Optimizing Engineered Heart Tissue for Therapeutic Applications as Surrogate Heart Muscle
Background— Cardiac tissue engineering aims at providing heart muscle for cardiac regeneration. Here, we hypothesized that engineered heart tissue (EHT) can be improved by using mixed heart cell populations, culture in defined serum-free and Matrigel-free conditions, and fusion of single-unit EHTs to multi-unit heart muscle surrogates.
Methods and Results— EHTs were constructed from native and cardiac myocyte enriched heart cell populations. The former demonstrated a superior contractile performance and developed vascular structures. Peptide growth factor-supplemented culture medium was developed to maintain contractile EHTs in a serum-free environment. Addition of triiodothyronine and insulin facilitated withdrawal of Matrigel from the EHT reconstitution mixture. Single-unit EHTs could be fused to form large multi-unit EHTs with variable geometries.
Conclusions— Simulating a native heart cell environment in EHTs leads to improved function and formation of primitive capillaries. The latter may constitute a preformed vascular bed in vitro and facilitate engraftment in vivo. Serum- and Matrigel-free culture conditions are expected to reduce immunogenicity of EHT. Fusion of single-unit EHT allows production of large heart muscle constructs that may eventually serve as optimized tissue grafts in cardiac regeneration in vivo.
Cell-based therapies aim at reconstructing biologically active tissues in vivo to restore the function of failing organs.1,2 Experimental and first clinical trials support the notion that hearts can at least partially be repaired by cell implantation. An alternative approach that might overcome some limitations of cell therapy is the implantation of in vitro engineered myocardium.3 Tissue engineered myocardium may not only function as an add-on to existing structures but also could theoretically be used as true myocardial replacement or in de novo construction of myocardium in congenital malformations (eg, Tetralogy of Fallot, single and hypoplastic ventricles). Several principally different cardiac tissue engineering approaches have been developed. These are: (1) seeding of cardiac myocytes on preformed polymeric scaffolds, which may function as organ blueprints;4–7 (2) stacking of cardiac myocyte monolayers to form cardiac muscle-like tissue without additional matrix material;8 and (3) entrapping of cardiac myocytes in a cardiogenic growth environment to support self-assembly into functional myocardium.9–11 All these tissue engineering concepts have been tested in animal models showing survival and growth of engrafted heart muscle surrogates.6–8,12
Yet several caveats must be considered before tissue engineering-based cardiac regeneration may be applied clinically. Optimal myocardial structure and function depends not only on the cardiac myocyte fraction but also on non-myocytes, which compose 70% of the total cell content of a heart.13,14 Consequently, engineered myocardium may benefit from a “physiological” cell composition. Culture in the presence of xenogenic serum may lead to impregnation of cell and matrix components or induce autoantigen presentation. Thus, identification of serum-free culture conditions will be of paramount importance, particularly when considering good laboratory practice and good manufacturing practice guidelines. So far, several culture medium supplements including peptide growth factors have been identified to support cardiac myocyte growth in vitro.15 Yet a serum-free cardiac tissue engineering approach has not been developed. Extracellular matrix from Engelbreth-Holm-Swarm tumors (also known as Matrigel) has been identified as an essential component in engineered heart tissue (EHT)10 and has recently also been used by others to support cardiac muscle development in vitro.16 However, its replacement with defined factors is equally essential as the replacement of serum. Eventually replacement of large myocardial tissue defects as well as reconstruction of missing myocardial structures will require large artificial heart muscle patches.
This study aimed at providing some solutions to the aforementioned bottlenecks. Here, we focus on improving a tissue engineering technology, namely the generation of EHT, which has been developed by our group over the past 10 years.9,11 This study demonstrates that: (1) EHTs exhibit an improved function when constructed from mixed rather than cardiac myocyte enriched heart cell populations; (2) supplementation of culture medium with growth factors allows EHT culture under serum-free conditions; (3) Matrigel can be replaced by insulin and triiodothyronine; and (4) EHTs can be generated in various geometries and sizes.
