A feasibility study of a multimodal stimulation bioreactor for the conditioning of stem cell seeded cardiac patches via electrical impulses and pulsatile perfusion

Abstract

BACKGROUND/OBJECTIVE:

Ischemic heart disease is a major cause of mortality worldwide. Myocardial tissue engineering aims to create transplantable units of myocardium for the treatment of myocardial necrosis caused by ischemic heart disease – bioreactors are used to condition these bioartificial tissues before application.

METHODS:

Our group developed a multimodal bioreactor consisting of a linear drive motor for pulsatile flow generation (500 ml/min) and an external pacemaker for electrical stimulation (10 mA, 3 V at 60 Hz) using LinMot-Talk Software to synchronize these modes of stimulation. Polyurethane scaffolds were seeded with 0.750 × 10⁶ mesenchymal stem cells from umbilical cord tissue per cm² and stimulated in our system for 72 h, then evaluated.

RESULTS:

After conditioning histology showed that the patches consisted of a cell multilayer surviving stimulation without major damage by the multimodal stimulation, scanning electron microscopy showed a confluent cell layer with no cell-cell interspaces visible. No cell viability issues could be identified via Syto9-Propidium Iodide staining.

CONCLUSIONS:

This bioreactor allows mechanical stimulation via pulsatile flow and electrical stimulation through a pacemaker. Our stem cell-polyurethane constructs displayed survival after conditioning. This system shows feasibility in preliminary tests.

Keywords

Bioreactors tissue engineering heart disease myocardium stem cells

1. Introduction

The prevalence of cardiovascular disease is rising with the westernization of developing countries [1,2]. According to a World Health Organization projection, global mortality by ischemic heart disease will rise to 13.4% by 2030, staying in the number one position of causes of death [3]. Ischemia due to coronary heart disease is dealt with in a catheter lab or in the operating room by revascularization. However, revascularization can only save tissue which is still vital. The only treatment options that remain for patients who have suffered myocardial necrosis and who subsequently lose global myocardial function are ventricular assist devices and transplantation [4].

To provide a new form of therapy for ischemic heart disease which has resulted in tissue necrosis, tissue engineers attempt to create myocardial tissue which can be transplanted in order to assist the heart. Research groups have been successful in creating rat heart tissue for transplantation which was successful in supporting infarcted rat hearts [5]. However, the use of human tissue for this endeavor remains a hurdle. While neonatal rat cardiomyocytes are a cell type which allows for beating patches to be created, the use of human cells is restricted to stem cells which must arduously be differentiated into the desired lineage. In our group we work with mesenchymal stem cells derived from umbilical cord tissue, which is a neonatal tissue providing stem cells with high proliferation capacity and low senescence in passaging [6].

Due to difficulties in cardiomyogenic differentiation of the available forms of human stem cells, bioreactors play an important role in this field of research. A bioreactor is a device which creates a specific environment for cultured tissues. As each tissue is different in function and microenvironment, individual bioreactors must be developed for the conditioning of specific tissues [7]. There are various forms of stimulation for myocardial tissue engineering, ranging from mechanical pull, push and different forms of fluid flow. Schachar and Cohen have suggested that future bioreactors in this specific field must incorporate perfusion and mechanical stimulation [8] to be effective.

The bioreactor described herein aims to simulate the environment found in the human heart with its pulsatile blood flow and electrical stimulation. It is with this environment that we aim to stimulate mesenchymal stem cells colonized on polyurethane scaffolds to differentiate into cardiomyocytes. Below we describe initial feasibility testing for our bioreactor.

2. Materials and methods

2.1. Bioreactor conception and construction

Our bioreactor encompasses a 300 × 200 × 250 mm (width × depth × height) system consisting of an actuation unit for pulsatile flow generation, a stimulation unit for electrical pulse generation and a control unit for synchronization of mechanical and electrical stimulation (see Fig. 1). Vertically the system is divided into a dry lower section and a fluid filled sterile upper section by a flexible silicone separation membrane. In the lower section a standard linear drive (NTI AG, Spreitenbach, Switzerland) is used to displace a cylinder head periodically. This results in air compression thus causing the separation membrane to bulge upwards and displace the sterile fluid above it. A Teflon^® scaffold carrier with a scaffold fixed between its two discs is placed in the sterile upper chamber. When activated the bioreactor subsequently causes the displacement of fluid in the sterile upper chamber and stimulates the patch which lies parallel to the separation membrane. The lid of the sterile chamber is outfitted with Luer lock connections to which sterile filters can be attached for sterile aeration of the chamber.

