SGM Journals
HOST RESPONSE

Antibodies raised in animals against the Streptococcus agalactiae proteins cα and R4 and normal human serum antibodies target distinct epitopes

Sylvester R. Moyo, Johan A. Maeland

  • 1Department of Medical Microbiology, Faculty of Medicine, University of Zimbabwe Medical School, PO Box A178, Avondale, Harare, Zimbabwe 2Department of Laboratory Medicine, Children's and Women's Health, Norwegian University of Science and Technology, N-7006 Trondheim, Norway
  • Correspondence Johan A. Maeland Johan.Melandmedisin.ntnu.no

    • Journal of Medical Microbiology 2003; 52(5):379–383 · https://doi.org/10.1099/jmm.0.05087-0

    View at publisher PubMed

    Abstract

    The targets for normal human serum antibodies that react with proteins cα and R4 isolated from group B streptococci (GBS; Streptococcus agalactiae) have been studied and compared with the targets for murine monoclonal and rabbit polyclonal antibodies raised against these proteins. The proteins were extracted by trypsin digestion and purified by precipitations and gel filtration and testing was based on enzyme immunoassays. The immune antibodies showed specificity for the corresponding protein, targeted that protein in Western blotting and recognized their targets after heat treatment (100 °C) of the proteins. Human antibodies in a commercial gammaglobulin preparation targeted a site(s) common to cα and R4. This target failed to bind the antibodies in Western blotting and was destroyed by heating. cα- and R4-reactive antibodies in sera from healthy pregnant women recognized the common, heat-labile determinant(s), but contained little or no antibodies against the heat-stable cα- or R4-specific determinants. These results are consistent with the notions that (i) the normal human antibodies and the immunization-induced animal antibodies targeted different sites on the cα and R4 proteins and that (ii) the natural human antibodies targeted conformational epitopes and the immune antibodies targeted linear epitopes. These findings are important for further clarification of GBS immunology and immunoprotection in humans.

    • Abbreviation: GBS, group B streptococcus.

    INTRODUCTION

    Group B streptococci (GBS; Streptococcus agalactiae) are important pathogens in humans, particularly in neonates, who frequently become colonized during delivery as a result of urogenital GBS carriage in the mother. Studies from different parts of the world have demonstrated variable GBS carriage rates in pregnant women of up to more than 30 % (Moyo et al., 2000). Among factors that determine the outcome when neonates become colonized by GBS (Schuchat, 1998), the level of protective serum antibodies transferred from the mother to the fetus seems to be important, notably antibodies against the capsular polysaccharide antigens (Baker & Kasper, 1976; Baker et al., 1980). Nine different GBS capsular antigen types are known and have been designated Ia, Ib and II–VIII. In addition to the type-specific anti-capsular antibodies, antibodies against surface-localized and strain-variable protein antigens may be important in immunoprotection according to results obtained in animal models and opsonophagocytic assays (Lachenauer & Madoff, 1996; Larsson et al., 1996; Madoff et al., 1991; Stålhammar-Carlemalm et al., 1993). These antigens include the cα and cβ proteins of the c protein fraction of GBS (Bevanger & Maeland, 1979). cα and a number of other proteins belong to a family of ladder-forming GBS proteins (Wästfelt et al., 1996), so designated on the basis of the banding patterns observed in immunoblotting. This group of proteins includes the R proteins R1–R4 (Flores & Ferrieri, 1996; Lancefield & Perlmann, 1952) and protein Rib (Stålhammar-Carlemalm et al., 1993), which is probably identical to R4 (Bevanger et al., 1995; Smith et al., 2002). Sequence analysis of some of the proteins has demonstrated similarity between them: up to 100 % identity for some stretches of the proteins (Lachenauer et al., 2000; Wästfelt et al., 1996).

