Evolution of long-distance signalling upon plant terrestrialization: comparison of action potentials in Characean algae and liverworts

Abstract

Background

In this review, we summarize data concerning action potentials (APs) – long-distance electrical signals in Characean algae and liverworts. These lineages are key in understanding the mechanisms of plant terrestrialization. Liverworts are postulated to be pioneer land plants, whereas aquatic charophytes are considered the closest relatives to land plants. The drastic change of the habitat was coupled with the adaptation of signalling systems to the new environment.

Scope

APs fulfil the ‘all-or-nothing’ law, exhibit refractory periods and propagate with a uniform velocity. Their ion mechanism in the algae and liverworts consists of a Ca²⁺ influx (from external and internal stores) followed by/coincident with a Cl^– efflux, which both evoke the membrane potential depolarization, and a K⁺ efflux leading to repolarization. The molecular identity of ion channels responsible for these fluxes remains unknown. Publication of the Chara braunii and Marchantia polymorpha genomes opened up new possibilities for studying the molecular basis of APs. Here we present the list of genes which can participate in AP electrogenesis. We also point out the differences between these plant species, e.g. the absence of Ca²⁺-permeable glutamate receptors (GLRs) and Cl^–-permeable SLAC1 channel homologues in the Chara genome. Both these channels play a vital role in long-distance signalling in liverworts and vascular plants. Among the common properties of APs in liverworts and higher plants is their duration (dozens of seconds) and the speed of propagation (mm s^–1), which are much slower than in the algae (seconds, and dozens of mm s^–1, respectively).

Conclusions

Future studies with combined application of electrophysiological and molecular techniques should unravel the ion channel proteins responsible for AP generation, their regulation and transduction of those signals to physiological responses. This should also help to understand the adaptation of the signalling systems to the land environment and further evolution of APs in vascular plants.

INTRODUCTION

Approximately 450 million years ago in the Middle Ordovician, plants began colonization of lands (Ishizaki, 2017). Charophytes are regarded as the ancestors of the early land plants (Bowman et al., 2017; Ohtaka and Sekimoto, 2022), with striking similarities and important differences, which were confirmed by comparative genome analysis after publication of the Chara braunii genome (Nishiyama et al., 2018). Paleobotanic, morphological and biochemical evidence, supported recently by genetic data, indicates that terrestrialization took place owing to adaptation of bryophytes, liverworts, mosses and hornworts to new environmental conditions. The adaptation processes covered morphological and physiological changes in pioneer plants. In comparison with a freshwater habitat, land plants experience drought stress, fast and profound temperature changes, mechanical stress due to wind and soil particles, UV radiation, toxic compounds and many other factors. Land plants also seem to be more susceptible than water plants to biotic stress factors such as herbivores, fungi, bacteria and viruses. Of great importance thus was the adaptation of signalling processes to the new environmental challenges. Among the signals generated by plants, there are relatively fast long-distance signals possessing an electrical component based on ion fluxes. Their postulated role is to ‘inform’ whole plants or their organs about local stress factors. The signals can also be evoked by stress factors such as illumination or temperature changes affecting the whole plant at the same time. The signals cover action potentials (APs) and variation potentials (VPs), also known as slow wave potentials (SWPs) (Trębacz et al., 2006; Beilby, 2007; Fromm and Lautner, 2007; Król et al., 2010; Gilroy et al., 2016; Hedrich et al., 2016; Huber and Bauerle, 2016; Vodeneev et al., 2016). In monocots, so-called system potentials (SPs) were occasionally reported (Zimmermann et al., 2009).

Action potentials in plants, similarly to those in excitable animal cells, are the responses to relatively mild, non-damaging stimuli. Individual plants can generate dozens of APs in a short time limited by refractory periods – absolute, during which subsequent excitation is not possible; and relative, whose duration depends on the stimulus strength (Trębacz et al., 2006; Król et al., 2010). Many plants generate APs spontaneously, without any visible stimulation. There are also plants generating long-lasting trains consisting of dozens of APs (Favre et al., 1999; Shepherd et al., 2008; Kisnieriene et al., 2012; Stolarz and Dziubinska, 2017). As in animal counterparts, plant APs fulfil the ‘all-or-nothing’ law, which means that an AP of relatively constant amplitude is generated, no matter how strong the stimulus was, provided it was higher than a certain threshold. The next important feature distinguishing APs from other signals is the relatively fast and uniform propagation over long distances. APs are generated by parts of plants, e.g. isolated leaves, petioles, stamens or pistils, or even individual isolated cells, in contrast to VPs, which can be evoked in whole plants (van Bell et al., 2014). VPs appear in response to application of damaging stimuli, e.g. cutting or burning, exhibit an irregular time course, propagate slowly and with decrement, and require vasculature as a route of transmission (van Bell et al., 2014; Huber and Bauerle, 2016). This limits their occurrence to vascular plants.

Unravelling the similarities and differences in the features of electrical signalling between Characeae and liverworts could offer invaluable insights into the event of terrestrialization. Their adaptations to reside in aqueous and terrestrial habitats render comparative studies especially relevant. Due to the vastly dissimilar morphological features, different electrophysiological techniques are more or less conventionally applicable to both plant groups, thus making direct comparisons strenuous. As a foundation of plant electrophysiology, APs are readily measured and well characterized in both Characean algae and liverworts, thus making these long-range electrical signals prospective targets for comparative analysis.

In this review, we made an attempt to summarize basic knowledge about APs in Characean algae and liverworts on the basis of electrophysiological data in the beginning of molecular studies.

UNIQUE CHARACEAN CELLS PROVIDE A PLETHORA OF APPROACHES FOR ELECTROPHYSIOLOGICAL INVESTIGATIONS

A group of branched, plant-like, multicellular green Characeae algae has been used as a model system in plant electrophysiology research for the past century. Found in fresh and brackish waters, the Characean algae include a continuous spectrum of organisms from very salt-sensitive to very salt-tolerant species. They grow in thin-segmented shoots that sprout whorls of branchlets every few centimetres (Fig. 1). The single cylindrical internodal cell with a diameter up to 1 mm and a length that can exceed 10 cm is the most studied part of Characean algae. The first ever recording of an AP with an intracellular microelectrode was performed in a Nitella cell (Umrath, 1930). Eventually, electrical measurements on intact or altered cells and cytoplasmic droplets conducted using various experimental techniques unique to Characean plants have formed the basis for modern plant electrophysiology. The gigantic size of the internodal cells facilitated the use of perfusion and permeabilization methods, creating access to both sides of the plasma membrane and tonoplast in a single-cell context (Tazawa, 1964; Shimmen and Tazawa, 1983; Beilby, 2016). Shimmen and MacRobbie (1987) employed the permeabilization technique to disintegrate the plasma membrane by removing external and cell wall-embedded Ca²⁺ by ethylene glycol tetraacetic acid (EGTA). While the chloroplasts were disrupted, the tonoplast and the vacuolar compartment were largely unchanged. Vacuolar perfusion of Characean cells with artificial solutions, enabling studies of various functions of the tonoplast, was explored by Japanese scientists (Mimura and Tazawa, 1983; Tazawa et al., 1987).

The simple morphology of Characean cells also offers a good experimental system to explore two-electrode current clamp and voltage clamp techniques. High temporal resolution recordings and analysis of real-time processes in the current clamp mode (such as changes in the AP excitation threshold, amplitude and velocity of repolarization) can indicate the effect of many bioactive substances on AP generation (Kisnieriene et al., 2018). The voltage clamp technique is routinely used in studies of voltage-sensitive ion channels, including measurements of the dependence of their activation and inactivation on voltage, dynamics of ion channel opening and closing, ion permeation and inhibition by both organic and inorganic substances. The main advantage of this technique is the possibility to measure the amount of the ionic current crossing a membrane at any given voltage in real time in vivo. This method is the most reliable source of information about the activity of separate ion transport systems in the whole cell in a steady state and during excitation. The two-electrode voltage clamp technique with a single impalement when the membrane is clamped to the desired voltage by passing just enough current through the cell using external Ag/AgCl wires was developed and elaborated (Findlay and Hope, 1964a; Lunevsky et al., 1983). This method preserves cell integrity and minimizes impalement damage to the cell wall, membranes and cytosol. Voltage clamp has mostly contributed to the determination of ion currents involved in generation of APs in Characean algae (Berestovsky and Kataev, 2005).

The application of the patch clamp technique in plants is hindered by the cell wall, which limits direct access to the channels of the plasma membrane. Despite this problem, the first published single-channel patch clamp recordings in plants were performed in turgid cells of Nitellopsis obtusa (Krawczyk, 1978). Challenges were alleviated by microsurgical or laser-based methods of cutting the cell wall to access the plasma membrane (Laver, 1991; Boer et al., 1994). For the tonoplast investigations, the cytoplasmic droplet technique was developed and has been used extensively in single-ion channel research for decades (Lühring, 1986; Sakano and Tazawa, 1986; Bertl, 1989; Katsuhara et al., 1991).

The simple structure of the Characean thallus allows direct measurements of cell to cell electrical signalling transmission along internodal cells. Tandem internodal cells provide a model system for complex multi-parametrical investigation of instantaneous alterations in generation, propagation and transmission of APs in vivo, thus contributing to understanding more complex laws of functionality, adaptation and information processing in higher plants.

