If the expression of mRNA in E. coli is characterised by the logic and reason implicit in the Jacob-Monod models, then in eukaryotes it is dominated by the irrationality of introns. Splicing restores sense, in a sophisticated process for removing the introns from RNA that may involve the recognition of numerous signals and a precise reaction between partner splice sites that are usually thousands of bases apart. In some genes, each of 80 or so introns has to be removed, without errors. Nothing is simple, though, in post-modern gene expression. Splicing is the one process that has emerged with its reputation enhanced after genome sequences produced the galling discovery that nematodes have as many protein-coding genes as mammals; the ability of splicing to bring together different partner sites (alternative splicing) allows most genes to produce more than one isoform of mRNA (and in most cases, more than one protein). These alternative splicing patterns are often regulated by signaling or development. Hence a system exhibiting precision and reliability retains considerable flexibility.
Splicing is not a biochemist's dream. With 100-300 proteins and several RNA molecules involved, it is too complicated to reproduce it with purified components, and reactions are done in crude nuclear extracts. These experiments have shown that the "spliceosome" is dynamic; it is assembled in stages on each intron, apparently, and then undergoes major conformational changes to activate itself. Unfortunately, it has been possible to analyse only a few points en route, where assembly can be stalled. In between these points, the reactions lose synchrony and individual complexes may not assemble via identical pathways. Furthermore, current methods for detecting the addition of components do not, in general, lend themselves to analysis of very short time intervals and they may require the separation of complexes from unbound components, a step that can lead to the loss of components that dissociate readily. In collaboration with Professor Bagshaw in Leicester, we are trying to develop methods for a kinetic analysis of spliceosome dynamics.
These problems are at their most exasperating when we try to explain the selection of splice sites. It has been clear for some years that sequences within the exons and introns (that is, not part of the conserved constitutive signal elements at the junctions) have important effects upon splice site selection or exon use. Some of these are bound by proteins that activate or repress splicing. We have shown that the best known of these proteins bind with low stringency and competitively to the pre-mRNA, and that their effects on 5' splice site selection arise because they indiscriminately strengthen or compete with the binding of U1 snRNPs, affecting the level of occupancy of multiple sites on the pre-mRNA by the snRNPs.
Clearly, the levels and positions of binding of activator and repressor proteins are likely to play an important role in determining whether splicing factors use splice sites. However, the plethora of sequences and proteins that bind to them suggests that introns and exons might be swathed in proteins that create zones within which splicing is inhibited or permitted. Little progress can be made biochemically until we know whether modifying proteins form assemblages on the pre-mRNA before the splicing factors. Nonetheless, we have recently developed a method (TOES) in which exons lacking sufficient positive signals for splicing can be activated in vitro and in vivo, and we are currently engaged in improving this method and investigating its mechanistic implications.
Key lab techniques: Targeted oligonucleotide enhancers of splicing (TOES); time-resolved psoralen cross-linking; quantitative RNase H assays to measure dissociation rates of U1 snRNPs etc.; use of RNase H to measure simultaneous occupancy of multiple sites.
Key lab reagents: Dual reporter vectors for splicing assays involving enzyme activity (luc/beta-gal) or fluorescence.
Lab contact: Ian Eperon: email@example.com 
Lab website: http://www.le.ac.uk/biochem/staff/eci/eci.html