MSU LIBRARIES .3— RETURNING MATERIALS: P1ace in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. EES‘VIVO DIVERSE SPECTRUM OF NEOPLASMS INDUCED IN 30-DAY-OLD RATS AND IN VITRO NEOPLASTIC TRANSFORMATION OF RAT MAMMARY EPITHELIAL CELLS FOLLOWING A SINGLE PULSE OF N-ETHYl-N-NITROSOUREA BY Gheorghe Stoica A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Doctor of Philososphy Department of Pathology 1984 @>1985 GHEORGHE STOI CA All Rights Reserved ABSTRACT IN VIVO DIVERSE SPECTRUM OF NEOPLASMS INDUCED IN 30-DAY-OLD RATS AND £1! VITRO NEOPLASTIC TRANSFORMATION OF RAT MAMMARY EPITHELIAL CELLS FOLLOWING A SINGLE PULSE OF N-ETHYL-N-NITROSOUREA BY Gheorghe Stoica The specific aim of these studies were to determine the i_rl vitro N-Ethyl-N-Nitrosourea (ENU)-induced spectrum of neo- plasms and the in vitro ENU-induced rat mammary epithelial cell neop l as tic trans formation . I. E Vivo Project. Male “and female 30-day-old Sprague- Dawley (CD) and Berlin Druckrey (BD-IV) rats were inoc- ulated intraperitoneally with a single dose of 45, 90 and 180 mg/kg of ENU. Previous experiments demonstrated that a single high dose of ENU (180 mg/kg i.p.) adminis- tered to 30-day-old female CD rats was capable of pro- duCing mammary tumors (MTs) in 100% of the animals with a 95% malignancy rate. The high rate of malignancy, lo— cal invasion and metastases to the lung were the parti- cular features of this model. The first two parts of the in vivo ”project were a continuation of further char- acterization of this animal model for human breast can- cer. To study the .spectrum of neOplasms and the inci- dence of MTs in the absence of ovarian hormones stimula- lation, CD male rats have been inoculated with similar doses of ENU as female CD rats. ENU induced mammary II. Gheorghe Stoica “tumors in 46% of the 30-day—old CD male rats with 3.3% lung metastases. This study also revealed that ENU in— ethyl > higher homologues 2) Preferential Reactivity: alkyl sulfate-oN N-nitroso—é O alkylalkane sulfonate —9N > O 3) Possible Sites of Reactions: All rings, nitrogens and oxygens (except sugar attachment sites) phosphodiester 2'-O of ribose Figure 1-1. General condlusions on the reactivity of direct acting alkylating agents toward both RNA and DNA. N indicates nitrogens and 0 represents oxygens.20 Studies of ethylation of nucleic acids with ENU revealed that approximately 80% of the total ethyl groups were on oxy- 27,28 gens. The predominant products were ethylphosphotri- esters (50-60%). All other oxygens were also modified. In addition to 06-EtG, the products included OZ-EtC, OZ-EtT (or U), and 04-EtT (or U). In the case of of RNA, the 2'-O or ribose was also ethylated.29'30 In their now classic experiments, Goth and Rajewsky ad- ministered (14C-ethyl) ENU to lO-day-old BD-IX rats under conditions that ultimately led to virtually all the animals developing tumors of the brain but no other tissue.16 They then isolated DNA from the brain and other tissues at various times after the single treatment with the carcinogen. They reported that OG-EtG was not removed rapidly from the DNA of the brain (target organ) whereas it was greatly reduced in other organs such as the liver. From this observation they evolved the concept that it was not the extent of alkylation of the O6 of G that was specifically important in carcinogene- sis, but rather the fact that O6-EtG was not removed enzyma- tically from a given cell population.21 Mutagenic Reactions of Alkylating Agents.11 Mutagenesis is the result of a change in transcription of a nucleic acid that can be a naturally occurring event and thus has probably contributed to biologic evolution. Mutation can occur with- out chemical modifiation when the polymerase that directs transcription has, for any reason, a high error rate, and the resulting mistakes are not corrected by a proofreading enzyme. Inasmuch as genetic change normally occurs at a very slow ‘ rate, it may be assumed that transcription is generally faith- ful. When a chemical modification is introduced, several types of effects are possible. 1) Transcription may be un- affected, as seems to be the case when N-7 alkyl G or N6-al- kyl A is present.3l’32~ 2) Transcription may be stopped by the modified base not being recognized at all, which is likely to result when substituents are introduced on sites involved in base pairing or when the substituent is bulky and steric factors play a role.33 3) Modifications can be mutagenic if the tautomeric equilibrium is changed by the modification and the modified base thus has the base-pairing properties of another base.34’35 4) A potential mutagenic modification can also be either lethal or have no effect, depending on whether the modification occurs at a site essential or non- essential for protein or nucleic acid structure or function. The major derivative formed upon ENU reaction with nucleic acids is ethylphosphotriesters. No evidence has thus far been obtained which suggests that this is a mutagenic action.11 All the in yitrg experiments on mutations resulting from alkylation of nucleosides can be summarized as follows: No predictions of base pairing can be made on the basis of chem- ical structure. The observed mispairing is not specific, but neither is it totally nonspecific. The derivatives that have been shown to mispair are substituted on the N-3 of C and U, the o2 and o4 of U (or T), the o2 of c, and the 06 of c.11'21 Repair of Alkleerivatives: Kineticsngossible Mechanisms and Biologic Consequences. Before postulating repair and its mechanisms, one must know and consider the chemical stability of the alkyl products.in DNA under physiologic conditions. It has long been known that 3- and 7- alkyl purine nucleosides are easily depurinated from DNA due to the great lability of their glycosyl bonds. . Figure 1-2 summarizes some of the data and conclusions regarding the stability of alkyl derivatives of DNA. Cells differ in their ability to remove modified nucleo- sides,36 and there have been attempts to correlate excision ability with enzyme activities. It has been postulated that enzymes exist that recognize conformational changes resulting from modifications of nucleic acids.11 Another enzyme acti- vity has been found in cell extracts that seems to excise 06- 37,38 alkyl G, but not 7-alky1 G, and the lack of enzymatic excision of OG-alkyl G has been correlated with the tumori- 39 ggenic potential of mammalian cells. With regard to bacteria, ;E, coli cells usually have limited capacity for removing 06- Stability of Alkyl Derivatives in DNA 1. In Vitro pH 7, 37C a) 0: alkyl G no dealkylation 02 alkyl T glycosyl bond break- 04 alkyl C age detected in 10 0 alkyl T days b) N-3 alkyl A glycosyl bond cleav- age t%=26 hours c) N-7 alkyl G glycosyl bond cleav- age tg5 = 155 hours d) Stability of glycosyl bond much greater in poly- nucleotides than in mononucleotides (t%=N-7 alkyl G in monomer = 6 hours). 2. In‘Vivo pH 7, 37C a) b) C) d) All alkyl bases excised (dealkylated); phospho- triesters stable. Loss of alkyl derivatives occurs most rapidly in first day; it varies greatly, depends on period observed. Complete excision (or dealkylation) not observed for O-alkyl derivatives after 5 days. There is a strong possibility that multiple enzy- mes are responsible for loss of alkyl bases. Figure 1-2. General conclusions regarding the stability of alkyl derivatives in DNA, under physiologic conditions. Alkyl refers to methyl or ethyl substituents. tL5 = half-life. methylguanine from DNA, but an inducible repair function is expressed after the exposure of cells to low concentrations of alkylating agents. This repair pathway, termed the adap- tive response, allows rapid and error-free repair of Oevmethyl -guanine residues.l Olsson and Lindahl4O has established that the methyl group is transferred to a cysteine residue on the enzyme protein and that guanine is thus generated in the DNA substrate. Rapid assay of the 06-methylguanine DNA trans- methylase using antibodies specific for 06-methylguanine to precipitate this base was described.41 The mammalian 06- methylguanine-DNA-transmethylase was also able to remove (at a slower rate) ethyl groups from OG-ethylguanine in DNA form- ing S-ethyl-cysteine, but did not attack any of the other methylated bases present in DNA methylated with MNU.41 The mammalian 06-methylguanine methyltransferase no longer seems active in some human tumor cell lines, in particular in lines from cells transformed with DNA tumor viruses, whereas normal diploid fibroblasts and other control cells express this re- pair function.42 The N-nitroso alkylating agents, particularly ENU, have an affinity for modifying all oxygens in nucleic acids. It may well be that the formation and fate of 06-alky1 G and other O-alkyl derivatives formed by the potent carcinogenic N-nitroso compounds are fortuitously good indicators of addi- tional reactions contributing to cancer. Activation of Oncogenes by N-Nitroso Compounds. Detection and isolation of transforming genes (oncogenes) from human tumors have provided the means to investigate, at the mole- 10 cular level, events involved in the development of human neo- plasia. A broad spectrum of human cancer contain such trans- forming genes, a group of at least three structurally differ- ent genes, H-ras, K-ras, and N-ras, which code for highly re- lated proteins designated p 21. Sukumar et al22 demonstrated that a single point mutation was responsible for the malig- nant activation of the H-ras-l locus in MNU-induced mammary carcinomas in Buf/N female rats. Molecular characterization of one of the genes revealed that the twelfth codon was GAA instead of GGA of the normal allele, encoding glutamic acid instead of glycine.22 These results indicate that chemical carcinogenesis represents an adequate model to study the role of transforming ras.genes in human neoplasia. 11 REFERENCES l. Druckrey H, Ivankovic S, Preussman R: Teratogenic and carcinoqenic effects in the offspring after single injec- tion of ethylnitrosourea to pregnant rats. Nature 210:1378- 1379, 1966. 2. Koestner A, Swenberg JA, Weschler W: Transplacental production with ethylnitrosourea of neoplasms of the nervous system in Sprague-Dawley rats. Am J Pathol 63:37-56, 1970. 3. Jones PL, Searle CE, Smith-Thomas W: Tumors of the nervous system induced in rats by the neonatal administration of N-ethyl-N-nitrosourea. J Pathol 109:123-139, 1973. 4. Koestner A, Swenberg JA, Weschler W: Experimental tumors of the nervous system induced by resorptive N-nitro- sourea compounds. Recent Advances in Brain Tumor Research, Progress in Experimental Tumor Research, Homburger Fed. 7:9- 30, 1972. 5. Stoica G, Koestner A, Capen C: Spectrum of neoplasms induced with single high doses of N-ethyl-N-nitrosourea in 30-day-old rats with special emphasis on mammary neoplasia. AntiCancer Research 4:5-12, 1984. 6. Stoica G, Koestner A, Capen C: Characterization of N-ethyl-N-nitrosourea-induced mammary tumors as an animal model for human breast cancer. Am J Pathol 110:161-169, 1983. 7. Stoica G, Koestner A: Diverse Spectrum of tumors in male Sprague-Dawley rats following single high doses of N-ethyl-N-nitrosourea. Am J Pathol 116:319-326, 1984. 8. Stoica G, Koestner A, Capen C: Testicular-like tumor in the ovary of rats induced by chemical carcinogens. In: preparation. 9. Magee PN, Swann PF: Nitroso compounds. Brit Med Bull 25:240, 1969. 10. Mandybur TI, Ormsby I, Buncher CR: Enhanced develop- ment of mammary tumors in rats following transplacental and neonatal exposure to ethylnitrosourea. Cancer Res 38:3182- 3185, 1978. 11. Singer B: The chemical effects of nucleic acid al- kylation and their relationship to mutagenesis and carcinogen- esis. Prog Nucleic Acids Res Mol Biol 15:219-284, 330-332, 1975. 12. Singer B, Fraenkel-Conrat H: The reaction of nitro- soguanidine (n-methyl-N'-nitro-N-nitrosoguanidine) with to- bacco mosaic virus and its RNA. Science 160:1235-1237, 1968. 13. Singer B, Fraenkel-Conrat H: The role of conforma- tion inwchemical mutagenesis. Prog_Nuc1eic Acids Res Mol Biol 9:1-30, 1969. 14. Lawley PD, Thatcher CJ: Methylation of deoxyribonu- cleic acid in cultured mammalian cells by N-methyl-N'-nitro- N-nitrosoguanidine. Biochem J 116:693-707, 1970. 15. Loveless A: Possible relevance of 0-6 alkylation of deoxyguanosine to the mutagenicity and carcinogenicity of nitrosamines and nitrosamides. Nature 223:206-207, 1969. 16. Goth R, Rajewsky MF: Molecular and cellular mecha- nisms associated with pulse-carcinogenesis in the rat nervous system by ethylnitrosourea: Ethylation of nucleic acids and elimination rates of ethylated bases from the DNA of differ- ent tissues. Z Krebsforsch; 82:37-64, 1974. l3 l7. Lawley PD, Warren W: Specific excisions of ethyla- ted purines from DNA of Escherichia coli treated with N-ethyl -N-nitrosourea. Chem Biol Interact 11:55-57, 1975. 18. Kleihues P, Magee PN: Alkylation of rat brain nu- cleic acids by N-methyl-N-nitrosourea and methyl methane-sul- fonate. J Neurochem 20:595-606, 1973. 19. Kleihues P, Margison GP: Carcinogenicity of N-methyl -N-nitrosourea: Possible role of excision repair of 06-methyl guanine from DNA. J Natl Cancer Inst 53:1839-1841, 1974. 20. Magee PN: Evidence for the formation of electrophil- ic metabolites from N-nitroso compounds. In: Origins of Hu- man Cancer, Book B, Mechanisms of Carcinogenesis. Ed. Hiatt, Watson, and Winsten, Cold Spring Harbor Laboratory, pp. 629- 637, 1977. 21. Gerchman LL, Ludlum D: The properties of 06-methyl -guanine in templates for RNA polymerase. Biochim Bithys Acta 308-310, 1973. 22. Sukumar S, Notario V, Martin-Zanca D, Barbacid M: Induction of mammary carcinomas in rats by nitroso-methylurea involves malignant activation of H-ras-l locus by single point mutations. Nature 306:658-661, 1983. 23. Goth R, Rajewsky M: Perspectives of O6-ethy1guanine in rathbrain DNA. Nature 269:625-626, 1977. 24. Margison GP, Kleihues P: Chemical carcinogenesis in the nervous system. Preferential accumulation of 06-methy1 -guanine in rat brain DNA during repetitive administration of MNU. Biochem J 148:521-525, 1975. 14 25. Frei JV, Lawley PD: Tissue distribution and mode of DNA methylation in mice by methyl methanesulfphonate and N-methyl -N’-nitroso-guanidine. Lack of thmic lymphoma induction and low extent of methylation of target tissue DNA at O6 of guanine. Chem Biol Interact 13:215-222, 1976. 26. Singer B: N-nitroso alkylating agents: Formation and persistence of alkyl derivatives in mammalian nucleic acids as contributing factors in carcinogenesis. J Natl Can- cer Inst 62:1329-1339, 1979. 27. Singer B: All oxygens in nucleic acids react with carcincinogenic ethylating agents. Nature 264:333-339, 1976. 28. Singer B: Sites in nucleic acids reacting with al- kylating agents of differing carcinogenicity or mutagenicity. J Toxicol Environ Health 2:1279-1295, 1977. 29. Singer B, Kusmierek JT: Alkylation of ribose in RNA reacted with ethylnitrosourea at neutrality. Biochemistry 15:5052-5057, 1976. 30. Kleihues P, Lantos PL, Magee PN: Chemical compounds in the nervous system. Int Rev Exp Pathol 15:153-232, 1976. 31. Ludlum DB: The properties of 7-methylguanine contain- ing templates for ribonucleic acid polymerase. J Biol Chem 245:477-482, 1970. 32. Engel JD, von Hippel PH: d(m6ATP) as a probe of the fidelity of base incorporation into polynucleotides by Escherichia coli DNA polymerase I. J Biol Chem 253:935-939, 1978. 33. Leffler S, Pulkrabek P, Grunberger'D: Template acti- vity Of calf thymus DNA modified by a dihydrodiol epoxide deri- vative of benzo(a)pYrene. Biochemistry 16:3133-3136, 1977. 34. Budowsky EI, Sverdlov ED, Spaskokukoiskaya TN: Tauto- merism of.nucleic acid bases and the template synthesis of poly- nucleotides. FEBS Lett 17:336-338, 1971. 35. Taniguchi T, Weissman C: Site-directed mutations in the initiator region of bacteriOphage QB cistron and their effect on ribosome binding. J M01 Biol 118:533-565, 1978. 36. Strauss BS, Karren P, Higgins NP: Alkylation damage and DNA excision repair in mammalian cells. J Toxicol Envi- ron Health 2:1395-1414, 1977. 37. Pegg AE, Hui G: Formation and subsequent removal of 06-methylguanine from deoxyribonucleic acid in rat liver and kidney after small doses of dimethyl nitrosamine. Biochem'J 173:739-748, 1978. 38. Pegg AE: Dimethylnitrosamine inhibits enzymatic re- moval of O6 -methylguanine from DNA. Nature 274:183-184, 1978. 39. Pegg AE: Formation and metabolism of alkylated nu- cleosides: Possible role in carcinogenesis by nitroso com- pounds and alkylating agents. Adv Cancer Res 25:195-269, 1977. 40. Olsson M, Lindahl T: Repair of alkylated DNA in Es: cherichia coli. Methyl group transfer from 06-methylguanine to approtein cysteine residue. J Biol Chem 255:10569-10571, 1980. 41. Pegg AE, Wiest L, Foote RS, Mitra S, Perry W: Puri- fication and properties of 06-methylguanine-DNA.transmethyl- ase from rat liver. J Biol Chem 258:2327-2333, 1983. 