Signaling and transport processes related to the carnivorous lifestyle of plants living on nutrient-poor soil

Abstract

In Eukaryotes, long-distance and rapid signal transmission is required in order to be able to react fast and flexibly to external stimuli. This long-distance signal transmission cannot take place by diffusion of signal molecules from the site of perception to the target tissue, as their speed is insufficient. Therefore, for adequate stimulus transmission, plants as well as animals make use of electrical signal transmission, as this can quickly cover long distances. This update summarises the most important advances in plant electrical signal transduction with a focus on the carnivorous Venus flytrap. It highlights the different types of electrical signals, examines their underlying ion fluxes and summarises the carnivorous processes downstream of the electrical signals.

Electrical signaling in plants

Long-range signaling in organisms within Eukaryotes is essential for flexible responses toward environmental threats and challenges. As such, plants’ systemic acquired adaptation is based on the ability to convert external stimuli perceived by only a few cells into mobile signals, which lead to adaptation reactions even in distant tissues or organs. In animal organisms, electrical signals and their transmission via nerve cells are the key to rapid adaptation. These electrical circuits are based on a dense network of nerve cells connected via axons, which operate as cables that conduct information encoded by the number and frequency of membrane potential changes (i.e. action potentials [APs]).

ADVANCES

In contrast, land plants do not possess specialized nerve cells with axons, but still their information transmission is dependent on long-range signaling. In plants, the vascular system (vascular bundle) provides for the distal signal propagation, including the long-distance transport of nutrients, metabolites and information given in the form of RNA, proteins, phytohormones (for review, see Kehr and Kragler, 2018; Liu and Chen, 2018; Morris, 2018; Koenig and Hoffmann-Benning, 2020; Tolstyko et al., 2020; Xia and Zhang, 2020) as well as electrical signals. The nature of this electrical signaling is based on changes in the activity of ion channels (Hodick and Sievers, 1988; Felle and Zimmermann, 2007; Fromm and Lautner, 2007; Hedrich, 2012; Hedrich and Neher, 2018), electrogenic membrane transporters, and active transport systems (Hills et al., 2012; Hoffmann et al., 2020; Reyer et al., 2020).

• After almost a century and a half of research on carnivorous plants, researchers following the footsteps of Charles Darwin have gained comprehensive knowledge of plant signaling and membrane transport using these research objects.

• The understanding of the ion fluxes and transport processes involved in APs has made great strides forward since that time.

• Among other results, the close interaction between calcium and the plant AP has been clearly demonstrated.

• The downstream signal processes following the AP, especially JA signaling, have been well elucidated.

Changes in plasma membrane (PM) potential or modulation of ion fluxes are among the earliest cellular responses toward external stimuli (for review, see Zimmermann et al., 2009; Choi et al., 2016; Huber and Bauerle, 2016; Sukhova et al., 2017; Szechyńska-Hebda et al., 2017; Hilleary and Gilroy, 2018; Volana Randriamandimbisoa et al., 2020). In plants, long-distance electrical signaling events in response to environmental stimuli are widely described in three different manifestations: (1) variation potentials (VPs; these highly variable signals are also called electro-potential waves or slow wave potentials), (2) system potentials (SPs), and (3) APs.

VPs and SPs are distinct from APs and are thought to be more closely related to the activity and regulation of the PM H⁺-ATPase, which adjust the very negative resting membrane potential in plant cells. Its inhibition might trigger VPs by depolarizing the membrane potential. Besides, Ca²⁺, glutamate, reactive oxygen species (ROS), anions, and hydraulic signals are also discussed as elicitors of VPs (Vodeneev et al., 2015; Nguyen et al., 2018; Blyth and Morris, 2019). On the other hand, sustained activation of the proton pump can create SPs via hyperpolarization of the membrane potential (Zimmermann et al., 2009). Unlike APs, VPs and SPs are not self-propagating and can vary in their amplitude and duration depending on the stimulus perceived (Choi et al., 2016). In contrast, APs follow the “all or nothing principle.” An AP is triggered as soon as a certain stimulus threshold is exceeded, and the following change in membrane potential is always similar with respect to kinetics and duration. Therefore, stimulus strength is not reflected by the amplitude of the AP but rather by its numbers and frequency.

Of these three electrical signals, APs propagate most rapidly within the plant body, with typical propagation speeds calculated in the centimeter-per-second range (Volkov et al., 2008; Vodeneev et al., 2015). In comparison, nerve transmission rates in mammals reach about 100 m s⁻¹ (Waxman and Bennett, 1972).

Ca²⁺ initiates plant APs

The predominant opinion on the initiation of the plant AP suggests that in response to an external stimulus initially the cytosolic concentration of the second messenger Ca²⁺ transiently increases. The origin of this Ca²⁺ influx from the apoplast is still under debate, although recently some Ca²⁺ permeable channels have been found to influence the cytosolic Ca²⁺ homeostasis (Tian et al., 2020). Among these, cyclic nucleotide-gated channels (Gao et al., 2014, 2016; Pan et al., 2019) and glutamate-like receptors (GLRs; Michard et al., 2011; Mousavi et al., 2013; Nguyen et al., 2018; Toyota et al., 2018; Wudick et al., 2018; Johnson et al., 2019; for structural analyses, see Alfieri et al., 2020; Gangwar et al., 2020) seem to play critical roles. The putative involvement of GLRs further indicates that glutamate may serve as a chemical trigger for electrical signaling. Furthermore, the recently discovered calcium-permeable OSCA (reduced hyperosmolality-induced [Ca²⁺]_i increase) channels are involved in plant osmosensing (Yuan et al., 2014; Zhai et al., 2020) and immunity (Thor et al., 2020). Besides OSCAs, only a few mechanosensitive (MS) ion channels have been functionally described in plants (Basu and Haswell, 2017). Among these, MS-Like (MSL) channels were characterized to exhibit a preferred conductance for anions (Basu et al., 2020; Li et al., 2020; Procko et al., 2021). Thus, it is unknown whether these MSLs are directly involved in Ca²⁺ fluxes induced by mechanical stress. The rather recently explored class of Piezo proteins in mouse and Drosophila represents the pore-forming subunits of MS channels and responds to mechanical stimuli. Many eukaryotes, including plant species, encode a single Piezo protein in their genome (Coste et al., 2010), and recent studies on plant Piezo homologs suggest a mechano-induced Ca²⁺ permeability (Zhang et al., 2019; Mousavi et al., 2020).

