Functional 3-D Cardiac Co-Culture Model Using Bioactive ...cho/Hussain_12.pdf · like scaffold,...
Transcript of Functional 3-D Cardiac Co-Culture Model Using Bioactive ...cho/Hussain_12.pdf · like scaffold,...
ARTICLE
Functional 3-D Cardiac Co-Culture Model UsingBioactive Chitosan Nanofiber Scaffolds
Ali Hussain, George Collins, Derek Yip, Cheul H. Cho
Department of Biomedical Engineering, New Jersey Institute of Technology, University
Heights Newark, New Jersey 07102; telephone: 973-596-5381; fax: 973-596-5222;
e-mail: [email protected]
ABSTRACT: The in vitro generation of a three-dimensional(3-D) myocardial tissue-like construct employing cells, bio-materials, and biomolecules is a promising strategy incardiac tissue regeneration, drug testing, and tissue engi-neering applications. Despite significant progress in thisfield, current cardiac tissue models are not yet able to stablymaintain functional characteristics of cardiomyocytes forlong-term culture and therapeutic purposes. The objectiveof this study was to fabricate bioactive 3-D chitosan nano-fiber scaffolds using an electrospinning technique andexploring its potential for long-term cardiac function inthe 3-D co-culture model. Chitosan is a natural polysaccha-ride biomaterial that is biocompatible, biodegradable, non-toxic, and cost effective. Electrospun chitosan was utilizedto provide structural scaffolding characterized by scale andarchitectural resemblance to the extracellular matrix (ECM)in vivo. The chitosan fibers were coated with fibronectin viaadsorption in order to enhance cellular adhesion to the fibersand migration into the interfibrous milieu. Ventricularcardiomyocytes were harvested from neonatal rats andstudied in various culture conditions (i.e., mono- andco-cultures) for their viability and function. Cellular mor-phology and functionality were examined using immuno-fluorescent staining for alpha-sarcomeric actin (SM-actin)and gap junction protein, Connexin-43 (Cx43). Scanningelectron microscopy (SEM) and light microscopy wereused to investigate cellular morphology, spatial organiza-tion, and contractions. Calcium indicator was used tomonitor calcium ion flux of beating cardiomyocytes. Theresults demonstrate that the chitosan nanofibers retainedtheir cylindrical morphology in long-term cell cultures andexhibited good cellular attachment and spreading in thepresence of adhesion molecule, fibronectin. Cardiomyocytemono-cultures resulted in loss of cardiomyocyte polarityand islands of non-coherent contractions. However, thecardiomyocyte-fibroblast co-cultures resulted in polarizedcardiomyocyte morphology and retained their morphologyand function for long-term culture. The Cx43 expression inthe fibroblast co-culture was higher than the cardiomyocytesmono-culture and endothelial cells co-culture. In addition,
fibroblast co-cultures demonstrated synchronized contrac-tions involving large tissue-like cellular networks. To ourknowledge, this is the first attempt to test chitosan nanofiberscaffolds as a 3-D cardiac co-culture model. Our resultsdemonstrate that chitosan nanofibers can serve as a potentialscaffold that can retain cardiac structure and function. Thesestudies will provide useful information to develop a strategythat allows us to generate engineered 3-D cardiac tissueconstructs using biocompatible and biodegradable chitosannanofiber scaffolds for many tissue engineering applications.
Biotechnol. Bioeng. 2012;xxx: xxx–xxx.
� 2012 Wiley Periodicals, Inc.
KEYWORDS: electrospinning; chitosan nanofibers; fibro-nectin adsorption; cardiomyocytes; co-cultures
Introduction
Heart disease is a preeminent cause of death in the world,responsible for about 40% of human mortality (Chen et al.,2008). Ischemic heart disease followed by an episode ofacute myocardial infarction is the most frequent cause ofleft-sided heart disease. Myocardial infarction can result inmyocyte slippage; the replacement of the damagedmyocardium with a non-contractile fibrous scar tissuecompromises heart contractile efficiency. The damage to theheart wall is permanent because, following extensive cellloss, the myocardium does not possess intrinsic regenerativecapability (Baig et al., 1998). As a result, the heart failsto pump sufficient amounts of blood to compensatethe ventricles undergoing remodeling and hypertrophy;predisposing the heart to congestive heart failure.Pharmacological therapeutics can reduce the heart’s workload via diuretics, nitrates, and humoral factor blockers suchas ACE inhibitors or B-blockers (Doughty et al., 1997;Sharpe et al., 1991). Interventional therapy, such asimplanting pacing devices can be performed to controlelectrical/mechanical asynchrony (Martinez et al., 2007).Nonetheless, pharmaceutics and surgery fall short frompreventing progression to end stage heart failure (Packer,2002). Ultimately, heart transplantation is the final option
Correspondence to: C. H. Cho
Additional supporting information may be found in the online version of this article.
Received 6 May 2012; Revision received 14 August 2012; Accepted 5 September 2012
Accepted manuscript online 18 September 2012;
Article first published online in Wiley Online Library
(wileyonlinelibrary.wiley.com).
DOI 10.1002/bit.24727
� 2012 Wiley Periodicals, Inc. Biotechnology and Bioengineering, Vol. xxx, No. xxx, 2012 1
for end-stage heart failure. The current survival rate of hearttransplants at 1 year after surgery is�85% and at 5 years it is75% (Stehlik et al., 2011). However, due to the scarcity ofheart donors and the complications related to immunesuppression and rejection, researchers are turning towardscardiac tissue engineering to develop alternative solutions torevive the distressed heart.
