From cardiogenesis to cardiac regeneration : focus on epicardium-derived cells
Winter, E.M.
Citation
Winter, E. M. (2009, October 15). From cardiogenesis to cardiac regeneration : focus on epicardium-derived cells. Retrieved from https://hdl.handle.net/1887/14054
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/1887/14054
Note: To cite this publication please use the final published version (if applicable).
Winter EM1, Hogers B1, Lie-Venema H1, van der Graaf LM1, Bax NAM1, Swildens J2, Maas S1, Nauerth A7, Dorjée AL3, Salvatori DCF1, Smits AM2, Klautz RJM4, van der Weerd L1,5, de Vries AAF2,
Doevendans PA6, Poelmann RE1, Gittenberger-de Groot AC1
The potential of epicardium in the first week after myocardial infarction
Submitted for publication
4
Abstract
BackgroundRecently, adult epicardium-derived cells (EPDCs) have been recognized as promising candidates for cellular cardiac regeneration therapy, improving cardiac function at 2 and 6 weeks after myocardial infarction (MI). We studied the potential underlying mechanism in early ischemia.Methods
EPDCs, derived from human adult atrial appendages, or control vehicle were injected into the infarcted myocardium of immunodeficient NOD/scid mice. Cardiac function (9.4T MRI) and histological properties were evaluated at day 2, 4 and 7.
Results
Engrafted EPDCs demonstrated a myofibroblast phenotype. EPDC transplantation improved cardiac function, with attenuated remodeling already starting at day 2, suggesting an early paracrine effect. Values did not surpass baseline, indicating preserved function, opposed to regained function. EPDC-injected hearts showed earlier (day2) upregulation of Wilms’ Tumor 1 (WT1) expression in the host epicardium and in cells scattered through the infarcted area. WT1 is necessary for epithelial-mesenchymal transformation and is a marker of undifferentiated EPDCs. Viral labeling in non-infarcted and infarcted hearts, without exogenous EPDC transplantation, revealed the capacity of endogenous epicardium to generate new EPDCs after MI.
Conclusions
Human adult EPDCs preserved cardiac function after MI mainly through early paracrine signaling. The earlier WT1 expression in EPDC-recipients, in combination with the knowledge that new EPDCs can be formed by adult host epicardium, strongly suggests that the profit of EPDC transplantation is based on stimulation of endogenous epicardium.
Introduction
Optimizing therapy of myocardial infarction (MI) with stem or progenitor cells is one of the major focuses of current cardiology. The aim of this novel experimental treatment is to repair or rescue the tissue that is often irreversibly damaged after coronary artery occlusion. Basic and clinical research largely concentrates on application of adult progenitor cells which are most often derived from the bone marrow, with various results 1-3. In our opinion the heart itself is, however, the most appropriate source 4, as the natural niche of these primitive ‘remnants’ of early development is located in the organ of interest 5-7. Moreover, they will already be committed to the cardiac lineage. Epicardium- derived cells (EPDCs) might be of special interest because they have an extensive and important role in cardiogenesis, with their absence being lethal 8,9. Embryonic EPDCs which are generated from the epicardium by epithelial-mesenchymal transformation (EMT), populate the developing myocardium as interstitial and adventitial fibroblasts, and as smooth muscle cells 10. They also regulate several processes like the formation of the compact myocardium 11. We demonstrated recently that EPDCs are present in the adult heart and can still be generated from the epicardium in vitro 12. After transplantation into the infarcted heart, adult EPDCs improved left ventricular (LV) function and attenuated adverse remodeling at 2 and 6 weeks after onset of the MI 13. The underlying mechanism was not elucidated, but an early protection and stimulation of the surrounding host tissue by the engrafted cells was suggested 13, comparable to the influence of the embryonic EPDCs on their cardiac environment 11.
To search for an explanation of the benefit of EPDC transplantation, we studied cardiac function and histology at day 2, 4 and 7 after MI. Functional data were assessed by magnetic resonance imaging (MRI) to reveal whether EPDCs exert a gain or a preservation of LV function. Histological evaluation at the described time points would demonstrate whether apoptosis, proliferation, inflammatory- cell influx, vascularization, and properties of epicardial cells and cardiomyocytes were altered as a result of EPDC engraftment. We also studied the murine host tissue for its expression of Wilms’
Tumor 1 (WT1), a marker of undifferentiated EPDCs, Platelet-derived growth factor (PDGF), necessary for embryonic EPDC formation and function 14-16, and Sonic Hedgehog (SHH). SHH is required for development 17,18 and maintenance of vasculature 19, processes which are highly regulated by embryonic EPDCs 18. Since transplantation of adult EPDCs transiently increased vascular density
13, while their natural fibroblast derivatives express SHH in the adult heart 19, we hypothesized that transplanted EPDCs might enhance SHH signaling. Moreover, we studied in vivo if EPDCs are generated from the host epicardium during myocardial healing.
To prevent rejection of transplanted allogeneic cells we applied non-obese diabetic severe combined immunodeficient (NOD/scid) mice. These mice, which are widely used in cellular transplantation studies, lack T and B cells and have alterations in innate immunity 20,21. Since the immune system is highly involved in scar formation, it remains to be investigated whether this type of mouse strain indeed demonstrates ‘normal’ cardiac healing as assumed. Since we had to compare EPDC- transplanted animals to baseline, this study provides on protein level several aspects of ‘standard’
infarct tissue development over time for the NOD/scid mouse, which can be used by others as a reference.
The goals of this study were to i) study the pattern of functional improvement by exogenous EPDCs over time early after MI, to ii) examine if and how transplanted EPDCs alter natural cardiac repair on a protein level, and to iii) investigate if endogenous EPDCs are generated by the host epicardium of the infarcted area. We hypothesized that adult EPDCs would preserve cardiac function by beneficial paracrine signaling. It was supposed that this would result from augmented cellular survival or proliferation, increased vascular density, enhanced PDGF and SHH signaling, and/or altered activity and WT1 expression of endogenous epicardium.
Materials and Methods
Culture of human adult epicardium-derived cells
Epicardial tissue was isolated from human adult atrial appendages, which were acquired as waste material during coronary artery bypass graft procedures, conformed to institutional ethical guidelines. The epicardial tissue was minced in small pieces and cultured in medium consisting of 45% DMEM, 45% M199, 10% fetal calf serum (FCS), 100 U/ml penicillin, 100 μg/ml streptomycin (all from Invitrogen, Paisley, UK), and 2 ng/mL basic fibroblast growth factor (bFGF, BD Biosciences), like described before 13. Cells from passage 2-4 were transduced with the adenoviral vector hAd5/F50.
CMV.eGFP (50 infectious units per cell) 13,22. Enhanced green fluorescent protein (eGFP) expression was confirmed before EPDCs were used for transplantation.
