SEARCH

SEARCH BY CITATION

Keywords:

  • local translation;
  • axon;
  • nerve injury;
  • regeneration;
  • retrograde transport

ABSTRACT

  1. Top of page
  2. ABSTRACT
  3. SUBCELLULAR mRNA LOCALIZATION IS A FUNDAMENTAL CHARACTERISTIC OF EUKARYOTIC CELLS
  4. IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN DENDRITES
  5. IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN AXONS BY TRANSGENIC OVEREXPRESSION
  6. IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN AXONS BY SUBCELLULAR KNOCKOUT
  7. SUMMARY AND IMPLICATIONS
  8. REFERENCES

Intracellular trafficking and localization of mRNA is a fundamental feature of living cells, suggesting that localized mRNA translation should enable subcellular regulation of the proteome. Such localized regulation may be of particular importance in highly polarized cells such as neurons, where the requirement for a specific protein can be at a site far distant from the nucleus. Although dendritic and synaptic protein syntheses are well-established phenomena, the apparent paucity of ribosomes in early studies on mature vertebrate axons generated significant skepticism regarding the possibility of protein synthesis within axons. Here, we summarize recent findings in genetically engineered mouse models that support a role for local translation in axonal expression of β-actin and importin β1 in injured adult sensory neurons in vivo. These definitive confirmations of mammalian axonal protein synthesis in both transgenic and subcellular knockout models should direct further attention to the diverse roles suggested for local protein synthesis in axonal physiology. © 2013 Wiley Periodicals, Inc. Develop Neurobiol 74: 210–217, 2014


SUBCELLULAR mRNA LOCALIZATION IS A FUNDAMENTAL CHARACTERISTIC OF EUKARYOTIC CELLS

  1. Top of page
  2. ABSTRACT
  3. SUBCELLULAR mRNA LOCALIZATION IS A FUNDAMENTAL CHARACTERISTIC OF EUKARYOTIC CELLS
  4. IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN DENDRITES
  5. IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN AXONS BY TRANSGENIC OVEREXPRESSION
  6. IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN AXONS BY SUBCELLULAR KNOCKOUT
  7. SUMMARY AND IMPLICATIONS
  8. REFERENCES

Cells adjust their local proteome (protein abundance and complexity) with high spatiotemporal precision in response to metabolic needs or extracellular signals. Subcellular localization of mRNA is now recognized as a widespread phenomenon in both prokaryotic and eukaryotic cells (Donnelly et al., 2010; Keiler, 2011), which suggests that local mRNA translation may have a general role in subcellular proteome specialization. Local protein synthesis may provide a number of advantages over the transport of pre-existing proteins from one part of the cell to another. First, translationally silenced mRNAs can be stored locally and translated to multiple protein copies only upon need, a mechanism that might be more energy efficient for specific protein usage. Second, the ectopic presence of proteins in nondesired parts of the cell during protein transport is avoided. Third, mRNAs can be targeted to different subcellular localizations with the help of localization motifs in their untranslated regions (UTRs) without changing the structure and function of the proteins they encode. Finally, properties that are unique to newly made proteins (such as minimal post-translational modification) may provide an additional layer of signaling information. Local translation of trafficked mRNAs may allow spatial or temporal compartmentalization of cellular responses to specific stimuli, or rapid responses to environmental or developmental signals (Andreassi and Riccio, 2009; Jung et al., 2012).

Localized mRNA regulation may be of particular importance in highly polarized cells such as neurons. For example, proteins that function at the axon terminus of an adult human sensory or motor neuron must be synthesized locally or transported over a meter of intracellular distance from the cell body. However, early studies detected little or no ribosomes in mature vertebrate axons, in striking contrast to the levels observed in cell body and dendrites, leading to the prevailing assumption that axons cannot synthesize proteins (reviewed in Twiss and Fainzilber, 2009). On the other hand, dendritic protein synthesis is well accepted by the neuroscience community and is thought to play key roles in synaptic plasticity and cognitive processes (Cajigas et al., 2010; Swanger and Bassell, 2011; Kindler and Kreienkamp, 2012). Nonetheless, a series of publications in recent years have presented evidence in support of local protein synthesis in axons of the peripheral nervous system (Donnelly et al., 2010). Many of these studies have proposed a role for local translation in axon regeneration and repair (reviewed in Gumy et al., 2010; Rishal and Fainzilber, 2010). Moreover, recent genome-wide analyses have suggested that the axonal transcriptome may be much larger than previously thought, encompassing literally thousands of different transcripts (Willis et al., 2005; Taylor et al., 2009; Zivraj et al., 2010; Gumy et al., 2011; Cajigas et al., 2012; Deglincerti and Jaffrey, 2012), thus raising the likelihood of widespread roles for local protein synthesis in axonal functions.

IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN DENDRITES

  1. Top of page
  2. ABSTRACT
  3. SUBCELLULAR mRNA LOCALIZATION IS A FUNDAMENTAL CHARACTERISTIC OF EUKARYOTIC CELLS
  4. IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN DENDRITES
  5. IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN AXONS BY TRANSGENIC OVEREXPRESSION
  6. IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN AXONS BY SUBCELLULAR KNOCKOUT
  7. SUMMARY AND IMPLICATIONS
  8. REFERENCES

Conclusive evidence in support of an in vivo physiological role for local translation ideally requires some way of selectively perturbing translation in a subcellular compartment of interest while not affecting production of the same gene product elsewhere in the cell. The identification of specific localization motifs in the UTR regions of certain mRNAs (Andreassi and Riccio, 2009) has enabled genetic approaches to this issue. A pioneering study by Mayford and colleagues on a CAM kinase II isoform in dendrites was the first to demonstrate such an approach (Miller et al., 2002). Targeted mutagenesis was used to replace part of the CaMKIIα 3′UTR with a segment from the 3′UTR of bovine growth hormone (BGH). The BGH transcript contains a distinct 3′UTR that is not dendritically localized, but has the same polyadenylation motif as CaMKIIα. The resulting knock-in mouse retained the complete protein-coding region of the gene, but lacked any dendritic localization signal in the mRNA. Indeed, CaMKIIα mRNA was restricted to the soma of hippocampal neurons in the mutant mice, together with a marked reduction of CaMKIIα protein in postsynaptic densities (Miller et al., 2002). The mice further exhibited a reduction in long-term potentiation (LTP), and impairments in spatial memory, associative fear conditioning, and object recognition memory (Miller et al., 2002). These results demonstrated that local translation of CaMKIIα is required for synaptic and behavioral plasticity. In a later study, Xu and colleagues reported dendritic localization differences between long and short 3′UTR variants for brain-derived neurotrophic factor (BDNF) (An et al., 2008). They further noted that the two isoforms arise from differential usage of two alternative polyadenylation sites, and took advantage of an allele generated by insertion of a lacZ cassette into the BDNF locus after the first polyadenylation site. This genetic manipulation in effect generated a selective knockout of the long 3′UTR BDNF isoform (An et al., 2008). They then showed that dendritic targeting of BDNF mRNA is impaired in this mutant mouse, with concomitant reduction of BDNF protein in hippocampal dendrites. The mutant mice reveal deficits in pruning and enlargement of dendritic spines, as well as selective impairment in LTP and in metabolic energy balance (An et al., 2008; Liao et al., 2012). Although it should be noted that BDNF localization is likely controlled by multiple motifs (Baj et al., 2011) and that the effect of the UTR deletion may also partially reflect selective stabilization of the long 3′UTR (Allen et al., 2013), these results clearly show that local translation of BDNF in dendrites is critical for some of its functions.

IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN AXONS BY TRANSGENIC OVEREXPRESSION

  1. Top of page
  2. ABSTRACT
  3. SUBCELLULAR mRNA LOCALIZATION IS A FUNDAMENTAL CHARACTERISTIC OF EUKARYOTIC CELLS
  4. IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN DENDRITES
  5. IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN AXONS BY TRANSGENIC OVEREXPRESSION
  6. IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN AXONS BY SUBCELLULAR KNOCKOUT
  7. SUMMARY AND IMPLICATIONS
  8. REFERENCES

