SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Protrudin is a membrane protein that regulates polarized vesicular transport. Now, we have identified a novel isoform of protrudin (protrudin-L) that contains an additional seven amino acids between the FFAT motif and the coiled-coil domain compared with the conventional isoform (protrudin-S) as a result of alternative splicing of a microexon (exon L). Protrudin-L mRNA was found to be mostly restricted to the central nervous system in mice, whereas protrudin-S mRNA was detected in all tissues examined. With the use of a splicing reporter minigene that produces two distinct fluorescent proteins in a manner dependent on the splicing pattern of protrudin transcripts, we found that most neurons express protrudin-L, whereas astrocytes express both protrudin isoforms and oligodendrocytes express only protrudin-S. Protrudin-L associated to a greater extent with vesicle-associated membrane protein–associated protein (VAP) than protrudin-S. Expression of protrudin-L in hippocampal neurons of protrudin-deficient mice also promoted neurite outgrowth more efficiently than protrudin-S. Our results suggest that protrudin-L is a neuron-specific protrudin isoform that promotes axonal elongation and contributes to the establishment of neuronal polarity.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Neurons are highly polarized cells and extend neurites that can achieve a length of >1 m and which contribute to the organization of neural networks. The molecular mechanisms that underlie neurite formation include both remodeling of the cytoskeleton and membrane trafficking (Pfenninger 2009; Tahirovic & Bradke 2009). During the establishment of neuronal polarity, membrane components are transported in a directional manner within the cell to ensure both the supply of new material to the growing processes and the distribution of different sets of molecules to distinct domains (Solecki et al. 2006; Arimura & Kaibuchi 2007). The small GTPase Rab11 is suggested to contribute to this process through the regulation of membrane recycling (Zerial & McBride 2001; Stenmark 2009).

We recently identified protrudin as a key regulator of Rab11-dependent vesicular trafficking during neurite extension through polarized membrane transport (Shirane & Nakayama 2006; Shirane et al. 2008). Protrudin contains a Rab11-binding domain (RBD) for interaction with Rab11, and three hydrophobic domains (HP-1, HP-2, and HP-3), a short sequence motif designated FFAT (two phenylalanines in an acidic tract), a coiled-coil domain, and a FYVE (Fab1, YOTB, Vac1, and EEA1) domain (Gillooly et al. 2001; Gil et al. 2012). Over-expression of protrudin in PC12 pheochromocytoma cells or hippocampal neurons was found to result in extensive neurite outgrowth, whereas the depletion of endogenous protrudin by RNA interference resulted in the inhibition of neurite outgrowth and in swelling of the cell soma in PC12 cells stimulated with nerve growth factor. Protrudin forms a complex with several proteins, including spastin, KIF5, and vesicle-associated membrane protein-associated protein (VAP) (Shirane & Nakayama 2006; Saita et al. 2009; Matsuzaki et al. 2011). Spastin is a member of the AAA (ATPases associated with various cellular activities) family of proteins and serves as a microtubule-severing enzyme. Over-expression or depletion of spastin in neurons was found to result in marked promotion or inhibition of neurite outgrowth, respectively (Salinas et al. 2007). KIF5 is a component of kinesin-1, a molecular motor that plays key roles in the polarized transport system of neurons. Mutations in the genes for spastin and KIF5A (a neuron-specific isoform of KIF5) are associated with autosomal dominant forms SPG4 and SPG10, respectively, of hereditary spastic paraplegia (AD-HSP) (Hazan et al. 1999; Reid et al. 1999). The protrudin gene has also been identified as a responsible gene for AD-HSP (SPG33 subtype) (Mannan et al. 2006). We previously showed that the association of protrudin with VAP is an important determinant of both the subcellular localization of protrudin and its ability to stimulate neurite outgrowth (Saita et al. 2009). Overall, these findings suggest that protrudin might play a key role in neuronal development and differentiation, and its dysfunction is likely related to the pathogenesis of AD-HSP.

Alternative splicing is an important mechanism of gene regulation and determinant of proteomic diversity (Johnson et al. 2003). In metazoans, alternative splicing thus contributes to the generation of alternative protein products that function in diverse cellular processes including cell growth, differentiation, and death (Moroy & Heyd 2007; Wang & Burge 2008). The decision as to which exons are included or excluded in the mature transcripts of a gene is based on the presence of RNA sequence elements and protein regulators (Black 2003). The splicing events are mediated by the spliceosome, a large structure in which five small nuclear ribonucleoprotein particles (snRNPs) and many auxiliary proteins cooperate to achieve the precise recognition of a splice site and then catalyze the two steps of the splicing reaction (Wahl et al. 2009). Tissue-specific alternative splicing events are regulated at the level of tissue-specific expression of splicing factors. The proteins generated as a result of alternative splicing contribute to the characteristic functions and properties of different cell types (Licatalosi & Darnell 2006; Li et al. 2007; Ellis et al. 2012).

