Download P Natl Acad Sci Usa 1994 Wild

Proc. Natl. Acad. Sci. USA
Vol. 91, pp. 12676-12680, December 1994
Biochemistry
Propensity for a leucine zipper-like domain of human
immunodeficiency virus type 1 gp4l to form oligomers correlates
with a role in virus-induced fusion rather than assembly of the
glycoprotein complex
CARL WILD*t, JOHN W. DUBAYt, TERESA GREENWELL*, TEASTER BAIRD, JR.§, TERRENCE G. OAS§,
CHARLENE MCDANAL*, ERIC HUNTERt, AND THOMAS MATTHEWS*
Departments of *Surgery and of §Biochemistry, Duke University, Durham, NC 27710; and tDepartment of Microbiology, University of Alabama at
Birmingham, Birmingham, AL 35294
Communicated by Peter K. Vogt, July 21, 1994
ABSTRACT
For a number of viruses, oligomerization is a
critical component of envelope processing and surface expression. Previously, we reported that a synthetic peptide (DP-107)
corresponding to the putative leucine zipper region (aa 553590) of the transmembrane protein (gp4l) of human immunodeficiency virus type 1 (HIV-1) exhibited a-helical secondary
structure and self-associated as a coiled coil. In view of the
tendency of this type of structure to mediate protein association, we speculated that this region of gp4l might play a role
in HIV-1 envelope oligomerization. However, later it was
shown that mutations which should disrupt the structural
elements of this region of gp4l did not affect envelope processing, transport, or surface expression (assembly oligomerization). In this report we compare the effects of amino acid
substitutions within this coiled-coil region on structure and
function of both viral envelope proteins and the corresponding
synthetic peptides. Our results establish a correlation between
the destabilizing effects of amino acid substitutions on coiledcoil structure in the peptide model and phenotype of virus
entry. These biological and physical biochemical studies do not
support a role for the coiled-coil structure in mediating the
assembly oligomerization of HIV-1 envelope but do imply that
this region of gp4l plays a key role in the sequence of events
associated with viral entry. We propose a functional role for the
coiled-coil domain of HIV-1 gp4l.
An essential component of human immunodeficiency virus
type 1 (HIV-1) infection involves the interaction of the
envelope glycoprotein (gpl20) with the CD4 protein expressed on the surface of the target cell (1-3). However, with
the exception of gpl20 shedding, processes which occur
subsequent to CD4/gpl2O interaction but prior to membrane
fusion remain poorly understood. The role of the transmembrane protein (gp4l) in this series of events has been the focus
of a number of research efforts (4-9). Reports have described
a leucine zipper-like (coiled-coil) motif located in the transmembrane protein of HIV-1 (10, 11). This domain is highly
conserved among HIV-1 isolates and analogous regions have
been identified in the transmembrane proteins of a number of
fusogenic viruses (11, 12). An alternating 4-3 (heptad) repeat
of ,-branched hydrophobic residues within the primary
amino acid sequence is characteristic of this particular structure. Although the functional role of this structural motif in
HIV-1 replication is yet to be established, studies in other
systems suggest that leucine zipper regions serve as control
elements by mediating protein multimerization (13-15). Because oligomerization is an essential component of HIV-1
envelope processing, transport, and expression (16-18),
there has been speculation that a leucine zipper structure
might play a role in this associative process (10, 11, 19). Carr
and Kim (20) reported that a pH-dependent conformational
change which induces fusion in the influenza virus system
includes formation of a coiled coil within the viral hemagglutinin and speculated that the leucine zipper-like domain of
gp4l might serve a similar role in the HIV-1 system.
To define the function of this region of gp4l, we have used
synthetic peptides and an envelope expression system. We
found that a synthetic peptide corresponding to aa 553-590 of
gp41 (DP-107) exhibits coiled-coil structure and accurately
models the leucine zipper domain of the transmembrane
protein. This peptide inhibits virus infection, and a strong
correlation exists between peptide structure and antiviral
activity (21). We and others have found that mutations within
this same region of the transmembrane protein, which should
disrupt the coiled-coil structure, had no effect on synthesis,
processing, transport, or surface expression of the HIV-1
envelope glycoprotein complex (22, 23). While these changes
did not affect assembly oligomerization, several of these
mutations resulted in gpl20/gp4l constructs which were
unable to support cell-cell fusion and infection (23, 24).