Several studies published in the early 1990s depicted long-term survival and differentiated phenotype of transplantedskeletal myoblasts in the ventricular myocardium (Koh et al.,1993a,b; Soonpaa et al., 1994). These studies paved the wayfor cardiac tissue engineering and shed light on the possibilityof replacing the injured myocardium with healthy contractilecells. The ex vivo generation of an implantable myocardialcontractile tissue-like unit is one of the most intenselyresearched topics in cardiac tissue engineering. The complex-ity of the myocardial tissue architecture demands theintricate design and integration of an extracellular matrix-like scaffold, cells, and biological factors. Cells sense thespatio-mechanical features of their microenvironmentthrough integrin and mechanoreceptor mediated linksbetween the external scaffold and the cell cytoskeleton(Ingber, 1997, 2003; Stupack, 2005). This allows the cells tosense the scaffold’s chemistry, dimension, orientation, andmechanical properties that influence their survival andfunction (Schussler et al., 2010). Furthermore, the cellularcomposition of the engineered unit must resemble the nativemyocardium’s ratio of fibroblasts to myocytes (70:30) inorder to closely replicate its physiological function (Banerjeeet al., 2006; Vliegen et al., 1991). Fibroblasts have been shownto propagate electrical signals over distances of 100mm viagap junctions and help synchronize cardiomyocyte contrac-tions (Gaudesius et al., 2003). The intercellular dynamismchanneled through gap junctions (connexin-43) and para-crine signaling between fibroblasts and myocytes affectmyocyte structure and function (LaFramboise et al., 2007).Hence, the biomaterial used for generating the scaffold, thecell species populating the scaffold, and the biological factorsenhancing the cells’ performance and survival within thescaffold contribute to the successful fabrication of amyocardial contractile tissue-like unit.
Electrospinning is a fabrication technique by whichsubmicron to nanometer fibers are produced as a non-woven mat from an electrostatically driven jet of polymersolution. Electrospinning is being extensively studied fortissue engineering applications because of its ability toproduce nano-structures with a very high surface area tomass ratio (40–100m2/g; Desai et al., 2008; Pham et al.,2006; Teo and Ramakrishna, 2006). In addition, the fibrousstructure forms a network of interconnected voids thatprovides an environment that is similar to the in vivoECM. Several synthetic or natural biomaterials have beenused to fabricate nanofiber scaffolds using the electrospin-ning technique for cardiac tissue engineering applications(Rockwood et al., 2008; Shin et al., 2004; Zong et al., 2005).Although some advances have been made in the area,synthetic polymer-based scaffolds are not ideal for in vivo
applications due to the toxicity of degradation products.Collagen is a natural polymer that has been commonly usedin cardiac tissue engineering but its various applicationshave been limited by weak mechanical properties as asupportive scaffold and rapid in vivo degradation (Haughet al., 2011; Venugopal et al., 2012).
Chitosan, a natural polysaccharide, is widely used in tissueengineering because of its biocompatibility, biodegradability,non-toxicity, and its pH dependent solubility facilitating itsprocessing into micro- and nano-scaffolds (Cho et al., 2008b;Kim et al., 2008; Madihally and Matthew, 1999). Thechemical structure of chitosan is similar to the glycosami-noglycans, such as hyaluronic acid, in the extracellular matrix,and its hydrophilicity enhances its interaction with growthfactors, cellular receptors, and adhesion proteins (Kumar andMajeti, 2000; Matthew et al., 2003). Although electrospunnanofiber scaffolds using chitosan have been recently studiedin many tissue engineering applications (Jayakumar et al.,2010; Ohkawa et al., 2004), there have been no reports onchitosan nanofiber scaffolds for cardiac tissue engineeringapplications. Furthermore, current cardiac tissue models arenot yet able to stably maintain functional characteristics ofcardiomyocytes in long-term culture.
In this study, we demonstrate for the first time thecytocompatiblity of three-dimensional (3-D) chitosannanofiber scaffolds for cardiac tissue engineering applica-tions and improved cardiac function in the 3-D co-culturemodels. Chitosan nanofibers were fabricated using anelectrospinning technique and their potential as ascaffold was studied for the development of 3-D cardiactissue constructs. In order to enhance cell attachmentand growth, fibronectin was incorporated onto thechitosan nanofibers by adsorption. Neonatal rat cardio-myocytes were cultured on the chitosan nanofibers and theirviability, morphology, and function was assessed. For long-term survival and function, cardiomyocytes were co-cultured with either fibroblasts or endothelial cells on thechitosan nanofibers and examined by phenotypic andfunctional analysis.
Materials and Methods
Electrospinning of Chitosan
An 8% (w/v) chitosan solution was prepared by dissolvingchitosan (medium molecular weight �200K, 75–85%deacetylation; Sigma, St. Louis, MO) in Trifluoroaceticacid (TFA; Fisher Chemicals, Pittsburgh, PA). The solutionwas stirred overnight at 408C. Methylene Chloride (MeCl;Fisher Chemicals) was added to form a final volume tovolume ratio of 80:20 (TFA:MeCl). The chitosan solutionwas fed into a 10mL disposable syringe fitted with an 18gauge needle. A DC voltage of 30 kV was applied to theneedle and the planar collector was placed 30 cm fromthe needle. The polymer solution was pumped at a rate of2mL/h and the process was performed at room temperature
2 Biotechnology and Bioengineering, Vol. xxx, No. xxx, 2012
and atmospheric humidity of 40–50%. Following vacuumdrying at room temperature, the chitosan nanofibers werecut into pieces to fit into 35-mm dish and neutralized with15N ammonium hydroxide: 100% ethanol (1:1 v/v ratio) for30min. The chitosan nanofibers were then washed withdistilled water three times for 15min each time. Thechitosan nanofibers were then sterilized under a UV lampfor 20min.
Fibronectin Adsorption on Chitosan Nanofibers
In order to evaluate fibronectin adsorption on chitosan, 24-well tissue culture dishes were coated with 1% chitosan anddried for 1 h before neutralization with 0.2M ammoniumhydroxide. Fibronectin (Sigma) solutions of differentconcentration (0, 1.2, 5, 10, and 20mg/mL in deionizedwater) were added into the dishes and incubated for 1 h atroom temperature. The amount of adsorbed fibronectin wascharacterized by fluorescent staining using anti-fibronectinantibody (Sigma).
