www.elsevier.com/locate/yviro

Virology 321 (2004) 274–286

Mechanism of feline immunodeficiency virus envelope

glycoprotein-mediated fusion

Himanshu Garg, Frederick J. Fuller, and Wayne A.F. Tompkins*

Department of Population Health and Pathobiology, College of Veterinary Medicine, North Carolina State University, Raleigh, NC 27606, USA

Received 13 October 2003; returned to author for revision 5 January 2004; accepted 7 January 2004

Abstract

Feline immunodeficiency virus (FIV) shares remarkable homology to primate lentiviruses, human immunodeficiency virus (HIV) and

simian immunodeficiency virus (SIV). The process of lentiviral env glycoprotein-mediated fusion of membranes is essential for viral entry

and syncytia formation. A detailed understanding of this phenomenon has helped identify new targets for antiviral drug development. Using a

model based on syncytia formation between FIV env-expressing cells and a feline CD4+ T cell line we have studied the mechanism of FIV

env-mediated fusion. Using this model we show that FIV env-mediated fusion mechanism and kinetics are similar to HIV env. Syncytia

formation could be blocked by CXCR4 antagonist AMD3100, establishing the importance of this receptor in FIV gp120 binding.

Interestingly, CXCR4 alone was not sufficient to allow fusion by a primary isolate of FIV, as env glycoprotein from FIV-NCSU1 failed to

induce syncytia in several feline cell lines expressing CXCR4. Syncytia formation could be inhibited at a post-CXCR4 binding step by

synthetic peptide T1971, which inhibits interaction of heptad repeat regions of gp41 and formation of the hairpin structure. Finally, using site-

directed mutagenesis, we also show that a conserved tryptophan-rich region in the membrane proximal ectodomain of gp41 is critical for

fusion, possibly at steps post hairpin structure formation.

D 2004 Elsevier Inc. All rights reserved.

Keywords: Feline immunodeficiency virus; Glycoprotein-mediated fusion; Lentiviruses

Introduction

Feline immunodeficiency virus (FIV) (Pedersen et al.,

1987) and human immunodeficiency virus (HIV) are

lentiviruses that share significant homology in their ge-

nomic organization and cause remarkably similar disease

in their respective hosts. Both infections are characterized

by a progressive depletion of CD4+ T cells leading to

immunodeficiency. The mechanism(s) mediating CD4+ T

cell loss has yet to be elucidated, but productive infection

does not appear to be the only cause (Alimonti et al.,

2003). Among other proposed mechanisms, env-mediated

cell fusion and death have been implicated in CD4+ T cell

depletion (Ferri et al., 2000). Enveloped viruses like HIV

and FIV utilize the env glycoprotein on their surface to

0042-6822/$ - see front matter D 2004 Elsevier Inc. All rights reserved.

doi:10.1016/j.virol.2004.01.006

* Corresponding author. Department of Population Health and

Pathobiology, College of Veterinary Medicine, North Carolina State

University, 4700 Hillsborough Street, Raleigh, NC 27606. Fax: +1-919-

515-4237.

E-mail address: Wayne_Tompkins@ncsu.edu (W.A.F. Tompkins).

mediate fusion between viral and cellular membranes. This

process is critical for viral entry and infection, and also

results in the formation of syncytia that is often seen in

vitro (Ferri et al., 2000). Models exploiting the potential of

HIV env glycoprotein to mediate syncytia formation have

been used to study the interaction between env glycopro-

tein and cell surface receptors on target cells (Gallo et al.,

2003; Jones et al., 1998), kinetics of env-mediated fusion

(Reeves et al., 2002), molecular determinants of the

fusogenic property of env glycoprotein (Salzwedel et al.,

1999), as well as designing and testing various inhibitors

of viral entry (Lawless et al., 1996).

The process of env-mediated fusion itself is complex,

involving several receptor ligand interactions and confor-

mational changes in the env glycoprotein. HIV env glyco-

protein is synthesized as a gp160 precursor and later

proteolytically cleaved to the gp120 surface unit and the

gp41 transmembrane protein (Chen, 1996; Wyatt and

Sodroski, 1998). Gp120 and gp41 are held together by

non-covalent interactions, and the presence of both proteins

is required to mediate fusion (Salzwedel et al., 1993). HIV

gp120 binds CD4 and a chemokine receptor, CCR5 or

H. Garg et al. / Virology 321 (2004) 274–286 275

CXCR4, on the surface of target cells (Choe et al., 1998),

while gp41 mediates fusion (Hart et al., 1991). HIV env-

mediated fusion has been studied extensively, and various

steps involved in this process have been identified (Gallo

et al., 2003). The fusion process is initiated when gp120

binds CD4 resulting in a conformational change in gp120

that allows co-receptor (CCR5/CXCR4) binding. Subse-

quent conformational changes in gp120 allow insertion of

the hydrophobic fusion domain of gp41 into the target

cell and exposure of the heptad repeat-1 (HR1) and

heptad repeat-2 (HR2) coiled domains of gp41 (Jones

et al., 1998). The gp41 molecule on the membrane

surface forms highly stable trimeric complexes of HR1

and HR2 running in antiparallel fashion (Lu et al., 1995).

An interesting phenomenon first reported for influenza

virus (Carr and Kim, 1993) is the interaction of trimeric

HR1 and HR2 coiled domains in a zipper-like fashion

(Wild et al., 1994b) to form the hairpin structure or the

six-helix bundle that brings the viral and cell membrane

into close contact resulting in fusion. This phenomenon

remains true for a variety of enveloped viruses including

HIV (Melikyan et al., 2000) and FIV (Medinas et al.,

2002). The importance of this coiling event in the

infection process is underscored by the recent demonstra-

tion that peptides corresponding to the HR1 and HR2

coiled domains are potent inhibitors of HIV env-mediated

fusion (Kilby et al., 1998; Wild et al., 1994a). Similar

peptides capable of inhibiting FIV fusion and entry have

been identified for FIV (Medinas et al., 2002).

Although a detailed understanding of HIV env-mediat-

ed fusion has been established, certain aspects of this

process such as the events post hairpin structure formation

are not clearly defined. In this context, a region of HIV

gp41 corresponding to the membrane proximal ectodo-

main also called the pre-transmembrane region (pre-TM)

has gained importance in the fusion process. This region

is known to contain highly conserved tryptophans in

various lentiviruses including FIV. Mutational analysis

shows that tryptophans in the pre-TM region are critical

Fig. 1. Schematic diagram of FIV gene orientation and primers used for clonin

rev (pFIVenv/rev) was cloned into pcDNA3 vector under the control of the

site and reverse primers (primers 2 and 3) had an EcoR1 overhang. The

pcDNA3 vector.

for HIV env-mediated fusion (Salzwedel et al., 1999),

suggesting that it may be a second fusion domain in HIV

(Suarez et al., 2000) and play a critical role in env-

mediated fusion at a post hairpin structure stage (Saez-

Cirion et al., 2002). In support of this, Giannecchini et al.

(2003) recently identified an octapeptide spanning the pre-

TM region in FIV that has anti-viral activity, most likely

by inhibiting fusion.

