Human immunodeficiency virus type 2 envelope glycoprotein: Differential CD4 interactions of soluble gpl20 versus the assembled envelope complex

VIROLOGY

187, 233-241

(1992)

Human lmmunodeficiency Virus Type 2 Envelope Glycoprotein: Differential CD4 Interactions of Soluble gpl20 Versus the Assembled Envelope Complex MARK I. MULLIGAN,*+’ G. DOUGLAS RITTER, JR.,t MARGERY A. CHAlKIN,+ GALINA V. YAMSHCHIKOV,* PRASANNA KUMAR,* BEATRICE H. HAHN,*+ RAYMOND W. SWEET,+ AND RICHARD W. COMPANSI*Department of Medicine, University of Alabama at Birmingham, 1900 University Boulevard, 229 THT, UAB Station, Birmingham, Alabama 35294: fDepartment of Microbiology, University of Alabama at Birmingham, 1918 University Boulevard, UAB Station, Birmingham, Alabama 35294; and +Department of Molecular Genetics, SmithKline Beecham Pharmaceuticak, King of Prussia, Pennsylvania 19406 Received September

23, 199 1; accepted

November

22, 199 1

Utilizing a recombinant vaccinia expression system, we investigated the biological properties and CD4 receptor interactions of the envelope glycoproteins of a noncytopathic human immunodeficiency virus type 2 strain, termed HIV-P/ST, and a highly cytopathic variant derived from it. The efficiency and host cell range of syncytium formation by the recombinant glycoproteins of both viruses were highly restricted compared to those of prototypic strains of HIV (HIV-P/ROD or HIV-l/IIIB). However, the glycoprotein of cytopathic but not wild-type ST generated numerous large syncytia in the human T-cell line Sup Tl from which it was derived. A single cell line (Molt 4 clone 8) was permissive to fusion by both wild-type and cytopathic ST envelopes, but only the glycoprotein of cytopathic ST could be inhibited with a soluble form of the viral receptor CD4 (sCD4). While these results indicated major differences in the envelope glycoprotein-CD4 receptor interactions of wild-type versus cytopathic ST, direct and competition binding assays utilizing soluble external glycoprotein (SU) and sCD4 surprisingly revealed equivalent low binding affinity for both viruses. From these experiments we conclude that relevant biological properties (e.g., CD4 binding, cytopathic potential, and sCD4 neutralization) of HIV viruses which differ in their pathogenic potential are reflected in the sCD4 interactions of the assembled native envelope complex (as on cell or virion surfaces) but not the soluble SU glycoprotein. 0 1992 Academic Press, Inc

INTRODUCTION

this virus had a restricted host cell range and was deficient in syncytium formation, down-modulation of cell surface CD4, killing of CD4-bearing cells, and replication kinetics. Sequence analysis of a biologically active molecular clone which retained the phenotype of the parental isolate revealed a highly conserved overall genomic organization and a conserved envelope glycoprotein as indicated by the number and location of cysteines and predicted N-linked oligosaccharide addition sites (Kumar et al., 1990). Interestingly, serial passage of this isolate in the human T-cell line Sup Tl generated highly cytopathic variants (ST/24.1 C and ST/ 24.2C) which were characterized by rapid replicative kinetics and marked syncytium formation (Hoxie et al., 1991). Subsequent molecular analysis of these cultures identified clones which retained the cytopathic phenotypes of the parental isolates (Hahn eta/., in preparation). Using a vaccinia expression system, we previously demonstrated that the deficiency in syncytium formation, impairment in cell entry, and restricted host cell range of wild-type HIV-2/ST were properties of the envelope protein itself and not dependent on other viral gene products (Mulligan et a/., 1990). Subsequent studies revealed that a recombinant form of the external glycoprotein (SU) of wild-type HIV-2/ST had a mark-

HIV strains isolated from advanced acquired immunodeficiency syndrome (AIDS) patients replicate rapidly and cause syncytium formation and cell death in lymphoid cell cultures. In contrast, viruses isolated from patients at earlier clinical stages show slower replication kinetics and minimal or no syncytium formation (Asjo eta/., 1986; Cheng-Meyer eta/., 1988; Tersmette et a/., 1988, 1989). The propensity of HIV for genetic variation and the diminished immune responsiveness of patients later in HIV disease likely interact to produce such biological variation. In ceil culture, many studies have implicated the viral envelope glycoprotein as a determinant of tropism and cytopathicity. The underlying mechanisms and biochemical basis of these differential biological properties are poorly understood. The recent derivation of cytopathic variants from an attenuated HIV-2 isolate, designated HIV-2/ST, affords an opportunity to explore the molecular nature of cytopathicity in a systematic manner. HIV-2/ST was originally recovered from the blood of a healthy Senegalese woman (Kong et a/,, 1988). In comparison to prototypic cytopathic isolates, such as HIV-l/IIIB and HIV-2/ROD, ’ To whom reprint requests should be addressed 233

0042.6822/92

$3.00

Copyright 0 1992 by Acaderr~ Press, Inc. All rights of reproduction in any form resewed.

