Optical modulation of excitation-contraction coupling in human-induced pluripotent stem cell-derived cardiomyocytes

Summary Non-genetic photostimulation is a novel and rapidly growing multidisciplinary field that aims to induce light-sensitivity in living systems by exploiting exogeneous phototransducers. Here, we propose an intramembrane photoswitch, based on an azobenzene derivative (Ziapin2), for optical pacing of human-induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs). The light-mediated stimulation process has been studied by applying several techniques to detect the effect on the cell properties. In particular, we recorded changes in membrane capacitance, in membrane potential (Vm), and modulation of intracellular Ca2+ dynamics. Finally, cell contractility was analyzed using a custom MATLAB algorithm. Photostimulation of intramembrane Ziapin2 causes a transient Vm hyperpolarization followed by a delayed depolarization and action potential firing. The observed initial electrical modulation nicely correlates with changes in Ca2+ dynamics and contraction rate. This work represents the proof of principle that Ziapin2 can modulate electrical activity and contractility in hiPSC-CMs, opening up a future development in cardiac physiology.


INTRODUCTION
In the cardiovascular field, optical stimulation is emerging as an alternative to traditional approaches for many research and therapeutic applications thanks to a series of key-enabling features: the lower energy consumption and release, and minimal invasiveness net of an extraordinary spatial and temporal resolution. [1][2][3] Optogenetics, 4-9 that has become widespread in neuroscience, could in principle be relevant in the cardiac field too. [10][11][12][13][14][15][16][17][18][19][20][21][22] However, the approach still has a limited clinical applicability mainly because cell optical sensitivity is obtained by transduction with gene constructs carried by viral vectors.
A possible alternative strategy to overcome these constraints relies on the use of light-sensitive transducers, based on both inorganic and organic semiconductors. [23][24][25][26][27][28][29][30][31][32] Those have been used recently as photoactive interfaces for cardiomyocyte (CM) optical stimulation. The generation of action potentials (APs) has been demonstrated by using planar graphene-based biointerfaces, possibly exploiting the photogeneration of charge carriers, and with polymer-silicon nanowires utilizing their photoelectrochemical properties. 25,30 Similar results were obtained also increasing the local temperature with illumination of absorbers in contact with CMs, using gold nanoparticles or nanorod electrodes and metasurface planar organic interfaces. 31,33,34 Organic semiconductors in particular allow to establish almost seamless abiotic/biotic interfaces, as they are soft materials that support both ionic and electronic transport, alike many biological molecules. 35 In this context, the triggering mechanism leading to cell activity upon light absorption can be capacitive, faradaic, or thermal. [36][37][38] An alternative approach, still based on organic molecules, exploits photochromic compounds that can be covalently bound [39][40][41] to an ion channel or not-covalently 42,43 to the plasma membrane. This methodology is gaining increasing interest due to the stimulation efficiency, its versatility, and the possibility of using two 1 ll OPEN ACCESS photon absorption that pushes the stimulation wavelengths to the near-infrared spectral region, improving tissue penetration. 44,45 In this work, we propose a new tool for optical pacing of human-induced pluripotent stem cell-derived CMs (hiPSC-CMs) exploiting Ziapin2, a recently synthetized intramembrane photochromic transducer that we successfully tested in bacteria, non-excitable cells, and neurons. [46][47][48][49][50] Ziapin2 has affinity for the hiPSC-CMs sarcolemma, and once partitioned, it undergoes trans-dimerization, which in turn leads to increased capacitance as the result of reduction in membrane thickness. Upon millisecond pulses of visible light, trans/cis isomerization causes a fast drop of capacitance due to membrane relaxation, resulting in a transient hyperpolarization followed by a delayed depolarization that triggers AP generation. The increase in AP frequency nicely correlates with changes in Ca 2+ dynamics and contraction rate, thus proving that Ziapin2 modulates excitation-contraction (E-C) coupling at a whole extent.

