R

P

AD

ARRA

KpPRK

I

spoP2apa

PtIalPpHi

N1mfP

0h

European Journal of Cell Biology 92 (2013) 129– 138

Contents lists available at SciVerse ScienceDirect

European Journal of Cell Biology

jou rn al h om epage: www.elsev ier .com/ locate /e jcb

eview

21-activated kinase 4 – Not just one of the PAK

nna E. Dart, Claire M. Wells ∗

ivision of Cancer Studies, New Hunts House, Guy’s Campus, King’s College London, London SE1 1UL, UK

a r t i c l e i n f o

rticle history:eceived 4 February 2013eceived in revised form 24 March 2013ccepted 25 March 2013

eywords:

a b s t r a c t

P21-activated kinase 4 (PAK4) is a member of the p21-activated kinase (PAK) family. Historically muchof the attention has been directed towards founding family member PAK1 but the focus is now shif-ting towards PAK4. It is a pluripotent serine/threonine kinase traditionally recognised as a downstreameffector of the Rho-family GTPases. However, emerging research over the last few years has revealedthat this kinase is much more than that. New findings have shed light on the molecular mechanism of

21-activated kinaseAKho family GTPasesinases

PAK4 activation and how this kinase is critical for early development. Moreover, the number of PAK4substrates and binding partners is rapidly expanding highlighting the increasing amount of cellular func-tions controlled by PAK4. We propose that PAK4 should be considered a signalling integrator regulatingnumerous fundamental cellular processes, including actin cytoskeletal dynamics, cell morphology andmotility, cell survival, embryonic development, immune defence and oncogenic transformation. Thisreview will outline our current understanding of PAK4 biology.

ntroduction

PAK4 is a member of the p21-activated kinase (PAK) family oferine/threonine kinases. This family contains six mammalian PAKroteins, PAK1-6 that have been subdivided into two groups basedn domain structure, sequence homology and regulation: group IAKs (PAK1-3) and group II PAKs (4–6) (Arias-Romero and Chernoff,008). PAK4 was the first Group II PAK to be cloned and char-cterised. It was originally identified as a cytoskeletal regulatoryrotein, specifically controlling filopodia formation downstream ofctivated Cdc42 (Abo et al., 1998).

Similarly to other PAK isoforms, PAK4 contains an N-terminalBD (p21-GTPase-binding domain) and a highly conserved C-erminal catalytic serine/threonine kinase domain. Unlike its Group

counterparts, PAK4 has much less sequence N-terminal to the PBDnd distinct from other Group II PAKs does not harbour a nuclearocalisation signal in this region (Fig. 1A). The central region ofAK4 also differs significantly from all other PAKs, containing three

utative proline-rich SH3 (Src homology 3) domain binding sites.owever as of yet, no binding partners have been identified that

nteract specifically with these proline-rich regions in the PAK4

Abbreviations: MRLCs, myosin regulatory light chains; MLC, myosin light chain;K, natural killer; NES, nuclear export signal; NLS, nuclear localisation signal; CRM-, chromosome region maintenance-1; UV, ultraviolet; ECM, extracellular matrix;iRNA, microRNA; SH3, Src homology domain; GEF, guanine nucleotide exchange

actor; IRES, internal ribosome entry site; EGFR, epidermal growth factor receptor;I3K, phosphatidylinositol 3-OH kinase; ATM, ataxia telangiectasia mutated.∗ Corresponding author. Tel.: +44 02078488769.

E-mail address: claire.wells@kcl.ac.uk (C.M. Wells).

171-9335/$ – see front matter © 2013 Elsevier GmbH. All rights reserved.ttp://dx.doi.org/10.1016/j.ejcb.2013.03.002

© 2013 Elsevier GmbH. All rights reserved.

sequence. There is also the presence of a RhoA GEF binding siteupstream of the PAK4 kinase domain which binds the substratesGEF-H1 (Callow et al., 2005) and PDZ-RhoGEF (Barac et al., 2004), aswell as the interacting partner Gab1 (Paliouras et al., 2009) (Fig. 1B).Through sequence homology it has been established that all thePAK proteins possess a potential integrin binding site within thekinase domain. However, only a direct interaction between PAK4and an integrin cytoplasmic tail has been demonstrated (Zhanget al., 2002) (Fig. 1). Interestingly, despite differences in sequencehomology between the kinase domains, and considerable sequencevariation outside the kinase domain PAK1 and PAK4 share a num-ber of substrates (Table 1) and PAK4 specific substrates have notbeen formally identified.

PAK4 is highly expressed throughout development (Qu et al.,2003) and is ubiquitously expressed at low levels in many adulttissues (Callow et al., 2002). Indeed, Pak4 gene deletion results inembryonic lethality (Qu et al., 2003). Human PAK4 has orthologuesin all metazoans but is not present in plants, fungi and protozoa.Interestingly, a splice variant of PAK4 lacking 154 residues (exon4) of the full length PAK4 (591 residues) has also been identifiedin human HeLa and U20S cells but is absent from mouse B16 cells(Baskaran et al., 2012). At this time, the physiological significanceof this shorter variant remains undetermined (Fig. 1B).

Regulation of PAK4 kinase activity

PAKs are a highly conserved group of effector proteins that arerecognised as crucial regulators of Rac and Cdc42 function. Yet, thepart played by the Rho GTPases in activating Group II PAKs hasbeen a subject of much debate. The interaction between PAKs and

130 A.E. Dart, C.M. Wells / European Journal of Cell Biology 92 (2013) 129– 138

Fig. 1. Domain structure of PAK family proteins. (A) All members of the p21-activated kinase (PAK) family share a common domain architecture consisting of an N-terminalp21-GTPase binding domain (GBD) and a C-terminal serine/threonine kinase domain. Within the Group I PAKs, the GBD contains a Cdc42/Rac interactive binding region(CRIB) which overlaps with an autoinhibitory domain (AID). The Group II PAKs contain a GBD that binds Rho GTPases and a putative sequence-related AID (likely to bepresent in PAK5 and 6 by virtue of conserved sequences with PAK4). Variable numbers of core PxxP motifs, ligands for SH3 domain-containing proteins, are enclosed withinthe sequences of all the PAKs. The N-termini of Group I PAKs bind directly to the SH3 domains of Nck1/2 (indicated in purple) and Grb2 whilst a proline-rich motif in thecentral portion of the Group I PAKs binds PIX (indicated in pink). These consensus binding sites are not present in the Group II PAKs. A �5 integrin binding domain hasbeen identified in the PAK4 sequence (indicated in red) and this region is highly homologous amongst PAK family members (indicated in brown). (B) The mammalian PAK4protein comprises of an N-terminal GBD and a C-terminal kinase domain. The central region of PAK4 includes three SH3 binding sites and a GEF-H1- and Gab1-interactingdomain (GID) adjacent to the kinase domain. Situated within the kinase domain resides a �5 integrin binding region. PAK4 autophosphorylation occurs at the activation loopp egulas

tkPT

hospho residue Ser 474 and is independent of its activation now believed to be rplice variant lacks exon 4.

he Rho GTPases occurs via the p21-binding domain (alternativelynown as the GTPase binding domain (GBD)) and specifically forAK4, it has been identified to bind Cdc42 and Cdc42-like GTPases,C10, TCL and Chp (Abo et al., 1998; Aspenstrom et al., 2004). To

ted by a pseudosubstrate sequence (R49PKPLV) within the N-terminal. The PAK4b

date, no functional significance of the interaction of PAK4 with theCdc42-like GTPases has been revealed. In addition, whilst it hasbeen shown that PAK4 can bind Rac, it is to a much lesser extentthan that of Cdc42 and on the basis of this weak interaction, is not

A.E. Dart, C.M. Wells / European Journal of Cell Biology 92 (2013) 129– 138 131

Table 1PAK4 kinase substrates.

Protein Site PAK substrate specificity Reference

LIM kinase T508 PAK1 and PAK4 Dan et al. (2001), Edwards et al. (1999)Slingshot phosphatase (SSH-1L) Not mapped Only PAK4 tested Soosairajah et al. (2005)GEF-H1 S810 PAK1 (S885), PAK2 (S885) and PAK4 Callow et al. (2005), Kosoff et al. (2013), Zenke

et al. (2004)Integrin �5 S759 and S762 PAK4 onlyb Li et al. (2010b)BAD S112 PAK1(S111), PAK2, PAK4 and PAK5 Cotteret et al. (2003), Gnesutta et al. (2001),

Jakobi et al. (2001), Jin et al. (2005), Ye et al.(2011)

Myosin light chain 9 S19 Only PAK4 tested Bright and Frankel (2011)Ran S135a Only PAK4 tested Bompard et al. (2010)Paxillin S272 PAK1c and PAK4 Nayal et al. (2006), Wells et al. (2010)Raf-1 S338 PAK1(S339), PAK2, PAK3 and PAK4 Beeser et al. (2005), Cammarano et al. (2005),

Chaudhary et al. (2000), Edin and Juliano(2005), Jin et al. (2005), King et al. (1998), Zanget al. (2002)

�-Catenin S675 PAK1 (S663, S675) and PAK4 He et al. (2008), Li et al. (2012)p120-Catenin S288 PAK5 and PAK4 Wong et al. (2010)PDZ-RhoGEF Not mapped Only PAK4 testedd Barac et al. (2004)

a This phosphorylation was observed with in vitro kinase assays using the kinase domain of Xenopus X-PAK4, the orthologue of human PAK4.b Unpublished results show that human PAK1 also interacts with various � subunits (Zhang et al., 2002).c PAK1 phosphorylation of paxillin at serine 273 is not fully validated see Dong et al. (2009).

lIGaraeGcfttd

ratpsaPg(ahd2(owmflbp(ltbRc

d However, PDZ-RhoGEF does not bind PAK1 or PAK2.

ikely to mechanistically regulate PAK4 (Abo et al., 1998). The Group PAKs comprise an autoinhibitory domain (AID) overlapping theBD and this region confers upon PAKs 1–3 an activation mech-nism whereby binding to active Rho GTPases via the GBD:AIDegion induces a conformational change relieving autoinhibitionnd allowing autophosphorylation (Eswaran et al., 2008). How-ver, although the Group II PAKs are able to bind Rho-family smallTPases, this interaction has never been demonstrated to signifi-antly increase the activity of PAK4 (Abo et al., 1998). Furthermore,or the Group II PAKs the kinase domain alone is more active thanhe full-length protein (Abo et al., 1998). Thus, studies so far suggesthat the regulation of Group I and Group II PAKs is mechanisticallyifferent.

