The situation turned out to be far more complex. The team found that the Dscam gene in fact produced numerous variants. And when they pored over the newly released sequence of the fruit fly genome, they realized that Dscam had the potential to generate a mind-boggling 38,016 distinct forms of proteins.

It's now known that Dscam is an extreme example of "alternative splicing"—a variable but carefully regulated adaptation of a routine RNA modification process that allows a single gene to give rise to multiple versions of a protein. While most cases are more modest than Dscam, the reality of alternative splicing has still managed to turn on its head the "one gene, one protein" principle that has guided genetics for more than half a century.

The ability of alternative splicing to exploit limited genetic information to generate a multitude of proteins is important because each cell in an organism depends on a highly specialized set of proteins to carry out its unique function. Even within a single cell, the set of required proteins varies as a function of the stage of development and changing environmental conditions. With the latest estimates putting the content of the human genome at only 20,000 to 25,000 genes, most scientists believe a one-to-one ratio of genes to proteins just cannot be enough.

		Robert Darnell Robert Darnell studies degenerative brain disorders that are provoked by an immune response to certain cancers. Douglas Black Douglas Black researches the regulation of pre-mRNA splicing in differentiated cells, particularly neurons. Lawrence Zipursky Lawrence Zipursky is interested in uncovering the mechanisms by which neurons make highly specific patterns of connections during development.

Douglas Black, another HHMI investigator at UCLA, points out that alternative splicing allows for much more complexity than the size of the genome suggests. It's too soon to know just how big the human proteome is, but "it's certainly much larger than the number of genes," he says. "The analyses that have been done seem to say that most genes have two or three splice variants per gene." And while not all of these transcripts (RNA segments) produce functional proteins, there are other genes that generate far greater diversity—hundreds or even thousands of forms.

The first example of alternative splicing in cells—there were earlier examples in viruses—was found in a gene called IgM in 1980. It was considered an anomaly. "But if you were paying attention, you started to see more and more examples of genes that were alternatively spliced," says Black. Computational biologists are now trying to assess the frequency of alternative splicing more accurately. "If you now ask how many of the known human genes are producing more than one splice variant, almost everyone says at least 50 percent," Black says. "Some people argue that 70 percent have alternative splicing. And while nothing as complicated as Dscam has been discovered in mammalian cells, there are many transcripts that can produce hundreds of proteins."

Cells use alternative splicing to increase protein diversity toward a host of biological ends. Some of the best-studied examples derive from fruit fly development, where the splicing of several related genes dictates whether an embryo will develop into a male or a female. Similarly, alternative splicing can allow one gene to generate different proteins in different tissues—many of the highly specialized proteins in the brain, for example, come from differential splicing of genes that are also expressed in other tissues. Cells can even modify splicing in response to changing conditions: An ion channel transcript studied by Black's lab produces a protein whose sensitivity to calcium depends on which exons (segments of DNA that encode a protein's amino acid sequence) are included.

Recent work in the lab of HHMI investigator Robert B. Darnell at the Rockefeller University shows that not only can alternative splicing tweak the structure of a single protein, but it may also be a means of regulating entire pathways. In the first genome-wide screen for the targets of a tissue-specific splicing factor, Darnell showed that Nova, a protein found only in the brain, controls the splicing of 49 mRNAs to produce proteins not found in other tissues. Almost all of these proteins help nerve cells transmit their signals across the synapse.

	PAGE 2 OF 6	Go Back | Continue