AIDS Research and Human Retroviruses 1992.8:1599-1610. Downloaded from online.liebertpub.com by Ucsf Library University of California San Francisco on 01/11/15. For personal use only.
AIDS RESEARCH AND HUMAN RETROVIRUSES Volume 8, Number 9, 1992 Mary Ann Fiebert, Inc., Publishers
Sulfated Polyester Interactions with the CD4 Molecule and with the Third Variable Loop Domain (v3) of gpl20 Are Chemically Distinct SETH
LEDERMAN,1
JOHN E. BERGMANN,2 3 AILEEN M. CLEARY,1 MICHAEL J. PAUL J. FUSCO,2 and LEONARD CHESS1
YELLIN,'
ABSTRACT Sulfated polyesters (SP) that inhibit HIV infection interact with both the gp 120 binding region of CD4 molecules and with the v3 domain loop of gp 120 molecules (gpl20/v3) but the contributions of these interactions to the inhibition of HIV e ;u-mediated fusion are presently unclear. In order to characterize the molecular mechanisms by which SP inhibit HIV env-mediated fusion, we studied the effect of SP treatment on env-mediated fusion of CD4+ cells driven by recombinant vaccinia virus (vPE-16) and on the binding of anti-HIV MAbs to cellular gpl20 or purified, rgpl20. SP were more effective than neutralizing anti-gpl20/v3 MAbs in inhibiting env-mediated fusion. In addition, SP interacted with the v3 loop of gpl20 to inhibit the binding of the neutralizing MAb 9284 but not the binding of 9305, a neutralizing anti-gpl20/v3 MAb that binds to an adjacent epitope. Because SP are polyanions, we compared the chemical properties of the SP-gpl20/v3 and SP-CD4 interactions. Whereas the ability of SP to inhibit the binding of MAb 9284 and rgpl20 was relatively independent of NaCl concentrations, the ability of SP to interfere with rCD4-rgpl20 binding depended on the NaCl concentration and was maximal at low NaCl concentrations. In addition, the SP-gpl20 interaction was found to be reversible, in contrast to the SP-/CD4 interaction which was previously shown to be relatively irreversible at low salt. These data are consistant with the notions that the interaction of SP with CD4 is primarily electrostatic, but that the interaction of SP with gpl20 has complex characteristics that implicate a role for protein conformation.
However, other regions of gpl20, besides those that partici¬ pate in CD4 binding are also involved in HIV infectivity.16 HIV
INTRODUCTION cells by free virus or cell-cell fusion expression of HIV env and cellular CD4.1-6 The envelope gene of HIV-1 encodes a polyprotein, gpl60, that iscleavedby host proteases into gpl 20 and gp41, which exist on the cell surface as noncovalently bound complexes.7 CD4 is a cell surface glycoprotein on human lymphocytes, macrophages, and other cells.8·9 HIV coat protein, gpl20, binds to CD4 with high avidity310 and binding to CD4 is necessary for efficient viral entry because agents that block gpl20-CD4 interaction inhibit HIV infection.1-6"-"
HIV-1 depends
INFECTION of on
infection in cell culture is inhibited by polyclonal'721 and monoclonal22-24 antibodies that react with the third variable region (v3) of gpl20, despite the fact that such antibodies do not inhibit gpl20-CD4 binding.25 Interestingly, divergent HIV strains have sequence variability in this region and strainspecific antibodies exhibit strain-restricted neutralization.21·26 In addition, HIV-infected individuals who have antibodies that react with the v3 sequences of particular HIV strains have been shown to be infected with HIV strains that have distinct and non-neutralized sequences in this region suggesting that HIV
Departments of 'Medicine and 2Anatomy and Cell Biology, Columbia University, 'Present address: Centocor, 244 Great Valley Parkway, Málveme, PA 19355.
1599
New York, NY 10032.
LEDERMAN ET AL.
1600 may evolve under the selective pressure of anti-v3 antibodies in
AIDS Research and Human Retroviruses 1992.8:1599-1610. Downloaded from online.liebertpub.com by Ucsf Library University of California San Francisco on 01/11/15. For personal use only.
VIVO.
Chemicals
27
Certain sulfated polyesters (SP) potently inhibit the infection of CD4+ cells by diverse HIV strains at the stages of virus adsorption and/or entry but the molecular targets of these agents are not precisely known.28-32 SP are known to inhibit the cellular association of radiolabelled HIV particles30'31 and the cellular incorporation of HIV surface antigens.33 The possibility that SP inhibit HIV infection by inhibiting CD4-gpl20 binding was suggested by experiments in which DS or heparin inhibited the attachment of rgp 120 to rCD4.34 Because the interaction of SP with rCD4 was relatively irreversible, it was possible to show that SP interact with rCD4 and not rgp 120 to inhibit their binding. In addition, direct binding studies using SP immobi¬ lized on fibers showed that SP bound CD4 molecules, but not gpl20 from cell lysates.35 SP also interact with gp 120 molecules to inhibit the binding of an anti-gpl20 MAb, 928423 to gpl20 on HIV-infected cells.36 The fact that the MAb 9284 interacts with the v3 loop of gp 120,22'23 a known neutralization epitope on gp 120, suggested that this interaction may have mechanistic significance in the ability of SP to inhibit HIV-1-mediated fusion.36 In addition, a recent report using a different set of anti-gp 120 MAbs confirmed the observation that SP inhibit the binding of anti-gp 120/v3 MAbs.37 If the interaction of SP with the neutralization epitope, v3 on gpl20 is important in the ability of SP to inhibit HIV infection, several lines of evidence suggest that this interaction
reversible.30'36 In the present study we first show in an HIV env expression system driven by recombinant vaccinia virus that SP were more is
effective than anti-gp 120/v3 MAbs in inhibiting gpl20-mediated fusion driven by vaccinia-env recombinant virus. SP inter¬ acted with gpl20/v3 to inhibit the binding of the anti-gp 120 MAb 9284 but not the binding of 9305, a neutralizing antigpl20/v3 MAb that binds to an adjacent epitope.23·24 Because SP are polyanions we compared the chemical properties of the interactions of SP with gpl20 and CD4. The ability of SP to inhibit the gpl20-MAb 9284 interaction was relatively insensi¬ tive to NaCl concentration, whereas the ability of SP to inhibit CD4-gpl20 binding was maximal at low NaCl concentration and was significantly attenuated at physiological NaCl concen¬ tration. These data demonstrate that SP interact with the gpl20 and CD4 molecules by interactions that are chemically distinct.
MATERIALS AND METHODS
Reagents Recombinant proteins: The rCD4 was the gift of Dr. Vicki Sato (Biogen Corp., Cambridge, MA). The preparation of rCD4 '2 The rgpl20 used was either a gift was previously described. from Biogen or purchased from American BioTechnologies (Cambridge, MA). The Biogen rgpl20 is a fusion protein combining a portion of tissue plasminogen activator with the HIV-1 (strain HXB-CG2) env gene segment corresponding to amino acids 44-500. The rgp 120 fusion protein was expressed in a baculovirus system and affinity purified on an rCD4 column (personal communication, Dr. Werner Meier, Biogen).