Cell Culture
Primary ventricular cardiomyocytes were isolated from1-day-old neonatal Wistar rats (Charles River Laboratories,Wilmington, MA) using a collagenase procedure asdescribed previously (Aoki et al., 1998) and cultured on0.1% gelatin-coated dishes. Cardiomyocyte culture mediumconsists of DMEM (Gibco, Gaithersburgh, MD) with 10%FBS (Biowest, Miami, FL), 2mM L-glutamine (Gibco),insulin/transferrin/selenious acid (ITS; 5, 5, 5 ng/mL,respectively) (Invitrogen, Grand Island, NY), and 2%penicillin/streptomycin (Gibco). Murine 3T3-J2 fibroblasts(purchased from Howard Green, Harvard Medical School,Boston, MA), were maintained in 60-mm tissue culturedishes in DMEM plus 10% FBS and 2% penicillin andstreptomycin. The 3T3-J2 fibroblasts cell line was used forcardiac co-cultures because of their easy access, propagation,and high induction of epithelial cell functions (Bhatiaet al., 1999; Cho et al., 2008a; Khetani et al., 2004). Primaryrat heart microvessel endothelial cells (MVEC; purchasedfrom VEC Technologies, Rensselaer, NY), were maintainedin DMEM supplemented with 10% FBS, 2mM L-glutamine,ITS, 2% penicillin/streptomycin, and 10 ng/mL vascularendothelial growth factor (VEGF, R&D Systems,Minneapolis, MN). Culture medium was changed every3 days.
Cell Seeding on 2-D Chitosan Films and 3-D ChitosanNanofiber Scaffolds
For the 2-D cell cultures, the chitosan or chitosan-FN filmswere prepared in 24-well tissue culture dishes, as describedabove. Cells were seeded at a density of�25,000 cells/cm2 onthe films in 0.5mL of culture medium. For 3-D cellcultures, the nanofibers were treated with fibronectin
solution (10mg/mL) for 1 h to adsorb fibronectin on thefibers. The nanofibers were then soaked in culture mediumfor about 5min prior to cell seeding. The suspendedcells (fibroblasts, cardiomyocytes, or endothelial cells) inculture medium were directly seeded at a density of�300,000 cells/cm2 on separate chitosan nanofibrousmats with 300mL of medium and incubated for about1 h before adding further medium to allow for bettercell entrapment and attachment. For the co-culturestudies (cardiomyocytes-fibroblasts or cardiomyocytes-endothelial cells co-culture) in 2-D and 3-D culture systems,cells were seeded at a ratio of 1 to 1, giving a total initial cellnumber (cardiomyocytesþ fibroblasts or cardiomyocytesþendothelial cells) of �50,000 cells/cm2 for 2-D and�600,000 cells/cm2 for 3-D cultures. The mats wereincubated in a 10% CO2, 378C incubator and the mediumwas changed every 3 days.
Immunofluorescence Analysis
The samples were washed with PBS and fixed with 4%paraformaldehyde for 20min at room temperature. Afterwashing with PBS, the samples were permeabilized in 0.2%Triton X-100 in PBS for 10min. The samples were thenwashed with PBS and incubated in blocking buffer (PBS/10% FBS/1% BSA) for 30min. The primary antibody wasthen added and incubated for 60min at room temperature.The samples were washed with PBS and then incubated withthe secondary antibody for 60min. The samples were thenwashed and examined with fluorescence microscopy (NikonTi-S) with a digital color camera. The primary antibodiesused were mouse anti-vinculin (Millipore, Billerica, MA,1:200 dilution), mouse anti-a-sarcomeric actin (Invitrogen,1:50 dilution), and rabbit anti-connexin 43 (Sigma, 1:1,000dilution). The secondary antibodies used were donkey anti-mouse IgG, alexa fluor 488, and donkey anti-rabbit IgG,alexa fluor 594 (Invitrogen, 1:1,000 dilution). For thedistribution of actin microfilaments, cells were stained byincubating fixed and permeabilized cultures with 0.1mg/mLrhodamine phalloidin (Sigma) for 30min. In some cases,cells were counterstained with 4,6-diamidino-2-phenylin-dole (DAPI; Invitrogen) for nuclear staining. In order toverify the adsorption of fibronectin on chitosan, the sampleswere stained for 1 h at room temperature with rabbit anti-fibronectin (Sigma, 1:400 dilution). After washing twicewith PBS, the samples were incubated with alexa fluor 488conjugated anti-rabbit IgG (Invitrogen, 1:1,000 dilution),and washed twice with PBS. Cell viability on the nanofiberscaffolds was examined by a live/dead viability/cytotoxicitykit (Invitrogen). The samples were observed with fluores-cence microscopy.
Scanning Electron Microscopy (SEM)
The electrospun chitosan scaffolds and the cell seededscaffolds were examined with SEM (Leo 1350 VP) forstructural analysis. The samples were fixed with 2.5%
Hussain et al.: Functional 3-D Cardiac Co-Culture Model 3
Biotechnology and Bioengineering
glutaraldehyde in PBS for 24 h at 48C. The samples were thenwashed with PBS and serially dehydrated with 50%, 70%,and 100% ethanol for 15min each. This was done to allowgradual dehydration of the cells preventing loss of cellularstructural integrity. The samples were then vacuum-driedfor about 6 h. The samples were coated with carbon andobserved with SEM.
Calcium Transient Ion Staining
Calcium indicator, fluo-4 AM (Invitrogen), was used toobserve calcium ion flux of beating cardiomyocytes. Thecells were incubated in 4mM calcium indicator incardiomyocyte medium for 45min at 378C. The cellswere then washed with cardiomyocyte medium twice,followed by incubation in cardiomyocyte medium foranother 30min at 378C. The stained cardiomyocytes werevisualized by fluorescence microscopy. Videos of thecalcium ion staining were captured using Nikon NIS-Elements image software. Pseudo-color images of thecalcium ion fluorescence intensity were captured every150ms. The intensity of the calcium ion flux was determinedby taking the average intensity from 12 pixel image points
from each frame and plotting the average intensity overtime.