Surgical procedure
All animal procedures were approved by the Animal Ethics Committee of the Leiden University and conformed to the Guide for the Care and Use of Laboratory Animals (National Institutes of Health publication No. 85-23, Revised 1996). A myocardial infarction was created in male non- obese diabetic severe combined immunodeficient (NOD/scid) mice (NOD.CB17-Prkdcscid/J, 001303, Charles River, Wilmington, MA, USA) of 12 weeks old, like described before 13,23. Shortly, animals were anesthetized with isoflurane 2% in 0.6 L/min oxygen, placed on a controlled heating pad and mechanically ventilated with a Harvard ventilator after non-invasive intubation. Tidal volume was 225 μl, respiratory frequency was 190/min, and a positive end expiratory pressure of 2 ml H2O was maintained. After opening of the thoracic cavity and the pericardial sac, the left anterior descending coronary artery (LAD) was visualized. Just below the bifurcation, the anterior branch was ligated with a 7.0 suture (Prolene, Johnson and Johnson, New Brunswick, NJ, USA). Successful occlusion of the LAD was confirmed by fading of the myocardium into yellow. Animals were randomized for transplantation of 20 μl culture medium (M199) with or without 4x105 EPDCs suspended (EPDC group and Medium or control group, respectively). The total volume was delivered through five small injections into the infarcted area and borderzones of the left ventricle, immediately after ligation of the LAD. Sham-operated animals underwent the same surgical procedure, but without ligation of the LAD and without injections. After recovery, animals received food and water supplied with antibiotics (Ciproxin and Polymixin B, 10mg/ml) and an antimycotic (Fungizone 10 mg/ml) ad libitum.
Because animals had to be sacrificed at day 2, 4 and 7 to be applied for immunohistochemical (IHC) analysis, while animals had to be alive up to day 7 for functional assessment, different animals were operated for these purposes. For technical reasons, only the animals determined for histological purposes received eGFP-positive EPDCs (as opposed to non-eGFP transduced EPDCs in case of functional analysis destiny).
Functional assessment by magnetic resonance imaging
Left ventricular function (n=13 for EPDC group, n=16 for Medium group) was measured non-invasively by magnetic resonance imaging (MRI; 9.4T) 13,23, using a newly developed retrospective triggering technique 24. With this method a carefully placed navigator slice retrospectively generates cardiac triggering time points and respiratory gating windows, which information is used to reconstruct cine MRI images from the image data afterwards. Standard retrospective gating (Bruker IntraGate) uses either a slice refocusing signal or a saturation slice refocusing signal as navigator scan, the latter making multi slice acquisition possible. Single slice acquisition could not be applied because scanning of each of the multiple single slices is started randomly regarding the phase of the cardiac cycle. To compute cardiac volumes from the different slices, the processing software requires the first movie frame to represent the same phase of the cycle. Therefore, IntraAngio was used, which is a tool to obtain a multi slice dataset via a slice-by-slice acquisition. IntraAngio employs one navigator slice at a fixed position resulting in identical movie frame starting points for each slice.
MRI was performed on a wide bore magnet (9.4T AVANCE II console, Bruker Biospin, Rheinstetten, Germany) equipped with an actively shielded gradient set of 1T/m and a 30-mm birdcage resonator.
Bruker ParaVision 4.0 software was used for data acquisition and image reconstruction. Mice were anesthetized with isoflurane (mixture of oxygen 0.3L/min and air 0.3L/min with 5% isoflurane for induction, and 1.5% for maintenance) and placed head-up in the animal holder. Respiratory rate was monitored via an air-pressure detection cushion connected to a Bruker BioTrig console. By subtly regulating isoflurane proportions, respiratory rate was stabilized at 50-60 times per minute. Body temperature was maintained at 35 °C. Scout images of a four-chamber view of the heart were used to position a set of contiguous 1 mm short-axis orientated images, together covering the entire long axis of the heart (8 or 9 slices). For the cine- modified fast gradient echo (FLASH) sequence we used the following parameters: RF excitation pulse 1 ms hermite, flip angle 15°, repetition time 8.5 ms, echo time 1.86 ms, spectral band width 75 kHz, echo position 30%, field of view 25.6 x 25.6 mm, matrix 192 x 192, in-plane resolution 133 μm, and number of repetitions 200. The navigator slice (2 mm thickness and flip angle 10°) was placed caudally to the apex oblique through the descending aorta, parallel to the image slices. Total scan time was 4 min 17 sec per slice. MRI images were analyzed with the Mass for Mice software package (Medis, Leiden, the Netherlands) 25. Manual delineation of endocardial borders provided end-diastolic and end-systolic volumes, after which ejection fraction was computed automatically.
Immunohistochemical analysis
Mice predetermined for immunohistochemical analysis were sacrificed at day 2, 4 and 7 after surgery, by excision of the heart under general anesthesia through intraperitoneal injection of Domitor (0.5 mg/kg, Pfizer, New York, NY, USA), Dormicum (5 mg/kg, Roche, Bazel, Switzerland), and Fentanyl (0.05mg/kg, Janssen-Cilag, Beerse, Belgium). Hearts were immersed in 4% paraformaldehyde in phosphate buffered saline (0.1 M, pH 7.4) at 4°C for 24 h, embedded in paraffin, and sectioned at 5 μm (n=5 per group at day 2, and n=3 per group at day 4 and 7). Before staining, paraffin sections were deparaffinated.
Besides standard histological hematoxilin and eosin (HE) and Masson’s Trichrome staining (MT), immune stainings were performed like described before 13 (see Table 1 for antibodies that were applied and for details about the procedures). If necessary, antigen retrieval was achieved by treatment with 0.1 mg/ml Pronase E (Merck, Darmstadt, Germany) for 6 min (‘P’ in Table 1), or heating during 12 min at 98°C in citric acid buffer (0.01 Mol/L, pH 6,0) (‘H’ in Table 1). Primary antibodies were incubated overnight, followed by 1 h incubation with secondary antibody, and 45 min with third antibody if required. For 3,3’-diaminobenzidine tetrahydrochloride (DAB, Sigma-Aldrich, St Louis, MO, USA) staining Mayer’s hematoxilin was used to mark the nuclei. When fluorescent antibodies were used, 4’6-diamidine-2-phenylidole-dihydrochloride (DAPI) was applied as nuclear counterstaining, and protein expression was analyzed with a DM-IRBE fluorescent microscope (Leica, Rijswijk, the Netherlands) in combination with a confocal spinning disc (BD-CARV II, Imsol, Preston, UK).
Viral labeling of the epicardium
To study the possible cellular contribution of the epicardium to the infarct region, we virally infected host epicardium with a novel far-red fluorescent protein, Katushka 26,27. This is characterized by deep tissue penetration 26 and it is 7- to 10-fold brighter than other far-red fluorescent proteins e.g.
