Silk protein fibroin from Antheraea mylitta for cardiac ... - CyberLeninka

Biomaterials 33 (2012) 2673e2680

Contents lists available at SciVerse ScienceDirect

Biomaterials journal homepage: www.elsevier.com/locate/biomaterials

Silk protein fibroin from Antheraea mylitta for cardiac tissue engineering Chinmoy Patra a, Sarmistha Talukdar b, Tatyana Novoyatleva a, Siva R. Velagala a, Christian Mühlfeld c, d, Banani Kundu b, Subhas C. Kundu b, *, Felix B. Engel a, ** a

Department of Cardiac Development and Remodelling, Max-Planck-Institute for Heart and Lung Research, Parkstrasse 1, 61231 Bad Nauheim, Germany Department of Biotechnology, Indian Institute of Technology, Kharagpur 721 302, India c Institute of Anatomy and Cell Biology, Justus-Liebig-University Giessen, Aulweg 123, 35385 Giessen, Germany d Institute of Functional and Applied Anatomy, Hannover Medical School, Carl-Neuberg-Str. 1, 30625 Hannover, Germany b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 1 November 2011 Accepted 17 December 2011 Available online 10 January 2012

The human heart cannot regenerate after an injury. Lost cardiomyocytes are replaced by scar tissue resulting in reduced cardiac function causing high morbidity and mortality. One possible solution to this problem is cardiac tissue engineering. Here, we have investigated the suitability of non-mulberry silk protein fibroin from Indian tropical tasar Antheraea mylitta as a scaffold for engineering a cardiac patch in vitro. We have tested cell adhesion, cellular metabolic activity, response to extracellular stimuli, cellto-cell communication and contractility of 3-days postnatal rat cardiomyocytes on silk fibroin. Our data demonstrate that A. mylitta silk fibroin exhibits similar properties as fibronectin, a component of the natural matrix for cardiomyocytes. Comparison to mulberry Bombyx mori silk protein fibroin shows that A. mylitta silk fibroin is superior probably due to its RGD domains. 3D scaffolds can efficiently be loaded with cardiomyocytes resulting in contractile patches. In conclusion, our findings demonstrate that A. mylitta silk fibroin 3D scaffolds are suitable for the engineering of cardiac patches. Ó 2011 Elsevier Ltd. All rights reserved.

Keywords: Non-mulberry silk Fibroin Adhesion Cardiac tissue engineering RGD peptide Connexin 43

1. Introduction Cardiovascular diseases (CVD) remain among the leading causes of death worldwide. The underlying mechanism of most CVD is the loss of cardiomyocytes that cannot be replaced with newly formed cardiomyocytes based on homeostatic mechanisms [1]. Cardiomyocyte death leads to reduced heart function, a diminished quality of life and inevitable progression to overt heart failure. Current treatment of congestive heart failure attempting to minimize cardiomyocyte death has reduced early mortality from myocardial infarction. However, despite considerable progress in the treatment of heart failure, mortality rates are still high with as much as 30% within one year [2]. One reason for this poor prognosisis the limited options to replace damaged or lost cardiac tissue other than organ transplantation. It is expected that the burden of CVD will increase in the future as life expectancy and the incidence of cardiac risk factors like obesity and diabetes increases. The high morbidity, mortality and cost associated with CVD necessitate the * Corresponding author. Department of Biotechnology, Indian Institute of Technology, Kharagpur-721 302, India. Tel.: þ91 3222 283764; fax: þ91 3222278433. ** Corresponding author. Max-Planck-Institute for Heart and Lung Research, Parkstrasse 1, 61231 Bad Nauheim, Germany. Fax: þ49 6032 705211. E-mail addresses: [email protected] (S.C. Kundu), [email protected] (F.B. Engel). 0142-9612/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.biomaterials.2011.12.036

development of improved treatments to prevent, treat and reverse heart disease. One solution to this problem might be the engineering of 3D cardiac tissues outside of the body that can be adjusted according to the needs of a patient before implantation. One of the first approaches was the embedment of cardiomyocytes in a mixture of type I collagen gel and Matrigel. These engineered cardiac tissues improved left ventricular function after myocardial infarction and proved that implantation of engineered tissue can indeed be used to treat CVDs [3]. However, this and similar approaches appear to be suboptimal regarding clinical applicability. Collagen gels [4] and foams [5] are characterized by mechanical weakness and random microstructures [6]. Synthetic polymers in addition exhibit often excessive stiffness [7,8]. An alternative approach is the use of the natural environment. For this purpose, hearts were decellularized with sodium dodecyl sulfate (SDS) and repopulated [9]. This approach resulted in functional 3D cardiac tissue, however, it exhibited high mechanical stiffness likely due to extracellular matrix compaction or through alterations of the extracellular matrix due to SDS treatment [9]. Thus, a new focus in the field is the development of advanced material strategies with an emphasis on mimicking native structural, mechanical, and transport properties by microfabricating an accordion-like honeycomb scaffold [10]. However, so far there is no approach available for the generation of clinically applicable 3D cardiac tissue patches.