Methods and Materials
Cell Isolation and Identification
Cells were isolated by fractionated DNase/Trypsin digestion as described earlier.10 Isolated cells were either plated on tissue culture dishes or Petri dishes that facilitated attachment of non-myocytes or did not allow cell attachment at all, respectively. Plating was performed for 1 hour. Cells in suspension, being depleted off non-myocytes (tissue culture dishes) or not depleted (Petri dishes), were pelleted by centrifugation, reconstituted in culture medium, counted in a Neubauer cell chamber, plated on collagen coated cover slips (100 000 cell/cm2), and cultured in the presence of BrdU (0.1 mmol/L) for 48 hours (DMEM, 1 g/L glucose, and supplements as indicated). Staining for cell type-specific antigens (α-sarcomeric actinin: cardiac myocytes; prolyl-4β-hydroxylase: fibroblasts; α-smooth muscle actin: smooth muscle cells), filamentous actin [TRITC-phalloidin], and DNA (propidium iodide; 5 μg/mL) was performed after formaldehyde fixation as described earlier.11 Finally, 400 to 1500 cells per group were analyzed to identify the respective cell species by laser scanning microscopy (LSM; LSM 5 Pascal; Zeiss). Endothelial cell (CD31-positive) number was assessed by FACS (3 independent counts of 10 000 cells; FACScan, Becton Dickinson).
EHTs were constructed as described previously.11 Briefly, acetic acid solubilized collagen type I was mixed with concentrated culture medium (2x DMEM, 20% horse serum, 4% chick embryo extract, 200 U/mL penicillin, 200 μg/mL streptomycin). The pH was neutralized by titration with 0.1 N NaOH. Matrigel was added (10% v/v) if indicated. Finally, cells were added to the reconstitution mixture, which was thoroughly mixed before casting in circular molds (inner diameter, 8 mm; outer diameter, 16 mm; height, 5 mm). Within 3 to 7 days, EHTs coalesced to form spontaneously contracting circular structures and were transferred on automated stretch devices11 or flexible holders for continuous culture under chronic strain.
Isometric Force Measurements
Force of contraction (twitch tension) of EHTs was measured under isometric conditions in standard organ baths as previously described to assess responsiveness to calcium (0.2 to 2.8 mmol/L) and isoprenaline (0.1 to 1000 nmol/L).11
RNA and DNA Isolation
RNA and DNA were isolated with Trizol (Invitrogen) and quantified by spectrophotometry.
Protein Isolation and Western Blotting
EHTs were washed thoroughly in ice cold phosphate-buffered saline and homogenized in lysis buffer containing 30 mmol/L Tris-hydroxyl aminomethane (pH 8.8), 5 mmol/L EDTA (pH 8), 3% SDS, 10% glycerol, 30 mmol/L NaF, and 2 μg/mL aprotinin. Protein was measured by a modified Lowry assay (Bio-Rad DC Protein Assay) using rat IgG as standard. Similar quantities of total proteins were loaded on 7% SDS-polyacrylamide gels and separated electrophoretically. Similar loading was confirmed by Ponceau staining after transfer onto polyvinylidene fluoride membranes. Blots were probed with monoclonal antibodies directed against sarcomeric actin (clone 5C5; Sigma). Primary antibodies were detected with appropriate horseradish peroxidase coupled secondary antibodies. Signals were enhanced with the ECL-plus kit (Amersham) and recorded with a ChemiDoc system (Syngene). Signals were quantified with GeneTools software (Syngene).
Morphological Evaluation of EHT
Formaldehyde-fixed EHTs were embedded in paraffin, sectioned, and stained with hematoxylin and eosin for light microscopic evaluation or processed for LSM as described previously.11 Living EHTs were incubated for 4 hour at 37°C with DiI-LDL (3,3′-dioctadecylindocarbocyanine iodide–low-density lipoprotein; 10 μg/mL) and analyzed by LSM after thorough washout off free DiI-LDL.
Data are presented as mean±standard error of the mean. Statistical differences were determined using paired 2-tailed Student t tests or ANOVA when appropriate. P<0.05 was considered statistically significant.
Statement of Responsibility
The authors had full access to the data and take full responsibility for their integrity. All authors have read and agree to the manuscript as written.
