Hemodynamic Stimulation Using the Biomimetic Cardiac Tissue Model (BCTM) Enhances Maturation of Human Induced Pluripotent Stem Cell Derived Cardiomyocytes

Abstract

Human induced pluripotent stem cell (hiPSC) derived cardiomyocytes (hiPSC-CMs) hold great promise for cardiovascular disease modeling and regenerative medicine. However, hiPSC-CMs are both structurally and functionally immature primarily due to their differentiation into cardiomyocytes occurring under static culture which only reproduces biomolecular cues and ignores the dynamic hemodynamic cues that shape early and late heart development during cardiogenesis. In order to evaluate the effects of hemodynamic stimuli on hiPSC-CM maturation we used the Biomimetic Cardiac Tissue Model (BCTM), to reproduce the hemodynamics and pressure-volume changes associated with heart development. Following 7 days of gradually increasing stimulation we show that hemodynamic loading results in (a) enhanced alignment of the cells and extracellular matrix (ECM), (b) significant increases in genes associated with physiological hypertrophy, (c) noticeable changes in sarcomeric organization and potential changes to cellular metabolism and (d) a significant increase in fractional shortening suggestive of a positive force frequency response (FFR). These findings suggest that culture of hiPSC-CMs under conditions that accurately reproduce hemodynamic cues results in structural organization and molecular signaling consistent with organ growth and functional maturation.

2. Introduction

Cardiovascular disease is the number one cause of death in the United States and throughout the world accounting for about 31% of all mortalities.(Benjamin et al. 2018, Mozaffarian et al. 2016, 2017) High mortality rates associated with cardiovascular disease can be attributed to the limited intrinsic regenerative capacity of cardiomyocytes that make up the heart. Damage and death of cardiomyocytes following events like myocardial infarctions result in the initiation of an inflammatory cascade of events that lead to cardiac tissue remodeling via cardiac fibroblast differentiation into myofibroblasts and the infarcted areas being replaced with nonfunctional fibrotic scar tissue.(Ma et al. 2014, Turner and Porter 2013) The resulting scar tissue does not contribute to contractile function and results in long-term complications that can result in sudden cardiac death, dilated cardiomyopathy, arrhythmias, and heart failure.(Louzao-Martinez et al. 2016, Dobaczewski, Chen, and Frangogiannis 2011, Khan and Sheppard 2006, Pauschinger et al. 1999)

In vitro studies play an important role in modeling of diseases and development of drugs and therapeutics. The lack of appropriate human cell types to construct in vitro models of cardiac tissue has limited the relevancy of years of research into mechanisms and potential therapies. A vast majority of these studies used neonatal or embryonic cardiomyocytes from animal and avian sources.(Miura and Yelon 2011, Warkman and Krieg 2007, Martinsen 2005) The use of adult cardiomyocytes is limited to acute studies due to the inability to maintain adult cardiomyocytes in in vitro culture without dedifferentiation.(Zhang et al. 2010, Bénardeau et al. 1997) The discovery of hiPSCs and advent of protocols to efficiently differentiate hiPSCs into hiPSC-CMs represents a major breakthrough in cardiovascular research that for the first time provided a human source of cardiomyocytes without the ethical concerns of embryonic stem cells. However, the relevancy of hiPSC-CMs is also limited as these cells represent an immature state that more closely resembles fetal or embryonic cardiomyocytes that are structurally and functionally different from human adult cardiomyocytes.(Bellin et al. 2012, Matsa, Burridge, and Wu 2014, Feric and Radisic 2016, Martins, Vunjak-Novakovic, and Reis 2014, Iglesias-Garcia, Pelacho, and Prosper 2013, Nakamura, Hirano, and Wu 2013)

