Probing plant signal processing optogenetically by two channelrhodopsins

Ethics

The laparotomy to obtain oocytes from X. laevis was carried out in accordance with the principles of the Basel Declaration and recommendations of Landratsamt Wuerzburg Veterinaeramt. The protocol under License number 70/14 from Landratsamt Wuerzburg, Veterinaeramt, was approved by the responsible veterinarian.

Molecular cloning

The pGEMHE vector⁵¹ was used for in vitro RNA synthesis and the following expression in X. aevis oocytes. The binary pCAMBIA3300 vector with the UBQ10 promoter or the pCAMBIA1300 vector with the CaMV 35S promoter was used for Agrobacterium infiltration and stable transformation of plants.

Substitutions were introduced by QuikChange site-directed mutagenesis PCR. The pCambia1300 NES-2×R-GECO1 was cloned using the USER cloning technique⁵². For other constructs, all of the fragments with suitable restriction sites were introduced into the pGEMHE vector or the binary vectors pCAMBIA3300 with T4 DNA ligase (Thermo Fisher Scientific).

Plasmid extractions from Escherichia coli cultures were carried out using the MiniPrep 250 Kit (QIAGEN) according to the manufacturer’s instructions. All constructs were verified by sequencing (Eurofins Genomics). The sequences of XXM 1.1 and XXM 2.0 are available in the Supplementary Information.

Molecular engineering and development of XXM 2.0

ChR2-XXM (D156H) was selected as a template to start molecular engineering owing to the high expression level and already enhanced Ca²⁺ permeability¹⁹. As no specific ion-selective filter is identified in ChR2, we anticipated that the three (extracellular, centre and intracellular) molecular gates (ECG, CG and ICG) might contribute to modulating ion-selective properties of ChR2, in addition to its gating function. Considering known substitution sites of ChR2 variants with modified ion selectivity in relation to the crystal structure indeed suggests a possibility to modulate ion selectivity using the ChR2 structure generated by PyMOL⁵³ (Extended Data Fig. 1b). Within the three-dimensional structure, the internal cavities were calculated using HOLLOW, which is a published python script⁵⁴. Some gating residues (H134 and E83 in ICG, and Q117 and R120 in ECG) and residues located in close proximity to the gate (E101 near ECG) were therefore substituted (Extended Data Fig. 1b). S63 (when combined with D156H) and E90 substitutions in CG have been reported to decrease Ca²⁺ conductance^55,56, and therefore were not tested here. The resulting variants were first tested by two-electrode voltage-clamp analysis with X. laevis oocytes for comparison of photocurrent amplitude (Extended Data Fig. 1c). Variants exhibiting high photocurrent amplitude were next studied to determine the ion selectivity (reversal potential shift comparison) by changing extracellular ion concentrations (Extended Data Fig. 1d, see details in the following section) for selecting the superior candidate with high Ca²⁺ conductance (Fig. 1). To further enhance its membrane trafficking, N-terminal truncation of ChR2 and addition of signal peptides as recently reported⁴ were explored and examined by fluorescence imaging along with a comparison of photocurrents in X. laevis oocytes.

Two-electrode voltage-clamp analysis with X.
laevis oocytes

The AmpliCap-MaxT7 High Yield Message Maker Kit (Epicentre Biotechnologies) was used to synthesize XXM, XXM 1.1, XXM 1.2, XXM 1.3 and XXM 2.0 complementary RNA (cRNA). All of the cRNAs were stored in nuclease-free water at −20 °C. Oocytes were injected with 30 ng cRNA and incubated in ND96 solution (96 mM NaCl, 2 mM KCl, 1 mM CaCl₂, 1 mM MgCl₂, 10 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), pH 7.4) at 16 °C for 2 days. Two-electrode voltage-clamp recordings were carried out at room temperature with a two-electrode voltage-clamp amplifier (TURBO TEC-03X, NPI Electronic). Electrode capillaries (diameter 1.5 mm, wall thickness 0.178 mm; Hilgenberg) were pulled by a vertical puller (PC-10, Narishige) and filled with 3 M KCl, with tip resistance of 0.4–1 MΩ. A USB-6221 DAQ interface (National Instruments) and WinWCP V5.3.4 software (University of Strathclyde, UK) were used for data acquisition. Blue light illumination was supplied by a 473-nm laser (Changchun New Industries Optoelectronics Tech).

