In vitro multistage carcinogenesis: A balance of genetic alterations and selective pressures.

Abstract

The long latency of human tumors and the difficulty of transforming human cell lines has necessitated the investigation of rodent models which have helped propagate a general theory of carcinogenesis. This author has addressed the temporal and physiological significance of oncogene activation and tumor suppressor loss during malignant progression of nontumorigenic murine keratinocytes. The data presented here indicate that mechanisms of carcinogenesis may be dependent upon both the model system studied, and on innate environmental selective pressures. 7,12-dimethylbenz(a) anthracene initiation of murine keratinocytes resulted in tumorigenic phenotypes with transferrable dominant transforming activity. However, mutated Ha-ras alleles were not detected in the tumorigenic cell lines despite the direct correlation between chemical initiation and ras activation in vivo. Somatic cell hybrid studies implicated the loss of multiple tumor suppressors in the progression of a normal cell through a benign state to frank malignancy. While all hybrids were suppressed to varying degrees, nontumorigenic cells suppressed tumor formation to a greater extent than benign cells. Interestingly, a well differentiated carcinoma cell line also suppressed tumor formation, The implication of tumor suppressor loss in the formation of a benign papilloma was a novel finding. Results from these experiments revealed that disruption of differentiation programs is concomitant with tumorigenicity, and that the ability to differentiate in vitro is a prerequisite for tumor suppression in vivo. Indicative of an inability to differentiate, benign and malignant cell lines failed to express keratins K1 or K10 in response to high extracellular calcium concentrations. While differential northern analysis revealed that keratins K6 and K18 were predominately expressed by either non-tumorigenic or malignant cells, respectively, subtractive hybridization identified five putative gene sequences (Suz 1-4 and 6) expressed exclusively by non-tumorigenic cells. In summary, somatic cell hybrid studies helped assess the phenotypic dominance of a given stage of cancer. The isolation of genes whose expression is abrogated in malignant cells may further our insight into the role of gene loss during malignant conversion. Finally, a better understanding of the mechanisms of endogenous and exogenous selective pressures will enable the design of improved in vitro models of carcinogenesis.