Therapeutic angiogenesis induced by human umbilical cord tissue-derived mesenchymal stromal cells in a murine model of hindlimb ischemia

RESEARCH Open Access The apeu ic angiogenesis induced by human umbilical co d issue-de i ed mesenchymal s omal cells in a mu ine model o hindlimb ischemia Ana Ri a S. Pe ei a 1† , Te esa F. Mendes 1† , Augus o Minis o 1,2 , Ma iana Teixei a 3 , Ma iana Filipe 3 , Jo ge M. San os 3 , Ri a N. Bá cia 3 , J. Goy i-O’Neill 4 , Faus o Pin o 1,5 , Ped o E. C uz 3 , Helde J. C uz 3 and Susana Cons an ino Rosa San os 1,5* Abs ac Backg ound: Mesenchymal s em cells de i ed om human umbilical co d issue, e med UCX®, ha e he po en ial o p omo e a ull ange o e en s leading o issue egene a ion and homeos asis. The main goal o his wo k was o in es iga e UCX® ac ion in expe imen ally induced hindlimb ischemia (HLI). Me hods: UCX®, ob ained by using a p op ie a y echnology de eloped by ECBio (Amado a, Po ugal), we e deli e ed ia in amuscula injec ion o C57BL/6 emales a e unila e al HLI induc ion. Pe usion eco e y, capilla y and colla e al densi y inc ease we e e alua ed by lase dopple , CD31 immunohis ochemis y and diaphonisa ion, espec i ely. The ac i a ion s a e o endo helial cells (ECs) was analysed a e EC isola ion by lase cap u e mic odissec ion mic oscopy ollowed by RNA ex ac ion, cDNA syn hesis and quan i a i e RT-PCR analysis. The UCX®-condi ioned medium was analysed on Gallios low cy ome e . The capaci y o UCX® in p omo ing ubulogenesis and EC mig a ion was assessed by ma igel ubule o ma ion and wound-healing assay, espec i ely. Resul s: We demons a ed ha UCX® enhance angiogenesis in i o ia a pa ac ine e ec . Impo an ly, a e HLI induc ion, UCX® imp o e blood pe usion by s imula ing angiogenesis and a e iogenesis. This is achie ed h ough a new mechanism in which du able and simul aneous up egula ion o ans o ming g ow h ac o β2, angiopoie in 2, ib oblas g ow h ac o 2, and hepa ocy e g ow h ac o , in endo helial cells is induced by UCX®. Conclusions: In conclusion, ou da a demons a e ha UCX® imp o e he angiogenic po ency o endo helial cells in he mu ine ischemic limb sugges ing he po en ial o UCX® as a new he apeu ic ool o c i ical limb ischemia. Keywo ds: UCX®, Mesenchymal s em cells, Angiogenesis, A e iogenesis, C i ical limb ischemia, Endo helial cells, Hindlimb ischemia Backg ound C i ical limb ischemia (CLI) is a se e e o m o pe iph- e al a e y disease (PAD) in which pa ien s wi h occlu- si e a e ial disease o he legs expe ience ch onic ischemic es pain, ulce , o gang ene [1]. This synd ome is associa ed wi h se e e p ognosis, wi h 1-yea mo ali y exceeding 25 % and abou 30–50 % majo limb ampu a- ion a 1 yea om diagnosis [1]. The limi a ions o su gical/endo ascula e ascula iza ion, due he dis ibu- ion and di useness o a e ial occlusions, a e well ecognized and ampu a ion, despi e i s associa ed mo bidi y and mo ali y a es, is o en ecommended [2]. The goal o limb sal age has s imula ed esea ch in o al e na i e me hods, including he apeu ic angiogenesis which can be achie ed ei he by local adminis a ion o * Co espondence: [email p o ec ed].p † Equal con ibu o s 1 Cen o Ca dio ascula da Uni e sidade de Lisboa, A . P o Egas Moniz 1649-028 Lisbon, Po ugal 5 Faculdade de Medicina da Uni e sidade de Lisboa, A . P o Egas Moniz 1649-028 Lisbon, Po ugal Full lis o au ho in o ma ion is a ailable a he end o he a icle © 2016 The Au ho (s). Open Access This a icle is dis ibu ed unde he e ms o he C ea i e Commons A ibu ion 4.0 In e na ional License (h p://c ea i ecommons.o g/licenses/by/4.0/), which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e Commons license, and indica e i changes we e made. The C ea i e Commons Public Domain Dedica ion wai e (h p://c ea i ecommons.o g/publicdomain/ze o/1.0/) applies o he da a made a ailable in his a icle, unless o he wise s a ed. Pe ei a e al. S em Cell Resea ch & The apy (2016) 7:145 DOI 10.1186/s13287-016-0410-4 p o-angiogenic g ow h ac o s and gene- o cell-based he apies [3]. Mesenchymal s em cells (MSCs) hold g ea p omise as a he apy o PAD, mainly due o hei pa ac ine ac i i y and immunosupp essi e capaci y [4]. MSCs a e known o home speci ically o hypoxic issues ollowing inju y [5] whe e hey po en ia e ascula g ow h h ough he elease o p o-angiogenic ac o s [6]. I was p e iously epo ed ha au ologous, allogeneic and xenogeneic MSC adminis a ion induced he apeu ic angiogenesis