Photocurrents of different XXM versions in response to blue light illumination (473 nm, 3 mW mm^−2) were compared at a holding potential of −90 mV in a standard recording solution (96 mM NaCl, 2 mM KCl, 1 mM BaCl₂, 1 mM MgCl₂, 10 mM HEPES, pH 7.6). For the reversal potential shift (mV) comparison, photocurrents of XXM and XXM 1.1 were recorded in both buffer A (119 mM N-methyl-d-glucamine, 0.8 mM BaCl₂, 5 mM HEPES, pH 7.6) and buffer B (80 mM BaCl₂, 5 mM HEPES, pH 7.6). Blue light (473 nm, 3 mW mm^−2) was applied to activate different XXM variants. In the experiment for the comparison of calcium currents between XXM and XXM 1.1, 10 mM 1,2-bis(o-aminophenoxy)ethane-N,N,N′N′-tetraacetic acid (BAPTA) was injected to block the endogenous Ca²⁺-activated chloride channels of oocytes. The Ca²⁺ currents triggered by 3 mW mm^−2 blue light were measured in the bath solution containing 80 mM CaCl₂ and 10 mM 3-(cyclohexylamino)-2-hydroxy-1-propanesulfonic acid (CAPSO) at pH 9 and a holding potential of −90 mV.

The action spectrum for XXM 2.0 was detected with light of different wavelengths. Light of different wavelengths (399.3 nm, 422 nm, 440.7 nm, 456 nm, 479.5 nm, 496 nm, 516 nm, 541 nm, 562 nm and 595 nm) was obtained by narrow bandwidth interference filters (Edmund Optics) together with a PhotoFluor II light source (89 North). Equal photon flux was set for each wavelength. Photocurrents were detected when treated with light of distinct wavelength and normalized to the maximal stationary currents triggered by blue light (456 nm). The light intensities were measured with a Plus 2 Power & Energy Meter (Laserpoint).

Agrobacterium transformation

The Agrobacterium tumefaciens strain GV3101 was collected by centrifugation after 1 day of culture and washed twice with sterile distilled water. All of the plasmids described for plant expression were transformed into A. tumefaciens by an electroporation protocol⁵⁷. Monoclonal colonies were selected from lysogeny broth (LB)–agar plates with 100 μg ml^−1 kanamycin, 25 μg ml^−1 gentamycin and 10 μg ml^−1 rifampicin at 28 °C and cultured in LB medium with 100 μg ml^−1 kanamycin, 25 μg ml^−1 gentamycin and 10 μg ml^−1 rifampicin at 28 °C. The colonies were confirmed by PCR.

Agrobacterium infiltration of N.
benthamiana leaves

For Agrobacterium infiltration, 30–37-day-old N. benthamiana plants grown in the greenhouse (40–60 kilolux light irradiation from 08:00–20:00, 24–26 °C) were used. Transient transformation of N. benthamiana plants was carried out according to the protocol of ref. ⁵⁸. Briefly, agrobacteria were cultured in LB medium with 150 µM acetosyringone at 28 °C for about 16–18 h, collected by centrifugation and washed twice with infiltration buffer (10 mM MgCl₂, 10 mM 2-morpholinoethanesulfonic acid (MES) (pH was adjusted to 5.6 by KOH), 150 μM acetosyringone). The final concentration was adjusted to 0.4 at an optical density of 600 nm (OD_600nm) in infiltration buffer. A 1-ml syringe was used to infiltrate the resuspended agrobacteria into the leaves through the abaxial epidermis. The infiltrated plants were grown in 650 nm red light (light intensity of about 30 µW mm^−2; cycles of 14 h light at 26 °C/10 h dark at 16 °C).