Here we report the results of experiments with peptide
analogs of the leucine zipper domain of gp4l which indicate
that the ability of this region to assume stable coiled-coil
structure is dramatically affected by substitutions at an
invariant isoleucine residue (Ile573). Our studies with intact
envelope indicate that while identical substitutions within
this region have little or no effect on HIV-1 envelope (assembly) oligomerization, these structure disrupting changes
greatly reduced or abrogated peptide antiviral activity and
reduced or blocked envelope-mediated virus entry. These
studies suggest that this region of gp4l does not play a major
role in HIV-1 envelope oligomerization, that the role of the
leucine zipper domain is related to the series of events
associated with virus entry, and that the coiled coil within this
region of gp4l is a critical component of this process.
MATERIALS AND METHODS
Peptide Synthesis and Circular Dichroism (CD). Peptides
were synthesized, purified, and characterized as described
(21). CD spectra were measured at peptide concentrations of
-10 ,M (21) determined from A280 (25). Thermal denaturation experiments were carried out over a concentration
range of 5-35 ,uM and the CD signal was monitored at 222 nm
(21).
Abbreviations: HIV, human immunodeficiency virus; RT, reverse
transcriptase; WT, wild type.
tTo whom reprint requests should be addressed at: P.O. Box 2926,
Duke University Medical Center, Durham, NC 27710.
The publication costs of this article were defrayed in part by page charge
payment. This article must therefore be hereby marked "advertisement"
in accordance with 18 U.S.C. §1734 solely to indicate this fact.
12676
Biochemistry: Wild et al.
Extracellular
Immunodominant
DP-107 s region
Fusion
peptide
NH P777i
553*
DP-107 (WT)
1-4
a
yy
m.
512 527
Proc. Natl. Acad. Sci. USA 91 (1994)
itated with HIV-positive serum. Immune complexes were
analyzed by SDS/8% PAGE.
Cell Fusion and Infectivity Assays. Peptide inhibition experiments were performed as described (21). In the infectivity
assay, cells were tested for virus production at 6 days
postinfection by measuring reverse transcriptase (RT) activity (see below), and the apparent infectious titer of the virus
was calculated for each concentration ofpeptide. Multiplicity
curves were generated by plotting the surviving virus fraction
(VN/Vo) against peptide concentration (where VN is the
infectious titer in the presence of peptide and VO is the titer
in the absence of peptide). The micro-RT assay was adapted
from Goff et al. (29) and Willey et al. (30) and was performed
as described (21). To determine the ability of envelope
mutants to support fusion, HeLa-T4 cells were plated in
35-mm dishes (Costar) and transfected with 1 pg of pSRHP
DNA. Two days after transfection, the cells were stained by
the May-Grunwald/Giemsa technique (23).
Intracellular
IKX\x
684 705
590* 604
rri
A
Transmembrane
region
NNLLRAIEAQQHLLQLTVWGZKQLQARILAVERYLKDQ
DP-121 (I573P) NNLLRAIEAQQHLLQLTVWGPKQLQARILAVERYLKDQ
DP-139 (1573S)
NNLLRAIEAQQHLLQLTVWGNKQLQARILAVERYLKDQ
DP-140 (1573A) NNLLRAIEAQQHLLQLTVWG&KQLQARILAVERYLKDQ
FIG. 1. HIV-1HXB2 transmembrane protein gp41 with position
573 highlighted. All peptides are acetylated at the amino terminus and
amidated at the carboxyl terminus. WT, wild type. (*) The numbering
convention differs from that employed in our earlier communication
(21). In this report the numbering corresponds to that used for the
HIV-lHXB2R isolate, with the DP-107 site designated 553-590. In the
earlier paper this same region was numbered according to the
HIV-1LAI isolate convention with the corresponding region labeled
558-595. Both the HIV-lHXB2R and the HIV-1Lum sequences are
numbered according to ref. 31. (**) Although this is a consensus
glycosylation site, it has not been established that this site is in fact
glycosylated.
