Rapid Propagation of Ca²⁺ Waves and Electrical Signals in the Liverwort Marchantia polymorpha

Abstract

In response to both biotic and abiotic stresses, vascular plants transmit long-distance Ca²⁺ and electrical signals from localized stress sites to distant tissues through their vasculature. Various models have been proposed for the mechanisms underlying the long-distance signaling, primarily centered around the presence of vascular bundles. We here demonstrate that the non-vascular liverwort Marchantia polymorpha possesses a mechanism for propagating Ca²⁺ waves and electrical signals in response to wounding. The propagation velocity of these signals was approximately 1–2 mm s^-1, equivalent to that observed in vascular plants. Both Ca²⁺ waves and electrical signals were inhibited by La³⁺ as well as tetraethylammonium chloride, suggesting the crucial importance of both Ca²⁺ channel(s) and K⁺ channel(s) in wound-induced membrane depolarization as well as the subsequent long-distance signal propagation. Simultaneous recordings of Ca²⁺ and electrical signals indicated a tight coupling between the dynamics of these two signaling modalities. Furthermore, molecular genetic studies revealed that a GLUTAMATE RECEPTOR-LIKE (GLR) channel plays a central role in the propagation of both Ca²⁺ waves and electrical signals. Conversely, none of the three two-pore channels were implicated in either signal propagation. These findings shed light on the evolutionary conservation of rapid long-distance Ca²⁺ wave and electrical signal propagation involving GLRs in land plants, even in the absence of vascular tissue.

Introduction

Despite the absence of a nervous system, plants are equipped with a plethora of sensors to monitor environmental changes and signal transmission mechanisms that span their entire structures. In response to biotic and abiotic stresses, long-distance signals are systemically transmitted from the stressed local sites to distant tissues. Such long-distance signals play important roles in inducing stress responses throughout the body. One type of long-distance signaling has been suggested to involve electrical signals, Ca²⁺ (Gilroy et al. 2016, Hedrich et al. 2016, Johns et al. 2021), reactive oxygen species (ROS) (Miller et al. 2009, Lew et al. 2020) and hydraulic signals (Malone and Stankovic 1991).

Electrical signals are a change in plasma membrane potential induced by the movement of ions into and out of the cell via ion channels/transporters. Various stimuli have been shown to induce electrical signals in streptophytes including bryophytes and angiosperms (Sibaoka 1969, Paszewski et al. 1982, Kupisz et al. 2017, Scherzer et al. 2019, Hagihara and Toyota 2020). Experiments using ion-selective microelectrodes and calcium-permeable channel inhibitors have indicated that Ca²⁺ influx is an important step in the generation of membrane depolarization (Trebacz et al. 1994, Ward et al. 1995, Kikuyama et al. 1997, Vodeneev et al. 2016, Kupisz et al. 2017). In addition to Ca²⁺ influx, Cl¯ and K⁺ play a role in electrogenesis (Trebacz et al. 1989, 1994, Vodeneev et al. 2016). In addition to electrical signals, given the presence of numerous Ca²⁺ sensor proteins in plants, plasma membrane Ca²⁺ influx into the cytoplasm triggered by a variety of stimuli plays a central role in signal transduction (Kurusu et al. 2013, Kudla et al. 2018, Thor 2019, Hashimoto et al. 2023).

Stimulus-induced membrane depolarization and Ca²⁺ influx into the cytoplasm not only occur in the stimulated cell but also propagate to distant cells. Propagation of the Ca²⁺ wave has been suggested to involve glutamate receptor–like (GLR) channels, which are ligand-gated and non-selectively permeable to cations, as well as two-pore channels (TPCs), which are voltage-activated and permeable to cations (Mousavi et al. 2013, Choi et al. 2014, Kiep et al. 2015, Nguyen et al. 2018, Toyota et al. 2018). Although many angiosperm species including Arabidopsis thaliana and rice Oryza sativa possess sole TPC1 gene (Furuichi et al. 2001, Kurusu et al. 2004, Peiter et al. 2005), liverworts and mosses possess type 1 (similar to angiosperm TPC1) and type 2 TPC genes (Hashimoto et al. 2022).

