Int. J. Cancer: 52,847-850 (1992) 0 1992 Wiley-Liss, Inc.

Publicationof the International Union Against Cancer Publication de I'Union Internationale Contre le Cancer

SMOKING-RELATED DNA ADDUCTS IN HUMAN GASTRIC CANCERS G.W. DYKE',^,^, J.L. CRAVEN^, R. HALL^ and R.C. GARNER' T a m e r Research Unit, Dept. of Biology, University of York, York YO1 5DD; and ?York District Hospital, WiggintonRoad, York, UK. DNA was extracted from the tumour tissue of 26 patients (18 ing butanol extraction, to examine gastric mucosal DNA for smokers, 8 non-smokers) undergoing surgery for gastric cancer, the presence of adducts. and analyzed for the presence of DNA adducts by the 32P-postlabelling method. Adducts were detected in all samples tested, MATERIAL AND METHODS and adduct levels ranged from 2 adducts/ lo* nucleotides to 60 adduct4 lo8 nucleotides. In male subjects, adduct levels were Patients significantly greater in the DNA of smokers than in that of Samples of gastric mucosa were obtained from 26 patients non-smokers. These results support epidemiological data sugundergoing surgery for proven gastric cancer. Patient details gesting that smoking is a risk factor for gastric cancer.

o 1992 Wiley-Liss, Inc.

The smoking of tobacco products has been implicated as a major risk factor in the aetiology of many human cancers (Doll and Peto, 1981). There is strong evidence to link cancer of the lung, mouth and throat with smoking (Doll and Peto, 1981), and while few would doubt these associations there is conflicting data on the relationship between smoking and gastric cancer. Case-control studies have often failed to demonstrate any increased risk in the smoking population (Wynder et al., 1963; Armijo et al., 1981; Jedrychowski et al., 1986); but they can be criticized, since they are generally retrospective and are frequently derived from areas of high gastric-cancer incidence where other environmental factors may have more relevance. More reliable information may therefore be gleaned from larger, prospective cohort studies. Such studies do show a consistent significantly increased risk in male smokers (Hirayama, 1975; Wu-Williams et al., 1990; Nomura et al., 1990). This increased risk is not seen in female smokers (Hirayama, 1975) but, in the male, the risk is apparent for all sub-sites (WuWilliams et al., 1990) and appears to be greatest in those who started smoking at an early age (Hirayama, 1988). The mechanism by which tobacco smoke initiates cancer is still uncertain. Tobacco smoke, however, contains many components thought to be carcinogenic, and it is suspected that many chemical carcinogens are capable of being converted to potent electrophilic species (Miller and Miller, 1981). These may then react with and covalently bind to cellular macromolecules such as DNA (Hemminki, 1983). The persistence of chemically modified DNA bases has been shown to be associated with the development of cancers in animal models (Goth and Rajewski, 1974; Nicholl et al., 1975); hence, identification of carcinogen-DNA adducts in human tissues may enable us to more fully understand the carcinogenic process. The technique of 32P-post-labelling(Randerath et al., 1981), coupled with the enhancement techniques of butanol extraction (Gupta, 1985) or nuclease P1 digestion (Reddy and Randerath, 1986), provides an assay system that is sensitive enough to quantitate DNA adducts in native human tissue. This technique has proved particularly useful in identifying adducts thought to be related to the smoking of tobacco products. Such adducts have been identified in the DNA of lung tissue (Phillips et al., 1988), placental tissue (Everson et al., 1986), cervical tissue (Phillips, 1990) and a variety of other tissues (Cuzick et al., 1990). A linear correlation has been established between the number of cigarettes smoked in a lifetime and the number of adducts found in the DNA of lung tissue (Phillips et al., 1988). To date there are no reports of smoking-related adducts in gastric mucosa, yet this is clearly of interest owing to the uncertainty concerning the role of tobacco in gastric carcinogenesis. We have used the 32P-post-labellingtechnique, follow-

regarding age, sex and smoking habits are shown in Table I. Immediately after removal of the surgical specimen, approximately 1 g of tissue was taken from an area where normal mucosa had obviously been replaced by tumour. Tissue specimens were kept frozen at -70°C until analyzed.

