Abstract
The immaturity of human induced pluripotent stem cell-derived
cardiomyocytes (iPSC-CMs) is a major limitation for their use in drug
screening to identify pro-arrhythmogenic or cardiotoxic molecules.
Here, we demonstrate an approach that combines lipid-enriched
maturation medium with a high concentration of calcium, nanopatterning
of culture surfaces and electrostimulation to generate iPSC-CMs with
advanced electrophysiological, structural and metabolic phenotypes.
Systematic testing reveals that electrostimulation is the key driver of
enhanced mitochondrial development and metabolic maturation and
improved electrophysiological properties of iPSC-CMs. Increased calcium
concentration strongly promotes electrophysiological maturation, while
nanopatterning primarily facilitates sarcomere organisation with minor
effect on electrophysiological properties. Transcriptome analysis
reveals that activation of HMCES and TFAM targets contributes to
mitochondrial development, whereas downregulation of MAPK/PI3K and SRF
targets is associated with iPSC-CM polyploidy. These findings provide
mechanistic insights into iPSC-CM maturation, paving the way for
pharmacological responses that more closely resemble those of adult
CMs.
Subject terms: Stem-cell differentiation, Heart development, Heart stem
cells
__________________________________________________________________
The immaturity of iPSC-derived cardiomyocytes limits their use in drug
testing. This study provides a systematic evaluation of how culture
medium, calcium, nanopatterning and electrostimulation distinctly
affect their structural, metabolic, and electrophysiological
maturation.
Introduction
The discovery of human induced pluripotent stem cells (iPSCs)
represents a breakthrough for medical research and clinical
application. As an unlimited source of cardiomyocytes (CMs) with a
patient-specific genetic background, iPSC-derived CMs (iPSC-CMs) can be
employed to explore disease mechanisms and drug effects. Additionally,
they hold the potential for regenerating lost myocardium in patients
with heart failure^[58]1. Numerous studies have demonstrated the
ability of iPSC-CMs to recapitulate clinical features of inherited
cardiomyopathies^[59]2, arrhythmias^[60]3,[61]4 and cardiotoxic drug
responses^[62]5, even with patient-specific sensitivities^[63]6–[64]8.
Despite these achievements, their immaturity remains a significant
limitation. In comparison to adult CMs, iPSC-CMs exhibit considerable
differences in their morphology, gene expression patterns, metabolism
and functionality^[65]1,[66]9, which may explain their low sensitivity
to hypoxia(-reperfusion) injury^[67]10,[68]11, or the lack of features
expected from a clinical phenotype of an inherited
disease^[69]12,[70]13. The use of iPSC-CMs to predict the
pro-arrhythmic activity of drugs in the Comprehensive in vitro
Proarrhythmia Assay (CiPA) has shown good correlation with the clinical
risk of torsade de pointes or QTc prolongation. However, discrepancies
have been reported for some multichannel blockers. For example,
although verapamil has a good safety profile in the clinic, it
abolishes the beating activity of iPSC-CMs at clinically relevant
concentrations^[71]14, probably due to differences in the expression of
genes encoding ion channels, such as SCN5A (Na[V]1.5), CACNA1C
(Ca[V]1.2), KCNH2 (hERG), and KCNQ1 (K[V]7.1)^[72]15, and in calcium
handling, as well as less structural maturity (e.g., fewer
mitochondrial networks, lack of T-tubules and intercalated
discs)^[73]1,[74]9 in iPSC-CMs compared with adult CMs. Whether the
establishment of more adult-like current patterns in iPSC-CMs affects
their drug response remains elusive^[75]16.
In recent years, several approaches have been developed to improve the
maturation of iPSC-CMs. The integration of fibroblasts and/or
endothelial cells into engineered heart tissue (EHT) models using
iPSC-CMs significantly enhanced both structural and functional
development, as demonstrated in various studies^[76]17,[77]18. However,
it is worth noting that 3D-tissue generation is challenging and the
experimental throughput is lower compared to 2D cultures^[78]19. Other
approaches to enhance iPSC-CM maturation include supplementation of the
culture medium with fatty acids (FA)^[79]13,[80]20–[81]22, hormones or
small molecules^[82]10,[83]23, micro- or nanopatterning (NP) of culture
surfaces^[84]24,[85]25, and electrostimulation (ES)^[86]26–[87]30. As
these stimuli have been investigated independently, the most effective
factor for enhancing iPSC-CM maturation remains unclear. It is
uncertain whether combined approaches could produce synergistic
effects, and the underlying mechanisms driving the advanced maturation
of iPSC-CMs remain to be elucidated.
Here, we systematically examined the effects of FA-enriched maturation
medium (MM), increased calcium concentrations in the medium, NP and ES
on the electrophysiological, structural and metabolic maturation of
iPSC-CMs through comprehensive functional and molecular analyses. While
NP primarily promoted structural maturation with limited impact on
electrophysiological properties, elevated Ca^2+ concentrations in the
medium strongly influenced the electrophysiological characteristics of
iPSC-CMs. Notably, ES emerged as the key driver of enhanced
mitochondrial development and metabolic maturation and improved
electrophysiological properties. The combined application of MM with a
high calcium concentration, NP and ES led to significant changes in the
sensitivity of iPSC-CMs to cardioactive drugs, yielding pharmacological
responses that more closely resemble those of adult CMs. Our data
provide mechanistic insights into the distinct roles of these stimuli
in shaping the cellular structure, metabolism and electrophysiology of
iPSC-CMs.
Results
We used the directed differentiation protocol to generate
ventricular-like CMs from iPSCs derived from 3 healthy
individuals^[88]31. On day 15, iPSC-CMs were digested and distributed
to 4 experimental groups (Fig. [89]1a). B27 medium, routinely used for
iPSC-CM culture, served as a control. To unravel the synergistic
effects of MM, NP, and ES on the maturation of iPSC-CMs, we
systematically applied NP and ES to MM in a stepwise parallel manner
(Fig. [90]1a). MM was designed based on a published FA-supplemented
medium that enhances the metabolic maturation of iPSC-CMs^[91]13, with
some modifications (Supplementary Table [92]1). NP was used to induce
cell alignment and ES was applied to induce a beating frequency of 2 Hz
(Supplementary Movie [93]1).
Fig. 1. Study design and structural characterisation of iPSC-CMs.
[94]Fig. 1
[95]Open in a new tab
a Schematic overview of the study design. Differentiated iPSC-CMs were
digested on day 15-16 (d15-16) and randomly divided into 4 experimental
groups to investigate the effects of maturation medium (MM),
nanopatterning (NP) and electrostimulation (ES). Extensive
characterisation of the cells was performed on day 42 (d42). For some
experiments, including calcium imaging, seahorse assays and
multi-electrode array measurements, iPSC-CMs were replated into the
corresponding assay plates on d42 and allowed to recover for another 7
days. Created in BioRender. Li, W. (2025)
[96]https://BioRender.com/p85e700. b Representative morphology of
iPSC-CMs under different conditions at d42. Scale bar, 200 µm for all
four groups. c Representative immunostaining for α-actinin, cardiac
ryanodine receptor (RYR2) and Hoechst33342. d Representative
immunostaining for connexin 43 (Cx43), phalloidin and Hoechst33342.
Data (b-d) are based on 3 independent experiments using 3 different
iPSC lines. e, f Quantification of sarcomere alignment based on z-disk
orientation (e) and nuclei elongation (f). The method used to analyse
the sarcomere alignment is shown in Supplementary Fig. [97]1a. Data
were obtained from 2 independent experiments using 2 iPSC lines (n = 4
images/group/experiment). Data are presented as averages of each
experiment. Orange, red, blue and green bars represent the four
experimental groups: B27, MM, MM + NP, and MM + NP + ES, respectively.
Symbols denote iPSC lines: circles for isWT7, and squares for iWTD2.
Source data are provided as a Source Data file.
Combined approach enhances structural maturation of iPSC-CMs
We observed that NP application induced changes in cell shape and a
significant increase in the alignment of iPSC-CMs in the MM + NP and
MM + NP + ES groups compared to the B27 and MM groups (Fig. [98]1b).
Striated pattern of sarcomeres, as demonstrated by the immunostaining
against the sarcomeric protein α-actinin, was observed in iPSC-CMs
under all conditions (Fig. [99]1c). Notably, highly organised
sarcomeres with well-defined striations, which are aligned into long,
continuous myofibrils (stained with phalloidin) that run the length of
the cell, were observed only in the MM + NP and MM + NP + ES groups
(Fig. [100]1c, d). Quantitative analysis revealed that the majority of
iPSC-CMs in the MM + NP and MM + NP + ES groups showed sarcomere
patterns at an angle of 90^o to the NP direction with elongated nuclei
along the NP direction, whereas iPSC-CMs in the B27 and MM groups
revealed random orientations of sarcomeres and nuclei in all directions
(Fig. [101]1c, e, f). In addition, the sarcomere length is
significantly longer in iPSC-CMs in the MM + NP and MM + NP + ES groups
compared to those in the B27 and MM groups (Supplementary
Fig. [102]1b). Co-immunostaining for α-actinin and cardiac ryanodine
receptor (RYR2) revealed that, in B27-cultured CMs, robust RYR2
staining was observed as punctate staining with a low degree of
α-actinin/RYR2 colocalisation. In contrast, CMs from the other three
groups revealed an augmented presence of striated patterns. The
α-actinin/RYR2 colocalisation was significantly enhanced in the
MM + NP + ES group compared to the B27 and MM groups (Fig. [103]1c,
Supplementary Fig. [104]1c). In the B27 and MM groups, the gap junction
protein connexin 43 (Cx43) was partially localised to perinuclear
regions and in the cytosol, whereas Cx43 membrane localisation was
increased in CMs of the MM + NP and MM + NP + ES groups (Fig. [105]1d).
These results provide evidence for the additive effects of NP and ES to
MM on the structural maturation of iPSC-CMs.
Combined approach improves electrophysiological maturation
To evaluate the effect of MM, NP and ES on electrophysiological
properties, we performed patch clamp and multi-electrode array (MEA)
studies to investigate the action potential (AP) and field potential
(FP) parameters of single and monolayer iPSC-CMs, respectively. We
observed 43% of iPSC-CMs in the MM + NP + ES group with a
‘notch-and-dome’ AP morphology, which was not seen in the other groups
(Fig. [106]2a). The resting membrane potential (RMP) was found to be
progressively more negative in CMs from the MM (−49.7 ± 8.5 mV),
MM + NP (−58.2 ± 7.4 mV) and MM + NP + ES (−65.6 ± 8.5 mV) groups
compared to the B27 group (−44.1 ± 9.8 mV), while the maximum AP
upstroke velocity (Vmax) was gradually increased in iPSC-CMs from
4.2 ± 1.4 V/s (B27) to 5.0 ± 1.1 V/s (MM), 6.6 ± 2.5 V/s (MM + NP) and
11.0 ± 7.4 V/s (MM + NP + ES). Similarly, a gradual increase in AP
amplitude (APA) was observed in the four groups (Fig. [107]2b,
Supplementary Table [108]2). The AP duration at 90% repolarisation
(APD[90]) was significantly shorter in iPSC-CMs paced at 0.5 Hz in the
MM + NP + ES group than in the B27 control (Fig. [109]2c). As the
transient outward K^+ current (I[to]) underlies the prominent phase 1
repolarisation of cardiac APs and the ‘notch-and-dome’ AP morphology,
we measured I[to] and found a significantly higher I[to] density in
iPSC-CMs from the MM, MM + NP and MM + NP + ES groups compared to the
B27 group, particularly in MM + NP + ES; and NP itself has less effect
on I[to] when comparing the MM + NP group with the MM group
(Fig. [110]2d, e).