Fig. 1.

Multimodal bioreactor for cardiac patch conditioning. The bioreactor consists of an actuation unit for fluid flow generation, a stimulation unit for electrical pulse generation and a control unit for synchronization of mechanical and electrical stimulation.

For electrical stimulation of the patches, two electrodes are integrated into the scaffold carrier. When in position the electrodes of the scaffold carrier are connected to analog electrodes leading through the wall of the bioreactor to an external dual chamber pacemaker (Medtronic Inc., Fridley, MN, USA). The internal software of the pacemaker is used to adjust the output to physiologic stimulation in the left ventricle (mode of operation: VOO).

The software LinMot-Talk (NTI AG, Spreitenbach, Switzerland) is used to set a defined movement curve of the piston for mechanical stimulation. In this manner it was possible to generate a physiological pressure-time curve and trigger pacemaker signaling at predefined time points along this curve (see Fig. 2).

Fig. 2.

Depiction of piston position in bioreactor relative to time as per programing in LinMot^® controller, indicating the time point of electrical stimulation.

2.2. Bioreactor sterilization

The bioreactor’s upper sterile chamber was separated into its individual parts, the individual pieces were washed with water and cleaned with a 70% isopropanol solution. These were then dried and bagged in gas sterilization bags. Gas sterilization was performed in-house in the sterilization department of the Klinikum Grosshadern, Ludwig Maximilian University of Munich (LMU).

2.3. Bioreactor functionality and sterility testing

The sterilized bioreactor was built together in an aseptic manner in a laminar flow chamber and a sterilized patch was sewn and fixed to the Teflon^® scaffold carrier (see Fig. 3b). This was mounted in the conditioning chamber. The conditioning chamber was filled with Medium 199 Earle (Biochrom AG, Berlin, Germany). The bioreactor was closed, the Luer lock air vents at the top of the bioreactor were outfitted with sterile filters (Nalge Nunc, Rochester, NY, USA) and the bioreactor was placed in a standard incubator at 37 °C and 5% CO₂ (Fig. 3a). The bioreactor was connected to its energy source and the engine was set to replicate a flow of 500 ml/min at 60 bpm (representing a physiological heart rate in the adult heart). The electrical stimulation was set to 10 mA, 3 V at a frequency of 60 Hz. The unseeded patch was conditioned for 72 h after which 2 samples of the medium were removed in an aseptic fashion and sent for bacteriological and mycological evaluation (testing for aerobic and anaerobic bacterial, candida and aspergillus species) by the Max von Pettenkofer-Institute of Hygiene and Medical Microbiology, LMU Munich.

Fig. 3.

(a) Bioreactor filled with medium in incubator prior to functionality testing; (b) Scaffold carrier with fixated PU scaffold – Teflon^® ring with medium demarcating the seeding area on the PU; c) Bioreactor with conditioning chamber removed and placed on right – cover of conditioning chamber on lower left.

2.4. Cell isolation and culture

All of the cells used in this study were isolated from the umbilical cords of newborn in the Perinatal Centre, Department of Gynecology and Obstetrics, Grosshadern Medical Centre, LMU Munich. The use of umbilical cords was approved by our institutional ethics committee and all families consented to use of the tissue. Isolation of umbilical cord derived Mesenchymal Stem Cells (MSC) was begun within 10 hours of childbirth. Umbilical cord arteries were removed from the cord and placed in a Petri dish with Phosphate Buffered Saline (PBS; PAA Laboratories GmbH, Pasching, Austria). Wharton’s Jelly was minced and placed in an enzyme solution (670 U/ml Hyaluronidase (AppliChem GmbH, Darmstadt, Germany), 300 U/ml Collagenase Type CLS (Biochrom AG, Berlin, Germany) in PBS) as described elsewhere [9]. The Wharton’s Jelly was incubated in the enzyme solution for 60 min at 37 °C and 5% CO₂. The enzyme solution was collected and centrifuged at 1200 rpm for 5 min. The supernatant was discarded, the cells re-suspended in 12 ml Growth Medium 1 (GM1) [9] and transferred to a 75 cm² cell culture flask (Corning Incorporated, Corning, USA). The arteries were removed before transferring to a further 75 cm² cell culture flask with the same volume of medium.