234

MULLIGAN

edly reduced affinity for sCD4 relative to cytopathic isolates of HIV-l (Ivey-Hoyle eta/., 1991). These observations suggested that envelope affinity for CD4 represented an important determinant of viral cytopathicity. Here, we have utilized the recombinant vaccinia system to express the envelope glycoprotein of a cytopathic HIV-2/STvariant and compared its fusion potential, host cell range, and CD4 binding affinity to those of wild-type ST as well as the prototypic HIV-l/IllB and HIV-2/ROD viruses. We observed marked differences in cell fusion and in sCD4 neutralization with wild-type and variant ST glycoproteins but, surprisingly, detected no differences in the binding affinity of their soluble SU glycoproteins for sCD4. These findings indicate that sCD4 interactions with soluble gpl20 versus the assembled envelope complex can differ significantly and that the interaction of the native glycoprotein complex with sCD4 is the biologically relevant one. MATERIALS

AND METHODS

Vaccinia vectors The construction of the recombinant vaccinia viruses (rVV) rVV/ST, rVV/ROD, rVVlBHl0, and rVV/ SC 1 1 (a negative control lacking an HIV env gene) was described previously (Mulligan eta/., 1990; Owens and Compans, 1989). A vaccinia recombinant (rVV/ST#2) expressing the entire envelope glycoprotein of a proviral clone (ST/24.1 C#2) derived from the cytopathic HIV-2/ST variant Sn24.1 C (Hahn et al,, in preparation) was similarly constructed after cloning into pSC1 lSalI (a gift of Casey Morrow). The rVV stocks were prepared and titered in TK-143B cells.

ET AL.

HIV2/ST envelope protein. Rabbit anti-HIV-2/ROD antiserum was purchased from American Biotechnologies, Inc. (Cambridge, MA). A monoclonal antibody to CD4 (OKT4) was purchased from Or-the Diagnostics, Inc. (Rariton, NJ) or produced in-house (ATCC cell line 8002). Soluble CD4 Recombinant CD4 (sCD4) was produced in CHO cells as previously described (Deen eta/., 1988). sCD4 was iodinated using Bolton-Hunter reagent (ICN, Costa Mesa, CA) to a specific activity of 2 x 1O7 cpm/ pugas previously described (Arthos et al., 1989). Envelope glycoprotein

To evaluate recombinant envelope protein expression and processing, monolayers of HEp-2 cells were infected with rVVs [at a multiplicity of infection (m.0.i.) of 5 PFU/cell] and incubated in DMEM with 2% NCS. At 16 hr postinfection, the cells were pulsed with [35S]methionine/cysteine (15 min, 100 &i/ml) and washed. After chase periods of O-l 7 hr, the cells were harvested and lysed as described (Mulligan et al,, 1990). Cell lysates as well as culture supernatants were incubated with specific antisera and Protein Aagarose (4”, 18 hr). The protein A complexes were washed three times, diluted 1 :l in sample reducing buffer (Sambrook et al., 1989), boiled for 2 min, and separated electrophoretically on 8.5% SDS-PAGE gels (Laemmli, 1970). After fixation, the gels were visualized by autoradiography. Syncytium

Cell lines

biosynthesis

formation

assay

Adherent cells [HEp-2, HeLa T4 (Maddon et a/., 1986)] were grown in Dulbecco’s minimal essential medium (DMEM) with 10% newborn calf serum (NCS). CD4+ (Sup Tl, H9, Molt 4 clone 8, CEM, CEMx174) and CD4- (HSB) suspension cell lines were grown in RPMI 1640 medium with 15% fetal calf serum. All cell lines tested free of mycoplasma. Fresh human or rhesus peripheral blood mononuclear cells (PBMCs) were separated from whole blood by gradient centrifugation (Histopaque, Sigma Co., St. Louis, MO)and PHAstimulated for 48 hr.

Monolayers of adherent cells or aliquots (3 X lo5 cells) of suspension lymphoid cells were infected with rVVs at a m.o.i. of 5 PFU/cell. After 24 hr of incubation in 96-well plates (37”, 5% CO,), syncytium formation was quantitated by observation of individual wells using the 4X objective of a Diaphot inverted microscope (Nikon, Inc., Garden City, NJ). The total number and relative sizes of the syncytia in each well were recorded by two observers. Syncytium size was estimated as the number of single cell diameters per median syncytium diameter in each culture. No substantial variations in scoring occurred between observers.

Antisera

sCD4 syncytium

Human anti-HIV-2 antisera were obtained through the AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH from S. Osmanov. Polyclonal anti-HIV-Z/ST antiserum was produced in rabbits inoculated with Drosophila recombinant-produced

Molt 4 clone 8 cells were infected with rVVs at an m.o.i. (0.1 PFU/cell) known to yield approximately 60 syncytia per well. Following adsorption for 1 hr at 37”, cells were washed and subsequently incubated in 96well plates (7.5 x 1O4 cells per well) in complete RPMI

inhibition

assay

CD4 BINDING

media containing varying dilutions of sCD4. After 15 hr, the syncytium numberwasdetermined bytwo independent observers. Production

and quantitation

of soluble SU proteins

Nearly confluent monolayers of HEp-2 cells in roller bottles were infected with rVV at an m.o.i. of 5 PFU/ cell. The cells were incubated in DMEM without serum for 72 hr (37”, 5% CO,). The media were collected and cellular debris was removed with a low-speed spin. High-speed centrifugation (1 hr at 23,000 rpm in SW28 rotor) was utilized to remove vaccinia virions from the medium. The spontaneously released soluble envelope proteins in the supernatant were then precipitated with 30% (w/v) ammonium sulfate (4” for 18 hr) and centrifuged at 2000 rpm for 15 min. The resuspended pellets were spin-dialyzed twice against PBS using an Amicon (Danvers, MA) centriprep 30 concentrator to remove residual ammonium sulfate. The resulting soluble SU preparations were resuspended as 100X concentrates in PBS and quantitated by immunoblot analysis. Samples were electrophoresed on 10% SDSPAGE gels adjacent to lanes with known amounts of recombinant gpl20 and transferred to nitrocellulose. The standards employed were recombinant IIIB and ST gpl20 produced in drosophila cells (Ivey-Hoyle et al., 1991) or recombinant ROD gp105 produced in a baculovirus expression system (American Bio-Technologies, Cambridge, MA). SU molecules were detected using rabbit anti-gpl20specific antisera followed by donkey anti-rabbit HRP-conjugated second antibody. gpl20 bands were visualized using an enhanced chemiluminescence detection system (Amersham Corp., Arlington Heights, IL). SU molecule concentrations utilized for the competition binding assay were based on two independent immunoblots. This semiquantitative method permits an estimation of the unpurified glycoprotein concentration with + 50% accuracy. CD4 binding assays Saturation and competition binding assays with the recombinant env proteins were performed essentially as previously described (Ivey-Hoyle et a/., 1991). Direct binding analysis. Increasing concentrations (0. l-l .O pmol) of ‘YsCD4 were incubated with a fixed concentration of each soluble vaccinia-expressed SU protein in buffer (PBS containing 0.25% milk and 0.05% NP-40) in a total volume of 100 ~1. After 4 hr at room temperature, the gpl20/sCD4 complexes were immunoprecipitated using HIV-l- or HIV-2-specific patient serum followed by capture on Protein A-Sepharose beads. The immunoprecipitates were washed 3X with ice-cold buffer and counted. Nonspecific binding

