Membrane Interactions of Synthetic Peptides Corresponding to Amphipathic Helical Segments of the Human Immunodeficiency Virus Type- 1 Envelope Glycoprotein*

The human and simian immunodeficiency virus en- velope glycoproteins, which mediate virus-induced cell fusion, contain two putative amphipathic helical seg- ments with large helical hydrophobic moments near their carboxyl-terminal ends. In an attempt to elucidate the biological role of these amphipathic helical segments, we have synthesized peptides corresponding to residues 768-788 and 826-864 of HIV-l/WMJ-22 gpl60. Circular dichroism studies of the peptides showed that the a helicity of the peptides increased with the addition of dimyristoyl phosphatidylcholine (DMPC) indicating that the peptides form lipid-asso-ciating amphipathic helixes. The peptides solubilized turbid suspensions of DMPC vesicles, and electron mi- croscopy of peptide-DMPC mixtures revealed the formation of discoidal complexes, suggesting that the pep- tides bind to and perturb lipid bilayers. The peptides were found to lyse lipid vesicles and caused carboxyfluorescein leakage from dye-entrapped egg phospha- tidylcholine liposomes. The peptides also lysed human erythrocytes and were found to be toxic to cell cultures. At of Eisenberg and co-workers (12, 13) and plots were generated using Macvector gene analysis software (International Biotechnologies, Inc., New Ha-ven, CT). The Schiffer-Edmundson helical wheels (19) were plotted using a program developed for vaxstations that orients the hydrophobic moment (approximately representing center of nonpolar face) toward the top of the wheel (29). The values for the hydrophobic moment (GES scale) are expressed as mean hydrophobic moment per residue. The values for the hydrophobicity of the amphipathic peptides are expressed as the mean hydrophobicity per residue on the nonpolar face and were calculated by averaging the hydrophobicity of the residues that map to the nonpolar face of the amphipathic helix. Synthesis, Purification, and Chnracterization of Peptides-Peptides were synthesized by the solid-phase procedure using an automated peptide synthesizer (Advanced Chemtech Inc., Louisville, KY) and Boc chemistry, as described previously (20). The amino acids, as their Boc derivatives, were attached to the benzhydylamine resin (BaChem, Torrence, CA) through a 4-oxymethyl-phenacetyl group using dicy-clohexylcarbodiimide (DCC) and 1-hydroxybenzotriazole (HOBT) (Sigma). Benzyl-based side chain protecting groups were used for Asp, Glu, and Lys residues; Arg residues were protected by the Tos group, whereas the His residues were protected by Bom group. Boc groups at each stage of the synthesis were removed by treatment with 50% trifluoroacetic acid in methylene chloride. Boc-amino acid deriv-atives were condensed with DCC/HOBT. The peptides were cleaved from the resin by cells. Cell fusion is readily observed in cocultures of rVVenv1-infected HSB cells and Molt4 cells, as well as rVVenvl-infected HSB cells and peptide-treated Molt4 cells. However, cell fusion was inhibited in cocultures of peptide-treated rVVenvl-in-fected HSB cells and Molt4 cells.

The human and simian immunodeficiency virus envelope glycoproteins, which mediate virus-induced cell fusion, contain two putative amphipathic helical segments with large helical hydrophobic moments near their carboxyl-terminal ends. In an attempt to elucidate the biological role of these amphipathic helical segments, we have synthesized peptides corresponding to residues 768-788 and 826-864 of HIV-l/WMJ-22 gpl60. Circular dichroism studies of the peptides showed that the a helicity of the peptides increased with the addition of dimyristoyl phosphatidylcholine (DMPC) indicating that the peptides form lipid-associating amphipathic helixes. The peptides solubilized turbid suspensions of DMPC vesicles, and electron microscopy of peptide-DMPC mixtures revealed the formation of discoidal complexes, suggesting that the peptides bind to and perturb lipid bilayers. The peptides were found to lyse lipid vesicles and caused carboxyfluorescein leakage from dye-entrapped egg phosphatidylcholine liposomes. The peptides also lysed human erythrocytes and were found to be toxic to cell cultures. At subtoxic concentrations, the peptides effectively inhibited the fusion of CD4+ cells infected with recombinant vaccinia virus expressing human immunodeficiency virus (H1V)-1 envelope proteins. Based on these results, and reported studies on the mutational analysis of HIV envelope proteins, we suggest that the amphipathic helical segments near the carboxyl terminus of HIV envelope proteins may play a role in lysis of HIVinfected cells and also may modulate the extent of cell fusion observed during HIV infection of CD4+ cells.