HcRed and mPlum 27. During surgery like described above, a volume of 20μl suspension 28 containing 2x106 transducing units of a lentivirus with the construct P635-turbo under the ubiquitine promotor (Katushka LV.ubic.F635+) 26,27,29 was injected into the pericardial cavity of mice with and without LAD ligation (n=6 and n=4 per group, respectively). In contrast to the other experiments, no cells or culture medium were injected into the myocardium. Because reporter fluorescence was optimal
Marker Category number (Company) Dilution AR EF Visualization
cTnI AB1627 (Chemicon, Temecula, CA, USA) 1:50 P No Fluorescent
Nkx2.5 (N-19) sc-8697 (Santa Cruz, Santa Cruz, CA, USA) 1:6000 H Yes DAB
sarcomeric myosin (MF-20) MF 20 (Hybridoma Bank, Iowa City, IA, USA) 1:2 -- Yes Fluorescent SERCA2a (2A7-A1) MA3-919 (Affinity Bioreagents, Golden, CO, USA) 1:100 P No Fluorescent
α/γ muscle actin (HHF35) M0635 (DAKO, Glostrup, Denmark) 1:100 -- No Fluorescent
α-SMA (1A4) A2547 (Sigma-Aldrich, St Louis, MO, USA) 1:3000 * -- No both
DDR2 (N-20) sc-755 (Santa Cruz, Santa Cruz, CA, USA) 1:50 ** P Yes both
Periostin Kind gift from R. Markwald (Charleston, SC, USA) 1:4000 H Yes DAB
CD45 (30-F11) 553076 (PharMingen, San Diego, CA, USA) 1:200 -- Yes DAB
Active Caspase-3 551150 (PharMingen, San Diego, CA, USA) 1:50 H Yes DAB
TUNEL 1684817 (Roche, Basel, Switzerland) 1:14 -- -- DAB
Ki-67 ab-833 (Abcam, Cambridge, UK) 1:300 H Yes DAB
PCNA (PC10) M 0879 (DAKO, Glostruk, Denmark) 1:1000 -- Yes DAB
CD31 (MEC 13.3) 550274 (PharMingen, San Diego, CA, USA) 1:50 P Yes DAB
CD31 (JC70A) M0823 (DAKO, Glostrup, Denmark) 1:25 P Yes Fluorescent
CD34 (MEC 14.7) mon1159 (Monosan, Uden, the Netherlands) 1:50 H Yes DAB
PDGF-AA Sc-128 (Santa Cruz, Santa Cruz, CA, USA) 1:150 -- Yes DAB
PDGFR-α P2110 (Sigma-Aldrich, St Louis, MO, USA) 1:200 H Yes DAB
PDGF-BB PA1-21093 (Affinity Bioreagents, Golden, CO, USA) 1:400 H Yes DAB
PDGFR-β 3162 (Cell Signaling, Danvers, MA, USA 1:50 H Yes DAB
SHH (N-19) sc-1194 (Santa Cruz, Santa Cruz, CA, USA) 1:250 H Yes DAB
WT1 (C-19) sc-192 (Santa Cruz, Santa Cruz, CA, USA) 1:2000 H Yes DAB
Secondary antibodies
Goat anti rat-biotin 559286 (PharMingen, San Diego, CA, USA) 1:200 -- -- -- Goat anti rabbit-biotin BA-1000 (Vector Labs, Burlingame, CA, USA) 1:200 -- -- -- Rabbit anti rat-biotin BA-4001 (Vector Labs, Burlingame, CA, USA) 1:200 -- -- -- Horse anti goat-biotin BA-9500 (Vector Labs, Burlingame, CA, USA) 1:200 -- -- -- Horse anti mouse-biotin BA-2000 (Vector Labs, Burlingame, CA, USA) 1:200 -- -- --
Rabbit anti mouse-HRP P0260 (DAKO, Glostrup, Denmark) 1:200 -- -- --
Goat anti mouse-Cy3 115-165-003 (Jackson ImmunoResearch, Suffolk, UK) 1:250 -- -- --
goat anti biotin-Cy3 ab6973 (Abcam, Cambridge, UK) 1:50 -- -- --
Goat anti rabbit-Cy3 111-165-144 (Jackson ImmunoResearch, Suffolk, UK) 1:250 -- -- --
Table 1. Characteristics of immunohistochemical stainings. For antigen retrieval (AR) slices were incubated with ponase (P) or heated (H) in citric acid buffer. For DAB staining the signal was enforced (EF) with the ABC-kit (PK-6100, Vector Labs, Burlingame, CA, USA), or for CD31 (MEC13.3) with the CSA-kit (K1500, DAKO, Glostrup, Denmark). Enforcement for fluorescent stainings was performed by application of an appropriate biotinylated secondary antibody and a Cyanine-3 (Cy-3) labeled third antibody.
Primary antibodies that were visualized with a Fluorescent label reacted with mouse and human tissue, besides for CD31 (JC70A human specific, MEC13.3 mouse specific) and were therefore reliable for evaluation of endogenous host tissue and exogenous engrafted human EPDC properties. cTnI: cardiac troponin I, SERCA2a: sarcoplasmic or endoplasmic reticulum Ca2+ ATP-ase, α-SMA: alpha smooth muscle actin, DDR2: discoidin domain receptor 2, TUNEL: terminal deoxynucleotidyl transferase biotin dUTP nick-end labeling, PCNA: proliferating cell nuclear antigen, PDGF: platelet derived growth factor, PDGFR: platelet derived growth factor receptor, SHH: Sonic hedgehog , WT1: Wilms’ Tumor 1, HRP: horse radish peroxidase. *: α-SMA dilution 1:3000 for fluorescent staining, 1:2000 for DAB staining; **: DDR2 dilution 1:50 for fluorescent staining, 1:250 for DAB staining.
after 72 hours in our hands, we assessed in anesthetized living animals at day 4 after surgery whether the virus was present in living cardiac tissue. The Optix-MX2 imager (ART Advanced Research Technologies Inc, Montreal, Canada) was used for visualization of in vivo fluorescence 30-33. This is a real-time in vivo imaging system based on time domain technology. It uses pulsing lasers to produce different excitation wavelengths (to enable excitation of the fluorescence entity) and a combined photomultiplier tube and single photon counter detection system to capture any fluorescent emission. Excitation was performed with a 635 nm pulsing laser and emission was detected with a 665 nm ( 20 band pass filter. Mice were anesthetized with 5% isoflurane for induction and 1.5-2%
isoflurane for maintenance in oxygen with a flow of 0.6 L/min, and placed in a supine position on an animal bed with integrated nose mask. The fluorescent intensity of the tissues was quantified using the Optiview software (version 2.2) provided as part of the Optix MX2 imaging package. Life time analysis was used to confirm Katushka origin 27. Intensities and life time were expressed in pseudo colors and projected on the bright field grayscale image of the mouse. The animals were sacrificed immediately afterwards. The extracted hearts were fixed during 2 hours in 0.25% glutaraldehyde and 2% paraformaldehyde in phosphate buffered saline, embedded in Tissue Tek (OCT compound, Sakura Finetek, Zoeterwoude, the Netherlands), and sectioned at 8 μm. Katushka-labeled epicardium as well as newly generated subepicardial and interstitial EPDCs could then be analyzed in sections by fluorescence microscopy (Leica microscope in combination with confocal spinning disc as described above).
