Matrix Biology 28 (2009) 188–193

Contents lists available at ScienceDirect

Matrix Biology

j ourna l homepage: www.e lsev ie r.com/ locate /matb io

Brief Report

MVL-PLA2, a phospholipase A2 from Macrovipera lebetina transmediterranea venom,inhibits tumor cells adhesion and migration

Amine Bazaa a,⁎, José Luis b, Najet Srairi-Abid a, Olfa Kallech-Ziri a, Raoudha Kessentini-Zouari a,Céline Defilles b, Jean-Claude Lissitzky b, Mohamed El Ayeb a, Naziha Marrakchi a,c

a Laboratoire des Venins et Toxines, Institut Pasteur de Tunis, Tunisiab INSERM UMR 911-CRO2, Aix-Marseille Université, Marseille, Francec Department of Biochemistry Faculty of Medicine of Tunis, Tunisia

⁎ Corresponding author. 13, Place Pasteur BelvedèTel.: +216 71 283 022; fax: +216 71 791 833.

E-mail address: amine.bazaa@gmail.com (A. Bazaa).

0945-053X/$ – see front matter © 2009 Elsevier B.V. Adoi:10.1016/j.matbio.2009.03.007

a b s t r a c t

a r t i c l e i n f o

Article history:Received 29 January 2009Received in revised form 23 March 2009Accepted 26 March 2009

Keywords:Snake venom PLA2Anti-tumorIntegrin inhibitionTumor cellsAdhesionMigration

Here, we report the purification and characterization of an acidic Asp49 phospholipase A2, named MVL-PLA2,with a molecular mass of 13,626.64 Da. The complete MVL-PLA2 cDNA was cloned from Macrovipera lebetinatransmediterranea venom gland cDNA library. MVL-PLA2 possesses 122 amino acid residues, including 14cysteines, and belongs to group II snake venom phospholipase A2 enzymes. MVL-PLA2 was not cytotoxic up to2 μM and completely abolished cell adhesion and migration of various human tumor cells. Chemical modi-fication with p-bromophenacyl bromide abolished the enzymatic activity of MVL-PLA2 without affecting itsanti-tumor effect, suggesting the presence of ‘pharmacological sites’ distinct from the catalytic site in snakevenom phospholipase A2. We demonstrated for the first time that the anti-tumor effect of MVL-PLA2 wasmediated byα5β1 andαv-containing integrins. This findingmay serve as starting point for structure–functionrelationship studies leading to design a new generation of specific anti-cancer drugs.

© 2009 Elsevier B.V. All rights reserved.

1. Introduction

Snake venoms are complex mixtures of molecules possessingvarious biological functions (Koh et al., 2006). Phospholipase A2enzymes (EC 3.1.1.4) are among the best-characterized components ofsnake venom. Based on their source, amino acid sequences anddisulfide bond patterns, snake venom PLA2s (svPLA2s) are classifiedinto group I and group II PLA2 (Six and Dennis, 2000). They areinteresting proteins, containing about 120–130 amino acids, that arecross-linked by seven disulfide bonds. In addition to their enzymaticactivity for the cleavage of ester bonds at the sn-2 position of 1,2-diacyl-3-sn-phosphoglycerides, svPLA2s display several pharmacolo-gical effects, including pre- or post-synaptic neurotoxicity, cardio-toxicity, myotoxicity and platelet aggregation modulation (see Ref.Kini (2003) and Valentin and Lambeau (2000) for a review).

The ability of these svPLA2 to exhibit such a diverse spectrum ofactivities is intriguing since they share significant structural homology(40–99% identity in their amino acid sequences) and since these acti-vities originate from a single conserved scaffold (Murakami and Arni,2003). There is no clear correlation between catalysis and pharmaco-

re BP74, 1002 Tunis, Tunisia.

ll rights reserved.

logical activity. Indeed, natural catalytically inactive (Lys49-PLA2) andthe chemically inactivated svPLA2 enzymes conserve some of theirbiological activities (Chioato et al., 2007; Chioato and Ward, 2003;Soares and Giglio, 2003). A growing body of evidence suggests thatthese activities may be mediated by interaction between svPLA2s andacceptors for endogenous PLA2 on the plasmamembrane of the targetcell through ‘pharmacological sites’ distinct from the catalytic site insvPLA2 (for a review, see Ref. Kini (2003) and Valentin and Lambeau(2000)).