235

BY HIV-2 ENVELOPE

was determined by incubating matched control samples with a 500-fold excess of unlabeled sCD4. The dissociation constant (K,) of binding was determined by Scatchard analysis. Competition binding analysis. Recombinant BH 10 gpl20 produced in Drosophila cells (Ivey-Hoyle et al., 1991; Culp et a/., 1991) was iodinated using BoltonHunter reagent (ICN, Costa Mesa, CA) to a specific activity of 1 X 1O7 cpm/pg as previously described (Arthos et a/., 1989). Equimolar concentrations (1.5 nM) of sCD4 and the 1251-gp120 were mixed with increasing concentrations of unlabeled soluble SU proteins in a total volume of 100 ~1. The binding reactions were allowed to proceed for 2 hr at room temperature. The gpl2O/sCD4 complexes were captured by immunoprecipitation using excess anti-CD4 monoclonal antibody OKT4 followed by Protein A-Sepharose beads. The samples were washed three times and counted. Relative binding affinities of the different vaccinia-expressed SU glycoproteins were determined by comparing the concentrations required to inhibit the lz51-gpl 2O/sCD4 interaction by 50% (IC,,).

RESULTS Expression and processing glycoproteins

of the envelope

We previously used a recombinant vaccinia expression system to demonstrate that the HIV-2/ST envelope glycoprotein was responsible for this virus’ attenuated phenotype (Mulligan et al., 1990). To investigate the contribution of the envelope protein of a highly cytopathic ST variant to the cytopathic behavior of that virus, we examined the envelope gene from a biologically active molecular clone of this isolate in the same system. To evaluate conditions for the expression and processing of the envelope proteins, human CD4HEp-2 cells were infected with the vaccinia recombinants. Following a 15-min pulse with radiolabeled methionine/cysteine, the envelope proteins were found only in cell lysates as uncleaved precursor polyproteins (arrow, Fig. 1A). The BHlO envelope precursor had a relative molecular mass (n/lJ of approximately 160 kDa (lane 5) while that of ST migrated slightly slower (lane 3). The ROD precursor protein (lane 2) had an M, of approximately 150 kDa and is known to possess a premature stop codon that truncates its TM protein by 139 amino acids (AA) (Guyader et a/., 1987). The cytopathic STvariant (hence designated ST#2) (lane 4) also codedforan envelope precursorwith fasterelectrophoretie mobility (similar to that of HIV-2/ROD), consistent with a premature termination codon in the molecular

236

MULLIGAN A

12345

B

1

2

3

4

5

ET AL.

6

7

8

9

10

180

84 58 48

36 26

expressed by vaccinia recombinants (as under Materials and Methods). (A) Cell FIG. 1. Radiolabeled HIV-2 or HIV-1 envelope glycoproteins lysates immunoprecipitated without a chase period. Lane 1, rW/SCl 1 (a negative control); lane 2, rVV/ROD; lane 3, rVV/ST; lane 4, rVV/ST#2; lane 5, rVV/BHlO. The uncleaved HIV glycoprotein precursors are indicated (arrow). Markers of approximate molecular weight (KDa) are on the left. (B) RIP/PAGE following a 17.hr chase period. Lanes l-5 are cell lysates as above. Lanes 6-l 0 are the corresponding culture media. The lysates contained precursor protein (indicated by PRE) as well as processed envelope subunits (indicated by SU and TM). During the chase period, SU proteins accumulated in culture media (lanes 7-l 0). Equivalent volumes of lysates or media were immunoprecipitated and electrophoresed for each rVV.

clone 147 AA upstream of the ST terminus (Hahn eta/., in preparation). Following a 17-hr chase period, the precursor proteins were partially processed to the mature SU (120130 kDa) and TM proteins (Fig. 1b, lanes 2-5). Within the cell lysates (lanes 2-5), over 50% (by visual estimation) of the total envelope protein remained uncleaved for all isolates, as has been reported previously (Owens and Compans, 1990; Willey et al., 1988). The TM proteins of wild-type ST and BH 10 viruses, which encode full-length TM proteins, appeared as diffuse bands with an M, of approximately 42 kDa (lanes 3 and 5, respectively). The truncated TM proteins of ROD and cytopathic ST were not well resolved (lanes 2 and 4). The SU proteins were also detectable in the culture media (lanes 7-l 0), indicating that the proteolytically processed glycoproteins were available for transport to and release from the cell surfaces. However, the fraction remaining associated with cells varied in an isolate-specific manner. Both ST#2 and ROD retained Iittle of their total SU protein with the cells (lanes 4 and 2) compared with wild-type ST and BHlO (lanes 3 and 5, respectively).