Extraction of DNA Tissue samples were thawed, homogenized and the DNA isolated by a standard solvent extraction procedure (Gupta, 1984). The purity of the extracted DNA was analyzed spectrophotometrically and samples re-extracted until the h260/280 ratio was close to 1.8. The concentration of the resulting DNA solution was calculated spectrophotometrically using 1A260 = 50 Labelling of samples For the purposes of labelling, 5 pg of each DNA sample was evaporated to dryness in a Univap C (Uniscience, Cambridge, UK) centrifugal evaporator and digested overnight at 37°C with 5 pg each of micrococcal nuclease (Sigma, Poole, UK) and spleen phosphodiesterase (Boehringer, Lewes, UK) in a total volume of 10 pl containing 20 mM sodium succinate, pH 6.0 and 10 mM calcium chloride. Adducted nucleotides were concentrated by butanol extraction according to the method of Gupta (1989, re-evaporated to dryness and dissolved in 10 pI of distilled water. Each sample was labelled with 100 pCi of (y-3ZP]-ATPsynthesized by the method of Johnson and Walseth (1979) using 2 to 3 U of T4 polynucleotide kinase (NBL, Cramlington, UK) per sample. Excess ATP was removed with 40 mU potato apyrase (Sigma) and labelled, adducted nucleotides were separated by chromatography on PEI-cellulose plates (Camlab, Cambridge, UK) as described by Gupta et al. (1982). The solvents used for separation were as follows. D1: 1 M sodium phosphate, pH 6.0; D3: 3.5 M lithium formate, 8.5 M urea, pH 3.5; D4: 0.8 M lithium chloride, 8.5 M urea, 0.5 M Tris, pH 8.0. Adduct spots were visualized by autoradiography (4 days at -8O'C) and counted as previously reported (Cuzick et al., 1990). Adduct levels were calculated by the method of Gupta (1985). Each sample was analyzed twice, using a different batch of ATP each time and the results averaged. RESULTS

32P-post-labellinganalysis was performed on DNA extracted from 26 samples of gastric mucosa. All autoradiographs demonstrated adduct spots to a greater or lesser extent. There were clear qualitative differences, however, between the pattern of adduct spots seen in smokers and those of nonsmokers. Figure 1 shows representative examples of chromato3To whom correspondence and reprint requests should be addressed. Received: February 25,1992 and in revised form July 24,1992.

848

DYKE E T A L . TABLE I - AGE, SEX A N D SMOKING HISTORY OF PATIENTS

Patient

Aee

1

54

7

78

3

72 55 74 65 71

L

4 5 6 7 8 9

82 76

11 12

51 66 78

13 14

66 76

15 16

64 68

17

73

18

74 77

10

19

20 21 22

69

Sex ( M I F )

M F M M M M F M F M

M

M

M

F M

M

F F F M

58

M

24 25

59 78 69 63

26

75

M F M M M

23

Smoking histon,

S NS S S S NS NS S S NS S S S NS S S S NS S NS S NS NS S S S

grams of a non-smoker (Fig. l a ) and a smoker (Fig. Ib) following butanol extraction and 72P-post-labelling.The chromatogram of the smoker is similar to that seen in other tissues (Cuzick et a/., 1990), but that of the non-smoker is not typical and appears to be specific for gastric mucosa. There were no significant qualitative differences between the adduct patterns in males and in females. Quantitation of adduct levels revealed a range from as low as 2 adducts/108 nucleotides to more than 60 adducts/lOx. Adduct levels are expressed as the average of 2 separate estimations, each being performed with a different batch of ATP. Duplicate analyses were of the same order of magnitude. Figure 2 illustrates the ranges of adducts found in smokers and non-smokers. There were significantly higher adduct levels found in the DNA of gastric mucosa obtained from smokers (mean level = 15.5 adductsilV nucleotides) than in DNA from non-smokers (mean level = 5.5 adductsi l o q nucleotides, p < 0.05). Stratifying the groups according to sex shows that this relationship is found only in the male population. DISCUSSION

