Abstract
Porphyromonas gingivalis is a key pathogen in periodontitis. Its outer membrane contains the RagAB transport complex, which has been implicated in protein uptake, essential for a proteolytic species. RagA is a 22-stranded β-barrel, and RagB is the corresponding 4-TPR lid, together forming a TonB-dependent system acting as a "pedal bin". Four different alleles were observed, of which ragAB-1 is more virulent than the others. Our aim was to map ragAB in 129 strains of P. gingivalis and related species available in our collection, supported by a newly introduced universal PCR for amplification/sequencing of all four ragA variants and to find reasons for the differences in virulence and/or fitness. Regarding the PCR method, by pairing established Long-PCR primers with our newly designed sequencing primers (ragA-F0, -F1, -R2, -R2a, -R4), it was possible to amplify and sequence all four ragA variants. The same was not possible for ragB due to high heterogeneity. The mapping allowed us to type all strains into ragAB-1-4. For each type, some strains (of mainly animal origin such as Porphyromonas gulae) with slightly different amino acid sequences were identified (designated ragAB-1a to -4a). In terms of function, the transfer of recently discovered SusCD information to the similar RagAB complex provided new insights. Substrate specificity as well as length of pedal could be the route to differential virulence (survival rate, fitness) as Rag-1 (closer related to Rag-3/4) and Rag-2 were found to be massively different here. In general, substrate-ligand-binding sites seem to be quite variable with the exception of Rag-1, probably indicating nutritional preferences. In addition, an insertion (8 aa long) found in loop L7 throughout RagA-2 could not only affect the dynamics of lid opening/closing but might also alter the associated substrate throughput rate.