As the field of microRNA research expands, it is becoming increasingly clear that these small RNAs play a significant role in cellular regulation. But according to a letter to the editor in the Jan. 14 issue of Cell, that role may be bigger than some might have thought.
According to the findings of David Bartel, a researcher at the Whitehead Institute, MIT scientist Christopher Burge, and Benjamin Lewis, a graduate student in both researchers’ labs, miRNAs may be responsible for regulating up to one-third of all human genes.
To reach this conclusion, the researchers used a computer algorithm they had designed, called TargetScan, to examine mRNAs from the human genome for sequences that would make them likely targets for miRNAs. They then looked to see if these sequences could be found in the corresponding genomes of the mouse, rat, dog, and chicken.
“What we find is actually quite a striking preservation of sequences corresponding to the 5’ end of the microRNAs, which we’ve called the seed,” Lewis told RNAi News this week.
In the Cell paper, Bartel and his colleagues state that they began with the “UTRs corresponding to mRNAs annotated in the UC Santa Cruz Genome Browser database, and a set of 62 unique seed matches that represented 148 human miRNA genes and defined the families of known miRNAs conserved in the” genomes of the five species.
“We identified 14,301 instances of conserved seed matches within the 3’ UTRs, thereby predicting 14,301 unique target sites,” the researchers wrote. “Because some UTRs had multiple conserved target sites for the same miRNA seed, this analysis implicated 12,893 unique miRNA-target regulatory relationships.”
Given the almost 13,000 regulatory relationships, more than 5300 human genes — 30 percent of the researchers gene set — were implicated as miRNA targets.
“One-third of our gene set … have significant conservation of sites corresponding to microRNAs in the context of neutral variation,” Lewis said. “It’s not that we’re just finding matches, it’s not just that we’re able to pair the mRNA to the microRNA or model that matching — we’re doing statistics on the probability of observing these matches relative to a background distribution that’s modeled with these control sequences. That’s been a really essential part of the work,” he said.
The Cell paper comes a little more than a year after Bartel, Burge, Lewis, and two other MIT researchers reported that with TargetScan, they had predicted more than 400 targets of mammalian miRNAs by examining the human, mouse, rat, and puffer fish genomes.
That research, which also appeared in Cell, found that the conserved 5’ region of mammalian miRNAs is key for target identification.
Lewis said that following their most recent Cell paper, he, Bartel, and Burge intend to continue refining their algorithm.
“In addition to identifying all these genes and targets, we’ve also been able to refine the rules for target recognition by a microRNA,” he said, noting that “it appears that there’s a significant conservation of an A sequence just proximal to a microRNA, which probably is recognized by the RISC complex.
“I plan to continue to refine the predictions, as well as our model of microRNA/mRNA recognition, and then also begin to examine the genes and the pathways regulated by microRNAs,” Lewis said. “I haven’t even really had a chance to explore that whole aspect of it.”
He noted that he conducted analyses of a few microRNAs, but “it was very difficult to say, ‘Here’s the set of genes we have and these genes are functionally coherent’ … because we only had a very small subset of genes that were probably being regulated by the microRNA. But now that we have a more comprehensive picture, stories are starting to form and that’s pretty exciting,” he said.