The mathematical modelling approach allowing confirmation of the mechanisms of Characean AP generation was first adopted when Beilby and Coster (1979) fitted the Hodgkin and Huxley model to Chara australis APs, replacing the inflow of Na⁺ as the depolarizing ionic flux with the outflow of Cl^–. Thiel and colleagues (Biskup et al., 1999; Wacke et al., 2003) also harnessed a paradigm from animal systems based on voltage-dependent production of a long-lived second messenger. It eventually developed into the Thiel–Beilby model of excitation in Characean algae, which supports the idea of two sources of Ca²⁺ in Chara APs (Beilby and Al Khazaaly, 2016; Kisnieriene et al., 2019).

The new era of excitability investigations started when the C. braunii genome was published (Beilby, 2019). Genetic data henceforth will link the wide-ranging electrophysiological studies that have been carried out for a century with the molecular understanding of electrical signalling (Quade et al., 2021).

Action potentials in giant Characean cells

The changes in the membrane potential induced by many biotic and abiotic factors vary between Characean genera, but they have been extensively explored in excitability investigations. Now it is well established that the internodal cells fire APs in response to such stimuli as electrical depolarization, mechanical stimulation, light, cold and heat shock (Beilby, 2007). Based on the thermodynamic properties of the membrane, it is predicted that APs can be triggered by an increase in pressure, a decrease in temperature and a decrease in pH, because these parameters shift the system state into a phase transition (Fillafer et al., 2021).

Action potentials are electrical events consisting of large transient changes in membrane polarization. The positive feedback and all-or-nothing law that APs follow in plants were proved to facilitate the large size and particular anatomy of Characean algae. During APs, an efflux of chloride ions from the cell together with an influx of calcium ions are responsible for the depolarization phase, with the main depolarization current driven by the Cl⁻ efflux.

There is an ongoing debate about the origin of Ca²⁺ ions entering the cytosol, as they may enter from the external environment through the apoplast or from intracellular resources, such as the endoplasmic reticulum (ER) (Kikuyama et al., 1993), the vacuoles to some extent (Lunevsky et al., 1983), etc. (Tazawa and Kikuyama, 2003). There is evidence that a Ca²⁺ influx via plasma membrane channels is involved and that Ca²⁺ release from internal stores is relevant. The coupling of the two processes may cause complex excitability regulation in Characean algae via the so-called calcium-induced calcium release (CICR). Patch clamp data suggest that the relationship between the number of Ca²⁺-discharging stores released and the size of the affected ensemble of Cl^– channels determines the magnitude of the local observed Cl^– efflux. Also, the two described chloride channel types showed bursting behaviour – a consequence of the quantal release of Ca²⁺ from stores in the vicinity of the plasmalemma (Thiel et al., 1993; Thiel and Dityatev, 1998). Now it is accepted that the initiation of AP generation is not entirely based on the time- and voltage-dependent activation properties of plasma membrane ion channels but on a complex signal transduction cascade (Beilby and Al Khazaaly, 2016).

In most Characean species, the AP peak does not reach a positive membrane potential (Beilby, 2007). AP peak potential depends on the Cl^– channels which tend to close near zero membrane potential (Beilby and Shepherd, 2006), and the timing of the activation of K⁺ channels by depolarization of the membrane potential difference (PD) (Beilby, 2007). In Chara and Nitellopsis, the AP peak potential remained relatively constant, despite the increase of Cl⁻ concentration in the medium up to 100 mm (Kisnieriene et al., 2019). However, it was shown that, if the external Ca²⁺ concentration is increased, the AP peak becomes positive (Findlay, 1962; Kisnieriene et al., 2019). Using tonoplast-free Chara cells, Shimmen and Tazawa (1980) demonstrated that the AP peak is not sensitive to the internal Cl⁻ concentration. In perfused cells without an intact tonoplast, the activation of Cl⁻ channels is hindered. A required second messenger cascade could be impaired during the disintegration of the tonoplast (Tazawa et al., 1987). Uniquely, when the electrode is impaled in the vacuole of Nitellopsis cells, positive AP peaks are routinely recorded (Kisnieriene et al., 2016; Lapeikaite et al., 2019). This may be related to the higher calcium concentration in the apoplast or a more conservative Ca²⁺ pump (Kisnieriene et al., 2019).

Membrane potential repolarization occurs upon an efflux of K⁺ ions from the cell (Beilby and Al Khazaaly, 2016). The increase in K⁺ permeability lags behind the rise in Cl⁻ permeability, reaching a peak approx. 2 s after the latter in Chara corallina (Thiel, 1995). The rates of AP repolarization are variable and depend not only on the functioning of the outward-rectifying potassium channels K⁺_out, but also on the activity of H⁺-ATPase and Ca²⁺-ATPase (Thiel et al., 1997; Kisnierienė et al., 2012, 2019). The Nitellopsis AP lasts longer compared with that of Chara, with a more gradual repolarization indicating a different composition and/or functioning of these ion transport systems (Kisnieriene et al., 2019).

The AP typically lasts about a few seconds and exhibits an absolute refractory period of around 3 s and a relative refractory period of about 6–60 s (Beilby and Coster, 1979). The refractory period that follows stimulation may result from inactivation of anion channels through phosphorylation (Johannes et al., 1998) or due to the kinetics of calcium channels governed by Ca²⁺ and inositol 1,4,5-trisphosphate (IP₃) (Wacke et al., 2003).

The propagation of plant APs was also revealed in Characean algae. APs with similar amplitude and temporal parameters are propagated independently of the stimulation strength if the membrane is sufficiently depolarized. The AP is conducted in both directions away from the place at which the plasma membrane is depolarized over the threshold. A propagation velocity of 10–20 mm s⁻¹ was established, and unidirectionality conditioned by channel inactivation was confirmed (Tabata and Sibaoka, 1987). The conduction rate is dependent on the conductivity of the medium (Sibaoka, 1958).

Plasmodesmata in Characean nodes provide pathways for direct communication between the cytoplasm of adjacent cells. In the short term (i.e. the duration of an AP), there is no discernible effect of elevated internal calcium on nodal resistance (Reid and Overall, 1992). In Characean algae, the AP transmission velocity was reported to be 11–44 mm s^–1, depending on the species (Beilby, 2007). The node resistance varies in different systems: approx. 1.7 kΩcm² in Nitella (Spanswick and Costerton, 1967), 0.06–0.12 kΩcm² in the nodes between young Chara branch cells and 0.2–0.51 kΩcm² in older branch cells (Reid and Overall, 1992). Sibaoka and Tabata (1981) investigated AP transmission across the nodes of C. braunii by inserting an additional electrode in one of the large nodal cells. The whole nodal cell was not excitable, but the adjacent area of the nodal cell (end-membrane) displayed APs.

Possibility to quantify the onset of electrical signalling by measuring the excitation threshold in characean cells

Action potentials are produced when the membrane potential exceeds a threshold value (E_th), above which voltage-gated ionic channels open and let a positive charge enter or a negative charge leave the cell, causing an abrupt membrane depolarization. APs are events, i.e. they have an onset time; consequently, they are inherent electrical signals characterized by an excitation threshold, which can be quantified. The Characean model system presents a unique possibility for direct evaluation of the excitation threshold due to the precise current clamping in the intact giant cell (Lapeikaite et al., 2019). E_th is variable between individual Characean cells and does not depend directly on the resting potential, which consists of an active ATP-driven potential and a diffusive passive potential due to the asymmetrical distribution mainly of potassium ions between the external medium and the cell interior (Kisnieriene et al., 2018; Lapeikaite et al., 2020). However, the excitation threshold depends on the activation of calcium-permeable transport systems (Beilby, 2007) and the dependence of chloride channels on voltage and [Ca²⁺]_cyt (Thiel et al., 1997; Biskup et al., 1999). Thus, the excitation threshold in Characean algae depends on the cytoplasmic calcium content. The increase in the cytoplasmic calcium concentration is caused by the influx of Ca²⁺ from the apoplast and/or internal stores. The latter is presumably governed by a second messenger signalling cascade. The generation of this second messenger, which was thought to be IP₃, is regarded to be voltage dependent and to proceed via an unknown mechanism (Wacke et al., 2003). In plants, no IP₃ receptors (IP₃Rs) have been identified; hence, other second messengers might translate depolarization to the Ca²⁺ release (Beilby, 2019).

It is known that transferring Chara from light into dark results in a slowly progressing elevation of the resting [Ca²⁺]_cyt (Plieth et al., 1998). On the basis of the rise in [Ca²⁺]_cyt, the model of the membrane excitation predicts that the depolarization of the AP occurs faster in dark- than in light-adapted cells (Baudenbacher et al., 2005).

The possibility of determining the excitation threshold and its alterations in N. obtusa cells allows linking the effect of many chemicals to the Ca²⁺-dependent initiation of excitation. The verapamil-induced lag of depolarization coincides with depolarized E_th (Koselski et al., 2021). Similar effects, indicating calcium signalling impairment, have been recorded under exposure to Ni²⁺ (Kisnieriene et al., 2016). Application of glutamate (Glu) and NMDA (N-methyl-d-aspartic acid), indicating possible binding to GLR-like channels, resulted in hyperpolarized E_th (Lapeikaite et al., 2019). It could be assigned to additional activation of Ca²⁺-permeable channels leading to an elevated [Ca²⁺]_cyt/transmembrane calcium gradient as suggested in Dionaea muscipula (Krol et al., 2006).