Until recently, studying the effects of carcinogens in humans has posed problems, due primarily to difficulty in detecting the low levels of adducts found in human DNA. The "P-postlabelling assay (Randerath et al., 1981), combined with the enhancement techniques of butanol extraction (Gupta, 1985) or nuclease P1 digestion (Reddy and Randerath, 1986), provides a highly sensitive method of detecting adducts in human tissue without any prior knowledge of their chemical structure. Using this assay, we have demonstrated the presence of a variety of adducts in the DNA of human gastric mucosa. This study has also demonstrated that the pattern of adducts differs in smokers and non-smokers and that, in male subjects, adduct levels are significantly higher in the gastric mucosa of smokers than that of non-smokers. The prcsence of smoking-related D N A adducts has been reported in a variety of human tissues, including placental tissue (Everson et al., 1986), lung (Phillips et a/., 1988), cervix (Phillips, 1990), bladder (Cuzicket al., 1990), pancreas (Cuzick

FIGURE1 -Typical autoradiographs of chromatograms from a non-smoker ( a ) and a smoker (b). Adducts were concentrated by butanol extraction, labelled and visualized by autoradiography at -80°C for 4 days.

et a/.. 1990) and liver (Cuzick et al., 1990). There is some evidence to suggest that adduct levels may be related to exposure to carcinogens as, in lung tissue, D N A adducts have been shown to have a linear relationship to the number of cigarettes smoked in a lifetime (Phillips et al., 1988). Our findings are of great interest in this respect, since the levels in gastric mucosa tend to be slightly higher than those found, by the use of this technique, in other tissues (Phillips et al., 1988; Cuzick et a/., 1990). This may be due to the use of butanol extraction as an enhancement procedure. The previously reported studies haw used nuclease P1 digestion to increase the sensitivity of the assay. A study comparing the relative recoveries of the 2 techniques in over 70 structurally diverse compounds, however, suggests that butanol extraction produces higher recover-

849

SMOKING-RELATED DNA ADDUCTS IN GASTRIC CANCERS TABLE I1 - RESULTS OF A PRELIMINARY STUDY COMPARING NUCLEASE P1 ENHANCEMENT WITH BUTANOL EXTRACTION I N GASTRIC MUCOSA

DNA Adducts/lOs Nucleotides

Sample

+

60 -

1

2

59 -

3

19 -

.1 29

4 5 6

+ 8

*+

A

a Non-smoker Smokers Females

4.9 1.0 7.5 2.0 1.5 0.7

12.5 5.5 16.5 2.9 4.3 6.5

Butanol extraction gave consistently higher adduct levels and hence the method was used subsequently in our study.

!

Non-smoker

DNA ddductri 10' nucleotide\ Butanol extraction

Nuclease PI

Smokers

Males

FIGURE 2 - DNA adducts in the gastric mucosa of smokers and non-smokers. In the male population, adduct levels (adducts/lOx nucleotides) are significantly higher in smokers ( p < 0.05). ies in the majority of cases (Gupta and Earley, 1988), and a small preliminary study performed upon gastric mucosal DNA samples showed consistently higher adduct levels after butanol extraction than after nuclease PI enhancement (Table 11). Thus. although it is technically more demanding and timeconsuming, there is reason to believe that butanol extraction may be a more reliable enhancement procedure, especially if the nature of the carcinogen-DNA adduct is unknown. Of course it is possible that there is a higher level of DNA adduction in the stomach than in other organs. One of the limitations of the 32P-post-labelling technique is that the chemical structure of the identified adduct is not always known. Thus, although the pattern of adduct spots seen in gastric niucosa is in many instances remarkably similar to that seen in other organs (Cuzick et a/., 1990), it is impossible to be certain that identical adducts are being detected, or that adducts are derived from the same source. It is likely that the stomach comes into contact with far more carcinogens than most other organs, and that these carcinogens are derived from a variety of sources. The normal gastric emptying time after a nical is 2 to 4 hours (Hill, 1986), and the stomach is therefore exposed for a considerable period of time to pre-formed dietary carcinogens. It is also the site at which dietary metabolites, some of which may be carcinogenic, are at their highest concentration (Hill, 1986); and, finally, it is possible that carcinogens may gain access to the stomach by way of duodenogastric reflux.

