Prevascularization of self-organizing engineered ...

Cell Tissue Res DOI 10.1007/s00441-012-1492-7

SHORT COMMUNICATION

Prevascularization of self-organizing engineered heart tissue by human umbilical vein endothelial cells abrogates contractile performance Claus Svane Sondergaard & Russell Witt & Grant Mathews & Skender Najibi & Lisa Le & Tracy Clift & Ming-Sing Si

Received: 2 May 2012 / Accepted: 17 August 2012 # Springer-Verlag 2012

Abstract Establishing vascularization is a critical obstacle to the generation of engineered heart tissue (EHT) of substantial thickness. Addition of endothelial cells to the formative stages of EHT has been demonstrated to result in prevascularization, or the formation of capillary-like structures. The detailed study of the effects of prevascularization on EHT contractile function is lacking. Here, we evaluated the functional impact of prevascularization by human umbilical vein endothelial cells (HUVECs) in self-organizing EHT. EHT fibers were generated by the self-organization of neonatal rat cardiac cells on a fibrin hydrogel scaffold with or without HUVECs. Contractile function was measured and force–length relationship and rate of force production were assessed. Immunofluorescent studies were used to evaluate arrangement and distribution of HUVECs within the EHT fibers. RT-PCR was used to assess the transcript levels of hypoxia inducible factor-1a (Hif-1α). EHT with HUVECs manifested tubule-like structures at the periphery

during fiber formation. After fiber formation, HUVECs were heterogeneously located throughout the EHT fiber and human CD31+ tubule-like structures were identified. The expression level of Hif-1α did not change with the addition of HUVECs. However, maximal force and rate of force generation were not improved in HUVECs containing EHT as compared to control EHT fibers. The addition of HUVECs may result in sparse microvascularization of EHT. However, this perceived benefit is overshadowed by a significant decrease in contractile function and highlights the need for perfused vascularization strategies in order to generate EHT that approaches clinically relevant dimensions.

C. S. Sondergaard : R. Witt : G. Mathews : S. Najibi : L. Le : T. Clift Department of Cardiothoracic Surgery, University of California at Davis Medical Center, Sacramento, CA 95817, USA

Successful fabrication of replacement tissues and organs would have a tremendous impact on the treatment of tissue and organ failure due to aging, disease or congenital defects. A fundamental barrier to tissue engineering is the lack of a vasculature for improved mass transfer and scale-up of constructs to clinicallyapplicable size (Ko et al. 2007; Novosel et al. 2011). This is especially important for engineered heart tissue (EHT) because of the high metabolic requirements of actively contracting cardiomyocytes (Balaban 2009). There have been a number of strategies described to prevascularize tissue-engineered constructs whereby a microvasculature is introduced either concurrently with the desired effector cells during the formative stages of the construct or as a separate angiogenic step either in vitro or in vivo by placing the construct in a preformed vascular bed

C. S. Sondergaard : M.-S. Si Institute for Regenerative Cures, University of California at Davis Medical Center, Sacramento, CA, USA M.-S. Si (*) Department of Cardiac Surgery, University of Michigan, 11-735 C.S. Mott Children’s Hospital 1540 E. Hospital Drive, SPC 4204, Ann Arbor, MI 48109, USA e-mail: [email protected]

Keywords Engineered heart tissue . Prevascularization . Tissue engineering . HUVECs

Introduction

Cell Tissue Res

(Borges et al. 2003; Chen et al. 2010; Laschke et al. 2011; Stevens et al. 2009). Prevascularization is thought to be beneficial because it eliminates the time needed for the recipient vasculature to penetrate all regions of the construct after in vivo implantation and therefore promotes more expedient perfusion. Coculture of endothelial cells with cardiomyocytes during EHT fabrication has resulted in the formation of a constructed microvasculature (Caspi et al. 2007; Lesman et al. 2010; Leung and Sefton 2010; Sekine et al. 2008; Sekiya et al. 2006; Stevens et al. 2009). It has, however, not been shown if prevascularization imparts an in vitro functional benefit to the resultant EHT. It is also unclear if this engineered microvasculature represents a bona fide vascular network with respect to function and density. The self-organization model of engineered heart tissue has several advantages over casting-mold and rigid-scaffold seeding methods (Baar et al. 2005; Huang et al. 2007). Most importantly, the self-organization approach may yield a more appropriate or optimal three-dimensional arrangement of cardiac cells. Histological studies of self-organized EHT have demonstrated an organized and directional cardiomyocyte arrangement that is very similar to that of native myocardium (Baar et al. 2005; Sondergaard et al. 2010). While there has been the demonstration of in vivo vascularization of implanted constructs, there have been no attempts to prevascularize self-organizing EHT in vitro (Birla et al. 2005). Therefore, the purpose of this study was two-fold: first, to attempt and characterize prevascularization of self-organizing EHT and second, to determine the effects of prevascularization on the in vitro contractile function of self-organizing EHT.

