INTRODUCTION

Top Abstract Introduction Materials & Methods Results & Discussion References

Entry of enveloped viruses into host cells requires fusion of the viral envelope with a host cell membrane. This reaction is carried out by the spike envelope glycoproteins found on the viral surface and can occur at either low or neutral pH, depending on the virus (2, 16). Viruses that require low pH for fusion bind to receptors at the cell surface and are then internalized via the endocytic pathway. Biochemical and biophysical studies have shown that in the low-pH environment of the endosome, the corresponding viral fusion proteins undergo conformational changes that transform them into active fusion proteins (25). These changes include exposure of a previously buried hydrophobic region of the protein called the fusion peptide. Interaction of the fusion peptide with the endosomal membrane initiates the fusion event (16). Viruses that do not require low pH are thought to enter cells by fusing directly at the plasma membrane. Considerably less is known about the mechanisms of viral fusion proteins that function at neutral pH (18).

The avian leukosis and sarcoma viruses (ALSVs) are members of the retrovirus family and, like most retroviruses, fuse with target cells at neutral pH (11). ALSVs are divided into five major subgroups designated A through E based on host range, receptor binding, and interference patterns. The ALSV envelope (Env) glycoprotein is synthesized as a glycosylated precursor, Pr95, that is proteolytically cleaved during transport to the plasma membrane, yielding the mature surface (SU) and transmembrane (TM) subunits, with the SU subunit containing receptor binding and subgroup specificity determinants (29). The SU and TM subunits remain associated by a disulfide bond(s) and are oligomerized as trimers (5). For ALSV subgroup C (ALSV-C), it has been shown that posttranslational cleavage is required for infectivity (23).

The overall organization of ALSV Env is similar to those of the fusion proteins of other viruses, especially those of other retroviruses and members of the paramyxovirus and orthomyxovirus families. Like ALSV Env, the fusion proteins of retro-, paramyxo-, and orthomyxoviruses are proteolytically processed from a precursor protein into the mature receptor binding and TM subunits. The TM subunits of these fusion proteins generally contain a stretch of hydrophobic amino acids at their amino termini. In several cases, this region has been implicated in viral fusion activity and has thus been termed the fusion peptide. In general, fusion peptides (i) are 15 to 25 amino acids in length, (ii) are relatively hydrophobic, (iii) can be modeled as an alpha helix with the bulky hydrophobic residues lying on one face of the helix, and (iv) are rich in alanine and glycine (31). Where studied, the fusion peptide can be labeled by photoactivatable phospholipids in target membranes (4, 15), and mutations that introduce changes at hydrophobic residues impair or abolish fusion activity (8, 10, 32, 36). Fusion peptides are not always at the N terminus; internal fusion peptides have been identified in the Semliki Forest virus (SFV) and vesicular stomatitis virus (VSV) fusion proteins (20, 32, 36). Internal fusion peptides generally contain a helix-breaking residue such as proline near their centers (31). Synthetic fusion peptides adopt either -helical or -sheet structures in membranes (16).

By analogy with other viral fusion proteins, a sequence in ALSV Env, residues 21 to 42 of the TM subunit, has been predicted to be its fusion peptide. Consistent with its suggested importance, this internal hydrophobic region is highly conserved among the different ALSV subgroups. Furthermore, Ebola virus, a member of the filovirus family whose envelope glycoprotein bears striking resemblance to the TM subunit of ALSV Env, contains a candidate fusion peptide at an analogous location (9). The candidate fusion peptide of ALSV Env, however, has not yet been subjected to experimental analysis.

