Alternative Splicing Goes Mainstream

In eukaryotic genetics, the one-gene/one-protein concept has, for the most part, breathed its last. Researchers have rallied behind mechanisms such as alternative splicing, which may allow a lowly 30,000-gene genome to produce the dizzying variety of proteins that some believe is necessary to produce beings as complex as humans. Alternative splicing--the post-transcriptional editing process that can result in various mRNAs--was previously seen as an interesting but relatively uncommon sideshow to the big-tent events covered in central dogma: DNA begets RNA begets protein. Now, thanks to a wave of new bioinformatics databases, techniques, and studies, alternative splicing has emerged as not only pervasive, but perhaps as a central regulatory mechanism. And it may not be alone. A previously unnoticed glut of premature stop codons in usable exons suggests a larger role for RNA decay mechanisms previously associated with quality control. Cousins such as alternative polyadenylation might play a similar role. These processes could band together as part of a vast regulatory network, controlling everything from developmental programs and cell differentiation to housekeeping functions and circadian rhythms. At the heart of this theory is a reassessment of how the gene-to-protein process works, especially with the confounding riddle of how a mouse and a human being can be so different, even though both have about 30,000 genes. "Thirty thousand genes are probably enough," says Miles Wilkinson of the M.D. Anderson Cancer Research Center in Houston. "The question isn't how many genes or even how many proteins, but how the regulatory system works to control the machinery. The regulatory system is probably far more complex than we ever imagined." MORE THAN A STOREHOUSE Such talk lends credence to John Mattick's theory of genetic regulation. Mattick, director of the Institute for Molecular Bioscience at the University of Queensland in Brisbane, Australia, has argued vociferously that nonprotein-coding RNA (including microRNAs, siRNAs and intronic RNA) are part of a vast multitasking regulatory mechanism. He points out that the intron-sparse yeast genome and the seemingly junk-dense human sequence are very similar. What differentiates them are the vast intronic regions shared by humans and other complex organisms. Mattick outlined his controversial theory in a paper published in 20011 in which he wrote, "The majority of phenotypic variation between individuals (and species) results from differences in the control architecture, not the proteins themselves." Probably an important cog in that as-yet-undefined regulatory system is the alternative splicing mechanism. Most genes contain at least one position for alternative splicing where the spliceosome can choose one of several alternative splice sites. Each splice variant mRNA may then code for a different protein. Alternative splicing is also seen as a method for enhancing evolutionary adaptability in an organism. "It's like a big storehouse of potential genetic information," says Douglas Black, a Howard Hughes Medical Institute investigator at the University of California, Los Angeles. "It's a mechanism that allows testing a new mutation without losing the wild-type version." Research shows that alternative splicing is far more pervasive than previously thought, perhaps too pervasive to be simply an evolutionary warehouse. In 1998, Peer Bork's group at the European Molecular Biology Lab in Heidelberg, Germany, mapped expressed sequence tags (ESTs) to human cDNAs and unexpectedly found that the observed putative splice variants exceed the 5% prediction offered by Nobel laureate Philip Sharp,2 one of the original discoverers of mRNA splicing. In 2001, they expanded the study to seven eukaryotic species and found that almost half the genes in animals could have alternative splice sites.3 Now, Bork says that even a prediction of 45% is a significant underestimate, based on his and others' work. "The best figure now is [that] somewhere around 60% of genes use alternative splicing," he says, referring to work done by Christopher Lee at UCLA.4 "That number has been increasing steadily at a linear rate in the last few years." One recent study, he points out, hints that almost all highly expressed genes use alternative splicing.5 Just as Bork's work seems to have proposed a solution to the genetic complexity question (that alternative splicing accounts for all those proteins), it also confounds the theory that humans owe their complexity to more alternative splicing in their genome. Other so-called lower species aren't very far behind in the preponderance of alternative splice sites. About 35% of murine ESTs showed them, as did at least 15% of Caenorhabditis elegans ESTs. When normalized, says Bork, the data for these organisms show similar splicing frequency to humans, meaning that alternative splicing by itself may not be enough to explain why humans drive cars and roundworms squirm in the dirt. Adapted from D.R. Mercatante et al., J Biol Chem, 276: 16411-7, 2001. GOOD GENE, BAD GENE: The BCL-X gene is a good example of how alternative splicing can make a big difference. If the gene is spliced to include all of Exon II, it will produce Bcl-XL mRNA, which results in a protein that inhibits apoptosis. If the gene is spliced without the non-shaded portion