The human immunodeficiency virus (HIV)' is the etiologic agent responsible for acquired immunodeficiency syndrome (AIDS), a chronic disease primarily affecting the immune and nervous systems. HIV primarily infects CD4+ T lymphocytes and macrophages. The viral envelope glycoproteins play an important role both in early and late events in viral infection (1,2). The HIV envelope glycoprotein is first synthesized as a precursor polyprotein gp160; it is then cleaved intracellularly to give rise to a surface (SU) glycoprotein designated as gp120, and a transmembrane (TM) glycoprotein designated as gp41 (3,4). The SU and TM proteins remain associated as a complex held together by noncovalent interactions (5). The SU protein is located on the outer surfaces of the viral envelope and infected cells, and functions as the viral attachment protein which binds to CD4 receptor on cell surfaces, whereas the TM protein serves as the membrane anchor. The amino terminus of the TM domain, generated upon cleavage of gp160, contains a stretch of hydrophobic amino acids termed the "fusion domain." This domain is necessary for fusion of the viral envelope with cellular membrane (6,7). When cell-free virus interacts with a potential target cell, the envelope glycoproteins specifically bind to the CD4 antigen, the primary cellular receptor for HIV. Subsequently, viruscell membrane fusion allows entry of the HIV nucleocapsid (8)(9)(10). Later in the viral life cycle, the HIV envelope glycoprotein is transported to the plasma membrane where it is available for virus assembly and budding. The surface-expressed envelope protein is also capable of inducing syncytium formation with neighboring uninfected CD4+ cells (2). Virusinduced cell fusion is a hallmark of HIV infection both in vivo and in cell culture and represents a major mechanism of virusinduced cell killing. Lysis of single virus-infected cells has also been observed with HIV infection (11).
Eisenberg and co-workers (12,13) have identified two segments with relatively large hydrophobic moments, which can fold into amphipathic a helical structures, near the carboxyl terminus of HIV gp160. By computer modelling, Venable et al. (14) concluded that these amphipathic helixes are structurally similar to regions of various proteins which form transmembrane selective channels. Some isolates of HIV-2 and simian immunodeficiency virus (SIV) carry mutations in their envelope genes which result in truncated TM proteins that lack the amphipathic helical segments (15,16). These viruses, as well as certain mutants of HIV-1 with truncated TM proteins, are infectious and induce cell fusion but do not cause extensive single cell killing (17,18). In this paper we show that synthetic peptides corresponding to the amphipathic helical segments near the carboxyl terminus of HIV-1 a 1 6 0 form CY helical structures in nonpolar solvents and in the presence of lipids. In an attempt to elucidate the biological role of these amphipathic helical segments, we have investigated the interaction of these peptides with lipid bilayers and determined their effects on HIV-induced cell fusion.

EXPERIMENTAL PROCEDURES
Hydrophobic Moment Plots, Helical Wheel Plots, and Analysis of Amphipathic Helixes-The nucleotide sequences of various HIV-1, HIV-2, SIV, and other lentiviruses were obtained from GenBank. Helical hydrophobic moments were calculated according to Eisenberg and co-workers (12,13) and plots were generated using Macvector gene analysis software (International Biotechnologies, Inc., New Haven, CT). The Schiffer-Edmundson helical wheels (19) were plotted using a program developed for vaxstations that orients the hydrophobic moment (approximately representing center of nonpolar face) toward the top of the wheel (29). The values for the hydrophobic moment (GES scale) are expressed as mean hydrophobic moment per residue. The values for the hydrophobicity of the amphipathic peptides are expressed as the mean hydrophobicity per residue on the nonpolar face and were calculated by averaging the hydrophobicity of the residues that map to the nonpolar face of the amphipathic helix.