2.5. Cell type verification

To verify that the cells isolated fulfilled the characteristics of MSC, flow cytometry was used. Cells from passages 1-2 were analyzed by dividing 0.8 × 10⁶ cells equally into 8 FACS tubes in PBS supplemented with 10% Fetal Bovine Serum (FBS; PAA Laboratories GmbH, Pasching, Austria). After washing steps, 100 μl cell suspensions were stained with allophycocyanin-conjugated murine antibodies against human CD45, CD73, CD90, CD105, CD34 and HLA-DR (all Becton Dickinson GmbH, Heidelberg, Germany) for 45 min at 4 °C. This testing panel allows for a classification of the cells as MSC when >95% of the cells are CD73⁺, CD90⁺, CD105⁺ and <2% are CD45⁺ and CD 34⁺, according to the International Society for Cellular Therapy consensus conference [10]. Propidium Iodide (PI) exclusion was performed after further washing steps by staining dead cells with 0.2 μg/ml PI (Sigma Aldrich GmbH, Deisenhofen, Germany). Flow cytometry was performed with a FACSCalibur^{^TM} flow cytometer (Becton Dickinson GmbH, Heidelberg, Germany). Cells were only used for experiments when the criteria defined above were confirmed for the cells.

2.6. Scaffolds

The scaffolds used for cell seeding and subsequent bioreactor stimulation were sprayed polyurethane scaffolds (Vasomer^®/PU; ITV, Denkendorf, Germany). PU was gamma sterilized prior to use [11].

2.7. Seeding scaffolds with mesenchymal stem cells

The PU scaffolds were sewn to the scaffold carrier in such a way that the PU was not tense. This was placed in a culture dish (Corning Incorporated, Corning, NY, USA) and a further Teflon^® ring with a diameter of 3 cm was placed in the center of the scaffold to provide the outer perimeter for cell seeding. Prior to seeding the culture dish was filled with 30 ml GM1 with a portion of the volume pipetted onto the area of the scaffold to be seeded later. The construct was placed in an incubator for 30 min at 37 °C and 5% CO₂ for preconditioning. The cells were seeded on the scaffold in the boundaries of the seeding perimeter ring suspended in GM1, after removal of the preconditioning fluid (seeding density 0.750 × 10⁶ cells/cm²).

2.8. Bioreactor conditioning

The seeded patch was left in the filled culture dish for 48 h of seeding time (medium exchange after 24 h). The conditioning chamber of the bioreactor which was assembled as above was filled with 250 ml of GM1 covering the patch. The bioreactor was set to a pulsatile stimulation of 500 ml/min and 10 mA, 3 V electrical stimulation – both pulsatile and electrical stimulation were set to 60 Hz. After 72 h of conditioning the patch was removed from the bioreactor and different sections of the patch allocated to different methods of evaluation. In this feasibility study 4 patches were seeded, stimulated and evaluated.

2.9. Syto9-PI stain

Syto9 (5 μM; Invitrogen GmbH, Karlsruhe, Germany) and Propidium Iodide stock (50 μM; Sigma Aldrich GmbH, Hamburg, Germany) were stored in dark boxes at −20 °C. Staining was performed in the first minutes after removal of the patch from the bioreactor. The colonized patches (n = 4) were carefully placed on 300 μl Syto9-PI solution mix and visualized using an Axio Observer fluorescence microscope (Carl Zeiss MicroImaging GmbH, Goettingen Germany). Syto9 is a DNA intercalating dye which is able to enter all cells. Propidium is only able to diffuse into cells with a reduced integrity of the cellular membrane - there it displaces Syto9. One can consider all cells stained with Syto9 (green fluorescence) to be vital and those stained with PI (red fluorescence) to be avital.

2.10. Histological evaluation

Initially the patches (n = 4) were placed in 4% formalin solution (Microcos GmbH, Garching, Germany) and left in the solution for 10 days at 4 °C before alcohol dehydration, paraffin embedding and microtome sectioning (Frigocut 2700; Leica, Brensheim, Germany) at a section thickness of 5 μm. The sections were deparaffinised and rehydrated, then stained with Haematoxylin and Eosin (H&E). The samples were visualized using a Leica DMR Microscope (Leica Microsystems GmbH, Wetzlar, Germany).

