Inhibition of complement-mediated opsonization and phagocytosis of Streptococcus pyogenes by D fragments of fibrinogen and fibrin bound to cell surface M protein.

The biological effects of the binding of fibrin(ogen) degradation products to M protein-bearing group A streptococci were investigated. Type 24 group A streptococci bind fibrinogen degradation products of the D family, but not fragment E. Binding appears to be mediated by M protein, since a large peptide of this molecule (pep M24) bound to fragments containing the terminal domains of the fibrinogen molecule (D, X, and Y), but not fragment E, and pep M24 inhibited the binding of digested fibrinogen to streptococcal cells. An M protein-binding site occurs on fragment D3 and, therefore, differs from several functional sites present on D1 but not D3, including the fibrin polymerization site, the two gamma chain crosslink sites, and the bindings sites for platelet fibrinogen receptor, staphylococcal clumping factor, and ionized calcium. Bound fibrinogen degradation products prevented deposition of C3 on the streptococcal cell surface, and, in consequence, prevented phagocytosis by neutrophils in nonimmune blood. The average affinity of D fragments for the streptococcal cell surface was approximately 30 times lower than that of native fibrinogen, and a terminal plasmic digest was approximately 50 times less potent in inhibiting opsonization by C3. However, physiologic concentrations of digested fibrinogen sufficed to inhibit opsonization and phagocytosis completely. Digests of crosslinked fibrin clot also inhibited opsonization, although slightly less effectively than did fibrinogen digests. The antiopsonic effect of fibrin(ogen) degradation products may be relevant to circumstances in which fibrin(ogen)olysis is occurring, e.g., exudation and suppuration.

Epithelial cell binding of group A streptococci by lipoteichoic acid on fimbriae denuded of M protein.

Group A streptococci were treated with various enzymatic and chemical agents in an attempt to dissociate the type-specific M protein from intact surface "fimbriae." Mild peptic digestion at pH 5.8, which was previously shown to extract serologically active M antigen from intact streptococci had little visible effect on the fimbriae even though virtually all of the M protein was removed as demonstrated by (a) increased susceptibility to phagocytosis, (b) lack of opsonic effect of homologous M antibody on the treated streptococci, and (c) loss of HCl-extractable M protein. These fimbriated streptococci which lacked M protein adhered to human oral mucosal cells equally as well as untreated, fimbriated organisms which retained their M protein. Removal of both fimbriae and M protein by digesting organisms with HCL at pH 2.0 at 94 degrees C. or with trypsin abolished their ability to bind mucosal cells. Electron microscopy of streptococci bound to epithelial cells demonstrated fimbriae radiating from the surface of the organisms to the membrane of the epithelial cells. It is apparent, therefore, that the determinants of streptococcal fimbriae involved in resistance to phagocytosis can be dissociated from those involved in epithelial cell binding. These results are consistent with our previous studies which suggested that fatty acids ester linked with glycerol teichoic acid rather than M protein of streptococci binds the organisms to epithelial cells.

Common protective antigens of group A streptococcal M proteins masked by fibrinogen.

The influence of fibrinogen on the opsonization of Group A streptococci by type-specific and cross-reactive anti-M protein antisera was investigated. As previously reported for type 24 streptococci, fibrinogen inhibited the complement-mediated opsonization of types 5, 6, and 19 organisms. Rabbit antisera against large peptide fragments of purified homologous M proteins (pep M proteins) overcame the anti-opsonic effect of fibrinogen in a dose-dependent manner. In the presence of optimal amounts of antibody, bacterial uptake by PMN was equal in serum and plasma, and greater than could be obtained in serum in the absence of antibody. Polyclonal anti-pep M sera contained antibodies directed against fibrinogen-binding as well as fibrinogen-nonbinding sites or regions of the M protein molecule. Three cross-reactive anti-pep M sera included antibodies directed against fibrinogen binding sites or regions of the cross-reacting M proteins. In the two sera studied in detail, these antibodies accounted for a large part of the cross-reacting anti-M antibody present in the sera. We suggest that fibrinogen binding sites on different serotypes of M protein may be structurally and therefore antigenically similar. Conservation of fibrinogen binding sites on M proteins may be related to their protective anti-opsonic function.

