Transcription Factor ZmNAC20 Improves Drought Resistance by Promoting Stomatal Closure and Activating Expression of Stress-Responsive Genes in Maize

Drought is a major environmental threat that limits crop growth, development, and productivity worldwide. Improving drought resistance with genetic engineering methods is necessary to tackle global climate change. It is well known that NAC (NAM, ATAF and CUC) transcription factors play a critical role in coping with drought stress in plants. In this study, we identified an NAC transcription factor ZmNAC20, which regulates drought stress response in maize. ZmNAC20 expression was rapidly upregulated by drought and abscisic acid (ABA). Under drought conditions, the ZmNAC20-overexpressing plants had higher relative water content and survival rate than the wild-type maize inbred B104, suggesting that overexpression of ZmNAC20 improved drought resistance in maize. The detached leaves of ZmNAC20-overexpressing plants lost less water than those of wild-type B104 after dehydration. Overexpression of ZmNAC20 promoted stomatal closure in response to ABA. ZmNAC20 was localized in the nucleus and regulated the expression of many genes involved in drought stress response using RNA-Seq analysis. The study indicated that ZmNAC20 improved drought resistance by promoting stomatal closure and activating the expression of stress-responsible genes in maize. Our findings provide a valuable gene and new clues on improving crop drought resistance.


Introduction
Drought limits plant growth and reduces crop production worldwide. With global warming, improving drought resistance is of great importance for crop breeding [1,2]. A variety of strategies have evolved to survive under drought conditions. Plants improve drought resistance mainly by promoting stomatal closure, altering root architecture, adjusting the contents of osmotic substances, and increasing the activities of the antioxidant enzymes [3]. When facing drought stress, plant cells sense the stimuli and activate stress response genes to cope with the stress [4]. It has been demonstrated in numerous studies that transcription factors, such as NAC, WRKY, basic leucine zipper (bZIP), homeodomain-leucine zipper (HD-Zip), drought-response elements binding proteins/Crepeat binding factor (DREB/CBF), and MYB, play crucial roles in enhancing drought resistance in plants [5][6][7]. Many stress-responsive genes are activated by transcription factors [6,7]. NAC transcription factors are one of the largest transcription factor families in the plant, with a highly conserved NAC domain at the N-terminal [8]. The name NAC is derived from three transcription factors: NAM (no apical meristem) in Petunia hybrida, ATAF1/2 (Arabidopsis thaliana activating factor), and CUC2 (cup-shaped cotyledon) in Arabidopsis thaliana [9][10][11]. NAC transcription factors are key targets for crop improvement creases drought resistance in Arabidopsis and rice [28]. ZmPP2C-A10 functions as a negative regulator of drought resistance. Deletion of an endoplasmic reticulum stress response element (ERSE) in ZmPP2C-A10 increases drought resistance [29]. ABA signaling is critical for plants to cope with drought stress [25,30]. There are many transcription factors involved in ABA signaling to regulate drought resistance in maize. The BRI1-EMS suppressor 1 (BES1)/brassinazole-resistant 1 (BZR1) transcription factor positively regulates drought resistance by binding to E-box to induce the expression of downstream stress-related genes. Heterologous expression of ZmBES1/BZR1-5 in transgenic Arabidopsis thaliana results in decreased ABA sensitivity, facilitates shoot growth and root development, and enhances drought resistance with lower malondialdehyde (MDA) content and relative electrolyte leakage (REL) under osmotic stress [31]. The expression of ZmbZIP33 is strongly upregulated by drought and ABA. Overexpression of ZmbZIP33 causes an accumulation of ABA content and improves drought resistance in Arabidopsis [32]. Lateral organ boundaries domain (LBD) proteins are plant-specific transcription factors. ZmLBD5 negatively regulates drought resistance by impairing ABA synthesis [33].
Stomata are essential structures for plants to control gas exchange and water status. Under drought conditions, plants can reduce water loss by adjusting the size of the stomatal aperture. Stomatal closure is regulated by various signaling molecules, such as ABA, ROS, and Ca 2+ [34]. Several critical factors that regulate stomatal closure in response to drought stress have been characterized in maize. ZmCPK35 (Ca 2+ -dependent protein kinases) and ZmCPK37 (calcium-dependent protein kinases) regulate the drought resistance of maize by regulating the activity of ZmSLAC1 (the S-type anion channel protein) in the guard cells of maize [35]. ZmSLAC1 is specifically expressed in maize guard cells and participates in regulating stomatal closure under drought conditions. zmslac1 mutant exhibits drought-sensitive phenotype. ZmCPK35 and ZmCPK37 are expressed in maize guard cells. ZmCPK35 and ZmCPK37 can interact with ZmSLAC1 on the cytoplasmic membrane and regulate ion channel activity. In the guard cells of zmslac1 and zmcpk37 mutants, S-type anion channel currents activated by ABA are significantly inhibited, while overexpression of ZmCPK35 and ZmCPK37 could significantly enhance ABA-activating S-type anion channel currents in guard cells [35]. The gene knockout mutant of zmmpkl1 (mitogen-activated protein kinase) is less sensitive to severe drought. The stomatal aperture of ZmMPKL1-overexpressing plants is higher than those of control plants, which leads to faster water loss. The stomatal aperture of the zmmpkl1 mutant is smaller than that of the control plant. ZmMPKL1 affects the response of plants to drought by regulating ABA homeostasis in plants [36].
In this study, we identified a NAC transcription factor gene ZmNAC20 in maize. ZmNAC20 was induced and upregulated by drought stress and ABA. ZmNAC20 was localized in the nucleus and enhanced drought resistance in maize. ZmNAC20 promoted stomatal closure in response to ABA. Our results suggest that ZmNAC20 plays an important role in drought stress response and stomatal movement in maize.