Stable transformation of N.
tabacum plants

N. tabacum (N. tabacum cultivar Petit Havana SR1) seeds sterilized by 6% NaOCl were germinated and grown in 500-ml sterile plastic boxes on agar plates (Murashige and Skoog medium including vitamins and MES (Duchefa Biochemie), 3% sucrose, 0.8% Gelzan (Sigma-Aldrich), pH 5.8 with KOH) in stable culture conditions of cycles of 14 h light at 26 °C/10 h dark at 16 °C. The A. tumefaciens strain GV3101 harbouring the pCAMBIA3300 vector with BASTA resistance or the pCAMBIA1300 vector with hygromycin resistance was used for N. tabacum transformation as described previously⁵⁹ with some minor modifications. Agrobacteria were collected and washed twice with sterilized MS solution (Murashige and Skoog medium including vitamins and MES (Duchefa Biochemie), 3% sucrose, pH 5.8 with KOH). The final concentration was adjusted to OD_600nm = 0.1. Sterilized leaves were cut into pieces of about 2 cm² and soaked in the resuspended agrobacteria solution for 20 min. The wet leaf pieces were dried, placed on plant growth medium and transferred after 3 days to a callus-inducing medium (Murashige and Skoog medium including vitamins and MES (Duchefa Biochemie), 3% sucrose, 0.8% Gelzan (Sigma-Aldrich), pH 5.8 with KOH, 20 µg ml^−1 dl-phosphinothricin (Duchefa Biochemie) or 30 mg l^−1 hygromycin B (Thermo Fisher Scientific), 500 µg ml^−1 ticarcillin disodium (Duchefa Biochemie), 100 mg l^−1 myo-inositol, 1 mg l^−1 thiamine hydrochloride, 1 mg l^−1 6-benzylaminopurine and 100 µg l^−1 1-naphthaleneacetic acid (Sigma-Aldrich)) and cultured under 650-nm light-emitting diode (LED) red light with light intensity of about 30 µW mm^−2. The pieces of leaves, explants and calli were transferred to new callus-inducing medium every 2 weeks. Generated shoots were decapitated and moved onto rooting medium (Murashige and Skoog medium including vitamins and MES (Duchefa Biochemie), 3% sucrose, 0.8% Gelzan (Sigma-Aldrich), pH 5.8 with KOH, 20 µg ml^−1 dl-phosphinothricin (Duchefa Biochemie) or 30 mg l^−1 hygromycin B (Thermo Fisher Scientific), 500 µg ml^−1 ticarcillin disodium (Duchefa Biochemie), 100 mg l^−1 myo-inositol, 1 mg l^−1 thiamine hydrochloride)) and, after root formation, were grown on soil in 650 nm red light (about 30 µW mm^−2, cycles of 14 h light at 26 °C/10 h dark at 16 °C).

The transgenic plants were verified by eYFP fluorescence or R-GECO1 fluorescence in leaves. Seeds of individual plants were collected and selected for BASTA or hygromycin resistance using selection medium (Murashige and Skoog medium including vitamins and MES (Duchefa Biochemie), 3% sucrose, 0.8% Gelzan, pH 5.8 with KOH, 20 µg ml^−1 dl-phosphinothricin (Duchefa Biochemie) or 30 mg l^−1 hygromycin B (Thermo Fisher Scientific)). Homozygous lines were used for experimental studies.

Confocal microscopy and image processing

A confocal laser scanning microscope (Leica SP5, Leica Microsystems CMS) controlled by Leica LAS AF (Version 2.7.3.9723, Leica Microsystems) was used to subcellularly localize the rhodopsin–eYFP fusions in plant cells or X. laevis oocytes. The yellow fluorescence was observed with a dipping 25× HCX IRAPO 925/0.95 objective in N. benthamiana leaves 3 days post Agrobacterium infiltration, in N. tabacum leaves after 45 days grown in red light (light intensity of about 30 µW mm^−2; cycles of 14 h light at 26 °C/10 h dark at 16 °C) and in X. laevis oocytes 2 days post injection. eYFP was excited at 496 nm and fluorescence was captured between 520 and 580 nm. N. benthamiana and N. tabacum leaf discs were placed upside down for yellow fluorescence detection. FIJI IMAGEJ-win64 software⁶⁰ was used for image processing.

Aequorin-based cytoplasmic free Ca²⁺ measurements

Co-infiltration of agrobacteria with 10 μM of coelenterazine (PJK Biotech) was carried out as described in the ‘Agrobacterium infiltration of N. benthamiana leaves’ section under red light conditions. After 2 days in the red light growth room, aequorin luminescence from the infiltrated leaves was measured by a homemade luminometer. Luminescence was detected by a photomultiplier (Photo Counting Module MP 1983 RS CPM, Perkin Elmer) controlled by IGI-MPRS232 (IGIsystems). Labview 14.0.0 (National Instruments) was used to control the shutter and LEDs. A 520-nm green light LED (from WINGER, WEPGN3-S1) with a light intensity of 50 μW mm^−2 was used to activate rhodopsins. To prevent the LED light from being detected by the photomultiplier, an additional shutter (Uniblitz, VCM-D1) was installed.