Although the detailed molecular basis of stimulus-induced Ca²⁺ influx in plants is still unsolved, an initial small depolarization occurs via Ca²⁺ influx through voltage- or mechanically gated Ca²⁺ channels (e.g. Hedrich and Becker, 1994). The transient rise in [Ca²⁺]_cyt in turn could trigger the efflux of Cl⁻ through Ca²⁺-dependent anion channels, leading to further depolarization of the membrane potential. The subsequent activation of voltage-gated, depolarization-activated K⁺ channels counteracts anion efflux and ensures that the cell repolarizes (Choi et al., 2016). Finally, the resting potential is established through an interplay between potassium conductivity and the PM-localized proton pump (Figure 1). However, despite such mechanistic ideas about how the plant AP is generated, in most cases, the exact molecular nature of the channels involved has remained undefined.

Figure 1.

Schematic model of the signaling cascade from prey perception to digestion in Dionaea. Touching the TH (center) is converted into an AP. The predominant opinion on the ion fluxes involved, which lead to a plant AP, are shown in the upper left corner. An initial Ca²⁺ influx subsequently triggers Cl⁻ efflux-associated depolarization of the membrane potential. This depolarization is compensated by the activation of K⁺ efflux. The following repolarization is dominated by an H⁺ efflux. Finally, the original Ca²⁺, Cl⁻, and K⁺ concentrations have to be re-adjusted again. Besides showing different ion permeabilities, the involved transport proteins also exhibit diverse transport activities, kinetics, and voltage dependencies. Thus, the mediated ion fluxes show overlapping phases, which are important for the fine tuning of the AP. The touch-induced AP leads to a calcium wave in the trap and when two APs are triggered within a short time interval, the Ca²⁺ threshold value is exceeded, and the trap closes quickly. Furthermore, the electrical signal also causes an induction of the plant hormone JA. After continuous electrical stimulation, the JA signaling pathway is activated, resulting in the expression of digestive enzymes and finally in prey digestion.

There are countless instances on the role of long-distance electrical signaling in plants. The fast movements of Mimosa pudica leaves and Dionaea muscipula traps after mechanostimulation are among the most vivid and impressive examples. Carnivorous plants consume animals to obtain mineral nutrients, including potassium, phosphorus, and nitrogen, to enhance growth, competitive ability and reproduction in nutrient-poor environments (Box 1). Whenever an animal prey touches one of the sensory hairs (trigger hairs [THs]) on the inner trap surface of the Venus flytrap, an AP is fired to trigger trap closure and prey capture (Figure 1). These electrical signals are particularly robust and easy to reproduce experimentally. For this reason, the first plant AP was measured in Dionaea already one and a half centuries ago (Burdon-Sanderson, 1882). With modern techniques, these APs can even be detected indirectly and without contact, since the underlying currents in the traps are so intense that magnetic fields are generated able to be measured at some distance with magnetic field detectors (Fabricant et al., 2021). Because of these robust APs, a great deal of work on electrical signals in plants has been done on the Venus flytrap. In this review, we therefore focus on carnivorous examples from the plant electrical signaling literature and following membrane transport mechanisms underlying the carnivorous lifestyle, including prey recognition and capture, digestion, and nutrient uptake.

Box 1.

Evolution and advantages of carnivory

To be defined as a carnivorous plant, the species must (1) have a clear adaptation by which it is able to attract, catch, or digest prey and (2) be able to absorb nutrients from animals and thus achieve an increased fitness (Givnish et al., 1984). Undoubtedly, part of their great appeal to scientists is that carnivorous plants have found an evolutionary way to evolve from (defenceless) food to (active) hunters. To this end, they are often equipped with remarkable lures, traps, stomachs, and in a few cases, exceptional speed of movement. Carnivorous plants evolved in five independent orders of flowering plants, including monocots and eudicots (Albert et al., 1992). The earliest evolution of a carnivorous lineage has been dated to 84.8 million years ago (mya; Nepenthes), with the most recent representative, Brocchinia reducta, estimated at only 1.9 mya (Ellison and Adamec, 2018). All these plants have found a special way of dealing with nutrient deficiency. Whether the carnivory has an advantage over the normal plant physiology can be explained by a cost/benefit calculation (Givnish et al., 1984). Only when the obtained nutrients derived from prey outweigh the additional expenditure (movement and production of specialized tissues/organs/proteins) and the associated loss of photosynthetically active tissue, the carnivorous lifestyle can prevail. In areas that are limited in nutrient availability, but not in the supply of energy (light) and water, the cost/benefit and thus carnivory can be achieved. Therefore, evolution has been able to produce unique plants over the course of time that has established a completely new food source. One of the major questions regarding the evolution of carnivory is whether carnivorous plants would possess entirely unique genes and if so, whether those genes were assimilated by the prey via, e.g. horizontal gene transfer (HGT). This question was answered through the increasing number of sequenced carnivore genomes and transcriptomes (Ibarra-Laclette et al., 2013; Leushkin et al., 2013; Bemm et al., 2016; Butts et al., 2016; Fukushima et al., 2017; Iosip et al., 2020; Palfalvi et al., 2020; Silva et al., 2020; ZulkapLi et al., 2021; for review, see Hedrich and Fukushima, 2021). It became clear that the evolution of carnivory was not associated with the “invention” of novel genes or HGT. Rather the repertoire of existing signaling pathways and innate immunity-related processes were neo-functionalized and transformed to fit the purpose of a carnivorous lifestyle.

