The N-terminus of m5C-DNA methyltransferase
Msp
I is involved
in its topoisomerase activity
Sanjoy K. Bhattacharya* and Ashok K. Dubey†
Department of Biochemical Engineering and Biotechnology, Indian Institute of Technology-Delhi, New Delhi, India
DNA cytosine methyltransferase MspI(M.MspI) must
require a different type of interaction of protein with DNA
from other bacterial DNA cytosine methyltransferases
(m5C-MTases) to evoke the topoisomerase activity that it
possesses in addition to DNA-methylation ability. This may
require a different structural organization in the solution
phase from the reported consensus structural arrangement
for m5C-MTases. Limited proteolysis of M.MspI, however,
generates two peptide fragments, a large one (p26) and a
small one (p18), consistent with reported m5C-MTase
structures. Examination of the amino-acid sequence of
M.MspI revealed similarity to human topoisomerase I at the
N-terminus. Alignment of the amino-acid sequence of
M.MspI also uncovered similarity (residues 245–287) to the
active site of human DNA ligase I. To evaluate the role of the
N-terminus of M.MspI, 2-hydroxy-5-nitrobenzyl bromide
(HNBB) was used to truncate M.MspI between residues 34
and 35. The purified HNBB-truncated protein has a mo-
lecular mass of  45 kDa, retains DNA binding and meth-
yltransferase activity, but does not possess topoisomerase
activity. These findings were substantiated using a purified
recombinant MspI protein with the N-terminal 34 amino
acids deleted. Changing the N-terminal residues Trp34 and
Tyr74 to alanine results in abolition of the topoisomerase I
activity while the methyltransferase activity remains intact.

sequence homology and are postulated to be similar in
structure and function, there are important differences with
respect to their kinetic, biological and chemical properties.
M.Dcm and M.BsuRI, members of the m5C-MTase family,
undergo self-methylation in the absence of substrate DNA
[15,16]. M.MspI catalyzes the exchange of tritium at C-5 of
cytosine from tritiated water in the absence of cofactor
AdoMet, not common in many members of m5C-MTase
family [17]. M.MspIandM.SssI MTases possess a
topoisomerase activity not found in other m5C-MTases
[18]. M.MspI induces a significant sequence-specific bend of
105 ° in DNA, which has not been reported for any
other m5C-MTase [19]. Neither topoisomerase activity
(S. K. Bhattacharya, unpublished observation) nor tritium
exchange [17,20] in the absence of cofactor has been
detected in M.HpaII, an isoschizomer of M.MspI. M.MspI
is a 418-amino-acid protein with known DNA and amino-
acid sequence [21]. It has one of the largest N-terminal
sequences of any bacterial m5C-MTase. N-Terminal
sequence here refers to the amino-acid sequence before
motif I. The N-terminal sequence of M.MspI spans residues
1–107 [10]. Whether the presence of a large N-terminal
sequence results in modulation of conformation/structure
or activity in bacterial m5C-MTases such as M.MspIhas
not been investigated. The phenomenon of sequence-specific
Correspondence to S. K. Bhattacharya, Department of Ophthalmic
Research/I31, Cole Eye Institute, Cleveland Clinic Foundation,
9500 Euclid Avenue, Cleveland, OH 44195, USA.
Fax: + 1 216 297 9892, Tel.: + 1 216 445 0424,
E-mail:

q
-
(lacZ)
–
M15] used for overexpression of the target gene
was placed at the downstream region of the T7 promoter
regulated by the lac operator in the expression vector
pMSP [22]. E. coli strain ER 1727 harboring recombin-
ant plasmid pMSP was cultivated in Luria broth
containing 150 lgÆmL
)1
ampicillin at 37 °C. M.MspI
was purified using column chromatography as described
previously [23]. The purified M.MspI was tested for
homogeneity by SDS/PAGE (10% gel) using Coomassie
Blue R250 staining.
Assay of MTase activity
Reactions of MspI MTase were performed in 30 mL
buffer A (potassium phosphate, pH 7.4, sodium EDTA
1m
M
, 2-mercaptoethanol 14 m
M
, glycerol 10%) con-
taining unlabeled 1459-bp BstXI DNA fragment of
/X174 as substrate (50 n
M
DNA) and 200 n
M
AdoMet