42. Lindahl T: -DNA repair enzymes. Annual Rev Biochem 51:78-81, 1982. CHAPTER 2 Diverse spectrum of tumors in male Sprague-Dawley rats following single high doses of N-ethyl-N-nitrosourea 16 17 SUMMARY In this study, 30-day-old male Sprague-Dawley rats, were inoculated intraperitoneally with a single dose of 45, 90 and 180 mg/kg of N-ethyl-N-nitrosourea (ENU). A wide spectrum of neoplasms occurred. The most common tumors were those of the mammary gland and of the nervous system. Although the inci- dence of mammary tumors was highest in the two high-dose groups (90 and 180 mg/kg ENU), the incidence of neurogenic tumors was highest in the 45 mg/kg dose group. Mammary tumor development led to early death and precluded development of tumors of the nervous system, which require a longer latency period. A variety of neOplasms of other organs have been as- sociated particularly with high doses of ENU, including amelo- blastic tumors, carcinomas of the thyroid, prostate, kidney, pancreas, intestine and lung, hemilymphatic tumors and sar- comas. It is concluded that large doses of ENU are capable of expanding the tumor spectrum of young male rats beyond the target organs generally affected with lower doses, as described in earlier reports. INTRODUCTION N-ethyl-N-nitrosourea (ENU) has extensively used as a transplacental neurocarcinogen capable of inducing neuro- genic tumors in nearly 100% of rats exposed prenatally under optimal conditions. Susceptibility increases with advancing gestation culminating at parturition time and declining post- natally. After 30 days of age it becomes difficult, if not impossible, to induce neurogenic tumors in rats with a single 18 dose of ENU}-4 In a recent study we used 30-day-old female Sprague-Daw- ley (CD) rats in an attempt to correlate molecular damage with tumor incidence in selected organs following single inocula- tion of high doses of ENU.5 A surprisingly high incidence of mammary tumors (MTs) (up to 100% in the 90 and 180 mg/kg dose groups) resulted in early death and precluded the development of tumors of the nervous system (a prime target organ), which require a longer latency period than MTs.6 These findings prompted us to repeat the experiment using 30-day-old male rats. In this report we describe the spectrum of neoplasms induced with high single doses of ENU in suscepti- ble young male CD rats. MATERIALS AND METHODS Carcinogen ENU crystalsa were freshly dissolved in phosphate/citrate -buffered saline at pH 4.2 (1 part buffer to 14 parts saline) before administration. The crystalline carcinogen was stored in a dessicator at -20 C. Animals Four groups of 30 rats (three experimental groups and one control group) were used (Table 2-1). All rats were 30-day -old male;Specific-pathogen-free (SPF) cesarean-delivered (CD) rats.b They were housed in plastic cages (two rats per cage) a Synthesized by Dr. MIng Chang, Veterinary Pathobiology De- partment, Ohio State University, Columbus, Ohio, b Charles River Laboratory, Portage, Michigan, 19 in temperature-controlled rooms (24 C) provided with 10 hours electric light per day. They were fed autoclaved Purina Lab Chow 5010 C and water ad libitum. Rats in the three experi- mental groups were inoculated intraperitoneally with freshly prepared ENU solutions in single high doses of 45, 90 and 180 mg/kg, respectively. The control group was inoculated with an equal volume of phosphate/citrate-buffered saline solution at pH 4.2. Rats were observed daily and weighed weekly after treatment. Most rats died or were sacrificed because of tumors of theunervous system, hemilymphatic system, or mammary glands or because of cachexia. A complete necrOpsy was performed shortly after spontaneous death or euthanasia of moribund ani- mals. The histologic sampling procedures were equal among all groups. Tissues were fixed in 10% buffered formalin, embedded in paraffin, sectioned at 4-6 u and stained with hematoxylin and eosin (H&E). Special stains, such as Masson's trichrome, periodic acid-Schiff (PAS), and Movat were used when indicated. Tumors of all organs were studied. Mammary neOplasms were classified according to the suggested classification of mam- mary tumors in rats by the World Health Organization.7’8 Mor- phologic criteria for malignancy were pleomorphism, layering of secretory epithelium, infiltration of surrounding tissue, high mitotic activity, anaplasia and metastases. Tissues for electron microsc0py were fixed in 4% glutar- aldehyde, postfixed in osmium tetroxide, and, after processing, embedded in Epon, cut with an ultramicrotome (l u sections), and stained with 1% toluidine blue. Blocks were selected for electron microscopic examination, and ultrathin sections were '1 20 contrasted with uranyl acetate and lead citrate. Survival times were analyzed by one-way analysis of variance. Signifi- cant differences between means were determined by the LSD (P =0005) o 21 RESULTS Incidence and Survival Time The incidence of tumors in all ENU treatment groups (80 -86%) was substantially higher than in the control group (7%) (Table 2-1). There was, however, no difference in either the tumor incidence or the number of tumors per rat among the three ENU dose groups. A significant difference existed in surviv- al time between the control group and treated groups and also among the treated groups. The mean survival time decreased with increasing doses of ENU (Table 2-1). Another difference among the three dose groups was the incidence of tumors of the mammary gland and nervous system (Table 2-2). The rats affec- ted with MTs per dose group were 7 (23%) in the 45 mg/kg group, 14 (46%) in the 90 mg/kg group, and 12 (40%) in the 180 mg/kg group. While 19% of the total number of MTs occurred in the 45 mg/kg group, 40% occurred in the 90 mg/kg group and 41% occurred in the 180 mg/kg group. The two highest ENU doses resulted not only in a higher incidence of mammary neOplasias but also in an increased incidence of carcinomatous and sar- comatous types of MTs. There was a direct correlation between dose of ENU and the percentage of malignant MTs (10%, 21% and 27% malginancy per dose group). The overall mean MT induction time was 140 days, with a mean of 105 days in the highest ENU dose group (180 mg/kg). The mean survival time of the rats. bearing malignant MTs was 280 days (:5) days. The Opposite was true for neurogenic tumors; 46% of the total number of neurogenic tumors developed in the low-dose group (45 mg/kg), 26% developed in the 90 mg/kg group, and 28% developed in the 22 Table 2-1. Number of tumors and survival time in male Spra- gue-Dawley rats treated with ENU intraperitone- ally at 30 days of age. No. of rats with tumors Mean Mean Dose (30 rats per group; number survival time of percentage in of tumors (days) ENU ‘parentheses) Aper rat ‘ (mean + SEM) 0 2(7) 0.1 491 + 4 45 mg/kg 24(80) 1.7 425 + 16 90 mg/kg 26(86) 1.2 367 + 19 180 mg/kg 24(80) 1.6 294 + 14 23 180 mg/kg group. The number of rats per group affected by neurogenic tumors was 20 (41%) in the 45 mg/kg group, 14 (28%) in the 90 mg/kg group, and 15 (31%) in the 180 mg/kg group. The mean survival time of rats bearing neurogenic tumors per dose group was 447 days for 45 mg/kg group, 386 days for 90 mg/kg group, and 350 days for 180 mg/kg group. Histologic Classification The histologic classification and the number of various types of neoplasms are listed in Table 2-2. Neurogenic Tumors. .Tumors of the nervous system are simi- lar to those described in detail in previous publications.]'-4 In addition to the neurogenic tumors mentioned (Table 2-2), 10% and 15% of the rats from the high-dose groups developed alterative changes such as focal malacia and gliosis. Mammary Tumors. Benign MTs (42% of all MTs) were either fibromas or fibroadenomas consisting of fibrotic encapsulated masses with various degrees of intercalated glandular struc- tures (fibroadenomas). Malignant MTs were either fibrosar- comas, adenocarcinomas, or carcinosarcomas. Fibrosarcomas consisted of anaplastic fusiform fibroblastic cells with whorl formation and a high rate of mitosis (5-6 per high power field). One fibrosarcoma had metastasized to the axillary lymph nodes, mediastinum and lung. Adenocarcinomas (Figure 2-1) consisted of irregular ductular and acinar structures intercalated with a dysplastic stroma often heavily infiltrated with mast cells. The multilayered cells exhibited an increased nuclear-cytoplas- mic ratio, pleomorphism and a high rate of mitosis. Marked necrosis and cystic ductular structures filled with eosinophil- 24 Table 2-2. Incidence and classification of tumors induced with ENU in male Sprague-Dawley rats inoculated at 30 days of age. Neoplasias Central Nervous S stem Oligodendro- gliomas Astrocytomas Mixed gliomas Ependimomas Peripheral Nervous System Anaplastic neurinomas Mammaryigland tumors Fibromas Fibroadenomas Fibrosarcomas Adenocarcinomas Carcinosarcomas Thyroid gland Follicular car- cinomas Follicular ade- nomas C-cell adenomas Pituitary gland ChromOphobe ade- noma Bone Osteosarcoma Odontogenic tumor AmeIoBlastic Odontoma Pancreas Islet cell carcinoma Exocrinexade- nocarcinoma Prostate gland Adenocarcinoma Skin Hemangiopericy- toma Liver Hepatocellular Number of tumors per dose/group (30 rats per grqu) 45 mg/kg 18 1 NNNN (DIDHIUIO \l N Hum I I-‘ I-‘ N N I 90 mg/kgg 12 I.bl m N N [‘0 NNWIUJNH H IPJ +4 H Ir—‘I-I 180 mg7kg 11 9 2 huh» NIHNNNIF‘ N MN N l-‘I-‘I INI l-‘NNI Table 2—2 'cont'd) Number of tumors per dose/group (30 rats per group) Neoplasias O 45 mg/kg 90 mg/kg, 180 mg/kg carcinomas - l l - Hemilymphatic System - 2 3 Lymphosarcoma - 2 Myeloid leuke- mia - - 2 Spleen - Hemangiosarcoma Large Intestine Carcinoma - Kidney - Renal carcinoma Adrenal gland - Pheochromocytoma- Lung - Adenomas - Squamous cell carcinomas - - Peritoneum - - MesothelIEma - - - l MA I I I I—‘l—‘N I II-‘I-‘t—‘I-‘r-‘Hl I I NNI I II-J I—‘NI II-‘ Totals 2 54 46 53 26 Figure 2-1. Mammary tumor adenocarcinoma showing ducts with multilayered neoplastic epithelial cells (H&E, x 120) Figure 2-2. Mammary tumor adenocarcinoma metastases to the lung exhibiting vascular neoplastic emboli (ar- rows). (H&E, x 120) 27 Figure 2-1 V‘.“' ' ‘ ,‘l ingl< L in . slab . “ hm E! a Figure 2-2 ic material were often associated with this type of tumor. Two of the rats with adenocarcinomas (one each in the 45 and 180 mg/kg ENU dose groups) had lung metastases characterized by multiple epithelial emboli and tumoral masses invading the lung parenchyma (Figure 2-2). The carcinosarcomas were char- acterized by malignancy of both epithelial and mesenchymal components. In addition, % of the rats from the treated groups free of mammary neoplasms developed mammary gland hy- perplasia. Electron micrographs of carcinoma cells exhibited the characteristic pattern of mammary epithelial cells. The cells consisted of a central nucleus with a large nucleolus. The free margin had numerous microvilli and desmosomal junctions. Free ribosomes and tubules of rough-surfaced endOplasmic reti- culum (RER) were widely distributed throughout all parts of the cell. Some of the tubules formed cisternae. Numerous electron-dense secretory granules and a moderate number of lipid bodies occurred intracytoplasmically and within inter- cellular spaces. Ameloblastic Odontomas. These tumors were detected in two rats (180 mg/kg ENU dose) with a survival time of 280 days. The tumors contained many developing teeth of various sizes and shapes resembling normal or atypical tooth germs, with enamel and cementum in varying quantities and, in addition, epithelial and mesenchymal odontogenic tissues with scattered island of bone and osteoid (Figure 2-3). Hepatocellular Carcinomas. These tumors were encountered in 2 rats (45 and 90 mg/kg ENU dose groups) with a survival Figure 2-3. Ameloblastic odontoma showing irregular struc- tures consisting of mesenchymal odontogenic tis- sue (O), dentin (D) and cementum (C). (H&E x 120) Figure 2—4. Hepatocellular carcinoma, (H&E, x 48) trabecular pattern. 7 3w .¢ ea..? ,, ._ aw... in. . .191. ”x, .. .V.. . _. Riva . a. ... . . N. ., ¥<. 30 Figure 2-3 Figure 2-4 31 time of 415 and 320 days, respectively. The tumors had a tra- becular pattern with moderate necrosis, telengiectasia and fi- brosis (Figure 2-4). In addition, multiple nodules and vascu- lar emboli consisting of well-differentiated hepatoblasts were detected in the lung. In other liver-tumor-free rats, hepa- tic changes consisting of focal cellular alterations, nodular hyperplasia, spongiosis hepatis, focal necrosis, biliary hy- perplasia and cholangiofibrosis were observed in 20%, 23% and 30% of the treated group rats, compared with foci of cellular alteration in only 5% of the control group rats. Thyroid Tumors. These tumors occurred in all dose groups andwcontrols in equal number except for the 45 mg/kg group, where four times the number of the other groups occurred. The mean survival time of rats with thyroid tumors was 450 (:15) days. Only 3 of the 14 tumors (all in experimental groups) had characteristics of malignancy. They were diag- nosed as follicular carcinomas (2 in the 45 mg/kg and 1 in the 90 mg/kg ENU dose groups) and were characterized by a papillary pattern with invasion of the capsule and adjacent tissue. No metastases to other organs was detected. Tumors of the Pancreas. These were detected in 4 ENU- inoculated rats. The mean survival time of rats bearing these tumors was 320 (:10) days. Three were islet cell tumors re- cognized macroscopically as solitary nodules 0.5-1 cm in dia- meter. The tumors (one in the 45 mg/kg group and two in the 180 mg/kg group) were islet cell carcinomas characterized by the lack of a capsule, a well-defined neuroendocrine pattern, cellular anaplasia, and local invasion. Invasion of the sur- 32 Figure 2-5. Pancreas. Islet cell carcinoma. Aggregates of exocrine acini incorporated into the neoplasia are clearly apparent. (H&E, x 120) 2-5 igure F 34 rounding exocrine pancreas, with aggregates of exocrine acini incorporated into the neoplasia, was clearly apparent (Figure 2-5). Hyperplasia of the islet cell of the pancreas was recog- nized in 10% of the additional rats from the high-dose groups. One tumor was an adenocarcinoma of the exocrine pancreas, with a high degree of anaplasia and many mitoses. The exocrine pancreatic adenocarcinoma had metastasized to the mesenteric lymph nodes and liver. Carcinoma of the Large Intestine. This type of tumor was seen in one rat, with metastases to the adjacent mesen- teric lymph node. The animal was terminated at 510 days of age. An osteosarcoma arising from sacral vertebrae was diag- nosed in l rat at the age of 330 days. This tumor metastasi- zed to the lung, right atrium and adrenal gland. Adenocarcinomas of the Prostate Gland. These tumors were detected in 3 rats. The mean survival time of rats bearing this type of tumor was 260 (:2) days. These tumors were ana- plastic and had metastasized to lymph nodes, liver and lung and invaded adjacent tissue. Mucosal papillary hyperplasia of the urinary bladder and prostate gland hyperplasia with squamous metaplasia were observed in 10% of the additional rats from the 2 high-dose groups. None of those changes were observed in the control group. Hemangiosarcoma of the spleen. This type of tumor was detected in one rat with a survival time of 320 days which had metastasized to the liver and kidney. Hemangiopericytoma. 3 ENU-exposed rats developed this type of tumor in the subcutis of the submaxillary region with 33 a mean survival time of 450 (:2) days. Mesothelioma. This tumor was detected in l rat with a survival time of 406 days. The tumor had spread throughout the serosal surface of the abdominal cavity. Renal Carcinomas. These tumors developed in 3 exposed rats, with a survival time of 260 (:2) days. One of the tumors (45 mg/kg group) was a solid carcinoma developing in the cor- tex and invading the adjacent renal parenchyma. The other two tumors (180 mg/kg group) were solid clear-cell carcinomas oc- cupying large areas of the corticomedulla at one pole of the kidney. They consisted of large polyhedral cells with small hyperchromatic nuclei and clear cytoplasm. Foci of small ag- gregates of polyhedral cells with eosinophilic granular cyto- plasm were interspersed throughout. Local invasion of the surrounding renal tissue was apparent. Primary Tumors of the Lung. These were seen in three instances: two were adenomas and one was a squamous cell car- cinoma characterized by the presence of numerous keratin pearls, invasion, necrosis and inflammation. The survival time for the rat bearing a squamous cell carcinoma was 253 days. Neoplasms of the Hemilymphatic Sygtem. These were seen in 9 rats: five lymphosarcomas and four myeloid leukemias. All nine hemilymphatic neoplasms occurred in ENU-exposed rats and affected liver, spleen, kidneys and lymph nodes. All the hemilymphatic neoplasms occurred in rats under 1 year of age (the mean survival time was 245 : 5 days). 36 DISCUSSION The results indicate that single postnatal ENU exposure of 30-day-old male rats has the potential of inducing a wide spectrum of neoplasms in multiple, histogenetically unrelated tissues. The most striking and unusual finding was the high inci- dence of MTs in male rats, not previously described as a con- sequence of transplacental or postnatal ENU exposure. In a previous experiment using female rats, 30% of the ENU-induced MTs were ovarian-hormone-independent.S Although the hormone responsiveness of the ENU-induced MTs in the male rats of the present study is not known, it is reasonable to assume that they are largely ovarian-hormone-independent, because they were transformed in the absence of substantial quantities of these hormones. £3 ziggg experiments provided evidence that N-methyl-N-nitrosourea (MNU) induced rat neoplastic mammary epithelial cells were able to grow in culture medium free of ovarian hormone.9 The inverse correlation between the doses of ENU and neurogenic tumor incidence can be explained by the shorter survival time in the high-dose groups precluding development of neurogenic tumors, which have a longer latency period than MTs. In the high-dose-treated groups we saw a high incidence of alterative changes of central nervous system, such as glial nodules associated with focal malacia, which can be considered early neoplastic proliferation. Fourteen rats from the 90 and 180 mg/kg dose groups in our experiments were terminated be- 37 cause of malignant MTs at 9 months of age (280 days), which is lower than the mean survival time of rats bearing neuro- genic tumors in the high-dose groups (386 and 350 days, re- spectively). Early neoplastic proliferations of the central nervous system were recognized in 8 of these rats. Given a longer survival time, it is probable that true tumors could have developed from these changes. Of special interest are the ameloblastic odontomas, a tumor type not previously reported in CD rats associated with ENU exposure. This tumor occurs in man and has been reported with relatively high frequency in some areas of Africa. Mali- gnant behavior has been recognized, and lung metastases have been described occasionally.lo In dogs, this tumor has been encountered with low frequency, behaving as a slow-growing benign neOplasm.ll Ameloblastic odontomas apparently arise from the primitive odontogenic epithelium, resembing the den- tal lamina. Tumors of the liver have not been reported in experiments using ENU as a carcinogen. The liver has been generally con- sidered to be a non-target organ in ENU carcinogenesis. How- ever, we observed hepatocellular carcinomas in 2 ENU-exposed rats. The spontaneous incidence of liver tumors in male CD rats was reported to be 1% at 24 months of age,14 a survival time which is considerably higher than that of rats from the ENU-treated groups. Some of the tumors listed in Table 2-2 are reported to 12-14 occur spontaneously in the rat. Tumors of the exocrine pancreas, bone and odontogenic apparatus were not encountered 33 spontaneously in historical controls of CD male rats at 24 months of age.l3'l4 Only 2 adenomas of the thyroid gland occurred in the control group. The reason no other tumors were noticed is the age of the animals (491 days) when the experiment was terminated. Spontaneous tumors in rats usu- ally occur later in life, peaking at 2 years and beyond.12' 13,14 The mean survival time for the ENU-exposed rats was ap- proximately 1 year. Most of the malignant tumors occurred in the high-dose groups, in which none of the rats survived the first year of life. It is a fair assumption that these tumors are treatment-related. The number of animals used in the ex- periment and the low incidence of most tumors preclude statis- tical analysis of these neoplasms other than mammary and neuro- genic tumors, where the high incidence is convincing. It is expected that a systemically acting carcinogenic agent might increase the incidence of naturally occurring tumors and shift their appearance to a younger age. Also, ex- posure to a carcinogen may render spontaneously occurring be- nign tumors more malignant and anaplastic. Neoplasias affec- ted in that manner were considered treatment-related. In sup- port of the assumption that malignancies are treatment-related is the presence of proliferative changes (pre-neoplastic a1- terations) in organs and tissues of treated animals without tumors. Such proliferative changes occurred in the brain, liver, pancreas, mammary gland, and urogenital system. None of these changes occurred in the control group. 