The Importance of Controlled Mismatch of Biomechanical Compliances of Implantable Scaffolds and Native Tissue for Articular Cartilage Regeneration

ORIGINAL RESEARCH published: 30 No embe 2018 doi: 10.3389/ bioe.2018.00187 F on ie s in Bioenginee ing and Bio echnology | www. on ie sin.o g 1No embe 2018 | Volume 6 | A icle 187 Edi ed by: Pina Zo lu una, Uni e si y o No e Dame, Uni ed S a es Re iewed by: Pie gio gio Gen ile, Newcas le Uni e si y, Uni ed Kingdom Pina Yilgo Hu i, Anka a Uni e si y, Tu key Jus in Lee B own, Pennsyl ania S a e Uni e si y, Uni ed S a es *Co espondence: Michael Gasik [email p o ec ed] Special y sec ion: This a icle was submi ed o Bioma e ials, a sec ion o he jou nal F on ie s in Bioenginee ing and Bio echnology Recei ed: 16 June 2018 Accep ed: 16 No embe 2018 Published: 30 No embe 2018 Ci a ion: Gasik M, Zühlke A, Haapa an a A-M, Muhonen V, Laine K, Bilo sky Y, Kellomäki M and Ki i an a I (2018) The Impo ance o Con olled Misma ch o Biomechanical Compliances o Implan able Sca olds and Na i e Tissue o A icula Ca ilage Regene a ion. F on . Bioeng. Bio echnol. 6:187. doi: 10.3389/ bioe.2018.00187 The Impo ance o Con olled Misma ch o Biomechanical Compliances o Implan able Sca olds and Na i e Tissue o A icula Ca ilage Regene a ion Michael Gasik1,2*, Alexand a Zühlke1, Anne-Ma ie Haapa an a3, Vi pi Muhonen4, Kaisa Laine3, Ye gen Bilo sky2, Minna Kellomäki3,5 and Ilkka Ki i an a4 1School o Chemical Enginee ing, Aal o Uni e si y Founda ion, Espoo, Finland, 2Seq e a L d., Helsinki, Finland, 3BioMediTech and Facul y o Biomedical Sciences and Enginee ing, Tampe e Uni e si y o Technology, Tampe e, Finland, 4Depa men o O hopaedics and T auma ology, Uni e si y o Helsinki, and Helsinki Uni e si y Hospi al, Helsinki, Finland, 5BioMediTech and Facul y o Li e Sciences and Medicine, Uni e si y o Tampe e, Tampe e, Finland Sca olds o a icula ca ilage epai ha e o be op imally biodeg adable wi h simul aneous p omo ion o hyaline ca ilage o ma ion unde a he complex biomechanical and physiological condi ions. I has been gene ally accep ed ha sca old s uc u e and composi ion would be he bes when i mimics he s uc u e o na i e ca ilage. Howe e , a epa a i e cons uc mimicking he ma u e na i e issue in a healing issue si e p esen s a biological misma ch o epa a i e s imuli. In his wo k, we s udied a new ecombinan human ype III collagen-polylac ide ( hCol-PLA) sca olds. The hCol-PLA sca olds we e assessed o hei ela i e pe o mance in simula ed syno ial luids o 1 and 4 mg/mL sodium hyalu ona e wi h applica ion o model- ee analysis wi h Bioma e ials Enhanced Simula ion Tes (BEST). Pu e PLA sca old was used as a con ol. The BEST esul s we e compa ed o he esul s o a p io in i o s udy wi h hCol-PLA. Collec i ely he da a indica ed ha a success ul a icula ca ilage epai equi e lowe s i ness o he sca old compa ed o su ounding ca ilage ye ma ching he s ain compliance bo h in s a ic and dynamic condi ions. This ensu es an op imal combina ion o load ans e and e ec i e oscilla o y nu ien s supply o he cells. The esul s encou age u he de elopmen o in elligen sca old s uc u es o op imal a icula ca ilage epai a he han simply ying o imi a e he espec i e o iginal issue. Keywo ds: a icula ca ilage, sca old, PLA, collagen, biomechanics, es ing, syno ial luid INTRODUCTION The need o de elop issue subs i u es and egene a ion pla o ms is one o he mos demanding and challenging applica ions in mode n issue enginee ing (Hubbell, 1995; Bu dick and Mauck, 2011). Th ee-dimensional bioma e ial s uc u es (sca olds) a e highly desi able ma ching he biomechanical p ope ies o he issue (Gomes and Reis, 2004) and closely mimicking in i o beha io [ acili a ing cell adhesion, g ow h, and issue o ma ion (Vol son e al., 2008)]. Such bioma e ials assis he body o ebuild he damaged issue and e en ually hey minimize associa ed Gasik e al. No el Sca olds o A icula Ca ilage Repai pain and healing ime (Wong and B onzino, 2007; Chung and Bu dick, 2008). The combined s a ic and dynamic biomechanical p ope ies o hese sca olds a e c ucial o he inal success o he ea men . Any p og ess in he de elopmen o sca olds should ensu e a high co ela ion be ween in i o condi ions and expec