Involvement of abscisic acid, ABI5, and PPC2 in plant acclimation to low CO2.

Phosphoenolpyruvate carboxylase (PEPC) plays a pivotal role in the photosynthetic CO2 fixation of C4 plants. However, the functions of PEPCs in C3 plants are less well characterized, particularly in relation to low atmospheric CO2 levels. Of the four genes encoding PEPC in Arabidopsis, PPC2 is considered as the major leaf PEPC gene. Here we show that the ppc2 mutants suffered a growth arrest when transferred to low atmospheric CO2 conditions, together with decreases in the maximum efficiency of PSII (Fv/Fm) and lower levels of leaf abscisic acid (ABA) and carbohydrates. The application of sucrose, malate, or ABA greatly rescued the growth of ppc2 lines under low CO2 conditions. Metabolite profiling analysis revealed that the levels of glycine and serine were increased in ppc2 leaves, while the abundance of photosynthetic metabolites was decreased under these conditions. The transcript levels of encoding enzymes involved in glycine or serine metabolism was decreased in ppc2 in an ABI5-dependent manner. Like the ppc2 mutants, abi5-1 mutants had lower photosynthetic rates and Fv/Fm compared with the wild type under photorespiratory conditions (i.e. low CO2 availability). However, the growth of these mutants was similar to that of the wild type under non-photorespiratory (low O2) conditions. The constitutive expression of ABI5 prevented the growth arrest of ppc2 lines under low CO2 conditions. These findings demonstrate that PPC2 plays an important role in the acclimation of Arabidopsis plants to low CO2 availability by linking photorespiratory metabolism to primary metabolism, and that this is mediated, at least in part, through ABA- and ABI5-dependent processes.

Introduction

Phosphoenolpyruvate carboxylase (PEPC; EC 4.1.1.31) is a ubiquitous enzyme in plants, algae, and bacteria (Chollet ). There are two functional forms of PEPC in higher plants, namely the photosynthetic and non-photosynthetic isoforms. In crassulacean acid metabolism (CAM) and C4 plants, PEPC plays a pivotal photosynthetic role in primary CO2 fixation, by catalyzing β-carboxylation of PEP with HCO3− to oxaloacetate and inorganic phosphate. Photosynthetic PEPC activity in C4 plants is essential for carbon assimilation via pre-fixation of CO2 in the bundle sheath, and thus decreases photorespiratory activity and contributes to greater water use efficiency and photosynthetic efficiency compared with C3 plants (Taylor ). The non-photosynthetic PEPCs play key roles in plant primary metabolism by replenishing the tricarboxylic acid (TCA) cycle to support carbon and nitrogen metabolism (Masumoto ). Moreover, they are also important for stomatal opening (Gehlen ) and supplying malate as a respiration substrate to symbiotic N2-fixing bacteroids in legume root nodules (Vidal and Chollet, 1997). PEPCs in C3 plants are believed to play minor roles in photosynthesis or photorespiration (von Caemmerer, 2013). Recently, 13C fixation NMR analysis in sunflower indicated that PEPC activity significantly increased along with decreases in net CO2 assimilation under high photorespiratory conditions (Abadie ; Abadie and Tcherkez, 2019). Moreover, OsPPC4 mutation in rice led to great accumulation of photorespiratory intermediates such as glycine, serine, and glycerate (Masumoto ). Since photorespiration is closely related to primary metabolism (Mouillon ; Rachmilevitch ), these studies imply that C3 PEPCs may play certain roles in photosynthesis and photorespiration.

In Arabidopsis, there are four genes encoding PEPC. Mutations in either PPC1, PPC2, or PPC3 lead to decreases in fresh weight and flowering time delay (Feria ). Moreover, PPC1 and PPC2 mutations greatly reduced malate and citrate synthesis, and severely suppressed ammonium assimilation, which finally leads to ppc1ppc2 growth arrest (Shi ). However, little is known about how these PPC genes regulate plant primary metabolism, and whether they regulate photosynthesis and plant development under stress conditions such as low CO2 and drought stress.

Abscisic acid (ABA) plays a prominent role in plant stress tolerance, and regulates various important plant development aspects, such as seed dormancy, germination, seedling establishment (Finkelstein ), and vegetative development (Sharp, 2002). ABA regulates these processes through affecting gene expression by modulating ABA-responsive transcription factors, such as B3-domain family proteins (e.g. ABI3 and VAL1) (Giraudat ; Suzuki ), APETALA2 (AP2) family proteins (e.g. ABI4) (Finkelstein ), and basic leucine zipper (bZIP) family proteins (e.g. ABI5) (Finkelstein and Lynch, 2000). ABI5 is a key component in ABA-triggered pathways during germination, seedling establishment, and vegetative growth (Lopez-Molina ); in addition, it has been reported to play certain roles in nitrogen assimilation and signaling (Signora ). ABI5 also positively regulates SGR1 and NYC1, two chlorophyll catabolism-related genes, through recognizing their upstream ABA-responsive elements (ABREs) (Sakuraba ). Hence, ABI5 has been proposed as a key regulator to monitor environmental conditions during seedling growth. However, it remains unknown whether ABI5 is responsive to CO2 changes and affects plant growth under different CO2 conditions.

Environmental changes have a rapid effect on plant intercellular CO2 concentration (Ci). Plant leaf photosynthesis under light conditions causes a quick drop of Ci to <200 ppm (Engineer ). In addition, drought stress triggers stomatal closure, thereby reducing CO2 uptake and lowering Ci, which results in reduced photosynthesis (Parry ). In this study, we demonstrated the crucial role of AtPPC2 in seedling growth under low CO2 conditions by linking photorespiration metabolism with primary metabolism, with the involvement of ABI5. ppc2 mutants showed retarded seedling growth, reduced CO2 assimilation, and suppressed ABI5 expression under low CO2 conditions. Metabolic analysis showed the accumulation of the photorespiratory intermediates glycine and serine, and a decrease of malate in ppc2 under low CO2 conditions. ABI5 overexpression rescued the growth arrest phenotype at low CO2. Taken together, our results demonstrate that PPC2 and ABI5 are key regulators of plant acclimation to low CO2, and positively contribute to carbon fixation and metabolism in C3 plants.

Materials and methods

Plant materials and growth conditions

All ppc mutant lines used in this study were in the Columbia (Col-0) background. The mutant lines ppc1 (SALK_088836) (Feria ), ppc2 (SALK_128516) (Shi ), and ppc3 (SALK_143289) (Feria ) were obtained from the Arabidopsis Biological Resource Center (http://abrc.osu.edu), and their homozygosity was confirmed by PCR (Supplementary Table S1 at JXB online).

Seeds were surface-sterilized and germinated. For A–Ci curves and drought stresses, plants were grown in pots and for all other experiments the plants were in Petri dishes. All plants in pots and Petri dishes were grown in CO2-controlled growth chambers (Percival), in which CO2 concentrations can be accurately and stably controlled in the range of 100–1500 ppm, with a light regime of 16 h light/8 h dark (light intensity 100 μmol m−2 s−1) and a relative humidity of 56%. For mild drought stress, a total of 64 five-day-old seedlings were transferred into a new pot and grown for 5 d and then the plants were not watered for 10 d.

