. . I I...“ ..... I .. I ..- o u can I III. . I I}. I h 0! II I v I. I‘d I I I w ..lea. I .In. \ II I .I . L . I I I .. . I.“ 35.5. . .-.“.I: II I. I I I I II I I... . I I I .I I . awn»! Jaw.) I I II. .Ih I... Hm. I .. I I I .o I II II I uco “Incl..PHNIIUJII3 [1an IIIII-I'IHIIIII I Id . I I I I I I I i IIIvatIIII . héfiuln‘flu III; ..I..Il. ..~ III I . I. I .L I I .I . .. . ..r ILI‘.“annIumI-dul 1” :3. In}. .I I . I .II ( II I k»: I I I IUIIVI OJIII vh?.|‘ II . . .nIlIInv II.‘ III I I I I ... u r. "I III a. I g .....I II. I- IIIJWIII HPIIIIIIIIIIIIJMQLW; I .I I I I I. :I II!- a II. 1&7”). ...IIRIIIIIIIIIIIIIII1NII3I IIIIIIWINHI? I I I I. I I 4 I.» II. It ..ItI (13. Isl I“ I I . I . . . .IUI..«qu.m.Im..IIIIH...KIr.urIII:TIiI VIIIH,.,I.IIIIII«Q.m..a.I.RIII IanIIIUII I . u I In.‘ IIIIRII flail." II I umIHhI...“ -..IIIIIIAI HUI :Ifl I I I .I I . . I31”? .. IIIIVIII LIJI I. II? -I I I .I. . II . . jfimfimg .. III”! ...In.... I, IMMIIIIJ I1 I III“ .1rv.!o IN.U\II.. II a L I3....IBrIIfiI 43.91%... I I I I I I ”...-I ‘ I a I .III I I'III,‘ .I . I I I\I\IIII o ‘3. II . II I I u ‘JHWIIXIIHHGMMIIIIIM'IIIIIIIIIINU II I HI . .I I .. I. I I INIHInI .III. IIIIIII: I IIII I . I I IIIII I I I I III). .5- a.IIfiIIwnH ”I III A IIIII I III I IIHIII 1‘1... I‘IeII'I.HI31¥IJJIu .‘vIIII I.r I In I I 1',” l: III“ 'lI I ”W. :I. IIIIIIIIIIII III I II“ l III I u - . I I I I ... I. IQI I I I IIIwnllflulluqulfluI “.III IMIIIIUI I- - III..\. B UIIIIH . III IIIIL‘ IIIIII IIIII . I.II noun”. IIIII.1[II..UHI\LII.III IIIIIIIII II...HIII:IIIIIIIIIIII|I (III . I! 8 I I. I I IIIIl'Ii Illl| I '1 VIII I‘:IIYHNIUII I . IIII I I I.IIIIIIIII III ...-I.) III I I II I, I I TI . .I IIIIII . IIIhSIIIIIIIIIdlIIIIIIIIIIIIIWIIIIIHIIINI IIIIIIIIIIII LIIIIIIRIIIIIII I .IIIIIIIIIIIIII I . . II I», III I III II 1|!“ )Itl II II III‘I . I 'Ilk I III III IIIII \ IIIIIIII. AIIWII Ir'.llI|l‘l|.Il.IIO‘-I Il‘llol'lIlYoI IIWII I .I I . I I II .I II III I I I l I l‘l‘l‘l‘ll E II- null-I'll II .IvIIII. II'XI IIIIII‘ICI I, I | I III I II I III ., I “II IIIIIIII -II IIIIIIIIIL IHIIIIIIUIIIIIIIIIII . II. uh. IIIIIII:I I III-II IIIIIIIVIIIIIIIIIIIIII ILLIMIIIII IIIIIIIILI I I I I I . . . , III III. III I. IIIII IIIIIIIu I.IIILwI..III III I I .IIII‘II l I 0.9 I9 M “HI” I IOI‘I III I?! I . I; I ILIL ! manna! Michigan State University This is to certify that the dissertation entitled . . Factors Involved in Target Spec1f1ci$y of the Hepatocarcinogen N-Z-Acetylamino- fluorene with Regard to Cell Type and Location within the Genome presented by Bonnie Lynn Baranyi has been accepted towards fulfillment of the requirements for PhoD. degreein PharmaCOIOgy ‘ (79’: Z a 1 V 5?: “5" " / L Major professor Date /// 7/321 MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 MSU LIBRARIES m RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. FACTORS INVOLVED IN TARGET SPECIFICITY OF THE HEPATOCARCINOGEN N-2-ACETYLAMINOFLUORENE WITH REGARD TO CELL TYPE AND LOCATION WITHIN THE GENOME BY Bonnie Lynn Baranyi A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Pharmacology and Toxicology 1982 ABSTRACT Factors Involved in Target Specificity of the Hepatocarcinogen N-2-Acetylaminofluorene with Regard to Cell Type and Location within the Genome by Bonnie Lynn Baranyi Carcinogenesis as a result of chemical exposure often exhibits organ-specific characteristics. The objective of this investigation has been to discern initial molecular events critical to target- specific carcinogenesis induced by the hepatocarcinogen. N-2-acetyl- aminofluorene (AAF). The following parameters were investigated: 1) the binding of AAF and its N-hydroxy metabolite (N-OH-AAF) to DNA of target cells, hepatic parenchymal (PC), and nontarget cells, hepatic nonparenchymal cells (NPC); 2) the location of carcinogen binding within the genome with regard to transcriptionally active as compared to inactive regions of chromatin in target and nontarget cells; 3) the effect of progressive carcinogen treatment on digestion of DNA from rat hepatic PC and NPC with base sequence-specific restriction endonucleases. AAF selectively binds to DNA of target PC due to a relative de- creased ability of the nontarget NPC to N—hydroxylate the procar- cinogen. Once N-hydroxylation has occurred, both cell types are able to carry out remaining steps in the metabolic activation and repair of adducts to a similar extent. Bonnie Lynn Baranyi Analysis of the binding of AAF to transcriptionally active vs. inactive regions of chromatin (as delineated by the nuclease DNase I) of target cells (PC) compared to nontarget cells (NPC) indicated that, at the time of peak binding, AAF selectively binds to DNA of trans- criptionally inactive regions of target cell (PC) DNA and to trans- criptionally active regions of nontarget cells (NPC). Pretreatment of rats daily with AAF for up to 5 days had no influence on the subse- quent binding of tracer doses of [ring-3H]-AAF to specific regions of DNA from PC and NPC. Treatment of rats with AAF (l5 mg/lOO g) for l, 3, 5 or 7 days did not alter the ability of the restriction enzyme, Eco Rl, to act at sites specific to this endonuclease, but resulted in inhibition of the restriction enzyme, Kpn I, to recognize sites for which this enzyme is specific immediately following treatment or 7 days following treat- ment. These results indicate that factors such as metabolic capabili- ties of target cells and structural characteristics of chromatin allowing for binding of carcinogen to discrete regions might be more important in determining carcinogenic susceptibility of a particular tissue than total binding and persistence of carcinogen. to Joseph, Mary and Jim ii ACKNOWLEDGEMENTS I sincerely thank Dr. Jay I. Goodman for his guidance and in- tellectual stimulation throughout the research toward this thesis project. I thank the members of my guidance committee, Drs. T.M. Brody, J.E. Trosko, H.J. Kung and R.A. Roth for their assistance during the course of this project. I thank Kimberly Thornton, Kathy Moloney and Sally Yoder for their excellent technical assistance, and Diane Hummel for her superior efforts in preparation of this thesis. TABLE OF CONTENTS Page DEDICATION ------------------------------------------------------- ii ACKNOWLEDGEMENTS ------------------------------------------------- iii LIST OF TABLES --------------------------------------------------- vi LIST OF FIGURES -------------------------------------------------- viii LIST OF ABBREVIATIONS -------------------------------------------- x INTRODUCTION ----------------------------------------------------- l I. Background --------------------------------------------- l 2. Organ Specificity of Chemical Carcinogens -------------- ll 3. Chromatin Structure and the Specificity of Carcinogens for Particular Genomic Locations ----------------------- l7 4. AAF-Induced Hepatocarcinogenesis ----------------------- 26 5. Interactions of AAF with DNA --------------------------- 3l 6. Experimental Objectives -------------------------------- 34 MATERIALS AND METHODS -------------------------------------------- 37 l. Animals: Maintenance and Carcinogen Treatment --------- 37 2. Isolation of Nuclei ------------------------------------ 38 3. Isolation of Distinct Liver Cell Populations ----------- 38 4. Nuclease Digestion of DNA from Nuclei ------------------ 39 5. Nick Translation of Transcriptionally Active DNA from Nuclei of Parenchymal and Non-Parenchymal Liver Cells-- 40 6. DNA Isolation ------------------------------------------ 41 A. Hydroxyapatite Column Chromatography -------------- 4l B. Isolation of DNA for Agarose Gel Electrophoresis-- 42 7. Isolation of Plasmid Containing Rat Albumin cDNA ------- 43 8. Agarose Gel Electrophoresis ---------------------------- 44 9. Southern Transfer of DNA ------------------------------- 46 10. 32F Nick Translation of cDNA Probe --------------------- 47 ll. Hybridization of Rat Liver DNA with 32P-labelled cDNA Probe -------------------------------------------------- 5] 12. Autoradiography ---------------------------------------- 52 l3. Other Methods ------------------------------------------ 52 iv TABLE OF CONTENTS (continued) Page RESULTS ---------------------------------------------------------- 54 l. Hydroxyapatite Column Chromatographic Isolation of DNA Eluted by Means of Centrifugal Force --------------- 54 2. Carcinogen Binding to Hepatic Macromolecules ----------- 63 3. Carcinogen Binding to DNA of Parenchymal and Non- Parenchymal Liver Cells -------------------------------- 66 4 Carcinogen Binding to DNA of Nuclei Isolated from Parenchymal and Non-Parenchymal Liver Cells ------------ 7l 5. Binding of Carcinogen to DNase I-Accessible and In- accessible Regions of Chromatin Derived from Parenchy- mal and Non-Parenchymal Cell Nuclei -------------------- 79 6. Nick Translation of Transcriptionally Active DNA in Intact PC and NPC Nuclei ------------------------------- 94 7. Effect of Continued AAF Treatment on Restriction Endo- nuclease Digestion of DNA ------------------------------ lOO DISCUSSION ------------------------------------------------------- l34 SUMMARY AND CONCLUSIONS ------------------------------------------ l56 BIBLIOGRAPHY ----------------------------------------------------- l6l 10 ll 12 13 LIST OF TABLES Page Recovery of DNA following hydroxyapatite column chromatography ----------------------------------------- 55 Assessment of RNA and protein contamination of DNA eluate from hydroxyapatite columns --------------------- 56 Recovery of radioactivity applied to the hydroxyapatite column ------------------------------------------------- 52 Binding of N-hydroxy-N-acetylaminofluorene (N-OH-AAF) to hepatic macromolecules ------------------------------ 64 Binding of carcinogen to acid-precipitable hepatic tissue and hepatic DNA --------------------------------- 65 Total binding of AAF and N-OH-AAF to liver cell DNA—--- 72 Binding of 3H-N-hydroxyacetylaminofluorene to DNA of rat liver ---------------------------------------------- 78 DNA digested from parenchymal cell nuclei by DNase I--- 82 DNA digested from nonparenchymal cell nuclei by DNase I 84 N-2-Acetylaminofluorene adducts in specific regions of chromatin DNA of different hepatic nuclei populations-- 90 N-Hydroxy-acetylaminofluorene adducts in specific re- gions of chromatin DNA of different hepatic nuclei populations -------------------------------------------- 92 Analysis of carcinogen binding to hepatic macromole- cules following pretreatment with AAF or vehicle- control ------------------------------------------------ 93 Analysis of carcinogen binding to DNase I accessible DNA of rat parenchymal and nonparenchymal liver cell nuclei following pretreatment with AAF or vehicle- control ------------------------------------------------ 95 vi LIST OF TABLES (continued) Table 14 15 16 17 18 19 20 21 22 23 Effect of daily N-2-acetylaminofluorene treatment on 1iver/body weight ratios ............................... Eco RI restriction endonuclease fragmentation of DNA from rats treated 1 day with N-2-acetylaminofluorene--- Eco RI and Kpn l restriction endonuclease fragmentation of DNA from rats treated 3 days with N-2-acetylamino- fluorene ----------------------------------------------- Eco R1 and Kpn l restriction endonuclease fragmentation of DNA from rats treated 5 days with N-2-acetylamino- fluorene ----------------------------------------------- Eco RI and Kpn l restriction endonuclease fragmentation of DNA from rats treated 7 days with N-2-acetylamino- fluorene ----------------------------------------------- Effect of daily N-2-acetylaminofluorene treatment followed by 7 days without treatment on liver/body weight ratios ------------------------------------------ Eco R1 and Kpn l restriction endonuclease fragmentation of DNA from rats treated l day with N-2-acetylamino- fluorene followed by 7 days without treatment ---------- Eco Rl and Kpn l restriction endonuclease fragmentation of DNA from rats treated 3 days with N-2-acetylamino- fluorene followed by 7 days without treatment ---------- Eco Rl and Kpn l restriction endonuclease fragmentation of DNA from rats treated 5 days with N-2-acetylamino- fluorene followed by 7 days without treatment ---------- Eco Rl and Kpn l endonuclease fragmentation of DNA from rats treated 7 days with N-Z—acetylaminofluorene or vehicle followed by 7 days without treatment ----------- vii Page 106 110 113 116 120 121 130 131 132 133 53‘s. E“! “‘1 Figure 10 11 12 LIST OF FIGURES Isolation of 32P-labelled nick-translated cDNA probe from pRSA-l3 cDNA -------------------------------------- Fractionation of chromatin by DNase II digestion and selective MgCl2 precipitation -------------------------- Binding of carcino en to DNA of rat liver following administration of Ering-3H]-N-hydroxyacetylamino- fluorene ----------------------------------------------- Binding of [ring-3H]-acetylaminofluorene to DNA of hepatic parenchymal (PC) and nonparenchymal (NPC) cells Binding of [ring-3H]-N-hydroxyacetylaminofluorene to DNA of hepatic parenchymal (PC) and nonparenchymal (NPC) cells -------------------------------------------- Binding of [ring-3HJ-acetylaminofluorene to DNA of hepatic parenchymal (NI) and nonparenchymal cell (NII) nuclei ------------------------------------------------- Binding of [ring-3H]-N-hydroxyacetylaminofluorene to DNA of hepatic parenchymal (N1) and nonparenchymal (NII) nuclei ------------------------------------------- Digestion of rat liver cell nuclei with pancreatic deoxyribonuclease I .................................... Binding of [ring-BHJ—acetylaminofluorene and [ring-3H]- N-hydroxyacetylaminofluorene to total DNA and DNase I- digested DNA from parenchymal cell nuclei -------------- Binding of [ring-3H]-acetylaminofluorene and [ring-3H]- N-hydroxyacetylaminofluorene to total DNA and DNase I- digested DNA from nonparenchymal cell nuclei ----------- Incorporation of 32P-labelled deoxyribonucleotide tri- phosphates into DNase I-accessible regions of DNA from hepatic parenchymal cell and nonparenchymal cell nuclei Incorporation of 32P-labelled deoxyribonucleotide tri- phosphates into DNase I-accessible regions of DNA from hepatic parenchymal cell and nonparenchymal cell nuclei viii Page 49 58 6O 67 69 73 76 8O 85 88 96 98 I_IST OF FIGURES (continued) Figure 13 14 15 16 17 18 19 20 21 22 Page Digestion of liver DNA of untreated rats by Eco RI ----- lOl Digestion of liver DNA of untreated rats by Kpn l ------ 104 Eco R1 digestion of liver DNA of rats treated l day with N-Z-acetylaminofluorene (AAF) or with vehicle ----- 108 Eco R1 and Kpn 1 digestion of liver DNA from rats treated 3 days with N-2-acetylaminofluorene (AAF) or with vehicle ------------------------------------------- lll Eco R1 and Kpn l digestion of liver DNA from rats treated 5 days with N—2-acetylaminofluorene (AAF) or with vehicle ------------------------------------------- 114 Eco Rl digestion of liver DNA from rats treated 7 days with N-2-acetylaminof1uorene (AAF) or with vehicle ----- ll8 Eco R1 and Kpn l digestion of liver DNA from rats treated 1 day with N-2-acetylaminofluorene (AAF) followed by 7 days without treatment ------------------- 122 Eco R1 and Kpn l digestion of liver DNA from rats treated 3 days with N-Z-acetylaminofluorene (AAF) followed by 7 days without treatment ------------------- 124 Eco R1 and Kpn l digestion of liver DNA from rats treated 5 days with N-2-acetylaminofluorene (AAF) followed by 7 days without treatment ------------------- 126 Eco R1 and Kpn l digestion of liver DNA from rats treated 7 days with N-2-acetylaminofluorene (AAF) or vehicle followed by 7 days without treatment ----------- 128 ix .AAF N-OH-AAF C8-gua-AAF C8-gua-AF NZ-gua-AAF DMN PC NPC NI NII ABBREVIATIONS N-2-acetylaminofluorene N-hydroxy-N-Z-acetylaminofluorene N-(deoxyguanosin-B-yl)-2-acetylaminofluorene N-(deoxyguanosin-B-yl)-2-aminof1uorene N-(deoxyguanosin-NZ-yl)-N-2-acetylaminofluorene dimethylnitrosamine parenchymal cells nonparenchymal cells nuclei of parenchymal cells nuclei of nonparenchymal cells '4 ‘— i ‘1 o ..,..'