Cellular Signalling 23 (2011) 731–738

Contents lists available at ScienceDirect

Cellular Signalling

j ourna l homepage: www.e lsev ie r.com/ locate /ce l l s ig

EGF-like peptide-enhanced cell motility in Dictyostelium functions independently ofthe cAMP-mediated pathway and requires active Ca2+/calmodulin signaling

Robert Huber a, Danton H. O'Day a,b,⁎a Department of Cell & Systems Biology, 25 Harbord Street, University of Toronto, Toronto, ON, Canada M5S 3G5b Department of Biology, University of Toronto at Mississauga, 3359 Mississauga Road, Mississauga, ON, Canada L5L 1C6

Abbreviations: EGFL, Epidermal growth factor-likCalmodulin-binding protein; Ca2+, Calcium; MHC, Myoskinase B; GEF, Guanine nucleotide exchange factor; GAPI3K, Phosphatidylinositol 3-kinase; PLA2, Phospholadenosine monophosphate; car, cAMP receptor; IP3,GDP, Guanine diphosphate; GTP, Guanine triphosphatesulfate polyacrylamide gel Electrophoresis.⁎ Corresponding author. Department of Biology, Unive

3359 Mississauga Road, Mississauga, ON, Canada L5L 1CE-mail addresses: robert.huber@utoronto.ca (R. Hub

(D.H. O'Day).

0898-6568/$ – see front matter © 2010 Elsevier Inc. Aldoi:10.1016/j.cellsig.2010.12.007

a b s t r a c t

a r t i c l e i n f o

Article history:Received 14 November 2010Received in revised form 14 December 2010Accepted 22 December 2010Available online 31 December 2010

Keywords:EGF-like peptideCell motilityDictyosteliumCalciumCalmodulinCytoskeleton

Current knowledge suggests that cell movement in the eukaryotic slime mold Dictyostelium discoideum ismediated by different signaling pathways involving a number of redundant components. Our previousresearch has identified a specific motility-enhancing function for epidermal growth factor-like (EGFL)repeats in Dictyostelium, specifically for the EGFL repeats of cyrA, a matricellular, calmodulin (CaM)-bindingprotein in Dictyostelium. Using mutants of cAMP signaling (carA−, carC−, gpaB−, gpbA−), the endogenouscalcium (Ca2+) release inhibitor TMB-8, the CaM antagonist W-7, and a radial motility bioassay, we showthat DdEGFL1, a synthetic peptide whose sequence is obtained from the first EGFL repeat of cyrA, functionsindependently of the cAMP-mediated signaling pathways to enhance cell motility through a mechanisminvolving Ca2+ signaling, CaM, and RasG. We show that DdEGFL1 increases the amounts of polymericmyosin II heavy chain and actin in the cytoskeleton by 24.1±10.7% and 25.9±2.1% respectively anddemonstrate a link between Ca2+/CaM signaling and cytoskeletal dynamics. Finally, our findings suggestthat carA and carC mediate a brake mechanism during chemotaxis since DdEGFL1 enhanced the movementof carA−/carC− cells by 844±136% compared to only 106±6% for parental DH1 cells. Based on our data, thissignaling pathway also appears to involve the G-protein β subunit, RasC, RasGEFA, and protein kinase B.Together, our research provides insight into the functionality of EGFL repeats in Dictyostelium and thesignaling pathways regulating cell movement in this model organism. It also identifies several mechanisticcomponents of DdEGFL1-enhanced cell movement, which may ultimately provide a model system forunderstanding EGFL repeat function in higher organisms.

e; CaM, Calmodulin; CaMBP,in II heavy chain; PKB, ProteinP, GTPase-activating protein;ipase A2; cAMP, 3′-5′-cyclicInositol 1,4,5-trisphosphate;; SDS-PAGE, Sodium dodecyl

rsity of Toronto atMississauga,6. Tel.: +1 905 828 3897.er), danton.oday@utoronto.ca

l rights reserved.

© 2010 Elsevier Inc. All rights reserved.

1. Introduction

Dictyostelium discoideum is a eukaryotic slime mold used as amodel system for the study of eukaryotic cell motility as well as anumber of other cell and developmental processes [1,2]. In particular,the mechanisms and proteins involved in regulating cell movementand chemotaxis in Dictyostelium are highly conserved with thosefound in higher organisms such as mammals [3]. In Dictyostelium,chemotactic stimulation by cAMP has been shown to cause an influxof calcium (Ca2+) into the cell from the external medium [4]. Ca2+

influx has also been shown to occur when intracellular stores of Ca2+

are depleted [5,6]. In addition, calmodulin (CaM), the primary sensorof Ca2+ fluxes within the cell, has been shown to be required forchemotaxis towards cAMP [7]. Chemotactic stimulation by cAMP hasalso been shown to affect cytoskeletal dynamics by increasing theassembly of myosin II heavy chain (MHC) and the polymerization ofactin in the cytoskeleton of chemotacting cells [8–11].

The movement of Dictyostelium amoebae has been extensivelystudied, however all lines of evidence suggest that the regulation ofmovement in thismodel eukaryote involvesmanypathways and likely anumber of redundant components, some of which have not yet beenidentified or characterized. Ras proteins are monomeric GTPases thatfunction as molecular switches, cycling between active GTP-bound andinactive GDP-bound states. They are activated by guanine-nucleotide-exchange factors (GEFs) and inactivated by GTPase-activating proteins(GAPs) [12]. Several Ras proteins have been identified in Dictyosteliumand some members of this family have been shown to be involved inregulating chemotaxis to cAMP [13–15].

