Citation
Pijnappels, D. A. (2009, June 18). Electrophysiological deterioration and resurrection in the scarred heart. Retrieved from
https://hdl.handle.net/1887/13851
Version: Corrected Publisher’s Version
License: Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden
Downloaded from: https://hdl.handle.net/1887/13851
Note: To cite this publication please use the final published version (if applicable).
Chapter V II
E pithelial-to-mesenchymal transformation alters electrical conductivity of WT1+ human epicardial cells
Daniël A. Pijnappels,1,* Noortje A.M. Bax,2,* Elizabeth M. Winter,2 Jerry Braun,4 Dirk L. Ypey,1 John van Tuyn,1 Saskia Maas,2 Heleen Lie-Venema,2 Antoine A.F. de Vries,3 Arnoud van der Laarse,1 Martin J. Schalij,1 Douwe E. Atsma,1 Adriana C. Gittenber- ger-de Groot2
*Authors contributed equally
Departments of Cardiology,1 Anatomy and Embryology,2 Molecular Cell Biology,3 and Cardiothoracic Surgery,4 Leiden University Medical Center, Leiden, The Nether- lands
Submitted for publication
Chapter V II
E pithelial-to-mesenchymal transformation alters electrical conductivity of WT1+ human epicardial cells
Daniël A. Pijnappels,1,* Noortje A.M. Bax,2,* Elizabeth M. Winter,2 Jerry Braun,4 Dirk L. Ypey,1 John van Tuyn,1 Saskia Maas,2 Heleen Lie-Venema,2 Antoine A.F. de Vries,3 Arnoud van der Laarse,1 Martin J. Schalij,1 Douwe E. Atsma,1 Adriana C. Gittenber- ger-de Groot2
*Authors contributed equally
Departments of Cardiology,1 Anatomy and Embryology,2 Molecular Cell Biology,3 and Cardiothoracic Surgery,4 Leiden University Medical Center, Leiden, The Nether- lands
Submitted for publication
Abstract
Epicardial cells are able to undergo epithelial-to-mesenchymal transformation (EMT), thereby contributing to cardiac development. Disturbances in this process of trans- formation are associated with seriously hampered cardiac function. However, there is lacking knowledge regarding the electrophysiological properties of epicardial cells and whether EMT influences electrical conductivity of epicardial cells. Therefore, these aspects were studied in a dedicated in vitro model.
We report that change in morphology and β-catenin staining pattern of cultured adult human epicardium-derived cells (EPDCs) indicated that spontaneous EMT oc- curred in vitro. Micro-electrode arrays were used to investigate electrical conduction across 2 rat cardiomyocyte (CMCs) fields separated by a strip of eGFP-labeled cob- blestone (c)-like EPDCs or spindle (s)-shaped EPDCs. In addition, patch-clamp expe- riments and immunocytochemical analysis for gap junction and ion channel protein expression were performed. All assessments on epicardial cells were compared to cardiac fibroblasts (cFBs) and CMCs. Application of cEPDCs (n=8) resulted in elec- trical coupling of 2 CMC fields within 24h, with a conduction velocity of 2.2±0.7 cm/s. In contrast, sEPDCs (n=7) were not able to electrically connect 2 CMC fields, like cFBs (n=8) (p<0.05). Both the expression of connexin (Cx)40, Cx43 and Cx45 and ion channels (SCN5A, CACNA1C and Kir2.1) were downregulated in sEPDCs as compared to cEPDCs (p<0.05).
Electrical conduction across epicardial cells is influenced by EMT, resulting in signifi- cantly reduced conductivity, associated with conduction block. After EMT, both gap junction and ion channel protein expression was downregulated. This study provides new insights in the importance of EPDCs in cardiac development, and EMT-related cardiac dysfunction.
Introduction
C
ardiogenesis is partly dependent on a proper development of the epicardium.During embryogenesis, the epicardium originates from the pro-epicardial organ (PEO), from which epicardial cells migrate and cover the primitive heart tube. Part of these epicardial cells undergo epithelial-to-mesenchymal transformation (EMT), thereby forming epicardium-derived cells (EPDCs). As a result of EMT, cell-cell and cell-matrix interactions change, allowing EPDCs to migrate into the subepicardium and subsequently into the myocardium.23,39 In the myocardium, EPDCs initially dif- ferentiate into interstitial fibroblasts17 and later on they form smooth muscle cells of the coronary vasculature as well as the coronary adventitial fibroblasts.10,37 Besides this cellular contribution to heart development, EPDCs also have a regulatory role in cardiogenesis by interacting with surrounding structures. Although the mechanisms underlying this regulatory process are largely unknown, it is likely that cell-cell com- munication is of importance. Moreover, several studies demonstrated a crucial role for EPDCs in growth of the compact myocardium and the organization of the myo- cardial architecture.14,16,23 Loss of proper EPDC-function results in a thin hypoplastic myocardium.25,34 Also, an inductive role for EPDCs in the development of the avian Purkinje fiber network of the ventricular conduction system has been found.13,17 Ho- wever, the aforementioned roles of EPDCs do not seem to be restricted to embryonic development, as adult rat epicardial cells delayed the dedifferentiation of rat ventricu- lar cardiomyocytes (CMCs) in vitro.12 In addition, van Tuyn et al. demonstrated that adult human EPDCs can undergo EMT spontaneously an
d obtain characteristics of smooth muscle cells in vitro.35 Interestingly, recent in vivo studies suggest that adult EPDCs can reactivate their embryonic program.38,40 In more detail, induced hyperplastic cardiac growth in adult zebrafish was associated with epicardial expression of embryonic markers such as raldh2 and tbx18. Also, epicardial cells proliferated to expand the epithelial covering of the ventricles. This study sug- gests that adult epicardium is a dynamic tissue that is still able to contribute EPDCs to the adult ventricular wall.38 Furthermore, recent study by Winter et al. demonstrated that adult human EPDCs injected into the infarcted myocardium, preserved cardiac function and reduced remodeling both early and late after the onset of infarction.40 These findings suggest that these cells could be suitable in cardiac cell therapy.
Cardiac cell therapy appears to be of therapeutic value especially if the transplanted cells functionally integrate with native, excitable cardiac tissue. Cell-cell coupling was shown to be an important determinant of integration, as absence of functional gap junctions between transplanted cells and native cardiomyocytes was associated with electrical disturbances.28 However, the working myocardium also contains a large
number of fibroblasts that, in a more passive manner, are involved in electrical con- duction. In other words, cardiac fibroblasts (cFBs) are not excitable and therefore only allow electrical conduction via their gap junctions without any contribution to the electrical current by generation of action potentials.