Although it has been shown that FIV env does not bind

CD4 (Hosie et al., 1993; Willett et al., 1997), an interesting

finding is that FIV and HIV share the chemokine receptor

CXCR4 for viral entry and syncytia formation (Richardson

et al., 1999). The laboratory-adapted Petaluma strain of FIV

(FIV-pet) can utilize either feline or human CXCR4 for

mediating fusion (Willett et al., 1997). Based on this,

several previous studies (Medinas et al., 2002; Willett et

al., 1998) have utilized Crandell feline kidney (CrFK) cells,

chronically infected FIV-pet (CrFKpet), and CXCR4

expressing HeLa cells as a model for studying FIV env-

mediated fusion. As most primary isolates of FIV do not

infect CrFK cells, the ability of FIV env from various

isolates to use CXCR4 as a receptor for viral entry or

syncytia formation remains controversial. Although it is

widely accepted that the laboratory adapted FIV-pet utilizes

CXCR4 to induce fusion in a variety of cells, including the

human cell line HeLa, the results with other primary

isolates have been varied (De Parseval and Elder, 2001;

Willett et al., 2002), suggesting the requirement of yet to be

identified receptor–co-receptor.

The objective of the present study was to develop a

suitable model to study FIV env-mediated fusion by a

primary isolate of FIV in lymphocytic target cell lines. For

this purpose we have cloned and expressed the env glyco-

protein from a primary isolate of FIV (FIV-NCSU1) in CrFK

cells. Cells expressing env glycoprotein from FIV-NCSU1

(CrFKenv/rev) showed fusion with the IL-2-dependent

feline T cell line FCD4E cells. Using a quantitative assay

based on syncytia formation between CrFKenv/rev cells and

FCD4E cells, we were able to block env-mediated fusion at

g FIV env. FIV env gene either alone (pFIVenv) or in combination with

CMV promoter. The forward primer (primer1) had a BamH1 overhang

PCR products were directionally cloned into BamH1 EcoR1-digested

H. Garg et al. / Virology 321 (2004) 274–286276

various receptor-ligand interactions and conformational

changes, and demonstrated the role of CXCR4, gp41-coiled

domains, and conserved tryptophan-rich regions of gp41 in

FIV env-mediated fusion. This simple model can be used to

address numerous questions concerning FIV env interac-

tions with cell surface receptors on T cells and their role in

FIV cell entry and pathogenesis.

Results

FIV env gene expression is rev dependent

To study the role of env glycoprotein in FIV pathogen-

esis, we cloned and expressed the env gene from the

pathogenic NCSU1 primary isolate of FIV (Yang et al.,

1996). The expression of retroviral genes is complex and

Fig. 2. Expression of FIV env glycoprotein in transfected cells. (A)

Western blot showing FIV env expression. Total cellular extracts from

transfected CrFK cells were run on SDS page gel, blotted onto PVDF

membrane, and probed with SU1–30 mAb against FIV gp120, which also

reacts with unprocessed gp160 protein. Lane 1: CrFK infected with

Petaluma strain of FIV; lane 2: CrFK transfected with pFIVenv/rev; lane 3:

CrFK transfected with pFIVenv; lane 4: CrFK non-transfected. (B)

Immunocytochemistry of transfected CrFK cells using serum from an

FIV-infected cat. CrFK cells were transfected with either pFIVenv or

pFIVenv/rev and selected for neomycin resistance. Selected cells were

stained with serum from FIV+ cat followed by HRP-conjugated anti-cat

antibody and developed with AEC substrate.

Fig. 3. Membrane-expressed FIV env induces syncytia formation with

FCD4E cells. (A) FIV env-expressing CrFK (CrFKenv/rev) cells were co-

cultured with the feline CD4+ T cell line (FCD4E). At 24 h post co-culture,

the plates were fixed and stained with Geimsa stain. (B) Two-color

fluorescent dye redistribution assay showing FIV env-mediated fusion

between CMFDA-labeled CrFKenv/rev cells and CMTMR-labeled FCD4E

cells. Labeled cells were co-cultured for 24 h, following which plates were

fixed with 3.7% paraformaldehyde and analyzed by fluorescent microsco-

py. Fused syncytial plaques (indicated by arrows) are positive for both dyes.

(C) Fusion mediated by in vitro-expressed FIV env is linearly related to the

number of target cells (FCD4E) added. CrFKenv/rev cells at 2 � 104 cells/

well were co-cultured with FCD4E cells at 2-fold serial dilutions starting at

104 cells/well. Fusion was quantified by counting the number of syncytial

plaques formed per well 24 h post co-culture. Data are mean F standard

deviation of quadruplicate wells.

involves multiple splicing of viral RNA. Rev, a regulatory

protein, reported in both HIV and FIV has been shown to

function as a transporter of full-length and partially spliced

HIV RNA out of the nucleus (Hadzopoulou-Cladaras et al.,

1989; Hammarskjold et al., 1989; Phillips et al., 1992). To

determine a similar function of rev in FIV we developed

two clones, one containing the open reading frame (orf) of

the env gene alone (pFIVenv) and other included the env

gene along with the 3Vexon of the rev gene (pFIVenv/rev)

(Fig. 1). Transfection of CrFK cells with these constructs

revealed that expression of FIV env as determined by

Western blotting and immunocytochemistry could be

achieved with the pFIVenv/rev construct but not with

Fig. 4. FIV env-mediated fusion can be blocked at the level of the env

glycoprotein. (A) CrFKenv/rev cells were co-cultured with FCD4E cells.

Serum from FIV+ or FIV� cats was added 1 h before co-culture. Fusion

was measured 24 h later and expressed as percent control (no treatment).

Data are mean F standard deviation. (B) CrFKenv/rev cells were seeded in

96-well plates at 2 � 104 cells/well. FCD4E cells were added at 5 � 103/

well. Fusion inhibitor T1971 or control peptide T1566 with no inhibitory

activity was added in serial 2-fold dilutions 1 h before co-culture. Fusion

was estimated 24 h later as described above. Data are mean F standard

deviation of quadruplicate wells.

H. Garg et al. / Virology 321 (2004) 274–286 277

pFIVenv (Figs. 2A and 2B). Env gene transcripts were

detected in both pFIVenv- and pFIVenv/rev-transfected cells

by RT-PCR (data not shown), suggesting that the role of rev

in env gene expression is at a posttranscriptional level. The

requirement of rev is absolutely critical for FIV env expres-

sion, and this simple mechanism of providing rev from the

same plasmid can be used for efficient expression of FIV

env from eukaryotic expression vectors. Using this con-

struct we established a stable cell line expressing FIV-

NCSU1 env termed CrFKenv/rev that was used in all

subsequent studies.

Membrane expressed env glycoprotein causes syncytia

formation with a feline CD4+ T cell line

To assess the biological activity of the expressed env

glycoprotein, a fusion assay was developed using env-

expressing CrFKenv/rev cells and the IL-2-dependent feline

CD4+ T cell line (FCD4E). Co-culture of CrFKenv/rev

cells with FCD4E cells resulted in the formation of syncy-

tia (Fig. 3A). To confirm that the syncytia seen in the co-

culture were formed due to fusion between CrFKenv/rev

(effector) cells and FCD4E (target) cells, a two-color

fluorescent dye redistribution assay was performed. Co-

culture of differentially labeled effector and target cells

resulted in the formation of syncytial plaques that were

double positive for the fluorescent dyes, indicating that the

fusion was in fact between effector and target cell (Fig.

