Chem.-Biol. fnteructionr, 9 (1975) 157-167 0 Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands

METHYLATION OF RAT LIVER MITOCHONDRIAL CLEIC ACID BY CHEMICAL CARCINOGENS AND TERATIONS IN PHYSICAL PROPERTIES

RAY WILKINSON*,

ANDREW

HAWKS

AND

ANTHONY

157

DEOXYRIBONUASSOCIATED AL-

E. PEGG

Courtcluld Institute of Biochemistry, Middlesex Hospital Medical School, Lot&n Britain)

WlP SPR (Gwat

(Received May 22nd, 1974) (Revision received August 6th, 1974)

SUMMARY

The formation of 7-methylguanine in rat liver mitochondrial DNA following the administration of the powerful carcinogen, dimethylnitrosamine, and the weak carcinogen, methyl methanesulphonate was measured and compared to the alkylation of nuclear DNA by these agents. At all doses tested mitochondrial DNA was alkylated more extensively than nuclear DNA by dimethylnitrosamine but both types of cellular DNA were alkylated to about the same extent by methyl methanesulphonate. The physical structure of rat liver mitochondrial DNA isolated from animals treated with these agents was investigated by electrophoresis in agarose gels and by isopycnic centrif’ugation in CsCl gradients. These procedures carried out in the presence of ethidium bromide, an intercalating dye, separate closed circular forms of mitochondrial DNA from open circular molecules (containing a single-strand breiik) and linear molecules. Administration of dimethylnitrosamine produced a considerable decrease in the amount of mitochondrial DNA which could be isolated in the closed circular form and at higher doses of dimethylnitrosamine no closed circular mitochondrial DNA could be found. Methyl methanesulphonate was less effective at reducing the amount of closed circular mitochondrial DNA. One explanation of these results is that dimethylnitrosamine leads to strand breaks in mitochondrial DNA and the possible use of this system to investigate carcinogen-induced breaks in DNA is discussed.

* Present address: Department of Microbiology, University of Colorado Medical Center, Denvr;:, Colo. 80220 (U.S.A.).

158

A general hypothesis concerning the ability of a wide range of chemicals to induce cancer suggests that an essential prerequisite for a compound to be carcinogenic is that it is converted within the ccl1 to an active electrophilic reactantr. Such unstable reacta& form covalent bonds with many cellular constituents and it is not yet known which of these reactions are vital in initiating the appearance of tumours. Most attention has been given to interactions with DNA, since damage to DNA could lead to the permanent inheritable changes seen in neoplasia. In this respect the binding ofmany types ofchemical carcinogens to nuclear DNA has been stud&P4. In addition to the altered nucleosides resulting from such reactions, the possibility that breaks in the DNA chain might also be produced has been put forward and some evidence for the existence of single-strand breaks in DNA after administration of carcinogens has heen published+s. In addition to the nuclear DNA of mammalian cells there is also DNA present in mitochondria. The occurrence, physicochemical properties and function of this DNA have been extensively studied during the past decades-11. The smaller size and homogeneity of mitochondrial DNA renders the structure of this DNA much more amenable to analytical methods presently available than nuclear DNA. It is now well establshed that the DNA in mitochondria from all mammalian tissues consists of circular duplex molecules of M.W. about 107 (refs. 9-11). The need for a complete analysis of the reactions of alkylating carcinogens with cellular DNA led to the present studies of the effects of such carcinogens ,on mitochondrial DNA. The only other studies of this nature so far reported have measured the fomation of 7-methylguanine in mitochondrial DNA of liver and kidney from rats treated with dimethylnitrosamine and N-methyl-N-nitrosourearzJs. 7-Methylguanirie is the major reaction product when these carcinogens interact with DNA14 amounting to some %I%of the total incorporation of methyl groups. It was reported that the alkylation at this site of DNA from mitochondria was more extensive than that cf nuclear DNA although only one concentration of the carcinogens was tested**m. The present paper describes the degree of alkylation of mitochondrial and nuclear DNA of rat liver produced by various doses of dimethylnitrosamine, a liver carcinogen, and methyl methanesulphonate, an alkylating agent which although carcinogenic under certain conditions 15e35,has never been shown to produce liver tumours in the rat. In addition, mitochondrial DNA from livers of rats treated with these compounds was analyzed using electrophoresis in agarose gels in the presence ofethidium br~mide~~. This procedure separates intact closed circular duplex mitochondrial DNA molecules from open circular duplex forms (which ran be derived from the former by singl~trand breaks).