Generation of constructs and transgenic plants

The coding sequences (CDSs) of PPC2 and ABI5 were amplified from Arabidopsis cDNA with primers PPC2-OE-F/PPC2-OE-R and ABI5-OE-F/ABI5-OE-R, respectively. The PCR products were cloned into pGreen-35S and pEarlyGate-35S-YFP (Earley ). The PPC2 promoter region amplified by primers Pro-PPC2-F and Pro-PPC2-R was cloned into the pEarlyGate-100-GUS vector (Earley ). All primers used for generation of construct are presented in Supplementary Table S1. The constructs were introduced into Arabidopsis by Agrobacterium tumefaciens-mediated transformation using the floral dipping method (Clough and Bent, 1998).

Determination of the chlorosis rate

Fifteen-day-old seedlings of Col-0 and the ppc2 mutant grown at 200 ppm and 400 ppm CO2 were analyzed for the chlorosis rate. Cotyledons which were almost bleached were defined as chlorotic cotyledons. The chlorosis rate was determined as:

PEPC activity assays

Proteins were extracted from leaves (0.1 g FW) of 15-day-old seedlings in 1 ml of extraction buffer [200 mM HEPES-NaOH (pH 7), 10 mM MgCl2, 5 mM DTT, and 2% (w/v) polyvinylpyrrolidone-40]. After centrifugation, the supernatant was immediately used for PEPC activity detection by a PEPC activity assay kit (Jiancheng Bioengineering Institute, Nanjing, China). One unit of PEPC activity was defined as 1 nmol of NADH oxidation per min per mg protein at 25 °C. The total protein content was determined by using a BCA (bicinchoninic acid) protein assay kit (Sangon Biotech).

A–Ci curve analysis

The measurements of A–Ci curves were performed by a closed infrared gas exchange analysis system (LI-COR 6400XT). Four- to five-week-old leaves were clamped in the 2 cm2 chamber with leaf temperature at 21 °C for measurement. The A–Ci curve measurements were performed at CO2 concentrations of 50, 100, 200, 300, 400, 600, and 800 ppm with a photosynthetic photon flux density of 2000 μmol m−2 s−1. The relative humidity was ~50% in all measurements. A–Ci curves under low oxygen conditions were performed by replacing air with pure N2.

Analyses of sucrose and starch contents

Sucrose and starch were extracted from 15-day-old seedlings and estimated by using the Sucrose Colorimetric/Spectrophotometric Assay kit (Comin Biotechnology Co., Ltd, Suzhou, China) and Starch Colorimetric/Spectrophotometric Assay kit (Comin Biotechnology Co., Ltd) according to the manufacturer’s instructions.

Analyses of chlorophyll content and chlorophyll fluorescence

Total leaf chlorophyll was extracted with 80% acetone at 4 °C for 24 h in darkness, and then the supernatant was used for the absorbance measurement with a spectrophotometer (BeckMan Coulter DU730). The total chlorophyll content was calculated with the following formula:

Chlorophyll fluorescence of 15-day-old seedlings was analyzed by using a FluorCam PAM as described in a previous report (Baker, 2008).

Endogenous ABA quantification

Crude extracts were prepared from leaves (0.1 g FW) of 15-day-old seedlings grown at 200 ppm and 400 ppm in 750 μl of 80:19:1 methanol:H2O:acetic acid buffer supplemented with internal standards for 6 h. After centrifugation, the extracts were filtered through a 0.22 μm filter and dried with N2 at room temperature, and then dissolved in 200 μl of methanol. ABA quantification was performed as previously described (Liu ).

Semi-quantitative PCR and quantitative real-time PCR

Total RNA was extracted from whole seedlings or leaves using TRIzol reagent (Invitrogen Life Technologies), according to the manufacturer’s instructions. The cDNA was reverse transcribed from 1 μg of total RNA using M-MLV reverse transcriptase (Promega) in a reaction buffer [50 mM Tris–HCl (pH 8), 75 mM KCl, 3 mM MgCl2, 10 mM DTT, 0.5 mM dNTP, and 0.5 μg oligo(dT)15]. Quantitative RT-PCR was performed by using a Universal SYBR® Green kit and the C1000 Touch Thermal Cycler real-time PCR detection system (Bio-Rad). The EF1α (AT5G60390) gene was used as a reference for mRNA normalization. The comparative cycle threshold (Ct) method was used to evaluate the relative gene expression levels. The primers used for the expression analysis are listed in Supplementary Table S1.

GUS histochemical analysis

β-Glucuronidase (GUS) histochemical analysis was carried out on transgenic lines expressing ProPPC2::GUS. Plants at various stages including emerging seedling, 7-day-old seedling, 15-day-old seedling, flowers, and siliques were stained in a GUS staining solution and imaged by a Nikon microscope.

Quantification of amino and organic acids

Crude extracts for amino acid determination were prepared from leaves (0.2 g FW) of 15-day-old seedlings in 8% (w/v) 5-sulfosalicylic acid for 1 h followed by centrifugation. The supernatants were filtered through a 0.22 μm filter. The amino acid contents were determined by LC-MS/MS with an Agilent 1290 Infinity II and Agilent 6460 (Kowalski ).

For organic acid determination, crude extracts were obtained from shoots (0.2 g FW) of 15-day-old seedlings grown at 200 ppm and 400 ppm. The samples were homogenized in 3 ml of 7:3 methanol:chloroform (–20 °C) for 2 h. The water-soluble metabolites were extracted from the chloroform phase by the addition of 2.4 ml of H2O. After shaking and centrifugation, the upper methanol–H2O phase was transferred and dried with N2 at room temperature. The extracts were dissolved in 200 μl of H2O and transferred to 0.45 μm cellulose acetate centrifuge tube filters. The determination of organic acid contents was performed by an AB SCIEX QTRAP 6500 Plus LC-MS/MS system as previously described (Ma ).

Luciferase assay

Arabidopsis protoplasts were isolated from 4- to 6-week-old plants following the method used in a previous report (Yoo ). Plasmid DNA (15 µg) was used for polyethylene glycol-calcium transformation (PEG4000), and the protoplast transformation was performed as previously described (Yoo ).

Cellular extracts of Arabidopsis protoplasts after transformation with different constructs were collected for dual-luciferase assays (Hellens ). Cellular extracts (30 μl) were used to detect the firefly and Renilla luciferase activities by a Mithras LB 940 Multimode Microplate Reader. Luciferase activity was normalized to the Renilla activity. All experiments were performed at least three times.

Nuclear and cytoplasmic protein extraction

Arabidopsis protoplasts were collected and homogenized in 100 μl of lysis buffer. After centrifugation, the supernatant for cytoplasmic protein extraction was boiled in SDS–PAGE loading buffer for 2 min. Insoluble nuclei for nuclear protein extraction were re-suspended in 1.5 ml of nuclei resuspension buffer with 0.2% Triton X-100 and then centrifuged. The resuspension was repeated three times (Xu ).