- . ¢ \ auo 1 .. .~ 0 u 4.. I r- «‘r u ”a. v . . A... A 4 . ‘- 0‘. ”C IV- " v- . 'D ' . ~. . «4. .. ‘ u I ... fir - ' u a‘ . .F‘ v n u. u '. u. -\ ‘_‘ . I“ I. H W. ." . u INTRODUCTION 1. Background Modern epidemiological studies indicate that 80-90% of all human cancers are caused by environmental factors (11,50). Chemicals seem to be the most probable causative factors, in view of the facts that oncogenic viruses are not highly contagious (50) and radiation is fairly uniformly distributed (50,86). Epidemiologists have found that the total percent of population mortality due to cancer varies little from country to country (86). However, mortality ascribed to cancers of specific tissues types varies a great deal between countries (86). As an example, cancers considered to be rare in the United States such as primary carcinoma of the liver, Burkitt's lymphoma or Kaposi's sarcoma are common in Africa (61). U.S. nonwhites resemble U.S. whites more than they do African groups of similar heredity in cancer susceptibility (61). With westernization of lifestyle in Africa, cancer patterns have become more like those of the Western world. Many cancers have been found to be higher among recent immigrants to the U.S. as opposed to native U.S. whites (61). Migrants from Poland, Czechoslovakia, Norway and the U.S.S.R. exhibit higher inci- dences of stomach cancer than U.S. whites and these incidences corre- late with those of their homeland (61). The lack of changes in cancer incidence following emigration may be due to a continuance of dietary 2 customs and habits once in the United States (61). Incidence of breast cancer was ranked highest among U.S. white females, yet quite reduced among women of Italy and Poland (61). However, data on Polish rnigrants indicate a rise in breast cancer incidence when women remain 1'1 the U.S. compared to the incidence observed in urban and rural Phaland. An extensive review by Haenszel and Kurihara on mortality 'from cancer among Japanese migrants to the U.S. revealed that U.S. Japanese experienced higher cancer mortality than native Japanese for cancer within the intestines, gall bladder, pancreas, lung, ovary, prostate, nervous system and breast (47). It must be kept in mind that although geographic variations in cancer incidence may be attributable to environment, environment can include lifestyle influences such as dietary, social and cultural habits (51). As noted by Higginson, "Clinical cancer is the end result of numerous influences at the cellular level, many of which cannot be investigated easily in epidemiologic studies..." (51). Carcinogenesis due to chemical exposure was first noted in humans over two hundred years ago when Percival Pott, a general surgeon in England, noted a high incidence of cancer of the scrotum in chimney sweeps who began their careers in early childhood (101). In question- ing the etiology of such cancers, he recognized that the problem was related to many years exposure to coal soot and tars, along with poor hygiene habits (101). Several years before Pott's observation, Dr. John Hill, a physician in London, published an article entitled, "Cautions Against the Immoderate Use of Snuff" (108), in which he reported cases of nasal polyps and lesions resembling cancer in 3 persons using tobacco snuff frequently and for prolonged periods (108). This report in 1761 was probably the first implication of tobacco as a cause of cancers. In the late part of the 19th century, Rehn observed the develop- ment of urinary bladder cancer among several workers following pro- longed exposure to aromatic amines in a German aniline dye factory (see reviews in Ref. 86 and 85). More recently, many associations have been made with cancer and occupational, medical and societal exposures to chemicals (29,50,84). The International Agency for Research on Cancer (IARC) has reviewed evidence on nearly 500 chemi- cals, and their association with cancer (52). Their objective was to ehufidate carcinogenic risk of chemicals to humans. Due to limited human evidence and uncertainties concerning extrapolations of animal data to man, only 18 chemicals, groups of chemicals and industrial processes fell into Group 1, i.e., substances that are carcinogenic for humans (52). Attempts to set up experimental animal models to study in depth factors associated with development of cancer following exposure to various chemicals dates back to the early 20th century. Yamagiwa and IChikawa are credited with the first successful production of cancer In experimental animals by producing malignant epithelial tumors in Ears of rabbits following coal tar applications (46). These types of Slnidies were carried a step further by Rous and Kidd in 1941 (109). These investigators set up a series of experiments for the purpose of Sluidying the relationship between benign growths which arise following aP131ications of tar to ears of rabbits and cancerous lesions. They 4 noted that the definition of cancer at that time did not allow for neoplasms which depended upon permissive conditions for existence and growth, nor the fact that some stage or stages of neoplastic growth may be reversible (109). A first set of experiments was designed to determine if tar-induced tumors which regress upon cessation of tar treatment reappear from the same cells when tar treatment is resumed. Rous and Kidd found that not only do tumors reappear in the same areas, but following a shorter latency period, along with appearance (H numerous new tumors (109). These studies suggested the condition of preneoplastic cells produced by the first tar treatment. In a second set of experiments, Rous and Kidd found that following cessa- tion of a round of treatment with tar, application of a noncarcino- mufic agent (turpentine) which induced a superficial inflammation and cell proliferation resulted in reappearance of original tumors and appearance of new ones (109). In addition, the process of wound healing itself caused neoplastic developments at the boundaries of scars following previous tar treatment (109). Lastly, when no further treatments were given following one period of tar application, the epithelium of rabbit ears healed, and any "carcinomatoids" regressed (109). Mottram (89) proposed several factors which appeared to be re- (Wired for production of abnormal cells following carcinogen exposure: '1) a "sensitizing factor" to produce an initial change from normality 1'1 a cell, 2) a specific cellular reaction to "fix" the consequent Changes following exposure to a "sensitizing factor", and 3) a de- VEloping factor to produce an abnormal population of cells through growth and multiplication. This investigator was primarily concerned 5 with the role of hyperplastic agents as developing factors. Experi- ments were designed in which one flank of a mouse was treated with benzpyrene followed by croton oil treatment (20 weeks) and the other flank was painted with benzpyrene followed by vehicle treatment. A predominance of malignant tumors developed on the flanks of mice treated with benzpyrene and croton oil (89). These results led Mottram to conclude that, ”The part played by hyperplasia and chronic irritation in the genesis of cancer is readily explained if prolonged stimulation of cell division is a necessary pre-requisite before cells will become neoplastic" (89). The existence of latency periods for carcinogenesis and the dormancy of altered cell populations following treatment with carcinogens were raised by these investigations. Berenblum and Shubik (12) reported a series of experiments de- signed to determine the importance of time sequence for application of carcinogen and hyperplastic agent, and to determine whether a single application of carcinogen was adequate to result in neoplastic changes in cells following treatment with croton oil. There was no difference in the total numbers of tumors which developed or the latent period for tumor induction when rats were pretreated with croton oil, then given benzpyrene followed by 20 weeks of croton oil treatment or when rats were treated similarly, but without croton oil pretreatment (12). These studies confirmed earlier results by Mottram (89) that croton ‘311 treatment following a single application of carcinogen can induce tumors (12) . From these early investigations, the notion of stages within the Cartfinogenic process evolved. Berenblum (13) explained his results in 6 terms first proposed by Friedewald and Rous (13) as an "initiation process“ in which normal cells were converted into latent tumor cells, and a promoting process in which latent tumor cells were caused to develop into frank tumors. Fundamental characteristics of the two stages of carcinogenesis were first coalesced into a unified theory by Berenblum (13). The initiation phase was considered to be a process specific to the chemical applied, as well as being irreversible and rapid. The promotion phase required multiple, repeated applications (12,13), was not specific to a chemical (12,13,109) and was reversible (109). These conclusions remain as the basis for the understanding of chemical carcinogenesis today. As knowledge has increased, further divisions within the two stages has occurred (122), as well as an attempt to understand the process of progression (97). Present day definitions of initiation, promotion and progression have taken into account the advances in field of molecular biology over the last 30 years. An initiating agent is defined as an agent (chemical, physical, or biological) which is capable of directly and irreversibly altering the native molecular structure of DNA (97). Alterations may or may not involve covalent binding of the agent to DNA, or may involve distortion of DNA structure, scission of the DNA Chain, elimination of a base or sugar, or errors in DNA repair (97). ‘A promoting agent is described as an agent which alters expression of genetic information of a cell (97). This event is generally consi~ dered to be epigenetic, brought about by such agents as hormones, IllYperplastic agents, drugs, etc. (97). Finally, progression is de- 'Flned as that stage of neoplastic development characterized by visible 6 terms first proposed by Friedewald and Rous (13) as an "initiation process" in which normal cells were converted into latent tumor cells, and a promoting process in which latent tumor cells were caused to develop into frank tumors. Fundamental characteristics of the two stages of carcinogenesis were first coalesced into a unified theory by Berenblum (13). The initiation phase was considered to be a process specific to the chemical applied, as well as being irreversible and rapid. The promotion phase required multiple, repeated applications (12,13), was not specific to a chemical (12,13,109) and was reversible (109). These conclusions remain as the basis for the understanding of chemical carcinogenesis today. As knowledge has increased, further divisions within the two stages has occurred (122), as well as an attempt to understand the process of progression (97). Present day definitions of initiation, promotion and progression have taken into account the advances in field of molecular biology over the last 30 years. An initiating agent is defined as an agent (chemical, physical, or biological) which is capable of directly and irreversibly altering the native molecular structure of DNA (97). Alterations may or may not involve covalent binding of the agent to DNA, or may involve distortion of DNA structure, scission of the DNA chain, elimination of a base or sugar, or errors in DNA repair (97). A promoting agent is described as an agent which alters expression of genetic information of a cell (97). This event is generally consi- dered to be epigenetic, brought about by such agents as hormones, hyperplastic agents, drugs, etc. (97). Finally, progression is de— fined as that stage of neoplastic development characterized by visible 7 karyotypic alterations within tumor cells (97). These karyotypic changes correlate with increased tumor growth rate, invasiveness, metastases and neoplastic biochemical and morphologic alterations (97). Additional investigations (123) have revealed that in some cases, a single exposure to carcinogen is adequate for initiation, and that the effects of initiators can be additive (97). Furthermore, a round of cell proliferation seems to be required to confer irreversi- bility on the initiation lesions (21,34,59). Proliferation may ”fix" some change in that it is now permanent. Thus, initiation may occur in at least 2 steps: 1) interaction of the carcinogen as an activated derivative with DNA, 2) fixation of the lesion by one round of cell replication (21). It has been shown that DNA containing MNU-, DMN- and N-OH-AAF-produced lesions can replicate in_vjyg_(see ref. 21). Promotion may occur as a result of an agent which provides a selective environment in which initiated cells can be expressed either by inhibited growth of surrounding cells but not of initiated cells, or by preferentially stimulating growth of initiated cells (34). In addition, there is evidence that promoters exhibit a threshold and maximum effect (97). These characteristics may allow for opportu- nities to control human exposure, and thereby reduce levels of and exposure to these substances rather than completely eliminating them from the environment. It is widely accepted that tumors evolve as clones of a single altered cell (110). In many cases the malignant cells in a primary tumor mass exhibit the same abnormal karyotype (llO). Tumors might arise from one of many neoplastic cells that had a selective growth advantage over normal cells. Some neoplastic cells may be eliminated 8 as a result of metabolic disadvantage or immunologic destruction (110), and the mutant which has more selective advantage may ultimately give rise to a new subpopulation of tumor cells. The biology of neoplastic cells remains uncertain. Initiation may involve altered gene ex- pression rather than structural mutation as indicated by the absence of unique gene products in tumor cells and the reversibility of transformation in some cell culture systems (110). Neoplastic traits generally reflect alterations in pre-existing genes which may result from mutations in regulatory genes or effects on gene dosage following chromosomal rearrangement. At present, evidence points to transcrip- tional control as the most frequent method for eukaryotic gene control (26). However, recent studies in mouse cells have indicated that treatment of cells with methotrexate induces a 350-fold increase in the gene which codes for dihydrofolate