We have previously shown that a synthetic EGFL peptide (DdEGFL1)from cyrA, an extracellular, CaM-binding protein (CaMBP) inDictyostelium (Suarez et al., manuscript submitted), enhances both

732 R. Huber, D.H. O'Day / Cellular Signalling 23 (2011) 731–738

random cell movement and cAMP-mediated chemotaxis [16]. Thiswas an interesting finding since genome analyses suggest thatDictyostelium possesses more epidermal growth factor-like (EGFL)domains than any other sequenced eukaryote [17]. The abundance ofthese domains in this model organism suggests functional signifi-cance. In mammals, certain EGFL repeats have been shown toenhance cell movement by binding to the EGF receptor [18,19].SadA, a novel adhesion receptor in Dictyostelium, possesses threeconserved EGFL repeats and has been proposed to function as a brakeduring vegetative cell movement [20]. Our initial study determinedthat DdEGFL1-enhanced cell movement requires active phosphati-dylinositol 3-kinase (PI3K) and phospholipase A2 (PLA2) [16], twoproteins that have been shown to mediate cAMP chemotaxis inparallel compensatory pathways [21,22].

In this paper we further investigated the signaling pathwaysmediating DdEGFL1-enhanced cell movement. We obtained adiversity of signaling mutants from the Dicty Stock Center (http://dictybase.org/StockCenter/StockCenter.html) to investigate theirresponse to DdEGFL1 peptides. Here we report that Ca2+ releasefrom intracellular stores and CaM are required for DdEGFL1-enhanced cell motility and that DdEGFL1-enhanced motility partial-ly requires RasG. In addition, we show that DdEGFL1 treatmentincreases the amount of polymeric MHC and actin in the cytoskel-eton. Our mutant analysis suggests that a brake mechanismmediated by the cAMP receptor exists in Dictyostelium. Takentogether, our research provides insight into how EGFL repeatsfunction to enhance cell movement in Dictyostelium and strengthensthe notion that this simple eukaryote can be used as a modelorganism for studying many aspects of cell movement observed inhigher systems.

2. Materials and methods

2.1. Cells, experimental peptides, chemicals

All D. discoideum strains were grown in the presence of Escherichiacoli on SM agar pH 6.5 at 22 °C in the dark for 24–30 h as previouslydescribed [16]. All mutant strains were obtained from the Dicty StockCenter (http://dictybase.org/StockCenter/StockCenter.html) andfresh stocks were used for all experiments. DdEGFL1 was synthesizedby a solid-phase method using the Fmoc strategy [16] and wasprovided as a gift from Dr. Yali Wang (Advanced Syntech Corporation,Mississauga, Ontario, Canada). The following chemicals were used inthis study; TMB-8, W-7, LY294002, quinacrine (EMD Biosciences Inc.,La Jolla, CA, USA), EGTA, BAPTA, BAPTA-AM (Sigma-Aldrich CanadaLtd., Oakville, Ontario, Canada).

2.2. Random cell motility radial bioassays

Random cell motility was analyzed using the radial bioassay[7,16,23,24]. Vegetative amoebae were harvested from SM agar andwashed three times in KK2 phosphate buffer pH 6.5 (2.31 g/Lpotassium phosphate monobasic, 1.3 g/L potassium phosphate diba-sic). Cells (~1.5×108 cells/ml) were plated in 0.1 μl volumes on 0.5%agar/KK2 containing the appropriate agents as detailed in the Resultsand Discussion. Pictures of cells were taken at 0 and 4 h using a Zeissinvertedmicroscope equipped with a Sony 950 digital camera. Imagesof cells were viewed and analyzed using the Northern Eclipse imageanalysis system (Empix Imaging, Mississauga, Ontario, Canada).Means ± standard error of the means (SEM) were calculated forcontrols and experimental conditions.

2.3. Protein expression analysis

Cells (~9×106 cells/ml) were starved in KK2 phosphate buffer ±DdEGFL1 (450 μM) and shaken at 22 °C and 160 rpm. Cells were

collected at designated times and lysed with a buffer containing50 mM Tris–HCl pH 7.4, 1% NP-40, 0.25% sodium deoxycholate,150 mM NaCl, 5 mM EDTA, 10 μg/ml leupeptin, 1 μg/ml pepstatin A,1 mM EGTA, 1 mM sodium orthovanadate, 10 mM sodium fluoride,and a protease inhibitor cocktail (Hoffmann-La Roche Limited,Mississauga, Ontario, Canada). Samples were stored at −80 °C forfuture use.

2.4. Cytoskeleton isolation

Cytoskeleton isolations were performed as previously described[25]. Vegetative AX3 amoebae were harvested from SM plates andwashed three times with KK2 phosphate buffer. Cells (~9×106 cells/ml) were starved ± DdEGFL1 (450 μM) for 4 h at room temperatureand spun at 160 rpm. At time 0 and time 4 h, cells were harvested andlysed with 400 μl of lysis buffer containing 100 mM PIPES, 5 mMEGTA, 5 mM MgCl2, pH 6.8, 0.5% Triton-X, and a protease inhibitorcocktail (Hoffmann-La Roche Limited, Mississauga, Ontario, Canada).Samples were placed on ice for 5 min. To standardize each sample,whole cell lysates were quantified with the Bradford assay to providea measure of the total protein in the lysate prior to the isolation of thecytoskeleton. The cytoskeleton was then isolated from 500 μg totalcell protein. Samples were spun for 6 min at 13000 rpm. Pellets wereresuspended in 400 μl of lysis buffer and spun for 6 min at 13000 rpm.This washing procedurewas performed two times. Sampleswere thenwashed once in lysis buffer that did not contain 0.5% Triton-X (washbuffer). The final pellet was resuspended in 100 μl of wash buffer andstored at −80 °C for future use.

2.5. SDS-PAGE and Western blotting

SDS-PAGE and Western blotting were carried out as previouslydescribed [16]. The following antibodies were used; mouse mono-clonal anti-MHC (1:100; 21-96-3, Developmental Studies HybridomaBank, The University of Iowa, IA, USA), mouse monoclonal anti-tubulin (1:1000; 12G10, Developmental Studies Hybridoma Bank, TheUniversity of Iowa, IA, USA), mouse monoclonal anti-actin (1:1000;Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA). Membranes weredevelopedwith the AmershamTM ECL PlusWestern Blotting DetectionSystem (GE Healthcare, Buckinghamshire, UK) and scanned using aStormTM 860 Phosphorimager/Fluorimager (Molecular Dynamics Inc.,Sunnyvale, CA, USA). Protein bands on the Western blots werequantified with ImageQuant Version 5.2 (Molecular Dynamics Inc.,Sunnyvale, CA, USA). Means ± SEM were calculated for controls andexperimental conditions.

2.6. Statistical analysis

Statistical analyses were conducted using R 2.11.1 for Windows(http://www.R-project.org). Two-sample and pairwise Wilcoxonrank sum tests were performed on the relevant data.

3. Results and discussion

3.1. Effect of DdEGFL1 on the random motility of mutants of cAMPreceptor-mediated signaling

The effect of DdEGFL1 on random cell motility was tested usingseveral mutants of cAMP receptor-mediated signaling. Dictyosteliumpossesses four cAMP receptors (car). The main receptor carA isexpressed maximally during the aggregation stage of development[26]. Three other receptors (B–D) have also been identified [27]. carCexpression levels rise towards the end of aggregation, while carB andcarD are maximally expressed during the slug and fruiting body stageof development [27]. We chose to focus our analysis on carA and carCsince we were interested in observing the response of cells to

733R. Huber, D.H. O'Day / Cellular Signalling 23 (2011) 731–738

DdEGFL1 during the early stages of development. carC has beenshown to mediate cAMP responses in carA− cells [28]. Therefore thefunctions of both receptors are partially redundant and they bothinteract with the same heterotrimeric G-protein to mediate chemo-taxis. Mutants of carA (DBS0236438; [29]) and carC (DBS0236440)were tested as well as a carA/carC doublemutant (DBS0236899; [30]).The total distance traveled by all three mutant strains in the absenceof DdEGFL1 was less than that of the parental strain DH1 (Fig. 1A).However, DdEGFL1 significantly enhanced the movement of the cellsof all four strains (p-valueb0.03). The carA− cells traveled the farthestin response to DdEGFL1, while the carA−/carC− cells travelled theleast. When expressed as a percent of control movement, the effect ofDdEGFL1 in all three mutant strains was greater than the parentalstrain DH1 (p-valueb0.05; Fig. 1B). DdEGFL1 enhanced themovementof carA−, carC−, and carA−/carC− cells by 456±19%, 341±20%, and844±136% respectively, compared to only 106±6% for DH1 cellssuggesting that DdEGFL1 does not require carA or carC to enhance cellmotility.

Binding of cAMP to its receptor catalyzes the exchange of GDP forGTP on theα2 subunit of the heterotrimeric G protein (Gα2) linked tothe cAMP receptor and causes the dissociation of Gα2 from the Gβ/γsubunit dimer [31]. There is only one Gβ subunit in Dictyostelium. The

Fig. 1. Effect of DdEGFL1 on random cell motility of mutants of cAMP receptor-mediatedsignaling. Vegetative cells were plated on 0.5% agar ± DdEGFL1 (450 μM). The totaldistance migrated (μm) was calculated 4 h after the initial plating. DH1, parental forcarA−, carC−, gpaB− (Gα2−), and gpbA− (Gβ−) cells; carA−, parental for carA−/carC−

cells. (A) Effect of DdEGFL1 on the total distance migrated after 4 h. Data presented asmean distance migrated ± SEM (n≥4). *p-valueb0.03 vs. control (no DdEGFL1)movement. (B) DdEGFL1-enhanced random cell motility expressed as a percentage ofcontrol movement (no DdEGFL1). Data presented as % control ± SEM (n≥4). Apairwise Wilcoxon rank sum test was performed on the data. Bars with specific letters(with asterisks) are statistically distinct from those groups with different letters (e.g.group A is significantly different from groups B, C, D, and E but not AE; p-valueb0.05).

movement of Gα2− cells (gpaB−; DBS0236575; [32]) and Gβ− cells(gpbA−; DBS0236531; [31]) in response to DdEGFL1 was analyzed(Fig. 1A). Gα2− cells not treated with DdEGFL1 migrated the sametotal distance as DH1 cells, whereas Gβ− cells migrated further.DdEGFL1 significantly enhanced the movement of Gα2− and Gβ−

cells (p-valueb0.01; Fig. 1A), however the response to DdEGFL1 wassimilar to DH1 cells (Fig. 1B). Peptide treatment enhanced the totaldistance travelled by Gα2− and Gβ− cells by 73±5% and 111±32%respectively, compared to 106±6% for DH1 cells suggesting thatDdEGFL1 does not require the same heterotrimeric G-protein utilizedin cAMP signaling.

Taken together, our analysis using mutants of cAMP signaling,suggests that both carA and carC function as a brake during cellmovement. A brake mechanism regulated by the novel adhesionreceptor SadA has been proposed in vegetative amoebae [20].Interestingly, this protein also possesses three conserved EGFL repeatsin its ectodomain. The average instantaneous velocity of sadA− cellswas found to be almost two times greater than that of parental cellssuggesting that SadA functions as a brake during cell movement via asignaling pathway that has not yet been identified [20]. Whencomparing the responses of carA− and carC− cells to DdEGFL1, theeffect was greater in carA− cells than in carC− cells suggesting thatcarA is the dominant regulator of the proposed brakemechanism. Thisobservation fits with the notion that carA is the dominant cAMPreceptor during chemotaxis in Dictyostelium. carA−/carC− cellsdisplayed the greatest increase in movement presumably sinceneither receptor was present to function as the brake to regulateand coordinate movement. In both the presence and absence ofDdEGFL1, Gα2− cells displayed similar rates of motility whencompared to parental DH1 cells suggesting that Gα2 it is not part ofthe brakemechanism. Although the enhancedmovement of Gβ− cellsin response to DdEGFL1 was similar to that of the parental DH1 cells,of all the mutants tested in this study, Gβ − cells possessed the fastestrates of movement under both control and DdEGFL1 treatmentconditions. It is possible that these mutant cells were moving at theirmaximum rate therefore not allowing us to observe a significantenhancement by DdEGFL1. Based on this, we can conclude that Gβ islikely involved in the brake mechanism.

