Suppression of OsMDHAR4 enhances heat tolerance by mediating H2O2-induced stomatal closure in rice plants

Background Monodehydroascorbate reductase (MDAR or MDHAR), which is responsible for growth, development and stress response in plants, is a key enzyme in the maintenance of the ascorbate acid (AsA) pool through the AsA–glutathione (AsA–GSH) cycle. High temperature affects a broad spectrum of cellular components and metabolism including AsA–GSH cycle in plants. In rice, however, the detailed roles of OsMDHAR4 in resistance against heat stress remains unclear. Results Here, we report that OsMDHAR4 protein was localized to the chloroplasts. OsMDHAR4 expression was detected in all tissues surveyed and peaked in leaf blade. OsMDHAR4 was responsive to multiple stresses and was relatively strongly induced by heat treatment. In comparison with wild type, the osmdhar4 mutant exhibited improved tolerance to heat stress, whereas OsMDHAR4 overexpression lines exhibited enhanced sensitivity to heat stress. Moreover, we found that suppression of OsMDHAR4 promoted stomatal closure and hydrogen peroxide accumulation, and overexpression of OsMDHAR4 increased stomatal opening and decreased hydrogen peroxide content in rice leaves. Conclusions Taken together, these results indicated that OsMDHAR4 negatively regulates tolerance to heat stress by mediating H2O2-induced stomatal closure in rice. Electronic supplementary material The online version of this article (10.1186/s12284-018-0230-5) contains supplementary material, which is available to authorized users.


Background
Extreme temperature is one of the serious threats affecting crop production and distribution worldwide (Yokotani et al., 2008). In plants, a transient increase of 10-15°C above ambient, is generally considered as heat shock or heat stress, which negatively affects plant growth, seed germination, photosynthesis, respiration, water relation, and membrane stability in plants (Wahid et al., 2007), and is often accompanied by the generation of reactive oxygen species (ROS), such as hydrogen peroxide (H 2 O 2 ), hydroxyl radical, superoxide anion radicals, and singlet oxygen (Liu and Huang, 2000;Mittler, 2002;Apel and Hirt, 2004;Song et al., 2014). The understanding of plant responses to heat stress in physiology, genetics, and molecular biology will be greatly helpful in improving the heat tolerance of plants through genetic engineering.
Compared with other ROS, H 2 O 2 is more stable, more diffusive, and have a long half-life (approximately 1 ms) and high permeability across membranes (Levine et al., 1994), so it can readily escape from the organelle where it was produced to the cytosol. Previous research has shown that H 2 O 2 plays a dual role in plants: at low concentrations, it acts as a secondary messenger involved in triggering tolerance to various biotic and abiotic stresses; but at high concentrations, it leads to programmed cell death (Quan et al., 2008). H 2 O 2 also acts as a key regulator in a broad range of physiological processes, such as photorespiration and photosynthesis (Noctor and Foyer, 1998), growth and development (Foreman et al., 2003), the cell cycle (Mittler et al., 2004), senescence (Peng et al., 2005), and stomatal movement (Bright et al., 2006).