We tested the role of the candidate fusion peptide of the Env of ALSV-A (Env A) by mutating three of its hydrophobic residues to charged residues. The effects of these mutations on the processing, cell surface expression, oligomerization, and receptor binding of Env as well as the ability of the mutant Envs to mediate infection and cell-cell fusion were examined. Our findings are fully consistent with the prediction that this hydrophobic region serves as the fusion peptide of ALSV Env.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results & Discussion References

Recombinant DNA and plasmids. Plasmids pCB6-Env A, pCB6-Env C, pCB6-Env Acl, and pCB6-Tva have been described previously (12, 13). Mutations within the putative fusion region were created by oligonucleotide-directed site-specific mutagenesis (17) on single-stranded preparations of Env A cDNA which had been cloned into the BamHI site of Bluescript. The oligonucleotides used were Glu 1 (5'CTTTGGGGTCCTACAGCTCGAGAATTTGCATCTATCTTAGCCCCGGGG3'), Glu 2 (5'GCATCTATCTTAGCCCCGGGGGAAGCTGCAGCGCAAGCCTTAAGA3'), Glu 3 (5'CAAGCCTTAAGAGAAGAAGAGAGGCTAGCCTGTTGGTCCG3'), Lys 1 (5'CTTTGGGGTCCTACAGCAAGAAAATTTGCATCGATCTTAGCCCCGGGG3'), Lys 2 (5'GCATCTATCTTAGCCCCGGGGAAAGCTGCAGCGCAAGCCTTAAGA3'), and Lys 3 (5'CAAGCCTTAAGAGAAAAAGAGCGCCTAGCCTGTTGGTCCG3'). The oligonucleotides were designed to encode the altered amino acid as well as a new restriction site for rapid detection of mutants. Mutant plasmids were sequenced to confirm the fidelity of mutagenesis and then subcloned into the BamHI site of pCB6. The murine leukemia virus (MLV) Gag-Pol expression plasmid pHIT60 and the lacZ expression plasmid pHIT111 were gifts from Alan Kingsman. The pOS8 plasmid, containing lacZ under the T7 promoter, was a gift from Bernard Moss.

Antibodies. The rabbit polyclonal antibodies against the carboxy-terminal cytoplasmic tails of Env A and Env C have been described previously (12). The rabbit polyclonal antiserum anti-Ngp37 was raised against a peptide corresponding to the first 17 residues of the amino terminus of the TM subunit. Antibodies were affinity purified against the corresponding peptide coupled to a SulfoLink column (Pierce Chemical Company, Rockford, Ill.) according to the manufacturer's instructions. The rat anti-MLV Gag monoclonal antibody was obtained from culture supernatants of a hybridoma cell line purchased from the American Type Culture Collection. Donkey anti-rabbit (Amersham, Arlington Heights, Ill.) and goat anti-rat (Jackson ImmunoResearch Laboratories, West Grove, Pa.) secondary antibodies coupled to horseradish peroxidase (HRP) were used for Western blot analyses.

Cells, transfections, production of pseudotyped viruses, and virus titration. Stable NIH 3T3 cell lines were established by transfection of pCB6 constructs, using the calcium phosphate precipitation method (33), and by selection of Geneticin-resistant colonies as described previously (12, 13). Single-cell clones of the fusion peptide mutant cell lines were obtained by limiting dilution. All stable NIH 3T3 cell lines and 293T cells were maintained in Dulbecco's modified Eagle's medium (DMEM) containing 10% supplemented calf serum (SCS; HyClone, Ogden, Utah) and 500 mg of Geneticin per liter. To induce higher levels of protein expression, cells were treated with sodium butyrate overnight: 5 mM for Env Acl; 10 mM for Tva, Glu 1, Glu 2, Lys 1, and Lys 2; 15 mM for Glu 3; 20 mM for Lys 3; and 25 mM for Env A.

