Disintegrins from the Venom of Vipera ammodytes ammodytes Efficiently Inhibit Migration of Breast Cancer Cells

Integrins are plasma membrane proteins, whose dysfunction frequently results in cancer pathology, and therefore they represent important targets of anti-tumour therapy. Snake venoms are a rich source of disintegrins (Dis), proteins that specifically bind integrins and thus interfere with their functions. In an attempt to discover new molecules for treatment of breast cancer, the major type of cancer in women, we isolated a dimeric Dis (Vaa-Dis) from the venom of the nosehorned viper. By cell viability testing we demonstrated that 50 nM and higher concentrations of Vaa-Dis were toxic to highly invasive human breast adenocarcinoma cell line MDA-MB-231. Wound-healing assay revealed that already at one order of magnitude lower concentrations Vaa-Dis efficiently inhibited MDA-MB-231 cell migration. This exposed a promising anti-metastatic potential of Vaa-Dis and a good perspective of these natural snake venom proteins for further research and development towards the application in breast cancer treatment.


Introduction
Cancer, one of the deadliest diseases worldwide, is caused by inherited or acquired mutations of the genetic material. The main characteristics of cancer include sustaining proliferative signalling, evading growth suppressors, resisting cell death, enabling replicative immortality, inducing angiogenesis, and activating invasion and metastasis. Such properties enable cancer cells an unlimited growth and spreading, invasion through the organism and finally causing its death. 1 Prevention of metastatic growth is an efficient therapy in cancer control. The key roles in the process of migration and cell viability are played by transmembrane proteins called integrins. 1,2 Integrins are cell adhesion receptors on the cells' surface that bind components of the extracellular matrix (ECM), various biological ligands and receptors on adjacent cells. They are heterodimeric proteins consisting of one αand one β-subunit. 18 different α-subunits and 8 βsubunits are known, which have been found to appear in 24 different α-β combinations, each of them possessing unique binding specificity. 2 Recent studies have exposed integrins as important factors in tumour cell survival, tumour growth and metastasis by establishing and breaking bonds between malignant cells and molecules in their surroundings. For this reason, integrins have become important targets of anti-tumour therapy. 2 Despite the fact that several therapeutics, integrin antagonists, have already been developed, a search for more efficient integrin-binding substances continues. [2][3][4] Snake venom (SV) proteins are broadly investigated as substances that can be used in diagnosis and treatment of human diseases. Among them, disintegrins (Dis) have been found to inhibit various cell functions by their interaction with different integrins, for example platelet aggregation, angiogenesis, tumour growth and metastasis. 5 Latinovi} et al.: Disintegrins From the Venom of Vipera ammodytes ammodytes ...
Dis are non-enzymatic cysteine-rich polypeptides. Their molecular mass ranges between 4 kDa and 15 kDa. They can be directly synthesized or formed by proteolysis of the P-II class metalloproteinases (MPs). SV Dis can be monomeric or dimeric, hetero-or homodimeric. In dimeric Dis, individual subunits are interconnected by disulphide bonds, which is crucial for stability and maintenance of a distinct globular structure. Dimerization also defines the configuration of the so called inhibitory loop, essential for the interaction with integrin receptor and, consequently, the biological activity. 6,7 In this work we focused on Dis from the venom of Vipera ammodytes ammodytes (Vaa-Dis). They were purified and biochemically characterized. We investigated their influence on cell viability and in vitro migration in a model of highly invasive triple-negative breast cancer and found that Vaa-Dis potently inhibited the migration of cancer cells.

1. Purification of Vaa-Dis
Raw Vaa venom was obtained from the Institute of Immunology, Zagreb, Croatia. Lyophilized venom was stored at -20 °C and before use dissolved in 50 mM Tris, 2 mM CaCl 2 , 300 mM NaCl, pH 7.0 (buffer A). 250 μl of the raw Vaa venom solution (16.67 mg) was applied on a Superdex 75 column 10/300 (GE Healthcare BioSciences AB, Sweden), equilibrated in buffer A and attached on a FPLC ÄKTA system (Amersham Biosciences, UK). Gel chromatography was performed at a constant flow rate of 0.5 ml/min. Concentration of proteins in the mobile phase was followed by measuring the absorbance at 280 nm (A 280 ). The B2 fraction was dialysed against 20 mM MES, 2 mM CaCl 2 , pH 6.5 (buffer B) and further fractionated on a SP Sepharose Fast Flow column (GE Healthcare Bio-Sciences AB, Sweden). Bound material was eluted by addition of 500 mM NaCl in buffer B. The material not retained by the cation-exchanger was collected and separated by reverse phase-high performance liquid chromatography (RP-HPLC) on a PLRP-S column (4.6 mm × 150 mm; 2.7 μm; 120 Å, Agilent Technologies, USA), equilibrated with 0.1% (v/v) trifluoroacetic acid (TFA) in water (solvent A), at a flow rate of 1 ml/min. The proteins were eluted from the column with a gradient of solvent B (90% (v/v) acetonitrile in 0.1% (v/v) TFA in water) as follows: 0-45% in 13 min and 45-55% in 10 min. A 215 was monitored to locate proteins in fractions, which were collected manually. Samples were dried using a vacuum concentrator SpeedVac (Savant, USA).

