Mitochondrial bioenergetics and distribution in living human osteoblasts grown on implant surfaces

Summa y. Osseoin eg a ion o implan s is c ucial o he long- e m success o o al implan s. The pe iimplan bone o ma ion by os eoblas s is s ongly dependen on he local mechanical en i onmen in he in e ace zone. Robus demands o ene gy a e placed on os eoblas s du ing he adhesion p ocess o solid su aces, and mi ochond ia a e capi al o ganelles in he p oduc ion o mos o he ATP needed o he p ocess. We ha e assessed he ela ionship be ween os eoblas di e en ia ion and mi ochond ial bioene ge ics in li ing cells g own on wo di e en i anium su aces, in o de o p o ide aluable in o ma ion o he design o ma e ial su aces equi ed o he de elopmen o he mos app op ia e os eogenic su ace o os eoblas ic ancho age. Combined backsca e ed and luo escence con ocal mic oscopy showed ha in la cells g own on a machined su ace, highly ene gized mi ochond ia we e dis ibu ed along he cell body. In con as , cells g own on he ough su ace emi ed long p o usions in sea ch o su ace oughness, wi h ac in s ess ibe s clea ly pola ized and highly ene gized mi ochond ia clus e ed a ocal adhesion si es. This epo using no mal human os eoblas ic cells indica es ha hese cells a e especially sensi i e o su ace cues h ough ene gy p oduc ion ha enhances he necessa y adhesion equi ed o a success ul osseoin eg a ion. Key wo ds: Ti anium, Su ace opog aphy, Os eoblas , Cell adhesion, Mi ochond ial memb ane po en ial In oduc ion Recen esea ch has sugges ed ha bones display an ex ao dina y adap a i e beha iou owa ds changing mechanical en i onmen , which is o en ega ded as pheno ype plas ici y. Fu he mo e, he pe iimplan issue o ma ion and mine aliza ion by os eoblas s a e s ongly dependen on he local mechanical en i onmen in he in e ace zone. The quali y o cell adhesion o he implan su ace de e mines issue in eg a ion, and he su ace oughness can di ec ly in luence os eoblas adhe ence, a achmen , sp eading and me abolism modi ying and con olling he osseoin eg a ion p ocess (Pie es e al., 2002; Ingbe , 2003a,b; Sade e al., 2005; Chiesa e al., 2007; Ga i e al., 2008). In mig a ing cells, ac in ilamen s a e p ima ily o ganized in o submemb ane meshwo ks ha appea di use o in small bundles, whe eas s a iona y cells usually display conspicuous a ays o ilamen bundles, “s ess ibe s”, which a e ancho ed a hei e mini in ma ix a achmen s known as ocal adhesions (Chen e al., 2003; Zime man e al., 2004; Diene e al., 2005; Hu e al., 2007). Robus demands o ene gy a e placed on os eoblas s in he adhesion p ocess o solid su aces, and mi ochond ia a e esponsible o he p oduc ion o he la ges pa o cellula ATP needed o endoe gonic p ocesses wi hin he cell (G igo iou e al., 2005; Anes i and Sco ano, 2006; Boldogh and Pon, 2007). To unde s and how mic o opog aphy modula es cell adhesion, in he p esen s udy we ha e de eloped an in i o model, using he human os eoblas ic cell line NHOs ®, o s udy mi ochond ial dis ibu ion and bioene ge ics du ing he ini ial phases o he adhesion p ocess o di e en i anium implan su aces, p epa ed as disks o in i o manipula ion . The e is a di ec co ela ion be ween he ene gized s a e o mi ochond ia and he mi ochond ial memb ane (∆ψµ) when analyzed in isola ed mi ochond ia (Sal ioli Mi ochond ial bioene ge ics and dis ibu ion in li ing human os eoblas s g own on implan su aces Me cedes Salido1, J. Ignacio Vilches-Pe ez2, Juan L. Gonzalez3and Jose Vilches1 1Depa men o His ology, School o Medicine, Labo a o y 57, Se icios Cen ales de In es igación en Ciencias de la Salud, Uni e si y o Cadiz, Cadiz, Spain, 2Depa men o O al su ge y, School o Den is y. Uni e si y o Se ille, Se ille, Spain and 3Depa men o S a is ics and Ope a ional Resea ch, School o