The microRNA408–plantacyanin module balances plant growth and drought resistance by regulating reactive oxygen species homeostasis in guard cells

Abstract

The conserved microRNA (miRNA) miR408 enhances photosynthesis and compromises stress tolerance in multiple plants, but the cellular mechanism underlying its function remains largely unclear. Here, we show that in Arabidopsis (Arabidopsis thaliana), the transcript encoding the blue copper protein PLANTACYANIN (PCY) is the primary target for miR408 in vegetative tissues. PCY is preferentially expressed in the guard cells, and PCY is associated with the endomembrane surrounding individual chloroplasts. We found that the MIR408 promoter is suppressed by multiple abscisic acid (ABA)-responsive transcription factors, thus allowing PCY to accumulate under stress conditions. Genetic analysis revealed that PCY elevates reactive oxygen species (ROS) levels in the guard cells, promotes stomatal closure, reduces photosynthetic gas exchange, and enhances drought resistance. Moreover, the miR408–PCY module is sufficient to rescue the growth and drought tolerance phenotypes caused by gain- and loss-of-function of MYB44, an established positive regulator of ABA responses, indicating that the miR408–PCY module relays ABA signaling for regulating ROS homeostasis and drought resistance. These results demonstrate that miR408 regulates stomatal movement to balance growth and drought resistance, providing a mechanistic understanding of why miR408 is selected during land plant evolution and insights into the long-pursued quest of breeding drought-tolerant and high-yielding crops.

The microRNA miR408 is repressed by ABA signaling and regulates PCY abundance in the guard cells thus modulating ROS homeostasis and stomatal movement. MicroRNA408 regulates PCY abundance in guard cells to modulate ROS homeostasis and stomatal movement, thereby balancing growth and drought resistance in Arabidopsis.

Introduction

MicroRNAs (miRNAs) are fast-evolving endogenous small RNAs that regulate organism function and behavior (Voinnet 2009; Song et al. 2019; Guo et al. 2022). There are only a handful of miRNA families that are deeply conserved in land plants, including miR408 (Cuperus et al. 2011; Pan et al. 2018; Guo et al. 2022), indicating that they exert functions fundamental to plant adaption to terrestrial habitats. In Arabidopsis (Arabidopsis thaliana), miR408 targets transcripts encoding 2 types of copper proteins, the small blue copper protein plantacyanin (PCY) and laccases, including LACCASE3 (LAC3), LAC12, and LAC13 (Abdel-Ghany and Pilon 2008; Cuperus et al. 2011; Pan et al. 2018). The first demonstrated biological role of miR408 was to promote vegetative growth, which was attributed to increase in chloroplast copper content, plastocyanin abundance, and photosynthesis (Zhang and Li 2013; Zhang et al. 2014; Pan et al. 2018). In Arabidopsis, the miR408–PCY module was also found to play a role in far-red light-induced seed germination and dark-induced leaf senescence (Jiang et al. 2021; Hao et al. 2022). In these cases, transcriptional regulation by SQUAMOSA PROMOTER BINDING PROTEIN-LIKE 7 (SPL7), ELONGATED HYPOCOTYL5 (HY5), and PHYTOCHROME INTERACTING FACTORs (PIFs), which are related to copper homeostasis and light signaling, respectively, has been shown to regulate the spatiotemporal accumulation of miR408 (Zhang et al. 2014; Jiang et al. 2021; Hao et al. 2022).

In addition to regulating plant growth and development, miR408 has been implicated in stress responses in diverse plant species. It was generally observed that miR408 abundance is influenced by a variety of factors that limit optimal growth, including mechanical stress, dehydration, chilling, salinity, and reactive oxygen species (ROS) (Trindade et al. 2010; Jovanovic et al. 2014; Ma et al. 2015; Candar-Cakir et al. 2016; Qin et al. 2023). Analyses of Arabidopsis plants with altered miR408 expression have led to the conclusion that miR408 increases tolerance to salinity and oxidative stress but decreases tolerance to drought and osmotic stress (Ma et al. 2015). These studies overall demonstrated the involvement of miR408 in abiotic stress responses, but the direct cellular mechanism by which miR408 coordinates plant growth and stress responses has not been elucidated. Thus, how this deeply conserved miRNA contributes to fundamental plant biology remains elusive.

Since occupying terrestrial habitats, land plants face the dilemma of allowing sufficient carbon dioxide to enter the leaf for photosynthesis and reducing concomitant water loss to avoid dehydration (Edwards et al. 1992; Zhao et al. 2019). Controlling gas exchange between the leaf and external atmosphere, stomata play a critical role in balancing photosynthesis and transpiration (Nilson and Assmann 2007). Osmoregulation of stomata in response to various environmental stimuli and internal signals is collectively controlled by ion transport across the plasma membrane and tonoplast inside guard cells to regulate turgor pressure, which is governed by sophisticated signal transduction cascades (Daloso et al. 2017; Jezek and Blatt 2017; Chen et al. 2020; Lawson and Matthews 2020). In most species, guard cells are the only epidermal cells containing chloroplasts (Lawson et al. 2008). Although the role of guard cell chloroplasts in stomatal movements has been a subject of debate (Lawson et al. 2008; Azoulay-Shemer et al. 2015), evidence indicates that this organelle contributes to stomatal movements by providing energy for proton pumping and starch turnover (Wang et al. 2014; Lawson and Matthews 2020; Lim et al. 2022). Moreover, guard cell chloroplasts are required for photoproduced and retrograde signaling-induced ROS generation, which is critical for cellular ROS homeostasis that ultimately controls stomatal closure (Asada 2006; Iwai et al. 2019; Zhao et al. 2019; Postiglione and Muday 2020).

Abscisic acid (ABA) is a powerful phytohormone that regulates stomatal movement. Decreased water availability increases ABA biosynthesis and accumulation in guard cells, which is perceived by a family of soluble receptor proteins, leading to activation of Sucrose nonfermenting Related Kinase 2 family members (SnRK2s) (Ma et al. 2009; Park et al. 2009; Umezawa et al. 2009; Vlad et al. 2009; Tischer et al. 2017). Activated SnRK2s transmit the ABA signal through signaling cascades that include bursts of ROS and calcium waves (Pei et al. 2000; Zhu 2016), which stimulate guard cell ion channels to ultimately control stomatal movements (Schmidt et al. 1995; Grabov and Blatt 1997, 1999; Geiger et al. 2009; Meyer et al. 2010; Demidchik 2018). ABA signal transduction requires a sophisticated transcription network to govern the expression of hundreds of stress-responsive genes. Among the key transcription factors (TFs) activated by ABA signaling are members of the ABA-responsive element (ABRE)-binding factor (ABF)/ABA-responsive element-binding protein (AREB), MYB, MYC, N-acetylcysteine (NAC), and WRKY families (Tuteja 2007). These TFs interact with their corresponding cis-acting elements to transcriptionally reprogram the ABA-responsive genes that lead to drought adaptation (Tuteja 2007). Using genome-wide targets of 21 ABA-related TFs determined by chromatin immunoprecipitation sequencing (ChIP-seq), a comprehensive ABA TF network was constructed in Arabidopsis (Song et al. 2016), which is useful for transcriptionally deciphering how plants modulate growth and resilience to stresses.

In this study, we identified the MIR408 promoter as a hotspot of ABA-mediated transcriptional repression in Arabidopsis. Through molecular and genetic analyses, we showed that repression of miR408 by ABA signaling leads to increased accumulation of PCY, which locates in endomembranes enclosing individual chloroplasts in the guard cells to promote ROS accumulation and stomatal closure. These findings indicate that the primary function of miR408 in the leaf is to modulate stomatal movement for balancing growth and drought resistance, an eternal dilemma faced by terrestrial plants.