Envelope incorporation into virus. To check incorporation of wild-type (wt) and mutant Envs into (MLV) ALSV pseudotyped viruses, 293T cell viral supernatants were centrifuged at 1,500 × g for 10 min at room temperature to remove cell debris. Viral supernatants (8 ml) were centrifuged at 4°C through 20% sucrose for 2 h at 100,000 × g in a Beckman SW41 rotor. Pellets were resuspended in Laemmli sample buffer, boiled, analyzed on a Western blot probed with either the anti-A tail or anti-MLV Gag antibody, and visualized by enhanced chemiluminescence (ECL; Amersham). All blots were exposed in the linear range. To confirm Env incorporation, wt, Glu 1, Glu 2, and Glu 3 viral pellets were resuspended in 200 µl of TE buffer (5 mM Tris-hydrochloride, 1 mM EDTA [pH 8.6]) with brief sonication. The samples were then layered onto a 15 to 60% linear sucrose gradient in TE buffer and centrifuged at 4°C for 3 h at 250,000 × g in a Beckman SW41 rotor. Fractions of 500 µl were collected, precipitated with chloroform-methanol (16a), and analyzed on a Western blot probed with the anti-A tail antibody and a donkey anti-rabbit-HRP secondary antibody. Blots were then incubated for 30 min at 70°C in stripping buffer (2% sodium dodecyl sulfate [SDS], 100 mM 2-mercaptoethanol, a 62.5 mM Tris [pH 6.8]) to remove primary and secondary antibodies from the membranes. Membranes were reprobed with secondary antibody, incubated with ECL detection reagents, and exposed to film to ensure removal of any potential cross-reacting antibodies. Blots were then reprobed with the anti-MLV Gag antibody and a goat anti-rat-HRP secondary antibody and visualized by ECL.

Cell surface labeling, immunoprecipitation, coimmunoprecipitation, and thermolysin digestion assay. After induction with sodium butyrate for 16 to 18 h, cells expressing either Env or Tva were labeled with the membrane-impermeant biotinylation reagent NHS-LC-biotin (Pierce). Cells were lysed, and proteins were immunoprecipitated or coimmunoprecipitated with the anti-A tail antibody as described previously (13). Immunoprecipitates were washed and processed for sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) or were used in a thermolysin digestion assay as described previously (13). The soluble Tva (sTva) used in the thermolysin assay was a gift from Paul Bates. Following electrophoresis, proteins were transferred to nitrocellulose and probed with HRP coupled to streptavidin (Pierce) to detect biotinylated proteins. The HRP signal was detected by ECL.

Analysis of oligomer formation. To determine the oligomeric state of the Env proteins, cell lysates, prepared as described in reference 14, containing biotinylated Env proteins were layered onto a 10 to 30% linear sucrose gradient in HEPES-buffered saline containing 40 mM octylglucoside and centrifuged in a Beckman SW41 rotor for 17 h at 275,000 × g at 4°C. The gradient was fractionated and processed for the detection of Env proteins as described previously (14).

Fusion assay. The fusion assay used is a modified version of that reported by Nussbaum et al. (22). 293T cells were cotransfected as described above with pOS8 and a pCB6-Env vector or pCB6-Tva; 16 to 18 h before the fusion assay, six-well dishes of Tva- or Env-expressing 3T3 cells were infected with the modified vaccinia virus Ankara (MVA) (34) at a multiplicity of infection of approximately 10 for 30 min at 37°C in DMEM supplemented with 2% SCS. Cells were then washed twice with DMEM-10% SCS and incubated for 16 to 18 h at 31°C in DMEM-10% SCS supplemented with 100 µg of cytosine -D-arabinofuranoside (Ara-C; Sigma, St. Louis, Mo.) per ml. Uninfected Env- or Tva-expressing 293T cells were lifted off the dish with calcium- and magnesium-free phosphate-buffered saline (PBS), centrifuged, resuspended in DMEM-10% SCS containing 100 µg of Ara-C per ml, and then added to vaccinia virus-infected Tva- or Env-expressing cells. The cell mixtures were incubated at 37°C for 12 h. Cells were analyzed for -galactosidase activity either by staining with X-Gal or by measuring -galactosidase activity as described below. MVA was a gift from Bernard Moss.

Analysis of -galactosidase activity. Following the fusion assay, cells were lifted off the dish, pelleted, and lysed in 50 µl of lysis buffer (1% Nonidet P-40 in 130 mM NaCl-20 mM HEPES [pH 7.4]). A 5- to 10-µl aliquot of the cell lysate was diluted in 90 µl of lysis buffer containing 0.1 mg of 4-methylumbelliferyl galactoside (MUG; Molecular Probes, Eugene, Oreg.) per ml and incubated at 37°C for 4 min. The reactions were quenched with 2 ml of 0.2 M glycine (pH 10.0), and the fluorescence in each sample was determined in an LS-5B fluorimeter (Perkin-Elmer, San Jose, Calif.) with the excitation and emission wavelengths set at 365 and 450 nm, respectively.