in animal models o hindlimb ischemia (HLI) [7]. In a me a-analysis o cell he apy o PAD, i was ound ha au ologous bone ma ow MSCs deli e y led o imp o ed indices o ischemia and pain- ee walking [8]. Al hough i s ha es ed om he bone ma ow, MSCs ha e since been iden i ied in many o he issues, namely adipose issue [9] and umbilical co d issue [10]. UCX® in pa icula a e MSCs ob ained om he human umbilical co d Wha on’s jelly ha a e isola ed, expanded and c yo- p ese ed acco ding o a pa en ed me hod (PCT/IB2008/ 054067; WO 2009044379) and p oduced acco ding o ad anced he apy medicinal p oduc (ATMP) guide- lines [11]. UCX® ul il he MSC c i e ia as de ined by he In e na ional Socie y o Cellula The apy (ISCT) [12]. UCX® cells a e ad an ageous in compa ison o o he sou ces due o he absence o in asi eness in hei collec ion p ocess, hei as e sel - enewal, highe cell yield and being mo e po en modula o s o he immune sys em han bone ma ow MSCs, making hem mo e a ac i e o allogeneic cellula he apies. Recen ly, i was demons a ed ha UCX® pa ac ine ac i i y ep esses T cell ac i a ion and p omo es he expansion o egula o y T cells (T egs) be e han bone ma ow MSCs [13]. Fu he mo e, by using an acu e a h i is in i o model i was ound ha UCX® can e- duce paw edema mo e e icien ly han bone ma ow MSCs. The use o a ch onic a h i is model showed ha UCX® induce as e emission o local and sys emic a h- i ic mani es a ions [13]. The UCX® issue egene a ion capaci y sugges ed o heuma oid a h i is was la e co - obo a ed o myoca dial in a c ion. By using a mouse model o myoca dial in a c ion, i was shown ha UCX® p ese e ca diac unc ion and a enua e ad e se issue emodelling a e in amyoca dial ansplan a ion [14]. In e es ingly, i was demons a ed ha his ca diop o ec- i e e ec was exe ed h ough pa ac ine mechanisms in ol ing angiogenesis p omo ion [14]. Fu he mo e, in i o s udies, pe o med wi h condi ioned medium (CM) p oduced by UCX® g own in classical wo-dimensional monolaye cul u es, ha e demons a ed he po en ial o induce ke a inocy e mig a ion in he ea ly s ages o wound healing [15]. Mo eo e , UCX® we e able o a ac bone ma ow MSCs in i o wi h po en ial o p omo e he o ma ion o g anula ion issue, con ac ion by myo- ib oblas s, angiogenesis, asculogenesis and epi heliza- ion [15]. These esul s s ongly sugges ha UCX® ha e he po en ial o p omo e a ull ange o e en s leading o issue egene a ion and homeos asis. In e es ingly, i was shown ecen ly ha he la e p oli e a i e and e- modelling s ages o wound healing p omo ed by he CM o he wo-dimensional monolaye cul u es we e imp o ed by using h ee-dimensional cul u e-de i ed CM [16]. An enhanced sec e ion o healing-inducing pa ac ine ac o s by UCX® including he ex acellula cellula ma ix me allop o einase-2 (MMP-2) and ma ix me allop o einase-9 (MMP-9), collagen I, ib o- nec in, laminin, collagen IV and angiogenic ac o s such as ascula endo helial g ow h ac o A (VEGF-A), g anulocy e-colony s imula ing ac o (G-CSF), ans o m- ing g ow h ac o β1 (TGF-β1), ib oblas g ow h ac o 2 (FGF-2), hepa ocy e g ow h ac o (HGF) and in e leukin 6 (IL-6) was obse ed in he CM o h ee-dimensional cul- u es based on sel -assembled sphe oids [16]. In his wo k, ou da a co obo a e ha UCX® enhance angiogenesis in i o and in i o, by modula ing endo- helial cells. In i o, we ound ha UCX® p omo e ubule o ma ion and endo helial cell mig a ion. Impo - an ly, ou esul s ha e demons a ed ha UCX® in he se ing o expe imen ally induced unila e al hindlimb ischemia (HLI) s imula e angiogenesis and colla e al es- sel de elopmen and he eby imp o e blood pe usion in he ischemic limb demons a ing he po en ial o UCX® as new he apeu ic ool o CLI. Fu he mo e, and o he bes o ou knowledge, his is he i s s udy in an ex- pe imen al model o hindlimb ischemia showing ha , in i o and in a sus ained way human MSCs up egula e he endo helial gene exp ession o se e al p o-angiogenic playe s (such as ans o ming g ow h ac o β2(Tg β2), angiopoie in 2 (Ang-2), Fg -2,Hg ), shedding ligh on he po en ial mechanisms o ac ion o his pa icula ATMP. Me hods E hics and egula ions All animal p ocedu es we e pe o med acco ding o Di ec - i e 2010/63/EU. The p ocedu es we e app o ed by he in- s i u ional Animal Wel a e Body and licensed by DGAV, he Po uguese compe en au ho i y o animal p o ec ion (license numbe 023861/2013). Cell cul u e UCX® we e cul u ed in s a ic monolaye s in α-MEM wi h 1 g/L glucose