Recently, an anti-tumoral activity was described for BthA-I-PLA2, anacidic PLA2 isolated from Bothrops jararacussu venom (Roberto et al.,2004) and for RVV-7, a basic 7 kDa toxin from Russell's viper venom(Maity et al., 2007). These two proteins exert an indirect anti-tumoractivity through their cytotoxicity. Furthermore, synthetic peptidesderived from the C-terminal of Lys49-PLA2 homologues from Agkistro-don piscivorus andBothrops asper reduced the tumormass by 36% (Arayaand Lomonte, 2007). Despite these very preliminary studies, little isknownabout the anti-tumor activityof the svPLA2 and theirmechanismof action on cancer cells remains elusive.

In this work, we report the purification and structural character-ization of MVL-PLA2, an acidic PLA2 from Macrovipera lebetinatransmediterranea venom. MVL-PLA2 displayed an inhibitory effect,independent from its catalytic activity, on tumor cell adhesion andmigration. This effect ofMVL-PLA2wasmediated via specific inhibitionof integrinsα5β1,αvβ3 andαvβ6. This is thus thefirst report revealing

Fig.1. cDNA and deduced amino acid sequences of MVL-PLA2 fromMacrovipera lebetina. The signal peptide consisting of 16 amino acid residues is showed in grey box. The stop codonis indicated by an asterisk. The sequence of the MVL-PLA2 was confirmed by the direct sequencing of N-terminal and tryptic peptides (boxes).

Fig. 2. Effect of MVL-PLA2 on tumor cells proliferation. IGR39 (A) and HT1080 (B) cellswere cultured for the indicated periods of time in the absence ( ) or in the presence( ) of 2 μMMVL-PLA2. Growth was quantified by staining cells with 0.1% crystal violet,solubilization with 1% SDS and measure of absorbance at 600 nm. The results (±S.D.)are from a representative experiment of two performed in triplicate.

189A. Bazaa et al. / Matrix Biology 28 (2009) 188–193

that integrins constitute specific targets for themechanism of action ofsvPLA2 on tumor cells.

2. Results and discussion

2.1. Purification and cloning of MVL-PLA2

The research of anti-tumor components from snake venom hasprovided excellent research tools used to develop novel therapeuticagents. In this study we have isolated and characterized a novel PLA2fromM. lebetina transmediterranea venom. This PLA2, designated MVL-PLA2, was purified to homogeneity by two chromatographic steps, in-cluding gelfiltration and reverse-phase chromatography (SupplementalFig. S1A and B). According to our previous venom profiling, MVL-PLA2seems to be the unique PLA2 in this venom (Bazaa et al., 2005). Thehomogeneity ofMVL-PLA2was verified by reverse-phase HPLC on a C18column (Supplemental Fig. S1C), SDS-PAGE analysis (data not shown),and MALDI-TOF mass spectrometry, that gave a molecular mass of13626.64±12Da (Supplemental Fig. S1D). Aminoacid sequences of the18 N-terminal residues and five other trypsin digestion peptides weredetermined by Edman degradation (Fig. 1). The overall sequence wasthen determined by a reverse biology strategy. The nucleotide sequenceof the 538 bp cDNA encoding MVL-PLA2 was deposited in the GenBankunder the accession number FM202092. The deduced sequence of theopen reading frame showed a highly conserved 16-amino acid signalpeptide followed by the 122 amino acid residues of the mature MVL-PLA2 (Fig. 1). All the peptide sequences obtained by Edman degra-dation matched the translated cDNA sequence. The primary sequenceanalysis showed that MVL-PLA2 contains all the D49-PLA2s con-served residues (Y24-G-C-Y-C-G-L-G-G32) involved in Ca2+ binding inthe active site (Murakami and Arni, 2003). The theoretical molecularmass (13,638.43 Da) is in accordance with that determined by MALDI-TOF, assuming all cysteines are involved in disulfide bonds. Based onits amino acid sequence, an isoelectric point of 4.69 was calculated.The multiple sequence alignment, in particular the arrangement ofthe cysteines, indicates that MVL-PLA2 belongs to group II PLA2(Supplemental Fig. 2). Its sequence presented higher similarities withVpaPLA2 (99%) and VP7 (98%) isolated from Vipera palaestinae (Krizajet al., 1996). Homologies of 90% and 88% were obtained with Am-modytin 1 C from Vipera aspis aspis and Ammodytin 1 D from Viperaammodytes meridionalis, respectively.