Statistical analysis
For functional data, differences between groups were analyzed with independent sample t-tests and a Bonferroni correction for multiple testing. Differences within groups over time were measured with paired sample t-tests and a Bonferroni correction for multiple testing. A p-value of <0.016 was considered significant.
Figure 1. Left ventricular function is preserved in EPDC-recipients. Figures a-d demonstrate differences over time between groups.
Figures d-i show functional changes over time within each group. Ejection fraction (EF) remains at a higher level in the EPDC group than in the Medium group, illustrated by a significant difference at day 4 and 7 in favor of the EPDC group (a). End-systolic volume (ESV) (b) and end-diastolic volume (EDV) (c) are significantly smaller in the EPDC group than in the Medium group, already at 2 days after surgery. EF decreases significantly from day 2 to day 4 and 7 in the Medium group (e), while deterioration over time only marginally reaches the level of a significant difference between day 2 and 7 in the EPDC group (g). ESV is continuously increased over time in the Medium group (e). In the EPDC group ESV is enlarged over time, but only when day 2 is compared to day 7 (h). EDV augments in size in the Medium group (f) as well as in the EPDC group (i). *: significant difference (p<0.016), #:
borderzone significant difference
Results
To study the effect of EPDC transplantation on cardiac function and host tissue properties, we assessed LV performance and characteristics by MRI and (immuno)histology, respectively. Eventually, we determined natural epicardial activity in case of MI using viral epicardial labeling.
Preservation of left ventricular function by EPDC transplantation
Ejection fraction (EF) was significantly higher in the EPDC group at day 4 and day 7 after MI than in the Medium group (Figure 1a) (EF of EPDC vs Medium: 49±1 vs 46±2 [day 2], 48±2 vs 36±2 [day 4, p<0.016], 44±3 vs 34±2 [day 7, p<0.016]). The EPDC group demonstrated a significantly smaller left ventricular end-systolic volume (ESV) and end-diastolic volume (EDV) than the Medium group at all three time points measured, being day 2, 4 and 7 after MI (Figure 1b, c) (ESV of EPDC vs Medium: 23±1 μl vs 30± 3 μl [day 2, p<0.016], 26±2 μl vs 39± 3 μl [day 4, p<0.016], 31±3 μl vs 48±5 μl [day 7, p<0.016]);
EDV of EPDC vs Medium: 45±2 μl vs 55±3 μl [day 2, p<0.016], 49±2 μl vs 60±3 μl [day 4, p<0.016], 54±2 μl vs 71±5 μl [day 7, p<0.016]). When changes of functional parameters over time were compared within each group, it was observed that EF deteriorated strongly over time in the Medium group (Figure 1d), but remained almost stable in the EPDC group (Figure 1g). ESV and EDV increased steadily over time in both groups (Figure 1e, f), but this ongoing enlargement, indicating pathological myocardial remodeling, was less apparent in the EPDC group (Figure 1h, i).
Figure 2. Increasing expression of smooth muscle and fibroblast markers in engrafted EPDCs over time. At day 2 (a-d) staining for α-smooth muscle actin (α-SMA, red, c) is negative in the enhanced green fluorescent protein (eGFP)-positive EPDCs (green, b). At day 4 (e-h), part of the engrafted cells (f) expresses the smooth muscle cells marker α-SMA (g) while another part is still negative. At day 7 (i-l) colocalization for eGFP and α-SMA is observed in almost all cells (l). Fibroblast marker DDR2 (red) is not expressed at day 2 (m-p). Staining for discoidin domain receptor 2 (DDR2) is positive in many of the EPDCs at day 4 (q-t), which is even more prominent at day 7 (u-x). Please note subtle expression of α-SMA and DDR2 in the infarcted area at day 4 and stronger expression in the entire region at day 7. Arrows indicate positive cells. Scale bars represent 50 μm. See Table 1 for details about the antibodies.
Engrafted EPDCs acquire a myofibroblast phenotype
The transplanted human adult EPDCs could be detected in each heart determined in the infarcted area of the left ventricle or in the borderzone. Confocal microscopic evaluation revealed that at each time point measured, the engrafted eGFP-positive EPDCs were negative for the cardiomyocyte markers sarcomeric myosin (MF20), cardiac troponin I (cTnI), sarcoplasmic or endoplasmic reticulum Ca2+
ATP-ase (SERCA2a), and α/γ muscle actin (HHF35) (not shown). Expression of the endothelial marker CD31 could not be detected in transplanted EPDCs in any of the animals at any time point either (not shown). The smooth muscle cell marker α-smooth muscle actin (α-SMA) was increasingly expressed by the engrafted EPDCs over time. By day 2 hardly any α-SMA positive EPDCs could be observed, at day 4 a large percentage of the engrafted cells expressed this marker, and at day 7 almost all human EPDCs were positive for α-SMA (Figure 2a-l). Similarly, antibody staining for discoidin domain receptor-2 (DDR2), a protein which is mainly expressed in fibroblasts, was not detected at day 2, but it was present in part of the engrafted cells at day 4, and in almost all EPDCs at day 7 (Figure 2m-x).
Histological properties of the infarcted heart
To describe how the infarct area of control- and EPDC-treated hearts developed, we used a panel of markers to delineate cellular infiltration, collagen deposition, cardiomyocyte conditions, fibroblasts, cellular turnover, vascularization, PDGF and SHH signaling, and epicardial behavior (see Table 1).
Figure 3. General characteristics of cardiac healing. Pictures demonstrate early repair of the infarcted left ventricle and are representative for Medium- and EPDC-transplanted hearts of immunodeficient NOD/scid mice. It is clearly visible that the infarcted myocardial wall is thinner (b-d) than the normal myocardial wall (a), and that cardiomyocytes within the injured area lose their nucleus (best visible at high magnification in j-l). The region with these necrotic cardiomyocytes having lost the nuclei (marked by dotted line) is engulfed over time by a cellular infiltrate (asterisks), and has almost disappeared by day 7 (d, h). Boxes in pictures e-h (100x magnification) represent areas that are shown at higher magnification in i-l (400x) and in consecutive sections in m-p (400x). The zone with the cellular infiltrate colocalizes with collagen deposition (light blue) at day 4 (k, o, respectively) and 7 (l, p, respectively). MI: myocardial infarction, HE: hematoxilin and eosin, MT: Masson’s Trichrome, Ep:
epicardium, End: endocardium, LV: left ventricle, RV: right ventricle. Scale bars in pictures a-d represent 600 μm, in pictures e-h
Cellular infiltrate and collagen deposition
Generally, ligation of the anterior branch of the LAD resulted in thinning of the injured left ventricular wall which increased over time (Figure 3a-d). At day 2 the infarcted cardiac muscle showed necrosis with loss of nuclei and cross striation. Myofibers were hypereosinophilic and myocytolysis was also present (Figure 3). The zone of necrotic cardiomyocytes with loss of nuclei was most prominent at day 4, and had almost disappeared at day 7 (Figure 3). By that time it was engulfed almost entirely by newly emerged small cells, phenotypically mainly recognized as macrophages and fibroblasts (Figure 3d, h, j). At day 2 this infiltrate was minimal and located in between cardiomyocytes at the epicardial and, to a lesser extent, the endocardial side of the ventricular wall (Figure 3b, f, j). At day 4 the number of infiltrating cells had dramatically increased (Figure 3c, g, k). They were positioned around the necrotic area embedded in a loose extracellular matrix (ECM). At day 7 they formed the majority of cells in the necrotic area (Figure 3d, h, j). MT staining revealed that collagen production spatiotemporally coincided with the presence of the infiltrate in the infarcted area (Figure 3m-p).
