Keywords

1 Introduction

Heart failure has many causes, but the most prevalent initiator is a myocardial infarction , in which myocardial tissue is deprived of oxygen for an extended period of time, usually through a blockage of one of the coronary arteries. In the ischemia induced by an infarction, one-quarter of the 4 billion cells in the left ventricle can be lost [1]. With this level of cell death, heart failure can develop as a result of the limited capacity of the injured myocardial tissue to recover or regenerate, leading to fibrosis and scar formation of the damaged myocardium, left ventricular dilation and thinning due to a resulting pressure and volume overload , and the inhibition of proper action potential propagation. Current treatments, whether they are pharmaceutical or medical device-based, act merely as palliative measures and the only effective long term treatment to date for heart failure is to replace the damaged myocardium via total heart transplantation [2].

With the limited number of donors and the inherent risks of surgery and immunogenicity, cellular therapy has become an attractive solution to the high incidence of heart failure post-infarction. However, direct injection of cells into the myocardium has shown limited efficacy due to poor grafting efficiency [3]. Low rates of retention may be a product of cell loss due to inability to create focal adhesions with neighboring cells, the inflammatory response to myocardial injury, or the hypoxic environment of the infarct zone. Thus, the use of engineered cardiac tissues has been a topic of increasing interest. Tissue engineering not only provides a platform for the delivery of a large number of cells to the injured myocardium, it provides a means to replace damaged myocardium. Tissue engineering allows for in vitro development of cellular organization, intracellular communication and ECM deposition while also isolating the cells from the inflammatory infarct environment when implanted in vivo (see Fig. 1).

Fig. 1
figure 1

Construction and characterization of cardiac patches

Full size image

In this chapter, we will review the currently used methods used to create engineered cardiac tissues, or “cardiac patches”, and their efficacy when delivered in vivo. We will begin by surveying the cells, scaffold and in vitro conditioning used to create these patches, and then move on to how they are characterized, and finally their efficacy to limit left ventricular remodeling post-infarction.

2 Cells for Cardiac Tissue Engineering

2.1 Neonatal Rat Cardiomyocytes

Until functional cardiomyocytes were able to be differentiated from human pluripotent stem cell sources, primary isolates of neonatal rat cardiomyocytes were the standard cell source for the development of cardiac patches [4–6]. These cells have been well characterized in vitro and can be obtained in large numbers, making them a useful tool for developing cardiac patches and interrogating the effects of in vitro conditioning on these tissues .

2.2 Pluripotent Stem Cell Derived Cardiomyocytes

Stem cells, whether taken from an embryo or reprogrammed from an adult cell, have enormous therapeutic potential due to their ability to be expanded in an undifferentiated state, and differentiated into cells of any lineage found in the body. In their undifferentiated state, pluripotent stem cells cannot be transplanted into patients as they can form teratomas. However, as the ability to not only differentiate pluripotent stem cells into functional cardiomyocytes but also in numbers large enough to be used therapeutically has become available, these cells have become the new standard cell source for creating cardiac patches.

2.2.1 Embryonic Stem Cell Derived Cardiomyocytes

Spontaneous differentiation of embryonic stem cells (ESCs) to cardiogenic cells was first observed in mouse ESCs in 1981 when these cells were cultured in 3D aggregates, called embryoid bodies [7]. Human ESCs were first obtained from a blastocyst in 1998 [8], but it wasn’t shown until 2001 that they could successfully be differentiated into functional cardiomyocytes that exhibited contractile force generation and recordable action potentials [9]. Human ESC-CMs are now commonly used as a cell source for cardiac tissue engineering [10, 11].