RESULTS
CD Studies of gp4l Peptides. We investigated the effect of
substitutions at position 573 of the HIV-1 envelope coiled-
coil domain on the structural stability of the peptides modeling this region (Fig. 1). These experiments were carried out
to establish a correlation between coiled-coil stability and the
fusogenic potential of the corresponding HIV-1 envelope
proteins.
The CD spectra for the WT and mutant peptides (Fig. 2)
show the dramatic effect of single amino acid changes on
coiled-coil structure. The mean residue ellipticities at 222 nm
and 37° C for DP-107 indicate that the WT peptide is a80%
folded at this temperature whereas DP-140 (1573A) and
DP-139 (I573S) are significantly less structured. DP-121
(1573P) had no discernible secondary structure. Thermal
denaturation experiments were carried out to determine the
stability of the mutant peptides compared with WT (Table 1).
The DP-107 peptide (WT) at -35 ,uM exhibited extremely
stable solution structure, with a Tm of 76° C. In contrast,
DP-140 (I573A) and DP-139 (1573S) had Tm values of 38° C and
22° C, respectively. As observed previously, DP-121 (I573P)
exhibited no stable solution structure. The difference in
stability for the alanine-substituted peptide compared with
the WT sequence was observed over a broad concentration
range (Fig. 2C). The effect of the serine substitution was such
that accurate Tm values could not be calculated at peptide
concentrations < 35 ,uM (Fig. 2C).
Inhibition of Virus-Mediated Cell-Cell Fusion and Virus
Infection by gp4l Peptides. The WT gp4l peptide and the
I573X mutants were assayed for their ability to inhibit virus
fusion (Table 1). DP-107 (WT) proved to be a potent inhibitor
of cell-cell fusion (IC9o of 5 pg/ml). DP-140 (I573A) was
substantially less active (IC90 of 17 pg/ml). DP-139 (1573S)
gave a slight reduction in syncytium formation at 50 ug/ml
Virus. The HIV-1LAI virus was propagated in CEM cells
cultured in RPMI 1640/10% fetal bovine serum. Supernatant
from infected cells was filtered and the infectious titer was
estimated in a microinfectivity assay using the AA5 cell line
to support virus replication (21). The infectious titer of the
HIV-1LAI stock (calculated according to ref. 26) was 106
median tissue culture infectious doses (TCID5o) per ml.
Plasmids and Transfections. The expression vector pSRHP
(23) contains the HIV env gene from pHXB2Dgpt (27) under
control of the simian virus 40 late promoter. Plasmid DNAs
were propagated in Escherichia coli DH1 cells, purified (28),
and transfected by use of DEAE-dextran.
Oligonucleotide Mutagenesis. Site-directed mutagenesis
[Altered Sites (Promega) mutagenesis system; ref. 23] was
confirmed by dideoxy sequencing with the Sequenase system
(United States Biochemical).
Protein Biosynthesis and Cell Surface Expression of WildType and Mutant Glycoproteins. Labeling of COS-1 cells
transfected with the pSRHP constructs was described previously (23). Cell surface expression of envelope was monitored with monoclonal antibody 9284 (DuPont), which is
specific for the V3 loop of gp120 (23).
Glycoprotein Oligomerization Analysis. Oligomerization
analyses were as described previously except that cells were
lysed with Triton buffer (23). Precleared cell lysates were
loaded onto 5-25% sucrose gradients and centrifuged for 18
hr at 40,000 rpm in a Beckman SW41 rotor at 4° C. Fourteen
fractions were collected from each tube and immunoprecip-
A
0
B
80
-c
70
a
0
60
FIG. 2. CD spectra of wild-type and mutant peptides. (A and B) Spectra for 10 AM
DP-107 (i), DP-121 (O), DP-139 (+), and
DP-140 (A) at 0C (A) and 370C (B). (C) Tm for
50
-10
.-
-1
2
I-°, 40
00-20
E-i 30
-20
12677
20
-30
10
-30
200 210 220 230 240 250
200 210 220 230 240 250
Wavelength (nm)
Wavelength (nm)
o
10
20
[Peptide] gM
40
the peptides as a function of amino acid
substitutions at position 573. The Tm values
for DP-107 (s) and DP-140 (A) are concentration dependent, as expected for selfassociating species. The solution structure
of DP-139 (+) is disrupted to such an extent
that Tm could not be determined at concentrations < 35 ,uM.
