DNA氧化損傷和結直腸癌發(fā)病風險的相關研究

周少玉，王澤民，James E. Klaunig

(印第安那大學 環(huán)境衛(wèi)生系，印第安那 布盧明頓 47408，美國)

There is increasing body of studies supporting the role of oxidative stress and damage in carcinogenesis. Oxidative stress results from overproduction of reactive oxygen species and impaired detoxification in living cells.Increased production of reactive oxygen species causes damage to macromolecules such as DNA, protein and lipids, thus modulates the functions of the molecules involved[1-2]. Reactive oxygen species also play important roles in regulating intracellular signaling serving as a second messenger.

Colorectal cancer is one of the most common cancers worldwide, and is the third most common cancer diagnosed in both men and women in the United States. According to the estimation of American Cancer Society, there will be 102480 new cases of colon cancer,and 40340 new cases of rectal cancer in the United States for 2013. While some (about 10%-15%) of CRC cases are attributed to inherited gene defects, majority (about 70%-80%) of the cases of CRC occur sporadically[3-4].

Association between oxidative stress and CRC has been extensively studied over last decade. A number of studies have demonstrated increased level of reactive oxygen species nitric oxide, DNA damage such as 8-oxodG, and lipid peroxidation in both colorectal tumors[5-7], indicating an increased oxidative stress status of CRC tissues. Significant increase in oxidative DNA damage in the form of 8-oxodG DNA has also been reported in leukocytes, serum, and urine of CRC patients[6, 8-10]. In addition, reactive oxygen metabolites and antioxidant capacity were estimated in serum and found to be associated with increased risk for CRC[11]. However, there is little report on the direct measurement of oxidative DNA while implicating the molecular mechanism of oxidative stress into the carcinogenesis of colorectal cancers.

The DNA damage can be prevented by intracellular antioxidant system or repaired by DNA base repair enzymes[1]. Environmental factors further modify the oxidative status. These factors include lifestyle, genetic makeup of an individual and vitamin supplements etc. It has been reported that smoking and alcohol are risk factors for CRC[12]and single nucleotide polymorphism influence the risk of CRC[13]. However, conclusion has never been made on the etiological role of oxidative stress because of complexity of environmental factors and inconsistent research.

The present study investigated the correlation between total anti-oxidant capacity and DNA damage and the associated risk of developing colorectal cancer. DNA damage was directly measured by Comet assay, and oxidative DNA damage was determined by formamidopyrimidine-DNA glycosylase(fpg)-modified comet assay using blood samples collected from healthy subjects, patients diagnosed with polyps and colorectal cancer. SNPs were examined on elected genes to correlate with the CRC risk.

1 Materials and Methods

1.1 Chemicals Deferoxaminemesylate,dimethylsulfoxide (DMSO),disodium EDTA,ethidium bromide,sodium chloride (NaCl),sodium hydroxide (NaOH),sodium lauryl sacrosinate,trizmabase,Triton X-100,HEPES potassium salt,potassium chloride (KCl),RPMI1640media,isopropanol and all other chemicals of analytical grade were purchased from Sigma Chemical Co.(St. Louis, MO, USA). Formamidopyrimidine-DNA glycosylase (Fpg) and Comet slides were purchased from Trevigen (Gaithersburg, MD, USA).

1.2 Study population and data collection The present project was initiated in the year 2005 by Indiana University (IU) and funded by the Department of Defense.The objective ofthis project is to developsensitive biomarkers/molecular signatures and predictive models for early diagnosis, prevention, and to provide better care and treatment of colorectal cancer patients. Participants were enrolled in the Multidisciplinary Gastrointestinal Oncology Clinics at the Indiana University Simon Cancer Center (IUSCC) and the IU Medical Centers (IU Hospital, IU Cancer Center, Wishard Hospital and IU Health Arnett Hospital)for colonoscopies from April 2009 to November 2010.The study included 175 participants (80 healthy controls, 67 polyps and 28 colorectal cancer cases). Tissue, blood and plasma samples were collected from all patients during this period. All participants were informed with prior written consent and given a life style risk factor questionnaire. The patients were diagnosed as healthy control, polyps and colorectal cancer based on the colonoscopy results followed by histopathological confirmation by pathologists. Clinical data including demographics, life style, diet, smoking, alcohol consumption, height, weight, family history of colorectal cancer, medications, prior cancer therapy and occupational exposure, were obtained from questionnaires, patient records, pathology reports, and physician notes.

