Immunization of Mice with Single PspA Fragments Induces Antibodies Capable of Mediating Complement Deposition on Different Pneumococcal Strains and Cross-Protection

PspA is an important candidate for a vaccine with serotype-independent immunity against pneumococcal infections. Based on sequence relatedness, PspA has been classified into three families comprising six clades. We have previously addressed the cross-reactivity of antibodies against PspA fragments containing the N-terminal and proline-rich regions of PspA from clades 1 to 5 (PspA1, PspA2, PspA3, PspA4, and PspA5) by Western blot analysis and reported that anti-PspA4 and anti-PspA5 were able to recognize pneumococci expressing PspA proteins from all of the clades analyzed. We have now analyzed the functional capacity of these antibodies to bind and to mediate complement deposition on intact bacteria in vitro. Our results show that both PspA4 and PspA5 elicit antibodies that are able to bind and to mediate complement deposition efficiently on pneumococcal strains bearing PspA proteins from clades 1 to 5. Moreover, mice immunized with PspA4 and PspA5 were protected against an intranasal lethal challenge with strains expressing PspA proteins from the two major families. PspA4 and PspA5 are thus able to induce antibodies with a high degree of cross-reactivity in vitro, which is reflected in cross-protection of mice. We have also analyzed the contribution of the nonproline (NonPro) block within the conserved proline-rich region to the reactivity of anti-PspA antibodies, and the results indicate that N-terminal α-helical region, the blocks of proline repeats, and the NonPro region can influence the degree of cross-reactivity of antibodies to PspA.Streptococcus pneumoniae is an important human pathogen, being responsible for millions of deaths worldwide every year. The pneumococcal disease burden could be greatly reduced by the use of the current seven-valent conjugate vaccine, but the high cost and restricted serotype coverage limit its widespread use, especially in developing countries. New-generation vaccines containing up to 13 serotypes are expected to increase vaccine coverage, but the serotype replacement in colonization and disease by nonvaccine serotypes observed with the use of the seven-valent conjugate vaccine (8-9, 11) further emphasizes the importance of the development of alternative vaccines. Protein antigens such as PspA (pneumococcal surface protein A) could be used to induce serotype-independent immunity at a low cost (24).PspA is present in all isolated pneumococcal strains and was shown to be an important virulence factor, interfering with complement deposition (19, 21, 25), killing by apolactoferrin (23), and immune adherence to erythrocytes (12). It has been shown to induce protection in mice in carriage, pneumonia, and fatal systemic models (2, 4, 16). Mature PspA is composed of a mosaic structure with four domains: an α-helical N-terminal domain, a proline-rich region, a choline-binding domain, and a short hydrophobic tail (10, 27-28). PspA shows variability in the surface-exposed N-terminal region, and a classification was proposed based on sequence relatedness of the C-terminal portion of the α-helix, the clade-defining region. It has been classified into three families encompassing six clades. Family 1 (Fam1) is composed of clades 1 and 2, Fam2 includes clades 3, 4, and 5, and Fam3, which is rarely isolated, comprises clade 6 (10). Since the degree of similarity seems to be reflected in cross-reactivity, it has been proposed that a broad-coverage vaccine should contain at least one fragment from each of the two major families.Immunization of healthy adults with a single recombinant fragment of PspA in a phase I clinical trial showed the induction of cross-reactive antibodies (14) that were able to induce passive protection in mice challenged intravenously (3). The natural exposure of adults to several pneumococcal strains might be responsible for the cross-reactivity detected, with the immunization with PspA acting as a booster dose.Because of the diversity observed in PspA, it is extremely important to analyze whether each fragment selected to compose a vaccine is indeed able to induce cross-protection. We have previously addressed the degree of cross-reactivity of antibodies to recombinant fragments including the N-terminal and proline-rich regions of PspA proteins from clades 1 to 5 (PspA1, PspA2, PspA3, PspA4, and PspA5) by Western blot analysis of 35 strains isolated in Brazil. As expected, we have observed higher cross-reactivity within the same clade. Within Fam1, anti-PspA1 serum also showed cross-reaction with PspA2-expressing strains, while anti-PspA2 showed reaction restricted to the same clade. Within Fam2, anti-PspA3 serum also showed reactivity restricted to PspA3-expressing strains, while anti-PspA5 and, more strikingly, anti-PspA4 sera showed a broad recognition capacity, being able to react with strains expressing PspA proteins from clades 1 to 5 (7). The ability of sera to recognize a pneumococcal strain by Western blot analysis does not necessarily correlate with their capacity to induce protection in vivo though. In fact, the levels of antibodies to PspA detected by enzyme-linked immunosorbent assay (ELISA) or through surface staining of the bacteria failed to provide a useful correlate of protection (22). Based on the strong evidence supporting the importance of complement in protection against pneumococcal disease, it was proposed that in vitro complement deposition mediated by antibody may be used as a surrogate assay for the prediction of protection induced by surface antigens of pneumococci (15). This work aimed at further characterizing antibodies against the PspA1, PspA2, PspA3, PspA4, and PspA5 N-terminal fragments in terms of their capacity to mediate C3 deposition on the surface of pneumococci expressing PspA proteins from different clades. Moreover, protection of mice against a lethal intranasal challenge with strains expressing PspA from Fam1 or Fam2 was also analyzed. The basis for the broad reactivity observed in the anti-PspA4 serum by Western blot analysis was also further investigated. Of the five PspA fragments analyzed, PspA4 was the only one containing a nonproline (NonPro) block within the proline-rich region. Not all native PspA proteins include this region: of 24 PspA sequences analyzed by Hollingshead and collaborators (10), 14 were shown to have this NonPro block. We have thus examined whether this region would be responsible for increased cross-reactivity.