Characterization of human antibody-binding sites on the external envelope of human immunodeficiency virus type 2

Transcription

1 Journal of General Virology (1991), 72, Printed in Great Britain 1261 Characterization of human antibody-binding sites on the external envelope of human immunodeficiency virus type 2 Frank de Wolf, 1. Rob H. Meloen, 2 Margreet Bakker, 1 Francis Barin 3 and Jaap Goudsmit ~ 1Human Retrovirus Laboratory, Academic Medical Centre, University of Amsterdam, Meibergdreef 15, 1105 AZ Amsterdam, 2Central Veterinary Laboratory, Lelystad, The Netherlands, and 3 Laboratoire de Virologie, CHRU, H6pital Bretonneau, Tours Cedex, France Antibody-reactive peptide scanning (Pepscan) using overlapping nonapeptides and human sera as probes allows the identification of amino acids contributing significantly to antigen-antibody interaction. Fivehundred and two overlapping nonapeptides derived from the human immunodeficiency virus type 2 strain Rod (HIV-2Rod) external envelope glycoprotein gpl25 were synthesized to serve as probes for reactivity with eight sera of HIV-2-infected individuals. Fifteen antibody-binding regions were identified, among which two amino-terminal regions [E3, amino acids (aa) 118 to 132; E4, aa 125 to 141] and four carboxy-terminal regions (El l, aa 303 to 324; El2, aa 340 to 358; El4, aa 436 to 452; El5, aa 486 to 507) were the most antigenic. The amino acids in binding sites E3 and E4 were highly variable among simian immunodeficiency virus (SIV), HIV-2 and HIV-1. The antibody-binding domains E14 and El5 were highly conserved among HIV-2 strains (94% and 86% identity of HIV-2sod to HIV-2~sy and HIV-2N~UZ, respectively). Both domains had more amino acids in common with SIV (88 % for El4, 64% for El5) than with HIV-1 (41% for El4, 45% for El5). Epidemiological studies revealed that the sera of African HIV-2-infected individuals bound the E11 and E15 peptides best (31%, 8/26). The sera of African HIV-l-infected individuals showed significant levels of cross-reactivity to the HIV-2 peptides, especially to the El5 peptide, whereas the sera of European HIV-l-infected individuals showed only moderate levels of cross-reactivity. Ifpeptides covering the E 15 epitope were used, African HIV-2-positive sera showed only a low level of cross-reactivity (4 %) to E15 of HIV-1 and SIV. African HIV-l-positive sera bound the HIV-1 El5 peptide best (81%, 52/64), but showed high levels of cross-reactivity to SIV E 15 (17 %, 11/64) and HIV-2 El5 (25%, 16/64). European HIV-1- positive sera showed a high level of reactivity to HIV-1 E 15 (91%, 50/55) and a low level of cross-reactivity to the HIV-2 (2%, 1/55), and none to the SIV El5 (0/55) peptide. These results indicate that the immunodominant regions of the HIV-2 external envelope (El 1 and E15) align with the most immunodominant regions of HIV-1. Introduction AIDS is caused mainly by infection by human immunodeficiency virus type 1 (HIV-1) (Wong-Staal & Gallo, 1985). The closely related but distinct retrovirus designated HIV type 2 (HIV-2) has also been detected in West African or West Africa-related patients with AIDS (Barin et al., 1985; Clavel et al., 1986). Although HIV-2 and HIV-1 have similar biological properties and a similar overall pattern of genetic organization (Clavel, 1987), HIV-2 is more closely related to the simian immunodeficiency virus (SIV) which causes immune deficiency in rhesus macaques (Chakrabarti et al., 1987; Franchini et al., 1987; Guyader et al., 1987). Alignment of the nucleotide sequences of HIV-1, HIV-2 and SIV reveals considerable homology between HIV-2 and SIV (Chakrabarti et al., 1987). These two viruses share about 75~ overall nucleotide sequence identity, but