Cardiogenesis and cardiomyocyte maturation involve a highly orchestrated series of events that take place over an extended period of time and continue even after birth. The heart begins to beat and pump blood early in embryogenesis when it is still a linear tube with no formed valves or chambers.(Keller, Hu, and Clark 1990, Clark and Hu 1990) This corresponds with cardiac differentiation of mesodermal precursors into immature cardiomyocytes. These immature cardiomyocytes first develop spontaneous action potentials and gradually begin contractions in a random and disorganized fashion which is sufficient to induce peristaltic blood flow within the linear heart tube. Gradually these contractions become stronger and more organized resulting in small changes in pressure and stretch as blood flows through the linear heart tube. The various hemodynamic parameters like ventricular blood pressure, stroke volume, and heart rate increase in proportion to the weight of the embryo such that blood flow indexed for the weight of the embryo remains constant.(Hu and Clark 1989, Clark et al. 1989) During cardiogenesis, the heart is extremely sensitive to biomechanical cues such that the heart does not develop properly in the absence of blood flow and any alterations to normal hemodynamic stresses result in cardiac malformations and congenital heart defects.(Sedmera et al. 1999, Tobita and Keller 2000, Clark, Hu, and Rosenquist 1984, Rychter 1978) Therefore, during cardiogenesis, blood flow and associated hemodynamic stresses play a critical role in the formation of the heart and the transformation of naïve cardiomyocytes into fully functional adult cardiomyocytes.(Srivastava 2006)

The dynamic environment and developmental cues associated with blood flow are contrary to the static environment human iPSC-CMs are typically differentiated, in that it is devoid of any hemodynamic cues. Moreover, a majority of established protocols achieve differentiation in 2D monolayers. The 2D differentiation process under static conditions results in cells that have similar resting membrane potentials, sarcomeric proteins, ion channels, and calcium handling and storage mechanisms to adult cardiomyocytes, but they are also smaller in size, exhibit underdeveloped and relatively disorganized sarcomeres, lack t-tubules, possess large pacemaker currents, and beat spontaneously.(Claire, D., and C. 2013, Synnergren et al. 2012, O’Hara et al. 2011, Poon, Kong, and Li 2011, ILANIT et al. 2006, Mummery et al. 2012) Using a 2D biochemically driven process fails to replicate the complexity of the environment in vivo that relies on a combination of ECM cues, electrical cues, mechanical cues, and the presence and interaction with other non-cardiomyocyte cell types including endothelial cells and cardiac fibroblasts, that is required for cardiomyocyte maturation.(Clark and Hu 1990, Hirota et al. 1983, Hu and Clark 1989, Christoffels and Moorman 2009, Moorman et al. 2003, Hove et al. 2003, Bowers and Baudino 2010) Therefore, it is conceivable that culturing hiPSC-CMs while reproducing the hemodynamics of cardiogenesis may be the most efficient method of generating mature cardiomyocytes from these pluripotent stem cells. Several groups have focused on mimicking geometric constraints,(Pilarczyk et al. 2016) patterning,(Carson et al. 2016) culture in dense collagen matrices,(Roberts et al. 2016) culture in 3D fibrin gels(Zhang et al. 2013) and ontomimetic differentiation(Kerscher et al. 2016) to enable progressive maturation of hiPSCs into hiPSC-CMs. While these efforts resulted in improvements in both structural and functional maturation of hiPSC-CMs, levels of functional maturation associated with adult cardiomyocytes was not achieved. Recent work by Ronaldson-Bouchard et al., used progressively increased electromechanical conditioning to mature early stage hiPSC-CMs and showed superior structural, functional, and metabolic maturation in comparison to other available methods.(Ronaldson-Bouchard et al. 2018) While this is an impressive achievement and moves us closer towards an adult cardiomyocyte, several functional measures including conduction velocities and contractile force generation fall short of what is seen in the adult myocardium. This could possibly be due to the fact that this process does not recreate hemodynamic loading associated with gradual increases in pressure-volume loading seen as the heart transitions from a linear heart tube to a neonatal and then postnatal heart.

To address the absence of hemodynamics in current protocols we used the BCTM which is the only system available that can recreate hemodynamic loading and unloading associated with pressure-volume changes in any chamber of the heart at any stage of development. We specifically focus on reproducing blood flow associated with the early stage heart. The novelty of this work is the ablity to superimpose flow, pressure and stretch associated with critical events during early stage heart development on iPS-CMs during in-vitro differentiation in a manner where the application of these stresses closely mimics what occurs in-vivo. To further enhance physiological significance, we coculture early stage hiPSC-CMs (immediately following mesodermal induction and cardiac differentiation) and neonatal fibroblasts in a 3D fibrin gel to more closely mimic cardiac tissue.

