Purification of a phospholipase A(2) from Daboia russelii siamensis venom with anticancer effects.

Venom phospholipases A2 (PLA(2)) are associated with neurotoxic, myotoxic, cardiotoxic, platelet aggregation, and edema activities. A PLA(2) (Drs-PLA(2)) was purified from Daboia russelii siamensis venom by a two-step purification procedure consisting of size-exclusion, followed by anion exchange high performance liquid chromatography (HPLC). The molecular weight of the Drs-PLA(2) was 13,679Da, which was determined by MALDI-TOF mass spectrometry. Its N-terminal amino acid sequence was homologous to basic PLA(2)s of viperid snake venoms. The Drs-PLA(2) had indirect hemolytic and anticoagulant activities, cytotoxic activity with a CC(50) of 65.8nM, and inhibited SK-MEL-28 cell migration with an IC(50) of 25.6nM. In addition, the Drs-PLA(2) inhibited the colonization of B16F10 cells in lungs of BALB/c mice by ∼65%.

INTRODUCTION

The superfamily of phospholipase A2 enzymes have been classified as 15 groups and many subgroups that include five distinct types of enzymes, namely the secreted PLA2 (sPLA2), the cytosolic PLA2 (cPLA2), the Ca2+ independent PLA2s (iPLA2), the platelet-activating factor acetylhydrolases (PAF-AH), lysosomal PLA2s, and a recently identified adipose-specific PLA2 (Duncan et al, 2008; Burke and Dennis 2009a; Ramar et al, 2010). Snake venom is one of the most abundant sources of secretory PLA2 (sPLA2), which are one of the potent molecules in snake venoms (Ritonja and Gubensek, 1985; Maung-Maung et al, 1995; Chakrabarty et al, 2000). sPLA2 are low molecular weight proteins with molecular masses ranging from 13-19 kDa and generally requires Ca2+ for their activities (Kini, 1997; Valentin and Lambeau, 2000). Snake venom sPLA2 are secreted enzymes belonging to only two groups that are based on their primary structure and disulfide bridge pattern (Six and Dennis, 2000; Rouault et al, 2003; Ramar et al, 2010). Those of group I are the same as pancreatic sPLA2 present in mammals and are found in venom of Elapidae snakes, while group II PLA2s belong to the Viperidae and are similar to mammals’ nonpancreatic, inflammatory sPLA2s (Lambeau and Lazdunski, 1999; Dennis, 2000). Despite a high identity of their amino acid sequences, they exhibit distinct pharmacological effects including pre- or post-synaptic neurotoxicity, myonecrosis, cardiotoxicity, anticoagulant, antiplatelet aggregation, hemorrhagic, hemolytic, and cytolytic activities (Kini and Evans, 1988; Kasturi and Gowda, 1989; Stefansson et al, 1989; Maung Maung et al, 1995; Huang et al, 1997; Kole et al., 2000; Chakrabarty et al, 2002; Dong et al., 2003; Kini, 2003). Recently, acidic PLA2s, basic PLA2s, and synthetic peptides derived from PLA2 homologues have been shown to possess antitumor and anti-angiogenic properties (Roberto et al, 2004; Araya and Lomonte, 2007; Maity et al, 2007; Bazaa et al, 2009; Zouari-Kessentini et al, 2009; Bazaa et al, 2010; Kessentini-Zouari et al, 2010).

The Russell's viper (Daboia species) is a common venomous Viperinae snake, usually found in many South Asian countries. The subspecies found in Thailand is Daboia russelli siamensis, which is also discovered in Myanmar, Cambodia, southern China, Taiwan, and Indonesia (Warrell, 1989). The key lethal component of D. r. siamensis venom is the phospholipase A2, Daboiatoxin, which produces neurotoxicity in mice and exhibits oedema-inducing and myonecrotic activities (PLA2, EC 3.1.1.4) (Ritonja and Gubensek, 1985; Maung-Maung et al, 1995; Risch et al, 2009). PLA2 is a multifunctional enzyme that specifically catalyzes the hydrolysis of the fatty acid ester bond at the position 2 of 1,2-diacyl-sn-3-phosphoglycerides to produce free fatty acids and lysophospholipids (Kini, 2003; Burke and Dennis, 2009b).

In this study, we reported the purification and inhibitory activities of a PLA2 from D. r. siamensis venom. The Drs-PLA2 displayed a cytotoxic effect and inhibited cell migration in human skin melanoma cells (SK-MEL-28). It also reduced tumor lung colonization of B16F10 melanoma cells in BALB/c mice.