Examination of the crystal structure of the PAK4 kinase domainevealed that the glycine-rich loop has conformational plasticityllowing it to function as a molecular sensor for ATP binding,hereby facilitating structural rearrangements to retain a com-etent active state (Eswaran et al., 2007). Such plasticity mightuggest that PAK4 would periodically exist in an unphosphorylatedctivation loop form. This could lead to dimerisation as reported forAK1 (Wang et al., 2011). However, this is unlikely given that recentel filtration analysis has shown that PAK4 behaves as a monomerBaskaran et al., 2012). Following these initial findings, two mech-nisms for a switch-like on/off regulation of PAK4 kinase activityave been proposed. The first of which suggests that Cdc42 canirectly release the autoinhibited state of PAK4 (Baskaran et al.,012). This study has identified a putative autoinhibitory domainAID) in the N-terminal of PAK4 (Baskaran et al., 2012), previ-usly believed not to be present in the sequences of Group II PAKsith the possible exception of PAK5 (Ching et al., 2003). In thisodel PAK4 is held in an inactive conformation by this AID formed

rom amino acids 20 to 68 until binding of GTP-Cdc42. This regu-atory mechanism would therefore suggest that PAK4 could onlye activated locally in the presence of the membrane-bound Cdc42roviding both spatial and temporal control to the kinase activityBaskaran et al., 2012). It would be an attractive possibility to specu-ate that the other Rho GTPases capable of binding PAK4 are unable

o induce these similar conformational changes and as such Cdc42inding would offer specificity to PAK4 activation. Indeed, otherho GTPases could afford PAK4 a means to be targeted to differentellular compartments.

The second regulatory mechanism proposes that PAK4 is autoin-hibited by a conserved pseudosubstrate sequence within theN-terminal region motif focussed around a critical proline residue(R49PKPLV) that binds to the kinase domain (Ha et al., 2012).Mutation of the pseudosubstrate relieves the constitutively autoin-hibited full-length PAK4 and results in elevated phosphorylation ofPAK4 substrates, such as BAD. Such autoinhibition might explainthe observation that truncated versions of PAK4 lacking N-terminalregions have increased catalytic activity (Abo et al., 1998; Wellset al., 2002). Furthermore, the autoinhibitory region is proline richand the authors reported that PAK4 can be activated through bind-ing to SH3 domain-containing proteins, such as Src, assuming thatthis releases the autoinhibition (Ha et al., 2012). Interestingly,both these studies found PAK4 to be constitutively phosphory-lated on Ser 474 in the activation loop, again in contrast to Group IPAKs. Taken together these studies support a hypothesis wherebyPAK4 activity is regulated by N-terminal-mediated conformationalchanges and not by activation loop phosphorylation. Whilst regu-lation of PAK4 activity is clearly complex, it is apparent that the RhoGTPases can function in localising the kinase to specific subcellularlocations. Initial studies showed that an interaction between Cdc42and PAK4 caused PAK4 to translocate to the Golgi apparatus (Aboet al., 1998). Since then Cdc42 has also been shown to spatiallyrestrict PAK4 to cell-cell adherens junctions in human bronchialepithelial cells (Wallace et al., 2010). As PAK4 activation must becoordinated with its transport to different cellular compartments, itis an appealing premise to imagine that Rho GTPases localise PAK4to a distinct area of the cell where an activation signal (presum-ably an SH3 domain-containing protein) is concentrated leading tokinase activation.

PAK4 and cytoskeletal regulation

The most well characterised function of PAK4 is in the reg-ulation of actin cytoskeletal organisation mediating changes incell morphology, adhesion and motility (Fig. 2). The first evidencethat PAK4 promoted cytosketetal changes was observed when

expression of PAK4 stimulated filopodia formation downstream ofactivated Cdc42 (Abo et al., 1998). Subsequently it was also demon-strated that overexpression of PAK4 lead to the dissolution of stressfibres and cell rounding which is accompanied with a loss of focal

132 A.E. Dart, C.M. Wells / European Journal of Cell Biology 92 (2013) 129– 138

F g cascc osurvt

ae

taPfSWvibteopcpe

eakeee2ttPb2pkr2fnmfira

ig. 2. PAK4 signalling pathways. Schematic depicting some of the varied signallinellular processes ranging from actin remodelling, apical junction formation and prhe text.

dhesions, in part the consequence of inhibiting GEF-H1 (Baract al., 2004; Callow et al., 2005; Qu et al., 2001; Wells et al., 2002).

PAK4 also has the capacity to regulate actin depolymerisationhrough phosphorylating LIMK1 which in turn phosphorylates thectin severing protein, cofilin (Ahmed et al., 2008; Dan et al., 2001).AK4 orchestrates a dual control of actin dynamics through theormation of a multiprotein complex consisting of LIMK1, Sling-hot phosphatase (SSH-1L) and 14-3-3� (Soosairajah et al., 2005).ithin this signalling pathway, PAK4 activates LIMK1 and inacti-

ates SSH1 resulting in actin filament turnover through an increasen ADF/cofilin activity (Soosairajah et al., 2005). It has recentlyeen shown that protein kinase D (PKD) phosphorylates PAK4 athe autophosphorylation site (Ser 474). The resultant downstreamffect of this phosphorylation event is an increase in the activityf LIMK1. A kinase dead PAK4 mutant blocks PKD-mediated phos-horylation of LIMK leading to inactive cofilin and an inhibition ofell migration (Spratley et al., 2011). Furthermore, PAK4 regulateshospho-LIMK1 levels via a novel PAK4-binding protein, DGCR6L,nhancing migration of human gastric cancer cells (Li et al., 2010a).

PAK4 is also intrinsically linked to cell adhesion dynamics (Wellst al., 2010). Activated PAK4 expression causes a loss of focaldhesions (Wells et al., 2002) whilst conversely, cells from Pak4nockout mice have an increased number of focal adhesions (Qut al., 2003) as do cells with reduced PAK4 expression (Wellst al., 2010). PAK4 is thought to drive cell adhesion disassemblyither through PAK4-mediated phosphorylation of paxillin at Ser72 (Wells et al., 2010) or via phosphorylation of the cytoplasmicail of integrin �5 (Li et al., 2010b). The integrin �5 cytoplasmicail contains a membrane proximal SERS motif that is critical forAK4 binding whilst the integrin binding site within PAK4 haseen mapped to the C-terminal amino acids 505–533 (Li et al.,010c) (Fig. 1B). However, it is not simply the interaction whichromotes integrin �v�5-mediated cell migration. Importantly, theinase activity of PAK4 is critical, with PAK4 phosphorylating Seresidues 759 and 762 contained within the SERS motif (Li et al.,010c). In this instance PAK4 increases integrin �v�5 turnover atocal adhesions as well as reducing integrin �v�5 clustering, con-ections between the integrin and F-actin, and adhesion complexaturation. The overall outcome of these events is to destabilise

ocal adhesions which in turn decreases cell adhesion strength facil-tating increased cell motility (Li et al., 2010b). There are similaresidues to the PAK4 target site in the sequences of integrins �6nd �8, thus these integrins are also putative PAK4 substrates.

ades mediated by the serine/threonine protein kinase. PAK4 contributes to manyival and anti-apoptotic pathways. Specific details of the pathways are described in

PAK4 and embryonic development

The absolute requirement for PAK4 during development is typ-ified by Pak4 gene deletion in mice. These Pak4 knockout miceare embryonic lethal, surviving only until embryonic day 11.5.This is primarily as a consequence of a defect in the foetal heart(Qu et al., 2003). The cardiac phenotype includes thinning of themyocardium and enlargement of the right atria and ventricles.Moreover, depletion of PAK4 in cardiomyocytes leads to a reductionin phosphorylated LIMK1/2, phosphorylated cofilin and total LIMK1levels (Nekrasova and Minden, 2012). Given that PAK4 activatesLIMK1 which in turn phosphorylates cofilin resulting in inhibi-tion of actin depolymerisation (Ahmed et al., 2008; Dan et al.,2001), disruption of this pathway would account for the disor-ganised sarcomeric structure and contraction of PAK4 knockdowncardiomyocytes (Nekrasova and Minden, 2012). More detailedexamination of the Pak4 knockout embryos has also revealedabnormalities in extraembryonic tissue and embryonic vascula-ture highlighting a role for PAK4 in placenta development, as wellas angiogenesis and vessel branching (Tian et al., 2009). A defi-ciency in angiogenesis is often coupled to a defect in endothelialcell migration. Furthermore, the developing heart is dependenton the correct migration of endothelial cells and so it wouldappear that the abnormally developed heart in combination withdecreased vessels throughout the embryo and extraembryonic tis-sues contribute to the early death of Pak4 knockout mice (Tianet al., 2009). In a complementary study using zebrafish embryosmorpholino-based knockdown of PAK4 also impaired vasculatureformation (Fiedler et al., 2011). Similar vascular defects as well asdisrupted somite development and other morphogenetic abnor-malities were too observed with knockdown of both maternal andzygotic pak4 expression in zebrafish embryos (Law and Sargent,2013). As in mammals, zebrafish pak4 is essential for normal devel-opment and is highly expressed during embryogenesis. Moreover,depleting maternal pak4 mRNA results in embryonic lethality. Themost striking phenotype of the translation-blocking morpholinoin zebrafish is the alterations in gene activity associated withprimitive myelopoiesis, specifically in the development of granu-locyte and leucocyte lineages. Since this phenotype, is to an extent,

recapitulated by the treatment of embryos with actin depoly-merising drugs, such as cytochalasin B, it suggests that maternalpak4 through maintaining actin microfilament integrity can con-trol expression of certain regulatory genes necessary for primitive

urnal

mimtr

rwisFAaooPtdfiIrtwarpbe

P

itebe1ldotickctmoet�ts

feaTitdaam

A.E. Dart, C.M. Wells / European Jo

yeloid cell differentiation (Law and Sargent, 2013). In the future,t would be of considerable interest to the field to generate a

orpholino-based knockdown of Xenopus pak4, another highlyractable model of vertebrate development to establish if this clearole in mammalian and fish embryonic development is conserved.