3. Materials and Methods

Cell Culture Chamber Fabrication:

Cell culture chambers were fabricated using standard soft-lithography using (poly)dimethyl siloxane (PDMS) (Sylgard 184, Dow Corning, Midland, MI). The chamber was created using a two-step process of bonding a thin PDMS membrane to the bottom of a square PDMS block with an ellipse shaped hole punched through the center measuring 15 mm by 30 mm. The PDMS block and thin membrane used a ratio of base to crosslinker of 10:1. The thin membrane was formed by spinning 10:1 PDMS to ~80 μm on a silicon wafer within a tabletop centrifuge retrofitted to hold and center the silicon wafers. Both components were cured overnight at 60 °C. Once cured, both pieces were cleaned and bonded using oxygen plasma (Harrick plasma systems, Ithaca, NY) optimized to create a permanent bond using 700 mTorr pressure, 30 W, and 20 s exposure. Next, two 10:1 PDMS post measuring 6.5 mm tall and 2 mm in diameter are placed on top of the thin membrane and along the ellipse’s long axis at a distance of 12 mm apart. The post are bonded by placing the bonded PDMS chambers on top of a hot plate set to 150 °C, then the post are dipped into liquid 10:1 PDMS and placed through a removable guide that allows for consistent post placement. After autoclaving the chambers are filled with 2% hot liquid agarose and allowed to cool and solidify. Circles around each post and a center lane connecting these circles are punched out leaving a dumbbell shape within the chamber. Finally the chamber is ready for seeding with fibrin encapsulated cells; the final product, once the gel has been added and the agarose removed can be seen in Figure 1B.

Fig 1:

(A) BCTM cell culture chamber with agarose mold prior to addition of cells within the fibrin gel. (B) BCTM cell culture chamber with the cell laden fibrin gel suspended between the two posts. This image was taken 2 days after removing the agarose mold. The gel originally takes the shape of the agarose mold but contracts over time as shown in the image. (C) The position of the posts within the cell culture chamber at the end of systole. The fiber experiences no strain at this point as the posts are at the initial resting position. (D) Position of posts within the cell culture chamber at the end of diastole. The thin membrane contacts with the insert below and deforms along its surface thus pulling the posts away from each other and applying uniaxial strain to the fiber. (E) Diagram representing how the BCTM reproduces the cardiac cycle. The BCTM imposes uniaxial strain onto the cells by using a flexible membrane that is stretched over an arched surface placed below the membrane. The yellow in this diagram represents the cell laden fibrin fiber.

hiPSC-CM Culture and Differentiation:

Human iPS cells were cultured within 6-well plates coated for 1 hour at room temperature with hESC-qualified Matrigel supplied from Corning (354277) at the manufacturer recommended concentration in order to maintain pluripotency. The cells are cultured using MTeSR1 (Stemcell Technologies 85850) and receive daily media changes. At 75–85% confluency cells are disassociated using Accutase (Stemcell Technologies 07920) and reseeded in MTeSR1 supplemented with 10 μM Y-27632, ROCK1 inhibitor (Selleckchem S1049) at a split ratio of 1:6. After 12 hours hiPS cells are switched to MTeSR1 without ROCK1 inhibitor. After 3 days hiPSCs are at a confluency of 75–85% and are switched to RPMI medium (Gibco 11875093) supplemented with B27 without insulin (Gibco A18956-01) and 10 μM WNT promoter CHIR99021 (Stemcell Technologies 72052); this is considered Day 1 of differentiation. After exactly 24 hours, cells are washed and switched to media without CHIR99021. Cells are maintained with daily media changes until Day 4 when the media is then supplemented with 5 μM WNT inhibitor IWP-2 (Stemcell Technologies 72124). Cells are cultured with IWP-2 until Day 6. On Day 8, cells are switched to RPMI medium supplemented with B27 with insulin (Gibo 17504044). At this point, the cells are referred to as hiPSC derived cardiomyocytes and have begun spontaneously beating. Medium is changed daily until Day 10, when it is refreshed thereafter every two days. Only plates that contain greater than 60% beating hiPSC-CMs are used for further study.

