ABSTRACT
The type I DNA methyltransferase M.
Eco
R124I consists of two methylation subunits (HsdM) and one DNA recognition
subunit (HsdS). When expressed independently, HsdS is insoluble, but this
subunit can be obtained in soluble form as a GST fusion protein. We show that
the HsdS subunit, even as a fusion protein, is unable to form a discrete
complex with its DNA recognition sequence. When HsdM is added to the HsdS
fusion protein, discrete complexes are formed but these are unable to methylate
DNA. The two complexes formed correspond to species with one or two copies of
the HsdM subunit, indicating that blocking the N-terminus of HsdS affects one of the HsdM binding sites. However, removal
of the GST moiety from such complexes results in tight and specific DNA binding
and restores full methylation activity. The results clearly demonstrate the
importance of the HsdM subunit for DNA binding, in addition to its catalytic
role in the methyltransferase reaction.
Type I DNA methyltransferases are complex multisubunit enzymes which bind to a
target recognition sequence with high affinity and methylate a specific base
within this sequence. The host DNA is thereby protected from restriction by the
corresponding endonuclease (
1
,
2
). The target recognition sequence for type I systems is asymmetrical,
consisting of two half-sites 3-5 bp in length, separated by a non-specific `spacer' sequence of 6-8 bp. Methylation occurs at the N6 of two specific
adenines on opposite strands, one within each half-site of the DNA recognition sequence. Type I restriction-modification (R-M) systems are encoded by three genes, encoding the
subunits HsdS, HsdM and HsdR (responsible for specificity, methylation and
restriction respectively). For methylation of the target sequence, only the
HsdS and HsdM subunits are required (
3
).
Sequence comparisons of type I R-M systems have shown that the HsdS subunit
consists of two highly variable domains of 150-180 amino acid residues and two or more regions that are well conserved
within a given family (
4
-
6
). The two variable regions of the HsdS subunit form independent target
recognition domains (TRDs), each being responsible for recognition of one half
of the bipartite DNA recognition sequence (
7
,
8
). Based on analysis of repeated sequences in HsdS, we have proposed a circular
model for the domain organization in HsdS, driven by the interaction with HsdM,
which thus locates the DNA binding domains of the specificity subunit (
9
).
The
Eco
R124I methyltransferase (M.
Eco
R124I) is amongst the best studied of all type I enzymes at the molecular level.
It consists of two copies of the HsdM subunit (each 58 kDa) and one HsdS
subunit (46 kDa), to form a trimeric enzyme (162 kDa) with a subunit
stoichiometry of M
2
S
1
(
10
). We have shown for M.
Eco
R124I that DNA binding confers considerable protection from proteolysis (
11
). Likewise, chemical modification experiments have shown that a large fraction
of the lysine residues on the surface of the protein are inaccessible in the
DNA-protein complex, including a number of sites in regions implicated in
intersubunit contacts, as well as likely DNA binding residues (
12
). These results all indicate that the methyltransferase undergoes a significant
conformational change when it binds to DNA.
X-Ray solution scattering and circular dichroism have been used to determine
the structural parameters of M.
Eco
R124I and its complex with DNA (
13
). A dramatic reduction is observed in the overall dimensions of the enzyme
following DNA binding, which was proposed to involve a large rotation of the
HsdM subunits to clamp the DNA, resulting in additional non-sequence-specific interactions outside the DNA recognition sequence. The
circular dichroism spectrum shows that this structural transition in the enzyme
is accompanied by considerable distortion of the DNA structure in the complex.
The DNA recognition subunit (HsdS) of M.
Eco
R124I is completely insoluble when expressed independently from a variety of
expression vectors, in contrast to the methylation subunit (HsdM), which is
highly soluble even at very high levels of expression (
14
). Studies to date have therefore been conducted entirely with the intact
methyltransferase. Expression as fusion proteins can frequently improve the
solubility of insoluble proteins (
15
,
16
). In this paper we adopt such an approach to allow us to investigate the DNA
binding properties of the individual subunits of M.
Eco
R124I.
JM109 cells containing the plasmid pHSDS124A (
17
)
were grown at 25oC until an OD
600
of 0.6 was obtained. The cells were then induced with 0.1 mM IPTG and grown for
4 h. The cells were harvested by centrifugation at 2000
g
at 4oC for 15 min. The supernatant was removed and the cells dried and frozen at
-20oC until required.