Medicine, Uni e si y o Cadiz, Cadiz, Spain His ol His opa hol (2009) 24: 1275-1286 O p in eques s o: Me cedes Salido, Depa men o His ology, School o Medicine, Labo a o y 57, Se icios Cen ales de In es igación en Ciencias de la Salud, Uni e si y o Cadiz, C/ D . Ma añon 3, 11002 Cádiz. Spain. e-mail: [email p o ec ed] h p://www.hh.um.es His ology and His opa hology Cellula and Molecula Biology 1276 e al., 1997; Dedo and Rou ogalis, 1999, Salido e al., 2007a). In o de o p o ide aluable in o ma ion o he design o ma e ial su aces ha a e equi ed o he de elopmen o he mos app op ia e os eogenic su ace o os eoblas ic ancho age in den al implan s, he ela ionship be ween os eoblas di e en ia ion on di e en i anium su aces and cellula bioene ge ics has been assessed in li ing cells in his s udy, by means o combined backsca e ed and luo escence analysis wi h con ocal mic oscopy. Thus, he dynamics o a physiological p ocess, as is he case o he ini ial phases o osseoin eg a ion, can be de e mined. We ha e, hus, used his in o ma ion o econs uc he h ee- dimensional in e ace be ween he os eoblas ic cells and he implan su ace. The esul s p esen ed he ein en o ce he ole o he ough subs a um su ace in a ec ing os eoblas ic cell adhesion, equi ed o he de elopmen o an app op ia e os eogenic su ace o os eoblas ic ancho age and ma u a ion, compa ed o a machined su ace in den al implan s. Ma e ial and me hods To e alua e he ela ionship be ween os eoblas di e en ia ion and bioene ge ics, NHOs cells we e g own on di e en i anium su aces, Ti CP machined and Ti CP Osseo i e®, p o ided as p e ab ica ed 2 cm x 1.5 mm disks, in medium supplemen ed wi h asco ba e o induce di e en ia ion. To e alua e changes in mi ochond ial memb ane po en ial (∆ψµ)) and in subcellula dis ibu ion in esponse o di e en sca olds, cells we e s ained wi h a ansmemb ane po en ial- sensi i e i al dye, JC1, which p e e en ly accumula es in mi ochond ia by ∆ψµ dependen mechanism, and con ocal imaged. Ac in cy oskele al o ganiza ion was assessed by immunos aining wi h hodamine phalloidin in o de o s udy he ole o he subs a um su ace in a ec ing os eoblas ic cell adhesion. Cell cul u e No mal human os eoblas ic NHOs ®cells (Camb ex, Walke s ille, MD, USA) we e seeded a a densi y o 5000 cells/cm2and incuba ed in Os eoblas G owing Medium, OGM, (Camb ex, Walke s ille, MD, USA) con aining 10% e al bo ine se um (Camb ex, Walke s ille, MD, USA), 1% gen amycin sulpha e/ ampho e icin B (Camb ex, Walke s ille, MD, USA) and 1% asco bic acid (Camb ex, Walke s ille, MD, USA), as ecommended by supplie s, a 37°C and 5% CO2un il he expe imen s s a ed. G ow h medium was changed daily a e seeding. Be o e he cells became 80% con luen hey we e subcul u ed wi h 2 ml o 0.25 mg/ml ypsin EDTA wa med o 37°C (Camb ex, Walke s ille, MD, USA) a e insing wi h 5 ml Hepes-BSS (Camb ex, Walke s ille, MD, USA) a oom empe a u e. Once cells we e de ached, ypsin EDTA was neu alized by adding 4 ml o ypsin neu alizing solu ion (Camb ex, Walke s ille, MD, USA). Ha es ed cells we e seeded on he di e en disks a a densi y o 5000 cells/cm2and immunos ained a e 48 h. G ow h medium was changed e e y day un il he expe imen s we e o e . NHOs cells a e assu ed o expe imen al use o en popula ion doublings, which we e no exceeded du ing he assay. Ti anium disks Be o e use, he disks we e imme sed in 100% e hanol o 10 min, ai d ied, and exposed unde u. . ligh o 30 min on each side, and inally insed in endo oxin ee phospha e bu e ed solu ion, and deposi ed on small s e ile Pe i dishes p io o cell seeding. Tissue cul u e Willco®(WillCo Wells, Ams e dam, The Ne he lands) wells, wi h a 0.17 mm glass bo om, we e used as he con ol su ace. Immunohis ochemis y and cy oskele al o ganiza ion A he end o he speci ic cul u