A cellularized biomaterial model of cardiac fibroblasts for evaluation of fibroblast induced changes in stiffness

Chronic heart failure (CHF) is a life-altering long-term condition that contributes a substantial burden to our healthcare system. It is caused by a maladaptive remodeling of the heart mediated through fibroblast synthesis, degradation, and modification of extracellular matrix (ECM). It is currently managed through pharmacologic intervention or medical device treatment, but can be reversed only through heart transplantation. Cell therapy is a new approach to treating CHF that promises to prevent and potentially reverse cardiac remodeling through interaction with cardiac fibroblasts. Adherent bone marrow derived stem cells (MSC) are one of the most promising candidates for use in cell therapies. The major challenge hindering standard clinical application of MSC therapy is limited understanding of how MSC interact with heart cells to reverse remodeling. Numerous techniques are available to harvest, isolate, and modify MSC, however these techniques are believed to influence the efficacy of the treatment. Current techniques for evaluating efficacy of MSC treatments are either prohibitively difficult or significantly limited in their ability to assess functional changes. Establishing an in vitro platform for evaluating the influence of MSC coculture on performance characteristics of cardiac fibroblasts is a logical and efficient step prior to successful clinical implementation of MSC therapy. The objective of this research was to develop a threedimensional (3D) tissue model that allows investigation of the underlying mechanism responsible for MSC mediated cardiac regeneration. The three phases of this work included: (1) development of a biomaterial substrate capable of sustaining fibroblast attachment, proliferation, and alignment, (2) development of a culture platform and seeding techniques capable of providing sufficient mass transport to sustain a relatively thick 3D scaffold populated with both fibroblasts and MSC (3) application of the substrate and culture platform to evaluate changes in mechanical properties and cell distribution resulting from MSC coculture with cardiac fibroblasts.

A CELLULARIZED BIOMATERIAL MODEL OF CARDIAC FIBROBLASTS FOR EVALUATION OF FIBROBLAST INDUCED CHANGES IN STIFFNESS by James Paul Kennedy A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Bioengineering University of Utah May 2013 Copyright © James Paul Kennedy 2013 All Rights Reserved The Uni v e r s i t y of Utah Graduat e School STATEMENT OF DISSERTATION APPROVAL The dissertation of James Paul Kennedy has been approved by the following supervisory committee members: Robert Hitchcock Vladimir Hlady Greg Burns Patrick Tresco David Grainger Chair Member Member Member Member 3/12/13 Date Approved 3/12/13 Date Approved 3/13/13 Date Approved 3/12/13 Date Approved 3/12/13 Date Approved and by Patrick Tresco the Department of Bioengineering Chair of and by Donna M. White, Interim Dean of The Graduate School. ABSTRACT Chronic heart failure (CHF) is a life-altering long-term condition that contributes a substantial burden to our healthcare system. It is caused by a maladaptive remodeling of the heart mediated through fibroblast synthesis, degradation, and modification of extracellular matrix (ECM). It is currently managed through pharmacologic intervention or medical device treatment, but can be reversed only through heart transplantation. Cell therapy is a new approach to treating CHF that promises to prevent and potentially reverse cardiac remodeling through interaction with cardiac fibroblasts. Adherent bone marrow derived stem cells (MSC) are one of the most promising candidates for use in cell therapies. The major challenge hindering standard clinical application of MSC therapy is limited understanding of how MSC interact with heart cells to reverse remodeling. Numerous techniques are available to harvest, isolate, and modify MSC, however these techniques are believed to influence the efficacy of the treatment. Current techniques for evaluating efficacy of MSC treatments are either prohibitively difficult or significantly limited in their ability to assess functional changes. Establishing an in vitro platform for evaluating the influence of MSC coculture on performance characteristics of cardiac fibroblasts is a logical and efficient step prior to successful clinical implementation of MSC therapy. The objective of this research was to develop a three­dimensional (3D) tissue model that allows investigation of the underlying mechanism responsible for MSC mediated cardiac regeneration. The three phases of this work included: (1) development of a biomaterial substrate capable of sustaining fibroblast attachment, proliferation, and alignment, (2) development of a culture platform and seeding techniques capable of providing sufficient mass transport to sustain a relatively thick 3D scaffold populated with both fibroblasts and MSC (3) application of the substrate and culture platform to evaluate changes in mechanical properties and cell distribution resulting from MSC coculture with cardiac fibroblasts. iv To my wife, Jill TABLE OF CONTENTS ABSTRACT........................................................................................................................... iii LIST OF FIGURES............................................................................................................. viii ACKNOWLEDGEMENTS.....................................................................................................x Chapters 1 INTRODUCTION.......................................................................................................1 1.1 Introduction...................................................................................................................1 1.2 The Cardiac Fibroblast.................................................................................................4 1.3 Cardiac Disease and Heart Failure............................................................................14 1.4 Cardiac Cell Therapy.................................................................................................19 1.5 Three-Dimensional Culture Models........................................................................ 22 1.6 Summary and Contributions..................................................................................... 33 1.7 References...................................................................................................................38 2 DESIGN OF SEMIAUTOMATEDTOOLING FOR FABRICATION OF POROUS POLYURETHANE FOAMS.................................................................. 49 2.1 Introduction................................................................................................................ 49 2.2 Spray Parameters.......................................................................................................50 2.3 Tooling Applications.................................................................................................53 3 THE MECHANICALLY ENHANCED PHASE SEPARATION OF SPRAYED POLYURETHANE SCAFFOLDS AND THEIR EFFECT ON THE ALIGNMENT OF FIBROBLASTS...............................................................64 3.1 Introducti on................................................................................................................ 65 3.2 Materials and Methods...............................................................................................66 3.3 Results........................................................................................................................67 3.4 Discussion ...................................................................................................................68 3.5 Conclusion..................................................................................................................70 3.6 Acknowledgements....................................................................................................70 3.7 Appendix.....................................................................................................................70 3.8 References...................................................................................................................70 4 ENGINEERED CHANNELS ENHANCE CELLULAR DENSITY IN PERFUSED SCAFFOLDS....................................................................................... 72 4.1 Introduction.................................................................................................................73 4.2 Methods.......................................................................................................................74 4.3 Results........................................................................................................................76 4.4 Discussion ...................................................................................................................78 4.5 Conclusion..................................................................................................................80 4.6 Acknowledgments......................................................................................................80 4.7 Appendix A .................................................................................................................80 4.8 References...................................................................................................................80 5 MSC COCULTURE INCREASES TISSUE STIFFNESS IN THREE­DIMENSIONAL CARDIAC FIBROBLAST MODEL.........................................82 5.1 Abstract.......................................................................................................................82 5.2 Introduction.................................................................................................................83 5.3 Materials and Methods...............................................................................................87 5.4 Results........................................................................................................................97 5.5 Discussion................................................................................................................ 100 5.6 Acknowledgements..................................................................................................105 5.7 References................................................................................................................ 105 6 CONCLUSION AND FUTURE DIRECTIONS.................................................. 121 6.1 Summary and Conclusions..................................................................................... 121 6.2 Applications and Future Directions....................................................................... 124 6.3 References................................................................................................................ 129 APPENDIX: ENGINEERING PRINTS OF PERFUSION BIOREACTOR..................134 vii LIST OF FIGURES 2.1 Spray assembly.......................................................................................................... 55 2.2 Nozzle assembly ........................................................................................................ 56 2.3 Spray surface assembly..............................................................................................57 2.4 Waste reservoir .......................................................................................................... 58 2.5 Spray plate ..................................................................................................................59 2.6 Spray substrate ........................................................................................................... 60 2.7 Scaffold frame............................................................................................................ 61 2.8 Mask........................................................................................................................... 62 2.9 Alignment pin and T-nut........................................................................................... 63 3.1 Representative SEM images of scaffolds................................................................67 3.2 Effective modulus of elasticity of scaffolds.............................................................68 3.3 Porosity of scaffolds..................................................................................................69 3.4 Confocal image of aligned fibroblasts..................................................................... 69 3.5 Surface plot of 2D FFT and intensity angle plot..................................................... 69 3.6 Orientation index of porous scaffolds...................................................................... 69 3.7 Cross sectional SEM of laminated scaffold.............................................................70 4.1 Polyester frame and lamination procedure..............................................................74 4.2 Fabrication of patterned substrates...........................................................................75 4.3 Custom molded silicone adapter...............................................................................75 4.4 Culture chamber and schematic of bioreactor.........................................................76 4.5 SEM imaging of scaffold sheets...............................................................................77 4.6 Hydraulic permeability testing..................................................................................77 4.7 Cross-sectional images of constructs....................................................................... 78 4.8 Spatial image analysis of cross sections.................................................................. 79 4.9 Cell density................................................................................................................ 79 5.1 Culture platform schematic..................................................................................... 109 5.2 Image processing method....................................................................................... 110 5.3 Initial phenotyping of cardiac cell population.......................................................111 5.4 3T3 Fb density with different seeding velocities and substrate coatings............ 112 5.5 Cardiac Fb uniformity with different seeding velocities and durations...............113 5.6 Mean cardiac Fb density with different seeding velocities and durations.......... 114 5.7 Treatment induced changes in stiffness................................................................. 115 5.8 Tensile modulus of MSC coculture constructs......................................................116 5.9 Cell characterization during culture....................................................................... 