Elsevier

Experimental Cell Research

Volume 319, Issue 15, 10 September 2013, Pages 2368-2374
Experimental Cell Research

Review Article
Rho GTPase signaling at the synapse: Implications for intellectual disability

https://doi.org/10.1016/j.yexcr.2013.05.033Get rights and content

Abstract

Intellectual disability (ID) imposes a major medical and social–economical problem in our society. It is defined as a global reduction in cognitive and intellectual abilities, associated with impaired social adaptation. The causes of ID are extremely heterogeneous and include non-genetic and genetic changes. Great progress has been made over recent years towards the identification of ID-related genes, resulting in a list of approximately 450 genes. A prominent neuropathological feature of patients with ID is altered dendritic spine morphogenesis. These structural abnormalities, in part, reflect impaired cytoskeleton remodeling and are associated with synaptic dysfunction. The dynamic, actin-rich nature of dendritic spines points to the Rho GTPase family as a central contributor, since they are key regulators of actin dynamics and organization. It is therefore not surprising that mutations in genes encoding regulators and effectors of the Rho GTPases have been associated with ID. This review will focus on the role of Rho GTPase signaling in synaptic structure/function and ID.

Introduction

The human brain is one of the most complex and fascinating organs, however it is also one of the most vulnerable ones. This is illustrated by the high incidence of significantly reduced cognitive and adaptive performance, grouped under the term intellectual disability (ID). ID, previously called mental retardation, is a common neurodevelopmental disorder arising before the age of 18 that affects approximately 1–3% of the population. ID is defined by an intelligence quotient (IQ) lower than 70, associated with deficits in conceptual, social and adaptive skills [1]. The severity of ID can be divided into mild (IQ between 50 and 69), moderate (IQ between 35 and 49), severe (IQ between 20 and 34) and profound ID (IQ lower than 20) [1]. Clinically, ID has been grouped into non-syndromic ID, which is characterized by impaired cognitive function without any other clinical features, and syndromic ID in which patients additionally present multiple biological or metabolic defects. The causes of ID are highly heterogeneous and include environmental factors that influence the development of nervous system, and/or genetic factors, such as chromosomal aneuploidies and single gene mutations [2]. Over the past two decades, great efforts have been made to identify ID genes, resulting in a list of approximately 450 ID causing genes [1]. Several of these genes are associated with severe brain abnormalities, such as neuronal heterotopia, lissencephaly, and microcephaly. A vast number of other genes however have been associated with ID disorders with no apparent gross abnormalities in brain structure and architecture. Because learning deficits are a constant feature of ID patients, it is tempting to attribute some of ID traits to alterations in synaptic function. Several lines of evidences point into this direction. First, a significant proportion of the ID-related proteins are enriched at pre- and/or post-synaptic compartments [3]. Second, alterations in dendritic spines, actin-rich structures on which most excitatory synapses in the brain are located, are observed in patients with ID. Such alterations are recapitulated in mouse models of ID, including Fragile X, which is the most common form of ID. Third, recent studies have provided functional evidence for alterations in synaptic strength in models of ID. These observations support the notion that many forms of ID, may share a common synaptic component [4]. For most of the ID genes identified, little information is available as to how mutations in these genes result in cognitive impairment. The functions of these genes vary largely including regulating cell adhesion, Rho and Ras signaling, gene transcription and chromatin remodeling. Given the actin-rich nature of dendritic spines, the Rho GTPase family, as key regulator of actin dynamics and organization, has recently received much attention in ID research [5], [6]. Fig. 1 Indeed, mutations in regulators and effectors of the Rho GTPases have been found to underlie various forms of ID. In this review, we first briefly discuss the role of Rho GTPase signaling in spine and synapse formation/plasticity and subsequently describe in more detail some of the Rho GTPase signaling pathways involved in ID.