However, the relationship demonstrated in this study between gastric mucosal D N A levels and smoking habits is strong and suggests a causal link. It is likely, therefore, that the stomach is exposed to tobacco carcinogens. These may enter the stomach in solution in saliva, or be ingested by the swallowing of expectorated bronchial secretions (Meyer et al., 1980). The fact that they actually bind to gastric mucosal DNA is of particular interest, since it is apparent from scrutiny of previous studies that not all tissues demonstrate such adduction. Cuzick et al. (1990) failed to find adducts in breast tissue, for example, whereas lung or bladder tissue from the same patient contained significant adduct numbers. The highest adduct levels have consistently been found in tissues for which smoking has been implicated as a risk factor in carcinogenesis (Phillips et al., 1988; Cuzick et a/., 1990), and may reflect some inherent difference in the ability of these tissues to handle and metabolize carcinogens. Should this be the case, the finding of higher adduct levels in the tissues of male smokers assumes a greater significance, since it may indicate a sex-specific ability to handle tobacco-related carcinogens. This would be in full accord with the existing epidemiological evidence identifying smoking as a risk factor for the development of gastric cancer in male but not in female smokers (Hirayama. 1975; WuWilliams et al., 1990). In summary. this study provides biochemical evidence to link gastric cancer DNA adducts with the smoking of tobacco products and is compatible with the known epidemiological links between smoking and gastric cancer. Clearly, the number of patients studied here is small, and the DNA was extracted from malignant tissue. It would be desirable to confirm these findings with larger numbers of subjects, especially female, subjects, and with non-neoplastic as well as tumour tissue. Although the incidence of gastric cancer is falling spontaneously (Howson eta/., 1986), it is still important to idcntify and possibly to eliminate proven risk factors. This study provides further evidence that smoking should be considered such a factor.

REFERENCES AKMIJO. R., ORELLANA. M.. MEDINA. E.. COULSON, A.H., SAYRt, J.W. and DLTELS.R., Epidemiology of gastric cancer in Chile: 1-Casecontrol study. Itif. J. Epiderniol., 10, 53-56 (1981). , JENKINS, D. and GARNER, R.C., DNA Cuzrck;.J.. R O L I T L ~ D G EM.N., adducts in different tissues of smokers and non-smokers./ti!. J. Caricer, 45,673-678 (1990). Do1.1.. R. and PETO,R., The cause.r of cancer, Oxford University Press. Oxford (1981 ). EVLRSON, R.B., RANDERATH, E., SANTELLA, R.M.. CEFALO, R.C., AVITTS.T.A. and RANDERATH, K., Detection of smoking-related covalent DNA adducts in human placenta. Science, 231,54-57 (1986). GOTH,R. and RAJEWSKI, M.J., Persistence of Oh-ethylguaninein rat brain DNA: correlation with nervous system-specific carcinogenesis by ethylnitrosourea. Proc. riut. Acad. Sci. (Wash.),71, 639-643 (1974). GLIPTA. R.C., Non-random binding of the carcinogen N-hydroxy-2-

acetylaminofluorene to repetitive sequences of rat liver DNA irz vivo. Proc. tiat. Acad. Sci. (Wash.),81, 6943-6947 (1984). GUPTA,R.C.. Enhanced sensitivity of "P-post-labelling analysis of aromatic carcinogen:DNA adducts. CaricerRes., 45,5656-5662 (1985). GUPTA, R.C. and EARLEY, K., j?P adduct assay: comparative recoveries of structurally diverse DNA adducts in the various enhancement procedures. Curcinogenesis, 9, 1687-1693 (1988). GLIPTA,R.C., REDDY, M.V. and RANDERATH, K., "P-post-labelling analysis of non-radioactive aromatic carcinogen-DNA adducts. Carcinogenesis, 3,1081-1092 (1982). HEMMINKI, K., Nucleic acid adducts of chemical carcinogens and mutagens. Arch. Toxicol., 52,249-285 (1983). HILL,M.J., Aetiology and micro-environment of gastric cancer. I n : M.I. Filipe and J.R. Jass (eds.). Gasfric carcinorntr, Churchill Livingstone (1986).