Despite the intriguing findings on the numerous effects of epicardial cells during car- diac development and their potency to preserve cardiac function after myocardial in- farction, knowledge about their electrical properties is lacking. Furthermore, it is un- known whether and to which extent EMT of epicardial cells influences their capacity to conduct electrical impulses. As illustrated by their wide spectrum of constructive, regulatory and therapeutic effects, this is of certain importance. We therefore deve- loped a controlled in vitro model to study the conductivity of human adult epicardial cells, before and after EMT, cultured in-between 2 fields of CMCs.
Materials and Methods
Animal Experiments and Human Specimens
Animal experiments were approved by the Animal Experiments Committee of the Leiden University Medical Center and conformed to the Guide for the Care and Use of Laboratory Animals as stated by the US National Institute of Health. In addition, all experiments with human tissue specimens were carried out according to the of- ficial guidelines of the Leiden University Medical Center and with the approval of the institutional ethical committee.
Isolation and Culturing of Cardiomyocytes and Cardiac Fibroblasts
Cardiomyocytes (CMCs) and cardiac fibroblasts (cFBs) were dissociated from hearts of 2-day old male Wistar rats of which the ventricles were minced and dissociated with collagenase and DNase, as described previously.31
Harvesting and Preparation of Human Epicardium-Derived Cells
Cultures of human epicardial cells were prepared as described previously.35 When outgrowth of epicardial cells was confluent, the cells were detached from the bottom of the culture dish with trypsin/EDTA (Invitrogen, Paisly, UK) solution and were di- vided into two subcultures. The first subculture was seeded in a high density and cul- tured in a 1:1 mixture of Dulbecco’s modified Eagle’s medium (DMEM) (Invitrogen) and medium 199 (M199) (Invitrogen) containing 100 U/ml penicillin (Invitrogen), 100 μg/ml streptomycin (Invitrogen) and 10% inactivated fetal calf serum (FCSi) (In- vitrogen), to maintain the epithelium-like morphology. The cells in this subculture of
epicardial cells will be referred to as cobblestone-like EPDCs (cEPDCs). The second subculture was seeded in a low density and cultured in aforementioned medium sup- plemented with 20 ng/ml basic fibroblast growth factor (bFGF; Sigma-Aldrich, St.
Louis, USA), to stimulate EMT. This subculture consists of epicardium-derived cells (EPDCs) and will be referred to as spindle-shaped EPDCs (sEPDCs). The purity of the human EPDC cultures was certified with immunohistochemical staining for Wilm’s Tumor-1 protein (WT1) (Calbiochem, San Diego, USA) at a dilution of 1:50.
Immunophenotyping
The surface antigen expression profiles of cEPDCs and sEPDCs were determined by flow cytometry as described previously.36 The antibodies used for flow cytometric analysis are listed in Table 1.
Immunofluorescence Microscopy
cEPDCs and sEPDCs were cultured on glass chamber-slides and subjected to immu- nofluorescent staining as described previously.31 The details of the antibodies used for immunofluorescence microscopy are listed in Table 1. To investigate their morpho- logy, the cells were stained for β-catenin and vimentin. Next, the cells were labeled with antibodies against connexins and ion channels. Antibodies for a voltage-gated sodium channel (SCN5A), inward rectifier potassium channel (Kir2.1) and voltage- gated L-type calcium channel (CACNA1C) were used. Incubation with primary and appropriate secondary antibodies (see Table 1) was carried out overnight and for 2 hr at room temperature, respectively. Nuclei were stained with Hoechst 33342 (diluted 1:1000 in PBS, Invitrogen). Finally, the slides were mounted with Vectashield (Vec- tor, Burlingame, USA). Examination of the slides was performed using a fluorescence microscope equipped with a digital camera (Eclipse, Nikon Europe, Badhoevedorp, the Netherlands). For quantitative analysis of protein expression, at least four images (100x magnification) per type of staining were taken in at least three coverslips. In each image, five areas were random selected to quantify the fluorescent signal (Ima- ge-Pro plus, version 4.1.0.0, Media Cybernetics, Siver Spring, USA).
Micro-Electrode Arrays
To study the functional effect of EMT on electrical conduction across epicardial cells we used a standardized in vitro model, described in our previous studies.30,32 Isolated CMCs were cultured in micro-electrode array culture dishes (MEA, Multichannel Systems, Reutlingen, Germany; number of titanium nitride electrodes: 60; inter-elec- trode distance: 200 μm; electrode diameter: 30 µm). In order to improve attachment of the cells to the glass surface, MEAs were glow-discharged and coated with collagen.
After 3 days of culture, the confluent, synchronously beating monolayer was divided into two fields of CMCs by an a-cellular channel of either (A) 250-270 µm or (B) 350- 370 µm wide, using a laser dissection microscope (P.A.L.M. microlaser system, inclu- ding PALM robosoftware 4.0, Microlaser Technologies GmbH, Bernried, Germany).
This a-cellular channel electrically separated the two CMC fields, and served as site for cell transplantation. In group A, either 50x103 eGFP-labeled cEPDCs or 50x103 sEPDCs were applied in a channel-crossing pattern after ensuring that no cells or cell debris were present in the channel and after confirming the presence of a conduction block between the two CMC fields. Group B received 75x103 eGFP-labeled cEPDCs or 75x103 sEPDCs. In addition, either 50x103 (group A) or 75x103 (group B) cFBs or CMCs were applied in-between the two CMC fields and these groups were compared to EPDC groups. Labeling of cells with eGFP by adenoviral transduction has been described in an earlier study.35 Cells of interest were applied with a pipette mounted in a micro-manipulator and a light microscope (20x magnification). Simultaneous high density mapping of these cultures was performed 24h and 48h after cell seeding, using a dedicated data acquisition system (sampling rate 5 kHz/channel, Multi Chan- nel Systems, Reutlingen, Germany). Electrograms were analyzed off-line using MC- Rack software (version 3.6.8, Multi Channel Systems).
