The presence of IgA that can neutralize HIV-1 in vitro has been demonstrated in serum and parotid saliva of HIV-1-seropositive patients [1–4]. Recent studies have shown that some individuals who are highly exposed to HIV-1 but who remain persistently HIV-1 seronegative have mucosal IgA HIV-1-specific antibodies [5,6]. This suggests that antibodies of IgA isotype have the potential to neutralize mucosal transmission. In serum and saliva, the anti-HIV IgA response is predominantly against the envelope protein of HIV-1 [7,8]. The identification and characterization of epitopes of the envelope protein that elicit neutralizing IgA antibodies may direct the design of vaccine protecting against mucosal transmission of HIV-1. Initial infection depends on binding of the viral envelope protein gp120 to CD4 and coreceptors. The loops V1, V2 and V3 of gp120, which are determinant of coreceptor usage, have been demonstrated to be target epitopes of neutralizing antibodies. However, these regions are highly variable between HIV-1 isolates. The CD4-binding domain, which contains the CD4-binding site and amino acids present in the C1, C2, C3, V4 and C4 regions of gp120, is well conserved between HIV-1 isolates [9–12]. Antibodies of the IgG isotype against the CD4-binding domain have been shown to neutralize a wide range of virus strains [11,12]. Kennedy et al.  have shown that a peptide representing a region of the C-terminal tail of gp41 (residues 731–752) was recognized by sera from HIV-1-seropositive individuals. Sera from HIV-1-infected individuals containing antibodies to the Kennedy epitope have neutralizing activity to HIV-1 .
The present study examines the presence of IgA antibodies against the CD4-binding domain of gp120 and to the Kennedy epitope in serum and saliva of HIV-1-infected patients.
Parotid saliva and serum samples were obtained with informed consent from HIV-1-antibody-positive patients at the Department of Infectious Diseases, University Hospital of Saint-Etienne, France. Patients were classified according to the Centers for Disease Control (CDC) staging system as asymptomatic. All patients were receiving antiviral combination therapy. In addition, 10 HIV-1-seronegative healthy individuals were recruited as controls.
Serum and parotid saliva collection procedures
The parotid saliva was collected by catheterization of Stensen's duct, as previously described . Upon receipt, within 1 h of collection, all samples were centrifuged at 400 × g for 10 min at 4°C and stored at −70°C until use. Before testing for antibodies, samples were heated for 30 min at 56°C.
Peptides were used as antigens in enzyme-linked immunosorbent assays (ELISA): the C3–V4 region represented by peptides ARP7035.35 (residues 367–386) SSGGDPEIVRHSFNCGGEFF, ARP7035.36 (residues 377–396) HSFNCGGEFFYCDTTQLFNS and ARP7035.37 (residues 387–406) YCDTTQLFN STWNGTEGNNT, and the Kennedy epitope represented by peptides 2038 (residues 731–750) PE GIEEEGGERDRDTSGRLV and 2039 (residues 741–762) RDRDTSGRLVHGFLAIIWVD.
Measurement of total immunoglobulins A, G and M
Saliva and serum IgA content was determined as previously described . Briefly, 96-well plates were coated with affinity-purified goat anti-human IgA (Cappel, Aurora, Ohio, USA) diluted in 0.1 mol/l carbonate buffer pH 9.6. IgA standard and IgA derived from sera and saliva were added to wells in serial dilutions and binding was detected by goat anti-human IgA conjugated to horseradish peroxidase (HRP) (Biorad, Hercules, California, USA). The peroxidase was detected by the addition of o-phenylenediamine dihydrochloride (OPD) (Sigma, Poole, UK) in citrate buffer, pH 5.6. The reaction was stopped after 10 min with 1 mol/l HCl and the activity was determined by measuring the absorbance at 492 nm. A calibrated human serum was used as a standard for the serum IgA determination. Purified colostrum secretory IgA (Sigma) was used as a standard for the salivary IgA determination. For each saliva sample, IgA was also detected with a mouse monoclonal antibody against the human secretory component (Sigma) followed by HRP–anti-mouse IgG. The method to measure total IgG and total IgM was the same as that described for the measurement of IgA. Plates were coated with goat anti-human IgG (Cappel) or goat anti-human IgM (5Fc Mu; Cappel). For the IgG and IgM standards, a calibrated human serum and an IgM standard, respectively, were used (Dade Behring, Deerfield, Illinois, USA) and detection was with HRP–anti-human IgG whole molecule (Sigma) or HRP-anti-human IgM (Mu; Biorad), respectively.
