Ben G. Szaro

Post-transcriptional control of neurofilaments: New roles in development, regeneration and neurodegenerative disease.

Neurofilament (NF) protein expression is coupled to axon development and the maintenance of neuronal homeostasis. Here, we present evidence that this tight regulation depends critically on post-transcriptionally regulated changes in NF mRNA transport, translation and stability. Recent studies have shown that post-transcriptional mechanisms modulate increases in NF gene transcription during axon regeneration to yield the final pattern of NF protein expression. Other recent work has found that post-transcriptional control of NFs shares elements with that of other axonal proteins and that its dysregulation contributes to amyotrophic lateral sclerosis. Such studies herald a novel approach to understanding how neurons coordinate the expressions of functionally related proteins and provide new insights into how the dysregulation of this control can lead to disease.

hnRNP K post-transcriptionally co-regulates multiple cytoskeletal genes needed for axonogenesis

The RNA-binding protein, hnRNP K, is essential for axonogenesis. Suppressing its expression in Xenopus embryos yields terminally specified neurons with severely disorganized microtubules, microfilaments and neurofilaments, raising the hypothesis that hnRNP K post-transcriptionally regulates multiple transcripts of proteins that organize the axonal cytoskeleton. To identify downstream candidates for this regulation, RNAs that co-immunoprecipitated from juvenile brain with hnRNP K were identified on microarrays. A substantial number of these transcripts were linked to the cytoskeleton and to intracellular localization, trafficking and transport. Injection into embryos of a non-coding RNA bearing multiple copies of an hnRNP K RNA-binding consensus sequence found within these transcripts largely phenocopied hnRNP K knockdown, further supporting the idea that it regulates axonogenesis through its binding to downstream target RNAs. For further study of regulation by hnRNP K of the cytoskeleton during axon outgrowth, we focused on three validated RNAs representing elements associated with all three polymers – Arp2, tau and an α-internexin-like neurofilament. All three were co-regulated post-transcriptionally by hnRNP K, as hnRNP K knockdown yielded comparable defects in their nuclear export and translation but not transcription. Directly knocking down expression of all three together, but not each one individually, substantially reproduced the axonless phenotype, providing further evidence that regulation of axonogenesis by hnRNP K occurs largely through pleiotropic effects on cytoskeletal-associated targets. These experiments provide evidence that hnRNP K is the nexus of a novel post-transcriptional regulatory module controlling the synthesis of proteins that integrate all three cytoskeletal polymers to form the axon.

Metamorphosis and the regenerative capacity of spinal cord axons in Xenopus laevis.

Throughout the vertebrate subphylum, the regenerative potential of central nervous system axons is greatest in embryonic stages and declines as development progresses. For example, Xenopus laevis can functionally recover from complete transection of the spinal cord as a tadpole but is unable to do so after metamorphosing into a frog. Neurons of the reticular formation and raphe nucleus are among those that regenerate axons most reliably in tadpole and that lose this ability after metamorphosis. To identify molecular factors associated with the success and failure of spinal cord axon regeneration, we pharmacologically manipulated thyroid hormone (TH) levels using methimazole or triiodothyronine, to either keep tadpoles in a permanently larval state or induce precocious metamorphosis, respectively. Following complete spinal cord transection, serotonergic axons crossed the lesion site and tadpole swimming ability was restored when metamorphosis was inhibited, but these events failed to occur when metamorphosis was prematurely induced. Thus, the metamorphic events controlled by TH led directly to the loss of regenerative potential. Microarray analysis identified changes in hindbrain gene expression that accompanied regeneration‐permissive and ‐inhibitory conditions, including many genes in the permissive condition that have been previously associated with axon outgrowth and neuroprotection. These data demonstrate that changes in gene expression occur within regenerating neurons in response to axotomy under regeneration‐permissive conditions in which normal development has been suspended, and they identify candidate genes for future studies of how central nervous system axons can successfully regenerate in some vertebrates.

A crucial role for hnRNP K in axon development in Xenopus laevis.

