0161-5890/92 $5.00 + 0.00 0 1992 Pergamon Press Ltd

Molecular Immunology, Vol. 29, No. 6, pp. 729-738, 1992 Printed in Great Britain.

CLONAL HUMAN

ANALYSIS OF MURINE B CELL RESPONSE TO THE IMMUNODEFICIENCY VIRUS TYPE 1 (HIVl)-gag p’ 7 AND ~25 ANTIGENS*

VI?RONIQUE ROBERT-HEBMANN,? STI?PHANE EMILIANI,?FRI?DDBRIC JEAN,$ MARIANAREsNIcoFF,t FRANC~ISTRAINCAR@and CHRISTIANDEVAUXty tCRBM du CNRS, Centre de tri des moltcules anti-HIV, Institut de Biologie, Brd Henri IV, 34060 Montpellier, France; SImmunotech S. A. Campus Universitaire de Luminy, 13288 Marseille, France; $Hybridolab, 28 rue du Dr Roux, 75724 Paris, France (First received 7 October 1991; accepted in revised form 30 November

1991)

Abstract-The antigenicity of HIV-gag p17 and ~25 proteins was analyzed using a panel of 52 monoclonal antibodies (mAb) derived from 17 independent fusion experiment protocols performed in 12 different laboratories. These mAb were tested for their capacity to bind peptides corresponding to sequences of HIVI-BRU-gag p17 and ~25. Thirty-five overlapping peptides (Pl to P35) totally covering the p17 and ~25 proteins were used. This study allowed us to identify four immunodom-

inant regions inducing B cell response, two on p17 corresponding to P2 and P13 (amino acids 1l-25 and 121-132, respectively) and two on ~25 corresponding to P21 and P28-P29-P30 (a.a. 201-218 and 285-320 respectively). According to secondary structure predictions, peptides P2 and P21 contained hydrophilic alpha helix folded regions whereas P13 sequence presented a beta turn propensity. These regions and the P28-30 region were also predicted to be easily accessible to mAb. Several other p25-derived peptides: P15 (a.a. 142-156), P16 (a.a. 148-162), P19 (a.a. 176-192), P22 (a.a. 219-233) and P23 (a.a. 233-253) were recognized by mAb. No pl7-derived peptide other than P2, P13 and P12 (a.a. 111-123) was found to react with mAb. Cross-blocking studies between mAb, suggested the existence of more than four distinct epitopic areas on p17 and eight on ~25.

INTRODUCTION

The Human Immunodeficiency Virus (HIV) is the primary aetiological agent of Acquired Immunodeficiency Syndrome (AIDS) and associated diseases (BarrbSinoussi et al., 1983; Gallo et al., 1984). This retrovirus belongs to the subfamily of lentiviruses. HIV and related viruses share a similar genetic organization (Alizon et al., 1986; Daniel et al., 1985; Guyader et al., 1987; Ohta et al., 1988); the viral genes encode core structural proteins (gag), enzymes (pal), envelope proteins (env) and several regulatory proteins (Cullen, 1991). The gag gene codes for a polyprotein, which is the precursor of internal proteins of the virion. Post-translational cleavage of the gag precursor of 55 kDa occurs at specific sites as a result of the activity of the virus-encoded protease (Peng et al., 1989). This processing yields the N-myristoylated and internally phosphorylated pl7 matrix protein, the phosphorylated ~25 capsid protein and the nucleocapsid protein ~15 further cleaved into p9

*This work was supported by institutional grants from CNRS, INSERM and ANRS. V. R. and S. E. were fellows of the Ministere de la Recherche et de 1’EnseignementSupt?rieur (MRES). M. R. was supported by a postdoctoral training grant from the Agence Nationale de Recherches sur le SIDA (ANRS). l/Author to whom correspondence should be addressed: Dr Christian Devaux.

and p6. Assembly of HIV1 particles and their maturation into infectious virions remain poorly understood in terms of mechanistic sequence of events, mutual interactions between partner proteins or between core proteins and genomic RNA and topology of virion structural elements (Gelderblom, 1991). During the past few years several laboratories including our own (Ferns et al., 1987; 1989; Janvier et al., 1990; Niedrig et al., 1988, 1991; Robert et al., 1991; Tatsumi et al., 1990a; Tersmette et al., 1989) have attempted to generate mAb against the HIV core proteins. Such mAb might be valuable tools to analyze the structure of capsid proteins and define their role in the assembly and maturation of viral particles into infectious viruses. Furthermore, these reagents might be used in order to serologically determine the extent of variability among HIV of unknown gag sequences. We here analyze the antigenicity of HIV1 -gag pl7 and ~25 by the combined use of 52 mAb (16 antiHIV-gag P17 and 36 anti-HIV-gag ~25 mAb) and 35 overlapping synthetic linear peptides totally covering the amino acid sequences of these two gag proteins. We determined the location of the HIV-gag antigenic regions, their spatial relationship and use probabilities of structure occurrence to predict the secondary structure of the p17 and ~25 molecules. The structural models of p17 (Andreassen et al., 1990) and ~25 (Argos, 1989; Langedijk et al., 1990) are revisited in light of these data.

