Introduction
At first glance, the nervous systems of vertebrates and invertebrates seem bilaterally symmetrical, but on closer inspection left–right asymmetries become apparent. Humans, for example, show gross anatomical differences between right and left temporal lobes, and visual and language faculties are asymmetrically distributed between the two hemispheres. How these asymmetries arise during development remains something of a mystery (for review, see ref 1). In the nematode Caenorhabditis elegans, the AWC and ASE chemosensory neuron pairs are bilaterally symmetrical based on anatomical considerations, but nevertheless display asymmetrical gene expression patterns. A recent study in Nature by Johnston and Hobert identifies a microRNA (miRNA) as a crucial mediator of this asymmetry in the ASE neurons2.
Of the 302 neurons in the C. elegans hermaphrodite, 32 neurons, including 14 bilateral pairs, comprise the chemosensory system. The ASE pair mediates responses to water-soluble compounds3. Despite expressing a set of genes in common, having similar morphologies and connecting to identical sets of postsynaptic partners, the left and right ASE neurons (ASEL and ASER) are not identical. ASEL alone expresses the transmembrane guanylyl receptor cyclase genes gcy-6 and gcy-7 and responds to sodium, whereas ASER expresses gcy-5 and primarily responds to chloride and potassium4, 5. C. elegans apparently maximizes the functionality of its limited repertoire of chemosensory neurons by further lateral differentiation.
Earlier this year, the Hobert laboratory showed that the ASEL/R asymmetry is mediated via competition between two protein complexes present in both ASEL and ASER neurons6. One complex contains the OTX-like homeodomain protein CEH-36 and the transcriptional coactivator LIN-49, and the other contains the Groucho-like transcriptional corepressor UNC-37 and the NKX6-like homeodomain protein COG-1. The outcome of this competition is biased by the expression of COG-1, which is present at far higher levels in ASER than in ASEL neurons6. Thus, the COG-1/UNC-37 complex outcompetes the CEH-36/LIN-49 complex in ASER neurons to repress ASEL-specific genes, thereby allowing the expression of ASER-specific genes. In contrast, low levels of COG-1 in ASEL allow the CEH-36/LIN-49 complex to activate the expression of ASEL-specific genes. Of course, this result immediately raised the question of how COG-1 expression is regulated in ASEL and ASER neurons.
Enter the lsy-6 miRNA. The miRNAs are small (
21 to 22-nucleotide) RNAs, which are processed from larger RNAs with extensive secondary structure (see ref. 7 for review). These small RNAs downregulate gene expression by base-pairing with cognate complementary sequences in the 3' untranslated regulatory regions (UTR) of target mRNAs, thereby preventing their translation. The first miRNA was identified over a decade ago in C. elegans8, 9, but only with the recent identification of conserved miRNAs in other multicellular organisms have miRNAs evolved from being a nematode-specific oddity to representing the newest general mechanism of gene regulation. Genome mining has predicted 100–300 miRNAs in different species, and several of these are expressed in a tantalizing tissue- and developmental stage–specific manner7. However, to date, the functions of only a handful of these miRNAs have been defined, and few targets have been identified in vivo.
Johnston and Hobert identified lsy-6 in their screen for mutants affecting ASEL/R asymmetry. In lsy-6 mutants, ASEL neurons adopt ASER-like properties, whereas shared ASE-like characteristics and ASER identity are unaffected. The authors could rescue the lsy-6 phenotype with a fragment that did not encode a protein product but instead encoded an RNA with a predicted hairpin structure. Within this hairpin lay a 23-nucleotide sequence that was highly conserved in the related nematode C. briggsae and likely was an miRNA. The authors used several approaches to confirm that it was indeed this miRNA that represented the lsy-6 locus. Most convincingly, they showed that a mutation that disrupted the hairpin structure also disrupted rescue of the lsy-6 phenotype, and moreover, that a compensatory mutation that restored the secondary structure also restored the rescuing ability of the gene. The authors found that lsy-6 was expressed in ASEL but not ASER neurons, and that misexpression of lsy-6 in ASER neurons was sufficient to induce ASEL characteristics.
