The Calcium Channel Subunit Alpha2delta2 Suppresses Axon Regeneration in the Adult CNS

Article

The Calcium Channel Subunit Alpha2delta2 Suppresses Axon Regeneration in the Adult CNS Highlights d

Large-scale transcriptome comparison identifies a novel regulator of axon growth

d

Alpha2delta2 functions as a developmental switch that limits axon growth

Authors Andrea Tedeschi, Sebastian Dupraz, Claudia J. Laskowski, ..., Marc Beyer, Joachim L. Schultze, Frank Bradke

Correspondence

d

Alpha2delta2 inhibits axon growth via calcium influx through Cav2 channels

[email protected]

d

Systemic PGB administration induces axon regeneration in the adult CNS

Using a multidisciplinary strategy that combines transcriptome profiling and genetic, molecular, and pharmacological approaches together with cutting-edge in vivo two-photon microscopy, Tedeschi et al. show that Cacna2d2 functions as a developmental switch that limits regenerative ability in the adult CNS.

In Brief

Accession Numbers GSE66128

Tedeschi et al., 2016, Neuron 92, 1–16 October 19, 2016 ª 2016 Elsevier Inc. http://dx.doi.org/10.1016/j.neuron.2016.09.026

Please cite this article in press as: Tedeschi et al., The Calcium Channel Subunit Alpha2delta2 Suppresses Axon Regeneration in the Adult CNS, Neuron (2016), http://dx.doi.org/10.1016/j.neuron.2016.09.026

Neuron

Article The Calcium Channel Subunit Alpha2delta2 Suppresses Axon Regeneration in the Adult CNS Andrea Tedeschi,1 Sebastian Dupraz,1 Claudia J. Laskowski,1 Jia Xue,2 Thomas Ulas,2 Marc Beyer,2 Joachim L. Schultze,2,3 and Frank Bradke1,4,* 1Axonal

Growth and Regeneration, German Center for Neurodegenerative Diseases, 53175 Bonn, Germany and Immunoregulation, LIMES-Institute, University of Bonn, 53115 Bonn, Germany 3Platform for Single Cell Genomics and Epigenomics, German Center for Neurodegenerative Diseases, 53175 Bonn, Germany 4Lead Contact *Correspondence: [email protected] http://dx.doi.org/10.1016/j.neuron.2016.09.026 2Genomics

SUMMARY

Injuries to the adult CNS often result in permanent disabilities because neurons lose the ability to regenerate their axon during development. Here, whole transcriptome sequencing and bioinformatics analysis followed by gain- and loss-of-function experiments identified Cacna2d2, the gene encoding the Alpha2delta2 subunit of voltage-gated calcium channels (VGCCs), as a developmental switch that limits axon growth and regeneration. Cacna2d2 gene deletion or silencing promoted axon growth in vitro. In vivo, Alpha2delta2 pharmacological blockade through Pregabalin (PGB) administration enhanced axon regeneration in adult mice after spinal cord injury (SCI). As PGB is already an established treatment for a wide range of neurological disorders, our findings suggest that targeting Alpha2delta2 may be a novel treatment strategy to promote structural plasticity and regeneration following CNS trauma. INTRODUCTION Lesioned axons fail to regenerate in the adult mammalian CNS, limiting recovery after injury. Both a non-permissive environment and a reduced intrinsic growth ability account for this regenerative failure (Liu et al., 2011). While several extracellular growth inhibitors expressed in the adult CNS have been characterized (Cregg et al., 2014; Schwab and Strittmatter, 2014), the molecular mechanisms underlying changes in axon growth ability are largely unclear. A few classes of adult neurons regenerate their axons under specific conditions. Activation of a transcriptional program after peripheral nerve lesion (PNL) allows dorsal root ganglion (DRG) neurons to mount a robust regenerative response to a second lesion of either the peripheral or central axon, a phenomenon called ‘‘conditioning’’ (Cho et al., 2013; Finelli et al., 2013; Neumann and Woolf, 1999; Puttagunta et al., 2014; Richardson and Issa, 1984; Ylera et al., 2009). Lens injury enables retinal ganglion neurons to regenerate their axons after optic nerve damage (Leon et al., 2000). Transcriptional analysis in these

neurons has identified a number of regeneration-associated genes, such as Arginase 1 (Arg1), cytoskeleton-associated protein 23 (Cap23), galanin (Gal), growth-associated protein 43 (Gap43), small proline-rich protein 1A (Sprr1a), tubulin alpha 1A (Tuba1a), and several pro-regenerative transcription factors, including activating transcription factor 3 (Atf3), Jun proto-oncogene (c-Jun), Kru¨ppel-like factor (Klf) family members, tumor protein 53 (p53), Smad family member 1 (Smad1), SRY (sex determining region Y)-box 11 (Sox11), and signal transducer and activator of transcription 3 (Stat3) (Chandran et al., 2016; Cho et al., 2015; Costigan et al., 2002; Omura et al., 2015; Tedeschi, 2012). However, our understanding of the molecular mechanisms that lead to neurons losing their ability to grow is still fragmentary. During development axons cease to grow as neurons form synapses and integrate into neuronal circuits (Hall and Sanes, 1993; Shewan et al., 1995). Therefore, the developmental transition from a growing to a transmitting phase may represent one of the first steps in the loss of axon growth and regeneration ability (Enes et al., 2010; Wang et al., 2007). A putative growth-inhibiting switch regulating this process has yet to be identified. We hypothesized that growth-inhibiting genes restraining axon growth at late stages of embryonic development impair regeneration in the adult CNS. Hence, their expression would not only positively correlate with the loss of axon growth ability during late embryonic development, but would also negatively correlate with gaining growth competence in the adult. To search for these genes, we sequenced the whole transcriptome of DRG neurons in both growth competent and incompetent states at different developmental stages, in diverse culture and in vivo experimental conditions. Transcriptome sequencing and bioinformatics analysis, followed by gain- and loss-of-function experiments, revealed that the voltage-gated calcium channels (VGCCs) subunit Alpha2delta2 (Cacna2d2) restrains axon growth at late stages of embryonic development and impairs axon regeneration in the adult CNS. Moreover, Alpha2delta2 blockade through Pregabalin (PGB), a drug used to treat neurological disorders, promoted axon regeneration in adult mice after spinal cord injury (SCI). Thus, our study provides insight into the molecular mechanisms underlying loss of axon growth and regenerative ability, with implications for repair after CNS trauma including SCI. Neuron 92, 1–16, October 19, 2016 ª 2016 Elsevier Inc. 1

Please cite this article in press as: Tedeschi et al., The Calcium Channel Subunit Alpha2delta2 Suppresses Axon Regeneration in the Adult CNS, Neuron (2016), http://dx.doi.org/10.1016/j.neuron.2016.09.026

Figure 1. Axon Growth Ability of Developing, Adult, and Regenerating DRG Neurons

A

B

C

D

E

F

G

RESULTS RNA-Sequencing Identifies Cacna2d2 as a Potential Negative Regulator of Axon Growth and Regeneration We began our search for potential negative regulators of axon growth and regeneration by sequencing the whole transcriptome of DRG neurons in three experimental paradigms that represent different stages of axon growth: a developmental transition from axon growth to synapse formation, an in vitro shift from arborizing to elongating growth, and regeneration following a conditioning lesion (Figures 1A and S1A). For each paradigm, changes in axon growth competence were validated in cell culture. In paradigm 1, axon length decreased and branching increased in cultured embryonic day (E) 17.5 DRG neurons, which have already formed synapses in vivo (Prasad and Weiner, 2011) compared to those at E12.5 that have not yet undergone synaptogenesis (Figures 1B and 1C). In paradigm 2, cultured adult DRG neurons regained axon growth competence within 24 hr af-

2 Neuron 92, 1–16, October 19, 2016

(A) Schematic for the three experimental paradigms. (B) Representative fluorescence images of E12.5 and E17.5 DRG neurons cultured for 12 hr and stained with a Tuj1 antibody. (C) Quantification of (B). (D) Representative fluorescence images of adult DRG neurons cultured for 12 or 24 hr and stained with a Tuj1 antibody. (E) Quantification of (D). (F) Representative fluorescence images of sham and preconditioned (PNL) adult DRG neurons cultured for 12 hr and stained with a Tuj1 antibody. (G) Quantification of (F). All values are plotted as mean and SEM (*p < 0.05, **p < 0.01, and ***p < 0.001, triplicate experiments, >100 neurons per condition). Scale bars, 200 mm.