ZmNAC20 Is Upregulated by Drought Stress and ABA
ZmNAC20 (Zm00001eb288360 in the version of Zm-B73-REFERENCE-NAM-5.0, Zm00001d038221 in version of Zm-B73-REFERENCE-GRAMENE-4.0, GRMZM2G180328 in version of B73 RefGen_v3), which has been named in the previous study [37], encodes a NAC-type transcription factor. ZmNAC20 belongs to a large number of members of the family in maize [37]. A previous study performed phylogenetic analysis using the amino acid sequences and indicated that the closest two identified homologous genes of ZmNAC20 (GRMZM2G180328) were Sb09g020750 in Sorghum bicolor and LOC_Os05g34830 in rice [13]. In the maizeGDB database (www.maizegdb.org, accessed on 7 February 2023), the GO annotations indicated that ZmNAC20 (Zm00001eb288360) was involved in multiple processes, including positive regulation of response to water deprivation, positive regulation of response to salt stress, regulation of defense response to fungus, and response to cold ( Figure S1). A previous study indicated that ZmNAC20 was upregulated by treatment with PEG, which could mimic drought stress [37]. We assumed that ZmNAC20 was likely involved in drought stress response.
We performed quantitative reverse-transcription PCR (qRT-PCR) to confirm whether ZmNAC20 expression was regulated by drought stress. When the maize seedlings developed three leaves, the aerial parts were cut and subjected to dehydration stress. The results showed that the expression of ZmNAC20 was significantly upregulated after dehydration stress ( Figure 1A), indicating that drought stress promoted ZmNAC20 expression. ABA is a critical hormone that regulates drought stress response in plants [25]. The leaves of maize growing for about 12 days were excised and treated with different concentrations of ABA for 1 h. By using qRT-PCR, we found that treatment with ABA elevated the expression level of ZmNAC20 ( Figure 1B). We also examined the expression of ZmNAC20 treated with 1 µM ABA for 0 h, 1 h, 2 h, and 3 h. ZmNAC20 was upregulated by ABA at different times ( Figure 1C). These results suggested that ZmNAC20 expression was regulated by both drought stress and ABA.
LOC_Os05g34830 in rice [13]. In the maizeGDB database (www.maizegdb.org, accessed on 7 February 2023), the GO annotations indicated that ZmNAC20 (Zm00001eb288360) was involved in multiple processes, including positive regulation of response to water deprivation, positive regulation of response to salt stress, regulation of defense response to fungus, and response to cold ( Figure S1). A previous study indicated that ZmNAC20 was upregulated by treatment with PEG, which could mimic drought stress [37]. We assumed that ZmNAC20 was likely involved in drought stress response.
We performed quantitative reverse-transcription PCR (qRT-PCR) to confirm whether ZmNAC20 expression was regulated by drought stress. When the maize seedlings developed three leaves, the aerial parts were cut and subjected to dehydration stress. The results showed that the expression of ZmNAC20 was significantly upregulated after dehydration stress ( Figure 1A), indicating that drought stress promoted ZmNAC20 expression. ABA is a critical hormone that regulates drought stress response in plants [25]. The leaves of maize growing for about 12 days were excised and treated with different concentrations of ABA for 1 h. By using qRT-PCR, we found that treatment with ABA elevated the expression level of ZmNAC20 ( Figure 1B). We also examined the expression of ZmNAC20 treated with 1 μM ABA for 0 h, 1 h, 2 h, and 3 h. ZmNAC20 was upregulated by ABA at different times ( Figure 1C). These results suggested that ZmNAC20 expression was regulated by both drought stress and ABA.