and 2 mM glu amine (Sigma-Ald ich, S . Louis, MO, USA), he ea e designa ed basal medium (BM), supplemen ed wi h 20 % oe al bo ine se um (FBS; Gibco®, Mad id, Spain), in a humidi ied incuba o a 37 °C and 7 % CO 2 . Pe ei a e al. S em Cell Resea ch & The apy (2016) 7:145 Page 2 o 12 Human umbilical ein endo helial cells (HUVECs) (Sciencell, Ca lsbad, CA, USA) we e cul u ed in M199 media (Sigma-Ald ich, S . Louis, MO, USA), he ea e designa ed endo helial basal medium (EBM) supple- men ed wi h 10 % FBS (Gibco™, The mo Fishe Scien- i ic, Wal ham, MA, USA), 50 μg/ml endo helial cell g ow h supplemen (ECGS) (Sigma-Ald ich), 100 μg/ml hepa in (Sigma-Ald ich) and 1 % penicillin-s ep omycin (10,000 U/mL, Sigma-Ald ich), he ea e designa ed endo helial g ow h medium (EGM). Cells we e g own in lasks coa ed wi h 0.2 % gela in (Sigma-Ald ich) un il 70 % con luence and used up o passage 6. P epa a ion o condi ioned media UCX® we e seeded in BM supplemen ed wi h 5 % FBS and g own un il 90 % con luence. A e washing cells wi h phospha e-bu e ed saline (PBS), BM (wi hou se um) was added o 24 hou s. The ea e , BM was e- placed and condi ioned o 48 hou s a 37 °C and 7 % CO 2 a e which i was collec ed, cen i uged a 300 g o 10 minu es o emo e cell deb is, il e ed (0.22 μm po e size, Me ck Millipo e, Bille ica, MA, USA), and concen- a ed using 5 kDa cu -o spin concen a o s (Agilen Technologies, San a Cla a, CA, USA) as pe manu ac- u e ’s ecommenda ions. Con ol condi ioned medium samples unde wen he same p ocedu e as desc ibed abo e, bu in he absence o cells. Quan i ica ion o sec e ed ac o s The quan i ica ion o HGF, VEGF-A, TGF-β1, IL-8, pla ele -de i ed g ow h ac o -AA (PDGF-AA) and FGF-2 in he condi ioned medium was pe o med using a comme cially a ailable ki (FlowCy omix™; eBioscience, San Diego, CA, USA) acco ding o he manu ac u e ’s ins uc ions. Samples we e acqui ed on a Gallios low cy ome e (Beckman Coul e , B ea, CA, USA) and he esul s ob ained using FlowCy omix P o 3.0 So wa e (eBioscience). Tubule o ma ion assay The ubule o ma ion assay was pe o med using he hick gel me hod o p epa a ion. B ie ly, a e hawing o e - nigh , he ma igel (Co ning, Co ning, NY, USA) was pla ed in o a p e-cooled 48-well pla e (Nunc™, The mo Fishe Scien i ic) (180 μl pe well) using a chilled pipe ip. A e polyme iza ion o ma igel a 37 °C, 5 % CO 2 , o 45 minu es, HUVECs we e inocula ed a a densi y o 4.5 × 10 4 cells/cm 2 on op o he ma igel in 350 μlo EBM. Cells we e hen incuba ed o 1 hou a 37 °C, 5 % CO 2 .1×10 6 UCX® we e esuspended in 350 μlBMand loaded on a 1-μm inse (B and, We heim, Ge many) and ca e ully placed in a 48-well pla e and incuba ed o 16 hou s a 37 °C, 5 % CO 2 . Con ols wi h no UCX® included HUVECs in (i) EBM, (ii) EGM and (iii) EBM supplemen ed wi h FGF-2 (50 ng/ml). A e incuba ion, inse s we e emo ed and all wells we e pho og aphed (×10 ampli ica ion) using a Nikon Eclipse Ti-U in e ed mic oscope wi h a conjuga ed Nikon DS-Qi1Mc came a (Nikon, Tokyo, Japan). Tubule o ma ion was quan i ied on ou andom ields pe eplica e, using he Angiogen- esis analyse om Image J. Wound-healing assay HUVECs we e pla ed o con luence in a 24-well pla e (p e iously coa ed wi h gela ine 0.2 %) wi h EBM supplemen ed wi h 5 % FBS. A e cul u ing o e nigh , wounds we e c ea ed in he monolaye by sc apping he pla e wi h a s e ile pipe e ip. A e ha , 1 × 10 6 UCX®, esuspended in EBM supplemen ed wi h 5 % FBS and loaded on a 1-μm inse (B and, We heim, Ge many), we e ca e ully placed in a 24-well pla e and incuba ed o 9 hou s a 37 °C, 5 % CO 2 . In con ols, he same p o- cedu e was pe o med in he absence o UCX®. Pho o- g aphs we e aken in P imo e (Ca l Zeiss Mic oscopy, Jena, Ge many) on a × 4 magni ica ion and h ee inde- penden wound a eas we e measu ed wi h ImageJ imme- dia ely a e wounding and 9 hou s la e . Mice Twen y- wo-week-old C57BL/6 emale mice, pu chased om Cha les Ri e Labo a o ies, Ba celona, Spain, we e used in all expe imen s. The animals we e anaes he ised wi h a ke amine-mede omidine cock ail (75 mg/kg BW and 1 mg/kg BW, espec i ely) in ape i oneally o he su gical p ocedu e as well as o he o he analysis p ocedu es. The anaes hesia was pa ially e e ed wi h a ipamezole (5 mg/kg BW). Pos ope a i ely, analgesia was pe o med (bup eno phine 100 μl/15–30 g BW e e y 8–12 h) and he animals we e closely moni o ed. HLI model A su gical p ocedu e was pe o med o induce unila e al HLI