Rockefeller University researchers identify protein that regulates RNA in nerve tissue

Members of the Laboratory of Neuro-Oncology, headed by Robert Darnell (pictured at right), and one of the Laboratories of Molecular Biophysics, headed by Stephen K. Burley, used cell cultures, animal models and X-ray crystallography to piece together the essential details of the Nova protein's function and structure. Pictured in the photo at left are: Ru Zhong, Giovanni Stefani, Kirk Jensen (Darnell Lab) and Kiran Musunuru (Darnell and Burley Labs). Co-authors not pictured: Kate Dredge, Ron Buckanovich, James Okano and Yolanda Yang (Darnell Lab) and Hal Lewis, Carmen Edo and Hua Chen (Burley Lab). Photos by Linne Ha (left) and Robert Reichert (right).

Protein may play a key role in nervous system function

Rockefeller University researchers have identified a protein that is responsible for regulating RNA splicing in nerve cells, a process essential for the development and operation of complex nervous systems, such as those found in mammals, including humans. The protein, called Nova-1, is the first splicing factor specific to one kind of tissue to be found in mammals.

The discovery that Nova-1 functions only in neurons suggests that the protein plays a key role in nervous system function. The research also sheds light on the expression of genes through a phenomenon known as RNA “alternative splicing,” in which a single gene can produce more than one protein.

In addition to clarifying Nova-1’s biological role, the scientists have used X-ray crystallography to solve the structure of an essential segment of the protein bound to the RNA it regulates. Knowing the structure gives researchers a powerful way to predict other RNAs that are regulated by Nova. The combined research, conducted in two laboratories at the university, is reported in separate papers in the Feb. 4 issue of Cell and the Feb. 24 issue of Neuron.

“This work gives us new insights into how RNA splicing is regulated in all mammals, including humans,” says Robert Darnell, M.D., Ph.D., an associate professor and head of the Laboratory of Molecular Neuro-Oncology, which conducted most of the work. “Until quite recently, it was thought that studying this kind of regulation in a mammalian system lay years in the future because the problem was too complex. But studying the disease antigen Nova has allowed us to link in vitro biochemistry and in vivo biology. We’ve been able to approach the question from two directions.”

The scientists first became interested in Nova-1 because the protein plays a key role in a rare disease called paraneoplastic opsoclonus myoclonus ataxia (POMA). Those suffering from POMA are unable to inhibit movement and suffer uncontrollable shaking. POMA is classified as a paraneoplastic neurological disorder (PND), diseases the Darnell lab studies. PNDs develop when cancer cells in the body prompt a tumor immune response that makes its way across the blood-brain barrier and disrupts the normal function of brain cells.

The brain is known as an “immune-privileged” site, meaning that proteins expressed only in the brain are not screened by the immune system as it goes through the process of learning which proteins are “self” and which are “foreign.” When a brain protein is expressed in a tumor elsewhere in the body, the immune system sees it as a foreign protein and mounts a strong response against it. This immune response, while good for eliminating the tumor, sometimes makes its way into the brain where it can attack those neurons that express the protein. In the case of POMA, the protein is Nova-1.

The exact nature by which the immune system attacks the brain is unclear, but POMA patients have high levels of antibodies against Nova-1 in their spinal fluid. These same antibodies, it turns out, can bind a segment of Nova-1 called the KH domain and inhibit the domain’s interaction with RNA. Researchers suspect that POMA patients’ profound motor dysfunction is caused at least in part by a direct inhibition of RNA binding by Nova-1 antibodies.

The Darnell lab found that the Nova-1 protein regulates RNA splicing in the nerve tissue of mammals. When the protein is present and performing its role, it directs alternative splicing of RNA, in which certain segments of pre-mRNA (top) are kept in the mRNA, while other segments are discarded during the transfer. Such cutting and stitching allows for translation to a wide diversity of proteins with very specific properties. If Nova-1 is absent, however, RNA splicing is deregulated, and the resulting protein's structure is changed. Darnell's lab hypothesizes that this altered receptor function directly contributes to nerve cell death. In a mouse model in which the gene encoding production of Nova-1 is deleted, the animal exhibits signs of severe neuronal dysfunction including debilitating shaking and weakness. These symptoms are present in patients with POMA. Diagram by Kirk Jensen.

