Final Report Summary - AXONSURVIVAL (Axon survival: the role of protein synthesis)
Background
Neurons in the brain and their long-projecting axons typically survive throughout the lifetime of an organism. In neurodegenerative diseases, axons commonly degenerate with devastating clinical consequences, therefore, it is important to understand how these long neuronal processes are established and maintained. Previously, we discovered that a new supply of proteins in axons, through the translation of localised messenger RNAs (mRNAs), is needed to keep axons alive in the vertebrate visual system in Xenopus (frog). We also found that cultured axons contain hundreds of different mRNAs and, among these, ribosomal protein (RP) encoding mRNAs were particularly enriched. Our project set out to uncover the mechanism(s) underlying axon survival and the establishment of axon architecture, with a particular focus on the role of RPs. We asked whether axonal translation occurs in CNS axons in the mammalian brain in vivo and, if so, whether it is critical for maintaining axon viability. The project has given rise to 27 publications (24 peer-reviewed papers and 3 book chapters) and the highlights of some of our achievements are summarized below.
Achievements
1. Adult and developing mammalian CNS axons synthesize hundreds of proteins locally.
We developed an Axon-TRAP approach in mouse that allowed us to characterize which mRNAs are being translated in the axons from the eye (retinal ganglion cell axons) and thereby discover the in vivo translatome of the axonal compartment for the first time. Our results provided direct evidence that adult CNS axons contain ribosomes and are translating mRNAs. We found that the axonal translatome is highly complex (>2500 genes) and that ribosomal protein (RP) mRNAs comprise a highly enriched category indicating that RPs are axonally synthesized in vivo. Interestingly, the axonal translatome is distinct from the somal translatome and it changes in-step with the axon architectural changes (arborisation, pruning, synaptogenesis) integral to establishing neural connectivity.
2. RNA docking and local translation required for establishing complex axon architecture.
The number of branches an axon has determines the complexity of its architecture and, importantly, the number of synaptic connections it can make. Low complexity branching (simple architecture) has been associated with neurodevelopmental diseases, such as autism. Our work has shown that RNA granules move to and dock at sites that predict new branch formation. New beta-actin proteins are synthesized at these sites and inhibition of this localised translation causes a severe loss of branch formation. This finding has broad implications both biological and clinical for understanding how complex neuronal architecture is established.
3. Late endosomes transport RNAs and serve as 'platforms' for the synthesis of proteins to sustain mitochondria and keep axons alive.
We discovered a new mechanism of RNA transport in axons whereby RNA granules ride along axons on endosomes. RNA-bearing endosomes frequently stop, or dock, on mitochondria and these sites become 'hotspots' of translation. Importantly, pro-survival proteins synthesized at these sites are required to maintain mitochondrial function and axon viability. Our work makes an important novel link between local translation, mitochondrial health and axon maintenance.
4. Different cues trigger distinct nascent proteomic signatures in axons within minutes.
We established a novel protocol that allowed us to detect and characterize the proteins synthesized in axons within 5 minutes of the addition of a cue. We compared three different cues (Netrin, BDNF, Sema3A) and found that each one elicited the synthesis of many proteins very quickly. This proteomics analysis, done on an unprecedented timescale and on tiny amounts of material, was made possible by a new single-pot method of proteomics. The results show that a localized cue can cause fast and highly localized changes in the local proteome at a subcellular level. The results have broad implications for understanding receptor-mediated (extrinsic cue) cell behaviours and support the notion that growth receptor-mediated translation supports axon maintenance.
5. Noncanonical modulation of eIF2 pathway controls increase in local translation during neural wiring.
The eIF2 pathway is well known to control the canonical stress response which leads to the repression of all but a small handful of proteins. While investigating the pathway controlling axonal translation we found, paradoxically, that Sema3A stimulation upregulates the translation of over 70 specific mRNAs via eIF2α. Thus our findings reveal the existence of a noncanonical eIF2 signaling pathway that controls selective changes in axon translation and that is required for neural wiring. The findings lead to a new model of translation regulation in axons and suggest that it is likely to be a widely employed pathway in cellular compartments in response to bursts of cue-induced protein synthesis predicted to rapidly overload the ER with unfolded nascent proteins.
6. The ALS neurodegenerative disease associated protein, FUS, disrupts protein synthesis in axons.
We discovered that FUS protein mutants associated with ALS disrupt axonal translation. FUS is an RNA-binding protein with a low complexity domain and reversible liquid-solid phase transition properties. Mutant FUS forms irreversible solid aggregates in cells and the mechanism by which it acts to cause neurodegeneration has been illusive. Our finding brings a new way of thinking about the problem by focusing attention on the possibility that the common denominator in neurodegenerative diseases is the dysregulation of translation.
7. On-site ribosome remodelling by locally synthesized ribosomal proteins in axons.
Since eukaryotic ribosomal proteins (RPs) are assembled into ribosomes in the nucleolus, our Axon-TRAP-RiboTag results showing the widespread synthesis of RPs in axons was intriguing but enigmatic. We have spent a significant part of the last two years addressing how RP mRNA translation is regulated in axons and investigating whether axonally synthesized RPs become incorporated into the axonal ribosomes. Using subcellular proteomic analysis and live-imaging, we show that locally synthesized RPs incorporate into axonal ribosomes in a nucleolus independent fashion. We discovered that axonal RP translation is regulated through a novel sequence motif, CUIC, that forms an RNA-loop structure in the region immediately upstream of the initiation codon. Furthermore, functional inhibition of axonal CUIC-regulated RP translation in vivo leads to defects in local translation activity and axon branching. Our results suggest that ribosomes in axons may not be be of fixed composition and, instead, may be able to exchange ribosomal proteins and repair subunits. A flexible mechanism of this sort could have far-reaching consequences enabling ribosomes to be tuned on-site.