2674

C. Patra et al. / Biomaterials 33 (2012) 2673e2680

In order to engineer a 3D tissue one aims to fabricate a biodegradable scaffold that mimics the in vivo extracellular matrix which enables cells to attach, migrate, grow and function properly [11]. A great variety of natural (collagen, hyaluronan, chitosan, cellulose, mulberry silk) and synthetically formulated materials have been tested for 3D scaffold constructions [12]. Synthetic materials are cheap, reproducible and easy to manufacture but their degradation products are often toxic and cause chronic inflammation [13]. Materials of natural origin are recognized by the in vivo biological environment and are degraded through established metabolic processes [14]. However, they often cause a host immunogenic response that results in implant rejection. Amongst the natural biomaterials available, silkworm silk protein fibroin possesses unique mechanical strength, biocompatibility [15e19] and relative ease in fabrication into diverse morphologies [20]. Antheraea mylitta silk protein fibroin scaffolds (2%) measuring 10 mm in diameter and 10 mm in height showed a compressive strength of 84 kPa [21]. This observed strength is higher than other reported naturally derived materials used for 3D scaffold fabrication (e.g. collagen and chitosan scaffolds have a mechanical strength of 15 kPa and 45 kPa, respectively) [22]. Scaffolds of 6.4 mm in diameter and 0.2 mm thickness showed compressive modulus of 5.2 kPa at frequency of 0.005 Hz, and dynamic stiffness of 8 kPa at frequency of 1.5 Hz [17]. Non-mulberry silk protein fibroin is mechanically superior to mulberry silk [21] and possesses RGD sequences (GenBank: AY136274.1) [23]. Its non-cytotoxic property and low level of inflammatory response [24,25] make it an excellent material for tissue engineering. In this study, we have evaluated the potential of silk protein fibroin of Indian tropical tasar silkworm A. mylitta (AM) as biomaterial for in vitro 3D cardiac tissue engineering using 3-days postnatal cardiomyocytes by analyzing cellular metabolic activity, cellto-cell communication, cell cycle activity and contractility. For comparison we utilized silkworm Bombyx mori silk fibroin (BM), gelatin and fibronectin.

2. Materials and methods 2.1. Rat neonatal cardiomyocyte isolation and cell culture Animal experiments were approved by the local Committee for Care and Use of Laboratory Animals (Regierungspräsidium Darmstadt, Gen. Nr. B 2/202) and conform to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996). Ventricular cardiomyocytes were isolated from 3-days-old (P3) Sprague Dawley rats as described [26]. For selective enrichment of cardiomyocytes, cells were preplated for 2 h (2 mM L-glutamine, 10% FBS and 100 U/mg/ml Pen/Strep in DMEM/F12 medium). Non-attached cells were collected, centrifuged, resuspended in neonatal cardiomyocyte medium (3 mM Na pyruvate, 2 mM L-glutamine, 0.1 mM ascorbic acid, 1 mg/ml insulin/transferrin/selenium, 0.2% BSA and 100 U/mg/ml Pen/Strep in DMEM/F12 medium) supplemented with 5% horse serum and seeded in 500 ml per 24-well for 48 h if not stated otherwise. Cells were cultured at 37 C in a 5% CO2/95% air humidified atmosphere.

2.2. Preparation of silk protein fibroin solution Indian tropical tasar A. mylitta silk larvae were obtained from Tasar Sericulture Farm, West Midnapur of West Bengal State of India. Silk protein fibroin from A. mylitta and B. mori silkworms and cocoons were prepared as described previously [27,28]. In brief, fibroin protein was collected by squeezing excised silk glands from fully-grown A. mylitta larvae after washing with distilled water. Fibroin was dissolved in 1%(w/v) SDS aqueous solution containing 10 mM Tris (pH 8.0) and 5 mM EDTA. Isolated fibroin protein was either used immediately or stored at 20 C. Fibroin protein from B. mori silkworm was prepared from the cocoon. Sericin protein was removed from cocoons by boiling in a 0.02 M Na2CO3 solution for 1 h. After thoroughly rinsing with distilled water the degummed fibroin fibers were dissolved in 9.3 M LiBr solution at 60 C and dialyzed (MWCO 12000) against distilled water for 3 days. Fibroin concentration was adjusted to 2% (w/v) by adding distilled water.