Now, we have identified a novel transcript of the mouse protrudin gene that is generated by alternative splicing. The protein product of this transcript (protrudin-L) is expressed specifically in the central nervous system and selectively in neuronal cells. Protrudin-L associates with VAP to a greater extent than does the originally identified isoform of protrudin (protrudin-S). We also show that protrudin-L regulates axonal elongation and contributes to the establishment of neuronal cell polarity.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Identification of a novel splicing isoform of protrudin

Immunoblot analysis of immunoprecipitates prepared with antibodies to protrudin from mouse cerebrum or cerebellum revealed the presence of two forms of protrudin with different electrophoretic mobilities, with the intensity of the slower-migrating band being much greater than that of the faster-migrating one (Fig. 1A). In contrast, only the faster-migrating band was detected in the thymus and liver. To investigate the mechanism underlying the difference in the mobility of these two forms of protrudin, we subjected protrudin cDNA prepared from the thymus or cerebrum to direct sequencing. Protrudin cDNA from the cerebrum, but not that from the thymus, was found to contain an additional 21 nucleotides between the sequences corresponding to exons IX and X (Fig. 1B). Comparison of this intervening sequence with the corresponding mouse genomic sequence revealed the presence in the latter of an unidentified exon (designated exon L) flanked by putative 5′ and 3′ splice sites (Fig. 1C). This additional sequence corresponding to exon L was predicted to encode seven amino acids. The isoform of protrudin that contains this additional seven amino acid sequence was designated protrudin-L, whereas that without this sequence was designated protrudin-S. Recombinant forms of mouse protrudin-L and protrudin-S showed electrophoretic mobilities identical to those of the slower- and faster-migrating bands of endogenous protrudin, respectively (Fig. 1A). Exon L is flanked by an atypical 3′ splice site that contains a polypurine (GGGAG) tract instead of the polypyrimidine (C and U/T) tract that usually resides upstream of the invariant AG sequence in 3′ splice sites (Fig. 1C). These findings suggested that exon L may be alternatively spliced in a tissue-specific manner.

image

Figure 1. Identification of a novel splicing variant of protrudin. (A) Immunoblot analysis with antibodies to protrudin. Lysates prepared from the indicated tissues of mice at 12 weeks of age were subjected to Immunoprecipitation with antibodies to protrudin. Lysates of HEK293T cells transfected with pcDNA3 vectors for mouse protrudin-S or protrudin-L were examined as controls. The resulting precipitates as well as cell lysates were subjected to immunoblot analysis with antibodies to protrudin. (B) Direct sequencing analysis of protrudin cDNA prepared from the thymus or cerebrum of mice at 4 weeks of age. The boxed region in the lower panel indicates an intervening sequence between exons IX and X of the protrudin gene. (C) Genomic structure of the mouse protrudin gene. Exon L encoding the intervening sequence in protrudin-L is shown in red. SS, splice site. Asterisks indicate an atypical purine tract in the 3′ splice site of exon L.

Download figure to PowerPoint

Protrudin-L is expressed specifically in neuronal tissues and cells

To examine the tissue distribution of protrudin-L and protrudin-S mRNAs in mouse, we carried out quantitative reverse transcription and polymerase chain reaction (qRT-PCR) analysis with PCR primers specific for each isoform. For absolute quantification of mRNA, the qRT-PCR data were compared with a standard curve obtained by PCR with plasmid DNA corresponding to protrudin-L or protrudin-S. Protrudin-L mRNA was detected almost exclusively in the cerebrum and spinal cord, being mostly undetectable in non-neural tissues (Fig. 2A). In contrast, protrudin-S mRNA was present in all tissues examined. A more detailed analysis of the brain revealed that protrudin-L mRNA was present in all regions examined, albeit in moderately different amounts (Fig. 2B). Similarly, protrudin-L mRNA was abundant in primary cultured cortical neurons and the neuroblastoma cell line Neuro2A, whereas it was virtually absent from NIH 3T3 cells (Fig. 2C). The purity of the cortical neurons was verified by qRT-PCR analysis of mRNAs for both the neuronal marker βIII-tubulin (Tubb3) and the astrocyte marker GFAP (glial fibrillary acidic protein) and the oligodendrocyte marker Olig2 (Fig. 2D). Together, these results thus suggested that protrudin-L mRNA is produced specifically in neuronal cells.

image

Figure 2. Tissue- and cell type–specific expression of protrudin isoforms. (A–C) Quantitative RT-PCR analysis of the absolute amounts of protrudin-S and protrudin-L mRNAs in the indicated mouse tissues (A), in mouse brain regions (B), and in primary cultured mouse cortical neurons or cell lines (C). (D) Quantitative RT-PCR analysis of Tubb3 (neuronal marker), GFAP (astrocyte marker), and Olig2 (oligodendrocyte marker) mRNAs in the cerebrum and cultured cortical neurons. The amount of each target mRNA was normalized by that of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA and is presented relative to the corresponding normalized value for the cerebrum. All data are means ± SEM of triplicates from representative experiments. **< 0.01, ***< 0.001 (Student's t-test (A–C) or Welch's t-test (D)).