Protein concentration was determined on the basis of
Bradford’s principle [24] using the Bio-Rad Coomassie plus
kit. Standard curves were established using BSA. Protein
samples were subjected SDS/PAGE (10% gels) using
Laemmli buffer [25]. Proteins were visualized by staining
with Coomassie Brilliant Blue R250.
Modification of M.
Msp
I with 2-hydroxy-5-nitrobenzyl
bromide (HNBB)
Modification reactions were carried out in 20 m
M
sodium
phosphate buffer (pH 10) at 25 °C. The enzyme was quickly
brought to pH 7.4 with buffer A [26]. HNBB (1–5 m
M
)was
addedto5l
M
MspI in a reaction volume of 40 mL. The
reaction mixture was analyzed for MTase activity after
HNBB treatment (HNBB is readily inactivated under these
conditions). A
410
was measured. An absorption coefficient
of 18 000
M
)1
Æcm
)1

M
Tris/HCl
(pH 8.0), 2.0 m
M
EDTA, 1 m
M
2-mercaptoethanol and
100 m
M
NaCl. The cleavage reaction was terminated by
adding 0.5 m
M
phenylmethanesulfonyl fluoride after
90 min of incubation at 37 °C. The fragments were purified
by FPLC column chromatography using Mono Q and
Mono S columns and a modification of a protocol for intact
M.MspI [23].
Peptide sequencing
After SDS/PAGE separation, peptides were transferred to
poly(vinylidene difluoride) membranes using a Novablot
multiphor semidried Western blotting apparatus. The first
20-amino-acid p26 and p18 band was sequenced using an
Applied Biosystems automated sequencer.
Assay of topoisomerase I activity
Topoisomerase I was assayed using CsCl-purified plasmid
pBR322 in topoisomerase assay buffer [50 m
M
Tris/HCl,
pH 7.5, 1 m
M

of 34 N-terminal amino acids; del34aa) was made using the
following set of primers: 5¢-CATATGgaatcaggtaaaaca-3¢
and 5¢-tgttttacctgattccCATATG-3¢ (forward and reverse
primers, respectively. These primers ensured introduction of
asitefortheNdeI enzyme, which has no recognition site in
the MspI gene sequence. The amplification was carried out
using Taq polymerase (Gibco/BRL) following the protocol
recommended by the manufacturer. The PCR product was
ligated in PGEMT vector using a PGEMT kit (Promega).
The amplified fragment was separated from the PGEMT
vector using NdeI and ligated with NdeI-digested pET3a
vector. The plasmid with the correct orientation of the
fragment was selected using restriction digestion; DNA
sequencing was subsequently performed to confirm that the
orientation was correct. The plasmid was introduced into
E. coli BL21 DE3 plysS and induced as described above to
produce the recombinant truncated protein. The truncated
protein was purified as described above.
RESULTS
Alignment of M.
Msp
I sequence with topoisomerase I
and DNA ligase
In a computer search with a basic local alignment search
tool [28] (BLAST network service at the National Center
for Biotechnology Information), no similarities were
detected between the MspI amino-acid sequence and
sequences of members of the topoisomerase, recombinase,
transpose or ligase families. Subsequently, the protein
sequences were obtained from GenBank and aligned using

Msp
I
Digestion of M.MspI with trypsin resulted in the generation
of two bands of molecular mass 26 kDa and 18 kDa,
designated p26 and p18, respectively (Figs 2A,B). The first
15 amino acids of fragment p26 are LKLIRSKLDLTQK
QA. The sequence obtained for the first 20 amino acids of
p18 is GIPQKRKRFYLVAFLNQNIH. This is in agree-
ment with the C-terminal portion of M.MspI that would
contain 166 amino acids. The molecular mass for this
fragment was calculated to be  17 kDa, which is in good
agreement with the experimental value of 18 kDa. Trypsin
digestion at two sites (between six and 251 amino acids) as
observed with N-terminal sequencing of the p26 and p18
fragments, the p26 fragment should contain 246 amino
acids which should correspond to an approximate molecu-
lar mass of 28 kDa consistent with the measured molecular
mass of this fragment.
2
Topoisomerase activity of HNBB-modified M.
Msp
I
Treatment with HNBB resulted in modification or trunca-
tion at tryptophan residues. There are only two tryptophan
residues in M.MspI, at positions 31 and 34, respectively.
Thus, modification of a tryptophan residue in M.MspI
should result in modification/truncation of the N-terminal
portion only. Indeed, on HNBB modification (50-fold
excess HNBB at pH 8.5), the N-terminal portion of the
protein is truncated, which was confirmed by SDS/PAGE