2.11. Scanning electron microscopy

Patches (n = 4) were placed in a fixative, consisting of 1.5 mM HCl (Pharmacy, Grosshadern Medical Centre, Munich, Germany), 8% gluataraldehyde (Sigma-Aldrich Chemie GmbH, Steinheim, Germany) and 1.13 mg/ml sodium cocodylate trihydrate (Sigma-Aldrich Chemie GmbH, Steinheim, Germany) in distilled water (Ampuwa, Fresenius Kabi Deutschland GmbH, Bad Homburg v.d. H., Germany), for a minimum of 24 h. Patch dehydration was performed via an ascending alcohol series. After dehydration the patches were critical point dried (Critical Point Dryer CPD 030; Bal-Tec GmbH, Schalksmuehle, Germany) and sputter coated with gold for 180 s at a pressure of 10⁻⁵ mbar (Sputter Coater SCD050; Bal-Tec GmbH, Schalksmuehle, Germany). The samples were then visualized using an EVO^® LS 10 scanning electron microscope (Carl Zeiss MikroImaging GmbH, Goettingen, Germany).

3. Results

3.1. Functionality and sterility testing

The bioreactor was found to work in the humid and warm environment of the incubator without pause for 72 h, continually stimulating the patches. Microbiological studies displayed that the medium in the bioreactor stayed free of aerobic and anaerobic bacteria, candida and aspergillus (microbiological culture time 10 days).

3.2. Cell vitality of stimulated patches

The majority of the area of the individual patches was covered with interlocking Syto9 stained vital cells. Interspersed between the vital cells individual PI stained cells with decreased cell membrane orientation were found (see Fig. 4a). In parallel to the SEM results a zone with a higher density of PI stained cells was found at the border of the seeded patch (see right side of Fig. 4b).

Fig. 4.

(a) Syto9-PI stain of conditioned patch: green indicates vital Syto9 stained cells and red indicates PI stained cells with a lower integrity of the cellular membrane. (b) MosaiX image of Syto9-PI stained conditioned patch (i.e. numerous fluorescence images stitched to create a larger reproduction of the patch) – on the top is an outer zone with a higher density of PI stained cells (see arrow) c+d) H&E images of conditioned patches (the arrows denote the cell layer) (In a size bar = 100 μm, in b = 5 mm, in c = 200 μm and in d = 50 μm).

3.3. Histology of stimulated patches

The cells were found to be distributed in a multilayer on the PU scaffold (see Fig. 4c and 5d). Individual cell nuclei were visible and the thickness of the cell layer remained similar throughout the evaluated section of the conditioned patch.

Fig. 5.

Scanning electron micrographs of conditioned patches a) Low magnification image with a patch cut apart in the center (tears are artifacts from the cutting in this case; 1 denotes the cell layer while 2 denotes the PU layer) b+c) Border zone of the patch with individual mesenchymal stem cells and their interaction with the PU fibers visible d+e) Center of the conditioned patch with individual cell borders visible (visible cell-cell interconnections are denoted by arrows); few spherical cells can be found on the patch surface (In a size bar = 200 μm, in b = 40 μm, in c = 40 μ m, in d = 100 μm and in e = 40 μm).

3.4. Scanning electron microscopy of stimulated patches

The cells found on the scaffold were packed in a dense confluent cell layer (see Fig. 5a and 5d). On the surface of the patches individual cell extensions and cell – cell connections could be identified (see Fig. 5e). Only few spherical cells (indicating dead cells falling from the patch) were found on the surface of center portion of the stimulated patches.

4. Discussion

This initial study of stem cell seeded patch conditioning displays the functionality and feasibility of our bioreactor. Its compact design and lightweight construction allows easy handling and availability for placement in a standard incubator. Not only did the bioreactor continuously function in the incubator environment, our bioartificial MSC-PU patches were able to maintain viability and structural integrity during dual pulsatile perfusion and electrical stimulation. We recently reported the feasibility and benefits of our bioartificial MSC-PU patches [11] with the current study presenting the next step in our investigation.

