Editorials |
From the Institute of Experimental and Clinical Pharmacology and Toxicology (T.E., W.H.Z.), University-Medical Center Hamburg Eppendorf, Germany; and the Department of Physiology (A.G.K.), University of Bern, Switzerland.
Correspondence to Thomas Eschenhagen, University Hospital Eppendorf, Institute of Experimental Pharmacology, Martinistrasse 52, 20246 Hamburg, Germany. E-mail t.eschenhagen{at}uke.uni-hamburg.de
See related article, pages 705712
Key Words: tissue engineering electrical coupling impulse propagation arrhythmia
The past 20 years have seen tremendous progress in the medical treatment of cardiac diseases, a progress that actually translated into decreased cardiovascular morbidity and mortality in clinical practice.1 Yet evidence accumulates that interventions intended to slow down the progression of chronic heart diseases have approached their limit. This may be one of the reasons why the recent progress in stem cell biology has stirred so much attention. In addition, recent studies suggesting the existence of cell populations in the adult organism that, albeit at a low rate and insufficiently, endogenously replace cardiac myocytes and endothelial and smooth muscle cells have challenged the dogma of the heart as a nonregenerating organ.25 Such cell populations could be accessed in biopsies and principally serve as a source to replace cardiac myocytes and other cardiac cell types lost after myocardial infarction or other pathologies. Human embryonic stem cell lines, on the other hand, can nowadays be generated from human blastocysts (accessible from IVF leftovers6) at a high success rate and have been unambiguously shown to generate human cardiac myocytes.79 Thus, despite many open questions and some controversy in the field (reviews in references 10,11), cardiac regenerative therapy has become a realistic perspective.
But what to do with stem cells or their derivatives, given that means will be provided to generate them at sufficient numbers? The most straightforward strategy is to infuse or inject these cells into the injured heart and to develop methods to increase the fraction of cells that home and survive in the heart and differentiate to the desired functional elements. This strategy has already been clinically evaluated, but gave mixed results,1216 potentially also because the homing rate after infusion and the survival rate after intramyocardial injections are very low.17 Given the creativity and technical expertise of invasive cardiology, significant progress in the efficacy of such delivery technique is to be expected. Alternatively, cells will be applied not as a suspension but as engineered cardiac muscle tissues generated in vitro. Several techniques have been developed over the past years that allow the generation of cardiac tissuelike 3-dimensional (3D) constructs from neonatal rat heart cells that coherently contract and develop contractile force similar to native heart tissue (for review see reference 18). The group around Shimizu and Okano has pioneered a technique which may be considered the most direct and simple approach in cardiac tissue engineeringthe production of matrix-free cardiac cell sheets that detach from thermosensitive plastic surfaces at room temperature and can be stacked to form up to 100-µm thick 3D cardiac tissue sandwiches.19 A previous study showed that these sandwiches survive when implanted on the heart of nude rats and are well vascularized.20 Yet a critical question in this as in all other tissue engineering and cell therapy approaches is whether the implanted biological material electrically integrates into the host myocardium and participate in the synchronous contraction of the whole heart.
The study by Furuta et al in this issue of Circulation Research addresses this pivotal question.21 Studying electrical coupling in the contracting heart is not trivial, however. First, donor and host cells/tissues have to be unambiguously identified and must be distinguishable from each other. Second, mechanical cardiac activity interferes with imaging at the high spatial resolution required to study coupling and needs to be suppressed, ideally without affecting electrical activity. Third, myocardial infarction in humans and in animal models, including rats and mice, are complicated by the fact that surviving myocardial islets and strands may obscure the analysis of donorhost coupling. The group of Rubart and Field has recently demonstrated functional coupling between genetically labeled donor fetal cardiac myocytes and unlabeled adult host myocardium using 2-photon laser confocal microscopy in Langendorff-perfused murine hearts to measure Ca2+ transients at cellular resolution.22
Furuta et al used a cardiac injury model in which the subepicardial muscle layer of the left ventricular free wall of nude rats was thermally ablated to a depth of 150 to 300 µm.21 Hearts were then mechanically arrested by cytochalasin D, and multisite electrical mapping of transmembrane potentials was performed by optical recordings of voltage-sensitive dye fluorescence using a fast CCD device. The authors tested the functional integration of the transplanted cell sheets by comparing subepicardial maps of electrical excitation in the absence of transplanted cell sheets with maps obtained at different times after transplantation. The authors show several important results: First, electrical coupling between the donor layers and the host heart can be observed after 1 week. Second, subepicardial electrical anisotropy (ie, fast impulse propagation parallel and slow propagation transverse to the main axis of the cardiac strands) closely approaching the degree of anisotropy of the donor heart is measured 4 weeks after the transplantation. Third, the values for supepicardial propagation velocity after 4 weeks are closely similar to values in the host heart and indicate full functional integration. In general, histological sections revealed a layer of donor cardiac myocytes in the cell sheet separated from the injured myocardium by a necrotic zone. Donor myocytes exhibited cross striations and intercellular junctions staining positive for connexin43. The dot-like appearance of these connections is similar to the patterns observed in the border zone of myocardial infarction.23,24 Morphometric analysis of serial sections will be necessary to characterize this donor-host interface more precisely.
