1394 Vol.
6, 1394 –1400, April 2000
Overexpression of the hOGG1 Gene and High 8-Hydroxy-2ⴕ-
deoxyguanosine (8-OHdG) Lyase Activity in Human
Colorectal Carcinoma: Regulation Mechanism of
the 8-OHdG Level in DNA1
Shohei Kondo, Shinya Toyokuni,2
Tomoyuki Tanaka, Hiroshi Hiai,
Hisashi Onodera, Hiroshi Kasai, and
Masayuki Imamura
Departments of Pathology and Biology of Diseases [S. K., S. T.,
T. T., H. H.] and Surgery and Surgical Basic Science [S. K., H. O.,
M. I.], Graduate School of Medicine, Kyoto University, Kyoto 606-
8501, Japan, and Department of Environmental Oncology, University
of Occupational and Environmental Health, Kita-Kyusyu 807-8555,
Japan [H. K.]
ABSTRACT
8-Hydroxy-2ⴕ-deoxyguanosine (8-OHdG) is one of the
most abundant oxidatively modified lesions in DNA. Our
previous study (Kondo et al., Free Radic. Biol. Med., 27:
401– 410, 1999) revealed that human colorectal carcinoma
cells are oxidatively stressed based on 8-OHdG determina-
tion. To elucidate 8-OHdG metabolism and its clinical sig-
nificance in colorectal carcinoma, we studied the 8-OHdG
repair system in DNA by measuring specific lyase activity
and hOGG1 expression using quantitative-competitive re-
verse transcription-PCR. In addition, we searched for the
presence of mutations and single nucleotide polymorphisms
of the hOGG1 gene by single-strand conformational poly-
morphism and sequencing analyses. It was found that
8-OHdG-specific lyase activity and hOGG1 expression were
significantly up-regulated in carcinoma, and a proportional
association between 8-OHdG levels and either 8-OHdG
lyase activity (r ⴝ 0.641, P < 0.05) or hOGG1 expression
(r ⴝ 0.702, P < 0.05) was present. Whereas no difference
was detected in the 8-OHdG level between early- and ad-
vanced-stage cancer, lyase activity (1.2-fold) and hOGG1
expression (1.6-fold) were significantly increased in ad-
vanced-stage cancer. No mutation was found in the 25 tu-
Received 9/20/99; revised 1/11/00; accepted 1/17/00.
The costs of publication of this article were defrayed in part by the
payment of page charges. This article must therefore be hereby marked
advertisement in accordance with 18 U.S.C. Section 1734 solely to
indicate this fact.
1
Supported in part by a Grant-in-Aid from the Japanese Ministry of
Education, Science, Sports and Culture, a grant from the Japanese
Owner’s Association, and a grant from the program for Promotion of
Basic Research Activities for Innovative Bioscience.
2 3
To whom requests for reprints should be addressed, at the Department
of Pathology and Biology of Diseases, Graduate School of Medicine,
Kyoto University, Yoshida-Konoe-cho, Sakyo-ku, Kyoto 606-8501, Ja-
pan. Phone: 81-75-753-4423; Fax: 81-75-753-4432; E-mail: toyokuni@
path1.med.kyoto-u.ac.jp.
Clinical Cancer Research
mors examined. Three kinds of single nucleotide polymor-
phism were observed, including that of codon 326 (Ser/Cys)
in exon 7. However, there was no correlation between any of
the three polymorphic patterns and either 8-OHdG level or
lyase activity. These results suggest that increased 8-OHdG
levels in colorectal carcinoma are attributed to increased
formation and are maintained by induced 8-OHdG repair
activity at appropriate high levels. Our results may offer a
unique approach in the development of preventive and ther-
apeutic interventions as well as new insights into the patho-
genesis of colorectal carcinoma.
INTRODUCTION
Oxidative stress caused by ROS3 plays a role in a variety of
tumor-associated cellular events such as carcinogenesis, UV- or
radiation-induced damage, chemotherapy-induced cytotoxicity,
and reperfusion injury (1, 2). Thus far, oxidative stress in tumor
cells has not attracted much attention.
In 1991, Szatrowski and Nathan (3) first reported that some
human cancer cell lines can produce large amounts of H2O2.
Other studies, including our own, have concluded that several
kinds of human cancer tissues such as lung (4), renal (5), and
colorectal carcinoma (6, 7) show higher levels of DNA oxida-
tion compared to their nontumorous counterparts, based on
8-OHdG determination. 8-OHdG, one of the major DNA base-
modified products, is induced by either hydroxyl radical, singlet
oxygen, or photodynamic action (8), and its pairing with ade-
nine as well as cytosine is known to be mutagenic, leading to
G:C to T:A transversion on DNA replication (9).
We (6) have previously found that colorectal carcinoma but
not adenoma cells are exposed to more oxidative stress than the
corresponding nontumorous epithelial cells. Furthermore, im-
munohistochemical methods revealed that the oxidative stress in
carcinoma cells is associated with cellular proliferation. Based
on these observations, we hypothesized that cancer cells are
under “persistent oxidative stress” (10). However, no unifying
concept has emerged thus far to explain the general metabolism
of oxidatively modified DNA products in tumor cells. Oxidative
stress might be defined as “ROS load minus total antioxidant
activities.” Total antioxidant activities are hypothetical concepts
including a variety of antioxidants and antioxidant enzymes
associated with glutathione, thioredoxin, or direct ROS metab-
The abbreviations used are: ROS, reactive oxygen species; 8-OHdG,
8-hydroxy-2⬘-deoxyguanosine; HPLC/ECD, high-pressure liquid chro-
matography/electrochemical detector; RT-PCR, reverse transcription-
PCR; SSCP, single-strand conformation polymorphism; SNP, single
nucleotide polymorphism.

