Research ArticleAutism

A Noncoding RNA Antisense to Moesin at 5p14.1 in Autism

Science Translational Medicine  04 Apr 2012:
Vol. 4, Issue 128, pp. 128ra40
DOI: 10.1126/scitranslmed.3003479

You are currently viewing the editor's summary.

View Full Text
As a service to the community, AAAS/Science has made this article free with registration.

Making (Anti-)Sense of Autism

Once only considered a potential therapeutic tool, antisense RNAs and their close cousins, the miRNAs, have been elevated in status: We now know them as naturally occurring molecules that—to add more prestige—have bona fide jobs in the cell. Now, one of these RNA regulators has shown up in an unexpected place: right in the middle of a genomic region known to harbor a risk gene for autism spectrum disorders (ASDs). Growing in frequency, ASD is caused by a still-murky interaction between genetic susceptibility and environmental insult. In a search for the genetic risk factors, Kerin et al. zoomed in on one region of the genome, unexpectedly unearthing a noncoding antisense RNA with incriminating characteristics.

A recent study to identify genetic risk factors for ASDs found a significant association with a marker on chromosome 5. The nearest genes, however, were located about a million base pairs away. Kerin et al. hypothesized that a previously undetected functional genetic element might lie near the site of the marker and, using bioinformatics techniques, determined that a noncoding RNA was likely transcribed from that site. The researchers showed that the RNA was indeed expressed; designated MSNP1AS, it is encoded by the antisense strand of a moesin pseudogene. The moesin gene itself, MSN, is X-linked and encodes a protein that plays a role in axon and dendrite development. MSNP1AS RNA directly binds to the complementary MSN transcript in human cells and, when overexpressed, decreases moesin protein levels. Kerin et al. found that MSNP1AS levels were strongly increased in postmortem brain samples from people with ASD compared to those without; furthermore, these higher levels correlated with the presence of the genetic marker for increased ASD risk. Despite significantly increased MSN RNA, moesin protein levels were not increased in postmortem brain samples from people with ASD, suggesting that the noncoding RNA MSNP1AS may play a role in reducing moesin protein. It is possible that overexpression of MSNP1AS at critical periods of brain development may lead to reduced levels of moesin protein, increasing the risk of ASD. Indeed, very low levels of moesin protein are observed in cerebral cortex of fetuses with Down syndrome.

This is a tantalizing beginning. Further work with animal models will be necessary to test whether MSNP1AS expression can down-regulate moesin during brain development and whether such changes indeed contribute to a risk for ASDs. Firming up the case against MSNP1AS could add some notoriety to the reputation of antisense RNAs.