Pak4 knockout mouse embryos also displayed impaired neu-onal differentiation and migration in the brain and neural tubeith decreased axonal outgrowth and incorrect neuronal target-

ng (Qu et al., 2003). More recently, active PAK4 has also beenhown to promote the neuroendocrine differentiation of LNCap-GC prostate cancer cells stimulated with cAMP (Park et al., 2012).ll these aspects of neuronal biology are tightly regulated by ECMdhesion and cytoskeletal reorganisation reflecting the crucial rolef PAK4 in these cellular processes, and highlighting the importancef the kinase in migration during normal development. ConditionalAK4 knockout mice in which PAK4 was specifically deleted fromhe central nervous system are viable but exhibited growth retar-ation and died prematurely (Tian et al., 2011). In depth analysisound that the brains of these mice had a significant reductionn the population of cortical and striatal neural progenitor cells.n vitro isolation and characterisation of the neural progenitor cellsevealed a decreased proliferative and self-renewing capacity ofhe cells, as assessed by neurosphere formation. In addition, thereas thinning of the cerebral cortex, impaired cortical neurogenesis

nd loss of neuroepithelial adherens junctions. PAK4 is evidentlyequired for the early development of the mammalian brain but atresent it remains to be seen which signalling pathways directedy PAK4 contribute to the functions of the kinase in the brain (Tiant al., 2011).

AK4 and transcriptional/translational regulation

The observation that Group II PAKs are localised in the nucleusmplied that these proteins could play a role in gene transcrip-ion pathways (Cotteret and Chernoff, 2006; Li et al., 2012; Yangt al., 2001). Indeed, PAK4 has been shown to constitutively cycleetween the nucleus and the cytoplasm (Li et al., 2012). The nuclearxport of PAK4 is controlled by three NES motifs through the CRM-/exportin pathway whilst its accumulation within the nucleus is

ikely mediated by two NLS motifs, NLS1 and 3, in an importin �5-ependent manner (Li et al., 2012; Siu et al., 2010). It is the abilityf PAK4 to shuttle between these cell compartments that allowshe kinase to modulate the nucleo-cytoplasmic trafficking of one ofts substrates, �-catenin (Li et al., 2012). �-Catenin is an importantomponent of canonical Wnt signalling and this pathway inducesey cell responses for example, during embryonic development andancer. PAK4 binds and phosphorylates �-catenin on Ser 675 andhrough this interaction regulates Wnt signalling via a two-fold

echanism. Firstly, within the cytoplasm, PAK4 phosphorylationf �-catenin inhibits its degradation, stabilising the protein andnhancing �-catenin-dependent transcriptional activity. Secondly,he nuclear import of PAK4 is associated with the nuclear import of-catenin; within the nucleus, �-catenin interacts with the TCF/LEF

ranscriptional complex driving activating Wnt target gene expres-ion, specifically cyclin D1 and c-myc (Li et al., 2012).

Another aspect of PAK4 regulation of gene expression arisesrom an association between the N-terminal domain, in the pres-nce of RNA, and translational machinery components knowns ribonucleoprotein (RNP) complexes (Baldassa et al., 2010).hrough this interaction, active full length PAK4 can influence cap-ndependent translation of specific IRES sequences in vivo, such ashe apoptotic c-Myc IRES. Furthermore, the N-terminal regulatory

omain appears critical to PAK4 subcellular localisation. The aminocid stretch 65–110 within the RNP interacting sequence contains

targeting element that serves as a NES and promotes cytoplas-ic localisation (Baldassa et al., 2010). It is therefore plausible that

of Cell Biology 92 (2013) 129– 138 133

the RNP association of PAK4 could direct the active shuttling ofthe protein kinase and it is appealing to speculate that this regioncould also be important in maintaining PAK4 in an inhibited statevia intramolecular interactions. Hence, it may be that the subcellu-lar localisation of PAK4 dictates the distinct mechanisms the kinaseplays in transcriptional and translational regulation.

PAK4 and cell cycle

An exciting area of PAK4 biology that has come to light in thelast couple of years is the control of cell cycle progression andregulation of mitosis (Bompard et al., 2010, 2013; Nekrasova andMinden, 2011). The levels of PAK4 expression peak significantlyand transiently in the early stages of the G1 phase of the cell cycle.Upon serum starvation, in cells depleted of PAK4, a decrease in thenumber of cells in G1 is concomitant with an increase of cells inG2/M phase. This observation suggests PAK4 plays a role in theinitiation of the cell cycle by coordinating a critical checkpoint.It has been proposed that to prevent improperly timed entranceof cells into the cell cycle, PAK4 downregulates the cell cycleregulatory protein p21 (CDKN1A) at the transcriptional level, pri-marily through rapid protein degradation. As a consequence, theamount of cyclin D1 being recruited to the assembling CDK4 com-plex would be abrogated (Nekrasova and Minden, 2011). Importantquestions to answer in the future will be to determine whetherthe regulation of p21 protein levels by PAK4 is kinase-dependentand if PAK4 can interact directly with and phosphorylate p21.Given that Cdc42 is also required for cell cycle progression butin the late G1 and G1/S transition phases (Yang et al., 2006), itwill be interesting to establish if PAK4 and Cdc42 have mutuallydependent roles in the G1 phase. At the start of mitosis, the cellmust undergo extensive cytoskeleton rearrangements to generate abipolar mitotic spindle that consequently allows the chromosomesto segregate and cytokinesis to proceed. To date, PAK4 has beenshown to impact upon two key aspects during mitosis. Firstly, it reg-ulates Ran activity by phosphorylating the GTPase in both the GTP-and GDP-bound form on Ser 135 (Bompard et al., 2010). Indeed,PAK4 and phosphorylated Ran dynamically colocalise to specificsubstructures of the microtubule spindle as mitosis advances. Inparticular, at metaphase, centrosome staining was increased forboth active PAK4 and phospho Ran. Phosphorylation of Ran at Ser135 is important for its subsequent interactions with downstreambinding partners since a phosphomimetic S135D mutant of Rancan no longer bind the GEF RCC1 and Ran GAP1. As a result, block-ing RCC1-mediated GDP/GTP exchange of Ran inhibits microtubuleaster nucleation in mitotic Xenopus egg extracts. Therefore, PAK4phosphorylation of Ran likely alters the RCC1-RanGAP1-mediatedregulation of the GTPase (Bompard et al., 2010). Moreover, loss ofPAK4 results in a G2/M block indicating that PAK4 function is notsimply restricted to the G1 phase of the cell cycle (Bompard et al.,2010; Nekrasova and Minden, 2011).

Secondly, PAK4 controls spindle positioning in the cell(Bompard et al., 2013). Depletion of PAK4 causes cells to be retainedin a long metaphase-like arrest distinguished by spindle rotation.The resultant bipolar metaphase-like spindles are not correctlycentred within the cell and have a defective astral microtubulenetwork. In addition, dynein does not efficiently accumulate atkinetochores, microtubule attachment to kinetochores is impairedand part of the dynactin complex, p150Glued is weakly localised atastral microtubules and their plus-end tips (Bompard et al., 2013).However, it is not yet known whether PAK4 could be regulating the

activity of the dynein/dynactin complex by phosphorylating one ormore of the motor protein components. Alternatively, the influ-ence of PAK4 on spindle anchorage could be via its substrate Ran(Bompard et al., 2010, 2013).

1 urnal

P

wgt2mcaast2m2ntpATmraTbPmi(

P

t(taPiremeetpictapmmcimicDditcpt

34 A.E. Dart, C.M. Wells / European Jo

AK4 and cell survival

Constitutively active PAK4 mutants (S445N and S474E) asell as endogenous PAK4 can induce anchorage-independent cell

rowth, a feature of oncogenic transformation that reveals a func-ion for PAK4 in cell growth (Callow et al., 2002; Gnesutta et al.,001; Li and Minden, 2005). PAK4 is thought to be required to pro-ote activation of cell survival pathways under stress and protect

ells from apoptosis. There are two mechanisms through which itchieves this: phosphorylation of the pro-apoptotic protein, BCL-2ntagonist of cell death (BAD) and inhibition of caspase-dependentignalling pathways. PAK4 phosphorylates BAD at Ser 112 blockinghe interaction of BAD with BCL family members, BCL-XL and BCL-

and consequently, inhibiting the release of cytochrome c fromitochondria and preventing apoptosis (Gnesutta and Minden,

003; Gnesutta et al., 2001). Furthermore, in response to tumourecrosis factor (TNF)-� treatment, UV irradiation or serum starva-ion, PAK4 abrogates the activation of initiator caspase 8, therebyroviding resistance to apoptosis (Gnesutta and Minden, 2003).s well as blocking pro-apoptotic signals, PAK4 can also augmentNF-�-induced pro-survival pathways. PAK4 is necessary for opti-al binding of the scaffold protein TRADD to the activated TNF-�

eceptor. Moreover, depletion of PAK4 leads to a reduction in thectivation of NF�B, ERK and JNK pathways (Li and Minden, 2005).he potency of PAK4 in manipulating cell survival is underlinedy it being the only Group II member to exhibit this dual activity.aradoxically, PAK4 has also been recognised as promoting pre-ature cellular senescence through the ERK-MAPK pathway and

nvolving the cell cycle regulatory proteins, p16INK4 and p19ARFCammarano et al., 2005).