How does lsy-6 act? In lsy-6 mutant animals, ASEL neurons expressed COG-1 at levels similar to those of ASER neurons. Interestingly, the 3' UTR of the cog-1 mRNA contained sequences that were complementary to lsy-6. Proof that cog-1 was indeed the target of lsy-6 was provided by demonstrating that expression of a 'sensor' containing cog-1 3' UTR sequences fused to the coding sequences of GFP was brought under lsy-6 control, and that mutations in the lsy-6 complementary site abolished this regulation. This result suggests that levels of COG-1 are regulated by lsy-6 post-transcriptionally in ASEL neurons (Fig. 1). The authors previously also observed regulation of COG-1 at the level of transcription6; however, this may occur simply because COG-1 autoregulates to maintain its expression (O. Hobert, personal communication). Thus, the lsy-6 miRNA may act to 'asymmetrize' initially symmetric cog-1 expression, leading to different subtype identities in ASEL and ASER neurons.
Figure 1: How lsy-6-mediated asymmetry determines ASEL/R subtype identities.
The ASEL and ASER chemosensory neurons acquire distinct subtype identities as a consequence of an antagonistic interaction between the COG-1/UNC-37 and CEH-36/LIN-49 protein complexes. This competition occurs through asymmetric expression of the NKX6-like homeodomain transcription factor COG-1. Higher levels of COG-1 in ASER neurons results in repression of ASEL-specific genes and expression of ASER-specific genes, whereas lower levels of COG-1 in ASEL neurons allows the CEH-36/LIN-49 complex to promote the expression of ASEL-specific genes. COG-1 levels are downregulated specifically in ASEL neurons by ASEL-expressed lsy-6 miRNA. This miRNA binds to a complementary site in the 3' UTR of the cog-1 mRNA to prevent translation (inset).
Full size image (43 KB)This satisfying story raises several intriguing issues. The lsy-6 miRNA was not identified in any of the extensive cloning or bioinformatics sweeps to identify all miRNAs predicted to be encoded by the C. elegans genome10, 11, 12. The lsy-6 sequences are conserved in C. briggsae, which is separated from C. elegans by 50–100 million years of evolution, although the authors do not report whether lsy-6 is asymmetrically expressed and shows conserved function in ASE neurons of C. briggsae. It is also unclear whether lsy-6-related sequences are found in other animals. However, the identification of lsy-6 raises the real possibility that there may be many more miRNAs yet to be identified and characterized.
This study yet again highlights the power of forward genetics in elucidating gene, or in this case miRNA, function. One issue with studying miRNA function has been that because miRNAs are so small, generating mutations via targeted reverse genetics approaches has proved to be technically challenging. Although it is possible that further iterations of bioinformatics software might yet identify lsy-6, given its fairly esoteric role, it is highly unlikely that its functions would have been easily identified via reverse genetics. In addition, in animals (as opposed to plants), miRNAs bind to target sequences that are not perfectly complementary)7, making the identification of target genes in silico extremely difficult. Individual miRNAs appear to regulate multiple target genes, suggesting that miRNA-mediated post-transcriptional regulation of gene expression may become the norm instead of the exception.
The identification of lsy-6-mediated regulation of L/R asymmetric gene expression is a significant advance, but the signal that underlies asymmetric expression of lsy-6 itself has yet to be discovered. When and where does this signal act? Are all cases of C. elegans nervous system asymmetry mediated via the same signal? Do similar signals act in other organisms? Genetic screens using the asymmetrically expressed lsy-6 as a marker may answer some of these questions.
Finally, this study defines the first miRNA function in the nervous system of any organism. The let-7 miRNA downregulates the hunchback-related gene hbl-1 in the nervous system in C. elegans, but the functional consequences of this regulation are unclear13, 14. Might miRNAs have different or unexpected functions in the nervous system as opposed to other tissues? Probably not. However, given that nervous system function critically depends on thousands of neuronal subtypes that must be generated in a strict temporally and spatially regulated manner, it is likely that the problem of regulating gene expression in the nervous system is especially complex. In Arabidopsis, most miRNA targets are believed to be transcription factors involved in pattern formation and cell differentiation15. Thus miRNAs are excellent candidates to determine cellular diversity in the nervous system. The exploration of miRNA functions is in its infancy. The Johnston and Hobert paper2 provides a first taste of the many surprises that miRNAs have in store for us in the future.