Synthesis, Purification, and Chnracterization of Peptides-Peptides were synthesized by the solid-phase procedure using an automated peptide synthesizer (Advanced Chemtech Inc., Louisville, KY) and Boc chemistry, as described previously (20). The amino acids, as their Boc derivatives, were attached to the benzhydylamine resin (BaChem, Torrence, CA) through a 4-oxymethyl-phenacetyl group using dicyclohexylcarbodiimide (DCC) and 1-hydroxybenzotriazole (HOBT) (Sigma). Benzyl-based side chain protecting groups were used for Asp, Glu, and Lys residues; Arg residues were protected by the Tos group, whereas the His residues were protected by Bom group. Boc groups at each stage of the synthesis were removed by treatment with 50% trifluoroacetic acid in methylene chloride. Boc-amino acid derivatives were condensed with DCC/HOBT. The peptides were cleaved from the resin by anhydrous hydrogen fluoride yielding a free carboxylic acid group at the carboxyl-terminal residue. To obtain the protected peptide, the Boc derivative of the first amino acid was directly attached to the benzhydylamine resin, and the resin-bound peptide was first treated with trifluoroacetic acid to remove the Boc group and then treated with acetic acid in the presence of DCC and HOBT. Upon cleavage with hydrogen fluoride, the released peptide was protected by an acetyl group at the amino terminus and an amide group at the carboxyl terminus. The crude peptides were first analyzed by HPLC under various solvent systems to identify optimum separation conditions. These conditions were translated to preparative systems and the peptides were purified by preparative HPLC using a Vydac C-4 column (25 X 2.2 cm). The purity of the peptides was ascertained by analytical HPLC and amino acid analysis. The concentrations of peptide solutions were determined by amino acid analysis and Pierce protein assay reagent.
Peptide-Lipid Interactions-Peptide-DMPC (dimyristoyl phosphatidylcholine, Avanti Polar Lipids, Birmingham, AL) complexes were prepared at molar ratios of 1:5 and 1:lO by previously described procedures (21). Briefly, an ethanolic solution of DMPC was evaporated under nitrogen and resuspended in phosphate-buffered saline (PBS, pH 7.4) by sonication. An appropriate amount of the peptide in PBS was added to the DMPC suspension. The absorbance at 400 nm was recorded after 20 h incubation at 25 "C. The peptide-DMPC mixtures were also examined by electron microscopy. Samples after 20-h incubation were placed on formvar-coated copper grids, stained with uranyl acetate, and examined with a Philips EM 301 electron microscope.
Circular Dichroism (CD)-The peptides were dissolved in PBS or 80% trifluoroethanol at a concentration of 15 p~; peptide-DMPC complexes were prepared as described above. The samples were placed in a 10-mm sample cell, and CD spectra were obtained using a Jasco J-500A spectropolarimeter connected to a DPN500 data processor unit. The instrument was calibrated with d-10-camphorsulfonic acid. Approximately 8-16 scans were averaged and corrected for the base line. The corrected data was used to estimate the a helical contents according to the calculations described by Greenfield and Fasman (22).
Erythrocyte Lysis-Freshly drawn heparinized human blood from healthy donors was centrifuged and washed three times in PBS (pH 7.4) to remove plasma and the buffy coat. After the last wash, the cells were suspended in an equal volume of PBS and the hematocrit was determined. A series of doubling dilutions of the peptides was prepared in PBS. Appropriate amounts of erythrocyte suspension were added to the diluted solutions of peptides to provide a final erthryocyte concentration of 1%. After a 30-min incubation at 37 "C, the samples were centrifuged for 5 min at 750 g, and the absorbance of the supernatant at 540 nm was measured. Complete lysis was obtained by the addition of Triton X-100 (final concentration, 1%) to the erythrocyte suspension.
Cells, Viruses, and Cell Fusion Assays-Vero cells (Clone 76, American Type Culture Collection) and CD4+ HeLa cells (24) were maintained in Dulbecco's modified minimal essential medium supplemented with 5% newborn calf serum. CD4+ T lymphocytic cell line Molt4 and CD4-HSB cells were maintained in RPMI supplemented with glutamine and 10% fetal calf serum. VVenvl, a recombinant vaccinia virus which expresses the HIV-1 envelope protein (25), was used to study the effects of peptides on HIV-induced cell fusion. Fusion was monitored either by infection of Molt4 or CD4+ HeLa cell monolayers with VVenvl or by cocultivation of VVenvl-infected HSB cells and Molt4 cells, according to previously described procedures (26). HSV-1 (macroplaque), a syncytium-inducing variant of HSV, was obtained from Dr. S. Chatterjee (University of Alabama at Birmingham) and effects of peptides on HSV-induced cell fusion was assayed according to previously described procedures (27).
