The present protocol generates mesh-shaped engineered cardiac tissues containing cardiovascular cells derived from human induced pluripotent stem cells to allow the investigation of cell implantation therapy for heart diseases.
The current protocol describes methods to generate scalable, mesh-shaped engineered cardiac tissues (ECTs) composed of cardiovascular cells derived from human induced pluripotent stem cells (hiPSCs), which are developed towards the goal of clinical use. HiPSC-derived cardiomyocytes, endothelial cells, and vascular mural cells are mixed with gel matrix and then poured into a polydimethylsiloxane (PDMS) tissue mold with rectangular internal staggered posts. By culture day 14 ECTs mature into a 1.5 cm x 1.5 cm mesh structure with 0.5 mm diameter myofiber bundles. Cardiomyocytes align to the long-axis of each bundle and spontaneously beat synchronously. This approach can be scaled up to a larger (3.0 cm x 3.0 cm) mesh ECT while preserving construct maturation and function. Thus, mesh-shaped ECTs generated from hiPSC-derived cardiac cells may be feasible for cardiac regeneration paradigms.
Numerous preclinical studies and clinical trials have confirmed the efficiency of cell-based cardiac regenerative therapies for failing hearts1,2,3. Among various cell types, human induced pluripotent stem cells (hiPSCs) are promising cell sources by virtue of their proliferative ability, potential to generate various cardiovascular lineages4,5, and allogenicity. In addition, tissue engineering technologies have made it possible to transfer millions of cells onto a damaged heart5,6,7,8.
Previously, we reported the generation of three-dimensional (3D) linear engineered cardiac tissues (ECTs) from hiPSC-derived cardiovascular lineages using a commercially available culture system for 3D bioartificial tissues5,7. We found that the coexistence of vascular endothelial cells and mural cells with cardiomyocytes within the ECT facilitated structural and electrophysiological tissue maturation. Furthermore, we validated the therapeutic potential of implanted hiPSC-ECTs in an immune tolerant rat myocardial infarction model to improve cardiac function, regenerate myocardium, and enhance angiogenesis5. However, the linear ECTs constructed by this method were 1 mm by 10 mm cylinders and therefore not suitable for the implantation in preclinical studies with larger animals or clinical use.
Based on the successful use of tissue molds to generate porous engineered tissue formation using rat skeletal myoblasts and cardiomyocytes9, human ESC-derived cardiomyocytes10 and mouse iPSCs11, we developed a protocol to generate scalable hiPSC-derived larger implantable tissue using polydimethylsiloxane (PDMS) molds. We evaluated a range of mold geometries to determine the most effective mold characteristics. Mesh-shaped ECTs with multiple bundles and junctions exhibited excellent characteristics in cell viability, tissue function and scalability compared to plain-sheet or linear formats that lacked pores or junctions. We implanted the mesh-shaped ECT in a rat myocardial infarction model and confirmed its therapeutic effects similar to implanted cylindrical ECTs12. Here we describe the protocol to generate a hiPSC-derived mesh-shaped ECT.
1. Maintenance of hiPSCs and cardiovascular differentiation
2. Cell harvest and lineage analysis on differentiation day 13‒15
3. Fabrication of PDMS tissue mold
4. ECT construction
Figure 1A,B shows the schematics of CM+EC and MC protocol. After inducing CMs and ECs from CM+EC protocol and MCs from MC protocol, the cells are mixed adjusting final MC concentrations to represent 10 to 20% of total cells. The 2 cm wide tissue mold is fabricated according to the design drawing from 0.5 mm thick PDMS sheet (Figure 2A,B). Six million of CM+EC+MC cells are combined with collagen I, and matrix and poured onto the tissue mold precoated with poloxamer 407 (Figure 2C). During preliminary experiments, we fabricated tissue molds with various patterns characterized by various post lengths and spacing and confirmed final geometries of ECTs (Figure 3). We selected the mold with 7 mm long posts with 2.5 mm wide intervals for this study.
Poloxamer 407 prevents cell adhesion to the mold and enabled the formation of characteristic mesh structure following rapid gel compaction in 14 days (Figure 4). This structure is maintained even after the ECT is released from its mold. The whole tissue is approximately 1.5 cm wide and 0.5 mm thick, and the width of each bundle in the mesh is approximately 0.5 mm in average.
It is possible to generate a 3 cm final width mesh ECT containing twenty-four million cells from a four times larger mold with the same staggered post design (Figure 5A,B). This larger mesh ECT can be removed from the mold easily and generates local active force equivalent to the smaller mesh ECT (Figure 5C).
