GABA Regulation of Stomatal Function in Arabidopsis thaliana

Abstract

Water scarcity limits crop yield. This is in part because of a reduction in plant photosynthetic capacity due to a trade-off between water loss through transpiration and CO₂ intake, which is ordinarily optimised through the opening and closing of stomata, micropores located on the surface of aerial part of plants. Stomatal pores are delineated by pairs of guard cells and stomatal movement is driven by osmolarity changes within guard cells compared to the surrounding cells, which is regulated by an elaborate network of transporters and signalling pathways. GABA (γ-aminobutyric acid) is a non-proteinogenic amino acid in plants, which is mainly synthesised from glutamate catalysed by Glutamate Decarboxylase (GAD) in the cytosol. There are 5 GAD genes (GAD1-5) identified in Arabidopsis thaliana, with GAD1 and GAD2 the most abundant transcripts. GABA has long been speculated as a signalling molecule in plants, with reports connecting GABA synthesis or metabolism to physiological phenotypes, such as accumulation of biomass, pollen tube elongation and tolerance to stress. The recent discovery that Aluminium-activated Malate transporters (ALMTs) may act as GABA receptors has identified a potential mechanism by which GABA affects membrane potential and can act as a signal in plants. A number of ALMTs are involved in stomatal movement, so here, we use stomatal guard cells as a model system to further investigate whether GABA acts as a signalling molecule in plants through the manipulation of GABA metabolism and ALMT expression in Arabidopsis. Initial findings within this thesis were that ablation of GAD2, the predominant GAD in leaves, led to significantly reduced endogenous GABA in leaves and enlarged stomatal pores, decreased water use efficiency, increased drought sensitivity and reduced sensitivity to abscisic acid (ABA) induced stomatal closing; these were restored to wildtype levels by reintroduction of GAD2 into leaves targeted exclusively into guard cells. Endogenous concentrations of GABA appeared to be negatively associated with stomatal opening in an ALMT9 dependent manner, which is a tonoplast localised anion transporter. These initial findings are the first clear genetic demonstration that GABA signalling can occur in planta. The impact of the other GADs on GABA signalling processes within stomata were further studied by employing the higher order gad1/2/4/5 knockout mutant, which has further reduced GABA production. Surprisingly, the quadruple mutant did not mimic the stomatal phenotype of gad2, instead it resembled WT in stomatal aperture, stomatal conductance and drought tolerance. However, when GAD2 was expressed in the quadruple mutant background it elevated stomatal conductance to near gad2 levels. We hypothesised that the divergent phenotypes of gad2 and gad1/2/4/5 were due to varied traits of GADs homologues, which may result in diverse GABA distribution tissue-wise and thus alter plants response to GABA. This was further explored on other higher order mutants generated from crossing the parental gad2-1 with gad1/2/4/5, comparing to phenotypes of single mutants of gad1, gad2 and gad4. The F₁ gad2-1 x gad1/2/4/5 plants mimicked the genotype of gad2-1. The filial generation has elevated stomatal conductance which consistent with the mutation of GAD2 being causation of more opened stomata. Stomatal conductance and aperture measurements on F₂ and F₃ plants suggests a synergistic effect of GAD homologues in mediating GABA signalling of stomatal movement. Our results indicate that knockout of both GAD4 and GAD5 contributes to the divergent phenotypes of gad2 and gad1/2/4/5. We were also able to show that GABA at a non-stressed concentration (0.5 mM) increases stomatal aperture during opening assays of WT in contrast to when GABA is applied at a stress level of GABA (2 mM) which limits opening. Increased opening due to 0.5 mM GABA was absent in several GAD(s) mutant plants, which again suggested an altered cellular homeostasis caused by various mutations in GADs. Finally, results from epidermal strip assays with pharmacological treatment of GABA and/or ABA suggests GAD1 and GAD4 are required for the full response to ABA for inhibition of stomatal opening. In summary, this thesis demonstrates that the GABA-ALMT9 interaction mediates a pathway by which GABA signalling occurs in the stomata of Arabidopsis. However, it also reveals a complexity in GABA regulation of stomatal movement where it is not a simple linear dose-response relationship and, rather it involves cross talk that is likely to involve multiple GAD homologues.