In each heart analyzed, zones of phenotypically healthy cardiomyocytes were observed in the close vicinity to the endocardial and epicardial surface (Figure 3d, h, j, p). This pattern was not different for Medium- and EPDC-transplanted hearts. In sham-operated hearts a similar but smaller infiltrate appeared at the epicardial side of the heart, at the location where the epicardium had been touched by the surgical instruments (not shown).
Cardiomyocytes and fibroblasts
In both groups the necrotic cardiomyocytes demonstrated a rapid decrease in expression of cardiomyocyte markers for SERCA2a (not shown), MF-20 (not shown), and cTnI (Figure 4a-d). In the infarcted area of Medium- and EPDC-transplanted hearts the smooth muscle cell marker α-SMA was initially (day 2) expressed almost only in the vascular wall of large vessels (Figure 4f), like observed in healthy cardiac tissue (Figure 4e). At day 4 and 7 this protein was observed in a large proportion of the infiltrate that had invaded the infarcted wall (Figure 4g, h). Similarly, expression of fibroblast specific DDR2 34,35 increased over time in the infarcted area of both groups (Figure 4j-l), as well as in the non- infarcted part of the LV (not shown). Periostin, which is known to be expressed by cardiac fibroblasts after MI 36, could be detected in the infarcted area already at 2 days after MI (Figure 4n-p), regardless of the fact whether EPDCs were transplanted. Nkx2.5, normally present in the nucleus of cardiomyocytes (Figure 4q), was transiently present in the cytoplasm of the necrotic cardiomyocytes in each heart determined (Figure 4r). This transcription factor was also strongly expressed in the cytoplasm of the small elongated shaped cells that were situated in the infarcted area around the necrotic cardiomyocytes (Figure 4s, t). At day 2 and 4 approximately two third of the small round cells of the infiltrate expressed the hematopoietic marker CD45 (Figure 4v, w), and at day 7, at which moment the number of the invading cells had dramatically increased, only one third of the population was positive for CD45 (Figure 4x). Again, no differences between EPDC- and Medium-injected hearts were observed.
Cellular turnover: apoptosis and proliferation
The number and distribution over time of apoptotic and mitotic figures was comparable between groups. Apoptotic cells could be detected at each time point determined in the infarcted area and border zone, as revealed by antibody staining against cleaved Caspase-3 (Figure 5b-d) and terminal deoxynucleotidyl transferase biotin dUTP nick-end labeling (TUNEL) assay (not shown). Most positive cells were observed within the infiltrate and not amongst the cardiomyocytes (Figure 5b-d). Ki67 staining showed a high proliferation rate in the infiltrate area, and absence of mitosis in the non- affected tissue. Proliferating cell nuclear antigen (PCNA) positivity, as a marker for proliferation and DNA-damage repair, could be observed in the entire healthy heart (Figure 5i), mainly in fibroblasts and endothelial cells, but also in some cardiomyocytes (not shown). The density of PCNA positive cells was increased in the border zone and infarcted area (Figure 5j-l). This was not altered in case of EPDC transplantation.
Vascularization
In the 2 day-old myocardial scar of EPDC- and Medium-transplanted hearts, the normal pattern of many small capillaries bordering all cardiomyocytes (Figure 6a, e) had changed into one where less expression of CD31 was observed (Figure 6b, f). At day 4 and 7 after LAD ligation (Figure 6c, d, g, h), vascular CD31 positive structures were absent in the central area of the injury in both groups. The influx that circumvented this necrotic core, however, contained many small CD31 and CD34 (not shown) positive vessels, sometimes with erythrocytes present in the lumen (not shown). EPDC transplantation had not changed the appearance of the vasculature.
PDGF and SHH signaling
PDGF-A was not present in the normal heart of immunodeficient NOD/scid mice, but expression of the ligand was induced in some of the infiltrating cells in both treatment groups (Figure 7a- d). Its receptor PDGFR-α was strongly upregulated in the infarcted area of the control- and EPDC transplanted hearts, and expressed by almost all infiltrating cells (Figure 7e-h). The vasculature of the non-injured hearts produced PDGF-B, while its receptor PDGFR-β was only expressed by interstitial cells (Figure 7i, m, respectively). In case of myocardial infarction, with or without EPDCs delivered to the injured area, both the ligand and its receptor were prominently present in the cells of the infiltrate including the vessels in the injured area (Figure 7j-l, n-p). In the healthy heart SHH protein could be observed in the vessel wall. It was absent in the epicardium. (Figure 7q). After MI it was expressed in the infiltrated zone including the epicardium (Figure 7r-t). This was not altered by EPDC transplantation.
Epicardial marker WT1
After MI, WT1 protein could not be detected in the infarcted area until day 4 (Figure 8b), from which moment it was clearly present in the epicardium and in cells scattered through the LV wall of the injured area (Figure 8c, d). In contrast, when EPDCs were transplanted, WT1-expressing cells could be observed in the entire harmed zone as well as in the outer epicardial layer already at 2 days after MI (Figure 8e). The protein remained present in the infarcted area in both groups until the last time point determined, day 7 after MI (Figure 8d, g). In both groups, the WT1-positive epicardial cells covering the infarcted area were morphologically clearly different from those in the healthy heart. Whereas normal epicardium consists of squamous cells, the injured area was covered by cubic-shaped epicardial cells (Figure 8).
Natural epicardial activity after MI
To investigate in vivo if the host epicardial layer of adults is activated in case of MI, we virally labeled the epicardium of hearts with and without MI. The strong Katushka signal in the cardiac region of the mice without and with MI (scanned in living animal, Figure 9a, b respectively) demonstrated that indeed living cardiac cells were infected with the virus. Histological determination revealed that in non-injured hearts, only the epicardium was labeled, which consisted of one layer of flattened cells (Figure 9d-f). In contrast, after MI, the Katushka positive epicardial layer covering the infarcted zone contained cubic cells (Figure 9g-i). Moreover, various Katushka-positive interstitial cells were detected in the inward infarcted zone (Figure 9j-o).