Biochemistry: Wild et al.
12678
Proc. Natl. Acad. Sci. USA 91 (1994)
Table 1. Effect of substitutions at residue 573 of HIV-1 gp4l on the physical characteristics and the biological activity
of the gp4l peptide analogs and assembly oligomerization and phenotype of gp120/41 env gene products
Peptides
Envelope mutants
Fusion blockade
Neutralization
Assembly
Virus
Syncytium
Sequence
oligomerization
formation*
Tm, OC
(IC90, ,ug/ml)
replicationt
(IC90, ,ug/ml)
DP-107 (WT)
76
5
10
+++
+++ (25-50)
+++
+++
DP-140 (I573A)
38
17
14
+
+ (3-7)
DP-139 (1573S)
22
>40
>30
+++
-(<3)
-t
DP-121 (1573P)
+++
>40
>30
- (<3)
NT
DP-107 (WT) sequence: NNLLRAIEAQQHLLQLTVWGIKQLQARILAVERYLKDQ.
*Range of nuclei/syncytium is shown in parentheses.
tlnfectivity of virus released from COS-1 cells following transfection with pHXB2 plasmid containing WT or mutant env
genes in H9 cells. Virus spread was quantitated by the presence of RT activity in the culture medium. + + +, Peak of RT
activity (6 x 104 cpm/25 ,ul) at 7 days; +, peak of RT activity at 18 days 8-fold less than WT; -, no detectable RT activity
at 21 days; NT, not tested.
tDue to the lack of observable solution structure, a Tm value (midpoint of the thermal denaturation curve) for the 1573P
peptide (DP-121) could not be determined.
and DP-121 (I573P) exhibited no biological activity even at
100 ,g/ml. Similar results were obtained in experiments
designed to determine inhibition of cell-free virus infection
(Table 1). In that assay DP-107 at 10 ,g/ml gave a 90%
reduction in infectious virus titer, whereas the alanine mutant
(DP-140) at 14 ,ug/ml gave a similar result. The serine
(DP-139) and proline (DP-121) mutants exhibited no inhibitory activity at concentrations of up to 30 ,g/ml.
It is interesting to note the correlation between the intermediate antiviral activity of the alanine mutant (DP-140) and
peptide solution structure. In Fig. 2A it is clear from both the
positions and relative intensities of the characteristic double
minima that although DP-140 exhibits less structure than the
WT peptide, the structural components are predominately
a-helical. Thus, DP-140 (1573A) efficiently models the coiledcoil region of gp4l but does not exhibit the same degree of
overall stability as the WT peptide. This decreased stability
results in the partial loss of antiviral activity. In contrast, Fig.
2A illustrates the direct disruptive effect that the serine
(DP-139) and proline (DP-121) substitutions have on a-helical
structure. Unlike the alanine mutant (DP-140) the spectra of
these two peptides show much less evidence of secondary
structure. This inability to model the secondary structural
components results in peptides which do not block virus
replication. We believe that it is this difference between
destabilization of the coiled-coil structure and disruption of
a-helical structural components that accounts for the observed differences in peptide antiviral activity.
Synthesis, Processing, and Assembly Oligomerization of the
Mutant Envelope Proteins Occur Normally. To investigate the
kinetics of transport and processing of envelope mutants, we
carried out pulse-chase experiments using the alanine, serine, and proline mutants (Fig. 3). Following a 30-min pulse of
A
ALA
WT
SER
PRO
1H 2H 4H
1H 2H 4H
P 1H 2H 4H P 1H2H 4H P 1H 2H ' P 1H 2H 4H
a
:
_
.:.3:. .........::.