Purification of serum and saliva IgA
IgA was separated from IgG by anion-exchange chromatography using a DEAE resin, grade DE52 from Whatman (Clifton, New Jersey). USA. Serum (200 μl), parotid saliva (1–2 ml) or pooled parotid saliva (10 ml) diluted 1:1 in buffer A (0.04 mol/l NaCl, 0.02 mol/l Tris-HCl, pH 7.2) were applied to a column packed with 0.5 g resin equilibrated with buffer A. IgG was collected from the column effluent. Retained IgG was recovered from the column by adding 2 ml buffer A. IgA was then eluted by adding 2.5 ml buffer B (0.34 mol/l NaCl, 0.02 mol/l Tris-HCl, pH 7.2). The eluate containing IgA and IgM was successively treated with protein G coupled to Sepharose beads to adsorb remaining IgG  and with an anti-human IgM (μ-chain-specific) agarose resin (Sigma) to remove IgM until the presence of IgG and IgM was undetectable by ELISA.
Construction of fusion proteins
An isopropyl-β-d-thiogalactopyranoside (IPTG)-inducible expression plasmid pMAL-c2E (New England BioLabs, Beverly, Massachusetts, USA) was used to generate recombinant maltose-binding protein (MBP) fusion proteins containing fragments of the HIV-1 envelope protein. The foreign sequence was inserted in-frame, downstream of the Escherichia coli malE gene, which encodes MBP. The plasmid pMBP24 consisted of a XbaI (nucleotide 1080)–BamHI (nucleotide 1387) fragment from p92/BR/014–1 (ARP239-12) cloned into pMAL-c2E . The corresponding protein MBP24 contained the amino acid sequence 362–466 (numbering is based on the sequence of the gp160 of the HIV-1 HXB2) including the CD4-binding site. The plasmid pMBP42 resulted from the insertion of an ApoI (nucleotide 1980)–PstI (nucleotide 2381) fragment from pGB8.C4 (ARP210) into pMAL-c2E . The MBP42 construct contained the amino acid sequence 668–801 (Kennedy epitope). The inserted sequences were verified by sequencing. Protein expression was induced by addition of 0.8 mmol/l IPTG for 2 h at 37°C. Cells were harvested by centrifugation and resuspended in column buffer (CB: 10 mmol/l Tris-HCl, pH 7.4, 200 mmol/l NaCl, 1 mmol/l ethylenediaminetetraacetic acid). Proteins were released in the supernatant by cell sonication. Purification followed the manufacturer's protocols. Purified MBP proteins were separated by sodium dodecylsulfate–polyacrylamide gel electrophoresis and detected by Western blotting using an anti-MBP rabbit antiserum (New England BioLabs) and/or specific monoclonal antibodies.
Purification of anti-gp160 IgA from HIV-infected patients
Anti-gp160 IgA was purified by using MBP constructs adsorbed to amylose agarose resin. Sonicated bacterial extract was centrifuged and 500 μl of supernatant containing HIV-1 recombinant MBP protein was applied to a column of amylose–agarose (500 μl) equilibrated in CB. After washing with CB, IgA purified from pooled saliva or sera as described above was diluted 1:1 in CB and applied to the column. The resin was then extensively washed with CB. Antibodies bound to the immobilized MBP construct were eluted with 2 ml 0.1 mol/l citrate buffer (pH 3.0). The eluate was immediately brought to pH 7.5 with 2 mol/l Tris-HCl (pH 9). The IgA content was determined by ELISA.
