The emerging role of GABA as a transport regulator and physiological signal

Abstract

While the proposal that γ-aminobutyric acid (GABA) acts a signal in plants is decades old, a signaling mode of action for plant GABA has been unveiled only relatively recently. Here, we review the recent research that demonstrates how GABA regulates anion transport through aluminum-activated malate transporters (ALMTs) and speculation that GABA also targets other proteins. The ALMT family of anion channels modulates multiple physiological processes in plants, with many members still to be characterized, opening up the possibility that GABA has broad regulatory roles in plants. We focus on the role of GABA in regulating pollen tube growth and stomatal pore aperture, and we speculate on its role in long-distance signaling and how it might be involved in cross talk with hormonal signals. We show that in barley (Hordeum vulgare), guard cell opening is regulated by GABA, as it is in Arabidopsis (Arabidopsis thaliana), to regulate water use efficiency, which impacts drought tolerance. We also discuss the links between glutamate and GABA in generating signals in plants, particularly related to pollen tube growth, wounding, and long-distance electrical signaling, and explore potential interactions of GABA signals with hormones, such as abscisic acid, jasmonic acid, and ethylene. We conclude by postulating that GABA encodes a signal that links plant primary metabolism to physiological status to fine tune plant responses to the environment.

γ-Aminobutyric acid (GABA) encodes a plant signal that links primary metabolism to physiological status to fine tune plant responses to the environment.

Introduction

The nonproteinogenic amino acid γ-aminobutyric acid (GABA) has been proposed to be an agent of cellular communication that emerged very early in evolution, being conserved across modern animals and plants (Shelp et al., 2006; Ben-Ari et al., 2007; Žárský, 2015; Ramesh et al., 2017). Cellular GABA metabolism (synthesis and catabolism) predominantly occurs via the GABA shunt pathway and is enacted by orthologous key enzymes in both kingdoms (Bouché et al., 2003; Bown and Shelp, 2016). GABA is primarily synthesized from glutamate by glutamate decarboxylase (GAD) in the cytosol, and is degraded by GABA transaminase (GABA-T) into succinic semialdehyde (SSA) in mitochondria, bypassing two stress-inhibited reactions of the mitochondrial-based tricarboxylic acid (TCA) cycle (Bouché et al., 2003; Bown and Shelp, 2016). Polyamine-derived GABA synthesis can also have a significant impact on plant function under certain scenarios (Zarei et al., 2016). GABA synthesis in plants is stimulated by stress, and its known or proposed roles—as a metabolite in plants—were traditionally thought to be confined to processes such as pH regulation, redox status, and carbon–nitrogen balance (Shelp et al., 1999; Batushansky et al., 2014; Bor and Turkan, 2019).

In mammals, GABA acts as a signal via receptor-mediated membrane hyperpolarization of neuronal cells (Owens and Kriegstein, 2002; Žárský, 2015). In plants, alongside its established roles as a metabolic bypass to sustain cellular energy production, GABA was proposed to fulfil a signaling role decades ago when it was found that GABA modulated the growth of pollen tubes, and disruption of GABA catabolism impacted fertilization (Palanivelu et al., 2003). Subsequent studies have shown that modulating GABA metabolism, or applying GABA as a supplement, impacts plant physiological response, for example, water use, drought tolerance, salt and osmotic responses (Krishnan et al., 2013; Li et al., 2016b, 2016c; Farooq et al., 2017; Cheng et al., 2018; Abdel Razik et al., 2020). Nevertheless, experimental evidence providing a mode of action for GABA, including the existence of a plant GABA receptor or elements in a GABA signaling pathway had not been obtained (Palanivelu et al., 2003; Bouche and Fromm, 2004; Michaeli and Fromm, 2015).

In 2015, the discovery that anion flux through plant-specific aluminum-activated malate transporters (ALMTs) was negatively regulated by GABA, altering plasma membrane potential and resulting in a downstream physiological response, represented a plausible mechanism by which GABA may act as a signal in plants (Ramesh et al., 2015; Gilliham and Tyerman, 2016). Moreover, it was subsequently shown that: GABA is likely to act directly upon ALMTs (Domingos et al., 2019; Long et al., 2019); ALMTs also facilitate the transport of GABA (Ramesh et al., 2018; Kamran et al., 2020); and, GABA responses of ALMTs are dependent upon the presence of specific amino acid residues within ALMTs (ALMTs share no homology to mammalian Cys-loop [GABA] receptors except a region of 12 amino acid residues predicted to bind GABA in GABA_A receptors; Ramesh et al., 2015, 2017, 2018; Long et al., 2019; Xu et al., 2021). Furthermore, revealing a mechanism by which GABA acts via ALMTs in guard cells—to regulate drought tolerance—(Xu et al., 2021), and that ALMTs are involved in GABA responses in pollen (Domingos et al., 2019), is the strongest evidence yet that GABA is an endogenous plant signal that links primary metabolism to physiological responses (Gilliham and Tyerman, 2016). Additional GABA targets have also been nominated in plants, including 14-3-3 proteins, H⁺-ATPases, and potassium channels (Lancien and Roberts, 2006; Ramesh et al., 2017; Su et al., 2019; Adem et al., 2020), which have the potential to form part of the GABA signaling network. Here, we elaborate upon the work cited above, and further recent outputs, to outline the case for GABA being a credible signaling molecule in plants, and the ways in which it may interact with other known signals to modulate plant physiology.