olism as well as different kinds of repair mechanisms for mod-
ified molecules (11).
In the present study, we focused on 8-OHdG metabolism in
the DNA of human colorectal carcinoma to determine whether
any deviation from nonneoplastic epithelial cells is present.
8-OHdG must be cleaved from the DNA strand to maintain
genomic integrity before subsequent replication. For this proc-
ess, activities of specific glycosylase and lyase have been
known. The responsible gene was recently cloned and assigned
as human OGG1 or MutM homologue gene (12–16). It removes
8-oxoguanine in double-stranded DNA by glycosylase activity
and further excises a phosphodiester bond at 3⬘ end of the
apurinic site by apurinic lyase activity via a ␤-elimination
reaction (12, 13, 16). Thus, we measured 8-OHdG-specific lyase
activity and hOGG1 expression in colorectal carcinoma and in
its nontumorous counterpart, with a subsequent analysis of their
correlation with clinical factors. Furthermore, we searched for
the presence of SNPs and mutations in the hOGG1 gene of
colorectal carcinoma because mutations in lung and renal car-
cinoma (17) as well as polymorphisms (18) were reported
recently.
MATERIALS AND METHODS
Tissue Samples. Twenty-five sporadic primary colorec-
tal adenocarcinoma specimens that were surgically resected at
the First Department of Surgery, Kyoto University Hospital or
its affiliated hospitals during the period between January 1997
and February 1999 were evaluated in this study. Histopatholog-
ical examination was carried out by two independent registered
pathologists (S. T. and H. H.) on H&E-stained sections. Both
the histological type and clinical stage of all of the specimens
were classified according to the WHO (19) and tumor-node-
metastasis (TNM) classification system (20), respectively. None
of the patients received chemotherapy or radiation before sur-
gery. Cancerous tissue, together with corresponding normal
mucosal tissue, was dissected separately immediately after sur-
gical resection. Control colorectal mucosal tissue was obtained
from an area at least 10 cm away from the margin of the
carcinoma. Thereafter, serosa and proper muscle layers were
carefully removed by using scissors. The specimens were snap-
frozen in liquid nitrogen and stored at ⫺80°C until RNA, DNA,
and protein extractions.
8-OHdG Determination by HPLC/ECD. 8-OHdG was
assayed by HPLC/ECD according to the method of Helbock et
al. (21), with minor modifications. To minimize artifactual
8-OHdG generation during sample preparation, DNA extraction
was achieved by using the chaotropic NaI method, and an iron
chelate, deferoxamine mesylate (Sigma Chemical Co., St. Louis,
MO), was added to suppress iron-catalyzed Fenton reaction.
Nuclear DNA was extracted from colorectal tissue (approxi-
mately 200 mg) using a DNA Extractor WB kit (Wako, Osaka,
Japan), denatured at 95°C for 5 min, and digested with 10 ␮l of
nuclease P1 (20 units/␮l; Sigma) in 10 mM sodium acetate
buffer (pH 4.5) at 37°C for 1 h, followed by incubation with 4
␮l of alkaline phosphatase (100 milliunits/␮l; Sigma) in 40 mM
Tris buffer (pH 8.5) at 37°C for 1 h. The solution was then
treated with an ion exchange resin (Muromac; Muromachi Ka-
gaku, Tokyo, Japan) to remove NaI and centrifuged at 5,000 ⫻
Clinical Cancer Research 1395
Table 1 Substrate and markers for 8-OHdG lyase activity assay
Substrate/marker Size
Substrate
Fa-GGTGGCCTGACGbCATTCCCCAA 22 bp
CCACCGGACTGCGTAAGGGGTT
Single-strand markers
F-GGTGGCCTGAC 11-mer
F-GGTGGCCTGACG 12-mer
a
F, FITC.
b
G, 8-OHdG.
g for 5 min. The supernatant was further filtered through a
0.22-␮m filter membrane (Millipore, Bedford, MA) and centri-
fuged again at 15,000 ⫻ g for 5 min. The filtrate was then
injected onto a HPLC column (ODS-80Ts; Tosoh, Tokyo, Ja-
pan) as described previously (22). 2⬘-Deoxyguanosine (Sigma)
and 8-OHdG (Wako) solutions were used as standards. The
molar ratio of 8-OHdG per 105 deoxyguanosine (8-OHdG/105
dG) was calculated.
8-OHdG-specific DNA Lyase Activity. The 8-OHdG-
specific DNA lyase activity assay was performed according to
the method of Yamamoto et al. (23), with minor modifications.