The following chemicals were purchased from Sigma (St. Louis, MO): dextran sulfate avg. 8,000 kD avg, dextran avg. 9,400 kD, chondroitin sulfate, and sodium azide. Bovine serum albumin (BSA) fraction V was purchased from U.S. Biochemi¬ cal Corp. (Cleveland, OH). Potassium phosphate monoblasic
(KH2P04), potassium phosphatate dibasic (K2HP04), and so¬ dium chloride (NaCl) were purchased from Fisher Scientific (Fair Lawn, NJ).
Monoclonal antibodies The murine IgG, MAbs; GS LAVgp 110-1 (110-1), and GS LAVgp41-l (41-1) were gifts of Dr. Lynn Goldstein (Genetic Systems, Seattle, WA). The murine IgG, anti-gpl20 MAbs, 9284, and NEA-9305 (termed 0.5ß in ref 22) were purchased from DuPont (MA). The MAbs; OKT4 (IgG2 anti-human CD4), OKT1 (IgG, anti-human CD5), and AB 2.06 (IgG, anti-human MHC Class II) were produced by hybridomas available from the American Type Culture Collection (Rockville, MD) and were purified from ascites fluid on protein A columns (Biorad, Rockville Center, NY). The MAb OKT4A-FITC was purchased from Ortho Pharmaceuticals (Raritan,NJ) and OKT4A was a gift of Mary Ann Talle (Ortho).
Cell lines Hela-T4 6c38 was a gift from Bruce Chesebro, Rocky Moun¬ tain National Labs (Hamilton, MT). A is a lymphoblastoid cell line.39 The cells were grown in Iscove's Dulbecco Modified Medium (IMDM) (Gibco, Grand Island, NY), and 10% fetal calf serum (Hyclone, CO) supplemented with 10 mg/mL peni¬
cillin-streptomycin (Gibco). Recombinant vaccinia virus The vaccinia viruses, vPE-1640 and vSC-841 and human tkcells42 were gifts from Dr. Bernard Moss, Laboratory of Viral Diseases, NIH (Bethesda, MD). The viruses were grown in Hela cell monolayers, released from cells by sonication, and titered on
tk~ cells.
Cell culture HeLa cells were maintained in a 50:50 mixture of Dulbecco's modified Eagles' medium and F-12 nutrient medium with 5% fetal calf serum. Human tk~ cells were grown in Dulbecco's modified Eagles' medium with 10% fetal calf serum. All cells were grown at 37°C in an atmosphere of 5% C02.
Fusion assay Sterile filtered poly-L-lysine (500 µ at 20 µg/mL in water from Sigma Chemical Co. P-2636) was added to each well of a Multi well™ tissue culture plate (Falcon #3047, Becton Dickenson, Lincoln Park, NJ). The polylysine was allowed to adsorb to the plastic for 30 min at room temperature. The liquid was removed and the dishes were allowed to dry for 1 h under a germicidal lamp. Hela6C cells were plated at a density of 3 x 105 cells per well and allowed to grow for 40 h at 37°C. The
INTERACTION OF SULFATED PE WITH
gpl20
v3 LOOP AND CD4 ARE CHEMICALLY DISTINCT
AIDS Research and Human Retroviruses 1992.8:1599-1610. Downloaded from online.liebertpub.com by Ucsf Library University of California San Francisco on 01/11/15. For personal use only.
cells were infected with vPE16 (a generous gift from Dr. Bernard Moss) at a multiplicity of 10 pfu/cell (assayed on human TK- cells as described by Mackett et al.42 After adsorption for 1 h at 37°C in 500 µ of medium, the virus inoculum was replaced with 1 mL of fresh medium containing SP at the concentration indicated in the text. The cells were incubated for an additional 5-7 h until fusion was evident in the control cultures.
ELISA
binding assays
The adherence of rgp 120 and rT4 to the U bottoms on untreated polystyrene plates (Corning, Corning, NY) was achieved by incubation of 50 µ of protein solutions at the given concentrations in 1 X phosphate-buffered saline (PBS) (145 mM NaCl, 77 mM K2HP04, and 23.5 mM KH2P04, pH 7.4.) for 2 h at room temperature. Control wells were treated with PBS without added protein. Wells were washed three times with PBS containing 0.05% Tween-20 (PBS-Tween) followed by block¬ ing with 0.15 mL of 1 % BSA at 4°C overnight. The wells were then washed three times in PBS-Tween. In experiments studying inhibitors, 50 µ of solutions containing the inhibitors plus the ligand were added to control and protein-coated wells and allowed to react for 30 min at room temperature. In experiments studying inhibitor pretreatment, the inhibitor was removed, the wells were washed three times, and the ligand was added for 30 min. In salt titration experiments, stock solutions of inhibitors ( 1 mM in 1 x PBS) were diluted 1:50 into test solutions of varying concentrations of PBS without BSA (1 x, 0.75 x, 0.5 x, 0.25 x, 0.1 x). Control solutions contained 1% BSA in H20 (BSA-H20)or 1 x PBS (BSA-PBS). To detect bound ligands, 50 µ of the MAbs OKT4 (5 µg/ml) or anti-gp 120 MAb in 1% BSA were added to wells and reacted for 60 min at 20°C. The wells were washed four times before the addition of 50 µ of 1:1000 affinity-purified goat anti-mouse IgG conjugated to alkaline phosphatase (Bio-rad Laboratories, Rich¬ mond, CA) in 1% BSA. After 60 min, wells were washed four times before the addition of 50 µ of substrate solution (pnitrophenylphosphate) prepared according to the manufacturer's instructions (Bio-rad). The enzyme reaction was allowed to progress at 20°C for 20-60 min before absorption at 405 nm was measured in a ELISA plate analyzer (VMAX, Molecular De¬ vices Corporation, Palo Alto, CA). Assays were performed in triplicate. Standard deviations (SD) were obtained by the STDEV formula (Excel, Microsoft, Redmond, WA). The stan¬ dard deviations of the subtracted means (A-B) obtained from the formula S.D. (A-B) SQR (SD(A)2 + SD(B)2). The percent inhibition of binding obtained from the formula % inhibition 100 x (C-I)/C where C (control) is the amount of binding in the absence of inhibitor and I is the amount of binding in the presence of inhibitor. =
=
Expression and analysis of cell surface gpl 60 BA cells were infected with 40 pfu/cell of the indicated vaccinia virus for 15 h shaking at 37°C. The cells were washed in IMDM media with 10% fetal calf serum. The effect of SP on surface gpl20 was studied by reacting these cells for 30 min at 20°C with the indicated concentrations of SP in 0.2% sodium azide containing IMDM media with 10% fetal calf serum. The cells were washed two times and reacted with saturating concen¬
1601
trations of the indicated MAbs followed by anti-mouse IgG coupled to FITC. Dead cells were excluded from the analysis by gating based on forward and side scatter before the fluorescence signals from 5000 living cells were collected on a FACscan (Becton Dickinson, Mountainview, CA).