Materials and methods Animals All procedures involving laboratory animals were done in accordance with protocols approved by the UC Davis Institutional Animal Care and Use Committee.

isolated cells were suspended in a plating medium (PM) consisting in 64 % (v/v) M199, 20 % F12K, 7 % fetal bovine serum, 7 % calf bovine serum, 1 % penicillinstreptomycin (all Invitrogen), 40 ng/ml hydrocortisone, 25 μg/ml 6-aminocaproic acid and 30 μg/ml ascorbic acid (all Sigma-Aldrich). EHT fibers were plated on polydimethylsiloxane (PDMS) elastomer (Dow Chemical, Pittsburgh, CA, USA)-coated 35-mm culture dishes with segments of 00 silk sutures (Ethicon, Somerville, NJ, USA) serving as anchors for the EHT fiber that were pinned 8 mm apart in the center of the plate. Thrombin (10U/ml) and fibrinogen (20 mg/ml; both Sigma-Aldrich) diluted in PM were mixed and layered on top of the PDMS to form a fibrin gel layer. Isolated cardiac cells re-suspended in PM were layered on top of the fibrin gel at 1×106 cells/plate with or without the addition of HUVECs (1×105 cells/plate, n06 per group). Cardiac cells self-organized into EHT fibers by days 7–10. Force measurements EHT contractile function was measured using a custom-built optical force transducer as previously described immediately after EHT fiber formation (Mathews et al. 2012). Briefly, one end of the EHT fiber was unpinned from the PDMS coating and attached to the force transducer load element. Preload was applied by increasing the pinned length of the fiber in a stepwise manner from 0 mm and up to 5 mm in 0.5-mm increments. Data were recorded using LabVIEW (National Instruments, Austin, TX, USA) for up to 30 s at each displacement and analyzed using a custom-written Python (v.2.6.6) script utilizing the Numpy (v.1.5.0b1), SciPy (v.0.8.0) and Matplotlib (v.1.0.1rc1) libraries as described (Mathews et al. 2012). Force measurements were normalized to the average cross-sectional area of the EHT fiber. High resolution images of all tested EHT fibers were taken digitally and were calibrated to the culture plate grid and used to estimate the fiber width averaged over three different positions along the construct; the average cross-sectional area for each construct was calculated assuming a cylindrical shape.

Cell culture RT-PCR Human umbilical vein endothelial cells were obtained from human umbilical cords, expanded and maintained as previously described under an approved UC Davis IRB protocol (Olson et al. 2011). Isolation of and plating of neonatal rat cardiac cells Cells were isolated from neonatal Sprague–Dawley rats (Charles River Laboratories, Wilmington, MA, USA) 2– 4 days of age as described (Sondergaard et al. 2010). The

Total mRNA was isolated from fibers and real-time PCR was done using SYBR Green PCR Master Mix according to the manufacturer’s recommendations (Applied Biosystems) and the following primers: Hypoxia inducible factor 1— forward: TGCTTGGTGCTGATTTGTGA, reverse: GGTCAGATGATCAGAGTCCA; GAPDH—forward: T C C T G C A C C A C C A A C T G C T TA G , r e v e r s e : AGTGGCAGTGATGGCATGGACT. The ΔΔCT method was used to quantify the fold difference in gene expression

Cell Tissue Res

using GAPDH as an internal control. Results were pooled from experimental replicates when appropriate. Immunofluorescence staining and imaging On day 4 after cell plating, forming EHT fibers were imaged on a bright field microscope at the periphery to identify sprout formation emanating from the edges of the forming constructs. Four pictures were taken at randomly selected positions at both ends and the lateral margins of the midpoint of the fiber. Pictures were scored with respect to the presence (score01) or absence (score00) of discernible sprouts and the total sum of the scores (between 0 and 4) for fibers with (n 06) or without HUVECs (n06) were compared. Following force measurements, EHT fibers were processed for frozen sections and stained with biotinylated mouse anti human CD31 (AbD Serotec, Raleigh, NC, USA) and mouse anti α-Sarcomeric Actin (Invitrogen) followed by the appropriate secondary Alexa488 conjugated antibody or Alexa594 conjugated Streptavidin, respectively (both Invitrogen). Native neonatal rat heart tissue processed in parallel was stained with a DyLight 594 conjugated GSL I-B4 lectin specific for rodent endothelial cells (Vector Laboratories, Burlingame, CA, USA). Slides were mounted with DAPI containing mounting media (Invitrogen). Prevascularization was estimated by identifying clusters of nucleated CD31+ cells in randomly selected fields of view at ×200–400 magnification and averaged over eight randomly selected fields of view. Statistics Mean values ± SD are given. Sprout density at the margin between groups was compared by the Wilcoxon Rank Sum Test. RT-PCR data were compared using Student’s t test. Mean contractile frequency, force generation and rate of force generation during systole and diastole, were compared by Mann–Whitney test. A value of p≤0.05 was taken to be significant.

Results EHT fibers containing HUVECs had a propensity to develop extensive tubule-like sprouts at the margin of the forming construct at day 4 post-plating, although the number of sprouting areas were not significantly different by Wilcoxon rank sum test (p