117 5.10 Collagen labeling during culture.............................................................................118 ix ACKNOWLEDGEMENTS I gratefully acknowledge my advisor and mentor, Dr. Robert Hitchcock, who provided me with the guidance, insight and support to complete this work. I thank my advisory committee members for their guidance, insight, and persistent efforts to enhance the quality of my education: Drs. David Grainger, Greg Burns, Patrick Tresco, and Vladimir Hlady. I acknowledge my fellow graduate students Richard Lasher, Monir Parikh, Tanner Coleman, Kylee North, and Chao Huang, for their friendship and technical support during this work. I humbly acknowledge the essential contributions of the undergraduate students who worked with me: Sean McCandless, Asad Rauf, Laura Williams, Jason Hansen, Jenn Hillam, Molly Person, Jessica Ashmead, Hendrik Stegall, John Lackey, Alexis Johnson, Joseph Goodrick, Martin Jensen, Ryan Russon, Mark Sedlacek, Karan Mehta, and Shaswat Chapagain. Finally, I thank my family, especially my wife Jill and my parents Jim and DeAnn for their continuous love, optimisim, and encouragement. CHAPTER 1 INTRODUCTION 1.1 Introduction Chronic heart failure is a significant problem facing the United States. In 2006, there were 5.8 million Americans living with heart failure [1]. The associated personal and economic burdens are enormous. The most common acute cause of heart failure is myocardial infarction (MI) [1]. Infarct-induced ischemia and hypoxia lead to myocyte depletion in the infarct zone and border regions, and local activation of inflammatory and fibroblast cell types. These two physiologic responses to MI often result in a self-sustaining series of detrimental structural changes to the heart. Typically termed maladaptive remodeling, this phenomenon is a biomechanical process that is attributed to changes in passive stiffness of the ventricular wall. More specifically, maladaptive remodeling is a consequence of changes in ventricular wall stiffness during two distinct phases of healing after MI. During the early post-MI phase, the stiffness of the infarct area is dramatically decreased. Rupture, dyskinesis, and infarct expansion arise from insufficient stiffness of the ventricular wall and lead to infarct expansion, dilated cardiomyopathy, decreased ejection fraction, and systolic heart failure. On the other hand, during the later post-MI healing phase, activated myofibroblasts (myoFb) increase matrix production and organization to stiffen the infarcted region. In many cases, increased matrix accumulation in regions remote to the infarct accompanies local matrix deposition. Increased passive stiffness in remote regions prevent adequate filling of the ventricle during diastole, and leads to diastolic heart failure. Both cases, systolic and diastolic heart failure, are results of pathologic changes in passive stiffness. Prevention of maladaptive remodeling and regeneration of myocardial function after MI relies on a delicate balance of passive stiffness and therefore is profoundly influenced through the matrix regulation activity of cardiac fibroblasts (CFb). Cardiac cell therapy appears to be a promising new treatment to prevent heart failure after MI [2]. The aim of this therapy is to enhance in situ regeneration of cardiac function. Modulation of the CFb response during post-MI healing is one potentially powerful approach that may lead to functional myocardial regeneration. Despite significant findings initially, cell therapy has been delayed in its transition to a standard treatment following MI. Delivery of bone marrow cells including mesenchymal stem cells (MSC) has been shown in animal models to increase positive clinical endpoints such as ejection fraction and survival [3, 4]. However, in multisubject human clinical trials, cell therapy has not significantly prevented or reversed cardiac remodeling [5]. To realize the potential of cardiac cell therapy, we need to elucidate the mechanism behind the apparent MSC-induced cardiac regeneration observed in animal models. One pathway to increased understanding of MSC-heart interactions is the development of better in vitro model systems. Two-dimensional (2D) coculture of MSC and CFb relies on subjective interpretation of expression of ECM components and regulators to draw speculative conclusions about changes in stiffness. The most recent 2 three-dimensional (3D) engineered tissue models allow investigation of survival, integration, differentiation, and coupling between stem cells and cardiac cells [6-8]. These models, however, are aimed at understanding how stem cells can help recover active contraction of cardiac tissue rather than modulating the passive stiffness. Prior to the work described in this dissertation, no model capable of directly coupling changes in stiffness with stem cell therapy has been described. The overarching purpose of the work described in this dissertation was to investigate the effects of MSC coculture with CFb mediated changes in stiffness. The objective of this work was accomplished through three distinct research phases: 1. Develop and characterize a 3D substrate capable of supporting and influencing fibroblast attachment and organization. 2. Design and validate a perfusion culture platform that supports a uniform fibroblast distribution throughout a 3D substrate of scalable thickness. 3. Apply the substrate and culture platform to evaluate MSC therapy on CFb-mediated changes in stiffness in vitro. To provide a conceptual framework to this work, the physiologic role of the cardiac fibroblast, the impetus and mechanics of heart failure, cardiac cell therapy, and state-of-the-art cell culture and tissue engineering models are reviewed below. 1.2. The Cardiac Fibroblast 1.2.1 The Fibroblast in Cardiac Physiology The human heart is a four chambered pump responsible for ensuring adequate metabolic exchange in our body. The primary function of the cardiovascular system is to 3 convectively transport blood (and the metabolites it contains) throughout the body. The role of the heart in this system is to generate a pressure gradient that will move blood through the vasculature. Pulsatile flow is generated through a biphasic process consisting of filling (diastole) and emptying (systole) of the heart. During diastole, the pressure of the blood entering the ventricle is stored as strain energy in the ventricular wall as the ventricle stretches to accommodate an increased blood volume. Systole begins when the heart contracts. The active mechanical contraction of the heart can be described as a sudden and significant increase in the stiffness of the heart wall. The balance between the stress and strain of the tissue during diastole is no longer stable. The effectively stiffer tissue causes the ventricular muscle to contract to balance the stress with a proportionally smaller strain. The contraction of the heart is responsible for expelling a fraction of the blood and driving it through the body. The ejection fraction is a function of the difference between passive mechanical properties, and active mechanical contraction. The majority of heart cells are either cardiac fibroblasts, or cardiac myocytes. Myocytes occupy the majority of the volume of the heart [9]. These cells are responsible for generation of active mechanical properties of the heart. Through an elegant excitation-contraction mechanism, electrical action potentials are converted to mechanical work - shortening of the cardiac myocyte. Tissue organization and specific timing converts small contractions of individual cells to macroscopic tissue contraction. The cardiac fibroblast is the other major cell type in the heart. Roughly 60-70% of all cells in the heart are fibroblasts [10]. These cells contribute to electrical, biochemical, and structural characteristics of cardiac muscle. Fibroblasts, unlike myocytes, are nonexcitable cells. They play a passive role in propagation of action potentials through 4 cardiac muscle by acting as obstacles to the orderly spread electrical excitation in the heart. Recent discoveries have also implicated fibroblasts in a more active role in cardiac electrophysiology through gap junction connections with each other and cardiac myocytes [11, 12]. CFb also maintain the structure and composition of cardiac extracellular matrix (ECM). Type I and III collagen account for 90% of the protein content of the cardiac ECM [13]. The ECM helps mechanically couple myocytes, and transfers the force generated by individual myocytes throughout the organ [14]. Moreover, the ECM composition and organization provides the passive stiffness and toughness necessary during systole. Coordinated action of both myocytes and fibroblasts is necessary to maintain cardiac function. The following sections describe in further detail the role of the cardiac fibroblast in maintaining the structure and function of the cardiac ECM. 1.2.2 The Myofibroblast In 1971, Gabbiani and colleagues observed what they termed a "modified fibroblast" and suggested that these cells have an important role in wound contraction [15]. The distinguishing features of these cells included increased expression of an intracellular fibrillar system, nuclear deformations (due to contraction of the cell), and increased intercellular connections [15]. This modified type of the cardiac fibroblast has come to be known as the myofibroblast. The most discernible structural feature of the myofibroblast when compared to the fibroblast is the presence of smooth muscle actin and myosin. The distinction between CFb and the myofibroblast is significant in terms of activity. Cardiac fibroblasts have been referred to as "silent" under normal circumstances 5 in the heart. The myofibroblast phenotype, on the other hand is typically associated with the active matrix synthesis, degradation, and modification activities normally associated with fibroblasts in general. The transition to a myofibroblastic phenotype occurs in response in vivo to mechanical stimulation, hypoxia, and inflammatory cytokines [16]. In vitro, increased substrate stiffness has been shown to cause fibroblasts to assume a myofibroblastic phenotype [17]. Anseth and colleagues utilized a light sensitive hydrogel to demonstrate that myofibroblast differentiation is nonterminal, and can be reversed. Myofibroblasts on a relatively stiff hydrogel substrate were induced to deactivate by decreasing their substrate modulus [18]. Importantly, it has been hypothesized that increased stiffness in vivo due to fibrosis results in positive feedback, with more fibroblasts becoming activated and assuming the myofibroblast phenotype [19]. A specific definition of myofibroblasts as a different cell type has yet to be generally established. Indeed, fibroblasts are pleomorphic in the sense that they change their structure and function depending on external stimuli [10]. Therefore, in this work the myofibroblast will be considered a phenotypic subset of the fibroblast. In the section below, the fibroblast mediated mechanisms of ECM turnover are discussed. Elevated levels of ECM modification are typically associated with what is considered a myofibroblast phenotype. 1.2.3 Fibroblast Maintenance of the Extracellular Matrix One primary physiologic role of cardiac fibroblasts (CFb) is to regulate extracellular matrix composition and properties through a complex network of interactive 6 7 mechanisms. One of the important functions of the ECM is to maintain the passive stiffness and elasticity of cardiac muscle. Indeed it has been shown that ECM components, even in hypertrophic disease conditions, are almost entirely responsible for passive stiffness of the heart muscle [20]. CFb respond to chemical and mechanical cues that result in matrix synthesis, matrix degradation, and ECM modification. These tasks are accomplished through ECM component synthesis (primarily type I and type III collagen), expression of matrix metalloproteinases (MMP's) and protease inhibitors (TIMP), and expression of matrix cross-linking enzymes enzymes (e.g., lysyl oxidases). CFb are responsible for production of procollagen fibers that will eventually become type I and type III collagen, the primary dictators of passive cardiac stiffness [21]. CFb synthesis of collagen is modulated through growth factors (e.g., TGF-P), signal peptides (e.g., angiotensin II), and cytokines (e.g., interleukins, TNF-a) [13]. The most potent regulators of collagen synthesis in CFb appear to be TGF-P, the rennin-angiotensin- aldesterone system (RAAS) and P-adrenergic system [22]. CFb also respond to mechanical stress by immediately increasing collagen deposition [16]. In vitro, CFb have been shown to increase matrix production in response to cyclic strain [23]. Another key role of the cardiac fibroblast is expression of matrix metalloproteinases (MMP). MMP are a family of zinc-dependent proteins expressed throughout the body. This class of proteases is responsible for cleaving extracellular matrix (ECM) molecules. In the heart, MMP play a critical role in maintaining the balance of ECM synthesis and degradation, as well as influencing ECM organization [24]. MMP are typically classified according to their substrate specificity. Cardiac fibroblasts express MMP-1, MMP-2, MMP-3, MMP-9, and MMP-14. Collagenase-type MMPs (MMP-1) cleave fibrillar collagen (type I, II, III). Gelatinases (MMP-2 and MMP- 9) cleave gelatin, nonfibrillar collagen and fibronectin. Stromelysins (MMP-3), like gelatinases, can cleave gelatins, nonfibrillar collagens, and fibronectin; however, they also cleave laminin and MMP zymogens. This makes the stromelysin class of MMP important for regulation of other classes of MMP. The last class of MMP found in the heart are membrane bound proteases. MMP-14, a membrane bound MMP, has the same substrate specificity to stromelysins, plus fibrillar collagen. Cytokines (TNF-alpha, IL-1), growth factors (TGF-beta), signal molecules (angiotensin, endothelin), reactive oxygen species (ROS), and mechanical stimuli all affect MMP transcription. MMP-1, 3 and 9 expression is typically low, and is upregulated by proinflammatory cytokines such as TNF-a and IL-6. MMP-2 and MMP-9 are expressed constitutively, however their expression can be increased by proinflammatory cytokines or decreased by angiotensin II through the TGF-P pathway. MMP-14 expression, remarkably, is increased by TGF-P expression. MMP-1 cleaves fibrillar collagen (types I, II, and III) and MMP-1 expression has been shown to mitigate pathologic cardiac remodeling in transgenic mice [25]. Conversely, increased expression of MMP-2 has been associated with early (0-7 day) left ventricular rupture after infarction and late stage (4 week) dilatation in a murine MMP-2 knock-out model [26]. Increased expression of MMP-14 has also been shown to favor pathologic remodeling of the heart [27]. However, in a different murine model of heart failure, treatment with a proinflammatory nuclear protein 2 weeks after infarction increased the activity of MMP-2 and MMP-9. In apparent 8 contradiction to previous findings, increased activity of MMP-2 resulted in higher ejection fractions and lower dilatation 7 weeks after infarction [28]. Because of the critical role of MMP in maintaining healthy, homeostatic condition of the ECM, they are tightly regulated not only in expression, but in function through a family of four different MMP inhibitors (TIMP). TIMP-3 is one of the major inhibitors found in cardiac tissue; it inhibits the function of all cardiac fibroblast-produced MMP. TIMP-3 is produced by most cardiac cells including myocytes, vascular smooth muscle cells, and fibroblasts. In addition to its function as an MMP inhibitor, TIMP-3 has MMP-independent effects including modulating Fb differentiation to myoFb, decreasing angiogenesis, signaling smooth muscle cell apoptosis, and preventing cardiomyocyte proliferation [29, 30]. Another fibroblast-mediated mechanism of ECM regulation in the heart is lysyl oxidase (LOX) production. LOX and LOX-like (LOXL) proteins are found throughout the body, including in the heart [31]. In the heart, LOX is an extracellular enzyme produced by CFb responsible for forming cross-links in types I and III fibrillar collagen. Crosslinking is the final step in collagen synthesis, and results in collagen fibrils that are less soluble, more resistant to degradation, and much stiffer [32]. LOX expression is stimulated through hypoxia and TNF-a (mediated through TGF-P signaling) [33]. The last, and indirect method of CFb maintenance of the ECM is through cellular signaling. Fibroblasts, in response to angiotensin II, increase production of TGF-P which in turn act as a potent autocrine signal for ECM modulation [34]. Further, CFb are the primary source of tumor necrosis factor-a (TNF-a) which serves as both an autocrine signaling molecule for fibroblast activity, and a paracrine signal molecule for numerous 9 other cell types [35]. Interleukin-6 (IL-6) is another CFb-produced molecule that influences cell activity in the heart [36]. In response to proinflammatory cytokines, human CFb has been shown to upregulate expression of cell surface markers such as ICAM-1 and E-selectin that facilitate leukocyte recruitment [37]. These cells in turn, express ROS that enhance the function of LOX. Until recently, the focus on cardiac tissue has been on the elegant excitation contraction mechanism of cardiac myocytes. Fibroblasts have been relegated to bystander status with little appreciation given for the role they play in maintenance and support of the myocardium. We are now beginning to understand just how important fibroblasts are in the production, maintenance and remodeling of the cardiac ECM. Myocytes play little role in the ECM synthesis and architecture yet rely on this matrix to act in synchrony to develop a functional contraction to pump blood through our vasculature. Understanding the role that fibroblasts play in creating and maintaining the exquisite balance of mechanical properties needed for optimum cardiac performance is paramount for the development of regenerative strategies for cardiac repair. 