2.2. MVL-PLA2 inhibits cell adhesion and migration of tumor cells

Snake venom PLA2s present a wide range of pharmacological ef-fects (see Ref. Kini (2003) for a review), including cytotoxicity ontumor cells (Araya and Lomonte, 2007; Maity et al., 2007; Robertoet al., 2004). In the case of MVL-PLA2, concentrations up to 2 µMduring 4 days did not induce detectable cytotoxicity on human cell

190 A. Bazaa et al. / Matrix Biology 28 (2009) 188–193

lines IGR39 (melanoma) and HT1080 (fibrosarcoma) (Fig. 2). How-ever, as shown in Fig. 3A, MVL-PLA2 completely blocked the adhesionof IGR39 cells to fibrinogen and fibronectin, while attachment tolaminin-1 and vitronectin was only slightly affected and no effectcould be observed on type I collagen. Moreover, no inhibition could beobserved on the integrin-independent substratum, poly- L -lysine,suggesting that the effect of MVL-PLA2 involves the integrin family ofadhesion receptors. The effect of MVL-PLA2 on IGR39 attachment tofibrinogen and fibronectin was dose-dependent, with a half-maximalinhibition (IC50) of 35 nM and 100 nM, respectively (Fig. 3B).

In order to colonize new organ sites, tumor cells have to access thelymphatic or blood vessel system, disseminate, extravasate and invadethe new organ parenchyma. These events require cell adhesion, butalso migration. We thus tested the ability of MVL-PLA2 to inhibit

Fig. 3. MVL-PLA2 inhibits tumor cell adhesion and migration. (A) Melanoma cells IGR39 weradded to 96-well microtiter plates coated with 10 μg/ml poly- L -lysine (PL), type I collagen (ml laminin-1 (Ln-1) and allowed to adhere for 1 h at 37 °C. After washing, adherent cells wereData shown aremeans (±SD) from 2 experiments performed in triplicate. They are expressedconcentrations of MVL-PLA2, adhesion of IGR39 cells to fibrinogen ( ) or fibronectin (performed in triplicate. (C) Upon treatment for 30 min at room temperature with MVL-Pmembranes precoated with 50 µg/ml fibrinogen (Fg), 10 µg/ml fibronectin (Fn) or type I costained with crystal violet. Membranes shown are from a representative experiment out offibronectin ( ). Data shown (±S.D.), from one experiment representative of 5 performedMALDI-TOF mass spectrometry analysis of MVL-PLA2 treated in the absence (left) or in the pPLA2 on IGR39 cell adhesion (F) and HT1080 cell migration (G). Data shown are means (±adhesion in the absence of peptide (control).

integrin-dependent migration of HT1080 cells, characterized by theirgreat invasive potential in vitro (Okada et al., 1994), by using hap-totaxis assays towards attractive proteins in modified Boyden cham-bers. As shown in Fig. 3C, MVL-PLA2 abolished HT1080 cells migrationtowards fibronectin and fibrinogen. This inhibition was also dose-dependent with IC50 of 10 nM and 110 nM for fibronectin andfibrinogen, respectively (Fig. 3D). To our knowledge, such inhibitoryactivities of secreted PLA2 on cell adhesion and migration have notbeen reported yet. Other studies have showed instead an increasedmotility of epithelial cells and an increase invasion of fibroblasts inresponse to porcine pancreatic and Naja naja PLA2 (Kundu andMukherjee, 1997; Minami et al., 1997).

To know whether PLA2 activity was required for the biologicaleffects of MVL-PLA2, its catalytic site was blocked by alkylation using

e preincubated with 1.7 μMMVL-PLA2 for 30 min at room temperature. Cells were thenColl I), fibronectin (Fn) or vitronectin (Vn), with 50 μg/ml fibrinogen (Fg) or with 2 μg/stainedwith crystal violet, solubilized by SDS and absorbancewasmeasured at 600 nm.as a percentage of adhesion in the absence of peptide. (B) Upon treatment with various

) was measured as described above. Data shown are means (±SD) from 4 experimentsLA2, HT1080 cell motility was measured in a modified Boyden chamber using porousllagen (Coll I). After 5 h at 37 °C, cells that migrated to the underside of the filter were3 to 5 performed. (D) Dose–effect of MVL-PLA2 on cell migration to fibrinogen ( ) orin triplicate, are expressed as a percentage of migration in the absence of peptide. (E)resence (right) of alkylating agent BPB. Effect of active (treated) or BPB-alkylated MVL-SD) from 2 experiments performed in triplicate. They are expressed as a percentage of

