Abstract:
Objective To investigate the genomic and biological characteristics of 89K-like pathogenicity islands (PAIs), and their roles in the pathogenesis of Streptococcus suis.
Methods Bioinformatics methods were used to analyze the genomic characteristics of the 89K-like PAIs. The transferability frequencies of the 89K-like PAIs were evaluated in conjugation assay. The antibiotic susceptibility profiles of the clinical strains and corresponding transconjugants were investigated by the minimum inhibitory concentration test. The pathogenicity of the clinical strains and corresponding transconjugants was evaluated using C57BL/6 mouse infection model and zebrafish infection model, respectively.
Results The sequence type (ST) of S. suis clinical strains GX54 and GX82 harboring 89K-like PAIs was ST665 and ST107, respectively. GX54 harbored a 75Kb PAI that exhibited 99.21% identity and 79.00% coverage to the 89K PAI carried by the ST7 epidemic strain SC84. In contrast, GX82 harbored an 87Kb PAI that exhibited 98.80% identity and 86.00% coverage to the 89K PAI. Both PAIs were integrated into the rplL site and contained a 15-bp att sequence 5′-TTATTTAAGAGTAAC-3′ in the flanking regions. Furthermore, both PAIs possessed a complete modular structure, including a functional NisK-NisR-like two-component signal transduction system and a type IV secretion system (T4SS). However, the SalK gene of the SalR-SalK-like two-component signal transduction system was truncated. The 75Kb PAI contained four additional genes, including two hypothetical protein genes, tet(O), and erm(B). On the other hand, the 87Kb PAI harbored 12 additional genes, ontaining plasmid-related recombinases genes, truncated replication protein genes, cat, tet(L), and erm(B). Additionally, genes encoding transposases and topoisomerases were also present. The transferability frequency of the 89K-like PAI harbored in strain GX54 was 4.61×10−5, significantly higher than the 8.46×10−6 of the 89K-like PAI harbored in strain GX82 (P<0.05). The 89K-like PAIs conferred the tetracycline and macrolide resistances to the corresponding transconjugants P1/7RIF-GX54 and P1/7RIF-GX82. Meanwhile, the mortalities of zebrafish infected with transconjugants P1/7RIF-GX54 and P1/7RIF-GX82 were also significantly higher than that of zebrafish infected with P1/7RIF (P<0.05). In addition, no significant difference in survival curves was observed between two transconjugants infected groups (P>0.05). The mortality of C57BL/6 mice infected with strain GX54 were significantly higher than that of C57BL/6 mice infected with strain GX82 (P<0.05). At 8 h post-infection, the bacterial loads in peripheral blood were similar between GX54 and GX82 infected mice (P>0.05). However, the production of TNF-α in serum of mice infected with strain GX54 was significantly higher than that of mice infected with strain GX82 (P<0.05). In addition, no significant difference in level of IL-6 was observed between serum of two infected groups (P>0.05). The virulence associated genes preferentially present in highly pathogenic S. suis strains were also found in strains GX54 and GX82, such as mrp, sly, epf, ofs, revS, nadR, SSU05_0473, neuB, and neuC.
Conclusion The 89K-like PAIs were firstly identified in S.suis ST665 and ST107 clinical strains. Although the 89K-like PAIs contributed to the increased pathogenicity of host strains, the virulence levels of S. suis strains cannot be effectively evaluated based on the presence of known virulence genes, including the 89K-like PAIs. The novel genetic elements and mechanisms were involved in the excessive TNF-α production at the early phase of infection which played a critical role in the pathogenesis of S. suis strains. The varied capacities to induce the TNF-α production at the early phase of infection between two clinical strains were not related to the differences in bacterial loads. An effective scheme for characterizing and subtyping the virulence level of S. suis strains is urgently needed.