Fig. 2. Assessment of action potential and field potential parameters in
iPSC-CMs.
[111]Fig. 2
[112]Open in a new tab
a Spontaneous action potential (AP) traces from the four groups. A
notch event (red arrow) is only present in the MM + NP + ES group. b
Quantification of spontaneous AP metrics: resting membrane potential
(RMP), maximum upstroke velocity (Vmax), and AP amplitude (APA).
n = 21, 25, 21 and 14 (Vmax) or 17 (RMP and APA) cells from 4
independent differentiations for the four groups, respectively. c
APD[90] quantification in APs paced at 0.5 Hz (n = 13, 24, 22 and 20
cells from 4 independent differentiations for the four groups,
respectively). d Representative I[to] traces recorded in iPSC-CMs from
the four groups. e Statistical analysis of I[to] in single cells
derived from 7 (B27), 9 (MM), and 6 (MM + NP, MM + NP + ES) independent
differentiations of 3 iPSC lines. The stimulation protocol is shown as
an inset. The holding potential was set at −90 mV. To inactivate I[Na],
a 20-ms pre-pulse to −35 mV was applied. I[to] was recorded by
increasing the test potential from −40 mV to +60 mV in 10 mV
increments, with each pulse lasting 600 ms. f Representative heatmaps
of field potential (FP) propagation. The colour scale indicates the
time at which the sodium peak of the FP signal propagates from the
start time (0 ms) to end points (10 ms). Scale bar indicates the
distance of the sodium peak propagation. g Quantification of FP
parameters: conduction velocity, spike amplitude and spike slope.
n = 24 cultures/group from 4 independent differentiations of 2 iPSC
lines. A schematic diagram of FP traces is included in Fig. [113]6a,
showing how the data for spike amplitude and slope were analysed.
Symbols in (b, c, g) denote iPSC lines: circles for isWT7, and squares
for iWTD2. Source data are provided as a Source Data file. Statistical
analysis was performed using Kruskal-Wallis test with Dunn’s multiple
comparison test (b, c, g) and two-way ANOVA with Sidak’s multiple
comparison test (e): *p < 0.05, **p < 0.01, ***p < 0.001,
****p < 0.0001 compared to the B27 group. Exact p values are provided
in the Source Data file. Data are presented in box plots indicating
median (middle line), 25th, 75th percentile (box) and min and max data
points (whiskers) in (b, c, g) and in line plots as mean ± SEM (e).
Intercellular electrotonic coupling and conduction velocity (CV) across
CMs are largely dependent on Cx43 expression at the gap
junction^[114]32. Consistent with the increased expression of Cx43 on
the cell membrane (Fig. [115]1d), heatmaps of electrical signal
propagation analysed using MEA illustrate the stepwise increase in CV
in iPSC-CMs in the MM (22.3 ± 3.7 cm/s), MM + NP (25.6 ± 4.3 cm/s) and
MM + NP + ES (27.8 ± 7.3 cm/s) groups, compared to the B27 condition
(12.5 ± 5.8 cm/s). Similar stepwise changes in spike amplitude and
slope were observed in the four groups (Fig. [116]2f, g, Supplementary
Table [117]3).
To analyse which changes in specific ion currents underlie the improved
electrophysiological functionality, we recorded I[Na], I[K1], and I[Kr]
using the patch clamp technique. We found that I[Na] density was
significantly higher in MM-cultured iPSC-CMs with a mean peak current
density of −81.6 ± 6.3 pA/pF at −25 mV compared to the B27 group
(−42.3 ± 4.7 pA/pF at −20 mV). Notably, NP induced only a small
increase in I[Na] density with a mean peak of −92.1 ± 6.9 pA/pF when
compared to the MM group, but I[Na] density was further significantly
induced by ES in the MM + NP + ES group with a mean peak of
−135.7 ± 9.8 pA/pF (Fig. [118]3a, b). These data are consistent with
the AP- (Vmax, APA) and FP-metrics (CV, spike amplitude and slope)
shown in Fig. [119]2.
Fig. 3. I[Na], I[K1] and I[Kr] recordings in iPSC-CMs.
[120]Fig. 3
[121]Open in a new tab
a Representative I[Na] traces recorded under 25 mM extracellular Na^+
concentration. b Statistical analysis of I[Na] in single cells derived
from 9 (B27) and 5 (MM, MM + NP, MM + NP + ES) independent
differentiations of 3 iPSC lines. The stimulation protocol is shown as
an inset. The holding potential was set at −100 mV. I[Na] was recorded
by increasing the test potential from −80 mV to +70 mV in 5 mV steps.
Each pulse lasted for 20 ms and the sweep interval was 2 s. c
Representative traces of 0.5 mM BaCl[2]-sensitive I[K1] for different
groups. d Statistical analysis of BaCl[2]-sensitive I[K1] in single
cells derived from 6 (B27, MM, MM + NP + ES) and 5 (MM + NP)
independent differentiations of three iPSC lines. The stimulation
protocol is shown as an inset. The holding potential was set at −40 mV.
I[K1] was recorded by increasing the test potential from −130 mV to
+10 mV in 10 mV steps, with each pulse lasting for 2 s. The sweep
interval was 10 s. The protocol was then repeated in the presence of
0.5 mM BaCl[2] and the Ba^2+-sensitive current was calculated as I[K1].
e Shown are traces of 1 µM E−4031-sensitive I[Kr] in the four groups.
f, g Averaged E−4031-sensitive I[Kr] step currents (f) and tail
currents (g) in single cells derived from 4 independent
differentiations of 2 iPSC lines. The I[Kr] pulse stimulation is shown
as an inset. The holding potential was set at −50 mV. I[Kr] step
currents were recorded by decreasing the test potential from +40 mV to
−40 mV in 10 mV steps with each pulse lasting for 2.5 s. The tail
currents were recorded in the following 4-s phase at −40 mV. The sweep
interval was 10 s. With respect to I[Na], I[K1], and I[Kr], we did not
observe significant differences among three iPSC lines used
(Supplementary Fig. [122]7). Source data are provided as a Source Data
file. Two-way ANOVA with Sidak’s multiple comparison test was used for
statistical analysis: *p < 0.05, **p < 0.01, ***p < 0.001,
****p < 0.0001 compared to the B27 group. Exact p values are provided
in the Source Data file. Data are presented as mean ± SEM.
The electrophysiological immaturity of iPSC-CMs compared to adult CMs
is partly attributed to the low density of the hyperpolarising K^+
current I[K1], which is important for stabilising the RMP^[123]33. We
found a very low I[K1] density in B27-cultured iPSC-CMs (−2.1 ± 0.3
pA/pF at −130 mV), and only a slight increase in I[K1] in the MM group
(−5.3 ± 0.8 pA/pF). Interestingly, NP induced a significant increase in
I[K1] in the MM + NP group (−11.7 ± 1.8 pA/pF), which was further
induced by ES (−16.0 ± 1.9 pA/pF in the MM + NP + ES group)
(Fig. [124]3c, d). HERG channels conducting the rapid delayed rectifier
K^+ current I[Kr] are involved in phase 3 repolarisation of cardiac
APs. Similar to I[K1], both E-4031-sensitive I[Kr] step and tail
current densities were only slightly induced by MM, but significantly
induced by NP and further enhanced by ES (Fig. [125]3e-g). These data
are consistent with the most negative RMP and the shortest APD[90] in
iPSC-CMs from the MM + NP + ES group (Fig. [126]2b, c).
Taken together, these findings highlight the distinct effects of the
three stimuli on specific ion currents. This is evidenced by the strong
influence of NP on I[K1] and I[Kr], with less or no effect on I[Na] and
I[to], and the robust effect of MM on I[Na], but less on I[Kr].
Importantly, the data underline that the combined approach
significantly enhanced electrophysiological maturation of iPSC-CMs.
Combined approach improves calcium handling and contractility
Since excitation-contraction coupling in CMs involves calcium cycling
to convert electrical signals into mechanical output (contraction), we
next examined L-type calcium channel (LTCC) current I[Ca-L] and calcium
transients in iPSC-CMs. I[Ca-L] densities exhibited similar reductions
in both the MM and MM + NP groups when compared to the B27 group, which
were further reduced by ES (Fig. [127]4a, b). To quantify intracellular
Ca^2+ dynamics, we performed Fura-2-based calcium imaging in iPSC-CMs
paced at 0.5 Hz (Fig. [128]4c-e). Significantly reduced diastolic and
systolic Ca^2+ levels were observed in the MM, MM + NP and MM + NP + ES
groups compared to the B27 group, but Ca^2+ transient amplitudes were
comparable between all groups despite the reduced I[Ca-L] density in
the MM, MM + NP and MM + NP + ES groups. The Ca^2+ transient decay time
constant (tau) is also significantly shortened in the MM group compared
to the control, whereas no further shortening was observed in the
MM + NP and MM + NP + ES groups. Application of 10 mM caffeine resulted
in significantly increased Ca^2+ release from the sarcoplasmic
reticulum (SR) in the MM, MM + NP and MM + NP + ES groups compared to
the B27 group (Fig. [129]4e). These data suggest a more efficient
coupling between I[Ca-L] and SR Ca^2+ release, enhanced Ca^2+ decay
kinetics and a higher SR calcium content in these three groups.
Fig. 4. Assessment of calcium handling in iPSC-CMs and effects of calcium
concentrations used in media.
[130]Fig. 4
[131]Open in a new tab
a Representative I[Ca-L] recordings in iPSC-CMs cultured under the four
conditions. b Statistical analysis of I[Ca-L] in single cells derived
from 6 independent differentiations of 3 iPSC lines. c Representative
Ca^2+ transient traces recorded at 0.5 Hz followed by application of
10 mM caffeine to induce the release of total SR calcium. d Statistical
analysis of calcium transient parameters: diastolic Ca^2+ level,
systolic Ca^2+ level, transient amplitude and decay time constant tau
measured in iPSC-CMs paced at 0.5 Hz. n = 15, 18, 17, and 20 cells for
B27, MM, MM + NP, and MM + NP + ES groups, respectively, from 2 iPSC
lines. e SR calcium release induced by 10 mM caffeine. n = 16, 14, 12,
and 18 cells for B27, MM, MM + NP, and MM + NP + ES groups,
respectively, from 2 iPSC lines. f–h Experimental scheme (f), and
effects of calcium concentrations on Ca^2+ transient amplitude and tau
(g, n = 38 cells/group from 2 iPSC lines), and SR Ca^2+ release induced
by 10 mM caffeine (h, n = 18 and 15 cells for the (RPMI + )B27 and
DMEM + B27 groups from 2 iPSC lines, respectively). f is created in
BioRender. Li, W. (2025) [132]https://BioRender.com/t88d878. i Effects
of calcium concentrations on I[Na] in single cells derived from 9
((RPMI + )B27) and 6 (DMEM + B27) independent differentiations of 3
iPSC lines. j Effects of calcium concentrations on I[Ca-L] in single
cells derived from 6 ((RPMI + )B27) and 4 (DMEM + B27) independent
differentiations of 3 iPSC lines. Symbols in (d, e, g, h) denote iPSC
lines: circles for isWT7, squares for iWTD2 and triangles for iBM76.
Source data are provided as a Source Data file. Statistical analysis
was performed using Kruskal-Wallis test with Dunn’s multiple comparison
test (d, e) or two-sided Kolmogorov-Smirnov test (g, h). Two-way ANOVA
with Sidak’s multiple comparison test was used in (b, i, j) *p < 0.05,
**p < 0.01, ***p < 0.001, ****p < 0.0001 compared to the B27 group.
Exact p values are provided in the Source Data file. Data are presented
as mean ± SEM (b, i, j) and in box plots indicating median (middle
line), 25th, 75th percentile (box) and min and max data points
(whiskers) in (d, e, g, h).
Previous studies reported that high calcium concentration induces
positive force-frequency behaviour, physiological twitch kinetics and
robust β-adrenergic response of EHT by improving Ca^2+
handling^[133]30. To investigate the contribution of the high calcium
concentration in MM to the maturation of iPSC-CMs, we cultured iPSC-CMs
in DMEM + B27 medium containing 1.8 mM calcium (Supplementary
Table [134]1) compared to (RPMI + )B27 medium containing 0.4 mM calcium
(Fig. [135]4f-j). Analysis of the calcium transient parameters revealed
only slightly increased Ca^2+ transient amplitudes but significantly
decreased tau in iPSC-CMs cultured in DMEM + B27 compared to
(RPMI + )B27 (Fig. [136]4g). Furthermore, caffeine-induced SR Ca^2+
release was significantly increased in the DMEM + B27 group compared to
the (RPMI + )B27 group (Fig. [137]4h). Interestingly, all changes were
similar to those in the MM group (Fig. [138]4d, e). We also found that
cultivation in DMEM + B27 led to an increase in I[Na] density to
−90.4 ± 7.2 pA/pF at −20 mV compared to −42.3 ± 4.7 pA/pF in the
RPMI + B27 group (Fig. [139]4i). I[Ca-L] density was significantly
reduced in the DMEM + B27 group (−6.2 ± 0.8 pA/pF at 10 mV) compared to
the B27 group (−8.4 ± 0.5 pA/pF) (Fig. [140]4j). Both I[Na] and I[Ca-L]
in the DMEM + B27 group are comparable to those in the MM group
(Figs. [141]3a and Fig. [142]4b). Overall, these results demonstrate
the strong influence of high Ca^2+ concentrations on the
electrophysiological properties of iPSC-CMs.
Movie-based analysis of iPSC-CM beating properties^[143]5 showed that
the changes in calcium handling were associated with improved
contractile function. Stopping ES in the MM + NP + ES group resulted in
cessation of beating, followed by regaining of spontaneous beating
activity within 15-30 minutes at a rate comparable to the other three
groups (Fig. [144]5a). Significantly shorter contraction and relaxation
times and beating duration were observed in the MM + NP + ES group
compared to the other groups (Fig. [145]5b; Supplementary Fig. [146]2a,
b). We found a similar trend in the MM and MM + NP groups compared to
the B27 control, but no significant difference between the two groups.
Consistent with this observation, all cultures in the MM + NP + ES
group successfully captured high-frequency (2 Hz) field stimulation,
whereas none of the B27 cultures demonstrated this capability
(Fig. [147]5c; Supplementary Fig. [148]2c). This improved contractile
function was accompanied by an increased gene expression ratio of
TNNI3/TNNI1 and MYL2/MYL7, whereas the expression of MYH6, encoding the
fast-twitch MHC isoform, was upregulated in response to sustained ES,
leading to a reduced MYH7/MYH6 ratio (Fig. [149]5d). Flow cytometry
analysis showed that in all three groups (MM, MM + NP, and
MM + NP + ES) there was a noticeable increase in cell volume and
granularity of iPSC-CMs compared to the B27 control. Compared to the MM
group, NP did not induce an additional increase in cell volume and
granularity, but the combination of MM + NP + ES induced further
significant increases (Fig. [150]5e), suggesting that ES plays an
important role in the hypertrophic growth of iPSC-CMs. Whereas a
comparable proportion of cardiac troponin T (cTNT)-positive cells was
found in all conditions, the highest cTNT mean fluorescence intensity
was observed in iPSC-CMs from the MM + NP + ES group (Fig. [151]5f, g).
These results demonstrate that MM, NP and ES individually and
synergistically induce electrophysiological and functional maturation
of iPSC-CMs.
Fig. 5. Analysis of contractility and structural development.
[152]Fig. 5
[153]Open in a new tab
a Analysis of beating rate. CMs of the MM + NP + ES group contracted
with a beating rate of 120 BPM (beats per minute) during the presence
of ES but regained spontaneous beating after ES was discontinued.
n = 15 cultures/group of 5 independent differentiations of 3 iPSC
lines. b Statistical analysis of beating properties: contraction time,
relaxation time and beating duration of iPSC-CMs under 0.5 Hz field
stimulation. n = 15 cultures/group of 5 independent differentiations of
3 iPSC lines. A schematic diagram of two beat traces showing how the
parameters were analysed is shown in Supplementary Fig. [154]2a. c
Heatmap of the ability of iPSC-CMs to adapt to increasing pacing
frequencies from n = 15 cultures/group of 5 independent
differentiations of 3 iPSC lines. The colour scale represents the
number of cultures. Representative beating traces are shown in
Supplementary Fig. [155]2b,c. d Expression of denoted marker genes for
structural maturation. n = 7 (TNNI1, TNNI3, MYL2, MYL7) and 9 (MYH6,
MYH7) independent differentiations of 3 iPSC lines. HPRT is used as a
housekeeping gene. e Quantification of cell volume (FSC-A) and
granularity (SSC-A) in the cTNT-positive CM populations using flow
cytometry analysis. n = 18 independent differentiations/group of 3 iPSC
lines. f, g Proportion of cTNT-positive cells (f) and quantification of
mean fluorescence intensity of cTNT (g). n = 17 (f) and 11 (g)
independent differentiations/group of 3 iPSC lines. Log transformation
was performed for statistical analysis (g). The gating strategy used to
analyse the flow cytometry data is shown in Supplementary Fig. [156]3.
Symbols denote iPSC lines: circles for isWT7, squares for iWTD2, and
triangles for iBM76. Source data are provided as a Source Data file.
Statistical analysis was performed using Kruskal-Wallis test with
Dunn’s multiple comparison test (b, d) and linear mixed model
(two-sided) with Tukey’s correction for multiple comparisons between
the 4 groups (e–f). Data are presented in box plots indicating median
(middle line), 25th, 75th percentile (box) and min and max data points
(whiskers).
Combined approach improves drug response of iPSC-CMs
To investigate whether the maturation state of iPSC-CMs influences
their drug response, we chose verapamil (calcium-channel blocker),
E-4031 (hERG-channel blocker) and isoprenaline (β-adrenergic stimulus)
as model substances to detect pro-arrhythmic activity based on changes
in FP parameters (Fig. [157]6; Supplementary Fig. [158]4a, b). We
observed beating arrest in cultures from the B27 (17/17), MM (6/17) and
MM + NP (6/18) groups at 1 µM verapamil, but no significant induction
of arrhythmias with all three substances (Supplementary Fig. [159]4c,
d). Concentration-dependent reductions in spike amplitude were observed
for verapamil in the B27 group and, to a lesser extent, in the MM and
MM + NP groups. In contrast, no effect of verapamil on beating activity
and spike amplitude was observed in the MM + NP + ES group
(Fig. [160]6b; Supplementary Fig. [161]4e). Verapamil-induced
shortening of FP duration (FPDc), which corresponds to QT-shortening in
the clinic, was comparable in iPSC-CMs from the MM, MM + NP and
MM + NP + ES groups. However, this effect was more pronounced compared
to iPSC-CMs cultured in B27 (Fig. [162]6c, d). Previous studies showed
that immature iPSC-CMs failed to produce APD prolongation after E-4031
treatment, even at high concentrations^[163]34. Similarly, we found
that E-4031 induced only minor changes in FPDc in B27-cultured
iPSC-CMs, whereas significant concentration-dependent FPDc prolongation
was detected in the MM, MM + NP and MM + NP + ES groups (Fig. [164]6e,
f). Furthermore, we observed a more pronounced positive-chronotropic
response of iPSC-CMs to isoprenaline in these three groups than in the
B27 group, which correlates with FPDc-shortening (Fig. [165]6g, h). The
EC[50] of isoprenaline for the chronotropic effect was also much lower
in the MM + NP (1.9 nM) and MM + NP + ES (1.8 nM) groups than in the MM
(4.3 nM) and B27 (6.8 nM) groups (Fig. [166]6h). These experiments
demonstrate the substantial impact of the maturation state of iPSC-CMs
on their response to various cardioactive drugs. They emphasise the
importance of utilising iPSC-CMs with more adult-like
electrophysiological properties for accurate drug risk assessment.
Fig. 6. Maturation status of iPSC-CMs affects their drug response.
[167]Fig. 6
[168]Open in a new tab
a Schematic diagram of FP traces showing how the data was analysed. b
Quantitative analysis of the effect of verapamil on spike amplitude.
n = 17 (B27, MM) and 18 (MM + NP, MM + NP + ES) cultures derived from 3
independent experiments using 2 iPSC lines. Experimental design is
shown in Supplementary Fig. [169]4. c Representative averaged field
potential (FP) traces of verapamil-treated iPSC-CMs showing the FPDc
(FP duration corrected by Fridericia’s formula) shortening. d
Quantitative analysis of the effect of verapamil on FPDc shortening
(ΔΔFPDc). n = 17 (B27, MM) and 18 (MM + NP, MM + NP + ES) cultures
derived from 3 independent experiments using 2 iPSC lines. e
Representative traces illustrating the FPDc prolongation induced by
increasing concentrations of E-4031. f Quantitative analysis of the
effect of E-4031 on FPDc. n = 10 (B27), 12 (MM), and 11 (MM + NP,
MM + NP + ES) cultures from 2 independent experiments. g, h
Quantitative analysis of concentration-dependent effect of isoprenaline
on FPDc (g) and beating rate (h). n = 18 (B27), 23 (MM, MM + NP), and
24 (MM + NP + ES) cultures derived from 3 (B27) or 4 (MM, MM + NP,
MM + NP + ES) independent experiments of 2 iPSC lines. Data are
normalised to the respective baseline of each group (b, d, f, g, h).
Symbols in (b, d, f, g) denote iPSC lines: circles for isWT7, and
squares for iWTD2. Source data are provided as a Source Data file.
Statistical analysis using two-way ANOVA with Dunnett’s post-test. Data
are presented as mean ± 95% CI (h) and in box plots indicating median
(middle line), 25th, 75th percentile (box) and min and max data points
(whiskers) in (b, d, f, g).
Combined approach downregulates MAPK/PI3K-AKT pathways
Previous studies have shown that FA-enriched media induce iPSC-CM
maturation by regulating key genes involved in FA metabolism,
mitochondrial function, calcium cycling, ion channels and
sarcomere^[170]13,[171]20. To gain insight into the molecular
mechanisms driving iPSC-CM maturation by NP and ES, we performed RNA
sequencing (RNA-seq) analysis. Surprisingly, NP had little synergistic
effect when combined with MM, whereas the addition of ES strongly
influenced gene expression (Fig. [172]7a-d; Supplementary Fig. [173]5a,
b). Comparing the MM and MM + NP groups, only 163 differentially
expressed genes were identified, of which 56 were upregulated and 107
downregulated in the MM + NP group. In contrast, 1370 significantly
upregulated and 1657 downregulated genes were identified in the
MM + NP + ES group compared to the MM group, of which 747 significantly
upregulated and 990 downregulated genes were also identified when
compared to the MM + NP group, indicating the synergistic effects of NP
and ES.
