e707.Full Neeew to Read

7/26/2019 e707.Full Neeew to Read

1/16

p21-Activated protein kinases and their emerging roles in glucose homeostasis

Yu-ting Alex Chiang1 and Tianru Jin1,2

1Department of Physiology, University of Toronto, Toronto, Canada; 2Division of Advanced Therapeutics, University Health

Network, Toronto, Ontario, Canada

Submitted 10 September 2013; accepted in final form 20 December 2013

Chiang YA, Jin T. p21-Activated protein kinases and their emerging roles inglucose homeostasis.Am J Physiol Endocrinol Metab306: E707E722, 2014. Firstpublished December 24, 2013; doi:10.1152/ajpendo.00506.2013.p21-Activatedprotein kinases (PAKs) are centrally involved in a plethora of cellular processes andfunctions. Their function as effectors of small GTPases Rac1 and Cdc42 has beenextensively studied during the past two decades, particularly in the realms of cellproliferation, apoptosis, and hence tumorigenesis, as well as cytoskeletal remodel-ing and related cellular events in health and disease. In recent years, a large numberof studies have shed light onto the fundamental role of group I PAKs, most notablyPAK1, in metabolic homeostasis. In skeletal muscle, PAK1 was shown to mediatethe function of insulin on stimulating GLUT4 translocation and glucose uptake,while in pancreatic-cells, PAK1 participates in insulin granule localization andvesicle release. Furthermore, we demonstrated that PAK1 mediates the cross talk

between insulin and Wnt/-catenin signaling pathways and hence regulates gutproglucagon gene expression and the production of the incretin hormone glucagon-like peptide-1 (GLP-1). The utilization of chemical inhibitors of PAK and thecharacterization ofPak1/mice enabled us to gain mechanistic insights as well asto assess the overall contribution of PAKs in metabolic homeostasis. This reviewsummarizes our current understanding of PAKs, with an emphasis on the emergingroles of PAK1 in glucose homeostasis.

GLP-1; insulin; PAK1; Wnt

Overview of the p21-Activated Protein Kinases

P21-ACTIVATED PROTEIN KINASES (PAKs) are serine/threonine

kinases that execute a multitude of cellular and physiologicalfunctions. PAK1, the prototype of the family, was initially

discovered in 1994 as a novel effector of small GTPases (114).

PAKs have been extensively studied as part of numerous

intricate and interconnecting signaling networks that regulate

important cellular activities, including cell proliferation, apo-

ptosis, and cytoskeleton dynamics (15, 20, 59, 122). Both the

disarrangement of thePak1gene and the dysfunctions of PAK

signaling under pathophysiological conditions have been iden-

tified, particularly for their involvement in tumorigenesis and

cancer metastasis (44, 97). The potential role of PAK signaling

in the development and progression of neuronal diseases, viral

pathogenesis, and immunity, as well as vascular disorders,

have been suggested as well (54, 75, 86, 87, 96, 127, 133, 189).In addition to the aforementioned functions of PAKs, there is

emerging evidence implicating the importance of PAKs in meta-

bolic processes, specifically in glucose homeostasis. PAK1 was

found to mediate the effect of insulin on stimulating glucose

uptake in skeletal muscle cells, in facilitating insulin secretion in

pancreatic -cells, and in controlling the production of the

incretin hormone glucagon-like peptide-1 (GLP-1) in the gut

endocrine L cells (29, 74, 84, 181, 195, 196).

Here we summarize the functions of PAKs, mainly focusingon the emerging roles of PAK1 in glucose homeostasis. Wehave reviewed the studies to date in the exploration of the

molecular and cellular functions of PAK1 in the muscle,pancreatic -cells, and intestine endocrine L cells, whichcollectively point to an essential metabolic role of PAK1.

These studies have also deepened our functional understandingof the PAK family, and continued efforts in this field may leadto the discovery of novel drug targets for type 2 diabetes (T2D)and other metabolic disorders.

The discovery of PAKs, their structural features, and the

activation mechanism.Guanine nucleotide binding proteins (Gproteins) act as molecular switches that transmit a wide variety

of extracellular stimuli to intracellular signaling cascades. Themajority of the G proteins fall into two categories, namely themonomeric forms (small GTPases) and the heterotrimeric

membrane-bound complexes (Fig. 1A). Small GTPases areable to bind to and hydrolyze guanosine triphosphate (GTP).

The Ras superfamily of small GTPases comprises over 150members. It is further classified into eight families (Fig. 1A).

One of them is the Ras homologous (Rho) family, whosemembers are also referred to as p21 GTPases since theirmolecular sizes are 21 kDa. Within the Rho family, PAKshave been identified as downstream effectors of the membersof Rac and Cdc42 subfamilies but not for members of theRhoA subfamily (Fig. 1A).

PAK1 was discovered as a binding partner and target of Rac

and Cdc42 GTPases in the rat brain in 1994 and was originallynamed p65PAK (114). The autophosphorylation and kinase

Address for reprint requests and other correspondence: T. Jin, Rm. 10-354Toronto Medical Discovery Tower, 101 College St., Toronto, Ontario, CanadaM5G 1L7 (e-mail: [email protected]).

Am J Physiol Endocrinol Metab 306: E707E722, 2014.First published December 24, 2013; doi:10.1152/ajpendo.00506.2013. Review

0193-1849/14 Copyright 2014 the American Physiological Societyhttp://www.ajpendo.org E707
mailto:[email protected]:[email protected]


2/16

A

B

SubfamilyFamilySuperfamily

Ras Rho

Ras

Rab

Arf

Ran

RhoA

Rac

Cdc42

Rac1

Rac2

Rac3

RhoGRap

Rheb

Rit

GTPase

PAKs

C

AID KDPBD CN

Pak1

Pak2

Pak3

Pak4

Pak5

Pak6

Group II

Group I

N C

N C

N C

N C

N C

N C

PBD

KD

PXXP SH3-binding motifs

AID

ED-rich regions

PXP SH3-binding motifs

N CPAK1

P

Nck Grb2

Akt

P

p35/Cdk5

Cdc2

Erk1/2

P

Thr423

Pdk-1

SAD-A

Ser57

Ser204

P

Rac

Cdc42

CHP

TC10

Wrch-1

Pix/Cool

hPIP1

CRIPAK

P P P P

Ser223

CK2

Merlin

P

Thr109

LKB1

Src

P

Tyr131

MonomericG proteins

HeterotrimericG proteins

Auto-Inhibitory Domain Kinase Domain

P21-Binding Domain

SKIPNischarin

Thr212Ser144 Ser199Ser21

Fig. 1. p21-Activated protein kinases (PAKs) are effectors of selected small GTPases. A: majority of GTPases are categorized in monomeric and heterotrimeric forms.The Ras superfamily contains 8 families, including the Rho family. Within this family, members of the Rac and Cdc42 subfamilies act as upstream activators of PAKs.

B: common features of PAKs include the p21-binding domain (PBD, orange), the catalytic kinase domain (KD, green), and multiple PXXP SH3-binding motifs (brown).The group I PAK proteins share a common autoinhibitory domain (AID, black). Members of group I PAKs also contain acid-rich domains known as ED-rich regions(light blue) as well as a noncanonical PXP Pix/Cool SH3-binding motif (dark blue) in the central region. C: illustration of the PAK1 molecule. The small p21 GTPases,including Rac, Cdc42, CHP, TC10, and Wrch-1, bind to the PBD (orange) and stimulates its activation. Ser21, Ser57, Ser144, Ser149, Ser199, Ser204, and Thr423 are7 autophosphorylation sites. The adapter proteins Nck and Grb2 bind to 2 PXXP motifs (brown) near the NH2terminus. The noncanonical PXP motif (light blue) interacts withthe adapter protein Pix/Cool. The endogenous inhibitors of PAK1 include hPIP1, Merlin, LKB1, CRIPAK, and Nischarin, with their sites of interaction as indicated (red line).

Review

E708 PAK AND GLUCOSE HOMEOSTASIS

AJP-Endocrinol Metab doi:10.1152/ajpendo.00506.2013 www.ajpendo.org


3/16

activity of p65PAK were shown to be stimulated in the pres-ence of activated Cdc42 and Rac, and sequence analysis ofp65PAK revealed its homology with the yeast protein Ste20(114). p65PAK was later renamed as PAK1, while the othertwo members of the family were designated as PAK2 andPAK3 (15, 16, 118, 175). Together, these three kinases com-prise the group I PAKs (Fig. 1B). After the initial discovery of

PAK13, three additional PAKs, namely PAK4 6, were iden-tified in mammals (73). They differ significantly from group IPAKs in their structural organization and regulation and arereferred to as group II PAKs (Fig. 1B) (47, 73). PAK4 wasidentified as a novel link between Cdc42 and actin cytoskele-ton, where activated PAK4 can be translocated into theGolgi and induced fibroblast filopodia formation and actinpolymerization (1). PAK6 was identified as an androgenreceptor (AR) interacting protein, and was suggested toserve as a negative regulator of the androgen/AR signaling(204). PAK5 is a brain kinase that promotes neurite growthand filopodia formation (37).

All PAK members share the NH2-terminal p21 bindingdomain (PBD) and the highly conserved COOH-terminal cat-alytic kinase domain (KD) (Fig. 1B). A unique feature of thegroup I PAKs is the autoinhibitory domain (AID), whichoverlaps but is not coincident with the PBD. The AID acts asan inhibitory switch and is critical for the autoinhibition ofgroup I PAKs. Group I PAKs also possess proline-rich PXXPand PXP SH3-binding motifs interspersed at the NH2terminusand central regions. On the other hand, in the group II PAKs,these SH3-binding motifs are only found in the central region(Fig. 1B). In the PAK1 molecule, these proline-rich motifsinteract with the Rac1/Cdc42 activator Pix/Cool (PAK inter-active exchange partner/Cloned out of library), as well ascertain adapter proteins, such as Nck and Grb2 (21, 115, 141).

As the most extensively studied member of the PAK family,

the crystal structure of PAK1 in an autoinhibited conformation

was determined in 2000 (100). The currently accepted mech-anism of group I PAK activation is a multistage model, wherebinding of Cdc42 or Rac to the PBD domain disrupts PAK-PAK dimerization, leading to a series of conformationalchanges that destabilize the folded structure of the inhibitoryswitch and thereby inducing the dissociation of the two PAKmolecules (100). Once freed from each other, the individual

PAK molecule may undergo sequential autophosphorylationsteps leading to its full activation.

PAK1 contains seven autophosphorylation sites (Fig. 1C)(113). Shortly after the discovery of PAK1, the Thr423 residuelocated within the activation loop was identified to act as acritical determinant of PAK1 activation by p21 GTPases andthe lipid sphingosine (209). The T423A PAK1 mutant, whichis unable to be autophosphorylated at this residue, showedsubstantially reduced kinase activity following Cdc42 activa-tion (209). With the use of biochemical methods, PAK1 auto-phosphorylation at Thr423 was demonstrated to occur as anintermolecular event, where an active PAK1 can trans-auto-phosphorylate another PAK1 molecule (92). Since then, PAK1Thr423 phosphorylation, assessed by Western blotting usingcommercially available antibodies, has served as an indicatorof its activation in numerous studies (29, 45, 52, 111, 126, 135,164, 197, 199, 212).