Results

The MIR408 promoter is an ABA transcriptional repression hotspot

Re-analyzing the ChIP-seq data for a set of 21 ABA-responsive TFs in Arabidopsis (Song et al. 2016), we found that 1,908 genes were targeted by at least 85% (18 of 21) of these TFs, including 1,898 protein-coding genes (Fig. 1A; Supplementary Fig. S1A). We identified Gene Ontology (GO) terms associated with these genes and found that GO terms related to water stress and cellular response to ABA were among those most significantly enriched (Supplementary Fig. S1B), suggesting that ABA-related genes are heavily targeted by multiple TFs. In addition, we identified 10 miRNA genes that were targeted by at least 85% of the ABA-responsive TFs (Fig. 1A; Supplementary Fig. S1A). We found that the proximal promoter of MIR408 was special among the 10 miRNA genes in that the binding peaks of the 18 TFs were concentrated in a region of approximately 500 bp (Fig. 1B). This observation suggests that the MIR408 promoter is a transcriptional regulation hotspot of the ABA-responsive TFs.

Figure 1.

Identification of the MIR408 promoter as an ABA transcriptional repression hotspot. A) The 1,908 genes heavily targeted by ABA TFs include 1,898 protein-coding genes and 10 miRNA genes. B) Binding patterns of 18 ABA TFs at the MIR408 promoter based on published ChIP-seq data. Chromosomal positions of pri-miR408 (black arrow) and the MIR408 promoter (horizontal line) are indicated at the bottom. C) Confirmation of transcriptional repression of MIR408 by 6 selected TFs using the LUC/REN dual reporter system transiently expressed in tobacco leaf epidermal cells. Images show the effects of GFP alone and GFP fusions with NF-YB2, GBF3, MYB44, HB7, ABF1, or ABF4 on LUC activity driven by the MIR408 promoter. Graphs show quantification of the LUC/REN ratio. Values are means ± SD from 10 independent transfection events for each combination. ***, P < 0.001 by Student's t-test. ABA, abscisic acid.

To ascertain the effects of individual ABA-responsive TFs on the MIR408 promoter, we employed a transient expression assay in Nicotiana benthamiana leaf epidermal cells using the firefly luciferase (LUC) and Renilla luciferase (REN) dual reporters (Liu et al. 2014). We generated the pMIR408:LUC-35S:REN reporter constructs using the MIR408 promoter to drive LUC and the constitutive cauliflower mosaic virus (CaMV) 35S promoter to drive REN (Supplementary Fig. S2A). In addition, pACTIN2:LUC-35S:REN was generated as a control reporter (Supplementary Fig. S2A). For effector constructs, we generated 35S:GFP (green fluorescent protein) as the negative control and 35S:HY5-GFP as the positive control, as well as constructs for 6 randomly selected ABA TFs, including 35S:NF-YB2-GFP, 35S:GBF3-GFP, 35S:MYB44-GFP, 35S:HB7-GFP, 35S:ABF1-GFP, and 35S:ABF4-GFP (Supplementary Fig. S2B). Consistent with previous studies (Zhang et al. 2014), we found that HY5-GFP, but not GFP alone, significantly increased the LUC/REN ratio of the pMIR408:LUC-35S:REN reporter following co-transfection of N. benthamiana leaf epidermal cells (Supplementary Fig. S2C). Remarkably, all 6 tested ABA-responsive TFs significantly decreased the LUC/REN ratio (Fig. 1C). However, LUC/REN ratio of the pACTIN2:LUC-35S:REN control reporter was not affected by the ABA TFs such as MYB44 (Supplementary Fig. S2D). These results suggest that the ABA-responsive TFs consistently repress the MIR408 promoter.

We selected MYB44, a previously characterized positive regulator of ABA signaling (Jung et al. 2008), for characterizing its regulation of the MIR408 promoter. A cluster of 3 MYB binding sites (MBS1 to 3) (Ogata et al. 1994; Jung et al. 2012; Shim et al. 2013) was identified in the MIR408 promoter that coincided with the MYB44 binding peak defined in the ChIP-seq data (Fig. 2A) (Song et al. 2016). In vitro MYB44 binding to MBS1 was validated by electrophoretic mobility shift assay (EMSA) using recombinant MYB44 and a DNA probe consisting of 3 concatenated MBS1s (Fig. 2B). When a mutated version (MBS1^m; Fig. 2A) was used, production of the shifted band was effectively abolished (Fig. 2B). Employing the transient LUC/REN assay, we confirmed that MYB44 represses the MIR408 promoter in an MBS-dependent manner (Fig. 2C). To validate MYB44 binding in vivo, we performed ChIP using a transgenic plant expressing 35S:MYB44-GFP with an anti-GFP antibody. qPCR analysis showed that MYB44 occupancy at the MIR408 promoter was significantly enriched in comparison to the wild type (Fig. 2D). Furthermore, RT-qPCR comparison of the wild type and MYB44-OX seedlings, which overexpress MYB44 via the 35S promoter (Supplementary Fig. S3), revealed a significant decrease of pri-miR408 abundance in MYB44-OX (Fig. 2E). These results indicate that MYB44 represses the MIR408 promoter via binding to the MBSs.

Figure 2.

MYB44 directly inhibits MIR408 expression. A) MYB44 occupancy at the MIR408 promoter is associated with 3 MBS motifs depicted as purple ovals. Sequences of the MBS motifs and the mutant versions used for EMSA and LUC/REN assays are shown at the bottom. B) EMSA analysis of MYB44 binding to MBS1. Labeled probe consisting of 3 concatenated MBS1 or MBS1^m was incubated with GST-tagged recombinant MYB44. Competitor, un-labeled MBS1 in 200-fold excess. C) Confirmation of MBS-dependent MYB44 repression of the MIR408 promoter using the LUC/REN dual reporter system transiently expressed in tobacco leaf epidermal cells. Images show the effects of GFP and MYB44-GFP on LUC activity driven by the native or mutated MIR408 promoter (MIR408^m) in which the 3 MBS motifs are substituted with the corresponding MBS^m shown in A). Graphs show quantification of the LUC/REN ratio. Values are means ± SD from 4 independent transfection events for each combination. n.s., not significant; ***, P < 0.001 by Student's t-test. D) Confirmation of MYB44 binding to the MIR408 promoter by ChIP-qPCR. ChIP was performed in the indicated genotypes using the anti-GFP antibody. Positions of the P1 and P2 amplicons are indicated in A). Data are means ± SDs from 3 replicate qPCR performed on the same DNA. E) RT-qPCR comparison of relative pri-miR408 transcript levels in the MYB44-OX lines in comparison to the wild type. Data are means ± SD from 3 qPCR replicates performed on the same cDNA. WT, wild type.

To test whether MIR408 expression is repressed by ABA and dehydration, we generated the pMIR408:LUC transgenic Arabidopsis plants. We found that LUC activity was drastically decreased in the pMIR408:LUC seedlings experiencing ABA and dehydration treatments in comparison to the untreated seedlings (Fig. 3, A to D). In contrast, LUC activity in the 35S:LUC control seedling was unaffected by the same treatments (Fig. 3, A to D). Moreover, RT-qPCR analysis showed that pri-miR408 level was substantially reduced by ABA and dehydration treatments (Fig. 3, E and F). These results indicate that ABA treatment and dehydration stress both inhibit the expression of MIR408. Taken together, our results indicate that the MIR408 promoter is a repression hotspot of ABA-responsive TFs.

Figure 3.

