Expression and testing in plants of ArcLight, a genetically–encoded voltage indicator used in neuroscience research
- 2.3k Downloads
It is increasingly appreciated that electrical controls acting at the cellular and supra-cellular levels influence development and initiate rapid responses to environmental cues. An emerging method for non-invasive optical imaging of electrical activity at cell membranes uses genetically-encoded voltage indicators (GEVIs). Developed by neuroscientists to chart neuronal circuits in animals, GEVIs comprise a fluorescent protein that is fused to a voltage-sensing domain. One well-known GEVI, ArcLight, undergoes strong shifts in fluorescence intensity in response to voltage changes in mammalian cells. ArcLight consists of super-ecliptic (SE) pHluorin (pH-sensitive fluorescent protein) with an A227D substitution, which confers voltage sensitivity in neurons, fused to the voltage-sensing domain of the voltage-sensing phosphatase of C iona i ntestinalis (Ci-VSD). In an ongoing effort to adapt tools of optical electrophysiology for plants, we describe here the expression and testing of ArcLight and various derivatives in different membranes of root cells in Arabidopsis thaliana.
Transgenic constructs were designed to express ArcLight and various derivatives targeted to the plasma membrane and nuclear membranes of Arabidopsis root cells. In transgenic seedlings, changes in fluorescence intensity of these reporter proteins following extracellular ATP (eATP) application were monitored using a fluorescence microscope equipped with a high speed camera. Coordinate reductions in fluorescence intensity of ArcLight and Ci-VSD-containing derivatives were observed at both the plasma membrane and nuclear membranes following eATP treatments. However, similar responses were observed for derivatives lacking the Ci-VSD. The dispensability of the Ci-VSD suggests that in plants, where H+ ions contribute substantially to electrical activities, the voltage-sensing ability of ArcLight is subordinate to the pH sensitivity of its SEpHluorin base. The transient reduction of ArcLight fluorescence triggered by eATP most likely reflects changes in pH and not membrane voltage.
The pH sensitivity of ArcLight precludes its use as a direct sensor of membrane voltage in plants. Nevertheless, ArcLight and derivatives situated in the plasma membrane and nuclear membranes may offer robust, fluorescence intensity-based pH indicators for monitoring concurrent changes in pH at these discrete membrane systems. Such tools will assist analyses of pH as a signal and/or messenger at the cell surface and the nuclear periphery in living plants.
KeywordsArcLight Electrical signalling Genetically-encoded voltage indicator pH-sensitive indicator Super ecliptic pHluorin
SEpHluorinA227D fused to Ci-VSD
Calcium sensor 12
CALCINEURIN B-LIKE PROTEIN 1. The Arabidopsis CBL1 protein (At4g17615) contains a myristolated glycine and a palmitolated cysteine, which tether the fluorescent fusion protein to the cytoplasmic surface of the plasma membrane
voltage sensing domain of Ciona intestinalis voltage-sensing phosphatase
Förster resonance energy transfer
genetically-encoded voltage indicator
induced transmembrane voltage
RAN GTPASE ACTIVATING PROTEIN 1
super-ecliptic(SE) pHluorin (pH-sensitive fluorescent protein)
super-ecliptic(SE) pHluorin (pH-sensitive fluorescent protein) containing an A227D substitution that confers voltage sensitivity in neurons
SAD1/UNC-84 DOMAIN PROTEIN 2. The Arabidopsis SUN2 protein (At3g10730) has one transmembrane domain that can localize SUN2-fusion proteins at the inner nuclear membrane surface
The WPP sequence contains a Trp(W)-Pro(P)-Pro motif consists of amino acids 28–131 of Arabidopsis RANGAP1 (At3g63130) that is sufficient for targeting fusion proteins to the outer nuclear membrane
Growth, development and appropriate responses to the environment require electrical controls and networks acting at multiple levels of organization within cells, tissues and whole organisms [1, 2, 3]. At the cellular level, changes in transmembrane potentials (electrical voltage gradients) and ion fluxes comprise an extensive system of bioelectrical communication that is integrated with molecular, chemical and mechanical signalling pathways [2, 4]. Together with classical methods for monitoring membrane potentials such as microelectrodes and patch clamp, a new generation of electrophysiological tools is being developed based on the concept of light-based or optical electrophysiology [4, 5]. An important group of these new tools consists of genetically-encoded, protein-based voltage indicators [6, 7, 8].