3.2. Effect of DdEGFL1 on the random motility of mutants ofRas-mediated signaling

Ras proteins are monomeric GTPases that are activated byguanine nucleotide exchange factors (GEFs) [12]. Members of thisprotein family have been shown to be involved in regulating EGF-induced cell motility and chemotaxis in normal and canceroushuman cells [33,34]. In Dictyostelium, rasC and rasG are expressed athigh levels in growing cells. rasD is expressed at low levels duringgrowth, but expression is induced after the cells initiatemulticellulardevelopment [35]. RasGEFA (encoded by the gefA gene) has recentlybeen shown to exchange GDP for GTP on RasC, but not on other Rasproteins [36]. In the absence of DdEGFL1, rasD− cells (DBS0236860;[37]) and rasG− cells (DBS0236862; [15])migrated further than cellsof the parental strain AX2 after 4 h (Fig. 2A). rasC− cells(DBS0236853; [14]) and rasC−/rasG− cells (DBS0236858; [15])migrated about the same distance as AX2 cells. Of the four strainsdescribed above, rasD− cellsmigrated the furthestwhen treatedwithDdEGFL1 (p-valueb0.03; Fig. 2A). However, when expressed as apercent of control movement, the increased movement was similarto the percent increase observed in AX2 cells (120±10% and 123±11%, respectively; Fig. 2B) suggesting that RasD is not involved inthe pathway that regulates DdEGFL1-enhanced cell motility. Thisconclusion is supported by a previous study that has shown thatRasD most likely regulates photosensory and thermosensoryresponses during the later stages of Dictyostelium development[37]. The effect of DdEGFL1 was less pronounced in rasG− and rasC−/

Fig. 2. Effect of DdEGFL1 on the random cell motility of mutants of Ras-mediatedsignaling. Vegetative cells were plated on 0.5% agar ± DdEGFL1 (450 μM). The totaldistance migrated (μm) was calculated 4 h after the initial plating. (A) DdEGFL1-enhanced random cell motility in various mutants of Ras-mediated signaling. AX2,parental for rasC−, rasD−, and rasG− cells; rasG− parental for rasC−/rasG− cells. Datapresented as mean distance migrated ± SEM (n≥4). *p-valueb0.03 vs. control (noDdEGFL1) movement. (B) DdEGFL1-enhanced random cell motility expressed as apercentage of control movement (no DdEGFL1). Data presented as % control ± SEM(n≥4). See Fig. 1 for statistical details.

Fig. 3. Effect of DdEGFL1 on the random cell motility of RasGEFA and PKB mutant cells.Vegetative cells were plated on 0.5% agar ± DdEGFL1 (450 μM). The total distancemigrated (μm) was calculated 4 h after the initial plating. (A) DdEGFL1-enhancedrandom cell motility in a RasGEFA mutant (gefA−). DH1, parental for gefA− cells. Datapresented as mean distance migrated ± SEM (n≥4). *p-valueb0.03 vs. control (noDdEGFL1) movement. (B) DdEGFL1-enhanced random cell motility expressed as apercentage of control movement (no DdEGFL1). Data presented as % control ± SEM(n≥4). *p-valueb0.05. (C) DdEGFL1-enhanced random cell motility in a PKB mutant(pkbA−). KAX3, parental for pkbA− cells. Data presented as mean distance migrated ±SEM (n≥4). *p-valueb0.03 vs. control (no DdEGFL1) movement. (D) DdEGFL1-enhanced random cell motility expressed as a percentage of control movement (noDdEGFL1). Data presented as % control ± SEM (n≥4). *p-valueb0.05.

734 R. Huber, D.H. O'Day / Cellular Signalling 23 (2011) 731–738

rasG− cells as the percent increases for these cells were 55±3% and79±19% respectively. Although rasC− cells traveled about the samedistance as AX2 cells in the absence of DdEGFL1, treatment withDdEGFL1 resulted in the largest percent increase in distance traveled(195±10%; p-valueb0.05). gefA− cells (DBS0236896; [13]) did nottravel as far as cells of the parental strain DH1, however treatmentwith DdEGFL1 significantly enhanced the movement of gefA− cellsby 170±26% compared to only 106±6% for the DH1 cells (p-valueb0.05; Fig. 3A, B). Together these results indicate that RasGmay be involved in the pathway regulating DdEGFL1-enhanced cellmovement. The ability of DdEGFL1 to still enhancemovement by 55%suggests that there are other components in the pathway thatcompensate for the absence of RasG. Our data also suggest that RasCand RasGEFA may participate in the brake mechanism regulated bycarA and carC (discussed in Section 3.1) since the effect of DdEGFL1on both strains was greater than the parental strain AX2. Thepotential relationship between RasC and RasGEFA is strengthened bya recent study that has shown that RasGEFA exchanges GDP for GTPon RasC, but not on other Ras proteins [36]. Presumably, when RasCand RasGEFA are not present then the cells are free to move. Thisbrake mechanism may help to orient the cells or prevent uncoor-dinated movement. In the rasC−/rasG− double mutant, the effect ofDdEGFL1 is less than rasC− cells, but greater than rasG− cellssuggesting that the dampened effect of DdEGFL1 due to the absence

of RasG is partially restored due to the absence of RasC. The possibleinvolvement of RasC, RasG, and RasGEFA in DdEGFL1-enhanced cellmovement is strengthened by the fact that all three proteins havepreviously been shown to be active during the early stages ofmulticellular development, specifically during cAMP chemotaxis[13–15].

3.3. Effect of DdEGFL1 on the random motility of pkbA− cells

In Dictyostelium, the serine/threonine kinase protein kinase B(PKB; encoded by the pkbA gene) is activated by phosphorylationthrough cAMP signal transduction. The protein is expressed invegetative cells and expression increases during the first few hoursof development [38]. In the absence of DdEGFL1, pkbA− cells(DBS0236784; [38]) did not travel as far after 4 h as the parentalstrain KAX3 (Fig. 3C). However, when plated on agar containingDdEGFL1, themovement of pkbA− cells was significantly enhanced by500±39% compared to only 202±58% for KAX3 cells (p-valueb0.03;Fig. 3C, D). These data suggest that the proposed brake mechanismmay also involve PKB, which has been shown to be phosphorylated inresponse to cAMP and is suggested to be activated through cAMPreceptors via a G-protein linked pathway [14,38]. This phosphoryla-tion has also been shown to be drastically reduced in rasC− cells [14]suggesting that the protein may interact with RasC in our proposedbrake mechanism.