M
......MEMO
_
_.:
60-_
gp1
gpl 20--
B
gp 120_to
1H 2H 4H
* .:
1H 2H 4H
:W
--woj
:~~~
*'W
FIG. 3. Kinetics of glycoprotein processing and release of gpl20
into culture medium. The autoradiographs show the precursor
(gpl60) and cleavage product (gpl20) immunoprecipitated from WT
and mutant glycoprotein-expressing COS-1 cells after a 30-min
pulse-label (P) and chases of 1, 2, and 4 hr (H). (A) Cell lysates. (B)
Culture supernatants.
pSRHP-transfected COS-1 cells, equivalent levels of gp160
were immunoprecipitated from both WT and mutant glycoprotein-expressing cells (Fig. 3A). During chases of 1, 2, and
4 hr, similar levels of cell-associated gpl20 were present in
cells expressing WT protein as well as alanine and serine gp4l
mutants. Immunoprecipitation of cell culture supernatants
indicated that each of the proteins was transported to the cell
membrane with similar kinetics (Fig. 3B). In cells expressing
the proline mutant, little cell-associated gpl20 was observed
but higher levels of gpl20 could be immunoprecipitated from
the culture supernatants, indicating that the cleaved products
(gp4l and gp120) of this mutant precursor exhibited reduced
association (Fig. 3B).
To confirm that the proline substitution did not affect
protein oligomerization, experiments were performed to directly assay this process. COS-1 cells were transfected with
the pSRHP expression vector containing the WT or 1573P
mutant env genes. Cells were either pulse labeled for 15 min
or pulse labeled and chased for 3 hr. Cell lysates were run on
5-25% sucrose gradients (23). After centrifugation, 14 fractions were collected, immunoprecipitated with HIV-positive
serum, and analyzed by SDS/PAGE. The WT glycoprotein
showed a distinct increase in sedimentation rate after the 3-hr
chase, consistent with the formation of glycoprotein multimers; gp160 bands were immunoprecipitated from fractions
further down the gradient than those in the pulse sample (Fig.
4). Identical results were obtained with the I573P glycoprotein (Fig. 4), with the exception of a small amount of gpl20
found in the chase gradient, indicating a rapid dissociation of
gp120 following cleavage of the 1573P precursor.
Mutations Within the Leucine Zipper Region of gp4l Affect
Fusogenic Potential but Not Cell Surface Expression. To determine the biological consequences of coiled-coil destabilizing
mutations at aa 573 of the HIV-1 envelope, the ability of mutant
viral proteins to mediate virus entry was examined (Fig. 5;
Table 1). When the WT env vector was transfected into
HeLa-T4 cells, syncytia containing 25-50 nuclei were observed
after 48 hr (Fig. SA). In contrast, none of the mutant envelope
glycoproteins were able to induce WT syncytia. The I573A
mutant glycoprotein exhibited an intermediate phenotype that
was characterized by a reduced number of small syncytia that
contained 3-7 nuclei (Fig. 5B). The 1573S and 1573P mutations
resulted in envelope complexes that were unable to support
syncytium formation (Fig. 5 C and D). Similarly, when the
mutant env genes were substituted into the infectious proviral
clone pHXB2D, the alanine substitution yielded a virus that was
severely impaired in its ability to replicate whereas the serine
mutation rendered the virus noninfectious (Table 1). Because of
defects in precursor processing and gpl2O/gp4l association, the
1573P mutant env gene was not tested in this assay. With the
exception of 1573P, the ability ofthe mutant glycoproteins to be
Biochemistry: Wild et al.
*. .
Proc. Natl. Acad. Sci. USA 91 (1994)
Wild-Type
Bottom
Pulse
gpl60 -_
gpl60
Bottom
Top
15
I10
*
....
...........
:2.r.Q.
.....
.....
.:
15
f..... ..,
O- g
_k
*:: ::
Chase
1573P Mutant
Top
*. '.'^'' w......
...
p, sF
12679
:.
.._.
m
:::
_
p2plt
......
.......
FIG. 4. Oligomerization of WT and 1573P mutant proteins. COS-1 cells were transfected with pSRHP containing either WT or mutant env
Two days after transfection, the cells were either pulse-labeled (Pulse) and lysed in Triton buffer or pulse-labeled and chased for 3 hr
before lysis (Chase). The lysates were layered onto 5-25% sucrose gradients and centrifuged, and 14 fractions were collected. Each fraction
was immunoprecipitated with serum from an HIV-1-infected individual and the immune complexes were analyzed by SDS/PAGE. Arrows
indicate the precursor glycoprotein gpl60 and the cleavage product gpl20.
genes.
expressed on the surface of the COS-1 cells transfected in
parallel with the different vectors was unaffected by the mutations (Fig. 5 E-H). These results support the conclusion that
coiled-coil formation is critical to the fusogenic process but is
unnecessary for oligomer assembly, transport, and surface
expression.