Stomatal pores that are located in the epidermis of plant leaves control the uptake of CO 2 for photosynthesis and the water loss during transpiration, and play a crucial role in abiotic stress tolerance (Schroeder et al., 2001;Hetherington and Woodward, 2003). In the process of ABA-dependent stomatal closure, H 2 O 2 plays a vital role as a signal molecule by elevating calcium levels in guard cells through the activation of plasma membrane calcium channels (Pei et al., 2000;Wang and Song, 2008). OsH-TAS, a RING finger ubiquitin E3 ligase, functions in rice heat tolerance through H 2 O 2 -induced stomatal closure and is mainly involved in the ABA-dependent pathway (Liu et al., 2016). The ABA-activated SnRK2 protein kinase OPEN STOMATA1 phosphorylates NADP (NADPH) oxidase (AtrbohF), which functions to produce ABA-induced ROS in guard cells (Sirichandra et al., 2009). There's another way of ABA-independent stomatal closure. DROUGHT AND SALT TOLERANCE (DST), a zinc finger transcription factor, negatively regulates H 2 O 2 -induced stomatal closure by directly regulating the expression of genes related to H 2 O 2 scavenging (Huang et al., 2009). OsS-RO1c suppresses DST to positively regulate H 2 O 2 -induced stomatal closure (You et al., 2013). ROS is normally scavenged by an enzymatic anti-oxidative system containing catalase (CAT), ascorbate peroxidase (APX), glutathione peroxidase (GPX) and superoxide dismutase (SOD), and by a non-enzymatic anti-oxidative system including ascorbic acid (AsA), glutathione (GSH), tocopherols (TOCs) and phenolic compounds to protect plant cells. The monodehydroascorbate reductase (MDAR or MDHAR), well-known as flavin adenine dinucleotide (FAD) enzyme, is involved in the ascorbate-glutathione cycle, and plays an important role in directly reducing monodehydroascorbate (oxidized ascorbate) to ascorbate using NAD(P)H as an electron donor (Apel and Hirt 2004). MDARs were found in many eukaryotes, including cucumbers, potatoes, soybean root nodules, and rot fungus (Gill and Tuteja, 2010), and were localized in chloroplasts, mitochondria, peroxisomes, and the cytosol in plants (Omoto et al., 2013), and in microsomes, mitochondria, the Golgi apparatus, and erythrocytes in animals (Sakihama et al., 2000).
The roles of MDAR have been extensively reported under abiotic and biotic stress. Overexpression of AtMDAR1 in tobacco confers enhanced tolerance to ozone, salt and polyethylene glycol stresses (Eltayeb et al., 2007). Overexpression of LeMDAR from tomato (Lycopersicon esculentum Mill.) enhanced tolerance to temperature and methylviologen-mediated oxidative stresses (Li et al., 2010).
In this study, a new gene encoding the chloroplastic MDHAR (OsMDHAR4) from the rice plant was cloned, and its function was analyzed. The osmdhar4 mutant displayed significantly improved heat tolerance with increased stomatal closure and reduced water loss by promoting H 2 O 2 accumulation. OsMDHAR4-overexpressing transgenic plants showed obviously enhanced sensitivity to heat stress with decreased stomatal closure and accelerated water loss speed. Our results suggest that the OsMDHAR4 gene negatively regulates resistance to heat stress in rice.
To study the subcellular localization of OsMDHAR4 protein, we fused the CDS of OsMDHAR4 to the N-terminal of yellow fluorescent protein (YFP) driven by the cauliflower mosaic virus 35S promoter. The empty and recombinant vectors were transiently transformed into the epidermis of Nicotiana benthamiana by agroinfiltration. Fluorescence microscopic analysis revealed that OsMDHAR4-YFP was mainly detected in chloroplast, partially colocalized with the red autofluorescence of chlorophyll. On the contrary, 35S-YFP was widely distributed in the nucleus and cytoplasm and was not colocalized with the red autofluorescence of chloroplasts (Fig. 1). This result indicated that OsMDHAR4 protein was localized to the chloroplasts.

Expression pattern of OsMDHAR4
A total of 6 representative tissues, including root, shoot, stem, sheath and blade of flag leaf, and panicle, were prepared for spatio-temporal expression analysis of OsMDHAR4 in rice Zhonghua11 (ZH11). The mRNA abundance of OsMDHAR4 was detected using quantitative RT-PCR (qRT-PCR). OsMDHAR4 transcripts were detected in all of the tissues surveyed, and the expression level was relatively lowest in root and peaked in the blade of the flag leaf (Fig. 2a).
To investigate the physiological and functional relevance of the OsMDHAR4 gene, we further examined the expression level of OsMDHAR4 under heat stress and other treatments, including polyethylene glycol (PEG), abscisic acid (ABA), H 2 O 2 , salt and cold, at the seedling stage. It was found that, after heat treatment, the OsMDHAR4 transcripts increased rapidly at 0.5 h (1.8-fold), after which a slight decrease occurred at 3 h, and then they continued to rise and peaked at 24 h (2.8-fold). Under the PEG treatment, the transcript abundance of OsMDHAR4 was nearly unchanged at most time points except for 3 h and 24 h, where two induced peaks (1.7 and 2.2-fold) appeared. In the ABA treatment, OsMDHAR4 was induced quickly at 0.5 h and 3 h, after which a dramatic decrease appeared at 6 h and 12 h, and recovered to initial level until 24 h. When treated with H 2 O 2 , the OsMDHAR4 expression showed a significant trend of rising followed by falling. The OsMDHAR4 transcript level was slightly suppressed by salt and cold treatments at 6 h and 12 h (Fig. 2b). These results demonstrated that OsMDHAR4 is responsive to multiple stresses.