Fig. 4. Effect of extracellular Ca2+ chelation, intracellular Ca2+ release inhibition, andcalmodulin (CaM) inhibition on DdEGFL1-enhanced random cell motility. Vegetativecells were plated on 0.5% agar ± DdEGFL1 (450 μM). The total distance migrated (μm)was calculated 4 h after the initial plating. (A) Effect of extracellular Ca2+ chelation onDdEGFL1-enhanced random cell motility. (B) Effect of intracellular Ca2+ releaseinhibition on DdEGFL1-enhanced random cell motility. (C) Effect of CaM inhibition onDdEGFL1-enhanced random cell motility. Data in all plots presented as mean %DdEGFL1-enhanced motility (control) ± SEM, n=4. See Fig. 1 for statistical details.

735R. Huber, D.H. O'Day / Cellular Signalling 23 (2011) 731–738

3.4. Ca2+ and calmodulin function in DdEGFL1-enhanced random cellmotility

Chemotactic stimulation via cAMP has been reported to increaseCa2+ entry across the plasmamembrane aswell as trigger the release ofCa2+ from intracellular stores, thereby effectively increasing thecytosolic concentration of Ca2+ [4,39]. Previous work has also shownthat Ca2+ uptake through plasma membrane channels is triggered bythe depletion of intracellular stores of Ca2+ [5,6]. Both inter- andintracellular Ca2+ signaling has also been shown to be required for EGFsignaling in tumor cells [40]. To assess the involvement of extracellularCa2+ signaling and/or influx in DdEGFL1-enhanced movement, cellswere treated with CaCl2 or one of the extracellular Ca2+ chelators EGTAor BAPTA. Previous studies have shown that 1 mM EGTA inhibitschemotaxis while 0.1–1 mM Ca2+ stimulates chemotaxis [41]. Exoge-nous Ca2+ (CaCl2; 1 mM) significantly increased DdEGFL1-enhancedrandomcellmotility by 20.9±12.4% (p-valueb0.03; Fig. 4A). Treatmentwith EGTA (1 mM) or BAPTA (1 mM) significantly inhibited DdEGFL1-enhanced cell movement by 41.4±2.6% and 67.2±1.7%, respectively(p-valueb0.03) suggesting that Ca2+ signaling is required for DdEGFL1-enhanced cell movement.

TMB-8 was used to investigate the involvement of intracellularCa2+ signaling in DdEGFL1-enhanced cell movement since treatmentwith this chemical has been shown to inhibit the release of Ca2+ frommembrane-bound stores within the cell [42,43]. TMB-8 significantlyinhibited DdEGFL1-enhanced random cell motility dose-dependentlyat 75 μM and 150 μM by 77.2±3.4% and 92.8±3.0%, respectively(p-valueb0.03; Fig. 4B). Exogenous calcium (CaCl2, 1 mM) partiallyrescued the inhibition at both concentrations (p-valueb0.03).Although TMB-8 has also been shown to inhibit ATP formation andcGMP formation in Dictyostelium [44], we show here that the effect ofTMB-8 on DdEGFL1-enhanced cell movement is primarily due to adefect in intracellular Ca2+mobilization and not altered respiration ormetabolism since the inhibited movement was partially rescued byCa2+ treatment. These results suggest that the exogenous Ca2+ wasable to compensate for the reduced amount of intracellular Ca2+ thatis normally required for DdEGFL1-enhanced cell movement. Inaddition, cells treated with similar concentrations of TMB-8 havepreviously been shown to adhere and extend pseudopods normally[43] and in our assays, cells treated with TMB-8 also displayed normalcell morphology (data not shown). To further verify the involvementof intracellular Ca2+ signaling, cells were also treated with themembrane-permeable Ca2+ chelator BAPTA-AM. Treatment withBAPTA-AM significantly inhibited DdEGFL1-enhanced motility by47.6±2.6% (p-valueb0.03) providing further evidence that intracel-lular Ca2+ signaling is required for the enhanced movement.

CaM is the primary Ca2+ binding protein that is expressed in alleukaryotic cells and is required for chemotaxis in Dictyostelium[7,45]. To investigate the involvement of CaM in DdEGFL1-enhancedcell movement, cells were treated with the CaM specific inhibitorsW7 and calmidazolium chloride. Treatment with W7 or calmidazo-lium chloride significantly inhibited DdEGFL1-enhanced motility by77.2±2.9% and 89.1±1.9%, respectively (p-valueb0.03; Fig. 4C),thus supporting a function for CaM in DdEGFL1-enhanced cellmovement.

3.5. Effect of DdEGFL1 on the random motility of iplA− cells

An inositol-1,4,5-triphosphate (IP3) receptor-like protein (IPL;encoded by the iplA gene) has been identified and characterized inDictyostelium [46]. It has been suggested that IPL acts either as aplasma membrane channel allowing Ca2+ to enter the cell or as amembrane channel of an intracellular Ca2+ store. The iplA transcriptis expressed at low levels during growth and increases during theearly stages of development, reaching peak levels after 9 h. Extracel-lular Ca2+ influx is abolished in iplA− cells, however the release of

stored Ca2+ into the cytoplasm remains intact [47]. In the absence ofDdEGFL1, there was no significant difference in the motility ofparental AX2 cells or mutant iplA− cells (DBS0236260; [46]; Fig. 5A).DdEGFL1 significantly enhanced the movement of both parental andmutant cells by a similar amount (103±9% and 139±25%, respec-tively; p-valueb0.0001; Fig. 5A, B). These observations indicate thatwith regards to Ca2+ signaling, intracellular Ca2+ release and notextracellular Ca2+ influx may be the predominant regulator of

Fig. 5. Effect of DdEGFL1 on the random motility of iplA− cells. Vegetative cells wereplated on 0.5% agar ± DdEGFL1 (450 μM). The total distance migrated (μm) wascalculated 4 h after the initial plating. AX2, parental for iplA− cells. (A) Effect of iplAknockout on DdEGFL1-enhanced random cell motility. Data presented asmean distancemigrated ± SEM (n≥9). *p-valueb0.0001 vs. control (no DdEGFL1) movement.(B) DdEGFL1-enhanced random cell motility expressed as a percentage of controlmovement (no DdEGFL1). Data presented as % control ± SEM (n≥9).

