Significance of the conserved amino acid sequence for human MTH1 protein with
antimutator activity
Significance of the conserved amino acid sequence for human MTH1 protein with antimutator activity
Jian-Ping
Cai
1,+
,
Hisaya
Kawate
1
,
Kenji
Ihara
1
,
Hiroyuki
Yakushiji
1
,
Yusaku
Nakabeppu
1
,
Teruhisa
Tsuzuki
1
and
Mutsuo
Sekiguchi
1,2,
*
1
Department of Biochemistry, Medical Institute of Bioregulation, Kyushu
University,
Fukuoka
812-82,
Japan
and
2
Department of Biology, Fukuoka Dental College,
Fukuoka
814-01,
Japan
Received December 12, 1996;
Accepted January 24, 1997
ABSTRACT
8-Oxo-7,8-dihydro-2
'
-deoxyguanosine 5
'
-triphosphate (8-oxo-dGTP) is produced during normal cellular metabolism, and
incorporation into DNA causes transversion mutation. Organisms possess an
enzyme, 8-oxo-dGTPase, which catalyzes the hydrolysis of 8-oxo-dGTP to the corresponding nucleoside monophosphate,
thereby preventing the occurrence of mutation. There are highly conserved amino
acid sequences in prokaryotic and eukaryotic proteins containing this and
related enzyme activities. To elucidate the significance of the conserved
sequence, amino acid substitutions were introduced by site- directed mutagenesis of the cloned cDNA for human 8-oxo-dGTPase, and the activity and stability of mutant forms of the
enzyme were examined. When lysine-38 was replaced by other amino acids, all of the mutants isolated carried the 8-oxo-dGTPase-negative phenotype. 8-Oxo-dGTPase-positive revertants, isolated from one
of the negative mutants, carried the codon for lysine. Using the same
procedure, the analysis was extended to other residues within the conserved
sequence. At the glutamic acid-43,
arginine-51 and glutamic acid-52 sites, all the positive revertants isolated carried codons for
amino acids identical to those of the wild type protein. We propose that Lys-38, Glu-43, Arg-51 and Glu-52 residues in the conserved region are essential to
exert 8-oxo-dGTPase activity.
INTRODUCTION
Oxygen radicals are produced during cellular metabolism and damage biologically important macromolecules. More than 20 different types of oxidatively altered purines and pyrimidines have been detected in DNA
(
1
,
2
). Among them, an oxidized form of guanine base, 8-oxo-7,8-dihydroguanine (8-oxoguanine) appears to be important in mutagenesis and in carcinogenesis (
3
-
5
). This oxidized base can pair with cytosine and with adenine at almost equal
efficiencies and, as a result, A:T to C:G as well as G:C to T:A transversion
mutations are induced (
6
-
9
).
To counteract such mutagenic effects of 8-oxoguanine, organisms are equipped with elaborate mechanisms. In
Escherichia coli
, two glycosylases encoded by the
mutM
and the
mutY
genes function to prevent mutation caused by 8-oxoguanine in DNA. MutM protein removes 8-oxoguanine paired with cytosine while MutY protein removes adenine paired with 8-oxoguanine (
10
-
13
).
Enzyme activities similar to those of MutM and MutY in mammalian cells were identified
(
14
,
15
). A significant amount of 8-oxoguanine is formed in chromosomal DNA of mammalian cells, and most
damaged bases are excised and excreted into the urine (
16
).
Oxidation of guanine proceeds also in forms of free nucleotides, and an oxidized form of dGTP, 8-oxo-dGTP, is a potent mutagenic substrate for DNA synthesis (
17
). Organisms possess a mechanism to prevent mutation due to misincorporation of 8-oxo-dGTP (
18
). MutT protein of
E.coli
hydrolyzes 8-oxo-dGTP to the monophosphate (
17
), and lack of the
mutT
gene increases the occurrence of A:T to C:G transversion 1000-fold over the wild type level (
19
,
20
). Human cells contain an enzyme similar to the MutT protein, and this enzyme
specifically hydrolyzes 8-oxo-dGTP to 8-oxo-dGMP, with a relatively low
K
m
value, as compared with other deoxyribonucleoside triphosphates (
21
,
22
). The human 8-oxo-dGTPase is likely to have the same antimutagenic capability as the MutT protein, since the elevated level of spontaneous
mutation frequency in
E.coli
mutT
-
cells was greatly reduced by expression of cDNA for the human enzyme (
22
,
23
). The human gene for 8-oxo-dGTPase, named
hMTH1
(for human
mutT
homologue), was found to be located on chromosome 7p22 (
23
).
