Burkholderia cenocepacia epigenetic regulator M.BceJIV simultaneously engages two DNA recognition sequences for methylation

Abstract

Burkholderia cenocepacia is an opportunistic and infective bacterium containing an orphan DNA methyltransferase (M.BceJIV) with roles in regulating gene expression and motility of the bacterium. M.BceJIV recognizes a GTWWAC motif (where W can be an adenine or a thymine) and methylates the N6 of the adenine at the fifth base position (GTWWAC). Here, we present a high-resolution crystal structure of M.BceJIV/DNA/sinefungin ternary complex and allied biochemical, computational, and thermodynamic analyses. Remarkably, the structure shows not one, but two DNA substrates bound to the M.BceJIV dimer, wherein each monomer contributes to the recognition of two recognition sequences. This unexpected mode of DNA binding and methylation has not been observed previously and sets a new precedent for a DNA methyltransferase. We also show that methylation at two recognition sequences occurs independently, and that GTWWAC motifs are enriched in intergenic regions of a strain of B. cenocepacia's genome. We further computationally assess the interactions underlying the affinities of different ligands (SAM, SAH, and sinefungin) for M.BceJIV, as a step towards developing selective inhibitors for limiting B. cenocepacia infection.

Figure 1.. Overall structure of M.BceJIVΔ29-DNA-Sinefungin complex.

M.BceJIV forms a homodimer complex with two DNAs and two Sinefungin molecules. Monomer A (magenta) receives the adenine to be methylated from DNA1 but its TRD reaches over to DNA2 and, conversely, monomer B (cyan) receives the adenine to be methylated from DNA2 but its TRD extends over to DNA1. Helices are labeled from αA to αH and strands are labeled as β1 to β8. Sinefungin is shown in a stick representation. The DNA substrate used for crystallization is shown and the recognition sequence is boxed and numbered.

Figure 2.. Structure of M.BceJIVΔ29 monomers bound to DNA and Sinefungin, and the electron density for DNA and Sinefungin molecules.

Structure of monomer A bound to DNA1 and Sinefungin_A (a) and of monomer B bound to DNA2 and Sinefungin_B (b) showing DNA distortions and the TRDs. Each monomer contributes to recognition of its own methylation-target DNA via its loop-45 through the minor groove and to the recognition of a separate DNA via its TRD through the major groove. 2Fo-Fc map contoured at 1.5 σ is shown for DNA1 and Sinefungin_A (c) and for DNA2 and Sinefungin_B (d). The recognition sequence is labeled.

Figure 3.. Molecular basis of target recognition.

(a), The first base pair (G1:C1) is specified by contacts with Gln207 of monomer A and Arg180 of TRD_B; the second base pair (T2:A2) makes contact with His205 of monomer A and van der Waals contacts with Arg180 of TRD_B; the third base pair (T3:A3) is specified by contacts with His205 and Arg203 of monomer A and van der Waals contacts with Leu182 and Phe183 of TRD_B; the fourth base pair (A4:T4) makes contacts with Arg203 and van der Waals contacts with Met136 of monomer A (from loop-45) and van der Waals contacts with Trp188 of TRD_B; at the fifth base pair (A5:T5), the thymine opposite the adenine to be methylated is specified by contacts with Gly137 of a short segment of monomer A (Met136-Gly137-Gly138 from loop-45), that embeds in place of the everted fifth adenine and sixth cytosines of the target strand, and van der Waals contacts with Trp188 of TRD_B; at the sixth base pair (C6:G6), the guanine opposite to the everted cytosine makes contacts with Gly137 of monomer A (from loop-45). (b), A close-up view of the flipped-out adenine that is accommodated in the catalytic cleft of monomer A and makes contacts with Asp59, Pro60 and Tyr62 of the conserved DPPY motif and with Tyr68 and Lys218 of monomer A. c, A close-up view of the flipped-out cytosine accommodated outside the catalytic cleft and makes contacts Asp148, His147, Thr106, Trp107, and Gln108 of monomer A. Monomer A is colored as magenta and the TRD_B is colored as cyan.

Figure 4.. Interactions of M.BceJIVΔ29 with Sinefungin and the flipped adenine.

A close-up view of specific M.BceJIVΔ29-Sinefungin interactions and residues shared between Sinefungin and the flipped-out adenine. Sinefungin is tightly bound within the MTase domain via extensive hydrogen bonding (dashed lines) and hydrophobic contacts.

Figure 5.. Addition of a second DNA methylation site in trans or in cis does not assist the methylation of the first DNA site.

(a) in trans experiment, HpaI digestion of 24 nM BamHI-linearized pUC18 containing an HpaI site only with increasing concentrations (0 nM to 50 nM) of a 19 mer DNA duplex in trans containing a M.BceJIV methylation motif and no HpaI site in the presence of 24 nM M.BceJIV and 20 μM SAM in a 50 s reaction. (b) in cis experiment, HpaI digestion of 24 nM BamHI-linearized pUC18 containing an HpaI site only (or both an HpaI site and another M.BceJIV methylation site in cis) with increasing reaction time (1 min to 5 min) in the presence of 24 nM or 48 nM M.BceJIV and 20 μM SAM.

Figure 6.. Distribution of GTWWAC motif sites across two strains of B. cenocepacia.

(a & b), Circos plot showing the genome-wide distribution of GTWWAC sites along the three chromosomes and a plasmid of B. cenocepacia J2315 and B. cenocepacia K56–2, respectively. (c & d), Density plot for the distribution of distances between each two neighboring GTWWAC sites in B. cenocepacia strains J2315 and K56–2. (e & f), Enrichment of GTWWAC in the intergenic regions (observed frequency vs. expected frequency) in the two strains.

Figure 7.. ITC analysis of M.BceJIVΔ29 with SAM, SAH and Sinefungin, and kinetics of M.BceJIV using a bioluminescence assay.

(a), ITC titration data for Sinefungin (left), SAM (middle) and SAH (right) with M.BceJIVΔ29. The equilibrium dissociation constants (K_D) were derived from the resulting binding isotherms. (b), Two-dimensional protein-ligand interaction diagrams generated using energy-minimized conformations of the crystal structure of M.BceJIVΔ29 bound to Sinefungin (left) and predicted structures of SAM and SAH in M.BceJIVΔ29 (middle and right, respectively) embedded in an explicit water environment. The Ligand Interaction script in Maestro (Schrödinger Inc., www.schrodinger.com) was used to generate these diagrams with default parameters. Among the protein residues within a 4 Å distance of the ligands, only those that interact with the ligands according to Maestro’s default parameters (see details in Table S2) are displayed. (c), Kinetics of SAH byproduct formation by varying concentrations of the methyl donor SAM. Summary of kinetic parameters are also shown.