Host-restricted glycoprotein

syncytium

formation

by ST#2

To evaluate cell fusion by the different envelope glycoproteins, we compared the abilities of the vaccinia recombinants to induce syncytia in 10 cell types (Table 1). CD4+ or CD4- cells were infected with each vaccinia recombinant and subsequently evaluated by phase-contrast microscopy. In Sup Tl cells, the line in which the cytopathic STvariants arose, ST#2 envelope induced numerous and large syncytia, in contrast to ST which gave only a low number of small syncytia (Table 1). However, both wild-type and cytopathic ST efficiently generated many large syncytia in the highly permissive Molt 4 clone 8 cell line. Neither glycoprotein generated syncytia in other cell lines or rhesus PBMCs, in agreement with our prior observations with rVV/ST (Mulligan et al., 1990). Interestingly, wild-type ST, but not cytopathic ST, generated syncytia in human PBMCs. The BHlO and ROD vaccinia recombinants produced numerous syncytia in most CD4+ cell lines as well as human PBMCs, and these syncytia were gener-

CD4 BINDING

237

BY HIV-2 ENVELOPE TABLE 1

SYNCMIUM

FORMATION ASSAY: NUMERATOR-AVERAGE

SYNCMIUM

Cell type

Sup Tl

M4 C8

H9

CEM

CEMxl74

rVV: ST ST/#2 ROD BHlO SC11

3115 9/>100 18/>100 14hlOO 0

12/>100 lO/>lOO 15/>100 40/> 100 0

0 0 1 l/>lOO 12/>100 0

0 0 3/20 9/>100 0

0 0 0 5128 0

SIZE; DENOMINATOR-NUMBER

HeLa T4

0 0 45150 >100/>100 0

OF SYNCMIA~

Human PBMC

Rhesus PBMC

HSB

HEp-2

1O/50 o/o lO/>lOO 12170 0

o/o o/o o/o 316 0

0 0 0 0 0

0 0 0 0 0

a Visual inspection of syncytia generated by each of five vaccinia recombinants (rVV) in each of 10 cell lines. The results are displayed as: numerator-syncytium size (No. single cell diameters per syncytium diameter); denominator-number of syncytia per well of 96-well plate. Cultures were scored 24 hr postinfection. Results are shown as mean of determinations by two observers each scoring two independent experiments (one with duplicate wells).

ally substantially larger than those observed with either ST recombinant. Interestingly, BHlO induced syncytia in two cell lines, CEM and CEMxl74, in which ROD generated few or no syncytia. In all instances, the absence of cell fusion did not result from a failure to express a functional envelope protein since syncytia formed rapidly (2-4 hr) upon coculture of the rVV-infected cells with uninfected Molt 4 clone 8 cells (not shown). The differential phenotypes of the various viral strains on the same and different cell lines emphasize the contributions of both isolate-specific and cell typedependent factors in syncytium formation. sCD4 inhibits syncytium but not wild-type ST

formation

by cytopathic

To analyze further the differential fusogenic potential of the two HIV-2/ST-derived envelope glycoproteins, we performed sCD4 syncytium inhibition assays utilizing vaccinia recombinant-infected cells expressing assembled glycoprotein complexes at their surfaces (Fig. 2). The cell line chosen for this experiment was Molt 4 clone 8, shown above to fuse efficiently with both ST and ST#2. While syncytium formation by the cell-associated envelope glycoprotein of ST was not inhibited at concentrations of sCD4 up to 100 pg/ml, cytopathic ST#2 was inhibited completely. The IC,, for sCD4 inhibition of ST#2 was about 10 pg/ml (230 nM). In this assay, BH 10 was similarly sensitive to sCD4 inhibition (not shown). The soluble SU glycoproteins of ST and ST#2 bind sCD4 with equivalent low affinity To investigate whether CD4 binding affinity of the SU glycoproteins correlated with the above differences in cell fusion and sCD4 inhibition, we performed direct and competition CD4 binding assays utilizing soluble

SU proteins and soluble CD4. The SU glycoproteins recovered from the culture media of vaccinia recombinant-infected cells were quantitated relative to purified HIV-l or HIV-2 SU protein standards (Fig. 3). In direct binding experiments with the SU proteins of BHlO, ROD, or ST, saturation of binding of ‘251-labeled sCD4 was achieved only for BHlO and ROD (Fig. 4a). Only minimal binding of ST occurred, although the input concentration of that SU protein was equal to that of BHlO and twice that of ROD, indicating a markedly reduced binding affinity for the wild-type ST protein. As determined by Scatchard analysis (Fig. 4b), the Kd values for BHlO and ROD were 1.5 and 3.1 nM, respectively (Table 2). Competition binding assays were then carried out to quantify the low affinity of the ST envelope protein and

1.5 2 m 5 1.0 E .-a ,” E 2

0.5

oh

0.0 0

sCD4

(mcglml)

FIG. 2. Inhibition of ST#2 but not ST syncytium formation by soluble CD4 (sCD4). Molt 4 clone 8 cells were infected with rVV/ST or rVV/ST#2, incubated in the presence or absence of dilutions of sCD4, and scored. The syncytium index is the (number of syncytia in presence of sCD4)/(number of syncytia in absence of sCD4). Results are shown as mean t standard deviation of determinations by two observers each scoring four independent experiments. The actual numbers of syncytia generated in this cell type by ST or ST#2 in the absence of sCD4 were similar.

238

MULLIGAN

BHlO 1

ST

ROD 234

ET AL.