39 Although neuroectodermal cells and epithelial cells of the mammary gland are the primary target cells for the onco- genic effect of ENU, high doses of ENU administered to young rats induce a wide spectrum of neOplasms. This indicates that ENU is capable of transforming a variety of cell types of dif- ferent organs. In addition to the high doses, the age of the rats may play an important role in explaining these results. During the first weeks of life, the cells of many organs have a high turnover rate, including those of the male mammary gland. The ENU-induced spectrum of neOplasms in multiple histo- genetically unrelated tissues in our experiment could be re- lated to the rate of DNA synthesis and cellular replication, which greatly augment the process of neoplastic transformation. 15,16,17 REFERENCES l. Druckrey H, Ivankovic S, Preussman R: Teratogenic and carcinogenic effects in the offspring after single injec— tion of ethylnitrosourea to pregnant rats. Nature 210:1378- 1379, 1966. 2. Koestner A, Swenberg JA, Weschler W: Transplacen- tal production with ethylnitrosourea of neoplasms of the ner- vous system in Sprague-Dawley rats. Am J Pathol 63:37-56, 1970. 3. Koestner A, Swenberg JA, Weschler W: Experimental tumors of the nervous system induced by resorptive N-nitro-' sourea compounds. Recent Advances in Brain Tumor Research, Progress in Experimental Tumor Research, Homburger Fed., 7: 9-30, 1972. 4. Jones PL, Searle CE, Smith-Thomas W: Tumors of the nervous system.induced in rats by the neonatal administration of N-ethyl-N-nitrosourea. J Pathol 109: , 1973. 5. Su CM, Brash DE, Chang MJW, Hart RW, D'Ambrosio SM: Induction of single-strand breaks plus alkali-labile bonds by N-nitrosourea in rat tissues in vivo: Ethylnitrosourea' versus benzoylnitrosourea. Mutat Res 108:1-12, 1983. 6. Stoica G, Koestner A, Capen C: Characterization of N-ethyl-N-nitrosourea induced mammary tumors in the rat. Am J Pathol 110:161-167, 1983. 7. Searr RW, Torloni H: Histological typing of breast tumors. World Health Organization, International Classifica- tion of Tumors, No. 2, 1968. 8. Young S, Hallowes RD: Tumors of the mammary gland, 41 Pathology of Laboratory Animals. Part I. Tumors of the Rat. Ed. Turusov, IARC, France, pp. 37-73,.l973. 9. Cohen LA: Isolation and characteristics of a seri- ally cultivated, neOplastic, epithelial cell line from the N-nitrosomethylurea induced mammary adenocarcinoma. £2.212E2 18:565-575, 1982. 10. Slavin G, Cameron DH: Ameloblastomas in Africans 'from Tanzania and Uganda: A report of 56 cases. Cancer 23: 31-37, 1969. ll. Langham RF, Mostosky UV, Schirmer RG: Ameloblastoma odontoma in the dog. Am J Vet Res 31:1873-1876, 1969. 12. Gilbert C, Gillman Y: Spontaneous neoplasms in the albino rat. S Afr J Med Sci 23:257-272, 1958. 13. Squire RA, Goodman DG, Valerio MG, Fredrickson TN, Strandbert JD, Levitt MH, Lingeman CH, Harshbarger JC, Dawe CJ: Tumors, Pathology of Laboratory Animals.' Vol 2. Ed. Benirschke, Garner, Jones. Springer-Verlag, New York, p. 1051, 1978. 14. Gart JJ, Chu KC, Tarone ER: Statistical issues in interpretation of chronic bioassay tests for carcinogenicity. 'J Natl Cancer Inst 62:957-974, 1979. 15. Russo J, Wilgus G, Tait L, Russo IH: Influence of age and parity on the susceptibility of rat mammary gland epi- thelial cells in primary cultures to 7, lZ-dimethylbenz(a)an- thracene. In Vitro 17:877-884, 1981. 16. Nagasawa H: Causes of age-dependency of mammary tumor induction by carcinogens in rats. Biomedicine 34:9-11: 1981. 17. Greiner WJ, Bryan AH, Malan-Shilbley LB, Janss HD: Aryl hydrocarbon hydroxylase and epoxide hydratase activi- ties: Age effects in mammary epithelial cells of Sprague-Daw- ley rats. J Natl Cancer Inst 64:1127-1133, 1980. CHAPTER 3 Testicular-like tumor in the ovary of rats induced by chemical carcinogens 43 44 SUMMARY Ovarian tumors which simulated testicular-like structures were detected in 18% of the Sprague-Dawley (CD) and 47% of BD- IV (Berlin Druckrey IV) rats inoculated with high doses of N- ethyl-N-nitrosourea (ENU) transplacentally and postnatally as compared to 2% (CD) and 3% (BD-IV) in controls. Diethylnitro- samine (DEN) administered intravenously postnatally did not produce tumors different from control levels. The carcinogens increased the incidence of ovarian tumors (OT) with a testis- like tubular pattern. Some of these OT (6%) produced metas- tases throughout the mesentery by implantation. Malignant mammary tumors (MT) associated with OT were recognized in 54% of exposed rats. Hormonal assays indicated that some of the tumors produced testosterone, estradiol or estrone. INTRODUCTION Since Pick (1905)23 first described a tumor with testi- cular-like structure in the human female and called it "ade- noma testiculare ovarii", the origin of and appropriate nomen- clature for this type of tumor has been a controversial issue. Investigators have described this tumor in the literature un- der different names such as: arrhenoblastoma,l'13'15'17'19' 20,22 10,21,32,33,34,35 29, androblastoma, Sertoli-like tumor, 33’35 granulosa cell tumor or "folliculoma lipidique",3 and 11'12'27'28 These tumors characteris- sex cord-stromal tumor. tically develop in young women during the third decade of life, or younger.1 Pathologists have distinguished three histolo- logical patterns of these tumors: differentiated tubular type, 45 undifferentiated pattern and intermediate pattern.l’38 Javert and Finn (1951) estimated that 22% of these neoplasms are mal- ignant. Generally, tumors of undifferentiated pattern, rather than those of predominantly tubular type, metastasize. Metas- tases appear late during the course of the disease and may be found in the liver, kidney, lung and paraaortic nodes.14 The histogenesis of these tumors remains debatable. Cur— tis4 suggested that this tumor represents a one-sided develop- ment of teratoma. The term arrhenoblastoma was first proposed by Meyer19 and was used by him in a morphological sense. Meyer suggested that arrhenoblastomas arise from certain indifferent, ambivalent, but male directed cells which he presumed had per- sisted in the rete ovarii since the intermediate phase of go- nadogenesis and which were capable of proliferating to pro- duce tubules.19 Subsequently, the term arrhenoblastoma was adopted for a group of virilizing ovarian neOplasms. The manifestations of masculinization vary in degree and not all arrhenoblastomas are recognizably virilizing tumors. Indeed, the androgenic activity of an arrhenoblastoma is not always related to its content of Leydig-like (interstitial) cells. In certain cases of tubular testicular adenoma, tubular andro- blastoma, or Sertoli testicular cell tumor, the hormonal ef- fects may be either estrogenic or androgenic.l Mackinley (1957) suggested that arrhenoblastomas and their mixed variants are granulosa-cell tumors in varying de- 32'35 termed the tumors ho- grees of differentiation.1 Teilum mologous androblastomas based upon the morphological identity of certain gonadal tumors in the male and female which led to 46 the concept of using the tumor classification for identical testicular and ovarian tumors. The term "androblastoma" (from "andros" meaning man) is a designation for neOplasms derived from the mesenchymal core of both testis and ovary and, in either sex, showing a tendency to reflect the specific testi- cular architecture irrespective of the qualitative hormone- 33 producing properties. Similar testicular-like tumors have been described in 25,26 7,12 39 . . . . . 36 animal ovaries, in mice, rats, agoutis, shrews, mares,15’20 Indian Deshi hens,9 and cats.12 Tubular adenomas of the ovary were observed spontaneously in old Wistar rats usually 900 to 1,000 days of age.17 In mice, mares, and In- dian Deshi hens, the tumors were associated with virilizing 9,15,25 changes. In rats and agoutis the tumors were not as- sociated with hormone production.7’l6 In our experiments, the rat strains Berlin Druckrey (BD- IV) and Sprague-Dawley (CD) were chosen for chronic carcino- genic experiments using N-ethyl-N-nitrosourea (ENU) and di- ethylnitrosamine (DEN) as chemical carcinogens to study the incidence and pathology of tumors of the nervous system, mam- mary gland, liver and kidney, well known target organs for 6'18'30’31 Besides the tumors of known tar- these chemicals. get organs (described previously), 21% of the experimental animals developed an unusual type of ovarian tumor with testi- cular-like structures (arrhenoblastomas). Although the ovary was not previously known to be a target organ for ENU or for DEN, our experiments proved that ENU especially was able to produce a relatively high yield of ovarian tumors with testi- 47 cular-like structure. The purpose of the present study was to define the frequency of ENU and DEN-induced ovarian neo- plasias in different rat strains, to characterize the morpholo- gy, metastasis behavior and hormonal characteristics of these relatively rare and controversial testicular-like tumors, and to investigate the possible relationship between these neo- plasms and other types of tumors, especially mammary tumors. MATERIALS AND METHODS Carcinogen ENU crystalsa were dissolved shortly before administration in phosphate/Citrate-buffered saline at pH 4.2 (1 part buffer to 14 parts saline). The crystalline carcinogen was stored in a dessicator at -20 C. DENb was freshly prepared in saline solution at pH 6.6. Animals All animals used (240 in experimental groups and 120 in control groups) were female specific-pathogen-free (SPF) Spra- gue-Dawley (CD)c and BD-IVd rats (Table 3-1). Eight experi- mental groups of 30 rats each were utilized: three groups of CD rats were inoculated IP at 30 days of age with freshly prepared ENU solution in doses of 180, 90 and 45 mg/kg; one group of rats were offspring of CD mothers inoculated intra- Synthesized by Dr. Ming Chang, Veterinary Pathobiology De- a partment, Ohio State University, Columbus, Ohio. b Fisher Company, Livonia, Michigan. c Charles River Laboratory, Portage, Michigan. d Harlan Industries, Indianapolis, Indiana. 48 venously (TPIV) with ENU on the 20th day of gestation; two groups of BD-IV rats were inoculated IP at 30 days of age with freshly prepared ENU solution at doses of 90 and 45 mg/kg; two groups of CD rats were inoculated intravenously at 30 days of age with freshly prepared DEN solution at doses of 160 and 80 mg/kg. Control animals (60 CD and 60 BD-IV rats) were inoculated IP at 30 days of age with saline solution. Animals were housed in plastic cages (two rats per cage) in a temperature-controlled room (24 C) with 10 hours of electric light per day. They were fed autoclaved Purina Lab Chow 5010 C and water 3g libitum. The rats were observed daily and weighed weekly after treatment until death. Most of the rats were sacrificed or died because of overgrowing mammary gland, nervous system, liver or kidney tumors, or because of cachex- ia. Morphological Studies A complete necropsy was performed shortly after spontan- eous death or euthanasia of moribund rats. The histological sampling procedures were identical in all groups. The tissues were fixed in 10% buffered formalin or in Bouin's solution and prepared for light microsc0pic examination. They were sectioned at 4-6 u and stained with hemotoxylin and eosin (H&E), periodic acid-Schiff (PAS) or osmium. The tissues for electron microsc0py were fixed in glutaraldehyde and osmic acid and, after proper processing, were embedded in Epon, cut at l u with an ultramicrotome and stained with 1% toluidine blue. Blocks were selected from thick sections for electron microsc0pic examination following sectioning and contrasting 49 with uranyl acetate and lead citrate. Hormonal Assays Blood samples were collected from rats prior to euthan- asia under anesthesia using ketamine hydrochloride; serum was separated by centrifugation and immediately frozen at -20 C. Hormonal assays for testosterone, estradiol and estrone were performed on serum samples collected from 8 experimental rats and 8 control rats of the same ages and strains. In addition, two samples of intracystic fluid were also examined for hor- monal content. Hormonal assays on serum samples were perform- ed by simple radioimmuno-assay techniques.24 The values of hormonal assays were estimated by standard error of mean. 50 RESULTS Tumor Induction A high percentage of rats in this experiment developed neoplasms of the target organs for the specific carcinogens (eg. mammary gland ( 100% in high dose CD rats; 45% in high dose BD-IV), central nervous system ( 80% in high dose CD rats; 50% in high dose BD-IV rats), and kidney (10% in high dose CD rats;0% in BD-IV rats)). The incidence and pathology of tumors in other organs have been reported previously.30’31 In addition, 21% of the rats developed tumors of the ovary. Ovarian tumors developed in 50 rats from the experimental groups (ENU - 48 tumors; DEN - 2 tumors) that received chemi- cal carcinogens and in 3 rats from the control groups (Table 3-1). Considering the ENU exposed rats separately, the CD rats had an 18% and the BD-IV a 47% incidence of ovarian tumors of testicular type. The ovarian tumors in the BD-IV rats exceeded in incidence all other tumors encountered in this strain. In the control groups, two rats (BD-IV) devel- oped granulosa cell tumors of the ovary and one (CD) rat de- velOped a testicular (Sertoli's cell)-1ike tumor. The sur- vival time ranged from 223 to 590 days in experimental rats with a mean survival time of 426 days for experimental groups and 850 days for control rats. Clinical Findings The majority of CD rats inoculated with ENU postnatally 2 -In 76% of the rats which developed mammary tumors (MT).3 developed OT, MT were also present. In some of the rats, zonal alopecia (inguinal and abdominal regions) were observed. 51 Table 3-1. Incidence of ovarian tumors in rats inoculated with chemical carcinogens Number of Rats Strain Number with of of Ovarian Carcinogen Rat Rats Tumors Percent ENU 45 mg/kg IP CD* 30 7 23 ENU 90 mg/kg IP CD 30 5 l7 ENU 180 mg/kg IP CD 30 7 23 ENU 20 mg/kg CD 30 3 10 TPIV Total 120 22 18 ENU 45 mg/kg IP BD-IV** 30 11 37 ENU 90 mg/kg IP BD-IV 30 15 50 Total 60 26 47 BEN 80 mg/kg IV CD 30 1 3 DEN 160 mg/kg IV CD' 30 1 3 Total 60 2 3 igtal_Experiment Groups 240 50 21 Control Group CD 60 1 2 BD-IV 60 2*** 3 Total-Control Groups 120 3 1 CD - Sprague-Dawley BD-IV - Berlin.DruCkrey *** Granulosa Cell Tumor 52 Tumors larger than 1 cm were detected by abdominal palpation. The rats which developed OT larger than 2-3 cm usually exper- ienced metrorrhagia. No other signs of virilization or femi- nization were observed clinically. Macroscopic Appearance The malignant MT associated with OT were recognized in 54% of experimental rats, the rest of the MT were classified as fibroadenomas or cystadenomas. The OT were unilateral, ranging from 0.2 to 5 cm in diameter (Figure 3-1). They were usually solid, encapsulated, lobulated tumoral masses, white- yellowish in color (Figure 3-2). Some contained multiple cysts filled with serosanguinous or hemorrhagic fluid. Cystic endometrial hyperplasia and polyps were detected in 27% of the rats bearing OT. Metastases of OT were detected as multi- ple miliary nodules disseminated throughout the mesenterium in 6% of the rats affected by OT. Two rats died because of ruptures of large (4-5 cm in diameter) hemorrhagic OT. The OT weighed from 2 to 40 grams. The Opposite ovary appeared smaller than normal. In one case an OT was associated with a vaginal leiomyosarcoma. Microscopic Appearance Tubular-like structures were present in all OT examined histologically. The histologic appearance of the tumors vari- ed considerably, consisting of cysts, cords, undifferentiated mesenchymal cells, or of solid cellular masses arranged in tubular structures of varying sizes resembling seminiferous tubules and lined by Sertoli-like cells (Figure 3-3). A thin fibrous capsule surrounded the mass. The neOplastic cells 53 Figure 3-1. Gross morphology of a testicular-like tumor in the ovary of a rat. The tumor metasta- sized by implantation throughout the mesen- terium. Serum testosterone was 2250 pg/ml. Figure 3-2. Sagittal section of the above tumor showing an encapsulated, solid lobulated mass. 54 -1 FIGURE 3 FIGURE 3-2 Figure 3-3. Testicular-like tumor. Sertoli cells arranged in tubules and pseudotubular structures. Bar = 50 um. Figure 3—4. Photomicrograph illustrating tubular arrange- ment of spindle-shaped Sertoli cells. Bar = 150 um. 56 FIGURE 3-3 FIGURE 3-4 57 were characterized by large spindle-shaped or polyhedral cells with basally-located nuclei and indistinct light eosinophilic foamy cytoplasm containing many indistinct vacuoles and pro- jecting out into a tubular lumen; usually the centers of such lumens were filled with similar appearing cells (Figure 3-4). Numerous intractyOplasmic lipid droplets were identified with osmium stain. Cords of cells were recognized especially in malignant tumors. The cords consisted of the same palisading cell type lining a thin fibrovascular stroma. Cell borders were indistinct due to the compact arrangement of tumor cells. The nuclei were oval and hyperchromatic. The tumor cells ex- hibited the heterochromatic Barr-body of the female nucleus. Fifty percent of the tumors were observed originating from the hilus region and ovarian parenchyma (Figure 3-5), 26% were seen arising from the ovarian capsule (surface epitheli- um), and 24% of the tumors were seen develOping intracystic- ally from the stromal wall (Table 3-2). In some of the smaller OT, atretic follicles and normal or degenerating corpora lu- tea were seen in addition to the tubular structures. Granu- losa cell tumors observed in control BD-IV rats exhibited a characteristic pattern with Call-Exner bodies. They were de- void of tubular structures. One testicular-like tumor (1.5 cm in diameter) in a control rat (CD, 870 days old) was in- distinguishable from induced OT. The metastases which occur- red in 6% of the OT in experimental groups appeared as miliary nodules seeding the mesentery and reproducing the same tubu- lar-like structure lined by Sertoli-like cells. The metas- tases spread by implantation from malignant OT through rup- tured tumor capsules. The opposite ovaries were usually b1 CC) Figure 3-5. Tubular structures with Sertoli cells arising in the hilus region of the rat ovary (arrow). Bar = 50 um. 59 Figure 3-5 60 Table 3-2. Classification of ovarian tumors and other related alterative changes induced by chemical carcinogens Number Percent I. Ovary: Testicular-like tumors 50 21 II. Origin 1. Hilus and ovarian parenchyma 25 50 2. Capsule 13 26 3. Intracystic walls 12 24 III. Metastases l. Mesentery 3 6 IV. Uterine, ovarian and mammary changes 1. Endometrial cystic hyper- plasia ll 27 2. Leiomyosarcoma l 2 3. Ovarian atrophy of germi- nal epithelium (Opposite ovary) 25 50 4. Mammary tumors 38 76 Adenocarcinomas 27 54 Fibroadenomas 8 l6 Cystadenomas 3 6 61 atrophic. Ultrastructural Evaluation (EM). As a rule, the tumor cell sheets were separated from the stromal connective tissue by a basal membrane (Figure 3-6,7). The cells were closely apposed to each other; occasionally intercellular spaces occurred. The cells were elongated, with their longitudinal axis often oriented at right angles to the basal membrane. Between adjacent cells, occasional desmosome -1ike contacts were formed. The nuclear shape varied and was either elongated or ovoid; most nuclei were moderately indent- ed (Figure 3-6). The cells had moderate numbers of organ- elles; however, there was an increase in smooth endoplasmic reticulum and in mitochondria with tubular cristae (Figure 3 -6). The Golgi cmplexes were inconspicuous. Lipid droplets were seen in the cytoplasm of all tumors. Ribosomes and polysomes occurrred regularly (Figure 3-8). Occasional an- nulate lamellae and small dispersed profiles of rough endo- plasmic reticulum were identified. The two granulosa cell tumors found in the control BD-IV rats exhibited distinct ultrastructural pattern characterized by chromatin rich nu- clei and numerous large lipid granules in the cytoplasm. No ultrastructural difference was observed between induced testi- cular-like tumors and the spontaneous one found in the control (CD) group. The mean standard error of testosterone levels in the experimental group was 37.5% higher than in the control group (Figure 3-9). The serum testosterone content was in one case of malignant metastasizing OT especially high (2250 pg/ml). Figure 3-6. 