MATERIALS AND METHODS

Venom collection

Daboia russelli siamensis venom was obtained from the Queen Saovabha Memorial Institute (QSMI, Thai Red Cross Society, Bangkok) and was pooled venom from an underdetermined number of snakes. The venom was extracted by allowing the snake to bite into a container covered with parafilm. The venom was centrifuged at 9,000xg for 15min at 4ºC and frozen at -20ºC until lyophilized. The lyophilized venom was stored at -20°C until used.

Purification of Drs-PLA2

Five milligrams of lyophilized crude D. r. siamensis venom was suspended in 0.2ml of 0.05M ammonium acetate buffer, pH 8.2 and filtered through a 0.45micron filter. A total of 200µl (25mg/ml) was injected into a Waters 300SW (PROTEIN-PAKTM, 7.5x300mm) size-exclusion column. The column was previously equilibrated with the elution buffer (0.05M ammonium acetate buffer, pH 8.2). The collection process required 60min at a flow rate of 0.5ml/min. A Waters 2487 Dual λ absorbance detector was used to monitor absorbencies at 280nm. Waters™ Breeze software was used to control the pumps and store data. Each fraction was screened for indirect hemolytic activity, cytotoxicity, and inhibition of cell migration. Fraction 8 had a major protein band at about 14kDa with indirect hemolytic, cytotoxic, and cell migration inhibition activities (Figure 1). Fraction 8 was lyophilized and further purified by a DEAE anion exchange HPLC column.

Figure 1.

Purification of Drs-PLA2. A. Size exclusion (SE) chromatographic profile of crude D. r. siamensis venom. The grey-shaded areas indicate the location of PLA2 activities using cytotoxicity and cell migration assays. B. SDS-PAGE analysis of venom fractions from SE HPLC column. Crude venom or venom fractions were run on 4-12% (w/v) bis-Tris Gel under non-reducing conditions at 200V for 50min. The gel was stained with RapidStain. Lane 1: SeeBlue Plus2 Markers (InvitrogenTM); lane 2: crude venom (7μg); lanes 3-10: fractions 2-9 (7µg). C. DEAE anion exchange HPLC profile of fraction 8 from the SE HPLC column. The grey-shade areas indicate the location of PLA2 activities using indirect hemolytic, cytotoxicity, and cell migration assays. D. SDS-PAGE analysis of venom fractions from DEAE HPLC column. Crude and venom fractions from DEAE HPLC column were run on a 4-12% (w/v) bis-Tris gel under non-reducing conditions at 200V for 50min. The gel was stained with RapidStain for 1hr and distained overnight with 18megaohm water. Lane 1: SeeBlue Plus2 Markers (InvitrogenTM); lane 2: crude venom (7μg); lane 3: fraction 8 from SE (7μg); lane 4: fraction 8.1 (1.4μg); lane 5: fraction 8.2 (1.2μg); lane 6: fraction 8.3 (6μg); lane 7: fraction 8.4 (1.6μg); lane 8: fraction 8.5 (1.4μg); lane 9: fraction 8.6 (2μg); lane 10: fraction 8.7 (1.2μg); lane 11: fraction 8.8 (1.6μg); lane 12: fraction 8.9 (1.4μg); lane13: fraction 8.10 (1 μg). E. SDS-PAGE analysis of fraction 8 and 8.3 (Drs-PLA2). Drs-PLA2 was run on 4-12% (w/v) bis-Tris Gel under reducing conditions. Lane 1: SeeBlue Plus2 Markers (InvitrogenTM); lane 2: reduced form of fraction 8 from SE (7μg); lane 3: reduced form of Drs-PLA2 (3μg). F. Mass spectrometry analysis of Drs-PLA2.

A Waters™ DEAE anion exchange HPLC column was used for the second purification step. Fraction 8 from the first purification step was suspended in 0.02M Tris-HCl buffer, pH 8.0 and further purified. Two hundred microliters of clear supernatant at a concentration of 11.4mg/ml were applied into a Waters DEAE (PROTEIN-PAKTM 5PW, 7.5x75mm) anion exchange column, which was previously equilibrated with 0.02M Tris-HCl buffer, pH 8.0. The fractions were eluted using 0.02M Tris-HCl, pH 8.0 with a 0-0.5M NaCl salt gradient. The collection required 60min at a flow rate of 1ml/min. Each fraction was tested for indirect hemolytic activity, cytotoxicity, and inhibition of cell migration. Fraction 8.3 (Drs-PLA2) had indirect hemolytic, cytotoxic, and cell migration inhibition activities. The molecular weight and purity of Drs-PLA2 were determined by SDS-PAGE and verified by mass spectrometry.

