Mass spectrometry guided venom profiling and bioactivity screening of the Anatolian Meadow ...

27 Pages·2015·2.61 MB·English

by  Bayram Göçmen

Toxicon107(2015)163e174
ContentslistsavailableatScienceDirect
Toxicon
journal homepage: www.elsevier.com/locate/toxicon
Mass spectrometry guided venom profiling and bioactivity screening
of the Anatolian Meadow Viper, Vipera anatolica
Bayram Goۍmen a,1, Paul Heiss b,1, Daniel Petras b, Ayse Nalbantsoyc,*,
Roderich D. Süssmuth b,**
aZoologySection,DepartmentofBiology,FacultyofScience,EgeUniversity,35100Bornova,Izmir,Turkey
bTechnischeUniversita€tBerlin,InstitutfürChemie,Strassedes17.Juni124,10623Berlin,Germany
cDepartmentofBioengineering,FacultyofEngineering,EgeUniversity,Bornova,35100Izmir,Turkey
a r t i c l e i n f o a b s t r a c t
Articlehistory: This contribution reports on the first characterization of the venom proteome and the bioactivity
Received1August2015 screening of Vipera anatolica, the Anatolian Meadow Viper. The crude venom as well as an isolated
Receivedinrevisedform dimeric disintegrin showed remarkable cytotoxic activity against glioblastoma cells. Due to the rare
27August2015
occurrence and the small size of this species only little amount of venom was available, which was
Accepted10September2015 profiledbymeansofacombinationofbottom-upandtop-downmassspectrometry.Fromthisanalysis
Availableonline16September2015
weidentifiedsnakevenommetalloproteases,cysteine-richsecretoryproteinisoforms,ametalloprotease
inhibitor,severaltypeA2phospholipases,disintegrins,asnakevenomserineprotease,aC-typelectin
Keywords:
and a Kunitz-type protease inhibitor. Furthermore, we detected several isoforms of abovementioned
Viperidae
AnatolianMeadowViper proteinsaswellaspreviouslyunknownproteins,indicatinganextensivecomplexityofthevenomwhich
Viperaanatolica wouldhaveremainedundetectedwithconventionalvenomicapproaches.
Snakevenomics ©2015ElsevierLtd.Allrightsreserved.
Top-downvenomics
Cytotoxicity
Bioactivityscreening
1. Introduction exampleiscaptopril,anlifesavinganti-hypertensivedrugderived
from a bradykinin potentiating peptide from Bothrops jararaca
Snake venom, which consists mainly of a mixture of proteins (Ferreira,1965;Rubinetal.,1978),apitviperfromthetropicaland
andpeptides,hasevolvedovereonsofyearstoimmobilizeandkill subtropical forests in southern Brazil, Paraguay and the North of
the prey as fast and as efficient as possible. Snake venoms also Argentina(Goncalves-Machadoetal.,2015).Nevertheless,notonly
causemanyhumanfatalities(>90,000)eachyear(Casewelletal., for the treatment of cardiovasculardiseases snake toxins maybe
2014).Apartfromtheseterribleconsequencesofsnakeenvenom- applicable,astherehavebeenreportedeffectsagainstneuropathic
ation, the toxins also represent an important natural source of pain(Woolf,2013)andtumors(Calderonetal.,2014).Particularly
potential leadstructuresfor thetreatmentofvarioushumandis- non-toxic doses of snake venom have been shown to reduce the
eases (Kang et al., 2011; McCleary and Kini, 2013). The medical size of solid tumors as well as inhibition of tumor angiogenesis
potentialofsnakevenomswasalreadyknownintheancientworld, (Swensonetal.,2004;Zhouetal.,2000).
whereitwasusedinsmalldosesasatraditionalmedicine(Vetter Ingeneral,snaketoxinstargetawiderangeofcellularmecha-
et al., 2011). Currently there are six FDA-approved drugs on the nisms,whicharetypicallyneuralreceptors,thecellmembraneand
marketwhicharebasedonvenomtoxins(King,2011)ofwhichthe thebloodcoagulationcascade(Chippaux,2006).Accordingly,the
majority originate from snakes. One of the most prominent toxinshavebeengroupedintoneurotoxins,cytotoxinsandhemo-
toxinscomposedofaround10proteinfamilies(three-fingertoxins;
protease inhibitors; C-type lectins; vascular-, endothelial- and
nerve growth factors; cysteine-rich secretory proteins (CRISPs),
* Correspondingauthor.
** Correspondingauthor. metallo-proteases (SVMP); serine-proteases (SVSP); L-amino acid
oxidases (LAAO); acetylcholine esterases (AChEs) and nucleotid-
E-mail addresses: [email protected] (A. Nalbantsoy), roderich.
[email protected](R.D.Süssmuth). ases) as well as peptides resulting from proteolytic cleavage
1 Bothauthorscontributedequallytothiswork.
http://dx.doi.org/10.1016/j.toxicon.2015.09.013
0041-0101/©2015ElsevierLtd.Allrightsreserved.
164 B.Goۍmenetal./Toxicon107(2015)163e174
(bradykinin-potentiating peptides, natriuretic peptides, dis- 2.2. Determinationofproteinconcentration
integrinsandSVMP-inhibitors)(Chippaux,2006;Fry,2015).
Overthepastdecades,theanalyticalcharacterizationofvenom Protein concentration was determined from diluted venom
has been continuously enhanced by technological developments. sample(4mg/mL)indeionizedwaterbyBradfordassay(Bradford,
Theimplementationofmassspectrometrictechniquesintoprote- 1976) using a UV/Vis spectrophotometer (VersaMax, Molecular
omic workflows as well as next-generation sequencing, have Devices, CA, USA) at awavelength of l ¼ 595 nm. Bovine serum
advancedtheanalysisofsnakevenomssignificantly.Sincevenom albuminwasusedasareference.
research evolved mainly from hypothesis-driven approaches,
focused on single toxins, to a systematic, untargeted analysis,
several so-called “bottom-up” approaches have been developed 2.3. Cellcultureandinvitrocytotoxicityassay
(Calvete, 2014). These techniques typically include a proteolytic
digestionoftheproteinstofacilitatetheiranalysisandidentifica- The following cell lines were used for determination of cyto-
toxicity: HeLa (human cervix adenocarcinoma), A-549 (human
tion (Calvete, 2013). The digestion step however multiplies the
alveolaradenocarcinoma),MCF-7(humanbreastadenocarcinoma),
numberofmoleculespresentinasampleanddisconnectstructural
CACO-2 (human colon colorectal adenocarcinoma), mPANC96
relation within highly homologous isoforms and eventually
occurring post-translational modified proteins. To overcome the (humanpancreasadenocarcinoma),PC-3(humanprostateadeno-
carcinoma), U87MG (human glioblastoma-astrocytoma) cancer
samplecomplexationandlossofinformationbyproteasedigestion
cells and as a non-cancerous cell lines, HEK (human embryonic
werecentlyintroducedatop-downmassspectrometricapproach
kidney).CelllineswerepurchasedfromATCC(Manassas,VA,USA).
in an exemplary study of King Cobravenom (Petras et al., 2015).