Quantitative Image and Statistical Analysis
Nikon Imaging Solutions (NIS)-elements imaging softwareprogram (v.3-448) was used to measure the diameterdistribution of the chitosan fibers. The measurements weredone from five different SEM images on different regions ofeach sample. A total of �150 measurements were analyzedfor the fiber diameter. To quantify the amount of adsorbedfibronectin on chitosan surfaces, average fluorescenceintensity from three to four random fields of images weremeasured and analyzed by the image software after stainingwith anti-fibronectin, alexa fluor 488. In order to analyze thecell attachment and spreading on chitosan and fibronectin-coated chitosan surfaces, the surface area of cultured cellswas measured and quantified by the image software. Threeto four random fields of image per sample were acquiredafter staining with actin/DAPI and quantified by the imageanalysis. Data from representative experiments are expressedas the mean� standard deviation (SD) of duplicate ortriplicate samples. All experiments were repeated at least 2–3times and similar results were reproduced. Statistical
Figure 1. SEM images of electrospun chitosan nanofiber scaffolds at (A) 5,000� original magnification and (B) 50,000� original magnification. C: Fiber diameter distribution of
the chitosan nanofibers. D: Photography of the chitosan nanofiber mat.
4 Biotechnology and Bioengineering, Vol. xxx, No. xxx, 2012
significance was determined by a two-tailed student’s t-test(�P< 0.01).
Results
Characterization of Electrospun Chitosan NanofiberScaffolds
Figure 1A and B shows SEM images of the nanofibrouschitosan non-woven mats fabricated using the electrospin-ning technique at 5,000� and 50,000� magnifications,respectively. The chitosan mats demonstrated homogeneouscylindrical morphology and well-formed fibers with afiber diameter ranging from 50 to 450 nm, and an averageof 188 nm, as illustrated in Figure 1C. The thickness of themats used in this study was �150mm. The randomorientation of the fibers produces many interconnectedspaces. The fibers did not dissolve and maintained theircylindrical morphology after neutralization with ammoni-um hydroxide.
Fibronectin Adsorption on the Chitosan Nanofibers
Cellular attachment to the fibers and infiltration into theinterfibrous spaces were enhanced by immobilizing fibro-nectin onto the chitosan nanofibers by adsorption. Toestablish the concentration at which chitosan is saturated byfibronectin adsorption, immunofluorescent staining ofadsorbed fibronectin using varying concentrations (0–20mg/mL) was performed. Figure 2A illustrates an increasein observed fluorescence intensity with fibronectin concen-tration, which suggests that fibronectin adsorption onchitosan coated wells is dependent on the concentration ofthe fibronectin solution. There was a steady increase in theamount of adsorbed fibronectin as the concentrationincreases and the adsorption plateaus beyond 10mg/mL.As a result, the 10mg/mL solution was used to adsorb
fibronectin on the chitosan nanofibers for improved celladhesion and migration.
Effect of Fibronectin Immobilization on CellularMorphology, Focal Adhesion Expression, andCytoskeletal Organization
In order to investigate the effect of fibronectin immobiliza-tion on chitosan material in cell cultures, the morphology,cytoskeletal protein distribution, and vinculin focal adhe-sion expression of various cell types were monitored on 2-Dchitosan films with and without fibronectin adsorption. Asshown in Figure 3, when the cells were cultured on chitosanfilm, they maintained a rounded morphology, minimalvinculin focal adhesion expression, and a diffused F-actincytoskeletal organization. In contrast, the cells cultured onfibronectin coated chitosan films demonstrated enhancedcellular spreading, significant increase in vinculin expressionand a well-organized fibrous F-actin cytoskeleton.
Characterization of Cellular Behavior in the ChitosanNanofiber Scaffolds
Assessment of cellular viability and morphology wasperformed to evaluate electrospun chitosan nanofibrousmat potential as cellular scaffolds. Fibroblasts, endothelialcells, and cardiomyocytes were seeded onto fibronectincoated chitosan mats and cultured over 3 weeks.Figure 4A–C depicts the live–dead cell staining of fibroblastscultured on the chitosan nanofiber scaffolds, indicating thatthe chitosan nanofibers do not adversely affect cell viability.Similar results were observed for other cell types, includingcardiomyocytes and endothelial cells (data not shown).Some cells formed filopodia-like extensions to attach to thefibers, assisting them in spreading inside the chitosannanofibrous scaffold (Fig. 4D–F). In addition, the SEMimages exhibit the formation of a film-like materialsurrounding the densely seeded areas, indicating the
Figure 2. Fibronectin (FN) adsorption on chitosan surfaces (2-D film and 3-D nanofibers). A: Relative fluorescence intensity and images of fibronectin adsorbed on chitosan
coated tissue culture dishes at various fibronectin concentrations by immunofluorescence staining. B,C: Phase and immunofluorescence staining for anti-fibronectin of chitosan
nanofibers adsorbed by fibronectin solution (10mg/mL) at 200� original magnification. Data are expressed as means�SD.
Hussain et al.: Functional 3-D Cardiac Co-Culture Model 5
Biotechnology and Bioengineering
secretion and immobilization of cell secreted ECMcomponents.
Cardiac Function in Mono- and Co-Cultures on the 2-DChitosan Films and 3-D Nanofiber Scaffolds
Cardiomyocyte morphology and gap junction formationwere monitored via alpha-sarcomeric actin (SM-actin) andconnexin-43 (Cx43) staining, respectively. Cardiomyocytes’SM-actin and Cx43 expression was examined on bothfibronectin adsorbed chitosan films (2-D) and fibronectinadsorbed chitosan nanofibers (3-D). In each condition,
cardiomyocytes were cultured in mono-cultures (cardio-myocytes only) and co-cultures (cardiomyocytes–fibroblastsor cardiomyocytes–endothelial cells).
