Scholarly article on topic 'PICK1 links AMPA receptor stimulation to Cdc42'

PICK1 links AMPA receptor stimulation to Cdc42 Academic research paper on "Biological sciences"

CC BY
0
0
Share paper
Academic journal
Neuroscience Letters
OECD Field of science
Keywords
{"Rho-family GTPase" / "Actin cytoskeleton" / "PDZ domain" / "BAR domain" / "Glutamate receptor"}

Abstract of research paper on Biological sciences, author of scientific article — Daniel L. Rocca, Jonathan G. Hanley

Abstract Rho-family GTPases control numerous cell biological processes via effects on actin dynamics, such as cell migration, cell adhesion, morphogenesis and vesicle traffic. In neurons, they are involved in dendritic spine morphogenesis and other aspects of neuronal morphology via regulation of the actin cytoskeleton. The Rho-family member Cdc42 regulates dendritic spine morphology via its effector N-WASP, which activates the actin-nucleating Arp2/3 complex. Excitatory synaptic transmission is known to regulate actin dynamics in dendritic spines to bring about changes in spine morphology or motility, however, the details of the signalling pathways that transduce glutamate receptor activation to Rho GTPase function are unclear. PICK1 is a PDZ and BAR domain protein that interacts with the Arp2/3 complex and the GTPase Arf1 to regulate actin polymerisation in dendritic spines. PICK1 also binds AMPA receptor subunits GluA2/3 and is involved in GluA2-dependent AMPAR trafficking. Here, we show that PICK1 binds Rac1 and Cdc42, via distinct but overlapping binding sites. Furthermore, AMPAR stimulation deactivates Cdc42 and alters its detergent solubility in neurons via a PICK1-dependent process. This work suggests a novel role for PICK1 in transducing AMPAR stimulation to Cdc42 function in neurons.

Academic research paper on topic "PICK1 links AMPA receptor stimulation to Cdc42"

W. Neuroscience "f5* Letters . 1

Contents lists available at ScienceDirect

Neuroscience Letters

journal homepage: www.elsevier.com/locate/neulet

Short communication

PICK1 links AMPA receptor stimulation to Cdc42

Daniel L. Rocca, Jonathan G. Hanley *

School of Biochemistry, Medical Sciences Building,University of Bristol, University Walk, Bristol BS8 1TD, UK

ELSEVIER

CrossMark

HIGHLIGHTS

PICK1 binds Racl and Cdc42.

AMPA receptors can interact with Cdc42 via PICK1.

AMPA stimulation increases Cdc42 detergent solubility in a PICKl-dependent manner.

ARTICLE INFO

ABSTRACT

Article history: Received 1 October 2014 Received in revised form 26 November 2014 Accepted 28 November 2014 Available online 2 December 2014

Keywords: Rho-family GTPase Actin cytoskeleton PDZ domain BAR domain Glutamate receptor

Rho-family GTPases control numerous cell biological processes via effects on actin dynamics, such as cell migration, cell adhesion, morphogenesis and vesicle traffic. In neurons, they are involved in dendritic spine morphogenesis and other aspects of neuronal morphology via regulation of the actin cytoskeleton. The Rho-family member Cdc42 regulates dendritic spine morphology via its effector N-WASP, which activates the actin-nucleating Arp2/3 complex. Excitatory synaptic transmission is known to regulate actin dynamics in dendritic spines to bring about changes in spine morphology or motility, however, the details of the signalling pathways that transduce glutamate receptor activation to Rho GTPase function are unclear. PICK1 is a PDZ and BAR domain protein that interacts with the Arp2/3 complex and the GTPase Arf1 to regulate actin polymerisation in dendritic spines. PICK1 also binds AMPAreceptor subunits GluA2/3 and is involved in GluA2-dependent AMPAR trafficking. Here, we show that PICK1 binds Rac1 and Cdc42, via distinct but overlapping binding sites. Furthermore, AMPAR stimulation deactivates Cdc42 and alters its detergent solubility in neurons via a PICK1-dependent process. This work suggests a novel role for PICK1 in transducing AMPAR stimulation to Cdc42 function in neurons.

© 2014 Published by Elsevier Ireland Ltd.

