(Received for publication, April 7, 1997, and in revised form, May 23, 1997)
From the Laboratory of Viral Diseases, NIAID,
National Institutes of Health, Bethesda, Maryland 20892 and the
¶ Department of Medicine, University of Texas Health Science
Center at San Antonio and South Texas Veterans Health Care System, San
Antonio, Texas 78284-7881
There is a close correspondence between the
ability of RANTES and macrophage inflammatory proteins 1 and 1
to
activate CC chemokine receptor 5 (CCR5) and the ability to inhibit
CCR5-dependent membrane fusion mediated by the envelope
glycoprotein of human immunodeficiency virus (HIV), type 1. This
finding suggests that some of the structural determinants for CC
chemokine/CCR5 interactions and CCR5 HIV-1 fusion co-receptor activity
may be shared. Recent studies using human CCR5/CCR2B chimeras have
suggested that the determinants of CCR5 co-receptor activity are
complex and may involve multiple extracellular receptor domains and
that viral co-receptor activity is dissociable from
ligand-dependent signaling responses. However, conclusive
evidence demonstrating an important role for the second and third
extracellular regions of human CCR5 is lacking. Furthermore, to
determine whether the determinants for CCR5 co-receptor activity
overlap with those required for agonist activity, studies that compare
the chemokine specificity for inhibition of envelope-mediated cell
fusion and the agonist profile of chimeric receptors are necessary. In
the present report, using a series of CCR5/CCR2B chimeras we ascribe an
important role for the second and third extracellular loop of CCR5 in
supporting the co-receptor activity of CCR5. We also provide evidence
that the intracytoplasmic tail of CCR5 does not play an important role in supporting HIV-1 entry. The hypothesis that the structural determinants for CC chemokine/CCR5 interactions and CCR5 HIV-1 fusion
co-receptor activity may be shared was confirmed by two novel
observations: first, the fusion activity supported by two hybrid
receptors could be inhibited by both RANTES and monocyte chemoattractant protein-1, chemokines specific to CCR5 and CCR2B, respectively; and second, the chemokine specificity for inhibition of
envelope-mediated cell fusion matched the agonist profile of these
hybrid receptors. These data shed new light on the structural determinants involved in these distinct activities of CCR5 and may have
important implications for the development of CCR5-targeted anti-viral
compounds.
CC chemokine receptor 5 (CCR5),1 a
seven-transmembrane domain (TMD7) G-protein coupled receptor for the CC
chemokines macrophage inflammatory protein (MIP)-1, MIP-1
, and
RANTES (regulated upon activation, normal T cell expressed and
secreted) (1-3), plays a critical role in transmission and
pathogenesis of human immunodeficiency virus (HIV)-type 1 infection.
When co-expressed with CD4, CCR5 serves as a co-receptor for entry of
macrophage (M)-tropic and dual-tropic strains of HIV-1 (4-8). The
importance of CCR5 in HIV-1 transmission is highlighted by the findings
that individuals homozygous for a 32-base pair deletion in CCR5 have
greatly reduced susceptibility to HIV-1 infection; the protein encoded
by the defective CCR5 gene cannot be detected on the cell surface and is nonfunctional as a fusion co-receptor (9-13). Of note, these individuals do not have any detectable immunological defect, suggesting that a strategy designed to mimic a CCR5 null mutation may prove to be
a viable therapeutic approach.
The observation that MIP-1, MIP-1
, and RANTES activate CCR5
(1-3) and also suppress infection and fusion by M-tropic HIV-1 strains
(9-14) suggests that the suppressive effects of chemokines are exerted
(at least in part) by blocking sites on CCR5 involved in interaction
with the HIV-1 envelope (Env) glycoprotein (gp). In support of this,
soluble gp120 complexed to CD4 has been shown to inhibit binding of
radiolabeled chemokine to CCR5-expressing cells (15, 16). Furthermore,
it is possible some of the HIV-inhibitory effects of the CC chemokines
may be due to down-modulation of CCR5 from the cell surface. Thus,
identification of CCR5-HIV-1 interaction sites and knowledge of whether
they overlap with those required for chemokine interaction are critical
for understanding HIV-1 transmission and pathogenesis and may help
guide the design of novel anti-HIV-1 compounds that target this
interaction.