Envelope Protein Synthesis and Processing-Monolayers of CD4+ HeLa cells grown in 24-well cluster plates were infected at a multiplicity of 1 with either VVenvl or wild-type vaccinia virus (IHD-J strain). After a 1-h adsorption period, the supernatant was removed and replaced with 400 p1 of each peptide dissolved in Dulbecco's modified minimal essential medium. At 8 h post-infection, the cells were labeled with [35S]methionine for 2 h and lysed in 0.5 ml of lysis buffer (50 mM Tris-HC1 (pH 7.5), 0.15 M NaC1, 1% Triton X-100, 1% sodium deoxycholate, 0.1% sodium dodecyl sulfate, and 20 mM EDTA). The lysates were immunoprecipitated using HIV immune globulins (obtained through the AIDS Research and Reference Reagent Program, AIDS Program, National Institute of Allergy and Infectious Diseases, National Institutes of Health, ERC Bioservices Corporation) and protein-A Sepharose according to previously described procedures (28).
Surface Expression of Envelope Proteins-Monolayers of CD4+ HeLa cells grown on 12-mm glass coverslips in 24-well cluster plates were infected at a multiplicity of 0.1 and treated with peptide solutions as described above. At 10 h post-infection, intact unfixed monolayers were examined by indirect immunofluorescence assay using HIV immunoglobulin and fluorescein isothiocyanate-conjugated goat antihuman Ig (Southern Biotechnology, Inc., Birmingham, AL) as described previously (28).

Amphipathic Helical Segments in HIV and SIV Envelope
Glycoproteins-Helical hydrophobic moment plots of the HIV gp160 (826-854) envelope glycoprotein, gp160, reveal the presence of two highly amphipathic segments. One of these segments is located at the carboxyl-terminal end of the molecule (Segment I; residues 826-854 ), whereas the other (Segment 11) corresponds to residues 768-788, near the carboxyl terminus of the molecule (Fig. 1). Similar segments with large hydrophobic moments are also observed near the carboxyl termini of various strains of HIV-2 and SIV, and representative examples are shown in Fig. 1. Interestingly, envelope glycoproteins of bovine (BIV) and feline (FIV) immunodeficiency viruses and other animal lentiviruses (not shown) did not display such amphipathic helical segments near their carboxyl termini.
Genetic variation in HIV is extreme, particularly in its envelope gene. We therefore analyzed the effect of such variation on the conservation of the amphipathic helical segments. A comparison of the amino acid sequences (deduced from nucleotide sequences of HIV envelope genes available from the GenBank data base) of HIV envelope proteins in the region that corresponds to the putative carboxyl-terminal amphipathic helical regions is shown in Figs. 2 and 3. Extensive sequence variation was observed with Segment I, which usually involved a substitution of a polar residue with another polar residue or a nonpolar residue with another nonpolar residue. However, several changes were nonconservative and involved a substitution of a charged residue with a nonpolar residue or vice versa. In certain isolates, an Arg+ residue at position 19 was replaced by a Val or T h r residue. In several isolates the sequence changes occurred in pairs, wherein a change of hydrophobic residue to a charged residue was accompanied by the change of a charged residue to a hydrophobic residue (Fig. 3). Importantly, all these changes occurred at the polar face of the helix without disrupting its amphipathic nature. In Segment 11, all but two changes were conservative in nature. In one instance, a His residue at position 2 which maps to the polar face of the helix (Fig. 3) was replaced by an Arg residue. In other instances, a Lys' residue located at the polar-nonpolar interface was replaced with a hydrophobic residue (Thr, Val, or Ala). None of these sequence variations altered the amphipathic helical motif.
These results suggest the presence of a strong selective pressure to conserve the amphipathic helical segments, despite the extensive sequence variation in HIV.

