Clostridium difficile TcdC protein binds four-stranded G-quadruplex structures

Abstract

Clostridium difficile infections are increasing worldwide due to emergence of virulent strains. Infections can result in diarrhea and potentially fatal pseudomembranous colitis. The main virulence factors of C. difficile are clostridial toxins TcdA and TcdB. Transcription of the toxins is positively regulated by the sigma factor TcdR. Negative regulation is believed to occur through TcdC, a proposed anti-sigma factor. Here, we describe the biochemical properties of TcdC to understand the mechanism of TcdC action. Bioinformatic analysis of the TcdC protein sequence predicted the presence of a hydrophobic stretch [amino acids (aa) 30-50], a potential dimerization domain (aa 90-130) and a C-terminal oligonucleotide-binding fold. Gel filtration chromatography of two truncated recombinant TcdC proteins (TcdCΔ1-89 and TcdCΔ1-130) showed that the domain between aa 90 and 130 is involved in dimerization. Binding of recombinant TcdC to single-stranded DNA was studied using a single-stranded Systematic Evolution of Ligands by Exponential enrichment approach. This involved specific binding of single-stranded DNA sequences from a pool of random oligonucleotides, as monitored by electrophoretic-mobility shift assay. Analysis of the oligonucleotides bound showed that the oligonucleotide-binding fold domain of TcdC can bind specifically to DNA folded into G-quadruplex structures containing repetitive guanine nucleotides forming a four-stranded structure. In summary, we provide evidence for DNA binding of TcdC, which suggests an alternative function for this proposed anti-sigma factor.

Figure 1.

Domain organization of C. difficile TcdC. (A) Schematic presentation of TcdC with three predicted domains; Hyd = hydrophobic membrane anchor, Dim = dimerization domain and OB-fold = conserved C-terminal domain containing a predicted oligonucleotide-binding fold. Bottom, spatial structure of the TcdC–C-terminal domain (aa 90–232) predicted by the automated simulation method I-TASSER (23) and represented with a ribbon diagram. N and C indicate amino-terminus and carboxy-terminus. (B) Limited proteolysis of TcdCΔ1-50 and TcdCΔ1-130. Histidine-tagged proteins were digested with chymotrypsin for 5, 30, 60 or 120 min, and the fragments were resolved by sodium dodecyl sulphate–polyacrylamide gel electrophoresis. Amino-terminal sequences were identified using an Edman degradation. Site of chymotrypsin cleavage (arrow 1) indicated in the primary aa sequence and TcdC predicted secondary structure elements. H = helix (red box), C = coil (blue line), S = sheet (pink arrow).

Figure 2.

Mobility shift assay of TcdC Δ1-89 selected binding sites. ssDNA selected at each round (see ‘Materials and Methods’) were used as probes in gel mobility analysis. The selection rounds are indicated. Shifted probes in round 2 and 3 were cut out and eluted as indicated and cloned after the last round.

Figure 3.

Binding characteristic of TcdC Δ1-89. (A) Binding to two selected clones (#2 and #5 corresponding to Table 3) was tested using 100 ng of TcdCΔ1-89 in a bandshift assay (shift indicated protein–DNA complex). (B) Determination of the dissociation constant of TcdCΔ1-89 for #5 HMW binding. Lanes 1–7 of the inset gel shows increasing concentrations of TcdCΔ1-89 3 ng, 6 ng, 12 ng, 25 ng, 50 ng and 100 ng. Band shift assays were performed as described under ‘Material and Methods’ section.

Figure 4.

Characterization of a TcdC-bound sequence element. (A) HMW is formed by intermolecular interactions, which are lost at 5 min 95°C in formamide. DNA was 5′-end labeled with ³²Pγ-phosphate for visualization using storage phosphor screen autoradiography. (B) Recognition of HMW product by ETC: a quadruplex-specific stain (see ‘Materials and Methods’ section). (C) Loading control of gel-samples loaded in panel B. Each DNA was 5′-end labeled with ³²Pγ-phosphate (see ‘Materials and Methods’ sections). (D) Point mutations affecting G-quadruplex formation. G-stretches and mutated positions are in bold underlined. Probes were ³²P-labelled and separated on a 10% polyacrylamide gel. Quadruplex is indicated.

Figure 5.

TcdC dimerization domain is required for DNA binding. TcdC dimer (Δ1-89) or monomer (Δ1-130) was incubated with oligo #5. Only the TcdC dimer was able to bind to the quadruplex DNA, as evidenced by the shifted DNA.