191A. Bazaa et al. / Matrix Biology 28 (2009) 188–193

p-bromophenacyl bromide (BPB) (Andriao-Escarso et al., 2002). Massspectra analysis of alkylated vs treated MVL-PLA2 (processed in thesame conditions but without BPB) revealed incorporation of one toone ratio (Fig. 3E). Moreover, the specific activity of treated MVL-PLA2, measured at 37 °C with pH-stat at pH 7, was 63.51±2.9 U/mg,whereas BPB-alkylated enzyme did not displayed any residual activity.All this demonstrated that alkylation procedure was successful withno evidences of non-inactivated enzyme. As illustrated in Fig. 3F andG, catalytically inactivated MVL-PLA2 was still able to inhibit bothadhesion of IGR39 cells to fibrinogen and migration of HT1080 cells tofibrinogen or fibronectin at the same rate than the active enzyme.Altogether these data demonstrated a clear dissociation of the anti-tumor effect of MVL-PLA2 and its catalytic activity. Furthermore arecent investigation shows that two cationic synthetic peptides,derived from the C-terminal region of Lys49-PLA2 homologues fromA. piscivorus and B. asper venoms caused in vivo a tumor mass re-duction by 36% (Araya and Lomonte, 2007). According to these ob-servations, we could suggest the presence of ‘pharmacological sites’distinct from the catalytic site in MVL-PLA2 responsible for tumor cellinhibition. More investigations should be carried out to identify thesepharmacological sites.

2.3. MVL-PLA2 effects are mediated by α5β1 and αv-containing integrins

In this work we have shown that MVL-PLA2 inhibits efficientlytumor cells adhesion and migration. Furthermore tumor cells are notonly able to adhere but also to spread on MVL-PLA2 used as matrix(data not shown). Since integrins are key actors during cell adhesion,

Fig. 4. MVL-PLA2 inhibits α5β1 and αv-containing integrins. (A) IGR39 cells were preincuantibodies raised against the indicated integrin subunits as described in the experimental secexpressed as a percentage of adhesion in the absence of peptide. (B) Adhesion assays were petype I collagen), α5β1 (K562/fibronectin), αvβ3 (HT29-D4/β3/fibrinogen), αvβ5 (HT29-Dadhesion, cells were preincubated without (open bar) or with 1 μM MVL-PLA2 (shade brepresentative of three performed in triplicate. (C) Cell adhesion through α5β1, αvβ3 or αpreincubated or not (control) with 1 mM GRGDSP peptide (RGD), were tested for adhesioexperiment representative of 3 performed in triplicate.

migration and spreading process, we postulated that MVL-PLA2 couldaffect the function of these adhesion receptors. Moreover, as IGR39cells attach to fibrinogen and fibronectin through αvβ3 and α5β1integrins respectively (not shown), at least these two integrins couldbe involved. Thus, we hypothesized that MVL-PLA2 could affect thefunction of this family of adhesion receptors. We therefore first per-formed adhesion assays to a panel of immobilized antibodies raisedagainst several integrin subunits used as matrix. As illustrated inFig. 4A, IGR39 cells preincubated with MVL-PLA2 were still able toadhere to antibodies against α2, α3, α6 and β1 subunits. However,adhesion of MVL-PLA2-pretreated cells was dramatically reduced(over 80%) in the case of anti-α5β1 antibody and completely abo-lished for anti-αv and anti-αvβ3 antibodies.

To further identify the targeted integrins, we next checked theMVL-PLA2 effect on various cell/ECM protein pairs involving uniqueintegrins. The MVL-PLA2 was not able to modify cell adhesionthrough α2β1 or α6β4 integrins, but completely blocked theadhesive function of α5β1. Interestingly, MVL-PLA2 was able todiscriminate between αv-containing integrins, since αvβ3 andαvβ6 were inhibited but not αvβ5 (Fig. 4B). As shown in Fig. 4C,MVL-PLA2 displayed a better efficiency in inhibiting the αvβ3integrin (IC50≈30 nM) compared to αvβ6 (IC50≈100 nM) andα5β1 (IC50≈600 nM). Since α5β1 and αv-containing integrins areRGD-dependent, we have tested RGD-peptides on the adhesion ofIGR39 cells to immobilized MVL-PLA2. As shown in Fig. 4D, theadhesion of cells treated with RGD-peptides was decreased by about90%. This result clearly demonstrates that the interaction betweenMVL-PLA2 and integrins involves a RGD-like sequence. The amino