Genetically-encoded voltage indicators (GEVIs) are composed of a fusion between a fluorescent protein (reporter) and a voltage-sensing domain (detector) . GEVIs have been developed by neurobiologists over the last two decades as a non-invasive method to optically monitor changes in transmembrane potential in single and multiple neurons and other cell types [6, 7, 8, 9]. One type of GEVI is based on Förster resonance energy transfer (FRET) between a pair of fluorescent proteins joined to a membrane-spanning voltage-sensing domain. Changes in membrane potential are thought to act through the voltage-sensing domain to induce more favourable alignment of the two fluorescent proteins, resulting in increased FRET efficiency [8, 9, 10]. By contrast, in monochromatic GEVIs, a transmembrane voltage-sensing domain is fused to a single fluorescent protein that reacts to a voltage change by showing alterations in fluorescence intensity. This has been proposed to result when membrane depolarization triggers movement of the voltage-sensing domain, resulting in deformation of the linked fluorescent protein in a manner that reduces fluorescence intensity .
One intensity-based GEVI is ArcLight [11, 12], which consists of super-ecliptic (SE) pHluorin (pH-sensitive fluorescent protein) [13, 14] containing an A227D substitution conferring voltage sensitivity in neurons  and the voltage-sensing domain of the voltage-sensing phosphatase of Ciona intestinalis (Ci-VSD) . The fluorescence intensity of ArcLight has been reported to change significantly in response to voltage changes at the plasma membrane in mammalian cells . In one study using human embryonic kidney (HEK293) cells, the fluorescence intensity of ArcLight decreased 35 % in response to a membrane depolarization of 100 mV .
One advantage of GEVIs as voltage indicators is that they can be fused to defined membrane targeting motifs, thus allowing electrophysiological analysis of internal cellular membranes that are largely inaccessible to classical tools for measuring membrane potential. Although the membrane potentials of multiple cells can in principle be measured using microelectrode arrays [16, 17], GEVIs also permit noninvasive detection of simultaneous changes in membrane potentials in populations of cells in intact tissues and organs .
We are interested in using GEVIs to study coordinated changes in the electrical potentials of plasma membranes and nuclear membranes of plant cells in response to environmental and developmental stimuli. Owing to their low background fluorescence and interesting developmental features, root cells provide a good experimental system for evaluating the feasibility of GEVIs to study the electrical behavior of different membrane systems in living plants . We described previously the generation of transgenic Arabidopsis thaliana (Arabidopsis) plants expressing FRET-based GEVIs in root cells . FRET-based GEVIs are stably expressed and well-tolerated by Arabidopsis and a recent study documents the successful use of Mermaid FRET sensors to monitor membrane voltage changes in response to exogenous application of potassium in a plant system . In view of former findings for ArcLight in mammalian cells showing large shifts in fluorescence intensity in response to voltage changes [11, 12], we have assembled and introduced into Arabidopsis constructs encoding ArcLight and several derivatives targeted to the plasma membrane and nuclear membranes of root cells. Here we describe the results of experiments designed to assess changes in the fluorescence intensity of ArcLight and derivatives situated in these two membrane systems in response to external ATP (eATP) and other stimuli expected to trigger changes in transmembrane potential .
Transgenic Arabidopsis plants expressing GEVIs and derivatives in root cells
The fluorescent proteins tested include: classic ArcLight (Fig. 1a), which - in the absence of any other membrane targeting sequence - is directed to the plasma membrane by the Ci-VSD (Fig. 1g, sector A); ArcLight joined at the N-terminus to the WPP domain of Arabidopsis RAN GTPASE ACTIVATING PROTEIN 1 (RANGAP1) (Fig. 1b) [22, 23], which promotes targeting to the outer nuclear membrane (Fig. 1g, sector B); and ArcLight fused at the N-terminus to the Arabidopsis SAD1/UNC-84 DOMAIN PROTEIN 2 (SUN2), which contains one transmembrane domain (Fig. 1c) and is able to target the protein to the inner nuclear membrane  (Fig. 1g, sector C).