Colorectal tissue was homogenized with 5 volume/weight of
buffer A [50 mM Tris-HCl (pH 7.4), 50 mM KCl, 3 mM EDTA,
5 mM magnesium acetate, and 3 mM ␤-mercaptoethanol] con-
taining protease inhibitors (5 ␮g/ml each of antipain, pepstatin
A, chymostatin, and leupeptin). The homogenate was centri-
fuged at 25,000 ⫻ g for 1 h at 4°C, and the cytosolic fraction
was obtained. Total protein concentration was determined using
the BCA protein assay kit (Pierce, Rockford, IL) and adjusted to
5 mg/ml in each sample. Table 1 illustrates the substrate and size
markers for the 8-OHdG-specific lyase activity assay. The sub-
strate is a double-stranded 22-bp oligonucleotide containing one
8-OHdG paired with deoxycytidine at the complementary
strand, and it was labeled with FITC. The cytosolic extract (50
␮g of protein) was incubated with 40 fmol of the DNA substrate
at 30°C for the indicated period of time. After two ethanol
precipitations, the pellet was dried and dissolved in 10 ␮l of
loading buffer (80% formamide, 10 mM NaOH, and 1 mM
EDTA). The solution was then denatured by heating at 95°C for
5 min. Sample solution (4 ␮l) was applied to a 20% denaturing
polyacrylamide gel containing 8 M urea in 1⫻ Tris-borate
EDTA buffer and run at 10 W for 30 min at room temperature.
After electrophoresis, the fluorescence intensity of each band
was measured using FMBio (Takara, Shiga, Japan). The nicked
substrate size was identified by coelectrophoresis of the 11-mer
and 12-mer oligonucleotides used as markers. The lyase activity
was calculated from the ratio of the excised fragment:total
substrate. Each assay was performed in duplicate.
RNA Preparation and Reverse Transcription Reaction.
Total RNA was extracted from frozen tumors as well as nontu-
morous counterparts by means of a modified acid guanidinium
phenol chloroform method (Isogen; Nippon Gene, Tokyo, Ja-
pan) according to the manufacturer’s protocol. cDNA was pre-
pared from 5 ␮g of total RNA by random priming using a
first-strand cDNA synthesis kit (Amersham Pharmacia Biotech,
Tokyo, Japan).
1396 Regulation of Oxidative Stress in Cancer
PCR Primers and Competitor Synthesis. PCR primers
were synthesized to amplify a 300-bp fragment of hOGG1
cDNA. The forward primer, SK1 (5⬘-ACACTGGAGTGGTG-
TACTAGCG-3⬘) was situated in exon 3, and the reverse primer,
SK2 (5⬘-GCCGATGTTGTTGTTGGAGG-3⬘), was situated in
exon 5. The PCR product spanned three exons to differentiate
the amplified products from genomic DNA. A heterologous
competitor for competitive RT-PCR was prepared using the
competitive DNA construction kit according to the manufactur-
er’s protocol (Takara). Briefly, each 3⬘-terminal site of SK1 and
SK2 primer was ligated with two additional 20-bp oligonucleo-
tides that had an annealing sequence to ␭-phage DNA template
spanning a 300-bp region. PCR amplification using these elon-
gated primers produced a 341-bp competitor fragment with the
same sequence as SK1 or SK2 primer. The PCR (MP3000;
Takara) conditions were as follows: denaturation at 94°C for 4
min followed by 30 cycles of denaturation at 94°C for 40 s,
annealing at 62°C for 40 s, and extension at 72°C for 40 s, with
a final extension reaction for 4 min. The competitor was serially
diluted to produce 103, 104, 105, 106, and 107 copies/␮l.
Quantitative RT-PCR Analysis. hOGG1 expression
was analyzed using a quantitative competitive PCR method with
the purified competitor. In the presence of PCR buffer [1.5 mM
MgCl2, 50 mM KCl, and 10 mM Tris-HCl (pH 8.3)], 0.25 mM
deoxynucleotide triphosphate, 5 pmol of SK1 and SK2 primer,
and 0.5 unit of Taq DNA polymerase (Takara), the reaction
mixture contained 1 ␮l of cDNA solution derived from 1 ␮g of
total RNA and 1 ␮l competitor of indicated dilution in a total
volume of 12.5 ␮l. PCR conditions were as follows: denatur-
ation at 94°C for 4 min followed by 28 cycles of denaturation at
94°C for 40 s, annealing at 60°C for 40 s, and extension at 72°C
for 40 s, with a final extension for 4 min. PCR products were
separated on 8% polyacrylamide gel, and the bands were visu-
alized by ethidium bromide staining. For quantitation, absorb-
ance of each band at 605 nm was measured using FMBio. Copy
number of the target cDNA was calculated by finding the
condition of equivalent PCR amplification of the original and
competitor products; namely, the intensity of each product was
plotted against the known copy number of the competitor, and
the equivalent point was determined as an intersection of the
two lines.
PCR-SSCP Analysis of the hOGG1 Gene. Mutation
and polymorphism in the hOGG1 gene were screened by PCR-
SSCP. Six pairs of primers for exons 1–7 of the hOGG1 gene
were used for SSCP analysis as described by Kohno et al. (18).
PCR amplification was carried out in a total reaction volume of
12.5 ␮l containing 50 ng of genomic DNA of sample tissue, 5
pmol of each of the primer pairs, 1.5 mM MgCl2, 50 mM KCl, 10
mM Tris-HCl (pH 8.3), 0.25 mM deoxynucleotide triphosphate,
and 0.5 unit of Taq DNA polymerase (Takara). PCR conditions
were as follows: denaturation at 94°C for 4 min followed by 30
cycles of denaturation at 94°C for 60 s, annealing at 55°C for
60 s, and extension at 72°C for 90 s, with a final 10-min
extension. After amplification, 2 ␮l of the product were elec-
trophoresed on 4% NuSieve GTG agarose gel (FMC Bioprod-
ucts, Rockland, ME) to confirm specific amplification of the
targeted fragment. The second PCR amplification was then
carried out in a total volume of 10 ␮l, in which 0.1 ␮l of the first
PCR product was used as a template, and 0.1 ␮l of [␣-32P]dCTP
Table 2 Summary of 8-OHdG level, 8-OHdG lyase activity, and
hOGG1 transcript amount in colorectal carcinoma and corresponding
nontumorous counterpart
Results are presented as means ⫾ SE (n ⫽ 25).
