Mutational analysis of the enzymatic domain of Clostridium difficile toxin B reveals novel inhibitors of the wild-type toxin

Abstract

Toxin B (TcdB), a major Clostridium difficile virulence factor, glucosylates and inactivates the small GTP-binding proteins Rho, Rac, and Cdc42. In the present study we provide evidence that enzymatically inactive fragments of the TcdB enzymatic domain are effective intracellular inhibitors of native TcdB. Site-directed and deletion mutants of the TcdB enzymatic region (residues 1 to 556), lacking receptor binding and cell entry domains, were analyzed for attenuation of glucosyltransferase and glucosylhydrolase activity. Five of six derivatives from TcdB(1-556) were found to be devoid of enzymatic activity. In order to facilitate cell entry, mutants were genetically fused to lfn, which encodes the protective antigen binding region of anthrax toxin lethal factor and mediates the cell entry of heterologous proteins. In line with reduced enzymatic activity, the mutants also lacked cytotoxicity. Remarkably, pretreatment or cotreatment of cells with four of the mutants provided protection against the cytotoxic effects of native TcdB. Furthermore, a CHO cell line expressing enzymatically active TcdB(1-556) was also protected by the mutant-derived inhibitors, suggesting that inhibition occurred at an intracellular location. Protection also was afforded by the inhibitor to cells treated with Clostridium sordellii lethal toxin (TcsL), which uses the same cosubstrate as TcdB but shares Rac only as a common substrate target. Finally, the inhibitor did not provide protection against Clostridium novyi alpha-toxin (Tcnalpha), which shares similar substrates with TcdB yet uses a different cosubstrate. This is the first report to demonstrate that the potential exists to inhibit toxins at their intracellular site of action by using inactive mutants.

FIG. 1.

LFnTcdB deletion and site-directed mutants used in this study. (A) Overview of deletion and site-directed mutants. Deletion mutants were generated by PCR, cloned in frame with lfn in pET15b, expressed in E. coli BL-21, and subsequently purified by Ni²⁺ affinity chromatography. Site-directed mutants were generated by the QuickChange method, by using complementary mutation-carrying oligonucleotides and pLMS200 as template. (B) SDS-PAGE analysis of His-tagged fusions. Approximately 10 μg of each isolated fusion protein was subjected to SDS-PAGE analysis. Samples and corresponding lanes are labeled within the figure. MWM, molecular weight markers; weights are shown in thousands at right.

FIG. 2.

Glucosylation activity of deletion and site-directed mutants on RhoA, Rac1, and Cdc42 and in vitro inhibition of Rac glucosylation. (A) Each mutant and TcdB were tested for glucosylation activity on recombinant substrates GST-RhoA, GST-Rac1, and GST-Cdc42, with UDP-[¹⁴C]glucose as cosubstrate. Following a 2-h incubation, the reaction mix was resolved by SDS-PAGE and exposed to film for 48 h. Samples and corresponding lanes are labeled within the figure. (B) LFnTcdB^1-500 was included in a standard assay for TcdB-mediated glucosylation of Rac to determine if the deletion mutant attenuated modification of substrate. Following coincubation in a standard glucosylation assay, relative levels of [¹⁴C]glucose incorporation were determined by densitometry.

FIG. 3.

Inhibition of TcdB cytopathic effects by TcdB mutants. HeLa cells were cotreated with TcdB and each TcdB fusion plus PA. The cells were monitored for 7 h, and cytopathic effects were determined by visualization. (A) Representative samples of actin condensation and cell rounding in the inhibitor assay: A, phosphate-buffered saline control; B, TcdB control; C, PA-LFnTcdB^1-500 plus TcdB; D, PA-LFn plus TcdB. (B) Summary of inhibitors capable of blocking TcdB CPEs. Symbols: solid bars, LFnTcdB^1-420; open bars, LFnTcdB^W102A; lightly stippled bars, LFnTcdB^C365W; checkered bars, LFnTcdB^33-556; hatched bars, LFnTcdB^1-500.

FIG. 4.

Sustained inhibition by supplemental treatments with LFnTcdB^1-500-PA. HeLa cells were cotreated with TcdB and LFnTcdB^1-500-PA in the test assay. Control assays included TcdB in the absence of inhibitor or LFnTcdB^1-500 and PA in the absence of TcdB. In the test sample, LFnTcdB^1-500-PA was added to the cells at 1-h intervals for 12 h. The cells were then monitored for 30 h and visualized for CPEs. Samples and corresponding lines are labeled within the figure.

FIG. 5.

Protection of CHO cells expressing TcdB^1-556. GeneSwitch-CHOpGene/TcdB^1-556 cells were induced with mifepristone in the presence or absence of LFnTcdB^1-500 plus PA. Cells were then observed for rounding and CPEs at the indicated time points. Samples and corresponding lines are labeled within the figure.

FIG. 6.

TcdB^1-500 inhibition of TcsL CPEs. HeLa cells were treated with TcdB^1-500 plus PA for 30 min prior to treatment with TcsL. To enhance TcsL cytopathic activity, cells were treated with the toxin by using an acid pulse where cells were subjected to TcsL in acid medium (pH 4.0) for 10 min followed by replacement with neutral medium (pH 7.4) and TcdB^1-500 plus PA. The cells were amended with inhibitor every 30 min for 12 h and then monitored for 18 h to determine CPEs. Samples and corresponding lines are labeled within the figure.