ed in i o issue egene a ion (F os , 2004; Wilson e al., 2006; Mollon e al., 2013). The non- oxic biodeg ada ion o he sca old should g adually ans e he s ess o he new g owing issue o e an app op ia e ime pe iod. As poin ed ou ecen ly (Panade o e al., 2016), he syne ge ic e ec o co ec mechanical s imula ion is g ea ly dependen on he sca olding ma e ial, i s en i onmen and he cell p esence. This shows he needs o consis en simul aneous analysis o compa e di e en bioma e ials and o ge conclusions abou hese ea u es. One o he mos challenging applica ions o biomedical sca olds is he a icula ca ilage (AC) epai . The damage and deg ada ion o AC a e no only p og essing wi h age, obesi y, o sys emic diseases, bu also in he young and ac i e popula ion due o physical causes, such as inju y. I un ea ed, hese de ec s may p og ess owa d os eoa h i is (OA), a ec ing o e 150 million people wo ldwide, mainly by degene a ion o he hyaline ca ilage in syno ial join lacking he abili y o sel - egene a ion (A ms ong and Mow, 1982). Na u al wound healing, in ull- hickness de ec s o ca ilage, o en leads o he o ma ion o ib oca ilage (A ms ong and Mow, 1982; Wilson e al., 2006; Mollon e al., 2013; Panade o e al., 2016), which is unc ionally and biomechanically in e io o he o iginal hyaline ca ilage making he issue mo e p one o u he de e io a ion and os eoa h i ic changes o he join . Ini ia ed icious cycle (Bende s e al., 2012) ul ima ely will call o a o al o pa ial join eplacemen . The e o e, chond o-conduc i e and -induc i e bioma e ials a e highly desi able o ea ca ilage lesions a ea ly s ages be o e mani es a ion o OA. Clinically used bioma e ials include a ious na u ally de i ed and syn he ic ma e ials ( on Recum, 1998; Ag awal and Pa JE, 2000). The ad an age o na u al ma e ials is hei in insic bioac i i y o he pu pose, al hough applica ion o animal- de i ed ma e ials (xenog a s) con ains ce ain isks, such as con amina ion and undesi ed immune esponse. This could be a oided by using bioabso bable syn he ic ma e ials no causing o eign body o hype sensi i i y eac ions hemsel es. Syn he ic ma e ials can be made biologically mo e ad an ageous and biocompa ible. On he o he hand, compa ed o he na u ally de i ed ma e ials, syn he ic polyme s a e usually lacking he desi ed in insic biological cues ha p omo e cell adhesion, p oli e a ion and issue eco e y. Howe e , any bioma e ial is always challenging o e alua e and op imize o clinical use and o he pu pose aiming on “p ecise medicine” solu ions. I is now widely an icipa ed ha he p esen le el o e alua e he mechanical unc ion o bioma e ial and issue enginee ing cons uc s is highly insu icien . Fo example, o 205 analyzed a icles on ca ilage issue enginee ing, men ioning o applied mechanical s imula ion, only 29% shows some quan i ied ma e ial p ope ies (Lujan e al., 2011). Co ec and de ailed bioma e ial es ing is a he ime-consuming and expe ise o p ope ly quan i y non-elas ic ma e ial beha io o issue is also sca ce in many dedica ed biology labs (Lujan e al., 2011). Syn he ic ma e ials wi h ib ous o igin a e o en used o AC epai applica ions. These sca olds ha e 75–85% po osi y and hey a e exposed o syno ial luid wi h sodium hyalu ona e (NaHA). Animal s udies a e needed o ensu e he biological unc ionali y o he sca olds be o e clinical use. Howe e , he ela ionship be ween he na u al issue and he sca old is challenging o measu e. The egula ions and he wo ldwide ends impose mo e p essu e o mo e om animal models in o in i o e alua ion (Di ec i e 2010/63/EC o Al e na i e Me hods, 2015). The e o e, in o de o de elop and op imize bioma e ials, one mus es ablish p o ocols o eliable compa ison o di e en ma e ials be o e in i o es s can be e hically jus i ied and hei esul s uly ex apola ed owa d sa e and e ec i e human use. The s uc u e, unc ions and biomechanical beha io o AC a e e y complex, highly aniso opic and ime- and loading his o y-dependen (Wilson e al., 2005). The a icula ca ilage consis s o a ela i ely small numbe o chond ocy es su ounded by a mul i-componen ma ix, which can be imaged as a composi e wi h 70–85% wa e and emaining p o eoglycans (p o eins wi h glycosaminoglycans a ached as a bo leb ush-like