FACS
analysis of rgpl20 binding
to
cellular CD4
CD4+ Jurkat (B 2.7 clone) cells were washed and resusat 8 x 106 cells/mL in PBS of varying tonicity 145 mM NaCl, 77 mM K2HP04, and 23.5 mM (1 x KH2P04, pH 7.4) and 25 µ aliquots (200,000 cells) were distributed to wells of a round-bottom microtiter plate (Titerek). To appropriate wells, 12.5 µ of solutions of indicated PBS tonicity containing 1 mM polyester (Dextran MW 9600 or Dextran Sulfate MW 8000) or control (no polyester) were added to yield a final polyester concentration of 250 µ . Next, 0.5 ng
pended
=
of rgp 120 or control (no gpl20) in 5 µ was added for a final concentration of 10 µg/ml rgpl20. The mixtures were brought up to a final volume of 50 µ with 7.5 µ of PBS with indicated tonicity. Greater than 95% of cells were viable by trypan blue exclusion after treatment at all PBS tonicity conditions indicated (down to 0.25 x ). The cells were incubated at room temperature (25°C) for 2 h, washed, stained with 5 µ of either OKT4-PE or OKT4a-FITC (Ortho Pharmaceuticals) and 5000 cells were analyzed by FACS as described above.
RESULTS Sulfated polyesters (SP) such as DS and PPS inhibit the HIV-1 infection of CD4+ cells, but the level at which these agents act is not precisely known. To study SP inhibition of HIV envmediated fusion we used an env recombinant vaccinia virus to induce fusion of CD4+ Hela cells. CD4+ Hela cells (clone 6c) were infected with the recombinant vaccinia virus (vPE-16) which expresses both HIV-1 env (gpl60) and ß-galactosidase or with a control vaccinia virus (vSC-8) that only expresses ß-galactosidase. As expected, env expression resulted in the cell-cell fusion of the CD4+ Hela cell monolayer and the formation of multinucleated giant cells (Fig. 1.). The sulfated polyesters PPS (Fig. 1) or DS (Fig. 2.) inhibited the envmediated fusion whereas non-sulfated controls had no effect (Figs. 1,2). In addition, treatment with DS or PPS did not inhibit the expression of ß-galactosidase by the vPE-16 infected HelaCD4+ cells indicating that the SP inhibited gpl20 mediated fusion and not the vaccinia infection under these conditions (not shown). Taken together, these results demonstrated that SP inhibited HIV env mediated cell-cell fusion of CD4+ cells driven by env recombinant vaccinia virus. The ability of SP to inhibit HIV env-mediated fusion was next compared to the ability of two anti-gp 120 MAbs that are known to inhibit HIV infection, 9284 and 9305. Although both 9284 and 9305 bind to adjacent and partially overlapping peptide epitopes from the v3 loop of gpl 20, 9305 was found to be more potent than 9284 in inhibiting other assays of HIV infectivity.23 When the CD4+ Hela cell monolayer fusion assay was allowed to proceed to points of massive cell fusion, SP conferred cytoprotection (Fig. 1.) but no protective effect of anti-gp 120
LEDERMAN ET AL.
AIDS Research and Human Retroviruses 1992.8:1599-1610. Downloaded from online.liebertpub.com by Ucsf Library University of California San Francisco on 01/11/15. For personal use only.
1602
Media
PPS
2.5/i.M Ä
Dextran
IO/aM
FIG. 1. SP inhibit HIV ercv-mediated fusion of Hela-CD4+ (6c clone) infected with vPE-16. Monolayersof Hela-CD4 + -6c were infected with 40 pfu of vPE-16 gpl60 recombinant vaccinia virus and treated with (a) media, (b) PPS, 2.5 µ or (c) Dextran, 10 µ . Cultures were examined for syncytia formation and photographed after 18 hr. MAbs was observed. However, when the monolayer fusion assay was allowed to proceed to a moderate level of syncytia, the anti-gp 120 MAb 930522 conferred protection of the CD4+ Hela monolayer from fusion in a dose-dependent manner (Fig. 2). In contrast, the anti-gp 120 MAb 928423 did not protect against snv-mediated fusion in this assay. The observation that 9305 was more potent than 9284 in inhibiting HIV env-mediated fusion is similar to their reported relative potencies in inhibiting HIV infectivity.23 However, these data demonstrate that SP are more potent than neutralizing anti-gp 120 MAbs in the monolayer CO4-env fusion model, particularly after progressive expression of the env proteins. In order to address the molecular mechanism by which SP inhibit HIV adsorption, we studied the effects of PPS treatment
the structure of gpl20 and gp41 on the surface of cells expressing the HIV-1 env gene. The infection of A cells, a lymphoblastoid cell line, with vPE16, but not VSC8, a control vaccinia virus, resulted in the cell surface expression of gpl20 and gp41 as detected by the specific immunofluorescence of MAbs in a cytofluorograph (Fig. 3). In order to determine whether SP interact with gpl20, we studied the effect of SP treatment on the binding of a panel of anti-gp 120 and anti-gp41 IgG, MAbs to vPE-16 infected BA cells expressing env (gpl20/gp41). In these experiments, env expressing cells were reacted with PPS and then washed prior to the addition of MAbs. Treatment of env expressing cells with PPS, did not affect the binding of an anti-gp41 MAb suggesting on
that PPS treatment did not down-modulate the surface exprès-
INTERACTION OF SULFATED PE WITH Uninfected control
Dextran IOOuM
AIDS Research and Human Retroviruses 1992.8:1599-1610. Downloaded from online.liebertpub.com by Ucsf Library University of California San Francisco on 01/11/15. For personal use only.