bated with 1 μM MVL-PLA2 and tested for adhesion on microtiter plates coated withtion. Data shown are means (±SD) from 2 experiments performed in triplicate and arerformedwith various cell/ECM protein pairs involving unique integrins:α2β1 (HT1080/4/vitronectin), αvβ6 (HT29-D4/fibronectin) and α6β4 (HT29-D4/laminin-1). Beforear) for 30 min at room temperature. Data shown (±SD) are from one experimentvβ6 was measured by using increasing concentrations of MVL-PLA2. (D) IGR39 cells,n on microtiter plates coated with 1 µM MVL-PLA2. Data shown (±SD) are from one

192 A. Bazaa et al. / Matrix Biology 28 (2009) 188–193

acid sequence of MVL-PLA2 does not contain any RGD sequence.However, a NGD similar motif is present which may be responsiblefor the inhibition of integrin function. Further studies must becarried out to check this hypothesis.

Snake venom contains pharmacological components able to dis-rupt cell–ECM interactions. The most known molecules are disin-tegrins (for a review, see Ref. McLane et al. (2004)) and C-type lectinproteins (CLPs) (Eble and Tuckwell, 2003; Horii et al., 2004; Pilorgetet al., 2007; Sarray et al., 2007). Here, we extend this observation toanother class of snake venom proteins, the PLA2s. To our knowledge,MVL-PLA2 is thus the first example of PLA2 reported to inhibit theintegrin receptors of tumor cells.

In conclusion, we have isolated and characterized for the first timean acidic Asp49-PLA2 with a potent anti-tumor activity. This venomprotein is able to prevent adhesion and migration of tumor cellsthrough α5β1 and αv integrins. It is noteworthy that this effect isindependent of the catalytic activity of the enzyme. It could thus beinteresting to check the effect of MVL-PLA2 on in vivo tumorigenesisand to realize structural and functional studies to clarify how MVL-PLA2 inhibits integrin receptors. This finding may serve as startingpoint for structure–function relationship studies leading to design anew generation of anti-cancer drugs.

3. Experimental procedures

3.1. Materials

Dulbecco's modified Eagle's medium (DMEM), RPMI 1640mediumand fetal calf serum (FCS) were purchased from Lonza (Walkersville,MD). Penicillin and streptomycinwere purchased from GIBCO (Cergy-Pontoise, France). The p-bromophenacyl bromide and human fibrino-gen and laminin-1were from Sigma (Mannheim, Germany). Rat type Icollagen was from Upstate (Lake Placid, NY) and human fibronectinfrom Chemicon (Temecula, CA). The RGDSP peptide was purchasedfrom Bachem. Human vitronectin was purified according to Yatohgoet al. (1988). Rat monoclonal antibody (mAb) 69.6.5 against αvintegrinwas produced as previously described (Lehmann et al., 1994).MousemAbs P1D6 (anti-α5) and LM609 (anti-αvβ3) were purchasedfrom Chemicon. Mouse mAbs Gi9 (anti- α2β1), C3VLA3 (anti-α3),Lia1/2 (anti-β1), SZ21 (anti-β3) and rat mAb GoH3 (anti-α6) werefrom Immunotech (Marseille, France). Rabbit anti-rat was purchasedfrom Sigma.

3.2. Determination of protein molecular mass and protein sequencing

The homogeneity and the apparent mass of the purified PLA2 weredetermined by SDS-PAGE and confirmed by mass spectrometryanalysis. The exact molecular mass of native MVL-PLA2 was deter-mined byMALDI-TOFMS using an Applied Biosystems Voyager-DE Promass spectrometer operated in linear mode as previously described(Bazaa et al., 2005). For peptides generation and protein sequencing,MVL-PLA2 was reduced and alkylated as previously described (Bazaaet al., 2005). The mixture was immediately desalted by reverse-phaseHPLC on a C18 column. Lyophilized pyridylethylated protein was thensubjected to enzymatic digestion with trypsin as described (Pawlaket al., 2006) and the sequence of digestion peptides was determinedby Edman degradation.