ter plating (Figures 1D and 1E) (Smith and Skene, 1997). In paradigm 3, conditioned adult DRG neurons extended long and sparsely branched processes 12 hr after plating compared to their unconditioned counterparts (Figures 1F and 1G) (Smith and Skene, 1997; Ylera et al., 2009). We analyzed the whole transcriptome using two independent approaches that, when combined, allowed gene prioritization (Figure S1A). The changes in paradigm 1 were more prominent than those found in paradigms 2 and 3. We therefore calculated the mean fold change differences across all transcripts, the mean effect size, and the sum effect size for each of the paradigms. This allowed for accurately adjusting the cut-off for each paradigm (Figures S1A and S1B). Unbiased hierarchical clustering of the most variable genes within the dataset (false discovery rate [FDR], p < 3 3 10-10, n = 1,560) (Figure S1A) identified a cluster of genes (cluster 1 n = 141) whose expression correlated both positively, with loss of axon growth and increased branching during later stages of embryonic development, and negatively, with gaining growth competence in the adult (Figure 2A). Two small additional clusters (cluster 2 n = 14; cluster 3 n = 64) with a similar pattern were also identified (Figure 2A). While having a proper patterning in paradigms 2 and 3, these additional clusters had only modest changes in paradigm 1. Gene ontology (GO) analysis, followed by network visualization of gene cluster 1, revealed a major hub associated with regulation of synaptic transmission, synaptic plasticity, and neurogenesis (Figure 2B). Other hubs related to nervous system development, cellular ion homeostasis, and transmembrane transport (Figure 2B and Table S1). It is important to note that inclusion of gene clusters 2 and 3 did not enrich the link to prior biological information (data not shown), further suggesting that

Please cite this article in press as: Tedeschi et al., The Calcium Channel Subunit Alpha2delta2 Suppresses Axon Regeneration in the Adult CNS, Neuron (2016), http://dx.doi.org/10.1016/j.neuron.2016.09.026

clusters 2 and 3 may not represent a biologically meaningful group. A statistical one-way ANOVA model followed by a defined filtering strategy (Figure S1A) identified 15 candidates by filtering for genes that were at least 2.5-fold upregulated in E17.5 compared to E12.5 DRGs and substantially downregulated in cultured adult neurons 24 hr after plating (at least 1.5-fold) and following a conditioning lesion (at least 1.4-fold; Figures S1A– S1C). To further narrow down to potential suppressors of axon growth and regeneration, we intersected both gene lists, resulting in five differentially expressed genes: Cacna2d2, docking protein 4 (Dok4), melanoma antigen family E, 2 (Magee2), sex comb on midleg-like 4 (Scml4), and teneurin transmembrane protein 1 (Tenm1) (Figure 2C). Extending the list of genes by including clusters 2 and 3 did not lead to the identification of additional candidate genes for further experimental analysis. Differential gene expression was validated using qRT-PCR. Except Magee2, the candidates had clear differential expression in all three paradigms (Figure S1D). The proteins encoded by these four genes have been functionally characterized. Cacna2d2 encodes the Alpha2delta2 subunit of VGCCs, a largely extracellular, membrane-associated protein (Bauer et al., 2010; Dolphin, 2012) that regulates VGCC density and vesicle release probability (Hoppa et al., 2012). Interestingly, Cacna2d2 expression pattern across all three paradigms was not mirrored by other subunits of VGCCs including Cacna2d1 and Cacna2d3 (Figure S1E). Dok4 encodes for a scaffolding protein with a role in the assembly of signaling complexes (Grimm et al., 2001). Scml4 encodes a putative polycomb group protein, a component of multiprotein complexes involved in transcriptional repression (Bornemann et al., 1998). Tenm1 encodes a transmembrane protein with a role in signal transduction (Minet et al., 1999). As we hypothesized that developing neurons lose axon growth ability while switching to a transmitting phase, and Cacna2d2 is the only gene among the four candidates suggested by our GO analysis to play a direct role in synaptic transmission (Table S2), we explored the relationship between Cacna2d2 expression and axon growth ability. We generated a polyclonal antibody that specifically recognizes murine Alpha2delta2 (Figure S2). Both immunofluorescence microscopy and immunoblot analysis revealed that Alpha2delta2 is upregulated in E17.5 when compared to E12.5 DRGs (Figures 2D–2F). Conversely, while Alpha2delta2 was expressed at the soma and along the axons (Figures S2F and S2G), its expression diminished as adult DRG neurons regained growth competence in culture (Figures 2G–2I). Cacna2d2 expression was reduced in DRGs after peripheral lesion, which promotes regeneration, but not after central branch lesion, when regeneration fails (Figures S3A–S3D). Thus, Alpha2delta2 expression inversely correlates with axon growth and regeneration. Cacna2d2 Inhibits Axon Growth and Peripheral Nerve Regeneration To determine whether Cacna2d2 suppresses axon growth, we first overexpressed Cacna2d2 by electroporating dissociated adult lumbar (L) DRG neurons with a mixture of GFP- and either control- (CTR) or Cacna2d2-expressing plasmids. Without affecting cell viability (Figures S3E–S3G), Cacna2d2 overexpression impaired axon elongation (Figures 3A and 3B). Second,

Cacna2d2 overexpression also restrained peripheral nerve regeneration. L5 adult DRGs were electroporated in vivo using GFP with either control or Cacna2d2-expressing plasmids with high percentage of co-expression (78.2% ± 3.6%; Figures S3H–S3M). Five days later, the sciatic nerves were crushed unilaterally, allowing to accurately and persistently mark the crush site (Figures S3N and S3O). Peripheral DRG axons robustly regenerated 4 days after injury in control animals (Figures 3C and 3D), as previously described (Shin et al., 2012). By contrast, peripheral axons overexpressing Cacna2d2 failed to regenerate over long distances (Figures 3C and 3D). Axon fragments observed distal to the crush site upon Cacna2d2 overexpression represented degenerating axons showing typical hallmarks of degeneration, i.e., discontinuous, dot-like structures along the axon (Figure 3C) (Ertu¨rk et al., 2007). Third, Cacna2d2 overexpression impaired the conditioning response. L5 adult DRGs were electroporated in vivo and 5 days later a PNL was performed. Seven days after PNL, the DRGs were dissected, dissociated, and plated. Cacna2d2-overexpressing neurons showed a drastic reduction in axon length and increased branching when compared to the control condition (Figures 3E and 3F). Next, we addressed whether Cacna2d2 gene deletion or silencing would be sufficient to drive axon growth in adult DRG neurons. The ducky-2J mutant mouse line contains a 2 bp deletion in exon 9 of the Cacna2d2 gene that causes a frameshift generating a non-functional protein (Donato et al., 2006). Mutant mice die as early as 4–5 weeks of age (Donato et al., 2006). Cultured mutant DRG neurons showed increased axonal length and diminished branching when compared to those originating from control littermates (Figures 3G and 3H). Similarly, downregulation of Cacna2d2 expression (Figure S4A) also promoted axon growth. The expression of the genes encoding other Alpha2delta subunits expressed in adult DRGs (Figure S4B) was not affected by Cacna2d2 silencing (Figure S4C). When overexpressed in adult DRG neurons, Cacna2d1 and Cacna2d3 subunits slightly inhibited axon growth without affecting branching (Figures S4D and S4E). In Cacna2d2-silenced neurons, axon length increased while branching diminished when compared to the control condition (Figure S4F). Increased axon growth was not due to off-target effects because transfecting a siRNA-resistant plasmid DNA expressing Alpha2delta2 restored the control phenotype (Figures S4F and S4G). Taken together, our data show that Cacna2d2 inhibits axon growth and regeneration. Cacna2d2 Suppresses Axon Growth via Calcium Influx through Cav2 Channels Recent findings have shown that Alpha2delta subunits are required for trafficking of Cav2 channels and that enhanced expression of the Alpha2delta2 subunit increases Cav2 channel density at the presynaptic active zone (Hoppa et al., 2012). Hence, we asked whether Cacna2d2 could inhibit axon growth by a similar mechanism (Figure 4A). Both immunoblot analysis and immunofluorescence microscopy showed that Cacna2d2 overexpression led to an increase in Cav2.1 channel expression in adult DRG neurons (Figures 4B–4D). Removal of extracellular calcium rescued axon growth defects as Cacna2d2-overexpressing adult DRG neurons cultured for 24 hr in the presence of the calcium chelator EGTA (0.5 mM) extended long and

Neuron 92, 1–16, October 19, 2016 3

Please cite this article in press as: Tedeschi et al., The Calcium Channel Subunit Alpha2delta2 Suppresses Axon Regeneration in the Adult CNS, Neuron (2016), http://dx.doi.org/10.1016/j.neuron.2016.09.026

A

C

B

D

E

F

G

H

I

Figure 2. RNA-Seq Identifies Cacna2d2 as Potential Negative Regulator of Axon Growth (A) Hierarchical clustering based on the most variable genes within the dataset (FDR, p < 3 3 10-10, n = 1,560, z transformed blue to red). The black box outlines a gene cluster (cluster 1) that met our search criteria. The gray boxes outline two small gene clusters with a proper patterning in paradigms 2 and 3 but only modest changes in paradigm 1. Each column represents a single sample, each row a single gene (n = 3 independent biological replicates per condition). (B) Network visualization of GO enrichment analysis using BiNGO and EnrichmentMap for the 141 genes originating from the gene cluster 1 in (A). Red nodes depict enriched GO terms, where color and size represent the corresponding FDR-adjusted enrichment p value (q value). Network edges represent mutual overlap between two GO terms. (legend continued on next page)

4 Neuron 92, 1–16, October 19, 2016

Please cite this article in press as: Tedeschi et al., The Calcium Channel Subunit Alpha2delta2 Suppresses Axon Regeneration in the Adult CNS, Neuron (2016), http://dx.doi.org/10.1016/j.neuron.2016.09.026

sparsely branched axons (Figure 4E). Similar results were obtained when Cacna2d2-overexpressing neurons were cultured in the presence of N- and P/Q-type channel blockers (Figure 4E). Conversely, treating control-electroporated DRG neurons with P/Q-type channel agonists or Ruthenium Red, a drug that stimulates exocytosis independently of calcium influx, led to short and arborized axons mimicking the effect of Cacna2d2 overexpression (Figure 4F). Clinically approved gabapentinoids, such as gabapentin and PGB, bind with high affinity and selectivity to Alpha2delta1/2 subunits (Gee et al., 1996; Gong et al., 2001). Therefore, we tested whether PGB administration could counteract axon growth inhibition induced by Cacna2d2 overexpression. Adult DRG neurons were electroporated and cultured for 24 hr in the presence of either vehicle or PGB (250 mM). Indeed, PGB restored elongating growth in Cacna2d2-overexpressing neurons, resulting in a morphology similar to that seen in control conditions (Figures 4G and 4H). The PGB action was mediated through binding of Alpha2delta2, as PGB failed to rescue axon growth defects in neurons overexpressing Alpha2delta2R282A, a mutant that does not bind PGB (Davies et al., 2006) (Figures 4G and 4H). We further tested whether PGB increases axon growth under normal conditions, rather than just rescuing deficit upon Cacna2d2 overexpression. Indeed, dissociated adult DRG neurons treated with PGB immediately after plating extended longer and sparsely branched axons compared to control neurons on a permissive substrate (Figures S4H and S4I) or growth inhibitory chondroitin sulfate proteoglycans (CSPGs) (Figure S4J). Altogether, these data suggest that Alpha2delta2 inhibits axon growth via calcium influx through N- and P/Qtype channels. Systemic PGB Administration Counteracts the Developmental Decline in Axon Growth Potential Associated with Presynaptic Differentiation Alpha2delta subunits regulate the differentiation of presynaptic terminals (Dickman et al., 2008; Kurshan et al., 2009) and promote synapse formation both in vitro and in vivo (Eroglu et al., 2009). Thus, we wondered whether systemic administration of PGB might counteract the developmental decline in axon growth ability associated with presynaptic terminal differentiation. Pregnant mice carrying E12.5 embryos received systemic administration of either vehicle or PGB three times a day. At E17.5, we analyzed the formation of primary (Ia) afferent terminals in the spinal cord. Parvalbumin- (Pv) positive Ia afferents connect to motorneuron (MN) pools in the ventral horn of the spinal cord during late embryonic development (Figure 5A) (Prasad and Weiner, 2011). Neuronal survival and differentiation of Pv-

positive DRG neurons expressing Alpha2delta2 (Figure S5A) was comparable between treatments (Figures S5B and S5C). Central Ia terminals developed a fine net-like pattern around the MN pool in the ventral horn of the spinal cord in control embryos (Figures 5B and S5D). By contrast, terminals expanded over a larger area in PGB-treated embryos (Figures 5B, 5C, and S5D). Moreover, Ia afferents failed to differentiate presynaptic terminals after systemically administering PGB (Figures 5D and 5E). A similar failure in terminal differentiation was also found in Cacna2d2 knockout (KO) embryos (Figure 5F). These results raise the possibility that PGB treatment keeps neurons in an active growth state. Consistent with this idea, cultured E17.5 DRG neurons derived from PGB-treated embryos extended longer and less branched neurites than those from vehicletreated embryos (Figures 5G and 5H). Consistently, an increase in axon length and decrease in branching was also observed in Cacna2d2 KO (Figure S5E) or silenced (Figures S5F–S5H) E17.5 DRG neurons when compared to the control condition. Thus, PGB chronic administration impairs presynaptic terminal differentiation during late embryonic development, and keeps DRG neurons in an elongating growth mode. Systemic PGB Administration Weakens Synaptic Transmission while Promoting Axon Growth To evaluate whether chronic PGB treatment weakens synaptic transmission between DRG neurons and their target cells in the adult spinal cord, age- and sex-matched adult mice received systemic administration of either vehicle (0.9% saline) or PGB (46 mg/kg body weight) three times/day. Seven days after the first injection, the left sciatic nerve was electrically stimulated at 5 Hz for 5 min. Such a stimulation triggers extracellularsignal-regulated kinase (ERK) activation in dorsal horn neurons of the ipsilateral spinal cord (Figures S6A and S6B) (Ji et al., 1999). Chronic PGB administration reduced ERK activation, suggesting that PGB dampens synaptic transmission in vivo (Figures S6C and S6D), as has been demonstrated between DRG and dorsal horn neurons in a co-culture system (Hendrich et al., 2012). Notably, the density of Synaptophysin (Syn)positive presynaptic terminals in the dorsal horn of the spinal cord was not altered by the treatment (Figure S6E). In addition, PGB chronic administration resulted in a diminished response to light touch (Figures S6F and S6G), further supporting the conclusion that PGB impairs synaptic functions in vivo. We then investigated whether chronic PGB administration leads to changes in axon growth ability in adult neurons. Adult mice received either PGB or vehicle for 7 days, and subsequently L4-5 DRGs were dissected, dissociated, and plated. DRG

(C) Normalized mean expression values of the five candidate genes identified after intersection of the 141 genes derived from the analysis of the most variable genes and the 15 genes derived from the ANOVA model. (D) Cryosections of E12.5 and E17.5 DRGs immunostained with Alpha2delta2 and neurofilament 160 antibodies (E12.5 n = 6 and E17.5 n = 6 embryos; 3–4 DRGs/embryo). Inset, higher-magnification view of a region in the main panel. DAPI-stained nuclei (blue). (E) Immunoblot shows Alpha2delta2 expression in E12.5 and E17.5 DRG. Tuj1 is shown as a loading control (n = 3 independent replicates per condition). (F) Quantification of (D) and (E). Mean and SEM (**p < 0.01 and ***p < 0.001; triplicate experiments). (G) Adult DRG neurons were cultured for 12 or 24 hr and stained with Alpha2delta2 and Tuj1 antibodies. (H) Immunoblot shows Alpha2delta2 expression in adult DRG neurons at 12 and 24 hr after plating (n = 3 independent replicates per condition). Tuj1 is shown as a loading control. (I) Quantification of (G) and (H). Mean and SEM (***p < 0.001; triplicate experiments). Scale bars, 50 mm. See also Figures S1 and S2; Tables S1 and S2.

Neuron 92, 1–16, October 19, 2016 5

Please cite this article in press as: Tedeschi et al., The Calcium Channel Subunit Alpha2delta2 Suppresses Axon Regeneration in the Adult CNS, Neuron (2016), http://dx.doi.org/10.1016/j.neuron.2016.09.026

A

B

C

D

E

F

G

H

Figure 3. Cacna2d2 Inhibits Axon Growth and PNS Regeneration (A) Representative fluorescence images of adult DRG neurons cultured for 24 hr after electroporation with GFP- plus either CTR- or Cacna2d2-expressing plasmids. Scale bar, 200 mm. (B) Quantification of (A). Mean and SEM (***p < 0.001, triplicate experiments, 100–110 neurons per condition). (C) Tile-scanned confocal images of longitudinal sections from sciatic nerves 4 days after crush injury. Five days before injury, L5 DRGs were electroporated in vivo with GFP- plus either CTR- or Cacna2d2-expressing plasmid DNA. Red arrowheads indicate regenerating axons. Empty arrowheads indicate degenerating axons. Yellow asterisks indicate the injury site. Scale bar, 1 mm. (D) Quantification of (C). Scatterplot with mean (***p < 0.001; CTR n = 5 and Cacna2d2 n = 6 animals; 8–10 axons/animal). (E) Representative fluorescence images of cultured adult DRG neurons after in vivo electroporation (as for panel C) followed by a PNL. Seven days after PNL, L5 DRG neurons were cultured for 12 hr and stained using a Tuj1 antibody. Scale bar, 200 mm. (legend continued on next page)