1.2.4 Fibroblast Contributions to Biomechanical Stiffness Fibroblasts are the most numerous stromal cell type in the heart. Fibroblasts have been shown to contribute to tissue stiffness through two different mechanisms: directly through tractional forces including cytoskeletal filament contraction [38] and indirectly through expression and assembly of matrix elements. Here both contributions of fibroblasts to the biomechanical stiffness of 3D tissue will be reviewed. 10 Some of the earliest studies investigating cell tractional forces were performed in collagen gels. Bell and colleagues developed a unique 3D model of dermal fibroblast (Fb) behavior by casting living fibroblasts inside a collagen gel [39]. They observed six to eight fold decreases in collagen expression in 3D collagen matrices when compared to 2D culture conditions. Further, the collagen that was produced in the 3D model appeared to be tightly bound to the surrounding matrix while collagen produced in 2D culture was passed to the culture medium [40]. In addition, they observed that upon incorporation into the gel lattice, fibroblasts began to extend processes and collect collagen fibrils. This compaction of collagen resulted in macroscopic contraction of the collagen gel [41]. Grinnell and colleagues modified this model to evaluate fibroblast mediated contraction of collagen gels. They utilized either floating, unconstrained collagen gels, or mechanically anchored gels to evaluate fibroblast mediated contraction of the gels [42]. Floating gels were unstressed throughout the experiments, while anchored gels developed internal stresses. Two different mechanisms of contraction of the gels were observed; in unstressed gels, tractional forces resulting from extension of Fb processes was responsible for contraction. Moreover, in unstressed gels, the fibroblasts maintained a dendritic morphology and did not produce high level of actin stress fibers [43]. In contrast to their behavior in unstressed gels, Fb in stressed gels were observed to elongate into a lamellar phenotype expressing higher levels of smooth muscle actin filaments [44]. Upon release of the gel, these cells also were observed to contract the collagen gel. Although the outcome of gel contraction was identical, the difference in Fb response to the two conditions indicated the possibility of a different mechanism responsible for each condition. Indeed it was observed that contraction of unstressed collagen is primarily due 11 to extension of cellular processes that create tractional forces, while contraction of stressed gels results from smooth muscle-like contraction of the Fb themselves [45]. Grinnell further demonstrated the difference in these mechanism by showing that contraction can be induced in floating gels by platelet derived growth factor (PDGF) stimulation, while contraction in the stressed state is not induced by PDGF [46]. Coupling cell-populated collagen gels to strain gauges in order to measure cell tractional forces allowed assessment of tractional force, however this approach is limited to measuring bulk forces, and cannot assess the contribution of individual cells [38]. Elegant methods for measuring individual cell tractional forces (CTF) have been developed to understand the forces that can be generated by fibroblasts to complement the work that was performed using cell-populated collagen gels. In 1980, Harris et al. described a method of culturing cells on a thin polymer film and observing wrinkling of the film as a measure of CTF [47]. This method was improved by anchoring the edges the silicone film, and incorporating fiduciary beads. The embedded beads allow quantification of surface deformation, and therefore permit estimation of the CTF with force resolution approaching nanonewtons [48]. The elasticity of the silicone film in conjunction with the surface deformation can be used to create a surface force map. An alternative to the use of silicone films is the use of polyacrylamide gels [49]. Mircobead markers embedded in the gel provide deformation data, while the concentration of acrylamide determines the stiffness of the gel. The low affinity of cells for polyacrylamide gels requires gel coating with ECM proteins such as fibronectin to allow mediate cell attachment. Use of micromachined force sensors is another approach to 12 measuring CTF. Du Roure and colleagues developed a micropost array that is able to measure post deformation and correlate the deformation directly to cellular forces [50]. In addition to the potential for active contraction, cardiac stromal cells contribute to stiffness of cardiac tissue through expression and assembly of matrix elements as described in the previous section. Collagen (primarily type I and type III) is the predominant constituent of cardiac extracellular matrix. By volume, it comprises 2-6% of the myocardium [51]. Despite the low volume fraction, collagen appears to be the primary contributor to myocardial stiffness due to its relatively high modulus. The modulus of a single procollagen molecule ranges from 2.9 to 9.0 GPa as measure using an X-ray diffraction technique [52]. Once assembled into macroscopic filaments, collagen fibers are considerably softer, measured at 0.7-3.7 MPa using tensile stress-strain analysis [53]. Collagen fibers are coupled to the cells themselves through lateral struts that are anchored through focal adhesion complexes [54]. Confocal microscopy has revealed that interstitial collagen fibers in the relaxed heart resemble a helical twisted ribbon with relatively high degree of tortuosity [55]. During diastole as the myocardium is strained, it appears that perimysial collagen fibers are not elongated, but rather undergo helical bending and uncrimping. The energy required to bend and uncrimp the collagen appears to accounts for the majority of the measured stiffness in the myocardium [56]. In a mathematical model of cardiac tissue these forces have been calculated to be on the order of 135-160 MPa [56] which correspond to roughly 100% of the modulus measured in rat ventricular tissue [57]. In the model described in this research, Fb are grown on a non-degradable, elastic polyurethane substrate. The goal of the research to evaluate tissue-level changes in 13 passive stiffness, so while this model as described in this work cannot discriminate between individual contributions of cells and matrix, it can provide an overall functional picture of changes in tissue stiffness. In reality, the overall changes in stiffness measured in this model are a composite value corresponding to contribution of both active cell tractional forces, and passive matrix stiffness. 1.3. Cardiac Disease and Heart Failure 1.3.1 Cardiac Disease Heart disease is a leading cause of death worldwide [58]. In America, an average of 2,300 people died every day from cardiovascular disease in 2006 [1]. Many cardiovascular diseases including infarction, hypertension, valvular diseases, and congenital heart disease lead to the same end stage outcome: heart failure [59]. Heart failure is currently only treatable with transplant and has a high associated mortality. Moreover, heart failure is a chronic condition that constitutes a major socio-economic burden to nations, and a significant personal burden to families and individuals. Rates of heart failure are increasing in developed countries due to longer life-spans and increased likelihood of surviving of acute myocardial injury (primarily infarction) [22]. Heart failure occurs when the cardiac output is unable to meet the metabolic demands of the body. There are two mechanisms of heart failure: systolic and diastolic [60]. Systolic heart failure occurs when the heart is unable to generate sufficient force to eject all of the blood from the heart. Ejection fraction is typically decreased during systolic heart failure. Systolic heart failure results from decreased efficiency of cardiac contraction. Ischemia/infarction can cause myocyte necrosis and apoptosis, which may 14 contribute to systolic HF. The primary impetus, however, for systolic HF is expansion of the dilatation of the ventricular chamber which increases wall stresses and creates further stress on functional myocytes [22]. Diastolic heart failure, on the other hand results from inadequate filling of the ventricle. This is almost exclusively a result of fibrosis, which in turn is a secondary effect of a primary cardiac pathology. During diastolic heart failure, increases in stiffness of the cardiac muscle prevent adequate filling of the ventricle during diastole. So while ejection fraction may remain high, cardiac output is decreased. Both systolic and diastolic heart failure are results of maladaptive structural remodeling of the heart. While there are many cardiovascular diseases that share HF as an end-stage outcome, here only HF induced through coronary artery disease and myocardial infarction is addressed. 1.3.2 Maladaptive Remodeling Studies of the biomechanical response of the left ventricle (LV) to myocardial infarction (MI) have identified infarct expansion (i.e., stretching) and infarct-induced fibrosis of remote tissue as important phenomenon that both initiate and sustain a progressive pathologic process that ultimately results in structural and functional changes of the heart [61, 62]. This maladaptive response is a complex biomechanical process caused by depletion of myocytes from the infarct zone and a biologic response to abnormal stress distributions in the heart [63]. Clinically, ventricular remodeling is a primary indicator of adverse outcomes such as chronic heart failure and death [64]. 15 Within minutes to hours of ischemia-induced hypoxia, cellular death occurs and cardiac tissue ceases to be an active, force-generating tissue, and assumes passive viscoelastic mechanical properties [65]. Systolic stiffness of the infarcted tissue is decreased drastically. During this phase, the tissue begins to be depleted of its myocyte population through necrosis (20% of cells) and apoptosis (80% of cells) [66]. Simultaneously, the disparity between the systolic stiffness of the unimpaired heart and infarcted region results in dyskinesis and stretching of the infarcted region [67]. This mismatch in stiffness dissipates the mechanical energy generated by the contractile region of the heart, resulting in decreased ejection fraction (EF). Stroke volume is maintained by increasing the end diastolic volume (EDV), and results in long-term cardiac dilatation. However, cardiac dilatation further increases the burden on functional myocytes through increased wall stresses according to Laplace's law [63]. Moreover, it is during this acute phase prior to infarct stiffening through neomatrix formation that the risk of catastrophic failure resulting from rupture of the ventricle is highest [61]. A rapid increase in the passive stiffness of the infarcted tissue is imperative for maintaining cardiac function. Within hours, the passive (diastolic) stiffness of the tissue begins to increase due to edema [65]. In an ovine model of MI, passive stiffness of infarcted tissue increased by approximately 250% after 4 hrs [68]. At about 1 week postinfarction in humans, the fibrotic phase begins resulting in accumulation of collagen produced by activated myofibroblasts. This local deposition of collagen results in significantly increased stiffness of the infarct. In the previously mentioned ovine model, the passive stiffness increased by 800% at 1 week post-MI, and 1600% at 2 weeks [68]. This increase in 16 stiffness is critical for minimizing infarct expansion, decreasing the risk of rupture, and restoring ejection fraction by decreasing the energy loss due to infarct stretching. Indeed insufficient stiffening of the infarcted region results in infarct expansion, wall thinning, and dilatated cardiomyopathy; all of which are predictive factors pointing toward systolic heart failure [61]. While stiffening of the infarcted region is critical for prevention of maladaptive remodeling, it is a double-edged sword. Stiffening of the infarct has been associated with stiffening of regions remote to the infarct as well. Gupta et al. showed that in the ovine model of MI, the passive stiffness of the remote myocardium had increased >300% at 1 week [68]. Using a rabbit model of hypertension-induced heart failure, Yamamoto et al. demonstrated that changes in the passive stiffness of cardiac tissue are due to fibrosis, rather than compensatory hypertrophy of the myocyte population that is associated with both hypertensive and MI induced heart failure [20]. In a sheep model of MI, Wilson et al. demonstrated that matrix metalloproteinase (MMP) and tissue inhibitors of metalloproteinases (TIMP) levels in the remote regions were altered compared to control values indicative of ECM remodeling remote to the infarct [69]. Accumulation of collagen in sites remote to the infarct results in pathologic fibrosis, and increased passive stiffness of the myocardium which in turn causes decreased EDV of the heart, decreased cardiac output, and diastolic heart failure. It is clear that changes in active and passive mechanical properties of both the infarct and noninfarcted myocardium are key parameters affecting ventricular remodeling and corresponding health outcomes after MI. During the early stages after MI it is essential to have a significant increase in passive stiffness of the infarcted zone. However, in later stages, increased fibrosis in the infarct leads to remote matrix 17 18 accumulation, global stiffening of the heart, pathologic fibrosis, and diastolic heart failure. 