Rho-family GTPases belong to the Ras superfamily of small (±21 kDa) GTPases. So far, 22 human members of the Rho family have been identified and they can be divided in 8 different subgroups: Rho (RhoA–RhoC); Rac (Rac1–Rac3, RhoG); Cdc42 (Cdc42, TC10, and TCL, Chp, Wrch-1); RhoD (RhoD and Rif); Rnd (Rnd1, Rnd2, RhoE/Rnd3); RhoH/TTF; RhoBTB (RhoBTB1 and RhoBTB2) and Miro (Miro-1 and Miro-2) [7]. Rho proteins are guanine nucleotide-binding proteins, which act as molecular switches cycling between an active GTP-bound form and an inactive GDP-bound form. The activity of Rho GTPases is spatio-temporally regulated by positive regulator guanine nucleotide exchange factors (GEFs), and negative regulators GTPases activating proteins (GAPs) and guanine nucleotide dissociation inhibitors (GDIs). In neurons the Rho GTPases, in particular Rac1, Cdc42 and RhoA, are best known for their effects on the actin cytoskeleton [8]. At the synapse Rac1, Cdc42 and RhoA have emerged as key regulators of spine formation and morphogenesis, and more recently have been implicated in synaptic plasticity [9]. In the central nervous system most of the excitatory synapses are located on dendritic spines. Spines are highly enriched in filamentous actin, and their ability to change shape depends on the rapid remodeling of the spine actin cytoskeleton [10], [11]. Rac1 and Cdc42 have been shown to promote the formation, growth and maintenance of spines, whereas RhoA induces spine retraction and loss [9]. Emerging evidence also points to a pivotal role for Rho GTPase signaling in the regulation of synaptic plasticity. Changes in the morphology of dendritic spines underlie long-term potentiation (LTP) and long-term depression (LTD), two processes that model the activity-dependent changes of synaptic strength and which are believed to represent the cellular basis of learning and memory. For both the formation of stable LTP and LTD, enduring actin-dependent structural changes in spine heads are required. These structural changes are strongly associated with the incorporation (LTP) or removal (LTD) of synaptic α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor (AMPARs) [12]. Indeed, inhibition of actin polymerization attenuates LTP maintenance, whereas LTD is associated with actin filament disassembly [13]. ADF/cofilin has emerged as a critical downstream Rho GTPase effector protein involved in spine enlargement (LTP) and shrinkage (LTD). ADF/cofilin, whose activity depends on its phosphorylation state, can be regulated by various Ca2+-dependent kinases (CaMKII), phosphatases (calcineurin and PP1) and Rho GTPase activity (LIMK). When ADF/cofilin is phosphorylated on Ser-3, and thus inactivated, it facilitates actin polymerization. Dephosphorylation of ADF/cofilin has been associated with spine shrinkage upon low-frequency stimulation induced LTD in hippocampal brain slices, which was dependent on NMDAR activation [14]. Conversely, ADF/cofilin was found to be phosphorylated, upon LTP signaling [15]. A recent study found that NMDAR-dependent LTP sets in motion two distinct signaling cascades that involve ADF/cofilin phosphorylation: a RhoA-dependent pathway, which is required for the initial induction of LTP, and a Rac1-dependent pathway that is required for the later LTP consolidation phase. The distinct functional roles of the RhoA and Rac1 cascade in LTP therefore suggest the sequential activity-dependent remodeling of actin involving first RhoA followed by Rac1 signaling [15]. One possible explanation would imply that RhoA and Rac1 each target a distinct pool of actin in dendritic spines and thereby contribute to different aspects of LTP-induced actin remodeling [11]. More recently an elegant study combined two-photon fluorescence lifetime imaging microscopy with glutamate uncaging to directly monitor the activity of RhoA and Cdc42 in single dendritic spines upon LTP stimuli in CA1 neurons. Both RhoA and Cdc42 were rapidly activated in a CaMKII-dependent manner. Whereas activated Cdc42 was restricted to the stimulated spine, activated RhoA diffused into neighboring spines [16]. Together these results indicate that a coordinated action of Rac, Cdc42 and RhoA is required for the actin-dependent structural changes underlying LTP [15], [17]. Further evidence into how actin remodeling and AMPAR trafficking are regulated in LTP/LTD have been gathered from recent studies in which Rho-GEFs, Rho-GAPs and downstream effectors have been implicated in various forms of synaptic plasticity (see below).