Positive and negative regulators of PAK1. Multiple positiveand negative regulators of PAK1 have been identified (Fig. 2).Other than Rac1 and Cdc42, PAK1 can also be activated bycertain atypical Rho members, including CHP, TC10, andWrch-1 (3, 156, 162, 172) (Fig. 2). The CHP/PAK1/PIXsignaling has been shown to regulate cell adhesion duringzebrafish embryonic development (174). TC10 was found toactivate PAK1 in assessing insulin-stimulated prolactin geneexpression in the rat pituitary GH4 cells (162). Wrch-1 is aWnt-1-stimulated small GTPase, which stimulates PAK1

Thr423 phosphorylation in the COS-7 cell line (172).

PAK1

Hormones

Positive regulators Negative regulators

p21GTPases

Tumorigenicfactors

Growthfactors

Endogenousinhibitors

Rac1Cdc42

CHP

TC10

Wrch-1

Thrombin

EGF

PDGF

IGF-1

Estrogen

Insulin

SKIP

MerlinLKB1

CRIPAK

Nisharin

hPIP1

Fig. 2. Positive and negative regulators of PAK1. Small p21 GTPases, including Cdc42 and Rac1, as well as atypical ones CHP, TC10, and Wrch-1, can stimulatePAK1. PAK1 can also be activated by other upstream factors, including tumorigenic factors, growth factors, and hormones such as insulin. Several endogenousinhibitors of PAK1 have been identified, including the kinase LKB1, the phosphatase SKIP, and other factors such as Merlin, CRIPAK, Nischarin, and hPIP1.

Review

E709PAK AND GLUCOSE HOMEOSTASIS



4/16

In addition to these small GTPases, other proteins includingtumorigenic factors, growth factors, and hormones, have beendemonstrated to activate PAK1 (Fig. 2). Thrombin-stimulatedPAK1 activation was observed in vascular smooth musclecells, and expression of the kinase-dead PAK1 mutant wasshown to attenuate stress fiber formation and cell migration(193). The epidermal growth factor (EGF) stimulates PAK1

activation in breast cancer cells, and expression of a dominant-negative PAK1 was shown to attenuate EGF-induced cellmigration (205). PAK1 serves as a convergence point forplatelet-derived growth factor (PDGF) and lysophosphatidicacid in regulating collagen matrix contraction in human fibro-blasts (146). Overexpression of PDGF is known to stimulateErk activity in medulloblastoma cells, and this was found to bemediated through the PAK1-MEK-Erk signaling cascade(208). PAK1 was also shown to mediate the function ofestrogen in stimulating FKHR phosphorylation, leading toFKHR nuclear exclusion and the inhibition of cell apoptosis(119). Notably, the growth factor insulin-like growth factor-1(IGF-1), as well as the metabolic hormone insulin, were found toactivate PAK1 as well (Fig. 2). IGF-1 activates RUNX2, atranscription factor that promotes endothelial cell migration, in-vasion, and proliferation. PAK1 inhibition attenuates RUNX2DNA-binding (142). Furthermore, insulin has been demon-strated to stimulate PAK1 activity in the skeletal muscle and inintestinal cells, which are detailed below in Role of PAK1 inGlucose Transport in the Skeletal Muscle and PAK1 as aLinker for the Cross Talk Between Insulin and Wnt SignalingPathways.

Several endogenous negative regulators of PAK1 have alsobeen recognized (Fig. 1Cand Fig. 2). The skeletal muscle andkidney enriched inositol phosphatase (SKIP) binds to the KD

of PAK1 and inhibits its scaffolding activity (66). Merlin, en-coded by the Nf2 tumor-suppressor gene, inhibits PAK1 activa-tion by targeting its PBD (93). The serine threonine kinase LKB1phosphorylates PAK1 at Thr109, leading to its inactivation (39).The cysteine-rich inhibitor of PAK (CRIPAK), identified in ayeast two-hybrid screen, was shown to suppress PAK1-medi-ated estrogen receptor transactivation in breast cancer cells

(170). The integrin binding partner Nischarin inhibits PAK1activity and PAK1-mediated cell migration through its directinteraction with the PAK1 COOH-terminal domain (2). Thehuman G-like WD-repeat protein (hPIP1) binds to the PAK1NH2-terminal regulatory domain, which abolishes Cdc42/Rac-stimulated PAK1 activity (202).

The mechanisms underlying Pak1 transcriptional regulationremain largely unknown, although the increase in Pak1 genecopy number has been reported in certain tumor cells (22, 129).One study has demonstrated that the microRNA miR-7 binds tothe 3=-untranslated region of Pak1 mRNA and may repressPak1 expression in human carcinoma cells (145).

An overview of cellular functions of PAK1.Here, we presenta brief overview of the cellular functions of PAK1 (Fig. 3). Anextensive body of knowledge exists for the downstream effectsof PAK signaling, which have been described in detail else-where (9, 20, 26, 44, 59, 122). PAK1 exerts its prosurvival andantiapoptotic functions via a battery of downstream targets,including Raf1, NFB, FKHR, BimL, and DLC1 (17, 35, 48,51, 119, 188) (Fig. 3, column 1). PAK1 is centrally involved incytoskeleton remodeling, either by phosphorylating kinasesthat regulate actin dynamics, such as LIMK, or by phosphor-ylating factors and kinases that are involved in actin-based cellmotility, such as myosin light chain (MLC), myosin heavychain (MHC), and MLC kinase (MLCK) (24, 28, 33, 36, 144,

(1)

PAK1

(2) (3) (4) (5)

LIMK

Myo3p

RLC

MLC

MHC

MLCK

PP2A

Cytoskeletonremodeling

Nck

Erk

CtBP1

Immunityand infection

histone H3

FKHR

CtBP1

SHARP

Snail

Transcription andmRNA splicing

-cat

Aurora

PIK1

histone H3

MEK1

Raf1

cyclinD1

Cell cycleprogression

NFB

Raf1

FKHR

BimL

DLC1

Cell survivaland apoptosis

NFB

Fig. 3. A general overview of cellular functions of PAK1. PAK1 regulates the following: column 1: cell survival and apoptosis; column 2: cytoskeletonremodeling;column 3: cell cycle progression; column 4: immunity and infection; as well as column 5: gene transcription and mRNA splicing.

Review




5/16

152, 190, 199, 201, 210) (Fig. 3, column 2). In mast cells,PAK1 was shown to directly interact with and activate thephosphatase PP2A in regulating F-actin rearrangement andmast cell granulation (163) (Fig. 3, column 2). Mast cells fromPAK1/ mice showed abnormal cortical F-actin structuresand granules and impaired allergic responses (163). The effec-tors of PAK1 in regulating cell cycle progression are listed in

Fig. 3, column 3, including the mitotic regulators cyclin D1,NF-B, Aurora, and Polo-like 1 (PlK1) (17, 35, 117, 211), thechromatin protein component histone H3 (99), and the com-ponents of the mitogen-activated protein kinase cascade MEK1and Raf1 (18, 50, 91, 171, 180). The targets that mediate thefunction of PAK1 in regulating immune functions and host-pathogen responses are listed in Fig. 3, column 4. PAK1activates Nck and Erk during T-cell activation, and PAK1phosphorylates CtBP1 during adenovirus entry (6, 105, 149,203) (Fig. 3, column 4). Finally, PAK1 is increasingly recog-nized to participate in nuclear events in regulating gene tran-scription and mRNA splicing (Fig. 3,column 5). Three nuclearlocalization signals have been identified in the PAK1 NH2terminus, which are required for its nuclear translocation (158).In the nucleus, PAK1 modulates transcriptional regulators,such as FKHR, CtBP, SHARP, and Snail, and directly regu-lates chromatin rearrangement through its interaction withhistone H3 (101). Notably, PAK1 has been shown to enhancethe transcriptional activity of the nuclear Wnt effector -cat incarcinoma and noncarcinoma cells (112, 134).

Loss-of-function and gain-of-function studies of PAK1 andother PAKs.The association between PAK activity and tumor-igenesis or metastasis has been intensively investigated (26, 90,97, 130). The underlying mechanisms include increased Pakgene copy numbers, hyperactivation of the kinase activity,dysregulation of upstream regulators, or enhanced nuclearlocalization of PAKs (8, 22, 60, 90, 116, 151, 187). Many

studies have been conducted using loss-of-function ap-proaches, such as dominant-negative forms of PAK1, RNAinterference, or chemical inhibitors (27, 49, 108, 134, 164).During the recent years, several small molecule inhibitors ofPAKs have been identified with varying degrees of specificity(38, 90, 103, 106). Among these, the chemical inhibitor 2,2=-dihydroxy-1,1=-dinaphthyldisulfide (IPA3) has demonstratedhigh specificity towards the inhibition of group I PAKs (38).

Gain-of-function approaches have also been utilized in anumber of studies. The expression of various forms of consti-tutively active PAK1 resulted in the formation of larger lamel-lipodia and increased mobility of fibroblasts (154). The T423Econstitutively active PAK1 mutant has been used in studyingcell proliferation (134, 153, 177), tumorigenesis (27, 72, 108),

and viral pathogenesis (82, 136).Pak1/ mice are viable, with various degrees of defects in

their immune, neuronal, pulmonary, and cardiac systems (5,13, 61, 88, 110, 120, 159, 169). In the absence of a challenge,fasting plasma glucose levels inPak1/mice are comparablewith those of the control littermates. However, in the presenceof a glucose or pyruvate challenge, impaired glucose disposalwas observed in thePak1/mice (29, 195).Pak2/mice areembryonic lethal at embryonicday 8.5(9, 59). The PAK3genein humans is associated with X-linked nonsyndromic mentalretardation (9, 59), and the Pak3/ mice show deficits inmemory retention (121).Pak1andPak3double knockout miceshow severely reduced postnatal brain growth, associated with

other neuronal deficits (63). Pak4/ mice are embryoniclethal at embryonic day 11, while Pak4/ embryos werefound to carry defects in neuronal differentiation and bloodvessel formation (143). Conditional knockout ofPak4 specif-ically in the nervous system resulted in neuronal defects, alongwith striking differences in brain anatomy in neonatal mice(178). Lastly, both thePak5/mice and thePak6/mice are

viable, and the gross morphology of various organs appears tobe normal (104, 124). It has been suggested that this may bedue to the potential redundant functions of PAK5 and PAK6(104). A comprehensive review on Pak knockout mouse mod-els has been presented recently by Kelly and Chernoff (88).

Role of PAK1 in Glucose Transport in the Skeletal Muscle

In 1996, Tsakiridis et al. (181) revealed the presence ofinsulin-responsive renaturable kinases in the mouse differenti-ated L6 muscle cells. Using histone VI-S as the substrate,Tsakiridis et al. (181) found that insulin is able to stimulate theactivity of PAK1. Utilizing chemical inhibitors for various cellsignaling molecules, Tsakiridis et al. (181) concluded that

insulin-stimulated PAK1 activation is dependent on the ty-rosine kinase activity and the activity of phosphoinositide3-kinase (PI3K).

