Natural avirulent Shigella boydii strain in the Brazilian Amazon lacks major virulence genes and present Type II, Type III and Type VI Secretion Systems

Background Among Shigella species, Shigella boydii has always displayed a smaller role to the overall Shigella burden, frequently placed at third in epidemiological studies and described as restricted to Southeast Asia. Here we characterize an S. boydii isolated from an epidemiological study enrolling 1,339 Brazilian children from the Amazon region, in which Shigella species solely was the fourth cause of bacterial diarrhea. S. boydii strain 183 was isolated from rotavirus co-infected children with acute diarrhea. Here we aimed to characterize this strain regarding virulence and, immune response in a pulmonary model. Methods An in vitro HEp-2 epithelial cell invasion assay was used to compare the invasive phenotype of S. boydii strain 183 with clinical and highly virulent S. flexneri strain, both isolated from Brazilian children. A murine pulmonary model was performed to assess lung damage by histopathological analysis. mRNA expression of immune response key genes was retrieved by multiplex real-time PCR and correlations were obtained by network analysis. Broad genome analysis was performed to confirm S. boydii identity and define its virulence profile. Results S. boydii strain 183 showed fewer invasion rates in vitro and tissue damage in vivo as compared to virulent S. flexneri 201. When compared to a survival challenge in mice, S. boydii had 100% survival against 10% of virulent S. flexneri. Overall, mRNA immune gene expression suggests a protective response against S. boydii strains 183, in contrast to the inflammatory response induced by the virulent S. flexneri strain 201. Network analysis with S. boydii strain 183 displayed IFN-γ protagonism, contrasting with the correlations centralized on TNF-α by the virulent S. flexneri strain 201. The genome showed a lack of effector proteins and enterotoxins in S. boydii strain 183, and sequencing analysis of Ipa invasins revealed mutations at functional sites. This avirulent S. boydii strain 183 presents the Type II Secretion System, T6SS, in addition to T3SS. Conclusions In addition to causing no disease, S. boydii strain 183 lacks effector proteins and enterotoxins. The presence of T6SS additional secretion system could provide an advantage to establish this strain among commensal bacteria. AUTHOR SUMMARY The Shigella genus is a human pathogen responsible to shigellosis and remains one of the significant causes of morbidity and mortality in children under five years old. This genus has four species, Shigella flexneri, Shigella sonnei, Shigella boydii, and Shigella dysenteriae. S. flexneri and S. sonnei are the most common in the worldwide infections; S. dysenteriae is rarely found, and S. boydii is responsible for 1% of the infections and is known to be restricted to Southeast Asia. Once S. boydii have a relatively small role in global Shigella disease, there are few studies regarding its virulence and mechanisms. Here we characterize an S. boydii isolated from Brazilian children from the Amazon region, and aimed to describe this strain regarding virulence. It is known that Shigella species use the Type 3 Secretion System (T3SS) to invade and colonize the human intestine. We found in S. boydii the presence of Type 2 Secretion System (T2SS), Type 6 Secretion System (T6SS), in addition to the T3SS. The T6SS have been described in S. sonnei only, granting a competitive advantage against S. flexneri mixed cultures. The presence of T6SS additional secretion system could provide a benefit to establish this strain among commensal bacteria.


186
DNA Sequencing: Genomic DNA was extracted (15) and DNA libraries were built using the 206 The same was performed with enzyme commission information and metabolic pathway maps; both 207 were extracted from the KEGG ftp server.

208
A volcano plot was constructed (19) for visual display of mRNA expression analysis results.
209 Briefly, the scheme was created with data obtained from the t-test and ANOVA, calculating the 210 unstandardized signal (e.g., log-fold-change) against the noise-adjusted/standardized signal (e.g., t-211 statistic or -log(10) (p-value) from the t-test 242 was identified as S. flexneri in pentaplex PCR (data not shown) and in whole-genome sequencing, re-243 named as Shigella flexneri str. 201 (NCBI accession # CP024981). This strain was isolated from 244 monoinfected children with shigellosis and carried the majority of the virulence genes ( Figure 1A) (5).
245 As both of the Shigella strains were wild-type bacteria, we also compared our results with Shigella 246 flexneri 5a (M90T), an invasion standard. All strains studied had ipaBCD, ial and uidA marker genes, 247 necessary for the invasive phenotype. An important finding was that strain 183 was negative for the 248 virF marker gene ( Figure 1A).

250
We assessed the ability of the S. boydii str. 183 to invade epithelial cells (EC) and multiply in 252 invasion reference, and while this bacterium displayed high invasiveness rates after one hour of cell 253 contact, the S. boydii str. 183 had success in invading but not multiplying in the EC cytoplasm ( Figure   254 1B). Surprisingly, this profile remained even after three hours of cell contact. The S. boydii str. 183 255 invasion rates did not increase, nor was their multiplication, suggesting that bacteria were unable to 256 escape from the phagosome. Conversely, S. flexneri str. 201 had higher invasion rates, with several 257 infected cells enclosing more than five bacteria per cell (p<0,001) after one hour of cell contact and 258 substantially increasing after three hours with more than ten bacteria per cell (p<0,001).

260
To illustrate the S. boydii str. 183 behavior in the HEp-2 cell assay, we showed the evolution 261 of Shigella invasion and multiplication inside the HEp-2 cell. After gentamicin treatment, lack of 262 invasion features was assessed by the absence of bacterial cell inside the invasion vacuole (left panel; 263 Figure 1C). The multiplication was noted by the occurrence of bacteria undergoing cell division, 264 featuring bacterial doublets (central panel; Figure 1C). Intense bacterial amplification was observed 265 in both virulent S. flexneri str. 201 and M90T (right panel; Figure 1C). In vitro HEp-2 cell invasion assay

283
We assessed the immune response to define patterns of response-associated pathology and 284 immunity. The gene expression of twenty-one immune-related genes were measured using multiplex 285 real-time PCR mRNA gene expression. To assess the kinetics of gene expression, we gathered data 286 from all strains in two-time points. Gene expression at 48 hours differed from that at 24 hours, and 287 we found up-regulation of the Interleukin 1β (IL-1β), Interleukin 6 (IL-6), and NOD1 (Nucleotide-288 binding Oligomerization Domain 1), consistent with the premise of an innate immune response to 289 bacteria (Table 1).

291
To determine the regulation range of the immune response genes, we measured qualitatively

440
We assessed whether our S. boydii str. 183 was able to subvert the host innate response. Of