s uc u e) and collagen (Hayes, 1972). P o eoglycans and wa e concen a ion a y h ough he dep h o he ca ilage issue. P o eoglycans can bind o agg ega e o a backbone o sodium hyalu ona e (NaHA) o molecula weigh o 2–4 MDa o o m a mac omolecule weigh ing up o 200 MDa (Kobayashi e al., 1994). The biomechanics o AC and syno ial luid is also complex and essen ially non-linea (Hayes and Mock os, 1971; Hayes and Bodine, 1978). No some many s udies ha e cohe en ly and sys ema ically analyzed AC p ope ies (Ahsan and Sah, 1999; Ko honen e al., 2006) due o a iabili y o he samples, local inhomogenei y and applied biomechanical me hods. Complex loading schemes a e associa ed wi h signi ican a ia ions o in e s i ial luid p essu e and luid low, complica ing he esul s in e p e a ion (Ahsan and Sah, 1999). The collagen- ich ma ix beha io is highly non-linea and equi es a he sophis ica ed models o be desc ibed as a composi e ma e ial, whe he wi h heo ies (Mäkelä and Ko honen, 2016). Syno ial luid is well- known o ha e non-New onian iscosi y s. i s composi ion, shea a e, mode o loading and he p esence o o he ac o s (King, 1966). Mos o he biomechanical p ope ies o AC issue epo ed expe imen ally a e ob ained wi h ei he con ined comp ession (Mow e al., 1980) o inden a ion (Kempson e al., 1971). These measu emen s da a a e commonly app oxima ed wi h biphasic (Mak e al., 1987) o iphasic (Lai e al., 1991) heo ies, o e en mo e simpli ied iscoelas ic models. Howe e , due o peculia i ies o he AC issue p ope ies (Lai e al., 1981), i is di icul o compa e esul s published wi h di e en s udies, using a ious specimen ypes, me hods and es ing de ices. I was also epo ed (Hosseini e al., 2014) ha luid low and low-dependen phenomena may domina e he AC beha io in di e en es ing egimes and hus i is impossible o de e mine in gene al equi ed eco e y ime. Agg ega e modulus in ange o 50–120 kPa was epo ed o human, bo ine and canine issues by di e en sou ces (Hayes and Mock os, 1971; Kempson e al., 1971; Hayes, 1972; Mow e al., 1980; Lai e al., 1981; A ms ong and Mow, 1982; Wilson e al., 2006), bu o en ull es da a F on ie s in Bioenginee ing and Bio echnology | www. on ie sin.o g 2No embe 2018 | Volume 6 | A icle 187 Gasik e al. No el Sca olds o A icula Ca ilage Repai we e no a ailable o compa e hese da a [inden a ion usually p oduces much la ge alues (Ahsan and Sah, 1999; Ko honen e al., 2006)]. Fo mal models o AC a e missing essen ial ea u es which limi hei p ac ical applica ion only o specimens analyzed in ha s udies. The e o e, i is a g ea o e simpli ica ion o cha ac e ize AC o sca olds o AC epai by se o one o wo numbe s wi hou exac da a on he es me hod and da a analysis. In his s udy we used highly po ous PLA mesh manu ac u ed om ine PLA ibe s. E en hough PLA i sel is a s i ma e ial, his s udied PLA mesh was op imized o ha e a ela i ely so na u e o sui be e as ca ilage epai ma ix. The hypo hesis was ha a sca old which is less s i han su ounding issue and which is ac ing in comp ession unde he equi emen o s ain compliance will ha e less s ess and he e o e luid p essu e which would cause luid o low in o he sca old o b ing mo e nu ien s o chond ocy es. The collagen componen was added o he PLA mesh o inc ease he hyd ophilic na u e o he sca olds and o p omo e cell p oli e a ion (Muhonen e al., 2016; Gasik e al., 2017). He e we epo esul s o his new xeno- ee, ecombinan human collagen-laden ( hCol) polylac ide (PLA) mesh sca olds ( hCol-PLA) de eloped o epai o ea ly ca ilage lesions o a oid os eoa h i ic changes, which ha e been designed, p oduced, and biomechanically op imized in i o and in i o alida ed in equine (unpublished da a) and po cine models (Muhonen e al., 2016; Gasik e al., 2017). The hCol-PLA sca olds we e assessed o hei ela i e pe o mance in simula ed syno ial luids o mimicking bo h human and e e ina y condi ions wi h applica ion o model- ee analysis wi h Bioma e ials Enhanced Simula ion Tes (BEST). The esul s o he sca old ma e ials selec ion we e also co ela ed wi h in i o es s, ca ied ou in a sepa a e s udy (Haapa an a e al., 2014; Muhonen e al., 2016), whe e his ma e ial combina ion was ound o wo k be e han he p e iously s udied plain PLA sca olds wi