Enzymatically-tailored pectins differentially influence the morphology, adhesion, cell cycle progression and survival of fibroblasts

9 pages
of 9
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Enzymatically-tailored pectins differentially influence the morphology, adhesion, cell cycle progression and survival of fibroblasts
  Enzymatically-tailored pectins differentially in fl uence the morphology, adhesion,cell cycle progression and survival of   fi broblasts Marie-Danielle Nagel a, ⁎ , René Verhoef  b , Henk Schols b , Marco Morra c , J. Paul Knox d , Giacomo Ceccone e ,Claudio Della Volpe f  , Pascale Vigneron a , Cyrill Bussy a , Marlène Gallet a , Elodie Velzenberger a ,Muriel Vayssade a , Giovanna Cascardo c , Clara Cassinelli c , Ash Haeger d , Douglas Gilliland e ,Ioannis Liakos e , Miguel Rodriguez-Valverde f  , Stefano Siboni f  a Domaine Biomatériaux-Biocompatibilité, UMR CNRS 6600, Université de Technologie de Compiègne, BP 20529, 60205 Compiègne Cedex, France b Wageningen University, Department of Agrotechnology and Food Sciences, Laboratory of Food Chemistry, PO Box 8129, 6700 EV Wageningen, The Netherlands c Nobil Bio Ricerche srl, Strada San Rocco 36, 14018 Villafranca d'Asti (AT), Italy d Centre for Plant Sciences, University of Leeds, Leeds LS2 9JT, United Kingdom e European Commission, Institute for Health and Consumer Protection, TP-203, BMS, Via E. Fermi, 21020 Ispra (VA), Italy f  Department of Materials Engineering and Industrial Technologies, University of Trento, via Mesiano, 38100 Povo (TN), Italy a b s t r a c ta r t i c l e i n f o  Article history: Received 12 February 2008Received in revised form 4 April 2008Accepted 17 April 2008Available online 25 April 2008 Keywords: Pectin modi fi ed hairy region bioactivityPectin/ fi bronectin interactionPectin grafted-biomaterial Improved biocompatibility and performance of biomedical devices can be achieved through the incorpo-ration of bioactive molecules on device surfaces. Five structurally distinct pectic polysaccharides (modi fi edhairy regions (MHRs)) were obtained byenzymatic liquefaction of apple (MHR-B, MHR-A and MHR- α ), carrot(MHR-C) and potato (MHR-P) cells. Polystyrene (PS) Petri dishes, aminated by a plasma deposition process,were surface modi fi ed by the covalent linking of the MHRs. Results clearly demonstrate that MHR-B inducescell adhesion, proliferation and survival, in contrast to the other MHRs. Moreover, MHR- α  causes cells toaggregate, decrease proliferation and enter into apoptosis. Cells cultured in standard conditions with 1%soluble MHR-B or MHR- α  show the opposite behaviour to the one observed on MHR-B and - α -grafted PS.Fibronectin was similarly adsorbed onto MHR-B and tissue culture polystyrene (TCPS) control, but poorly onMHR- α . The Fn cell binding site (RGD sequence) was more accessible on MHR-B than on TCPS control, butpoorly on MHR- α . The disintegrin echistatin inhibited  fi broblast adhesion and spreading on MHR-B-graftedPS, which suggests that MHRs control  fi broblast behaviour via serum-adhesive proteins. This study providesa basis for the design of intelligently-tailored biomaterial coatings able to induce speci fi c cell functions.© 2008 Elsevier B.V. All rights reserved. 