Fig. 7. RNA sequencing of iPSC-CMs cultivated under MM, MM + NP and
MM + NP + ES conditions.
[174]Fig. 7
[175]Open in a new tab
a–d Volcano plots (a–c) and Venn analyses (d) of significantly
differentially expressed genes (DEGs, p < 0.01) between iPSC-CMs from
MM, MM + NP and MM + NP + ES groups. e Enrichment map illustrating
clustered pathways identified in gene set enrichment analysis (GSEA)
based on canonical pathway database. The colour scale represents
normalised enrichment score (NES). Pathways were filtered based on max.
size of 500 genes, NES ≤ −1.9, false discovery rate (FDR) q-value ≤ 0.2
and cluster size of ≥ 2 pathways. f Enrichment plot of SRF_Q4 gene sets
obtained with GSEA. g Heatmaps of the expression of SRF target genes
significantly downregulated in MM + NP + ES. The colour scale
represents z-score. h Western blot of total SRF. n = 4 independent
experiments using 2 iPSC lines. Source data are provided as a Source
Data file. Statistical analysis was performed using the Wald test of
DESeq2 (two-sided) in (a–c, g). Exact p values in g are provided in the
Source Data file.
Pathway enrichment analysis of the downregulated genes in the
MM + NP + ES group mainly mapped to MAPK/PI3K-AKT, TNFR2-NFκB,
G-protein-coupled receptor (GPCR), and cytokine/chemokine signalling
(Fig. [176]7e; Supplementary Fig. [177]5d; Supplementary Data [178]1).
Using the transcription factor targets (TFT) collection
([179]https://www.gsea-msigdb.org/gsea/msigdb/human/genesets.jsp?collec
tion=TFT), we identified the SRF cluster, which includes many genes
involved in cell cycle regulation and cell proliferation (Fig. [180]7f,
g; Supplementary Fig. [181]5c). We also found a decrease in SRF protein
levels in CMs from the MM + NP + ES group compared to the other groups
(Fig. [182]7h). These findings encouraged us to evaluate the expression
of genes that regulate cell cycle progression. Notably, activators of
G2/M checkpoints including cyclins (CCNB1-3), cyclin-dependent kinase 1
(CDK1) were downregulated, whereas CDK inhibitors (CDKN1A, CDC20) were
upregulated in the MM + NP + ES group compared to the other two groups
(Fig. [183]8a, b). Interestingly, genes (ANLN, SEPTIN7/2) encoding
activators of cytokinesis were also downregulated. We did not observe
any significant changes in the gene sets (CCND1-3, CCNE1/2, CCNA1/2,
CDK2/4/6, CDKN2A-D, CDKN1B/C, CDH1) controlling the G1 and S phase
progression and the G1/S checkpoint (Fig. [184]8a, b). These data
suggest a cell cycle arrest after S phase and before exit from M phase
in the MM + NP + ES group, which may lead to bi-nucleation or nuclear
polyploidy. To confirm this, we examined DNA content and found that the
number of diploid iPSC-CMs was significantly reduced and polyploid
cells significantly increased in the MM + NP + ES group compared to the
other three groups (Fig. [185]8c, f). Interestingly, we observed no
difference in the proportion of 5-ethynyl-2’-deoxyuridine
(EdU)-incorporated iPSC-CMs between all four groups, which can detect
DNA synthesis during S phase (Fig. [186]8d, g). However, the proportion
of Ki67^+ iPSC-CMs and Ki67^- polyploid iPSC-CMs was higher in the
MM + NP + ES group than in the other groups (Fig. [187]8e, h, i). Ki67
is widely expressed throughout the entire cell cycle, except in G0, and
reaches a maximum in S/G2^[188]35. These results indicate that the
downregulation of MAPK/PI3K-AKT and SRF-related genes is involved in
G2/M arrest and polyploidy development of iPSC-CMs.
Fig. 8. Cell cycle regulation of iPSC-CMs cultivated under MM, MM + NP and
MM + NP + ES conditions.
[189]Fig. 8
[190]Open in a new tab
a Scheme illustrating the expression of cyclin-CDK complexes and
respective inhibitors during cell cycle. b Heatmaps of cyclin-CDK
complexes, CDK inhibitors, and cytokinesis related genes including
those significantly differentially expressed genes among the three
groups. The colour scale represents z-score. c–e Flow cytometry plots
showing cellular DNA content (c), DNA synthesis activity (d), and Ki67
activity (e). f Quantification of diploid and polyploid iPSC-CMs
(cTNT-positive populations). Data from 17 independent experiments of 3
iPSC lines. g Proportion of EdU-incorporated iPSC-CMs. Data from 4
independent experiments of 2 iPSC lines. h, i Quantification of diploid
and polyploid CMs in Ki67-positive (h) and negative (i) populations.
Data from 7 independent experiments of 3 iPSC lines. Symbols in (f–i)
denote iPSC lines: circles for isWT7, squares for iWTD2, and triangles
for iBM76. Source data are provided as a Source Data file. Statistical
analysis was performed using the Wald test of DESeq2 (two-sided) in (b)
or using linear mixed model (two-sided) with Tukey’s correction for
multiple comparisons between the 4 groups (f–i). Exact p values in (b)
are provided in the Source Data file. Data are presented in box plots
indicating median (middle line), 25th, 75th percentile (box) and min
and max data points (whiskers) in (f–i).
Gene expression profile has links to metabolism and electrophysiology
The pathway enrichment analysis of genes upregulated in the
MM + NP + ES group revealed that the combined approach induced the
upregulation of genes involved in electron transport chain (ETC), TCA
cycle, mitochondrial biogenesis, NRF2 signalling, glucose metabolism,
N-glycan biosynthesis, FA oxidation, tRNA aminoacylation, etc.
(Fig. [191]9a; Supplementary Fig. [192]5e; Supplementary Data [193]2).
These gene clusters were only slightly upregulated in the MM + NP group
compared to the MM group (Supplementary Fig. [194]5e).
Fig. 9. Upregulation of TFAM and HMCES target genes contributes to
mitochondrial development induced by ES.
[195]Fig. 9
[196]Open in a new tab
a Enrichment maps illustrating clustered pathways identified in gene
set enrichment analysis (GSEA) based on canonical pathway database. The
colour scale represents normalised enrichment score (NES). Pathways
were filtered based on max. size of 500 genes, NES ≥ 1.5, false
discovery rate (FDR) q-value ≤ 0.2 and cluster size of ≥ 2 pathways. b,
c Enrichment plots and most regulated genes in TFAM and HMCES clusters.
The colour scale represents z-score. d Quantification of Tom20
intensity detected in cTNT-positive CM populations. n = 10 independent
experiments with 3 different iPSC lines. Log transformation was
performed for statistical analysis. e, f Seahorse mean traces (e) and
determined parameters (f) were performed with sequential addition of
oligomycin (ATP synthase inhibitor), carbonyl
cyanide-p-(trifluoromethoxy) phenylhydrazone (FCCP; mitochondrial
uncoupler) and rotenone/antimycin A (complex 1 and 2 inhibitor). n = 4
(B27, MM + NP + ES) and 5 (MM, MM + NP) independent experiments using 2
iPSC lines. Orange, red, blue and green lines represent the four
experimental groups: B27, MM, MM + NP, and MM + NP + ES, respectively.
g Relative expression of OPA1, PPARGC1α and PPARα determined using
real-time PCR. HPRT is used as control. Data from 7 (PPARα, PPARGC1α)
or 8 (OPA1) independent experiments of 3 iPSC lines. Symbols in (d, f,
g) denote iPSC lines: circles for isWT7, squares for iWTD2, and
triangles for iBM76. h Heatmaps of selected genes encoding ion channels
including those significantly differentially expressed genes among the
three groups. The colour scale represents z-score. Source data are
provided as a Source Data file. Statistical analysis was performed
using linear mixed model (two-sided) with Tukey’s correction for
multiple pairwise comparisons between the 4 groups (d, f),
Kruskal-Wallis test with Dunn’s multiple comparison test (g), or the
Wald test of DESeq2 (two-sided) in (b, c, h). Exact p values in (b, c,
h) are provided in the Source Data file. Data are presented as
mean ± SEM (e) and in box plots indicating median (middle line), 25th,
75th percentile (box) and min and max data points (whiskers) in (d, f,
g).
Using the TFT collection we identified two clusters, TFAM
(Fig. [197]9b) and HMCES (Fig. [198]9c), which were enriched in the
MM + NP + ES group (Supplementary Fig. [199]5c). In these two clusters,
mitochondrial DNA (mtDNA)-encoded NADH dehydrogenase subunits (MTND2-6)
and pseudogenes (MTND4P12, MTND5P11, MTND6P4), cytochrome c oxidase
subunits (MT-CO1/3) and pseudogenes (MT-CO1P2, MT-CO1P12, MT-CO3P12),
cytochrome b (MTCYB), 12S rRNA (MT-RNR1) and tRNAs
(MT-TP/-TM/-TE/-TI/-TW/-TC/-TY/-TR/-TN/-TQ) were upregulated in the
MM + NP + ES group compared to the MM group.
Subsequently, we found increased mitochondrial mass in CMs from the
MM + NP + ES group compared to the other groups (Fig. [200]9d).
Furthermore, CMs from the MM + NP + ES group showed a higher expression
of OPA1, PPARGC1α and PPARα (Fig. [201]9g), confirming an enhanced
mitochondrial development in response to ES. Measurements of oxygen
consumption rate as a surrogate for mitochondrial function showed an
increased basal and maximal respiration, ATP production and spare
capacity of CMs from the MM group compared to the B27 control. There is
only a marginal additional increase in these parameters in the
MM + NP + ES and MM + NP groups compared to the MM group (Fig. [202]9e,
f). These findings indicate that the combination of MM + NP + ES leads
to upregulation of oxidative phosphorylation, activation of
mtDNA-encoded components and an overall enhancement of mitochondrial
development and function.
Further examination of individual changes in key ion channel components
revealed a clear correlation between gene expression and channel
function (Fig. [203]9h). We found upregulation of genes contributing to
I[to] (KCNA7, KCNC3, KCNC4, KCND2, KCND3), I[K1] (KCNJ2, KCNJ4, KCNJ6,
KCNJ11, KCNJ12), I[Kr] (KCNH2, KCNE2) and I[Ks] (KCNQ1) currents, as
well as genes encoding calcium-activated (KCNN1) and voltage-gated
(KCNS3, KCNAB2, KCNIP3) potassium channels in the MM + NP + ES group
compared to the MM group. In contrast, genes encoding the LTCCs
(CACNA1C, CACNA1D, CACNA1A, CACNA2D3, CACNA2D4) and regulating the LTCC
activity (CACNG7) were downregulated, whereas the genes encoding the
T-type calcium channels (CACNA1H, CACNA1I) were upregulated. No
significant differences in SCN5A expression were found between the
three groups, but SCN1B was upregulated and SCN3B was downregulated in
the MM + NP + ES group. In addition, the expression of the genes
encoding calcium handling proteins were not significantly altered.
Collectively, these results are consistent with the significantly
increased I[Na], I[to], I[K1], and I[Kr] but decreased I[Ca-L] currents
in iPSC-CMs from the MM + NP + ES group.