Figure 4. Representative photographs demonstrating myocardial properties of normal hearts without myocardial infarction (MI), and of EPDC- and Medium-recipients at day 2, 4, and 7 after MI. Pictures show the center of the myocardial wall (for day 4 and 7 the interface of the infiltrate and the necrotic cardiomyocytes), besides for figure a-d where the border of the infarcted area and healthy myocardium is shown. Expression of cardiomyocyte marker cardiac Troponin I (cTnI) disappeared immediately in the ischemic area (a-d). Smooth muscle cell marker α-smooth muscle actin (α-SMA), normally present in arterioles, arteries and large venes (e), and discoidin domain receptor 2 (DDR2) and Periostin, normally expressed by interstitial fibroblasts (i, m), are present in the elongated infiltrating cells in the infarcted area of Medium- and EPDC-transplanted hearts, increasing in number over time (f-h for α-SMA, j-l for DDR2, n-p for Periostin). Nkx2.5 is a transcription factor present in the nucleus of healthy cardiomyocytes (q), but it is transiently present in the cytoplasm of necrotic cardiomyocytes (r). It can also be observed in the cytoplasm of many mainly elongated shaped cells of the infiltrate (s,t). CD45 is only detected in the injured hearts, mainly present in small round cells within the infiltrate (u-x). Scale bars represent 50 (a-d), or 45 μm (e-x). Arrows indicate cellular presence of the antigen. See
Figure 5. Apoptosis and proliferation during early myocardial healing (similar for Medium- and EPDC-injected mice). While in the healthy heart almost no apoptotic and proliferating cells can be detected (a and e, respectively), the infarcted area contains many small cells that are positive for Caspase-3 in the first week after myocardial infarction (MI) (b-d) and ki67 (f-h), indicating increased apoptosis and proliferation, respectively. In contrast, proliferating cell nuclear antigen (PCNA) is abundantly expressed in the healthy left ventricular wall (i), as well as in the infiltrate of the infarcted area, probably mainly marking DNA-damage repair (j-l). Arrows indicate cells that are positive for the respective antigen. Scale bars represent 45 μm. See Table 1 for details about the antibodies.
Figure 6. The normal pattern of many small capillaries characteristic of healthy myocardial tissue (a, e) is lost by day 2 after myocardial infarction (MI) (b, f). However, many small CD31 positive vessels (arrows) are detected in the infiltrate at day 4 and 7. Pictures are representative for Medium- and EPDC-recipients. Pictures e-h represent magnifications of boxes in a-d. End:
endocardium, Ep: epicardium. Scale bars represent 120 μm (a-d) or 45 μm (e-h). See Table 1 for details about the antibodies.
Discussion
Functional MRI assessment 24 confirmed our hypothesis 13 that transplantation of EPDCs preserved LV function early after MI. EF of the EPDC group did not deteriorate as fast over time as EF of the Medium group, but values did not surpass baseline. This indicated that EPDCs did not exert a gain of function after an initial loss. Likewise, EDV and ESV increased in the EPDC group during the first 7 days after MI, but to a lesser extent than as observed for the Medium group. Volumes of the EPDC-injected hearts were already smaller at day 2, by which time new tissue could not have been regenerated. This implied, together with the pattern of attenuated deterioration as opposed to functional recovery, that the beneficial effect of EPDCs must be explained by paracrine stimulation, protection, or improved function of the remaining endogenous host tissue 13. Moreover, viable human myocardium or vasculature was not generated by the engrafted EPDCs, which is in line with the current ideas on cardiomyocyte lineage. They acquired a myofibroblast phenotype with DDR2 and α-SMA expression 37. It could very well be that this early effect is critical for acquiring the long-term benefit, as has been demonstrated for week 2 and 6 13. Considering the Frank-Starling mechanism 38, a small but early benefit in LV volumes can result in large functional profit on the long term. To elucidate the immediate beneficial paracrine signaling of the engrafted EPDCs on the host tissue, we extensively studied the process of myocardial healing on a histological level. We thereby hypothesized that EPDCs might alter infiltrate formation, collagen deposition, cell turn-over, vascularization, PDGF and SHH signaling, and murine EPDC behavior. Of note, by our thorough investigation of the control group, we provided a detailed demonstration of normal cardiac healing for immunodeficient NOD/scid mice, which currently lacks in literature.
General signs of cardiac healing
General signs of cardiac healing were comparable between the EPDC and the control group, with an expanding infiltrate engulfing the necrotic cardiomyocytes 39, concurrent with collagen deposition 40, and small vessel formation 41. Both proliferation and invasion of systemically derived cells accounted for the occurrence of the dense cellular infiltrate, as was indicated by increased mitosis (Ki67) and expression of CD45, respectively. It consisted mainly of macrophages 42 and myofibroblasts 37,40, which were characterized by α-SMA, and by the fibroblast markers DDR2 34 and Periostin 35.
Interestingly, in both groups, a considerable number of elongated shaped cells located at the interface of the necrotic cardiomyocytes and the infiltrate demonstrated the cardioprotective 43 cardiomyocyte marker Nkx2.5 in their cytoplasm. This is peculiar, since as a transcription factor it is normally only present in the nucleus of cardiomyocytes. The origin of these cells is therefore uncertain. They might be dying cardiomyocytes with their nuclear Nkx2.5 positive content present in the cytoplasm. But they could also denote myofibroblasts expressing an epitope to which the Nkx2.5 antibody bound non-specifically. However, atypical binding was not observed in the non-injured area of the heart. We speculate on a third connotation by suggesting that these cells with Nkx2.5 protein in their cytoplasm but with the shape of fibroblasts represent ischemic cardiomyocytes that trans- or dedifferentiate into myofibroblasts during the healing process. This needs to be further elucidated, but it is supported by the fact that the origin of myofibroblasts in the infarcted area is still uncertain
35,37,44.
Cellular turn-over
Apoptosis, identified by TUNEL and cleaved caspase-3 positivity 45, could be detected within the infiltrate of both control- and EPDC-transplanted hearts. It probably reflected the decrease in cellularity needed for the transition of the granulation tissue into scar 41,46,47. On the other hand, PCNA activity, characteristic for DNA-damage repair and proliferation 48, was extremely high in the
infiltrate of both groups. Since less ki67 positive mitotic cells were detected, the majority of the PCNA positivity most likely represented DNA-damage repair. The fact that DNA-damage repair was high in both groups early after MI, and not only in the EPDC group as was the case after 14 days 13, suggested that EPDCs have a late paracrine effect which is different from the early effect. We suppose that the prolonged presence of DNA-damage repair observed at 14 days after transplantation 13 instigated maintenance of the vasculature in the granulation tissue, since vascular density was also increased in EPDC-recipients at a later stage after MI 13.