Overexpression of ZmNAC20 Enhances Drought Resistance in Maize
To understand whether increased transcript levels of ZmNAC20 enhanced drought resistance, we generated two transgenic maize which overexpressed ZmNAC20 under the constitutive promoter ZmUbiquitin1 (Ubi). The two transgenic maize lines were named ZmNAC20-OE1 and ZmNAC20-OE2. Using qRT-PCR assay, we found that the transcript levels of ZmNAC20 in the two transgenic maize lines were significantly higher than that in the wild type (WT, B104) ( Figure 1D). After drought treatment, it was found that the leaves of ZmNAC20-overexpression plants were slightly yellow and curly, whereas the leaves of WT plants were severely wilted or even began to die (Figure 2A). After 3 days of rewatering, the leaves of ZmNAC20-overexpressing plants gradually expanded and turned green, and the survival rate reached about 80%, much higher than the WT plants ( Figure 2A,B). Furthermore, both the relative water content, the fresh weight, and the dry weight of ZmNAC20-overexpressing plants were significantly higher than those of WT plants ( Figure 2C,D and Figure S2). These results demonstrated that overexpression of ZmNAC20 improved drought resistance in maize.

Overexpression of ZmNAC20 Promotes Stomatal Closure
Stomata are key structures to cope with drought in plants [3,38]. Plants regulated water loss mainly through modulation of the opening and closure of stomata [25]. Rapid stomatal closure is an effective way to reduce water loss and improve the drought resistance of plants [3,25]. The detached leaves of ZmNAC20-overexpressing plants showed a slower water loss rate than those of WT plants ( Figure 3A). After 180 min of dehydration stress treatment, the leaves of WT had severely curved, whereas the leaves of ZmNAC20overexpressing plants displayed a slightly curled phenotype ( Figure 3B). These results suggest that overexpression of ZmNAC20 could retain more water in leaves and enhance drought resistance in maize. Next, we wondered if ZmNAC20 regulates water loss by modulation of stomatal closure. ABA plays a pivotal role in stomatal closure in response to drought stress, and ZmNAC20 was upregulated by ABA ( Figure 1B,C). We performed the experiment to test whether ZmNAC20 affected the stomatal closure in response to ABA. The stomatal aperture was determined by the ratio of stomatal width to length. It was shown that overexpression of ZmNAC20 reduced the size of the stomatal aperture under 1 µM and 10 µM ABA treatment ( Figure 3C,D). The stomata in ZmNAC20-overexpressing plants closed more rapidly than those of WT plants after ABA treatment. Stomatal closure in ZmNAC20-overexpressing plants was more sensitive to ABA compared with that of WT plants. These results indicated that ZmNAC20 promoted stomatal closure and impaired water loss, resulting in enhanced drought resistance in maize.