Accession numbers

The gene accession numbers in this study are available at TAIR (The Arabidopsis Information Resource): PPC1 (AT1G53310), PPC2 (AT2G42600), PPC3 (AT3G14940), ABI3 (AT3G24650), ABI4 (AT2G40220), ABI5 (AT2G36270), GGAT1 (AT1g23310), GGAT2 (AT1g70580), SGAT1 (AT2G13360), GLDT1 (AT1G11860), GLDP1 (At4g33010), and SHMT1 (At4g37930).

Results

The ppc2 mutant showed growth arrest under low CO2 conditions

To determine whether the Arabidopsis PEPCs are involved in plant growth regulation under low CO2 conditions, we evaluated the growth performance of T-DNA insertion lines of plant-type PEPCs, ppc1 (Salk_088836) (Feria ), ppc2 (Salk_128516) (Shi ), and ppc3 (Salk_143289) (Feria ), on sucrose-free half-strength Murashige and Skoog (1/2 MS) medium under low CO2 (200 ppm) conditions for 15 d, as well as the control ambient CO2 condition (400 ppm). These single mutants were determined as knockout mutants by our analysis (Fig. 1A, B) and previous studies (Shi ; Feria ). Under ambient CO2 conditions, no obvious morphological differences were observed among the ppc and Col-0 seedlings (Fig. 1C). Under low CO2 conditions, the cotyledons of Col-0, ppc1, and ppc3 turned pale green without obvious morphological differences among them. Interestingly, ppc2 mutant plants exhibited a smaller size and chlorosis when compared with Col-0, ppc1, and ppc3 (Fig. 1C, D). The chlorotic rate was 87.5% in ppc2, and only 10–20% in Col-0, ppc1, and ppc3 (Fig. 1D). These results suggest that PPC2 is required for seedling growth under low CO2 conditions.

Fig. 1.

Growth arrest phenotype of ppc2 mutant seedlings under low CO2 conditions. (A) Genotyping and (B) expression analysis of PPC1, PPC2, and PPC3 in their corresponding single mutants of ppc1, ppc2, and ppc3. ACTIN7 (AT5G09810) was used as a control. (C) Phenotype and (D) statistical analysis of the chlorosis rate of wild-type (WT), ppc1, ppc2, ppc3, and ppc2ppc3 seedlings grown on sucrose-free 1/2 MS medium at 400 ppm or 200 ppm CO2 for 15 d. Data shown are mean ±SEM (n=4). Each replicate has at least 60 seedlings. Asterisks indicate significant differences between genotypes (*P<0.05; ***P<0.005 by Student’s t-test; ns, no significant difference). (E) Expression levels of PPC2 in PPC2-overexpressing ppc2 plants determined by RT-PCR. RNA was extracted from leaves of 15-day-old seedlings. ACTIN7 (AT5G09810) was used as a control. (F) Growth phenotypes of seedlings of the WT, ppc2, and PPC2 complementary lines (COM-1, COM-2, and COM-3) grown on sucrose-free 1/2 MS medium at 400 ppm or 200 ppm CO2 for 15 d. Each replicate has at least 60 seedlings. Scale bars=1 cm in (C) and (F).

To determine whether the growth arrest of ppc2 was due to a defect in seed germination, we assessed the seed germination rates of ppc and Col-0 seeds under low CO2 conditions on sucrose-free 1/2 MS medium. Similar germination rates and status were observed among Col-0, ppc1, ppc2, and ppc3 (Supplementary Fig. S1), revealing that the ppc2 growth arrest phenotype occurred after germination, and PPC2 is involved in seedling development under low CO2 conditions. To determine whether PPC1 or PPC3 has effects on seedling growth under low CO2 conditions in a PPC2-dependent manner, we crossed the ppc1 and ppc3 mutants with ppc2. ppc2ppc3 and ppc1ppc2 double mutants were obtained, but we could not obtain the ppc1ppc2 double mutant seeds due to severe growth arrest, consistent with a previous study (Shi ). The ppc2ppc3 double mutant showed a similar growth phenotype and seed germination to the ppc2 single mutant under low CO2 conditions (Fig. 1C; Supplementary Fig. S1), suggesting that PPC3 possibly is not essential for seedling growth under low CO2 conditions.

To confirm that PPC2 is responsible for the growth retardation observed in ppc2, the PPC2 CDS driven by the Cauliflower mosaic virus (CaMV) 35S promoter was introduced into the ppc2 mutant. The growth arrest and cotyledon chlorotic phenotypes were all rescued in the lines by PPC2 expression (Fig. 1E, F). These results demonstrate that PPC2 is the causal gene, and is a major regulator of seedling growth under low CO2 conditions.

PPC2 encodes a major PEPC in Arabidopsis leaves

The expression levels of these three PEPC genes under low CO2 conditions were determined by qPCR analysis. Only PPC2 was induced by low CO2 (Fig. 2A), consistent with a previous study (Li ). To confirm that PPC2 is a functional PEPC in Arabidopsis, we measured the total PEPC activity in ppc2 and Col-0 under low and ambient CO2 conditions. The ppc2 mutant lost most of its PEPC activity in leaves under both 400 ppm and 200 ppm CO2 conditions (Fig. 2B). Moreover, low CO2 treatment increased the PEPC activity in Col-0 but not in ppc2 leaves (Fig. 2B). These results suggest that PPC2 is a major PEPC in Arabidopsis leaves and is responsive to low CO2.

Fig. 2.

PPC2 is low CO2 inducible and encodes a major PEPC in Arabidopsis leaves. (A) Expression levels of PPC1, PPC2, and PPC3 in 15-day-old seedlings of Col-0 grown on sucrose-free 1/2 MS medium at 400 ppm or 200 ppm CO2. Expression levels are expressed relative to that of EF1α (AT5G60390). Data shown are mean ±SEM (n=3). Asterisks indicate significant differences between genotypes (*P<0.05; ns, no significant difference). (B) Total leaf PEPC activity in 15-day-old wild-type and ppc2 seedlings under ambient and low CO2 conditions. Data shown are mean ±SEM (n=3). Different letters indicate significant differences using Tukey’s test at P≤0.05. (C) Tissue-specific expression of PPC2 in leaves, flowers, and siliques. a–c, GUS activity in leaves of 3-day-old (a), 5-day-old (b), and 15-day-old (c) seedlings; d, guard cells of cotyledon; e and f, GUS activity in calyxes; g, guard cells of calyx; h, GUS activity in siliques; i, guard cells of siliques. Scale bars=50 μm in d and g, and 1 mm in a–c, e, f, h, and i. (D, E) Subcellular localization of PPC2–YFP co-expressed with the ER marker HDEL–mCherry (D) and nuclear-localized FLC–YFP (E) in protoplasts from 4-week-old Col-0. BF, bright field. YFP, yellow fluorescent protein. Scale bar=20 μm. (F) Western blot analyses of PPC2–YFP and FLC–YFP levels in the cytoplasmic and nuclear fractions of Arabidopsis protoplasts expressing PPC2-YFP or FLC-YFP.