Figure 1: Protocols to differentiate cardiovascular cells from human induced pluripotent stem cells. Schematic diagrams of protocols used to induce cardiomyocytes and vascular endothelial cells (A, CM+EC protocol) and to induce vascular mural cells (B, MC protocol).CM = cardiomyocyte; EC = endothelial cell; MC = vascular mural cell; iPSC = induced pluripotent stem cell; MG = basement membrane matrix; ActA = Activin A; Wnt3a, BMP4, Bone morphogenetic protein 4; bFGF = basic fibroblast growth factor; VEGF = vascular endothelial cell growth factor; FBS = fetal bovine serum. This figure is adapted from reference Masumoto et al.5. Please click here to view a larger version of this figure.
Figure 2: Fabrication of PDMS tissue mold and ECT construction. (A) Representative image of an ECT mold fabricated from 0.5 mm thick PDMS sheets. (B) Design of the PDMS tissue mold with 7 mm long internal posts arranged in staggered position; front view (upper panel) and side view (lower panel). (C) The schematic protocol of ECT construction from hiPSC-derived cardiovascular cells and matrix gel. Please click here to view a larger version of this figure.
Figure 3: Impact of tissue mold designs on final ECT geometries. (A) Definitions of post length (PL) and horizontal post spacing (HPS). (B) A tissue mold without rectangular posts (PL0) and formed a sheet ECT. (C‒E) Tissue molds with different PL/HPS and mesh ECTs. (F) A tissue mold with long parallel posts and formed a multiple linear ECT. Please click here to view a larger version of this figure.
Figure 4: Formation of mesh ECT following gel compaction. Representative series of images of a mesh ECT maturation from day 0 to 14 are shown. Cell/matrix is poured in the PDMS tissue mold (Day 0_0 h), then culture medium is added one hour later (Day 0_1 h). On day 1, elliptical pores are observed around loading posts. The construct showed rapid gel compaction and matured into a mesh tissue thereafter. On day 14, the tissue is released from the mold. The unloaded ECT maintains the mesh geometry. Please click here to view a larger version of this figure.
Figure 5: Representative images of a larger mesh ECT intended for large animal preclinical trials and clinical trials. (A) Design of the 4 cm x 4 cm PDMS tissue mold with 7 mm long posts and (B) an image of the mold. (C) Representative image of a 3 cm x 3 cm larger mesh ECT. Please click here to view a larger version of this figure.
Supplemental Figure 1: Pathological and electrophysiological evaluation of mesh ECTs. (A) Representative image for a bundle in a mesh ECT stained with Hoechst 33342 (blue) for live cells and Ethidium Homodimer III (red) for dead cells. Scale bar: 250 μm. (B) Representative image of a three-dimensional confocal image of a bundle in a mesh ECT stained with cardiac troponin T (green). Local cardiomyocyte orientations within the bundle are visualized as small lines, where line color indicates magnitude in the circumferential (green), radial (red), and axial (blue) directions. (C) The spherical histogram displays local cardiomyocyte orientations. The volume of each ray represents the relative count for each direction and the thick red line shows the mean CM orientation. (B) and (C) are adapted from reference13 with revision. (D) Representative image of contractile force measurement. A segment was cut off at the red dotted line in a 1.5 cm x 1.5 cm mesh ECT and attached to the muscle testing system using 10-0 nylon suture. The white arrowhead indicates force transducer and the yellow arrowhead indicates high-speed length controller. (E) Representative waveforms of active stress at different pacing frequencies from 1.5 Hz to 4 Hz. Please click here to download this figure.
Supplemental Video 1: Intrinsic beating of a junction in a mesh ECT on day 14. Please click here to download this video.
Supplemental Video 2: Intrinsic beating of a bundle in a mesh ECT on day 14. Please click here to download this video.
Supplemental Video 3: Intrinsic beating of a mesh ECT on day 14 after released from the tissue mold. Please click here to download this video.
Following the completion of our investigation of a linear format, hiPSC derived ECT5, we adapted the protocol to mix hiPSC-derived CMs, ECs, and MCs to facilitate the in vitro expansion of vascular cells within ECTs and subsequent in vivo vascular coupling between ECTs and recipient myocardium.
To facilitate the generation of larger, implantable mesh ECT geometries we used thin PDMS sheets to design the 3D molds with loading posts arrayed at staggered positions. During preliminary experiments, we noted that ECTs adhered to the PDMS posts during gel compaction and so we modified our method to coat each mold with the surfactant poloxamer to prevent cell adhesion which is a critical step of the present protocol9. ECTs remain detached from the internal loading posts and the mold bottom during in vitro maturation, which facilitates gel compaction and ECT removal from the mold. Another critical step of the protocol is to evenly pour the cell-gel matrix mixture into the PDMS mold before the matrix solidifies. It is critical to accomplish the procedure within a short time. Bubbles in the cell-gel matrix mixture should also be avoided because it may cause structural vacuolation and functional disruption of the ECTs.