AK4 and immune defence

SAGE (serial analysis of gene expression) analysis has shownhat PAK4 (along with PAK2) is expressed in the lymph nodesArias-Romero and Chernoff, 2008). It is therefore unsurprisinghat this kinase has recently been recognised as influencing severalspects of immune cell biology. Evidence in support of a role forAK4 in signalling pathways that regulate the immune responsenitially came from studies of podosomes, the specialised actin-ich adhesive structures found in haematopoietic cells (Gringelt al., 2006). In human macrophages, the kinase activity of PAK4odulates podosome size through altering localised F-actin lev-

ls whilst the PBD of PAK4 influences podosome number (Gringelt al., 2006). PAK4 has also been linked to phagocytosis (an essen-ial mechanism whereby immune cells can engulf and destroyathogens) where loss of PAK4 expression leads to a reduction

n MRLC phosphorylation at Ser 19 during Fc�R-mediated phago-ytosis in macrophages (Bright and Frankel, 2011). Furthermore,he MRLC isoform, MLC9 can be phosphorylated by PAK4 in vitrot this same residue. Internalisation through engagement of thehagocytic receptor requires actin cytoskeletal reorganisation andyosin-based contractility. It therefore seems likely that PAK4 byeans of phosphorylating MLC9 and perhaps other MRLC isoforms

ould regulate stages of phagocytosis that require myosin activ-ty, such as in the closure of phagocytic cups. Recently, through a

ass spectrometry-based phosphoproteome approach, PAK4 wasdentified as being phosphorylated on Ser 181 downstream ofo-stimulation of the co-activating receptors, 2B4 (CD244) andNAM-1 (CD226) in Natural killer (NK) cells. NK cells provideefence during the innate immune response by directly killing

nfected or transformed cells. NK cell activation is triggered by liga-

ion of activating and inhibitory receptors giving rise to signallingascades coordinated by protein kinases. Since reduced Ser 181hosphorylation was also observed following activating Fc recep-or, CD16 engagement (Konig et al., 2012), PAK4 is unique amongst

of Cell Biology 92 (2013) 129– 138

the regulated kinases of NK cells in that ligation of the activatingreceptors does not generate a convergence of signals. Moreover, therole of the Ser 181 phosphorylation site in NK cell effector functionsis yet to be determined, although it is expected to be involved inactin cytoskeleton rearrangements and/or adhesion (Konig et al.,2012). Since this is the first description of PAK4 phosphorylation atSer 181, it will be interesting in the future to establish the signifi-cance of this regulation on PAK4 activity and function.

Often during the course of infectious disease, the functions ofhost proteins are subverted by the viral or bacterial species toinvade and colonise mammalian cells and ultimately, evade thehost immune system. The signalling pathways of PAK4 are noexception to this hijacking by microbial virulence factors. In thecourse of Mycobacterium tuberculosis and Salmonella typhimuriuminfections, PAK4 is targeted by Akt1 which leads to GEF-H1 phos-phorylation and subsequently, affects actin dynamics throughmodulation of RhoA and Rac1 (Kuijl et al., 2007). Pathogen sur-vival also frequently relies on subversion of host innate immuneresponses and to this end PAK4-dependent signalling pathwaysare manipulated by both Klebsiella pneumoniae (Frank et al., 2013)and Streptococcus pneumoniae (Ha et al., 2008). More specifi-cally, K.pneumoniae triggers an EGFR-PI3K-AKT-PAK4-ERK-GSK3�signalling pathway which attenuates IL-1�-dependent nucleartranslocation of NF-�B, thereby blocking the transcription ofinflammatory mediators. Moreover, K.pneumoniae infection ofA549 lung carcinoma cells results in a sustained increase in PAK4Ser 474 phosphorylation (Frank et al., 2013). Similarly, the viru-lence factor pneumolysin produced by S. pneumoniae induced Ser474 phosphorylation of PAK4. During infection, the bacterial com-ponent pneumolysin triggers MUC5AC mucin gene expression tostimulate the host defence response of mucus production. PAK4knockdown was shown to enhance MUC5AC transcription whereasa constitutively active form of PAK4 inhibits this induction indi-cating that the kinase acts as a negative regulator. Furthermore,PAK4 is acting as an upstream kinase controlling JNK activation tonegatively influence MUC5AC induction by pneumolysin (Ha et al.,2008). Hence, by identifying the part played by PAK4 in pathogensurvival, it will provide insight into the roles of PAK4 under physi-ological conditions in the host. In addition, PAK4 could represent atarget in drug design against certain infectious diseases.

PAK4 and disease

Given the key role played by PAK4 in neuronal development (Quet al., 2003; Tian et al., 2011), it is perhaps unsurprising that thiskinase has been associated with the progression of neurologicaldisorders. Aggregation of �-synuclein is a significant contributoryfactor in the deterioration of the central nervous system of patientssuffering from Parkinson’s disease. Mutations in �-synuclein causethe protein to from oligomeric or fibrillar aggregates (Stefanis,2012). It has been shown that the oligomeric form of �-synucleinbut not the monomeric form or �-synuclein can block PAK4 kinaseactivity as assessed by levels of autophosphorylation. Further-more, the phosphorylation of a recognised PAK4 substrate, LIMK1is reduced in brainstem protein extracts from �-synuclein (A30P)transgenic mice (Danzer et al., 2007). Inhibition of PAK activitymay represent a general mechanism in the pathology of diseasescoupled to the production of aggregated protein species since acti-vation of PAKs 1, 2 and 3 are repressed in brain lysates from patientswith Alzheimer’s disease (Zhao et al., 2006). It would be interest-ing in the future to see whether PAK4 activity is also suppressed inAlzheimer’s disease.

PAK4 also functions heavily in vascular development(Nekrasova and Minden, 2012; Tian et al., 2009) and con-versely, has been identified as influencing the pathophysiology ofvascularity during myocardial infarction (Fiedler et al., 2011). The

urnal

kicefswp

fitcepgeia2ccecfislcaiitoifdMmdAwtaesiogfi4saK

abagadg2ncwoa

A.E. Dart, C.M. Wells / European Jo

inase is a target of miR-24, a microRNA (miRNA) highly expressedn cardiac endothelial cells and whose levels are elevated uponardiac ischaemia. Repression of PAK4 mRNA by miR-24 promotesndothelial cell apoptosis, blocks endothelial capillary networkormation on matrigel and prevents cell sprouting from endothelialpheroids. The increased apoptosis observed in endothelial cellsas found to be the result of a reduction in miR-24-mediated BADhosphorylation through PAK4 (Fiedler et al., 2011).

PAK4 has been found to be overexpressed, genetically ampli-ed and/or point mutated in a number of cancer cell lines andumours including breast (Callow et al., 2002; Liu et al., 2010), pan-reas (Chen et al., 2008; Kimmelman et al., 2008), colon (Parsonst al., 2005), lung (Callow et al., 2002), ovarian (Siu et al., 2010),rostate, melanoma, CNS, renal, leukaemia (Callow et al., 2002),astric (Ahn et al., 2011) and oral squamous cell carcinoma (Begumt al., 2009). Indeed, the genomic region of the Pak4 locus (19q13.2)s frequently amplified in colorectal and pancreatic cancers as wells aggressive breast cancers with basal-like features (Chen et al.,008; Parsons et al., 2005; Yu et al., 2009). In the MCF10 cell series (aell line model that recapitulates the different stages of breast can-er progression using MCF10A, AT1, DCIS.com and CA1a cells) thexpression levels of PAK4 correlate with the tumorgenicity of theell lines (So et al., 2012). The PAK4 gene is also commonly ampli-ed in high grade serous and endometrioid cancer and as such PAK4ilencing in ovarian cancer cell lines has been shown to result in aoss of viability (Davis et al., 2013). More specifically, there is PAK4opy number gain in 21%, amplification in 8%, and high amplitudemplification in 5% of high grade serous and endometrioid ovar-an tumours (Gorringe et al., 2010). Recently, studies have usedmmunohistochemistry of tissues to demonstrate there is a posi-ive correlation between PAK4 overexpression and the progressionf certain cancers. One such example is that of ovarian cancer wherencreased expression of PAK4 and the phosphorylated Ser 474orm of PAK4 are associated with metastasis, shorter overall andisease-free survival of patients and resistance to chemotherapy.oreover, strong PAK4 immunoreactivity was found in carcino-as of advanced stages (stages III and IV) and poor histological

ifferentiation (grade 3) as well as metastatic foci (Siu et al., 2010).nother study highlighting the pathological expression of PAK4ith malignancy has been demonstrated with glioblastoma mul-

iforme (GBM) (Kesanakurti et al., 2012). Immunohistochemistrynalysis of a GBM tissue microarray showed an extensive, elevatedxpression of PAK4 in gliomas compared to no significant expres-ion in normal brain tissues. Importantly, an association betweenncreasing PAK4 expression and glioma pathological grades wasbserved, with moderate levels in grade 2 and high levels inrade 3 and 4 tumours. Moreover, patient tumour biopsies con-rmed increased upregulation of PAK4 and phospho-PAK4 (Ser74) (Kesanakurti et al., 2012). Collectively, these studies stronglyuggest PAK4 has an oncogenic role in the initiation, progressionnd clinical outcome of several cancer types (Davis et al., 2013;esanakurti et al., 2012; Siu et al., 2010).