3B). For most other experiments, the non-fluorescent syn-

cytial plaque-forming assay was used. A two-fold serial

dilution of target cells in the assay showed a linear

relationship between the number of target cells added and

number of syncytia formed (Fig. 3C). Based on this

experiment, 5–2.5 � 103 target cells/well gave the most

reproducible results.

Blocking of FIV env-mediated fusion

To validate the specificity of env-mediated fusion, either

pooled serum from FIV-infected cats (FIV+ serum) or gp41-

specific fusion inhibitor T1971 was incorporated in the

assay as possible blocking agents. FIV+ serum (shown to

have antibodies to FIV env by immunocytochemistry in Fig.

2B) blocked env-mediated fusion in a dose-dependent

manner (Fig. 4A). The fusion event mediated by HIV and

FIV env involves interaction among coiled domains of gp41

at a post gp120 binding stage. Peptides corresponding to the

coiled domains of HIV and FIV gp41 have been shown to

be potent inhibitors of env-mediated fusion (Kilby et al.,

1998; Medinas et al., 2002). One of these inhibitors, T1971,

shown to have significant fusion inhibition against FIV in

an earlier study (Medinas et al., 2002) blocked syncytia

formation in the CrFKenv/rev-FCD4E syncytia forming

assay in a dose-dependent manner, while T1566, a peptide

derived from a region outside the coiled domains of FIV

gp41, had no inhibitory activity (Fig. 4B). This study

confirmed the biological activity and specificity of the

cloned env glycoprotein.

CXCR4 is involved in FIV env-mediated fusion of FCD4E

cells

It has been reported that FIV utilizes CXCR4 as a

receptor for viral entry and cell fusion (Hosie et al., 1998;

Willett et al., 1997). To determine if FIV-NCSU1 utilizes

CXCR4, antihuman CXCR4 antibodies were assessed for

their ability to block fusion. Antihuman CXCR4 monoclo-

nal antibodies (mAb) 12G5 and 44717 failed to block

CrFKenv/rev- or CrFKpet-induced fusion with FCD4E

(feline) cells up to a concentration of 30 Ag/ml, but

interestingly, both antibodies completely blocked fusion of

CrFKpet with HeLa (human) cells at 3 Ag/ml (data not

shown). Previous studies (Egberink et al., 1999; Hosie et al.,

1998) and our own observation demonstrated that mAb

44717 cross-reacts with feline CXCR4, whereas 12G5 fails

to bind feline cells. We confirmed these findings by flow

cytometric analysis of FCD4E cells (data not shown).

Failure of cross-reacting mAb 44717 to block fusion medi-

H. Garg et al. / Virology 321 (2004) 274–286278

ated by FIV env from both primary FIV-NCSU1 and lab-

adapted FIV-pet strain in feline cells may be due to weak

binding affinity or more likely due to failure to mask the

FIV env-binding site on the feline CXCR4 receptor. This is

supported by our observation that fusion mediated via feline

CXCR4 (by either NCSU1 or Petaluma) could not be

inhibited up to a concentration of 30 Ag/ml of mAb 44717

while the cell surface staining of feline cells was saturated at

5 Ag/ml. Also, fusion via human CXCR4 (CrFKpet-HeLa

model) was completely inhibited at 10-fold lower concen-

tration of 3 Ag/ml (data not shown).

To further explore a role for CXCR4 in FIV env-medi-

ated fusion, the CXCR4 antagonist AMD3100 that has been

shown to block infection of feline cells by various isolates

of FIV (Egberink et al., 1999) was used in the fusion assay.

Interestingly, AMD3100 efficiently blocked syncytia forma-

tion between FIV-NSCU1 env-expressing CrFKenv/rev cells

and FCD4E cells in a dose-dependent manner (Fig. 5A). To

determine that AMD3100 was in fact binding to feline

CXCR4, a competitive assay was done to block binding

of cross-reacting anti-CXCR4 mAb 44717 to FCD4E cells

by excess of AMD 3100 and analyzed by flow cytometry.

AMD3100 at 1 AM completely blocked binding of mAb

44717 to FCD4E cells (Fig. 5B). These results are in

accordance with previous reports of FIV env binding to

Fig. 5. CXCR4 is involved in FIV env-mediated cell fusion. (A) CrFKenv/rev cell

AMD3100 added at the indicated concentration 1 h before co-culture. The cells

analyzed for syncytia formation. Data are mean F standard deviation of quadrup

FCD4E cells. FCD4E cells were stained with either isotype control or anti-CXCR4

by flow cytometry.

CXCR4 and confirm that feline CXCR4 is involved in

binding and fusion mediated by FIV-NCSU1 env.

CXCR4 is required but not sufficient for fusion by a primary

isolate (NCSU1) of FIV

Having established that CXCR4 was involved in bind-

ing and fusion between CrFKenv/rev cells and FCD4E

cells, we assessed the ability of FIV-NCSU1 env to induce

syncytia formation with other CXCR4 expressing feline or

human cell lines. We tested FCD4E (CD4+ feline T cell

line), Fet J cells (CD8+ feline T cell line), 3201 cells

(CD4+CD8+ feline lymphoma cell line), and HeLa cells

(human epithelial cell line), all of which express CXCR4

on their surface, for their ability to fuse with CrFK cells

expressing either the FIV-NCSU1 env (CrFKenv/rev) or

chronically infected with FIV-pet (CrFKpet). Interestingly,

only FCD4E cells fused with CrFKenv/rev cells, while all

four cell lines fused with CrFKpet cells (Figs. 6A and 6B).

These results are consistent with the observation that of all

the cell lines tested, FIV-NCSU1 is capable of infecting

only FCD4E cells, whereas the cell culture-adapted FIV-

pet isolate has a wider tropism (Verschoor et al., 1995).

FACS analysis of CXCR4 expression revealed that FCD4E

cells, the only cell line capable of fusion with FIV-NCSU1

s were co-cultured with FCD4E cells in the presence of CXCR4 antagonist

were co-cultured for 24 h following which plates were fixed, stained, and

licate wells. (B) AMD3100 blocks binding of anti-CXCR4 mAb 44717 to

antibody 44717 in the presence or absence of 1 AMAMD3100 and analyzed

Fig. 6. CXCR4 is required but not sufficient to mediate fusion by primary isolate of FIV. Cells either expressing FIV-NCSU1 env (CrFKenv/rev) or infected

with FIV-pet (CrFKpet) were tested for their potential to fuse with various cell lines in a co-culture assay as described in Materials and methods. CrFKenv/rev

(A) or CrFKpet (B) cells were seeded at 5 � 104 cells/well; FCD4E, 3201, and Fet-J cells were added at 2-fold serial dilution starting at 104 cells/well and

incubated for 24 h following which the plates were fixed, stained, and photographed. For fusion with HeLa cells, HeLa cells were seeded at 104 cells followed

by 5 � 102 CrFKenv/rev or CrFKpet. Either 104 or 5 � 103 dilution wells shown for Fet-J, 3201, and FCD4E. (C) Flow cytometric analysis of CXCR4

expression on various cells lines. Cells were incubated with antihuman CXCR4 Ab 44717 or isotype control followed by FITC-conjugated secondary Ab and

analyzed by flow cytometry.