Human Primary Fibroblast Culture:

Neonatal primary dermal fibroblast were purchased from ATCC (PCS-201-010) and cultured in primary fibroblast medium (ATCC PCS-201-030) supplemented with Fibroblast Growth Kit-Low serum (ATCC PCS-201-041). Cells were cultured according to ATCC recommendations and were not used in any study past passage 15.

Fibrin Gel Encapsulation:

At Day 12 the hiPSC-CMs are disassociated using 0.25% Trypsin/EDTA and combined with human fibroblasts disassociated using 0.05% Trypsin/EDTA at a final density of 2.0×10⁶ fibroblast and 4.0×10⁶ hiPSC-CMs per 300 μL fibrin solution. The fibrin solution is composed of DMEM-HG supplemented with 10% FBS, 50 μg/mL L-Ascorbic acid (Sigma A5960), 1X nonessential amino acids (Gibco 11140-050), 5 mg/mL aminocaproic acid (Acros 103301000) and contains a final concentration of 2 mg/mL fibrinogen (Millipore 341576), 0.9 units thrombin (Millipore 605195) per mg fibrin (0.54 units/gel), and 2 mM CaCl₂. The fibrin encapsulated cells are then placed within a 37 °C, 5% CO₂ incubator for 30 minutes to complete the fibrin polymerization. Following polymerization and until the end of experimentation cells are maintained with DMEM-HG supplemented with 10% FBS, 50 μg/mL Ascorbic acid, 1X nonessential amino acids, and 5 mg/mL aminocaproic acid. Once within the fibrin gel, the cells are maintained through daily media changes. Cells are allowed to spread and adapt to the 3D architecture as well as recover from the trypsinization and encapsulation process for 5 days before beginning BCTM stimulation.

BCTM Stimulation:

The mechanism of action of the BCTM revolves around using dynamic pressure differences above and below the cell culture chamber membrane in order to reproduce the cardiac cycle.(Rogers, Fast, and Sethu 2016). The BCTM consists of media reservoir set to a specified height in order to create a downward filling pressure imposed onto the cell culture member to mimic a diastolic filling pressure. A programmable pneumatic pump is used to increase the pressure below the cell culture membrane in order to generate a systolic pressure. The cycling of the pneumatic pump coupled with one-way flow control values allows the actuation of the culture membrane to displace fluid in one direction with each contraction, Figure 1E. The BCTM has changed from its previous published iterations through its adaptation to the culture of 3D tissue fibers suspended between two posts as shown in Figure 1B. Another notable difference in the current design is the presence of a dome shaped insert below the cell culture chamber and aligned between the two posts that deforms the membrane as it is pushed down onto the insert during the diastole phase of the cycle. By recording the post movements during the cardiac cycle, seen in Figure 1C–D, we are able to calculate strain.

BCTM Stimulation Conditions:

In order to adapt to the mechanical stress, the cells are placed under gentle conditions and ramped up over time, Figure 2. Gradually increasing the stress was found to be necessary for hiPSC-CMs to survive under hemodynamic loading.(Rogers, Fast, and Sethu 2016) On day 0, the stimulation is set to 5 mmHg peak-systolic pressure, 1 mmHg end-diastolic pressure, and a maximum strain of 0.2%. After 24 hours, day 1, the stimulation is increased to 10 mmHg peak-systolic pressure, 3 mmHg end-diastolic pressure, and a maximum strain of 0.7%. On day 2 the stimulation is increased to 15 mmHg peak-systolic pressure, 5 mmHg end-diastolic pressure, and a maximum strain of 1.2%. Finally, on day 3 the stimulation is increased to 30 mmHg peak-systolic pressure, 10 mmHg end-diastolic pressure, and a maximum strain of 2.3%. This final condition is maintained until the end of the 7 day experiment. Throughout the entire experiment the cycle frequency is maintained at 1 Hz. A frequency of 1Hz was used to replicate a normal human resting heart rate. The pressure and stretch stimulus was detrmined to ensure cell survival as higher levels of mechanical stimulation have been found to be detrimental to cell survival. The 7 day stimulation represents the initial phase of embryonic heart development during which blood flow and pressure-volume changes are initiated. Also, 7 days is sufficient time to clearly see changes in cardiomyocyte structural and functional maturation.