In the 2-D systems, the cardiomyocyte mono-culture(Fig. 5A and D) exhibited low expression of SM-actinand the cardiomyocytes lost their structural polarity andacquired a rounded morphology. Gap junction proteinCx43 expression was minimal in the mono-culturesystem. In contrast, the cardiomyocytes co-cultured with3T3-J2 fibroblasts maintained a highly polar morphologyand the SM-actin was strongly expressed along theaxis of morphological polarity (Fig. 5B and E). Inaddition, Cx43 expression was the highest in the fibroblast
Figure 3. Morphology, F-actin distribution, and vinculin focal adhesion expression of cardiomyocytes, fibroblasts, and endothelial cells cultured on 2-D chitosan or
fibronectin coated chitosan (chitosan-FN) films on day 2. A: F-actin (red)/DAPI (blue) staining, (B) Cell spreading area, (C) Vinculin focal adhesion expression and representative
immunofluorescence images of cardiomyocytes for vinculin focal adhesion. Arrows indicate the expression of vinculin focal adhesion. 400� original magnification. Data are
expressed as means�SD. �P< 0.01 (chitosan vs. chitosan-FN).
6 Biotechnology and Bioengineering, Vol. xxx, No. xxx, 2012
co-culture which enabled the cardiomyocytes to contract ina tissue-like synchronized manner (Supplementary Video1). The cardiomyocytes co-cultured with endothelial cells(Fig. 5C and F) demonstrated a spherical morphology withlower levels of SM-actin and Cx43 expression than those inthe fibroblast co-culture as well as isolated contractions(Supplementary Video 2).
To assess the chitosan nanofibers’ potential as acardiac tissue engineering scaffold, the same cardiomyocytemono-culture and co-culture studies were performed onthe 3-D chitosan nanofibers. The cardiomyocyte mono-culture (Fig. 6A and D) and cardiomyocytes–endothelialcell co-culture (Fig. 6C and F) did not have any visibleSM-actin or Cx43 expression. The cardiomyocyte-fibroblastco-culture resulted in elongated networks of contractingcardiomyocytes (Supplementary Videos 3 and 4) withthe highest expression of SM-actin and Cx43 (Fig. 6Band E).
Intracellular Calcium Ion Flux Analysis
The intracellular calcium ion staining was used to monitorcalcium ion flux of beating cardiomyocytes. Figure 7 showsthe pseudo color time-frame images of the 3-D
cardiomyocyte co-culture system. The co-culture systemwith fibroblasts demonstrated elongated cardiomyocytemorphology and a rhythmic and frequent change inintracellular calcium ion concentrations (SupplementaryVideo 5). The image analysis of the calcium flux showed afrequency of 17� 3 contractions per minute. In contrast, the2-D and 3-D mono-cultures of cardiomyocytes lost theirfunctionality and their polarized morphology after 4 days ofculture (data not shown).
Discussion
The aim of this study was to fabricate 3-D chitosannanofiber scaffolds using an electrospinning technique andto test for the first time the feasibility of 3-D chitosannanofibers as scaffolds for cardiac tissue engineeringapplications. The results demonstrate that the chitosannanofibers retained their cylindrical morphology in long-term cell cultures and exhibited good cellular attachmentand spreading in the presence of adhesion molecule,fibronectin. In both 2-D and 3-D cultures on the chitosanscaffolds, cardiomyocyte-fibroblast co-cultures resulted inpolarized cardiomyocyte morphology with high levels ofSM-actin and Cx43 expression over long-term culture
Figure 4. A–C: Live/dead cell staining of 3T3-J2 fibroblasts seeded on fibronectin-coated chitosan (chitosan-FN) nanofiber scaffolds after 4 days of culture. The fibroblasts
were stained with calcein-AM (green) for live cells and ethidium homodimer (red) for dead cells. 200� original magnification. D–F: SEM images of cardiomyocytes, fibroblasts, and
endothelial cells cultured on chitosan-FN nanofiber scaffolds after 3 weeks of culture. Solid arrows indicate cell morphology on the scaffolds; dashed arrows indicate ECM-like
material secreted by the cells.
Hussain et al.: Functional 3-D Cardiac Co-Culture Model 7
Biotechnology and Bioengineering
periods. In addition, the fibroblast co-cultures demonstrat-ed synchronized contractions involving large tissue-likecellular networks, indicating the maintenance of long-termand stable function of cardiomyocytes gap junctions. To ourknowledge, the results of the present study provided the firstevidence that 3-D chitosan nanofibers can be used as apotential scaffold that can retain cardiomyocyte morpholo-gy and function. Further details of the chitosan nanofiberscaffolds on their physical, chemical, and thermal propertieswill be reported separately.