DISCUSSION
Synthetic peptides can be used to model the coiled-coil
structures present in a number of biological systems (32). Our
work with gp4l peptides indicates that aa 553-590 of the
transmembrane protein of HIV-1 form a coiled-coil structure
(21). CD experiments illustrate the dramatic effect substitutions at the conserved isoleucine (I573X) can have on the
stability of this structure. Dubay et al. (23) suggested that
changes in envelope phenotype, which resulted from nonconservative amino acid substitutions at this position of the
transmembrane protein (Ile to Ser or Glu), reflected how
changes in hydrophobic character influenced protein function (23). However, the data presented here indicate that
these pheonotypic differences actually reflect the disruptive
effect of these changes on coiled-coil structure.
Based on the effect of primary sequence substitutions in
the peptide model, we expected that, if coiled-coil formation
were the driving force in assembly oligomerization, analogous mutations in this region of the intact envelope protein
FIG. 5. HeLa T4 cell-cell fusion and surface immunofluorescent
labeling of WT and mutant glycoproteins. (A-D) HeLa T4 cells were
transfected with pSRHP containing either WT or mutant env genes.
Two days after transfection, the cells were fixed and stained (MayGrunwald/Giemsa). (E-H) COS-1 cells were transfected with the
same plasmids, seeded on coverslips, and stained unfixed with
mouse monoclonal anti-V3 loop antibody and Texas Red-conjugated
goat anti-mouse antibody. (A and E) WT. (B and F) 1573A. (C and
G) 1573S. (D and H) I573P. (A-D, x390; E-H, x780.)
would in some way alter this process. However, as shown
here, both conservative and nonconservative changes at this
position of the TM had no detectable effect on assembly
oligomerization. Even after insertion of the helix-breaking
proline residue, we were unable to detect any differences in
the oligomeric state of the mutant and WT proteins. Hence,
it seems unlikely that formation of coiled coil is critical for
assembly oligomerization or for intracellular transport and
surface expression.
In contrast, amino acid substitutions in this region of the
HIV-1 transmembrane protein had a dramatic effect on the
ability ofthe envelope glycoprotein to mediate virus entry. This
defect in biological activity might be expected for the proline
substitution, since stability of the gpl20/gp41 complex is reduced, but not for the serine- or alanine-substituted proteins.
The effect these changes had on envelope function and the
observation that only structured peptides modeling this region
of gp4l inhibited virus replication lead us to conclude that the
structural components found in this region of the transmembrane protein play a critical role in the events which culminate
in virus entry. These observations have allowed us to develop
a model which proposes a functional role for the coiled-coil
region of gp4l as well as a mechanism by which peptides
mimicking this domain inhibit virus infection.
For HIV-1 it has been established that following the
interaction of gpl20 with CD4, the viral envelope undergoes
conformational changes which expose previously inaccessible regions of gp4l (33). Work by several groups (34, 35)
indicates that one of these epitopes corresponds to the
carboxyl terminus of the DP-107 sequence. We propose that
prior to CD4 binding, the coiled-coil region of gp4l is
constrained in a way that precludes its functional role in virus
entry. However, once binding has occurred, conformational
changes take place which cause residues 553-590 to undergo
the transition from a monomeric, non-coiled-coil domain to a
multimeric, coiled-coil structure (Fig. 6). Our results indicate
that coiled-coil formation is a critical component of virus
entry and very likely mediates the formation of the fusogenic
viral complex. A similar process has been proposed to take
place within the HA2 protein of influenza virus (20). While it
is conceivable that envelope binding to the CD4 receptor
alone might trigger the necessary conformational changes, it
is likely that other biochemical processes are necessary.
Further, we propose that it is during this entry event that the
DP-107 peptide and its structural analogs exert their inhibitory effect and that this inhibition is related to the ability of
the synthetic peptide to accurately mimic and interact (in a
coiled-coil structure) with the complementary region of the
virus transmembrane protein (Fig. 6). This aberrant interaction would give rise to a gp4l/peptide hybrid structure that
12680
Biochemistry: Wild et al.
Proc. Natl. Acad. Sci. USA 91 (1994)
Target Cell
"Trigger
Event"
i .',..