Fig. 2:

Pressure-volume loops that hiPSC-CMs were subjected to during the course of the 7 day study. The systolic/diastolic ratio was maintained at 40/60% and the frequency was maintained at 1 Hz throughout the duration of the experiment.

Staining:

At the end of the experiment the fibrin encapsulated cells are removed from the BCTM and fixed using fresh 4% paraformaldehyde at room temperature for 1 hour before moving to 4 °C. Samples are then placed in OCT solution and frozen for cryosectioning of 10 micron thick slices. Sections were immunolabeled with rabbit anti-cardiac troponin T (Abcam, ab91605) and mouse monoclonal anti-actin (alpha-sarcomeric) antibody (Sigma, A2172). Nuclei are counter stained with DAPI. Images were taken using an Olympus IX83 epifluorescence microscope.

TEM Imaging:

At the end of the 7 day stimulation protocol the cell fibers were removed from the BCTM and fixed using a fixative composed of 2% Paraformaldehyde and 2.5% Glutaraldehyde in a 0.1 M sodium cacodylate solution for 1 hour at room temperature and then moved to 4 °C overnight. Samples were then taken to the UAB High Resolution Imaging Facility where they were further processed, sliced, embedded, and sectioned. Slices were imaged using a Tecnai Spirit T12 Transmission Electron Microscope.

RNA qRT-PCR:

After 7 days of stimulation, the cells are removed from the BCTM, carefully slid off of the posts, placed within a liquid nitrogen safe cryovial with 700 μL Trizol (Invitrogen), and then flash frozen and maintained in liquid nitrogen until processing. Once enough samples are collected for gene analysis they are thawed at room temperature and immediately placed within a beadmill (BeadBug D1030) containing 1.0 mm silica beads (Benchmark D1031-10) and milled for 30 seconds at speed 4000. RNA is then isolated using standard phenol-chloroform extraction methods for quantification of transcripts GADPH, GATA4, NKX2.5, MYLK, MYL2, KCNH2, CAMK2B, and MYH7. Complementary DNA for mRNAs was obtained from 2 μg total RNA in a 20 μl reaction using TaqMan Reverse Transcription Reagents (ThermoFisher Scientific) and 100 pmole of oligo(dT)15 primer. This mixture was incubated at 37 °C for 2 hours. Quantitative RT-PCR was performed with primers (Table 1) using Eppendorf realplx² Real-Time PCR system (Eppendorf). Complementary DNA synthesized from 100 ng total RNA was combined with Maxima SYBR Green qPCR Master Mix (Thermofisher Scientific) and 0.5 μM each of forward and reverse primers. Cycling conditions were as follows: 95 °C for 10 min followed by 40 cycles of amplification (95 °C denaturation for 15 sec, and 60 °C annealing-extension for 1 min). To avoid the influence of genomic contamination forward and reverse primers for each gene were located in different exons. Relative quantitative expression of gene of interest was calculated after normalizing to the housekeeping gene GAPDH.

Table 1:

Genes with corresponding forward and reverse primers used for the qRT-PCR.