Assessment of Cell Composition
Cell species were identified by staining for α-sarcomeric actinin (cardiac myocytes; Figure 1a), prolyl-4-hydroxylase (fibroblasts; Figure 1b), α-smooth muscle actin (smooth muscle cells; Figure 1c), and CD31 (endothelial cells; analyzed by FACS). The native cell isolate contained 47±1% cardiac myocytes and 49±3% fibroblasts. Preplating increased the cardiac myocyte fraction to 63±2% and decreased the fibroblast fraction to 33±3% (Figure 1d, 1e). Smooth muscle cell (3% to 4%) and endothelial cell (2% to 3%) fractions were not affected by this procedure.
Differences of EHTs With Native and Cardiac Myocyte-Enriched Cell Fractions
Baseline force of contraction (systolic force) was markedly higher in EHTs with native heart cell populations (n=17) when compared with EHTs with cardiac myocyte enriched heart cells (n=13; 0.7±0.06 versus 0.3±0.03 mN; Figure 2a). At the same time, resting tension (diastolic force) was increased in EHTs with native heart cell populations (0.4±0.04 versus 0.2±0.04 mN; Figure 2a). In addition, inotropic responses to calcium (Figure 2b) and isoprenaline (Figure 2c, 2d) were higher in the latter. DNA content was higher in native heart cell EHTs (14±0.5 versus 12±0.7 μg; P<0.05) despite a similar input cell number (2.5×106 cells/EHT), but no microscopically distinguishable difference in EHT cell number was observed. The only apparent morphological difference was an increased number of vascular structures in native heart cell EHTs (Figure 3a to 3d).
Withdrawal of horse serum and chick embryo extract after development of contractile EHTs (starting on culture day 7) resulted in rapid ceasing of contractile activity.17 Earlier studies had shown that EHTs benefit from addition of growth factors.11 Thus, we performed an initial series of experiments to test whether growth factors (alone or in combination) can maintain cardiac myocytes in monolayer culture in the absence of serum. All tested factors prevented atrophy of cardiac myocytes to some extent, but only the combination of all factors resulted in a cardiac myocyte phenotype essentially undistinguishable from cardiac myocytes cultured in 10% fetal calf serum (Figure 4). Consequently, a mixture of these factors with additional baseline supplements maintained contractile EHTs in the absence of serum (Figure 5).
Withdrawal of Matrigel
Matrigel has been essential for the construction of EHTs with cardiac myocyte enriched heart cell populations.10 The latter effect is most likely explained by the high growth factor content in Matrigel. Based on our observation that insulin elicited strong beneficial effects on EHT development and function,11 we tested whether its addition to the culture medium would suffice to substitute Matrigel. Triiodothyronine (T3) supports sarcomere development and was also tested as a Matrigel replacement. Interestingly, addition of Matrigel was not essential for EHTs with native heart cell populations. However, contractile force and responsiveness to calcium and isoprenaline was markedly better if Matrigel was included in the original EHT reconstitution mixture. T3 and insulin alone did not or only slightly increase force of Matrigel-free EHTs (Figure 6a); however, simultaneous addition of both factors allowed generation of strongly contracting EHTs. Notably, addition of insulin and T3 during the first culture day (24 hours) was sufficient to elicit its long-term effects (Figure 6b). Essentially, T3 and insulin supplements not only compensated the Matrigel effect but improved EHT contractility above control (EHTs with Matrigel) levels (Figure 6c, 6d). These effects did not stem from enhanced EHT cell number (no difference in DNA content and no apparent structural differences; data not shown) but may be the result of enhanced protein content in T3/insulin-treated EHTs when compared with nontreated Matrigel-free EHTs (593±13 versus 479±15 μg/EHT, n=4 per group; P<0.05). Notably, sarcomeric actin content was higher in the former EHTs (Figure 6e) and we observed a clear positive correlation of actin content and force of contraction (Figure 6f).
Constructing Complex EHTs
Small tissue-engineered cardiac muscle may not suffice to function as surrogate heart tissue in vivo. Here, we used the propensity of EHTs to fuse during in vitro culture to create several EHT geometries, namely, star-shaped EHT patches (fusion predominately in the center of the EHTs; Figure 7a; supplemental video 1), tubular constructs (fusion along the side of EHT rings; Figure 7b; supplemental video 2), EHT networks by cutting open single EHTs in one point and weaving them together (Figure 7c; supplemental video 3), or EHT “ropes” by twirling several EHTs around each other (Figure 7d; supplemental video 4). Fusion was established within 24 to 48 hours, yielding in-unison contracting novel heart muscle constructs. EHTs could be cultured for at least 1 month under the described in vitro conditions. Single and complex EHTs developed thin cardiac myocyte networks (10 to 20 μm) and thick myocardial structures (&150 μm), which were localized preferentially in the center and close to the surface, respectively (Figure 8).