are only distantly related to HIV-1, with about 40~o overall identity. At the protein level, the gag and pol proteins of HIV-I, HIV-2 and SIV are antigenically cross-reactive, consistent with the strong conservation of the gag andpol genes (Franchini et al., 1987). The envelope glycoprotein of HIV-1 shows 39~o identity with that of HIV-2 and 34~ with that of SIV, whereas the envelope protein of HIV-2 shows 75 ~ identity with that ofsiv (Chakrabarti et al., 1987). In contrast to the envelope glycoprotein divergence, cross-reactivity between envelope proteins of HIV-1 and HIV-2 was demonstrated by Western blotting (Tedder et al., 1988). Cross-neutralization betbetween HIV-2 and HIV-1 has also been reported (Weiss et al., 1988; Bfttiger et al., 1989, 1990) SGM

2 1262 F. de Wolf and others Antibodies detected in sera of HIV-2-infected individuals and binding to specific domains of the transmembrane part of the envelope glycoprotein of HIV-2 have been described. A 22 amino acid (aa) synthetic peptide corresponding to a SIV sequence (aa 581 to 603) having 86% aa identity with the HIV-2 sequence bound 100% (20/20) of human HIV-2-positive sera and only one of 20 HIV-l-positive (Norrby et al., 1987). Using the sera of HIV-2-infected individuals and the antibody-reactive peptide scanning (Pepscan) method (Geysen et al., 1984, 1985), we identified nine antibody-binding domains of the transmembrane glycoprotein of HIV-2 (Goudsmit et al., 1989). For these epitopes, moderate to high degrees of similarity were observed with the aligned sequences of SIV, whereas only weak to moderate similarity was found with the aligned sequences of HIV-1. Of these nine domains, eight were located at the very same position as HIV-1 gp41 epitopes for HIV-1 antibodies, indicating that the HIV-1 and HIV-2 transmembrane glycoproteins are presented similarly to the immune system. Several binding sites for human antibodies have been identified on the HIV-1 external envelope glycoprotein gp120 (Ho et al., 1987; Palker et al., 1987; Rusche et al., 1988; van Tijn et al., 1989; Neurath et al., 1990a, b). Only one epitope (aa 307 to 321) located at the third variable (V3) domain of HIV-I gpl20 has been associated convincingly with virus neutralization and cell fusion inhibition (Goudsmit, 1988; Rusche et al., 1988; Neurath et al., 1990a, b). In the present study we report the mapping of linear binding sites on the HIV-2 external glycoprotein gp125 for antibodies from sera of HIV-2-infected individuals with the use of Pepscan analysis and the antibody reactivity to the most antigenic epitopes in sera of HIV-1- and HIV-2-infected individuals measured with a synthetic peptide enzyme immunoassay (EIA). Methods Sera. Sera were obtained from the following groups: (i) 26 HIV-2- infected individuals, from both Europe and West Africa [the sera of each of these individuals reacted with the HIV-2 transmembrane envelope glycoprotein on an immunoblot (Goudsmit et al., 1989)]; (ii) 64 HIV-l-seropositive individuals from Africa; (iii) 49 HIV-1 and HIV-2 seronegative individuals from Africa; (iv) 55 HIV-l-seropositive homosexual men participating in a cohort study in Amsterdam, The Netherlands (de Wolf et al, 1988); (v) 48 HIV-1- and HIV-2- seronegative homosexual men, participating in the same cohort study. HIV-1/HIV-2 seroreactivity was measured by a commercially available EIA (Abbott Laboratories). Pepscan. The sera of eight HIV-2-infected individuals were used for Pepscan analysis. Five of these eight patients (two male, three female) were symptom-free when the sample was collected. Of the remaining three (two male, one female), one had leprosy, one was diagnosed as having Pneumocystis carinii pneumonia