in he mice. B ie ly, an incision in he skin o e lying he high o he igh hindlimb o each mouse was made and he dis al ex e nal iliac a e y and he emo al a e y and eins we e liga ed and excised. The ein was liga ed bo h o inc ease he se e i y o he ischemia as well as o inc ease he echnical ep oducibili y o he model, as isola ion o he emo al a e y alone o en esul s in ea - ing o he ein esul ing in haemo hage. UCX® adminis a ion A dose o 2 × 10 5 UCX® was injec ed in amuscula ly, 5 hou s pos -ischemia induc ion on he gas ocnemius muscle ( igh hindlimb) in a olume o 50 μl pe animal, di ided in o wo injec ions o 25 μl. As a con ol, UCX® ehicle (PBS) was adminis e ed a he same condi ions. Pe ei a e al. S em Cell Resea ch & The apy (2016) 7:145 Page 3 o 12 Lase dopple pe usion imaging The lase dopple pe usion image (Moo LDI V6.0, Moo Ins umen s L d, Axmins e , UK) was used o as- sess limb pe usion. Hai was emo ed 1 day be o e lase dopple analysis using an elec ical sha e ollowed by depila o y c eam. Blood low was measu ed bo h in he ischemic leg and he con ala e al one, be o e HLI in- duc ion (PRE-HLI), immedia ely pos -HLI (POST-HLI), and a day 7, 14 and 21 pos -HLI (d7 POST-HLI, d14 POST-HLI and d21 POST-HLI, espec i ely). Colou - coded images o issue pe usion we e eco ded and poo o no pe usion was displayed as da k blue, and he highes pe usion le el was displayed as ed. Mean lux alues we e calcula ed using he Moo LDI V6.0 image p ocessing so wa e. To accoun o a iables such as empe a u e and ambien ligh , blood pe usion is exp essed as he a io o ischemic o non-ischemic limb. The mice we e placed on a 37 °C hea ing pad o educe hea loss du ing measu emen s. Immunohis ochemis y and capilla y densi y analysis Mice we e sac i iced a day 90 pos -HLI ( o assu e capil- la y s abiliza ion a e he ischemic inju y). The gas o- cnemius muscles o bo h legs we e ha es ed, placed in ans e se o ien a ion on a small co k disc wi h he help o 10 % agacan h, snap ozen in liquid ni ogen- cooled isopen ane and s o ed a -80 °C un il sec ioned. Se en-mic ome e sec ions we e labeled wi h CD31 monoclonal an ibody (Pha mingen, San Diego, CA, USA). A e ixa ion in ace one o 10 minu es, hyd ogen pe oxidase (0.3 % dilu ed in me hanol) was added o 30 minu es, a oom empe a u e (RT) and ollowed by wo washes in PBS o a o al o 10 minu es. Blocking solu ion (5 % abbi se um in PBS) was applied o 30 minu es, a RT, and he slides we e hen incuba ed o 1 hou a RT wi h a monoclonal an ibody agains mouse CD31 a 1:500, dilu ed in 1 % bo ine se um albu- min (BSA) in PBS. A e h ee washes in PBS o a o al o 30 minu es, a seconda y bio inyla ed abbi an i- a IgG an ibody was added a 1:200 in 1 % BSA in PBS and 5 % abbi se um o 30 minu es, a RT. Washes we e pe o med as be o e and labeled a idin-conjuga ed pe - oxidase complex (Vec as ain ABC ki ; Vec o Labo a o - ies, Bu lingame, CA, USA) was used o colo de elopmen acco ding o he manu ac u e ’s ecom- menda ions o 30 minu es, a RT. A e insing in PBS ( h ee imes o 5 minu es), DAB pe oxidase subs a e ki (Dako, Glos up, Denma k) was added o 5 minu es o localize he immune complexes. The sec ions we e coun e s ained wi h haema oxylin (Me ck, Kenilwo h, NJ, USA) o 10 seconds and moun ed wi h en elan (Me ck). Omission o he i s an ibody was used as a nega i e con ol. Analysis o issue samples was conduc ed using a Leica DM2500 up igh b igh ield mic oscope (Leica Mic osys ems, We zla , Ge many). Capilla y densi ies, i.e. numbe o capilla ies pe numbe o myocy es, we e measu ed in wo di e en sec ions o ou dis inc ana omic a eas o each specimen using he ImageJ so wa e. Con as agen pe usion and diaphonisa ion Nine y days pos -ischemia induc ion ( o assu e ascula s abiliza ion a e he ischemic inju y), mice we e deeply anaes he ised and he o so and limbs we e sha ed. A medial ho aco omy was pe o med o expose he hea and a needle (26 gauge), a ached o an au oma ic in- jec o , was in oduced in he le en icle. An incision was pe o med in he igh a ium o allow enous d ainage. Mice we e p ima ily pe used wi h hepa inized se um (3000 IU/L) un il he blood was comple ely e- mo ed om ci cula ion. A asodila a ion mix u e o ad- enosine (1 mg/L) and papa e ine (4 mg/L) was subsequen ly adminis e ed, igh be o e he con as agen , o 2 minu es. The con as used was a mix u e o ba ium sul a e (50 %) and gela in (5 %). This solu ion was kep wa m un il he injec ion ime o a oid