H. Garg et al. / Virology 321 (2004) 274–286 279

Fig. 7. Time dependent inhibition of fusion by AMD3100 and T1971.

CrFKenv/rev cells were co-cultured with FCD4E cells as described in

Materials and methods. Inhibitors AMD3100 at 1 AM and T197 at 0.3 AMwere added at the indicated time points post cell mixing starting at 0 h.

After a total incubation of 24 h, the plates were fixed and the number of

syncytial plaques counted. Data are mean F standard deviation of

quadruplicate wells (*P < 0.01 refers to comparison between AMD3100

and T1971).

Fig. 6 (continued).

H. Garg et al. / Virology 321 (2004) 274–286280

env, had the lowest CXCR4 expression (Fig. 6C), suggest-

ing that the level of CXCR4 expression was not a limiting

factor in FIV-NCSU1 env-mediated fusion. These findings

suggest that in contrast to the cell culture-adapted FIV-pet

isolate, the primary FIV isolate NCSU1 may utilize a

receptor–co-receptor in addition to CXCR4 for cell entry

and fusion, as has been suggested for other primary

isolates (De Parseval and Elder, 2001).

FIV env-mediated fusion is rapid and follows kinetics

similar to HIV env glycoprotein

To determine if the fusion process mediated by FIV env

follows similar kinetics as HIV env-mediated fusion, either

T1971 or AMD3100 was added at different time points post

co-culture. As shown in Fig. 7, the fusion process could be

completely blocked when T1971 or AMD3100 was added

up to 45 min post co-culture. At time points 1.5 and 2 h,

AMD3100 was significantly less effective at blocking

syncytia formation than T1971 (P < 0.01), suggesting that

T1971 acts at time points post CXCR4 binding. Our data

indicate that while some of the cells have completed fusion

as early as 1 h, it takes up to 4 h for all cells to complete

fusion under our assay conditions. The kinetics of fusion in

the presence of AMD3100 or T1971 were similar, suggest-

ing that the limiting factor in FIV env-mediated fusion was

gp120 binding to receptor–co-receptor and not gp41 hairpin

structure formation, which was rapid after CXCR4 engage-

H. Garg et al. / Virology 321 (2004) 274–286 281

ment (Fig. 7). Similar findings have been reported for HIV

(Gallo et al., 2001).

Conserved tryptophans in the pre-transmembrane region of

FIV gp41 are essential for fusion

Sequence analysis of HIV and FIV env glycoproteins

revealed significant homology in the gp41 region and not in

the gp120 region. A region of HIV gp41 N terminal to the

Fig. 8. The conserved tryptophan-rich region in gp41 is critical for fusion. (A) Schem

that is conserved in various lentiviruses and among various isolates of FIV. (B) Exp

which all three tryptophanswere replaced by alanine. CrFK cells were transiently tran

by Western blotting and fusion by a co-culture assay as described above.

membrane-spanning region called the pre-TM region has

been shown to have conserved tryptophans (W) that are

critical for membrane fusion (Salzwedel et al., 1999). Based

on the observation that the pre-TM region of FIV gp41 also

has conserved tryptophans (Fig. 8A), we used site-directed

mutagenesis to replace W at positions 766, 769, and 772

with Alanine (A). The mutant env/rev construct thus gen-

erated [W(1–3)A] was transfected into CrFK cells and

tested for its fusion-inducing ability with FCD4E cells. As

atic diagram of FIV gp41 showing the location of the tryptophan-rich region

ression and fusion potential of wild-type (WT) and mutant env W(1–3)A in

sfectedwithWTorW(1–3)A env constructs and analyzed for env expression

H. Garg et al. / Virology 321 (2004) 274–286282

shown in Fig. 8B, the W(1–3)A env showed similar levels

of expression and processing as the wild type (WT), but was

completely defective in forming syncytia with FCD4E cells.

As with HIV, this region appears to be critical for fusion by

FIV env and that the presence of tryptophans is required for

this property.

Discussion

The present study was undertaken to develop an in vitro

model to study interactions of membrane-expressed FIV env

with receptors on T cells that leads to env-mediated fusion.

We have used this model to study env receptor interactions

as well as kinetic analysis of FIV env-mediated fusion.

Several critical questions regarding FIV env expression

and function are answered in our study. Firstly, failure to

express FIVenv from a plasmid lacking 3Vrev-coding regionsuggests that the expression of FIV env, similar to HIV env,

is rev dependent. Rev is involved in the transport of

unspliced and partially spliced env mRNA from the nucleus

(Malim et al., 1989). Similar constructs have been used to

express HIV env glycoprotein in transfected cells (Ham-

marskjold et al., 1989). In some studies with HIV, rev has

been provided in trans from another plasmid to obtain env

expression (Moir and Poulin, 1996). In our system, provid-

ingly rev from the same plasmid is sufficient to get efficient

express env. Using this construct we were able to study the

various steps involved in FIV env-mediated fusion and

correlate it to findings with HIV.

The first step in HIV env-mediated fusion is the binding

of gp120 to CD4 and CXCR4 on T cells (Choe et al.,

1998). Although several studies have ruled out a role of

CD4 as a receptor for FIV (Hosie et al., 1993; Willett et al.,

1997), CXCR4 has been well established as a necessary

receptor or co-receptor for the laboratory-adapted Petaluma

isolate of FIV (Hosie et al., 1998; Richardson et al., 1999),

while results with primary isolates have been varied. In

support of this, a recent report by Willett et al. (2002)

shows that expression of feline CXCR4 alone was sufficient

for viral entry and fusion by laboratory-adapted FIV-pet,

but not the primary isolate FIV-Glasgow. Similarly, studies

by De Parseval and Elder (2001) have shown a yet

unidentified 40-kDa protein on T cells that bind recombi-

nant FIV gp120 from the primary isolate FIV-PPR. In the

same study, CXCR4 agonist SDF-1 failed to inhibit recom-

binant FIV-PPR gp120 binding to T cells. In both these

studies, the authors fail to show direct evidence that the env

glycoprotein from primary isolates bind to or utilize feline

CXCR4. Using CXCR4 antagonist AMD3100 we show

that CXCR4 is in fact a receptor (co-receptor) for FIV-

NCSU1 and plays a necessary role in FIV env-mediated cell

fusion by both primary- and laboratory-adapted strains of

FIV. Cross-reacting antihuman CXCR4 antibody failed to

block fusion in our model even though excess AMD3100

blocked binding of this antibody to FCD4E cells. We

believe that the failure to inhibit fusion via cross reacting

antibody is likely due to failure to mask the FIV env-

binding site on feline cells. This is supported by the report

of Willett et al. (1998) showing that the second extracellular

loop of CXCR4 that is required for FIV env binding

contains the majority of differences between feline and

human CXCR4.