h s i e s uc u e (Pulliainen e al., 2007). MATERIALS AND METHODS Ma e ials Analyzed The sca olds es ed we e made o syn he ic polyme ibe s. The polylac ide sca old (PLA) was p ocessed o medical g ade poly-(L/D)-lac ide PLA96/4 (Co bion Pu ac, Go inchem, NED), manu ac u ed o mel spun ibe s and a e wa ds ca ded and needle punched in o meshes (po osi y ∼90–93%) in Tampe e Uni e si y o Technology (Tampe e, Finland). The used PLA was a highly pu i ied, medical g ade polyme (Länsman e al., 2006) wi h a esidual monome con en o <0.5%. The PLA meshes we e washed wi h e hanol, d ied, packed and s e ilized by gamma i adia ion 25 kGy. A pa o PLA sca olds was asep ically doped (Haapa an a e al., 2014) wi h a ecombinan human collagen III (Fib oGen, Inc., San F ancisco, USA) solu ion and he s uc u e was eeze-d ied (ma ked as hCol-PLA). The hCol- PLA sca olds we e u he c osslinked wi h 14 mM 1-e hyl-3- (3-dime hylaminop opyl)-ca bodiimide hyd ochlo ide (EDC) + 6 mM N-hyd oxysuccinimide (NHS) (Sigma-Ald ich, Helsinki, Finland) in 95% e hanol, washed and subsequen ly eeze-d ied again. The a io be ween he PLA and collagen componen s in he hCol-PLA sca olds was 86/14 ol. % o PLA and collagen, espec i ely. All he specimens o PLA and hCol-PLA sca olds we e cu in o ec angula pieces ∼5×5 mm (±1 mm) in size wi h he hickness o he o iginal ma e ials as supplied. The exac size o he specimens was measu ed wi h a non-con ac me hod using a lase mic ome e (Me aLigh , CA, USA) wi h ±1µm esolu ion and he samples we e weigh ed wi h a balance be o e and a e he es . The measu ing and weighing p ocess was epea ed h ee imes; on he d y sample, on he imme sed sample and on he sample a e he measu emen . The samples we e imme sed in dis illed wa e o ensu e ha he sample was comple ely we be o e inse ing o he sample holde as possible apped ai bubbles may al e he es esul s imp ope ly. The media-simula ed syno ial luid (SSF)— o he es s was p epa ed as wo solu ions wi h di e en concen a ions o sodium hyalu ona e (NaHA). Sodium hyalu ona e o molecula weigh 1.68 MDa (Nu ihyl R , Con ip o Bio ech, Czech Republic) was dissol ed in 200 mL o cold dis illed wa e o mimic “no mal” (4 mg/mL) and “os eoa h i ic” (1 mg/mL) solu ions (Fam e al., 2007). Expe imen al Me hods The iscosi y o he SSF solu ions was de e mined using SV- 10 ib o- iscosime e (A&D Co. L d., JAP) consis ed o wo ib a ing gold pla es imme sed in he solu ion. Abou 45 mL o he SSF solu ion was pou ed in a cu e e and hea ed o 40◦C. The cu e e was hen placed a he iscome e and he iscosi y wi h empe a u e was measu ed simul aneously upon ee cooling. Viscosi y cu es we e well- i ed o e e y composi ion wi h he A henius equa ion. The biomechanical analysis was ca ied ou using wo dynamic mechanical analysis (DMA) 242C and 242E machines (Ne zsch Ge ä ebau GmbH, Ge many) wi h a specially de eloped bioma e ials enhanced simula ion es (BEST; Seq e a L d., Finland) p o ocol (Gasik, 2014, 2017a,b), adjus ed o simula ed ca ilage condi ions (Hayes and Bodine, 1978; Mow e al., 1980; Lai e al., 1981). The comp essi e mode sample holde and he specimen we e ully imme sed in he he mally con olled ba h wi h media (∼30 mL). This esembles he g adien s o de o ma ion, po e p essu e and luid low simila o ibial ca ilage condi ions as has been shown wi h o he expe imen s and compu e simula ions (Ko honen e al, 2002; Milan e al., 2010). Th ee di e en p o ocols we e applied o imme sion es s: c eep (n=38 o PLA and n=40 o hCol-PLA), equency scans (0.01–20 Hz) as n=38 o PLA and n=28 o hCol-PLA, and s ain sweeps up o 25–50 µm a 1 Hz) as n=20 o PLA and n=30 o hCol-PLA. A p econdi ioning s ep was applied (Piole i and Rako omanana, 2000) o all specimens by an axial con inemen by ∼5µm o ini ial de o ma ion (o se ), ollowing he 15 min equilib a ion unde a small o ce o 0.05 N o s abilize he dimensions and empe a u e. This was ound o supp ess ini ial swelling (whe e p esen ; as explained below) hus all he c eep de o ma ion and compliance a e o igina ed om ze o. A hese condi ions i was obse ed ha de o ma ion o he po ous, ully sa u a ed ib ous s uc u es p oceeded wi hou excessi e F on ie s in Bioenginee ing and Bio echnology | www. on ie sin.o g 3No embe 