Extracellular Matrix-Mediated Maturation of Human Pluripotent Stem Cell-Derived Cardiac Monolayer Structure and Electrophysiological Function

Abstract

Background: Human pluripotent stem cell-derived cardiomyocytes (hPSC-CMs) monolayers generated to date display an immature embryonic-like functional and structural phenotype that limits their utility for research and cardiac regeneration. In particular, the electrophysiological function of hPSC-CM monolayers and bioengineered constructs used to date are characterized by slow electric impulse propagation velocity and immature action potential profiles.

Methods and results: Here, we have identified an optimal extracellular matrix for significant electrophysiological and structural maturation of hPSC-CM monolayers. hPSC-CM plated in the optimal extracellular matrix combination have impulse propagation velocities ≈2× faster than previously reported (43.6±7.0 cm/s; n=9) and have mature cardiomyocyte action potential profiles, including hyperpolarized diastolic potential and rapid action potential upstroke velocity (146.5±17.7 V/s; n=5 monolayers). In addition, the optimal extracellular matrix promoted hypertrophic growth of cardiomyocytes and the expression of key mature sarcolemmal (SCN5A, Kir2.1, and connexin43) and myofilament markers (cardiac troponin I). The maturation process reported here relies on activation of integrin signaling pathways: neutralization of β1 integrin receptors via blocking antibodies and pharmacological blockade of focal adhesion kinase activation prevented structural maturation.

Conclusions: Maturation of human stem cell-derived cardiomyocyte monolayers is achieved in a 1-week period by plating cardiomyocytes on PDMS (polydimethylsiloxane) coverslips rather than on conventional 2-dimensional cell culture formats, such as glass coverslips or plastic dishes. Activation of integrin signaling and focal adhesion kinase is essential for significant maturation of human cardiac monolayers.

Keywords: cardiac electrophysiology; cell culture techniques; hardness; induced pluripotent stem cells; integrin alpha5beta1; myocytes, cardiac; troponin I.

Figure 1

Electrical wave propagation in mature hiPSC-CM monolayers. A. Left, optical activation map of spontaneously initiated electrical wave propagation in an iCell™ cardiomyocyte monolayer cultured on PDMS+matrigel. Right, single pixel signals of optical action potentials recorded from the pacemaker site and a more distal site in the monolayer. We used a CCD camera and the voltage sensitive dye FluoVolt™. B. Action potential impulse propagation velocity slowed as pacing frequency increased. The conduction velocities were: 0.7Hz= 42.4±2.3cm s⁻¹, 1Hz=36.4±1.5cm s⁻¹, 1.2Hz=33.8±0.7cm s⁻¹, 1.5Hz=29.2±0.8cm s⁻¹, 1.8Hz=27.0±0.71cm s⁻¹, 2 Hz=25.9±0.9cm s⁻¹, and 2.5Hz=23.3±0.87cms⁻¹. C. Action potential duration calculated at 80% repolarization (APD₈₀) shortened as pacing cycle length shortened. APD₈₀ values were: 0.7Hz=629.9±16.6ms, 1Hz=569.2±14.9ms, 1.2Hz=511.2±14.4ms, 1.5Hz=452.7±9.8ms, 1.8Hz=398.1±6.9ms, 2.0Hz=363.2±3.3ms, and 2.5Hz=293.3±7.3ms. n=6 monolayers for 0.7–1.8Hz and n=4 for 2.0 and 2.5Hz. Inset shows representative action potential recordings at different frequencies. All data are presented as mean ± SEM.

Figure 2

Effects of ECM on hiPSC-CM Monolayer Impulse Propagation. A. Four different ECM combinations were tested to determine the effects on hiPSC-CM monolayer structure and function. B. Activation maps of calcium impulse propagation in the different plating conditions. Each color represents a different activation time with time zero appearing in yellow. C. Quantification of impulse propagation. * denotes statistical difference where P<0.0001, † denotes statistical difference where P=0.003 analyzed by one way ANOVA, Bonferroni’s multiple comparisons test, n=4–14 monolayers per group. D. Quantification of impulse propagation during electrical stimulation at 1Hz. ‡denotes difference where P=0.0003, §denotes difference where P=0.01 analyzed by one way ANOVA, All data are presented as mean ± SEM.

Figure 3

Mature hiPSC-CM action potential and sodium channel characteristics. A. Representative action potential recordings from monolayers plated on fibronectin on glass (left) and matrigel on PDMS (right). Middle panel of A shows a faster time scale of the AP upstroke. Bottom panel of A shows the first derivative of the AP upstroke (dV/dt). B–E. AP parameters demonstrate significant electrophysiological maturation of monolayers plated on matrigel on PDMS (red). * denotes significant difference by Student’s t-test, P<0.05, N=5 monolayers −6 monolayers with n=24 or n=37 individual cellular recordings. F. Representative I_Na recordings of hiPSC-CMs cultured on fibronectin on glass (black, n=12) and cardiomyocytes cultured on matrigel on PDMS (red, n=19). G. Current-Voltage (I–V) relationship of sodium current in each condition shows elevated I_Na in cardiomyocytes cultured on matrigel coated PDMS. Comparisons made by Student’s t-test, P values as indicated. H. RT-PCR analysis of SCN5a expression. *P<0.05, Student’s t-test. All data are presented as mean ± SEM.