Gene	Forward primer	Reverse primer
MYH7	GAGGACAAGGTCAACACCCT	CGCACCTTCTTCTCTTGCTC
MYLK	AGTT GT GGAGGGAAGT GCT G	GG AC TC C C T GATT G AC T GGT
CAMK2B	CACCGACGAGTACCAGCTCT	GTT GAT GAT C TT GGC T GC AT
GAPDH	AAGGT GAAGGT C GGAGTCAA	AAT G AAGGGGTC ATT GAT G G
G AT A4	TCTAAGACACCAGCAGCTCCTTCA	ACT G AC T G AG AAC GT C T GGG AC AC
NKX2.5	C ACC AT GGGGAGAAAAAAGATT C A	TGTTGCCCATCATTCAGAAAGTC
MYL2	CCAACTCCAACGTGTTCTCCATGT	CAT C AATTT CTT C ATTTTT C ACGTTC A
KCNH2	GGGAGC TT GGGGC C T GAC C	TCCCCTCCCCCGCCTCACAC

Fractional Shortening:

Immediately after removing the hiPSC-CM fibers from the BCTM their intrinsic (spontaneous) and paced contractions at 1.0 Hz, 1.3 Hz, 1.5 Hz, and 2.0 Hz were recorded using a video camera. A Myopacer field stimulator (Ionoptix) with adapted carbon electrodes provided field stimulation at 4.0 V/cm with 40 ms duration. The field was set to alternate polarity to avoid build up on the electrodes. The carbon electrodes were placed near the bottom of the cell culture surface and between the post and outside edge of the cell culture chamber so that the field ran along the entire length of the fiber. HBSS (Gibco 14025-092) warmed to 37 °C was used during the recordings. Fractional shortening was estimated using video analysis and Lagrangian strain calculations for finitely small displacement as previously described by Bonet et al(Bonet and Wood 1997) and previously used by our group.(Nguyen et al. 2009, Nguyen et al. 2015) Briefly, if p1 and p2 are two nearby points defined on the first frame and d1 denotes the Euclidian distance between p1 and p2 at the first frame (i.e., reference frame), then dj denotes the Euclidian distance between these two points at time frame j. Then, the engineering strain S, at frame j, is defined as: S_j=(d_j-d₁)/d₁

A custom algorithm was written to estimate planar Lagrangian strains. First, the peak displacement was determined by tracking two points based on their appearance features through the image frames and calculating the displacement between the two points over the image frames. The maximum displacement was then selected as the reference frame. Given that the contractions are unidirectional due to alignment of the fibers, two pairs of points on the first (reference) frame were determined and their tracked pairs on the selected frame with maximum displacement. Lagrangian strain for each of the points was estimated using the previously mentioned equation. Finally, the mean Lagrangian strain was calculated and expressed as % shortening. To account for drift due to macroscopic movement of the fiber, the analysis was performed on a per beat basis where the measurements correspond to maximum and minimum displacements per contraction.

Statistical Analysis:

An unpaired t-test was used for the analysis of the gene expression and fractional shortening data with two-tailed significance set at p ≤ 0.05 for both experiments.

4. Results

Cardiomyocyte Sarcomeric Structure:

Immunofluorescence microscopy was performed to visualize sarcomeric organization in both stimulated and static samples. Images show that both stimulated and static samples had organized sarcomeres with striated organization of the sarcomeric actin (α-SA) and cardiac troponin T (cTnT) (Figure 3).

Fig. 3: (Left).

Static Controls and (Right) BCTM Stimulated samples with the entire fiber stained with cardiac troponin T (cTnT), alpha sarcomeric actin (α-SA), and DAPI with magnified images of α-SA (bottom) cTnT (middle), and combined image (top row).

Cardiomyocyte Ultrastructural Morphology:

TEM was performed to evaluate changes in intracellular and extracellular structure and organization. Images clearly show that the ECM associated with the stimulated samples is more aligned and composed of longer, thicker fibrils (Figure 4, top, right) in comparison to the shorter, thinner, and more disorganized fibrils in static controls (Figure 4, top, left). Static control cells were inundated with lipid deposits and unorganized regions of sarcomeric proteins, denoted with a black and white arrow respectively (Figure 4, middle, left) whereas the stimulated samples did not show lipid deposits but had more highly organized sarcomeres (Figure 4, middle, right). Highly magnified image of sarcomeres shows that Z-disks, A-bands, and I-bands are clearly visible in both the static controls and stimulated samples (Figure 4, bottom) with evidence of the beginnings of the formation of an M-line in the stimulated sample (black arrow).