    Since antibodies raised in animals against ladder-forming GBS proteins are protective in animal models (Lachenauer & Madoff, 1996; Larsson et al., 1996; Madoff et al., 1991; Stålhammar-Carlemalm et al., 1993), we speculated that corresponding human serum antibodies might be important in defence against GBS disease in humans. For that reason, we recently measured antibodies against the cα and R4 proteins of GBS (Moyo et al., 2001), the ladder-forming proteins that occur at the highest frequency (Kvam et al., 1995). We found that 70 % or more of pregnant women tested from Norway and Zimbabwe had appreciable levels of antibodies against these proteins, with highest levels in the group of Norwegian women (Moyo et al., 2001). Our previous data also suggested that the specificity of the cα- and R4-reactive human antibodies might be different from that of immune antibodies raised in animals against these proteins (Moyo et al., 2001). The present study describes the results of further work on the two categories of antibodies, notably their specificity, since this might be important in the context of functional capability of the antibodies and means to discriminate between the natural and immune antibodies in serological testing.

    METHODS

    Human sera.

    Sera from 30 pregnant Norwegian women and 30 pregnant Zimbabwean women were selected arbitrarily from larger collections of sera from pregnant women described in a previous report (Moyo et al., 2001). The sera were collected during antenatal check-up visits from women without known GBS or any other infectious disease. Only sera that tested negative for hepatitis B virus and human immunodeficiency virus were used (Moyo et al., 2001). A commercial preparation of human gammaglobulin (165 mg ml−1; Pharmacia) was used in a number of experiments.

    Bacterial strains.

    Strains 335 (Ia/cα) and 65604 (III/R4), our prototype strains for the cα and R4 proteins, were respectively used for preparation of the cα and R4 antigens. The bacteria were preserved at 4 °C on sealed blood-agar plates and cultured at 37 °C on blood-agar plates or in Todd–Hewitt broth as described previously (Bevanger & Maeland, 1979). Harvested bacteria were washed with PBS (pH 7.2).

    Antigens.

    The procedures for extraction and purification of the cα and R4 proteins synthesized by the GBS strains 335 and 65604 have been described previously (Moyo et al., 2001). Briefly, the proteins, both of which are resistant to degradation by trypsin (Bevanger et al., 1992, 1995), were extracted by trypsin digestion of whole bacterial cells, precipitated with 5 % (w/v) trichloroacetic acid and then with ammonium sulphate (72 % saturation). The final precipitate was chromatographed by gel filtration on a Sephacryl S-200 HR column (Pharmacia), which was equilibrated and eluted with PBS containing 0.02 % sodium azide. The proteins were eluted at a position corresponding to the void volume fractions. The fractions with peak activity as antigen in ELISA were combined and kept at −20 °C. The proteins appeared immunologically homogeneous, as evidenced by immunoblotting and other antibody-based tests (Moyo et al., 2001).

    Antisera.

    The preparation and properties of rabbit antisera against whole cells of the GBS strains 335 and 65604 and of antisera against purified cα and R4 have been described previously (Bevanger & Maeland, 1977; Bevanger et al., 1992, 1995). Briefly, rabbits were given several intravenous injections with formalin-killed bacteria over a period of 3 weeks (Bevanger & Maeland, 1979) or were immunized intradermally with the cα or R4 protein, which had been purified by immunosorbent chromatography (Bevanger et al., 1992, 1995). Animals were bled 1–2 weeks after the last injection. For the generation of cα-specific mAb, BALB/c mice were immunized with killed A909 (Ia/cα/cβ) bacteria (Bevanger et al., 1992) and, for anti-R4 mAb, with immunosorbent-purified R4 protein (Bevanger et al., 1995). Clones that synthesized cα- or R4-specific monoclonal IgG were used for production of the ascites used in this study.

    Indirect ELISA.

    The test was performed as described previously (Moyo et al., 1999, 2001). Briefly, cα or R4 protein recovered from the gel-filtration column was used to coat microtitre plates (Sero-Wel; Bibby Sterilin) after determining optimal antigen dilution for coating by checkerboard titration (Moyo et al., 1999). The coating was preformed at 4 °C for 20 h and incubation with antibodies at 20 °C for 1 h. Alkaline phosphatase-conjugated anti-immunoglobulins (Sigma) were used and p-nitrophenyl phosphate was used as substrate. PBS containing 0.05 % (v/v) Tween 20 (PBST) was used as diluent and for washings. A405 was recorded. When twofold dilutions of sera were tested, the ELISA titre was defined as the reciprocal of the highest serum dilution that showed an A405 of at least 0.200 above the background signal, which was determined by testing serum in wells that contained no antigen. For measuring antibody levels in sera from pregnant women, sera were tested at a dilution of 1 : 200 (Moyo et al., 2001). Antibody levels were expressed as ELISA ratios obtained by dividing the mean A405 recorded for individual human sera tested in duplicates by that recorded for the gammaglobulin preparation tested in duplicate at a dilution of 1 : 2000 on the same microtitre plate.