M.MspI control column; lane 2, M.MspI
treated with 50-fold excess of HNBB at
pH 8.5; lane 3, HNBB-treated M.MspI puri-
fied on a Q-Sepharose column; lane 4, HNBB-
treated M.MspI purified on an SP-Sepharose
column. (B) Topoisomerase activity of
HNBB-modified protein. Lane 1, M.MspI
control; lane 2, HNBB-treated protein; lane 3,
pBR 322 control; lane 4, EcoRI digest. (C)
MTase activity of M.MspI and HNBB-treated
protein. The amino acid residues 1–34 of
M.MspI are given below panel (A); the arrows
below W (tryptophan) indicate the site of
modification by HNBB.
2494 S. K. Bhattacharya and A. K. Dubey (Eur. J. Biochem. 269) Ó FEBS 2002
proteolytic fragments possessed either of these activities
(Fig. 4A,B).
Effect of mutation on enzyme activities
The purified mutant enzymes W34A and Y74A were
compared for topoisomerase and MTase activity. Both
possessed more than 80% MTase activity (Fig. 4C), but
both were completely inactive with respect to topoisom-
erase I/relaxation activity, even when 200 lg of the mutant
enzyme was used, a more than 20-fold excess over the
amount needed to obtain relaxation by the wild-type
enzyme(Fig. 4D).
Characterization of truncated
Msp
I (del34aa)
The truncated M.MspI protein (del 34aa) was purified to

with tyrosine mutant Y74A; lane 5, DNA control.
Ó FEBS 2002 N-Terminal of MspI and topoisomerization (Eur. J. Biochem. 269) 2495
interesting to note that the better matches were obtained
with eukaryotic than prokaryotic sequences. The best match
was with human DNA topoisomerase I and human DNA
ligase I. This better match with eukaryotic sequences,
human proteins in particular, has also been observed
for NaeI restriction endonuclease [31]. The amino-acid
sequence 245–287 of M.MspI shows similarity to the active-
site sequence of DNA ligases, which has also been found for
NaeI restriction. In NaeI [31], as well as in M.MspI, the
sequence with similarity around the active site of the DNA
ligase sequence differs from human ligase active site in one
important respect: the lysine that forms the adenylated
intermediate essential for catalysis by the DNA ligase active
site in NaeI has been replaced with a leucine (L43) at this
position, whereas in M.MspI it has been replaced with a
histidine (H271). Using a
BLAST
search, sequence similarity
was not observed between NaeI MTase and topoisomer-
ases. Changing L43 to K43, however, enables NaeI
restriction to possess topoisomerase activity [31]. The three
members of the restriction-modification family that show
topoisomerase activity (or with the potential to do so on
mutation of a single residue), NaeI restriction endonuclease,
MspIandSssI methylase, recognize 5¢-CGCCGGC-3¢,
5¢-CCGG-3¢ and 5¢-CG-3¢, respectively. The common
element in their recognition sequences is 5¢-CG-3¢.Itisalso
noteworthy that the regions in NaeI restriction endonuc-

of 85 residues of the N-terminal region does not affect either
the MTase activity or specific DNA binding in M.EcoRII,
which possesses one of the largest N-terminal sequences (98
residues) [36]. This is commensurate with our observation
that a HNBB-treated protein does retain MTase activity.
However, deletion of 97 amino acids in M.EcoRII resulted
in a decrease in enzyme activity. Further deletions caused
complete loss of activity. The N-terminus is a variable
region present in many prokaryotic DNA (cytosine-5)
methylases which plays no role in determining enzyme
specificity, although it does contribute to the interaction
with both AdoMet and DNA; it has been investigated in
detail for the EcoRII methylase [36].
ÔPromiscuous domainsÕ are widespread components of
many proteins; the fusions found may simply represent
permutations and combinations of a set of common
components and may not imply interactions [37]. Even
though the organisms that harbor these members of
restriction-modification (r-m) systems (R.NaeI, M.MspI
and solitary M.SssI) are capable of being human parasites,
it is unlikely that there could have been a fusion of two genes
conferring MTase and topoisomerase activity in which the
topoisomerase/ligase part was derived from a eukaryotic
counterpart. Equally intriguing is the fact that topoisomer-
ase activity is associated with the methylase/endonuclease
with a 5¢-CG-3¢ in the recognition sequence. In the light of
the similarities at the amino-acid level observed between
MspI and ligase and topoisomerase, and the experimental
observations presented here, it is debatable whether it was a
single progenitor protein with different regions that diverged