1.3 Sample collection Whole blood samples were collected from patients, who were fasted for 8 hours,prior to colonoscopy. An aliquot of 10 μl whole blood was mixed with 0.5 ml of comet solution (RPMI 1640containing 10% FBS, 10%DMSO, and 1 mMdeferoxamine),was step-frozen and stored at -80℃ for later analysis of oxidative DNA damage.Aliquots of plasma samples were prepared from blood samples by centrifuging at 1500g for 20 min, and stored at -80℃. Remaining blood samples were also stored at -80℃ for DNA isolation and SNP analysis.

1.4 Measurement of total antioxidant capacity The Trolox Equivalent Antioxidant Capacity (TEAC) is an assay to measure the overall antioxidants status in biological samples.TEAC was measured in plasma based on the ability of antioxidants in the plasma sample to scavenge the preformed radical monocation of 2,2'-azinobis-(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS·+), compared to water soluble analogue of vitaminE[14].Plasma samples (25 μl) or Trolox standards were incubated to react with ABTS·+(100 μl) at room temperature for 5 minutes, and the absorbance measured at 734 nm using TECAN Infinite M200 spectrophotometer (Tecan US, Durham, NC).TEAC was quantified from a Trolox standard curve, and expressed as Trolox equivalents.

1.5 Measurement of direct and oxidative DNA damage by Comet assay Comet assay or single cell electrophoresis measures the DNA double strand breaks as well as single strand breaks. In this study, we used the modified method of comet assay as reported previously from our lab[15-16]. Briefly, the frozen blood samples in comet solution were thawed and 7μl of blood was mixed with 70 μl of 1% low-melting agarose.This suspension was applied and spread on CometSlides (Trevigen, Gaithersburg, MD) and incubated at 4 ℃ for solidification. Cells were lysed for 1 hour in lysis solution (100mM Na2EDTA, pH 10, 2.5M NaCl, 10mM Trizma base, 1% sodium lauryl sarcosinate, 1% Triton-X 100, 10% DMSO, and 1mM deferoxamine), incubated in alkali buffer (0.3M NaOH and 1mM Na2EDTA, pH>13) for 40 min and electrophoresed in same alkali buffer (25 V, 300 mA) for 30 min at 4℃.Slides were neutralized by incubating in neutralization buffer (0.4M Tris, pH 7.5) at 4 ℃ for 5 min, rinsed with deionized water and air dried for overnight in dark room. Slides were stained with ethidium bromide (25μl, 20μg/ml) and 100 randomly selected comets per sample were evaluated using Komet 5.5 imaging Software (Kinetic Imaging Ltd., Liverpool, UK) under fluorescence microscope (Nikon, Japan).

Oxidative DNA damage was determined by modified alkaline comet assay as reported previously[16]. Briefly, lysed cells were washed in Fpg solution (40mM 4-[2-hydroxyethyl]piperazine-1-ethanesulfonic acid, 100mM KCl, 0.5mM Na2EDTA, and 0.2% bovine serum albumin; pH 8.0) and incubated either with Fpg enzyme or Fpg solution alone for 40 min at 37℃. Rest of the procedures was same as described above.Direct and oxidative DNA damage was measure dusing the formula,[tail mean-head mean] × tail%DNA/100 and expressed as comet (Olive) tail moment.

1.6 Single nucleotide polymorphism analysis

1.6.1 DNA Isolation Genomic DNA was isolated from 3 ml of whole blood samples using the QIAamp DNA Blood Midi Kit (Qiagen Inc., Valncia, CA) according to manufacturer’s instructions. Isolated DNA was quantitated by UV spectrophotometry and stored at -80℃ until genotyping.

