Download Genetic polymorphisms and alternative splicing of the

Oncogene (1998) 16, 3219 ± 3225
 1998 Stockton Press All rights reserved 0950 ± 9232/98 $12.00
http://www.stockton-press.co.uk/onc
Genetic polymorphisms and alternative splicing of the hOGG1 gene, that is
involved in the repair of 8-hydroxyguanine in damaged DNA
Takashi Kohno1, Kazuya Shinmura1,3, Masahiko Tosaka1, Masachika Tani1, Su-Ryang Kim2,
Haruhiko Sugimura3, Takehiko Nohmi2, Hiroshi Kasai4 and Jun Yokota1
1
Biology Division, National Cancer Center Research Institute, 1 ± 1, Tsukiji 5-chome, Chuo-Ku, Tokyo 104; 2Division of Genetics
and Mutagenesis, National Institute of Health Sciences, 18 ± 1, 1 chome, Kamiyoga, Setagayaku, Tokyo 158; 3First Department of
Pathology, Hamamatsu University School of Medicine, 3600 Handa-cho, Hamamatsu, 431 ± 31; 4Department of Environmental
Oncology, University of Occupational and Environmental Health, 1 ± 1, Iseigaoka, Yahatanishi-ku, Kitakyushu 807, Japan
The hOGG1 gene encodes a DNA glycosylase that
excises 8-hydroxyguanine (oh8Gua) from damaged DNA.
Structural analyses of the hOGG1 gene and its
transcripts were performed in normal and lung cancer
cells. Due to a genetic polymorphism at codon 326,
hOGG1-Ser326 and hOGG1-Cys326 proteins were produced in human cells. Activity in the repair of oh8Gua
was greater in hOGG1-Ser326 protein than in hOGG1Cys326 protein in the complementation assay of an E. coli
mutant defective in the repair of oh8Gua. Two isoforms
of hOGG1 transcripts produced by alternative splicing
encoded distinct hOGG1 proteins: one with and the other
without a putative nuclear localization signal. Loss of
heterozygosity at the hOGG1 locus was frequently (15/
23, 62.2%) detected in lung cancer cells, and a cell line
NCI-H526 had a mutation leading to the formation of
the transcripts encoding a truncated hOGG1 protein.
However, the oh8Gua levels in nuclear DNA were similar
among lung cancer cells and leukocytes irrespective of
the type of hOGG1 proteins expressed. These results
suggest that the oh8Gua levels are maintained at a steady
level, even though multiple hOGG1 proteins are
produced due to genetic polymorphisms, mutations and
alternative splicing of the hOGG1 gene.
Keywords: 8-hydroxyguanine; DNA repair; genetic
polymorphism; alternative splicing; hOGG1 gene
Introduction
8-hydroxyguanine (oh8Gua), a major form of oxidative
DNA damage induced by reactive free radicals, is
highly mutagenic in vitro and in vivo (Kasai and
Nishimura, 1991; Michaels and Miller, 1992; Grollman
and Moriya, 1993). The presence of oh8Gua in DNA
causes G:C to T:A transversion, since oh8Gua directs
the incorporation of cytosine and adenine nucleotides
opposite the lesion. To prevent the mutagenic e€ect of
oh8Gua, three DNA repair enzymes (MutM, MutY and
MutT proteins) exist in Escherichia coli (E. coli) (Tchou
et al., 1991; Boiteux et al., 1992; Au et al., 1989;
Michaels et al., 1991, 1992; Maki and Sekiguchi, 1992).
Recently, the OGG1 gene of Saccharomyces cerevisiae
(yOGG1) was cloned as being a functional yeast
Correspondence: J Yokota
Received 12 August 1997; revised 26 January 1998; accepted 27
January 1998
homologue of the bacterial mutM gene (van der
Kemp et al., 1996; Nash et al., 1996). Subsequently,
a human homologue of the yeast OGG1 gene, hOGG1,
was isolated based on the homology search of
expressed sequence tags (Aburatani et al., 1997; Arai
et al., 1997; Lu et al., 1997; Radicella et al., 1997;
Roldan-Arjona et al., 1997; Rosenquist et al., 1997).
Since hOGG1 proteins eciently released the oh8Gua
opposite the pyrimidine from DNA and cleaved the AP
site in the same manner as yOGG1 protein, it was
concluded that hOGG1 is a functional human
homologue of the yOGG1 gene (Aburatani et al.,
1997; Lu et al., 1997; Radicella et al., 1997; RoldanArjona et al., 1997; Rosenquist et al., 1997; Shinmura
et al., 1997).
Genetic backgrounds in control of the repair of
damaged DNA have been shown to be involved in the
susceptibility to cancer development (reviewed in
Perera, 1996). Therefore, it is worth investigating
genetic polymorphisms of the hOGG1 gene, since a
population with decreased enzyme activity of the
hOGG1 gene would be at high risk of developing
cancer in life because of an incomplete repair of
oxidative DNA damage. The hOGG1 gene has been
mapped to chromosome 3p26.2 (Arai et al., 1997), a
region showing loss of heterozygosity (LOH) in a
variety of human cancers (Yokota and Sugimura, 1993;
Gazder 1994). In particular, 3p25-p26 is a common
region of LOH in lung cancer (Hibi et al., 1992, Hosoe
et al., 1994), and G:C to T:A transversions in the p53
gene occur frequently in lung cancer (Greenblatt et al.,
1994). Therefore, it is possible that the hOGG1 gene
functions as a tumor suppressor (Lu et al., 1997). In
addition, previous Northern blot hybridization analysis
of the hOGG1 gene indicated that alternative splicing
of the hOGG1 gene occurs in human cells (Arai et al.,
1997). Thus, major forms of the hOGG1 transcripts
should be elucidated for further functional analysis of
the hOGG1 gene products.