Cell cultures were electrically stimulated via an external pipette electrode producing bipolar rectangular pulses (1.5x threshold, pulse width: 10 ms), placed in close con- tact to the cell culture and at least 5 mm apart from the measurement sites. Cultures were stimulated for at least 30 s, before recordings were started. Conduction velo- cities were calculated from averaged local activation times recorded at eight fixed measuring points, distributed equally over the two lines of electrodes adjacent to the channel.31
Intracellular Measurements
CMCs (1x106) were co-cultured with eGFP-labeled cEPDCs (50x103) or sEPDCs (50x103) on collagen-coated glass coverslips in densities allowing measurements in both single cells and cell clusters. After identification of the eGFP-label using fluorescence micros- copy, changes in membrane potentials of these eGFP-labeled cells were recorded by glass patch-electrodes. Whole-cell recordings were performed 24h after application of eGFP-labeled cells, at 25°C using a L/M-PC patch-clamp amplifier (3 kHz filtering) (List- Medical, Darmstadt, Germany). The pipette solution contained (in mM) 10 Na2ATP, 115 KCl, 1 MgCl2, 5 EGTA, 10 HEPES/KOH (pH 7.4). Tip resistance was 2.0-2.5 MΩ, and seal resistance >1 GΩ. The bath solution contained (in mM) 137 NaCl, 4 KCl, 1.8 CaCl2, 1 MgCl2, 10 HEPES (pH 7.4). For data acquisition and analysis pClamp/Clampex8 soft- ware (Axon Instruments, Molecular Devices, Sunnyvale, USA) was used.
Statistics
Statistical analysis of the electrical conductivity and immunofluorescence data were performed using SPSS14.0 for Windows (SPSS Inc., Chicago, USA). P-values <0.05 were considered statistically significant. Electrical conductivity data were compared with the one-way or two-factor mixed ANOVA test with Bonferroni correction for multiple comparisons, and expressed as mean±SD. Statistical analysis of the immu- nofluorescence data was performed with the one-way ANOVA test with Bonferroni correction for multiple comparisons, and expressed as mean±SEM. For all antigens with the exception of vimentin, the fluorescent signal was compared to the signal in cEPDCs (100%). For vimentin the signal was compared to the signal in sEPDCs (100%).
Results
Analysis of Cell Surface Marker Profile
Epicardial cells from atrial appendages of several human adults were expanded in culture. Purity of the cultures was certified by WT1 staining (Figure 1D,E). In both cEPDCs (Figure 1D) and sEPDCs (Figure 1E) WT1 was predominantly localized in the nuclei. The surface antigen profiles were analyzed by flow cytometry (Table 1).
Consistent with a previous report of van Tuyn et al.35 sEPDCs abundantly expressed endoglin (CD105) at their plasma membrane (Figure 1A, B). The surface of cEPDCs was not decorated with CD105 but contained substantial amounts of vascular cell adhesion molecule-1 (VCAM-1; CD106) (Figure 1A, B, C). For the other surface mar- kers that were tested, cEPDCs and sEPDCs yielded similar results, i.e. there were low surface levels of hyaluronate receptor (CD44), membrane cofactor protein of the complement system (MCP; CD46) and major T-cell antigen (Thy1; CD90). Neither cEPDCs nor sEPDCs expressed the hematopoietic marker CD34 or endothelial mar- kers such as platelet-endothelial cell adhesion molecule-1 (PECAM-1; CD31) and vascular endothelial (VE)-cadherin at their surface (Figure 1).
Immunofluoresence Microscopy
To evaluate whether EMT in epicardial cells may have an effect on their conductivity, these cells were analyzed by immunofluorescence microscopy using markers for con- nexins (Figure 2) and ion channels (Figure 3). Changes in cellular morphology during EMT were confirmed using immunofluorescent staining for β-catenin (Figure 2A1- A4) and vimentin (Figure 3A1-A4). cEPDCs displayed intense β-catenin staining, es- pecially at sites of cell-cell contact, confirming their epithelial nature (Figure 2A1)35.
Figure 1: Flow cytometric analysis of cultured adult human EPDCs before (cEPDCs) and after (sEPDCs) EMT (A). Histograms of CD105 (B) and CD106 (C) are shown with isotype control (dashed line) and the specific signal (solid line). EPDC cultures were stained using WT1 to certify the purity of the culture (D,E). Scale bar, 20μm.
After EMT, the levels of β-catenin are reduced by 30.92% (cEPDC [n=145] vs sEPDCs [n=100], P<0.05, Figure 2A5) and the protein was redistributed to the cytoplasm in sEPDCs (Figure 2A2) where it was also found in cFBs [n=90] (Figure 2A3).
In CMCs [n=92] intense expression of β-catenin was concentrated at sites of cell-cell contact (Figure 2A4). Cx40 (Figure 2B1, B2) and Cx45 (Figure 2D1, D2) were weakly present in the cytoplasm of cEPDCs and sEPDCs with some Cx45 staining in the nucleus (Figure 2D1, D2). However, the Cx40 and Cx45 levels were higher in cEPDCs [n=145, n=165] by 72.15% [n=145] and 27.55% [n=165], respectively than in sEPDCs [n=90, n=115] (P<0.05; Figure 2B5, D5). Cx43 was present in the cytoplasm of cEP- DCs and between adjacent cEPDCs. The punctated pattern of the staining reflected the presence of gap junctions (Figure 2C1). Cx43 levels were reduced in the cytoplasm of sEPDCs and between adjacent sEPDCs (Figure 2C2) compared to cEPDCs. After EMT the amount of Cx43 decreased by 45.99% (cEPDC [n=120], sEPDCs [n=145],
Antigen Source Clone Isotype Label Species
Β-catenin BD 14 IgG1 - Mouse
CACNA1C (A-20) SC - - - Goat
CD31 CLB HEC/75 IgG1 FITC Mouse
CD34 BD 8G12 IgG1 PE Mouse
CD44 BD G44-26 C26 IgG2b PE Mouse
CD46 BD E4.3 IgG2a - FITC Mouse
CD90 BD 5E10 IgG1 - FITC Mouse
CD105 Bio SN6 IgG1 - PE Mouse
CD106 BD 51-10C9 IgG1 - PE Mouse
Cx40 (C-20) SC - - - Goat
Cx43 (C-363-382) SC - - - Rabbit
Cx45 (C-19) SC - - - Goat
Kir2.1 (N-18) SC - - - Goat
SCN5a (C-20) SC - - - Goat
VE-cadherin SC F8 IgG1 - Mouse
Vimentin SA V9 IgG1 - Cy3 Mouse
Goat IgG MP - - - Alexa Fluor 568 Rabbit
Mouse Ig Dako - - - FITC Rabbit
Mouse Ig BD - - - FITC Goat
Mouse IgG1 BD X56 IgG1 PE Rat
Rabbit IgG MP - - Alexa Fluor 568 Goat
Table 1. Abbreviations used in this table: BD, BD Biosciences; Bio, Biocarta; CLB, Sanquin; Dako, Dako Cytomation; MP, Molecular Probes; SA, Sigma Aldrich; SC, Santa Cruz; FITC, fluorescein isothiocya- nate; PE, phycoerythrin. All antibody preparations were used at the concentrations recommended by the suppliers. For an explanation of the abbreviations of the antigens, see the main text of the paper.