Download figure to PowerPoint

Reporter system to monitor alternative splicing of protrudin mRNA

To visualize the pattern of alternative splicing for protrudin mRNA at the single-cell level by fluorescence imaging analysis, we took advantage of a bichromatic fluorescence splicing reporter system that emits green or red fluorescence when exon L is included in or excluded from protrudin mRNA, respectively (Orengo et al. 2006; Kuroyanagi et al. 2010; Takeuchi et al. 2010). The reporter minigene, designated protrudin splicing reporter (PSR), contains a 1.8-kb fragment of the mouse protrudin gene that extends from the 3′ portion of exon IX to the 5′ region of exon X and thus includes the alternative exon L (Fig. 3A). The sequences of exons L and X in PSR were modified by insertion of two nucleotides (+2) and by substitution to eliminate a generated stop codon (+1), respectively. The genomic fragment was cloned into a vector encoding the fluorescent proteins mVenus (green fluorescence) and mCherry (red fluorescence) in tandem but with different reading frames. Alternative splicing in cells harboring the PSR minigene would be expected to result in the expression of only mVenus when exon L is included in the transcript (PSR-L), whereas only mCherry should be expressed when exon L is excluded from the transcript (PSR-S).

image

Figure 3. Construction of a splicing reporter minigene for protrudin. (A) Schematic representation of the structure of the protrudin splicing reporter (PSR) minigene. A genomic fragment spanning exons IX and X of the mouse protrudin gene is linked to cDNA encoding mVenus and mCherry under the control of a CAGGS/CMV promoter. Exons L and X are modified by insertion of two nucleotides (+2) or by substitution (+1), respectively. Alternative splicing of this minigene gives rise to two transcripts designated PSR-L (including exon L) and PSR-S (excluding exon L). The PSR-L mRNA encodes mVenus in frame followed by a stop codon, whereas mVenus is not produced from PSR-S mRNA as a result of a frame shift. PSR-S mRNA encodes mCherry in frame. The red crosses in the PSR-S mRNA indicate amino acid substitutions introduced to eliminate arginine clusters encoded by the mVenus nucleotide sequence in the PSR-S frame. (B) Expression pattern of the fluorescent proteins encoded by PSR-L and PSR-S in the mouse cell lines Neuro2A and NIH 3T3 transfected with the PSR minigene. Fluorescence emitted by mVenus and mCherry corresponds to the splicing pattern for PSR-L (L) and PSR-S (S), respectively. Scale bars, 10 μm. (C) RT-PCR analysis of total RNA from NIH 3T3 and Neuro2A cells infected with the PSR minigene. PCR was carried out with a primer pair targeted to PSR, with the positions of amplification products corresponding to PSR-L and PSR-S being schematically indicated. The analysis was carried out with or without the RT reaction as well as with a negative control (N.C.). PSR, PSR-S, and PSR-L denote the RT-PCR products from vectors containing indicated genes. Marker lanes contain a 20-bp ladder.

Download figure to PowerPoint

We introduced the PSR construct into Neuro2A cells (which produce both protrudin-L and protrudin-S mRNAs) and into NIH 3T3 cells (which produce protrudin-S mRNA exclusively) and then monitored the cells for green and red fluorescence of the marker proteins. As expected, Neuro2A cells emitted both green and red fluorescence, whereas NIH 3T3 cells emitted only red fluorescence. However, nucleolar aggregation of mCherry produced from the PSR-S transcript was unexpectedly observed, probably as a result of the generation of a nucleolar localization signal consisting of several clusters of arginine residues within the mVenus sequence in the frame of PSR-S. To prevent such aggregation, we substituted nucleotides to replace the arginine clusters with other amino acids in the PSR-S frame, without changing the amino acid sequence of mVenus in the PSR-L frame. This modified reporter system gave rise to a diffuse pattern of mCherry fluorescence throughout the cell without affecting mVenus localization (Fig. 3B). To validate the fidelity of the reporter system, we monitored the spliced mRNAs derived from PSR in Neuro2A and NIH 3T3 cells by RT-PCR analysis (Fig. 3C). The results were consistent with those of the imaging analysis (Fig. 3B) as well as with those of the qRT-PCR analysis of nontransfected cells (Fig. 2C): Both protrudin-L and protrudin-S transcripts were produced in Neuro2A cells, whereas only protrudin-S transcripts were produced in NIH 3T3 cells. The splicing reporter system thus allowed us to visualize the alternative splicing pattern of protrudin mRNA at the single-cell level.

