Characterization of an actin-targeting ADP-ribosyltransferase from Aeromonas hydrophila

Abstract

The mono-ADP-ribosyltransferase (mART) toxins are contributing factors to a number of human diseases, including cholera, diphtheria, traveler's diarrhea, and whooping cough. VahC is a cytotoxic, actin-targeting mART from Aeromonas hydrophila PPD134/91. This bacterium is implicated primarily in diseases among freshwater fish species but also contributes to gastrointestinal and extraintestinal infections in humans. VahC was shown to ADP-ribosylate Arg-177 of actin, and the kinetic parameters were K(m)(NAD(+)) = 6 μM, K(m)(actin) = 24 μM, and k(cat) = 22 s(-1). VahC activity caused depolymerization of actin filaments, which induced caspase-mediated apoptosis in HeLa Tet-Off cells. Alanine-scanning mutagenesis of predicted catalytic residues showed the predicted loss of in vitro mART activity and cytotoxicity. Bioinformatic and kinetic analysis also identified three residues in the active site loop that were critical for the catalytic mechanism. A 1.9 Å crystal structure supported the proposed roles of these residues and their conserved nature among toxin homologues. Several small molecules were characterized as inhibitors of in vitro VahC mART activity and suramin was the best inhibitor (IC(50) = 20 μM). Inhibitor activity was also characterized against two other actin-targeting mART toxins. Notably, these inhibitors represent the first report of broad spectrum inhibition of actin-targeting mART toxins.

FIGURE 1.

Sequence alignment of actin-binding mART toxins. The multiple-sequence alignment was performed with T-Coffee (58), and the figure was prepared with ESPript (59) and modified manually. The fully conserved residues are shaded. Those residues with global similarity scores over 0.7 are boxed and were calculated using the Risler matrix method. Structural elements are based on the reported crystal structure of VahC. Residue numbering is according to the VahC primary sequence (top) and iota primary sequence (bottom). The active site loop and motifs 1–3 for ADPRT activity are shown in boldface type. Regions 1–6 (R1 to R6) are shown as potential actin binding regions in the sequences. The gray asterisks show the actin binding residues (from the Ia actin structure, PDB entry 3BUZ).

FIGURE 2.

VahC targets all actin isoforms and inhibits growth by ADP-ribosylation of Arg-177. A, the mART activity was assessed by incubating active VahC with α, β, and γ isoforms of actin and FITC-labeled NAD⁺. The reaction mixtures were separated on a 12.5% SDS-polyacrylamide gel, and FITC-ADP-ribosylated actin was visualized with UV light. B, VahC produced in S. cerevisiae causes a reduction in yeast growth, which is restored when expressed in the ACT-RA yeast strain, which ubiquitously expresses a mutant actin, R177A. C, Ala-scanning mutagenesis of the predicted catalytic signature was conducted, and the catalytic activity was characterized in vitro using an established fluorescence-based assay and expressed as a percentage of native VahC activity (black bars). The WT mART activity was 22.3 ± 2.0 s⁻¹ and is shown as 100% (black bars). Cell-based toxicity is expressed as a percentage of yeast growth (white bars). All mutants showed a marked decrease in enzyme activity, which corresponded to the restoration of yeast growth. Statistical analysis by one-way analysis of variance and Tukey's test revealed that all mutants were significantly different (p < 0.05) from controls in both catalytic activity and yeast cytotoxicity. Error bars, S.D.

FIGURE 3.

Intracellular expression and production of VahC cause actin cytoskeleton disruption and induced caspase-mediated apoptosis in mammalian cells. A, episomal expression of the vahC gene in HeLa Tet-Off cells. Shown is a Western blot on HeLa Tet-Off cells producing VahC and VahC E/E, vector alone, and the VIP-2 domain of VgrG1 using sera from mice immunized with VIP-2 from A. hydrophila ATCC 7966. B, disruption of actin cytoskeleton by expression of the vahC gene in the HeLa Tet-Off cell system. HeLa Tet-Off cells were transfected with the pBI-EGFP vector containing the vahC gene or the mutated vahC E/E gene. The vector alone and the vector carrying the VIP-2 domain of VgrG1 from A. hydrophila ATCC 7966 were stained with Alexa 568-conjugated phalloidin and mounted with a mounting solution containing DAPI. Images were acquired by fluorescence microscopy (×40 magnification). C, quantification of phalloidin staining by flow cytometry. HeLa Tet-Off cells expressing and producing VahC, VahC E/E, VIP-2 domain from VgrG1, and vector alone were stained with Alexa Fluor-568-conjugated phalloidin. Cells were acquired in a FACScan based on EGFP expression. The quadrants show the differences in percentage of cells positive for phalloidin in different groups. D, induction of apoptosis in HeLa Tet-Off cells by VahC. Cells expressing and producing VahC and VahC E/E, VIP-2 domain of VgrG1, and the vector alone were tested for rates of apoptosis using cytoplasmic nucleosome ELISA (top) and a colorimetric caspase-3 assay (bottom). Treatment with camptothecin was used as a positive control. Statistical differences were measured by one-way analysis of variance followed by Tukey's multiple comparison test. Error bars, S.D.

FIGURE 4.

Inhibition of VahC, Photox, and SpvB mART activity. A, structures of inhibitor compounds effective in reducing mART activity of VahC, Photox, and SpvB. B, IC₅₀ determinations of VahC by suramin (white circles), V31 (black circles), V7 (white squares), and V8 (black squares). Activity loss was assessed as described under “Experimental Procedures.” C, relative inhibition of Photox (gray), VahC (black), and SpvB (white). Activities were compared in the presence and absence of each inhibitor at its respective IC₅₀ concentration against VahC using reaction conditions described under “Experimental Procedures.” Error bars, S.D.

FIGURE 5.

High resolution (1.9 Å) crystal structure of VahC. A, crystal structure of VahC (PDB entry 4FML) shown as a schematic diagram. Secondary structural elements (α-helices; β, β-strands) are shown and numbered in succession. Adjacent to the VahC structure is an expanded view of the active site with the important binding/catalytic residues shown in stick format. Motif 1 residues are in violet, motif 2 residues are in orange, motif 3 residues are in magenta, and the proposed critical residues of the active site loop are in yellow. B, structural comparison of VahC (green) with SpvB (PDB entry 2GWM; yellow) based on an iterative three-dimensional alignment of protein backbone C_α atoms. C, structural comparison of VahC (green) with ι-toxin catalytic domain, Ia (PDB entry 1GIQ; magenta) based on an iterative three-dimensional alignment of protein backbone C_α atoms. The structural differences are highlighted by orange arrows.

FIGURE 6.

Electrostatic potential representation of the Type IV mART toxins that ADP-ribosylate Arg-177 in α-actin. A, surface potential of the Ia/C2I subtype members that include ι-toxin (Ia) (magenta ribbon and surface potential models) and VIP2 and C2I (PDB entries 1QS1 and 2J3Z, respectively; surface potential models only). The five regions (1–5) of iota believed to be involved in actin substrate recognition are shown as numbered circles (green). B, surface potentials of the SpvB/VahC subtype toxins are shown, and the iota C-terminal catalytic domain is also shown for comparison. Potential regions that may be involved in actin substrate recognition are also highlighted with green circles. These regions (Regions 2 and 5) correspond to those regions highlighted for Ia/C2I (above) except that Region 6 is unique to the SpvB/VahC subtype. The C/ denotes the catalytic subunit of the toxin only. Molecular surfaces are colored by the relative electrostatic potential (red, negative or acidic; blue, basic or positive). Surface potentials were calculated using PyMOL APBS software (60).