2018 | Volume 6 | A icle 187 Gasik e al. No el Sca olds o A icula Ca ilage Repai de o ma ion o he ibe s hemsel es and wi hou subs an ial dec ease in po osi y. Whe eas, in i o p essu es exp essed on heal hy a icula ca ilage may each 1–10 MPa a peak, he e he luid low is uncon ined bu he sca old as AC issue unde goes de o ma ion simila o one in clinical condi ions (Lai e al., 1981, 1991; Milan e al., 2010). Da a Analysis Expe imen al da a has been con e ed in o biomechanical alues and analyzed wi h an applica ion o idempo en ype analysis wi hou use o a ma e ial model (Gunawa dena, 1996; Piole i and Rako omanana, 2000; Li ino e al., 2001). This gi es an ad an age o e commonly epo ed moduli unc ional dependence as i allows ex ac ion o he ime-in a ian da a sui able o u u e p edic ion o he ma e ial beha io . The da a quali y epo ed o he same ma e ial migh be also con using, as no exac in o ma ion is gi en o condi ioning changes, and usually no solid p oo shown, e.g., abou sui abili y o he small s ain heo y o ma e ial linea i y (No is, 2008; Lujan e al., 2011). Such condi ions a e o en assumed by de aul , despi e i is o common knowledge ha “elas ic modulus” canno be uniquely de ined o ma e ial which does no ollow linea elas ici y model. Idempo en p ocessing, common in compu e echnology (Gunawa dena, 1996) p ese es he da a s uc u e and o iginal a iables wi hou demand o explici knowledge o hei unc ional dependencies. I obeys causali y p inciple ( esponse always comes a e he s imulus applied) and espec s he bounda ies o he modynamics (no iola ion o conse a ion laws). One essen ial ad an age in es ing o bioma e ials wi h his me hod is in aking in o accoun non-local e ec s—on he con a y o con en ional ma hema ical analysis, whe e he de i a i e o a unc ion is always local. Hence, he p edic o s can be used in in silico simula ions o calcula e, o example, mechano- egula i e index wi hou necessi y o explici local luid low de e mina ion (Maslo , 1970; Gasik, 2017b). S ess was calcula ed as Piola–Ki chho s ess, om he a io o ac ing o ce o he o iginal su ace a ea A0o he specimen, σ =F/A0.The s ain was calcula ed as Lag ange ( ue loga i hmic) s ain, ela ed o he ini ial hickness H0a he beginning o he c eep: ε( ) =ln(1 +1L( )/H0), whe e 1L( ) is he obse ed change in he specimen hickness wi h expe imen ime . Whe eas, o he s ain de ini ions can be also used, his one has a igo ous he modynamic a ionale (Xiao, 1995; Luba da and Chen, 2008). The a io o he s ain ε( ), o cons an s ess, σ0, is he c eep compliance C( ) =ε( )/σ0, which is he main eadou om he c eep expe imen s. Fo dynamic loading, he s ain ampli ude is om a ha monic signal wa e o m ex eme, aking in o accoun load his o y: εω( )=1 2ln H0+1L( )+a0 H0+1L( )−a0(1) whe e a0is he applied de o ma ion ampli ude a some ins an equency ω. In his o ma , he i e e sible c eep o simila de o ma ion is aken in o accoun o long expe imen imes ( ). All expe imen s we e p e o med in iplica e and o e e y equency o s ain es 10 cycles we e used wi hin e e y un. Hence o ime dependencies one should conside bo h long ime ( eal ime o he es —minu es and hou s) and sho ime ( ime span wi hin one o ew dynamic cycles—seconds). Tes eadou s om DMA expe imen s ha e been p ocessed wi h model- ee idempo en me hods (Gasik and Bilo sky, 2018). In gene al, he e is no explici ma hema ical o mula w i en as he calcula ion is i e a i ely p og essing o e e y da a poin collec ed. This allows inclusion o specimen his o y wi hou a need o assump ion o ime ke nels (he edi a y in eg als). In his wo k he app oach was used o ind o ins ance agg ega e modulus, ma e ial memo y, s a ic and dynamic pe meabili y. RESULTS Ma e ials P epa a ion and P elimina y Analysis The iscosi y o SSF composi ions was app oxima ed as unc ion o empe a u e wi h A henius equa ion, leading o 8.14·10−4·exp(3022.2/T) and 1.746·exp(1143/T) in mPa·s, o 1 and 4 mg/mL NaHA, espec i ely (co ela ion 2=0.9869– 0.9953). A 25◦C his gi es 20.5 and 80.9 mPa·s alues o hese SSF ( o compa ison, wa e iscosi y is 0.89 mPa·s). The exac alues a e no explici ly equi ed because he DMA de ice and sample holde dynamics ha e been ecalib a ed o e e y ype o SSF, and hus media iscosi y changes