8-OHdG Amount of hOGG1
8-OHdG level lyase activity transcript
(8-OHdG/105dG) (%) (Copies/␮g RNA)
Na 0.64 ⫾ 0.05 19.72 ⫾ 1.13 4.55 ⫾ 0.54 ⫻ 105
b b b
T 1.34 ⫾ 0.11 36.16 ⫾ 1.79 30.58 ⫾ 3.21 ⫻ 105
a
N, nontumoros counterpart, T, carcinoma.
b
P ⬍ 0.001, Wilcoxon signed rank test.
(3000 Ci/mmol; Amersham, Arlington Heights, IL) in the pres-
ence of 0.025 mM dCTP was used instead of 0.25 mM. PCR
products were mixed with 100 ␮l of loading buffer (95% form-
amide, 20 mM EDTA, 0.05% xylene cyanol, and 0.05% brom-
phenol blue). Before loading, samples were denatured by incu-
bation at 95°C for 5 min and cooled on ice. Each sample (2 ␮l)
was loaded onto 6% nondenaturing polyacrylamide gel with and
without 10% glycerol and electrophoresed at 40 W for 4 – 6 h,
followed by autoradiography.
DNA Sequencing of the hOGG1 Gene. The detected
band of altered mobility from SSCP analyses was further eval-
uated by sequencing analyses. DNA fragments extracted from
excised bands of altered mobility were reamplified by PCR with
the same pair of primers and subcloned into the PCR II vector
using the TA cloning kit (Invitrogen, San Diego, CA). Sequenc-
ing was performed using an ABI Prizm 377 DNA sequencer
(Perkin-Elmer, Branchburg, NJ), and the results were compared
with the original DNA sequence.
Statistical Analyses. Results of 8-OHdG-specific lyase
activity and hOGG1 expression were analyzed by Wilcoxon’s
signed rank test. Comparison between the unpaired two groups
was achieved by using the Mann-Whitney U test. Correlation
between the two variables was estimated by Pearson’s correla-
tion coefficient. P ⬍ 0.05 was considered statistically signifi-
cant.
RESULTS
8-OHdG Determination by HPLC/ECD. The 8-OHdG
levels in the DNA of colorectal carcinoma were significantly
higher than those of the nontumorous counterparts (P ⬍ 0.001),
as described in our pervious report (Ref. 6; Table 2).
8-OHdG-specific Lyase Activity. We initially studied
the time course of the reaction at 30°C. During the first 120 min,
the fluorescence intensity of the excised substrate (12-mer)
increased constantly. However, thereafter it gradually plateaued
in the reactions of both carcinoma and its nontumorous coun-
terpart. The entire time course up to 24 h was consistent with a
semilogarithmic correlation, as shown in Fig. 1. We selected a
60-min incubation period for this assay. A representative elec-
trophoretic pattern is shown in Fig. 2. Colorectal carcinoma
revealed higher 8-OHdG-specific lyase activity than its nontu-
morous counterpart (P ⬍ 0.001; Table 2).
Optimization of Quantitative Competitive RT-PCR.
We first tested the effect of PCR cycles on the amount of
products as follows: coamplification of hOGG1 cDNA derived

Fig. 1 Time course study of 8-OHdG-specific lyase activity. A semi-
logarithmic correlation was observed. See “Materials and Methods” and
“Results” for details. Results are presented as the means ⫾ SE.
Fig. 2 Representative electrophoretic pattern of 8-OHdG-specific
lyase activity. A cytosolic fraction of the tissue was used as the enzyme,
and 5⬘ fluorescence-labeled double-stranded oligonucleotide containing
one 8-OHdG (Table 1) was used as the substrate. See “Materials and
Methods” for details. 4, 7, 8, and 14, case numbers of four paired cases
of colorectal cancer and nontumorous counterparts. M11, 11-mer single-
strand marker; M12, 12-mer single-strand marker; E(⫺), no cytosolic
fraction was added to the substrate; T, colorectal carcinoma tissue; N,
nontumorous mucosal tissue.
from 1 ␮g of total RNA and 104 copies of the prepared com-
petitor fragment was performed for various PCR cycles between
cycles 24 and 42 by competitive PCR. The two products were
amplified with the same efficiency, based on the result that the
ratio of target cDNA-derived product:competitor-derived prod-
uct was constant for each of the examined PCR cycles (Fig. 3).
Twenty-two cycles of PCR amplification were necessary to
detect hOGG1 products by fluorescence analyses after ethidium
bromide staining. The amplification plateaued after 36 cycles
and lost the proportional relationship with the substrate level.
Quantitation of hOGG1 mRNA. The amount of target
cDNA was determined on the basis of selecting an equivalent
amount of the competitor PCR signal. Fig. 4 is a representative
example of hOGG1 expression detected by quantitative com-
petitive RT-PCR. The amount of hOGG1 transcript in nontu-
morous counterpart and carcinoma was 4.55 ⫾ 0.54 ⫻ 105 and
30.58 ⫾ 3.21 ⫻ 105 copies/␮g RNA, respectively, and the
difference was statistically significant (P ⬍ 0.001; Table 2).