729

V. ROBERT-HEBMANN et al.

730 MATERIALS

AND METHODS

Virus The virus HIVl-BRU kindly provided by L. Montagnier (Institut Pasteur, Paris, France) (Barre-Sinoussi et al., 1983) was propagated in CEM cells cultured in RPM1 1640 supplemented with 10% FCS and antibiotics, then concentrated from the cell free supernatant by 100,000 g ultracentrifugation for 2 hr. The viral pellet was resuspended hundredfold concentrated in NTE buffer (10 mM Tris, 100 mM NaCl, 1 mM EDTA pH 7.4). Viral particles (1 mg/ml) were inactivated by 56°C incubation for 30 min. and treated with 0.1% Triton x 100. Monoclonal

antibodies

The anti-HIVl-NDK (RL4.11.14, RL4.72.1, RL4.138.1, RL16.30.1, RL16.24.5, RL16.45.1) and antiHIVZROD (M01.34.1, M09.42.2, M09.50.2) mAb producing hybridomas have been obtained by members of this laboratory. The C.V.K. mAb has been obtained by J. Chassagne and provided by F. Barr6Sinoussi (Institut Pasteur, Paris, France). mAb K7.17, L14.17, K3.24, K5.24 and L6.24 have been provided by M. Tatsumi (NIH, Tokyo, Japan). mAb M26 and M33 have

been provided by F. Di Marzo Veronese under transfer agreement with NIH/ADAMHA (NIHIADAMHA, Bethesda, MD). mAb 8D2, 8H7, 3H7 and 12B4 have been provided by H. Gelderblom (Robert Koch Institut, Berlin, Germany). mAb 32/5.8.42, 32j1.24.89 and 32/5.17.76 have been provided by L. Papsidero (Cellular Products Inc., Buffalo, NY), mAb CLB-14, CLB-16, CLB-21 and CLB-47 have been provided by J. Huisman (Netherlands Red Cross Blood Transfusion Service, Amsterdam, The Netherlands). mAb lE8G2, lD8F6, 3D3, 1 lH9, 4H2Bl and lD9 originally described by R. Ferns have been provided by the MRC AIDS directed programme (MRC, London, England, U.K.). mAb 14D4E11, 3DlOG9, 15F8C7, 1 ID1 lF2, IG5C8,9A4C4, 1 lClOB10,23A5G7 and 9B5C12 have been provided by B. Mandrand (UM 103 CNRS-Biomerieux, France) under transfer agreement with Agence Nationale de Recherches sur le SIDA (ANRS, Paris, France). mAb 311/01, 406/01, 714/01, 1109jOl and 211/84 have been provided by J. Larroque (Biosoft, Clonatec, Paris, France). mAb 31-11, 15-21, 47-2 and 169-2 have been obtained by F. Traincard (Hybridolab, Paris, France). The main characteristics of these mAb are summarized in Table 1.

Table 1. mAb used in this study Hybridoma designation

isotype”

RL16.24.5 RL16.45.1 K7.17

IgGl IgGl IgG2a

X63.Ag8.653 X63.Ag8.653 SP2/O.Ag14

HIVl-NDK HIVl-NDK HIVl-KB

Pl7lP55 Pl7lP55 Pl7lP55

L14.17 C.V.K. 32/5X42 32/l .24.89 M33 llH9

IgGl IgG2a IgG2a IgG2b IgG2a IgGl

SP2/O.Ag14 X63.Ag8.653 SP2/O.Agl4 SP2/O.Ag14 NSl-Ag4.1 X63.Ag8.653

HIVl-BRU(LAV) HIVl-BRU(LAV) HIV1 HIV1 HIVl-(HTLVIIIB) HIVl-CBLl

Pl7lP55 Pl7lP55 Pl7lP55 Pl7lP55 Pl7lP55 P17/p55

4H2B 1

IgGl

X63.Ag8.653

HIVl-CBLl

Pl7/P55

lD9

IgG2a

X63.Ag8.653

HIVl-CBLl

P17

3H7 31 l/O1 31-11 15-21 M01.34.1 RL16.30.1 RL4.11.14

IgG3 IgGl IgGl IgGl IgG2a IgG2a IgGl

NSl-Ag4.1 SP2/O.Ag14 X63.Ag8.653 X63.Ag8.653 X63.Ag8.653 X63.Ag8.653 X63.Ag8.653

HIV1 -(HTLVIIIB) HIVl-(HTLVIIIB) HIVl-BRU(LAV) HIVl-BRU(LAV) HIVZ-ROD HIV1 -NDK HIVl-NDK

Pl7/P55 P17 P17 Pl7/P55 ~161~55 ~251~55 ~251~55

RL4.72.1

IgGl

X63.Ag8.653

HIV1 -NDK

~251~55

RL4.138.1 K3.24 K5.24 L6.24 CLB-14 CLB-16 CLB-2 1

IgGl IgG2a IgG2b IgGl IgG2b

X63.Ag8.653 SP2/O.Ag14 SP2/O.Ag14 SP2/O.Ag14 SP2/O.Ag14 SP2/O.Ag14 SP2/O.Ag14

HIVl-NDK HIVl-KB HIV1 -KB HIVl-BRU(LAV) HIV1 HIV1 HIV1

k

IgG2b IgG2b

Parental myelomab

Western blot Immunogen’

reactivity

Reference Robert et al., 1991 Robert et al., 1991 Tatsumi et al., 1990a; Tsunetsugu-Yokota et al., 1991 Tatsumi et al., 1990~ Chassagne et al., 1986 Papsidero et al., 1989 Papsidero et al., 1989 Di Marzo Veronese et al., 1988 Ferns et al., 1987; 1989; Spence et al., 1989b Ferns et al., 1987; 1989; Spence et al., 19896 Ferns et al., 1987; 1989; Spence et al., 1989b Niedrig et al., 1988; 1989 Larroque et al., unpublished Traincard et al., unpublished Traincard et al., unpublished Devaux et al., unpublished Robert et al., 1991 Tatsumi et al., 199Oa; Tsunetsugu-Yokota et al., 1991 Tatsumi et al., 1990~; Tsunetsugu-Yokota et al., 1991 Tsunetsugu-Yokota et al., 1991 Tatsumi et al., 1990~ Tatsumi et al., 1990~ Tatsumi et al., 1990~ Tersmette et al., 1989 Tersmette et al., 1989 Tersmette et al., 1989 Langedijk et al., 1990

Epitope mapping on HIVI-gag proteins

731

Table I-continued Hybridoma designation

Ig isotype’