1.3.3 The Role of the Cardiac Fibroblast in Remodeling Cardiac fibroblasts are jointly responsible with acute phase inflammatory cells for wound healing after cardiac injury. The successful recapitulation of cardiac function after the initial insult is largely dependent on the activity of CFb. Upon infarction, inactive MMP found in the interstitial space become are zymogenically activated as surrounding matrix is degraded in response to necrotic and apoptotic signals from myocytes. This immediate activation results in a local and significant degradation of collagen. Takahashi et al. demonstrated that within 3 hrs after infarction, 50% of collagen in the infarct region was degraded in rats [70]. Increased MMP activity and resultant decreases in the collagen volume fraction in the heart have been correlated with ventricular dilatation and systolic dysfunction [71]. Collagen cross-linking is another important factor that impacts cardiac function. Fibroblasts affect collagen cross-linking through LOX production and secretion. While it has been shown that collagen production increases in the infarct region [72], this does not necessarily translate to an increase in tissue stiffness (see [68]). Several studies have correlated increases in collagen cross-linking with increases in apparent myocardial stiffness [73, 74]. These increases in apparent myocardial stiffness, and not just collagen content are hypothesized to be responsible for diastolic dysfunction of the heart. The passive stiffness of the myocardium is paramount in determining post-MI outcomes. Scar formation in myocyte depleted regions is a dynamic process consisting of 19 continuous collagen turnover, cytokine expression, and myofibroblastic activation persisting in humans for decades [75]. ECM and therefore the cardiac fibroblast clearly have an important role in maintaining the geometry and elasticity of the heart in both the acute and long term after infarction. Improved understanding of the mechanisms responsible for fibroblast-mediated changes in stiffness will provide a foundation for novel therapeutic approaches to treating heart failure. 1.4. Cardiac Cell Therapy Under normal conditions after MI, myocytes are depleted from the wound site via necrosis or apoptosis. Fibroblasts replace the missing myocytes with a stiff, dense scar. Little regeneration of cardiac muscle occurs. Chronic mechanical stimulation and autocrine signaling through TGF-P and TNF-a results in maladaptive remodeling and heart failure initiated by changes in ventricular fibroblast activity. Starting around the year 2000, a novel approach to regenerating cardiac muscle began to be explored: cardiac cell therapy. The complex biologic and mechanical impetus of ventricular remodeling makes it a challenging pathology to treat. Indeed, medical device and pharmacologic treatments are used to manage and minimize cardiac remodeling, but are generally unable to reverse its course [62]. Heart transplantation remains the most effective method for treating maladaptive remodeling, but is limited by a persistent shortage of donor organs. Cardiac cell therapy is a promising new treatment that is currently in the early phases of clinical trials [2, 76]. In a single case study performed by Okano and colleagues, a myoblast "cell-sheet" was delivered to a patient with progressive heart failure. After implantation, this treatment improved the patient's EF from 26% to 46%, allowed explantation of his left ventricular assist system, and removed him from the heart transplant waiting list [77]. Cell therapy utilizes exogenous cells to stimulate cardiac regeneration. Early attempts to regenerate cardiac function focused on delivering cells that would replace the myocytes lost to ischemic injury. In the last 12 years, however, the cardiac cell therapy paradigm has shifted largely away from replacement of myocytes, to cytoprotective and regenerative paracrine support of cardiac function [66, 78]. Utilization of mesenchymal stem cells (MSC) in animal models, has shown considerable promise as a treatment to prevent postinfarction remodeling. Moreover, the low immunogenicity of MSC allows the use of allogenic donor cells and proprietary expansion methods of MSC make this cell type attractive to biotech companies interested in commercialization [5]. In small animal models of MI, MSC have been shown to effectively treat some of the major histologic and functional pathologies associated with ventricular remodeling. Mias et al. showed that in a rat MI model, injection of MSC pretreated with melatonin significantly increased ejection fraction (EF), and decreased wall thinning and collagen accumulation [4]. Godier-Furnemont and colleagues utilized a decellularized human tissue patch to deliver TGF-P pre-conditioned mesenchymal precursor cells to a nude rat MI model [79]. Histologically, they observed increased angiogenesis; functionally, they observed increased fractional shortening (FS) of the heart. Another group genetically modified MSC to increase prostaglandin I synthase (PG1S) transcription. Increased PG1S production resulted in increased FS and EF, while simultaneously decreasing fibrosis, wall thinning and apoptosis [80]. Gnecchi et al. showed that MSC's modified with ATK- 20 1, a cytoprotective factor encoding gene, were shown to result in smaller infarct sizes and decreased myocyte apoptosis after infarction in rats [81]. Despite the numerous animal studies performed, and their favorable results, the mechanism behind the observed efficacy of cell therapy in animal models is still poorly understood. There are, in fact, at least five proposed paracrine mechanisms that may be responsible for the histologic and functional improvements observed in animal studies: (1) Decreased remodeling of the myocardium, (2) paracrine mediated angiogenesis, (3) modulation of the immune response, (4) cytoprotection of residual myocytes in the infarct region and border zones, and (5) recruitment of resident cardiac progenitor cells [82-85]. This lack of mechanistic understanding becomes significantly more important when the results of early human clinical trials are considered. Unlike most animal models, the efficacy of cell therapy in humans has been questionable. A meta analysis of 13 clinical trials with a total of 811 patients revealed that bone marrow cell therapy did not improve postinfarction remodeling [5]. Further mechanistic investigation of cell therapy is paramount for moving cell therapies from the trial phase to a standard treatment in the clinic. Certain approaches to cell therapy have proposed that the fibroblast may be an important therapeutic target for cell therapy. Mias et al. showed that fibroblast MMP expression was altered by paracrine factors from injected MSC [4]. While this study is key, there is no direct link between the expression of MMP and mechanical properties, which are more directly responsible for the function of the heart. Development of a model that couples changes in bulk mechanical properties based on cell injection 21 22 approaches will accelerate our ability to design cell-based therapies that more effectively prevent maladaptive remodeling of the heart. 1.5. Three-Dimensional Cell Culture Models In vitro cell culture is a common tool used to evaluate the response of cells to stimuli. Practically, cell culture is an essential component of all drug and device development. Indeed ISO 10993 - the international standard for medical device biocompatibility - requires that cell culture be used to evaluate cytotoxicity for all medical devices adhering to this standard. The primary components of cardiac cell culture strategies are: (1) the substrate, (2) the environmental culture conditions, (3) the culture medium, and (4) the cells themselves [86]. Since the first attempts at cell culture, researchers have been defining and refining these elements of cell culture. Current tissue culture plastic (TCP) based cell culture models are the result of decades of discovery and development. In more recent history, focus has shifted from development of cell culture models that are cultured on a flat substrate like TCP, to development of three-dimensional (3D) cell culture models [87]. This has been a particularly challenging paradigm shift to implement, because both the substrate and the culture conditions must change drastically to accommodate 3D cell culture. This section will review some of the key developments in substrate design, and culture platform innovation that have facilitated development of 3D models. Further, recent applications of 3D tissue models to understanding cardiac cell therapy will be reviewed. 23 1.5.1 The Substrate Cardiac cells, including fibroblasts, are anchorage-dependent, and therefore require a substrate for attachment to facilitate survival, activity, and proliferation. In the realm of tissue engineering there are currently four different approaches to providing a 3D substrate: (1) utilization of naturally occurring substrates from animals, (2) cell entrapment within hydrogels, (3) lamination of cell sheets with a self-assembled substrate, and (4) utilization of porous, engineered substrates [88-90]. Each of these approaches provides engineering control over two key substrate properties: (1) material composition (2) substrate microarchitecture, including porosity, alignment, and mechanical properties. 1.5.1.1 Material Composition Evaluation of cardiac fibroblast (CFb) response to material composition in the literature is very limited. Exclusive cardiac fibroblast culture in 3D substrates are almost nonexistent. Most studies that have cultured CFb on different materials have included the CFb as part of a coculture including myocytes. Lasher et al. utilized fibrin gel to coculture cardiac myocytes and cardiac fibroblasts [91]. Vimentin staining revealed cardiac fibroblasts distributed between myocytes throughout the construct. These findings indicate that CFb are able to maintain vitality when cultured in serum containing DMEM-based media. Moreover, electrical stimulation did not appear to affect CFb numbers [91]. Freed et al. evaluated CFb cultured on a synthetic, elastomeric polymer substrate. Her group stained for F-actin in a mixed neonatal rat heart cell population. They observed that F-actin expressing cells (primarily CFb) filled the void spaces in porous poly (glycerol- 24 sebacate) substrates [57]. Our own group demonstrated that cardiac fibroblasts will attach and proliferate on poly (ether-urethane) substrates in serum containing media (See Chapter 3). Further we observed that cell attachment after perfusion seeding significantly increased with a composite fibronectin-coated polyurethane substrate (unpublished data). 1.5.1.2 Architecture Substrate microarchitecture profoundly influences cell response to the substrate. Further, substrate microarchitecture can be used to influence functional properties of the tissue model such as stiffness and hydraulic permeability. The geometric configuration of the substrate on the micro scale is determined by material fabrication and processing techniques. Therefore, a discussion of substrate design parameters must necessarily be coupled with a discussion of fabrication methods and techniques. The key architectural features that will be discussed here are: (1) porosity, (2) alignment/anisotropy, (3) stiffness, and (4) hydraulic permeability. 1.5.1.2.1 Porosity. Porosity is defined as the ratio of void space to total volume of a porous material. This parameter affects how much space is available for cells to occupy, diffusion of nutrients through the substrate matrix, and the effective mechanical properties of the construct [92]. Techniques for evaluating porosity include mercury-intrusion porosimetry, optical analysis, and gravimetric analysis (based on material densities). These methods provide insight to pore size, pore morphology, and porosity respectively. Electrospinning is common biomaterial fabrication technique for creating fibrous, nonwoven materials with fiber diameters typically on the order nanometers [93]. A polymer dissolved in solution is drawn through a charged spinneret to a conductive surface by a high electrical potential, typically on the order of 10 kV [94, 95]. Baker et al. utilized an electrospinning approach to prepare fibrous of controlled porosity [96]. Sacrificial poly (ethylene oxide) (PEO) fibers were cospun with permanent poly (s-caprolactone) fibers. After spinning, the PEO fibers were removed by ethanol immersion for 3 hrs, resulting in increased porosity that was dependent on the fraction of PEO fibers in the original material. Cell culture studies revealed that cell infiltration into the interior of the 3D substrate increased with increased fraction of sacrificial fibers [96]. Phase separation is a technique that is frequently used to fabricate porous biomaterials. This fabrication technique provides significantly more flexibility to the ability to control porosity of the resultant material. Rowlands et al. used a thermal phase separation (TIPS) technique to form polyurethane/PLGA substrates with pores ranging from 0.1-200 p,m in diameter [97]. Woodhouse and colleagues created porous polyurethane substrates using TIPS. The resultant material was found to have nominal porosities ranging from 14% to 35% depending on fabrication conditions. When embryonic stem cell-derived cardiac cells were seeded onto these substrates, they were observed to infiltrate approximately 72 p,m into the tissue, with significantly decreased numbers beyond this depth [98]. Another variant to the phase separation approach is the sprayed technique. In sprayed phased separation, two spray nozzles are simultaneously employed to deposit a solvent solution and nonsolvent onto a surface. Khorasani et al. described an elegant sprayed phase separation technique where porous substrates were fabricated by spraying onto a rotating mandrel. SEM was used to evaluate morphological changes that occurred in response to changes in the working distance of the spray nozzles and the speed of the mandrel [99]. Papenburg et al. took the phase separation fabrication techniques a step further. They 25 introduced a notable micromolding enhancement to the phase separation approach. This method utilizes a patterned substrate to provide geometric patterns to a thin sheet of phase separated material [100]. Moreover, they investigated the use of several different alcohols as nonsolvents in substrate formation. They determined that the type of alcohol affected porosity and pore structure. Using this method, they were able to independently control surface patterns of thin sheets through the mold geometry, and porosity through variation of the composition of the nonsolvent [100]. In addition to these approaches, Papenburg and colleagues also proposed, simultaneously to my own research, lamination of thin phase separated sheets to build thick 3D constructs [101]. Control over porosity has evolved from simple use of sacrificial components, to elegant physiochemical-control mechanisms that provide multiple levels of input and capabilities to substrate fabrication techniques. The work described herein utilizes a sprayed phase separation approach to substrate fabrication because of the high level of control over porosity and the ease of automation of this approach. 1.5.1.2.2 Alignment. Substrate alignment is a key feature that relates both to cell response and bulk mechanical properties. Contact guidance is a well established hypothesis that postulates that cells will be guided through contacting external geometric cues [102]. Using advanced quantification techniques [103], Freed and colleagues demonstrated that cardiac fibroblasts would align according to geometric cues of laser-ablated synthetic substrates [57]. Using micromolding techniques, Papenburg et al. demonstrated that cells can be aligned according to CAD-defined geometric patterns [100]. In Chapter 3, utilization of postspray elongation technique is discussed in detail 26 that provides alignment cues to fibroblasts. Further, a FFT-based method for assessing cellular alignment in response to the substrate is described. 