When 8-OHdG levels of carcinoma, together with those of their
nontumorous counterparts, were plotted against either 8-OHdG-
specific lyase activity or hOGG1 transcript amount, a propor-
tional association was observed between the 8-OHdG content
Clinical Cancer Research 1397
Fig. 3 Kinetics of coamplification of the hOGG1 gene and a compet-
itor DNA. cDNA reverse-transcribed from 1 ␮g of total RNA and
competitor (104 copies) was coamplified using the same primer pair for
24 – 42 cycles at increments of 3 cycles to evaluate the amplification
efficiency. The densitometric values from target (T) and competitor (C)
were plotted against the number of amplification cycles and used for
calculation of the T:C ratio for each cycle. The T:C ratio was constant.
After 36 cycles, PCR amplification plateaued. See “Results” for details.
Fig. 4 Competitive RT-PCR analysis of hOGG1 expression in colo-
rectal carcinoma and nontumorous mucosal tissue from the same pa-
tient. An increasing amount of target cDNA was coamplified with serial
dilutions of known amounts of the competitor fragment. The 300-bp
(target cDNA) and 341-bp (competitor) PCR products were analyzed on
an 8% polyacrylamide gel. The signals of target cDNA and competitor
were determined by measuring the intensity of ethidium bromide fluo-
rescence. An aliquot of the nontumorous counterpart contains 7.94 ⫻
105 copies/␮g RNA, whereas an aliquot of the carcinoma sample con-
tains 2.52 ⫻ 106 copies/␮g RNA, as determined by calculating the
equivalent point (Œ) of both PCR products of target DNA or the
competitor.
and either 8-OHdG-specific lyase activity or hOGG1 transcript
amount by Pearson’s correlation coefficient [r ⫽ 0.642 and P ⬍
0.05 (Fig. 5A) and r ⫽ 0.702 and P ⬍ 0.05 (Fig. 5B), respec-
tively].
To examine whether there is an association between the
levels of 8-OHdG-specific lyase activity or hOGG1 expression
and clinical staging, 25 cases of colorectal carcinoma were
divided into two groups according to the TNM staging classi-
fication: (a) the early-stage group (stage 0, stage I, and stage II;
n ⫽ 10); and (b) the advanced-stage group (stage III and stage
IV; n ⫽ 15). A significant difference was detected between the
two clinical groups for 8-OHdG-specific lyase activity and
hOGG1 expression by the Mann-Whitney U test (P ⫽ 0.03 and
P ⫽ 0.02, respectively; Fig. 6). However, 8-OHdG levels of
1398 Regulation of Oxidative Stress in Cancer

Fig. 5 8-OHdG-specific lyase activity (A) and

expression of hOGG1 transcript (B) in colorectal
tissues are proportionally associated with 8-OHdG
level. Correlations were evaluated statistically by
Pearson’s correlation coefficient (r). A, r ⫽ 0.641,
P ⬍ 0.05; B, r ⫽ 0.702; P ⬍ 0.05.

Fig. 6 Comparison between early-stage and ad-

vanced-stage colorectal carcinoma for (A)
8-OHdG-specific lyase activity and (B) hOGG1
expression. Results are presented as the means ⫾
SE, with P determined by the Mann-Whitney U
test. ⴱ, P ⬍ 0.05; ⴱⴱⴱ, P ⬍ 0.001.

Table 3 Genotype in exon 7 and 8-OHdG repair enzyme activity (25). There is an increasing interest in genomic instability as an
8-OHdG lyase activity (%) important factor for cancer promotion and progression as well as
carcinogenesis (26, 27). In particular, DNA repair systems for
Frequency Normal replication error have been studied extensively since the eluci-
Genotype (%) mucosa Carcinoma
dation of a gene responsible for hereditary nonpolyposis colo-
Ser326/Ser326 32 19.8 ⫾ 1.2  32.5 ⫾ 2.3 
a a rectal cancer (28).
Ser326/Cys326 44 18.6 ⫾ 1.4 
34.5 ⫾ 1.6 
Cys326/Cys326 24 18.9 ⫾ 2.1  32.1 ⫾ 1.4 
We have been working on the metabolism of cancer cells
a from the standpoint of ROS to develop novel therapeutic as well
Not statistically significant.
as preventive approaches to human cancers. We have reported
previously that higher levels of ROS-modified products,
8-OHdG, 4-hydroxy-2-nonenal histidine adducts, and 3-nitroty-
early-stage carcinoma and those of advanced-stage carcinoma rosine were observed in colorectal carcinoma cells (but not in
were not statistically different (means ⫾ SE, 1.31 ⫾ 0.13/105dG adenoma cells) than in the nontumorous counterparts, leading to
and 1.28 ⫾ 0.05/105dG, respectively). the conclusion that carcinoma cells are more oxidatively
SSCP Analysis and Sequencing of the hOGG1 Gene. stressed (6). Several reports have indicated the presence of
No mutation was detected in the 25 tumors. However, polymor- higher levels of 8-OHdG in the DNA of cancer cells than in that
phisms were observed in the 5⬘-noncoding region (⫺18 bp from of nontumorous counterparts (5, 7, 29, 30). In the present study,
the first codon; G/T), exon 1 (codon 98; G/A), and exon 7, in we attempted to determine: (a) whether higher levels of
agreement with previous reports (18, 24). Among these poly- 8-OHdG in cancerous DNA result from increased generation of
morphisms, only the polymorphism in exon 7 was accompanied 8-OHdG or from decreased repair of this oxidative lesion; (b)
by an amino acid substitution. However, there was no difference whether there is a mutation in the hOGG1 gene of colorectal
in the 8-OHdG-specific lyase activity among the three geno- carcinoma; and (c) whether polymorphism of hOGG1 is asso-
types as shown in Table 3. ciated with 8-OHdG-specific lyase activity.