The rigorous evidence summarized above for local protein synthesis in dendrites contrasted strongly with the prevailing assumption that axons lack meaningful levels of translational machinery (Twiss and Fainzilber, 2009). Thus, despite increasing in vitro evidence from culture systems (Donnelly et al., 2010), considerable skepticism remained regarding the in vivo relevance of local protein synthesis in axons. Twiss and colleagues took up the challenge to address this issue in a transgenic mouse model for axonal localization of β-actin (Donnelly et al., 2011; Willis et al., 2011). The β-actin zipcode sequence is one of the most characterized cis-acting elements for axonal targeting of an mRNA (Patel et al., 2012). This element, a conserved bipartite sequence present at the beginning of the β-actin 3′UTR, interacts with the RNA-binding protein ZBP-1 to mediate axonal localization of β-actin (Welshhans and Bassell, 2011). Willis et al. (2011) linked a diffusion-limited GFP reporter with the 3′UTRs of β-actin or of the nonlocalizing γ-actin, and expressed the reporter constructs under control of an injury-activated neuronal promoter in transgenic mice. The GFP-3′UTR-β-actin mice revealed axonal localization of reporter mRNA in adult sensory neurons, and moreover, crush lesion of the nerve enhanced accumulation of the reporter mRNA in injured axons proximal to the lesion site, whereas in contrast GFP-3′-UTR-γ-actin mice did not show any axonal expression (Willis et al., 2011). Donnelly et al. (2011) then assessed endogenous β-actin transcript levels in the transgenic axons and showed that overexpression of GFP-3′UTR-β-actin, but not GFP-3′UTR-γ-actin, in adult sensory neurons caused a reduction in endogenous β-actin levels. The transgenic GFP-3′UTR-β-actin neurons had reduced axon regeneration after nerve injury, and this deficit could be rescued by ZBP1 transfection. A similar regeneration deficit was observed in ZBP+/− heterozygous mice, and the phenotype could be rescued by expression of exogenous ZBP1 (Donnelly et al., 2011). Thus, ZBP1 is required for optimal axon outgrowth following injury in adult sensory neurons, and increases in ZBP1 levels can enhance nerve regeneration. The transgenic β-actin 3′UTR was further found to compete in vivo with other ZBP1 cargo mRNAs such as GAP-43 (Yoo et al., 2013), indicating that the regeneration deficits in the transgenic animals were likely due to reduced axonal localization of other ZBP1 cargos, in addition to β-actin (Fig. 1). In further studies, Twiss and colleagues went on to show that axonal translation of β-actin mRNA primarily supports axon branching, while axonal translation of GAP-43 mRNA supports elongating growth in sensory neurons (Donnelly et al., 2013). Taken together, this series of publications provides strong evidence for local protein synthesis in adult sensory axons of transgenic mice.

image

Figure 1. Transgenic overexpression of an axon-localizing mRNA affects axonal regeneration (Reproduced with permission from Perry and Fainzilber, EMBO J 2011, 30, 4520–4522, based on the data of Donnelly et al., EMBO J 2011, 30, 4665–4677). A: ZBP1 is required for the axonal localization of β-actin, GAP-43, and likely additional mRNAs. B: Introduction of an exogenous mRNA that competes with endogenous transcripts for binding to ZBP1 reduces the supply of endogenous mRNAs to the axon and attenuates axon regeneration after injury. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN AXONS BY SUBCELLULAR KNOCKOUT

  1. Top of page
  2. ABSTRACT
  3. SUBCELLULAR mRNA LOCALIZATION IS A FUNDAMENTAL CHARACTERISTIC OF EUKARYOTIC CELLS
  4. IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN DENDRITES
  5. IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN AXONS BY TRANSGENIC OVEREXPRESSION
  6. IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN AXONS BY SUBCELLULAR KNOCKOUT
  7. SUMMARY AND IMPLICATIONS
  8. REFERENCES