and one as having the AIDSrelated complex. Sera of the individual diagnosed with leprosy reacted with both HIV-2 gp40 and HIV-1 gp41. Pepscan analysis was performed with overlapping nonapeptides derived from the external envelope glycoprotein of HIV'2Rod (Chakrabarti et al., 1987) as described before (Geysen et al., 1984, 1985; Goudsmit et al., 1988 a, b). In short, Pepscan required the synthesis of each overlapping peptide in the gp125 sequence. The gp125 sequence of 511 amino acids can be read as 502 overlapping nonapeptides, in which peptide 1 consists of residues 1 to 9, peptide 2 of residues 2 to 10, and so on. EIA reactions were carried out directly with serum and conjugate solutions in polystyrene microtitre wells. The rod tips, with the synthesized peptides still bound to them, were incubated with blocking buffer, subsequently with serum and then washed. The bound antibody was detected with the use of horseradish peroxidase (HRP)-labelled goat anti-human IgG and o-phenylenediamine (OPD) as substrate. Prior to re-testing, bound antibody was removed from the peptides by three washes in 8 u-urea and then with PBS. Absorbance values were plotted against the position of the amino-terminal amino acid of the peptide in the total sequence. We considered the mean absorbance plus two times the standard deviation to be the cutoff value. Single isolated reactivities were ignored in the present analysis; at least two nonapeptides overlapping by one amino acid had to be reactive to allow a particular region to be considered as an antibody-binding domain (Goudsmit et al., t990). HIV-2 EIA. Peptides were synthesized as described elsewhere (van Tijn et al., 1989). Antibody reactivity to these peptides was determined by using a non-competitive solid phase EIA. Polystyrene microtitre wells were coated with peptide (100 ng/well) diluted in PBS. After 16 h at room temperature, the wells were washed with 0.1 ~ Tween-20 in PBS, incubated for 1 h at 37 C with 4% normal goat serum (NGS) in PBS-Tween-20 and again washed. Subsequently wells were incubated for 1 h at 37 C with serum (1 : 100 dilution in 4~ NGS in PBS-Tween- 20) and washed. Bound antibody was detected using HRP-labelled goat anti-human IgG and OPD as substrate (Goudsmit et al., 1989). Sera were considered to be positive if, for each synthetic peptide, the absorbance values (at 450 nm) were greater than the mean A~5o plus four times the standard deviation of the control groups. As the control group for the African or African-related sera, the HIV-1/HIV-2- negative African sera were used; as control group for the sera of the Dutch homosexual men we used sera from the HIV-1/HIV-2-negative Dutch homosexual men. Alignment of HIV-2, SIV and HIV-I. Alignment of HIV-2Rod envelope protein with SIVMac and HIV-1Br. corresponds to that described by Chakrabarti et al (1987). Alignment of the HIV-2 strains Rod, Isy and NIHZ corresponds to that described by Myers et al. (1990). Results Pepscan analysis Pepscan analysis of sera from eight HIV-2-infected individuals identified a total of 15 antibody-reactive sites on the HIV-2 gp125 (Table 1). Antibody reactivity observed by peptide EIA All sera from the different risk groups were tested by EIA against synthetic peptides representing the Pepscanselected regions of the external glycoprotein of HIV- 2god (Table 2). We focused on the antibody reactivity to