In vitro PAK4 overexpression frequently displays essentialspects of oncogenic transformation and related changes in cellehaviour; these include regulation of cell survival and prolifer-tion. Activated PAK4 is able to promote anchorage-independentrowth in immortalised fibroblasts in soft agar focus formationssays (Callow et al., 2002; Qu et al., 2001) and conversely a kinaseead mutant of PAK4 is able to inhibit the anchorage-independentrowth of a human colon cancer cell line, HCT116 (Callow et al.,002). During tumorigenesis, the cells must rapidly expand inumbers. Indeed, both wild-type and constitutively active PAK4

an stimulate ovarian cancer cell proliferation and conversely,hen PAK4 is depleted can reduce proliferation. This regulation

f cell proliferation in ovarian cancer cell lines by PAK4 is medi-ted through c-Src phosphorylation of the EGFR (Tyr 845), which

of Cell Biology 92 (2013) 129– 138 135

enhances cyclin D1 and CDC25A expression (Siu et al., 2010). Sim-ilarly, in gestational trophoblastic disease, which is characterisedby malignant choriocarcinoma, knockdown of PAK4 in choriocarci-noma cell lines decreases proliferation. However, unlike in ovariancancer cell lines this was not through dysregulation of cyclin D1and CDC25A gene expression but through changes in p16 andCDK6 levels (Zhang et al., 2011). In addition, the proliferation oflaryngeal carcinoma cells can be decreased upon PAK4 knock-down; Hep-2 cells depleted of PAK4 are arrested in the S phaseof the cell cycle, a phenotype resulting from the activation of theATM/Chk1/2/p53 pathway (Sun et al., 2013). PAK4 has also beenshown to regulate the anchorage-dependent and -independentproliferation of colon carcinoma cell lines harbouring K-RAS andB-RAF mutations (Tabusa et al., 2013). In stark contrast to other can-cers, knockdown of PAK4 in mutant K-RAS HCT116 colorectal cellsreduced proliferation independently of RAF/MEK/ERK/cyclin D1and PI3K/AKT signalling. However, the signalling pathway throughwhich PAK4 promotes colon cancer cell proliferation is yet to beidentified. Interestingly, the suppression of proliferation observedupon depletion of PAK4 in this study was shown to occur irrespec-tive of PAK4 protein levels indicating that gene amplification and/orincreased protein expression of PAK4 may not be a determinant oftumour cell proliferation (Tabusa et al., 2013).

Overexpressed PAK4 can protect cells from apoptosis andtherefore, has a role in promoting cell survival during tumourprogression (Gnesutta and Minden, 2003; Gnesutta et al., 2001).Equally, cells lacking PAK4 have an elevated susceptibility to apo-ptosis (Li and Minden, 2005). Indeed, elevated anoikis is observedwhen PAK4 is silenced in human glioma xenograft cell lines andthis phenotype is attributed to the direct interaction as well asfunctional association of PAK4 with the matrix metalloproteinase,MMP-2 (Kesanakurti et al., 2012). Injection of NIH3T3 fibroblastsstably overexpressing wild-type and constitutively active PAK4into athymic mice leads to tumour formation and oncogenic V12Rastumour formation is attenuated in a PAK4 null background (Liuet al., 2008). Similarly, stable knockdown of PAK4 in prostate can-cer and ovarian cancer cell lines reduced tumour growth in athymicmice (Park et al., 2012; Siu et al., 2010). In a xenografted laryn-geal carcinoma Hep-2 tumour mouse model, the survival rate ofmice with tumours lacking PAK4 was found to be considerablyhigher when compared to mice with tumours expressing PAK4(Sun et al., 2013). Interestingly, suppression of PAK4 in gliomaxenograft cells impaired intracranial tumour growth in mice andthis inhibition corresponded with decreased levels of phospho-EGFR, �3 integrin and MMP-2 (Kesanakurti et al., 2012). Moreover,PAK4-overexpressing immortalised mouse mammary epithelialcells lose their differentiated architecture and when implanted intothe mammary fat pads of athymic mice form tumours (Liu et al.,2010). Taken together, these studies strongly link PAK4 to tumourformation and suggest that PAK4 is both essential and sufficient formalignant cell growth (Liu et al., 2008, 2010; Park et al., 2012; Siuet al., 2010).

PAK4 is not only associated with primary tumour growth buthas also been implicated in cancer cell metastasis. The centralcomponents of cancer cell metastasis are migration and inva-sion; be that through the basement membrane into the ECM orthrough endothelial cells in order to enter the lymph or blood-stream. Both of these steps are driven by reorganisation of the actincytoskeleton and regulation of cell: substratum adhesion dynam-ics, processes known to be regulated by PAK4. Prostate cancer cellsdepleted of PAK4 are less responsive to HGF stimulation and dis-play reduced motility (Wells et al., 2010) clearly demonstrating a

link between PAK4 and migration. Moreover, ovarian cancer cellmigration and invasion are promoted by PAK4 kinase activity andinvolve a PAK4-directed downstream signalling pathway compris-ing of c-Src, MEK-1/ERK1/2 and MMP-2 (Siu et al., 2010). In line

1 urnal

wcieebeheboobgphacdb(sPipCa32P

T

iptbehdwcihtec7opP

ubddtkMatcmar

36 A.E. Dart, C.M. Wells / European Jo

ith these studies, it was also shown that overexpression of aonstitutively active form of PAK4 enhances the migratory andnvasive capacities of pancreatic ductal cells in vitro (Kimmelmant al., 2008) whereas siRNA-mediated downregulation of PAK4 inndometrial cells and glioma xenograft cells significantly lessensoth cell viability and invasiveness (Kesanakurti et al., 2012; Kimt al., 2013). Furthermore, PAK4 deficient PC3 prostate cancer cellsave reduced LIMK1-driven cell motility in response to HGF (Whalet al., 2012). Equally, PAK4 contributes to oncogene ErbB2-inducedreast cancer cell invasiveness (Rafn et al., 2012). Overexpressionf PAK4 in hepatocellular carcinoma tissue is correlated with thatf the CDK5 kinase-associated protein, CDK5RAP3 which itself haseen documented to play a prometastatic role in hepatocarcino-enesis (Mak et al., 2011). More recently, an in vivo SILAC-basedroteomics approach to investigate mouse skin cancer progressionas highlighted the importance of dysregulated cell growth and celldhesion as key drivers of malignancy (Zanivan et al., 2013). Using ahemical-induced multistage skin carcinogenesis model with threeistinct stages – precancerous, mutant H-RAS skin, pre-malignant,enign papilloma lesions and malignant squamous cell carcinomaSCC) tumours – this study identified a PAK4-PKC/Src network to beignificantly perturbed in SCC and not in papilloma. Furthermore,AK4 was revealed to phosphorylate RAF1 at Ser 43 and Src at Ser 17n SCC but not in benign tumours and notably, showed increasedhosphorylation at the functionally uncharacterised site Ser 181.omplementing the phosphoproteomic data, PAK4-depleted A431nd SCC9 skin cancer cell invasiveness was highly decreased inD collagen-based and organotypic invasion assays (Zanivan et al.,013). This study and the many others strongly implicate aberrantAK4 signalling in invasive disease.

argeting PAK4

Since the number of studies identifying PAK4 as a key factorn cancer progression is rapidly increasing it is perhaps not unex-ected that it has become a popular candidate as a potential drugarget. A pan-PAK ATP-competitive inhibitor (PF-3758309) haseen identified as potently blocking PAK4 kinase activity (Murrayt al., 2010). PF-3758309 treatment will block the growth of severaluman tumour xenografts (Murray et al., 2010) but unfortunatelyuring clinical trials on patients with advanced solid tumours thereere undesirable pharmacokinetic characteristics. As a result, the

linical trial has been terminated (clinicaltrials.gov). A second PAK4nhibitor LCH-7749944 has initial promising results suppressinguman gastric cancer cell proliferation and impeding the migra-ory and invasive capacity of human gastric cancer cells (Zhangt al., 2012). However, this study only focussed on gastric cancerell lines and it will be of interest to demonstrate whether LCH-749944 could potentially be effective as a therapeutic target forther cancers (Zhang et al., 2012). In contrast an autoinhibitoryseudosubstrate peptide might represent a useful means to targetAK4 activity in cells.

Whilst the exploitation of PAK4-specific inhibitors may offers currently the best approach to cancer therapeutics, we shoulde cautious. Firstly, it will be important to delineate the kinase-ependent from the kinase-independent functions of PAK4 beforeeveloping it further as a drug target. This is particularly true inhe case of competitive ATP inhibitors. PAK4 is not simply a proteininase and it has other roles independent of its catalytic activity.oreover, PAK4 point mutations may lead to inhibitor resistance

s recently highlighted by Whale et al. (2012). A somatic muta-ion within the PAK4 kinase domain (E329K) identified in colon

arcinoma patients (Parsons et al., 2005) enhances HGF-drivenigratory capacity of prostate cancer cells. In addition, this mutant

lso moderately increases the kinase activity of PAK4 and confersesistance to competitive ATP inhibitors (Whale et al., 2012). This

of Cell Biology 92 (2013) 129– 138

work highlights the importance of developing multiple and alter-native therapeutic strategies (for example allosteric PAK inhibitors(Viaud and Peterson, 2009)) to offset the risks associated with clin-ical drug resistance.

Although not yet therapeutically explored, an alternative strat-egy to target PAK4 may lie in the use of miRNAs. Recently, onespecific miRNA, miR-24 upon overexpression has been linked tomodulation of actin adhesion dynamics and epithelial differentia-tion/stratification in keratinocytes. The mode of action of miR-24is likely to be through repression of its target PAK4. Given this keyrole for miR-24 and PAK4 in skin morphogenesis and the hyper-proliferative nature of skin cancers, it may represent a potentialtherapeutic tool (Amelio et al., 2012). Levels of miR-199a/b-3p (ahighly expressed liver miRNA) are consistently reduced in hepato-cellular carcinoma; a decrease that is linked with poor survival ofpatients. miR-199a/b-3p represses PAK4 expression which in turnis able to suppress hepatocellular carcinoma growth both in vitroand in vivo (Hou et al., 2011). Similarly, miR-145 targets PAK4 andlevels of miR-145 have been reported to be downregulated in coloncancer cells (Wang et al., 2012).