GABA as a stomatal guard cell signal regulating plant water loss

Stomatal guard cells delineate the stomatal pores on plant aerial surfaces and respond to environmental signals by regulating the stomatal pore aperture to modulate plant water loss and carbon assimilation (Hetherington and Woodward, 2003; Kim et al., 2010; Murata et al., 2015; Xu et al., 2021). It has been shown numerous times that water loss was minimized from a variety of plants when GABA was applied as a treatment (Krishnan et al., 2013; Li et al., 2016a; Farooq et al., 2017; Abdel Razik et al., 2020). GAD1 and GAD2 are the major GAD isoforms in roots and leaves of Arabidopsis (Arabidopsis thaliana), respectively, and their knockout leads to negligible GABA concentrations in tissues (Mekonnen et al., 2016); further, it was proposed that depletion of GABA concentration in gad1/gad2 leaves led to plants that were more drought prone (Mekonnen et al., 2016).

The greater stomatal conductance and drought sensitivity of gad1/gad2 mutants were initially attributed to their more open stomatal phenotype and greater stomatal density (Mekonnen et al., 2016). The minor developmental phenotype of gad1/gad2 is likely due to the smaller leaves of the line tested compared to the wild-type, as other GAD mutants do not share this feature (Xu et al., 2021), so it is unlikely that GABA plays a significant role in stomatal development. The greater stomatal aperture of gad1/gad2 was proposed to be due to H⁺-ATPases mediating a greater proton (H⁺) efflux across the plasma membrane leading to greater pore opening and inhibition of stomatal closure; this was inferred after it was observed that gad1/gad2 roots had a greater acidification capacity of the surrounding media (Mekonnen et al., 2016). Interestingly, when direct microelectrode-based measurements of gad1/gad2 roots were made, their H⁺ efflux capacity was diminished compared to wild-type plants but was increased in GABA overaccumulating mutants, and that the membrane potential was relatively depolarized in gad1/gad2 roots when exposed to 100-mM NaCl (Su et al., 2019). It was suggested that GABA inhibited NaCl stimulated K⁺-efflux from roots and that this was associated to a greater ability to quench reactive oxygen species (ROS), whereas gad1/gad2 had greater K⁺-efflux, which was proposed to occur via guard cell outwardly rectifying K⁺ channel (GORK; Su et al., 2019). GABA has previously been implicated in activating transcription of 14-3-3 proteins, which are known activators of H⁺-ATPases and GORK (Alsterfjord et al., 2004; Lancien and Roberts, 2006; Van Kleeff et al., 2018). Furthermore, GORK has been shown to be activated by ROS (Demidchik et al., 2010; Wang et al., 2017), and GABA has been implicated in ROS detoxification through an unidentified mechanism (Wu et al., 2021). In a further study, 10-mM GABA was shown to activate K⁺-efflux from roots in a GORK-dependent manner, and it was argued that GORK shared the same putative GABA-sensitive motif found in ALMTs (Adem et al., 2020; Wu et al., 2021). These seemingly contradictory observations raise several questions including: Is there a dose dependent effect of GABA on H⁺-efflux and K⁺-efflux from roots?; Is the impact of GABA on ion fluxes different in roots and guard cells?; How does GABA regulate H⁺-ATPases?; Is GORK directly regulated by GABA or indirectly via GABA’s impact on ROS, or another route?; and, Are there alternative explanations for the stomatal and root phenotypes of GABA-depleted mutants?

In Ramesh et al. (2015), it was proposed that negative regulation of anion efflux via ALMT would indirectly reduce the activity of the plasma membrane H⁺-ATPase. ALMT activity is a prime candidate for contributing to the short circuit (equal and opposite charge exchange) that maintains H⁺-ATPase activity by preventing it stalling at extremely hyperpolarized membrane potentials. This hypothesis is compatible with the above observations of altered membrane potential, and is a possible explanation for the inconsistencies observed for K⁺ and H⁺-fluxes between studies if the H⁺-ATPase and K⁺ channels are not direct targets of GABA. The hypothesis that GABA regulation of ALMT constituted a physiological signal was furthered in Xu et al. (2021) using the stomatal guard cell as an experimental system.

Similar to gad1/gad2 mutants, gad2 mutants exhibited high stomatal conductance and drought sensitivity; however, gad2 mutants do not share the developmental differences of gad1/gad2 when compared to wild-type plants, for example, smaller rosettes or higher stomatal densities (Mekonnen et al., 2016; Xu et al., 2021). Furthermore, the high stomatal conductance and drought sensitivity of gad2 plants were complemented by the additional loss of ALMT9 (Xu et al., 2021). ALMT9 is a tonoplast localized anion transporter that catalyzes malate and chloride (Cl^–) uptake across the vacuolar membrane of the guard-cell to contribute to the osmotic increase that is required for stomatal opening (Kovermann et al., 2007; De Angeli et al., 2013). The loss of ALMT9 impairs light-induced stomatal opening and led to almt9 mutants being more drought tolerant (De Angeli et al., 2013); ablation of ALMT9 also abolished the ability of GABA to inhibit stomatal opening, which was restored by native ALMT9 complementation (Xu et al., 2021). This signifies that GABA inhibits stomatal opening via acting on ALMT9 (Figure 1A). Attempted complementation of almt9 plants with ALMT9^F243C/Y245C (containing mutations within the putative GABA interacting motif first characterized in the wheat [Triticum aestivum] TaALMT1; Ramesh et al., 2015, 2017; Long et al., 2019) failed to restore the sensitivity of stomatal opening to GABA, but instead phenocopied the higher stomatal conductance of the gad2 mutant (Xu et al., 2021). These data are consistent with ALMT9 being the predominant “GABA receptor” in guard cells, and when GABA synthesis is inhibited, ALMT9 is deregulated resulting in increased opening and pore aperture, and an increase in drought sensitivity of the plant (Figure 1A;Xu et al., 2021). An assay is now required to demonstrate whether GABA has a direct effect on ALMT9, as has been demonstrated for TaALMT1 via electrophysiology (Long et al., 2019), or whether GABA acts on ALMT9 via a distinct mechanism. What is not in doubt though is that ALMT9 (and the putative GABA binding domain) is required for the response to GABA, even if other signaling elements are involved. It is noted that a range of candidates for interaction with GABA were nominated by Ramesh et al. (2017) based on the presence of a putative GABA binding site within a number of plant proteins, with none of these other candidates yet being examined in planta.