In this report, we performed structural analysis of
the hOGG1 gene and its transcripts in normal and
cancerous cells of the lung. Our data indicated that,
due to genetic polymorphisms and alternative splicing
of the hOGG1 gene, multiple isoforms of hOGG1
protein are produced in human lung cells. We also
found frequent LOH at the hOGG1 locus in primary
lung tumors and an intragenic mutation of the hOGG1
gene in a lung cancer cell line. However, the oh8Gua
levels were maintained at a steady level in nuclear
DNA of peripheral leukocytes and lung cancer cells
irrespective of the type of hOGG1 proteins expressed.
Polymorphisms and alternative splicing of the hOGG1 gene
T Kohno et al
3220
Therefore, enzyme activities of the hOGG1 proteins
may be high enough to maintain the oh8Gua levels at a
steady level in human cells.
Results
Genetic polymorphisms of the hOGG1 gene
Genomic organization of the hOGG1 gene was
determined by sequencing of a lambda phage clone
and three bacterial arti®cial chromosome (BAC) clones
covering the hOGG1 gene locus. The hOGG1 gene
consisted of 7 exons and 6 introns, and the initiator
ATG and terminator TAG sequences were located in
the ®rst and seventh exons, respectively. This data was
consistent with the one determined by Aburatani et al.,
(1997).
Then, we searched for polymorphisms of the hOGG1
gene using genomic DNAs prepared from noncancerous lung tissues of 45 unrelated patients with lung
cancer. Six pairs of intron based primers were used to
amplify exons 1 ± 7 for the PCR-single strand
conformation polymorphism (SSCP) analysis. By the
direct sequencing of DNA fragments corresponding to
shifted bands, ®ve di€erent types of DNA polymorphisms were identi®ed (Table 1). A C/G polymorphism at
position 1245 in exon 7 was associated with an
exchange of an amino acid, serine/cysteine, in codon
326. A silent polymorphism was found at codon 98 in
exon 2, while three others were present in the
noncoding region of the hOGG1 gene. Allele frequencies of these polymorphisms are shown in Table 1.
Genetic alterations at the hOGG1 locus in lung cancer
We next undertook PCR ± SSCP analysis against 45
primary lung tumors. Somatic mutation of the hOGG1
gene was assessed by comparing the electrophoretic
patterns of DNAs from 45 primary lung tumors with
those from corresponding normal lung tissues. However, mutations of the hOGG1 gene were not detected
in these materials.
Six, one, 19 and 19 of the 45 individuals were
heterozygous for the polymorphisms at the 5'-noncoding region, codon 98, codon 326, and intron 4,
respectively, and in total 23 of the 45 (51.1%)
individuals were heterozygotes for the polymorphisms
of the hOGG1 gene. Therefore, we could assess LOH at
the hOGG1 locus in these cases. LOH was detected in
15 of the 23 cases (65.2%). Of the 19 cases
heterozygous for the polymorphism at codon 326,
four cases lost the hOGG1-Ser326 allele and nine cases
lost the hOGG1-Cys326 allele.
PCR ± SSCP analysis was also performed in 52 lung
cancer cell lines. A G to A transition at the last
nucleotide of exon 1 was detected in a lung cancer cell
line, NCI ± H526 (Table 1). Arginine at codon 46 was
replaced by glutamine by this transition. Direct
sequencing of the PCR products covering exon 1
from NCI ± H526 cells revealed that NCI ± H526 was
homozygous for the variant allele. Since the corresponding normal tissue DNA for this cell line was not
available, we could not determine whether it was a rare
genetic polymorphism or a somatic mutation.
Analyses of the hOGG1 transcripts
hOGG1 transcripts were analysed by RT ± PCR using
a set of primers which are located at the 5'- and 3'non-coding regions, respectively. Multiple sizes of
RT ± PCR products including two major forms were
ampli®ed from human lung cancer cells and normal
lung cells, and the ratio of the transcripts varied
among cells (Figure 1a). The relationship between the
types of polymorphisms and the pattern of transcripts
was not observed. We cloned two major forms of
transcripts and determined the nucleotide sequences.
The larger one was an alternatively spliced form
containing 244 bp of the intron 6 sequence, while the
smaller one was the same as the one we and others
previously isolated (Arai et al., 1997; Lu et al., 1997;
Radicella et al., 1997; Rosenquist et al., 1997).
Southern blot hybridization analysis using oligonucleotides derived from the intron 6 sequence as a
probe con®rmed this result (Figure 1a). In amino acid
sequence encoded by the larger transcripts, 29 amino
acids at the C-terminal were replaced by other eight
amino acids, and a putative nuclear localization signal
(NLS) was lacked as reported by Aburatani et al.,
(1997).
Notably, two other distinct transcripts were detected
as major ones in NCI ± H526 cells in addition to the
two forms of transcripts described above. We cloned
these transcripts and determined the nucleotide
sequences. These transcripts contained the 521 bp
intron 1 sequence, resulting in a premature termination just after exon 1. Southern blot hybridization
analysis using oligonucleotides derived from the intron
1 sequence as a probe con®rmed this result (Figure 1a).
This result indicated that NCI ± H526 cells expressed
truncated hOGG1 proteins in addition to two fulllength isoforms of hOGG1 proteins with and without
NLS.