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE)

Each fraction was applied to NuPAGE® Novex 4-12% (w/v) SDS-PAGE gels (InvitrogenTM). A XCell SureLockTM system with MOPS SDS running buffer (20x) diluted to 1x at 200V for 50min using a Bio-Rad PowerPac Basic system (Bio-Rad Laboratories, CA, USA). SeeBlue Plus2 markers ranging from 6-191kDa were used as standards. The gel was stained with RapidStain (InvitrogenTM, CA, USA).

Mass spectrometry (MALDI-TOF-TOF)

Drs-PLA2 was desalted using C18 Zip-tips (Millipore ZTC18S096). A 1μl of 2, 5- dihydroxybenzoic acid matrix was spotted on a MALDI Plate and 0.5μl of desalted fraction 8.3 was added onto the matrix. MALDI-TOF analysis was performed using a Bruker AUTOFLEX II-TOF (Bruker Daltonics) Mass spectrometer in positive mode. A Peptide Standard 2 (Bruker Daltonics) including insulin-5735.89, ubiquitin-8573.6, cytochrom-12370.83, myoglobin-16948.86, cytochrome-6183.08, and myoglobin-8459.67 was performed in a reflectron mode.

N-terminal sequencing

Drs-PLA2 (0.3mg/ml) was transferred from an SDS-PAGE onto a PVDF membrane (Milipore Immobion-P#PVH00010) using a Semi-Dry Transblot Cell (Bio-Rad) at 100mV for 1hr. The membrane was stained with Coomassie R-250 stain for 5min. The sample membrane was sent to the Iowa State University for N-terminal amino acid sequencing.

Anticoagulant activity

The anticoagulant activity of crude venom, fraction 8 from size exclusion purification, and Drs-PLA2 from anion exchange purification were measured using the Sonoclot analyzer by a modification of the procedure of Sánchez et al (2010). Briefly, a cuvette containing glass beads as the clotting activator was placed into the cuvette holder which maintained the temperature at 37oC. A pre-warmed 13μl of 0.25M CaCl2 were added to one side of the cuvette. A 10µl of Drs-PLA2 at the concentration of 0.3mg/ml (24µM) was added to the opposite side of the cuvette. A constant volume of 360µl of normal 10% (v/v) citrated human whole blood was added to the cuvette. The activated clot time (ACT) and clot rate (CR) were measured and the data were analyzed from Signature Viewer, software provided by Sienco, Inc. on an iMAC computer and analyzed by Microsoft Excel 2007. The negative control consisted of whole blood incubated with 0.02M Tris-HCl, pH 8.0 and 0.25M CaCl2.

Several studies have demonstrated that PLA2s inhibit blood coagulation by binding to FXa, which is the target protein in the coagulation cascade (Stefansson, 1990; Kerns et al, 1999; Kini, 2005). To test whether Drs-PLA2 specifically inhibits factor Xa, anticoagulation was assayed utilizing FX deficient plasma by a method of Suntravat et al (2010). Briefly, 300µl of FX deficient plasma were added to a cuvette without glass beads (Sienco, Inc, USA). Coagulation was activated by the addition of 10μl of pre-warmed 0.30M CaCl2 and 10µl of 6nM FXa pre-incubated at 37oC for 30min with 10µl of 0.1mg/ml (8µM) of Drs-PLA2. The controls included FX deficient plasma, 0.30M CaCl2 and 3nM FXa without Drs-PLA2 (positive control), and FX deficient plasma, 0.30M CaCl2 without 3nM FXa (negative control).

Indirect hemolytic assay

The crude venom and fractions were tested for PLA2 activity. The hemolysis indirect was tested on human washed red blood cells (1.2%, v/v) agarose plates (0.8%, w/v) with calcium chloride (0.01M-1%, v/v). The egg yolk solution was added to the agarose medium at a final concentration of 1.2% (v/v). Three millimeters diameter wells were made into the agarose plates. Twenty-five microlitres of D. r. siamensis and Crotalus atrox crude venoms were tested as positive hemolytic controls at a concentration of 1mg/ml. Fractions 8 (Figure 1A) and 8.3 (Figure 1C) were tested at a concentration of 0.5 and 0.3mg/ml, respectively. PBS was used as a negative hemolytic control. The plates were incubated at 37oC for 24hr. The specific activities were estimated as the ratio between the diameter (millimeters) of the hemolytic halo and the amount of protein added per well (micrograms).