Concluding remarks

PAK4 is a pluripotent kinase regulating many biological pro-cesses ranging from cell motility to survival. To date, it is the bestcharacterised of the Group II PAKs with more diverse roles for thiskinase rapidly emerging every year, such as in immune defence,a previously unrecognised function for the kinase. Nevertheless,there is now a significant body of work linking PAK4 to the crit-ical control of the cytoskeleton. As more substrates and bindingpartners are discovered for PAK4, it will be crucial to delineate thefunctional importance of these interactions in vivo, especially giventhe often shared substrate specificity with PAK1. As we begin tounravel the regulatory mechanism of kinase activity for the GroupII PAKs, it will be necessary to establish the physiological relevanceof PAK4 activation in terms of spatial and temporal control. Thenext step will be to find physiologically relevant activators of PAK4.Attractive candidates would be PKD (Spratley et al., 2011), Src (Haet al., 2012) and CDK5RAP3 which has been demonstrated to inter-act with the N-terminal region of PAK4 (Mak et al., 2011). Althoughsmall molecule inhibitors of PAK4 have yet to prove successful,the continued understanding of PAK4 biology will be a valuableresource in strengthening the rationale for new therapeutic inter-ventions. The future of PAK4 research is an exciting prospect as weexpand our knowledge of the roles of this kinase and validate itseffectiveness as a drug target.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.ejcb.2013.03.002.

References

Abo, A., Qu, J., Cammarano, M.S., Dan, C., Fritsch, A., Baud, V., Belisle, B., Minden, A.,1998. PAK4, a novel effector for Cdc42Hs, is implicated in the reorganization ofthe actin cytoskeleton and in the formation of filopodia. EMBO J. 17, 6527–6540.

Ahmed, T., Shea, K., Masters, J.R., Jones, G.E., Wells, C.M., 2008. A PAK4-LIMK1 path-way drives prostate cancer cell migration downstream of HGF. Cell. Signal. 20,1320–1328.

Ahn, H.K., Jang, J., Lee, J., Se Hoon, P., Park, J.O., Park, Y.S., Lim, H.Y., Kim, K.M., Kang,W.K., 2011. P21-activated kinase 4 overexpression in metastatic gastric cancerpatients. Transl. Oncol. 4, 345–349.

Amelio, I., Lena, A.M., Viticchie, G., Shalom-Feuerstein, R., Terrinoni, A., Dinsdale,D., Russo, G., Fortunato, C., Bonanno, E., Spagnoli, L.G., Aberdam, D., Knight,R.A., Candi, E., Melino, G., 2012. miR-24 triggers epidermal differentiation bycontrolling actin adhesion and cell migration. J. Cell Biol. 199, 347–363.

Arias-Romero, L.E., Chernoff, J., 2008. A tale of two PAKs. Biol. Cell 100, 97–108.

urnal

A

B

B

B

B

B

B

B

B

C

C

C

C

C

C

C

C

D

D

D

D

E

E

E

E

F

F

G

A.E. Dart, C.M. Wells / European Jo

spenstrom, P., Fransson, A., Saras, J., 2004. Rho GTPases have diverse effects on theorganization of the actin filament system. Biochem. J. 377, 327–337.

aldassa, S., Calogero, A.M., Colombo, G., Zippel, R., Gnesutta, N., 2010. N-terminalinteraction domain implicates PAK4 in translational regulation and reveals novelcellular localization signals. J. Cell. Physiol. 224, 722–733.

arac, A., Basile, J., Vazquez-Prado, J., Gao, Y., Zheng, Y., Gutkind, J.S., 2004. Directinteraction of p21-activated kinase 4 with PDZ-RhoGEF, a G protein-linked Rhoguanine exchange factor. J. Biol. Chem. 279, 6182–6189.

askaran, Y., Ng, Y.W., Selamat, W., Ling, F.T., Manser, E., 2012. Group I and IImammalian PAKs have different modes of activation by Cdc42. EMBO Rep. 13,653–659.

eeser, A., Jaffer, Z.M., Hofmann, C., Chernoff, J., 2005. Role of group A p21-activatedkinases in activation of extracellular-regulated kinase by growth factors. J. Biol.Chem. 280, 36609–36615.

egum, A., Imoto, I., Kozaki, K., Tsuda, H., Suzuki, E., Amagasa, T., Inazawa, J.,2009. Identification of PAK4 as a putative target gene for amplification within19q13.12-q13.2 in oral squamous-cell carcinoma. Cancer Sci. 100, 1908–1916.

ompard, G., Rabeharivelo, G., Cau, J., Abrieu, A., Delsert, C., Morin, N., 2013. P21-activated kinase 4 (PAK4) is required for metaphase spindle positioning andanchoring. Oncogene 32, 910–919.

ompard, G., Rabeharivelo, G., Frank, M., Cau, J., Delsert, C., Morin, N., 2010. SubgroupII PAK-mediated phosphorylation regulates Ran activity during mitosis. J. CellBiol. 190, 807–822.

right, M.D., Frankel, G., 2011. PAK4 phosphorylates myosin regulatory light chainand contributes to Fcgamma receptor-mediated phagocytosis. Int. J. Biochem.Cell Biol. 43, 1776–1781.

allow, M.G., Clairvoyant, F., Zhu, S., Schryver, B., Whyte, D.B., Bischoff, J.R., Jallal, B.,Smeal, T., 2002. Requirement for PAK4 in the anchorage-independent growthof human cancer cell lines. J. Biol. Chem. 277, 550–558.

allow, M.G., Zozulya, S., Gishizky, M.L., Jallal, B., Smeal, T., 2005. PAK4 medi-ates morphological changes through the regulation of GEF-H1. J. Cell Sci. 118,1861–1872.

ammarano, M.S., Nekrasova, T., Noel, B., Minden, A., 2005. Pak4 induces prematuresenescence via a pathway requiring p16INK4/p19ARF and mitogen-activatedprotein kinase signaling. Mol. Cell. Biol. 25, 9532–9542.

haudhary, A., King, W.G., Mattaliano, M.D., Frost, J.A., Diaz, B., Morrison, D.K., Cobb,M.H., Marshall, M.S., Brugge, J.S., 2000. Phosphatidylinositol 3-kinase regulatesRaf1 through Pak phosphorylation of serine 338. Curr. Biol. 10, 551–554.

hen, S., Auletta, T., Dovirak, O., Hutter, C., Kuntz, K., El-ftesi, S., Kendall, J., Han,H., Von Hoff, D.D., Ashfaq, R., Maitra, A., Iacobuzio-Donahue, C.A., Hruban, R.H.,Lucito, R., 2008. Copy number alterations in pancreatic cancer identify recurrentPAK4 amplification. Cancer Biol. Ther. 7, 1793–1802.

hing, Y.P., Leong, V.Y., Wong, C.M., Kung, H.F., 2003. Identification of anautoinhibitory domain of p21-activated protein kinase 5. J. Biol. Chem. 278,33621–33624.

otteret, S., Chernoff, J., 2006. Nucleocytoplasmic shuttling of Pak5 regulates itsantiapoptotic properties. Mol. Cell. Biol. 26, 3215–3230.

otteret, S., Jaffer, Z.M., Beeser, A., Chernoff, J., 2003. p21-Activated kinase 5 (Pak5)localizes to mitochondria and inhibits apoptosis by phosphorylating BAD. Mol.Cell. Biol. 23, 5526–5539.

an, C., Kelly, A., Bernard, O., Minden, A., 2001. Cytoskeletal changes regulated bythe PAK4 serine/threonine kinase are mediated by LIM kinase 1 and cofilin. J.Biol. Chem. 276, 32115–32121.

anzer, K.M., Schnack, C., Sutcliffe, A., Hengerer, B., Gillardon, F., 2007. Functionalprotein kinase arrays reveal inhibition of p-21-activated kinase 4 by alpha-synuclein oligomers. J. Neurochem. 103, 2401–2407.

avis, S.J., Sheppard, K.E., Pearson, R.B., Campbell, I.G., Gorringe, K.L., Simpson, K.J.,2013. Functional analysis of genes in regions commonly amplified in high-gradeserous and endometrioid ovarian cancer. Clin. Cancer Res. 19, 1411–1421.

ong, J.M., Lau, L.S., Ng, Y.W., Lim, L., Manser, E., 2009. Paxillin nuclear-cytoplasmiclocalization is regulated by phosphorylation of the LD4 motif: evidence thatnuclear paxillin promotes cell proliferation. Biochem. J. 418, 173–184.

din, M.L., Juliano, R.L., 2005. Raf-1 serine 338 phosphorylation plays a key rolein adhesion-dependent activation of extracellular signal-regulated kinase byepidermal growth factor. Mol. Cell. Biol. 25, 4466–4475.

dwards, D.C., Sanders, L.C., Bokoch, G.M., Gill, G.N., 1999. Activation of LIM-kinaseby Pak1 couples Rac/Cdc42 GTPase signalling to actin cytoskeletal dynamics.Nat. Cell Biol. 1, 253–259.

swaran, J., Lee, W.H., Debreczeni, J.E., Filippakopoulos, P., Turnbull, A., Fedorov,O., Deacon, S.W., Peterson, J.R., Knapp, S., 2007. Crystal structures of the p21-activated kinases PAK4, PAK5, and PAK6 reveal catalytic domain plasticity ofactive group II PAKs. Structure 15, 201–213.

swaran, J., Soundararajan, M., Kumar, R., Knapp, S., 2008. UnPAKing the classdifferences among p21-activated kinases. Trends Biochem. Sci. 33, 394–403.

iedler, J., Jazbutyte, V., Kirchmaier, B.C., Gupta, S.K., Lorenzen, J., Hartmann, D.,Galuppo, P., Kneitz, S., Pena, J.T., Sohn-Lee, C., Loyer, X., Soutschek, J., Brand,T., Tuschl, T., Heineke, J., Martin, U., Schulte-Merker, S., Ertl, G., Engelhardt, S.,Bauersachs, J., Thum, T., 2011. MicroRNA-24 regulates vascularity after myocar-dial infarction. Circulation 124, 720–730.

rank, C.G., Reguerio, V., Rother, M., Moranta, D., Maeurer, A.P., Garmendia,J., Meyer, T.F., Bengoechea, J.A., 2013. Klebsiella pneumoniae targets an

EGF receptor-dependent pathway to subvert inflammation. Cell. Microbiol.,http://dx.doi.org/10.1111/cmi.12110 [Epub ahead of print].

nesutta, N., Minden, A., 2003. Death receptor-induced activation of initiator cas-pase 8 is antagonized by serine/threonine kinase PAK4. Mol. Cell. Biol. 23,7838–7848.

of Cell Biology 92 (2013) 129– 138 137

Gnesutta, N., Qu, J., Minden, A., 2001. The serine/threonine kinase PAK4 pre-vents caspase activation and protects cells from apoptosis. J. Biol. Chem. 276,14414–14419.