Electrical signals control prey capture

The capture organs of carnivorous plants derive from modified leaves and function with two different types of trapping mechanisms. Some carnivorous plant species use immobile (passive) traps, which do not rely on an organ movement for capture purpose. Among these, trap types based on the principle of a pitfall use a nonadhesive, slippery surface, and the physical property of gravity to catch prey. The so-called pitfall traps are found in Cephalotus, Nepenthes, Darlingtonia, Heliamphora, and Sarracenia. They form pitcher-shaped leaves with slippery edges and are constitutively filled with a digestive fluid. The secretion of digestive enzymes is elevated after prey capture, which serves as a chemical stimulus (Owen and Lennon, 1999; An et al., 2002; Bohn and Federle, 2004; Bauer et al., 2013; Gao et al., 2015; Bohm et al., 2016a; Lee et al., 2016; Rottloff et al., 2016; Scherzer et al., 2017). Brocchinia reducta and Catopsis berteroniana are so-called tank bromeliads, which have erected, overlapping leaves that store rainwater for carnivory as well as for water usage. Another variety of trap is found in the corkscrew plants of the genus Genlisea. Some species of this genus possess the smallest carnivorous genomes and grow in moist terrestrial to semi-aquatic environments and form so-called eel traps (Leushkin et al., 2013). These highly modified underground leaves trap and digest minute microfauna. Although proposed a century earlier by Charles Darwin, carnivory was not demonstrated in the genus until 1998 (Barthlott et al., 1998). In contrast, flypaper-type traps exhibit strong adhesive properties brought about by specialized gland cells to capture the insect. A very simple kind of adhesive trap can be found in the leaves of Pinguicula grandiflora, which have specialized stalked glands that exude a sticky mucilage to catch small prey (Legendre, 2000).This group of carnivorous plants that use adhesive properties to catch prey further includes, for example, the flypaper traps of Drosera, Drosophyllum, Triphyophyllum, Roridula, Byblis, and Philcoxia. These adhesive traps are the most common category of trapping mechanisms. While most of these traps do not perform any movement (called passive adhesive traps), Drosera and some species of Pinguicula and Byblis combine the adhesion properties with relatively “slow” movements (called: active adhesive traps). In certain species of Drosera, glandular hairs called tentacles move toward captured prey and secret the digestive fluid. To sense the caught prey and react with movement, Drosera rotundifolia elicits an AP when the tentacles are mechanically stimulated. These APs are significantly faster in duration (∼10 s) and in travelling (0.5 cm s⁻¹) across the tentacle tissue (Williams and Spanswick, 1976) than APs observed in nonexcitable plants such as Arabidopsis (Arabidopsis thaliana; ∼80 s and 0.04 cm s⁻¹, respectively; Mousavi et al., 2013). This use of an electrical signal enables a fast response to the prey by triggering further tentacle movements within half a minute to immobilize the insect. Interestingly, also a particularly fast movement is found in the “catapult” traps of Drosera glanduligera and a few related species in which outer tentacles can snap within several hundred milliseconds to smash preys onto sticky droplets of inner tentacles (Poppinga et al., 2012; Ellison and Adamec, 2018; Kocab et al., 2020).

The carnivorous group with active/motile traps uses organ movement for capture purposes. The fastest plant movement is found in converted leaves of the carnivorous genus Utricularia. This genus has developed aquatic suction traps that can catch the smallest aquatic animals within half a millisecond (Plachno et al., 2020; Whitewoods, 2020). Snap-trap leaves are found in Dionaea and Aldrovanda and consist of a sister leaf pair with stored energy in the open state to enable fast snap closure. Regardless of whether the trap movements are in the low millisecond range or are so slow that they are hardly visible, carnivorous plants with active/motile traps must be “alerted” to detect prey in order to start their movement. However, similar to the animal system, this perception can also be overridden in plants by anesthetics (Box 2; De Luccia, 2012; Weinrich et al., 2012; Weinrich and Worcester, 2013; Rinaldi, 2014; Craddock et al., 2017; Turin and Skoulakis, 2018; Yokawa et al., 2018, 2019; Pavel et al., 2020; Pavlovic et al., 2020; Draguhn et al., 2021). In Dionaea, the prey is sensed by unique mechanically sensitive THs (Figure 1). The mechanical stimulation upon bending of these specialized sensory TH is translated into the electrical signal of an AP (Burdon-Sanderson, 1882; Escalante-Perez et al., 2011; Volkov, 2019). The duration of Dionaea’s APs is very short (∼1 s) and they travel 6–25 cm s⁻¹ across the trap to allow and ensure prey capture (Volkov et al., 2008). This mechanic-electrical translation is a yes-no answer, and thus virtually “digital” (Masi et al., 2016; Krausko et al., 2017; Hedrich and Neher, 2018). If the prey’s touch is below the detection threshold, no stimulus translation and transmission will occur. Recently, the force required to bend/stimulate a sensory hair of Dionaea by an insect was determined, showing that even forces in the low micronewton range are sufficient to be translated into APs (Scherzer et al., 2019; Burri et al., 2020; see also Box 3). In further micro-CT (computed tomography)-based analyses “hotspots” of sensory cells in the TH base were reported with higher mechanotransduction activity compared to other regions (Saikia et al., 2021).

Box 2.

