We have developed an eent in vitro transformation system using N-methyl-N-nitrosourea that aflows us to study the role of hormones and growth factors in mouse mammary tumorigenesis. Ulizing this system, we reported earlier that mammary tumors induced in vitro with N-methyl-N-nitrosourea in the presence of mamogenic hormones (progesterone and prolactin) contain predominately an activated c-Ki-ras protooncogene with a G35 --A35 transitional mutation in the 12th codon. Mammary tumors induced in the presence of another mitogen, lithium (Li), do not have a mutation in the c-Ki-ras protooncogene. By using an expression cloning system, a pl d cone containing a 1. transversion mutations at the 61st codon of the H-ras protooncogene (9-12). A majority ofthymic lymphomas induced with MNU, on the other hand, contain a G35 --A35 mutation in the N-ras protooncogene (13,14).The long-term goal of our laboratory is to understand the role of mammogenic hormones and related growth factors in the growth and differentiation of MECs and the manner in which these factors influence the induction of mammary neoplasias of different phenotypes and genotypes. To achieve our goal, during the past several years we have developed a defined serum-free cell culture system in which mouse MECs embedded in a three-dimensional collagen gel matrix can be grown, induced to differentiate, and be neoplastically transformed with chemical carcinogens (15). Using this system, we have observed that the types of mammary lesions induced by carcinogens are greatly influenced by the mitogens present around the time of carcinogen treatment. Earlier we reported on an in vitro system for the induction of preneoplastic hyperplastic alveolar nodules (HANs) and carcinomas from MECs exposed to the direct-acting chemical carcinogen MNU in the presence of different mitogens (16). When mouse MECs were grown in the presence of the mammogenic hormones progesterone and prolactin (PPRL) during MNU administration, the predominant types of lesions induced were a high incidence ofHANs and carcinomas with squamous metaplasia. In contrast, when epidermal growth factor was used as a mitogen during the carcinogen treatment, only a low incidence of ductal hyperplasia was detected, although the extent of MEC proliferation between the two groups was equivalent. The genetic analysis of these lesions indicated that the activation ofthe protooncogene was also dependent on the mitogen used around the time of carcinogen treatment. The majority (80%o) of the HANs and carcinomas induced with MNU in the presence of PPRL had an activation ofthe protooncogene c-Ki-ras by a specific G35 -* A35 point mutation at codon 12. The activation of the protooncogene was determined to be an early event in this carcinogenesis process because the activation was detected in preneoplastic lesions (17). In contrast, activation of c-Kiras was absent in all the ductal hyperplasias induced by MNU in the presence of the mitogen epidermal growth factor. Involvement of the same type of c-Ki-ras mutation ha...