Gorringe, K.L., George, J., Anglesio, M.S., Ramakrishna, M., Etemadmoghadam, D.,Cowin, P., Sridhar, A., Williams, L.H., Boyle, S.E., Yanaihara, N., Okamoto, A.,Urashima, M., Smyth, G.K., Campbell, I.G., Bowtell, D.D., 2010. Copy numberanalysis identifies novel interactions between genomic loci in ovarian cancer.PLoS ONE 5, e11408.

Gringel, A., Walz, D., Rosenberger, G., Minden, A., Kutsche, K., Kopp, P., Lin-der, S., 2006. PAK4 and alphaPIX determine podosome size and number inmacrophages through localized actin regulation. J. Cell. Physiol. 209, 568–579.

Ha, B.H., Davis, M.J., Chen, C., Lou, H.J., Gao, J., Zhang, R., Krauthammer, M., Halaban,R., Schlessinger, J., Turk, B.E., Boggon, T.J., 2012. Type II p21-activated kinases(PAKs) are regulated by an autoinhibitory pseudosubstrate. Proc. Natl. Acad. Sci.U.S.A. 109, 16107–16112.

Ha, U.H., Lim, J.H., Kim, H.J., Wu, W., Jin, S., Xu, H., Li, J.D., 2008. MKP1 regulatesthe induction of MUC5AC mucin by Streptococcus pneumoniae pneumolysin byinhibiting the PAK4-JNK signaling pathway. J. Biol. Chem. 283, 30624–30631.

He, H., Shulkes, A., Baldwin, G.S., 2008. PAK1 interacts with beta-catenin and isrequired for the regulation of the beta-catenin signalling pathway by gastrins.Biochim. Biophys. Acta 1783, 1943–1954.

Hou, J., Lin, L., Zhou, W., Wang, Z., Ding, G., Dong, Q., Qin, L., Wu, X., Zheng, Y., Yang,Y., Tian, W., Zhang, Q., Wang, C., Zhuang, S.M., Zheng, L., Liang, A., Tao, W., Cao,X., 2011. Identification of miRNomes in human liver and hepatocellular carci-noma reveals miR-199a/b-3p as therapeutic target for hepatocellular carcinoma.Cancer Cell 19, 232–243.

Jakobi, R., Moertl, E., Koeppel, M.A., 2001. p21-activated protein kinase gamma-PAK suppresses programmed cell death of BALB3T3 fibroblasts. J. Biol. Chem.276, 16624–16634.

Jin, S., Zhuo, Y., Guo, W., Field, J., 2005. p21-activated kinase 1 (Pak1)-dependentphosphorylation of Raf-1 regulates its mitochondrial localization, phosphoryla-tion of BAD, and Bcl-2 association. J. Biol. Chem. 280, 24698–24705.

Kesanakurti, D., Chetty, C., Rajasekhar Maddirela, D., Gujrati, M., Rao, J.S., 2012.Functional cooperativity by direct interaction between PAK4 and MMP-2 in theregulation of anoikis resistance, migration and invasion in glioma. Cell DeathDis. 3, e445.

Kim, S.H., Kim, S.R., Ihm, H.J., Oh, Y.S., Chae, H.D., Kim, C.H., Kang, B.M., 2013.Regulation of P21-activated kinase-4 by progesterone and tumor necrosis

factor-alpha in human endometrium and its increased expression in advanced-stage endometriosis. J. Clin. Endocrinol. Metab. 98, E238–E248.

Kimmelman, A.C., Hezel, A.F., Aguirre, A.J., Zheng, H., Paik, J.H., Ying, H., Chu, G.C.,Zhang, J.X., Sahin, E., Yeo, G., Ponugoti, A., Nabioullin, R., Deroo, S., Yang, S., Wang,X., McGrath, J.P., Protopopova, M., Ivanova, E., Zhang, J., Feng, B., Tsao, M.S., Red-ston, M., Protopopov, A., Xiao, Y., Futreal, P.A., Hahn, W.C., Klimstra, D.S., Chin,L., DePinho, R.A., 2008. Genomic alterations link Rho family of GTPases to thehighly invasive phenotype of pancreas cancer. Proc. Natl. Acad. Sci. U.S.A. 105,19372–19377.

King, A.J., Sun, H., Diaz, B., Barnard, D., Miao, W., Bagrodia, S., Marshall, M.S., 1998. Theprotein kinase Pak3 positively regulates Raf-1 activity through phosphorylationof serine 338. Nature 396, 180–183.

Konig, S., Nimtz, M., Scheiter, M., Ljunggren, H.G., Bryceson, Y.T., Jansch, L., 2012.Kinome analysis of receptor-induced phosphorylation in human natural killercells. PLoS ONE 7, e29672.

Kosoff, R., Chow, H.Y., Radu, M., Chernoff, J., 2013. Pak2 kinase restrains mast cellFc{epsilon}RI receptor signaling through modulation of Rho protein guaninenucleotide exchange factor (GEF) activity. J. Biol. Chem. 288, 974–983.

Kuijl, C., Savage, N.D., Marsman, M., Tuin, A.W., Janssen, L., Egan, D.A., Ketema, M.,van den Nieuwendijk, R., van den Eeden, S.J., Geluk, A., Poot, A., van der Marel, G.,Beijersbergen, R.L., Overkleeft, H., Ottenhoff, T.H., Neefjes, J., 2007. Intracellularbacterial growth is controlled by a kinase network around PKB/AKT1. Nature450, 725–730.

Law, S.H., Sargent, T.D., 2013. Maternal pak4 expression is required for primitivemyelopoiesis in zebrafish. Mech. Dev. 130, 181–194.

Li, X., Ke, Q., Li, Y., Liu, F., Zhu, G., Li, F., 2010a. DGCR6L, a novel PAK4 interactionprotein, regulates PAK4-mediated migration of human gastric cancer cell viaLIMK1. Int. J. Biochem. Cell Biol. 42, 70–79.

Li, X., Minden, A., 2005. PAK4 functions in tumor necrosis factor (TNF) alpha-inducedsurvival pathways by facilitating TRADD binding to the TNF receptor. J. Biol.Chem. 280, 41192–41200.

Li, Y., Shao, Y., Tong, Y., Shen, T., Zhang, J., Gu, H., Li, F., 2012. Nucleo-cytoplasmicshuttling of PAK4 modulates beta-catenin intracellular translocation andsignaling. Biochim. Biophys. Acta 1823, 465–475.

Li, Z., Lock, J.G., Olofsson, H., Kowalewski, J.M., Teller, S., Liu, Y., Zhang, H., Stromblad,S., 2010b. Integrin-mediated cell attachment induces a PAK4-dependent feed-back loop regulating cell adhesion through modified integrin alpha v beta 5clustering and turnover. Mol. Biol. Cell 21, 3317–3329.

Li, Z., Zhang, H., Lundin, L., Thullberg, M., Liu, Y., Wang, Y., Claesson-Welsh,L., Stromblad, S., 2010c. p21-activated kinase 4 phosphorylation of integ-rin beta5 Ser-759 and Ser-762 regulates cell migration. J. Biol. Chem. 285,23699–23710.

Liu, Y., Chen, N., Cui, X., Zheng, X., Deng, L., Price, S., Karantza, V., Minden, A., 2010.

The protein kinase Pak4 disrupts mammary acinar architecture and promotes

mammary tumorigenesis. Oncogene 29, 5883–5894.Liu, Y., Xiao, H., Tian, Y., Nekrasova, T., Hao, X., Lee, H.J., Suh, N., Yang, C.S., Minden, A.,

2008. The pak4 protein kinase plays a key role in cell survival and tumorigenesisin athymic mice. Mol. Cancer Res. 6, 1215–1224.

1 urnal

M

M

N

N

N

P

P

P

Q

Q

R

S

S

S

S

S

S

T

38 A.E. Dart, C.M. Wells / European Jo

ak, G.W., Chan, M.M., Leong, V.Y., Lee, J.M., Yau, T.O., Ng, I.O., Ching, Y.P., 2011.Overexpression of a novel activator of PAK4, the CDK5 kinase-associated pro-tein CDK5RAP3, promotes hepatocellular carcinoma metastasis. Cancer Res. 71,2949–2958.

urray, B.W., Guo, C., Piraino, J., Westwick, J.K., Zhang, C., Lamerdin, J., Dagostino,E., Knighton, D., Loi, C.M., Zager, M., Kraynov, E., Popoff, I., Christensen, J.G.,Martinez, R., Kephart, S.E., Marakovits, J., Karlicek, S., Bergqvist, S., Smeal, T.,2010. Small-molecule p21-activated kinase inhibitor PF-3758309 is a potentinhibitor of oncogenic signaling and tumor growth. Proc. Natl. Acad. Sci. U.S.A.107, 9446–9451.

ayal, A., Webb, D.J., Brown, C.M., Schaefer, E.M., Vicente-Manzanares, M., Horwitz,A.R., 2006. Paxillin phosphorylation at Ser273 localizes a GIT1-PIX-PAK complexand regulates adhesion and protrusion dynamics. J. Cell Biol. 173, 587–589.

ekrasova, T., Minden, A., 2011. PAK4 is required for regulation of the cell-cycleregulatory protein p21, and for control of cell-cycle progression. J. Cell. Biochem.112, 1795–1806.