e ime, cells we e washed wice wi h p ewa med phospha e-bu e ed saline, (PBS), pH 7.4, ixed wi h 3.7% pa a- o maldehyde (PFA) solu ion in PBS o 10 min a oom empe a u e, and washed wice wi h p ewa med PBS. The cells we e hen pe meabilized wi h 0.1% T i on x- 100 (Sigma, S Louis, Missou i, USA) o 5 min and washed wice wi h p ewa med PBS. To educe non- speci ic backg ound s aining, 1% bo ine se um albumine (BSA) in PBS was added o he su aces o 20 min, and cells we e immunos ained o 20 min. wi h hodamine phalloidin, 12.5 µl o me hanolic s ock solu ion (Sigma, S Louis, Missou i, USA) in 500 µl PBS o each sample. A e disca ding he s aining solu ion, disks we e insed wi h p ewa med PBS h ee imes p io o moun ing wi h ec ashield®(Vec o Labs. Bu lingame CA, USA) and 0.17 mm co e slip in one holed polyca bona e slide de ices specially designed in ou labo a o y and ab ica ed by means o con ol p ecision sys ems (Mecap ec, Cadiz, Spain). Mi ochond ial pe meabili y po en ial Cells we e s ained wi h he ca ionic dye JC-1 (Mi o PT TM®, Immunohis ochemis y Technologies, Blooming on, MN, USA), which exhibi s po en ial- dependen accumula ion in mi ochond ia. A low memb ane po en ials, JC-1 con inues o exis as a monome and p oduces a g een luo escence (emission a 527 nm). A high memb ane po en ials, JC-1 o ms “J- agg ega es" (emission a 590 nm) and p oduces a ed luo escence. Cells we e cul u ed in Willco®wells wi h glass bo om as con ol g oup o he assay, and on he di e en disk su aces p o ided, no exceeding a inal numbe o 106cells/ml, and a e disca ding he cul u e medium, 1x Mi oPT®s aining solu ion ob ained om a 100x s ock was added o he wells, 0.5 ml pe well. Cells we e hen incuba ed a 37°C o 15 min in a CO2 incuba o and, a e disca ding he medium, washed Mi ochond ial esponse o su ace induced cues 1277 wice wi h 1-2 ml o assay bu e wa med o 37°C. A e disca ding he wash, a d op o assay bu e was added o he specimens p io o immedia e examina ion in he in e ed con ocal mic oscope Leica TCS SL (Leica, Da ms aad, Ge many), equipped wi h an HCX PL APO CS 63.0x1.30 glyce ol imme sion objec i e, wi h an incuba ion sys em consis ing o a cube ha comple ely co e s he mic oscope and allows us o keep cells a 37°C in a con olled a mosphe e wi h a mixed ai /CO2 low o 4 l/h and 5% CO2du ing image collec ion and analysis. Con ocal examina ion. Image collec ion and analysis o cy oskele on The cells and disks we e simul aneously isualized using a Leica TCS-SL con ocal mic oscope equipped wi h a 63.0x1.30 glyce ol objec i e, allowing simul aneous acquisi ion o hodamine phalloidin s aining o ac in cy oskele on (exci a ion 554 nm / emission 573 nm) and su ace e lec ance. A leas ou isola ed disks we e analyzed in each g oup o p o ilome ic in o ma ion, and a minimum o ou disks wi h cells g owing on hem we e analyzed in each g oup, in o de o assess su ace in luence on cy oskele al o ganiza ion and cell mo phology. A leas 50 cells pe disk we e analysed. Images we e collec ed and p ocessed o quan i a i e analysis using he imaging so wa e p o ided by he Leica TCS SL sys em. P o ilome ic s udies, o he quan i ica ion o maximal and minimal alues in su ace p o ile, dis ance be ween peaks in each su ace and wid h o g oo es and alleys in he di e en su aces we e assessed. All samples we e exposed o lase o a ime in e al no mo e han 5 min o a oid pho obleaching. The exci a ion beam spli e selec ed was a DD 488/543. The lase was se o he lowes powe able o p oduce a luo escen signal. Maximum ol age o pho omul iplie s was used o dec ease he equi ed lase powe as much as possible. O se was main ained a 0. A pinhole o 1 Ai y uni was used. Images we e acqui ed a a esolu ion o 1024x1024. Se ies we e acqui ed in he xyz mode. Image collec ion and analysis o ∆ψµ The cells and disks we e simul aneously isualized using a Leica TCS-SL con ocal mic oscope equipped wi h a 63.0x1.30 glyce ol objec i e, unde he condi ions desc ibed abo e, allowing simul aneous acquisi ion o su ace e lec ance and JC1 s ained mi ochond ia. JC-1 was exci ed a 490 nm and he emission luo escence was collec ed in TRITC (590 nm) and FITC (530 nm) channels simul aneously. Fo quan i a i e analysis, a leas 120 egions o in e es (ROIs) we e selec ed in each g oup o quan i y changes in ∆ψµ. All o he ROIs a e cells selec ed unde he ollowing c i e ia: well-de ined limi s, clea iden i ica ion o nucleus and absence o in e sec ion wi h neighbou ing cells. Size, numbe o pixels and luo escence in ensi y in he ed (high memb ane po en ial) and g een (low memb ane po en ial) channels in each ROI we e calcula ed. S a is ical analysis The s a is ical analysis was pe o med wi h SPSS p og am. A one-way ANOVA analysis was used o compa e he mean alues o cell a eas and ed pixel in ensi y, s anda d de ia ion and skewness. The no mali y o he g oups was con as ed wi h Kolmogo o -Smi no es and he a iances homogenei y wi h Coch an’s C es . Pos -hoc con as s, HSD es o Tukey, we e ca ied ou o de ec he di e ences be ween g oups. A alue o p≤0.05 was conside ed signi ican . Resul s CLSM cha ac e iza ion o su aces Con ocal backsca e ed imaging o disk su aces, ob ained by e lec ion o he 488 lase line, showed a ma kedly i egula su ace in he TiCP Osseo i e®wi h p ominen peaks and ca es, as shown (Fig. 1A). Some o hem p esen ed well-de ined con ou s, while some o he s o med a wide abecula su ace wi h nume ous c a e - like s uc u es and andomly dis ibu ed ca i ies and bo de s. The TiCP machined su ace displayed pa allel g oo es, caused by he wa iness inhe en o he machining ope a ion, which de e mined a ela i ely smoo h su ace (Fig. 1C). Th ee dimensional econs uc ions, ob ained om se ial acquisi ion o s acks, we e pe o med o he assessmen o os eoblas - su ace in e ace (Fig. 1B,D). The p o ilome ic measu emen esul ed in a minimum alue o 1.2 µm o he dep h o he alleys and a maximum alue o 4.15 µm o he op o he embossed s uc u e. The mean dis ance be ween ops eached 18.28 µm, he mean wid h o he alleys was 2.85 µm, he mean alue ound o he g oo e on op o each embossed ing was 1.51 µm. Ti CP Osseo i e®disks showed a ma ked inc ease in he su ace oughness, wi h p ominen peaks and ca es, some o hem wi h well-de ined con ou s and some o he s o ming a wide abecula su ace wi h andomly dis ibu ed peaks. The dis ance be ween peaks, consequen ly, was highly a iable, wi h alues anging om 24.45 µm up o 183.67 µm. The mean diame e o he ca es was ound o be la gely a iable, anging om 8.01 and 39.27 µm. As e ealed in he p o ilome ic s udy, his pa e n no ably inc eased he mean oughness o he su ace, ha showed a minimum alue o 1.9 µm in dep h and a maximum peak alue o 8.79 µm. CLSM obse a ion o di e ences in os eoblas pheno ypes on di e en su aces Cells g owing on he TiCP Osseo i e®su ace, whe e he dis ance be ween peaks inc eases, p esen ed an inc easingly pola ised sp eading wi h emission o ei he wide lamellipodia o long ilopodia, depending on Mi ochond ial esponse o su ace induced cues he su ace oughness and he dis ance be ween subjacen peaks, and showed he longes mo phology wi h a mean axial a io ha ose o 3.5 (Fig. 2). The ac in cy oskele on adop ed an o ganiza ion cha ac e ized by ac in bundles unning ei he pe pendicula ly o sligh ly obliquely, wi h appa en ly segmen ed con ac ile uni s de ined opog aphically by ancho age poin s, o ganized acco ding o he su ace s uc u e. Pa allel ac in ilamen s we e obse ed along he cellula p ocesses and ancho ing o ocal con ac s o ela i e opog aphical peaks was obse ed. As shown, cells clea ly i ed o su ace