Epitopes of streptococcal M proteins shared with cardiac myosin.

We present evidence that M proteins from three different serotypes of group A streptococci share epitopes with cardiac myosin. Rabbit antisera evoked by a purified fragment of type 5 M protein crossreacted with myosin, but not alpha-tropomyosin, actin, or myosin light chains. In enzyme-linked immunosorbent assays, the myosin-crossreactive antibodies were totally inhibited by type 5 M protein and partially inhibited by types 6 and 19 M proteins. The affinity-purified myosin antibodies opsonized type 5 streptococci, indicating that they were directed against protective M protein epitopes on the surface of the organisms. Immunoblot analyses demonstrated the binding of the crossreactive antibodies to myosin heavy chains. Sera from patients with acute rheumatic fever showed significantly stronger reactions with myosin than did sera from their siblings, hospitalized controls, or patients with poststreptococcal glomerulonephritis.

Multiple, heart-cross-reactive epitopes of streptococcal M proteins.

We present evidence that a highly purified pepsin extract of type 5 streptococcal M protein (pep M5) contains at least three epitopes that are cross-reactive with sarcolemmal membrane proteins of human myocardium. The tissue-cross-reactive determinants of pep M5 are also partially shared with pep M6 and pep M19. Three rabbits immunized with a single 300 micrograms dose of pep M5 developed significant levels of heart-cross-reactive antibodies, as determined by indirect immunofluorescence tests. All three sera also contained antibodies that cross-reacted with pep M6 and pep M19. The heart tissue--specific antibodies that were eluted from sarcolemmal membranes opsonized types 5, 6, and 19 streptococci, indicating that they were directed against protective M protein epitopes on the surface of virulent organisms. Immunofluorescence inhibition tests, using purified M proteins as soluble inhibitors of heart-cross-reactive antibodies, revealed the number and M protein serotype distribution of the tissue-cross-reactive epitopes. Immunoblot analyses demonstrated the sarcolemmal membrane proteins containing the various cross-reactive antigenic determinants.

Human cytotoxic T lymphocytes evoked by group A streptococcal M proteins.

Purified group A streptococcal M proteins were used to stimulate peripheral blood lymphocytes from normal adult volunteers. The activated lymphocytes were cytotoxic against cultured human heart cells, as well as liver cells, fibroblasts, and K562 cells, but showed only minimal cytotoxicity against several animal cell types. The cytotoxic activity evoked by type 5 M protein was dose and time dependent. Rabbit antisera against pep M5 that contained heart-crossreactive antibodies partially inhibited cytotoxicity against heart cells, but had no effect on other target cells, suggesting that a fraction of the effector lymphocytes may be recognizing M protein-crossreactive cell surface antigens. All of the cytotoxic activity was recovered from a CD3+ population of lymphocytes obtained from a fluorescence-activated cell sorter, and CD4+ and CD8+ cells were also cytotoxic. M protein-responsive T cell clones were generated that showed specificity for heart and K562 cells, in addition to clones that were cytotoxic against both cell lines. Our data show that streptococcal M protein evokes cytotoxic T lymphocytes against multiple human but not animal target cells. Some of the effector cells may be specific for cultured myocardial cells, but their role in the pathogenesis of rheumatic carditis will require further studies of lymphocytes from patients with acute rheumatic fever and carditis.

Sequence of myosin-crossreactive epitopes of streptococcal M protein.

Group A streptococcal M proteins contain epitopes that crossreact with sarcolemmal membrane proteins of human myocardium and myosin. In the present study, synthetic peptide copies spanning the entire 197-residue pepsin extracted fragment of type 5 M protein were used to localize the myosin-crossreactive epitopes of the molecule. Peptide 84-116 inhibited by 75% the binding of myosin-crossreactive antibodies evoked by pep M5, as determined by ELISA. Immunoblot inhibition studies confirmed that peptide 84-116 almost totally inhibited the binding of pep M5 antibodies to the heavy chain of human cardiac myosin. None of the remaining synthetic peptides, including peptide 1-35, which contains protective epitopes, inhibited antibodies binding to myosin. Two of three rabbits immunized with peptide 84-116 developed low but significant levels of antibodies crossreactive with myosin. Identification of the primary structures containing tissue-crossreactive as opposed to protective epitopes should not only allow the development of safe and effective M protein vaccines, but may also provide insights into the pathogenesis of rheumatic heart disease.

Localization of protective epitopes of the amino terminus of type 5 streptococcal M protein.

We have used a set of overlapping chemically synthesized peptides representing the amino terminus of type 5 streptococcal M protein to localize protective, as opposed to nonprotective and tissue-crossreactive epitopes that might be appropriate for vaccine formulations. Rabbit antisera raised against SM5(1-35) reacted in high titer with pep M5 by ELISA and opsonized type 5 streptococci. None of the antisera crossreacted with human heart tissue or myosin. Antisera against SM5(26-35) reacted with SM5(1-35) and pep M5 but failed to opsonize type 5 streptococci. Particle-phase ELISA indicated that SM5(26-35) antibodies were directed against nonprotective determinants of pep M5 that were not exposed on the surface of viable organisms. Opsonization and ELISA inhibition assays showed that, of the SM5(1-35) antibodies that reacted with M5, all were inhibited by SM5(14-35), whereas none was inhibited by SM5(26-35), suggesting that the protective epitopes of SM5(1-35) resided between residues 14 and 26. This was confirmed by subsequent chemical synthesis of this region; SM5(14-26) totally inhibited SM5(1-35) antibodies that reacted with pep M5 in ELISA, and completely inhibited opsonization of type 5 streptococci by SM5(1-35) antibodies. SM5(14-26) evoked high titers of type-specific, opsonic antibodies against type 5 streptococci, confirming the protective immunogenicity of this 13-residue peptide of type 5 M protein.

Protective antigenic determinant of streptococcal M protein shared with sarcolemmal membrane protein of human heart.

We present definitive evidence that at least one protective antigenic determinant on type 5 M protein of group A streptococci evokes antibody that is cross-reactive with human heart tissue. One of nine rabbits immunized with a peptide fragment of type 5 M protein (pep M5) produced antibody that cross-reacted by immunofluorescence with sarcolemmal membranes of human heart. The cross-reactive antibody could be removed by absorbing the antiserum with sarcolemmal membranes, types 5 and 19 streptococci, or their pepsin-extracted M proteins, but with no other serotypes tested. Although each of the pep M5 immune sera was opsonic for type 5 streptococci, only the heart-reactive antiserum opsonized type 19 streptococci. The opsonization of type 19 streptococci was abolished by absorbing the antiserum with sarcolemmal membranes isolated from human heart tissue. Purified heart-reactive antibodies eluted from sarcolemmal membranes opsonized both types 5 and 19 streptococci, indicating that the heart cross-reactive determinant of type 5 M protein is cross-protective. The cross-reactive antigen was purified by affinity chromatography from detergent extracts of sarcolemmal membranes and determined to be a complex protein composed of four subunits apparently linked by disulfide bonds.

Heterogeneity of type-specific and cross-reactive antigenic determinants within a single M protein of group A streptococci.

The heterogeneity of a pepsin extract of type-24 M protein (pep M24) was demonstrated by absorption of type-specific and cross-reactive human antisera with M protein fragments and heterologous serotypes of M proteins, pepsin extract of type-5 M protein (pep M5) and pepsin extract of type-6 M protein (pep M6). 2 of 12 individuals immunized with pep M24 developed significant rises in antibody titers against pep M5 and pep M6, as measured by the enzyme-linked immunosorbent assay. The sam individuals also developed opsonic antibodies against type-6, but not type-5, streptococci, which suggested the development of cross-protective immunity. Inhibition studies of one of these sera with the heterologous pep M proteins showed that the cross-reactive antibodies against pep M6 could not be blocked by high concentrations of pep M24, the immunizing antigen; these antibodies could be blocked, however, by cyanogen bromide-derived peptide fragments of pep M24, which suggested that the cross-reactive antibody was raised against an inaccessible site(s) in the pep M24 molecule. Inhibition studies of type-specific immune sera with pep M24 and peptides derived therefrom indicated that the M protein molecule contained multiple distinct as well as identical type-specific antigenic determinants that are unequally distributed among the seven derived peptide fragments.

Toxic effects of streptococcal M protein on platelets and polymorphonuclear leukocytes in human blood.

Purified M protein isolated from Group A streptococci produced cytotoxic reactions in normal human blood in vitro. In the presence of M antigen, platelets aggregated, fused, and lysed. Polymorphonuclear leukocytes (PMN) surrounded the platelet aggregates, then became highly vacuolated and lysed. In addition, PMN progressively lost their capacity to phagocytose unrelated bacteria and to migrate in glass capillary pipettes. Platelet-PMN reactions were directly proportional to the type-specific precipitin reactivity of each M preparation and could be removed with homologous M antibody, only. Moreover, the reactivity of M protein was abolished by enzymatic digestion with trypsin, but not with lysozyme, strongly suggesting that cell-wall mucopeptide was not involved. Preliminary studies showed that platelet-PMN reactions require heat-stable and heat-labile serum factors, presumably antibody and complement. It is suggested that cytotoxic determinants are uncovered by the extraction and purification process and are intimately associated with the type-specific M determinant, possibly in a molecular complex.

Protective immunogenicity and T lymphocyte specificity of a trivalent hybrid peptide containing NH2-terminal sequences of types 5, 6, and 24 M proteins synthesized in tandem.

The protective immunogenicity of a hybrid peptide containing tandem copies of types 5, 6, and 24 M protein epitopes was investigated. An NH2-terminal peptide of type 24 M protein was chemically synthesized and then extended to include NH2-terminal peptides of types 6 and 5 M proteins yielding a 34-residue hybrid peptide containing a cysteine residue at its COOH-terminus. When conjugated via the cysteine residue to keyhole limpet hemocyanin (KLH), emulsified in CFA, and injected into rabbits, the synthetic hybrid evoked opsonic antibodies against types 5, 6, and 24 streptococci without stimulating tissue crossreactive immunity. The trivalent hybrid also was capable of priming T lymphocytes in vivo that responded to each of the native serotypes of M protein as well as to the synthetic hybrid peptide in vitro. The primed T cells failed to respond to the individual component peptides contained in the hybrid peptide, suggesting that the hybrid peptide confers conformations resembling the presentations of each of the subpeptides in the respective serotypes of M protein. The brisk immune responses to the type 6 peptide contained in the middle of the tandem hybrid indicates that with judicious placement between proline residues, potentially hidden peptides are readily accessible to the immune system. These results suggest that synthetic tandem peptides can be tailored in a fashion in which each of the component sets of protective epitopes can be made optimally immunoaccessible and immunogenic.

Protective immunity evoked by oral administration of attenuated aroA Salmonella typhimurium expressing cloned streptococcal M protein.

Attenuated strains of Salmonella have been used effectively as vaccines against typhoid fever. We have investigated the use of such strains to deliver cloned antiphagocytic virulence determinants of unrelated bacteria. The aroA strain of S. typhimurium SL3261 was transformed with a low-copy plasmid vector pMK207, which contains the cloned gene spm5 encoding streptococcal M protein, the major virulence factor of these organisms. The transformed SL3261 expressed type 5 M protein in the cytoplasmic fraction, and when fed orally to BALB/c mice, evoked both serum and salivary IgA, IgG, and IgM antibodies directed against type 5 M protein. The orally immunized mice were completely protected against both intranasal and intraperitoneal challenge infections with virulent S. typhimurium SL1344 or M5 streptococci. These studies provide evidence that an attenuated strain of Salmonella can be used effectively as a general vaccine vehicle to deliver antiphagocytic virulence determinants of unrelated bacteria.

Autoimmune sequence of streptococcal M protein shared with the intermediate filament protein, vimentin.

The crossreactivity of antibodies against a renal autoimmune epitope of Streptococcus pyogenes M protein with glomerular mesangial cells was investigated. The antibodies directed against the amino acid sequence Ile-Arg-Leu-Arg of the nephritogenic type 1 M protein reacted in a fibrillar pattern with mesangial cells cultured from isolated glomeruli. In Western blots of urea-extracted mesangial proteins, the antibodies reacted with a 56-kD protein. Monoclonal and polyclonal antibodies identified the 56-kD mesangial protein as vimentin. Two synthetic peptides of human vimentin containing the sequence Arg-Leu-Arg reacted with the autoimmune antibodies raised against a streptococcal M protein peptide. These results provide evidence that the intermediate filament protein vimentin shares autoimmune epitopes with streptococcal M protein.

Protective and heart-crossreactive epitopes located within the NH2 terminus of type 19 streptococcal M protein.

M protein was purified to homogeneity from limited pepsin digests of intact type 19 streptococci (pep M19). The purified pep M19 when emulsified in CFA and injected into rabbits evoked type-specific and crossreactive opsonic antibodies, as well as heart-crossreactive antibodies. The NH2-terminal primary structure of pep M19 was determined and a peptide copying the first 24 amino acids [SM19(1-24)C] was chemically synthesized. Rabbits that were immunized with the unconjugated peptide developed antibodies that recognized the native pep M19, as determined by ELISA, and opsonic antibodies against type 19 streptococci, as determined by in vitro opsonophagocytosis tests. The synthetic peptide also evoked antibodies that crossreacted with a 60-kD sarcolemmal membrane protein of human myocardium. By using overlapping synthetic subpeptides as immunoinhibitors, the opsonic and heart-crossreactive epitopes of SM19(1-24)C were localized to SM19(11-24)C. Our data confirm the presence of heart-crossreactive epitopes within the primary structure of pep M19 and show that these potentially harmful autoimmune epitopes may be located in the NH2-terminal regions of certain M proteins. We conclude that continued efforts to identify the primary structures of protective and heart-crossreactive epitopes will be necessary to elucidate the pathogenesis of acute rheumatic heart disease and to develop safe and effective streptococcal vaccines.

Type-specific immunogenicity of a chemically synthesized peptide fragment of type 5 streptococcal M protein.

We determined the antigenic specificity and protective immunogenicity of two chemically synthesized peptides of type 5 streptococcal M protein. The synthetic peptides, designated S-M5(1-20) and S-M5(20-40), represent the amino-terminal amino acid sequence of the native pepsin-extracted M5 molecule, which is known to contain at least one heart cross-reactive epitope. Initial studies showed that neither of the synthetic peptides was able to bind purified heart-reactive M5 antibodies. In addition, S-M5(1-20), but not S-M5(20-40), contained type-specific antigenic determinants as measured by enzyme-linked immunosorbent inhibition assays. When covalently linked to tetanus toxoid, S-M5(1-20), but not S-M5(20-40), evoked significant levels of type-specific, opsonic (and presumably protective) antibodies in rabbits without evoking heart cross-reactive antibodies.

Superantigenicity of streptococcal M protein.

M proteins that define the serotypes of group A streptococci are powerful blastogens for human T lymphocytes. The mechanism by which they activate T cells was investigated and compared with the conventional T cell mitogen phytohemagglutinin, and the known superantigen staphylococcal enterotoxin B. Although major histocompatibility complex (MHC) class II molecules are required for presentation, there is no MHC restriction, since allogeneic class II molecules presented the bacterial protein to human T cells. Type 5 M protein appears to bind class II molecules on the antigen-presenting cells and stimulate T cells bearing V beta 8 sequences. Our results indicate that this streptococcal M protein is a superantigen and suggest a possible mechanism of its role in the pathogenesis of the postinfectious autoimmune sequelae.

Cell membrane-binding properties of group A streptococcal lipoteichoic acid.

Lipoteichoic acid (LTA) was extracted from group A streptococci, previously treated with hot HCl, by the phenol method. The extracted LTA was loaded on an isoelectric (IE) focusing column and two fractions were collected; one at pH 4.65 and the other at pH 2.95. Chemical analysis demonstrated that the unfractionated LTA contained alanine and glycerolphosphate at molar ratio of 1:10, and ester-linked lipids, but no detectable sugars or amino-sugars. The two IE fractions contained lipids but lacked alanine. The LTA and its IE fractions spontaneously adsorbed to human erythrocytes (sensitization) causing them to agglutinate in the presence of rabbit anti-LTA. The RBC-sensitizing and antigenic activities of IE fractions were equal to, or greater (for IE fraction at pH 4.65) than the unfractionated LTA, indicating that alanine is not involved in the sensitizing activity of LTA. Mild ammonia-hydrolysis abolished the RBC-sensitizing activity of LTA and its IE fractions. Chloroform-methanol-soluble material of the ammonia-hydrolysate lacked antigenic activity but blocked sensitization of erythrocytes by LTA. The water-soluble material of the hydrolyzed LTA retained antigenic activity, was not able to block sensitization by LTA, and its sensitizing activity was restored after esterification with fatty acids. These experiments indicate that ester-linked fatty acids (palmitic acid being the major one) are involved in the spontaneous adsorption of LTA to erythrocytes. The LTA, its lipid moiety, and anti-LTA blocked adherence of group A streptococci to human epithelial cells, suggesting that small amounts of LTA may reside on the streptococcal surface to mediate attachment and colonization of these organisms on mucosal surfaces in vivo.