In A. thaliana, electrical signals and Ca²⁺ waves have been shown to propagate at the fastest rate (1–2 mm s⁻¹) along the vascular bundles (Mousavi et al. 2013, Nguyen et al. 2018, Toyota et al. 2018, Bellandi et al. 2022). A recent model has put forth the idea that substances diffuse upwards through the vascular bundle from the site of stimulation, leading to alterations in plasma membrane potential and an increase in [Ca²⁺]_cyt (Bellandi et al. 2022, Gao et al. 2023).

Signal propagation has also recently been observed in non-vascular bryophytes. In the moss Physcomitrium patens, osmotic pressure and H₂O₂ evoke Ca²⁺ waves at a much slower propagation rate, approximately 5 µm s⁻¹, which is notably slower compared to vascular plants (Storti et al. 2018, Koselski et al. 2023). In liverworts, electrical signals have been reported to propagate at a rate of 0.4–1.5 mm s⁻¹ (Trebacz et al. 1994, Koselski et al. 2021), while the dynamics of intracellular Ca²⁺ has not yet been explored. Nonetheless, the molecular mechanism underlying long-distance signal propagation in the absence of vascular bundles remains enigmatic.

Marchantia polymorpha holds particular significance due to several characteristics that render it suitable for investigating these molecular mechanisms. Notably, it is relatively amenable to genetic transformation, with the haploid generation encompassing most of its life cycle. Furthermore, the complete sequencing of the Marchantia genome has been achieved, revealing a genome that, despite retaining a multitude of genes encoding essential signaling factors conserved among land plants, demonstrates reduced genetic redundancy when compared to other land plants (Bowman et al. 2017).

In the present study, we provide groundbreaking evidence for the presence of wound-induced rapid Ca²⁺ wave and electrical signal propagation mechanisms in the non-vascular plant, M. polymorpha. We induced localized damage to the thallus as a stressor and observed that the propagation velocity of both Ca²⁺ waves and electrical signals in Marchantia was approximately 1–2 mm s⁻¹, a rate akin to what has been reported for vascular plants. Furthermore, Ca²⁺ wave and electrical signal propagation were tightly coupled. Genetic studies notably highlight the crucial role of the sole GLR but do not implicate either type 1 or type 2 TPCs in the long-distance propagation of the Ca²⁺ wave and electrical signals. This compelling evidence suggests that the long-distance signaling mechanism-mediated by GLRs are evolutionarily conserved in both vascular and non-vascular land plants.

Results

Propagation of wound-induced Ca²⁺ waves and electrical signals in Marchantia thallus

We employed plastochron 2 stage plants, in which the thallus had been divided into four branches (Supplementary Fig. S1, Solly et al. 2017). In the asexual reproduction cycle, a Marchantia gemma develops dichotomous branching thallus from the two distinct meristematic zones (Supplementary Fig. S1, Solly et al. 2017). In this study, we designate the wounded branch as ‘branch 1’, while both branch 1 and branch 2 originate from the same meristematic zone of the gemma. In addition, ‘branch 3’ and ‘branch 4’ refer to branches that arise from different meristematic zones than ‘branch 1’ and ‘branch 2’. The central region of the thallus is referred to as the intermediate site.

To visualize Ca²⁺ dynamics in Marchantia, we introduced a fusion protein consisting of GCaMP6f, a Ca²⁺ sensor and mCherry, a Ca²⁺-independent fluorescent protein (Waadt et al. 2017). The fluorescence intensity of GCaMP6f and mCherry expressed with the MpEF1α promoter was particularly high in the notch area and gemma cup (Supplementary Fig. S2). This is consistent with the expression pattern of MpEF1α (Sauret-Güeto et al. 2020). Variations in the fluorescence intensity from site to site are likely due to variations in the expression levels of the fluorescent protein. When a branch of thallus was locally wounded by a needle, the increases in GCaMP6f fluorescence intensity were induced in both the wounded and unwounded sites (Supplementary Fig. S2, Video S1). In contrast, mCherry fluorescence intensity did not change before or after the wounding (Supplementary Fig. S2). To better visualize increased fluorescence intensity in areas of low GCaMP6f expression, binarized images were created showing areas of significantly increased GCaMP6f fluorescence intensity in black, compared to pre-wounding levels. The [Ca²⁺]_cyt elevation propagated distally from the wounded site in a wave-like manner (Fig. 1A, B).

