Relationship between 7,12-dimethyl- and 7,8,12-trimethylbenz[a]anthracene DNA adduct formation in hematopoietic organs and leukemogenic effects

41

Cancer Letters, 31(1987141-49 Elsevier Scientific Publishers Ireland Ltd.

RELATIONSHIP BETWEEN 7,12-DIMETHYLAND 7,8,12TRIMETHYLBENZblANTHRACENE DNA ADDUCT FORMATION IN HEMATOPOIETIC ORGANS AND LEUKEMOGENIC EFFECTS

MIRIAM FALZON+, VANESSA T. VU, PETER P. ROLLER and SNORRI S. THORGEIRSSON Laboratory ItX3.A.I

of Experimental

Carcinogenesir,

National

Cancer Institute.

Bethesdu,

MD 20892

(Received 9 February 19871 (Revised version received 1 May 19871 (Accepted 3 June 198’71

SUMMARY

The leukemogens 7,12dimethylbenz[a)nthracene (DMBA) and 7,8,12trimethylbenz[crlanthracene (TMBA] bind covalently in vivo to DNA of LongEvans rats in the hematopoietic organs, spleen and bone marrow, and in the liver, a non-target organ. Both TMBA and DMBA depleted bone marrow cells and both agents bound persistently to the DNA of bone marrow and of liver, and less to that of spleen. The three main DMBA:deoxyribonucleoside adducts in spleen, bone marrow and liver were the same as those found previously in the liver (Dipple et al. (1983) Cancer Res., 43, 4132). There were no organ-specific or age-dependent differences in the relative amounts of adducts formed. There appears to be no direct correlation between the susceptibility of an organ to carcinogenesis and the nature and relative amount of the specific adducts formed, at least for the three organs studied here.

INTRODUCTION

The polycyclic aromatic hydrocarbons, DMBA and TMBA, are powerful carcinogens. DMBA is one of the most potent skin carcinogens known [14]. The compound also induces mammary gland carcinoma [8] and leukemia [9]. The reactive metabolites of DMBA which bind to mouse embryo cell DNA [5,7,13,16] and to mouse skin [3] have been characterized. The adducts are identical in both systems, namely, bay-region anti-dihydrodiol epoxide: deoxyguanosine (anti-DE : dG) an anti-dihydrodiol epoxide :deoxyadenosine untCDe : dA) and a bay-region s ylr-dihydrodiol epoxide : deoxyadenosine (syn*To whom reprint requests should be sent at: National Cancer Institute, Building 37, Room 3C38, Bethesda, MD 39893, U.S.A. 03943833/87/393.38 0 1987 Elsevier Scientific Publishers Ireland Ltd. Published and Printed in Ireland

42

DE:dAl adduct (the syn and anti define dihydrodiol epoxides wherein the benzylic hydroxyl group and epoxide oxygen are cis and trans to one another, respectively [3,7]1.TMBA has been less extensively studied than DMBA. The compound also induces leukemia and mammary gland carcinoma [9]. To date, neither the metabolites nor the DNA adducts of TMBA have been characterized. In this study, the covalent binding of DMBA to DNA was investigated in spleen and bone marrow, which are hematopoietic target organs, and compared to that in liver, which is a non-target organ. The specific DMBADNA adducts in these organs were identified to determine whether there is any correlation between the individual adducts formed and tumor initiation in these organs. The in vivo tumorigenic effects of both DMBA and TMBA have been extensively studied [9,10]. Huggins et al. [9] report that the two agents produce a similar incidence of leukemia in rats. We have compared the covalent binding of DMBA to DNA to that of TMBA to determine whether there is any correlation between the covalent binding level in target and non-target organs and the tumorigenic effects of the two agents. MATERIALS AND METHODS

Chemicals

Chemicals [G-*HpMBA and [G-3H]TMBA were obtained from Amershaml Searle Corp., Arlington Heights, IL and Chemsyn Science Labs., Lenexa, KS, respectively. Unlabeled DMBA and TMBA were a generous gift of Dr. C.B. Huggins (The University of Chicago, Chicago, IL). Proteinase K, deoxyribonuclease I (bovine pancreas) and phosphodiesterase I (Crotaeus atrox venom) were acquired from Sigma Chemical Co. (St. Louis, MO). Ribonuclease A (bovine pancreas1 and alkaline phosphatase (calf intestine) were purchased from Boehringer-Mannheim (Indianapolis, IN). Sep-Pak C,, cartridges were obtained from Waters Associates (Milford, MA). All other chemicals and reagents used were of the highest purity commercially available. In uivo treatment