3.3. Chemical modification

Modification of His48 with p-bromophenacyl bromide (BPB) wascarried out as previously described (Andriao-Escarso et al., 2002).Excess reagent was usually removed from the acidic PLA2 prepara-tions by ultrafiltration through an Amicon YM3 membrane, followedby lyophilization. Control experiments were performed in parallel bytreating PLA2 in the same conditions in the absence of BPB.

3.4. Total RNA isolation and Reverse Transcription (RT)-PCR

Total RNA was extracted from pooled venom glands of M. lebetinatransmediterranea and the first strand of cDNA was reverse-tran-scribed as described in a previous study (Sanz et al., 2006) from1 µg oftotal venom RNA using a standard method and the following modifiedoligo-dT 5′-CCAGTGAGCAGAGTGACGAGGACTCGAGCTCAAGCTT16-3′(Qt) purchased from Sigma-Genosys. The svPLA2-coding DNA wasamplified by PCR using venom cDNAs as template and the followingpairs of primers. (i) Forward primer, 5′-TCTGGATTCAGGAGGATGAGG-3′, which was designed according to the highly conserved cDNAregions of the group II PLA2s from snake venoms (Chen et al., 2004),and reverse primer, 5′-CCAGTGAGCAGAGTGACG-3′, based on theadopter sequence (the first 18 nucleotides of the Qt primer used in RT)in the cDNA library. The PCR protocol included an initial denaturationand activation of HotStartTaq polymerase at 94 °C for 10 min and32 cycles of three-step PCR (94 °C for 1 min, 57 °C for 1 min, and 72 °Cfor 1 min) and a final extension at 72 °C for 10 min to ensure that allthe products were double stranded. A 600 bp fragment was speci-fically amplified, as shown by 1% agarose gel electrophoresis.

3.5. Cloning and sequencing of RT-PCR products

The amplified fragment was purified using the PCR purification kit(Qiagen) and cloned in a pGEM-T vector (Promega) whichwas used totransform E. coli DH5α competent cells (Invitrogen). Positive clones(blue/white colony screening), selected by growing the transformedcells in LB (Luria–Bertani) broth containing 100 μg/ml ampicillin,were confirmed by PCR-amplification using the above primers. Thesequence of the inserts was subjected to sequencing on an AppliedBiosystems model 377 DNA sequencing system in both directionsusing T7 and SP6 sequencing primers.

3.6. Cell culture

The human cell lines derived from fibrosarcoma (HT1080), mela-noma (IGR39), colonic adenocarcinoma (HT29-D4) and theβ3 integrin-expressing HT29-D4 (HT29-D4/β3) were routinely cultured in DMEMcontaining 10% FCS. Human leukemia (K562) cells were cultured inRPMI 1640mediumcontaining10% FCS. All cell linesweremaintained at37 °C in 5% CO2.

3.7. Cell adhesion and migration assays

Adhesion assays were performed as previously described (Berthetet al., 2000). Briefly, cells in single cell suspension were added to 96-well plates coated with purified extracellular matrix (ECM) proteinsand allowed to adhere to the substrata for 1 h (HT1080 and IGR39cells) or 2 h (HT29-D4, HT29-D4/β3 and K562 cells) at 37 °C. Ad-hesion of K562 cells was performed in the presence of 1 mM MnCl2and 100 nM phorbol 12-myristate 13-acetate in order to activateα5β1integrin. After washing, adherent cells were fixed, stained with 0.1%crystal violet and lysed with 1% SDS. Absorbance was then measuredat 600 nm. For adhesion assays on antibodies, 96-well plates werecoated with 50 µl of rabbit anti-rat IgG (50 µg/ml), overnight at 4 °C.Wells were washed once with PBS and 50 µl of anti-integrin blockingantibodies (10 µg/ml) were added for 5 h at 37 °C. Then, wells wereblocked with a solution of PBS/0.5% BSA and adhesion assay wascontinued as above.

In vitro cell migration assays were performed in modified Boydenchambers (NeuroProbe Inc., Bethesda, MD) with porous membranesprecoated with 10 µg/ml of fibronectin or type I collagen or with50 µg/ml fibrinogen for 5 h at 37 °C as previously described (Sarrayet al., 2007), except that cells were stained with 0.1% crystal violet.Cell migration was then quantified by measuring the absorbance at600 nm.