2.2.2 Induced Pluripotent Stem Cell Derived Cardiomyocytes

In 2006 it was discovered that the introduction of four transcription factors (Oct3/4, Sox2, c-Myc, and Klf4) into adult murine fibroblasts reprogrammed these cells into an embryonic state, called induced pluripotent stem cells (iPSCs) [12]. This was a landmark discovery, as the reprogramming of adult cells into an embryonic state not only alleviates the ethical concerns of using embryonic cardiomyocytes but also allows for potentially autologous cell transplantation, eliminating the immunogenicity obstacle of allogeneic transplantation. Human iPSCs were reported in 2007 using the same factors or a new combination of Oct4, NANOG, LIN28, and SOX2 [13, 14], and were differentiated into functional cardiomyocytes in 2009 [15]. They have been used in parallel with hESC-CMs for tissue engineering applications, and will likely become the preferred cell source for cardiac patches for the foreseeable future [10].

2.2.3 Differentiation Methods

Initially, pluripotent stem cell derived cardiomyocytes were differentiated via the embryoid body method, where undifferentiated pluripotent stem cells are cultured in 3D aggregates on top of irradiated mouse embryonic fibroblast feeder cells and allowed to spontaneously differentiate into cells from all three embryonic germ layers, among them being cardiomyocytes. However, as can be expected, this method results in low percentages and yields of cardiomyocytes and requires purification steps in order to obtain high percentages of cardiomyocytes. In efforts to efficiently obtain high yields of cardiomyocytes from stem cell differentiation, Zhang et al developed a new method in which cells are cultured between two layers of matrigel [16], resulting in a significant increase in differentiation efficiency to cardiomyocytes. Recently, small molecule methods of reprogramming and differentiation under defined conditions have been developed to eliminate the need for viral vectors and to increase reprogramming efficiency [17].

2.3 Cardiac Progenitor Cells

The heart has a limited capacity to regenerate via the presence of sparsely distributed resident cardiac progenitor cells (CPCs) and cardiac side population (SP) cells within the myocardium, approximately one CPC for every 30,000–40,000 cells in the myocardium [18]. These cells have been isolated based on their expression of a number of cell surface markers and gene expression, including c-kit [19], Sca-1 [20], and islet-1 [21] as well as from their migration of progenitor cells out of excised cardiac tissue when cultured in vitro [22]. Cells selected through these methods display the capacity to self-renew and differentiate into cardiomyocyte as well as endothelial phenotypes and can be used to create cardiac patches [11, 23].

2.4 Non-Cardiomyocyte Cells and Co-Culture of Cells for Cardiac Patches

Cardiomyocytes are not the only cell type used to create engineered tissues for cardiac repair. Non-cardiomyocyte cells are required for the creation of cardiac patches that utilize compacting biopolymer hydrogels, as cardiomyocytes do not contract collagen and fibrin hydrogels. Despite improvements in differentiation techniques, pluripotent stem cell-derived CM populations are only 60–90 % pure CMs, with the remaining 10–40 % of cells primarily being fibroblast-like cells, though those cells remain poorly characterized. Neonatal rat cell isolates also contain between 50–60 % of a heterogeneous population of non-CM cells, consisting primarily of fibroblasts, but also containing smooth muscle cells and endothelial cells from the vasculature and other interstitial cells from the myocardium [24]. The addition or inclusion of non-CM cells into cardiac patches has proven to be beneficial. Co-culture of hESC and hiPSc-derived cardiomyocytes with endothelial cells with or without mesenchymal stromal cells has resulted in an increase in contractile force generation by the resulting cardiac patches [10], and the force generated per cardiomyocyte has shown to be higher with less pure populations of hESC-CMs [25]. Other non-CM cells used in cardiac tissue engineering include blood outgrowth endothelial cells (BOECs) [26], human umbilical endothelial cells (HUVECs) [27], pericytes (PCs) [28], and human dermal fibroblasts (HDFs) [29]. These cells are used either in combination with cardiomyocytes or by themselves in engineered tissues, all with the goal of rescuing the injured myocardium post-infarction.

3 Scaffolds for Cardiac Tissue Engineering

Cardiac patches are typically formed by the in vitro seeding and culturing cells in 3-dimensional scaffolds. The choice of scaffold influences cell survival, phenotype, and function through its physical and chemical properties, including stiffness , microstructure, and surface chemistry. Two general classes of scaffolds are utilized to create cardiac patches: prefabricated scaffolds, in which cells are seeded onto a pre-existing scaffold structure; and biopolymer scaffolds, in which cells are present during scaffold polymerization and become entrapped in a fibrous network .