A certain degree of sequence homology has been noted in
E.coli
MutT and human MTH1 protein (
22
,
24
). Genes for analogous functions were isolated from
Proteus vulgaris
and
Streptococcus pneumoniae
, bacteria distantly related to
E.coli
(
25
,
26
). The products of the latter two genes carry enzyme activity specifically
degrading dGTP to dGMP and may be functionally related to the
E.coli
MutT protein. More recently, cDNAs for mouse and rat 8-oxo-dGTPase protein were isolated and their structures elucidated (
27
,
28
). Like the human MTH1 proteins, these mammalian proteins suppress the
occurrence of transversion mutations in
E.coli
mutT
-
cells, when appropriately expressed in the mutant cells. These six proteins of
prokaryotic and eukaryotic origins are of a similar size, and alignment of the
sequences showed that all carry a conversed sequence in almost the same region
(Fig.
5
). In the conserved region (from the 36th to the 58th amino acid for hMTH1), 10
among 23 amino acid residues are identical. It is likely that this region
constitutes an active center for the enzyme, and the secondary structure of the
MutT protein, elucidated by NMR analysis (
29
), supports this view.
To understand further the significance of the conserved amino acid sequence, we carried out a systematic site-directed mutagenesis study. Use of negative and positive screening procedures led to identification of the amino acid residues essential for enzyme activity.
MATERIALS AND METHODS
Chemicals
[[alpha]-
32
P]dCTP, [[alpha]-
32
P]dGTP and
125
I-labeled protein A were obtained from Amersham Japan (Tokyo). A DNA
labeling kit was purchased from Nippon Gene (Toyama, Japan). Restriction
enzymes, T4 DNA ligase, T4 DNA polymerase were obtained from Toyobo Co. (Osaka, Japan). 5-Bromo-4-chloro-3-indolyl-[beta]- D-galactopyranoside, phenyl-[beta]-D-galactopyranoside and rifampicin were obtained from Sigma, and isopropyl-[beta]-D-thiogalactopyranoside was from Wako Pure Chemical Industries (Osaka, Japan). The altered sites II
in vitro
mutagenesis system was purchased from Promega. Sources of other materials are
given in the text.
Bacterial strains
Escherichia coli
strain CC101 [
ara
[Delta](
lac proB
)
XIII
] carrying an
F lacI
-
Z
-
proB
+
episome (
30
), was a gift from J. H. Miller and
mutT-
mutation was introduced into CC101 by P1 transduction to obtain CC101T.
Escherichia coli
strain JM109 {
endA
1,
recA
1,
gyrA
96,
thi
-1
,
hsdR
17 (r
k
-
, m
k
+
),
relA
1,
supE
44, [lambda]
-
,
[Delta](
lac-proAB
), [
F
-
,
traD
36
, proA
+
B+
,
lacIq
Z
[Delta]M15]}
and ES1301 (
mutS
-
) {
lacZ
53,
mutS
201::Tn
5
,
thyA
36,
rha
-5,
metB
1,
deoC
, IN(
rrn-rrnE
)} were purchased from Promega.
Plasmids
Plasmid pALTER-1 was purchased from Promega and used for site-directed mutagenesis. Plasmid pTT100 was constructed from pTrc99A by
removing the
lacI
q
gene sequence by digestion with
Tth
111I and
Pvu
II and resealing (
23
). Plasmid pALTER-MTH-ss was constructed by inserting a 327 bp
Ssp
I-
Sac
I fragment, encoding the conserved amino acid sequence of hMTH1 protein, into
the
Sma
I-
Sac
I site of pALTER-1. Plasmid pTT100- hMTH1(K38X) carrying the mutant cDNA was constructed by inserting a
236 bp
Nco
I-
Sac
I mutagenized DNA fragment into
Sac
I-
Nco
I site of pTT100-hMTH1.
Oligonucleotides
Oligonucleotide primers were obtained from Greiner Labortechnik Co. Ltd (Tokyo, Japan). A mixture of 27mer oligonucleotides carrying every
possible combination of nucleotides of the target 3-nucleotide region were prepared and used as primers for site-directed mutagenesis.
