Are PP2Cs with intrinsically disordered regions CO₂ sensors that control stomatal movements?

Abstract

A study has suggested that specific protein phosphatases are central CO₂ sensors that control CO2 regulation of stomatal conductance, which could not be unequivocally confirmed here.

Dear Editor,

Guard cells play a critical role in the regulation of photosynthesis and water use efficiency by controlling the size of stomatal pore apertures. Stomata respond to several environmental stimuli to regulate their apertures, including light, temperature, humidity, vapor pressure difference, CO₂ levels, dehydration status, pathogens and hormonal signals, to optimize carbon absorption and water loss in plants. The response to daily shifts in leaf CO₂ concentrations is important because stomata of C3 and C4 plants must open when the CO₂ concentration in leaves (Ci) is low due to photosynthesis, and close when Ci is elevated due to respiration, for example in the dark, thus preventing unnecessary water loss. In addition, the rising atmospheric CO₂ concentration is resulting in reduced stomatal apertures globally and affecting water use efficiencies of diverse plant species, due to this CO₂ response pathway (Raschke 1979; Assmann 1999; Medlyn et al. 2001; Franks et al. 2017). Recognizing the proteins that sense CO₂ has been a longstanding challenge in plant biology. In the present study, we have investigated the relevance of a recently reported direct CO₂ sensing mechanism by two specific Arabidopsis thaliana type 2C protein phosphatases (PP2Cs) that have been proposed to act as CO₂ sensors that regulate stomatal movements (Zhang et al. 2022). Several central early regulators of CO₂-induced stomatal movements have been identified. The Raf-like protein kinase High leaf Temperature 1 (HT1), the Mitogen-activated protein kinases MPK12 and MPK4, and the Convergence of Blue light and CO₂ signaling (CBC) kinases together form a signaling module that is necessary for the guard cell CO₂ response, with mutations in these genes strongly disrupting stomatal CO₂ signaling (Hashimoto et al. 2006; Des Marais et al. 2014; Hõrak et al. 2016; Jakobson et al. 2016; Hiyama et al. 2017; Tõldsepp et al. 2018; Takahashi et al. 2022). A CO₂/bicarbonate-triggered protein-protein interaction of MPK12/MPK4 with HT1, resulting in the inhibition of the Raf-like HT1 protein kinase, has been identified as a primary CO₂ sensor, based on strong genetic phenotypes, and protein biochemical, physiological, modeling, and imaging analyses (Takahashi et al. 2022). However, these findings do not exclude additional roles of CO₂/bicarbonate-regulated proteins in stomatal CO₂ signal transduction. The activity of the Slow anion channel-associated 1 (SLAC1) S-type anion channel, for example, is enhanced by increasing the CO₂/HCO₃⁻ concentration in heterologous cells and under physiological CO₂ concentrations in stomatal guard cells, together with structural resolution into the underlying mechanisms, suggesting that SLAC1 is a secondary CO₂/HCO₃⁻ sensor (Zhang et al. 2018; Li et al. 2022).

A recent study reported that specific PP2Cs with an intrinsically disordered region act as stomatal CO₂ sensors (Zhang et al. 2022). Zhang et al. (2022) revealed that an intrinsically disordered region (IDR) in the fungal PP2C, Ptc2, is required for CO₂-dependent phase separation and CO₂-responsive activation of the phosphatase activity. Furthermore, Zhang et al. (2022) found that the Arabidopsis PP2C5 (also known as AP2C3) and PP2C74 are the only PP2Cs among the ∼80 PP2Cs encoded in the Arabidopsis genome (Schweighofer et al. 2004; Xue et al. 2008; Singh et al. 2010), that specifically have an intrinsically disordered region similar to the fungal Ptc2. The authors reported that the plant PP2C5(AP2C3) phosphatase (like fungi) functions as a direct CO₂ sensor in vitro. The Arabidopsis PP2C5(AP2C3) was reported to show strongly increased PP2C activity upon CO₂ elevation (Zhang et al. 2022). This led to a new model that IDR-containing PP2Cs in plants, including PP2C5(AP2C3) and PP2C74, function as CO₂ sensors that regulate stomatal movements (Zhang et al. 2022). In their study, PP2C5(AP2C3) phosphatase activity was dramatically stimulated as the CO₂ concentration increased from ambient CO₂ 0.04% (400 ppm) to 0.1% (1,000 ppm) and further increased at ∼0.5% (5,000 ppm) CO₂ (Zhang et al. 2022).