MM + ES alone induces electrophysiological maturation and polyploidy of
iPSC-CMs
As MM + NP had little synergistic effect on gene expression profile
when compared to MM alone, it is interesting to test whether MM + ES
alone can induce the maturation of iPSC-CMs similar to MM + NP + ES. To
study this, we compared iPSC-CMs cultured under MM and MM + ES
conditions. We observed no significant differences in cell morphology,
cTNT intensity, and cell size (FSC-A) (Fig. [204]10a-c). Consistent
with the changes observed in the MM + NP + ES group, iPSC-CMs in the
MM + ES group showed increased granularity (SSC-A), and Tom20 intensity
(Fig. [205]10c, d). These data indicate that ES alone enhances
structural maturation, whereas the addition of NP may contribute to
cell alignment, cell size and cTNT intensity. We observed the
‘notch-and-dome’ AP morphology in iPSC-CMs (32%) of the MM + ES group
(Fig. [206]10e) comparable to those in the MM + NP + ES group. iPSC-CMs
in the MM + ES group had a more negative RMP, increased Vmax and APA,
and shortened APD[90] compared to the MM group (Fig. [207]10f), but
with a less extent when compared to the MM + NP + ES group
(Fig. [208]2b). Similarly, I[to] density was increased in the MM + ES
group compared to the MM group (Fig. [209]10g). Investigation of the
contractile function revealed a slight shortening of the contraction
time, relaxation time and beating duration of iPSC-CMs in the MM + ES
group compared to MM (Fig. [210]10h), but these changes were smaller
compared to those observed with MM + NP + ES versus MM (Fig. [211]5a).
Similar to the MM + NP + ES group, all iPSC-CM cultures in the MM + ES
group captured the pacing frequency of 2 Hz (Fig. [212]10i).
Fig. 10. Effect of ES on iPSC-CM maturation.
[213]Fig. 10
[214]Open in a new tab
a Representative morphology of iPSC-CMs under MM and MM + ES
conditions. b–d Quantification of cTNT intensity (b), cell volume
(FSC-A) and granularity (SSC-A) in the cTNT-positive CMs (c), and Tom20
intensity (d). n = 8 independent experiments with 3 different iPSC
lines. Log transformation (b, d) was performed for statistical
analysis. e A notch event (red arrow) in an iPSC-CM of the MM + ES
group. f Quantification of action potential (AP) metrics: resting
membrane potential (RMP), maximum upstroke velocity (Vmax), AP
amplitude (APA), and AP duration at 90% repolarization (APD[90]). For
RMP, Vmax and APA: n = 15 (MM) and 22 (MM + ES) cells from 3 (MM) and 4
(MM + ES) independent differentiations of 2 iPSC lines. For APD[90]:
n = 15 (MM) and 20 (MM + ES) cells from 3 (MM) and 4 (MM + ES)
independent differentiations of 2 iPSC lines. g Statistical analysis of
I[to] in single cells derived from 5 independent differentiations of 2
iPSC lines. The stimulation protocol is shown as an inset. h
Statistical analysis of beating properties: contraction time,
relaxation time and beating duration of iPSC-CMs under 0.5 Hz field
stimulation (n = 20 cultures/group from 5 independent differentiations
of 3 iPSC lines). i Heatmap of the ability of iPSC-CMs to adapt to
increasing pacing frequencies (n = 9 cultures/group from 3 independent
differentiations of 3 iPSC lines). The colour scale represents the
number of cultures. j, k Shown are flow cytometry plots for DNA content
(j) and Ki67 activity (k). l-n Quantification of diploid and polyploid
cells in total iPSC-CMs (l) and in Ki67-positive (m) and negative (n)
populations. n = 8 independent experiments with 3 different iPSC lines.
Symbols in (b–d, f, h, l–n) denote iPSC lines: circles for isWT7,
squares for iWTD2, and triangles for iBM76. Source data are provided as
a Source Data file. Statistical analysis was performed using linear
mixed model and t-test (two-sided) (b–d, l–n), Two-way ANOVA with
Sidak’s multiple comparison test (g), or two-sided Kolmogorov-Smirnov
test (f, h). Data are presented as mean ± SEM (g) and in box plots
indicating median (middle line), 25th, 75th percentile (box) and min
and max data points (whiskers) in (b–d, f, h, l–n).
To establish whether ES is the pivotal stimulus for the development of
polyploidy of iPSC-CMs, DNA content was measured in iPSC-CMs. The
proportion of diploid iPSC-CMs was significantly reduced, while the
number of polyploid CMs was significantly increased in the MM + ES
group compared to the MM group (Fig. [215]10j, l). Additionally, we
observed the increase in the number of Ki67^+ and Ki67^- polyploid
iPSC-CMs in the MM + ES group compared to the MM group (Fig. [216]10k,
m, n), similar to those in the MM + NP + ES group (Fig. [217]8h, i).
Taken together, these results demonstrate that ES alone can enhance
mitochondrial development, electrophysiological function and polyploidy
of iPSC-CMs.
Discussion
In this study, we systematically investigated the collective impact of
MM, elevated calcium concentration in the medium, NP and ES on the
maturation of iPSC-CMs. Our findings demonstrate that the concurrent
application of MM with a high concentration of calcium, NP and ES
serves as an efficient strategy to enhance the structural,
electrophysiological, metabolic and functional maturation of iPSC-CMs.
This maturation process involves the modulation of MAPK/PI3K-AKT
signalling as well as the regulation of TFAM/HMCES and SRF target
genes, which is of significant relevance for the utilisation of
iPSC-CMs in disease modelling and drug testing.
The core of the maturation strategy is the application of
FA-supplemented MM in iPSC-CMs, which has been reported in several
studies^[218]13,[219]16,[220]20,[221]22. Previous studies have shown
that MM induces the metabolic transition from glucose-based energy
production to FA β-oxidation^[222]13,[223]20 and changes in the
expression of genes associated with calcium cycling, ion channels and
structural proteins^[224]13,[225]16,[226]22. This metabolic transition
is essential for increased Ca^2+ transient kinetics, I[K1] density, and
iPSC-CM hypertrophy and improved AP parameters, mitochondrial density
and function, contractility and drug response not only in
2D-cultures^[227]13,[228]20 but also in 3D-tissues generated with
iPSC-CMs and fibroblasts^[229]16,[230]22. In line with these studies,
we show here that MM induces enhanced structural (cell size, cTNT and
Tom20 intensity), electrophysiological (increased I[Na], I[to], I[K1]
and I[Kr] density, improved FP parameters and Ca^2+ cycling) and
metabolic (mitochondrial respiration) maturation of iPSC-CMs, leading
to improved contractility and drug response. The effect of MM on I[Na],
I[Ca-L] and Ca^2+ cycling is largely due to the high concentration of
calcium in MM.
One of the contributions of the NP surface to the maturation of
iPSC-CMs is the improvement of cell alignment, granularity, sarcomere
length and contractile behaviour, but their structural maturation is
still less pronounced compared to 3D tissue. 3D tissue allows symmetric
contractions of EHTs with a significantly improved structure, including
T-tubule formation and colocalisation with RYR2, sarcomere organisation
and length^[231]16–[232]18,[233]28. This difference may be due to the
need of 2D monolayers to attach to the surface, where there is a
significant mechanical mismatch. The NP surface with a Young’s modulus
of ~7 mPa is much higher compared to diastolic adult human myocardium
in the range of 8-15 kPa^[234]36 and is therefore still a major
limitation for its use in 2D culture when compared to 3D tissue or 2D
cultures with other techniques (e.g., on PDMS with 8 kPa)^[235]24 that
allow consistent fractional shortening across the tissue.
By stepwise addition of NP and ES to MM, we found that the combination
of NP with MM had only a limited impact on the metabolic maturation of
iPSC-CMs, as demonstrated by subtle changes in the expression profile
of gene clusters related to energy metabolism as well as mitochondrial
development and function, such as ETC (also known as oxidative
phosphorylation), TCA cycle, FA oxidation, and glucose metabolism
(Supplementary Fig. [236]5e). Strikingly, the addition of ES to MM + NP
resulted in significant upregulation of these gene clusters, likely due
to the increased FA β-oxidation in iPSC-CMs to meet the energy demand
for the persistent beating activity at 2 Hz.
An important finding of our study is that GSEA analysis of the
upregulated genes in MM + NP + ES using the transcription factor target
database maps to the enrichment of HMCES- and TFAM-related target gene
sets. HMCES (5-hydroxymethylcytosine binding, embryonic stem
cell-specific) may safeguard the genomic and mtDNA integrity of
iPSC-CMs during oxidative stress responses triggered by elevated levels
of reactive oxygen species due to the use of FA. This protective
mechanism involves the formation of stable DNA-protein crosslinks with
abasic DNA damage to prevent error-prone repair
pathways^[237]37–[238]39. TFAM (mitochondrial transcription factor A)
is essential for the transcription, replication and packaging of mtDNA
into nucleoids. It is indispensable for the meticulous regulation of
mitochondrial biogenesis, ensuring the seamless adaptation of the
mitochondrial population to precisely match the energy demand of the
cell^[239]40. This is supported by our observation that most
mtDNA-encoded essential components of the ETC as well as rRNAs and
tRNAs required for the translation of mtDNA-encoded proteins were
significantly upregulated by the ES stimulation (Fig. [240]9).
PGC-1α (PPARGC1α) is the master regulator of mitochondrial energy
metabolism, respiration and biogenesis through interaction with its
various coactivators ERR, PPAR and NRF1/2^[241]41. Together with ERR,
it controls mitochondrial dynamics via activation of genes involved in
mitochondrial fission and fusion including OPA1 and MFN2^[242]42, which
were found to be upregulated in the MM + NP + ES group. Together with
PPARα, which is also upregulated in the MM + NP + ES, PGC-1α regulates
genes involved in mitochondrial FA oxidation and many other cellular
lipid metabolic pathways^[243]40,[244]41. Furthermore, PGC-1α together
with NRF1/2 promotes mitochondrial biogenesis by activating
TFAM^[245]40,[246]41.
Another interesting discovery of our study revealed a significant
downregulation of genes involved in MAPK/PI3K signalling pathways in
the MM + NP + ES group. This aligns with the substantial downregulation
of the MAPK and PI3K-AKT pathways in the postnatal heart when compared
to the neonatal heart^[247]43. Thus, MAPK/PI3K-AKT inhibition promotes
iPSC-CM maturation, partially mediated by the upregulation of
PGC-1α^[248]43. Further research is needed to determine whether the
downregulation of MAPK/PI3K signalling pathways in the MM + NP + ES
group is involved in PGC-1α and TFAM activation through the control of
AMPK^[249]41,[250]44, AKT^[251]45 and/or mTORC1^[252]46 pathways, and
is therefore essential for the metabolic maturation of iPSC-CMs.
Most strikingly, our RNA-seq and cell cycle analysis data show that
downregulated MAPK/PI3K-AKT signalling is involved in the
downregulation of genes important for the G2/M transition and
cytokinesis, leading to polyploidy (nuclear polyploidy and/or
multinucleation) of iPSC-CMs. Human CMs are diploid during the first
years of life and gradually become polyploid over time. By the second
decade of life, approximately 25% of CM are multinucleated and 57%
polyploid^[253]47. In CMs, the major cyclin-CDK complexes controlling
cell cycle progression are CCND-CDK4/6 (G1 phase), CCNE-CDK2 (G1/S
transition), CCNA-CDK2/1 (S and G2 phase and S/G2 transition) and
CCNB-CDK1 (G2/M transition and M phase), the activity of which is
inhibited by CDK inhibitors^[254]35,[255]48,[256]49. The CCNB-CDK1
complex is not expressed in cell cycle-arrested adult CMs, and it is
also not required for CM hypertrophy^[257]49. In our study, we found a
decrease in the expression of CCNB-CDK1 in iPSC-CMs from the
MM + NP + ES group. Despite this decrease, these cells exhibited
ongoing DNA synthesis, an increased proportion of polyploid CMs and
higher cell volume and granularity. The downregulated CCNB-CDK gene
expression is associated with the upregulation of CDK inhibitor p21 and
MEIS1 that are negatively regulated by TBX20^[258]50. Better
understanding of how MAPK/PI3K-AKT signalling controls cell cycle
progression, including the expression of TBX20, MEIS1 and p21, may
provide valuable mechanistic insights to promote iPSC-CM maturation or
to stimulate adult CM regeneration.
