To a greater or lesser extent, all malignant cells lose the ability to undergo terminal differentiation, which provides them with a selective advantage over surrounding normal cells. Human keratinocytes (HKcs), the normal counterparts of squamous cell carcinomas (SqCCs), are an ideal system for studying differentiation in vitro. The terminal differentiation of keratinocytes involves two basic processes: 1) Growth arrest, and 2) Induction of specific structural proteins and enzymes, such as keratins 1 and 10, involucrin, and transglutaminase type I, that are characteristic of mature squamous epithelial cells. HKcs can be induced to terminally differentiate in vitro by the addition of serum and calcium to the culture medium. SqCC lines do not stop proliferating under these conditions, and fail to express differentiation-specific keratins, involucrin, and transglutaminase I. Although the cell biology of keratinocyte differentiation has been studied extensively, the genetic controls of this process are unknown. The long-term goal of this proposal is to identify the genes that control terminal differentiation of human keratinocytes, and the genetic lesions that lead to loss differentiation in squamous cell carcinomas. Preliminary studies using somatic cell hybrids between one particular SqCC (FaDu) and HKc demonstrated that loss of expression of the differentiation-specific-protein transglutaminase I in vitro, and the ability to form tumors in vivo are recessive traits. Hybrid cell lines regained expression of the differentiation-specific proteins, and were not tumorigenic. Hybrids between the two SqCC lines FaDu and A253, and between FaDu and the SqCC line CE-48, conformed the recessive nature of the defect in TGase I induction, and revealed at least two complementation groups for loss of TGase I induction in response to differentiating signals. Identifying and characterizing the genetic lesions underlying these complementation groups is the goal of this proposal. Three strategies will be used to identify and characterize the genetic lesions that cause disruption of TGase I expression, and therefore of normal differentiation in SqCCs. First, the number of complementation groups for TGase I will be determined by pair-wise hybridizations of a panel of SqCC lines. Northern and Southern analysis of the SqCC lines with a TGase I cDNA probe will be used to initially characterize the nature of the defect in these lines. Finally, recessive genes will be identified by the differential display of mRNA transcripts from SqCC lines in different complementation groups for loss of TGase I induction. Once clones of interest have been isolated, they will be partially sequenced, and this information will dictate the strategy used to characterize the gene products.