    ELISA inhibition.

    The test was performed in order to examine whether antigens in a bacterial extract would neutralize the anti-cα and/or anti-R4 antibodies. Undiluted or diluted bacterial extract, or PBST in the positive control, was mixed with an equal volume of antiserum or the gammaglobulin preparation, which had been diluted so as to give an A405 of approximately 1.000 for the positive control. The mixture was incubated at 20 °C for 1 h and then tested in the indirect ELISA as described above. The antibody-neutralizing capacity of an extract is expressed as percentage reduction caused by the extract of the A405 recorded for the uninhibited positive control.

    Western blotting.

    Western blotting was performed according to Laemmli (1970) and a previous report from our laboratory (Moyo et al., 1999). Briefly, undiluted material from the gel-filtration column which contained the cα or R4 protein was subjected to SDS-PAGE, transferred to PVDF membranes (Bio-Rad) and probed with gammaglobulin (1 : 2000), individual human sera (1 : 200), polyclonal rabbit antibodies (pAb; 1 : 500) or mAb (1 : 1000). Probing with the appropriate horseradish peroxidase-conjugated anti-Ig and colour development were performed as described previously (Moyo et al., 1999).

    RESULTS AND DISCUSSION

    Both the murine monoclonal and rabbit polyclonal anti-cα and anti-R4 antibodies showed strong reactivity in ELISA against the homologous protein, but no or very weak binding activity against the heterologous protein (Table 1). Normal rabbit serum showed no reactivity against the cα and R4 proteins (not shown). In contrast to the immune antibodies, antibodies in a human gammaglobulin preparation recognized both antigens (Table 1).

    Table 1. ELISA titres of cα- and R4-reactive antibodies in rabbit antisera, murine mAbs in ascitic fluid and a human gammaglobulin preparation

    In order to compare further the human and immune antibodies, ELISA inhibition was performed. The R4 protein strongly inhibited the rabbit anti-R4 antibodies, whereas cα did not, confirming the R4-specificity of the pAb (Fig. 1). By contrast, both cα and R4 inhibited the human antibodies that recognized the R4 protein, although not completely within the range of concentrations of the inhibiting proteins. Fig. 1 shows the results of a representative experiment, which was confirmed on repetition. Concordant results were obtained in analogous inhibition tests using anti-cα antibodies and solid-phase-bound cα, i.e. neutralization of the anti-cα immune antibodies only by cα and neutralization of the human antibodies recognizing cα by both cα and R4 (not shown). Taken together, these results support the notion that the immune antibodies recognized a protein-specific epitope(s), whereas the normal human serum antibodies recognized a different epitope(s) that is shared by cα and R4.

    Figure image not available in archive

    Fig. 1. ELISA competition using R4 protein-coated microtitre plates and rabbit anti-R4 antibodies inhibited by R4 (○) or cα (•) extract or human gammaglobulin inhibited by the R4 (□) or cα (▪) extract.