Cell 69, 915–926.
5. Schlappi, M., Raina, R. & Fedoroff, N. (1994) Epigenetic reg-
ulation of the maize Spm transposable element: novel activation of
a methylated promoter by TnpA. Cell 77, 427–437.
6. Engler, P., Weng, A. & Storb, U. (1993) Influence of CpG me-
thylationandtargetspacingonV(D)Jrecombinationinatrans-
genic substrate. Mol. Cell. Biol. 13, 571–577.
2496 S. K. Bhattacharya and A. K. Dubey (Eur. J. Biochem. 269) Ó FEBS 2002
7. Gartler, S.N. & Riggs, A.D. (1983) Mammalian X-chromosome
inactivation. Annu. Rev. Genet. 17, 155–190.
8. Li, E., Beard, C. & Jaenisch, R. (1993) Role for DNA methylation
in genomic imprinting. Nature 266, 362–365.
9. Posfai, J., Bhagwat, A.S., Posfai, G. & Roberts, R.J. (1989) Pre-
dictive motifs derived from cytosine methyltransferases. Nucleic
Acids Res. 17, 2421–2435.
10.Kumar,S.,Chang,X.,Klimasauskas,S.,Mi,S.,Posfai,J.,
Roberts, R.J. & Wilson, G.G. (1994) The DNA (cytosine-5)
methyltransferases. Nucleic Acids Res. 22, 1–10.
11. Adams, R.L.P. (1995) Eukaryotic DNA methyltransferases:
structure and function. Bioessays 17, 139–145.
12. Yoder, J.A., Yen, R.W.C., Vertino, P.M., Bestor, T.H. & Baylin,
S.B. (1996) New 5¢ regions of the murine and human genes for
DNA (cytosine-5)-methyltransferase. J. Biol. Chem. 271, 31092–
31097.
13. Liu, Y., Oakeley, E.J., Sun, L. & Jost, J.P. (1998) Multiple
domains are involved in the targeting of the mouse DNA
methyltransferase to the DNA replication foci. Nucleic Acids Res.
26, 1038–1045.
14. Chuang, S H.L., Ng, H H., Chia, J N. & Li, B.F.L. (1996)
Characterisation of independent DNA and multiple Zn-binding

23. Dubey, A.K., Mollet, B. & Roberts, R.J. (1992) Purification and
characterization of the MspI DNA methyltransferase cloned and
overexpressed in E. coli. Nucleic Acids Res. 20, 1579–1585.
24. Bradford, M.M. (1976) A rapid and sensitive method for the
quantitation of microgram quantities of protein utilizing the
principle of protein-dye binding. Anal. Biochem. 72, 248–254.
25. Jost, J.P. & Saluz, H.P. (1991) A Laboratory Guide to in Vitro
Studies of Protein–DNA Interactions. Birkhauser-Verlag, Berlin.
26. Bhattacharya, S.K. & Dubey, A.K. (2000) Chemical modification
studies with MspI DNA methyltransferase. J. Biochem. Mol. Biol.
Biophys. 4, 109–120.
27.Means,G.E.&Feeney,R.E.(1971)Chemical Modification of
Proteins. Holden-Day, Inc, San Francisco.
28. Altschul, S.F., Gish, W., Miller, W., Myers, E.W. & Lipman, D.J.
(1990) Basic local alignment search tool. J. Mol. Biol. 215,
403–410.
29. D’Arpa, P., Machlin, P.S., Ratrie, H. III, Rothfield, N.F.,
Cleveland, D.W. & Earnshaw, W.C. (1988) cDNA cloning of
human DNA topoisomerase I: catalytic activity of a 67.7-kDa
carboxyl-terminal fragment. Proc.NatlAcad.Sci.USA85, 2543–
2547.
30. Barnes, D.E., Johnston, L.H., Kodama, K., Tomkinson, A.E.,
Lasko, D.D. & Lindahl, T. (1990) Human DNA ligase I cDNA:
cloning and functional expression in Saccharomyces cerevisiae.
Proc. Natl Acad. Sci. USA. 87, 6679–6683.
31. Jo, K. & Topal, M.D. (1995) DNA topoisomerase and re-
combinase activities in NaeI restriction endonuclease. Science 267,
1817–1820.
32. Cheng, X., Kumar, S., Posfai, J., Pflugarth, J., W. & Roberts, R.J.
(1993) Crystal structure of the HhaI DNA methyltransferase