To investigate how the above PP2Cs function in stomatal CO₂ signal transduction, we first analyzed single mutant plants pp2c5(ap2c3) (SALK_109986). In infrared thermal imaging, pp2c5 plant leaf temperatures were comparable to wild-type (WT) Col-0 under low CO₂ (100 ppm) (Fig. 1a). High leaf temperature 1 (ht1-2) mutant plants were included in the same pots as controls, as ht1-2 mutant plants are known to exhibit increased leaf temperatures at low CO₂ due to disrupted low CO₂-induced stomatal opening (Hashimoto et al. 2006). Stomatal conductance measurements performed during CO₂ transitions (400 → 800/900 → 100 ppm) showed CO₂ responsiveness (Fig. 1b and c; Figure S1). In two of the experimental sets, a larger stomatal conductance response was observed during exposure to 100 ppm CO₂ in the pp2c5(ap2c3) mutant (e.g., Figure S1a), whereas in other experiments, no clear phenotype was observed in intact leaf and whole plant gas exchange experiments (Fig. 1b; Figure S1b). An increased stomatal conductance in response to the transition to low CO₂ was not consistently reproduced across independent experiments (>four independent experimental sets conducted by KP, P-KH, and HJ).

Figure 1.

Stomatal CO₂ response and thermal imaging of pp2c mutants. a, d, f) Infrared thermal images of wild-type (WT) Col-0, pp2c5(ap2c3) (SALK_109986), pp2c74 (SAIL_682_D10), and ap2c1 (SALK_065126) single mutant plants, shown alongside ht1-2 (high leaf temperature 1-2) and wild-type controls grown in the same pots. Pseudo-colored temperature scales (°C) are shown to the right of each image. Thermal image (a) and stomatal conductance (b and c) of WT and pp2c5(ap2c3) single mutant. Panels b and c represent independent experimental sets performed by different experimenters. Thermal image (d) and stomatal conductance (e) of WT and pp2c74. Thermal image (f) and stomatal conductance (g) of WT and ap2c1. Gas exchange analyses of pp2c5 and pp2c74 mutants were conducted in parallel using the same Col-0 wild-type plants as controls. Plants were grown in potting soil (Sungro Horticulture, Professional Growing Mix, MA, USA) under controlled chamber conditions (AR-41L2, Percival Scientific, USA): 12 h light/12 h dark cycles, 400 ppm CO₂, 65 to 68% relative humidity, and 21 °C. Thermal images were captured after 45 min exposure to low CO₂ (100 ppm); warmer leaf temperatures indicate reduced transpiration (eg during stomatal closure) and cooler temperatures indicate increased transpiration (eg stomatal opening; pseudo-colored scales (°C) shown to the right). Plants were 3 to 4 weeks old at the time of analysis. Stomatal conductance was measured using portable gas-exchange systems (LI-6400XT (panels c, e and g), LI-6800 LI-COR (B), Lincoln, NE, USA) with 150 μmol m⁻² s⁻¹ light, airflow of 500 μmol s⁻¹, 21 °C leaf temperature, and 65 to 68% relative humidity. Leaves were equilibrated at ambient CO₂ (400 ppm) until stomatal conductances were stable, followed by sequential shifts to 800 or 900 and 100 ppm CO₂. Plants were 5 to 6 weeks old at the time of analysis. Stomatal conductance data represent mean ± SEM, n = 3 to 4 plants per genotype in each figure panel (Fig. 1e, n = 2 plants, see Figure S2 for independent replicates). Experiments were repeated in at least 3 independent experimental sets by several co-authors (KP, MK, P-KH and HJ; see text). Genotype blinded experiments were conducted with support of three individuals: (1) different “blinders”, (2) co-authored experimenter and (3) different “unblinders”, who are not all co-authors.

A pp2c74 knockdown mutant with a T-DNA insertion in the 5′-UTR of PP2C74 (SAIL_682_D10) that shows an average 209-fold reduction in PP2C74 transcript level was identified to examine the role of PP2C74 in stomatal movements (Table S1 and Figure S2a). In pp2c74 mutant plants, thermal imaging showed typical biological variability, but no consistent effect on leaf temperature under low CO₂ (100 ppm) (Fig. 1d). Gas-exchange analyses performed during CO₂ transitions (400 → 800 → 100 ppm) revealed no clear consistent differences in stomatal conductance responses between pp2c74 and WT plants (Fig. 1e & Figure S2b; 3 independent experimental sets with n = 3 to 4 plants per genotype in each set by KP, see Methods in figure legends and Supplementary Methods for details).