12

3

12345678

FIG. 3. Quantitative immunoblot of the soluble SU glycoproteins. Vaccinia-expressed SU proteins were prepared and quantitated by immunoblot analysis as described under Materials and Methods. Lanes 1 and 2 are known amounts of recombinant standards: for BHlO, 3 and 1 rig/lane, respectively; for ROD and ST, 20 and 10 rig/lane, respectively. The positions of molecular weight standards are indicated on the left. For comparison, the experimental SU preparations were loaded as follows: BH 10, lanes 3 and 4, 2 and 1 ~1; ROD, lane 3, 2 pl; ST, lanes 3 and 4, 2 and 1 ~1; ST#2, lanes 5 and 6, 2 and 1 ~1; SC1 1 (control), lanes 7 and 8, 2 and 1 ~1, respectively. Very similar results were obtained in duplicate immunoblots. On the basis of these results and on additional standards not shown, the concentrations of the various SU proteins are accurate within 50%.

to compare it with the SU protein of cytopathic ST (Fig. 5; Table 2). Consistent with the direct binding assays, both Drosophila BHl 0 (dBH 10) and vaccinia-produced BH 10 glycoprotein, as well as both baculovirus- and vaccinia-produced ROD glycoprotein, efficiently competed with the binding of ‘251-labeled dBH 10 SU protein to sCD4. In contrast, the SU protein of noncytopathic ST was unable to compete efficiently even at concentrations 20-fold greater than the IC,, of dBHl0 (the practical upper concentration limit for the vaccinia SU protein preparations). Surprisingly, the soluble SU protein of the cytopathic ST variant likewise competed minimally at high concentrations. A comparison of the relative affinities of the SU proteins for sCD4, as approximated from the IC,,values, is shown in Table 2. In relation to purified dBH 10 SU protein, the affinities of the ST and ST#2 proteins were reduced more than 20-fold and appear to be similar. By extrapolation of the curves (Fig. 5), the reduction in their sCD4 affinities probably approaches 1OO-fold relative to dBHl0.

DISCUSSION Utilizing vaccinia recombinants to obtain high levels of expression, we compared the biological and biochemical properties of the envelope glycoproteins of an attenuated HIV-2 virus (HIV-2/ST), a cytopathic variant derived from it, and prototypic cytopathic viruses HIV-l/IIIB (BHlO) and HIV-2/ROD. The ability of the vaccinia-expressed cell-associated envelope proteins to form syncytia essentially paralleled that previously observed for the parental viruses. The envelope proteins of the cytopathic IIIB and ROD isolates were quite potent inducers of fusion as reflected in the size and number of syncytia and in the host range. ROD was not active, or only weakly active, in three cell types that supported syncytium formation by IIIB. The ST isolates were quite restricted in their ability to induce syncytia in CD4+ cells. ST#2 produced large and numerous syn-

cytia in Sup Tl cells, the cell line from which the virus was derived, and both ST and ST#2 generated many, large syncytia in the highly permissive Molt 4 clone 8 cell line. Cell fusion in that cell line by ST#2 was sensitive to sCD4 inhibition, whereas fusion by the cell-associated envelope complex of wild-type ST was not. Despite the differences in cell fusion and sCD4 neutralization demonstrated by the cell-associated, assembled glycoprotein complexes, the soluble SU proteins of both ST viruses had an equivalent low affinity for sCD4. From these data, we conclude that the differential cytopathicity of the ST and ST#2 viruses is a property of their envelope glycoproteins. However, both cell-specific and isolate-specific factors influence the level of cytopathology that results from the envelope protein/target cell interaction. More importantly, our data also indicate that the cytopathic potential of the ST#2 glycoprotein complex appears to correlate with sensitivity to sCD4 inhibition but not with the sCD4 affinity of its soluble SU glycoprotein. CD4 binding affinities have previously been reported for the SU proteins of IIIB, ROD, and ST by ourselves and others (Ivey-Hoyle et al., 1991; Laskey et al., 1987; Moore, 1990). Our present results with some of the vaccinia-produced soluble env proteins differ quantitatively from the previous results but yield the same conclusion, namely a rank order of relative affinity of IIIB > ROD $ ST. Our results indicate a 2-to 3-fold lower CD4 affinity for ROD (whether expressed by vaccinia or baculovirus systems) relative to that for IIIB but contrast with other reports of a lo- to 25-fold lower affinity (Moore, 1990; Moore and Morikawa, 1991). The vaccinia-produced wild-type ST glycoprotein had an affinity about 1OO-fold less than that of the BH 10 SU protein, compared to the >200-fold difference we previously observed with a purified recombinant ST protein produced in Drosophila cells (Ivey-Hoyle et al., 199 1). These differences in affinity may reflect factors such as differential glycosylation and protein quantitation.

CD4 BINDING A

239

BY HIV-2 ENVELOPE TABLE 2

0.06 1

AFFINITYMEASUREMENTSOF

SU PROTEINSFORSCD~~

l

0.05z E ,0 ,0 Iii 2 x2 ?i tn

0.040.04-

. l

.

0.03-

0.020.02-I

BHIO ROD ST

l

0.0

0.4

0.2

0.6

0.8

I .o

sCD4 (nM)

l l

0.00

0.01

0.02

0.03

0.04

0.05

BHIO ROD

SU protein

Expression system

Measured Kd (nWb

Relative CD4 affinity by competitionC

HIV-l/BHlO HIV-l/BHlO HIV-2/ROD HIV2/ROD HIV-2/ST HIV-2/ST#2

Drosophila Vaccinia Baculovirus Vaccinia Vaccinia Vaccinia

1 .o 1.5 NDd 3.1 NMe NM

1.0 2.5 1.7 4.0 >20 >20

a Values shown are means of two experiments (except vaccinia BH 10 and baculovirus ROD-one experiment). ’ Results of saturation binding analysis using “%CD4 and soluble SU proteins. c Results of competition binding assays using 1251-gpl20, soluble SU proteins, and sCD4 (see Materials and Methods). d ND, not determined. e NM, not measurable.

bled cytopathic ST envelope complex at the cell surface is consistent with our observed sensitivity of cells expressing cytopathic ST glycoprotein to syncytium inhibition by sCD4. Taken together, these data indicate that the high affinity of the assembled native envelope complex of cytopathic ST as opposed to the low affinity of its soluble SU proteins derives from conformation imposed by the multimeric presentation of the TM-associated SU glycoprotein on the surface of the virus and virus-infected cells.