Detection of anti-gp160 IgA antibodies
Microtitre plates were coated with 150 μl 0.5% dextrin in Tris-buffered saline (TBS) overnight at 4°C. The plates were washed with TBS containing 0.5% Tween 20 and blocked with the same buffer plus 2% bovine serum albumin (BSA) for 1 h at 37°C. Sonicated bacterial extract containing MBP–fusion protein was centrifuged and 100 μl of the supernatant (500 μg protein/ml) was added to the wells for 3 h at 37°C. After washing, 100 μl purified serum or saliva IgA was added at the appropriate concentration to the wells and incubated overnight at 4°C. After washing, the antibodies were detected with successively HRP- conjugated goat anti-human IgA, α-chain specific (Biorad) and OPD substrate. Absorbance was measured at 492 nm. The cut-off value was defined as the mean value of absorption derived for negative control samples (obtained from HIV-1-seronegative persons) plus 2 SD. For peptide ELISA, microtitre plates were coated with 0.5 μg peptide per well and processed as described above.
Testing of gp120–soluble CD4 binding
Binding of gp120 to soluble CD4 (sCD4) was measured by the procedure of Moore . D7324, an affinity-purified polyclonal antibody (Aalto Bioreagents, Dublin, Eire) raised in sheep against the conserved C-terminus of gp120, was adsorbed onto assay plate-wells overnight at 5 μg/ml in sodium bicarbonate buffer, pH 9.6. Unbound antibodies were removed by washing four times with phosphate-buffered saline (PBS) containing 0.05% Tween 20. The wells were then blocked for 1 h at 37°C with 100 μl PBS–Tween 20 containing 2% BSA. After washing, recombinant gp120 BRU (produced from recombinant vaccinia virus ARP256-infected cells) was bound by incubation at 0.8 μg/ml in PBS for 1 h at 37°C. Unbound recombinant gp120 was washed from the plate and recombinant sCD4 (obtained from the NIH AIDS Research and Reference Reagent Program) was applied for 3 h at 37°C (0.5 μg/ml) in the presence or absence of purified IgA. Bound sCD4 was detected by mouse CD4V4 antibody (clone L120; Becton- Dickinson, San Jose, California, USA) followed by sheep anti-mouse IgG conjugated to horseradish peroxidase (Sigma). Colour development was then achieved with OPD substrate and absorbances were measured at 492 nm.
Reactivity of IgA antibodies from sera and saliva
The binding ability of IgA purified from sera and saliva from HIV-1-seropositive patients was measured to peptides ARP7035.35, ARP7035.36 and ARP7035.37, which contained residues involved in CD4 binding to gp120 [9–12], and to peptides 2038 and 2039, which represented the Kennedy epitope. IgA antibodies purified from saliva were secretory, as shown by measuring the amount of IgA in ELISA with an anti-human IgA and with an anti-human secretory component, using purified secretory IgA from colostrum as a standard (not shown). Total IgA concentrations used in this study were chosen after preliminary experiments showing that specific HIV reactivity against the peptides and MBP recombinant proteins was lower in sera than in saliva (Table 1). Serum and saliva IgA antibodies reacted to overlapping peptides corresponding to residues 367–406 of HIV-1 gp120. The serum and saliva IgA antibody response to peptides representing the Kennedy epitope was mainly directed to the peptide 2039. IgA antibodies purified from serum and saliva of five HIV-1-seropositive patients (A–E) were tested for their ability to bind to recombinant proteins MBP24 and MBP42 (Fig. 1). MBP24 contained the CD4-binding site and MBP42 the Kennedy epitope, respectively. Serum and saliva IgA of HIV-1-seropositive patients recognized MBP24 and MBP42. Serum and saliva IgA of patient B showed the strongest reactivity against both constructs.