We next examined the ap2c1 single mutant (SALK_065126), given that AP2C1 is a close homolog of PP2C5(AP2C3) and has been shown to act redundantly with PP2C5(AP2C3) (Brock et al. 2010). ap2c1 mutant plants were not clearly different from WT in both thermal imaging (Fig. 1f) and CO₂-regulated stomatal conductance assays (Fig. 1g). These experiments were conducted in 3 sets (KP), with n = 3 to 4 plants per genotype in each set. In general, pp2c5(ap2c3), pp2c74, and ap2c1 single mutant plants showed functional stomatal CO₂ responses, with some experiments showing slight variation towards the end of gas exchange experiments. In contrast, parallel-grown ht1-2 control plants showed higher leaf temperatures in the same pots in infrared imaging (Fig. 1 a, d and f), consistent with its known disruption of stomatal opening in response to low CO₂ (Hashimoto et al. 2006; Takahashi et al. 2022).

Because AP2C1 and PP2C5(AP2C3) are closely related and functionally redundant (Brock et al. 2010), we next asked whether overlapping functions might mask phenotypes in the single mutants. To address this possibility, we obtained an ap2c1/pp2c5 double mutant from Brock et al. (2010) (Figure S6; see Methods in figure legends and Supplementary Methods for details). ap2c1/pp2c5 double mutant leaves did not show consistently different temperatures of leaves compared to parallel-grown wild-type controls under low CO₂ (100 ppm) in thermal imaging experiments (Fig. 2a). In gas-exchange measurements, 4 of 8 independent experimental sets, with n = 3 to 4 plants investigated per genotype in each set, responses were similar to the parallel-grown CO₂ responses of wild-type plants (Fig. 2b; Figure S3a and S3b, including genotype-blinded experiments). These separate experimental sets were conducted independently by KP, PKH, MK, and HJ). In 4 of 8 experimental sets, an increased stomatal opening response was observed in the ap2c1/pp2c5 double mutant at low (100 ppm) CO₂ (Fig. 2c) (separate experimental sets conducted independently by KP, MK).

Figure 2.

Stomatal conductance and thermal imaging of ap2c1/pp2c5 double and ap2c1/pp2c5/pp2c74 triple mutant alleles, and biochemical analysis of PP2C5 phosphatase activity changes at low vs. high CO₂. (a) Infrared thermal images of wild-type (WT) Col-0 control, ap2c1/pp2c5 double, and the ht1-2 control after 45 minutes of exposure to low CO₂ (100 ppm). Temperatures (°C) are indicated in pseudo-color scales to the right. (b and c) Stomatal conductance measurements in leaves of WT and the first ap2c1/pp2c5 double mutant allele, ap2c1 (SALK_065126) and pp2c5 (SALK_109986) (Brock et al. 2010), during sequential CO₂ transitions (400 ppm → 800 ppm → 100 ppm), (±SEM, n = 3 plants; panels b and c: LI-6400XT). (d) Stomatal conductance of WT controls and a second independent ap2c1/pp2c5 allele ap2c1 (CS886675) and pp2c5 (SALK_102925C) under the same CO₂ regime, displaying wild-type–like responses (±SEM, n = 3 plants; LI-6400XT). Gas exchange analyses of 1st allele and 2nd allele double mutants were conducted in parallel using the same Col-0 wild-type plants as controls in the same experimental set. The exonic insertion sites of the 2nd allele double mutant allele were confirmed by sequencing. (e) Thermal imaging and (f) stomatal conductance of WT control plants and ap2c1/pp2c5/pp2c74 triple mutant plants, pp2c5 (SALK_109986), pp2c74 (SAIL_682_D10), and ap2c1 (SALK_065126). Plants that were 3 to 4 weeks old were used for these whole-plant gas exchange experiments. In (e) Temperatures (°C) are indicated in pseudo-color scales to the right. In (f) A temperature-controlled multi-cuvette gas exchange device (Plant Invent, Tartu, Estonia) was used to measure stomatal conductance in four plants in parallel. The plants were placed inside gas exchange cuvettes and exposed to a temperature of 24 °C, 70% humidity, light intensity of 150 μmol m⁻² s⁻¹, and 400 ppm CO₂ for about 90 min to acclimate the plants. After the stomatal conductance stabilized, the concentration of CO₂ was first raised to 800 ppm and then to 100 ppm as depicted. Data represent mean ± SEM. n = 3 to 4 plants per genotype in each experimental set. (g and h) In vitro protein phosphatase activity of recombinant PP2C5(AP2C3) under low (50 ppm) and high (2,000 ppm) CO₂. The protein phosphatase activity of PP2C5(AP2C3) was measured using the Serine/Threonine Phosphatase Assay System (V2460, Promega) (Zhang et al. 2022). Reactions contained 5 µL of 100 µM Ser/Thr phosphopeptide (RRA(pT)VA) mixed with or without recombinant PP2C5 protein (0.5, 1, or 2 µg) in 50 µL PPase-2C reaction buffer [50 mM imidazole (pH 7.2), 0.2 mM EGTA, 5 mM MgCl₂, 0.02% β-mercaptoethanol, 0.1 mg/mL BSA], following the manufacturer's instructions. Reactions were incubated for 30 min at 30 °C under 50 ppm or 2,000 ppm CO₂, as described by Zhang et al. (2022), in ½-area flat-bottom 96-well plates (Costar). After incubation, 50 µL of the molybdate dye/additive mixture was added, and the plate was incubated for 30 min at room temperature. Absorbance was measured at 600 nm using a Spark plate reader (TECAN), and activity was calculated from a phosphate standard curve (0 to 2,000 pmol free phosphate) (n = 3 assays per PP2C protein concentration). Data were analyzed using one-way ANOVA followed by Tukey's multiple comparison test to evaluate differences among all protein-by-CO₂ treatment groups. Significance was defined as adjusted P < 0.05. Different letters above bars indicate groups that are significantly different from each other, and groups sharing at least one letter are not significantly different. Error bars represent mean ± SEM. For every PP2C5 protein concentration tested (0, 0.5, 1, and 2 μg), absorbance and activity at 50 ppm vs. 2,000 ppm CO₂ share the same letter, indicating no significant CO₂-dependent change in PP2C5 activity. Significant differences were observed only in PP2C5 protein amount, reflecting increased reaction rates at higher enzyme concentrations.