Disruption of OsMDHAR4 increased heat tolerance
As shown in Fig. 2b, the expression of OsMDHAR4 was significantly induced by heat treatment, and the positive roles of MDHAR have been extensively reported under abiotic stress such as ozone, salt, PEG (Sharma and Davis 1997;Eltayeb et al.,2007) and drought (Sharma and Dubey 2005). However, the role of the MDHAR gene in rice remains largely unknown. To evaluate the function of OsMDHAR4 in heat stress response, we firstly searched for the Rice Mutant Database (http://rmd.ncpgr.cn/). Then, the osmdhar4 mutant (RMD_03Z11DA36), a transfer DNA (T-DNA) insertion line in the japonica rice ZH11 background was obtained (Wu et al., 2003;Zhang et al., 2006). DNA sequencing and genotyping revealed that the T-DNA was inserted in the first intron of the OsMDHAR4 gene, 254 bp downstream of the translation initiation site ( Fig. 3a and Additional file 2: Figure S2). A qRT-PCR assay showed that OsMDHAR4 was expressed at a very low level in the osmdhar4 plants (Fig. 3b). Under normal growth conditions, the osmdhar4 mutant had no obvious differences compared with the wild type (WT, ZH11) (Fig. 3c, left). osmdhar4 and WT seedlings (5.5-to 6.5-leaf stage) were treated at high temperature (45°C) for 72 h and subsequently returned to 26°C for recovery. Survival rates were used as a measure of heat tolerance. After the heat treatment and recovery, 80.5% of the osmd-har4 seedlings survived compared with 45.6% of the WT seedlings (Fig. 3c, d). Subsequently, we measured the water loss rates of detached leaves from osmdhar4 and WT. The results showed that the detached leaves of osmdhar4 lost water more slowly than the WT leaves ( Fig. 3e). This decreased water loss of the osmdhar4 mutant might lead to an increased tolerance to high temperature.

Overexpression of OsMDHAR4 decreased heat tolerance
To further validate the functions of OsMDHAR4 in rice, we constructed an overexpression vector of OsMDHAR4 driven by 35S promoter. The OsMDHAR4-overexpressing vector was transformed into rice ZH11. PCR analysis using OsMDHAR4-specific primers (OsMDHAR4-OE-F and R) and vector-specific primers (Hyg-F and R) confirmed the presence of the transgene in the two Error bars indicate the SE based on three biological replicates. Statistical differences among the samples are labeled with different letters according to the LSD test (P < 0.05, one-way ANOVA). b Expression of OsMDHAR4 under abiotic stress conditions. Seedlings (3.5-to 4.5-leaf stage) were subjected to heat (45°C), PEG (20% [w/v] PEG 6000), ABA (100 μM), H 2 O 2 (100 mM), salt (250 mM NaCl), and cold (4°C). Relative expression levels of OsMDHAR4 were examined by qRT-PCR. The rice Actin1 gene was used as the internal control. Error bars indicate the SE based on three biological replicates. *, P < 0.05, **, P < 0.01, by Student's t-test independent overexpression lines OsMDHAR4-OE-1 and OsMDHAR4-OE-5 (Additional file 3: Figure S3). Both the two transgenic lines, in which the OsMDHAR4 expression level was significantly increased approximately triple and twice of that in WT (Fig. 4a), were selected for heat tolerance testing (45°C for 48 h at the 5.5-to 6.5-leaf stage). Under normal conditions, we did not observe any phenotypic differences between these two overexpression lines and the WT (Fig. 4b, top). Under the heat stress treatment, both OsMDHAR4-OE-1 and OE-5 were more sensitive than the WT, which was consistent with the behavior of the osmdhar4 mutant in heat treatment. The survival rates ranged from 46.9 to 60.0% for the OsMDHAR4-overexpressing lines, and 90.0 to 90.6% for the WT plants after recovery, but there was no significant difference between OsMDHAR4-OE-1 and OE-5 (Fig. 4b, c). Consistent with this result, detached leaves of OsMDHAR4-overexpressing lines lost water more quickly than the WT leaves (Fig. 4d). Taken together, all of these results suggested that OsMD-HAR4 plays a negative role in heat tolerance in rice.
