AIDS RESEARCH AND HUMAN RETROV1RUSES Volume 8, Number 11, 1992 Mary Ann Liebert, Inc., Publishers

A CD4-Derived Peptide Carrier Blocks Acute HIV-1 Infection In Vitro and Binds to gpl20 in the Presence of Walter-Reed Stage 1-6 HIV+ Sera VICTOR GHETIE, TOM WHEELER, DEVRI SCOTT, JONATHAN W. UHR, and ELLEN S. VITETTA

ABSTRACT A peptide containing amino acid residues 41-84 of the CD4 molecule was synthetized and coupled through a thioether bond to human serum albumin. This conjugate bound to gpl20 with an affinity that was half that of CD4 and blocked the HIV infection in vitro with an efficacy tenfold lower than that of CD4. More importantly, the CD4 peptide-human serum albumin conjugate could bind to gpl20 in the presence of HIV+ sera from 18 Walter Reed stage 1-6 patients.

gpl20. To this end, 18 HIV+ sera which could be assigned to a WR stage based on peripheral blood CD4+ T-cell

INTRODUCTION

carrier and

Soluble envelope glycoprotein (HIV)

CD4 (rCD4) binds to the gpl20 of the human immunodeficiency virus and thereby prevents its attachment to CD4+ cells. Hence, soluble rCD4 can inhibit HIV infection in vitro. '~6 The binding of rCD4 to gp 120 is inhibited by anti-gp 120 antibodies present in the sera of patients with AIDS. These anti-gp 120 antibodies are directed against epitopes near or within the CD4-binding regions of the gpl20 molecule712 and would be expected to limit the therapeutic effectiveness of rCD4 in vivo. To circumvent the problem of inhibition by these antibodies, we have generated a CD4 peptide (residues 41-84) conjugated to ovalbumin (OVA) at a ratio of three peptides per molecule of OVA |CD4(41-84)3-OVA]. We have previously reported that this multivalent conjugate, but not rCD4, binds to gpl20 in the '2 presence of sera from 8 AIDS patients. recombinant

METHODS Since it has been reported by others13'4 that the levels of anti-gp 120 antibodies present in the sera of HIV+ individuals

related to their stage of disease, we evaluated HIV+ sera from individuals of different Walter Reed (WR) stages for their ability to inhibit the interaction between CD4 or CD4(41-84),are

Department of Microbiology and the Cancer Hines Boulevard, Dallas, TX 75235-9048.

clinical disease were obtained from the AIDS Six sera were from WR1 patients, two from WR3, five from WR5, and five from WR6. CD4(41-84) was coupled to human serum albumin (HSA) (Sigma, St. Louis, MO) at a ratio of three peptides per molecule of HSA using sulfocounts and

Repository.

succinimidyl 4-(A/-maleidomethyl)cyclohexane-l carboxylate (Pierce. Rockford, IL) (sulfo-SMCC)

to introduce a thioether bond between HSA and the CD4 peptide. Briefly, 10 mg HSA dissolved in 1 ml of 0.05 M phosphate buffer with 0.003 M Na2EDTA, pH 7.5 (PBE) was treated with 2 mg sulfo-SMCC dissolved in 0.1 ml of PBE. After 1 h at room temperature the excess of sulfo-SMCC was removed by gel filtration on Sephadex G-25 equilibrated with PBE and after concentration to 5 mg/ml, the derivatized HSA was mixed with a ten fold molar excess of reduced CD4(41-84) peptide (alanine labeled with 3H) at 4 mg/ml. The CD4 peptide was reduced with 0.2 M dithiothreitol (DTT) at pH 7.5 in 3 M urea for 1 h and the excess DTT was removed by gel filtration on Sephadex G-10 equilibrated with PBE containing 3M urea. The coupling reaction took place at room temperature for 2-3 h and unreacted peptide was removed by either gel filtration on Sephadex G-50 or HPLC on TSK3000 column, both equilibrated with PBE. The purified conjugate gave a major band a 83 kD on SDS-PAGE which is also consistent with a CD4 peptide/HSA molar ratio of 3 as

Immunobiology Center, University of Texas Southwestern

1945

Medical Center at Dallas, 5323

Harry

GHETIE ET AL.