Primary isolates of FIV, including NCSU1, neither infect

nor mediate fusion in CXCR4-expressing CrFK cells (Hoh-

datsu et al., 1996), which suggests that a second receptor

may be required for entry by these viruses. The differences

between FIV-pet and primary isolates have been largely

attributed to variation in the env gene sequence, more

specifically changes in the V3 loop of FIV-pet resulting in

a net positive charge in this region have been implicated in

the broader tropism and fusogenicity of FIV-Pet (Verschoor

et al., 1995). We show that while CrFK cells infected with

FIV-pet (CrFKpet) would fuse with a variety of cell lines

expressing CXCR4, such as Fet J, 3201, HeLa, and FCD4E

cells, cells expressing FIV-NCSU1 env (CrFKenv/rev) fused

only with FCD4E cells. This is again in agreement with the

tropism of NCSU1, which productively infects only FCD4E

cells or primary lymphocytes (English et al., 1993). This

suggests that though CXCR4 is required, it is not sufficient

for fusion by primary isolates of FIV and that another

receptor or co-receptor expressed on FCD4E cells may be

involved. Although FCD4E cells express CD4, it is not a

receptor for FIV as has been shown previously (Hosie et al.,

1993). Our findings confirm that FIV-NCSU1 binds feline

CXCR4, but expression of CXCR4 alone was not sufficient

to induce fusion and that another receptor–co-receptor may

be involved.

Following HIV gp120 binding to CD4/CXCR4, a con-

formational change in the env glycoprotein occurs that

results in insertion of the N terminal fusion domain of

gp41 into the target cell membrane and exposure of the

coiled domains HR-1 and HR-2 of gp41 (Wild et al.,

1994b). Interaction among the coiled domains of gp41 in

a zipper-like fashion (Wild et al., 1994b) brings the mem-

branes close together facilitating fusion. Lentiviral env

glycoprotein is present as a trimer on the membrane surface,

and hence interaction of the coiled domains results in

formation of a six-helix bundle structure. Peptides corr-

esponding to either the HR-1- or HR-2-coiled domains of

HIVor FIV gp41 have the unique property of inhibiting six-

helix bundle formation and thus viral entry (Wild et al.,

1994a). These peptides have evolved into a new generation

of anti-HIV drugs premiered by T-20 (Enfuvirtide) (Moyle,

2003). Gp41-specific peptide inhibitors are also powerful

tools to study env interactions post receptor–co-receptor

binding. We utilized an FIV-specific fusion inhibitor T1971,

previously shown to block CrFKpet fusion with HeLa cells

(Medinas et al., 2002), to block fusion in our model in a

dose-dependent manner. Inhibition by gp41-specific fusion

inhibitors is unique because it inhibits env response at a

post-CXCR4 binding stage (Gallo et al., 2001).

H. Garg et al. / Virology 321 (2004) 274–286 283

Utilizing CXCR4 antagonist AMD3100 and gp41-spe-

cific fusion inhibitor T1971, we studied the kinetics of FIV-

NCSU-1 env-mediated fusion by inhibiting this process at

different time points post co-culture. As shown in Fig. 7, we

see a lag phase of around 45 min in fusion mediated by FIV-

NCSU1, which is not too different from 20 to 30 min

reported for HIV (Melikyan et al., 2000). T1971 was more

efficient than AMD3100 at inhibiting fusion at time points

1.5 and 2 h. This correlates with the proposed sequence of

events involved in env-mediated fusion in that CXCR4

binding precedes and induces gp41 hairpin formation.

Nevertheless, the overall kinetics of fusion in the presence

of T1971 or AMD 3100 were similar, suggesting that the

limiting factor in FIV env-mediated fusion is not gp41

hairpin structure formation, which probably occurs rapidly

after receptor–co-receptor binding.

The events after hairpin formation that may be important

for fusion in HIV have not been well characterized. However,

a tryptophan-rich region of gp41 proximal to the membrane-

spanning domain termed pre-transmembrane region has

gained importance following a report by Salzwedel et al.

(1999) that conserved tryptophans in this region are critical

for fusion. Further, others have shown that this region is a

novel fusogenic domain (Suarez et al., 2000) and that it has

lectin-like properties and binds to sphingomyelin- and cho-

lesterol-rich structures in cell membranes (Saez-Cirion et al.,

2002). Sequence analysis showed that tryptophans in this

region were conserved among various lentiviruses and also

within various strains of FIV (Fig. 8A). We used site-directed

mutagenesis to replace tryptophans (W) in this region with

alanine (A) resulting in a fusion-defective mutant. We hy-

pothesize that the hydrophobic nature ofW in this region may

facilitate fusion at a step post six-helix bundle formation by

Fig. 9. Proposed sequence of events leading to FIV env-mediated fusion. Step1

receptor–co-receptor on the surface of target cells. Step 2: a conformational change

target membrane and exposure of the coiled domain. Step 3: the coiled domains int

4: the pre-TM region of gp41 may mediate mixing of the lipid components of the

unable to mediate fusion].

mediating mixing of lipid components of effector and target

membranes. Interestingly, Giannecchini et al. (2003) have

recently reported that an octapeptide spanning this region has

anti-FIV activity by inhibiting virus entry. The authors have

suggested that the peptide inhibits viral entry by binding to

components on the cell surface rather than any region in the

viral gp41 protein. They also found that tryptophans were

critical for the inhibitory activity of the peptides.

Using the CrFKenv/rev-FCD4E syncytia model, we have

delineated various steps that are involved in cell fusion and

syncytia formation by the env glycoprotein from FIV-

NCSU1 primary isolate (Fig. 9). The first step involves

gp120 binding to CXCR4 and possibly another receptor–

co-receptor on the cell surface. This is followed by confor-

mational changes in the env protein that result in exposure

of gp41 heptad repeat domains and six-helix bundle forma-

tion. Finally, our data indicate that the tryptophan-rich pre-

TM region has a critical role in FIV env-mediated fusion

probably at a step post six-helix bundle formation.

Materials and methods

Antibodies and reagents

Antihuman CXCR4 antibodies 12G5 and 44717 were

obtained from BD Biosciences (San Jose, CA). Anti-FIV

gp120 antibody SU1–30 was obtained from Custom Mono-

clonals (West Sacramento, CA). Gp41 fusion inhibitor

T1971 and control peptide T1566 were a kind gift from

Robyn Medinas (Trimeris Inc., Durham, NC). CXCR4

antagonist AMD3100 was a kind gift from Dr Edward

Hoover, (Colorado State University).

: FIV gp120 binds to CXCR4 (inhibited by AMD3100) and an unknown

in the env glycoprotein allows insertion of gp41 hydrophobic domain in the

eract with each other (inhibited by T1971) to form the hairpin structure. Step

effector and target membranes resulting in fusion [the W(1–3)A mutant is

logy 3

Cell lines and cell culture

The CrFK or CrFKpet cell line (Crandell feline kidney

cells persistently infected with a CrFK-adapted strain of FIV-

Petaluma; stock number ATCC-CCL-94; American Type

Culture Collection [ATCC]) was maintained in Dulbecco’s

modification of Eagle’s medium (high glucose; Mediatech,

Herndon, VA) supplemented with 10% fetal bovine serum

(FBS) (Hyclone, Logan, UT), streptomycin (100 Ag/ml),

penicillin (100 IU/ml), L-glutamine (2 mM), HEPES (15

mM), and sodium pyruvate (2 mM) at 37 jC in 5% CO2.