1. Introduction Biomaterial surface properties control cellular and host responses toimplanted biomedical devices [1]. Thus, through the last decade, to im-prove implant integration in host tissues, extensive efforts have concen-trated on surface engineering and immobilization of bioactive molecules.Extracellular matrix (ECM) components have important signalling andregulatory functions in tissue evolution and safety. Hence, naturallyderivedbiopolymerssuchasstructuralproteinsandpolysaccharideshavebeen widely utilized for biomaterial coating, tissue engineering matricesor drug delivery systems [2 – 6]. Moreover, it has been shown that thecontrolofcellbehaviouronsubstratais,forseveralcelltypes,atleastpartlysugar-mediatedandsomeadhesivepeptidesequencescanberecognizedby polysaccharide molecules on the cell membrane [7 – 10]. Our previousworkpresentedtheresultsofamultidisciplinaryeffortaimedatproducinga proof-of-concept on the use of pectic polysaccharides in the surfacemodi fi cation of medical devices [11]. We demonstrated that the engineering of plant-derived pectins can be a valuable tool to preparenovel and  fi nely tuned polysaccharides (differing in charge, molecularweight, degree of branching and acetylation) to be used in the surfacemodi fi cation of medical devices and materials [11].Pectinsarehighlystructurallycomplexandconsistof  “ smooth ” and “ hairy ”  regions, the precise structure and relative amounts of whichdiffer depending on the cell type and plant of srcin [12]. The smoothregion consists of homogalacturonan (HGA) and xylogalacturonan(XGA), both built of a (1 → 4)- α - D -Gal  p A backbone. XGA is substitutedbysingle β - D -Xyl  p -(1 → 3)units[13].BothHGAandXGAcanbemethyl-esteri fi ed at position C-6, HGA can be substituted by O-Acetyl moie-ties at the O-2 and -3 positions [14,15]. The hairy region consists of  rhamnogalacturonan-I(RG-I)whichhasabackboneof  → 2)- α - L  -Rha  p - Biochimica et Biophysica Acta 1780 (2008) 995 – 1003  Abbreviations:  AG-I, arabinogalactan type I; AmPS, aminated polystyrene; DA,degree of acetylation; DM, degree of methylation; ECM, extracellular matrix; Fn,fbronectin; HGA, homogalacturonan; HR, hairy region; MHR, modi fi ed hairy region; PS,polystyrene; RG-I, rhamnogalacturonan-I; RGD, Arg-Gly-Asp; TCPS, tissue culturepolystyrene; XGA, xylogalacturonan; XPS, X-ray photoelectron spectroscopy ⁎  Corresponding author. Tel.: +33 3 44 23 44 21; ^ fax: +33 3 44 20 48 13. E-mail address:  marie-danielle.nagel@utc.fr (M.-D. Nagel). 0304-4165/$  –  see front matter © 2008 Elsevier B.V. All rights reserved.doi:10.1016/j.bbagen.2008.04.002 Contents lists available at ScienceDirect Biochimica et Biophysica Acta  journal homepage: www.elsevier.com/locate/bbagen  (1 → 4)- α - D -Gal  p A-(1 → repeatsbranchedatpositionO-4of therham-nose moiety [16,17]. The side chains attached to position O-4 of the rhamnosecanbeclassi fi edintotwodifferenttypes:a)arabinogalactantype I (AG-I) composed of a predominately linear  β -(1 → 4)-galactanbranched with single  α - L  -Ara  f   units at position O-3 of the galactosylunits[18]andb)linearorbranchedarabinanscomposedof  α - L  -(1,5)- L  -Ara  f   units branched at position O-3 or O-2 with α -linked- L  -arabinosylside chains [14]. Biotechnological techniques are being developed fortheinplantaengineeringofrhamnogalacturonansidechainstructures[19,20] and techniques for tailoring side chain structures in vitro byspeci fi c enzymes are well established [21]. A simple procedure to isolate RG-I rich fractions on semi-large scale is to use commercialpectinase preparations to liquefy the tissues of fruit and vegetables[12,22]. The enzymes degrade the HGA part (smooth regions) of thepectinpresent,leavingthehairyregions(HR).Kokkonenetal.[23]haveshown that by modifying the HR side chains, it is possible to triggerbone-cellattachmentonanarti fi cialmaterial,andourpreviousstudieshave demonstrated that immobilized tailored pectins modulate invitro bone-cell and macrophage behaviour [24].Here, we report that both grafted and soluble pectin MHRs differ-entiallyinteractwith fi broblastsinvitroandstudythepotentialroleof serum-adhesive proteins in mediating the cell responses. To addressthis, polystyrene (PS) Petri dishes, aminated by a plasma depositionprocess (AmPS), were surface modi fi ed by the covalent linking of different enzymatically-modi fi ed