1946

|

500-

c

o

S .g

400 -

c

_c °

O

40

20

60 Concentration

80

(3 o

100

300-

o

(nM)

FIG. 1. Inhibition of the binding of CD4-peroxidase to gpl20 by rCD4 (•), CD4(41-84)3-HSA ( A ), HSA (o). In the case of CD4(41-84)3-HSA the nM concentration represents the total conjugate concentration and not that of the peptide. Hence, the IC5U is similar to that reported previously using a similar

f

200-

rx

100 -

OVA-based-peptide.12

12

3

Walter Reed

calculated

by

measurement of

specific radioactivity

of the

conjugate.

4

5

6

Stage

FIG. 3. The ability of HIV+ sera to inhibit the rCD4 (•) and CD4(41-84)3-HSA (O) to gpl20.

binding

of

HIV infection but the IC50 was over 1000 times higher (7 x 10~7 M) than that of rCD4. These results suggest that CD4(41-84)3-HSA and rCD4 bind to gpl20 with comparable

RESULTS The ability of CD4(41-84)3-HSA to bind to gpl20 was determined using an ELISA assay to measure its ability to inhibit '2 the binding of rCD4-peroxidase to gp 120. As shown in Figure 1, CD4(41-84)3-HSA inhibited the binding of CD4-peroxidase to gpl20 only twofold less effectively than CD4. More importantly, when CD4 or CD4(41-84)3-HSA were mixed with virus and then added to the cells, CD4(41-84)3-HSA blocked acute infection of H9 cells with HIV-1 (HTLV-IIIB) albeit 11-fold less effectively than rCD4 (IC50 5.5 x 10~9 M vs. 4.8 x 10~10 M) as shown in Figure 2. In these experiments human H9 cells were infected with HIV-1 in the presence or absence of rCD4, CD4(41-84)3-HSA or CD4(41-84) peptide and cultured for 8 days. The cells were washed and recultured for 4 h in the absence of inhibitors. The release of p24 was measured using an ELISA assay kit (Abbott Laboratories). As noted in the Figure 2, even the free CD4(41-84) peptide blocked =

affinity.

Previous studies have shown that HIV+ sera (WR6) block the but not CD4 (41-84)3-OVA to gpl20.,2 We, therefore, determined whether WR sera 1-5 would behave similarly. The ability of HIV+ sera to inhibit the binding of either rCD4 or CD4(41-84)3-HSA to gpl20 was expressed as the reciprocal of the dilution of serum giving 50% inhibition of binding. The binding of both rCD4 and CD4(41-84)3-HSA to gpl20 was measured using an ELISA assay in which either the rCD4 (rCD4 assay) or the rabbit IgG anti-HSA [CD4(41-84)3HSA assay] were labeled with peroxidase. Both procedures have been described previously.12 Figure 3 shows significant inhibition of the binding of rCD4 to gpl20 by sera from patients of all WR stages. In contrast, the same sera inhibited the binding of CD4(41-84)3-HSA to gpl20 10-fold less effectively, confirming our previous results using WR6 sera.12 There also was a correlation between the WR stage of the disease and the ability of the sera from these patients to interfere with the binding of rCD4 to gpl20. Hence, sera from WR1 and WR6 patients inhibited the binding of rCD4 to gp 120, albeit it to a lesser extent than those sera from WR3 and WR5 patients. The decreased ability of WR1 and WR6 sera versus WR3 and WR5 sera to inhibit the binding of rCD4 to gpl20 is in accord with a recent report that the ability of a CDA-Pseudomonas exotoxin to kill HIV-infected target cells was inhibited to a lesser extent when sera from patients with early or late stage disease were used.I4 The ability of the sera to block the binding of rCD4 versus

binding of rCD4,

CD4(41-84)3-HSA to gpl20 was not statistically significant although it was close to significance (Spearman correlation coefficient r

CONCENTRATION

=

0.457; p

=

0.057).

(M)

FIG. 2. Inhibition of infectivity of H9 cells with HTLV-IIIB by rCD4 (o), CD4(41-84)3-HSA (A) and the CD4Í41-84)peptide ( ). Values indicate the IC50. See legend to Figure 1 for the definition of M.