Anesthetics interfere with electrical signaling

Animal studies show that stimulus conduction can be suppressed by anesthetics. Recent studies on noncarnivorous plants such as pea (Yokawa et al., 2018), Mimosa (De Luccia, 2012), and garden cress (Yokawa et al., 2018), as well as on carnivorous plants such as Dionaea (Volkov et al., 2014, Yokawa et al., 2019) and Drosera (Yokawa et al., 2018) have shown that the effect of anesthetics in plants exhibits parallelisms to the animal system, resulting in the absence of APs and thus in prevented or reduced perception of ambient stimuli. However, it remains a mystery how and why different chemical compounds show comparable effects in animals and plants (Rinaldi, 2014; Draguhn et al., 2021).

In general, several mechanisms and sites of action are discussed: anesthetics (1) show a direct interaction with proteins (Rinaldi, 2014; Craddock et al., 2017); (2) might affect the lipid membrane structure leading to modulated channel function (Weinrich et al., 2012; Weinrich and Worcester, 2013; Pavel et al., 2020); and (3) interfere with electron transfer pathways (Turin and Skoulakis, 2018). Eventually, there is always an impairment of the stimulus transmission and finally the absence of electrical signals. The figure illustrates how a very common anesthetic, diethyl-ether, affects the AP generation in Dionaea. A trap was dry-impaled and exposed to a saturated ether-atmosphere. Continuous mechanical stimulation of a TH (arrows) resulted in gradually decreasing AP amplitudes until finally no AP could be triggered. Such behavior has never been observed under control conditions. The decrease in amplitude and the absence of APs can be explained by an impact of ether at the channel activity of proteins involved in AP generation and transmission, because the resting potential itself is not influenced. Which proteins, and by which mechanism they are affected, must be clarified. The "wake-up-phase" of anesthetized Dionaea traps has already been shown. The effect diminishes over time and mechanostimulation leads to small AP reactions already after 100 s of ether removal, and AP amplitudes return to initial values after about 10 min. Interestingly, ether also inhibits the downstream signaling processes of mechanostimulation (Pavlovic et al., 2020). This fact illustrates the trap’s complete disconnection from the environment (comparable to the unconsciousness of animals) and shows that during anesthesia no “unconscious” stimuli are perceived at the level of the downstream signal cascades. Compared to a good surgical anesthesia, even a plant cannot remember anything afterward.

Box 3.

The Venus flytrap protects itself from wasting energy

To achieve the best possible cost/benefit ratio, Dionaea must control and optimize its expenditures. Therefore, several ingenious protective mechanisms have been established during Dionaea’s development.

Mechanism 1: The trap only snaps when two APs are triggered in quick succession (Darwin, 1875). An accidental AP caused by a nondigestible object, such as rain, wind, or dead matter falling into the trap will result in a single AP, which is usually not sufficient to close the trap and thus to exhaust the stored energy.

Mechanism 2: THs of a large trap are quite stiff and need a greater force to be bent compared with smaller traps. In this way, Dionaea avoids large traps initiating an unprofitable digestion process for a prey that is too small. A better cost/benefit ratio is achieved in smaller traps, which exhibits a higher sensitivity (Scherzer et al., 2019). Dionaea, therefore, ensures that the trap always has an optimal size in relation to the prey. However, if particularly strong but small animals trigger two APs in large traps, traps will close quickly, but the small animals can escape through the gaps between the outer hairs (see figure illustration). In fact, not all Venus flytrap attempts are crowned with success, and many animals can escape in time before the trap is completely closed (Ellison and Gotelli, 2009). The third important protective mechanism ensures that the trap does not start a costly digestion without benefit from a prey.

Mechanism 3: If only two APs are triggered, the trap closes quickly, but without a trapped insect, only these two APs will remain. In this case, the trap registers the unsuccessful attempt and opens again after a few hours. When a prey has been caught, it will continue to move, hoping to escape. This panicked movement ultimately leads to countless more APs in the trap over many hours. With this information, the trap now senses success and starts the complete digestive process and fine-tunes the amount of digestive enzymes related to the number of APs fired (Bohm et al., 2016a) and therefore to the strength and size of the prey. With these three sophisticated mechanisms, Dionaea avoids unnecessary energy consumption and fine-tunes its digestion to achieve an optimal cost/benefit ratio.

Ca²⁺ and the electrical signal go hand-in-hand

In order to investigate how the mechanical stimulus of bending the sensitive TH is translated into an electrical AP, a recent study (Iosip et al., 2020) analyzed the organ-specific expression pattern of ion transporters. Compared to trap tissue that is not designed to be mechanosensitive, including digestive glands and the trap rim, the expression profile of TH appeared quite unique. Particularly, MS channels (Ranade et al., 2015) of the MSL-type (Basu and Haswell, 2017; Guerringue et al., 2018) and OSCAs (Yuan et al., 2014; Murthy et al., 2018; Thor et al., 2020) were found to be highly expressed. In noncarnivorous plants like Arabidopsis, these touch sensors are activated in response to wound-, mechano-, or osmostimulation (Deng et al., 2020; Farmer et al., 2020). Recently, the Dionaea MSL homolog DmFLYC1 was reported as a mechanically activated chloride-permeable channel (Procko et al., 2021). In cooperation with stretch-activated Ca²⁺ permeable OSCAs, these channels could accompany the initial depolarization phase of the AP (Figure 1).