Our results showed clearly that hOGG1 expression and
DISCUSSION 8-OHdG-specific lyase activity are increased in carcinoma cells.
Major advances in discovering genetic alterations of hu- A proportional correlation was obtained between the 8-OHdG
man colorectal cancer have been made during the last decade level and hOGG1 expression or 8-OHdG-specific lyase activity.

This could represent an appropriate feedback mechanism. An
increase in 8-OHdG-specific lyase activity under oxidative
stress has been reported under several pathological conditions
such as smoking (31), ionizing irradiation (32), and exposure to
transition metal (33). Moreover, levels of 8-OHdG were directly
proportional to the lyase activity in these studies, whereas oxi-
dative stress of these kinds is usually temporary. In cancer,
however, oxidative stress is persistent, associated with its own
metabolic activity (10).
Therefore, our data demonstrate that an increased load of
ROS is responsible for higher levels of 8-OHdG in colorectal
carcinoma. Levels of oxidative stress in colorectal carcinoma
appear to be suitable for both cellular proliferation and genomic
instability. It is established that low levels of oxidative stress
promote cellular proliferation instead of inducing apoptosis and
necrosis (11). We also found that hOGG1 expression and
8-OHdG-specific lyase activity were associated with the clinical
stage of colorectal carcinoma (Fig. 6). These results, together
with the finding that 8-OHdG levels did not differ between
early- and advanced-stage cancers, suggest that carcinoma cells
adjust the levels of oxidative stress to a suitable level by mod-
ulating the repair activity. It was also reported that the human
MutT homologue, an enzyme for hydrolyzing 8-hydroxy-dGTP
in the nucleotide pool, is overexpressed in human renal cell
carcinoma or in lung cancer cell lines (5, 34).
Whereas hOGG1 expression levels were increased by
4 –11-fold in colorectal carcinoma, 8-OHdG-specific lyase ac-
tivity showed an increase of only 1.5–3.0-fold. It seems that
8-OHdG-specific lyase activity does not reflect hOGG1 expres-
sion levels. There are three possible explanations for this. First,
an inhibitor of hOGG1 protein or an 8-OHdG-binding protein in
the cytosol might be present (35, 36). Second, we may have
underestimated the enzyme activity because of the different
localization of each isoform. hOGG1 is known to possess four
major alternative splicing isoforms [type1a, type 1b, type 1c,
and type 2 (16, 37)], with each type distributed to different
organelles. The main hOGG1 isoform is type 1a, which contains
nuclear transfer signal sequence and is localized in the nucleus
after translation (16, 37), and the other hOGG1 types without
the nuclear transfer signal (type 1b, type 1c, and type 2) might
be localized in intracellular organelles. We thought that using
the cytosolic compartment as an enzyme source in the lyase
activity assay rather than using crude extract containing nuclear
protein was a better choice because the system excludes nuclear
DNA fragments. On the other hand, transcripts of all of the
hOGG1 mRNA isoforms were amplified in the RT-PCR reac-
tion due to the presence of common primer annealing site in
hOGG1 gene. It remains unclear how different isotypes of
hOGG1 respond to oxidative stress at the transcriptional level.
Third, the half-life of hOGG1 mRNA or protein might be
different. In a recent report, Chevillard et al. (17) showed that
there was no significant difference in hOGG1 expression be-
tween tumorous and nonneoplastic tissues in human kidney
cancer . Although the reason for this discrepancy is not clear,
this difference might be due to an organ-specific tumor metab-
olism.
The hOGG1 gene was found to be localized on chromo-
some 3p25 (12, 13, 16). Because several kinds of human carci-
noma reveal a frequent allelic deletion at chromosome 3p, the
Clinical Cancer Research 1399
hOGG1 gene has been a good candidate for a responsible tumor
suppressor gene. However, a small percentage of mutations
detected were in carcinomas of the lung (17, 18), kidney (17),
and stomach (24). There was no hOGG1 mutation in the 25
cases of colorectal carcinoma in the present study. Three kinds
of SNPs were observed, and the patterns and frequencies were
in accordance with a recent study (18). Reports indicate the
presence of a C/G SNP at codon 326 in exon 7, with serine/
cysteine amino acid substitution and higher 8-OHdG-specific
lyase activity in purified hOGG1-Ser326 protein than in purified
hOGG1-Cys326 (18, 24). In this study, no significant difference
in lyase activity was found in any of the hOGG1 genotypes
(Ser326/Ser326, Cys326/Ser326, and Cys326/Cys326) of carcinoma
cells or in the noncancerous counterparts (Table 3). It is possible
that there is another enzyme responsible for 8-OHdG-specific
DNA lyase activity, hence additional careful studies are neces-
sary.
We have reached the following conclusions: (a) increased
load of ROS is responsible for high 8-OHdG levels in human
colorectal carcinoma; (b) there is a general proportional corre-
lation between 8-OHdG levels and 8-OHdG-specific lyase ac-
tivity or hOGG1 expression; (c) however, 8-OHdG levels are
maintained at a constant high level among early- and advanced-
stage colorectal carcinoma, and this adjustment appears to be
regulated by hOGG1 transcription; and (d) no mutation in
hOGG1 was detected in the present study. Furthermore, poly-
morphism of the gene was independent of 8-OHdG-specific
lyase activity. Further understanding of tumor metabolism in
“persistent oxidative stress” might facilitate the development of
a new therapeutic and preventive strategy for colorectal carci-
noma.