Repression of MIR408 expression by ABA and dehydration treatments. A) Representative luminescence images of the 35S:LUC or pMIR408:LUC seedlings treated with (+ABA) or without (−ABA) 0.5 μM ABA. B) Quantification of relative luminescence. Data are mean ± SD from 5 independent experiments each containing 4 seedlings. n.s., not significant; ***, P < 0.001 by Student's t-test. C) Representative luminescence images of the 35S:LUC or pMIR408:LUC seedlings treated with (+dehydration) or without (−dehydration) dehydration, which was applied by opening the Petri dishes for 30 min in which 10-d-old seedlings were growing. D) Quantification of relative luminescence. Data are mean ± SD from 5 independent experiments each containing 3 seedlings. n.s., not significant; ***, P < 0.001 by Student's t-test. E) and F) RT-qPCR analysis of pri-miR408 level in response to ABA E) and dehydration stress F). Data are means ± SD from 3 replicate qPCR performed on the same cDNA. ABA, abscisic acid.

ABA enhances PCY accumulation in the guard cells

In Arabidopsis, miR408 has 4 validated target genes that all encode cuproproteins, namely, PCY, LAC3, LAC12, and LAC13 (Abdel-Ghany and Pilon 2008; Cuperus et al. 2011). Interrogation of the Arabidopsis eFP Browser, the online database of gene expression (Winter et al. 2007) showed that, while LAC13 expression was almost undetectable, PCY, LAC3, and LAC12 were mainly expressed in the leaves, the root, and the inflorescence stem, respectively (Supplementary Fig. S4A). In the seedlings, the transcript level of PCY, but not that of LAC3, LAC12, and LAC13, was induced by ABA (Fig. 4A). This expression pattern was verified by RT-qPCR analysis, which showed that transcript abundance of PCY was highly induced by both ABA and dehydration treatments (Fig. 4B).

Figure 4.

ABA enhances PCY expression in the guard cells. A) ABA-induced changes in transcript levels of the 4 miR408 target genes obtained from Arabidopsis eFP browser. B) RT-qPCR analysis of PCY transcript levels in seedlings following ABA (left) and dehydration (right) treatments. Data are means ± SD from 3 replicate qPCR performed on the same cDNA. C) The pPCY:PCY-GFP, pLAC3:LAC3-GFP, pLAC12:LAC12-GFP, and pLAC13:LAC13-GFP seedlings were treated with or without 10 μM ABA for 1 h and subjected to fluorescence microscopy. Bar, 20 μm. D) Quantification of relative GFP fluorescence intensity. Data are mean ± SD from 15 seedlings. n.s., not significant; ***, P < 0.001 by Student's t-test. E) GUS activity in the epidermis of various green tissues of the pPCY:GUS plants. Bar, 50 μm. ABA, abscisic acid.

As miRNAs may regulate target genes by inhibiting translation in addition to transcript cleavage, we further confirmed PCY as the main miR408 target in the leaf at the protein level. To this end, we generated the pPCY:PCY-GFP, pLAC3:LAC3-GFP, pLAC12:LAC12-GFP, and pLAC13:LAC13-GFP transgenic plants. In these plants, GFP was fused to the PCY, LAC3, LAC12, and LAC13 coding sequences on the C-terminus and driven by the respective native promoters (Fig. 4C). We found that the GFP fusion proteins for the 3 laccases were expressed in tissues with lignin deposition. For example, LAC3-GFP mainly accumulated on the Casparian strip in the root endodermal cells as previously reported (Zhuang et al. 2020) and LAC12-GFP accumulated in the xylem of the inflorescence stem (Supplementary Fig. S4, B to E). In the leaf expressing GFP fusions with the 3 laccases, only nonspecific autofluorescence in the inner side of guard cells due to thickened cell walls (Fahy et al. 2017) was observed (Fig. 4C). In contrast, the PCY-GFP signals were detected in the guard cells and the pavement cells of the leaf (Fig. 4C). Consistent with the transcript profiles, we found that ABA treatment enhanced the PCY-GFP signals but not the other 3 fusion proteins (Fig. 4, C and D).

To independently verify the guard cell expression pattern of PCY, we generated the pPCY:GUS reporter line in which the β-glucuronidase (GUS) gene was fused with the native PCY promoter. Consistent with previous reports (Dong et al. 2005), staining for GUS activity revealed that PCY was preferentially expressed in the guard cells of green tissues, including the rosette leaf, the petiole, and the pedicel (Fig. 4E). Taken together, these observations indicate that repression of miR408 by ABA signaling is a mechanism to enhance PCY accumulation in the guard cells of photosynthetic tissues.

PCY promotes ROS accumulation in the guard cells

PCY is a compact cuproprotein consisting of an N-terminal signal peptide and a C-terminal type I copper binding motif (Jiang et al. 2021). Previously we have reported that PCY is associated with the storage vacuoles in the cotyledons of non-germinated seeds (Jiang et al. 2021) and with the endomembrane in mesophyll cells of senescing leaves (Hao et al. 2022). Using the pPCY:PCY-GFP plants, we inspected the subcellular localization of PCY in the guard cells by fluorescence microscopy. This analysis also included the Q4-GFP (Cutler et al. 2000) and COPT5-GFP (Jiang et al. 2021) marker lines for labeling the endoplasmic reticulum membrane and the tonoplast, respectively. While PCY-GFP signal was more diffused than that of Q4-GFP and COPT5-GFP, their overall distribution patterns around the chloroplast were similar (Fig. 5, A and B; Supplementary Fig. S5), suggesting that PCY locates in the endomembrane surrounding the chloroplast and/or in association with the chloroplast membranes.

Figure 5.

PCY surrounds the chloroplast to promote ROS accumulation. A) Subcellular localization of PCY-GFP in the guard cells. Seedlings expressing pPCY:PCY-GFP were treated with 10 μM ABA for 1 h and subjected to fluorescence microscopy. Bar, 10 μm. B) Enlarged view of an individual chloroplast showing PCY-GFP signals (green) accumulate in the endomembrane surrounding the chloroplast (magenta). Bar, 0.5 μm. C) Representative fluorescence images showing DCF signals in individual guard cells of the indicated genotypes with and without ABA treatment. DCF signals are in green and chlorophyll autofluorescence in magenta. Bar, 5 μm. D) Quantification of the DCF autofluorescence in guard cells. DCF intensity was normalized against the chlorophyll autofluorescence of the same genotype with the relative DCF intensity in the wild type without ABA treatment set to 1. Data are mean ± SD from 12 stomata. Different letters denote samples with significant differences (one-way ANOVA, P < 0.001). E) Comparison of relative DCF intensity changes by ABA treatment. DCF intensity after ABA treatment was divided by that without ABA treatment. Different letters denote genotypes with significant differences (one-way ANOVA, P < 0.001). WT, wild type; ABA, abscisic acid.

In light, photosynthetic electron transport in the chloroplast is the main source of cellular ROS, which could be influenced by various stress conditions such as ABA-induced growth arrest (Asada 2006; Iwai et al. 2019). Moreover, ROS are ubiquitous signaling molecules mediating stomatal movement, including ABA-mediated stomatal closure (Huang et al. 2019; Postiglione and Muday 2020). The association of PCY with the chloroplast in the guard cells prompted us to investigate whether the miR408–PCY module participates in regulating ROS homeostasis. We monitored cellular ROS levels by fluorescence microscopy using a frequently utilized cell-permeable ROS sensor, 2′,7′-dichlorofluorescein diacetate (H₂DCF-DA), which is converted to the highly fluorescent DCF after oxidation (Pei et al. 2000). Consistent with previous reports (Iwai et al. 2019; Postiglione and Muday 2020), we found that the DCF signals were mainly detected in the chloroplasts of the guard cells (Fig. 5C; Supplementary Fig. S6).

