In vivo actin cross-linking induced by Vibrio cholerae type VI secretion system is associated with intestinal inflammation

Abstract

Type VI secretion systems (T6SSs) have recently been recognized as potential virulence determinants of many Gram-negative bacterial pathogens. Although mechanistic studies are lacking, T6SS-dependent phenotypes can be observed in various animal models of infection. Presumably translocation of T6SS effectors into target cells is involved in virulence, but few such effectors have been identified. A hallmark of T6SS function is the in vitro secretion of Hcp and VgrG proteins, which are thought to form part of an extracellular secretion apparatus. One well-characterized effector domain is the C-terminal actin cross-linking domain (ACD) of the VgrG-1 protein, constitutively secreted by the T6SS of Vibrio cholerae strain V52. Previous work indicated that translocation of VgrG-1 occurred only after endocytic uptake of bacteria into host cells. VgrG-1-induced actin cross-linking impaired phagocytic activity of host cells, eventually causing cell death. To determine whether V. cholerae T6SS is functional during animal infection, derivatives of V52 were used to infect infant mice. In this infection model a diarrheal response occurred, and actin cross-linking could be detected. These host responses were dependent on a functional T6SS and on the ACD of VgrG-1. Gene expression and histologic studies showed innate immune activation and immune cell infiltration in the intestinal lumen. The T6SS-dependent inflammatory response was also associated with increased recovery of V. cholerae from the intestine. We conclude that the T6SS of V52 induces an inflammatory diarrhea that facilitates replication of V. cholerae within the intestine.

Fig. 1.

Fluid accumulation elicited by V. cholerae T6SS. Fluid accumulation was determined by calculating ratio of intestinal weight to remaining body weight. (A) controls for RtxA (ATM-1) and CtxAB (ATM-13) and (B) additional T6SS mutants. Groups marked with asterisks have values that are significantly higher than unmarked groups (Student’s t test, P < 0.05). Bars indicate mean values for each group. Four to seven mice were used per group. Full genotypes are listed in Table 1.

Fig. 2.

In vivo actin cross-linking. Infant mice were infected for 4 h with various V52 strains of ΔhlyA hapA background. Single-cell suspensions were prepared from harvested intestines by passage through cell strainers and collected by centrifugation. Samples were analyzed by Western blot against actin.

Fig. 3.

Histology of infant mouse small intestine. Infant mice were infected with various V52 strains of ΔrtxA hlyA hapA background. Intestinal sections were stained with H&E (40×). Sections are from (A) mock-infected mice, (B and C) T6SS mutant-infected mice, and (D) ATM-1 infected mice.

Fig. 4.

Heat map of differentially expressed genes. Significant genes were determined by the SAM algorithm with a 2-fold change cutoff. (A) Genes with higher expression and (B) genes with lower expression are shown. Values depicted are log₂ expression values from three independent experiments.

Fig. 5.

Quantitative PCR verification of differentially expressed genes. Relative expression was calculated by normalization to GAPDH expression and then normalization to mock-infected expression. Values given are means ± SD of biologic triplicates.

Fig. 6.

Colony forming units (CFU) recovered from small intestine of infant mice after overnight infection. Groups preinfected for 4 h with ATM-15 (ATM-1ΔlacZ) or ATM-16 (ATM-6ΔlacZ) are indicated. Bars represent geometric means for each group. Four to five mice were used per infection group. Asterisks indicate groups with significantly higher means than unmarked groups (Mann-Whitney test, P < 0.05).