6 Neuron 92, 1–16, October 19, 2016

Please cite this article in press as: Tedeschi et al., The Calcium Channel Subunit Alpha2delta2 Suppresses Axon Regeneration in the Adult CNS, Neuron (2016), http://dx.doi.org/10.1016/j.neuron.2016.09.026

neurons derived from PGB-treated animals extended longer and less branched neurites 12 hr after plating (Figures 6A and 6B), a phenotype markedly similar to conditioned neurons. Thus, PGB administration weakens synaptic transmission in vivo and returns adult DRG neurons to a growth-competent state. Systemic Administration of PGB after SCI Promotes Axon Regeneration In Vivo To test whether systemic PGB administration can promote regeneration after injury, we imaged central DRG axons, also known as dorsal column (DC) axons, with time-lapse twophoton microscopy. Adult transgenic mice expressing GFP in a small number of DRG neurons were injected with either vehicle or PGB as described above. One week after the first injection, a chronic spinal window was implanted at thoracic (T) 12 level (Figures S6H–S6J) and DC axons were crushed. Animals were imaged daily over 3 days. While the majority of DC axons degenerated in control animals, numerous axons reached the site of injury and several had already grown beyond the lesion 3 days after injury in PGB-treated animals (Figures 6C and 6D). We then determined whether sustained PGB administration, beginning 1 hr after injury, promotes axon regeneration in more severe models of SCI that completely transect DC axons (Figures S7A–S7E). To visualize DC axons in the spinal cord, we injected adeno-associated virus (AAV) expressing enhanced GFP (eGFP) into the left sciatic nerve to transduce L4-6 DRGs. Using two-photon imaging of the unsectioned adult spinal cord (Figures S7B and S7C), we observed that PGB administration promotes regeneration of DC axons across the lesion comparably to a conditioning lesion (Figures 7A–7C and S7F). Similar results were also obtained in a blind replication study in which AAV injection was performed 2 weeks before termination to exclude the possibility that nerve manipulation before SCI would promote regeneration (Figure 7D). PGB administration also induced regeneration of DC axons in a more severe SCI model, the dorsal hemisection (Figures S7G and S7H). After crossing the site of the lesion, regenerating axons exhibited tortuous trajectories in PGB-treated animals (Movie S1). PGB treatment had no obvious effect on scarring (Figures 7E and 7F). PGB administration starting 2 weeks after SCI also promoted some degree of sensory axon regeneration (Figures 7G–7I), albeit less prominent when compared to animals that were injected starting 1 hr after injury. Collectively, these data show that PGB administration after SCI enhances axon regeneration. DISCUSSION Axons stop growing after reaching target areas. The developmental decline in axon growth and regeneration ability coincides with presynaptic terminal differentiation and synapse formation. Here we provide evidence that manipulation of a developmental switch centered on Alpha2delta2, identified using a systematic

and unbiased approach, promotes axon growth and regeneration in the adult CNS. A Molecular Signature Arising in Response to Changes of Intrinsic Neuronal Properties Our transcriptome profiling of DRG neurons demonstrated that the transition from a growth to a transmitting phase correlates with profound changes in intrinsic neuronal properties. Conversely, such changes in adult DRG neurons were relatively subtle. By combining the whole dataset, we identified four potential negative regulators of axon growth and regeneration. The candidate genes encode proteins with related functions, including cell adhesion, synaptic transmission, synaptic plasticity, and signal transduction. Accumulating evidence suggests that changes in intrinsic neuronal properties, in particular at the synaptic level, are shared among different CNS disorders (Mitchell, 2011). Along this line, Cacna2d3 has been associated with CNS disorders including autism (De Rubeis et al., 2014). However, none of the genes that we identified have yet been associated with axon growth or regeneration failure. This may be explained by the more comprehensive coverage of the transcriptome provided by RNA sequencing (RNA-seq) (Nagalakshmi et al., 2008) compared to previous hybridization-based microarray technologies (Costigan et al., 2002). In addition, our approach of combining three different experimental paradigms for screening facilitated gene prioritization. Furthermore, our screening approach aimed at dissecting the molecular mechanisms mediating the decline in axon growth ability and its relationship with regeneration failure in the adult, rather than identifying genes promoting axon growth and regeneration, which others have targeted (Costigan et al., 2002; Kadoya et al., 2009). As neurons readily make synapses (Friedman et al., 2000), we propose that a mechanism must be in place to inhibit the formation of inappropriate presynaptic specializations when growth cones are still navigating through non-target areas. Transcriptional and microRNA-mediated regulatory programs may allow expression of synaptogenic genes and pathways only when growth cones arrive in close proximity to their targets. At this stage, synaptogenesis could override the intrinsic axon growth program (Williams et al., 2010), thereby contributing to a loss of regenerative ability. Further dissection of the regulatory programs underlying changes in intrinsic axon growth ability will be an important direction for future investigations. The Role of Alpha2delta2 and Neurotransmission in Suppressing Axon Growth Spatial and temporal arrangement of calcium transients, combined with specific expression patterns of VGCCs, may be instrumental to shift the balance between rapid axon growth and presynaptic terminal differentiation (Gomez and Spitzer, 1999). Consistent with this idea, reduced excitability and impaired neurotransmitter release keep the axons in a growth

(F) Quantification of (E). Mean and SEM (**p < 0.01; triplicate experiments; 75-100 neurons per condition). (G) Representative fluorescence images of dissociated DRG neurons from Cacna2d2 KO and control littermates cultured for 12 hr and stained with a Tuj1 antibody. Scale bar, 200 mm. (H) Quantification of (G). Mean and SEM (***p < 0.001; CTR n = 3 and Cacna2d2 KO n = 4 animals; 130–170 neurons per condition). See also Figures S3 and S4.

Neuron 92, 1–16, October 19, 2016 7

Please cite this article in press as: Tedeschi et al., The Calcium Channel Subunit Alpha2delta2 Suppresses Axon Regeneration in the Adult CNS, Neuron (2016), http://dx.doi.org/10.1016/j.neuron.2016.09.026

A

B

D

E

F

C

G

H

(legend on next page)

8 Neuron 92, 1–16, October 19, 2016

Please cite this article in press as: Tedeschi et al., The Calcium Channel Subunit Alpha2delta2 Suppresses Axon Regeneration in the Adult CNS, Neuron (2016), http://dx.doi.org/10.1016/j.neuron.2016.09.026

mode, resulting in axons growing beyond the target area (Wang et al., 2007). As Cacna1c deletion failed to promote CNS axon regeneration (Enes et al., 2010), mechanisms independent of Cav1.2 channel activity likely control intrinsic axon growth ability in the CNS. N- and P/Q-type channels are primarily responsible for initiation of synaptic transmission (Catterall, 2011) and Alpha2delta subunits increase VGCC density and vesicle release probability (Hoppa et al., 2012). Thus, Alpha2delta subunits may function as a molecular switch to enhance synaptic transmission, thereby suppressing axon growth. In mammals, four genes encoding Alpha2delta subunits have been identified. While Cacna2d1, Cacna2d2, and Cacna2d3 are differentially expressed in neurons (Klugbauer et al., 1999), Cacna2d4 is mainly non-neuronal (Qin et al., 2002). Our analysis identified only Cacna2d2 as a potential suppressor of axon growth and regeneration in DRG neurons. Neither Cacna2d1 nor Cacna2d3 expression profiles mirrored Cacna2d2 expression in all three paradigms. In fact, Cacna2d2 expression not only positively correlated with the loss of axon growth ability during late embryonic development, but also negatively correlated with gaining growth competence in the adult. Hence, Cacna2d2 is regulated in the opposite way than previously suggested for Cacna2d1 (Costigan et al., 2002). Of note, overexpressing Cacna2d1 and Cacna2d3 had only a small effect on reducing axon growth compared to Cacna2d2 overexpression. The mechanisms regulating expression of these genes are not completely understood. It is likely that Cacna2d2 may be regulated as part of a cell-autonomous axon growth program. We show that removal of extracellular calcium and N- and P/Q-type channel blockers are both sufficient to rescue axon growth defects induced by Cacna2d2 overexpression, providing evidence that the presynaptic function of Alpha2delta2 is important for regulating axon growth ability. Whether Cacna2d2 alters axon growth and regeneration programs by regulating neurotransmitter release or different signaling components awaits clarification. Axon Regeneration versus Synapse Formation: Potential Relevance to CNS Disorders and SCI Genetic strategies that promote robust axon regeneration in the adult CNS have been recently developed. Gene-targeting approaches have been widely used, including studies on tuberous sclerosis 1 (Tsc1), phosphatase and tensin homolog (Pten), and Neurofibromin 1 (Nf1) (Park et al., 2008; Romero et al., 2007), genes known to play critical roles in synaptic function. Tsc1

null mutations cause cognitive and synaptic impairments (Goorden et al., 2007). Moreover, Pten conditional deletion in neurons drastically weakens synaptic transmission (Fraser et al., 2008), and Nf1 mutations inhibit synapse function (Costa et al., 2002). Thus, experimental evidence underscores a remarkable convergence between axon regeneration ability and abnormal synaptic structure and function. This relationship is also supported by recent studies showing that regenerating sensory axons are stabilized by formation of synaptic-like contacts on non-neuronal cells that stop them from growing further as they penetrate into the CNS (Di Maio et al., 2011; Filous et al., 2014). Moreover, a spared synaptic branch suppresses axon regeneration following highly localized laser injury in mice (Lorenzana et al., 2015). Interestingly, a similar phenomenon is also observed in C. elegans (Wu et al., 2007), suggesting that mechanisms controlling axon re-growth and regeneration failure may be conserved throughout evolution. Changes in the composition of the extracellular matrix (ECM) contribute to synapse formation (Dityatev and Schachner, 2003) and regeneration failure (Fawcett, 2015; Gaudet and Popovich, 2014). Several of the inhibitory signals present in the lesion site might have synaptogenic activity (Silver et al., 2014). The Alpha2delta subunits are largely extracellular, membrane-associated proteins. They contain several domains, including a von Willebrand factor A (VWA) domain, commonly involved in protein-protein interactions, especially between cell-adhesion proteins and the ECM (Whittaker and Hynes, 2002). Mutagenesis experiments suggest that the VWA domain indeed controls the interaction of Alpha2delta with ECM components, including thrombospondins and tenascin, interactions that are necessary for synapse formation (Eroglu et al., 2009; Kurshan et al., 2009). A motif located N-terminally to the VWA domains of both Alpha2delta1 and Alpha2delta2 subunits is required for the binding of the anti-epileptic drugs gabapentin and PGB (Davies et al., 2006), which powerfully inhibit excitatory synapse formation both in vitro and in vivo (Eroglu et al., 2009). Our results show that chronic PGB administration reduces synaptic transmission in vivo and promotes axon regeneration after SCI. Axon regeneration is one of several steps needed to restore function (Ramer et al., 2014). The concept that axon regeneration and consolidation of functional neuronal circuits are temporally distinct cellular programs may impact on rehabilitative training. Rehabilitation stabilizes newly formed functional circuits, promoting functional recovery upon regeneration. Combining