3 Antibody-binding sites on HIV-2 external envelope 1263 Table 1. Epitopes of the HIV-2Rod external envelope-b&ding sera from eight HIV-2-infected individuals Identity (%) with Epitope No. of sera (aa Pepscan- SIV HIV-1 HIV-2 HIV-2 residues) Amino acid sequence reactive Mac* Bru* Isy t NIHZ t E1 (78-88) TVTEQAIEDVW E2 ( ) MKCSSTESSTGNNTTS E3 ( ) GNNTTSKSTSTTTTT E4 ( ) STSTTTTTPTDQEQEIS E5 ( ) DNCSGLGEEETINCQ E6 ( ) TGLERDKKKQYNET E7 ( ) SKDVVCETNNSTNQTQC E8 ( ) TSVITESCDK E9 ( ) ALLRCNDTN El0 ( ) STCTRMMETQTS El I ( ) LHCKRPGNKTVKQIMLMSGHVF El2 ( ) CWFKGKWKDAMQEVKETLA El3 ( ) FAAPGKGSDPEVAYM El4 ( ) YLPPREGELSCNSTVTS El5 ( ) LVEITPIGFAPTKEKRYSSAHG * Chakrabarti et al. (1987). l" Myers et al. (1990). synthetic peptide analogues of epitopes E3, E4, E11, El2, El4 and especially El5, as the most antigenic epitope of HIV'2Rod. Antibody reactivity to HIV-2 El5 was found in 31~o of the HIV-2 immunoblot-positive African sera. Twenty-five per cent of the HIV-l-positive African sera showed reactivity to HIV-2 El5, whereas 2% of the HIV-l-positive Dutch sera showed reactivity to El5 of HIV-2. Reactivity to El5 of SIVMac was low in the sera of all groups, except for the HIV-l-positive African sera. As expected, antibody reactivity to HIV-1 E 15 occurred most frequently in HIV-l-positive African and Dutch sera. Only one (4%) of the HIV-2-positive sera was positive for HIV-1 El5. In the HIV-1/HIV-2- negative control sera from both African individuals and Dutch homosexual men the frequency of sera reactive to El5 of HIV-2, HIV-1 or SIV was low. Cross-reactivity for HIV-2 El5 and HIV-1 El5 was most frequently found in the HIV-l-positive African sera, whereas there was no cross-reactivity for these synthetic peptides in the HIV-2 immunoblot-positive African sera. Sera of the HIV-1-positive African patients also showed more cross-reactivity for El5 of SIV. No cross-reactivity for E 15 HIV-2 and E 15 SIV was found in sera from the HIV-2 immunoblot-positive patients. Remarkably, 10% of HIV-1- and HIV-2-negative African sera reacted with HIV-2 E4, with positive absorbance values ranging from to > (Table 2). Alignment with HIV-1 and SIV Alignment of the HIV-2Ro~ strain, the SIVMac strain and the HIV-1Bru strain (Chakrabarti et al., 1987) points to a high degree of similarity (80 to 100%) between the aligned sequence of HIV-2Rod and SIVMa for epitopes El, E6, E8, E9, El0 and El4, and moderate similarity (50 to 80%) for E5, E7, E11, El2, El3 and El5. Epitopes E2, E3 and E4 diverged most between HIV-2Ro~ and SIVMa c (Table 1). Only moderate similarity (50 to 80~ aa identity) with the aligned sequences of HIV-2Rod and HIV-1Bru was found for E8 and E9. The other HIV-2Rod epitopes showed less than 50% aa identity with the aligned sequences of HIV-1Bru. On the assumption that the secondary structure of the external envelope of HIV-2Rod and that of HIV-1Br u are comparable (Leonard et al., 1990), the positions of their respective epitopes are depicted in Fig. 1. The partially overlapping HIV-2Ro~ epitopes E3 and E4 in the aminoterminal half of the external envelope glycoprotein are localized at a position comparable to the first variable (V1) domain of HIV-1B~ u. Epitope Ell of HIV-2Rod is localized at a position comparable to the V3 region, and epitopes E14 and E15 at positions comparable to the third and fourth conserved domains of HIV-1Br u. Epitope El5 of HIV-2Rod, the most antigenic epitope, encompasses the cleavage site of gp125-gp36 of HIV-2. This epitope is situated at the very same location as the epitope for HIV-I antibodies, although the primary sequences showed only 45% aa identity. Alignment with different HIV-2 strains Alignment of the HIV-2 strains Rod, Isy and NIHZ (Myers et al., 1990) showed an overall identity for the external envelope glycoprotein of 79% between