Wounding the branch 1 induced Ca²⁺ waves propagated within the two branches (branches 1 and 2) emerging from the same meristematic zone of the gemmae in 92.8% (39/42) of the samples examined (Supplementary Fig. S3A). The rest of the samples (7.2%; 3/42), the Ca²⁺ waves, were confined only to branch 1, with no propagation extending to branch 2 (Supplementary Fig. S3A). Interestingly, Ca²⁺ waves did not propagate to branches 3 and 4 beyond the intermediate site in all samples tested (Fig. 1, Supplementary Figs. S1, S3A). The velocity of the Ca²⁺ waves, measured from positions 1 to 2, was 1,512 ± 585 µm s⁻¹ (n = 8). However, this velocity decreased as the wave traversed between branches and subsequently accelerated upon reaching the interior of the branch (Fig. 1C, D).

To investigate whether electrical signal propagation is induced by wounding in Marchantia, we measured the surface potentials of the thallus at four positions. Electrode 1 was positioned in proximity to the wounded site, electrode 2 was positioned in the intermediate site between branches 1 and 2 and electrode 3 was symmetrically placed with electrode 1 on branch 2. Electrode 4 was situated on branch 3. (Fig. 1E). The wounding induced changes in surface potentials in both wounded and unwounded branches (Fig. 1F). As with the Ca²⁺ waves, the change in surface potential did not propagate beyond the intermediate site to the two opposite branches (Fig. 1F). An identical response was observed in all eight samples tested.

The initial positive change in surface potential is not dependent on the alteration of membrane potential due to the influx and efflux of ions directly beneath the measuring electrode. The measuring electrodes are electrically interconnected through the electrolyte within the apoplast and medium, establishing an electrical circuit. Consequently, a shift in the surface potential at the stimulation site is detected as a positive surface potential change at an electrode distant from the stimulation site through the electrical circuit (Zawadzki and Trebacz 1985). Therefore, we focused on the timing of the decrease in surface potential. The decrease in surface potential was induced sequentially from positions 1 to 3, providing clear evidence that mechanical injury triggers the propagation of electrical signals in Marchantia. Even with an increased number of electrodes aimed at examining the dynamics of the electrical signals in more detail, a decrease in surface potential was induced in the order of proximity to the wounded site (Supplementary Fig. S4). Notably, the velocity of this electrical signal measured from positions 1 to 2 was 1,517 ± 958 µm s⁻¹ (n = 8), a rate nearly identical to the propagation velocity of Ca²⁺ waves in Marchantia and electrical signals in vascular plants.

Wound-induced [Ca²⁺]_cyt elevation and surface potential changes are linked

To elucidate the connection between electrical signals and Ca²⁺ waves, we conducted concurrent measurements of surface potential and [Ca²⁺]_cyt at two distal positions, surface potential at electrode 1 (E1)–electrode 2 (E2) and [Ca²⁺]_cyt in the region of interest 1 (ROI1)–ROI2 (Fig. 2A). We set ROIs in close proximity to the electrodes since the fluorescence of GCaMP6f cannot be observed directly beneath the electrodes. To assess whether there is a discrepancy in the timing of the increase in GCaMP6f fluorescence intensity based on the ROI’s location near the electrode, we measured the fluorescence intensity of GCaMP6f in ROIs at various positions near the electrode (Supplementary Fig. S6). There was no significant difference in the timing of the increase in GCaMP6f fluorescence intensity when ROIs were set at different positions in the vicinity of the electrode (Supplementary Fig. S6). We presumed that, under the current measurement conditions, the placement of the ROI close to the electrode would not significantly affect the measurement of signal arrival timing. Therefore, we arbitrarily positioned the ROI near the electrode and conducted the measurements. The negative changes in surface potential resulting from mechanical injury and the elevation in [Ca²⁺]_cyt occurred almost simultaneously (Fig. 2A). To ascertain whether the electrical signal and the Ca²⁺ wave exhibit concurrent arrival, we assessed the latency time (time 1) from the moment of wounding to the occurrence of negative surface potential changes and the elevation in [Ca²⁺]_cyt at positions 1 and 2. Our analysis indicates that the latency between the [Ca²⁺]_cyt elevation and the negative surface potential changes is closely similar in value. (Fig. 2B). A correlation was identified in the duration time (time 2) of these two signals (Fig. 2C). Furthermore, in order to confirm the concurrent propagation of the electrical signal and the Ca²⁺ wave, we assessed the propagation velocity of the two signals from position 1 to position 2 in each plant (Fig. 2D). To determine the equivalence of the velocity of the Ca²⁺ wave and the electrical signal, we calculated the differences between their propagation velocities in each plant. The analysis demonstrated that the difference in velocities between the two signals was statistically indistinguishable from zero (Fig. 2E). Collectively, these findings strongly suggest an association between negative surface potential changes and [Ca²⁺]_cyt elevation.