(no vPE-16)
gpl20 v3
LOOP AND CD4 ARE CHEMICALLY DISTINCT
1603
mAb9305-
¿g/ml
Dextran sulfate
O.I56MM
0.313/iM
0.625/iM
1.25/j.M
ICVM
FIG. 2. Effect of SP or anti-gp 120 MAbs on HIV ercv-mediated fusion of CD4+ Hela cells. Monolayers of Hela-CD4+-6c were infected with 40 pfu of vPE-16 recombinant vaccinia virus, treated with the indicated agents (Dextran, Dextran sulfate, or MAb
9305) and photographed after 15 h.
sion oí env (Fig. 3b). Similarly, the binding of an anti-gpl20 MAb, 110-1 which binds to gpl20 close to the CD4 binding region (Lynn Goldstein, personal communication) was mini¬ mally affected by PPS treatment (Fig. 3c). In addition, PPS treatment minimally affected the binding of the anti-v3 gpl20 MAb 9305 (Fig. 3e). Taken together, these data suggested that the surface expression of gpl20 was not down-modulated by PPS treatment and in addition, these data were not consistant with the interpretation that PPS treatment resulted in the detach¬ ment of gpl20 from gp41 molecules. In contrast, PPS treatment drastically decreased the binding of the anti-gpl20 MAb, 9284, despite the proximity of the 9284 and 9305 epitopes on the v3 domain23 (Fig. 3d). Taken together, these observations sug¬ gested that PPS interacted with a specific site on gpl20 overly¬ ing the 9284 epitope. The fact that this epitope represents a target of neutralizing antibodies suggested that the interaction of SP with gpl20 at this site may contribute to their ability to inhibit HIV env-mediated fusion. To investigate whether the ability of sulfated and nonsulfated polymers to inhibit 9284 binding correlated with the ability of such compounds to inhibit HIV infection, the effect of poly¬ meric compounds on 9284 binding was studied. The binding of 9284 to env expressing cells was not affected by treatment with dextran (9.4 kD), a nonsulfated polysaccharide, or chondroitan sulfate, a SP that is not active against HIV infection in vitro. The correlation between the ability of sulfated polyesters to inhibit 9284 binding and HIV infection strengthened the suggestion that
SP interaction with the v3 domain on gpl20 may contribute to the ability of these agents to inhibit HIV infection. In order to directly determine whether SP interact with gp 120 in the absence of other cell surface structures, the effect of SP on the binding of anti-gp 120 mABs to rgp 120 was studied in solid-phase assays. PPS inhibited the binding of the anti-gp 120 MAb, 9284, but not the other anti-gpl20 MAbs, to immobilized rgpl20 (Fig. 4a,b). The fact that PPS reacted with rgpl20 to inhibit the binding of MAb 9284 but not other isotype control anti-gp 120 MAbs in a similar manner to the experiments on env expressing cells, suggested that SP interact with a specific epitope of gpl 20. In addition to the interaction of SP with gp 120 presented here, it is also known that SP also bind to CD4 molecules to inhibit the interaction of CD4 and gpl20.35'36 In order to compare the chemical characteristics of SP interactions with rCD4 or gpl20, the next series of experiments studied the effect of varying NaCl concentrations on the SP-mediated inhibition of binding of MAb 9284 to gpl20 and rCD4 to rgpl20. The PPS-mediated inhibi¬ tion of 9284 binding to rgp 120 was independent of NaCl concentrations between 0 and 150 mM NaCl (Fig. 5a). In contrast, the ability of SP to interfere with rCD4-gpl20 binding was sensitive to NaCl concentration and was maximal at low NaCl concentration (Fig. 5b). Importantly, the ability of SP to inhibit rCD4-rgpl20 binding was significantly decreased at physiological salt concentrations. In addition, we studied the ability of SP to inhibit the binding of rgp 120 to CD4+ cells at
LEDERMAN ET AL.
1604
300
AIDS Research and Human Retroviruses 1992.8:1599-1610. Downloaded from online.liebertpub.com by Ucsf Library University of California San Francisco on 01/11/15. For personal use only.
low and physiological NaCl concentrations. Similarly to the results obtained in the solid-phase assays, SP inhibited the binding of rgp 120 to cellular CD4 molecules only at low NaCl concentrations (Fig. 6). Taken together, these results show that SP binding to the 9284 epitope on gpl20 is less sensitive to increases in NaCl concentration than the SP-CD4 interaction. We have previously shown that the SP-CD4 interaction is stable to repeated washing. We, therefore, studied the stability of the SP-gpl20 interaction by comparing the ability of SP to inhibit 9284-gpl20 binding when SP were continuously present in the binding reaction with conditions in which the immobilized rgp 120 molecules were first treated with SP and then washed with SP-free solution. In contrast to the SP-CD4 interaction, the SP-gpl20 interaction was readily reversed by washing the rgp 120 coated surface with SP-free solution (Fig. 7.). Taken together, these data demonstrate that SP interact with CD4 molecules in a relatively irreversible manner that is sensitive to ionic shielding and SP react with gp 120 molecules in a relatively reversible manner that is independent of tonicity in the range studied.
DISCUSSION In the present work we studied the mechanism by which sulfated polyesters inhibit HIV infection at the molecular level. Both DS and PPS, but not nonsulfated control polymers, inhibited the ercv-dependent fusion of CD4+ Hela cells and were more effective than neutralizing anti-gp 120/v3 MAbs. PPS specifically inhibited the binding of the anti-gp 120 MAb 9284 and not other anti-gp 120 MAbs to gpl20 expressed on cell surfaces or to rgp 120 in solid-phase assays. Because previous work has shown that MAb 9284 interacts with the v3 loop of gpl20,22'23 these data suggested that SP interact with the the v3 domain loop of gpl 20 to inhibit 9284 binding. These data do not exclude the possibility that SP might interact with MAb 9284 to inhibit 9284-gpl20 binding. However, taken together with the fact that SP inhibit the binding of other anti-gp 120/v3 MAbs37 these data suggest that SP interact with gpl20/v3 to inhibit the binding of these anti-gpl20/v3 MAbs. MAbs 9284 and 9305 bind to adjacent and slightly overlap¬ ping epitopes on the V3 loop22,23 with similar affinities.23 Whereas PPS potently inhibited 9284 binding to gpl20 on cells expressing env and in solid-phase assays, PPS only slightly inhibited the binding of 9305 to gpl20 on cells expressing env and did not appreciably affect the binding of 9305 to rgp 120 in the solid phase assay. Taken together, these data suggest that SP bind to the V3 loop of gpl20 and inhibit the binding of certain
10°
"IO1
102 FL1
103
104
FIG. 3. SP inhibit the binding of anti-gp 120 MAbs to cell surface gpl20. Shown are FACS histograms (5000 gated cells) of cell fluorescence intensity of BA cells infected with vPE-16 (except where labeled vSC-8). Cells are labeled with (a) 9301, an anti-gp 120 MAb that is not reactive with cell surface gpl20 and 2.06 (anti-la) (indicated by arrows), (b) gp41 -1 (anti-gp41), (c) gpl 10-1 (anti-gpl20), (d) 9284 (anti-gpl20), and (e) 9305
(anti-gpl20). PPS, pentosan polysulfate; DEX, dextran; CS,
Chondroitan sulfate.
INTERACTION OF SULFATED PE WITH
gpl20 v3, LOOP AND CD4 ARE CHEMICALLY DISTINCT
1605
1.2 D
£CO
1.0
AIDS Research and Human Retroviruses 1992.8:1599-1610. Downloaded from online.liebertpub.com by Ucsf Library University of California San Francisco on 01/11/15. For personal use only.