ACKNOWLEDGMENTS
We thank the Surgical Department of Kyoto Police Hospital for
providing us with the colorectal specimens and Noriko Shibata (Depart-
ment of Pathology and Biology of Diseases, Graduate School of Med-
icine, Kyoto University, Kyoto, Japan) for technical assistance.
REFERENCES
1. Halliwell, B., and Gutteridge, J. M. C. Free Radicals in Biology and
Medicine, 3rd ed. Oxford, United Kingdom: Clarendon Press, 1999.
2. Toyokuni, S. Oxidative stress and cancer: the role of redox regula-
tion. Biotherapy, 11: 147–154, 1998.
3. Szatrowski, T. P., and Nathan, C. F. Production of large amounts of
hydrogen peroxide by human tumor cells. Cancer Res., 51: 794 –798,
1991.
4. Jaruga, P., Zastawny, T. H., Skokowski, J., Dizdaroglu, M., and
Olinski, R. Oxidative DNA base damage and antioxidant enzyme ac-
tivities in human lung cancer. FEBS Lett., 341: 59 – 64, 1994.
5. Okamoto, K., Toyokuni, S., Uchida, K., Ogawa, O., Takenawa, J.,
Kakehi, Y., Kinoshita, H., Hattori-Nakakuki, Y., Hiai, H., and Yoshida,
O. Formation of 8-hydroxy-2⬘-deoxyguanosine and 4-hydroxy-2-
nonenal-modified proteins in human renal-cell carcinoma. Int. J. Cancer,
58: 825– 829, 1994.
6. Kondo, S., Toyokuni, S., Iwasa, Y., Tanaka, T., Onodera, H., Hiai,
H., and Imamura, M. Persistent oxidative stress in human colorectal
carcinoma, but not in adenoma. Free Radical Biol. Med., 27: 401– 410,
1999.
7. Oliva, M. R., Ripoll, F., Muniz, P., Iradi, A., Trullenque, R., Valls,
V., Drehmer, E., and Saez, G. T. Genetic alterations and oxidative
metabolism in sporadic colorectal tumors from a Spanish community.
Mol. Carcinog., 18: 232–243, 1997.
1400 Regulation of Oxidative Stress in Cancer
8. Kasai, H. Analysis of a form of oxidative DNA damage, 8-hydroxy-
2⬘-deoxyguanosine, as a marker of cellular oxidative stress during
carcinogenesis. Mutat. Res., 387: 147–163, 1997.
9. Shibutani, S., Takeshita, M., and Grollman, A. P. Insertion of spe-
cific bases during DNA synthesis past the oxidation-damaged base
8-oxodG. Nature (Lond.), 349: 431– 434, 1991.
10. Toyokuni, S., Okamoto, K., Yodoi, J., and Hiai, H. Hypothesis:
persistent oxidative stress in cancer. FEBS Lett., 358: 1–3, 1995.
11. Toyokuni, S. Reactive oxygen species-induced molecular damage
and its application in pathology. Pathol. Int., 49: 91–102, 1999.
12. Roldan-Arjona, T., Wei, Y. F., Carter, K. C., Klungland, A.,
Anselmino, C., Wang, R. P., Augustus, M., and Lindahl, T. Molecular
cloning and functional expression of a human cDNA encoding the
antimutator enzyme 8-hydroxyguanine-DNA glycosylase. Proc. Natl.
Acad. Sci. USA, 94: 8016 – 8020, 1997.
13. Radicella, J. P., Dherin, C., Desmaze, C., Fox, M. S., and Boiteux,
S. Cloning and characterization of hOGG1, a human homolog of the
OGG1 gene of Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA,
94: 8010 – 8015, 1997.
14. Rosenquist, T. A., Zharkov, D. O., and Grollman, A. P. Cloning and
characterization of a mammalian 8-oxoguanine DNA glycosylase. Proc.
Natl. Acad. Sci. USA, 94: 7429 –7434, 1997.
15. Arai, K., Morishita, K., Shinmura, K., Kohno, T., Kim, S. R.,
Nohmi, T., Taniwaki, M., Ohwada, S., and Yokota, J. Cloning of a
human homolog of the yeast OGG1 gene that is involved in the repair
of oxidative DNA damage. Oncogene, 14: 2857–2861, 1997.
16. Aburatani, H., Hippo, Y., Ishida, T., Takashima, R., Matsuba, C.,
Kodama, T., Takao, M., Yasui, A., Yamamoto, K., and Asano, M.
Cloning and characterization of mammalian 8-hydroxyguanine-specific
DNA glycosylase/apurinic, apyrimidinic lyase, a functional mutM ho-
mologue. Cancer Res., 57: 2151–2156, 1997.
17. Chevillard, S., Radicella, J. P., Levalois, C., Lebeau, J., Poupon,
M. F., Oudard, S., Dutrillaux, B., and Boiteux, S. Mutations in OGG1,
a gene involved in the repair of oxidative DNA damage, are found in
human lung and kidney tumours. Oncogene, 16: 3083–3086, 1998.