opog aphy and e en emi ed long ilopodial b anches, when necessa y, o success ul ancho age media ed by ac in “s ess ibe s” and ocal adhesion con ac s on selec ed peaks. When os eoblas s g owing on TiCP machined su aces we e examined, signi ican di e ences we e obse ed, no only ega ding size and shape, bu also o cy oskele al o ganiza ion. Cells g own on Ti CP machined su aces la ened and adop ed a la gely adial mo phology, and appea ed o be p edominan ly plana and andomly o ien ed wi h a mean axial a io o 1.1 (Fig. 3). Cells ancho ed o he su ace by displaying long dend i ic ilopodia, bo h along he g oo es and c ossing o e hem in a pe pendicula way. Con ol cells, g owing on he glass su ace, did no show a pa e ned mo phology. Ins ead, a numbe o cells showed long, hin ilopodia, while some o he s la ened and sp ead o e he glass su ace, hus p esen ing a high a iabili y in ac in cy oskele on o ganiza ion, wi h ac in ibe s ha appea ed unning in all di ec ions (Fig. 6D). Mi ochond ial bioene ge ics and dis ibu ion On TiCP Osseo i e®su aces, NHOs ®cells p esen ed a high densi y o ilamen ous mi ochond ia. Mos o he highly ene gized mi ochond ia, s ained in ed (Fig. 4), accumula ed in he long ilopodial b anches. The simul aneous isualiza ion o mi ochond ial labelling and he backsca e ed imaging o he su ace allowed us o co ela e he mi ochond ial dis ibu ion wi h he i egula subs a um mic o opog aphy, ull o peaks and ca es, hus con i ming mi ochond ial pola iza ion o he selec ed ancho age poin s. In cells g owing on TiCP machined su aces, highly ene gized mi ochond ia illed he cell body (Fig. 5). No 1278 Mi ochond ial esponse o su ace induced cues Fig. 1. A. Con ocal backsca e ed imaging o TiCP ough su ace (Osseo i e®), showing he i egula dis ibu ion o peaks and ca es ha con igu e an i egula su ace o bone cell a achmen . B. P o ilome ic 3D econs uc ion. In con as , as shown in C, he machined su ace p esen ed pa allel g oo es ha con igu e an almos uni o m su ace, qui e egula in dep h, as shown in he p o ilome ic 3D econs uc ion in D. In o de o cla i y in e p e a ion o images, a glow LUT pale e is p esen ed, i.e. he da ke zones a e he deepe ones. clus e ing o signi ican pola iza ion a ound con ac si es we e obse ed, and SD alues (Table 1) con i med qui e a simila mi ochond ial dis ibu ion o ha displayed in he con ol g oup, whe e os eoblas s, which adhe e o glass su ace, showed sca e ed mi ochond ia une enly dis ibu ed. The e is also a ilamen ous ne wo k o 1279 Mi ochond ial esponse o su ace induced cues Fig. 2. A. Simul aneous imaging o disk su ace and hodamine-phalloidin s ained NHOs ®cell g own on he ough su ace o 48 h. Cells i ed o he su ace wi h he emission o ilopodia and lamellipodia. As shown in B, when glow imaging was employed, he dis ibu ion and concen a ion o ac in s ess ibe s is condi ioned by su ace i egula i ies. C. Me ged image ob ained om a se ial acquisi ion o 50 s acks in he xyz mode, combining backsca e ed and luo escence imaging. Long ilopodia emi ed om he cell body a e looking o app op ia e adhesion si es. D. De ailed image o a so called “cell oo ” in a ocal adhesion si e, wi h ac in s ess ibe s clea ly pola ized o he disk su ace. In his case, an in e ed glow pale e is p esen ed. pola ized mi ochond ia su ounding he nucleus (Fig. 6). S a is ical analysis When analyzing he a iable a ea o he selec ed 1280 Mi ochond ial esponse o su ace induced cues Fig. 3. A Simul aneous imaging o disk su ace and hodamine-phalloidin s ained NHOs ®cell g own on he machined su ace o 48 h. Cells a e p edominan ly la and s ess ibe s un in di e se di ec ions, as clea ly shown in Bwhen glow imaging was employed. In his case, cells a e g owing c ossing o e he g oo es in a pe pendicula way, wi h a disc e e emission o