Cardiac fibroblasts are the most abundant non-cardio-myocyte cells in the mature heart. Their functions includedeposition of the extracellular matrix (ECM), paracrinesignaling, and propagation of the electrical stimuli. In thisstudy, we used murine 3T3-J2 fibroblasts cell line (derivedfrom Swiss mouse embryo by Howard Green) for cardiac co-cultures because of their easy access, propagation, and highinduction of parenchymal cell functions (e.g., primaryhepatocytes; Bhatia et al., 1999; Khetani et al., 2004). It wasreported that co-cultures of primary parenchymal cells (e.g.,cardiomyocytes, hepatocytes) with non-parenchymal stro-mal cells are able to maintain their phenotype and functionfor a longer period of time, depending on co-culture cell
type (Cho et al., 2008a; Driesen et al., 2005; Goulet et al.,1988; Narmoneva et al., 2004), but the effects of co-culturesare not species-specific (Clement et al., 1984). Potentialmechanisms include gap junctional communications onheterotypic cell–cell contacts (Gaudesius et al., 2003; Kizanaet al., 2006; Rook et al., 1992) and production of theextracellular matrices (Guillouzo et al., 1984; Loreal et al.,1993). However, details of the regulatory mechanism on therole of cell-cell interactions in co-cultures are still unknown.The key feature of our approach is the co-culturing ofcardiomyocytes with either fibroblasts or endothelial cellswithin a 3-D scaffold for long-term functionality of thecardiomyocytes. SM-actin expression was solely found incardiomyocytes and was expressed the most in the fibroblastco-cultures. Cx43 is the gap junction protein that is mainlyfound in ventricular cardiomyocytes (van Veen et al., 2001).The Cx43mediates fibroblasts heterogeneous coupling, suchas between cardiomyocytes and fibroblasts (Snider et al.,2009). These gap junctions with fibroblasts are known topropagate electrical stimuli for 100mm (van Veen et al.,2001). The results from both the 2-D and 3-D culturesindicate that fibroblast co-cultures resulted in high levels ofSM-actin and Cx43 expression, suggesting fibroblasts are
Figure 5. Morphology and phenotypic characteristics of cardiomyocytes on 2D Chitosan-FN film. A,D: Cardiomyocytes cultured alone, (B,E) cardiomyocytes cocultured with
3T3-J2 fibroblasts, and (C,F) cardiomyocytes co-cultured with microvascular endothelial cells after 7 days of culture. Cardiomyocytes were immunostained for a-sarcomeric actin
(SA-actin) and connexin-43 (Cx43) gap junction expression. Neonatal cardiomyocytes (CM), 3T3-J2 fibroblasts (FB), microvascular endothelial cells (EC). 200� original
magnification.
8 Biotechnology and Bioengineering, Vol. xxx, No. xxx, 2012
essential in maintaining cardiomyocytes viability andfunction in vitro. Optimal cardiac tissue for therapeuticapplications should involve a mixture of cardiac cells thatmore closely resemble native tissue but the optimal cell typeto be used still remains uncertain. Thus, in order to mimiccardiac structure and function in vitro, a basic understand-ing of the role of each of these interactions will be essential toprogress in the field.
Another advantage of our system is the use of electrospunchitosan to create nano tomicro-sized fibers that reproducesthe spatial dimensionality of the fibrous component of theECM. The fact that cardiomyocytes are able to survive andcontract on chitosan nanofibers mats has never beenreported before, as per our knowledge. It is projected thatthese mats can be layered on top of each other to create aprevascularized thick tissue-like structure composed ofcardiomyocytes, fibroblasts, and endothelial cells (Sasagawaet al., 2010). The fibroblasts will enhance the electricalsynchronization of the cardiomyocytes, while the endothe-lial cells have the potential to facilitate vascularization intothe graft to avoid diffusion limitation of oxygen andnutrients. Chitosan can interact electrostatically with cellssince cells carry an overall slightly negative surface chargeand chitosan’s free amine group can become protonated
allowing ionic interactions (Chatelet et al., 2001; Prasitsilpet al., 2000). Our data suggests that cells cultured onchitosan surfaces maintained rounded morphology withpoor cell adhesion. However, cells cultured on fibronectincoated chitosan surfaces exhibited a typical elongated shapewith improved cell adhesion. Fibronectin is a large ECMglycoprotein which facilitates cell adhesion and spreadingvia a5b1 and avb3 integrin receptors in cells (Amaral et al.,2009). The integrins recognize and interact with RGD celladhesion domains initiating cell signaling pathways thatcontrol cell survival, proliferation, differentiation, andremodeling of the ECM (Garcıa and Boettiger, 1999). Theamine groups present in chitosan are engaged in fibronectinadsorption (Amaral et al., 2005). The functional activity offibronectin is conserved because there is minimum proteinunfolding conserving the cell adhesion sites (Garcıa andBoettiger, 1999).
In conclusion, results of this study demonstrate thatchitosan nanofibers can be used as scaffolds for thedevelopment of 3-D cardiac tissue constructs that moreclosely resemble native heart tissue. Our results suggest thatcardiac co-culture model is a promising system for themaintenance of long-term survival and function ofcardiomyocytes. The engineered 3-D cardiac co-culture
Figure 6. Morphology and phenotypic characteristics of cardiomyocytes in 3-D Chitosan-FN nanofiber scaffolds. A,D: Cardiomyocytes cultured alone, (B,E) cardiomyocytes
co-cultured with 3T3-J2 fibroblasts, and (C,F) cardiomyocytes cocultured with microvascular endothelial cells after 19 days of culture. Cardiomyocytes were immunostained for a-
sarcomeric actin (SA-actin) and connexin-43 (Cx43) gap junction expression. Neonatal cardiomyocytes (CM), 3T3-J2 fibroblasts (FB), microvascular endothelial cells (EC). 200�original magnification.
Hussain et al.: Functional 3-D Cardiac Co-Culture Model 9
Biotechnology and Bioengineering
model using chitosan nanofiber scaffolds will be useful forthe design and improvement of engineered tissues for therepair of myocardial infarcts, tissue engineering applica-tions, and drug testing.
We thank Dr. Junichi Sadoshima from University of Medicine and
Dentistry of New Jersey (UMDNJ)-New Jersey Medical School for the
technical support on isolating primary cardiomyocytes from neonatal
rat heart. We thank Dr. Haesun Kim from Rutgers University for
providing neonatal rat hearts. This study was partially supported by
theWallace H. Coulter Foundation and the startup funding fromNew
Jersey Institute of Technology.
References
Amaral I, Lamghari M, Sousa S, Sampaio P, Barbosa M. 2005. Rat bone
marrow stromal cell osteogenic differentiation and fibronectin adsorp-
tion on chitosan membranes: The effect of the degree of acetylation.
J Biomed Mater Res A 75(2):387.
Amaral I, Unger R, Fuchs S, Mendonca A, Sousa S, Barbosa M, P go A,
Kirkpatrick C. 2009. Fibronectin-mediated endothelialisation of
chitosan porous matrices. Biomaterials 30(29):5465–5475.
