Seeds of Arabidopsis (Arabidopsis thaliana, ecotype Col-0) were grown in soil composed of equal parts of sterilized sand, loam soil, and a compost mixture of peat moss, perlite, and vermiculite (5:3:2, v/v/v). Seeds were grown under white fluorescent light (130 μmol photons/m²/s) at 24 ± 1°C. For the germination and seedling growth experiment, Arabidopsis seeds were grown on Murashige and Skoog (MS) media (Duchefa Biochemie, Haarlem, Netherlands) supplemented with 1% sucrose in a growth chamber at 24 ± 1°C and under a 16-h light/8-h dark period. Seeds were vernalized at 4°C for 2 days to synchronize germination.

To examine germination, primary root growth, and cotyledon greening, 200 seeds of wild-type plants, gork and pp2ca mutant plants were plated on MS agar medium containing ABA, NaCl, and mannitol. Germination was determined by radicle emergence. Primary root growth was measured 7 days after plating, and green cotyledons were measured 10 days after plating. Plates were placed in a growth chamber at 24°C for a 16-h light/8-h dark cycle.

Total RNA isolation and RT-PCR assays were conducted as described by Lim et al. (2015b), using leaf tissues of wild-type and gork mutant plants. To synthesize cDNA, we used 1 μg of the total RNA and a Transcript First Strand cDNA Synthesis kit (Roche, Indianapolis, IN, USA), according to the manufacturer’s protocol. For the RT-PCR assay, the cDNA was amplified using the iQ SYBR Green Supermix (Bio-Rad, Hercules, CA, USA) with GORK specific primers (F: CAGGAAACTTCCTTTGCACGAC; R: TCAAATGAATCATCGGAAAAATC). The AtACT8 gene was used for internal control.

For the subcellular localization analysis, we used full-length coding regions of GORK and PP2CA without the stop codon and p326GFP vector. To express each transient construct, Agrobacterium tumefaciens strain GV3101 was infiltrated to epidermal cells of Nicotiana benthamiana p19 strain (1:1 ratio; OD₆₀₀ = 0.5). Four days after such transient expression, tobacco leaves were cut and epidermal cells were observed with a confocal microscope (510 UV/Vis Meta, Zeiss, Oberkochen, Germany).

To verify the interaction between GORK and PP2CA, we used a yeast two-hybrid (Y2H) analysis (Lee et al., 2007). The C-terminal region of GORK and the full-length cDNA of PP2CA were inserted into pGBKT7 and pGADT7 vectors, respectively, and transformed into the AH109 yeast strain using the lithium acetate-mediated method (Ito et al., 1983). To evaluate the interaction between GORK and PP2CA, yeast cells containing each construct were spotted on selection media (SC-adenine-histidine-leucine-tryptophan) and grown for 4 days.

BiFC analysis was conducted as described previously (Lim and Lee, 2016). To generate BiFC constructs, open reading frames without a stop codon for GORK and PP2CA were inserted into Pro35S:SPYNE and Pro35S:SPYCE vectors, respectively (Waadt et al., 2008). To express each transient construct, Agrobacterium tumefaciens strain GV3101 was infiltrated into epidermal cells of N. benthamiana (OD₆₀₀ = 0.5). Four days after agroinfiltration, tobacco leaves were cut and epidermal cells were observed with a confocal microscope (510 UV/Vis Meta, Zeiss).

Previous studies have shown that GORK is involved in stomatal closure and water loss by transpiration (Hosy et al., 2003; Lefoulon et al., 2016). Even though it is widely recognized that clade A PP2Cs negatively modulates GORK channel activity and that CPK3 positively modulates GORK channel activity (Corratgé-Faillie et al., 2017; Lefoulon et al., 2016), the biological role of GORK in early stages of a plant’s life cycle remains unclear. To investigate the biological functions of GORK in seeds and seedlings, we identified T-DNA insertion mutants gork-1 (SALK_082258) and gork-2 (SALK_144737) as knockout alleles (Fig. 1A). First, we performed RT-PCR analysis to verify the knockout of gork mutants (Fig. 1B). GORK transcripts were detected in wild-type plants but not in gork mutants. The identified gork mutants showed no obvious phenotype under normal growth conditions (Fig. 1C). However, growth of wild-type and gork mutant plants was inhibited in the presence of 0.5 and 0.8 μM of ABA. Additionally, primary root lengths of gork mutants were significantly longer than that of wild-type plants (Fig. 1C and D).

Earlier, we demonstrated that PP2CA inhibits the activity of SnRK2.6 kinase, which consequently inhibits the SLAC1 anion channel (Lee et al., 2009, 2013). We conducted a Y2H assay to isolate the clade A PP2C interacting with GORK. We used the GORK C-terminal region as bait and clade A PP2Cs as prey. We performed a one-by-one Y2H assay with clade A PP2Cs (Fig. 2A); specifically, the yeast with GORK and PP2CA grew on the selective media, suggesting that GORK interacts with PP2CA. It has been demonstrated that the domain of clade A PP2Cs, including specific amino acids, plays an important role in the interaction with RCARs and SnRK2 kinases (Fuchs et al., 2013); however, little is known about the interaction with GORK. To elucidate which domain is important in the interaction between GORK and PP2CA, we generated truncated forms of PP2CA and performed Y2H. Growth of yeast with GORK was only detected with full-length PP2CA (Fig. 2B), indicating that full-length PP2CA is necessary for the interaction to occur. We used BiFC analysis to determine where the interaction complex was located inside the cell and green fluorescent protein (GFP) analysis to determine where each protein was located inside the cell (Fig. 2C). Co-expression of GORK-SPYNE with PP2CA-SPYCE was detected as yellow fluorescence predominantly in plasma membrane. The GFP signals of GORK-GFP and PP2CA-GFP were generated in the plasma membrane and nucleus, respectively. Such results agree with the Y2H analysis and indicate a physical interaction between GORK and PP2CA. In addition, it also suggests that GORK and PP2CA are both targeted at the plasma membrane, where they regulate K⁺ efflux thus causing stomatal closure (Corratgé-Faillie et al., 2017; Lefoulon et al., 2016).

PP2CA has been reported as one of the core ABA signaling component, where it acts as a negative ABA regulator (Baek et al., 2018; Lee et al., 2013; Lim et al., 2015a). To investigate the in vivo function of GORK in response to ABA, we compared the germination and seedling growth of wild-type plants, gork and pp2ca mutant plants (Fig. 3). In the absence of exogenous ABA, we did not find differences between the phenotypes in relation to seed germination and seedling development. However, when grown on media containing ABA, gork mutants showed significantly higher germination in comparison to wild-type plants, whereas pp2ca mutants showed significantly lower germination rate than wild-type plants (Fig. 3A). We also compared primary root growth and seedling establishment of wild-type, gork, and pp2ca mutant plants (Fig. 3B-D). ABA-induced inhibition of primary root growth in gork mutants was significantly lower in comparison to wild-type and pp2ca plants (Fig. 3B). Additionally, the number of green cotyledons in gork mutant plants was significantly higher in comparison to wild-type plants; whereas in pp2ca mutants, the number of green cotyledons was significantly lower in comparison to wild-type plants (Fig. 3C and D). These results suggest that knockout of GORK decreases sensitivity to ABA during germination, which leads to reduced primary root growth and hinders seedling establishment.

ABA signal and the abiotic stress response share common signaling components (Lim et al., 2015a; Sah et al., 2016). We therefore investigated the involvement of GORK and PP2CA in plant responses to abiotic stress, including high salinity and osmotic stress (Figs. 4 and 5). First, we evaluated the plants response to high salinity by comparing different phenotypes (wild-type, gork and pp2ca mutant plants) treated with NaCl (Fig. 4). Under regular growth conditions, there was no significant difference between wild-type and mutant plants. However, in the presence of NaCl, pp2ca mutants showed significantly lower germination rate compared to wild-type and gork plants (Fig. 4A). Moreover, the number of green cotyledons was significantly higher in gork mutant plants and significantly lower in pp2ca mutant plants in comparison to wild-type plants (Fig. 4B and C).

To further examine whether gork and pp2ca mutants showed different responses to osmotic stress, we measured germination and the number of green cotyledons in seeds treated with mannitol (Fig. 5). In contrast to the ABA and NaCl treatments, seed germination of gork and pp2ca mutants did not differ from that of wild-type plants (Fig. 5A). However, the number of green cotyledons was significantly higher in gork mutants and significantly lower in pp2ca mutants in comparison to the number of green cotyledons found on wild-type plants (Fig. 5B and C). These results indicate that both GORK and PP2CA affect the plant responses to high salinity and osmotic stress. Together, these results are consistent with earlier studies and with the genetic analyses performed. In conclusion, the shaker type K⁺ channel GORK is related to the ABA response as it interacts with clade A PP2C, affecting the plant’s response to high salinity and osmotic stress. However, gork mutants do not appear to be related in regulating turgor pressure in guard cells in germination and seedling growth. Identification of additional regulators of GORK, such as other kinases or phosphatases, will help clarify the exact functions of GORK in the ABA signaling pathway in seeds and seedlings.