Ziapin2 photoisomerization modulates hiPSC-CMs membrane potential
The molecule was tested at two different concentrations (5 and 25 mM). In both cases, once added to hiPSC-CMs cultures, Ziapin2 successfully partitioned into the sarcolemma causing a significant increase in capacitance ( Figure S1A). The effect was concentration dependent, as we observed a +23% rise (14.2 G 1.2 pF vs 17.5 G 0.9 pF; p = 0.05089) for 5 mM and +56% (14.2 G 1.2 pF vs 22.2 G 1.3 pF; p = 0.00063) for 25 mM, while no alterations were detected in hiPSC-CMs incubated with the vehicle (DMSO, data not shown). At equilibrium, this phenomenon changes the membrane electrical time constant (RC) but not the resting potential. According to previous experimental evidences, 46,48,49 the modulation has been ascribed to the thinning of the bilayer caused by the dimerization of Ziapin2 molecules inside the membrane.
Upon photostimulation, the azobenzene isomerization removes the geometric constraint of the dimer leading to membrane relaxation and a partial return toward steady-state capacitance values with a reduction of À14% (17.5 G 0.9 pF vs 15.0 G 0.7 pF; p = 0.08) and À12% (22.2 G 1.3 pF vs 19.5 G 1.1 pF; p = 0.14) for 5 and 25 mM Ziapin2, respectively ( Figure S1B).
Whole-cell patch clamp experiments in current clamp mode (I = 0) revealed that the photoinduced drop in capacitance correlates with a modulation of the membrane potential. In particular, the Ziapin2-mediated photostimulation (with 20 and 200 ms single light pulses) results in a transient hyperpolarization that is occurring within few ms after the light onset, consistent with the linear inverse relation between capacitance and voltage. The recovery of the potential by capacitive currents is followed by a depolarization which is often large enough to generate an AP (see representative traces in Figure 1A).
Both hyperpolarization and depolarization peaks display much higher amplitude at 25 mM concentration than 5 mM ( Figure 1B). For instance, with 20 ms light stimulation, hyperpolarization is six times larger at 25 mM concentration, even if the relative change in capacitance is comparable (À14% vs À12% for 5 and 25 mM Ziapin2, respectively, see Figure S1B). This is surprising, since according to the simple plane capacitor model, if there are no other changes in the circuit, DV/V = DC/C. To explain this discrepancy, we conjecture that equal amplitude, but faster changes in capacitance, could give rise to larger changes in membrane potential. This is confirmed by a numerical simulation based on the equivalent circuit model for the membrane ( Figure S2). To account for the different kinetics, we propose that at higher concentration, Ziapin2 dwells in a more heterogeneous membrane environment, characterized by different viscosity, order, and phospholipid polarity, where a faster isomerization might occur.
In agreement with capacitance data, no light-dependent effects were noticed in vehicle-treated hiPSC-CMs ( Figure 1).

Ziapin2 photoisomerization induces AP firing in hiPSC-CMs
To assess the impact of the molecule on hiPSC-CMs electrical activity, we evaluated the ability of light stimulation to generate APs in hiPSC-CMs loaded with Ziapin2. Remarkably, the most striking effect was ll OPEN ACCESS iScience 26, 106121, March 17, 2023 iScience Article observed in CMs exposed to the highest photochromic concentration (72 G 4% of responding cells for Ziapin2 25 mM vs 37 G 3% for Ziapin2 5 mM, Figure 2A).
As a next step, we compared spontaneous and light-evoked AP features. Since the most significant effects occurred in hiPSC-CMs loaded with Ziapin2 25 mM and subjected to 20 ms light stimuli, here we report only data within these experimental conditions.
In dark conditions, no significant differences were measured in maximum diastolic potential ( Figure S3A) and maximal upstroke velocity (dV/dt max , Figure S3B), while AP amplitude (APA, Figure S3C) was slightly decreased in Ziapin2-exposed CMs (70.3 G 1.2 mV vs 70.8 G 2.9 mV; p = 0.04). Ziapin2 photoisomerization did not exert any detectable modification of these AP parameters ( Figures S3B and S3C), while a light-triggered increase in the number of AP per unit time was observed (0.14 G 0.02 Hz vs 0.27 G 0.02 Hz for spontaneous and light-evoked, respectively; p < 0.0001, Figure 2B). This light-induced frequency increment was associated with a shortening of the AP duration; plotting the SD of these two parameters, it was possible to notice a greater clustering of both the indicators upon illumination ( Figure 2C). This suggests that Ziapin2 can potentially play a pivotal role in making homogeneous activity of cardiac cells and tissues by reducing the overall beating variabilities. Interestingly, a punctual AP generation was obtained when a stimulation train of 20 ms light pulses at 0.5 Hz was applied, observing just few failures under rate-controlled conditions ( Figure 2D). We attributed the missed stimulations to the low temperature (24 C) used in our experiments. It is indeed known that a more physiological temperature would speed up the dynamics of cellular processes. Accordingly, we monitored the electrical activity of hiPSC-CMs at 37 C ( Figure S4). As expected, The experiments were carried out at room temperature (24 C). Data were collected from three independent differentiations and are represented as mean G SEM *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001. iScience Article in these conditions, there was a higher spontaneous electrical activity ($0.78 Hz, Figure S4A, left panel) that increased upon Ziapin2-mediated photostimulation ($0.95 Hz, Figure S4B, left panel). Notably, considering a sampling time of 2 s, we observed that the APs under control conditions are randomly distributed ( Figure S4A, right panel), while upon photoexcitation they cluster at the light onset ( Figure S4B, right panel). This means that the spontaneous APs are generated in deterministic way and that we are imposing a 0.5 Hz stimulation frequency.
From this small subset of data, we also extrapolate the AP duration at 90% of repolarization (APD 90 , directly correlated with the frequency increase) and APA. In Figure S4C are reported data acquired at both 24 C and 37 C to show that, as expected, the increase in temperature mainly affects APD 90 , leaving the amplitude unaltered.

Ziapin2 photoisomerization elicits Ca 2+ dynamics in hiPSC-CMs
In CMs, the AP triggers the E-C coupling, namely the physiological process that converts an electrical stimulus to a mechanical response. It is well established that the contraction of cardiac cells is triggered by a considerable increase in intracellular Ca 2+ concentration. We thus evaluated Ziapin2 effect on cytoplasmic Ca 2+ levels performing fluorescence-based measurements with the widely used ratiometric Ca 2+ indicator Fura-2 AM ( Figure 3A).
To monitor if the molecule photoisomerization had an impact on Ca 2+ dynamics, we first measured the spontaneous Ca 2+ activity by keeping Ziapin2-loaded hiPSC-CMs in dark conditions, and subsequently, iScience Article we subjected the same cells to a single light stimulation pulse using an arc lamp light source coupled with a high-speed OptoScan monochromator that allowed us to accurately select the excitation wavelength.
In accordance with the patch-clamp data, we observed a significantly higher frequency of Ca 2+ transients (CaTs) (+15.4%, p = 0.0001, Figure 3B) when hiPSC-CMs were subjected to light stimulation while no significant changes were detected in the amplitude (p = 0.6, Figure 3C), in the rise-time (p = 0.93, iScience Article Figure 3D), or in the decay kinetics (p = 0.67, Figure 3E) of the CaTs. Moreover, the diastolic Ca 2+ levels were also comparable among the two groups ( Figure 3F).
Since changes in the shape of the CaT might reflect alterations in cardiac ion channels or cardiac signaling pathways, we can speculate that is unlikely that a primary membrane transporter like Na + /Ca 2+ exchanger (NCX) or the plasma membrane Ca 2+ ATPase is altered upon Ziapin2 photoisomerization, suggesting that the molecule acts by fully respecting the physiological behavior of the cell.