Another important finding of our study is that NP and ES
synergistically induce the maturation of different ion channels. We
found that NP combined with MM strongly increased I[Kr] and I[K1], but
had little effect on I[to], I[Na] and I[Ca-L]. However, the addition of
ES to MM + NP led to significant changes in all currents (larger I[Na],
I[to], I[K1], and I[Kr], but smaller I[Ca-L]) and maturation of
electrophysiology (more negative RMP, shorter APD, higher Vmax, APA, CV
and spike amplitude, and the ‘notch-and-dome’ AP morphology) in
iPSC-CMs, similar to adult CMs^[259]9,[260]33,[261]51. The EC[50]
values for isoprenaline chronotropy in this study fell within the
nanomolar range (1-47 nM) of EC[50] values reported for EHTs cultured
in MM^[262]16,[263]22. Slightly higher EC[50] values have been reported
for EHTs subjected to pacing (98 nM)^[264]28, and for foetal cardiac
tissue (30 nM)^[265]28 and adult human heart slices (180 nM)^[266]52.
Notably, significant differences in EC[50] values for isoprenaline
chronotropy were observed in two different iPSC lines (1 nM vs.
47 nM)^[267]16. The enhanced electrophysiological functionality of
iPSC-CMs contributes to their improved predictive value for risk
assessment of cardioactive drugs, especially in the case of
multi-channel blockers such as verapamil, ranolazine or
alfuzosin^[268]15,[269]16. Several studies reported that verapamil
inhibits the beating activity of iPSC-CMs^[270]15,[271]16,[272]53,
probably because depolarisation in immature iPSC-CMs does not rely
exclusively on I[Na], as in adult CMs, but also on I[Ca-L]^[273]15. In
line with these studies^[274]15,[275]16, our data demonstrate that
increased maturation reduces verapamil-induced beating arrest of
iPSC-CMs. The RNA-seq data suggest that the increased I[to], I[K1], and
I[Kr] and the decreased I[Ca-L] in the MM + NP + ES group may be due to
the upregulation of genes encoding different potassium channels and the
downregulation of genes encoding LTCCs. In addition, the downregulation
of PI3K-AKT signalling may contribute to the reduced I[Ca-L]
densities^[276]54. Furthermore, the similar Ca^2+ transient amplitudes
of all groups together with the decreased I[Ca-L] in iPSC-CMs from MM,
MM + NP and MM + NP + ES groups in comparison to the B27 condition
suggest an enhanced excitation-contraction coupling gain, an important
indicator of improved calcium handling^[277]23. The colocalisation of
RYR2 with α-actinin and the ability to adapt to high-frequency
stimulation support an improved calcium handling, especially in CMs
from the MM + NP + ES group. Future studies should investigate how the
automaticity of iPSC-CMs in the MM + NP + ES is affected, which is
controlled by the coupled system of Ca^2+ and membrane clocks^[278]55
and may involve T-type calcium channels and HCN channels^[279]56.
Interestingly, we observed no changes in the expression of SCN5A,
coding for the pore-forming α-subunit of the sodium channel (Na[V]1.5),
but an upregulation of SCN1B and a downregulation of SCN3B, which
encode the β-subunits (Na[V]-β1/3) of the sodium channel that interact
with the α-subunits^[280]57. While SCN1B is highly expressed in adult
CMs, SCN3B is highly expressed in the embryonic heart^[281]58. Previous
studies have shown that co-expression of SCN1B with SCN5A increases the
density of I[Na]^[282]58, and the β1-subunit modulates the cell surface
localisation, gating, and kinetics of α-subunits^[283]59. In addition,
Na[V]-β1 also regulates voltage-gated potassium channels, including
K[V]4.3 and associates with the cardiac intercalated disc proteins
N-cadherin and Cx43^[284]59, contributing to the enhanced electrical
signal conduction. Future studies should focus on whether/how the
downregulation of MAPK/PI3K-AKT signalling in iPSC-CMs regulates the
gene expression related to ion channel (for example, K^+ and Na^+
channels) maturation and function^[285]60 as well as the formation of
Cx43 gap junction plaques^[286]61 and intercalated
discs^[287]43,[288]62.
Finally, when we compared the expression pattern of marker genes,
important for ventricular CM maturity, cardiac ion channels, and genes
related to cell cycle activity and mitochondrial development in
iPSC-CMs of the MM + NP + ES group with published RNA-seq datasets of
human foetal ventricle and adult heart samples, we found that the
maturity level of iPSC-CMs was intermediate between that of human
foetal and adult CMs (Supplementary Fig. [289]6).
Taken together, we demonstrate that the combined application of MM, NP
and ES synergistically induces structural, electrophysiological,
metabolic and functional maturation of iPSC-CMs and provide first
insights into the mechanism driving advanced maturation of iPSC-CMs.
Cultivation in FA-enriched MM strongly improves mitochondrial
development and electrophysiological functionality of iPSC-CMs.
Although the addition of NP to MM had little effect on the gene
expression profile, it induced a specific increase in I[K1] and I[Kr]
current densities and cell alignment. The MM + NP + ES combination
induces molecular changes that occur during cardiac development,
leading to increased structural maturation, polyploidy, improved
mitochondrial development, and current patterns of I[Na], I[to], I[K1],
I[Kr] and I[Ca-L] more similar to adult human CMs. These changes
translated into an altered sensitivity of iPSC-CMs to cardioactive
drugs, suggesting the efficacy of our maturation approach to improve
the predictive power of iPSC-CMs in drug screening. Furthermore, the
improved maturation of iPSC-CMs highlights the potential of our
combined approach to recapitulate clinical phenotypes that require an
advanced development state of iPSC-CMs.
Methods
Directed differentiation and pro-maturation culture of iPSC-CMs
In this study, three human iPSC lines were used, which were
reprogrammed from somatic cells of three healthy individuals
previously. iWTD2.1 (also known as UMGi001-A.1, FB2-iPS1) and iBM76.3
(UMGi005-A.3, MSC3-iPS3) were generated from dermal fibroblasts and
mesenchymal stem cells, respectively, using STEMCCA
lentivirus^[290]31,[291]63. isWT7.22 (UMGi020-B clone 22) was generated
from dermal fibroblasts using the integration-free CytoTune-iPS 2.0
Sendai Reprogramming Kit^[292]64. All three cell lines were
authenticated by karyotyping and pluripotency assessment, as published
previously^[293]31,[294]64. Regular mycoplasma testing was conducted
via PCR analysis using specific primers (for:
5’-ACACCATGGGAGCTGGTAAT-3’ and rev: 5’-CTTCWTCGACTTYCAGACCCAAGGCAT-3’),
confirming the absence of contamination. The iPSC generation and
application in research were approved by the Ethics Committee of the
University Medical Centre Göttingen (21/1/11 and 10/9/15) and TU
Dresden (EK422092019).
All iPSCs were cultured on Geltrex (Thermo Fisher Scientific, A1413301)
coated 6-well plates in Essential 8 (E8) medium (Thermo Fisher
Scientific, A1517001) with daily medium change. Cells were passaged or
differentiated when they were ~85% confluent. To initiate
differentiation, cells were cultured in RPMI 1640 medium (Thermo Fisher
Scientific, 72400021) with Glutamax and HEPES, 0.5 mg/mL human
recombinant albumin, and 0.2 mg/mL L-ascorbic acid 2-phosphate and
treated with 4 µM CHIR99021 (Merck Millipore, 361559), an inhibitor of
GSK3β. After 48 h, CHIR99021 was removed and the cells were treated
with 5 µM IWP2 (Wnt antagonist II, Merck Millipore, 681671) for another
two days. The first beating cells were detected on day 8 post
differentiation. From day 8, cells were cultivated in B27 medium
containing RPMI 1640, supplemented with 1x B27 with insulin (Thermo
Fisher Scientific, 17504044).
On day 15 after differentiation, the cells were digested. Cells were
first incubated with 1 mg/mL collagenase B (Worthington Biochemical,
CLSAFB) for 1 h at 37°C, then detached iPSC-CM clusters were gently
collected in a 15-mL Falcon tube and dissociated with 0.25%
trypsin/EDTA (Thermo Fisher Scientific) for 8 min. Dissociated iPSC-CMs
were resuspended in cardio-digestion medium (80% B27 medium, 20% foetal
calf serum, and 2 µM thiazovivin (Merck Millipore, 420220)). The
resuspended cells were seeded into Geltrex-coated 6-well plates for the
B27, MM, MM + ES and DMEM + B27 groups, or onto Geltrex-coated ø25 mm
nanopatterned (NP) coverslips (NanoSurface Coverglass, Curi Bio) in
6-well plates for the MM + NP and MM + NP + ES groups at a density of
300,000-500,000 cells/well. NP coverslips are made of glass coverslips
covered with a structured polyethylene glycol diacrylate polymer
(800 nm groove width, 800 nm ridge width, and 600 nm height) with a
surface stiffness of approximately 7 mPa^[295]25. Cells were maintained
in B27 medium for 6 days with medium changes every 2 days. On day 21,
all the non-B27 groups were switched from B27 medium to maturation
medium (MM, Supplementary Table [296]1) for 7 days, with medium changes
every 2 days. On day 28, the MM + NP + ES and MM + ES groups were
subjected to 2 Hz electric field stimulation (2 ms pulse duration, 8 V)
using a C-Pace EP (IonOptix) together with a 6-well C-dish (IonOptix)
for 14 days. On day 42 post differentiation, cells from all four groups
were harvested directly for analysis or dissociated, replated and
cultured for another 7 days for further analysis (Fig. [297]1a).
Patch clamp analysis
iPSC-CMs at day 42 from the four groups were dissociated using a
previously described method^[298]65,[299]66. Briefly, for MM + NP and
MM + NP + ES groups, NP coverslips with cells were transferred to a
3.5-cm dish and then treated for 10 min with 2 mL of 20 U/mL papain
(Sigma-Aldrich, 76220) dissolved in 1.1 mM EDTA-buffered B27 medium
containing 2.5 µM blebbistatin (Sigma-Aldrich, B0560). The cells were
gently centrifuged at 50 × g for 1 min. After aspirating the
supernatant, the cell pellet was gently resuspended in B27 medium
containing 2.5 µM blebbistatin and stored at 4°C until measured.
All automated patch-clamp experiments were performed at room
temperature (RT) using Patchliner Quattro with PatchControlHT software
(Nanion technologies GmbH) with low (for I[Na], I[K1], and I[Ca-L]) and
medium resistance (for I[to]) NPC-16 chips. The intracellular pipette
and extracellular bath solutions for I[Na], I[K1], I[Ca-L], and I[to]
are listed in Supplementary Table [300]4. To record I[Na], cells were
depolarised from a holding potential of −100 mV using voltage steps
from −80 to +70 mV for 20 ms in 5 mV steps. The sweep interval was 2 s.