1.6.2 Genotyping We selected SNPs in genes which previously showed a significant association with oxidative stress (rs1001179, rs1001179, rs4880, rs1799983,rs1065411, rs1695), DNA damage and repair (rs2307 486,rs1052133, rs25487, rs1042522),inflammation (rs1800796, rs4073, rs689466, rs5275),vitamin D status (rs1544410,rs1801725, rs2181874) and metabolism (rs2069522).The details of these SNP assays are listed in Table 1. Allelic discrimination assays of all SNPs variants were performed using primers supplied by Applied Biosystems (Foster City, CA). 100 ng of DNA samples were used in standard 96 well reaction plates and the probe fluorescence signals were detected by 7900HT Fast Real Time PCR System (Applied Biosystems, Foster City, CA), following the manufacturer’s instructions.

Table1ListofgenesthatwereselectedforanalysisofSNPs

SymbolGeneSNPOxidative stresCATCatalasers1001179SOD2Super oxide dismutasers4880NOS3Nitric oxide synthase 3rs1799983GSTM1Glutathione S-transferase M1 rs1065411GSTP1Glutathione S-transferase P1rs1695DNA damage & repairAPEX1APEX nuclease (multifunctional DNA repair enzyme) 1rs2307486OGG18-oxoguanine DNA glycosylasers1052133XRCC1X-ray repair complementing defective repair in Chinese hamster cells 1rs25487TP53Tumor protein P53rs1042522InflammationIL-6Interleukin-6rs1800796IL-8Interleukin-8rs4073PTGS2Prostaglandin-endoperoxide synthase 2 (Prostaglandin G/H synthase and cyclooxygenase)rs689466PTGS2Prostaglandin-endoperoxide synthase 2 (Prostaglandin G/H synthase and cyclooxygenase) Ex10rs5275Vitamin D statusVDRVitamin D receptorrs1544410CASR Calcium Sensing Receptorrs1801725CYP24A1cytochrome P450 (CYP) 24A1rs2181874MetabolismCYP1A2cytochrome P450 (CYP) 1A2rs2069522

1.7 Statistical analysis of data Values for TEAC, alkaline DNA damage and oxidative DNA were presented as mean ± SD. Differences among multiple groups were evaluated using a one-way ANOVA with Dunnett’s post hoc test.Data of SNPs was presented as frequency and percentage for each allele, and differences among health, polyp, and colorectal cancer were analyzed by multi-table Chi Square. The level of significance was selected to beP<0.05.

2 Results

2.1 Demographic data We analyzed 175 blood samples for this study.The subjects included health individuals with average age 52.4 years (20-70 years), and patients diagnosed with polyps with average age 62.9 years (39-84) and CRC 53.7 years (27-86) at the time of diagnosis.

2.2 Direct and oxidative DNA damage and antioxidant capacity There have been a number of studies on DNA damage in colorectal cancers. However, most of those studies focused on the analysis of 8-oxodG, indirectly indicating DNA damage. In this study, we utilized Comet assay to directly measure DNA damage. Direct DNA damage in leukocyte cells collected from health subjects, and patients diagnosed with polyp and colorectal cancer was measured by alkaline comet assay. Values were presented as the means ± SD in all the samples. As shown in Figure 1, there was a marginally significantly higher direct DNA strand breakage in colorectal cancer group compared to health population (P=0.071).

*P=0.071 in comparison with control.Fig 1 DNA strand breakage among health, polyp and colorectal cancers

We then analyzed samples with formamidopyrimidine DNA glycosylase(fpg) modified Comet assay, which indicates oxidative DNA damage. As shown in Figure 2, oxidative stress induced DNA damage was significantly higher in colon and colorectal cancers when compared to healthy controls.However, there was no significant difference observed in polyps compared to health group. Because of the small size, rectal cancers were not analyzed as a separate group. Furthermore, the total antioxidant capacity (TEAC) was analyzed. TEAC was measured in serum using a method based on the scavenging of the 2,2'-azinobis-(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) radical ABTS·+. Data was presented as means ± SD in all the samples. Statistical analysis was done by a one-way ANOVA with Dunnett’s post hoc test. As shown in Figure 3, there was a positive trend on polyps but not on colorectal cancer (P< 0.05).