Inhibition of interaction of gp120 with soluble CD4 by IgA antibodies purified from sera and saliva of HIV-1-seropositive patients
IgA were purified from sera and parotid saliva obtained from HIV-1-infected patients and tested for their ability to inhibit gp120–sCD4 interaction. Representative experiments are shown in Fig. 2. Figure 2a–d shows data obtained with 11 total serum IgA and 10 total saliva IgA collected from HIV-1-seropositive patients. Two serum IgA and two saliva IgA inhibited significantly the binding of gp120 to sCD4. Figure 2e,f, used paired sera and saliva obtained from five HIV-1-seropositive patients. Two of the five serum IgA samples and one of the five saliva IgA samples were able to reduce by 50% the interaction between gp120 and sCD4.
A total of 26 sera and 25 saliva samples from HIV-1-seropositive individuals have been tested for their capacity to inhibit the gp120–sCD4 interaction. Six serum IgA and five saliva IgA inhibited by approximately 50% the binding of gp120 to sCD4. Purified serum and saliva IgA obtained from control subjects did not inhibit the gp120–sCD4 interaction.
Inhibition of the gp120–soluble CD4 interaction by IgA purified by affinity binding to recombinant proteins containing the CD4-binding site and the Kennedy epitope
Pooled sera IgA or pooled saliva IgA was purified by affinity using agarose-immobilized MBP24 or MBP42 and its ability to inhibit the gp120–sCD4 interaction was examined. A representative experiment is shown in Fig. 3. Pool 1 and pool 2 contained serum and saliva samples, respectively, obtained from the same group of seven patients. Serum and saliva IgA antibodies purified by affinity using agarose-immobilized MBP24 decreased the interaction of gp120 with sCD4. As expected, serum and saliva IgA purified by affinity using agarose-immobilized MBP42 had no effect on the interaction of gp120 with sCD4.
This study demonstrates that IgA purified from serum and saliva of HIV-1-seropositive individuals can inhibit the gp120–sCD4 interaction. ELISA peptide assays indicated that IgA directed to a region involved in CD4 binding was present in serum and saliva of HIV-1-seropositive individuals, confirming previous studies [4,20]. In our study, the IgA reactivity against the MBP fusions with envelope proteins and the peptides is rather low. However, this may be because the gp160 sequence is highly variable among HIV-1 isolates. In addition, the use of peptides or recombinant proteins produced in E. coli to measure the antibody response allows the detection of linear epitopes but not conformation-dependent epitopes. The gp120-binding site for the CD4 cellular receptor is known to be a discontinuous conformational epitope involving the conserved residues 256–262, 368–370, 384, 427–430 and 447–457 [9–12]. The MBP24 construct used in this study contained amino acid residues 360 to 463 and was able to bind sCD4 (N. Vincent, unpublished data). Using MBP24 as a substrate, an original method has been developed to purify by affinity IgA able to recognize the CD4-binding domain. Serum and saliva IgA purified on agarose-immobilized MBP24 inhibited the interaction of gp120 with sCD4. MBP24 represents a large fragment of gp120 (104 residues); consequently, IgA purified on agarose-immobilized MBP24 may contain antibodies with different epitope specificities. Recently, it has been demonstrated that the region represented by MBP24 also contains residues implicated in coreceptor binding . MBP42 and the peptide representing the Kennedy epitope were recognized by serum and saliva IgA. A recent report suggested that the Kennedy epitope may be looped out on the surface of the virion . The capacity of this region to elicit neutralizing antibody is still a subject of controversy [23–25].
Previous studies have demonstrated that vaginal mucosal transmission of HIV-1/SIV chimeric virus in macaques could be prevented by passive immunization with a mixture of anti-gp160 monoclonal antibodies of the IgG isotype, including one against the CD4-binding domain of gp120 . A similar study indicated that the administration of a human monoclonal antibody against the CD4-binding domain (b12) was able to protect macaques against the vaginal transmission of HIV-1/SIV chimeric virus , which suggests that the receptor CD4 is involved in the vaginal transmission of HIV-1. In the female genital tract, antibody against the CD4-binding domain may thus inhibit the attachment of HIV-1 to intraepithelial or subepithelial dendritic cells and prevent the infection of mucosal T cells. Therefore, it will be of interest to examine if the anti-HIV IgA antibodies found in vaginal secretions of HIV-exposed uninfected sex workers react with the CD4-binding domain of gp120. IgA is considered to be more efficient than IgG in immunity against pathogens at the mucosal surface. This was emphasized by a recent study where class-switching of IgG monoclonal antibody 2F5 to its corresponding IgA isotype gave it the capacity to interfere with HIV-1 entry across a mucosal epithelial layer in vitro .
