Molecular Psychiatry (2015) 20, 1037–1045 © 2015 Macmillan Publishers Limited All rights reserved 1359-4184/15 www.nature.com/mp

REVIEW CRMPs: critical molecules for neurite morphogenesis and neuropsychiatric diseases

TT Quach1,2, J Honnorat1,3,4, PE Kolattukudy5, R Khanna6 and AM Duchemin7

Neuronal polarity and spatial rearrangement of neuronal processes are central to the development of all mature nervous systems. Recent studies have highlighted the dynamic expression of Collapsin-Response-Mediator (CRMPs) in neuronal dendritic/ axonal compartments, described their interaction with cytoskeleton proteins, identified their ability to activate L- and N-type voltage-gated calcium channels (VGCCs) and delineated their crucial role as signaling molecules essential for neuron differentiation and neural network development and maintenance. In addition, evidence obtained from genome-wide/genetic linkage/proteomic/ translational approaches revealed that CRMP expression is altered in human pathologies including mental (schizophrenia and mood disorders) and neurological (Alzheimer’s, prion encephalopathy, epilepsy and others) disorders. Changes in CRMPs levels have been observed after psychotropic treatments, and disrupting CRMP2 binding to calcium channels blocked neuropathic pain. These observations, altogether with those obtained from genetically modified mice targeting individual CRMPs and RNA interference approaches, pave the way for considering CRMPs as potential early disease markers and modulation of their activity as therapeutic strategy for disorders associated with neurite abnormalities.

Molecular Psychiatry (2015) 20, 1037–1045; doi:10.1038/mp.2015.77; published online 16 June 2015

The ability of the nervous system to process continuously maintenance of axodendritic arborization and discuss their changing information from the internal and external environ- potential involvement in human nervous system diseases. ments requires the development of precise networks of neurons and structural plasticity of their dendrites and axons. Considering the intricacies of specific axodendritic connectivity, it is surprising CRMPS: CHARACTERISTICS AND EXPRESSION that neurons can undergo such precise structural reorganization in CRMPs are five homologous cytosolic phosphoproteins.6–8 CRMP1, response to environmental change. However, evidence indicates 2, 3 and 4 display 75% homology with each other,9–11 whereas that far from being static structures, axons and dendrites are the phylogenetically divergent CRMP5/CRAM shares only 50% plastic and this plasticity, which is part of a lifelong process, is homology.12 CRMP1, 2 and 4 have two distinct alternatively mediated by various proteins such as neurotrophic/transcription spliced isoforms in their N termini,13 extending the number of factors, neurohormones, guidance cues, membrane receptors, active molecules in this family. All CRMPs assemble in heterophilic voltage-gated calcium channels (VGCCs) and involves rearrange- oligomers in vivo, bind to tubulin14 and are phosphorylated by ment of cytoskeletal proteins that support the structural changes.1 various kinases that regulate their activity.15 They act as signaling However, despite the rich information on mechanisms underlying phosphoproteins modulating cytoskeletal organization and neuronal differentiation, axonogenesis and dendritogenesis, the regulating neurite formation and retraction. Site-specific phos- key molecules involved in the initiation of neurites through phorylation of CRMPs determines their affinity for binding to cyto- formation of lamellopodia/filopodia, specification and maturation skeletal proteins and their neurite-promoting/collapsing activities. – of neurites, and genesis of spines still remain elusive.2 3 Because CRMP1 appears to be a target of RhoA signaling.16 The Sema3A17 such a large fraction of the genome is expressed in the brain and and reelin18 signaling pathways are regulated by the phosphor- our knowledge of mechanisms involved in neuronal connectivity ylation of CRMP1 by Cdk5 at Thr-509 and Ser-522, or by Fyn at is still in its infancy, it has been postulated that more mole- Tyr-504 that lowers its binding affinity to tubulin. Similarly, non- cular cues and pathways specialized for axonal/dendritic/spine phosphorylated CRMP2 binds strongly to tubulin leading to genesis and synaptic network formation remain to be iden- microtubule formation, whereas its phosphorylation by Rho kinase tified. Consistent with this notion, Collapsin-Response-Mediator suppresses its binding to tubulin and Numb.14 Two other kinases Proteins (CRMPs) can now be considered as building blocks for working in tandem, Cdk5 and GSK-3β, also control the interaction neuronal differentiation and neurite network organization and of CRMP2 with tubulin. In addition, CRMP2 regulates the stability remodeling.4–5 In this review, we outline the current evidence of actin filaments.19 CRMP3 is phosphorylated by Fyn in vitro15 but showing that CRMPs are important in the establishment and is unique among CRMPs in that the three GSK-3β phosphorylation

1Lyon Neuroscience Research Center, INSERM U1028/CNRS UMR 5292, F-69372 Lyon, France; 2Department of Neuroscience, The Ohio State University, Columbus, OH, USA; 3French Reference Center on Paraneoplastic Neurological Syndrome, Hospices Civils de Lyon, Hôpital Neurologique, Neurologie B, Bron, France; 4Université de Lyon, Université Claude Bernard Lyon 1, F-69372 Lyon, France; 5Burnett School of Biomedical Sciences, College of Medicine, University of Central Florida, Orlando, FL, USA; 6Department of Pharmacology and Neuroscience, Graduate Interdisciplinary Program, College of Medicine, The University of Arizona, Tucson, AZ, USA and 7Department of Psychiatry, College of Medicine, The Ohio State University, Columbus, OH, USA. Correspondence: Dr AM Duchemin, Department of Psychiatry, College of Medicine, The Ohio State University, Columbus, OH 43210, USA. E-mail: [email protected] Received 7 November 2014; revised 29 April 2015; accepted 8 May 2015; published online 16 June 2015 CRMPs in neuron development and diseases TT Quach et al 1038

Table 1. Summary of relative CRMP expression pattern in mouse nervous system

Cortex Hippocampus Dentate gyrus Striatum Cerebellum Pons DRG

CRMP1 ePN + + + ? +++ ? ? Adult ++ +++ ++ ? +++ ? ? CRMP2 ePN ++ ++ ++ ++ ++ ? ++ Adult ++ +++ ++ + ++ ? + CRMP3 ePN − +++ +++ − +++1 ++ Adult − +++ +++ −−++1 − CRMP4 ePN ++ +++ +++ + + +2 ? Adult − ++−−−+ CRMP5 ePN + + + ? + ? ? Adult + ++ ++ + ++ ? ? Abbreviations: DRG, dorsal root ganglia; ePN, early postnatal. (1), inferior olive complex and reticular formation; (2), pontine nuclei. Data from refs 9,12,13,23,29,31,32,44,46,48,127–129; +++, high level; ++, moderate level; +, low level; − , undetectable signal; ?, not determined.