1. Introduction

Rho-family GTPases are proteins of fundamental importance in integrating intracellular signalling pathways. They are molecular switches, cycling between an active GTP-bound state and an inactive GDP-bound state, and once activated bind to a wide range of effectors to initiate a diverse array of signalling pathways that control numerous cell biological processes via effects on actin dynamics, such as cell migration, cell adhesion, morphogenesis and vesicle traffic [10,23]. Rho GTPases are activated mainly through cell-surface receptors via guanine nucleotide exchange factors (GEFs), which promote GTP loading. Conversely, GTPase-activating proteins (GAPs) deactivate Rho GTPases by enhancing

* Corresponding author. Tel.: +44 01173311944. E-mail address: jon.hanley@bristol.ac.uk (J.G. Hanley).

their enzymatic activity and returning the protein to a GDP-bound state [3,25,31].

In neurons, Rho-family GTPases such as RhoA, Rac1 and Cdc42 are involved in dendritic spine morphogenesis and other aspects of neuronal morphology via regulation of the actin cytoskeleton [21,31,32]. For example, Cdc42 regulates dendritic spine morphology via its effector N-WASP, which promotes actin polymerisation via activation of the Arp2/3 complex [14,33]. Glutamatergic synap-tic transmission is known to regulate actin dynamics in dendritic spines [8,19], however, the details of the signalling pathways and molecular mechanisms that transduce glutamate receptor activation to Rho GTPase function are unclear.

Postsynaptic AMPA-type glutamate receptors (AMPARs) mediate most fast excitatory synaptic transmission and are crucial for many aspects of brain function, including learning, memory and cognition [6,17]. AMPARs undergo constitutive and activity-dependent trafficking to, and removal from, synapses. Changes in synaptic AMPAR number, subunit composition and/or channel

http://dx.doi.org/10.1016/j.neulet.2014.11.046 0304-3940/© 2014 Published by Elsevier Ireland Ltd.

properties result in long-term potentiation (LTP) or long-term depression (LTD) of synaptic efficacy [1,13].

P1CK1 is a PDZ and BAR domain protein that interacts with the actin-nucleating Arp2/3 complex and inhibits actin polymerisation [12,24]. P1CK1 binds AMPAR subunits GluA2/3 and is required for GluA2-dependent AMPAR trafficking in hippocampal neurons during synaptic plasticity, and also following pathological insults such as oxygen/glucose deprivation [7,30]. P1CK1 also restricts dendritic spine size via Arp2/3 inhibition, and is involved in spine shrinkage during LTD [20]. ABP/GR1P is a family of multi-PDZ domain scaffold proteins that also interact with AMPAR subunits GluA2/3, and are involved in AMPAR trafficking [5,26].

Here, we show that P1CK1 binds Rac1 and Cdc42, via distinct but overlapping binding sites. Furthermore, AMPAR stimulation deactivates Cdc42 and alters its detergent solubility in neurons via a P1CK1-dependent process.

viruses were prepared as directed by the manufacturers instructions (Invitrogen). Infections were carried out around 20 h before cell lysis and were carried out as described before (Hanley et al., 2002)

2.4. Co-immunoprecipitations

Co-immunoprecipitations were carried out from dissociated cortical cultures as previously described [15]. Briefly cortical neurons were lysed in lysis buffer (0.5% TX-100,150 mM NaCl, 20 mM Tris pH 7.5 plus protease inhibitors), the Triton X-100 content was then diluted to 0.25% using 150 mM NaCl and 20 mM Tris, pH 7.5, before immunoprecipitation with 2 |ig control IgG, anti-PlCK1 or anti-GluA2 antibodies. Bound proteins were detected by western blotting.

2. Materials and methods

2.1. Plasmids

Both pRK5-myc-Rac1 and pRK5-flag-Cdc42 were kind gifts from Prof. Kate Nobes. GST-Rhotekin, GST-PAK-CRIB and pcDNA3.1 myc-RhoA were kind gifts from Prof. Harry Mellor. All constructs were expressed in COS7 cell lines following transfection using Lipo-fectamine 2000 as per manufacturers instructions (Invitrogen). Sindbis virus constructs encoding the short peptides pep2-SVKl, pep2-SVKE and pep2-EVKl and EGFP after an IRES cassette were a kind gift from Prof. Jeremy Henley [29]. GST-R2C and his6-PlCK1 were expressed in Escherichia coli BL21 and have been described previously (Hanley et al., 2002).