In 2020, for the first time, Dionaea was stably transformed with the calcium reporter GCaMP (Suda et al., 2020). The resulting and fascinating images show that the mechanical stimulation of a TH can be associated with a calcium wave that spreads concentrically from the site of stimulation and even travels across the second trap lobe. However, the extent to which this calcium signal is related to the electrical transmission of APs is questioned, as previous measurements of the AP propagation speed (6–25 cm s⁻¹; Sibaoka, 1966; Volkov et al., 2008) do not correspond to the recently determined calcium wave speed (2 cm s⁻¹). The previously reported velocities may be overestimated because electrical stimulation interferes with the AP measurements. To address this divergence, we performed a new set of experiments in which the propagation of the calcium wave and the AP were analyzed in parallel, using the calcium reporter plants described above. Here, the propagation velocity of the AP from the sensory hair was measured using two separate surface potential electrodes on the trap (Figure 2, A and B). In this process, we have noticed that the propagation speed of the AP is decisively influenced by temperature (Figure 2C), which can also cause divergence in speed estimations. Furthermore, these experiments clearly demonstrated the coexistence and synchronous velocities of the calcium signal and the AP (1.91 ± 0.23 cm s⁻¹ for AP propagation- and a Ca²⁺ wave speed of 2.06 ± 0.34 cm s⁻¹ (mean ± sd, n ≥ 4)), confirming the assumption that calcium plays a decisive role in plant electrical signaling (for Ca²⁺ review, see Martins et al., 2013; Brunet and Arendt, 2016; Zimmermann et al., 2016).

Figure 2.

Calcium wave and AP propagation go hand in hand. A, Model of the experimental setup used for parallel AP and Ca²⁺ measurements in the Venus flytrap. Ca²⁺ signal (indicated in red) originates at the stimulated TH and the propagation is monitored using calcium reporter GCaMP (cf. Suda et al., 2020) whose fluorescence is recorded with 50 frames per second. The AP propagation velocity is traced using two surface potential electrodes placed with a known distance at the trap surface and a reference electrode located in the soil. B, This cartoon illustrates simultaneous measurements of the Ca²⁺ wave (red) and the AP transmission (black) in the trap tissue. By bending a TH, both signals propagate from the point of origin (TH) to known positions (Δx) with measured time delays (Δt) resulting in the calculation of comparable velocities for signals of: Ca²⁺ velocity (2.06 ± 0.34 cm s⁻¹, n = 4) and AP velocity (1.91 ± 0.23 cm s⁻¹, n = 16, mean ± sd). C, AP propagation velocity is temperature-dependent. While at 10°C, the AP travels at a speed of only 1.3 cm s⁻¹, warming leads to an increase in velocity by a factor of 1.5 ± 0.2 (mean ± sd, n = 15) until at 40°C the AP reaches an average speed of 6.2 cm s⁻¹.

The co-existence and synergy of a Ca²⁺- and electrical wave not only result in prey recognition, but also subsequently lead to trap closure and finally to prey capture (Figure 1). With the calcium reporter-expressing Dionaea plants, Suda et al. (2020) were able to define calcium as a key component in trap closure, as the increase of the calcium concentration during mechanical stimulation is additive and has to reach a critical threshold to trigger trap closure. This threshold is only achieved when two APs are elicited with a frequency higher than 1/30 Hz (two stimuli within 30 s), which is consistent with the observation of two separate mechanical stimuli are required for trap movement (Darwin, 1875; Forterre et al., 2005; Scherzer et al., 2019; Volkov, 2019). In contrast, a new study (Burri et al., 2020) suggests that an AP can also be triggered by a single very slow mechanical stimulus, providing Dionaea with the ability to detect very slow prey as assumed by the authors. In this context, Brown and Sharp (1910) already observed the closing of the trap after just one stimulation, but only at an exceptionally high temperature.

How is the interaction of the electrical and Ca²⁺-signal converted into the fast closing of the traps and which components are involved?

Certainly, no energy-releasing metabolic processes accompany the fast trapping movement (Escalante-Perez et al., 2014). Instead, plant movement is usually achieved by turgor changes and thus water redistribution, which relies on osmotic and hydration/dehydration processes. However, the turgor changes by swelling and shrinking of cells and tissues are often limited in speed by the transport of water through the plant body. Therefore, it is predicted that snap-traps use a mechanical instability to achieve a much faster movement. This mechanism is based on pre-stored energy in the metastable state of the open trap, which makes it possible to maintain elastic energy until a given stimulus allows to overcome the barrier. This results in the sudden release of stored potential energy as kinetic energy (movement; Markin et al., 2008; Volkov et al., 2008; Forterre, 2013; Volkov et al., 2014; Morris and Blyth, 2019). Although good models exist describing the underlying physical principles including snap buckling and elastic instability, the detailed molecular mechanisms and the biological signaling cascades leading to a reallocation of energy and thus to rapid changes in turgor and/or cell wall elasticity remain to be demonstrated.