ekrasova, T., Minden, A., 2012. Role for p21-activated kinase PAK4 in developmentof the mammalian heart. Transgenic Res. 21, 797–811.

aliouras, G.N., Naujokas, M.A., Park, M., 2009. Pak4, a novel Gab1 binding partner,modulates cell migration and invasion by the Met receptor. Mol. Cell. Biol. 29,3018–3032.

ark, M.H., Lee, H.S., Lee, C.S., You, S.T., Kim, D.J., Park, B.H., Kang, M.J.,Heo, W.D., Shin, E.Y., Schwartz, M.A., Kim, E.G., 2012. p21-Activatedkinase 4 promotes prostate cancer progression through CREB. Oncogene,http://dx.doi.org/10.1038/onc.2012.255 [Epub ahead of print].

arsons, D.W., Wang, T.L., Samuels, Y., Bardelli, A., Cummins, J.M., DeLong, L., Silli-man, N., Ptak, J., Szabo, S., Willson, J.K., Markowitz, S., Kinzler, K.W., Vogelstein, B.,Lengauer, C., Velculescu, V.E., 2005. Colorectal cancer: mutations in a signallingpathway. Nature 436, 792.

u, J., Cammarano, M.S., Shi, Q., Ha, K.C., de Lanerolle, P., Minden, A., 2001. ActivatedPAK4 regulates cell adhesion and anchorage-independent growth. Mol. Cell. Biol.21, 3523–3533.

u, J., Li, X., Novitch, B.G., Zheng, Y., Kohn, M., Xie, J.M., Kozinn, S., Bronson, R., Beg,A.A., Minden, A., 2003. PAK4 kinase is essential for embryonic viability and forproper neuronal development. Mol. Cell. Biol. 23, 7122–7133.

afn, B., Nielsen, C.F., Andersen, S.H., Szyniarowski, P., Corcelle-Termeau, E., Valo,E., Fehrenbacher, N., Olsen, C.J., Daugaard, M., Egebjerg, C., Bottzauw, T., Koho-nen, P., Nylandsted, J., Hautaniemi, S., Moreira, J., Jaattela, M., Kallunki, T., 2012.ErbB2-driven breast cancer cell invasion depends on a complex signaling net-

work activating myeloid zinc finger-1-dependent cathepsin B expression. Mol.Cell 45, 764–776.

iu, M.K., Chan, H.Y., Kong, D.S., Wong, E.S., Wong, O.G., Ngan, H.Y., Tam, K.F., Zhang,H., Li, Z., Chan, Q.K., Tsao, S.W., Stromblad, S., Cheung, A.N., 2010. p21-activatedkinase 4 regulates ovarian cancer cell proliferation, migration, and invasionand contributes to poor prognosis in patients. Proc. Natl. Acad. Sci. U.S.A. 107,18622–18627.

o, J.Y., Lee, H.J., Kramata, P., Minden, A., Suh, N., 2012. Differential expression ofkey signaling proteins in MCF10 cell lines, a human breast cancer progressionmodel. Mol. Cell. Pharmacol. 4, 31–40.

oosairajah, J., Maiti, S., Wiggan, O., Sarmiere, P., Moussi, N., Sarcevic, B., Sampath, R.,Bamburg, J.R., Bernard, O., 2005. Interplay between components of a novel LIMkinase-slingshot phosphatase complex regulates cofilin. EMBO J. 24, 473–486.

pratley, S.J., Bastea, L.I., Doppler, H., Mizuno, K., Storz, P., 2011. Protein kinaseD regulates cofilin activity through p21-activated kinase 4. J. Biol. Chem. 286,34254–34261.

tefanis, L., 2012. Alpha-synuclein in Parkinson’s disease. Cold Spring Harb. Perspect.Med. 2, a009399.

un, X., Liu, B., Wang, J., Li, J., Ji, W.Y., 2013. Inhibition of p21-activated kinase 4

expression suppresses the proliferation of Hep-2 laryngeal carcinoma cells viaactivation of the ATM/Chk1/2/p53 pathway. Int. J. Oncol. 42, 683–689.

abusa, H., Brooks, T., Massey, A.J., 2013. Knockdown of PAK4 or PAK1 inhibits theproliferation of mutant KRAS colon cancer cells independently of RAF/MEK/ERKand PI3K/AKT signaling. Mol. Cancer Res. 11, 109–121.

of Cell Biology 92 (2013) 129– 138

Tian, Y., Lei, L., Cammarano, M., Nekrasova, T., Minden, A., 2009. Essential rolefor the Pak4 protein kinase in extraembryonic tissue development and vesselformation. Mech. Dev. 126, 710–720.

Tian, Y., Lei, L., Minden, A., 2011. A key role for Pak4 in proliferation and differenti-ation of neural progenitor cells. Dev. Biol. 353, 206–216.

Viaud, J., Peterson, J.R., 2009. An allosteric kinase inhibitor binds the p21-activatedkinase autoregulatory domain covalently. Mol. Cancer Ther. 8, 2559–2565.

Wallace, S.W., Durgan, J., Jin, D., Hall, A., 2010. Cdc42 regulates apical junction for-mation in human bronchial epithelial cells through PAK4 and Par6B. Mol. Biol.Cell 21, 2996–3006.

Wang, J., Wu, J.W., Wang, Z.X., 2011. Structural insights into the autoactivationmechanism of p21-activated protein kinase. Structure 19, 1752–1761.

Wang, Z., Zhang, X., Yang, Z., Du, H., Wu, Z., Gong, J., Yan, J., Zheng, Q., 2012. MiR-145 regulates PAK4 via the MAPK pathway and exhibits an antitumor effect inhuman colon cells. Biochem. Biophys. Res. Commun. 427, 444–449.

Wells, C.M., Abo, A., Ridley, A.J., 2002. PAK4 is activated via PI3K in HGF-stimulatedepithelial cells. J. Cell Sci. 115, 3947–3956.

Wells, C.M., Whale, A.D., Parsons, M., Masters, J.R., Jones, G.E., 2010. PAK4: apluripotent kinase that regulates prostate cancer cell adhesion. J. Cell Sci. 123,1663–1673.

Whale, A.D., Dart, A., Holt, M., Jones, G.E., Wells, C.M., 2012. PAK4 kinase activityand somatic mutation promote carcinoma cell motility and influence inhibitorsensitivity. Oncogene, http://dx.doi.org/10.1038/onc.2012.233 [Epub ahead ofprint].

Wong, L.E., Reynolds, A.B., Dissanayaka, N.T., Minden, A., 2010. p120-catenin is abinding partner and substrate for Group B Pak kinases. J. Cell. Biochem. 110,1244–1254.

Yang, F., Li, X., Sharma, M., Zarnegar, M., Lim, B., Sun, Z., 2001. Androgen receptorspecifically interacts with a novel p21-activated kinase PAK6. J. Biol. Chem. 276,15345–15353.

Yang, L., Wang, L., Zheng, Y., 2006. Gene targeting of Cdc42 and Cdc42GAP affirmsthe critical involvement of Cdc42 in filopodia induction, directed migration,and proliferation in primary mouse embryonic fibroblasts. Mol. Biol. Cell 17,4675–4685.

Ye, D.Z., Jin, S., Zhuo, Y., Field, J., 2011. p21-Activated kinase 1 (Pak1) phosphorylatesBAD directly at serine 111 in vitro and indirectly through Raf-1 at serine 112.PLoS ONE 6, e27637.

Yu, W., Kanaan, Y., Bae, Y.K., Gabrielson, E., 2009. Chromosomal changes in aggres-sive breast cancers with basal-like features. Cancer Genet. Cytogenet. 193,29–37.

Zang, M., Hayne, C., Luo, Z., 2002. Interaction between active Pak1 and Raf-1 is neces-sary for phosphorylation and activation of Raf-1. J. Biol. Chem. 277, 4395–4405.

Zanivan, S., Meves, A., Behrendt, K., Schoof, E.M., Neilson, L.J., Cox, J., Tang, H.R., Kalna,G., van Ree, J.H., van Deursen, J.M., Trempus, C.S., Machesky, L.M., Linding, R.,Wickstrom, S.A., Fassler, R., Mann, M., 2013. In vivo SILAC-based proteomicsreveals phosphoproteome changes during mouse skin carcinogenesis. Cell Rep.3, 552–566.

Zenke, F.T., Krendel, M., DerMardirossian, C., King, C.C., Bohl, B.P., Bokoch, G.M., 2004.p21-activated kinase 1 phosphorylates and regulates 14-3-3 binding to GEF-H1,a microtubule-localized Rho exchange factor. J. Biol. Chem. 279, 18392–18400.

Zhang, H., Li, Z., Viklund, E.K., Stromblad, S., 2002. P21-activated kinase 4 interactswith integrin alpha v beta 5 and regulates alpha v beta 5-mediated cell migration.J. Cell Biol. 158, 1287–1297.

Zhang, H.J., Siu, M.K., Yeung, M.C., Jiang, L.L., Mak, V.C., Ngan, H.Y., Wong, O.G., Zhang,H.Q., Cheung, A.N., 2011. Overexpressed PAK4 promotes proliferation, migrationand invasion of choriocarcinoma. Carcinogenesis 32, 765–771.

Zhang, J., Wang, J., Guo, Q., Wang, Y., Zhou, Y., Peng, H., Cheng, M., Zhao, D., Li, F., 2012.LCH-7749944, a novel and potent p21-activated kinase 4 inhibitor, suppresses

proliferation and invasion in human gastric cancer cells. Cancer Lett. 317, 24–32.

Zhao, L., Ma, Q.L., Calon, F., Harris-White, M.E., Yang, F., Lim, G.P., Morihara, T.,Ubeda, O.J., Ambegaokar, S., Hansen, J.E., Weisbart, R.H., Teter, B., Frautschy, S.A.,Cole, G.M., 2006. Role of p21-activated kinase pathway defects in the cognitivedeficits of Alzheimer disease. Nat. Neurosci. 9, 234–242.