Insulin-stimulated glucose uptake by skeletal muscle cells isa key process in maintaining glucose homeostasis, and this isaccomplished by the recruitment of GLUT4 vesicles to theplasma membrane. Actin cytoskeleton reorganization is in-volved in insulin-stimulated GLUT4 vesicle translocation, aspharmacological agents that disrupt actin dynamics, such aslatrunculin B, inhibited GLUT4 translocation to the plasmamembrane and led to reduced glucose uptake in the musclecells (23, 31, 166, 183). Insulin is known to regulate actindynamics in cultured muscle cells and in the mouse L6 myo-

tubes (179, 182). The binding of insulin to the insulin receptorresults in the phosphorylation of the insulin receptor sub-strate-1 (IRS-1), which in turn recruits and activates PI3K. Theinsulin signaling cascade bifurcates at PI3K into two armsconsisting of Rac1 and Akt (150). Insulin-induced actin cyto-skeleton reorganization in muscle cells requires PI3K activity(94, 183, 184), and insulin-stimulated GLUT4 translocationand actin remodeling can be blocked either by depleting Rac1or by inducing conditions that mimic insulin resistance incultured muscle cells (74, 185). Expression of a constitutivelyactive Rac1 elicited GLUT4 vesicle translocation even in theabsence of insulin, suggesting that the presence of active Rac1suffices in stimulating GLUT4 translocation, at least in the invitro setting (185). Treating cultured muscle cells with glucose

oxidase mimics insulin resistance in vitro, and this leads to theinhibition of Rac1 activation, accompanied by the loss ofRac1-mediated PAK1 Thr423 phosphorylation (74). Theseobservations highlighted the requirement of Rac1 signaling ininsulin-stimulated muscle glucose uptake, involving the pro-cess of cytoskeleton remodeling. Consistent with the observa-tions made in the above in vitro investigations, insulin wasdemonstrated to stimulate Rac1 activation in mouse skeletalmuscle, and mice with muscle-specific deletion of Rac1 showreduced insulin-induced GLUT4 vesicle translocation in themuscle tissue (186).

Another small GTPase, namely RalA, has been reported tofunction as the effector of Rac1 in mediating its stimulatory

Review




6/16

effect on muscle glucose uptake. Depletion of RalA using thesmall interfering RNA-mediated knockdown approach wasshown to reduce Rac1-stimulated GLUT4 vesicle plasmamembrane translocation in the mouse L6 myotubes (128).

Observations described above indicated the essential role ofRac1 in mediating the stimulatory effect of insulin on GLUT4membrane translocation and that PAK1 is among the down-

stream targets of insulin and Rac1 in the muscle. The genera-tion of Pak1/ mice made the direct assessment of thecontribution of PAK1 in muscle glucose uptake available.Following insulin injection, muscle samples collected fromPak1/ mice showed significantly reduced GLUT4 vesicletranslocation to the plasma membrane, when compared withthe muscle samples from age-matched wild-type littermates(195). Cofilin has been reported to act downstream of thePAK1 effector LIMK and is critically involved in the regula-tion of actin filamentsin a marsupial kidney epithelial cell line(40). In line with the role of PAK1/LIMK/cofilin signaling incytoskeletal reorganization in epithelial cells, the skeletal mus-cle from Pak1/ mice was shown to lack the response toinsulin treatment in stimulating cofilin phosphorylation (195).

A very recent study by Sylow et al. (166) provided extendedevidence that Rac1-PAK1 signaling centrally regulates muscleglucose uptake in intact muscle tissue. More importantly, thissignaling axis was rendered defective under the insulin-resis-tant state (166). First, they demonstrated that insulin indeedstimulates Rac1 and PAK1 activation in both mouse andhuman skeletal muscle (166). Pharmacological inhibition ofRac1 resulted in the loss of insulin-stimulated glucose uptake,independent of Akt status, in isolated mouse skeletal muscle(166). Muscle-specific deletion of Rac1 in mice resulted inreduced insulin-stimulated muscle glucose uptake, associatedwith attenuated insulin-stimulated PAK1 Thr423 phosphoryla-tion in the muscle (166). Importantly, skeletal muscle isolated

from high-fat diet-fed mice showed reduced Rac1 expression,associated with blunted PAK1 Thr423 phosphorylation inresponse to insulin treatment (166). These observations in theexperimental animals were at least partially recapitulated inhumans, where skeletal muscle insulin resistance induced viaintralipid infusion resulted in a loss of insulin-stimulated PAK1Thr423 phosphorylation during a hyperinsulinemic-euglyce-mic clamp study (166).

Similar to PAKs, the Rho-associated coiled-coil-containingprotein kinases (ROCKs) are serine/threonine kinases that alsoact as downstream effectors of certain small GTPases, andROCKs have been shown to participate in insulin signaling andenergy homeostasis (62). Inhibition of ROCK1 impairedGLUT4 translocation, while ROCK1 overexpression lead to an

enhanced insulin-sensitizing effect on glucose transport inadipocytes (34). In the myoblasts, ROCK1 depletion lead toattenuated insulin-stimulated glucose transport, associated withdefects in actin cytoskeleton remodeling (34). Taken together,these observations indicate the involvement of additional smallGTPase effectors, such as ROCKs, in GLUT4 translocationand glucose transport in the muscle (34, 62, 74).

In addition to insulin activation, muscle contraction repre-sents another mechanism by which muscle glucose uptake canbe stimulated (55). The activity of Rac1 was observed to beelevated during contraction-stimulated glucose uptake in skel-etal muscle, and this was concomitant with enhanced PAK1Thr423 phosphorylation (167). PAK activation and muscle

glucose uptake following contraction were attenuated by eitherRac1 inhibition or the use of F-actin depolymerizing agents invitro, as well as in muscle-specific Rac1/ mice (166, 167).Furthermore, PAK1 Thr423 phosphorylation was enhanced inmouse and human skeletal muscle following exercise, whereaslipid infusion in healthy volunteers impaired PAK1 Thr423phosphorylation. These observations collectively suggest that

Rac/PAK1 signaling plays a role in muscle glucose disposalduring exercise (166, 167).

As aforementioned, Akt2 is another arm of insulin signalingafter its bifurcation at PI3K (150). In 2002, a 160-kDa substrateof Akt was identified in 3T3L1 adipocytes, known as AS160(85). It is a protein that contains a Rab GTPase-activatingprotein (GAP) domain. Upon insulin stimulation, Akt2 trans-mits the signal to AS160, which in turn activates the smallGTPases Rab8A and Rab13, and ultimately induce GLUT4vesicle release (71, 165). Both the Rac1 arm and the Akt2 armupregulate glucose uptake in the skeletal muscle, and it hasbeen proposed that each arm act independently of each other inpromoting GLUT4 vesicle translocation (150, 168). Further-more, the joint requirement of both arms in insulin-stimulatedGLUT4 translocation has been suggested, where pharmacolog-ical inhibition of both Akt2 and Rac1 signaling cascades wasshown to be required to completely block insulin-stimulatedglucose uptake in mouse muscle cells (168). In addition, bothAkt2 and Rac1 signaling pathways were found to be dysfunc-tional in the muscle tissue of the insulin-resistant ob/obmousemodel (168). Interestingly, in examining the involvement ofPAK1 as the effector of Rac1 signaling, muscle tissue isolatedfrom Akt2/ mice appear to have blunted insulin-stimulatedPAK1 Thr423 phosphorylation, while pharmacological inhibi-tion of Akt2 does not alter PAK1 activation (168). To explainthis obvious discrepancy, it has been proposed that althoughthe Rac1 and Akt2 arms do not cross talk with each other in

regulating glucose uptake, the presence of Akt2 is still requiredfor the activation of the Rac1 cascade (168).

The essential role of Rac1 in regulating actin dynamicsunderlying muscle glucose uptake has been demonstrated innumerous investigations (31, 74, 150, 185). Although PAK1has been shown to regulate actin filament dynamics in anumber of other cell types, including neuronal cells and pan-creatic -cells (detailed in the sections below) (57, 98, 140,163), the direct effect of PAK1 on actin remodeling in musclecells remains to be examined.

Role of PAK1 in Insulin Secretion

Following food consumption or glucose stimulation, insulin

is released from pancreatic -cells in two distinct phases,known as biphasic insulin secretion. The first phase is triggeredby the elevation of intracellular calcium levels, leading to thefusion of predocked insulin-containing granules to the plasmamembrane. The second phase involves the mobilization ofstored granule pools to the cell surface, which represents thesustained insulin release following glucose stimulation.

As demonstrated by Wang et al. (196), Cdc42/Rac1 signal-ing is evidently involved in the insulin secretion process, asCdc42 knockdown selectively attenuated the second phase ofglucose-stimulated insulin secretion (GSIS) in rodent pancre-atic -cell lines and in isolated mouse and human pancreaticislets. In the mouse pancreatic -cell line MIN6, Cdc42 acti-

Review




7/16

vation was detected in response to D-glucose treatment but notto potassium chloride (KCl) or the nonmetabolizable glucoseanalogs (196). Furthermore, Cdc42 activation in response tohigh glucose treatment preceded Rac1 activation at the plasmamembrane, and PAK1 Thr423 phosphorylation was observedto be Cdc42-dependent but Rac1 independent (196). Knock-down of PAK1 expression in the mouse -cell line using the

small interference RNA knockdown approach abolishedCdc42-mediated Rac1 activation (196). These observationscollectively implicate the existence of the Cdc42-PAK1-Rac1axis during the second phase of GSIS in the pancreas (196),which differs from the Cdc42-Rac1-PAK1 signaling observedin the skeletal muscle cells in terms of the sequential order ofactivation of the signaling components (74).

To further assess the role of PAK1 in insulin secretion andto explore its pathophysiological implications, Wang et al.(195) conducted further investigations. They demonstrated thatPAK1 protein levels are reduced in islets of T2D patients(195). In human islets and a rodent pancreatic -cell line, IPA3pretreatment was shown to elicit a defect in the second phaseof GSIS (195). Wang et al. (195) then revealed that Pak1/

islets exhibited attenuated second phase of GSIS, associatedwith impaired glucose and insulin tolerance in the Pak1/

mice. To identify downstream effectors of PAK1 during GSIS,Wang et al. (195) performed PAK1 small interference RNAknockdown in a pancreatic -cell line and observed an abol-ished glucose-induced Erk activation. This event was alsopartially recapitulated in pancreatic islets isolated from thePak1/ mice (195, 196).

Although PAK1 has been widely accepted as a positiveregulator of cell proliferation, the Pak1/ mice lack notabledefects in pancreatic -cell mass and abnormalities in isletmorphology, as observed by both Wang et al. (195) and by ourgroup (29).

The importance of F-actin remodeling in GSIS has beenrecognized for40 yr, where F-actin filaments were shown toact as barriers at the plasma membrane to prevent the dockingand fusion of insulin granules under unstimulated conditions(14, 131, 191). Rac1 and a number of other small GTPaseshave been implicated in various functions of pancreatic -cells,including the secretion of insulin at different steps (95). In amurine -cell line, Rac1 activation was shown to be stimulatedby high glucose treatment (102). The expression of a domi-nant-negative Rac1 in -cells abolished its cytosol-to-mem-brane translocation upon high glucose stimulation, which wasassociated with reduced GSIS (102). Very recently, Asahara etal. (12) demonstrated the role of Rac1 in regulating GSIS viamodulating F-actin using both in vitro and in vivo approaches.

In the pancreatic -cell line MIN-6, Rac1 activation wasobserved following high glucose treatment (12). Rac1 knock-down leads to the persistence of intact F-actin in the presenceof high glucose, associated with reduced insulin secretionfollowing glucose stimulation (12). The study utilizing isletsisolated from -cell-specific Rac1/ mice also recapitulatedthe observations of defective insulin secretion following highglucose treatment, without significant alterations in islet mor-phology (12).