OsMDHAR4 alters the stomatal aperture status of rice seedlings The function of OsMDHAR4 in the control of water loss prompted us to investigate the stomatal aperture status, the major factor affecting the water-holding capacity in rice leaves. The stomatal apertures of the osmdhar4 mutant, OsMDHAR4-OE-1 and the WT plants were checked by scanning electron microscopy. No significant differences in the percentages of the three types of c 5.5-to 6.5-leaf stage WT and the osmdhar4 mutant plants were growing in barrels filled with well-mixed soil. The left and right half of each barrel was planted with WT and mutant plants, respectively. Plants were treated with high temperature (45°C) for 72 h, followed by culturing under normal conditions for 7 d. The photograph was taken before heat stress and after recovery. d Comparison of survival rates of WT and the osmdhar4 mutant after heat stress. Error bars indicated the SE based on three biological replicates. *, P < 0.05, by Student's t-test. e Water loss from detached leaves of WT and the osmdhar4 mutant at indicated time points. FW, Fresh Weight stomata and stomatal conductance between the osmd-har4 mutant and the WT were observed before heat stress. However, after heat treatment (45°C for 24 h at 5.5to 6.5-leaf stage), the results showed that 56.8% of stomata were completely closed in the osmdhar4 mutant, whereas only 25.0% were completely closed in the WT plants. On the other hand, only 10.5% of stomata were completely open in the osmdhar4 mutant, but 46% were completely open in WT. The percentage of partially open stomata didn't differ significantly from each other (Fig. 5a, b). Furthermore, the stomatal conductance was dramatically decreased in the osmdhar4 mutant compared the WT plants after heat treatment (Fig. 5c). These results were in agreement with the slower water loss of detached leaves from the osmdhar4 mutant. Either before or after heat treatment, OsMDHAR4-OE-1 plants had lower percentages of completely closed stomata than WT (before, 32.2% for OsMD-HAR4-OE-1, 40.4% for the WT; after, 17.2% for OsMDHAR4-OE-1, 30.4% for the WT), while the percentages of completely open were higher than the WT (before, 24.6% for OsMDHAR4-OE-1, 20.1% for the WT; after, 36.6% for OsMDHAR4-OE-1, 25.1% for the WT; Fig. 5d). Moreover, compared with the WT, OsMDHAR4-OE-1 had significantly higher stomatal conductance after heat treatment (Fig. 5e). These results were also in agreement with the faster water loss of the OsMDHAR4-OE-1 plants.
OsMDHAR4 modulates H 2 O 2 homeostasis of rice seedlings ROS are continuously produced as the byproducts of various metabolic pathways and are scavenged by different antioxidative defense components in plants. The equilibrium between ROS production and scavenging may be perturbed by adverse environmental factors, such as heat stress (Apel and Hirt, 2004). As H 2 O 2 is a very stable ROS with a long half-life (Levine et al., 1994) and a signal molecular that induces stomatal closure (McAinsh et al., 1996), so we measured the H 2 O 2 content in the rice seedling leaves. A higher level of H 2 O 2 accumulation was detected in the osmdhar4 mutant either before or after heat treatment (45°C for 24 h at 5.5-to 6.5-leaf stage, Fig. 6a). The DAB staining results also support the quantitative analysis (Fig. 6b). In addition, more H 2 O 2 accumulation was detected in the guard cells of the osmdhar4 mutant under normal conditions, as indicated by the ROS indicator, H2DCFDA (Additional file 4: Figure S4). Thus, ROS homeostasis in osmdhar4 seedlings was perturbed. These results indicated that the increased stomatal closure in the osmdhar4 mutant was probably due to accumulation of H 2 O 2 in guard cells. As expected, in the OsMDHAR4-OE-1 seedling leaves, visibly less H 2 O 2 accumulation was detected under normal conditions or after heat stress through the quantitative analysis and DAB staining (Fig. 6c, d). Taken together, our results indicated that OsMDHAR4 may negatively regulate H 2 O 2 -induced stomatal closure. Fig. 4 Overexpression of OsMDHAR4 showed increased sensitivity to heat stress. a Transcript level of OsMDHAR4 was measured by qRT-PCR in two OsMDHAR4-overexpresing lines (OsMDHAR4-OE-1 and OsMDHAR4-OE-5, ZH11 background). The rice Actin1 gene was used as the internal control. Error bars indicate the SE based on three biological replicates.