ha e been au oma ically included in he es da a. A he beginning o he es s i was disco e ed ha PLA ma e ials exhibi e y high swelling a ios e en i hey we e comple ely soaked in liquid be o e he es s. This is usually aced in hyd ophobic ma e ials a ee swelling due o epulsi e o ces and p essu e a ia ions (Benne hum and Weins ein, 2004). I was o mula ed (Benne hum and Cushman, 1996) ha he mac oscopic solid s ess enso ( ela ed o isible bulk swelling) is combined o a he modynamic solid p essu e, a solid s ess enso , a s ess due o he in e ac ion o he solid and liquid phases, and a s ess due o he in e ac ion o he in e ace wi h he solid phase as well as kine ic componen o cons i uen s. By changing he condi ioning p essu e, i was ound ha he o ce o ∼50 mN (equi alen o applied ∼1 kPa s ess) is equi ed o supp ess he swelling bu no o cause p e-comp ession o he sample, Figu e 1. This condi ioning p essu e was used in all hese expe imen s o ge consis en esul s. Pseudo-S a ic Expe imen s The esul s o one se o c eep measu emen s a 0.2 N (∼4 kPa applied s ess) a e shown in Figu e 2. The le el o applied s ess o 4 kPa (∼30 mmHg) was conside ed o be a limi which does no cause ca ilage-adjacen so issues nec osis (Goode and Shinn, 1977). The na u e o a c eep es is pseudo-s a ic (change o s ain in ime a cons an applied s ess) and i is o en used o e alua e iscoelas ic na u e o ma e ials and o app oxima e i wi h some models (Bilo sky and Gasik, 2015). He e one may see ha addi ion o NaHA o media does no a ec compliance o PLA much, bu has a g ea e ec on hCol-PLA. I is no able ha compliance o hCol-PLA in 1 mg/mL SSF is he highes , and in 4 mg/mL is a e age be ween 1 mg/mL and wa e (0 F on ie s in Bioenginee ing and Bio echnology | www. on ie sin.o g 4No embe 2018 | Volume 6 | A icle 187 Gasik e al. No el Sca olds o A icula Ca ilage Repai FIGURE 1 | The e ec o condi ioning o ce (0.006 and 0.05N) on swelling and compac ion (a 0.2N) o PLA sca old ( wo expe imen al cu es o ei he o ces combina ion). FIGURE 2 | C eep compliance a 0.2 N o P LA and hCol-PLA in di e en media (numbe s in pa en heses show NaHA concen a ion in SSF, mg/mL). mg/mL NaHA). The highe is he compliance, he mo e easily he ma e ial de o ms unde cons an load. Hence an obse a ion can be d awn ha addi ion o c oss-linked collagen o PLA makes i “s i e ” when es ed in wa e bu makes li le di e ence when es ing in 1 mg/mL SSF. The compa ison o he da a om pseudo-s a ic (c eep) analysis (Figu e 3) shows ha s i ness o he hCol-PLA sca olds is inc eased by se e al imes when compa ed o PLA, and his e ec is independen on he ype o SSF used. Also hCol-PLA ma e ial in s a ic condi ions has lowe pe meabili y (Figu e 4), which in combina ion suppo s a ision ha syno ial luid will likely be kep in hCol-PLA be e han in PLA—a he same loading, wi hin he same ime span PLA will lose luid o a g ea e ex en . FIGURE 3 | S a ic agg ega e modulus (kPa) o he sca olds s. NaHA concen a ion in SSF. Ba s he e and u he indica e s anda d e o unless s a ed o he wise. F equency Expe imen s Beha io o ma e ials unde cons an de o ma ion bu a ied equency is di e en om pseudo-s a ic one. These di e ences in he case o ib ous po ous ma e ials a e due: (1) oscilla ing mobili y o luid wi hin a ib ous s uc u e, (2) ine ia e ec s associa ed wi h hys e esis be ween incoming and ou going luid low, and (3) non-linea i y in luid iscous p ope ies and possible non-linea i y in coupled de o ma ion o he ib ous skele on o he sca old. One o he expe imen al c i e ia o obse e he di e ences is he loss angen [ an(δ)], which is de ined as he a io o imagina y o eal pa o elas ic moduli o s i ness. Highe loss angen o hCol-PLA s. PLA (Figu e 5) was obse ed o all equencies. Figu e 6 shows a 3D plo o hese dependencies o F on ie s in Bioenginee ing and Bio echnology | www. on ie sin.o g 5No embe 2018 | Volume 6 | A icle 187 Gasik e al. No el Sca olds o A icula Ca ilage Repai FIGURE 4 | S a ic (c eep) pe meabili y o sca olds s. NaHA concen a ion in SSF, indica ing ha s a ic pe meabili y inc eases a highe NaHA concen a ion. FIGURE 5 | Loss angen o he sca olds in di e en SSF (NaHA concen a ion as numbe s, mg/mL) s. equency a 25 µm de o ma ion ampli ude. an(δ) s. applied