According to viability assays, approximately 97% of cells were alive within the day 14 constructs (Supplemental Figure 1A). Supplemental Figure 1B shows the whole mount confocal image stained with cTnT representing myofiber alignment parallel to the local bundle long axis (Supplemental Figure 1C)13. All constructs start intrinsic spontaneous beating in vitro within 72 h and continued beating throughout the duration of culture (Supplemental Videos 1‒3). We analyzed electromechanical properties of 1.5 cm x 1.5 cm ECTs using a custom isolated muscle testing system (Supplemental Figure 1D). ECTs displayed maximal pacing capture rates of 4 Hz and generated an average active stress of 0.55 mN/mm2 at 2 Hz, 5 V pacing protocol (n = 11, standard error of means = 0.063; Supplemental Figure 1E).
Although we selected 7 mm long internal loading posts with 5 mm intervals, it is possible to vary the final ECT geometry by arranging the length of loading posts and the interval between adjacent posts (Figure 3). Moreover, it is possible to expand the scale of 1.5 cm x 1.5 cm ECTs to larger formats such as 3 cm x 3 cm large mesh ECT. According to force measurement, 3 cm x 3 cm ECTs showed electromechanical properties similar to 1.5 cm x 1.5 cm ECTs12. The flexibility of the tissue shapes and the scalability with preserved tissue function is significances of the present protocol with respect to already existing methods.
One limitation of the present protocol is that the PDMS molds are hand-assembled from PDMS sheets. Although the construction of millimeter-unit molds would be feasible by hand-assembly, molds from photolithography or casting from master molds would be more suitable for expanded and stable ECT generation.
The authors have nothing to disclose.
This work was financially supported by the Kosair Charities Pediatric Heart Research Program at the University of Louisville and the Organoid Project at the RIKEN Center for Biosystems Dynamics Research. HiPSCs used in our published protocols were provided by the Center for iPS Cell Research and Application, Kyoto University, Kyoto, Japan.
Materials | |||
Cell Culture Dishes 100×20 mm style | Falcon/ Thomas scientific | 9380C51 | |
Multiwell Plates For Cell Culture 6well 50/CS | Falcon / Thomas scientific | 6902A01 | |
Sylgard 184 Silicone Elastomer Kit | Dow Corning | 761036 | |
Reagents | |||
Accumax | Innovative Cell Technologies | AM-105 | |
BMP4, recombinant (10µg) | R&D | RSD-314-BP-010 | |
Collagen, Type I solution from rat tail | Sigma | C3867 | |
Growth factor-reduced Matrigel | Corning | 356231 | |
Human VEGF (165) IS, premium grade | Miltenyi | 130-109-385 | |
Pluronic F-127, 0.2 µm filtered (10% Solution in Water) | Molecular Probes | P-6866 | |
Recombinant human bFGF | WAKO | 060-04543 | |
Recombinant Human/Mouse/Rat ActivinA (50µg) | R&D | 338-AC-050 | |
rh Wnt-3a (10µg) | R&D | 5036-WN | |
Versene solution | Gibco | 15040066 | |
Culture medium and supplements | |||
10x MEM | Invitrogen | 11430 | |
2 Mercaptro Ethanol | SIGMA | M6250 | |
B27 supplement minus insulin | Gibco | A1895601 | |
DMEM, high glucose | Gibco | 11965084 | |
Fetal Bovine Serum (500ml) | Any | ||
Fetal Bovine Serum (500ml) | Any | ||
L-Glutamine | Gibco | 25030081 | |
NaHCO3 | Any | ||
PBS 1x | Gibco | 10010-031 | |
Penicillin-Streptomycin (5000 U/mL) | Gibco | 15070-063 | |
RPMI1640 medium | Gibco | 21870092 | |
αMEM | Invitrogen | 11900024 | |
Flowcytometry | |||
anti-TRA-1-60, FITC, Clone: TRA-1-60, BD Biosciences | BD / Fisher | 560380 | |
anti-Troponin T, Cardiac Isoform Ab-1, Clone: 13-11, Thermo Scientific Lab Vision | Fisher | MS-295-P0 | |
BD FACS Clean Solution | BD | 340345 | |
BD FACSFlow Sheath Fluid | BD | 342003 | |
BD FACSRinse Solution | BD | 340346 | |
EDTA | Any | ||
Falcon Tube with Cell Strainer Cap (Case of 500) | Corning | 352235 | |
Fetal Bovine Serum (500ml) | Any | ||
LIVE/DEAD Fixable Aqua Dead Cell Stain Kit, for 405 nm excitation | Molecular Probes | L34957 | |
PDGFRb; anti-CD140b, R-PE, Clone: 28D4, BD Biosciences | BD / Fisher | 558821 | |
Saponin | Sigma-Aldrich | 47306-50G-F | |
VEcad-FITC; anti-CD144, FITC, Clone: 55-7H1, BD Biosciences | BD / Fisher | 560411 | |
Zenon Alexa Fluor 488 Mouse IgG1 Labeling Kit | Molecular Probes | Z25002 |