Properties of Amphipathic Helical Peptides from HIV-1 gp160
helix sequences from a large number of published protein sequences and classified them into seven classes. We therefore compared the structural features of the two COOH-terminal amphipathic HIV gp160 peptides with the above mentioned classes of amphipathic helical segments. The HIV gp160 amphipathic peptides appear to be distinct from the seven previously described groups of amphipathic helixes, although they share several common features with amphipathic helixes found in the apolipoproteins, as well as the amphipathic helixes found in calmodulin-binding proteins. Like the apolipoprotein class of amphipathic helixes, the gp160 amphipathic helixes contain a cluster of positively charged residues at the polar-nonpolar interface. However, they are distinct from the apolipoprotein class of amphipathic helixes in that: (i) the nonpolar face of the helixes are more hydrophobic than the apolipoprotein class of amphipathic helixes. (ii) The helixes contain arginine residues in the center of the polar face and at the polar-nonpolar interface. (iii) The center of the most amphipathic of the gp160 helixes (Segment I) contains a proline residue. It is therefore possible that these three unique features of the amphipathic helixes of gp160 produce a bent helix that can penetrate deeply into the interior of the cell (or viral) membrane, perhaps inducing instability and leakage. The gp160 peptides also share several common features with the amphipathic peptides found in calmodulin-binding proteins, in that both helixes have a very broad polar face which is highly charged, and contain several positively charged residues on their polar face. However, the gp160 amphipathic helixes exhibit a much a greater hydrophobic moment. Interestingly, the amphipathic helixes from lytic polypeptides and polypeptide hormones also contain a predominance of positively charged residues at the center of their polar face, although they have a much narrower polar face. The precise function of the gp160 amphipathic helical segments is presently not clear. We have therefore investigated the properties of synthetic peptide analogs corresponding to these amphipathic helical segments, in an attempt to elucidate their possible role in HIV envelope protein function. Lipid Affinity of gp160 Amphipathic Helical Peptides-Two peptides corresponding to residues 826-854 (Peptide I) and 768-788 (Peptide 11) of gp160 from HIV-1 (WMJ-22) ( Fig. 2 and 3) were synthesized by the solid-phase method. Acetylated Peptide 11, Ac(768-788)NH2, was also synthesized to investigate the effects, if any, of end group modifications. An analytical HPLC profile of the purified peptides is shown in Fig.  4. The CD spectra of the peptides in aqueous and nonpolar solvents revealed minima at 222 and 208 nm, characteristic of LY helical structures (Fig. 5). Peptide I1 displayed a greater CY helical content both in buffer and in trifluoroethanol, compared with Peptide I. The acetylated Peptide I1 had a greater CY helical content compared with its nonacetylated counterpart, consistent with the idea that end group modifications influence helix stability. In the presence of a lipid such as DMPC, the peptides showed an increase in the a helical content which is consistent with the surface-active amphipathic helical nature of the peptides. The peptides also clarified turbid DMPC suspension as evidenced by a decrease in the absorbance at 400 nm. Electron microscopy of the peptide-DMPC solution suggested that the peptides bind to DMPC snd form discoidal complexes (Fig. 6). These results are consistent with the idea that the peptides form lipid-associating amphipathic CY helixes. PC vesicles with the amphipathic peptides resulted in the leakage of liposomal contents in a time-dependent and dosedependent manner (Fig. 7). Peptide I1 was significantly more potent in inducing liposomal dye leakage in that 100% leakage was accomplished by 0.2 PM peptide, whereas only a 25% leakage was observed with Peptide I at this concentration. The minimum concentration of Peptide I required to induce 100% dye leakage was 3 PM. Although similar amounts of acetylated and nonacetylated Peptide I1 were required to induce 100% leakage, the leakage was found to be complete within a minute by acetylated Peptide I1 as opposed to 8 min with the nonacetylated Peptide 11.
Effects of gpl60 Amphipathic Peptides on Eucaryotic CelLs-Biological membranes are far more complex than synthetic phospholipid bilayers and are likely to respond differently to membrane perturbing agents. We therefore investigated the effects of the gp160 amphipathic peptides on human erythrocytes. All of the gp160 amphipathic peptides were found to lyse human erythrocytes (Fig. 8). The percent of erythrocytes lysed increased gradually with increasing amounts of the peptide, and 100% lysis was observed at a peptide concentration of 6 PM acetylated Peptide I1 and 25 PM Peptide II; at the highest concentration tested (25 PM), Peptide I induced only a 25% lysis. These results suggest that the gp160 amphipathic peptides interact with and disrupt biological membranes, although the concentrations required to disrupt the biological membranes are much greater than the concentrations required to disrupt synthetic phospholipid bilayers. We have also investigated the effects of gp160 amphipathic peptides on cultured, nucleated cells (e.g. Vero cells, CD4+ HeLa cells, and Molt-4 cells) and found them to be cytotoxic at 100 PM or higher concentrations (data not shown).
Effects in Fig. 9. Approximately a 6040% inhibition of fusion was observed at -25 pM, and a nearly complete inhibition was observed at -50 p~ with Peptide 11. The gp160 amphipathic peptides did not inhibit the synthesis, processing, or surface expression of HIV envelope glycoproteins in rWenvl-infected HeLa or Molt4 cells (data not shown), suggesting that the inhibition of fusion was not due to toxic effects of the peptides. Virus-induced cell fusion was also studied in a two-cell coculture assay where a CD4+ cell was co-cultivated with a vaccinia-em recombinant-infected non-CD4 expressing cell. FIG. 8. Erythrocyte lysis by a160 peptides. A 1% erythrocyte suspension was exposed to varying concentrations of Peptide I (triangles), Peptide I1 (circles), or acetylated Peptide I1 (squares), After a 30-min incubation at 37 "C, the amount of hemoglobin released was determined spectrophotometrically, by measuring the absorbance at 540 nm, and the results are expressed as percent erythrocytes lysed. Complete lysis was obtained by the addition of Triton X-100 (final concentration, 1%) to the erythrocyte suspension.
We used a two-cell assay involving rVVenv-1 infected CD4-HSB cells and CD4 expressing M o b 4 cells to dissect the components of the fusion process affected by the amphipathic peptides. Uninfected Molt-4 cells or rVVenvl-infected HSB cells were incubated overnight with gp160 amphipathic peptides, prior to cocultivation. Extensive cell fusion was observed when neither cell type was pretreated with the peptides (Fig. 1OA). A significant reduction in fusion was observed when rVVenv1-infected HSB cells were pretreated with the gp160 amphipathic peptides (Fig. lOC), whereas pretreatment of CD4+ Molt-4 cells with the peptides did not result in fusion inhibition (Fig. 10B). These results indicate that the amphipathic peptides exert their inhibitory effect on the cell membranes expressing the HIV envelope glycoproteins and not on uninfected cells.
In order to determine whether the peptide-mediated effects are virus-specific, we also studied the effects of the gp160 amphipathic peptides on HSV-induced cell fusion (data not shown). effects of gp160 amphipathic peptides are not virus-specific. The precise mechanisms involved in peptide-mediated inhibiton of cell fusion induced by different viruses remain to be elucidated.

DISCUSSION
The carboxyl termini of the HIV and SIV envelope glycoproteins contain two amphipathic helical segments. Available evidence indicates that these carboxyl-terminal amphipathic segments of gp160 may fold into a helical structures and associate with membranes when present as part of the viral glycoprotein complex (30, 31). It is possible that the membrane association of the gp160 amphipathic helical segments may modulate membrane-related events (e.g. virus budding, cell fusion, or single-cell lysis) in the viral life cycle.
Our studies on the synthetic amphipathic peptide analogs of HIV-l/WMJ-22 gp160 demonstrate that they lyse synthetic phospholipid vesicles and human erythrocytes, and are cytotoxic to cultured cells. Miller et al. (32) have recently synthesized a peptide corresponding to Segment I of HTLV-III/LAV and also observed that this peptide is lytic to prokaryotic and eukaryotic cells. Unlike these in vitro studies, where the amphipathic peptides were exogenously added to the cells, the gp160 amphipathic segments are located intracellularly in HIV-infected cells. Studies with other cytolytic amphipathic peptides suggest a requirement for the formation of peptide oligomers for their lytic effects (34,35). The HIV envelope glycoproteins exist in the membranes as oligomers (either trimers or tetramers) and tend to concentrate in localized areas of the plasma membrane such as the sites of virus budding (36)(37)(38)(39). Therefore, the gp160 amphipathic helical segments may interact with the cytoplasmic leaflets of the plasma membranes and play a role in the lysis of virusinfected cells.