    In Western blotting, we observed no antibody binding when the gammaglobulin preparation or individual human sera were probed against cα and R4. In these tests, all of the sera examined were used at dilutions that showed an A405 of > 1.000 in ELISA. The antibodies raised in animals resulted in ladder-like banding patterns, which are known features of these proteins (Bevanger et al., 1992, 1995; Madoff et al., 1991; data not shown). Other investigators have used Western blotting in order to measure human serum antibodies against R4, but with electrophoresis performed in non-denaturing polyacrylamide minigels (Fasola et al., 1996), which may explain the discrepancy. To test further the possibility that the failure to detect binding of the human serum antibodies by cα and R4 in Western blotting was due to denaturation of the proteins in hot SDS, we tested whether heating itself would affect antibody binding. As seen from Fig. 2, heating of cα in PBS at 100 °C for 15 min nearly eliminated its ability to bind the human antibodies but not its ability to bind the murine or rabbit immune antibodies, not even when the heating was prolonged for up to 60 min. Analogous results were obtained when the R4 protein was exposed to heating and then tested for binding of the corresponding immune or natural human antibodies. The antigen preparations were exposed to heating for 30 min in a step-wise fashion in the range 40–100 °C. Binding of the human antibodies was unaffected by heating of cα and R4 up to 60 °C but, on the basis of A405 recordings, was reduced by 25 % by heating at 70 °C and by 60 % at 80 °C and was maximally affected at temperatures ⩾ 90 °C. These results were interpreted to mean that the cα and R4 epitopes that were recognized by the human antibodies were destroyed by the denaturing effect of the heating of the proteins, whereas the epitopes targeted by the immune antibodies resisted the denaturation. The results further support the above notion that the natural human and immune antibodies target different epitopes. Similar discrepancies between natural and induced antibodies have been described before (Guilbert et al., 1982).

    Figure image not available in archive

    Fig. 2. ELISA results (A405) for rabbit anti-cα antibodies (open columns), monoclonal anti-cα antibodies (hatched columns) and a human gammaglobulin preparation (filled columns) probed against microtitre plates coated with cα that had been untreated or heat-treated (100 °C) for the times indicated.

    Experiments were performed to test the extent to which heat-denaturation of cα and R4 would affect antibody binding by the proteins when individual human sera were tested, altogether using 60 sera from pregnant women from Norway and Zimbabwe. Individuals showed great variation of antibody levels (Table 2; Fig. 3), with generally higher levels in the Norwegian population than in the Zimbabwean population, as observed earlier (Moyo et al., 2001). Compared with the levels of signal in tests with the untreated antigens, the levels were reduced significantly with the heat-treated antigens, down to background levels for the majority of sera. The results indicate that the major component of the natural cα- and R4-reactive human antibodies was directed against the heat-sensitive determinant(s) common to cα and R4.

    Figure image not available in archive

    Fig. 3. Distribution of ELISA ratios for antibodies in sera from 60 pregnant women, 30 from Norway and 30 from Zimbabwe, tested against R4 protein that had been untreated (open columns) or heat-treated (100 °C, 30 min) (hatched columns).

    Table 2. Binding of antibodies in sera from healthy pregnant women by untreated and heat-treated cα and R4 proteins Proteins were heat-treated at 100 °C. Binding was measured as the ELISA ratio as described in Methods. In all cases, the difference between antibody levels with untreated and heat-treated antigen was significant (P < 0.05; Mann–Whitney U test).

    This study has shown that immune anti-cα and anti-R4 antibodies recognize protein-specific epitopes that resist denaturation by SDS and/or heating, consistent with targets determined by amino acid sequences. Natural human cα- and R4-reactive antibodies recognize targets that are shared by the two proteins and are destroyed by the denaturing procedures, consistent with binding sites determined by conformation. It has been shown that conformation also plays an important role in the binding by cα of immune antibodies, in this case conformational effects of variations in the number of cα repeat units (Madoff et al., 1996). In the present study, we can exclude this type of protein structure variation, as all the final experiments were performed using the same cα and R4 protein preparations. Rather, our results focus on the route of immunization as an important factor, parenteral for the immune antibodies and, presumably, mucosal presentation for the normal human antibodies. Mucosal immunization with GBS induced immune responses in both mice (Hordnes et al., 1995) and humans (Hordnes et al., 1996), with a major contribution by broadly cross-reacting antibodies (Hordnes et al., 1996), which could have included cα-/R4-reactive antibodies corresponding to the normal human antibodies measured in the present study. It will be important to test whether humans immunized during the course of invasive GBS disease or by parenteral vaccination (Paoletti et al., 2000) generate antibodies against the heat-stable or heat-labile epitopes. Our observation that the putative confirmation-determined epitope was heat-labile should facilitate the design of tests to discriminate between the two categories of antibodies.