Cytotoxicity assay of crude venom and Drs-PLA2 on cancer cell lines

Human skin melanoma (SK-MEL-28), human urinary bladder carcinoma (T24), human lung bronchus carcinoma (ChaGo-K-1), human fibrosarcoma (HT-1080), and murine skin melanoma (B16F10) cell lines were obtained from American Type Tissue Culture Collection (ATCC). The SK-MEL-28 and HT-1080 cell lines were maintained with Eagle's minimum essential medium (EMEM) containing 10% (v/v) fetal bovine serum (FBS), 50U/ml penicillin, and 50µg/ml streptomycin. The T24 cells were maintained with McCoy's 5A minimum essential medium containing 10% (v/v) FBS, 50U/ml penicillin, and 50µg/ml streptomycin. The B16F10 cells were maintained with Dulbecco's modified Eagle's medium containing 10% (v/v) FBS, 50U/ml penicillin, and 50µg/ml streptomycin. The ChaGo-K-1 cells were maintained with RPMI 1640 medium supplemented containing 10% (v/v) FBS, 0.02M sodium bicarbonate, 100U/ml penicillin, and 100μg/ml streptomycin. The cultured medium was replaced daily. Cells were incubated at 37oC in a 5% (v/v) CO2 humidified incubator.

Cytotoxic activity of D. r. siamensis venom (1mg/ml) was performed according to the procedure of Umthong et al (2009) by measuring cell viability using MTT (3-(4, 5-Dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide). To identify the most sensitive cancer cell lines for further screening cytotoxic and cell migration inhibition activities of venom fractions, cytotoxic activity of crude venom was determined on SK-MEL-28, T24, ChaGo-K-1, and HT-1080 cells. Cells were cultured on 96-wells flat-bottom microtiter plates at ∼105 cells/well in triplicate and incubated at 37oC in 5% (v/v) CO2 for 24hr. Crude venom was added to each cell suspension at various concentrations and incubation times. SK-MEL-28 cells were the most sensitive, and were further used for the determination of the cytotoxic and cell migration inhibition activities of venom fractions.

Each fraction collected from HPLC (100μg/ml), or Drs-PLA2 at various concentrations was added to SK-MEL-28 cell suspension at 37oC for 72hr. Then, 10μl of MTT (5mg/ml) was added to each well. After incubation for 4hr at 37oC, MTT was aspirated and 150μl of DMSO was added to lyse the cells. The absorbance at 570nm was read using a Beckman CoulterTM model AD 340 reader. Doxorubicin (5μg/ml), an anticancer drug was used as the positive control. The negative control was cells treated with 0.02M Tris-HCl, pH 8.0. The percentage of cell viability was calculated relative to the negative control, which was defined as 100%. The 50% cytotoxic concentration (CC50) of sample is defined as the protein concentration generating cell viability of 50%. The values of the percentages of cell viability were plotted against venom concentrations, and the CC50 was determined. Experiments were performed in triplicate.

Cell migration assays of the Drs-PLA2

Cell migration inhibition was determined by a wound-healing assay according to the procedure of Galán et al (2008). Briefly, SK-MEL-28 cells were plated (5x105 cells/ml) on a 24-well plate. After 24hr, the confluent monolayer was scratched with a sterile pipette tip creating the scratch wound of 16-23mm width. The detached cells were washed away and renewed with 0.9ml of EMEM mixed with crude venom at various concentrations, venom fractions collected from HPLC (6.25μg/ml), or Drs-PLA2 at various concentrations. The cells were then incubated at 37oC in a CO2 incubator for 0, 3, 6, 12, 24, and 48hr. After the incubation period, cell migration was observed under an inverted microscope (ULWCD 0.3 Olympus CK2, Japan). Echistatin (10μg/ml), a disintegrin known to inhibit SK-MEL-28 cell migration was used as the positive control (Sánchez et al, 2009). The negative control was cells treated with 0.02 M Tris-HCl, pH 8.0. The percentage of cell migration inhibition was calculated by the following equation: [(C-E)/C] x100, where C is the distance of the wound scratch (mm) at zero time of the negative control, and E is the distance of the wound scratch (mm) at the final incubation time for the venom. The 50% inhibitory concentration (IC50) is defined as the protein concentration that inhibits cell migration by 50%. The values of the percentages of cell migration inhibition were plotted against protein concentrations, and the IC50 was determined. Experiments were performed in triplicate.