CsCl was purchased from Harshaw Chemical Co., Solon, Ohio, U.S.A. A stock

159 solution of 118 g/l00 ml of 0.01 M Tris-HCl, 0.001 M EDTA of pH 8.2 at 20” was prepared and used as required. Dimethylnitrosamine and methyl methanesulphonate were obtained from Eastman-Kodak Chemicals, Kirby, England. Ethidium bromide was a gift from Boots Pure Drug Co., Nottingham, England. Agarose was purchased from Sea Kern Marine Colloids, Inc., Rockland, Maine, U.S.A. [t4C]Dimethylamine (22 mCi/mmole) was purchased from the Radiochemical Centre, Amersham, England. [r‘C]Dimethylnitrosamine was kindly prepared by Dr. P. F. Swann from the (r4C]dimethylamine. Animals

Wistar strain rats were bred in the Institute and fed on Rowett research diet 86. Rats (170-180 g body wt) were fasted overnight prior to the injection of drugs and isolation of mitochondria. Dimethylnitrosamine and methyl methanesulphonate were diluted in 0.9% NaCl solution and a volume of 0.5 to 1.0 ml was administered by intraperitoneal injection. The rats were killed by cervical dislocation 4 or 5 h later and the livers removed and used for preparalion of mitochondria. Preparation of mitochondria

Mitochondria were prepared by differential centrifugation, essentially by the method of Schneider and Kuff” as modified by LefIler et af.18. Livers from 4 rats were rinsed and finely minced with scissors in ice-cold 0.25 M sucrose. All subsequent operations were carried out at 4”. The livers were then homogenized19 in 10 vol. of 0.25 M sucrose, and the homogenate centrifuged at 2500 rev./min for 15 min in a Mistral 6L centrifuge (MSE, Crawley, England) to remove nuclei and whole cells. The supernatant was removed to within 1 cm of the pellet with a syringe. The supernatant was then centrifuged for 15 min at 10 000 rev./min (9000 x g) in the 8 x 50 rotor of the MSE 18 centrifuge to pellet the mitochondria. The supernatant was removed and the pellet resuspended in 0.25 M sucrose, 0.001 M EDTA, pH 7.4 using a hand-held glass homogenizer with a teflon pestle (Thomas, Philadelphia, U.S.A. ; Size C). The mitochondrial suspension was centrifuged at 2500 rev./min for 15 min and the supernatant removed with a syringe as before. The mitochondria were pelleted at 10 000 rev./min for 15 min, resuspended and then washed twice by centrifugation and resuspension. The final mitochondrial pellet was suspended in 0.25 M sucrose, 0.001 M EDTA, 0.01 M Tris-HCI, pH 8.2 and made up to a volume of 2.25 ml. Mitochondrial proteinso, RNA21 and DNA22 content were then estimated. The DNA content of the mitochondrial fraction was about 0.A.5 pg/mg protein and the RNA content was 5-7 pg/mg protein in reasonable agreement with published values17sr8. Preparation of mitochondrial DNA The mitochondrial suspension (2.25 ml) was transferred

to a Spinco 40 rotor cellulose nitrate centrifuge tube and 0.25 ml of 10% (w/v) N-lauroylsarcosinate in 1 mM EDTA, 10 mM Tris-HCl, pH 8.2 at 20” was added to lyse the mitochondria and release the DNA. A stock solution of CsCl was then added to bring the volume to 12.5 ml and the density to 1.70 g/ems. The sample was then centrifuged at 33 000 rev./

160 min (70 000 x g) for 65 h at 20” in the 50 Ti rotor of a Spinco L2 centrifuge. Under these conditions the DNA bands in the centre of the tube, protein floats at the top of the gradient and RNA is pelleted. The fractions conraining DNA were collected and recycled~. The 6nal DNA pellet was dissolved in the appropriate solution for further analysis. Prepmatrn of nuclear DNA The pellet of nuclei obtained during the isolation of mitochondria was resuspended in 0.25 M sucrose, filtered through fine mesh gauze and washed twice. DNA was extracted from the nuclear pellet either by deproteinization with phenol’*, or by fruition in CsCl as follows. An aliquot of nuclear sus~nsion (confining about 0.1 mg DNA) was lysed with N-lauroylsarcosinate as described above for mitochondria. The DNA was then sheared by a 3-see treatment in an Ultraturrax homogenizer

(Janke and Kunlde K.G.) and isolated by centrifugation in CsCl density gradients as described above. Determ&ution

ofextent ofmethylation

DNA isolated from rats treated with [WJdimethylnitrosamine or [W]methyl methanesulphonate was hydrolyzed in 1 N HCl at 100”for 1 h and the released purine bases separated as previously describedr4. The guanine com~ition of the mit~hondrial DNA was taken as 19.4 moies[ 100 moles bases” and the degree of alkylation was calculated as the amount of 7methylguanine produced per 100 moles of guanine present. Gel electrophoresis ofmitochondrial DNA Separation of closed circular mit~hond~l

DNA (type I) from open circular

forms(type II) and linear forms (type III) was carried out by electrophoresis on agarose slab gels as described by Aaij and Borst 1s. The gels used routinely contained 0.5 % agarose and lo&ml ethiclium bromide. Increasing the ethidium bromide concentration to 100 &ml did not significantly increase the resolution. The gels were examined under short-wave ultravio~t light and photography with a Polaroid camera (Model MP3), using L&z orange and ultraviolet filters and black and white film 42. The gels and photographs were scanned with a Joyce-Loebl gel scanner. BunaYngof mitochondriui DNA in CsCI in the presence of ethidium bromide

~it~hondrial DNA (2-5 pg) was ~ntrifug~ to ~~librium in a solution of 0.01 M Tris-HCI, 0.001 M EDTA of pII 8.2 and containing 0.1 mg ethidium bromide/ ml and sufficient CsCl to bring the density to 1.58 g/cmi The total volume of 3 ml was centrifuged as described by Nassr* and the tube then photographed as described above. RESULTS

Methylation of rat liver DNA Rats were injected intraperitoneally with {l~c~imethylnitro~m~e

or [14C]-

161 TABLE I ALKYLATION OF LIVER NUCLEAR AND MITOCHONDRIAL

DNA

FOLLOWING ADMINISTRATION OF DIMETHYL-

NITRO!JAMINF,AND METHYL METHANFSULPHONATE

Rats were injected intraperitoneally with [W]dimethylnitrosamine (DMN) and [r%]methyl methanesulphonate (MMS) and killed 4-5 h later. DNA was isolated from the liver dnd the percentage of guanine converted to 7-methylguaninedeterminedas describedin the text. (At least 150 cpm were present in the 7-methylguanine of the mitochondrial DNA analysed). Treatment (per kg body wt.)

3.4 mg DMN 12 mg DMN 27 mg DMN 60 mg MMS 120 mg MMS

Alkylution of DNA ( % guanine converted into 7-methylguanine) Nuclear DNA

Mitochondrial DNA

0.05 0.23 0.72 0.04 0.09

0.14 0.45 1.21 0.06 0.10

methyl methanesulphonate and killed 5 or 4 h later, respectively. Nuclear anti mitochondrial DNA were prepared from the livers and the percentage of guanine converted to 7-methylguanine determined (Table I). (Nuclear DNA was isolated either by extraction with phenol or by CsCl isopycnic centrifugation. Both methods gave similar results, which were in agreement with other published estimates for the formation of 7-methylguanine in rat liver by these agent@.) Results for 3 doses of dimethylnitrosamine (3.4, 12.0 and 27.0 mg/kg body wt. respectively) are shown in Table I. Alkylation of nuclear and mitochondrial DNA was roughly proportional to the dose of dimethylnitrosamine but, at all doses tested, mitochondrial DNA was alkylated more extensively than nuclear DNA. At the lowest dose of dimethylnitrosamine the formation of ‘I-methylguanine in mitochondrial DNA was 2.8 times that of the nuclear DNA and at the highest dose this ratio was 1.7. It was not possible to carry out this experiment a sufficient number of times to establish the statistical significance of this apparent fall in the preferential alkylation of mitochondrial DNA with increasing dose of dimethylnitrosamine because of the small amount of mitochondrial DNA present in rat liver and the high cost of the radioactively labelled carcinogen but it is clear that there is preferential alkylation of the 7-position of guanine in mitochondrial DNA at all doses tested. For the same reason we were unable to determine the amounts of the other minor alkylated products of the reaction of dimethylnitrosamine2~4~s with mitochondrial DNA and the possibility that there may also be qualitative differences in the relative amounts of these products formed in mitochondrial DNA as opposed to nuclear DNA cannot be ruled out. The alkylation of DNA produced by two doses (60 and 120 mg/kg body wt.) of methyl methanesulphonate also showed approximate proportionality between dose and degree of alkylation. However, the increased methylation of mitochondrial compared with nuclear DNA was only 1.2$-fold and this difference is unlikely to be sig nificant.

162

Fig. 1. Separation of mitochondrial DNA from animals treated with dimethylnitrosamine (DMN) by gel electrophoresis. Mitochondrial DNA was isolated from animals treated as indicated and 3 pg was applied to the gel. All gels contained 0.5 % agarose and 10 pg/ml ethidium bromide and were run for 3 h at 70 mA. After electrophoresis the gels were photographed and the photographs scanned for refkted light giving the tracings shown; .(a} open circular DNA; (b) closed circular DNA; (c} RNA; (4 top of gel.

a

Fig. 2. Separation of mitochondrial DNA from animals treated with dimethylnitrosamine (DMN) and methyl methaaesulphonate (MMS) by centrifugation in CsCl gradients containing ethidium bromide. Mitochondriaf DNA was isoWed from animals treated as indicated and aboilt 3 pg was centrifugedin C&l ~adien~ containing 100 $g ethidium bromide per ml. After ~ntrjf~~ation the tubes were photographed giving the photographs shown. The photographs were then scanned for reflectedlight as describedfor Fig. 1.; (a) open circularDNA, (@ closed circular DNA.

163 Separation of diflerentforms of mitochondrialDNA Mitochondrial DNA was extracted from the livers of rats treated with dimethylnitrosamine or methyl methanesulphonate and subjected to electrophoresis on agorose gels in the presence of ethidium bromide is. This procedure separates the closed circular duplexes of mitochondrial DNA (Type I) from open circular duplexes (Type II) and linear DNA (Type III). In our experiments, any linear mitochondrial DNA would have run with the open circular form, Type II. As can be seen from Fig. 1, administration of dimethylnitrosamine led to a reduction in the amount of closed circular form. Some decrease in the amount of closed circular mitochondrial DNA was seen with the lowest (3.4 mg/kg) dose of dimethylnitrosamine tested. This amount was further decreased as the dose of dimethylnitrosamine was increased until after a dose of 27 mg/kg no closed circular forms were detected. A similar loss of closed circular DNA was seen when the mitochondrial DNA from rats, which had received 12 mg/kg dimethylnitrosamine, was studied by centrifugation in CsCl gradients in the presence of ethidium bromide24 (Fig. 2). The possible repair of the mitochondrial DNA was investigated by analysing the DNA at later time points after treatment of the rats with dimethylnitrosamine. Rats killed up to 24 h after treatment showed no significant alterations in the amounts of closed and open circular mitochondrial DNA present at 5 h. Further work is required before this finding can be interpreted as showing the absence of repair of the damage produced to mitochondrial DNA by the carcinogen. However, it is unlikely to be due to the continued production of damage to the mitochondrial DNA by the presence of dimethylnitrosamine since it is known that virtually all of a dose of dimethylnitrosamine similar to that used in these experiments is metabolised within 5-6 h (refs. 14, 36). No quantitative estimation of the relative decline in Type I DNA could be made

couv?oL

~2omg/koMhAs

3.hgkg

DMU

Fig. 3. Separation of mitochondrial DNA from animals treated with methyl methanesulphate (MMS) by gel electrophoresis. Isolation and separation of mitochondrial DNA were carried out 9s described in Fig. 1.

