Clostridioides difficile specific DNA adenine methyltransferase CamA squeezes and flips adenine out of DNA helix

Abstract

Clostridioides difficile infections are an urgent medical problem. The newly discovered C. difficile adenine methyltransferase A (CamA) is specified by all C. difficile genomes sequenced to date (>300), but is rare among other bacteria. CamA is an orphan methyltransferase, unassociated with a restriction endonuclease. CamA-mediated methylation at CAAAAA is required for normal sporulation, biofilm formation, and intestinal colonization by C. difficile. We characterized CamA kinetic parameters, and determined its structure bound to DNA containing the recognition sequence. CamA contains an N-terminal domain for catalyzing methyl transfer, and a C-terminal DNA recognition domain. Major and minor groove DNA contacts in the recognition site involve base-specific hydrogen bonds, van der Waals contacts and the Watson-Crick pairing of a rearranged A:T base pair. These provide sufficient sequence discrimination to ensure high specificity. Finally, the surprisingly weak binding of the methyl donor S-adenosyl-L-methionine (SAM) might provide avenues for inhibiting CamA activity using SAM analogs.

Fig. 1. Activity of CamA.

a, b The formation of byproduct SAH was measured in a bioluminescence assay, by varying concentrations of methyl donor SAM (a) or substrate DNA (b) (N = 2). The dependence of the velocity of product SAH formation per enzyme molecule [SAH]/[E] on substrate concentration was analyzed according to the Michaelis–Menten equation. c Summary of CamA kinetic parameters. The DNA substrate used is also shown. Data represent the mean ± SD of two independent determinations, with duplicates assayed for each of the two determinations (N = 2). Source data are provided as a Source Data file. d ITC measurements of dissociation constants (K_D) and stoichiometry of CamA for SAM, SAH, and sinefungin, with N number of independent determinations (N = 3 for SAH, N = 2 each for SAM and sinefungin) (Supplementary Fig. 2). e Influence of cofactor on CamA-DNA binding dissociation (two independent determinations N = 2; Supplementary Fig. 3b, c).

Fig. 2. Structure of CamA-DNA complex.

a DNA oligos used for co-crystallization are connected by two different joints in the crystals (T:T mismatch linking complexes B and C, and T:A in a Hoogsteen base pair linking complexes A and B). The CamA recognition sites (CAAAAA) are indicated by arrows from 5′ to 3′, and the three target adenines are each separated by ~1.5 turns of DNA (14 bp). b Three CamA-DNA complexes were formed in the crystallographic asymmetric unit. c Omit electron density map (contoured at 4.5 σ above the mean) for the DNA molecule of complex B. The base-flipped adenine is visible. d Schematic illustration of CamA with relative locations of motifs I and IV and TRD. Residue 320 before αK indicates the linker point between the two domains. e, f Two orthogonal views of CamA, colored in spectrum from blue (N-terminus) to red (C-terminus). g The N-terminal catalytic domain folds into a nine-stranded sheet. Loop-1 follows strand β4 and loop-2 is between strands β6 and β7. h Antiparallel β-structure of C-terminal TRD. i Electrostatic surface of CamA with blue for positive and red for negative charges. j Seven loops (L1–L7) line the basic surface of the cleft for DNA binding.

Fig. 3. CamA-bound DNA conformation.

a The 6-bp recognition sequence is numbered from 1 to 6. B-form DNA (gray) is superimposed with CamA-bound DNA molecule (orange). Note the flipped-out A6 and rearranged base pairs between A5 and T6. b-c Conformational deviations of CamA-bound DNA molecule from B-form DNA. d Conformational differences among the six base pairs of the recognition sequence. e Discontinuous π-stacking between T5 and T6. f Inter-phosphate distances of CamA-bound DNA with the largest variations from three phosphate groups 5′ and 3’ respectively to the target adenine (P₋₃ and P₃). g Lys173 sits in the center of equilateral triangle of phosphate groups P₋₁, P₁, and P₂. h Under the single turnover conditions where the enzyme is present at or above the DNA substrate concentration, only one methylation event occurs (N = 2). Enzyme concentrations used are within a linear range (left). Source data are provided as a Source Data file. i Effects of single-base pair substitutions. No activity was observed for ssDNA, using the unsubstituted and unannealed strands separately. Data represent the mean ± SD of N number of independent determinations (N = 5 for control, N = 4 each for C1G, A2G, A3G, A4G, A6G, and G7A substitutions, N = 6 each for A5G and No SAM control, and N = 2 each for A6G/G7A and ssDNA substrates) performed in duplicate. Source data are provided as a Source Data file. j Schematic illustration of CamA-DNA interactions (for an enlarged version, see Supplementary Fig. 6b): residues in cyan background are from the N-terminal catalytic domain, and residues in (light and dark) orange from the C-terminal TRD. The base-specific contacts are placed between the two strands and the phosphate contacts are depicted above or below the strand. mc, main-chain-atom-mediated contacts.

Fig. 4. CamA-mediated base-specific recognition.

a Interactions with C1:G1 base pair. Interatomic distances are shown in angstroms. b–d Interactions with A2:T2, A3:T3, and A4:T4 base pairs at both major (right) and minor (left) grooves. e Interactions with the orphan T5. f View from the DNA minor groove, with intercalation of TRD residues that occupy the space normally occupied by A5. g Interactions with the rearranged A5:T6 base pair. h Lack of protein-mediated contacts with G7:C7 base pair. i The flipped-out A6 bound in the active-site cage. The nearby SAM-binding pocket is empty in the current structure. j Interaction with the target A6 in the aromatic cage. k, l A modeled SAM molecule in the cofactor binding pocket. Label S27 indicates the first ordered residue in the current structure.

Fig. 5. Cofactor-induced conformational change.

a The ternary structure of CamA-DNA-SAH (in stick model). The ordered N-terminal residues are in magenta. b The omit electron density map (contoured at 5.0σ above the mean) for the bound SAH. c Surface presentation of the ternary complex of SAH (in green and magenta), DNA (in orange) and SAH. d The enlarged binding pocket of SAH, where the edge of adenine moiety and the ribose hydroxyl groups were visible. e Intramolecular interactions between the N-terminal residues (in magenta) and the loop prior to helix αE. f, g Interactions with SAH adenine moiety. h Interactions with SAH ribose moiety. i Interactions involving residues of motif I. Note that Phe200 is behind SAH and away from the viewer. j Interaction with SAH aminocarboxypropyl moiety. k Interactions between the methyl donor (sulfur atom of SAH) and methyl acceptor (N6 of target Ade). l Conformational change of Lys25 and Lys173. m Superimposition of active sites of complexes B, in the SAH-free (PDB 7LNI) and SAH-bound (PDB 7LT5) states. The bound target adenine is sandwiched between two tyrosine residues, with Tyr30 undergoing a small rotation. n The largest movement is Gly28 and Tyr31, which move from an open to the closed conformation upon SAH binding.

Fig. 6. Comparison of three orphan adenine methyltransferases.

a Schematic of three Class I groups of amino-MTases, showing altered orders of motifs responsible for SAM binding (motif I), catalysis (motif IV) and target DNA substrate binding (TRD). b Sequence alignment of motif I and motif IV for the three orphan adenine MTases. c–e Enzyme-bound DNA conformations in Dam (c), CcrM (d), and CamA (e). f Dam interacts with guanine G4 of the non-target strand. g CcrM interacts with guanine G1 of the target strand. The lowercase m in black circle is the methylated adenine of the parental strand immediately after DNA replication. h CamA interacts with guanine G1 of the non-target strand. The underlined A in red in each case is the methylation target.