SDS-PAGE Analysis
Venom samples were analysed using 15% (m/v) polyacrylamide gels in the presence of SDS (SDS-PAGE) un-der non-reducing and reducing conditions according to Laemmli. 8 Proteins in gels were visualized by Page-Blue TM (Thermo Scientific, USA) as instructed by the manufacturer. Molecular mass standards were from Fermentas (Lithuania).

3. N-terminal Amino Acid Sequence Analysis
Proteins were N-terminally sequenced by automated Edman degradation on a Procise 492A protein sequencing system (Applied Biosystems, USA).

4. Culturing of MDA-MB-231 Cells
To carry out the migration assays the highly invasive breast cancer cell line MDA-MB-231 (ATCC, USA) was used. Cells were grown in RPMI-1640 medium (Invitrogen, USA) supplemented with 10% (v/v) fetal bovine serum (FBS) at 37°C in an atmosphere of 5% (v/v) CO 2 . Adherent cell monolayers were routinely cultured in T25 and T75 tissue culture flasks (Corning, USA) and passaged in ratios of 1 : 3 to 1 : 4.

5. Cell Viability Testing
The in vitro cytotoxic potency of Vaa-Dis was evaluated using the PrestoBlue TM viability assay (Invitrogen, USA) essentially as described previously. 9 Cells were trypsinized and counted by Trypan blue (Thermo Scientific, USA) exclusion assay 10 using a hemocytometer (Fortuna, Germany). Cells were seeded in 96-well plates (TPP, Switzerland) at a density of 5000 cells/well in 100 μL RPMI-1640 medium and left to attach overnight. Cells were then treated with different concentrations of Vaa-Dis (0.005; 0.05; 0.5; 5; 50; 500 nM) for 0, 24, 48, and 72 h. After the treatment, 10 μL of PrestoBlue TM was added to each well, and the plates were incubated for 30 min at 37°C . Cell supernatants were transferred to black microtiter plates (Corning, USA) and fluorescence was measured at an excitation of 560 nm and emission of 590 nm on an Infinite M1000 microplate reader (Tecan, Switzerland) to determine metabolic activity of the cells. Wells containing only the cell culturing medium and PrestoBlue TM were used as blank reference standards. Experiments were performed in triplicate. The normalization of cell viability was calculated as the ratio of sample absorbance to control absorbance (cells in media without Vaa-Dis).

6. Wound-healing Assay
The wound-healing assay (WHA) was used to estimate the potency of Vaa-Dis to inhibit cell migration (anti-migratory potency). MDA-MB-231 cells were trypsinized, counted as specified above and plated in 48-well cell culture dishes (TPP, Switzerland) at a density of 2 × 10 5 cells/well. Wounds (millimetre gaps) were then scratched in each cell monolayer using a pipette tip. Dead cells were removed by washing with D-PBS buffer and solutions of Vaa-Dis in RPMI-1640/10% FBS at different concentrations (2.5, 5.0, 7.5 and 15.0 nM) were added. Control samples contained only media. The influence of Vaa-Dis on cell migration was determined by observing the width of the gap under a CKX41 inverted microscope equipped with an E-450 camera (Olympus, Japan) after 1 h, 6 h, 12 h and 24 h of incubation. The gap was analysed using the ImageJ software (Softonic International S.A., Spain). The width of the gap was measured as specified (Suppl. Fig. 1) in three different wells for each concentration and incubation time.

7. Statistical Analysis
Each experiment was performed in at least three independent repeats. If not stated otherwise, data are presented as a mean percent difference from control with cor-responding standard error of the mean. Statistical tests were run by GraphPad Prism 6 (GraphPad Software Inc., La Jolla, CA, USA) using two-way ANOVA followed by Dunett's multiple comparisons test. Statistical significance is as follows: not significant for p > 0.05 (no mark in Fig. 2); significant for p ≤ 0.05 (* in Fig. 2), p ≤ 0.001 (** in Fig. 2) and p ≤ 0.0001 (*** in Fig. 2).

