How do RNA-binding proteins control splice site selection?

Most genes contribute to the life of an organism by encoding proteins. The level of protein expressed depends on the level of transcription (in which an RNA copy of the gene is made) and the level of translation (when the RNA copy is used as a template for protein synthesis). In animals and plants, there is a step in between in which most of the RNA is spliced out of the original RNA copy, leaving a much smaller RNA sequence to be translated. In complex organisms, particularly vertebrates, the RNA copies of many genes can be spliced in a number of different ways. This means that a number of different proteins can be produced from one gene. This is amazing, and it happens to the greatest extent in the human brain. Unlike the other processes, splicing does not so much control the LEVEL of protein made, but rather it determines WHICH protein is made. We can squeeze around 8-10-fold more proteins out of our genes than would have been expected, and the different variants are expressed in different parts of the body, different cell types, at different stages in the life of a cell, or in response to ageing or disease. This incredible extra layer of flexibility has been achieved by weakening the simple system for recognising splice sites that is seen in lower eukaryotes (like yeast). Instead, our genes are full of sequences that could be splice sites. How does the cell recognise the right sites? How does it switch when required from one set of sites to another? These processes are controlled by a large number of proteins that bind to the RNA and activate or repress potential splice sites. How do they do this? This is a really critical process, essential for life, required for memory, diurnal rhythm, development and almost every other healthy biological process, a cause and contributor to disease when it goes wrong, a potential target for therapies... but the answer to the last question is that we do not know, despite years of investigations. Splicing is complex. There have recently been stunning advances in understanding the process of splicing the RNA after the right sites have been identified, but our understanding of how the sites are identified has barely changed in 25 years. Our conceptual principles have been outstripped by data. One of the most unsettling things we have come to realise is that some portions of the RNA can be bound by numerous proteins, activators and repressors, all of which have effects on the outcome, but they cannot all fit on at once. Do the proteins bind independently, weakly and transiently, and the outcome is a matter of chance that a particular protein is bound at a particular moment, do they bind stably in defined combinations, where several combinations might somehow permit splicing and others block it, or do activators and repressors actually bind competitively at a single crucial site? How does the next step work? How do activators activate or repressors repress? Do they form direct contacts with spliceosomal proteins or alter the flexibility and freedom of movement of the RNA or each other? How do they contact each other? We can answer these questions. We have developed a way of looking at single molecules of RNA in nuclear extracts (which support splicing). By labelling just two types of protein with fluorescent dyes, we can determine whether they both bind the same molecule of RNA and how many of them are bound. We have also tested whether a particular activator protein communicates with splice sites by 3D diffusion or by propagating complexes along the RNA. Based on such experiments, we have recently published a breakthrough paper that describes evidence for new molecular mechanisms for the activator. By testing lots of proteins in pairs, we can determine their binding patterns, and then their modes of communication. We propose to use innovative methods to look at the properties and interactions of the proteins, and then observe their binding in real time.