Depolarization of membrane PD to the excitation threshold may set off trains of APs, leading to further loss of K⁺ and Cl^–. Repetitive firing in N. obtusa cells when the membrane potential reaches a certain level of depolarization after application of various compounds has been observed (Kisnierienė et al., 2009). The occurrence of a spontaneous AP can be mostly explained by the depolarization of the membrane PD to and above the excitation threshold. This depolarization is a result of increased background conductance and progressive inhibition of the proton pump (Shepherd et al., 2008). From the evolutionary perspective, the E_th is maintained at distant voltage ranges from a resting membrane potential to ensure that only significant environmental stimuli could elicit APs.

Excitability of the Characean tonoplast

Cells of Characeae have large central vacuoles occupying up to 90 % of the cell volume. The existence of two membranes, the plasmalemma and the tonoplast, in Characean cells makes the analysis of excitability more complex. Microelectrodes can be positioned in the cytoplasm and in the vacuole, and then the membrane PD is measured separately across the plasma membrane and the tonoplast (Kikuyama, 1986).

Since the first attempts to record tonoplast APs in the large excitable cells of Nitella and Chara algae failed, it was thought that the tonoplast is electrically unexcitable, and the excitation current arises only across the plasmalemma (Findlay, 1959). Later it was revealed in freshwater Chara that the plasma membrane AP is always accompanied by a tonoplast AP (Findlay, 1970; Lunevsky et al., 1983). The idea that there must be a coupling mechanism between the two membranes was widely accepted (Findlay, 1970; Kikuyama and Tazawa, 1976; Shimmen and Nishikawa, 1988). The membrane potential of the vacuolar membrane hyperpolarizes from –10 mV (protoplasmic side negative) to approx. –50 mV (protoplasmic side negative) during an AP (Kikuyama, 1986). The tonoplast component of the AP in N. obtusa is larger than that in C. australis (Kisnieriene et al., 2019). The value of E_th initiating tonoplast excitation varies between –10 and –30 mV (Lunevsky et al., 1983). High [Ca²⁺]_cyt also acts as a signal inducing a tonoplast AP. Microinjection of Ca²⁺ into the protoplasm induces an AP at the vacuole membrane. It has been shown that, in a solution lacking Ca²⁺ and containing Ba²⁺, Mg²⁺ or Mn²⁺, Chara cells lose their excitability in time, resulting in disappearance of the tonoplast AP (Kikuyama, 1986). Thus, it was assumed that [Ca²⁺]_cyt plays an essential role in the generation of APs in the tonoplast. Furthermore, a rise in Ca²⁺ has been shown to cause a transient change in both plasma membrane and tonoplast potentials in Nitella (Shimmen and Nishikawa, 1988).

The Cl^– permeability of the tonoplast increases during its excitation. The electrochemical gradient across the tonoplast can facilitate movement of Cl^– from the cytosol to the vacuole through channels or transporters, even if the vacuolar Cl^– concentration is higher than that of the cytosol. It was suggested that Ca²⁺, being the most probable candidate for inducing the tonoplast AP in vivo, is also required for the regulation of Cl^– channels in the tonoplast, thus the Cl^– channel activation by Ca²⁺ is assumed to be responsible for generating the tonoplast AP (Findlay and Hope, 1964b; Kikuyama and Tazawa, 1976). After confirmation that the tonoplast surrounds cytoplasmic droplets formed by cutting an internodal cell and immersing its end in vacuolar sap-like medium (Lühring, 1986), tonoplast ion channels participating in excitation were investigated. In tonoplast-free cells, the mechanism of Cl^– channel activation seems to be impaired (Tazawa et al., 1987)

The increased Ca²⁺ concentration on the protoplasmic side of the vacuolar membrane causes the vacuolar membrane to depolarize for as long as Ca²⁺ is present (Wayne, 1994). Since chloride channels are activated by free calcium ions, inactivation of these channels is determined to a large extent by an asymptotic decrease in the Ca²⁺ concentration due to adsorption of these ions by cytoplasm molecules and cellular organelles, in particular by mitochondria, and due to active transport to the vacuole (Lunevsky et al., 1983).

Similarly to the plasmalemma AP, the repolarization phase of the tonoplast AP also seems to depend on the activity of Ca²⁺-activated K⁺ channels (Lunevsky et al., 1983). Some likely candidates have been reported via patch clamp studies of the cytoplasmic droplet membrane (Lühring, 1986; Laver and Walker, 1991; Pottosin and Andjus, 1994).

The AP at the tonoplast is usually associated with that at the plasmalemma, although some Nitella and Nitellopsis cells experience tonoplast APs only (Findlay, 1962; Kisnieriene et al., 2019). Tonoplast excitation and a separate set of ion transport systems may be favoured by evolution to ensure an additional way to regulate AP variability and the role of excitation in osmotic regulation.

Genes involved in Characean electrical signalling

The C. braunii genome appears to be a good research target to fill the gap in knowledge between Chlorophyta and land plants and to identify patterns in the evolution of ion transport system functions and regulation mechanisms in plants. The genome project has renewed the interest in Characean algae and has provided a new meaning and a new perspective for electrophysiological investigations. Electrophysiological evidence has proved the existence of specific voltage-gated and mechano-activated channels as well as pump currents in Characean cells (Tsutsui et al., 1987) despite the lack of knowledge about the molecular identity and genes of channels underlying these activities, but the new resources are providing new possibilities (Beilby, 2016).

The Ca²⁺ ion channels, which are the primary routes for Ca²⁺ influx into the cytoplasm during the generation of APs, remain unidentified. Typical animal-like Ca²⁺channels, such as the four-domain voltage-activated Ca²⁺ channels (VDCCs) and IP₃Rs, are absent in the C. braunii genome. Homologous sequences for voltage-gated Ca²⁺ channels are, however, present in green algae up to Charophytes (represented by Klebsormidium flaccidum, http://www.plantmorphogenesis.bio.titech.ac.jp/~algae_genome_project/klebsormidium/), but they do not carry the major depolarization current, although it may have a role in mediating the initial Ca²⁺ influx (Edel et al., 2017). This situation suggests that VDCCs have been lost during the colonization of land but before the evolution of more advanced Streptophyta.

Glutamate receptor-type ligand-gated cation channels were not found in the Chara genome, even though N. obtusa appears to respond to classical GLR agonists such as amino acids or NMDA (Lapeikaite et al., 2019, 2020).

Characean AP is highly intertwined with response to mechanical stress and stoppage of cytoplasmic streaming (Brunet and Arendt, 2017), thus mechanosensitive ion transport systems should participate in AP generation. Indeed, the C. braunii genome encodes several putative touch/mechano-sensitive (MS) channels: two members of the MscS-like (MSL) family (CHBRA34g00550 and CHBRA470g00310) and an orthologue of the eukaryote-specific Piezo-type channel CHBRA630g00070. Two copies of the Ca²⁺ release-activated Ca²⁺ (CRAC) channel (CRAC-C) family ORAI1 (CHBRA282g00400) and ORAI2 (CHBRA52g00600) have been distinguished. These genes could facilitate auto-amplified calcium-induced calcium release in Charophyta, although the mode of generating the CICR is unclear since the ORAI calcium sensors STIMs (stromal interaction proteins) involved in animal CICR are absent in plants. No animal-like voltage-gated Na⁺ channels were identified, as was expected from electrophysiological data.

Extensive investigations of the ionic basis of Characean AP resulted in the conclusion that Ca²⁺-activated anion channels must reside at the plasma membrane. More primitive organisms (e.g. Chlamydomonas) utilize Ca²⁺- and sometimes Na⁺-based APs only (Harz and Hegemann, 1991; Taylor, 2009), thus, anion channels are an evolutionary new player in AP generation. Patch clamp studies on intact Characean cells confirmed that Cl^– transients are caused by ensembles of two types of Cl^– channels which are abundant (10–20 channels µm^–2) in the plasma membrane, i.e. 15 pS and 38 pS (with 100 mm Cl^– in the pipette) (Okihara et al., 1991; Homann and Thiel, 1994). Cl^– channels with a conductance of approx. 21 pS were also detected in cytoplasmic droplets surrounded by the tonoplast (Tyerman and Findlay, 1989). Cl^– channels, which depend on voltage and extracellular Ca²⁺, play a principal role in voltage-triggered membrane excitation (Thiel et al., 1993).

Genome analysis confirmed that a single aluminium-activated malate transporter (ALMT)-type anion channel gene, CHBRA577g00190, is present in C. braunii. The channel in Chara is Ca²⁺ activated and voltage sensitive. Besides ALMT, five chloride channels (CLCs) for extrusion of Cl^– were found (CHBRA19g00410, CHBRA262g00230, CHBRA431g00080, CHBRA437g00060 and CHBRA50g00900). It is interesting that S-type anion channel 1 (SLAC1) was found to be absent. However, five OSCA-type Ca²⁺-dependent Cl^– channels (CHBRA19g00410, CHBRA262g00230, CHBRA431g00080, CHBRA437g00060 and CHBRA50g00900) as well as one anoctamin-like channel (CHBRA1226g00020) has been identified.

