Two faces of the same coin

Today, it is unremarkable that uncontrolled cell proliferation or cancer is the outcome of a developmental process gone awry. Just over 15 years ago, however, the initial finding that regulators of cell growth were also important for normal development, and that these two processes were driven by common signalling pathways, created a stir both within and outside the developmental biology community.

In 1987, a series of papers reporting the sequences of developmentally important genes, such as sevenless and decapentaplegic (Dpp), provided some of the earliest molecular links between development and cancer. Genetic studies in Drosophila melanogaster had already shown that the sevenless gene was required for proper differentiation of photoreceptors in the developing eye. At this time, it was generally appreciated that cell–cell interactions were fundamentally important for developmental processes, and the product of the sevenless gene was predicted to translate information from cell–cell contacts into the appropriate developmental outcome. But the molecular nuts and bolts of how this was achieved remained obscure. After sequencing the sevenless gene, Gerry Rubin and colleagues discovered that it encoded a predicted transmembrane protein bearing a tyrosine kinase domain that is highly related to domains that are present in viral oncogenes and hormone receptors. Based on this homology, they correctly prophesized that the molecular mechanisms of signalling would be the same in both cases — a prediction that came to fruition a few years later.

As other cloning and sequencing studies were published in the same year, the connections between development and growth control became clearer. Dpp, which specifies dorso ventral patterning in the fruit fly, and Vg1, a maternal transcript in frog eggs, were both found to encode members of the transforming growth factor-β (TGF-β) family of growth regulators. In another groundbreaking paper from the Nusse laboratory, published in the same year, the cancer–development link was extended to mammals. These authors set out to clone the Drosophila melanogaster homologue of the mouse mammary oncogene, int-1, by genomic hybridization, and they showed that it was identical to the Drosophila patterning gene wingless. They speculated that the int-1 oncogene might shift the balance between proliferation and differentiation in mouse epithelial cells and changes in this homeostasis might lead to tumorigenesis. Although these early studies did not reveal the molecular basis for these observations, what emerged quite clearly was that the overlap between cancer, growth-factor signalling and developmental regulation was a fundamental characteristic of cell division and differentiation.

Building on their earlier work on sevenless, Rubin and colleagues carried out a genetic screen to identify mutations that enhanced the phenotype associated with sevenless mutations, hoping that this approach would allow them to identify components of its signalling pathways. In 1991, Michael Simon, along with colleagues from the Rubin laboratory, reported that two of the isolated enhancers of sevenless encoded known signalling molecules — the GTPase Ras, and its guanine nucleotide exchange factor, Cdc25. Ras was not only a characterized oncogene that was known to have transforming ability, but it had also been shown to function downstream of other protein tyrosine kinases and to regulate development in the worm. As predicted 4 years earlier, the parallels between sevenless and protein tyrosine kinases seemed to extend to their signalling pathways. At this stage, however, the possibility remained that Ras might influence sevenless through an independent pathway, and conclusive proof that Ras acts in the sevenless pathway only came with subsequent biochemical analysis. However, biochemical links between Ras and other receptor tyrosine kinases did exist, so these genetic data also supported a physiological role for Ras in receptor tyrosine kinase signalling.

During the past decade, the number of developmental regulators that are also associated with cancer has multiplied, as has our understanding of how these processes are intertwined. These early studies — a testament to the combined predictive power of genetics and sequence homology — were profoundly influential in establishing a conceptual framework for considering cancer and development as related processes or as 'two faces of the same coin'.