A random walk in oncogene space: the quest for targets.

Abstract

Introduction Oncogenes code for components of cellular growth-regulatory signals (1, 2). Such signals typically start at the cell surface, traverse the cytoplasm, and then terminate in the nucleus. They are propagated by protein-protein interactions between modular recognition domains and controlled by conformational changes and protein phosphorylation. The oncogenic potential of these growth-regulatory signals is linked to a mutational gain of function in one of the signaling components. While genetic analyses have identified the specific biochemical activities of these mutated signal components, structural studies have assigned these activities to distinct protein domains. As components of signaling pathways, oncogenes are bound to interact with each other. Conceptually, these interactions define a genetic space with hierarchical order that reflects the inherent directionality of signaling: each upstream mediator feeds its signal to downstream targets. Exploring oncogene space is essentially a quest for targets. The first part of this review will explore the oncogenic determinants of a common cytoplasmic signaling pathway. The results show that whereas a growth regulator can be oncogenic in one cell type, it can enhance differentiation in another. The second part of this review will focus on nuclear effects of growth signals. Signaling networks are branched trees with cross-talk between different components that ultimately regulate a transcription factor. The transcription factor then converts the signal into a program of gene expression that affects the specific phenotype of the cell. Oncogenic signals corrupt normal programs of gene regulation, and, as a result, the patterns of gene expression in the neoplastic cell differ from those of its normal progenitor. Genes that are differentially expressed in the cancer cell control the malignant properties of that cell. An understanding of the neoplastic cellular phenotype requires the identification of these differentially expressed genes that are targets of oncogenes and of tumor suppressor genes. The ultimate goal of the search for targets is to isolate a few genes that, when differentially expressed, will transform a normal cell into a cancer cell. This is an attainable goal if the number of these oncogenic effector genes is small. The aim becomes unrealistic if the number of genes required to induce and maintain the neoplastic phenotype is large.