Although PAK1 is evidently involved in the second phase ofGSIS, the exact mechanism by which PAK1 evokes F-actinremodeling in response to glucose is still not completelyunderstood. An early study identified the role of the Cdc42-

PAK1-Rac1 axis in regulating F-actin rearrangement and in-sulin secretion in -cells (196). It was further demonstratedthat glucose-induced cortical F-actin remodeling also in-volves the Cdc42-PAK1-MEK-Erk pathway and that glu-cose-induced insulin granule exocytosis can be blocked bythe PAK inhibitor IPA3 (83, 84). The downstream effects ofthe PAK1-MEK-Erk signaling cascade may include MLCK,

an actin-associated kinase that has been implicated in -cellinsulin secretion (11, 83).

In a recent live cell imaging study, Kalwat et al. (84) havefurther elucidated the mechanistic requirement of PAK1 inglucose-stimulated insulin granule exocytosis. Pretreatment ofmouse pancreatic-cells with IPA3 inhibited glucose-inducedF-actin remodeling and ablated glucose-stimulated insulingranule localization to the plasma membrane (84). In mouse-cells, glucose treatment stimulated Raf1 phosphorylationand activation, while IPA3 treatment or Cdc42 inhibitionabolished this activation (84). Similarly, IPA3 treatmentblunted MEK to Erk signaling, while the presence of a MEK-Erk inhibitor disrupted F-actin remodeling and resulted indefective insulin granule release (84). Altogether, these in vitrofindings implicate the functional requirement of PAK1 inrelaying the signal of Cdc42 to the downstream Raf1-MEK-Erk cascade, which potentiates insulin granule mobilizationthrough F-actin remodeling (84).

Despite the substantial evidence supporting the notion thatCdc42-PAK1 signaling participates in insulin granule mobili-zation, the effect is still needed to further clarify how Cdc42activation, in response to high glucose treatment, leads toPAK1 activation in pancreatic -cells The synapses of amphidsdefective (SAD-A) kinase is a serine/threonine kinase closelyrelated to the AMP-activated kinases (4). SAD-A is exclu-sively expressed in the brain and pancreas and has beenimplicated in synaptic function and neuronal development

(68). In examining the role of SAD-A in insulin secretion, Nieet al. (126) first showed that treatment with high glucoseinduces SAD-A activation and SAD-A can stimulate PAK1Thr423 phosphorylation and PAK1 kinase activity in a murinepancreatic-cell line. SAD-A was also shown to interact withrecombinant GST-PBD of PAK, which can be completelyabolished by a point mutation that inactivates the SAD-Akinase activity (126). Overexpression of SAD-A led to en-hanced cortical F-actin formation, while expression of a ki-nase-dead SAD-A K48M mutant induced cortical actin fila-ment disintegration (126). Importantly, short hairpin RNA-mediated SAD-A knockdown attenuated GSIS in the mouse-cell line MIN-6, and adenoviral expression of either K48MSAD-A or the dominant-negative PAK1 mutant K299R was

able to significantly reduce GSIS in mouse islets (126). Takentogether, SAD-A is at least among the direct upstream activa-tors of PAK1 in regulating insulin granule exocytosis in-cells. SAD-A has been postulated as the downstream effectof PKA and calmodulin-dependent protein kinase kinase(CaMMK) (126), while the relationship between SAD-A andRac1 remains to be investigated.

PAK1 as a Linker for the Cross Talk Between Insulin andWnt Signaling Pathways

Although it is still under debate after intensive investigationsfor a number of years, certain epidemiological studies have

Review




8/16

implicated an association among obesity, metabolic syndrome,and the development of various types of tumors, includingthose in the mammary gland and the large intestine (7, 19, 157,192, 200). It has been postulated that the molecular basis forthis association is the aberrant insulin signaling, resulting in thestimulation of cell proliferation and tumorigenesis in the con-text of T2D (43, 109, 164, 198). The canonical Wnt signaling

pathway positively regulates cell proliferation, and its dysregu-lated activation is strongly associated with tumors of variousorigins. The clinical and epidemiological correlations betweenT2D and the cancer risk have been summarized in numerousexcellent review articles (42, 46, 69, 70, 77, 79, 81, 132, 160).

Activation of the canonical Wnt signaling pathway occursupon the binding of a Wnt ligand to the Frizzled receptor andlow-density lipoprotein receptor-related protein 5 and 6 (LRP5/6), the coreceptor of the Wnt ligands. This triggers a series ofsignaling events, ultimately leading to the nuclear translocationof free-catenin (-cat) (78). Nuclear -cat will subsequentlyinteract with a member of the TCF transcription factor family,forming the bipartite transcription factor -cat/TCF. Amongthe four known TCF members, TCF7L2 has been recognizedas an important diabetes risk gene(56). A number of pro-proliferative genes, such as c-Myc and cyclin D1, are knowndownstream targets of-cat/TCF. In addition, our group hasdemonstrated that -cat/TCF positively regulates the expres-sion of the proglucagon gene (gcg) in the gut endocrine L cells.PAK1 has been shown to serve as a downstream effector ofinsulin and IGF-1 (89, 142, 147, 181). In both breast cancerand colon cancer patients, increased PAK1 copy number hasbeen reported (22, 129). In addition, the upregulation of PAKactivity has been associated with the progression and metasta-sis of various types of tumors (26, 90, 97, 130). It is thereforeplausible to propose that PAKs may act as the linker betweeninsulin/IGF-1 and Wnt signaling pathways. In supporting this

hypothesis, a battery of in vitro investigations conducted byour group and others have shown that PAK1 mediates thestimulatory effect of insulin on the proliferation of intestinaltumorous and nontumorous cells, involving the -cat/TCFtranscriptional activity (29, 89, 147, 164).

Both insulin and IGF-1 were shown to stimulate PAK1Thr423 phosphorylation in different cell lineages (29, 147,164, 181), and we have demonstrated that intraperitonealinjection of insulin in the FVB mice resulted in insulin-stimulated PAK1 Thr423 phosphorylation in the liver, fat,heart, as well as small and large intestines (164). In theintestinal carcinoma cell lines, expression of dominant-nega-tive K299R PAK1 attenuated insulin-stimulated c-Myc andcyclin D1 protein expression, while expression of constitu-

tively active T423E mutant PAK1 increased c-Myc and cyclinD1 protein levels following insulin stimulation (164). Theexpression of K299R dominant-negative PAK1 also led toreduced insulin-stimulated binding of-cat/TCF to the humanc-Myc gene promoter (164). Furthermore, PAK1 knockdownresulted in a drastic reduction of c-Myc and cyclin D1 expres-sion, accompanied by attenuated Erk1/2 phosphorylation in thepresence and absence of insulin stimulation (164). Finally,insulin treatment was shown to increase-cat phosphorylationat Ser675, an event that is positively correlated with -catnuclear translocation and -cat/TCF transcriptional activity(58, 164, 173). In contrast, PAK1 knockdown attenuated insu-lin-stimulated -cat Ser675 phosphorylation, associated with

reduced binding of -cat to the c-Myc gene promoter, andrepressed c-Myc gene expression (164). Recently, in vitrobiochemical studies have demonstrated a direct interactionbetween PAK1 and -cat, where PAK1 was shown to stimulate-cat Ser675 phosphorylation in breast cancer and colon can-cer cells (10, 213). These findings collectively implicate thatmechanistically PAK1 acts as a linker between insulin and Wnt

signaling pathways.

Role of PAK1 in Proglucagon Gene Expression, GLP-1Production, and Secretion

The role of Wnt signaling pathway in metabolic homeostasishas captured our attention during the last decade (30, 79).Hence, we have been exploring the metabolic effect of thecross talk between Wnt and insulin signaling pathways (125,155, 206, 207).

The proglucagon gene (gcg) is expressed in pancreatic cells, intestinal endocrine L cells, and certain neuronal cells inthe brainstem. Thegcg gene encodes several important peptidehormones, including the incretin hormone GLP-1, which is

mainly produced in the gut and brain but not in the pancreas(25). Lithium treatment, which mimics Wnt ligand activation(161), was shown to stimulate gcg transcription and GLP-1synthesis in the intestinal L cells, while the expression of adominant-negative TCF7L2 abolished lithium-induced gcgtranscription and GLP-1 production. More recently, we foundthat the Wnt ligand Wnt3a can also stimulate gcg promoteractivity and increasegcg mRNA levels in the mouse intestinalendocrine L cell line GLUTag (29). Although insulin wasshown to repress gcg expression in pancreatic -cells (137,138), it activatesgcg transcription in the gut GLP-1-producingcells, and this activation was found to be PI3K dependent butAkt independent (207). Furthermore, insulin treatment was

shown to stimulategcg promoter activity, and this stimulationdepends on the TCF binding motif within the G2 enhancerelement. The G2 enhancer element is known to mediate-cat/TCF- or Wnt ligand-stimulated gcg promoter activity. Indeed,insulin treatment enhanced binding of-cat/TCF7L2 to the G2enhancer element (207). Thus the cross talk between insulinsignaling and the Wnt signaling pathway effectors -cat/TCF7L2 represents a novel mechanism underlying the tran-scriptional regulation of gut gcg expression and GLP-1 pro-duction (76).

Considering that PAK1 functions as a mediator for the crosstalk between insulin and Wnt signaling pathways in the gutendocrine and nonendocrine cells as well as elsewhere (147,164), we asked whether PAK1 plays a role in regulating gut

and braingcg expression and GLP-1 production. We first of allverified that insulin treatment indeed stimulates PAK1 Thr423phosphorylation in both the mouse intestinal endocrine L-cellline GLUTag and a hypothalamic gcg-expressing neuronal cellline (29). We then found that insulin treatment also stimulated-cat Ser675 phosphorylation in these two cell lines (29). Inthe GLUTag cell line, IPA3 was shown to block the stimula-tion of insulin on -cat Ser675 phosphorylation, associatedwith attenuated gcg promoter activity and gcg mRNA expres-sion following insulin stimulation (29). We then expanded theinvestigation to investigate the physiological implications ofPAK1 in GLP-1 production and function utilizing the Pak1/

mice (29). Pak1/ mice exhibited impaired tolerance to

Review




9/16

glucose as well as pyruvate challenge (29). Furthermore,Pak1/ mice show drastically lower levels of intestinal andbraingcg mRNA, associated with reduced distal ileum GLP-1content, and decreased circulating insulin and GLP-1 levelsfollowing oral glucose challenge (29). These in vivo observa-tions demonstrated that PAK1 ablation leads to the impairmentof glucose homeostasis, as well as defective insulin and GLP-1

responses following glucose challenge (29). As PAK1 regu-lates gutgcg expression, the reduction in gutgcg expression inthe Pak1/ mice likely contributes, at least partially, to theoverall glucose intolerance (29).

In the intestinal endocrine L cells, PAK1 was also verified tobe the target of Cdc42, where Cdc42 and PAK are required forinsulin-induced actin remodeling (107). In the mouse gutGLUTag cell line, chemical reagents that disrupt filamentousactin potentiated GLP-1 secretion, whereas the knockdown ofeither Cdc42 or PAK1 attenuated actin remodeling and GLP-1secretion (107). Hence, activation of the Cdc42/PAK1 signal-ing cascade and the consequent actin remodeling events alsopositively regulate GLP-1 secretion, at least in the in vitrosetting (107).

Summary, Disputes, and Perspectives

Originally identified as effectors of selected small GTPasesincluding Cdc42 and Rac1, the PAK family of protein kinaseshave been shown to carry out a plethora of functions in variouscell lineages, including cell proliferation and apoptosis, cyto-skeleton remodeling, cell cycle progression, and gene tran-scription. The exploration of these general cellular functions ofPAKs, especially its prototype PAK1, has been further ex-panded into the field of metabolic research, revealing thefundamental role of PAK1 in glucose homeostasis duringhealth and disease. PAK1 is expressed in multiple metabolic

tissues, and it evidently functions as an important metabolicregulator. A schematic overview of the current knowledge onthe role of PAK1 in glucose homeostasis is presented in Fig. 4,focusing on its role in glucose uptake in the skeletal muscle, ininsulin secretion in the pancreatic islets, and in the productionand function of the incretin hormone GLP-1 in the intestine.