reductase (DHFR), the target enzyme for methotrexate (7). In addition, treatment of methotrexate- exposed cells with the tumor promoter, l2-0-tetradecanoyl-phorbol-l3- acetate increases the gene copy number for DHFR 16-fold (134). Thus, changes in gene dosage following drug or chemical exposure may be an important factor in neoplastic development. The concept that cancer is the result of some change in the genetic material of cells, the somatic mutation theory, has been a central hypothesis in fundamental cancer research (92). Early re- search suggested that chromosomal mutation may be related to car- cinogenesis. More recently, studies have been performed dealing with alterations of DNA, RNA and proteins by carcinogens, and how these events may be justified within the theory of somatic mutation. Fahmy 9 and Fahmy (30) found the mutations arising from direct intramolecular DNA damage, such as point mutations and chromosomal breaks may only be a small contribution to the types of DNA damage that occurs due to chemical carcinogens. In a series of investigations on a variety of chemical carcinogens, it was found that alkylating agents and nitroso agents often directly damage DNA resulting in point mutations and chromosomal breaks (30). However, hydrocarbons and aromatic amines 1° nduced small deletions of a chromosome in Drosophila, but not by point mutations or chromosomal breaks (30). The investigators suggest that these carcinogens may have induced the chromosomal deletions via ‘i Interference with enzymes or regulators involved in DNA synthesis, replication or repair (30). They suggest that these results are in 1 ‘i he with the somatic mutation theory, which can be expanded to in- C1 ude events which involved indirect mechanisms for induction of mUtations, rather than exclusively direct DNA attack (30). Cairns presented evidence that somatic mutation may not be an i deally inclusive explanation for cancer (20). He explained that Seubjects exhibiting deficiency in DNA repair capabilities should e>031 ain some aspects of the development of cancer. However, in view of rapid advances being made in the field of molecular biology in terms of the nature of genetic transposition (20) and alteration of gene expression (32), alternative explanations must be considered to eXplain the diverse nature of mechanisms for chemically-induced Carcinogenesis. In view of the multiple explanations for carcino- genesis, it is probable that there are multiple causes. It is IDOssible to view the development of cancer as the result of an inter- aQtion of factors such as faulty differentiation as a consequence of § ‘ltered DNA structure due to chemical damage or by viral modifications Q ‘1: the genome equivalent to mutations (102). Thus, 3 major theories Q ‘1 carcinogenesis (somatic mutations, viral modifications, faulty q fifferentiation) may be integrated to explain the end result of Q encer, rather than each theory being mutually exclusive. ll 2. Organ Specificity of Chemical Carcinogens It has become evident from the extensive research that has been done in the field of chemical carcinogenesis that many carcinogens exhibit specificity for one or several organs. Epidemiologic evi- dence, discussed above, for environmental factors being induced at 1 east in part for cancers in various regions of the world indicates that cancers of particular tissues are predominant in certain parts of the world (47,51,61). Thus, regional predominance of certain types of cancers may be the result of human exposure to different industrial chemicals or to various substances in the diet or environment of people in these regions. From the time carcinogen exposure occurs (either via the environ- ment or through administration to an experimental animal) to the development of malignant tumors, a number of events may occur within Various tissues. Organ specificity of chemical carcinogens may be i '17:] uenced by the distribution of the carcinogen within the body, the r‘eTative activities of toxifying and detoxifying enzymes in various organs, the presence of cellular components to which an ultimate Care-i nogen may bind before it reaches the critical intracellular ta r‘Qet (presumed to be DNA), and the repair capacity of various ti sSues (79). In some cases, the route of administration may in- F1 uence the site of tumor development due to factors sited above. when given in the diet, N-OH-AAF produced tumors of the forestomach (88) . When injected intraperitoneally, N-OH-AAF produced peritoneal 8a‘F‘Comas (88). This is probably due to N-OH-AAF being one metabolic s hep closer than the parent compound, AAF, to the ultimate carcino- Sen ‘1 c form. 12 Dimethylnitrosamine (DMN) causes primarily liver, kidney and some lung tumors (79). Since the nitrosamines appear to distribute uni- formly throughout the body water following oral administration to rats (79), it is unlikely that distribution of the carcinogen 2.9139. is involved to any great extent in organotropism of these carcinogens. Similarly, N-methylnitrosourea (MNU) can induce tumors in a variety of tissues depending upon conditions of administration, yet this car- cinogen rapidly distributes throughout total body water following administration (79). However, streptozotocin, a glucose derivative of MNU, has been shown to selectively accumulate in pancreatic islet cel ls of mice, an area in which tumors develop following administra- tion of single, large doses (79). Metabolic activation and detoxification appear to contribute extensively to organotropism of carcinogens. The observation that ma ny carcinogens of diverse chemical structure caused similar types of tMinors led investigators to investigate binding of carcinogens to ce] 1 ular components, and metabolic activation of carcinogens (85). DMN i s more readily activated by hamster lung slices than by rat lung pr‘epaarations, and these results correlated with the greater suscepti- bi 1 ‘i ty of hamster respiratory tract to carcinogenesis from DMN (79). AAF: the parent compound from which N-OH-AAF is derived, does not Cause tumors of the forestomach upon oral administration, nor does th‘i S precarcinogen cause peritoneal sarcomas upon i.p. injection (88). G” i "ea pigs are resistant to AAF-induced carcinogenesis, and there is evi dence that this may be due to a low rate of N-oxidation of AAF in 1: he guinea pig relative to other species (69,87). When N-OH-AAF is 13 administered to guinea pigs, tumors develop at the site of admini- stration (87). Mouse and hamster are susceptible to AAF-induced hepatocarcinogenesis, and both species were found to N-hydroxylate AAF 111. 11:19 (87). Metabolic detoxification may be important in reducing the suscep- tibility of a tissue to carcinogenicity by a chemical. It is well ficticnun that the liver is capable of metabolically activating carcino- ggeerus to reactive forms (69.85-87.103). These reactive metabolites may then undergo conjugation reactions, such as N-glucuronidation, to render them more stable and water-soluble, and thus, excretable (LESEB,69,103). However, these water-soluble metabolites may be trans- F><3rvted through the kidney to the bladder, wherein the acidic environ- rrzeerrt of the urine (pH of 4-6 in dog and human urine) may cause acid h.Ydrolysis of the glucuronide conjugate (58,103). Such is the case with some glucuronide conjugates of arylamine N-hydroxylation pro- duets (58,103). These hydrolysis products can be protonated at the N- hJV'dY‘oxy position, forming a highly reactive arylnitrenium ion capable of binding to nucleophilic macromolecules of the urinary bladder epi thelium ((58,103). Thus, a precarcinogen which undergoes extensive metEibolism at one site is carcinogenic predominantly at a distant ss-i 1:62,, Induction of different metabolic pathways in the liver may reduce or ‘i ntensify the carcinogenicity of some chemical carcinogens. When r773“13$3 were fed 3-methylcholanthrene in the diet (0.003% w/w), the ‘i r‘F-T-‘idence of tumors following AAF administration decreased by two- th‘i Y‘ds (88). A predominance of l-, 3-, 5- and 7-ring-hydroxymetabo- 1 tes occurred, while very little of the N-hydroxy metabolite was (U ... O. .- ‘ Uzi. \- u: ‘n 14 produced (88). The ring hydroxy metabolites of AAF have little to no carcinogenic potential (88). weeks of diethylnitrosamine (DEN) administration to rats via drinking When phenobarbital was fed during 10 water, a decrease in hepatic tumor yield was observed compared to 10 weeks of DEN treatment without concomitant phenobarbital administra- Induction of detoxifying metabolic pathways may have been ti on (143). However, when phenobarbi- responsible for decreased tumor incidence. ta'l treatment was begun one week following cessation of DEN treatment, an increased tumor incidence was observed (143). In this case, phenobarbital was most probably acting as a promoter, resulting in expression of damage caused by DEN pretreatment. More recently, it has been shown that pretreatment of rats with polybrominated biphenyls ( PBBs) followed by AAF administration decreased the incidence of mammary tumors in female rats (117). Subsets of cells within a target organ may be differentially S”Sceptible to carcinogensis by various compounds. Several factors may influence the susceptibility of subsets of liver cells to carci- nogens. Cells closest to the portal triad (zone 3 of the hepatic acTV‘IUS model) are higher in microsomal enzyme activity (this may, in part , be due to differences in oxygen tension across the acinus) (129) Substrates for detoxifying metabolic reactions such as glutathione 0nd ugation may not be evenly distributed across the acinus (129). us(Zeptibility of protein synthesis to toxic damage may show regional i netJualities. There can be a graded clearance of toxic substances ‘F "7°") hepatic cells within an acinus to the blood (129). Treatment of h fits with the hepatocarcinogen N-nitrosomorpholine results in 15 development of large hepatocyte subpopulation with enlarged nuclei (139). The process of centrifugal elutriation, i.e., isolation of populations of cells based on cell size and density, allows for iso- lation of subpopulations of hepatocytes of varying size and ploidy. By employing this procedure on suspensions of liver cells from rats treated with N-nitrosomorpholine, higher proportions of hypertrophied and polyploid cells were found in fractions of elutriated hepatocytes consisting of the large cell subpopulations following N-nitrosomor- pholine treatment (139). Additionally, when 5 simple aliphatic nitrosamines were fed to Fischer F344 rats, varying patterns of carcinogenesis were observed (76). Using centrifugal elutriation to separate populations of hepatic parenchymal and nonparenchymal cells, Lewis and Swenberg found damage to nonparenchymal target cells as a r‘esult of treatments of rats with the carcinogen 1,2-dimethylhydrazine to Persist, while damage was repaired in parenchymal nontarget cells ( 74) . Nitrosodimethylamine gave rise to hemangiosarcomas of the 1 ‘iver, while nitrosodiethylamine gave rise to hepatocellular car- (2 Thomas at an approximately equivalent dose (76). Nitrosomethyl- ethyl amine induced both hemangiosarcomas and hepatocellular carcino- mas 3 all 5 nitroso compounds caused esophageal tumors (76). Nitroso- di ‘h—propylamine induced tumors of the esophagus and forestomach, but not of the liver (76). Following treatment of rats with AAF in the diet for long periods or 1”or short periods followed by treatment with phenobarbital, car- Qi hOrnas develop primarily from hepatocytes as opposed to other liver c e1 1 populations (1,33,38,119,105). When rats are maintained on a O . . . ‘ 02% (w/w) AAF-containing diet, a dosage which causes minimal oval 16 cell proliferation, altered foci become apparent at 3 weeks of treat- ment (145). Ultrastructural studies have revealed that cell orga- nelles within cells of altered foci are characteristic of hepatocytes (145). These changes include abundant endoplasmic reticulum, glycogen and microbodies, prominent nucleoli, increased mitochondria and de- creased parallel-arrayed rough endoplasmic reticulum (145). When rats are maintained on a 0.05% (w/w) AAF diet, different types of changes occurred in organelles of hepatocytes (38), including permanent There is a permanent decrease in the amount of rough endo- c hanges. p1asmic reticulum (38). Nucleoli and Golgi apparatus are hypertro- phi ed, and mitochondria are decreased in size (38). There is evidence to suggest that hepatocellular carcinomas may arise from cells of a 1 tered foci (1,28,98,105). Convincing evidence reported by Rabes gt 2.1. (,105) demonstrated that some cells within enzyme-altered foci are h ‘istologically and biochemically compatible with early carcinomas. In view of the early, rapid proliferation of "oval" cells (33, I 1 9.1 20) seen when rats begin an AAF-containing diet, malignant tumors i n the liver are derived almost exclusively from hepatic parenchymal Ce} 1 s (146). Though hepatocytes constitute 90% of the liver weight, they represent only 65% of the total number of cells present. Blood- borne environmental carcinogens must traverse the Kupffer cell barrier 01: the sinusoids to reach the liver parenchyma, yet the target of many carcinogens is the hepatocyte (33). Therefore, there is a need for e 1 “Cidation of events which result in carcinogenesis targeted to h epatocytes after AAF exposure. l7 Covalent interaction of carcinogens with DNA is thought to be a critical step in the initiation of chemically-induced carcinogenesis. lilkylation of the N-7 position of guanine by nitrosamines, while producing the major alkylation product, does not correlate with organ specificity of nitrosamines (79). The formation and persistance of the 06-methylguanine adduct may play a more important role in car- cinogenesis (10). Lewis and Swenberg found that following admini- stration of 1,2-dimethylhydrazine which induces primarily malignant hemangioendotheliomas, initial alkylation of DNA bases was higher in hepatic parenchymal (nontarget) cells (74). However, removal of 06- methylguanine was significantly slower from the nonparenchymal (tar- get) cell DNA, resulting in an accumulation of this adduct in the target cell DNA (74). These results suggest that selective repair in Various cell populations at a target organ may be an important factor ‘5 n carcinogenesis. However, Beland gt a_l_. have recently reported that F0] lowing biweekly treatment of rats with N-OH-AAF, certain DNA adducts (N-deoxyguanosin-B-yl-2-aminof1uorene) persisted and accumu- ] ated in both target (liver) and nontarget (kidney) tissue DNA (10). TheY‘efore, persistance of DNA adducts alone may not be sufficient for C - O a "C 1 nogeneSi s . :3 ‘ Chromatin Structure and the Specificity of Carcinogens for Particular Genomic [Stations Carcinogen attack may be specific with regard to target site in 65’s other than speCifiCity at the organ and cellular level. In view 1: the importance of carc1nogen modification of DNA, it 15 worthwhile 1: . . . . 