After wounding, the positive surface potential change was induced preceding the [Ca²⁺]_cyt elevation at position 4 (Fig. 2A). This positive surface potential change was also induced at the site where the negative surface potential change and [Ca²⁺]_cyt elevation did not occur (Supplementary Fig. S3). Positive change in surface potential after wounding appears to be independent of the [Ca²⁺]_cyt elevation and the negative surface potential changes.

Critical importance of Ca²⁺ channels and K⁺ channels in the propagation of electrical signals and Ca²⁺ waves

To understand which ion channels are involved in Ca²⁺ wave and electrical signal propagation, we treated Marchantia with LaCl₃, a Ca²⁺ channel inhibitor, and tetraethylammonium chloride (TEA), a potassium channel inhibitor.

When measuring the surface potential, the plants were placed in plastic chambers divided into multiple sections, and the section containing the inhibitor was partially treated by sealing it with Vaseline as a barrier to prevent its diffusion into the other sections. Surface potentials were measured in the inhibitor-treated section.

In Ca²⁺ imaging, the observation of plants within a plastic enclosure posed challenges. Therefore, the inhibitor treatment was administered by immersing the plant in a solution containing the inhibitor. Immersion in the solution led to an elevated percentage of individuals in which the Ca²⁺ waves were restricted to propagation solely within the wounded branches. This phenomenon was observed in 36.3% of the tested individuals, even when they were immersed in the solution without the inhibitor.

LaCl₃ inhibited both Ca²⁺ wave and electrical signal propagation (Fig. 3A–C, Supplementary Video S2), suggesting that the Ca²⁺ influx from the apoplast through the plasma membrane is responsible for [Ca²⁺]_cyt elevation and membrane potential depolarization. The propagation of Ca²⁺ waves necessitates the propagation of electrical signals.

Despite the critical importance of Ca²⁺ channels, possible involvement of K⁺ channels have remained poorly understood. TEA also suppressed both electrical signals and Ca²⁺ waves, suggesting the importance of K⁺ channels in the long-distance propagation of Ca²⁺ waves and electrical signals.

MpGLR is required for rapid long-distance signal propagation

Marchantia polymorpha has a single GLR that is phylogenetically close to the clade3 GLRs in Arabidopsis thaliana (Wudick et al. 2018). We have generated Mpglr-1^ko by genome editing and investigated whether MpGLR is required to propagate wound-induced Ca²⁺ waves and electrical signals (Supplementary Fig. S7A, B). In Mpglr-1^ko, wound-induced Ca²⁺ wave and electrical signal propagation were entirely suppressed (Fig. 4A, B, Supplementary Video S3). On the other hand, the [Ca²⁺]_cyt elevation and surface potential changes induced at the local wounded site was not suppressed in Mpglr-1^ko (Fig. 4C, D). In addition, slow Ca²⁺ wave propagation was observed at wounded sites in Mpglr-1^ko (Fig. 4C, Supplementary Video S4). To further validate whether GLR is essential for signal propagation, we developed a line in which proEF1α:MpGLR was introduced into the Mpglr-1^ko mutant. The incorporation of GLR complementary DNA into the plant body was confirmed by conducting PCR with primers designed to amplify the intron-containing region (Supplementary Fig. S7A, C). Subsequently, after generating the complementation line, GCaMP6f was introduced, resulting in different expression levels of GCaMP6f in the Takaragaike-1 and GLR-complemented lines. Wound-induced Ca²⁺ wave and electrical signal propagation were observed in the complementation line (Fig. 4A, Supplementary Video S3). Similar to the wild type (WT), Ca²⁺ waves did not propagate beyond the intermediate site to branches 3 and 4 in the complementation line. The velocity of signal propagation seemed to be faster in the complementation line than in the WT, possibly due to the expression of MpGLR by the EF1α promoter. In all of the eight tested complementation lines, Ca²⁺ waves propagated throughout branch 1 and branch 2 within one or two frames immediately after wounding. Consequently, accurate measurement showed that the velocity of signal propagation in the complementation line was not feasible. These results strongly suggest that MpGLR plays a crucial role in modulating the propagation of wound-induced Ca²⁺ waves and electrical signals. However, other ion channels, distinct from MpGLR, are likely responsible for [Ca²⁺]_cyt elevation and potential surface changes at the local wounded site.

MpTPCs are not required for long-distance signal propagation

Since TPC has been reported to be required for Ca²⁺ wave propagation in Arabidopsis (Choi et al. 2014, Kiep et al. 2015), we investigated whether MpTPCs are involved in Ca²⁺ wave and electrical signal propagation. Marchantia polymorpha possesses three homologs of TPC proteins, which are localized at the vacuolar membrane. Among these, only MpTPC1 exhibits slow vacuolar (SV) channel activity, akin to the TPCs found in vascular plants. In contrast, MpTPC2 and MpTPC3 belong to a distinct group separated from the TPCs found in vascular plants and notably lack SV channel activity (Hashimoto et al. 2022).

We examined the electrical signals and [Ca²⁺]_cyt kinetics in the mutants defective in MpTPC1 and MpTPC2/MpTPC3 (Hashimoto et al. 2022). In Mptpc1-1^ko and Mptpc2-2 tpc3-2^ko double mutant, wound-induced Ca²⁺ wave propagation was observed as in the wild-type (WT) (Fig. 5A, Supplementary Video S5). No significant differences were observed in the signal propagation velocity (Fig. 5B). Wound-induced electrical signals also propagated at a comparable velocity to WT (Fig. 5C). These results suggest that the MpTPCs are not involved in Ca²⁺ wave and electrical signal propagation.

Discussion

Vascular plants propagate long-distance Ca²⁺ and electrical signals from localized stress sites to distant tissues through the vascular bundle. Various models have recently been proposed for the mechanisms underlying long-distance signaling, grounded in the presence of vascular bundles (Farmer et al. 2014, Bellandi et al. 2022, Gao et al. 2023, Grenzi et al. 2023). However, in the present study, we demonstrated that M. polymorpha lacking vasculature exhibits a wound-induced rapid long-distance signaling. In this signaling paradigm, Ca²⁺ waves and electrical signal propagation are intricately intertwined. Remarkably, we observed that the propagation velocity of both Ca²⁺ waves and electrical signals across the thallus was approximately 1–2 mm s⁻¹ (Figs. 1–2), a rate reminiscent of what is observed in the vascular bundles of angiosperms (Mousavi et al. 2013, Nguyen et al. 2018, Toyota et al. 2018). The propagation of these two signals was also suppressed in Mpglr-1^ko mutant (Fig. 4), suggesting that a rapid long-distance signaling mechanism involving GLR is conserved in Marchantia.

In recent years, several models have been considered for the mechanism of long-distance signal propagation. For example, a calcium-induced calcium release model, in which [Ca²⁺]_cyt elevation itself induces [Ca²⁺]_cyt elevation in adjacent cells, has been proposed (Evans et al. 2016, Choi et al. 2017). On the other hand, some models do not require [Ca²⁺]_cyt elevation or depolarization of the plasma membrane potential for signal propagation. One model is that wound-induced change in turgor pressure in the vascular bundle propagates (hydraulic signal), and the pressure change activates the GLR by unknown mechanism, inducing [Ca²⁺]_cyt and depolarization of the plasma membrane potential (Farmer et al. 2014, Grenzi et al. 2023). Another model is that substances that activate GLR and induce [Ca²⁺]_cyt elevation and depolarization of the membrane potential diffuse thorough vascular bundle from the wounded sites (Bellandi et al. 2022, Gao et al. 2023). In this study, we have shown that a rapid Ca²⁺ wave and electrical signal propagation mechanism involving GLR is conserved in non-vascular plants (Fig. 4). When the signal propagated from the wounded branch 1 to branch 2, its propagation velocity decelerated between the two branches and subsequently accelerated (Fig. 1C, D). These results imply that rapid Ca²⁺ wave and electrical signal propagation in Marchantia cannot be attributed to the direct diffusion of substances that induce plasma membrane depolarization or [Ca²⁺]_cyt elevation from the wounded sites to distal sites. Whether GLRs can be activated by alterations in plasma membrane potential or by an increased concentration of glutamate in the apoplast has been a topic of an ongoing debate (Moe-Lange et al. 2021, Grenzi et al. 2023). Testing these possibilities may lead to the understanding of the mechanism of GLR activation in non-vascular plants.