Fig. 6. Effect of PI3K and PLA2 inhibitors on MHC, tubulin, and actin expression duringDdEGFL1 treatment. Cells were starved±DdEGFL1 (450 μM), LY294002 (75 μM; PI3Kinhibitor), and quinacrine (40 μM; PLA2 inhibitor) for 2 h and 4 h. (A) Western blotsprobed with anti-MHC, anti-tubulin, and anti-actin. (B) Quantification of MHC,tubulin, and actin expression in DdEGFL1-treated cells after treatment with LY294002and quinacrine. Protein bands were quantified and plotted as a percentage of theexpression in DdEGFL1-treated cells. Data presented as the mean % control ± SEM(n=4). *p-valueb0.05.

736 R. Huber, D.H. O'Day / Cellular Signalling 23 (2011) 731–738

DdEGFL1 enhanced cell movement. Although DdEGFL1-enhanced cellmovement was enhanced by exogenous Ca2+ and inhibited bychelators of extracellular Ca2+, DdEGFL1 enhanced the movement ofiplA− cells to a similar amount as parental AX2 cells. Since Ca2+

uptake is reduced in iplA− cells, the cells may compensate for thereduced influx of Ca2+ by releasing more stored intracellular Ca2+ tofacilitate the increased cell movement.

3.6. Effect of PI3K and PLA2 inhibition on myosin II heavy chain, tubulin,and actin expression during DdEGFL1 treatment

In a previous study, we showed that DdEGFL1-enhanced cellmotility was abolished in cells treated with both LY294002 (PI3Kinhibitor) and quinacrine (PLA2 inhibitor) suggesting that DdEGFL1-enhanced cell movement could not compensate for PI3K and PLA2inhibition [16] and strengthens the notion that both proteinsmediate pathways that are essential for cell movement in Dictyos-telium [21,22]. In this study we investigated the effect of PI3K andPLA2 inhibition on the expression of MHC, tubulin, and actin(Fig. 6A). Inhibitor treatment did not significantly affect the amountsof tubulin or actin in whole cell lysates of DdEGFL1 treated cells,however treatment with both inhibitors significantly decreased theexpression of MHC after 2 and 4 h in peptide treated cells by 47.8±8.9% and 37.6±11.6%, respectively (p-valueb0.05; Fig. 6B). Pre-sumably this decrease in expression could prevent cell movement bydecreasing the amount of MHC that can be assembled into thecytoskeleton. In contrast, whole cell expression levels of actin werenot affected by inhibitor treatment. Although the PI3K and PLA2pathways have been shown to regulate actin polymerization duringcell movement [21,22], our data suggest that actin expression isregulated independently of the two pathways.

3.7. Effect of DdEGFL1 on myosin II heavy chain and actin levels

Previous studies have shown that the cytoskeleton undergoesdynamic changes during chemotactic stimulation and that theamount of actin in detergent insoluble cytoskeletons increases afterstimulation with cAMP [8,10,48,49]. Since DdEGFL1 increases cellmotility and chemotaxis, we tested its effects on the amount of MHCand actin in whole cell lysates and in cytoskeletal fractions to see if itcould induce similar changes in the cytoskeleton of treated cells. Cellswere starved for 4 h in the presence or absence of DdEGFL1. Starvationand/or DdEGFL1 treatment had no effect on MHC, tubulin, or actin

expression in whole cell lysates (Fig. 7A, B). After 4 h of treatment, thecytoskeleton was isolated and assayed for total amounts of MHC andactin in treated and untreated cells (Fig. 7C). Western blots were alsoprobed for tubulin. The absence of tubulin verified that the isolatedfraction was composed primarily of MHC and actin. After 4 h,DdEGFL1 treatment significantly increased the amount of polymericMHC and actin in the cytoskeleton by 24.1±10.7% and 25.9±2.1%respectively (p-valueb0.03; Fig. 7D) suggesting that DdEGFL1enhances cell movement by increasing the amount of MHC andactin in treated cells. Coupled with our PI3K and PLA2 inhibitoranalysis, these data show that MHC expression and assembly into thecytoskeleton are required for DdEGFL1-enhanced cell movement.

In Dictyostelium, Ca2+ immobilization has been shown to inhibitcAMP-induced actin polymerization and cGMP production during cellaggregation suggesting that Ca2+ signaling is critically important forthe cytoskeletal changes that occur during chemotactic stimulationwith cAMP [10,42,43,50,51]. Furthermore, it has been suggested thatCa2+ signaling may play a dual role by regulating both the initiationand termination of cytoskeletal responses to chemotactic stimulation[10]. Since treatment with TMB-8 and W7 significantly inhibitedDdEGFL1-enhanced cell movement, cells were treated with DdEGFL1and one of TMB-8 or W7 to determine the effect of inhibition on theamount of MHC and actin in whole cell lysates and in the cytoskeleton(Fig. 7A). Actin and tubulin protein expression in whole cell lysateswere not significantly affected by treatment with TMB-8 or W7(Fig. 7B). However, treatment with TMB-8 or W7 decreased theexpression of MHC by 42.8±11.0% and 20.1±11.5%, respectively (p-valueb0.03; Fig. 7B), suggesting that intracellular Ca2+ release andCaM activity are required for MHC expression. Treatment with eitherTMB-8 or W7 prevented the DdEGFL1-mediated increase of MHC andactin in the cytoskeleton (Fig. 7C, D). Interestingly, TMB-8 treatment