CLB-47 3215.17.76 M26 1E8G2

IgG2b IgG2a IgGl IgGl

lD8F6

Parental myefoma*

Immunogenc

Western blot reactivity

SP2/O.Ag14 SP2/O.Agl4 NSl-Ag4.1 X63.Ag8.653

HIV1 HIV1 H~Vl-(HTLVIIIB) HIVl-CBLI

P251P55 P25 P25lP55 P25lP55

IgGi

X63.Ag8.653

HIVl-CBLl

P251P55

3D3

IgG2b

X63.Ag8.653

HIVl-CBLl

~251~55

8D2 8H7 12B4 21l/84 406/O1 714/01 1109/01 14D4E11 3DlOG9 15F8C7 llDllF2 lG5C8 9A4C4 1lClOBl0 23A5G4 9B5Cl2 47-2 169-2 M09.42.2 M09.50.2

IgG2a IgG3 IgGl IgGl IgG2b IgGl IgGl IgGl IgGl IgGl IgGl IgG2b IgGl IgGl IgGl IgG2b IgGl IgGl IgGl IgGI

NS 1-Ag4.1 NSl-Ag4.1 NSl-Ag4.1 SP2/O.Agl4 SP2/O.Agl4 SP2/O.Agl4 SP2/O.Agl4 SP2/O.Ag14 SP2/O.Agl4 SP2/0.Ag14 SP2/O.Ag14 SP2/O.Ag14 SP2/O.Ag14 SP2jO.Ag14 SP2/O.Ag14 SP2/O.Agl4 X63.Ag8.653 X63.Ag8.653 X63.Ag8.653 X63.Ag8.653

HIV 1-(HTLVIIIB) HIVl-(HTLVIIIB) HIVl-(HTLVIIIB) HIVl-(HTLVIIIB) HIVl-(HTLVIIIB) HIVl-(HTLVIIIB) HIVl-(HTLVIIIB) ~25 HIVI-(HTLVIIIB) ~25 HIVI-(HTLVIIIB) ~25 HIVl-(HTLVIIIB) ~25 HIVI-(HTLVIIIB) ~25 HIVI-(HTLVIIIB) ~25 HIVI-(HTLVIIIB) ~25 HIVI-(HTLVIIIB) ~25 HIVl-(HTLVI~IB) HrVl-(HTLVIIIB)HIV2 HIVl-BRU(LAV) HIV2-ROD HIV2-ROD HIVZ-ROD

~251~55 P25/P55 ~251~55 P25lP55 P25/P55 P25/P55 P25lP55 P25lP55 ~251~55 ~251~55 P25/P55 P25lP55 P25/P55 ~251~55 P25 P25lP55 P25iP55 ~261~55 ~261~55 ~261~55

Reference Tersmette et al., 1989 Papsidero et al., 1989 Di Marzo Veronese et al., 1988 Ferns et al., 1987; 1989; Spence et al., 1989a Ferns et al., 1987; 1989; Spence et al., 1989a Ferns et al., 1987, 1989; Spence et al., 1989a Niedrig et al., 1988, 1989 Niedrig et al., 1988, 1989 Niedrig et al., 1988, 1989 Larroque et al., unpublished Larroque et al., unpublished Larroque et al., unpublished Larroque et al., unpublished Janvier et al., 1990 Janvier et al., 1990 Janvier et al., 1990 Janvier et al., 1990 Janvier et al., 1990 Janvier et al., 1990 Janvier et al., 1990 Janvier et al., 1990 Mandrand et al., unpublished Traincard et al., unpublished Traincard et al., unpublished Devaux et al., ~published Devaux et af., unpublish~

aImmunoglobuiin isotypes were determined by Ouchterlony immun~iffusion technique and/or ELISA. *Hybridomas were constructed by fusion between mouse myeloma NSl-Ag4.1, SP2/O.Ag14 or X63.Ag8.653 and spleen cells from BALB/c mice immune to HIV antigens. ‘Mice were immunized with either inactivated HIV1 or HIV2 viruses or purified p25 protein from HIV1 strain HTLVIIIB (~25 HIVl-(HTLVIIIB)).

Recombinant proteins and oligopeptides

The HIVl-BRU-gag p17 and p25 recombinant proteins have been provided by H. Kolbe (Transgkne, Strasbourg, France). Polyhedrin fusion-HIV-gag (~17, ~25 and most of p9) protein recovered from baculovirus infected cells was provided by P. Boulanger (Institut de biologie, Montpellier, France). The 35 HIVl-gag linear peptides used (purity over 85%) have been synthesized at Neosystem (Neosystem, Strasbourg, France) according to the sequence published for the HIVl-BRU isolate and provided by the ANRS.

times with PBS-0.05% Tween 20. Bound immunoglobulins (Ig) were detected by adding 100 ,ul of goat-antimouse IgG H + L peroxidase conjugate (lo3 fold diluted) (Immunotech S. A., Marseille, France) for 1 hr, followed by washing and subsequent incubation with 0-phenylenediamine (OPD) (Sigma, Main, MO) as substrate. A reaction with a peptide in a series was considered positive in direct ELISA when absorbance was over three times the background (background: mean absorbance of the three peptides in a series giving the lowest absorbance).

Enzyme -linked imm~~osorben t assay (ELBA

Peptide inhibition of mAb binding assay

)

ELISA plates (Nunc, Paisley, Scotland, U.K.) were coated overnight with 10 pg/ml of r~ombinant protein or peptide in 100 mM sodium carbonate buffer pH 9.6. After coating, plates were washed and saturated with phosphate buffered saline (PBS) containing 1% bovine serum albumin (BSA). One hundred microliters of purified mAb (20 pg/ml), ascitic fluid (200 fold diluted), or hybridoma culture supernatant (neat), were incubated overnight at 4”C, then the plates were washed three

ELISA plates were coated with HIVl-BRU and BSA saturated as previously described (Tatsumi et al., 1990b). Mixture of mAb (lowest concentration of mAb giving an OD 492 nm of 2.0 when enzymatically detected in direct binding assay) and peptide (concentration ranging from lo-‘OM to 10e4M) in PBS-0.2% BSA solution were added to the HIV coated wells. After 1 hr incubation at room temp, plates were washed three times and bound Ig were detected as described above using either

732

V. ROBERT-HEBMANN

streptavidin peroxidase/OPD substrate or goat-antimouse IgG H + L peroxidase conjugate/OPD substrate, depending whether mAb were biotinylated or not.

et al.

algorithms, BISANCE

respectively. Programs (CIT12, Paris, France).