hairy regions from apple (MHR-A,MHR- α  and MHR-B), carrot (MHR-C) and potato (MHR-P) pectins.Swiss 3T3  fi broblast adhesion, growth, progression in cell cycle andprogrammed cell death (apoptosis) when cultured on MHR-coated-PSor on tissue culture polystyrene (TCPS) in the presence of solubleMHRs were investigated. The capability of   fi bronectin (Fn) to adsorbonto MHR-B and MHR- α  was compared. Cell morphology and adhe-sion were examined on MHR-B in the presence of echistatin, a disin-tegrin which prevents both α 5 β 1 and  α v β 3 integrins binding to theirrespective ligands: Fn and vitronectin.Thisinnovativeapproachallowsthedirectanalysisofpectinstructure – cellfunctionrelationshipswhichmightbeofconsiderableimportancefor the biomaterial  fi eld. 2. Materials and methods  2.1. Pectic substrata Different pectin modi fi ed hairy regions (MHRs) were prepared as described bySchols et al. [21,22] by treatment of apple (A), carrot (C) and potato (P) pulp with commercial enzyme preparations. The resulting suspension was centrifuged, the juiceultra fi ltrated, and the ultra fi ltration retentate was lyophilized to yield the MHRs. MHR-A, MHR-C, MHR-P and MHR- α  (a new batch of MHR-A), were prepared by using theexperimental pectolytic enzyme preparation Rapidase C600, whilst MHR-B was madeby the use of commercial pectinase/hemicellulase preparation Rapidase Liq+. Bothenzymes were from DSM Food Specialities, Delft, The Netherlands. Characterization of the polymers was performed as previously described [11]. In general the difference between MHR is re fl ected by the sugar composition and the molecular weight dis-tribution. The sugar composition of MHR is presented in Table 1. Since MHR is lackinglongHGAchainexposingcharges,theviscosityofMHRsinsolutionismuchlowerwhencompared to common pectins.Surface modi fi cation of polystyrene (PS) Petri dishes was carried out as follows: PS was fi rst surface-functionalized by the introduction of amino groups via deposition from ally-lamine plasma, as described by Morra et al. [11]. As a second step, MHRs were covalently coupled to the surface amino groups on PS. In brief, 5 mL of an unbuffered 0.5% solution of MHRwasplaced intotheaminated PS and coupledbycarbodiimidemediatedcondensationbetweencarboxylgroupspresentintheMHRandaminogroupsofthesurface.Forthestudyofcellbehaviour incontactwithMHRsinsolution,MHRswerediluted incompletemediumand passed through a 0.45  μ  m  fi lter before incubation with cells. X-ray photoelectronspectroscopy (XPS) analyses were performed according to methods described in  “ Practicalsurface analysis ”  [25].ThepresenceofepitopesintheMHRpectinsampleswasusedtoassesstheef  fi cacyof thegraftingprocessusingELISAsandindicatedthatthegraftingwassuccessful.AbundantepitopesthatoccurredinallsamplesincludedthoserecognizedbymonoclonalantibodiesLM5to1,4-galactan[29]andLM6to1,5-arabinan[30],HGAepitopesboundbymonoclonal antibodies JIM5 and JIM7 [31] and the arabinogalactan-protein glycan epitope bound bymonoclonal antibody JIM13 and that bound by MAC207 [32] were assessed.  2.2. Cell culture Swiss 3T3 albino mouse embryo  fi broblasts (ATCC ref. CCL-92) were grown inDMEMmedium supplementedwith 10%foetalbovineserum(FBS) (Gibco), L  -glutamine(4mMGibco)andantibiotics(penicillin/streptomycin,Gibco)at37°Cunderhumidi fi edatmosphere (10% CO 2 , 90% air). For cell behaviour experiments cells were seeded at10,000 cells/cm 2 and for the echistatin test at 20,000 cells/cm 2 .  