of Exon II, it produces Bcl-XS mRNA, which in turn creates an apoptosis-inducing protein. Many cancers have a high incidence of Bcl-XL, while successful chemotherapy results in a higher proportion of Bcl-XS. ANOTHER REGULATOR: NONSENSE The puddle got even muddier in January. Steve Brenner and colleagues at UC-Berkeley went fishing for connections between alternative splice sites and a cellular mechanism called nonsense-mediated mRNA decay. NMD is a process whereby mRNAs with premature stop codons, usually resulting from nonsense mutations, are identified and gobbled up by RNA-degrading enzymes. The Berkeley team compared the human genome to a refined EST database and then overlaid that onto the RefSeq database, which is by far the clearest and least error-prone genetic record, according to Brenner. What Brenner found was remarkable: More than a third of alternatively spliced exons had premature stop codons.6 Conventional wisdom states that NMD does the cell's quality control work. When a defective piece of mRNA is produced, NMD wipes it out to keep it from producing mutant proteins that harm the organism. But Brenner suggests that NMD, coupled with alternative splicing, might have an even more important role than that. "It's a regulatory mechanism for gene control," says Brenner. "Instead of turning a gene on or off in the genome, the gene can always be on." When a protein isn't needed anymore, says Brenner, the splicing machinery changes the splice site and plugs a premature stop codon into the mRNA being produced. NMD then clears that piece of unnecessary mRNA. Widespread premature stop codons in the human genome change the nature of this quality control mechanism. "We propose that regulated unproductive splicing and translation (RUST), through the coupling of alternative splicing and NMD, may be a pervasive, underappreciated means of regulating protein expression," he says. If Brenner is right, then in some cases, at least, the cell produces the proteins it needs to function through a system of subtraction by NMD, not just by addition from the genome. Therefore, mRNA regulation might be as important to the cellular life as the DNA transcription process. "It's probably the most important paper in the alternative splicing field in the last few years," says Bork. "The implications are astounding." Of course, Brenner could be right in a small way while his grand hypothesis could be wrong. RUST could be happening on a much smaller level; a biomolecular oddity, but not a new paradigm in genetic regulation. "The pessimist's view of his theory, which I don't agree with, is that this process is happening, but it's only cellular noise, not a purposeful process," says M.D. Anderson's Wilkinson. "I think Brenner has come up with a great idea to explain some very interesting lab results, but it will take a long time to prove whether or not it's true." Beyond bioinformatics, RUST has to be tested on a gene-by-gene basis, says Wilkinson. So far, such work has shown encouraging results, he adds. ALTERNATIVE ALTERNATIVES Alternative polyadenylation, splicing's kissing cousin, may also play a regulatory role. Prescott Deininger, a geneticist at Tulane University, New Orleans, suggests that a retrotransposon known as LINE1 (long interspersed elements) uses this process to limit its genome-shuffling effects.7 Transcripts that are polyadenylated earlier than normal are stable, but they don't carry the full coding capacity of normal RNA; this may be an important regulatory mechanism as well. Clinton MacDonald, who studies polyadenylation factors in testes at Texas Tech University Health Sciences Center in Lubbock, Texas, says he sees the fields of alternative splicing and alternative polyadenylation converging. "There's more and more evidence that these two processes often work together to produce another layer of regulation." With each layer, the complexity deepens. Other biochemists are attacking the genetic complexity issue from a different perspective. Brenton Graveley, of the University of Connecticut Health Center in Farmington, has taken a one-gene-at-a-time approach, focusing efforts on (among others) the prolific alternative splicer Dscam (see sidebar, at bottom). A host of teams are trying to decipher the link between alternative splicing and cancer. For instance, a team led by Surinder Batra at the University of Nebraska Medical Center, Omaha, studies the MUC4 gene, which is normally expressed in the lungs, trachea, and testes. But when pancreatic cancer develops, the gene splices at a different site and starts running wild in the pancreas.8 "The grand question in such cases," says Wilkinson, "is whether the relationship is causal or correlative. Lots of funny things happen to cells during cancer, including [changes] in alternative splicing." If the alternative splicing is contributing to cancer progression (the Nebraska researchers have yet to show that), a new drug target may be in the works. The only thing that's known with absolute certainty is that certainty is a long way off. Bork suggests that a better understanding will take an assay of every protein, in every tissue type, on a minute-by-minute basis, over an organism's entire lifespan. "It might happen with Drosophila in my lifetime," he says. "But it's not going to happen with humans until a long time later." Sam Jaffe can be reached at sjaffe@the-scientist.com. References 1. J.S. Mattick, "The evolution of controlled multitasked gene networks: The role of introns and other noncoding RNAs in the development of complex organisms," Mol Biol Evol, 18:1611-30, 2001. 