P<0.05, Figure2C5). Comparison of cardiac connexin levels between neonatal rat CMCs [n=145] and cFBs [n=100] revealed that Cx43 amounts were ±1.6 and ±3.2 times higher in cEPDCs, respectively (both P<0.05, Figure 2C).
sEPDCs cEPDCs
cFBs
CMCs
Figure 2: Epithelium-to-mesenchymal transformation is accompanied by a decrease in β-catenin and connexins expression levels. Immunofluorescence microscopy of cEPDCs (before EMT) and sEPDCs (af- ter EMT) labeled with antibodies directed against β-catenin (A) Cx40 (B), Cx43 (C), and Cx45 (D). CMCs and cFBs were used as controls. The expression of β-catenin is strongly expressed at the cell borders of cEPDCs and CMCs. Expression of β-catenin is redistributed to cytoplasm in sEPDCs and cFBs. Quantifi- cation of the fluorescence (panels A5-D5) showed a reduction in cardiac connexin levels following EMT.
*P<0.05, vs all groups, #P<0.05. Scale bar, 20 μm.
The organization of the intermediate cytoskeleton filaments, stained with vimentin, was a reference for cellular morphology. In cEPDCs [n=95, 80.11%] the intermediate filaments are tightly packed reflecting their epithelial nature (Figure 3A1).
sEPDCs cEPDCs
cFBs
CMCs
Figure 3: Immunofluorescent staining of ion channels in epicardial cells before EMT (cEPDCs) and after EMT (sEPDCs). (B-D): Immunofluorescence analysis of Kir2.1, SCN5a, and CACNA1C in EPDCs, CMCs and cFBs. Expression of ion channels was reduced by EMT. Vimentin was used to determine cell morpho- logy (A). *P<0.05, vs all groups, #P<0.05. Scale bar, 20 μm.
After EMT, in the sEPDCs [n=77, 100%] the intermediate filaments are more visible (Figure 3A2) compared to cEPDCs. The intermediate filament organization in cFBs was similar to that in the sEPDCs, but staining for vimentin in cFBs was reduced by 36.54% compared to sEPDCs [n=70] (Figure 3A2, A3, A5). Staining for vimentin was weakly present in the cytoplasm of rat CMCs [n=65, 12.73%] (Figure 3A4).
Staining for the inward rectifier potassium channel (Kir2.1) revealed presence of this channel in the cytoplasm of cEPDCs [n=114], while the expression levels were rather heterogeneous among cells (Figure 3B1). Expression levels of Kir2.1 were reduced in sEPDCs by 34.48% (P<0.05, n=145) compared to cEPDCs (100%) (Figure 3B1, B2, B5).
Expression of SCN5A, encoding the voltage-gated fast sodium channel, was present in both cEPDCs [n=155] and sEPDCs [n=108], however, in cEPDCs the expression of SCN5A levels were higher (38.49%, P<0.05, Figure 3C1, C5) and the distribution pattern was more distinct as compared to sEPDCs (Figure 3C2). SCN5A was also detected in CMCs [n=90] and cFBs [n=80], however, the amount of SCN5A in CMC was ±1.4 times higher than in cEPDCs (p<0.05) (Figure 3C1, C4, C5).
Low amounts of voltage-dependent L-type calcium channels (CACNA1C) were pre- sent in the cytoplasm of cEPDCs [n= 150] and sEPDCs [n=100], although cEPDCs contained 20.08% more of this protein than sEPDCs (P>0.05; Figure 3D1, D2, D5).
Comparison of expression levels between all different cell types revealed that the ex- pression of CACNA1C is almost equal in cEPDCs and CMCs (Figure 3D1, D4). Fur- thermore, in cEPDCs the amount of CACNA1C expression is ±2.0 times higher than in cFBs (Figure 3D1, D3, D5).
Electrical Conduction Across EPDCs before and after EMT
To study the effect of EMT on electrical properties of epicardial cells, we examined the conduction properties of these cells, both before and after EMT, using multi-elec- trode arrays. Two days after culture, each spontaneously beating monolayer of CMCs was divided in two electrically isolated fields by a laser-dissected a-cellular channel.
The channel width was varied (either 270±15 µm or 360±20 µm) to allow further characterization of conduction across the cells of interest. After two days, all cultures were thoroughly screened for inhomogeneities as assessed by light-microscopy and extracellular electrogram recordings. As a result, over 70% of the cultures were inclu- ded for further studies. Next, cEPDCs (pre-EMT) (n=8) were applied in-between the two CMC fields, which resulted in a confluent strip of cells connecting both fields, within 24h after application. Multi-electrode recordings of electrical conduction across these cEPDCs and adjacent CMC fields showed persistent electrical interac- tion between the CMC fields. Electrograms recorded from the site of cEPDCs showed their ability to conduct electrical impulses over a 270 µm wide channel, resulting in
electrical activation of the distal CMC field. Further analysis of electrogram characte- ristics confirmed a substantial conduction delay between the two CMC fields, which resulted from relatively slow conduction across these cEPDCs (4.2±0.9 cm/s) (Figure 4A, C). In order to study the functional effects of EMT, the separated CMC fields were now connected by transplanted sEPDCs (post-EMT) (n=8). Similar to cEPDCs, these sEPDCs were also acting as a conductive cellular bridge in-between the two CMC fields within 24h after application. However, EPDC-related conduction delays were significantly increased, and consequently, conduction velocity across sEPDCs was significantly decreased as compared to cEPDCs, now reaching values of 1.8±1 cm/s. To put these values in perspective, we also applied cFBs or CMCs in-between the two CMC fields.
Transplantation of cFBs (n=9) resulted also in electrical conduction from one CMC field to the other. Electrogram analysis confirmed the occurrence of significant con- duction delays, comparable to those found after transplantation of sEPDCs. Con- duction velocity across cFBs (2.0±0.8 cm/s) was comparable to those across sEPDCs, however, significantly lower as compared to conduction velocities across cEPDCs.