hicken- ing. Con as agen was pe used manually un il he ee blanched. Righ a e he injec ion he mice we e ans- e ed o a cold chambe , so ha he con as agen became solidi ied. All he solu ions we e injec ed wi h a pe usion a e o 0.7 mL/minu e. Then, he skin o each mouse was emo ed om he lowe body and diaphoni- sa ion was pe o med by using a modi ied e sion o he Spal eholz echnique. B ie ly, he mice we e ixa ed, decalci ied, whi ened, washed, dehyd a ed by eeze subs i u ion and placed in o a acuum pump wi h Spal eholz solu ion (benzyl benzoa e and me hyl salicyl- a e) un il anspa ency was acqui ed. Colla e al essels quan i ica ion Mice we e kep in Spal eholz solu ion du ing image acquisi ion, in o de o achie e a homogenous densi y be ween he issues and he media, wi h minimal abso p- ion o e lec ion o ligh . Mice en i e limbs we e pho o- g aphed in a magni ie wi h a ligh sou ce. A e acquisi ion, images we e aligned and s i ched oge he using Adobe Pho oshop CS6®, and en i e limb pho o- g aphs we e ob ained. In o de o exclude he emo al a e y and all enous s uc u es om his quan i ica ion, only he colla e al essels we e manually segmen ed by highligh ing hem using Adobe Pho oshop CS6®. We conside ed as colla e al essels all he essels wi h a diam- e e be ween 20 and 300 μm. Fo e e y mouse, an ana- omically de e minable egion comp ising he liga ion si e was selec ed and de ined as a egion o in e es (ROI). Colla e al essels densi y (CVD) was quan i ied in equi a- len ROIs co esponding o 20 % o he o al limb a ea. Pe ei a e al. S em Cell Resea ch & The apy (2016) 7:145 Page 4 o 12 The CVD was calcula ed as he a io be ween he ascula and he ROI a eas. To exclude a ia ions in he ana omy, pe usion and diaphoniza ion p ocedu es, he CVD alue o he non- ischemic limb ob ained o each mouse was assumed o co espond o 100 %. Acco ding o his assump ion, he CVD pe cen age in he ischemic limb was calcula ed ela i ely o he non-ischemic one. The pe cen age o CVD inc ease was de e mined as he di e ence be ween he CVD pe cen age among he ischemic and non- ischemic limbs. Means o CVD pe cen ages and s and- a d de ia ions we e calcula ed o each expe imen al g oup. All densi y measu emen s we e pe o med using ImageJ so wa e. Lase cap u e mic odissec ion o capilla ies Mice we e sac i iced a day 70 pos -HLI. Twel e- mic ome e sec ions o he gas ocnemius muscles we e labeled wi h CD31 monoclonal an ibody (Pha mingen). The sec ions we e s o ed a –80 °C un il mic odissec- ion. The immunohis ochemis y p o ocol desc ibed abo e was modi ied o imp o e RNA p ese a ion by using high sal bu e , 2 M NaCl in PBS 1× (a 4 °C) in all incuba ion and washing s eps [17]. B ie ly, slides we e placed in ace one, o 5 minu es, ai -d ied, ehyd a ed wi h 2 M NaCl/PBS (4 °C) and incuba ed, o e nigh , a 4 °C wi h a monoclonal an ibody agains mouse CD31, a 1:500, in 2 M NaCl/PBS. A e wo washes in 2 M NaCl/PBS o a o al o 6 minu es, a seconda y bio inyl- a ed abbi an i- a IgG an ibody was added a 1:200 in 2 M NaCl/PBS and 5 % abbi se um o 30 minu es, a 4 °C. Washes we e pe o med as be o e and labeled a idin-conjuga ed pe oxidase complex (Vec as ain ABC ki ; Vec o Labo a o ies) was used o colo de elopmen acco ding o he manu ac u e ’s ecommenda ions o 30 minu es, a 4 °C. A e insing, DAB pe oxidase subs a e ki (Vec o Labo a o ies) was added o 5 mi- nu es o localize he immune complexes. Sec ions we e dehyd a ed in ice-cold 90 % e hanol ollowed by 100 % e hanol and allowed o d y. Ten housand capilla ies we e mic odissec ed using a Zeiss PALM Mic oBeam Lase Mic odissec ion Sys em (Ca l Zeiss Mic oscopy) equipped wi h a pulsed solid-s a e 355 nm lase . Dissec ed capilla ies we e ca apul ed in o a mic o uge ube adhesi e cap. RNA ex ac ion, cDNA syn hesis, p e-ampli ica ion and RT-PCR To al RNA om he mic odissec ed capilla ies was isola ed using an RNeasy Mic o Ki (QIAGEN, Hilden, Ge many), including DNase ea men o emo e po en ial genomic DNA con amina ion. Fo syn hesis and p eampli ica ion o cDNA, RT 2 P eAMP cDNA Syn hesis Ki (QIAGEN) was used wi h wo ounds o p e-ampli ica ion using he ollowing p ime s: Fg -2_F (5′- ACTCCAGTTGGTATGTGGCACTGA-3′); Fg -2_ R(5′-AACAGTATGGCCTTCTGTCCAGGT-3′); Tg b- 2_F (5′-GCTTTGGATGCGGCCTATTGCTTT-3′); Tg b-2_R (5′-CTCCAGCACAGAAGTTGGCATTGT-3′); Ang-2_F (5′-ATCCAACACCGAGAAGATGGCAGT-3′); Ang-2_R (5′-AACTCATTGCCCAGCCAGTACTCT- 3′); Hg _F (5′-GCATTCAAGGCCAAGGAGAAGGT T-3′); Hg _R (5′-TCATGCTTGTGAGGGTACTGCG