2.3. Fabrication of 3D silk fibroin scaffolds The silk fibroin solutions (2% w/w) from both A. mylitta silk gland and B. mori cocoons were cast into moulds, frozen at 20 C for 8 h and then lyophilized to yield porous 3D silk fibroin scaffolds, having a diameter of 13 mm and a thickness of 2 mm [18]. The lyophilized silk fibroin scaffolds were analyzed by scanning electron microscopy. The size of 30 random pores per scaffold was determined by Image J software. 2.4. Coating procedure for 2D cell culture Glass coverslips (B 12 mm, Karl Hecht GmbH, Germany) were placed in 24-well tissue culture plates, washed once with 70% ethanol and twice with 100% ethanol for 3 min. Air-dried coverslips were incubated with 100 ml of 10 mg/ml fibronectin (PromoCell, Germany)/PBS or 0.5% (w/v) silk fibroin/PBS or 1% (w/v) gelatin (Sigma)/ water for 2 h at 37 C. Solutions were aspirated and coverslips were treated for 30 min with UV light before cardiomyocytes were seeded. 2.5. Immunohistochemistry Paraffin sections (5 mm) were deparaffinized and rehydrated following standard protocols [29]. Antigen retrieval was achieved by boiling sections in 1 mM EDTA (pH 8.0) for 8 min. Cultured cells were fixed in 3.7% formaldehyde (Sigma) for 15 min. All antibodies were diluted in blocking buffer as indicated, and all manipulations were carried out at room temperature. Samples were permeabilized for 10 min with 0.5% Triton X-100/PBS. For BrdU incorporation assays, cells were incubated for 30 min in 2 N HCl/1%Triton X-100 after permeabilization. Samples were blocked for 20 min with 5% goat serum/0.2% Tween 20/PBS and incubated for 1 h with primary antibodies (mouse monoclonal anti-sarcomeric a-actinin, rat monoclonal anti-BrdU, 1:100, Abcam, rabbit polyclonal anticonnexin 43 and anti-troponin I, 1:50, Santa Cruz Biotechnology). Immune complexes were detected with ALEXA 488- or ALEXA 594-conjugated antibodies (1:200, Molecular Probes). DNA was visualized with DRAQ-5 (Cell Signaling) or 40 ,60 diamidino-2-phenylindole (DAPI, Sigma) (0.5 mg/ml PBS). Fluorescent images were obtained using a Leica fluorescent microscope or a Zeiss LSM 710 confocal laserscanning microscope. 2.6. Cell adhesion assay Coated coverslips were seeded with 1.2 105 cells. Cell adhesion was quantified by determining the average number of attached cells from 10 randomly chosen microscopic fields at indicated time points. 2.7. Cellular metabolic activity To examine cell viability and cytocompatibility on different 2D matrices 0.4 105 cells were seeded and cultured up to 8 days with media changes every 2 days. MTT assay (Sigma) was performed after 2, 4, 6 and 8 days. 2.8. Arg-Gly-Asp-Ser (RGDS) assays Isolated cardiomyocytes were incubated for 30 min in culture medium containing either 300 mM of an Arg-Gly-Asp-Ser (RGDS) peptide (Calbiochem, La Jolla, CA) or an Arg-Gly-Glu-Ser (RGES) peptide (American peptide Company, Sunnyvale, CA) as control at 37 C incubator in a 5% CO2/95% air humidified atmosphere. Subsequently cells were seeded for 24 h or 48 h in presence of the peptides. 2.9. Cell cycle activity Coated coverslips were seeded with 0.8 105 cells. After 48 h, cells were cultured either in neonatal medium containing 5% horse serum or 50 ng/mlFGF1 (R&D Systems). SB 203580 HCl (p38i) (Tocris) was added freshly every day for 2 days and BrdU (30 mM, Sigma) for the last 24 h. Cell cycle activity was quantified 48 h after stimulation by counting >500 cardiomyocytes from >7 randomly chosen microscopic fields per experiment. 2.10. Generation of cardiac tissue Silk fibroin scaffolds were placed in 6-well plates, treated with 70% ethanol/ dH2O for 30 min, washed with PBS and incubated for at least 2 h in neonatal medium. Scaffolds were then seeded with 5 106 cardiomyocytes suspended in 100 ml neonatal medium supplemented with 10% fetal calf serum (FCS) per 1 cm3 scaffold and were incubated for 2 h at 37 C in a 5% CO2/95% air humidified atmosphere. Subsequently, the amount of medium was increased to immerse the loaded scaffolds, which were incubated up to 3 weeks with media changes every second day.


2.12. Statistical analysis Data are expressed as the mean SEM of at least three independent experiments. Statistical significance of differences was evaluated by one way ANOVA followed by Bonferroni’s post-hoc test (GraphPad Prism). p < 0.05 was considered statistically significant.

3. Results 3.1. Attachment of cardiomyocytes One criterion for choosing a material to engineer tissue is its cell adhesive properties. Therefore, we compared the properties of A. mylitta (AM) and B. mori (BM) silk fibroin in 2D culture with gelatin and fibronectin in regards to adhesion of primary 3-days postnatal rat cardiomyocytes. Gelatin is a cheap matrix that is commonly used for in vitro cardiomyocyte cultures even though it has rather poor adhesive properties. Fibronectin is a natural matrix for cardiomyocytes during heart development and a wellestablished coating material for neonatal cardiomyocyte attachment [30,31]. Therefore, it has been used in this study as positive control.

6h Troponin I/DAPI

30 h Troponin I/DAPI

MTT assay was performed to determine the cytocompatibility of silk with cardiomyocytes. This assay is an indirect measurement of the cellular metabolic activity and was performed on neonatal cardiomyocytes cultured for up to 8 days on the indicated 2D substrates. Cardiomyocytes grown on AM and fibronectin exhibited similar metabolic activity at all investigated time points, which was significantly higher than the activity of cardiomyocytes grown on

48 h Troponin I/DAPI

B

250 200

:6h

150

ns

*

100

: 30 h

: 48 h ns

ns

*

* *

*

50 0

Relative number of cardiomyocytes

C Relative number of non-myocytes

Fibronectin

Silk (BM)

Silk (AM)

G

el

Gelatin

A

3.2. Metabolic activity of cardiomyocytes

at in AM Fi br B on M ec G tin el at in A Fi M br B on M ec ti G n el at in AM Fi br B on M ec tin

Movies of beating cardiomyocytes or beating scaffolds were recorded with a Sony HDR-SR12 camcorder and formatted (Wondershare or iSkysoft video converter).

After cell isolation and cardiomyocyte enrichment, cells were allowed to attach for 6, 30 and 48 h. Cell counting experiments after troponin I/DAPI staining revealed that the vast majority of nonmyocytes (troponin I-negative) attached within the first 6 h independently of the utilized matrices (Fig. 1A,C). In contrast, cardiomyocyte attachment was dependent on time and matrix. After 6 h significantly more cardiomyocytes attached on AM compared to BM and gelatin. Moreover, there was no difference between AM and the positive control fibronectin (Fig. 1A,B). At 48 h the majority of cardiomyoctes had attached to all four matrices whereas significantly more cardiomyocytes attached on gelatin, AM and fibronectin compared to BM. Taken together, these data demonstrate that AM is a good substrate for cardiomyocyte attachment. This was further supported by the observation that spreading of cardiomyocytes after 30 h was clearly visible on AM and fibronectin but not on BM or gelatin (Fig. 1A, arrow heads).