Figure 1.

Proposed model of GABA-improved WUE in plants. A, A proposed model of GABA supplementation reducing stomatal opening in dicots A. thaliana, adapted from (Xu et al., 2021). Increases in cellular GABA have an inhibitory effect on anion uptake through tonoplast ALMT9 into guard-cell vacuoles, reducing stomatal opening and water loss through stomatal pores (A). B and C, GABA supplementation reduces stomatal opening (B) and increases transient iWUE (C) of detached barley leaves. B, Leaves detached from 2- to 3-week-old H. vulgare L. Barke seedlings were fed through the xylem sap, and gas exchange was recorded using a LI-COR LI-6400XT Portable Photosynthesis System following protocols as described in Xu et al. (2021); GABA was applied into xylem sap solution to a final concentration of 8 mM as indicated by an arrow, allowing for a 30 min pretreatment for uptake of GABA through leaf petiole, followed by 30 min dark (shaded region) to close stomata and 30 min light (white region, 1,000 µmol m⁻² s⁻¹) to reopen stomata. C, iWUE of detached leaves was calculated as the ratio of photosynthetic rate (Supplemental Figure S1) versus stomatal conductance within a range that showed significant difference between control and GABA treatments (Leakey et al., 2019). Data represent means ± sd, n = 3; statistical analysis was determined by two-sided Student’s t test, ^*P < 0.05 (B and C). D, A proposed model of GABA-enhanced WUE in monocot H. vulgare. Similar to Arabidopsis (Xu et al., 2021), GABA supplement did not affect stomatal closing, but reduced the extent of opening of barley leaves; GABA’s proposed mode of action in barley guard cells based on observation from (B) and GABA’s function in Arabidopsis guard cells from (A); GABA may be associated with negative regulation of anion uptake through unidentified anion channels, perhaps, for example, tonoplast-localized HvALMT(s) in guard cells to reduce opening extent of stomatal pores; however, it is unknown whether GABA acts in subsidiary cells in this regulation, as stomatal opening of barley plants is modulated by ionic influx into guard cells and efflux from subsidiary cells (Chen et al., 2017).

There were a number of other significant observations in regard to the nature of GABA as a signal stemming from Xu et al. (2021). First, overproduction of GABA in wild-type plants improved water use efficiency (WUE) and led to an improvement in drought resilience (Xu et al., 2021). This suggests that GABA metabolism can be manipulated to improve stress tolerance in plants over and above wild-type levels. Second, Xu et al. (2021) showed supplementation to epidermal peels of GABA or muscimol (a GABA analog) suppressed stomatal movement in response to multiple opening (e.g. light and coronatine; Melotto et al., 2006; Shimazaki et al., 2007; Sussmilch et al., 2019) or closing signals (e.g. dark, low-dose abscisic acid [ABA], and H₂O₂; Shimazaki et al., 2007; Sussmilch et al., 2019). This differentiates it from many of the more well-defined guard-cell signals, such as ABA, hydrogen peroxide (H₂O₂), and calcium (Ca²⁺; Kim et al., 2010; Murata et al., 2015), as GABA itself does not stimulate stomatal movement when its treatment falls within the physiological range (Xu et al., 2021; i.e. under nonstressed and stressed conditions, e.g. ∼1 µmol g⁻¹ fresh weight [FW] and ∼ 2 µmol g⁻¹ FW, equivalent to 1–2 mM respectively; Ramesh et al., 2015, 2018; Deng et al., 2020; Xu et al., 2021). When GABA was fed to leaves through the petiole to corroborate the findings in epidermal peels it was found that GABA only impacted stomatal opening, not closure (Xu et al., 2021), indicating the loss of the mesophyll in epidermal peels impairs the ability to reproduce the standard physiological response of intact plants (Lee and Bowling, 1992; Lawson et al., 2008, 2014). This finding also suggests that under the conditions tested GABA neither regulate stomatal closure, nor activate GORK. The physiological conditions where GABA impacts closure in planta are yet to be determined. However, it was found that GABA was unable to inhibit, in epidermal peels, closure of knockout mutants of ALMT12 (otherwise known as QUAC1—Quick-activating Anion Channel 1; Xu et al., 2021) and so this is likely to represent a mechanism by which GABA could inhibit guard cell closure.