2.2. Antibodies

The antibodies used were as follows: anti-myc (9E10, Santa Cruz); anti-flag (M2, Sigma); anti-Cdc42 (clone44/CDC42, BD biosciences); anti-Rac1 (610,650, BD biosciences); anti-PlCK1 (75-040, Neuromab) and anti-ß-tubulin (clone TUB2.1, Sigma).

2.5. GSTpulldowns

These were carried out as previously described [24]. GST-P1CK1, GST-PAK-CR1B or GST-Rhotekin were incubated with lysates prepared from COS7 cells expressing epitope-tagged GTPases or with purified his6-tagged proteins. Bound proteins were detected by western blotting.

2.6. Cdc42 activation assays

Cortical neurons were stimulated with 100 |im AMPA for 5 min before lysis in 0.5% TX-100, 150 mM NaCl, 10 mM HEPES pH 7.4 and protease/phosphatase inhibitors. Subsequently, GTP-bound Cdc42 was isolated from lysates via a one-step batch purification using GST-PAK-CR1B pulldowns followed by western blotting. Total Cdc42 was also determined using a fraction of the neuronal extracts before pulldowns.

2.7. Western blot analysis and quantification

2.3. Primary neuronal culture and Sindbis virus transduction

Primary cortical neuronal cultures were prepared from E18 wis-tar rats as previously described (Hanley and Henley, 2005). Sindbis

Western blots from five independent experiments were scanned and analysed by densitometry using ImageJ. error bars represent s.e.m., and two-tailed t-tests were carried out to determine significant differences between two conditions.

Fig.1. PICK1 interacts with Cdc42 and Rac1 but not RhoA.

(A) P1CK1 binds Rac1 and Cdc42. GST-pulldowns were carried out from lysates prepared from COS7 cells expressing myc-taggedRac1(V12), Rac1(N17), Cdc42(V12)orCdc42(N17) using GST-P1CK1, GST-PAK-CR1B or GST alone. Bound proteins were detected by western blotting using anti-myc.

(B) P1CK1 does not bind RhoA. GST-pulldowns were carried out from lysates prepared from COS7 cells expressing myc-tagged RhoA(V14) or RhoA(N17) using GST-P1CK1, GST-Rhotekin or GST alone. Bound proteins were detected by western blotting using anti-myc.

(C) Cdc42 and Rac1 interact with P1CK1 in neurons. Lysates prepared from cultured cortical neurons were immunoprecipitated with anti-P1CK1 antibody or non-immune 1gG as a control, and bound proteins were detected by western blotting using specific antibodies as shown.

3. Results and discussion

To investigate the interaction of P1CK1 with Rho-family GTPases, we carried out pulldown assays using GST-P1CK1 and lysates prepared from COS cells expressing epitope-tagged Cdc42, Rac1 or RhoA. Since an important functional feature of Rho-family GTPases is that they bind downstream effector proteins preferentially in their active, GTP-bound state [4], we tested constitutively active (CA, V12) and dominant negative (DN, N17) mutant GTPases. p21 activated kinase (PAK) is a known effector for Cdc42 and Rac and binds CA but not DN mutants of both GTPases ([35] and Fig. 1A). GST-P1CK1 binds CA and DN mutants equally well for both Rac1 and Cdc42 (Fig. 1A), suggesting that P1CK1 is not a Rac1/Cdc42 effector, but perhaps plays a scaffolding role to localise the GTPases to specific subcellular locations. We carried out equivalent experiments for RhoA, using the known effector protein Rhotekin as a positive control [22]. GST-P1CK1 does not interact with either RhoA mutant, demonstrating specificity for the interaction with Rac1 and Cdc42 (Fig. 1B). To confirm that Cdc42 and Rac1 interact with P1CK1 in neurons, we carried out co-immunoprecipitations (co-1Ps) from lysates prepared from cultured cortical neurons using anti-P1CK1 antibodies. Both Rac1 and Cdc42 show a robust interaction with P1CK1 (Fig. 1C), demonstrating that both GTPases interact with P1CK1 in neurons.