Fig. 4:

Static controls (Left) and stimulated samples (Right). (Top) Visualization of ECM. (Middle) Representative images of sarcomeres and associated structures within cardiomyocytes. Arrows indicate lipid deposits (black) and an example of unorganized sarcomeres (white) both of which were common throughout the static controls. (Bottom) High magnification images of representative sarcomeres from the static controls and stimulated samples. The black arrow on the stimulated sample marks a white area along the middle of the sarcomere that was found consistently in the stimulated sarcomeres and what could be the formation of the M-line.

Expression of Genes involved in Structural and Functional Maturation:

Static and stimulated samples were profiled for changes in genes associated with cardiomyocyte hypertrophy and cardiac specific structural maturation (Figure 5). Several genes associated with hypertrophic growth were upregulated in the stimulated samples. Our results confirm that GATA4, Myosin regulatory light chain 2 (MYL2), Calcium/calmodulin-dependent protein kinase type II beta chain (CAMK2B), myosin heavy chain beta (MYH7) were significantly upregulated in stimulated samples vs the static controls. We also found no statistically significant changes in NKX2.5, Myosin light chain kinase (MYLK), and potassium voltage-gated channel subfamily H member 2 (KCNH2). Although a mixed population of fibroblast and cardiomyocytes make up the cell samples used for gene analysis, the genes chosen are highly cardiac specific.

Fig. 5:

Static controls vs. BCTM stimulated samples normalized to GAPDH expression levels. Statistically significant upregulation was observed in genes GAT4, MYL2, CAM2KB, and MYH7 in stimulated samples (N=5) vs. static samples (N=4).

Functional Maturation of Cardiomyocytes:

Video analysis of the fractional shortening shows that stimulated samples achieved greater fractional shortening than the samples cultured under static conditions. It can also be noted that the static samples demonstrated a decrease in fractional shortening at progressively higher pacing frequencies while the fractional shortening of stimulated samples peaked at the 1.3 Hz pacing frequency before progressively decreasing at higher pacing frequencies, Figure 6. Importantly, significantly higher fractional shortening measured in the stimulated samples (N=6) in comparison to the static samples (N=5) at frequencies: 1.3 Hz, 1.5 Hz, and 2.0 Hz.

Fig. 6:

Data collected from videos made of the fibers undergoing electrical stimulation at frequencies: 1 Hz, 1.3 Hz, 1.5 Hz, and 2.0 Hz. Videos were also made of the intrinsic (spontaneous) fractional shortening of each fiber (N=4). The stimulated samples showed significantly higher fractional shortening at 1.3, 1.5, and 2.0 Hz. The intrinsic contractions were not significantly different.

5. Discussion/Conclusion

Given the importance of mechanical loading during embryonic heart development, we sought to determine if hemodynamic stimulation could potentially impact differentiation of hiPSC-CMs into a more mature phenotype. This involved the culture of early stage (immediately following cardiac differentiation of mesodermal precursors) hiPSC-CMs under peristaltic flow and pressure, stretch which is gradually increased to 20 mmHg and 3% stretch within the BCTM. The results provide preliminary validation that recreating mechanical loading during early cardiac differentiation induces changes associated with the stage of embryonic heart development linked to an increase in chamber size and functional maturation. Immunofluorescence microscopy provides clear evidence of physiological hypertrophy and enhanced organization of hiPSC-CMs subject to stimulation within the BCTM (Figure 3). Visualization of cTnT and α-SA shows significant improvements in the sarcomeric alignment and localization of actin and troponin in clear alternating patterns in the stimulated samples but not in the static controls. It also appears that each sarcomere is wider and longer than the sarcomeres in the static controls, although this was not quantified.