In other constructs, we tested the importance of the transmembrane Ci-VSD in voltage-sensing by replacing it with either an Arabidopsis CALCINEURIN B-LIKE PROTEIN 1 (CBL1) plasma membrane targeting peptide  at the N-terminus (Fig. 1d), which situates the fluorescent reporter at the cytoplasmic surface of the plasma membrane (Fig. 1g, sector D); an N-terminal WPP domain (Fig. 1e), which places the fluorescent reporter at the cytoplasmic surface of the outer nuclear membrane (Fig. 1g, sector E); or an N-terminal fusion to inner nuclear membrane protein SUN2 (Fig. 1f), which positions the fluorescent reporter in the perinuclear space (Fig. 1g, sector F).
For comparative purposes, we used transgenic plants expressing the intensity-based free calcium concentration sensor Case12 (Calcium sensor 12)  (Fig. 2g) and mCitrine, which has been modified to reduce environmental sensitivity , joined to either Ci-VSD  (Fig. 2h) or CBL1 (Fig. 2i). A GST-tagged SEpHluorinA227D (Fig. 2j) was expressed in E. coli and isolated to test as a soluble variant of ArcLight. The amino acid sequences of wild-type GPF, mCitrine, SEpHluorin and SEpHluorinA227D are shown in Additional file 1: Figure S1.
Transgenic plants expressing Case12 displayed diffuse fluorescence that was particularly strong at the root tip whereas fluorescence was localized at the plasma membrane in root cells of transgenic plant expressing Ci-VSD-mCitrine and CBL1-mCitrine (Additional file 2: Figure S2). Expression of ArcLight and derivatives did not noticeably affect the phenotype of the transgenic plants, which grew and reproduced normally (data not shown).
External ATP (eATP)
All of the observed responses to eATP depended on the fluorescent proteins being in a cellular context because soluble GST-SEpHluorinA227D protein did not display any changes in fluorescence intensity when ATP was added to the solution (Additional file 4: Figure S4, top). In addition, negligible responses to eATP application were observed in plants expressing environmentally-insensitive mCitrine fused to either Ci-VSD or CBL1 (Additional file 5: Figure S5).
ITMV and Light
To determine further effects on ArcLight fluorescence, we tested two additional stimuli that might be expected to provoke changes in membrane potential: induced transmembrane voltage (ITMV) [30, 31] and light [32, 33]. For ITMV experiments, seedlings were placed in a chamber flanked by two electrodes and subjected to an electric pulse of 2.5 V. For experiments using additional light, seedlings were placed in an agarose-pad-chamber and illuminated with various wavelengths of light in addition to continuous illumination at 500/20 nm, which is the excitation wavelength of ArcLight.
Our study was designed to test the feasibility of using the fluorescence intensity-based GEVI ArcLight, which has been used as a voltage indicator in neurons, to monitor voltage changes at the plasma membrane and nuclear membranes in root cells. The membrane-associated fluorescent reporters were expressed well in Arabidopsis root cells. The voltage-sensing Ci-VSD conferred good targeting to the plasma membrane in the absence of additional targeting motifs. For reasons that are not completely clear, the Ci-VSD tended to promote protein aggregation and/or interfere with the specificity of nuclear envelope targeting when a nuclear membrane targeting sequence was also present.
As expected from previous work in neural cells, ArcLight and Ci-VSD-containing derivatives situated in these membrane systems responded robustly to eATP treatments by displaying transient reductions in fluorescence intensity. However, similar reductions in fluorescence intensity were observed with ArcLight derivatives lacking the voltage sensor Ci-VSD, indicating that the observed responses did not rely on voltage-sensing ability of the fluorescent protein. Therefore, decreased fluorescence intensity of ArcLight in response to eATP application in root cells is best interpreted as reflecting the pH sensitivity of its SEpHluorin base. In neurons, the pH sensitivity of ArcLight is less of a concern because H+-fluxes and pH changes during neuronal activity are of minor importance. By contrast, H+-ions contribute substantially to depolarisation and electrical activities in plants .