Cross-competition assay was performed as previously described (Devaux et al., 1985; Pierres et al., 1981). Briefly, mixture of appropriately adjusted biotinylated mAb (lowest concentration of biotinylated mAb needed to reach an absorbance of 2.0 OD 492 nm after incubation with avidin-enzyme-substrate in direct assay) and various amounts of competitor mAb were incubated for 1 hr in wells of HIVl-BRU or polyhedrin fusion-HIVgag protein coated plates. The plates were washed three times and incubated with streptavidin peroxidase (Immunotech S. A., Marseille, France) and substrate as described above. Identical data were obtained when plates were coated with either HIVl-BRU or polyhedrin fusion-HIV-gag protein. Prediction of p 17 and ~25 protein secondary structure HIVl-BRU-gag p17 and ~25 structure predictions were done according to the probabilities of structure occurrence frequency after Chou and Fasman (1978) to predict potential alpha helices, beta pleated sheets and beta turn secondary structures. Flexibility, accessibility and hydrophilicity were predicted after Karplus and Schulz (1985) Janin (1979) and Hopp and Woods (1981)

P3 IRPGQzKYKIxH1v

GEm~mLRPGG

run

Ma@rg Of mAb binding site Fifty-two anti-HIV-gag mAb, whose characterization is summarized in Table 1, were tested for their reactivity with 35 overlapping synthetic linear peptides (PI to P35) made according to the sequence published for HIVlBRU and covering the entire p17 and ~25 proteins (Fig. 1). A representative binding experiment is shown in Fig. 2. The anti-p25 mAb RL4.72.1 strongly reacted with the peptide P22 that mimicks an amino acid sequence of the ~25 protein and did not show reactivity with any of the other 21 ~25 derived peptides tested nor the thirteen p17 derived peptides (Fig. 2A and data not shown). The specificity of this reaction was confirmed by a competitive inhibition of RL4.72.1 binding to HIVl-BRU by peptides. As shown in Fig. 2B, the binding of biotinylated RL4.72.1 mAb to HIVIBRU was totally inhibited by concn of 10 .hM ~25 recombinant proteins and 10-5M P22 peptide. No inhibition was observed using a concentration of 1O- 4M P2 1 peptide. Table 2 summarizes the binding results of all tested mAb to HIVl-BRU inactivated particles, recombinant p17 and ~25 proteins and the 35 peptides (Pl to P35).

P7 GQLQPSLOTGSEEIRSL

IBRFAV&LLETSE

KHIVWAkLERFAV

LETSEGC&.GOLO

PlO EIKD'XEA

EEIRSLkVATLY LYNTVAp~YCVH(RI

P12 sxI;pAOCMAADTG Pll Pi3 DTt3HSSOVSQNY IEEEORKS?XKA

PlO LLXIEEE

Pi4 P17 Pi9 PIVaNIaGu4VHoAIS AWVKVVEEKAFSPEVIP~ SEGATPQDLNTMLMVG Pi5 P18 P20 HVHOAISPRTUJAWV IP?@SALSEGATPQDL GHaAArml Pi6 SPRTINAWVKWEEK

220 240 201 IKETINEEAAEWlRVHPVHAGPIAPGOaEPRGSDIAG~STLEQI~

HAGPIAif&7EPRG P21 IKETINEEAABWDRVHPV P20 IKmINEE

P28 YKTL. P29 YKTLRAEOAS P30 KTIRASQASOE'T

through

RESULTS

Cross-inhibition of mAb binding assay

hS3~ASkGGELDR

were

260

P24 BOI~NPPIPVGEI

GSDIAG&~EOIGMTNN

260

300

PIPVGEIYP;RWIILGLH;IVRMYSPTSIIDIROGPKEPFRDYVDRF P26 SRwIIyGL~Iw

IPVGEIk%IILGL

P28 IROGPKEPFRDYYDRF P29 FRDYVDRF

P27 LM;IVF+lYSPTSILDIRO

P31 P34 ETLLVONANPD(XTILKAL ~ACWVGGF'GliKA P32 P35 DCXTILKALGPAATLS GHKARVLAEAMSOVTN P33 KALGPMTLEEtMTAW

Fig. 1. Amino acid (a.a.) sequence of the 35 peptides used in this study and their location relatively to the HIVl-BRU-gag p17 and p25 proteins sequence. The a.a. numbering was made according to the HIV Sequence Database Group T-10 (Los Alamos National Laboratory, NM). A permutation L-V (underlined a.a.) relatively to the HIVl-BRU corresponding sequence was introduced in P26 sequence to improve the solubility of this peptide.

Epitope mapping on HIVl-gag proteins

733

mAb reacted with the P30 peptide. The last four mAb recognized the P15-P16, P19, P22 and P23 peptides respectively. Study of HIV 1-BRU-gag p 17 and p 25 protein secondary structure

CONCENTRATION OF COMPETITOR PEPTIDES AND RECOMBINANT PROTEIN (H)