Transplantation of CMCs (n=15) resulted in a confluent cell strip, which electrically coupled the adjacent CMC fields, associated with conduction delays comparable to those measured in the CMC fields. As a result, the transplanted CMCs were able to conduct the electrical impulses as fast as the host CMC fields (18±1.9 cm/s), thereby forming a homogeneous syncytium with adjacent cardiac tissue. Importantly, follow- up till 48h after transplantation did not show any significant differences concerning conduction velocity across transplanted cEPDCs (n=8), sEPDCs (n=8), cFBs (n=9), or CMCs (n=15) (respectively, 4.3±1.4 cm/s, 2.0±0.9 cm/s, 2.1±1.1 cm/s, 18.3±1.5 cm/s) (Figure 4A, C). Except from CMCs, all other three cell types showed decremental con- duction, defined here as decreased electrogram amplitudes combined with increased conduction times as compared to control cultures (CMCs).
To further define this depressed conduction across EPDCs, we used a similar model, but with increased channel width (360 µm), to assess the magnitude of decremental conduction in EPDCs. In this model, transplantation of cEPDCs (n=7) resulted in electrical restoration between the two CMC fields, although with extensive conduc- tion delay and slow conduction (2.2±0.7 cm/s) (Figure 4B, D). Interestingly, trans- plantation of sEPDCs did not result in electrical interaction between the two adjacent CMC fields. In fact, these sEPDCs (n=7) now imposed a cellular conduction block, thereby prohibiting electrical impulse conduction from one CMC field to the other.
Of note, transplantation of cFBs (n=7) showed the same phenomenon, in which these cells were not able to conduct the electrical impulse across the whole width of the strip, but were active as a conduction block between the two CMC fields. In contrast,
CMCs (n=12) transplanted in-between adjacent cardiac tissue resulted in electrical coupling of the CMC fields, associated with conduction velocities (18±1.2 cm/s) com- parable to those across the surrounding native CMCs (20±1.5 cm/s). Follow-up at 48h did not show any significant differences as compared to conduction velocities across cells measured 24h after transplantation (Figure 4B, D).
Figure 4: Overview of conduction velocities, and corresponding conduction delays, measured in EPDCs in-between two adjacent fields of CMCs (A,B). Conduction velocity was significantly decreased in sEP- DCs, thereby resembling the conduction properties of those measured in cardiac fibroblasts. In fact, with increasing distance (360 μm) sEPDCs were no longer able to conduct the electrical impulse across the channel, resulting in asynchronized beating of the two CMC fields. No significant differences were found during follow-up. Extracellular electrograms derived from co-cultures of labeled EPDCs and neonatal rat cardiomyocytes, 24h and 48h after plating (C,D). These electrograms clearly show the decremental nature of conduction across EPDCs, regardless of EMT. However, as conduction across these EPDCs depends on electrotonic interaction, the decrease in connexin levels that occurs in these cells during EMT is expected to result in conduction block over a certain distance, which can be appreciate in panel D. #P<0.05, vs sEPDCs.
Intracellular Measurements of Cardiac Impulse Conduction across cEPDCs and sEPDCs
To further investigate the decremental conduction across EPDCs, we performed whole-cell membrane potential measurements to assess electrical coupling of the- se cells, 24h and 48h after plating. Experiments were first conducted on single cells (controls) and then on cells adjacent to 3-4 spontaneously and synchronously beating CMCs, as assessed by phase-contrast microscopy.
Measurements in both single cEPDCs (n=4) and sEPDCs (n=3) showed steady de- polarized resting membrane potentials (as compared to CMCs (31)) of, respectively, -12±6 mV and -7±4 mV (ns), with no significant differences at follow-up (Figure 5A).
However, recordings of invading cardiac impulses in EPDCs adjacent to beating CMCs, demonstrated significant differences between cEPDCs and sEPDCs.
Figure 5: Intracellular recording from single EPDCs and EPDCs adjacent to clusters of CMCs at 24h and 48h after plating. Resting membrane potentials are measured in a single cEPDCs or sEPDCs. No significant differences were found for these values (A). Resting membrane potentials measured in a single EPDC ad- jacent to excitable CMCs (3-4) was significantly lower than that in CMC-flanked sEPDCs (B1), consistent with decreased intercellular conductance in the latter cell type. No significant differences were found bet- ween 24h or 48h follow-up (B2). In addition, the amplitude of the invading electrical impulse was signifi- cantly higher in cEPDCs as compared to sEPDCs, again indicating that gap junction coupling was decreased upon EMT in EPDCs. No significant differences were found during follow-up. *P<0.05, vs cEPDCs.
Conducted impulses in the cEPDCs were characterized by more negative resting membrane potentials (-29±8 mV) (Figure 5B1) and higher impulse amplitudes (35±5 mV) (Figure 5B2), as compared to those measured in sEPDCs (-18±6 mV, and 23±8 mV, respectively). No significant differences were found in resting membrane poten- tials and impulse amplitudes at follow-up (Figure 5B).
Discussion
As far as we know, this is the first study describing the electrical behavior of adult human epicardial cells before and after EMT. The key findings of the present study are: 1) Adult epicardial cells are able to connect to functionally active CMCs and to conduct electrical impulses over significant distances, although this is characterized by slow and decremental conduction. 2) Epithelial-to-mesenchymal transformation in adult epicardial cells is associated with a decrease in conduction velocity, which is consistent with a decrease in connexin and ion channels protein expression levels.
Immunophenotypic Characterization of Adult EPDCs
Epicardial cells isolated from adult atrial appendages are characterized by flow cyto- metric analysis before and after EMT. Both cEPDCs and sEPDCs are characterized by the expression of vimentin, consistent with their mesothelial origin.11 The mor- phological changes of epicardial cells during EMT are accompanied by a reduction of β-catenin level. Flow cytometric analysis of cEPDCs and sEPDCs revealed that cEP- DCs were characterized by expression of CD106 (VCAM-1) and absence of CD105 (endoglin). In contrast, sEPDCs were positive for CD105 and negative for CD106.
Previous studies revealed that VCAM-1 enhances adhesion, concomitant with in- creased association of β-catenin with intercellular junctions and VCAM-1 restrict TGFβ-stimulated EMT by inhibiting TGFβ3-mediated loss of β-catenin from inter- cellular junctions.11 These findings confirm that the surface expression of VCAM-1 and β-catenin is supportive for the epithelial nature of cEPDCs. Epicardial cells loose their epithelial characteristics under influence of TGFβ.7 Endoglin (CD105) is an an- cillary TGFβ-receptor that can induce EMT.8,33 Thus, the phenotypic signs of EMT were confirmed by expression of CD105 and decreased expression of β-catenin at the plasma membrane of sEPDCs.