1.5.1.2.3 Mechanical properties. Substrate mechanical properties are often key determinants of construct function. In addition, substrate mechanical properties can have profound effects on cellular activity [17, 104]. For soft tissue (such as cardiac tissue) tensile properties are of primary interest. In general, mechanical properties are measured through tensile testing, and normalization of the results. Cardiac tissue is a highly anisotropic tissue, meaning that, structural and functional parameters are inhomogeneous. Tensile properties of cardiac tissue are generally referred to in terms of the longitudinal or preferred modulus, and the transverse modulus. Several approaches to recapitulating the anisotropy of cardiac tissue in biomaterial substrates have been proposed. One method for controlling mechanical properties of cell substrate is through material composition. Zhang et al. developed custom polyurethane base formulations, and demonstrated different tensile properties over time base on material composition [105]. Rowlands et al. utilized different blends of polyurethane and PLGA, and demonstrated moduli ranging from 15 to 93 kPa depending on the material composition [97]. Geometric composition is another approach to control mechanical properties. Baker et al. achieved anisotropic geometry by electro-spinning onto a rotating mandrel to create fiber alignment. This fiber alignment resulted in an approximately 10-fold difference between the elastic modulus in the preferred direction and transverse direction [96]. Moroni et al. utilized a 3D fiber deposition approach controlled by CAD-CAM techniques to create substrates with geometrically controlled moduli ranging from 0.26 MPa to 13.7 MPa [106]. In Chapter 3, a postspray elongation technique is used to create geometric base 27 tensile anisotropy. This anisotropy is evaluated both optically through electron microscopy, and mechanically through tensile testing. 1.5.1.2.4 Hydraulic permeability. The last architectural feature of biomaterial for substrates that will be discussed here is hydraulic permeability. Hydraulic permeability is defined as the permeability of a porous medium to aqueous solutions. This architectural feature is of special interest to 3D cell culture because media perfusion is often used to enhance mass transport in the interior of thick substrates. Hydraulic permeability is measured by monitoring flow of media through a substrate while maintaining a constant head pressure. Hydraulic permeability, like mechanical properties, can be either isotropic, or anisotropic. Geometric features such as channels can be included in the porous matrix to influence hydraulic permeability [107-109]. Papenburg et al. utilized the micromolding approach to design corrugations into thin sheets which were then rolled up resulting in channel-like geometries running through the materials [110]. Chapter 4 includes a detailed discussion hydraulic permeability of various types of substrates including the substrate developed for use in this work. 1.5.2. Culture Conditions Advances in design, processing, and fabrication of 3D substrates has necessitated concurrent advances in cell culture platforms. Native human tissue is sustained through a complex cooperation between multiple physiologic systems. As engineered tissues become more complex, the framework required to support the tissue will increase in complexity. Engineered tissue that is identical to native tissue will, by definition, require the same type of complex environmental support that is found in the body. The design of 28 advanced bioreactors and culture conditions is an essential step toward achieving more native like tissues. In an important review, Barron et al. defines a bioreactor as "a system that simulates physiological environments for the creation, physical conditioning, and testing of cells, tissues, precursors, support structures, and organs in vitro" [111]. In the realm of cardiac tissue engineering, there are three important environmental conditions that can be simulated through utilization of bioreactors: mechanical stimulation, electrical stimulation, and enhanced metabolic exchange through perfusive flow [111]. In addition, an essential function of a bioreactor for evaluating 3D cardiac tissue is to support a uniform cell distribution throughout a 3D substrate [112, 113]. The greatest impediment to obtaining a uniform cell distribution in a 3D model is maintaining proper mass transport. Oxygen, nutrients, and metabolites must be brought near enough to cells to allow diffusive transport to exchange the spent resources with fresh ones. Convective fluid transport is the most versatile way of supplementing the transport capabilities of diffusion. Without convective supplementation, diffusion can only provide support to cardiac tissue that is ~100 p,m thick [114]. In a mathematical model of oxygen transport and consumption, Radisic et al. demonstrated that perfusion of a channeled substrate can provide sufficient oxygen to support a 2 mm thick construct that is densely packed with cardiac cells [115]. In an in vitro model, Carrier et al. demonstrated that perfusion increased spatial uniformity of cardiac cells in a synthetic 3D substrate [116]. Liu and colleagues demonstrated that perfusion significantly increased the number of cells on an elastomeric polyurethane substrate when compared to static controls [117]. Moreover, perfusion provides important stimuli to cardiac cells. Maidhof 29 30 and colleagues showed that excitation threshold increased significantly without perfusion or electrical stimulation [118]. There are two distinct phases that occur during perfusion culture that influence cell distribution: seeding and steady state culture. The seeding phase of culture occurs in the beginning to help cells attach to the substrate. In a 3D substrate, efficient seeding is necessary to ensure that cells are anchored uniformly throughout the material. Maidhof et al. explored the important seeding parameters in elastomeric channeled substrates. They determined that flow velocity and duration of seeding of cardiac myocytes are important seeding parameters [108]. They evaluated seeding velocities ranging from 6 mm/min to 60 mm/min. They determined the highest flow (60 mm/min) resulted in the highest seeding efficiency (87%). When they investigated how long to maintain the seeding velocity, they found that 2 hrs provided a higher seeding efficiency than 1 hr. During the steady state culture phase, the flow velocity is the primary perfusion parameter. McCoy et al. developed a mathematical model to understand the effect of flow rate on cell attachment in a bioreactor [119]. Under high flow conditions, cells were detached from the substrate surface by the flow-induced shear stress. The work of McCoy et al. and Maidhof et al. help establish general constraints for seeding and steady state perfusion parameters: high velocity is good for a limited duration while cells are attaching, but lower velocities are better long term to prevent cells from detaching. 1.5.3 3D Cell Culture Applications 3D cell culture is only possible through the use of appropriate substrates and culture platforms. Development of 3D culture techniques, while challenging, has been primarily motivated by two different potential applications: therapeutic use of 3D cell culture (engineered tissue) to treat diseased or damaged tissue [88, 120, 121], and utilization of advanced cell culture as a bridge between 2D in vitro and in vivo testing [87, 122, 123]. Many of the design requirements for engineered tissue and 3D models overlap, however, there are fundamental differences. For example, cell culture models do not have the stringent requirements for systemic biocompatibility that are necessary for implanted tissues. While this can simplify some aspects of the design, the advantage is tempered by inability of in vitro models to undergo modification after implantation. For example, one approach to creating therapeutic engineered tissue relies on in vitro angiogenesis to provide vascularization [124, 125]. 3D models, on the other hand, will undergo angiogenesis and vascularization only if the culture conditions are designed specifically for this purpose. At the present, there has been only one attempt at developing 3D cardiac fibroblast models. Gaile et al. seeded CFb into collagen gels that were either free floating, or constrained at the edges and observed changes in mRNA levels of collagen (type I and III), TGF-P, and a-smooth muscle actin (a-SMA) [17]. No changes in collagen or TGF-P expression were detected. A significant increase in a-SMA was detected at 6 and 24 hrs, but became nonsignificant after 48 hrs. Further Gaile and colleagues monitored the stiffness of the constructs after 6 hrs and 120 hrs. They noted that the elastic modulus was similar to fibrotic cardiac tissue after 120 hrs and proposed that a 3D cell model of cardiac fibroblasts could be used to better understand fibrosis in the heart [17]. 31 While engineered tissue models have primarily been focused on cardiac myocytes, some models have been developed that utilize a mixed cell population of both myocytes and fibroblasts. Freed and colleagues demonstrated that coculture of myocytes and CFb resulted in higher levels of MMP 2 expression than CM enriched populations [126]. In another study, Freed and colleagues observed that cardiac fibroblasts when cocultured with myocytes aligned to match scaffold geometry [57]. Beyond the scope of cardiac fibroblasts, the concept of a 3D in vitro model for evaluating stem cell therapy has been pioneered by Radisic and colleagues [6]. Using a collagen sponge substrate and an electrical stimulation bioreactor, Song et al. investigated the functional integration of injected stem cells and surrogate heart tissue. They observed that embryonic stem cell derived cardiomyocytes (ESC-CM) cultured in low glucose were able to significantly improve the excitation threshold of the surrogate cardiac tissue [7]. On the other hand, these same cells also decreased the maximum rate of synchronous contraction of these constructs. In another important study, Radisic's group used surrogate cardiac tissue to understand embryonic stem cell (ESC) response to a cardiac environment [8]. They observed that ESC can survive and even proliferate in a cardiac environment. Despite this initial positive finding, they also observed that ESC did not functionally integrate with the surrogate heart tissue. The findings of these landmark studies have pushed researchers to look for other potential mechanisms of cardiac regeneration through cell therapy, and demonstrate the utility of 3D cell models in evaluating and designing stem cell therapies. The last type of in vitro 3D model that will be discussed here is explanted cardiac tissue models. Pillenkamp developed a model of cardiac tissue by explanting and 32 sectioning murine hearts. These 300 p,m thick slices of cardiac tissue were functionally maintained using static cell culture techniques for up to 24 hrs before evidence of apoptosis became evident [127]. These models were used to assess the influence of ESC-CM injection on force of contraction of the tissue [127]. The results indicated that ESC-CM injection improved contractile force of the constructs. Pillenkamp et al., like Radisic's group, observed that ESC-CM did not functionally integrate with surrounding cardiac tissue [128]. Another group modified the tissue slice model by utilizing a gas-liquid interface bioreactor to increase the time that the tissue slice could survive ex vivo [129]. The gas-liquid interface culture platform allowed Habeler and colleagues to functionally preserve the tissue slice for up to 80 days. Further application of this model demonstrated that 60 days after injection, ESC in the heart slice had adapted a cardiac myocyte like phenotype [130]. The future of more complex experimental treatments for heart failure will rely on better methods to understand these treatments. Models like surrogate heart tissue and heart slices have demonstrated their ability to make a significant impact on the development of novel treatments however, these models rely on complex substrates, and advanced bioreactors to simulate the necessary environmental signals. Further expansion of 3D tissue models to include cardiac fibroblasts will broaden our capacity to evaluate the next generation of treatments for chronic heart disease. 1.6 Summary, Overview, and Contributions There is an important opportunity for advanced cell culture platforms to make a significant impact on the standard treatment of heart failure by helping successfully 33 design stem cell therapy treatments. In this chapter the physiologic role of the cardiac fibroblast, the impetus and mechanics of heart failure, cardiac cell therapy, and state-of-the- art cell culture and tissue engineering models were reviewed. The overarching objective of the research described in this dissertation was to directly correlate MSC treatment with CFb mediated changes in stiffness. This objective was addressed through the development of culture materials, a specialized bioreactor and utilization of newly developed methods to coculture fibroblasts and MSCs. In Chapter 2 the design requirements and specifications for foundational equipment used in all further chapters are described. Chapter 2 includes a description of pertinent design spray parameters, as well as the mechanical design for the spray assembly and substrate fixturing. In Chapter 3, novel methods for fabricating a porous polyurethane substrate that mimics the mechanical anisotropy of ventricular heart tissue are described. Dual spray nozzles were mounted onto a three axis computer numeric control robot (see appendix A for engineering drawings of spray platform). An elastomeric poly(ether-urethane) dissolved in N-dimethylacetamide (DMAc) was sprayed through one nozzle while a water-alcohol mixture serving as the nonsolvent was sprayed simultaneously through the other. The polymer begins to precipitate on the spray substrate as the solvent is diluted in the nonsolvent. After spray deposition, the biomaterial was uniaxially strained in custom frames that held the material in an elongated conformation during the curing period. After 24 hrs the biomaterial had relaxed and cured sufficiently to be removed from the frames while permanently maintaining the elongated architecture. Electron microscopy, gravimetric porosity analysis, and mechanical testing were used to evaluate the 34 35 architecture of the biomaterial. To validate that the architecture observed through characterization testing would influence cell behavior, 3T3 murine dermal fibroblasts were seeded and cultured on 250 ^m thick sheets. A two-dimensional fast fourier transform (2D FFT) was used to evaluate cell alignment with the direction of postspray elongation. Fibroblasts were observed to increase alignment of cytoskeletal actin filaments as the degree of postspray elongation increased. To demonstrate further flexibility of the substrate, multiple thin sheets were laminated after spraying. After the curing period, thick robust sheets were observed with little discernible evidence of lamination boundaries when imaged with an electron microscope. In Chapter 4 two different methods for supporting cell populations in the interior of a 1 mm thick laminated substrate were developed and evaluated: (1) development of channel-like features running along the length of the substrate, and (2) development of perfusion seeding and culture. Phase separation molding techniques were adapted to the sprayed phase separation technique described in Chapter 3. Computer aided design (CAD) techniques and rapid stereo-lythographic prototyping were employed to fabricate corrugated spray substrates. Sprayed phase separation of the polymer onto this surface yielded reciprocal corrugations on one surface of the porous polyurethane material. Electron microscopy characterization confirmed that the corrugation pattern closely resembled the CAD pattern. Lamination of six corrugated sheets created a composite substrate approximately 1 mm thick with channel-like features arrayed in parallel along the length of the material. Hydraulic permeability testing confirmed the functional performance of the channels. One of the fundamental limitations of 3D materials for cell culture is inadequate mass transport near the cells. A perfusion bioreactor was design to allow continuous perfusion through the channels and porous matrix of the substrate. Human mesenchymal stem cells were seeded either using a surface or perfusion approach. Following a 7 day culture period, cellularized samples were evaluated for cell density and cell