A second ap2c1/pp2c5 exon insertion sequence-confirmed double mutant allele was generated (ap2c1: SAIL_590_B11; pp2c5 (ap2c3): SALK_102925C) and showed no clearly increased stomatal conductance in response to low CO₂ compared to WT (Col 0) control plants (Fig. 2d and Figure S3c; independent experimental sets of MK and HJ, including genotype-blinded analyses).

Analysis of guard cell transcriptome datasets indicates that PP2C5 and AP2C1 are expressed in Arabidopsis guard cells, with PP2C5 showing the highest transcript abundance in most of the Yang et al. (2008) and Pandey et al. (2010) datasets (Figure S4) (Winter et al. 2007). Additional guard cell transcriptome datasets from independent laboratories show that PP2C74 can be the second-highest expressed among these PP2Cs (Wang et al. 2011; Bates et al. 2012) and, in some conditions, the most strongly expressed among the three PP2Cs in guard cells (Mustroph et al. 2009). Together, these datasets support guard-cell expression of all three PP2Cs. We therefore next tested the hypothesis that if PP2C5(AP2C3) and PP2C74 act as CO₂ sensors (Zhang et al. 2022), combining them with ap2c1 may uncover a stronger phenotype. We next generated and investigated a pp2c5/pp2c74/ap2c1 triple mutant line (Figure S7). Infrared thermal imaging experiments showed typical biological variability, but no consistent difference from wild type control plants grown in the same pots (Fig. 2e). Stomatal conductance responses to CO₂ concentration transitions of 400 → 800 → 100 ppm resembled WT control plants in three experimental sets (eg Figure 2f; completed independently by KP and MK). In contrast to the previously analyzed mutants, in one experiment, pp2c5/pp2c74/ap2c1 triple mutant leaves showed a slightly lower stomatal conductance at the end of the 100 ppm exposure than WT plants (Figure S5; unpaired two-tailed Welch's t-test P = 0.0407). Together, these experiments suggest that the investigated PP2Cs are not essential for CO₂ regulation of stomatal movements under the imposed conditions, but may cause small effects on stomatal conductance in some experiments.

To evaluate biochemical experiments reporting that the PP2C5(AP2C3) protein functions as a strongly CO₂-activated protein phosphatase in in vitro experiments (Zhang et al. 2022), we carefully pursued the experimental methods for these experiments as described by Zhang et al. (2022). We analyzed purified PP2C5 activity under low (∼50 ppm) and high (∼2,000 ppm) CO₂ concentrations (Fig. 2g and h). In these experiments, PP2C5 activity showed no clear, strong, and consistent activation upon CO₂ concentration elevation (n = 3 total experiments with three to four PP2C5 protein levels tested in each experiment; Fig. 2g). The determined PP2C5 activities showed only small average differences between low and high CO₂ (Fig. 2h). This finding contradicts the reported strong activation of PP2C5 phosphatase activity in vitro at high CO₂ and puts into question whether PP2C5 enzymatic activity itself is clearly modulated by CO₂ concentration changes.