Studies on the naturally occurring variants and mutants of HIV and SIV also support the idea that the amphipathic helical segments in the viral envelope proteins may play a role in lysis of virus-infected cells (15)(16)(17)(18). Analysis of HIV-1 mutants with deletions in the carboxyl terminus of gp160 shows that mutations which disrupt the putative amphipathic helical segments result in viruses with reduced cytopathicity (18). For example, a variant of HIV-1, designated as X10-1, codes for a mutant envelope protein in which the 14 carboxylterminal residues have been replaced with 15 exogenous residues (K-R-R-R-R-W-V-F-Q-S-H-L-R-Y-L) and is fully capable of virus expression and syncytium formation but is no longer able to kill human T cells (40). Likewise, another mutant X9-3, in which the last five amino acids of gp160 were substituted with 153 amino acids of the 3' Orf also had a diminished cytopathic potential (40). These changes disrupt the amphipathic nature of one of the two putative carboxylterminal amphipathic helical segments in gp160 (Segment I). Sakai et al. (41) have molecularly cloned multiple virus genotypes from a cytopathic HIV isolate, which differ markedly in their infection kinetics and cytopathogenic properties. Evidence was obtained that the weakly cytopathogenic isolates contained changes in the 3' end of the envelope gene and 3' Orf regions. In light of earlier studies which show that the 3' Orf does not play a role in viral cytopathology, these results support the notion that the carboxyl-terminal regions of gp16O may contribute to viral cytopathology. Nevertheless, the amphipathic segments of gp160 are not likely to be the sole determinants of cell lysis in HIV-infected cultures (42, 43). Studies on mutant envelope glycoproteins, wherein the amphipathic helical segments have been disrupted by site-specific mutagenesis, without modifying other viral components, should provide a better understanding of their role in the viral life cycle.
Several naturally occurring variants of HIV-2, SIV, and mutants of HIV-1 which show TM protein truncations and fail to express the amphipathic helical segments in gp160, replicate well in culture, and release infectious virus particles containing truncated transmembrane proteins (15)(16)(17)(18)40,41). These variants also induce syncytium formation in CD4+ cells. These reports, therefore, suggest that the amphipathic helical segments in gp160 are not required for virus budding

Properties of Amphipathic Helical Peptides from
HIV-1 gp160 7127 or syncytium formation. We have shown that the gp160 amphipathic peptides exert a significant inhibitory effect on cell fusion induced by HIV-1 and an unrelated herpes simplex virus as well. It is therefore possible that the amphipathic helical segments may modulate the extent of cell fusion observed during HIV infections. Certain mutant HIV-1, HIV-2, and SIV glycoproteins with truncated cytoplasmic sequences that lack the amphipathic helical segments show a greater syncytium-forming ability compared with wild-type HIV-2 envelope proteins (44,45).' Furthermore, one mutant HIV-1 envelope protein carrying an extensive carboxyl-terminal deletion was found to be capable of inducing fusion of CD4+ murine cells, which are usually refractory to HIV-induced cell fusion (45). The presence of such fusion-modulatory sequences may provide certain growth advantages to the virus. The syncytia formed during HIV infection eventually die, possibly as a consequence of osmotic swelling (46), membrane permeability changes (47), and changes in ion flux (48). Although syncytium formation typically involves fusion of virus-infected cells with adjacent uninfected CD4+ cells, regions of plasma membranes on single infected cells may also undergo auto fusion, leading to cell death (49). Thus modulation by the amphipathic segments may limit the extent of fusion and cell death prior to optimum virus replication.
The mechanisms by which the gp160 amphipathic peptides inhibit virus-induced cell fusion are presently unclear. Since the gp160 amphipathic peptides also inhibit cell fusion induced by both HIV and HSV, fusion inhibition by amphipathic peptides may not involve sequence-specific proteinprotein interactions. In a two-cell fusion assay, an inhibitory effect was observed only when the cells expressing viral glycoproteins (but not the CD4 receptor) were pretreated with the peptides. These results indicate that: ( a ) the peptides may exert their effects on the membranes expressing viral glycoproteins and render them fusion-incompetent or ( b ) the conformation of the effector membranes (viral membranes or cell membranes carrying viral fusion protein) may be more critical than the conformation of the target cell membranes during virus-induced cell fusion. The lipid interactions of the amphipathic helical segments of gp160 and their membranerelated effects during HIV infections remain to be elucidated.