0.06

Bound (pMol) FIG. 4. (A) Direct (saturation) binding analysis of the vaccinia-expressed SU proteins for sCD4. Binding was carried out as described under Materials and Methods. Saturation of binding was achieved for BH 10 (solid circles) and ROD (solid diamonds) but not for ST (solid triangles). (B) Scatchard analysis of the binding curves for BH 10 and ROD SU oroteins.

Companion studies utilizing FITC-conjugated sCD4 and flow cytometry to directly measure the CD4 binding affinity of the native envelope complex of HIV on infected cells have independently evaluated a prototype strain of HIV-1 as well as wild-type and cytopathic HIV-2/ST viruses (Hoxie et al., 1991). sCD4 had a high affinity for IIIB and a very low affinity for ST, in agreement with our present and previous (Ivey-Hoyle et al., 1991) studies with recombinant soluble SU proteins from these isolates. However, ST/24.2C-a second and phenotypically indistinguishable cytopathic ST isolate derived from the same biological clone as ST/ 24.1 C-bound to sCD4 with a high affinity comparable to that observed for IIIB. This high affinity of the assem-

6000

I

6000

oD 0

10

20

30

40

50

60

70

60

-

dBHl0

-ROD

BHiO

-

ST ST#2

-

SC11

90100

SU (nM) FIG. 5. Competition binding assay for the soluble SU proteins. Serial dilutions of the indicated unlabeled SU proteins were mixed with a fixed amount of lzsl-gpl 20 (Drosophila-produced BH 10, dBH 10). sCD4 was then added and the reactions were incubated for 2 hr at room temperature. gpl20/sCD4 complexes were collected by immunoprecipitation as described under Materials and Methods. Unlabeled dBH 10 protein (open circles) was used as a known standard competitor in this assay. SC1 1 (open squares) is a control vaccinia recombinant preparation lacking SU protein,

240

MULLIGAN

An analogous but inverse relationship of sCD4 affinity to envelope glycoprotein presentation (soluble versus an assembled complex) appears to occur for several other HIV isolates. Primary HIV-1 isolates that underwent only limited passage in cell culture were reported to be quite refractory to inhibition by sCD4 proteins (Daar et a/., 1990). A reduced sensitivity to sCD4 neutralization was previously observed for some isolates of HIV-2 and SIV (Clapham eta/., 1989; Looney et al., 1990; Sekigawa et al., 1990) and the affinities of their SU proteins were reported to be lower than that observed for the prototypic IIIB envelope protein (Moore, 1990; Ivey-Hoyle et a/., 1991). These results suggested that the relative insensitivity of HIV-l primary isolates to sCD4 derived from a similarly reduced sCD4 affinity. Surprisingly, the soluble SU proteins of several such isolates were found to have high affinities similar to IIIB (Ashkenazi et al., 1991; Brighty et a/., 1991). The explanation of this apparent discrepancy now appears to be that the assembled envelope complex on the virion, as opposed to the soluble SU protein, has a reduced affinity for sCD4 (J. Moore, et a/., 1992). Thus, the relative sensitivities of the ST and BHlO vaccinia recombinants and the primary HIV-1 isolates to sCD4 inhibition is directly reflected in the affinity of sCD4 for the assembled envelope complex, rather than for the soluble SU protein. The observations for the ST viruses establish a correlation between virion affinity for CD4 and cytopathicity. However, this correlation is highly host cell restricted and other isolate-specific glycoprotein properties are operative as well. Viruses derived from two cytopathic ST cultures-as well as the ST#2 vaccinia recombinant-induce syncytia only in the Sup Tl cell line in which they were selected and in the highly permissive Molt 4 clone 8 line. In every cell line examined, generation of syncytia by the vaccinia recombinants paralleled previous observations with the parental viruses. Interestingly, vaccinia-expressed wild-type ST envelope did cause cell fusion in one cell type (human PBMCs), in which neither infection by the original ST isolate nor vaccinia-expressed ST#2 glycoprotein caused syncytia. We postulate that in this one instance, a qualitative cell fusion deficiency present in the wild-type ST virus was no longer observed when the envelope glycoprotein was quantitatively overexpressed in the vaccinia system. While the cytopathic ST virus was identified based on its ability to induce fusion of Sup Tl cells, its glycoprotein was apparently even less competent for human PBMC fusion, to the extent that vaccinia overexpression could no longer generate syncytia. This observation emphasizes the exquisite adaptability of HIV to its cellular environment and the cell-specific nature of env gene alterations.

ET AL.