4 1264 F. de Wolf and others KY SGFA I I ~a3, P T Y Y WH ~ ~ "7,, R/ Vv E D A T N ~N~#"~C E T ~ N AC~TX F I ~ T N ~ _ YR ~RM\ EIO G SL ~ ~ M ~ _ ES' ~'%'~I ~LPT LKV C PK' STEF~ AQ~ 12 -QRpRKN I PQYH I'~R ~V " ~EI~" ~(k LNW ' EN K.r :~" ~E\_ l~1 HP N ~U R v R M N pa" ~-CHIKQll / ~ FoA w DN? ~.L NTWN K E4- ~ //"7/i,T EA RT R G k'~ JIM/.v ot~.,.,~wrn..%~...n" wo,,oj t%,o.. " r,c Ela OWON DR N RTACFLPI v.c,o,v,v.,v.,w T i NT on:i= ii yrlelg Dy l~. )'1 Fig. 1. Secondary structure of the external envelope of HIV-2R~ as compared to the secondary structure of the external envelope of HIV-1 (Leonard et al, 1990) and location of the respective epitopes E1 to El5. Table 2. Antibody reactivity to selected HIV-2Ro ~ HIV-Ia~, and SIVMa c synthetic peptides Percentage of sera positive (A45o range) HIV- 1Br, SIVMac Percentage of sera positive (A45o range) to HIV-2Rod Risk group El5* El5t El5 E14 E12 Ell E4 E3 HIV-2 African sera, n = 26 (3" 114) (0.276) ( ( ) > 3.000) HIV-1 African sera, n=64 ( ( ) ( ( ( ( > 3.000) 1'324) 0-906) 0-644) 0.840) HIV-I/2- African sera, n=49 (0.893) (0-732) (0.811) HIV-1 Dutch sera, n=55 ( (1.164) ( ( > 3.000) 1.130) 2.705) HIV-1/2- Dutch sera, n=48 (0.508) (1-308) (0.217) 4 0 (1.492) ( ( ) 1.819) 10 0 ( > 3.000) 9 13 ( ( > 3.000) 0.515) 2 2 (0-249) (0-164) * aa , VVKIEPLGVAPTKAKRRVVQRE. t" aa , LVEITPIGLAPTNVKRYTTGGT. HIV-2lsy and HIV-2Rod and 81 ~ between HIV-2NIHZ and HIV-2Roa. Epitopes E2, E3 and E4 diverged most between the different HIV-2 variants (identity less than 50~; Table 1), followed by epitope Ell. Discussion Pepscan analysis of sera from eight HIV-2-infected individuals revealed 15 antibody-binding domains on the

5 Antibody-binding sites on HIV-2 external envelope 1265 Table 3. Antibody-binding domains on HIV-1 and HIV-2 external envelope Peptide EIA* Pepscan Peptide EIA HIV-1 (No. positive/ HIV-2 (No. positive/ (No. positive/ aa number total number) aa number total number) total number) / / / / / / /8 0/ /8 1/ / / / / / / / / / / / /8 8/ / /8 0/ / / / /8 0/ / / /8 8/26 * Neurath et al. (1990b). external envelope glycoprotein of HIV-2. Two partially overlapping epitopes, i.e. E3 (aa 118 to 132) and E4 (aa 125 to 141), located at the amino terminus and four epitopes, E11 (aa 303 to 324), El2 (aa 374 to 388), El4 (aa 436 to 452) and El5 (aa 486 to 507), at the carboxyl terminus of gp125, were the most antigenic. The results of the Pepscan analysis and the synthetic peptide EIA differed, particularly with respect to reactivity to E3, E 12 and E14. In the African HIV-2-positive sera no reactivity to these domains was found using the synthetic peptide EIA. This discordance is probably due to differences in techniques (e.g. differences in coating) or to conformational differences. The number of antibody-binding domains on the HIV-2 external glycoprotein identified by the Pepscan method is in contrast with the number of antibodybinding domains on the HIV-I external glycoprotein, detected by the same method (van Tijn et al., 1989). Only three domains located in the first constant region (Cl), the first variable (V1) and the third variable region (V3) of gpl20 of HIV-IIIIB(BH10) were identified in sera from HIV-l-infected individuals from Europe. Using synthetic peptides overlapping the complete envelope of niv-lni B and mouse antisera, Ho et al. (1987) identified six antibody-binding domains on gp 120 of HIV-1, two of which (aa 298 to 314 and aa 458 to 484/503 to 532) were shown to be important determinants for antibody neutralization. Fifteen domains with antibody-binding capacity on the external envelope glycoprotein of HIV- IlIIB(BHI0) were recently described and antibody reactivity to a domain encompassing aa 303 to 338 was seen most frequently in sera from symptomatic and asymp- tomatic HIV-l-infected individuals (Neurath et al., 1990b). Epitopes E3, E4 and Ell of HIV-2Rod external glycoprotein showed moderate to low similarity to the aligned