We then determined the PPC2 expression pattern by expressing PPC2pro::GUS in Col-0. GUS staining showed that PPC2 expression was high in leaves, hypocotyls, flowers, and siliques, but low in roots (Fig. 2C). At the cellular level, PPC2 was highly expressed in guard cells. The PPC2 subcellular localization was also determined by expressing 35S::PPC2-YFP in Col-0 protoplasts. Yellow fluorescence protein (YFP) fluorescence revealed PPC2 localization in the cytoplasm, nucleus, and also the endoplasmic reticulum (ER; Fig. 2D), which was confirmed by western blot and co-expression with the ER marker HDEL (Fig. 2E, F). All these results demonstrate a specific role for PPC2 in plant growth regulation under low CO2 conditions.

ppc2 seedlings showed reduced carbon assimilation under low CO2 conditions

To determine whether the growth arrest and cotyledon chlorosis of the ppc2 mutant under low CO2 conditions are due to defects in carbohydrate accumulation, we detected the starch content by iodine staining and quantification in 15-day-old ppc2 and Col-0 seedlings at the end of the illumination period (22.00 h) and darkness period (06.00 h) under different CO2 concentrations. Under ambient CO2 conditions, no obvious difference was found in starch accumulation between ppc2 and Col-0 plants (Supplementary Fig. S2A, B). However, under low CO2 conditions, starch accumulation was significantly reduced in ppc2 at both time points (Supplementary Fig. S2A, B). The sucrose content of ppc2 and Col-0 under the same conditions was then measured. ppc2 seedlings had reduced sucrose content under both low and ambient CO2 conditions compared with Col-0 (Supplementary Fig. S2B). Moreover, the synthetic substrates of starch and sucrose, such as G6P, F6P, G1P, ADPG, UDPG, and Suc6P, were also lower in ppc2 under low CO2 conditions, but comparable with Col-0 under ambient CO2 conditions (Supplementary Fig. S2B), consistent with the starch and sucrose contents in ppc2 (Supplementary Fig. S2A, B). These results suggested that PPC2 mutation led to decreased photosynthetic carbohydrate accumulation under low CO2 conditions.

Because ppc2 plants were chlorotic under low CO2 conditions, we measured the chlorophyll and carotenoid contents in ppc2 and Col-0 seedlings. Under ambient CO2 conditions, there were no significant differences in the total chlorophyll and carotenoid contents between ppc2 and Col-0 (Fig. 3A, B). Under low CO2 conditions, both chlorophyll and carotenoid contents were greatly reduced in ppc2. We next determined the maximum quantum yield of PSII (Fv/Fm) in ppc2 and Col-0 seedlings grown at 200 ppm and 400 ppm CO2 using the chlorophyll florescence detector FluorCam FC800. Under ambient CO2 conditions, there was no difference in Fv/Fm between Col-0 and ppc2. Low CO2 led to obvious Fv/Fm decreases in both Col-0 and ppc2, but the decrease was more remarkable in ppc2 than in Col-0 (Fig. 3C, D). These results indicated the involvement of PPC2 in photosynthesis regulation under low CO2 conditions.

Fig. 3.

Carbon assimilation in the ppc2 mutant under low CO2 conditions. (A, B) Chlorophyll content (A) and carotenoid content (B) in 15-day-old seedlings of the wild type (WT) and ppc2 mutant at 400 ppm and 200 ppm CO2. Data shown are mean ±SEM (n=3). (C) Fv/Fm was monitored by Closed FluorCam FC800 in WT and the ppc2 mutant seedlings grown under 400 ppm or 200 ppm CO2 conditions for 15 d. (D) Fv/Fm value comparison between the WT and ppc2 mutant. Data shown are mean ±SEM (n=3). (E, F) Time-resolved stomatal conductance in response to CO2 changes in WT and the ppc2 mutant plants. (E) Data shown in (F) were normalized. Different letters indicate significant differences using Tukey’s test at P≤0.05. Asterisks indicate significant differences between genotypes (***P<0.005 by Student’s t-test; ns, no significant difference).

Because PPC2 was highly expressed in guard cells (Fig. 2C), it was necessary to clarify whether the reduced carbon assimilation rate under low CO2 conditions was induced by compromised low CO2-induced stomatal opening. The CO2 shift from 400 ppm to 100 ppm triggered a dramatic increase in stomatal conductance in both Col-0 and ppc2 mutant plants. However, there were no significant differences in stomatal response to low CO2 between the ppc2 mutant and Col-0 (Fig. 3E, F), indicating a minor role for PPC2 in regulating stomatal opening. This also suggested that the reduced carbon assimilation in ppc2 at low CO2 conditions was not caused by less CO2 uptake, but by defects in CO2 utilization.

ppc2 mutant seedlings showed growth arrest under mild drought stress conditions

Drought stress promotes stomatal closure, and thus decreases plant CO2 uptake and Ci (Parry ). We then determined whether ppc2 had phenotypes under water shortage similar to those under low CO2 conditions. When plant watering was reduced, ppc2 exhibited growth arrest, while under normal growth conditions no significant differences were observed between ppc2 and Col-0 (Fig. 4A). Leaf temperature is an indicator of stomatal status (Jones, 1999). We found that drought stress increased leaf temperature in both ppc2 and Col-0 compared with normal growth conditions, but no significant temperature differences between them were found under both normal and drought stress conditions (Supplementary Fig. S4). These results further support that drought stress promotes stomatal closure, and that ppc2 plants retain normal stomatal responses. Furthermore, the ppc2 mutant exhibited a reduced Fv/Fm value under mild drought stress conditions (Fig. 4B), similar to low CO2 conditions. These results further suggest that PPC2 is required for CO2 utilization and plant growth under low CO2 conditions caused by stresses such as drought.

Fig. 4.

Growth arrest phenotype of the ppc2 mutant seedlings under mild drought stress conditions. (A) Morphological phenotype and (B) Fv/Fm monitored by Closed FluorCam FC800 of 10-day-old WT and ppc2 mutant seedlings grown under normal and mild drought stress conditions for 10 d. Each pot had 64 seedlings. Scale bar=2 cm in (A) and (B).

Application of exogenous sucrose or malate greatly rescued ppc2 seedling growth arrest under low CO2 conditions

Considering the reduced photosynthetic carbohydrate accumulation and carbon assimilation in the ppc2 mutant under low CO2 conditions, we further determined whether sucrose application could rescue the retarded growth of ppc2 under low CO2 conditions. Interestingly, exogenous sucrose (25 mM) treatment completely rescued the seedling growth retardation and cotyledon chlorosis in ppc2 under low CO2 conditions (Fig. 5A).

Fig. 5.