2.11. Video image analysis

2675

250

Gelatin

BM

200

AM

Fibronectin

150 100 50 0

0

250 200 150

10

20 30 40 Seeding time [h]

Gelatin

BM

AM

Fibronectin

50

100 50 0 0

10

20 30 40 Seeding time [h]

50

Fig. 1. Cardiomyocytes efficiently attach on silk fibroin. (A) Cardiomyocytes were seeded for indicated times on different matrices and subsequently stained for troponin I (cardiomyocytes, green) and DAPI (nuclei, blue). Arrow heads indicating spread cardiomyocytes on AM and fibronectin. (B,C) quantitative analysis of the number of attached troponin Ipositive cardiomyocytes (B) and non-myocytes (C) (n ¼ 3, mean SEM, *: p < 0.05). Scale bars: 50 mm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

2676


gelatin or BM at days 2, 4 and 6 (Fig. 2A). At 8 days of culture we observed no significant difference among cultures plated on gelatin, AM and fibronectin. However, the metabolic activity of cultures on BM remained significantly lower. The distribution of cardiomyocytes was at day 8 comparable among the different matrices (Fig. 2B). Thus, AM had in contrast to BM no negative effect on the metabolic activity of cardiomyocytes and was comparable to the natural matrix protein fibronectin.

determine the effect of RGDS peptide incubation on cell attachment isolated and enriched cardiomyocytes were allowed to attach for 24 h after 30 min incubation with RGDS or RGES peptides. Incubation with the RGDS peptide inhibited significantly the number of attached cardiomyoctes on AM but not on gelatin or fibronectin (Fig. 3C,D). In addition, we observed that spreading of cardiomyocytes was inhibited on AM as well as fibronectin (Fig. 3C). 3.4. Characterization of growth factor-induced cell cycle activity

3.3. RGD-dependency of cardiomyocyte attachment In order to engineer cardiac tissue it is important to determine that cardiomyocytes seeded on the chosen matrix respond normal to extracellular stimuli. Therefore, we have stimulated cardiomyocytes with horse serum (HS) as well as fibroblast growth factor 1 (FGF1) plus an inhibitor of the mitogen-activated protein kinase p38 (p38i) that are known to induce cell cycle activity [35]. Stimulation with horse serum results usually in a low activation of DNA synthesis in cardiomyocytes seeded on gelatin (12.1 4.4%) (Fig. 4). We observed no significant difference for cardiomyocytes seeded on AM (18.4 4.5%) and BM (15.6 3.8%). However, there was a trend towards a better induction on AM. Stimulation with FGF1/p38i is a strong inducer of cardiomyocyte cell cycle activity and induced BrdU incorporation in 32.0 1.7% of cardiomyocytes seeded on gelatin (Fig. 4A,C). There was no significant difference to cells seeded on BM (36.8 7.0%). In contrast, stimulation of cardiomyocytes on AM and fibronectin was significantly more effective (44.8 6.6%, 46.5 5.5% respectively). These data suggest, that AM does not interfere with the responsiveness of cardiomyocytes to extracellular signals.

Naturally occurring AM contains, like fibronectin, several RGD sequences [23,32]. Previously, it has been shown that presence of RGD sequences in natural as well as synthetic material promotes cell adhesion and spreading via activation of cell surface integrin receptors [33]. In contrast, gelatin and BM are devoid of RGD sequence. To evaluate whether the observed advantages of AM and fibronectin over BM and gelatin are indeed due to the RGD sequences we performed blocking assays with RGDS peptides. To block all cell surface molecules that can interact with RGD sequences we incubated the enriched cardiomyocyte cultures before seeding for 30 min with RGDS peptides. As a control we used the highly similar RGES peptide that does not block RGD binding sites. We performed MTT assays to measure cellular activity at 24 and 48 h and set in the control (incubation with RGES) the "Relative cellular metabolic activity" for each matrix (gelatin, AM, fibronectin) to 100%. Incubation of cells with RGDS peptides had in comparison to incubation with the RGES peptides no effect on cellular metabolic activity of cell cultures on gelatin (Fig. 3). In contrast, there was a significant reduction in activity on AM (100 10.2% to 66 10.5%) and fibronectin (100 7.8% to 73.4 6.0%) at 24 h (Fig. 3A). At 48 h RGDS peptide incubation reduced the metabolic activity in cultures grown on AM (100 23.1% to 59.7 8.9%) but not in cultures grown on fibronectin (Fig. 3B). In conclusion, our data indicate that the cellular activity of cardiomyocytes on AM depends on its RGD sequences. The RGD sequence can mediate cell adhesion and is thus often employed to improve adhesion on biomaterials [33,34]. Thus, the measured reduced cellular metabolic activity after RGDS peptide incubation might be due to reduced cardiomyocyte attachment. To

In order to generate a cardiac tissue it is not only important that cardiomyocytes attach and react to extracellular stimuli but also are able to couple with their neighboring cells and beat synchronously exhibiting a high contractility. In order to exhibit a high contractility it is crucial that cardiomyocytes exhibit well-differentiated sarcomeres that are aligned in parallel. Immunofluorescence analysis based on troponin I demonstrates that sarcomeres in cardiomyocytes grown on AM and fibronectin contain clearly visible

8 days

2 days ns

5

**

ns

** **

**

in AM br B on M ec tin G el at in AM Fi br B on M ec tin G el at in AM Fi br B on M ec tin G el at in AM Fi br B on M ec tin

0

el

Fi

G

Silk (AM)

* ** 4 days

Fibronectin

ns 10

ns

**

8 days in culture alpha Actinin/DAPI alpha Actinin/DAPI

Gelatin

6 days

B

Silk (BM)

15

at

Relative cellular metabolic activity

A

3.5. Cell-to-cell communication and contractility

Fig. 2. Cardiomyocytes exhibit high metabolic activity on A. mylitta silk fibroin. (A) Cardiomyocytes were cultured on the indicated matrices for up to 8 days. Relative cellular metabolic activity was determined by means of an MTT assay. (B) Examples of cardiomyocyte cultures at 8 days stained for sarcomeric a-actinin (cardiomyocytes, green) and DAPI (nuclei, blue) used for quantitative analyses in A. Note that the majority of the attached cells are cardiomyocytes (n ¼ 3, mean SEM, *: p < 0.05, **: p < 0.01). Scale bars: 50 mm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)


Gelatin ns

150

*

**

C Control

RGDS

R G

Gelatin

D S

l tro C on

D S

tro

48 h seeding AM

Gelatin ns

150

R G

R G

C on

D S

tro

l

0

C on

Fibronectin ns

*

Silk (AM)

100 50

D S

l tro

Fibronectin ns

*

100 50

Fibronectin

R G

on C

R G

24 h seeding AM

Gelatin ns

150

D S

tro C

R G

on

D S

tro on C

D

l

0

l

S G R

on C

D

l tro

S R

G

D

l on

tro

S D G R

C

on

tro

l

0

C


Fibronectin

50

B


24 h seeding AM

100

l


A

2677

Fig. 3. Cardiomyocyte attachment on A. mylitta silk fibroin depends on its RGD sequence. Cardiomyocytes were seeded for indicated times on different matrices after 30 min incubation with RGES (control) or RGDS peptides. (A,B) relative cellular metabolic activity was determined by means of an MTT assay and was set for each matrix to 100% in the control. (C) Cardiomyocyte cultures were stained for troponin I (cardiomyocytes, green) and DAPI (nuclei, blue). Scale bars: 50 mm. (D) Quantitative analysis of (C). Number of cells attached to a matrix in the control was set to 100% (n ¼ 3, mean SEM, *: p < 0.05, **: p < 0.01). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

on BM ec tin

0

br

B on M ec tin

0

20

Fi

20

40

in AM

ns

60

at

40

FGF1+p38i ns *

el

60

C

G

5% HS

% of BrdU-positive cardiomyocytes

B

Fi br

Fibronectin

in AM

Silk (BM)

at

Silk (AM)

el

Gelatin

G

TropinI/DAPI/BrdU FGF1+p38i 5% HS

A

cardiomyocytes [36]. Immunofluorescence analyses revealed that cardiomyocytes grown on AM and fibronectin expressed along cellular junctions high amounts of connexin 43 (Fig. 5b). In contrast, cardiomyocytes grown on gelatin or BM expressed low or non-detectable levels of connexin 43. Counting the number of

% of BrdU-positive cardiomyocytes

aligned sarcomeres (Fig. 5A). In contrast, the sarcomeres in cardiomyocytes grown on gelatin as well as BM appeared immature and were mainly not aligned. For intercellular coupling the gap junctional protein connexin 43 is essential as it regulates electrical signal propagation between

Fig. 4. A. mylitta silk fibroin promotes growth factor-induced cell cycle activity. Cardiomyocytes were stimulated with 5% horse serum or FGF1 þ p38 inhibitor and analyzed for DNA synthesis (BrdU). (A) Examples of cardiomyocytes stained for troponin I (green), BrdU (red), and DAPI (nuclei, blue). (B,C) quantitative analysis of the number of BrdU-/troponin I-positive cardiomyocytes after stimulation with 5% horse serum (HS) (B) or FGF1 þ p38 inhibitor (C) (n ¼ 3, mean SEM, *: p < 0.05). Orange arrow: BrdU-positive cardiomyocyte. Yellow arrow: BrdU-positive non-myocyte. Scale bars: 50 mm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

2678


A

Silk (AM)

Silk (BM)

Fibronectin

C

Troponin I

beats/min

Gelatin

60 40 20 0

in

lat

Ge

B

Silk (AM)

Silk (BM)

BM ectin ron Fib

Fibronectin

Connexin43/ DAPI/Actinin

Connexin 43

Gelatin

AM

Fig. 5. A. mylitta silk fibroin promotes cell-to-cell communication and contractility. (A) Examples of cardiomyocytes seeded on different matrices and subsequently stained for troponin I (green) and DAPI (blue). Note that cardiomyocytes on A. mylitta silk fibroin contain well-established and aligned sarcomeres (yellow arrows). (B) Examples of cardiomyocytes seeded on different matrices and subsequently stained for connexin 43 (red), sarcomeric a-actinin (green) and DAPI (blue). Note that cardiomyocytes on A. mylitta silk fibroin express connexin 43 (yellow arrow heads) suggesting well-established cell-to-cell communication. (C) quantitative analysis of the beating frequency. Scale bars: 20 mm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

contraction and relaxation cycles per min indicated that the beating frequency of cardiomyocytes does not depend on the matrices (Fig. 5C). However, in accordance to the connexin 43 data, cardiomyocyte cultures grown on gelatin and BM did not beat synchronously. On gelatin one can observe that neighboring cardiomyocytes beat independently of each other (Supplementary Movies 1). On BM, even at high cellular density, only a subset of cardiomyocytes couple with each other whereas the remaining cardiomyocytes beat independently (Supplementary Movies 2). In contrast, cardiomyocytes on AM or fibronectin contracted synchronously (Supplementary Movies 3 and 4). In summary, AM promotes sarcomere maturation and alignment, cell-to-cell communication and synchronous contractions. Supplementary video related to this article can be found at doi: 10.1016/j.biomaterials.2011.12.036.