Employing the previously characterized MIR408-OX and amiR408 plants that overexpress miR408 via the 35S promoter and silence MIR408 via an artificial miRNA (Zhang and Li 2013; Zhang et al. 2014), we found that the DCF fluorescence in MIR408-OX and amiR408 guard cells was significantly lower and higher than that in the wild type, respectively (Supplementary Fig. S6). Using PCY-OX plants that overexpress PCY via the 35S promoter (Jiang et al. 2021), we found that the intensity of DCF fluorescence was significantly increased in comparison to that in the wild type (Fig. 5, C and D). Conversely, in the PCY-KO plants that knockout PCY (Jiang et al. 2021), DCF fluorescence was significantly lower than that in the wild type (Fig. 5, C and D). These results indicate that the miR408 and PCY levels are associated with ROS levels in the guard cells.

Consistent with previous reports (Watkins et al. 2017), we observed a significant increase in the chloroplast DCF signals following ABA treatment of the wild-type plants (Fig. 5, C and D). This increase in chloroplast DCF signals by ABA treatment was further elevated in the PCY-OX guard cells (Fig. 5, C and E). In contrast, the extent of ABA-induced increase in DCF signals was significantly lesser in the PCY-KO guard cells compared to in the wild type (Fig. 5, C and E). These results indicate that induction of PCY is partially responsible for the increased ROS accumulation in the guard cell chloroplasts in response to ABA signaling.

PCY enhances stomatal closure

Since ROS ultimately controls stomatal movement, we investigated the effects of PCY on stomatal opening and closure. In peeled leaf epidermis, we found that PCY-KO plants showed a significant increase in stomatal aperture, calculated as the ratio between stomatal width and length (Roelfsema and Prins 1995), in comparison to the wild type (Fig. 6, A and B; Supplementary Fig. S7). However, PCY-OX plants exhibited a significant decrease in stomatal aperture compared to the wild type (Fig. 6, A and B; Supplementary Fig. S7). These observations indicate that the PCY levels are inversely associated with the sizes of leaf stomatal aperture during regular growth.

Figure 6.

PCY enhances stomatal closure. A) Representative images showing stomatal apertures of the indicated genotypes under the indicated treatments. The epidermis of 4-wk-old leaves was peeled and imaged immediately as the control. The epidermis was soaked in the stomatal opening buffer under light to open the stomata. For the +ABA treatment, 10 μM ABA was added to the stomatal opening buffer for 1 h. For the +NAC treatment, 1 mm NAC was added for 1 h before the addition of ABA. Bar, 10 μm. B) Quantification of stomatal apertures. Data are mean ± SD from 102 stomata. Different letters denote genotypes with significant differences within each treatment (one-way ANOVA, P < 0.05). C) Representative infrared images of 5-wk-old plants under normal and drought conditions. The images were pseudo-colored to show leaf temperature distribution. D) Average leaf temperature quantified from infrared thermography. Data are means ± SD from 15 individual leaves. Different letters denote genotypes with significant differences within each treatment (one-way ANOVA, P < 0.001). WT, wild type; ABA, abscisic acid.

Under light, potassium treatment triggers the opening of stomata (Fischer 1968). After treating the peeled epidermis with potassium-containing stomatal opening buffer, we found that PCY-KO and PCY-OX plants, respectively, showed a significant increase and a decrease in stomatal aperture compared to the wild type (Fig. 6, A and B; Supplementary Fig. S7). Consistently, we found that the MIR408-OX and amiR408 plants exhibited larger and smaller stomatal apertures in comparison to the wild type (Supplementary Fig. S6). The closing of open stomata is rapidly induced after application of exogenous ABA. Indeed, we found that addition of ABA to the stomatal opening buffer caused stomatal closure in the wild type (Fig. 6, A and B). In comparison, PCY-OX and PCY-KO plants exhibited significantly smaller and larger stomatal apertures, respectively, following the same ABA treatment (Fig. 6, A and B; Supplementary Fig. S7). To confirm that the reduction in stomatal aperture in PCY-OX plants was due to ROS hyper-accumulation, the cell-permeable ROS scavenger NAC was used. Under both conditions for inducing stomata opening and closure, we found that the addition of NAC successfully increased the stomatal aperture of PCY-OX plants to a degree comparable to that of PCY-KO plants (Fig. 6, A and B). Together these results indicate that PCY-mediated ROS homeostasis reduces stomatal opening but enhances stomatal closure.

Leaf surface temperature reflects the strength of transpiration, with higher leaf temperatures indicating lower levels of water loss through the stomata (Wang et al. 2018; Hong et al. 2020). We measured the leaf temperatures under the control and drought conditions using infrared imaging. Because plants actively slow their growth as an adaptive strategy to stresses (Zhang et al. 2020), the leaf temperatures under the control condition were higher than those recorded under drought (Fig. 6, C and D). Moreover, we found that the PCY-OX and PCY-KO canopies exhibited significantly higher and lower temperatures than the wild type (Fig. 6, C and D). These differences were also observed in plants grown under dehydration stress (Fig. 6, C and D). These observations revealed that PCY negatively impacts transpiration and confirmed that PCY-mediated ROS homeostasis is critical for stomatal movement.

miR408 and PCY reciprocally modulate growth and drought resistance

As stomatal movement directly controls stomatal conductance (Lawson and Matthews 2020), we tested whether PCY regulates stomatal conductance to CO₂ (g_sc). We found that the g_sc of the PCY-OX leaf was significantly lower than that of the wild type; however, the g_sc of the PCY-KO leaf was significantly higher than that of the wild type (Fig. 7A). Since net photosynthetic rate (NPR) positively correlates with g_sc (Gago et al. 2016), we also measured NPR in the fully expanded leaves of the wild-type, PCY-OX, and PCY-KO plants. This analysis confirmed that NPR of PCY-OX and PCY-KO plants was significantly lower and higher than that of the wild type, respectively (Fig. 7B). Furthermore, we found that these alterations to NPR were exhibited over the entire recorded range of actinic light intensity (Fig. 7C). These results indicate that PCY functions to reduce stomatal conductance and photosynthetic gas exchange.

Figure 7.

PCY negatively regulates plant growth. A) Comparison of g_sc of 5-wk-old WT, PCY-OX, and PCY-KO plants. Data are means ± SD from 10 individual fully expanded leaves. Different letters denote genotypes with significant differences (one-way ANOVA, P < 0.05). B) Comparison of NPR of the indicated genotypes. Data are means ± SD from 10 individual fully expanded leaves. Different letters denote genotypes with significant differences (one-way ANOVA, P < 0.001). C) Light-intensity-dependent NPR of WT, PCY-OX, and PCY-KO leaves. Data are means ± SD from 10 individual fully expanded leaves. D) Morphology of 5-wk-old plants grown under the short-day condition. Left, top view of a representative WT, PCY-OX, and PCY-KO plant. Right, morphology of individual leaves from the plants shown on the left. Rosette leaves and cotyledons are shown from left to right. Bar, 1 cm. E) and F) Quantification of the entire foliage area E) and fresh weight F) of 5-wk-old plants of the indicated genotypes. Values are means ± SD of 15 individual plants. Different letters denote genotypes with significant differences (one-way ANOVA, P < 0.001). WT, wild type.

We monitored plant growth patterns to confirm the negative effects of PCY on vegetative development. In comparison to the wild type, PCY-KO seedlings showed more vigorous growth, resulting in significantly higher fresh weight and cotyledon area than that of the wild type (Supplementary Fig. S8). Conversely, the PCY-OX seedlings showed reduced growth vigor, resulting in significantly lower fresh weight and cotyledon area than that of the wild type (Supplementary Fig. S8A and B). In the adult stage, an enhancement of vegetative development was readily noticeable for the PCY-KO plants, which exhibited large individual leaves, significantly larger foliage area and higher fresh weight than the wild type (Fig. 7, D to F). On the contrary, with significantly smaller foliage area and lower fresh weight (Fig. 7, D to F), the growth vigor of PCY-OX was severely reduced compared with that of the wild type. These results are in line with the observed changes in the photosynthetic parameters of the PCY-OX and PCY-KO plants (Fig. 7, A to C; Supplementary Fig. S8C). Thus, consistent with previous reports that elevated miR408 level leads to a higher capacity for CO₂ fixation, improved photosynthetic performance, and more vigorous vegetative growth (Zhang and Li 2013; Zhang et al. 2014; Pan et al. 2018), miR408 and PCY reciprocally modulate plant growth.