equency a 25 µm de o ma ion ampli ude along wi h expe imen ime (no e he angen is p ac ically cons an wi h he ime and depends essen ially on equency only). I is seen ha addi ion o NaHA o he media and espec i e inc ease in iscosi y also leads o inc ease in he loss angen ( he mo e, he highe is he NaHA concen a ion). Fo hCol-PLA loss angen is oughly 2–3 imes highe han o PLA in all SSF (Figu e 6). This indica es mo e ac i e in e ac ion o luid low wi h hCol-PLA han PLA and is likely associa ed wi h a ines collagen ib ils ne wo k be ween he PLA-based ibe s in hCol- PLA (Länsman e al., 2006; Muhonen e al., 2016; Gasik e al., 2017). No able is ha his in e ac ion is only due o p esence o NaHA, as such di e ences a e no seen when only wa e is used (Figu e 6). S ain Sweep Expe imen s One o he mo e physiologically ele an dynamic condi ions is applica ion o a iable de o ma ion unde cons an equency. This can be depic ed as change o he gai loads keeping no mal walking condi ions (∼1 Hz) (Hayes and Mock os, 1971; Hayes, 1972; Mow e al., 1980; A ms ong and Mow, 1982). Thus, he las es sequence was applied o simula e changes in p ope ies o sca olds up o 50 µm o dynamic de o ma ion wi h epea ing o he load cycles. A cons an equency, loss angen is no F on ie s in Bioenginee ing and Bio echnology | www. on ie sin.o g 6No embe 2018 | Volume 6 | A icle 187 Gasik e al. No el Sca olds o A icula Ca ilage Repai FIGURE 6 | Loss ac o [ an(δ)] o sca olds in di e en SSF s. applied equency a 25 µm displacemen ampli ude. No e plo s o wa e (0 mg/mL NaHA; su aces No. 5 and 6) a e shi ed ahead o ime scale o imp o e isual eadabili y o he plo s. FIGURE 7 | A e age dynamic modulus o sca olds a 1 Hz s. applied de o ma ion ad di e en SSF (numbe s indica ing NaHA concen a ion, mg/mL). No e log scale o modulus. signi ican ly de o ma ion- o s ess-dependen so he e majo pe o mance comes om dynamic s i ness and luid exchange. The absolu e alue o a e age dynamic elas ic modulus a 1 Hz is shown in Figu e 7 as di ec ly ob ained om he DMA signal. I is seen ha his modulus sligh ly dec eases wi h de o ma ion. Howe e , wi h inc eased numbe o loading cycles and ue s ain a ia ions due o changes in geome y, ue (co ec ed) elas ic modulus sligh ly inc eases. Also, highe NaHA concen a ion shows highe s i ness o all ma e ials bu i is no ewo hy his s i ness inco po a ed luid mo emen unde dynamic load and he e o e una oidably includes some iscous and ine ia con ibu ion, as shown abou o loss angen (Figu es 5,6). F on ie s in Bioenginee ing and Bio echnology | www. on ie sin.o g 7No embe 2018 | Volume 6 | A icle 187 Gasik e al. No el Sca olds o A icula Ca ilage Repai FIGURE 8 | Compa ison o s a ic agg ega e modulus (Figu e 3) wi h in a ian dynamic modulus a 1 Hz. Numbe s indica e concen a ion o NaHA in SSF, mg/mL. FIGURE 9 | Compa ison o a io o dynamic in a ian modulus o dynamic pe meabili y s. NaHA concen a ion (no e log scale). Analyzed in a ian modulus a 1 Hz is shown in Figu e 8 s. espec i e s a ic agg ega e modulus (Figu e 3) o espec i e NaHA concen a ion. The a io o dynamic modulus o dynamic pe meabili y is shown in Figu e 9 o 1 Hz condi ion. He e hCol-PLA is a leas simila o be e (a 1 mg/mL NaHA) han PLA sca olds. DISCUSSION The wo s udied sca olds con ained he same kind o PLA mesh s uc u es. The PLA sca olds we e s udied as such and in he hCol-PLA sca olds he collagen componen was added in o he s uc u e o gi e he highly po ous PLA sca old inc eased hyd ophilic na u e and o p omo e cell p oli e a ion. F om he pos -p ocessing o he expe imen al da a, many addi ional alues ha e been ob ained wi hou assump ion o a ma e ial model (a p op ie a y pa en -pending me hod). He e da a o agg ega e modulus (in s a ic and dynamics) and pe meabili y a e shown as an example. The ele ancy o s a ic condi ions esul s o clinical condi ions is ha hCol-PLA sca olds a e be e suppo i e o weigh -bea ing and unde go smalle de o ma ion han pu e PLA. In combina ion o lowe pe meabili y his sugges s syno ial luid o s ay likely hCol-PLA mo e han in PLA, whe eas he la e will lose mo e luid a he same loading. As a simple decision-aiding c i e ion, one migh conside he a io o agg ega e modulus o F on ie s