Fig. 2. Representative experiment of anti-HIV-gag mAb epitope mapping. (A) Reactivity of RL4.72.1 mAb with the different p25 peptides using a direct ELISA (see M.M. for details). The results are expressed as optical density versus peptide number. (B) Competitive inhibition of binding of anti-HIVl-gag ~25 RL4.72.1 mAb by soluble synthetic peptides and p25 recombinant protein. Symbols: 0, ~25 recombinant protein; 0, P22 peptide; n , P21 peptide. The binding site of a given mAb was considered to be mapped only when a reactivity for a single or two overlapping peptides was observed in direct ELISA and absorbance was over 3 times the background (background: mean absorbance of the three peptides in a series giving the lowest absorbance with the same mAb in direct ELISA) and when this peptide inhibited the binding between mAb and HIVl-BRU coated plates. In all other cases it was considered not determined (ND). With the exception of RL16.24.5 (an anti-HIVl-NDK mAb) and MO 1.34.1 (an anti-HIV2-ROD mAb) used as controls, all mAb (50/52) bind HIVl-BRU viral particles. Thirteen mAb reacted with the p17 and 36mAb with the ~25 recombinant protein. Despite its specificity for the p17 protein of HIVI-CBLI and its capacity to bind HIVl-BRU viral particles, mAb lD9 did not significantly react with the p17 recombinant protein used in this study. The mAb were then tested for reactivity with peptides. According to the criteria retained for considering that a given mAb reacts with a given peptide, the binding sites of 29 mAb were considered to be mapped. The large majority (S/10) of the anti-p17 mAb that reacted with a peptide demonstrated a specificity either for the P2 (3/10) or the PI3 (6/10) peptide. One mAb (3H7) however, reacted with P12. Several anti-p25 mAb (7/19) derived from independent fusion experiments reacted with the P21 peptide. Another group of anti-p25 mAb reacted with P28 or P28 and P29 peptides. Three MlMM 29,6c

The p17 secondary structure predictions (Fig. 3; a.a. 1-132) suggest that peptide P2 (a.a. I l-25) starts with an hydrophilic and accessible alpha helix region, flanked in its C-terminal portion by a sequence with high beta sheet propensity followed by a beta turn. The P13 peptide (a.a. 121-132) corresponds to a sequence of p17 with a high beta turn propensity and was predicted to be flexible and accessible. A region encompassing the PIO, Pll and a part of P12 peptides was also predicted to be hydrophilic, accessible and mainly composed of alpha helix folded structures. The ~25 secondary structure predictions suggest the existence of several hydrophilic and accessible regions including the one corresponding to the P21 peptide organized in alpha helix. Our prediction fits correctly with the ~25 structural model proposed by Langedijk et al. (1990). A better fit was, however, obtained for the beta pleated sheets when predictions were performed after Deleage and Roux algorithms (data not shown). It is noteworthy that the region containing the P26 peptide corresponding to the puff in the Argos’ model (Argos, 1989) turned out to be composed of hydrophobic sequences and demonstrated a high beta pleated sheet propensity. This last result totally fits the ~25 modified model proposed by Langedijk et al. (1990). Topology of the epitopes analyzed by cross-blocking studies

Several anti-HIV mAb were biotinylated and their binding to HIVl-BRU coated ELISA plates tested in the presence of various amounts of competitor mAb. Using a similar approach, we previously reported the existence of at least six spatially distinct epitopic regions on HIVl-NDK ~25 protein (Tatsumi et al., 1990a). Here the results illustrated in Fig. 4 show bidirectional crossinhibition of binding among several groups of anti-p25 mAb thereby defining clusters of epitopes either identical or distinct, but very close to each other. For example, 7/7 mAb reacting with the P21 peptide inhibited the binding of biotinylated 14D4Ell and 47-2 mAb to HIVl-BRU. Yet, our results support the possibility that further heterogeneity exists within this group of epitopes since their capacity to inhibit the binding of mAb from other groups varies significantly. Altogether our results support the existence of 8 distinct clusters of epitopes on ~25 and suggest that the number of distinct epitopic regions of this molecule is even higher. Five mAb not precisely mapped using the peptide’ approach, exhibited a pattern of reactivity in crossblocking studies compatible with binding to epitopes located in the vicinity of the P22 (M26), P23 (CBL-14 and L6.24), P28 (169-2) and P30 (K3.24) corresponding

134

V. ROBERT-HEBMANNet al. Table 2. Specificity

of anti-HIV

mAb

Binding to recombinant mAb designation _____ RL16.24.5 RL16.45.1 K7.17 L14.17 C.V.K. 3215X.42 32/1.24.89 M33 llH9 4H2Bl lD9 3H7 31 l/O1 31-11 15-21 M01.34.1 RL16.30.1 RL4.11.14 RL4.72.1 RL4.138.1 K3.24 K5.24 L6.24 CLB-14 CLB-16 CLB-2 1 CLB-47 3215.17.76 M26 lE8G2 lD8F6 3D3 8D2 8H7 12B4 21 l/84 406/O 1 714/01 1109/01 14D4Ell 3DlOG9 15F8C7 llDllF2 lG5C8 9A4C4 1 lClOBl0 23A5G4 9B5C12 47-2 169-2 M09.42.2 M09.50.2

-

Binding to HIV1 -BRU”

P17

_d

-

-

+ + + + + + + + + f + + + + _d

+ + + + + + + + + -

NT _ _ -

+ + + + + + + + + + + + + + + + + + -t + + + + + -I+ + + + + + + + + + +

+ + + + _ _ _ NT NT _ NT NT _ -

P25

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +

Known assignmentb ____~ NT” NT ND NT NT 12-19/10&105 17-22 NT 101-115 119-132 119-132 113-122 NT 124-132 124-132 NT NT ND 150-240 15&240 NT NT NT 133-163/24&251 133-163/21 l-250 312-356 or 164-172 133-163/21 l-250 NT NT 250-310 143-157 177-182 203-217 203-217 293-302 NT NT NT NT 209-217 260-267 179-188 26&267 209-217 26(t267 260-267 209-217 NT 209-2 18 ND NT NT

Reactive peptide’ ND’ ND ND P2( 1 l-25) ND P2( 1 l-25) P2( 1 l-25) P13(121-132) ND P13(121-132) P13(121-132) P12(111-123) P13(121-132) P13(121-132) P13(121-132) ND ND ND P22(219-233) ND ND ND ND ND ND ND ND ND ND ND P15-P16 (142-162) P19 (176-192) P21(201-218) P21(201-218) P28(285-304) ND P23(233-253) P21(201-218) P21(201-218) P21(201-218) P28(285-304) ND P30(302-320) P21(201-218) P30(302-320) P30(302-320) P28(285-304) ND P21(201-218) ND P28-P29(285-310) P28-P29(285-310)

“mAb were tested for binding to HIVl-BRU, recombinant p17 and p25 proteins using the ELISA technique described in Materials and Methods. %ee Table 1 for published assignment references. ‘See legend of Fig. 2 for details, amino acids number are shown between bracket. dmAb RL16.24.5 and M01.34.1 that failed to react with HIVl-BRU were shown to react with HIVl-NDK and HIV2-ROD, respectively. ‘NT: Not tested; ND: Not determined.