Figure 4. Cacna2d2 Inhibits Axon Growth via Calcium Influx through Cav2 Channels (A) Schematic representation of the Alpha2delta2 signaling cascade. (B) Immunoblot shows Alpha2delta2 and Cav2.1 expression in adult electroporated DRG neurons cultured for 24 hr. Tuj1 is shown as a loading control (n = 3 independent replicates per condition). (C) Representative fluorescence images of adult DRG neurons cultured for 24 hr after electroporation as in (B). Scale bar, 5 mm. (D) Imaris reconstruction and quantification of (C). Mean and SEM (**p < 0.01; triplicate experiments; at least 30 neurons per condition). (E) Quantification of maximal axon length and branching of CTR and Cacna2d2-overexpressing adult DRG neurons cultured for 24 hr in the presence of vehicle, EGTA, or N- and P/Q-type channels blockers. Mean and SEM (*p < 0.05, **p < 0.01; triplicate experiments; 67–120 neurons per condition). (F) Quantification of maximal axon length and branching of adult DRG neurons cultured for 24 hr after electroporation with either CTR or Cacna2d2-expressing plasmid DNA. CTR electroporated neurons were also cultured in the presence of the P/Q-type agonists Roscovitine (25 mM) and GV-58 (20 mM) or Ruthenium Red (100 mM). Mean and SEM (*p < 0.05, **p < 0.01, and ***p < 0.001; ns, not significant; triplicate experiments; 80–130 neurons per condition). (G) Representative fluorescence images of electroporated adult DRG neurons cultured for 24 hr in the presence of vehicle or PGB (250 mM). Scale bar, 200 mm. (H) Quantification of (G). Mean and SEM (**p < 0.01, ***p < 0.001; ns, not significant; triplicate experiments; 110–150 neurons per condition).

Neuron 92, 1–16, October 19, 2016 9

Please cite this article in press as: Tedeschi et al., The Calcium Channel Subunit Alpha2delta2 Suppresses Axon Regeneration in the Adult CNS, Neuron (2016), http://dx.doi.org/10.1016/j.neuron.2016.09.026

A

B

C

D

E

G

H

F

Figure 5. Systemic PGB Counteracts the Developmental Decline in Axon Growth Ability (A) Schematic of proprioceptive afferent-motorneuron connectivity in the mouse spinal cord. (B) Coronal sections of E17.5 spinal cords immunostained with a Parvalbumin antibody. PGB (46 mg/kg body weight) or an equal volume of vehicle was chronically administered between E12.5 and E17.5. (C) The area occupied by nerve terminals (dashed box in B) was quantified. Mean and SEM (***p < 0.001; vehicle n = 6 and PGB n = 6 embryos; 3 regularly spaced lumbar sections/embryo). (D) Coronal sections of E17.5 spinal cords (ventral horn) immunostained with Parvalbumin and Synaptophysin 1 antibodies. (E) Quantification of (D). Mean and SEM (*p < 0.05; vehicle n = 3 and PGB n = 3 embryos; 3–4 regularly spaced lumbar sections/embryo). (F) Quantification as in (E). Mean and SEM (**p < 0.01; CTR n = 7 and Cacna2d2 KO n = 6 embryos; 2–3 regularly spaced lumbar sections/embryo). (G) Representative fluorescence images of E17.5 DRG neurons originating from vehicle or PGB-treated embryos. Neurons were cultured for 12 hr and stained with a Tuj1 antibody. (H) Quantification of (G). Mean and SEM (***p < 0.001; triplicate experiments; 240–350 neurons per condition). Scale bars, 100 mm. See also Figure S5.

intensive training with regenerative strategies early after injury, however, results in aberrant fiber patterns and no functional recovery (Wahl et al., 2014). Therefore, gradual interruption of PGB treatment in parallel with rehabilitation may be necessary to create favorable conditions for functional connectivity at times

10 Neuron 92, 1–16, October 19, 2016

when regenerating axons approach target areas. This may be a critical point to consider, as PGB is already administered to SCI patients as a first-tier treatment to manage neuropathic pain. Although anticonvulsants have recently been shown to exert a beneficial effect on recovery in SCI individuals as shown by

Please cite this article in press as: Tedeschi et al., The Calcium Channel Subunit Alpha2delta2 Suppresses Axon Regeneration in the Adult CNS, Neuron (2016), http://dx.doi.org/10.1016/j.neuron.2016.09.026

A

C

B

D

Figure 6. Systemic PGB Weakens Synaptic Transmission and Promotes DC Axon Regeneration In Vivo (A) Representative fluorescence images of cultured adult DRG neurons originating from vehicle or PGB-treated animals (vehicle n = 3 and PGB n = 3 animals). Neurons were fixed 12 hr after plating and stained using a Tuj1 antibody. (B) Quantification of (A). Mean and SEM (**p < 0.01, ***p < 0.001; triplicate experiments; 215–220 neurons per condition). (C) In vivo multiphoton imaging of injured GFP-positive DC axons degenerating or regenerating over time. Images were manually stitched. Asterisk indicates the lesion epicenter (R, rostral; C, caudal). (D) Schematic drawing and quantification of (C). Scatterplot with mean (***p < 0.001; vehicle n = 8 and PGB n = 8 animals; 6–15 axons/animal). Scale bars, 200 mm. See also Figure S6.

increase in total motor score (Cragg et al., 2016), it is still unclear which anticonvulsants were administered to patients. Therefore, it will be important to assess whether patients receiving PGB after SCI improve neurological functions. To maximize the positive outcome, however, the treatment should be started soon after trauma and not after complications such as neuropathic pain have developed.

In summary, manipulation of a developmental switch centered on Alpha2delta2 allows adult sensory neurons to return to a growth-competent state, promoting axon regeneration in the adult CNS. Our results highlight the importance of dissecting the molecular mechanisms underlying changes in neuronal properties for the development of safe therapies aimed at promoting structural plasticity and regeneration after CNS injury.

Neuron 92, 1–16, October 19, 2016 11

Please cite this article in press as: Tedeschi et al., The Calcium Channel Subunit Alpha2delta2 Suppresses Axon Regeneration in the Adult CNS, Neuron (2016), http://dx.doi.org/10.1016/j.neuron.2016.09.026

A

C

B

D

F

E

G

H

I

(legend on next page)

12 Neuron 92, 1–16, October 19, 2016

Please cite this article in press as: Tedeschi et al., The Calcium Channel Subunit Alpha2delta2 Suppresses Axon Regeneration in the Adult CNS, Neuron (2016), http://dx.doi.org/10.1016/j.neuron.2016.09.026