Oligophrenin-1 (OPHN1), the first identified Rho-linked ID gene, encodes a Rho-GAP protein that negatively regulates RhoA in neurons [18]. Besides RhoA, OPHN1 also interacts with Homer 1b/c [18], endophilin A1 and A2/3 (EndoA1, A2/3) [19], amphiphysins, and CIN85 [20]. The protein is abundantly expressed in the brain, where it is found in neurons of all major regions, including hippocampus and cortex, and is present in axons, dendrites and spines [18]. Families with mutations in the OPHN1 gene suffer from syndromic ID with cerebellar abnormalities, of which cerebellar hypoplasia is the most recurring phenotype [21], [22], [23], [24].

All OPHN1 mutations identified to date appear to result in a loss of protein function. Recent studies have provided insight as to how mutations in OPHN1 may impact neuronal function. Knockdown of Ophn1 in CA1 pyramidal neurons in hippocampal slices resulted in a significant decrease in dendritic spine length. This phenotype was mimicked by using a constitutive active RhoA mutant and was rescued by inhibiting a key effector of RhoA, Rho-kinase (ROCK) [18]. In accordance, Nadif Kasri and colleagues demonstrated that knocking down Ophn1 in hippocampal neurons inhibited AMPAR- and NMDAR-mediated currents and impeded synaptic maturation and both structural and functional LTP [25]. These defects were dependent on Rho-GAP activity, and the interaction of OPHN1 with Homer 1b/c [25], [26]. Conversely, overexpressing OPHN1 selectively enhanced AMPAR-mediated synaptic transmission and increased spine density [25]. Using a peptide derived from the C-terminus of the AMPAR GluA2 subunit that blocks AMPA receptor endocytosis [27], they have further demonstrated that OPHN1 can regulate synaptic function and structure by stabilizing synaptic AMPA receptors. A model was therefore proposed in which OPHN1 during early development contributes to excitatory synapse maturation at the CA3–CA1 synapse by stabilizing AMPARs. More recently OPHN1 was also found to play an important role in metabotropic glutamate receptor-induced LTD (mGluR-LTD). OPHN1 expression was rapidly increased at the synapse upon mGluR1 stimulation, an increase in OPHN1 that was required for the proper expression of mGluR-LTD. Interestingly, the role of OPHN1 in mediating mGluR-LTD could molecularly be dissociated from its role in basal AMPAR-mediated synaptic transmission [26]. Whereas the former required OPHN1's interaction with EndoA2/3, the latter requires OPHN1's Rho-GAP activity and interaction with the Homer 1b/c proteins. Recent studies in Ophn1 KO mice provided further evidence for the importance of OPHN1 in spine morphology and the regulation of AMPA receptor trafficking [20], [28]. Disruption of OPHN1 in mice was shown to increase the number of filopodia-like spines and impaired NMDA receptor-dependent LTD. Interestingly, the deficits in AMPAR endocytosis and LTD could be reversed completely using a Rho kinase inhibitor. At the pre-synaptic compartment, a reduction in endocytosis of synaptic vesicles was observed in cortical neurons of Ophn1 KO mice. Additionally, Nakano-Kobayashi and colleagues demonstrated that OPHN1 controls pre-synaptic vesicle cycling in CA3 neurons by forming a complex with EndoA1 [19]. Besides controlling excitatory synaptic transmission, OPHN1 is also important for inhibitory synaptic transmission. In addition, impairments in presynaptic vesicle recycling and a reduction in the readily releasable pool were recently observed in inhibitory synapses in the dentate gyrus of Ophn1 KO mice [29]. Interestingly, pharmacologic treatment with Rho kinase inhibitors rescued these deficits. Finally, OPHN1 has been implicated in circadian clock regulation through its interaction with the nuclear receptor Rev-erbα in the hippocampus [30]. Accordingly, they observed circadian rhythm deficits in Ophn1 KO mice. This is particularly interesting since sleep disturbances are common in children and adults with ID [31]. Taken together these studies clearly illustrate the multifaceted function of OPHN1 in spine morphology, synaptic plasticity and vesicle recycling at the pre- and post-synaptic terminals.