The high stomatal conductance, low WUE and drought sensitivity of the gad2 mutant could be complemented to wild-type levels by guard cell specific expression of GAD2Δ (a constantly active form of GAD2 with truncation of a Ca²⁺/Calmodulin [Ca²⁺/CaM] binding domain), but not by mesophyll-cell complementation of GAD2Δ (Turano and Fang, 1998; Zik et al., 1998; Akama and Takaiwa, 2007; Xu et al., 2021). This suggests, on first examination, that the generation of GABA within the guard cell cytosol is sufficient to constitute a signal, and that mesophyll GABA accumulation does not overtly contribute to stomatal regulation. However, full-length GAD2 complementation driven by a constitutive 35S promoter recovered the higher stomatal conductance of gad2 to wild-type levels under normal conditions, whereas gain of GAD2 in guard cells only complemented gad2 under water-deficit stress (Xu et al., 2021). This indicates the importance of posttranslational control in shaping GABA signals, and that different cell types are likely to contribute to the nature of the signal under different conditions.

GABA synthesis is stimulated by acidification of the cytosolic pH and Ca²⁺/CaM-dependent activation of GAD (Carroll et al., 1994; Crawford et al., 1994; Turano and Fang, 1998; Zik et al., 1998). GABA breakdown is catalyzed by GABA-T in mitochondria (Clark et al., 2009). Both synthesis and degradation elements (GADs and GABA-T, respectively) have distinct expression patterns in plants (Clark et al., 2009; Renault et al., 2010; Scholz et al., 2015). It is possible, therefore, that GABA metabolomic levels may be differentially controlled in different cell types. Intracellular pH and Ca²⁺ signals, the key regulators of GAD-catalyzed GABA synthesis (Zik et al., 1998) are known to be spatially and temporally regulated in response to the environment (Behera et al., 2018; Li et al., 2021). It can therefore be expected that cellular GABA signals are dynamically shaped in plant tissue, and this will need to be investigated with the application of technologies such as intensity-based GABA sensing fluorescence reporters (e.g. iGABASnFR) in planta (Marvin et al., 2019; Fromm, 2020).

GABA’s impact on stomatal pore movement occurs across a range of crop plants and relatives, including broad bean (Vicia faba), soybean (Glycine max), Nicotiana benthamiana, and barley (Hordeum vulgare; Xu et al., 2021). Here, we show GABA supplementation of detached, but intact, barley leaves reduces stomatal opening and does not affect stomatal closure (Figure 1B). The mechanism behind this is unknown. The only well-characterized barley ALMT, HvALMT1, is localized at plasma membrane and is expressed in roots and guard cells (Gruber et al., 2010). While its anion transport capacity was inhibited by GABA (Ramesh et al. 2015), its overexpression resulted in greater closure with no opening phenotype (Gruber et al., 2010), and RNAi knockdown resulted in diminished closure in the dark and greater water loss (Xu et al., 2015)—opposite to that observed for almt9 (Xu et al., 2021) and for GABA treated barley (Figure1B). Here, we show that GABA improves intrinsic WUE (iWUE) of barley (Figure 1C). Therefore, it is not likely that HvALMT1 is the target that leads to GABA reducing stomatal opening (Gruber et al., 2010; Ramesh et al., 2015; Xu et al., 2015). As the barley gas exchange response to GABA parallels the phenotype of Arabidopsis, it is entirely possible that GABA’s effects are actioned through inhibition of barley tonoplast ALMTs that have a role in opening pores. While a simple bioinformatic search can reveal the barley ALMTs, without functional characterization it would not be possible to identify the correct target or indeed those present on the tonoplast (David et al., 2019). Additional questions are relevant here, as barley—a cereal monocot crop—has a different stomatal morphology. In barley, the stomatal complex is formed by dumbbell-shaped guard cells flanked by subsidiary cells; it is possible then that GABA signals may regulate ionic flux across both barley guard- and/or subsidiary cell membranes (Figure 1D;Merilo et al., 2014; Chen et al., 2017), and this would need to be tested. Nevertheless, it appears that the cell-type manipulation of GABA metabolism has potential to improve the drought tolerance of crop plants, considering the responsiveness to GABA in crops, and that overproduction of GABA leads to improved WUE in Arabidopsis (Xu et al., 2021).

GABA regulation of pollen tube growth

Pollen–pistil interactions guide pollen tube apical growth during sexual production in flowering plants (Higashiyama and Takeuchi, 2015), which involves the regulation of complex signaling and ionic flux networks as extensively reviewed previously (Lamport et al., 2018; Johnson et al., 2019). GABA has been proposed to form a gradient within pistil to accelerate the pollen tube growth toward the ovule as part of growth guidance (Palanivelu et al., 2003; Yu et al., 2014). Palanivelu et al. (2003) gave an initial indication that GABA may have a signaling role in plants. However, conclusive genetic proof that GABA signaling modulates plant fertility, and by which mechanism, is yet to be revealed despite recent evidence that implicate ALMTs have a role in the responses of pollen tubes to GABA (Domingos et al., 2019). Further, the existence of the pistil GABA gradient would benefit from confirmation, or otherwise, via other techniques, as well as further investigation into the molecular mechanism of its establishment and the response of pollen to GABA. Here, we concentrate on the evidence for GABA having a role in regulating pollen tube growth and fertility (Figure 2).

Figure 2.