Examination of ultrastructural morphology using TEM revealed important distinctions between the static controls and stimulated samples (Figure 4). It appears that mechanical stimulation greatly enhanced the alignment of the ECM and enabled the formation of longer and thicker ECM fibrils. This is potentially beneficial to cardiomyocyte maturation by providing alignment cues and coordinating force transmission along the length of the engineered muscle fiber. Stimulation also increased the overall alignment of sarcomeres and prevented disorganized regions that can be observed in the static controls. High magnification images show organization of sarcomeres including Z-disks, A-bands, and I-bands in both static controls and stimulated samples. Closer examination shows the beginnings of the formation of M-lines in the stimulated samples. Surprisingly, we found lipid deposits within hiPSC-CMs in the static controls. The absence of lipid deposits in the stimulated samples could possibly be explained by a switch to fatty acid metabolism due to the higher energetic demand of stimulated cells to achieve physiological hypertrophy and growth. However, at this stage of development (embryonic/neonatal) glycolysis is still the primary source of energy associated with proliferation and organ growth. Additional studies are necessary to determine the significance of lipid accumulation in the static controls and evaluate possible differences in metabolic activity between the static controls and stimulated samples. TEM is commonly used to evaluate ultrastructure morphology of cardiomyocytes. Adult cardiomyocytes are defined by organized bands and lines (A-band, M-line, Z-line, I-bands) that make up the sarcomere and are flanked by sarcoplasmic reticulum and a high number of mitochondria that typically run along the length of the sarcomere. The presence of each band along with increased alignment places our stimulated samples closer to what would be expected from fetal cardiac tissue (Ronaldson-Bouchard et al. 2018), which confirms early and progressive maturation, and supports our hypothesis.

Gene expression studies provide further evidence of physiological hypertrophy in the BCTM stimulated samples (Figure 5). The upregulation of GATA4 is significant due to its presence in developing and adult cardiomyocytes where it has been linked to hypertrophic growth of both immature and mature cardiomyocytes. MYL2 and MYH7 are both part of the myosin complex involved in binding to actin filaments and are integral to proper contractile function. Upregulation of these two genes implies the creation of new sarcomeric structures and possibly explains the increase in fractional shortening observed in the stimulated samples. Human CAMK2B is a calcium handling protein that was upregulated in the stimulated samples and suggests improved calcium handling and hypertrophy in stimulated samples. The lack of a change in KCNH2 could be due to the stimulation period being too short or that electrical stimulation may also be required for upregulation of voltage regulated channels, as seen in other studies using electrical stimulation for maturation.(Eng et al. 2016)

Finally, fractional shortening measurements suggest significant improvements in percent shortening of stimulated samples in comparison to static controls (Figure 6). Interestingly, the fractional shortening data suggests that stimulation within the BCTM induces a positive force frequency response in the stimulated samples, increase in contractile force with increasing frequency. The stimulated samples demonstrated a peak fractional shortening at 1.3 Hz which gradually decreased with increasing frequencies creating a bell curve indicative of the force frequency response measured in mature cardiac tissue. A similiar bell curve peaking at 1.5 Hz is present in normal healthy cardiac tissue. It is possible that this is an example of a positive force frequency response, but it should be noted that the increases in fractional shortening between pacing frequencies in the stimulated samples were not significantly different. However, it is clear that the fractional shortening was significantly increased in the stimulated samples when compared to static culture at most pacing frequencies.

In summary, the BCTM was used to successfully recreate hemodynamic loading associated with early heart development. 3D engineered cardiac tissue fibers constructed using early stage hiPSCCMs/neonatal fibroblast co-cultures stimulated for 7 days within the BCTM show noticeable and quantifiable changes in structure, alignment, organization, hypertrophic gene expression, and function when compared to engineered fibers maintained under static culture. These results suggest that the BCTM can be used to recreate hemodynamic cues associated with the embryonic development and can possibly be used in conjunction with electrical stimulation to mimic conditions associated with the postnatal (adult) heart.

Abbreviations used in this paper:

hiPSC-CM

Human induced pluripotent stem cell derived cardiomyocytes

BCTM

Biomimetic Cardiac Tissue Model

ECM

Extracellular Matrix

FFR

Force Frequency Response

PDMS

(Poly)dimethyl siloxane

MYL2

Myosin regulatory light chain 2

CAMK2B

Calcium/calmodulin-dependent protein kinase type II beta chain

MYH7

Myosin heavy chain beta

MYLK

Myosin light chain kinase

KCNH2

Potassium voltage-gated channel subfamily H member 2