62 Electron micrograph of tubules composed of in- dented nuclei of Sertoli cells with cytoplasm containing mitochondria (M), free ribosomes (R) and secretory granules (S). Uranyl ace- tate-lead citrate. (x 6000). 63 Figure 3-6 64 Figure 3-7. Electron micrograph of a ductular structure illustrating Sertoli cells lined by a thin basement membrane (B). The cytOplasm con- tains vesicles (V) and secretory granules (S). Uranyl acetate-lead citrate. (x 6000). 65 5. .. a». at . A.» m 3-7 Figure Figure 3-8. 66 Higher magnification electron micrograph of Sertoli cells lined on a thin basement mem- brane (B). The nuclei contain homogeneous disperse chromatin. The cytOplasm contain large vesicles (V), rough endoplasmic pro- files (RE) and mitochondria (M). The cells are attached to each other by desmosomes (arrows). (x 10000). 67 Figure 3-9. Figure 3-10. 68 Serum testosterone in rats with testicular (Sertoli's cell)-like tumors of the ovary related to tumor diameter. The mean value in 8 rats with ovarian tumors (636 : 109 pg/ ml) was elevated 46% above controls (437 : 27 pg/ml) (mean = solid line; standard error = shaded area). Serum estrone in rats with testicular (Ser- toli's cell)-like tumors of the ovary related to tumor diameter. The mean concentration in 8 rats with ovarian tumors (174 : 11 pg/ml) was elevated above serum estrone levels in control rats (156 : 9 pg/ml) but the differen- ces were not significant. 0'3 (“9) UBLBNVIO HOWfll 0'9 0'9 0 69 SERUM TESTOSTERONE (pg/ml) 0| 0'? N A 8 o 8 8 8 o o o o o 1 / 7 1 A X‘H 4 O O O O O / / A 0 Figure 3-9 q: 0E? or (“I U313MIO norm or “L ‘\ Figure 3-10 Figure 3-11. 70 Serum estradiol in rats with testicular (Ser- toli's cells)-like tumors of the ovary related to tumor diameter. The mean value in 8 rats with ovarian tumors (108 : 6 pg/ml) was elevated significantly above serum estradiOl levels in control rats (90 : 4 pg/ml) (mean = solid line; standard error = shaded area). '9’ 3.9 3 seem: ESTRADIOL (pg/mu 8 $ 7 71 O to 2b 3b 4:0 5.0 Tunas DIAMETER (cm) Figure 3-11 72 The testosterone content of the intracystic tumoral fluid in the other two malignant OT was also elevated (1027 and 5702 pg/ml). The estrone level in malignant OT was 50% (Figure 3- 10) and the estradiol level was 75% above the mean control (Figure 3-11). The latter two hormones were not elevated in the cystic fluid. DISCUSSION The carcinogens (ENU and DEN) used in our experiments increased the OT incidence and shifted their occurrence to a younger age (mean survival time of 426 days). The tumor type was uniformally of a testis-like tubular pattern with a high rate of malignancy. The BD-IV rat is considered a strain which is relatively resistant to chemical carcinogens.6 In- deed, in comparison with CD rats, BD-IV rats developed a lower incidence of nervous system and mammary gland tumors.30'31 Surprisingly the opposite was true for ovarian tumors, BD-IV rats had a significantly higher ovarian tumor rate than CD rats (Table 3-1). With the highest dose of ENU used (90 mg/ kg IP), 50% of the female BD-IV rats developed testicular-like tumors in the ovaries. The OT in the BD-IV rats exceeded in incidence all other tumors encountered in this strain. A good correlation between the incidence of OT and MT was established. Of 22 CD female rats bearing OT, 18 had also mammary adenocarcinomas. In BD-IV rats exposed to ENU, 10 out of 26 rats had OT and MT concurrently. Nine of the MT were adenocarcinomas. Four rats bearing OT and MT (for which hor- monal assays were done) expressed a higher than normal serum 73 estradiol. None of the OT-bearing rats with increased serum testosterone developed MT. Hormonal assays were not performed in all OT bearing rats. However, the hormonal increase in 8 rats bearing OT combined with cystic endometrial hyperplasia ‘ (38% of rats bearing OT) and malignant MT is a good indication that some of the OT were functional. The association of OT with malignant MT in BD-IV rats (9 out of 10) which, contrary to CD rats, are not prone to develOp MT, could be related to increased ovarian hormone production in rats bearing OT, which probably stimulated the progression of malignant MT. Histologically and ultrastructurally the cells populat- ing the seminiferous-like structures are of epithelial origin. They are large cells with ovoid and grooved nuclei, indistinct foamy cytoplasm and secretory lipid droplets, similar to the Sertoli tumor cells described in dogs.2 The clinical, histo- logical and hormonal findings illustrate that some of the OT are functional. Increased serum-testosterone levels in some and increased estrone and estradiol levels in other rats (Fig- ure 3-9,10,ll) contrast the opinions of other authors who claim that these types of tumors are nonfunctional in rats.3' 7 The induction of OT in 30-day-old female CD and BD-IV rats with chemical carcinogens is possible because the ovary remains an embryonic organ for a period of time after the fe- male reaches puberty. At the 30th day of age and thereafter, growth and develOpment of ovarian components continue, espec- ially toward the formation of new follicles. The granulosa and Sertoli cells have the same embryological anlage, derived 74 from mesenchymal blastema which forms the sex cords (granulosa cells in females and Sertoli cells in males).40 A high rate of DNA synthesis and cellular replication occur in the ovaries, constituting an ideal target for alkylating carcinogens. For unknown reasons the neoplastic transformation differentiates mostly towards testicular-like tumors resembling Sertoli cells. As a consequence of this differentiation the hormonal behavior varies since Sertoli cells may produce estrogens as well as 1,3,5,7 testosterone. Alterations in enzyme pathways in neo- plastic tissue should explain why some of these tumors result in high levels of testosterone and others do not.5’8'20'37'40 Controversy continues regarding the histogenesis of these tumors, called arrhenoblastomas, androblastomas, Sertoli-Ley- dig tumors, tubular adenomas or sex cord stromal tumors by different authors. Since this is a relatively rare tumor in man and there is not a good biological model, it has been difficult to clarify the pathogenesis. Indeed, some of these tumors arise from the hilus region, probably as Meyer suggested, 19 from the male-directed cells that have persisted in the rete ovaries after gonadogenesis. However, a good proportion of the tumors also arise from the ovarian capsule or from the stromal wall, which Opens the alternative possibility that they develop despite their frequent testicular morphology, directly from ovarian stroma. There is no reason to believe that only rudimentary testicular cords are capable of forming tubular structures, inasmuch as the ovarian mesenchyme is al- iso capable of differentiating into tubular structures, sex 1,21,33,40 <:ords, Sertoli cells and Leydig cells. Clarifica- 75 tion of the origin and pathogenesis of this controversial type of human ovarian neoplasm may likely result from more careful scrutiny of spontaneous and induced ovarian neoplasms of animals and a comparison of the morphology and physiology of human and animal ovaries. REF ERENCES 1. Ashley DJB: In: Evans' Histological Appearance of Tumours, ed 3. New York, Churchill and Livingston, 1978, pp 663-672. 2. Capen CC, Martin SL: Ultrastructural evaluation and pathophysiology of testicular sertoli cell neOplasm from dogs with a syndrome of hyperestrogenism. In: Arecenaux (ed): iiEE Annual Proceedings of the Electron Microscopy Society of Ameri- gg, Baton Rouge, Claitor's Publishing Division, 1973, pp 384- 385. 3. Cotchin E: Spontaneous tumors of the uterus and ovar- ies in animals. In: .Blaustein (ed): Pathology of the Female Genital Tract, New York, Springer-Verlag, 1977, pp 836-838. 4. Curtis AH: Another case of arrhenoblastoma. Am J Obstet Gynecol 52:128-132, 1946. 5. Devlin TM: Textbook of Biochemistgy, New York, Wiley and Sons, 1982, pp 741. 6. Druckrey H, Ivankovic S, Preussman R, Zulch KH, Men— nel HD: Selective induction of malignant tumors of the ner- vous system by resorptive carcinogens. In: Kirsch, Grossi- Paoletti and Paoletti (eds): Experimental Biology of Brain Tumors, Springfield, Thomas, 1972, pp 85-147. 7. Engle ET: Tubular adenomas and testis-like tubules of the ovaries of aged rats. Cancer Res 6:578-582, 1946. 8. Engel LL: Recent studies on estrogen metabolism. In: Pincus and Vollmer (eds): Activities of Steroids in Re- lation to Cancer, New York, Academic Press, 1960, pp 111. 9. Gupta BN, Langham RF: Arrhenoblastoma in an Indian Deshi hen. Avian Dis 12:441-444, 1968. 10. Hayes DM, Hunter RJ: Androblastoma of the ovary with heterotropic elements. J Pathol 109:267-270, 1973. ll. Hertig AT, Gore H: Tumors of the female sex cords. 3. Tumors of the ovary and Fallopian tubes. Atlas of Tumor Pathology, sect. IX, fasc. 33 (Armed Forces Institute of Path- ology, Washington 1961). 12. Hofman W, Arbiter D, Scheele D: Sex cord stromal tumor of the cat:- So-called androblastoma with Sertoli-Ley- idig cell pattern. Vet Pathol 17:5o8-513, 1980. 13. Hughesdon PE, Fraser IT: Arrhenoblastoma of ovary. ACta Gynecol Scand Suppl 4, 32:1-78, 1955. 14. Javert TC, Finn WF: Arrhenoblastoma, the incidence of malignancy and the relationship to pregnancy, to sterility and to treatment. Canoer 4:60-77, 1951. 15. Jubb RW, Kennedy PC: Pathology of Domestic Animals, ed. 2, New York, Academic Press, 1970, vol 1, pp 394-401. 16. Kullander S: Studies in spayed rats with ovarian tissue auto-transplanted into the spleen. Acta Endrocrinol 24:307-332, 1957. 17. Langley FA: "Sertoli" and "Leydig" cells in relation to ovarian tumors. J Clin Pathol 7:10-17, 1954. 18. Magee PN, Montesano R, Preussman R: N-Nitroso com- pounds and related carcinogens. In: Advances in Cancer Res- search MOnograph Series No. 173, Chemical Carcinogens, pp 504-505, 1974. 19. Meyer R: The pathology of some special ovarian tumors 78 and their relation to sex characteristics. Am J Obstet Gyne- go: 22:691-71 , 1931 20. Mills JHL, Fretz PB, Clark EG, Ganjam VL: Arrheno- blastoma in a mare. JAZMA 171:754-757, 1977. 21. Moulton JE: Tumors in Domestic Animals, Berkeley, University of California Press, 1978, pp 331-336. 22. Norris EH: Arrhenoblastoma: Malignant ovarian tumor associated with endocrinological effects. Am J Cancer 32:1- 29, 1938. 23. Pick L: Uber Neubildungen am Genitale bei Zwittern nebst Beitragen zur Lehre von den Adenomen des Hodens und Ei- erstockes. Arch Gynaekol 76:191-281, 1905. 24. Powell JE, Stevens VC: Simple radioimmunoassay of 5 unconjugated ovarian steroids in single samples of serum plasma. Clin Chem 19:210-215, 1973. 25. Russell ES, Fekete E: Analysis of W-series pleio- tropism in the mouse: Effects of WVWv substitution on defi- nitive germ cells and on ovarian tumorigenesis. J Natl Can- cer Inst 21:365-381, 1958. 26. Russfield AB: Pathology of endocrine glands, ovary and testis of rats and mice. In: Cotchin and Roe (eds): Pathology of Laboratory Rats and Mice, Oxford and Edinburgh, Blackwell Scientific Publications, 1967, pp 391-468. 27. Scully RE: Sex-cord stromal tumors. In: Blaustein (ed): Pathology of the Female Genital Tract, New York, Springer-Verlag, 1977, pp 505-525. 28. Serov SF, Scully RE, Sabin LH: International Histo- logical Classification of Tumors, No. 9. Histological Typing 79 of Ovarian Tumors, Geneva, World Health Organization, 1973. 29. Siller WG: A Sertoli cell tumor causing feminization in a Brown Leghorn capon. J Endocrinol 14:197-203, 1956. 30. Stoica G, Koestner A, Capen C: Spectrum of neoplasms induced with high single doses of N-ethyl-N-Nitrosourea in 30 -day-old rats with special emphasis on mammary neoplasia. AntiCancer Research 4:5-12, 1984. 31. Stoica G, Koestner A, Capen C: Characterization of N-ethyl-N-nitrosurea-induced mammary tumors in the rat. Am. J Pathol 110:161-169, 1983. 32. Teilum G: Arrhenoblastoma-androblastoma. Homologous ovarian and testicular tumors. I. With discussion of "meso- nephroma ovaries". Acta Path Microbiol Scand 23:242-251, 1946. 33. Teilum G: Estrogen-producing Sertoli cell tumors (an- droblastoma tubulare lipoides) of the human testis and ovary. HomolOgous ovarian and testicular tumors. III. J Clin Endo- crinol 9:301-318, 1949. 34. Teilum G: Classification of testicular and ovarian androblastoma and Sertoli cell tumor. A survey of comparative studies with consideration of histogenesis, endocrinology, and embryological theories. Cancer 2:769-781, 1958. 35. Teilum G: Special tumors of Ovary and Testis. Com- parative Pathology and Histological Identification, Philadel- phia, Lippincott, 1971. 36. Weir BJ: Some observations on reproduction in the female agouti, Dasyprocta aguti. J Reprod Fertil 24:203-211, 1971. 37. Wiest WG, Zander J, Holmstrom EG: Metabolism of pro- 80 gesterone-4-Cl4 by an arrhenoblastoma. J Clin Endocrinol Metab 19:297-305, 1959. 38. Williamson ME, Leestma JE, Black WC, King DW: Histo- logic Patterns in Tumor Pathology, New York, Harper and Rowe, 1968. 39. Wilcox DE, Mossman HW: The common occurence of "tes- tis" cords in the ovaries of a shrew (Sorex vagrans, Baird). Anat Rec 92:183-194, 1945. 40. Witschi E: Embryology of the Ovary. In: -Grady and Smith (eds): The Ovary, Baltimore, Williams and Wilkins, 1962, pp 1-10. 41. Zander J: Steroids in the human ovary. J Biol Chem 232:117-124, 1957. CHAPTER 4 Characteristics of normal rat mammary epithelial cells and N-ethyl-N-nitrosourea -induced mammary adenocarcinoma cells grown in culture 81 82 SUMMARY The characteristics of normal mammary epithelial cells derived from Lewis and Sprague-Dawley rats and ENU-induced adenocarcinoma cells derived from Sprague-Dawley and Fisher rats and grown in monolayer cultures were compared. After collagenase treatment the rat mammary epithelial cell aggre- gates were placed in a hormone-supplemented medium. The nor- mal mammary epithelial cells (NE) attached to the surface of the culture dish within 50 hours, whereas the mammary adeno- carcinoma cells (MA) attached within 24 hours and grew as a cell multilayers. After the colonies of NE and MA cells be- came confluent, the culture entered a steady state in which the cells from the upper layer were shed into the medium. The rate of proliferation and squame detachment in confluent cultures was increased by the presence of epidermal growth factor. Rhodanile blue staining and transmission electron micrsc0py showed that the shed cells were partially keratin- ized. In addition, the normal cells were unable to grow in soft agar or to form tumors when inoculated into appropriate hosts. The Opposite was true in each case for the mammary adenocarcinoma cells. Karyotypes of normal and neoplastic mammary rat epithelial cells revealed a hypodiploid modal num- ber of chromsomes. INTRODUCTION The administration of relatively high doses of certain :polycyclic aromatic hydrocarbons, e.g., DMBA, MCA or alkylated Initrosoureas e.g., MNU, ENU to female rats invariably results 1-4 .in a high incidence of mammary carcinomas. The carcinogen- 83 ic efficiency of these chemicals and the morphological and physiological similarities of the induced carcinomas to human breast cancer provides the basis for the attractiveness of this system as an important model to examine mechanisms of mammary gland carcinogenesis. To date, chemical carcinogene- sis of the rat mammary gland has been examined primarily :2 3:39; few laboratories have attempted to examine this neo- plastic process ip yitgg. A prerequisite for a thorough analysis of rat mammary gland carcinogenesis lg vitro is the availability of cell culture techniques capable of maintaining and propagating normal and neoplastic rat mammary epithelial cells. During the past two year, we have developed techniques to excise and isolate normal rat mammary epithelial cells and mammary carcinoma cells derived from ENU-induced rat mammary carcinomas. In this commmunication, we describe these tech- niques as well as characteristics of these cells maintained in cell cultures. MATERIALS AND METHODS Source and Isolation of Cells. Normal mammary tissue (NE) was obtained from the whole fat pad (inguinal and axillary) of nor- mal female Sprague-Dawleya rats (SNLR/CDl) and normal Lewisa rats (SNLR/Ll), at fifty days of age. Under ketamine hydro- chloride anesthesia,b the whole fat pad containing the mammary gland was removed and transferred to a small Erlenmeyer flask a Charles River Laboratories, Portage, Michigan. b Vetalar, Parke Davis, Morris Planes, New Jersey. 