*Statistical significance as P<0.05 in comparison with Control group. Fig 2 Oxidative DNA damage among health,polyp and colorectal cancers

2.3 Association of smoking, alcohol and CRC with DNA damage We analyzed smoking and alcohol status and correlated with DNA damage. There was an overall higher direct and oxidative DNA damage in smoking colorectal group compared to smoking control and polyps groups (Table 2). Further gender analysis showed that female smoking group with colorectal cancer had higher level of directly DNA damage compared to other smoking group. But, this is not the case for male groups. Overall, no significant difference in direct DNA damage or oxidative DNA damage was found in alcoholic colorectal patients compared to healthy control (data not shown).

*Statistical significance as P<0.05 in comparison with Control group.Fig 3 Total antioxidant capacity (TEAC assay)

Table2AssociationofDNAdamageandsmokingstatusamongdifferentgroups

GroupsNumber(n)TEACDNA breakageOxidative DNA damageSmoking Males Healtyy Control155.83±0.150.95±0.081.55±0.39Polyps235.85±0.101.07±0.101.48±0.17CPC75.64±0.201.09±0.202.15±0.57Smoking FemalesHealthy Control215.66±0.150.90±0.081.62±0.25Polyps145.46±0.131.11±0.151.48±0.52CPC65.72±0.161.49±0.32*1.78±0.41AllControl365.65±0.091.11±0.071.32±0.16Smoking Polyps375.79±0.091.07±0.091.48±0.16Smoking CRC135.69±0.121.32±0.20*1.94±0.33*

*P<0.05 compared to control group.

2.4 Oxidative DNA repair and antioxidant enzyme SNPs and CRC We analyzed the SNPs variants in genes which previously showed a significant association with oxidative stress, DNA damage and repair, inflammation, vitamin D status and metabolism as listed in Table 1. Allelic discrimination assays of all SNPs were performed using primers from Applied Biosystems (Foster City, CA), and the probe fluorescence signals were detected by 7900HT fast Real-time PCR System (Applied Biosystems, Foster City, CA). Figure 4 shows allelic discrimination plot genotyping oxoguaninegly cosylase 1 (Ogg1) with probe rs1052133 for the DNA samples.

Fig 4 Allelic discrimination plot showing the genotyping of Ogg1 for all the samples

We observed that the mutant phenotype CC in DNA damage repair enzyme OGG1 (Ser326Cys), 8-Oxoguanine DNA glycosylase,was significantly higher in colorectal cancers (78.9%) than controls and polyps, which were 56.3% and 52.7%, respectively (Table 3, and Fig 4), while the mutant phenotype KK for GSTM1, an antioxidant enzyme, was significantly higher in colorectal cancers(22.9%) than healthy people and people with polyps (Table 3). No statistical difference was found in other genes analyzed in the present study.

Table3FrequencyofSNPsinOgg1andGSTM1amongdifferentgroups

PhenotypeOGG1(S326C)SSSCC%oftotalGSTM1 (N173K)NNNKKK%oftotalControl6.836.956.355.630.913.6Polyp5.541.852.742.030.415.0Colorectal5.315.878.9*37.114.022.9*

*P< 0.05 compared to control group.

3 Discussion

Colorectal cancer is one of the most frequent human cancers, and is the third most common cancer diagnosed in both men and women in the United States[17]. Increasing evidence supports a critical role of oxidative stress and reactive oxygen species generation in the pathological progress of all the three stages, initiation, promotion and progression, of human cancers including colorectal cancer. Investigation on oxidative stress and antioxidant status has been a focus of many researchers for a long time in order to predict and prevent human colorectal cancers. However, dissecting the exact role of oxidative stress in the development of colorectal cancer is complicated and remains a challenge.No any specific biochemical marker has been identified yet up to date.