The HeLa cell line (stock number ATCC-CCL-2; ATCC) was

maintained in DMEM (high glucose) supplemented as de-

scribed above. The feline T cell line (FCD4E) (an interleukin

2 [IL-2]-dependent feline CD4+ lymphocyte cell line de-

scribed by English et al., 1993) was maintained in RPMI

1640 medium (Mediatech) supplemented with 10% FBS,

streptomycin (100 Ag/ml), penicillin (100 IU/ml), L-gluta-

mine (2 mM), HEPES (15 mM), sodium pyruvate (2 mM), h-mercaptoethanol (2.5 � 10�5 M), and recombinant human

IL-2 (rhIL-2) (100 U/ml; AIDS Reference and Reagent

Program) at 37 jC in 7% CO2. Other feline T cell lines Fet-

J and 3201 (Hohdatsu et al., 1996) were maintained as

described for FCD4E cells in the absence of IL-2.

Cloning of FIV env

Plasmid pFIVenv was generated by PCR amplification of

the env gene from a plasmid containing the molecular clone

JSY3 of FIV-NCSU1 (Yang et al., 1996) using primer

BamH1FIVenv> (ATTGGATCCGCAACAATAATTATGG-

CAGA) and <FIVenvEcoR1 (GCCGGAATTCAGTCTGA-

GATACTTCATCAT). For plasmid pFIVenv/rev, which

contained the 3V exon of rev, the forward primer was the

same as above (BamH1FIVenv>) and the reverse primer was

located at the end of the LTR region <FIVrevEcoR1

(ATAGAATTCAAGTTCTCGGCCCGGATTC) (underlined

sequences shows restriction enzyme sites). PCR-amplified

products were digested with BamH1 and EcoR1 and cloned

into BamH1- and EcoR1-digested pcDNA3 vector (Invitro-

gen). The cloned products were sequenced at the automated

DNA sequencing facility at University of North Carolina

(Chapel Hill, NC).

Transfection of cells

Subconfluent monolayers of CrFK cells were transfected

with 1 Ag of plasmid DNA of either pFIVenv or pFIVenv/

rev using Effectene transfection system (Qiagen Inc., Valen-

cia, CA) as per manufacturer’s instructions. Forty-eight

hours after transfection, the cells were seeded at low density

in selection medium containing G418 at 700 Ag/ml (Gib-

coBRL, Gaithersberg, MD). The cells were selected by

limiting dilution for at least 3 weeks before analysis and

use in experiments. Selected cells were maintained in 400

Ag/ml G418 throughout the study.

H. Garg et al. / Viro284

Detection of env expression

Immunocytochemistry and Western blotting were used to

detect expression of FIV env in transfected cells. For

immunocytochemistry, transfected and selected CrFK cells

were seeded in 96-well plates and grown till subconfluent.

Cells were fixed with methanol 0.03% H2O2 for 30 min

followed by blocking with 0.1% bovine serum albumin

(BSA) in phosphate-buffered saline (PBS) for 1 h. Cells

were than stained with serum from an FIV-infected cat

diluted 1:1000 in blocking buffer (0.1% BSA in PBS) for

2 h. Subsequently, cells were stained with horseradish

peroxidase (HRP) conjugated goat anti-cat antibody (Cappel

Research products, Durham, NC) for 1 h. Finally, the plates

were developed with 3-amino-9-ethylcarbazole (AEC) sub-

strate kit (Biogenex, San Ramon, CA) and analyzed by

microscopy. For Western blotting, total cellular lysates were

run on a 4–12% polyacrylamide gel, transferred to PVDF

membrane, and blocked with Tris-buffered saline (TBS)

containing 0.05% Tween 20 with 1% gelatin, overnight.

Blocked membranes were blotted with monoclonal antibody

to FIV env, SU1–30, at 1 Ag/ml followed by HRP-conju-

gated goat anti-mouse antibody (Pierce Biotechnology Inc.,

Rockford, IL) at 0.1 Ag/ml in blocking buffer. The mem-

branes were developed with enhanced chemiluminescence

(ECL) system using ECL Western blotting kit (Pierce

Biotechnology Inc.). The developed membrane was ana-

lyzed by Lumimager (Boheringer, Germany).

FIV syncytial assay

The FIV syncytia forming assay utilized FCD4E cells and

CrFK cells transfected with pFIVenv/rev (CrFKenv/rev).

CrFKenv/rev cells were plated in a 96-well flat-bottom plates

at a concentration of 2 � 104 cells/well in 100 Al of DMEM

medium and allowed to adhere overnight. The following day,

the culture medium was aspirated and FCD4E cells, 5 � 103

in 100 Al of medium, were added to give reproducible

numbers of syncytia. After 24 h, the plates were fixed and

stained with crystal violet (stock stain: 2.5 g of crystal violet,

1.25 g of Giemsa stain, 500 ml of 80% methanol; working

solution: 200 ml of stock and 200 ml of 80% methanol).

Stained syncytial plaques (fused cells that are five cell

diameters or greater) were counted using an inverted micro-

scope. When the number of syncytia formed within 24 h was

compared to the number of FCD4E cells plated, there was a

linear correlation between the number of target cells plated

and the number of syncytia produced. For studies with CrFK

cells chronically infected with Petaluma strain of FIV (CrFK-

pet) and HeLa cells, the fusion assay used has been described

previously (Medinas et al., 2002). Briefly, HeLa cells were

plated at 104 cells/well in 96-well plates, and the following

day 5� 102 CrFKpet cells were added. Plates were fixed and

stained the next day as described above. Various treatments

such as serum, peptides, or AMD3100 were added in a 2-fold

serial dilution series in 100 Al of FCD4E cell medium to the

21 (2004) 274–286

H. Garg et al. / Virology 321 (2004) 274–286 285

CrFK cells before co-culture. For anti-CXCR4 mAb treat-

ment, the FCD4E or HeLa cells were pre-incubated with

serial dilutions of the Ab for 1 h before co-culture with CrFK

cells. For fusion assays with different feline cell lines, either

CrFKenv/rev or CrFKpet was seeded in 96-well plates at

2 � 104 cells/well. FCD4E, Fet-J, or 3201 was added 24

h later at 2-fold dilutions starting at 2 � 104 cells/well.

Plates were stained and observed for syncytia formation

24 h later as described above.

Flow cytometry

Different cell lines were stained with antihuman CXCR4

antibody 44717 that cross-reacts with feline CXCR4, fol-

lowed by FITC-conjugated goat anti-mouse IgG antibody

(BD Biosciences). Stained cells were analyzed by flow

cytometry using FACS Calibur (BD Biosciences). For inhi-

bition of anti-CXCR4 antibody binding to feline cells, feline

cells were incubated with either isotype control or 44717

antibody in the presence or absence of AMD3100 (0.5 Ag/ml) for 1 h followed by washing and incubation with FITC-

conjugated goat anti-mouse antibody and analyzed as above.