Fig. 7. Effect of intracellular Ca2+ release and CaM inhibition on the amounts of MHC, actin, and tubulin in whole cell lysates and in the cytoskeleton of DdEGFL1 treated cells. Cellswere starved ± DdEGFL1 (450 μM)± TMB-8 (75 μM) orW-7 (50 μM) for 4 h at which time the cells were harvested and lysed. Whole cell lysates and cytoskeletal fractions werethen isolated, separated by SDS-PAGE, and analyzed byWestern blotting. (A) Effect of intracellular Ca2+ release and CaM inhibition on the amounts of MHC, actin, and tubulin inwhole cell lysates of DdEGFL1 treated cells. Western blots probed with anti-MHC, anti-actin, and anti-tubulin. Protein bands were quantified and plotted (B). Data presented asmean % control ± SEM, n=6. *p-valueb0.03. (C) Effect of intracellular Ca2+ release and CaM inhibition on the amounts of MHC, actin, and tubulin in the cytoskeleton of DdEGFL1treated cells. Western blots probed with anti-MHC, anti-actin, and anti-tubulin. Protein bands were quantified and plotted (D). Data presented as the mean % control ± SEM,n=6. *p-valueb0.03.

Fig. 8. Proposed model for DdEGFL1-enhanced cell movement and the carA/carC mediated brake. (A) DdEGFL1 functions independently of the cAMP mediated signaling pathway tofine-tune Dictyostelium cell movement by activating a signaling pathway that releases Ca2+ from intracellular stores (e.g. endoplasmic reticulum). Released Ca2+ could then activateCaM ultimately leading to actin polymerization and MHC assembly in the cytoskeleton. RasG may also be involved in facilitating the enhanced movement. (B) Proposed model for abrake mechanism mediated by carA, carC, and Gβ. Our data suggest that RasC and its associated GEF, RasGEFA, interact with PKB to mediate the brake.

737R. Huber, D.H. O'Day / Cellular Signalling 23 (2011) 731–738

738 R. Huber, D.H. O'Day / Cellular Signalling 23 (2011) 731–738

actually decreased the amount of MHC in the cytoskeleton by 20.4±12.2% when compared to cells not treated with DdEGFL1 (Fig. 7D)presumably by inhibiting the expression of the protein (as discussedabove). Together, our data suggest that both intracellular Ca2+ releaseand CaM mediate DdEGFL1-enhanced cell movement by regulatingthe polymerization of actin and the assembly of MHC in thecytoskeleton.

3.8. Proposed model for EGFL repeat enhanced cell movement and thebrake mechanism

An analysis of our results has allowed us to develop a model forEGFL repeat enhanced cell movement in Dictyostelium and the brakemechanism regulated by carA and/or carC (Fig. 8). Our results suggestthat DdEGFL1 functions independently of the cAMP mediatedsignaling pathway to fine-tune Dictyostelium cell movement(Fig. 8A). DdEGFL1 binds to a receptor on the cell surface initiating asignaling cascade that releases Ca2+ from intracellular stores (e.g.endoplasmic reticulum). Released Ca2+ could then activate CaMultimately leading to actin polymerization and MHC assembly in thecytoskeleton. Based on our data, RasG may also be involved inDdEGFL1-enhanced cell movement. During chemotaxis, cAMP bindsto either carA or carC and activates the PI3K and PLA2 mediatedsignaling pathways. Activation results in actin polymerization andMHC assembly in the cytoskeleton allowing the cell to directionallyrespond to cAMP. PI3K, PLA2, intracellular Ca2+ release, and CaM alsoregulate MHC expression. Our data also suggests that carA, carC, andGβ regulate a signaling pathway that functions as a brake during cellmovement (Fig. 8B). Our proposed model suggests that RasC and itsassociated GEF, RasGEFA, interact with PKB to mediate the brake.

4. Conclusions

Since the mechanism underlying the motility-enhancing functionof EGFL repeats has not been studied in Dictyostelium, our focus in thisstudy was to sort out the proteins and signaling events involved in theresponse as well as determine whether the components involved incAMP signal transduction were also important for EGFL repeatsignaling. Since cells cannot move when PI3K and PLA2 areinactivated, the pathways mediated by these proteins must be thedominant regulators of cell movement in Dictyostelium. However, themutant analysis and pharmacological approach we have employed inthis study suggest that EGFL repeat signaling serves as a fine-tuningmechanism for basic cell movement. We have shown that DdEGFL1functions independently of the cAMP-induced chemotactic pathwaymediated by carA and carC. Our data show that intracellular Ca2+

release and CaM are required for DdEGFL1-enhanced cell motility anddemonstrate a link between Ca2+/CaM signaling and cytoskeletondynamics during DdEGFL1-enhanced cell movement. Our data alsoindicate that RasG may be involved in mediating the response toDdEGFL1 stimulation. Finally, our findings suggest that carA and carCmediate a brake mechanism involving Gβ during chemotactic cellmovement. Based on our data, this mechanism also appears to involveRasC, RasGEFA, and PKB.

Acknowledgements

This work was supported by a grant (DHO'D) and scholarship (RH)from the Natural Sciences and Engineering Research Council of

Canada and an Ontario Graduate Scholarship (RH). We would like tothank Dr. Yali Wang (Advanced Syntech Corporation, Mississauga,Ontario, Canada) for providing the DdEGFL1 peptide.

References

[1] S.J. Annesley, P.R. Fisher, Mol. Cell. Biochem. 329 (2009) 73.[2] H. Urushihara, Exp. Anim. 58 (2009) 97.[3] J.S. King, R.H. Insall, Trends Cell Biol. 19 (2009) 523.[4] U. Wick, D. Malchow, G. Gerisch, Cell Biol. Int. Rep. 2 (1978) 71.[5] D.E. Clapman, Cell 80 (1995) 259.[6] M.J. Berridge, Biochem. J. 312 (1995) 1.[7] M.L. Gauthier, D.H. O'Day, Cell. Signal. 13 (2001) 575.[8] S. Yumura, Y. Fukui, Nature 314 (1985) 194.[9] A. Hall, V. Warren, J. Condeelis, Dev. Biol. 136 (1989) 517.