Mutations in ARHGEF6, also known as alphaPIX or Cool-2, have been found in patients with non-syndromic ID. The first mutation was identified in a male ID patient carrying a reciprocal X;21 translocation breakpoint located between exons 10 and 11 of the ARHGEF6 gene. Additionally, mutations were detected in the first intron of this gene [32]. ARHGEF6 codes for a Cdc42/Rac1 exchange factor, which was initially isolated as a PAK-interacting protein. More recent studies have unraveled the complex nature of its GEF activity towards Rac1 and Cdc42. In its dimeric form, ARHGEF6 functions as a Rac1-specific GEF, whereas in its monomeric form, ARHGEF6 can act as a GEF for Rac1 and Cdc42, but only upon binding of PAK to its SH3 domain [33]. In the brain, ARHGEF6 is highly expressed in hippocampus compared to cortex and cerebellum. Previous studies showed that, in cultured neurons, overexpressed ARHGEF6 localizes specifically at the post-synaptic compartment of excitatory synapses, suggesting an important synaptic function of ARHGEF6 [34]. Subsequently, several lines of evidence have demonstrated a crucial role of ARHGEF6 in regulating spine morphology, synaptogenesis and synaptic plasticity. Knocking down Arhgef6 by RNAi in cultured neurons resulted in a decrease of large mushroom spines as well as an increase of elongated spines and filopodia-like protrusions. This phenotype could be rescued by a constitutively active form of PAK3 [34]. More recently, Arhgef6 KO mice showed longer and more branched dendrites and an increase in the density of spines in CA1 pyramidal neurons. Surprisingly, the increase in dendritic spines on apical dendrites was accompanied by a reduction in excitatory synapses. In addition, Arhgef6 KO mice showed a significant impairment in early-phase LTP (E-LTP) and an increase in NMDAR-dependent LTD. These changes in synaptic plasticity were attributed to a dramatic reduction in the levels of the active form of Rac1 and Cdc42 in the hippocampus, in accordance with the Rho-GEF function of ARHGEF6. Furthermore, behavioral characterization showed impairments in complex spatial learning, associated with deficits in coping with altered learning tasks or sensory stimuli [35]. These deficits caused by the loss of ARHGEF6 may mimic, at least in part, the human ID phenotype.

PAK3 encodes p21-activated kinase 3, a member of the PAK family of serine/threonine protein kinases which are downstream effectors for Rac/Cdc42 Rho GTPases. Different mutations of PAK3 have been found in non-syndromic X-linked ID patients [36], [37], [38]. PAK3 is prominently expressed in different regions of the brain [39]. Activation of PAK by Rac1 or Cdc42 leads to the phosphorylation of LIM Kinase (LIMK). In turn, activated LIMK phosphorylates and inactivates cofilin, resulting in actin depolymerization [40], [41]. Several studies, using knockdown and overexpression approaches in neuronal cell cultures and transgenic animal models, have elucidated the role of PAK3 in spine morphogenesis and synaptic plasticity in vitro and in vivo. In rat hippocampal organotypic slice cultures, antisense- and RNAi-mediated suppression of PAK3 levels, or expression of a dominant negative PAK3 (dnPAK3) mutant carrying the human (MRX30) mutation, decreased the number of mature spines and promoted thin, immature spines. These changes in dendritic spines were accompanied by impaired AMPAR transmission and LTP [42]. These spine phenotypes were largely recapitulated in transgenic mice expressing dnPAK3 (dnPAK). At the functional level, these mice also displayed enhanced LTP and reduced LTD in the cortex. Furthermore, behavioral tests showed impaired memory consolidation [43]. Mice lacking the PAK3 gene, showed no obvious deficits in neuronal structures. Functionally however, these mice were selectively impaired in late-phase hippocampal LTP (L-LTP) and showed a dramatic decrease in the levels of the active form of cAMP-responsive element-binding protein (CREB) in the hippocampus [44]. The reduction of active CREB, which is required for activity-dependent transcription, may therefore contribute to the L-LTP phenotype. A more recent investigation indicated a dual role of PAK3 in regulating activity-dependent and learning-associated structural plasticity. Dubos et al. found that PAK3 negatively regulated spine growth, as loss of PAK3 function caused uncontrolled and excessive growth of new spines. This suggests that PAK3 may be involved in homeostatic regulation of spine morphology upon synaptic activity. On the other hand, they showed that PAK3 was activated in spines during LTP induction and was specifically recruited to dendritic spines by synaptic activity. PAK3 inhibition prevented stimulated spines to maintain in a stable state. This indicates that PAK3 is involved in regulating activity-mediated rearrangement of synaptic connectivity [45].