sites identified in other CRMPs are not conserved in this molecule. activation of L- and N-type VGCCs by CRMP3 provide strong CRMP4, which binds to F-actin and regulates F-actin bundling,20 is support for an inductive role of CRMP3 in neurite initiation and also a physiological partner of GSK-3β.21 Of note, GSK-3β is a dendritic development and plasticity, at least partly via Ca2+ ubiquitous kinase having a pivotal role in neurodevelop- influx29 (Figure 1d). In contrast, CRMP5 depletion by RNA ment and is suspected to be involved in neuropsychiatric interference enhances the length and number of neurites, disorders such as schizophrenia,22 bipolar disorders23 and whereas its overexpression induces mitophagy and reduces Alzheimer’s disease (AD).24 CRMP5 is a substrate for Cdk525 and dendritic length. Further, overexpression of mutated-Threo-516 26 is tyrosine-phosphorylated by Fes/Fps and TrkB. In view of these CRMP5 (ref. 34) and/or its truncated dominant-negative isoform observations, we can conclude that, depending on the character- lacking the tubulin-binding domain35 has no effect on dendrite istics of their phosphorylation status, CRMPs may associate with formation or length, demonstrating the critical role of the tubulin- 27 various partners and signaling pathways, some of which may binding domain in the neurite outgrowth inhibition induced by exclude direct interaction between them in different brain areas. CRMP5. The interaction of CRMP5 with tubulin has important 28 29 30 In addition, CRMP2, CRMP3 and CRMP4 have been shown to ramifications for the function of CRMP2. Consistent with the modulate calcium flux, another mechanism of neurite regulation. 10 31 notion that the activities of CRMPs are inter-related, dendritic According to in situ hybridization, immunostaining and arborization induced by CRMP2 is considerably enhanced by 32 fi replacement analyses, all ve CRMPs are highly expressed CRMP5 deficiency.36 There is also evidence that CRMP4 regulates in the brain during early postnatal development and continue to dendritic outgrowth: CRMP4–short hairpin RNA-transfected be present in specific regions into adulthood, especially in areas and CRMP4 − / − cultured hippocampal neurons show increased with extensive neuronal plasticity (Table 1). They are found in dendritic branching without significant changes in primary axons and dendrites of the neurons where they contribute to dendrite number.37 This effect is partly dependent on Sema3A specific signaling mechanisms involved in the regulation of axonal signaling through F-actin bundling.20 Phosphorylated CRMP1 and and dendritic development and maintenance. CRMP2 are localized in the dendrites of cortical neurons where they regulate dendritic branch trajectories.38 Mice with abnormal phosphorylation of CRMP1 and CRMP2 present a curling dendrite CRMPS INFLUENCE NEURITE INITIATION AND DENDRITIC 38 OUTGROWTH phenotype, suggesting that these two molecules are critical for normal dendrite patterning in cortical neurons. The roles of molecules critical for dendritic specification have been examined extensively in cultured primary neurons. In vitro studies examining the selective contribution of CRMPs to dendritic CRMPS AND AXON SPECIFICATION development revealed their differential roles. In cultured hippo- It has been assumed that axon–dendrite polarity involves campal neurons, CRMP3 is essential, not only for dendrite exten- polarized transport of various cargoes, actin instability, complex sion, a late-stage event in the differentiation of neurons, but also microtubule orientation and distinct repertoires of extracellular in lamellopodia formation, the early stage of neurite initiation matrix/structural/signaling proteins. Some proteins are specific for (Supplementary Information, SI). Although neurons from control either dendritic or axonal specification: for example, down- wild type and CRMP3+/ − mice polarize and grow normally, all regulation of myelin-associated protein-2 inhibits dendrite forma- neurons lacking CRMP3, identified by their expression of the LacZ- tion, whereas downregulation of Tau1 inhibits axon formation. reporter gene (SI-A), from CRMP3 − / − littermates do not initiate However, several molecules including CRMP2 can affect both neurite formation.29 They remain spherical, even after 10 days dendrites38 and axons.39 In hippocampal neurons, overexpression in vitro. This defect in neuritogenesis is largely due to a failure to of CRMP2 induces supernumerary axons, and increases axonal evolve from stage 1 to stage 2, defined as neurite initiation. The branching, whereas its inhibition with a dominant-negative first step in neuronal polarization requires the establishment of protein suppresses the formation of the primary axon.39 Interest- lamellopodia, considered as the principal means of initiating the ingly, CRMP2 can convert established dendrites to axon. It is formation of primary neurites. Consistent with this finding, CRMP3 involved in the Sra-1/WAVE/kinesin cargo complex and is required overexpression promotes lamellopodia formation and significantly for brain derived neurotrophic factor (BDNF) and neurotrophin 3 increases both the number and length of dendrites (Figure 1). In (NT3)-induced axon outgrowth/branching.19,40 How might CRMP2 cortical neurons, Ena/VASP deletion also prevents neurite initiation fulfill this function? Axon elongation is achieved through micro- but by inhibiting filopodia formation,33 the second step in tubule assembly and extension, actin reorganization and mem- establishing the primary neurites. The correlation between levels brane genesis at the leading edge of the growth cone. CRMP2 of CRMP3 expression, dendrite/spine morphogenesis and the may not only enhance microtubule polymerization but may also

Molecular Psychiatry (2015), 1037 – 1045 © 2015 Macmillan Publishers Limited CRMPs in neuron development and diseases TT Quach et al 1039 complex formation is critical for outgrowth inhibition via an actin-dependent process.30 Ectopic expression of the short isoform of CRMP4 enhances axonal extension, whereas its dephosphorylated long isoform mediates the inhibition of myelin-associated inhibitor- induced axonal outgrowth illustrating that the long isoform can serve as short isoform antagonist. The role of CRMP1 in axono- genesis was demonstrated by the abolition of axonal extension induced by NT3 in chick dorsal root ganglia neurons using specific antibodies or CRMP1-mutant proteins.43 In CRMP1 − / − mice, axonal lengths in cerebellar explant cultures are not altered,44 suggesting selective functional activity of CRMP1 depending on cell types and/or species. Finally, despite their role in dendrito- genesis, overexpression or deletion of CRMP3 in cultured hippocampal CA1 neurons29 and deletion of CRMP5 in cerebellar explant cultures36 yield no apparent alteration of axon morphol- ogy, suggesting that some CRMPs may affect only dendrites.

PHENOTYPES OF CRMP-DEFICIENT MICE Although the cultured neuron is a very valuable model, one must be cautious when extrapolating results to in vivo situations. Polarity and asymmetry can be altered by disruption of cell–cell and cell–extracellular matrix interactions. Moreover, not all neurons select an axon from primary neurites as in the case of hippocampal neurons.45 In the living animal, the difference in axon-versus-dendrite genesis may be linked to the presence of extracellular cues affecting the immature neuron.33 Gene- targeting inactivation has enabled elucidation of unique functions of individual CRMPs in different cell types in vivo, although adaptive compensation by related CRMP members must be kept in mind. Figure 1. Effects of CRMP3 on dendrites of hippocampal neurons. ● CRMP1 − / − mice display abnormal granule cell migration and (a) Overexpression of CRMP3 increases dendritic arborization. 45 Dissociated 1 d.i.v. hippocampal neurons were transfected with apoptosis in the cerebellum, and abnormal GAP43/PSD95/ Flag-CRMP3 or Flag-vector, fixed at 4 days in vitro (div) and immuno- myelin-associated protein-2/Golgi staining in the CA1 area of stained for Flag (red) and dendritic marker myelin-associated hippocampus consistent with a structural alteration of protein-2 (MAP2; green). (A) Representative untransfected neuron. neurites.46 Behaviorally, CRMP1 − / − mice exhibit abnormalities (B) Neurons expressing control Flag-vector. (C) Neurons expressing in long-term potentiation (sSI), motor activity, emotional Flag-CRMP3. Merged images showed transfected neurons in yellow behavior, long-term retention and prepulse inhibition (bottom panels). Some Flag-CRMP3-expressing neurons display an (sSI); the latter is rescued by treatment with chlorpromazine, an increase of more than three times in the number of dendrites. (b)A proposed model for the regulation of dendrite morphology by antipsychotic drug. These observations raise the possibility that CRMP3. (1) Combinatorial stimulation of CRMP3 or CRMP1 − / − -associated dendrite and spine dysfunction may activity by intrinsic and extrinsic factors. (2) The localization and involve deficits in information processing. Interestingly, in vivo distribution of FL-CRMP3-positive puncta in dendrites suggests that microdialysis revealed an increased release of dopamine in the —similar to CRMP2—CRMP3 could be associated with vesicles or prefrontal cortex, which might contribute to the observed large carrier protein complexes, and actively transported to hyperactivity of CRMP1 − / − mice.47 – dendrites. (3 4) Structural and biochemical studies support the ● CRMP2 − / − mice: our attempt to target the CRMP2 gene was idea that CRMP3 activity might be regulated by phosphorylation or unsuccessful (Quach et al., unpublished observation) suggesting other post-translational modifications induced by (7) neurotrophic factors/guidance cues. Activated CRMP3 binds to different partners a critical role of CRMP2 in embryonic development. and serve as adaptor to a variety of signaling pathways including ● Analysis of CRMP3 − / − mice supports a role for CRMP3 in Ca2+ influx (CaV). CRMP3 contributes to (8) neurite initiation, (5) hippocampal dendritic organization and function. CRMP3 − / − dendrite arborization and (6) spine genesis. These effects may mice display behavioral alteration concordant with hippo- involve lamellopodia formation. Reproduced from Quach et al.29 campal functional impairment such as a decrease in long- term potentiation48 and prepulse inhibition.49 They present abnormal hippocampal dendrites characterized by decreased modulate vesicle/membrane trafficking/remodeling through inter- dendritic length and number of branching points, abnormal action with its partners, depending on specific post-translational undulation of apical primary dendrites and abnormal dendritic 48 modifications19,41–42 and secreted trophic factors. In hippocampal spine structure. The abnormal undulation, also present in L1 neurons, BDNF and NT3 induce axonal growth and branching knockout mice,50 is most likely due to altered signaling of activity by inhibiting CRMP2 phosphorylation by GSK-3β.39–40 CRMP3-dependent guidepost proteins directing point-to-point Conversely, Sema-3A induces axonal retraction by increasing navigation of growth cones on the spatial array. L1 is the co- CRMP2 phosphorylation by GSK-3β.15 Short interfering RNA- receptor of the Neuropilin1 receptor complex for Sema 3A. mediated knockdown of CRMP4 in the dorsal root ganglia Intriguingly, while dendrites are altered but remain present attenuates neurite outgrowth inhibition induced by myelin in vivo, the absence of neurites is the common phenotype of substrate, identifying CRMP4 as an inhibitor of axonal CRMP3 − / − cells in vitro.29 The apparent difference in cellular outgrowth.27 The long variant CRMP4 increases Ca2+ current in phenotypes obtained from the identical genotype may be whole-cell voltage clamp experiments and the CRMP4-RhoA related to functional redundancy between proteins and partial