*, P < 0.05, by Student's t-test. b 5.5-to 6.5-leaf stage ZH11 and OsMDHAR4-overexpressing plants were subjected to 45°C high temperature environment for 48 h, followed by growing in normal conditions for 7 d. Surviving seedlings were photographed. c Comparison of the survival rates of ZH11 and OsMDHAR4-overexpressing plants after heat stress. Error bars indicated the SE based on three biological replicates. *, P < 0.05, by Student's t-test. d Water loss from detached leaves of ZH11and OsMDHAR4-overexpressing plants at indicated time points. FW, Fresh Weight Discussion MDAR has been reported to be distributed widely across kingdoms pointing towards a universal role. In plants, it has been found to localize in chloroplasts, glyoxysomes, mitochondria, and cytosol. Chloroplastic MDAR localizes on the stroma of the chloroplast (Sano et al., 2005) and catalyzes the conversion of monodehydroascorbate (MDA) to AsA. AsA is a major antioxidant molecule that directly neutralizes ROS. Previous research has indicated that overexpression of chloroplastic MDAR from tomato enhanced tolerance to temperature and methyl viologen-mediated oxidative stress (Li et al., 2010). However, the physiological role of the chloroplastic MDAR in rice in response to heat stress is not reported.
In this study, OsMDHAR4, one new gene encoding monodehydroascorbate reductase was isolated from rice, and its amino acid sequence showed high identity and similarity to the MDAR sequences from several other plant species (Additional file 1: Figure S1). Bioinformatics studies predicted that ZmMDAR and BdMDAR, the two most homologous MDARs from Zea mays and Brachypodium distachyon, were located in the cytoplasm. However, intracellular localization studies using YFP fusion confirmed chloroplast localization of OsMDHAR4 (Fig. 1). Plant chloroplasts are the most significant generators of ROS and proposed to be heat sensors. A large number of chloroplasts are distributed in mesophyll cells (Ishikawa and Shigeoka, 2008;Maruta et al., 2012). OsMD-HAR4 transcripts peaked in the blade of flag leaf, in which a lot of mesophyll cells were distributed, while it showed the lowest transcript level in the root (almost with no chloroplasts) (Fig. 2a), the results further confirmed the finding that OsMDHAR4 protein was located in chloroplast. Since 1995, several studies have reported that MDAR genes are regulated by abiotic stresses (Leterrier et al., 2005;Grantz et al., 1995;Eltelib et al., 2011). We found that OsMDHAR4 transcripts were induced by heat and PEG (2.8-fold and 2.2-fold). Under both ABA and H 2 O 2 treatment, the OsMDHAR4 expression levels were increased in first and decreased at last, but salt and cold treatments had almost no effect on these (Fig. 2b). So, we assumed that OsMD-HAR4 may function in rice heat resistance.
Many scholars have focused on conferring abiotic stress tolerance through the overexpression of MDAR genes. Overexpression of LeMDAR gene alleviated photoinhibition of PSI and PSII and enhanced the tolerance to various abiotic stresses by elevating AsA level (Li et al., 2010).
Overexpressing AeMDHAR gene increased MDHAR enzyme activity compared to untransformed plants under both NaCl and control conditions (Sultana et al., 2012). OsMDHAR-expressing yeast cells displayed enhanced tolerance to H 2 O 2 by maintaining redox homoeostasis, proteostasis, and the ascorbate (AsA)-like pool (Kim et al., 2016). In the above studies, we can find that all the MDAR genes play a positive role in response to different abiotic stresses. In our study, we found that disruption of OsMDHAR4 reduced the water loss rates and increased heat tolerance (Fig. 3), and overexpression of OsMDHAR4 accelerated the water loss rates and decreased heat tolerance (Fig. 4). Unlike previous researches, these results showed that OsMD-HAR4 plays a negative role in heat tolerance in rice.