Upon food intake, the rise in plasma glucose levels leads tothe postprandial release of insulin. The insulin/IRS-1/PI3Ksignaling not only activates Akt2, which utilizes AS160 andRab proteins to stimulate glucose uptake (176), but also acti-vates the Cdc42-Rac1-PAK1 signaling axis in regulatingGLUT4 vesicle translocation (Fig. 4, left). Actin remodeling isa key process underlying GLUT4 vesicle translocation, whichprovides the mechanistic basis for glucose uptake. The Rac1/

PAK1 signaling plays a fundamental role in insulin-stimulatedGLUT4 vesicle mobilization, and deficiency of either Rac1 orPAK1 leads to impaired glucose uptake in response to insulinstimulation. However, it remains unclear as to whether PAK1,or other downstream effectors of Rac1, or both, mediate theactin remodeling events in the skeletal muscle cells. In addi-tion, the relationship between the contributions of PI3K-Akt2signaling vs. PI3K-Rac1-PAK1 signaling in muscle glucoseuptake in response to insulin or exercise needs to be furtherexplored.

It should be pointed out that, although the majority of studiesto date have focused on assessing PAK1 in the muscle as aprotein kinase, one laboratory has recently suggested a poten-

tial novel mechanism for PAK1 in regulating glucose uptake(66). The inositiol phosphatase SKIP has been shown to inhibitinsulin/PI3K/Akt signaling and repress insulin-stimulatedGLUT4 translocation as well as glucose uptake in the mouseL6 myoblasts (64, 65, 67). It has been postulated that theunderlying mechanism involves the binding of SKIP to PAK1,which blocks PAK1 scaffolding function but not its kinase

activity in mouse myoblasts (66). Hence, the potential dualfunction of PAK1 in insulin-stimulated muscle glucose uptakeis worth furtherer exploration.

In pancreatic -cells, the elevation of plasma glucose stim-ulates postprandial insulin release, and the underlying mecha-nisms include the activation of the Cdc42-Rac1-PAK1 signal-ing cascade in the-cells, and the secretion of the gut incretinhormones, including GLP-1 (Fig. 4, rightand middle). Deple-tion or knockdown of Cdc42 or PAK1 abolished the secondphase of GSIS in the rodent pancreatic-cell line as well as inmouse and human islets (195, 196). In line with these in vitrofindings,Pak1/ mice were shown to exhibit reduced plasmainsulin levels following glucose challenge (29, 195).

As aforementioned, the role of PAK1 as a scaffold has beendemonstrated in muscle cell glucose uptake (66). A very recentinvestigation further showed that overexpression of a kinase-dead PAK1 in Hela cells and in the colon cancer cell lines ledto increased MEK and Erk phosphorylation, without affectingthe phosphorylation status of PAK1 downstream targets (194).In pancreatic -cells, whether the scaffolding function ofPAK1 participates in regulating insulin secretion remains anopen question.

It should be noted that the hierarchical relationship betweenPAK1 and the small GTPases could be cell type or tissuespecific. Both the Cdc42-PAK1-Rac1 axis and the Cdc42-PAK1-Raf1-MEK-Erk axis have been demonstrated to partic-ipate in the regulation of GSIS in pancreatic -cells (196). In

the skeletal muscle cells, as presented in Fig. 4, left, PAK1 wasshown to act downstream of both Cdc42 and Rac1 in regulatinginsulin-stimulated glucose uptake (74). In addition to the vary-ing sequence of activation events, the functions of Cdc42 andRac1 have been found to vary in other cell lineages. Cdc42 andRac1 have been demonstrated to act in antagonistic manners, inthe generation of reactive oxygen species (41), and in theregulation of cell proliferation (32, 123).

Finally, in the gut endocrine L cells, PAK1 was shown toregulate both the production and secretion of GLP-1 (29, 107)(Fig. 4, middle). PAK1-deficient mice express reduced levelsofgcg mRNA in their gut and brain (29). We have revealedthat PAK1 links insulin with the Wnt signaling pathwayeffector -cat, leading to the activation ofgcg expression and

GLP-1 production in the gut GLP-1-producing cell lineGLUTag (29).

As GLP-1 functions as an incretin, it is difficult to evaluatethe deleterious effect of PAK1 deficiency in the gut endocrineL cells vs. PAK1 deficiency in the pancreatic -cells on insulinsecretion in the Pak1/ mice. To properly dissect the meta-bolic function of PAK1 in the gut endocrine L cells vs. thepancreatic -cells, it will be necessary to generate cell typespecific knockout animal models.

In addition to the skeletal muscle, pancreatic islets, andintestines, PAK1 is also expressed in tissues including thebrain, liver, heart, and adipocytes, all of which are fundamen-tally involved in metabolic and glucose homeostasis (29).

Review




10/16

Whether PAK1 in those tissues is also involved in regulating

glucose homeostasis under physiological and pathophysiolog-ical conditions is an interesting direction for future research.

The subtle structural differences among the three membersof group I PAKs (Fig. 1B) suggest that they may exert bothunique and overlapping functions. It is notable that the meta-bolic defects observed in thePak1/mice are not drastic. Wehave noticed that the impairment in whole body glucosedisposal was absent when the Pak1/mice are in the C57BL/6129 hybrid background (29). Even after the Pak1/ micehave been backcrossed to the C57BL/6 genetic background, inthe absence of glucose challenge, the basal blood glucoselevels were comparable between the Pak1/ mice and thecontrol littermates, as demonstrated by our group and by Wang

and colleagues (29, 195). It is reasonable to speculate that

certain metabolic functions of PAK1 in the Pak1

/

mice arecompensated by the other two group I PAKs. Indeed, redun-dant functions of PAK1 and PAK3 have been observed in thebrain neurons (53, 63). Hence, it will be imperative to exploreand define the unique contributions of each of the three groupI PAKs, as well as their redundant functions, with continued invitro and in vivo investigations.

Mechanistic exploration of the involvement of PAK1 inmetabolic homeostasis represents an expansion of our funda-mental knowledge of the general cellular functions of PAKproteins. The investigations conducted exploring their role incytoskeleton remodeling led us to appreciate the role of PAK1in muscle cell glucose uptake and in -cell insulin secretion.

PAK1

-cat

Gcgexpression

Bloodstream

L cells

Intestine

Muscle Pancreas

Cdc42

PAK1

Cdc42

PAK1

PI3K

Akt Cdc42

Rac1

AS160PAK1

GLUT4translocation

Rac1 Insulin

secretion

GLP-1secretion

GLP-1

Elevated insulin Elevated glucose

Incretin effect

Glucose uptake

Insulin response

FEEDING

Fig. 4. A summary of current knowledge on the role of PAK1 in glucose homeostasis. Upon feeding, the rise in blood glucose level leads to elevated circulatinginsulin levels. In the muscle, insulin stimulates GLUT4 translocation via the phosphoinositide 3-kinase (PI3K)-Cdc42-Rac1-PAK1 signaling axis and thePI3K-Akt-AS160 axis in promoting muscle glucose uptake (left). In the pancreas, increased blood glucose level elicits the postprandial release of insulin formthe-cells, involving the Cdc42-PAK1-Rac1 signaling cascade (right). The postprandial insulin response is augmented by the incretin hormone glucagon-likepeptide-1 (GLP-1; middle). In the gut L cells, insulin activates the PAK1--cat signaling pathway, leading to the stimulation ofgcg transcription and GLP-1production.

Review




11/16

The studies deciphering their role in regulating gene expressionprompted us to examine the role of PAK1 in gut gcgexpressionand GLP-1 production. In the cancer research field, PAKs havebeen shown to function as effectors of growth factors, and theyare evidently involved in stimulating cell proliferation andinhibiting cell apoptosis. In the Pak1/ mice, we, however,did not observe noticeable proliferative defects in the pancre-

atic -cells and elsewhere (29). Again, this suggests that theother two group I PAKs may compensate the function of PAK1in the Pak1/ mice. Hence, we will need to employ bothgain-of-function and loss-of-function approaches, under bothin vitro and in vivo settings, to further elucidate whether andhow members of group I PAKs play redundant roles in con-trolling -cell proliferation. It should be pointed out that PAK3may coordinate the stress response (148). A very recent studyindicated that PAK3 stimulates cell cycle exit and differentia-tion of embryonic pancreatic -cells (139). PAK3 is essentialfor maintaining glucose homeostasis in response to high-fatdiet challenge (139). Hence, group I PAK members may exertnot only redundant and compensatory function but also oppo-site functions in pancreatic -cells. This is worthy of furtherinvestigation.

In summary, extensive in vitro and in vivo investigations todate have uncovered that PAK1 exerts critical functions in themuscle, pancreas, and gut in regulating glucose disposal. Fu-ture exploration, including the expansion of the study intoother PAK members as well as additional metabolic tissues,will aid in our understanding of the complex regulatory net-works underlying glucose homeostasis during health and dis-ease.

ACKNOWLEDGMENTS

We regret that we cannot cite other excellent publications related to thetopic in this review article.

GRANTS

The research on PAK1 and Wnt signaling pathway in the laboratory of Dr.T. Jin has been supported by operating grants from Canadian Institutes ofHealth Research (CIHR; MOP-89987 and MOP-97790) and an operating grantfrom Canadian Diabetes Association (CDA; OG-3-10-3040).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

Author contributions: Y.-t.A.C. and T.J. conception and design of research;Y.-t.A.C. prepared figures; Y.-t.A.C. drafted manuscript; Y.-t.A.C. and T.J.edited and revised manuscript; Y.-t.A.C. and T.J. approved final version ofmanuscript.

REFERENCES

1. Abo A, Qu J, Cammarano MS, Dan C, Fritsch A, Baud V, Belisle B,Minden A. PAK4, a novel effector for Cdc42Hs, is implicated in thereorganization of the actin cytoskeleton and in the formation of filopodia.

EMBO J 17: 65276540, 1998.2. Alahari SK, Reddig PJ, Juliano RL. The integrin-binding protein

Nischarin regulates cell migration by inhibiting PAK. EMBO J 23:27772788, 2004.

3. Alan JK, Berzat AC, Dewar BJ, Graves LM, Cox AD. Regulation ofthe Rho family small GTPase Wrch-1/RhoU by C-terminal tyrosinephosphorylation requires Src. Mol Cell Biol 30: 43244338, 2010.

4. Alessi DR, Sakamoto K, Bayascas JR. LKB1-dependent signalingpathways. Annu Rev Biochem 75: 137163, 2006.

5. Allen JD, Jaffer ZM, Park SJ, Burgin S, Hofmann C, Sells MA, ChenS, Derr-Yellin E, Michels EG, McDaniel A, Bessler WK, Ingram DA,

Atkinson SJ, Travers JB, Chernoff J, Clapp DW. p21-activated kinaseregulates mast cell degranulation via effects on calcium mobilization and

cytoskeletal dynamics. Blood 113: 26952705, 2009.

6. Amstutz B, Gastaldelli M, Klin S, Imelli N, Boucke K, Wandeler E,Mercer J, Hemmi S, Greber UF. Subversion of CtBP1-controlledmacropinocytosis by human adenovirus serotype 3. EMBO J27: 956

969, 2008.