850

DYKE E T A L .

HIRAYAMA, T., Epidemiology of cancer of the stomach With special reference to its recent decrease in Japan. Cancer Res., 35, 3460-3463 ( 1975). HIRIYAMA, T., Actions suggested by gastric cancer epidemiological studies in Japan. In: P.I. Reed and M.J. Hill (eds.), Gastric Carcinogenesis, Excerpta Medica (1988). HOWSON,C.P., HIYAMA, T. and WYNDER, E.L., The decline in gastric cancer: epidemiology of an unplanned triumph. Epiderniol. Rev., 8, 1-21 (1986). JEDRYCHOWSKI, W., WAHRENDORF. J., POPIELA, T. and RACHTAN, J., A case-control study of dietary factors and stomach-cancer risk in Poland. hi.J. Cancer, 37, 837-842 (1986). JOHNSON, R.A. and WALSETH,T.F., The enzymatic preparation of [cI-~*P]-ATP. [w3'P]-GTP, 13*P]cAMPand [32P]cGMPand their use in assay of adenylate and guanylate cyclases and cyclic nucleotide phosphodiesterases. Advanc. Cvclic Nucleotide Res., 10,135-17 (1979). MEYER,M.B., LUK,G.D., SOTELO.J.M.. COHEN,B.H. and MENKES, H.A.. Hypothesis: the role of the lung in stomach cancer. Amer. Rev. resp. Dis., 121,887-892 (1980). MILLER,E.C. and MILLER,J.A., Searches for ultimate chemical carcinogens and their reactions with cellulat macromolecules. Cancer, 47,2327-2345 (1981). NICHOLL,J.W., SWANN, P.F. and PEGG,A.E., Effect of dimethylnitrosamine on persistence of methylated guanines in rat liver and kidney DNA. Nature (Lotid.), 254,261-263 (1975).

NORUMA, A., GROVE,J.S., STEMMERMANN. G.N. and SEVERSON, R.K., A prospective study of stomach cancer and its relation to diet. cigarettes and alcohol consumption. Cancer Rex, 50,627-631 (1990). PHILLIPS, D.H., HEWER,A., MALWLM, A.D.B., WARD,P. and COLEMAN, D.V., Smoking and DNA damage in cervical cells. Lancet. i, 417 (1990). PHILLIPS, D.H., HEWER,A., MARTIN, C.N., GARNER, R.C. and KING, M.M., Correlation of D N A adduct levels in human lung with cigarette smoking. Nature (Lond.),336,790-792 (1988). RANDERATH, K., REDDY,M.V. and GUPTA,R.C., 32P-labellingtest for DNA damage. Proc. nat. Acad. Sci. (Wash.), 78,6126-6129 (1981). RANDERATH, E., MILLER,R.H., MITTAL, D., AVITTS,T.A., DUNSFORD, H.A. and RANDERATH, K., Covalent D N A damage in tissues of cigarette smokers as determined by 3*P-post-labelling assay. J. not. Cancerlnst., 81,341-347 (1989). REDDY.M.V. and RANDERATH, K., Nuclease P1-mediated enhancement of sensitivity of 32P-post-labelling test for structurally diverse DNA adducts. Carcinogenesis, 7,1543-1551 (1986). WU-WILLIAMS, A.H., Yu, M.C. and MACK,T.M., Life-style, workplace and stomach cancer by sub-site in young men of Los Angeles county. Cancer Res., 50,2569-576 (1990). WYNDER, E.L., KMET,J., DUNGAL, N. and SEGI,M., An epidemiological investigation of stomach cancer. Carrcer, 16, 1461-1496 (1963).