Increased reactive oxygen species generation and DNA damage has been reported in both tumor tissue and leukocytes of colorectal cancer patients[5-7]. However, limited studies have been conducted on direct measurement of DNA damage in relation to colorectal cancers.In the present study,we employed Comet assay, and directly measured DNA strand breakage of leukocytes from normal individuals, patients diagnosed with polyps and colorectal cancers. We found significantly higher DNA damage in the leukocytes of colorectal cancer patients compared to normal control, suggesting a status of oxidative stress. Furthermore, we employed formamidopyrimidine DNA glycosylase modified (oxidative DNA damage) Comet assay to evaluate the oxidative DNA damage, and found a similar pattern as the direct DNA damage, i.e., higher level of oxidative DNA damage was found in colorectal cancer patients than the normal control group. All these demonstrated that an increased oxidative stress is associated with colorectal cancer.

Although oxidative stress is proposed to be involved in the mechanism of carcinogenesis, it is still being debated whether oxidative stress plays a causal role in pathological progression.Our results in the present study with human population clearly demonstrated a positive association between oxidative stress and development of colorectal cancer. Since tumor cells possess high capacity of generating reactive oxygen species, we cannot determine whether increased DNA damage in the colorectal cancer subject is a cause or consequence of colorectal cancers. Indeed, elevated level of oxidative lesions has been well documented in tumor tissues in animal studies and population investigation, yet the effectiveness of intervention of cancer using antioxidants has been limited in a number of clinical trials[18-20]. From the perspective of causative etiology, we would expect increased oxidative stress in early stage of carcinogenesis.Results from the subjects of polyp sin the present study showed no increase in the level of both DNA breakage damage or oxidative damage compared to normal subjects, suggesting that the increased oxidative DNA damage might be a consequence of late carcinogenesis rather than a causative factor.

Oxidative DNA damage is the consequence of imbalance between oxidants and defense mechanisms. There are a number of defense mechanisms in the living organism to remove harmful reactive oxygen species. Oxidative stress occurs only when the critical balance between oxidants and antioxidants is disrupted due to the depletion of antioxidants or excessive accumulation of the reactive oxygen species. Interestingly, we found higher antioxidant capacity (TEAC) in polyps but not in colorectal cancers when compared to normal control (Fig 3). We postulate this as an oxidative stress-induced defense mechanism. Oxidative stress condition induces increased antioxidant capacity in people with polyps, therefore, confers a protection against DNA damage under stress. As such, no increased DNA damage was found in the polyps subjects. However, as continuous and persistent oxidative stress occurs, oxidative DNA damage may arise as seen in the late stage of colorectal cancer subjects found in the present study. It is worthy to mention that we measured both DNA damage and antioxidant capacity simultaneously, which may provide a better profile explaining the cause of accumulation of both direct and oxidative DNA damage found in the colorectal cancer subjects. Nonetheless, our finding could not conclusively differentiate between increased DNA damage as a cause or the consequence of the carcinogenesis of colorectal cancers.

Many environmental and lifestyle factors, such as alcohol and smoking，affect oxidative status. Cigarette smoking is an important environmental factor contributing to the development of a variety of human cancer, and it has been well established that tobacco smoke contains thousands of chemicals and causes inflammation[2, 21]. It is also known that reactive oxygen species are generated during the combustion of tobacco products, thus cause oxidative DNA damage resulting in genemu tations or modifying signaling transduction pathways that lead to cancer[22-25]. We have found both direct DNA damage and oxidative DNA damage were significantly higher in the smoking colorectal patients than smoking control subjects. Further stratified analysis of gender found that increased DNA damage was associated with female smoking colorectal cancer patients, but not male smoking subject. We are however unable to explain the gender difference in light of the DNA damage since the increased DNA damage level is not the consequence of single mechanism, but may be influenced by a wide range of environmental factors.