We report that hnRNP K, an RNA-binding protein implicated in multiple aspects of post-transcriptional gene control, is essential for axon outgrowth in Xenopus. Its intracellular localization was found to be consistent with one of its known roles as an mRNA shuttling protein. In early embryos, it was primarily nuclear, whereas later it occupied both the nucleus and cytoplasm to varying degrees in different neuronal subtypes. Antisense hnRNP K morpholino oligonucleotides (MOs) microinjected into blastomeres suppressed hnRNP K expression from neural plate stages through to at least stage 40. Differentiating neural cells in these embryos expressed several markers for terminally differentiated neurons but failed to make axons. Rescue experiments and the use of two separate hnRNP K MOs were carried out to confirm that these effects were specifically caused by knockdown of hnRNP K expression. For insights into the involvement of hnRNP K in neuronal post-transcriptional gene control at the molecular level, we compared effects on expression of the medium neurofilament protein (NF-M), the RNA for which binds hnRNP K, with that of peripherin, another intermediate filament protein, the RNA for which does not bind hnRNP K. hnRNP K knockdown compromised NF-M mRNA nucleocytoplasmic export and translation, but had no effect on peripherin. Because eliminating NF-M from Xenopus axons attenuates, but does not abolish, their outgrowth, hnRNP K must target additional RNAs needed for axon development. Our study supports the idea that translation of at least a subset of RNAs involved in axon development is controlled by post-transcriptional regulatory modules that have hnRNP K as an essential element.

Xenopus laevis peripherin (XIF3) is expressed in radial glia and proliferating neural epithelial cells as well as in neurons

Neuronal intermediate filament (nIF) proteins form the most abundant component of the axonal cytoskeleton. Thus, understanding their function and the regulation of their expression is essential for comprehending how axonal structure is regulated. Although most vertebrate nIF proteins are classified as type IV intermediate filament (IF) proteins, additional nIF proteins exist in frogs (Xenopus laevis), cyprinid fishes, and mammals (called XIF3, plasticin, and peripherin, respectively) that are classified as type III. Expression of a type III nIF protein is correlated strongly with the earliest phases of axonal outgrowth in fishes but less so in mammals. To understand better how the correlation between type III nIF protein expression and early phases of axonal outgrowth has changed during evolution, the authors examined XIF3 expression in Xenopus laevis. In Xenopus, the association between XIF3 expression and early axonal outgrowth was especially strong. For example, during early axonal development, XIF3 expression preceded and was more abundant and widespread than that of any of the type IV nIF proteins. As axons matured, neuronal expression of XIF3 gradually became more restricted while that of type IV nIF proteins increased. These results support the idea that type III nIF proteins play a special role during early phases of axonal outgrowth. In addition to finding XIF3 in neurons, the authors also unexpectedly found it in regions of the central nervous system that contain proliferating cells and radial glia. As a framework for interpreting variations in nIF expression in different vertebrate species, the authors built phylogenetic trees to clarify relationships among vertebrate nIF proteins. These trees supported the classification of XIF3, plasticin, and peripherin as orthologs (products of the same genetic locus, evolving separately only since the species lineages diverged). Thus, XIF3, plasticin, and peripherin probably should be referred to as Xenopus, fish, and mammalian peripherin, respectively. This finding argues that differences in expression of these three proteins in frogs, fishes, and mammals are the result of regulatory changes to the peripherin ancestral gene along each lineage. The expression of a peripherin ortholog in Xenopus glia may represent either an adaptation that arose since the divergence of Xenopus from mammals or, alternatively, a feature retained from an ancestral IF protein that was expressed originally both in neurons and in glia. J. Comp. Neurol. 423:512–531, 2000.