distribution. Perfusion seeding and culture were observed to increase cell density and provide a more homogenous distribution of cells through the substrate. The presence of channel-like features in the substrate increased cell density in perfusion seeded samples when compared to nonchanneled materials. In Chapter 5, the perfusion culture system was modified and utilized in conjunction with the channeled substrate to evaluate the effect of MSC coculture on CFb-mediated changes in stiffness. Appendix B contains engineering drawings and design considerations for the individual perfusion cassette cell culture platform used in Chapter 5. Chapter 5 first describes design and validation of the culture platform. Dermal fibroblasts were used to evaluate fibronectin preconditioning of the substrate and optimize seeding flow velocity for cell density and distribution. Cardiac fibroblasts were then used in a second seeding optimization investigation seeding velocity and duration. The purpose of the culture platform is to allow direct evaluation of treatment induced changes in stiffness. To validate that the model is capable of detecting increases in stiffness, dermal fibroblasts were cultured with and without ascorbic acid. Stiffness assessment demonstrated a significant increase between acellular substrates and fibroblast seeded constructs, and again between fibroblast seeded constructs without ascorbic acid supplementation, and constructs with ascorbic acid supplementation. To validate that the model was capable of detecting treatment induced decreases in stiffness, a common ACE inhibitor that has been shown to inhibit fibroblast mediated fibrosis was 36 used to treat cardiac fibroblast seeded constructs. After cell culture, the stiffness of the ACE inhibitor treated group was significantly less stiff than untreated CFb group. To test whether MSC coculture influence stiffness, CFb and MSC were cultured together on the substrate. End-point stiffness assessment revealed that MSC coculture resulted in increased stiffness. Confocal imaging revealed that MSC were present and type III collagen was expressed in higher abundance in the coculture samples. Chapter 6 discusses the conclusions from this work, specifically that MSC increase construct stiffness when cocultured with CFb. These findings seem to indicate that clinical application of MSC to alter is stiffness is a feasible approach to treating passive biomechanical heart pathologies. Moreover, future development of stem cell and pharmacologic therapeutics be evaluated using the model platform described in this work to quantitatively evaluate treatment induced changes in stiffness. 1.6.1 Contributions Chapters 3 and 4 are comprised of multiple author peer-reviewed publications. In Chapter 3 James Kennedy, Sean McCandless, and Robert Hitchcock were responsible for conception, design of the study, and experiments. Richard Lasher and James Kennedy were responsible for development of image analysis algorithms. James Kennedy analyzed the data and primarily wrote the manuscript. Robert Hitchcock, Sean McCandless, and Richard Lasher assisted in writing and provided critical review of the manuscript. All authors approved the final version of the manuscript. In Chapter 4, James Kennedy, Sean McCandless, Asad Rauf, and Robert Hitchcock were responsible for conception, design of the study, and experiments. James 37 38 Kennedy, Jennifer Hillam and Laura Williams performed cell culture experiments. James Kennedy analyzed the data and primarily wrote the manuscript. Sean McCandless, Asad Rauf, and Robert Hitchcock assisted in writing and provided critical review of the manuscript. All authors approved the final version of the manuscript. 1.7 References [1] Lloyd-Jones D, Adams RJ, Brown TM, Carnethon M, Dai S, De Simone G, et al. Heart disease and stroke statistics--2010 update: a report from the American Heart Association. Circulation. 2010;121:e46-e215. [2] Segers VF, Lee RT. Stem-cell therapy for cardiac disease. Nature. 2008;451:937-42. [3] Dai W, Hale SL, Martin BJ, Kuang JQ, Dow JS, Wold LE, et al. Allogeneic mesenchymal stem cell transplantation in postinfarcted rat myocardium: short- and long-term effects. Circulation. 2005;112:214-23. [4] Mias C, Lairez O, Trouche E, Roncalli J, Calise D, Seguelas MH, et al. Mesenchymal stem cells promote matrix metalloproteinase secretion by cardiac fibroblasts and reduce cardiac ventricular fibrosis after myocardial infarction. Stem Cells. 2009;27:2734-43. [5] Menasche P. Cardiac cell therapy: lessons from clinical trials. J Mol Cell Cardiol. 2010;50:258-65. [6] Nunes SS, Song H, Chiang CK, Radisic M. Stem cell-based cardiac tissue engineering. J Cardiovasc Transl Res. 2011;4:592-602. [7] Song H, Yoon C, Kattman SJ, Dengler J, Masse S, Thavaratnam T, et al. Interrogating functional integration between injected pluripotent stem cell-derived cells and surrogate cardiac tissue. Proc Natl Acad Sci U S A. 2010;107:3329-34. [8] Dengler J, Song H, Thavandiran N, Masse S, Wood GA, Nanthakumar K, et al. Engineered heart tissue enables study of residual undifferentiated embryonic stem cell activity in a cardiac environment. Biotechnol Bioeng. 2011;108:704-19. [9] Nag AC. Study of non-muscle cells of the adult mammalian heart: a fine structural analysis and distribution. Cytobios. 1980;28:41-61. [10] Camelliti P, Borg TK, Kohl P. Structural and functional characterisation of cardiac fibroblasts. Cardiovasc Res. 2005;65:40-51. 39 [11] Goldsmith EC, Hoffman A, Morales MO, Potts JD, Price RL, McFadden A, et al. Organization of fibroblasts in the heart. Dev Dyn. 2004;230. [12] Duffy HS. Fibroblasts, myofibroblasts, and fibrosis: fact, fiction, and the future. J Cardiovasc Pharmacol. 2011;57:373-5. [13] Agocha A, Sigel AV, Eghbali-Webb M. Characterization of adult human heart fibroblasts in culture: a comparative study of growth, proliferation and collagen production in human and rabbit cardiac fibroblasts and their response to transforming growth factor-beta1. Cell Tissue Res. 1997;288:87-93. [14] Curtis MW, Russell B. Micromechanical regulation in cardiac myocytes and fibroblasts: implications for tissue remodeling. Pflugers Arch. 2011;462:105-17. [15] Gabbiani G, Ryan GB, Majne G. Presence of modified fibroblasts in granulation tissue and their possible role in wound contraction. Experientia. 1971;27:549-50. [16] Daskalopoulos EP, Janssen BJ, Blankesteijn WM. Myofibroblasts in the infarct area: concepts and challenges. Microsc Microanal. 2012;18:35-49. [17] Galie PA, Westfall MV, Stegemann JP. Reduced serum content and increased matrix stiffness promote the cardiac myofibroblast transition in 3D collagen matrices. Cardiovasc Pathol. 2011;20:325-33. [18] Wang H, Haeger SM, Kloxin AM, Leinwand LA, Anseth KS. Redirecting valvular myofibroblasts into dormant fibroblasts through light-mediated reduction in substrate modulus. PLoS One. 2012;7:e39969. [19] Weber KT, Sun Y, Bhattacharya SK, Ahokas RA, Gerling IC. Myofibroblast-mediated mechanisms of pathological remodelling of the heart. Nat Rev Cardiol. 2013;10:15-26. [20] Yamamoto K, Masuyama T, Sakata Y, Nishikawa N, Mano T, Yoshida J, et al. Myocardial stiffness is determined by ventricular fibrosis, but not by compensatory or excessive hypertrophy in hypertensive heart. Cardiovasc Res. 2002;55:76-82. [21] Eghbali M, Robinson TF, Seifter S, Blumenfeld OO. Collagen accumulation in heart ventricles as a function of growth and aging. Cardiovascular Res. 1989;23:723-9. [22] Leonard BL, Smaill BH, Legrice IJ. Structural remodeling and mechanical function in heart failure. Microsc Microanal. 2012;18:50-67. [23] Webb K, Hitchcock RW, Smeal RM, Li W, Gray SD, Tresco PA. Cyclic strain increases fibroblast proliferation, matrix accumulation, and elastic modulus of fibroblast-seeded polyurethane constructs. J Biomech. 2006;39:1136-44. [24] Turner NA, Porter KE. Regulation of myocardial matrix metalloproteinase expression and activity by cardiac fibroblasts. IUBMB Life. 2012;64:143-50. 40 [25] Foronjy RF, Sun J, Lemaitre V, D'Armiento JM. Transgenic expression of matrix metalloproteinase-1 inhibits myocardial fibrosis and prevents the transition to heart failure in a pressure overload mouse model. Hypertens Res. 2008;31:725-35. [26] Hayashidani S, Tsutsui H, Ikeuchi M, Shiomi T, Matsusaka H, Kubota T, et al. Targeted deletion of MMP-2 attenuates early LV rupture and late remodeling after experimental myocardial infarction. Am J Physiol Heart Circ Physiol. 2003;285:H1229-35. [27] Spinale FG, Mukherjee R, Zavadzkas JA, Koval CN, Bouges S, Stroud RE, et al. Cardiac restricted overexpression of membrane type-1 matrix metalloproteinase causes adverse myocardial remodeling following myocardial infarction. J Biol Chem. 2010;285:30316-27. [28] Limana F, Esposito G, D'Arcangelo D, Di Carlo A, Romani S, Melillo G, et al. HMGB1 Attenuates Cardiac Remodelling in the Failing Heart via Enhanced Cardiac Regeneration and miR-206-Mediated Inhibition of TIMP-3. PLoS One. 2011;6:e19845. [29] Vanhoutte D, Heymans S. TIMPs and cardiac remodeling: 'Embracing the MMP-independent- side of the family'. J Mol Cell Cardiol. 2009;48:445-53. [30] Liu H, Chen SE, Jin B, Carson JA, Niu A, Durham W, et al. TIMP3: a physiological regulator of adult myogenesis. J Cell Sci. 2010;123:2914-21. [31] Molnar J, Fong KS, He QP, Hayashi K, Kim Y, Fong SF, et al. Structural and functional diversity of lysyl oxidase and the LOX-like proteins. Biochim Biophys Acta. 2003;1647:220-4. [32] Lopez B, Gonzalez A, Hermida N, Valencia F, de Teresa E, Diez J. Role of lysyl oxidase in myocardial fibrosis: from basic science to clinical aspects. Am J Physiol Heart Circ Physiol. 2010;299:H1-9. [33] Voloshenyuk TG, Hart AD, Khoutorova E, Gardner JD. TNF-alpha increases cardiac fibroblast lysyl oxidase expression through TGF-beta and PI3Kinase signaling pathways. Biochem Biophys Res Commun. 2011;413:370-5. [34] Lee AA, Dillmann WH, McCulloch AD, Villarreal FJ. Angiotensin II stimulates the autocrine production of transforming growth factor-beta 1 in adult rat cardiac fibroblasts. J Mol Cell Cardiol. 1995;27:2347-57. [35] Yokoyama T, Sekiguchi K, Tanaka T, Tomaru K, Arai M, Suzuki T, et al. Angiotensin II and mechanical stretch induce production of tumor necrosis factor in cardiac fibroblasts. Am J Physiol. 1999;276:H1968-76. [36] Ancey C, Corbi P, Froger J, Delwail A, Wijdenes J, Gascan H, et al. Secretion of IL- 6, IL-11 and LIF by human cardiomyocytes in primary culture. Cytokine. 2002;18:199-205. 41 [37] Turner NA, Das A, O'Regan DJ, Ball SG, Porter KE. Human cardiac fibroblasts express ICAM-1, E-selectin and CXC chemokines in response to proinflammatory cytokine stimulation. Int J Biochem Cell Biol. 2011;43:1450-8. [38] Wang JH, Lin JS. Cell traction force and measurement methods. Biomechanics and Modeling in Mechanobiology. 2007;6:361-71. [39] Bell E, Ivarsson B, Merrill C. Production of a tissue-like structure by contraction of collagen lattices by human fibroblasts of different proliferative potential in vitro. Proc Natl Acad Sci U S A. 1979;76:1274-8. [40] Nusgens B, Merrill C, Lapiere C, Bell E. Collagen biosynthesis by cells in a tissue equivalent matrix in vitro. Collagen and Related Research. 1984;4:351-63. [41] Coulomb B, Dubertet L, Merrill C, Touraine R, Bell E. The collagen lattice: a model for studying the physiology, biosynthetic function and pharmacology of the skin. The British Journal of Dermatology. 1984; 111 Suppl 27:83-7. [42] Grinnell F. Fibroblasts, myofibroblasts, and wound contraction. J Cell Biol. 1994;124:401-4. [43] Grinnell F. Fibroblast-collagen-matrix contraction: growth-factor signalling and mechanical loading. Trends in Cell Biology. 2000;10:362-5. [44] Grinnell F. Fibroblast biology in three-dimensional collagen matrices. Trends in Cell Biology. 2003;13:264-9. [45] Rhee S, Grinnell F. Fibroblast mechanics in 3D collagen matrices. Advanced Drug Delivery Reviews. 2007;59:1299-305. [46] Grinnell F. Fibroblast mechanics in three-dimensional collagen matrices. Journal of Bodywork and Movement Therapies. 2008;12:191-3. [47] Harris AK, Wild P, Stopak D. Silicone rubber substrata: a new wrinkle in the study of cell locomotion. Science. 1980;208:177-9. [48] Beningo KA, Wang YL. Flexible substrata for the detection of cellular traction forces. Trends in Cell Biology. 2002;12:79-84. [49] Pelham RJ, Jr., Wang Y. Cell locomotion and focal adhesions are regulated by substrate flexibility. Proc Natl Acad Sci U S A. 1997;94:13661-5. [50] du Roure O, Saez A, Buguin A, Austin RH, Chavrier P, Silberzan P, et al. Force mapping in epithelial cell migration. Proc Natl Acad Sci U S A. 2005;102:2390-5. [51] MacKenna DA, Omens JH, McCulloch AD, Covell JW. Contribution of collagen matrix to passive left ventricular mechanics in isolated rat hearts. Am J Physiol. 1994;266:H1007-18. 42 [52] Sasaki N, Odajima S. Stress-strain curve and Young's modulus of a collagen molecule as determined by the X-ray diffraction technique. J Biomech. 1996;29:655- 8. [53] Christiansen DL, Huang EK, Silver FH. Assembly of type I collagen: fusion of fibril subunits and the influence of fibril diameter on mechanical properties. Matrix Biology : Journal of the International Society for Matrix Biology. 2000;19:409-20. [54] Robinson TF, Geraci MA, Sonnenblick EH, Factor SM. Coiled perimysial fibers of papillary muscle in rat heart: morphology, distribution, and changes in configuration. Circ Res. 1988;63:577-92. [55] MacKenna DA, Omens JH, Covell JW. Left ventricular perimysial collagen fibers uncoil rather than stretch during diastolic filling. Basic Res Cardiol. 1996;91:111-22. [56] MacKenna DA, Vaplon SM, McCulloch AD. Microstructural model of perimysial collagen fibers for resting myocardial mechanics during ventricular filling. Am J Physiol. 1997;273:H1576-86. [57] Engelmayr GC, Jr., Cheng M, Bettinger CJ, Borenstein JT, Langer R, Freed LE. Accordion-like honeycombs for tissue engineering of cardiac anisotropy. Nat Mater. 2008;7:1003-10. [58] Lloyd-Jones D, Adams R, Carnethon M, De Simone G, Ferguson TB, Flegal K, et al. Heart disease and stroke statistics--2009 update: a report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Circulation. 2009;119:480-6. [59] Cutler MJ, Rosenbaum DS, Dunlap ME. Structural and electrical remodeling as therapeutic targets in heart failure. J Electrocardiol. 2007;40:S1-7. [60] Brower GL, Gardner JD, Forman MF, Murray DB, Voloshenyuk T, Levick SP, et al. The relationship between myocardial extracellular matrix remodeling and ventricular function. Eur J Cardiothorac Surg. 2006;30:604-10. [61] Connelly CM, Vogel WM, Wiegner AW, Osmers EL, Bing OH, Kloner RA, et al. Effects of reperfusion after coronary artery occlusion on post-infarction scar tissue. Circ Res. 1985;57:562-77. [62] Gorman RC, Jackson BM, Burdick JA, Gorman JH. Infarct restraint to limit adverse ventricular remodeling. J Cardiovasc Transl Res. 2011;4:73-81. [63] Pfeffer MA, Braunwald E. Ventricular remodeling after myocardial infarction. Experimental observations and clinical implications. Circulation. 1990;81:1161-72. [64] Udelson JE, Konstam MA. Relation between left ventricular remodeling and clinical outcomes in heart failure patients with left ventricular systolic dysfunction. J Card Fail. 2002;8:S465-71. 43 [65] Holmes JW, Borg TK, Covell JW. Structure and mechanics of healing myocardial infarcts. Annu Rev Biomed Eng. 2005;7:223-53. [66] Forrester JS, White AJ, Matsushita S, Chakravarty T, Makkar RR. New paradigms of myocardial regeneration post-infarction: tissue preservation, cell environment, and pluripotent cell sources. JACC Cardiovasc Interv. 2009;2:1-8. [67] Sunagawa K, Maughan WL, Sagawa K. Effect of regional ischemia on the left ventricular end-systolic pressure-volume relationship of isolated canine hearts. Circ Res. 1983;52:170-8. [68] Gupta KB, Ratcliffe MB, Fallert MA, Edmunds LH, Jr., Bogen DK. Changes in passive mechanical stiffness of myocardial tissue with aneurysm formation. Circulation. 