'..',.5.,
i
( Coiled-coil
precursor
....
region
Fusion
Fusion
gp
domain
CInfcted
ldomain
Cell
lacks the requisite components to successfully complete the
fusion process.
Although the specific function of the HIV-1 coiled-coil
structure is unknown, several roles are possible. By analogy
to influenza virus, coiled-coil formation might serve as a
driving force to position the fusion domain of the transmembrane protein for insertion into the target cell membrane (20).
A second possible function involves the participation of the
coiled-coil structure in the formation of a fusion pore. This
dynamic event (initiated by insertion of the fusion peptide
into the target cell membrane) may involve the continuous
recruitment of subunit components (the coiled-coil dimer
serving as a building block) until a functional fusion pore is
formed (12, 35). Amphipathic a-helical structures are known
to function as membrane-spanning domains in a variety of
systems (36). A third scenario involves a combination of the
roles described above, with coiled-coil formation serving as
the driving force for insertion of the fusion peptide into the
target cell membrane, with the resultant structure serving as
a building block in formation of a fusion pore. The heptad
repeat of hydrophobic amino acid residues which is predictive of coiled-coil structure has been observed in the surface
proteins of a number of enveloped viruses (12, 19, 35, 37).
While no direct evidence exists, we speculate that this
common structural feature might play a similar functional
role in diverse virus systems.
The series of events outlined above deals primarily with the
role of the leucine zipper domain of gp4l in the processes of
virus entry. We believe that this region serves an important
function in this event and might serve as a target for therapeutic intervention. It should be emphasized that the proposed model attempts to define the role of a single component
in what is undoubtedly a complex, multicomponent process.
We thank S. Roberts, L. Stoltenberg, and S. Hellenbrand for
technical assistance. This research was funded by grants from the
National Institutes of Health (5-ROI-AI30411 to T.M. and 5-R37AI-33319 to E.H.). HIV culture at the University of Alabama at
Birmingham was carried out in the Center for AIDS Research Central
Virus Culture Core Facility under Program Grant P30-AI27767 from
the National Institutes of Health. C.W. is a Scholar of the American
Foundation for AIDS Research (Grant 70036-14-RF).
1. Dalgleish, A. G., Beverly, P. C. L., Clapham, P. R., Crawford, D. H.,
Greaves, M. F. & Weiss, R. A. (1984) Nature (London) 312, 763-767.
2. Maddon, P., Dalgleish, A., McDougal, J. S., Clapham, P., Weiss, R. &
Axel, R. (1986) Cell 47, 333-348.
3. Sattentau, Q. J. & Weiss, R. A. (1988) Cell 52, 631-633.
FIG. 6. Transition proposed to
take place within the coiled-coil
:5 \ 11 1
~
~~~~~domain of gp4l on goinga from a
(Left) to fusioncompetent
state.
to
interaction (Right)
with CD4,
thePrior
coiledCoiIedcoil region
.lii.5tt
, .,., l,of gp4l is constrained
in such a way that its structural
domain
elements are not realized. Following CD4 binding the gpl20/gp4l
complex undergoes a series of
gp41
<
conformational changes which regp4l
/#
sult in the transition from random
coil to coiled coil and a reorganipInfeted Ce!lI
.zation of gp4l into a fusogenic
;..Vi
state.
.on
Qgpl2O5\} nOnnonfusogenic
4. Bosch, M. L., Earl, P. L., Fargnoli, K., Picciafuoco, S., Giombini, F.,
Wong-Staal, F. & Franchini, G. (1989) Science 244, 694-697.
5. Dalgleish, A. G., Chanh, T. C., Kennedy, R. C., Kanda, P., Clapham,
P. R. & Weiss, R. A. (1989) Virology 165, 209-215.
6. Earl, P. L., Koenig, S. & Moss, B. (1991) J. Virol. 65, 31-41.
7. Helseth, E., Olshevsky, U., Gabuzda, D., Ardman, B., Haseltine, W. &
Sodroski, J. (1990) J. Virol. 64, 6314-6318.
8. Kowalski, M. L., Bergeron, L., Dorfman, T., Haseltine, W. & Sodroski,
J. (1991) J. Virol. 65, 281-291.
9. Syu, W. J., Lee, W. R., Du, B., Yu, Q. C., Essex, M. & Lee, T. H.
(1991) J. Virol. 65, 6349-6352.