Female 28dayold or ‘IO-day-oldLong-Evans rats (Charles River, Kingston, NY) were allowed food and water ad libitum throughout the study. DMBA (52 Cilmmoll or TMBA (4.9 Cilmmoll was injected intravenously. The tritiated compound, initially dissolved in benzene/hexane, was dried under a stream of nitrogen and dissolved in DMSO. It was then mixed with unlabeled material constituted in a fat emulsion. At least 3 animals per age group and per time point were used. Each animal received 2 mCi of radioactivity at a dose of 35 mglkg. The animals were sacrificed 2, 6, 24 and 48 h post-treatment. The liver and spleen were removed, weighed and immediately frozen in liquid nitrogen. Bone marrow cells were removed from the femur and tibia and washed in PBS. The tissues were then stored at - 70°C until DNA isolation.

43 Bats were killed and tissues (Westinghouse FG040 lamps).

were

processed

under

yellow

lights

Isolation and enzymic digestion of DNA The liver and spleen from treated animals were ground to a fine powder in a mortar kept on dry ice. Bone marrow cells were allowed to thaw out. DNA isolation was carried out as described by Vu et al. [17]. The radioactivity in an aliquot of the DNA from each rat was counted in an LS-9000 liquid scintillation counter (Beckman Instruments, Inc., Fullerton, CA). The covalent binding is expressed as pmol boundlmg DNA. The isolated DNA from the individual rats were digested into component nucleosides as described by Vu et al. [17]. Enzymes, buffer salts and any incompletely undigested DNA were removed by passing the digests through Sep-Pak C,, cartridges. Elution was carried out as described by Bigger et al. [3]. Fractions containing the modified nucleosides were evaporated to dryness under nitrogen and the residue was reconstituted in 300 d methanol. Toluene (50 4 of a lo-fold dilution in methanol) was added as a UV-absorbing marker. The samples were then analyzed by HPLC followed by radioactivity quantitation. Analysis of adducts HPLC analyses were performed using a Waters Associates system (Milford, MA) equipped with two M6OOOA pumps, a model 660 solvent programmer, a 440 UV detector set at 254 nm (for detection of toluenel, a S-pm LC-l&DB guard column (Supelco, Bellafonte, PA1 and a bm Altex Ultrazphere ODS (0.46 x 25 cm) column (Beckman Instruments, Inc., Berkeley, CA). Separation of the DMBA:deoxyribonucleoside adducts was carried out as described by Sawicki et al. [15]. The HPLC profile of previously characterized DMBA-DNA adducts from mouse embryo cells (provided by Dr. A. Dipple, NCI-Frederick Cancer Research Facility, Frederick, MD) was obtained under the same chromatographic conditions. The DMBA-DNA adducts in rat spleen, bone marrow and liver were identified by normalizing to the position of the internal standard toluene in the HPLC profile from the characterized DMBA-DNA adducts. RESULTS Binding of DMBA and TUBA to vat tissues in viva The levels of DMBA- and TMBA-DNA binding to spleen, liver and bone marrow are shown in Table 1. Significant binding was detectable in the three tissues studied. Following DMBA or TMBA administration in 23-dayold rats, the level of binding in spleen and bone marrow had already reached a maximum 2 h after administration of the chemical. With TMBA, these levels remained essentially constant over the period of the experiment in the three organs studied. With DMBA, the level of binding remained constant

26 70

23 70

28 70

Spleen

Liver

Bone marrow

6

35 f 3.9 39 * 1.0 41 f 3.1 33 f 3.7 14 * 1.2 16 f 1.4

2

51 f 4.5 49 f 4.5

14 f 1.3 13 f 1.6

13 f 0.7 12 f 2.1

DMBA Time (h)’

11 f 1.6 14zko.3

36 f 1.0 41 f 2.7

22 f 2.3 17 f 1.0

24

“Denotes period between administration of the carcinogen and sacrifice. bBone marrow was severely depleted at this time point.