domains of the HIV-2 strains Isy and NIHZ (Myers et al., 1990) and are located in regions comparable to the V1 domain of HIV-1 (E3 and E4) and the V3 domain of HIV-1 (El 1). Thus, epitopes E3, E4 and E11 apparently represent variable antigenic domains of HIV- 2 gp125, E11 being situated in a region comparable to the principal neutralizing domain of HIV-1. Type-specific antibody reactions against a 35 aa peptide of HIV- 2SBL66692, a peptide equivalent to Ell of HIV-2Ro d, correlate with neutralization in HIV-2 antibody-positive serum samples (B6ttinger et al., 1990). In that study, cross-reactivity to this peptide in HIV-l-positive samples was found in 23 ~ of the sera and reactivity to an HIV- 1rob peptide in HIV-2-positive samples was found in 33 ~ of the sera. In contrast, in our study, cross-reactions to E3, E4 and E11 in HIV-l-positive African sera and in HIV-l-positive sera of Dutch homosexual males were seen infrequently. Reactivity to synthetic peptides representing E3, E4 and E11 was also infrequent in HIV- 1 and HIV-2 antibody-negative sera from African individuals and Dutch homosexuals, except for E4 in 10 ~ of the negative African sera; the latter observation is as yet unexplained. Epitope E 12 represents a relatively conserved domain of HIV-2 gpl25. Homology with the aligned regions of HIV-1, however, is low and no comparable Pepscanreactive region of HIV-1 has been described. Remarkably, in the pool of HIV-2 antibody-positive sera no reactivity to E 12 was found in the synthetic peptide EIA; the reactivity in the HIV-l-positive African and Dutch sera was seen relatively infrequently and absorbance values of positive sera were low. These results correspond to those of a study where reactivity to a peptide corresponding to aa 331 to 361 of HIV-1 gp120 was found in immunized rabbits, but was not in a pool of human anti-hiv-l-positive sera (Neurath et al, 1990a). Epitopes El4 and El5 of HIV-2Roa were most similar to the aligned domains of the HIV-2 strains Isy and NIHZ, and compared to HIV-1 are localized in the third (El4) and fourth (E15) conserved regions. However, similarity between HIV-2 and HIV-1 for these domains was relatively low. As with epitopes E3 and El2, no reactivity to El4 was found in the pool of HIV-2-positive sera, when the synthethic peptide EIA was used. Only a few HIV-l-positive African sera there were found to be reactive at a low level. Epitope E 15 was most reactive in the Pepscan analysis of HIV-2-positive sera and 31 ~ of the sera in the anti- HIV-2-positive pool were reactive in synthetic peptide EIA. These results agree with those obtained in a

6 1266 F. de Wolf and others separate study done by others (A. Baillou, personal communication). Also, cross-reactivity was found mainly in the pool of HIV-l-positive African sera. On the other hand, almost no reactivity to a comparable HIV-1Bru peptide was seen in the pool of anti-hiv-2-positive sera, whereas 81% of the HIV-l-positive African, and 91% of the Dutch sera were reactive. Cross-reactivity to a comparable peptide of SIVMa c was found only in the HIV- 1-positive African sera. E 15 of HIV-2 encompasses the cleavage site of the envelope glycoprotein (Chakrabarti et al., 1987) and a comparable HIV-1 epitope (aa 507 to 518) was described recently as highly immunogenic but non-neutralizing (Palker et al, 1987; Neurath et al., 1990a). The results of our study indicate that this epitope is highly immunogenic in HIV-2 also. Neurath et al. (1990b) found 15 antibody-binding domains on the HIV-1 external glycoprotein when they used a synthetic peptide ELISA and sera from HIV-1- infected individuals. These HIV-1 antibody-binding domains align partially with the HIV-2 antibody-binding