3.1 Prefabricated Scaffolds

Synthetic biodegradable, elastomeric polymers including polycaprolactone (PCL) [30], poly(glycerol-sebacate) (PGS) [31], and poly(lactide-co-glycolide) (PLGA) [32] copolymers and natural materials such as collagen I [33], alginate [34], and gelatin [35] are materials that have been used to create prefabricated foam or nanofiber scaffolds. Foam scaffolds allow for control of pore size and structure through the use of different material processing techniques. Highly porous scaffolds have proven successful in creating cardiac patches with high cell viability, but they are often isotropic and do not result in cellular alignment. Nanofiber scaffolds can be created through electrospinning [36], in which fiber diameter and orientation can be controlled, mimicking organization of collagen fibers in the native myocardium. These scaffolds can be fabricated in advance and can incorporate surface modifications to enhance cell adhesion and activity, such as the conjugation of RGD peptides [23].

3.2 Biopolymer Scaffolds

Biopolymers are another class of scaffolds used in cardiac tissue engineering. These are native protein fibril networks in hydrogel form made from either collagen I or fibrin [5, 37, 38]. Rather than being seeded onto pre-formed scaffolds, cells are suspended in a gel-forming solution and fibrillogenesis occurs around the cells, entrapping them in a hydrated, fibrillar network. Initially, collagen and fibrin gels are extremely soft andisotropic. Over time, entrapped non-CM cells compact the gel, resulting in a denser fibril network. Eventually, the entrapped cells not only compact the gel, but begin to degrade the fibrils (in the case of fibrin gels) and replace them with cell-produced matrix. By constraining the compaction of the gel, fibrillar and cellular alignment can be induced in these gels. Induction of alignment allows the native architecture of the myocardium to be mimicked and has been shown to result in an increase in both contractile force generation by entrapped cardiomyocytes and the presence of gap junctions between cardiomyocytes [5]. Collagen and fibrin biopolymer scaffolds can be altered by the addition of Matrigel [10, 25, 39]. Other biopolymer scaffold materials include alginate, which can be 3D printed into various geometries [23].

3.3 Tissue-Based Scaffolds

Decellularized heart tissue has been used as a scaffold to culture cardiac cells. Heart tissues are decellularized by perfusion with SDS followed by Triton-X to remove cells while leaving most of the heart ECM content and structure intact [40]. Cells can then be reperfused into the heart [40], or entrapped in an ECM-hydrogel matrix to create cardiac patches [41].

3.4 Scaffold-Free Tissues

Scaffold-free approaches have been used to create cardiac patches in the form of cell sheets [42] and cell aggregate patches [43]. Cell sheets are created by seeding cells onto temperature-responsive membranes, allowing them to deposit cell-produced ECM and form cellular connections, and releasing them from the culture surfaces through temperature reduction. Cell sheets can then be stacked to create thin 3D tissues or layered with cell sheets of endothelial cells to facilitate rapid vascularization upon implantation [42]. Cell aggregate patches are prepared by culturing cells on a low-attachment plate placed on an orbital shaker, allowing for large numbers of cells to aggregate in solution and form tissues.

4 In vitro Cellular Conditioning

In addition to seeding cells onto or into scaffolds, external stimulation of the cells is often required to induce them to convert the scaffold into a functional cardiac tissue. This may be achieved through mechanical, electrical, or pharmacological stimulation.

4.1 Mechanical Stimulation

Cardiomyocytes in the native myocardium experience cyclic mechanical stretch as part of the cardiac cycle. Taking inspiration from this, in vitro mechanical stimulation has been widely used to stimulate alignment, hypertrophy, and maturation of cardiomyocytes in engineered tissues. Stimulation through cyclic stretching has been achieved for ring-shaped cardiac patches using two methods: looping the patch around one fixed post and another coupled to a motorized stretching device [44], or through pneumatically-controlled distension of a latex mandrel on which the patch is mounted [37]. In both instances, cyclic stretching resulted in over a two-fold increase in contractile force generation.