© 2015 Macmillan Publishers Limited Molecular Psychiatry (2015), 1037 – 1045 CRMPs in neuron development and diseases TT Quach et al 1040 compensatory mechanisms of repair in vivo.29 Guidance cues/ involved in neuronal migration and neurite/synapse formation. extracellular matrix /neurotrophic factors that possibly activate Recent studies of gene copy number variants and exome other pathways leading to neurite formation in CRMP3 − / − sequencing of single-nucleotide polymorphisms (SNPs) impli- neurons may be absent in cultured cells. cate modulating cytoskeleton proteins, NMDA receptors ● In CRMP4 − / − mice, the distance from the soma to the first and VGCCs, converging on pathways regulating synaptic bifurcation point in the apical dendrites of CA1 pyramidal plasticity.55 It is not surprising that CRMP1, also called neurons is significantly decreased, confirming an inhibitory role dihydropyriminidase-like 1 (DPYSL1), which interacts with of CRMP4 on dendrite bifurcation and branching.37 Otherwise, and DISC1,18 and CRMP2 (DPYSL2), which affects not only 2+ these mice have no observable changes in the CNS structure. neurites but also neurotransmitter release and modulates Ca 28 ● In CRMP5 − / − mice, anti-calbindin immunostaining revealed fluxes through activation of NMDA receptors, have been altered cell body sizes and dendrites.26 In addition, the dendritic included in the list of schizophrenia susceptibility genes. CRMP1 branching induced by BDNF is markedly attenuated in these was identified as altered protein in a search for schizophrenia- 56 animals. Significant deficiencies in long-term depression (sSI) specific epitopes from patients’ post-mortem brain fractions. and abnormal limb-clasping reflexes are consistent with A genetic analysis corroborated the association of anhedonia 56 abnormal central synaptic function. These mice also develop with the CRMP1 gene in a Finnish population cohort. abnormalities in motor conduction velocity, polyaxonal Furthermore, CRMP1 immunoreactivity is increased in lympho- ensheathment and non-myelinating Schwann cell structure, blastoid cells derived from patients diagnosed with schizo- phrenia, suggesting that CRMP1 may be a potential biomarker and mimic a paraneoplastic neurological disorder syndrome 56 found to be associated with anti-CRMP5 auto-antibodies.51 of the disease. However, the precise mechanisms by which endogenous CRMP1 disrupts brain circuitry in patients, and However, the mechanisms underlying the structural changes 48 are unknown. It is possible that signaling changes originate in how the deletion of its gene disrupts brain circuitry in rodents the axon and initiate dysfunction in glial cells, which in return ultimately leading to behavioral similarities found in compara- leads to axonopathy. Complementary to gene deletions, tive phenotypic analyses, such as enhanced mesocortical dopaminergic transmission and decrease in prepulse inhibition transgenic mice overexpressing CRMPs may enhance our (also present in CRMP3 − / − mice49), remain to be determined. understanding of the molecular pathways and mechanisms Multiple proteome-wide analyses have identified significant underlying human pathology. changes in levels of CRMP2 in the brain of schizophrenia patients.57–64 Corroborating these data, a more-than-6000-SNP- linkage scan of 831 pedigrees, confirmed by an independent HUMAN NERVOUS SYSTEM DISEASES: INSIGHTS FROM Ashkenazi case–control and a family-based association of CRMP-ALTERED LEVELS/ACTIVITY AND ANIMAL MODELS European-Caucasian pedigrees, found a schizophrenia suscept- ibility 1-Mb region on 8p21 that contains the CRMP2 gene.65 Previous multicenter genome-wide linkage ● 52 Schizophrenia is considered a neurodevelopmental disorder studies have consistently showed strong evidence implicating characterized by ventricular enlargement and decreased 8p21 susceptibility locus for schizophrenia66–68 and CRMP2 volume of the frontal cortex and hippocampus, with a reduction gene in particular.69 Recent findings of association of several in size of neuronal soma, shortened length of axons and functional gene variants of CRMP2/DPYSL2 with schizophrenia dendrites and loss of spines in the prefrontal cortex and suggest that alterations of CRMP2 transcription and translation 53–54 hippocampus. Altered prepulse inhibition is a common increase the risk for the disease.70 However, despite finding an neurophysiological endophenotype. Transcriptomic studies association of CRMP2 polymorphism with schizophrenia in argue that the molecular underpinnings of the disease are Japanese patients, Koide et al.71 did not replicate it in another even more varied than its symptomatic diversity, and that group. Explanations for the conflicting linkage findings may changes in the transcriptome can affect developmental come from various factors present in (a) the disorder itself (its processes. Once compensatory mechanisms are exceeded, multifactorial genetic–allelic difference, phenotypic hetero- alterations manifest as a specific behavioral phenotype geneity and possible epigenetic nature), (b) methods (popula- corresponding to the symptoms of the disease. Discovery of tion pedigree, gender, ethnicity and sample sizes) and (c) NRG1/Akt/dysbindin-1, and DISC-1 as susceptibility genes for technology (SNPs are different between studies: for example, schizophrenia, suggested an etiopathogenic role for molecules the SNPrs 12155555 that shows strong association to

Figure 2. Suppression of inflammatory and neuropathic pain by uncoupling CRMP-2 from the presynaptic Ca2+ channel complex. (a) Normalized binding of CaV2.2 to 15-mer peptides (overlapping by 12 amino acids) encompassing full-length CRMP-2 overlaid with spinal cord lysates. The sequence of peptide 96, designated CBD3, is shown. (b) Left: representative differential interference contrast/fluorescence images showing robust penetration of fluorescein isothiocyanate-TAT-CBD3 into dorsal root ganglia (DRGs; arrowheads) but not other cells (arrows). Nuclei are stained with Hoechst dye in the bottom image. Right: representative current traces from a dorsal root ganglia (DRG) incubated for 15 min with TAT-Scramble (10 μM; black) or TAT-CBD3 (10 μM; cyan) in response to the voltage steps illustrated at the top. (c) Representative current clamp recordings from small-to-medium (430440 μm) diameter lumbar four to five DRG neurons demonstrating that application of 10 μM CBD3 (cyan) reduces the number of elicited action potentials; grouped data are shown in the bar graphs on the right. (d) Representative traces of spontaneous EPSCs (sEPSCs) in lamina II neurons in spinal cord slices before treatment (left) or after application of 10-μM TAT-Scramble (middle) or 10-μM TAT-CBD3 (right). Bottom, enlarged traces. Voltage-clamp recordings (holding voltage, –70 mV) were used to record synaptic responses. Right: ratio of sEPSC frequency and amplitude. *Po0.05 compared with baseline. (e) The effect of peptides on number of flinches on formalin-induced phase 1 (0–10 min) and phase 2 (15–60 min). (f) Paw withdrawal threshold (PWT; in mN) of rats injected once with the antiretroviral drug ddC and treated with TAT peptides at 7 days following ddC administration. *Po0.05 versus ddC or TAT-Scramble (analysis of variance (ANOVA) with Student–Newman–Keuls post hoc test). (g) Visualization of FITC fluorescence in DRGs, isolated 15 min after injection of FITC-TAT-CBD3 and immunolabeled with a NeuN-specific antibody (NeuN; red). Cells were also stained with Hoechst (blue), which labels cell nuclei. (h) Cartoon depicting the main hypothesis which proposes that uncoupling the CRMP-2-CaV2.2 interaction leads to a physiologically relevant decrease in Ca2+ current and neurotransmitter release and, in turn, suppresses persistent inflammatory and neuropathic hypersensitivity. Reproduced from Wang et al.115 and Brittain et al.121