We further examined whether the miR408–PCY module affects plant responses to drought stress. To facilitate phenotypic comparison, we designed 2 drought treatments with different severity. In the so-called moderate drought condition, we kept the plants well-watered for 4 wk and then withheld watering for 15 d (Supplementary Fig. S9). In the so-called severe drought condition, we kept the plants well-watered for 3 wk before stopping watering for 21 days (Supplementary Fig. S9). Under the moderate drought stress, the MIR408-OX and PCY-KO leaves exhibited apparently more severe wilting symptoms than the wild type (Fig. 8A). Three days after re-watering, while a near 100% survival rate, or the percentage of plants that are green and turgid, was recorded for the wild type, the average survival rates for MIR408-OX and PCY-KO were approximately 8% (Fig. 8B). Under the severe drought treatment, we found that the amiR408 and PCY-OX plants displayed much less wilting than the wild type (Fig. 8C). After re-watering, while the wild type exhibited an average survival rate of merely 3%, nearly all the amiR408 (92.6%) and PCY-OX (95.5%) plants survived and remained green (Fig. 8D). In addition to the survival rate, we examined the expression pattern of 4 hallmark dehydration inducible genes. During water withdrawing, transcript abundance of these genes all increased, but to higher levels in amiR408 and PCY-OX plants and lower levels in MIR408-OX and PCY-KO plants in comparison to the wild type (Supplementary Fig. S10). Collectively, these results indicate that miR408 and PCY reciprocally regulate drought avoidance through modulating transpiration.

Figure 8.

PCY and miR408 reciprocally modulate drought resistance. A) Representative WT, MIR408-OX, and PCY-KO plants before and after moderate drought treatment. After treatment, plants were re-watered for 3 days. B) Survival rates of the indicated genotypes after moderate drought and re-watering. Values are means ± SD determined from 3 biological replicates each containing 9 pots (81 plants). Different letters denote genotypes with significant differences (one-way ANOVA, P < 0.001). C) Representative WT, amiR408, and PCY-OX plants before and after severe drought treatment. D) Survival rates of the indicated genotypes after severe drought and re-watering. Values are means ± SDs determined from 3 biological replicates each containing 9 pots (81 plants). Different letters denote genotypes with significant differences (one-way ANOVA, P < 0.001). WT, wild type.

The miR408–PCY module is required for ABA signaling

Finally, we investigated whether ROS homeostasis and stomatal movement controlled by the miR408–PCY module are pertinent to ABA signaling. MYB44 is one of the well-established regulators of ABA signaling with drastic impacts on plant growth and abiotic stress responses (Jung et al. 2008; Shim et al. 2013; Wang et al. 2023). Moreover, a recent study showed that MYB44 promotes ROS accumulation to confer resistance against bacterial pathogens (Wang et al. 2023). Therefore, MYB44 offers an excellent example to genetically test whether the miR408–PCY module is required for ABA signaling. Consistent with the previous reports, we found that the DCF fluorescence intensity was significantly increased in MYB44-OX guard cells relative to the wild type (Fig. 9, A and B, Supplementary Fig. S11). On the contrary, DCF fluorescence was significantly decreased in the myb44 mutants relative to in the wild type (Fig. 9, A and B; Supplementary Fig. S11). We found that the MYB44-OX plants exhibited smaller stomatal apertures (Fig. 9, C and D; Supplementary Fig. S11), retarded vegetative growth (Fig. 9, E to G), and enhanced resistance to severe drought (Fig. 10, A to C) in comparison to the wild type. Consistently, the myb44 mutants exhibited larger stomatal apertures (Fig. 9, C and D; Supplementary Fig. S11), accelerated growth (Fig. 9, E to G), and decreased drought resistance (Fig. 10, A to C) than the wild type. Taken together, these results indicate that MYB44 relays ABA signaling to regulate ROS accumulation, growth, and stress resistance.

Figure 9.

The MYB44-MIR408-PCY pathway regulates stomatal opening and growth. A) Representative fluorescence micrographs showing DCF (green) and chlorophyll (magenta) autofluorescence in the guard cells. Bar, 5 μm. B) Quantification of relative DCF intensity. DCF intensity was normalized against the chlorophyll autofluorescence of the same genotype with the relative DCF intensity in WT set to 1. Data are mean ± SD from 15 stomata. Different letters denote genotypes with significant differences (one-way ANOVA, P < 0.01). C) and D) Comparison of stomatal apertures in the 5 indicated genotypes. Representative stomata in 4-wk-old leaves soaked in the stomatal opening buffer under light C) and the quantification results D) are shown. Bar, 20 μm. Data are mean ± SD from 102 stomata. Different letters denote genotypes with significant differences (one-way ANOVA, P < 0.05). E) Morphology of 5-wk-old plants of the indicated genotypes. Bar, 1 cm. F) and G) Quantification of foliage area F) and fresh weight G). Values are means ± SD determined from 15 individual plants. Different letters denote genotypes with significant differences (one-way ANOVA, P < 0.001). WT, wild type.

Figure 10.

The MYB44-MIR408-PCY pathway regulates drought resistance. A) Representative WT, MYB44-OX, myb44-2, MYB44-OX MIR408-OX, and myb44-2 PCY-OX plants after moderate and severe drought treatments. B) Survival rates of the indicated genotypes after moderate drought and re-watering. Values are means ± SD determined from 3 biological replicates each containing 9 pots (81 plants). Different letters denote genotypes with significant differences (one-way ANOVA, P < 0.001). C) Survival rates of the indicated genotypes after severe drought and re-watering. Values are means ± SD determined from 3 biological replicates each containing 9 pots (81 plants). Different letters denote genotypes with significant differences (one-way ANOVA, P < 0.001). WT, wild type.

To test whether the miR408–PCY module is in the same genetic pathway with MYB44, we generated the MYB44-OX MIR408-OX and myb44 PCY-OX lines (Supplementary Fig. S12). In the MYB44-OX MIR408-OX double overexpression plants, DCF signals in the guard cells were significantly decreased; however, the stomatal aperture and the growth vigor were substantially increased compared to that of MYB44-OX plants (Fig. 9; Supplementary Figs. S13 and S14), indicating that MIR408-OX was able to rescue the growth defects of MYB44-OX. Compared to myb44, the myb44 PCY-OX lines exhibited stronger DCF signals, smaller stomatal aperture, and reduced vegetative growth (Fig. 9; Supplementary Figs. S13 and S14), indicating that PCY is downstream of MYB44. Thus, MYB44, MIR408, and PCY act in the same pathway to regulate ROS accumulation and plant growth.

Consistent with previous reports (Jung et al. 2008; Song et al. 2016), we found that the expression of MYB44 was significantly induced by ABA and drought treatments (Supplementary Fig. S15A). Following ABA treatment, while DCF signals were increased across the genotypes, MIR408-OX and PCY-OX plants were still able to reverse the ROS hyper-accumulation and hypo-accumulation defects of MYB44-OX and myb44, respectively (Supplementary Fig. S15B and C). Consistently, the nearly 100% survival rate of MYB44-OX plants was lowered to about 5% by MIR408-OX in the MYB44-OX MIR408-OX plants subjected to moderate drought treatment (Fig. 10, A and B). In contrast, the 3% survival rate of myb44 plants subjected to severe drought treatment increased to 94% in myb44 PCY-OX plants (Fig. 10, A and C). Examination of the expression profiles of dehydration inducible genes showed the same trends (Supplementary Fig. S10). These results collectively indicate that MYB44 requires the miR408–PCY module to relay ABA signaling for regulating ROS homeostasis and drought resistance.