    Both cα and R4 belong to a GBS protein family that includes members with large, tandemly arranged repeats (Michel et al., 1992; Wästfelt et al., 1996) and with considerable sequence similarity between different proteins (Lachenauer et al., 2000; Michel et al., 1992; Wästfelt et al., 1996). This similarity may have resulted in conformational structures in different proteins that can be recognized by the same natural human serum antibodies as a result of antigenic mimicry. Findings described previously (Moyo et al., 2001) as well as the data shown in Table 1 argue against a contaminating non-cα/R4 antigen shared by the GBS strains 335 and 65604 as a target for the normal human serum antibodies.

    Although the results described in the present study and in a previous study should broaden our knowledge of GBS immunology, they leave several challenges for further work, including investigations on the levels and specificity of cα-/R4-reactive antibodies in mucosal secretions, factors that determine the generation of cross-reacting versus protein-specific antibodies and the biological functions of the cross-reacting versus the protein-specific antibodies.

    Acknowledgments

    We are grateful to Randi V. Lyng for technical assistance, Professor Lars Bevanger for fruitful discussions and James Mudzori for providing human sera used in this study. We appreciate the support from the University of Zimbabwe Research Board and the Norwegian Quota programme for students from Developing Countries and Central and Eastern Europe.

    References

    1. Baker, C. J. & Kasper, D. L. (1976). Correlation of maternal antibody deficiency with susceptibility to neonatal group B streptococcal infection. N Engl J Med 294, 753–756.
    2. Baker, C. J., Webb, B. J., Kasper, D. L., Yow, M. D. & Beachler, C. W. (1980). The natural history of group B streptococcal colonization in the pregnant woman and her offspring.II. Determination of serum antibody to capsular polysaccharide from type III, group B Streptococcus. Am J Obstet Gynecol 137, 39–42.
    3. Bevanger, L. & Maeland, J. A. (1977). Type classification of group B streptococci by the fluorescent antibody test. Acta Pathol Microbiol Scand B 85, 357–362.
    4. Bevanger, L. & Maeland, J. A. (1979). Complete and incomplete Ibc protein fraction in group B streptococci. Acta Pathol Microbiol Immunol Scand B 87, 51–54.
    5. Bevanger, L., Iversen, O.-J. & Naess, A. I. (1992). Characterization of the α-antigen of the c proteins of group B streptococci (GBS) using a murine monoclonal antibody. APMIS 100, 57–62.
    6. Bevanger, L., Kvam, A. I. & Maeland, J. A. (1995). A Streptococcus agalactiae R protein analysed by polyclonal and monoclonal antibodies. APMIS 103, 731–736.
    7. Fasola, E. L., Flores, A. E. & Ferrieri, P. (1996). Immune response to the R4 protein antigen of group B streptococci and its relationship to other streptococcal R4 proteins. Clin Diagn Lab Immunol 3, 321–325.
    8. Flores, A. E. & Ferrieri, P. (1996). Molecular diversity among the trypsin resistant surface proteins of group B streptococci. Zentbl Bakteriol 285, 44–51.
    9. Guilbert, B., Dighiero, G. & Avrameas, S. (1982). Naturally occurring antibodies against nine common antigens in human sera.I. Detection, isolation and characterization. J Immunol 128, 2779–2787.
    10. Hordnes, K., Digranes, A., Haugen, I. L., Helland, D. E., Ulstein, M., Jonsson, R. & Haneberg, B. (1995). Systemic and mucosal antibody responses to group B streptococci following immunization of the colonic-rectal mucosa. J Reprod Immunol 28, 247–262.
    11. Hordnes, K., Tynning, T., Kvam, A. I., Jonsson, R. & Haneberg, B. (1996). Colonization in the rectum and uterine cervix with group B streptococci may induce specific antibody responses in cervical secretions of pregnant women. Infect Immun 64, 1643–1652.