ZmNAC20 Regulates Multiple Biological Pathways under Drought Stress
Because ZmNAC20 was a NAC transcription factor, we assumed that ZmNAC20 might localize in the nucleus. To test the hypothesis, p35S:ZmNAC20-GFP and the nucleus localized construct 35S:H2B-mCherry were together transformed into leaves of Nicotiana benthamiana by agrobacterium-mediated infiltration. The ZmNAC20-GFP signal was colocalized with the H2B-mCherry signal; a strong yellow fluorescence signal was only detected in the nucleus, whereas the signals of the control construct 35S:GFP were detected throughout the cell (Figure 4). These results indicated that ZmNAC20 was localized in the nucleus.
To uncover the molecular pathways regulated by ZmNAC20 under drought stress, we performed RNA-seq analysis using the detached leaves of ZmNAC20-OE1 and WT treated with dehydration stress for 3 h. Three biological repeats of the WT (WT-A, WT-B, WT-C) and ZmNAC20-OE1 (OE1-A, OE1-B, OE1-C) were used for RNA-seq analyses. Twofold gene transcript levels with the adjusted p-value (q value) cutoff of 0.05 between ZmNAC20-OE1 and WT was used as the criterion for differential expression. Compared with WT, 1361 genes were upregulated and 913 genes were downregulated in ZmNAC20-OE1 ( Figure 5A, Tables S1 and S2). To reveal ZmNAC20-mediated pathways, we performed Gene Ontology (GO) enrichment analysis and KEGG pathway enrichment analysis. GO analysis showed that the 1361 upregulated genes were mainly involved in multiple biological processes, including responses to abiotic stimulus, photosynthesis, hormone, cold, water deprivation, salt stress, and osmotic stress ( Figure 5B). The 913 downregulated genes were mainly involved in the processes of RNA modification, ribosome biogenesis, ribosome assembly, rRNA processing, protein import into the nucleus, and translation ( Figure S3). KEGG enrichment analysis indicated that the upregulated genes participated in the pathways of photosynthesis, biosynthesis of secondary metabolites, carbon fixation in photosynthetic organisms, benzoxazinoid biosynthesis, carotenoid biosynthesis, carbon metabolism, and flavonoid biosynthesis ( Figure 5C). The downregulated genes were involved in the pathways of ribosome biogenesis in eukaryotes monoterpenoid biosynthesis, tryptophan metabolism, and linoleic acid metabolism ( Figure S4). These results suggested that ZmNAC20 regulated multiple biological pathways under drought stress in maize.

Overexpression of ZmNAC20 Enhances Drought Resistance in Maize
To understand whether increased transcript levels of ZmNAC20 enhanced drought resistance, we generated two transgenic maize which overexpressed ZmNAC20 under the constitutive promoter ZmUbiquitin1 (Ubi). The two transgenic maize lines were named ZmNAC20-OE1 and ZmNAC20-OE2. Using qRT-PCR assay, we found that the transcript levels of ZmNAC20 in the two transgenic maize lines were significantly higher than that in the wild type (WT, B104) ( Figure 1D). After drought treatment, it was found that the leaves of ZmNAC20-overexpression plants were slightly yellow and curly, whereas the leaves of WT plants were severely wilted or even began to die (Figure 2A). After 3 days of rewatering, the leaves of ZmNAC20-overexpressing plants gradually expanded and turned green, and the survival rate reached about 80%, much higher than the WT plants (Figure 2A,B). Furthermore, both the relative water content, the fresh weight, and the dry weight of ZmNAC20-overexpressing plants were significantly higher than those of WT plants ( Figures 2C,D and S2). These results demonstrated that overexpression of ZmNAC20 improved drought resistance in maize. Phenotype of ZmNAC20-overexpressing transgenic maize. (A) Drought resistance of transgenic maize overexpressing ZmNAC20. Images were taken before and after drought stress treatment and rewatering. WT represented the wild-type B104 maize and served as the control. ZmNAC20-OE1 and ZmNAC20-OE2 represented the two ZmNAC20-overexpressing maize lines. Bars, 5 cm. (B) Statistical data for the survival rate of WT, ZmNAC20-OE1, and ZmNAC20-OE2. The survival rate Figure 2. Phenotype of ZmNAC20-overexpressing transgenic maize. (A) Drought resistance of transgenic maize overexpressing ZmNAC20. Images were taken before and after drought stress treatment and rewatering. WT represented the wild-type B104 maize and served as the control. ZmNAC20-OE1 and ZmNAC20-OE2 represented the two ZmNAC20-overexpressing maize lines. Bars, 5 cm. (B) Statistical data for the survival rate of WT, ZmNAC20-OE1, and ZmNAC20-OE2. The survival rate was obtained from 30 seedings in three independent tests. (C) The relative water content of maize leaves under drought stress treatment. The second leaves were cut from drought-treated plants. SE values were calculated from three biological replicates and more than 10 plants were examined in each replicate. (D) Fresh weight of maize leaves under drought stress treatment. SD values were calculated from three biological replicates and more than 10 plants were examined in each replicate. Asterisks indicate significant differences: ** p < 0.01 (Student's t-test).