7. Andersen DK. Diabetes and cancer: placing the association in perspec-

tive. Curr Opin Endocrinol Diabetes Obes 20: 8186, 2013.8. Aoki H, Yokoyama T, Fujiwara K, Tari AM, Sawaya R, Suki D, HessKR, Aldape KD, Kondo S, Kumar R, Kondo Y. Phosphorylated Pak1level in the cytoplasm correlates with shorter survival time in patients

with glioblastoma. Clin Cancer Res 13: 66036609, 2007.9. Arias-Romero LE, Chernoff J. A tale of two Paks. Biol Cell 100:

97108, 2008.

10. Arias-Romero LE, Villamar-Cruz O, Huang M, Hoeflich KP, Cher-noff J.Pak1 kinase links ErbB2 to -catenin in transformation of breastepithelial cells. Cancer Res 73: 36713682, 2013.

11. Arous C, Rondas D, Halban PA.Non-muscle myosin IIA is involved infocal adhesion and actin remodelling controlling glucose-stimulated

insulin secretion. Diabetologia 56: 792802, 2013.12. Asahara S, Shibutani Y, Teruyama K, Inoue HY, Kawada Y, Etoh H,

Matsuda T, Kimura-Koyanagi M, Hashimoto N, Sakahara M, Fu-jimoto W, Takahashi H, Ueda S, Hosooka T, Satoh T, Inoue H,Matsumoto M, Aiba A, Kasuga M, Kido Y. Ras-related C3 botulinumtoxin substrate 1 (RAC1) regulates glucose-stimulated insulin secretionvia modulation of F-actin. Diabetologia 56: 10881097, 2013.

13. Asrar S, Meng Y, Zhou Z, Todorovski Z, Huang WW, Jia Z.Regulation of hippocampal long-term potentiation by p21-activated pro-tein kinase 1 (PAK1). Neuropharmacology 56: 7380, 2009.

14. Aunis D, Bader MF. The cytoskeleton as a barrier to exocytosis insecretory cells. J Exp Biol 139: 253266, 1988.

15. Bagrodia S, Cerione RA. Pak to the future. Trends Cell Biol 9:350355, 1999.

16. Bagrodia S, Taylor SJ, Creasy CL, Chernoff J, Cerione RA. Identi-fication of a mouse p21Cdc42/Rac activated kinase. J Biol Chem 270:

2273122737, 1995.17. Balasenthil S, Vadlamudi RK. Functional interactions between the

estrogen receptor coactivator PELP1/MNAR and retinoblastoma protein.J Biol Chem 278: 2211922127, 2003.

18. Beeser A, Jaffer ZM, Hofmann C, Chernoff J. Role of group A

p21-activated kinases in activation of extracellular-regulated kinase bygrowth factors. J Biol Chem 280: 3660936615, 2005.

19. Belardi V, Gallagher EJ, Novosyadlyy R, Leroith D. Insulin and IGFsin obesity-related breast cancer.J Mammary Gland Biol Neoplasia2013.

20. Bokoch GM. Biology of the p21-activated kinases. Annu Rev Biochem72: 743781, 2003.

21. Bokoch GM, Wang Y, Bohl BP, Sells MA, Quilliam LA, Knaus UG.Interaction of the Nck adapter protein with p21-activated kinase (PAK1).

J Biol Chem 271: 2574625749, 1996.22. Bostner J, Ahnstrm Waltersson M, Fornander T, Skoog L, Nor-

denskjld B, Stl O. Amplification of CCND1 and PAK1 as predictorsof recurrence and tamoxifen resistance in postmenopausal breast cancer.Oncogene 26: 69977005, 2007.

23. Brozinick JT, Hawkins ED, Strawbridge AB, Elmendorf JS. Disrup-tion of cortical actin in skeletal muscle demonstrates an essential role ofthe cytoskeleton in glucose transporter 4 translocation in insulin-sensitive

tissues.J Biol Chem 279: 4069940706, 2004.

24. Buss F, Kendrick-Jones J, Lionne C, Knight AE, Ct GP, PaulLuzio J.The localization of myosin VI at the golgi complex and leadingedge of fibroblasts and its phosphorylation and recruitment into mem-

brane ruffles of A431 cells after growth factor stimulation. J Cell Biol

143: 15351545, 1998.

25. Campbell JE, Drucker DJ. Pharmacology, physiology, and mecha-nisms of incretin hormone action. Cell Metab 17: 819837, 2013.

26. Chan PM, Manser E.PAKs in human disease.Prog Mol Biol Transl Sci106: 171187, 2012.

27. Chen L, Liu H, Liu J, Zhu Y, Xu L, He H, Zhang H, Wang S, Wu Q,Liu W, Liu Y, Pan D, Ren S, Xu J, Gu J. Klotho endows hepatomacells with resistance to anoikis via VEGFR2/PAK1 activation in hepa-

tocellular carcinoma. PLoS One 8: e58413, 2013.28. Chew TL, Masaracchia RA, Goeckeler ZM, Wysolmerski RB.Phos-

phorylation of non-muscle myosin II regulatory light chain by p21-

Review




12/16

activated kinase (gamma-PAK). J Muscle Res Cell Motil 19: 839854,

1998.

29. Chiang YA, Shao W, Xu XX, Chernoff J, Jin T.P21-activated proteinkinase 1 (Pak1) mediates the cross talk between insulin and -catenin on

proglucagon gene expression and its ablation affects glucose homeostasis

in male C57BL/6 mice.Endocrinology 154: 7788, 2013.

30. Chiang YT, Ip W, Jin T. The role of the Wnt signaling pathway inincretin hormone production and function. Front Physiol 3: 273, 2012.

31. Chiu TT, Jensen TE, Sylow L, Richter EA, Klip A. Rac1 signallingtowards GLUT4/glucose uptake in skeletal muscle. Cell Signal 23:

15461554, 2011.

32. Choi SY, Chacon-Heszele MF, Huang L, McKenna S, Wilson FP,Zuo X, Lipschutz JH.Cdc42 deficiency causes ciliary abnormalities andcystic kidneys. J Am Soc Nephrol 24: 14351450, 2013.

33. Chu J, Pham NT, Olate N, Kislitsyna K, Day MC, LeTourneau PA,Kots A, Stewart RH, Laine GA, Cox CS, Uray K. Biphasic regulationof myosin light chain phosphorylation by p21-activated kinase modulates

intestinal smooth muscle contractility. J Biol Chem 288: 1200 1213,

2013.

34. Chun KH, Araki K, Jee Y, Lee DH, Oh BC, Huang H, Park KS, LeeSW, Zabolotny JM, Kim YB. Regulation of glucose transport byROCK1 differs from that of ROCK2 and is controlled by actin polym-

erization.Endocrinology 153: 16491662, 2012.

35. Dadke D, Fryer BH, Golemis EA, Field J. Activation of p21-activated

kinase 1-nuclear factor kappaB signaling by Kaposis sarcoma-associatedherpes virus G protein-coupled receptor during cellular transformation.

Cancer Res 63: 88378847, 2003.

36. Dan C, Kelly A, Bernard O, Minden A. Cytoskeletal changes regulatedby the PAK4 serine/threonine kinase are mediated by LIM kinase 1 and

cofilin.J Biol Chem 276: 3211532121, 2001.

37. Dan C, Nath N, Liberto M, Minden A. PAK5, a new brain-specifickinase, promotes neurite outgrowth in N1E-115 cells. Mol Cell Biol 22:

567577, 2002.

38. Deacon SW, Beeser A, Fukui JA, Rennefahrt UE, Myers C, ChernoffJ, Peterson JR. An isoform-selective, small-molecule inhibitor targetsthe autoregulatory mechanism of p21-activated kinase. Chem Biol 15:

322331, 2008.

39. Deguchi A, Miyoshi H, Kojima Y, Okawa K, Aoki M, Taketo MM.LKB1 suppresses p21-activated kinase-1 (PAK1) by phosphorylation of

Thr109 in the p21-binding domain. J Biol Chem 285: 1828318290,

2010.

40. Delorme V, Machacek M, DerMardirossian C, Anderson KL, Witt-mann T, Hanein D, Waterman-Storer C, Danuser G, Bokoch GM.Cofilin activity downstream of Pak1 regulates cell protrusion efficiency

by organizing lamellipodium and lamella actin networks. Dev Cell 13:

646662, 2007.

41. Diebold BA, Fowler B, Lu J, Dinauer MC, Bokoch GM.Antagonisticcross-talk between Rac and Cdc42 GTPases regulates generation of

reactive oxygen species. J Biol Chem 279: 2813628142, 2004.

42. Djiogue S, Nwabo Kamdje AH, Vecchio L, Kipanyula MJ, FarahnaM, Aldebasi Y, Seke Etet PF. Insulin resistance and cancer: the role ofinsulin and IGFs. Endocr Relat Cancer20: R1R17, 2013.

43. Dossus L, Lukanova A, Rinaldi S, Allen N, Cust AE, Becker S,Tjonneland A, Hansen L, Overvad K, Chabbert-Buffet N, Mesrine S,Clavel-Chapelon F, Teucher B, Chang-Claude J, Boeing H, DroganD, Trichopoulou A, Benetou V, Bamia C, Palli D, Agnoli C, GalassoR, Tumino R, Sacerdote C, Bueno-de-Mesquita HB, van Duijnhoven

FJ, Peeters PH, Onland-Moret NC, Redondo ML, Travier N, San-chez MJ, Altzibar JM, Chirlaque MD, Barricarte A, Lundin E,Khaw KT, Wareham N, Fedirko V, Romieu I, Romaguera D, NoratT, Riboli E, Kaaks R. Hormonal, metabolic, and inflammatory profilesand endometrial cancer risk within the EPIC cohorta factor analysis.Am

J Epidemiol177: 787799, 2013.

44. Dummler B, Ohshiro K, Kumar R, Field J. Pak protein kinases andtheir role in cancer. Cancer Metastasis Rev 28: 5163, 2009.

45. Eggers CM, Kline ER, Zhong D, Zhou W, Marcus AI. STE20-relatedkinase adaptor protein (STRAD) regulates cell polarity and invasion

through PAK1 signaling in LKB1-null cells. J Biol Chem 287: 18758

18768, 2012.

46. Esposito K, Chiodini P, Colao A, Lenzi A, Giugliano D. Metabolicsyndrome and risk of cancer: a systematic review and meta-analysis.

Diabetes Care 35: 24022411, 2012.

47. Eswaran J, Soundararajan M, Kumar R, Knapp S. UnPAKing theclass differences among p21-activated kinases. Trends Biochem Sci 33:394403, 2008.

48. Friedland JC, Lakins JN, Kazanietz MG, Chernoff J, Boettiger D,Weaver VM. alpha6beta4 integrin activates Rac-dependent p21-acti-vated kinase 1 to drive NF-kappaB-dependent resistance to apoptosis in3D mammary acini. J Cell Sci 120: 37003712, 2007.

49. Frost JA, Khokhlatchev A, Stippec S, White MA, Cobb MH. Differ-

ential effects of PAK1-activating mutations reveal activity-dependentand -independent effects on cytoskeletal regulation. J Biol Chem 273:2819128198, 1998.

50. Frost JA, Steen H, Shapiro P, Lewis T, Ahn N, Shaw PE, Cobb MH.Cross-cascade activation of ERKs and ternary complex factors by Rhofamily proteins. EMBO J16: 64266438, 1997.