Table 1 Polymorphism of the hOGG1 gene
Location
5'-noncoding region (± 23 bp from the initiation codon)
5'-noncoding region (± 18 bp from the initiation codon)
Exon 1
Exon 2
Intron 4 (± 15 bp from exon 5)
Exon 7
a
mRNA positiona
(codon)
246
251
405 (46)
562 (98)
1245 (326)
Sequence
A/G
G/T
G/A
G/A
C/G
C/G
Amino acid
Allele frequencyb
Arg/Gln
Lys/Lys
0.94/0.06
0.96/0.04
1.00/0.00c
0.99/0.01
0.57/0.43
0.57/0.43
Ser/Cys
According to EMBL accession no. AB000410. bAllele frequency was calculated from 45 unrelated individuals with lung carcinoma. cNot
detected in 45 unrelated individuals, but detected in NCI-H526 cell line
Polymorphisms and alternative splicing of the hOGG1 gene
T Kohno et al
hOGG-plasmid
H526
H69
PC13
Lu135
PC10
H23
Lu65
Normal lung
hOGG-plasmid
(–)
a
We also examined the hOGG1 transcripts in lung
cancer cell lines by Northern blot hybridization
analysis (Figure 1b). Again, the relationship between
the types of polymorphisms and the pattern of the
transcripts was not observed, except that the size of the
hOGG1 transcripts in NCI ± H526 cells was approximately 0.5 kb greater than those in other cell lines.
5' (-23)
Intron 4
Codon 326
A A A A A A A/G
C C C G G C G/C
S S S C C C S/C
kbp
— 2.0
— 1.6
EtBr
— 1.0
kbp
— 2.0
— 1.6
Exon 2
— 1.0
Activities of hOGG1 variants to repair oh8Gua residues
in DNA
We constructed three vectors producing full-length
hOGG1-Ser326, hOGG1-Cys326 and hOGG1-Gln46 proteins with NLS. The ability of hOGG1-Ser326, hOGG1Cys326 and hOGG1-Gln46 proteins to suppress spontaneous mutagenesis was assayed by the generation of
rifampicin resistant (RifR) mutants in YG5132, a mutM
mutY double mutant of E. coli strain CC104, that is
defective in the repair of oh8Gua in DNA and has an
extremely high G:C to T:A mutation rate (Arai et al.,
1997; Suzuki et al., 1997) (Table 2). The number of
RifR colonies was reduced by either of three types of
GST-hOGG1 fusion proteins. However, it is noted that
the degrees of suppression by the hOGG1-Cys326 and
hOGG1-Gln46 proteins were signi®cantly smaller than
that by the hOGG1-Ser326 protein.
kbp
— 2.0
— 1.6
Intron 6
— 1.0
kbp
— 2.0
— 1.6
Intron 1
E. coli strain
— 4.4
hOGG1
— 2.4
— 1.4
β-actin
Lower activity of hOGG1-Cys326 protein than that of
hOGG1-Ser326 protein in oh8Gua DNA repair
prompted us to examine the association of the Cys326
allele with the susceptibility to human lung cancer. We
undertook a comparative analysis of the distribution of
the genotypes and the allele frequencies of the
polymorphism at codon 326 between 42 healthy
individuals and 45 lung cancer patients (Table 3). In
both groups, the distribution of the polymorphic alleles
Table 2 Complementation of an E. coli (mutM mutY) mutant with
hOGG1-overproducing plasmids
Lu65
H23
PC10
Lu135
PC13
H69
b
H526
— 1.0
Genotypic and allelic frequency of the polymorphism at
codon 326 in healthy individuals and lung cancer patients
CC104
YG5132b
YG5132b
YG5132b
YG5132b
YG5132b
Plasmid
No. of Rif r cells per
108 cellsa
±
±
pGEX-1lT
GST-hOGG1-Ser326
GST-hOGG1-Cys326
GST-hOGG1-Gln46
1.3+0.3
2766+672
1520+456
3.1+2.8c,d
20.9+4.9c
42.7+12.6d
a
Mean+s.e. is shown. bStrain YG5132 is a mutM mutY double
mutant of E. coli CC104 (Aria et al., 1997; Suzuki et al., 1997). cP
value for the di€erence between GST-hOGG1-Ser326 and GSThOGG1-Cys326 is 0.0013. dP value for the di€erence between GSThOGG1-Ser326 and GST-hOGG1-Gln46 is 0.013
— 2.0
Figure 1 Analysis of the hOGG1 transcripts. (a) RT ± PCR
analysis of the hOGG1 transcripts. PCR products were
fractionated by 0.8% agarose gel, stained with ethidium
bromide, transferred to a nylon membrane, and hybridized to
the oligonucleotide probes derived from exon 2, intron 6 and
intron 1. A hOGG1 cDNA clone (Arai et al., 1997) was also
subjected to this analysis as a control (lanes designated as
`hOGG-plasmid'). The positions of size markers in kbp are shown
on the right. (b) Northern blot analysis of the hOGG1 gene. Three
mg of poly(A) RNA was electrophoresed, and transferred to a
nylon membrane. The membrane was sequentially hybridized to
the hOGG1 (upper) and b-actin cDNA probes. b-actin cDNA
was used for standardizing the RNA amounts on the membrane
(lower). Sizes in kb are shown on the right
3221
Polymorphisms and alternative splicing of the hOGG1 gene
T Kohno et al
3222
was consistent with Hardy-Weinberg equilibrium. The
distribution of the three genotypes, namely two forms
of homozygotes (Ser326/Ser326 and Cys326/Cys326) and a
heterozygote (Ser326/Cys326), as well as the frequency of
each allele were not signi®cantly di€erent between the
two groups (P40.05 by w2-test).