Host range differences were also observed between the vaccinia-expressed env proteins of the highly cytopathic isolates ROD and BHl 0. Similar host-dependent restriction of cytopathic effects in CD4+ cells that cannot be attributed to lack of envelope expression has been described for other HIV-l, HIV-2, and SIV isolates (Koenig et al., 1989; Ashorn et a/., 1990; Clapham et al., 1991). Thus, although a high affinity between CD4 and the envelope complex may promote efficient fusion, other host-specific determinants are apparently operative in an isolate-specific manner. Similarly, isolate-specific glycoprotein determinants in addition to CD4 affinity-e.g., differences in postbinding properties such as the association between SU and TM-appear to restrict host range and perhaps influence sCD4 sensitivity as well. While studies of primary human lymphocytes infected with low passage HIV isolates may best approximate the interaction of HIV with its in viva circulating targets, otherin vitro systems permit detailed biochemical analyses of glycoprotein-mediated viral entry and cytopathic effects. The ST virus/Sup Tl model and vaccinia-env expression systems are useful tools with which to gain insights into the molecular mechanisms of HIV cell tropism and cytopathology. In addition, the finding that CD4 affinity is apparently subject to selection pressures in vitro suggests that similar selection pressures may also occur in viva and may contribute to the emergence of more virulent strains in infected individuals over time. ACKNOWLEDGMENTS This work was supported by Public Health Service Grants Al00912, Al-27290, AI-28147, and Al-25291 from the National Institute of Allergy and Infectious Disease. We thank K. Wynne for manuscript preparation; L. Melsen and E. Arms for photographic assistance; R. Owens for rVV/SCll and rW/BHlO; C. Morrow for pSC1 l-%/l; J. Hoxie for HIV-2/ST.lC and HIV-1 patient serum; P. Fultz for rhesus PBMCs; and B. Hellmig for providing recombinant Drosophila SU proteins and OKT4 monoclonal antibody.

REFERENCES ARTHOS, J., DEEN, K. C., CHAIKIN, M. A., FORNWALD,J. A., SATHE, G., SAT~ENTAU, Q. J., CLAPHAM, P. R., WEISS, R. A., MCDOUGAL, J. S., PEITROPAOLO,C., AXEL, R., TRUNEH, A., MADDON, P. J., and SWEET, R. W. (1989). Identification of the residues in Human CD4 critical for the binding of HIV. Cell 57, 469-481. ASHKENAZI, A., SMITH, D. H., MARSTERS, S. A., RIDDLE, L., GREGORY, T. J., Ho, D. D., and CAPON, D. J. (1991). Resistance of primary isolates of human immunodeficiency virus type 1 to soluble CD4 is independent of CD4-rgpl20 binding affinity. Proc. Nat/. Acad. Sci. USA 88,7056-7060. ASHORN, P. A., BERGER,E. A., and Moss, B. (1990). Human immunodeficiency virus envelope glyprotein/CD4 mediated fusion of nonprimate cells with human cells. J. Viol. 64, 2149-2156. ,&JO, B., ALBERT, J., KARLSSON,A., MORFELDT-MAMSON, L., BIBERFELD.

CD4 BINDING

BY HIV-2 ENVELOPE

G., LIDMAN, K., and FENYB, E. M. (1986). Replicative properties of human immunodeficiencyvirus from patients with varying severity of HIV infection. Lancer ii, 660-662. BRIGHP/, D. W., ROSENBERG,M., CHEN, I. S. Y., and IVEY-HOYLE, M. (1991). Envelope proteins from clinical isolates of human immunodeficiency virus type 1 that are refractory to neutralization by soluble CD4 possess high affinity for the CD4 receptor. Proc. Nat/. Acad. Sci. USA 88, 7802-7805. CHENG-MEYER, C., SETO, D., TATENO, M., and LEVY, J. A. (1988). Biologic features of HIV-l that correlate with virulence in the host. Science 240, 80-82. CLAPHAM, P. R., WEBER, J. N., WHITBY, D., MCINTOSH, K., DALGLEISH, A. G., MADDON, P. J., DEEN, K. C., SWEET, R. W., and WEISS, R. A. (1989). Soluble CD4 blocks the infectivity of diverse strains of HIV and SIV for T cells and monocytes but not for brain and muscle cells. Nature (London) 337, 368-370. CLAPHAM, P. R., BLANC, D., and WEISS, R. A. (1991). Specific cell surface requirements for the infection of CD4-positive cells by human immunodeficiencyvirus types 1 and 2 and by simian immunodeficiency virus. tirology 181, 703-715. CULP, J. S., JOHANSEN, H., HELLMIG, B., BECK, J., MATHEWS, T. J., DELERS, A., and ROSENBERG,M. (1991). Regulated expression allows high level production and secretion of HIV-l gpl20 envelope glycoprotein in Drosophila Schneider cells. BioITechnology 9, 173-177. DAAR, E. S., LI, X. L., MOUDGIL, T., and Ho, D. D. (1990). High concentrations of recombinant soluble CD4 are required to neutralize primary HIV-1 isolates. Proc. Nat/. Acad Sci. USA 87,6574-6578. DEEN, K. C., MCDOUGAL, J. S., INACKER,R., FOLENA-WASSERMAN,G., ARTHOS, J., ROSENBERG,J., MADDON, P. J., AXEL, R., and SWEET, R. A. (1988). A soluble form of CD4 (T4) protein inhibits AIDS virus infection. Nature (London) 331, 82-84. GUYADER, M. M., EMERMAN, P., SONIGO, F., CLAVEL, L., MONTAGNIER, and ALIZON, M. (1987). Genome organization and transactivation of the human immunodeficiency virus type 2. Nature (London) 326, 662-669. HOXIE, J. A., BRASS, L. E., PLETCHER, C. H., HAGGARTY, B. S., COLLORAN, R., and HAHN, B. (1991). Cytopathic variants of an attenuated isolate of human immunodeficiency virus type 2 exhibit increased affinity for CD4. J. Viral. 65, 5096&5101. IVEY-HOYLE, M., CULP, J. S., CHAIKIN, M. A., HELLMIG, B. D., MATHEWS, T. J., SWEET, R. W., and ROSENBERG,M. (1991). Envelope glycoproteins from biologically diverse isolates of immunodeficiency viruses have widely different affinities for CD4. Proc. Nat/. Acad. Sci. USA 88, 512-516. KOENIG, S., HIRSCH, V. M., OLMSTED, R. A., POWELL, D., MAURY, W., RABSON, A., FAUCI, A. S., PURCELL, R. H., and JOHNSON, P. R. (1989). Selective infection of human CD4+ cells by SIV: Productive infection associated with envelope glycoproteininduced fusion. Proc. Nat/. Acad. Sci. USA 86, 2443-2447. KONG, L. I., LEE, S., KAPPES,J. C., PARKIN, J. S., DECKER, D., HOXIE, J. A., HAHN, B. H., and SHAW, G. M. (1988). West African HIV-2-related human retrovirus with attenuated cytopathicity. Science 240, 1525-l 529. KUMAR, P., HUI, H., KAPPES,J. C., HAGGARTY,B. S., HOXIE, J. A., ARYA, S. K., SHAW, G. M., and HAHN, B. H. (1990). Molecularcharacterization of an attenuated human immunodeficiency virus type 2 isolate. /. Viral. 64, 890-901. IAEMMLI, U. K. (1970). Cleavage of structural proteins during the