ThePC3celllinewasobtainedfromDr.K.Korkmaz(EgeUniversity,
However, even equipped with a set of toolboxes, including high
BioengineeringDepartment,Bornova-Izmir,Turkey).Allcellswere
throughput techniques capable of a rapid and systematic venom
cultivatedinDulbecco'smodifiedEagle'smediumF12(DMEM/F12),
analysis, only few venoms from over 600 biomedically relevant
supplemented with 10% fetal bovine serum (FBS), 2 mM/L gluta-
snakespecieshavebeenthoroughlystudied(Brahmaetal.,2015;
Calvete, 2013). Furthermore, most of these studies only reflect a mine,100U/mLofpenicillinand100mg/mLofstreptomycin(Lonza,
Visp,Switzerland).Thecellswereincubatedat37(cid:3)Cinahumidi-
partialpictureofthewhole-venomofacertainsnakespecies,since
toxinstructuresandvenomcompositionshowconsiderableintra- fiedatmosphereof5%CO2.
specific and geographic variability (Casewell et al., 2014; Reyes- Cytotoxicityofcrudevenomwasdeterminedbyfollowingthe
generalprocedurebasedoncellviabilityusingamodifiedcolori-
Velascoetal.,2015;Vonketal.,2013).
metric MTT [3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetra-
The distribution of poisonous snakes inTurkey includes three
zoliumbromide)]assay(Mosmann,1983;Yalcinetal.,2014).The
families: Colubridae, Viperidae and Elapidae. Particularly, viper
speciesarequiteabundantandsomespeciesareendemic(Go€çmen optical density (OD) was measured in triplicates at l ¼ 570 nm
(withareferencewavelengthl¼690nm)byUV/Visspectropho-
etal.,2006a).TheAnatolianMeadowViper,Viperaanatolica,which
was first recognized as a differentiated taxon among the Euro- tometry(Thermo,Bremen,Germany).Allcelllineswerecultivated
Asiaticvipersin1983(Bo€hmeandJoger,1983)canonlybefound for 24 h in 96-well microplates with an initial concentration of
1(cid:1)105cells/mL.Subsequently,theculturedcellsweretreatedwith
inanarrowandlimitedareanearElmali,intheAntalyaprovince
(Goۍmenetal.,2014).Tothebestofourknowledge,nostudieson differentdosesofvenomandincubatedfor48hat37(cid:3)C.Theplant-
derived compound parthenolide (a sesquiterpene lactone) was
the venom composition of V. anatolica have been reported in
usedasapositivecytotoxiccontrolagent.Percentagesofsurviving
literature so far. Thus, V. anatolica venom might be a source of
cellsineachcultureweredeterminedafterincubationwithvenom.
untapped bioactive peptides and proteins with interesting phar-
Theviability(%)wasdeterminedbythefollowingformula:
macologicalproperties.Asapartofourongoingresearchonsnake
venoms from Turkish species, the purpose of this work was to
profiling V. anatolica venom composition and to assess potential %Viablecells¼[(absorbanceoftreatedcells)(cid:4)(absorbanceof
blank)]/[(absorbanceofcontrol)(cid:4)(absorbanceofblank)](cid:1)100
cytotoxicactivitiesagainstdifferentcancercelllines.
2. Materialandmethods
2.4. Determinationofhalfmaximalinhibitoryconcentration(IC50)
2.1. Collectionandpreparationofvenomsamples
Incellculturestudiesforuntreatedcelllines(negativecontrols)
The study was approved by the Ege University, Local Ethical cytotoxicitywassetto0%.TheIC50valueswerecalculatedbyfitting
CommitteeofAnimalExperiment(Number:2010/43)andaspecial
thedatatoasigmoidalcurveandusingafourparameterlogistic
permission (2011/7110) for field studies was accepted from the
model and presented as an average of three independent mea-
Republic of Turkey, Ministry of Forestry and Water Affairs. All surements. The IC50 values were reported at 95% confidence in-
specimenswerereleasedbackintotheirnaturalenvironmentafter
terval and calculations were performed using Prism 5 software
investigation.
(GraphPad5, San Diego, CA, USA). The values of the blank wells
V. anatolica individuals were collected from late April to mid-
weresubtractedfromeachwelloftreatedandcontrolcellsandhalf
October 2014 in Cıglıkara forests, North West of Kohu Mt. in
Elmalı,AntalyaprovincesinTurkey,inaltitudesbetween1650and maximalinhibitionofgrowth(IC50)werecalculatedincomparison
tountreatedcontrols.
1750m(MASL).CrudeV.anatolicavenomwasextractedfromtwo
male and one female adult, using a paraffin-covered laboratory
beaker without exerting pressure on the venom glands. Venom 2.5. Morphologicalstudies
sampleswerepooledinonetubeandcentrifugedat2000(cid:1)gfor
10minatþ4(cid:3)Ctoremovecelldebris.Supernatantswerecollected, Themorphologicalstudiesofthecellswereperformedwithan
immediately frozen at (cid:4)80 (cid:3)C, and then lyophilized. Lyophilized inverted microscope (Olympus, Tokyo, Japan) compared to the
sampleswerestoredat4(cid:3)C. controlgroup48haftertreatment.