51. Frost JA, Swantek JL, Stippec S, Yin MJ, Gaynor R, Cobb MH.Stimulation of NFkappaB activity by multiple signaling pathways re-quires PAK1. J Biol Chem 275: 1969319699, 2000.

52. Fu YM, Yu ZX, Lin H, Fu X, Meadows GG. Selective amino acidrestriction differentially affects the motility and directionality of DU145and PC3 prostate cancer cells. J Cell Physiol 217: 184193, 2008.

53. Furnari MA, Jobes ML, Nekrasova T, Minden A, Wagner GC.Functional deficits in PAK5, PAK6 and PAK5/PAK6 knockout mice.PLoS One 8: e61321, 2013.

54. Galan Moya EM, Le Guelte A, Gavard J. PAKing up to the endothe-lium. Cell Signal 21: 17271737, 2009.

55. Goldsteinm S, Mullick V, Huddlestun B, Levine R. Action of muscularwork on transfer of sugars across cell barriers; comparison with action ofinsulin.Am J Physiol 173: 212216, 1953.

56. Grant SF, Thorleifsson G, Reynisdottir I, Benediktsson R, Man-olescu A, Sainz J, Helgason A, Stefansson H, Emilsson V, HelgadottirA, Styrkarsdottir U, Magnusson KP, Walters GB, Palsdottir E,Jonsdottir T, Gudmundsdottir T, Gylfason A, Saemundsdottir J,Wilensky RL, Reilly MP, Rader DJ, Bagger Y, Christiansen C,Gudnason V, Sigurdsson G, Thorsteinsdottir U, Gulcher JR, KongA, Stefansson K.Variant of transcription factor 7-like 2 (TCF7L2) geneconfers risk of type 2 diabetes. Nat Genet38: 320323, 2006.

57. Gu S, Kounenidakis M, Schmidt EM, Deshpande D, Alkahtani S,Alarifi S, Fller M, Alevizopoulos K, Lang F, Stournaras C. Rapidactivation of FAK/mTOR/p70S6K/PAK1-signaling controls the earlytestosterone-induced actin reorganization in colon cancer cells. CellSignal25: 6673, 2013.

58. Hino S, Tanji C, Nakayama KI, Kikuchi A. Phosphorylation ofbeta-catenin by cyclic AMP-dependent protein kinase stabilizes beta-catenin through inhibition of its ubiquitination.Mol Cell Biol 25: 90639072, 2005.

59. Hofmann C, Shepelev M, Chernoff J. The genetics of Pak. J Cell Sci117: 43434354, 2004.

60. Holm C, Rayala S, Jirstrm K, Stl O, Kumar R, Landberg G.Association between Pak1 expression and subcellular localization andtamoxifen resistance in breast cancer patients. J Natl Cancer Inst 98:671680, 2006.

61. Hoover WC, Zhang W, Xue Z, Gao H, Chernoff J, Clapp DW, GunstSJ, Tepper RS.Inhibition of p21 activated kinase (PAK) reduces airwayresponsiveness in vivo and in vitro in murine and human airways. PLoSOne7: e42601, 2012.

62. Huang H, Lee DH, Zabolotny JM, Kim YB. Metabolic actions ofRho-kinase in periphery and brain. Trends Endocrinol Metab 2013.

63. Huang W, Zhou Z, Asrar S, Henkelman M, Xie W, Jia Z. p21-

Activated kinases 1 and 3 control brain size through coordinatingneuronal complexity and synaptic properties. Mol Cell Biol31: 388403,2011.

64. Ijuin T, Mochizuki Y, Fukami K, Funaki M, Asano T, Takenawa T.Identification and characterization of a novel inositol polyphosphate5-phosphatase.J Biol Chem 275: 1087010875, 2000.

65. Ijuin T, Takenawa T. Regulation of insulin signaling and glucosetransporter 4 (GLUT4) exocytosis by phosphatidylinositol 3,4,5-trispho-sphate (PIP3) phosphatase, skeletal muscle, and kidney enriched inositolpolyphosphate phosphatase (SKIP).J Biol Chem287: 69916999, 2012.

66. Ijuin T, Takenawa T. Regulation of insulin signaling by the phospha-tidylinositol 3,4,5-triphosphate phosphatase SKIP through the scaffold-ing function of Pak1. Mol Cell Biol 32: 35703584, 2012.

67. Ijuin T, Takenawa T. SKIP negatively regulates insulin-inducedGLUT4 translocation and membrane ruffle formation. Mol Cell Biol 23:12091220, 2003.

Review




13/16

68. Inoue E, Mochida S, Takagi H, Higa S, Deguchi-Tawarada M,Takao-Rikitsu E, Inoue M, Yao I, Takeuchi K, Kitajima I, Setou M,Ohtsuka T, Takai Y. SAD: a presynaptic kinase associated with syn-aptic vesicles and the active zone cytomatrix that regulates neurotrans-

mitter release. Neuron 50: 261275, 2006.

69. Inoue M, Tsugane S. Insulin resistance and cancer: epidemiologicalevidence.Endocr Relat Cancer 19: F1F8, 2012.

70. Ip W, Chiang YT, Jin T.The involvement of the wnt signaling pathway

and TCF7L2 in diabetes mellitus: The current understanding, dispute,and perspective. Cell Biosci 2: 28, 2012.

71. Ishikura S, Klip A. Muscle cells engage Rab8A and myosin Vb ininsulin-dependent GLUT4 translocation.Am J Physiol Cell Physiol295:

C1016C1025, 2008.

72. Ito M, Nishiyama H, Kawanishi H, Matsui S, Guilford P, Reeve A,Ogawa O.P21-activated kinase 1: a new molecular marker for intraves-ical recurrence after transurethral resection of bladder cancer. J Urol178:

10731079, 2007.

73. Jaffer ZM, Chernoff J. p21-activated kinases: three more join the Pak.Int J Biochem Cell Biol 34: 713717, 2002.

74. JeBailey L, Wanono O, Niu W, Roessler J, Rudich A, Klip A.Ceramide- and oxidant-induced insulin resistance involve loss of insulin-

dependent Rac-activation and actin remodeling in muscle cells. Diabetes

56: 394403, 2007.

75. Jhaveri KA, Debnath P, Chernoff J, Sanders J, Schwartz MA. The

role of p21-activated kinase in the initiation of atherosclerosis. BMCCardiovasc Disord12: 55, 2012.

76. Jin T.Mechanisms underlying proglucagon gene expression. J Endocri-nol 198: 1728, 2008.

77. Jin T.The WNT signalling pathway and diabetes mellitus. Diabetologia51: 17711780, 2008.

78. Jin T, George Fantus I, Sun J. Wnt and beyond Wnt: multiplemechanisms control the transcriptional property of beta-catenin. Cell

Signal 20: 16971704, 2008.

79. Jin T, Liu L. The Wnt signaling pathway effector TCF7L2 and type 2diabetes mellitus. Mol Endocrinol 22: 23832392, 2008.

81. Jinjuvadia R, Lohia P, Jinjuvadia C, Montoya S, Liangpunsakul S.The association between metabolic syndrome and colorectal neoplasm:

systemic review and meta-analysis. J Clin Gastroenterol 47: 3344,

2013.

82. Johnston JB, Barrett JW, Chang W, Chung CS, Zeng W, Masters J,Mann M, Wang F, Cao J, McFadden G. Role of the serine-threoninekinase PAK-1 in myxoma virus replication. J Virol77: 58775888, 2003.

83. Kalwat MA, Thurmond DC.Signaling mechanisms of glucose-inducedF-actin remodeling in pancreatic islet cells. Exp Mol Med 45: e37,

2013.

84. Kalwat MA, Yoder SM, Wang Z, Thurmond DC. A p21-activatedkinase (PAK1) signaling cascade coordinately regulates F-actin remod-

eling and insulin granule exocytosis in pancreatic cells. Biochem

Pharmacol85: 808816, 2013.

85. Kane S, Sano H, Liu SC, Asara JM, Lane WS, Garner CC, LienhardGE.A method to identify serine kinase substrates. Akt phosphorylates anovel adipocyte protein with a Rab GTPase-activating protein (GAP)

domain. J Biol Chem 277: 2211522118, 2002.

86. Ke Y, Lei M, Solaro RJ. Regulation of cardiac excitation and contrac-tion by p21 activated kinase-1. Prog Biophys Mol Biol 98: 238250,

2008.

87. Kelly ML, Astsaturov A, Chernoff J. Role of p21-activated kinases incardiovascular development and function. Cell Mol Life Sci 2013.

88. Kelly ML, Chernoff J. Mouse models of PAK function. Cell Logist2:8488, 2012.

89. Kennedy LM, Pham SC, Grishok A. Nonautonomous regulation ofneuronal migration by insulin signaling, DAF-16/FOXO, and PAK-1.

Cell Rep 4: 9961009, 2013.

90. Kichina JV, Goc A, Al-Husein B, Somanath PR, Kandel ES. PAK1 asa therapeutic target. Expert Opin Ther Targets 14: 703725, 2010.

91. King AJ, Sun H, Diaz B, Barnard D, Miao W, Bagrodia S, MarshallMS.The protein kinase Pak3 positively regulates Raf-1 activity throughphosphorylation of serine 338. Nature 396: 180183, 1998.

92. King CC, Gardiner EM, Zenke FT, Bohl BP, Newton AC, HemmingsBA, Bokoch GM. p21-activated kinase (PAK1) is phosphorylated andactivated by 3-phosphoinositide-dependent kinase-1 (PDK1). J Biol

Chem 275: 4120141209, 2000.

93. Kissil JL, Wilker EW, Johnson KC, Eckman MS, Yaffe MB, Jacks T.Merlin, the product of the Nf2 tumor suppressor gene, is an inhibitor ofthe p21-activated kinase, Pak1. Mol Cell 12: 841849, 2003.

94. Kotani K, Yonezawa K, Hara K, Ueda H, Kitamura Y, Sakaue H,Ando A, Chavanieu A, Calas B, Grigorescu F. Involvement of phos-phoinositide 3-kinase in insulin- or IGF-1-induced membrane ruffling.

EMBO J 13: 23132321, 1994.95. Kowluru A.Small G proteins in islet beta-cell function.Endocr Rev31:

5278, 2010.96. Kreis P, Barnier JV. PAK signalling in neuronal physiology. CellSignal21: 384393, 2009.

97. Kumar R, Gururaj AE, Barnes CJ. p21-activated kinases in cancer.Nat Rev Cancer6: 459471, 2006.

98. Kundumani-Sridharan V, Singh NK, Kumar S, Gadepalli R, RaoGN.Nuclear factor of activated T cells c1 mediates p21-activated kinase1 activation in the modulation of chemokine-induced human aorticsmooth muscle cell F-actin stress fiber formation, migration, and prolif-eration and injury-induced vascular wall remodeling. J Biol Chem 288:2215022162, 2013.

99. Lee SR, Ramos SM, Ko A, Masiello D, Swanson KD, Lu ML, BalkSP. AR and ER interaction with a p21-activated kinase (PAK6). Mol

Endocrinol16: 8599, 2002.100. Lei M, Lu W, Meng W, Parrini MC, Eck MJ, Mayer BJ, Harrison

SC. Structure of PAK1 in an autoinhibited conformation reveals amultistage activation switch. Cell 102: 387397, 2000.

101. Li F, Adam L, Vadlamudi RK, Zhou H, Sen S, Chernoff J, MandalM, Kumar R. p21-activated kinase 1 interacts with and phosphorylateshistone H3 in breast cancer cells. EMBO Rep 3: 767773, 2002.