Figure 4

Potassium current density (I_K1) in hiPSC-CM single cells. A. Representative I_K1 recordings in single hiPSC-CM cultured on fibronectin on glass (n=7, black, top) and hiPSC-CM cultured on matrigel on PDMS (n=5, red, bottom). B. I–V relationship of I_K1 in each condition shows significantly elevated current density in hiPSC-CM cultured on matrigel on PDMS (red). Comparison by student’s t-test, p value as indicated. C. Western blot probing for Kir2.1 demonstrates expression only in hiPSC-CMs cultured on matrigel on PDMS (lane 1 in each blot, 2 individual monolayers for each condition). D. intracellular calcium flux measurements in hiPSC CMs cultured on rigid plastic bottom dishes (matrigel coated) show very significant impact of E4031 blockade on spontaneous beating frequency and calcium transient duration 80 (CaTD₈₀). Quantification shows greater effect of E4031 on the beat frequency and CaTD₈₀ in immature iPSC CMs cultured on rigid plastic bottom dishes compared to PDMS bottom dishes. Unpaired t-test, for effect on beat frequency: *P=0.000000001, †P=0.01 and for effect on CaTD₈₀ ‡P=0.0000001, §P=0.0003 data expressed as mean±SEM.

Figure 5

ECM Effect on Cx43 Expression and cardiomyocyte size. A. Immunostaining of α-actinin (green) and Cx43 (red). DAPI (blue) marks nuclei. Monolayers on matrigel coated PDMS exhibit the greatest amount of Cx43 at the cell-cell borders. B. Western blotting for Cx43 and total myosin confirms the immunofluorescence results of panel A. Quantification of total Cx43 protein expression shows elevated expression in hiPSC-CM cultured on matrigel on PDMS (n=5 monolayers in each condition). *Student’s t-test, P<0.05. All data are presented as mean ± SEM. C&D. hiPSC-CM (iCell™) cell size was determined 5 days post thaw by immunoflurescent staining for N-cadherin. There was no difference in hiPSC-CM size between the fibronectin+glass group and the matrigel+glass group (938.7±61.2μm², n=79 vs. 917.8±64.2μm², n=66). Myocytes plated on fibronectin+PDMS were larger (1403.4±66.9μm², n=132) and myocytes plated on matrigel+PDMS were even larger (2130.3±99.9μm², n=130) than those plated on rigid glass coverslips. One way ANOVA, Bonferroni’s multiple comparison test; †P<0.0001 and ‡P<0.001

Figure 6

Molecular markers of maturation. A. Flow cytometry analysis for quantification of the population of binucleated iCell cardiomyocytes. B. The proportion of binucleated cells was significantly greater in the PDMS group compared to matrigel coated glass (35.24±3.0% vs. 21.68±1.08%, n=5 per group; *denotes significant difference, t-test, P=0.003). Furthermore, the incidence of Ki67 positive binucleated cells was less in the PDMS group (13.4±1.1% vs. 20.5±1.1%, n=5 per group, †denotes significant difference, t-test, P=0.002. C. (top panel) immunostaining for Ki67 (red) and α-actinin (green) shows decreased proliferative activity in BJ-hiPSC-CMs cultured on PDMS compared to glass coverslips (0.87±0.29 CM/60X field compared to 6.2±0.90 CM/60Xfield; n=8 and n=10. ‡P=0.0001). C, (Bottom panel), immunostaining for sarcomeric actin (red) and N-cadherin (green) shows hypertrophy and elongation of BJ-iPSC-CMs cultured on PDMS compared to glass coverslips (cell area=3,678.59±171.6μm² compared to 2,071.44±116.7μm²;n=84 and n=83. §P=0.000003). D. Western Blotting for cTnI protein expression. On glass coverslips the cTnI/GAPDH ratio=0.39±0.11au and on PDMS coverslips the cTnI/GAPDH ratio=0.6464±0.005au, ║denotes significant difference, P<0.05, unpaired t-test. All data mean ± SEM.

Figure 7

Integrin signalling via Focal Adhesion Kinase promotes maturation of hiPSC-CM monolayers. A. RT-PCR analysis indicates elevated expression of ITGA5 and ITGB1 integrin receptor genes in mature monolayers (red). RT-PCR performed in triplicate for 5 individual monolayers for each group. B. PTK2 gene expression is elevated in mature monolayers (red). C. β1 integrin localization (red) in hiPSC-CM monolayers. More extensive receptor expression is apparent on PDMS. D. Nab (Neutralizing antibody) again β1 integrin receptors blocks FAK activation and cTnI expression in iCell CMs. *significant difference, t-test, P<0.05; †t-test, P=0.02; n=3 monolayers per group. E. Pharmacological inhibition of FAK activity using FAK inhibitor-14 prevents cTnI protein expression. F. FAK inhibition prevents PDMS induced hypertrophic growth of hiPSC-CMs. Cell areas: PDMS control=4,410.8±217.3μm²; 10μmol L⁻¹ =2,547.7±104.7 μm²; 100μmol L⁻¹ =1,057.1±59.0 μm²). ‡denotes significant difference from GLASS (see supplemental figure 7), §denotes significant difference within PDMS group compared to control media, One way ANOVA, P<0.0001.

Figure 8

Integrin activation promotes Nav1.5 sodium channel expression. A. Treatment of hiPSC-CM monolayers chronically with a mouse monoclonal antibody for the α5β1 integrin receptor heterodimer (fibronectin receptor) causes increase of total FAK expression and activation (B). C. Sodium channel (Nav1.5) expression is also induced by integrin activation via α5β1 integrin receptor antibody treatment. Data is expressed as Whiskers plots (mean, maximum and minimum values are shown), ANOVA was used to test for significance with *P=0.016 (A), †P=0.001 (B), ‡P=0.003 (panel C) to indicate difference.