"We don't know this yet, but one can guess that different exons may change the way these proteins interact or send messages," Darnell says. "If that is so, by changing these exons, you can modulate the quality of the synapse in a very powerful way—'powerful' meaning very regulated." Darnell expects that as similar studies are done with other splicing factors, they too will be found to regulate similarly coherent groups of proteins.

Cells use alternative splicing to increase protein diversity toward a host of biological ends.

Mutations that alter splice sites often cause entire exons to be excluded, severely damaging the encoded protein; these types of mutations are often associated with human disease. In the case of alternative splicing, however, a cell usually splices a single transcript in multiple ways to generate an assortment of proteins. So some mutations that alter a splice site or a nearby regulatory sequence have subtle effects—shifting the ratio of the resulting proteins without entirely eliminating any form. Several human diseases illustrate that these sorts of mutations still have the potential to be devastating. Alternative splicing errors are known to contribute to some growth deficiencies; a urogenital disorder known as Frasier syndrome; a type of cystic fibrosis; and a condition known as frontotemporal dementia and Parkinsonism. In the latter case, mutations disrupt the normal splicing pattern of a transcript called tau, which produces a protein that helps give a nerve cell its shape. Normally, the cell uses the transcript to produce six different forms of tau protein. Interfering with splicing, however, can result in an excess of some of those forms, causing tau to clump together in the brain and bring about progressive dementia.

		View Image Medical illustrator Graham T. Johnson's conception of alternative splicing: Two RNA polymerases (green globular proteins) traverse a gene (orange double helix) as they move along the DNA from the top of the picture toward the foreground at the bottom. The RNA polymerases each make a copy of the gene in the form of pre-mRNA. Each exon (ribbon-like structure) is shown in a different color, and introns are darkened and gray. As the gene is transcribed, spliceosomes (yellow globular structures) excise the introns and select exons, and splice together the remaining exons. Excised "lariats," coated with various RNA-binding proteins, drift away from the transcription complex. Image: Graham T. Johnson

As it becomes clear that alternative splicing is more the rule than the exception, scientists are realizing that, to make effective use of the enormous amount of data being generated by genome-sequencing projects, they must understand how the cell's splicing machinery processes that information. "Proteins are the major workhorse of the organism. If we want to have a really big-picture view of how organisms develop and function, we need to know what all the proteins are," says Brenton R. Graveley, an associate professor at the University of Connecticut Health Center. "In cases like Dscam, where you have 38,000 different proteins that are made from that gene, knowing just one of them is not terribly useful."

Christopher Burge, an associate professor at the Massachusetts Institute of Technology, likens the state of the splicing field to an earlier phase in the ongoing quest to decipher the genetic code. "Back in the early sixties," he says, "they didn't know whether it was a triplet code, a doublet code, a tetramer code, or not of fixed length. And I think that's where we are with splicing. We don't yet know what the code looks like."

To delve into that code, Burge has scoured human and mouse transcript databases—enormous collections of gene sequences generated from mRNAs—for features that distinguish alternatively spliced exons. The presence of a splice site, Burge says, is not sufficient to know that splicing will occur; nearby enhancer and repressor sequences—short segments of RNA that serve as landing pads for regulatory proteins—are equally crucial. The splicing of a single exon, he estimates, is likely promoted by at least three to seven enhancer sequences.

By analyzing the complete sequence information of genes known to be alternatively spliced in both mice and humans, Gene Yeo, a graduate student in Burge's lab (now a fellow at the Salk Institute for Biological Studies), developed a profile of a "typical" alternatively spliced exon and later identified another 2,000 exons that fit the description. Of these, at least 70 percent appear to undergo alternative splicing in both humans and mice.

	PAGE 3 OF 6	Go Back | Continue