Aoki H, Izumo S, Sadoshima J. 1998. Angiotensin II activates RhoA in
cardiac myocytes: A critical role of RhoA in angiotensin II-induced
premyofibril formation. Circ Res 82(6):666–676.
Baig M, Mahon N, McKenna W, Caforio A, Bonow R, Francis G, Gheor-
ghiade M. 1998. The pathophysiology of advanced heart failure.
Am Heart J 135(6):S216–S230.
Banerjee I, Yekkala K, Borg TK, Baudino TA. 2006. Dynamic interactions
between myocytes, fibroblasts, and extracellular matrix. Ann N Y Acad
Sci 1080:76–84.
Bhatia SN, Balis UJ, Yarmush ML, Toner M. 1999. Effect of cell–cell
interactions in preservation of cellular phenotype: Cocultivation of
hepatocytes and nonparenchymal cells. Faseb J 13(14):1883–1900.
Chatelet C, Damour O, Domard A. 2001. Influence of the degree of
acetylation on some biological properties of chitosan films. Biomater-
ials 22(3):261–268.
Chen Q, Harding S, Ali N, Lyon A, Boccaccini A. 2008. Biomaterials in
cardiac tissue engineering: Ten years of research survey. Mater Sci Eng
R Rep 59(1–6):1–37.
Cho CH, Berthiaume F, Tilles AW, Yarmush ML. 2008a. A new technique
for primary hepatocyte expansion in vitro. Biotechnol Bioeng
101(2):345–356.
Cho CH, Eliason JF, Matthew HW. 2008b. Application of porous glycos-
aminoglycan-based scaffolds for expansion of human cord blood stem
cells in perfusion culture. J Biomed Mater Res A 86(1):98–107.
Clement B, Guguen-Guillouzo C, Campion JP, Glaise D, Bourel M,
Guillouzo A. 1984. Long-term co-cultures of adult human hepatocytes
with rat liver epithelial cells: Modulation of albumin secretion
and accumulation of extracellular material. Hepatology 4(3):373–
380.
Desai K, Kit K, Li J, Zivanovic S. 2008.Morphological and surface properties
of electrospun chitosan nanofibers. Biomacromolecules 9(3):1000–
1006.
Doughty M, Whalley B, Gamble M, MacMahon P, Sharpe M. 1997. Left
ventricular remodeling with carvedilol in patients with congestive heart
failure due to ischemic heart disease. J Am Coll Cardiol 29(5):1060–
1066.
Driesen RB, Dispersyn GD, Verheyen FK, van den Eijnde SM, Hofstra L,
Thone F, Dijkstra P, Debie W, Borgers M, Ramaekers FC. 2005. Partial
cell fusion: A newly recognized type of communication between
dedifferentiating cardiomyocytes and fibroblasts. Cardiovasc Res
68(1):37–46.
Garcıa A, Boettiger D. 1999. Integrin–fibronectin interactions at the
cell-material interface: Initial integrin binding and signaling. Bioma-
terials 20(23–24):2427–2433.
Gaudesius G, Miragoli M, Thomas S, Rohr S. 2003. Coupling of cardiac
electrical activity over extended distances by fibroblasts of cardiac
origin. Circ Res 93(5):421.
Goulet F, Normand C,Morin O. 1988. Cellular interactions promote tissue-
specific function, biomatrix deposition and junctional communication
of primary cultured hepatocytes. Hepatology 8(5):1010–1018.
Guillouzo A, Delers F, Clement B, Bernard N, Engler R. 1984. Long term
production of acute-phase proteins by adult rat hepatocytes co-
cultured with another liver cell type in serum-free medium. Biochem
Biophys Res Commun 120(2):311–317.
Haugh MG, Murphy CM, McKiernan RC, Altenbuchner C, O’Brien FJ.
2011. Crosslinking and mechanical properties significantly influence
cell attachment, proliferation, and migration within collagen glycos-
aminoglycan scaffolds. Tissue Eng Part A 17(9–10):1201–1208.
Ingber DE. 1997. Tensegrity: The architectural basis of cellular mechan-
otransduction. Annu Rev Physiol 59:575–599.
Ingber DE. 2003. Mechanosensation through integrins: Cells act locally but
think globally. Proc Natl Acad Sci USA 100(4):1472–1474.
Jayakumar R, Prabaharan M, Nair SV, Tamura H. 2010. Novel chitin and
chitosan nanofibers in biomedical applications. Biotechnol Adv
28(1):142–150.
Khetani SR, Szulgit G, Del Rio JA, Barlow C, Bhatia SN. 2004. Exploring
interactions between rat hepatocytes and nonparenchymal cells using
gene expression profiling. Hepatology 40(3):545–554.
Kim I, Seo S, Moon H, YooM, Park I, Kim B, Cho C. 2008. Chitosan and its
derivatives for tissue engineering applications. Biotechnol Adv 26(1):
1–21.
Figure 7. Intracellular calcium ion flux of beating cardiomyocytes in 3-D co-
culture of cardiomyocytes and fibroblasts after 7 days of culture. A: Pseudo-color
images of calcium ion changes in fluorescence intensity, monitored at 0.7-s time
intervals with fluo-4, AM indicator. 200� original magnification. B: Frequency of
intracellular calcium ion flux in the 3-D co-culture. The fluorescence images of calcium
ion flux were pseudo-colored, with red representing high calcium ion concentrations
and blue representing low calcium ion concentrations.
10 Biotechnology and Bioengineering, Vol. xxx, No. xxx, 2012
Kizana E, Ginn SL, Smyth CM, Boyd A, Thomas SP, Allen DG, Ross DL,
Alexander IE. 2006. Fibroblasts modulate cardiomyocyte excitability:
Implications for cardiac gene therapy. Gene Ther 13(22):1611–1615.
Koh G, Klug M, Soonpaa M, Field L. 1993a. Differentiation and long-term
survival of C2C12 myoblast grafts in heart. J Clin Invest 92(3):1548.