domains determined by Pepscan analysis (Table 3). When the synthetic peptide EIA results on HIV-2 antibody binding are taken into account, the immunodominant regions of the HIV-2 external envelope (E 11 and E 15) appear to be located in an area comparable to the most immunodominant regions of the external envelope of HIV-1. This study was supported in part by the Netherlands Foundation for Preventive Medicine (grant no ) and by the Working Party on AIDS Research of the European Community (Concerted Action on the Pathophysiology and Immunology of HIV-related Diseases). References BARIN, F., M'BOUP, S., DENIS, F., KANKI, P., ALLAN, J. S., LEE, T. H. & ESSEX, M. (1985). Serological evidence for virus related to simian T- lymphotropic retrovirus III in residents of West Africa. Lancet ii, BOTTIGER, B., KARLSSON, A., NAUCLf/R, A., ANDREASSON, P.-A., MENDES COSTA, C. & BIBERFIELD, G. (1989). Cross-neutralizing antibodies against HIV-1 (HTLV-IIIB and HTLV-IIIRF) and HIV- 2 (SBL-6669 and a new isolate SBL-K 135). Aids Research and Human Retrovirus 5, BOTTIGER, B., KARLSSON, A., ANDREASSON, P.-A., NAUCLI~R, A., MENDES COSTA, C., NORRBY, E. & BIBERFIELD, G. (1990). Envelope cross-reactivity between human immunodeficiency virus types l and 2 detected by different serological methods: correlation between cross-neutralization and reactivity against the main neutralizing site. Journal of Virology 64, CHAKRABARTI, t., GUYADER, M., ALIZON, M., DANIEL, M. D., DESROSIERS, R. C., TIOLLAIS, P. & SONIGO, e. (1987). Sequence of simian immunodeficiency virus from macaque and its relationship to other human and simian retroviruses. Nature, London 328, CLAVEL, F. 0987). HIV-2, the West African AIDS virus. AIDS 1, CLAVEL, F., GUETARD, n., BRUN-VEZINET, F., CHAMARET, S., REY, M.-A., SANTOS-FERREIRA, M. O., LAURENT, A. G., DAUGUET, C., KATLAMA, C., ROUZIOUX, C., KLATZMANN, D., CHAMPALIMAUD, J. L. & MOlCI'AGNIER, L. (1986). Isolation of a new human retrovirus from West African patients with AIDS. Science 233, DE WOLF, F., LANGE, J. M. A., HOUWELING, J. T. M., Cotrrimto, R. A., SCHELLEKENS, P. T., VAN DER NOORDAA, J. & GOUDSMIT, J. (1988). Numbers of CD4 + cells and the levels of core antigens of and antibodies to the human immunodeficiency virus as predictors of AIDS among seropositive homosexual men. Journal of Infectious Diseases 158, FRANCHINI, G., GURGO, C., Guo, H. G., GALLO, R. C., COLLALTI, E., FARGNOLI, K. A., HALL, L. F., WONG-STAAL, F. & REITZ, i. S., JR. (1987). Sequence of simian immunodeficiency virus and its relationship to the human immunodeficiency viruses. Nature, London 328, GEYSEN, H. M., MELOEN, R. H. & BARTELING, S. J. (1984). Use of peptide synthesis to probe viral antigens for epitopes to a resolution of a single amino acid. Proceedings of the National Academy of Sciences, U.S.A. 81, GEYSEN, H. M., BARTELING, S. J. & MELOEN, R. H. (1985). Small peptides induce antibodies with a sequence and structural requirement for binding antigen comparable to antibodies raised against the native protein. Proceedings of the National Academy of Sciences, U.S.A. 82, GOUDSMIT, J. (1988). Immunodominant B-cell epitopes of the HIV-1 envelope recognized by infected and immunized hosts. AIDS 2, (supplement 1), $41-$45. GOUDSMIT, J., DEBOUCK, C., MELOEN, R. H., SMIT, L., BAKKER, M., ASHER, D. M., WOLFF, A. V., GIBBS, C. J. & GAJDUSEK, D. C. (1988 a). Human immunodeficiency virus type 1 neutralizing epitope with conserved architecture elicits early type-specific antibodies in experimentally infected chimpanzees. Proceedings of the National Academy of Sciences, U.S.A. 85, GOUDSMIT, J., BOUCHER, C. A. B., MELOEN, R. H., EPSTEIN, L. G., SMIT, L., VAN PER HOEK, L. & BAKKER, M. (1988b). Human