Site-directed mutagenesis
Site-directed mutagenesis was performed according to the Promega technical
manual of altered sites II
in vitro
mutagenesis system, with modifications (Fig.
2
). The mutagenized DNA was used to transform
E.coli
ES1301 (
mutS
-
), and transformed cells were grown in an LB plate containing 50 [mu]g/ml ampicillin. Two types of plasmids derived from template DNA and from
newly synthesized DNA were recovered from the cells, and the plasmid mixture
was applied to
E.coli
JM109. Transformants were selected on LB plate containing 50 [mu]g/ml ampicillin, and Amp
r
colonies were picked and plated on paired plates containing either 50 [mu]g/ml ampicillin or 12.5 [mu]g/ml tetracycline. Amp
r
/Tet
s
colonies were isolated and cultured for preparation of plasmid DNA. The DNA was
digested with a set of restriction enzymes
Sac
I and
Nco
I, and subcloned into pTT100-hMTH1. Plasmid pTT100-hMTH1(K38X) thus obtained was applied to
E.coli
CC101T and cultured on agar medium containing minimal A salts, 0.2% glucose, 0.05% phenyl-[beta]-D-galactopyranoside and 5-bromo-4-chloro-3-indolyl-[beta]-D-galactopyranoside (40 [mu]g/ml) (
31
). After incubation for 5 days at 37oC, papillae formation was examined and appropriate samples were subjected to further analyses.
Assay of 8-oxo-dGTPase
The reaction mixture (12.5 [mu]l) contained 20 mM Tris-HCl, pH 8.0, 4 mM MgCl
2,
40 mM NaCl, 20 [mu]M 8-oxo-dGTP (including [[alpha]-
32
P]8-oxo-dGTP), 80 [mu]g/ml bovine serum albumin (BSA), 8 mM dithiothreitol (DTT), 2% glycerol and an extract of cells
carrying plasmids to be examined. The reaction was run at 30oC for 20 min and terminated by adding 2.5 [mu]l 50 mM EDTA. An aliquot (2 [mu]l) of the reaction mixture was applied to a PEI-cellulose sheet. 8-Oxo-dGMP produced was separated from 8-oxo-dGTP on TLC with 1 M LiCl for 90 min
and quantitated in Fujix Bio-image analyzer BAS2000 (Fuji Photofilm Co., Ltd, Tokyo) (
22
). One unit of 8-oxo-dGTPase was defined as the amount of enzyme that produced 1 pmol/min
8-oxo-dGMP at 30oC (
21
).
Western blot analysis
Western blotting was done as described (
32
,
33
), but with minor modification. Five or ten micrograms of protein of the
bacterial cell crude extracts were subjected to 15% SDS-polyacrylamide gel electrophoresis, and proteins were electrotransferred
to nitrocellulose membrane (BA-83, Schleicher & Schuell, Dassel, Germany) at 25 V for 1 h in transfer buffer (48 mM
Tris, 39 mM glycine, 1.3 mM SDS, 20% methanol, pH 9.3). The nitrocellulose
filter was soaked in blocking solution (5% BSA, 10 mM Tris-HCl, pH 7.4, 0.9% NaCl, 0.05% Tween 20) at 52oC for 1 h and then incubated overnight at 4oC with anti-MTH1 (1 [mu]g/ml) (
33
). The filter was rinsed in buffered saline containing 0.05% Tween 20 to remove excess antibodies, then reacted with 1 [mu]Ci/ml of
125
I-labeled protein A for 1 h on ice. The filter was washed with radioimmune
precipitation assay buffer (50 mM Tris-HCl, pH 8.0, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 150 mM NaCl) at room
temperature to remove unbound protein A. The filter was air-dried, exposed to an imaging plate for 3 h and then data were processed using a Fujix Bio-image analyzer BAS2000.
RESULTS
Isolation of negative mutants with Lys-38 substitution
Substitution of the lysine residue at amino acid position 38 was done as this
residue is present in all the known MTH1-related proteins and locates within the highly conserved region of the
protein (Fig.