Live-cell imaging and all-optical physiology measurements

Live-cell imaging experiments were carried out using transiently transformed N. benthamiana leaves, transgenic N. tabacum leaves (Ret-XXM 2.0 with R-GECO1, Ret-ACR1 2.0 with R-GECO1, R-GECO1, Ret-XXM 2.0 with pHuji and pHuji) or transiently transformed N. benthamiana mesophyll protoplasts (R-GECO1 and Ret-XXM 2.0 with R-GECO1). The A. tumefaciens strain GV3101 harbouring the corresponding pCAMBIA vectors and A. tumefaciens strain K19 were cultured at 28 °C overnight. The infiltration solution for co-expression of Ret-XXM 2.0 with R-GECO1 contained A. tumefaciens strain K19 (OD_600nm = 0.3) and A. tumefaciens strain GV3101 harbouring the pCAMBIA plasmids (pCAMBIA3300 vector carrying Ret-XXM 2.0, pCAMBIA1300 vector carrying NES-2× R-GECO1; OD_600nm = 0.4 for both). Control plants were infiltrated with infiltration solution containing A. tumefaciens strain K19 (OD_600nm = 0.3) and A. tumefaciens strain GV3101 harbouring the pCAMBIA1300 vector carrying NES-2× R-GECO1 (OD_600nm = 0.4). [Ca²⁺]_cyt measurements were carried out 3 days post infiltration. Mesophyll protoplasts were prepared from the transiently transformed N. benthamiana leaves from which the abaxial epidermis was peeled off. The leaf pieces without the main vein were incubated in enzyme solution (1% BSA, 0.05% pectolyase Y23, 0.5% cellulase R-10, 0.5% macerozym R-10, 1 mM CaCl₂, 10 mM MES, 500 mM d-sorbitol, pH 5.6 with Tris) for 2 h. Following enzymatic digestion, cells were filtered through a 100-µm mesh. Protoplasts were collected by low-speed centrifugation (80g) without acceleration at 4 °C. Protoplasts were washed twice using precooled wash solution (1 mM CaCl₂, 500 mM d-sorbitol, pH 5.6 with Tris) and finally resuspended in precooled wash solution and stored on ice until use. Leaf disc samples with a diameter of 5 mm were prepared by peeling the abaxial epidermis off and gluing leaf discs upside down with medical adhesive (ULRICH Swiss) on custom-made recording chambers. The samples were allowed to recover in the dark in bath solution (1 mM KCl, 1 mM CaCl₂, 10 mM MES, and 1,3-bis(tris(hydroxymethyl)methylamino)propane (BTP), pH  6.0) at room temperature (about 25 °C) overnight before R-GECO1 or pHuji fluorescence measurement and all-optical experiments were carried out.

The microscope setup to carry out live-cell imaging is described in detail elsewhere⁶¹. R-GECO1²⁵ and pHuji²⁶ were excited with 570 nm excitation light. VisiView software (Version 2.1.1) was used to simultaneously control R-GECO1 imaging and triggering of local green light (532 nm, 180 μW mm^−2) illumination by a solid-state laser (Changchun New Industries Optoelectronics Tech) or global green light (520 nm, 9 μW mm^−2) illumination by a homemade LED device (LED from WINGER, WEPGN3-S1). XXM 2.0 activation by green light was carried out in the 5 s interval time during R-GECO1 or pHuji imaging. Green light illumination was terminated more than 1 s before R-GECO1 was excited to avoid photoswitching effects of R-GECO1 during optogenetic stimulation as reported previously⁶². A dichroic mirror (HC593 (F38-593), AHF Analysetechnik) combined with a high-speed filter wheel equipped with bandpass filters for R-GECO1 or pHuji (ET 624/20 nm) was used to detect the red fluorescence. During the detection of [Ca²⁺]_cyt signal triggered by XXM 2.0 stimulation when a different light condition was used, a 2 s interval time was set during R-GECO1 illuminations. It should be noted that the light pulse protocols used to set defined [Ca²⁺]_cyt signatures must be customized for each particular cell system or plant line and cannot act as a blueprint as features such as the expression level and cell type used or the Ca²⁺ homeostasis will probably influence the signature.

To avoid undesirable XXM 2.0 or ACR1 2.0 activations during bright-field imaging, the microscope white light source was covered by a primary red filter (Lee filter 106). Simultaneous plasma membrane potential and R-GECO1 fluorescence recordings were carried out in R-GECO1, Ret-XXM 2.0 with R-GECO1, Ret-ACR1 2.0 with R-GECO1 samples. Current-clamp-based voltage recordings were carried out by microelectrode impalement as described elsewhere⁶¹. Glass microelectrodes filled with 300 mM KCl were connected to the microelectrode amplifier (TEC-05X; NPI Electronic) equipped with head stages of more than 10¹³â€‰Î© input impedance. The reference electrodes were filled with 300 mM KCl. A piezo-driven micromanipulator (Sensapex) was used to direct the glass electrode. The current-clamp protocols were applied by WinWCP V5.3.4 software (University of Strathclyde, UK). R-GECO1 fluorescence intensities were recorded using FIJI IMAGEJ-win64 software⁶⁰.