2.3. Echistatin assay Beforeseeding,cellswereincubated5minatroomtemperaturewith5 μ  g/mLechistatin(Sigma)incompletemedium.Then,cellswereplatedonTCPSorMHR-B-coatedPetridishesand incubated at 37 °C,10% CO 2 . Cell morphologies were observed at 1, 2 and 3 h.  2.4. Cell harvesting, percentage of adherent cells and proliferation index 48 hour post-seeding, poorly adherent cells and aggregates were collected in culturesupernatants and adherent celllayers washed 3 times with complete mediumcontaining25 mM HEPES buffer. The washes and poorly adherent cell suspension were pooled andcentrifuged for 4 min at 200 ×g at room temperature. Aggregates were dissociated byincubation for 5 min at 37 °C in 0.25% trypsin+1 mM EDTA solution (Gibco). At the sametime,adherentcellsweredetachedbyincubationfor5minat37°Cin0.25%trypsin+1mMEDTAsolution(Gibco).Inbothcases,trypsinreactionwasstoppedbyaddinghalfavolumeofFBS.CellswerecountedinaMalassezhaemocytometer.Thepercentageofadherentcellswascalculatedasfollows:(adherentcellnumber/totalcellnumber)100.Theproliferationindex was the ratio between total cell counts and cell number seeded.  2.5. Cell cycle experiments Cell cycle analysis was performed on adherent or poorly adherent cells. Cells werewashed with 1 mLof PBS containing 5 mMEDTAthen fi xed for45 min at4 °Cin 1 mLof 75% ethanol in PBS with 5 mM EDTA. Cells were washed and suspended in PBS, 5 mMEDTA containing 0.1% Triton X-100 (Promega), mixed with 40  μ  g RNase A (Sigma) and25  μ  g propidium iodide (PI, Sigma), and incubated for 15 min protected from light. Thestained samples were analysed in an Epics XL-MCL   fl ow cytometer (Beckman Coulter).Histograms were analysed using Wincycle software.  2.6. Apoptosis quanti  fi cation Annexin-Vlabellingrevealstheearlystagetranslocationofphosphatidylserinefromtheinner to the outer of the bilayer plasma membrane. We used the Annexin-V-FITC kit fromBeckmanCoulterandfollowedthemanufacturer'sinstructions.Brie fl y,48hourpost-seeding,cells were harvested as described in Section 2.4 then cells were washed with PBS andresuspended in the binding buffer at 5.10 5 to 5.10 6 cells/mL.1 μ  L of Annexin-V solution and5 μ  LofPIsolutionwereaddedto100 μ  Lofcellsuspension.Cellswereincubatedfor15minoniceinthedark,thendilutedbyadding400 μ  Lofbindingsolution.ThepositivepopulationforAnnexin-V binding was detected with an Epics XL-MCL  fl ow cytometer (Beckman Coulter).  Table 1 Sugar and substituent composition (in mol%, rhamnose/galacturonic acid ratio and total sugar in w/w%) and XPS analysis of the different MHR samplesMHR Fuc Rha Ara Gal Glc Xyl GalA GlcA O-Me O-Ac Rha/GalA Total w/w% O/C XPS data-B 0 11 11 20 3 18 37 nd 34 11 0.30 78 0.5-P 1 10 14 28 4 nd 42 nd 57 59 0.24 70 0.58-C 2 22 8 32 2 2 30 nd 16 46 0.75 82 0.57-A 0 9 51 10 0 8 22 2 42 60 0.41 86 0.53- α  1 15 25 15 1 12 31 nd 37 45 0.49 86 0.52O-Me: Moles MeOH per 100 mol GalA (max 100%).O-Ac: Moles Ac per 100 mol GalA (max 200%).nd: non detectable.MHR-B, MHR-A and MHR- α  were obtained by enzymatic liquefaction of apple.MHR-C was obtained by enzymatic liquefaction of carrot.MHR-P was obtained by enzymatic liquefaction of potato.996  M.-D. Nagel et al. / Biochimica et Biophysica Acta 1780 (2008) 995 – 1003   2.7. Characterization of Fn adsorption and conformation Lyophilized human plasmatic Fn (Roche) was reconstituted with sterile distilledH 2 O to 1 mg/mL. Working solution was adjusted at 2  μ  g/mL in PBS. TCPS and MHR-grafted 96 well plates were coated with Fn for 45 min at 37 °C, thenwashed 3 times for10 min at roomtemperature in PBS-0.1% Tween.Blockingwas performed in BSA-Tweensolution (0.25% BSA and 0.1% Tween 20 in PBS) for 30 min at 37 °C.Samples were incubated in a solution of anti-Fn antibodies for 1 h at 37 °C(polyclonal Dako ref A0245: 1/5000; MAbs Chemicon 1926, 1935,1936, 1892: 1/5000;dilution in the BSA-Tween solution). After rinsing with the BSA-Tween solution,samples were incubated with horse-radish peroxydase-conjugated secondary anti-bodies (Sigma A6154 orUptima UP446330,1/10,000inthe BSA-Tweensolution) for1hat 37 °C. After rinsing, 100  μ  L of the substrate (5 mg OPD (Sigma) dissolved in 10 mL citrate buffer and 4  μ  L H 2 O 2  30% (v/v)) was added on each sample and incubated for2 min atroomtemperature,protectedfrom light.Reactionwasstoppedwith 100 μ  LHCl1 M, and absorbance ( λ =490 nm) was read in a microwell plate reader. 