Role of EPDCs in Conductivity
The electrophysiological properties of cultured adult epicardial cells and their role in electrical conduction have not been studied in detail until now. During embryo-
genesis, in the splanchnopleuric mesoderm two crescent-shaped heartforming fields develop.23,26,39 From a second heart field, also positioned in this splanchnopleuric me- soderm, that can be divided into two fields related to the cranio-caudal axis of the pri- mary heart, cells are added to both the outflow (anterior heart field (AHF)) and inflow (posterior heart field (PHF)) of the heart.1 The PHF contributes to the development of the cardiac conduction system,21 and the epicardium that covers the heart develops also from the PHF.23,25 Due to this shared PHF-origin in early development, these cells can have several characteristics in common. Through differentiation these cells will distinguish themselves from each other. After the heart is covered by epicardium, these cells undergo EMT and migrate into the subepicardial space and into the myo- cardium, where they will differentiate into interstitial fibroblasts, smooth muscle cells (SMCs) and fibroblasts of the coronary vasculature. EPDCs are also involved in the induction of Purkinje fiber differentiation of the ventricular conduction system.18,20 Recent literature suggests that the epicardium is also a source of cardiac progenitors.
This suggestion is based on the fact that WT1-positive pro-epicardial cells are likewi- se descendants of Nkx2.5+/Isl+ precursors like most cardiomyocytes.42 This finding supports the common origin of CMC and epicardial cells from the PHF. van Tuyn et al. described that adult EPDCs and CMCs share the expression of GATA4 and car- diac troponinT.42 Although the origin, the role in development, and the presumed dif- ferentiation of pro-epicardial cells into CMCs suggest a role in conduction, still little is known about these aspects. The present in vitro electrophysiological study demon- strates that cultured adult cEPDCs and sEPDCs can connect to functionally active CMCs and are able to conduct a cardiac action potential, which capacity is affected by EMT, as reflected by a decrease in conduction velocity. EMT in EPDCs is probably the onset for differentiation. The differentiation of EPDCs to electrically active SMCs of the coronary vasculature is dependent on signals from the endothelium of the co- ronary vessels.23 EPDCs can also differentiate into interstitial fibroblasts, which are unexcitable cells that form supportive layers within the myocardium, while modula- ting cardiac action potential propagation.4,15 The present data show that conduction velocity of sEPDCs is decreased compared to that in cEPDCs, but this velocity is comparable to that found in cFBs. Therefore, EMT of epicardial cells might represent the onset of differentiation into cardiac fibroblast-like cells.
Role of Epithelial-to-Mesenchymal Transformation in Conductivity
EMT is a critical process in the development of the heart. Not only for the develop- ment of cardiac structures, like the cardiac valves,27 but also in the development of gap junctions.29,41 Previous studies have shown that classical cadherins support gap junctional stabilization. Cadherins are cell surface molecules anchored via catenins
to the cytoskeleton.24 Cadherins and catenins are localized at cell-cell adherent junc- tions, especially in cells with an epithelial-like morphology. In the heart, Cx43 colo- calizes with β-catenin in the junctional membrane2 and the association of β-catenin and Cx43 is required for the development of gap junctions.41 The present data show a decrease in β-catenin and Cx43 expression levels during EMT. Previous studies have shown that N-cadherin and β-catenin control targeting of Cx43 to adherens juncti- ons and that the stabilization of Cx40 and Cx43 can be regulated by the N-cadherin/
β-catenin complex.22 The present results show that expression of CX43 and Cx40 in EPDCs is downregulated by EMT. This downregulation is associated with a decrease of β-catenin, thereby probably decreasing the adhesion and connexin expression le- vels.
Previously, it was shown that EMT of human epithelial cells is accompanied by downregulation of P-cadherin and Cx26 promoter activity,9 and that in embryonic carcinoma cells, EMT is followed by repression of Cx43.9 In cardiac-specific tamoxi- fen-inducible Cre transgenic mice with a floxed N-cadherin gene, it is shown that loss of N-cadherin leads to alterations in Cx40 and Cx43. These alterations result in loss of functional gap junctions with consequent cellular uncoupling and diminished conduction velocity.22 Likewise, the decrease in conduction velocity in human adult epicardial cells by EMT is associated with less cell-cell coupling through downregu- lation of β-catenin and consequent decreased expression levels of connexins and ion channels.
Little is known about the effect of EMT on ion channel formation or stabilization. The present data show that epicardial cells are not excitable, and that ion channel protein expression levels in epicardial cells are decreased under influence of EMT. Earlier it has been shown that metanephrogenic mesenchyme-to-epithelium transition (MET) induced profound expression changes of ion channels.19 This effect is mediated by E- cadherin and β-catenin, factors that play a crucial role in early epithelial polarization by mediating cell-cell adherens junctions. E-cadherin is also important for the inte- gration and retention of Na+-K+-ATPase in membrane-cytoskeleton complexes.19 In atrial myocyte cultures intracellular measurements show that transforming growth factor beta1 (TGF-β1), which is also an EMT stimulator of epicardial cells,7 decrea- ses cardiac muscle L-type Ca2+ channels.3 TGF-β1 also decreases epithelial sodium channel functionality and thereby decreases the electrical current in renal collecting ducts.5 Although epicardial cells appear unable to generate intrinsic action potentials, the downregulation of ion channel protein expression levels after EMT suggests that these decreases in protein expression may contribute to the EMT-related changes in conduction velocity.
Intracellular electrical recordings show steady, but depolarized resting membrane
potentials in single adult EPDCs before and after EMT. This depolarized condition does not allow a cell to be excitable. Since in this study, ion channel functionality and activity in EPDCs is assessed only indirectly, and it remains to be investigated in more detail. Immunofluoresecence microscopy showed the presence of a variety of ion channels in EPDCs, which could be involved in the generation of action poten- tials. However, the relatively low velocity by which the action potential is conducted across EPDCs and the decremental nature of conduction indicates that EPDCs pas- sively conduct action potentials from the CMCs in stead of actively contributing to the conduction process by excitation. Nevertheless, records of invading cardiac elec- trical impulses in EPDCs adjacent to CMCs, demonstrate significant differences with regard to impulse conduction between cEPDCs and sEPDCs. In more detail, resting membrane potential and amplitudes of invading electrical impulses are influenced by EMT. Resting membrane potential is mainly determined by the inwardly rectifying K+ channel, Kir2.1.6 Although it is not yet clear whether this ion channel is functional in EPDCs. We conclude from the records that both cEPDCs and sEPDCs receive a hyperpolarizing influence from the adjacent CMCs through electrical coupling, with a stronger influence on the cEPDCs.