Molecular Psychiatry (2015), 1037 – 1045 © 2015 Macmillan Publishers Limited CRMPs in neuron development and diseases TT Quach et al 1041 schizophrenia (P = 0.000063)63 was not tagged by Koide et al.71). of schizophrenia susceptibility genes inhibits the expression Levels of CRMP4 (DPYSL3) are decreased in the CA2/CA3 areas of CRMP4.73 CRMP4 is also a target for the NMDA-regulated of the hippocampus of schizophrenia patients,72 and these microRNA-132, which is downregulated in the prefrontal cortex changes may be under the controls of microRNAs altered in the of subjects with schizophrenia.74 Interestingly, CRMPs were disease. The overexpression of microRNA-137 coded by one identified in a proteomic profile of rat cerebral cortex proteins

© 2015 Macmillan Publishers Limited Molecular Psychiatry (2015), 1037 – 1045 CRMPs in neuron development and diseases TT Quach et al 1042 changed after chronic administration of two antipsychotic disorder,94 supports common etiopathogenic mechanisms for drugs, clozapine and risperidone:75 ‘Alterations in CRMP2 and disease symptomatology. CRMP4 were the most spectacular consequences of treatment ● Alzheimer’s disease: Since various forms of memory require an with both drugs’ the authors reported. Increased CRMP2 levels intact hippocampus, it is not surprising that it is one of the in schizophrenia correlate with antipsychotic exposure.62 earliest and most vulnerable brain regions affected by the Protein levels in patients may depend on brain area, stage of disease.95 In fact, hippocampal synaptic dysfunction and the disease and psychotropic use. In addition, chronic treat- memory impairment in AD are recapitulated in AD transgenic ment of rats with phencyclidine, which can mimic schizo- models with neuritic senile plaques (β-amyloid) and neurofi- phrenia symptoms, downregulates CRMP2 and upregulates brillary tangles (NFTs), which support the neuropathological CRMP4 levels in the medial prefrontal cortex.76 Despite these concept of a ‘dendritic-dysconnectivity’ syndrome96 associated supportive observations, the primary extent to which CRMP with aberrant NMDA-R activation, repressor element 1-silencing expression is associated with schizophrenia awaits further transcription factor (REST) expression97 and calcium signaling confirmation. perturbation.98 Interestingly, increased phosphorylated ● Mood disorders: major depressive disorder (MDD) is associated CRMP2 has been detected in the soluble fraction of AD- with hippocampal atrophy characterized by decreased density affected human brain tissue and in NTFs.99 In transgenic mouse in dendrites and dendritic spines that can be reversed models, phosphorylated CRMP2 is increased in hippocampus by antidepressant treatments.77 The role of stress, and espe- well before plaque and tangle formation.100 Although the cially early-life trauma, leading to hippocampal atrophy is relationship between pathophysiological mechanisms and AD well documented in the development of depression. In an symptoms remains to be clarified, these findings position animal model of early-life stress, the levels of CRMP5—an phosphorylated CRMP2 as a potential biomarker for early AD, inhibitor of dendritic growth—are increased while tubulin and especially since it seems specific to this type of dementia.101 In microtubule-associated protein are decreased.78 In MDD, the an animal model of AD, suppression of CRMP2 phosphorylation structural dendrite alterations associated with the depressive is associated with amelioration of β-amyloid-induced cognitive symptoms seem related to a deficiency in neurotrophic dysfunction and hippocampal axon degeneration.102 factors.79 Levels of the brain and serum BDNF are decreased ● Human prion diseases, which include sporadic/familial Creutz- in patients with depression.80–81 Furthermore, it has been feld–Jacob disease, are disorders of protein conformation in shown that (1) BDNF may be involved in the MDD gene– which PrPc, the normal cellular conformer, is converted to the environment interaction model82 and (2) BDNF+/ − knockout83 abnormal protease-resistant PrPSc.103-104, In animal models, a and BDNF Val66/Met knock-in84 mice display reduction in specific increase in CRMP4 expression105 and dendritic neurite length, altered dendritic spines and depression-like degeneration106 are the two earliest morphological features, behaviors. The role of CRMPs in BDNF effects is supported by detectable in mice before clinical symptoms are evident. In the the facts that (1) BDNF blocks phosphorylation of CRMP2 via the late stage of the disease, the amount of CRMP2-ΔC that lacks PI3-kinase/AKt/GSK3-β pathway40 and (2) knockdown of the sequence from Ser-518 to the C terminus is increased CRMP2 causes a marked inhibition of BDNF-induced axon significantly.107 In cultured neurons, it was found that PrP106–126 elongation and branching,40 indicating that CRMP2 is necessary induces dystrophic changes in dendrites, an effect prevented for at least some of BDNF neurotrophic activities. In addition, by overexpression of CRMP3.108 It is worthwhile to note that in CRMP2 levels are decreased in the brain of patients with humans, the hippocampus, where CRMP3 is highly expressed, is depression.59 As antidepressants release neurotrophic factors protected from the extensive neuronal loss of the spongiform and/or modulate protein kinases such as GSK-3β,85 their degeneration until the final stage of disease.108 activities on dendrites and spines may be mediated in part by ● Epilepsy: epilepsy can be considered as a disorder of ion channel CRMPs. Indeed, proteomic analysis shows that chronic treat- dysfunction and abnormal brain networks.109 The pathogenic ment with two antidepressants, venlafaxine and fluoxetine, role of the genetic mutations of CDKL5, SYN1, ARHGEF9, DCX increases the level of CRMP2 in hippocampus by 4.0 and 2.6 and MeCP2 involves neuronal migration, neurite pruning, protein times, respectively.86 Tianeptine, an atypical antidepressant transporters, synaptogenesis, neurotransmitter release and func- with neurotrophic activity, also increases CRMP2 expression.87 tions of receptors. In patients with temporal lobe epilepsy, there The rapid antidepressant ketamine, an NMDA receptor antago- is a reduction in fiber density in the limbic structure110 and nist, also requires release of BDNF and inhibition of GSK3β thickness of the entorhinal cortex111 with abnormal global suggesting a common final pathway for antidepressant network efficiency.112 CRMP1 expression is decreased in the activity.88 Together, these observations suggest that CRMP2 temporal neocortex of patients with intractable temporal lobe may be involved in MDD and/or its treatment. In bipolar epilepsy113 as well as in pilocarpine-induced animal model of disorder, CRMP2 protein levels are decreased in the frontal epilepsy.113 Proteomic analysis of hippocampal tissue from cortex and CA2/CA3 areas of the hippocampus.59,70 CRMP4 is patients with mesial temporal lobe epilepsy revealed a drastic also decreased in the hippocampus.72 Genome-wide associa- decrease in CRMP2 protein.114 It is tempting to suggest that tion studies have detected several genetic markers for bipolar abnormal expression of the two proteins impairs neurite disorder, including CACNA1C,89 that encodes a subunit of the outgrowth and axonal pathfinding, contributing to the abnormal L-type voltage-dependent calcium channel shown to be structural connectivity found in epilepsy, although whether they associated with the dendritic arborization activity of CRMP3 in are a cause or an effect of seizures remained to be established. the hippocampus.29 The mood stabilizer lithium is a potent The anti-epileptic drug lacosamide binds directly to CRMP2 and GSK-3β inhibitor that inhibits CRMP2 phosphorylation in this property is proposed as one of its mechanisms of action.115 cultured rat cortical neurons.90 Lithium neurite elongation and ● Neuropathic pain (NP): sprouting after nerve injury probably branching activities may result in part from changes in CRMP2 induced by neurotrophic factors, for example, nerve growth phosphorylation. In bipolar patients, lithium treatment factor (NGF) for the sympathetic system,116–117 and maladaptive increases total grey matter and hippocampal volumes and spine remodeling, for example, in diabetic neuropathy,118 may improves white matter integrity altered by the disease,91–93 explain the neuron hyperexcitability and decrease in pain suggesting structural changes of neurites. Evidence of shared threshold characteristics of NP. Although N-type VGCCs (CaV2.2) genetic variations between psychiatric disorders, with high are genetically119 and clinically120 validated targets for pain correlation between schizophrenia and bipolar disorder, and management, the CaV2.2-selective blockers have untoward side moderate correlation between MDD and schizophrenia/bipolar effects. Interestingly, a CRMP-2-derived peptide (tat-CBD3;

Molecular Psychiatry (2015), 1037 – 1045 © 2015 Macmillan Publishers Limited CRMPs in neuron development and diseases TT Quach et al 1043 Figure 2; Supplementary Information, SI-B) that disrupts the 10 Wang LH, Strittmatter SM. A family of rat CRMP genes is differentially expressed CaV2.2/CRMP2 interaction proved to be antinociceptive in in the nervous system. J Neurosci 1996; 16: 6197–6207. animal models of inflammation and NP.121–124 Adeno- 11 Byk T, Ozon S, Sobel A. The Ulip family phosphoproteins: common and specific 254 – associated virus-encoded CBD3 delivered to rat peripheral properties. Eur J Biochem 1998; :14 24. 4 12 Inatome R, Tsujimura T, Hitomi T, Mitsui N, Hermann P, Kuroda S et al. Identifi- sensory neurons provides sustained relief ( 6 weeks) of NP. The cation of CRAM, a novel unc-33 gene family protein that associates with CRMP3 positive effect and relative lack of toxicity of the adeno- and protein tyrosine kinase(s) in the developing rat brain. J Biol Chem 2000; 275: associated virus-targeted expression of CBD3 peptide in this 27291–27302. animal model provide evidence of therapeutic promise for this 13 Quinn CC, Chen E, Kinjo TG, Bell AW, Eliott RC, McPherson PS et al. TUC-4b, type of approach to treat chronic NP125 and other disorders with a novel TUC family variant, regulates neurite outgrowth and associates with similar pathophysiological mechanisms. vesicles in the growth cone. J Neurosci 2003; 23: 2815–2823. 14 Fukata Y, Itoh TJ, Kimura T, Menager C, Nishimara T, Shiromizu T et al. CRMP2 binds to tubulin heterodimers to promote microtubule assembly. Nat Cell Biol CONCLUDING REMARKS 2002; 4: 583–591. There is no doubt that CRMPs have important roles in nervous 15 Uchida Y, Ohshima T, Yamashita N, Ogawara M, Sasaki Y, Nakamura F et al. system development and disease. They are intracellular mediators Sema3A signaling mediated by Fyn-dependent tyrosine phosphorylation of CRMP2 at tyrosine 32. J Biol Chem 2009; 284: 27393–27401. of neurotrophic factors/guidance cues regulating neurite structure/ 16 Leung T, Ng Y, Cheong A, Ng CH, Tan L, Hall C et al. P80 ROK alpha binding spine formation and are essential for dendrite patterning and protein is a novel variant of CRMP1 which associates with CRMP2 and modulates directional axonal pathfinding. The array of activities provided by RhoA-induced neuronal morphology. FEBS Lett 2002; 523: 445–449. the various members of the family depending on their amino-acid 17 Yamashita N, Morita A, Uchida Y, Nakamura F, Usui H, Ohshima T et al. sequences, their isoforms and their phosphorylation status encom- Regulation of spine development by semaphorin 3A through cyclin-dependent pass a variety of interrelated mechanisms for regulating dendrite kinase 5 phosphorylation of collapsin response mediator protein 1. J Neurosci 14 – and axon growth and retraction. Altogether, these observations 2007; : 12546 12554. 18 Yamashita N, Uchida Y, Ohshima T, Hira S, Nakamura F, Taniguchi M et al. CRMP1 indicate that CRMPs are important in maintaining the constantly mediates reelin signaling in cortical neuronal migration. J Neurosci 2006; 26: remodeling neuronal network that supports neuronal function. As 13357–13362. such, their involvement in neuropsychiatric diseases is not 19 Kawano Y, Yoshimura T, Tsuboi D, Kawabata S, Kaneko-Kawano T, Shirataki H surprising. Now that their involvement has been demonstrated et al. CRMP2 is involved in kinesin-dependent transport of the Sra-1/WAVE1 and is supported by data from animal models, and genetic linkage complex and axon formation. Mol Cell Biol 2005; 25:9920–9935. scanning, we expect more studies will focus on their role in the 20 Rosslenbroich V, Dai L, Baader SL, Gieselmann V, Kappler J. Collapsin response 310 – pathophysiology of these diseases. In a parallel effort, evidence mediator protein-4 regulates F-actin bundling. Exp Cell Res 2005; :434 444. 21 Cole AR, Knebel A, Morrice NA, Robertson LA, Irving AJ, Connolly CN et al. GSK-3 from pharmacological studies shows that CRMP activities are phosphorylation of the Alzheimer epitope within collapsin response mediator 75,86–87,90 modulated by various psychotropic drugs and suggests proteins regulates axon elongation in primary neurons. J Biol Chem 2004; 279: that they are compelling intracytoplasmic targets. Drug discovery 50176–50180. technologies targeting the ‘CRMP-activity/disease-change’ equation 22 Cole A. GSK-3 substrates in mood disorders and schizophrenia. FEBS J 2013; 280: are currently explored for neurological diseases,99,125,126 and may 5213–5227. bring new therapeutic interventions for psychiatric diseases as well. 23 Brennand JK, Simone A, Tran N, Gage FH. Modeling psychiatric disorders at the cellular and network levels. Mol Psychiatry 2012; 17: 1239–1252. 24 Medina M, Avila J. New insights into the role of GSK-3 in Alzheimer’s disease. CONFLICT OF INTEREST Expert Opin Ther Targets 2014; 18:69–77. The authors declare no conflict of interest. 25 Uchida Y, Ohshima T, Sasaki Y, Suzuki H, Yanai S, Yamashita N et al. Sema- phorin3A signaling is mediated via sequential Cdk5 and GSK3-β phosphorylation of CRMP2: implication of common phosphorylating mechanism underlying axon ACKNOWLEDGMENTS guidance and Alzheimer’s disease. Genes Cell 2005; 10:165–179. 26 Yashimata N, Mosinger B, Roy A, Miyazaki M, Ugagin K, Sasaki Y et al. CRMP5 We thank Dr A Frostholm, Department of Neuroscience, The Ohio State University, for regulates dendritic development and synaptic plasticity in the cerebellar her critical reading. Purkinje cells. J Neurosci 2011; 31: 1773–1779. 27 Yamashita N, Goshima Y. CRMPs regulate neuronal development and plasticity 16 – REFERENCES by switching their phosphorylation status. Mol Neurobiol 2012; : 933 944. 28 Brittain JM, Piekarz AD, Wang Y, Kondo T, Cummins TR, Khanna R. An atypical 1 Tessier-Lavigne M, Kolodkin AL. Neuronal guidance. Cold Spring Harb Perspect role for CRMP2 in neurotransmitter release via interaction with presynaptic Biol 2011; 3:1–14. voltage-gated calcium channels. J Biol Chem 2009; 284: 31375–31390. 2 Barnes AP, Polleux F. Establishment of axon-dendrite polarity in developing 29 Quach TT, Wilson SM, Rogemond V, Chounlamountri N, Kolattukudy PE, neurons. Ann Rev Neurosci 2009; 32: 347–381. Martinez S et al. Mapping CRMP3 domains involved in dendrite morphogenesis 3 Cheng P, Poo MM. Early events in Axon/Dendrite polarization. Annu Rev Neurosci and voltage-gated calcium channel regulation. J Cell Sci 2013; 126: 4262–4273. 2012; 35:181–201. 30 Alabed Y, Pool M, Ong Tone S, Fournier A. Identification of CRMP4 as a con- 4 Charrier E, Reibel S, Rogemond V, Aguera M, Thomasset N, Honnorat J. Collapsin vergent regulator of axon outgrowth inhibition. J Neurosci 2007; 27: 1702–1711. response mediator proteins (CRMPs): involvement in nervous system develop- 31 Bretin S, Reibel S, Charier E, Maus-Motti M, Auvergon N, Thevenoux A et al. ment and adult neurodegenerative disorders. Mol Neurol 2003; 28:51–64. Differential expression of CRMP1, CRMP2A, CRMP2B and CRMP5 in axons or 5 Schmidt EF, Strittmatter SM. The CRMP family of proteins and their role in dendrites of distinct neurons in the mouse brain. J Comp Neurol 2005; 486:1–17. Sema3A signaling. Adv Exp Med Biol 2007; 600:1–11. 32 Quach TT, Mosinger B, Ricard D, Copeland NC, Gilbert DJ, Jenkins JA et al. Col- 6 Goshima Y, Nakamura F, Strittmatter P, Strittmatter SM. Collapsin-induced lapsin response mediator protein 3/unc33-like protein 4 gene: organization, growth cone collapse mediated by an intracellular protein related to UNC-33. chromosomal mapping and expression in the developing mouse brain. Gene Nature 1995; 376: 509–514. 2000; 242:175–182. 7 Minturn JE, Fryer HJ, Geschwind DH, Hockfield S. Toad-64 a gene expressed early 33 Kwiatkowski AV, Rubinson DA, Dent EW, Van Veen JE, Leslie JD, Zhang J et al. in neuronal differentiation in the rat is related to unc-33, a C-elegans gene Ena/VASP is required for neuritogenesis in the developing cortex. Neuron 2007; involved in axon outgrowth. J Neurosci 1995; 15: 6757–6766. 56:441–455. 8 Byk T, Dobransky T, Sobel A. Identification and molecular characterization of 34 Brot S, Smaoune H, Youssef-Issa M, Malleval C, Benetollo C, Besancon R et al. Unc-33-like phosphoprotein (Ulip), a putative mammalian homolog of the axo- CRMP5 phosphorylation at Threonine 516 regulates neurite outgrowth inhibi- nal guidance-associate unc-33 gene product. J Neurosci 1996; 16:688–701. tion. Eur J Neurosci 2014; 40: 3010–3020. 9 Quach TT, Rong Y, Belin MF, Duchemin AM, Akaoka H, Ding S et al. Molecular 35 Brot S, Auger C, Bentata R, Rogemond V, Menigoz S, Chounlamountri N et al. cloning and expression of a new unc-33 like cDNA from rat brain and its relation CRMP5 induces mitophagy, thereby regulating mitochondrion numbers in to paraneoplastic neurological syndromes. Mol Brain Res 1997; 46: 329–332. dendrites. J Biol Chem 2014; 289:2261–2276.