102. Li J, Luo R, Kowluru A, Li G. Novel regulation by Rac1 of glucose-and forskolin-induced insulin secretion in INS-1 -cells. Am J Physiol

Endocrinol Metab 286: E818E827, 2004.103. Li X, Liu F, Li F. PAK as a therapeutic target in gastric cancer. Expert

Opin Ther Targets 14: 419433, 2010.104. Li X, Minden A.Targeted disruption of the gene for the PAK5 kinase in

mice.Mol Cell Biol 23: 71347142, 2003.105. Liberali P, Kakkonen E, Turacchio G, Valente C, Spaar A, Perinetti

G, Bckmann RA, Corda D, Colanzi A, Marjomaki V, Luini A. Theclosure of Pak1-dependent macropinosomes requires the phosphorylationof CtBP1/BARS. EMBO J27: 970981, 2008.

106. Licciulli S, Maksimoska J, Zhou C, Troutman S, Kota S, Liu Q,Duron S, Campbell D, Chernoff J, Field J, Marmorstein R, Kissil JL.FRAX597, a small molecule inhibitor of the p21-activated kinases,inhibits tumorigenesis of NF2-associated schwannomas. J Biol Chem

288: 2910529114, 2013.107. Lim GE, Xu M, Sun J, Jin T, Brubaker PL. The rho guanosine

5=-triphosphatase, cell division cycle 42, is required for insulin-inducedactin remodeling and glucagon-like peptide-1 secretion in the intestinalendocrine L cell. Endocrinology 150: 52495261, 2009.

108. Liu J, Liu H, Zhang W, Wu Q, Liu W, Liu Y, Pan D, Xu J, Gu J.N-acetylglucosaminyltransferase V confers hepatoma cells with resis-tance to anoikis through EGFR/PAK1 activation. Glycobiology 23:10971109, 2013.

109. Liu J, Visser-Grieve S, Boudreau J, Yeung B, Lo S, Chamberlain G,Yu F, Sun T, Papanicolaou T, Lam A, Yang X, Chin-Sang I. Insulinactivates the insulin receptor to downregulate the PTEN tumour suppres-sor. Oncogene 2013 Sep 2 [E-pub ahead of print].

110. Liu W, Zi M, Naumann R, Ulm S, Jin J, Taglieri DM, Prehar S, GuiJ, Tsui H, Xiao RP, Neyses L, Solaro RJ, Ke Y, Cartwright EJ, Lei

M, Wang X. Pak1 as a novel therapeutic target for antihypertrophictreatment in the heart. Circulation 124: 27022715, 2011.111. Luo ZG, Wang Q, Zhou JZ, Wang J, Luo Z, Liu M, He X,

Wynshaw-Boris A, Xiong WC, Lu B, Mei L. Regulation of AChRclustering by Dishevelled interacting with MuSK and PAK1.Neuron35:489505, 2002.

112. Lv Z, Hu M, Zhen J, Lin J, Wang Q, Wang R. Rac1/PAK1 signalingpromotes epithelial-mesenchymal transition of podocytes in vitro viatriggering -catenin transcriptional activity under high glucose condi-tions.Int J Biochem Cell Biol 45: 255264, 2013.

113. Manser E, Huang HY, Loo TH, Chen XQ, Dong JM, Leung T, LimL. Expression of constitutively active alpha-PAK reveals effects of thekinase on actin and focal complexes.Mol Cell Biol17: 11291143, 1997.

114. Manser E, Leung T, Salihuddin H, Zhao ZS, Lim L. A brainserine/threonine protein kinase activated by Cdc42 and Rac1. Nature

367: 4046, 1994.

Review




14/16

115. Manser E, Loo TH, Koh CG, Zhao ZS, Chen XQ, Tan L, Tan I,Leung T, Lim L.PAK kinases are directly coupled to the PIX family ofnucleotide exchange factors. Mol Cell 1: 183192, 1998.

116. Mao X, Onadim Z, Price EA, Child F, Lillington DM, Russell-JonesR, Young BD, Whittaker S. Genomic alterations in blastic naturalkiller/extranodal natural killer-like T cell lymphoma with cutaneous

involvement.J Invest Dermatol 121: 618627, 2003.

117. Maroto B, Ye MB, von Lohneysen K, Schnelzer A, Knaus UG.

P21-activated kinase is required for mitotic progression and regulatesPlk1. Oncogene 27: 49004908, 2008.

118. Martin GA, Bollag G, McCormick F, Abo A. A novel serine kinaseactivated by rac1/CDC42Hs-dependent autophosphorylation is related to

PAK65 and STE20. EMBO J14: 4385, 1995.

119. Mazumdar A, Kumar R. Estrogen regulation of Pak1 and FKHRpathways in breast cancer cells. FEBS Lett535: 610, 2003.

120. McDaniel AS, Allen JD, Park SJ, Jaffer ZM, Michels EG, Burgin SJ,Chen S, Bessler WK, Hofmann C, Ingram DA, Chernoff J, ClappDW. Pak1 regulates multiple c-Kit mediated Ras-MAPK gain-in-func-tion phenotypes in Nf1/ mast cells. Blood112: 46464654, 2008.

121. Meng J, Meng Y, Hanna A, Janus C, Jia Z. Abnormal long-lastingsynaptic plasticity and cognition in mice lacking the mental retardation

gene Pak3. J Neurosci 25: 66416650, 2005.

122. Molli PR, Li DQ, Murray BW, Rayala SK, Kumar R.PAK signalingin oncogenesis. Oncogene 28: 25452555, 2009.

123. Molnar A, Theodoras AM, Zon LI, Kyriakis JM. Cdc42Hs, but notRac1, inhibits serum-stimulated cell cycle progression at G1/S through amechanism requiring p38/RK. J Biol Chem 272: 1322913235, 1997.

124. Nekrasova T, Jobes ML, Ting JH, Wagner GC, Minden A. Targeteddisruption of the Pak5 and Pak6 genes in mice leads to deficits in learning

and locomotion. Dev Biol 322: 95108, 2008.

125. Ni Z, Anini Y, Fang X, Mills G, Brubaker PL, Jin T. Transcriptionalactivation of the proglucagon gene by lithium and beta-catenin in

intestinal endocrine L cells. J Biol Chem 278: 13801387, 2003.

126. Nie J, Sun C, Faruque O, Ye G, Li J, Liang Q, Chang Z, Yang W,Han X, Shi Y.Synapses of amphids defective (SAD-A) kinase promotesglucose-stimulated insulin secretion through activation of p21-activated

kinase (PAK1) in pancreatic -Cells. J Biol Chem 287: 2643526444,

2012.

127. Nikolic M. The Pak1 kinase: an important regulator of neuronal mor-phology and function in the developing forebrain. Mol Neurobiol 37:

187202, 2008.

128. Nozaki S, Ueda S, Takenaka N, Kataoka T, Satoh T. Role of RalAdownstream of Rac1 in insulin-dependent glucose uptake in muscle cells.Cell Signal 24: 21112117, 2012.

129. Ong CC,Jubb AM, Haverty PM, Zhou W, Tran V, Truong T, TurleyH, OBrien T, Vucic D, Harris AL, Belvin M, Friedman LS, Black-wood EM, Koeppen H, Hoeflich KP. Targeting p21-activated kinase 1(PAK1) to induce apoptosis of tumor cells.Proc Natl Acad Sci USA 108:

71777182, 2011.

130. Ong CC, Jubb AM, Zhou W, Haverty PM, Harris AL, Belvin M,Friedman LS, Koeppen H, Hoeflich KP. p21-activated kinase 1:PAKed with potential. Oncotarget2: 491496, 2011.

131. Orci L, Gabbay KH, Malaisse WJ. Pancreatic beta-cell web: itspossible role in insulin secretion. Science 175: 11281130, 1972.

132. Orgel E, Mittelman SD. The links between insulin resistance, diabetes,and cancer. Curr Diab Rep 13: 213222, 2013.

133. Pacheco A, Chernoff J. Group I p21-activated kinases: emerging rolesin immune function and viral pathogenesis. Int J Biochem Cell Biol 42:

1316, 2010.

134. Park MH, Kim DJ, You ST, Lee CS, Kim HK, Park SM, Shin EY,Kim EG. Phosphorylation of -catenin at serine 663 regulates itstranscriptional activity. Biochem Biophys Res Commun 419: 543549,

2012.

135. Parrini MC, Camonis J, Matsuda M, de Gunzburg J. Dissectingactivation of the PAK1 kinase at protrusions in living cells. J Biol Chem

284: 2413324143, 2009.

136. Pascua PN, Lee JH, Song MS, Park SJ, Baek YH, Ann BH, Shin EY,Kim EG, Choi YK. Role of the p21-activated kinases (PAKs) ininfluenza A virus replication. Biochem Biophys Res Commun 414:

569574, 2011.

137. Philippe J. Glucagon gene transcription is negatively regulated byinsulin in a hamster islet cell line. J Clin Invest84: 672677, 1989.

138. Philippe J. Insulin regulation of the glucagon gene is mediated by aninsulin-responsive DNA element. Proc Natl Acad Sci USA 88: 7224

7227, 1991.

139. Piccand J, Meunier A, Merle C, Jia Z, Barnier JV, Gradwohl G.Pak3 promotes cell cycle exit and differentiation of beta-cells in the

embryonic pancreas and is necessary to maintain glucose homeostasis in

adult mice. Diabetes 63: 203215, 2013.

140. Podkowa M, Christova T, Zhao X, Jian Y, Attisano L. p21-Activated

kinase (PAK) is required for bone morphogenetic protein (BMP)-induceddendritogenesis in cortical neurons. Mol Cell Neurosci 57C: 8392,

2013.

141. Puto LA, Pestonjamasp K, King CC, Bokoch GM. p21-activatedkinase 1 (PAK1) interacts with the Grb2 adapter protein to couple to

growth factor signaling. J Biol Chem 278: 93889393, 2003.

142. Qiao M, Shapiro P, Kumar R, Passaniti A.Insulin-like growth factor-1regulates endogenous RUNX2 activity in endothelial cells through a

phosphatidylinositol 3-kinase/ERK-dependent and Akt-independent sig-

naling pathway. J Biol Chem 279: 4270942718, 2004.

143. Qu J, Li X, Novitch BG, Zheng Y, Kohn M, Xie JM, Kozinn S,Bronson R, Beg AA, Minden A.PAK4 kinase is essential for embryonicviability and for proper neuronal development. Mol Cell Biol23: 7122

7133, 2003.

144. Ramos E, Wysolmerski RB, Masaracchia RA. Myosin phosphoryla-tion by human cdc42-dependent S6/H4 kinase/gammaPAK from pla-

centa and lymphoid cells. Recept Signal Transduct7: 99110, 1997.145. Reddy SD, Ohshiro K, Rayala SK, Kumar R.MicroRNA-7, a homeo-box D10 target, inhibits p21-activated kinase 1 and regulates its func-

tions.Cancer Res 68: 81958200, 2008.

146. Rhee S, Grinnell F. P21-activated kinase 1: convergence point in PDGF-and LPA-stimulated collagen matrix contraction by human fibroblasts. J

Cell Biol 172: 423432, 2006.

147. R TB, Holien T, Fagerli UM, Hov H, Misund K, Waage A, SundanA, Holt RU, Brset M. HGF and IGF-1 synergize with SDF-1 inpromoting

Documents

e707.Full Neeew to Read