Limitations
In this study we compared adult human EPDCs from multiple donors were compared with neonatal rat CMCs and rat cFBs. Given possible species- and individual-specific differences, ideally EPDCs, CMCs and cFBs from the same human patient(s) should have been used for our experiment. However, there are considerable technical and ethical objections to their use. Furthermore, the use of double patch clamp techni- ques could refine the coupling resistance measurements presented in this study.
Conclusions
Electrical impulse transmission across adult human epicardial cells is characterized by slow and decremental conduction. Importantly, EMT in adult human epicardial cells is associated with a decrease in conduction velocity, which is most likely me- diated by decreases in connexin and ion channel protein expression levels, strongly affecting their intercellular conductivity. This is of importance as epicardial cells are crucial for proper cardiac development and might be applicable for cardiac regene- ration therapy.40 Future experiments with adult epicardial cells in models of cardiac injury40 and disturbed EMT25 may further emphasize the importance of the present findings with regard to cardiac development and repair.
Acknowledgements
We gratefully acknowledge Jan Lens (Department of Anatomy and Embryology, LUMC, Leiden) for his assistance in the preparation of the figures.
Reference List
1. Abu-Issa R, Waldo K and Kirby ML. Heart fields: one, two or more? Dev Biol 272: 281-285, 2004.
2. Ai Z, Fischer A, Spray DC, Brown AM and Fishman GI. Wnt-1 regulation of connexin43 in cardiac myocytes. J Clin Invest 105: 161-171, 2000.
3. Avila G, Medina IM, Jimenez E, Elizondo G and Aguilar CI. Transforming growth factor-beta1 de- creases cardiac muscle L-type Ca2+ current and charge movement by acting on the Cav1.2 mRNA.
Am J Physiol Heart Circ Physiol 292: H622-H631, 2007.
4. Camelliti P, Borg TK and Kohl P. Structural and functional characterisation of cardiac fibroblasts.
Cardiovasc Res 65: 40-51, 2005.
5. Chang CT, Hung CC, Chen YC, Yen TH, Wu MS, Yang CW, Phillips A and Tian YC. Transforming growth factor-beta1 decreases epithelial sodium channel functionality in renal collecting duct- cells via a Smad4-dependent pathway. Nephrol Dial Transplant 23: 1126-1134, 2008.
6. Chilton L, Ohya S, Freed D, George E, Drobic V, Shibukawa Y, Maccannell KA, Imaizumi Y, Clark RB, Dixon IM and Giles WR. K+ currents regulate the resting membrane potential, prolifera- tion, and contractile responses in ventricular fibroblasts and myofibroblasts. Am J Physiol Heart Circ Physiol 288: H2931-H2939, 2005.
7. Compton LA, Potash DA, Mundell NA and Barnett JV. Transforming growth factor-beta induces loss of epithelial character and smooth muscle cell differentiation in epicardial cells. Dev Dyn 235:
82-93, 2006.
8. Dallas NA, Samuel S, Xia L, Fan F, Gray MJ, Lim SJ and Ellis LM. Endoglin (CD105): a marker of tumor vasculature and potential target for therapy. Clin Cancer Res 14: 1931-1937, 2008.
9. de Boer TP, van Veen TA, Bierhuizen MF, Kok B, Rook MB, Boonen KJ, Vos MA, Doevendans PA, de Bakker JM and van der Heyden MA. Connexin43 repression following epithelium-to-mesenchyme transition in embryonal carcinoma cells requires Snail1 transcription factor. Differentiation 75:
208-218, 2007.
10. Dettman RW, Denetclaw W, Ordahl CP and Bristow J. Common epicardial origin of coronary vascular smooth muscle, perivascular fibroblasts, and intermyocardial fibroblasts in the avian heart. Dev Biol 193: 169-181, 1998.
11. Dokic D and Dettman RW. VCAM-1 inhibits TGFbeta stimulated epithelial-mesenchymal trans formation by modulating Rho activity and stabilizing intercellular adhesion in epicardial mesothe lial cells. Dev Biol 299: 489-504, 2006.
12. Eid H, Larson DM, Springhorn JP, Attawia MA, Nayak RC, Smith TW and Kelly RA. Role of epi- cardial mesothelial cells in the modification of phenotype and function of adult rat ventricular myocytes in primary coculture. Circ Res 71: 40-50, 1992.
13. Eralp I, Lie-Venema H, Bax NAM, Wijffels MC, Van der Laarse A, DeRuiter MC, Bogers AJ, Van Den Akker NM, Gourdie RG, Schalij MJ, Poelmann RE and Gittenberger-de Groot AC. Epicar-
dium-derived cells are important for correct development of the Purkinje fibers in the avian heart.
Anat Rec 288A: 1272-1280, 2006.
14. Eralp I, Lie-Venema H, DeRuiter MC, Van Den Akker NM, Bogers AJ, Mentink MM, Poelmann RE and Gittenberger-de Groot AC. Coronary artery and orifice development is associated with proper timing of epicardial outgrowth and correlated Fas ligand associated apoptosis patterns. Circ Res 96:526-534, 2005.
15. Gaudesius G, Miragoli M, Thomas SP and Rohr S. Coupling of cardiac electrical activity over ex- tended distances by fibroblasts of cardiac origin. Circ Res 93: 421-428, 2003.
16. Gittenberger-de Groot AC, Vrancken Peeters M-PFM, Bergwerff M, Mentink MMT and Poelmann RE. Epicardial outgrowth inhibition leads to compensatory mesothelial outflow tract collar and ab- normal cardiac septation and coronary formation. Circ Res 87: 969-971, 2000.
17. Gittenberger-de Groot AC, Vrancken Peeters M-PFM, Mentink MMT, Gourdie RG and Poelmann RE. Epicardium-derived cells contribute a novel population to the myocardial wall and the atrio- ventricular cushions. Circ Res 82: 1043-1052, 1998.
18. Gourdie RG, Wei Y, Kim D, Klatt SC and Mikawa T. Endothelin-induced conversion of embryonic heart muscle cells into impulse-conducting purkinje fibers. Proc Natl Acad Sci U S A 95: 6815- 6818, 1998.
19. Huber SM, Braun GS, Segerer S, Veh RW and Horster MF. Metanephrogenic mesenchyme-to-epi- thelium transition induces profound expression changes of ion channels. Am J Physiol Renal Physiol 279: F65-F76, 2000.
20. Hyer J, Johansen M, Prasad A, Wessels A, Kirby ML, Gourdie RG and Mikawa T. Induction of purkinje fiber differentiation by coronary arterialization. Proc Natl Acad Sci U S A 96: 13214- 13218, 1999.
