Group B Streptococci (GBS) are the leading cause of neonatal sepsis and meningitis. Maternal IgG antibodies to the GBS PS protect the neonate from invasive GBS disease. Such antibodies are deficient in mothers of infants who develop GBS type III disease. A case-control study was designed by NICHD to confirm the amounts of antibody needed for protection. As part of this study GBS strains were collected from seven hospitals across the U.S. to study serotype prevalence. We developed a new inhibition ELISA for serotyping GBS strains, because existing methods are not quantitative, and need high titered antisera against each GBS serotype. We used a GBS hyperimmune IVIG solution containing antibodies to GBS types Ia, Ib, II and III, but the assay can also be done using rabbit antisera as we have done for serotype type V. Our results are comparable with conventional methods and inhibition ELISA is more sensitive and can be used to quantitate capsular PSs. We have serotyped over 1600 strains from cases of invasive disease and from healthy controls. Approximately 40% of the disease isolates are type Ia. There is yet no standard method to determine GBS anti-capsular PS IgG antibody concentrations. Four different PS preparations were compared as antigens in ELISA : free PS, free PS mixed with methylated human serum albumin, PS conjugated to biotin or PS conjugated to human serum albumin. All four Ps preparations measured as low as 0.05 microg/ml of IgG antibody. Antibody concentrations estimated by all Ps preparations were similar for GBS type Ia. Estimated antibody concentrations were different depending on the PS preparation used for GBS types III and V. Free GBS type III PS or PS mixed with mHSA provides a sensitive and specific antigen for ELISA. The GBS type III PS is structurally similar to the pneumococcal type -14 PS, differing only in the presence of sialic acid residues in type III. Our studies show that conjugation type III PS to biotin or to HSA may alter its conformation, exposing regions of the PS reactive with PN-14 antibodies. Differences in estimated type V antibody concentrations between methods may also be due to differential expression of epitopes resulting from the method of attachment of PS to the ELISA plate. To study this, we have treated the type V PS with neuraminidase to cleave the terminal sialic acid residues to expose the core PS. We will immunize rabbits with the desialyated type V PS. We are in the process of correlating ELISA to opsonophagocytosis assays.