Age (days)

Tissue

-b -b

32 2 1.9 33 + 2.4

16 f 1.7 17 f 1.6

32 f 4.4 27 f 3.0

47 f 5.3 54 f 2.9

35 f 3.9 30 f 2.9

43 2 pmoleslmg DNA

TMBA Time (hr 24

33 f 2.4 42 f 6.2 55 f 4.7 66 f 3.2 26k2.7 30*3.3

6

33 f 4.1 34 f 3.5 66 f 7.1 53 f 6.9 34 + 1.9 33 f 4.0

-b -b

52 f 3.5 63 f 7.9

46 f 4.3 49 f 5.3

43

Rats were given an intravenous injection of (G-“HJDMBAor [G?HHJTMBA (35 mg/kg). At the specified time points, the organs were removed and DNA was isolated and analyzed as described in Materials and Methods. The data are means f S.D. from at least 3 animals.

CARCINOGEN-DNA BINDING AT VARIOUS TIME INTERVALS IN FEMALE LONG-EVANS RATS

TABLE 1

FRACTION

NUMBER

Fig. 1.HPLC profile for DMBAdeoxyribonucleoside standards from mouse embryo cells (0) and DMBAdeoxyribonucleoside adducts from the spleen of a female 28day-old Long-Evans rat treated with [G-%IDMBA and killed 6 h post-treatment (0 1.Qualitatively similar profiles were obtained from spleen and liver 2, 6, 24 and 48 h following DMBA treatment and from bone marrow 2 and 6 h following DMBA treatment (bone marrow cells were depleted at the later time points). The arrow denotes the position of elution of an added toluene UV marker.

over 24 h only in bone marrow. In the spleen, there was a gradual decline in the level of DMBA-DNA adducts with time, while in the liver peak binding occurred after 6 h (Table 1). The level of DMBA adduct binding to bone marrow DNA is significantly lower (P < 0.051 than the level of binding to spleen DNA (Table 11. In contrast, following administration of TMBA, there is no signficance (P > 0.01) in the level of binding between spleen and bone marrow over the first 6 h (Table 1). Both DMBA and TMBA caused a dramatic depletion of bone marrow cells which was noticeable after 24 h. The bone marrow cells were almost totally depleted after 48 h, no DNA was isolated and thus no covalent binding levels are reported for this time point. The covalent binding of DMBA and TMBA occurred to a similar extent and followed the same pattern in 28-day-old and 70dayold Long-Evans rats (Table 11, with the exception of the profile in spleen following TMBA administration. In the latter case, there was a gradual increase in the level of TMBA-DNA adducts over 48 h. Identity of DMBA-DNA vivo

adducts formed after administration of DMBA

in

DMBA-DNA adducts formed in mouse embryo cells have been characterized [5,7,13,15]. Comparison to the HPLC retention times of these

Bone Marrow

W-n anti-DE:dG

I

62446

2

6 24

Time (h)

2

62446

2

6

Time (h)

62446

2 6 24

Time (h)

.iver

Bans Marrow

syn-DE:dA 20r

0s

2 62446

anti-DE:dA 40$ 20-

0

2

62446

2

anti-DE:dG

0

ml 2

6 2446

2

6 2446

2

6 24

Time (h)

2

62446

2

6

Time (h)

2

62446

2

6 24

Time (h)

syn-Dt3.4

T OI

2

62446

anti-DE:dA 40-

20-

0

2

6 2446

Fig. 2. Profile for the DMBAdeoxyribonucleoside adducts from 28dayeld (A) or ‘IO-day-old(B) female Long-Evans rats treated with [Gl”H]DMBA. The rats were killed 2. 8, 24 and 48 h following treatment. No values are reported for bone marrow 24 and 48 h post-treatment because of depletion of bone marrow cells at these time points. Values are expressed as a percentage of the total radioactivity from a particular HPLC run and represent the mean f S.D. from 3 individual rats.

