Search

Mice with skin tumors induced either by 7,12-dimethylbenz[a]anthracene complete carcinogenesis or subcutaneous injection of a carcinogenic keratinocyte cell line showed moderate to severe splenomegaly as a result of an increase in splenic granulocyte-macrophage and erythroid (erythroid burst-forming unit) progenitors. To test whether the observed alterations involve the release of soluble factors by the epidermal component of skin tumors, we used an in vitro approach. A series of mouse keratinocyte cell lines resembling progressive stages of skin carcinogenesis and carrying either normal or activated Ha-ras genes were assayed for their ability to produce the factors required for colony growth of hematopoietic-committed progenitors. Only the conditioned media of keratinocytes harboring activated Ha-ras genes were able to support the growth of granulocyte-macrophage colony-forming units. In addition, preincubation of normal bone-marrow cells with conditioned media from the transformed epidermal cell lines stimulated in vitro amplification of the hematopoietic granulocyte-macrophage progenitor compartment. To identify the possible factors responsible for the activities detected in the keratinocyte-conditioned media, we performed northern blot analysis using the cytokine probes granulocyte colony-stimulating factor, macrophage colony-stimulating factor, granulocyte-macrophage colony-stimulating factor, stem cell factor, interleukin-1 alpha, interleukin-3, and tumor necrosis factor-alpha. The cell lines expressed different cytokine mRNA combinations that positively correlated with the colony-stimulating activity detected in the corresponding conditioned medium. These results suggest that transformed epidermal tumor cells in vivo may alter normal hematopoiesis as a consequence of the production of cytokines that act in autocrine or paracrine loops probably related to tumor growth.

We have studied in this report the expression of keratins in mouse epidermal keratinocytes transformed in culture by a chemical carcinogen (PDV) and in cell lines (PDVCM1, PDVCM2, PDVCV1, and PDVC57) derived by tumor transplantation of PDV cells in syngeneic C57Bl/6 mice. PDV, PDVCM1, PDVCM2, and PDVCV1 cell lines are weakly to moderately tumorigenic, giving rise to squamous cell carcinomas, although not at all injection points, whereas PDVC57 cells are more malignant, inducing highly anaplastic carcinomas at 100% of injection sites. All the cell lines synthesize anomalously simple epithelial keratins, substantial amounts of K8, and minor quantities of K18 and K19, but the level of expression is increased in PDVC57. We have found that in PDVC57 cells upregulation of K8 is linked to down-regulation of the normal keratins produced by epidermal keratinocytes in culture (i.e., K5, K6, K14, and K17). On the other hand, K8 does not generally colocalize with K13, a keratin also aberrantly expressed by epidermal cell cultures when induced to differentiate by high Ca2+ medium. K13, normally synthesized in internal stratified epithelia, is anomalously induced in mouse epidermal tumors and has been used as an early marker of carcinoma progression. In tumors induced by the cell lines upon injection in mice, K8 is found in the less differentiated regions as opposite to K13, restricted to the differentiating areas of the tumors. In PDV, PDVCM1, PDVCM2, and PDVCV1 carcinomas the overall expression of K13 is higher than that of K8. However, this relation is inverted in tumors induced by PDVC57 cells, in a good correlation with the tumoral phenotypes produced by the cell lines. Our results suggest that upregulation of the simple epithelial keratin K8, as found in transformed epidermal cell lines and tumors, is a late marker of malignant progression and is associated with the loss of the differentiated phenotype.

Activation of a Harvey ras (H-ras) protooncogene is a frequent event associated with mouse epidermal carcinogenesis. We report that the transfection of a human H-ras oncogene into an immortalized mouse epidermal cell line (MCA3D) induces the anomalous expression of cytokeratins (CKs) 8 and 18 characteristic of simple epithelia. The comparison of various transfectant cell clones indicated a direct correlation between the levels of CK8 expression and the mutated H-ras p21s. The expression of simple epithelial CKs is also described in cell lines derived from mouse skin carcinomas (HaCa4, CarC) and in keratinocytes transformed in vitro by a chemical carcinogen (PDV, PDVC57), all of which contain altered H-ras genes. The induction of CK8 and CK18 occurs at the mRNA level and, although both CK8 and CK18 mRNAs are expressed, CK18 protein does not accumulate whereas CK8 is incorporated into intermediate filaments. Immunofluorescence studies show that the pattern of CK8 protein expression is heterogeneous; some cells express very low amounts of CK8, whereas others synthesize relatively high levels of this protein. However, selection of strongly CK8-positive cells was found in one case where a more malignant population of cells (PDVC57) was derived by tumor transplantation of PDV. Our results suggest that activation of a H-ras gene can alter the normal differentiation program of epidermal cells and that the ability to synthesize CK8 and CK18 could be related to tumor progression.

Keratins have been demonstrated to be suitable markers of changes taking place during epithelial neoplasia. Therefore, we analyzed 18 mouse skin tumors (nine papillomas and nine squamous cell carcinomas), induced either by two-stage carcinogenesis with 7,12-dimethylbenz[a]anthracene(DMBA)/12-O-tetradecanoylphorbol-13-acetat e or complete carcinogenesis with DMBA, by immunofluorescence with a monoclonal antibody to keratin (K) 8 (TROMA-1). Immunoperoxidase staining and immunoblotting were also used on selected tumor samples to further explore for the presence of K8. All of the papillomas tested were negative for the presence of K8, whereas the carcinomas were positive. The level of K8 expression in carcinomas showed a positive correlation with the degree of malignancy. Northern blot analysis using a K8 cDNA probe suggested that control of K8 expression in mouse skin tumors occurs at the transcriptional level. Double-label immunofluorescence staining using TROMA-1 and RK13 antibodies demonstrated that K8 did not generally colocalize with K13, a keratin normally found in internal stratified epithelial but aberrantly expressed in mouse epidermal tumors. Furthermore, tumors expressing high levels of K8 showed a reduced expression of K13. Histological examination of immunoperoxidase-stained tumors demonstrated that K8-positive cells were mainly found in anaplastic areas, whereas K13 foci were restricted to well-differentiated regions. Our results demonstrate that K8 expression is a marker of late stages of carcinoma progression in the mouse skin carcinogenesis model.

Our previous results on the genotoxic effect of 8-methoxypsoralen-induced lesions on pBR322 suggested an important involvement of an inducible error-free repair pathway in the repair of plasmid lesions. We present herein further results obtained in order to explore that possibility, together with a more general report on the subject. pBR322 treated with increasing concentrations of 8-MOP plus fixed UVA light irradiation was used to transform several E. coli strains differing in their repair capacities, and plasmid survival and mutagenesis were determined. Survival results suggested that crosslinks were completely lethal in pBR322 whereas monoadducts were partially removed from plasmid DNA mainly through an error-free excision pathway. A mutagenic repair pathway did not show a significant contribution to the total repair process. Cell preirradiation stimulated plasmid recovery in recA+ strains, including the umuC strain, thus confirming our previous results indicating that an inducible error-free repair had occurred. Globally, our results showed a strong requirement on the excision pathway for the repair of psoralen-damaged plasmid DNA. In contrast, the recA dependent pathway was needed only for SOS induction. After a theoretical correction of the data for estimating the effect only due to 8-MOP adducts, a different pattern of repair mechanisms appeared to be involved.

We have studied the lethality produced on pBR322 by near-UV radiation and by 8-Methoxypsoralen plus near-UV (PUV treatment). Samples of pBR322 DNA were irradiated with increasing fluences of 360 nm-light either in the absence or presence of 400 molecules of 8-Methoxypsoralen (8-MOP) per plasmid molecule. We have estimated to what extent the global lethality of PUVA treatment is due to the presence of psoralen adducts in DNA or to radiation itself. In order to analyse the involvement of DNA repair mechanisms in the removal of plasmid lesions, several strains of E. coli (differing in their repair capacities) were used as recipients of the treated plasmids. Results showed that excision and recombination participate in the repair of near-UV-induced plasmid lesions. Repair of PUV-induced lesions showed an even greater requirement of the excision pathway. Besides, a slight increase on plasmid mutation frequencies was observed after near-UV or PUV treatment in wild type and uvrA cells. Estimation of the contribution of 8-MOP to the global lethality of PUV treatment showed that only the excision pathway was involved in removing psoralen adducts from plasmid DNA, suggesting the involvement of the recombinational pathway in the repair of near-UV-derived lesions.

The effect of temperature on the activities of acyl-CoA:lysolecithin acyltransferase and acyl-CoA hydrolase has been studied in microsomal preparations. The enzymes had different thermal stabilities, the hydrolase being more stable. The temperature dependence of enzyme activities was studied either in 4 M NaCl-washed microsomes or NaCl-washed detergent-treated microsomes. Both preparations showed little increase in acylation rate when the temperature was increased, whereas hydrolase activity was increased markedly. This increase in hydrolytic activity was higher when lysolecithin was absent. Based on this behavior of the enzymes, the relative accuracies of the spectrophotometric and radioactive assay methods are discussed.

The genotoxic effect of 8-methoxypsoralen damages (monoadducts and crosslinks) on plasmid DNA was studied. pBR322 DNA was treated with several concentrations of 8-methoxypsoralen plus fixed UVA light irradiation. After transformation into E. coli cells with different repair capacities (uvrA, recA and wild-type), plasmid survival and mutagenesis in ampicillin- and tetracycline-resistant genes were analysed. Results showed that crosslinks were extremely lethal in all 3 strains; indeed, it seemed that they were not repaired even in proficient bacteria. Monoadducts were also found to be lethal although they were removed to some extent by the excision-repair pathway (uvrA-dependent). Damaged plasmid DNA appeared to induce mutagenic repair, but only in the wild-type strain. In order to study the influence of the SOS response on plasmid recovery, preirradiation of the host cells was also performed. Preirradiation of the uvrA or wild-type strains significantly increased plasmid recovery. Consistent with the expectations of SOS repair, no effect was observed in preirradiated recA cells. Plasmid recovery in the excision-deficient strain was mainly achieved by the mutagenic repair of some fraction of the lesions, probably monoadducts. The greatest increase in plasmid recovery was found in the wild-type strain. This likely involved the repair of monoadducts and some fraction of the crosslinks. We conclude that repair in preirradiated repair-proficient cells is carried out mainly by an error-free pathway, suggesting enhancement of the excision repair promoted by the induction of SOS functions.