Epitope mapping on NIVi-gag

proteins

735

6 4 2 0 -2 -‘4

Fig. 3. Prediction of ~IVI-BRU-gig p17 and p25 protein secondary structure and antigeni~ty. Abscissa axis represents amino acid sequence. A vertical bar is positioned in between the amino acids 132 of p17 and 133 of ~25. (A) (B) and (C) ~17 (a.a. l-l 32) and ~25 (a.a. 133-363) protein structure predictions after Chou and Fasman algorithm; (A) alpha helix prediction (cut off value = 1.0); (B) beta sheet prediction (cut off value = 1.0); (C) turn prediction (cut off value = 0.5). (D) (E) and (F) ~17 and p25 protein antigenicity predictions; (D) flexibility (cut off value = 1.O); (E) accessibility (cut off value = 4.0); (F) hydrophilicity (cut off value = 0).

sequences on ~25. It is noteworthy that published assignment of CLB-I4 is amino acids 133-163/240-251 and part of this site corresponds to a sequence that maps

within the P23 peptide (a.a. 233-253). Unidirectional inhibition of binding observed using P21-reacting mAb as competitor suggests a complex spatial organization of the various clusters of epitopes. These mAb inhibited the binding of the anti-P28reacting mAb to ~25 but not that of mAb reacting with P22 or P30 peptides. Since P21 and P28-like sequences are on opposite sides of the ~25 molecule according to the Langedijk’s model (Langedijk eb al., 1990), it suggests that binding of an anti-P21 reacting mAb to ~25 induces a conformational change in the molecule. Cross-blocking experiments performed with anti-p17 mAb (data not shown) revealed that all mAb reacting with P2 bind epitopes in the same epitopic region, whereas mAb reacting with PI3 define two distinct epitopic regions. Another independent epitopic region is defined by mAb 3H7 and localized in P12 peptide. Yet, several mAb did not bind epitopes located within one of these four epitopic regions. DISCUSSION The main purpose of this study was the identification of B-cell epitopes of HIVl-gag. The availability of series of overlapping linear synthetic peptides to p17 and ~25 sequences of HIVl-BRU through the ANRS and mAb provided by several laboratories permitted us to further

define the map of antigenic gag regions. Among a panel of 52 mAb specific for gag antigens, 50 recognized the p17 or ~25 proteins of HIVl-BRU. The last two mAb are an HIVI-NDK strain-specific mAb (Robert et al., 1991) and an anti-HIVZ-ROD mAb (C.D., unpublished data) used as controls. The advantage of this study over others performed with mAb, resides in the number of mAb tested, the genetic diversity among HIV isolates used as immunogen and the fact that these mAb were obtained in several independent immunization and fusion experiments performed in different laboratories. Although several regions of ~17 were predicted to be good candidates for inducing B-cell response (Andreassen et al., 1990; this work) only two of them reacted with mAb, the 1l-25 (P2) and Ill-132 (PI2 and P13) amino acid regions. This might be a consequence of the limited number of anti-p17 mAb reacting with HIVIBRU tested in this study (14mAb) and the failure to reveal some binding sites (4/14) by the peptide approach. Yet, only these two regions were recognized by immunoglobulins of mice immunized with HIV (F. T., unpublished data). Therefore, the restricted response observed might also be a consequence of a mechanism of immunolo~cal tolerance to self antigens presenting structural homologies with HIV-gag p17 protein (Nossal, 1989). Interestingly, sequence homology between HIVl-gag p17 and self antigen have already been described in the region 92-109 (PlO-PI 1) (Naylor et al., 1987; Sarin et al., 1986). The fact that this region, predicted to be hydrophilic, accessible and alpha helix

V, ROBERT-HEBMANN et ul.

RI.4

178 I

I

I

I

3UI. I - 76 ,2liW

I

1

I

I

I

1

I

I

I

I Nl

I

I

!

h-l

I

I

/MI

I I

*

1%

I

I

h-3

! h7 /

Fig. 4. Cross-inhibition of binding of anti-p25 mAb. The cross-inhibition between two mAb for binding to HIVI-BRU was evaluated as described in the experimental section using the ELISA system. The data are represented as a diagram where absence of competition is shown as an empty square and partial and complete competition are shown as hatched and closed squares, respectively. Each Figure has been deducted from repeated titrations of a given inhibitor mAb. Some mAb were

not available in large enough amounts to be biotinylated or used as competitor and therefore do not appear in the Figure. NT: Not tested.

folded was not recognized by any mAb might support this hypothesis. When antigenic sites defined using anti-p25 mAb were superimposed on the structural model proposed by Argos (1989) and predicted by homology between ~25 and the coat protein VP2 of picornaviruses, the major B cell-inducing-response sequence, corresponding to P21 peptide, maps into an alpha helix folded region. This prediction is in agreement with our own. Another region we predicted to contain B cell epitopes that included P28, P29 and P30 peptides are also recognized by several anti-HIVl-gag p25-reacting mAb obtained after HIV1 or HIV2 immunization. This region was also considered by Argos to be accessible. The majority (44152) of mAb tested in this study were obtained by immunizing mice with inactivated viral particles. Among the 8 hybridoma cell lines derived from mice immunized with recombinant proteins, the majority produced mAb that bind the P21, P28 or P30 peptides and therefore did not differ in their specificity from those obtained by hybridizing cells from mice immunized with inactivated virus. These results are in agreement with those of other authors (FCigerstam et al., 1990; Hinkula et al., 1990; Niedrig et al., 1991) who have analyzed the