To examine the molecular determinants of human CCR5 that are important in supporting HIV-1 entry, recent studies (17-19) have examined the co-receptor activity of chimeric molecules created by exchanging corresponding regions between CCR5 and CCR2B, a TMD7 receptor for monocyte chemoattractant protein (MCP)-1 and MCP-3 that shares 72% sequence identity with CCR5 but does not function as a co-receptor for M-tropic strains (although it does function for dual-tropic strains such as 89.6; Refs. 4, 5, and 8). Both Rucker et al. (17) and Atchison et al. (18) have shown that the NH2-terminal domain of CCR5 can confer M-tropic co-receptor activity to CCR2B. Using chimeras in which the NH2 terminus or any of the three extracellular loops (designated e1, e2, and e3) of CCR5 were individually substituted with the corresponding regions of CCR2B, these molecules were reported to have co-receptor activity similar to wild type CCR5 (17, 18). However, additional data demonstrate that HIV-1 fusion activity is sensitive to modifications of the CCR5 NH2 terminus (17, 19). These studies have not demonstrated directly the importance of the e2 and e3 loop or the intracytoplasmic tail of CCR5 in supporting viral entry. Furthermore, to clarify whether the structural determinants necessary for chemokine signaling overlap with those required for HIV-1 entry, it is important that chimeric receptors be tested in both signaling assays and assays that can analyze the ability of chemokines to inhibit co-receptor activity.
In the present report, we use chimeric human CCR5/CCR2B molecules to ascribe an important role for the e2 and e3 loops in supporting the co-receptor activity of CCR5. We also provide evidence that the intracytoplasmic tail of CCR5 does not play an important role in supporting HIV-1 entry. We describe two CCR5/CCR2B chimeras that display hybrid specificities of the parental receptors, in assays of both agonist-induced mobilization of intracellular calcium and inhibition of fusion co-receptor activity. These data shed new light on the structural determinants involved in these distinct activities of CCR5.
The coding regions of CCR5 and CCR2B
were amplified by polymerase chain reaction (PCR) from genomic DNA,
subcloned into the SmaI restriction site of a pBluescript
SKII+ vector (Stratagene) that had been modified by deleting the
sequences in the multiple cloning site that span from the
PstI to AccI restriction sites of pBluescript
SKII+. The conserved ClaI and EcoRI (see Fig. 1) restriction sites of CCR5 and CCR2B were used to construct the chimeras
(see Fig. 2). The chimeras are named according to the origin of their
four extracellular segments. For example, chimera 5552 is an example
where the fourth extracellular domain (e3 loop) of CCR5 has been
exchanged with that from CCR2B. For HIV-1 Env-mediated cell fusion
assays using the vaccinia-based system, the receptor constructs were
transferred into the NotI and XhoI sites of
pBluescript KSII+ so that the translation initiation sites are adjacent
to the T7 promoter; expression was achieved by vaccinia-encoded
bacteriophage T7 RNA polymerase (see below). Point mutants were
introduced into the e3 loop of chimera 5552 by 1) digesting chimera
5552 in pBluescript KSII+ with restriction endonucleases
NotI and HincII; 2) using PCR with the T3 primer
in conjunction with a primer that included an EcoRI linker
and the sequence encoding the desired mutation(s). The PCR template was
the 5552 pBluescript KSII+ plasmid that had been digested with
NotI and HincII; and 3) digesting the amplified PCR product with the restriction endonucleases EcoRI and
XhoI and then ligating it into a CCR5-pBluescript KSII+
plasmid that had also been digested with EcoRI and
XhoI. The fidelity of all the receptor constructs was
verified by DNA sequencing.
Creation of Cell Lines Stably Expressing Wild Type and Chimeric Human CCR5 and CCR2B Receptors
Receptor DNA was transferred into the NotI and XhoI sites of the hygromycin-selectable, stable episomal vector pCEP4 (Invitrogen). Human embryonic kidney (HEK) 293 cells (107) grown to log phase in Dulbecco's modified Eagle's medium with 10% fetal bovine serum were electroporated with 20 µg of plasmid DNA. Multiple hygromycin-resistant colonies were picked and expanded in 150 µg/ml of hygromycin (Calbiochem).
Intracellular [Ca2+] MeasurementsHEK 293 transfectants (107/ml) were suspended in Hanks' buffered
saline solution with Ca2+ and Mg2+ and 10 mM Hepes, pH 7.4, containing 2.5 µM Fura-2
for 30 min at 37 °C in the dark. The cells were subsequently washed
twice in phosphate-buffered saline (PBS) and then resuspended in
Hanks' buffered saline solution with Ca2+ and
Mg2+ at 2 × 106 cells/ml. 2 ml of the
cell suspension were placed in a continuously stirred cuvette
maintained in a fluorimeter F4500 (Hitachi).