PDGF signaling
We hypothesized that transplantation of EPDCs might influence the PDGF expression pattern in the infarcted area, since both the ligands and receptors are important for proper formation and function of EPDCs 14-16,49, but no differences were observed between groups. We can, however, not exclude that possible modifications were too subtle to detect with immunohistochemical analysis, especially since expression was already extremely high in control hearts. The increased signaling is needed for myocardial healing 50,51. Both ligands PDGF-A and –B and their receptors PDGFR-α and -β are involved in collagen deposition 50,51. They are mitogenic 52, chemotactic 50,51, and stimulate growth factor production by macrophages 50. Moverover, PDGF-B and PDGFR-β signaling is important for the recruitment and development of vasculature in the infiltrate 53,54, which itself is critical for resolution of the inflammatory process 51. More detailed future research which evaluates active signaling, will demonstrate if EPDCs do not alter PDGF signaling in cardiac healing indeed.
SHH signaling
It is known from previous research that EPDCs could preserve or enhance vascular density, as was observed in a 14-day old MI 13. Since SHH signaling is required for maintenance 19 and growth stimulation 17,55 of adult vasculature, we speculated that SHH signaling would be increased in EPDC- recipients early after MI. This was of particular interest because embryonic epicardium regulates development of coronary vasculature through SHH signaling 17,18, while fibroblasts of the adult heart, which are derivatives of EPDCs 8,10, express SHH 19. However, we did not detect any differences between groups. Interestingly though, we observed that SHH was expressed by the host epicardium that covered the infarcted zone, which has not been reported before. It suggested an active role of epicardium in cardiac healing. We suppose that SHH signaling contributed to the late paracrine effect of engrafted EPDCs, since increased SHH expression in the adult heart results in large tortuous vessels
56, which was observed in EPDC-recipients 2 weeks after MI 13. Endogenous EPDC activation
We hypothesized that transplanted donor-EPDCs influence activation and migratory behavior of endogenous EPDCs. To study possible reactivation of endogenous EPDCs in the infarcted area we evaluated WT1 protein expression. WT1 is a nuclear transcription factor involved in development of several mesodermal organs 57, including the heart 58-60. It is expressed by undifferentiated embryonic EPDCs, being lost as soon as the cells have fully differentiated 59, and it is crucial for proper EPDC functioning 58,59. In the adult heart, presence of WT1 could therefore indicate dedifferentiated 61 or newly generated EPDCs. We demonstrated that the protein was indeed present in the infarcted area
61 as well as in the epicardium covering this injured zone. Interestingly, in EPDC-recipients the protein could be discerned already at day 2, while the control group did not exhibit expression before day 4. It could thus be concluded that donor-EPDCs, delivered into the myocardium, triggered a temporal shift forward of WT1 expression. The earlier WT1 expression in the EPDC group coincided with the timeline seen for functional preservation: the attenuation in adverse LV remodeling started at day 2. A role of WT1 in protection from ischemia is therefore strongly suggested.
Figure 7. After myocardial infarction (MI), expression of platelet derived growth factor (PDGF) and sonic hedgehog (SHH) is increased in Medium- and EPDC-transplanted hearts, without any differences between treatment groups. PDGF-A and PDGF receptor-α (PDGFR-α) are hardly detected in the normal heart (a, e), but PDGF-A and PDGFR-α are present within the infiltrate (b-d and f-h respectively). Normally, PDGF-B is observed in the large coronary vasculature, and its receptor PDGFR-β in interstitial cells (i and m, respectively). After MI, the ligand is still present in the vessels, but also in many cells of the infiltrate (j-l), where its receptor PDGFR-β is also abundantly expressed (n-p). The healthy heart shows SHH expression in its vasculature but not in the epicardium (q). However, after MI, SHH is also expressed in the cellular infiltrate and epicardial layer (r-t). Scale bars represent 45 μm. See Table 1 for details about the antibodies. Arrows indicate positive cells.
Figure 8. Photographs of Wilms’ Tumor 1 (WT1) expression in the center of the left ventricular wall (1) and the epicardium (2) in normal hearts and in hearts with myocardial infarction (MI) that received Medium (Med) or EPDC transplantation. The healthy heart, which hardly contains WT1-expressing cells in the ventricular wall (a1), demonstrates some squamous WT1-positive epicardial cells (a2). At day 2 after MI, WT1 is not observed in the more cubic epicardial cells of the Medium-recipients (b2), but at day 4 and 7 many WT1 positive cells are observed in the infiltrate including endothelial cells (c1, d1) as well as in the activated epicardial layer (c2, d2). Transplantation of adult EPDCs alters this process, as demonstrated in picture e1 and e2: already by day 2 many WT1 expressing cells are detected in the infiltrate (e1) as well as in the epicardium (e2), which is maintained until day 7 (f1, f2, g1, g2). Arrowheads indicate WT1 positive interstitial cells, arrows indicate cubic WT1 positive epicardial cells. Scale bars
Figure 9. EPDCs are generated by host tissue epicardium after MI. Bioluminescent scanning reveals that Katushka-expressing cells are present in the cardiac region 4 days after injection into the pericardial cavity of mice without (a) and with (b) myocardial infarction (MI). The fluorescent signal is absent in a non-treated control mouse (c). Histological evaluation of mice without (d-f) and with (g-o) MI at day 4 after labeling of the pericardial cavity demonstrated that the flattened outer epicardial cells of healthy myocardium express Katushka (d-f). Katushka positive epicardial cells that cover the infarcted area are cubic (g-i). Regions with moderate (j-l) and profound (m-o) ingrowth of Katushka labeled cells are detected within the infarcted zone. Arrows indicate
We can speculate on the origin of these WT1-positive cells. WT1 expression might have indicated renewed generation of WT1-positive EPDCs from the host epicardium by EMT, and also upregulation of WT1 by cells already present in the infarcted area 61. The cubic shape of the epicardial cells in the infarcted zone, indicative for EMT, was suggestive for the former. Moreover, advanced migration of extracardiac-derived WT1 positive cells into the infarcted area could not be excluded, with some hematopoietic progenitors expressing WT1 60,62. Wt1 is known to protect from apoptosis 57,63, instigate robust capillary formation 59,64, and stimulate proliferation 60,61,63,65. Since we did not detect any differences regarding these processes early after MI, we consider it less likely that WT1 was only upregulated by resident cells. Although we can not fully exclude that our methods were too coarse to detect a potential decrease in apoptosis or increase in vascular density and mitosis, accelerated invasion of bone-marrow derived cells or generation of new EPDCs are more plausible explanations.
Therefore, we investigated by lentiviral fluorescent (Katushka 27) labeling of the host epicardium whether new EPDCs were regenerated after MI. We first demonstrated by in vivo fluorescence determination with the Optix-MX2 imager, that living cells in the cardiac region, supposedly the pericardial cavity, were successfully infected. Histological evaluation revealed that in the normal hearts only the outer layer of flat epicardial cells was positive, as was expected since the tightly joined epicardial cells form a barrier that prevents transduction of the underlying layers 28. Strikingly, at day 4 after MI, areas of Katushka expressing cells were detected within the infarcted area, implicating that they were generated by the epicardium. Since WT1 is known to enable EMT 59,66, it is therefore strongly suggested that the earlier WT1 expression in the EPDC-recipients designates accelerated EPDC formation from the host tissue. Future research will further elucidate this interesting and conceptually new finding.