[10] S. Yumura, Cell Struct. Funct. 18 (1993) 379.[11] T.J. Mitchison, L.P. Cramer, Cell 84 (1996) 371.[12] M.S. Boguski, F. McCormick, Nature 366 (1993) 643.[13] R.H. Insall, J. Borleis, P.N. Devreotes, Curr. Biol. 6 (1996) 719.[14] C.J. Lim, G.B. Spiegelman, G. Weeks, G. EMBO J. 20 (2001) 4490.[15] P. Bolourani, G.B. Spiegelman, G. Weeks, Mol. Biol. Cell 17 (2006) 4543.[16] R. Huber, D.H. O'Day, Biochem. Biophys. Res. Commun. 379 (2009) 470.[17] G. Glöckner, L. Eichinger, K. Szafranski, J.A. Pachebat, A.T. Bankier, P.H. Dear, R.

Lehmann, C. Baumgart, G. Parra, J.F. Abril, R. Guigó, K. Kumpf, B. Tunggal, E. Cox,M.A. Quail, M. Platzer, A. Rosenthal, A.A. Noegel, Nature 418 (2002) 79.

[18] A.K. Iyer, T.T. Kien, C.W. Borysenko, M. Cascio, C.J. Camacho, H.C. Blair, I. Bahar, A.Wells, J. Cell. Physiol. 211 (2007) 748.

[19] A.K.V. Iyer, K.T. Tran, L. Griffith, A. Wells, J. Cell. Physiol. 214 (2008) 504.[20] P. Fey, S. Stephens, M.A. Titus, R.L. Chisholm, J. Cell Biol. 159 (2002) 1109.[21] L.F. Chen, M. Iijima, M. Tang, M.A. Landree, Y.E. Huang, Y. Xiong, P.A. Iglesias, P.N.

Devreotes, Dev. Cell 12 (2007) 603.[22] P.J.M. van Haastert, I. Keizer-Gunnink, A. Kortholt, J. Cell Biol. 177 (2007) 809.[23] D.H. O'Day, Can. J. Microbiol. 25 (1979) 1416.[24] D.D. Browning, T. The, D.H. O'Day, Cell. Signal. 7 (1995) 481.[25] D. Reines, M. Clarke, J. Biol. Chem. 260 (1985) 14248.[26] P.S. Klein, T.J. Sun, C.L. Saxe, A.R. Kimmel, R.L. Johnson, P.N. Devreotes, Science 241

(1988) 1467.[27] C.L. Saxe, R. Johnson, P.N. Devreotes, A.R. Kimmel, Dev. Genet. 12 (1991) 6.[28] R.H. Insall, R.D.M. Soede, P. Schapp, P.N. Devreotes, Mol. Biol. Cell 5 (1994) 703.[29] M.J. Caterina, J.L.S. Milne, P.N. Devreotes, J. Biol. Chem. 269 (1994) 1523.[30] J.-Y. Kim, R.D.M. Soede, P. Schaap, R. Valkema, J.A. Borleis, P.J.M. van Haastert, P.N.

Devreotes, D. Hereld, J. Biol. Chem. 272 (1997) 27313.[31] L.Wu, R. Valkema, P.J.M. van Haastert, P.N. Devreotes, J. Cell Biol. 129 (1995) 1667.[32] M.-Y. Chen, P.N. Devreotes, R.E. Gundersen, J. Biol. Chem. 269 (1994) 20925.[33] K. Kato, Y. Ueoka, T. Tamura, J. Nishida, N. Wake, Eur. J. Cancer 34 (1998) 737.[34] J.T. Price, T. Tiganis, A. Agarwal, D. Djakiew, E.W. Thompson, Cancer Res. 59 (1999)

5475.[35] M. Khosla, S.M. Robbins, G.B. Spiegelman, G. Weeks, Mol. Cell. Biol. 10 (1990) 918.[36] H. Kae, A. Kortholt, H. Rehmann, R.H. Insall, P.J.M. van Haastert, G.B. Spiegelman,

G. Weeks, EMBO Rep. 8 (2007) 477.[37] A.Wilkins, M. Khosla, D.J. Fraser, G.B. Spiegelman, P.R. Fisher, G.Weeks, R.H. Insall,

Genes Dev. 14 (2000) 1407.[38] R. Meili, C. Ellsworth, S. Lee, T.B.K. Reddy, H. Ma, R.A. Firtel, EMBO J. 18 (1999)

2092.[39] J. Bumann, B. Wurster, D. Malchow, J. Cell Biol. 98 (1984) 173.[40] J.A. Bryant, R.S. Finn, D.J. Slamon, T.F. Cloughesy, A.C. Charles, Cancer Biol. Ther. 3

(2004) 1243.[41] D. Malchow, R. Böhme, U. Gras, Eur. Biophys. J. 9 (1982) 131.[42] G.N. Europe-Finner, P.C. Newell, FEBS Lett. 171 (1984) 315.[43] M.A. Lydan, D.H. O'Day, J. Cell Sci. 90 (1988) 465.[44] G.N. Europe-Finner, H.S. Tillinghast, S.J. McRobbie, P.C. Newell, J. Cell Sci. 79

(1985) 151.[45] D. Chin, A.R. Means, Trends Cell Biol. 10 (2000) 322.[46] D. Traynor, J.L.S. Milne, R.H. Insall, R.R. Kay, EMBO J. 19 (2000) 4846.[47] R.H. Schaloske, D.F. Lusche, K. Bezares-Roder, K. Happle, D. Malchow, C.

Schlatterer, BMC Cell Biol. 6 (2005) 13.[48] P.C. Newell, Bioessays 5 (1986) 208.[49] S.J. McRobbie, P.C. Newell, J. Cell Sci. 68 (1984) 139.[50] G.N. Europe-Finner, P.C. Newell, J. Cell Sci. 82 (1980) 41.[51] S.J. McRobbie, P.C. Newell, Biochem. Biophys. Res. Commun. 115 (1983) 351.