Proposed model of GABA action on pollen tube growth. Ca²⁺, H⁺, and Cl⁻ form gradients within the pollen tubes to facilitate its growth to ovules. Glutamate-dependent Ca²⁺ channels (GLR1.2/3.7) and cyclic nucleotide gated channels (CNGC16/18) conduct apical Ca²⁺ influx and form a Ca²⁺ gradient; this triggers calcium-dependent protein kinases (CPK2/6/20)-dependent activation of anion efflux (Cl⁻ and malate [mal²⁻]) via ALMT12-14 (Gutermuth et al., 2018); and H⁺-ATPases (AHA6/8/9) to establish a pH gradient across the pollen tube plasma membrane (Chen et al., 2020; Hoffmann et al., 2020). We propose that cytosolic GABA (together with Ca²⁺/CPKs) fine tune apical anion efflux via ALMT12-14 and may be associated with H⁺ gradient and membrane potential regulation within pollen tubes via its actions on the H⁺-ATPase. Such cytosolic GABA homeostasis may be shaped via uptake by unidentified GABA transporter(s) (and/or ALMT) from the apoplast (pistil; Ramesh et al., 2018) and via degradation by GABA-T into SSA in mitochondria (Palanivelu et al., 2003). All these elements work together to facilitate pollen tube growth to ovules during sexual production. Adapted from Domingos et al. (2019) combined with information from Gutermuth et al. (2018); Ramesh et al. (2018); Domingos et al. (2019); Hoffmann et al. (2020).

In vitro assays have shown that micromolar GABA stimulates pollen tube elongation of Arabidopsis and tobacco (Nicotiana tabacum), while >10 mM suppresses pollen tube growth (Palanivelu et al., 2003; Yu et al., 2014). The supplementation of low concentrations of muscimol (the potent GABA analogue) mimic the high dose effects of GABA, that is, it produces an inhibitory effect on pollen tube growth of Arabidopsis and grapevine (Vitis vinifera), which can be attenuated by bicuculline, a competitive GABA receptor antagonist (Ramesh et al., 2015, 2017). Domingos et al. (2019) determined that both GABA and muscimol significantly reduce anionic currents from Arabidopsis pollen protoplasts by up to 90% when supplied from the cytoplasmic side, but not the extracellular face, and that this effect of muscimol was abolished in almt12 mutants (Domingos et al., 2019). The inhibitory effect of externally supplied muscimol on the growth of pollen tubes was also abolished in almt12; further, the length of almt12 pollen tubes was similar to the length of those from wild-type plants supplied with muscimol (Domingos et al., 2019). Interestingly, anion efflux at the tip of almt12 mutants was not decreased, and ALMT12 has not yet been localized to the plasma membrane (Gutermuth et al., 2018; Domingos et al., 2019). However, GABA also inhibits ALMT13- and 14-mediated malate efflux from Xenopus laevis oocytes (Ramesh et al., 2015), as was shown for ALMT12 induced currents from COS-7 cells (Domingos et al., 2019), and ALMT13 and 14 have been localized to the pollen tube plasma membrane (Gutermuth et al., 2018). ALMT12-14 catalyze anionic (Cl⁻ and malate) efflux from pollen protoplasts, which play an important role in establishing anionic gradients formed within pollen tubes, which are proposed to have a role in navigating their growth (Gutermuth et al., 2018). Other factors reported to regulate ALMT12 and ALMT14 anion transport capacity in pollen are cytosolic Ca²⁺ and the Ca²⁺-dependent kinase CPK6 (and other CPK; Gutermuth et al., 2018). Oscillations in tip cytosolic Ca²⁺ concentration, anion efflux and growth are synchronous (Gutermuth et al., 2018). GAD activity is also stimulated by cytosolic Ca²⁺ concentration. Thus, the opposing effects of Ca²⁺ and GABA on ALMT12-14 activity may contribute to the oscillations in pollen tube tip anion efflux and growth modulation (Figure 2;Ramesh et al., 2015; Domingos et al., 2019). This said, the reproductive phenotypes of either almt12/almt13 or almt12/almt14 are minor in terms of seed set and pollen tube growth regulation in the conditions tested (Gutermuth et al., 2018; Domingos et al., 2019). This is similar to the impact of ALMT12 on stomatal pore control by GABA—impacts were detected in vitro but these did not extrapolate to intact plant phenotypes. When GABA breakdown is impaired in the Arabidopsis gaba-t/pop2 mutant, aberrant pollen tube growth to ovules and lower seed set than the wild-type was recorded. Collectively, this suggests that GABA metabolism could impact reproductive processes through other proteins in addition to ALMTs. One possibility is GABA modulation of the plasma membrane proton (H⁺) ATPases (i.e. AHA6/8/9), which are essential for pollen tube elongation and reproductive production (Figure 2;Su et al., 2019; Hoffmann et al., 2020). In root tissue, GABA accumulation has been positively correlated with the magnitude of H⁺ efflux under salt by contrasting the Arabidopsis (GABA-overaccumulation) gaba-t/pop2 mutant that exhibits net H⁺ release and the (GABA-deficiency) gad1/2 mutant, which exhibits net H⁺ uptake (Su et al., 2019). As such, GABA may also modulate H⁺-ATPase catalyzed H⁺ flux either directly (or through GABA activated 14-3-3 proteins?; Alsterfjord et al., 2004; Lancien and Roberts, 2006), or indirectly through regulation of ALMT (Ramesh et al., 2015, 2018), which will impact the pollen tube growth (Figure 2). To our knowledge, the impact of GABA on H⁺ fluxes of pollen tubes has not been examined.

