Mini-ReviewEnhancing non-coding RNA information content with ADAR editing
Introduction
Cells of the mammalian nervous system receive, process, and distribute vast amounts of information, with high fidelity, sensitivity, and resolution. Synaptic regions, in particular, appear to challenge the classic thermodynamic limits of electrophysiological information flow with the sheer density of their information content. Yet such nearly infinite computing power is orchestrated through the actions of very finite numbers of proteins deployed into the molecular machineries of cognition. To reach beyond these limits while respecting the laws of physics and information theory, the nervous system takes advantage of evolutionary innovation in the transcriptome.
Traditional views for articulating the form and functionality of the nervous system adhere to the primacy of protein, as defined by the Central Dogma of Molecular Biology [17]. However, recent discoveries revealing tens of thousands of non-coding (nc) RNAs and transcripts of unknown function in metazoan genomes [14], [23], [37], together with their extensive and specific expression in defined brain regions [59], [69], articulate a new substrate for nervous system information coding and processing. This emerging ncRNA landscape provides a computational matrix able to increase information densities within the spatially and thermodynamically limited real estate of the brain [82].
The diverse portfolio of non-coding transcripts expressed throughout the brain could hardly avoid the adenosine de-amination acting on RNA (ADAR) editing system, considering that most ncRNAs contain secondary structure as mature products, or proceed through double-stranded (ds) RNA intermediates. Mammalian genomes code for three ADAR enzymes, ADAR1, ADAR2, and ADAR3, whose catalytic activity de-aminates adenosine residues located in certain double-stranded regions of RNA to produce an inosine, which is recognized by protein machineries as a guanosine. Recent reviews present the many aspects of ADAR biochemistry, genetics, and biology that now serve as the basis to study the impact of this intriguing activity on cellular and organismal behavior [34], [54], [58]. Nervous system tissues prominently express ADARs in all metazoans, resulting in a repertoire of alterations in mRNAs coding for important proteins in neuronal transmission and synaptic plasticity [57], [29]. In order to achieve optimal levels of function, key components of electrical and chemical signaling machineries may depend on the precise modulation of ADAR editing activity in response to cell identity, fluctuations in the micro-region of synapses, or changes in the environment of the neuron. For example, the Drosophila Shaker potassium channel exhibits striking tissue specific expression of differentially edited isoforms, epistatic differences in biophysical channel phenotypes, and coupling between editing and the expression distributions of a number of the isoforms [33]. ADARs recognize minute differences in secondary structures among similar RNAs [73], permitting the elaboration of a codable language of signaling and recognition between the ADARs and the transcriptome.
A partial understanding of how ADAR chooses its targets from millions of RNA structures in the transcriptome has emerged, using comparative genomics [29], bioinformatics [16], and sequencing of ADAR targets in the transcriptome [50]. An examination of the wider non-coding transcriptome's secondary structure landscape reveals a profile similar to that of the coding transcriptome, with an abundance of the same RNA structural motifs that promote binding and de-amination by ADAR enzymes. One class of RNAs with a particularly favorable structural motif for ADAR, ALU containing RNAs, has already been shown to contain thousands of editing sites, distributed fairly equally among coding and non-coding members [4], [39], [47]. Thus, potential ADAR targets abound by the thousands in the non-coding transcriptomes of nervous system cell types, leading to a map expansion for the significance of ADAR action in the brain.
These developments place ADAR in the path of interaction with thousands of previously unknown members of the neuronal transcriptome, potentially opening new dimensions of nervous system information flow. We argue that this map expansion will include a widespread role for the ADAR–ncRNA interface in the acquisition, computational processing, and distribution of information to macromolecular networks throughout the nervous system. This review explores these concepts, including the mechanisms of interactions between brain enriched ADARs and the ncRNA transcriptome, and how these interactions could serve as a computational matrix to enhance information processing by the organism.
Section snippets
The shapes and contours of an ADAR interactive transcriptome
The results of the Human Genome project [43], [88] disappointed the neuroscience community because the surprisingly low estimates of the number of human coding genes could not explain the robustly expanding complexity of mammalian nervous systems compared to so-called “lower” organisms with similar gene numbers, but far simpler brains. Sequencing data indicated that less than 2% of the human genome coded for mRNAs, yielding a total number of proteins uncomfortably close to that of flies or even
ADAR's machinery of transcriptome recognition drives information processing
Within a cell's proteome, hundreds or even thousands of distinct RNA binding regions compete for RNA epitopes as they form during transcription, processing, or transport. The behavior of each family of RNA binding domains depends on the modulation of surrounding conditions, including a constant influx of changes from the environment, signaling events such as protein phosphorylation, and statistical distributions of interactions at the RNA–protein interface [20]. Overlapping affinity profiles of
ADAR's impact on the nervous system non-coding transcriptome
Perhaps no other tissue matches the diversity, heterogeneity, and complexity of the non-coding transcriptome expressed in the nervous system. Adjacent cells in local micro-regions, with otherwise identical morphological features can express highly divergent levels of ncRNAs. While precise characterization awaits deep sequencing at the single cell level, Nelson et al. have proposed redefining nervous system cell types using cell specific transcriptome signatures [61]. Non-coding transcriptomes
The ADAR information processing cascade: coupling environmental stimuli and stress to ncRNA editing levels and downstream information signaling
Interesting recent work by Barkai and colleagues provides evidence that stress response in yeast responds primarily to evolutionarily coded predictions made from information contained in extra-cellular signaling cascades, rather than direct measurement of conditions within the cell. In wild type yeast, modest increases in temperature result in a more permissive environment, with reduced stress signaling and curtailed stress responsive gene expression. On the other hand, conditional lethal
Conclusion
Even though the non-coding transcriptome remains sparsely surveyed, especially in the nervous system, existing examples provide a preview of the broad scope of ncRNA function. Discovery of ADAR editing events outside the well established mRNA targets associated with the synaptic proteome suggests that ADAR edits many ncRNAs involved in nervous system information processing pathways. Together with the high levels of ADAR expression localized to brain regions, these facts suggest that ADAR
Acknowledgements
The authors would like to thank Prakriti Mudvari for editorial assistance, and Mark Mazaitis for the production of the graphics.
References (93)
- et al.
Mitochondrial criticality: a new concept at the turning point of life or death
Biochim. Biophys. Acta
(2006) - et al.
Relief of microRNA-mediated translational repression in human cells subjected to stress
Cell
(2006) - et al.
Cellular stress and RNA splicing
Trends Biochem. Sci.
(2009) - et al.
Transcriptional scaffolds for heterochromatin assembly
Cell
(2009) - et al.
Temperature-controlled structural alterations of an RNA thermometer
J. Biol. Chem.
(2003) - et al.
Structure and sequence determinants required for the RNA editing of ADAR2 substrates
J. Biol. Chem.
(2004) - et al.
Retention and repression: fates of hyperedited RNAs in the nucleus
Curr. Opin. Cell Biol.
(2005) - et al.
Long-range Ca2+ signaling from growth cone to soma mediates reversal of neuronal migration induced by slit-2
Cell
(2007) - et al.
Drosophila DDP1, a multi-KH-domain protein, contributes to centromeric silencing and chromosome segregation
Curr. Biol.
(2004) - et al.
RNA editing in regulating gene expression in the brain
Biochim. Biophys. Acta
(2008)