ilopodia. C. Some o he cells g owing in pa allel o machined g oo es wi h s ess ibe s co e ing he whole cell body, as highligh ed in D. ROIs, a highly signi ican di e ence (p<2•E-10) appea ed be ween cells g own on he di e en su aces es ed, and cells g owing on he ough, TiCP Osseo i e® su ace, showed he highes alues, ollowed by cells g owing on glass su ace and inally, he smalles ones, cells g owing on he machined su ace. Despi e his ac , he di e ences be ween ed pixel means be ween g oups we e no signi ican (p=0.369). On he con a y, highly signi ican di e ences we e ound when compa ing s anda d de ia ions (p<2•E-14), and skewness (p=0.011) (Table 1). 1281 Mi ochond ial esponse o su ace induced cues Fig. 4. A. Li ing cell imaging o JC1 s ained os eoblas s simul aneously acqui ed wi h he backsca e ed imaging o Osseo i e®disk su ace 48 h a e seeding. Cells emi ed long p o usions in sea ch o su ace p o ube ancies. In ed, highly ene gized mi ochond ia ha accumula e in ocal adhesion si es. B. Shown in de ail, long ilopodia a e b anching o con ac wi h su ace peaks, which appea in clea g ey. Red s ained mi ochond ia clus e in he ene gy equi emen zones. C. The cell shows he ypical mo phology obse ed in he Osseo i e®g oup, wi h long ilopodial emission ull o highly ene gized mi ochond ia, i ing o disk su ace, in g een. D. Mi ochond ial clus e ing, ed, in ocal adhesion si es. Cells show an elonga ed mo phology de e mined by ela i e su ace peaks. The esul s o pos -hoc compa isons be ween g oups, analyzing he di e ences in model-p edic ed means, a e shown in Table 2. B ie ly, highly signi ican di e ences we e ound in cell a ea when compa ing he 1282 Mi ochond ial esponse o su ace induced cues Fig. 5. A. Li ing cell imaging o JC1 s ained os eoblas s simul aneously acqui ed wi h he backsca e ed imaging o he machined disk su ace 48 h. a e seeding. Cells a e p edominan ly la , and highly ene gized mi ochond ia, in ed, a e dis ibu ed along he cell body. B. Os eoblas s g own on he machined su ace, wi h a ela i ely high cellula densi y ha o ien ed along he g oo es, some o he cells emi ing small ilopodia wi h low p esence o ed s ained mi ochond ia. C. Polygonal os eoblas s g owing on he machined su ace. The cellula densi y is high, bu no clea adhesion si es a e obse ed. Highly ene gized mi ochond ia a e sca e ed along he cell body. D. Shown in de ail, combined backsca e ed image o machined disks clea ly showing he g oo ed pa e n and JC1 s ained cells. Mi ochond ia a e dis ibu ed nea p olonga ions bu also in he cell body. Cells appea o be la and sp ead. ough, Osseo i e®, and machined g oups wi h he con ol g oup, bu no when compa ing Osseo i e®and machined g oups. Signi ican di e ences be ween all h ee g oups appea ed when compa ing s anda d de ia ions o ed pixel dis ibu ion, mainly when compa ing cells g owing on Osseo i e®disks wi h he emaining g oups. S a is ically signi ican di e ences be ween Osseo i e®and machined 1283 Mi ochond ial esponse o su ace induced cues Fig. 6. A. Con ol cells g owing on a glass su ace and JC1 s ained a e 48h. Cell mo phology e eals a non-pola ized sp eading in he sea ch o di e en ia ion cues. Highly ene gized mi ochond ia a e sca e ed and une enly dis ibu ed. In B, an elonga ed os eoblas g owing on glass is emi ing ilopodia wi h highly ene gized mi ochond ia. Due o he absence o su ace signaling, cell mo phology is di e se, as shown also in C, whe e he cellula body adop ed a nea ly polygonal shape and mi ochond ia sca e along he cell. D. Rhodamine-phalloidin s ained NHOs ®cell g owing on glass. Ac in ibe s a e i egula ly dis ibu ed along he cell body. g oups we e ound when compa ing skewness o ed pixel dis ibu ion, bu no when compa ing machined o Osseo i e®g oups wi h he con ol g oup.