Mechanisms of action of the oncogene SRSF1: an inter-disciplinary approach

The process known as pre-mRNA splicing is one of the most important steps in gene expression. Almost all protein-coding genes contain long stretches of sequence (introns) that are removed by the process of splicing before the mRNA is translated into protein. Removing these introns requires recognition of the correct signals at both ends of the intron. This is clearly a task of some difficulty, since the sequences that specify the sites of splicing are very short and many similar sequences are scattered throughout the sequence of any gene. Moreover, it is clear that the candidate sites compete with each other; thus, mutations to one site can result in the use of another candidate, or different molecules of RNA may use different sites for splicing, a process known as alternative splicing. Alternative splicing is a mechanism that is used by most genes to generate an average of seven different forms of mRNA per gene; alternative splicing patterns may be restricted to specific tissues or may coexist within a cell. There is rapidly accumulating evidence that changes in alternative splicing either cause serious diseases or are required for development of the condition.

One of the most important proteins involved in controlling alternative splicing is known as SRSF1. This is a proto-oncogene, meaning that slight increases in the level of expression can produce tumours. It binds to a very large number of transcripts, and few introns that have been tested by experiment do not respond to increased levels. The effect of SRSF1 is to cause parts of the RNA that otherwise might be spliced out to be spliced into the final mRNA and to change the selection of alternative sites at the ends of an intron in such a way as to reduce the intron length. In this way, it affects which parts of the RNA end up in the mRNA that gets translated and thus it affects the sequence of the protein encoded by the mRNA. It seems that in many (but not all) transcripts involved in cell growth control the outcome is to favour the production of proteins that spur on tumour development.

If we understood how SRSF1 worked, we might be able to develop chemical drugs that impede its action. The problem at the moment is that its actions are so complicated and its binding to the other proteins involved in splicing is so weak that it is really not at all clear how it works. We need to do is to find out whether it does get involved in the basic splicing reactions and whether this is connected to its effects on selection. The most urgent step is to find out how many copies bind to a pre-mRNA molecule undergoing splicing and whether regulation by SRSF1 involves new proteins being recruited, singly or in hordes, or whether sequences that enhance splicing just improve binding of a molecule that then acts just like any other molecule of SRSF1.The next thing we need to do is to find out where SRSF1 contacts the pre-mRNA, and whether those enhancers alter the number of contacts; does it work deeply inside the catalytic machinery or does it just help it start and direct where to splice? We know SRSF1 is affected by phosphorylation of side chains, in large numbers. We want to find out whether all of these changes are important and whether they have to be reversible. There is a suggestion that the protein undergoes major changes in shape after phosphorylation. This is very important, because one line for therapy might be to find chemicals that stop it changing. Hence, we need to know whether it really does change shape and how this connects to its activity. If we had a good assay for monitoring the protein's conformation, it might be possible to screen large collections of chemicals to find inhibitors.

All these are critical questions for unravelling the actions of SRSF1, but they are beyond normal methods available within any one discipline. We have recently developed methods that combine physics, biology and chemistry, and we will use them to solve all these questions.