84 containing Eagle's Minimum Essential Medium (MEM) with Earle's Salts and 2 mM L-glutam1ne.c Fragments of the whole mammary gland were transferred to a 60 mm Petri dish and under a dis- secting microscope remnants of connective tissue, fat and lymph nodes were removed using a small forceps with fine tips and an iris scissors. After dissection, the tissue contained mostly mammary epithelium and fat tissue. The fragments were trans- ferred to a concave glass plate and minced with crossed sterile scalpels. The finely minced tissue was then placed in a 100 ml Erlenmeyer flask containing 50 ml MEM, 0.1% collagenase,d 4,000 I.U. penicillinC and 2,000 mg streptomycinc and was incubated at 37 C in a gyratory water bath for 40 minutes. The cell sus- pension was then pipetted vigorously to further disperse cells and then was filtered through one layer of 74 um opening nylon mesh.d The filtrate was centrifuged at 100 x g to remove cell debris, and the resulting pellet was resuspended in 4 m1 of Dul- becco's Modified Eagles Medium (DMEM)c supplemented with 25% horse serum (HS)c and a hormone combination consisting of 20 ug /ml each of hydrocortisone, insulin, and prolactin,6 and 10 ng /ml epithelial growth factor (EGF),f and 1% antibiotics (peni- cllin and streptomycinc). Initially, the cells and cell clus- ters were plated at a high density into 60 mm tissue culture dishes and incubated in a humidified 10% CO incubator at 37 C. 2 Six hours after the cellular components started to settle down *flley were examined under an inverted microscope. Free cells, I Gibco, Grand Island, New York. Worthington Biochemical Corp., Freehold, New York. Sigma Chemical Co., St. Louis, Missouri. Collaborative Research, Waltham, Massachusetts. HUDQJ) 85 cell aggregates and ductular structures, some closely resem- bling terminal end buds, were photographed using a Nikon F5 UFX Automatic Camera9 and Panatomic X film.h At 8 to 10 hour intervals, the suspended cells in each dish were removed by decanting to another dish. Fresh growth medium was added to these dishes. This process was repeated 5 times. The initial dishes usually contained mesenchymal cells, but the cells plated out in subsequent dishes were mostly epithelial cells. Most of the epithelial colonies developed from cell aggregates and duct-like structures which attached to the bottom of the dishes, between 20-50 hours post-seeding. Every 2 days half the medium was removed and fresh growth medium was added to every dish. Ten days later multiple epithelial colo- nies of various sizes surrounded by mesenchymal cells were re- cognized in several dishes, but mostly in dishes that were se- eded at later time periods. Using a marker objective mounted on the inverted Nikon microscope the epithelial-like colonies were identified and the mesenchymal cells were removed using a rubber policeman and differential trypsinization (gradually increasing trypsin concentrations from 0.01 to 0.05% with 0.2% EDTA). The procedures were repeated several times until only epithelial-like colonies remained. Confluency in culture dishes was obtained in 2 months. _1nduction of Mammary Tumors. Mammary adenocarcinomas (MA) were .induced in female Sprague-Dawleya and Fishera rats by a single 4-6 :i.p. injection of ENU as previously described. Between 60 9 Nikon, Tokyo, Japan. Eastman Kodak Co., Rochester, New York. 86 days and 80 days post-inoculation,-the rats developed multiple mammary tumors. The rats were euthanized using CO and mam- 2. mary tumors were excised surgically under aseptic conditions, cleared of surrounding connective tissue, and cut into 2 mm3 segments for cell culture preparation. The remainder of the tumor was fixed in buffered formalin, sectioned and stained with hematoxylin and eosin (H&E) for histOpathological examin- ation. Tissue dissociation and isolation of neoplastic epi- thelial cells was achieved in the same manner as described above for normal mammary epithelium. Confluency of the ade- nocarcinoma cells in culture dishes was obtained in 2 weeks. Routine Tissue Culture Techniques. Once growth was established, the cells were transferred to T75 flasks and maintained as des- cribed above. At later passages (P10) the concentration of ser- um was loWered from 25 to 10%. For subculture of normal and neoplastic mammary epithelial cells, growth medium was aspira- ted and the cells washed once with Hanks? Buffered Saline Solu- tion‘(HBSS)c and cells dissociated with 0.25% trypsin and 0.0 2% EDTA.c Treatment with trypsin was for 3 to 5 minutes at 37 C. The treatment was repeated twice. Epithelial cells (NE and MA) strongly attached to the dishes for the first passages and therefore were difficult to remove. The cells were never to- tally removed from the flask in which they achieved confluency for the first time; these cells were maintained for future sub- cultures in case of cell loss. The medium was changed every other day. The cells were counted using a hemocytometer. For the first passages, the cell aggregates were difficult to dis- rupt during preparation of the cell suspension; the cells were 87 lysed in 0.1 M citric acid containing 0.1% crystal violet at 37 C for 1 hour and the nuclei were counted. The cells were frozen in liquid nitrogen (N2) in DMEM + 2 mM glutamine sup- plemented with 20% HS and 10% DMSO as a cryOprotectant. To determine seeding efficiency, the cells were seeded at 4.2 x 105/60 mm Petri dish and scored after 4 hours when the maximum number ofcells had attached but before mitosis started.7 The formula for seeding efficiency is: No. of Cells Recovered x 100 = Seeding Efficiency No. of Cells Seeded Plating efficiency was calculated by seeding, at a low density, 600 and 1000 cells/60 mm Petri dish. The cells grew as discrete colonies. The colonies were counted after 8 days and the results were expressed as the plating efficiency. The formula for plating efficiency is: No. of Colonies Formed x 100 = Plating Efficiency No. of Cells Seeded. The population doubling time was calculated expressing the cell number as a function of log2 by the multiplication of the common logarithm of the cell count by the factor 3.33.8 The formula used for the calculation of population doubling time is: Number of population doublings = log (N/N ) x 3.3 Where: N = number of cells in the gfgwth gessel at the end of a period of growth. N0 = number of cells planted in the growth vessel. ~ '88 The squames accounted for the calculation of population dou- bling time. Cloning and Anchorage Independent Growth. The mammary adeno- carcinoma epithelial cells were cloned using dilution cloning technique.9 The ability of MA rat mammary epithelial cells tO'form colonies in agar was determined as described by Mac- Pherson10 with minor modifications. The basal layer, which contained 5 ml of 2% Sea Plaque Agarosei supplemented with DMEM, 25% HS, and 0.2% sodium bicarbonate was formed in 60 mm plastic Petri dishes. Formation of the basal layer was fol- lowed by the addition of a second layer (1.5 ml) 0f 0.33% Sea 4 dil- Plaque Agarose supplemented as above, containing 7.5 x 10 utions of cells/plate. Plates were incubated at 37 C in an atmosphere of 10% C02. After 5 to 6 weeks, colonies greater than 40 cells were counted using an inverted phase microscope. Oncogenicity. Suspensions of cells diluted to l and 2 x 106 cells in 0.05 ml of culture medium were inoculated subcutan- eously into the interscapular region of newborn Sprague-Daw- ley and Fisher rats and nude mice were examined three times weekly for progressively growing tumors at the site of inoc- ulation. Six to eight weeks later, the animals were sacri- ficed and all palpable nodules and tumors were surgically ex- cised and prepared for routine tissue cultures and histologi- cal examination. Chromosome Analysis. Chromosome analysis was performed accord- ing to methods described in the literaturell'12 with minor modi- fications. Briefly, cells grown in T75 flasks in a logarithmic phase of growth were treated with 0.04 ug Colcemid/mlc for 1 89 hour before being harvested with trypsin solution. The cell suspension was then centrifuged at 1000 x g for 10 minutes. The supernatant was removed and washed twice with HBSS. After the last wash the supernatant was removed and, without dis- turbing the cell pellet, a hypotonic solution (distilled water) was added, the pellet resuspended, and the tube was incubated at 37 C for 15 minutes. After hypotonic treatment, the cells were fixed in methanol:glacial acetic acid (3:1) for 15 minutes and slides were prepared by dropping the cell suspension with a narrowed tip Pasteur pipet onto the slide; the preparation was subsequently rapidly air dried. G-banding was performed by a modification of the method of Yunis12 using a trypsiniza- tion time of about 15 seconds. Electron Microscopy. The cells grown in Petri dishes or T75 flasks were harvested by trypsinization or by using a fine, sharp spatula. To obtain perpendicular sections through the cell multilayers, the cells were fixed directly in a Petri dish and then processed. The entire film of cells was removed and fixed in 0.1% glutaraldehyde for 3 hours. Then the cells were refixed in 3% glutaraldehyde carefully to avoid disrupting the pellet. The pellet was fixed at room temperature for 2 to 5 hours. This second fixation hardened the pellet so that it could be handled more easily for dehydration and embedding. The pellet was then washed in 0.1M phosphate buffer, post- fixed in 1% osmium tetroxide for 1 hour, washed again with phosphate buffer and three times in 30% alcohol for 10 minutes each time. Each block was stained with uranyl acetate and lead citrate. The sections were examined on a Phillips 201 Electron 90 Microscope. For scanning electron microscopy (SEM), cultures were fixed in Petri dishes with glutaraldehyde and processed as de- signated above, except that the uranyl acetate was omitted. After dehydration with an ethanol series, fragments of the dish were cut out under alcohol and critical point dried in liquid C02. The specimens were coated with gold-palladium and examin- ed in a Cambridge 84-10 Scanning Electron Microscope. I Scoring of Detergent-Insoluble Cells and Cornified Cell Enve- lopes. This procedure was performed according to the technique used by Green.13 The squames shed into the medium were harvest- ed and centrifuged. The sedimented squames were resuspended in a small volume of buffer and counted in a hemocytometer chamber. The rate of squame production was calculated for the interval during which the accumulation took place every 3 days and cumu- lated at 20 days when the cells were subcultured. The cells shed were suspended in 12 m1 of 1% sodium dodecylsulfate (SDS); after incubation of 5 minutes, the suspension was centrifuged at 2000 rpm for 10 minutes, the supernatant was removed and the pellet was resuspended in 0.20 ml of buffer. A small aliquot was removed and the cells counted in a hemocytometer chamber using Nikon phase Optics. To count cornified envelopes, S-mer- captoethanol was added to the detergent-treated suspension to 1%, and after 5 minutes an aliquot was counted in the hemocyto- meter. The cell ghosts were then incubated at 90 C for 10 min- utes, centrifuged at 2000 rpm for 10 minutes and prepared for EM as described above. Rhodanile Blue Staining for Keratin. Cultures grown on Lab- 91 Tek tissue culture chamber slidesi and slides made from squame suspensions were fixed in 10% buffered formalin, stained with 1% Rhodanile Blue (a 1:1 mixture of 2% Rhodamine B and 2% Nile Blue A in water)j for 3 minutes and then destained in running tap water for about 30 seconds to one minute, until the excess red dye was eliminated from non-keratinocyte areas, leaving them light blue, and the keratinocyte colonies bright red.13’l4 i Ifiiles Laboratories, NapervilIe, IIIinois. j Matheson Co., Inc., Norwood, Ohio. 92 RESULTS A. Primary Cultures of Normal Rat Mammary Gland Two normal primary epithelial cell lines (SNLR/Ll and SNLR/CDl) from young virgin female Sprague-Dawley and Lewis rats were obtained. 1. Growth Characteristics. Examination under an inverted mi- crosc0pe revealed that the normal mammary epithelial cells grew from aggregates of ductular structures in DMEM with HS and hormone supplements (Figure 4-1). They attached to the surface of the culture dish within 20-50 hours and grew out very slowly as a cell multilayer. They grew well on plastic substrate but poorly on glass. Three distinct morphological cell types were observed after 5 days in culture. The first cell type was a cuboidal shaped cell, possibly of ductular or alveolar origin (Figure 4-2). The second type was a spindle-shaped, fibro- blast-like cell usually located between islands of cuboidal cells or even beneath the cuboidal epithelial colonies. The third type was a larger, ovoid-shaped lipocyte with numerous large intracytoplasmic lipid drOplets. The lipocytes and fibroblast-like cells were gradually eliminated during the first 2-5 weeks of culturing. The epithelial colonies con- tinued to expand, gradually coalescing and finally forming a confluent cell multilayer. The cells reached confluency in a 60 mm Petri dish in about 2 months. The final cell pOpulation was of epithelial type (NE): consisting of cuboidal-shaped cells forming a typical "cobblestone" appearance with dome formation (Figure 4-3). The NE cells had a strong cell to cell adherence and were resistant to trypsinization. The cell growth 93 Figure 4-1. Inverted microscopic appearance of NE free cells, cell aggregates and ductular structures, some closely resembling terminal end buds (arrow), six hours post-seeding after collagenase treat- ment. (x 60). Figure 4-2. NE cells colonies appearance, free of mesenchymal components, 4 weeks post-seeding, after repeated mechanical and differential trypsinization pro- cedures. (x 60). 94 Figure 4-1 Figure 4—2 '95 Figure 4-3. Appearance of confluent NE cell culture with cobblestone-like pattern and domes formation 2 months post-seeding. (x 120). Figure 4-4. Scanning electron microscopy of NE cells expanding large pseudopod-like membranes at the periphery and centrally forming stratified cell layers of squamous keratinized cells. (x 200). Z96 Figure 4-3 Figure 4-4 '97 characteristics illustrated that the population doubling time (PDT) decreased with increased passages (Table 4-1). For NE cells, PDT decreased from 95 (Lewis rat) and 100 hours (Spra- gue-Dawley rat) initially to 33 and 22 hours, respectively, after multiple passages. The seeding efficiency and plating efficiency increased with the passage numbers for NE Lewis rat cells. The seeding efficiency for NE Sprague-Dawley rat <:ells decreased slightly with passage numbers and the plating (efficiency increased with the passage numbers. The cells tuould not grow readily at low cell inocula and grew very slowly Ilntil a critical cell number (105/m1) was reached. The NE <:ells (Sprague-Dawley rat) which were maintained in culture up tr) 80 passages, and became transformed, greatly changed their pflaenotype. The cells no longer formed tight colonies, became aware adherent-independent, and piled-up, forming spheroids wTLich floated freely in the medium. Scuanning Electron Microscopy (SEM). Scanning electron micro- scopy confirmed these observations. The NE cells grew in a morualayer at the periphery, and had expanding large pseudOpod '1iJce membranes with centrally forming stratified cell layers ShOMfling varying degrees of differentiation toward squamous kera- tinized.cells (Figure 4-4). The cell monolayer was recognized by time cuboidal-Shape and the many microvillous projections on the snarface (Figure 4-5). The second layer, more differentiated ce1143. were flattened with polygonal outlines and fewer, short- ened microvilli. IEEEEEEission Electron Microscopy (TEM). Ultrastructural exam- lnatiOnrevealed three structural markers that may be used to Table 4-1. 923 Comparative characteristics of normal and neOplastic rat mammary epithelial cells Cell Line Passage Population Seeding Squame Plating Saturation Modal Tutori- Doubling Efficiency Production Efficiency Density Chromosome geniczty Time (4 hrs. (8 days (no.5cells N mber (Hrs.) post-sced- post-secd- x 10 sq. (per 50 ing) ing) mm.) metaphases % ‘ ‘ Ra:sx?i:s SILK/CD 1- s so 33(32-77)+ 1/5 (neoplastic) 9 70 38(28-70) 15 25 26 36(32-60) 3/5 33 22 36 21 96 6 6 6 38(32-80) 4/5 3’5 STLR/CDF 1" 3 82 (neoplastLC) S 80 15 74 37 25 5 2 46(36-88) 4/5 3/3 30 60 SXLR’L 1"' 6 95 (normal) 10 86 20 0.2 16 60 20 25 73 2 4 3 40(32-46) 0/5 0,- stR/CD 1' 6 108 (normal) 20 90 0.1 4 38(22-80) 30 60 2 40 40 64 60 25 50 4 4 36(32-88) 5/10 072 80 22 44 ‘ Derived from a Sprague-Dawley female rat “ Derived from a Fisher female rat "' Derived from a Lewis female rat * Range in parenthesis 99 identify mammary epithelial cells: desmosomes, tonofibrils, and intracytoplasmic vacuoles.15 Ultrathin sections, cut per- pendicular to the surface of the outgrowth, formed a strati- fied squamous epithelium (Figure 4-6). The basal layer ex- hibited multiple well developed villous projections. The cells from the upper layers showed either short villous pro- jections or an absence of villi. Cells were connected by numerous desmosomes with attached tonofilaments (Figure 4-7). Similar tonofilaments were also present in the cytoplasm and especially as a perinuclear rim.- The number of mitochondria, smooth endoplasmic reticulum, rough endoplasmic reticulum, free ribosomes and Golgi apparatus was variable. Squame Production in Confluent (NE) Cell Cultures. Squame production was observed in NE cells when EGF was present in the medium. The cells from stratified epithelium (2-3 layers) gradually lost their nuclei, became more flattened, partially keratinized, contained a large number of phagosomes and an increased density of microfilaments resembling the filaments observed in normal human mammary epithelial cells.15 These were the squames or keratinized cells which shed into the medium, depending on how long they were kept at confluency. About 2% of the total cells were shed into the medium during a 20-day post-seeding period. Chromosome Analysis. The NE cells maintained a near diploid chromosome constancy (modal 38) with a range between 32 to 80 (Table 4-1). The chromosome pattern was characteristic of the rat species. 100 Figure 4-5. Scanning electron microscopy of NE cells from the basal layers showing a cobblestone-like pattern with numerous microvilli projections on the surface. (x 2600). Figure 4-6. Ultrathin section, cut perpendicular to the surface of the outgrowth revealed that the NE cells formed a stratified squamous epithelium. The second cell layer showed increased intracytoplasmic perinuclear tonofilaments. (x 700). 3“ :2; I ' '— .‘ mares“ - .. Figure 4-5 Figure 4-6 102 B. Primary Cultures of ENU-Induced Rat Mammary Gland Adeno- carcinoma Cells Two primary neOplastic epithelial cell cultures were ob- tained from adenocarcinomas induced in Sprague-Dawley and Fisher rats by inoculation with ENU (MA cells). Growth Characteristics. After collagenase-treatment, the MA cells grew in culture from cell aggregates and continued to grow as a cell multilayer and also showed a faster rate of growth than NE cells for the first passages. The MA cells formed domes and piled-up colonies, forming spheroids of cells floating into the medium after approximately 5 passages (Fig— ure 4-8) . Within the epithelial colonies, cells were mainly cuboidal, possessing large nuclei with prominent nucleoli and Striking perinuclear granulation, closely adherent to one another and resistant to trypsinization procedure. The PDT for MA cells decreased from 80 and 82 to 40 and 22 hours, and the seeding efficiency increased with passages. The plating efficiency and saturation density was slightly higher than that of the NE cells (Table 4-1) . SC3&1ning Electron Microscppy (SEM) . The MA cells formed cell InIll-Iizilayers and multifocal spheroids. The MA cells continued to divide at confluency, settled on. top of the formed layer, forming spheroids (Figure 4-9) . Trflmmission Electron Microscopy (TEM) . Ultrastructural examin- ation of MA cells revealed the same markers as were. observed with NE cells. The MA cells formed a stratified squamous epithelium up to 5 layers thick. An increased accumulation of filaments and electron dense bodies in the cytoplasm of shed cells with 103 gradual degeneration and disappearance of nuclei, ribosomes, mitochondria and alteration of desmosomes has been observed. Occasionally, blister-like structures (domes) were formed be- tween the upper stratified squamous epithelium (Figure 4-10). Round, dense intracytoplasmic bodies, probably representing li- pid or phagosomes, were constantly present; they were large and more numerous in stratified cells. Some of these electron dense bodies formed large intracytoplasmic aggregates in the stratified cells which caused the superficial layer to have a dome-like appearance (Figure 4-11). Squame Production in MA Cell Cultures. The process of squames formation in cultures containing EGF was increased in MA (es- pecially in STLR/CDFl) cells than in NE cells (Table 4-1). Addition of 10 ng/ml of EGF to the tissue culture medium sub- stantially increased the rate of cell replication and squame formation compared with tissue culture medium without EGF (Fig- gure 4-12). The percentage of squames harvested during 20 days period ranged from 6 to 25% for MA cells. About 70% of the squames harvested consisted of cells with pycnotic nuclei or without nuclei. Treatment of squames with 1% sodium dodecyl- sulfate (detergent) and 1% B-mercaptoethanol (reducing agent) showed cell ghosts with thick cornified envelopes. The well preserved marginal band was clearly visible on electron photo- micrographs (Figure 4-13). Rhodanile blue staining indicated that the shed cells were partially keratinized. Chromosome Analysis. The MA cells derived from Sprague-Dawley rat were in a hypodiploid range with a modal number of 38 and a range of 22 to 80. The MA cells derived from a Fisher rat 104 Figure 4-7. Transmission electron microscopy appearance of NE cells showing numerous surface microvillous projections (M), desmosomes with attached tono- filaments (D), and intracytoplasmic vacuoles (V). (x 22000). Figure 4-8. Inverted microscope appearance of MA cells forming a large piled-up colony. (x 60). 