Ca²⁺ signals were detected in N. tabacum leaves utilizing the R-GECO1 reporter following Pst treatment using a perfusion system (780 μl min^−1), which prevents motion and touch-induced imaging artefacts^47,63,64,65. Pst inoculated from LB–agar plates containing 10 µg ml^−1 rifampicin was cultured in LB medium with 10 µg ml^−1 rifampicin for 1.5 days (28 °C, 200 r.p.m.) and subsequently subcultured in 500 ml LB medium containing 10 µg ml^−1 rifampicin for 16 h. Pst cells were washed twice using sterile deionized water and suspended in bath solution (1 mM KCl, 1 mM CaCl₂, 10 mM MES and BTP, pH 6.0), resulting in a final perfusion solution with an OD_600nm of 0.5. Leaf disc samples from R-GECO1 transgenic N. tabacum plants without abaxial epidermis were glued with the adaxial side down to the coverslip in custom-made chambers using Medical Adhesive B (Ulrich Swiss) and allowed to recover in bath solution overnight before Ca²⁺ imaging.

Membrane voltage recordings in mesophyll cells

Microelectrodes for mesophyll cell impalement were pulled from borosilicate glass capillaries (inner diameter 0.58 mm, outer diameter 1.0 mm, Hilgenberg) using a horizontal laser puller (P2000, Sutter). Microelectrodes filled with 300 mM KCl having an electrode resistance of 60–110 MΩ were connected by Ag/AgCl wires to the microelectrode amplifier (Axon geneclamp 500 or VF-102; BioLogic). Reference electrodes were filled with 300 mM KCl, and plugged with 2% agar in 300 mM KCl. The NA USB-6221 interface (National Instruments) was used to digitalize data. Cells were impaled by an electronic micromanipulator (NC-30, Kleindiek Nanotechnik) and current-clamp protocols were applied with the WinWCP V5.3.4 software (University of Strathclyde, UK).

Plant growth conditions and sample collection

All of the WT and transgenic N. tabacum plants were grown under constant red light (650 nm, 30 μW mm^−2, 26 °C) for 45 days. For ACR1 2.0 and XXM 2.0 stimulation, 0 h, 1 h, 4 h and 24 h additional global green light (520 nm, 9 μW mm^−2) were applied for the experimental groups. For osmotic stress treatment, 35% PEG was used to water Ret-eYFP #1 transgenic N. tabacum plants. This PEG concentration was selected after initial experiments with different PEG concentrations that would cause wilting phenotypes in a similar time frame (after 4–5 h) to that for the green-light-treated Ret-ACR1 2.0 plants. For Pst treatment, the Pst was washed twice with sterile deionized water and suspended in 10 mM MgCl₂ containing 0.04% Silwet L-77 with a final OD_600nm of 0.5 and sprayed on the entire Ret-eYFP #1 transgenic N. tabacum plants. Pst treatment was applied by spray inoculation to prevent wounding effects taking place that would unequivocally occur during infiltration by a syringe. The same amount (about 25 ml for each plant) of 10 mM MgCl₂ containing 0.04% Silwet L-77 was sprayed on Ret-eYFP #1 transgenic N. tabacum as the negative control. At t = 0 h before the additional green light illumination, PEG, Pst or 10 mM MgCl₂ was applied. For plants treated with 0 h, 1 h and 4 h green light illumination, the fifth leaves of N. tabacum were collected at different time points (t = 0 h, 1 h, 4 h and 8 h) and frozen in liquid nitrogen quickly for metabolite measurement and transcriptomics analysis. For plants treated with 24 h global green light, PEG, Pst or 10 mM MgCl₂, the fifth leaves from N. tabacum at different time points (t = 0 min, 3 min, 10 min, 0.5 h, 1 h, 2 h, 4 h, 8h and 24 h) were used for detection and quantification of ROS, electrolyte leakage estimations, chlorophyll fluorescence detection, metabolite measurement and transcriptomics analysis.

Detection of necrosis in leaf discs

A tissue puncher (Stiefel Disposable biopsy punch, diameter of 6 mm) was used to prepare leaf discs from the fifth leaf of 45-day-old N. tabacum plants. Leaf discs were washed twice with deionized water and transferred into 24-well plates containing 0.4 ml ultrapure H₂O that contained 1 mM CaCl₂, 10 mM CaCl₂, 5 mM EGTA (pH 7.0, KOH), 5 mM K₄BAPTA or 5 mM K₄BAPTA plus 10 mM NaCl as indicated in Extended Data Fig. 5c,d. Samples were placed in the dark for 1 h before exposing them to the light condition (growth chamber with constant red light (650 nm, 30 μW mm^−2) plus constant green light (520 nm, 9 µW mm^−2, 26 °C)). Images were captured after 24 h treatment.

Chlorophyll fluorescence measurements

N. tabacum plants treated as described in the ‘Plant growth conditions and sample collection’ section were used to quantify photosynthesis performance with a pulse-amplitude modulation fluorometer. The fifth leaf was fixed and monitored with a Maxi pulse-amplitude modulation fluorometer (AVT 033); chlorophyll fluorescence measurements were recorded with IMAGING WIN v.2.41a FW MULTI RGB (Walz). The dark-adapted N. tabacum leaf was exposed to actinic light with intensity 7 (photosynthetically active radiation as 146 μmol m^−2 s^−1). The maximal fluorescence yield of a dark-adapted sample (Fm) and dark-level fluorescence yield (Fo) were detected. The quantum yield of the dark-adapted leaf samples is a measure of the potential quantum yield of the samples, which was calculated according to the equation: Yield = (Fm − Fo)/Fm.

Electrolyte leakage estimations

The membrane integrity of N. tabacum leaf cells was estimated by electrolyte leakage of leaf samples. Ion conductivity was measured as described previously⁶⁶. Seven leaf discs (5 mm diameter) were detached from the fifth leaves of 45-day-old plants and equilibrated together in 0.3 ml of ultrapure H₂O after washing twice with ultrapure H₂O. Ion conductivity was quantified 20 min after leaf disc equilibration in ultrapure H₂O (EC1) using a LAQUAtwin EC-11 conductivity meter (Horiba). The samples were then heated at 99 °C for 1 h to measure the final electrical conductivity (EC2) when the samples reached room temperature again. The relative electrolyte leakage was calculated, as a percentage, by the formula: EL = EC1/EC2 × 100.

Detection of ROS

Chemical detection of ROS in green light (520 nm, 9 μW mm^−2)-treated N. tabacum leaves was carried out by 3,3-diaminobenzidine (DAB, Sigma-Aldrich) staining as described previously⁶⁷. The fifth leaves of N. tabacum plants were stained with fresh DAB staining solution (10 mM Na₂HPO₄, 1 mg ml^−1 DAB, 0.05% Tween 20, pH 3.0) by application of negative pressure for 5 min in dark. After 5 h incubation (shaking speed of 80–100 r.p.m.), the stained leaves were moved into fresh chlorophyll destaining solution (ethanol/acetic acid/glycerol, 3:1:1) and bathed in hot water (about 90–95 °C) for 15 min. Finally, the stained leaves were put into cold fresh chlorophyll destaining solution for 30 min. Images were taken with a plain white background under uniform lighting.

The chemiluminescent ‘superoxide probe’ luminol can be applied to indicate the ROS production⁶⁸. Superoxide released from leaf tissues was detected by the luminescence of luminol with Skanlt software (Version 6.1) according to the method described previously⁶⁹ with minor modifications. Leaf discs were prepared from the fifth leaves of 45-day-old N. tabacum plants using a tissue puncher (Stiefel Disposable biopsy punch, diameter of 6 mm). Leaf discs were washed with deionized water twice and transferred into the 96-well assay plate (black plate, clear bottom with lid, Corning) and incubated in the dark overnight to recover. Water was replaced with 200 µl of luminol–peroxidase working solution (30 mg l^−1 luminol (Sigma) and 20 mg l^−1 horseradish peroxidase (Sigma)) in each well containing leaf discs. Samples were kept in the dark for 1 h before measurement. Luminescence was measured in a microplate reader (Luminoskan Ascent, Thermo Labsystems) and 5 min global constant green light (9 µW mm^−2) illumination was applied during the rest periods.

The method for in vivo measuring the production of H₂O₂ amperometrically from mesophyll cells in parallel with intracellular membrane potential recordings was described previously⁷⁰. Measurements were carried out in standard bath solution (1 mM KCl, 1 mM CaCl₂, 10 mM MES, and BTP, pH 6.0) with a platinum–iridium electrode (MicroProbes) cut back to an active (uninsulated) area of about 1 mm length. ROS detection was carried out by Patch-Master software V2x90 (HEKA). The platinum–iridium disc was gently placed in close proximity to mesophyll cells and held at a constant voltage of 600 mV with an amperometry amplifier (VA 10X, NPI Electronic). Oxidation of H₂O₂ at the active microelectrode surface resulted in a positive current signal, which was low-pass-filtered at 1 Hz and recorded with Patch-Master software V2x90 (HEKA). The electrode was calibrated in freshly prepared bath solutions with defined H₂O₂ concentrations. Green light (520 nm, about 9 µW mm^−2) illumination on the N. tabacum leaves was carried out by green LEDs (WINGER, WEPGN3-S1).

