All nucleated eukaryotic cells exhibit ligand-induced, transcription-dependent changes in phenotype. Even the single-celled yeasts respond to polypeptide mating factors by alterations in cellular proliferation and mating capability that are dependent on receptor-mediated transcriptional activation. In multi cellular organisms, individual cells constantly respond to a complex environ ment in which gene expression and'consequent cellular phenotype are medi ated by ligands such as growth factors, peptide hormones, neurotransmitters, antigens, morphogens, and other hydrophilic agents that bind to cell-surface, transmembrane receptors. Researchers studying area as seemingly diverse as mitogenesis, endocrinology, immunology, neurobiology, and developmental biology are all confronted with a similar set of questions: How is the informa tion that a ligand-receptor interaction has occurred at the cell surface transmit ted across the membrane? What are the intracellular mechanisms that relay, from the membrane to the nucleus, the information that a cell-surface, ligand-receptor interaction has occurred? What genes are influenced as a result of this new information, and how is their expression selectively mod ulated? What are the roles of these new gene products, expressed in response to ligand-receptor interaction, in effecting the subsequent alteration in phe notype? One major approach to these questions has been to identify those genes whose expression is initially induced by ligands, then work" upstream" to the signalling mechanisms that lead to their expression and" downstream" to their roles in cellular responses.

The first experiment reporting the cloning of cDNAs for mitogen-inducible messages, using differential hybridization to filter-replicas of cDNA libraries, demonstrated the feasibility of this approach and caught the interest of biologists studying the molecular basis of the mitogenic response (1). The subsequent identification of the cellular homologues of two oncogenes, vjas and v-myc, as mitogen-responsive genes likely to play a role in the growth response (2-4), demonstrated a satisfying logical relationship between normal and abnormal growth control. This relationship fired the imagination and enthusiasm of molecular oncologists, and greatly increased interest in identifying additional genes that shared the characteristics of rapid, transient induction by mitogens.