The Scientist

There's a new potential role for microRNAs, according to a report in this week's Nature: modulating dendritic protein building in the hippocampus, a process known to be involved in learning and memory. The findings suggest that microRNAs may help regulate synaptic strength at individual synapses, the authors say. This study is one of the first to examine microRNAs in mature neurons, said Kenneth Kosik of the Neuroscience Research Institute at the University of California, Santa Barbara, who did not participate in the research. Almost all data on microRNAs in the nervous system come from development, Kosik added. MicroRNAs are small, non-coding RNA molecules that exert control over many developmental pathways in animals and plants by preventing the translation of messenger RNA into protein. Previous work has shown that mRNAs can be translated locally at ribosomes found in dendrites, first author Gerhard Schratt of Harvard Medical School and Children's Hospital Boston told The Scientist. Since microRNAs can suppress mRNA translation by binding to complementary sequences, researchers have suggested that they may inhibit translation at synapses, Schratt said, but no data have confirmed this hypothesis. Schratt and his colleagues found that overexpressing a brain-specific microRNA called miR-134 in a culture of mature rat hippocampal neurons caused a significant decrease in the size of dendritic spines, the neuron's major sites of excitatory synaptic contact. Expression of an antisense inhibitor of miR-134, on the other hand, caused an increase in dendritic spine size. To see how miR-134 might prevent the growth of these dendritic spines, the researchers searched for possible target mRNAs of miR-134. Among 48 genes suspected to be involved in dendritic protein synthesis, they found three that contained sequences partially complementary to miR-134. They chose to focus on one called Lim-domain-containing protein kinase 1 (Limk1), whose protein product is involved in building dendritic spines. Limk1 knockout mice show abnormalities in dendritic spine structure that are "almost a phenocopy of miR-134 overexpression," Schratt said. The authors found several pieces of evidence suggesting that miR-134 represses dendritic spine synthesis by binding Limk1 mRNA: miR-134 and Limk1 interact in vitro and co-localize within dendrites, and miR-134 overexpression in neurons reduces translation of Limk1. These interactions disappear if Limk1 is engineered to lack the sequence complementary to miR-134. To see if miR-134 and Limk1 interact locally at dendrites, Schratt and colleagues imaged fluorescence-tagged Limk1. They found that, in dendrites, wild-type Limk1's expression levels are 18 to 28% lower than mutant Limk1, which cannot interact with miR-134. Since miR-134 does not block translation of Limk1 completely, however, other microRNAs or proteins must also help repress Limk1, Schratt said. When the researchers replaced wild-type Limk1 with the mutant form, miR-134's effect on spine size was rescued, demonstrating that miR-134 regulates dendritic spine size through the wild-type Limk1, the researchers note. Finally, they found that treatment with brain-derived neurotrophic factor (BDNF) can release some of miR-134's repression of Limk1 ? suggesting that normal neuronal activity could contribute to Limk1 translation, and thus spine synthesis and memory formation, Schratt said. The study is "exciting but it leaves a lot of questions open," said Eric Miska, of the Gurdon Institute at the University of Cambridge, who was not a co-author of the study. Overexpression studies using short, non-coding RNA can suffer from off-target effects, Miska said, so it's hard to pin down exactly how a microRNA is interacting with a specific mRNA. "There may be up to a thousand targets for a given microRNA," he said. "We don't know which are the real targets of this microRNA in vivo." Conversely, Limk1 mRNA has binding sites for several microRNAs besides miR-134, said James Eberwine of the University of Pennsylvania Medical Center, also not a co-author. "Various cellular functions could be attributed to the other microRNAs." Also, the paper doesn't prove that miR-134 is acting in dendritic spines, Miska told The Scientist. "They show localization, but there's no functional data that suggests it has to be in the spine," he said. "It would really be nice to know whether it's translation in the dendrite or translation in the cell soma -- or both -- that's being modulated by the microRNA," Eberwine agreed. Either way, miR-134 interactions with Limk1 could be affecting dendrite outgrowth, he said, but follow-up experiments could isolate dendrites from the rest of the neuron to prove that miR-134 suppresses translation of Limk1 specifically in dendrites. "I think it's going to be very difficult to prove what a microRNA does in vivo," Miska said. Although still unavailable, mouse microRNA knockouts will likely be "the gold standard to prove the function of a particular microRNA" in mammals, he said. mlp@nasw.org Links within this article G.M. Schratt et al., "A brain-specific microRNA regulates dendritic spine development," Nature 439:283-289, 19 Jan 2006. http://www.nature.com/nature Kenneth Kosik http://www.lifesci.ucsb.edu/mcdb/faculty/kosik/ I.Ganguli, "Linking microRNAs to hepatitis C," The Scientist, September 2, 2005. http://www.the-scientist.com/article/display/22759/ D. Steinberg, "MicroRNA target practice," The Scientist, June 20, 2005. http://www.the-scientist.com/article/display/15538/ J. Eberwine et al., "Local translation of classes of mRNAs that are targeted to neuronal dendrites," PNAS 98:7080-7085, 19 Jun 2001. PMID: 11416191 K.S. Kosik, A.M. Krichevsky. "The Elegance of the MicroRNAs: A neuronal perspective," Neuron, September 15, 2005. PMID: 16157272 Y. Meng et al., "Abnormal spine morphology and enhanced LTP in LIMK-1 knockout mice," Neuron 35:121-33, 3 Jul 2002. PMID: 12123613 Eric Miska http://www.gurdon.cam.ac.uk/~miskalab/ Jim Eberwine http://www.med.upenn.edu/ins/faculty/eberwine.htm