All-trans retinal and carotenoid measurements

All-trans retinal and carotenoids were measured according to a protocol published elsewhere⁴. The fifth leaf of transgenic N. tabacum plants grown for 45 days in red light was triturated in liquid nitrogen and 200 mg leaf material was extracted with 500 µl of chloroform. The extract was centrifuged for 5 min at 18,400g and 50 µl of the organic phase was evaporated in a SpeedVac at 40 °C and dissolved in 50 µl of a 1:1 ethanol and chloroform mixture. A 5 µl volume of dissolved solution was analysed by ultrahigh-performance liquid chromatography (UPLC) combined with ultraviolet and tandem mass spectrometry detection using a Waters Acquity UPLC system coupled to a Waters Quattro Premier triple-quadrupole mass spectrometer equipped with an electrospray interface. Ten plants were used for retinal and 12 plants were used for carotenoid quantification.

Phytohormone measurement

All of the samples were prepared as described in the ‘Plant growth conditions and sample collection’ section. Ground samples (150 mg) were lyophilized in a laboratory freeze dryer (CHRIST, Laboratory freeze dryer Alpha 1-2) and subsequently used for phytohormone extraction. The extraction and chromatographic separation was carried out as described previously⁷¹, using 5 ng of dihydro-JA, JA–norvaline, [¹⁸O₂]OPDA, [D₄]SA and [D₆]ABA as phytohormone internal standard. The extraction solution contained ethylacetate (p.a.) and formic acid (p.a.) (99:1 in volume) to which 5 ng phytohormone internal standard was added. All samples were fixed on a TissueLyser with shaking for 3 min at a speed of 23 Hz. After that, samples were centrifuged and the supernatant was dried in a SpeedVac at 45 °C and finally dissolved in 40 μl liquid (acetonitrile (for high-performance liquid chromatography)/water (MilliQ), 1:1 (v/v)). Phytohormones were analysed by UPLC–electrospray interface–tandem mass spectrometry using a Waters Acquity I-Class UPLC system coupled to an AB Sciex 6500+ QTRAP tandem mass spectrometer (AB Sciex), operated in negative ionization mode as described elsewhere^72,73. Analyst (Version 1.6.3) software and MultiQuant (Version 3.0.2) software from Sciex were used for mass spectrometry detection of hormones and metabolites.

Proline quantification

The proline content of N. tabacum leaves was measured according to the spectroscopic method of ref. ⁷⁴. Samples were prepared as described in the section ‘Plant growth conditions and sample collection’. Ground samples (150 mg) were lyophilized in a laboratory freeze dryer (CHRIST, Laboratory freeze dryer Alpha 1-2). The dry samples were mixed in 40% ethanol and incubated at 4 °C overnight. The supernatant was collected after centrifugation at 13,500g for 5 min. A 500 µl volume of the ethanol extraction or 100 µl standard solution was mixed with 1,000 µl reaction mix (1% ninhydrin (w/v) in 60% acetic acid (v/v) and 20% ethanol (v/v)). After incubation at 95 °C for 20 min and subsequent centrifugation for 1 min at 9,000g, the samples (supernatant) were subjected to absorption measurement at 520 nm with a spectrophotometer (Hitachi U-1500). Proline concentration was determined according to the standard curve, and concentrations were calculated on the basis of dry weight.

Transcriptomics analysis

Three replicates of leaf samples from two batches of N. tabacum plants were collected for RNA sequencing. The experimental design for transcriptomics is shown in Extended Data Fig. 8a. Samples at 0 h, 1 h, 4 h and 8 h from plants growing in red light conditions were used as the biological controls to compare the expression levels of Ret-XXM 2.0 and Ret-ACR1 2.0 transgenic plants during or after the global green light illumination. In these experiments PEG-treated plants and Pst inoculation served as possible physiological controls to ACR1 2.0 and XXM 2.0 activation, and spraying leaves with buffer (MgCl₂) served as an additional control to Pst-sprayed plants (Extended Data Fig. 8a). RNA extraction was carried out using ground samples (150 mg) with the Macherey-Nagel NucleoSpin RNA Plant Kit (https://www.takarabio.com/documents/User%20Manual/UM/UM_TotalRNAPlant_Rev_07.pdf). DNase1 (Thermo Fisher) was used to digest DNA. RNA sequencing was carried out by Novogene (UK) with an Illumina NovaSeq 6000 Sequencing System. Paired-end 150 bp was the read length. Data processing (fastp) and mapping to the N. tabacum genome (kallisto)⁷⁵ was carried out using Amalgkit (https://github.com/kfuku52/amalgkit). Functional annotations of N. tabacum genes for subsequent bioinformatic analyses were retrieved from the Dicots PLAZA 5.0 repository^76,77.