4.2 Electrical Stimulation

Ventricular cardiomyocytes in the heart beat under the continuous regulation of their electrical activity through the cardiac conduction system originating at the sinoatrial node. In vitro, cardiomyocytes in cardiac patches beat spontaneously, but often at an irregular and variable rate, and not in synchrony with cardiomyocytes in other regions of the patch. Electrical stimulation of cardiac patches throughout culture has resulted in increases in the maximum frequency at which these patches can be stimulated and elicit a contractile response synchronous with pacing, a reduction in the voltage threshold at which cardiomyocytes contract, increased contractile force and also stimulated cardiomyocyte hypertrophy [45].

4.3 In vitro Perfusion

Due to the high metabolic demand of cardiomyocytes, diffusion limitations hamper the ability to make the thicker cardiac patches that are necessary to achieve clinical relevance. To increase oxygen and nutrient availability to cells, different methods of in vitro perfusion have been utilized in cardiac patches. Pulsatile perfusion of cell culture medium through porous pre-formed scaffolds has proven effective in increasing cell viability and myofibril assembly through media flow [6] and activates the ERK 1/2 signaling pathway [46]. In biopolymer hydrogels, microchannels have been included in the gel [47], and perfusable microvessels have been created within the gel to allow for medium flow through the gel during culture [48–50]. Though promising, this latter approach has yet to be combined with cardiomyocyte culture in vitro.

4.4 Medium Supplementation

Supplementation of culture medium with pharmacological agents can also be used to stimulate development of cardiac patches. The addition of platelet-derived growth factor BB (PDGF-BB) to cultures of collagen/matrigel scaffolds containing neonatal rat cardiomyocytes protected cells from apoptotic death, thus increasing the final contractile performance of the patch [51]. Insulin and ascorbic acid have been used to promote matrix deposition by entrapped non-CM cells [37].

4.5 Combination Treatments

In addition to exposing cardiac patches to one form of in vitro conditioning, combinations of the above mentioned techniques have been used in combination to further condition tissues. Electrical stimulation has been used in concert with perfusion and resulted in an increase in cell elongation, enhanced expression of gap junction protein connexin-43, increase in cell number, and an increase in contractile performance [52, 53]. However, the combination of stretch and β-adrenergic stimulation resulted in no additional benefit [53].

5 Characterization of Cardiac Patch Function

One additional benefit of using engineered tissues to treat heart failure post-infarction is that these engineered tissues and the cells entrapped in them can be functionally characterized in vitro, allowing for quality control and a more thorough understanding of the benefits of cell transplantation. Functional performance of patches is primarily assessed through contractile force measurements and optical mapping of electrical conductivity.

5.1 Contractile Force Measurements

Cardiac patches should, once developed, function as a piece of cardiac muscle tissue and generate contractile stresses close to physiological values. Quantification of this contractile activity is widely used to assess the quality of a cardiac patch. Contractile activity can be assessed through the magnitude of patch contraction both spontaneously and in response to pacing, the maximum frequency by which a patch can be paced and still elicit a contractile response in phase with the electrical stimulus, the force-frequency relationship of contractile amplitude with increasing pacing frequency, and the contractile response of the patch to exposure to pharmacological agents. Such agents include β-adgrenergic agonists, gap junction blocker 1-Heptanol [54], or cholinergic agonist carbacol [39].

5.2 Optical Mapping

Electrical activity of cardiac patches can be assessed through optical mapping. To do this, tissues are submerged in medium containing a voltage sensitive dye, such as Di-4ANEPPS or ANNINE-6. Tissues are then point-stimulated and the action potential propagation is recorded by a high speed camera. This data can be analyzed to assess both conduction velocity (the rate at which action potentials propagate across the tissue), and action potential duration (the time it takes for a cell to re polarize to either 50 or 90 % of its resting potential). This data can be used to obtain the conduction velocities of action potentials across the patch, as well as the action potential duration. These values can then be compared to physiological values to assess the functional quality of the cardiac patch [55]. Additionally, calcium transients can be similarly imaged with a calcium sensitive dye [56].