It has been well established that inherited factors such as familial adenomatous polyposis (FAP) and hereditary nonpolyposis colorectal carcinoma (HNPCC) contribute to the development of colorectal cancer[26-29]. More recently, single nucleotide polymorphisms are realized to interact with environmental factors predisposing to a number of human cancers[2]. We determined SNPs in genes that have significant association with oxidative stress such asCAT,SOD2,NOS3,GSTM1, andGSTP1. We also determined SNPs in genes in DNA damage and repair (APEX1,OGG1,XECC1, andTP53).All these genes are highly related to oxidative DNA damage. In addition, we determined SNPs in genes of inflammation, metabolism, and genes that influence vitamin D status. We indeed found that colorectal cancer patients had a higher percentage carrying mutant genotype in antioxidant geneGSTM1compared to normal subjects. Also a significantly higher percentage of mutant genotype in DNA repair geneOGG1was observed in the colorectal cancer patients than normal population. These positive associations strongly support a role of DNA damage and repair in the development of colorectal cancers.

In conclusion, we reported here both increased direct and oxidative DNA damage are associated with colorectal cancer. Environmental factor such as smoking, and genetic background such as DNA repair and antioxidant enzymes are strongly associated with high risk of colorectal cancers. This study suggests that high level of oxidative stress induced DNA damage is associated with increased risk of colorectal cancer, and may be a potential good biomarker.However, the relatively small cohort of subjects limits our further analysis of interactions between environmental factors and genetic background in the role of DNA damage and modification on colorectal cancer development. Nonetheless, the measurement of both direct and oxidative DNA damage combined with the simultaneous determination of antioxidant capacity provides a new clue into the role of oxidative stress and DNA damage in the development of colorectal cancer.

[References]

[1] Klaunig J E, Kamendulis L M, Hocevar B A. Oxidative stress and oxidative damage in carcinogenesis[J]. Toxicol Pathol, 2010,38(1):96-109.

[2] Klaunig J E, Wang Z,Pu X,et al. Oxidative stress and oxidative damage in chemical carcinogenesis[J]. Toxicol Appl Pharmacol,2011,254(2): 86-99.

[3] Souglakos J. Genetic alterations in sporadic and hereditary colorectalcancer:implementations for screening and follow-up[J]. Dig Dis,2007,25(1):9-19.

[4] Per?e M. Oxidative stress in the pathogenesis of colorectal cancer: cause or consequence?[J].Biomed Res Int, 2013,2013: 725710.

[5] Haklar G, Sayin-Ozveri E, Yüksel M,et al. Different kinds of reactive oxygen and nitrogen species were detected in colon and breast tumors[J]. Cancer Lett, 2001, 165(2):219-224.

[6] Guz J, Foksinski M, Siomek A, et al. The relationship between 8-oxo-7,8-dihydro-2'-deoxyguanosine level and extent of cytosine methylation in leukocytes DNA of healthy subjects and in patients with colon adenomas and carcinomas[J]. Mutat Res, 2008,640(1-2): 170-173.

[7] Rainis T, Maor I, Lanir A, et al. Enhanced oxidative stress and leucocyte activation in neoplastic tissues of the colon[J]. Dig Dis Sci, 2007,52(2): 526-530.

[8] Chang D, Wang F, Zhao Y S, et al. Evaluation of oxidative stress in colorectal cancer patients[J]. Biomed Environ Sci, 2008, 21(4): 286-289.

[9] Sato T, Takeda H, Otake S, et al. Increased plasma levels of 8-hydroxydeoxyguanosine are associated with development of colorectal tumors[J]. J Clin Biochem Nutr, 2010, 47(1): 59-63.

[10] Dincer Y, Himmetoglu S, Akcay T, et al. Prognostic significances of oxidative DNA damage evaluated by 8-hydroxy-deoxyguanosine and antioxidant enzymes in patients undergoing resection of gastric and colon carcinoma[J]. Neoplasma, 2007, 54(2):131-136.

[11] Leufkens A M, van Duijnhoven F J, Woudt S H, et al. Biomarkers of oxidative stress and risk of developing colorectal cancer: a cohort-nested case-control study in the European Prospective Investigation Into Cancer and Nutrition[J]. Am J Epidemiol, 2012,175(7): 653-663.