EXPERIMENTAL PROCEDURES Reagents All drugs except nerve growth factor (NGF, Sigma), Roscovitine (Sigma), and GV-58 (Alomone Labs) were purchased from Tocris Bioscience. Antibodies The following antibodies were used, mouse anti-bIII tubulin (Tuj1, MMS-435P, Covance), mouse anti-Synaptophysin 1 (101011, Synaptic systems), rabbit anti-Parvalbumin (PV25, Swant), mouse anti-Parvalbumin (P3088, Sigma), rabbit anti-Neurofilament 160 kD (MAB5254, Millipore), mouse anti-NeuN (A60, Chemicon), mouse anti-Glyceraldehyde 3-phosphate Dehydrogenase (GAPDH) (ACR001P, Acris), Alexa Fluor 488 Phalloidin (A12379, Invitrogen), rabbit anti-Cav2.1 (ACC001, Alomone Labs), rabbit anti-Alpha2delta1 (ACC-015, Alomone Labs), rabbit anti-Alpha2delta3 (ACC-103, Alomone Labs), rabbit anti- Bax (2772, Cell Signaling), rabbit anti-Bcl-xL (2762, Cell Signaling), rabbit anti-pERK (4370P, Cell Signaling), and rabbit anti-GFAP (Z0334, Dako). A custom-made polyclonal antibody recognizing murine Alpha2delta2 was generated (Charles River). Animals and Surgeries All animal experiments were performed in accordance to the Animal Welfare Act and the guidelines of the Landesamt fu¨r Natur, Umwelt und Verbraucherschutz (LANUV). Female C57BL/6J mice (7–8 weeks old, Charles River) were used for all experiments except those specifying Cacna2d2 KO (Ducky 2J) or GFP-M mice. Ducky 2J (Stock Nr. 008625) and GFP-M (Stock Nr. 007788) mice were purchased from the Jackson Laboratories. Details on surgical procedures are described in the Supplemental Experimental Procedures. DRG Culture DRGs were dissected and collected in ice-cold Hank’s balanced salt solution (HBSS, GIBCO). After removing the surrounding connective tissue, the ganglia were transferred into a sterile tube and washed twice with HBSS. The ganglia were incubated in Neurobasal-A medium (GIBCO) containing collagenase type I (3,000 U/mL, Worthington) at 36.5
C for 15 min, followed by 30 min with trypsin (0.25%, GIBCO). Embryonic ganglia were incubated at 36.5
C in the presence of both collagenase and trypsin for 30 min. Serum was then added to stop trypsin digestion. Ganglia were dissociated by gently pipetting up and down. The cell suspension was filtered using a nylon cell strainer (70 mm, BD Falcon) and centrifuged at 900 rpm for 5 min. Dissociated neurons were re-suspended in Neurobasal-A or Neurobasal medium supplemented with B-27 (GIBCO) and were plated at low density on laminin- (5 mg/mL, Roche) or CSPGs-coated (10 mg/mL, Millipore) coverslips. Embryonic neurons were plated in the presence of NGF (12.5 ng/mL). In a set of experiments utilizing E17.5 neurons, PGB (46 mg/kg body weight) or an equal amount of vehicle (0.9% saline) was administered (intraperitoneal injections, three times/day) to pregnant mice (E12.5 to E17.5). The culture was maintained in a humidified atmosphere containing 5% CO2 in air at 36.5
C. When needed, the indicated chemicals were added to the culture medium while plating the neurons. Roscovitine, GV-58, and Ruthenium Red were added 2 hr after plating, when electroporation medium was replaced

with Neurobasal-A medium. For the siRNA gene silencing experiment, dissociated DRG neurons were electroporated with pre-designed siRNA oligos (Ambion) using the Amaxa Nucleofector (VPG-1001, Amaxa) system (program G-013). For the Cacna2d1, Cacna2d2, or Cacna2d3 overexpression or rescue experiments, dissociated DRG neurons were electroporated with a mixture of GFP (2.5 mg) and Cacna2d1-3 (4 mg) or empty (4 mg) expressing plasmid DNA. Electroporated neurons were plated at low density on laminin-coated coverslips. The electroporation medium was replaced with fresh medium 2 hr after plating. Cellular reconstruction was made using Imaris software (Version 7.7.1, Bitplane). The punctate isosurface in Figure 4D was created by adding a ‘‘spot’’ component in the surpass view. After selecting the source channel (i.e., Cav2.1) and point style (i.e., sphere), a manually set intensity threshold was used to separate objects from the background. All images were processed using the same thresholding parameters. RNA Isolation 3 3 104–1 3 107 DRG neurons were harvested and subsequently lysed in TRIZOL (Invitrogen) and total RNA was extracted according to the manufacturer’s protocol. The precipitated RNA was solved in RNase-free water. The quality of the RNA was assessed by measuring the ratio of absorbance at 260 nm and 280 nm using a Nanodrop 2000 Spectrometer (Thermo Scientific) and by visualization of 28S and 18S band integrity on an agarose gel. Generation of cDNA Libraries and Sequencing Total RNA was converted into libraries of double-stranded cDNA molecules as a template for high-throughput sequencing using the Illumina TruSeq RNA Sample Preparation Kit v2. Briefly, mRNA was purified from 100–500 ng of total RNA using poly-T oligo-attached magnetic beads. Fragmentation was carried out using divalent cations under elevated temperature in Illumina proprietary fragmentation buffer. First strand cDNA was synthesized using random oligonucleotides and SuperScript II. Second strand cDNA synthesis was subsequently performed using DNA Polymerase I and RNase H. Remaining overhangs were converted into blunt ends via exonuclease/polymerase activities and enzymes were removed. After adenylation of 30 ends of DNA fragments, Illumina adaptor oligonucleotides were ligated to prepare for hybridization. DNA fragments with ligated adaptor molecules were selectively enriched using Illumina PCR primers in a 15 cycles PCR reaction. Size-selection and purification of cDNA fragments with preferentially 200 bp in insert length was performed using SPRIBeads (Beckman-Coulter). Size distribution of cDNA libraries was measured using the Agilent high sensitivity DNA assay on a Bioanalyzer 2100 system (Agilent). cDNA libraries were quantified using KAPA Library Quantification Kits (Kapa Biosystems). After cluster generation on a cBot, 100 bp paired-end (PE) reads for Sham and PNL samples and 79 bp PE reads for time series samples were generated on a HiSeq1500 and demultiplexed using CASAVA 1.8. RNA-Seq Data Analysis Alignment to the mouse reference genome mm10 from UCSC was performed by TopHat2 (Kim et al., 2013) using standard settings (TopHat version 2.0.9, Bowtie version 2.1.0). Aligned BAM files (accepted hits) were imported into Partek Genomics Suite (PGS) software (version 6.6, Partek) for further analysis. mRNA quantification was performed using mm10 RefSeq Transcripts

Figure 7. Systemic PGB Administration after SCI Promotes Axon Regeneration (A) Experimental scheme. (B) Multiphoton scan of the unsectioned adult spinal cord 4 weeks after SCI. DC axons were labeled by injecting AAV-eGFP into the left sciatic nerve. Asterisk indicates the lesion epicenter (R, rostral; C, caudal). (C) Quantification of regenerating axons and comparison with the conditioning paradigm (*p < 0.05, **p < 0.01, ***p < 0.001; ns, not significant; LS, lesion site; vehicle n = 6, PGB n = 6 and PNL n = 10 animals). (D) Results of a blind replication experiment (*p < 0.05, **p < 0.01; ns: not significant; LS: lesion site; vehicle n = 9 and PGB n = 7 animals). (E) Sagittal sections of the injured spinal cord immunostained using a GFAP antibody. (F) Quantification of (E) (ns, not significant; vehicle n = 4 and PGB n = 4 animals; 3 regularly spaced sections/animal). (G) Experimental scheme. (H) Automated multiphoton tile scanning of the unsectioned adult spinal cord as in (B). Asterisk indicates the lesion epicenter (R, rostral; C, caudal). (I) Quantification of regenerating axons (*p < 0.05, **p < 0.01; ns, not significant; LS, lesion site; vehicle n = 8, PGB n = 8 animals). Scale bars, 100 mm. See also Figure S7 and Movie S1.

Neuron 92, 1–16, October 19, 2016 13

Please cite this article in press as: Tedeschi et al., The Calcium Channel Subunit Alpha2delta2 Suppresses Axon Regeneration in the Adult CNS, Neuron (2016), http://dx.doi.org/10.1016/j.neuron.2016.09.026

(2014-01-03) as annotation file. Afterward, raw gene counts throughput samples were normalized using DESeq algorithm in R (Dillies et al., 2013). The accession number for the data reported in this paper is GEO: GSE66128. Statistical Analysis Statistical analysis was performed using Prism (v. 6.0; GraphPad Software) as follows: unpaired two-tailed Student’s t test (Figures 1C, 1E, 1G, 2F, 2I, 3B, 3D, 3F, 3H, 4D, 5C, 5E, 5F, 5H, 6B, and 7F; Figures S1D, S3B, S3J, S3M, S5B, S5E, S5G, S5H, S6D, S6E, and S6G), one-way ANOVA followed by Dunnett posttest (Figure 4E; Figures S3G, S4A, S4C, S4E, and S4I), one-way ANOVA followed by Tukey posttest (Figures 4F and 4H; Figures S1D, S4F, and S4J), two-way ANOVA followed by Bonferroni posttest (Figure 6D). Permutation test (Figures 7C and 7D and 7I; Figure S7H) was performed using a custom script implemented in Python (2.7.3 version) including Pandas and NumPy libraries. For all analyses performed, significance was defined as *p < 0.05, **p < 0.01, and ***p < 0.001. ACCESSION NUMBERS

Cho, Y., Sloutsky, R., Naegle, K.M., and Cavalli, V. (2013). Injury-induced HDAC5 nuclear export is essential for axon regeneration. Cell 155, 894–908. Cho, Y., Shin, J.E., Ewan, E.E., Oh, Y.M., Pita-Thomas, W., and Cavalli, V. (2015). Activating Injury-Responsive Genes with Hypoxia Enhances Axon Regeneration through Neuronal HIF-1a. Neuron 88, 720–734. Costa, R.M., Federov, N.B., Kogan, J.H., Murphy, G.G., Stern, J., Ohno, M., Kucherlapati, R., Jacks, T., and Silva, A.J. (2002). Mechanism for the learning deficits in a mouse model of neurofibromatosis type 1. Nature 415, 526–530. Costigan, M., Befort, K., Karchewski, L., Griffin, R.S., D’Urso, D., Allchorne, A., Sitarski, J., Mannion, J.W., Pratt, R.E., and Woolf, C.J. (2002). Replicate high-density rat genome oligonucleotide microarrays reveal hundreds of regulated genes in the dorsal root ganglion after peripheral nerve injury. BMC Neurosci. 3, 16. Cragg, J.J., Haefeli, J., Jutzeler, C.R., Ro¨hrich, F., Weidner, N., Saur, M., Maier, D.D., Kalke, Y.B., Schuld, C., Curt, A., and Kramer, J.K. (2016). Effects of Pain and Pain Management on Motor Recovery of Spinal CordInjured Patients: A Longitudinal Study. Neurorehabil. Neural Repair 30, 753–761.

The accession number for the RNA-seq data reported in this paper is GEO: GSE66128.

Cregg, J.M., DePaul, M.A., Filous, A.R., Lang, B.T., Tran, A., and Silver, J. (2014). Functional regeneration beyond the glial scar. Exp. Neurol. 253, 197–207.

SUPPLEMENTAL INFORMATION

Davies, A., Douglas, L., Hendrich, J., Wratten, J., Tran Van Minh, A., Foucault, I., Koch, D., Pratt, W.S., Saibil, H.R., and Dolphin, A.C. (2006). The calcium channel alpha2delta-2 subunit partitions with CaV2.1 into lipid rafts in cerebellum: implications for localization and function. J. Neurosci. 26, 8748–8757.

Supplemental Information includes Supplemental Experimental Procedures, seven figures, two tables, and one movie and can be found with this article online at http://dx.doi.org/10.1016/j.neuron.2016.09.026. AUTHOR CONTRIBUTIONS A.T. conceived the project; A.T. and F.B. designed research; A.T., S.D., and C.J.L. performed research; A.T. and S.D. analyzed the data; J.X., T.U., M.B., and J.L.S. performed RNA-seq and bioinformatic analysis; F.B. supervised the research; A.T. and F.B. wrote the paper. J.L.S. provided feedback and contributed to editing the manuscript. ACKNOWLEDGMENTS We would like to thank M. Hu¨bener, W. Jackson, A. Kania, S. Mukherjee, G. Tavosanis, R. Wedlich-So¨ldner, and members of the F.B. Laboratory for critically reading and discussing the manuscript; W. Sun, D. Dietrich, C. Hoogenraad, S. Schoch-McGovern, J.J. Cragg, C.R. Jutzeler, A. Curt, and J.K. Kramer for discussion of our work; K. Keppler for manufacturing the spinal window; L. Meyn for technical assistance; and C. Mo¨hl for permutation test analysis. We also would like to thank J. Weiner and Y. Yoshida for discussion on Pv immunohistochemistry. This work was supported by IRP, WfL, and DFG (F.B.). J.L.S. was supported by SFB745 and SFB645. Received: April 11, 2016 Revised: July 21, 2016 Accepted: September 1, 2016 Published: October 6, 2016 REFERENCES Bauer, C.S., Tran-Van-Minh, A., Kadurin, I., and Dolphin, A.C. (2010). A new look at calcium channel a2d subunits. Curr. Opin. Neurobiol. 20, 563–571. Bornemann, D., Miller, E., and Simon, J. (1998). Expression and properties of wild-type and mutant forms of the Drosophila sex comb on midleg (SCM) repressor protein. Genetics 150, 675–686. Catterall, W.A. (2011). Voltage-gated calcium channels. Cold Spring Harb. Perspect. Biol. 3, a003947. Chandran, V., Coppola, G., Nawabi, H., Omura, T., Versano, R., Huebner, E.A., Zhang, A., Costigan, M., Yekkirala, A., Barrett, L., et al. (2016). A SystemsLevel Analysis of the Peripheral Nerve Intrinsic Axonal Growth Program. Neuron 89, 956–970.

14 Neuron 92, 1–16, October 19, 2016

De Rubeis, S., He, X., Goldberg, A.P., Poultney, C.S., Samocha, K., Cicek, A.E., Kou, Y., Liu, L., Fromer, M., Walker, S., et al.; DDD Study; Homozygosity Mapping Collaborative for Autism; UK10K Consortium (2014). Synaptic, transcriptional and chromatin genes disrupted in autism. Nature 515, 209–215. Di Maio, A., Skuba, A., Himes, B.T., Bhagat, S.L., Hyun, J.K., Tessler, A., Bishop, D., and Son, Y.J. (2011). In vivo imaging of dorsal root regeneration: rapid immobilization and presynaptic differentiation at the CNS/PNS border. J. Neurosci. 31, 4569–4582. Dickman, D.K., Kurshan, P.T., and Schwarz, T.L. (2008). Mutations in a Drosophila alpha2delta voltage-gated calcium channel subunit reveal a crucial synaptic function. J. Neurosci. 28, 31–38. Dillies, M.A., Rau, A., Aubert, J., Hennequet-Antier, C., Jeanmougin, M., Servant, N., Keime, C., Marot, G., Castel, D., Estelle, J., et al.; French StatOmique Consortium (2013). A comprehensive evaluation of normalization methods for Illumina high-throughput RNA sequencing data analysis. Brief. Bioinform. 14, 671–683. Dityatev, A., and Schachner, M. (2003). Extracellular matrix molecules and synaptic plasticity. Nat. Rev. Neurosci. 4, 456–468. Dolphin, A.C. (2012). Calcium channel auxiliary a2d and b subunits: trafficking and one step beyond. Nat. Rev. Neurosci. 13, 542–555. Donato, R., Page, K.M., Koch, D., Nieto-Rostro, M., Foucault, I., Davies, A., Wilkinson, T., Rees, M., Edwards, F.A., and Dolphin, A.C. (2006). The ducky(2J) mutation in Cacna2d2 results in reduced spontaneous Purkinje cell activity and altered gene expression. J. Neurosci. 26, 12576–12586. Enes, J., Langwieser, N., Ruschel, J., Carballosa-Gonzalez, M.M., Klug, A., Traut, M.H., Ylera, B., Tahirovic, S., Hofmann, F., Stein, V., et al. (2010). Electrical activity suppresses axon growth through Ca(v)1.2 channels in adult primary sensory neurons. Curr. Biol. 20, 1154–1164. Eroglu, C., Allen, N.J., Susman, M.W., O’Rourke, N.A., Park, C.Y., Ozkan, E., Chakraborty, C., Mulinyawe, S.B., Annis, D.S., Huberman, A.D., et al. (2009). Gabapentin receptor alpha2delta-1 is a neuronal thrombospondin receptor responsible for excitatory CNS synaptogenesis. Cell 139, 380–392. Ertu¨rk, A., Hellal, F., Enes, J., and Bradke, F. (2007). Disorganized microtubules underlie the formation of retraction bulbs and the failure of axonal regeneration. J. Neurosci. 27, 9169–9180. Fawcett, J.W. (2015). The extracellular matrix in plasticity and regeneration after CNS injury and neurodegenerative disease. Prog. Brain Res. 218, 213–226.

Please cite this article in press as: Tedeschi et al., The Calcium Channel Subunit Alpha2delta2 Suppresses Axon Regeneration in the Adult CNS, Neuron (2016), http://dx.doi.org/10.1016/j.neuron.2016.09.026

Filous, A.R., Tran, A., Howell, C.J., Busch, S.A., Evans, T.A., Stallcup, W.B., Kang, S.H., Bergles, D.E., Lee, S.I., Levine, J.M., and Silver, J. (2014). Entrapment via synaptic-like connections between NG2 proteoglycan+ cells and dystrophic axons in the lesion plays a role in regeneration failure after spinal cord injury. J. Neurosci. 34, 16369–16384. Finelli, M.J., Wong, J.K., and Zou, H. (2013). Epigenetic regulation of sensory axon regeneration after spinal cord injury. J. Neurosci. 33, 19664–19676. Fraser, M.M., Bayazitov, I.T., Zakharenko, S.S., and Baker, S.J. (2008). Phosphatase and tensin homolog, deleted on chromosome 10 deficiency in brain causes defects in synaptic structure, transmission and plasticity, and myelination abnormalities. Neuroscience 151, 476–488. Friedman, H.V., Bresler, T., Garner, C.C., and Ziv, N.E. (2000). Assembly of new individual excitatory synapses: time course and temporal order of synaptic molecule recruitment. Neuron 27, 57–69. Gaudet, A.D., and Popovich, P.G. (2014). Extracellular matrix regulation of inflammation in the healthy and injured spinal cord. Exp. Neurol. 258, 24–34. Gee, N.S., Brown, J.P., Dissanayake, V.U., Offord, J., Thurlow, R., and Woodruff, G.N. (1996). The novel anticonvulsant drug, gabapentin (Neurontin), binds to the alpha2delta subunit of a calcium channel. J. Biol. Chem. 271, 5768–5776.