Atypical sequence of the 3′ splice site is essential for selective inclusion of exon L in neuronal cells

The PSR minigene manifested different expression patterns in neuronal versus non-neuronal cells, with exon L being included specifically in neuronal cells. Given that exon L is flanked by an atypical 3′ splice site containing a polypurine tract instead of the conventional polypyrimidine sequence that flanks exon X (Fig. 4A), we examined whether the presence of this atypical splice site is responsible for the selection of exon L in neuronal cells. To this end, we replaced the atypical nucleotides (GGGAG) in this splice site of PSR with the corresponding sequence (TTCTC) of the 3′ splice site that flanks exon X. The resulting construct, designated PSR (EL-S), was introduced into Neuro2A and NIH 3T3 cells, and its splicing pattern was compared with that of wild-type (WT) PSR by fluorescence imaging. Whereas NIH 3T3 cells harboring PSR (WT) manifested only mCherry fluorescence (reflecting protrudin-S–type splicing), those expressing PSR (EL-S) showed both mVenus and mCherry fluorescence (Fig. 4B). In contrast, the pattern of fluorescence did not differ substantially between Neuro2A cells harboring PSR (WT) and those harboring PSR (EL-S), with both constructs conferring expression of both mVenus and mCherry. These results suggested that the atypical 3′ splice site of exon L is responsible for the neuron-specific selection of this exon—or, in other words, this splice site is essential for the skipping of this exon in non-neuronal cells.

image

Figure 4. Atypical sequence of the 3′ splice site is responsible for selective inclusion of exon L of the protrudin gene in neuronal cells. (A) Schematic representation of the 3′ splice site flanking exon L in the PSR minigene. Mismatched nucleotides in the 3′ splice site relative to the conserved consensus sequence are shown in red, and they are replaced in the EL-S construct with the underlined nucleotides corresponding to the typical 3′ splice site flanking exon X. (B) Expression patterns of the wild-type (WT) and mutant (EL-S) versions of the PSR minigene in Neuro2A and NIH 3T3 cells as revealed by fluorescence microscopy. Scale bars, 10 μm.

Download figure to PowerPoint

Alternative splicing pattern of protrudin transcripts in neural-lineage cells

Given that our results suggested that protrudin-L is specifically expressed in neurons, we, furthermore, investigated the expression patterns of protrudin-L and protrudin-S in cells of the neural lineage, including neurons, astrocytes, and oligodendrocytes, with the use of the PSR system. Lentivirus-mediated delivery of the PSR minigene into primary cultured cells derived from embryonic mouse cerebrum revealed that neurons expressed only protrudin-L (Fig. 5A), whereas astrocytes expressed both protrudin-L and protrudin-S (Fig. 5B) and oligodendrocytes expressed only protrudin-S (Fig. 5C). The splicing pattern of protrudin transcripts was thus found to differ between neuronal cells and glial cells as well as between astrocytes and oligodendrocytes.

image

Figure 5. Expression patterns of protrudin-S and protrudin-L in neurons, astrocytes, and oligodendrocytes. Mouse cortical neurons, astrocytes, and oligodendrocytes were infected with a lentiviral vector harboring the PSR minigene and then subjected to immunofluorescence staining for mVenus and either the neuronal marker βIII-tubulin (A), the astrocyte marker GFAP (B), or the oligodendrocyte marker myelin basic protein (MBP) (C). The fluorescence of mCherry was monitored directly. N, neuron; A, astrocyte; O, oligodendrocyte. Scale bars, 20 μm.

Download figure to PowerPoint

VAP interacts preferentially with protrudin-L

We previously showed that protrudin interacts with VAP and KIF5 through its FFAT motif and coiled-coil domain, respectively, both of which are located adjacent to the sequence encoded by exon L (Fig. 6A). Each of these proteins promotes the stimulatory effect of protrudin on neurite extension. The proximity of the binding sites (FFAT motif and coiled-coil domain) to the sequence encoded by exon L suggested that the insertion might affect the binding affinity of protrudin for VAP or KIF5. Co-immunoprecipitation analysis revealed that endogenous VAP-A preferentially with FLAG epitope–tagged protrudin-L compared with protrudin-S (Fig. 6B). Similarly, association of VAP-B with protrudin-L was greater than that with protrudin-S (Fig. 6C), although the difference in the binding to VAP-B was less pronounced compared with that to VAP-A. In contrast, hemagglutinin epitope (HA)–tagged KIF5A interacted with FLAG-tagged protrudin-L and protrudin-S to similar extents (Fig. 6D). These results suggested that the neuron-specific insertion of seven amino acids encoded by exon L increases the binding affinity of protrudin for VAP.

image

Figure 6. Protrudin-L associates with VAP to a greater extent than does protrudin-S. (A) Domain structure of mouse protrudin-S and protrudin-L. The seven amino acids encoded by exon L are indicated by the black box adjacent to the FFAT motif. Arrows indicate the regions that associate with VAP or KIF5. (B, C) Co-immunoprecipitation analysis of FLAG-protrudin and VAP-A or VAP-B. Lysates prepared from HEK293T cells transiently transfected with an expression vector for 3 × FLAG-tagged protrudin-S or protrudin-L were subjected to immunoprecipitation with antibodies to FLAG. The resulting precipitates, as well as a portion (1.25% of the input for IP) of the cell lysates, were subjected to immunoblot analysis with antibodies to FLAG and to HSP90 (loading control) as well as with those to VAP-A (B) or to VAP-B (C). (D) Co-immunoprecipitation analysis of FLAG-protrudin and HA-KIF5A. Lysates prepared from HEK293T cells transiently transfected with expression vectors for 3 × FLAG-tagged protrudin-S or protrudin-L and for 2 × HA-tagged KIF5A were subjected to immunoprecipitation with antibodies to FLAG. The resulting precipitates, as well as a portion (2.5% of the input for IP) of the cell lysates, were subjected to immunoblot analysis with antibodies to HA, to FLAG, and to HSP90.

Download figure to PowerPoint

Protrudin-L promotes polarized neurite outgrowth

Given that protrudin promotes neurite extension, we further investigated the role of protrudin in establishment of the polarity of neurite extension. Dissociation cultures of hippocampal neurons prepared from WT or protrudin-deficient mice were immunostained for the axonal marker Tau1 and the dendrite marker MAP2 (microtubule-associated protein 2) (Fig. 7A). The axon of protrudin-deficient neurons was significantly shorter than that of WT neurons (Fig. 7B). In contrast, the total length of all dendrites for protrudin-deficient neurons was significantly greater than that for WT neurons (Fig. 7C). The axon/dendrite ratio was thus significantly smaller for protrudin-deficient neurons than for WT neurons (Fig. 7D), suggesting that protrudin contributes to the establishment of polarity during neuritogenesis.

image

Figure 7. Protrudin is required for polarized neurite outgrowth. (A) Morphology of primary cultured hippocampal neurons derived from WT or protrudin knockout (KO) mouse embryos. The neurons were cultured on poly-L-lysine–coated plates for 3 days, fixed, and subjected to immunofluorescence staining with antibodies to Tau1 (green) and to MAP2 (red). Scale bars, 100 μm. (B, C). The length of the longest axon (B) or the total length of all dendrites (C) for WT or protrudin-deficient primary neurons. The length of neurites was measured with the use of NeuronJ software. (D) The axon/dendrite ratio of neurons from WT or protrudin-deficient mice. The ratio of maximum axon length to total dendrite length was measured for each neuron. All quantitative data are means ± SEM for a total of 61 neurons examined. *< 0.05, **< 0.01 (Student's t-test).

Download figure to PowerPoint

We next investigated whether protrudin-L and protrudin-S differ in the ability to establish neuronal polarity. Hippocampal neurons from embryos deficient in protrudin were transfected with a vector encoding protrudin-L or protrudin-S. After culture for 1 day, when the ectopic proteins were expressed at a sufficient level, the cells were evaluated for neurite length (Fig. 8A). For some reason, it was technically difficult to culture the primary neurons more than 1 day under this condition. The neurites were categorized on the basis of their length as short (0–50 μm), midlength (51–100 μm), or long (>100 μm), and their length distributions were compared. In comparison with control neurons (those transfected with the empty vector), neurons expressing protrudin-L or protrudin-S showed an increase in the frequency of long neurites, and this increase was significantly greater for neurons expressing protrudin-L than for those expressing protrudin-S (Fig. 8B, C).

image

Figure 8. Expression of protrudin-L promotes polarized neurite outgrowth. (A) Experimental scheme for examination of the ability of protrudin-S or protrudin-L to promote neurite outgrowth. Hippocampal neurons were isolated from protrudin-deficient embryos at embryonic day 18 (E18), transfected with a vector encoding a fusion protein of Venus and either protrudin-S or protrudin-L, and cultured for 24 h before analysis of neurite length with the use of NeuronJ software. (B) Morphology of primary neurons transfected with the Venus vector alone or with the vectors encoding Venus-tagged protrudin-S or protrudin-L. Scale bars, 50 μm. (C) Relative frequency of neurites categorized by length in the transfected neurons. Data are means ± SEM for a total of 474 neurons examined in three independent experiments. *< 0.05 (one-way anova followed by the Tukey–Kramer test).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

We have shown that a novel splicing variant of protrudin, designated protrudin-L, is expressed exclusively in the nervous system. Protrudin-L was found to promote polarized neurite outgrowth and to associate preferentially with VAP (Saita et al. 2009). The 3′ splice site flanking the newly identified exon L of the protrudin gene is atypical in that it contains a polypurine sequence (weak splice site) rather than a polypyrimidine tract (strong splice site). Mutational analysis revealed that this atypical splice site is essential for the neuron-specific inclusion of exon L in the mature transcript. The mechanism underlying this specificity remains unknown, however.

The splicing process is executed by the spliceosome, a ribonucleoprotein megaparticle that assembles around splice sites at each intron. Each splice site consists of a version of the consensus sequence that is recognized by spliceosomal components, with strong splice sites being recognized and used more efficiently than weak ones (Mount 1982). The vicinity of competing strong and weak splice sites along a nascent precursor mRNA gives rise to alternative splicing. In general, strong splice sites are associated with constitutive splicing and full usage of the site (Takeuchi et al. 2010). The extent of usage of weak splice sites varies in a manner dependent on the cellular context.

Alternative splicing plays an important role in the control of neural gene expression (Gehman et al. 2011; Wang et al. 2013). Several key events of neural development, such as cell-fate determination, axon guidance, and synaptogenesis, are controlled by alternative splicing (Doherty et al. 1992; Wang et al. 2013). Development of the vertebrate midbrain depends on fibroblast growth factor 8 (FGF8), and two alternatively spliced isoforms of this protein, one derived from the use of an alternative 3′ splice site in the second exon, have markedly different activities in midbrain patterning (Lee et al. 1997; Liu et al. 1999). Postsynaptic density protein 95 (PSD95) is essential for synaptic maturation and plasticity, and the polypyrimidine tract binding proteins PTB and nPTB regulate alternative splicing of PSD95 mRNA during neural development (Zheng et al. 2012). Many molecules that contribute to neurite outgrowth and axon guidance—including cell adhesion molecules, kinases, specialized cytoskeletal components, and receptors for neurotrophins and guidance cues—are also the products of alternative splicing. The large number of alternative exons of the gene for Down syndrome cell adhesion molecule (DSCAM) underlies the generation of up to 38 016 potential protein isoforms, which represents the highest product diversity of any known gene (Schmucker et al. 2000). It remains unclear, however, how exons are silenced to allow the inclusion of only a single exon in each isoform of this protein. Multiple splice variants of the glutamate receptor subunit NMDAR1 contribute to the control of neuronal plasticity at the level of receptor trafficking (Mu et al. 2003; Perez-Otano & Ehlers 2005).

Changes in splicing pattern are directed by regulatory proteins that bind the precursor mRNA and promote or inhibit particular splicing choices. The most well characterized such positive regulatory factors are the serine-arginine-rich proteins (SR proteins) (Kim et al. 2011). Negative regulatory factors include members of the heterogeneous nuclear ribonucleoprotein (hnRNP) family such as hnRNP A1 and hnRNP I (also known as PTB). Some proteins either enhance exon inclusion or repress it in a manner dependent on the position of their binding sites relative to the target exon (Mauger et al. 2008; Russo et al. 2010). Some splicing factors are expressed exclusively in neurons (Li et al. 2007).

Collectively, our findings indicate that neuron-specific alternative splicing results in the insertion of a seven-amino acid sequence adjacent to the FFAT motif in protrudin. This insertion promotes the association of protrudin with VAP and thereby increases the stimulatory effect of protrudin on polarized neurite extension. Our identification of a neuron-specific splicing isoform of protrudin (protrudin-L) should now provide an impetus to the identification of splicing regulators that recognize the atypical splice site that gives rise to this isoform, as well as possibly to that of regulators responsible for the generation of neuron-specific splicing variants of other proteins.

Experimental procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Construction of expression plasmids

Mouse cDNAs encoding protrudin-S or protrudin-L (GenBank Accession Numbers NM_177319.3 and NM_001164531.1, respectively) were generated by PCR with PrimeSTAR HS DNA polymerase (TaKaRa, Shiga, Japan) and specific primer pairs listed in Table 1 from cDNA preparations of the thymus and cerebrum, respectively, and they were subcloned into the p3 × FLAG-CMV-7.1 (Sigma, St Louis, MO, USA), pVenus-C1 (Clontech, Palo Alto, CA, USA), or pcDNA3 (Invitrogen, Carlsbad, CA, USA) vectors. Construction of vectors encoding mouse KIF5A was described previously (Matsuzaki et al. 2011).

