Group B Streptococcus (GBS) is the leading cause of neonatal sepsis and meningitis in the United States. The surface-associated C protein alpha antigen of GBS is thought to have a role in both virulence and immunity. We previously cloned the C protein alpha antigen structural gene (named bca for group B, C protein, alpha) intoEscherichia coil. Western blots of both the native alpha antigen and the cloned gene product demonstrate a regularly laddered pattern of heterogeneous polypeptides. The nucleotide sequence ofthe bca locus reveals an open reading frame of 3060 nucleotides encoding a precursor protein of 108,705 Da. Cleavage of a putative signal sequence of 41 amino acids yields a mature protein of 104,106 Da. The 20,417-Da N-terminal region of the alpha antigen shows no homology to previously described protein sequences and is followed by a series of nine tandem repeating units that make up 74% of the mature protein. Each repeating unit is identical and consists of 82 amino acids with a molecular mass of 8665 Da, which is encoded by 246 nucleotides. The size of the repeating units corresponds to the observed size differences in the heterogeneous ladder of alpha C proteins expressed by GBS. The C-terminal region of the alpha antigen contains a membrane anchor domain motif that is shared by a number of Gram-positive surface proteins. The large region of identical repeating units in bca defines protective epitopes and may play a role in generating phenotypic and genotypic diversity of the alpha antigen.

Members of a family of repeat-containing surface proteins of group B streptococci (GBS) defined by the alpha C and Rib proteins exhibit size variability and cross-reactivity and have been studied as potential vaccine components. We report evidence of horizontal DNA transfer with subsequent recombination as a mechanism generating diversity within this antigen family. Alp2 and Alp3 are additional members of the alpha C protein family identified in strains of the emerging GBS serotypes V and VIII. Each contains an overall genetic organization highly similar to that of the alpha C and Rib proteins, including a tandem repeat region and conserved N-and C-terminal regions. Among different strains, protein size varies according to the number of tandem repeats within the corresponding gene. Unlike the alpha C and Rib proteins, however, the newly described alpha-like proteins contain other regions, including one similar to the IgA-binding region of the GBS beta C protein, a nontandem repeat region, and an isolated repeat highly homologous to the alpha C repeat. Sequence analysis of the regions flanking the alpha C protein gene on a 13.7-kb insert reveals several ORFs that are likely to be involved in basic metabolic pathways. Analysis of corresponding flanking regions in other GBS strains, including the parent strains of the newly described alpha-like proteins, shows striking conservation among all strains studied. These findings indicate that the alpha-like proteins are encoded by mosaic variants at a single genomic locus and suggest that recombination after horizontal DNA transfer is a means of generating diversity within this protein family.

Group B streptococci (GBS) are the most common cause of neonatal sepsis, pneumonia, and meningitis. The alpha C protein is a surface-associated antigen; the gene (bca)

Variable expression of repeating units of the protective alpha C proteins among clinical isolates of group B streptococci (GBS) may have implications for vaccine development. In this study, alpha C protein genes containing various numbers of repeats (1, 2, 9, and 16) were cloned in a T7 overexpression vector in Escherichia coli. Expression was induced by isopropyl-␤-D-thiogalactopyranoside, and proteins were purified by anionexchange, gel filtration, or affinity chromatography or by isoelectric focusing. Rabbits were immunized with purified 1-, 2-, 9-, or 16-repeat proteins. All proteins appeared to be highly immunogenic. Enzyme-linked immunosorbent assay inhibition with 9-repeat protein as the coating antigen and 9-repeat-antigen-elicited antiserum showed that a 200-fold-higher concentration of 1-repeat antigen than of 9-or 16-repeat antigen was required for 50% inhibition of antibody-antigen binding. The concentration of 2-repeat antigen required for 50% inhibition was intermediate relative to the concentrations of 1-and 9-repeat antigens. These results suggested that antibodies to 9-repeat antigen recognized predominantly a conformational epitope(s) contained in proteins with higher numbers of repeats (9 or 16) but lost considerable binding affinities for an epitope(s) contained in alpha C proteins with fewer repeats (1 or 2). Similar results were obtained with antiserum to 16-repeat antigen. However, antibodies to 1-and 2-repeat antigens recognized 1-, 2-, 9-, and 16-repeat antigens with equal binding affinities. This finding suggested that 1-and 2-repeat-elicited antibodies recognized an epitope(s) on individual repeats. Loss of repeating units from the alpha C proteins may result in decreased protection because the loss of epitopes (including conformational epitopes) gives the microorganisms the opportunity to escape host antibodies. If 1-and 2-repeat-elicited antibodies bind all alpha C proteins with equal affinity, regardless of their repeat number, they may prevent GBS strains with fewer repeats from escaping host immunity. Protection data obtained with antisera to the proteins with different repeat numbers support this hypothesis: mouse pups challenged with GBS strain A909 were better protected when immunized with 1-or 2-repeat-elicited antiserum (76 and 75%, respectively) than when immunized with 9-or 16-repeatelicited antiserum (41 and 48%, respectively).