Inhibition of lung tumor colonization in vivo

B16F10 murine skin melanoma cells (106cells) were mixed with Drs-PLA2 at the concentration of 100µg/kg body weight and then incubated at 37oC for 1hr. A 200μl of the sample mixture was injected into the tail veins of 18-20gm BALB/c mice. After 19 days, the mice were sacrificed to detect tumors in the lungs. Mice injected with only B16F10 cells were used as the positive control group. Mice injected with only Dulbecco's modified Eagle's medium were used as the negative control group. The tumors were counted and compared to the positive control group.

Statistical analysis

All results were expressed as mean ± standard deviation (SD). Their significance was analyzed by the Student's t-test. The level of significance was at P <0.05.

RESULTS

Crude venom was fractionated by size exclusion chromatography and nine fractions were collected (Figure 1A). The high molecular weight proteins (97kDa) were observed in fractions 2, 3, and 4. Low molecular weight proteins at molecular weights of about 14 and 6kDa were found in fractions 4 through 9. Fractions 2 through 9 were screened for cytotoxic and migration inhibition activities using the most sensitive cell towards the cytotoxic action of crude D. r. siamensis venom. Fraction 8 showed the most intense protein band at about 14kDa (Figure 1B, lane 9). Fraction 8 had the highest cytotoxicity with a CC50 of 1.47μg/ml, the migration inhibition activity with an IC50 of 1.22μg/ml, and the indirect hemolytic activity (data not shown). Fraction 8 was further purified by anion exchange HPLC chromatography. Ten different fractions were collected (Figure 1C). The molecular weight of fraction 8.3 was verified by mass spectrometry (13,679Da) (Figure 1F). The characteristics of the crude venom, fractions 8 and Drs-PLA2 used in this study are shown in Table 1.

Table 1

Comparison of N-terminal sequence homology between Drs-PLA2 and other snake venom PLA2s.

Snake PLA2	Accession No. or Ref.	Organism	N-terminal sequence	Identity (%)
Drs-PLA2*	-	D. r. siamensis	NLFQFARMINGKLGAFSV†	-
Basic PLA2	AAZ53185	D. r. limitis	NLFQFARMINGKLGAFSV	100
Rv4RV7 (basic PLA2)	1OQS	Vipera russelli formosensis	NLFQFARMINGKLGAFSV	100
Viperotoxin (basic PLA2)	Q02471.1	D. r. siamensis	NLFQFARMINGKLGAFSV	100
Basic PLA2	2I0U	V. nikolskii	NLFQFAKMINGKLGAFSV	95
Vaspin basic subunit	CAE47300	V. aspis zinnikeri	NLFQFAKMINGKLGAFSV	95
Basic PLA2 (B chain)	Q8JFG0	V. a. aspis	NLFQFAKMINGKLGAFSV	95
Vipoxin complex	1AOK	V. ammodytes meridionalis	NLFQFAKMINGKLGAFSV	95
Basic PLA2 I	B60512	V. aspis	NLFQFALMINGKLGAFSV	95
Vaspin basic subunit	CAE47291.1	V. a. aspis	NLFQLAKMINGKLGAFSV	89
Vaspin B isoform 1	AAO86503.1	V. a. aspis	NLFQSAKMINGKLGAFSV	89
Basic PLA2-II	ABD24037.1	D. r. russellii	NLFQFARMIDAKQEAFS	77
Basic PLA2	AAZ53178.1	D. r. siamensis	NLFQFARLIDAKQEAFS	71
Basic PLA2	AAP37177.1	D. r. siamensis	NLFQFARLIDAKQEAFS	71
Acidic PLA2	ACD43469	D. r. siamensis	NLFQFGDMINKKTGRFGL	61
Acid Daboiatoxin (DbTx)	1	D. r. siamensis	NFFQFAEMIVKMTGKEAV	50
Acidic PLA2	2	Cerastes cerastes	NLYQFGKMIKHKTGKSAL	44
Acidic PLA2	AAP41217.1	Echis carinatus	NLYQFGRMIWNRTGKL	43

This work

The sequences were aligned using BLASTP 2.2.25 program of GenBank.