Phylogenetically Conserved Binding of Specific K Homology Domain Proteins to the 3-Untranslated Region of the Vertebrate Middle Neurofilament mRNA*

As axons mature, neurofilament-M (NF-M) expression rises, contributing to maturation of the axonal cytoskeleton and an expansion in axon caliber. This increase is partly due to a rise in NF-M mRNA stability. Such post-transcriptional regulation is often mediated through the binding of specific proteins to the 3′-untranslated region (3′-UTR) of mRNAs. Vertebrate NF-M 3′-UTRs are remarkably well conserved, prompting us to test whether similar proteins bind the 3′-UTRs of different vertebrate NF-Ms. Identification of such proteins could lead to insights into the regulation of NF-M expression during development and in response to trauma or disease. Ultraviolet cross-linking analysis of proteins isolated from adult frog (Xenopus laevis), mouse, and rat brains revealed three ribonucleoprotein complexes (97, 70, and 47 kDa) that were present in all species and bound specifically to NF-M 3′-UTRs. Affinity purification of NF-M 3′-UTR-binding proteins from rat brain followed by mass spectrometry and immunoprecipitation assays identified heterogeneous nuclear ribonucleoprotein (hnRNP) K and hnRNP E1 as the proteins forming the 70- and 47-kDa complexes, respectively. These RNA-binding proteins of the KH domain family recognize CU-rich motifs identical to ones present in NF-M 3′-UTRs. Ultraviolet cross-linking assays performed on Xenopus embryos at different stages of neural development demonstrated that whereas hnRNP K binding occurred at all stages, hnRNP E binding occurred only at the most mature stages of axon development. Since hnRNP E is known to stabilize mRNAs, these results raise the hypothesis that these proteins may contribute to the increases in cytoplasmic levels of NF-M mRNA that accompany axonal maturation.

Increased expression of multiple neurofilament mRNAs during regeneration of vertebrate central nervous system axons.

Characteristic changes in the expression of neuronal intermediate filaments (nIFs), an abundant cytoskeletal component of vertebrate axons, accompany successful axon regeneration. In mammalian regenerating PNS, expression of nIFs that are characteristic of mature neurons becomes suppressed throughout regeneration, whereas that of peripherin, which is abundant in developing axons, increases. Comparable changes are absent from mammalian injured CNS; but in goldfish and lamprey CNS, expression of several nIFs increases during axon regrowth. To obtain a broader view of the nIF response of successfully regenerating vertebrate CNS, in situ hybridization and video densitometry were used to track multiple nIF mRNAs during optic axon regeneration in Xenopus laevis. As in other successfully regenerating systems, peripherin expression increased rapidly after injury and expression of those nIFs characteristic of mature retinal ganglion cells decreased. Unlike the decrease in nIF mRNAs of regenerating PNS, that of Xenopus retinal ganglion cells was transient, with most nIF mRNAs increasing above normal during axon regrowth. At the peak of regeneration, increases in each nIF mRNA resulted in a doubling of the total amount of nIF mRNA, as well as a shift in the relative proportions contributed by each nIF. The relative proportions of peripherin and NF‐M increased above normal, whereas proportions of xefiltin and NF‐L decreased and that of XNIF remained the same. The increases in peripherin and NF‐M mRNAs were accompanied by increases in protein. These results are consistent with the hypothesis that successful axon regeneration involves changes in nIF subunit composition conducive to growth and argue that a successful injury response differs between CNS and PNS. J. Comp. Neurol. 461:262–275, 2003.

Xefiltin, a Xenopus laevis neuronal intermediate filament protein, is expressed in actively growing optic axons during development and regeneration