105 Figure 4-8 106 Figure 4-9. Scanning electron microscopy appearance of MA cells showing squamous differentiation and mito- sis on the top of differentiated layer from which piled-up colonies will form. (x 200). Figure 4-10. Transmission electron microscopy appearance of MA cells with blister-like structure (dome) formed between the upper stratified squamous epithelium. (x 7000). 1'37 Figure 4-10 108 Figure 4-11. Transmission electron microscopy appearance of in- tracytoplasmic electron dense body aggregates in in the stratified cells which occasionally caused a "dome-like" appearance of the tissue culture. (x 7000). Figure 4-13. Photomicrograph.of squames showing marginal bands (post-treatment with detergent and reducing agents). (x 7000). 109 Figure 4-12 Figure 4-13 110 Figure 4-12. Graphic representation of the EGF effect on MA cell cultures. The squames formation was sig- nificantly increased on tissue cultures contain- ing EGF in the medium. lll +EGF :IQJRV7.no no 4. go no. 32300 .jmo m2<30m 5 E 20 DAYS Figure 4-12 112 were in a near diploid number with a modal number of 46 and a range of 36 to 88 (Table 4-1). Soft Agar Assay. MA cells grew readily in soft agar culture at later passages (after passage 10). Cell passages under 10 grew slowly and formed smaller and fewer colonies (0.05%). MA cells formed colonies in soft agar which were counted after a period of 5 weeks. Colonies containing 40 or more cells were scored, and ranged in size from 1.2 to 200 um. Two of the multicellular colonies were removed with a Pasteur pipet and successfully replated in monolayer culture. NE cells did not form colonies in soft agar. Oncogenic Potential; Oncogenic potential differed with the passage number of MA cells (Table 4-1). Earlier passages of MA cells (up to 10 passages) inoculated into isologous new- born rats and nude mice gave poor tumor yield but tumor yield increased up to 60-80% at passages beyond 15. The histological appearance of these tumors was carcinoma and adenocarcinoma. One of the NE cell cultures (Sprague-Dawley rat) at passage 80 gave an incidence of 50% tumors which indicates that they be- came spontaneously transformed in vitro. The other NE primary cells (Lewis rat) up to passage 20 didn't show any indication of spontaneous transformation. The inoculum was negative, even in concentrations of 6 x 106 per newborn rat. 113 DISCUSSION The results of the present study indicate successful de- monstration of in vitro growth of NE cells and for the first time MA cells derived from ENU-induced rat mammary adenocarci- noma. The cells grew in DMEM containing 25% HS, supplemented with hormones and EGF. Both the NE cells and the MA cells grew as a cell multilayer rather than monolayer, forming up to 5 layers of stratified squamous epithelial cells. EGF increased proliferation and squame formation and induced keratinization in cultured MA cells, an observation heretofore not reported in rat MA cells. The colonies usually eXtended laterally by cell proliferation and movement, while differentiation began in the center and expanded on the whole surface at confluency, where upper cells formed cornified envelopes which were insol- uble in detergents and reducing agents. EGF has been reported to increase the rate of cell replication in human breast can- 16-19 21,22 cer cells in vitro and of mice mammary gland explants and to induce differentiation with squame and keratin formation of cultured human epidermal cells.l3'23'24 At confluency, an increased number of cells were in terminal differentiation phase and when these cells were subcultured they no longer attached to the dish substrate. Those cells accounted for the greatly increased percentage of squame production after 2-3 days when the medium was changed for the first time and the squame quantitated. Then the curve of squame production con- tinued to decrease slightly to the 12th day postculture and raised again toward the 20th day when the cells were subcul- tured (Figure 4-12). 114 This phenomenon in addition to strong adherence to the substrate and to each other makes it difficult to evaluate growth parameters with accuracy. As mentioned earlier, ter- minal differentiation occurs and this can lead to loss of up to 25% (MA) of the cell population in 20 days in confluent cultures. The light microscopic and ultrastructural charac- teristics of the cultured mammary epithelial cells (NE and MA) resemble those observed in the epidermis in vivol3'l4’23' 24 but distinct differences in the morphology of cornified squames indicate an incomplete maturation process. The squames formed in culture had Only loosely packed filaments rather than a typical dense keratohylin matrix. Cultures of normal or neoplastic cells grew as a cell multilayer with formation of domes and piling up (MA) similar to features described in human and rodent epithelial cells.15’ 25’26'28 Overlapping and multilayering in culture vessels reveals that the cultures are not conventional monolayers. Multilayering occurs in the initial outgrowth from some ex- planted epithelial tissues 13'14'23'24 but has not commonly been reported in primary cultures of dissociated normal epi- thelium. This seems to be a characteristic of the growing pattern of some epithelial cells rather than a consequence of excessive crowding. Dome Formation. This is a characteristic of many epithelial 23,26,28,29,30 cell monolayers. Some authors have suggested that o o o .29 domes represent in VltrO analogues of mammary secretory ac1n1 and that their presence is due to local variations in transcel- lular transport mechanisms.28 As mentioned above, the cells, during earlier passages, form numerous droplets of unidentified ‘“3 fine 115 structures (possible phagosomes) which sometimes accumulated in the stratified cells and could have mediated dome formation. Also, domes were recognized in the stratified cells which were ready to shed, cells which probably lost their transcellular capabilities. Why domes form when and where they do is still an insufficiently explained phenomenon. Growth characteristics provide insufficient information to distinguish between normal and malignant epithelial cells. However, the ability of cells to grow in soft agar and to form colonies appears to be a reliable indicator of malignancy of ENU-induced MA cells. Consistent with this view is our observation that the NE cells failed to grow in soft agar whereas MA cells grew in the agar. Interestingly, the NE cells (SNLR/CDl) when transplanted to isologous hosts or nude mice formed palpable anaplastic carcinomas (P80). It appears that at least a portion of the NE cells are very sensitive to spontaneous neoplastic trans- formation in vitro. MA cells from the earlier passages (up to 10) usually do not produce tumors when they are inoculated into isologous newborn rats. It seems that neoplastic cells originating from ENU-induced adenocarcinomas required multiple passages in vitro until they can induce a high yield of tumors in the appropriate hosts. The chromosomal pattern was typical of rodentia species. The NE cells (or even those NE cells that spOntaneously trans- formed in vitro) maintained a near diploid modal chromosome num- ber (38 to 40 chromosomes). The MA cells maintained primarily hypodiploid modal number (38 chromosomes). Quantitaion of DNA 116 by cytofluorometry (see Chapter 5) revealed that rat NE cells in culture had a reduced amount of DNA compared to peripheral blood lymphocytes and fresh, dissociated mammary epithelial cells. Karyotypic analysis showed that these cells were also hypodiploid. Thus in the process of becoming established in cell cultures, rat NE cells lost DNA and chromosomes. These changes may be essential for stable, ip'yitgg growth. The origin of mammary epithelial cells involved in car- cinogenesis continues to be a debatable topic. Information regarding the region of the ductal tree from which the epithel- ium was obtained may be important with respect to an under- standing of carcinogenesis. The ductal system starts at the skin level as large ducts in the nipple branching downward into smaller ductules and, during pregnancy, presumably de- velOping terminal alveoli.31 In ultrastructural studies of terminal ductules, investigators have not been able to agree on the number of cell types comprising the epithelium.ls’26'32 However, a cell has been consistently recognized containing bundles of cytoplasmic filaments resembling epidermal tonofil- aments.15’27'28’31’32 These cells were also connected by des- mosomes. The cells observed in our experiments had similar characteristics. It is possible that they derived from fila- ment-containing cells of the duct. It appears from the available data that rodent mammary tissue cultures, even with the inconsistency of frequent :2 2:: Egg spontaneous neoplastic transformation, remain a valuable source of information for the understanding of the complexity of malignant mammary neoplasia. 117 REFERENCES l. Huggins CB: Experimental leukemia and mammary can- cer. In: Cancer: Induction, Prevention and Cure. Chicago, University of Chicago Press, 1979, pp 73-98. 2. Gallino PM, Pettigrew HM, Grantham: N-nitroso-meth- ylurea as mammary gland carcinogen in rat. J Natl Cancer EEEE 54:401-414, 1975. 3. Cohen LA, Tsuang J, Chan PC: Characteristics of rat normal mammary epithelial cells and dimethylbenzanthra- cene-induced mammary adenocarcinoma cells grown in monolayer culture. In Vitro 10:51-62, 1974. 4. Stoica G, Koestner A, Capen C: Characterization of N-ethyl-N-nitrosourea-induced mammary tumors in the rat. fig J Pathol 110:161-169, 1983. 5. Stoica G, Koestner A, Capen C: Neoplasms induced with high single doses of N-ethyl-N-nitrosourea in 30-day- old Sprague-Dawley rats with special emphasis on mammary neo- plasia. AntiCancer Res 4:5-12, 1984. 6. Stoica G, Koestner A: Diverse spectrum of tumors in male Sprague-Daney rats following single high doses of N-ethyl-N-nitrosourea. Am J Pathol 116:319-326, 1984. 7. Freshney RI: Culture of Animal Cells. New York, Alan R. Liss, 1973. 8. Pastan IH: Cell Culture. In: Jakoby (ed): Meth- ods in Enzymology, Vol LVIII, New York, Academic Press, 1973, p 150. 9. Puck TT, Marcus PI: A rapid method for viable cell titration and clone production with HeLa cells in tissue cul- 118 ture: The use of x-irradiated cells to supply conditioning factors. Proc Natl Acad Sci USA 41:432-437, 1955. 10. MacPherson I: Soft agar techniques. In: Kruse and Patterson (eds): Tissue Culture Methods and Applications, New York, Academic Press, 1972, pp 276-280. 11. Woranooj U, Hsu TC: The C- and G- banding patterns of Rattus norvegicus chromosomes. J Natl Cancer Inst 49:1425- 1431, 1972. 12. Yunis JJ: New chromosome techniques in the study of human neoplasia. Human Pathol 12:540-549, 1981. 13. Allegra JC, Lippman ME: Growth of a human breast cancer cell line in a serum-free hormone-supplemented medium. Cancer Res 38:3823-3829, 1978. 14. Cohen LA: Isolation and characterization of a seri- ally cultivated neoplastic epithelial cell line from the N- ILitroso-methylurea induced rat mammary adenocarcinoma. £2 Vitro 18:565-575, 1982. 15. Behring GC, Hackett AJ: Human breast tumor cell liuaes: Identity evaluation by ultrastructure. J Natl Can- cer Inst 53:621-629, 1974. 16. Osborne CK, Hamilton B, Titus G, Livingston RB: E:EPidermal growth factor stimulation of human breast cancer calls in culture. Cancer Res 40:2361-2366, 1980. 17. Osborne CK, Hamilton B, Nover M, Ziegler J: Anta- Sflbnism between epidermal growth factor and phorbol ester tu- mor‘promoters in human breast cancer cells. J Clin Invest 57:943-951, 1981. 18. Imai Y, Leung CKH, Friesen HG, Shiu RPC: Epider- 119 mal growth factors receptors and effect of epidermal growth factor on growth of human breast cancer cells in long-term culture. Cancer Res 42:4394-4398, 1982. 19. Stoker MGP, Pigott D, Taylor-Papadimitriou J: Re- sponse to epidermal growth factor of cultured human mammary epithelial cells from benign tumors. Nature 264:764-767, 1976. 20. Turkington RW: The role of epithelial growth fac- tor in mammary gland development in vitro. Exp Cell Res 57: 79-85, 1969. 21. Tonelli QJ, Sorof S: Epidermal growth factor re- quirements for development of cultured mammary gland. Nature~ 285:250-252, 1980. 22. Turkington RW: Stimulation of mammary carcinoma cell proliferation by epithelial growth factor. Cancer Res 29:1457-1458, 1969. 23. Fusenig NE, Breitrkreutz D, Lueder M, Boukamp P, Worst PKM: Keratinization and structural organization in epidermal cell cultures. In: Schweiger (ed): Internation- al Cell Biology 1980-1981, New York, Springer-Verlag, 1981. 24. Fusenig NE, Worst PKM: Mouse epidermal cell cul- tures. II. Isolation, characterization and cultivation of epidermal cells from perinatal mouse skin. EprCell Res 93:443-457, 1975. 25. Greiner JW, DiPaolo JA, Evans CH: Carcinogen-induc- ed phenotypic alterations in mammary epithelial cells accom- panying the develOpment of neoplastic transformation. 932: cer Res 43:273-278, 1983. 120 26. Bennett DC, Peachey LA, Durbin H, Rudland PS: A possible mammary stem cell line. Cell 15:283-298, 1978. 27. Russo J, Furmanski P, Bradley R, Wells P, Rich MA: Differentiation of normal human mammary epithelial cells in culture: An ultrastructural study. Am J Anatomy 145:57-78. 28. Voyles BA, McGrath CM: Markers to distinguish nor- mal and neoplastic mammary epithelial cells in zitrg: Com- parison of saturation density, morphology and concanavalin A reactivity. Int J Cancer 18:498-509, 1976. 29. Pickett PB, Pitelka DR, Hamamoto ST, Misfeldt DS: Occluding junctions and cell behavior in primary cultures of normal and neOplastic mammary gland cells. J Cell Biol 66: 316-332, 1975. 30. McGrath CM: Cell organization and responsiveness to hormones in zitgg: Genesis of domes in mammary cell cul- tures. Am Zool 15:231-236, 1975. 31. Flaxman BA, Van Scott EJ: Growth of normal mammary gland epithelium in vitro. Cancer Res 32:2407-2412, 1972. 32. Whitehead RH, Bertoncello I, Webber LM, Pedersen JS: A new human breast carcinoma cell line (PMC42) with stem cell characteristics. I. Morphologic characterization. J Natl Cancer Inst 70:649-661, 1983. CHAPTER 5 N-ethyl-N-nitrosourea-induced in vitro neoplastic transformation of rat mammary epithelial cells 121 122 SUMMARY Normal mammary epithelial cells (NE), originating from Sprague-Dawley (CD) and Lewis female rats, were grown in Dul- becco's Modified Eagles Medium (DMEM) containing 25% horse serum (HS) and hormone supplements. At passage 3, after most of the cells were selectively separated as epithelial cell monolayer, the cells were treated with N-ethyl-N-nitrosourea (ENU) in various doses (25, 50, 100 ug ENU/ml) to study the process of in vitro mammary epithelial neoplastic transforma- tion. The immediate effect of ENU on NE cells was monitored for cellular DNA content measured by flow cytometry, DNA syn- thesis by (3H) thymidine incorporation assay and chromosomal aberrations using Giemsa banding techniques. The NE cytofluorometric histograms showed a hypodiploid population. Depending on doses, after ENU exposure, the NE showed a second peak of cells with DNA content in tetraploid or octaploid range. The ENU-exposed NE cells (NET) regained their hypodiploid pattern after 75 to 120 hours. DNA synthe- sis histograms showed that NET cells (25, 50 and 100 ug ENU/ ml) increased DNA synthesis above the control and, depending on doses, blocked mitosis up to 75 to 120 hours. The chromo- somal study showed a dose and time effect relationship. Most of the chromosomal aberrations, isochromatid breaks, were at 6 hours post-exposure. Post-ENU exposure, the cell counts and cytology of NET cells showed an increased number of cells and multinucleation compared with NE. The sequence of phenotypic alteration (termed "stage I-V") was monitored by daily observation. The NET, after a period 123 of approximately 30 days post-exposure, showed a marked pro- liferation of morphologically altered cells (piled-up colonies) which subsequently acquired the capacity to form colonies in soft agar and finally became tumorigenic (stage V) when the cells were transferred into an isologous host. The histology of these tumors resembled highly anaplastic carcinomas. INTRODUCTION Ethylnitrosourea (ENU)-induced carcinogenesis in zitrg has been investigated by direct application of carcinogen to fibroblasts and neurogenic cells}.3 ENU is a potent alkyl- ating agent, and in contrast to some other carcinogens it is thought to act directly, not requiring enzymatic activation. 