164 with the equipment available. Rough calculations of the area under the tracings corresponding to the various peaks can be made but since the base line was not accurately defined these cannot be used directly in order to determine the relationship between the reduction in unbroken circles and the degree of alkylation of the mitochondrial DNA. It would be possible to measure the relative changes in the DNA of a particular class after treatment with carcinogen by use of the equipment described by Watson et ~1.37. This technique would however be only approximately quantitative in relating the amount of one class of DNA to that of another in the same: gel or gradient as the fluorescence varies with the type of DNA in the bands’. Administration of 120 mg/kg methyl methanesulphonate led to very little reduction in the amount of closed circular mitochondrial DNA (Fig. 3). This dose of methyl methanesulphonate methylated mitochondrial DNA almost to the same degree as 3.4 mg/kg of dimethylnitrosamine which produced a clear decrease in the amount of closed circular DNA. There was also very little loss detectable in closed circular DNA after methyl methanesulphonate treatment when the DNA was analyzed with CsCl gradients with the presence of ethidium bromide (Fig. 2). CONCLUSIONS

The present results on the formation of 7-methylguanine in nuclear and mitochondrial DNA by dimethylnitrosamine confirm the previous reportls. that mitochondrial DNA is preferentially alkylated, although the ratio of alkylation was about 1:2 for all of the doses of dimethylnitrosamine tested rather than I :4.5 as previously reportedrs. The reason for the preferential alkylation of mitochondrial DNA is not clear but may be related to the easier access of the alkylating species generated by metabolism of dimethylnitrosamine to mitochondrial DNA than to nuclear DNA. This could be caused by (I) a lesser mrmber of protective molecules (proteins, etc.) closely bound to the mitochondrial DNA and preventing access of, or preferentially reacting with, the methylating agent; (2) the easier entry of the methylating species into the mitochondrion than into the nucleus; or (3) the greater proximity of mitochondrial DNA to the site of generation of the methylating agent produced from dimethylnitrosamine. Methyl methanesulphonate which acts as a methylating agent by a direct chemical reaction not requiring metabolic intervention, produced only a very slight preferential alkylation of mitochondrial DNA over nuclear DNA. It is noteworthy that in previous studies of the alkylation of liver RNA and DNA by various methylating agents it was found that dimethylnitrosamine alkylated RNA to a greater extent than nuclear DNA but that methyl methanesulphonate alkylated nuclear DNA to a greater extent than RNA14. Admirutration of dimethylnitrosamine led to an increase in the content of closed circular forms of mitochondrial DNA. Although accurate quantitation was not possible with the equipment available, the decrease in the open circular forms appeared to be approximately proportional to the administered dose of dimethylnitrosamine, which in turn is directly related to the degree of alkylation of the mitochondrial DNA. Although other interpretations of these data are possible the simplest

165 explanation is that reaction with the carcinogen leads to single-strand breaks in the mitochondrial DNA producing the open circular form from the normal closed circles. The possible occurrence of single-strand breaks in nuclear DNA after the administration of dimethylnitrosamine has recently been reported’. A large dose of methyl methanesulphonate (120 mg/kg) gave only a slight indication of a decrease in closed circular forms of mitochondrial DNA molecules although a dose of 3.4 mg/kg dimethylnitrosamine which produced only a slightly greater amount of alkylation of mitochondrial DNA gave a clear decrease. This result suggests that dimethylnitrosamine is more effective than methyl methanesulphonate in producing the change in the physical structure of mitochondrial DNA not only on the basis of the applied dose as is obvious from Figs. l-3 but also on the basis of the degree of alkylation at the 7-position of guanine. Methyl methanesulphonate was reported to produce single-strand breaks in nuclear DNA of cells in culture25-27 and rat liver6 although in the latter case the lesion appeared to be repaired more rapidly than that produced by dimethylnitrosamine6v7. Some theories requiring the involvement of cytoplasmic genes in carcinogenesis28329have been put forward but at present there is no convincing reason for the hypothesis that mitochondrial DNA is a more important target molecule for the action of chemical carcinogens than nuclear DNA. Although “abnormal” catenated forms of mitochondrial DNA were first observed in tumour cellsso such forms can also be isolated from non-neoplastic cells 1OJlJz. On the other hand, the significance of alkylation, or the production of single-strand breaks in nuclear DNA, is also not firmly established, although under intensive investigation and there is no reason to exclude mitochondrial DNA from such considerations. The present technique for studying mitochondrial DNA could be of value in investigating the possible interactions of carcinogens with DNA in viva in tissues in which tumours are produced. It has the advantage that neither carcinogen nor DNA need be radioactively labelled. The alkaline sucrose gradient technique for studying single-strand breaks in DNA 697~5-27has the disadvantage that extensive labelling of the DNA has been used in order to detect the small amount of material which can be separated on the gradients. Such labelling can only be achieved in certain tissues and the effect of the incorporation of components of very high specific radioactivity on the stability of the DNA is not known. The method also suffers from the disadvantages that it is not known how many breaks are needed in order to prevent the DNA from sedimenting through the gradient in the same way as control DNA and the strongly alkaline conditions might actually generate breaks from alkali-labile sites33B34.In contrast, only one break per molecule is required to change closed circular mitochondrial DNA into the open circular form and the separation is carried out under neutral conditions with unlabelled mitochondrial DNA which can be isolated from many tissues. Thus, providing the agent causing single-strand breaks in mitochondrial DNA can enter the mitochondrion, this procedure should provide a sensitive test for the formation of such agents from carcinogens (or other compounds) in V&L