[Ca2+]i changes elicited at varying
concentrations of human CC chemokines (R&D) from transfected HEK 293 cells was determined. Fluorescence was monitored at
ex1 = 340 nm,
ex2 = 380 nm, and
em = 510 nm, and the data are presented as the
relative ratio of fluorescence at 340 and 380 nm.
HeLa, BSC-1, and NIH 3T3 cell lines (ATCC) were cultured in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum. Recombinant vaccinia viruses were grown in BSC-1 cells, and purified stocks were prepared by standard procedures (20).
Assays of HIV Fusion Co-receptor ActivityCo-receptor
activity was determined by a vaccinia-based reporter gene assay
quantitating Env-mediated cell fusion (21). To prepare target cells,
plasmid DNA (in pBluescript KSII+) containing the insert sequences
encoding wild type CCR5, CCR2B, or chimeric/mutant constructs linked to
the T7 promoter were transfected into NIH 3T3 cells using DOTAP
lipofection reagent (Boehringer Mannheim). After 4 h of incubation
at 37 °C, the transfected cells were coinfected with the vTF7-3
encoding bacteriophage T7 RNA polymerase under the control of a
natural vaccinia virus early/late promoter (22) and vCB-3 (23) encoding
human CD4 under control of a synthetic strong early/strong late
vaccinia promoter; the multiplicity of infection was 10 plaque-forming
unit/cell for each virus. The control in each of these experiments
represented NIH 3T3 cells transfected with the vector lacking inserts
and infected with the same recombinant vaccinia viruses. For
preparation of effector cells expressing HIV-1 Env, HeLa cells were
co-infected with the recombinant vaccinia virus vCB-21R encoding
-galactosidase under control of the T7 promoter (24) and one of the
following vaccinia recombinants containing the indicated Env linked to
the synthetic strong early/strong late promoter (25): vCB-43, Ba-L
(M-tropic); vCB-39 ADA (M-tropic); vCB32, SF-162 (M-tropic); vCB-28,
JR-FL (M-tropic); and vCB-16, IIIB uncleaved (Unc, rendered
nonfusogenic by deletion of the gp120/gp41 cleavage site). For the
dual-tropic 89.6 Env, the gene was cloned into plasmid pSC59 containing
the synthetic strong early/strong late vaccinia
promoter2 and was expressed
on HeLa cells by transfection with DOTAP and infection with vCB-21R.
All populations of cells (effectors and targets) were incubated
overnight at 31 °C to allow expression of the vaccinia-encoded
proteins, then washed, and resuspended in Earle's modified Eagle's
medium with 2.5% fetal bovine serum. Duplicate samples containing
105 target cells (expressing CD4 plus the indicated
co-receptors) and 105 effector cells (expressing Env) were
mixed in a 96-well microtiter plate in the presence of 40 µg/ml of
cytosine arabinoside (Ara-C) and incubated at 37 °C for 2 h.
The
-galactosidase produced as a result of cell fusion between
effector and target cells was quantitated by a colorimetric assay of
detergent-treated cell lysates, using a 96-well plate
spectrophotometer.
Rabbit polyclonal antiserum was raised against a
synthetic peptide corresponding to the first 24 amino acids of the
NH2 terminus of CCR5 (1-3) chemically coupled to keyhole
limpet hemacyanin. Cell surface expression of CCR5 or mutant constructs
containing the NH2 terminus of CCR5 was examined using the
same transfected/infected 3T3 cells that were used as CD4+ targets in
the fusion assay. Cells were resuspended in PBS containing 0.1% bovine
serum albumin. Each sample was incubated with the anti-CCR5 antiserum
at a 1:50 dilution for 45 min at room temperature. Cells were washed
with PBS and incubated with a 1:10 dilution of fluorescein
isothiocyanate-conjugated F(ab)2 goat anti-rabbit IgG (Boehringer
Mannheim) for 30 min at room temperature. Stained cells were washed,
resuspended in PBS, and immediately analyzed using a flow cytometer
(Becton Dickenson). As a control for background staining, samples of
the same cell population were stained using the prebleed rabbit serum.