47

characterized adducts enabled identification of the adducts formed in the spleen, bone marrow and liver of Long-Evans rats in vivo. Figure 1 shows the HPLC elution profile of the DMBAdeoxyribonucleoside adducts. The major peaks have been identified in earlier studies p,13,15] as anti-DE:dG, syn-DE:dA and anti-DE:dA. Adducts anti-DE:dG and anti-DE:dA are the predominant adducts, syn-DE:dA formation occurs to a significantly lesser extent (Fig. 1). Figure 2 shows the profile of adduct formation, expressed as a percentage of the total radioactivity recovered during a particular HPLC run, in spleen and liver at 2,6,24 and 48 h, and in bone marrow at 2 and 6 h post-treatment. The total radioactivity recovered from bone marrow DNA 24 and 48 h post-treatment was too low to allow characterization of the adducts. The overall pattern of adduct formation was similar in the three tissues studied. Moreover, the contribution of each of the three adducts to the overall binding, when expressed in terms of pmoleslmg DNA, was independent of the gross DMBA-DNA binding. Thus, there are little or no qualitative differences in the elution profiles obtained from different tissues in 28-day-old and ‘IO-day-old rats, even though there are minor quantitative differences. DISCUSSION

This study shows that the leukemogenic agents DMBA and TMBA bind covalently to DNA in the hematopoietic organs (spleen and bone marrow) and in liver. DMBA has previously been shown to bind to primary mouse embryo cell DNA [5,7,X] and mouse skin DNA [3]. The covalent interaction of TMBA to DNA has not been reported to date. DMBA and TMBA produce a similar incidence of leukemia [9] and similar chromosomal abnormalities [16] in Long-Evans rats. Bone marrow cells are the target cells for the hematopoietic carcinogenic activity of DMBA [2]. The similarity between the in vivo effects of the two agents suggests that the bone marrow is also the target cell for TMBA carcinogenesis. For both TMBA and DMBA, the DNAadduct covalent binding levels did not correlate with the organ specificity of these carcinogens. Daniel and Joyce [4] also report a lack of correlation between the levels of DMBA-DNA adduct formation in the mammary gland. a target organ, and the liver, a non-target organ. Similarly, the profile of overall (for TMBA and DMBA) and specific (for DMBA) adduct formation was independent of the age of the animals. This is in contrast to the agedependent incidence of leukemia in Long-Evans rats reported by Huggins et al. [9]. Thus, even though the tissues studied contain a heterogenous cell population, and it is thus possible that the binding to DNA is not random within various cell types in an organ or within the genome of individual cells, these data indicate that DNA binding alone does not explain the organ specificity of these hydrocarbons, even though it may be one of the factors contributing to this organ specificity. As with the overall covalent binding of DMBA, there appears to be no

48

direct correlation between the susceptibility of an organ to carcinogenesis and the nature and relative amount of the specific adducts formed, at least in the three organs studied here. Similar findings have been reported for benzo@lpyrene in mice, where an identical metabolite: DNA adduct profile was observed in lung and forestomach, which are target organs, and in liver, which is a non-target organ [l,ll]. However, following pre-treatment with the antioxidant butylated hydroxyanisole, which inhibits benzo[alpyrene:DNA metabolism, the decrease in the levels of benzoluJpyrene:DNA adduct formation did correlate with the reduced tumor incidence in susceptible tissues [1,12]. Similarly, the specific activities of benzo(o] pyrene adducts in lung and liver in different mouse strains show a direct correlation to the susceptibility to benzo[@yrene-induced pulmonary adenomas [1,12]. This is in contrast to the DMBA:DNA adduct profile, where no individual adducts are formed in notably greater quantities in the Swiss NIH 3T3 mice, which strain is more sensitive to initiation by DMBA, when compared to the less sensitive C57BL mice [S]. The overall conclusion that can be drawn from this study is that other factors, perhaps connected with tumor promotion (such as sensitivity to hormones) may be important in determining the in vivo tumorigenic effects of DMBA and TMBA. ACKNOWLEDGMENTS

