By using the antibodies we developed last year, we have shown that 1) steady state levels of notch-1 protein, but not of notch-1 mRNA, are downregulated during terminal differentiation of Friend murine erythroleukemia cells, indicating possible post-translational mechanisms controlling the amount of notch-1; 2) this phenomenon is observed independently of the agent used to induce terminal differentiation (hexamethylene-bis-acetamide [HMBA] or sodium butyrate); 3) levels of notch-1 protein are transiently upregulated very early during HMBA-induced Friend cell differentiation (1 h) through a pathway that seems to require Ca2+- independent isoforms of protein kinase C. The mechanism and possible role of this early upregulation are currently under investigation. In addition, 4) similar patterns of notch-1 expression are observed when SY5Y human neuroblastoma cells are differentiated with retinoic acid and when Jurkat T cells are activated in vitro. The hypothesis we are currently testing is that the early upregulation of notch-1 controls the timing of cell differentiation, by allowing the cell to remain in the cell cycle and become competent to receive a second signal which induces growth arrest and terminal differentiation. Preliminary evidence in support of our hypothesis was obtained by treating Friend cells with several different phosphorothioate antisense notch-1 oligonucleotides in addition to HMBA. Such treatment significantly reduces (compared to sense oligonucleotides) both the total number of cells and the percentage of differentiated cells. If this hypothesis is confirmed, manipulating notch-1 levels or its signal transduction pathway may be used as an adjuvant to treatments intended to cause terminal differentiation and cell death in human malignancies. We are currently developing stably transfected clones of several cellular differentiation models with antisense notch-1 constructs and with oncogenic notch-1 constructs to study the relationship between notch activation, cell cycle and cell differentiation. We have also generated monoclonal antibodies directed to the putative ligand-binding region of human notch-1. These antibodies will be used to study: 1) levels of notch-1 expression in hematopoietic malignancies and 2) signal transduction from wild-type notch-1 in in vitro differentiation models. Published reports suggest that a number of human solid tumors considerably overexpress notch-1 compared to untransformed tissues. Thus, our eventual objective is to test whether one or more of these monoclonal antibodies has potential diagnostic and/or therapeutic applications in human malignancies. Finally, we have developed 2 strains of notch-1 antisense transgenic mice. We are currently expanding colonies of these strains and we will have generated homozygotes by this fall. These mice may provide us with the first model of partial notch-1 deficiency during postnatal life. In addition, such a model may help us predict the possible effects of manipulating notch-1 expression in vivo for therapeutic purposes.