Nova’s function

Darnell’s laboratory set out to understand Nova-1’s function, and they focused on the protein’s three KH domains, the parts thought to bind RNA. Using a technology known as SELEX, Darnell and his colleagues were able to determine that Nova prefers to bind RNA stretches that contain repeats of the nucleic acid sequence UCAY. (Nucleic acids are the building blocks of DNA and RNA.) Upon learning this, the scientists searched RNA databases for repeats of UCAY and zeroed in on a sequence found in a molecule called the inhibitory glycine receptor, which contains several UCAY repeats. The researchers hypothesized that Nova-1 was acting through this sequence in the glycine receptor to control alternative splicing.

Alternative splicing is a phenomenon that enables cells to produce a wide variety of proteins from a finite number of genes. This capacity is essential for the development and operation of complex nervous systems such as those found in mammals. The initial transcript of any gene, known as pre-mRNA, is pieced together to produce a mature mRNA that can code for a protein. In alternative splicing, different pieces of this pre-mRNA are stitched together to produce different mRNAs, and thus different proteins.

To test their theory that Nova-1 regulates glycine receptor alternative splicing, Kirk Jensen, first author and postdoctoral fellow in the lab, and his collaborators carried out two lines of studies–one in cell cultures, the other in animals. In cell cultures, they demonstrated that Nova-1 could control the alternative splicing of a glycine receptor “mini-gene.” Also, Nova-1 would not bind with the glycine receptor if the UCAY sequence had been slightly altered. This demonstrated that Nova regulated RNA specifically through UCAY sequences.

In the animal study, the scientists generated and studied a special breed of mice, known as “knockout mice,” that lack the gene responsible for encoding Nova-1. These mice look normal at birth but do not grow as much as their littermates and, like humans with POMA, suffer from debilitating shaking and weakness. When the researchers looked at the pattern of splicing in the glycine receptor, they saw it was different from the splicing pattern in a normal mouse–another sign that Nova acts as a splicing regulator.

Nova’s KH domain structure

In addition to deducing Nova’s function, Rockefeller researchers also wanted to find out exactly how Nova and its targets are put together. Darnell’s laboratory collaborated with one of the Laboratories of Molecular Biophysics, headed by Rockefeller Professor and Howard Hughes Medical Investigator Stephen K. Burley, M.D., D. Phil.,to determine the co-crystal structure of Nova and the RNA that it targets to regulate alternative splicing.

“We had some pretty solid evidence that Nova regulates alternative splicing of the glycine receptor by binding the UCAY repeats,” says Kiran Musunuru, a biomedical fellow in the Darnell and Burley laboratories. “So we wanted to see the molecular mechanism by which the Nova KH domain interacted with UCAY and understand why Nova recognizes that sequence instead of others.”

What they found was surprising. Most proteins that bind with RNA do so along one of two surface areas on the protein, the alpha helix or the beta sheet, because they offer favorable conditions for connection. The KH domain of Nova, however, uses a different method. Every KH domain contains two loops: an “invariant loop” that is the same in all domains, and another, “variant loop” with differing lengths. The Nova KH domain uses these two loops as a molecular vise that holds the RNA strand in place on a platform between them. The platform is made up of amino acids projecting from two helices and one of the beta strands. Since the three amino acids lack affinity for water, the researchers have dubbed it the “hydrophobic platform.”

Such a structure, which binds using two loops and a hydrophobic platform rather the alpha helix or beta sheet, has never before been reported. “For those of us familiar with protein structures, this is quite a novel way for a protein to hold an RNA strand in place,” says first author Hal Lewis.

The interdisciplinary research environment at Rockefeller allowed scientists from two different labs to piece together the protein’s vital information, says Burley, the Richard M. and Isabel P. Furlaud Professor, who also serves as Deputy for Academic Affairs. “This research has been unusual in that it’s extended into a lot of different scientific domains,” he says, “With two labs working different angles, we were able to get a more complete picture of how Nova normally works and what is happening when it doesn’t.”

In addition to Darnell and Jensen, co-authors of the Neuron paper are graduate fellow Kate Dredge, Giovanni Stefani, Ru Zhong, Ron Buckanovich, James Okano and Yolanda Yang. The research reported in Neuron was supported in part by the National Institutes of Health, an Irma T. Hirschl Career Scientist Award, the Breast Cancer Research Program and the Ataxia Telangectasia Children’s Project.

In addition to Burley, Darnell, Musunuru, Lewis and Jensen, co-authors of the Cell paper are Carmen Edo and Hua Chen. The research reported in Cell was supported in part by The Rockefeller University, the National Institutes of Health and the Breast Cancer Research Program.