The importance of maintaining higher levels of AsA over DHA has been reported in previous studies. In transgenic tobacco plants overexpressing cucumber ascorbate oxidase, decreased AsA/DHA ratio increased ozone sensitivity (Sanmartin et al., 2003), whereas a higher AsA/DHA brought about by overexpression of DHAR and MDAR in cytosol resulted in enhanced tolerance to salt stress (Eltayeb et al., 2006). Higher AsA/DHA ratio, higher photosynthetic activity and lower H 2 O 2 contents were reported in transgenic tobacco plants expressing ascorbate oxidase gene (Yamamoto et al., 2005). According to this, we measured the AsA and DHA contents in leaves. Consistently with previous study , the Fig. 6 Accumulation of H 2 O 2 in the osmdhar4 mutant and OsMDHAR4-overexpressing plants. a Detection of the H 2 O 2 content of WT and the OsMDHAR4 mutant under normal and heat stress conditions. b DAB staining in the seedling leaves of WT and the osmdhar4 mutant under normal and heat stress conditions. c Detection of the H 2 O 2 content of WT and OsMDHAR4-OE-1 plants under normal and heat stress conditions. d DAB staining in the seedling leaves of WT and OsMDHAR4-OE-1 plants under normal and heat stress conditions. Error bars indicate the SE based on three biological replicates. *, P < 0.05, and **, P < 0.01, by Student's t-test. FW, Fresh Weight osmdhar4 mutant showed lower AsA contents and AsA/ DHA ratio than the WT before or after heat treatment,and the OsMDHAR4-OE-1 plants showed higher AsA contents and AsA/DHA ratio than the WT before or after treatment (Additional file 5: Figure S5). However, the changes of ascorbate redox state were not be able to explain the increased heat tolerance in the osmdhar4 mutant and incremental heat sensitivity in OsMD-HAR4 overexpression lines.
Stomata control the exchange of gases-most importantly water vapor and CO 2 -between the interior of the leaf and the atmosphere, and are involved in responses to abiotic stresses (Hetherington and Woodward, 2003). Stomatal closure degree makes major contribution to the ability of the plant to hold water in rice leaves. H 2 O 2 is an essential signaling molecule involved in the regulation of stomatal movement (Wang and Song, 2008;Yao et al., 2013). Mutation of DST results in the down regulation of peroxidase 24 precursor (Prx-24, a scavenger of H 2 O 2 ), might lead to the accumulation of H 2 O 2 in guard cells and trigger stomatal closure, and enhances drought and salt tolerance (Huang et al., 2009). Arabidopsis mutants lacking either or both a cytosolic and chloroplastic ascorbate peroxidase (APX), which were responsible for H 2 O 2 removal, were found to be more tolerant to salinity stress (Miller et al. 2007). Recently,one study showed that overexpression of Prx-24 (the target gene of DST) enhanced the sensitivity to drought and salt stress in rice (Cui et al., 2015). In our case, OsMDHAR4 might also be a target regulated by some transcript factors in response to heat stress, more research will be required to address this hypothesis. Liu et al. (2016) found that the oshtas mutant (gain-of-function) showed a strongly enhanced tolerance to heat stress with lower water loss rates, more closed stomata and significantly higher H 2 O 2 contents after the treatment as compared with the WT. Our data also showed that the osmdhar4 mutant had more closed stomata and significantly reduced stomatal conductance after heat treatment. Meanwhile, the osmdhar4 mutant accumulated more H 2 O 2 as compared with the WT before or after heat treatment. All of the above was found to be opposite in the OsMDHAR4 overexpression line (OsMD-HAR4-OE-1, Figs. 5, 6). These findings suggested that OsMDHAR4 might play a negative regulation role in H 2 O 2 -induced stomatal closure.

Conclusions
Our results demonstrated that OsMDHAR4 negatively regulates tolerance to heat stress in rice by promoting H 2 O 2 -induced stomatal closure. This study increases our insights into the molecular mechanisms of rice responses to heat stress and may ultimately be helpful for enhancing heat-tolerance via genetic engineering in crop breeding programs.

Plant materials and stress treatments
Overexpression transgenic plants and the osmdhar4 mutant were all based on the Oryza sativa L. ssp. Japonica Zhonghua 11 rice background.