Maung-Maung et al., 1995; Risch et al., 2009; Ritonja and Gubensek, 1985

Zouari-Kessentini et al., 2009

Bold letter indicates different residues as compared to Drs-PLA2.

Fractions from all chromatographic fractionation were applied to NuPAGE® Novex 4-12% (w/v) Bis-Tris SDS-PAGE gels. Drs-PLA2 showed a single band at about 14kDa using SDS-PAGE under non-reducing and reducing conditions (Figure 1D, lane 6 and 1E, lane 3). These results suggest that Drs-PLA2 is a purified monomeric protein. The monomeric Drs-PLA2 is similar to the PLA2 isolated from Cerastes cerastes venom (Zouari-Kessentini et al, 2009).

Mass analysis was performed on Drs-PLA2 using a Bruker AUTOFLEX II-TOF (Bruker Daltonics) resulting with a mass of 13,679Da, which falls in the range of many venom PLA2s (Figure 1F).

The N-terminal amino acid sequencing of Drs-PLA2 was NLFQFARMINGKLGAFSV displaying 100% sequence homology with basic PLA2s (Table 1). In addition, it is very similar (95% identity) to other basic PLA2s from viperid species (Table 1).

Crude venom, fraction 8, and the Drs-PLA2 were examined for coagulant activity using the Sonoclot analyzer. The Sonoclot signatures of crude venom, fraction 8, and the Drs-PLA2 are shown in Figure 2. Crude D. r. siamensis venom known to have the coagulant activity significantly shortened ACT (0.32±0.03min) and increased the CR (86.50±7.78U). Fraction 8 has the signature of coagulant activity with a shortened the ACT (0.66±0.03min) and enhanced the CR (44.00±5.66U). In contrast, the Drs-PLA2 had a signature of anticoagulant activity with an extended ACT (3.88±0.78min) and slower CR (5.60±0.36U). The positive blood control incubated with 0.02M Tris, pH 8.0 buffer had an average ACT of 3.46±0.30min, and a CR of 15.10±2.72U.

Figure 2.

Whole Blood Sonoclot Signatures of D. r. siamensis venom and venom fractions. Data from the Sonoclot Signatures further analyzed by Microsoft Excel 2007 obtained when human citrated whole blood was activated by crude venom, fraction 8 from size exclusion purification, and Drs-PLA2. The ACT is the time (min) in which whole blood begins to clot. The CR is defined as the rate of fibrin polymerization, which is the slope in the linear part of the curves and is defined as the change clot signal with change in time (U = Δsignal/Δtime). The solid line represents the ACT. The dashed line represents the CR.

To identify FXa as a target protein of Drs-PLA2, we demonstrated that Drs-PLA2 significantly prolonged the ACT (3.21±0.11min) and lowered the CR (1.70±0.28U). The positive control had an average ACT of 2.73±0.22 min and a CR of 3.37±0.91 U.

All samples displayed indirect hemolytic activity with Drs-PLA2 having the highest specific activity of 2mm/µg±0.2 followed by fraction 8 at 1.4mm/µg ±0.1. D. r. siamensis and C. atrox crude venoms had specific activities of 0.8±0.1 and 1.0mm/µg ±0.2, respectively.

Cytotoxicity assay of crude venom on cancer cell lines

Four different cell lines (SK-MEL-28, T24, ChaGo-K-1, and HT-1080 cells) were treated with a 100µg/ml of crude D. r. siamensis venom. The SK-MEL-28 cell line was the most sensitive (15-20% cell viability) to crude venom at all three incubation times (data not shown). All other cell lines exhibited 50-60% cell viability (data not shown). Thus, SK-MEL-28 cells were further used for the determination of cytotoxic and cell migration inhibition activities of the venom fractions.

Effects of crude venom at the various concentrations (6.25, 12.5, 25, 50, and 100µg/ml) and various incubation times on SK-MEL-28 cell viability were determined. Crude venom reduced cell viability in a dose-dependent manner. Crude venom at the concentration of 100µg/ml showed the lowest cell viability (20%) at 24hr and slightly decreased in the percentage of cell viability at 48hr and 72hr (data not shown).

Cytotoxicity and cell migration assays of the Drs-PLA2

Purified Drs-PLA2 decreased SK-MEL-28 percent cell viability in a dose-dependent manner (0.012µg/ml-98% ±10; 0.060µg/ml-94% ±5; 0.30µg/ml-80% ±3; 1.5µg/ml-40% ±10; and 7.5 µg/ml-20% ±5). The cytotoxic activity of the PLA2 was compared to crude venom. The CC50 value of the Drs-PLA2 (0.90µg/ml) was approximately 6 times less than the CC50 value of crude venom (5.63µg/ml). In addition, the Drs-PLA2 inhibited the migration of SK-MEL-28 with an IC50 of 0.35μg/ml, which was 11.3x lower than crude venom (3.97 µg/ml). Effects of the Drs-PLA2 at various incubation times on cell migration inhibition were determined. The percent inhibition of SK-MEL-28 cell migration of echistatin (positive control) at 3-12hr showed no significant difference to the percent inhibition of the Drs-PLA2. Subsequently, similar inhibition activity was found at 24hr. After a 48hr incubation period, cell migration inhibition activity of the Drs-PLA2 was higher than the positive control (Figure 3A). The actual images of SK-MEL-28 cell migration with the Drs-PLA2 at 24hr are shown in Figure 3B.

Figure 3.

Inhibition of Cell Migration by Drs-PLA2. A. Cell migration inhibition activity of the Drs-PLA2 on SK-MEL-28 cells using a wound- healing assay. Cell migration inhibition was evaluated by incubating 5x105 cells with Drs-PLA2 (6μg/ml) for 3, 6, 12, 24, or 48hr incubation periods. Echistatin (10μg/ml) was used as the positive control. The negative control consisted of cells treated with 0.02M Tris-HCl, pH 8.0. The results are expressed as the percentage of cell migration inhibition with respect to activity of the negative control, and as mean ±SD (n=3). B. SK-MEL-28 cells with the Drs-PLA2 (1.2μg/ml) at 24hr. (1) SK-MEL-28 cells were treated with 0.02M Tris-HCl, pH 8.0 at 24hr. (2) SK-MEL-28 cells were treated with Drs-PLA2 at 0hr, (3) 12hr, and (4) 24hr.

The Drs-PLA2 was further tested for anti-metastatic property using an in vivo skin melanoma cell colonization assay. The Drs-PLA2 significantly reduced tumor nodules by ∼65% compared with the positive control group (Table 2; Figure 4).

Table 2

Comparative analysis of tumor foci per lung in BALB/c mice using purified Drs-PLA2 compared to controls

	Control	Drs-PLA2 (100 μg/kg)
Mice no.	13	19
Minimum tumors	21	0
Maximum tumors	90	40
Mean tumors	49.15	17.42
Standard deviation	24.54	13.91
Tumor inhibition (%)a	-	64.50
P valueb	-	0.0002

The percent tumor inhibition was calculated by the following equation: [(E/C) x 100]-100, where E is the mean tumors of the Drs-PLA2 group, and C is the mean tumors of the control group.

P value as compared to the control. The level of significance was at P < 0.05.

Figure 4.

Inhibition of Lung Tumor Colonization by Drs-PLA2. The effect of the Drs-PLA2 on B16F10 lung tumor colonization in BALB/c mice at 100µg/kg mouse. The B16F10 cells (2x105) were injected in the lateral tail vein in the absence or presence of the Drs-PLA2. The lungs were isolated from the mice 19-days post-injections and observed for tumor colonization. A. Medium-treated mice (control); B. B16F10 cells in medium; C. 100µg of the Drs-PLA2/kg mouse.

DISCUSSION

Heart attack, stroke, and cancers are among the most serious medical problems worldwide. Venomous snakes contain an array of molecules with many different biological activities, which could have beneficial applications in medicine and biomedical research. In this study, we reported the purification and characterization of Drs-PLA2 isolated from crude D. r. siamensis venom. The Drs-PLA2 (13,679 Da) was purified from crude venom by a two-step purification procedure (Figure 1A, 1B and 1C) with approximately 27.4% yield (data not shown). The N-terminal sequence (18 amino acids) showed 100% homology to basic PLA2s from the viperid snakes (Table 1). However, it should be noted that this homology is based on the comparison of amino acid sequences at the N-terminal region of the molecules, without taking into account the remaining sequence of our PLA2. A basic PLA2 has been identified in the venom of D. r. siamensis (viperotoxin) sharing 100% N-terminus sequence homology with Drs-PLA2 (Table 1); however, viperotoxin is a basic protein with an approximate pI of 8.96. The pI for Drs-PLA2 appears to be slightly acidic since it bound to the anion exchange column, but was easily eluted with slightly above a 10% (w/v) NaCl (0.5M) solution (Figure 1C). These results suggest that viperotoxin is different from Drs-PLA2.

Although many PLA2s exhibit a high degree of similarity in their amino acid sequences, they display different biological effects. Tsai et al (2007) reported that PLA2s isolated from Myanmar Russell's viper venom (D. r. siamensis) and Eastern India Russell's viper (D. russelli) venom share 97-100% amino acid sequence identity, but they differ in their enzymatic and lethality effects.

Previous studies have indicated that several PLA2s isolated from the families Crotalidae, Viperidae, and Elapidae possessed anticoagulant properties (Mukherjee, 2007; Pereanez et al, 2009; Garcia Denegri et al, 2010). In this study, based on the ACT and CR data, fraction 8 from size exclusion purification exhibited a coagulant signature with a shortened ACT and an increased CR (Figure 2). In contrast, the Drs-PLA2 from the anion exchange purification had an anticoagulant activity with an extended ACT and a decreased CR. These results indicated that low molecular weight procoagulant molecules were completely removed from the first purification step (Figure 1E). Since most snake venoms contain PLA2 isoforms, it would be important to purify PLA2s from venom of individual snakes. In addition, the Drs-PLA2 had an anticoagulation effect on FX deficient plasma by inhibiting FXa, which is the target protein for anticoagulant PLA2s as previously reported (Stefansson, 1990; Kerns et al, 1999).

In recent studies, cytotoxicity, inhibition of cell migration, and tumor metastasis properties of PLA2s have been widely investigated. For example, a basic PLA2, RVV-7 (7.2kDa) isolated from D. r. russellii venom had strong cytotoxic activity on B16F10 melanoma cells. It also inhibited tumor growth in BLJ6 mice (Maity et al, 2007). Furthermore, an acidic Asp49 PLA2 (13,626.64Da) isolated from Macrovipera lebetina transmediterranea venom completely inhibited cell adhesion and migration of various human tumor cells mediated by α5β1 and αv integrins, but lacked cytotoxicity (Bazaa et al, 2009). Two acidic PLA2s purified from Cerastes cerastes venom, CC-PLA2-1 and CC-PLA2-2, inhibited HT-1080 cell migration towards fibrinogen and fibronectin (Zouari-Kessentini et al, 2009). In our study, crude venom had a higher cytotoxic activity on SK-MEL-28 cells than T24, HT-1080, and ChaGo-K-1 (data not shown). In addition, the PLA2 inhibited migration of SK-MEL-28 cells (Figure 3A and 3B). Although, there are studies demonstrating the anti-tumor activity of PLA2s in vitro, to the best of our knowledge, with the exception of the study by Maity et al (2007), there is insufficient in vivo evidence of the effects of secreted PLA2s on tumors. To confirm the metastasis property in our study, we determined the inhibition of skin melanoma cell colonization by Drs-PLA2, in vivo. It was observed that the pretreatment of B16F10 melanoma cells with the Drs-PLA2 inhibited growth of tumors in BALB/c mice by ∼65%.

Phospholipases A2s are enzymes that spark interest in the medical field because of their participation in a large number of human inflammatory diseases (Rodrigues et al, 2009). The activity and expression of several PLA2 isoforms are augmented in numerous human cancers, signifying that these enzymes have a vital role in both tumor development and progression and can be targets for anti-cancer drugs. On the other hand, some PLA2s isolated from Viperidae venoms are capable of inducing antitumoral activity (Rodrigues et al, 2009), suggesting that these molecules can be a new class of anticancer agents and provide new molecular and biological insights of cancer development. Furthermore, the initiation of the coagulation system in cancer patients is a well-known phenomenon accountable for recurrent thrombosis, which is the second most common cause of death in cancer patients (Xie et al, 2005). Thromboembolic complications consist of a wide range of clinical tribulations, from pulmonary embolism and deep venous thrombosis to superficial thrombophlebitis and clotting of central venous catheters (Versteeg et al, 2004; Xie et al, 2005; Furie and Furie, 2006).

CONCLUSIONS

In this study, we report for the first time a PLA2 purified from Thailand D. r. siamensis venom that had cytotoxic, anticoagulant, and antitumor activity, in vivo and in vitro. With the increase in protein engineering, it is now possible to produce a smaller fragment of this molecule that could preserve or improve its anticoagulant and antitumor activities, which could be useful in medical applications.