1994;89:2315-26. [69] Wilson EM, Moainie SL, Baskin JM, Lowry AS, Deschamps AM, Mukherjee R, et al. Region- and type-specific induction of matrix metalloproteinases in post-myocardial infarction remodeling. Circulation. 2003;107:2857-63. [70] Takahashi S, Barry AC, Factor SM. Collagen degradation in ischaemic rat hearts. Biochem J. 1990;265:233-41. [71] Janicki JS, Brower GL, Henegar JR, Wang L. Ventricular remodeling in heart failure: the role of myocardial collagen. Adv Exp Med Biol. 1995;382:239-45. [72] Cleutjens JP, Verluyten MJ, Smiths JF, Daemen MJ. Collagen remodeling after myocardial infarction in the rat heart. Am J Pathol. 1995;147:325-38. [73] Norton GR, Tsotetsi J, Trifunovic B, Hartford C, Candy GP, Woodiwiss AJ. Myocardial stiffness is attributed to alterations in cross-linked collagen rather than total collagen or phenotypes in spontaneously hypertensive rats. Circulation. 1997;96:1991-8. [74] Woodiwiss AJ, Tsotetsi OJ, Sprott S, Lancaster EJ, Mela T, Chung ES, et al. Reduction in myocardial collagen cross-linking parallels left ventricular dilatation in rat models of systolic chamber dysfunction. Circulation. 2001;103:155-60. [75] Sun Y, Weber KT. Infarct scar: a dynamic tissue. Cardiovasc Res. 2000;46:250-6. [76] Passier R, van Laake LW, Mummery CL. Stem-cell-based therapy and lessons from the heart. Nature. 2008;453:322-9. [77] Sawa Y, Miyagawa S, Sakaguchi T, Fujita T, Matsuyama A, Saito A, et al. Tissue engineered myoblast sheets improved cardiac function sufficiently to discontinue LVAS in a patient with DCM: report of a case. Surg Today. 2012;42:181-4. [78] Menasche P. Stem cell therapy for chronic heart failure: lessons from a 15-year experience. C R Biol. 2011;334:489-96. 44 [79] Godier-Furnemont AF, Martens TP, Koeckert MS, Wan L, Parks J, Arai K, et al. Composite scaffold provides a cell delivery platform for cardiovascular repair. Proc Natl Acad Sci U S A. 2011;108:7974-9. [80] Lian WS, Cheng WT, Cheng CC, Hsiao FS, Chen JJ, Cheng CF, et al. In vivo therapy of myocardial infarction with mesenchymal stem cells modified with prostaglandin I synthase gene improves cardiac performance in mice. Life Sci. 2011;88:455-64. [81] Gnecchi M, He H, Liang OD, Melo LG, Morello F, Mu H, et al. Paracrine action accounts for marked protection of ischemic heart by Akt-modified mesenchymal stem cells. Nat Med. 2005;11:367-8. [82] Moreno-Gonzalez A, Korte FS, Dai J, Chen K, Ho B, Reinecke H, et al. Cell therapy enhances function of remote non-infarcted myocardium. J Mol Cell Cardiol. 2009;47:603-13. [83] Chimenti I, Smith RR, Li TS, Gerstenblith G, Messina E, Giacomello A, et al. Relative roles of direct regeneration versus paracrine effects of human cardiosphere-derived cells transplanted into infarcted mice. Circ Res. 2010;106:971-80. [84] Hamdi H, Planat-Benard V, Bel A, Puymirat E, Geha R, Pidial L, et al. Epicardial adipose stem cell sheets results in greater post-infarction survival than intramyocardial injections. Cardiovasc Res. 2010;93:483-91. [85] Bel A, Planat-Bernard V, Saito A, Bonnevie L, Bellamy V, Sabbah L, et al. Composite cell sheets: a further step toward safe and effective myocardial regeneration by cardiac progenitors derived from embryonic stem cells. Circulation. 2010;122:S118-23. [86] Vunjak-Novakovic G, Tandon N, Godier A, Maidhof R, Marsano A, Martens T, et al. Challenges in Cardiac Tissue Engineering. Tissue Engineering Part B: Reviews. 2009:140-568. [87] Hollister S, Wang Y, Wang G, Chen L, Li H, Yin T, et al. Scaffold engineering: a bridge to where? Biofabrication. 2009;1:012001-14. [88] Eschenhagen T, Zimmermann WH. Engineering myocardial tissue. Circ Res. 2005;97:1220-31. [89] Naito H, Melnychenko I, Didie M, Schneiderbanger K, Schubert P, Rosenkranz S, et al. Optimizing engineered heart tissue for therapeutic applications as surrogate heart muscle. Circulation. 2006;114:I72-8. [90] Zimmermann WH, Cesnjevar R. Cardiac tissue engineering: implications for pediatric heart surgery. Pediatr Cardiol. 2009;30:716-23. 45 [91] Lasher RA, Pahnke AQ, Johnson JM, Sachse FB, Hitchcock RW. Electrical stimulation directs engineered cardiac tissue to an age-matched native phenotype. J Tissue Eng. 2012;3:1-15. [92] Hollister SJ. Porous scaffold design for tissue engineering. Nat Mater. 2005;4:518- 24. [93] Huang Z, Zhang Y, Kotaki M, Ramakrishna S. A review on polymer nanofibers by electrospinning and their applications in nanocomposites. Composites Science and Technology. 2003;63:2223-53. [94] Doshi J, Reneker D. Electrospinning process and applications of electrospun fibers. Journal of Electrostatics. 1995;35:151-60. [95] Reneker D, Yarin A. Electrospinning jets and polymer nanofibers. Polymer. 2008;49:2387-425. [96] Baker BM, Gee AO, Metter RB, Nathan AS, Marklein RA, Burdick JA, et al. The potential to improve cell infiltration in composite fiber-aligned electrospun scaffolds by the selective removal of sacrificial fibers. Biomaterials. 2008;29:2348-58. [97] Rowlands AS, Lim SA, Martin D, Cooper-White JJ. Polyurethane/poly(lactic-co-glycolic) acid composite scaffolds fabricated by thermally induced phase separation. Biomaterials. 2007;28:2109-21. [98] Fromstein JD, Zandstra PW, Alperin C, Rockwood D, Rabolt JF, Woodhouse KA. Seeding bioreactor-produced embryonic stem cell-derived cardiomyocytes on different porous, degradable, polyurethane scaffolds reveals the effect of scaffold architecture on cell morphology. Tissue Eng Part A. 2008;14:369-78. [99] Khorasani MT, Shorgashti S. Fabrication of microporous polyurethane by spray phase inversion method as small diameter vascular grafts material. J Biomed Mater Res A. 2006;77:253-60. [100] Papenburg B, Vogelaar L, Bolhuis-Versteeg L, Lammertink R, Stamatialis D, Wessling M. One-step fabrication of porous micropatterned scaffolds to control cell behavior. Biomaterials. 2007;28:1998-2009. [101] Papenburg BJ, Liu J, Higuera GA, Barradas AM, de Boer J, van Blitterswijk CA, et al. Development and analysis of multi-layer scaffolds for tissue engineering. Biomaterials. 2009;30:6228-39. [102] Guido S, Tranquillo RT. A methodology for the systematic and quantitative study of cell contact guidance in oriented collagen gels. Correlation of fibroblast orientation and gel birefringence. J Cell Sci. 1993; 105 ( Pt 2):317-31. 46 [103] Ayres CE, Jha BS, Meredith H, Bowman JR, Bowlin GL, Henderson SC, et al. Measuring fiber alignment in electrospun scaffolds: a user's guide to the 2D fast Fourier transform approach. J Biomater Sci Polym Ed. 2008;19:603-21. [104] Marsano A, Maidhof R, Wan LQ, Wang Y, Gao J, Tandon N, et al. Scaffold stiffness affects the contractile function of three-dimensional engineered cardiac constructs. Biotechnol Prog. 2010;26:1382-90. [105] Zhang C, Wen X, Vyavahare NR, Boland T. Synthesis and characterization of biodegradable elastomeric polyurethane scaffolds fabricated by the inkjet technique. Biomaterials. 2008;29:3781-91. [106] Moroni L, de Wijn JR, van Blitterswijk CA. 3D fiber-deposited scaffolds for tissue engineering: influence of pores geometry and architecture on dynamic mechanical properties. Biomaterials. 2006;27:974-85. [107] Zhang T, Wan LQ, Xiong Z, Marsano A, Maidhof R, Park M, et al. Channelled scaffolds for engineering myocardium with mechanical stimulation. J Tissue Eng Regen Med. 2011;6:748-56. [108] Maidhof R, Marsano A, Lee EJ, Vunjak-Novakovic G. Perfusion seeding of channeled elastomeric scaffolds with myocytes and endothelial cells for cardiac tissue engineering. Biotechnol Prog. 2010;26:565-72. [109] Kaihara S, Borenstein J, Koka R, Lalan S, Ochoa E, Ravens M, et al. Silicon micromachining to tissue engineer branched vascular channels for liver fabrication. Tissue Engineering. 2000;6:105-17. [110] Papenburg BJ, Bolhuis-Versteeg LA, Grijpma DW, Feijen J, Wessling M, Stamatialis D. A Facile Method to Fabricate Poly(L-lactide) Nano-fibrous Morphologies by Phase Inversion. Acta Biomater. 2010;6:2477-83. [111] Barron V, Lyons E, Stenson-Cox C, McHugh PE, Pandit A. Bioreactors for cardiovascular cell and tissue growth: a review. Ann Biomed Eng. 2003;31:1017-30. [112] Freed LE, Vunjak-Novakovic G. Tissue engineering bioreactors. Principles of Tissue Engineering. 2000;2:143-56. [113] Thevenot P, Nair A, Dey J, Yang J, Tang L. Method to Analyze Three­Dimensional Cell Distribution and Infiltration in Degradable Scaffolds. Tissue Engineering Part A. 2008;14:319-31. [114] Radisic M, Marsano A, Maidhof R, Wang Y, Vunjak-Novakovic G. Cardiac tissue engineering using perfusion bioreactor systems. Nat Protoc. 2008;3:719-38. [115] Radisic M, Deen W, Langer R, Vunjak-Novakovic G. Mathematical model of oxygen distribution in engineered cardiac tissue with parallel channel array perfused 47 with culture medium containing oxygen carriers. American Journal of Physiology- Heart and Circulatory Physiology. 2005;288:1278-89. [116] Carrier RL, Rupnick M, Langer R, Schoen FJ, Freed LE, Vunjak-Novakovic G. Perfusion improves tissue architecture of engineered cardiac muscle. Tissue engineering. 2002;8:175-88. [117] Liu C, Abedian R, Meister R, Haasper C, Hurschler C, Krettek C, et al. Influence of perfusion and compression on the proliferation and differentiation of bone mesenchymal stromal cells seeded on polyurethane scaffolds. Biomaterials. 2011;33:1052-64. [118] Maidhof R, Tandon N, Lee EJ, Luo J, Duan Y, Yeager K, et al. Biomimetic perfusion and electrical stimulation applied in concert improved the assembly of engineered cardiac tissue. J Tissue Eng Regen Med. 2012;6:e12-23. [119] McCoy RJ, Jungreuthmayer C, O'Brien FJ. Influence of flow rate and scaffold pore size on cell behavior during mechanical stimulation in a flow perfusion bioreactor. Biotechnol Bioeng. 2012;109:1583-94. [120] Sekine H, Shimizu T, Okano T. Myocardial tissue engineering: toward a bioartificial pump. Cell Tissue Res. 2012;347:775-82. [121] Zimmermann W, Eschenhagen T. Cardiac tissue engineering for replacement therapy. Heart Failure Reviews. 2003;8:259-69. [122] Rouwkema J, Gibbs S, Lutolf MP, Martin I, Vunjak-Novakovic G, Malda J. In vitro platforms for tissue engineering: implications for basic research and clinical translation. J Tissue Eng Regen Med. 2011;5:e164-7. [123] Bursac N, Papadaki M, Cohen RJ, Schoen FJ, Eisenberg SR, Carrier R, et al. Cardiac muscle tissue engineering: toward an in vitro model for electrophysiological studies. Am J Physiol. 1999;277:H433-44. [124] Asakawa N, Shimizu T, Tsuda Y, Sekiya S, Sasagawa T, Yamato M, et al. Pre­vascularization of in vitro three-dimensional tissues created by cell sheet engineering. Biomaterials. 2010;31:3903-9. [125] Shimizu T, Sekine H, Yang J, Isoi Y, Yamato M, Kikuchi A, et al. Polysurgery of cell sheet grafts overcomes diffusion limits to produce thick, vascularized myocardial tissues. FASEB J. 2006;20:708-10. [126] Nichol JW, Engelmayr GC, Jr., Cheng M, Freed LE. Co-culture induces alignment in engineered cardiac constructs via MMP-2 expression. Biochem Biophys Res Commun. 2008;373:360-5. 48 [127] Pillekamp F, Reppel M, Dinkelacker V, Duan Y, Jazmati N, Bloch W, et al. Establishment and characterization of a mouse embryonic heart slice preparation. Cell Physiol Biochem. 2005;16:127-32. [128] Pillekamp F, Halbach M, Reppel M, Pfannkuche K, Nazzal R, Nguemo F, et al. Physiological differences between transplanted and host tissue cause functional decoupling after in vitro transplantation of human embryonic stem cell-derived cardiomyocytes. Cell Physiol Biochem. 2009;23:65-74. [129] Habeler W, Peschanski M, Monville C. Organotypic heart slices for cell transplantation and physiological studies. Organogenesis. 2009;5:62-6. [130] Habeler W, Pouillot S, Plancheron A, Puceat M, Peschanski M, Monville C. An in vitro beating heart model for long-term assessment of experimental therapeutics. Cardiovasc Res. 2009;81:253-9. CHAPTER 2 DESIGN OF SEMIAUTOMATED TOOLING FOR FABRICATION OF POROUS POLYURETHANE FOAMS 2.1 Introduction Through all of the work described in this dissertation, substrates were fabricated using custom, semiautomated tooling. Development of this tooling was performed prior to the research described in Chapters 3-5. The fundamental purpose of this tooling was to enable production of biomaterials to be used as three-dimensional substrata for cell culture. Beyond this requirement, consideration was given to reproducibility of output materials independent of the operator, and a high degree of flexibility to adapt this process to the unique requirements of small batch production. A combination of custom, and off-the- shelf components were used to build the spray tooling. The primary components of the spray assembly are a computer, off-the-shelf syringe pump, CNC robot (Fig 2.1 MaxNC, Gilbert AZ), the nozzle assembly (Fig 2.2), and the spray surface assembly (Fig 2.3). The spray surface assembly is comprised of a spray plate (Fig 2.4) on top of a waste reservoir (Fig 2.5). A silicone substrate (Fig 2.6) was used in Chapter 4 for sprayed micromolding of the material. Polyester frames (Fig 2.7) were used in the spray 50 surface assembly to facilitate handle-ability of the material after production. A mask (Fig 2.8) was aligned using alignment pins anchored to t-nuts on the surface of the CNC (Fig 2.9). At the end of this chapter, the engineering assembly drawings, part drawings, and bill of materials are included to describe the geometric and mechanical configuration of this device. Validation of the ability of this system to produce porous polyurethane substrates for use as cell scaffolding in cell culture is included in Chapter 3 and 4. 2.2 Spray Parameters The important inputs to the spray process include the volume of polymer solution on the deposited on surface, the rate of deposition, the spray pressure, the spray pattern, and curing conditions. These parameters and their influence on the overall process are described in detail in the following sections. 2.2.1 Polymer Deposition The volume of polymer deposited on the spray surface is determined by the concentration of the polymer solution and the volume of this solution that is pumped during spray. The polymer deposition was normalized to the area of the surface, yielding a coverage parameter measured in g/in2. For this work, a coverage value of 0.65 g/in2 was found to yield acceptable substrates. The coverage parameter can be altered in a custom Matlab program that will automatically calculate the corresponding pump settings. This value is limited on the upper side by accumulation of solvent that is trapped within the precipitated material, resulting in resolubilization of the deposited polymer, and loss of control over the porous architecture, and on the lower limit by the handle-ability of the material after deposition. 2.2.2 Rate of Deposition The rate of deposition is a function both of flow rate from the pump, and speed of the CNC while it moves through the spray pattern (discussed below). Precipitation of the polymer appears to happen relatively quickly with respect to the deposition rate, therefore, we used deposition rates that corresponded to the maximum linear travel speed of the CNC. The deposition rate is a user input parameter that can be altered for more sensitive processes requiring more time for precipitation or curing during fabrication. 