The decrease of ArcLight fluorescence in response to pH changes following eATP treatment can be understood as follows: The transient depolarization induced by eATP is accompanied by a large increase in free cytoplasmic calcium ion concentration ([Ca2+]cyt), as shown by the transient increase in fluorescence of Case12. Both depolarisation and [Ca2+]cyt transient are the result of cation channel activities, which are mainly K+-channels, but these are rather nonspecific and can also conduct H+ ions. Since there is a membrane potential (negative inside the cell with respect to the outside) and a pH-gradient between the outer medium and cytoplasm (the pH of the apoplast is normally between 4.5 and 6.5 , whereas cytoplasmic pH is usually around 7.3 ), protons run down their electrochemical gradient upon cation channel opening, enter the cell, and acidify its internal contents. The SEpHluorin component of ArcLight responds to H+-ion plumes near the membrane and to cytoplasmic acidification, resulting in reductions in fluorescence. In this scenario, ArcLight responds primarily to the downstream consequence of a membrane voltage change (decreased pH) and not directly to the voltage change itself.
In summary, although ArcLight and the derivatives tested here do not provide a direct sensor for voltage changes in plants, they can potentially be used as fluorescence intensity-based, membrane-localized indicators of pH changes at the cell surface and nuclear periphery. These fluorescence intensity-based pH indicators display robust responses and, following further validation and calibration, may provide facile alternatives to ratiometric-based pH indicators based on GFP . The development of monochromatic GEVIs for use in plant systems will require the identification of fluorescent reporter proteins that are less sensitive than ArcLight to changes in pH.
Transgenic plants expressing fluorescent protein reporters
The nucleotide sequence of ArcLight, codon-optimized for expression in Drosophila melanogaster , was obtained from Dr. Michael Nitabach (Yale University) and then codon-optimized by our lab for Arabidopsis and synthesized by GeneScript. Transgenic constructs (Fig. 2) were produced using standard molecular biology techniques. The transgenes plus promoter (either Ubi10  or Rps5 ) were assembled on modified pBC plasmids (Stratagene, Cat. Nr. 212215) between Sal1 and XhoI sites, and the entire transgene was inserted into the Sal1 site of binary vector pZP221 . The respective binary vectors containing each transgene construct were introduced via Agrobacterium-mediated transformation into Arabidopsis thaliana ecotype Columbia-0 using the floral dip method . All transgenic lines used in this study were generated in our lab. The Ubi10 promoter drives expression in the whole root, including root hairs. The RPS5 promoter is expressed primarily in the division zone-transition zone. The data shown are from lines that showed the best expression.
Seeds of transgenic plants were surface sterilized in 1.5 ml Eppendorf tubes by shaking them for 20 min in 1 ml of 70 % ethanol solution containing Triton X-100 (50 μl per 100 ml 70 % ethanol). The seeds were centrifuged in an Eppendorf centrifuge for 1 min, the supernatant removed, and the seeds were resuspended in 1 ml of 100 % ethanol and immediately pipetted onto a filter paper disk in a sterile hood. After air-drying for 1 h, seeds were sprinkled onto sterile, solid Murashige and Skoog (MS) medium in petri dishes, stratified by storing the plate at 4 °C for 3 days, and then transferred to a light incubator (23 °C, 16 h light, 8 h dark) for about one week of growth before being used for experiments as described.
Expression of SEpHluorinA277D in E. coli
We expressed GST-tagged SEpHluorinA277D in E. coli strain BL21 (NEB, USA) using the expression vector pET-42a(+), which contains a GST-tag and multiple cloning site (Novagen, USA). The SEpHluorinA227D sequence was amplified using PCR as a BamHI/HindIII fragment and cloned in frame into pET-42a(+) and introduced into BL21 cells. Production of GST-tagged SEpHluorinA227D protein was induced with isopropyl β-D-1-thiogalactopyranoside (IPTG) on agar plates overnight. The GST-tagged SEpHluorinA277D was isolated using the BugBuster GST-Bind Purification Kit (Novagen 79794–3 REF) and purified using the small scale batch method according to the manufacturer’s instructions.