Interestingly, in Fmr1 KO mice, a model for Fragile X, inhibition of PAK3 activity by dnPak or PAK inhibitor could reverse the abnormalities in dendritic spines, synaptic plasticity and behavior [46], [47]. This indicates that Rac/Cdc42-PAK pathway may have therapeutic potential for ID. In addition, PAK signaling has also been found to play an important role in Alzheimer disease (AD) and Huntington disease (HD), which has recently been reviewed in detail [48].

As mentioned above, PAK and Rho-kinases both stimulate LIMK. It is therefore interesting that LIMK1 is one of the genes heterozygously deleted in Williams-syndrome, a rare (1 in 25,000) genetic disorder [49]. Patients with this syndrome suffer from a variety of abnormalities, including mild ID. Of particular interest is the fact that Limk1 knockout mice show abnormal spine morphology, abnormal synaptic plasticity, and impaired spatial learning. Together these studies indicate that regulation of the actin cytoskeleton by Rho GTPases through ARHGEF6/Rac-Cdc42/PAK3/LIMK1 pathway plays a crucial role in controlling local spine growth and stability, which subsequently contribute to learning and memory processes.

The MEGAP gene was identified as a gene that is disrupted in patients suffering from 3p syndrome. These patients exhibit microcephaly, growth failure, heart and renal defects, hypotonia, facial abnormalities and ID [50]. Due to the finding of heterozygous MEGAP-truncating mutations in three healthy families, the implication of MEGAP in ID is currently under debate [51].

The MEGAP (mental-disorder-associated GAP) gene product was initially identified as a WAVE-associated protein (WRP) and as a SLIT-ROBO interacting protein (srGAP3) [52], [53]. The MEGAP mRNA transcript is predominantly expressed in fetal and adult brain, and is enriched in the neurons of the hippocampus, cortex and amygdala [50]. The RhoGAP domain of MEGAP strongly enhances intrinsic Rac1 GTPase activity and to a lower extent Cdc42 [50]. Besides the RhoGAP domain, MEGAP contains two other conserved domains, a WAVE1 binding SH3 domain and a lipid membrane binding IF-BAR domain. Recently MEGAP was shown to regulate key aspects of synapse development and function [54]. Using a MEGAP conditional KO mouse they showed that, in hippocampal cultured neurons, only removing MEGAP in early development decreases filopodia and spine formation. This phenotype could be rescued by overexpressing either the IF-BAR truncation or full-length MEGAP. Knockout of Megap at later stages did not result in any spine phenotype. These results were confirmed in vivo, strongly supporting the idea that MEGAP is required during the early (filopodial) stages of spine development and that the IF-BAR domain is important in this function. In addition, loss of Megap led to significant impairments in long-term memory as measured in multiple behavioral tests such as novel object recognition, water maze, and passive avoidance [54]. Interestingly, mice expressing a WRP binding impaired mutant form of WAVE-1, were also deficient in long-term memory [55]. Together these findings suggest that signaling through MEGAP and WAVE-1 to the actin cytoskeleton is important for normal neuronal function.

Although much of the ID research has been directed towards the formation of excitatory synapses, Rho GTPase regulatory proteins also function in the formation inhibitory synapses. ARHGEF9 is the best-characterized Rho-family GEF involved in the formation of inhibitory synapses.

