Recognizing RNA

Courtesy of Phillip Zamore, University of Massachusetts Medical School, and Science.

Researcher:

Phillip Zamore, Gretchen Stone Cook Professor of Biomedical Sciences, University of Massachusetts Medical School, Worcester

Project:

Looking for endogenous small-interfering RNAs in flies, changes in which could influence gene expression.

Problem:

How do you distinguish siRNAs from microRNAs, which are roughly the same size?

Solution:

Zamore used both Solexa and 454 technologies to sequence short, noncoding RNAs from 5,000 fly heads. How do you isolate fly heads? Freeze the flies in liquid nitrogen and sift them with a flour sifter. To remove microRNAs from the equation, he took advantage of a chemical difference between the two molecules.

MicroRNAs typically associate with Argonaute-1, while siRNAs associate with Argonaute-2. Ago-2 recruits a methyltransferase, which adds a methyl group to the 2'-OH on the 3'-end of the siRNA. Ago1 doesn't do this, so Zamore and his team used sodium periodate to oxidize unmethylated RNAs (microRNAs) while leaving siRNAs intact. Then it was a simple matter to sequence what was left (Science, 320:1077-81, 2008).

"We basically destroyed the microRNAs with this oxidation step, so they wouldn't show up," Zamore explains. "This dramatically increased our sensitivity to find the siRNAs, and it worked." His team discovered a collection of RNAs almost exclusively 21-nucleotides long, as expected for Dicer products. "It was the sharpest length distribution I have ever seen for a small RNA. This is like a skyscraper on a histogram - 80-something percent were 21 nucleotides."

Cost:

Though prices vary, Zamore says each Solexa sequencing run typically costs about $2,000; 454 runs cost $16,000.