In our previous experiments we compared polyurethane, Collagen Cell Carriers, ePTFE and Spider Silk Protein 1-RGD coated ePTFE regarding their properties as scaffolds for umbilical cord derived mesenchymal stem cells. Initially we were able to identify PU as the most receptive of scaffolds. Cells spread nicely on the scaffolds and did not clump at high cell seeding densities as they did on all other scaffolds. Further experiments showed us that cell seeding of PU was most feasible at a cell density of 0.750 × 10⁶/cm² with the cells spreading in a multicellular layer over the complete PU scaffold. We showed that seeding the PU scaffolds at this density did not have a significant effect on the elastic modulus of the bioartificial patch and that the patches could survive for a minimum of 35 days with continuously increasing mitochondrial activity. Similar results regarding the advantageous properties of PU for cell seeding have also been reported by others [12].

After bioreactor stimulation MSC were found to be aligned in a multilayer formation with a confluent cell surface on the stimulated patches. The histological and scanning electron microscopic presentation (alignment and texture) of the patches was identical to when the patches were not stimulated electrically and mechanically. In scanning electron micrographs minimal cell separation with few spherical cells on the surface of the cell layer were noted. Syto9-PI staining showed that the cells stayed vital during stimulation on the scaffold. Merely a border zone of our stimulated patches was found to have a high density of PI stained dead cells. We assume that after seeding minimal movement of the Teflon^® ring, used as a seeding perimeter, causes disruption of the cell layer in this zone. Further modification of the seeding process (for example by the use of a seeding ring made of a different material or made to be thinner than the current one) is expected to resolve this issue.

We plan to incrementally extend the time of stimulation and then further evaluate the differentiation of the stem cells on the scaffold. Over a decade ago myocardial tissue engineering bioreactors were mainly based on stretch as form of stimulation [13,14]. These bioreactors were able to condition neonatal rat cardiomyocytes and allowed hypertrophy and better alignment of the cells. Later on pulsatile perfusion began to play a role [15] and only recently have bioreactors been developed for electrical stimulation of cardiac patches [16–18].

Aiming to differentiate human mesenchymal stem cells cardiomyogenically, we must create a very precise environment for stimulation. For this reason the bioreactor developed in our group combines two forms of stimulation, namely pulsatile flow conditions and electrical stimulation (see Fig. 2). The pulsatile flow simulates hemodynamic pulsatility in the heart as well as myocardial contraction. Furthermore the pulsatile action allows for a continuous convective flow in the conditioning chamber for cell metabolism. Synchronous to the pulsatile action our bioreactor stimulates electrically at the same frequency as the pulsatile thrusts below the patch.

MSC have been found to have differentiation capacity towards a cardiomyogenic lineage [19]. In experiments published by Hollweck et al. the treatment of umbilical cord derived MSC with oxytocin led to the development of several proteins which are found in cardiac myocytes. The cells contain proteins involved in cardiac contraction: cardiac actin, cardiac actinin, sarcomeric actin, sarcomeric actinin and troponin T. They furthermore contained the protein connexin-43 which is involved in gap junctions vital to cell-coupling in the cardiac contraction. While the experiments led to the production of cardiac proteins no functional activity such as contractility could be shown. We believe that for full differentiation (including the functional component) it is important to recreate the heart’s precise in vivo conditions. We hypothesize that the constant stimulation with an electrical current as well as with a pulsatile fluid flow should support the development of functionality in the cardiac differentiated MSC. To this length we believe that the electrical stimulus is more important than the mechanical one. In further experiments we hope to be able to show this by combining chemical, mechanical and electrical stimulation in various combinations and then investigating protein production in the cells as well as testing cell functionality.

5. Conclusion

Our bioreactor system allows for exact definition of a desired stroke-time profile and electrical stimulation triggering at any chosen time in respect to mechanical stimulation. We aimed to show that this bioreactor doesn’t have direct adverse effects on our constructs. Our results display that we can stimulate our constructs via pulsatile flow and electrically without disrupting our construct’s integrity. These results warrant further investigation of our novel bioreactor incorporating initial investigations of cellular differentiation. If we could eventually develop a cardiac patch consisting of differentiated human cardiomyocytes a step would have been made towards advancing a new treatment for patients whose myocardium is terminally damaged after myocardial infarctions.

Footnotes

Acknowledgements

The authors would like to thank the Perinatal Centre Großhadern, LMU Munich, for providing umbilical cords.

Ethical approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. This article does not contain any studies with animals performed by any of the authors.

Conflict of interest

All authors declare that they have no conflict of interest.