2. P.A. Sharp, "Split genes and RNA splicing," Cell, 77:805-15, 1994. 3. D. Brett et al., "Alternative splicing and genome complexity," Nat Genet, 30:29-30, 2002. 4. B. Modrek, C. Lee, "A genomic view of alternative splicing," Nat Genet, 30:13-9, 2002. 5. Z. Kan et al., "Selecting for functional alternative splices in ESTs," Genome Res, 12:1837-45, 2002. 6. B.P. Lewis et al., "Evidence for the widespread coupling of alternative splicing and nonsense-mediated mRNA decay in humans," Proc Natl Acad Sci, 100:189-92, 2003. 7. V. Perepelitsa-Belancio, P. Deininger, "RNA truncation by premature polyadenylation attenuates human mobile element activity," Nat Genet, advanced online publication, Nov. 16, 2003. 8. A. Choudhury et al., "Alternate splicing at the 3'-end of the human pancreatic tumor-associated mucin MUC4 cDNA," Teratogen Carcin Mut, 21:83-96, 2001. THE 38,0000-PROTEIN GENE When Brenton Graveley read about a Drosophila melanogaster gene with almost 100 splicing sites and 38,000 putative protein products, he immediately knew what he would be doing over the next few years, if not the rest of his working career. Graveley, a biochemist by trade, taught himself fly genetics so he could work on the provocative Dscam gene. "That gene was the most amazing thing I'd ever seen," says Graveley, an associate professor at the University of Connecticut Health Center, Farmington. "I started working on an experiment on it that day." ©2001 Geneticcs Society of America FROM ONE, MANY: The Drosophila melanogaster Dscam gene has 115 exons. Four of them (4, 6, 9, and 17) are clusters of multiple alternative exons (with 12, 48, 33, and 2 alternative exons, respectively). Each alternative exon is mutually exclusive; the gene must choose only one alternative exon from each of those four exon clusters. This diagram shows exon 4 and its 12 alternatives. When the alternatives are multiplied, 38,016 distinct protein products result. The secret to Dscam's allure is its amazing prolificacy. Named for its human homolog, the Down Syndrome Cell Adhesion Molecule, Dscam can produce more than 38,000 different proteins through alternative splicing. Moreover, it appears that most of those proteins are actually produced at some point in a fly's life. Little else is known about the gene besides its ability for producing diverse proteins. It appears to promote neuron movement during developmental phases. Homologs have very little functional similarities. Human DSCAM is found on chromosome 21 and its products may be the cause of neurodevelopmental defects associated with the disease from which it gets its name.2 But the human gene has few known alternatively spliced products. Other insects appear to splice and dice the gene. The mosquito, for instance, can produce 32,000 proteins. And, Drosophila virilis may make even more proteins than D. melanogaster, almost 40,000, according to Graveley. With colleague Alicia Celotto, Graveley determined that alternative splicing in Dscam is a regulated process and not just noise.3 The researchers concentrated on exon 4, which has 12 alternative exon variants; this stretch can be edited 12 different ways for inclusion in the finished mRNA. They harvested fruit flies at different times in development and then used reverse transcriptase PCR to determine the variant of exon 4 that was used in each developmental phase. Splicing preferences fluctuated throughout the developmental stages. It turned out that two variants changed the most: the second variant (4.2) and the eighth (4.8). Exon variant 4.2, for instance, was present in only 1% of the RNA emanating from the Dscam gene for samples 0-12 hours old. By the first instar larval stage, it was present in 20%; by adulthood, variant 4.2 accounted for about 40% of the mRNA sequences. "It clearly shows that there is some regulatory mechanism deciding when to choose certain alternative exons," says Graveley. Nevertheless, Graveley's research merely hints at the function of exon 4.2. "We still don't know what all those proteins actually do." References 1. D. Schmucker et al., "Drosophila Dscam is an axon guidance receptor exhibiting extraordinary molecular diversity," Cell, 101:671-84, 2000. 2. G.M. Barlow et al., "DSCAM: an endogenous promoter drives expression in the developing CNS and neural crest," Biochem Biophys Res Commun, 299:1-6, 2002. 3. A.M. Celotto, B.R. Graveley, "Alternative splicing of the Drosophila Dscam pre-mRNA is both temporally and spatially regulated," Genetics, 159:599-608, 2001. function sendData() { document.frm.pathName.value = location.pathname; result = false if (document.frm.score[0].checked) result = true; if (document.frm.score[1].checked) result = true; if (document.frm.score[2].checked) result = true; if (document.frm.score[3].checked) result = true; if (document.frm.score[4].checked) result = true; if (!result) alert("Please select") return result; } .myradio{background-color : #94bee5} Please indicate on a 1 - 5 scale how strongly you would recommend this article to your colleagues? Not recommended 1 2 3 4 5 Highly recommended Please register your vote

Alternative Splicing Goes Mainstream

Interested in reading more?

Become a Member of