map of antigenic gag regions with mAb produced by hybridomas derived from mice immunized with recombinant p25 protein. A major difference between our results and the proposed Argos’ model is related to the puff region. The predicted puff was expected to be well exposed and highly antigenic. However, the P26 peptide which corresponds to this region was not recognized by any mAb. This result is in agreement with our predictions suggesting that this sequence is hydrophobic, unflexible and only composed of beta strands. This result correlates to the data reported by Langedijk et al. (1990) and their proposed modified model for ~25. Finally, our data demonstrate a complex spatial organization of the various clusters of epitopes on ~25 and suggest that binding of mAb to some epitopes located in the P21 region (a.a. 201--218) induces confo~ational changes in the molecule that affect the accessibility to the P28 region (a.a. 285-304). Availability of a large number of mAb with a precisely determined specificity together with a large number of HIV isolates from different origins will allow a better evaluation of the antigenic polymorphism among this group of viruses. This work is currently under progress.

Epitope mapping on HIVl-gag proteins AcknowZedgement~-We wish to thank the ANRS, MRC, NIH, lnstitut Pasteur and Netherlands Red Cross, the private companies Biosoft, Biomerieux, Cellular products, Hybridolab, Immunotech and Transgene and all our colleagues F. Barry-Sinoussi, F. Barin, P. Boulanger, J. Chassagne, F. Di Marzo Veronese, R. Ferns, H. ~lderblom, J. Huisman, H. Kolbe, J. Larroque, B. Mandrand, L. Montagnier, L. Papsidero, M. Tatsumi and R. Weiss who have kindly offered their reagents to make this work possible. We also thank M. Nicolas for excellent secretarial assistance.

737

in monitoring HIV isolate variation. J. Gen. Yirol: 68, 1543-1551. Ferns R. B., Partridge 3. C., Spence R. P., Hunt N. and Tedder R. S. (1989) Epitope location of 13 anti-gag HIV-l monoclonal antibodies using oligo~ptides and their cross reactivity with HIV-2. AIDS 3, 829-834. Gal10 R. C., S~ahuddin S. Z., Popovic M., Shearer G. M., Kaplan M., Haynes B. F., Palker T. J., Redfield R., Oleske J., Safai B., White G., Foster P. and Markham P. D. (1984) Frequent detection and isolation of cytopathic retroviruses (HTLV-III) from patients with AIDS and at risk for AIDS. Science 224, 500-503. REFERENCE Gelderblom R. (1991) Assembly and mo~hoIogy of HIV: potential effect of structure on viral function. AIDS 5, Alizon M., Wain-Hobson S., Montagnier L. and Sonigo P. 617-638. (1986) Genetic va~ability of the AIDS virus: nucleotide sequence analysis of two isolates from African patients. Cell Guyader M., Emerman M., Sonigo P., Clavel F., Montagnier L. and Alizon M. (1987) Genome organization and trans46, 63-74. activation of the human immunodeficiency virus type 2. Andreassen H., Bohr H., Bohr J., Bnmak S., Bugge T., Nature 326, 662-669. Cotterill R. M. J., Jacobsen C., Kusk P., Lautrup B., Hinkula J., Rosen J., Sundqvist V. A., Stigbrand T. and Petersen S. B., Saermark T. and Uhich K. (1990) Analysis Wahren B. (1990) Epitope mapping of the HIV-l gag region of the secondary structure of the human immunodeficiency with monoclonal anti~dies. Molec. ~rnrn~~. 27, 395-403. virus (HIV) proteins ~17, gpl20, and gp41 by computer modeling based on neural network methods. J. AIDS 3, Hopp T. P. and Woods K. R. (1981) Prediction of protein antigenic determinants from amino acid sequences. 615-622. P.N.A.S., U.S.A. 78, 3824-3828. Argos P. (1989) A possible homology between immunoJanin J. (1979) Surface and inside volumes in globular proteins. deficiency virus p24 core protein and picornaviral VP2 coat Nature 227, 491-492. protein: prediction of HIV p24 antigenic sites. Embo J. 8, Janvier B., Archinard P., Mandrand B., Goudeau A. and Barin 779-785. F. (1990) Linear B-cell epitopes of the major core protein of Barre-Sinoussi F., Chermann J. C., Rey F., Nugeyre M. ‘I’., human immunodeficiency virus types 1 and 2. J. Viral. 64, Chamaret S., Gruest J., Dauguet C., Axler-Blin C., Vezinet42.584263. Brnn F., Rouzioux C., Rozenbaum W. and Montagnier L, (1983) Isolation of a T-l~pho~opic retrovirus from a Karplus P. A. and Schulz G. E. (1985) Prediction of chain flexibility in proteins. ~at~rwisse~sc~a~t 72, 212-213. patient at risk for acquired immune deficiency syndrome Langedijk J. P. M., Schalken 3. J., Tersmette M., Huisman (AIDS), Science 220, 868871. 3. G. and Meloen R. H. (1990) Location of epitopes on the Chassagne J., Vexrelle P., Dionet C., Clavel F., Barre-Sinoussi major core protein p24 of human immunodeflciency virus. F., Chermann 3. C., Montagnier L., Gluckman J. C. J. Gen. Viral. 71, 2609-2614. and Klatzmann D. (1986) A monoclonal antibody against Naylor P. H., Naylor C. W., Badamchian M., Wada S., LAV-gag precursor: use for viral protein analysis and Goldstein A. L., Wang S. S., Sun D. K., Thornton A. H. and antigenic expression in infected cells. J. Immu~. 136, Sarin P. S. (1987) Human immunode~ciency virus contains 1442-144s. an epitope immuno~active with thymosin aI and the 30Chou P. Y. and Fasman G. D. (1978) Empirical predictions of amino acid synthetic ~17 group-specific antigen peptide protein conformation, Ann Rw. Biochem. 47, 251-276. HGP30. P.N.A.S., U.S.A. 84, 2951-2955. Cullen B. R. (1991) Human immunodeficiency virus as a Niedrig M., Rabanus J. P., L’Age Stehr J., Gelderblom H. R. prototypic complex retrovirus. J. l&of. 65, 1053-1056. and Pauli G. (1988) Monoclonal antibodies directed against Danief M. D., Letvin N. L., King N. W., Kannagi M., Sehgal human immun~eficiency virus (HIV) gag proteins with P. K., Hunt R. D., Kanki P. J., Essex M. and Desrosiers specificity for conserved epitopes in HIV-I, HIV-2 and R. C. (1985) Isolation of T-cell tropic HTLV-III-iike retrosimian immunodeficiency virus. J. Gen. Viral. 69,2109-2114. virus from macaques. Science 228, 1201-1204. Niedrig M., Hinkula J., Weigelt W., L’Age-Stehr J., Pauli G., Devaux C. A., Nadler P. I., Miller G. G. and Sachs D. H. Rosen J. and Wahren B, (1989) Epitope mapping of mono(1985) Genetic control of immune response to staphylococclonal antibodies against human immunodeficiency virus cal nuclease. XII: Analysis of nuclease antigenic determitype 1 structural proteins by using peptides. J. Viral. 63, nants using anti-nuclease monoclonal antibodies. Molec. 3525-3528. cell. Biochem. 68, 31-40. Niedrig M., Hinkula J., Harthus H. P., Briiker M., Hopp L., Di Matzo Veronese F., Copeland T. D., Oroszlan S., Gallo Pauli G. and Wahren B. (1991) Characterization of murine R. C. and Sarngadharan M. G. (1988) Biochemical and immunological analysis of human immunodeficiency virus monoclonal antibodies directed against the core proteins of human immunode~ciency virus types 1 and 2. J. Viral. 65, gag gene products ~17 and ~24. .T. Yirof. 62, 795-801. Flgerstam L. G., Frostell A., Karlsson R., Kullman M., 4529-4533. Larsson A., Malmqvist M. and Butt H. (1990) Detection of Nossal G. J. V. (1989) Immunologic tolerance: collaboration antigen-antibody interactions by surface plasmon resonance: between antigen and lymphokines. Science 245, 147-153. Application to epitope mapping. J. Molec. Recognition 3, Ohta Y., Masuda T., Tsujimoto H., Ishikawa K. I., Kodama 208-214. T., Morikawa S., Nakai M., Honjo S. and Hayami M. (1988) Ferns R. B., Tedder R. S. and Weiss R. A. (1987) CharacterIsolation of simian immunodeficiency virus from Africa ization of monoclonal antibodies against the human green monkeys and seroepidemiology survey of the virus in immunodeficiency virus (HIV) gag products and their use various non human primates. ht. J. Cancer, 41, 115-122.