E—
o
0.8o CM
§5
0.6-
C
0.4-
O -O
AIDS RESEARCH AND HUMAN RETROVIRUSES Volume 8, Number 9, 1992 Mary Ann Fiebert, Inc., Publishers
Sulfated Polyester Interactions with the CD4 Molecule and with the Third Variable Loop Domain (v3) of gpl20 Are Chemically Distinct SETH
LEDERMAN,1
JOHN E. BERGMANN,2 3 AILEEN M. CLEARY,1 MICHAEL J. PAUL J. FUSCO,2 and LEONARD CHESS1
YELLIN,'
ABSTRACT Sulfated polyesters (SP) that inhibit HIV infection interact with both the gp 120 binding region of CD4 molecules and with the v3 domain loop of gp 120 molecules (gpl20/v3) but the contributions of these interactions to the inhibition of HIV e ;u-mediated fusion are presently unclear. In order to characterize the molecular mechanisms by which SP inhibit HIV env-mediated fusion, we studied the effect of SP treatment on env-mediated fusion of CD4+ cells driven by recombinant vaccinia virus (vPE-16) and on the binding of anti-HIV MAbs to cellular gpl20 or purified, rgpl20. SP were more effective than neutralizing anti-gpl20/v3 MAbs in inhibiting env-mediated fusion. In addition, SP interacted with the v3 loop of gpl20 to inhibit the binding of the neutralizing MAb 9284 but not the binding of 9305, a neutralizing anti-gpl20/v3 MAb that binds to an adjacent epitope. Because SP are polyanions, we compared the chemical properties of the SP-gpl20/v3 and SP-CD4 interactions. Whereas the ability of SP to inhibit the binding of MAb 9284 and rgpl20 was relatively independent of NaCl concentrations, the ability of SP to interfere with rCD4-rgpl20 binding depended on the NaCl concentration and was maximal at low NaCl concentrations. In addition, the SP-gpl20 interaction was found to be reversible, in contrast to the SP-/CD4 interaction which was previously shown to be relatively irreversible at low salt. These data are consistant with the notions that the interaction of SP with CD4 is primarily electrostatic, but that the interaction of SP with gpl20 has complex characteristics that implicate a role for protein conformation.
However, other regions of gpl20, besides those that partici¬ pate in CD4 binding are also involved in HIV infectivity.16 HIV
INTRODUCTION cells by free virus or cell-cell fusion expression of HIV env and cellular CD4.1-6 The envelope gene of HIV-1 encodes a polyprotein, gpl60, that iscleavedby host proteases into gpl 20 and gp41, which exist on the cell surface as noncovalently bound complexes.7 CD4 is a cell surface glycoprotein on human lymphocytes, macrophages, and other cells.8·9 HIV coat protein, gpl20, binds to CD4 with high avidity310 and binding to CD4 is necessary for efficient viral entry because agents that block gpl20-CD4 interaction inhibit HIV infection.1-6"-"
HIV-1 depends
INFECTION of on
infection in cell culture is inhibited by polyclonal'721 and monoclonal22-24 antibodies that react with the third variable region (v3) of gpl20, despite the fact that such antibodies do not inhibit gpl20-CD4 binding.25 Interestingly, divergent HIV strains have sequence variability in this region and strainspecific antibodies exhibit strain-restricted neutralization.21·26 In addition, HIV-infected individuals who have antibodies that react with the v3 sequences of particular HIV strains have been shown to be infected with HIV strains that have distinct and non-neutralized sequences in this region suggesting that HIV
Departments of 'Medicine and 2Anatomy and Cell Biology, Columbia University, 'Present address: Centocor, 244 Great Valley Parkway, Málveme, PA 19355.
1599
New York, NY 10032.
LEDERMAN ET AL.
1600 may evolve under the selective pressure of anti-v3 antibodies in
AIDS Research and Human Retroviruses 1992.8:1599-1610. Downloaded from online.liebertpub.com by Ucsf Library University of California San Francisco on 01/11/15. For personal use only.
VIVO.
Chemicals
27
Certain sulfated polyesters (SP) potently inhibit the infection of CD4+ cells by diverse HIV strains at the stages of virus adsorption and/or entry but the molecular targets of these agents are not precisely known.28-32 SP are known to inhibit the cellular association of radiolabelled HIV particles30'31 and the cellular incorporation of HIV surface antigens.33 The possibility that SP inhibit HIV infection by inhibiting CD4-gpl20 binding was suggested by experiments in which DS or heparin inhibited the attachment of rgp 120 to rCD4.34 Because the interaction of SP with rCD4 was relatively irreversible, it was possible to show that SP interact with rCD4 and not rgp 120 to inhibit their binding. In addition, direct binding studies using SP immobi¬ lized on fibers showed that SP bound CD4 molecules, but not gpl20 from cell lysates.35 SP also interact with gp 120 molecules to inhibit the binding of an anti-gpl20 MAb, 928423 to gpl20 on HIV-infected cells.36 The fact that the MAb 9284 interacts with the v3 loop of gp 120,22'23 a known neutralization epitope on gp 120, suggested that this interaction may have mechanistic significance in the ability of SP to inhibit HIV-1-mediated fusion.36 In addition, a recent report using a different set of anti-gp 120 MAbs confirmed the observation that SP inhibit the binding of anti-gp 120/v3 MAbs.37 If the interaction of SP with the neutralization epitope, v3 on gpl20 is important in the ability of SP to inhibit HIV infection, several lines of evidence suggest that this interaction
reversible.30'36 In the present study we first show in an HIV env expression system driven by recombinant vaccinia virus that SP were more is
effective than anti-gp 120/v3 MAbs in inhibiting gpl20-mediated fusion driven by vaccinia-env recombinant virus. SP inter¬ acted with gpl20/v3 to inhibit the binding of the anti-gp 120 MAb 9284 but not the binding of 9305, a neutralizing antigpl20/v3 MAb that binds to an adjacent epitope.23·24 Because SP are polyanions we compared the chemical properties of the interactions of SP with gpl20 and CD4. The ability of SP to inhibit the gpl20-MAb 9284 interaction was relatively insensi¬ tive to NaCl concentration, whereas the ability of SP to inhibit CD4-gpl20 binding was maximal at low NaCl concentration and was significantly attenuated at physiological NaCl concen¬ tration. These data demonstrate that SP interact with the gpl20 and CD4 molecules by interactions that are chemically distinct.
MATERIALS AND METHODS
Reagents Recombinant proteins: The rCD4 was the gift of Dr. Vicki Sato (Biogen Corp., Cambridge, MA). The preparation of rCD4 '2 The rgpl20 used was either a gift was previously described. from Biogen or purchased from American BioTechnologies (Cambridge, MA). The Biogen rgpl20 is a fusion protein combining a portion of tissue plasminogen activator with the HIV-1 (strain HXB-CG2) env gene segment corresponding to amino acids 44-500. The rgp 120 fusion protein was expressed in a baculovirus system and affinity purified on an rCD4 column (personal communication, Dr. Werner Meier, Biogen).