The present study addressed several issues pertaining to the improvement of so far developed tissue engineering concepts and eventually its clinical application. Namely, use of native, ie, not enriched, cell populations improved the contractile performance of EHT. Serum-free and Matrigel-free culture conditions could be established to reduce the fraction of immunogenic EHT components. Finally, new means to construct more complex EHT structures were exploited demonstrating the possibility to generate large contractile cardiac surrogate tissue in vitro.
The use of mixed cell populations in cardiac tissue engineering appears straightforward if one wishes to construct a true heart muscle equivalent. Yet, non-myocytes tend to overgrow cardiac myocytes in monolayer cultures. This is apparently not the case in EHTs, which can be cultured in the absence of otherwise commonly used cytostatic agents. Interestingly, large vascular structures formed in EHTs containing native heart cell populations, which may facilitate engraftment in vivo. The finding that non-myocytes facilitate EHT development also imposes a potential caveat for future application of stem cells in cardiac tissue engineering. Essentially, our data suggest that stem cells would have to be differentiated into various cardiac cell types to establish and maintain a cardiogenic environment.
The development of a serum-free chemically defined tissue engineering concept will be essential to correspond to good laboratory practice and good manufacturing practice guidelines. The choice of growth factors in our serum-free culture mix was based on previous experience11 and findings by others.15 Further optimization of the described medium is warranted. This relates especially to the identification of the “right” concentration of a single factor at the “right” time and eventually the “right” combination of factors. The importance of timing and combination of growth supporting factors is exemplified by the finding that Matrigel can be replaced by adding insulin and T3 for only 24 hours at the beginning of EHT culture. In fact, the necessity to only transiently stimulate cells within EHTs with T3 and insulin could indicate a crucial phase for EHT development. We did indeed observe a marked cell loss during the first three culture days in standard EHT cultures (unpublished observation). The cause of cell loss appeared to be mostly apoptotic cell death; however, an involvement of necrosis and autophagia cannot be ruled out. Consequently, substances that protect cells, and especially cardiac myocytes, from apoptotic cell death may impose a beneficial effect on EHT development. Several groups have described anti-apoptotic and hypertrophic effects of insulin.18 In contrast, T3 may not directly protect from cell death but induce expression of α-myosin heavy chain molecules and thereby play a role in re-establishing sarcomere structures in freshly isolated heart cells.19 Eventually, large-scale screening will be necessary to define an optimal cardiogenic cocktail to facilitate the generation of structurally and functionally optimal EHTs. A “physiological” cell population in EHTs may also contribute to the allocation of essential factors. This could explain why contractile EHTs could be generated in the current study also in the absence of Matrigel, albeit with a low calcium and isoprenaline responsiveness. This was formerly not possible when enriched heart cell populations were used.10
The potential to fuse single EHTs to more complex and essentially larger contractile heart muscle units is intriguing and may offer a solution to the common problem of size limitation in cardiac tissue engineering. Here, large constructs consist of single small units that remain largely diffusible for oxygen and metabolites. Core necrosis was not observed in single-unit or in multi-unit EHTs, and cell debris stemming from apoptosis during early stages of EHT development may in fact be scavenged by macrophages within EHT.11 More detailed studies on cell survival and eventually cell seeding efficiency in EHTs are presently underway. Single-unit EHTs developed maximal forces of 1.5 mN in the present study. Exemplary force measurements in multi-unit EHTs demonstrated maximal forces of 1.5 to 3 mN. However, the latter measurements were hampered by the nonuniform geometry of the tissue constructs. This may in fact lead to an underestimation of force development. Ultimately, the utility of structurally and functionally optimized EHTs in cardiac regeneration must be tested in vivo.