Results and Discussion
Using gel-filtration chromatography we separated in the first step the raw venom of Vaa into seven fractions (A-F in Fig. 1a). The efficient inhibition of platelet aggregation induced by ADP or collagen by the gel filtration fraction B2 has been suggested to be due to the presence of Dis, which may obstruct this process by binding to α IIb β 3 fibrinogen receptor. 11 For this reason, we decided to analyse the fraction B2 further and split it on a strong cation exchanger into a fraction retained by the exchanger (bound) and a flow-through fraction (unbound) (Fig. 1b). The latter was chromatographed on an RP-HPLC column.
Vaa-Dis was found in a sharp peak eluting at 40% of solvent B (Fig. 1c). SDS-PAGE analysis of this peak under non-reducing conditions revealed only one protein band of about 17 kDa. Under reducing conditions the protein displayed a lower apparent molecular mass of about 13 k-Da (Fig. 1d). This was expected as the usual structure of Viperidae Dis is a cystine-crosslinked dimer, 6,8 which dissociates to monomers in the presence of reducing agents. The homogeneity of Vaa-Dis sample was inspected by Nterminal amino acid sequencing using Edman degradation. Two N-terminal sequences, NSANP and NSGNP, were obtained, both characteristic for Dis. 12 Considering also the SDS-PAGE analysis, this means that the Vaa-Dis sample did not contain any non-Dis protein.  12 This is a hint that our sample may consist of two homodimeric rather than one heterodimeric Dis. VA6 is an RGD motifcontaining Dis. 5,10 It has been demonstrated that such Dis efficiently bind to α IIb β 3 receptor on platelets and in this way inhibit platelet aggregation. 12 The observed inhibition of ADP-and collagen-induced platelet aggregation by the Vaa venom gel filtration fraction B2 11 thus also suggests the same interpretation of our results. Some RGD-containing Dis have been demonstrated to inhibit α v β 3 (vitronectin receptor)-mediated migration of endothelial and cancer cells -two examples of such SV Dis are triflavin and DisBa-01. 13 As one of our main research topics is discovering new substances to oppose breast cancer, the major cause of death in women population worldwide, 14 the possibility that Vaa-Dis also affects the migration of cancer cells directed our subsequent experiments. To this end, resolving the question of whether Vaa-Dis is a heterodimeric Dis or a mixture of two homodimeric Dis was secondary so we did not go further to answer it. We study metastatic breast cancer on a model cell line, highly invasive human breast adenocarcinoma MDA-MB-231 cells. We found that Vaa-Dis is not cytotoxic to these cells at concentrations lower than 50 nM (Fig.  2a). At 50 nM and higher concentrations, Vaa-Dis induced a significant drop in cell viability, possibly by effecting proliferation and/or dying of the cells. Both processes are namely regulated via integrins, 2,4 which are present in higher amounts in cancer cells than in the healthy cells. 5 The level of expression of integrins α v β 3 , α v β 5 , α 5 β 1 , α 6 β 4 , α 4 β 1 and α v β 6 has been found in positive correlation with the ability of cells to migrate. 13 In MDA-MB-231 cells the integrin α v β 5 and α v β 3 are particularly highly expressed. 13 Drugs should not be toxic to healthy cells therefore the effect of Vaa-Dis on migration of cancer cell was evaluated at sub-cytotoxic concentrations of Vaa-Dis. As evident from the WHA results presented in Suppl. Fig. 1 and Fig.  2b, Vaa-Dis significantly slowed down the in vitro migration of MDA-MB-231 cells already at the concentration of 2.5 nM, well below concentrations at which cytotoxic effects were detected (Fig. 2a) and in the same concentration range as rhodostomin, triflavin and trigramin, all SV Dis. 15 The wound-healing assay on MDA-MB-231 cells was performed in the presence of Vaa-Dis at concentrations, which did not affect cell viability. Each result is an average of three independent experiments. Units are arbitrary (arb. U). Statistical significance is displayed as follows: no mark for p > 0.05 (not significant), * for p ≤ 0.05, ** for p ≤ 0.001 and *** for p ≤ 0.0001.

Conclusions
We have described the isolation and characterisation of novel Dis from the venom of Vaa. Due to a potent cancer cell anti-migratory activity, Vaa-Dis sample is promising for further research and development towards the use in breast cancer therapy.