B.Goۍmenetal./Toxicon107(2015)163e174 165
2.6. Preparationofvenomsamplesforproteomicanalysis manually, dried in avacuum centrifuge, chemically reduced with
dithiothreitol(DTT)andsubmittedtoSDS-PAGE.SubsequentCoo-
For LC-MS analysis and semi-preparative HPLC fractionation, massiestainedbandswereexcisedfromthegelandsubjectedtoin-
crudevenomwasdissolvedinaqueous1%formicacid(HFo)(v/v)to gelreduction(10mMDTTin25mM(NH4)HCO3,pH8.3,for45min
afinalconcentrationof10mg/mL,andcentrifugedat20,000gfor at 65 (cid:3)C) and alkylation (50 mM iodacetamide in 50 mM (NH4)
5min.Fortop-downmeasurementsdisulfidebondswerereduced. HCO3, pH 8.3, for 30 min at 25 (cid:3)C), followed by in-gel trypsin
Therefore 10 mL of venom (10 mg/mL) was mixed with 10 mL of digestion(12hat37(cid:3)Cwith66ngsequencing-gradetrypsin/mLin
tris(2-carboxyethyl)phosphine (TCEP, 0.5 M) and 30 mL of citrate 25 mM (NH4)HCO3,10% ACN; 0.25 mg/sample). Tryptic peptides
buffer (0.1 M, pH 3), after which the reaction mixture was incu- weredriedinavacuumcentrifuge,re-dissolvedin15mLof5%ACN
batedfor30minat65(cid:3)C.Thereafter,thesamplewasmixedwithan containing0.1%HFo,andsubmittedtoLC-MS/MSanalysisusinga
equalvolumeof1%formicacidandcentrifugedat20,000(cid:1)gfor an Orbitrap XL hybrid mass spectrometer(Thermo, Bremen, Ger-
5min.Then10mLofbothreducedandnon-reducedsampleswere many) equipped with an HPLC system (Agilent, Waldbronn, Ger-
submittedtotop-downmeasurements.Avolumeof100mLofnon- many).LCseparationwasperformedonaGraceVydac218MSC18
reduced venom was subjected for semi-preparative HPLC column (2.1 (cid:1) 15 mm, 5 mm) with a flow rate of 0.3 mL/min. A
fractionation. gradientwasappliedusingof0.1%HFoinwater(solutionA)andin
ACN (solution B). The gradient started isocratically with 5% B for
2.7. Top-downvenomics 2min,followedbyanincreaseover10minfrom5to40%B,40e99%
B over 15 min, and held at 99% B for 5 min with final re-
LC-ESI-HR-MS/MS experiments were performed on an LTQ equilibrationphaseat5%Bfor5min.MSexperimentswereper-
Orbitrap XL mass spectrometer (Thermo, Bremen, Germany) formed in the Orbitrap analyser with R ¼ 15,000at m/z 400 and
coupled to an Agilent 1260 HPLC system (Agilent, Waldbronn, maximumfillingtimeof200msforbothsurveyandfirstproduct
Germany). A Suppelco Discovery 300 Å C18 (2 (cid:1) 150 mm, 3 mm ionscans.MS/MSfragmentationofthetwomostintenseionswas
particlesize)columnwasused.Typicallytheflowratewassetto performedintheLTQusingCID(30msactivationtime);thecolli-
0.3 mL/min and a gradient of 0.1% HFo inwater(solution A) and sionenergywassetto35%.Precursor-ionisolationwasperformed
0.1%HFoinacetonitrile(ACN)(solutionB)wasused.Thegradient withinamasswindowof2m/z.Dynamicexclusionwassetupfora
startedisocratically(5%B)for1min,followedbyanincreasefrom5 3m/zwindowsforupto50precursorionswitharepeatof2within
to40%Bover40min,40e70%over20min,awashoutat70%Bfor 30s.DenovoannotationofMSMSspectrawasperformedwiththe
10min,andare-equilibrationphaseat5%B.ESIsettingswere40L/ DeNovoGUItool(Muthetal.,2014).Manuallyprovedsequencetags
min sheath gas; 20 L/min auxiliary gas; spray voltage, 4.8 kV; weresearchedagainstaViperidaenon-redundantproteindatabase
capillary voltage, 46 V; tube lens voltage, 135 V and capillary ofUniProtKB/TrEMBLusingBLASTP(Altschuletal.,1990).
temperature,330(cid:3)C.Forinformationdependentacquisition(IDA),
fourscaneventsweresetwith2microscansand500msmaximal 2.9. Relativetoxinquantification
fill time. The survey scan was performed with mass resolution
R¼100,000(atm/z400).ForMS/MSRwassetto60,000(atm/z The relative abundances (percentage of the total venom pro-
400).EverycyclecontainedtwoCIDscansofthetwomostabun- teins)ofthedifferentproteinfamilieswerecalculatedastheratioof
dantionsofthesurveyscan.Normalizedcollisionenergywassetto thesumoftheareasofthereverse-phaseUV214-chromatographic
35%forCID.Thedefaultchargestatewassetto8þandtheacti- peakscontainingproteinsfromthesamefamilytothetotalareaof
vationtimeto30msec.Theprecursorselectionwindowwassetto venomproteinpeaksinthereverse-phasechromatogramaccord-
2 m/z. Dynamicexclusionwasperformed witha 3 m/zexclusion ingtoCalveteetal.(Calvete,2014).Ifmorethanoneproteinwas
windowforprecursorionswith1repeatwithin10s.Theexclusion present in a reverse-phase fraction, their proportions were esti-
list contained maximal 50 ions for a duration of 20 s. For the mated by the relative abundance of deconvoluted top-down
deconvolution of isotopically resolved spectra the XTRACT algo- spectra.Ifco-elutingproteins wereobservedbySDS-PAGE which
rithm of Xcalibur (Thermo, Bremen, Germany) was used. For werenotaccessiblebymassspectrometry,theirrelativeabundance
isotopicallynotresolvedspectratheZscorealgorithm(Zhangand wasestimatedbyopticalsignalstrengthoftheCoomassie-stained
Marshall,1998)implementedinthemagictransformer(MagTran) bands(Goncalves-Machadoetal.,2015).