To further control for nonspecific staining, vector-expressing cells
were stained with anti-CCR5 serum as well as prebleed serum; no
difference was observed in the fluorescence intensity between these two
samples.
To analyze the effect of chemokines on cell fusion activity, the CC chemokines RANTES or MCP-1 (R&D) were added at a final concentration of 50 or 500 nM to the target cells expressing CD4 plus the indicated chemokine receptor constructs; the cells were incubated for 30-45 min at 37 °C and then mixed with Env-expressing effector cells. To monitor the effect of chemokines on the vigorous co-receptor activity of wild type CCR5, HeLa cells were coinfected with vCB-21R and HIV-1 Env-encoding vaccinia virus in the presence of Ara-C (40 µg/ml). The drug inhibits vaccinia virus late gene expression and therefore reduces the expression of Env on HeLa cell, thereby reducing the fusion activity and facilitating measurement of inhibitory effects of chemokines.
In the vaccinia-based cell fusion assay, CCR5 functioned as a fusion co-receptor for Envs from several macrophage-tropic HIV-1 strains, as well as the dual-tropic strain 89.6; among the Envs tested, CCR2B functioned only for 89.6 (see Fig. 2). These results are consistent with previous reports using a variety of experimental systems (4-8).
To study the molecular determinants of CCR5 that support HIV-1 fusion co-receptor activity, we exploited the fact that most of the divergent regions between CCR5 and CCR2B are clustered in the extracellular amino acids (the NH2-terminal segment before TMD1 and the e1, e2, and e3 loops) and the carboxyl-terminal cytoplasmic tail (Fig. 1). The contributions of these divergent regions to CCR5 co-receptor activity were assessed by creating a series of chimeras in which the following regions of CCR5 were replaced with the corresponding regions of CCR2B (Fig. 1): 1) the NH2 terminus to the end of TMD3, including the e1 loop (chimera 2255); 2) the second intracellular loop to the fifth amino acid in the e3 loop, including the e2 loop (chimera 5525); 3) the sixth amino acid in the e3 loop to the end of the carboxyl-terminal cytoplasmic tail (chimera 5552). Because the first five amino acids of the e3 loop of CCR5 and CCR2B are identical, chimera 5552 is suited to test the importance of the e3 loop of CCR5 in supporting HIV-1 entry. We also created a complementary set of chimeras in which these regions of CCR2B were replaced with the corresponding regions of CCR5 (5522, 2252, and 2225). These chimeras are identical to a set of chimeras previously reported by Rucker et al. (17), who also used the conserved EcoRI and ClaI restriction endonuclease sites to switch homologous regions between CCR5 and CCR2B.
Fig. 2 shows that these CCR5/CCR2B
chimeras exhibited a 75-100% decrease in co-receptor activity for
Envs from the M-tropic strains. The fusion co-receptor defects of the
chimeric molecules were not due to loss of cell surface expression, as
judged by one direct (flow cytometry) and two indirect criterias.
First, all chimeras except 5522 exhibited some co-receptor activity for the 89.6 Env (Fig. 2). Second, flow cytometry analysis using an antiserum directed against the CCR5 NH2-terminal
extracellular region demonstrated that all chimeras containing this
region were expressed at levels comparable with wild type CCR5 (Fig. 2
legend). Finally, the three receptor constructs that include the
NH2 terminus of CCR2B (2255, 2225, and 2252) all mobilized
calcium in response to one or more CC chemokines (summarized in Fig. 2
and illustrated in detail in Fig. 4). Our results therefore suggest
that the fusion co-receptor activity of CCR5 involves several regions,
including the NH2 terminus and/or the e1 loop, the e2 loop,
and the e3 loop.
Our results show many parallels with previous studies of CCR5/CCR2B
chimeras, in which the structural determinants on CCR5 involved in
fusion co-receptor activity were found to be complex and to involve
multiple extracellular regions (17, 18). There is good agreement on the
loss of M-tropic co-receptor activity with chimeras 5522, 2255, 2252, and 2225. However, there are some significant differences between our
findings and the previous studies. In particular, we observed
considerable loss of M-tropic fusion co-receptor activity for chimeras
5525 and 5552, suggesting importance for the e2 loop and, to a greater
extent, the e3 loop; these results are in contrast to the report of
minimal effects of such changes on fusion co-receptor activity with a
M-tropic Env (17). Atchinson et al. (18) have also analyzed
human-murine CCR5 chimeras; however, because murine and human CCR5 have
identical sequences in the e3 loop, chimeras between murine/human CCR5
cannot reveal the importance for this region in viral entry.
