Text]
  12. Durell, S. R., Martin, I., Ruysschaert, J. M., Shai, Y., and Blumenthal, R. (1997) Mol. Membr. Biol. 14, 97-112[Medline] [Order article via Infotrieve]
  13. Chan, D. C., and Kim, P. S. (1998) Cell 93, 681-684[CrossRef][Medline] [Order article via Infotrieve]
  14. Sackett, K., and Shai, Y. (2002) Biochemistry 41, 4678-4685[CrossRef][Medline] [Order article via Infotrieve]
  15. Suarez, T., Nir, S., Goni, F. M., Saez-Cirion, A., and Nieva, J. L. (2000) FEBS Lett. 477, 145-149[CrossRef][Medline] [Order article via Infotrieve]
  16. Suarez, T., Gallaher, W. R., Agirre, A., Goni, F. M., and Nieva, J. L. (2000) J. Virol. 74, 8038-8047[Abstract/Free Full Text]
  17. Saez-Cirion, A., Nir, S., Lorizate, M., Agirre, A., Cruz, A., Perez-Gil, J., and Nieva, J. L. (2002) J. Biol. Chem. 277, 21776-21785[Abstract/Free Full Text]
  18. Weissenhorn, W., Dessen, A., Calder, L. J., Harrison, S. C., Skehel, J. J., and Wiley, D. C. (1999) Mol. Membr. Biol. 16, 3-9[CrossRef][Medline] [Order article via Infotrieve]
  19. Jones, P. L., Korte, T., and Blumenthal, R. (1998) J. Biol. Chem. 273, 404-409[Abstract/Free Full Text]
  20. Furuta, R. A., Wild, C. T., Weng, Y., and Weiss, C. D. (1998) Nat. Struct. Biol. 5, 276-279[CrossRef][Medline] [Order article via Infotrieve]
  21. Koshiba, T., and Chan, D. C. (2003) J. Biol. Chem. 278, 7573-7579[Abstract/Free Full Text]
  22. Munoz-Barroso, I., Durell, S., Sakaguchi, K., Appella, E., and Blumenthal, R. (1998) J. Cell Biol. 140, 315-323[Abstract/Free Full Text]
  23. Wild, C. T., Shugars, D. C., Greenwell, T. K., McDanal, C. B., and Matthews, T. J. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 9770-9774[Abstract/Free Full Text]
  24. Wild, C., Oas, T., McDanal, C., Bolognesi, D., and Matthews, T. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 10537-10541[Abstract/Free Full Text]
  25. Jiang, S., Lin, K., Strick, N., and Neurath, A. R. (1993) Nature 365, 113[Medline] [Order article via Infotrieve]
  26. Chen, C. H., Matthews, T. J., McDanal, C. B., Bolognesi, D. P., and Greenberg, M. L. (1995) J. Virol. 69, 3771-3777[Abstract]
  27. Markosyan, R. M., Cohen, F. S., and Melikyan, G. B. (2003) Mol. Biol. Cell 14, 926-938[Abstract/Free Full Text]
  28. Chan, D. C., Fass, D., Berger, J. M., and Kim, P. S. (1997) Cell 89, 263-273[CrossRef][Medline] [Order article via Infotrieve]
  29. Tan, K., Liu, J., Wang, J., Shen, S., and Lu, M. (1997) Proc. Natl. Acad. Sci. U. S. A. 94, 12303-12308[Abstract/Free Full Text]
  30. Weissenhorn, W., Dessen, A., Harrison, S. C., Skehel, J. J., and Wiley, D. C. (1997) Nature 387, 426-430[CrossRef][Medline] [Order article via Infotrieve]
  31. Campbell, S. M., Crowe, S. M., and Mak, J. (2002) AIDS 16, 2253-2261[CrossRef][Medline] [Order article via Infotrieve]
  32. Guyader, M., Kiyokawa, E., Abrami, L., Turelli, P., and Trono, D. (2002) J. Virol. 76, 10356-10364[Abstract/Free Full Text]
  33. Viard, M., Parolini, I., Sargiacomo, M., Fecchi, K., Ramoni, C., Ablan, S., Ruscetti, F. W., Wang, J. M., and Blumenthal, R. (2002) J. Virol. 76, 11584-11595[Abstract/Free Full Text]