tool was used for charge distribution deconvolution. For protein
identifications, deconvoluted mass spectra were analyzed manu- 3. Resultsanddiscussion
ally.TheresultingdenovogeneratedsequencetagswereBLASTed
against the NCBI non-redundant Viperidae database (http://blast. 3.1. Viperaanatolica
ncbi.nlm.nih.gov/Blast.cgi) using the BLASTP algorithm (Altschul
etal.,1990). TheAnatolianMeadowViperisanviperspeciesendemicinan
area less than 100 km2 east of Elmali, in Southwestern Anatolia,
2.8. Bottom-upvenomics Turkey(showninFig.1B).Thespeciesisextremelyrareandlisted
ascriticallyendangeredontheIUNCRedlist2015(VarolToketal.,
Forbottom-upanalysis,thevenomwasdissolvedtoaconcen- 2015). It took us several years and expeditions to finally capture
trationof10mg/mLinaqueous1%HFoand5%acetonitrile(ACN).A threeindividuals(2maleand1female)inlengthsbetween20and
volumeof100mLwassubjectedtosemipreparativereverse-phase 40cm. Fig.1Ashowsaphotographofafemaleandamale indi-
(RP) HPLC separation on a Suppelco Discovery 300 Å C18 vidual in their natural habitat. Venom of each individual was
(4.6(cid:1)150mm,3mmparticlesize)column.Theflowratewassetto extractedtwiceandfreezedried,whichyieldedatotalamountof
1mL/min.Alineargradientof0.1%HFoinwater(solutionA)and ~1 mg dried venom. The peptide and protein concentration was
0.1%HFoinACN(solutionB),isocratically(5%B)for5min,followed determinedbymeansofaBradfordassayas~38%(wt%).
bylineargradientsof5e40%Bfor95min,40e70%for20min,70%B Inordertoobtainadetailedpictureofthevenomcomposition
for10min,andfinallyre-equilibrationat5%Bfor10minwasused. and to assess the cytotoxicity of isolated toxins we performed a
Peakdetectionwasperformedatl¼214nmusingadiodearray detailed proteomic analysis and fractionation of the venom. It is
detector (DAD). Chromatographic fractions were collected importanttonotethatthelimitedaccesstothesnakes,duetotheir
166 B.Goۍmenetal./Toxicon107(2015)163e174
Fig.1. A:PhotographofViperaanatolicaeThephotowastakenduringafieldtripinOctober2014atKohuMt.,Elmali,Antalya,Turkey.Theindividualontherightisanadultfemale;
thesecondindividualisanadultmale.B:GeographicdistributionofViperaanatolica.TheAnatolianMeadowViperisanendemicviperspeciesinanarealessthan100km2eastof
Elmali,insouthwesternAnatolia,Turkey,bluecircle.(Forinterpretationofthereferencestocolourinthisfigurelegend,thereaderisreferredtothewebversionofthisarticle.)
lowabundance and the minimal amounts of venomavailable for sametimeoccurinlowconcentrations.
mass spectrometric analysis, made a classic venomic protocol In order to obtain useful fragment spectra of toxins in subse-
(HPLCfractionation,SDS-PAGE,ingeltrypsindigestionfollowedby quent MS2 experiments it was required to reduce potentially
MSMSdenovosequencing)aimingatathoroughproteomicchar- occurringdisulfidebridgesoftoxinspriortotop-downanalysis.We
acterization an extremely challenging task (Calvete, 2013, 2014; thusreducedthevenomchemicallyusingTCEPpriortoasecond
Goncalves-Machado et al., 2015). This is exemplified by the fact LCMS run with data dependent MS2 acquisition. Because a
thatthetypicalamountofvenomappliedtoonesemi-preparative comprehensive analysis of top-down data relies strongly on the
HPLCrunequalstheamountofvenomwehadavailabletoperform availabilityof transcriptomeand/orgenomedata,whichyetdoes
thecompleteanalyticalcharacterization.Inordertoobtainagen- not exist for V. anatolica and is still limited for closely related
eraloverviewofthemolecularmassesofthevenomcomponents, Viperidae,wemanuallyanalyzedtheESImassspectraandsearched
including low abundant and low molecular mass compounds denovogeneratedsequencetagsagainstaViperidaedatabaseusing
whichmightnotbegatheredduringSDS-PAGE,wethusdecidedto the Blast algorithm. Using this method we could identify a PLA2
perform an initial top-down mass spectrometric analysis of the fragmentwiththesequencetag(1871)ALFSYSDYGCYCGWG(1931).
crudevenom. Asthespectrumwasobtainedfromchemicallyreducedvenomthe
observedfragmentationofthePLA2couldhaveoccurredduringthe
3.2. Top-downvenomics sample preparation before mass spectrometric analysis. Further-
morewecoulddirectlyidentifyadisintegrinlikecompoundwith
Intop-downapproachestheproteinsarenotdigestedpriorto thesequencetag(5085)-DYCTGJS-(221).Fig.3exemplarilyshows
mass spectrometric analysis (Ge et al., 2002). Instead they are the deconvoluted top-down MS1 and MS2 spectra of the
directlyionizedandthusrenderthemolecularmassoftheintact disintegrin-like venom protein with a molecular weight of
protein. Ideally, if the protein is further fragmented in the gas 6047.5Da(peak11inFig.2).Inthetop-downMSanalysisthenon-
phase,aminoacidsequencesandsitespecificmodificationscanbe reduced disintegrin like compounds (peak 11) instead showed
determined. Therefore isotope resolution is typically required, in average molecular masses of 13,982.8, 14,001.7 and 14,017.7 Da
ordertodeconvolutethecomplexfragmentspectra.Thisislimited which correspond to the SDS-PAGE analysis of the isolated peak
by the resolution provided by the mass analyzer used, typically with a band height at ~ 14 kDa and thus indicates that the dis-
<50kDaforOrbitrapanalyzerswithR¼100,000at400m/z.Asthe integrin in peak 11 is a hetero-dimer. Furthermore peak 4 (see
majorityofthetoxinspresentinvipervenomsexceedthismolec- Fig. 2) could be identified as a snake venom metalloprotease in-
ularweight(>50kDa,SVMPs,SVSPs,LAAOsetc.),onecanstillmake hibitor,whichiscommonlypresentinViperidaevenoms(Munekiyo
useofdeconvolutionbasedonchargedistributionofproteinMS1 andMackessy,2005;Wagstaffetal.,2008).Nevertheless,through
spectra. thedifficultiesinionizinganddeconvolutinghighmolecularmass
AninherentlimitationofSDS-PAGEisthatsmallsizedpeptides components (>50 kDa) which represent the majority of the
and proteins are not efficiently resolved which makes a direct V.anatolicavenomcontent,thepeptide/proteinIDsobtainedfrom
HPLC-supported mass spectrometric analysis the most effective the initial top-down analysis are limited. A future solution to in-
waytogatherstructuralinformation(Petrasetal.,2015).InFig.2 crease the number of protein IDs obtained by top-down MS/MS
the UV chromatogram of the semi-preparative fractionation and wouldbeapre-fractionationbysizeexclusionchromatographyand
the corresponding Coomassie-stained SDS-PAGE analysis of the thus an enrichment of the low abundant low molecular mass
fractionsaswellasthetotalioncurrent(TIC)oftheLCMSanalysis toxins, whichare accessiblefortop-downMS/MSelucidationbut
ofnativevenomisshown.Duetothedifferenceincolumnsizethe remainedunselectedduringtheIDAMS/MSexperiments.Afurther
retentiontimesareslightlyshiftedbetweentheLCMSanalysisand generalimprovementforthevenomproteomeanalysiswouldbea
thesemiprep-HPLCrun.Intheinitialtop-downmassprofiling~100 transcriptome analysis of V. anatolica venom gland tissue, which
toxin masses were detected which are listed in Table 1 (All would enable the use of spectra-database comparisons for data
deconvolutedmassspectra are shownin thesupplementalinfor- analysis, which is more effective than the generation of de novo
mation).Intotalweobserved22molecularmasses<1kDaand53 sequencetagsandasubsequentBlastsearch.
molecularmassesbetween1and9kDa,whichrepresents~75%of
thepeptideandproteinspecies.Furthermoretheinitialtop-down 3.3. Bottom-upvenomics
analysisrenderedanother27compoundsbetween9and60kDa.
Interestingly, the majority of venom toxins seem to consist of Besides the initial top-down mass spectrometric characteriza-
molecular masses in the low mass range (<9 kDa) which, at the tionrenderingintactvenomcomponents,whichalreadyshoweda
B.Goۍmenetal./Toxicon107(2015)163e174 167
Fig.2. VenomprofileofViperaanatolica.Reverse-phaseHPLCseparationofV.anatolicavenomproteins.AshowstheSDS-PAGEanalysisoffractionscollectedfromthesemi-
preparativeHPLCrundisplayedinB.Excisedproteinbandswereidentifiedbybottom-updenovosequencing.PanelCshowsthetop-downMS1analysis.[Molecularmasses,
sequencetagsanddatabasehitsarelistedinTable2.Extractedanddeconvolutedmassspectraareshowninthesupplementalmaterial.ThedeconvolutedMS1andMS2spectraof
peak11areshowninFig.4].
detailedpictureofthestructuraldiversityofthevenom,weiden- deconvoluted top-down spectra of proteins was taken as a mea-
tified the remaining toxin families through the application of a sure.Ifaninefficientionizationofproteinswasobservedthesignal
bottom-up approach. By means of HPLC fractionation, SDS-PAGE strengthofSDS-PAGEbandsservedasameasure.Hence,themost
separation followed by in-gel digestion and LC-MS/MS analysis abundant toxin family is represented by snake venom metal-
wecouldgenerate46denovosequencetags(Table1),representing loproteases(SVMP,41.5%),followedbytwocysteine-richsecretory
seventoxinfamilies.Thepercentageofthevenomproteincontent protein isoforms (CRISP, 15.9%) with molecular masses of
ispresentedinthepiechartinFig.4.Fordeterminationofrelative 24,645.0 Da and 24,544.1 Da, metalloprotease inhibitor (SVMPi,
abundance(in%)ofvenomproteinstheintensitiesofUVpeakareas 9.3%),A2phospholipases(PLA2,8.1%),disintegrins(2.0%),asnake
were used. In case of coelution the relative intensity of venom serine protease (SVSP, 1.6%), a C-type lectin (1.1%) and a
168 B.Goۍmenetal./Toxicon107(2015)163e174
Table1
VenompeptidesandproteinsidentifiedfromViperaanatolica.ProteinassignmentofRP-HPLCfractions(Fig.4)byLCMSandMS/MSanalysis.Peaknumberingcorrespondsto
theUVandMSchromatogramsshowninFig.2.SequencetagsareobtainedbydenovoanalysisofintactproteinandtrypticpeptideMS/MSspectra.ProteinIDsareobtainedby
BLASTPanalysisofthesequencetagsagainstaviperidnon-redundantproteindatabase.MolecularweightwasdeterminedbySDS-PAGE(*)andtop-downMSanalysis(#)and
areshownasaveragemasses.
Peak Identifiedsequencetag ProteinID BlastE-valueNCBIaccessionM[kDa]*M[Da]# Method
number number
1 unknown(~peptide) 429.0;621.0;835.1; Top-downMS1
1071.0;1241.1
2 unknown(~peptide) 579.2;816.4;1269.6; Top-downMS1
1600.8;2049.0
3 unknown(~peptide) 969.4;1026.5 Top-downMS1
4 pQKW Metalloproteaseinhibitor 443.2 Denovofrom
Top-downMS2
5 unknown(~peptide) 474.2;598.4;752.4;858.3;Top-downMS1
1128.6;1747.7
6 unknown(~peptide) 454.3;499.4;726.8 Top-downMS1
7 unknown(~peptide) 452.3;606.3 Top-downMS1
8 unknown(~peptide) 808.4;1128.6;3943.8 Top-downMS1
9 (423.02)-YGGCGGNANNFK-COOH ~Kunitz-typeserine 2.0E-08 P0DKL8.1 14 6737.9 Denovofrom
proteaseinhibitor Trypsin-digest,
Top-downMS1,
MassfromSDS-Page
unknown(~peptide) 452.3;497.4;707.4;781.4;Top-downMS1
3137.5;4014.8;
4848.2;5123.4
10 unknown(~Protein) 14;56 MassfromSDS-Page
unknown(~peptide) 1115.6;3119.5;3233.5 Top-downMS1
11 NH2-FJNAGTJCQY-(227.02) ~DisintegrinVA6 3.0E-04 P0C6A5.1 14 13,982.8;14,001.7; Denovofrom
14,017.7 Trypsin-digest,
Top-downMS1,
MassfromSDS-Page
(532.05)-DYCDGJSSDGVDR-COOH ~DisintegrinVB7A 2.0E-02 P0C6A6.1
(1001.56)-YQTGJSSDCPR-COOH ~DisintegrinVB7B 6.0E-04 P0C6A7.1
(5085)-DYCTGJS-(221) ~DisintegrinVB7A 1.2E-02 6047.5(reduced) Denovofrom
Top-downMS2
12 unknown(~peptide) 14 1457.7;6817.2 Top-downMS1;Mass
fromSDS-Page
13 unknown(~peptide) 14 MassfromSDS-Page
unknown(~peptide) 1101.5;1357.7;6965.2; Top-downMS1
7001.2;7227.1
14 (241.03)-JCFGDQJNTYDK-COOH ~Neutralphospholipase 1.0E-05 P34180.2 14 13,639.9(5461.6) DenovofromTrypsin-digest,
(1871)ALFSYSDYGCYCGWG(1931).A2ammodytinI2 7.0E-12 CAE47236.1 Top-downMS1,
~ammodytinI2(C)isoform Top-downMS2,Mass
fromSDS-Page
unknown(~Protein) 6948.3;6982.2 Top-downMS1