18. Kohno, T., Shinmura, K., Tosaka, M., Tani, M., Kim, S. R., Sugimura,
H., Nohmi, T., Kasai, H., and Yokota, J. Genetic polymorphisms and
alternative splicing of the hOGG1 gene, that is involved in the repair of
8-hydroxyguanine in damaged DNA. Oncogene, 16: 3219 –3225, 1998.
19. Jass, J. R., and Sobin, L. H. Histological Typing of Intestinal
Tumors (International Histological Classification of Tumors, WHO).
New York: Springer Verlag New York, Inc., 1989.
20. Union Internationale Contre le Cancer. TNM Classification of Ma-
lignant Tumors. New York: Wiley-Liss, Inc., 1997.
21. Helbock, H. J., Beckman, K. B., Shigenaga, M. K., Walter, P. B.,
Woodall, A. A., Yeo, H. C., and Ames, B. N. DNA oxidation matters:
the HPLC-electrochemical detection assay of 8-oxo-deoxyguanosine
and 8-oxo-guanine. Proc. Natl. Acad. Sci. USA, 95: 288 –293, 1998.
22. Toyokuni, S., Tanaka, T., Hattori, Y., Nishiyama, Y., Ochi, H.,
Hiai, H., Uchida, K., and Osawa, T. Quantitative immunohistochemical
determination of 8-hydroxy-2⬘-deoxyguanosine by a monoclonal anti-
body N45.1: its application to ferric nitrilotriacetate-induced renal car-
cinogenesis model. Lab. Invest., 76: 365–374, 1997.
23. Yamamoto, F., Kasai, H., Bessho, T., Chung, M. H., Inoue, H.,
Ohtsuka, E., Hori, T., and Nishimura, S. Ubiquitous presence in mam-
malian cells of enzymatic activity specifically cleaving 8-hydroxygua-
nine-containing DNA. Jpn. J. Cancer Res., 83: 351–357, 1992.
24. Shimura, K., Kohno, T., Kasai, H., Koda, K., Sugimura, H., and
Yokota, J. Infrequent mutations of the hOGG1 gene, that is involved in
the excision of 8-hydroxyguanine in damaged DNA, in human gastric
cancer. Jpn. J. Cancer Res., 89: 825– 828, 1998.
25. Vogelstein, B., Fearon, E. R., Hamilton, S. R., Kern, S. E., Preis-
inger, A. C., Leppert, M., Nakamura, Y., White, R., Smits, A. M., and
Bos, J. L. Genetic alterations during colorectal-tumor development.
N. Engl. J. Med., 319: 525–532, 1988.
26. Cifone, M. A., and Fidler, I. J. Increased metastatic potential is
associated with increasing genetic instability of clones isolated from
murine neoplasms. Proc. Natl. Acad. Sci. USA, 78: 6949 – 6952,
1981.
27. Loeb, L. A. Mutator phenotype may be required for multistage
carcinogenesis. Cancer Res., 51: 3075–3079, 1991.
28. Kinzler, K. W., and Vogelstein, B. Lessons from hereditary colo-
rectal cancer. Cell, 87: 159 –170, 1996.
29. Malins, D. C., and Haimanot, R. Major alterations in the nucleotide
structure of DNA in cancer of the female breast. Cancer Res., 51:
5430 –5432, 1991.
30. Olinski, R., Zastawny, T., Budzbon, J., Skokowski, J., Zegarski, W.,
and Dizdaroglu, M. DNA base modifications in chromatin of human
cancerous tissues. FEBS Lett., 309: 193–198, 1992.
31. Asami, S., Hirano, T., Yamaguchi, R., Tomioka, Y., Itoh, H., and
Kasai, H. Increase of a type of oxidative DNA damage, 8-hydroxygua-
nine, and its repair activity in human leukocytes by cigarette smoking.
Cancer Res., 56: 2546 –2549, 1996.
32. Bases, R., Franklin, W. A., Moy, T., and Mendez, F. Enhanced
excision repair activity in mammalian cells after ionizing radiation. Int.
J. Radiat. Biol., 62: 427– 441, 1992.
33. Yamaguchi, R., Hirano, T., Asami, S., Chung, M. H., Sugita, A.,
and Kasai, H. Increased 8-hydroxyguanine levels in DNA and its repair
activity in rat kidney after administration of a renal carcinogen, ferric
nitrilotriacetate. Carcinogenesis (Lond.), 17: 2419 –2422, 1996.
34. Kennedy, C. H., Cueto, R., Belinsky, S. A., Lechner, J. F., and
Pryor, W. A. Overexpression of hMTH1 mRNA: a molecular marker of
oxidative stress in lung cancer cells. FEBS Lett., 429: 17–20, 1998.
35. Hazra, T. K., Izumi, T., Maidt, L., Floyd, R. A., and Mitra, S. The
presence of two distinct 8-oxoguanine repair enzymes in human cells:
their potential complementary roles in preventing mutation. Nucleic
Acids Res., 26: 5116 –5122, 1998.
36. Monden, Y., Arai, T., Asano, M., Ohtsuka, E., Aburatani, H., and
Nishimura, S. Human MMH (OGG1) type 1a protein is major enzyme
for repair of 8-hydroxyguanine lesions in human cells. Biochem. Bio-
phys. Res. Commun., 258: 605– 610, 1999.
37. Takao, M., Aburatani, H., Kobayashi, K., and Yasui, A. Mitochon-
drial targeting of human DNA glycosylases for repair of oxidative DNA
damage. Nucleic Acids Res., 26: 2917–2922, 1998.