In addition to the transgenic models described above, in vivo evidence for local translation in axons was recently achieved by a subcellular knockout of importin β1, a critical facilitator of both nucleocytoplasmic and axonal transport. In the classical nuclear import pathway, an importin α binds the nuclear localization signal (NLS) within the cargo protein directly. Its affinity for NLS is increased by interaction with an importin β (of which importin β1 is the main representative in mammalian cells) which then facilitates transport of the complex through the nuclear pore (Perry and Fainzilber, 2009). Importins also play a role in retrograde signaling from axon to soma after nerve injury, by mediating binding of injury signaling proteins to dynein motors (Hanz et al., 2003). Local translation of importin β1 was proposed to be a critical initiating step for this mechanism, by switching affinity of the importin α/dynein complex for signaling cargos. Moreover, a number of additional components of the complex were also described to arise from local translation, including an accessory cargo adapter (Perlson et al., 2005), a regulator of importin binding properties (Yudin et al., 2008), and the transcription factor STAT3, a retrograde cargo of the complex (Ben-Yaakov et al., 2012).

Importin β1 is an essential gene, as shown by very early embryonic mortality in a gene trap mouse model (Miura et al., 2006). Validation of its role in retrograde injury signaling therefore required a subcellular knockout that would remove it from axons without affecting its essential roles in nucleocytoplasmic transport in the neuronal cell body. Perry et al. (2012) identified an axon-localizing region in the 3′UTR of importin β1 and showed that targeting this region enables selective depletion of importin β1 from axons without perturbing its essential cell body functions. A long isoform of importin β1 3′UTR was shown to localize GFP and Dendra reporter proteins to axons in vitro and in transgenic mice, whereas short variants of the UTR had no such effect (Perry et al., 2012). Conditional targeting of the long 3′UTR region in mice caused subcellular loss of importin β1 mRNA and protein in adult sensory axons without reducing its cell body levels (Fig. 2). Importantly, the sensory neuron cell body transcriptome was not significantly affected in uninjured animals, indicating that subcellular knockout of importin β1 in axons had little or no effect on nuclear functions of this essential gene (Perry et al., 2012). In striking contrast, microarray analyses showed that axonal loss of importin β1 delayed and reduced the upregulation or downregulation of over 60% of the large gene ensembles changed upon injury. Moreover, both behavioral and histological analyses revealed a delay in the regeneration of injured sensory neurons in the importin β1 long 3′UTR null animals (Perry et al., 2012). Thus, localized translation of importin β1 mRNA enables separation of cytoplasmic and nuclear transport functions of importins and is required for efficient retrograde signaling in injured sensory axons.

image

Figure 2. Importin β1 3′UTR deletion in sensory axons (Reproduced with permission from Perry et al., Neuron 2012, 75, 294–305). A: Targeting strategy for importin β1 3′UTR. The loxP insertion sites are marked in red and the location of the 3′ and 5′ homology arms are in black. The PGK-neo selection cassette is inserted downstream of the region to be deleted (orange arrows) and flanked by FRT sites (green) that can be deleted using FLP recombinase. Three SV40 polyA signals are inserted immediately downstream of the floxed region (yellow boxes). B: Quantification of relative importin β1 transcript levels in DRG and sciatic nerve extracts of wild type, PGK-Cre (whole body knockout) and Adv-Cre (sensory neuron specific knockout) targeted mice. Note the significant decrease in message levels in the knockout nerves compared with wild type, coupled with increase in knockout DRGs, consistent with accumulation of message in ganglia due to the lack of an axon-targeting element. β actin served as an internal control, average ± SEM, n = 3, ∗p < 0.01, ∗∗p < 0.001 (unpaired two sample t-test). C: Sciatic nerve cross sections taken 6 h after nerve crush and immunostained for importin β1 (red) and the neuronal marker NF-H (green). Scale bar 20 μm. D: Quantification of axonal importin β1 immunoreactivity in sciatic nerve sections using CellProfiler, average ± SEM, n ≥ 1000, ∗∗p < 0.01 (unpaired two sample t-test). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

SUMMARY AND IMPLICATIONS

  1. Top of page
  2. ABSTRACT
  3. SUBCELLULAR mRNA LOCALIZATION IS A FUNDAMENTAL CHARACTERISTIC OF EUKARYOTIC CELLS
  4. IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN DENDRITES
  5. IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN AXONS BY TRANSGENIC OVEREXPRESSION
  6. IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN AXONS BY SUBCELLULAR KNOCKOUT
  7. SUMMARY AND IMPLICATIONS
  8. REFERENCES