in Bioenginee ing and Bio echnology | www. on ie sin.o g 8No embe 2018 | Volume 6 | A icle 187 Gasik e al. No el Sca olds o A icula Ca ilage Repai pe meabili y: he highe i is, he be e he sca old wi hs ands s a ic loads. Figu e 10 shows ha in his espec hCol-PLA ma e ial is by 1–2 o de s o magni ude supe io o PLA a all es ed SSF composi ions. A dynamic condi ions, such as walking, he si ua ion e e ses: unde dynamic loading one has o aim on mo e ac i e luid exchange o p o ide biomechanical s imulus o chond ocy es, o ensu e luid and nu ien s supply and emo al o me abolic p oduc s, p omo ing issue egene a ion (Ge isch and Chaplain, 2006; Si ichokechaiwu e al., 2010). Ca ilage issue is a ascula and i s ex acellula ma ix c ea es u he ba ie s o nu ien /was e exchange by di usion. Thus, lowe dynamic modulus (Figu e 8) and be e luid di usi i ies a e desi ed. These ea u es ha e o be, howe e , compa ible o he abo e equi emen s o s a ic condi ions as ca ilage mus wo k well in bo h hese ex emes. The iscosi ies o SSF a e subs an ially highe han wa e : abou 20 and 100 imes o 1 and 4 mg/mL, espec i ely. This means ha e en small changes in pe meabili y, i.e., a ea u e o he ma e ial s uc u e, will a ec changes in pe mi i i y, a ea u e o a speci ic luid low hough he ma e ial s uc u e. High loss angen (Figu es 5,6) means mo e dissipa ion o FIGURE 10 | Compa ison o a io o s a ic agg ega e modulus (Figu e 3) o pe meabili y (Figu e 4) s. NaHA concen a ion (no e log scale). applied mechanical ene gy (inelas ic losses) which is impo an o keep high damping p ope ies o a icula ca ilage (Mow e al., 1980; Lai e al., 1981; A ms ong and Mow, 1982). The e o e, hCol-PLA is supe io o PLA also in his p ope y, whe he o “a h i ic” (1 mg/mL NaHA) o “no mal” (4 mg/mL NaHA) syno ial luids. A p oo -o -concep animal s udy was pe o med in domes ic pigs (Sus sc o a domes ica, 4-mon hs-old, n=20) in a sepa a e s udy, epo ed elsewhe e (Muhonen e al., 2016). B ie ly, he animals we e andomized in o h ee g oups: (1) hCol-PLA sca old ea men , (2) comme cial sca old ea men and (3) spon aneous epai . A ci cula ull- hickness chond al lesion wi h a diame e o 8 mm was c ea ed in he igh medial emo al condyle. The emo ed ca ilage issue was collec ed and u he p ocessed o chond ocy e isola ion and subsequen p oli e a ion. A e 3 weeks, he lesion was app oached again, cleaned and epai ed wi h one o he cons uc s, i.e., hCol- PLA o comme cial sca old wi h chond ocy es, o le un ea ed. Only one lesion pe animal was pe o med and he animals we e allowed ee weigh -bea ing and un es ic ed mo emen a e he ope a ions. The epai issue was e alua ed a e 4 mon hs. Hyaline ca ilage was epo ed o be o med mos equen ly in he hCol-PLA ea men g oup. He e he analysis o he ca ilage epai sco es (Haapa an a e al., 2014; Muhonen e al., 2016) was addi ionally pe o med using BUGS—Bayesian in e ence Using Gibbs Sampling (US FDA, 2010), a o m o a Ma ko Chain Mon e Ca lo sampling. The esul s o using no mal o Poisson dis ibu ions o he o al no malized ICRS (In e na ional Ca ilage Repai Socie y) sco es show hCol-PLA ha ing s a is ically signi ican highe a e age sco e (0.515) s. 0.38 o comme cial sca old and 0.288 o spon aneous healing con ol g oup (Muhonen e al., 2016). A schema ic o he ad an age o lowe dynamic s i ness sca old o AC epai is depic ed in Figu e 11. The e a sca old implan ed in o a ca ilage de ec should exhibi he same s ain compliance (no ea s, wis ing o buckling). Fo he same s ain, ma e ial wi h a highe appa en s i ness will gene a e mo e in e nal s esses and hence highe luid p essu e (Hayes, 1972; Mow e al., 1980; Lai e al., 1991). This will lead o p e e en ial luid mo emen ou o ca ilage which will no allow cells and issue egene a ion (“d y-ou ”). Fo an opposi e, lowe FIGURE 11 | Possible mechanism o bene icial compliance misma ch o sca olds s. na i e issue when submi ed o a mechanical load: when a sca old wi h luid is oo s i , highe gene a ed luid p essu e leads o luid low ou o sca old (le ), whe eas o he same de o ma ion condi ions o lowe s i ness sys em ( igh ) luid is d i en in o he sca old. F on ie s in Bioenginee ing and Bio echnology | www. on ie sin.o g 9No embe 2018 | Volume 6 | A icle 187