Exogenous ABA partially rescues ppc2 seedling growth arrest under low CO2 conditions. (A) Phenotype of 15-day-old seedlings of wild-type (WT) and the ppc2 mutant under 200 ppm CO2 conditions on sucrose-free 1/2 MS medium supplemented with mock (ddH2O), 25 mM sucrose, 1.5 mM malate, or 3 mM malate. Each replicate has at least 60 seedlings. (B) ABA contents in WT and the ppc2 mutant seedlings grown under different CO2 concentrations. Data shown are mean ±SEM (n=3). (C) Phenotype and (D) statistical analysis of chlorosis rate of 15-day-old WT and the ppc2 mutant seedlings grown at 200 ppm CO2 on sucrose-free 1/2 MS medium with mock (ddH2O), 0.1 μM ABA, and 0.2 μM ABA, respectively. Data shown are mean ±SEM (n=4). Each replicate has at least 60 seedlings. Asterisks indicate significant differences between genotypes (**P<0.01 by Student’s t-test). Different letters indicate significant differences using Tukey’s test at P≤0.05. Scale bar=1 cm in (A) and (C).

PEPC catalyzes oxaloacetate (OAA) synthesis from PEP and HCO3−. OAA is then rapidly converted into malate by malate dehydrogenase. Our primary metabolic analysis showed that PPC2 mutation led to a decrease in malate (Supplementary Fig. S2B), particularly under low CO2 conditions. We then explored whether the ppc2 mutant growth arrest phenotype under low CO2 conditions was caused by, or at least contributed by, malate defect. Application of 1.5 mM malate did not have any impact on the growth of either ppc2 or Col-0, and also could not eliminate the ppc2 cotyledon chlorosis phenotype (Fig. 5A). However, 3 mM malate greatly relieved the ppc2 growth arrest phenotype compared with controls under our growth conditions. These results support PPC2 functioning in both carbon assimilation and metabolic pathway.

Exogenous ABA partially rescued ppc2 seedling growth arrest under low CO2 conditions

ABA is synthesized from C40 oxygenated carotenoids (Ruiz-Sola and Rodriguez-Concepcion, 2012). The carotenoid content was noticeably reduced in ppc2 under low CO2 concentrations (Fig. 3B). To determine whether ABA biosynthesis is blocked in ppc2 under low CO2 conditions, we quantified the ABA level in ppc2 and Col-0 seedlings grown under low and ambient CO2 conditions by UFLC-ESI-MS (Liu ). Interestingly, the ABA content in the ppc2 mutant was greatly reduced under low CO2 conditions, but increased under ambient CO2 conditions compared with that in Col-0 (Fig. 5B). It can be speculated that the decrease in ABA level might be a cause of the ppc2 growth arrest under low CO2 conditions.

To prove this, exogenous ABA (0.1 μM or 0.2 μM) was added in sucrose-free 1/2 MS medium plates to observe the growth performance of Col-0 and ppc2 under low CO2 conditions. ABA treatment inhibited the growth and increased the chlorotic seedling ratio in Col-0, while, in the ppc2 mutant, ABA treatment largely recovered the seedling growth arrest phenotype (Fig. 5C, D). Collectively, our results demonstrate that PPC2 mutation affects ABA biosynthesis, and the ppc2 seedling growth arrest phenotype is at least partly due to the decrease in ABA level.

ABI5 overexpression rescued ppc2 seedling growth arrest under low CO2 conditions

It has been reported that ABI3, ABI4, and ABI5 (the downstream targets of the ABA signaling pathway) are required for ABA modulation of seed germination and post-germination development (Finkelstein and Lynch, 2000; Lopez-Molina ). To determine whether ABI transcription factors function in ppc2 seedling growth arrest under low CO2 conditions, we first determined their expression levels by qPCR. Low CO2 treatment induced ABI3 and ABI4 expression in both Col-0 and the ppc2 mutant. However, ABI5 induction by low CO2 conditions was greatly suppressed in ppc2 (Fig. 6A). Furthermore, exogenous ABA rescued ABI5 expression in the ppc2 mutant under low CO2 conditions (Fig. 6B). These results indicated that the decrease in ABI5 expression might be a major cause of ppc2 growth arrest under low CO2 conditions. To confirm this possibility and reveal the role of ABI5 in seedling growth under low CO2 conditions, we evaluated the growth performance of the abi5-1 mutant and wild-type Wassileskija (Ws) plants under low CO2 conditions on sucrose-free medium. No obvious morphological differences were observed between Ws and abi5-1 seedlings (Fig. 6C). However, the Fv/Fm value of abi5-1 was significantly less than that of Ws under low CO2 conditions (Fig. 6D), indicating that ABI5 mutation led to reduction of maximum quantum efficiency of PSII under low CO2 conditions. We further overexpressed ABI5 in ppc2 driven by the constitutive CaMV 35S promoter. ABI5 overexpression greatly rescued the seedling growth arrest phenotype of three randomly selected ppc2 transgenic lines under low CO2 conditions (Fig. 6E, F; Supplementary Fig. S3).

Fig. 6.

ABI5 overexpression greatly rescues ppc2 growth arrest under low CO2 conditions. (A) Expression levels of ABI3, ABI4, and ABI5 relative to that of EF1α (AT5G60390) in the wild type (WT) and the ppc2 mutant grown on sucrose-free 1/2 MS medium at 400 ppm or 200 ppm CO2 for 15 d. Data are shown as mean ±SEM (n=3). (B) ABI5 expression levels in ppc2 and WT seedlings treated with mock (ddH2O), 0.1 μM ABA, and 0.2 μM ABA at 200 ppm, respectively. Data shown are mean ±SEM (n=3). Different letters indicate significant differences using Tukey’s test at P≤0.05. (C) Fv/Fm monitored by Closed FluorCam FC800 in Ws and the abi5-1 mutant seedlings grown at 400 ppm or 200 ppm CO2 for 15 d. (D) Maximum photosynthetic yields (Fv/Fm) of Ws and the abi5-1 mutant at different CO2 concentrations. Data shown are mean ±SEM (n=3). Asterisks indicate significant differences between genotypes (*P<0.05 by Student’s t-test; ns, no significant difference). (E) Phenotype and (F) statistical analysis of the chlorosis rate of the WT, the ppc2 mutant, and ABI5-overexpressing ppc2 transgenic lines (ppc2ABI5oe-7, ppc2ABI5oe-8, and ppc2ABI5oe-9). Data shown are mean ±SEM (n=3). Each replicate has at least 60 seedlings. Different letters indicate significant differences using Tukey’s test at P≤0.05. Scale bar=1 cm in (C) and (E).

