Structural vaccinology to thwart antigenic variation in microbial pathogens

Pathogenic microbes evolve continuously to adapt to the immune responses of their infected hosts. For some pathogens, this presents a formidable challenge, because host adaptive immune responses may neutralize key virulence attributes of the microbe, thereby establishing immunity to subsequent infection (1). Under such immune selection, microbes evolve variants with altered antigenic but unaltered functional properties, thereby restoring virulence for invasion of hosts with neutralizing immunity because of prior exposure with another isolate of the same pathogen (2). How then can one find the key protective antigens of microbes, and what technologies may be applied to thwart their evasive strategies? Answers to these queries are provided in a paper by Nuccitelli et al. (3) in PNAS. Large-scale DNA sequencing of microbial genomes combined with bioinformatic analyses of their predicted products provide access to the genetic information of all isolated variants of a pathogen (4). Reverse vaccinology involves the exhaustive testing of recombinant translational products as vaccine candidates in search of subunits that elicit specific protective immunity in animal models (Fig. 1). Applied to the Gram-positive pathogen Streptococcus agalactiae, these approaches led to the identification of pilus genes that are critically important for the disease process but display variation among different isolates (5). Combining structural biology approaches with domain mapping of protective epitopes, Nuccitelli et al. (3) string together six antigenic variants of the same functional unit into a single recombinant antigen that elicits antibody-mediated vaccine protection against all of S. agalactiae isolates examined.

Fig. 1.

Structural vaccinology to develop vaccines against variant microbes. Microbes evolve variants of protective surface antigens to infect hosts with existing immunity. Key protective antigen genes are identified in the genome sequence of pathogen isolates on the basis of their variant sequence. Reverse vaccinology provides for the testing of isolated recombinant products in animal models. Structural vaccinology involves analysis of the 3D structure and vaccine testing of individual domains. Variant domains that elicit protective immunity are assembled into a single hybrid antigen to elicit protection against all isolates of the pathogenic microbe.

Lancefield group B streptococci (GBS or S. agalactiae) are a leading cause of neonatal sepsis, meningitis, and pneumonia (6). Two clinical syndromes have been described. Early-onset disease occurs in neonates 7 d of age or less, resulting from intrapartum transmission of GBS replicating in the genital tract and amniotic fluid (6). Late-onset disease occurs on or after day 7 and may be acquired by routes other than mother-to-infant transmission (6). American obstetricians reduced the incidence of early-onset GBS infection from 1.7 per 1,000 live births in 1990 to 0.35 per 1,000 live births in 2005 (a total of 1,425 cases in the United States) (6). This success is based on rigorous antenatal culture screening of pregnant women combined with intrapartum antibiotic prophylaxis for GBS carriers (7). Nevertheless, neonatal GBS infections continue to represent a global health concern that, because of resource limitations, cannot be addressed with culture screening and antibiotic prophylaxis alone (8).

Using a mouse model of infection, Lancefield and Freimer (9) provided evidence for protective immunity against GBS, which is based on type-specific antibodies directed against capsular polysaccharide (9). Indeed, mothers with antibodies specific for GBS capsular polysaccharides provide passive transfer of type-specific immunity for their newborn babies (10). Purification of multiple different capsular polysaccharides, their conjugation to protein carriers, and assembly of multivalent vaccines are dauntingly laborious as well as time- and resource-consuming tasks. Nevertheless, the job can get done, and pneumococcal capsule conjugates serve as an example of a multiple-decade, successful effort in establishing vaccine safety and efficacy (11). To avoid the aforementioned manufacturing and developmental issues of conjugate vaccines, multiple research teams isolated surface protein antigens that are expressed by many or most GBS isolates. This effort has also met with success, which is documented in preclinical studies with α- and β-C proteins (12), Rib (13), BipA, and others (14). Nevertheless, a universal protein vaccine for all GBS isolates, composed perhaps of many of these surface protein antigens, has not yet emerged.

All GBS isolates examined to date elaborate pili, long filamentous structures that protrude from the bacterial surface (5). Genome sequence information from many GBS isolates revealed three pilus islands, designated PI-1, -2a, and -2b (15). GBS isolates harbor one or two of the three islands, either PI-1 and -2a or PI-2a and -2b (15). Each pilus island represents a cluster of three subunit genes that encode one major [backbone protein (BP)] and two minor pilins of which ancillary protein 1 (AP1) functions as the tip adhesin, whereas AP2 represents the base subunit for pilus attachment to the cell wall envelope (16). Sortase enzymes catalyze pilus assembly and attachment to the cell wall envelope. With the exception of the backbone protein for PI-2a (BP2a), which elaborates at least six immunological variants, the ancillary and backbone proteins of GBS pili display sequence conservation (15). When used as subunit vaccines in preclinical trials with a neonatal mouse challenge model, BPs elicit higher levels of GBS disease protection than APs (15). The protective antigen status of BPs is attributable to pilus structure: sortases form pili from single subunits of AP1 or AP2 but incorporate several hundred BP pilins (16). Vaccine protection seems to be based on opsono-phagocytic clearance of GBS whose pili can be decorated with antibodies.

By solving the crystallographic structure of BP2a (from GBS strain 515), Nuccitelli et al. (3) discern four Ig-like domains (D1-D2-D3-D4), which could be anticipated to function as independent folding units (17, 18). Individual domains are tested as subunit vaccines; however, only D3 elicits protection similar to full-length BP2a (3). Nuccitelli et al. (3) then cobble together the D3 domains of six variant BP2a molecules that are connected by short Gly-Ser-Gly-Ser spacers. Expression in Escherichia coli generates a stable hybrid product that provides protection against all six prototypic pilus variant GBS strains examined (3). The use of this technology, designated structural vaccinology, can likely be expanded well beyond the boundaries of GBS vaccines. The availability of genome sequences from many isolates of bacterial pathogens permits rapid identification of genes with variable amino acid sequences in their products. Assuming that these genes encode secreted proteins accessible to the immune system, this approach may, by default, identify protective antigens of pathogens that are subject to immune selection. Applying structural vaccinology to candidate protective antigens may, thus, promote rapid development of designer vaccines to elicit protection against all microbial variants captured in hybrid vaccine proteins.

As with all technologies, reverse vaccinology and structural vaccinology have limitations. What if variable genes encode products that generate secreted carbohydrates or glycolipids? Clearly, immune responses against polypeptides cannot elicit antibodies that recognize carbohydrates, which often constitute protective antigens of bacterial pathogens. Furthermore, what works well for the Ig-like domains of pilin subunits may not apply to other protein antigens [for example, the N-terminal domains of group A streptococcal M proteins (19) or the variant surface proteins of trypanosomes (20)]. Nevertheless, structural vaccinology represents an important advance that hopefully will be tested soon in clinical trials for safety and efficacy against its founding paradigm, neonatal GBS disease. A multivalent GBS pilus vaccine that captures all pathogen isolates in a single hybrid antigen could facilitate vaccine manufacturing and ease the path to licensure (8). This vaccine would address key problems in GBS disease prevention: massive antenatal use of prophylactic antibiotics with concomitant increases in bacterial drug resistance (7), emergence of strains with new capsular types, and commitment of our society to save the lives of newborn babies or prevent the debilitating, life-long sequelae associated with GBS disease.