Ion fluxes during plant AP

As mentioned above, not only carnivorous plants show electrical signal transmission, but also nonexcitable plants, such as Arabidopsis. Even if the duration of the AP and its propagation speed differ, both signals seem to be based essentially on the same ion fluxes. In general, the plant membrane potential is established and maintained by PM-located proton pumps of the AHA-type (Reyer et al., 2020). These ATPases transport H⁺ under ATP consumption from the cytosol into the apoplast. Thus, they (1) create a proton gradient across the PM and (2) provide an excess of negative charge inside the cells. Besides the H⁺ gradient, the resting potential of the cells is further determined by potassium conductivity (Hedrich, 2012; Reyer et al., 2020). In addition, the cytosol of cells contains about 100 mM K⁺, 10–15 mM Cl⁻, and Ca²⁺ in the nanomolar range (Roberts, 2006). While the extracellular K⁺ concentration is extremely variable (0.05–100 mM), the apoplastic concentrations of Cl⁻ and Ca²⁺ fluctuate only in the low millimolar range, resulting in outward- and inward-directed electrochemical gradients for these two ions, respectively. Compared to animals, genomes of vascular plants do not encode voltage-dependent Na⁺ channels (Hedrich, 2012) and as a consequence, plant cells cannot depolarize the PM by opening sodium channels as known from animal APs (Raghavan et al., 2019). Instead, plants take advantage of inward Ca²⁺ and outward Cl⁻ gradients leading to a depolarization of up to 120 mV (Hedrich and Becker, 1994; Levchenko et al., 2005; Toyota et al., 2018; Dayal et al., 2019; Farmer et al., 2020; Figure 1). Similar to the animal system, the opening of voltage-dependent K⁺_out channels of the GORK-type will drive K⁺ efflux, starting to repolarize the plant cell in the post depolarization phase (Cuin et al., 2018; Hedrich and Neher, 2018). This repolarization will be continued by the voltage-dependent activation of the PM H⁺-ATPase (Reyer et al., 2020). Finally, initial ionic gradients will be restored by anion influx via NRT-type H⁺/anion symporters (Sun et al., 2014; Wang et al., 2018), while released K⁺ will be taken up again by HAK5-type H⁺/K⁺ symporters (Scherzer et al., 2015) in concert with voltage-gated K⁺_in channels (Iosip et al., 2020). In parallel, energy consuming transport proteins such as the P-type Ca²⁺-ATPase together with Ca²⁺/H⁺ exchangers of the CAX family reset the Ca²⁺ gradients and facilitate the establishment of resting conditions for cells to be ready for firing the next AP (Demidchik et al., 2018; Figure 1, upper left corner).

Adding up the electrical signals

In noncarnivorous plants, APs are usually triggered by stress factors (biotic or abiotic) and lead to an adaptation response. In this case, however, the response time is usually of secondary interest, since the stress factors do not require extremely rapid countermeasures (Wang et al., 2020a for review see Gupta et al., 2020; Wang et al., 2020b). As carnivorous plants seek to prevent the escape of captured prey, some of them have evolved responses on a time scale comparable to that of animals. Therefore, the network of electrical and chemical signaling pathways can be very well illustrated by considering the predominant processes that occur during the hunting cycle of Dionaea. If a prey is successfully caught, the fighting animal continues to touch the TH, eliciting numerous APs (Figure 1). Thereby a further slow, growth-related movement of the trap leaves is induced, which results in a hermetically sealed trap within the next 5–6 h (Pavlovic et al., 2010; Escalante-Perez et al., 2011; Scherzer et al., 2017). Unlike the Ca²⁺-threshold-dependent initial rapid closure, this slow trap-sealing phase is controlled by a chemical signaling cascade via the touch/wound hormone jasmonate (JA; Farmer et al., 2014; Pavlovic and Mithofer, 2019; Farmer et al., 2020). The process of hermetic trap sealing can even be initiated experimentally without prior mechanostimulation and fast trap closure, when coronatine—a mimic of the biologically active JA-derivate JA-isoleucine (JA-Ile)—is applied directly on open traps. In planta, the electrical signal of two APs typically marks the onset of the chemical signaling cascade to follow. Two APs activate JA signaling via the transcriptional regulation of the JA receptor Coronatine Insensitive 1 (COI1) and the coreceptor Jasmonate ZIM Domain 1 (JAZ1) (Bohm et al., 2016a; Monte et al., 2018, 2020), which is maintained by a further continuous stimulation by the prey in its attempt to escape the trap. Mimicking a series (more than three) of trigger-hair stimulations, the JA signaling pathway is induced, mediating the expression of genes encoding prey-degrading hydrolases in the secretory gland cells. Interestingly, the induction of chemical signals is proportional to the number of mechanical/electrical stimulations (Libiakova et al., 2014; Bemm et al., 2016; Bohm et al., 2016a; Pavlovic et al., 2017; Scherzer et al., 2017; Iosip et al., 2020). By coupling the electrical and chemical signals, the trap can fine-tune the amount of digestive enzymes secreted. A large and strong prey that struggles for a very long time and triggers many APs also needs a larger amount of enzymes to be fully digested (Box 3). When the prey finally stops triggering electrical APs the digestion process is under the tightly regulated control of the resulting chemical signal (Darwin, 1875; Libiakova et al., 2014).

Exocytosis floods the stomach with digestive enzymes

In the traps of Dionaea and Aldrovanda, the digestive process is fully inducible, whereas pitcher plants maintain permanent levels of digestive fluid and sticky-trapped plants produce already mucilage without a prey-induced mechanoelectrical signaling cascade (Ellison and Adamec, 2018). Research on these carnivorous plants of the Caryophyllales glade exhibited that a conserved JA signaling pathway induces enzyme secretion during prey capture (Bemm et al., 2016; Yilamujiang et al., 2016; Kocab et al., 2020; Hedrich and Fukushima, 2021). In contrast, a recent study on butterwort Pinguicula from the order Lamiales showed a JA-independent regulation of the onset and activity of digestive enzymes suggesting an order dependent role of the phytohormone JA during the digestion process (Kocab et al., 2020). Nevertheless, many digestive proteins found are known to play roles in plant defense responses of noncarnivorous plants against microbial pathogens, such as pathogenesis-related (PR) proteins (Buch et al., 2014). The findings that many proteases, and chitinases found in the pitcher fluids can also be classified as PR proteins led to the conclusion that the offensive carnivory mechanism evolved from an existing plant defensive mechanism against herbivory (Hedrich and Neher, 2018; Pavlovic and Mithofer, 2019).