We propose that GABA fluxes across pollen tubes may also be an important regulatory factor in their growth. This stems from the observation that crossing pop2 mutant pollen onto wild-type pistils, or wild-type pollen onto pop2 mutant pistils, obtained wild-type-like self-crossed fertilization levels of seed set (Palanivelu et al., 2003), and the following: (1) GABA signals are likely to act from the cytosolic side (Long et al., 2019; Xu et al., 2021); (2) TaALMT1 can faciliate GABA transport across the plasma membrane (Ramesh et al., 2018); (3) extracellular (i.e. pistil) GABA fine tunes pollen tube growth (Palanivelu et al., 2003; Yu et al., 2014; Ramesh et al., 2015); and (4) GABA-T-catalyzed GABA breakdown is essential for plant fertility (Palanivelu et al., 2003). A competent GABA flux could compensate for the lack of GABA breakdown ability in either pollen or pistil tissue to maintain cytosolic GABA within pollen tubes, unless GABA overaccumulates on both sides (e.g. ♂ pop2×♀ pop2; Palanivelu et al., 2003). Therefore, the combined impact of GABA catabolism via GABA-T and transport via GABA transporters (i.e. ALMTs and/or unidentified proteins) could contribute to fine-tuning pollen tube elongation within pistil to ovules, as summarized in Figure 2. This could be tested by stimultaneously knocking out GABA-T and GABA transporter genes expressed in pollen, together with in vivo intensity-based GABA fluorescence imaging using iGABASnFR in planta.

This is an intriguing combination of evidence and restrained speculation that indicates GABA could be a signal important for regulating pollen tube growth in coordination with other well-defined signals and transport networks as defined below. In summary, we propose that GABA has the potential to act as a signal that regulates pollen tube growth via its impact on anion efflux and potentially pH, and that this may occur through a Ca²⁺ mediated pathway (Figure 2).

Communication between glutamate and GABA signals in plants

GABA and glutamate are intimately linked through the synthesis of GABA via GAD. Not only is glutamate the substrate for GABA synthesis but also glutamate may stimulate Ca²⁺ entry into cells to activate GAD. Both glutamate and GABA have been implicated in playing a role in plant responses to wounding; thereby, we will discuss the potential relationship between GABA and glutamate in the context of wound signaling.

In response to wounding, plants generate long-distance electrical signaling, such as systemic surface potential changes and action potentials (APs; Hedrich et al., 2016; Farmer et al., 2020). Glutamate-dependent Ca²⁺ channels (i.e. GLR3.3 and 3.6) mediate wound-induced transient long-distance Ca²⁺ signal transduction and surface electrical changes via plasmodesmata that later stimulate distal jasmonate biosynthesis and systemic ROS propagation; this has been recently reviewed (Johns et al., 2021).

Wounding caused by the robotic caterpillar MecWorm on Arabidopsis (on leaf 8, the typical leaf for testing signal transduction to younger leaves) is also known to provoke systemic GABA accumulation in distal leaves (i.e. leaves 5, 11, and 13; Farmer et al., 2013; Scholz et al., 2015, 2017). It is unclear whether such systemic GABA accumulation is linked to glutamate-dependent Ca²⁺ activation of GAD(s), but the role of tonoplast-localized two pore calcium channel protein 1 in increasing cytosolic Ca²⁺ was ruled out (Scholz et al., 2017; Toyota et al., 2018). Cellular GABA metabolic status has been observed to affect stress (i.e. NaCl and hypoxia) triggered H⁺ flux, membrane potential changes and ROS signaling, where greater GABA accumulation is associated with faster restoration from stress-depolarized membrane potential and less ROS production (Su et al., 2019; Wu et al., 2021). Therefore, the question arises of whether GABA can facilitate the recovery of local cell membrane potential and/or mitigation of ROS damage if both are primed by glutamate-activated (GLR-mediated) Ca²⁺ influx during wound responses (Lew et al., 2020; Fichman et al., 2021).

Similar to surface potential changes, wound-stimulated APs involves long-distance transmission (Felle and Zimmermann, 2007; Zimmermann et al., 2009; Hedrich et al., 2016). APs can be propagated in barley by the application of many substances, such as NaCl, KCl, CaCl₂, glutamate, and GABA (Felle and Zimmermann, 2007). Amongst these, glutamate and GABA were proposed to act on putative “receptors” to prime Ca²⁺ influx, Ca²⁺-dependent Cl⁻ efflux, and initiate APs together with transient apoplastic pH regulation (Felle and Zimmermann, 2007). Later, Hedrich et al. (2016) proposed that AP are excited by membrane depolarization via anion efflux through R-type anion channels (e.g. ALMT12/QUAC1), followed by depolarization-activated K⁺ release through GORK and/or shaker-like outwardly-rectifying K⁺ channel (SKOR) to rehyperpolarize membrane potential. Indeed, the Arabidopsis GORK knock-out mutant (gork) had impaired APs in magnitude and duration when generated by electrical stimulation (Cuin et al., 2018), and GABA-stimulated K⁺ efflux was abolished in the root epidermis of gork1 mutants (Adem et al., 2020).

Although it has been noted that both glutamate and GABA may facilitate long-distance electrical signal transmission through plants, such as APs (Felle and Zimmermann, 2007), it is unclear whether they interact to shape such signals. On one hand, intracellular Ca²⁺ signal modulated by glutamate-dependent GLR may shape GAD activity and GABA signals (Zik et al., 1998; Toyota et al., 2018; Shao et al., 2020; Xu et al., 2021); on the other hand, GABA may be associated with apoplastic pH balance and cellular H⁺ flux via ALMTs and/or H⁺-ATPases that work together with glutamate to regulate the activity of GLR-mediated Ca²⁺ influx, membrane potential changes and ROS propagations (Ramesh et al., 2018; Kamran et al., 2020; Shao et al., 2020; Wu et al., 2021).