241

assembly of the head of the bacteriophage T4. Nature (London) 227, 680-685. LASKEY,L. A., NAKAMURA, G., SMITH, D. H., FENNIE, C., SHIMASAKI,C., PATZER, E., BERMAN, P., GREGORY,T., and CAPON, D. J. (1987). Delineation of a region of the human immunodeficiency virus type 1 gpl20 glycoprotein critical for interaction with the CD4 receptor. Cell 50, 975-985. LOONEY, D. J., HAYASHI, S., NICKLAS, M., REDFIELD,R. R., BRODER,S., WONG-STAAL, F., and MITSUYA, H. (1990). Differences in the interaction of HIV-1 and HIV-2 with CD4. /. Acquired immune. Defic. Syndr. 3, 649-657. MADDON, P. J., DALGLEISH,A. G., MCDOUGAL, J. S., CLAPHAM, P. R., WEISS, R. A., and AXEL, R. (1986). The T4 gene encodes the AIDS virus receptor and is expressed in the immune system and the brain. Cell 47, 333-348. MOORE, J. P. (1990). Simple methods for monitoring HIV-l and HIV-2 gpl20 binding to sCD4 by ELISA; HIV-2 has a 25.fold lower affinity than HIV-l for sCD4. A/DS 4, 297-305. MOORE, J. P., MCKEATING, J. A. HUANG, Y., ASHKENAZI, A., and Ho, D. D. (1992). Virions of primary human immunodeficiency virus type 1 isolates resistant to soluble CD4 (sCD4) neutralization differ in sCD4 binding and glycoprotein gpl20 retention from sCD4-sensitive isolates. J. Viroi. 66, 235-243. MOORE, 1. P.. and MORIKAWA,Y. (1991). Binding of recombinant HIV-1 and HIV-2 SU glycoproteins to sCD4. /. Acquired immune. Defic. Syndr. 4, 442-443. MULLIGAN, M. J., KUMAR, P., HUI, H., OWENS, R., RITTER, G. D., JR., HAHN, B. H., and COMPANS, R. W. (1990). The env protein of an infectious noncytopathic HIV-2 is deficient in syncytium formation. AIDS Res. Human Retroviruses 6, 707-720. OWENS, R. J., and COMPANS, R. (1989). Expression of the human immunodeficiency virus envelope glycoprotein is restricted to basolateral surfaces of polarized epithelial cells. 1. Viroi. 63, 978982. OWENS, R. J., and COMPANS, R. (1990). The human immunodeficiency virus type 1 envelope glycoprotein precursor acquires aberrant intermolecular disulfide bonds that may prevent normal proteolytic processing. virology 179, 827-833. SAMBROOK, J., FRITSCH, E. F., and MANIATIS, T. (1989). “Molecular Cloning: A Laboratory Manual.” 2nd ed., pp. 18.53. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. SEKIGAWA, I,, CHAMOW, S. M., GROOPMAN, J. E., and BYRN, R. A. (1990). CD4 immunoadhesin, but not recombinant soluble CD4, blocks syncytium formation by human immunodeficiency virus type 2-infected lymphoid cells. J. Viral. 64, 5194-5198. TERSME~E, M., DEGOEDE, R. E. Y., AL, B. J. M., WINKEL, I. N., GRUTERS,R. A., CUYPERS,H. T., HUISMAN, H. G., and MIEDEMA, F. (1988). Differential syncytium-inducing capacity of human immunodeficiency virus isolates: Frequent detection of syncytium-inducing isolates in patients with acquired immunodeficiency syndrome (AIDS) and AIDS-related complex. J. Krol. 62, 2026-2032. TERSME~E, M., GRUTERS, R. A., DEWOLD, F., DEGOEDE, R. E. Y., LANGE, J. M. A., SCHELLEKENS,P. T. A., GOUDSMIT, J., HUISMAN, H. G., and MIEDEMA, F. (1989). Evidence for a role of virulent human immunodeficiency virus (HIV) variants in the pathogenesis of acquired immunodeficiency syndrome: Studies on sequential HIV isolates. /. Viroi. 63, 21 18-2125. WILLEY, R. L., BONIFACINO, J. S., Porrs, B. J., MARTIN, M. A., and KLAUSNER,R. D. (1988). Biosynthesis, cleavage, and degradation of the human immunodeficiency virus 1 envelope glycoprotein gpl60. Proc. Nat/. Acad. Sci. USA 85, 958OC9584.