3 independentexperiments were performed in quadruplicate.  2.8. Statistical analysis All statistical evaluations were performed using GraphPad InStat software.Continuous variables are expressed as means±standard deviation. As most of thecontinuous values measured had a non-Gaussian distribution, non-parametric Mann – Whitney and Kruskal – Wallis tests were performed for comparisons. A value of   p b 0.05was considered as signi fi cant. 3. Results Thesugarandsubstituentcompositionofthevariouspecticmaterialsis presented in Table 1. The differences between the three apple MHR samplesaremainlyduetotheamountofarabinosepresent.AccordingtoSchols et al. [22], arabinose is mainly derived from branched arabinansidechains.TheshapeofarabinansisthemaindifferencebetweenMHR-A, - α and -B. Furthermore MHR-B has a lowerdegree of acetylation (DA)and is relatively enriched in galactose and xylose, compared to the twoother apple MHRs. With respect to its side chains, potato MHR-P can becharacterized by its high amounts of galactose, which is present in theform of  β -(1 → 4)-galactan inpotato RG-I [26]. MHR-C also contains highamounts of galactose, however it is known that pectin extracts fromcarrot rich in arabinogalactan type II are composed of 1 – 3/1 – 6 and 1 – 3 – 6-linked galactose [21,27,28]. In general the Rha/GalA ratio is a good indication for the amount of RG-I and HGA present. From this ratio it isobvious that MHR-C is most rich in RG-I and MHR-P contains the mostHGA. With respect to the DA and degree of methylation (DM), appleMHR-B has a lowered DA compared to the other two apple MHR samples. In contrast to all other samples, the DM of MHR-C issigni fi cantly lower, the DM of potato MHR-P has the highest value.XPS analyses of the  fi ve parent MHRs are shown in Table 1. Allsampleshavesimilarcomposition as indicated bythe O/C ratiovalues.Surface compositional changes of Petri dishes were analysed afterplasma deposition process and MHR-grafting (Table 2). As expected,TCPS surface contained only carbon (98.5 at.%) with a little amount of oxygen( b 2 at.%).The plasma treatment increased boththe N/C and O/C ratios of AmPS surface whilst the MHRs covalent coupling increasedthe O/C and decreased the N/C ratios of all the surfaces.The presence of epitopes in the MHR pectin samples was used toassesstheef  fi cacyofthegraftingprocessusingELISAsandindicatedthat  Table 2 Behaviour of Swiss 3T3  fi broblasts on MHR-coated Petri dishes 48 hour post-seeding, and XPS analysis The significant differences between TCPS, AmPS and MHR-B are indicated by grey boxes and those between MRH-P, -C, -A and - α  are indicated by brackets. Poorly adherent cells are initalics. ns: non-significant. nd: non-detectable. 997 M.-D. Nagel et al. / Biochimica et Biophysica Acta 1780 (2008) 995 – 1003  the grafting was successful. Abundant epitopes that occurred in allsamples included those recognized by monoclonal antibodies LM5to 1,4-galactan and LM6 to 1,5-arabinan. HGA epitopes bound bymonoclonal antibodiesJIM5and JIM7 were present in samples MHR- α ,MHR-P and MHR-A. The arabinogalactan-protein glycan epitope boundby monoclonal antibody JIM13 was present in all samples and thatbound by MAC207 occurred in samples MHR-C, MHR-A and MHR-B.  3.1. Cell behaviour in contact with MHR-grafted PS  In all cases, TCPS and AmPS Petri dishes were employed ascontrols. Morphological analysis performed 48 hour post-seeding(Fig. 1), showed that cells had spread well on MHR-B, TCPS andAmPS. A few areas of MHR-P were still covered by spread cells butsome cell aggregates were observable. Areas of spread cells were Fig.1.  