Development of two animal models to study the function of Vibrio parahaemolyticus type III secretion systems

Abstract

Vibrio parahaemolyticus is an emerging food- and waterborne pathogen that encodes two type III secretion systems (T3SSs). Previous studies have linked type III secretion system 1 (T3SS1) to cytotoxicity and T3SS2 to intestinal fluid accumulation, but animal challenge models needed to study these phenomena are limited. In this study we evaluated the roles of the T3SSs during infection using two novel animal models: a model in which piglets were inoculated orogastrically and a model in which mice were inoculated in their lungs (intrapulmonarily). The bacterial strains employed in this study had equivalent growth rates and beta-hemolytic activity based on in vitro assays. Inoculation of 48-h-old conventional piglets with 10(11) CFU of the wild-type strain (NY-4) or T3SS1 deletion mutant strains resulted in acute, self-limiting diarrhea, whereas inoculation with a T3SS2 deletion mutant strain failed to produce any clinical symptoms. Intrapulmonary inoculation of C57BL/6 mice with the wild-type strain and T3SS2 deletion mutant strains (5 × 10(5) CFU) induced mortality or a moribund state within 12 h (80 to 100% mortality), whereas inoculation with a T3SS1 deletion mutant or a T3SS1 T3SS2 double deletion mutant produced no mortality. Bacteria were recovered from multiple organs regardless of the strain used in the mouse model, indicating that the mice were capable of clearing the lung infection in the absence of a functional T3SS1. Because all strains had a similar beta-hemolysin phenotype, we surmise that thermostable direct hemolysin (TDH) plays a limited role in these models. The two models introduced herein produce robust results and provide a means to determine how different T3SS1 and T3SS2 effector proteins contribute to pathogenesis of V. parahaemolyticus infection.

FIG. 1.

(A) The growth rates of V. parahaemolyticus strains (wild-type strain NY-4 [serotype O3:K6] and T3SS1 knockout KO, T3SS2 KO, and T3SS1/2 KO mutant strains) were compared over 24 h in broth culture. The value at each hour is an average of three biological replicates (variance not shown). The final OD₆₀₀ was not different for the different strains (P > 0.05). Ctrl, control. (B) The hemolytic activity of the supernatants was evaluated by measuring the optical absorbance at 541 nm 4 h after mixing culture supernatant with human red blood cells. The values (bars) are the means plus standard errors of the means (error bars) for three biological replicates. For simplicity, the values shown here were normalized by dividing the values by the value for the lysis positive control. The ΔtdhA strain and the negative control (Neg Ctrl) had significantly lower hemolytic activity (P < 0.05) than the other strains (indicated by an asterisk). The outcome of the analysis did not change when the values were normalized by CFU/ml (except the negative control was not included in the latter analysis).

FIG. 2.

(A) The presence or absence of diarrhea was evaluated every 2 h during the first 12 h postinoculation and every 3 h thereafter. The average diarrhea severity score (see Materials and Methods) was estimated for six piglets in each treatment group. Piglets infected with wild-type strain NY-4 or T3SS1 KO mutant strain had significantly higher clinical scores (P < 0.003) than piglets infected with T3SS2 KO strain or piglets in the PBS control group at 2 and 4 h postinoculation (p.i) (indicated by three asterisks). N.S, not significant. (B) Rectal swabs were cultured every 2 h during the first 12 h postinoculation and every 3 h thereafter. Bars show the percentages of piglets shedding V. parahaemolyticus. Numeric values below bars indicate hours p.i.

FIG. 3.

(A) Piglets infected with wild-type strain NY-4 and T3SS1 KO mutant strain showed a mild to severe colonic submucosal and serosal edema (arrows). (B) No histological lesions were observed in the piglets infected with the T3SS2 KO strain. The sections were stained with hematoxylin and eosin. Magnification, ×100.

FIG. 4.

Four groups of C57BL/6 adult mice (10 mice in each group) were inoculated with different strains of V. parahaemolyticus. The survival rate was evaluated for two independent replicates shown in panels A and B by Kaplan-Meier and log-rank tests. Values that are significantly different (P < 0.0001) are indicated by three asterisks. Values that are not significantly different (N.S) are indicated.

FIG. 5.

Histopathological examination of the mouse lung specimens. (A) At 8 h postinoculation for all infected groups (wild-type strain NY-4 and T3SS2 KO, T3SS1 KO, and T3SS1/2 KO mutant strains), the mice presented diffuse capillary congestion with perivascular edema and large areas of alveolar hemorrhage. The alveolar lumen is severely distended by large aggregates of neutrophils intermixed with a few foamy macrophages and cellular debris. (B) At 12 h postinfection (NY-4, T3SS1 KO, and T3SS1/2 KO strains), there was severe neutrophilic infiltration. The bronchi were completely plugged by abundant suppurative exudates and infiltrated by large peribronchiolar cuff of neutrophils and macrophages. (C) Section in control animals (PBS only) are characterized by diffuse capillary hyperemia. The sections were stained with hematoxylin and eosin. Magnification, ×200.

FIG. 6.

Average CFU recovered from lung (A), gastrointestinal tract (B), and liver (C) for mice from which bacteria were recovered. Mice infected with wild-type strain NY-4 and T3SS2 KO mutant strain died or were euthanized between 8 and 12 h postinoculation. The mice inoculated with the T3SS1 KO mutant strain were euthanized at the same time to have matched bacterial counts over time. The number of animals with recoverable bacteria/number of inoculated mice (n/n°) is shown on the x axes.