The cell pellets were resuspended in 0.3 M TNE supplemented with 0.1 mM PMSF, 1
mM benzamidine, 2 mM DTT and 1 mg/ml lysozyme. The resuspended cells were
frozen in liquid nitrogen and allowed to thaw slowly. Once thawed, the samples
were disrupted by sonication at 0oC. The insoluble debris was removed by centrifugation at 40 000
g
. PEI-cellulose (5 g) was added to the supernatant and the suspension stirred slowly
for 1 h at 4oC. The bound nucleic acids were removed by centrifugation at 5000
g
in a swing out rotor at 4oC. The supernatant was removed and stored awaiting additional purification.
GST fusion extracts were purified using a 10 ml column containing glutathione
covalently attached to a Sepharose 6B matrix (Pharmacia). The crude preparation
was loaded onto the column in buffer A (0.15 M TNE, 2 mM DTT). Subsequently,
the bound protein was eluted by the addition of buffer B (0.15 M TNE, 10 mM
glutathione, 2 mM DTT). Fractions were analysed by SDS-PAGE, those containing GST-HsdS being pooled for further purification.
The pooled fractions were further purified using a 5 ml Econopac heparin
cartridge (BioRad). The sample was loaded in buffer A (0.15 M TNE, 2 mM DTT)
and the bound protein subsequently eluted with a linear gradient of buffer B (1
M TNE, 2 mM DTT). The GST-HsdS was passed down a Q ion exchange column for final purification and
eluted in the same manner as described for the heparin chromatography. The
samples were then concentrated and dialysed into a buffer containing 50%
glycerol, 150 mM NaCl, 50 mM Tris-HCl, pH 8.2, 5 mM DTT. Protein concentrations were estimated by UV
spectroscopy.
Aliquots of 1 nmol 30 bp DNA were mixed with 50 [mu]Ci [[gamma]-
32
P]ATP (6000 Ci/mmol) and 50 U T4 polynucleotide kinase (New England Biolabs) in
25 [mu]l kinase buffer (50 mM Tris-HCl, pH 7.6, 10 mM MgCl
2
, 5 mM DTT, 10 mM spermidine) and incubated for 30 min at 37oC. The solution was made up to 400 [mu]M with non-labelled ATP and incubated for an additional 30 min at 37oC. The enzyme was heat inactivated at 68oC and the solution made up to 0.15 M NaCl. The
unincorporated ATP was separated from the labelled oligonucleotides using a
Nuctrap push column (Stratagene).
The reconstituted methyltransferase was incubated overnight at 37oC with the radiolabelled 30 bp fragment (100 nM DNA, 12.5 nM protein) in
binding buffer (50 mM Tris-HCl, pH 7.8, 10 mM MgCl
2
, 1 mM DTT) supplemented with 1 mM AdoMet. Identical experiments were carried
out with the native methyltransferase and the GST-HsdS-HsdM complex and, as controls, with GST-HsdS, HsdM and cleaved GST-HsdS respectively, under the same conditions.
Samples were heat inactivated, made up in
Eco
RI reaction buffer (50 mM NaCl, 100 mM Tris-HCl, 10 mM MgCl
2
, 0.025% Triton X-100, pH 7.5, at 25oC) and digested with
Eco
RI. After this time, the reactions were again heat inactivated. Finally,
formamide loading dye was added to the samples, which were denatured by
heating, before loading onto a 24% sequencing gel. The gels were fixed in 10%
acetic acid, 50% methanol, then dried and the bands visualized by
autoradiography.
Gel retardation experiments were performed using non-denaturing gel electrophoresis. Proteins were incubated at various
concentrations with the radiolabelled 30 bp DNA duplex containing the M.
Eco
R124I recognition sequence and the complexes were allowed to form at 4oC for a period of 20 min. At this point the samples were loaded onto a 6%
non-denaturing acrylamide gel running at 4oC in 1* TAE buffer (40 mM Tris-acetate, pH 7.4, 1 mM EDTA). After an appropriate time
the gels were dried under vacuum and visualized by autoradiography.
HsdS was expressed as a fusion with glutathione
S
-transferase (GST) from the plasmid pHSDS124A (
17
). In order to achieve acceptable levels of expression and solubility of the
fusion protein, it was necessary to grow the bacterial cells at 25oC. At this temperature, despite relatively low levels of expression, the
majority of the fusion protein was found in the soluble fraction of the
bacterial lysate.