ARHGEF9 is located on the X chromosome and the first mutation, located in exon2, was detected in a patient with severe ID and clinical symptoms of hyperekplexia and epilepsy [56]. Subsequent studies have shown that loss of function mutations in the ARHGEF9 gene cause ID associated epilepsy [57], [58], [59], [60]. ARHGEF9 encodes a Cdc42 GEF protein collybistin (Cb), which is highly expressed throughout the adult brain and is specifically enriched in neuronal dendrites [61], [62]. In rodents, four distinct splice variants (Cb1–Cb3) have been described. Three of these differ in their C-termini (Cb1–3). Cb2 exists with (Cb2SH3+) and without a SH3 domain (Cb2SH3-). In the human brain, only a single isoform (hPEM2, human homolog of Posterior End Mark-2) corresponding to Cb3, was found to be expressed [56], [63], [64]. All Cb isoforms contain a Dbl-homology (DH) domain, which is known to be responsible for the GDP/GTP exchange activity, and a pleckstrin-homology (PH) domain. Cb is required for formation and maintenance of postsynaptic gephyrin scaffolds and the synaptic localization of gephyrin-dependent GABAAR subtypes [56], [65]. The PH domain of Cb was shown to selectively bind with phosphatidylinositol-3-phosphate (PI3P) and overexpressed Cb2SH3- mutant lacking the PH-domain interfered with gephyrin clustering at inhibitory postsynaptic sites [58]. Interestingly, the GEF activity of Cb seems not to be essential for gephyrin clustering [56], [66], [67]. Recently, it was proposed that the interaction of the Cb PH domain and the plasma membrane may be sufficient for establishing and maintaining gephyrin clustering-dependent GABAergic synapses, since any of the Cb isoforms can rescue the impairment in GABAergic neurotransmission induced by Cb knockdown [68]. Studies comparing Cb isoforms with or without SH3 domain further revealed that the SH3 domain exerts an autoinhibitory function, as Cb2SH3- was constitutively active and promoted the clustering of gephyrin while CbSH3+ isoforms did not. Interestingly, SH3 domain containing Cb isoforms were found to be activated by neuroligin 2 (NL2), as the binding of NL2 and GABAARα2 released the autoinhition of Cb by its SH3 domain [69], [70]. Recently it was reported that Arhgef9 KO mice displayed deficits in spatial learning and increased anxiety-like behavior. These mice showed a reduced density of both synaptic gephyrin clusters and GABAAR in hippocampus, amygdala and cerebellum. Consequently, GABAergic transmission in the hippocampus was significantly impaired, as both the frequency and the amplitude of GABAergic miniature inhibitory post synaptic currents (mIPSC) from CA1 pyramidal neurons were dramatically reduced in Arhgef9 KO mice. In addition, in Arhgef9 knockout mice, theta-burst induced LTP was increased. Blocking GABAR by applying picrotoxin eliminated the difference between Arhgef9 KO and control mice, indicating that the reduction of GABAergic inhibition contributed to the changes in LTP [71].

Section snippets

Conclusions

Spine abnormalities are associated with numerous neurodevelopmental, neuropsychiatric, and neurodegenerative disorders [9], [72], [73]. It is therefore not surprising that Rho GTPase signaling, as a critical controller of the actin cytoskeleton pathway, has emerged as a major signaling pathway affected in these disorders. More specifically, Rho GTPase signaling plays an important role as a key signaling integrator regulating both, synaptic structure and function through the control of AMPAR

Conflict of interest

None.

Funding

This work was supported by the MERE-GLU, an EU FP7 Marie Curie Re-integration Grant (PEOPLE-2010-RG, 277091 to N.N.K).

References (73)

  • C. Körber

    Effects of distinct collybistin isoforms on the formation of GABAergic synapses in hippocampal neurons

    Mol. Cell. Neurosci.