ZmNAC20 Regulates Multiple Biological Pathways under Drought Stress
Because ZmNAC20 was a NAC transcription factor, we assumed that ZmNAC20 might localize in the nucleus. To test the hypothesis, p35S:ZmNAC20-GFP and the nucleus localized construct 35S:H2B-mCherry were together transformed into leaves of Nicotiana benthamiana by agrobacterium-mediated infiltration. The ZmNAC20-GFP signal was colocalized with the H2B-mCherry signal; a strong yellow fluorescence signal was only detected in the nucleus, whereas the signals of the control construct 35S:GFP were detected throughout the cell (Figure 4). These results indicated that ZmNAC20 was localized in the nucleus.

ZmNAC20 Regulates the Expression of Stress-Responsive Genes
RNA-seq analysis revealed that overexpression of ZmNAC20 affected the expression of many genes. We selected 9 upregulated genes involved in drought stress response for detailed analysis. The gene symbol name used in this study referred to the maizeGDB database (www.maizegdb.org, accessed on 7 February 2023). It is well known that the transcription factors, such as APETALA2/Ethylene Responsive Element Binding Protein (AP2/EREBP), WRKY, and bZIP, play key roles in drought stress response [7,[39][40][41]. Zm00001eb168040 (ZmEREB106), Zm00001eb276700 (ZmEREB145), and Zm00001eb021440 (ZmEREB188) are putative AP2/EREBP transcription factor genes. Zm00001eb344160 (ZmWRKY106) is a transcription factor that includes a WRKY DNA-binding domain. Zm00001eb366900 (Zm-bZIP75) is a putative bZIP transcription factor. Overexpression of ZmNAC20 promoted the expression of the four genes by qRT-PCR ( Figure 6). After treatment with ABA, ZmEREB106, ZmEREB145, ZmEREB188, and ZmbZIP75 were upregulated by ABA, whereas ZmWRKY106 was downregulated by ABA (Figure 7). To uncover the molecular pathways regulated by ZmNAC20 under drought stre we performed RNA-seq analysis using the detached leaves of ZmNAC20-OE1 and W treated with dehydration stress for 3 h. Three biological repeats of the WT (WT-A, WT WT-C) and ZmNAC20-OE1 (OE1-A, OE1-B, OE1-C) were used for RNA-seq analys Twofold gene transcript levels with the adjusted p-value (q value) cutoff of 0.05 betwe ZmNAC20-OE1 and WT was used as the criterion for differential expression. Compar with WT, 1361 genes were upregulated and 913 genes were downregulated in ZmNAC OE1 ( Figure 5A, Tables S1 and S2). To reveal ZmNAC20-mediated pathways, we p formed Gene Ontology (GO) enrichment analysis and KEGG pathway enrichment ana sis. GO analysis showed that the 1361 upregulated genes were mainly involved in mu ple biological processes, including responses to abiotic stimulus, photosynthesis, h mone, cold, water deprivation, salt stress, and osmotic stress ( Figure 5B). The 913 dow regulated genes were mainly involved in the processes of RNA modification, riboso biogenesis, ribosome assembly, rRNA processing, protein import into the nucleus, a translation ( Figure S3). KEGG enrichment analysis indicated that the upregulated gen participated in the pathways of photosynthesis, biosynthesis of secondary metabolit carbon fixation in photosynthetic organisms, benzoxazinoid biosynthesis, carotenoid b synthesis, carbon metabolism, and flavonoid biosynthesis ( Figure 5C). The downreg lated genes were involved in the pathways of ribosome biogenesis in eukaryotes mon terpenoid biosynthesis, tryptophan metabolism, and linoleic acid metabolism (Figure S These results suggested that ZmNAC20 regulated multiple biological pathways und drought stress in maize. Dehydrin is a well-known protein which accumulates massively under drought stress and helps plants improve resistance to stresses [42]. Previous studies have demonstrated many dehydrins involved in drought resistance improvement [43]. For example, Medicago truncatula MtCAS31 (cold acclimation-specific 31) encodes a dehydrin which acts as a positive regulator of drought response [43,44]. In the ZmNAC20-overexpressing lines, Zm00001eb376710 (ZmDHN2), encoding a dehydrin protein showed increased transcript level compared with the wild-type B104, suggesting that ZmNAC20 promoted ZmDHN2 expression ( Figure 6). Moreover, we found that ZmDHN2 was upregulated by ABA ( Figure 7). Zm00001eb283860 (ZmGA2OX6) encodes a gibberellin 2-beta-dioxygenase that may deactivate gibberellin (GA). A previous study showed that low GA activity promoted stomatal closure and enhanced drought resistance [45]. Overexpression of ZmNAC20 increased the transcript level of ZmGA2OX6, suggesting that ZmNAC20 might regulate drought resistance through modulation of GA signaling via ZmGA2OX6 ( Figure 6). In addition, ZmGA2OX6 was also upregulated by ABA, indicating that ABA participated in the expression of ZmGA2OX6 (Figure 7). Zm00001eb130550 (ZmSWEET17A) encodes a sugar transporter. SUGAR WILL EVENTUALLY BE EXPORTED TRANSPORTER (SWEET) proteins promote the transport of different sugars over cellular membranes and control both inter and intracellular distribution of sugars [46]. In Arabidopsis thaliana, knockout of SWEET17 resulted in impaired drought resistance [46], and mutation of SWEET11 and SWEET12 also exhibited reduced drought resistance [47]. ZmSWEET17A was regulated by ZmNAC20 and ABA ( Figures 6D and 7D), suggesting that ZmNAC20 might regulate drought resistance by modulating the ZmSWEET17A-mediated sugar transporter. Zm00001eb076200 (ZmPx15) encodes a peroxidase. The water-deficit condition caused the excessive production and accumulation of reactive oxygen species (ROS), which damaged the plant cells. Peroxidase acted as an antioxidant to scavenge ROS [48]. Overexpression of ZmNAC20 enhanced the transcript level of ZmPx15, implying that ZmNAC20 might improve drought resistance by modulation of ZmPx15 ( Figure 6). ABA also promoted the expression of ZmPx15, suggesting that ABA regulated the expression of ZmPx15 (Figure 7). These results suggested that ZmNAC20 regulated expressions of various genes involved in drought stress response. ZmNAC20 improved drought resistance by modulation of multiple genes, which constituted a complex network.