We are grateful to Dr. A. Dipple of the NC&Frederick Cancer Research Facility, Frederick, MD, for supplying the characterized 7,12dimethylbenz[a]anthracene adducts, and to Dr. C.B. Huggins (The University of Chicago, Chicago, IL) for his gift of 7,12dimethylbenz[u]anthracene and 7,8,12-trimethylben@u[alanthracene. We thank Mrs. F. Williams for preparing the manuscript. REFERENCES Anderson, M.W., Boroujerdi. M. and Wilson, AGE. (1981) Inhibition in vivo of the formation of adducts between metabolites of bensdalpyrene and DNA by butylated hydroxyanisole. Cancer Res.. 41, 4309-4315. Ball. J.K. (1988) Role of bone marrow in induction of thymic lymphomas by neonatal injection of 7.12dimethylben~aJanthracene. J. Natl. Cancer Inst., 41, 553-558. Bigger, C.A.H., Sawicki, J.T.. Blake, D.M., Raymond, L.G. and Dipple, A. (1983) Products of binding of 7.12dimethylben~akuithracene to DNA in mouse skin. Cancer Res., 43, 56475851. Daniel, F.B. and Joyce, NJ. (1984) 7,12Dimethylbenz(a~nthracene adducts in SpragueDawley and Long-Evans female rats: the relationship of DNA adducts to mammary cancer. Carcinogenesis. 5,1021- 1028. Dipple, A. and Nebzydoski, J.A. (1978) Evidence for the involvement of a diol-epoxide in the binding of 7.12-dimethylben~a~nthracene to DNA in cells in culture. Chem.-Biol. Interact., 20,17-28. Dipple. A., Pigott, M.A., Bigger, C.A.H. and Blake, D.M. (1984) 7,12Dimethylbene[a]anthracene-DNA binding in mouse skin: response of different mouse strains and effects Of VsriOUSmodifiers of earcinogenesis. Carcinogenesis, 5, lO87-11090.

49

Dipple, A., Pigott, M., Moschel, R.C. and Costantino, N. (19831 Evidence that binding of 7,12dimethylbemjahmthracene to DNA in mouse embryo cell cultures results in extensive substitution of both adenine and guanine residues. Cancer Res., 43,4132 - 4135. 8 Huggins, C.B. and Fukunishi, R. (19631 Mammary and peritoneal tumors induced by intraperitoneal administration of 7,12dimethylbenqa]anthracene in newborn and adult rats. Cancer Res., 23,785-789. 9 Huggins, C., Grand, L. and Oka, H. (19701 Hundred day leukemia: preferential induction in rat by pulse doses of 7.8,12-trimethylben$a[alanthracene. J. Exp. Med., 131,321-330. 10 Huggins. C.B., Grand, L. and Ueda, N. (19821 Specific induction of erythroleukemia and myelogenous leukemia in Sprague-Dawley rats. Proc. Natl. Acad. Sci. U.S.A., 79, 54115414. 11 Ioannou, Y.M., Wilson, A.G.E. and Anderson, M.W. (19821 Effect of butylated hydroxyanisole on the metabolism of benzo(alpyrene and the binding of metabolites to DNA, in vitro and in vivo, in the forestomach, lung and liver of mice. Carcinogenesis, 3, 739 - 745. 12 Ioannou, Y.M.. Wilson, A.G.E. and Anderson, M.W. (19821 Effect of butylated hydroxyanisole, a-angelica la&one, and /3naphthoflavone on bensofalpyrene:DNA adduct formation in vivo in the forestomach, lung and liver of mice. Cancer Res., 42, 1199- 1204. 13 Moschel, R.C.. Pigott,, M.A.. Costantino, N. and Dipple, A. (1983) Chromatographic and fluorescence spectroscopic studies of individual 7,12dimethylbenz[a]anthracenedeoxyribonucleoside adducts. Carcinogenesis, 4, 1201- 1204. 14 Pataki, J. and Huggins, C. (19691 Molecular site of substituents of benz[a]anthracene related to carcinogenesis. Cancer Res., 29, 506 - 509. 15 Sawicki, J.T., Moschel, R.C. and Dipple, A. (19831 Involvement of both syn- and antidihydrodiol epoxhles in the binding of 7,12dimethylhenz(a]nthracene to DNA in mouse embryo cell cultures. Cancer Res.. 43, 3212 - 3218. 16 SugiYama, T. and Brillantes, F.P. (19701 Cytogenetic studies of leukemia induced by 6,8,12and 7.8,12-trimethylbenalaz[alanthracene.J. Exp. Med., 131, 33I- 341. 17 vu. Y.T., Moller, M.E., Grantham, P.H., Wirth, P.J. and Thorgeirsson, S.S. (19851 Association between DNA strand breaks and specific DNA adducts in murine hepatocytes following in viva and in vitro exposure to N-hydroxy-2-acetylaminofluorene and N-acetoxy-2acetYlaminofhiorene. Carcinogenesis, 6, 45- 52. 7