Discussion

As one of the most conserved miRNAs in land plants, the involvement of miR408 in copper scavenging under limited availability has long been proposed (Abdel-Ghany and Pilon 2008). Subsequent studies further implicated miR408 in regulating chloroplast copper homeostasis, photosynthesis, leaf senescence, and stress tolerance (Zhang et al. 2014; Ma et al. 2015; Pan et al. 2018; Hao et al. 2022). While these studies have established a role for miR408 in leaf development and function, the underlying cellular mechanism and control logic remain largely unclear. In this study, we identified miR408–PCY as a regulatory module operating in the guard cells to modulate stomatal movement and balance leaf growth and drought responses.

miR408 regulates stomatal movement via PCY

We showed that the primary target for miR408 in the leaf is PCY (Fig. 4), which specifically accumulates in the guard cells in association with the chloroplast to regulate ROS accumulation (Fig. 5). By measuring stomatal aperture, leaf surface temperature, and gas exchange, we showed that PCY participates in the regulation of stomatal closure (Fig. 6). Phenotypically, we confirmed that PCY-regulated stomatal movement is both relevant to photosynthesis and leaf growth under water-sufficient conditions (Fig. 7) and ABA-mediated drought resistance (Fig. 8). Moreover, the miR408–PCY module is sufficient to rescue the growth and drought tolerance defects caused by skewed expression of the upstream regulator MYB44 (Figs. 9 and 10). These results are in line with previous studies showing that miR408 enhances leaf growth (Zhang and Li 2013; Zhang et al. 2014; Pan et al. 2018) but compromises tolerance to drought stress (Ma et al. 2015). Thus, miR408 acts as a core component in the reciprocal regulation of growth and drought responses by modulating stomatal movement via regulating PCY expression in the guard cells (Fig. 11).

Figure 11.

A proposed model for miR408 in balancing growth and drought resistance. Under normal growth conditions, miR408 is activated by growth-prone TFs HY5 and SPL7. Accumulation of miR408 silences PCY and reduces ROS to increase stomatal conductance, thus favoring plant growth. Upon drought stress, ABA-responsive TFs, such as MYB44, are activated to transcriptionally repress miR408 expression, which allows PCY and hence ROS to accumulate in high levels in the guard cells, thus inducing stomatal closure, growth suppression, and drought resistance.

Stomata are ancient structures present in the common ancestor of land plants, prior to the divergence of bryophytes and tracheophytes (Lawson and Matthews 2020; Clark et al. 2022). The acquisition of stomata is a key step for early land plants to adapt to and spread through the terrestrial environments. Almost all water lost through transpiration and CO₂ absorbed for photosynthesis pass through the stomatal pores and stomatal movement are therefore the major determinant of plant productivity and fitness under varying environments (Lawson and Matthews 2020; Wang et al. 2022). Elucidation of a role for miR408 in regulating stomatal movement, a vital process in leaf development and function, provides a plausible explanation as to why this miRNA is selected during the long journey of land plant evolution and universally present in all examined land plants (Cuperus et al. 2011; Pan et al. 2018; Guo et al. 2020). Further investigations of miR408's involvement in regulating stomatal movement in other plants should provide more insights into the function and evolution of this conserved miRNA.

Control logic for the miR408–PCY module

In a reconstructed miRNA gene network of Arabidopsis, MIR408 was classified as a typical party hub, which has all its targets locating within a module pertinent to the same biological process (Gao et al. 2022). The input-heavy structure encompassing a party hub entails that MIR408 is regulated by a myriad of TFs to spatiotemporally determine miR408 expression (Gao et al. 2022). Indeed, we have experimentally shown that transcriptional activation of MIR408 by SPL7 and HY5 respectively via the GTAC copper responsive motif and the G-box in the MIR408 promoter is important for miR408 to promote photosynthesis and leaf growth (Zhang and Li 2013; Zhang et al. 2014; Pan et al. 2018). In this study, we further identified the MIR408 promoter as a hot spot of ABA-mediated transcriptional repression (Figs. 1 to 3). It should be noted that all the identified cis-elements in the MIR408 promoter are densely concentrated in an approximately 500 bp region (Fig. 1B) (Zhang et al. 2014; Jiang et al. 2021; Hao et al. 2022). It is possible that the arrangement of different cis-elements with contrasting functions in close proximity may facilitate the MIR408 promoter in acting as a toggle switch to determine the expression status of miR408 (Fig. 11).

Taken together, the control logic for the miR408–PCY module becomes clear. Under growth-friendly environments, TFs prone to growth, such as SPL7 mediating copper homeostasis (Yamasaki et al. 2009; Zhang and Li 2013) and HY5 mediating light signaling (Zhang et al. 2014), activate miR408 expression by binding to their respective cis-elements. Silenced expression of PCY in the guard cells in turn leads to open stomata to facilitate gas exchange and hence photosynthesis (Figs. 6 and 7) (Zhang and Li 2013; Zhang et al. 2014, 2017; Pan et al. 2018). Under stressful conditions, ABA-responsive TFs collectively repress miR408 expression by binding to the alternative stress-responsive cis-elements in the MIR408 promoter, allowing PCY to accumulate in the guard cells to promote stomatal closure as an adaptive strategy (Figs. 5, 6, and 9). Thus, miR408 acts as an integrator of environmental signals to control PCY accumulation, thereby modulating stomatal movement to balance growth and drought resistance.

The miR408–PCY module coordinates copper and ROS homeostasis

PCY belongs to the phytocyanin family of blue copper proteins and was originally identified as a regulator of male reproduction in Arabidopsis (Dong et al. 2005). In this study, we showed that PCY promotes ROS accumulation (Fig. 5), which is one of the primary signals regulating stomatal movement (Pei et al. 2000; Postiglione and Muday 2020). Although the biochemical function of PCY needs further investigation, several lines of evidence indicate that PCY modulates ROS homeostasis by inversely affecting copper allocation to the chloroplast. Overexpression of MIR408, thus downregulation of PCY, was found to increase copper content in the chloroplast and enhance accumulation of plastocyanin, the main cuproprotein in the chloroplast (Zhang et al. 2014; Pan et al. 2018). Degradation of PCY upon light illumination was linked to copper reallocation from the storage vacuoles to the differentiating proplastid during seed germination (Jiang et al. 2021). In senescing leaves, induced PCY in the endomembrane was found as a possible recipient for copper emanating from degrading chloroplasts (Hao et al. 2022). In this study, we observed a direct association of PCY with the chloroplast in the guard cells (Fig. 5). These observations suggest that PCY likely competes for copper with cuproproteins in the chloroplast and increased PCY results in a copper deficient state in the chloroplast, causing the downregulation of chloroplastic cuproproteins such as plastocyanin and the chloroplast-located copper/zinc superoxide dismutase.

Plastocyanin is a water-soluble electron carrier located in the thylakoid lumen that uses copper as the critical cofactor to mediate electron flow from Photosystem II (PSII) to Photosystem I (Katoh 1960; Katoh and Takamiya 1961). Catalyzing light-driven water oxidation and plastoquinone reduction, the photoproduction of ROS is unavoidably associated with PSII (Asada 2006; Pospíšil 2016). For examples, singlet oxygen is produced via the interaction of molecular oxygen with the triplet excited state of chlorophyll in the PSII antennae complex or the recombination of the charge-separated radical pair in the PSII reaction center. Hydrogen peroxide could be produced on the PSII electron donor side by incomplete water oxidation. Superoxide anion radical is produced on the PSII electron acceptor side by electron leakage to molecular oxygen (Asada 2006; Pospíšil 2016). Because the downregulation of plastocyanin likely leads to an increase in the reduced forms of PSII, ROS production is expected to be enhanced under this circumstance. It is also feasible that PCY-induced downregulation of chloroplastic cuproproteins would affect the copper/zinc superoxide dismutase, which facilitates efficient removal of the superoxide generated by photosynthetic electron transport (Pilon et al. 2011). Therefore, the miR408–PCY module could participate in coordinating copper and ROS homeostasis in the guard cell chloroplasts to regulate stomatal movement.