The study by Furuta et al,21 although clearly supporting general evidence for electrical coupling between donor and host myocytes,22 leaves also a number of open questions and caveats.
One limitation is the model of thermal injury, which had been chosen for simplicity and clarity, but has little validity for cardiac pathology. Coupling between cell sheets and the surface of the heart may be quite different between otherwise healthy hearts and chronically-diseased scarred hearts, both for mechanical and biological reasons. Similarly, functional consequences of cell sheet implantations cannot be evaluated in this model because contractile function is unaltered. Both questions have to be studied in more relevant models such as chronic myocardial infarction after LAD ligation.
An important unresolved issue relates to the topology and mechanism of the apparently robust electrical coupling. The hypothesis proposed by the authors states that the cell sheet on top of the injured zone couples to the host myocardium in the intact borderzone outside the injured area and subsequently conducts the electrical impulse across. This hypothesis may well correspond to the underlying events, but alternative mechanisms should be discussed. In addition to the aforementioned lack of a systematic morphometric analysis, the pacing protocol was not selected to fully answer the question of excitation of the donor cell sheets and the electrical activity of the host myocardium below the infarcted zone that is visible on the excitation maps. Thus, the optical mapping of electrical excitation of the donor sheets, showing fast propagation across the transplanted zone, may be explained by both circumscribed electrical coupling in the border zone or diffuse electrical coupling of the whole region forming the interface between host and donor myocardium. With the second hypothesis, the electrotonic host-to-donor interaction, which occurs within &1 length constant (0.15 mm to 0.3 mm), would implicate that the excitation of the donor myocardium follows the excitation of the underlying host myocardium. Such a more diffuse electrical coupling would explain the observed high degree of electrical anisotropy in the donor layer observed 4 weeks after transplantation in the presence of a relatively disordered cell arrangement observed in histology. Along the same line of reasoning, the high degree of anisotropy observed after 4 weeks may reflect a real morphological and functional differentiation of the donor cell layer, or as an alternative, the continuous improvement of cell-to-cell communication. An experiment which might shed light on the topology of electrical transmission between host and donor cells would consist in producing transmural propagation by endocardial stimulation. A further technical aspect that might affect data interpretation relates to the "amplitude" of the optical action potential recorded from surfaces of 3D. The optical action potential, its amplitude and shape, is affected by the mass of cells contributing to changes in fluorescent light at a given instant in time.25,26 It therefore depends on the staining procedure (perfusion versus diffusion), on the mass, types and biological states of non-myocytes adsorbing dye, and on 3D propagation itself. As a consequence, the decrease in optical "amplitude" of action potentials located below the injured area, represented by "white zones" in Figure 3 from Furuta et al, does not exclude normal electrical activity in this region.
It will be very important to follow the mechanism and the phenotypical evolution of electrical host-to-donor coupling at cellular resolution. Besides the myocytemyocyte type of coupling proposed by the authors, a myocytefibroblastmyocyte type of coupling27 or a mixture between the two cannot be ruled out as an explanation of the observed electrical synchrony, especially in the presence of a diffuse donor-to-host interface and the small observed width of the interface.
Finally, the study did not answer another maybe even more important question of all tissue engineering approaches, namely that of the critical size of the implants. How many sheets can be stacked before approaching diffusion limits? In a former study, 4 sheets gave thicker constructs than 1 to 3, but more sheets failed to improve tissue thickness. On the other hand, how many cell sheets will be required to improve the function of a failing human heart?
Despite these open questions the present study provides convincing and important evidence that engineered cardiac tissue constructs electrically couple to the host myocardium when implanted onto the surface of the heart. The question of coupling belongs to the major challenges in this field, because electrical integration is the prerequisite for synchronous mechanical work and protects against arrhythmia. Fifty percent of cardiac deaths in heart failure are arrhythmogenic,28 and proarrhythmic effects could offset any benefit in terms of contractile function. Thus, it is reassuring that, in the present study, coupling occurred rapidly, reproducibly in all experiments, and without specific handling tricks. This implies a very high degree of electrical remodeling activity of cardiac tissues and provides hope that proarrhythmic risks may be not the limiting problem in this and similar tissue engineering approaches.
| Acknowledgments |
|---|
| Footnotes |
|---|
| References |
|---|
|
|
|---|
Related Article:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Circulation Research Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2006 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |