Optimized photo-stimulation of halorhodopsin for long-term neuronal inhibition

Abstract

Background: Optogenetic silencing techniques have expanded the causal understanding of the functions of diverse neuronal cell types in both the healthy and diseased brain. A widely used inhibitory optogenetic actuator is eNpHR3.0, an improved version of the light-driven chloride pump halorhodopsin derived from Natronomonas pharaonis. A major drawback of eNpHR3.0 is related to its pronounced inactivation on a time-scale of seconds, which renders it unsuited for applications that require long-lasting silencing.

Results: Using transgenic mice and Xenopus laevis oocytes expressing an eNpHR3.0-EYFP fusion protein, we here report optimized photo-stimulation techniques that profoundly increase the stability of eNpHR3.0-mediated currents during long-term photo-stimulation. We demonstrate that optimized photo-stimulation enables prolonged hyperpolarization and suppression of action potential discharge on a time-scale of minutes.

Conclusions: Collectively, our findings extend the utility of eNpHR3.0 to the long-lasting inhibition of excitable cells, thus facilitating the optogenetic dissection of neural circuits.

Keywords: Halorhodopsin; Inhibition; Optogenetic; Transgenic; eNpHR3.0.

Fig. 1

Blue light accelerates the recovery of eNpHR3.0-mediated currents from inactivation in a duration- and power-dependent manner. a Sample voltage-clamp recording illustrating that prolonged (10 s) photo-stimulation at 594 nm (5 mW) induces pronounced inactivation of eNpHR3.0-mediated currents. Note that the recovery from inactivation is slow (test pulse at Δt = 15 s). b Sample trace from another cell demonstrating that blue light (500 ms, 488 nm, 5 mW) accelerates the recovery from inactivation. Also note the outward current induced by blue light. c Recovery of eNpHR3.0-mediated currents is enhanced by blue light. Inset, recovery is defined as the ratio of current amplitudes induced by the test (at Δt) versus initial pulse, measured relative to I_late (i.e., recovery = A₂/A₁). Dotted lines represent mono-exponential fits to population data. Each cell was tested for all values of Δt either without (Control, n = 7 cells) or with (Rescue, n = 8 cells) an intervening photo-stimulation at 488 nm (500 ms). In a and b, current responses to − 10-mV voltage steps used to monitor access resistance are clipped for clarity (#). d Independent of the degree of inactivation (1 − I_late/I_peak), time constants of recovery are lower for rescue as compared to control trials. Each symbol represents a single cell. e Recovery from inactivation depends on the duration of the 488-nm rescue pulse (blue lines). All traces are from a single cell. f Quantification. g Recovery from inactivation depends on the power of the 488-nm rescue pulse at a constant duration of 1 s. All traces are from a single cell. h Quantification. Data are presented as mean ± SEM

Fig. 2

Co-stimulation at 594 nm and 488 nm attenuates the inactivation of eNpHR3.0-mediated currents during prolonged photo-stimulation in a mean power-dependent manner. a Sample voltage-clamp recording from a single cell illustrating eNpHR3.0-mediated currents in response to photo-stimulation at 594 nm (5 mW) alone (top) or in combination with 488 nm at variable power levels (middle and bottom). Power levels indicated refer to 488-nm light. b Dependence of inactivation on the power of 488-nm light. c The rescue effect of 488-nm light on the inactivation of eNpHR3.0-dependent currents depends on its mean, rather than peak, power. Top: continuous 488-nm stimulation (left) is equally effective in attenuating inactivation as compared to pulsed (1 kHz, 20/80% on/off) stimulation at constant mean power (right). Bottom: continuous 488-nm stimulation (left) is more effective in attenuating inactivation as compared to pulsed (1 kHz, 20/80% on/off) stimulation at constant peak power (right). d For quantification, I_late measured during pulsed stimulation was normalized to I_late obtained for the respective continuous-stimulation trials. Each symbol represents a single cell. Data are presented as mean ± SEM. **P < 0.01

Fig. 3

488-nm light alone enables efficient and stable long-term photo-stimulation of eNpHR3.0. a Sample voltage-clamp recordings from a single cell illustrating the power-dependence of HR-mediated currents evoked by photo-stimulation at 594 nm (top) or 488 nm (bottom), delivered at 1 mW (left), 3 mW (middle), or 5 mW (right). Note that photo-currents evoked at 488 nm display lower peak amplitudes, but high temporal stability across the entire power range examined. At the end of each trial, 488-nm light (5 mW) was used to accelerate the recovery from inactivation (note the difference in onset kinetics of evoked currents depending on the degree of previous inactivation). Current responses to − 10-mV voltage steps used to monitor access resistance are clipped for clarity (#). b–d Late (I_late, b) and peak (I_peak, c) current amplitudes as well as the ratio of I_late versus I_peak (d) normalized to the respective values at 594 nm and 1 mW (n = 7 cells). e Sample traces from a single cell photo-stimulated at 594 nm and/or 488 nm and a constant total light power of 5 mW. f Quantification of I_late measured during the photo-stimulation regimes indicated normalized to I_late obtained by photo-stimulation at 594 nm (5 mW) alone. Note that each combination of 594 nm plus 488 nm tested (at constant total power) considerably outperformed photo-stimulation at 594 nm alone (dotted line). Each symbol represents a single cell. Data are presented as mean ± SEM. **P < 0.01, ***P < 0.001

Fig. 4

Wavelength-dependent inactivation and recovery of eNpHR3.0 in X. laevis oocytes. a Sample photo-current traces of eNpHR3.0 upon stimulation for 60 s at 590 nm, 532 nm, or 473 nm at constant intensity (2.6 mW/mm²). b, c Quantification of the initial peak current (I_peak), the remaining current at the end of illumination (I_late), and the ratio I_late/I_peak. d I_late/I_peak upon 60-s-long illumination at 590 nm, 532 nm, or 473 nm at different light intensities (n = 5 cells). e Sample photo-current trace of eNpHR3.0. Inactivation was induced by a 60-s light pulse at 590 nm (2.6 mW/mm²). Recovery was probed by 10-ms light pulses (590 nm, 2.6 mW/mm²) at 1, 5, 10, 20, 40, 60, 120, and 300 s after the initial 60-s illumination. f Quantification of eNpHR3.0 recovery (n = 8 cells). g Peak-scaled sample traces from three different cells demonstrating that blue (473 nm) or violet (400 nm) light (2 s, 1 mW/mm²) accelerates the recovery from inactivation (at 5 s). Note the outward current induced by blue light. h Quantification of recovery as in g. i Peak-scaled sample traces from one oocyte illustrating eNpHR3.0 photo-currents induced by illumination at 590 nm alone (2.6 mW/mm²) or by co-illumination with either 473 nm (1 mW/mm²) or 400 nm (1 mW/mm²). j Quantification of I_late/I_peak as in i. k Sample traces from one oocyte illustrating eNpHR3.0 photo-currents induced by illumination at 473 nm alone (6.6 mW/mm²), by co-illumination at 590 nm (2.6 mW/mm²) and 400 nm (1 mW/mm²) or by co-illumination at 532 nm (6.6 mW/mm²) and 400 nm (1 mW/mm²). l, m Quantification of I_peak, I_late, and I_late/I_peak as in k. All measurements were performed in Ringer’s solution (pH 7.6) at a holding potential of − 40 mV. Data are presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001. c, h, j, m Asterisks indicate significance levels of post hoc t tests with Bonferroni correction following one-way ANOVA (h) or one-way repeated-measures ANOVA (c, j, m)

Fig. 5

Mechanistic insight of the inactivation and recovery of eNpHR3.0. a Decreasing extracellular proton concentration enhances inactivation of eNpHR3.0. Currents were measured in the same oocyte in Ringer’s solution at pH 5.6, pH 7.6, or pH 9.6. b Increasing extracellular chloride concentration reduces inactivation of eNpHR3.0. Currents were measured in the same oocyte at different chloride concentrations. Buffers with different chloride concentrations were achieved by mixing Ringer’s solution (pH 7.6) and NMG-Asp solution (pH 7.6) at different ratio. c Recovery of eNpHR3.0-mediated photo-currents in Ringer’s solution at pH 5.6 (n = 4 cells), pH 7.6 (n = 8 cells), or pH 9.6 (n = 4 cells) at a holding potential of − 40 mV. d Recovery of eNpHR3.0-mediated photo-currents at an extracellular chloride concentration of 6 mM (n = 5 cells), 16 mM (n = 6 cells), 60 mM (n = 6 cells), or 121 mM (n = 5 cells). pH was set to 7.6 and holding potential to − 40 mV. Dotted lines in c and d represent bi-exponential fits to population data. e Recovery of eNpHR3.0 (pH 7.6) at holding potentials of − 100 mV (n = 7 cells), − 40 mV (n = 5 cells), or + 20 mV (n = 5 cells). Five hundred ninety-nanometer light at an intensity of 2.6 mW/mm² was applied for 60 s in a and b, while in c–e, additional 10-ms 590-nm light pluses at the same intensity were delivered at 1, 5, 10, 20, 40, 60, 120, and 300 s after the initial 60-s illumination, as in Fig. 4e. Data are presented as mean ± SEM. *P < 0.05, ***P < 0.001. a, b Asterisks indicate significance levels of post hoc t tests with Bonferroni correction following one-way repeated-measures ANOVA

Fig. 6

Photo-stimulation of eNpHR3.0 at 488 nm enables efficient long-term hyperpolarization and inhibition. a Sample current-clamp measurements from a single cell (biased to about − 65 mV at rest) repetitively challenged with an inward current of constant amplitude either without (5 mW, left) or with photo-stimulation at 594 nm (5 mW, middle) or 488 nm (right), respectively. Note that yellow-light stimulation initially abolished action potential firing (#), which recovered during prolonged photo-stimulation periods. Also note that blue-light stimulation suppressed firing for the entire 1-min period on the background of a stable hyperpolarization. Insets: current responses to the last test pulse at higher temporal magnification (scale bars, 25 mV, 0.5 s). b Number of action potentials (AP) as a function of the test-pulse number. c Time-course of membrane potential (gray—period of photo-stimulation). Data are presented as mean ± SEM

Fig. 7

Wavelength and duration dependence of chloride loading due to photo-stimulation of eNpHR3.0. a Sample gramicidin perforated-patch recording of membrane potential in response to puff application of isoguvacine (Iso, 100 μM, 2 s) before (Control) and after photo-stimulation at 488 nm for 30 s (5 mW). Dotted lines indicate resting (V_rest) and peak (V_peak) membrane potential measured before photo-stimulation. Note that V_peak approximates E_GABA under our recording conditions. b Quantification of V_rest and V_peak before and after photo-stimulation. c, d As in a and b, but photo-stimulation was performed for 120 s at 488 nm (5 mW). e, f As in a and b, but photo-stimulation was performed for 120 s at 594 nm (5 mW). Experiments were performed at P4–10. Data are presented as mean ± SEM. n.s. not significant, **P < 0.01, ***P < 0.001

Fig. 8

Proposed photo-cycle of NpHR. An extracellular chloride ion is bound to the Schiff base lysine of NpHR at resting state, with K_m = 16 mM [11]. Photon absorption (with maximum at 580 nm) triggers the isomerization of retinal and starts the photo-cycle, containing intermediates K (omitted here), L, N, and O. The chloride ion is released into the cytosol during the transition from N to O, and uptake of a chloride ion from the extracellular side takes place in the recovery from O to the initial state. HR without a bound chloride ion is prone to deprotonation of the Schiff base in the L state (indicated by dashed line), leading to formation of M. This intermediate is long-lived and absorbs similarly as HR₄₁₀ (or M₄₁₂ in BR) from Halobacterium salinarum. The uptake of the proton for reprotonation of the M intermediate is very slow in dark (open arrow) but fast after absorption of a blue photon (blue arrow). Our data support deprotonation of the chloride-free L state (indicated by broken line)