oh8Gua levels in DNA from lung cancer cell lines and
peripheral leukocytes
Next, we examined the oh8Gua levels in nuclear DNA
of human cells that express either or both of hOGG1Ser326 and hOGG1-Cys326 by the high-performance
liquid
chromatography-electrochemical
detector
(HPLC ± ECD) method. NCI ± H526, that expresses
hOGG1-Gln46, was also subjected to this analysis. We
analysed three cases of peripheral leukocytes expressing either hOGG1-Ser326 or hOGG1-Cys326 protein,
three lung cancer cell lines expressing one of three
di€erent hOGG1 proteins, and a lung cancer cell line
expressing both hOGG1-Ser326 and hOGG1-Cys326
proteins (Table 4). The oh8Gua levels in leukocytes
expressing hOGG1-Ser326 protein ranged from 0.12 ±
0.14 per 105 guanine, while those in leukocytes
expressing hOGG1-Cys326 protein ranged from 0.13 ±
0.17 per 105 guanine. The mean oh8Gua levels were
not signi®cantly di€erent between the two groups
(P40.05 by t-test). Four lung cancer cell lines also
showed the similar oh8Gua levels irrespective of the
type of hOGG1 protein expressed.
Discussion
Five di€erent types of genetic polymorphisms were
detected in the hOGG1 gene locus. A polymorphism in
Table 3 Genotypes and allele frequencies of the polymorphism at
codon 326 of the hOGG1 gene in healthy individuals and lung cancer
patients
Healthy individualsa
n
(%)
Genotype
Ser326/Ser326
Ser326/Cys326
Cys326/Cys326
Allele:
Ser326
Cys326
a
n=42; bn=45.
a,b
Lung cancer patientsb
n
(%)
15
20
7
(35.7)
(47.6)
(16.7)
16
19
10
(35.6)
(42.2)
(20.0)
50
34
(59.5)
(40.5)
51
39
(56.7)
(43.3)
All the subjects are Japanese from Tokyo
Table 4 oh8Gua levels in DNA from lung cancer cell lines and
peripheral leukocytes
Name
L1/L2/L3
L4/L5/L6
Lu135
NCI ± H526
NCI ± H82
Lu65
Cell
Leukocytes
Leukocytes
SCLC
SCLC
SCLC
LCC
Polymorphism
8-OH-Gua level
Codon 46 Codon 326 (per 105 Gua)
Arg
Arg
Arg
Gln
Arg
Arg
Ser/Ser
Cys/Cys
Cys
Ser
Ser
Ser/Cys
0.13+0.01a
0.15+0.02a
0.16
0.16
0.15
0.15
SCLC: Small cell lung carcinoma, LCC: Large cell carcinoma.
Mean+s.e. is shown
a
exon 7 resulted in a serine-cystein exchange in codon
326, while four others were not associated with the
exchange of amino acids. Therefore, at least two types
of hOGG1 proteins, hOGG1-Ser326 and hOGG1-Cys326,
are produced in human cells. Interestingly, comparative
functional analysis of the two polymorphic forms of
the hOGG1 proteins revealed that activity to suppress
spontaneous mutagenesis in an E. coli strain, that is
defective in the repair of oh8Gua in DNA, was
signi®cantly stronger in hOGG1-Ser326 protein than in
hOGG1-Cys326 protein. Therefore, it is possible that the
activity in oh8Gua DNA repair di€ers among human
cells due to the polymorphic genotype in this region of
the hOGG1 gene.
In this study, crude comparison of the distribution
of the polymorphism at codon 326 did not detect the
di€erence between healthy individuals and lung cancer
patients, but obviously, confounding factors such as
age and smoking dosage should be incorporated to the
interpretation. A larger case-control study is in
progress. Furthermore, the other four genetic polymorphisms of the hOGG1 gene were not examined in
detail in this study. It is possible that those
polymorphisms are associated with the di€erences in
cellular enzyme activity by changing mRNA stability,
eciency in transcription and translation. Therefore,
comparative analysis of the multiple polymorphisms at
the hOGG1 locus among a large number of healthy
individuals and patients with a variety of cancers
should be done to assess the involvement of the
hOGG1 gene in human carcinogenesis.
Multiple alternative splicing isoforms of the hOGG1
gene, including two major isoforms, were detected in
both normal lung cells and lung cancer cell lines. This
is consistent with the previous result that multiple sizes
of mRNA transcripts were detected in human organs
and lung cancer cell lines (Arai et al., 1997). One of
major form hOGG1 transcripts encoded a protein with
a nuclear localization signal (NLS) that was translated
from the transcripts composed of exons 1 ± 7, while the
other was an isoform lacking NLS translated from the
transcripts with intron 6. Aburatani et al. (1997)
reported another isoform lacking NLS, in which exon
7 was replaced by exon 8 located at the downstream of
exon 7. hOGG1 proteins without NLS are expected to
have di€erent intracellular localization. Recently,
hMTH protein, a 8-hydroxydeoxyguanosine triphosphatase, was shown as being present in nucleus,
mitochondria and cytosol (Kang et al., 1995). Therefore, each isoform of the hOGG1 proteins may have
distinct intracellular localization and distinct functions
in human cells. It is also possible that some isoforms
are involved in the regulation of the enzyme activity of
the major isoform of hOGG1 protein.
The hOGG1 gene is located at chromosome 3p26.2
(Arai et al., 1997), a commonly deleted region in lung
cancer cells (Hibi et al., 1992, Hosoe et al., 1994). Since
the VHL tumor suppressor gene at 3p25 is not altered
in human lung cancer cells, a target gene at this
chromosomal region is still unidenti®ed (Sekido et al.,
1994). Thus, we examined whether the hOGG1 gene is
frequently altered in lung cancer cells. LOH at the
hOGG1 locus was frequent, while intragenic mutation
of the hOGG1 gene was infrequent in human lung
cancer cells. Therefore, it is unlikely that the hOGG1
gene is commonly inactivated by chromosomal
Polymorphisms and alternative splicing of the hOGG1 gene
T Kohno et al
deletions at this region. Infrequent biallelic inactivation
of the hOGG1 gene in lung cancer cells limited the
importance of the hOGG1 gene in lung carcinogenesis.