    (2012)
  • A. Poulopoulos

    Neuroligin 2 drives postsynaptic assembly at perisomatic inhibitory synapses through gephyrin and collybistin

    Neuron

    (2009)
  • L. Saiepour

    Complex role of collybistin and gephyrin in GABAA receptor clustering

    J. Biol. Chem.

    (2010)
  • J.C. Fiala et al.

    Dendritic spine pathology: cause or consequence of neurological disorders?

    Brain. Res. Rev.

    (2002)
  • H. Van Bokhoven

    Genetic and epigenetic networks in intellectual disabilities

    Annu. Rev. Genet.

    (2011)
  • J. Chelly et al.

    Genetics and pathophysiology of mental retardation

    Eur. J. Hum. Genet.: EJHG

    (2006)
  • Y. Humeau et al.

    X-linked mental retardation: focus on synaptic function and plasticity

    J. Neurochem.

    (2009)
  • A Pavlowsky et al.

    Major synaptic signaling pathways involved in intellectual disability

    Mol. Psychiatry.

    (2012)
  • N. Nadif Kasri et al.

    Rho-linked genes and neurological disorders

    Pflügers Archiv: Eur. J. Physiol.

    (2008)
  • P. Valnegri et al.

    Synaptic dysfunction and intellectual disability

    Adv. Exp. Med. Biol.

    (2012)
  • P. Aspenström et al.

    Rho GTPases have diverse effects on the organization of the actin filament system

    Biochem. J.

    (2004)
  • E.-E. Govek et al.

    The role of Rho GTPase proteins in CNS neuronal migration

    Dev. Neurobiol.

    (2011)
  • S.E. Newey et al.

    Rho GTPases, dendritic structure, and mental retardation

    J. Neurobiol.

    (2005)
  • L. a Cingolani et al.

    Actin in action: the interplay between the actin cytoskeleton and synaptic efficacy

    Nat. Rev. Neurosci.

    (2008)
  • K.-I. Okamoto et al.

    Rapid and persistent modulation of actin dynamics regulates postsynaptic reorganization underlying bidirectional plasticity

    Nat. Neurosci.

    (2004)
  • C.S. Rex

    Different Rho GTPase-dependent signaling pathways initiate sequential steps in the consolidation of long-term potentiation

    J. Cell. Biol.

    (2009)
  • H. Murakoshi et al.

    Local, persistent activation of Rho GTPases during plasticity of single dendritic spines

    Nature

    (2011)
  • H. Okada et al.

    Signaling through actin to regulate spine formation and function

    Open Neurosci. J.

    (2009)
  • E.-E. Govek

    The X-linked mental retardation protein oligophrenin-1 is required for dendritic spine morphogenesis

    Nat. Neurosci.

    (2004)
  • M. Khelfaoui

    Inhibition of RhoA pathway rescues the endocytosis defects in Oligophrenin1 mouse model of mental retardation

    Hum. Mol. Genet.

    (2009)
  • C. Bergmann

    Oligophrenin 1 (OPHN1) gene mutation causes syndromic X-linked mental retardation with epilepsy, rostral ventricular enlargement and cerebellar hypoplasia

    (2003)
  • N. Philip

    Mutations in the oligophrenin-1 gene (OPHN1) cause X linked congenital cerebellar hypoplasia

    J. Med. Genet.

    (2003)
  • V. Des Portes

    Specific clinical and brain MRI features in mentally retarded patients with mutations in the Oligophrenin-1 gene

    Am. J. Med. Genet. A.

    (2004)
  • G. Zanni

    Oligophrenin 1 mutations frequently cause X-linked mental retardation with cerebellar hypoplasia

    Neurology

    (2005)
  • N. Nadif Kasri et al.

    The Rho-linked mental retardation protein oligophrenin-1 controls synapse maturation and plasticity by stabilizing AMPA receptors

    Genes Dev.

    (2009)
  • G. Ahmadian

    Tyrosine phosphorylation of GluR2 is required for insulin-stimulated AMPA receptor endocytosis and LTD

    EMBO J.

    (2004)
  • Cited by (78)

    View all citing articles on Scopus
    View full text