Further studies are necessary to evaluate the biological importance of IgA antibodies directed to the CD4-binding domain and to the Kennedy epitope. For example, it will be interesting to examine the capacity of IgA purified by affinity on agarose-immobilized MBP-constructs to interfere with the transport of HIV-1 through mucosal epithelial cells grown as a monolayer. It also could be useful to develop immortalized mucosal B cell lines secreting neutralizing anti-HIV IgA directed to the CD4-binding domain for immunoprophylaxis experiments.
We would like to thank all the blood and saliva donors for making this study possible. We also thank Dr C. Defontaine, Dr A. Frésard and A. M. Lantner for supply of human samples from patients. We thank the MRC AIDS Reagent Project for envelope genes of HIV-1 isolates, synthetic peptides and virus ARP256, and the NIH AIDS Research and Reference Reagent Program for peptides and recombinant sCD4.
Sponsorship: Financial support was provided by SIDACTION, ANRS and EUROVAC.
Note: Nadine Vincent and Etienne Malvoisin contributed equally to this work and are co-first authors.
1. Burnett PR, VanCott TC, Polonis VR, Redfield RR, Birx DL. Serum IgA-mediated neutralization of HIV type 1. J Immunol
2. Kozlowski PA, Chen D, Eldridge JH, Jackson S. Contrasting IgA and IgG neutralization capacities and responses to HIV type 1 gp120 V3 loop in HIV-infected individuals. AIDS Res Hum Retroviruses
3. Moja P, Tranchat C, Tchou I, Pozzetto B, Lucht F, Desgranges C, et al. Parotid IgA of HIV-1-infected patients mediated neutralization of HIV-1. J Infect Dis
4. Matsuda S, Noda M. Detection of IgA-binding sites on human immunodeficiency virus type 1 envelope glycoproteins, gp120 and gp41. Microbiol Immunol
5. Mazzoli S, Trabattoni D, Lo Caputo S, Piconi S, Ble C, Meacci F, et al. HIV-specific mucosal and cellular immunity in HIV-seronegative partners of HIV-seropositive individuals. Nat Med
6. Kaul R, Trabattoni D, Bwayo JJ, Arienti D, Zagliani A, Mwangi FM, et al. HIV-1-specific mucosal IgA in a cohort of HIV-1-resistant Kenyan sex worker. AIDS
7. Kozlowski PA, Jackson S. Serum IgA subclasses and molecular forms in HIV infection: selective increases in monomer and apparent restriction of the antibody response to IgA1 antibodies mainly directed at env glycoproteins. AIDS Res Hum Retroviruses
8. Cartry O, Moja P, Quesnel A, Pozzetto B, Lucht F, Genin C. Quantification of IgA and IgG and specifities of antibodies to viral proteins in parotid saliva at different stages of HIV-1 infection. Clin Exp Immunol
9. Olshevsky U, Helseth E, Furman C, Li J, Haseltine W, Sodroski J. Identification of individual human immunodeficiency virus type 1 gp120 amino acids important for CD4 receptor binding. J Virol
10. Thali M, Moore JP, Olshevsky U, Furman C, Gabuzda D, Posner M, et al. Characterization of discontinuous human immunodeficiency virus type 1 gp120 epitope recognized by a broadly reactive neutralizing human monoclonal antibody. J Virol
11. Thali M, Moore JP, Furman C, Charles M, Ho DD, Robinson J, et al. Characterization of conserved human immunodeficiency virus type 1 gp120 neutralization epitopes exposed upon gp120–CD4 binding. J Virol
12. Schutten M, McNight A, Huisman RC, Thali M, McKeating JA, Moore JP, et al. Further characterization of an antigenic site of HIV-1 gp120 recognized by virus neutralizing human monoclonal antibodies. AIDS