To check the expression levels of the OsMDHAR4 gene under various abiotic stresses or phytohormone treatment, Zhonghua 11 rice plants were grown in Yoshida solution for approximately 3 weeks under normal conditions. The seedlings at the 3.5-to 4.5-leaf stage were treated with abiotic stresses, including heat stress (exposing plants to 45°C), simulated drought stress (treated with 20% [w/v] PEG 6000), ABA treatment (100 μM ABA), oxidative stress (treated with 100 mM H 2 O 2 ), salt stress (treated with 250 mM NaCl), and cold stress (seedlings were transferred to a growth chamber at 4°C), followed by sampling at the designated time points (0 h, 0.5 h, 3 h, 6 h, 12 h and 24 h). We prepared each RNA sample with shoots from at least four seedlings and ensured that every sampled seedling was of the same growth stage.
Before treatment, OsMDHAR4-overexpressing transgenic plants were firstly selected by germinated seeds on 1/2 MS medium containing 50 mg L − 1 hygromycin, WT and the osmdhar4 mutant were grown on normal 1/2 MS medium 1 d later. For heat stress testing, the uniformly germinated seeds were sown in the same barrels filled with a well-mixed soil, which was then cultured in a growth chamber with a 13-h-light (28°C)/11-h-dark (25°C) photoperiod, 70% humidity. The 5.5-to 6.5-leaf stage seedlings were transferred to a growth chamber with a 13-h-light (45°C)/11-h-dark (45°C) photoperiod, 70% humidity for 48 h or 72 h. The treated seedlings were transferred back into the previous growth conditions for recovery, and seedlings with newly growing leaf blades were then counted as surviving plants.

Plasmid construction and Rice transformation
To generate the overexpression construct, the full-length CDS of OsMDHAR4 was amplified from the cDNA of Zhonghua 11 by PCR. The PCR fragment was inserted into vector pEXT06/g (target gene driven by 35S promoter). The resulting construct was then confirmed by sequencing and transformed into Zhonghua 11, a japonica rice that can be easily transformed, by the Agrobacterium tumefaciens-mediated co-cultivation method. Transformed calli were selected on hygromycin medium. The primers used in this work are listed in Additional file 6: Table S1.

Subcellular localization of OsMDHAR4
Coding sequence of the OsMDHAR4 gene was amplified by PCR and directionally inserted into pCAMBIA1300-35S:YFP-NOS vector. The cultures of the Agrobacterium tumefaciens strain EHA105 harboring empty and recombinant vector were used to infect the healthy leaves of Nicotiana benthamiana (4 weeks old). The fluorescence signals were observed using a confocal microscope (Leica TCS SP5 or Zeiss LSM710). The primers used in this work are listed in Additional file 6: Table S1.

RNA extraction and qRT-PCR analysis
The TRIZol reagent (Invitrogen, USA) was used according to the manufacturer's instructions to extract total RNA. Before reverse transcription, total RNA was treated with gDNA Eraser (TaKaRa, Japan) for 5 min at 42°C to degrade possibly contaminated residual genomic DNA. The cDNA templates were synthesized using PrimeScript™ reagent Kit (TaKaRa, Japan) according to the manufacturer's instructions. Quantitative real-time PCR was performed on an optical 96-well plate with a CFX96 Real-time PCR Detection System (Bio-Rad, USA) using SYBR Premix Ex Taq (TaKaRa, Japan). The PCR thermal cycling protocol was as follows: 95°C for 10 s, followed by 40 cycles at 95°C for 5 s and 60°C for 30 s. Gene-specific primers for OsMDHAR4 were designed using the Roche Web site (Roche Applied Science). The rice Actin1 was used as the internal reference (Zhang et al., 2012), and data analyses with the 2 -ddCt method were performed as described (Livak and Schmittgen, 2001). The primers used in this work are listed in Additional file 6: Table S1.