Seven genes of CNGC-type cyclic nucleotide-gated cation channels, i.e. CHBRA38g00220, CHBRA209g00540, CHBRA211g00290, CHBRA38g00230, CHBRA161g00350, CHBRA161g00370 and CHBRA107g00140, were found. They could be candidates for Ca²⁺ transport systems and be associated with regulation of cytoplasmic Ca²⁺ concentration and stress tolerance. However, one has to bear in mind that plants do not utilize a canonical cyclic nucleotide-based signalling system known in animal cells (Edel et al., 2017).

The role of specific potassium channels in the repolarization and maintenance of the resting membrane potential has been extensively studied in Characean algae. Single ion channels with conductance of about 50 pS were reported in the N. obtusa plasma membrane (Azimov et al., 1987). In C. corallina, K⁺-conducting channel (40 pS) activity during an AP was identified. Such channels should play a role in conducting an outward current in the repolarization phase of an AP (Homann and Thiel, 1994). It is thought that high-conductance (170 pS) tonoplast Ca²⁺-dependent K⁺ channels may also be involved in the generation of tonoplast AP, i.e. its repolarization phase (Laver and Walker, 1991). The calcium-activated potassium channel slowpoke (BK) CHBRA373g00030 was found in the Chara genome. Shaker-type voltage-gated K⁺ channels KOR CHBRA826g00040 and AKT5 CHBRA136g00080 in the C. braunii genome probably mediate the depolarization-activated potassium efflux of the AP repolarization phase. The ATP-activated inward-rectifier K⁺ channel IRK1 CHBRA588g00200 was also found. It can be involved in the restoration of K⁺concentration after AP.

As the largest calcium store in plant cells, vacuoles can release large amounts of endogenous Ca²⁺ into the cytoplasm through slow vacuolar Ca²⁺ channels located on the vacuole membrane when plants are stimulated. The channel is encoded by TPC1 with the characteristics of slow activation and slow inactivation. Two-pore channnel (TPC)-CT, CHBRA57g00010 as well as TPC-NT CHBRA1050g00030 were shown in the Chara genome, although functional experiments have not confirmed their physiological activity (Koselski et al., 2021). Thus, the molecular identity of possible Ca²⁺-permeable channels in the tonoplast still needs to be established.

Studies on Characean cells have also contributed to the exploration of P-type H⁺ pumps (Mimura, 1995). The negative resting potential across the plasma membrane is generated by P-type H⁺-ATPases. Three genes, AHA4, AHA2 and HA9, encoded in the C. braunii genome, i.e. CHBRA141g00090, CHBRA351g00130 and CHBRA89g00370, were found. Blast analysis showed that CHA1 of C. australis is closely related to P-type H⁺ -ATPases g6273 and g30033 of C. braunii (Zhang et al., 2020). Sequence analysis indicated it to be a P3A ATPase, with the diversity in the regulation domain and the proton transport cavity, which showed a new perspective on the evolution pattern of plasma membrane H⁺-ATPases.

Even though the influx of Ca²⁺ (approx. 4 × 10⁻¹⁴ mol cm⁻² per AP) is believed to be orders of magnitude smaller than the efflux of Cl⁻ and K⁺ (approx. 4 × 10⁻⁹ mol cm⁻² per AP), precise regulation of [Ca²⁺]_cyt is of vital importance (Fillafer et al., 2018). Ca²⁺-ATPase governs Ca²⁺ re-sequestration after excitation to inner sources or extracellular storage spaces such as cell walls. Prolongation of repolarization of AP can be attributed to inhibition of Ca²⁺-ATPase activity. Seven genes of P-type Ca²⁺-ATPase, namely ACA11a, ACA11b, ECA1, ACA4, ECA4, ECA3 and ACA9, were determined in Chara; two of them are endomembrane type and another two are situated in the ER (Nishiyama et al., 2018).

Physiological relevance of Characean action potentials

Plant APs are different from those in animal cells because they are osmotically active due to extrusion of both Cl^– and K⁺ ions. While in guard cells oscillations in voltage are directly linked with turgor decrease and stomatal closure (Minguet-Parramona et al., 2016), in Characean cells the osmotic effect of a single AP is limited due to their large volume (Barry, 1970). Nitella cells clearly regulate not the turgor pressure but the osmotic pressure, and only brackish Characeae can regulate turgor (Tazawa et al., 1987). An AP was shown to be generated upon the increase in turgor but not upon the decrease; this also agrees with the assumption that an increase in membrane tension activates Ca²⁺ channels (Kaneko et al., 2009). It was reported in the brackish Characeae Lamprothamnium succinctum that hypotonic treatment induces a large increase in [Ca²⁺]_cyt while hypertonic treatment does not (Okazaki et al., 1987). Moreover, Chara cells do not lose excitability in the course of plasmolysis (Fillafer et al., 2018). It is suggested that the activation of Ca²⁺ channels is triggered by membrane stretching but not by membrane compression. When the cytoplasm of the Characeae cell, especially that of Nitella flexilis, is rehydrated rapidly, [Ca²⁺]_cyt increases through Ca²⁺ release from intracellular stores. The release may be triggered by the stretching of endomembranes caused by osmotic swelling of the Ca²⁺ store lumens.

During Aps, [Ca²⁺]_cyt increases from the low resting level of approx. 10^–7m to approx. 10^–6m (Williamson and Ashley, 1982). The appearance of a large amount of calcium ions as signalling molecules triggers plant cell responses to stress factors. In Characeae, APs trigger intracellular events and instantaneous cessation of cytoplasmic streaming. Myosin XI produces an intracellular sap flow known as cytoplasmic streaming or cyclosis in internodal cells by moving on actin filaments while binding organelles via its tail domain. Cytoplasmic streaming facilitates the distribution of molecules and vesicles throughout long cells. It was noticed early on that a temporary halt in cyclosis is caused by all stimuli that produce APs (mechanical shock, bending and a sudden drop in temperature). This allows wound healing following a disruption of the cell wall and cell membrane, preventing leakage of the pressurized cytoplasmic contents and hindering the propagation of stressors along these giant cells (Williamson and Ashley, 1982; Beilby, 2016; Brunet and Arendt, 2017). The membrane repair mechanisms are quickly activated upon local rupture and are directly dependent on calcium ions. A controlled influx of calcium upon mechanical stress allows prevention before the actual damage occurs, and thus the effects of membrane rupture are anticipated. The cessation is a result of inhibition of the actin–myosin interaction by Ca²⁺. When a Chara cell is incubated with cytochalasin and polymerization of actin is blocked, cytoplasmic streaming comes to a halt but the cell remains excitable (Foissner and Wasteneys, 2014).

Another phenomenon that is directly affected by APs is pH banding that aids carbon fixation (Bulychev et al., 2004). A long Characean cell generates external bands of alternating pH along its axis which, combined with the vigorous cytoplasmic streaming, enhances production and distribution of photosynthetically active compounds (Bulychev and Komarova, 2017). A generated AP dissipates the spatial heterogenicity of pH and, consequently, transiently inactivates photosynthesis. However, at the same time, the differences in the photosynthetic activity of a cell are made more spatially diverse, and non-photochemical quenching is drastically increased (Krupenina et al., 2008), which points to the function of AP in enhancement of protection of the photosynthetic apparatus. The mechanism by which AP influences photosynthetic activity involves an increase in the cytoplasmic Ca²⁺ concentration, which is directly related to the generation of an AP (Bulychev and Komarova, 2014) as well as inactivation of H⁺-ATPase (Sukhov et al., 2014).

The cell membrane undergoes thermodynamic state changes during AP propagation. A synchronized coupled soliton pressure pulse in the cell membrane accompanies the AP. It is demonstrated that excitation of C. braunii internodes is coupled by out-of-plane displacements of the cell surface in the micrometre range (approx. 1–10 μm). The onset of cellular deformation coincides with the depolarization phase of the AP (Fillafer et al., 2018).

Altogether, APs in Characean algae are directly linked to calcium signalling, osmotic regulation, cytoplasmic streaming and photosynthetic activity.

ACTION POTENTIALS IN LIVERWORTS: CONOCEPHALUM CONICUM AND MARCHANTIA POLYMORPHA

Gametophytes of liverworts consisting of relatively uniform cells connected with plasmodesmata constitute a suitable system of AP transmission. Liverworts have not developed conducting tissues, roots and active stomata, which are among the basic morphological differences in comparison with vascular plants. A dominating generation of liverworts is a haploid gametophyte. Moreover, molecular analyses indicate that many genes, among them those involved in long-distance signalling, are less redundant than in higher plants – sometimes single genes encode an ion channel protein (Bowman et al., 2017). The basal liverwort Marchantia polymorpha was relatively easy to transform, and the number of available mutants is growing (Tsuboyama and Kodama, 2018). Liverworts can be propagated vegetatively by gemmae or by regeneration of thallus fragments, which makes the mutations stable over a long time. All these points indicate that liverworts are convenient subjects for electrophysiological and molecular studies. In this review, we focus on electrophysiological aspects of liverwort excitation.

Action potentials in Conocephalum conicum

Examination of APs in liverworts began in the early 1980s by the plant electrophysiology research team from Maria Curie-Skłodowska University, Lublin, Poland. Initially C. conicum was chosen as a model plant. Already the first experiments disclosed that C. conicum thalli were excitable – they generated a single AP or a series of APs in response to different stimuli (Paszewski et al., 1982; Favre et al., 1999). In non-immersed thalli, APs spread with velocities in the range of 0.4–1.5 mm s^–1 (Zawadzki and Trębacz, 1985). APs in C. conicum exhibited refractory periods. The absolute refractory period was approx. 1.5 min, whereas the relative refractory period was in the range of 3.5–8.5 min (Dziubińska et al., 1983). Due to their relatively large dimensions, the cells of C. conicum proved appropriate for application of intracellular microelectrodes. The transmembrane potential (resting potential) was registered in the range from –100 to –180 mV (Dziubińska et al., 1983).

Conocephalum conicum cells were found to be light sensitive (Trębacz and Zawadzki, 1985; Trębacz, 1989). Illumination caused transient depolarization whose magnitude depended on light intensity. Darkening evoked a reverse effect – vanishing hyperpolarization. After exceeding a threshold light intensity, a slow depolarization phase, called a generator potential (GP), developed into a fast depolarization – an initial phase of an AP. Increasing light intensity lowered the duration of GPs but had no impact on the AP amplitude, which was a manifestation of the ‘all-or-nothing’ law (Trębacz and Zawadzki, 1985). After exceeding the threshold of excitation, which was manifested by the fast depolarization, light could be turned off, but it had no effect on the AP development, which continued in the same way as during illumination (Trębacz and Zawadzki, 1985). It was demonstrated that photosynthetic pigments (mainly chlorophylls) play a role as light receptors. GPs totally disappeared after application of 3-(3,4-dichlorophenyl)-1,1-dimethyl urea (DCMU), an inhibitor of the photosynthetic electron transfer chain (ETC). This conclusion was confirmed in experiments demonstrating that the magnitude of GPs has a similar action spectrum to photosynthesis with the characteristic maxima in red and blue light bands. ETC within photosystem II (PSII) seems crucial in GP formation because electron donors to PSI did not restore GP after DCMU treatment (Trębacz and Zawadzki, 1985). It is worth mentioning that DCMU had no impact on plant excitability. Conocephalum conicum cells treated with DCMU keep on generating APs in response to electrical stimuli (Trębacz et al., 1989a). It was demonstrated for the first time in plants that responses to sub-threshold light and electrical stimuli can add and evoke APs (Trębacz and Zawadzki, 1985).

Different ion channel inhibitors were applied in search of the ion mechanism of APs in C. conicum (Trębacz et al., 1989b). The starting point in this study was a sequence of ion fluxes postulated as the basis of APs in Characean algae, as mentioned in the first part of the review. The hypothesis to be verified proposed that AP in C. conicum is initiated by a Ca²⁺influx followed by a Cl^– efflux, which both cause depolarization, and the K⁺ efflux is responsible for repolarization.

Special attention was paid to the Ca²⁺ component of APs. Lanthanum and manganese ions proved effective inhibitors of APs in Conocephalum, which indicates participation of the calcium flux in an early phase of AP (Trębacz et al., 1989b). The anion channel inhibitors NPPB [(5-nitro2-3-phenylpropyl-amino)benzoic acid], Zn²⁺, A9C (anthracene-9-carboxylic acid) and ethacrynic acid caused a decrease in AP amplitude but none of them abolished excitability (Trębacz et al., 1997).

Unexpectedly, the potassium channel inhibitor TEA (tetraethylammonium) totally blocked APs in C. conicum evoked by electrical stimulation instead of the expected prolongation of the repolarizing phase, which occurs in many other excitable systems (Trębacz et al., 1989b).

The activity of the electrogenic proton pump proved necessary for excitability in C. conicum. DNP (2,4-dinitrophenol), CCCP (carbonyl cyanide-m-chlorophenylhydrazone) and NH₄Cl caused a gradual decrease in the membrane potential difference accompanied by vanishing series of APs. When the resting potential stabilized at the level of –50 to –80 mV, no APs could be evoked, no matter how strong the stimuli were (Trębacz et al., 1989b). This indicates that the existence of the metabolic component of the membrane potential is necessary for AP generation.

The sequence of ion fluxes in Conocephalum consistent with that in Characean giant cells was confirmed in experiments with application of ion-selective microelectrodes (Trębacz et al., 1994). During the AP, the cytoplasmic Ca²⁺ concentration [Ca²⁺]_cyt increased transiently from approx. 230 to 480 nm. A transient decrease in Cl^– and K⁺ in the cytoplasm during an AP was also registered, indicating an efflux of these ions (Trębacz et al., 1994). However, a very high resistance of the ion-selective microelectrodes was responsible for their inertia and did not allow precise measurements of the kinetics and absolute values of the fast ion fluxes.

To characterize in more detail the calcium component of APs in C. conicum, chloride and potassium fluxes were blocked by ion channel inhibitors. Combined application of the chloride and potassium channel inhibitors, A9C and TEA, revealed a new type of C. conicum responses to light and cold stimuli (Trębacz et al., 1997; Król and Trębacz, 1999; Krol et al., 2003). These responses were named voltage transients (VTs). VTs did not fulfil the ‘all-or-nothing’ principle; on the contrary, their amplitudes depended on the stimulus strength, i.e. the light intensity or rate of cooling. Their kinetics was different from those of APs – both depolarization and repolarization phases were faster than during an AP. With strong stimuli, or even mild stimuli but applied after long breaks, an overshoot was registered, i.e. the peaks of VTs were at positive membrane potentials, which was not the case in APs (Trębacz et al., 1997). The most striking difference between VTs and APs was the inability of VTs to propagate (Król and Trębacz, 1999; Król et al., 2003). Additionally, VTs cannot be evoked by electrical stimuli. Experiments with treatments affecting Ca²⁺ fluxes indicated that VTs were presumably a calcium component of APs (Trębacz et al., 1997; Król and Trębacz, 1999). The amplitudes of VTs depended on the Ca²⁺ gradient according to Nernst’s law. VTs were suppressed by La³⁺ and Gd³⁺, and enhanced by Sr²⁺, regarded as a factor liberating calcium ions from internal stores (Król and Trębacz, 1999; Król et al., 2003).

Both APs and VTs appeared after treatment of C. conicum cells with and glycine (Gly). At millimolar Glu and Gly concentrations, a single or a series of APs were generated. The responses were blocked by inhibitors of animal GLRs: 6,7-dinitroquinoxaline-2,3-dione (DNQX) and d-amino-5-phosphonopentanoic acid (AP5) (Krol et al., 2007).

All these results taken together can be interpreted as follows: Ca²⁺ fluxes initiate APs. The sources of calcium ions entering the cytoplasm seem to be both external and internal. Whether there is a mechanism resembling ‘calcium-induced calcium release’ (CICR) is to be examined in detail. Ca²⁺ fluxes through the plasma membrane could occur owing to opening of Glu channel receptors homologous to animal cells and plants, including arabidopsis. Other calcium-permeable channels can be activated too (see below). An increase in [Ca²⁺]_cyt can activate calcium- and voltage-dependent slow vacuolar (SV) channels, which in turn may release Ca²⁺ from the vacuole lumen to the cytosol, causing a further increase in [Ca²⁺]_cyt. Vacuoles of C. conicum possess numerous SV channels (Trębacz et al., 2007; Schönknecht and Trębacz, 2008; Koselski et al., 2019). SV channels in arabidopsis and in many other vascular plants are encoded by a single gene TWO-PORE-CHANNEL 1 (TPC1) (Peiter et al., 2005; Guo et al., 2016; Kintzer and Stroud, 2016). It was recently demonstrated that SV currents were involved in long-distance electrical signals (VPs rather than APs). Mutants with a knocked-out TPC1 gene exhibited a drastic decrease in the VP transmission rate (Choi et al., 2014; Kiep et al., 2015). In liverworts, pharmacological inhibition of SV channels did not affect the velocity of AP propagation. This indicates that SV channels are not crucial in the mechanism of AP (Koselski et al., 2021). A direct proof will be presented in an ongoing study. As in Characean algae, calcium fluxes are followed by Cl^– and K⁺ effluxes. An increase in plasma membrane conductance caused by Cl^– and/or K⁺ fluxes is probably the cause of the overshoots observed in VTs which in untreated cells are short-circuited by Cl^– and K⁺ currents and the accompanying decrease in the plasma membrane resistance. The inability of VTs as a putative calcium component of the AP to propagate and to be evoked by electrical stimulation is consistent with the absence of voltage-dependent calcium channels homologous to animal VVCa channels (Edel et al., 2017). The main question in this context is: which type of ion channel is voltage dependent to cause a regenerative depolarization and spread of the APs based on local circuits as in animal excitable cells? The candidates responsible for Ca²⁺ influx are described below. The role of ‘a calcium spreading system’ might be played by chloride-permeable channels. Whether they are [Ca²⁺]_cyt activated, like those postulated in Characean cells, needs to be examined.

In C. conicum, there are at least two other candidates which may be responsible for the membrane potential depolarization and exceeding the threshold of excitation. These could be potassium channels. Since the equilibrium potential for K⁺ in C. conicum is positive relative to the resting membrane potential (due to the activity of the electrogenic proton pump), opening of potassium channels may lead to a substantial depolarization. This may explain the inhibition of APs by TEA. The plasma membrane H⁺-ATPase itself can be another player responsible for cell depolarization, especially in responses to stimuli affecting ATP supply, e.g. cooling or changing light intensity. Thus, exceeding the excitation threshold and initiation of an AP is possible due to opening of Ca²⁺, Cl^–, but also K⁺ channels, as equilibrium potentials for all these ions are positive with respect to the resting potential (Trębacz et al., 1994). Temporal suppression of the proton pump can evoke a similar effect – a depolarization of the resting potential leading to exceeding the excitation threshold and generation of an AP.

In search of a membrane-localized cold sensor in liverworts, we examined membrane potential changes in response to sudden and gradual cooling and applied menthol which activates transient receptor potential (TRP) channels in animal temperature-sensing cells (Kupisz and Trębacz, 2011a, 2011b). It was demonstrated that cooling caused depolarization of the resting potential and, when fast enough, evoked an AP. The rate of cooling but not the absolute temperature was a genuine stimulus for exceeding the threshold of excitation (Kupisz and Trębacz, 2011a). Menthol caused effects dependent on its concentration: at approx. 1 mm, it evoked a transient depolarization, 5 mm was sufficient to trigger an AP and, at 10 mm, menthol evoked a series of several APs. Unlike in mammals, there was no difference in the concentration threshold between the two chiral forms of menthol, (+)menthol and (–)menthol. The menthol-evoked APs and VPs were blocked by calcium channel and proton pump inhibitors. Alternate application of menthol and cooling in short time intervals allowed the conclusion to be drawn that menthol and temperature drop affect different membrane transport systems because no typical refractory periods were observed (Kupisz and Trębacz, 2011a, 2011b). Liverworts, like most other plants, harbour no TRP-like channel genes, thus the question of which channels/transporters are affected by menthol remains open. Further experiments indicated that menthol may also interact with the lipid phase of the plasma membrane (Kupisz et al., 2015).

Action potentials in C. conicum are affected by the flavonoids quercetin, apigenin and genistein (Pawlikowska-Pawlęga et al., 2000, 2007, 2009, 2011). Combined application of these flavonoids and verapamil indicates their interaction with calcium transporting systems. This can be an evolutionarily new way for calcium signalling regulation in liverworts, as there is no convincing evidence of flavonoid biosynthesis in Characean algae (Bowman et al., 2017).

Physiological consequences of action potentials in Conocephalum conicum

It was clearly demonstrated by Dziubinska et al. (1989) that excitation affects the respiration rate in C. conicum thalli. Plants were kept in darkness and the respiration rate was measured by an infrared gas analyser. Non-damaging electrical and damaging cutting stimuli were applied, and APs were registered. Approximately 6 s after the onset of the AP, an increase in the respiration rate began and reached a level approx. 70 % higher than in intact plants. Wounding caused an even higher increase, up to 100 %, and the generation of a series of APs was accompanied by a series of respiration rate enhancements (Dziubinska et al., 1989). When plant excitability was blocked by TEA, no significant increase in respiration was registered, no matter whether electrical stimulation or cutting was applied (Dziubinska et al., 1989). The fast onset of the respiration rate increase suggests glycolysis as part of the respiration pathway, which is subjected to regulation by ion fluxes (mainly Ca²⁺) accompanying APs. Glycolysis takes place in the cytoplasm and can be a target of Ca²⁺-dependent regulation factors. Detailed coupling of respiration and APs is to be clarified. An excess of the energy stored in ATP may be used to re-establish ion homeostasis and, for instance, to support healing processes after wounding.

Action potentials in C. conicum interfere with reactive oxygen species (ROS) and their scavenging systems (Dziubińska et al., 1999; Brankiewicz, 2015). Mechanical stimulation (local cutting) evoked a burst in peroxidase activity, reaching its maximum approx. 6 min after the stimulation. In plants that did not generate APs after wounding (only local responses), there was no significant increase in peroxidase activity (Dziubińska et al., 1999).

In C. conicum, hydrogen peroxide (H₂O₂) and hydroxyl radicals applied to intact thalli triggered either one or a vanishing series of APs appearing on the background of a slow decrease in the membrane potential difference. Relatively high ROS concentrations should be used to evoke APs; the effect was induced by H₂O₂ at concentrations of approx. 10 mm and by hydroxyl radicals at concentrations >1 mm. Experiments with application of ion channel inhibitors indicated that both these ROS affected the activation of ion channels involved in the formation of APs in C. conicum (Brankiewicz, 2015).

Long-distance electrical signals in Marchantia polymorpha

Marchantia polymorpha is a liverwort closely related to C. conicum. The shape and dimensions of the thallus are similar. The main advantage of M. polymorpha over C. conicum is that its genome was recently sequenced (Bowman et al., 2017) and mutants with knocked out ion channel genes are becoming available (Hashimoto et al., 2022). Quite recently, M. polymorpha was chosen as the suject of an electrophysiological study of long-distance signals. It was demonstrated that it too is excitable. Marchantia polymorpha generates APs upon illumination. In contrast to Conocephalum, Marchantia responds vigorously to darkening. Two different responses were registered: typical action potentials such as those evoked by light and AP-like responses consisting of transient depolarization followed by a long-lasting plateau (Kupisz et al., 2017). The participation of the proton pump in these responses to light/dark transitions was examined. The proton pump inhibitor FCCP (4-(trifluoromethoxy)phenylhydrazone) depolarized the cells and blocked all these responses. On the other hand, fusicoccin, which is a proton pump activator, hyperpolarized the membrane potential and abolished APs and AP-like responses (Kupisz et al., 2017). This was interpreted as the possibility to generate APs only in a narrow ‘window’ of the resting potential. This may be coupled with activation of ion channels in that range of voltages.

The calcium component of APs in M. polymorpha was studied by Koselski et al. (2021). Verapamil (20 mm) reduced the amplitudes of electrically triggered APs. Intracellular stores, among them the vacuole, can be a source of Ca²⁺ ions entering the cytoplasm during APs, in addition to the calcium flux from the apoplast. Special attention was paid to the examination of the role of vacuolar TPC/SV channels in AP electrogenesis (Koselski et al., 2021). The M. polymorpha genome possesses three genes encoding this type of channel: MpTPC1, MpTPC2 and MpTPC3 (Hashimoto et al., 2022). All these channel proteins are localized in the tonoplast. Experiments with application of the patch clamp technique clearly demonstrated that, out of the three channel genes, only MpTPC1 encoded an active SV channel (Hashimoto et al., 2022). Its basic features, e.g. the slow kinetics of activation, calcium dependence and cation permeability, are similar to those of AtTPC1 channels in Arabidopsis thaliana (Hedrich et al., 2018; Hashimoto et al., 2022). The other two channels do not pass detectable ion currents. The mechanism of their activation remains unknown.

Mammalian cells possess two TPC proteins, TPC1 and TPC2, located in the intracellular compartments endosomes and lysosomes (She et al., 2018). These channels proved crucial in the aetiology of many severe diseases such as virus infections, arrhythmias and carcinogenesis (Patel and Kilpatrick, 2018; Penny et al., 2019). Medicines, among them the TPC blockers tetrandrine and NED-19, produce positive therapeutic effects (Kintzer and Stroud, 2016). The effects of these substances on TPCs in M. polymorpha were examined using the patch clamp technique on isolated vacuoles and with intracellular microelectrodes on intact cells (Koselski et al., 2021). A similar approach was applied to the internodal cells of N. obtusa (Koselski et al., 2021). Tetrandrine (20 μm) caused reduction in the AP amplitude and substantial prolongation of the repolarization phase. A similar long-lasting plateau was observed in APs registered in a group of M. polymorpha thalli treated with NED-19 (75 μm). In another group of plants, the duration of the AP did not significantly differ from the control. Patch clamp experiments revealed that tetrandrine did not block TPC/SV channels, whereas NED-19 even caused an increase in the open probability of the channel and so-called flickering (Koselski et al., 2021). These results indicate that the effects of these chemicals were not coupled with inhibition of the TPC/SV channels and the putative vacuolar component of the AP. The prolongation of the repolarization phase of the AP may indicate interference with outward-rectifying potassium channels.

Genes encoding ion channels in Marchantia polymorpha

The known genome of M. polymorpha (https://marchantia.info) allows finding genes encoding ion channels that can participate in APs. The mechanism of AP initiation in Marchantia, if similar to that in the cells of Characean algae (Lunevsky et al., 1983), should assume an increase in [Ca²⁺]_cyt. In plant cells, the structures fulfilling this role can be assigned to calcium-permeable channels activated by ligands, i.e. Glu and/or cyclic nucleotides, by voltage or by mechanical stress. The dependence of long-distance electric signals in plants on Glu-activated channels (AtGLR3.3 and AtGLR3.6) was well documented in arabidopsis (Mousavi et al., 2013; Salvador-Recatala et al., 2014). In Marchantia, the GLR-like channel (Mp1g01040) belongs to the GLR 3.4 subfamily. The role of this channel in AP electrogenesis is currently being examined in our lab. What is unknown is the involvement of CNGCs in APs despite the evidence that, in plant genomes, CNGC genes encode ligand-gated calcium-permeable channels (Duszyn et al., 2019). In A. thaliana, CNGC19 is at least partially responsible for systemic Ca²⁺ signals evoked by the herbivore Spodoptera litura (Meena et al., 2019). The role of CNGCs in Marchantia can be played by five proteins belonging to the CNGC family (Mp3g14660, Mp4g04110, Mp4g11640, Mp5g07780 and Mp6g01920).