Nifedipine (10 µM) was used to block I[Ca-L]. I[K1] was recorded using
test potentials of 2 s duration between −130 and 10 mV from a holding
potential of −40 mV. The sweep interval was 10 s. The protocol was
repeated in the presence of 0.5 mM BaCl[2] and the Ba^2+-sensitive
current was calculated as I[K1]. To record I[Ca-L], cells were
depolarised for 100 ms to voltages between −80 and +60 mV from a
holding potential of −90 mV, and the sweep interval was 3 s. I[to] was
recorded by increasing the test potential from −40 mV to +60 mV in
10 mV steps from a holding potential of −90 mV with a 20 ms pre-pulse
to −35 mV to inactivate I[Na]. Each pulse lasted for 600 ms, and the
sweep interval was 10 s. CdCl[2] (0.5 mM) was used to block sodium and
calcium currents.
For action potential (AP) recordings and I[Kr] measurements using
manual patch clamp technique, CMs from all four groups were used
directly after overnight recovery in cardio-digestion medium in order
to minimise the time that CMs from the MM + NP and MM + NP + ES groups
spent in non-NP and non-ES conditions. The pipette and extracellular
solutions used for AP and I[Kr] recordings are listed in Supplementary
Table [301]4. All manual patch-clamp experiments were performed at RT
using a ruptured whole-cell patch clamp with a HEKA EPC10 amplifier and
Patchmaster (HEKA Elektronik).
To assess resting membrane potential (RMP), maximum upstroke velocity
(Vmax), and action potential amplitude (APA), spontaneous APs were
recorded in Tyrode’s solution without current injection. To assess AP
duration (APD), a negative current was injected into the CMs to
maintain the RMP at approximately −80 mV prior to application of 0.5 Hz
pacing stimulation. Signals were filtered with 2.9 and 10 kHz Bessel
filters. At least 5 consecutive stable spontaneous APs and paced APs
were averaged to determine RMP, Vmax, APA and APD at 90% repolarisation
(APD[90]) using Fitmaster (HEKA Elektronik) and LabChart 8 software
(ADInstruments).
The holding potential of the I[Kr] recording was set at −50 mV. I[Kr]
was recorded using test potentials of 2.5 s duration from +40 to −40 mV
in 10 mV decrements. This was followed by a 4 s phase at −40 mV to
elicit the I[Kr] tail current. The pulse interval for each sweep was
10 s. I[Kr] was defined as the E-4031-sensitive current by subtracting
the current recorded after application of 1 µM E-4031 from the current
recorded before application.
Multi-electrode array
All multi-electrode array (MEA) recordings were performed using a
Maestro Edge equipped with AxIS Navigator software (Axion BioSystems)
at 37 °C, 5% CO[2] at a sampling rate of 12,500 Hz. Approximately
200,000 cells were resuspended in 20 µL of cardio-digestion medium and
seeded onto the electrode distribution area of Geltrex-coated CytoView
6-well MEA plates (Axion BioSystems). To evaluate the drug response,
iPSC-CMs were seeded into Geltrex-coated CytoView 24-well MEA plates
(Axion BioSystems) at a density of 25,000 cells/well. Around one hour
after seeding, 1 mL of cardio-digestion medium was gently added into
every well. iPSC-CMs were recovered for 7 days in the same medium used
during the maturation period. To avoid the influence of different media
during recording, iPSC-CMs of all conditions were incubated in MM for
one hour before starting measurements. MEA drug testing was performed
using a sequential addition protocol with concentration increments
after baseline activity recording in each well (Supplementary
Fig. [302]4). Vehicle controls for all conditions were performed on
each assay plate. After drug addition, cells were incubated at 37 °C,
5% CO[2] for 12 min, and the response was recorded for 2 min. The main
metrics including conduction velocity (CV), spike amplitude, spike
slope, inter-beat interval, and corrected field potential duration
(FPD[C], corrected by Fridericia’s formula) were further analysed using
AxIS Navigator, Cardiac Analysis Tool and AxIS Metric Plotting tool
(Axion BioSystems). In addition, we quantified the number of quiescent
and arrhythmic cultures after treatment with verapamil, E-4031 and
isoprenaline based on beating rate variation using Cardiac Analysis
Tool. Spontaneous beating frequency was defined as the reciprocal of
the averaged inter-beat interval. The mainstream CV values were
averaged for CV quantification.
Calcium transient measurement
iPSC-CMs at day 42 were dissociated, replated onto Geltrex-coated ø25
mm coverslips at a density of 200,000 cells per well of a 6-well plate
and recovered for 7 days in the respective media. For calcium transient
measurement using Fura-2^[303]67, cells were loaded with 2.5 µM Fura-2
(Thermo Fisher Scientific, F1221) in B27 medium (B27 group) and MM
medium (the other three groups) at 37°C for 30 min and washed twice
with the corresponding medium. Cells were incubated for 10 min to
achieve complete de-esterification of intracellular Fura-2. For calcium
transient measurement using Fluo-4, cells were loaded with 5 µM Fluo-4
AM (Thermo Fisher Scientific, [304]F14217) and 0.02% (w/v) Pluronic
F-127 (Thermo Fisher Scientific, P3000MP) at 37°C for 30 min and washed
twice. Intracellular calcium was recorded at 35 °C using a 40x
objective on an Olympus IX70 microscope equipped with the IonOptix
system with the IonWizard core software (IonOptix). Fura-2 stained
samples were excited at 340 and 380 nm with a switching frequency of
200 Hz and the emitted fluorescence was collected at 510 nm. Cytosolic
calcium levels were defined as the ratio of fluorescence intensity at
340 and 380 nm (340/380 nm). Fluo-4 stained cells were exited at 488 nm
and the emitted fluorescence was collected at 510 nm. Ca^2+ transients
were recorded in Tyrode’s solution containing (in mM): NaCl 138, KCl 4,
CaCl[2] 1.8, MgCl[2] 1, NaH[2]PO[4] 0.33, HEPES 10, and glucose 10 (pH
adjusted to 7.3 with NaOH). To normalise Ca^2+ transient frequency,
iPSC-CMs were field stimulated (6 V, 10 ms) at a pacing rate of 0.5 Hz
using a MyoPacer (IonOptix). To assess the calcium content, 10 mM
caffeine was applied to the CMs under 0.5 Hz pacing. Monotonic
transient analysis was performed using LabChart 8 software. For Fura-2
stained cells, diastolic and systolic Ca^2+ levels, Ca^2+ transient
amplitude (systolic level minus diastolic level), and decay rate (tau)
of Ca^2+ transients were determined. For Fluo-4 stained cells,
fluorescence intensities at baseline (F[0]) and transient peak (F[1])
were measured and changes in calcium levels (ΔF) were calculated (F[1]
minus F[0]). Normalised Ca^2+ transient amplitude (ΔF/F[0]), and tau
were determined.
Movie-based contraction analysis
Movie-based contraction analysis was performed on day 42 as previously
described^[305]5. Briefly, movies (1024 × 1024 pixel, 60 FPS, length
30 s, 2 movies per culture, 3 cultures per batch) were recorded using a
Hamamatsu Orca Flash 4.0 V3 camera (Hamamatsu) with 60 FPS, 1024 × 1024
pixels resolution. Movies were exported as MPEG4 files and analysed
using Maia motion analysis software (QuoData–Quality & Statistics GmbH)
using a block size of 10.7 µm (16 pixels), frameshift of 67 ms, and
maximum distance shift of 4.69 µm (7 pixels). Contraction and
relaxation peaks in the raw beating traces were assessed manually.
Western blot
On day 42, iPSC-CMs were scraped off, pelleted, snap-frozen in liquid
nitrogen and stored at −80 °C. Cells were lysed by homogenisation in
RIPA buffer (150 mM NaCl, 50 mM Tris, 1.0% NP-40, 0.5% sodium
deoxycholate, 0.1% SDS, 1 mM EDTA, 10 mM NaF, and 1 mM PMSF)
supplemented with inhibitors of proteases (cOmplete mini, EDTA-free,
Roche) and phosphatases (PhosSTOP, Roche) and incubated for 30 min at
4 °C with gentle rotation. Lysates were centrifuged at 14,000 rpm
(19,500 × g) for 20 min at 4 °C and protein concentration was
determined by BCA assay following the manufacturer’s instruction.
Proteins were subjected to SDS-PAGE and transferred to nitrocellulose
membranes. Membranes were blocked with 5% non-fat milk in TBS-T
overnight at 4 °C. Afterwards, the membranes were incubated with
primary antibodies overnight at 4 °C, followed by incubation with
horseradish peroxidase-conjugated goat anti-mouse or anti-rabbit
secondary antibodies (Supplementary Table [306]5) at RT for 1 h.
Proteins were visualised by chemiluminescence using the Super Signal
West Dura Chemiluminescent Substrate kit in combination with the Fusion
FX Spectra Imaging System (Peqlab) and images were analysed using
FusionCapt Advance software (Vilber).
Flow cytometry
Cells were singularised using collagenase and trypsin, fixed in 4%
paraformaldehyde (PFA) for 20 min at RT and stored in PBS containing 1%
BSA at 4 °C. For staining, iPSC-CMs were permeabilised in PBS
containing 1% BSA and 0.1% Triton-X for 10 min at RT. Staining was
performed with specific antibodies (Supplementary Table [307]5). cTNT
was detected using either mouse anti-cTNT (Thermo Fisher Scientific,
MS-295-P1) or directly coupled cTNT-APC (Miltenyi Biotec, 130-120-543).
For cTnT and Tom20 staining, cTNT-APC and anti-Tom20 (Santa Cruz,
sc-17764) antibodies were used. Negative controls were performed using
either the respective secondary antibodies (for samples detected with
the non-coupled primary antibodies) or isotype controls (for samples
detected with the directly coupled primary antibodies). After
incubation with primary antibodies, cells were washed with PBS
containing 1% BSA, followed by incubation with secondary antibodies,
and Hoechst 33342 (5 µg/mL). To assess EdU-incorporation, PFA-fixed
iPSC-CMs were incubated with mouse anti-cTNT antibody in Click-iT™
permeabilisation and wash reagent (Thermo Fischer Scientific,
[308]C10645) overnight, EdU click reaction was performed according to
manufacturer’s instructions, and DNA was stained with Draq5 (abcam,
ab108410, 10 µM). To determine the activity of Ki67, iPSC-CMs were
stained with antibodies cTNT-APC and Ki67-FITC (Miltenyi Biotec,
130-117-691) for 1 h at 4 °C. DNA was stained with Hoechst 33342.
Afterwards, cells were resuspended in PBS containing 1% BSA and
analysed on an LSRII or FACS Canto II flow cytometer using FACSDiva
software version 8.0.2 (BD Biosciences). At least 10,000 events were
recorded for each sample. Flow cytometry data were then analysed using
FlowJo v10.10 (BD Biosciences).
Immunofluorescence staining
For immunofluorescence staining, iPSC-CMs from the B27 and MM groups
were dissociated and replated onto Geltrex-coated ø25 mm glass
coverslips at day 28 at a density of 200,000 cells per well of a 6-well
plate. For staining of RYR2, α-actinin and Hoechst 33342, cells at day
42 were fixed with 4% PFA for 20 min and permeabilised with PBS
containing 1% BSA and 0.1% triton-X 100 for 10 min at RT. For detection
of Cx43 and Hoechst 33342, iPSC-CMs were fixed in methanol-acetone
(7:3, v/v, 10 min at −20 °C). After fixation, cells were washed with
PBS and incubated in PBS containing 1% BSA at 4 °C for at least 2 h.