Besides the induction of the JA signaling cascade after two APs, measurements with the Ca²⁺-sensitive dye FURA-2—loaded iontophoretically into a gland cell complex—also showed an increase in calcium in these secretory glands after the third AP (Escalante-Perez et al., 2011). With the translation of the electrical signal into a Ca²⁺ and JA signal (Figure 1), no defense reaction starts in these secretory gland cells, unlike in noncarnivorous plants, but the release of a decomposing digestive juice (Figure 3). This leads to a swelling of gland cells (Escalante-Perez et al., 2011) and the formation of secretory vesicles. The fusion of secretory vesicles with the PM could be prevented by Ca²⁺ channel inhibitors, which indicate that the Ca²⁺ induced fusion process is an early event in gland activation (Schulze et al., 2012; Paszota et al., 2014; Scherzer et al., 2017). The early vesicles loaded with H⁺ and Cl⁻ ions hyper-acidify the “green stomach”-like digestive organ, whereas subsequent ones carry digestive enzymes like chitinases, proteases, nucleases, phosphatases, and lipases (Bemm et al., 2016; Bohm et al., 2016a; Fukushima et al., 2017; Krausko et al., 2017; Scherzer et al., 2017; Saganova et al., 2018; Jaksova et al., 2020; Figure 3). This process could be demonstrated by using carbon fiber electrodes (for review, see Shin et al., 2019), thereby amperometric signals could be detected at the earliest about 6 h after mechanical or JA stimulation. This suggests that oxidizable compounds are present only in the late vesicles and are most likely peptides/proteins with function in digestion. Compared to animal secretion, Dionaea has a very slow time scale, which indicates a high diffusion resistance of the glandular cell wall, which is not present in the animal system. In contrast to fast synaptic signals in the animal nervous system, this slow time scale of release reflects the biology of the insect processing flytrap: as soon as Dionaea catches a prey via its fast haptoelectric sensor system, the exocytotic release, and slow diffusion of the digestive fluid perfectly serves the long-term nutrient requirements of the plant.

Figure 3.

Gland cell activation by the JA and Ca²⁺ signaling pathways to enable prey digestion and nutrient uptake. The expression of transport proteins to acidify the formed stomach after trap closure as well as the expression of digestive enzymes, like hydrolases etc., are induced via the JA signaling pathway (blue). Although, the associated transport proteins are not known in detail, one could suggest the involvement of H⁺ ATPases and chloride selective channels, which lead to an efflux of HCl into the stomach (left gland). Together with early exocytotic vesicles loaded with proton and chloride ions (green and yellow), the stomach is acidified in order to enable digestive enzyme activity and prey decomposition. The digestive fluid, including a diversity of digestive enzymes, is secreted into the trap and the prey is degraded. Thereby, the fusion process of exocytotic vesicles is triggered by the second messenger calcium (red). The released nutrients (e.g. phosphate, nitrogen species, and potassium) as well as organic compounds and prey derived building blocks are taken up by the gland cells via channel-based transport processes or endocytosis (right gland). DmAMT1 displays an ammonium channel and therefore transports the nitrogen source into the cytosol of the gland cells. The macronutrient potassium is taken up with high capacity and low affinity by the K⁺ selective DmKT1 channel and the proton-driven, high-affine transporter DmHAK5. Both modules enable a consistent potassium uptake and are activated via the Ca²⁺-dependent sensor/kinase complex CBL/CIPK. The absorption of prey-derived sodium ions is mediated by the DmHKT1 channel and is further sequestered into the trap vacuole to prevent the toxification of photosynthetically active tissue. The nutrient uptake by transport proteins is furthermore supplemented by endocytosis (gray) that facilitates the absorption of whole proteins as well as the re-absorption of the secreted digestive enzymes. Endocytoses enables the Venus flytrap to finalize the uptake process and to prevent a loss of compounds during the following re-opening. ATP: adenosine triphosphate; expr: expression; CBL: Calcineurin B-like protein; CIPK: CBL-interacting protein kinase.

Nutrient uptake

Membrane transport processes play an important role in the information transmission during the hunting cycle of the Venus flytrap. Not only ion fluxes during mechanical prey detection and electrical signal transmission or the compound release from secretory vesicles during digestion depend on membrane transport processes, but also the absorption of nutrients from the prey (Figure 3). The cost/benefit models for carnivores in general describe that the special properties and mechanisms required for carnivory are associated with high costs (Box 1; Givnish et al., 1984; Ellison and Gotelli, 2009; Ellison and Adamec, 2011; Givnish, 2015). For the organism to benefit from carnivory, these costs have to be balanced by the gain of nutrients from the prey, and the micro- and macronutrients (including nitrogen species, phosphate, potassium) obtained must be relocated to tissues requiring them. This is supported by transcriptomic analysis of the Dionaea traps that dramatically shifted toward signal transduction and nutrient transport upon insect feeding (Bemm et al., 2016). When comparing the transport components in digestive traps of Dionaea with the root-expressed orthologous proteins from noncarnivorous plants, a large intersection of channels, transporters, and regulatory components is found, which performs similar functions in the glands of Dionaea as they do in the roots of noncarnivorous plants (Bemm et al., 2016). This indicates that carnivorous plants in their evolution have not acquired completely new genes, but neo-functionalized existing genes for use in prey capture and nutrient absorption (Box 1; Albert et al., 1992; Ibarra-Laclette et al., 2013; Butts et al., 2016; Palfalvi et al., 2020; Silva et al., 2020; ZulkapLi et al., 2021). Transport proteins and regulatory components for the nitrogen source ammonium and the macronutrient potassium—that are transcriptionally induced after more than five mechanoelectrical stimuli—could be identified in glands of Venus flytraps (Scherzer et al., 2013, 2015). The acidic protease mix of the secretion leads to the release of ammonium from the organic compound glutamine. The resulting cation is absorbed into the glands via the ortholog of the Arabidopsis root hair-located ammonium transporter, DmAMT1, and thus is available to the plant as a macronutrient (Scherzer et al., 2013; Figure 3).