    12. Kvam, A. I., Efstratiou, A., Bevanger, L., Cookson, B. D., Marticorena, I. F., George, R. C. & Maeland, J. A. (1995). Distribution of serovariants of group B streptococci in isolates from England and Norway. J Med Microbiol 42, 246–250.
    13. Lachenauer, C. S. & Madoff, L. C. (1996). A protective surface protein from type V group B streptococci shares N-terminal sequence homology with the alpha C protein. Infect Immun 64, 4255–4260.
    14. Lachenauer, C. S., Creti, R., Michel, J. L. & Madoff, L. C. (2000). Mosaicism in the alpha-like protein genes of group B streptococci. Proc Natl Acad Sci U S A 97, 9630–9635.
    15. Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685.
    16. Lancefield, R. C. & Perlmann, G. E. (1952). Preparation and properties of a protein (R antigen) occurring in streptococci of group A, type 28 and in certain streptococci of other serological groups. J Exp Med 96, 83–97.
    17. Larsson, C., Stålhammar-Carlemalm, M. & Lindahl, G. (1996). Experimental vaccination against group B streptococcus, an encapsulated bacterium, with highly purified preparations of cell surface proteins Rib and α. Infect Immun 64, 3518–3523.
    18. Madoff, L. C., Michel, J. L. & Kasper, D. L. (1991). A monoclonal antibody identifies a protective C-protein alpha-antigen epitope in group B streptococci. Infect Immun 59, 204–210.
    19. Madoff, L. C., Michel, J. L., Gong, E. W., Kling, D. E. & Kasper, D. L. (1996). Group B streptococci escape host immunity by deletion of tandem repeat elements of the alpha C protein. Proc Natl Acad Sci U S A 93, 4131–4136.
    20. Michel, J. L., Madoff, L. C., Olson, K., Kling, D. E., Kasper, D. L. & Ausubel, F. M. (1992). Large, identical, tandem repeating units in the C protein alpha antigen gene, bca, of group B streptococci. Proc Natl Acad Sci U S A 89, 10060–10064.
    21. Moyo, S. R., Maeland, J. A. & Bevanger, L. (1999). Comparison of three different methods in monoclonal antibody-based detection of Streptococcus agalactiae protein serotype markers. APMIS 107, 263–269.
    22. Moyo, S. R., Mudzori, J., Tswana, S. A. & Maeland, J. A. (2000). Prevalence, capsular type distribution, anthropometric and obstetric factors of group B Streptococcus (Streptococcus agalactiae) colonization in pregnancy. Cent Afr J Med 46, 115–120.
    23. Moyo, S. R., Maeland, J. A. & Mudzori, J. (2001). Antibodies against Streptococcus agalactiae proteins cα and R4 in sera from pregnant women from Norway and Zimbabwe. Clin Diagn Lab Immunol 8, 1110–1114.
    24. Paoletti, L. C., Madoff, L. C. & Kasper, D. L. (2000). Surface structures of group B Streptococcus important in human immunity. In Gram-positive Pathogens, pp. 137–153. Edited by V. A. Fischetti, R. P. Novick, J. J. Ferretti, D. A. Portnoy & J. I. Rood. Washington, DC: American Society for Microbiology.
    25. Schuchat, A. (1998). Epidemiology of group B streptococcal disease in the United States: shifting paradigms. Clin Microbiol Rev 11, 497–513.
    26. Smith, B., Flores, A., Dechaine, J., Krepela, J., Bergdall, A. & Ferrieri, P. (2002). The group B streptococcal protein R4 gene, its presence in clinical reference laboratory strains, and analysis of R4 protein pepsin sensitivity. In Abstracts of the 15th Lancefield International Symposium on Streptococci and Streptococcal Diseases, Goa, India. Abstract P4.07.
    27. Stålhammar-Carlemalm, M., Stenberg, L. & Lindahl, G. (1993). Protein Rib: a novel group B streptococcal cell surface protein that confers protective immunity and is expressed by most strains causing invasive infections. J Exp Med 177, 1593–1603.
    28. Wästfelt, M., Stålhammar-Carlemalm, M., Delisse, A.-M., Cabezon, T. & Lindahl, G. (1996). Identification of a family of streptococcal surface proteins with extremely repetitive structure. J Biol Chem 271, 18892–18897.