RRSP and RID Effector Domains Dominate the Virulence Impact of Vibrio vulnificus MARTX Toxin

Abstract

Background: The bacterial pathogen Vibrio vulnificus causes severe septic foodborne infections. The multifunctional autoprocessing repeats-in-toxins (MARTX) toxin is an important secreted virulence factor. The effector domain region is essential for lethal intestinal infection in mice, but the contribution of each of the 5 effector domains to infection has not been investigated.

Methods: V. vulnificus mutants with varying effector domain content were inoculated intragastrically to mice, and the time to death was monitored to establish the contribution of each effector domain to overall virulence. Each strain was also tested for bacterial dissemination from the intestine to internal organs and for inhibition of phagocytosis.

Results: The effector domain region was required for V. vulnificus to inhibit phagocytosis by J774 macrophages, but no single effector domain was required. No single MARTX effector domain was necessary for bacterial dissemination. Nonetheless, overall survival of infected mice differed with respect to the infecting V. vulnificus strain. Removal of rid or rrsp significantly reduced the virulence potential of V. vulnificus, while deletion of duf1 or abh accelerated the time to death.

Conclusion: Rho GTPases inactivation domain and Ras/Rap1-specific endopeptidase each exert greater effects on virulence than other MARTX domains, suggesting that modulation of the Rho/Ras family of GTPases is a critical function of the toxin during intestinal infection.

Keywords: Vibrio vulnificus; rtxA; MARTX; foodborne; intragastric; pathogenesis; phagocytosis; toxin; virulence.

Figure 1.

Multifunctional autoprocessing repeats-in-toxins (MARTX) toxin variants. Schematic representation of the MARTX toxins produced by strains used in this study. ABH, alpha/beta hydrolase domain; Bla, heterologous beta-lactamase domain; COOH, protein C-terminus; CPD, cysteine protease domain; DUF1, domain of unknown function in the first position; MCF, makes caterpillars floppy-like domain; NH₃, protein N-terminus; RID, rho GTPase inactivation domain; RRSP, Ras/Rap1-specific endopeptidase domain; WT, wild type.

Figure 2.

Individual multifunctional autoprocessing repeats-in-toxins (MARTX) toxin effector domains modulate Vibrio vulnificus time to death. Mice (n = 12–15) were intragastrically infected with 6 × 10⁶–8 × 10⁶ colony-forming units of the indicated strains and monitored for survival over 48 hours. Kaplan-Meier plots were compared by the log-rank test, and P values are specified when difference were statistically significant. Note that the same wild-type (WT) data are displayed in multiple graphs, which are separated for visual clarity.

Figure 3.

No single multifunctional autoprocessing repeats-in-toxins (MARTX) toxin effector domain is required for bacterial survival or dissemination. Numbers of colony-forming units (CFU) of indicated bacterial strains recovered from individual mice (n = 5–6). CFU were quantified 6 hours after intragastric inoculation, by selective plating of organ homogenate from the intestine (A) or liver and spleen (B). Data were analyzed by 1-way analysis of variance with the Tukey post hoc test for comparisons of each strain to the wild-type (WT) strain. The dotted line indicates the detection limit of 500 CFU per organ (for the intestine) or 100 CFU per organ (for the liver/spleen). *P < .05.

Figure 4.

Multifunctional autoprocessing repeats-in-toxins (MARTX) toxin effector domain region is required for inhibition of phagocytosis. A, J774 cells do not lyse under the tested conditions. B and C, The phagocytic activity of J774 cells was tested following exposure to the indicated Vibrio vulnificus strains (B) or proteins (C) to test the necessity and sufficiency of MARTX toxin effector domains. Data are pooled from 2 independent experiments. Results are normalized relative to ΔrtxA1. Bars represent means ± standard deviations. Each experiment was conducted using biological triplicate samples. Data were analyzed by 1-way analysis of variance with the Tukey post hoc test for multiple comparisons. LF_N, Bacillus anthracis lethal factor; PA, anthrax toxin–protective antigen; PBS, phosphate-buffered saline; RRSP, Ras/Rap1-specific endopeptidase; WT, wild type. *P < .05 and **P < .01.

Figure 5.

Hydropathic nature of key residue in membrane localization domain (MLD) L1 differs among Vibrio vulnificus strains. Alignment of the first 42 amino acids from the MLD of multifunctional autoprocessing repeats-in-toxins toxin RID domains. The arrow indicates residue 15 from the basic-hydrophobic motif that mediates plasma membrane localization. H1, helix 1; H2, helix 2; L1, loop 1.