© 2015 Macmillan Publishers Limited Molecular Psychiatry (2015), 1037 – 1045 CRMPs in neuron development and diseases TT Quach et al 1044 36 Brot S, Rogemond V, Perrot V, Chounlamountri N, Auger C, Honnorat J et al. 62 Beasley CL, Pennington K, Behan A, Wait R, Dunn MJ, Cotter D. Proteomic CRMP5 interacts with tubulin to inhibit neurite outgrowth, thereby modulating analysis of the anterior cingulate cortex in the major psychiatric disorders: the function of CRMP2. J Neurosci 2010; 30: 10639–10654. Evidence for disease-associated changes. Proteomics 2006; 6: 3414–3425. 37 Niisato E, Nagai J, Yamashita N, Abe T, Kiyonari H, Goshima Y et al. CRMP4 63 Clark D, Dedova I, Cordwell S, Matsumoto I. A proteome analysis of the anterior suppresses apical dendrite bifurcation of CA1 pyramidal neurons in the mouse cingulate cortex gray matter in schizophrenia. Mol Psychiatry 2006; 11: 459–470. hippocampus. Dev Neurobiol 2012; 72:1447–1457. 64 Pennington K, Dicker P, Dunn MJ, Cotter DR. Proteomic analysis reveals protein 38 Yamashita N, Ohshima T, Nakamura F, Kolattukudy P, Honnorat J, Goshima Y. changes within layer 2 of the insular cortex in schizophrenia. Proteomics 2008; 8: Phosphorylation of CRMP2 is involved in proper dendritic field organization. 5097–5107. J Neurosci 2012; 32:1360–1365. 65 Fallin MD, Lasseter VK, Liu Y, Avramopoulos D, McGrath J et al. Linkage and 39 Inagaki N, Chihara K, Arimura N, Menager C, Kawano Y, Matsuo N et al. CRMP2 association on 8p21.2-p21.1 in Schizophrenia. Am J Med Genet B Neuropsychiatr induces axons in cultured hippocampal neurons. Nat Neurosci 2001; 4:781–782. Genet 2011; 156:188–197. 40 Yoshimura T, Kawano Y, Arimura N, Kikuchi A, Kaibuchi K. GSK-3β regulates 66 Blouin BA, Dombroski SK, Nath VK, Lasseter VK, Wolyniec PS, Nestadt G et al. phosphorylation of CRMP2 and neuronal polarity. Cell 2005; 120: 137–149. Schizophrenia susceptibility loci on chromosomes13q32 and 8p21. Nat Genet 41 Nishimura T, Fukata Y, Kato K, Yamaguchi T, Matsuura Y, Kamiguchi H et al. 1998; 20:70–73. CRMP-2 regulates polarized Num-mediated endocytosis for axon growth. Nat 67 Gurling HM, Kalsi G, Brynjolson J, Sigmundsson T, Sherrington R, Mankoo BS Cell Biol 2003; 5: 819–825. et al. Genomewide genetic linkage analysis confirms the presence of suscept- 42 Chi XX, Schmutzier BS, Brittain JM, Wang Y, Hintgen C, Nicol GD et al. Regulation ibility loci for schizophrenia on 1q32.2, 5q33.2 and 8p21.22 and of N-type voltage-gated calcium channels (Cav2.2) and transmitter release by provides support for linkage to schizophrenia on chromosomes 111q23.3 and CRMP2 in sensory neurons. J Cell Sci 2009; 122: 4351–4362. 20q21.1-11-23. Am J Hum Genet 2001; 68: 661–673. 43 Quach TT, Duchemin AM, Rogemond V, Aguera M, Honnorat J, Belin MF et al. 68 Holmans PA, Riley B, Pulver AE. Genomewide linkage scan of schizophrenia in a Involvement of collapsin response mediator proteins in the neurite extension large multicenter pedigree sample using single nucleotide polymorphisms. Mol produced by neurotrophins in dorsal root ganglion neurons. Mol Cell Neurosci Psychiatry 2009; 14:786–795. 2004; 25: 433–443. 69 Nakata K, Ujike H, Sakai A, Takaki M, Imamura T, Tanaka Y et al. The human 44 Charrier E, Mosinger B, Meisserel C, Aguera M, Rogemond V, Reibel S et al. dihydropyrimidinase-related protein 2 gene on chromosome 8p21 is associated Transient alterations in granule cells proliferation, apoptosis and migration in with paranoid-type schizophrenia. Biol Psychiatry 2003; 53: 571–576. postnatal developing cerebellum of CRMP1 − / − mice. Genes Cells 2006; 11: 70 Liu Y, Pham X, Zhang L, Chen PL, Burzynski G, McGaughey DM et al. Functional 1337–1352. variants in DPYSL2 sequence increase risk of Schizophrenia and suggest a link to 45 Ghogha A, Bruun DA, Lein P. Inducing dendritic growth in cultured sympathetic mTOR signaling. G3 (Bethesda) 2015; 5:61–72. neurons. J Vis Exp 2012; 61: e3546. 71 Koide T, Aleksic B, Ito Y, Usui H, Yoshimi A, Inada T et al. A two-stage case-control 46 Su SK, Chien WL, Fu WM, Yu IS, Huang HP, Huang PH et al. Mice deficient in association study of the dihydropyrimidinase-like 2 gene (DPYSL2) with CRMP1 exhibit impaired long-term potentiation and impaired spatial learning schizophrenia in Japanese subjects. J Hum Genet 2010; 55: 469–472. and memory. J Neurosci 2007; 27:2513–2524. 72 Föcking M, Dicker P, English JA, Schubert KO, Dunn MJ, Cotter DR. Common 47 Yamashita N, Takashi A, Takao K, Yamamoto T, Kolattukudy, Miyakawa T. Mice proteomic changes in the hippocampus in schizophrenia and bipolar disorder lacking CRMP1 manifest hyperactivity, impaired learning and memory, and and particular evidence for involvement of cornu ammonis regions 2 and 3. Arch impaired prepulse inhibition. Front Behav Neurosci 2013; 7: 216. Gen Psychiatry 2011; 68:477–488. 48 Quach TT, Massicote G, Belin MF, Honnorat J, Glasper ER, Devries AC et al. CRMP3 73 Hill MJ, Donocik JG, RA Nuamah, CA Mein. Sainz-Fuertes R1 Transcriptional is required for hippocampal CA1 dendritic organization and plasticity. FASEB J consequences of schizophrenia candidate miR-137 manipulation in human 2008; 2: 401–409. neural progenitor cells. Schizophr Res. 2014; 153: 225–230. 