3.6. Characterization of 3D cardiac tissue constructs After obtaining promising results in 2D, we attempted to engineer 3D cardiac tissues. 3D AM scaffolds were fabricated, loaded with cardiomyocytes and analyzed. 3D AM scaffolds were prepared by lyophilization of a 2% AM solution. A quantitative analysis of scanning electron micrographs of the cell-free scaffolds demonstrated that the pore size varies from 30 mm2 to 160 mm2 (Fig. 6A,B). To evaluate the cell attachment and distribution on the scaffold we performed laser scanning microscopic analyses of scaffolds that were stained for cardiomyocytes (troponin-I) and nuclei (DRAQ5) 8 days after cardiomyocyte loading. The walls of the scaffold were visible in blue due to autofluorescence. This analysis revealed that cardiomyocytes were attached uniformly on the wall of the pores

(Fig. 6C) containing well-organized and aligned sarcomeres (Fig. 6D; arrow head). Finally, we were interested whether cardiomyocyte loading of 3D AM scaffolds results indeed in contractile 3D tissue. Immunofluorescence analyses of 5 mm thin paraffin sections taken from the center of scaffolds 8 days after loading stained for sarcomeric aactinin and connexin 43 demonstrated that cardiomyocytes attached throughout the scaffold expressing connexin 43 (Fig. 6E). Hematoxylin and Eosin (H&E)-stained sections at 8 and 20 days confirmed the distribution throughout the scaffold. This suggested that AM scaffolds are non-toxic and allow long-term culture of cardiomyocytes (Fig. 6F). Accordingly, we observed that cardiomyocyte-loaded AM scaffolds beat in culture for at least 20 days. Quantitative analysis shows that scaffolds beat synchronously with a frequency of 51.4 5.6 per minute at day 8 (Fig. 6G and Supplementary Movie 5 and 6). Supplementary video related to this article can be found at doi: 10.1016/j.biomaterials.2011.12.036.

4. Discussion The generation of cardiac tissue is highly challenging as the heart is constantly contracting and thus demands special mechanical properties from the utilized scaffold. Additional strategies to repair the heart after injury include regeneration through stem cells [37] or induction of cardiomyocyte proliferation [1], organ- or xenotransplantation [38]. Similar to tissue engineering these approaches are mainly far from translation to clinic or they appear to be not yet efficient in improving heart function. The relative paucity of techniques currently available to repair the


Troponin I

Troponin I/Draq5

D

Troponin I/Draq5

* * *

Connexin43

60 40 20 0 30-50 51-100 > 100 Pore area [µm 2]

Connexin43/ DAPI/Actinin

* *

*

8 days

20 days

Connexin43/Actinin

G

60 beats/min

E

80

20 days

Pore count (%)

*

B

*

F H&E staining

C

8 days

A

2679

40 20 0 AM-scaffold

Fig. 6. 3D A. mylitta silk fibroin scaffolds for cardiac tissue engineering. (A) Example of bright field and scanning electron microscopic pictures of A. mylitta silk fibroin scaffolds. Scale bar: 2 mm (red); 50 mm (yellow). (B) Quantitative analysis of pore sizes of 3D scaffolds. (C) Projection of confocal images of a cardiomyocyte loaded scaffold stained for troponin I (cardiomyocytes, green) and Draq5 (nuclei, blue). Scale bar: 30 mm. (D) Note that cardiomyocytes attach and exhibit well-organized and aligned sarcomeres (white arrow heads). 5 mm thick projection. Blue stars: scaffold. Scale bar: 15 mm. (E) Hematoxylin and Eosin (H&E)-stained sections of 3D loaded scaffolds at 8 and 20 days indicating the biocompatibility of A. mylitta silk fibroin. Scale bar: 50 mm. (F) Examples of cardiomyocyte-loaded scaffold sections at 20 days stained for connexin 43 (red), sarcomeric a-actinin (green) and DAPI (blue). Note that cardiomyocytes on A. mylitta silk fibroin express connexin 43 (yellow arrow heads) suggesting cell-to-cell communication. Scale bar: 25 mm. (G) Quantitative analysis of the beating frequency of cardiomyocyte loaded A. mylitta silk fibroin 3D scaffolds at 8 days. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

mammalian heart necessitates the development of more effective clinical strategies. Our data suggest that silk fibroin from A. mylitta provides a new platform for the development of cardiac therapies. Silk fibroin from B. mori has extensively been studied and shown to provide a surface on which a variety of cell types like osteoblast, keratinocyte, fibroblasts, nerve cells, mesenchymal stem cells, endothelial cells can grow on [15]. It can be used in its pure form or as films, membranes, gels, sponges, nanoparticles and 3D scaffolds making it useful for enzyme immobilization, drug delivery and tissue engineering [15]. In conclusion, silk fibroin has been demonstrated due to its mechanical properties, biodegradability, cytocompatibility, low immunogenicity and controllable porosity to have a great potential to solve a wide range of biomedical problems [15,16]. In previous years Kundu and co-workers have established silk fibroin from Indian tropical tasar silkworm A. mylitta (AM) and shown that it has excellent properties as a biomaterial in terms of cell adhesion, migration, differentiation and proliferation of a number of different cell types [17]. AM is superior to BM in respect to mechanical properties like elasticity and tensile strength [21]. These properties make it an ideal candidate material for the generation of contractile tissue like cardiac patches. For regenerative therapy it is important that the biomaterial is tolerated by the host immune system. The body responds to foreign material depending on its surface topography not its chemical composition [39]. Smooth surfaces induce strong inflammatory reactions, which lead to scarring and the development of a fibrotic capsule around the implant. In contrast, rough surfaces diminish innate immune reaction and prevent implant rejection after transplantation. Surface topography analysis has shown that film as well as thread from A. mylitta fibroin is rougher than fibroin from B. mori [40]. In order to develop a biomaterial for tissue engineering it is important to get adequate interaction between cells and polymer [41]. Cardiomyoctes are characterized by poor attachment to

random surfaces and rather slow attachment and spreading even to components of its natural in vivo matrix. Our data clearly demonstrate that cardiomyocytes attach and spread much faster on AM than BM. As cardiomyocytes on AM responded to extracellular stimuli in the same way as cardiomyocytes attached to fibronectin the interaction between cells and AM appears to be adequate. The observed differences between AM and BM might be due to the fact that AM in contrast to BM possesses repeating RGD sequences [23]. The RGD sequence is the most often employed peptide to stimulate cell adhesion on different biomaterials, as it is a ligand of the cell adhesion integrin receptor family. It has been shown that about half of the 24 integrin receptors recognize the RGD sequence [33,34]. This interaction not only has a role for cell anchoring but also is also important for wound healing, immune response and cell differentiation. Incorporation of RGD sequences in BM has demonstrated to promote cell adhesion properties in chrondocytes [42]. Our data demonstrate that RGD domains in AM are required for cardiomyocyte attachment by incubating cardiac cells with RGDS peptides. 3D AM scaffold loaded with cardiomyocytes spontaneously beat. Cardiomyocytes on 2D AM express higher levels of the gap junctional protein connexin 43 than on BM. This expression was similar to expression in cardiomyocytes on fibronectin and was also observed in 3D AM scaffolds. Moreover, sarcomeres in cardiomyoctes loaded on 3D AM scaffolds were aligned. These data suggest that AM enables cardiomyocyte maturation which might promote contractile strength and electrical coupling of the implant with the host myocardium. In the future it will be important to further optimize the loading of the AM scaffold with cardiomyocytes but also nonmyocytes, to investigate how pore size of the scaffold will affect cell behavior including migration,to challenge the AM-based cardiac tissue to improve its contractility and finally to perform in vivo studies.

2680


5. Conclusion A. mylitta silk fibroin (AM) enables efficient attachment of cardiomyocytes without affecting their response to extracellular stimuli. Cardiomyocytes growing on AM express connexin 43, exhibit aligned sarcomeres and couple electrically with each other resulting in synchronous beating. The generated cardiac tissues are stable and spontaneously beat for at least 20 days. In conclusion, our findings together with the known in vivo properties of silk suggest that AM 3D scaffolds may be suitable for the establishment of therapies for cardiac disease requiring mechanical support and therefore warrants further preclinical investigation. Acknowledgements We are grateful to Petra Freund, Sandra Rühl, Ingrid HauckSchmalenberger and Gerd Magdowski for technical support. This work was supported by a grant from the Alexander von Humboldt Foundation (Sofja Kovalevskaja Award to F. B. E.), the Excellence Cluster Cardio-Pulmonary System (DFG), the International Research Training Group 1566 (PROMISE, DFG), the Department of Biotechnology and Department of Science and Technology, Government of India (S. C. K.). S. C. K. is also grateful to F. B. E. and the Max-Planck-Institute for Heart and Lung Research (MPIHL) for supporting his visit to the MPIHL. References [1] van Amerongen MJ, Engel FB. Features of cardiomyocyte proliferation and its potential for cardiac regeneration. J Cell Mol Med 2008;12:2233e44. [2] Roger VL, Go AS, Lloyd-Jones DM, Adams RJ, Berry JD, Brown TM, et al. Heart disease and stroke statisticse2011 update: a report from the American Heart Association. Circulation 2011;123:e18e209. [3] Zimmermann WH, Melnychenko I, Wasmeier G, Didie M, Naito H, Nixdorff U, et al. Engineered heart tissue grafts improve systolic and diastolic function in infarcted rat hearts. Nat Med 2006;12:452e8. [4] Feng Z, Matsumoto T, Nakamura T. Measurements of the mechanical properties of contracted collagen gels populated with rat fibroblasts or cardiomyocytes. J Artif Organs 2003;6:192e6. [5] Akhyari P, Fedak PW, Weisel RD, Lee TY, Verma S, Mickle DA, et al. Mechanical stretch regimen enhances the formation of bioengineered autologous cardiac muscle grafts. Circulation 2002;106:I137e42. [6] Radisic M, Yang L, Boublik J, Cohen RJ, Langer R, Freed LE, et al. Medium perfusion enables engineering of compact and contractile cardiac tissue. Am J Physiol Heart Circ Physiol 2004;286:H507e16. [7] Radisic M, Park H, Martens TP, Salazar-Lazaro JE, Geng W, Wang Y, et al. Pretreatment of synthetic elastomeric scaffolds by cardiac fibroblasts improves engineered heart tissue. J Biomedical Materials Research Part A 2008;86: 713e24. [8] Boublik J, Park H, Radisic M, Tognana E, Chen F, Pei M, et al. Mechanical properties and remodeling of hybrid cardiac constructs made from heart cells, fibrin, and biodegradable, elastomeric knitted fabric. Tissue Eng 2005;11: 1122e32. [9] Ott HC, Matthiesen TS, Goh SK, Black LD, Kren SM, Netoff TI, et al. Perfusiondecellularized matrix: using nature’s platform to engineer a bioartificial heart. Nat Med 2008;14:213e21. [10] Park H, Larson BL, Guillemette MD, Jain SR, Hua C, Engelmayr Jr GC, et al. The significance of pore microarchitecture in a multi-layered elastomeric scaffold for contractile cardiac muscle constructs. Biomaterials 2011;32: 1856e64. [11] Wang Y, Singh A, Xu P, Pindrus MA, Blasioli DJ, Kaplan DL. Expansion and osteogenic differentiation of bone marrow-derived mesenchymal stem cells on a vitamin C functionalized polymer. Biomaterials 2006;27:3265e73. [12] Willerth SM, Sakiyama-Elbert SE. Combining stem cells and biomaterial scaffolds for constructing tissues and cell delivery. In: Girard Lisa, editor. Stembook; 2008. doi/10.3824/stembook.1.1.1. [13] Laflamme MA, Murry CE. Heart regeneration. Nature 2011;473:326e35.