Similar parallels can also be found in the transport proteins for potassium. Dionaea, for example, uses AKT1 and HAK5-type potassium uptake systems in traps, like Arabidopsis in roots (Scherzer et al., 2015). Potassium is, in addition to the cation sodium, largely represented in the Dionaea prey and both cations are released during digestion (Punzo, 1990; Ajai et al., 2013; Finke, 2013). Interestingly, the orthologs, DmKT1 and DmHAK5, are localized in gland cells of the Venus flytrap, which act during nutrient absorption from the prey. Besides transcriptional regulation, many transport proteins are also subject to post-translational regulation. In planta studies showed that the protein kinase CIPK23, which interacts with calcineurin B-like (CBL) protein, is an important regulatory component of potassium uptake in roots of noncarnivorous plants (Hedrich and Kudla, 2006; Xu et al., 2006). In fact, the Arabidopsis CBL/CIPK complex was able to activate not only DmKT1 but also DmHAK5 from the Venus flytrap, confirming the cross-species conservation of signaling complexes and thus illustrating the importance of such regulatory mechanisms (Imes et al., 2013; Zipfel and Oldroyd, 2017; Tang et al., 2020). Thus, this Ca²⁺-sensor/kinase complex can be identified as the key regulatory complex for potassium uptake in Dionaea. DmKT1 acts with a high transport rate and a low affinity for potassium, whereas the highly potassium-affinitive DmHAK5 mediates potassium uptake down to very low concentrations after prey digestion. Therefore, DmHAK5 utilizes the proton gradient derived from the acidic secretion fluid for its transport purposes (Scherzer et al., 2015). The glands of the Venus flytrap thus exhibit a biphasic system for the high- and low-affinity potassium absorption induced by prey capture, which can be controlled in detail at the transcriptional level—induced by the frequency of electrical signals—and at the post-translational level via a Ca²⁺ signaling cascade (Figure 3).

The fact that all the soluble components of the prey must be completely absorbed in order to avoid osmotic problems within the traps means that, in addition to the uptake of all essential nutrients, stress-related and toxic ions must also be absorbed. As the animal prey of Dionaea has a high Na⁺ load, this must also be ingested. Traditionally, plants protect themselves from damage by depositing the toxic substances in the vacuoles. Since the traps are less photosynthetically active, the sodium could cause less damage there. Furthermore, in the trap parenchyma cells, sodium could be used as an osmotically active substance for maintaining the cell turgor. Possibly, this sodium content, which increases by each prey catch, terminates the trapping capacity of an individual trap. This sodium uptake is mediated by a transport protein from the HKT1-family that provides a potent mechanism for mediating salt tolerance in Arabidopsis and crop plants (Hamamoto et al., 2015; Borjigin et al., 2020; Houston et al., 2020; Huang et al., 2020). This stress-related gene is functionally and locally converted in Dionaea gland cells to constantly guarantee the prey-derived sodium uptake and to meet the carnivorous challenges (Bohm et al., 2016a, 2016b).

Furthermore, analysis with fluorescent endocytosis markers revealed, that in some carnivorous plants, including Nepenthes, Drosera, Dionaea, Aldrovanda, and Cephalotus, endocytosis could be involved in uptake processes in gland cells. After mimicking prey-digestion within a few hours, numerous endocytotic vesicles were formed in the cytoplasm (Adlassnig et al., 2012). On one hand the ionic nutrient uptake by transport proteins could be supplemented by the endocytosis process, especially for larger building blocks like peptides, and on the other hand endocytosis is a way to re-absorb any secreted fluid, including enzymes and thus remain available as part of the carnivorous physiology (Figure 3).

Trap reopening

After the digestion and resorption phase, the reopening of the Dionaea traps shows that all substances and ions of the prey are absorbed (Darwin, 1875). The only residues are parts of the animal exoskeleton that can be blown away by the next gust of wind. In contrast to the rapid trap closure, very little is known about the opening process. It takes about half a day and is thus significantly slower than the closing process (Volkov et al., 2014). Stuhlman (1948) observed the slow trap opening and suspected growth processes from its cause. Through the later analysis of time-lapse videos, Fagerberg and Howe (1996) divided the reopening into several different phases. One reason for the slow time course of this movement may be that the potential energy for the next trap closure, similar to a spring, has to be stored first. This could be illustrated by means of several mathematical models that describe morphing processes of the trap in detail (Yang et al., 2010; Li and Zhang, 2011; Markin and Volkov, 2012; Sachse et al., 2020). Even if the Venus flytrap already has an application as a “soft machine” in modern robotics (Esser et al., 2020; Li et al., 2021), the responsible “motor cells” that mediate this process are still unknown in its plant prototype. The elucidation of the slow opening movement will further enrich and complete our understanding of the carnivorous cycle of the fascinating Venus flytrap.

Concluding remarks

As discussed above, recent studies have substantially enriched the field of carnivorous membrane transport and signaling processes. Although these new data certainly help to better understand the fascinating but complex world of plant signaling, many areas on the scientific map are still unexplored (see Outstanding Questions).

We await with excitement and curiosity new findings and future work, which will add additional pieces to the puzzle elucidating the secret mechanisms underlying plant carnivory.

OUTSTANDING QUESTIONS

• What is the molecular nature of the fast closure?

• What is the mechanism underlying trap reopening?

• How does the plant terminate its digestion in terms of the secretion of lytic enzymes and the absorption of nutrients?

• Why and when do old traps die?

• Which channels exactly control the plant action potential?

• What is the exact physiological mode of action of anaesthetics in plants and which molecular components are affected?

 

Both authors participated in writing the article.

The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (https://academic.oup.com/plphys/pages/general-instructions) is: Sönke Scherzer (soenke.scherzer@uni-wuerzburg.de).