Swiss 3T3 fi broblast morphology 48 hour post-seeding onTCPS, AmPS controls and MHRs-coated Petri dishes. Cells had spread well on MHR-B, TCPS and AmPS. A fewspreadcellswerestilldetectableonMHR-P.FewerspreadcellswereobservedonMHR-CandMHR-A.OnMHR- α ,allcellswereroundandaggregated.Observationswereperformedusinganinverted microscope Olympus CKX41. Scale bar=100  μ  m. Fig. 2.  Swiss 3T3  fi broblast morphology 48 hour post-seeding on TCPS in complete culture medium supplemented with 1% soluble MHR-B or MHR- α . When cultured with MHR-B,cellswerepoorlyattached,remainedroundandformedaggregates.CellswereadherentandspreadonthebottomofTCPSPetridishesincontrolcultureconditions(withoutMHR)aswell as with MHR- α . Observations were performed using an inverted microscope Olympus CKX41. Scale bar=100  μ  m.998  M.-D. Nagel et al. / Biochimica et Biophysica Acta 1780 (2008) 995 – 1003  reduced and formation of cell aggregates increased on MHR-C andMHR-A, and in the case of MHR- α  all cells remained round andaggregated.In order to compare cell adhesion on the different substrata,poorly adherent cells in culture supernatants and adherent cellsremaining after enzymatic dissociation were counted. Table 2 showsresults of cell behaviour on MHRs. Percentages of adherent cells wereroughly similar on MHR-B, TCPS and AmPS but signi fi cantly lower onMHR-P, -C, -A and - α . Percentages of adherent cells on MHR-P weresigni fi cantly higher than thoseon MHR-C, -A and - α . Cells cultured onMHR- α  had the lowest percentage of adherent cells.Proliferation indices on MHR-B, TCPS and AmPS did not differsigni fi cantly.Bycontrast,thoseobtainedonMHR-P,-C,-Aand- α weresigni fi cantly lower. Results on MHR-A, -C and - α  were similar but allsigni fi cantly different from those gained on MHR-P. In order to extendthecellgrowthanalysis,cellcycleprogressionwasanalysed. Cell cyclestatus on MHR-B, TCPS and AmPS were signi fi cantly different to othertreatments with percentages of cells being lower in G1 phase andhigher in S phase, than those obtained on MHR-P, -C, -A and - α .Within the latter group of pectins MHR-P and MHR- α  also resulted insigni fi cantly different cell cycle percentages. There were no differ-ences in cell cycle progression between adherent and poorlyadherentcells.Theobservationoflowproliferationindicesaswellasthepresenceof cell aggregates encouraged us to study early apoptosis by the assess-ment of cell-surface expression of phosphatidylserine using Annexin-V binding. The percentages of apoptotic adherent cells (PI − /Annexin-V+)were similaron MHR-B, TCPS, AmPS, MHR-P, -Cand -A.Those of poorlyadherent cells on MHR-P, -C, -A and - α , as well as percentages of ad-herent cells on - α , were signi fi cantly higher. Fig. 3.  Fibronectinadsorption/conformationonTCPS andAmPScontrols,and MHR-Band - α coatedmulti-wellplatesassessedbyanELISA-modi fi edmethod.AntiFnpolyclonalantibody(pAb)usedforaglobalrecognitionofadsorbedFnshowedthatFnwassimilarlyadsorbedonTCPSandMHR-B,butsigni fi cantlylessonMHR- α (a).Monoclonalantibodies(mAbs)toFncellrecognitionsite(RGD)showedbetteraccessibilityonMHR-BandverypooronMHR- α comparedtoTCPSandAmPScontrols(b).MAbstogelatin(c),C-terminus(d)andN-terminus(e)sitescon fi rmed on MHR-B a similar or improved accessibility and on MHR- α  a poor accessibility to these sites compared to controls.  ⁎  p b 0.05,  ⁎⁎  p b 0.01, and  ⁎⁎⁎  p b 0.001.  Table 3 BehaviourofSwiss3T3 fi broblastsculturedonTCPSwithout(control)orwith ^ 1%MHR-Bor MHR- α  in solution, 48 hour ^ post-seeding Significant differences between TCPS and MHR- α  are indicated by grey boxes.A-V: Annexin-V; PI: Propidium Iodide. 999 M.-D. Nagel et al. / Biochimica et Biophysica Acta 1780 (2008) 995 – 1003
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks