Anion channelrhodopsins for inhibitory cardiac optogenetics

Optical control of the heart muscle is a promising strategy for cardiology because it is more specific than traditional electrical stimulation, and allows a higher temporal resolution than pharmacological interventions. Anion channelrhodopsins (ACRs) from cryptophyte algae expressed in cultured neonatal rat ventricular cardiomyocytes produced inhibitory currents at less than one-thousandth of the light intensity required by previously available optogenetic tools, such as the proton pump archaerhodopsin-3 (Arch). Because of their greater photocurrents, ACRs permitted complete inhibition of cardiomyocyte electrical activity under conditions in which Arch was inefficient. Most importantly, ACR expression allowed precisely controlled shortening of the action potential duration by switching on the light during its repolarization phase, which was not possible with previously used optogenetic tools. Optical shortening of cardiac action potentials may benefit pathophysiology research and the development of optogenetic treatments for cardiac disorders such as the long QT syndrome.


Results
We expressed the transmembrane domains of GtACR1, GtACR2 and Arch fused to C-terminal enhanced yellow fluorescence protein (EYFP) under the mouse ubiquitin C promoter in cultured neonatal rat ventricular cardiomyocytes (NRVMs) using lentiviral delivery. The position of the maximal absorption wavelength of GtACR1 (515 nm) is red-shifted compared to that of GtACR2 (470 nm), which is advantageous because it allows using light stimuli of higher penetration depth. Therefore, we chose GtACR1 for comparison with Arch. Expression of both constructs in cardiomyocytes was confirmed by immunofluorescent microscopy (Fig. 1).
First, we characterized photocurrents generated by Arch and GtACR1 by whole-cell voltage clamp of individual NRVMs. Figure 2a,b show series of hyperpolarizing photocurrents recorded from cells expressing Arch and GtACR1, respectively, evoked by a 1-s light pulse of varied intensity. The mean peak amplitude of GtACR1 Scientific RepoRts | 6:33530 | DOI: 10.1038/srep33530 photocurrents was ~12 fold greater than that of Arch currents, although the expression level of the former, assessed by measuring the tag fluorescence, was ~3 fold lower (Fig. 2c). The current rise and decay rates are shown in Fig. 2d. The membrane hyperpolarization is proportional to the amount of charge transported across the membrane. Figure 2e shows the dependence of charge transported during 1 s of illumination on the light intensity measured in cells expressing Arch or GtACR1. Almost 4 orders of magnitude less light was needed to reach the same value for GtACR1 than for Arch. The amplitude of Arch currents showed a weak voltage dependence, whereas that of GtACR1 currents was very steep (Fig. 2f), making the latter protein a dynamic hyperpolarizing tool at membrane potentials above the Nernst equilibrium potential for Cl − .
To test the performance of Arch and GtACR1 as tools for inhibition of cardiomyocyte function without perturbing the intracellular ion concentrations, we used extracellular recording from star-shaped aggregates of synchronously beating cells in growth medium. Expression of either Arch or GtACR1 did not change the frequency of APs in the dark, which was 67 ± 3 (mean ± sem; n = 11 cell clusters) and 68 ± 3 (mean ± sem; n = 29 cell clusters) APs/min, respectively, compared to 65 ± 2 (mean ± sem; n = 6 cell clusters) for control (uninfected) NRVMs. In cells expressing Arch switching on the light resulted in immediate suppression of electrical activity, EYFP tag fluorescence is shown in the left column (green channel), the cardiomyocyte marker (MYBPC3, myosin binding protein C-3) fluorescence is shown in the middle column (red channel), and merged images are shown in the right column.
but it was resumed at irregular intervals during a 30-s illumination period even at the highest intensity used (6.7 mW mm −2 ; Fig. 3a). In cells expressing GtACR1 illumination caused a decrease in AP amplitude; 50% inhibition was observed at ~10 μ W mm −2 (Fig. 3b). Figure 3c summarizes the results obtained with Arch-and GtACR1-expressing cells. In control (uninfected) NRVMs both the AP frequency and amplitude were unaffected by light even at the maximal intensity: 96 ± 3% and 108 ± 6%, respectively, as compared to the previous dark  The light dependence of AP number during the illumination period relative to the previous dark period (left axis, empty symbols) and AP amplitude (right axis, filled symbols) cells expressing Arch (black) and GtACR1 (red). Error bars, sem (n = 5-9 cells).
period (mean ± sem; n = 6 cell clusters). Normal electrical activity of the GtACR1-expressing culture restored within 2.0 ± 0.4 s at 7.7 mW mm −2 (mean ± sem; n = 6 cell clusters) and 1.5 ± 0.2 s at 0.7 mW mm −2 (mean ± sem; n = 5 cell clusters) after switching off the light of the maximal intensity. In Arch-expressing NRVMs no significant increase in AP frequency was observed after switching off the light relative to the preillumination period, in contrast to the results obtained in cardiomyocytes electrically coupled to Arch-expressing human embryonic kidney (HEK293) cells 13 .
GtACR2 also generated robust photocurrents in NRVMs (Fig. 4a), but more light was required to achieve the same degree of AP inhibition as compared to GtACR1-expressing cells (Fig. 4b, hatched bars). The decay rate of GtACR2-generated photocurrent at 0 mV was 0.019 ± 0.004 ms −1 (n = 5 cells), i.e., ~2 fold higher than that of GtACR1-generated currents (Fig. 2d). We also compared AP inhibition in GtACR1-and GtACR2-expressing cells by patch clamp recording in the current clamp mode. For both proteins, the light sensitivity probed by patch clamping was higher than that determined noninvasively by extracellular recording (Fig. 4b, solid bars), which can be explained by a steeper Cl − gradient used in patch clamp experiments. Measurements of the intracellular Cl − concentration in the heart muscle by ion-selective microelectrodes yielded values from 14 to 20 mM depending on the animal species, cell type and developmental stage [15][16][17] . We tested the dependence of the photocurrent amplitude and the reversal potential (E rev ) on the Cl − concentration in the pipette. When it was increased from 4 to 40 mM, the E rev followed the Nernst equilibrium potential for Cl − , as expected, which resulted in ~2-fold reduction of the photocurrents ( Supplementary Fig. S1).
Besides complete silencing of cardiac electrical activity, optogenetics can potentially be used for shaping AP morphology 18 . In particular, changing the AP duration (APD) is desired for probing and potentially curing pathological conditions resulting from APD abnormalities, such as the life-threatening long QT syndrome, which results from an abnormally slow repolarization phase of the AP 19 . To explore the possibility of optical shortening of the APD, we delivered light pulses at different times during an ongoing spontaneous AP using threshold-based closed-loop control. Expression of GtACRs did not significantly influence the APD at 50% repolarization (APD50) measured in the dark (272 ± 34, 270 ± 32 and 269 ± 21 ms in control, GtACR1-and GtACR2-expressing cells, respectively; n = 5-7 cells). When light was switched on during the depolarization phase, it resulted in cutting off the nascent AP ( Supplementary Fig. S2), as was previously observed using NpHR 20 . Moreover, GtACRs permitted temporally precise truncation of the AP at any time point during the repolarization phase ( Fig. 5a and Supplementary Fig. S2), which could not be achieved with NpHR 20 . Variation of the light intensity allowed fine tuning of the AP shape (Fig. 5b,c). The AP amplitude in experiments in which the light was switched on during the repolarization phase was 100 ± 0.8% (mean ± sem, n = 13 recordings) as compared to that measured in the dark.

Discussion
In previous studies the proton-pumping rhodopsin Arch was used to suppress cardiomyocyte electrical activity via cell delivery, i.e. by coupling of Arch-expressing HEK293 cells to cardiomyocytes 13 . We showed that AP generation could be suppressed immediately after switching on the light upon lentiviral expression of Arch in cardiomyocytes, but it resumed at irregular intervals during the 30-s illumination period even at the highest light intensity used (6.7 mW mm −2 ). Expression of the chloride-pumping rhodopsin NpHR enabled inhibition of individual cardiac APs by short light pulses of lower intensity, but the activity was not monitored for longer illumination periods 20 . For a complete suppression of AP generation during several seconds with NpHR a light intensity of 30 mW mm −2 was required 9 . GtACRs generated much larger currents and permitted AP inhibition at much lower light intensities than Arch. In addition to conducting many ions per photocycle in contrast to only one transported by rhodopsin ion pumps such as Arch or NpHR, ACRs also become more efficient upon the membrane depolarization (e.g. during the depolarization phase of an AP) because of the steep voltage dependence of their currents. Unlike rhodopsin ion pumps, ACRs transport ions passively, i.e., they generate hyperpolarizing currents only above the Nernst equilibrium potential for Cl − . This means that their efficiency as optogenetic silencing tools will depend on the intracellular Cl − concentration which may vary between species, cell types and other conditions. Nevertheless, currents generated by GtACR1 at 40 mM Cl − in the pipette (which is twice as high as the intracellular Cl-concentration measured in the heart with ion-selective microelectrodes 15 ) were still ~5 fold larger than maximal Arch photocurrents. Furthermore, our data obtained by extracellular recording unambiguously demonstrated that complete inhibition of cardiomyocyte electrical activity by light could be achieved at physiological intracellular Cl − concentrations.
The potentially most attractive application of GtACRs in optogenetics is their use for temporally precise truncation of APs during the repolarization phase. In a previous study NpHR failed to deliver this benefit 20 . Moreover, shortening of the AP duration observed when the light pulse was delivered during the depolarization phase was accompanied by a reduction of the AP amplitude 20 . Therefore, the results obtained with NpHR are better described as AP inhibition rather than AP shaping.
Shortening of the AP duration to treat the long QT syndrome appears to be a therapeutic therapeutic goal for which ACRs are particularly suitable. Using intensity-modulated rather than continuous light stimuli may help recreate more realistic morphology of shortened APs. Although continuous illumination of ACR-expressing cardiomyocytes led to a decrease in the AP amplitude rather than frequency (Fig. 3c), potentially closed-loop optogenetics can also be used to reduce the frequency by imposing a desired refractory period after each AP, which can lead to the development of novel therapeutic treatments for tachyarrhythmias.
A disadvantage of GtACRs as compared to rhodopsin ion pumps is a slower decay rate of their photocurrents. It is of little importance when a complete inhibition of cardiac electrical activity for an extended period of time is required (i.e., for analysis of the heart conduction system as in ref. 9). GtACR2 exhibits ~2 fold faster photocurrent decay compared to GtACR1 and is therefore more suitable for the latter goal. Screening of natural ACR variants from other cryptophyte algae and/or molecular engineering of known ACRs is desirable for improving their temporal resolution.

Methods
Expression constructs. The polynucleotide sequences encoding the seven-transmembrane (7TM) domains of GtACR1 (295 amino acid residues), GtACR2 (291 amino acid residues) and Arch (a gift of E.S. Boyden; 258 amino acid residues) were subcloned to the pFUGW vector backbone 21 in frame with an enhanced yellow fluorescence protein (EYFP) tag and verified by sequencing.
Neonatal rat ventricular cardiomyocyte isolation and transduction. Cardiomyocytes were isolated following the procedure as previously described 22 . In accordance with the animal protocol approved by the Animal Welfare Committee at University of Texas Health Science Center at Houston, 1-2 day old Sprague-Dawley rat pups were euthanized by decapitation. Hearts were exposed, excised, minced, and digested through a series of agitations in buffer containing collagenase type II (Worthington, Collagenase type II 305 U/mg) and pancreatin (Sigma). Cells from each digestion were collected and pooled, and pre-plated on Nunc plates (Thermo Fisher Scientific) in Complete Medium (40% DMEM, 40% Ham's F-10, 20% fetal calf serum) for 2-4 hours to reduce fibroblasts and enrich for cardiomyocytes. Following the pre-plate step, the non-adherent cells (cardiomyocytes) were aspirated from the plate, pelleted, resuspended in Complete Medium, and plated on coverslips or MatTek tissue culture plates (fibronectin-coated; Roche Applied Science). 36 hours later cardiomyocytes were washed with Ham's F-10 and incubated with lentiviral preparations of Arch, GtARC1 or GtARC2 for 12-16 hours. Then, lentivirus was aspirated and the cardiomyocytes were maintained in Complete Medium before experiments were performed. Exogenous all-trans retinal (1 μ M final concentration) was supplied immediately after transduction to improve light responsiveness of cardiac cells that express channelrhodopsins 23 . Confocal immunofluorescence microscopy. 48 hours after viral transduction, cardiomyocytes were fixed in 2% paraformaldehyde and incubated with primary antibodies overnight at 4 °C in 1XPBS containing 5% normal goat serum and 0.075% TritonX-100. Primary antibodies used were: MyBPC3 (1:250, Santa Cruz) and GFP (1:500, GF28R, Thermo Scientific). Secondary antibodies used were donkey anti-mouse conjugated to Alexa Fluor 488 and donkey anti-rabbit conjugated to Alexa Fluor 568 (1:500, LifeTechnologies). Images were obtained with a Nikon A1 confocal microscope (Nikon, Melville, NY) equipped with 40x oil, numerical aperture 1.3 objective.
Fluorescence measurements. EYFP fluorescence was measured with a CoolSnap HQ2 CCD camera (Photometrics, Tucson, AZ) using a Nikon Eclipse Ti-U microscope. Excitation was from an X-Cite 120Q light source (EXFO, Mississauga, Ontario), and fluorescence was detected using a Nikon B-2E/C filter set. All images were taken with the same acquisition parameters and analyzed with the ImageJ1.42q software. . The light intensity was attenuated with the built-in Polychrome system or with neutral density filters. Maximal quantum density at the focal plane of the 40x objective lens was 7.7 mW/mm 2 for 510-nm light and 6.7 mW/mm 2 for 560-nm light. Threshold-based closed-loop control was implemented by triggering a 5-V pulse at ~95% of the AP peak to open the shutter after a time delay set by an S44 electrical stimulator (Grass Medical Instruments, Quincy, MA). Data were analyzed using pClamp 10 and Origin 7 (OriginLab Corporation, Northampton, MA) software. Extracellular recording. Extracellular electrical recording was performed on clusters of synchronously beating NRVMs in the growth medium at 37 °C (controlled by an automatic temperature controller TC-324B, Warner Instruments Corporation, Hamden, CT) 7-10 days after transduction using the same equipment as described for patch clamping. The pipette was filled with the Tyrode's solution and placed in the vicinity of a cell cluster so that it did not perturb its contraction. The electrical activity was recorded for 30 s in the dark to assess the baseline, after which a 30-s light pulse was applied followed by another 30-s dark period to monitor the recovery of beating after illumination. For analysis, the low-pass filter at 0.3 Hz and a high-pass filter at 3 Hz were applied to the recorded traces using pClamp 10.

Statistics.
Statistical data are shown as mean ± sem. The n values in the legends indicate the number of independent experiments (individual cardiomyocytes or their clusters). For analysis of significance unpaired Student's t-test was used, and p values less than 0.05 were considered significant.