Flow cytometry. Approximately 2 × 10⁵ transiently transfected 293T cells were incubated on ice with the anti-Ngp37 antibody in a total volume of 100 µl of PBS-2% fetal calf serum (FCS) for 30 min. Cells were washed twice in PBS and then incubated with fluoresceinated goat anti-rabbit antibody (Jackson ImmunoResearch Laboratories) in 100 µl of PBS-2% FCS on ice. After 30 min, cells were washed twice with PBS, resuspended in PBS containing 2% paraformaldehyde, and analyzed at the University of Virginia FACS (fluorescence-activated cell sorting) Core Facility, using a FACScan flow cytometer (Becton Dickinson Immunocytometry Systems, San Jose, Calif.).

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials & Methods Results & Discussion References

Design of mutant Env proteins. Sequences are available for the Env proteins of four of the five major ALSV subgroups. Analysis of these sequences revealed a highly conserved hydrophobic region near the amino terminus of the TM subunit (amino acids 21 to 42; Fig. 1A and B). This region is also strikingly similar in relative location and overall hydrophobicity to the predicted fusion peptide of the Ebola virus envelope glycoprotein (9). By analogy with other viral fusion proteins, this region is predicted to be the fusion peptide of ALSV Env. Similar to many fusion peptides, this region displays a hydrophobic face when modeled as an alpha helix (Fig. 1C). To examine the importance of this region, we engineered six mutations that replaced the hydrophobic residues Ile-21, Val-30, and Ile-39 (marked as positions 1, 2, and 3) on the hydrophobic face of the modeled helix with the charged residues glutamic acid or lysine. We predicted that insertion of a charged residue would disrupt the hydrophobicity of the fusion peptide and therefore impair its ability to interact with membranes and initiate fusion. We chose the charged amino acids glutamic acid and lysine since they have side chain volumes similar to those of the replaced amino acids valine and isoleucine and display either a negative or a positive charge. We refer to the mutant Env proteins as Glu 1, Glu 2, Glu 3, Lys 1, Lys 2, and Lys 3, with the name designating the amino acid and position of the substitution (Fig. 1C). Note that although we have modeled this region as an alpha helix, it is quite possible that part or all of this region adopts alternate structures either before or during fusion (16).

View larger version (19K): [in this window] [in a new window]   FIG. 1.   ALSV putative fusion peptide region. (A) Overlay of the hydropathy plots of the TM subunits of Env glycoproteins of subgroups A (Schmidt-Ruppin), C (Prague), D (Schmidt-Ruppin), and E (RAV-0) ALSV. Arrow points to the candidate fusion peptide region. *, TM domain. (B) Amino acid sequence of the candidate fusion peptide region in four ALSV subgroups. , identical amino acid at that position. (C) Putative fusion peptide region of ALSV-A modeled as an alpha helix. The hydrophobic face is outlined and has an average hydrophobicity index of 1.01. The position 1, 2, and 3 amino acids are circled. As discussed in the text, this region may adopt other structures either before or during fusion.