2’4 2'5 it is con- Since the half-life in ziyg is 30 minutes, sidered to act relatively rapidly both in gizg and in vitro. Possible targets in the cell are DNA, various species of RNA, and a‘variety of proteins.4’6'7 The interval between primary carcinogen-cell interaction and the onset of tumor growth represents a largely unidenti- fied phase in the process of carcinogenesis, both at the mole- cular and cellular levels. In this phase many characteristic phenotypic and constitutional alterations occur in the presump- tive cancer cells. For a better understanding of this phase of neoplastic transformation in yiyg experiments would be bene- ficial. However, this approach is hampered by the complex cel- lular composition of intact tissues, together with the fact that only a minor fraction of their constituent cells undergo the alterations ultimately resulting in the expression of a 124 "malignant" phenotype. The problem is further aggravated when dealing wiht epithelial systems because of the difficulty to separate the epithelial cells from the mesenchymal compo- nent and grow them as continuous cell lines i§_yitgg. Cul- turing of mammary epithelial cells (NE) i2_yi££g, therefore, constitutes one of the most difficult problems in the study of in vitro neoplastic transformation. In recent years a number of papers dealing with in zitrg neoplastic epithelial transformation, using mostly tissue ex— plants were rep'orted.8-13 There have been only a few attempts of in yitgg neoplastic transformation of mammary gland epithel- ium. Dao used rat organ mammary culture exposed to 7,12-di- methylbenz(a)anthracene as a chemical carcinogenl4 while Co- hen used dissociated mammary cells exposed to nitrosomethyl- urea.15 The multistage process of carcinogenesis is well recogniz- ed. Cellular changes associated with the initiation and pro- motion of neoplasia are studiedmost easily in yitgg, where the environmental factors can be controlled and cellular re- sponses monitored. Such studies have been difficult to inves- tigate with epithelial cells in gitgg since there was no satis- factory model system. Recently, a culture system has been described that satis- fies the criteria required for an in vi££g_model of carcino- gen-induced mammary gland epithelial cell neoplasia. The cells grown in yitgg have been identified by morphologic, scanning and transmission electron microsc0pic criteria as epithelial in origin.16 The cultured cells, of ductular or 125 alveolar origin, have been derived from target cell types of a tissue known to be susceptible to ENU-induced neoplasia in yizg. In this system, cells were treated in primary culture with ENU soon after isolation from the animal. The NE cells survived in vitro for extended periods of time (months rather than days) which is important for analysis of long term changes following exposure to carcinogens. One should be aware, how- ever, of the frequent "spontaneous" neoplastic transformation in in 22552 rat systems. In the present study, we investigated the earliest ENU- induced structural and morphological changes in cultured rat mammary epithelial cells. DNA content, rate of DNA synthesis, chromosomal aberrations and cell counts were analyzed at vari— ous times up to 120 hours post-ENU exposure. The multistep phenotypic alterations of rat mammary epithelial cells were recorded and morphologically altered cells were implanted into compatible hosts for in vivo carcinoma development. MATERIALS AND METHODS Source and Isolation of Cells Epithelial cells for primary normal mammary gland (NE) tissue culture were obtained from the whole mammary gland fat pad (inguinal and axillary) of normal female, specific-patho- gen-free (SPF) Sprague-Dawley (CD) and Lewis ratsa at 50 days of age by previously described techniques.l6 Briefly, after F" P.) CA collagenase (0.1%) treatment, the cell suspension was filtered through one layer of 74 micron opening nylon meshb and seeded in a 60 mm Petri dish containing DMEMC supplemented with 25% HSC and a hormone combination consisting of 20 ug/ml each of hydrocortisone, prolactin and insulin,d and 10 ng/ml of epi- thelial growth factor (EGF)e, and 1% antibiotics (penicillin and streptomycin).c The cells were incubated in a humidified 10% CO2 atmos- phere at 37C. At about 8 to 10 hour intervals, the suspended cells in each dish were removed from the incubator and decant- ed to another dish. This process was repeated about 5 times. For further separation of epithelial cells, differential tryp- sinization and mechanical separation was applied to the dishes containing epithelial colonies. In about 2 months, dishes with pure epithelial colonies were obtained. Carcinogen Pulse At passage 3 (P3), the tissue culture dishes were treated with 25, SO, 100.and 500 ug/ml of the carcinogen and mutagenl7 acylalkylnitrosamide N-ethyl-N-nitrosourea (ENU). ENU cry- stals were dissolved shortly before administration in phos- phate/citrate-buffered saline at pH 4.2 (1 part buffer to 14 parts saline) and added to the dishes containing medium with- out serum. The duration of treatment was for a 2 hour period, Charles River Laboratory, Portage, Michigan. Cistron Corporation, Elmsford, New York. GIBCO, Grand Island, New York. Sigma Chemical Company, St. Louis, Missouri, Collaborative Research, Waltham, Massachusetts. (DL‘LOU‘D’ 127 after which the supernatant was removed, the cells washed twice with Hank's Buffered Saline Solution (HBSS)c and refed with complete medium. Characterization of NE Treated Cell Lines (NET) The morphologic appearance of neoplastic cell lines was monitored by phase-contrast microscopyf and recorded photograph- ically. Tumors obtained after reimplantation of NET cultUred cells were characterized by their histologic appearance. Flow Cytometric Deterimination of Cellular DNA Content NE cells P3, 1 x 106 were diluted in hypotonic prOpidium 18 Peri- iodide solution (0.05 mg/ml in 0.1% sodium citrate). pheral blood leukocytes (PBL) and a suspension of freshly dis- aggregated rat mammary tissue cells were treated in an identi- cal manner to determine where the 2N peak occurred in the DNA histograms. Samples with greater than I% cell clumps were discarded. The instrument used was the Coulter Cell Sorter Model TPS-Large Laser-Two Color System.9 The argon-ion laser emitted light at 488 nm and the output power was set at 400 nm. Longpass filters used were 515 and 630 nm. Amplification was set so that the 2N modal peak of normal PBL and disaggre- gated rat mammary tissue peaks occurred in channel number 6. The coefficients of variation for the PBL and mammary cells were 12.1% and 14.2% respectively. Data were collected for each sample such that 10,000 cells were counted in the peak channel. Samples were collected at 24, 48, 72 and 96 hours f Nikon Diaphot-TMD, Nippon Kogaku K.K., Tokyo, Japan. 9 Coulter Electronics, Hialeah, Florida. 128 after exposure to 0, 50, 100 and 500 ug ENU/ml of culture medium. DNA Synthesis Primary NE at P3 and P4 were seeded in 60 mm Petri dishes at concentrations of 7 x 105/ml. When the cells reached ap- proximately 50-60% confluency, the treatment with ENU was applied as described above in doses of O, 25, 50 and 100 ug ENU/ml of culture medium. Three dishes from each dose level were pulse labeled with (3H) thymidine (2uCi/ml) at 24 hour intervals up to 120 hours. The cells were incubated at 37C for 1 hour, trypsinized and processed for the (3H) thymidine incorporation assay. The collected cells were rinsed several times with HBSS, then incubated with 1 ml of 5% trichlorace- tic acid (TCA) at 37C for 1 hour and centrifuged at 4430 g for 15 minutes at 37C. TCA-insoluble material was dissolved in 0.2 ml of 0.2 N NaOH for 15 minutes at 37C. Finally, 100 mi- croliter aliquots (TCA-soluble and TCA-insoluble in NaOH) were transferred to scintillation vials and counted in 9.00 ml of an aqueous counting scintillation fluid. Cell Counts Three dishes from each dose level were counted using try- pan blue dye exclusion technique and hemocytometer at 24 hour intervals up to 120 hours. Chromosome Analysis Chromosome analysis was performed according to established 19'20 with minor modifications. Briefly, 66113 in a methods logarithmic phase of growth were treated with Colcemid 0.04 ug/mlC for 1 hour then harvested with 0.25% trypsinC solution. 129 The cell suspension was then centrifuged at 114 g for 10 min- utes and the supernatant was removed and, without disturbing the cell pellet, distilled water was added and the pellet re- suspended, and incubated at 37C for 15 minutes. Following hypotonic treatment, the cells were fixed in methanol:g1acial acetic acid (3:1) for 15 minutes and slides were prepared by dropping the cell suspension with a narrowed Pasteur pipet from a height of 3-4 feet above the slide and rapidly air dried. G-banding was performed by a modification of the meth- 20 od of Yunis with a trypsinization time of about 15 seconds. ENU-induced chromosomal aberrations were scored accord- ing to World Health Organization (WHO)21 and others.3’23 Colonnyormation and Cloning in Semisolid Agar Medium The ability of rat mammary epithelial cells to form colo- 25 with nies in agar was determined as described by MacPherson minor modifications. The basal layers contained 5ml of 2% Sea Plaque Agaroseh supplemented with DMEM, 25% HS and 0.2% sodium bicarbonate in 60 mm plastic Petri dishes. A second layer (1.5 m1) of 0.33% Sea Plaque Agarose supplemented as above, containing 7.5 x 104 dilutions of cells/plate was added. Plates were incubated at 37C in a 10% C02 atmosphere. Five 1 to six weeks after plating, colonies containing greater than 40 cells were counted using an inverted phase microsc0pe. Cloned sublines of established "parental" tumorigenic cell lines were obtained by aspiration of single colonies from the cloning medium with a PaSteur pipet followed by trypsinization for l-2 minutes then transferred to a 60 mm Petri dish for further culture. 130 Oncogenicity Suspensions of cells (stage V NET and control NE) dilu— ted to l to 2 x 106 cells in 0.05 ml of culture medium were inoculated subcutaneously into the interscapular region of 10 newborn CD and Lewis rats and lO athymic nude mice. The rats were examined three times weekly for progressively growing tumors at the site of inoculation. Six to eight weeks later, the animals were sacrificed and all palpable nodules and tu- mors were surgically excised and prepared for routine tissue culture, electron microscopy and histological examination as described above. Statistics Quantitative changes were assessed for significance by the two way analysis of variance followed by the Duncan's Multiple Range Test post-test. Changes were considered sig- nificant' if p\<0. os. h Marine Colloids DivISion, FMC Corporation, Rockland, Maine. 131 RESULTS Untreated NE in Culture (COntrols) The NE cells were obtained from CD female rats and main- tained in culture as was described in Materials and Methods. After approximately two months, a pure epithelial cell culture consisting of cuboidal cells arranged in a "cobblestone-like" pattern and showing dome formation at confluency was obtained (Figure 5-1). The morphological and electron microscopic characterization of these cells were previously reported.16 The NE cells had a slow rate of growth for the first 3 pass- ages. Although the growth rate increased in later passages it remained slower than that of the NET cells. The pOpulation doubling time (PDT) decreased with increased passages, and the plating efficiency remained low when compared to NET cells. The NE cells did not form colonies in soft agar and did not produce tumors when implanted into appropriate hosts, with the exception of one cell line which continued to grow, forming piled-up colonies at P60' grew in soft agar and induced tumors (P80) in 50% of newborn CD rats upon inoculation. Five other primary epithelial cells were maintained in culture for about 2 -4 passages when they degenerated and died. Another NE cul- ture, derived from a Lewis rat, continued to grow and at P35 showed no indication of.transformation (4 months in culture). ENU-Treated Cells (NET) and DevelOpment of NeOplastic Mammary Gland Cell Lines In contrast to the control cultures, NET cells, from 2 cell lines exposed to ENU in vitro, underwent a characteristic and reproducible sequence of phenotypic alterations preceeding 132 Figure 5-1. Inverted microscopic appearance of confluent nor- mal mammary epithelial cells culture with cobble- stone-like pattern and dome formation. ( x 120). 1_33 Figure 5-1 134 the expression of neoplastic properties (Figure 5-2). The duration of the individual phases of this process, termed l 8 . . . ' was Similar in both cultures. There were "stage I-V", variations in time sequences relative to the dose of ENU. For instance the recovery period following exposure to 500 ug ENU/ml was longer when compared to doses of 25, 50 and 100 ug ENU/ml; at doses of 500 ug/ml 50% of the cells died out and small epithelial-like colonies were observed to develop among the larger sensecent cells after a latency period of 12 days. Stage I During the early phase of observation (up to 6 days), primary NET cells showed slight differences from the corres- ponding controls. An increased number of multinucleated cells (binucleated and trinucleated) in NET cells (50 and 100 ug ENU /ml) was observed after 72 hours post-treatment in contrast to controls (Figure 5-3). Depending on dose, the NET cells showed a slight to severe altered morphology, consisting of large, polyhedral or heterogeneous cells with loose cell to cell ad- herence (senescent cells). The monolayer failed to maintain the tight "cobblestone-like" pattern. The multinucleated and senscent cells died out after a short period of time (approx- mately 5-10 days). Stage II After a few days (7-15 days), isolated colonies of small cuboidal epithelial cells developed and maintained this morph- ology throughout stage II (Figure 5-4). These cells had a fas- ter rate of growth than NE cells although multilayer growth and formation was similar to the homogeneous cuboidal appearance of 135 Figure 5-2. Diagrammatic representation of the in vitro-in vivo system for pulse carcinogenesis by ENU in rat mammary epithelial cells. In Culture Carcinogen 136 Gradual appearance ol malignant properties Pulse (Morphology. Increased rate at proliferation. Loss of anchorage de- pendenceL Mammary epithelial cells (primary cul- ture) pleimorphic population Stationary (or slowly prolifer- P ating) mammary 3 E.C. Formation oi confluent layers ol E.C. (cuboidal mm P. l fit ~ ~ 2 months ENU Pulse in Vivo Tumor formation after reim- implantation ol cultured cells ENU pulse (180 mglkg ip) 50 days O 1.5 months Tumors alter reim- plantation into newborn rats and athymic mice (5- 6 wke.) .v@ e 1 Slow prolileration oi E.C. (cuboidal type) as conlluent cell layers Dillicuit to trypsinize Figure 572 Transplantable MT into newborn CD rats and nude mice Colony formation in salt agar Focal prolifer- ation (piled up loci) Pa ( 30 days post treatment) Reduced adherence to continent cell layer 137 Figure 5-3. Stage I NET cells (ENU 100 ug/ml) 72 hours post -treatment showing increased numbers of senescent and multinucleated cells. (x 200). Figure 5-4. Stage II NET cells (ENU 100 ug/ml) 7 days post- exposure. Epithelial colony (center) surrounded by sensecent cells. (x 200). 138 ‘ ,\~ "5! Figure 5-4 139 NE cells. This phase extended from 7 to 30 days post ENU- exposure. Stage III A subsequent phase of moderate proliferation resulted in the formation of piled-up foci (spheroids) of epithelial cells, 30 days post-exposure to ENU (Figure 5-5). These "piled-up" foci persisted and slowly grew larger and were easily removed when flushed gently with medium. These cells maintained their previous morphologic appearance, grew as cell multilayers with a tendency to form piled-up colonies. The piled-up colonies were sensitive to trypsin during this period. The overall PDT was decreased and the maximum cell number per unit was increas- ed in this phase._ Implants of stage III NET cells into newborn rats failed to grow. Soft Agar Assay (Stage IV) Spheroids from NET cells were collected by flushing and were recultured for 2 more passages in liquid medium and then seeded into soft agar. The NET cells formed colonies in soft agar cultures and were scored after a period of 5-6 weeks. The average initial cloning efficiency was 0.2% (P6) and then increased to 4% (P8). The colonies diameters ranged in size from 1.2 to 200 um (Figure 5-6). Two of the multicellular colonies were removed with a Pasteur pipet, trypsinized and replated in liquid media. The cells grew as a cell multilayer and retained the ability to pile-up. Tumorigenicity (Stage V) The tumorigenicity of the cells was assessed by subcu- 6 taneous inoculation of cell suspensions at l to 2 x 10 cells Figure 5-5. Figure 5-6. 140 Stage III NET cells (ENU 100 ug/ml) 30 days post -exposure. A focus of active cellular prolifera- tion precluding development of piled-up colony. (x 48). Soft agar colony of NET cells 5 weeks post-seed- ing. (x 48). I—--.. __,. 141 Figure 5—5 Figure 5-6 142 into newborn isologous rats. After 6 weeks post-inoculation, tumors developed at the site of inoculum in 8/10 Lewis and 9/ 10 CD inoculated rats and 5/10 nude mice. The gross, histo- logic and electron microsc0pic examination of the tumors re- covered from the implanted rats revealed a carcinomatous pat- tern (Figure 5-7,8). The cell lines obtained from the reim- planted NET cells not only phenotypically resembled the ori- ginal cells, but they also maintained the characteristic of piling-up. The time to move from stage I to stage V took an average of 70-90 days or 10 passages. Characteristics of the Immediate Changes in NE Post ENU- Treatment Cytometric measurements of cellular DNA content. Figure 5-9 shows the DNA histograms following the addition of ENU to the cell cultures. The ordinate represents the frequency (number of cells per channel) while the abscissa represents the chan- nel number (from 1 tol27). Columns represent ENU concentra- tion; from left to right are 0, 50, 100 and 500 ug ENU/m1. Rows represent 24, 48, 72 and 96 hours following exposure to ENU. In the untreated cells the modal population was channel number 4 which was hypodiploid compared with the normal cells that had the 2N peak in channel 6. Although a distinct peak was not present, a shoulder was evident in the area of channel 8 representing tetraploid cells. On days 2 and 3 the untreated cells showed a distinct second peak with modal channels of 18 and 20, respectively, which was slightly greater than an octa- ploid population of cells. At 24 hours following the addition of 50 ug ENU/ml to the cultures the hypodiploid peak was still 143 Figure 5-7. Gross appearance of tumors 5 weeks after the im- 6 plantation of 2 x 10 NET cells subcutaneously into nude mice. Figure 5-8. Histological appearance of a solid carcinoma harvested from a nude mouse transplanted with 2 x 106 NET cells. (H&E x 200). 144 Figure 5-7 Figure 5-8 Figure 5-9. Cytometric DNA histogram following the addition of ENU to the cell cultures. 0 pg/ml 146 50 ,ug/ml lOO ,ug/mi 500 lug/mi [LL— DMZ}: 3 DAY 3:? 2 l EL 1K 1. DAY4§L L L \ i 3 r fivirfilrrnvr VIIIVIY‘IIIIII ’l‘l‘rrirrfiwr 'I'WI'I'Irr'; Cytofluoromeiric DNA histograms of normal rat mammary gland cells exposed to various doses of ENU. Figure 5-9 147 present but there were also other peaks that appeared. The modal channel numbers for these peaks were 11 and 20, respec— tively. On days 2 through 4 there was a recovery towards the distributions seen with the untreated cells. Similar changes occurred where 100 and 500 ug ENU/ml were used. With these higher doses the modal channel numbers for the peaks beyond the tetraploid region were similar to those modal channel numbers seen at the 50 ug ENU/ml dose. In Figure 5-10 the ordinate shows the percent of resting cells. This was the number of cells that appeared in channels l through 11 divided by the number of cells counted, expressed as a percentage. This figure was used to demonstrate the rate at which the treated cells returned to their resting or hypo- diploid state. Over the 4 days of the experiment, about 60% of the untreated cells were in this hypodiploid region. There was a dose dependent decrease with time in the fraction of hypodi- ploid cells and the rate of recovery to the hypodiploid state was inversely related to dose (p (0.05) . DNA Synthesis The diagrammatic representation of (3H) thymidine incor- poration assay of the normal mammary gland epithelial cells treated with ENU is illustrated in Figure 5-11. Peak rates of thymidine incorporation occurred at 96 hours following treat- ment with ENU and in control cultures. Cells treated with ENU however, had rates of incorporation greater than control cells (p 0.05). The histogram of cell counts processed (Figure 5-12) in parallel with incorporation assay (Figure 5-11) illustrated Figure 5-10. 148 Diagrammatic representation of the DNA content illustrating the percent of resting cells ex- posed to various doses of ENU during a 4 day period. This figure was used to demonstrate the rate at which the treated cells returned to their resting or hypodiploid state. °/o RESTING CELLS 149 IOO~ ’0 l00,ug 80" // / . // ’ 0N9 0 / . fi’OSOpg / I 40- ’/ l . ,’ / OSOOpg o’ / / . l 20 / WSNO O I 1' I I I 2 3 4 DAYS POST- EXPOSURE Figure 5-10 150 Figure 5-11. Diagrammatic representation of (3H) thymidine incorporation assay of the NE cells exposed to various doses of ENU. COUNTS 400 360 320 280 240 200 ISO IZO 80 4O 0 Control 0 ENU 25/Ug/ml A ENU SO/ug/ml a ENU lOOpg/ml 24 . 