5
). Figure
1
shows the strategy for site-directed mutagenesis and subsequent mutant selection. Oligonucleotides with all possible combinations for a given codon were synthesized and used as primers for generating the complementary strand. To
facilitate selective replication of the mutated plasmids, a set of
oligonucleotide fragments exerting ampicillin resistance and tetracycline
susceptibility were also introduced. Double-stranded DNAs were produced by
in vitro
repair reaction and then propagated in
E.coli
cells defective in mismatch repair. From the progeny plasmid a DNA fragment
carrying the mutagenized region was excised and introduced into pTT100-hMTH1 to replace the corresponding region with the normal sequence. The
resulting plasmids were applied to
E.coli mutT
-
cells and colonies were examined.
Properties of mutant proteins
Western blot analysis was done with lysates of
E.coli
cells producing mutant forms of proteins (Fig.
3
). Using a polyclonal antibody against hMTH1 protein, a major band was detected
in all the samples, except one carrying a termination codon, at a position
corresponding to the authentic 18 kDa human protein. Quantitative estimation of
amounts of proteins was made by scanning the gels. Based on the standard curve
obtained from the purified hMTH1 protein, we estimated the amount of mutant
protein in each extract (Table
2
). In all cases, immunologically reactive proteins were produced in quantities
similar to or more than the wild type protein.
These mutants were assayed for 8-oxo-dGTPase activity and the result is shown also in Table
2
. Little or no activity was found in extracts of cells harboring either one of
the mutant clones; specific activities of mutants were <4% of the wild type.
To determine the
in vivo
biological activities of mutant proteins, mutation frequencies of cells
producing various forms of hMTH1 protein were determined. As shown in Table
3
, the mutation frequency of CC101T (
mutT
-
) harboring the vector plasmid was ~250-fold higher than that of CC101 (
mutT
+
) carrying the same vector. When the normal cDNA (Lys-38) was placed on the vector plasmid and properly expressed, this high
level of mutation induction was completely suppressed. Mutants with tyrosine or
leucine substitution at the Lys-38 site showed no capability of suppression, while those with arginine or
alanine substitution exerted 10% of the suppressive effect of the wild type.
Positive revertants at the Lys-38 sites
Analyzing clones isolated by negative selection, we identified mutants with 11
different amino acid substitutions. Since it is laborious to isolate all the
possible mutants using these procedures, positive selection was used to obtain
revertants from individual negative mutants. The procedure was essentially the
same as that described above except that one of the negative mutants was used
as the template for mutagenic DNA synthesis. For this, we used the K38R mutant,
in which lysine at codon 38 was converted to arginine. It was shown by Western
blot analysis that the mutant produces a sufficient amount of hMTH1 protein
(Table
2
). After mutagenization, clones showing no papillae formation were selected
(Fig.
2
D) and, among the 619 examined, eight positive revertants were isolated. DNA
sequence analysis revealed that all carried the codon for lysine at position 38.
Figure
4
shows data on the quantitative assay of 8-oxo-dGTPase activity exerted by wild type, one of negative mutants and
its positive revertant. The enzyme activity of Arg-38 mutant (K38R) was <10% of the wild type. As expected, the enzyme level of cells harboring
the positive revertant (K38R-K) was essentially the same as that of cells carrying the wild type DNA.
Suppression of
E.coli
mutT
mutator phenotype by expression of wild type or mutant hMTH1
Escherichia coli
strain
Mutation frequency
CC101 (
mutT
+
) with pTT100 (vector)
2.7 +- 0.7 * 10
-8
CC101T (
mutT
-
) with pTT100 (vector)
6.7 +- 0.3 * 10
-6
CC101T (
mutT
-
) with pTT100-hMTH1 (wild type)
1.9 +- 0.8 * 10
-8
CC101T (
mutT
-
) with pTT100-hMTH1 (K38Y)
5.9 +- 1.1 * 10
-6
CC101T (
mutT
-
) with pTT100-hMTH1 (K38L)
5.4 +- 0.6 * 10
-6
CC101T (
mutT
-
) with pTT100-hMTH1 (K38R)
1.9 +- 0.1 * 10
-7
CC101T (
mutT
-
) with pTT100-hMTH1 (K38A)
1.5 +- 0.5 * 10
-7
Analyses of other conserved residues
ACKNOWLEDGEMENTS
We thank Dr K. Sakumi for discussion, Y. Fujii and T. Ohtsubo for participating
in some experiments, and M. Ohara for helpful comments. This work was supported
by grants from the Ministry of Education, Science and Culture, Japan, and HFSP.