Expression, processing, and oligomerization of mutant Env proteins. Because the designed mutations are within a highly conserved region that may be important for protein structure, we first characterized the expression, processing, and oligomerization of the mutant Envs. Stable NIH 3T3 cell lines expressing each of the mutant proteins were established. After induction with sodium butyrate for 16 to 18 h, cells were lysed and analyzed by Western blotting with an antibody against the cytoplasmic tail of Env A to assess total Env expression. Alternatively, cells were labeled with the membrane-impermeant reagent NHS-LC-biotin, lysed, and immunoprecipitated with anti-A tail antibody to assess cell surface forms of Env. As shown in Fig. 2A, a Pr95 polypeptide, similar in mobility to wt Pr95, was expressed in all of the mutant cell lines. The proteins were processed to the mature form since a TM subunit similar in mobility to the wt TM was detected in each mutant. Glu 1, Glu 2, Lys 1, and Lys 2 were processed and transported to the cell surface similar to wt Env since only mature SU and TM proteins were detected at the cell surface (Fig. 2B). The position 3 mutants were also expressed as mature protein at the cell surface. However, in these cases processing was less efficient, since some precursor (a band similar in size to Pr95) was transported to the surface. We next examined oligomerization of the mutant proteins. Cells were labeled with the membrane-impermeant biotinylating reagent, lysed, and subjected to sucrose density centrifugation (Fig. 3). All of the mutant proteins sedimented to a position similar to that of wt Env, consistent with trimer formation (5, 14). For all mutants except Glu 3 and Lys 3, only the mature protein was observed in the trimer fractions. For Glu 3 and Lys 3, some trimerized Pr95 was found as well.

View larger version (48K): [in this window] [in a new window]   FIG. 2.   Expression of mutant proteins in stable NIH 3T3 cell lines. (A) Lysates of Env-expressing NIH 3T3 cells were boiled and reduced in sample buffer, separated by SDS-PAGE (11.5% gel), transferred to nitrocellulose, and Western blotted with anti-A tail antibodies. (B) A parallel set of Env-expressing cells were biotinylated with the membrane-impermeant reagent NHS-LC-biotin, lysed, and immunoprecipitated with anti-A tail antibodies as described in Materials and Methods. Immunecomplexes were analyzed by SDS-PAGE (11% gel), transferred to nitrocellulose, probed with streptavidin-HRP, and detected by ECL. The SU, TM, and Pr95 proteins migrated as shown.

View larger version (88K): [in this window] [in a new window]   FIG. 3.   Sucrose density centrifugation analysis of cell surface Env proteins. Env-expressing NIH 3T3 cells were biotinylated as described in the legend to Fig. 2B, lysed, and subjected to sucrose gradient centrifugation as described in Materials and Methods. After centrifugation, fractions were collected, immunoprecipitated with an anti-A tail antibody, subjected to SDS-PAGE (9% gel), and transferred to nitrocellulose. Biotinylated proteins were detected as described in the legend to Fig. 2B. Fraction 1 indicates the top of the gradient.

Ability of mutant Envs to interact with receptor and change conformation. Tva is the receptor for ALSV-A (1, 35) and appears to be sufficient to mediate viral entry. We previously showed that Tva binds specifically to Env A but not to Env C as measured by a coimmunoprecipitation assay (12). We also showed that upon interaction with sTva, a soluble form of the Tva ectodomain, Env A undergoes a conformational change, demonstrated by the generation of a specific thermolysin digestion product of the SU subunit termed SU* (13). We tested whether the mutant Env proteins were able to bind Tva and, if so, whether they could undergo the receptor-induced conformational change.

View larger version (106K): [in this window] [in a new window]   FIG. 4.   Receptor binding activity of mutant Env proteins. NIH 3T3 cells expressing Tva were biotinylated and lysed as described in the legend to Fig. 2B. Unlabeled lysates of cells expressing Env A, Env C, or mutant Env A proteins were mixed with the biotinylated Tva lysate and then immunoprecipitated with anti-A tail (or anti-C tail) antibodies. Samples were resolved on SDS-PAGE (12% gel) and blotted to nitrocellulose. Biotinylated proteins were detected as described in the legend to Fig. 2B.

View larger version (67K): [in this window] [in a new window]   FIG. 5.   Thermolysin digestion of envelope proteins in the absence and presence of sTva. (A) NIH 3T3 cells expressing Env proteins were biotinylated with the membrane-impermeant reagent NHS-LC-biotin, lysed, immunoprecipitated with anti-A or anti-C tail antibody, and washed. Immunecomplexes were then subjected to thermolysin digestion in the presence or absence of sTva as described previously (13). Samples were boiled, separated by SDS-PAGE (11% gel), transferred to nitrocellulose, probed with streptavidin-HRP, and visualized by ECL. The SU and TM subunits and SU* migrate as indicated. (B) The thermolysin digestion assay was repeated with the Glu 3 and Lys 3 mutants, using three times as much Env protein.