10. Delwart, E. L., Mosialos, G. & Gilmore, T. (1990) AIDS Res. Hum.
Retroviruses 6, 703-706.
11. Gallaher, W. R., Ball, J. M., Garry, R. F., Griffin, M. C. & Montelaro,
R. C. (1989) AIDS Res. Hum. Retroviruses 5, 431-440.
12. Chambers, P., Pringle, C. R. & Easton, A. J. (1990) J. Gen. Virol. 71,
3075-3080.
13. Guiltinan, M. J., Marcotte, W. R., Jr., & Quatrano, R. S. (1990) Science
250, 267-271.
14. Landschulz, W. H., Johnson, P. F. & McKnight, S. L. (1988) Science
240, 1759-1764.
15. Vinson, C. R., Sigler, P. B. & McKnight, S. L. (1989) Science 246,
911-916.
16. Earl, P. L., Doms, R. W. & Moss, B. (1990) Proc. Natl. Acad. Sci. USA
87, 648-652.
17. Pinter, A., Honnen, W. J., Tilley, S. A., Bona, C., Zaghouani, H.,
Gorny, M. K. & Zolla-Pazner, S. (1989) J. Virol. 63, 2674-2679.
18. Einfeld, D. & Hunter, E. (1988) Proc. Natl. Acad. Sci. USA 85,
8688-8692.
19. Buckland, R. & Wild, F. (1989) Nature (London) 338, 547 (lett.).
20. Carr, C. M. & Kim, P. S. (1993) Cell 73, 823-832.
21. Wild, C., Oas, T., McDanal, C., Bolognesi, D. & Matthews, T. (1992)
Proc. Natl. Acad. Sci. USA 89, 10537-10541.
22. Cao, J., Bergeron, L., Helseth, E., Thali, M., Repke, H. & Sodroski, J.
(1993) J. Virol. 67, 2747-2755.
23. Dubay, J., Roberts, S. J., Brody, B. & Hunter, E. (1992) J. Virol. 66,
4748-4756.
24. Chen, S.-L., Lee, C.-N., Lee, W.-R., McIntosh, K. & Lee, T.-H. (1993)
J. Virol. 67, 3615-3619.
25. Edlehoch, H. (1967) Biochemistry 6, 1948-1954.
26. Reed, L. J. & Muench, H. (1938) Am. J. Hyg. 27, 493-497.
27. Fisher, A. G., Ratner, L., Mitsuya, H., Marselle, L. M., Harper, M. E.,
Broder, S., Gallo, R. C. & Wong-Staal, F. (1986) Science 233, 655-659.
28. Maniatis, T., Fritsch, E. F. & Sambrook, J. (1982) Molecular Cloning: A
Laboratory Manual (Cold Spring Harbor Lab. Press, Plainview, NY).
29. Goff, S., Traktman, P. & Baltimore, D. (1981) J. Virol. 38, 239-248.
30. Willey, R. L., Smith, D. H., Lasky, L. A., Theodore, T. S., Earl, P. L.,
Moss, B., Capon, D. J. & Martin, M. A. (1988) J. Virol. 62, 139-147.
31. Myers, G., Korber, B., Wain-Hobson, S., Smith, R. F. & Pavlakis,
G. N. (1993) Human Retroviruses and AIDS 1993 (Los Alamos Natl.
Lab., Los Alamos, NM).
32. O'Shea, E. K., Rutkowski, R. & Kim, P. S. (1989) Science 243,538-542.
33. Sattentau, Q. J. & Moore, J. P. (1991) J. Exp. Med. 174, 407-415.
34. Sattentau, Q. J., Moore, J. P., Vignaux, F., Traincard, F. & Poignard, P.
(1993) J. Virol. 67, 7383-7393.
35. Buckland, R., Malvoisin, E., Beauverger, P. & Wild, F. (1992) J. Gen.
Virol. 73, 1703-1707.
36. Kono, K., Nishii, H. & Takagishi, T. (1993) Biochim. Biophys. Acta 1164,
81-90.
37. Britton, P. (1991) Nature (London) 353, 394 (lett.).