Incubation with primary antibodies (Supplementary Table [309]5) was
performed overnight at 4 °C in PBS containing 1% BSA. After washing the
coverslips in PBS, samples were incubated with secondary antibodies and
Hoechst 33342 in PBS containing 1% BSA for 1 h at RT. Coverslips were
washed with PBS, deionised water and mounted onto glass slides using
Fluoromount-G mounting medium (Thermo Fisher Scientific). Imaging was
performed using LSM880 confocal microscope and ZEN software (Carl
Zeiss).
To analyse the colocalisation of RYR2 and α-actinin, samples were
imaged using an LSM880 confocal microscope (at 60x magnification).
Colocalisation was quantified using Cell Profiler v4.2.6 (Broad
Institute). For this, composite images were loaded, single channels
were separated using the ColorToGray module, colocalisation of the red
(RYR2) and green (α-actinin) channels was measured for the entire image
covering 2-4 cells using the MeasureColocalisation module, and data
were exported using the ExportToSpreadsheet function. Colocalisation
was measured for 3 independent experiments with 6 images per experiment
and conditions.
To study cell alignment, we imaged iPSC-CMs stained for α-actinin and
nuclei and the NP surface on the coverslip at different z levels using
confocal microscopy. The NP direction was defined as 0^o in the MM + NP
and MM + NP + ES groups, whereas 0^o was randomly defined in the B27
and MM groups. The z-discs and nuclei were identified using the
IdentifyPrimaryObjects module, and z-disk orientation (stained with
α-actinin) and the direction of the elongated nuclei (stained with
Hoechst 33342) were calculated with respect to 0^o using Cell Profiler
software.
Seahorse measurements
The Seahorse system with Wave Desktop software (Agilent) was used to
determine the metabolic activity of iPSC-CMs. Cells were singularised
using collagenase B and trypsin on day 42 and replated in 96-well
Seahorse assay plates (Agilent, 103792-100) at a density of 15,000
cells/well in cardio-digestion medium. The next day, the medium was
changed to the appropriate culture medium (B27 medium for the B27
group, MM for three other groups) and cells were recovered for 7 days
with medium changes every other day. Seahorse recordings were performed
using Seahorse Agilent XF Analyser (Agilent) according to the
manufacturer’s protocols. The Seahorse XF Cell Mito Stress Test Kit
(Agilent, 103010-100) with sequential addition of oligomycin, FCCP and
rotenone/antimycin A was used to study mitochondrial respiration. All
data were normalised to the total amount of protein per well after
lysis of cells in RIPA buffer.
Real-time PCR
On day 42, the cells (1 × 10^6 cells/well) were washed with ice-cold
PBS and scraped from the culture plate or NP coverslips. Two wells of
each condition were pooled as one sample that was centrifuged at
2000 × g at 4 °C. All steps were performed on ice. Cell pellets were
lysed in 1 mL TRIzol^TM (Thermo Fisher Scientific) and homogenised
through continuous pipetting (30x) using a metal syringe. Lysates were
centrifuged for 5 min at 12,000 × g at 4 °C, the clear supernatant was
transferred into a new tube, and 0.2 mL chloroform was added. Samples
were mixed thoroughly, incubated for 3 min on ice and centrifuged for
15 min at 4 °C and 15,000 × g to separate phases. The upper phase
containing the isolated RNA was carefully collected and mixed with
doubled volume of 95% ethanol. The solution was loaded on RNA columns
of RNeasy isolation kit (Qiagen), RNA was purified according to
manufacturer’s protocol, and eluted in RNAse-free water. Concentration
was determined using a Nanodrop^TM spectrophotometer (Thermo Fisher
Scientific). Synthesis of cDNA was performed using the iScript cDNA
Synthesis Kit (Bio-Rad) with 200 ng total RNA per reaction (20 µL)
according to the manufacturer’s protocol. For the detection of gene
expression, samples and specific primers (Supplementary Table [310]6)
were prepared using SsoAdvanced Universal SYBR Green Supermix (BioRad)
and real-time PCR was performed in Hard-Shell Optical 96-well plates
using the CFX96 Real-time PCR system (Bio-Rad). Initial denaturation
was done at 95 °C for 30 s, followed by 45 amplification cycles (15 s
at 95 °C, 1 min elongation at 60 °C). A melting curve was obtained over
the temperature range 65-95 °C. HPRT was used as a reference gene.
Efficacies of all primers were determined before use through serial
dilution of cDNA and the specificity based on gel electrophoresis as
well as melting curves. Primer efficacies were calculated using CFX
Manager software (Bio-Rad). For quantification of gene expression, we
used Eq. ([311]1) ^[312]68 to calculate relative expression of each
gene to HPRT.
[MATH:
Relativ<
mi>eexpression=<
mrow>1/(1+efficacy)Ct(targe<
/mi>t)1<
/mn>/(1+efficac
y)Ct(reference)
:MATH]
1
RNA sequencing and DEG analysis
iPSC-CMs from three independent differentiations of 2 iPSC lines
(iWTD2.1, isWT7.22) were used for RNA sequencing analysis. mRNA was
isolated from an average of 600 ng total RNA by poly-dT enrichment
using the NEBNext Poly(A) mRNA Magnetic Isolation Module (NEB)
according to the manufacturer’s instructions. Samples were then
directly subjected to the strand-specific RNA-seq library preparation
workflow (Ultra II Directional RNA Library Prep, NEB). Ligation was
performed using the NEB Next Adaptor from the NEB Next Multiplex Oligos
for Illumina Kit. After ligation, the adaptors were depleted by XP bead
purification (Beckman Coulter), where the bead solution was added to
the samples in a ratio of 0.9:1. Unique dual indexing was done during
the subsequent PCR enrichment (12 cycles) using amplification primers
carrying the same sequence for i7 and i5 index (Primer 1: AAT GAT ACG
GCG ACC ACC GAG ATC TAC AC NNNNNNNN ACA TCT TTC CCT ACA CGA CGC TCT TCC
GAT CT, Primer 2: CAA GCA GAA GAC GGC ATA CGA GAT NNNNNNNN GTG ACT GGA
GTT CAG ACG TGT GCT CTT CCG ATC T). After two further XP bead
purifications (0.9:1), the libraries were quantified using the Fragment
Analyser (Agilent). Libraries were sequenced on an Illumina NovaSeq
6000 in 100 bp paired-end mode with an average of 50 million fragments
per library.
FastQC ([313]http://www.bioinformatics.babraham.ac.uk/) was used to
perform a basic quality control of the resulting sequencing data.
Fragments were aligned to the human reference genome hg38 with the
support of the Ensembl 104 splice sites using the aligner STAR
(2.7.10b). Counts per gene and sample were obtained based on the
overlap of the uniquely mapped fragments with the same Ensembl
annotation using featureCounts (v2.0.1). Normalisation of raw fragments
based on library size and testing for differential expression between
the different conditions was performed using the DESeq R package
(v1.38.3). Sample to sample Euclidean distance, Pearson’ and Spearman
correlation coefficient (r) and PCA based on the top 500 genes with the
highest variance were computed to explore the correlation between
biological replicates and different libraries. To identify
differentially expressed genes (DEGs), counts were fitted to the
negative binomial distribution and genes were tested between conditions
using the Wald test of DESeq2 (two-sided). TPM values were generated
using Kallisto (v0.46.1).
Gene set enrichment analysis (GSEA) was performed based on normalised
count data using GSEA software v.4.3.2^[314]69 and canonical pathways
(CP) collection (C2, CP: c2.cp.v2023.2.Hs,
[315]https://www.gsea-msigdb.org/gsea/msigdb/human/genesets.jsp?collect
ion=CP) or TFT collection (C3, TFT: c3.tft.v2023.2,
[316]https://www.gsea-msigdb.org/gsea/msigdb/human/genesets.jsp?collect
ion=TFT), two Human Molecular Signatures Database (MSigDB) collections
provided by the Broad institute with following parameters: weighted
scoring, meandiv normalisation, max_probe mode, maximum gene set size
500 genes, minimum set size 15 genes, and 1000 permutations. GSEA
results were applied for the creation of enrichment maps using
Cytoscape software according to^[317]70. Briefly, nodes were included
based on false discovery rate (FDR) q-value < 0.2 and normalised
enrichment score (NES) ≥ 1.5 for upregulated and ≤ −1.9 for
downregulated genes. Edges cut-off was set to 0.375 and pathway cluster
names were determined after manual examination of all individual nodes
in clusters.
To benchmark the gene expression pattern of iPSC-CMs in the
MM + NP + ES group with human heart tissue, our own data and published
bulk RNA-seq datasets ([318]GSE62913) from foetal ventricles
([319]GSM1536186, [320]GSM1536187) and adult heart samples
([321]GSM1536192, [322]GSM1536193) have been were aligned using STAR to
the homo sapiens reference hg38 and using Ensembl annotation 104. A
list of 210 genes, including marker genes for ventricular CMs, ion
channels, and genes related to cell cycle activity and mitochondrial
development that were affected by ES in our study, was used for
comparison. Fragments were counted using FeatureCounts to assign
fragments to gene features. The counts table was processed using R and
the libraries DESeq2 (v1.38.3) and edgeR (v3.40.2). DEseq2 standard
methods were used to normalise the data. Edge methods were applied to
the normalised data to correct for possible effects of different
library preparation methods. The normalised and corrected data were
then visualised using the R library ggplot2.
Statistical analysis and reproducibility
Statistical analysis was conducted using R Studio v2024.04.2 (Posit
Software, PBC) by applying a linear mixed model (“lmerTest::lmer”
fitting of the raw values), considering experimental groups and cell
lines as fixed variables and independent batches as a random variable.
Residuals were evaluated for normal distribution based on plots of
residuals vs. fitted values and Shapiro-Wilk test (check_normality
function), and for homogeneity with Bartlett’s test (check_homogeneity
function). The analysis using the linear mixed model followed by
pairwise comparisons (using “emmeans(~group) %>% pairs()”) represents a
two-sided statistical test. In detail, the function emmeans (‘Estimated
marginal means’) estimates the mean values per group. These are
forwarded to the pairs function, which performs the pairwise
comparisons of the mean values. If more than 2 groups are analysed,
Tukey’s post-test is performed to correct for multiple comparisons. In
case of 2 groups, significance was determined using a two-sided
standard t-test.
If the mixed model could not be applied due to violation of normality
or homogeneity, statistical analysis was performed using Kruskal-Wallis
test with Dunn’s multiple comparison test (comparison of > 2 groups),
two-sided Kolmogorov-Smirnov (comparison of 2 groups), or two-way ANOVA
with Sidak’s post-test for multiple comparisons using GraphPad Prism
10, indicated in figure legends. Results were considered statistically
significant at p < 0.05.
Reporting summary
Further information on research design is available in the [323]Nature
Portfolio Reporting Summary linked to this article.
Supplementary information
[324]Supplementary Information^ (1.7MB, pdf)
[325]41467_2025_58044_MOESM2_ESM.pdf^ (84.3KB, pdf)
Description of Additional Supplementary Files
[326]Supplementary Data 1^ (59.8KB, xlsx)
[327]Supplementary Data 2^ (46.6KB, xlsx)
[328]Supplementary Data 3^ (40.2KB, xlsx)
[329]Supplementary Movie 1^ (14MB, mp4)
[330]Reporting Summary^ (7.4MB, pdf)
[331]Transparent Peer Review file^ (2.5MB, pdf)
Source data
[332]Source Data^ (853.2KB, xlsx)
Acknowledgements