However, since lung cells have high concentrations of
reactive oxygen species generating oh8Gua especially by
inhalation of tobacco smoke, it is possible that lung
cells that have lost one allele of the hOGG1 gene have a
reduced capacity to repair oh8Gua in DNA and it may
accelerate the acquisition of G:C to T:A mutations.
A nucleotide change was detected at the last
nucleotide of exon 1 in NCI ± H526 cells. This
nucleotide change causes the conversion of an amino
acid at codon 46, which is located in a domain highly
conserved in the yeast counterpart (Arai et al., 1997).
Activity to suppress spontaneous mutagenesis of
hOGG1-Gln46 protein in an E. coli strain was weaker
than those of hOGG1-Ser326 and hOGG1-Cys326
proteins. Furthermore, this nucleotide change caused
abnormal splicing of the hOGG1 gene. Due to a partial
inactivation of the 5' splice site of intron 1, readthrough transcripts of intron 1 encoding a truncated
hOGG1 protein were produced in NCI ± H526 cells.
This result is compatible with the previous report that
base changes at the terminal nucleotide position of
exons lead to the abnormal splicing of the genes
(Huang et al., 1993; Garuti et al., 1996). Guanine at
the last nucleotide of exon resides in the consensus
sequence at the 5' splice site, which is recognized and
bound by U1 small nuclear RNA in the 5' splice site
selection (Chabot, 1996). Therefore, G to A change at
this site could weaken the splicing signal, allowing the
maturation of both full-length and readthrough
transcripts from primary transcripts in NCI ± H526
cells. Interestingly, this nucleotide change was not
detected in 45 normal DNAs. Therefore, this variant
can be a somatic mutation and not a genetic
polymorphism, although this cannot be concluded
because of the lack about the sequence in information
of the normal DNA corresponding to the cell line.
We demonstrated that multiple isoforms of hOGG1
protein are produced in human cells due to polymorphisms, mutations and alternative splicing of the
hOGG1 gene. The result of structural and functional
analysis in this study suggested the presence of
di€erence in the enzymatic activity and property
among hOGG1 isoforms. We preliminarily compared
oh8Gua DNA glycosylase and AP-lyase activities,
which are known as two major enzyme activities of
hOGG1 protein (Shinmura et al., 1997), among
puri®ed GST-hOGG1-Ser326, GST-hOGG1-Cys326 and
GST-hOGG1-Gln46 proteins. However, their activities
were not signi®cantly di€erent among the three
isoforms. Therefore, the enzyme activities, which
account for the di€erence in ability to suppress
spontaneous mutagenesis of a E. coli mutant defective
in the repair of oh8Gua, have not been elucidated.
Other factors including stability of the proteins or
interaction with other proteins involved in the repair
process may be signi®cant for the di€erence. In
contrast to the fact that multiple isoforms of hOGG1
protein are produced in human cells, the oh8Gua levels
in nuclear DNA were similar among lung cancer cell
lines and peripheral leukocytes irrespective to the type
of hOGG1 proteins expressed. Thus, it is possible that
intracellular activities of hOGG1 proteins may be high
enough to maintain the oh8Gua contents at a steady
level in nuclear DNA under conditions without severe
oxidative stress. Intercellular variation of the oh8Gua
levels may become distinct, only when formation of
oh8Gua is highly induced by reactive oxygen species.
Therefore, it would be important to analyse the
oh8Gua levels under conditions with high oxidative
stress to elucidate whether each isoform of hOGG1
proteins have enough activities. Alternatively, it is also
possible that other proteins than hOGG1 play more
signi®cant roles in the repair of oh8Gua in human cells.
This idea is supported by the previous result that
human cell extract includes repair enzymes for oh8Gua,
whose enzymatic property is di€erent from that of
hOGG1 protein (Bessho et al., 1993). We have to also
consider the possibility that intercellular variation of
the oh8Gua levels was masked due to contamination of
arti®cially produced oh8Gua during preparation of
deoxynucleoside samples. Although the oh8Gua contents measured by our method is much lower than
those by other methods (Malins et al., 1991; Asami et
al., 1996; Nakajima et al., 1996), we can not completely
exclude the possibility that a signi®cant portion of the
oh8Gua detected in this study was produced arti®cially
during preparation.
Recently, we have demonstrated the presence of the
considerable interindividual variations (7.2-fold) in the
oh8Gua levels among the leukocytes of healthy
individuals (Asami et al., 1996). The result of this
paper indicated that the intervariations in the oh8Gua
levels were not directly correlated with the genetic
polymorphisms of the hOGG1 gene. Life-style, age, sex,
environment and genetic di€erences other than hOGG1
may have more signi®cant e€ects on the level of
oh8Gua in human cells. However, it is possible that
interindividual di€erence in capacity and property for
the repair of oh8Gua is associated with the type of
hOGG1 proteins expressed. A population with reduced
capacity can be at high risk of developing cancer due
to incomplete repair of oh8Gua, especially when the
formation of oh8Gua is increased by exposure to
mutagen, carcinogen and g-ray irradiation. Therefore,
functional analysis of the multiple isoforms of hOGG1
proteins in vivo as well as in vitro should be done in
more detail to assess whether each of multiple forms of
hOGG1 protein has distinct activities in the repair of
oh8Gua in damaged DNA.