Taken together, the transgenic and localized knockout studies summarized above strongly support mRNA localization and local protein synthesis as functionally important mechanisms in the injury response of adult sensory axons. The animal models generated in the studies of Donnelly et al. (2011) and Perry et al. (2012) will be useful to identify additional neuronal subtypes and physiological functions requiring local translation. In this context, it is interesting to note that a number of recent studies have also proposed roles for local translation in presynaptic and axonal compartments in the CNS (Akins et al., 2009). In two examples, local translation of beta-catenin was suggested to regulate synaptic vesicle release at presynaptic terminals (Taylor et al., 2013); and Fragile X granules were found to be expressed both axonally and presynaptically in central circuits (Akins et al., 2012). Other studies have highlighted roles for axonal mRNAs and local translation in motor neuron physiology (Akten et al., 2011; Fallini et al., 2011; Rathod et al., 2012). In a particularly interesting recent development, retrograde transport of pseudorabies virus was shown to depend on local protein synthesis in axons (Koyuncu et al., 2013). Perturbation of axonal mRNA localization and translation may therefore have significant consequences in central neurons in vivo, and this will be an interesting topic for future studies.

Finally, a pressing issue that arises from these studies is the need to understand the source and regulation of functional translational machineries in axons. Early morphological and immunohistochemical studies struggled to detect ribosomes in vertebrate axons (reviewed in Twiss and Fainzilber, 2009), and Koenig suggested that part of the difficulty in identifying ribosomes in myelinated axons is due to their concentration along axonal edges in relatively insoluble plaque-like structures (Koenig, 2009). Additional support for this notion has come from recent work of Flanagan and colleagues, who reported that the translation machinery associates with a transmembrane receptor at axonal membranes (Tcherkezian et al., 2010). Other recent studies have used diverse techniques to report new evidence for the occurrence of ribosomes in axons or presynaptic compartments. Michaelevski et al. (2010) carried out proteomics analyses on injured rat sciatic nerve and found that translation machinery accounted for one of the principal groups of proteins transported to the lesion site after injury. Holt and colleagues expressed a GFP-tagged ribosomal L10a protein in Xenopus retinal ganglion cells (RGC), and were able to immunoprecipitate locally translated mRNAs together with GFP-L10a from RGC axons, indicating that the axonal complexes contained functional ribosomes and that the associated mRNAs are indeed locally translated in Xenopus RGC axons in vivo (Yoon et al., 2012). Finally, Taylor et al. (2013) used a microfluidics system to isolate presynaptic terminals of hippocampal neurons, thus enabling their imaging separately from higher somatodendritic signals, and allowing visualization of ribosomal RNA in the axonal compartment. Diverse sources have been proposed for axonal ribosomes or mRNAs, including the possibility of specialized ribosome-containing structural domains in axons (Koenig, 2009), or the intriguing possibility of lateral transfer from adjacent glia upon an appropriate stimulus (Court et al., 2008, 2011; Sotelo et al., 2013). The latter may be part of a continuum of mechanisms by which glia provide trophic and metabolic support to axons (Nave, 2010; Funfschilling et al., 2012). Clarification of this issue may be facilitated by the availability of engineered mouse models with specific epitope tag insertions at endogenous ribosomal protein loci (Sanz et al., 2009; Walker et al., 2012). A broad effort to generate such mice, as well as additional knockout models of axon localization motifs in candidate gene UTRs, will be important to assess the overall scope and significance of local translation in neuronal functions.

M.F. is the incumbent of the Chaya Professorial Chair in Molecular Neuroscience at the Weizmann Institute of Science. The authors have no conflict of interest to declare.

REFERENCES

  1. Top of page
  2. ABSTRACT
  3. SUBCELLULAR mRNA LOCALIZATION IS A FUNDAMENTAL CHARACTERISTIC OF EUKARYOTIC CELLS
  4. IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN DENDRITES
  5. IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN AXONS BY TRANSGENIC OVEREXPRESSION
  6. IN VIVO EVIDENCE FOR LOCAL TRANSLATION IN AXONS BY SUBCELLULAR KNOCKOUT
  7. SUMMARY AND IMPLICATIONS
  8. REFERENCES