Defense against stress and active growth inhibition are 2 complementary strategies employed by plants to cope with abiotic stresses such as drought (Zhang et al. 2020). Understanding the mechanisms of how plants balance growth and stress responses is the key to increasing crop productivity under stress conditions. Further elucidation of the miR408–PCY module as a component in this reciprocal regulation, such as measurement of water use efficiency, root architecture, and osmotic adjustment across various soil water regimes, should provide insights into how plant growth and stress responses are coordinated and pave the way for engineering stress-resistant and high-yielding crops.

Materials and methods

Plant materials and growth conditions

Arabidopsis (A. thaliana) ecotype Col-0 was used as the wild type in this study. The MIR408-OX, amiR408, PCY-KO, and PCY-OX lines have been described previously (Zhang and Li 2013; Zhang et al. 2014; Jiang et al. 2021). The myb44 T-DNA insertion lines (SALK_039074 and SALK_008606C) were obtained from the Nottingham Arabidopsis Stock Centre. To overexpress MYB44 in Arabidopsis, the respective coding sequences were cloned into the pJim19 and pJim19-eGFP binary vector driven by the CaMV 35S promoter, to generate the MYB44-OX and MYB44-GFP constructs. The resulting constructs were introduced into the wild-type background using Agrobacterium-mediated transformation. To obtain the double mutants, the resulting MIR408-OX and PCY-OX constructs (Zhang and Li 2013; Zhang et al. 2014; Jiang et al. 2021) were introduced into the MYB44-OX or myb44 mutants via Agrobacterium-mediated transformation to generate MYB44-OX MIR408-OX and myb44 PCY-OX plants, respectively. The promoter region of MIR408 was cloned into the p1305.1-LUC vector (35S:LUC) to generate the pMIR408:LUC vector. The resulting 35S:LUC and pMIR408:LUC vectors were introduced into the wild-type background using Agrobacterium-mediated transformation. Transformants were selected with 50 mg L⁻¹ hygromycin (Roche), allowed to propagate to the T₃ generation, and multiple homozygous lines were identified for subsequent experiments. Primers used for cloning and genotyping are listed in Supplementary Table S1.

To grow Arabidopsis seedlings, seeds were surface sterilized and grown on the half-strength Murashige and Skoog medium containing 1% (w/v) sucrose. The plates were vernalized at 4°C for 3 d in the dark and then seedlings were transferred to soil grown in long-day (16 h light/8 h dark) or short-day (10 h light/14 h dark) conditions with 50% relative humidity, 120 μmol m⁻² s⁻² light intensity provided by white fluorescence bulbs and a temperature of 23°C/21°C. Nicotiana benthamiana plants used for transient expression experiments were grown in long-day (16 h light/8 h dark) conditions with a light intensity of 200 μmol m⁻² s⁻², a relative humidity of 50% and a temperature of 25°C/21°C.

Drought tolerance assays

Seedlings were grown on half-strength MS medium for 7 d and were transferred to pots containing the same amount of soil and water as above in short-day conditions. For drought treatment, plants were grown in soil with sufficient water for 21 or 15 d and then water was withheld for 15 d (moderate drought) or 21 d (severe drought) followed by re-watering. After the 3-d recovery period, the survival rates are counted based on surviving plants from 3 biological replicates each containing 9 pots or 81 plants. The control plants were grown in soil with sufficient water for 5 wk.

Leaf infrared thermography

To monitor the leaf temperature, thermal imaging was performed as described previously with slight modification (Chen et al. 2021). Wild-type, PCY-OX, and PCY-KO seedlings grown in MS medium for 7 d were transferred into the soil with sufficient moisture. For normal conditions, seedlings were continuously watered. For drought treatment, seedlings were not watered and the soil naturally dried. 28 d thereafter, infrared energy emitted from the leaves was recorded using a VarioCAM HD infrared camera (InfraTec). Leaf temperature was deduced and temperature distribution images were generated using the FLIR Tools⁺ professional software. For each genotype, plants in 6 different pots were analyzed and similar results were obtained.

Analysis of stomatal aperture

ABA-mediated stomatal closure was analyzed as described previously with minor modification (Xie et al. 2019; Yang et al. 2022a). Briefly, the 5th or 6th leaves of 5-wk-old short-day grown plants were detached. The epidermal strips were immediately peeled from the abaxial surface of leaves and floated in stomatal opening buffer (50 mm KCl and 10 mm MES-KOH, pH 6.15) at room temperature under illumination of 120 μmol m⁻² s⁻¹ for 3 h to reopen the stomata. 10 μM ABA (Sigma-Aldrich, A1049) and control solution were then added to the stomatal opening buffer, and the epidermal strips were further incubated for 1 h under the same conditions. To study the effect of NAC (Sigma-Aldrich, A7250) on PCY-induced stomatal closure, the epidermal strips with reopened stomata were transferred to a stomatal opening buffer containing 1 mm NAC for 1 h, followed by supplementation with or without 10 μM ABA. Stomatal aperture images were taken using a DM 6 B microscope (LEICA). The inner width and length of each stoma were measured by ImageJ, and stomatal apertures are represented as the ratio between the width and the length of a stoma. At least 100 stomata of each genotype were measured.

ChIP-seq and ChIP-qPCR

The ChIP-seq data, TF target gene lists, and visualized peaks in ABA treatment conditions to determine ABA-related TFs binding at the MIR408 promoter were obtained from http://neomorph.salk.edu/dev/pages/shhuang/aba_web/pages/TFnetwork_ABA.php and analyzed as described previously (Song et al. 2016). GO analysis was performed using AgriGO (http://systemsbiology.cau.edu.cn/agriGOv2/) and visualized using R scripts.

ChIP was performed as described previously (Zhou et al. 2018). Briefly, 10-d-old wild-type and MYB44-GFP seedlings were fixed in phosphate buffer saline (pH 7.4) with 1% formaldehyde (v/v) under vacuum for 10 min and then homogenized in liquid nitrogen. Extracted chromatin was sonicated to ∼200 to 500 bp at 4°C using a Bioruptor (Diagenode). For immunoprecipitation, a GFP antibody (Abcam, ab290; dilution, 1:250) and a rabbit IgG antibody (Abcam, ab171870; dilution, 1:250) together with protein A-Sepharose beads (GE Healthcare) were used. After ChIP, equal amount of input and immunoprecipitated DNA was subjected to qPCR analysis using primers listed in Supplementary Table S1. Quantification was based on 3 replicate qPCR reactions performed on the same DNA.

Promoter activity assays

Effects of ABA-responsive TFs on MIR408 promoter were tested using the REN/LUC dual luciferase assays as previously described with modifications (Liu et al. 2014). Briefly, the promoter region of MIR408 and ACTIN2 was cloned into modified p1305.1-LUC-35S-REN vector to generate the reporter construct, respectively. The coding region of selected ABA-responsive TFs and HY5 (positive control) were cloned into the pJIM19 vector to generate the HY5-GFP, NF-YB2-GFP, GBF3-GFP, MYB44-GFP, HB7-GFP, ABF1-GFP, and ABF4-GFP effector vectors. The sequence-confirmed reporter and effector constructs were transformed into Agrobacterium strain GV3101, which was then used to co-infiltrate N. benthamiana leaf epidermal cells. Two days after transfection, dual luciferase reactions were carried out using the Dual-Glo Luciferase Assay System (Promega) following the manufacturer's protocol. Chemiluminescence was detected using a PyLoN2048B luminescence imaging system (Lumazone) and quantified using a Multimode Reader LB 942 luminometer (Berthold). Replicates were individual transfection events.