15 unknown(~Protein) 7175.4;7212.4;7293.5 Top-downMS1
16a unknown(~Protein) 35 MassfromSDS-Page
16b (890.35)-VAAJCFGENJNS-(548.34) ~Neutralphospholipase 5.0E-06 P34180.2 14 13,637.9 DenovofromTrypsin-digest,
A2ammodytinI2 Top-downMS1,Mass
fromSDS-Page
(354.21)-CFGENJNTYDKK-COOH ~Neutralphospholipase 5.0E-09 P34180.2 14 7276.4;7903.7 DenovofromTrypsin-digest,
A2ammodytinI2 Top-downMS1,Mass
fromSDS-Page
unknown(~Protein) 3772.2;5549.1;5586.1; Top-downMS1
5677.2;5713.1
17 NH2-SVNPTASNMJK-COOH ~Cysteine-richvenom 5.0E-07 B7FDI0.1 25 24,645.0 DenovofromTrypsin-digest,
protein Top-downMS1,Mass
fromSDS-Page
NH2-SVDFDSESPR-COOH ~Cysteine-richvenom 3.0E-05 B7FDI0.1
protein
NH2-FJDAYPEAAANAER-COOH ~Cysteine-richvenom 7.0E-06 B7FDI0.1
protein
(225.09)-EJQNEEPDJHNSJR-COOH ~Cysteine-richvenom 2.0E-05 B7FDI0.1
protein
18 (439.24)-ECGENJYMSTSEVK-COOH~Cysteine-richvenom 1.0E-05 B7FDI0.1 25 24,544.1 DenovofromTrypsin-digest,
protein Top-downMS1,Mass
fromSDS-Page
19a NH2-AYJGTMCQPK-COOH ~H3metalloproteinase 1.2E-02 AGL45259.1 55 46,397.0 DenovofromTrypsin-digest,
precursor1 Top-downMS1,Mass
fromSDS-Page
19b unknown(~Protein) 35 MassfromSDS-Page
19c (186.09)-DFTTESPR-COOH ~Cysteine-richvenom 8.6E-02 B7FDI0.1 25 DenovofromTrypsin-digest,
protein Top-downMS1,Mass
fromSDS-Page
19 unknown(~Protein) 1375.8;6067.3;6166.3 Top-downMS1
20a unknown(~Protein) 60 MassfromSDS-Page
B.Goۍmenetal./Toxicon107(2015)163e174 169
20b NH2-JQGJVSWGS-(487.26) ~Snakevenomserine 5.0E-05 E5AJX2.1 35 DenovofromTrypsin-digest,
proteasenikobin Top-downMS1,Massfrom
SDS-Page
NH2-JMGWGTJTTTK-COOH ~Snakevenomserine 4.5E-02 E5AJX2.1
proteasenikobin
(699.24)-JJJEEWVJDQR-COOH ~Snakevenomserine 8.9E-01 E5AJX2.1
proteasenikobin
NH2-FPNGDJKDLMJJR-COOH ~Snakevenomserine 4.4E-01 E5AJX2.1
proteasenikobin
20c (1350.51)-JVCDGGDDPGTR-COOH ~ammodytinI1(A) 9.5E-02 CAE47176.1 14 DenovofromTrypsin-digest,
variant(PLA2) MassfromSDS-Page
21 NH2-TAJDFDGSVJGK-COOH ~metalloproteinase 5.7E-01 ADI47580.1 55 DenovofromTrypsin-digest,
MassfromSDS-Page
20/21 unknown(~Protein) 3181.2;4688.4; Top-downMS1
5505.0:7376.8;
7833.8;12,891.8
22 unknown(~Protein) 8172.7;16,346.4, Top-downMS1
16,397.1
23aþ24a NH2-SSDJGFEDYSQJDR-COOH ~Zincmetalloproteinase 7.6E-01 P0DJE2.3 70 DenovofromTrypsin-digest,
-disintegrin-like MassfromSDS-Page
ammodytagin
(895.05)-JJTFDDSFGEWR-COOH ~Zincmetalloproteinase 5.3E-01 P0DJE2.3
-disintegrin-like
ammodytagin
NH2-JPECJJNKPJR-COOH ~Zincmetalloproteinase 4.4E-01 P0DJE2.3
-disintegrin-like
ammodytagin
24bþ25 (373.93)-GDDTJDSFGEWR-COOH ~metalloproteinaseF1 8.0E-05 AJC52543.1 60 DenovofromTrypsin-digest,
MassfromSDS-Page
NH2-YDYSEDPDYGFGD-(449.1) ~Zincmetalloproteinase 1.6E-01 P0DJE2.3
-disintegrin-like
ammodytagin
NH2-VNJJNEMYJPLNJR-COOH ~H3metalloproteinase 2.6E-01 AGL45259.1
23bþ24c NH2-VPJVGVEJWDHGDJJK-COOH ~H3metalloproteinase 4.0E-03 AGL45259.1 45 48,693.0 DenovofromTrypsin-digest,
MassfromSDS-Page,
Top-downMS1
NH2-VNJJNEFYLPJNJR-COOH ~H3metalloproteinase 3.7E-01 AGL45259.1
(174.31)-NJFGEDYSQJDR-COOH ~H3metalloproteinase 1.5E-02 AGL45259.1
23e25 unknown(~Protein) 4452.7;6299.4; Top-downMS1
10,102.4;11,831.3;
24,411.4
26 NH2-VPJVGVEJWDVPJTPR-COOH ~metalloproteinaseH4-A 2.2E-02 AHB62069.1 DenovofromTrypsin-digest
26 unknown(~Protein) 8594.1;16,106.5 Top-downMS1
27 unknown(~Protein) 50 MassfromSDS-Page
unknown(~Protein) 6433.0;9356.4; TICTop-downMS1
16,586.1
29be30b NH2-TWFNJNCEER-COOH ~C-typelectin 1.0E-04 Q696W1.1 20 21,892.7 DenovofromTrypsin-digest,
Top-downMS1,Mass
fromSDS-Page
28e29 unknown(~Protein) 50 49,184;24,561.5 TICTop-downMS1,
MassfromSDS-Page
30a unknown(~Protein) 48,281; TICTop-downMS1
30e32 unknown(~Protein) 27,058.0 TICTop-downMS1
31 (388.99)-PDDPDYGFVDJGTK-COOH ~Zincmetalloproteinase 1.0E-08 P0DJE2.3 50 DenovofromTrypsin-digest,
-disintegrin-like MassfromSDS-Page
ammodytagin
NH2-TLGJAPVSGMCQPK-COOH ~metalloproteinaseH4-A 3.0E-06 AHB62069.1
NH2-HDNAQJJTGJDJN-(213.07) ~metalloproteinaseH4-A 7.0E-08 AHB62069.1
32 NH2-MTJEGFJTGLDJNGR-COOH ~Zincmetalloproteinase 2.1E-01 P0DJE2.3 50 DenovofromTrypsin-digest,
-disintegrin-like MassfromSDS-Page
ammodytagin
NH2-DVGJAPVSGMCQPK-COOH ~metalloproteinaseH4-A 2.0E-04 AHB62069.1
(323.01)-ATSEQQR-COOH ~metalloproteinaseH4-A 1.1E-01 AHB62069.1