193A. Bazaa et al. / Matrix Biology 28 (2009) 188–193

3.8. Statistical analysis

Results are presented as the mean value±standard deviation (SD)obtained with the indicated number of samples. Statistic significanceof results was evaluated using Student's t-test. A p valueb0.05 wasconsidered to indicate significance.

Acknowledgments

This work was supported in part by IPT budget and MRST/IPT/LVT-LR005 SP 05 grant, the CMCU (Comité Mixte de CoopérationUniversitaire France-Tunisie), the Conseil Régional and the Cancéro-pôle of Provence-Alpes-Côte d'Azur.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.matbio.2009.03.007.

References

Andriao-Escarso, S.H., Soares, A.M., Fontes, M.R., Fuly, A.L., Correa, F.M., Rosa, J.C.,Greene, L.J., Giglio, J.R., 2002. Structural and functional characterization of an acidicplatelet aggregation inhibitor and hypotensive phospholipase A(2) from Bothropsjararacussu snake venom. Biochem. Pharmacol. 64, 723–732.

Araya, C., Lomonte, B., 2007. Antitumor effects of cationic synthetic peptides derived fromLys49 phospholipase A2 homologues of snake venoms. Cell Biol. Int. 31, 263–268.

Bazaa, A., Marrakchi, N., El Ayeb, M., Sanz, L., Calvete, J.J., 2005. Snake venomics:comparative analysis of the venom proteomes of the Tunisian snakes Cerastescerastes, Cerastes vipera and Macrovipera lebetina. Proteomics 5, 4223–4235.

Berthet, V., Rigot, V., Champion, S., Secchi, J., Fouchier, F., Marvaldi, J., Luis, J., 2000. Roleof endoproteolytic processing in the adhesive and signaling functions of αvβ5integrin. J. Biol. Chem. 275, 33308–33313.

Chen, Y.H., Wang, Y.M., Hseu, M.J., Tsai, I.H., 2004. Molecular evolution and structure–function relationships of crotoxin-like and asparagine-6-containing phospholi-pases A2 in pit viper venoms. Biochem. J. 381, 25–34.

Chioato, L., Aragao, E.A., Lopes Ferreira, T., Medeiros, A.I., Faccioli, L.H., Ward, R.J., 2007.Mapping of the structural determinants of artificial and biological membranedamaging activities of a Lys49 phospholipase A2 by scanning alanine mutagenesis.Biochim. Biophys. Acta 1768, 1247–1257.

Chioato, L., Ward, R.J., 2003. Mapping structural determinants of biological activities insnake venom phospholipases A2 by sequence analysis and site directed mutagen-esis. Toxicon 42, 869–883.

Eble, J.A., Tuckwell, D.S., 2003. The alpha2beta1 integrin inhibitor rhodocetin binds tothe A-domain of the integrin alpha2 subunit proximal to the collagen-binding site.Biochem. J. 376, 77–85.

Horii, K., Okuda, D., Morita, T., Mizuno, H., 2004. Crystal structure of EMS16 in complexwith the integrin alpha2-I domain. J. Mol. Biol. 341, 519–527.

Kini, R.M., 2003. Excitement ahead: structure, function andmechanism of snake venomphospholipase A2 enzymes. Toxicon 42, 827–840.

Koh, D.C., Armugam, A., Jeyaseelan, K., 2006. Snake venom components and theirapplications in biomedicine. Cell. Mol. Life Sci. 63, 3030–3041.

Krizaj, I., Bdolah, A., Gubensek, F., Bencina, P., Pungercar, J., 1996. Protein and cDNAstructures of an acidic phospholipase A2, the enzymatic part of an unusual, two-component toxin from Vipera palaestinae. Biochem. Biophys. Res. Commun. 227,374–379.

Kundu, G.C., Mukherjee, A.B., 1997. Evidence that porcine pancreatic phospholipase A2via its high affinity receptor stimulates extracellular matrix invasion by normal andcancer cells. J. Biol. Chem. 272, 2346–2353.

Lehmann, M., Rabenandrasana, C., Tamura, R., Lissitzky, J.C., Quaranta, V., Pichon, J.,Marvaldi, J., 1994. A monoclonal antibody inhibits adhesion to fibronectin andvitronectin of a colon carcinoma cell line and recognizes the integrins alpha v beta3, alpha v beta 5, and alpha v beta 6. Cancer Res. 54, 2102–2107.