49 Demyanenko GK, Tsai AY, Maness PF. Abnormalities in neuronal process 74 Miller BH, Zeier Z, Xi L, Lanz TA, Deng S, Strathmann J et al. MicroRNA-132 extension, hippocampal development, and the ventricular system of L1 dysregulation in schizophrenia has implications for both neurodevelopment and knockout mice. J Neurosci 1999; 19: 4907–4920. adult brain function. Proc Natl Acad Sci USA 2012; 109: 3125–3130. 50 Camdessanche JP, Ferraud K, Boutahar N, Lassabliere F, Mutter M, Touret M et al. 75 Kedracka-Krok S, Swiderska B, Jankowska U, Skupien-Rabian B, Solich J, Buczak K The CRMP5 onconeural protein is expressed in Schwann cells under axonal et al. Clozapine influences cytoskeleton structure and calcium homeostasis in rat signals and regulates axon-Schwann cell interaction. J Neuropathol Exp Neurol cerebral cortex and has a different proteomic profile than risperidone. J Neuro- 2012; 71: 298–311. chem 2015; 132: 657–676. 51 Rapoport JL, Gield JN, Gogtay N. Neurodevelopmental model of schizophrenia: 76 Pickering C, Ericson M, Söderpalm B. Chronic phencyclidine increases synapsin-1 update 2012. Mol Psychiatry 2012; 17:1228–1238. and synaptic adaptation proteins in the medial prefrontal cortex. ISRN Psychiatry 52 Penzes P, Cahill ME, Jones KA, Vanderleuwen JE, Woolfrey KM. Dendritic spine 2013; 620361. pathology in neuropsychiatric disorders. Nat Neurosci 2011; 14: 285–293. 77 Mateus-Pinheiro A, Patrício P, Bessa JM, Sousa N, Pinto L. Cell genesis and 53 Jones LB, Johnson N, Byne N. Alterations in MAP2 immunocytochemistry in area dendritic plasticity: a neuroplastic pas de deux in the onset and remission from 9 and 32 of schizophrenic prefrontal cortex. Psychiatry Res 2002; 114:137–148. depression. Mol Psychiatry 2013; 18: 748–750. 54 Hall J, Trent S, Thomas KL, O'Donovan MC, Owen MJ. Genetic risk for schizo- 78 Dimatelis JJ, Stein DJ, Russel VA. Chronic exposure to light reverses the effect of phrenia: convergence on synaptic pathways involved in plasticity. Biol Psychiatry maternal separation on proteins in the prefrontal cortex. J Mol Neurosci 2013; 51: 2015; 77:52–58. 835–843. 55 Bader V, Tomppo L, Trossbach SV, Bradshaw NJ, Prikulis I, Leliveld SR et al. 79 Duman RS, Monteggia LM. A neurotrophic model for stress-related mood Proteomic, genomic and translational approaches identify CRMP1 for a role in disorders. Biol Psychiatry 2006; 59: 1116–1127. schizophrenia and its underlying traits. Hum Mol Genet 2012; 21: 4406–4418. 80 Molendjik ML, Spinhoven P, Polak M, Elzinger BM. Serum BDNF concentrations 56 Quach TT, Glasper ER, Devries AC, Honnorat J, Kolattukudy PE. Altered prepulse as peripheral manifestations of depression: evidence from a systematic review inhibition in mice with dendrite abnormalities of hippocampal neurons. Mol and meta-analysis on 179 association (N: 9484). Mol Psy 2014; 19:791–800. Psychiatry 2008; 13:656–658. 81 Brunoni AR, Lopes M, Fregni F. A systematic review and meta-analysis of 57 Hensley K, Venkova K, Christov A, Gunning W, Park J. CRMP2: an emerging clinical studies on major depression and BDNF levels: implications for the pathologic feature and therapeutic target for neurodisease indications. Mol role of neurplasticity in depression. Int. J. Neuropsychopharmacol 2008; 11: Neurobiol 2011; 43:180–191. 1169–1180. 58 Martins de Souza D, Dias-Neto E, Schmitt A, Falkai P, Gormanns P, MacCarrone G 82 Wong ML, Dong C, Andreev V, Arcos-Burgos M, Licinio J. Prediction of sus- et al. Proteome analysis of schizophrenia brain tissue. World J Biol Psychiatry ceptibility to major depression by a model of interactions of multiple functional 2010; 11: 110–120. genetic variants and environment factors. Mol Psychiatry 2012; 17:624–633. 59 Johnston-Wilson NL, Sims CD, Hofmann JP, Anderson L, Shore AD, Torrey EF 83 Magarinos AM, Li CJ, Bath KC, Jing FS, Lee FS, McEven BS. Effect of brain-derived et al. Disease-specific alterations in frontal cortex brain proteins in schizophrenia, neurotrophic factor haploinsufficiency on stress-induced remodeling of hippo- bipolar disorder, and major depressive disorder. Mol Psychiatry 2000; 5:142–149. campal neurons. Hippocampus 2011; 21:253–264. 60 Edgar PF, Douglas JE, Cooper GJ, Dean B, Kydd R, Faull RL. Comparative pro- 84 Chen ZY, Jing D, Bath KG, Ieraci A, Khan T, Siao CJ et al. Genetic variant BDNF teome analysis of the hippocampus implicates chromosome 6q in schizo- (Val66Met) polymorphism alters anxiety-related behavior. Science 2006; 314: phrenia. Mol Psychiatry 2000; 5:85–90. 140–143. 61 Prabakaran S, Swatton JE, Ryan MM, Huffaker SJ, Huang JT, GriffinJLet al. 85 Schmidt HD, Duman RS. The role of neurotrophic factors in adult hippocampal Mitochondrial dysfunction in schizophrenia: evidence for compromised brain neurogenesis, antidepressant treatments and animal models of depressive-like metabolism and oxidative stress. Mol Psychiatry. 2004; 9:684–697. behavior. Behav Pharmacol 2007; 18: 391–418.