The potential interconnection and reciprocal regulation between Ca²⁺ waves and electrical signals have been a topic of discussion (Gilroy et al. 2016, Choi et al. 2017). Ca²⁺ has been suggested to be involved in membrane depolarization in various plants (Vodeneev et al. 2016). Inhibition of Ca²⁺ influx from outside the cell by La³⁺ suppressed wound-induced electrical signal propagation in Marchantia (Fig. 3). Both wound-induced Ca²⁺ waves and electrical signal propagation depend on MpGLR (Fig. 4). These results indicate that GLR-mediated rapid long-distance signaling involving Ca²⁺ waves and electrical signals appears to be highly conserved among vascular and non-vascular land plants.

[Ca²⁺]_cyt elevation and surface potential changes induced at the local wounded site were not suppressed in Mpglr-1^ko or La³⁺, TEA treatment (Figs. 3, 4), and slow Ca²⁺ wave propagation was observed at wounded sites in Mpglr-1^ko (Fig. 4B, Supplementary Video S4). In the vicinity of the wounded sites, other ion channels, distinct from MpGLR, may play a crucial role in local plasma membrane depolarization and propagation of slow and localized Ca²⁺ wave. The slow and localized Ca²⁺ wave may also be attributed to the diffusion of Ca²⁺ from the wounded cells into the surrounding cells. Given the much less genetic redundancy in M. polymorpha in comparison with angiosperms such as A. thaliana (for example, only 1 GLR in Marchantia, while 20 GLRs in Arabidopsis), the present experimental system would provide an excellent model to identify various ion channels involved in and elucidate the molecular mechanisms for rapid long-distance signaling conserved in land plants.

The long-distance signal propagation was inhibited not only by a Ca²⁺ channel blocker but also by TEA, an inhibitor for K⁺ channels (Fig. 3). TEA inhibits depolarization of cell membrane potential in Conocephalum conicum and Physcomitrium (Physcomitrella) patens (Trebacz et al. 1989, Johannes et al. 1997). In Conocephalum, the equilibrium potential of K⁺ is higher than the resting membrane potential, and opening of the K⁺ channel may induce depolarization of the membrane potential (Trebacz et al. 1994, Kisnieriene et al. 2022). Furthermore, simultaneous measurement of [Ca²⁺]_cyt and surface potential revealed that these two signals are tightly interconnected (Fig. 2). These results underscore the critical role of K⁺ channels in mediating membrane depolarization, and both [Ca²⁺]_cyt elevation and depolarization of the plasma membrane potential are required for the propagation of these two signals.

Wound-induced Ca²⁺ waves and electrical signals failed to propagate to half of the body through intermediate sites (Fig. 1). Marchantia exhibits a reproductive strategy involving the formation of clones known as gemmae, which proliferate within specialized structures called gemmae cups during asexual reproduction. Gemmae remain in a dormant state characterized by arrested cell division and differentiation. Notably, gemmae possess two apical meristems, which become active upon breaking dormancy, initiating growth from the apical meristems (Shimamura 2016). The cells located in the intermediate sites where Ca²⁺ waves fail to propagate are presumed to be gemmae cells by origin. These gemmae cells differ from other thallus cells in which they have remained in a dormant state and did not originate from the apical meristem. The present study has demonstrated that the sole MpGLR is crucial for the propagation of Ca²⁺ waves and electrical signals (Fig. 4). Additionally, studies involving the overexpression of PLASMODESMATA-LOCATED PROTEIN 5 (PDLP5) and the examination of plasmodesmata (PD)-associated β-1,3-glucanase-deficient mutants with reduced PD conductance have indicated that Ca²⁺ wave propagation at non-vascular sites is inhibited (Toyota et al. 2018). It is plausible that gemmae cells have lower expression levels of factors associated with rapid long-distance signaling. Consequently, this reduced expression may account for the failure of Ca²⁺ waves to propagate in gemmae cells.