To further compare, the P1CK1-Cdc42 interaction with that of P1CK1-Rac1, we analysed the binding of purified his6-tagged flagCdc42 and mycRac1 to a range of P1CK1 truncations. Wild-type GST-P1CK1 binds both GTPases, demonstrating that the interactions are direct, with no requirement for intermediary protein components. 1nterestingly, the two GTPases show distinct patterns of binding to the P1CK1 mutants, indicating that Cdc42 and Rac1 have overlapping, but not identical binding sites on P1CK1 (Fig. 2). Both GTPases require the presence of the BAR domain, indeed

Cdc42 binds the isolated BAR domain and binding is unaffected by deletion of either acidic region (ACT, ANT) or deletion of an extreme C-terminal region (1-379) of the full-length protein. However, Cdc42 binding is abolished in the absence of the PDZ domain when the C-terminal region is present (105-416), suggesting an intramolecular inhibition of the interaction. 1t has previously been suggested that P1CK1 forms an intramolecular interaction between the PDZ and BAR domains [18,24], and also that the C-terminal tail interacts with the BAR domain [16]. 1n contrast, Rac1 does not bind the isolated BAR domain, but requires the presence of both BAR and C-terminal regions for the interaction (Fig. 2).

These results demonstrate that P1CK1 directly interacts with Cdc42 and Rac1 via the BAR domain, with additional sequence determinants that suggest overlapping but distinct binding sites on P1CK1. Rac1 has previously been shown to interact with Arfaptin BAR domain, which shows some homology to that of P1CK1 [27,28]. The structure of the Rac-Arfaptin complex has been defined, and indicates that the GTPase sits on the concave face of the crescent-shaped BAR domain [28]. 1f a similar conformation exists for P1CK1, this would suggest that GTPase binding and curved membrane binding to the BAR domain would be mutually exclusive. P1CK1-bound GTPase would therefore, likely be cytosolic unless associated with a transmembrane protein.

Since P1CK1 is a well-established AMPAR accessory protein [12], we explored an association between AMPAR stimulation and Cdc42. 1nitially, we investigated whether P1CK1 can form a triple complex with Cdc42 or Rac1 and GluA2 C-terminus. GST-GluA2 C-terminus (GluA2C) does not bind Cdc42 or Rac1 in the absence of P1CK1, but when his6-P1CK1 is added, a robust interaction with both GTPases is observed (Fig. 3A). Furthermore, both Cdc42 and P1CK1 are present in GluA2 immunoprecipitations from neuronal lysate, strongly suggesting the presence of a GluA2-P1CK1-Cdc42 tripartite complex in vivo (Fig. 3B). These experiments demon-

Fig. 2. Cdc42 and Rac1 have distinct but overlapping binding sites on P1CK1.

Upper panel: GST pulldowns were carried out using purified his6flagCdc42 or his6mycRac1 and truncation mutants of P1CK1 as GST fusions as depicted. Bound proteins were detected by western blotting using anti-myc or anti-flag.

Lower panel: diagram showing truncation mutants of P1CK1 used, and a summary of the results. A tick indicates a positive interaction, whereas across indicates no binding.

strate that Cdc42 can associate with AMPARs via PICK1, and suggest that either Cdc42 regulates AMPAR trafficking, or AMPARs regulate Cdc42 function via PICK1. To test the latter hypothesis, we used GST-PAK pulldown assays to determine the effect of AMPAR stimulation on Cdc42 activation in cultured neurons. Bath application of AMPA for 5 min causes a significant reduction in GTP-bound Cdc42 (Fig. 3C). In addition, we noted an increase in the detergent solubility of Cdc42 after AMPAR stimulation (Fig. 3C), suggesting that AMPAR stimulation displaces Cdc42 from specific membrane compartments or protein complexes. Since cell lysis and western analysis were carried out after just 5 min of drug treatment, it is highly unlikely that this difference in Cdc42 immunoreactiv-ity could be explained by an increase in protein translation or a reduction in protein degradation.

To investigate whether PICK1 could mediate these effects of AMPA on Cdc42, we used Sindbis virus to express peptides mimicking the C-terminus of GluA2 that block AMPAR-PDZ domain interactions, and hence, disrupt the link between Cdc42 and AMPARs. Pep2-SVKI represents the wild-type sequence and disrupts PDZ interactions with PICK1 and ABP/GRIP, whereas pep2-EVKI is selective for PICK1. Pep2-SVKE does not bind PDZ domains, and hence, serves as a negative control [29]. Interestingly, the presence of pep2-EVKI increases the detergent-solubility of Cdc42 under basal conditions, which occludes the effect of subsequent AMPAR stimulation. However, pep2-EVKI has no effect on AMPA-induced Cdc42 deactivation, which is similar to pep2-SVKE expressing neurons (Fig. 3D). In contrast, pep2-SVKI has no effect

Fig. 3. PICK1 links AMPAR stimulation to Cdc42 deactivation.