Materials and methods
Samples
Thirty-three NSCLC cell lines (A427, A549, Lu65, Lu99,
NCI ± H23, NCI ± H157, NCI ± H322, NCI ± H441, NCI ±
H520, NCI ± 596, NCI ± H1155, PC3, PC7, PC9, PC10,
PC13, Ma1, Ma2, Ma3, Ma10, Ma12, Ma17, Ma24, Ma25,
Ma26, Ma29, Ma31, LC1-Sq, RERF ± LCD, RERF ±
LCOK, RERF ± LCMS, ABC1 and EBC1) and 19 small
cell lung carcinoma cell lines (NCI ± H69, NCI ± H82,
NCI ± H209, NCI ± H526, NCI ± H774, NCI ± H841, N230,
N231, N417, Lu24, Lu130, Lu134, Lu135, Lu138, Lu139,
Lu140, Lu141, SBC-5 and MS18) were used in this study
(Kohno et al., 1995). Detailed information on these cell
lines can be obtained upon request. High-molecular-weight
DNA was prepared from tumors and adjacent noncancerous tissues of 45 patients with SCLC or NSCLC, and from
peripheral blood samples of 42 unrelated healthy indivi-
3223
Polymorphisms and alternative splicing of the hOGG1 gene
T Kohno et al
3224
duals as described previously (Shiseki et al., 1994, 1996;
Baba et al., 1995). Poly(A) RNA was prepared by the Fast
Track mRNA isolation kit (Invitrogen Corp.).
Determination of the exon-intron organization of the hOGG1
gene
High Density BAC Colony Membranes (Research Genetics) were screened with a 32P-labeled hOGG1 cDNA probe
according to the supplier's protocol. Three BAC clones,
168D44, 176M11 and 187A24 were positive for the probe
and BAC DNAs for these clones were isolated according to
the supplier's protocol. Phage DNA of G27 (Arai et al.,
1997) and PCR products from BAC clones were used for
DNA sequencing. Sequence analysis was performed along
the exons, at the exon-intron boundaries, and at the 5'- and
3'-¯anking regions of the exons by f-mol DNA sequencing
system (Promega) or using the 373S DNA sequencer with
DNA sequencing kit (Applied Biosystems).
PCR ± SSCP and PCR ± LOH analyses
Exons 1 ± 7 of the hOGG1 gene were ampli®ed for SSCP
analysis
with
six
sets
of
PCR
primers:
5'TGAATTCGTCTTTGGGCGTCGACGA-3' and 5'-ACAGGCTTCTCAGGCTCAGT-3' for exon 1, 5'-ATTGAGTGCCAGGGTTGTCA-3' and 5'-TGGCAAAACTGAGTCATAGAG-3' for exon 2, 5'-AACAGCAGGTACCTGTCCTA-3' and ACAGATCTTGAAAGCTGATGG for exon 3, 5'-GAGAGCTCACTTACTAGCCT-3'
and 5'-AGTAGAGAGGGCAGCTCCTA-3' for exon 4, 5'AAGCAAGATGCTGGCCACAT-3' and 5'-TGGTGGAAGAGTCCACTGG-3' for exons 5 and 6, and 5'-ACTGTCACTAGTCTCACCAG-3' and 5'-TGAATTCGGAAGGTGCTTGGGGAAT-3' for exon 7. PCR ± LOH and
PCR ± SSCP analyses were performed as described previously (Sameshima et al., 1992, Otsuka et al., 1996).
Brie¯y, PCR was carried out by using 50 ng genomic DNA
as a template, in a 20 ml reaction mixture containing 0.25 m
of [a732P]dCTP (3000 Ci/mmol, 10 Ci/ml). After 35 cycles
of 958C (60 s) for denaturation, 558C (60 s) for annealing,
728C (90 s) for extention, PCR products were denatured
and electrophresed on neutral 5% polyacrylamide gel with
and without 5% (vol/vol) glycerol. Gels were dried and
exposed to X-ray ®lms at 7808C. DNA fragments
corresponding to the band were reampli®ed by PCR. PCR
products were puri®ed using a QIA quick-spin PCR
puri®cation kit (QIAGEN Inc.), and directly sequenced
with fmol DNA cycle sequencing system (Promega) or using
the 373S DNA sequencer with DNA sequencing kit (Applied
Biosystems). In PCR ± LOH analysis, the signal intensity of
polymorphic alleles was quanti®ed and calculated by the
scanning densitometer and data analysis system (The
Discovery Series, Quantity One, pdi, New York). LOH
was considered to be present if the radiographic signal of
one allele was at least 50% reduced in the tumor DNA when
compared to its corresponding normal allele.
RT ± PCR and Northern blot hybridization analyses of the
hOGG1 gene
Randomly primed cDNAs were reverse-transcribed from
0.5 mg of mRNAs using SuperScript II RNase H-Reverse
Transcriptase (Gibco BRL) in a 20 ml mixture according to
the supplier's protocol. One ml of the cDNA conversion
mixture was ampli®ed by PCR in a 10 ml reaction mixture
using a set of primers, 5'-TGAATTCGTCTTTGGGCGTCGACGA-3' and 5'-TGAATTCGGAAGGTGCTTGGGGAAT-3'. A hOGG1 cDNA clone (Arai et al.,
1997) was also subjected to this analysis as a control. After
35 cycles of 958C (60 s) for denaturation, 528C (60 s) for
annealing, 728C (90 s) for extention, PCR products were
fractionated by 1.0% agarose gel electrophoresis, and were
transferred to Hybond N-plus membranes (Amersham).