Ziapin2 photoisomerization enhances hiPSC-CMs contraction rate
As a final step, we investigated the effect of Ziapin2 on hiPSC-CMs contraction behavior by exploiting a custom MATLAB code that allowed us to obtain a quantitative analysis from previously acquired highspeed movies (Figure 4).
In dark condition, Ziapin2-loaded hiPSC-CMs had a contraction rate of 0.13 G 0.006 Hz. Under 1 Hz pulsed illumination, the rate increased by 59% during the entire acquisition window ( Figure 4B), while it remained constant in vehicle-treated hiPSC-CMs subjected to the same stimulation protocol ( Figure S5). The augmented frequency occurs net of variations in contraction amplitude (calculated as area reduction, p = 0.74, Figure 4C) and contraction velocity (extracted as derivative of the contraction amplitude, p = 0.57, Figure 4D). Interestingly, the effect on the contraction frequency persisted in the 25-30 s following light offset, albeit with a 27% decrease (from 0.22 G 0.006 Hz to 0.16 G 0.0059 Hz, Figure 4B). There are two iScience Article possible explanations for this; first, the experiments were performed at 24 C, a temperature far from the canonical physiological 37 C that could have slowed down the cell behavior. 29 Secondly, even though the hiPSC-CMs are a consolidated tool and hold an enormous potential in cardiovascular research, their E-C coupling process resembles more the one of fetal/neonatal CMs, thus it is possible that the Ca 2+ clearance machinery in these cells is substantially slower in comparison to adult CMs. 51,52 Remarkably, the lower the spontaneous hiPSC-CMs contraction rate, the higher the light-induced frequency increase. To better characterize the involvement of Ziapin2 in the light-induced increase in contraction frequency, we measured the latency, calculated as the interval between the onset of stimulation and the contraction of the CMs, in a small subset of quiescent cells ( Figure 4E).
The timescale was found to be distributed around 500 ms in a non-stochastic way, compatibly, at least in these experimental conditions, with expected times for the electromechanical coupling, thus confirming the link between optical stimulus and increased contraction.

Conclusions & perspectives
The possibility to manipulate the cellular activity with a precise spatial and temporal punctuality is a key enabling technology for biology and medicine. In this work, we report the first proof-of-principle demonstration of Ziapin2 as a light-sensitive tool to control cardiac cell excitability and contractility. Our approach is innovative for a series of reasons: i) unlike electrical stimulation, it enables a non-contact and reversible excitation with high spatiotemporal precision; ii) it avoids genetic manipulation and the pitfalls related to the introduction of exogenous genetic material using viral vectors, thus overcoming the major drawback of optogenetics; iii) Ziapin2 primarily targets the cellular passive properties without affecting ionic conductances, an issue shared by both optogenetics and photoswitchable ligands, still allowing a millisecond control of V m ; iv) the molecule effect is heatless, thus preventing increases in temperature that could be potentially harmful to the cells over the long term.
Ziapin2 could be used as a non-invasive optical tool to control cardiac electrical activity for bio-hybrid robotics studies (in which CMs are often the biological substrate), thus enabling or simplifying design and fabrication of the actuators. [53][54][55] Also, the possibility to pattern light in cardiac organoids or microtissues will allow to address fundamental questions of cardiac biology.

Limitations of the study
Further studies are needed to elucidate the biophysical mechanisms ruling the photostimulation process. In particular, future research will be devoted to the study of the triggering mechanism that leads to AP generation and to exclude the potential involvement of membrane transporters like NCX, Na + or K + -ATPase, as well as membrane receptors (e.g., b-adrenergic receptors). Moreover, some chemical improvement should be faced to increase the molecular absorption cross section leading to a reducing of the required light power density. Similarly, to further corroborate our findings, we will extend this approach to fully differentiated adult cardiomyocytes, either human-or mouse-derived. This will be an important step to show that our molecules do work in a more complex and physiological scenario.
If successful, this could open a broader field of applications for light-mediated clinical treatment of cardiac arrhythmias. In this context, new curative strategies that are less invasive, long-lasting, and less power consuming than the existing ones are a conditio sine qua non to reduce the burden for the recipients.

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following:

ACKNOWLEDGMENTS
The authors would like to thank Prof. Antonio Zaza for the interesting and inspiring discussions on this work.

DECLARATION OF INTERESTS
The authors declare no competing interests. iScience

Lead contact
Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Francesco Lodola (francesco.lodola@unimib.it).

Materials availability
This study did not generate any unique new reagents.
Data and code availability d All the data reported in this paper will be shared by the lead contact upon request. d All original code has been deposited to Github and is publicly available as of the date of publication. The DOI is listed in the key resources table.
d Any additional information required to re-analyze the data reported in this paper is available from the lead contact upon request.