antibody response to a strain-specifc HIV-1 gpl20 epitope associated with cell fusion inhibition. AIDS 2, GOUDSMIT, J., MELOEN, R. H., BRASSEUR, R. & BARIN, F. (1989). Human B-cell epitopes of HIV-2 transmembrane protein are similarly spaced as in HIV-1. Journal of Acquired Immune Deficiency Syndromes 2, GOUDSMIT, J., MELOEN, R. H. & BRASSEUR, R. (1990). Map of sequential B cell epitopes of the HIV-1 transmembrane protein using human antibodies as probe. Intervirology 31, GUYADER, M., EMERMAN, M., SONIGO, P., CLAVEL, F., MONTAGNIER, L. & ALIZON, M. (1987). Genome organization and transactivation of the human immunodeficiency virus type 2. Nature, London 326, Ho, D. D., SARNGADHARAN, M. G., HIRSCH, M. S., SCHOOLEY, R. T., ROTA, T. R, KENNEDY, R. C., CHANH, T. C. & SATO, V. L. (1987). Human immunodeficiency virus neutralizing antibodies recognize several conserved domains on the envelope glycoprotein. Journal of Virology 61, LEONARD, C. K., SPELLMAN, M. W., RIDDLE, L., HARRIS, R. J., THOMAS, J. N. & GREGORY, T. J. (1990). Assignment of interchain disulfide bonds and characterization of potential glycosylation sites of the type 1 recombinant human immunodeficiency virus envelope glycoprotein (gpl20) expressed in Chinese hamster ovary cells. Journal of Biological Chemistry 265, MYERs, G., RABSON, A. B., BERZOFSKY, J. A., SMITH, T. F. & WONG- STAAL, F. (editors) (1990). Human Retroviruses and AIDS. Los Alamos: Los Alamos National Laboratory. NEURATH, A. R., STRICK, N. & LEE, E. S. Y. (1990a). B cell epitope mapping of human immunodeficiency virus envelope glycoproteins with long (19- to 36-residue) synthetic peptides. Journal of General Virology 71, NEURATH, A. R., STRICK, N., TAYLOR, P., RUBINSTEIN, P. & STEVENS, C. E. (1990b). Search for epitope-specific antibody responses to the human immunodeficiency virus (HIV-I) envelope glycoproteins signifying resistance to disease development. AIDS Research and Human Retroviruses 6, NORRBY, E., BIBERFELD, G., CHIOIDI, F., VON GEGERFELDT, A., NAUCL~R, A. PARKS, E. & LERNER, R. (1987). Discrimination between antibodies to HIV and to related retroviruses using sitedirected serology. Nature, London 329,

7 Antibody-binding sites on HIV-2 external envelope 1267 PALKER, T. J., MATI'I-IEWS, T. J., CLARK, M. E., CIANCIOLO, G. J., RANDALL, R. R., LANGLOIS, A. J., WHITE, G. C., SAFAI, B., SNYDERMAN, R., BOLOGNESI, D. P. & HAYNES, B. F. (1987). A conserved region at the COOH terminus of the human immunodeficiency virus gpl20 envelope protein contains an immunodominant epitope. Proceedings of the National Academy of Sciences, U.S.A. 84, RUSCHE, J. R., JAVAHERIAN, K., MCDANAL, C., PETRO, J., LYNN, D L., GRIMAILA, R., LANGLOIS, A., GALLO, R. C., ARTHUR, L. O., FISCHINGER, P. J., BOLOGNESI, D. P., PUTNEY, S. D. & MATrHEWS, T. J. (1988). Antibodies that inhibit fusion of human immunodeficiency virus-infected cells bind a 24-amino acid sequence of the viral envelope, gpl20. Proceedings of the National Academy of Sciences, U.S.A. 85, TEDDER, R. S., HUGHES, T., O'CONNOR, T., N'JIE, H. & WHITTLE, H. (1988). Envelope cross-reactivity in Western blot for HIV-1 and HIV-2 may not indicate dual infection. Lancet ii, VAN TIJN, D. A., BOUCHER, C. A. B., BAKKER, M. & GOUDSM1T, J. (1989). Antigenicity of linear B-cell epitopes in the C1, VI and V3 region of HIV-1 gpl20. Journal of Acquired Immune Deficiency Syndromes 2, WEISS, R. A., CLAPHAM, P. R., WEBER, J. N., WHITBY, D., TEDDER, R. S., O'CONNOR, T., C~ARET, S. &Mor, rragnmr, L. (1988). HIV- 2 antisera cross-neutralize HIV-1. AIDS 2, WONG-STnnL, F. & GALLO, R. C. (1985). Human T-lymphotropic retroviruses. Nature, London 317, (Received 13 December 1990; Accepted 18 February 1991)