Shigella sonnei infection of zebrafish reveals that O-antigen mediates neutrophil tolerance and dysentery incidence

Shigella flexneri is historically regarded as the primary agent of bacillary dysentery, yet the closely-related Shigella sonnei is replacing S. flexneri, especially in developing countries. The underlying reasons for this dramatic shift are mostly unknown. Using a zebrafish (Danio rerio) model of Shigella infection, we discover that S. sonnei is more virulent than S. flexneri in vivo. Whole animal dual-RNAseq and testing of bacterial mutants suggest that S. sonnei virulence depends on its O-antigen oligosaccharide (which is unique among Shigella species). We show in vivo using zebrafish and ex vivo using human neutrophils that S. sonnei O-antigen can mediate neutrophil tolerance. Consistent with this, we demonstrate that O-antigen enables S. sonnei to resist phagolysosome acidification and promotes neutrophil cell death. Chemical inhibition or promotion of phagolysosome maturation respectively decreases and increases neutrophil control of S. sonnei and zebrafish survival. Strikingly, larvae primed with a sublethal dose of S. sonnei are protected against a secondary lethal dose of S. sonnei in an O-antigen-dependent manner, indicating that exposure to O-antigen can train the innate immune system against S. sonnei. Collectively, these findings reveal O-antigen as an important therapeutic target against bacillary dysentery, and may explain the rapidly increasing S. sonnei burden in developing countries.Author summary: Shigella sonnei is predominantly responsible for dysentery in developed countries, and is replacing Shigella flexneri in areas undergoing economic development and improvements in water quality. Using Shigella infection of zebrafish (in vivo) and human neutrophils (in vitro), we discover that S. sonnei is more virulent than S. flexneri because of neutrophil tolerance mediated by its O-antigen oligosaccharide acquired from the environmental bacteria Plesiomonas shigelloides. To inspire new approaches for S. sonnei control, we show that increased phagolysosomal acidification or innate immune training can promote S. sonnei clearance by neutrophils in vivo. These findings have major implications for our evolutionary understanding of Shigella, and may explain why exposure to P. shigelloides in low and middle-income countries (LMICs) can protect against dysentery incidence.