Site directed mutagenesis

PCR-mediated site-directed mutagenesis was used to

induce a tryptophan (W) to alanine (A) mutation in the pre-

TM region of FIVenv. The following primer pairs were used

to generate two fragments: Fragment 1, BamH1FIVenv>

(sequence shown above) and <W(1 – 3)A-antisense

(CGCTCCTACCGCATCTTCCGCCTTTTGTAATT

GTTGTATCCC); Fragment 2, W(1–3)A-sense> (GCGGA-

AGATGCGGT AGGAGCGATAGGAAATATTCCACAA-

TAC) and <FIVrevEcoR1 (sequence shown above).

Fragments 1 and 2 were then used in an overlap PCR to ge-

nerate a mutant construct that was digested with EcoR1 and

BamH1 and cloned into pcDNA3 vector as described above.

The presence of mutation in the construct was confirmed by

sequencing.

Two-color fluorescent dye redistribution assay

For the two-color fluorescent dye redistribution assay,

CrFKenv/rev or CrFK control cells were labeled with

cytoplasmic dye 5-chloromethylfluorescein diacetate

(CMFDA) at a concentration of 0.5 Am in PBS for 30

min at 37 jC followed by washing twice with medium and

seeded in 96-well plates at 2 � 104 cells/well. The cells

were allowed to adhere overnight following which FCD4E

cells labeled with cytoplasmic dye 5- and 6-{[(4-choloro-

methyl) benzoyl] amino} tetramethyl rhodamine (CMTMR)

at a concentration of 5 Am were added to the wells at 5 �103/well. The cells were co-cultured for 24 h following

which the plates were washed to remove non-adherent cells

and fixed with 3.7% paraformaldehyde. Fixed plates were

analyzed by fluorescent microscope.

Acknowledgments

We thank Robyn Medinas (Trimeris Inc.) for providing

T1971 and T1566 peptides and Dr Edward Hoover

(Colorado State University) for providing AMD3100. We

also thank Dr Mary Tompkins and Dr Edward Havell (North

Carolina State University) for critical review of the

manuscript.

References

Alimonti, J.B., Ball, T.B., Fowke, K.R., 2003. Mechanisms of CD4+ T

lymphocyte cell death in human immunodeficiency virus infection and

AIDS. J. Gen. Virol. 84, 1649–1661.

Carr, C.M., Kim, P.S., 1993. A spring-loaded mechanism for the confor-

mational change of influenza hemagglutinin. Cell 73, 823–832.

Chen, S.S., 1996. Expression and processing of the human immunodefi-

ciency virus type 1 envelope glycoprotein. Intervirology 39, 242–248.

Choe, H., Martin, K.A., Farzan, M., Sodroski, J., Gerard, N.P., Gerard, C.,

1998. Structural interactions between chemokine receptors, gp120 Env

and CD4. Semin. Immunol. 10, 249–257.

De Parseval, A., Elder, J.H., 2001. Binding of recombinant feline immu-

nodeficiency virus surface glycoprotein to feline cells: role of CXCR4,

cell-surface heparans, and an unidentified non-CXCR4 receptor.

J. Virol. 75, 4528–4539.

Egberink, H.F., De Clercq, E., Van Vliet, A.L., Balzarini, J., Bridger, G.J.,

Henson, G., Horzinek, M.C., Schols, D., 1999. Bicyclams, selective

antagonists of the human chemokine receptor CXCR4, potently inhibit

feline immunodeficiency virus replication. J. Virol. 73, 6346–6352.

English, R.V., Johnson, C.M., Gebhard, D.H., Tompkins, M.B., 1993. In

vivo lymphocyte tropism of feline immunodeficiency virus. J. Virol. 67,

5175–5186.

Ferri, K.F., Jacotot, E., Geuskens, M., Kroemer, G., 2000. Apoptosis and

karyogamy in syncytia induced by the HIV-1-envelope glycoprotein

complex. Cell Death Differ. 7, 1137–1139.

Gallo, S.A., Puri, A., Blumenthal, R., 2001. HIV-1 gp41 six-helix bun-

dle formation occurs rapidly after the engagement of gp120 by

CXCR4 in the HIV-1 Env-mediated fusion process. Biochemistry

40, 12231–12236.

Gallo, S.A., Finnegan, C.M., Viard, M., Raviv, Y., Dimitrov, A., Rawat,

S.S., Puri, A., Durell, S., Blumenthal, R., 2003. The HIV Env-mediated

fusion reaction. Biochim. Biophys. Acta 1614, 36–50.

Giannecchini, S., Di Fenza, A., D’Ursi, A.M., Matteucci, D., Rovero, P.,

Bendinelli, M., 2003. Antiviral activity and conformational features of

an octapeptide derived from the membrane-proximal ectodomain of the

feline immunodeficiency virus transmembrane glycoprotein. J. Virol.

77, 3724–3733.

Hadzopoulou-Cladaras, M., Felber, B.K., Cladaras, C., Athanassopoulos,

A., Tse, A., Pavlakis, G.N., 1989. The rev (trs/art) protein of human

immunodeficiency virus type 1 affects viral mRNA and protein ex-

pression via a cis-acting sequence in the env region. J. Virol. 63,

1265–1274.

Hammarskjold, M.L., Heimer, J., Hammarskjold, B., Sangwan, I., Albert,

L., Rekosh, D., 1989. Regulation of human immunodeficiency virus

env expression by the rev gene product. J. Virol. 63, 1959–1966.

Hart, T.K., Kirsh, R., Ellens, H., Sweet, R.W., Lambert, D.M., Petteway Jr.,

S.R., Leary, J., Bugelski, P.J., 1991. Binding of soluble CD4 proteins to

human immunodeficiency virus type 1 and infected cells induces re-

lease of envelope glycoprotein gp120. Proc. Natl. Acad. Sci. U.S.A. 88,

2189–2193.

Hohdatsu, T., Hirabayashi, H., Motokawa, K., Koyama, H., 1996. Com-

parative study of the cell tropism of feline immunodeficiency virus

isolates of subtypes A, B and D classified on the basis of the env gene

V3–V5 sequence. J. Gen. Virol. 77, 93–100.

H. Garg et al. / Virology 321 (2004) 274–286286

Hosie, M.J., Willett, B.J., Dunsford, T.H., Jarrett, O., Neil, J.C., 1993. A

monoclonal antibody which blocks infection with feline immunodefi-

ciency virus identifies a possible non-CD4 receptor. J. Virol. 67,

1667–1671.

Hosie, M.J., Broere, N., Hesselgesser, J., Turner, J.D., Hoxie, J.A., Neil,

J.C., Willett, B.J., 1998. Modulation of feline immunodeficiency virus

infection by stromal cell-derived factor. J. Virol. 72, 2097–2104.

Jones, P.L., Korte, T., Blumenthal, R., 1998. Conformational changes in cell

surface HIV-1 envelope glycoproteins are triggered by cooperation be-

tween cell surface CD4 and co-receptors. J. Biol. Chem. 273, 404–409.

