New & Noteworthy

Measure Twice, Cut Once

April 23, 2014

If this carpenter is worth his salt, he will take a tip from Rnt1p and measure twice before cutting. Image from Wikimedia Commons

As any seasoned carpenter knows, if you are going to cut a piece of wood, you want to do it right the first time!  There is no second chance.

This means that good carpenters are very, very careful.  They use a clamp to hold the wood in place and measure where to cut not once but twice.  Now they have a good shot at getting a length of lumber they can use.

As shown in a new paper in Molecular Cell by Liang and coworkers, it turns out that our cells do something similar when cutting their RNA.  The yeast enzyme Rnt1p measures a piece of double stranded RNA twice to make sure it cuts in the right place.  And, like a second-rate carpenter, if it measures the RNA only once it often cuts the RNA in the wrong place.   

This is almost certainly not just a yeast thing.  Rnt1p is a member of the conserved RNAse III family, which is present in all domains of life except Archaea.

In higher organisms, RNAse III enzymes such as Dicer produce the small interfering RNAs (siRNA) and microRNAs (miRNA) that have important roles in gene regulation via RNA interference. S. cerevisiae doesn’t use RNA interference, but Rnt1p is still important for maturation of small nuclear RNAs, small nucleolar RNAs, and ribosomal RNA, and also for degradation of some specific mRNAs.

Most RNase III enzymes recognize the RNA they are to cut by certain secondary structures like loops.  Liang and coworkers used X-ray crystallography on Rnt1p in complex with an RNA substrate to learn how Rnt1p recognizes its substrate and “knows” where to cut it. The RNA had a double-stranded stem capped by a 4-nucleotide loop, a so-called tetraloop, that had a conserved G residue at the 2nd position.

Rnt1p cleaves this RNA at a fixed distance from the tetraloop, and it cleaves the two strands unequally so that they have 2-nucleotide 3’ overhanging ends. The crystal structure showed that two of the five RNA-binding motifs (RBMs) in Rnt1p form a pocket that clamps down on the conserved G residue in the tetraloop. This clamp is fastened so tightly that the RNA structure is changed.  It is like the clamp distorting the carpenter’s piece of wood.

When Liang and coworkers deleted one of these two Rnt1p RBMs, or mutated the conserved G in the substrate, the substrate was no longer held or cleaved.  Clamping the RNA was critically important for the reaction. 

They also showed both by structural modeling and by mutational analysis that other parts of Rnt1p interact with the RNA stem structure. Clamping the RNA and interaction with the rest of the substrate puts the cleavage site at a fixed position relative to the Rnt1p active site.

This tight binding and measurement by protein-RNA interactions would seem to be good enough to ensure accurate cleavage. But it’s not the whole story.

Another domain of Rnt1p, the N-terminal domain (NTD), was known to contribute to substrate selection, but it was unclear exactly how it did this. Surprisingly, the crystal structure showed that it, too, contacts the tetraloop. When Liang and colleagues deleted the NTD, the RNA substrate was still cleaved but there was a mixture of products, cleaved at several different sites. So it too is needed for precise cleavage.

The overall conclusion is that two different domains contact the tetraloop, each acting like a ruler. The protein-protein and protein-RNA interactions stiffen each ruler such that the cleavage site is always precisely measured before cutting.  Just like our carpenter friend, to get the right cut, Rnt1p needs to measure twice before cutting. The same knowledge that is handed down through generations of carpenters is also deeply ingrained in our biochemistry! 

by Maria Costanzo, Ph.D., Senior Biocurator, SGD