Photorespiratory intermediates were increased in ppc2 under low CO2 conditions

It has been suggested that PEPC activity is linked to photorespiration by supplying malate to the TCA cycle to sustain glutamate and glutamine metabolism (Masumoto ; Shi ). We then determined amino acid contents in ppc2 and Col-0 under both ambient and low CO2 conditions. Under low CO2 conditions, the ppc2 mutant exhibited reduced glutamate but increased glutamine, leading to increases in the glutamine to glutamate ratio (Fig. 7A, B) and the levels of β-alanine and arginine, with decreases in the levels of alanine, asparatic acid, and proline (Supplementary Fig. S5). Glycine and serine levels have been recognized as indicators of carbon flux through photorespiration, and a higher glycine/serine ratio indicates higher photorespiration (Novitskaya ). Low CO2 concentrations would increase photorespiration. The ppc2 mutant had greater levels of the photorespiratory intermediates glycine and serine under photorespiratory (low CO2, 21% O2) conditions (Fig. 7A), and greater glycine levels at 400 ppm CO2 (Supplementary Fig. S5), consistent with the reduced photosynthesis and growth arrest phenotypes of ppc2 under low CO2 conditions (Figs 3E, 5A). We found that low CO2 (photorespiratory conditions) triggered similar increases in the glycine/serine ratio in both ppc2 and Col-0 compared with ambient CO2 conditions (Fig. 7C). Under ambient CO2 conditions, PPC2 mutation did not show significant effects on amino acid and organic acid contents, and only glycine, valine, and tyrosine were slightly increased (Supplementary Figs S2B, S5). These results suggest that PPC2 functions in both primary metabolism and photorespiratory metabolism under low CO2 conditions through modulation of the carbon–nitrogen balance.

Fig. 7.

Metabolite analysis of the ppc2 mutant and WT under low CO2 conditions. (A) Differences in leaf metabolite levels between the ppc2 mutant and wild-type (WT) seedlings under 200 ppm CO2 conditions. PEPC, phosphoenolpyruvate carboxylase. GGAT, glutamate:glyoxylate aminotransferase. GLD, glycine decarboxylase, including GLDP, GLDT, and GLDH. SHMT, serine hydroxymethyltransferase. SGAT, serine:glyoxylate aminotransferase. RuBP, ribulose-1,5-disphosphate. F6P, fructose 6-phosphate. G6P, glucose 6-phosphate. G1P, glucose 1-phosphate. UDPGlc, UDP-glucose. ADPGlc, ADP-glucose. Suc6P, sucrose 6-phosphate. Pyr, pyruvate. 2-OS, 2-oxosuccinamate. OAA, oxaloacetate. PEP, phosphoenlpyruvate. (B) Glutamine to glutamate ratio in 15-day-old WT and ppc2 seedlings at different CO2 concentrations. Data shown are mean ±SEM (n=3). (C) Ratio of glycine to serine ratio in 15-day-old WT and ppc2 seedlings at different CO2 concentrations. Data shown are mean ±SEM (n=3). Different letters indicate significant differences using Tukey’s test at P≤0.05.

ABI5 regulated photorespiratory enzyme expression levels

In the photorespiratory pathway, GGAT (glutamate:glyoxylate aminotransferase) transfers -NH3+ from glutamate into glyoxylate to generate glycine, and SGAT1 (serine:glyoxylate aminotransferase) transfers -NH3+ from serine, alanine, and asparagine into glyoxylate to produce glycine. GLDP1 and GLDT1 are components of the glycine decarboxylase complex, which catalyzes glycine into CH2-THF. To further explore whether the greater serine and glycine levels in ppc2 under low CO2 conditions were caused by enzyme expression changes involved in glycine and serine synthesis and metabolism during photorespiration, we checked the expression levels of GGAT1, GGAT2, SGAT1, GLDP1, GLDT1, and SHMT1 (Peterhansel ) in ppc2 under both low and ambient CO2 conditions. Except for GGAT2, the expression levels were significantly reduced in ppc2 under low CO2 conditions (Fig. 8A). Among them, GGAT1 and SGAT1 were slightly induced by low CO2 in Col-0 (Fig. 8A). SHMT1 expression in ppc2 was significantly reduced under both ambient and low CO2 conditions (Fig. 8A).

Fig. 8.

ABI5 regulates the expression levels of major photorespiratory enzymes related to glycine and serine synthesis and metabolism. (A) Expression levels of photorespiratory enzyme genes in wild-type (WT) and the ppc2 mutant leaves. RNAs were extracted from the leaves of 15-day-old seedlings. EF1α (AT5G60390) was used as an internal control. Data shown are mean ±SEM (n=3). (B) Expression levels of photorespiratory genes in WT, ppc2 mutant, and ABI5-overexpressing ppc2 plants (ppc2ABI5oe-7, ppc2ABI5oe-8, and ppc2ABI5oe-9). RNAs were extracted from the leaves of 15-day-old seedlings at different CO2 concentrations. EF1α (AT5G60390) was used an internal control. Data shown are mean ±SEM (n=3). Different letters indicate significant differences using Tukey’s test at P≤0.05. (B) Expression levels of photorespiratory genes in the WT (Ws) and abi5-1 mutant leaves. RNAs were extracted from the leaves of 15-day-old seedlings. EF1α (AT5G60390) was used as an internal control. Data shown are mean ±SEM (n=3). (D) A schematic representation of the dual-luciferase reporter system. ABI5 or GFP (control) driven by CaMV 35S as effector was co-transformed with reporters, REN (Renilla luciferase) driven by 35S and LUC (firefly luciferase) driven by the promoter regions of photorespiratory genes. (E) ABI5 activation of SGAT1, GLDT1, and SHMT1 by dual-luciferase reporter assays. LUC values were normalized to those of REN. Data shown are mean ±SEM (n=3). Asterisks indicate significant differences between genotypes (**P<0.01 by Student’s t-test; ns, no significant difference).

ABI5 is a transcription factor that directly binds to the promoter regions of its targets to activate their expression. ABI5 expression was reduced in ppc2, and ABI5 overexpression rescued the ppc2 growth arrest phenotype under photorespiratory conditions. Thus, we speculated that ABI5 might regulate enzyme expression levels that function in glycine and serine metabolism, such as GGAT1, SGAT1, GLDP1, GLDT1, and SHMT1. Interestingly, there were several ABREs in their promoter regions (Supplementary Fig. S6). SGAT1, GLDT1, and SHMT1 expression was reduced in abi5-1 under low CO2 conditions, and their expression in ppc2 was completely or greatly recovered by ABI5 overexpression (Fig. 8B, C), suggesting that these three genes may be the targets of ABI5. We then performed dual-luciferase assays to determine whether ABI5 activates the promoters of SGAT1, GLDT1, and SHMT1 that drive LUC expression in Arabidopsis protoplasts. ABI5 expression greatly activated SGAT1 and GLDT1 expression, but could not activate SHMT1 expression (Fig. 8D, E), demonstrating that ABI5 regulates photorespiration by modulating photorespiratory enzyme expression levels, and SGAT1 and GLDT1 are the direct targets, and SHMT1 is an ABI5 indirect target.

