The broad antibacterial activity of the natural antibody repertoire is due to polyreactive antibodies

Abstract

Polyreactive antibodies bind to a variety of structurally unrelated antigens. The function of these antibodies, however, has remained an enigma, and because of their low binding affinity their biological relevance has been questioned. Using a panel of monoclonal polyreactive antibodies, we showed that these antibodies can bind to both Gram-negative and Gram-positive bacteria and acting through the classical complement pathway can inhibit bacterial growth by lysis, generate anaphylatoxin C5a, enhance phagocytosis, and neutralize the functional activity of endotoxin. Polyreactive antibody-enriched, but not polyreactive antibody-reduced, IgM prepared from normal human serum displays antibacterial activity similar to that of monoclonal polyreactive IgM. We conclude that polyreactive antibodies are a major contributor to the broad antibacterial activity of the natural antibody repertoire.

Figure 1

Properties of polyreactivity antibody 2E4. (A) Dose-dependent binding of 2E4 to different antigens as measured by ELISA and the calculated dissociation constants (Kd) . (B) Polyreactive antibody 2E4 binds to a variety of Gram-positive and Gram-negative bacteria. (C) Non-polyreactive MAb2507 only binds to its cognate antigen, E. coli O157: H7. (D) Polyreactive antibody 2E4 (right panel), but not non-polyreactive MAb2507 (left panel), fixes complement as measured by FACS analysis with anti-C3. Abbreviation: So, S. oralis; Sm, S. mitis; Sg, S. gordonii; An, A. naeslundii; Ec, E. coli.

Figure 2

Polyreactive 2E4 is bactericidal in the presence of complement. Growth of 2E4-treated E. coli BL21 as measured by the incorporation of ³H-TdR in the presence of different concentrations of (A) untreated and (B) heat-inactivated complement. (C) Bacteria colony formation on agar plates in the presence of complement with or without 2E4. (D) Lysis of 2E4-treated bacteria as measured by the release of ³H-TdR in the presence of different concentrations of complement. (E) Growth of 2E4-treated bacteria as determined by the incorporation of ³H-TdR in the presence of different concentrations of C4^+/+ or C4^−/− serum or (F) heat-inactivated C4^+/+ or C4^−/− serum. (G) Lysis of 2E4-treated E. coli BL21 as measured by the release of H³-TdR in the presence of different concentrations of C4^+/+ or C4^−/− serum. Bars represent s.d.

Figure 3

Lysis of Pseudomonas aeruginosa (PSA-10) by MAb2E4 and three additional polyreactive antibodies that inhibit the growth of E. coli BL21 in the presence of complement. (A) MAb 2E4 (right panel), but not MAb8512 (left panel), binds to PSA-10. (B) Enhanced lysis of 2E4-treated PSA-10 as compared to MAb8512-treated PSA-10 in the presence of two different concentrations of complement as measured by the release of ³H-adenine. (C) Polyreactive MAbs ZH-6, ZH-14 and ZH-20, but not non-polyreactive MAb2507, bind to E. coli BL21. (D) Inhibition of growth of antibody-treated E. coli BL21 as measured by incorporation of ³H-TdR in the presence of complement (left) or heat-inactivated complement (right). (E) Lysis of antibody-treated E. coli BL21 as measured by the release of ³H-TdR in the presence of complement.

Figure 4

Polyreactive antibodies in the presence of complement do not lyse Gram-positive bacteria (i.e., Streptococcus mitis, Streptococcus oralis) but generate the anaphylatoxin, C5a. (A) Polyreactive antibody 2E4 fixes complement as measured by FACS analysis with anti-C3 antibody (red line) as compared to non-binding MAb8512 (green shadow). (B) Neither polyreactive 2E4-treated nor non-polyreactive MAb8512-treated Gram-positive bacteria are lysed in the prescence of complement (1:10 dilution) as determined by ³H-TdR release. (C) Monoclonal antibody treated Gram-positive bacteria were incubated with guinea pig complement (1:10 dilution) followed by incubation with human C5. MAb 2E4-treated, but not MAb8512-treated, Gram-positive bacteria generated anaphylatoxin C5a as measured by ELISA.

Figure 5

Polyreactive antibodies enhance phagocytosis and inhibit the activity of LPS. (A) Immunofluorescence staining shows bacteria E. coli BL21 (green) being taken up by macrophages (red) with blue nuclei. Left panel: bacteria treated with MAb2507 and complement. Right panel: bacteria treated with MAb2E4 and complement. (B) Phagocytosis of 2E4-treated bacteria in the prescence of complement as determined by FACS analysis and measured by mean fluoresce intensity (MFI). (C) Human monocyte MMP-1 is increased by incubation with LPS (Salmonella abortus-equi, SAE) in the presence of PBS or non-binding MAb8512, but is decreased by polyreactive 2E4. (D) MMP-1 is not produced by monocytes incubated with PBS or with different polyreactive (i.e., ZH-14, ZH-6 and ZH-20) or non-polyreactive MAbs (i.e., 8512 and 2507) in the absence of LPS. (E-G) Monocyte MMP-1 production is increased by incubation with LPS from (E) SAE, (f) E. coli O157: H7 or (G) Pseudomonas aeruginosa (PSA) in the presence of PBS or non-binding MAb8512, but is decreased by polyreactive MAbs ZH-14, ZH-6 and ZH-20. Non-polyreactive MAb2507 decreased only the production of MMP-1 induced by LPS from E. coli O157: H7, its cognate antigen (F). (H) Polyreactive-enriched, but not polyreactive-reduced, human IgM inhibits LPS-induced production of MMP-1.

Figure 6

Bactericidal activity of polyreactive antibody-enriched human IgM. (A) Binding of polyreactive antibody-enriched, as compared to polyreactive antibody-reduced, IgM to different antigens and to (B) E. coli BL21. (C) Polyreactive-enriched, but not polyreactive-reduced, IgM lyses E. coli BL21 in the presence of complement as measured by the release of ³H-TdR. Equal concentrations (50 μg/ml) of polyreactive-enriched and polyreactive-reduced IgM were used in all experiments.