(A) Both Cdc42 and Rac1 form a triple complex with PICK1 and GluA2 in vitro. GST-pulldowns were carried out from lysates prepared from COS7 cells expressing flag-tagged Cdc42(V12) or myc-tagged Rac1(V12) using GST-GluA2 C-terminus (GluA2C) in the absence or presence of purified his6PICK1, or GST alone. Bound proteins were detected by western blotting using anti-PICK1, anti-flag or anti-myc.

(B) Cdc42 forms a triple complex with PICK1 and GluA2 in neurons. Lysates were prepared from dissociated cortical neurons, and immunoprecipitations carried out using anti-GluA2 or non-immune mouse IgG as control. Bound proteins were detected by western blotting using anti-GluA2, anti-PICK1 and anti-Cdc42.

(C) AMPAR stimulation increases the detergent solubility of Cdc42 and reduces the proportion of GTP-bound Cdc42. Dissociated cortical neurons were treated with 100 |xM AMPA or vehicle for 5 min. Lysates were prepared and GTP-bound Cdc42 was isolated by GST pulldown using GST-PAK. GST-PAK bound Cdc42-GTP and unbound Cdc42 in the lysate were detected by western blotting using anti-Cdc42. Tubulin serves as a loading control. Representative western blots are shown, and graphs show pooled data for total detergent-soluble Cdc42 (left graph) and for the proportion of Cdc42 that is GTP-bound (right graph). n = 5.

(D) PICK1 PDZ domain interactions are involved in AMPAR-induced changes in detergent solubility of Cdc42. Dissociated cortical neurons were transduced with Sindbis virus expressing pep2-SVKE-IRES-EGFP, pep2-SVKI-IRES-EGFP or pep2-EVKI-IRES-EGFP. Cultures were treated with AMPA(+) orvehicle(-), and processed for biochemistry as in (B). A representative western blot is shown, and graphs show pooled data for total detergent-soluble Cdc42 (left graph) and for the proportion of Cdc42 that is GTP-bound (right graph). n = 5.

on detergent solubility, but causes a decrease in GTP-bound Cdc42, which occludes the effect of AMPA application.

These results suggest that P1CK1 is involved in the change in subcellular localisation of Cdc42 that occurs as a result of AMPAR stimulation. Since disrupting GluA2-P1CK1 binding causes the same change in detergent solubility as stimulating AMPARs, these results are consistent with a model whereby a Cdc42-P1CK1-AMPAR complex is associated with TX-100 resistant membrane compartments or protein complexes under basal conditions, and dissociates upon AMPAR stimulation. Following dissociation of P1CK1 from membrane-bound AMPARs, Cdc42 would have a more cytosolic distribution, and would consequently be more detergent-soluble. Our results also suggest a role for ABP/GR1P in AMPAR-dependent changes in Cdc42 activation. Since pep2-SVK1 but not pep2-EVK1 occlude the effect of AMPAR stimulation, dissociation of ABP/GR1P from AMPARs may be involved in AMPA-induced Cdc42 deactivation. A potential explanation for this result is that P1CK1 binds ABP/GR1P [18] and can therefore, associate with AMPARs independently of the P1CK1 PDZ domain, but via the ABP/GR1P PDZ domain. 1t is therefore, possible that a GluA2-GR1P-P1CK1-Cdc42 complex is involved in regulating Cdc42 activity. Although there are no reports of Cdc42, GAPs or GEFs that associate with ABP/GR1P, GRASP-1 is a Ras GEF that binds directly to GR1P1, indicating that such a mechanism is feasible [34]. Although the mechanistic details are likely to be different, this model may have some features in common with the functional effect of RhoGD1s on Rho-family GTPases. RhoGD1s associate with cytosolic GTPases, blocking their association with membranes, and maintaining the GTPase in an inactive, GDP-bound state [9].