6 In vivo Assessments of Cardiac Patches

Although cardiac patches can be fully characterized in vitro and can be conditioned in vitro to obtain functional values close to that of the native myocardium, cardiac patches can only truly be effective if their implantation results in the restoration of cardiac function post-infarction. Some studies have made progress towards achieving this goal, utilizing both small and large animal models, and acute or delayed implantation. The timing of patch implantation is a crucial aspect of in vivo assessments as a way to help determine the mechanisms by which any benefits provided by the patch occur. Acute transplantation is transplantation of the patch immediately after infarction, delivering it during the initial inflammatory phase before fibrosis begins to occur. Chronic, or delayed transplantation, requires a second surgery and applies the patch after the inflammatory phase has passed and either before or after mature scar formation, depending on the time of implantation.

6.1 Rodent Models

Rodent models of myocardial infarction treatments are commonly used as the first initial model due to the smaller size of patch required for treatment. Acute placement of a cardiac patch using fibrin as the scaffold and syngeneic neonatal rat cardiac cells, with in vitro stretch conditioning of the patch prior to implantation, resulted in minimal scar formation and restoration of cardiac function 4 weeks post-infarct, with 36 % cardiomyocyte retention [37]. A similar patch made with a collagen/Matrigel scaffold rather than fibrin and implanted 2 weeks post-infarction limited further scar formation and preserved cardiac function from the time of implantation [4]. Transplantation of a layered neonatal rat cell sheet patch into a nude rat infarct model 14 days post-infarction resulted in significant improvements in cardiac function, though cell retention was minimal after 4 weeks [57]. Few studies to date have been published assessing the functional consequences of a human stem cell derived cardiac patch in an infarcted rat model. Acute transplantation into non-infarcted rats of scaffold-free tissue patches utilizing hESC-CMs and HUVECs resulted in rapid vascularization of the patches in vivo. Patches containing only an enriched CM population resulted in poor CM retention [58].

6.2 Large Animal Models

The use of large animal models of myocardial infarction is a necessary step towards bringing cardiac patches to clinical relevance Few studies to date have been able to generate the large numbers of human stem cell derived cardiomyocytes required to create cardiac patches large enough to be implemented in large animal models. Human ESC-CMs in cell sheet form were transplanted into a porcine ischemia-reperfusion model 4 weeks after ligation in a feasibility and safety study, and it was found that the cell sheet treatment resulted in a reduction in LV dilation and an improvement in cardiac function both 4 and 8 weeks post-transplantation [59].

7 Current Deficiencies and Obstacles

Despite the many advances in cardiac tissue engineering , there remain deficiencies in the field and many obstacles still to be overcome. Contractile force generation and cellularity of cardiac patches average far below physiological values of 44 mN/mm2 and 20–40 M cells/gram for human ventricular myocardium and 56.4 N/mm2 rat ventricular myocardium [60]. Diffusional limitations may contribute to this, and may be overcome by the development of a microvascular network- withing the patch during its fabrication. As mentioned earlier, progress has been made, but microvascular network have yet to be created and perfused in the presence of a cardiomyocyte-containing patch of size relevant to even a rat infarct. Other issues that will be addressed in the coming years are maturation mismatch of donor cells to recipient hearts. Currently used stem cell-derived cardiomyocytes are a mixture of atrial, ventricular, and pacemaker phenotypes, and all exhibit a fetal or neonatal phenotype. Immature phenotypes also have different electrophysiological, metabolic, and contractile properties than adult CMs. It has yet to be determined if transplantation of immature cells will elicit arrhythmic events, or if cells need to be differentiated to maturity prior to transplantation. However, more mature cells may not survive transplantation with the same efficiency as less mature cells.