Environmental Factor, August 2011, National Institute of Environmental Health Sciences
Designing a human RNA-binding domain outlined in JBC paper of the week
By Robin Arnette
Hall is head of the Laboratory of Structural Biology Macromolecular Structure Group. A 2D model of her human PUF RNA-binding domain appears on the monitor behind her in the photo. According to the JBC website, only the top one percent of papers submitted to the journal earn the honor of being selected as a paper of the week. (Photo courtesy of Steve McCaw)
NIEHS principal investigator Traci Hall, Ph.D.,(http://www.niehs.nih.gov/research/atniehs/labs/lsb/ms/index.cfm) is interested in how a family of proteins, called PUF RNA-binding domains, recognizes RNA targets. The name PUF was derived from the fruit fly Drosophila Pumilio protein and roundworm Caenorhabditis elegans fem-3 mRNA binding factor (FBF), and hints at the proteins' universality and importance to living systems.
Starting 10 years ago, Hall's group with Phillip Zamore, Ph.D., at the University of Massachusetts Medical School, determined the first crystal structures of a human PUF protein, quickly followed by structures using RNA targets from the fly version, since human target RNAs had not been identified. Earlier this year, they published work with details on how it recognizes targets in human sequences and in the June 8 online(http://www.ncbi.nlm.nih.gov/pubmed/21653694) issue of the Journal of Biological Chemistry (JBC), they published an article that is the culmination of studies to design specific RNA-binding domains. The article has also been selected as the paper of the week and is featured in the July 29 hardcopy issue of the journal.
Hall said the reason why her lab studies RNA-binding domains is that these proteins are important factors in gene regulation and can help researchers understand how the environment impacts human health. She said the findings in her latest study, which were made possible by her use of leading-edge three-dimensional imaging in the NIEHS Laboratory of Structural Biology, may also have clinical applications in the fight against cancer (see text box).
“A key strategy is identifying the pathways that allow our bodies to respond at the molecular level,” Hall said. “Without this basic knowledge, it's difficult to determine how the environment interferes with these pathways and how we might combat negative consequences.”
PUF the magic shape sorter
Hall said an illustration of how these proteins work would be to think about it as a sophisticated shape sorter. Imagine the PUF RNA-binding domain possesses a line of connected pockets shaped like circles, squares, and triangles, and the RNAs that bind to it are chains of these shapes. If the pockets on the contraption match the shapes on the RNA, then it has a proper fit and the RNA grabs on. If one shape doesn't match, then the contraption can ignore it and look for another RNA that does match.
However, Hall added, if the PUF RNA-binding domain came across a hexagon, it wouldn't have a pocket that specifically fit that shape. The hexagon could fit in a big pocket that would accept any shape, but the match would be less than optimal.
Hall explained what happens in the real world of cells this way. “RNA has four types of bases - adenine (A), uracil (U), guanine (G) and cytosine (C). In the PUF protein's natural code, we could identify code sequences for, or sets of side chains that could recognize, A, U, and G, but nature didn't show us a code for C. We weren't completely thwarted by that, because one of the binding pockets could tolerate any base, and we could use this pocket for a single C and still design a specific RNA-binding protein. But, if we wanted to recognize an RNA sequence with more than one C, we were out of luck, and any protein that was designed to recognize a sequence containing one C would recognize not just one, but four related RNA sequences.”
With this thought in mind, Hall's group successfully collaborated with the lab headed by Zefeng Wang, Ph.D.,(http://www.med.unc.edu/pharm/people/primaryfaculty/zefeng-wang/wang) at the University of North Carolina at Chapel Hill, to use yeast to select a code for C, giving the ability to design a PUF domain that can recognize any RNA sequence. Now, researchers can design RNA-binding domains to recognize over 64,000 different sequences instead of roughly 9,000. To demonstrate the utility, they attached splicing machinery to a new RNA-binding domain using the C code to make artificial splicing factors.
Citation: Dong S, Wang Y, Cassidy-Amstutz C, Lu G, Bigler R, Jezyk MR, Li C, Hall TM, Wang Z.(http://www.ncbi.nlm.nih.gov/pubmed/21653694) 2011. Specific and modular binding code for cytosine recognition in Pumilio/FBF (PUF) RNA-binding domains. J Biol Chem 286(30):26732-26742.
Splicing in the fight against cancer
Hall said the application of their findings could lead to viable cancer treatment options. Changing the RNA splice form changes the protein that is produced and can change its function. Scientists could use artificial splicing factors in cancer cells to force them toward apoptosis or cell death. In theory, if the splicing factor were expressed in tumor cells from a patient who is taking anti-cancer drugs, the malignant cells would experience a higher rate of cell death.
An artificial splicing factor could also be used to eliminate the lifeline of cancer cells. Large tumors need a blood supply to survive, so along with a regimen of pharmaceutical compounds that inhibits angiogenesis, or the ability to build the blood vessels, a splicing factor could prevent the formation of the proteins needed in the process.
According to Hall, designing a specific RNA-binding protein was tricky business before the discovery of PUF proteins' RNA-binding properties. She was fortunate because nature made this particular family of RNA-binding proteins very amenable for single-stranded RNA sequence recognition. The first crystal structures revealed most of the code. With the help of yeast in the current study, the code is solved.
“In the past, we tried using computers to predict what side chains would be able to recognize a C,” Hall continued. “It was exciting to see the code on paper, but the crystal structure let us see how it actually worked in 3D.”