Maity, G., Mandal, S., Chatterjee, A., Bhattacharyya, D., 2007. Purification and characteriza-tion of a low molecular weight multifunctional cytotoxic phospholipase A2 fromRussell's viper venom. J. Chromatogr. BAnalyt. Technol. Biomed. Life Sci. 845, 232–243.

McLane, M.A., Sanchez, E.E., Wong, A., Paquette-Straub, C., Perez, J.C., 2004. Disintegrins.Curr. Drug Targets Cardiovasc. Haematol. Disord. 4, 327–355.

Minami, T., Tojo, H., Zushi, S., Shinomura, Y., Matsuzawa, Y., 1997. Phospholipase A2stimulates rat gastric epithelial cell line (RGM-1)migration. Inflamm.Res. 46,103–107.

Murakami, M.T., Arni, R.K., 2003. A structure based model for liposome disruption andthe role of catalytic activity in myotoxic phospholipase A2s. Toxicon 42, 903–913.

Okada, T., Okuno, H.,Mitsui, Y.,1994. A novel in vitro assay system for transendothelial tumorcell invasion: significance of E-selectin and alpha 3 integrin in the transendothelialinvasion by HT1080 fibrosarcoma cells. Clin. Exp. Metastasis 12, 305–314.

Pawlak, J., Mackessy, S.P., Fry, B.G., Bhatia, M., Mourier, G., Fruchart-Gaillard, C., Servent,D., Menez, R., Stura, E., Menez, A., Kini, R.M., 2006. Denmotoxin, a three-finger toxinfrom the colubrid snake Boiga dendrophila (Mangrove Catsnake) with bird-specificactivity. J. Biol. Chem. 281, 29030–29041.

Pilorget, A., Conesa, M., Sarray, S., Michaud-Levesque, J., Daoud, S., Kim, K.S., Demeule,M., Marvaldi, J., El Ayeb, M., Marrakchi, N., Beliveau, R., Luis, J., 2007. Lebectin, aMacrovipera lebetina venom-derived C-type lectin, inhibits angiogenesis both invitro and in vivo. J. Cell. Physiol. 211, 307–315.

Roberto, P.G., Kashima, S., Marcussi, S., Pereira, J.O., Astolfi-Filho, S., Nomizo, A., Giglio, J.R.,Fontes, M.R., Soares, A.M., Franca, S.C., 2004. Cloning and identification of a completecDNA coding for a bactericidal and antitumoral acidic phospholipase A2 from Bothropsjararacussu venom. Protein J. 23, 273–285.

Sanz, L., Bazaa, A., Marrakchi, N., Perez, A., Chenik, M., Bel Lasfer, Z., El Ayeb, M., Calvete,J.J., 2006. Molecular cloning of disintegrins from Cerastes vipera and Macroviperalebetina transmediterranea venom gland cDNA libraries: insight into the evolutionof the snake venom integrin-inhibition system. Biochem. J. 395, 385–392.

Sarray, S., Srairi, N., Hatmi, M., Luis, J., Louzir, H., Regaya, I., Slema, H., Marvaldi, J., ElAyeb, M., Marrakchi, N., 2003. Lebecetin, a potent antiplatelet C-type lectin fromMacrovipera lebetina venom. Biochim. Biophys. Acta 1651, 30–40.

Sarray, S., Delamarre, E.,Marvaldi, J., El Ayeb,M.,Marrakchi, N., Luis, J., 2007. Lebectin andlebecetin, two C-type lectins from snake venom, inhibit α5β1 and αv-containingintegrins. Matrix Biol. 26, 306–313.

Six, D.A., Dennis, E.A., 2000. The expanding superfamily of phospholipase A(2)enzymes: classification and characterization. Biochim. Biophys. Acta 1488, 1–19.

Soares, A.M., Giglio, J.R., 2003. Chemical modifications of phospholipases A2 from snakevenoms: effects on catalytic and pharmacological properties. Toxicon 42, 855–868.

Valentin, E., Lambeau, G., 2000. Increasing molecular diversity of secreted phospho-lipases A(2) and their receptors and binding proteins. Biochim. Biophys. Acta 1488,59–70.

Yatohgo, T., Izumi,M., Kashiwagi, H., Hayashi,M.,1988. Novel purificationof vitronectin fromhuman plasma by heparin affinity chromatography. Cell Struct. Funct. 13, 281–292.