Normalization and DEG analysis were carried out employing the DIANE package using DESeq2 and default parameters⁷⁷. DESeq2 uses a Wald test, in which the shrunken estimate of log fold change is divided by the standard error to produce a z-statistic. This z-statistic is then compared against a standard normal distribution⁷⁸. A prefiltering step eliminated genes exhibiting rowMeans over all conditions ≤ 5 counts, reducing the number of input genes from 69,500 to 42,196. Unless stated otherwise, |log₂[fold change]| ≥ 2 with a false discovery rate of 0.01 was taken as the cutoff for DEG identification. Only a small number of DEGs were identified when comparing Ret-ACR1 2.0 or Ret-XXM 2.0 transgenes with Ret-eYFP control plants grown under non-stimulating red light conditions (Supplementary Table 2, tab 1), demonstrating that ACR1 and XXM expression have virtually no impact on the transcript profiles of plants and plants growing in red light conditions are proper biological controls. Likewise, GO enrichment analysis on DEGs was carried out using the DIANE suite and corresponding enrichment plots were created using the srplot web interface (https://www.bioinformatics.com.cn/en). Venn diagrams were generated with the GOVenn script of the GOPlot package⁷⁹ and subsequent GO analysis on Venn subsets was carried out using gprofiler2⁸⁰. g:Profiler functional enrichment analysis is conducted using the g:GOSt tool that carries out over-representation analysis via the hypergeometric test⁸⁰. Finally, heat maps were generated with the pheatmap R package (version 1.0.12; https://CRAN.R-project.org/package=pheatmap). N. tabacum genes were further annotated manually according to their A. thaliana orthologues⁸¹ and corresponding gene symbols from the Aramemnon database⁸².

Quantitative real-time PCR

Quantification of gene transcripts was carried out by real-time PCR as described elsewhere⁸³. The samples were prepared as described in the section ‘Plant growth conditions and sample collection’. Ground samples (100 mg) were used for RNA extraction by the NucleoSpin RNA Plant Kit (Macherey-Nagel). cDNA was synthesized from 2.5 g of total RNA using oligo(dT) primer (Thermo Fisher Scientific) and M-MLV Reverse Transcriptase (Promega). All quantitative real-time PCR reactions were carried out with the Eppendorf Mastercycler ep realplex 2 system and Eppendorf Mastercycler ep realplex (Version 2.2) software, in a 20 µl reaction volume containing 2 µl diluted cDNA, 0.8 µM primer pairs and 10 µl ABsolute qPCR SYBR Green Capillary Mix (Thermo Scientific). Information on the genes and primers used is provided in Supplementary Table 7. Transcripts were normalized to that of 10,000 molecules of actin.

Surface potential recording on N.
tabacum leaves

The design for long-range electrical signal measurements in N. tabacum leaves is shown in a diagram in Extended Data Fig. 10a. The surface potential recordings were carried out on 6–7-week-old N. tabacum leaves according to a previously described protocol⁸⁴ with minor modifications. A USB-6221 interface (National Instruments) was used to digitalize the electrical signals, which were recorded with WinWCP V5.3.4 software (University of Strathclyde, UK). The electrode silver wires (Ag/AgCl) connected to a microelectrode amplifier (Axon geneclamp 500) were wrapped around the petiole of the fifth leaves gently. Electrode gel (Auxynhairol) was used to cover the surface of the wrapped wires to aid connectivity between the electrodes and the petiole. The reference electrode consisting of an Ag/AgCl electrode was placed in a 200-ml pipette tip filled with electrode gel (Auxynhairol), which was inserted in the soil of the pots the N. tabacum plants grew in. Nine hours after mounting the electrodes, the surface potential was recorded when applying a 600-ms green light (532 nm, 5.3 mW mm^−2) pulse at the main vein. A popular Technology Enhanced Clad Silica multimode optical fibre (diameter of 1,500 µm, 0.39 NA, Thorlabs) was placed directly on the top of main vein for light application. To prevent scattering of light and to guarantee local green light application, the optical fibre was covered by a non-transparent black plastic pipe up to the tip. The electrical signals were monitored at exactly 5 cm away from the illumination spot.

Significance analysis

Student’s t-test or ANOVA was used to analyse significant differences between groups. Significance analysis among more than three groups was carried out with one-way ANOVA using IBM SPSS statistics (version 26.0). For the post hoc multiple comparisons, the homogeneity of variances was tested first. If the variances were homogeneous (P > 0.05), the Tukey test was used for significance analysis. If the variances were not homogeneous (P < 0.05), either the Dunnett T3 or Games-Howell test was chosen, depending on whether the sample sizes were equal or not. Different letters indicate significant differences among the samples (lowercase letters indicate P values at the 0.05 level and capital letters indicate P values at the 0.01 level). Significance analysis among two groups was carried out with a two-sided Student’s t-test. *P ≤ 0.05, **P ≤ 0.01 and ***P ≤ 0.001. The significance analysis is performed with a 95% confidence of interval. All of the P values are listed in Supplementary Table 8.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.