ZmNAC20 Regulates the Expression of Stress-Responsive Genes
RNA-seq analysis revealed that overexpression of ZmNAC20 affected the expression of many genes. We selected 9 upregulated genes involved in drought stress response for detailed analysis. The gene symbol name used in this study referred to the maizeGDB database (www.maizegdb.org). It is well known that the transcription factors, such as APETALA2/Ethylene Responsive Element Binding Protein (AP2/EREBP), WRKY, and bZIP, play key roles in drought stress response [7,[39][40][41]. Zm00001eb168040 (ZmEREB106), Zm00001eb276700 (ZmEREB145), and Zm00001eb021440 (ZmEREB188) are putative AP2/EREBP transcription factor genes. Zm00001eb344160 (ZmWRKY106) is a transcription

Discussion
Drought is among the major environmental factors that affect plant growth and crop yield [3]. To obtain stress resistance, many transcription factor genes are activated under drought conditions. The transcription factors receive the upstream drought stress signa and activate expressions of many stress-responsive genes [49]. These molecular responses help plants rapidly adapt to the water-deficit stress environment [25]. The transcription factors-mediated networks are critical and necessary for plants to survive under drough conditions [5]. It has been demonstrated that NAC transcription factors play crucial roles in drought stress response [8,13,14,17,20]. In this study, we reported a NAC transcription factor gene ZmNAC20, which was upregulated by drought stress (Figure 1A,B). Overex pression of ZmNAC20 resulted in an increased survival rate and a higher level of relative water content ( Figure 2B,C), suggesting that ZmNAC20 functioned as a positive factor in improving drought resistance. Previous studies have reported that several maize