Confocal images of seedlings growing on sterile, solid MS medium in petri dishes were acquired using the Leica TCS LSI microscope equipped with a 5x Z16 APO A zoom system (Leica Microsystems CMS GmbH, Germany, purchased from Major Instruments, Taiwan).
Fluorescence imaging and data processing
Fluorescence changes were recorded with a fast CCD imaging system, MiCAM02-HR, which is specialized for both calcium ion and membrane voltage imaging applications (Brainvision Inc., Japan, purchased from Major Instruments, Taiwan), mounted on an Axiovert25 inverted fluorescence microscope (Carl Zeiss GmbH, Germany) equipped with 5x/0.12 (CP-Acromat), 20x/0.8 (Plan-Apochromat) and 40x/0.9 Pol (EC Plan-Neofluar) objectives and either an ET YFP filter cube (Ex ET 500/20, beam splitter T515p, Em ET 535/30) or an FITC filter cube (Ex HQ 480/40x, beam splitter Q 505 LPe, Em HQ 535/50 m). The light source for the microscope was a xenon short arc lamp without reflector, model XBO 150 W/CR OFR (OSRAM GmbH, Germany) housed in an OptoSource illuminator (Cairn Research Ltd., U.K. purchased via Major Instruments, Taiwan). Displays of results were obtained using Metamorph (Meta Imaging Series Software, Molecular Devices, USA) (Figs. 4, 8; Additional file 6: Figure S6) or data analysis software BV_Analyzer (ver1312) (Brainvision Inc., Japan) (Figs. 5, 6, 7; Additional file 3: Figure S3; Additional file 4: Figure S4; Additional file 5: Figure S5). In the latter case, the fractional fluorescence changes (dF/Fmax) shown in the figures were calculated using the processing function of BV_Analyzer. All experiments were performed multiple times and representative results are shown. After seedlings recovered from the experiments, they could be transferred to soil, where they grew and reproduced normally.
External ATP (eATP)
For addition of eATP to seedlings, we used a microscope slide-sized, open-top bath chamber (a gift from Dr. Kai Konrad, University of Würzburg, Germany) made of 3 mm thick plexiglass (7.6 × 2.5 cm) with an oval, bevelled indentation (4.5 × 1.8 cm) and a large cover slip (6 × 2.5 cm) glued on the bottom. This chamber was filled with 1 % agarose in 1x imaging solution [5 mM potassium chloride, 10 mM MES hydrate, 10 mM calcium chloride, adjusted to pH5.8 with Tris(hydroxymethyl)aminomethane] [29, 43] to make tight fitting agar blocks. After solidification at room temperature, the agar block was removed and transferred to a round petri dish, where it was kept in 1x imaging solution until use, when it was cut into approximately 1 cm broad slices.
To mount an Arabidopsis seedling (1–2 weeks old) for fluorescence imaging, 400 μl 1x imaging solution was pipetted into the open-top chamber and an intact seedling removed from MS medium was positioned lengthwise into the imaging solution. A 1 cm slice of the agar block (as prepared above) was then placed on top of the area of the extended root to be imaged, leaving the root tip protruding on one side and the leaves on the other side. With fine forceps, the seedling was pulled carefully until the root tip was just under the agar block. Excess solution around the agar block was removed. The chamber was then mounted on the inverted microscope. A silicon tube was attached on one end to a Gilson Pipetman and - after filling the tube with 50 μl 2 mM ATP in 1x imaging solution or buffer for buffer-only control experiments as indicated in the figures - the other end was attached to a holder that is positioned just above the edge of the agar block under which the root has been positioned (Additional file 7: Figure S7A). Fluorescence recording was then started and the eATP solution was pipetted into the bath chamber at a specified time.