The following chemicals were purchased from Sigma (St. Louis, MO): dextran sulfate avg. 8,000 kD avg, dextran avg. 9,400 kD, chondroitin sulfate, and sodium azide. Bovine serum albumin (BSA) fraction V was purchased from U.S. Biochemi¬ cal Corp. (Cleveland, OH). Potassium phosphate monoblasic
(KH2P04), potassium phosphatate dibasic (K2HP04), and so¬ dium chloride (NaCl) were purchased from Fisher Scientific (Fair Lawn, NJ).
Monoclonal antibodies The murine IgG, MAbs; GS LAVgp 110-1 (110-1), and GS LAVgp41-l (41-1) were gifts of Dr. Lynn Goldstein (Genetic Systems, Seattle, WA). The murine IgG, anti-gpl20 MAbs, 9284, and NEA-9305 (termed 0.5ß in ref 22) were purchased from DuPont (MA). The MAbs; OKT4 (IgG2 anti-human CD4), OKT1 (IgG, anti-human CD5), and AB 2.06 (IgG, anti-human MHC Class II) were produced by hybridomas available from the American Type Culture Collection (Rockville, MD) and were purified from ascites fluid on protein A columns (Biorad, Rockville Center, NY). The MAb OKT4A-FITC was purchased from Ortho Pharmaceuticals (Raritan,NJ) and OKT4A was a gift of Mary Ann Talle (Ortho).
Cell lines Hela-T4 6c38 was a gift from Bruce Chesebro, Rocky Moun¬ tain National Labs (Hamilton, MT). A is a lymphoblastoid cell line.39 The cells were grown in Iscove's Dulbecco Modified Medium (IMDM) (Gibco, Grand Island, NY), and 10% fetal calf serum (Hyclone, CO) supplemented with 10 mg/mL peni¬
cillin-streptomycin (Gibco). Recombinant vaccinia virus The vaccinia viruses, vPE-1640 and vSC-841 and human tkcells42 were gifts from Dr. Bernard Moss, Laboratory of Viral Diseases, NIH (Bethesda, MD). The viruses were grown in Hela cell monolayers, released from cells by sonication, and titered on
tk~ cells.
Cell culture HeLa cells were maintained in a 50:50 mixture of Dulbecco's modified Eagles' medium and F-12 nutrient medium with 5% fetal calf serum. Human tk~ cells were grown in Dulbecco's modified Eagles' medium with 10% fetal calf serum. All cells were grown at 37°C in an atmosphere of 5% C02.
Fusion assay Sterile filtered poly-L-lysine (500 µ at 20 µg/mL in water from Sigma Chemical Co. P-2636) was added to each well of a Multi well™ tissue culture plate (Falcon #3047, Becton Dickenson, Lincoln Park, NJ). The polylysine was allowed to adsorb to the plastic for 30 min at room temperature. The liquid was removed and the dishes were allowed to dry for 1 h under a germicidal lamp. Hela6C cells were plated at a density of 3 x 105 cells per well and allowed to grow for 40 h at 37°C. The
INTERACTION OF SULFATED PE WITH
gpl20
v3 LOOP AND CD4 ARE CHEMICALLY DISTINCT
AIDS Research and Human Retroviruses 1992.8:1599-1610. Downloaded from online.liebertpub.com by Ucsf Library University of California San Francisco on 01/11/15. For personal use only.
cells were infected with vPE16 (a generous gift from Dr. Bernard Moss) at a multiplicity of 10 pfu/cell (assayed on human TK- cells as described by Mackett et al.42 After adsorption for 1 h at 37°C in 500 µ of medium, the virus inoculum was replaced with 1 mL of fresh medium containing SP at the concentration indicated in the text. The cells were incubated for an additional 5-7 h until fusion was evident in the control cultures.
ELISA
binding assays
The adherence of rgp 120 and rT4 to the U bottoms on untreated polystyrene plates (Corning, Corning, NY) was achieved by incubation of 50 µ of protein solutions at the given concentrations in 1 X phosphate-buffered saline (PBS) (145 mM NaCl, 77 mM K2HP04, and 23.5 mM KH2P04, pH 7.4.) for 2 h at room temperature. Control wells were treated with PBS without added protein. Wells were washed three times with PBS containing 0.05% Tween-20 (PBS-Tween) followed by block¬ ing with 0.15 mL of 1 % BSA at 4°C overnight. The wells were then washed three times in PBS-Tween. In experiments studying inhibitors, 50 µ of solutions containing the inhibitors plus the ligand were added to control and protein-coated wells and allowed to react for 30 min at room temperature. In experiments studying inhibitor pretreatment, the inhibitor was removed, the wells were washed three times, and the ligand was added for 30 min. In salt titration experiments, stock solutions of inhibitors ( 1 mM in 1 x PBS) were diluted 1:50 into test solutions of varying concentrations of PBS without BSA (1 x, 0.75 x, 0.5 x, 0.25 x, 0.1 x). Control solutions contained 1% BSA in H20 (BSA-H20)or 1 x PBS (BSA-PBS). To detect bound ligands, 50 µ of the MAbs OKT4 (5 µg/ml) or anti-gp 120 MAb in 1% BSA were added to wells and reacted for 60 min at 20°C. The wells were washed four times before the addition of 50 µ of 1:1000 affinity-purified goat anti-mouse IgG conjugated to alkaline phosphatase (Bio-rad Laboratories, Rich¬ mond, CA) in 1% BSA. After 60 min, wells were washed four times before the addition of 50 µ of substrate solution (pnitrophenylphosphate) prepared according to the manufacturer's instructions (Bio-rad). The enzyme reaction was allowed to progress at 20°C for 20-60 min before absorption at 405 nm was measured in a ELISA plate analyzer (VMAX, Molecular De¬ vices Corporation, Palo Alto, CA). Assays were performed in triplicate. Standard deviations (SD) were obtained by the STDEV formula (Excel, Microsoft, Redmond, WA). The stan¬ dard deviations of the subtracted means (A-B) obtained from the formula S.D. (A-B) SQR (SD(A)2 + SD(B)2). The percent inhibition of binding obtained from the formula % inhibition 100 x (C-I)/C where C (control) is the amount of binding in the absence of inhibitor and I is the amount of binding in the presence of inhibitor. =
=
Expression and analysis of cell surface gpl 60 BA cells were infected with 40 pfu/cell of the indicated vaccinia virus for 15 h shaking at 37°C. The cells were washed in IMDM media with 10% fetal calf serum. The effect of SP on surface gpl20 was studied by reacting these cells for 30 min at 20°C with the indicated concentrations of SP in 0.2% sodium azide containing IMDM media with 10% fetal calf serum. The cells were washed two times and reacted with saturating concen¬
1601
trations of the indicated MAbs followed by anti-mouse IgG coupled to FITC. Dead cells were excluded from the analysis by gating based on forward and side scatter before the fluorescence signals from 5000 living cells were collected on a FACscan (Becton Dickinson, Mountainview, CA).