Together, these data demonstrate that AP2C1, PP2C5, and PP2C74 are not robust and essential CO₂ sensors that control CO₂-regulated stomatal movements in Arabidopsis guard cells. Upon careful evaluation of all independent experimental sets, the apparent enhancement in stomatal opening at low CO₂ observed in some experiments was weak and variable, and did not represent a reproducible phenotype. Additionally, occasional increases in basal stomatal conductance were observed in some pp2c5 single-mutant and ap2c1/pp2c5 double-mutant experimental sets. However, these effects were modest and not consistently reproducible across independent experiments, mutant alleles, or gas-exchange platforms. Given that PP2C5 has been reported to interact with and inactivate stress-activated MAP kinases, including MPK3, MPK4, and MPK6 (Brock et al. 2010), and that MAP kinase activity is not in first order required for stomatal CO₂ sensing (Takahashi et al. 2022), such interactions in guard cells may contribute to the modest and variable increases in stomatal conductance observed in some experimental sets. In addition, pp2c5/pp2c74/ap2c1 triple mutant leaves showed either no difference from wildtype CO₂ responses (n = 3 independent experimental sets) or a slight reduction in stomatal opening at low CO₂ in one experimental set. As three different gas exchange analyzers were used in the present study (see figure legends), it is interesting that with the new generation Li-COR 6800 and PlantInvent gas exchange systems, which have improved constant humidity and leaf temperature control (Garen et al. 2022) as well as control of other parameters, no aberrant phenotypes were observed in any of the mutants in the present study. These new gas exchange systems are known to employ more robust approaches to controlling and monitoring leaf temperature, humidity, and vapor pressure deficit, thereby enabling more precise stomatal conductance measurements (Hsu et al. 2021; Garen et al. 2022; Pankasem et al. 2024). Our findings cannot strictly exclude the possibility that these PP2Cs may have a modulatory influence that occasionally alters stomatal opening under low CO₂, or that these PP2Cs may have a modulatory role in stomatal movements under specific conditions, given their expression in guard cells. The contribution of these PP2Cs to the direct CO₂ regulation of stomatal movements is not essential, as functional CO₂ responses were observed in experiments in the physiological range from 100 ppm, 400 ppm, up to 900 ppm (Figs. 1 and 2; Figures S1, S3, and S5). In several experiments, mutant responses were similar to those of wild-type Col-0, whereas in others they showed slight deviations. Nevertheless, all mutants exhibited robust stomatal closure at 800 to 900 ppm CO₂ and reopened at low CO₂. This indicates that the core CO₂ signaling pathway remains intact even in the absence of these PP2Cs. Furthermore, elevated CO₂ did not clearly or strongly increase recombinant PP2C5 activity. Earlier studies showed partial redundancy between AP2C1 and PP2C5(AP2C3) in other physiological processes, including stress hormone production, pathogen resistance, baseline stomatal conductance, and wound-induced responses (Brock et al. 2010). The present study does not support a role of the investigated Arabidopsis PP2Cs with intrinsically disordered regions as essential stomatal CO₂ sensors. The proposal that PP2C5 acts as a CO₂ sensor (Zhang et al. 2022) is of interest because it raises the possibility of a class of CO₂-regulated plant proteins. Reporting our independent findings provides additional context for the community. We hope this information will help guide future work and models of stomatal CO₂ signaling using analyses of intact plant leaf stomatal responses. Future studies aimed at mapping how protein kinases, phosphatases, carbonic anhydrases, ion channels, transporters, and other proteins are regulated will be necessary for a complete understanding of the stomatal CO₂ signal transduction pathway. We hope this clarification will help uncover presently unknown stomatal CO₂ signaling mechanisms while continuing to explore additional hypotheses.

Author contributions

K.P. and J.I.S. designed the research. K.P., M.K., Y.T., H.J., and P.-K.H. performed the research. J.I.S. contributed to reagents/analytic tools. K.P., M.K., Y.T., H.J., and P.-K.H. analyzed the data. K.P. and J.I.S. wrote the paper.

Supplementary material

Supplementary material is available at Plant Physiology online.

Funding

This research was funded by National Science Foundation (NSF), United States grant MCB-2401310 to J.I.S. H.J. was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (RS-2023 00239562). Y.T. was supported by a Japan Science and Technology Agency (JST) PRESTO grant (JPMJPR21D8).

Data availability

The data supporting the findings of this study are included within the article and its supplementary materials.