Effects of ABA and dehydration on MIR408 promoter were determined using luciferase assays as described previously (Yang et al. 2022b). For ABA treatment, the 35S:LUC and pMIR408:LUC seedlings were grown on half-strength MS medium supplemented with the indicated concentrations of ABA for 7 d. For dehydration treatment, the 10-d-old seedlings were removed from half-strength MS medium and placed in open Petri dishes under room temperature, and the luminescence was measured at the indicated times. The seedlings were sprayed with 200 μg mL⁻¹ potassium luciferin (Gold Biotech). Luciferase activity was quantified using a PyLoN2048B luminescence imaging system (Lumazone) with a 5 min exposure time. Three independent experiments were performed each including 15 seedlings.

RT-qPCR

Total RNA was extracted using an RNA extraction kit (Huayueyang) according to the manufacturer's instructions, and reverse transcribed using a Fastking RT kit (Tiangen). qPCR was carried out using the SYBR Green master mix on a QuantStudio 3 Real-Time PCR System (Applied Biosystems) using primers listed in Supplementary Information Table S1. ACTIN7 was used as the internal control and normalization standard. Quantification of differential expression was based on 3 replicate qPCR reactions performed on the same cDNA.

GUS staining

The histochemical staining of GUS activity was performed as described previously (Zhang and Li 2013). Briefly, the rosette leaves, petioles, and pedicels were detached and immersed in GUS staining buffer [0.1 m NaH₂PO₄ (pH 7.0), 10 mm EDTA, 0.1% Triton X-100, 0.5 mm K₃Fe(CN)₆, 0.5 mm K₄Fe(CN)₆·3H₂O, 0.5 mg/mL X-Gluc] and incubated overnight at 37°C. Following the removal of the GUS staining buffer, tissues were washed with several changes of 75% ethanol until chlorophyll was no longer visible. Images of the GUS staining pattern were taken with a DM 6 B microscope (LEICA).

ROS staining

DCF imaging and quantification were performed as described previously with minor modifications (Watkins et al. 2017). Epidermal peels from leaves of 4-wk-old plant were prepared and then covered with stomatal opening buffer for 3 h under white light to fully open stomata. This was followed by incubation with 10 μM ABA or ethanol for 1 h. Leaf peels were then removed and incubated with H₂DCF-DA for 15 min. Guard cells in the epidermal peels were analyzed using an LSM 800 laser scanning confocal microscope (ZEISS).

Fluorescence microscopy for protein subcellular localization

Fluorescence microscopy analysis of proteins stably expressed in Arabidopsis was performed on 7-d-old seedlings. Cotyledon epidermal cells expressing PCY-GFP, Q4-GFP (ABRC #CS84728) (Cutler et al. 2000), and COPT5-GFP (Jiang et al. 2021) were analyzed for GFP and chlorophyll autofluorescence using an LSM 800 laser scanning confocal microscope (ZEISS) as previously described (Yang et al. 2022b). The GFP signal was excited at 488 nm and detected at 490 to 575 nm with a laser intensity of approximately 1% and a gain value around 750. The chlorophyll autofluorescence was excited at 488 nm and detected at 637 to 750 nm with a laser intensity of approximately 1% and a gain value around 800.

Measurements of stomatal conductance and photosynthetic performance

Five-week-old plants grown under the short-day condition were used for measurements of photosynthetic gas exchange. Stomatal conductance and other parameters were measured from 10 individual leaves. The measurements were performed on a fixed area of the youngest fully expanded rosette leaves using the LI-6800 Portable Photosynthesis System (Li-Cor) outfitted with a standard 2 cm² leaf chamber for small plants according to manufacturer's instructions. The measurements were performed using the internal light source of LI-6800 at a leaf temperature of 22°C and a relative humidity of 50%. The CO₂ level in the leaf chamber was set to 400 µmol mol⁻¹. Light-intensity-dependent NPR curve was measured with a range of intensities from 0, 30, 60, 100, 150, 200, 250, 300, 350, 400, 500, 600, 800 to 900 µmol m⁻¹ s⁻¹.

Statistical analysis

Statistical analyses were conducted as described in the text and figure legends. Statistical data are provided in Supplementary Data Set S1.

Accession numbers

Sequence data from this article can be found in the Arabidopsis Genome Initiative or GenBank/EMBL databases under the following accession numbers: MIR408 (At2g47015), PCY (At2g02850), and MYB44 (At5g67300). T-DNA insertion mutants are myb44-1 (SALK_008606C) and myb44-2 (SALK_039074).

Author contributions

L.L. conceived this study. Y.Y., L.X., C.H., M.W., Y.T., and Y.Z. carried out the experiments. Y.Y. and Y.S. analyzed the data. Y.Y. and L.L. wrote the manuscript.

Supplementary data

The following materials are available in the online version of this article.

Supplementary Figure S1. Analysis of target genes of ABA-responsive TFs. Supports Fig. 1.

Supplementary Figure S2. Constructs and controls for the dual luciferase assay. Supports Fig. 2.

Supplementary Figure S3. Characterization of MYB44 related lines. Supports Figs. 2 and 9.

Supplementary Figure S4. Expression patterns of MIR408 related genes. Supports Fig. 4.

Supplementary Figure S5. Comparison of PCY localization with endomembrane markers. Supports Fig. 5.

Supplementary Figure S6. miR408 promotes stomatal opening. Supports Fig. 9.

Supplementary Figure S7. Characterization of PCY-OX and PCY-KO lines regarding stomatal closure. Supports Fig. 6.

Supplementary Figure S8. PCY negatively regulates seedling growth. Supports Fig. 7.

Supplementary Figure S9. Design for drought treatments. Supports Figs. 8 and 10.

Supplementary Figure S10. Transcript levels of representative drought responsive genes. Supports Fig. 8.

Supplementary Figure S11. Analyses of MYB44 using independent lines. Supports Fig. 9.

Supplementary Figure S12. Characterization of independent MYB44-OX MIR408-OX and myb44 PCY-OX lines. Supports Fig. 9.

Supplementary Figure S13. Genetic analysis of the MYB44-MIR408-PCY pathway using independent lines. Supports Fig. 9.

Supplementary Figure S14. Analysis of MYB44-OX MIR408-OX and myb44 PCY-OX seedlings. Supports Fig. 9.

Supplementary Figure S15. miR408–PCY is required for ABA-induced ROS accumulation. Supports Fig. 9.

Supplementary Table S1. Oligonucleotide sequences of the primers used in this study.

Supplementary Data Set 1. Tables for statistical analysis.

Funding

This work was supported by a grant from the National Natural Science Foundation of China (32370368). Y.Y. was supported in part by the Postdoctoral Fellowship of Peking-Tsinghua Center for Life Sciences.

Data availability

The data underlying this article are available in the article and in its online supplementary material.

Dive Curated Terms

The following phenotypic, genotypic, and functional terms are of significance to the work described in this paper:

• LAC12 Gramene: AT5G05390

• LAC12 Araport: AT5G05390

• LAC13 Gramene: AT5G07130

• LAC13 Araport: AT5G07130

• SPL7 Gramene: At5g18830

• SPL7 Araport: At5g18830

• MIR408 Gramene: At2g47015

• MIR408 Araport: At2g47015

• NAC CHEBI: CHEBI:7421

• leaf AmiGo: PO:0025034