Koh G, Soonpaa M, Klug M, Field L. 1993b. Long-term survival of AT-1
cardiomyocyte grafts in syngeneic myocardium. Am J Physiol Heart
Circ Physiol 264(5):H1727.
Kumar R, Majeti N. 2000. A review of chitin and chitosan applications� 1.React Funct Polym 46(1):1–27.
LaFramboise WA, Scalise D, Stoodley P, Graner SR, Guthrie RD, Magovern
JA, Becich MJ. 2007. Cardiac fibroblasts influence cardiomyocyte
phenotype in vitro. Am J Physiol Cell Physiol 292(5):C1799–C1808.
Loreal O, Levavasseur F, Fromaget C, Gros D, Guillouzo A, Clement B.
1993. Cooperation of Ito cells and hepatocytes in the deposition of an
extracellular matrix in vitro. Am J Pathol 143(2):538–544.
Madihally SV, Matthew HW. 1999. Porous chitosan scaffolds for tissue
engineering. Biomaterials 20(12):1133–1142.
Martinez C, Tzur A, Hrachian H, Zebede J, Lamas G. 2007. Pacemakers and
defibrillators: Recent and ongoing studies that impact the elderly. Am J
Geriatr Cardiol 15(2):82–87.
Matthew HW, Cho CH, Birla RK, Khanna HJ. 2003. Polysaccharides as
biomaterials. Biodegradable polymers, Vol. 2. The polymers in bio-
technology & medicine (PBM) series 2. London: Citus Books. p 35–68.
Narmoneva D, Vukmirovic R, Davis M, Kamm R, Lee R. 2004. Endothelial
cells promote cardiac myocyte survival and spatial reorganization:
Implications for cardiac regeneration. Circulation 110(8):962.
Ohkawa K, Cha D, Kim H, Nishida A, Yamamoto H. 2004. Electrospinning
of chitosan. Macromolecul Rapid Commun 25(18):1600–1605.
Packer M. 2002. The impossible task of developing a new treatment for
heart failure. J Card Fail 8(4):193–196.
Pham QP, Sharma U, Mikos AG. 2006. Electrospinning of polymeric
nanofibers for tissue engineering applications: A review. Tissue Eng
12(5):1197–1211.
Prasitsilp M, Jenwithisuk R, Kongsuwan K, Damrongchai N, Watts P. 2000.
Cellular responses to chitosan in vitro: The importance of deacetyla-
tion. J Mater Sci Mater Med 11(12):773–778.
Rockwood DN, Akins RE, Jr., Parrag IC, Woodhouse KA, Rabolt JF. 2008.
Culture on electrospun polyurethane scaffolds decreases atrial natri-
uretic peptide expression by cardiomyocytes in vitro. Biomaterials
29(36):4783–4791.
Rook MB, van Ginneken AC, de Jonge B, el Aoumari A, Gros D, Jongsma
HJ. 1992. Differences in gap junction channels between cardiac myo-
cytes, fibroblasts, and heterologous pairs. Am J Physiol 263(5 Pt 1):
C959–C977.
Sasagawa T, Shimizu T, Sekiya S, Haraguchi Y, Yamato M, Sawa Y, Okano
T. 2010. Design of prevascularized three-dimensional cell-dense tissues
using a cell sheet stacking manipulation technology. Biomaterials
31(7):1646–1654.
Schussler O, Chachques JC, Mesana TG, Suuronen EJ, Lecarpentier Y, Ruel
M. 2010. 3-dimensional structures to enhance cell therapy and engineer
contractile tissue. Asian Cardiovasc Thorac Ann 18(2):188–198.
Sharpe N, Smith H, Murphy J, Greaves S, Hart H, Gamble G. 1991. Early
prevention of left ventricular dysfunction after myocardial infarction
with angiotensin-converting-enzyme inhibition. Lancet 337(8746):
872–876.
ShinM, Ishii O, Sueda T, Vacanti JP. 2004. Contractile cardiac grafts using a
novel nanofibrous mesh. Biomaterials 25(17):3717–3723.
Snider P, Standley K, Wang J, Azhar M, Doetschman T, Conway S. 2009.
Origin of cardiac fibroblasts and the role of periostin. Circul Res
105(10):934.
Soonpaa M, Koh G, Klug M, Field L. 1994. Formation of nascent interca-
lated disks between grafted fetal cardiomyocytes and host myocardium.
Science 264(5155):98.
Stehlik J, Edwards LB, Kucheryavaya AY, Benden C, Christie JD, Dobbels F,
Kirk R, Rahmel AO, Hertz MI. 2011. The Registry of the International
Society for Heart and Lung Transplantation: Twenty-eighth Adult
Heart Transplant Report—2011. J Heart Lung Transplant 30(10):
1078–1094.
Stupack DG. 2005. Integrins as a distinct subtype of dependence receptors.
Cell Death Differ 12(8):1021–1030.
Teo WE, Ramakrishna S. 2006. A review on electrospinning design and
nanofibre assemblies. Nanotechnology 17(14):R89–R106.
van Veen T, van Rijen H, Opthof T. 2001. Cardiac gap junction channels:
Modulation of expression and channel properties. Cardiovasc Res
51(2):217.
Venugopal JR, Prabhakaran MP, Mukherjee S, Ravichandran R, Dan K,
Ramakrishna S. 2012. Biomaterial strategies for alleviation of myo-
cardial infarction. J R Soc Interface 9(66):1–19.
Vliegen H, Van der Laarse A, Cornelisse C, Eulderink F. 1991. Myocardial
changes in pressure overload-induced left ventricular hypertrophy: A
study on tissue composition, polyploidization and multinucleation.
Eur Heart J 12(4):488.
Zong X, Bien H, Chung CY, Yin L, Fang D, Hsiao BS, Chu B, Entcheva E.
2005. Electrospun fine-textured scaffolds for heart tissue constructs.
Biomaterials 26(26):5330–5338.
Hussain et al.: Functional 3-D Cardiac Co-Culture Model 11
Biotechnology and Bioengineering