Carbon assimilation was reduced in ppc2 mature plants under photorespiratory conditions

Recently, a potential role for PEPC in C3 plant metabolism under high photorespiratory (low CO2, 21% O2) conditions was proposed (Abadie and Tcherkez, 2019; Tcherkez and Limami, 2019). Here, we also found that PPC2 is involved in seedling development by modulating photorespiratory metabolism under low CO2 conditions. To further investigate PPC2 function under photorespiratory conditions, we determined the CO2 assimilation rate under different Ci conditions in ppc2 and Col-0 mature leaves under ambient air conditions. Compared with Col-0, the ppc2 mutant exhibited a lower CO2 assimilation rate under low CO2 (50–400 ppm) concentrations, and displayed a similar CO2 assimilation rate under high CO2 (400–800 ppm) concentrations as observed in the A–Ci curve (Fig. 9A). In addition, no significant difference was observed in the maximum photosynthetic electron transport rates (ETRs) calculated from the A–Ci curves between Col-0 and the ppc2 mutant (Fig. 9B), indicating that ppc2 mutation did not alter the photosynthetic capacity. The reduction in the initial A–Ci curve was recovered when the measurements were performed under very low oxygen conditions that restricted photorespiration (Fig. 9C). These results indicated that the altered CO2 assimilation of ppc2 under low CO2 conditions was associated with the simultaneous photorespiratory CO2 loss. Moreover, PPC2 expression rescued the reduced photosynthetic rate in response to low Ci conditions in ppc2 (Fig. 9D). Together with the fact that PEPC is an important enzyme in the glycolytic pathway that links with photorespiration and respiration in plants, our results suggest that PPC2 is involved in photorespiration under relatively low CO2 conditions. The phenotype of ppc2 is at least partially due to the reduction of net carbon assimilation, which may result from the low capacity to utilize the photorespiratory metabolites under relatively low CO2 conditions when PPC2 is mutated.

Fig. 9.

Carbon assimilation of the abi5-1 mutant and the effect of ABI5 overexpression on carbon assimilation of the ppc2 mutant under low CO2 conditions. (A, C) A–Ci curves of 30-day-old wild type (WT) and the ppc2 mutant plants under different CO2 conditions balanced with air (A) or low oxygen conditions (C). The light intensity for the measurement was set at 2000 μmol m−2 s−1. Data shown are mean ±SEM (n=3). (B) Maximum photosynthetic electron transport rate (ETR) of 30-day-old WT and ppc2 mature plants. Data shown are mean ±SEM (n=5). (D) PPC2 expression complements the reduced carbon assimilation in the ppc2 rosette leaves grown under 400 ppm CO2 conditions. COM-1, COM-2, and COM-3 are PPC2 complementary lines. Data shown are mean ±SEM (n=3). (E) Maximum photosynthetic ETR of 30-day-old Ws and abi5-1 mutant plants. Data shown are mean ±SEM (n=5). (F, G) A–Ci curves of 30-day-old abi5-1 and Ws under different CO2 conditions balanced with air (F) or low oxygen (G). Data shown are mean ±SEM (n=3). (H) ABI5 overexpression complements the reduced carbon assimilation of ppc2 plants grown under 400 ppm CO2 conditions. Data shown are mean ±SEM (n=3). (I) Dry weight of WT, ppc2, and ABI5-overexpressing ppc2 transgenic lines (ppc2ABI5oe-7, ppc2ABI5oe-8, and ppc2ABI5oe-9) grown under ambient CO2 conditions. Data shown are mean ±SEM (n=6). Asterisks indicate significant differences between genotypes (*P<0.05; **P<0.01; ***P<0.005 by Student’s t-test; ns, no significant difference). Different letters indicate significant differences using Tukey’s test at P≤0.05.

ABI5 overexpression rescued the reduced CO2 assimilation in mature ppc2 plants under photorespiratory conditions

We also determined the A–Ci curves of 30-day-old abi5-1 plants under ambient air and low oxygen conditions. Under ambient air conditions, ABI5 mutation greatly reduced the CO2 assimilation rate under low CO2 concentrations (50–400 ppm), but not under high CO2 concentrations (600–800 ppm) compared with Ws (Fig. 9F). The maximum photosynthetic ETRs inferred from the A–Ci curves were similar between Ws and abi5-1 under ambient air conditions (Fig. 9E). Under low oxygen conditions, the slopes of the A–Ci curve showed no significant difference between abi5-1 and Ws (Fig. 9G). These results demonstrate that ABI5 is involved in photorespiratory metabolism.

We next determined the A–Ci curves of ABI5-overexpressing ppc2 plants grown under ambient CO2 conditions. The reduced CO2 assimilation of ppc2 at low CO2 was completely rescued by ABI5 overexpression (Fig. 9H). Consistent with this, ABI5 overexpression restored and even slightly increased the ppc2 mutant reduced dry weight biomass under ambient conditions compared with Col-0 (Fig. 9I).

Discussion

PPC2 is essential for plant acclimation to low CO2

CO2 is the major source for photosynthesis and is pivotal for plant growth. High CO2 usually increases plant growth and reproduction, whereas low CO2 decreases plant growth by changing some physiological responses (e.g. water use efficiency), reducing biomass production, and delaying development (Gerhart and Ward, 2010). However, the mechanisms underlying the effects of low CO2 concentrations on plants are still unclear. Here, we report that Arabidopsis PPC2, which encodes a PEPC involved in plant primary metabolism for production of C4-dicarboxylic acids, is essential for plant growth under low CO2 conditions. ppc2 seedlings showed chlorosis and growth arrest under low CO2 (200 ppm) conditions (Fig. 1C, D), which could be rescued by PPC2 expression (Fig. 1E, F). Moreover, there were no significant differences in the germination rate between ppc2 and Col-0 under low CO2 conditions (Supplementary Fig. S1), suggesting that the phenotypes of the ppc2 mutant occurred at the seedling development stage. Compared with Col-0, ppc2 mutant seedlings accumulated less photosynthetic carbohydrates in leaves, such as sucrose, starch, and their upstream precursors (Supplementary Fig. S2), and exogenous sucrose or malate application greatly recovered the ppc2 growth arrest phenotype (Fig. 5A). PPC2 mutation greatly decreased the total leaf PEPC activity under low CO2 conditions (Fig. 2B; Supplementary Fig. S2B). Furthermore, among these three Arabidopsis plant-type PEPCs, only PPC2 was induced by low CO2, consistent with the previous study (Li ), and PPC1 or PPC3 mutation did not affect plant growth at low CO2 (Figs 1C, 2A) (Li ; Shi ; Feria ). All these results suggest that PPC2 is essential for plant growth under low CO2 conditions, and is the major PEPC in Arabidopsis leaves.

PPC2 participates in photorespiration by linking with primary metabolism under photorespiratory conditions

Recent studies in sunflower have shown that malate content closely correlates with photorespiration by metabolomics analysis, and C3 PEPC CO2 fixation increases under high photorespiratory conditions (low CO2, 21% O2) (Abadie ; Abadie and Tcherkez, 2019), indicating that non-photosynthetic PEPC may still contribute to photorespiration in C3 plants. Studies in Arabidopsis have reported the novel function of C3 PEPC in regulating the balance of carbon–nitrogen metabolism (Masumoto ; Shi ). However, the molecular mechanisms remain elusive. Here, we for the first time clarified the special role of PPC2 in photorespiration by regulating the carbon–nitrogen balance under low CO2 conditions.