Human immune response to immunization with a structurally defined polypeptide fragment of streptococcal M protein.

We tested the ability of pepsin-extracted, highly purified M protein to induce type-specific immunity in experimental animals and humans. M protein was prepared from limited peptic digests of whole group A type 24 streptococci and was purified to chemical homogeneity as judged by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, quantitative amino acid analysis, and Edman degradation. For vaccination, the lyophilized M24 protein preparation (pep M24) was precipitated in aluminum hydroxide. When injected into laboratory animals, alum-precipitated pep M24 produced type-specific protective antibodies and was free of non-type-specific immunoreactivity. In man, skin tests with 1-microgram doses of pep M24 were negative in all 37 adults tested. 12 adult human volunteers received two-four subcutaneous injections of 100-200 micrograms of alum-precipitated pep M24 at intervals of at least 2 wk. The immune response to pep M24 was measured by a variety of assays designed to detect (a) type-specific humoral antibodies (opsonophagocytic, long chain, and mouse protection tests); (b) total humoral antibodies (complement fixation and enzyme-linked immunosorbent assay); (c) cellular immunity (skin tests); and (d) heart cross-reactive antibodies (immunofluorescence). Type-specific opsonic antibodies developed in 10 of the 12 vaccinees, and positive delayed-type skin tests developed in 11. Immune sera from two of the vaccinees were effective in mouse-protection tests against challenge with M24 but not M6 streptococci. None of the volunteers developed heart-reactive antibodies or antibodies to non-type-specific M protein antigens. Alum-precipitated pep M24 was well-tolerated in man, and no serious local or systemic reactions were observed. Thus, pep M24 induces type-specific, protective antibodies in doses that are well-tolerated in man.

Antiadhesive properties of a quaternary structure-specific hybridoma antibody against type 1 fimbriae of Escherichia coli.

The relationship between the structure and biological function of type 1 fimbriae of Escherichia coli was investigated using a set of monoclonal antibodies directed against conformation-specific antigenic determinants. Of three monoclonal antibodies tested, only one (clone CD3) prevented adhesion of the vaccine strain to epithelial cells or guinea pig erythrocytes. The antibody produced by CD3, but not that produced by the other two hybridoma clones (AA8 and GG1), precipitated isolated fimbriae by double diffusion in agar gel and was shown to bind in a highly discrete, periodic manner along the length of each of the fimbriae by immunoelectron microscopy. Immunoelectroblots of type 1 fimbrial subunits and polymers electrophoresed in SDS-gels indicated that the antibodies in AA8 and GG1 reacted only with fimbrial monomers (mol wt 17,000), whereas the antibody in CD3 reacted only with polymers of mol wt 102,000 (hexamers) or higher. ELISA inhibition assays demonstrated that dissociated fimbrial subunits lost their reactivity with antibody CD3 but gained reactivity with antibodies AA8 and GG1. Conversely, when allowed to reassemble in vitro in the presence of 5 mM MgCl2, the reassembled fimbriae lost their reactivity with antibodies AA8 and GG1 but regained reactivity with antibody CD3. These results demonstrated that certain antigenic epitopes are dependent on quaternary structural determinants, whereas others are independent of quaternary fimbrial structure and also are inaccessible for antibody binding in fimbriae once they have been assembled. These monoclonal antibodies should prove useful in studies of the structural determinants of the biological function of type 1 fimbriae as well as in studies of fimbrial synthesis, transport, and assembly.