[12] Tsong W H, Koh W P, Yuan J M, et al.Cigarettes and alcohol in relation to colorectal cancer: the Singapore Chinese Health Study[J]. Br J Cancer, 2007, 96(5): 821-827.

[13] Stern M C, Conti D V, Siegmund K D, et al. DNA repair single-nucleotide polymorphisms in colorectal cancer and their role as modifiers of the effect of cigarette smoking and alcohol in the Singapore Chinese Health Study[J]. Cancer Epidemiol Biomarkers Prev, 2007, 16(11): 2363-2372.

[14] Miller N J, Rice-Evans C, Davies M J,et al. A new method for measuring antioxidant activity[J]. Biochem Soc Trans, 1993,21(2): 95.

[15] Harvilchuck J A, Pu X, Klaunig J E, et al.Indicators of oxidative stress and apoptosis in mouse whole lung and Clara cells following exposure to styrene and its metabolites[J]. Toxicology, 2009, 264(3): 171-178.

[16] Pu X, Kamendulis L M, Klaunig J E.Acrylonitrile-induced oxidative stress and oxidative DNA damage in male Sprague-Dawley rats[J]. Toxicol Sci, 2009, 111(1): 64-71.

[17] Leufkens A M, Van Duijnhoven F J, Boshuizen H C, et al.Educational level and risk of colorectal cancer in EPIC with specific reference to tumor location[J]. Int J Cancer, 2012, 130(3):622-630.

[18] Bairati I, Meyer F, Gélinas M, et al.A randomized trial of antioxidant vitamins to prevent second primary cancers in head and neck cancer patients[J]. J Natl Cancer Inst, 2005, 97(7): 481-488.

[19] Hennekens C H, Buring J E, Manson J E, et al. Lack of effect of long-term supplementation with beta carotene on the incidence of malignant neoplasms and cardiovascular disease[J]. N Engl J Med, 1996, 334(18): 1145-1149.

[20] Poulsen H E. Oxidative DNA modifications[J]. Exp Toxicol Pathol, 2005,57(1):161-169.

[21] Hecht S S. Cigarette smoking and lung cancer: chemical mechanisms and approaches to prevention[J]. Lancet Oncol, 2002, 3(8):461-469.

[22] Cederblad L, Thunberg U, Engstr?m M, et al. The combined effects of single-nucleotide polymorphisms, tobacco products, and ethanol on normal resting blood mononuclear cells[J]. Nicotine Tob Res, 2013,15(5):890-895.

[23] Cheng G, Shi Y, Sturla S J, et al. Reactions of formaldehyde plus acetaldehyde with deoxyguanosine and DNA: formation of cyclic deoxyguanosine adducts and formaldehyde cross-links[J]. Chem Res Toxicol, 2003, 16(2): 145-152.

[24] Wang M, Cheng G, Sturla J, et al.Identification of adducts formed by pyridyloxobutylation of deoxyguanosine and DNA by 4-(acetoxymethylnitrosamino)-1-(3-pyridyl)-1-butanone, a chemically activated form of tobacco specific carcinogens[J]. Chem Res Toxicol, 2003, 16(5): 616-626.

[25] Boysen G, Kenney P M, Upadhyaya P, et al. Effects of benzyl isothiocyanate and 2-phenethyl isothiocyanate on benzo[a]pyrene and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone metabolism in F-344 rats[J]. Carcinogenesis, 2003,24(3): 517-525.

[26] Pineda M, González S, Lázaro C, et al. Detection of genetic alterations in hereditary colorectal cancer screening[J]. Mutat Res, 2010, 693(1-2): 19-31.

[27] Gryfe R.Inherited colorectal cancer syndromes[J]. Clin Colon Rectal Surg, 2009,22(4):198-208.

[28] Gryfe R.Overview of colorectal cancer genetics[J]. Surg Oncol Clin N Am, 2009, 18(4):573-583.

[29] Burn J,Mathers J, Bishop D T.Genetics, inheritance and strategies for prevention in populations at high risk of colorectal cancer (CRC) [J]. Recent Results Cancer Res,2013,191:157-183.