DISCUSSION The inhibitory been attributed to

activity in WR6 HIV+ sera has previously anti-gp 120 antibodies; absorption of the sera

ACTIVITY OF A CD4-PEPTIDE CARRIER CONJUGATE Table 1. Inhibition Patient number: AIDS Repository serum

panel

3 5 6 7 8 9 1 21 19 20 22 24 29 2 4 18 23 26

on

of the

WR stage

1947

Binding of CD4 and CD4 + by HIV Sera

(41-84)3-HSA to gpl20

Inhibition of CD4 binding

(reciprocal liter)

1400 1200 3000 1700 700 4000 1000 1000 900 1500 750 1700 1700 1000 2000 0 800 600

222 118 357 143 167 167 500 235 667 313 435 833 500 286 286 10 10 100

gp 120-Sepharose or protein A-Sepharose removed this activ-

ity.12 Since only small amounts of the WR sera were available

study, we were unable to carry out similar absorption experiments. However, we explored our prior observation in an indirect fashion by measuring the anti-gp 120 titers of the sera using an ELISA technique. Wells of a microtiter plate were coated with recombinant gpl20 and then incubated with various dilutions of the WR sera. The binding of anti-gp 120 antibodies in these sera to gpl20 was determined using a peroxidase-conjugated-goat anti-human Ig. Titers were defined as the reciprocal of the serum dilution giving half maximal absorption at 405 nm. As shown in Table 1 the titers of anti-gp 120 antibodies in the 18 HIV+ sera tested varied from 0 to 1/4000 and showed only a moderate correlation with their CD4/gpl20 inhibition titers (Pearson correlation coefficient on log,,, transformed data 0.633; p =£ 0.005). Hence, antifor the current

of CD4 peptide binding (reciprocal liter)

Inhibition

Reciprocal of anti-gpl20 liter

0 0 0 0 40 0 0 50 0 60 10 100

100 0 0 0 30 0

84)3-HSA to gain access to the CD4-binding site on gpl20. In regard, it has been demonstrated that the percentage of anti-gp 120 antibodies (from a pool of asymptomatic HIV+ sera) directed against the CD4-binding site represents only 16% of the total amount of the total anti-gp 120 antibody. ' ' The binding of the CD4(41-84)3-HSA to gpl20 in the presence of these antibodies may be explained by the fact that the hydrophobic peptide protrudes from the surface of the hydrophilic HSA and hence, can gain access to the CD4-binding site on gpl20. In conclusion, we have shown that CD4(4 l-84)3-HSA blocks this

acute infection of H9 cells with HTLVIIIB and that it reacts with

in the presence of sera from 18 WRl->6 HIV + individuals. The antibodies which block gpl20 binding are not directed against the CD4-binding site on gpl20, but may bind to adjacent regions on the gpl20 molecule.

gpl20

even

=

gp 120 antibodies present in seventeen of the 18 HIV+ sera may have been responsible for the inhibition of the binding of CD4 to gpl20 as we have reported previously, but we cannot prove this definitively until larger amounts of sera are available.8 In any case, since only a portion of these antibodies may be directed against epitopes in or near the CD4-binding site, only a moderate degree of correlation would even be predicted. Furthermore, the only HIV+ serum without measurable anti-gp 120 antibodies failed to inhibit the binding of rCD4 to gpl20. In contrast to the situation with rCD4, there was absolutely no correlation between the anti-gp 120 titers of the sera and their modest inhibitory titers on the binding of CD4(41-84)3-HSA suggesting that the anti-gp 120 antibodies did not interfere with the binding of CD4(41-84)3-HSA to gpl20. The lack of correlation between these titers and the moderate correlation between anti-gp 120 titers and rCD4 inhibitory titers suggests that the anti-gp 120 antibodies which inhibit binding of CD4 to gpl20 are not directed against the CD4-binding site on gpl20, but rather to adjacent region(s) on gpl20 which, when masked with antigp 120 antibodies, block the ability of CD4, but not CD4(41-

ACKNOWLEDGMENTS We thank Mr. Y. Chinn for technical assistance, Ms. R. Reiber and Ms. S. Richardson for secretarial assistance and Dr. O. Ramilo for statistical analyses. We thank the AIDS Repository for providing all the test sera. This study was supported by NIH Grant AI-27336.