FACS
analysis of rgpl20 binding
to
cellular CD4
CD4+ Jurkat (B 2.7 clone) cells were washed and resusat 8 x 106 cells/mL in PBS of varying tonicity 145 mM NaCl, 77 mM K2HP04, and 23.5 mM (1 x KH2P04, pH 7.4) and 25 µ aliquots (200,000 cells) were distributed to wells of a round-bottom microtiter plate (Titerek). To appropriate wells, 12.5 µ of solutions of indicated PBS tonicity containing 1 mM polyester (Dextran MW 9600 or Dextran Sulfate MW 8000) or control (no polyester) were added to yield a final polyester concentration of 250 µ . Next, 0.5 ng
pended
=
of rgp 120 or control (no gpl20) in 5 µ was added for a final concentration of 10 µg/ml rgpl20. The mixtures were brought up to a final volume of 50 µ with 7.5 µ of PBS with indicated tonicity. Greater than 95% of cells were viable by trypan blue exclusion after treatment at all PBS tonicity conditions indicated (down to 0.25 x ). The cells were incubated at room temperature (25°C) for 2 h, washed, stained with 5 µ of either OKT4-PE or OKT4a-FITC (Ortho Pharmaceuticals) and 5000 cells were analyzed by FACS as described above.
RESULTS Sulfated polyesters (SP) such as DS and PPS inhibit the HIV-1 infection of CD4+ cells, but the level at which these agents act is not precisely known. To study SP inhibition of HIV envmediated fusion we used an env recombinant vaccinia virus to induce fusion of CD4+ Hela cells. CD4+ Hela cells (clone 6c) were infected with the recombinant vaccinia virus (vPE-16) which expresses both HIV-1 env (gpl60) and ß-galactosidase or with a control vaccinia virus (vSC-8) that only expresses ß-galactosidase. As expected, env expression resulted in the cell-cell fusion of the CD4+ Hela cell monolayer and the formation of multinucleated giant cells (Fig. 1.). The sulfated polyesters PPS (Fig. 1) or DS (Fig. 2.) inhibited the envmediated fusion whereas non-sulfated controls had no effect (Figs. 1,2). In addition, treatment with DS or PPS did not inhibit the expression of ß-galactosidase by the vPE-16 infected HelaCD4+ cells indicating that the SP inhibited gpl20 mediated fusion and not the vaccinia infection under these conditions (not shown). Taken together, these results demonstrated that SP inhibited HIV env mediated cell-cell fusion of CD4+ cells driven by env recombinant vaccinia virus. The ability of SP to inhibit HIV env-mediated fusion was next compared to the ability of two anti-gp 120 MAbs that are known to inhibit HIV infection, 9284 and 9305. Although both 9284 and 9305 bind to adjacent and partially overlapping peptide epitopes from the v3 loop of gpl 20, 9305 was found to be more potent than 9284 in inhibiting other assays of HIV infectivity.23 When the CD4+ Hela cell monolayer fusion assay was allowed to proceed to points of massive cell fusion, SP conferred cytoprotection (Fig. 1.) but no protective effect of anti-gp 120
LEDERMAN ET AL.
AIDS Research and Human Retroviruses 1992.8:1599-1610. Downloaded from online.liebertpub.com by Ucsf Library University of California San Francisco on 01/11/15. For personal use only.
1602
Media
PPS
2.5/i.M Ä
Dextran
IO/aM
FIG. 1. SP inhibit HIV ercv-mediated fusion of Hela-CD4+ (6c clone) infected with vPE-16. Monolayersof Hela-CD4 + -6c were infected with 40 pfu of vPE-16 gpl60 recombinant vaccinia virus and treated with (a) media, (b) PPS, 2.5 µ or (c) Dextran, 10 µ . Cultures were examined for syncytia formation and photographed after 18 hr. MAbs was observed. However, when the monolayer fusion assay was allowed to proceed to a moderate level of syncytia, the anti-gp 120 MAb 930522 conferred protection of the CD4+ Hela monolayer from fusion in a dose-dependent manner (Fig. 2). In contrast, the anti-gp 120 MAb 928423 did not protect against snv-mediated fusion in this assay. The observation that 9305 was more potent than 9284 in inhibiting HIV env-mediated fusion is similar to their reported relative potencies in inhibiting HIV infectivity.23 However, these data demonstrate that SP are more potent than neutralizing anti-gp 120 MAbs in the monolayer CO4-env fusion model, particularly after progressive expression of the env proteins. In order to address the molecular mechanism by which SP inhibit HIV adsorption, we studied the effects of PPS treatment
the structure of gpl20 and gp41 on the surface of cells expressing the HIV-1 env gene. The infection of A cells, a lymphoblastoid cell line, with vPE16, but not VSC8, a control vaccinia virus, resulted in the cell surface expression of gpl20 and gp41 as detected by the specific immunofluorescence of MAbs in a cytofluorograph (Fig. 3). In order to determine whether SP interact with gpl20, we studied the effect of SP treatment on the binding of a panel of anti-gp 120 and anti-gp41 IgG, MAbs to vPE-16 infected BA cells expressing env (gpl20/gp41). In these experiments, env expressing cells were reacted with PPS and then washed prior to the addition of MAbs. Treatment of env expressing cells with PPS, did not affect the binding of an anti-gp41 MAb suggesting on
that PPS treatment did not down-modulate the surface exprès-
INTERACTION OF SULFATED PE WITH Uninfected control
Dextran IOOuM
AIDS Research and Human Retroviruses 1992.8:1599-1610. Downloaded from online.liebertpub.com by Ucsf Library University of California San Francisco on 01/11/15. For personal use only.