21. Jongbloed MR, Mahtab EAF, Blom NA, Schalij MJ and Gittenberger-de Groot AC. Development of the cardiac conduction system and the possible relation to predilection sites of arrhythmogenesis.
ScientificWorldJournal 239-269, 2008.
22. Li J, Patel VV, Kostetskii I, Xiong Y, Chu AF, Jacobson JT, Yu C, Morley GE, Molkentin JD and Radice GL. Cardiac-specific loss of N-cadherin leads to alteration in connexins with conduction slowing and arrhythmogenesis. Circ Res 97: 474-481, 2005.
23. Lie-Venema H, van den Akker NMS, Bax NAM, Winter EM, Maas S, Kekarainen T, Hoeben RC, DeRuiter MC, Poelmann RE and Gittenberger-de Groot AC. Origin, fate, and function of epicardium-derived cells (EPCDs) in normal and abnormal cardiac development. ScientificWorld Journal 7: 1777-1798, 2007.
24. Luo Y, High FA, Epstein JA and Radice GL. N-cadherin is required for neural crest remodeling of the cardiac outflow tract. Dev Biol 299: 517-528, 2006.
25. Mahtab EAF, Wijffels MCEF, van den Akker NMS, Hahurij ND, Lie-Venema H, Wisse LJ, DeRuiter MC, Uhrin P, Zaujec J, Binder BR, Schalij MJ, Poelmann RE and Gittenberger-de Groot AC. Car- diac malformations and myocardial abnormalities in podoplanin knockout mouse embryos: cor relation with abnormal epicardial development. Dev Dyn 237: 847-857, 2008.
26. Männer J. Extracardiac tissues and the epigenetic control of myocardial development in vertebrate embryos. Ann Anat 188: 199-212, 2006.
27. Mercado-Pimentel ME and Runyan RB. Multiple transforming growth factor-beta isoforms and re- ceptors function during epithelial-mesenchymal cell transformation in the embryonic heart. Cells Tissues Organs 185: 146-156, 2007.
28. Leobon B, Garcin I, Menasche P, Vilquin JT, Audinat E, Charpak S. Myoblasts transplanted into rat infarcted myocardium are functionally isolated from their host. Proc Natl Acad Sci U S A 100: 7808-7811, 2003
29. Nishii K, Kumai M and Shibata Y. Regulation of the epithelial-mesenchymal transformation through gap junction channels in heart development. Trends Cardiovasc Med 11: 213-218, 2001.
30. Pijnappels DA, Schalij MJ, Ramkisoensing AA, van Tuyn J, de Vries AA, Van der Laarse A, Ypey DL and Atsma DE. Forced alignment of mesenchymal stem cells undergoing cardiomy
ogenic differentiation affects functional integration with cardiomyocyte cultures. Circ Res 103:
167-176, 2008.
31. Pijnappels DA, Schalij MJ, van Tuyn J, Ypey DL, de Vries AA, van der Wall EE, Van der Laarse A and Atsma DE. Progressive increase in conduction velocity across human mesenchymal stem cells is mediated by enhanced electrical coupling. Cardiovasc Res 72: 282-291, 2006.
32. Pijnappels DA, van Tuyn J, de Vries AA, Grauss RW, Van der Laarse A, Ypey DL, Atsma DE and Schalij MJ. Resynchronization of separated rat cardiomyocyte fields with genetically modified human ventricular scar fibroblasts. Circulation 116: 2018-2028, 2007.
33. Schmidt-Weber CB, Letarte M, Kunzmann S, Ruckert B, Bernabeu C and Blaser K. TGF-{beta} sig- naling of human T cells is modulated by the ancillary TGF-{beta} receptor endoglin. Int Immunol 17: 921-930, 2005.
34. Van Loo PF, Mahtab EAF, Wisse LJ, Hou J, Grosveld F, Suske G, Philipsen S and Gittenberger- de Groot AC. Transcription Factor Sp3 knockout mice display serious cardiac malformations. Mol Cell Biol 27: 8571-8582, 2007.
35. van Tuyn J, Atsma DE, Winter EM, van der Velde-van Dijke, Pijnappels DA, Bax NA, Knaan- Shanzer S, Gittenberger-de Groot AC, Poelmann RE, Van der Laarse A, van der Wall EE, Schalij MJ and de Vries AA. Epicardial cells of human adults can undergo an epithelial-to-mesenchymal transition and obtain characteristics of smooth muscle cells in vitro. Stem Cells 25: 271-278, 2007.
36. van Tuyn J, Knaan-Shanzer S, van de Watering MJ, de Graaf M, Van der Laarse A, Schalij MJ, van der Wall EE, de Vries AA and Atsma DE. Activation of cardiac and smooth muscle-specific genes in primary human cells after forced expression of human myocardin. Cardiovasc Res 2005.
37. Vrancken Peeters M-PFM, Gittenberger-de Groot AC, Mentink MMT and Poelmann RE. Smooth muscle cells and fibroblasts of the coronary arteries derive from epithelial-mesenchymal transfor - mation of the epicardium. Anat Embryol 199: 367-378, 1999.
38. Wills AA, Holdway JE, Major RJ and Poss KD. Regulated addition of new myocardial and epicar- dial cells fosters homeostatic cardiac growth and maintenance in adult zebrafish. Deve
lopment 135: 183-192, 2008.
39. Winter EM and Gittenberger-de Groot AC. Cardiovascular development: towards biomedical ap - plicability : Epicardium-derived cells in cardiogenesis and cardiac regeneration. Cell Mol Life Sci 64: 692-703, 2007.
40. Winter EM, Grauss RW, Hogers B, van Tuyn J, van der Geest RJ, Lie-Venema H, Vicente-Steijn R, Maas S, DeRuiter MC, deVries AA, Steendijk P, Doevendans PA, Van der Laarse A, Poelmann RE, Schalij MJ, Atsma DE and Gittenberger-de Groot AC. Preservation of left ventricular function and attenuation of remodeling after transplantation of human epicardium-derived cells into the in- farcted mouse heart. Circulation 116: 917-927, 2007.
41. Wu JC, Tsai RY and Chung TH. Role of catenins in the development of gap junctions in rat cardiomyocytes. J Cell Biochem 88: 823-835, 2003.
42. Zhou B, Ma Q, Rajagopal S, Wu SM, Domian I, Rivera-Feliciano J, Jiang D, von GA, Ikeda S, Chien KR and Pu WT. Epicardial progenitors contribute to the cardiomyocyte lineage in the developing heart. Nature 454: 109-113, 2008.