Discrepancies have also been reported regarding the CCR5 extracellular
NH2-terminal region; although the importance of this domain
has been clearly demonstrated by analysis of chimeras and site-directed
mutants (17-19), varying results have been reported for the activities of certain chimeras with similar structures. Presumably differences in
the alternative expression and assay systems account for these discrepancies. For example, the target cells used by Rucker et al. (17) were quail QT6, whereas we used NIH3T3 cells; the gene reporter assay systems are also different (luciferase versus
-galactosidase). Atchison et al. (18) quantitated HIV-1
entry into COS cells transiently co-transfected with CD4 and chemokine
receptors by measuring intracellular expression of the viral capsid
protein p24 by flourescence-activated cell sorting. The assay
(chloramphenicol acetyltransferase activity) and cell expression system
(HeLa-CD4 and CF2Th canine thymocytes) used by Farzan et al.
(19) is also different from those used in this study.
In interpretating our finding of diminished M-tropic fusion co-receptor activity for chimera 5552, it must be noted that this molecule contains not only the e3 loop of CCR2B but also TMD7 and the carboxyl-terminal cytoplasmic tail. The residues in TMD7 of CCR5 and CCR2B are identical; however, in the the 51 residues of the carboxyl-terminal cytoplasmic tail of CCR5, 20 differences exist between the two receptors, of which only three are conservative (Fig. 1). The possibility must therefore be considered that these differences contribute to the reduction of M-tropic co-receptor activity in chimera 5552.
It is striking that a chimera containing the NH2 terminus of CCR2B and the extracellular loops of CCR5 (chimera 5222) can fully support HIV-1 entry (17-19), whereas chimera 5522 (this study and Ref. 17) does not. Conceivably, despite the high degree of sequence homology between CCR5 and CCR2B, "artificial" receptors created by exchanging different regions of CCR5 and CCR2B may have unusual folding patterns that could account for the unpredictable co-receptor activities that we and others have observed.
Importance of the CCR5 e3 Loop for M-tropic Fusion Co-receptor ActivityIn comparing the extracellular domains of CCR5 and
CCR2B, the e3 loop is one of the most conserved. In this region only
six of the 23 residues of CCR5 are different from CCR2B, and of these one is a conserved substitution (Fig. 1); by contrast in the e2 loop,
19 of 30 residues are different. To examine the importance of
individual residues in the e3 loop of CCR5 for M-tropic co-receptor activity, we created site-directed mutants of chimera 5552 in which the
CCR2B-specific amino acids in this region were replaced with those
present at the corresponding positions in CCR5 (Fig. 1). For example,
in chimera 5552:2,5 we changed the CCR2B-specific residues Glu and
Gln to the CCR5-specific residues Ser and Arg, respectively. Flow
cytometry analysis verified that each of the mutants was expressed at
the cell surface at levels comparable with wild type CCR5 (see Fig.
3 legend). The effects of these restorative mutations on fusion cofactor activity are shown in Fig. 3
for three M-tropic Envs. Substituting the amino acid Glu
Ser
(chimera 5552:
2) had no effect on M-tropic co-receptor function,
whereas substituting amino acid Gln
Arg (5552:
5) resulted in
modest restoration of activity. Although chimera 5552:
2,5 contains
the amino acid substitutions present in both chimeras 5552:
2 and
5552:
5, this chimera also had a minimal effect on restoring the
co-receptor activity of CCR5 (Fig. 3). We also created a construct
(chimera 5552:
1-5) in which the the sequence is entirely that of
CCR5 from amino acids 1-278. Nearly complete restoration of M-tropic
fusion co-receptor activity was observed; indeed in some experiments
complete restoration of fusion co-receptor activity was observed (data
not shown). Because chimera 5552:
1-5 has the carboxyl-terminal tail
of CCR2B, the differences in this intracytoplasmic region must not be
significant for co-receptor function. We conclude that amino acid
residues in the e3 loop of CCR5 that differ from those of CCR2B make
important contributions to CCR5's fusion co-receptor activity for
M-tropic Envs. It is conceivable that additional residues in the e3
loop of CCR5 may also play an important role in supporting HIV-1
entry.
Agonist Selectivity of CCR5/CCR2B Chimeras
We next examined
the structural determinants on CCR5 involved in chemokine agonist
activity. Using transfectants of the HEK293 cell line, the CCR5/CCR2B
chimeras were analyzed for their ability to mobilize calcium in
response to chemokine ligands (Figs. 2 and
4). As previously reported, the wild type
receptors CCR5 (1-3) and CCR2B (26-28) were activated in response to
distinct nonoverlapping sets of agonists. Thus MIP-1, MIP-1
and
RANTES potently activated CCR5, whereas MCP-1 and MCP-3 were inactive
(Figs. 2 and 4A); MCP-1 and MCP-3 were potent activators of
CCR2B, whereas MIP-1
, MIP-1
, and RANTES were inactive (Figs. 2
and 4B and data not shown). Chimeras 2252 and 2255 demonstrated hybrid agonist specificities compared with the wild type
receptors. Thus MIP-1
, RANTES, MCP-1, and MCP-3 were potent agonists
for chimeras 2252, whereas the signaling response with MIP-1
was
weak (Figs. 2 and 4C). MIP-1
, MIP-1
, and RANTES were
potent agonists for chimera 2255, and MCP-1 and MCP-3 showed activity
at higher concentrations (Figs. 2 and 4D). Chimera 2225 had
a chemokine selectivity similar to that of wild type CCR2B; MCP-1 was
active, whereas MIP-1
, MIP-1
, and RANTES were not (Figs. 2 and
4E). With chimeras 5552, 5522, and 5525 we were unable to
detect intracellular calcium mobilization with any of the CC chemokines
(Fig. 2). We note that hybrid agonist selectivities (MCP-1 and
MIP-1
) have been reported for chimera 2255 (Ref. 18) as well as for
another chimera containing the NH2 terminus of CCR2B and
the extracellular loops of CCR5 (19).
The hybrid agonist specificities observed for
chimeras 2252 and 2255 suggested the importance of testing the effects
of CC chemokines on fusion co-receptor activity; such analyses have not
been reported in previous studies of CCR5/CCR2B chimeras. Fig.
5 shows results with the ADA Env, which
displayed the greatest activity of the M-tropic Envs for these
chimeras. With wild type CCR5, dose-dependent inhibition of
fusion was observed with RANTES but not with MCP-1, as expected (Fig.
5A). Consistent with the hybrid agonist activities described
above, fusion mediated by chimeras 2252 and 2255 was inhibited in
dose-dependent fashion by both RANTES and MCP-1 (Fig. 5,
B and C). These results suggest that at least
some common structural determinants of CCR5 are involved in both
agonist activity and co-receptor function.
CCR5 ligands therefore probably exert their suppressive effects on HIV-1 fusion and entry by interfering with Env-CCR5 interactions. This notion is supported by the direct demonstration of gp120 blocking of radiolabeled chemokines to CCR5-expressing cells (15, 16). It is also possible that CC chemokines trigger down-regulation of CCR5 (as has been shown for other chemokine/receptor interactions (29)), rendering it unavailable for use as a fusion co-receptor. It should be emphasized that our studies do not resolve the questions of whether HIV-1-CCR5 interactions result in receptor activation and whether such activation is essential for viral fusion. Recent studies argue against the latter requirement (19). Our findings do imply that antagonists capable of blocking ligand interactions with fusion co-receptors may also block HIV-1 entry; however, because the determinants for HIV-1 fusion and receptor signaling are not totally convergent, a narrower spectrum of antagonists that specifically block CCR5-Env interactions can be envisioned.
In summary, our results extend previous reports (17-19) illustrating the complex nature of the HIV-1-Env interaction sites on CCR5 and show that these sites may overlap with those required for CC chemokine-mediated signaling. We provide evidence that multiple regions of CCR5 influence agonist selectivity and that these selectivity determinants overlap with those involved in fusion co-receptor activity. We ascribe an important role for the e2 and e3 loop for CCR5 co-receptor function. Additional studies will be required to develop a more detailed model of the HIV-1 Env and CC chemokine interaction sites on CCR5, including studies with the recently described CCR5 antagonists (30)
We thank R. A. Clark for helpful discussions and continuous support and R. A. Clark, E. Lolis, and J. Allan for critical reading of the manuscript. We thank John Yewdell, Randy Brutkiewicz, Gustav Russ, and Shiv A. Prasad for assistance in flow cytometry studies.