738

V. ROBERT-HEBMANN et ui

Papsidero L. D., Sheu M. and Ruscetti F. W. (1989) Human immunodeficiency virus type l-neutralizing monoclonal antibodies which react with p17 core protein: characterization and epitope mapping. d. Viral. 63, 267-272. Peng C., Ho B. K., Chang T. W. and Chang N. T. (1989) Role of human immunode~ciency virus type l-specific protease in core protein maturation and viral infectivity. f. Vird. 63, 2550-2556. Pierres M., Devaux C., Dosseto M. and Marchetto S. (1981) Clonal analysis of B- and T-cell responses to Ia antigens. Immunogenetics 14, 481-495. Robert V., Resnicoff M., Chermann J. C. and Devaux C. (199 1) Characterization of monoclonal antibodies identifying type and strain-specific epitopes of human immunodeficiency virus type 1. Mol. Cell. Biochem. 102, 115-123. Sarin P. S., Sun D. K., Thornton A. H., Naylor P. H. and Goldstein A. L. (1986) Neutralization of HTLV-III/LAV replication by antiserum to thymosin CL~.Science 232, 113551137. Spence R. P., Jarvill W. M., Ferns R. B., Tedder R. S. and Parker D. (1989a) The cloning and expression in Escherichia coli of sequences coding for ~24, the core protein of human immunodeficiency virus, and the use of the recombinant protein in characterizing a panel of monoclonal antibodies against the viral p24 protein. J. Gen. Viral. 70, 2843-285 1.

Spence R. P., Walker J., Jarviil W. M., Ferns R. B., Tedder R. S., Sattentau Q., Weber J., Parry N. R. and Highfield P. E. (1989b) The expression in Escherichia coli of sequences coding for the pi 8 protein of human immunodefi~iency virus and the use of the recombinant protein in characterizing a panel of monoclonal antibodies against the viral pl8 protein. f. Gen. Viroi. 70, 2853-2863. Tatsumi M., Devaux C., Kourilsky F. and Chermann J. C. (1990a) Characterization of monocional antibodies directed against distinct conserved epitopes of human immunodeficiency virus type 1 core proteins. Molec. cell. Biochem. 96, 127-136.

Tatsumi M., Jean F., Robert V., Chermann J. C. and Devaux C. (1990b) Use of monocional antibodies for the detection and quantitation of HIV1 core protein ~25: comparative evaluation of in vitro HIV1 infection by immunofluorescence, antigen capture ELISA and reverse transcriptase assays. Res. Viral. 141, 649-661. Tersmette M., Winkel I. N., Groenink M., Gruters R. A., Spence R. P., Saman E., Van Der Groen G., Miedema F. and Huisman J. G. (1989) Detection and subtyping of HIV-l isolates with a pane1 of characterized monoclonal antibodies to HIV p24g”g. Virology 171, 149-155. Tsunetsugu-Yokota Y., Tatsumi M., Robert V., Devaux C., Spire B., Chermann J. C. and Hirsch 1. (1991) Expression of an immunogenic region of HIV by a filamentous bacteriophage vector. Gene 99, 261-265.