MicroRNAs Have Hand in Shaping Synapse

Emerge as Regulators of Protein Translation in Brain’s Dendritic Spines

As if biology weren’t complex enough, the discovery of microRNAs has, over the past few years, introduced a new twist into how cells work. The function of these small, noncoding bits of RNA is still enigmatic, but already they are emerging as a hidden hand in events in both plant and animal cells. To explain the administration of a cell, scientists have always looked to a few basic processes—transcribing genes into RNA, splicing RNA in different ways, translating RNA into proteins, modifying proteins to change their function—all of which seemed to account for even the most precise molecular feats. But an explosion of studies in the past few years has pointed to microRNAs as important managers in the cell, overseeing the pattern of protein expression by preventing RNA translation.

Photo by Rachel Eastwood

MicroRNA may help control synaptic plasticity by preventing protein translation until it is needed at synapses, according to research by (left to right) Michael Greenberg, Christina Kane, and Gerhard Schratt.

Michael Greenberg’s lab, for example, has spent years piecing together a program of gene transcription activated in neurons in response to cell activity. A longstanding mystery in the neuroscience field is how gene transcription in the nucleus, which is a global activity, translates into local changes at particular synapses. Many researchers have thought that new proteins must be made at the synapses themselves to control their development and ongoing adjustment. Yet in a study in the Jan. 19 Nature, research fellow Gerhard Schratt in Greenberg’s lab offers evidence that a specific microRNA helps to regulate the translation of a protein at synapses involved in forming dendritic spines, the small peninsulas at the cell surface where synapses form. It suggests that microRNA may help control synaptic development and plasticity by silencing messages until they are needed.

Tracking Translation
For several years, researchers have studied messenger RNAs that travel to the synapse from the cell’s interior, but it has been difficult to prove that this cargo is used to make proteins only when needed at the synapse. Greenberg, HMS professor of neurology at Children’s Hospital Boston, and his colleagues reasoned that if they could first identify messenger RNAs that are translated in response to cellwide stimulation, they could then determine if these RNAs are present at synapses. In a paper previously published in the Journal of Neuroscience, his team, led by Schratt and former graduate student Elizabeth Nigh, stimulated neurons with a growth factor, called brain-derived neurotrophic factor (BDNF), and identified a subset of messenger RNAs that are translated in response to it. They then offered evidence that at least some of the translation occurs at synapses.

If synapses contain silos of messenger RNAs that are released when the cell is stimulated, what is keeping them at bay? Greenberg said that the most obvious idea was something binding to the 3' untranslated region (UTR), a cap on one end of messenger RNAs that does not encode for a protein. “Three-prime UTR regions are involved in regulating translation as well as stability of the message,” said Greenberg.

Image courtesy of Michael Greenberg

In rat neurons, increasing levels of the microRNA miR-134 causes dendritic spines to shrink (b), while dampening its activity with antisense miR-134 causes spines to grow (c), compared with controls (a). The arrows point to small spines (b) and enlarged spines (c).

It is known that proteins can bind to this region to regulate messenger RNA, but Schratt looked at microRNA as well. “MicroRNAs in non-neuronal cells were emerging as a mechanism for suppressing translation,” Greenberg explained. “So on that basis, it’s a reasonable hypothesis to say local translation might involve microRNAs that suppress translation in a regulated way.”

The team focused on one micro-RNA, miR-134, that is detectable only in the brain and is expressed during development in rats around the time that synapses form. When the researchers overexpressed miR-134 in neurons, the dendritic spines shrank. When they interfered with miR-134 function using an antisense approach, the size of spines increased. “The size is thought to correlate with the strength of the synapse,” Greenberg said. It seemed that the microRNA was binding to and suppressing something that made spines grow.

One of the messenger RNAs with a binding site for miR-134 encodes a protein called Limk1 that is known to control cell structure by regulating the behavior of the cytoskeletal protein actin. Actin molecules can mill around the cell as loners or they can gang up to form filaments that give the cell shape; Limk1 promotes the filament state. Mice that lack Limk1 have abnormal spines that look similar to those overexpressing miR-134, and loss of Limk1 in humans is associated with Williams syndrome, a genetic disorder that causes cognitive impairment.

The team put together several pieces of evidence to show that miR-134 interacts with and suppresses Limk1. When they introduced miR-134 into cells, expression of Limk1 was blocked, but it was unaffected when cells were exposed to other microRNAs. By fusing the Limk1 3' UTR to a firefly luciferase gene that allows for detection and then measuring its expression in the cell, the team found that expression levels were higher if the piece of Limk1 was mutated so that it could not bind to miR-134. And using this mutant form of Limk1, the researchers were able to prevent the effect of miR-134 on dendritic spine size. Stimulating cells with BDNF could overcome the dampening effects of miR-134 on Limk1, showing that the microRNA’s inhibition can be overcome when synapses are stimulated.

To see if the interaction happens at synapses, the team used a fluorescent reporter of protein synthesis that was engineered to stick to the inner surface of the cell membrane and break down quickly to prevent diffusion. By analyzing the fluorescent signal with confocal microscopy, the researchers found that Limk1 expression was reduced along the length of the dendrites compared to the mutant form of Limk1 that could not bind miR-134.

One Word: Plasticity
Kenneth Kosik, the Harriman professor of neuroscience research at the University of California, Santa Barbara, said that “it’s a very lovely paper” addressing a major question in the field of synaptic plasticity: to what extent is plasticity regulated by local translation of RNA into proteins? Though this paper does not completely prove this elusive fact, “it strengthens the case for a highly regulated system of local translation in dendrites.” The findings suggest that microRNAs may help the cell keep messages silent as they travel from the nucleus to the synapses.

Complicating this picture is that microRNAs do not always perfectly match their targets in sequence and that multiple microRNAs may bind to a single messenger RNA and vice versa. “It’s an entire level of regulation that had not been noticed before,” Kosik said, “but has the complexity of these other layers we already knew about.”

—Courtney Humphries

top