166 ACKNOWLEDGEMEh”fS

This research was supported by the Cancer Research Campaign and the Sir Michael SobeU Feliowship of the Cancer Research Campaign. We thank Professor P. N. Magee and Dr. P. F. Swann for their advice and interest.

REFERENCES I J. A. Miller, Carcinogen&s by Ghemicals: An overview-G. H. A. Clowes Memorial lecture, Cancer Res., 30 (1970) 559-576. 2 P. D. Lawley, Effects of some chemical mutagens and carcinogens on nucleic acids, Progr. Nucf. Acid Res. Mol. hid., 5 (1966) 89-131. 3 J. A. Miller and E. C. Miller, The metabolic activation of carcinogenic aromatic amines and amides, Prog. Exptl. Tumor Res., 11 (1969) 273-301. 4 P. N. Magee, P. F. Swann and A. E. Pegg, Molecular mechanisms of carcinogenesis, in E. Gruadmann (Ed.), Handbuch der Alfgemeinet~Pafhologie, Springer, Heidelberg, 1974, in press. 5 P. J. O’Connor, M. J. Capps and A. W. Craig, Comparative studies of the hepatocarcinogen N,Ndimethylnitrosamine in viva: reaction sites in rat liver DNA and the significance of their relative stabilities, f&it. J. Cancer, 27 (1973) 153-166. 6 R. Cox, 1. Damjanov, S. E. Abanobi and D. S. R. Sarma, A method for measuring DNA damage and repair in the liver in viva, Cancer Res., 33 (1973) 211+2121. 7 1. Damjanov, R. Cox, D. S. R. Sarma and E. Farber, Patterns of damage and repair of liver DNA induced by carcinogenic methylating agents in viva, Cancer Res., 33 (1973) 2122-2128. 8 J. I. Goodman, Hepatic DNA repair synthesis in rats fed 3’-methyl4dimethylaminoazobenzene, Rio&em. Biophys. Res. Commun., 56 (1974) 52-59. 9 M. M. K. Nass, Mitochondrial DNA: advances, problems, and goals, Science, 165 (1969) 25-35. 10 P. Borst, Mitochondrial nucleic acids, Ann. Rev. Biochem., 41 (1972) 333-376. 11 P. Borst, Mitochondrial nucleic acids, Biochem. Sot. Trutts., (1974) in press.

I2 V. Wunderlich. M. Schiitt, M. Battger and A, Graffi, Preferential alkylation of mitochondrial deoxyribonucleic acid by N-methyl-N-nitrosourea, Biochem. J., 118 (1970) 99-109. 13 V. Wunderlich, I. Tetzlaff and A. Graffi, Studies on nitrosodimethylamine: Preferential methylation of mltochondrial DNA in rats and hamsters, Chem.-Biol. Interactions, 4 (1971/72) 81-89. 14 P. F. Swann and P. N. Magee, Nitrosamine-induced carcinogerresis: The alkylation of nucleic acids of the rat by N-methyl-N-nitrosourea, dimethylnitrosamine, dimethyl sulphate, and methyl methanesulphonate, Lliochem.J., 1IO (1968) 39-47. 15 P. F. Swann and P. N. Magee, Induction of rat kidney tumours by ethyl methanesulphonate and nervous tissue tumours by methyl methanesulphonate and ethyl methanesulphonate, Narure, 223 (1969) 947-949. 16 C. Aaij and P. Borst, The gel electrophoresis of DNA, Biochint. Biophys. Acra, 269 (1972) 192-200. 17 W. C. Schneider and E. L. Kuff, The isolation and some properties of rat liver mitochondrial deoxyribonucleic acid, Proc. Nat!. Acud. Sci. (U.S.), 54 (1965) 1650-1658. 18 A. T. LeWer, E. Creskog, S. W. Luborsky, V. McFarland and P. T. Mora, Isolation and characterization of rat liver mitochondrial DNA, J. Mol. Eiol., 48 (1970) 455-468. 19 W. N. Aldridge, R. C. Emery and B. W. Street, A tissue homogenizer, Biochem. J., 77 (1960)