Kilby, J.M., Hopkins, S., Venetta, T.M., DiMassimo, B., Cloud, G.A., Lee,

J.Y., Alldredge, L., Hunter, E., Lambert, D., Bolognesi, D., Matthews,

M.R., Johnson, M.R., Nowak, M.A., Shaw, G.M., Saag, M.S., 1998.

Potent suppression of HIV-1 replication in humans by T-20, a peptide

inhibitor of gp41-mediated virus entry. Nat. Med. 4, 1302–1307.

Lawless, M.K., Barney, S., Guthrie, K.I., Bucy, T.B., Petteway Jr., S.R.,

Merutka, G., 1996. HIV-1 membrane fusion mechanism: structural stud-

ies of the interactions between biologically-active peptides from gp41.

Biochemistry 35, 13697–13708.

Lu, M., Blacklow, S.C., Kim, P.S., 1995. A trimeric structural domain

of the HIV-1 transmembrane glycoprotein. Nat. Struct. Biol. 2,

1075–1082.

Malim, M.H., Hauber, J., Le, S.Y., Maizel, J.V., Cullen, B.R., 1989.

The HIV-1 rev trans-activator acts through a structured target se-

quence to activate nuclear export of unspliced viral mRNA. Nature

338, 254–257.

Medinas, R.J., Lambert, D.M., Tompkins, W.A., 2002. C-Terminal gp40

peptide analogs inhibit feline immunodeficiency virus: cell fusion and

virus spread. J. Virol. 76, 9079–9086.

Melikyan, G.B., Markosyan, R.M., Hemmati, H., Delmedico, M.K., Lam-

bert, D.M., Cohen, F.S., 2000. Evidence that the transition of HIV-1

gp41 into a six-helix bundle, not the bundle configuration, induces

membrane fusion. Cell Biol. 151, 413–423.

Moir, S., Poulin, L., 1996. Expression of HIV env gene in a human T cell

line for a rapid and quantifiable cell fusion assay. AIDS Res. Hum.

Retrovir. 12, 811–820.

Moyle, G., 2003. Stopping HIV fusion with enfuvirtide: the first step to

extracellular HAART. J. Antimicrob. Chemother. 51, 213–217.

Pedersen, N.C., Ho, E.W., Brown, M.L., Yamamoto, J.K., 1987. Isolation

of a T-lymphotropic virus from domestic cats with an immunodeficien-

cy-like syndrome. Science 235, 790–793.

Phillips, T.R., Lamont, C., Konings, D.A., Shacklett, B.L., Hamson, C.A.,

Luciw, P.A., Elder, J.H., 1992. Identification of the Rev transactivation

and Rev-responsive elements of feline immunodeficiency virus.

J. Virol. 66, 5464–5471.

Reeves, J.D., Gallo, S.A., Ahmad, N., Miamidian, J.L., Harvey, P.E., Shar-

ron, M., Pohlmann, S., Sfakianos, J.N., Derdeyn, C.A., Blumenthal, R.,

Hunter, E., Doms, R.W., 2002. Sensitivity of HIV-1 to entry inhibitors

correlates with envelope/coreceptor affinity, receptor density, and fusion

kinetics. Proc. Natl. Acad. Sci. U.S.A. 99, 16249–16254.

Richardson, J., Pancino, G., Merat, R., Leste-Lasserre, T., Moraillon, A.,

Schneider-Mergener, J., Alizon, M., Sonigo, P., Heveker, N., 1999.

Shared usage of the chemokine receptor CXCR4 by primary and labo-

ratory-adapted strains of feline immunodeficiency virus. J. Virol. 73,

3661–3671.

Saez-Cirion, A., Nir, S., Lorizate, M., Agirre, A., Cruz, A., Perez-Gil, J.,

Nieva, J.L., 2002. Sphingomyelin and cholesterol promote HIV-1 gp41

pretransmembrane sequence surface aggregation and membrane restruc-

turing. J. Biol. Chem. 277, 21776–21785.

Salzwedel, K., Johnston, P.B., Roberts, S.J., Dubay, J.W., Hunter, E., 1993.

Expression and characterization of glycophospholipid-anchored human

immunodeficiency virus type 1 envelope glycoproteins. J. Virol. 67,

5279–5288.

Salzwedel, K., West, J.T., Hunter, E.A., 1999. Conserved tryptophan-rich

motif in the membrane-proximal region of the human immunodeficien-

cy virus type 1 gp41 ectodomain is important for Env-mediated fusion

and virus infectivity. J. Virol. 73, 2469–2480.

Suarez, T., Nir, S., Goni, F.M., Saez-Cirion, A., Nieva, J.L., 2000. The pre-

transmembrane region of the human immunodeficiency virus type-1

glycoprotein: a novel fusogenic sequence. FEBS Lett. 477, 145–149.

Verschoor, E.J., Boven, L.A., Blaak, H., van Vliet, A.L., Horzinek, M.C.,

de Ronde, A., 1995. A single mutation within the V3 envelope neutral-

ization domain of feline immunodeficiency virus determines its tropism

for CRFK cells. J. Virol. 69, 4752–4757.

Wild, C.T., Shugars, D.C., Greenwell, T.K., McDanal, C.B., Matthews,

T.J., 1994a. Peptides corresponding to a predictive alpha-helical domain

of human immunodeficiency virus type 1 gp41 are potent inhibitors of

virus infection. Proc. Natl. Acad. Sci. U.S.A. 91, 9770–9774.

Wild, C., Dubay, J.W., Greenwell, T., Baird Jr., T., Oas, T.G., McDanal, C.,

Hunter, E., Matthews, T., 1994b. Propensity for a leucine zipper-like

domain of human immunodeficiency virus type 1 gp41 to form

oligomers correlates with a role in virus-induced fusion rather than

assembly of the glycoprotein complex. Proc. Natl. Acad. Sci. U.S.A.

91, 12676–12680.

Willett, B.J., Picard, L., Hosie, M.J., Turner, J.D., Adema, K., Clapham,

P.R., 1997. Shared usage of the chemokine receptor CXCR4 by the

feline and human immunodeficiency viruses. J. Virol. 71, 6407–6415.

Willett, B.J., Adema, K., Heveker, N., Brelot, A., Picard, L., Alizon, M.,

Turner, J.D., Hoxie, J.A., Peiper, S., Neil, J.C., Hosie, M.J., 1998. The

second extracellular loop of CXCR4 determines its function as a recep-

tor for feline immunodeficiency virus. J. Virol. 72, 6475–6481.

Willett, B.J., Cannon, C.A., Hosie, M.J., 2002. Upregulation of surface

feline CXCR4 expression following ectopic expression of CCR5: impli-

cations for studies of the cell tropism of feline immunodeficiency virus.

J. Virol. 76, 9242–9252.

Wyatt, R., Sodroski, J., 1998. The HIV-1 envelope glycoproteins: fusogens,

antigens, and immunogens. Science 280, 1884–1888.

Yang, J.S., English, R.V., Ritchey, J.W., Davidson, M.G., Wasmoen, T.,

Levy, J.K., Gebhard, D.H., Tompkins, M.B., Tompkins, W.A., 1996.

Molecularly cloned feline immunodeficiency virus NCSU1 JSY3 indu-

ces immunodeficiency in specific-pathogen-free cats. J. Virol. 70,

3011–3017.