Conclusion and Future Aspects
The present study suggests that tissue-engineered cardiac muscle should be composed of a “physiological” heart cell population and can be cultured in the absence of serum as well as Matrigel. Fusion of single EHTs may eventually allow enlargement of EHT to a clinical scale. However, several issues remain to be addressed before tissue engineering may be translated from a purely experimental stage to a clinically applicable treatment. These include the identification of clinically applicable cells, replacement of all non-human components, and an unambiguous proof of the cardiac tissue engineering concept in large animals.
Sources of Funding
This study was supported by the German Research Foundation (Deutsche Forschungsgemeinschaft; DFG Es 88/8-2 to T.E., DFG GRK 750 A1 to W.H.Z., and DFG RO 1306/2-2 to S.R. and W.H.Z.), the German Ministry for Education and Research (BMBF 01GN 0124 and BMBF 01GN 0520 to T.E. and W.H.Z.), the Deutsche Stiftung für Herzforschung (F29/03 to W.H.Z), the Novartis Foundation (W.H.Z.), the European Union (EUGeneHeart to T.E.), and the Foundation Leducq (T.E. and W.H.Z.).
Presented at the American Heart Association Scientific Sessions, Dallas, Tex, November 13–16, 2005.
Murry CE, Field LJ, Menasche P. Cell-based cardiac repair: reflections at the 10-year point. Circulation. 2005; 112: 3174–3183.
Eschenhagen T, Zimmermann WH. Engineering myocardial tissue. Circ Res. 2005; 97: 1220–1231.
Leor J, Aboulafia-Etzion S, Dar A, Shapiro L, Barbash IM, Battler A, Granot Y, Cohen S. Bioengineered cardiac grafts: A new approach to repair the infarcted myocardium? Circulation. 2000; 102: III56–II61.
Li RK, Jia ZQ, Weisel RD, Mickle DA, Choi A, Yau TM. Survival and function of bioengineered cardiac grafts. Circulation. 1999; 100: II63–II9.
Shimizu T, Yamato M, Isoi Y, Akutsu T, Setomaru T, Abe K, Kikuchi A, Umezu M, Okano T. Fabrication of pulsatile cardiac tissue grafts using a novel 3-dimensional cell sheet manipulation technique and temperature-responsive cell culture surfaces. Circ Res. 2002; 90: e40.
Eschenhagen T, Fink C, Remmers U, Scholz H, Wattchow J, Weil J, Zimmermann W, Dohmen HH, Schafer H, Bishopric N, Wakatsuki T, Elson EL. Three-dimensional reconstitution of embryonic cardiomyocytes in a collagen matrix: a new heart muscle model system. FASEB J. 1997; 11: 683–694.
Zimmermann WH, Schneiderbanger K, Schubert P, Didie M, Munzel F, Heubach JF, Kostin S, Neuhuber WL, Eschenhagen T. Tissue engineering of a differentiated cardiac muscle construct. Circ Res. 2002; 90: 223–230.
Zimmermann WH, Didie M, Wasmeier GH, Nixdorff U, Hess A, Melnychenko I, Boy O, Neuhuber WL, Weyand M, Eschenhagen T. Cardiac grafting of engineered heart tissue in syngenic rats. Circulation. 2002; 106: I151–I157.
Brutsaert DL. Cardiac endothelial-myocardial signaling: its role in cardiac growth, contractile performance, and rhythmicity. Physiol Rev. 2003; 83: 59–115.
Radisic M, Park H, Shing H, Consi T, Schoen FJ, Langer R, Freed LE, Vunjak-Novakovic G. Functional assembly of engineered myocardium by electrical stimulation of cardiac myocytes cultured on scaffolds. Proc Natl Acad Sci U S A. 2004; 101: 18129–18134.
Eschenhagen T, Didie M, Munzel F, Schubert P, Schneiderbanger K, Zimmermann WH. 3D engineered heart tissue for replacement therapy. Basic Res Cardiol. 2002; 97 (Suppl 1): I146–I152.
Gustafson TA, Bahl JJ, Markham BE, Roeske WR, Morkin E. Hormonal regulation of myosin heavy chain and alpha-actin gene expression in cultured fetal rat heart myocytes. J Biol Chem. 1987; 262: 13316–13322.