[14] Li L, Sabir MI, Xu XX. A review on biodegradable polymeric materials for bone tissue engineering applications. J Mater Sci 2009;44:5713e24. [15] Omenetto FG, Kaplan DL. New opportunities for an ancient material. Science 2010;329:528e31. [16] Omenetto FG, Kaplan DL. SnapShot: silk biomaterials. Biomaterials 2010;31: 6119e20. [17] Talukdar S, Nguyen QT, Chen AC, Sah RL, Kundu SC. Effect of initial cell seeding density on 3D-engineered silk fibroin scaffolds for articular cartilage tissue engineering. Biomaterials 2011;32:8927e37. [18] Talukdar S, Mandal M, Hutmacher DW, Russell PJ, Soekmadji C, Kundu SC. Engineered silk fibroin protein 3D matrices for in vitro tumor model. Biomaterials 2011;32:2149e59. [19] Bhardwaj N, Nguyen QT, Chen AC, Kaplan DL, Sah RL, Kundu SC. Potential of 3D tissue constructs engineered from bovine chondrocytes/silk fibroinchitosan for in vitro cartilage tissue engineering. Biomaterials 2011;32: 5773e81. [20] Rockwood DN, Preda RC, Yucel T, Wang X, Lovett ML, Kaplan DL. Materials fabrication from Bombyx mori silk fibroin. Nat Protoc 2011;6:1612e31. [21] Mandal BB, Kundu SC. Non-bioengineered silk gland fibroin protein: characterization and evaluation of matrices for potential tissue engineering applications. Biotechnol Bioeng 2008;100:1237e50. [22] Nazarov R, Jin HJ, Kaplan DL. Porous 3-D scaffolds from regenerated silk fibroin. Biomacromolecules 2004;5:718e26. [23] Datta A, Ghosh AK, Kundu SC. Differential expression of the fibroin gene in developmental stages of silkworm, Antheraea mylitta (Saturniidae). Comp Biochem Physiol B Biochem Mol Biol 2001;129:197e204. [24] Zhang YQ, Zhou WL, Shen WD, Chen YH, Zha XM, Shirai K, et al. Synthesis, characterization and immunogenicity of silk fibroin-L-asparaginase bioconjugates. J Biotechnol 2005;120:315e26. [25] Meinel L, Hofmann S, Karageorgiou V, Kirker-Head C, McCool J, Gronowicz G, et al. The inflammatory responses to silk films in vitro and in vivo. Biomaterials 2005;26:147e55. [26] Engel FB, Hauck L, Cardoso MC, Leonhardt H, Dietz R, von Harsdorf R. A mammalian myocardial cell-free system to study cell cycle reentry in terminally differentiated cardiomyocytes. Circ Res 1999;85:294e301. [27] Mandal BB, Kundu SC. A novel method for dissolution and stabilization of non-mulberry silk gland protein fibroin using anionic surfactant sodium dodecyl sulfate. Biotechnol Bioeng 2008;99:1482e9. [28] Altman GH, Diaz F, Jakuba C, Calabro T, Horan RL, Chen J, et al. Silk-based biomaterials. Biomaterials 2003;24:401e16. [29] Harlow E, Lane D. Staining tissues. In: Cuddihy J, Kuhlman T, editors. Using antibodies: a laboratory manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press; 1998. p. 151e220. [30] Borg TK, Rubin K, Lundgren E, Borg K, Obrink B. Recognition of extracellular matrix components by neonatal and adult cardiac myocytes. Dev Biol 1984; 104:86e96. [31] George EL, Georges-Labouesse EN, Patel-King RS, Rayburn H, Hynes RO. Defects in mesoderm, neural tube and vascular development in mouse embryos lacking fibronectin. Development 1993;119:1079e91. [32] Mandal BB, Das S, Choudhury K, Kundu SC. Implication of silk film RGD availability and surface roughness on cytoskeletal organization and proliferation of primary rat bone marrow cells. Tissue Eng Part A 2010;16:2391e403. [33] Hersel U, Dahmen C, Kessler H. RGD modified polymers: biomaterials for stimulated cell adhesion and beyond. Biomaterials 2003;24:4385e415. [34] Barczyk M, Carracedo S, Gullberg D. Integrins. Cell Tissue Res 2010;339: 269e80. [35] Engel FB, Schebesta M, Duong MT, Lu G, Ren S, Madwed JB, et al. p38 MAP kinase inhibition enables proliferation of adult mammalian cardiomyocytes. Genes Dev 2005;19:1175e87. [36] Guerrero PA, Schuessler RB, Davis LM, Beyer EC, Johnson CM, Yamada KA, et al. Slow ventricular conduction in mice heterozygous for a connexin43 null mutation. J Clin Invest 1997;99:1991e8. [37] Wollert KC, Drexler H. Cell therapy for the treatment of coronary heart disease: a critical appraisal. Nat Rev Cardiol 2010;7:204e15. [38] Ekser B, Cooper DK. Update: cardiac xenotransplantation. Curr Opin Organ Transplant 2008;13:531e5. [39] Ratner BD, Bryant SJ. Biomaterials: where we have been and where we are going. Annu Rev Biomed Eng 2004;6:41e75. [40] Acharya C, Ghosh SK, Kundu SC. Silk fibroin film from non-mulberry tropical tasar silkworms: a novel substrate for in vitro fibroblast culture. Acta Biomater 2009;5:429e37. [41] Thull R. Surface functionalization of materials to initiate autobiocompatibilization in vivo. Materialwiss Werkst 2001;32:949e52. [42] Yanagisawa S, Zhu Z, Kobayashi I, Uchino K, Tamada Y, Tamura T, et al. Improving cell-adhesive properties of recombinant Bombyx mori silk by incorporation of collagen or fibronectin derived peptides produced by transgenic silkworms. Biomacromolecules 2007;8:3487e92.