AA-3′); 18s_F (5′-GCCCTATCAACTTTCGATGGTA GT-3′); 18s_R (5′- CCGGAATCGAACCCTGATT-3′). RT-PCR was pe o med acco ding o he manu ac- u e ’sp o ocolusingPowe SYBR®G een(In i o- gen, Ca lsbad, CA, USA) and an Applied Biosys ems (Wal ham, MA, USA) 7500 Fas Real-Time PCR o he same a ge s desc ibed abo e. The housekeeping gene used o no malize was 18S. The RT-PCR p o- g am consis ed o an ini ial dena u a ion s ep, a 95 ° C, o 10 minu es ollowed by 50 cycles, a 95 °C, o 15 seconds and a 60 °C, o 1 minu e. The ela i e quan i ica ion was pe o med acco ding o he com- pa a i e me hod (2 -ΔΔC ; Applied Biosys ems Use Bulle in no. 2P/N 4303859), wi h he non-ischemic muscle as an in e nal calib a o . The o mula used is 2 -ΔΔC =2 -[ΔC (sample) - ΔC (calib a o )] ,whe eΔC (sample) = C (sample) –C ( e e ence gene). Fo he in e nal calib a o ΔΔC = 0 and 2 0 =1. Fo he emaining samples he alue o 2 -ΔΔC indica es he old change in gene exp ession ela i e o he calib a- o . The ΔC alue o each sample is he a e age o iplica es. S a is ical analysis Fo he wound-healing assay, he alues assume no mal dis ibu ion and unequal a iances and he e o e an inde- penden wo- ailed es was pe o med. Fo dopple ana- lysis, an independen wo- ailed es was also pe o med a days 7, 14 and 21. The alues assume no mal dis ibu- ion. Only a day 7, we canno assume equal a iances. Fo capilla y and colla e al analysis, he alues also assume no mal dis ibu ion and equal a iances and he e o e an independen wo- ailed es was also pe o med. Fo he mas e junc ions and segmen leng h, one-way independ- en analysis o a iance (ANOVA) was de eloped as da a ollowed a no mal dis ibu ion, hough equali y o a iances could no be assumed o he i s . The e o e, a Games-Howell co ec ed pos hoc es was used o iden- i y di e ences be ween g oups. Fo he meshes a ea, as no mali y could no be assumed and so a K uskal-Wallis non-pa ame ic es was de eloped ollowed by i s pai ed analysis o iden i y di e ences be ween g oups. The e ec size and powe was de e mined by using he G-Powe so wa e in all ype o analyses. p< 0.05 was in e p e ed o deno e s a is ical signi icance. Pe ei a e al. S em Cell Resea ch & The apy (2016) 7:145 Page 5 o 12 Resul s UCX® p omo e ubulogenesis and endo helial cell mig a ion We i s e alua ed he e ec o UCX® on he capilla y s uc u e o ma ion. A ma igel ubule o ma ion assay was used as an in i o model. HUVECs we e seeded on o ma igel in endo helial basal medium (EBM) and co-cul u ed wi h UCX®. As a con ol HUVECs we e seeded in EBM, whe e we do no expec o isualize capilla y-like s uc u es o med by HUVECS. Fo ha eason, wo posi i e con ols we e added and HUVECs we e seeded bo h in endo helial g ow h medium (EGM) ha con ains se e al p o-angiogenic ac o s such as VEGF and FGF-2 and in EBM supplemen ed wi h FGF-2 (EBM-FGF-2). As shown in Fig. 1a and quan i ied in Fig. 1b, UCX® induce ubule o ma ion by HUVECs and hese s uc u es consis en ly showed a signi ican in- c ease in he numbe o mas e junc ions, o al mesh a ea and o al segmen leng h when compa ed o hose om HUVECs cul u ed in EBM o EGM. Wi h he ex- cep ion o he o al mesh a ea, all he measu ed pa ame- e s we e also signi ican ly di e en be ween HUVECs cul u ed wi h UCX® in EBM and HUVECs cul u ed in EBM supplemen ed wi h FGF-2. Nex , we assessed he mig a o y capaci y o HUVECs seeded in EBM supple- men ed wi h 5 % o se um co-cul u ed wi h UCX® wi h- ou cell o cell con ac . Wi h his objec i e, a wound- healing assay was pe o med and as a con ol HUVECs we e seeded in he same cul u e condi ions bu in he absence o UCX®. As shown in Fig. 2a and b, he wound a ea dec eased mo e apidly in he p esence o UCX® han in he con ol condi ion, sugges ing ha UCX® p omo e he mig a ion o HUVECs ia a pa ac ine sec e ion. Taken oge he , hese esul s clea ly show ha , in i o, UCX® induce angiogenic p ocesses in endo he- lial cells. A B C Fig. 1 UCX® p omo e ubulogenesis. Ma igel assay was pe o med by seeding human umbilical ein endo helial cells (HUVECs) in endo helial basal medium (EBM), endo helial g ow h medium (EGM), EBM supplemen ed wi h ib oblas g ow h ac o 2 (FGF-2) (EBM + FGF-2) o EBM co-cul u ed wi h UCX®. (a) Rep esen a i e images om he di e en expe imen al condi ions a e shown. (b) Quan i a i e e alua ion demons a ed signi ican ly enhanced mas e junc ions, o al mesh a ea and o al segmen leng h in capilla y-like s uc u es o med in HUVECs co-cul u ed wi h