Lorenzana, A.O., Lee, J.K., Mui, M., Chang, A., and Zheng, B. (2015). A surviving intact branch stabilizes remaining axon architecture after injury as revealed by in vivo imaging in the mouse spinal cord. Neuron 86, 947–954. Minet, A.D., Rubin, B.P., Tucker, R.P., Baumgartner, S., and ChiquetEhrismann, R. (1999). Teneurin-1, a vertebrate homologue of the Drosophila pair-rule gene ten-m, is a neuronal protein with a novel type of heparin-binding domain. J. Cell Sci. 112, 2019–2032. Mitchell, K.J. (2011). The genetics of neurodevelopmental disease. Curr. Opin. Neurobiol. 21, 197–203. Nagalakshmi, U., Wang, Z., Waern, K., Shou, C., Raha, D., Gerstein, M., and Snyder, M. (2008). The transcriptional landscape of the yeast genome defined by RNA sequencing. Science 320, 1344–1349. Neumann, S., and Woolf, C.J. (1999). Regeneration of dorsal column fibers into and beyond the lesion site following adult spinal cord injury. Neuron 23, 83–91. Omura, T., Omura, K., Tedeschi, A., Riva, P., Painter, M.W., Rojas, L., Martin, J., Lisi, V., Huebner, E.A., Latremoliere, A., et al. (2015). Robust Axonal Regeneration Occurs in the Injured CAST/Ei Mouse CNS. Neuron 86, 1215– 1227.

Gomez, T.M., and Spitzer, N.C. (1999). In vivo regulation of axon extension and pathfinding by growth-cone calcium transients. Nature 397, 350–355.

Park, K.K., Liu, K., Hu, Y., Smith, P.D., Wang, C., Cai, B., Xu, B., Connolly, L., Kramvis, I., Sahin, M., and He, Z. (2008). Promoting axon regeneration in the adult CNS by modulation of the PTEN/mTOR pathway. Science 322, 963–966.

Gong, H.C., Hang, J., Kohler, W., Li, L., and Su, T.Z. (2001). Tissue-specific expression and gabapentin-binding properties of calcium channel alpha2delta subunit subtypes. J. Membr. Biol. 184, 35–43.

Prasad, T., and Weiner, J.A. (2011). Direct and Indirect Regulation of Spinal Cord Ia Afferent Terminal Formation by the g-Protocadherins. Front. Mol. Neurosci. 4, 54.

Goorden, S.M., van Woerden, G.M., van der Weerd, L., Cheadle, J.P., and Elgersma, Y. (2007). Cognitive deficits in Tsc1+/- mice in the absence of cerebral lesions and seizures. Ann. Neurol. 62, 648–655. Grimm, J., Sachs, M., Britsch, S., Di Cesare, S., Schwarz-Romond, T., Alitalo, K., and Birchmeier, W. (2001). Novel p62dok family members, dok-4 and dok-5, are substrates of the c-Ret receptor tyrosine kinase and mediate neuronal differentiation. J. Cell Biol. 154, 345–354. Hall, Z.W., and Sanes, J.R. (1993). Synaptic structure and development: the neuromuscular junction. Cell 72 (Suppl ), 99–121. Hendrich, J., Bauer, C.S., and Dolphin, A.C. (2012). Chronic pregabalin inhibits synaptic transmission between rat dorsal root ganglion and dorsal horn neurons in culture. Channels (Austin) 6, 124–132. Hoppa, M.B., Lana, B., Margas, W., Dolphin, A.C., and Ryan, T.A. (2012). a2d expression sets presynaptic calcium channel abundance and release probability. Nature 486, 122–125. Ji, R.R., Baba, H., Brenner, G.J., and Woolf, C.J. (1999). Nociceptive-specific activation of ERK in spinal neurons contributes to pain hypersensitivity. Nat. Neurosci. 2, 1114–1119. Kadoya, K., Tsukada, S., Lu, P., Coppola, G., Geschwind, D., Filbin, M.T., Blesch, A., and Tuszynski, M.H. (2009). Combined intrinsic and extrinsic neuronal mechanisms facilitate bridging axonal regeneration one year after spinal cord injury. Neuron 64, 165–172. Kim, D., Pertea, G., Trapnell, C., Pimentel, H., Kelley, R., and Salzberg, S.L. (2013). TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol. 14, R36. Klugbauer, N., Lacinova´, L., Marais, E., Hobom, M., and Hofmann, F. (1999). Molecular diversity of the calcium channel alpha2delta subunit. J. Neurosci. 19, 684–691. Kurshan, P.T., Oztan, A., and Schwarz, T.L. (2009). Presynaptic alpha2delta-3 is required for synaptic morphogenesis independent of its Ca2+-channel functions. Nat. Neurosci. 12, 1415–1423. Leon, S., Yin, Y., Nguyen, J., Irwin, N., and Benowitz, L.I. (2000). Lens injury stimulates axon regeneration in the mature rat optic nerve. J. Neurosci. 20, 4615–4626. Liu, K., Tedeschi, A., Park, K.K., and He, Z. (2011). Neuronal intrinsic mechanisms of axon regeneration. Annu. Rev. Neurosci. 34, 131–152.

Puttagunta, R., Tedeschi, A., So´ria, M.G., Hervera, A., Lindner, R., Rathore, K.I., Gaub, P., Joshi, Y., Nguyen, T., Schmandke, A., et al. (2014). PCAFdependent epigenetic changes promote axonal regeneration in the central nervous system. Nat. Commun. 5, 3527. Qin, N., Yagel, S., Momplaisir, M.L., Codd, E.E., and D’Andrea, M.R. (2002). Molecular cloning and characterization of the human voltage-gated calcium channel alpha(2)delta-4 subunit. Mol. Pharmacol. 62, 485–496. Ramer, L.M., Ramer, M.S., and Bradbury, E.J. (2014). Restoring function after spinal cord injury: towards clinical translation of experimental strategies. Lancet Neurol. 13, 1241–1256. Richardson, P.M., and Issa, V.M. (1984). Peripheral injury enhances central regeneration of primary sensory neurones. Nature 309, 791–793. Romero, M.I., Lin, L., Lush, M.E., Lei, L., Parada, L.F., and Zhu, Y. (2007). Deletion of Nf1 in neurons induces increased axon collateral branching after dorsal root injury. J. Neurosci. 27, 2124–2134. Schwab, M.E., and Strittmatter, S.M. (2014). Nogo limits neural plasticity and recovery from injury. Curr. Opin. Neurobiol. 27, 53–60. Shewan, D., Berry, M., and Cohen, J. (1995). Extensive regeneration in vitro by early embryonic neurons on immature and adult CNS tissue. J. Neurosci. 15, 2057–2062. Shin, J.E., Cho, Y., Beirowski, B., Milbrandt, J., Cavalli, V., and DiAntonio, A. (2012). Dual leucine zipper kinase is required for retrograde injury signaling and axonal regeneration. Neuron 74, 1015–1022. Silver, J., Schwab, M.E., and Popovich, P.G. (2014). Central nervous system regenerative failure: role of oligodendrocytes, astrocytes, and microglia. Cold Spring Harb. Perspect. Biol. 7, a020602. Smith, D.S., and Skene, J.H. (1997). A transcription-dependent switch controls competence of adult neurons for distinct modes of axon growth. J. Neurosci. 17, 646–658. Tedeschi, A. (2012). Tuning the orchestra: transcriptional pathways controlling axon regeneration. Front. Mol. Neurosci. 4, 60. Wahl, A.S., Omlor, W., Rubio, J.C., Chen, J.L., Zheng, H., Schro¨ter, A., Gullo, M., Weinmann, O., Kobayashi, K., Helmchen, F., et al. (2014). Neuronal repair. Asynchronous therapy restores motor control by rewiring of the rat corticospinal tract after stroke. Science 344, 1250–1255.

Neuron 92, 1–16, October 19, 2016 15

Please cite this article in press as: Tedeschi et al., The Calcium Channel Subunit Alpha2delta2 Suppresses Axon Regeneration in the Adult CNS, Neuron (2016), http://dx.doi.org/10.1016/j.neuron.2016.09.026

Wang, C.L., Zhang, L., Zhou, Y., Zhou, J., Yang, X.J., Duan, S.M., Xiong, Z.Q., and Ding, Y.Q. (2007). Activity-dependent development of callosal projections in the somatosensory cortex. J. Neurosci. 27, 11334–11342. Whittaker, C.A., and Hynes, R.O. (2002). Distribution and evolution of von Willebrand/integrin A domains: widely dispersed domains with roles in cell adhesion and elsewhere. Mol. Biol. Cell 13, 3369–3387. Williams, M.E., de Wit, J., and Ghosh, A. (2010). Molecular mechanisms of synaptic specificity in developing neural circuits. Neuron 68, 9–18.

16 Neuron 92, 1–16, October 19, 2016

Wu, Z., Ghosh-Roy, A., Yanik, M.F., Zhang, J.Z., Jin, Y., and Chisholm, A.D. (2007). Caenorhabditis elegans neuronal regeneration is influenced by life stage, ephrin signaling, and synaptic branching. Proc. Natl. Acad. Sci. USA 104, 15132–15137. Ylera, B., Ertu¨rk, A., Hellal, F., Nadrigny, F., Hurtado, A., Tahirovic, S., Oudega, M., Kirchhoff, F., and Bradke, F. (2009). Chronically CNS-injured adult sensory neurons gain regenerative competence upon a lesion of their peripheral axon. Curr. Biol. 19, 930–936.