P1CK1 inhibits actin polymerisation by direct binding to the Arp2/3 complex. Activated (GTP-bound) Arf1 attenuates Arp2/3-P1CK1 binding, hence, P1CK1 is a downstream effector of the small GTPase in this pathway. 1n the current study, we show that Cdc42 binds P1CK1 in a GTP-independent manner and that P1CK1 is upstream of Cdc42. Hence, P1CK1 can function in multiple ways to regulate actin dynamics in neurons, and is one of a growing number of BAR domain containing proteins that have critical roles in controlling the actin cytoskeleton in multiple cell types [2]. 1t will be of great interest to determine the precise function of Rac1/Cdc42 binding to P1CK1 in the control of actin-dependent processes in neuronal function.

1n conclusion, our results suggest that AMPAR activation regulates Cdc42 subcellular localisation and function via P1CK1. This could provide a mechanism for the regulation of local actin polymerisation in dendritic spines to regulate spine dynamics or morphology. Consistent with this hypothesis, it has been shown that AMPAR stimulation blocks spine motility and causes alterations in actin polymerisation leading to morphological changes in spines that are believed to correspond to spine stabilization and maturation [8,11].

Acknowledgement

This work was funded by the Medical Research Council. References

[1] V. Anggono, R.L. Huganir, Regulation of AMPA receptor trafficking and synaptic plasticity, Curr. Opin. Neurobiol. 22 (2012) 461-469.

[2] P. Aspenstrom, BAR domain proteins regulate Rho GTPase signaling, Small GTPases 5 (2014).

[3] A. Bernards, J. Settleman, GAP control: regulating the regulators of small GTPases, Trends Cell Biol. 14 (2004) 377-385.

[4] A.L. Bishop, A. Hall, Rho GTPases and their effector proteins, Biochem. J. 348 (Pt 2) (2000)

241-255.

S.P. Braithwaite, H. Xia, R.C. Malenka, Differential roles for NSF and GR1P/ABP in AMPA receptor cycling, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) 7096-7101. R. Dingledine, K. Borges, D. Bowie, S.F. Traynelis, The glutamate receptor ion channels, Pharmacol. Rev. 51 (1999) 7-61.

R.M. Dixon, J.R. Mellor,J.G. Hanley, P1CK1-mediated glutamate receptor subunit 2 (GluR2) trafficking contributes to cell death in oxygen/glucose-deprived hippocampal neurons, J. Biol. Chem. 284 (2009) 14230-14235.

M. Fischer, S. Kaech, U. Wagner, H. Brinkhaus, A. Matus, Glutamate receptors regulate actin-based plasticity in dendritic spines, Nat. Neurosci. 3 (2000) 887-894.

R. Garcia-Mata, E. Boulter, K. Burridge, The 'invisible hand': regulation of RHO GTPases by RHOGD1s, Nat. Rev. Mol. Cell Biol. 12 (2011) 493-504. A. Hall, Rho family GTPases, Biochem. Soc. Trans. 40 (2012) 1378-1382. S. Halpain, A. Hipolito, L. Saffer, Regulation of F-actin stability in dendritic spines by glutamate receptors and calcineurin, J. Neurosci. 18 (1998) 9835-9844.

J.G. Hanley, P1CK1: a multi-talented modulator of AMPA receptor trafficking, Pharmacol. Ther. 118(2008) 152-160.

J.M. Henley, E.A. Barker, O.O. Glebov, Routes, destinations and delays: recent advances in AMPA receptor trafficking, Trends Neurosci. 34 (2011) 258-268.

F. 1rie, Y. Yamaguchi, EphB receptors regulate dendritic spine development via intersectin, Cdc42 and N-WASP, Nat. Neurosci. 5 (2002) 1117-1118.

N. Jaafari, J.M. Henley, J.G. Hanley, P1CK1 mediates transient synaptic expression ofGluA2-lacking AMPA receptors during glycine-induced AMPA receptor trafficking, J. Neurosci. 32 (2012) 11618-11630. W. Jin, W.P. Ge, J. Xu, M. Cao, L. Peng, W. Yung, D. Liao, S. Duan, M. Zhang, J. Xia, Lipid binding regulates synaptic targeting of P1CK1, AMPA receptor trafficking, and synaptic plasticity, J. Neurosci. 26 (2006) 2380-2390. H.W. Kessels, R. Malinow, Synaptic AMPA receptor plasticity and behavior, Neuron 61 (2009) 340-350.