Infectivity of mutants. MLV particles efficiently incorporate ALSV Env into their envelopes (19). The resulting MLV(ALSV) pseudotypes are capable of infecting cells, but only if they express the appropriate ALSV receptor. We tested the ability of the mutant Envs to mediate entry of MLV(ALSV-A) pseudotypes into cells expressing Tva as a correlate of their fusion activity. MLV(ALSV-A) particles containing a lacZ reporter gene were made in 293T cells by using the three-plasmid transient transfection system (26). Briefly, the expression plasmids pHIT60 (containing MLV gag-pol), pHIT111 (containing lacZ), and pCB6-Env (either wt A, wt C, Acl, or a fusion peptide mutant) were cotransfected into 293T cells. Culture supernatants were collected 2 days posttransfection and assessed for particle production and envelope incorporation. Virus particles were pelleted from equal amounts of culture supernatants, resuspended in sample buffer, separated by SDS-PAGE, and Western blotted with the anti-A tail antibody (Fig. 6A). The same blot was then reprobed with the anti-MLV Gag antibody (Fig. 6B). All viral supernatants contained approximately the same number of viral particles since similar levels of Gag were detected. All mutant Env proteins were incorporated into virus at or near wt levels; equivalent amounts of TM protein (or Pr95 in the case of the Acl [13], a cleavage site mutant that is not processed to the mature SU and TM subunits) were detected in all of the samples.

View larger version (62K): [in this window] [in a new window]   FIG. 6.   Env incorporation into pseudotyped virus particles. Equal amounts of 293T cell viral supernatants were centrifuged through 20% sucrose as described in Materials and Methods. (A) Pellets were resuspended in sample buffer, boiled, analyzed on a Western blot probed with the anti-A tail antibody, and visualized by ECL. (B) The same blot was reprobed with the anti-Gag antibody and visualized by ECL.

View larger version (68K): [in this window] [in a new window]   FIG. 7.   Sedimentation analysis of pseudotyped virus particles. Equal amounts of 293T cell viral supernatants were pelleted through 20% sucrose as described in the text. Pellets were resuspended in 200 µl of TE buffer and subjected to centrifugation on a continuous 15 to 60% sucrose gradient as described in Materials and Methods. After centrifugation, fractions were chloroform-methanol precipitated, run on an SDS-12.5% gel, analyzed on a Western blot probed with the anti-A tail antibody, and visualized by ECL. Blots were then stripped as described in Materials and Methods, reprobed with the anti-Gag antibody, and visualized by ECL. 1 denotes the top of the gradient. A band is variably seen in fraction 1 of blots probed with the anti-Gag antibody. We do not think this is Gag because it migrates somewhat faster and always more broadly than Gag. No Pr95 was seen in these viral particles.

Fusion activity of mutant Envs determined by cell-cell fusion assay. The vaccinia virus/bacteriophage T7 system has been used previously to measure the fusion activity of viral envelope glycoproteins (7, 22). The assay measures the cytoplasmic activation of a reporter gene upon fusion of two different cell populations. We used a modified version of this system to measure fusion between cells expressing Env and cells expressing Tva. Briefly, Tva-expressing 3T3 cells were infected with a modified vaccinia virus Ankara (MVA), which encodes bacteriophage T7 RNA polymerase (34). MVA is a highly attenuated vaccinia virus strain that can assemble and replicate efficiently only in avian and BHK-21 cells (3a, 28). The infected cells were then overlaid with 293T cells which had been cotransfected with an Env expression plasmid and a plasmid with lacZ under the control of the T7 promoter (pOS8). Upon fusion, the lacZ gene is activated and the amount of -galactosidase produced can be measured either by in situ staining or by an enzymatic assay of detergent cell lysates.