REFERENCES 1. Smith DH, Byrn RA, Marsters SA, Gregory T, Groopman JE, and Capon DJ: Blocking of HIV-1 infectivity by a soluble, secreted form oftheCD4 antigen. Science 1987;238:1704-1707. 2. Fisher RA, Bertonis JM, Meier W, Johnson VA, Costopoulos DS, LiuT, Tizard R, Walker BD, Hirsch MS, Schooley RT, and Flavell RA: HIV infection is blocked in vitro by recombinant soluble CD4. Nature 1988:331:76-78.

1948

McDougal JS, InackerR, Folena-Wasserman G, Arthos J, Rosenberg J, Maddon PJ, Axell R, and Sweet RW: A soluble form of CD4 (T4) protein inhibits AIDS virus infection. Nature 1988;331:82-84. 4. Hussey RE, Richardson NE, Kowalski M, Brown NR, Chang H-C, 3. Deen KC,

Siliciano RF. Dormían T, Walker B, Sodroski J, and Reinherz EL: A soluble CD4 protein selectively inhibits HIV replication and syncytium formation. Nature 1988;331:78-81. 5. Traunecker A, Luke W, and Karjalainen K: SolubleCD4 molecules neutralize human immunodeficiency virus type 1. Nature

1988;331:84-86. 6.

7.

8.

9. 10. 11.

Weber JN, Whitby D, Mclntosh K, Dalgleish AG, Maddon PJ, Deen KC, Sweet RW, and Weiss RA: Soluble CD4 blocks the infectivity of diverse strains of HIV and SI V for T cells and monocytes but not for brain and muscle cells. Nature 1989:337: 368-370. Skinner MA, Langlois AJ, McDanal CB, McDougal JS, Bolognesi DP, and Matthews TJ: Neutralizing antibodies to an immunodominant envelope sequence do not prevent gpl20 binding to CD4. J Virol 1988;62:4195-4200. Schnittman SM, Lane HC, Roth J, Burrows A, Folks TM, Kehrl JH, Koenig S, Berman P. and Fauci AS: Characterization of gpl20 binding to CD4 and an assay that measures ability of sera to inhibit this binding. J Immunol 1988;141:4181-4186. Callahan LN and Norcross MA: Inhibition of soluble CD4 therapy by antibodies to HIV. Lancet 1989;2:734-735. Callahan LN, BagnatoM, PhelanM. and Norcross M: New Orleans Meeting Abstract. FASEB J 1990;A:2015. Kang CY, Nara P. Chamat S, Caralli V, Ryskamp T, Haigwood N,

Clapham PR,

GHETIE ET AL. Newman R, and Köhler H: Evidence for non-V3-specific neutralizing antibodies that interfere with gpl20/CD4 binding in human immunodeficiency virus 1-infected humans. Proc Nati Acad Sei

(USA) 1991;88:6171-6175.

Slaughter C, Wheeler HT, Uhr JW, and VitettaES: CD4 peptide protein conjugates, but not recombinant human CD4, bind to recombinant gpl20 from the human immunodeficiency virus in the presence of serum from AIDS patients. Proc Nati Acad Sei

12. Ghetie V,

(USA) 1991;88:5690-5693. 13. HoDD, McKeatingJA, Li XL, MoudgilT, DaarES, SunN-C, and Robinson JE: Conformational epitope on gpl20 important in CD4 binding and human immunodeficiency virus type 1 neutralization identified by a human monoclonal antibody. J Virol 1991 ;65:489493. 14. Pitts TW. Bohanon MJ, Leach MF, McQuade TJ, Marschke CK, Merritt JA, Wierenga W, and Nicholas JA: Soluble CD4-PE40 is cytotoxic for a transfected mammalian cell line stably expressing the envelope protein of human immunodeficiency virus (HIV-1), and cytotoxicity is variably inhibited by the sera of HIV-1-infected patients. AIDS Res Human Retroviruses 1991 ;7(9):741-750.

Address

reprint requests

to:

Victor Ghetie

Department of Microbiology University of Texas Southwestern Medical Center Blvd. Dallas, TX 75235-9048

5323

Harry Hines