2.2.3 Spray Pressure Air pressure is required to aerosolize the polymer and nonsolvent solution. However, upon precipitation, the polyurethane materials fabricated in this work were very fragile and easily susceptible to plastic deformation. Artifacts in substrate geometry resulting from air flow induced deformation were minimized through minimizing the air pressure, and maximizing the working distance between the nozzles (see Fig 2.2) and the spray surface (Fig 2.3, see below for more details on spray pattern). A minimum of 10 PSI was required for adequate dispersal of the polymer and nonsolvent. Depending on the viscosity and surface tension of the solutions used, this value may potentially be reduced even lower. Regardless of the solutions used, air pressure should be minimized to decrease deformation of the substrate geometry by air flow. 51 2.2.4 Spray Pattern The spray pattern is determined by the nozzle itself, and the working distance between the nozzle and the deposition surface. Initial spray experiments using dye loaded solutions sprayed onto absorbent paper demonstrated a linear increase (R2=0.9995) in spray diameter as a function of the working distance from 7.6 cm to 17.8 cm. The maximum working distance with the spray plate (Fig 2.5) and waste reservoir (Fig 2.4) loaded into the assembly was 14.0 cm. A working distance of 13 cm was selected and calculated to produce a spray diameter of 5.1 cm at the surface. Due to the side-by-side arrangement of the nonsolvent and solvent spray nozzles (see Fig 2.2), nonuniform border regions exist around the border of the sprayed material. These regions, depending on their location receive either excess polymer solution, or excess nonsolvent. To ensure consistent material properties in the final biomaterial, a mask (Fig 2.8) was designed and incorporated into the spray surface assembly to separate the consistent center regions from the nonhomogenous border regions. The travel distance in the X and Y axis of the CNC robot, in conjunction with the offset of the nozzles in the cross head and working distance provide a maximum consistent spray region of 13.3 cm x 8.9 cm. Beyond the available space to create a uniform spray deposition, the spray pattern itself was controlled. The CNC traversed the spray in a serpentine pattern. The pitch of the each traverse is a user input parameter, limited on the low end by the ability of an off-the- shelf syringe pump to consistently deliver low volumes, and on the high end by the diameter of the spray pattern by the nozzle. For this research, a 17 mm step over was found to provide good coverage, while working within an optimal flow rate for the pump. In addition to the pitch, the number of passes is a variable user input. We used two coats 52 to create substrates that were approximately 200 p,m thick in order to return the spray nozzles to the start position at the end of a run, although fewer or more coats seemed to work equally well. 2.2.5 Curing Conditions After spray, the majority of the solvent and nonsolvent are collected in the waste reservoir. Removal of residual solvent is necessary for curing the material. Removal of solvent can be accomplished through evaporation or through diffusion when submerged in a solvent-miscible nosolvent. For all the polyurethane material used in this research, an aqueous nonsolvent was used. It was observed that relative humidity impacted the curing process when solvent was allowed to evaporate. To control for variations in relative humidity, samples were cured in water to remove residual solvent. Another important environmental parameter influencing the rate of solvent removal is ambient temperature. It was observed that low temperatures appeared to decrease the rate of solvent removal, resulting in slower kinetics of precipitation, and generally, higher porosity. All experiments were performed at room temperature, and relied on nonsolvent composition to influence the kinetics of precipitation, but future applications could utilize temperature as well as nonsolvent composition to influence material architecture using this system. 2.3 Tooling Applications In the research described in this dissertation, the material used with system was an elastomeric poly (ether-urethane), however, while this was the only material used in this 53 54 research, the tooling described in this chapter has also been used to deposit drug loaded poly (lactic-glycolic acid) (PLGA) coatings, and fabricate composite polyurethane /collagen blends. Future adaption of the spray parameters described in this section can be used to fabricate a wide variety of porous biomaterial sheets using the tooling described here. REV DESCRIPTION DWN DATE 01 INITIAL RELEASE JK 2/27/13 Item No. Description QTY. 1 CNC C om p u te r Mill (MAXNC, P/N MAXNC1 5 1 2 Spray nozzle assembly - See Page 2 1 3 Spray substrate - See P a g e 3 1 ► > Fig 2.1 Spray assembly UNIVERISTY OF UTAH 20 S 2030 E Rm 108 Salt Lake City, 84112 TITLE: Base Assembly SIZE DWG. NO. A N /A REV 01 PART NO: N /A SPEC: N /A MODEL FILE: Base Assembly SCALE: 1:4 SHEET 1 OF 9 2.000 2X 0 .107 THRU 6-32 UNC THRU Item No. Description Q u a n tity 1 6-32 McM aster-Carr P/N 91735A 144 2 2 Iw a ta Eclipse P/N HP-BCS 2 3 Nozzle M o u n t 6061 Aluminum 1 0 .4 5 0 THRU TITLE: UNIVERISTY OF UTAH 20 S 2030 E Rm 108 Salt Lake City, 84112 Spray Nozzle Assembly SIZE DWG. NO. A N/A REV 01 Fig. 2.2 Nozzle assembly PART NO: N/A SPEC: N/A MODEL FILE:Spray nozzle assembly SCALE:T7l I SHEET 2 OF 9 Item N umber Description Quanity 1 Aluminum T-Nut - see p a g e 9 2 2 Aluminum A lig nm e n t Pin - see p a g e 9 2 3 Polyethylene Waste Resevior - see p a g e 4 1 4 Aluminum Spray Plate - see p a g e 5 1 5 VST50 Platinum Cure Silicone - see p a g e 6 1 6 Polyester Frame - see p a g e 7 1 7 Aluminum Mask - see p a g e 8 1 UNIVERISTY OF UTAH 20 S 2030 E Rm 108 Salt Lake City, 84112 TITLE: Spray Substrate AD SIZE DWG. NO. A N /A REV 01 Fig 2.3 Spray surface assembly PART NO: N /A SPEC: N/A MODEL FILE: Spray assembly SCALE: 1:4 SHEET 3 OF 9 .500 3.500 7.000 .063 SECTION B-B SCALE 1 :2 NOTE: ALL SLOTS ARE .25 WIDE AND THRU ALL Fig 2.4 Spray plate UNIVERISTY OF UTAH 20 S 2030 E Rm 108 Salt Lake City, 84112 TITLE: Spray Plate SIZE DWG. NO. REV A N /A 01 PART NO :N /A SPEC:N/A MODEL FILE: Spray Plate 1 SCALE: 1:2 SHEET 5 OF 9 1.000 7.000 -*■1.750 .375 5.500 SECTION A-A SCALE 1 : 2 3.500 <!> 12.000 .750 4X R .250 - .750 s \ THRU ALL 2X 0 .250 Fig 2.5 Waste reservoir UNIVERISTY OF UTAH 20 S 2030 E Rm 108 Salt Lake City, 84112 TITLE: Waste Reservoir SIZE DWG. NO. A N /A REV 01 PART NO: N /A SPEC: N/A MODEL FILE: w a s te reservoir SCALE: 1:2 SHEET 4 OF 9 ,076 100X .015 .014 101X .020 DETAIL C SCALE 4 : 1 UNIVERISTY OF UTAH 20 S 2030 E Rm 108 Salt Lake City. 84112 TITLE: Spray Substrate SIZE DWG. NO. A N /A REV 01 PART NO: N /A SPEC: N/A Fig. 2.6 Spray substrate M.QDEL FILE: Silicone p a tte rn 500w 875p 350d SCALE: 1:1 SHEET 6 OF 9 NOTE: THICKNESS IS .014 Fig 2.7 Scaffold frame UNIVERISTY OF UTAH 20 S 2030 E Rm 108 Salt Lake City. 84112 TITLE: Frame SIZE DWG. NO. REV A N /A 01 PART NO: N /A SPEC: N/A MODEL FILE: Large frame SCALE: 1:1 SHEET 7 OF 9 NOTE: THICKNESS IS .0625 Fig 2.8 Mask UNIVERISTY OF UTAH 20 S 2030 E Rm 108 Salt Lake City, 84112 TITLE: Mask SIZE DWG. NO. A N /A PART NO: N /A SPEC: N/A MODEL FILE: mask2 SCALE: 1:2 REV 01 SHEET 8 OF 9 r 0.250 0 .1 2 5 - * " 0 .201 THRU ALL 1/4-20 UNC THRU ALL T-Nut .120 T Fig 2.9 Alignment pin and t-nut .300 .110 UNIVERISTY OF UTAH 20 S 2030 E Rm 108 Salt Lake City, 84112 TITLE: A lig nm e n t Pin a n d T-Nut SIZE DWG. NO. A N /A PART NO: n / a MODEL FILE: SCALE: 3:2 REV 01 SPEC. n /A a lig nm e n t pin SHEET 9 OF 9 On CHAPTER 3 THE MECHANICALLY ENHANCED PHASE SEPARATION OF SPRAYED POLYURETHANE SCAFFOLDS AND THEIR EFFECT ON THE ALIGNMENT OF FIBROBLASTS Reprinted from Biomaterials, Vol 31 Issue 6, Kennedy, J. P. McCandless, S. P. Lasher, R. A. Hitchcock, R. W, The mechanically enhanced phase separation of sprayed polyurethane scaffolds and their effect on the alignment of fibroblasts 1126-1132, (2010), with permission from Elsevier 65 Biomaterials 31 (2010) 1126-1132 ELSEVIER Contents lists available a t ScienceDirect Biomaterials jo u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / b i o m a t e r i a l s The mechanically enhanced phase separation of sprayed polyurethane scaffolds and their effect on the alignment of fibroblasts James P. Kennedy, Sean P. McCandless, Richard A. Lasher, Robert W. Hitchcock* Department o f Bioengineering, University of Utah. 20 S 2030 E, Rm 108 Salt Lake City. UT 84112, USA A R T I C L E I N F O Article history: Received 3 August 2009 Accepted 9 October 2009 Available online 30 October 2009 Keywords: Scaffold Cell alignment Mechanical properties Porosity Polyurethane Tissue engineering A B S T R A C T This p a p e r r e p o r ts a m e th o d to fab rica te an iso tro p ic scaffolds o f tu n ab le p o ro sity an d mech an ical p ro p e rtie s . Scaffolds w e re fab ricated usin g a c om p u te r c o n tro lled sp ray ed p h a s e s e p a ra tio n tech n iq u e. Following fabrication, th e sh e e ts w e re e lo n g a te d 0, 35 o r 70% o f th e ir original len g th to in d u ce varying d e g re e s o f scaffold a lig nm en t an d aniso tro p y . The n o n so lv e n t u sed in th e p h a s e s e p a ra tio n w a s sh ow n to affect p o ro sity a n d th e e la stic m o d u lu s. Mo u se em b ry o NIH-3T3 fib ro b lasts w e re c u ltu red o n the scaffolds to in v estig a te cell r e sp o n s e to th e a n iso tro p y o f th e scaffold. A 2D FFT m e th o d w a s u sed to q u a n tify c ellu la r a lig nm en t. Cells w e re sh ow n to align th em s e lv e s w ith th e scaffold. This sh e et-lik e scaffold m ate ria l c a n be u s e d in single plys o r c an be lam in a te d to fo rm p o ro u s 3D com p o s ite scaffolds. P u b lish ed by Elsevier Ltd. 1. Introduction Tissue engineering and regenerative medicine have the poten­tial to develop novel biosynthetic materials for improved treat­ment, maintenance, and regeneration of diseased or damaged tissue. Development of materials th a t utilize tissue engineering strategies requires d esign considerations such as cell type, seeding and attachment, as well as molecular signals, and macromolecular matrix in order to develop constructs th a t improve or replace function of natural tissue [1-3]. Many types of engineered tissue rely on a provisional or permanent scaffold to generate a th re e ­dimensional framework for cell atta chm en t and tissue organiza­tion. Both n atural and synthetic scaffold materials are used in tissue engineering [4,5]. Synthetically derived cell scaffolds can be permanent or degradable and facilitate expression and organiza­tion of the extracellular matrix (ECM). Architectural cues in these scaffolds have been shown to affect the morphology, organization, and phenotypic expression of cells in vitro [6-8]. In order to develop effective implants, the tissue engineer needs to be able to specify and tune the scaffold's morphological features for different applications. In addition, scaffold architecture must be designed to provide cues for cellular organization and form th e basis for engineered tissue constructs that mimic tissue specific organiza­tion and physical properties. : Corresponding author. E-mail address: [e-mail] (R.W. Hitchcock). Cardiac tissue is an example of highly structured tissue th a t relies on cellular organization for its function [9,10]. Cell scaffold materials can help cardiac tissue development by: (1) providing cues th a t induce alignment of cardiac myocytes, (2) allowing sufficient n u trien t and cell infiltration n ecessary to form a 3D tissue construct, (3) modulating the cell type distribution of co-cultures, and (4) mimicking anisotropic mechanical stiffness of the heart. Scaffolds used for cardiac tissue engineering applications require development of design specifications th a t include scaffold align­ment, structure, porosity, and stiffness all of w hich will influence cellular development, overall tissue organization, and bioreactor integration [11-13]. Various methods have been employed to fabricate scaffolds of varying porosity and anisotropy. For example, microfabrication techniques have been used to fabricate scaffolds with aligned structure [14,15]. Electrospinning methods have been employed with post process elongation to produce anisotropic fibrous scaf­fold architecture [16]. In order to create scaffolds th a t allow adequate nutrient and oxygen diffusion, methods such as 3D fiber deposition, sacrificial fiber electrospinning, and phase separation have been utilized to generate scaffolds of controlled porosity [17-20]. Spray phase separation (SPS) is method for creating scaf­folds with control over alignment, porosity, and stiffness; however this method has not been directly applied for cardiac tissue scaf­folds [19]. SPS fabricated scaffolds are produced using a method that simultaneously sprays a polymer solution and a nonsolvent onto the surface. The nonsolvent mixes with the solvent and the 0142-9612/$ - see front matter Published by Elsevier Ltd. doi: 10.1016/j .biomaterial s.2009.10.024 66 J.P. Kennedy et al. / Biomaterials 31 (2010) 1126-1132 1127 polymer causing the polymer to precipitate. Some groups have used SPS methods to fabricate materials of varying porosity for drug delivery devices [21 ] and vascular graft materials [19,22]. SPS fabrication is a promising method for controlling scaffolds prop­erties such as alignment, porosity, stiffness, and anisotropy, which are key features for directing cellular development, overall tissue organization, and bioreactor integration [11-13]. We hypothesized th a t post spray elongation o f SPS scaffolds would generate scaffold microstructure alignment and in turn induce cellular alignment. Furthermore, we hypothesized th a t the nonsolvent (NS) EtOH concentration would affect scaffold porosity. In addition, we hypothesize tha t this material may be laminated into thick scaffold material. Here we report an SPS method for fabricating polyurethane scaffolds for tissue engineering applications. 2. Materials and methods 2.1. Scaffold fabrication Scaffolds w ere fabricated using an SPS m ethod. A 4% polyurethane solution was prepared by dissolving Tecoflex SG80 polyether polyurethane (Lubrizol Advanced Materials Inc. Cleveland, OH) in dimethylacetamide (DMAc) (Sigma Aldrich, St. Louis, MO). Polyether polyurethane was selected because of its known biocompat­ibility, ease of processing, and extensive use in the past for scaffold materials [23-25]. The polymer solution was sealed in glass storage containers prior to use and used immediately after opening to ensure that minimal solvent was lost through evaporation. Deionized water (referred to as 0% ethanol), 50% EtOH, and 70% EtOH solutions were used as the nonsolvent for precipitation. Spray nozzles (Excel ES4, Porter Cable, Jackson, TN) were mounted onto a custom crossbar attached to the spindle head of a computer-controlled desktop milling machine (MaxNC 12, MAXNC, Gilbert AZ) to provide X-Y-Z control o f spray pattern. Custom G-code was used to move the spray head and control the spray pattern and spray time. The spray nozzles traversed the substrate at a distance of 20 cm and a speed o f 0.85 cm/s in a serpentine pattern yielding a total spray time of 2.5 min. Aluminum frames were designed to facilitate manipulation and mechanical alignment of the scaffold material during and after spraying. The frames were fabricated from 0.32 cm thick 6016-T6 aluminum (McMAster-Carr, Princeton NJ) with an outside dimension of 5.3 cm x 3.7 cm and an inside dimension of 2.86 cm x 2.54 cm. Stainless steel hypodermic tubing sliders (Small Parts Inc., Miramar, FL) were assembled and attached to frames with UV cure adhesive to facilitate elongation and constrain the maximum elongation to 3.86 cm (35% elongation) or 4.86 cm (70% elongation). Frames without sliders were used to make scaffolds with no in-process mechanical alignment. Stainless steel screens (Type 316 Mesh #60, Small Parts Inc.) were attached to the opposite ends of the frame to provide a rigid, yet porous surface for scaffold adherence. A sheet of 0.32 cm thick aluminum (McMaster-Carr) was coated with silicone (VST 50 silicone elastomer. Factor II Inc., Lakeside, AZ) to prevent the scaffold from sticking to the aluminum. Three frames (one of 0%, 35% and 70% elongation) w ere clamped to this silicone coated aluminum backing which was subsequently placed inside the spray chamber. The polymer solution and nonsolvent were sprayed simultaneously onto the fra