About 8 years ago, the fragile X syndrome gene FMR1 was shown to encode an RNA-binding protein, but its physiological function has remained elusive. The most popular theory is that FMR1 regulates translation because it is associated with polysomal RNA. However, until now, there has been little information on the RNA targets that are bound by FMR1. Two papers in Cell report complementary studies aimed at finding such targets, and both converge on a small set of transcripts — encoded by genes involved in several aspects of neuronal function — that represent very strong candidates for regulation by FMR1.

The study by Victoria Brown, Peng Jin and colleagues involved two microarray experiments. In the first, mRNA isolated from mouse brain was compared with mRNA that had been immunoprecipitated with an FMR1 antibody, the aim being to find mRNA that was bound to FMR1. Of the >25,000 genes screened, 432 were identified that were enriched by immunoprecipitation. The second study compared polysomal RNA from fragile X with control human cell lines. The authors reasoned that if FMR1 regulates translation by binding to polysomal RNA, the absence of FMR1 in the fragile X cell lines should lead to differences in the abundance of specific polysome-associated RNAs. In this experiment, >35,000 genes were compared, and 251 showed substantial differences in abundance between the polysomal RNA fractions of the fragile X and control cell lines. A partial comparison of the sets of genes revealed 14 that were identified in both experiments, and could therefore be targets for regulation by FMR1.

The second paper reports a biochemical approach to finding the optimal RNA target sequence for FMR1. After several rounds of selection from a pool of random RNA sequences, Jennifer Darnell et al. identified an RNA molecule that bound to FMR1 with high affinity. They then defined the minimal sequence required for binding, and concluded that FMR1 requires a specific RNA structure called a G quartet, which is present in only a small percentage of transcripts. They tested 12 such transcripts for binding to FMR1, and found six that bound. So, there is more to find out about the requirements for FMR1 binding, but the presence of a G quartet is a good pointer.

Most satisfyingly, 8 of the 14 RNAs identified in the first study contained sequence that would be expected to form a G quartet, which reinforces the conclusions of both papers. Attention can now shift towards the genes that have been identified in these studies — and they are a tantalizing collection. Several genes are implicated in synaptic function or other aspects of neuronal biology. The study of these genes will help to pin down whether the key function of FMR1 is to regulate the translation of specific RNAs. Looking beyond fragile X syndrome, this twin-track approach is likely to be valuable for identifying the targets of other RNA-binding proteins.