Ca²⁺ wave and electrical signal propagation was induced in Mptpc1^ko mutants as in WT (Fig. 5). In Arabidopsis tpc1 mutant, the wound-induced systemic [Ca²⁺]_cyt elevation is suppressed at distal sites (Kiep et al. 2015). TPC opening is facilitated by Ca²⁺ binding to the EF-hand at the cytosolic (Beyhl et al. 2009, Schulze et al. 2011). It has been proposed that TPCs may be activated initially by a rise in Ca²⁺ levels, potentially regulated by factors such as GLRs, and subsequently amplify the signal by modulating Ca²⁺ release from the vacuole (Kiep et al. 2015). Conversely, recent studies have reported the induction of Ca²⁺ wave and electrical signal propagation in tpc1 mutants (Moe-Lange et al. 2021, Bellandi et al. 2022). In Marchantia, neither MpTPC1 (type 1 TPC) nor MpTPC2 and MpTPC3 (type 2 TPCs) seem to be indispensable for the propagation of the Ca²⁺ waves and electrical signals themselves.

In summary, this study demonstrates that GLR-mediated long-distance signaling mechanisms are conserved in liverworts, non-vascular plants and presumably in all land plants. However, many questions remain unanswered regarding the activation of GLR during signal propagation and the factors that play a role in signal propagation itself. Marchantia, characterized by its relatively low genetic redundancy and the absence of vascular bundles, serves as a promising model organism for investigating the factors involved in long-distance signaling and gaining insights into the propagation mechanisms.

Materials and Methods

Plant material and growth conditions

Marchantia polymorpha accession Takaragaike-1 (Tak-1; male) was used as wild-type. Plants were grown on the half-strength Gamborg’s B5 solid medium containing 1% agar, 1% sucrose and 1 g l⁻¹ MES (pH 5.5, adjusted by KOH) under continuous light (approximately 40 µmol m⁻² s⁻¹) at 21°C. Fourteen-day-old plants were used for Ca²⁺ imaging and surface potential measurements.

Plasmid construction and transformation

GCaMP6f-mCherry was cloned into the NotI-AscI site of pENTR/D-TOPO (Invitrogen). The entry clones were subcloned into pMpGWB303 vector (Ishizaki et al. 2015) using the Gateway LR reaction. Complementary DNA (cDNA) of MpGLR was cloned into pENTR/D-TOPO using primers 5ʹ-TTTTGCGGCCGCTCAAATGACACTATTTCTGTCC-3ʹ and 5ʹ-TTTTGGCGCGCCTCAAATGACACTATTTCTGTCC-3ʹ. The entry clones were subcloned into pMpGWB337 vector (Nishihama et al. 2016) using the Gateway LR reaction. To obtain Mpglr-1^ko mutant, the gRNA was designed using a web tool CRISPRdirect (https://crispr.dbcls.jp) and cloned into the pMpGE013 vector (Sugano et al. 2018). These vectors were transformed into Agrobacterium GV2260. Transformation into Marchantia was performed using the G-Agar trap method; transformants were selected using hygromycin (Tsuboyama and Kodama 2018). To confirm that the MpGLR cDNA had been introduced, DNA was extracted from the plants and PCR was performed using two primers, 5ʹ-AGCTCCACCAAGCCCAAGTC-3ʹ and 5ʹ-CGCTTGGTGGTTATCGTTAC-3ʹ.

Wounding assay

A robotic arm, Do-bot MG400 (product model; DR-MG-4R005-02E), was equipped with a 20-cm-long, 5-mm-square plastic rod. A needle (tungsten, 0.5 mm diameter) was attached to the end of the rod. Wounding was inflicted by the robotic arm, which was inserted vertically just above the thallus so that the needle penetrated the thallus completely.