326-327. 20 0. H. Lowry, N. J. Rosebrough, A, L. Farr and R. J. Randall, Protein measurement with Folin phenol reagent, J. Eiol. Chetn., 193 (1951) 265-275. 21 Z. Dische, Colour reactons of nucleic acid components, in E. Chargaff and J. N. Davidson (Ed%), The Nttcleic Acids Vol. 1. Academic Press, New York, 1955, pp. 285305.

22 K. Burton, A study of the conditions and mechanism of the diphenylamine reaction for the calorimetric estimation of deoxyribonucleic acid, Biochem. J., 62 (1956) 315-323. 23 A. Hawks, E. Farber and P. N. Magee, Equilibrium centrifugation studies of colon DNA from mice treated with the carcinogen 1,2dimethylhydratine, Chem.-B/o/. fttrernctiotts, 4 (1971-72) 144-148. 24 M. M. K. Nass, Mitochondrial DNA, II. Structure and physicochemical properties of isolated DNA, J. Mol. Biol., 42 (1%9) 529-545.

167 25 M. B. Coyle and B. S. Strauss, Characteristics of DNA synthesized by methyl methane-sulphonate-treated HEp-2 cells, Chem.-Biol. Interactions, 1 (1969/70) 89-98. 26 M. Fox and S. R. Ayad, Characteristics of repair synthesis in P388 cells treated with methyl methanesulphonate, Chem.-Biol. Interactions, 3 (1971) 193-211. 27 J. J. Roberts, J. M. Pascoe, B. A. Smith and A. R. Crathorn, Quantitative aspects of the repair of alkylated DNA in cultured mammalian cells, IL Non-semiconservative DNA synthesis (“repair synthesis”) in HeLa and Chinese hamster cells following treatment with alkylating agents, Chem.-Biul. Interactions, 3 (1971) 49-68. 28 C. D. Darlington, The plasmagene theory of the origin of cancer, Brif. J. Cancer, 2 (1948) 1% 126. 29 A. Graffi, 1st die Umache der malignen Entartung eine Mutation der Mitochondrien?, Zbl.

Gyniikol., 90 (1968) 945-958. 30 D. A. Clayton and J. Vinograd, Circular dimer and catenate forms of mitochondrial DNA in human leukaemic leukocytes, Nature, 216 (1967) 652-657. 31 M. M. K. Nass, Abnormal DNA patterns in animal mitochondria: Ethidium bromideinduced breakdown of closed circular DNA and conditions leading to oligomer accumulation, Proc. Nat/. Acud. Sci. (U.S.), 67 (1970) 19261933. 32 M. M. K. Nass, Reversible generation of circular dimer and higher multiple forms of mitochondrial DNA, Nature, 223 (1969) 1124-l 129. 33 F. Wang-Staal, J. Mendelsohn and M. Goulian, Ribonucleotides in closed circular mitochondrial DNA from HeLa cells, B&hem. Biophys. Res. Commun., 53 (1973) 140-148. 34 W. Wickner. D. Brutlag, R. Schekman and A. Komterg, RNA synthesis initiates in vitro conversion of Ml3 DNA to its replicative form, Proc. Nut!. Acad. Sci. (U.S.), 69 (1972) 965-%9. 35 P. Kleihues, C. Mende, and W. Reucher, Turnouts of the peripheral and central nervous system induced in BD-rats by prenatal application of methyl methanesulphonate, Europeun J. Cancer, 8 (1972) 641-645. 36 D. F. Heath, The decomposition and toxicity of dialkylnitrosamines in rats, B&hem. J., 85 (1962) 72-91. 37 R. Watson, W. Bauer and J. Vinograd, An optical system for the photography of fluorescent bands in preparative ultracentrifuge tubes, Anal. B&hem., 44 (1971) 200-206.