13. Kennedy RC, Henkel RD, Pauletti D, Allan JS, Lee TH, Essex M, et al. Antiserum to a synthetic peptide recognizes the HTLV-III envelope glycoprotein. Science
14. Brolinden P-A, von Gegerfelt A, Clapham P, Rosen J, Fenyö E-M, Wahren B, et al. Identification of human neutralizing-inducing regions of the human immunodeficiency virus type 1 envelope glycoproteins. Proc Natl Acad Sci USA
15. Gaspari MM, Brennan PT, Solomon SM, Elson CO. A method of obtaining, processing and analyzing human intestinal secretions for antibody content. J Immunol Meth
16. Weiblen BJ, Schumacher RT, Hoff R. Detection of IgM and IgA HIV antibodies after removal of IgG with recombinant protein G. J Immunol Meth
17. Gao F, Morrison SG, Robertson DL, Thorton CL, Craig S, Karlson G, et al. Molecular cloning and analysis of functional envelope genes from human immunodeficiency virus type 1 sequence subtypes A through G. J Virol
18. Vella C, Smith MH, Farrar GH, Jones DH, Daniels RS. A molecular and serologic study of the envelope gene of the British isolate: HIV-1 GB8.
Vaccine, 1995, 13
19. Moore JP. Simple methods for monitoring HIV-1 and HIV-2 gp120 binding to soluble CD4 by enzyme-linked immunosorbent assay: HIV-2 has a 25–fold lower affinity than HIV-1 for soluble CD4. AIDS
20. Skott P, Lucht E, Juler I, Dillner J, Björling E. Salivary sIgA response in HIV-1 infection. J Acquir Immune Defic Syndr
21. Basmaciogullari S, Babcock GJ, van Ryk D, Wojtowicz W, Sodroski J. Identification of conserved and variable structures in the human immunodeficiency virus gp120 glycoprotein of importance for CXCR4 binding. J Virol
22. McLain L, Brown JL, Cheung L, Reading SA, Parry C, Jones TD, et al. Different effects of a single amino acid substitution on three adjacent epitopes in the gp41 C-terminal tail of a neutralizing antibody escape mutant of human imunodeficiency virus type 1. Arch Virol
23. Buratti E, McLain L, Tisminetzki S, Cleveland SM, Dimmock NJ, Baralle FE. The neutralizing antibody response against a conserved region of human immunodeficiency virus type 1 gp41 (amino acid residues 731–752) is uniquely directed against a conformational epitope. J Gen Virol
24. Newton SMC, Joys TM, Anderson SA, Kennedy RC, Hovi ME, Stocker BAD. Expression and immunogenicity of an 18-residue epitope of HIV1 gp41 inserted in the flagellar protein of a Salmonella live vaccine. Res Microbiol
25. Evans DJ, McKeating J, Meredith JM, Burke KL, Katrak K, John A, et al. An engineered poliovirus chimaera elicits broadly reactive HIV-1 neutralizing antibodies. Nature
26. Mascola, JR Stiegler, G VanCott, TC Katinger, H Carpenter, CB Hanson CE, et al. Protection of macaques against vaginal transmission of a pathogenic HIV-1/SIV chimeric virus by passive infusion of neutralizing antibodies. Nat Med
27. Parren PWHI, Marx PA, Hessel AJ, Luckay A, Harouse J, Cheng-Mayer C, et al. Antibody protect macaques against vaginal challenge with a pathogenic R5 simian/human immunodeficiency virus at serum levels giving complete neutralization in vitro. J Virol
28. Wolbank S, Kunert R, Stiegler G, Katinger H. Characterization of human class-switched polymeric (immunoglobulin M [IgM] and IgA) anti-human immunodeficiency virus type 1 antibodies 2F5 and 2G12. J Virol