Besides ligand-gated calcium channels, generation of plant APs can be dependent on voltage-gated and mechanosensitive calcium channels. Although voltage-dependent calcium currents were recorded in plants (Demidchik et al., 2018), plant genome analysis indicates lack of typical genes encoding this type of channel (VDCC). In turn, genes encoding stretch-activated channels were recently characterized in the touch-sensitive structures of carnivorous plants (Procko et al., 2021). The visualization of calcium wave propagation in the trap of the carnivorous plant Dionaea muscipula (Suda et al., 2020) confirms that calcium plays the main role in AP generation and propagation. The stretch-activated Ca²⁺-permeable channel component (Mp2g20150) is present in the Marchantia genome, but BLAST analysis indicated no significant homology with the genes from Dionaea. The Marchantia genome also possesses genes encoding mechanosensitive MSL-type channels (Mp1g10420, Mp1g16810, Mp8g14680, Mp2g23620, Mp7g10770, Mp2g12070 and Mp4g04820) and PIEZO-type channels (Mp2g12070). The latest studies conducted by Mousavi et al. (2021) demonstrated that PIEZO channels from arabidopsis are calcium channels required for root mechano-transduction.

One of the ions whose efflux seems to be an important component of depolarization of the membrane potential during AP is chloride, since the Nernst potential for chloride is much more positive than the membrane potential values observed at the resting potential (Trębacz et al., 1994; Kupisz et al., 2017; Koselski et al., 2021). An important factor which should be considered in searching for a suitable channel responsible for anion efflux during AP is its Ca²⁺ and/or voltage dependence. The chloride channels that may participate in the generation of an AP can be divided into two groups based on the rate of activation – rapid (R-type) and slow (S-type) (Schroeder and Keller, 1992). One of these channels, SLAC1 (S-type anion channel 1), exhibits a multistep Ca²⁺-dependent and Ca²⁺-independent regulation (Pantoja, 2021). MpSLAC1 genes encoding such channels (Mp5g19130) were found in Marchantia. Also, two other proteins (Mp5g19130 and Mp5g18420) can act as SLAH3, one of the homologues of SLAC1. In contrast to SLAC1 exhibiting weak rectifying properties, the voltage dependence of SLAH3 indicates outward rectifying properties of the channel (Hedrich, 2012). In addition to SLAC/SLAH channels, some role in excitability is assigned to R-type channels called QUAC (QUick activated Anion Channels), which belong to the family of aluminium-activated malate transporters (ALMT). In arabidopsis, QUACs are represented by ALMT12 required for signal-dependent stomatal movements in guard cells (Meyer et al., 2010). In Marchantia, there are four genes (MpALMT1–4) encoding ALMT channels (Mp3g25480, Mp3g01810, Mp7g05600 and Mp4g14650). The potential candidate for a chloride channel participating in excitation of the cell can be the calcium-activated anion channel recently identified in animals as TMEM16A (anoctamin 1) and BEST1 (bestrophin 1) (Kunzelmann, 2015). A single copy of the TMEM gene was found in arabidopsis (At1g73020), but its function is still unknown (Hedrich, 2012). In turn, BEST-like channels in arabidopsis (At3g61320 and At245870) do not possess Ca²⁺-binding domains (Hedrich and Geiger, 2017). In Marchantia, proteins homologous to TMEM16A (Mp3g10180) and BEST (Mp5g15730 and Mp7g14460) are also present.

In addition to calcium and chloride ion channels, whose opening leads to depolarization of the membrane potential during APs in plants, activation of potassium channels is also important in these responses. The opening of depolarization-activated K⁺release channels leads to repolarization of the membrane potential. Such voltage-gated outward-rectifying potassium channels are represented in plants by GORK and SKOR channels. GORK channels modulate the shape of an AP and limit its amplitude and duration in A. thaliana (Cuin et al., 2018). Genes encoding outward-rectifying potassium channels are present in Marchantia (MpORK), and a product of the gene (Mp2g04950), according to BLAST analyses, shows 54 % identity with GORK and SKOR from arabidopsis. Another type of a well-characterized potassium channel from arabidopsis taking part in APs is AKT2. In contrast to GORK, AKT2 is a weak inward rectifier modulating plant tissue excitability (Hedrich, 2012; Cuin et al., 2018). In Marchantia, the gene encoding AKT1 was identified as the channel-encoding MpAKT1 (designated as Mpg22750), whose putative role can be potassium uptake. The potassium homeostasis in Marchantia cells can also be maintained by two-pore channels. One of the channels, TPC1 (in Marchantia designated as Mp1g26020), is localized in the vacuole membrane and was electrophysiologically characterized in our laboratory (Koselski et al., 2017; Hashimoto et al., 2022). The results proved the outward-rectifying properties of MpTPC1, i.e. its calcium dependence and permeability to cations (Koselski et al., 2017; Hashimoto et al., 2022). The permeability of the channels to calcium recorded in different plants (Hedrich et al., 2018) and the control of the membrane potential by TPC1 (Jaślan et al., 2019) indicate a possibility of participation of these channels in tonoplast APs. A study of TPC1 in arabidopsis has indicated that TPC1 is essential for propagation of systemic wound- and salt stress-evoked calcium waves (Choi et al., 2014; Kiep et al., 2015). In addition to MpTPC1, other channels possess two pores and are present in Marchantia, i.e. MpTPC2 (Mp3g12680) and MpTPC3 (Mp7g04900). Both these gene products are localized in the tonoplast (Hashimoto et al., 2022), but the mechanism of their activation remains unknown. The Marchantia genome contains genes encoding channels with high potassium selectivity (MpTPK1 designated as Mp1g24190 and MpTPK2 designated as Mp4g00270) homologous to the TPK channel in the arabidopsis vacuole, but their function and localization have not been examined yet.

COMPARISON OF ACTION POTENTIALS IN CHARACEAE AND LIVERWORTS

Despite substantial morphological differences, Characean algae and liverworts exhibit similarities in long-distance signalling, i.e. APs. The main features of the APs distinguishing these signals from VPs and SPs, i.e. fulfilling the ‘all-or-nothing’ law, propagation without decrement, refractory periods, etc., are common in both these groups of plants. The sequence of ion fluxes is also similar. Direct and indirect evidence indicates an important role for Ca²⁺fluxes in the initial (depolarization) phase of APs. In both cases, participation of Ca²⁺ fluxes from the apoplast and internal stores during APs is postulated. Generation of APs in response to electrical stimulation (weak electric current passing through a part of a plant) is a common feature, indicating a voltage-dependent component of APs, too (Fig. 1).

There are also noticeable differences among APs in these plants. APs in Characeae typically last several seconds in comparison with dozens of seconds in liverworts. Refractory periods in liverworts last longer than in Characeae algae (3–60 s vs. 1.5–10 min), which is probably connected with regulation of ion channel activity (Fig. 1). The AP parameters in liverworts are close to those in vascular plants, with the exception of special cases such as the known carnivorous plants or Mimosa pudica (Trębacz et al., 2006; Huber and Bauerle, 2016). APs in Characean cells propagate faster than in liverwort thalli (11–44 mm s^–1 vs. 0.4–1.5 mm s^–1). It is noteworthy that the apoplast of terrestrial plants, among them liverworts, is usually deficient in charge carriers, i.e. ions. Submerging C. conicum in solutions with ion concentrations comparable with APW (artificial pond water – a standard control solution for Characean AP investigations) causes acceleration of APs to approx. 6 mm s^–1, which is consistent with predictions of the ‘cable theory’ (Trębacz and Zawadzki, 1985; Zawadzki and Trębacz, 1985). Basic parameters characterizing APs in the alga N. obtusa and liverworts M. polymorpha and C. conicum are presented in Table 1. APs in C. conicum and M. polymorpha usually cover the whole thallus, which is a manifestation of effective intercellular transmission covering thousands of cells. In Characean algae, where giant internodal cells are connected with nodes consisting of much smaller cells, APs usually stop at the nodes. Transnodal transmission of APs is relatively rare. This is one of the examples of specialization of giant cells. Another important aspect of Characean cells is the exceptionally specialized and adaptive cytoplasmic streaming. It is vigorous in intact internodal cells and hardly noticeable in liverworts. Cessation of the cytoplasmic streaming evoked by the AP is postulated to be important in preventing the detrimental consequences of stress factors, such as wounding, affecting these semi-autonomous cells (Brunet and Arendt, 2017). No wonder that there is a mechanism, probably in plasmodesmata, which limits AP propagation and arrests the cyclosis in cells directly affected by the stimuli. One such mechanisms can be the so-called ‘valving’. It was demonstrated that even the small osmotic pressure gradient between Characean tandem internodal cells connected with a node blocks the transport through plasmodesmata (Zawadzki and Fensom, 1986). Ion fluxes during APs in internodal cells are strong enough to evoke such osmotic potential gradient. Valving was demonstrated in C. conicum thalli, too (Trębacz and Fensom, 1989).

The existence of externally generated acidic and basic bands is characteristic in giant Characean cells but is not observed in liverworts. Again, this may be associated with the large dimensions of the internodal cells, the degree of their specialization and the opportunity to enhance photosynthetic performance selectively.

The role of ion fluxes through the vacuolar membrane (the tonoplast) in the AP mechanism seems to be different in both groups of plants. Partially, it can be a result of different experimental possibilities offered by the large dimensions of Characean cells in comparison with ‘normal’ cells of liverworts. In Characeae, the vacuolar component of the AP can be detected by direct monitoring of the electrical potential difference across the tonoplast. Moreover, in many instances, the vacuolar components of APs can be distinguished as an additional repolarization phase of the AP registered when the microelectrode tip is localized in the vacuole (Fig. 1). In C. conicum, no such difference was observed. Combined insertion of H⁺-selective and ordinary microelectrodes allowed distinguishing cytoplasmic vs. vacuolar tip localization (two orders of magnitude difference in pH), but no difference in the time-course of APs was observed (Trębacz et al., 1994). In the model vascular plant A. thaliana, a vacuolar AP can be generated after activation of TPC1/SV channels (Jaślan et al., 2019). Its coupling with ion fluxes through the plasma membrane during an AP requires further examination.

As already mentioned in this review, the molecular basis of APs in algae and liverworts remains unknown. The publication of the genomes of C. braunii and M. polymorpha should accelerate this type of study. The previous sections presented candidates for the roles of genes encoding ion channel proteins in both these basal plants. One can consider which of these gene products are good candidates to carry ion fluxes in a sequence derived from the electrophysiological studies: Ca²⁺ influx, Cl^– efflux (through Ca²⁺-dependent chloride channels) and K⁺ efflux. This assignment is complicated by the absence of VDCCs in both these groups of plants. Thus, VDCCs cannot play a role analogous to sodium channels in neurons. On the other hand, there is strong evidence that [Ca²⁺]_cyt transiently increases during APs in algae and liverworts (Williamson and Ashley, 1982; Trębacz et al., 1994). The question is: which of the Ca²⁺-permeable ion channels can do the job? The C. braunii and M. polymorpha genomes possess candidates to encode Ca²⁺-permeable channels, among them CNGCs and touch-sensitive MSL and PIEZO. Somewhat unexpected is the absence of GLR homologues in the Chara genome. In Marchantia and in vascular plants, these channels play an important role in long-distance signalling. Thus, the question of which ion channel(s) in the plasma membrane is(are) activated during the depolarization phase of APs (especially in Characean algae) remains to be answered. Ca²⁺ can be liberated to the cytoplasm from the internal stores, among them the biggest one – the vacuole. Both Chara and Marchantia harbour genes encoding calcium-permeable TPCs (one gene in Chara and three genes in Marchantia). The problem is that these channels require high, non-physiological [Ca²⁺]_cyt for activation and are suited to pass Ca²⁺ (and other cations) in a reverse direction, i.e. from the cytoplasm to the vacuole. Moreover, in Characean cells, these channels have not been characterised by the patch clamp method yet despite the existence of the well-utilized cytoplasmic droplet technique which has revealed multiple K⁺and Cl^– conductances in Characean tonoplasts (Lühring, 1986; Sakano and Tazawa, 1986; Bertl, 1989; Katsuhara et al., 1991).

There is a much broader spectrum of possibilities in the case of putative Cl^–-permeable channel genes. However, again, there is no homologue of SLAC1 in the Chara genome. It was found in another Characeae alga, Klebsormidium flaccidum (Jiang et al., 2021). SLAC1 channels participate in signalling in vascular plants. In Chara, the role of a Ca²⁺-dependent Cl^– channel can be played by a member of the ALMT family, which seems to be voltage sensitive. The group of anion-permeable channels also includes anoctamin-like and CLC-like channel genes. In contrast to Chara, the Marchantia genome contains SLAC1 and SLAH3 homologues. The list of channel genes that may be responsible for the Cl^– efflux during an AP is supplemented by the QUAC/ALMT, TMEM and BEST homologues. Some of them are believed to be [Ca²⁺]_cyt dependent.

We have selected CNGC and ALMT channel proteins whose homologues are present in C. braunii and M. polymorpha to perform a comparative analysis. Figure 2 shows phylogenetic trees of CNGC and ALMT channels in C. braunii, M. polymorpha, Klebsormidium nitens, Physcomitrella/Physcomitrium patens and A. thaliana. The channel proteins of A. thaliana were selected based on their documented role in cellular signalling (Pantoja, 2021). High homology among those channels in C. braunii and M. polymorpha is observed. It remains to be determined whether the channels participate in AP electrogenesis.

It is a general belief that the K⁺ efflux is responsible for AP repolarization. Homologues of K⁺ outward-rectifying channel genes were found in the C. braunii and M. polymorpha genomes. In Marchantia, the MpAKT1 gene can encode such a channel as well as CHBRA KOR or CHBRA AKT5 in Chara. The ATP-activated inward-rectifier K channel CHBRAIRK1 can be involved in membrane potential regulation in Chara. Table 2 summarizes ion channel genes – possible candidates to play roles in AP generation that exist or are absent in Chara and/or Marchantia genomes.

Summarizing, what we can conclude based on examination of APs in Characean algae and liverworts is an abrupt rather than a continuous change in dynamics and probably the molecular mechanism of those signals. The main principles remained unchanged during terrestrialization: (1) spreading of Ca²⁺ waves and (2) substantial efflux of Cl^– and K⁺. However, the channels fulfilling the tasks may be different and/or differentially fine-tuned. In contrast to the above-mentioned high homology among CNGC and ALMT channels, is an absence of GLR and SLAC1-like channel genes in Chara. After terrestrialization, sudden changes in temperature, illumination, osmolarity and mechanical stress including wounding became more pronounced in liverworts and other terrestrial plants. The response to changes in water potential suggests that one of the roles of APs is probably an energetically low-cost turgor regulation. Further evolution of the long-distance signals in plants was related to appearance of conducting tissues – convenient transmission ‘cables’. Variation potentials evolved as specific wound-induced signals. Wounding, no matter whether evoked by abiotic or biotic factors, became the main stressor on the lands. Glutamate and some other amino acids released from damaged cells in combination with GLR channel/receptors constitute good coupling factors between wounding and the response. In vascular plants, signals with an electrical component (AP or VP) are supplemented with chemically based ways to respond, including ROS. Liverworts seem to be equipped with the basic set of ion channels homologous to those in higher plants. Ongoing studies on liverworts and other lower plants with a combination of electrophysiological and molecular methods will throw more light on the evolution of long-distance signals in terrestrial vascular plants.

FUTURE PERSPECTIVES

The examination of electrical signals in Characeae algae has a much longer history than the study of liverworts. The morphology of the giant internodal cells of these algae facilitated application of sophisticated electrophysiological techniques such as open vacuole, permeabilized cells, microinjection. etc. The patch clamp technique was applied in plants from both these groups – in liverworts mainly to characterize vacuolar ion channels. Perhaps the most important data concerning the ion background of APs in Characean cells were obtained by application of the voltage clamp method (Lunevsky et al., 1983). In liverworts, the attempts to apply this technique failed due to the high current leakage through plasmodesmata.

Deciphering the genomes of the representatives of Characean algae and liverworts opened up new perspectives especially for examination of electrical phenomena in liverworts. Characean algal cells are multinuclear, and elaboration of efficient methods of transgenization will take time. On the other hand, M. polymorpha is easy to transform (Tsuboyama and Kodama, 2018). Pioneer results on this plant with mutations in ion channel genes have already been published (Hashimoto et al., 2022). The era of combining genetic and electrophysiological methods has already begun, and one must hope that the molecular nature of APs will be recognized soon. Optical methods requiring transgenization such as GCaMP or genetically encoded voltage indicators (GEVIs) will add a lot to our knowledge (Matzke and Matzke, 2015). This would help to understand not only the molecular mechanism underlying generation of APs but also regulation and transduction of these signals as well as their downstream physiological implications. Conclusions of such studies should have a positive impact on our understanding of plant terrestrialization and further evolution of long-distance signals in vascular plants.

The following conclusions can be drawn. (1) APs in Characean algae and liverworts exhibit differences in dynamics, e.g. duration and rate of transmission. (Table 1). (2) The ion mechanism of APs in both groups of plants consists of Ca²⁺ and Cl^– fluxes responsible for depolarization and K⁺ efflux participating in repolarization. (3) The electrogenic proton pump, H⁺-ATPase, establishes a highly negative resting membrane potential and participates in AP repolarization. (4) Initiation of an AP, i.e. exceeding the threshold of excitation, can be a result of Ca²⁺ and/or Cl^– channel activation, but can also be by an increase in K⁺ permeability and a transient shut down of the proton pump. (5) Ion fluxes across the tonoplast follow those across the plasma membrane. Activation of vacuolar TPCs is considered. (6) There is a high homology among CNGC and ALMT channel proteins in C. braunii and M. polymorpha which can mediate Ca²⁺ influx and Cl^– efflux, respectively (Fig. 2). (7) On the other hand, homologues of GLR and SLAC1 channels, which in terrestrial plants play key roles in shaping long-distance electrical signals, are absent in Chara (Table 2).

FUNDING

This work was supported by DAINA1 – POLISH-LITHUANIAN FUNDING INITIATIVE, project ‘Long-distance electrical signaling systems in plants – adaptation to the change from water to terrestrial environment’, funded by the National Science Centre, Poland, no. 2017/27/L/NZ1/03164 and by the Research Council of Lithuania, no. S-LL-18-1.

LITERATURE CITED