First, under photorespiratory conditions, PPC2 may regulate carbon–nitrogen balance. Glycolysis pathway metabolites and amino acid levels were greatly affected (Fig. 7A; Supplementary Figs S2B, S4), and the photorespiratory intermediates glycine and serine were significantly accumulated in the ppc2 mutants (Fig. 7A; Supplementary Fig. S5), suggesting that PPC2 is involved in photorespiratory metabolism under low CO2 conditions. Glutamate, which plays a central signaling and metabolic role in regulating carbon and nitrogen assimilatory pathways (Forde and Lea, 2007), was decreased in ppc2 under photorespiratory conditions. The increase in glutamine further reduced ammonium assimilation released by glycine decarboxylation in photorespiration, and thus probably contributed to glycine accumulation at low CO2. Exogenous malate application not only greatly rescued the growth arrest of ppc2 under low CO2 conditions (Fig. 5A), but also reduced the high accumulation of glycine in the ppc1/ppc2 double mutant (Shi ). These phenomena are in accordance with the previous report that malate dehydrogenase mutants exhibit an alteration in photorespiratory metabolism (Tomaz ). Because of PPC2 mutation, PEP flux into the glycolysis pathway was reduced, while that into the shikimate pathway was increased under low CO2 conditions, leading to greater phenylalanine, tyrosine, and tryptophan levels (Fig. 7A; Supplementary Fig. S5). These results demonstrate that PPC2 controls carbon and nitrogen metabolism balance under low CO2 conditions.

Secondly, under photorespiratory conditions, PPC2 affects the expression patterns of photorespiratory enzyme related to glycine and serine synthesis and metabolism. In the photorespiratory pathway, GGATs and SGAT1 transfer -NH3+ from glutamate and serine, respectively, into glyoxylate to synthesize glycine; GLDP1 and GLDT1 decarboxylate glycine into CH2-THF; and subsequently SHMT1 transfers the C1 moiety to another glycine to result in serine formation (Peterhansel ). Mutation of either SGAT1 or SHMT1 leads to serine and glycine accumulation (Somerville and Ogren, 1980; Kuhn ). Furthermore, glycine decarboxylase overexpression resulted in lower glycine and serine contents (Timm ). In agreement with the above results, SHMT1, SGAT1, GLDP1, and GLDT1 expression was significantly reduced in ppc2 under low CO2 conditions, which increased glycine and serine levels.

Thirdly, decreased CO2 assimilation in ppc2 under low CO2 conditions was caused by increased photorespiratory carbon loss. The ppc2 mutant exhibited decreased CO2 assimilation at relatively lower CO2 concentrations (50–400 ppm) (Fig. 9A). However, after reducing the O2 concentration to inhibit photorespiration, the reduced CO2 assimilation in ppc2 under low CO2 conditions was recovered (Fig. 9C). Moreover, our results showed that the reduced carbon assimilation of ppc2 during low CO2 conditions was not related to the stomatal status, because the stomatal conductance at the steady state and in response to the low CO2 shift remained normal in ppc2 (Fig. 3E, F). In this sense, PPC2 is specifically involved in photorespiration under low CO2 conditions (Fig. 9D).

ABA regulates photorespiration under low CO2 conditions through ABI5

ABA induces gene expression and plays a prominent role in establishment of stress tolerance. However, little is known about the relationships of ABA to low CO2 stress and the simultaneous photorespiration. In this study, we found that ABA biosynthesis was heavily blocked under low CO2 conditions, possibly due to the decrease in accumulation of carotenoids (Fig. 3B), which are ABA biosynthesis precursors (Ruiz-Sola and Rodriguez-Concepcion, 2012). PPC2 mutation further reduced ABA synthesis (Fig. 5B). Application of a small amount of ABA largely recovered ppc2 growth (Fig. 5C, D), suggesting that low ABA concentration is required for stimulating photosynthesis and plant growth under low CO2 conditions. In addition, it has been reported that exogenous ABA induces the photorespiratory rate in barley by increasing GOX (glycolate oxidase) activity (Popova ). Our data also showed that high ABA treatment led to leaf chlorosis in Col-0 (Fig. 5C, D). These results suggest that a greater ABA concentration could reduce photosynthesis under low CO2 conditions, possibly by promoting carbon flux through the photorespiratory cycle.

ABI5 has been proposed as a key player in monitoring environmental conditions during seedling growth (Lopez-Molina ), and to function as an intermediate in ABA signaling to regulate seed germination and seedling growth. Our results show that ABI5 plays a key role in seedling growth under photorespiratory conditions by regulating photorespiratory enzyme expression, and ABA regulates plant growth under low CO2 conditions through modulating ABI5 expression. abi5-1 seedlings grown at low CO2 showed lower Fv/Fm values (Fig. 6C, D). CO2 assimilation in the mature abi5-1 mutant was reduced under low CO2 conditions, and could be fully recovered by non-photorespiratory (low O2) conditions (Fig. 9F, G). In the ppc2 plants, ABA synthesis was reduced and ABI5 was repressed by low CO2 conditions, while ABA treatment greatly rescued ABI5 expression and ppc2 growth arrest (Figs 5D, 6B). Moreover, the expression levels of photorespiratory enzymes related to glycine and serine metabolism decreased in ppc2 and abi5-1 seedlings under low CO2 conditions (Fig. 8A, C). Further experiments showed that ABI5 may regulate SGAT1 and GLDT1 by directly binding to their promoters, and regulates SHMT1 indirectly (Fig. 8E). ABI5 overexpression rescued the expression of the above genes (Fig. 8C) and the reduction of dry weight biomass in ppc2 mature plants (Fig. 9H, I).

In summary, our study reveals that PPC2 is essential for plant acclimation to low CO2 and plays an important role in carbon assimilation via regulating carbon and nitrogen metabolism balance under photorespiratory low CO2 conditions. We also identified the important role of ABA in photorespiration and the novel function of ABI5 in regulating expression of photorespiratory enzyme under low CO2 conditions. Our work demonstrates the key role of C3-PEPC PPC2 in photorespiration under low CO2 conditions, and may offer clues for future studies to understand the mechanism of C3 PEPCs in photosynthesis regulation and for potential application in crop improvement against photorespiration.

Supplementary data

Supplementary data are available at JXB online.

Fig. S1. Seed germination of the wild type and ppc mutants.

Fig. S2. Photosynthetic carbohydrates were reduced in the ppc2 mutant seedlings under low CO2 conditions.

Fig. S3. RT-PCR analysis of ABI5 expression level in WT, ppc2, and ABI5-overexpressing ppc2 plants.

Fig. S4. Leaf temperature of WT and ppc2 mutant seedlings by thermal imaging under normal and mild drought stress conditions.

Fig. S5. Amino acid contents in WT and ppc2 mutant seedlings.

Fig. S6. Predicted ABRE cis-elements in the promoter regions of photorespiratory genes.

Table S1. List of primers used in this study.

Click here for additional data file.