Table 1. Sequences of PCR primers used in the study
For qRT-PCR or RT-PCR
GeneForward primer (5′–3′)Reverse primer (5′–3′)
Protrudin-S AAAGATGCAATTGAGGAGGA TCTTCTTCAGCACGGAGAACG
Protrudin-L AGACCCACCTGGTGGTGCTG ACACACACACAGTCTCTCTC
Tubb3 ATTCTGGTGGACTTGGAACCT ACTCTTTCCGCACGACATCT
GFAP ACAGACTTTCTCCAACCTCCAG CCTTCTGACACGGATTTGGT
Olig2 GGGAGGTCATGCCTTACGC CTCCAGCGAGTTGGTGAGC
GAPDH CATGGCCTTCCGTGTTCCTA GCGGCACGTCAGATCCA
PSR ATGGTGGAGGAAGCTGAGGAG GGGTCCGCTCATCCTCTGC
For the construction of PSR
NameSequence (5′–3′)
PSR-1 TTAGAATTCGCCACCATGGTGGAGGAAGCTGAGGAGGCTG  
PSR-2 CTGCTCCTACTTACCAGCAtaCCACCAGGTGGGTCTCCTG  
PSR-3 CAGGAGACCCACCTGGTGGtaTGCTGGTAAGTAGGAGCAG  
PSR-4 ATAGCGGTCGACGGGTCcGCTCATCCTCTGCTGGGC  
mVenus-F ATACTCCTCGAGGAGTGAGCAAGGGCGAGGAGGAC  
mVenus-R ATAGCGGTCGACTTACTTGTACAGCTCGTCCATGCC  
mCherry-F ATACTCCTCGAGgaGTGAGCAAGGGCGAGGAGGAC  
mCherry-R TTAGGATCCTTACTTGTACAGCTCGTCCATGCC  
For removing mVenus arginine clusters in PSR-S frame
NameSequence (5′–3′)
Mut-mVen-1-F aTTcACaTCcCCaTCCAGCTCGACCAGGATGG  
Mut-mVen-1-R GAtGGgGAtGTgAAtGGCCACAAGTTCAGC  
Mut-mVen-2-F CTCcCCcGACACGCTGAACTTGTGG  
Mut-mVen-2-R AAGTTCAGCGTGTCgGGgGAGGGCGAG  
Mut-mVen-3-F GACaTAGCCcTCcGGCATGGCGGACTTGAAG  
Mut-mVen-3-R CATGCCgGAgGGCTAtGTCCAGGAGCGCAC  
Mut-mVen-4-F GAGCTGCACtCCcCCaTCCTCGATGTTGTG  
Mut-mVen-4-R AACATCGAGGAtGGgGGaGTGCAGCTCGCC  
Mut-mVen-5-F CAGCACcGGtCCGTCcCCGATGGGGGTG  
Mut-mVen-5-R ATCGGgGACGGaCCgGTGCTGCTGCCC  
Mut-mVen-6-F AGTGATCCCcGCcGCcGTCACGAACTCCAG  
Mut-mVen-6-R AGTTCGTGACgGCgGCgGGGATCACTCTCG  
For the construction of EL-S
NameSequence (5′–3′)
EL-S-F

CAGCATACCACCAGGTGGGTCTCCTGCgagaaAGGGTGC

AGGGAAGAAGAAGA

EL-S-R

TTCTTCCCTGCACCCTttctcGCAGGAGACCCACCTGGT

GGTATGCTGGTAAG

Construction of the protrudin splicing vector (PSR)

A genomic region spanning exons IX to X of the mouse protrudin gene was amplified by PCR and cloned into pCAGGS or pCSII-CMV vectors harboring both mVenus and mCherry cDNAs in different reading frames (Fig. 3A). Mutation of the mVenus nucleotide sequence in the PSR-S frame as well as of the 3′ splice site flanking exon L was achieved by PCR. All PCR primers are listed in Table 1.

Cell isolation, culture, transfection, and infection

HEK293T and Neuro2A cells were cultured under a humidified atmosphere of 5% CO2 at 37 °C in Dulbecco's modified Eagle's medium (DMEM, Invitrogen) supplemented with 10% fetal bovine serum (Invitrogen). NIH 3T3 cells were cultured under a humidified atmosphere of 5% CO2 at 37 °C in DMEM supplemented with 10% calf serum (Invitrogen). Neurons, astrocytes, and oligodendrocytes were isolated from the hippocampus or cerebral cortex of C57BL/6J mouse or protrudin knockout mouse (M.S. and K.I.N., manuscript in preparation) embryos at embryonic day 18 (E18) with the use of Nerve-Cell Culture System/Dissociation Solutions (Sumitomo Bakelite, Tokyo, Japan), and they were cultured in Neuron culture medium (Sumitomo Bakelite) at a density of 3 × 105 or 5 × 104 cells per well in 24-well plates coated with poly-l-lysine. Primary culture of neurons without other cell type were established with the presence of 2.5 μm Ara-C. Primary cultured mouse neurons were transfected with the use of anAmaxaNucleofector instrument (program O-5) and an Amaxa Mouse Neuron Nucleofector Kit (Lonza, Cologne, Germany). NIH 3T3 cells were transfected with the use of an AmaxaNucleofector instrument (program A-24) and an Amaxa Cell Line Nucleofector Kit R (Lonza, Walkersville, MD, USA). Other cell types were transfected with the use of the FuGene HD reagent (Promega, Fitchburg, WI, USA). For lentiviral infection, HEK293T cells were transiently transfected with pCAG-HIV gp, pCMV-VSVG-RSV Rev, and pCS2-CMV-based vectors by the calcium phosphate method and then cultured for 48 h. The lentiviruses released into the culture medium were then used to infect neurons, astrocytes, and oligodendrocytes.

Quantitative RT-PCR analysis

Total RNA was isolated from cell lines or from tissues of 12-week-old C57BL/6J mice with the use of an RNeasy Mini Kit (Qiagen, Tokyo, Japan) or an RNeasy Lipid Tissue Mini Kit (Qiagen), respectively. Portions of the RNA were subjected to RT with the use of a Quantitect Kit (Qiagen), and the resulting cDNA was subjected to real-time PCR analysis with the use of a StepOnePlus real-time PCR system (Applied Biosystems, Foster City, CA, USA) and SYBR Premix Ex Taq (TaKaRa). Primers for PCR are listed in Table 1.

RT-PCR analysis

Total RNA was isolated from Neuro2A or NIH 3T3 cells harboring the PSR minigene with the use of Isogen (Nippon Gene, Tokyo, Japan). Portions of the RNA were subjected to RT with a Quantitect Kit, and the resulting cDNA was subjected to PCR with Taq polymerase (TaKaRa) and specific primers listed in Table 1.

Antibodies

Antibodies to protrudin were generated as described previously (Shirane & Nakayama 2006). Antibodies to FLAG (M2 and polyclonal), to βIII-tubulin, and to VAP-B were obtained from Sigma; those to HA (HA.11) were from Covance (Princeton, NJ, USA); those to green fluorescent protein (for detection of mVenus) were from Frontier Science (Hokkaido, Japan); those to GFAP were from Dako (Glostrup, Denmark); and those to MBP and MAP2 were from Abcam (Cambridge, MA, USA); those to HSP90 and VAP-A were from BD Biosciences (San Jose, CA, USA); those to Tau1 were from Millipore (Billerica, MA, USA) Alexa Fluor 488-, Alexa Fluor 546-, or Alexa Fluor 633-conjugated goat antibodies to mouse, rabbit, or rat immunoglobulin G were obtained from Molecular Probes (Eugene, OR, USA). Horseradish peroxidase–conjugated goat antibodies to mouse or rabbit immunoglobulin G were from Promega (Madison, WI, USA).

Immunoprecipitation and immunoblot analysis

Cell lines cultured for 1 day after transfection or tissues from 12-week-old C57BL/6J mice were homogenized in a solution (Triton lysis buffer) containing 40 mm HEPES-NaOH (pH 7.5), 150 mm NaCl, 0.5% Triton X-100, 10 mm MgCl2, 1 mm Na3VO4, 25 mm NaF, 1 mm EDTA, 1 mm phenylmethylsulfonyl fluoride, leupeptin (10 μg/mL), and 10% glycerol. After incubation for 10 min on ice, each homogenate was centrifuged at 20 400 g for 10 min at 4 °C, and the protein concentration of the resulting supernatant was determined with the Bradford assay (Bio-Rad, Hercules, CA, USA). The cell and tissue extracts were then subjected to SDS-polyacrylamide gel electrophoresis, and the separated proteins were transferred to an Immobilon-P membrane (Millipore) and probed with antibodies. Immune complexes were detected with the Super Signal reagent (Pierce, Rockford, IL, USA). Chemiluminescence images were scanned with an LAS-4000 instrument (GE Healthcare, Waukesha, WI, USA). For immunoprecipitation, cell extracts or tissue extracts were incubated with both antibodies and protein G-Sepharose 4 Fast Flow (Amersham Biosciences, Uppsala, Sweden).

Immunostaining, fluorescence microscopy, and quantitation of neurite length

Cells were fixed for 10 min at 37 °C with 4% paraformaldehyde in phosphate-buffered saline, incubated consecutively with primary antibodies and Alexa Fluor 488-, Alexa Fluor 546-, or Alexa Fluor 633- labeled secondary antibodies in phosphate-buffered saline containing 0.5% Triton X-100 and covered with a drop of GEL/MOUNT (Biomeda, Hayward, CA, USA) and examined with a confocal fluorescence microscope (LSM510 META; Carl Zeiss, Inc.) with 63 × /1.4 oil or a fluorescence microscope (Olympus BX51; Olympus) with 20×. The length of neurites was quantitated with the use of NeuronJ software (http://www.imagescience.org/meijering/software/neuronj/).

Statistical analysis

Quantitative data are presented as means ± SEM and were analyzed with Student's or Welch's t-test, or by one-way analysis of variance (anova) followed by the Tukey–Kramer multiple-comparison test. A P value of less than 0.05 was considered statistically significant.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

We thank S. Nagata (Kyoto University) and H. Sumimoto (Kyushu University) for providing the pEFBOS-HHg vector; A. Miyawaki (RIKEN, BSI) for providing Venus cDNA; H. Miyoshi (RIKEN, BRC) for providing the CSII-CMV-MCS vector; as well as A. Hamasaki and other laboratory members for technical assistance. This study was supported in part by a Japan Society for the Promotion of Science KAKENHI Grant-in-Aid for Young Scientists (S) to M.S.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information
FilenameFormatSizeDescription
gtc12109-sup-0001-figS1.pdfapplication/PDF408K

Figure S1 Characterization of PCR primer pairs for detection of protrudin-S and -L mRNAs.

Figure S2 Removal of arginine clusters encoded by the unmodified mVenus nucleotide sequence in the PSR-S frame of the PSR construct.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.