0 Consider organization of DNA in the mammalian cell and how 18 carcinogens may gain access to DNA. In eucaryotic cells, chromosomes at metaphase, which are visible under a light microscope, are com- plexes of DNA packaged tightly with histone and nonhistone proteins (71). At stages of the cell cycle other than metaphase, DNA and associated proteins are present as the more diffuse chromatin (36, 63,71). At present, it is recognized that DNA of eucaryotic chromo- somes is arranged in at least 3 levels of organization (71). Firstly, the fundamental unit of a chromatin fiber is a nucleosome (36,63,71). Nucleosomes assemble in a chain to form the chromatin fiber (36,63, 7 'l ). Folding and/or coiling of the chromatin fiber into a compact c hromosome is the third level of organization (71). Recent studies of chromatin structure have revealed that the nucleosome consists of a specific length of DNA wrapped around an octamer of histone proteins termed the core particle (36,63,71). When chromatin fibers are unfolded in a low ionic strength environ- ment, nucleosomes are arranged along DNA as "beads on a string" (36, 63 ,71). Each nucleosome particle is approximately 100 A in diameter ( 36 .71). The nucleosome consists of a core particle and spacer DNA ( 35 .63,7l). The core particle contains an octamer of histones com- poSed of 2 each of the lysine-rich histones H2A and H23, and two each 01: the arginine-rich histones H3 and H4 (36,71). Recent research has Shown that the arginine-rich histones (H3 and H4) are essential for F01 ding of DNA in a nucleosomal manner (36). The length of DNA asso- ciated with the core particle is 140 base pairs (bp), and is invariant a"‘th species (36,71). The linker (or spacer) region of DNA can vary betWeen 10 to 70 bp in length, and is dependent on the species (36,71). 19 X-ray crystallographic data have revealed that the nucleosome core particle is probably a flat disk, approximately 100 A in diameter and 50 A in height, with DNA wrapped around the outside of the disk (36, 71). Dissociation of the nucleosome in 2 M NaCl has yielded 2 tetra- mers consisting of 1 molecule of each histone, suggesting that nucleo- somal core particles possess a 2-fold axis of symmetry (71). There is evidence that a fifth histone protein, H1, is associated with DNA at the end of the nucleosome core, or between nucleosomal cores, and may form a link between 2 "beads" in a chromatin fiber (63). Removal of H 1 has little effect on nucleosomal cores, but does affect the con- Formation of the chain of "beads" (63). There is controversy over the arrangement of nucleosomal core particles within the chromatin fiber. Some evidence suggests a solenoid-type arrangement with a pitch of 100 A and a diameter of 300 A (71). Others suggest the nucleosomes arrange as clusters (71). Hi stone Hl appears to stabilize the arrangement possibly by cross- 1 ‘ining nonadjacent nucleosomes (36,71). The amino acid composition of the H1 histone subspecies has been found to vary a great deal more than that of the other histones (36). This variable histone is asso- ciated with the variable section of DNA within a nucleosome (36). It i s conceivable that variability in these regions is related to SpeCies-specificity. Laemmli (71) has found that through treatment of metaphase chromosomes with competing polyanions (dextran sulfate), the chromosomal DNA held together by so-called scaffolding nonhistone pro“tr-fins, can be isolated free of histone proteins. Their results have shown that a set of nonhistone proteins form a central scaffold- ‘nQ crosslinking DNA into a radial distribution of regular loops 20 (10-30 um) (71). In this model, histones would serve to compact the DNA loops (71). The result is compaction of genetic information into the structure of a chromosome for distribution to daughter cells during mitosis. With an increased understanding of the organization of genetic i nformation in a eucaryotic cell, investigations on biological acti- v 'i ty of chromatin could be undertaken. Generally, 10-20% of the entire genome is considered to be transcriptionally active chromatin ( 1 17,141). The exact mechanism for control of gene expression remains to be elucidated. It has been noted, however, that transcriptionally active regions of the genome are organized into nucleosomes, as are the transcriptionally inactive regions (141). Yet there is some a ‘I teration of this basic structure that renders some regions of c: hromatin transcriptionally active. Weisbrod and Weintraub have reported that there appear to be a group of proteins associated with the globin gene of erythrocytes which is preferentially sensitive to DNase I digestion (.142). When these proteins of the high mobility QVOUp (HMG) proteins are removed, globin gene is no longer prefer- ents ally susceptible to DNase I. The investigators suggest that somewhere along the DNA near the globin gene, a recognition event ocCurs followed by a preparation event in which DNase I-sensitive St‘F‘ucture is propagated along the chromatin region to be transcribed ( 1 42 ). These HMG proteins may be involved in propagating transcrip- ti c>l"|ally active chromatin structure (142). Endonucleases have been used as tools to investigate the nature f chromatin structure in transcriptionally active and inactive 21 regions of the genome. Weintraub and Groudine (140) have found that pancreatic deoxyribonuclease l (DNase I) can digest DNA sequences corresponding to globin genes in chick erythrocyte nuclei and oval- bumin genes in hen oviduct nuclei, when only 10% of the total nuclear DNA is digested. However, treatment of erythrocyte nuclei with DNase I did not remove ovalbumin gene sequences (140). Furthermore, when s taphylococcal nuclease (also known as micrococcal nuclease) was used to digest erythrocyte nuclei, there was no preferential digestion of active gene sequences (140). The erythrocyte genome was arranged in a nucleosomal structure, yet the DNase I sensitivity of globin gene sequences suggests that nucleosomes in active chromatin regions differ conformationally in some respect. These results were subsequently confirmed in analagous studies by Garel and Axel (42). Gottesfeld and Bonner (44) have employed spleen deoxyribonuclease I I (DNase II) to fractionate chromatin into regions of differing degrees of transcriptional activity followed by selective MgCl2 precipitation. DNase II digests chromatin into segments 100-200 nucTeotides in length. The nuclease-resistant chromatin (PI) is De] Teted. Portions of the remaining chromatin can be selectively DreCipitated in the presence of a divalent cation (Mgz+) resulting in the P2 pellet, and the MgZ+-soluble portions (52) which are thought to Contain transcriptionally active regions of chromatin (44). The Q] obin gene sequence has been localized in the $2, or putative trans- cr‘. Dtionally active, region of Friend leukemia cell chromatin both phi Or to and after induction of hemoglobin synthesis (17,138). These Ys esults support those obtained with DNase I, i.e., there are unique 22 structural characteristics of transcriptionally active chromatin that confer upon them a susceptibility to endonucleases. Hyperacetylated Iristones have been found associated with DNase I-sensitive chromatin, suggesting that this mode of histone modification may be important in (jestermining transcriptional activity (118). However, some evidence 'ilnciicates this type of modification is not sufficient for gene acti- vation (141). Additional evidence indicates that DNase-I sensitive chromatin in higher and lower eucaryotes contains genes transcribed by 61‘1 1 .3 of the RNA polymerases (135). There are conflicting electron tri‘irzroscopic reports on whether or not transcribing chromatin remains 'i '1 a nucleosomal conformation (81). Questions remain to be answered concerning the spacing of RNA polymerase along transcribing DNA, and whether the active structure is in a dynamic equilibrium with native c hromatin structure (81). Staphylococcal nuclease has been shown to digest preferentially 1 inker vs. nucleosomal core DNA (36,75). Early in incubation of Chromatin with staphylococcal nuclease, linker regions became acid 801 uble (75). However, core DNA can be digested by this enzyme at a much slower rate (36,75). These endonucleases can be used as probes for localization of Carcinogen binding within the genome of target cells. The binding of representatives of various classes of chemical can"(Z‘inogens to DNase I-sensitive regions of chromatin DNA has been extensively investigated. When bronchial epithelial cells and fibro- b‘ as ts (target and nontarget cells, respectively) in culture were treated with the polycyclic aromatic hydrocarbon, benzpyrene, 23 carcinogen adducts were formed in DNase I—sensitive (transcriptionally active) regions of chromatin of both cell populations (4). Binding of benzpyrene to DNase I resistat areas of chromatin in target cells occurs rapidly, and adducts persist (4). However, in chromatin of f’ilaroblasts, adduct formation occurs more slowly and adduct removal c)<:<:urs more rapidly from DNase I resistant regions (4). These obser- vations may be important in determining the eventual transformation of '1 IJIig epithelial cells into carcinoma cells (4). Similar studies have been undertaken employing the nitroso com- pound, dimethylnitrosamine (DMN) (102). Ramanathan 9131. found that, 'f’ta'llowing treatment of rats with DMN, DNase I-sensitive regions of ‘1 'i\Ier chromatin contained a higher concentration of methylated pro- ducts compared to DNase I accessible regions of chromatin (102). S tudies of the removal of methylated products from fractions of liver c hromatin revealed that methylated products were nearly completely ‘1 casrt from DNase I-sensitive chromatin 2 weeks following administra- 13‘T<)ri, while 14% of the methylated products remained in nuclease- ? "accessible regions (102). These results suggest that conformational (illéi1.ities that render regions of chromatin accessible to DNase I may a1 80 be responsible for increased susceptibility of these regions to (zaii‘czinogen attack, and to repair enzymes. The aromatic amine, N-OH-AAF, has been found to bind preferen- ti a1ly to DNase I-resistant regions of chromatin from liver cell nuC1ei following treatment of rats with this carcinogen in vivo (84). T . . . . t‘EE concentration of carCinogen adducts remains in DNase I-resistant ChV‘Omatin regions up to 72 hours following treatment (84). Ramanathan 24 ital. (106) have performed analogous experiments with N-OH-AAF, and found that there was a 4-fold concentration of adducts in DNase I- inaccessible regions of rat liver chromatin. This ratio persisted up to one week following carcinogen treatment. The biological signifi- cance of carcinogen localization in these regions of chromatin remains to be more fully elucidated. The fact remains, however, that repre- sentatives of various classes of known chemical carcinogens interact nonrandomly with DNA of target and nontarget cells, and this nonrandom i nteraction appears to be the result of the unique conformation of c hromatin as determined by DNase I accessibility. Drugs may interact with specific regions of chromatin delineated by DNase I susceptibility. Recently, it has been shown that the antitumor antibiotic, bleomycin, causes single strand breaks in DNase I sensitive regions of hen oviduct cell nuclei containing ovalbumin gene sequences, but not in DNase I-insensitive regions of chromatin in o\"iduct cell containing globin gene sequences (70). Thus, chromatin WT th a more open configuration (DNase I-sensitive) may be more suscep- ti b1e to attack by a variety of damaging agents than inactive, con- densed chromatin. DNase II selectively attacks regions of transcriptional activity i n Chromatin. Following treatment of rats in yi_v_o_ with the nitroso Compounds methylnitrosourea (MNU) and dimethylnitrosamine (DMN), a] k.ylation was found to be greater in DNase II digested, MgClz- 30‘ Uble portions, the putative transcriptionally active regions of rat 1 7 Ver chromatin (35). These results are in agreement with similar Studies using DNase I to fractionate rat liver chromatin following carcinogen treatment. 25 In contrast to the DNase I studies, a 16-fold concentration of carcinogen adducts has been found in putative transcriptionally active regions of rat liver chromatin digested by DNase II and solubilized in fifig(212 following treatment of rats with N-OH-AAF in vivg_(ll4). Simi- ‘lair‘ results for preferential carcinogen binding to transcriptionally ai<:1:ive regions of chromatin were confirmed following fractionation of Dd-—()H-AAF-treated rat liver chromatin by the selective MgCl2 chromatin precipitation procedure (115), a sucrose gradient chromatin fraction— ation procedure (90), and the glycerol gradient chromatin fractiona- ‘t:‘i<)n procedure (115). The latter study (115) revealed that carcinogen inreass preferentially lost from transcriptionally active regions of rat 1 iver chromatin after 10 days. The discrepancy between DNase I and DNase II studies following treatment of rats with the aromatic amine N—OH-AAF may indicate that these endonucleases may act by different mechanisms. Staphylococcal nuclease has been a useful tool for investigation of Specificity of carcinogen binding and repair of adducts from Sspefiific regions of chromatin. When normal human fibroblasts are tr‘eated with the ultimate carcinogen N-acetoxy-AAF, carcinogen adducts c’<:<2lni~ preferentially in regions of nuclease sensitivity, i.e., linker '5E3S3‘itans (131). When 3H-thymidine was present in the media to allow for radioactive labelling of repair-incorporated nucleotides, Tlsty and Lieberman (131) found that repair synthesis initially occurred in 1 i rilser regions, but at later times, 3H-thymidine-labelled repair pat-ches were relocated within nucleosomal core regions. Similar r . . . . EEGlrrangement occurred follow1ng repair of UV-light-induced DNA 26 damage, indicating that chromatin structure, rather than type of car- cinogen, may be the determining factor in carcinogen adduct repair. Kaneko and Cerutti (60) reported on an analogous study in which tnitiding to DNA and persistance of adducts in Staphylococcal nuclease sensitive and resistant areas of chromatin from normal human fibro- la‘laasts was investigated following treatment of cells with a lower dose (J'f’ N-acetoxy-AAF than that used in the former study (131). In accord \nl‘iifli results of Tlsty and Lieberman (131), the initial concentration of carcinogen adducts was higher in linker vs. core (60), as was the 'i l1'ltlal loss (presumably due to repair) of adducts. With increasing 't:‘irne for repair, carcinogen adducts continued to be rapidly removed ‘f’iccnn linker regions, and were slowly removed from core regions (60). 'I‘tiease results provided no support for nucleosomal rearrangement <::<>ricomitant with repair of adducts, or as a result of the repair process. The 5- to 60-fold increase in carcinogen concentration to ‘n'fli<:h fibroblasts were exposed in the studies of Tlsty and Lieberman ( 1 31) may have resulted in adduct concentration to an extent that nuCTeosomal rearrangement was induced. Nevertheless, results from these studies illustrate that arrangement of chromatin into nucleo- 5;<3nnEil core and linker regions can influence carcinogen binding and repair. 