UCX® when compa ed o he o he expe imen al condi ions. Fo mas e junc ions, ollowing Games-Howell pos hoc es he mean di e ences be ween UCX® and he o he expe imen al condi ions we e: EBM (mean di . =24.27; p= 0.01); EGM (mean di . = 19.67; p= 0.03) and EBM + FGF-2 (mean di . = 13.8; p= 0.034). The e ec size o basal was 2.08 and powe 0.98. Fo he meshes a ea, ollowing pai wise compa isons, he equi alen o a pos hoc K uskal-Wallis es , signi ican di e ences be ween UCX® and EBM o UCX® and EGM we e obse ed wi h KW = 22.66, p< 0.01 o KW = 18.77, p=0.01, espec i ely. Fo segmen leng h, ollowing Games-Howell pos hoc es , UCX® show signi ican di e ence when compa ed o he o he g oups: EBM (mean di . = 4920; p< 0.001); EGM (mean di . = 3346; p= 0.01) and EBM + FGF-2 (mean di . = 2223; p= 0.016). E ec size o basal was 3.85 and powe 1. Scale ba , 500 μm Pe ei a e al. S em Cell Resea ch & The apy (2016) 7:145 Page 6 o 12 UCX® sec e e p o-angiogenic ac o s Since i was al eady demons a ed ha he bene icial e ec s p omo ed by UCX® esul om a pa ac ine sec e- ion, we decided o e alua e he sec e ion o six ep e- sen a i e ophic ac o s c i ical o angiogenesis in he condi ioned medium o an UCX® monolaye seeded in basal medium (BM) o 48 hou s. As a con ol, he same ophic ac o s we e measu ed in BM ha had no been in con ac wi h cells. The exp ession o HGF, VEGF-A, TGF-β1, in e leukin 8 (IL-8), pla ele -de i ed g ow h ac o -AA (PDGF-AA) and FGF-2 we e analysed. Ou esul s show a signi ican inc ease o HGF, TGF-β1, IL-8 and PDGF-AA in he condi ioned medium o UCX® when compa ed o he con ol (Table 1), demons a ing ha UCX® sec e e p o-angiogenic ac o s o he ex acel- lula medium. UCX® inc ease pe usion eco e y a e HLI In o de o e alua e he he apeu ic po en ial o UCX® in a HLI mouse model, he pe usion eco e y was assessed o e ime a e ischemia induc ion and UCX® adminis- a ion. Fi e hou s a e su gical induc ion o unila e al HLI, UCX® o hei ehicle (as a con ol) we e adminis- e ed in he ischemic muscle and pe usion was mea- su ed o e ime. As shown in Fig. 3a and quan i ied in Fig. 3b, a d ama ic educ ion in blood low was obse ed in he ischemic limb immedia ely a e su ge y, in com- pa ison o he con ala e al limb, ollowed by a g adual no maliza ion o ischemic blood low o e ime. Impo - an ly, UCX® ma kedly imp o ed blood low eco e y a 7, 14 o 21 days pos -HLI, compa ing wi h con ol mice, demons a ing ha UCX® adminis a ion con ibu es o- wa d he unc ional eco e y o ischemic issues. A B Fig. 2 UCX® s imula e endo helial cell mig a ion. Human umbilical ein endo helial cells (HUVECs) we e seeded o con luence in endo helial basal medium (EBM) supplemen ed wi h 5 % FBS and cul u ed o 16 hou s a e wounding. Immedia ely a e wound, HUVECs we e co-cul u ed wi h UCX® o no (con ol) o 9 hou s. (a) Rep esen a i e images a e shown. The le and igh panels conce n HUVEC monolaye immedia ely a e o 9 hou s pos -wounding, espec i ely. (b) The wound a ea was measu ed immedia ely a e and 9 hou s pos -wounding by using ImageJ so wa e ( h ee independen measu emen s we e pe o med o each expe imen al g oup; (2.03) = 4.66; *p< 0.05); e ec size was 3.82 and powe 0.93) Pe ei a e al. S em Cell Resea ch & The apy (2016) 7:145 Page 7 o 12 UCX® inc ease capilla y and colla e al essel densi y a e HLI As blood low eco e y depends on bo h angiogenesis and a e iogenesis we examined whe he UCX® would a ec capilla y and colla e al essel densi ies in hindlimb muscles. Fi e hou s a e su gical induc ion o unila e al HLI, UCX® o hei ehicle (as a con ol) we e adminis- e ed in he ischemic muscle and mice we e sac i iced 90 days pos -HLI induc ion. Capilla y densi y was assessed h ough quan i ica ion o CD31-posi i e capil- la ies on his ological sec ions o gas ocnemius muscle. As expec ed, he capilla y densi y was g ea e in he ischemic e sus he non-ischemic hindlimb in bo h ex- pe imen al g oups (Fig. 4b). No ably, he le el o capil- la y densi y in he ischemic hindlimb ea ed wi h UCX® was signi ican ly highe han he one obse ed in he con ol ischemic hindlimb, as shown in Fig. 4a and quan i ied in Fig. 4b. In o de o e alua e he colla e al essel densi y (CVD), mice we e diaphonised and an equi alen egion o in e es (ROI), co esponding o 20 % o he limb a ea, was selec ed o CVD quan i ica- ion (Fig. 4c). A signi ican ly highe CVD inc ease was obse ed in UCX®- ea ed mice when compa ed o he con ol (Fig. 4d). Taken oge he , ou esul s