Molecular Psychiatry (2015), 1037 – 1045 © 2015 Macmillan Publishers Limited CRMPs in neuron development and diseases TT Quach et al 1045 86 Khawaja X, Xu J, Liang JJ, Barrett JE. Proteomic analysis of proteins changes 108 Quach TT, Wang Y, Khanna R, Chounlamountri N, Auvergnon N, Honnorat J et al. developing in rat hippocampus after chronic antidepressants treatment: impli- Effect of CRMP3 expression on dystrophic dendrites of hippocampal neurons. cations for depressive disorders and future therapies. J Neurosci Res 2004; 75: Mol Psychiatry 2011; 16:689–691. 451–460. 109 Hirose S. A new paradigm of channelopathy in epilepsy syndromes: intracellular 87 Hensley K, Venkova K, Christov A, Gunning W, Park J. Collapsin response med- trafficking abnormality of ion channels dysfunction and channel molecules. iator protein 2: an emerging pathologic feature and therapeutic target for Epilepsy Res 2006; 70S: S206–S217. neurodisease indications. Mol Neurobiol 2011; 43:180–191. 110 Bonilha L, Nesland T, Martz GU, Joseph JE, Spampinato MW, Tabesh A. Medial 88 Duman RS, Aghajanian GK. Synaptic dysfunction in depression: Potential temporal lobe epilepsy is associated with neuronal fiber loss and paradoxal therapeutic targets. Science 2012; 338:68–72. increase in structural connectivity in limbic structures. J Neurol Neurosurg 89 Ferreira MA, O'Donovan MC, Meng YA, Jones IR, Ruderfer DM, Jones L et al. Psychiatry 2012; 83:903–909. Collaborative genome-wide association analysis supports a role for ANK3 and 111 Bernhardt BC, Worsley KJ, Besson J, Concha L, Lerch JP, Evans AC et al. Mapping CACNA1C in bipolar disorders. Nat Genet 2008; 40: 1056–1058. limbic network organization in temporal lobe epilepsy using morphometric 90 Caberlotto L, Carboni L, Zanderigo F, Andreetta F, Andreoli M, Gentile G et al. correlations insights on the relation between mesiotemporal connectivity and Differential effects of glycogen synthase kinase 3 (GSK3) inhibition by lithium cortical atrophy. Neuroimage 2008; 42:512–524. or selective inhibitors in the central nervous system. Naunyn Schmiedebergs 112 DeSalvo MN, Douw L, Tanaka N, Reinsberger C, Stufflebeam SM. Altered struc- Arch Pharmacol 2013; 386:893–903. tural connectome in temporal lobe epilepsy. Radiology 2014; 270:842–848. 91 Lyoo IK, Dager SR, Kim JE, Yoon SJ, Friedman SD, Dunner DL et al. Lithium- 113 Luo J, Zeng K, Zhang C, Fang M, Zhu Q, Wang L et al. Down-regulation of CRMP1 induced gray matter volume increase as a neural correlate of treatment in patients with epilepsy and a rat model. Neurochem Res 2012; 37: 1381–1391. response in bipolar disorder: a longitudinal brain imaging study. Neuropsycho- 114 Czech T, Yang JW, Csaszar E, Kappler J, Baumgardner C, Lubec G. Reduction of pharmacology 2010; 35: 1743–1750. hippocampal CRMP2 in patients with mesial temporal lobe epilepsy. Neurochem 92 Hajek T, Cullis J, Novak T, Kopecek M, Höschl C, Blagdon R et al. Hippocampal Res 2004; 29: 2189–2196. volumes in bipolar disorders: opposing effects of illness burden and lithium 115 Wang Y, Brittain J, Jarecki BW, Wilson S, Wang B, Hale R et al. In silico docking treatment. Bipolar Disord 2012; 14:261–270. and electrophysiological characterization of Lacosamide binding sites on CRMP2 93 Gildengers AG, Butters MA, Aizenstein HJ, Marron MM, Emanuel J, Anderson SJ identifies a pocket important in modulating sodium channels slow inactivation. et al. Longer lithium exposure is associated with better white matter integrity in J Biol Chem 2010; 285: 25296–25307. older adults with bipolar disorder. Bipolar Disord 2014; 17:248–256. 116 Bridges D, Thompson SWN, Rice ASC. Mechanisms of neuropathic pain. Br J 94 Cross-Disorder Group of the Psychiatric Genomics Consortium. Genetic rela- Anaesthesia 2001; 87:12–26. tionship between five psychiatric disorders estimated from genome-wide SNPs. 117 Cohen SP, Mao J. Neuropathic pain: mechanisms and their clinical implications. Nat Genet 2013; 45:984–994. Br Med J 2014; 348: f7656. 95 Frisoni GB, Sabattoli F, Lee AD, Dutton RA, Thompson PM. In vivo neuro- 118 Tan AM, Samad OA, Fischer TZ, Zhao P, Persson AK, Waxman SG. Maladaptive pathology of the hippocampal formation in Alzheimer’s disease: a radial map- dendritic spine remodeling contributes to diabetic neuropathic pain. J Neurosci ping MR-based study. Neuroimage 2006; 32:104–110. 2012; 32: 6795–6807. 96 Mucke L, Selkoe DJ. Neurotoxicity of amyloid β-protein: synaptic and network 119 Kim C, Jun K, Lee T, Kim SS, McEnery MW, Chin H et al. Altered nociceptive dysfunction. Cold Spring Harb Perspect Med 2012; 2:1–17. response in mice deficient in the alpha(1B) subunit of the voltage-dependent 97 Lu T, Aron L, Zullo J, Pan Y, Chen Y, Yang TH et al. REST and stress resistance in calcium channel. Mol Cell Neurosci 2001; 18:235–245. aging and Alzheimer’s disease. Nature 2014; 507:448–454. 120 Cao YQ. Voltage-gated calcium channels and pain. Pain 2006; 126:5–9. 98 Woods NK, Padmanabhan J. Neuronal calcium signaling and Alzheimer disease. 121 Brittain JM, Duarte DB, Wilson SM, Zhu W, Ballard C, Johnson PL et al. Sup- Adv Exp Med Biol 2012; 740: 1193–1217. pression of inflammatory and neuropathic pain by uncoupling CRMP-2 from the 99 Soutar MP, Thornhill P, Cole AR, Sutherland C. Increased CRMP2 phosphorylation presynaptic Ca2+ channel complex. Nat Med 2011; 17:822–829. is observed in Alzheimer's disease; does this tell us anything about disease 122 Ju W, Li Q, Allette YM, Ripsch MS, White FA, Khanna R. Suppression of development? Curr Alzheimer Res 2009; 6:269–278. pain-related behavior in two distinct rodent models of peripheral neuropathy 100 Cole AR, Noble W, Van Aalten L et al. CRMP2 hyperphosphorylation is by a homopolyarginine-conjugated CRMP2 peptide. J Neurochem 2012; 124: an early event in Alzheimer’s disease progression. J Neurochem 2007; 103: 869–879. 1132–1144. 123 Ripsch MS, Ballard CJ, Khanna M, Hurley JH, White FA, Khanna R. A peptide 101 Williamson R, van Alten L, Mann DM, Platt B, Plattner F, Bedford L et al. CRMP2 uncoupling CRMP2 from the presynaptic Ca(2+) channel complex demonstrates hyperphosphorylation is characteristic of Alzheimer's disease and not a feature efficacy in animal models of migraine and AIDS therapy-induced neuropathy. common to other neurodegenerative diseases. J Alzheimers Dis 2011; 27: Transl Neurosci 2012; 3:1–8. 615–625. 124 Piekarz AD, Due MR, Khanna M, Wang B, Ripsch MS, Wang R et al. CRMP2 102 Wang Y, Yin H, Li J, Zhang Y, Han B, Zeng Z et al. Amelioration of β-amyloid- peptide mediated decrease of high and low voltage-activated Ca2+ channels, induced cognitive dysfunction and hippocampal axon degeneration by curcu- attenuation of nociceptor excitability, and anti-noception in a model of AIDS min is associated with suppression of CRMP-2 hyperphosphorylation. Neurosci therapy-induced painful peripheral neuropathy. Mol Pain 2012; 8:1–19. Lett 2013; 557: 112–117. 125 Fischer G, Pan B, Vilceam D, Hogan OH, Yu H. Sustained relief of neuropathic 103 Marshall KE, Marchante R, Serpell LC. The relationship between amyloid struc- pain by AAV-targeted expression of CBD3 peptide in rat DRG. Gene Ther 2014; ture and cytotoxicity. Prion 2014; 8:2. 21:44–51. 104 Prusiner SB. Biology and genetics of prions causing neurodegeneration. Ann Rev 126 Petratos S, Ozturk E, Azari MF, Kenny R, Lee JY, Magee KA et al. Limiting multiple Genet 2013; 47:601–623. sclerosis related axonopathy by blocking Nogo receptor and CRMP-2 phos- 105 Auvergnon N, Reibel S, Touret M, Honnorat J, Baron T, Giraudon P et al. Altered phorylation. Brain 2012; 135: 1794–1818. expression of CRMPs in the brain of bovine spongiform encephalopathy- 127 Tsutiya A, Ohtani-Kaneko R. Post-natal alteration of CRMP4 mRNA expression in infected mice during disease progression. Brain Res 2009; 1261:1–6. the mouse brain. J Anat 2012; 221: 341–351. 106 Ishikura N, Clever JL, Bouzamondo-Berstein E, Prusiner SB, Huang EJ, 128 Ricard D, Rogemond V, Charrier E, Bagnard D, Belin MF, Thomasset N et al. DeArmond SJ. Notch-1 activation and dendritic dystrophy in prion disease. Isolation and expression of human Ulip6/CRMP5: coexistence with Ulip2/CRMP2 Proc Natl Acad Sci USA 2005; 102:886–891. in Sema3A-sensitive oligodendrocytes. J Neurosci 2001; 21: 7203–7214. 107 Shinkai-Ouchi F, Yamakawa Y, Hara H, Tobiume M, Nishijima M, Hagiwara K. 129 Marques JM, Rodrigues RJ, Valbuena S, Rozas JL, Selak S, Marin P et al. CRMP2 Identification and structural analysis of C-terminally truncated CRMP2 in a tethers kainate receptor activity to cytoskeleton dynamics during neuronal murine model of prion diseases. Proteome Sci 2010; 8:53–65. maturation. J Neurosci 2013; 33: 18298–18310.

Supplementary Information accompanies the paper on the Molecular Psychiatry website (http://www.nature.com/mp)

© 2015 Macmillan Publishers Limited Molecular Psychiatry (2015), 1037 – 1045