Neurons in the brain and their long-projecting axons typically survive throughout the lifetime of an organism. In neurodegenerative diseases, axons commonly degenerate with devastating clinical consequences, therefore, it is important to understand how these long neuronal processes are established and maintained. Previously, we discovered that a new supply of proteins in axons, through the translation of localised messenger RNAs (mRNAs), is needed to keep axons alive in the vertebrate visual system in Xenopus (frog). We also found that cultured axons contain hundreds of different mRNAs and, among these, ribosomal protein (RP) encoding mRNAs were particularly enriched. Our project set out to uncover the mechanism(s) underlying axon survival and the establishment of axon architecture, with a particular focus on the role of RPs. We asked whether axonal translation occurs in CNS axons in the mammalian brain in vivo and, if so, whether it is critical for maintaining axon viability. The project has given rise to 27 publications (24 peer-reviewed papers and 3 book chapters) and the highlights of some of our achievements are summarized below.
Achievements
1. Adult and developing mammalian CNS axons synthesize hundreds of proteins locally.
We developed an Axon-TRAP approach in mouse that allowed us to characterize which mRNAs are being translated in the axons from the eye (retinal ganglion cell axons) and thereby discover the in vivo translatome of the axonal compartment for the first time. Our results provided direct evidence that adult CNS axons contain ribosomes and are translating mRNAs. We found that the axonal translatome is highly complex (>2500 genes) and that ribosomal protein (RP) mRNAs comprise a highly enriched category indicating that RPs are axonally synthesized in vivo. Interestingly, the axonal translatome is distinct from the somal translatome and it changes in-step with the axon architectural changes (arborisation, pruning, synaptogenesis) integral to establishing neural connectivity.
2. RNA docking and local translation required for establishing complex axon architecture.
The number of branches an axon has determines the complexity of its architecture and, importantly, the number of synaptic connections it can make. Low complexity branching (simple architecture) has been associated with neurodevelopmental diseases, such as autism. Our work has shown that RNA granules move to and dock at sites that predict new branch formation. New beta-actin proteins are synthesized at these sites and inhibition of this localised translation causes a severe loss of branch formation. This finding has broad implications both biological and clinical for understanding how complex neuronal architecture is established.
3. Late endosomes transport RNAs and serve as 'platforms' for the synthesis of proteins to sustain mitochondria and keep axons alive.
We discovered a new mechanism of RNA transport in axons whereby RNA granules ride along axons on endosomes. RNA-bearing endosomes frequently stop, or dock, on mitochondria and these sites become 'hotspots' of translation. Importantly, pro-survival proteins synthesized at these sites are required to maintain mitochondrial function and axon viability. Our work makes an important novel link between local translation, mitochondrial health and axon maintenance.
4. Different cues trigger distinct nascent proteomic signatures in axons within minutes.
We established a novel protocol that allowed us to detect and characterize the proteins synthesized in axons within 5 minutes of the addition of a cue. We compared three different cues (Netrin, BDNF, Sema3A) and found that each one elicited the synthesis of many proteins very quickly. This proteomics analysis, done on an unprecedented timescale and on tiny amounts of material, was made possible by a new single-pot method of proteomics. The results show that a localized cue can cause fast and highly localized changes in the local proteome at a subcellular level. The results have broad implications for understanding receptor-mediated (extrinsic cue) cell behaviours and support the notion that growth receptor-mediated translation supports axon maintenance.
5. Noncanonical modulation of eIF2 pathway controls increase in local translation during neural wiring.
The eIF2 pathway is well known to control the canonical stress response which leads to the repression of all but a small handful of proteins. While investigating the pathway controlling axonal translation we found, paradoxically, that Sema3A stimulation upregulates the translation of over 70 specific mRNAs via eIF2α. Thus our findings reveal the existence of a noncanonical eIF2 signaling pathway that controls selective changes in axon translation and that is required for neural wiring. The findings lead to a new model of translation regulation in axons and suggest that it is likely to be a widely employed pathway in cellular compartments in response to bursts of cue-induced protein synthesis predicted to rapidly overload the ER with unfolded nascent proteins.
6. The ALS neurodegenerative disease associated protein, FUS, disrupts protein synthesis in axons.
We discovered that FUS protein mutants associated with ALS disrupt axonal translation. FUS is an RNA-binding protein with a low complexity domain and reversible liquid-solid phase transition properties. Mutant FUS forms irreversible solid aggregates in cells and the mechanism by which it acts to cause neurodegeneration has been illusive. Our finding brings a new way of thinking about the problem by focusing attention on the possibility that the common denominator in neurodegenerative diseases is the dysregulation of translation.
7. On-site ribosome remodelling by locally synthesized ribosomal proteins in axons.
Since eukaryotic ribosomal proteins (RPs) are assembled into ribosomes in the nucleolus, our Axon-TRAP-RiboTag results showing the widespread synthesis of RPs in axons was intriguing but enigmatic. We have spent a significant part of the last two years addressing how RP mRNA translation is regulated in axons and investigating whether axonally synthesized RPs become incorporated into the axonal ribosomes. Using subcellular proteomic analysis and live-imaging, we show that locally synthesized RPs incorporate into axonal ribosomes in a nucleolus independent fashion. We discovered that axonal RP translation is regulated through a novel sequence motif, CUIC, that forms an RNA-loop structure in the region immediately upstream of the initiation codon. Furthermore, functional inhibition of axonal CUIC-regulated RP translation in vivo leads to defects in local translation activity and axon branching. Our results suggest that ribosomes in axons may not be be of fixed composition and, instead, may be able to exchange ribosomal proteins and repair subunits. A flexible mechanism of this sort could have far-reaching consequences enabling ribosomes to be tuned on-site.