33a (260.16)-DHJDJJCJJNQPJR-COOH ~Zincmetalloproteinase 1.5E-02 Q0NZX6.1 65 32,076.0 DenovofromTrypsin-digest,
-disintegrin MassfromSDS-Page,
jararin Top-downMS1
(626.8)-NPESSJJNQPJR-COOH ~metalloproteinase 1.6E-01 ADI47709.1
33b NH2-HDNAQJJTGJD-(440.24) ~Zincmetalloproteinase 8.0E-05 U5PZ28.1 50 57,085.0 DenovofromTrypsin-digest,
-disintegrin MassfromSDS-Page,
BlatH1 Top-downMS1
NH2-VTSSGDDTJDSFGEGER-COOH ~metalloproteinase 5.0E-07 ADI47715.1
NH2-YYJNEMYJPJNJR-COOH ~metalloproteinase 2.0E-05 ADI47590.1
34a unknown(~Protein) 60 26,398.0 TICTop-downMS1,
MassfromSDS-Page
34b unknown(~Protein) 35 27,603.0 TICTop-downMS1,
MassfromSDS-Page
35a (493.29)-VSWGSCAQK-COOH 6.6E-02 Q58G94.1 170 DenovofromTrypsin-digest
(continuedonnextpage)
170 B.Goۍmenetal./Toxicon107(2015)163e174
Table1(continued)
~Thrombin-likeenzyme
gyroxinB2.1
35bþ36a NH2-JJEWSER-COOH ~metalloproteinaseF1 1.7E-02 AJC52543.1 50 25,225.0 DenovofromTrypsin-digest,
MassfromSDS-Page,
Top-downMS1
(37.19)-QJYYTPR-COOH ~H3metalloproteinase 7.0E-03 AGL45259.1
(339.04)-JTPEEKR-COOH ~H3metalloproteinase 4.9E-01 AGL45259.1
36a unknown(~Protein) 27 TICTop-downMS1,
MassfromSDS-Page
36b unknown(~Protein) TICTop-downMS1,
MassfromSDS-Page
Fig.3. Top-downspectrumofchemicallyreducedpeak11,adisintegrin-likevenomcompound.PanelAshowsthedeconvolutedMS1spectrumofpeak11withamolecularmassof
6047.5Da(averagemass).InBthedeconvolutedMSspectraofthe7timeschargedprecursor(864.94m/z)isshown.Inthelowmassrangethey-ionsandinthehighmasszoomthe
b-ionseriesisannotated,theresultingsequencetagisdisplayedabove(JstandsforeitherIorL).TheresultingdatabasehitisshowninTable2.
B.Goۍmenetal./Toxicon107(2015)163e174 171
Fig.4. Semi-quantitativevenomcompositionofViperaanatolica.Thepiechartrepresentstherelativeoccurrenceofdifferenttoxinfamilies.Inouranalysisweidentifiedasnake
venommetalloproteaseinhibitor(SVMPi);aKunitz-typeproteaseinhibitor;disintegrins;PhospholipasesoftypeA2(PLA2s);aCRISP,cysteine-richsecretoryproteins;SVMPs,snake
venommetalloproteinase;aSVSP,snakevenomserineproteases,andaC-typelectin.
Kunitz-typeproteaseinhibitor(0.3%). lines: CACO-2, human colon carcinoma epithelial cells; MCF-7,
AsthefirstlineofquantificationreliesontheUVsignalfromthe human breast adenocarcinoma epithelial cells; U87MG, human
HPLCrun,themolarextinctioncoefficientεofthepeptidesplaysa glioblastoma-astrocytoma epithelial-like cells; PC3, human pros-
crucialroleforthepeakareasofthetoxins.Thus,ourquantification tate epithelial cells; HeLa, human cervical epithelial carcinoma
is based on the simple assumption that these compound-specific cells; MPanc-96, humanpancreatic fibroblast cells; A549, human
values are nearly equal. This indeed might be the case for com- lungepithelialcells;HEK293,humanembryonicepithelialkidney
pounds of the same protein family but is rather unlikely for the cell; Vero, African green monkey fibroblast-like kidney cells. We
comparisonofpeptideswithhighmolecularmassproteins.Espe- thereforeperformedinitialcytotoxicityassaysbymeansofanMTT
ciallytheconcentrationoftheSVMPi,containingaTrpresidueat assaywhichmeasuresthemitochondrialreductaseactivityofcells
the C-terminus, might be overestimated as ε-value is most likely andthereforeisameasureforviabilityofcells.InFig.5itcanbe
higherthanoftheothercompounds.Thusthevenomcomposition seenthatcrudevenomoftheAnatolianMeadowViperinhibitscell
hastobeconsideredassemi-quantitative. viability in a dose-dependent manner. The IC50 values for all
affectedcelllinesareshowninTable2.MCF-7,MPanc-96andA549
3.4. Medicalimplications were herby the most resistant cell lines against all venom doses
tested (IC50 > 50 mg/mL), while PC3, HEK293, CACO-2, Vero and
Incomparisonwithotherviperspecies,effectsthroughenven- HeLacellswereonlymoderatelyaffected(IC50~4.9;5.5and7.2;
omation caused by V. anatolica are generally low and not life 11.7 and 13.2 mg/mL respectively). The highest activity was
threatening.AftertwoaccidentalbitecaseswithV.anatolicaduring observedagainstbraincancercells(U87MG)withanIC50valueof
our field studies, the subjects (healthy, 51 and 28 years old, both ~0.8mg/mL,whichismorethanoneorderofmagnitudelowerthan
male)reportedonlylocalpainanditchatthebittenareawithout against the non-cancerogenous cell lines (HEK293 and Vero).