Ca²⁺ imaging

For Ca²⁺ imaging, GCaMP6f-mCherry-expressing lines were used. The images were acquired by a Nikon SMZ25 equipped with Nikon DS-Ri2 camera, P2-SHR Plan Apochromatic 1× objective lens and Nikon INTENSILIGHT C-HGFIE lamp. For GCaMP6f image acquisition, a 470/40 nm excitation filter, a 500 nm long-pass dichroic and a 535/50 nm emission filter were used, and for mCherry images acquisition, a 545/15 nm excitation filter, 570 nm long-pass dichroic and 620/30 nm emission filter were used. To increase the number of images that could be acquired per unit time, fluorescence images of GCaMP6f and mCherry were not taken simultaneously, but only fluorescence images of GCaMP6f were taken to calculate the Ca²⁺ wave velocity. In this case, fluorescence images of mCherry were taken for 20 seconds before acquiring images of GCaMP6f.

Image analyses

Image analysis was performed using ImageJ Fiji (Schindelin, 2012). Fluorescence intensity values were obtained as the average intensity within a circular ROI. GCaMP6f intensity was divided by mCherry intensity, and the calculated GCaMP6f/mCherry ratio was normalized by subtracting the average ratio values of the 10 frames before wounding from the ratio value at each time. The velocity was calculated by dividing the lag time for a significant fluorescence increase of GCaMP6f within each ROI by the distance between the ROIs. The time at which the fluorescence intensity of GCaMP6f exceeded the average fluorescence intensity + 3*SD of the 10 frames before wounding was considered the time at which a significant increase in fluorescence occurred. ΔF images and videos were created by subtracting the average image of the 10 frames before wounding from each frame. Binarized images were created by subtracting the average +2*SD image of the 10 flames before wounding from each frame. To calculate the duration time, the derivative of GCaMP6f fluorescence intensity and surface potential was used. The duration time of [Ca²⁺]_cyt elevation was calculated as the time from when the derivative of the fluorescence intensity of GCaMP6f exceeded the average values + 3*SD of the 10 values before wounding to the time when it dropped below this threshold. The duration time of surface potential was calculated as the time from when the derivative of surface potential dropped below the average values −3*SD of the 10 values before wounding to the time when it exceeded the threshold.

Surface potential measurement

Individual thalli together with agar were placed in small Petri dishes (35 mm). The Petri dish containing the thallus was covered with a stretched transparent foil to avoid evaporation. Thin sharp Ag/AgCl electrodes were gently inserted into the thallus through the foil, one day before the experiment. The reference electrode was placed on the other part of the plant, behind the narrowing through which neither Ca²⁺ waves nor electrical signals were able to pass. Electrical signals (surface potential changes) were registered with an iWorx 304T (iWorx, USA) amplifier running under Labscribe 2.0 software (iWorx, USA).

For the concurrent measurement of [Ca²⁺]_cyt and surface potential, a custom-designed device was employed (Supplementary Fig. S5). A 5-µl volume of 0.8% agarose gel containing 10 mM KCl was placed on the surface of thallus, through which an electrode was brought into contact. The reference electrode was placed in the medium. The gain of the device was set to 60× and connected to an Arduino Uno A/D converter, and the voltage was measured every 0.5 s. Arduino IDE is used to send/receive data and write programs to/from Arduino Uno.

Inhibitor treatment

In 10 mM HEPES buffer (pH 5.5, adjusted by KOH), 5 mM LaCl₃ and 5 mM TEA were dissolved. In surface potential measurements, the plants were placed in a plastic case divided into three compartments, one of which was filled with solution. The compartment containing the inhibitor was sealed with Vaseline to prevent the inhibitor from leaking into other compartments. The solution was applied for 30 min and then removed before measurements were started. In Ca²⁺ imaging, plants were completely immersed in the solution for 30 min. The solution was removed by aspiration, and the Ca²⁺ imaging was performed 2 h later. In the electrical signal measurement, elliptic well ca. 3–5 mm made of Vaseline was formed on one of the branches of the thallus. The well was filled with a standard solution supplemented with an ion channel inhibitor. Electrodes were placed inside and outside the wells. Stimuli were applied 30 min after filling the wells with the solution.

Supplementary Data

Supplementary Data are available at PCP online.

Data Availability

The data underlying this article are available in the article and in its online supplementary material.

Funding

Grant-in-Aid for Scientific Research on Innovative Areas JP22H04734 from MEXT, Japan; JSPS KAKENHI Grant Number 23H02505 and Bilateral Joint Research Project #JPJSBP120214601.

Acknowledgments

We thank Dr. Rainer Waadt for providing GCaMP6f-mCherry plasmid.

Disclosures

The authors have no conflicts of interest to declare.

References

Author notes