References

Shiba

Nochioka

Miura

Kohno

and Shimokawa

, Trend of westernization of etiology and clinical characteristics of heart failure patients in Japan, Circulation Journal 75(4) (2011), 823–833.

Woo

K.S.

Yip

T.W.

Kwong

S.K.

Chook

and L.C.

Kum

, Commentary: Smoking and atherosclerotic diseases in Asia–the implication in global atherosclerosis prevention, International Journal of Epidemiology 34(5) (2005), 1045–1046.

Mathers

C.D.

and Loncar

, Projections of global mortality and burden of disease from 2002 to 2030, PLoS Medicine 3(11) (2006), e442.

Lund

L.H.

Matthews

and Aaronson

, Patient selection for left ventricular assist devices, European Journal of Heart Failure 12(5) (2010), 434–443.

Zimmermann

W.H.

Melnychenko

Wasmeier

Didie

Naito

Nixdorff

, Engineered heart tissue grafts improve systolic and diastolic function in infarcted rat hearts, Nat Med. 12(4) (2006), 452–458.

Hass

Kasper

Bohm

and Jacobs

, Different populations and sources of human mesenchymal stem cells (MSC): A comparison of adult and neonatal tissue-derived MSC, Cell Communication and Signaling: CCS 9 (2011), 12.

Portner

Nagel-Heyer

Goepfert

Adamietz

and N.M.

Meenen

, Bioreactor design for tissue engineering, J Biosci Bioeng 100(3) (2005), 235–245.

Shachar

and Cohen

, Cardiac tissue engineering, ex-vivo: Design principles in biomaterials and bioreactors, Heart Fail Rev 8(3) (2003), 271–276.

Seshareddy

Troyer

and M.L.

Weiss

, Method to isolate mesenchymal-like cells from Wharton’s Jelly of umbilical cord, Methods in Cell Biology 86 (2008), 101–119.

10.

Dominici

Le Blanc

Mueller

Slaper-Cortenbach

Marini

Krause

, Minimal criteria for defining multipotent mesenchymal stromal cells, The International Society for Cellular Therapy Position Statement. Cytotherapy 8(4) (2006), 315–317.

11.

Herrmann

F.E.

Lehner

Hollweck

Haas

Fano

Fehrenbach

, In vitro biological and mechanical evaluation of various scaffold materials for myocardial tissue engineering, Journal of Biomedical Materials Research Part A 102(4) (2014), 958–966.

12.

Rockwood

D.N.

Akins

R.E.

Jr Parrag

I.C.

Woodhouse

K.A.

and Rabolt

J.F.

, Culture on electrospun polyurethane scaffolds decreases atrial natriuretic peptide expression by cardiomyocytes in vitro, Biomaterials 29(36) (2008), 4783–4791.

13.

Fink

Erguen

Kralisch

Remmers

Weil

and Eschenhagen

, Chronic stretch of engineered heart tissue induces hypertrophy and functional improvement, The FASEB Journal 14 (2000), 669–679.

14.

Liu

Montazeri

Jedlovsky

van Wert

Zhang

, Bio-stretch, a computerized cell strain apparatus for three-dimensional organotypic cultures, In Vitro Cellular & Developmental Biology - Animal (1999), 87–93.

15.

Brown

M.A.

Iyer

R.K.

and Radisic

, Pulsatile perfusion bioreactor for cardiac tissue engineering, Biotechnology Progress 24 (2008), 907–920.

16.

Barash

Dvir

Tandeitnik

and Ruvinov

, Electric field stimulation integrated into perfusion bioreactor for cardiac tissue engineering, Tissue Engineering Part C, Methods 16(6) (2010).

17.

Tandon

Marsano

Maidhof

Numata

Montouri-Sorrentino

Cannizzaro

, Surface-patterned electrode bioreactor for electrical stimulation, Lab Chip 10(6) (2010), 692–700.

18.

Mooney

Mackle

J.N.

Blond

D.J.

O’Cearbhaill

Shaw

Blau

W.J.

, The electrical stimulation of carbon nanotubes to provide a cardiomimetic cue to MSCs, Biomaterials 33(26) (2012), 6132–6139.

19.

Hollweck

Hartmann

Eblenkamp

Wintermantel

Reichart

Überfuhr

, Cardiac differentiation of human Wharton’s Jelly stem cells – Experimental comparison of protocols, The Open Tissue Engineering and Regenerative Medicine Journal 4 (2011), 95–102.