4 ‘ AAF-Induced Hepatocarci nogenesi s 2-Acetylaminofluorene (AAF) was originally developed by the U.S. Depiirtment of Agriculture in 1940 to be used as an insecticide (146). I\I\F: was found to have low acute toxicity in rats, mice and rabbits, b . . ”11 produced numerous tumors in various organs of rats upon long-term 27 feeding (146). This compound was never marketed, but has since been used as a model compound for study of carcinogenesis induced by aromatic amines. Extensive investigations have shown AAF to be car- cinogenic for liver, urinary bladder, mammary gland, earduct, salivary gland, lungs, forestomach and small intestine (146, reviewed in ref. 37). There are a number of changes in rat liver which consistently occur during and after long-term feeding of a 0.05% (w/w) AAF-contain- i ng diet to rats. At 3-4 weeks of AAF treatment, rat livers appear pale, and light microscopy reveals vacuolization of the hepatocyte cytoplasm, an increase in agranular endoplasmic reticulum, and a d 1' stortion and dilation of bile canaliculi (136). By 12 weeks of AAF Feeding, nodular hyperplasia appears, distorting the normal pattern of 1 i ver structure, along with the continuation of changes that had Occurred at 3-4 weeks of treatment, as well as glycogen accumulation a Nd the appearance of large lysosomes (136). At 8-10 months of treatment, livers were visibly enlarged and finely nodular; the 1 Obular structure was entirely lost, and there was evidence of wide- 3 Dread hyperplasia of hepatic parenchymal cells, along with loss and d ‘3 1ation of agranular and granular endoplasmic reticulum (136). By the end of 10 months on the AAF diet, there was 100% incidence of 1 ‘3 Ver tumors in treated rats (38). Studies involving maintenance of rats on an AAF-containing diet ( 0 - 04% w/w) revealed that the earliest change observed was prolifera- ti on of oval cells in portal areas (33). Autoradiographic studies of EI‘DIDearance and proliferation of oval cells as rats are maintained on a 0 ~ 05% AAF-containing choline-devoid diet revealed that oval cells do 28 not arise from hepatocytes, but rather, arise from a few portally- situated oval cells, and contain a-fetoprotein (119,120). By 5 weeks of maintenance of rats on an AAF-containing diet, hyperplasia of hepatic parenchymal cells was observed (148). This hyperplasia was only observed within hepatic nodules (148). In summary, AAF-induced hepatocellular carcinomas develop in several stages as rats are maintained on an AAF-containing diet (145). Foci, small islands (8-10 cells in diameter) of altered hepatocytes, develop within 3 weeks of carcinogen exposure (145). These cells 5 tain positively for gamma-glutamyl transpeptidase, and are deficient i n adenosine triphosphatase, glucose-6-phosphatase, glycogen accumu- ‘I ation, hyperbasophilia following staining with toluidine blue, and are resistant to iron accumulation (145). Functionally, cells of foci characteristically became resistant to toxic effects of chemicals requiring metabolic activation (145). This properly may allow for the Selective growth advantage of these altered cells. When carcinogen is r‘emoved at this stage, altered foci disappear (145). However, if Carcinogen treatment is continued, islands of foci develop into r‘Odules, which are round, elevated groups of cells approximately the S ‘3 ze of several lobules that compress normal, surrounding liver tissue ( 1 45). The cells of nodules continue to exhibit the same functional abhormalities and enzyme alterations as those of foci (145). Nodules tie\Ielop 2 to 4 weeks following foci development (145). They display Few oncologic properties compared to carcinomas (145), however, it is Q 1 ear that they are composed of abnormal hepatocytes which are resis- ta Int to the cytotoxic effects of carcinogens. Finally, after continued 29 exposure to AAF hepatocellular carcinomas arise, probably from hyper- plastic nodules (145) which display enzyme and functional alterations. Additionally, these carcinomas display progressive growth upon cessa- tion of carcinogen, tumor invasiveness and metastasis (145). From early studies by the Millers and co-workers (87,88), it became clear that AAF was metabolized to an active form as part of its carcinogenic action. The N-hydroxy metabolite of AAF produced tumors at the site of injection (peritoneum), or in the forestomach following ‘i nclusion of AAF in the diet of rats, while similar treatment with AAF d ‘i d not produce tumors at these sites (87,88). Furthermore, guinea p 1‘ gs were resistant to AAF-induced carcinogenesis, yet developed tumors at the site of administration when N-OH-AAF was given (87). R‘i ng-hydroxylated metabolites produced to a great extent following pretreatment of rats with 3-methylcholanthrene had little to no car- C1° nogenic activity (88). Thus, the minor metabolite, N-OH-AAF, was Considered to be an important metabolite of AAF in the carcinogenic process. It is now recognized that the hepatic microsomal P-450 monooxygenase system metabolizes AAF to its N-hydroxy derivative ( 67 ,69,85). The sulfate ester of N-OH-AAF has been found to be important in AAF-induced carcinogenesis. From early studies, the level of liver cytosolic sulfotransferase activity paralleled the level of suscepti- b‘i ‘lity to tumor induction by N-OH-AAF (28). These studies suggested that another metabolic step was necessary to convert N-OH-AAF to a carcinogenic form. Female rats are less susceptible to AAF-induced lr‘epatocarcinogenesis than male rats, and female rats possess one-fifth 30 the cytosolic sulfotransferase activity of male rats (28). More recently, inhibition of sulfate conjugation of N-OH-AAF by penta- chlorophenol or low availability of sulfate has been shown to result in an increase in the percent of the administered dose excreted as a glucuronide conjugate compared to the percent of dose normally ex- creted as such (83). In addition, inhibition of sulfate conjugation ‘in vivo by injecting rats with pentachlorophenol prior to injection wi th N-OH-AAF resulted in a 26% reduction of total DNA binding, most notably reducing the total amount of N-acetylated adducts produced ( 82). Other enzymic pathways have been recognized for conversion of AAF to a reactive electrophile (85). N-OH-AAF may undergo a peroxidase- catalyzed oxidation resulting in a free nitroxide radical, followed by d 1’ smutation of two of these radicals to yield N-acetoxy-AAF and 2- n 'i trosofluorene, both being reactive electrophiles (85). Secondly, cytosolic acetyltransferase may transfer the acetyl group from N-OH- AAF to the oxygen of the hydroxylamine, resulting in formation of the lr‘eatctive N-acetoxy-aminofluorene (85). Lastly, AAF can be converted to the O-glucuronide and/or the N-glucuronide conjugates (58,85). Tl“Iese forms may enable AAF to be stably transported to another tissue, sMich as the urinary bladder, wherein the conjugate can be hydrolyzed 5’“ elding a reactive electrophilic species (58). The alternative metabolic pathways discussed above may be important in activating AAF to a reactive form in extrahepatic tissues (85). 31 5. Interactions of AAF with DNA The assumption that DNA is a critical target for AAF-induced carcinogenesis had led many investigators to study the nature of the interaction of ultimate reactive form of AAF with DNA. Initial studies revealed that guanines within DNA were the primary sites for adduct formation following AAF treatment (see ref. 69, 85 and 85 for r~£a\riew). AAF reacted with DNA to yield 2 types of adducts: 1) an arylamidation product as the result of addition of a nitrenium ion to the C8 of guanine (N-(guanin-8-y1)acetylaminofluorene or N-(guanin-8- y1 )-aminof1uorene), and 2) an arylation product as the result of addition of a carbon cation to the 2-amino group of guanine (3-(deoxy- g uanosin-Nz-yl)-N-acetylaminof1uorene) (40). Approximately 84% of the reaction products with native DNA are arylamidation products, whereas only 16% of the products are derived from an arylation reaction ( 40,67,69). _I_rly_j_v_o_, approximately 65-70% of the arylamidation product was found in a deacetylated form (54,64,69). The arylation product, 3-(dexoyguanosin-N2-yl)-N-acetylaminof1uorene, does not occur i n RNA (65,69). N-(deoxyguanosin-8-yl)-AAF is the predominant adduct i h RNA (the acetylated form predominates) (54,69). The half-life of aciducts in RNA is approximately 3 days (54), while the half-life for the C8 adduct in DNA is approximately 7-10 days (54,69). The N2-gua- AAF adduct in DNA has been found to be persistant in DNA (65). Per- Centages of persisting adducts may vary depending on the strain of rat L"Sec! for investigations (136). Furthermore, adducts may accumulate and persist in nontarget tissues for a carcinogen. In male rat liver, lelowing carcinogen treatment, the NZ-gua-AAF adduct accumulated and 32 persisted for up to 2 weeks of carcinogen treatment (10). However, the C8-gua-AF adduct accumulated and persisted in the liver (target tissue) and kidney (nontarget tissue) of male rats (10). These results suggest accumulation and persistance of adducts alone is not sufficient for tumor induction. The identification of products of carcinogen reaction with nucleic acids led investigators to question the consequences of such 1' nteractions. Through biochemical and physical analyses, it was found that the 3-dimensional confirmation of nucleic acids was modified by AAF (72). These studies led to proposal of the "base displacement model" in which the attachment of the reactive form of AAF to the C-8 position of guanosine results in rotation of the guanine base around the glycosidic bond from the a_n_t_i_ (normal Watson-Crick orientation) to the gm conformation (72). The fluorene residue swings into the helix and is stacking in a coplanar manner with adjacent bases (.72). It has been found that base modifications by carcinogens can alter base- pairing ability, resulting in reduction of recognition of tRNA codons ( 45), and reduction in DNA template capacity, probably due to an aIDIDarent decrease in RNA chain size (116). These interactions result 1 "I local regions of denaturation, as shown by a decreased thermal S tability and intrinsic viscosity of AAF-reacted DNA, as well as e1 ution from a hydroxyapatite column at lower salt concentrations ( 72). Analysis of carcinogen adducts released from digestion by a S ‘3 ngle-strand-specific nuclease of Neurospora crassa of duck reti- QL-I locyte DNA reacted j_n_ BLED). with [9-]4CJ-acetoxy-AAF revealed that tI"|e areas containing the C8-gua-AAF adduct were selectively 33 susceptible to the single-strand-specific nuclease digestion (the N2 adduct remained in undigested DNA) (149). These results strengthened the contention that the C8-gua-AAF adduct caused major conformation distortions in the double helix, while the NZ-gua-AAF adduct does not (149). Kriek and Spelt (68) demonstrated that calf-thymus DNA con- taining C8-gua-AAF reaction products from N-OH-AAF treatment is hydrolyzed 3 times more slowly by nuclease S1 of Aspergillus oryzae compared to DNA containing C8-gua-AAF adducts following reaction with N—acetoxy-AAF. These results indicate that C8-gua-AAF residues result 1" n smaller regions of denaturation in DNA than do CB-gua-AAF residues (:16513). Further studies indicate that carcinogen residues in DNA follow- “i ng reaction with N-0H-AAF _i_rl M are less accessible to adduct- 8 pecific antibodies in native DNA than in denatured DNA lending Support to the base-displacement model (126). Very recently, it has been shown that AAF adducts present in poly(dG-dC) DNA can cause this polymer to change from the standard 8 form at the helix to a newly d‘i scovered 2 form, which is a left-handed helix (in). Circular ‘1 ‘i chroism studies indicate that, in this form, AAF-guanine residues 31"e in the syn conformation and may remain paired with cytosine residues ( 1 11). In this form, guanine is more accessible to attack by carcino- gens, and because there is no base displacement, and therefore, no Clehaturation, this conformation of carcinogen-modified DNA may be more 3 table (111). The biological consequences of this observation remain to be investigated. 34 From the extensive work that has been done on binding and per- sistance of adducts following AAF treatment, several conclusions can be drawn concerning AAF adducts. The acetylated adducts, C8-gua-AAF and Nz-gua-AAF, are probably formed through N,0-sulfation of N-OH-AAF i vivo (82). However, the deacetylated adduct, CB-guaAF, may be 'Formed through an alternative activating pathway (82). The persistant aadducts include CB-guaAF and Nz-gua-AAF (the latter persists for the 'longest periods) (10). The C8-gua-AAF adduct was lost rapidly from ‘liver, while the C8-gua-AF adduct accumulated in target and nontarget 'tissue with progressive carcinogen treatment (10). However, recent (evidence suggests that prolonged carcinogen administration (28 days) Iresults in a decreased ability for removal of acetylated and deacety- ‘lated C-8 adducts (99). Lastly, the CB-gua-AAF adduct causes the sgreatest DNA conformational distortion, while the CB-gua-AF adduct <:auses less distortion (68), and the NZ-gua-AAF adduct results in even 1:255 conformation distortion of the DNA double helix (149). In view (Dif'the unique characteristics of each AAF adduct, it is possible that JQUAF-induced carcinogenesis may depend on effects caused by all three iacjducts, rather than any 1 or 2 of the individual adducts. (5.. Experimental Objectives The objective of this thesis project has been to discern initial rnolecular events critical to target-specific carcinogenesis induced by ‘Cilnemicals. AAF and its N-hydroxy metabolite were used as model IWepatocarcinogens to investigate events related to carcinogenesis ‘tliargeted to hepatic parenchymal cells of the liver. These studies ‘5 ncluded a determination of the binding of the parent carcinogen, AAF, 35 and its N-hydroxy metabolite, to DNA of target cells, parenchymal cells (PC), and at the nontarget cells, nonparenchymal cells (NPC). DNA was isolated from centrifugally elutriated PC and NPC populations of rat carcinogen treatment of rats. Alternatively, DNA was isolated from nuclei of PC (NI nuclei) or from nuclei of NPC (NII nuclei) following the same carcinogen treatment regimens. In both cases, total carcinogen binding to DNA was assessed to determine any speci- f-‘i’ city for carcinogen binding to DNA of target and nontarget cell populations. In view of the apparent nonrandom nature of binding to DNA ex hibited by several carcinogens, studies were undertaken to investi- Qa te whether or not AAF binding to DNA was specific for regions of the genome in target cells and nontarget cells. DNase I was used as a 13001 to enable analysis of carcinogen binding to DNA at transcrip- t"? Onally active and inactive areas of the genome at target parenchymal :91 1 (NI) and nontarget nonparenchymal cell (NII) cell nuclei. Ca Y‘cinogen binding to DNase I-accessible and inaccessible regions of DNA from NI and N11 nuclei was determined at the time of peak binding and 3 days later following treatment of rats with AAF and its N- hycl boxy metabolite. The influence of daily exposure to AAF for l, 3, and 5 days on binding at tracer doses of [ring'3H]-AAF to DNase I- actlessible and inaccessible regions of DNA of target and nontarget Q31 1 nuclei were also investigated. The DNA of target (N1) and nontarget (NII) nuclei was nick- tY‘anslated by incorporation of 32P-labelled deoxyribonucleotide (h 36 triphosphates into DNase I-sensitive regions of