W. Lu, E.B. Ziff, P1CK1 interacts with ABP/GR1P to regulate AMPA receptor trafficking, Neuron 47 (2005) 407-421.

R.A. McKinney, Excitatory amino acid involvement in dendritic spine formation, maintenance and remodelling, J. Physiol. 588 (2010) 107-116. Y. Nakamura, C.L. Wood, A.P. Patton, N. Jaafari,J.M. Henley, J.R. Mellor,J.G. Hanley, P1CK1 inhibition of the Arp2/3 complex controls dendritic spine size and synaptic plasticity, EMBO J. 30 (2011) 719-730. M. Negishi, H. Katoh, Rho family GTPases and dendrite plasticity, the neuroscientist: a review journal bringing neurobiology, Neurol. Psych. 11 (2005)187-191.

T. Reid, T. Furuyashiki, T. 1shizaki, G. Watanabe, N. Watanabe, K. Fujisawa, N.

Morii, P. Madaule, S. Narumiya, Rhotekin, a new putative target for Rho

bearing homology to a serine/threonine kinase, PKN, and rhophilin in the

Rho-binding domain, J. Biol. Chem. 271 (1996) 13556-13560.

A.J. Ridley, Rho GTPases and actin dynamics in membrane protrusions and

vesicle trafficking, Trends Cell Biol. 16 (2006) 522-529.

D.L. Rocca, S. Martin, E.L. Jenkins, J.G. Hanley, 1nhibition of Arp2/3-mediated

actin polymerization by P1CK1 regulates neuronal morphology and AMPA

receptor endocytosis, Nat. Cell Biol. 10 (2008) 259-271.

K.L. Rossman, C.J. Der, J. Sondek, GEF means go: turning on RHO GTPases with

guanine nucleotide-exchange factors, Nat. Rev. Mol. Cell Biol. 6 (2005)

167-180.

K. Takamiya, L. Mao, R.L. Huganir, D.J. Linden, The glutamate receptor-interacting protein family of GluR2-binding proteins is required for long-term synaptic depression expression in cerebellar Purkinje cells, J. Neurosci. 28 (2008) 5752-5755.

R. Takeya, K. Takeshige, H. Sumimoto, 1nteraction of the PDZ domain of human P1CK1 with class 1 ADP-ribosylation factors, Biochem. Biophys. Res. Commun. 267 (2000) 149-155.

C. Tarricone, B. Xiao, N. Justin, P.A. Walker, K. Rittinger, S.J. Gamblin, S.J. Smerdon, The structural basis of arfaptin-mediated cross-talk between Rac and Arf signalling pathways, Nature 411 (2001) 215-219. A. Terashima, L. Cotton, K.K. Dev, G. Meyer, S. Zaman, F. Duprat, J.M. Henley,

G.L. Collingridge, J.T. 1saac, Regulation of synaptic strength and AMPA receptor subunit composition by P1CK1, J. Neurosci. 24 (2004) 5381-5390.

A. Terashima, K.A. Pelkey, J.C. Rah, Y.H. Suh, K.W. Roche, G.L. Collingridge, C.J. McBain, J.T. 1saac, An essential role for P1CK1 in NMDA receptor-dependent bidirectional synaptic plasticity, Neuron 57 (2008) 872-882. K.F. Tolias, J.G. Duman, K. Um, Control ofsynapse development and plasticity by Rho GTPase regulatory proteins, Prog. Neurobiol. 94 (2011) 133-148. M. Watabe-Uchida, E.E. Govek, L. Van Aelst, Regulators of Rho GTPases in neuronal development, J. Neurosci. 26 (2006) 10633-10635.

A.M. Wegner, C.A. Nebhan, L. Hu, D. Majumdar, K.M. Meier, A.M. Weaver, D.J. Webb, N-wasp and the arp2/3 complex are critical regulators of actin in the development of dendritic spines and synapses, J. Biol. Chem. 283 (2008) 15912-15920.

B. Ye, D. Liao, X. Zhang, P. Zhang, H. Dong, R.L. Huganir, GRASP-1: a neuronal RasGEF associated with the AMPA receptor/GR1P complex, Neuron 26 (2000) 603-617.

Z.S. Zhao, E. Manser, PAK and other Rho-associated kinases - effectors with surprisingly diverse mechanisms of regulation, Biochem. J. 386 (2005) 201-214.