  34. Struck, D. K., Hoekstra, D., and Pagano, R. E. (1981) Biochemistry 20, 4093-4099[CrossRef][Medline] [Order article via Infotrieve]
  35. Gazit, E., Miller, I. R., Biggin, P. C., Sansom, M. S., and Shai, Y. (1996) J. Mol. Biol. 258, 860-870[CrossRef][Medline] [Order article via Infotrieve]
  36. Surewicz, W. K., Mantsch, H. H., and Chapman, D. (1993) Biochemistry 32, 389-394[CrossRef][Medline] [Order article via Infotrieve]
  37. Ishiguro, R., Matsumoto, M., and Takahashi, S. (1996) Biochemistry 35, 4976-4983[CrossRef][Medline] [Order article via Infotrieve]
  38. Tatulian, S. A., and Tamm, L. K. (2000) Biochemistry 39, 496-507[CrossRef][Medline] [Order article via Infotrieve]
  39. Zachowski, A. (1993) Biochem. J. 294, 1-14
  40. Silvius, J. R. (2003) Biophys. J. 85, 1034-1045[Abstract/Free Full Text]
  41. Devaux, P. E., and Zachowski, A. (1994) Chem. Phys. Lipids 73, 107-120
  42. Kliger, Y., Gallo, S. A., Peisajovich, S. G., Munoz-Barroso, I., Avkin, S., Blumenthal, R., and Shai, Y. (2001) J. Biol. Chem. 276, 1391-1397[Abstract/Free Full Text]
  43. Salzwedel, K., West, J. T., and Hunter, E. (1999) J. Virol. 73, 2469-2480[Abstract/Free Full Text]
  44. Kliger, Y., Peisajovich, S. G., Blumenthal, R., and Shai, Y. (2000) J. Mol. Biol. 301, 905-914[CrossRef][Medline] [Order article via Infotrieve]
  45. Nieva, J. L., Nir, S., Muga, A., Goni, F. M., and Wilschut, J. (1994) Biochemistry 33, 3201-3209[CrossRef][Medline] [Order article via Infotrieve]
  46. Pecheur, E. I., Sainte-Marie, J., Bienvene, A., and Hoekstra, D. (1999) J. Membr. Biol. 167, 1-17[CrossRef][Medline] [Order article via Infotrieve]
  47. Jackson, M., and Mantsch, H. H. (1995) Crit. Rev. Biochem. Mol. Biol. 30, 95-120[Medline] [Order article via Infotrieve]
  48. Lu, M., and Kim, P. S. (1997) J. Biomol. Struct. Dyn. 15, 465-471[Medline] [Order article via Infotrieve]
  49. Lu, M., Blacklow, S. C., and Kim, P. S. (1995) Nat. Struct. Biol. 2, 1075-1082[CrossRef][Medline] [Order article via Infotrieve]
  50. Judice, J. K., Tom, J. Y., Huang, W., Wrin, T., Vennari, J., Petropoulos, C. J., and McDowell, R. S. (1997) Proc. Natl. Acad. Sci. U. S. A. 94, 13426-13430[Abstract/Free Full Text]
  51. Otaka, A., Nakamura, M., Nameki, D., Kodama, E., Uchiyama, S., Nakamura, S., Nakano, H., Tamamura, H., Kobayashi, Y., Matsuoka, M., and Fujii, N. (2002) Angew. Chem. Int. Ed. Engl. 41, 2937-2940[CrossRef][Medline] [Order article via Infotrieve]
  52. Peisajovich, S. G., Epand, R. F., Pritsker, M., Shai, Y., and Epand, R. M. (2000) Biochemistry 39, 1826-1833[CrossRef][Medline] [Order article via Infotrieve]
  53. Martin, I., Defrise-Quertain, F., Decroly, E., Vandenbranden, M., Brasseur, R., and Ruysschaert, J. M. (1993) Biochim. Biophys. Acta 1145, 124-133[Medline] [Order article via Infotrieve]
  54. Rafalski, M., Lear, J. D., and DeGrado, W. F. (1990) Biochemistry 29, 7917-7922[CrossRef][Medline] [Order article via Infotrieve]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us