chromatin to determine the degree of transcriptional activity within the nuclei of the 2 liver cell populations. Lastly, a series of experiments were performed to determine the effect of carcinogen modification of DNA from target and nontarget cells on the ability of several restriction endonucleases, enzymes which cleave at specific base sequences, to recognize and cleave at these sequences. Changes in the molecular weight of particular re- s triction fragments could be monitored as an indication of impaired restriction enzyme cleavage. These fragments could be identified on the basis of localization of albumin gene sequences within these fragments by hybridization with a radioactively-labelled cDNA probe Complementary for the rat albumin gene. DNA from target and nontarget (N I and N11, respectively) rat liver cell nuclei was analyzed in this ma h ner following treatment of rats for 1, 3, 5 or 7 days with AAF. Another set of rats treated for l, 3, 5 or 7 days was left untreated For 7 days followed by restriction enzyme analysis of DNA from target and nontarget cell nuclei to determine the effect of a period of he31>air on restriction enzyme recognition of carcinogen-modified DNA. From these types of experiments, it is hoped that the importance not only of cell target of a carcinogen may be determined, but also, the DNA target region(s) critical to initiation of carcinogenesis in a pay"ticular cell population may be elucidated. MATERIALS AND METHODS 1 . Animals: Maintenance and Carcinogen Treatment Male, Sprague-Dawley rats (150-200 9), purchased from Spartan Farms (Haslett, M1) were used in these studies. Animals were housed 2 per cage in a room with a controlled 12 hour light cycle beginning at 7 p.m. Rats were given food (Lab-Blox, Chicago, IL) and water a_d 'l “i bitum. When indicated, rats were injected intraperitoneally (i.p.) wi th [ring-3HJ-N-2-acetylaminofluorene or [ring-3]-N-hydroxy-N-acety1- Z—aminofluorene, 51.0 mCi/mmole and 50.8 mCi/mmole, respectively (Midwest Research Institute, Kansas City, MO). [ring-3HJ-N-hydroxy- ac:e‘l:ylaminofluorene was prepared in 0.9% sodium chloride in a concen- thation of 1.8 umole/0.5 ml injection volume/100 g body weight. Eh Ti ng-3HJ-N-2-acetylaminofluorene was prepared in corn oilzDMSO (6:1, V/V ) containing 14% ethanol in a concentration of 1.8 mole carcino- 9el'1/0.5 ml injection volume/100 g body weight. In some studies rats were injected i.p. with N-2-acetylaminof1uorene (Aldrich Chemical, Mi 1 waukee, NI) dissolved in corn oilzDMSO (6:1, v/v) in a dose of 15 mg/ 0.75 ml injection volume/100 g body weight or corn oilzDMSO (6:1, V/ \I ) in a dose of 0.75 ml injection volume/100 g body weight as v eh ‘icle control. 37 38 2. Isolation of Nuclei Liver cell nuclei were isolated from rats by the method of Blobel and Potter (14) as follows. Livers were homogenized in 3 volumes of ice-cold 250 mM sucrose, 50 mM Tris (pH=7.5), 25 mM KCl, 5 mM MgCl2 (STKM) and filtered through cheese cloth. Total liver cell nuclei were isolated by washing twice in 2% Triton X-100 and STKM followed by centrifugation at 750 x g for 10 minutes. The nuclear pellet was washed once in STKM. The method of Bushnell _e_t_a_l_. (19) was used to ‘i solate nuclei of hepatic parenchymal cells (class Nl nuclei) and nuclei of hepatic non-parenchymal cells (class NII nuclei). Following f’ ‘i ltration of homogenate through cheese cloth, 12.5 ml of filtered homogenate and 12.5 ml of 2.3 M sucrose in TKM were mixed, and under- 1 ayed with 10 ml of 2.3 M sucrose in TKM followed by centrifugation at 23 .000 rpm at 5°C for 1 hour. The sw27 Rotor (Beckman Instrument Co.) W61 3 used for this procedure. Following centrifugation the NI nuclei De 1 let was isolated and washed once in STKM. The 750 x g supernatant We 3 added to 255 m1 of STKM and 5.8 ml 20% Triton X-100, and centri- Fu Qed at 5000 rpm at 5°C, 20 minutes. Pelleted NII nuclei were washed tw ‘i ce in STKM followed by centrifugation at 4000 rpm, 5°C, 10 mi hutes. 3 - Isolation of Distinct Liver Cell Populations This procedure was performed in collaboration with Dr. James S"'Wenberg (Chemical Industry Institute of Toxicology, Research Triangle park, NC). Populations of hepatic parenchymal (PC) and non-paren- chymal (NPC) cells were isolated by a method for centrifugal elutri- ation detailed elsewhere (Lewis and Swenberg, 74). Livers from rats 39 were perfused with collagenase through the portal vein. The mixed liver cell suspension was washed twice in Hepes buffered saline solu- tion (HBSS), then slowly injected into a large mixing chamber of the Beckman JE-6 elutriation rotor containing a Sanderson separation chamber in a Beckman 02-21 refrigerated chamber. The rotor speed was held constant at 1,500 rpm and the flow rate varied. The nonparen- chymal cells were separated at 18 ml/minute, and the hepatocytes were 'i’ solated at 100 ml/minute. Elutriated nonparenchymal cells (NPC) were composed of Kupffer and endothelial cells, in contrast to the popula- t 'i on of NPC from which NII nuclei were derived. NII nuclei were 0 btained from Kupffer and endothelial cells as well as from bile duct cells and fat-storing cells. 4 - Nuclease Digestion of DNA from Nuclei DNase I was used as a tool to delineate transcriptionally active DNA from inactive DNA. Parenchymal (NI) and nonparenchymal cell nu c‘lei (NII) were digested by pancreatic deoxyribonuclease I according to a method of Weintraub and Groudine (140). Nuclei were washed twice i n 10 mM Tris (pH=7.4), 10 mM NaCl, 3 mM MgC12 (TNM). Nuclei were 8“ S pended by gentle homogenization in TNM such that 1 ml suspension contained 1 mg DNA. DNase I was added (20 09/125 units/mg DNA) 1:01 lowed by incubation at 0, 5 and 10 minutes at 37°C with mild shak- ing . The reaction was terminated by addition of ice-cold trichloro- at:etic acid to a final concentration of 10%. Suspensions were centri- fuged at 2500 rpm, 5°C, 5 min. Supernatants were boiled for 5 minutes E‘chi cooled on ice to preferentially solubilize DNA digested by DNase 40 I. Supernatants were centrifuged at 5000 rpm, 5°C, 10 minutes. Supernatants were removed for DNA analysis. 5. Nick Translation of Transcriptionally_Active DNA from Nuclei of Parenchymaf and NonparenchymaT Liver Cells It has been shown recently that DNA of hen oviduct cell nuclei can be nicked by the endonuclease DNase I and translated with E. _c_9_]_j_ [)PGA polymerase I using 32P-labelled deoxyribonucleotide triphosphates resulting in radioactive labelling in regions of transcriptionally active chromatin (73). This procedure was adapted to nick translate DNA of target (PC) and nontarget (NPC) liver cells. DNase I was added in a sufficiently low concentration so as to introduce single- 5 trand nicks into active regions of chromatin. Subsequent addition of DIVA polymerase I results in incorporation of 32 P-labelled deoxyribo- n ucleotide triphosphates in nicked areas opposite an intact template. Hepatic NI and N11 nuclei were isolated by discontinuous sucrose g"‘adient centrifugation and suspended in nick-translation buffer (50 ""M Tris, pH=7.9, 5 mM M9012, 10 mM 2-mercaptoethanol, 50 ug/ml BSA) in a concentration of 1 mg nuclei DNA/m1. DNase I was added (0.3 ug/ml) and the suspensions were incubated at 37°C for 5 minutes. Deoxyribo- “LI (2 leotide triphosphates (dATP, dCTP, dGTP and dTTP) were added to a fi hal concentration of 4 pm each and 26 pmoles (769 mCi/nmole) of each 32 P —nucleotide triphosphate was added. E. coli DNA polymerase I was add ed (10 units/ml) and incubation proceeded for 5 minutes at 15°C. F01 lowing termination of the reaction on ice, nuclei were centrifuged at 4000 rpm for 10 minutes at 5°C. Nuclei were washed 3 times in nick translation buffer to remove unincorporated nucleotides. .l a 3 P 41 32P-labelled DNA from NI and N11 nuclei was isolated by Marmur ex- traction (6,49,80). DNA was digested by the restriction endonuclease Eco Rl (0.15 units/pg DNA) and applied to a 0.9% agarose gel and run eat 50 V overnight. DNA in the gel was stained with ethidium bromide (().5 ug/ml) and photographed while being illuminated with a UV trans- i lluminator (Ultra-Violet Products, Inc., San Gabriel, CA). The gel ‘VJEiS dried under vacuum onto gel backing paper (Bio-Rad). ‘The dried gel was exposed to X-ray film (Kodak X-OMat AR) for 4-24 hr, then developed. 6 - DNA Isolation A. Hydroxyapatite Column Chromatography The method of Beland gt_al, (9) was used to isolate DNA from pa renchymal and nonparenchymal cells or nuclei with the following modifications. Hydroxyapatite (Biogel DNA-Grade HTP, Bio-Rad) was prepared in 1 g portions/mg DNA. Hydroxyapatite was washed and fines we re decanted twice in 0.014 M Na2P04 (pH=6.8) at 25°C. The hydroxy- aF36.‘l:.ite slurries were poured into glass wool-plugged plastic spin ‘tz'i lfirltales (Reeve-Angel) and placed in centrifuge tubes. Columns were pa-‘=::losms mm: AHdv cepoEocgu ucmumwm um; mmcm_o:z .APFnzav mesh 25 om cw umpocwscmp mm: cowpommc we» use .wpomme use :owummm_u HH wmmzo An :wumeoggo co cowpmcowpoocd .N mszmwd 59 N weaned .wnumziHH on some msp mm ummmmcaxm use mafiammm .cmum_ mzmv m use Acowuomecwiumoa s; wpw mcwucwn xmma to weep me» we umcwsswpmo mm: <20 me cma ucson :mmOCVULoo mo pczoEm esp .A oo_\mo_cE: w._v a<Fm>wuomammc .msma oesopm; use :maov m__mu Aodzv Pegasucmcoacoc new Auev Posxzocmcma owpmam; mo <2: op mcmcozFeocwsmpzpmomumzmimcvsc to oceucwm .e manure 68 PCD NPC. 7/// ”'4 90 post-iniulion Figure 4 M————:f 18 l L O 0 Hours 9" N (90lx)VNa Bw/ianppn «Iowa 69 .Lon m an ompcmmmcamc we cocsm use -ucmum .mme m scam mspm> some msp mm ummmmsaxm «so mupzmmm .cmum_ axon m use .Acowpumh inviumoa c; NV mcwccwn xmma we we?“ we“ pm umcwscmpmu we: wuomammc .mcmc cocoon; use :maov mppmo Acozv Fmezgocmcmq120c cam Auev Fmezco -cmsma uwpmqm; mo <20 ou wcmcozpmocwEmPxpmomxxosuxciz-mzmimcetc to mcwncwm .m mczmwd 70 % Us: D U.— m weaned cozuzfilnon 950: Ne. N J4 *-—L *-—£ 1 L c: (N) l a Q” (golX)VNQ Bw/aanppn solown 71 at the time of peak binding (2 hr for N-OH-AAF), or 3 days later. As with AAF, at 3 days after peak binding, there was a similar extent of loss of adducts from DNA of both cell populations, i.e., 54% of the initially-bound adducts remained in DNA of PC, and 57% remained in the DNA of NPC at 3 days after peak binding. When these data are adjusted for the total amount of DNA within these cell populations of the whole liver, there is 19 and 24 times the amount of carcinogen bound to the DNA of target cells (PC) com- pared to DNA of nontarget cells (NPC) as shown in Table 6, at the time of peak binding as well as 3 days later, respectively. In addition, this increased concentration of carcinogen adducts in DNA of PC compared to that of NPC results following N-OH-AAF administration, though not as great as following AAF treatment. 4. Carcinogen Binding to DNA of Nuclei Isolated from Parenchymal and Nonparenchymal LiVer Cells A procedure developed by Bushnell eE_al, (19) was used to isolate nuclei derived from PC and NPC by sedimentation through a discontinu- ous sucrose gradient. Nuclei isolated in this manner were used for preparation of highly purified DNA following carcinogen treatment of rats jg_vivo. Experiments carried out on the binding of AAF and its N-hydroxy metabolite to DNA of PC and NPC were repeated on isolated DNA of PC and NPC nuclei populations (NI and N11 nuclei, respectively) following treatment of rats with 1.8 pmoles/100 g of either carcinogen. Figure 6 shows there was no significant difference in binding of carcinogen to DNA isolated from NI or NII nuclei at 18 or 90 hours following 72 Table 6 Total Binding of AAF and N-OH-AAF to Liver Cell DNA DNA Binding (pmol/total DNA) Hepatocyte/Sinusoidal Carcinogen/Time . Hepatocyte Sinusoidal Cell 09“ Ratlo AAF 18 hr 230:43 12:4 19.2 90 hr 95:21 4:1 23.8 N-0H-AAF 2 hr 141:32 17:3 8.3 72 hr 76:31 10:3 7.5 73 Figure 6. Binding of [ring-3H]acety1aminofluorene to DNA of hepatic parenchymal (N1) and nonparenchymal cell (NII) nuclei (open and hatched bars, respectively). Following a single i.p. injection of [ring-3H]AAF (1.8 pmoles/100 g). NI and N11 nuclei populations were isolated from whole liver homogenate (25% w/v) by discontinuous sucrose gradient sedimentation as described in Methods. The amount of carcinogen bound per mg DNA was determined at the time of peak binding (18 hr post-injection), and 3 days later. Results are expressed as the mean value from 6 rats. Standard error is repre- sented by a bar. 74 NI N11 90 3H-AAF 1 Tc #1, P p (c O 5 O 5 2 I I 3 o3osms we; ucopmccmasm one .oom .cws m Low Ea; comm pm ummzmwcpcmo mam; mcowm -cmamzm .eOF to :owpmcpcmocoo _mcee m ow w_ A>\zv emu m Ease umHmFomw mew: wm_u:: Fmpoh .H mmmm_u=:onwsxxomu uwpmmcucoa cue: wm_o:: rpmu sm>wp no; mo cowpmmmwo .m menace m mczmwd Ant-35F; mirp Go On 9. On ON 9 o 81 a d u A u a , t i l O O O 0 V N 03193910 VNO X 1 O Q 20:.ww05 . 3020 deep 82 TABLE 8 DNA Digested from Parenchymal Cell Nuclei by DNase Ia Digestion Time 5 Minutes 10 Minutes Control 45:20b 68:27 3H-AAFC 18 hr 38:11 58:13 90 hr 68:33 81:36 3H-N-0H-AAFC 2 hr 33: 8 64:10 72 hr 56:26 66:14 aPC nuclei (Class NI) were incubated with DNase I (125 units/mg DNA) for 5 or 10 minutes at 37°C. Incubation was terminated by the addi- tion of ice-cold TCA, final concentration 10% w/v. bPercent of total PC DNA. Data represent the mean : S.E. of the results obtained from 4-6 rats. cMale Sprague-Dawley rags were injected i. p. with 1. 8 pmoles/100 g H-AAF (35.5 mCi/mMole) H- N- 0H- AAF (55. 6 mCi/mMole). 83 active DNA in hepatic parenchymal cells as compared to nonparenchymal cells. Alternatively, parenchymal cell nuclei (NI) may be more per- meable to DNase I than are NII nuclei. Similar results for nuclease digestion of DNA from nonparenchymal cells are summarized in Table 9. Following treatments of rats with AAF or N-hydroxy AAF, there was no significant difference in the amount of DNA digested as compared to control except at 5 minutes digestion of DNA from rats 90 hours following treatment with AAF, which was significantly different from control (p<0.05). These results from studies on DNase I digestion of DNA from target (NI) and nontarget (NII) nuclei suggest that the initial presence of adducts in DNA of nontarget cells and subsequent repair of adducts does not influence the ability of DNase I to cleave DNA, i.e., does not enhance or inhibit enzyme activity. Following treatment of rats with [ring-3HJ-AAF or -N-0H-AAF (1.8 pmole/100 g), the DNA from N1 and NII chromatin which was nuclease- accessible and nuclease-resistant was analyzed for carcinogen binding. As shown in Figure 9, there is significantly less carcinogen bound per mg DNA to DNase I-accessible DNA compared to that bound per mg DNA of total undigested NI DNA at the time of peak binding as well as 3 days following. This trend is noted following treatment of rats with N-OH- AAF, although this difference is not statistically significant (Figure 9). These results suggest that regions of the target cell DNA that are transcriptionally active (DNase-I sensitive) are not necessarily the same regions to which carcinogen selectively binds. 