Southern blot hybridization was performed with oligonucleotides probes (5'-ACAGGCTTCTCAGGCTCAGT-3' in
inron 1, 5'-AAAGTCCTGCACACTGGA-3' in exon 2 and
5'-ACTGTCACTAGTCTCACCAG-3' in intron 6) labeled
with polynucleotide kinase according to the standard
procedure (Maniatis, 1982). Northern blot hybridization
analysis of the hOGG1 gene was performed as described
previously (Arai et al., 1997). Brie¯y, 3 mg of poly(A) RNA
were fractionated in 1% denaturing formaldehyde/agarose
gels and transferred to Hybond N-plus membranes
(Amersham). The membrane was sequentially hybridized
to the hOGG1 (nucleotide 554 ± 916 of hOGG1 cDNA) and
b-actin cDNA probes according to the standard procedure
(Maniatis 1982). b-actin cDNA was used for standardizing
the RNA amounts on the membrane.
Complementation of an E. coli (mutM mutY) mutant with
hOGG1-overproducing plasmid
Plasmids expressing GST-hOGG1-Ser326, GST-hOGG1Cys326 and GST-hOGG1-Gln46 fusion proteins were constructed by PCR ampli®cation of hOGG1 cDNA (Arai et
al., 1997), cDNAs synthesized from mRNA of PC10 and
NCI ± H526 cells using a set of primers, 5'-TGAATTCATGCCTGCCCGCGCGCTT-3' and 5'-TGAATTCCTAGCCTTCCGGCCCTTT-3'. PCR products were inserted
into the EcoRI site of pGEX ± 1lT (Pharmacia biotech),
and nucleotide sequences were determined on both strands
of the cloned inserts. Detailed protocol for the complementation assay was described previously (Arai et al., 1997).
Brie¯y, overnight cultures of YG5132 transformed with
each hOGG1-overproducing plasmid and of control were
analysed for rifampicin resistant mutation events. In each
group, two plates were prepared, and average numbers of
rifampicin resistant RifR colonies per 108 cells were
recorded. t-test was used for statistical analyses.
Quanti®cation of oh8Gua levels in lung cancer cell lines and
peripheral leukocytes DNA
Human peripheral blood samples were obtained from six
healthy male volunteers, who were homozygous either for
Cys or Ser at codon 326. They are never-smokers of 28 ± 33
years old, but no information was available on the history
of passive smoking. Bu€y coat fractions of blood samples
(20 ml) and lung cancer cell lines (36106 cells) were frozen
until used for the determination of the amount of oh8Gua
in DNA. Detailed protocol for the quanti®cation of
oh8Gua levels in DNA was described previously (Asami
et al., 1996). Brie¯y, DNA was extracted using the DNA
Extractor WB Kit (Wako junyaku, Tokyo, Japan), and
digested with nuclease P1 and acid phosphatase in a 10 mM
sodium acetate solution (378C for 30 min). After incubation, the mixture was treated with the ion exchange resin
Muromac (Muromachi kagaku, Tokyo, Japan) and was
centrifuged at 15 000 g for 5 min. The supernatant was
®ltrated by a ®lter tube (Millipore; Samprep C; 0.2 mm),
and injected onto an high-performance liquid chromatography column (Beckman; Ultrasphere-ODS; 5 mm,
4.66250 nm) equipped with an electrochemical detector
(ESA Coulochem II: detector 1, 0.15 V; detector 2, 0.30 V).
As standard samples, 20 ml each of deoxyguanosine
(0.5 mg/ml) and 8-hydroxydeoxyguanosine (5 ng/ml) solutions were injected. The value of oh8Gua was calculated as
the number per 105 guanine.
Acknowledgements
We thank the following scientists for providing cell lines:
Dr Y Hayata of Tokyo Medical College, Drs T Terasaki
Polymorphisms and alternative splicing of the hOGG1 gene
T Kohno et al
and S Hirohashi of National Cancer Center Research
Institute, Japan, Dr M Takada of Izumisano Municipal
Hospital. Cell lines were also obtained from ATCC. We
gratefully acknowledge Dr Shinya Asami of University of
Occupational and Environmental Health for excellent
technical assistance in the HPLC ± ECD analysis, and Dr
Shoichiro Tsugane of National Cancer Center Research
Institute, East for providing DNA samples from healthy
individuals. We also thank Manami Ishii and Mina
Takahashi for technical assistance. This work was
supported in part by a Grant-in-Aid for the 2nd-term
Comprehensive 10-Year Strategy for Cancer Control from
the Ministry of Health and Welfare, Grants-in-Aid from
the Ministry of Education, Science, Sports and Culture of
Japan, and from the Vehicle Racing Commemorative
Foundation. K Shinmura and M Tosaka are recipients of
a Research Resident Fellowship from the Foundation for
Promotion of Cancer Research.
References
Aburatani H, Hippo Y, Ishida T, Takashima R, Matsuba C,
Kodama T, Takao M, Yasui A, Yamamoto K, Asano M,
Fukusawa K, Yoshinari T, Inoue H, Ohtsuka E and
Nishimura S. (1997). Cancer Res., 57, 2151 ± 2156.
Arai K, Morishita K, Shinmura K, Kohno T, Kim S-R,
Nohmi T, Taniwaki S, Ohwada S and Yokota J. (1997).
Oncogene, 14, 2857 ± 2861.
Asami S, Hirano T, Yamaguchi R, Tomioka Y, Itoh H and
Kasai H. (1996). Cancer Res., 56, 2546 ± 2549.
Au KG, Clark S, Miller JH and Modrich P. (1989). Proc.
Natl. Acad. Sci. USA, 86, 8877 ± 8881
Baba S, Masago S, Takahashi T, Kasama T, Sugimura H,
Tsugane S, Tsutsui Y and Shirasawa H. (1995). Hum. Mol.
Genet., 4, 1083 ± 1087.
Bessho T, Tano K, Kasai H, Ohtsuka E and Nishimura S.