Measurement of H 2 O 2 contents
Leaves harvested from 3.5-to 4.5-leaf stage seedlings, with or without heat treatment, were used to measure H 2 O 2 contents. The contents were measured spectrophotometrically after reaction with potassium iodide (KI). The reaction mixture consisted of 0.5 ml of 0.1% trichloroacetic acid (TCA), leaf extract supernatant, 0.5 ml of 100 mM potassium phosphate buffer (pH 7.8) and 1 ml reagent (1 M KI, w/v in fresh double-distilled water). The blank control consisted of 1 ml 0.1% TCA and 1 ml KI in the absence of leaf extract. After 1 h of reaction in darkness, the absorbance was measured at 390 nm. The amount of H 2 O 2 was calculated using a standard curve prepared with known concentrations of H 2 O 2 .
DAB staining for H 2 O 2 in leaves and detection of H 2 O 2 production in guard cells 3,3′-Diaminobenzidine (DAB) staining was performed following a published method with some modifications . Rice seedlings at 5.5-to 6.5-leaf stage were treated with a 45°C high temperature for 24 h or not treated, then the leaves (3-5 mm in width) at the same position were detached and immersed in 1% 3,3′-diaminobenzidine (DAB) solution in HCl-acidifed (pH 3.8). After 30 min under vacuum the samples were incubated at room temperature for 24 h in the dark. The samples were then bleached by boiling in ethanol in order to remove the chlorophyll and reveal the brown spots, which are indicative of the reaction of DAB with H 2 O 2 . The samples were observed and imaged under a scanner.
H 2 O 2 production in guard cells was detected using 2′,7′-dichlorodihydrofluorescein diacetate (H2DCFDA; Molecular Probes) as described previously (Huang et al., 2009) with small modifications. Leaves of 5.5-to 6.5-leaf stage seedlings were immerged in 0.01% Tween-20 and vacuum-infiltrated for 10 min. After rinsing twice with distilled water, leaves were incubated in 3% (w/v) cellulase RS (Yakult Honsha, Japan) for 5 h at 40°C without shaking to facilitate peeling off the epidermal strips. The epidermal strips were peeled off from leaves using tissue forceps. After washing with loading buffer (10 mM Tris-HCl, 50 mM KCl at pH 7.2), the epidermal strips were incubated in staining buffer (loading buffer containing 50 mM H2DCFDA) for 10 min at room temperature in the dark. The epidermal strips were washed with distilled water to remove the excess dye. The fluorescence was examined using a confocal laser-scanning microscope (Leica TCS SP5, Germany). All confocal images were taken under identical conditions and the guard cell region was selected to quantify the mean grey value of guard cells.

Observation of guard cells and detection of water loss rates
Leaves of 3.5-to 4.5-leaf stage plants with or without heat treatment for 24 h were fixed with 2.5% (v/v) glutaraldehyde, and stomatal images (20 kV, 2000×) were obtained using scanning electron microscopy (KYKY-EM3200).
Plants germinated under normal conditions for 4 weeks. The leaves were detached from various lines with same age and position, and weighed immediately as the initial fresh weight. They were then placed in clean filter papers, and incubated at 25°C. The decreases in fresh weight were recorded at every 30 min for 6.5 h or 7 h. Water loss was presented as percentage of fresh weight loss versus the initial fresh weight.

Measurement of the AsA and DHA contents
Leaves harvested from 3.5-to 4.5-leaf stage seedlings, with or without heat treatment, were used to measure the total AsA (tAsA), AsA and DHA contents determined by the spectrophotometric method described previously (Gillespie and Ainsworth, 2007), with minor modifications. The frozen leaf samples were ground with inert sand and 10% trichloroacetic acid (TCA) solution using a mortar and pestle. The homogenate was centrifuged at 12,000 rpm for 20 min. The tAsA contents were determined in a reaction mixture consisting of crude extract and 150 mM KH 2 PO 4 buffer (pH 7.4) containing 5 mM EDTA and 10 mM dithiothreitol (DTT) for the reduction of DHA to AsA. The reaction mixtures were incubated at room temperature for 10 min and 0.5% N-ethylmaleimide (NEM) was added. AsA was assayed in a similar manner, except that deionized H 2 O was substituted for DTT and NEM. The color was developed in both the reaction mixtures by the addition of 10% TCA, 44% o-phosphoric acid, α,ά-dipyridyl in 70% ethanol, and 30% FeCl 3 . The reaction mixtures were incubated at 37°C for 1 h and quantified spectrophotometrically at 525 nm.