The following questions regarding the potential interaction of GABA and glutamate with APs would be important to examine: (1) Does GABA regulation of ALMT12 play a role? (Domingos et al., 2019; Xu et al., 2021); (2) Does ALMT-facilitated GABA efflux affect apoplastic pH? (Ramesh et al., 2018; Kamran et al., 2020); (3) What is the impact of cellular GABA metabolism on H⁺ and K⁺ (via GORK) flux regulation, and does this impact GLR activated Ca²⁺ waves and surface potential changes? (Su et al., 2019; Adem et al., 2020; Wu et al., 2021); and (4) Does GLR based activation of Ca²⁺ influx activate GABA synthesis and does this impact AP regulation?

Cross talk between GABA and plant hormones

Emerging evidence suggests that GABA as a signaling molecule interacts with other signals to coordinate particular physiological processes. In terms of guard cell signaling, ABA closes stomata via activation of Open Stomata 1/Snf1-Related protein Kinase 2.6- and/or CPK(s)-dependent phosphorylation on SLow Anion Channel-associated 1 (SLAC1)/SLAC1-homolog protein 3 (SLAH3) and ALMT12 to release anions (Mori et al., 2006; Geiger et al., 2011; Brandt et al., 2012, 2015; Imes et al., 2013; Gutermuth et al., 2018). ABA also phosphorylates tonoplast-localized ALMT4 to activate anion release from guard cell vacuoles to facilitate stomatal closure (Eisenach et al., 2017). GABA can attenuate ABA-induced stomatal closure at low doses (2.5 µM), presumably acting via the inhibition of ALMT12, since the loss of ALMT12 function in the almt12 mutant reduced stomatal sensitivity to both signals (Meyer et al., 2010; Imes et al., 2013; Xu et al., 2021). However, it is unknown whether GABA attenuates ABA’s effect also via reducing ALMT4-mediated anion release from vacuoles in this process. This could play out in a physiological scenario when cellular GABA increases to reduce the sensitivity of the guard cell to low ABA concentrations. However, GABA has no impact on the effect of high concentrations of ABA (25 µM) on stomatal closure implicating that reduced anion efflux via ALMT12 by GABA may not reverse guard-cell membrane depolarization and anion efflux through SLAC1/SLAH3 in such circumstances (Geiger et al., 2011; Brandt et al., 2012, 2015; Kollist et al., 2014; Xu et al., 2021). Collectively, this suggests that GABA homeostasis may fine adjust tissue sensitivity to cellular signals when the stimulus is of low intensity, but not antagonize the plant response when these signals are of sufficient magnitude. Intriguingly, a high dose of ABA (25 µM) does not fully close stomata on epidermal peels of gad2 mutants (Xu et al., 2021). The open stomata phenotype here was proposed to be due to deregulation of ALMT9 in gad2 mutants as discussed in section above; as such ALMT9 appears not to be a target of ABA. This suggests that some GABA-mediated processes may be not overwritten by amplifying other signals, and therefore provides an opportunity to engineer GABA responses in plants for altered outcomes to environmental stress.

Wound or herbivory attack on leaves stimulates systemic jamsonate (JA) and GABA biosynthesis in plants, as discussed above. JA accumulation promotes biosynthesis of secondary metabolites (e.g. glucosinolates) and proteinase inhibitors to repel herbivory attack, such as the Arabidopsis herbivore—Arion lusitanicus and rice (Oryza sativa) root-feeding insects—Diabrotica balteata and Lissorhoptrus oryzophilus (Falk et al., 2014; Lu et al., 2015; Wang et al., 2019). GABA production reduces insect growth and survival (e.g. Spodoptera littoralis larvae), probably due it its effects on invertebrate (insect) ionotropic GABA receptors at neuromuscular junctions (Bown et al., 2006; Scholz et al., 2015, 2017; Tarkowski et al., 2020). GABA depletion (in gad1/gad2) or overaccumulation (in pop2-5) does not alter JA biosynthesis stimulated by S. littoralis and MecWorm feeding (Scholz et al., 2015, 2017), and this implicates that endogenous GABA metabolism does not regulate of JA synthesis. But mutation in JAsmonate Resistant 1 (JAR1), in jar1, did cause greater GABA accumulation when attacked by S. littoralis (Scholz et al., 2015), and JAR1 encodes a jasmonate-amido synthetase that catalyzes the formation of JA-Ile that structurally is an amino acid (Ile) conjugated JA and directly facilitates the JA-signaling core target interaction (i.e. SCF^COI1-JAZ1; Staswick et al., 2002; Katsir et al., 2008). The loss of the key JA-Ile receptor (in coi1) and lowering JA-Ile stimulation (in cml37) both resulted in greater susceptibility to S. littoralis (Scholz et al., 2014). Taken together, JA signaling may affect the levels of wound-stimulated GABA production in plants or render the plant more susceptible to insect attack, damage and consequently stimulate more production of GABA.

Exogenous application of 10-mM GABA stimulates ethylene biosynthesis in sunflower (Helianthus annuus L.) and the Caryophyllaceae Stellaria longipes via up-regulation of ethylene signaling genes—1-Aminocyclopropane-1-carboxylic acid (ACC) synthase (ACS) and ACC oxidase (ACO) (Kathiresan et al., 1997, 1998; Booker and DeLong, 2015). Salt stress increased ethylene biosynthesis at 24 h and GABA production at 48 h in Caragana intermedia roots (Shi et al., 2010). Interestingly, 10-mM GABA supplement suppressed this early 24-h ethylene accumulation, whilst promoting ethylene production and further enhancing endogenous GABA accumulation 48 h post treatment (Shi et al., 2010). Similarly, GABA treatment has also been found to affect ethylene production in poplar (Populus tomentosa Carr) with a low dose of GABA (0.25 mM) enhancing ethylene synthesis, again through up regulation of ACS and ACOs (Ji et al., 2018). Together, this suggests that the GABA metabolism appears to affect ethylene synthesis response to salt stress in plants, and different plant species may vary in sensitivity to endogenous GABA in order to stimulate ethylene synthesis.

In plants, ethylene is a key hormone that controls (climacteric) fruit ripening and malate impacts fruit flavour (Alexander and Grierson, 2002; Liu et al., 2015; Hu et al., 2019; Wege, 2020). The downregulation of ethylene biosynthesis via silencing ACS an ACO genes is associated with low ethylene production in apple (Malus domestica) fruits (Dandekar et al., 2004; Defilippi et al., 2004). The malate content in apple fruit is expected to be significantly reduced at 2 weeks postharvest; however, it remains unchanged in low-ethylene transgenic apple fruits that can be reversed by exogenous ethylene application (Dandekar et al., 2004; Defilippi et al., 2004). Exogenous GABA treatment (10 mM) increases GABA and malate contents, but lowers ethylene synthesis in apple fruit during storage up to 70 d (Han et al., 2018). Malate storage in apple and tomato (Solanum lycopersicum) fruit is respectively linked with MdALMT9/MdMa1 and SlALMT9 (Ye et al., 2017; Li et al., 2020). MdALMT9/MdMa1, an ortholog of ALMT9 from Arabidopsis, encodes a tonoplast-localized channel catalyzing malate uptake into the vacuoles and facilitating malate accumulation in apple fruit (Li et al., 2020). Moreover, MdALMT9/MdMa1 contains identical amino-acid residues (FIYPIWAGEDLH) of the GABA regulation motif within Arabidopsis ALMT9, in which the mutation of two aromatic residues (F243 and Y245) abolished its GABA sensitivity in planta (Ramesh et al., 2017; Li et al., 2020; Xu et al., 2021), implicating that MdALMT9 might have GABA sensitivity as well. Intriguingly, both ethylene and GABA have been demonstrated to negatively regulate malate efflux through TaALMT1 at wheat root apices (Tian et al., 2014; Ramesh et al., 2015). Accordingly, the equilibrium between ethylene and GABA signaling may regulate fruit taste via the modulation of tonoplast-localized ALMT-mediated malate storage within fruit during ripening and postharvest storage. This provides a mechanistic link between GABA and ethylene that goes beyond the proposed association of GABA and ethylene production with malate metabolism (Defilippi et al., 2004; Han et al., 2018). Experiments would have to be performed to explore the explicit link between ALMT and GABA in apple and the other species mentioned in this section to determine whether this anion channel family provides the mechanism of cross talk between GABA and other signals.

Conclusion

Recent research has shown that GABA can fulfill a signaling role in plants that ultimately may regulate key growth, development and stress tolerance processes. As GABA synthesis increases during stress, to sustain energy production via the TCA cycle (Bown and Shelp, 2016; Gilliham and Tyerman, 2016), GABA has the potential to modulate other signals; cross talk of GABA therefore has the potential to fine tune plant physiology rather than initiating a physiological response per se. This appears to the case with the interaction with known signals such as ethylene and ABA, and in the regulation of guard cell movement, while GABA has the potential to fulfill a more overt role in signaling for pollen tube growth and fertilization and wound signaling—but this remains to be demonstrated. Many questions remain, such as: Are there additional GABA responsive elements beyond ALMTs (e.g. GORK and H⁺-ATPases)?; How does GABA flux through ALMTs relate to GABA signaling, and does this occur through two states of the same protein as suggested by Long et al. (2019) , or through alternative transporters?; and how does GABA flux regulate GABA distribution in tissues? Future research will no doubt explore these, and additional questions related to GABA’s physiological role in plants.

Advances

• GABA regulation of ALMT-mediated anion flux represents a class of physiological signal in plants that links primary metabolism to environmental responses.

• GABA mode of action in guard cells demonstrates that GABA modulates physiological responses rather than stimulates a response per se.

• Glutamate and GABA are intimately linked, with glutamate being a GABA precursor, and both modulate ion transport, indicating the concentrations of these metabolites may represent part of a homeostatic mechanism of membrane potential control and signaling.

• As a signal, GABA likely interacts with other hormonal signals to shape physiological processes.

Outstanding questions

• Does cytosolic GABA directly interact with ALMT9 and 12 in stomatal regulation?

• How does cellular GABA equilibrium fine tune the ionic flux and gradient during pollen tube growth?

• Do Arabidopsis ALMT9 or 12-14 facilitate GABA transport in stomatal and pollen tubes, as shown for TaALMT1?

• Does the GABA inhibitory effect on stomatal closing signals (e.g. dark, ABA, and H₂O₂) have a physiological role?

• How does GABA activate GORK-mediated K⁺ efflux?

• Are there other protein targets for GABA in plants beyond ALMTs?

Supplemental data

The following materials are available in the online version of this article.

Supplemental Figure S1 . GABA supplementation does not affect the photosynthetic rate of detached barley leaves.

 

B.X. and M.G. developed and wrote the paper, and N.S. conducted the experiments in Figure 1. All the authors provided edits.

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 Bo Xu (b.xu@adelaide.edu.au).