(1993). J. Biol. Chem., 268, 19416 ± 19421.
Boiteux S, Gajewski E, Laval J and Dizdaroglu M. (1992).
Biochemistry, 31, 106 ± 110.
Chabot B. (1996). Trends in Genetics, 12, 472 ± 478.
Gazder AF. (1994). Anticancer Res., 13, 261 ± 268.
Greenblatt MS, Bennett WP, Hollstein M and Harris CC.
(1994). Cancer Res., 54, 4855 ± 4878.
Grollman AP and Moriya M. (1993). Trends in Genetics, 9,
246 ± 249.
Hibi K, Takahashi T, Yamakawa K, Ueda R, Sekido Y,
Ariyoshi Y, Suyama M, Takagi H, Nakamura Y and
Takahashi T. (1992). Oncogene, 7, 445 ± 449.
Hosoe S, Shigedo Y, Ueno K, Tachibana I, Osaki T, Tanio
Y, Kawase I, Yamakawa K, Nakamura Y and Kishimoto
T. (1994). Lung Cancer, 10, 297 ± 305.
Huang CH, Reid M, Daniels G and Blumenfeld OO. (1993).
J. Biol. Chem. 268, 25902 ± 25908.
Garuti R, Lelli N, Barozzini M, Tiozzo R, Dotti MT,
Federico A, Ottomano AM, Croce A, Bertolini S and
Calandra S. (1996). J. Lipid Res., 37, 1459 ± 1467.
Kang D, Nishida J, Iyama A, Nakabeppu Y, Furuichi M,
Fujiwara T, Sekiguchi M and Takeshige K. (1995). J. Biol.
Chem., 270, 14659 ± 14665.
Kasai H and Nishimura S. (1991). Oxidative stress: Oxidants
and Antioxidant. Academic Press: London. pp. 99 ± 116.
Kohno T, Otsuka T, Takano H, Yamamoto T, Hamaguchi
M, Terada M and Yokota J. (1995). Hum. Mol. Genet., 4,
667 ± 674.
Lu R, Nash HM and Verdine GL. (1997). Cur. Biol., 7, 397 ±
407.
Maki H and Sekiguchi M. (1992). Nature, 355, 273 ± 275.
Malins DC and Haimanot R. (1991). Cancer Res., 51, 5430 ±
5432.
Maniatis T, Fritsch FF and Sambrook J. (1982). Molecular
Cloning: A Laboratory Manual. Cold Spring Harbor Press:
New York.
Michaels ML, Cruz C, Grollman AP and Miller JH. (1992).
Proc. Natl. Acad. Sci. USA, 89, 7022 ± 7025.
Michaels ML and Miller JH. (1992). J. Bacteriology, 174,
6321 ± 6325.
Michaels ML, Pham L, Cruz C and Miller JH. (1991).
Nucleic Acids Res., 19, 3629 ± 3632.
Nakajima M, Takeuchi T and Morimoto K. (1996).
Carcinogenesis, 17, 787 ± 791.
Nash HM, Bruner SD, Scharer OD, Kawate T, Addona TA,
Spooner E, Lane WS, Verdine GL, van der Kemp PA,
Thomas D, Barbey R, de Oliveira R and Boiteux S. (1996).
Curr. Biol., 6, 968 ± 980.
Otsuka T, Kohno T, Mori M, Noguchi M, Hirohashi S and
Yokota J. (1996). Genes Chrom. Cancer, 16, 113 ± 119.
Perera FP. (1996). J. Natl. Cancer Inst, 88, 496 ± 509.
Radicella JP, Dherin C, Desmaze C, Fox MS and Boiteux S.
(1997). Proc. Natl. Acad. Sci. USA, 94, 8010 ± 8015.
Roldan-Arjona T, Wei Y, Carter KC, Klungland A,
Anselmino C, Wang R, Augustus M and Lindahl T.
(1997). Proc. Natl. Acad. Sci. USA, 94, 8016 ± 8020.
Rosenquist TA, Zharkov DO and Grollman AP. (1997).
Proc. Natl. Acad. Sci. USA, 94, 7429 ± 7434.
Sameshima Y, Matsuno Y, Hirohashi S, Shimosato Y,
Mizoguchi H, Sugimura T, Terada M and Yokota J.
(1992). Oncogene, 7, 451 ± 457.
Sekido Y, Bader S, Latif F, Gnarra JR, Gazder AF, Linehan
WM, Zbar B, Lerman MI and Minna JD. (1994).
Oncogene, 9, 1599 ± 1604.
Shiseki M, Kohno T, Nishikawa R, Sameshima Y,
Mizoguchi H and Yokota J. (1994). Cancer Res., 54,
5643 ± 5648.
Shiseki M, Kohno T, Adachi J, Okazaki T, Otsuka T,
Mizoguchi H, Noguchi M, Hirohashi S and Yokota J.
(1996). Genes Chrom. Cancer, 17, 71 ± 77.
Shinmura K, Kasai H, Sasaki A, Sugimura H and Yokota J.
(1997). Mut. Res., 385, 75 ± 82.
Suzuki M, Matsui K, Yamada M, Kasai H, Sofumi T and
Nohmi T. (1997). Mut. Res., in press.
Tchou J, Kasai H, Shibutani S, Chung MH, Laval J and
Grollman ANS. (1991). Proc. Natl. Acad. Sci. USA, 88,
4690 ± 4694.
van der Kemp PA, Thomas D, Barbey, R, de Oliveira R and
Boiteux S. (1996). Proc. Natl. Acad. Sci. USA, 93, 5197 ±
5202.
Yokota J and Sugimura T. (1993). FASEB J., 7, 920 ± 925.
3225