48 72 96 léO TIME (hr) Figure 5-11 152 Figure 5-12. Diagrammatic representation of the NE cell counts exposed to various doses of ENU. CELL COUNTS x IO6 x I05 (M J5 L” OS‘QiIHOC) 43 U) axioma— L” l I ll 153 - )D—---1D / - / . / / __ d ,o’ / / / / A -—-—°( .. 0 Control _ e 25 ,ug/ml .. A 50 Ng/ml __ a lOONg/ml 1 AJ I I J 24 48 72 96 I20 TIME (hr) Figure 5-12 r: 154 that the number of cells, treated with 25 and 50 ug ENU/ml, were slightly increased above the control throughout the test period (24 to 120 hours). There was a significant difference at 96 hours with 50 ug ENU/ml (p'<0.05). The cytologic examination of the cells exposed to 50 and 100 ug ENU/ml showed multinucleated cells which increased pro- gressively from 24 hours up to 120 hours. Cytogenetic Findings The NE cells showed a near normal diploid chromsome num- ber with a modal number of 38 ranging between 36-60. The find- ings from chromosome studies after ENU exposure of NE cell cul- tures are illustrated in Table 5-1. In each experiment 2 or more doses of ENU were tested. A dose—related effect at 25, 50, 100 and 200 ug ENU/ml was demonstrated. At 500 ug ENU/m1 up to 72 hours none or few mitoses were identified and only 50 meta- phases were recorded at 72 to 96 hours. A dose as low as 25 ug ENU/m1 produced chromosome damage in 16% of the cells. In cultures exposed to ENU, 24 hours before harvest, abnormal metaphases were almost half the number of those treated at 6 hours. At 96 hours post-treatment, abnormal cells decreased to near control values. The predominant type of aberrations observed in this ex- periment, regardless of dose and time, were the isochromatid and chromatid breaks (Table 5-1). Most of the exchanges were seen when cells were exposed to ENU, 24 hours before harvest. Small DNA-containing particles called double minutes (dm) were observed in the metaphases of cells exposed to ENU and increased with higher doses and after 72 hours and 96 hours post-eXposure. 155 Table 5-1. Effect of END 33 Vitro on chromosomes of rat mammary cultures (P4 and P5). Types 6?_§Eerrations/IOO cells Dose No. Percent Single 150- Exchanges Mul- D.M. ENU Experi- Cells Chromatid chromatid tiple (metaphases ug/ml ments with Breaks Breaks Breaks with D.M. Aberrations Control 2 9(8-10) 4 S ENU added 6 hours before harvest 25 2 15(14-16) 6 9 50 2 25(20-30) 5 lo 5 3 2 100 2 45(40-50) 20 10 S 5 S 200 2 50(48-52) 30 15 5 5 ENU added 24 hours before harvest 25 2 10(9-11) 10 50 2 12(10-14) 5 3 2 2 100 2 25(15-26) 10 2 10 3 200 2 27(17—29) 12 4 10 1 $00 2 MI ENU added 72 hours before harvest 100 2 15(10-18) 5 S 1 .4 200 2 20(15-22) 5 5 5 S 500 2 20'(lS-25) 8 10 2 8 ENU added 96 hours before harvest 100 2 10(6-14) 4 2 4 200 2 12(6-16) 6 2 2 2 500 2 18(8-10) 8 2 2 6 MI - Mitotic InhibitIon 0 00M. - Counted only 50 metaphases - Double Minutes 156 'Ihe number of "dm" per cell was between 2 and 5. Fewer ex- czhanges (Figure 5-13) and multiple breaks were mostly related vvith the higher doses of ENU. Giemsa banding studies of NET cell line karyotypes con- ssistenly revealed the presence of chromosomal markers. Tri- somy of chromosome #10 was found with higher frequency (30%) (Efligure 5-14). Trisomy of chromosome #19 appeared as an in- cziriental finding on the same metaphase (Figure 5-14). Karyotyping of NET cells recovered from tumors after re- iJnlplantation into appropriate hosts, revealed many abnormali- tuiees, such as translocation, dm, and loss of sex chromosomes. 'Ftiee chromosome numbers were increased in reimplanted NET cells slicywing a modal number of 48 with a range of 36 to 96 chromo— somes per metaphase. H 01 \I Figure 5-13. NET cell metaphase, 72 hours post-ENU exposure, revealing isochromatid exchanges (arrow). (x 1200). r1 ’ - ‘ i‘ 3 f" y a I" . . - . C ,r ,J‘ “a 0.: i Q a * i “ 1' .s 7T " 0 T: Figure 5-13 Figure 5-14. NET cell metaphase karyotype (stage IV) showing trisomy (l) of chromosome #10, trisomy (2) of chromosome #19 and double minutes (3). (x 1200). ('9’! "is as. '160 3‘} II ii a h. 33 If it) it at :r s: u it it tr '= h ,3 2 a " ‘ d 161 DISCUSSION 7 . “6’27 the present study As it was demonstrated in yiyg, indicates that mammary gland tissue constitutes a target organ for the mutagenic and direct acting carcinogen effect of ENU. In the in vitro system, the average interval between ENU ex- posure and the first observation of tumorigenicity (morpho- logically transformed mammary epithelial cells) was approxi- mately 70-90 days. This was similar to the mean induction time of in yiyg mammary tumor development induced with high doses of ENU in CD female rats.26’27 3‘ Quantitation of DNA by cytofluorometry revealed that rat mammary cells in culture had a reduced amount of DNA compared with PBL and fresh, dissociated mammary tissue. Karyotypic analysis showed that these Cells were also hypodiploid. Thus in the process of becoming established in cell culture, rat mammary epithelial cells lose DNA and chromosomes. These changes may be necessary for stable, in yitrg growth. The DNA frequency distributions of untreated mammary epi- thelial cells in culture had the appearance of a randomly di- viding cell population. The shoulder on the hypodiploid peak represented the tetraploid population. This population was not more apparent due to the relatively low concentration of tetraploid cells and the low amplification setting used for the cytofluorometer. A high amplification setting probably would have made the tetraploid peak more apparent. Also pre- sent in the untreated cells were some cells with a near-octa- ploid amount of DNA. These cells probably represented multi- nucleated cells that were seen on phase contrast microscopy 162 and stained preparations. Cell clumps did not interfere with the study since we discarded any sample with greater than 1% cell clumps. Since the untreated and treated NE cells had identical peak modal channel numbers, it is not likely that ENU altered the binding of propidium iodide for DNA. The effect of ENU on DNA frequency distributions of rat mammary epithelial cells was to increase, in a bimodal fashion, the number of cells with greater than the resting, hypodiploid amount of DNA. The modal channels for the 2 additional peaks were around 11 and 20. These peaks may represent tetraploid and octaploid cells, respectively. Subjective microscopic e- valuation of the cell cultures showed that treated-cultures had more multinucleated cells than controls. The results of the experiment studying incorporation of thymidine showed that ENU-treated cells incorporated thymidine at a greater rate than control cells. This serves as indirect evidence that ENU stimulated DNA synthesis. These data coupled with the results showing accumulation of near tetraploid and near octaploid cells following treatment with ENU suggested that ENU stimulated DNA synthesis and blocked cell division. These effects of ENU were only temporary; the rate of thymidine incorporation decreased along with control cells and the DNA distributions returned toward pre-exposure distributions by the end of the experiment. ’ The explanation of the observations described above most likely resides in the fact that alkylating agents, although de- pendent on proliferation, are not cell-cycle specific, and the drugs may act at any stage of the cycle. However, the toxicity 163 is usually expressed when the cell enters the S phase and progression through the cycle is blocked at the G2 (premi- totic) phase.28 Cells appear more sensitive in late G1 or S then in G2, mitosis, or early G Polynucleotides are more 1' susceptible to alkylation in the unpaired state than in he- lical form. During replication of DNA, positions of the molecule are unpaired. The cells accumulated behind the block at G2 may have a double complement of DNA while continuing to synthesize other cellular components, such as protein and RNA. This may result in unbalanced growth, with the formation of enlarged or giant cells that may continue to synthesize DNA, making as much as four or five times the normal complement.29 These mechanisms of action of ENU may account for the results in the present study. ENU may block those tetraploid cells from dividing and stimulate them to synthesize more DNA, bringing them up to a near octaploid state prior to dissipa— tion of the effect. Chromosome analysis of in yitrg rat NE exposed to ENU in- dicated both dose and time effect. The greatest number of cells showing chromosomal damage occurred when cells were treated 6 hours before harvest (Table 5-1). Types of breaks were prin- cipally chromatid and isochromatid breaks suggesting that one mechanism of action of ENU could be on the chromsome structure. Trisomy of chromsome 10 was the most consistent marker associated with in yitrg ENU-induced neoplastic transformation of rat mammary epithelial cells, though trisomy of 19 chromosome was also found. Excess chromosome 10 is presumed to arise by mitotic nondysjunction and rearrangements, but its significance in the neoplastic process is not known. Comparison of chromo- 164 somal findings in these early stages with those of gross tumors would be important. However, good metaphases are dif- ficult to obtain directly from solid tumors. The presence of "dm" in metaphases was first considered 30'31 but re- to be a specific finding in neurogenic tumors, cently these were recognized in many other tumor types espec- ially in breast cancer.33 Their origin and significance re- main debatable. Recent observations have demonstrated that there is a close interrelationship between "dm" and so-called 33 homogeneously staining regions (HSR). Thus, it has been shown that "dm" can be transformed into HSR and vice versa,30' 31 but they also can originate from other regions.34 The im- portance of these "dm" in carcinogenesis is not known. ENU-treatment of NE cells showed a dose-effect relation- ship expressed by various phenotypic and numeric and struc- tural chromsomal aberrations. However, the significance of these changes is unknown and may have only indirect relevance to carcinogenicity, for example as an indicator of cell dam- age. Phenotypic Changes In regard to phenotypic changes, a common feature of stage III was a tendency toward squamous differentiation and irregularity of nuclear size. The cells at this stage inoc— ulated into isologous hosts did not induce tumors. NeOplastic cell lines transformed in yitrg invariably derived from stage III foci. This stage probably is equivalent with the in_yiyg Proliferative nodules (wholemount preparation) observed at the ‘UErminal end buds (TEB) of rat mammary gland post ENU—treatment. 165 26'35'36 Evidence suggests that malignant mammary tumors may develop from these proliferative nodules. However there is no direct evidence that other types of carcinomas arise from islands of altered squamous cells in yiyg. More work is needed to determine if the described in yi- tgg epithelial cell neoplastic transformation could be related with activation of the "ras" oncogene as it was demonstrated with nitroso-methylurea.37 The ENU-induced reproducible sequence of in yitrg neo- plastic transformation.constitutes a valuable model for the study of earliest neoplastic alterations of human breast can- cer in particular and carcinomas in general. 166 REFERENCES l. Laerum O, Rajewsky MF: Neoplastic transformation of fetal rat brain cells in culture after exposure to ethylnitro- sourea in vivo. J Natl Cancer Inst 55:1177-1187, 1975. 2. Soukup SW, Au W: The effect of ethylnitrosourea on chromosome aberrations in vitro and in vivo. Humangenetik 29: 319-328, 1975. 3. Au W, Soukup SW, Mandybur TI: Excess chromosome #4 in ethylnitrosourea-induced neurogenic tumor lines of the rat. J Natl Cancer Inst 59:1709-1716, 1977. 4. Druckrey H: Chemical structure and action in trans- placental carcinogenesis and teratogenesis. In: Transplacen- tal Carcinogenesis, Publication No. 4, Lyon, International Agency for Cancer Research, 1973. 5. Rice JM: An overview of transplacental chemical car- cinogenesis. Teratology 8:113-135, 1973. 6. Magee PN: Mechanisms of transplacental carcinogene- sis by nitroso compounds. In: Transplacental Carcinogenesis, Publication No. 4, Lyon, International Agency for Research on Cancer, 1973. 7. Singer B: N-nitroso alkylating agents: Formation and persistence of alkyl derivatives in mammalian nucleic acids as contributing factors in carcinogenesis. J Natl Cancer Inst 62:1329-1339, 1979. 8. Wigley CB: Transformation in vitro of adult mouse salivary gland epithelium; a system for studies on mechanisms of initiation and promotion. In: Franks and Wigley (eds): Neoplastic Transformation in Differentiated Epithelial Cell 167 Systems In Vitro, New York, Academic Press, 1979, pp 3-36. 9. Fusenig N, Breitkreutz D, Boukamp P, Lueder M, Irm- scher G, Worst PKM: Chemical carcinogenesis in mouse epider- mal cell cultures: Altered expression of tissue specific func- tions accompanying cell transformation. In: Franks and Wig- ley (eds): Neoplastic Transformation in Differentiated Epi- thelial Cell Systems In Vitro, New York, Academic Press, 1979, pp 37-98. 10. Indo K, Maiyaji H: Neoplastic transformation and ab- normal differentiation in foetal rat keratinising epidermal cells in culture. In: Franks and Wigley (eds): Neoplastic Transformation in Differentiated Epithelial Cell Systems In Vipgp, New York, Academic Press, 1979, pp 99-112. 11. Summerhayes IC: Influence of donor age on in vitro transformation of bladder epithelium. In: Franks and Wigley (eds): Neoplastic Transformation in Differentiated Epithelial Cell Systems In Vitro, New York, Academic Press, 1979, pp 137- 162. 12. Franks LM, Wigley CB: In vitro transformation in the respiratory tract. In: Franks and Wigley (eds): Neoplastic Transformation in Differentiated Epithelial Cell Systems In Vitro, New York, Academic Press, 1979, pp 163-172. 13. Kuroki T, Drevon C, Saint Vincent L, Montesano R: Properties of the IAR-series of liver epithelial cells trans- formed by chemical carcinogens. In: Neoplastic Transformation in Differentiated Epithelial Cell Systems In Vitro, New York, Academic Press, 1979: pp 173-188. 14. Dao TL, Sinha D: Mammary adenocarcinoma induced in 163 organ culture by 7,12-dimethylbenz(a)anthracene. J Natl Can- cer Inst 49:591-593, 1972. 15. Cohen LA, Tsuange J, Chan PC: Characteristics of rat normal mammary epithelial cells and dimethylbenzanthracene -induced mammary adenocarcinoma cells grown in monolayer cul- ture. In Vitro 10:51-62, 1974. 16. Stoica G, Koestner A: Characteristics of normal rat mammary epithelial cells and N-ethyl-N-nitrosourea-induced mam- mary adenocarcinoma cells grown in culture. In Preparation 17. Gichner T, Veleminsky J: The mutagenic activity of l-alkyl-l-nitrosoureas and l-alkyl-3-nitro-1-nitrosoguanidines. Mutat Res 4:207-212, 1967. 18. Krishan A: Rapid flow cytofluorometric analysis of mammalian cell cycle by propidium iodide staining. J Cell Sig; 66:188-193, 1975. 19. Wooranooj U, Hsu TC: The C-and G- banding patterns of Rattus norvegicus. J Natl Cancer Inst 49:1425-1431, 1972. 20. Yunis JJ: New chromosome techniques in the study of human neoplasia. Human Pathology_12:540-549, 1981. 21. Committee for a Standard Karyotype of Rattus norvegio cus: Standard karyotype of the Norway rat Rattus norvegicus. Cytogenet Cell Genet 12:199-205, 1973. 22. Buckton KE, Evans HJ: Methods for the analysis of human chromosome aberrations. World Health Organization, pp 18-22, 1973. 23. DiPaolo JA, Pepescu NC: Relationship of chromosome changes to neoplastic cell transformation. Am J Pathol 85:709- 738,.1976. 169 24. Levan G, Mitleman F: G- banding analysis in a seri- ally transplanted Rous rat sarcoma. Hereditas 80:140-145, 1976. 25. MacPherson I: Soft agar techniques. In: Kruse and Patterson (eds): Tissue Culture Methods and Applications, New York, Academic Press, 1972, pp 276-280. 26. Stoica G, Koestner A, Capen C: Characteristics of N -ethyl-N-nitrosourea induced mammary tumors in the rat. Am_g Pathol 110:161-169, 1983. 27. Stoica G, Koestner A, Capen C: Neoplasms induced with high single doses of N-ethyl-N-nitrosourea in 30-day-old Sprague-Dawley rats, with special emphasis on mammary neopla- ia. AntiCancer Research 4:5-12, 1984. 28. Wheeler GP: Some biological effects of alkylating agents. Fed Proc 26:885-892, 1967. : Alkylating agents. In: Holland and Frei (eds): Cancer Medicine, Philadelphia, Lea and Feibiger, 1979. PP 791-806. : Mechanisms of action of nitrosoureas. In: Sartorelli and Johns (eds): Antineoplastic and Immuno- suppressive Agents, Pt. II, Handbuch der Experimentellen Phar- makologie, Vol 38, Berlin, Springer—Verlag, 1975, pp 65-84. 29. Calabresi P, Parks RE: Chemotherapy of neoplastic diseases. In: Goodman and Gillman's The Pharmacological Ba- sis of Therapeutics, ed 6, New York, Macmillan Publishing Co., 1975, pp 1256-1270. 30. Balaban-Mahlenbaum G, Gilbert F: Double minute chromosomes and the homoqeneously staining regions of chrom- osomes in human neuroblastoma cell line. Science 198:739-741, 1977. 179 31. Balaban-Mahlenbaum G, Gilbert F: Relationship be- tween homoqeneously staining regions and double minute chro- mosomes of a human neuroblastoma cell line. Cancer Genet Cy- togenet 2:339-348, 1980. 32. Barker PE, Hsu TC: Double minutes with special re- ference to breast tumors. J Natl Cancer Inst 62:257-262, 1978. 33. Biedler JL, Spengler BA: Metaphase chromosome ano- maly: Association with drug resistance and cell-specific pro- ducts. Science 191:185-l87, 1976. 34. Joachim M, Dahlenfors R, Ekedahl C: On double-min- utes and their origin in a benign human neoplasm, a mixed sali- vary gland tumor. AntiCancer Research 2:261-264, 1982. 35. Russo J, Saby J, Isenberg WM, Russo IH: Pathogene- sis of mammary carcinomas induced in rats by 7,12-dimethyl- benzene anthracene. J Natl Cancer Inst 59:435-445, 1977. 36. Russo IH, Russo J: Developmental stages of the rat mammary gland as determinant of its susceptibility to 7,12-di- methylbenzanthracene. J Natl Cancer Inst 5131439-1449, 1978. 37. Sukumar S, Notario V, Zanca-Martin D, Barbacid M: Induction of mammary carcinomas in rats by nitroso-methylurea involves malignant activation of H-ras-l locus by single point mutations. Nature 306:658-661, 1983. VITA The author was born in Homorod, Brasov, Romania in 1941. He attended elementary school in Homorod and Liceum (high school) in Rupea where he graduated in 1961. He enrolled in the veterinary curriculum at the Institute of Agronomy of Yassy, graduating with Magna cum Laudae in 1966. He was ap- pointed assistant professor in the Department of Veterinary Clinical Medicine in the Veterinary School of Yassy. After 2 years he transferred to the Veterinary Diagnostic Labora- tory in Brasov as a pathologist. He worked also as Chief Veterinary Territorial Inspector at Rupea up to 1977 when he immigrated to the U.S.. In the USA he worked for 1 year as an industrial employee in Indianapolis. In 1978 he became a graduate student at Ohio State University Department of Vet- erinary Pathobiology where he obtained his Master's Degree in Pathology. In 1982 he transferred to Michigan State Uni- versity where he completed his PhD in pathology in 1984. The author was a member of the Phi Zeta Society and was the recipient of an Upjohns Fellowship. 171