Purification of the fusion protein was achieved initially by single step
chromatography on a glutathione-Sepharose affinity column. However, it was necessary to follow this
procedure with additional purification using both Q-Sepharose and heparin in order to remove nucleic acid and contaminating
proteins that co-purify with the GST-HsdS fusion protein. The purified protein could be seen as
essentially a single ~65 kDa band by SDS-PAGE (Fig.
1
).
In previous experiments (
10
) we have shown that M.
Eco
R124I binds specifically to a 30 bp oligonucleotide containing its cognate DNA
recognition sequence with high affinity (
K
d
[approx]10
-8
M). The same oligonucleotide duplex was used to compare the binding affinity of
GST-HsdS before and after thrombin cleavage by gel retardation analysis.
Initially, binding reactions with GST-HsdS were performed at concentrations similar to those used for the
experiments with the native methylase, but these resulted in no observable
shift in the DNA. At higher concentrations of both protein and DNA, binding was
observed, but this was in the form of aggregated species that did not enter the
gel. At very high protein concentrations (a 20-fold excess of protein), there was a faint band observable that migrated
into the gel but this represented only a few per cent of the total (Fig.
2
). Similar experiments were performed where the GST was first cleaved from the
HsdS moiety by thrombin, before the addition of DNA, and in this case not even
the aggregated species was observed (data not shown). Cleavage of GST-HsdS
in situ
in the DNA binding reaction gave an identical result.
Since HsdS is soluble when co-expressed with HsdM during the preparation of wild-type M.
Eco
R124I (
14
), we decided to investigate the DNA binding properties of both GST-HsdS and HsdS in the presence of HsdM. The GST-HsdS fusion protein and the 30 bp DNA fragment were first combined
at a ratio of 4:1 and the effect of increasing amounts of HsdM on DNA binding
was assessed by gel retardation analysis. Three shifted complexes were observed
(Fig.
3
). One of these complexes remained at the top of the gel and represented an
aggregated species. In addition, two species migrated as discrete complexes in
the gel, the slower migrating species being predominant at the higher
concentrations of HsdM. This suggests that these two bands represent
nucleoprotein complexes with different numbers of HsdM subunits bound (see
Discussion).
Having shown that full DNA binding activity could be obtained from the HsdS-HsdM complex, we subsequently investigated the ability of this complex to
methylate its DNA recognition sequence. A synthetic oligonucleotide substrate
was constructed containing overlapping recognition sites for
Eco
RI and M.
Eco
R124I. The target adenine for methylation on one strand of the DNA duplex is
identical to that modified by the
Eco
RI methylase. Thus methylation of this base by an active M.
Eco
R124I methyltransferase will result in protection from cleavage by
Eco
RI, which can readily be assayed on a denaturing gel using an end-labelled substrate (see Fig.
5
).
Gel retardation analysis has shown that the GST-HsdS fusion protein is unable to bind to DNA to form a specific complex;
rather, it forms a high molecular weight complex which is unable to enter the
gel. However, this interaction is weak and requires a large excess of protein
(at stoichiometric ratios the DNA is almost entirely unbound). Presumably, this
is due to aggregation of the protein and/or multiple (non-specific) interactions with the DNA. When the GST moiety is cleaved from
HsdS, no binding at all can be observed, reflecting the total insolubility of
the native subunit.
However, cleavage of the GST-HsdS fusion protein prior to (or concomitant with) the addition of
stoichiometric quantities of HsdM gives rise to a multisubunit protein complex
with native DNA binding characteristics. High affinity binding to the DNA
recognition sequence is then observed and the mobility of the resulting DNA-protein complex on non-denaturing gels is identical to that formed by the native
methyltransferase.
The results differ from those of Kusiak
et al
. (
17
), who reported that DNA binding activity was possible with the HsdS subunit in
the absence of HsdM. However, the significance of these findings is unclear.
Although the concentrations of DNA and protein were not defined, the shifted
species represented only a very small fraction of the total DNA; thus the
binding appeared to be exceptionally weak, even though the gel was greatly
overexposed. Moreover, the mobility of the HsdS complex was identical to that
of intact methylase complex, despite enormous differences in molecular weight
(46 kDa compared with 162 kDa) and nor did the addition of HsdM to the HsdS
complex give rise to the expected `supershift'.
We are grateful to Dr David Hornby for freely providing the HsdS expression
plasmid (pHSDS124A) used in these experiments and to Damian Watts for technical
assistance.
REFERENCES
Return