84 TABLE 9 DNA Digested from Nonparenchymal Cell Nuclei by DNase Ia Digestion Time 5 Minutes 10 Minutes Control 9:4 27:16 3H-AAFC 18 hr 16:83 19: 7 90 hr 34:6 33: 6 3H-N-0H-AAFC 2 hr 12:3 17: 6 72 hr 9:2 16: 5 aNPc nuclei were incubated with DNase I (125 units/mg DNA) for 5 or 10 minutes at 37°C. Incubation was terminated by the addition of ice-cold TCA, final concentration 10% w/v. bPercent of total NPC DNA. Data represent mean : S.E. of the results obtained from 4-6 rats. cMale Sprague-Dawley rags were injected i.p. with 1.8 pmoles/100 g H-AAF (35.5 mCi/mmole) or 3H-N-0H-AAF (55.5 mCi/mmole). dSignificantly different from control as deter- mined by Student's t test, p<0.05. 85 Figure 9. Binding of [ring-3HJ-acetylaminofluorene and [ring-3H]-N- hydroxyacetylaminofluorene to total DNA and DNase I-digested DNA from parenchymal cell nuclei. Solid bars represent binding to DNase I- digested DNA at the time of peak binding (18 or 2 hr after 3H—AAF or 3H-N-OH-AAF treatments, respectively) and 3 days later. Open bars represent binding to total DNA at the indicated time periods. The data shown represent the mean : SEM of the values obtained from 3-6 rats. The standard error for the binding to DNase I-digested DNA overlapped that of the binding to total DNA for the 3H-N-OH-AAF treatment group at 3 days after peak binding. *Significant at p<0.05 as determined by Student's t-test. 86 p nuw n1. mi. m 5 303720 96:33.0 10:3 Figure 9 87 Binding of carcinogen to DNase I-accessible regions of nontarget cell NII DNA compared to that of total, undigested NII DNA is repre- sented in Figure 10. At the time of peak binding for AAF or its N- hydroxy metabolite, there is a greater quantity of carcinogen adducts per mg nuclease-accessible DNA as compared to the amount bound per mg undigested DNA. However, 3 days following AAF treatment, no detect- able carcinogen remained bound to DNase I-accessible regions of DNA. Approximately 42% of the initially bound carcinogen remained in total, undigested NII DNA 3 days following AAF treatment. Three days follow- ing treatment of rats with N-0H-AAF, carcinogen bound to DNase I- accessible regions of NII DNA was detectable (13% of that initially bound). Approximately 28% of the initial carcinogen adducts remained in the DNA of total, undigested NPC genome. Thus, transcriptionally active (DNase I-accessible) areas of nontarget cell (NII) DNA are highly sensitive to carcinogen attack initially, however, damage within these regions is rapidly repaired. Data from studies on AAF binding to DNase I-accessible and in- accessible regions of DNA from NI and NII nuclei populations are compiled in Table 10. Carcinogen adducts accumulate in nuclease- resistant regions of target cell nuclei (NI) DNA at the time of peak binding following AAF treatment, as well as at 3 days following. However, AAF preferentially attacked nuclease-accessible regions of nontarget cell nuclei (NII) cell DNA at the time of peak binding. The carcinogen adducts in these nuclease-accessible regions were rapidly repaired, while carcinogen adducts remained in DNA of DNase I-in- accessible regions of DNA from NII nuclei. 88 Figure 10. Binding of [ring-3HJ-acetylaminof1uorene and [ring-3H]- N-hydroxyacetylaminofluorene to total DNA and DNase I-accessible DNA from nonparenchymal cell nuclei. Solid bars represent binding to DNase I-accessible DNA at the time of peak bindin (18 or 2 hr after 3H-AAF or 3H-N-0H-AAF treatments, respectively) and 3 days later. Open bars represent binding to total DNA at the indicated time periods. The data shown represent the mean : S.E. of the values obtained from 3-6 rats. 89 «b 0 on O O 0 ”M01” adduct/my DNA‘X 106) to O 3H-AAF - 3H-N-OH-AAF -’fi p----fi ' L} L ' ' P. k 3 D P k 3 Bincring ay 3113:1119 Day Figure 10 90 TABLE 10 N-2-Acetylaminofluorene Adducts in Specific Regions of Chromatin DNA of Different Hepatic Nuclei Populations Nuclei Nuclease Nuclease Po ulation Accessible Inaccessible p (pmoles/mg DNA) (pmoles/mg DNA) Parenchymal 18 hb 4.53: 2.7C 26.24:8.03 90 h 2.05: 1.2 9.81 Nonparenchymal 18 h 33.70:17.4 5.51:2.77 90 h 0.00 9.24 aNuclei were incubated with DNase I (125 units/mg DNA) for 10 min at 37°C. Nuclease-accessible DNA remained soluble in a supernatant. Nuclease-inaccessible DNA was TCA-insoluble, pelleted at 5000 rpm. bRats were injected i.p. with 1.8 pmoles [ring-3HJ-N-2- acetylaminofluorene/lOO g and sacrificed 18 and 90 hr following injection. CMean value of 4-6 rats : SEM. 91 Similar results following treatment of rats with N-OH-AAF are summarized in Table 11. At 2 and 72 hours after treatment of rats with N-OH-AAF, there was significantly more carcinogen bound to nuclease inaccessible regions of DNA from N1 nuclei (p<0.05). Car— cinogen adducts appeared to concentrate in nuclease accessible areas (the putative transcriptionally active regions) of DNA from NII nuclei at 72 hours following injection, there was a 7.5-fold loss of car- cinogen adducts in the nuclease accessible regions of NII nuclei DNA, while there was a 3-fold loss of the adducts in nuclease-inaccessible regions of NII DNA. The amount of carcinogen bound per mg DNA of nuclease-accessible and inaccessible regions of NI and NII nuclear DNA is higher in every case following N-0H-AAF treatment in comparison to AAF treatment (Tables 10 and 11). N-0H-AAF it is one metabolic step closer to the ultimate reactive species (85,86). To determine whether pretreatment of rats with daily doses of AAF alters the pattern of binding for AAF to DNase I-accessible and -in- accessible regions of DNA derived from N1 and NII nuclei populations studies were done in which rats were given daily i.p. injections of AAF in corn oilzDMSO (15 mg/0.75 m1/100 g) for 1, 3 or 5 days. At 24 hours following the last dose of AAF, a tracer dose of [ring-3H]-AAF (91.8 pCi/l.8 pmole/100 g) was given 18 hours (time of peak binding) prior to sacrifice. Results presented in Table 12 indicate that rats treated for 3 and 5 days with AAF contained significantly fewer car- cinogen residues bound to hepatic macromoles (p<0.05), suggesting that 92 TABLE 11 N-Hydroxy-Acetylaminofluorene Adducts in Specific Regions of Chromatin DNA of Different Hepatic Nuclei Populations N clei Nuclease Nuclease Po Elation Accessible Inaccessible p (pmoles/mg DNA) (pmoles/mg DNA) Parenchymal 2 hb 7.15:2.5C 22.14:12.83 72 h 4.91:1.2 13.89: 4.0 Nonparenchymal 2 h 53.9l:8.9 21.19 72 h 7.15:3.29 6.85: 2 l aNuclei were incubated with DNase I (125 units/mg DNA) for 10 min at 37°C. Nuclease-accessible DNA remained soluble in a 5000 rpm supernatant. Nuclease-inacces- sible DNA was TCA-insoluble, pelleted at 5000 rpm. bRats were injected i.p. with 1.8 pmoles [ring-3H]-N- hydroxy-acetylaminofluorene/l00 g and sacrificed 18 and 90 hr following injection. cMean value of 4-6 rats : SEM. dSignificantly different from nuclease accessible, p<0.05 as determined by Student's t test for 2 means. 93 TABLE 12 Analysis of Carcinogen Binding to Hepatic Macromolecules Following Pretreatment with AAF or Vehicle-Controla Days Treatedb ControlC Pretreatment (pmoles adduct/mg DNA) 3 869:124d 3,993:808 5 1,008: 8id 3,945:444 aMale Sprague-Dawley rats were injected i.p. wi h a tracer dose (91.86 pCi/l.8 pmole/100 9) [ring- H]- AAF following 3 or 5 days pretreatment with AAF or vehicle. Portions of 25% w/v liver homogenate (0.5 ml) were precipitated with 5% TCA and washed once, followed by solubilization of pellet in 88% formic acid. bRats were injected i.p. with 15 mg/0.75 m1 injec- tion solution/100 g AAF in corn oilzDMSO (6:1, v/v) daily for 3 or 5 days. cRats were injected i.p. with 0.75 ml corn oil: DMSO (6:1, v/v)/100 g daily for 3 or 5 days. dSignificantly different from control, p<0.05, as determined by Student's t test. 94 pretreatment with AAF results in a decreased ability to metabolically activate [ring-3H]-AAF. This observation has been noted previously (114). Carcinogen binding to DNase I-accessible regions of DNA from PC and NPC nuclei following carcinogen pretreatment is summarized in Table 13. Following 1, 3 or 5 days pretreatment with AAF (15 mg/100 9), there was no difference from vehicle-injected rats in carcinogen binding to DNase I-accessible regions of DNA from NI or NII nuclei. These results suggest that the presence of carcinogen adducts in DNA of PC or NPC nuclei did not inhibit or enhance the ability of DNase I to recognize and/or cleave at its target sites within DNA. 6. Nick Translation of Transcriptionally Active DNA in Intact PC and NPC Nuc1ei DNA of PC and NPC nuclei was nick-translated to determine the degree of transcriptional activity within the nuclei of the 2 liver cell populations. NI and NII nuclei were isolated from untreated rats. DNase I (Sigma Chemical Co.) was used to nick DNA selectively in transcriptionally active regions of DNA. These areas were then filled in with E, coli polymerase I and 32 P-labelled deoxyribonucleo- tide triphosphates. Following agarose gel electrophoresis of equal amounts of DNA derived from N1 and NII nuclei digested by the re- striction endonuclease, Eco RI, autoradiography after 24 hours reveals a great deal more transcriptional activity in DNA of PC nuclei com- pared to equal amounts of DNA from NPC nuclei (Figure 11). After only 4 hours autoradiography, labelled transcriptionally active areas of only the NI nuclear DNA were detectable (Figure 12), providing further 95 TABLE 13 Analysis of Carcinogen Binding to DNase I Accessible DNA of Rat Parenchymal and Nonparenchymal Liver Cell Nuclei Following Pretreatment with AAF or Vehicle-Controla Days b Parenchymal Cell Nonparenchymal Cell Pretreatment Nuclei (NI) Nuclei (NII) 1 c 21.1:8.7d 22.5: 12.5 Control l9.0:0.9 36.0 3 12.8:4.2 22.8: 5.4 Control 12.7:1.0 26.2: 13.8 5 18.1:7.5 133.9: 27.8 Contro1 13.8:0.6 267.2:126.9 aMale Sprague-Dawley rats were injected3i.p. with a tracer dose (91.8 pCi/l.8 pmole/100 9) [ring- H]-AAF following 3 or 5 days pretreatment with AAF or vehicle. Nuclease-accessible DNA remained soluble in a 5000 rpm supernatant following termination of DNase I digestion at NI and N11 nuclei with TCA. bRats were injected i.p. with 15 mg/0.75 ml injection solu- tion/100 g AAF in corn oilzDMSO (6:1, v/v) daily for 1, 3 or 5 days. cRats were injected i.p. with 0.75 ml corn oilzDMSD (6:1, v/v)/100 g daily for l, 3 or 5 days. dmoles carcinogen adduct x 10'6/mg DNA. Data represent mean : SEM of 3 rats. 96 Figure 11. Incorporation of 32P-labelled deoxyribonucleotide triphosphates into DNase I-accessible regions of DNA from hepatic parenchymal cell and nonparenchymal cell nuclei. Parenchymal cell nuclei (NI) and nonparenchymal cell nuclei. Parenchymal cell nuclei (MI) and nonparenchymal cell nuclei (NII) were isolated as described in Methods. Nuclei were suspended in 50 mM Tris (pH=7.9), 5 mM MgC12, 10 mM 2-mercaptoethanol, 50 pg/ml BSA, in a concentration of lung nuclei DNA/m1. DNA was nicked by DNase I (0.65 units/m1), and nucleotide triphosphates (26 pmoles of 32P-labelled nucleotide tri- phosphates) were incorporated by E, coli DNA polymerase I (10 units/m1) as described in Methods. Purified DNA was di ested by Eco R1 (1 unit/pg DNA) and equal amounts (50 and 100 pg) of NI and NII DNA were run on a 0.9% agarose gel. Molecular weight of 32P- labelled fragments ranged from less than 2.2 kilobases to greater than 17.5 kilobases. The dried gel was autoradiographed for 24 hr. 97 Figure 11 98 Figure 12. Incorporation of 32P-labelled deoxyribonucleotide triphosphates into DNase I-accessible regions of DNA from hepatic parenchymal cell and nonparenchymal cell nuclei. 32P-labelled DNA prepared from parenchymal cell and nonparenchymal cell nuclei as described above (see legend, Figure 11). The dried agarose gel was autoradiographed for 4 hours. NII a b C 99 d 9 Figure 12 ....1145Ith — 9.1 100 evidence that liver parenchymal cells, the functional cells of the liver, contain a much greater amount of transcriptionally active DNA than do nonparenchymal cells. 7. Effect of Continued AAF Treatment on Restriction Endonuclease Digestion of DNA The following experiments were performed to determine whether or not restriction endonucleases, enzymes that recognize specific base sequences could recognize and/or cleave DNA that has been modified to various degrees by carcinogen treatment of rats in_vjv9, The specific endonucleases used to cleave DNA of rat liver were: 1) Eco R1 which cleaves the following base sequence at the point indicated by the arrow: G+AATTC and 2) Kpn 1 which cleaves the following base sequence at the point indicated by the arrow: GGTAC+C. The digestion of DNA by Eco R1 from liver of an untreated rat is illustrated in Figure 13. Following digestion of DNA from rat liver by Eco Rl, fractionated DNA was electrophoresed on an agarose gel to separate restriction fragments on the basis of molecular weight. Lanes a and b of Figure 13 represent the ethidium bromide stained agarose gel. Eco R1 digests rat genomic DNA (b) into a multitude of fragments ranging from less than 2.2 kilobases (Kb) to greater than 17.5 Kb (determined by standard molecular weight markers of Lane a). Eco R1 restriction fragments of rat liver DNA containing albumin gene sequences is illustrated in Lane c of Figure 13. Following Southern 32 transfer of DNA from agarose gels to nitrocellulose filters, P- labelled cDNA for the rat albumin gene hybridizes to homologous 101 .UoOml #0 mXoo m oogoocmowoocopzo woo; mpopo omNVopsoxx .Aogv momoooppx cw oopoopocp .m smoocgp < mocoo to mpzmwoz Lopooopoe moassopoo op o>e=o ocoooopm ocp co ooppopo mo; pcoEmocw ooppmoop comm so» oocopmpo copposmwe och .Ao mcoov moocmooom moomopoeo; mcpcpopcoo mpcmsmosw coppowopmoc pm oom op ooNpopco>5 mo; Acoccom mosoo .Lo mo pepm .m_pp pa: to coppmompa apdpano oscoaao co 8:8 Ac: N.N .ao 4.: .8: N.@ .ox m.mpv o moo; op oopoopoop mosoocopm pompoz copoompos to AEov mooopmpo oopposmpe mop ocpEsmpoo op A_E\mn m.ov oopEoco Espowcpo op mowcpopm mcpzoppow omsoocmopogo coop .pom omocomo pm.o a co admocoeaocpoopa ma: Am: oav <2: eopmoopo .oopm pa mead: N toe App Eocp <23 pm oom xo mpoc oopoocpco eo pp mo coppmompo .mp oczmpe 102 m— mgompm .50. 85me m—N=O—omhom¢mup __——_ _—_ - a _ _ _ m in m 8. . .2 Sin m to... o 1 8.. .o - mm . m H 8.34 < .8 “ Hoo— 103 sequences in restriction fragments of rat genomic DNA covalently bound to the nitrocellulose filter. The distance from the origin to which these fragments have migrated is compared with molecular weight standards, and the molecular weights of albumin gene-containing re- striction fragments can be determined on the ordinate. Albumin gene- containing Eco R1 restriction fragments have molecular weights of 1.7, 2.3, 2.8 and 3.8 kilobases (Figure 13). Digestion of DNA from untreated rat liver by the restriction endonuclease Kpn l is represented in Figure 14. Lanes a and b of the agarose gel stained in ethidium bromide represent standard molecular weight markers, the same as those employed for the studies presented in Figure 13, and rat genomic DNA digested by Kpn 1, respectively. Hybridization of 32P—labelled rat albumin gene cDNA to Kpn l restric- tion fragments of rat liver DNA is represented in Lane c of Figure 14, illustrating that Kpn 1 produces albumin gene-containing fragments of DNA having molecular weights of 2.3, 2.9, 3.9, 4.7, 9.7 and 16.0 Kb. Rats were given daily i.p. injections of 15 mg AAF/100 g or 0.75 ml corn oilzDMSO (6:1, v/v) for l, 3, 5 or 7 days. The effect of AAF treatment on liver to body weight ratios is illustrated in Table 14. The toxic effect of the carcinogen on rats is noted as early as after 3 days AAF treatment, a time at which the liver/body weight ratio of treated rats is significantly higher than control rats (p<0.05). The increasing liver/body weight ratio with progressive AAF treatment is due to a decreasing body weight occurring simultaneously with an increasing liver weight. No change is seen in vehicle—injected rats (Table 14). 