Purification and characterisation of a novel DNA methyltransferase, M.AhdI

Abstract

We have cloned the M and S genes of the restriction-modification (R-M) system AhdI and have purified the resulting methyltransferase to homogeneity. M.AhdI is found to form a 170 kDa tetrameric enzyme having a subunit stoichiometry M2S2 (where the M and S subunits are responsible for methylation and DNA sequence specificity, respectively). Sedimentation equilibrium experiments show that the tetrameric enzyme dissociates to form a heterodimer at low concentration, with K(d) approximately 2 microM. The intact (tetrameric) enzyme binds specifically to a 30 bp DNA duplex containing the AhdI recognition sequence GACN5GTC with high affinity (K(d) approximately 50 nM), but at low enzyme concentration the DNA binding activity is governed by the dissociation of the tetramer into dimers, leading to a sigmoidal DNA binding curve. In contrast, only non-specific binding is observed if the duplex lacks the recognition sequence. Methylation activity of the purified enzyme was assessed by its ability to prevent restriction by the cognate endonuclease. The subunit structure of the M.AhdI methyltransferase resembles that of type I MTases, in contrast to the R.AhdI endonuclease which is typical of type II systems. AhdI appears to be a novel R-M system with properties intermediate between simple type II systems and more complex type I systems, and may represent an intermediate in the evolution of R-M systems.

Figure 1

Purification of M.AhdI. SDS gel electrophoresis showing the purification of the methylase complex. Lane 1, molecular weight marker; lane 2, mixed cell lysates; lane 3, insoluble fraction; lane 4, soluble fraction; lane 5, supernatant after ammonium sulphate precipitation. Lanes 6–8, heparin column; lane 6, sample loaded; lane 7, run-through; lane 8, pooled fractions. Lanes 9–11, gel filtration; lane 9, sample loaded; lane 10, methylase peak; lane 11, S subunit peak.

Figure 2

(a) Dynamic light scattering of the M.AhdI enzyme at a concentration of 4 µM. (b) Rayleigh light scattering data obtained from analysis of the elution profile of the M.AhdI enzyme from a Superose 12 size exclusion column.

Figure 3

Sedimentation equilibrium. The top graph shows the sedimentation equilibrium data for the methyltransferase at a series of different concentrations and rotor speeds. Each data set has been fitted with a global model to describe the data and is shown by the black line, with the residuals to the fit plotted below. The model is of two heterodimers of 85 kDa self-associating to form a tetramer of 170 kDa.

Figure 4

Sedimentation velocity. Sedimentation velocity data were recorded for a 1.2 µM methyltransferase solution. The data were analysed to give a sedimentation coefficient distribution curve (black line) and were fitted to a two species model (S = 4.57 and 7.82). The sum of the two species is shown by the red line and an offset baseline shown in blue.

Figure 5

Gel retardation assay. Increasing concentrations of M.AhdI (0–6 µM, in 0.4 µM steps) were incubated with a 30 bp duplex probe containing (a) the AhdI recognition sequence and (b) a non-specific sequence, both at a concentration of 1 µM, prior to electrophoresis on a native 6% acrylamide gel. (c) The fraction of bound DNA from experiment (a) is plotted against the total M.AhdI concentration (expressed as monomer). The data were then fitted to a model that allows for dissociation of the tetramer into dimers (see text). The fitted curve corresponds to values of K₁ = 1 µM and K₂ = 50 nM.

Figure 6

Methylation assay. (a) Protection from R.AhdI endonuclease indicates methylation of the plasmid by M.AhdI. Lane 1, 1 kb marker; lane 2, uncut plasmid; lane 3, linearised plasmid; lane 4, linearised plasmid cut with R.AhdI; lane 5, linearised plasmid challenged with R.AhdI after incubation with heat killed M.AhdI; lane 6, linearised plasmid challenged with R.AhdI after incubation with native M.AhdI; lanes 7–30, time course of methylation with samples taken every 5 min. (b) Enzyme progression curve, showing the concentration of unmethylated DNA (nM) against incubation time with the M.AhdI enzyme. The data have been fitted to the integrated form of the Michaelis–Menten equation.

Figure 7

(a) Sequence alignment of residues 168–187 of the S subunit of AhdI and the consensus sequence found at the start of the central conserved region of all type I S subunits. (b) Comparison of the sequences of the S subunits of a type I MTase (EcoR124I) and AhdI. Conserved sequences (darker colours) are repeated in type I S subunits and include a limited region of sequence homology shared by all type I MTases. TRDs (lighter colours) responsible for recognition of DNA half-sites are unrelated in sequence. (c) Circular model for the subunit/domain structure of type I enzymes of the form M₂S₁ [adapted from Kneale (6)], showing the approximate two-fold symmetrical disposition of subunits and the orientation of the enzyme on the DNA. (d) Proposed symmetrical model (M₂S₂) for M.AhdI.