show ha A B Fig. 3 UCX® inc ease pe usion eco e y. UCX® o hei ehicle (as a con ol) we e adminis e ed in he ischemic gas ocnemius muscle 5 hou s a e HLI induc ion. (a) Rep esen a i e lase dopple low images be o e (PRE-HLI), immedia ely a e (d0 POST-HLI) and a 7, 14 and 21 days pos -HLI induc ion (d7 POST-HLI, d14 POST-HLI, d21 POST-HLI). (b) Quan i a i e e alua ion o blood low exp essed as a a io o ISC o NISC limb demons a ed signi ican ly enhanced limb blood pe usion in UCX®- ea ed mice a 7, 14 and 21 days pos -HLI. (n = 16 o each expe imen al g oup; D7: (22.69) = 4.26; ***p< 0.001; e ec size was 1.51 and powe 0.98; D14: (30) = 4.7; ***p< 0.001; e ec size was 1.66 and powe 0.99; D21: (30) = 7.22; ***p< 0.001; e ec size was 2.56 and powe 0.99). HLI hindlimb ischemia, ISC ischemic, NISC non-ischemic Table 1 P o-angiogenic ac o s p esen in he condi ioned medium om UCX® o basal medium UCX® CM (pg/ml) Mean±SEM Basal CM (pg/ml) Mean±SEM - es HGF 2,64 ± 0,718 0,001 ± 0 0,0144 VEGF-A 2,025 ± 1,043 0,001 ± 0 0,1099 TGF-β1 0,983 ± 0,122 0,001 ± 0 0,0005 IL-8 0,765 ± 0,069 0,001 ± 0 0,0001 PDGF-AA 0,123 ± 0,045 0,001 ± 0 0,0412 FGF-2 0,0317 ± 0,009 0,02 ± 0 0,2874 CM condi ioned medium Pe ei a e al. S em Cell Resea ch & The apy (2016) 7:145 Page 8 o 12 UCX® signi ican ly augmen capilla y densi y and CVD inc ease a e ischemic inju y. UCX® up egula e he endo helial Tg -β2, Ang-2, Fg -2 and Hg exp ession in esponse o HLI To elucida e he mechanism o ac ion o UCX® in ou HLI model we in es iga ed i UCX® could change he endo helial gene exp ession. Fo his pu pose, 5 hou s a e su gical induc ion o unila e al HLI, UCX® o hei ehicle (as a con ol) we e adminis e ed in he ischemic muscle and mice we e sac i iced a day 70 pos -HLI in- duc ion. The gas ocnemius muscle sec ions we e s ained o CD31 and isualized using a Lase Cap u e Mic odissec ion mic oscope. CD31-posi i e cells we e dissec ed, isola ed and analysed by quan i a i e RT-PCR o he exp ession o se e al p o-angiogenic ac o s.Ou esul s show ha ansc ip s o Tg -β2, Ang-2, Fg -2 and Hg we e clea ly up egula ed in endo helial cells isola ed om muscle o he ischemic limb, compa ing wi h hose ound in endo helial cells om he con ala e al limb. This is obse ed exclusi ely in mice ea ed wi h UCX®. Con ol mice show he opposi e end, down egula ing he exp ession o he angiogenic genes in endo helium om he ischemic limb, when compa ed wi h he con ala e al limb (Fig. 5). These esul s sugges ha he mechanism o ac ion o UCX® in he HLI mouse model in ol es he up egula ion o se e al angiogenic genes in he endo helial cells p esen in he ischemic muscle. Discussion CLI is a mani es a ion o pe iphe al a e ial disease ha desc ibes pa ien s wi h ch onic ischemic es pain, ulce s o gang ene [1]. A ailable he apies a e limi ed and pa- ien s may equi e ampu a ion. The goal o limb sal age has s imula ed esea ch in o al e na i e me hods, includ- ing he apeu ic angiogenesis. S em cell he apy holds g ea po en ial o he apeu ic angiogenesis, bu i s clin- ical ansla ion has been slow due o (i) leng hy p oce- du es pe o med ex i o in which s em cells a e manipula ed and o en lose iabili y, di e en ia e o AB CD Fig. 4 UCX® inc ease capilla y and colla e al densi ies. UCX® o hei ehicle (as a con ol) we e adminis e ed in he ischemic gas ocnemius muscle 5 hou s a e HLI induc ion. (a) Rep esen a i e sec ions om con ol and UCX®- ea ed ischemic gas ocnemius muscles a 90 days pos -HLI. Capilla ies and myocy es we e iden i ied by CD31 immunohis ochemis y and haema oxylin, espec i ely. Scale ba , 125 μm. (b) Quan i a i e analysis, a 90 days pos -HLI, e ealed inc eased capilla y densi y (capilla ies/myocy e) in UCX®- ea ed ischemic gas ocnemius muscles compa ed o con ol ischemic ones. (n = 3 and n = 7 o con ol and UCX®, espec i ely; (8) = 4.44; **p= 0.0057; e ec size was 3.064 and powe 0.97). (c) Illus a i e images o se- lec ed egions o in e es (ROI) o con ol and UCX®- ea ed mice. ISC and NISC limbs a e shown. Scale ba , 1 mm (d) Da a a e ep esen ed as he pe - cen age o colla e al essel densi y (CVD) inc ease o he ISC limb ela i ely o he NISC one. A 90 days pos -HLI, UCX®- ea ed mice p esen ed signi ican ly highe CVD inc ease (%) when compa ed o con ol mice (n = 5 o each expe imen al g oup; (8) = 7.63; ***P=0.000062; e ec size was 4.82 and powe 0.99). ISC ischemic, NISC non-ischemic Pe ei a e al. S em Cell Resea ch & The apy (2016) 7:145 Page 9 o 12