104 .Aaxv moaaaop_x e_ eopaopaep .Ad-c=o osoooopm pzmpoz Lopooopoe mop :o oocpELopoo moo: mmocoooom memo cpsoopo mcpcpop -coo mpooEmocp :oppopcpmoc p cox to mpgmpoz copzoopoe och .Am ocopv Amflwp poo wo coppmompo opoposou .Ampzmpoz copooopos to» ooomop mp osompe oomv o woo; op oopoopocp mocoooopm pompmz copooopos to oucopmpo oowpocmpe mop mochopoo op Ape\mo m.ov oopeoco Ezpopcpo op mcpopopm mcwzoppow oooooooopoco :o2p .pom mmocomo em.o o op oomosogoospoopo was <29 oopmompo .uomm po mono; N Low App Soto <2o ._ cox 2o mpos oopomcpcz mo <29 oo>pp mo ooppmomwo .op osompe 105 e_ oceape .52 85me m—N—Zo—omnomvmm— _— ___ _ p _ _ p _ _ _ -m “2 5. .2. 10m Hon e. 9:1 I < - moo— 106 TABLE 14 Effect of Daily N-2-Acetylaminofluorene Treatment on Liver/Body Weight Ratiosa Day Treatedb Controlc 1 0.044:0.003 0.045:0.003 3 0.059:0.010d 0.047:0.004 5 0.055:0.006d 0.042:0.006 7 0.058:0.002d 0.047:0.001 aDetermined as liver weight/body weight on day of sacrifice. Data represent mean :_SEM of 3-6 rats. bMale, Sprague-Dawley rats were injected i.p. with 15 mg/100 g 2-acety1aminofluorene in corn oilzDMSO (6:1, v/v) for l, 3, 5 or 7 days. cMale, Sprague-Dawley rats were injected i.p. with 0.75 m1/100 9 corn oilzDMSO (6:1, v/v) for l, 3, 5 or 7 days. dSignificantly different from control as determined by Student's t test, p<0.05. 613. tar .. in. A“ «wok. 107 Following treatment of rats for 1, 3, 5 or 7 days, DNA was iso- lated, electrophoresed in agarose gels, transferred to nitrocellulose and hybridized with 32 P-labelled rat albumin gene cDNA. Autoradio- graphs of albumin gene-containing Eco R1 and Kpn l restriction frag- ments of rat DNA are depicted in Figures 15-18. Following treatment of rats with AAF for 1 day, there was no difference in the pattern of Eco R1 restriction fragments containing albumin gene sequences in treated vs. control rats in DNA either from target (NI) or nontarget (NII) nuclei (Figure 15, Table 15). One additional albumin gene-containing restriction fragment was observed in DNA of N11 nuclei (Figure 15, Lanes c, d, e). Following 3 days of treatment with AAF, there was no difference from control in the ability for Eco R1 to recognize and cleave at specific sites in DNA of target (NI) or nontarget (NII) nuclei (Figure 16, Table 16). However, 3 albumin gene-containing Kpn I restriction fragments observed in control rat DNA were not present in DNA of NI and NII nuclei from AAF-treated rats. Control DNA from NII nuclei fractionated by Kpn 1 exhibited one additional labelled restriction fragment of 8.0 Kb not observed in DNA of NI nuclei (Figure 16, Table 16). Similarly, at 5 days of AAF treatment, the distribution of labelled Eco R1 restriction fragments of DNA from NI and NII nuclei was not influenced by AAF treatment (Figure 17, Table 17). However, 2 fragments (9.7 and 2.2 Kb) and 5 fragments (9.3, 4.4, 3.7, 3.0, and 2.25 Kb) were no longer observed in DNA of NI or NII nuclei, respec- tively, following carcinogen treatment (Table 17). 108 Figure 15. Eco R1 digestion of liver DNA of rats treated 1 day with N-2-acetylaminof1uorene (AAF) or with vehicle. Male Sprague- Dawley rats were injected i.p. with AAF (15 mg/100 g) and sacrificed 24 hours later. Nuclei of hepatic parenchymal cells (N1) and nonparenchymal cells (NII) were isolated by discontinuous sucrose gradient centrifugation. DNA was purified free of protein and RNA (see Methods). DNA (40 pg) was digested by Eco R1 and electropho- resed, then transferred and hybridized as indicated (see legend to Figure 13). Lanes a, c and d represent DNA from rats treated 1 day with AAF. Lanes b and e represent DNA from vehicle control rats (0.75 ml corn oilzDMSO per 100 9). Sizes of restriction fragments labelled with 32P-labelled cDNA for rat albumin gene are indicated in kilobases. 109 Eco R1 ...ilOS __ L55 110 TABLE 15 Eco R1 Restriction Endonuclease Fragmentation of DNA from Rats Treated 1 Day with N-2-Acetylaminofluorenea b NI NII DNA Fragment c d Treated Control Treated Control 11.5 11.5 A 3.55e 3.55 3.4 3.4 B 2.7 2.7 2.7 2.7 C 1.75 1.75 2.05 2.05 D 1.5 1.5 1.5 1.5 E 1.2 1.2 Not Detected Not Detected aPurified DNA from rats was digested with Eco R1 (0.15 units/pg DNA) for 2 hours, 37°C. bDNA fragments as noted in Figure 14. CRats were injected i.p. with 15 mg/100 g N—2-acetylaminof1uorene 24 hours prior to sacrifice. dRats were injected i.p. with 0.75 m1/100 9 corn oilzDMSO (6:1, v/v) 24 hours prior to sacrifice. eData expressed in terms of kilobases. 111 .9099- pm mxoo m omsoosmopoos -opoo msmz mpopo omNTopsoxz .mmmooopps op ompoopocw mso mcmm cwsoopo pos soe <29o omppmoop 2pp3 omppmoop mpcmsmosm coppopspmms to mmNpm .m oco o mmcoo op ompcmmmsoms mp <29 pospooo Nmeo coppmmmao p :92 .o ooo o .o mmcoo cw ompommmsoms ma mpos ompomsp to HHz oco H2 Eose <29 to ooppmmmao p :92 .m use o mmcmo op ompommmsoms we Amxmo m sot .9929 "ppo csoo FE mn.9v mpos pospcoo-m_ow2m> Eosm <29 to coppmmmpo ’2 com com .m ooo o .o .o mmcoo 9p ompcmmmsoms mp p2 com 29 mpos ompomsp to sz oco H2 Eosw <29 mo 9: 9a to coppmm9p9 .Aop ocm m— mmsompd .mocmmmp mmmv ompoowocp mo omNpopsozc oco omssmmmcosp cmcp .ommmsogoospompm oco p :92 so pm oom 2o ompmmmpo mo; <29 .Amoocpmz mmmv cpmposo ocm <29 mo mmsp ompepsoo mm; zoom wo <29 oco .ompopomp msmz Appzv mppmo poszcocmsooco: ooo Apzv mppmo posxgocmsoo oppooms co pmpooz .mxoo m soc Am 99_\9E 9pc e<< opp: .o.p ompomwcp msmz mpos 2mp3m9umomosom mpoz .mpop2m> opp: so A9<pp to cowpmmmwo p :92 com _9 oom .op msomwd 19 112 up ossops __ z. _z __ z . _z — cav- E. Ouw 113 .mmmoooppx mo mEsmp op ommmmsoxm opo9m .mopwpsoom op sopso mxoo 9 soe 2>\> .pnov omzouppo :sao 9 oop\pe ms.o :pp: .a.p aopoohep osoz mpa: .mopepsoom op sopso mzoo 9 sow momsoopmocpEopxpmomiNiz 9 99_\9E mp 2pp3 .o.p ompomwop msmz mpomo o .9_ new 9_ mmsz9p9 op omppmoop mo mp29pm3 sopzompoz .oonm pa aseoe N so: 2:29 9: _\app:: :9 H :as so 2:29 a: p \mppee m_.ov H: com app: aapmoape ma; apes sass <29 ampepseaa 9 9.N 9.N 9.N ompompmo poz d 9.9 9.N 9.N 9.N 9 9.9 9.9 9.9 9.9 9 N.e 9.9 N.< N.o 9 9.9 N.9 ompompmo poz 9.9 ompompmo poz 9 9.9— ompompmo poz 9.9_ ompompmo poz < H cox 9.N 9.N N._ N._ 9 9.N 9.N ompompmo poz ompompmo poz 9 9.9 9.9 9.9 9.9 9 9.9 9.2 9.9 m9.9 9 ompompmo poz ompompmo poz ompompmo poz ompompmo poz < Hm oom pospcou ompomsh Pospcou oompomsh o opcms9msd <29 sz Hz momsozpeoopEopzpmoppom9mms .m 9:9 9 mm:92 :w <29 9o p::o59 m:p moezp p:mmms9ms 9 9:9 9 mm:92 .9 .9 .m 9:9 o .9 .9 mm:92 :p 9mp:mmms9ms mp p9 oo9 99 p9s 9mp9msp mo pmpoo: 992 9:9 Hz Eosm <29 99 99 to :oppmm9p9 .Ae— 9:9 9p mms99p9 op m9:m9mp mmmv 9mp99p9:p m9 9mNp9psoa: 9:9 9mssmpm:9sp :m:p .9mmmso29ospompm 9:9 _ :92 so p9 com a: 9mpmm9p9 993 <29 .Am9o2pm2 mmmv 9mwmpsz9 <29 9:9 .9mp9Fomp msmz Aszc memo p9ex:o:ms99:o: 9:9 292V mppmo P9592ocms99 opp99m: co wm_o:2 .9999 9 so» A>\> puo .9929uppo :soo 9 99_\_E 99.9v mpo92m> so A9 99_\9E mpv 9<< :ppz .9.p 9mpom9:w msmz mp9s xmpz99imo99s99 mp9: .mpop:m> :ppz so A9<pp mo :oppmm9w9 F :92 9:9 p9 oom .NP ms99p9 115 my U 99.9 I o9 l 9.9 l 9.9 I 9.9 I... . 9.91... ....“ 9.991 . 3. I __z pp osaap: ..a... 98:5 mail! a .wgmniii. .9 mount“ ...... g __ as _ as 116 2>\> .pu9v 99:9 poo eseo 9 99p\pe 99.9 cps: .9.p 99po9999 asp: mp9: .99 9:9 99 mms99p9 :9 9mppmo9F 99 9p:9pm3 s9_:om_o2 .mm99oopp2 mo mesmp :9 9mmmms9xm 9p99m .mopmwso9m op sops9 9999 9 so: 9 .mowepso9m op so9s9 9999 9 so: m:mso:_eo:pE999pmo9iNuz 9 999\9E mp :ppz .9.9 9mpom9:p msmz 9p99o 9 .9099 pa asso: 9 see 2929 9: p299999 99 9 :92 so 2929 9: p \99999 9_.99 92 oo9 :poz 99p99999 99: mp9s ease <29 ampeps999 99.9 espoopae poz 9.N aopoopoe poz s 9.9 aopoopoa poz ompoapae poz empoopoa poz 9 :.9 aopompoe poz empoapoe poz empoopaa poz 9 9.9 ampoapae poz 9.9 9.9 o 9.9 99989999 992 9.9 aopoopoe poz 9 9.9, 9.9_ 9.99 9.9_ 9 9.99 9.99 9.99 9.99 9 992 99._ 99._ empoapoa poz 9 --- --- 99.9 99.9 9 ampoapae poz espoapee 992 9.9 9.9 o 9.9 9.9 9.9 9.9 9 9.9 9.9 9.9 9.9 9 99.9 9 99 oo9 pospco9 9mp9msh Fosp:o9 9mp9msh 9 o 9p99999ss <29 9:2 92 m:mso:_eo:pE9pxpmo9> .9999 9929nppo osoo 9 99999: 99.9 9992 .9.9 oopoo9ep oso: 99999 .mopepso99 op sops9 9999 9 sow mcmsozpeo:pE9_9pmo9nNiz 9 99999: 99 :p93 .9.p 9mpom9:p msmz 9p9mo .99 9:9 99 9ms99p9 :9 9m__m99— 99 9p:9pm3 s9poompo2 899 pa 9s9o: 9 so: 9929 9: p99p999 99 .9 o H :92 so 9929 9: 9 9999:: 9_.99 H: oo9 goo: oopmo9oo 99: apos eoss <29 oopspsaso 9.9 9.9 99.9 99.9 9 9.9 9.9 99.9 99.9 9 9.9 9.9 99.9 99.9 o 9.9 9.9 oopoopoo poz oopwopoo poz 9 oopoopoo poz oopoopoo poz oopoopoo poz oopoopoo poz < 99 oo9 Fosp:o9 9mp9msh _ospcoo o9mpoms9 9 op:mE99s9 <29 :92 Hz m:mso999o:pE9_9pmo<-N- z 299: 9999 9 oopoosp 9p99 Eosm <29 to :o9p9p:m599s9 m99m_99:09:m coppUWspmma H :92 9:9 He com 9 92: 121 TABLE l9 Effect of Daily N-2-Acetylaminofluorene Treatment Followed by 7 Days Without Treatment on Liver/Body Weight Ratiosa Day Treatedb Controlc l 0.050:0.002 3 0.050:0.003 5 0.05l:0.005 7 0.045:0.007 0.045:0.00l aDetermined as liver weight/body weight on day of sacrifice. bMale, Sprague-Dawley rats were injected i.p. with l5 mg/lOO g N-2-acetylaminofluorene in corn oil: DMSO (6:l, v/v) for l, 3, 5 or 7 days, followed by 7 days without treatment. CMale, Sprague-Dawley rats were injected i.p. with 0.75 ml/lOO 9 corn oilzDMSO (6:1, v/v) for 7 days followed by 7 days without injection. l22 .9099. 99 9999 m so9 9oco9s9o999so999 osoz 99o9o 999999s99: .99xv 99999o99x :9 999 999999999 99999 .co9u9om99 9 oax 99999 9-9 99:99 999 .oo9999999 9m oom sop99 9-9 99:99 :9 9999o99c9 99 999s 9op9ogpnm<< Eoom 99999: 992 999 Hz Eosm <29 9o :o9p9om9o .999 9:9 m9 9999999 o9 9999999 999v 999999so9; 9:9 9oss99999s9 99:9 .9o9oso99osuoo9o 9:9 9 cog so 99 oom 99 99999999 993 999 ovv <29 .999ocpoz 999v 99999s99 993 <29 9:9 .99999o99 osoz 999o>99ooo9os .99999 9959cocos9aco: 9:9 9959cocos99 Eosm 9o>9so9v 99999: 992 999 Hz .9999 9 Lo9 999E99os9 99o9993 9oo999999e 99:9 .99 009995 99v 9<< £993 .9.9 9opooho9 osoz 999s 999399-9999som 999: .999299ms9 99o9992 9999 9 9o 993o99o9 99<99 9o :o9p9om99 9 sax 9:9 99 oom .99 ms9999 123 99 as9999 124 .9999. 99 9999 9 so» 9ozo9gmo999souo9 oso3 9po9o 99999 -9s99: .9oxv 99999o99x c9 os9 9ou9o99c9 99999 .999299999 sax 9:93o99o9 9-9 99:99 999 .9992 -99999 99 oom 9:93o99o9 9-9 99:99 99 999999999 99 999s 9op9ospum<< 9o 9o9o9: 992 9:9 92 Eos9 <29 99999999 .999 999 m9 9ms9999 o9 9999999 999v 9op9o9999 99 999999so9; 9:9 9oss99999s9 99:9 .9o9osogoospoo9o 9:9 9 any so 99 oom 99 99999999 993 99: 09v <29 .999ogpoz 999v 99999s99 993 <29 999 .99999o99 oso3 999o>999999ms .999oo 9959socos999o: 9:9 9959cocos9o Eos9 99>9so9v 9o9o99 992 999 Hz .9999 9 s99 9:99999s9 9999993 9999999999 9999 .99 99_\99 99v 999 9993 .9.9 99999999 mso3 999s 999399-9999s99 9992 .999599999 99o9993 9999 9 9o 9o3o99o9 99<99 9o co9p9om99 9 cox 9:9 99 com .09 os9999 125 99 9s9999 0.0...ro’.c \o...v. ..99 . . 9. C J. o \ . o. .. . m. 9 9 a o .. 126 .9909. 99 9999 9 so» 99:99s9o999so999 oso3 99o9o 999999so9z .onv 9999oo99x c9 os9 999999999 99999 .9.9 99:99 :9 999co9osaos 99 9 :99 so 99 oom 99 99999999 999 9<< 5993 9999os9 999s 9o 9o9o9: 992 999 92 Eosw <29 .999 999 m9 9ms9999 o9 9999999 999v 9999o99o9 99 999999s99; 9:9 99sso99c9s9 99:9 .9o9wso99os9oo9o 9:9 9 :99 so 99 oom 99 99999999 993 99: 09V <29 .999og9oz 999V 99999s99 993 <29 9:9 .99999o99 oso3 999m>99o999os .999oo 9929cocos999o: 9:9 99s9gocos99 Eosm 9o>9so9v 99999: 992 999 Hz .9999 9 s99 9999999s9 9999993 9999999999 9999 99 999999 999 999 9993 .9.9 99999999 oso3 999s 999399-9999s9m 9992 .999599os9 99oz993 9999 9 99 9o3o99o9 99<99 9o co999om99 9 sax 999 9a oom .99 os9999 127 99 9s=999 128 .9999. 99 9999 m so» 99:99smo999so999 oso3 99o9o 9o9999s99z .onv 9999oo99x :9 os9 9999o99c9 99999 .9 9:9 9 .o 99:99 :9 :3oc9 99 999s 9os9coo sosw <29 9o99mm99 9 cog 9:9 9m com .9 9:9 9 .o .9 99:99 :9 :3o99 99 :o9999999 9 :9x so 99 oom 9:93o99o9 999s 9999os9um<< 9o 9o9o9: 992 9:9 92 Eosm <29 .999 9:9 99 9os9999 o9 99:9999 999V 9999o99o9 99 999999so9: 9:9 9oss999:9s9 :999 .9o9os -ogoos9oo9o 9:9 9 :ox so 99 oou 99 99999999 993 99: 09v <29 .999o99oz 999V 99999s99 993 <29 9:9 .99999o99 mso3 999o>99ooa9os .999oo 9929co:os99:o: 9:9 9959go:os99 Eosw 9o>9so9v 99999: 992 999 Hz .9999 9 s99 9999999s9 9999993 9999999999 9999 99 99999E 99v 999 9993 .9.9 99999999 9s93 999s 999399-9999s99 9992 .9:oE99os9 99og993 9999 9 9o 9o3o99o9 o9o9go> so 99<99 9o :o999om99 9 :99 9:9 99 oom .99 os9m99 129 mm mssm9u 130 TABLE 20 Eco R1 and Kpn 1 Restriction Endonuclease Fragmentation of DNA from Rats Treated 1 Day with N-2-Acety1aminofluorene Followed by 7 Days Without Treatmenta NI NII DNA Fragment TreatedC Treated Eco R1 9.7d A 5.4 Not detected B 3.90 4.0 C 2.9 3.2 D 2.35 2.35 E Not detected Not detected Kpn 1 17.5 17.5 A Not detected Not detected 8 Not detected Not detected C Not detected 5.2 D Not detected 3.4 E Not detected 2.5 F Not detected Not detected aPurified DNA from rats was digested with Eco Rl (0.15 units/l ug DNA) or Kpn 1 (8 units/1 ug DNA) for 2 hours at 37°C. bMolecular weights as labelled in Figures 14 and 15. cRats were injected i.p. with 15 mg/100 g N-Z-acetylamino- fluorene for 1 day followed by 7 days without treatment. d Data expressed as kilobases. 131 TABLE 2] Eco R1 and Kpn 1 Restriction Endonuclease Fragmentation of DNA from Rats Treated 3 Days with N-2-Acetylaminof1uorene Followed by 7 Days Without Treatmenta b NI NII DNA Fragment TreatedC Treated Eco R1 A Not detacted Not detected B 4.0 4.0 C 3.0 3.0 D Not detected Not detected E 1.9 1.9 Kpn 1 19.5 19.5 A Not detected Not detected 8 Not detected Not detected C Not detected Not detected D Not detected Not detected E Not detected Not detected F Not detected Not detected aPurified DNA from rats was digested with Eco R1 (0.15 units/l ug DNA) or Kpn l (8 units/l pg DNA) for 2 hours at 37°C. b Molecular weights as labelled in Figures 14 and 15. CRats were injected i.p. with 15 mg/100 g N-Z-acetylamino- fluorene for 3 days followed by 7 days without treatment. d Data expressed as kilobases. 132 TABLE 22 Eco R1 and Kpn 1 Restriction Endonuclease Fragmentation of DNA from Rats Treated 5 Days with N-2-Acetylaminofluorene Followed by 7 Days Without Treatmenta b NI NII DNA Fragment Treatedc Treated Eco R1 A 4.5“l 5.7 B 3.8 3.8 C 2.8 Not detected D 2.4 2.4 E 1.9 1.85 Kpn 1 19.5 21.0 A Not detected Not detected B Not detected Not detected C Not detected Not detected D Not detected Not detected E Not detected Not detected F Not detected Not detected aPurified DNA from rats was digested with Eco R1 (0.15 units/1 pg DNA) or Kpn l (8 units/1 ug DNA) for 2 hours at 37°C. bMolecular weights as labelled in Figures 14 and 15. cRats were injected i.p. with 15 mg/100 g N-2-acetylamino— fluorene for 5 days followed by 7 days without treatment. dData expressed as kilobases. 133 .mmmmnopwx mo mscw“ cw ummmmcaxm «poem .pcmeummgp uzoguwz mxmu u an umzopfioy meu n so» me oop Lmq A>\> .Fuov omzonpwo :goo PE mn.o saw: .a.w uwpummcw mew: mpmm u .pcmsummgp pzocwwz mzmu u »a nmzo~pom mama N com mcmcozpmocwsmfiaumum-muz m oop\ms m_ saw: .a.w umpomncw mng mummo .m_. 9.6 .3. mm&:m_...._ cw UmP—wnm— mm mwcmwmz przom—OZ a .UONM on maze; N Loo A Lo wcmgoa—mocwsmpxumu