E2F1 has both oncogenic and tumor-suppressive properties in a transgenic model.

Using a transgenic mouse model expressing the E2F1 gene under the control of a keratin 5 (K5) promoter, we previously demonstrated that increased E2F1 activity can promote tumorigenesis by cooperating with either a v-Ha-ras transgene to induce benign skin papillomas or p53 deficiency to induce spontaneous skin carcinomas. We now report that as K5 E2F1 transgenic mice age, they are predisposed to develop spontaneous tumors in a variety of K5-expressing tissues, including the skin, vagina, forestomach, and odontogenic epithelium. On the other hand, K5 E2F1 transgenic mice are found to be resistant to skin tumor development following a two-stage carcinogenesis protocol. Additional experiments suggest that this tumor-suppressive effect of E2F1 occurs at the promotion stage and may involve the induction of apoptosis. These findings demonstrate that increased E2F1 activity can either promote or inhibit tumorigenesis, dependent upon the experimental context.

Suppression or elevation of cytosolic phospholipase A2 alters keratinocyte prostaglandin synthesis, growth, and apoptosis.

Prostaglandins and other arachidonic acid (AA) metabolites are synthesized by keratinocytes in response to tumor promoters and are produced at very high levels in tumors. After phorbol ester treatment, AA is hydrolyzed from keratinocytes primarily by the cytosolic form of phospholipase A2 (cPLA2), which exhibited a strong substrate preference for phosphatidylcholine over phosphatidylethanolamine and AA over other fatty acids. Phorbol esters increase cPLA2 activity but not the level of expression. To dissociate increased cPLA2 activity from other phorbol ester effects and thus determine the effects of altered AA release on cell growth, the murine keratinocyte cell line, HEL-30, was stably transfected with the sense or antisense cDNA for cPLA2. The resulting cell lines displayed corresponding over- or underexpression and up to 23-fold differences in cPLA2 activity between them. Phorbol ester caused a 15-fold difference in AA release between sense and antisense transfectants. Prostaglandin E2 levels correlated with AA release levels. The sense transfectants showed an enhanced proliferative capacity, based on increased cell number over time and [3H]thymidine incorporation. The antisense transfectants had significantly (>60%) reduced growth rates, compared with both parental cells and sense transfectants. The extent of apoptosis was determined in tumors from cell lines grown in graft chambers in vivo. The number of apoptotic cells was significantly greater in tumors from the sense transfectants, based on terminal deoxynucleotidyl transferase-mediated dUTP-biotin nick end labeling staining, compared with the parental or antisense lines. These data are in agreement with a recent study (M. C. Stern et al., Mol. Carcinog., 20: 137-142, 1997) showing a correlation between increased apoptosis and tumor progression in this model system. These results suggest that the elevated eicosanoid synthesis that is observed in skin carcinomas contributes to the growth and progression of these tumors.

Dissociation of sensitivities to tumor promotion and progression in outbred and inbred SENCAR mice.

The sensitivity of outbred SENCAR mice and inbred SENCAR (SSIN) mice to multistage carcinogenesis was studied. Tumors were induced using either 7,12-dimethylbenz[a]anthracene or N-methyl-N'-nitro-N-nitrosoguanidine as initiators and 12-O-tetradecanoylphorbol-13-acetate or benzoyl peroxide as promoting agents. Although the number of papillomas per mouse was higher in SSIN than in outbred SENCAR mice, the number of carcinomas observed in the SSIN strain was significantly lower regardless of the initiator or promoter used. It was also observed that the expression of markers of premalignant progression (i.e., dysplasia, expression of keratin K13, and loss of keratin K1 expression) was markedly suppressed in SSIN papillomas. After 50 wk of promotion with 12-O-tetradecanoylphorbol-13-acetate, the pattern of expression of K13 and K1 in SSIN mice was comparable to the pattern observed in outbred SENCAR mice after 10 to 20 wk of promotion with 12-O-tetradecanoylphorbol-13-acetate. It was also observed that 67% of the tumors induced in SSIN mice by initiation with 7,12-dimethylbenz[a]anthracene exhibited a mutation in codon 61 of the Ha-ras-1 gene. This latter finding suggests that the differences observed in tumor progression between the inbred strain and the outbred stock are not related to a genetic alteration in the Ha-ras-1 gene but rather to an independent event that we have postulated to involve a putative suppressor gene. The data reported here suggest that the putative gene(s) that confers susceptibility to tumor promotion was segregated from the gene(s) involved in tumor progression during selection and inbreeding of the SENCAR mouse stock.