Cell culturing and differentiation
Two different batches of hiPSC-CMs were used for this study, a commercial one (Pluricyteâ CMs provided by Ncardia, Gosselies, Belgium) and a laboratory handle one. The latter was produced by directed cardiac differentiation of previously generated control hiPSCs. 29,56,57 In brief, cardiac induction and generation of hiPSC-CMs were obtained as previously reported, using a chemically-defined serum-free protocol, which is based on activation (CHIR99021) and inhibition (IWR1) of the Wnt pathway in RPMI-B27 medium. [58][59][60] In this work, CMs were differentiated from two different control iPSC lines (one male and one female) between the 18 th and the 35 th passage in culture. Importantly, employed cell lines were regularly tested for being free of major chromosomal abnormalities by karyotype analysis. CMs were used for the experiments 25-30 days after spontaneous contracting activity was started, a differentiation stage at which they expressed the repertoire of sarcomeric proteins, calcium regulators, and ionic channels necessary for their correct functionality. Purity of differentiated CM populations was regularly checked before each experiment to be greater than 90% (not shown). The commercial batch was thawed and cultured following the manufacturer's protocol. In terms of plating density, cells for both batches were seeded onto the material at a confluence comprised between 50% and 60% on fibronectin coated (15 mg/ml in PBS buffer solution) 18-mm round glass coverslip (VWR, Radnor, USA). The cells were maintained in incubator at 37 C and 5% CO 2 .

METHOD DETAILS
Ziapin2 synthesis and uptake process Ziapin2 was synthetized as previously described. 46,48,49 Micromolar concentrations (5 or 25 mM) of the compound were added to hiPSC-CMs cultures directly into the culture medium. Subsequently the cells were placed in the incubator at 37 C and 5% of CO 2 . After 7 minutes the medium was gently washed out and the petri dish was rinsed with fresh extracellular solution.

Capacitance recordings
Capacitance measurements were performed as previously described. 49,61 Briefly, a double sinusoidal voltage clamp signal was applied to the cell in whole-cell configuration. The response current signal was acquired, and membrane capacitance and resistance were extracted fitting the current with a custom MATLAB program. The capacitance value was extracted in dark condition and during light stimulation (20 ms pulse). The shorter pulse was used in order to consider mainly the effect related to Ziapin2 photoisomerization. The following parameters were extracted from each CaT: diastolic Ca 2+ (CaD), CaT amplitude (CaT Amp), CaT rise-time (t peak ) and CaT decay kinetics (t decay ).

Acquisition and analysis of hiPSC-CMs contractile behavior
The Nikon Eclipse Ti inverted microscope described before was used to stimulate (Lumencor Spectra X, l ex = 470 nm, 20X objective, Pobj = 30 mW/mm 2 ) and acquire video of the CMs contraction a frequency. An objective with a lower magnification was used to increase the amount of recorder cells resulting in a more robust analysis. Consequently, the resulting power density was lower due to the bigger area.
Contractile behavior was analyzed using a custom-built algorithm, implemented in MATLAB R2020a. The approach is based on the contraction-induced retraction of cell body towards nucleus and is effective even if there is no displacement of cell extremities or if the cellular edge cannot be detected properly.
The user defines one or more regions of interest (ROI) using bounding boxes. Each bounding box should contain one single cell. For each ROI a set of features is identified and tracked across video frames by means of Kanade-Lucas-Tomasi algorithm. 62,63 Since the ROI is delimiting a cell, the tracked features are expected to belong to cell body. Averaging the estimated motion fields of all these feature points returns a mean geometric transformation, which can be applied to the bounding box delimiting the ROI. The area of this bounding box is measured over time and presents minima at cellular contractions. The number of minima per time interval yields an estimate of the cell contraction rate. Contraction amplitude is measured as percent reduction of bounding box area at contraction and averaged over all contractions within a specific running observation window (e.g., pre). The box area derivative when cells contract evaluates cell contraction velocity. Data were analyzed with Origin 9.0 (OriginLab Corporation).

QUANTIFICATION AND STATISTICAL ANALYSIS
Data were expressed as mean G SEM. Normal distribution was assessed using D'Agostino-Pearson's normality test. To compare two sample groups, either the Student's t-test or the Mann-Whitney U-test was used. To compare more than two sample groups, one-way ANOVA or Kruskal-Wallis were used. p < 0.05 was considered statistically significant. The analysis was carried out using Origin (OriginLab) and Prism (GraphPad).