RNA targeting and translation in axons

See allHide authors and affiliations

Science  23 Mar 2018:
Vol. 359, Issue 6382, pp. 1331-1332
DOI: 10.1126/science.aat1498

Neurons are among the largest and most complex cells in nature, often extending very long axons, which in adult mammals, including humans, can reach up to one meter in length. These extraordinary morphological features pose a challenging problem as to how information codified in the nucleus can reach the periphery of the cell in a timely manner to respond to extrinsic stimuli. Similar to virtually all eukaryotic cells, neurons have adopted the strategy of localizing RNA asymmetrically. The nature of the transcripts targeted to dendrites and axons have been extensively studied, and they encode synaptic proteins, cytoskeleton components, ion channels, mitochondrial and ribosomal proteins, and proteins required for plasma membrane biogenesis. However, the mechanism underlying local translation has remained elusive. On page 1416 of this issue, Terenzio et al. (1) add a new piece to the puzzle and show that local translation to produce the protein mammalian target of rapamycin (mTOR) precedes the burst of protein synthesis associated with the regeneration of injured axons. mTOR is a serine/threonine kinase that plays a central role in regulating protein synthesis (2).

Peripheral localization of transcripts is a widespread phenomenon that mediates many cellular processes. In neurons, coding and noncoding RNAs are targeted to dendrites and axons, where messenger RNAs (mRNAs) are rapidly translated in response to extrinsic stimuli. Local protein synthesis has been shown to mediate synaptic development and plasticity in dendrites, whereas in axons, it is necessary for axon extension and steering in response to guidance cues (3). Although polyribosomes were visualized at the base of dendritic spines more than 30 years ago, the presence of the translational machinery in axons has been hotly debated. This was mostly because in axons, ribosomes are found close to the plasma membrane, which makes visualization by using classical microscopy techniques difficult. Because of their localization, it has even been proposed that axons may “borrow” ribosomes from surrounding cells, such as Schwann cells that produce the myelin sheath that coats axons (4).

Comparative analyses of RNAs localized in either dendrites, axons, or cell bodies showed expression patterns that only partially overlap and differ depending on cell type and developmental stage (57). Interestingly, transcripts that are highly expressed in cell bodies are not necessarily enriched in axons or dendrites, indicating that RNAs do not reach the peripheral compartments by passive transport but are sorted and delivered with an active mechanism. How are transcripts selected to be transported to axons? At least two mechanisms must be taken into account, one intrinsic to the RNA and dependent on its structure and a second related to the extrinsic signals that trigger transcript localization. Although the information necessary for RNA transport can be stored anywhere along the transcript, most elements that regulate mRNA targeting are found within the 3′ untranslated regions (UTRs). The first localization element of a neuronal transcript was identified in the 3′UTR of the mRNA encoding β-actin and was named “zipcode” because it was necessary for delivering the mRNA to axons in response to neurotrophins (8). A number of localization elements have since been found in the 3∼UTRs of transported transcripts (5, 9). Although the ribonucleotide sequences of localization elements described so far show little resemblance, it is possible that the folding of the RNA may form common secondary structures.

RNA transport and translation in regenerating axons

In developing neurons, mRNA localization and translation mediate axon growth. Although adult neurons are less dependent on this mechanism to maintain nerve homeostasis, after injury many transcripts, including mTOR, are transported to axons and translated at the lesion site. Local translation of mTOR increases protein synthesis in a positive loop that supports nerve growth and regeneration.


The main aim of asymmetric distribution of mRNA is to compartmentalize signaling. For example, the guidance cue netrin-1 induces localized synthesis of β-actin in growth cones, which mediates the steering of retinal axons toward the guidance cue (10). Interestingly, RNA localization and local protein synthesis is regulated with a high degree of signal specification. In developing sensory neurons, for example, distinct transcripts are targeted to axons in response to different neurotrophins (11). A potential mechanism entails that each extrinsic signal activates specific RNA-binding proteins (RBPs) that act as a hub to recruit and transport different sets of transcripts. This has been demonstrated for splicing factor glutamine-rich (SFPQ), an RBP that regulates the transport of functionally related transcripts in response to neurotrophins (12).

A prototypical example of the advantages brought about by signal compartmentalization in neurons is provided by the regenerative response that follows nerve injury. Compared with developing neurons, adult neurons have fewer ribosomes in axons, and lower levels of local protein synthesis are required for their maintenance. However, a sudden change of circumstances, such as traumatic injury, has a profound impact on gene expression and dramatically increases RNA localization to axons (9), ensuring that regenerating axons receive a constant supply of newly synthesized proteins. The initial response to axon damage requires the activation of the signaling protein extracellular signal–regulated kinase (ERK) and an increase of intracellular calcium at the site of the lesion. These events are necessary for membrane resealing, and although short-lived, they can also influence gene expression. After this acute response, Terenzio et al. found that a sustained retrograde propagation of the injury signal to the nucleus induces the targeting of newly synthesized transcripts, including those encoding mTOR to the lesion site of the axons. The combination of increased transport of mRNAs and high local translation of mTOR results in the synthesis of proteins necessary for nerve regeneration (see the figure).

A number of fundamental questions remain unanswered. It is still unknown how RNAs are sorted in the nucleus and “tagged” for transport to dendrites and axons. A potential mechanism may involve small noncoding RNAs and/or the UTRs of targeted transcripts that in combination with specific RBPs, such as SFPQ, could act as a scaffold to tag mRNA for transport. Indeed, the 39UTR of the transcript interacts with the encoded cell surface protein CD47, driving localization to the plasma membrane (13). Genome-wide analyses of 3′UTR expression, such as poly(A)-seq and 3′end-seq, will help to obtain a comprehensive picture of the features shared among 3′UTRs of peripherally localized transcripts.

Noncoding RNAs, including microRNAs, are also present in both dendrites and axons. It is not known whether their role is confined to regulating mRNA stability and translation of localized transcripts or whether they represent the missing link between extrinsic signals applied at the periphery of the cell and nuclear functions. mRNAs encoding transcription factors are known to be transported and translated in both axons and dendrites (6, 14, 15). However, their biological importance is uncertain, mostly because it is difficult to understand how they could elicit a transcriptional response that is distinct from the one induced by the much larger fraction of the same transcription factors residing in the nucleus. It is possible that noncoding RNAs, and perhaps UTRs, form a tag that determines the binding of axon-derived transcription factors to specific promoters in a manner akin to many enhancer RNAs, which interact with transcription factors and RNA polymerase II on the promoters of regulated genes.

The mechanistic links between extrinsic signals and translational activation in developing and regenerating axons are still largely unknown. Although Terenzio et al. made the important discovery that local translation of mTOR regulates protein synthesis in response to injury in adult axons, whether this is a mechanism shared with other neurons and at different developmental stages remains unclear. It should also be noted that the incorrect processing and delivering of mRNA has been linked to the pathogenesis of many human neurological disorders, to the point that it has been proposed that most, if not all, neurodegenerative diseases fundamentally are disorders of RNA metabolism (15). Further understanding of the basic mechanisms underlying mRNA localization in dendrites and axons will lay the foundations for developing new therapeutic approaches for many neural disorders.


View Abstract

Navigate This Article