By now, there's no point explaining why researcher study small RNAs — their importance has been proven time and again, and they've opened the doors to understanding a range of biological mechanisms that have long stumped scientists. But years after the discovery of small interfering RNAs and microRNAs, researchers continue to stumble across what appear to be entirely new flavors of small RNAs. For many of them, getting a clear grasp of function continues to elude scientists.
To some extent, that may be because RNA research is playing catch-up to DNA research. Peter Stadler, a scientist at the University of Leipzig, says the impetus for his team to launch a project to annotate noncoding RNA across the Trichoplax adhaerens genome was the organism's recent sequencing. "If you look in the genome paper that was published," Stadler says, "the word 'RNA' doesn't show up at all. It's completely out of the scope of the annotation that had been done." That's certainly not specific to the Trichoplax genome, and Stadler hopes to see more effort put into understanding small RNAs across whole genomes.
Among researchers surveyed for this article, there was a clear consensus on the major breakthrough for small RNA research — and that's next-generation sequencing platforms. "It was really difficult to analyze small RNAs before," says Gunter Meister, a group leader in RNA biology at the Max Planck Institute. Ultra-deep sequencing, he says, has enabled scientists to see beyond the abundant small RNAs to the relatively rare ones.
At the bench
Sequencing and a number of other technologies have been deployed to investigate small RNAs, with a primary goal being to understand the function these genetic snippets play. Since the central element for small RNAs is binding to an argonaute protein, some researchers like Meister are using large-scale immunoprecipitation to look at the full range of small RNAs that may be involved.
Molly Hammell, a postdoc in Victor Ambros' lab at the University of Massachusetts who collaborates with scientists performing similar types of immunoprecipitation experiments, says this type of work "started in the past year or two" with the technological advances that allowed researchers to study RISC proteins. "I think that's going to be tremendously helpful," she says.
Hammell is also working with knockout organisms to evaluate function. It was in the Ambros lab that an early project knocked out many of the microRNAs in C. elegans and resulted in very little phenotypic effect, leading to the theory that miRNAs might work in tandem to accomplish their silencing. Hammell will be starting work in flies to see what can be found in that system.
At Harvard, Neil Kubica is a postdoc in Winston Kuo's lab, where he also performs gain- and loss-of-function experiments. Using model mice, Kubica works to derive precursor embryonic cells and then "try to overexpress a panel of microRNAs … to see if we can push the fate of these explanted stem cells" in a particular direction, he says. Kubica also studies cell signaling in cancer, and to that end performs screens of miRNAs to "find upstream activators or repressors of the pathway," he adds. "We think there are going to be a lot of interactions between microRNAs and cell signaling."
Le Kang, a professor at the Institute of Zoology at the Chinese Academy of Sciences, recently published the results of a study looking at small RNAs in locust. His team used sequencing to do the heavy lifting for the project, which was designed to determine whether these RNAs play a role in phase changes of locust, Kang says. He adds that his team was especially looking for miRNAs, piRNAs, and endo-siRNAs in the study, which also turned up an unknown class of longer small RNAs — 26 nucleotides to 29 nucleotides — for which his group will perform follow-up characterization studies.
Of course, biological experiments often come from, and lead to, computational studies. After finding, say, an miRNA, says Kubica, "you're left with a bioinformatics and experimental challenge [of] knowing what the difference is between a predicted target and a target that's actually functional."
Small RNA research has seen a boom in target prediction and other tools recently as the field takes off. A significant challenge, says Leipzig's Stadler, is that in RNAs "the sequence changes fairly rapidly," making them that much harder to pin down in discovery or annotation efforts. In his work with the Trichoplax genome, he says, it became clear that "you need specialized annotation tools for these housekeeping noncoding RNAs because you really do not find them with standard Blast searches." His team's work lent computational evidence to what had already been suggested experimentally: the absence of miRNAs in the Trichoplax genome.
Hammell at the UMass med school worked on a project that resulted in mirWIP — the name comes from "miRNA targets by weighting immunoprecipitation-enriched parameters" — an algorithm to "improve microRNA target predictions," she says. One of the pieces of the project was looking at RISC-IP transcripts to check the binding strength in particular miRNA regions. "We think the better they can bind, the more of an effect they'll have on their target," Hammell says.
Another new tool comes from Yasubumi Sakakibara at Keio University, where a team recently introduced base-pairing profile local alignment, or BPLA, kernels for analysis of functional noncoding RNA sequences. The Sakakibara group tested the method on C. elegans in a search for small nucleolar RNAs, which Sakakibara says are harder to find because they have fewer structural features than other small RNAs. The project turned out 48 candidate snoRNAs, 14 of which were new — a major finding, Sakakibara notes, given that "only about 100 snoRNAs have been discovered even in the well-studied [organisms]." Sakakibara says his team will continue with this work. "Our real goal is to establish a framework for comprehensive finding of novel, functional RNAs hidden in large genomes of various organisms," he says.
One of the great things about small RNA research is the frequency of novel findings. Michael Levine, a professor at the University of California, Berkeley, was just as surprised as anybody when a project from his lab turned up a new miRNA-related pattern in Ciona intestinalis. A study of miRNAs in the organism demonstrated that half of the small RNAs in question were connected to moRs, or miRNA-offset RNAs. These sequences of 18 nucleotides to 20 nucleotides "arise immediately adjacent to micro-RNA sequence," Levine says.
While moRs have been seen before, they've been so rare that "they were dismissed as incompletely processed RNAs or spurious products," Levine says. "In Ciona it's really clear: they're quite abundant." At this point, the function of these moRs is a mystery, though Levine speculates that they could indicate that moRs are more universal than previously thought, that they occur in many other organisms but that most have "some kind of degradation pathway to eliminate the moRs … but in Ciona that pathway is either impaired or absent."
In Meister's lab at Max Planck, recent findings suggest a role for snoRNAs, which are not well understood. Though they were identified years ago, their function for the most part remains elusive. Meister found hundreds of snoRNAs in the human brain, and posits that they get cleaved into smaller RNAs that in turn act like miRNAs. A genomic region commonly deleted in patients with Prader-Willi syndrome, for instance, has no genes — but does have two snoRNAs, Meister says.
Future research will require better tools, the scientists are quick to point out. Most important: "We don't really have screening systems," says Stadler. "We have no systems to look at the large number of these molecules from any kind of phenotypic or functional perspective … and to a large extent lack techniques in the lab to manipulate these molecules in a way that's cost-effective."