Changes in keratin expression during 7,12-dimethylbenz[a]anthracene-induced hamster cheek pouch carcinogenesis.

This study was undertaken to explore the expression of keratins in the hamster cheek pouch carcinogenesis model, using monospecific keratin antibodies and a technique that allows immunoblotting analysis of tissues embedded in paraffin. Changes in keratin expression were correlated with histopathological changes and with the expression of the enzyme gamma-glutamyl transpeptidase. The right cheek pouch of 20 male golden Syrian hamsters was treated with 0.5% 7,12-dimethylbenz[a]anthracene for 16 weeks. As previously described by other laboratories, this treatment resulted in hyperplastic and dysplastic lesions and benign and malignant tumors. The keratins assayed in this study were K14 (Mr 55,000), K1 (Mr 67,000), and K13 (Mr 47,000). The normal hamster cheek pouch epithelium expressed K14 in the basal layer and K13 in the suprabasal and differentiated layers, whereas K1 was not detected by either immunohistochemistry or immunoblotting. Concomitant with 7,12-dimethylbenz[a]anthracene-induced hyperplasia, there were some topographical alterations in the distribution of K14. In this case, K14 was no longer restricted to the basal layer but was also expressed in differentiated cells. The same pattern was also observed in dysplastic lesions and in squamous cell carcinoma. Furthermore, expression of the K13 differentiation-associated keratin was preserved in this hyperplastic epithelium during all the stages of carcinogenesis, including either anaplastic or differentiated areas. In contrast, after 2 weeks of 7,12-dimethylbenz[a]anthracene treatment, K1 expression started as a weak and patchy pattern in suprabasal cells, becoming stronger and more homogeneous at 8 and 16 weeks of treatment. However, K1 was almost absent in squamous cell carcinoma, where only small very well differentiated areas were stained. We also observed gamma-glutamyl transpeptidase-positive foci in earlier stages of carcinogenesis, concomitant with the expression of the K1 keratin. However, it was not possible to find a perfect topographical correspondence between the two events. Alterations in the pattern of keratin expression appear to be a common feature during the development of squamous cell carcinoma in different systems and could be an excellent tool to study carcinogenesis and chemoprevention.

Overexpression of insulin-like growth factor-1 induces hyperplasia, dermal abnormalities, and spontaneous tumor formation in transgenic mice.

Transgenic animals were developed to assess the role of insulin-like growth factor 1 (IGF-1) in skin growth, differentiation and organization, as well as its importance in tumor formation. Expression of a human IGF-1 cDNA was targeted to the interfollicular epidermis of transgenic mice using a human keratin 1 promoter construct (HK1). Transgenic animals (HK1.IGF-1 mice) could be identified at birth by early ear unfolding and excessive ear and skin growth compared to non-transgenic littermates. Further examination of the skin from these mice showed epidermal hyperplasia and hyperkeratosis, marked thickening of the dermis and hypodermis, and early hair follicle generation in newborns. The severity of this phenotype correlated with transgene expression both of which subsided with age. Adult HK1.IGF-1 mice developed spontaneous tumors following treatment with 12-O-tetradecanoylphorbol-13-acetate (TPA) alone and exhibited an exaggerated epidermal proliferative response following treatment with the tumor promoter compared to non transgenic littermates. Additionally, HK1.IGF-1 transgenic mice developed papillomas faster and in markedly greater numbers compared to non-transgenic littermates in standard initiation-promotion experiments. The data presented suggest an important role for IGF-1 in the process of multistage carcinogenesis in mouse skin.

Deregulated expression of cell-cycle proteins during premalignant progression in SENCAR mouse skin.

It is now evident that several genes encoding regulatory activities that control the mammalian cell cycle, particularly some that control the progression of quiescent cells through G1 and into S phase, are targets for alterations that underlie the development of neoplasms. Here, we made a sequential study of alterations in cell cycle protein expression and complex formation among cyclin, cyclin dependent kinases (CDKs) and CDK inhibitors (CKIs) during premalignant progression in SENCAR mouse skin tumors. Changes in the level of expression were observed in positive (cyclin D1, D2, and E2F family members) and negative regulators (p16Ink4a, p57Kip2) of the cell cycle. Also, we observed the formation of cyclin/CDK/CKI complexes. The amounts of these proteins and complexes increased substantially at specific times during promotion but not during malignant conversion to carcinomas. These data show that deregulation of growth control occurs in benign tumors and that subsequent mutations not involved cell-cycle regulation are probably necessary to induce invasive behavior.

Estrogen and progesterone regulation of proliferation, migration, and loss in different target cells of rabbit uterine epithelium.

We have explored the possibility that estrogens and progesterone could have different target uterine cell populations according to their cell cycle stage and localization in glands vs. lumen. Experiments were carried out in which rabbits were injected with [3H]thymidine for 3 days to label nuclei of dividing cells, then either 17 beta-estradiol, progesterone, or vehicle were administered. 17 beta-Estradiol induced a decrease in the percentage of cells with labeled nuclei or labeling index of either luminal or glandular epithelium. Since this steroid has been shown to have a significant proliferative effect on glands, the data suggest that its effect is exerted on unlabeled quiescent cells, which are then recruited into the cell cycle. Progesterone, on the other hand, was found to induce a significant increase in labeling index of both luminal and glandular epithelium. Therefore, it is concluded that dividing cells are a target for this hormone. Analysis of the number of nuclear grains according to cell location in luminal vs. glandular epithelia and the effect of hormone administration confirmed that each ovarian hormone acts on different target cell populations. Short and long term administration of estrogens resulted in a larger internal circumference of the uterus due to an increase in the number of luminal cells, whereas the number of glands and glandular cells per section did not appear to change. These findings, in combination with previous research, suggest that endometrial gland cells migrate towards the lumen and estrogen administration decreases the rate of cell loss in the luminal epithelium. The concept of cell migration is supported by experiments in which single administration of [3H]thymidine to rabbits was followed by determination at different times of the geographical distribution of cells with labeled nuclei. There was observed, as a function of time, a decrease in the number of labeled cells in the bottom of the glands with a concomitant increase in the same parameter in the upper part of the glands and luminal epithelia. Estradiol administration changed these kinetics.

An animal model for oral cancer.

Human head and neck squamous cell carcinogenesis (SCC) is a common malignancy that appears to be related to continuous exposure to putative carcinogens or promoters such as tobacco and alcohol. To understand the mechanisms of the development of head and neck cancer and to test the efficiency of new therapeutic approaches, the characterization of an animal model system is necessary. The check-pouch carcinogenesis model in Syrian golden hamsters is probably the best known animal system that is most closely comparable with the development of premalignant and malignant lesions in human oral cancer. Furthermore, it is one of the most well-characterized animal system models for SCC. Our first approach to understanding the cellular and molecular changes that occur in the hamster cheek-pouch carcinogenesis process was to compare this model to the mouse-skin system, in which a number of critical events have been well characterized. We examined the sequential expression of hyperplasia, micronucleated cells, ornithine decarboxylase activity, polyamine levels, transglutaminase I activity, epidermal growth factor receptor levels, expression of several keratins, gamma-glutamyl transpeptidase, and nucleolar organizer regions. We suggest that these markers can be used to understand mechanisms of carcinogenesis and, in addition, can serve as alternative shorter end points in studies of chemoprevention. We also present preliminary molecular studies in the experimental oral model. We obtained a partial sequence of exon 2 of the Ha-ras gene and detected an A182----T transversion in codon 61 in hamster cheek-pouch SCC induced by 7,12-dimethylbenz(a)anthracene.

Optimal fluorescence excitation wavelengths for detection of squamous intra-epithelial neoplasia: results from an animal model.

Using the hamster cheek pouch carcinogenesis model, we explore which fluorescence excitation wavelengths are useful for the detection of neoplasia. 42 hamsters were treated with DMBA to induce carcinogenesis, and 20 control animals were treated only with mineral oil. Fluorescence excitation emission matrices were measured from the cheek pouches of the hamsters weekly. Results showed increased fluorescence near 350-370 nm and 410 nm excitation and decreased fluorescence near 450-470 nm excitation with neoplasia. The optimal diagnostic excitation wavelengths identified using this model - 350-370 nm excitation and 400-450 nm excitation - are similar to those identified for detection of human oral cavity neoplasia.

Genetic analysis of neoplasia induced by N-nitroso-N-methylurea in Xiphophorus hybrid fish.

Interspecific crosses within the genus Xiphophorus have historically been used to study the genetic aspects of melanoma formation. Melanomas typically occur as a result of deregulation of polymorphic, naturally occurring macromelanophore pigment patterns. Hybrid crosses also have been used to study the inducibility of melanoma by physical sources (such as UV light) and chemicals (such as N-methyl-N-nitrosourea, MNU). We previously defined a genomic region that is implicated in fish melanomagenesis and identified a candidate tumor suppressor gene (CDKN2X) within this genomic area. Highly significant associations between BC(1)-hybrid CDKN2X genotypes and UV-induced melanoma formation exist in a backcross produced from 2 inbred parental lines. However, when BC(1) hybrids are exposed to MNU as the tumor induction agent, a significant association between inheritance of CDKN2X alleles and tumor development is not observed. These data suggest there is mechanistic and genetic heterogeneity in melanomas derived from different etiologies within BC(1) hybrid fish.

A proposed classification scheme for Xiphophorus melanomas based on histopathologic analyses.

We studied the histopathologic characteristics of melanomas induced in the Xiphophorus model. This fish model has been used for several decades to study the molecular and genetic mechanisms underlying its susceptibility to melanoma induction. Numerous distinct interspecific hybrid crosses currently are being used in research on carcinogenesis. We previously reported that tumors were induced in such hybrid crosses after treatment with N-methyl-N-nitrosourea or UV radiation. In this report, we describe the histopathologic features of Xiphophorus melanomas and propose a new classification system. We suggest that melanomas in these fishes can be classified as follows: melanocytic melanomas; melanophorous-macromelanophorous polymorphic melanomas; spindle cell type melanomas; epithelioid cell melanomas; and amelanotic melanomas. The new classification of Xiphophorus melanomas should allow correlations between histopathologic characteristics and carcinogen treatment, and between histopathologic characteristics and the genetic background of the hybrid fish.

Morphological changes of the nucleolar organizer regions induced by 7,12-dimethylbenz[a]anthracene in the hamster cheek pouch.

The staining of nucleolar organizer regions (NORs), which equate to rDNA transcription, was applied to chemically induced-lesions of the hamster cheek pouch. The cheek pouches of 16 male golden Syrian hamsters were treated three times a week with 0.5% 7,12-dimethylbenz[a]anthracene (DMBA) in mineral oil for 16 weeks. The percentage of gathered-type NORs with high activity nucleoli increased in the pouch epithelium during DMBA treatment and reached the highest values in malignant tumors. The percentage of dispersed-type NORs also increased in the malignant lesions. However, the absolute number of NORs was not affected by DMBA treatment. These results suggest that DMBA induces modification of NOR activity at the early stages of carcinogenesis and shows the potential of this model for studying NOR alterations in neoplasia.

Molecular genetic alterations as intermediate end points in chemoprevention studies.

It is generally accepted that carcinogenesis is related to the accumulation of genetic damage in somatic cells. In support of this concept, numerous studies have identified and characterized genes that are mutated or deleted in carcinogenesis. Some of these genetic alterations occur very frequently, and therefore it is theoretically possible to use them as diagnostic tools or as intermediate end points for chemoprevention studies. This is particularly relevant for some animal models, in which certain genetic alterations are found at a frequently close to 100%. In contrast, specific genetic alterations normally occur less frequently in human cancer, probably because humans are genetically more heterogeneous and are exposed to multiple carcinogens, tumour promoters, and other modifiers of carcinogenesis. Some genetic alterations occur early in tumour development, as in the case of ras mutations in some chemical carcinogenesis models. Similarly, p53 mutations also appear to be a frequent and early event in ultraviolet light (UV)-induced skin tumours in mice and possibly in humans. However, in other neoplasias such as prostate cancer, p53 alterations occur at late stages of tumour development. Alterations that occur early in neoplastic development may be valuable as early markers of tumour development, whereas those that occur late in development may be useful for determining the action of chemopreventive agents on tumour progression. Although the use of genetic markers as intermediate end points of cancer prevention studies has not been completely developed, it has great potential. The development of simple and sensitive molecular techniques such as the polymerase chain reaction and in situ hybridization has opened the possibility of using these markers in large-scale cancer prevention studies.

The hamster cheek pouch carcinogenesis model.

The Syrian golden hamster cheek pouch carcinogenesis model is probably the best-known animal system that closely compares to events involved in the development of premalignant and malignant human oral cancers. Furthermore, it is one of the most well-characterized models for squamous cell carcinomas (SCCs). However, stages of carcinogenesis (initiation, promotion, and progression) have not been well-defined in this system. Basic understanding of the mechanism(s) of carcinogenesis in this organ is instrumental for the development of new strategies for chemoprevention and early chemointervention. To understand the important early events that occur in the hamster cheek pouch carcinogenesis model, we compared it to the mouse skin model, where a number of critical events have been well characterized. We determined that approximately 60% of the hamster cheek pouch SCCs have a mutation in codon 61 of the Ha-ras gene. We also established a two-stage carcinogenesis protocol in this model using a single dose of dimethylbenz(alpha)anthracene (DMBA) and multiple doses of benzoyl peroxide for 45 weeks. Twenty-five percent of tumors developed with this protocol had the same mutation in codon 61 of the Ha-ras gene, confirming that this mutation, as in the mouse skin model, is initiation-related. We examined the sequential expression of hyperplasia, micronucleated cells, ornithine decarboxylase (ODC) activity, polyamine levels, transglutaminase I activity, epidermal growth factor receptor (EGF-R) levels, keratins, gamma-glutamyltranspeptidase (GGT), transforming growth factor-beta 1 (TGF-beta 1), leukoplakia, and carcinomas induced during carcinogenesis.(ABSTRACT TRUNCATED AT 250 WORDS)

The mouse skin carcinogenesis model.

Skin carcinogenesis can be divided into at least three major stages: initiation, promotion, and progression. In the mouse skin model, the first stage is thought to involve the interaction of a tumor initiator with the genetic material of stem cells, leading to an irreversible alteration in growth control or differentiation, probably by activation of the Ha-ras oncogene. The major effect of all skin-tumor promoters seems to be the specific expansion of the initiated stem cells. The correlation between the abilities of tumor promoters to induce sustained hyperplasia and their tumor-promoting activities is very good. We found that the appearance of alpha-glutamyl transpeptidase (GGT) and keratin 13 and the lack of expression of keratins 1 and 10 are good markers for skin tumor progression. These alterations occur when papillomas change from a diploid to an aneuploid state, mainly as a result of developing trisomies 6 and 7. To evaluate the role of GGT in skin-tumor progression, we transfected a functional GGT cDNA into two cell lines that normally produce papillomas when grafted into the skin of nude mice. When injected subcutaneously, all of the GGT-transfected clones formed malignant tumors, whereas only 24% of vector-transfected cells did. When GGT-transfected clones were placed into grafts, the grafts had an average mass almost three times that of grafts of vector-transfected cells. Our recent studies also suggest that the ribonucleoprotein telomerase and the gap-junctional proteins connexins (Cxs) are also important in skin-tumor progression. A progressive increase in telomerase activity was associated with the increased level of genomic instability during tumor progression. In addition, the level and expression of Cx26, Cx43, and Cx31.1 were significantly altered during skin tumor promotion and progression. The differences of various mouse stocks and strains in susceptibility to multistage skin carcinogenesis seem to be related more to alterations in tumor promotion than to tumor initiation; however, the critical events have not been determined. Results with an inbred strain of SENCAR mice, which are very sensitive to papilloma formation by the two-stage protocol, also suggest that susceptibility is related to promotion. Despite the high incidence of papillomas in these inbred SENCAR mice, the number of malignant tumors was extremely low, suggesting that sensitivity to promotion and progression are independent in these mice.

Induction of short-term markers of tumor promotion by organic peroxides.

Experiments from different laboratories have shown that benzoyl peroxide (BzPo) and other organic peroxides are effective tumor promoters in the mouse skin two-stage carcinogenesis system. In the present paper we have studied the short-term effect of six other organic peroxides, which have not been previously assayed as skin tumor promoters. These compounds were chosen for their molecular diversity, the type of radical predicted to be formed, solubility and availability. The parameters evaluated in this study include a series of short-term markers of tumor promotion, hyperplasia, induction of dark basal keratinocytes and induction of ornithine decarboxylase activity. After single applications the biological activity of the compounds was: m-chloroperoxybenzoic acid greater than di-m-methylbenzoyl peroxide greater than dicumyl peroxide greater than O,O-t-butyl-O-(2-ethylhexyl)mono-peroxycarbonate greater than benzoyl peroxide greater than di-m-chlorobenzoyl peroxide greater than di-t-butyl peroxide greater than t-butyl hydroperoxide. After multiple applications, the order of activity of the compounds was: dicumyl peroxide greater than di-m-methyl-benzoyl peroxide greater than O,O-t-butyl-O-(2-ethylhexyl)monoperoxy carbonate greater than m-chloroperoxybenzoic acid greater than di-m-chlorobenzoyl peroxide greater than t-butyl hydroperoxide greater than benzoyl peroxide greater than di-t-butyl peroxide. The difference of activity among the different compounds did not seem to correlate directly with the chemical stability of the compound; it is more likely that the activity depends on different factors such as percutaneous absorption, metabolism, and the rate of free radical formation in vivo. The data presented here further support the association between free radicals and tumor promotion since all of the compounds, with the exception of one, were active in inducing the short-term markers of tumor promotion. It will also establish conditions for future tumor experiments.

Genetic susceptibility to papilloma progression in SENCAR mice.

Previous results showed that in an inbred line (SSIN) derived from outbred SENCAR mice there is a dissociation between susceptibility to papilloma development and the malignant conversion of these into squamous cell carcinomas (SCC). To extend this conclusion, we designed an interstrain breeding experiment using the two-step carcinogenesis protocol in order to study the susceptibility to tumor progression of F1 offspring. The strains used were SSIN, BALB/c, both known for their resistance to papilloma progression, and SENCAR. Both the SSIN X SENCAR and SENCAR X SSIN F1s showed a promotion sensitivity similar to that of the SSIN mice. This behavior was also seen in the SSIN X (SSIN X SENCAR) and SSIN X (SENCAR X SSIN) backcrossed animals, suggesting that susceptibility to 12-O-tetradecanoylphorbol-13-acetate promotion under these protocol conditions is inherited as a dominant trait. The BALB/c X SENCAR F1s showed an average response that was intermediate between the two parental strains/stocks. Regarding the progression, all F1s showed a cumulative number of SCCs similar to the SENCAR progenitor. We also investigated the previously described switch of keratin 1 to 13 as a marker of premalignant progression, which is significatively delayed in SSIN mice compared with SENCAR mice. The SSIN X SENCAR F1s expressed this switch in a way similar to the SENCAR mice. These findings suggest that susceptibility to tumor progression is inherited as a dominant autosomal trait. The putative gene(s) that confers susceptibility is present in the SENCAR stock and was probably lost in the selection and inbreeding of the SSIN mice.