View larger version (57K): [in this window] [in a new window]   FIG. 8.   Cell-cell fusion assay of Glu 1, Glu 2, Lys 1, and Lys 2. 293T cells were cotransfected with pOS8 and pCB6 vectors expressing Env proteins; 48 h posttransfection, cells were lifted off the dish and split into three samples that were either checked for surface expression of envelope proteins or used in the cell-cell fusion assay. A portion of transfected 293T cells was overlaid onto Tva-expressing 3T3 cells which had been infected with MVA. After 12 h at 37°C, samples were either fixed and stained with X-Gal (A) or lysed and analyzed in the MUG -galactosidase fluorometric assay (B). Results of the -galactosidase fluorometric assay are expressed as fluorescence units and represent averages from three replicates; bars show standard deviations. All samples were read in the linear range of fluorescence emission. (C) A portion of transfected 293T cells were biotinylated with NHS-LC-biotin, lysed, and immunoprecipitated with anti-A tail antibody. Samples were analyzed as described in the legend to Fig. 2B. (D) Another subset of transfected 293T cells was incubated on ice with the anti-gp37 antibody in a total volume of 100 µl PBS-2% FCS for 30 min. Cells were washed twice in PBS and then incubated with a fluoresceinated goat anti-rabbit antibody (Jackson ImmunoResearch Laboratories) in 100 µl of PBS-2% FCS on ice. After 30 min, cells were washed twice with PBS, resuspended in PBS containing 2% paraformaldehyde, and analyzed on a Becton Dickinson flow cytometer.

View larger version (25K): [in this window] [in a new window]   FIG. 9.   Cell-cell fusion assay of Lys 1 mutant in 3T3 cells. NIH 3T3 cells expressing Env A, Env Acl, or Lys 1 were seeded in six-well dishes (105 cells per well) and incubated overnight. Cells were then infected with MVA at a multiplicity of infection of 1 as described in Materials and Methods and incubated overnight at 31°C in the presence of 100 µg of Ara-C per ml and sodium butyrate. Cells were then either analyzed for surface expression of Env proteins or used in the cell-cell fusion assay. (A) One well of a six-well dish of infected 3T3 cells was biotinylated with NHS-LC-biotin, lysed, and immunoprecipitated with anti-A tail antibody. Samples were analyzed as described in the legend to Fig. 2B. (B) Infected 3T3 cells were overlaid with 293T cells which had been cotransfected with pOS8 and pCB6-Tva 2 days earlier. After 24 h cells, were lifted off the dish, lysed and analyzed in the MUG -galactosidase fluorometric assay. The data represent the averages of three replicates; bars show standard deviations.

View larger version (14K): [in this window] [in a new window]   FIG. 10.   Cell-cell fusion assay of Glu 3 and Lys 3. 293T cells were cotransfected with pOS8 and either pCB6-Env A, Env Acl, Glu 3, Lys 3, or pCB6 alone; 48 h posttransfection, cells were lifted off the dish and either checked for cell surface expression or used in the cell-cell fusion assay. (A) FACS analysis of transfected 293T cells. (B) Transfected 293T cells were overlaid onto MVA-infected Tva-expressing 3T3 cells. After 24 h at 37°C, samples were lysed and analyzed in the MUG -galactosidase fluorometric assay. The data represent the averages of three replicates; bars show standard deviations.

Summary. The viral fusion proteins of orthomyxo-, paramyxo-, and retroviruses are arranged similarly, with a mature transmembrane and surface subunit being generated by proteolytic processing of a precursor protein. The newly revealed N terminus of the TM subunit contains the fusion peptide, a hydrophobic domain that interacts with the target membrane to initiate fusion (16). Internal fusion peptides are not as common but have been identified in the E1 and G proteins of SFV and VSV, respectively (4, 20, 32, 36). Mutations within these latter fusion peptides affect membrane fusion. Photolabeling experiments have shown that, like the hemagglutinin fusion peptide (15), the VSV fusion peptide interacts hydrophobically with target membranes (4).