Conclusions
In conclusion, we demonstrated that transplantation of adult human EPDCs into the infarcted heart improved cardiac performance by preventing additional loss already at day 2 after MI, which was instigated through early paracrine signaling. It was strongly suggested that accelerated generation of new murine EPDCs from the host epicardium contributed to the functional profit, since i) it was demonstrated that host epicardium could form new EPDCs, and ii) WT1 expression, highly associated with EMT and EPDC formation, was only expressed in EPDC-recipients at day 2 (as opposed to day 4 in the control group).
Reference List
1.
Passier R, Van Laake LW, Mummery CL.
Stem-cell-based therapy and lessons from the heart. Nature. 2008; 453:322- 329.
2.
Segers VF, Lee RT. Stem-cell therapy for cardiac disease. Nature. 2008; 451:937- 942.
3.
Dimmeler S, Burchfield J, Zeiher AM.
Cell-based therapy of myocardial infarction. Arterioscler Thromb Vasc Biol. 2008; 28:208-216.
4.
Lyngbaek S, Schneider M, Hansen JL, Sheikh SP. Cardiac regeneration by resident stem and progenitor cells in the adult heart. Basic Res Cardiol. 2007.
5.
Garry DJ, Olson EN. A common progenitor at the heart of development.
Cell. 2006; 127:1101-1104.
6.
Bruneau BG, Black BL. The heart’s Da Vinci code: a Renaissance at Keystone.
Development. 2007; 134:1631-1633.
7.
Srivastava D, Ivey KN. Potential of stem- cell-based therapies for heart disease.
Nature. 2006; 441:1097-1099.
8.
Winter EM, Gittenberger-de Groot AC.
Cardiovascular development: towards biomedical applicability : Epicardium- derived cells in cardiogenesis and cardiac regeneration. Cell Mol Life Sci.
2007; 64:692-703.
9.
Lie-Venema H, Van Den Akker NM, Bax NA, Winter EM, Maas S, Kekarainen T, Hoeben RC, DeRuiter MC, Poelmann RE, Gittenberger-de Groot AC. Origin, fate, and function of epicardium- derived cells (EPDCs) in normal and abnormal cardiac development.
ScientificWorldJournal. 2007; 7:1777- 1798.
10.
Gittenberger-de Groot AC, Vrancken Peeters MP, Mentink MM, Gourdie RG, Poelmann RE. Epicardium-derived cells contribute a novel population to the myocardial wall and the atrioventricular cushions. Circ Res. 1998; 82:1043-1052.
Poelmann RE. Epicardial outgrowth inhibition leads to compensatory mesothelial outflow tract collar and abnormal cardiac septation and coronary formation. Circ Res. 2000;
87:969-971.
12.
van Tuyn J, Atsma DE, Winter EM, van der Velde-van Dijke I, Pijnappels DA, Bax NAM, Knaan-Shanzer S, Gittenberger- de Groot AC, Poelmann RE, van der Laarse A, van der Wall EE, Schalij MJ, de Vries AAF. Epicardial cells of human adults can undergo an epithelial-to- mesenchymal transition and obtain characteristics of smooth muscle cells in vitro. Stem Cells. 2007; 25:271-278.
13.
Winter EM, Grauss RW, Hogers B, van Tuyn J, van der Geest R, Lie-Venema H, Vicente Steijn RV, Maas S, DeRuiter MC, deVries AA, Steendijk P, Doevendans PA, van der Laarse A, Poelmann RE, Schalij MJ, Atsma DE, Gittenberger- de Groot AC. Preservation of left ventricular function and attenuation of remodeling after transplantation of human epicardium-derived cells into the infarcted mouse heart. Circulation.
2007; 116:917-927.
14.
Van Den Akker NM, Lie-Venema H, Maas S, Eralp I, DeRuiter MC, Poelmann RE, Gittenberger-de Groot AC.
Platelet-derived growth factors in the developing avian heart and maturating coronary vasculature. Dev Dyn. 2005;
233:1579-1588.
15.
Van Den Akker NM, Winkel LC, Nisancioglu MH, Maas S, Wisse LJ, Armulik A, Poelmann RE, Lie-Venema H, Betsholtz C, Gittenberger-de Groot AC. PDGF-B signaling is important for murine cardiac development: its role in developing atrioventricular valves, coronaries, and cardiac innervation. Dev Dyn. 2008; 237:494-503.
16.
Kang J, Gu Y, Li P, Johnson BL, Sucov HM, Thomas PS. PDGF-A as an epicardial mitogen during heart development. Dev Dyn. 2008; 237:692-701.
17.
Lavine KJ, White AC, Park C, Smith CS, Choi K, Long F, Hui CC, Ornitz DM. Fibroblast growth factor signals
18.
Lavine KJ, Ornitz DM. Fibroblast growth factors and Hedgehogs: at the heart of the epicardial signaling center. Trends Genet. 2008; 24:33-40.
19.
Lavine KJ, Kovacs A, Ornitz DM.
Hedgehog signaling is critical for maintenance of the adult coronary vasculature in mice. J Clin Invest. 2008;
118:2404-2414.
20.
Meyerrose TE, Herrbrich P, Hess DA, Nolta JA. Immune-deficient mouse models for analysis of human stem cells. Biotechniques. 2003; 35:1262-1272.
21.
Shultz LD, Schweitzer PA, Christianson SW, Gott B, Schweitzer IB, Tennent B, McKenna S, Mobraaten L, Rajan TV, Greiner DL, . Multiple defects in innate and adaptive immunologic function in NOD/LtSz-scid mice. J Immunol. 1995;
154:180-191.
22.
Knaan-Shanzer S, van der Velde I, Havenga MJ, Lemckert AA, de Vries AA, Valerio D. Highly efficient targeted transduction of undifferentiated human hematopoietic cells by adenoviral vectors displaying fiber knobs of subgroup B. Hum Gene Ther. 2001;
12:1989-2005.
23.
Winter EM, Grauss RW, Atsma DE, Hogers B, Poelmann RE, van der Geest RJ, Tschope C, Schalij MJ, Gittenberger- de Groot AC, Steendijk P. Left ventricular function in the post-infarct failing mouse heart by magnetic resonance imaging and conductance catheter:
a comparative analysis. Acta Physiol (Oxf). 2008; 194:111-122.
24.
Heijman E, de GW, Niessen P, Nauerth A, van EG, de GL, Nicolay K, Strijkers GJ. Comparison between prospective and retrospective triggering for mouse cardiac MRI. NMR Biomed. 2007;
20:439-447.
25.
van der Geest RJ, Buller VG, Jansen E, Lamb HJ, Baur LH, van der Wall EE, de Roos A, Reiber JH. Comparison between manual and semiautomated analysis of left ventricular volume parameters from short-axis MR images. J Comput Assist