any other complications. Similar symptoms were reported by Remarkably, the crude venom cytotoxic effects on glioblastoma
Krecsak et al. (Krecsak et al., 2011) for the closely related small- cellsaswellastheselectivitytocancercellsarenotablyhigherthan
sized insectivorous snakes viper Vipera (Acridophaga) ursinii. On of the plant-derived sesquiterpene lactone parthenolide, a prom-
theotherhand,inacasestudyaboutaccidentswithanotherviper ising candidate in anti-cancer research. Microphotography of
speciestobefoundinTurkey,theblunt-nosedviper,Macrovipera treated cells (shown in the supplemental information) showed
lebetinaobtusa,strongclinicalsymptomsandphysiologicaldamage similarresultsincomparisontotheMTTassay.Untreatedcellswere
werereported(Goۍmenetal.,2006b).Thedifferenceinthetoxicity homogeneously distributed in the culture plates showing typical
canbeexplainedbythebiggersizeandhigheramountofvenom morphological characteristics. Whereas morphological changes
injected by M. lebetina obtusa, but also the higher abundance of wereobservedaftertreatmentwithcrudevenomfor48hvaried
PLA2(~34%vs.~8%V.anatolica)(IgciandDemiralp,2012)mightbe depending on the origin of the cell lines. Increasing venom con-
a reason for the stronger necrotic effects as well as swelling- centrationsresultedinahighernumberofrounded,detachedcells,
associatedsymptoms. elevated mobility and distribution of cells on the surface, multi-
cellular aggregate formation and growth inhibition compared to
untreated cells. In addition, cell disorganization and large areas
3.5. Cytotoxicityscreening
without cells were observed with increasing venom
concentrations.
To assess the potential of the toxins against cancer cells, we
Becausetheamountoffractionatedvenomgenerallywerevery
assessedthecytotoxicityofcrudevenomagainstthefollowingcell
172 B.Goۍmenetal./Toxicon107(2015)163e174
Fig.5. Viabilityofcancerandnon-cancerouscelllinesaftercrudevenomtreatmentfor48h.CellviabilitywasdeterminedbyMTTassay,controlwasexposedtovehicleonlywhich
wastakenas100%viability.CACO-2,humancoloncarcinomaepithelialcells;MCF-7,humanbreastdenocarcinomaepithelialcells;U87MG,humanglioblastoma-astrocytoma
epithelial-like cells; PC3, humanprostateepithelial cells;HeLa, human cervical epithelial carcinomacells;MPanc-96,humanpancreaticfibroblastcells;A549,human lung
epithelialcells;HEK293,humanembryonicepithelialkidneycell;Vero,Africangreenmonkeyfibroblast-likekidneycells.
Table2 inthesupplementalinformation).Hereby,theonlyactivefraction
IC50valuesofV.anatolicacrudevenomandParthenolidetreatedcelllines.CACO-2, (Peak11),containingadimericdisintegrin,hadasignificantcyto-
human colon carcinoma epithelial cells; MCF-7, human breast adenocarcinoma
ehpuimthaenliaplrcoesltlast;eUe8p7iMthGe,lihaulmcealnls;glHioebLlaa,sthoummaa-nasctreorcvyictaolmeapeitphiethliealliacla-rlickineocmellas;cPeCll3s;, 0to.5x1ic±e0f.f0e4ctmgo/nmlgwlihoibclhasitsomsliaghtcleyllbsettweriththaanntheIC5IC050voaflucerudoef
MPanc-96,humanpancreaticfibroblastcells;A549,humanlungepithelialcells; venom (0.75 mg/mL). However comparing the fractions activity
HEK293, human embryonic epithelial kidney cell; Vero, African green monkey withtheactivityofparthenolidewefoundatwofoldhighercyto-
fibroblast-likekidneycells. toxicity. If the molecular mass ratio between parthenolide and
Celllines V.anatolicavenom[mg/mL] Parthenolide[mg/mL] disintegrin (248 Da: ~14.0 kDa) are considered, then the molar
CACO-2 7.21±0.12 0.69±0.08 activityofpeak11(~36nM)isaround100timeshigherthanthatof
MCF-7 >50 1.17±0.06 parthenolide.Incomparisontothecytotoxiceffectsofrecombinant
U87MG 0.75±0.010 1.93±0.05 disintegrins,r-viridistatin2andr-mojastin1,peak11wassignifi-
PHCe3La 41.38.515±±0.01.813 13..2383±±00..0172 ocafn1t0ly.6manorde8a.7ctimvMe(aLguaciennstaBeXtPaCl.-,320p1a5n)c.rTehateicdciseilnlstewgritinhcICo5lo0mvabliuse-
MPanc-96 >50 0.58±0.02
A549 >50 1.66±0.09 tatin, isolated from Bothrops colombiensis, on the other hand
HEK293 5.52±0.21 0.63±0.04 showedsimilarIC50againstT24andSK-Mel-28cells(4.4mMand
Vero 11.65±0.16 2.58±0.10 33nM)(Sanchezetal.,2009).
Cytotoxicactivityofdisintegrinsagainstglioblastoma,hasbeen
described to be caused by targeting specific integrins ultimately
low(0.04e9.03mg)theisolatedtoxinswereonlytestedagainstthe resulting in decrease of tumor growth, affecting invasion and
mostsensitivecellline(U87MG)ofthecrudevenomcytotoxicity migration of tumor cells in carcinogenesis (Calvete et al., 2005;
results.Theresultsforpeak11collectedbysemipreparativeHPLC Macedo et al., 2015). Cell adhesion and migration are important
areshowninFig.6(Theresultsoftheotherfractionscanbefound stages in metastasis development in which integrins, a class of
Fig.6. A:ViabilityofU87MGcellsaftertreatmentwithfraction11for48h.CellviabilitywasdeterminedinanMTTassay,normalizedtoacontrol(100%viability).B:Effectofthe
isolateddisintegrinonU87MGcells.Cellsweretreatedwithfractionatedvenom(fraction11)for48hat37(cid:3)C.1:untreated,2:treatedwithpeak112mg/ml,3:treatedwith
parthenolide,1.25mg/ml.