To test soluble GST-SEpHluorinA227D, 200 μl of the isolated protein in elution buffer EB (BugBuster GST-Bind Purification Kit, Novagen) (protein concentration approximately 30 μg/ml) was placed in the Bügelkammer (see section on ITMV) and 50 μl eATP solution, which also contained 30 μg/ml GST-tagged SEpHluorin to maintain the protein concentration, was added from a silicon tube positioned just above the edge of the cover slip.
ITMV (induced transmembrane voltage)
For application of voltage pulses, we used a Grass SD stimulator (Grass, USA) connected to two electrodes mounted on a slide holder (Bügelkammer, Krüss GmbH, Germany). The two electrodes of the Bügelkammer are mounted on separate plastic stirrups and can be clamped down individually over the sample, thus positioning the electrodes at a distance of 200 μm. A 24 × 40 mm cover slip is placed in the open Bügelkammer, 200 μl 1x imaging solution [29, 43] is placed on the cover slip, and the first electrode is clamped down on the cover slip. A seedling is placed horizontally in the imaging solution above the first electrode and the second electrode is clamped down onto the cover slip such that the root is under the second electrode. An 18x18 mm cover slip is then placed on top of the root for stabilization (Additional file 7: Figure S7B). The area of the root between the two electrodes was imaged (see for example, Fig. 8a). To test soluble GST-SEpHluorinA227D, the isolated protein in elution buffer EB (BugBuster GST-Bind Purification Kit, Novagen) (protein concentration approximately 30 μg/ml) was placed in the Bügelkammer and covered with an 18x18 mm cover slip.
To test the influence of additional light pulses on fluorescence intensity of ArcLight and derivatives (during continuous illumination at 500/20 nm, the excitation wavelength of ArcLight), a SPECTRA X Light Engine (Lumencor, USA, purchased from Major Instruments, Taiwan) was used for illumination at 390/18 nm (violet), 438/24 nm (blue), 475/28 nm (cyan), 589/15 nm (near red) and 632/22 nm (far red). The additional light at these wavelengths was shone at an angle of 30° and from a distance of 10 cm on a seedling mounted under an agar block in the open-top chamber as described for eATP experiments. The light intensity was set at 10 % on the SPECTRA X Light Engine and focused on the imaging plane. Automated turning on and off at certain wavelengths of the SPECTRA X Light Engine and at specific time points was achieved by using GhostMouse software (http://tw.vrbrothers.com) or was performed by hand. To test soluble GST-SEpHluorinA227D, the isolated protein in elution buffer EB (BugBuster GST-Bind Purification Kit, Novagen) (protein concentration approximately 30 μg/ml) was placed in the Bügelkammer, covered with a 18x18 mm cover slip, and illuminated with the SPECTRA X Light Engine as described above.
We thank Map Chen for helpful advice and discussions; Jason Fu for microscope administration; Dr. Hiroo Fukuda, University of Tokyo, for SUN2; Dr. Roger Deal, Emory University, for WPP; Dr. Thomas Knöpfel, Imperial College, London, for V3.1-mCitrine; and and Dr. Michael Nitabach, Yale University, for the sequence of ArcLight. We are grateful to the anonymous reviewers of the first version of this paper for many helpful comments.
Financial support for this work has been generously provided by Academia Sinica (AS) and the Institute of Plant and Microbial Biology (AS).
- 8.Patti J, Isacoff E. Measuring Membrane Voltage with Fluorescent Proteins. Cold Spring Harb Protoc. 2013;7:606–13.Google Scholar
- 20.Grefen C, Karnik R, Larson E, Lefoulon C, Wang Y, Waghmare S, et al. A vesicle-trafficking protein commandeers Kv channel voltage sensors for voltage-dependent secretion. Nat Plants. 2015, DOI: 10.1038/NPLANTS.2015.108
- 43.Loro G, Costa A. Imaging of mitochondrial and nuclear Ca2+ dynamics in Arabidopsis roots. Cold Spring Harb Protoc. 2013;8:781–5.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.