Neurofilaments are an important structural component of the axonal cytoskeleton and are made of neuronal intermediate filament (nIF) proteins. During axonal development, neurofilaments undergo progressive changes in molecular composition. In mammals, for example, highly phosphorylated forms of the middle- and high-molecular-weight neurofilament proteins (NF-M and NF-H, respectively) are characteristic of mature axons, whereas nIF proteins such as alpha-internexin are typical of young axons. Such changes have been proposed to help growing axons accommodate varying demands for plasticity and stability by modulating the structure of the axonal cytoskeleton. Xefiltin is a recently discovered nIF protein of the frog Xenopus laevis, whose nervous system has a large capacity for regeneration and plasticity. By amino acid identity, xefiltin is closely related to two other nIF proteins, alpha-internexin and gefiltin. alpha-Internexin is found principally in embryonic axons of the mammalian brain, and gefiltin is expressed primarily in goldfish retinal ganglion cells and has been associated with the ability of the goldfish optic nerve to regenerate. Like gefiltin in goldfish, xefiltin in Xenopus is the most abundantly expressed nIF protein of mature retinal ganglion cells. In the present study, we used immunocytochemistry to study the distribution of xefiltin during optic nerve development and regeneration. During development, xefiltin was found in optic axons at stage 35/36, before they reach the tectum at stage 37/38. Similarly, after an orbital crush injury, xefiltin first reemerged in optic axons after the front of regeneration reached the optic chiasm, but before it reached the tectum. Thus, during both development and regeneration, xefiltin was present within actively growing optic axons. In addition, aberrantly projecting retinoretinal axons expressed less xefiltin than those entering the optic tract, suggesting that xefiltin expression is influenced by interactions between regenerating axons and cells encountered along the visual pathway. These results support the idea that changes in xefiltin expression, along with those of other nIF proteins, modulate the structure and stability of actively growing optic axons and that this stability is under the control of the pathway which growing axons follow.

Heterogeneous Nuclear Ribonucleoprotein K, an RNA-Binding Protein, Is Required for Optic Axon Regeneration in Xenopus laevis

Axotomized optic axons of Xenopus laevis, in contrast to those of mammals, retain their ability to regenerate throughout life. To better understand the molecular basis for this successful regeneration, we focused on the role of an RNA-binding protein, heterogeneous nuclear ribonucleoprotein (hnRNP) K, because it is required for axonogenesis during development and because several of its RNA targets are under strong post-transcriptional control during regeneration. At 11 d after optic nerve crush, hnRNP K underwent significant translocation into the nucleus of retinal ganglion cells (RGCs), indicating that the protein became activated during regeneration. To suppress its expression, we intravitreously injected an antisense Vivo-Morpholino oligonucleotide targeting hnRNP K. In uninjured eyes, it efficiently knocked down hnRNP K expression in only the RGCs, without inducing either an axotomy response or axon degeneration. After optic nerve crush, staining for multiple markers of regenerating axons showed no regrowth of axons beyond the lesion site with hnRNP K knockdown. RGCs nonetheless responded to the injury by increasing expression of multiple growth-associated RNAs and experienced no additional neurodegeneration above that normally seen with optic nerve injury. At the molecular level, hnRNP K knockdown during regeneration inhibited protein, but not mRNA, expression of several known hnRNP K RNA targets (NF-M, GAP-43) by compromising their efficient nuclear transport and disrupting their loading onto polysomes for translation. Our study therefore provides evidence of a novel post-transcriptional regulatory pathway orchestrated by hnRNP K that is essential for successful CNS axon regeneration.

Dynamic endogenous association of neurofilament mRNAs with K-homology domain ribonucleoproteins in developing cerebral cortex.

The low, middle, and high molecular mass neurofilament subunit proteins (NF-L, NF-M, and NF-H) co-polymerize to form neurofilaments (NFs). During development, NF subunit expression is highly regulated, and in neurodegenerative disease, aberrant regulation of this expression can lead to the formation of harmful aggregates. NF expression in both development and disease is under significant post-transcriptional control, but the specific ribonucleoproteins (RNPs) involved are only poorly understood. Previously, mass spectrometry on affinity purified proteins from rat brain identified three K-homology (KH) domain RNPs - hnRNP K, hnRNP E1, hnRNP E2 - as being capable of binding NF-M RNA. In the current study, to determine whether these RNPs associate with NF mRNAs endogenously, we performed a co-immunoprecipitation assay on homogenates of postnatal and developing rat cerebral cortex. We found that all three NF mRNAs indeed associated endogenously with these RNPs and that the degree of this association changed during postnatal development, a period when NF expression is under significant post-transcriptional control. The degree of these associations changed independently of the abundance of either the RNPs or the NF messages, indicating that the RNA-protein interactions themselves are directly regulated. This study is consistent with a model whereby these RNPs and NF mRNAs are components of a dynamic post-transcriptional regulatory module that influences the cytoskeletal compositions of neurons.