(no vPE-16)
gpl20 v3
LOOP AND CD4 ARE CHEMICALLY DISTINCT
1603
mAb9305-
¿g/ml
Dextran sulfate
O.I56MM
0.313/iM
0.625/iM
1.25/j.M
ICVM
FIG. 2. Effect of SP or anti-gp 120 MAbs on HIV ercv-mediated fusion of CD4+ Hela cells. Monolayers of Hela-CD4+-6c were infected with 40 pfu of vPE-16 recombinant vaccinia virus, treated with the indicated agents (Dextran, Dextran sulfate, or MAb
9305) and photographed after 15 h.
sion oí env (Fig. 3b). Similarly, the binding of an anti-gpl20 MAb, 110-1 which binds to gpl20 close to the CD4 binding region (Lynn Goldstein, personal communication) was mini¬ mally affected by PPS treatment (Fig. 3c). In addition, PPS treatment minimally affected the binding of the anti-v3 gpl20 MAb 9305 (Fig. 3e). Taken together, these data suggested that the surface expression of gpl20 was not down-modulated by PPS treatment and in addition, these data were not consistant with the interpretation that PPS treatment resulted in the detach¬ ment of gpl20 from gp41 molecules. In contrast, PPS treatment drastically decreased the binding of the anti-gpl20 MAb, 9284, despite the proximity of the 9284 and 9305 epitopes on the v3 domain23 (Fig. 3d). Taken together, these observations sug¬ gested that PPS interacted with a specific site on gpl20 overly¬ ing the 9284 epitope. The fact that this epitope represents a target of neutralizing antibodies suggested that the interaction of SP with gpl20 at this site may contribute to their ability to inhibit HIV env-mediated fusion. To investigate whether the ability of sulfated and nonsulfated polymers to inhibit 9284 binding correlated with the ability of such compounds to inhibit HIV infection, the effect of poly¬ meric compounds on 9284 binding was studied. The binding of 9284 to env expressing cells was not affected by treatment with dextran (9.4 kD), a nonsulfated polysaccharide, or chondroitan sulfate, a SP that is not active against HIV infection in vitro. The correlation between the ability of sulfated polyesters to inhibit 9284 binding and HIV infection strengthened the suggestion that
SP interaction with the v3 domain on gpl20 may contribute to the ability of these agents to inhibit HIV infection. In order to directly determine whether SP interact with gp 120 in the absence of other cell surface structures, the effect of SP on the binding of anti-gp 120 mABs to rgp 120 was studied in solid-phase assays. PPS inhibited the binding of the anti-gp 120 MAb, 9284, but not the other anti-gpl20 MAbs, to immobilized rgpl20 (Fig. 4a,b). The fact that PPS reacted with rgpl20 to inhibit the binding of MAb 9284 but not other isotype control anti-gp 120 MAbs in a similar manner to the experiments on env expressing cells, suggested that SP interact with a specific epitope of gpl 20. In addition to the interaction of SP with gp 120 presented here, it is also known that SP also bind to CD4 molecules to inhibit the interaction of CD4 and gpl20.35'36 In order to compare the chemical characteristics of SP interactions with rCD4 or gpl20, the next series of experiments studied the effect of varying NaCl concentrations on the SP-mediated inhibition of binding of MAb 9284 to gpl20 and rCD4 to rgpl20. The PPS-mediated inhibi¬ tion of 9284 binding to rgp 120 was independent of NaCl concentrations between 0 and 150 mM NaCl (Fig. 5a). In contrast, the ability of SP to interfere with rCD4-gpl20 binding was sensitive to NaCl concentration and was maximal at low NaCl concentration (Fig. 5b). Importantly, the ability of SP to inhibit rCD4-rgpl20 binding was significantly decreased at physiological salt concentrations. In addition, we studied the ability of SP to inhibit the binding of rgp 120 to CD4+ cells at
LEDERMAN ET AL.
1604
300
AIDS Research and Human Retroviruses 1992.8:1599-1610. Downloaded from online.liebertpub.com by Ucsf Library University of California San Francisco on 01/11/15. For personal use only.
low and physiological NaCl concentrations. Similarly to the results obtained in the solid-phase assays, SP inhibited the binding of rgp 120 to cellular CD4 molecules only at low NaCl concentrations (Fig. 6). Taken together, these results show that SP binding to the 9284 epitope on gpl20 is less sensitive to increases in NaCl concentration than the SP-CD4 interaction. We have previously shown that the SP-CD4 interaction is stable to repeated washing. We, therefore, studied the stability of the SP-gpl20 interaction by comparing the ability of SP to inhibit 9284-gpl20 binding when SP were continuously present in the binding reaction with conditions in which the immobilized rgp 120 molecules were first treated with SP and then washed with SP-free solution. In contrast to the SP-CD4 interaction, the SP-gpl20 interaction was readily reversed by washing the rgp 120 coated surface with SP-free solution (Fig. 7.). Taken together, these data demonstrate that SP interact with CD4 molecules in a relatively irreversible manner that is sensitive to ionic shielding and SP react with gp 120 molecules in a relatively reversible manner that is independent of tonicity in the range studied.
DISCUSSION In the present work we studied the mechanism by which sulfated polyesters inhibit HIV infection at the molecular level. Both DS and PPS, but not nonsulfated control polymers, inhibited the ercv-dependent fusion of CD4+ Hela cells and were more effective than neutralizing anti-gp 120/v3 MAbs. PPS specifically inhibited the binding of the anti-gp 120 MAb 9284 and not other anti-gp 120 MAbs to gpl20 expressed on cell surfaces or to rgp 120 in solid-phase assays. Because previous work has shown that MAb 9284 interacts with the v3 loop of gpl20,22'23 these data suggested that SP interact with the the v3 domain loop of gpl 20 to inhibit 9284 binding. These data do not exclude the possibility that SP might interact with MAb 9284 to inhibit 9284-gpl20 binding. However, taken together with the fact that SP inhibit the binding of other anti-gp 120/v3 MAbs37 these data suggest that SP interact with gpl20/v3 to inhibit the binding of these anti-gpl20/v3 MAbs. MAbs 9284 and 9305 bind to adjacent and slightly overlap¬ ping epitopes on the V3 loop22,23 with similar affinities.23 Whereas PPS potently inhibited 9284 binding to gpl20 on cells expressing env and in solid-phase assays, PPS only slightly inhibited the binding of 9305 to gpl20 on cells expressing env and did not appreciably affect the binding of 9305 to rgp 120 in the solid phase assay. Taken together, these data suggest that SP bind to the V3 loop of gpl20 and inhibit the binding of certain
10°
"IO1
102 FL1
103
104
FIG. 3. SP inhibit the binding of anti-gp 120 MAbs to cell surface gpl20. Shown are FACS histograms (5000 gated cells) of cell fluorescence intensity of BA cells infected with vPE-16 (except where labeled vSC-8). Cells are labeled with (a) 9301, an anti-gp 120 MAb that is not reactive with cell surface gpl20 and 2.06 (anti-la) (indicated by arrows), (b) gp41 -1 (anti-gp41), (c) gpl 10-1 (anti-gpl20), (d) 9284 (anti-gpl20), and (e) 9305
(anti-gpl20). PPS, pentosan polysulfate; DEX, dextran; CS,
Chondroitan sulfate.
INTERACTION OF SULFATED PE WITH
gpl20 v3, LOOP AND CD4 ARE CHEMICALLY DISTINCT
1605
1.2 D
£CO
1.0
AIDS Research and Human Retroviruses 1992.8:1599-1610. Downloaded from online.liebertpub.com by Ucsf Library University of California San Francisco on 01/11/15. For personal use only.
E—
o
0.8o CM
§5
0.6-
C
0.4-
O -O