Discussion
Drought is among the major environmental factors that affect plant growth and crop yield [3]. To obtain stress resistance, many transcription factor genes are activated under drought conditions. The transcription factors receive the upstream drought stress signal and activate expressions of many stress-responsive genes [49]. These molecular responses help plants rapidly adapt to the water-deficit stress environment [25]. The transcription factors-mediated networks are critical and necessary for plants to survive under drought conditions [5]. It has been demonstrated that NAC transcription factors play crucial roles in drought stress response [8,13,14,17,20]. In this study, we reported a NAC transcription factor gene ZmNAC20, which was upregulated by drought stress (Figure 1A,B). Overexpression of ZmNAC20 resulted in an increased survival rate and a higher level of relative water content ( Figure 2B,C), suggesting that ZmNAC20 functioned as a positive factor in improving drought resistance. Previous studies have reported that several maize ZmNACs genes, including ZmNAC111 [13], ZmNAC33 [20], ZmNAC55 [14], ZmNAC49 [17], ZmNST3 [15], ZmNUT1 [21], ZmNAC080308 [22], ZmSNAC13 [23], ZmNAC84 [24], and ZmNAC48 [16], contributed to improving drought resistance. In maize, a large number of ZmNACs genes have been identified, but the function of most of them has remained elusive until now. Considering the important function of NACs in drought stress response, more efforts on revealing the function of ZmNACs genes in maize should be made in future studies.
Stomata are important structures to control water loss and gas exchange [38,50]. Under drought conditions, stomatal closure is regulated by multiple signals, such as ABA, ROS, Ca 2+ , and H 2 S, [34,38,51]. Among these signals, ABA is the center molecule to control stomatal closure in response to drought stress [25,26]. ZmNAC20 was upregulated by drought stress and ABA ( Figure 1A,B). Overexpression of ZmNAC20 accelerated the closure of stomata in response to ABA. The ZmNAC20-overexpressing transgenic maize showed a smaller stomatal aperture and lost less water than the wild-type B104 ( Figure 3A,C). These results suggested that ZmNAC20 contributed to promoting ABA-mediated stomatal closure in response to drought stress. ZmNAC20 partially improved drought resistance by promoting stomatal closure. In maize, ZmNAC48 contributes to improving drought resistance by promoting stomatal closure, whereas ZmNAC49 enhances drought resistance by the decline of stomatal conductance and stomatal density [16,17]. ZmNST3 increases drought resistance by activating the expression of genes related to cellulose synthesis in the secondary cell wall [15]. ZmNUT1 improves drought resistance by affecting water transportation via regulating the expression of the cellulose synthase gene and cysteine proteolytic enzyme gene [21]. These findings indicated that ZmNACs transcription factors improved drought resistance by modulation of various pathways, suggesting that the function of NACs has diversity and complexity.
Using RNA-seq analysis, we obtained a list of genes regulated by ZmNAC20 (Tables S1 and S2). GO enrichment analyses indicated that ZmNAC20 regulated multiple pathways that are mainly involved in abiotic stimulus, response to water deprivation, and response to hormones ( Figure 5B). RNA-seq analysis indicated that ZmNAC20 regulated the expression of various stress-responsive genes, such as the transcription factor genes ZmEREB106, ZmEREB145, ZmEREB188, ZmWRKY106, and ZmbZIP75, the dehydrin gene ZmDHN2, the GA dioxygenase gene ZmGA2OX6, the sugar transporter gene ZmSWEET17A, and antioxidant enzyme gene ZmPx15 (Figures 6 and 7). These results suggested that ZmNAC20 improved drought resistance by modulation of multiple pathways. ZmNAC20 formed a network to cope with drought stress. The molecular response induced by ZmNAC20 promoted plants to adapt to the water-deficit environment.
There are many limitations when using transcription factors to improve drought resistance in crops. We should consider the fact that many transcription factors inhibit plant growth and development, while improving drought resistance. Stomatal closure prevents water loss and enhances drought resistance but also limits CO 2 intake and photosynthesis [3]. For example, overexpression of ZmNAC49 enhances drought resistance in maize, but also significantly decreases stomatal conductance and stomatal density. ZmNAC49 overexpression affects the expression of genes related to stomatal development [17]. Overexpression of ZmNAC48 improves drought resistance by enhancing stomatal closure [16]. Less studies focus on revealing the mechanism of plants maintaining growth under drought conditions. It is necessary to identify the genes that increase drought resistance but do not limit plant growth and development. Previously, a study reported that overexpression of a wheat ABA receptor gene TaPYL4 improves drought resistance by reducing stomatal conductance but does not affect wheat growth and development. This is because overexpression of TaPYL4 increases water use efficiency and enhances photosynthesis [52]. Future studies should pay more attention to those transcription factors that can improve drought resistance without penalizing plant growth. In this study, we did not detect whether ZmNAC20 affected water use efficiency and photosynthesis. Further research will be focused on revealing the function of ZmNAC20 to maintain growth under drought conditions. Taken together, our work revealed the function of ZmNAC20 in drought stress response in maize. The study provides valuable insights into the role of ZmNAC20 in promoting drought resistance and shed light on the molecular mechanisms underlying plant adaptation to water-deficit stress.

Plant Materials and Growth Conditions
The overexpression transgenic plants in this study were derived from the maize (Zea mays L.) inbred line B104, which is a public, transformable maize inbred line and has been widely used in recent studies [15,[53][54][55]. To analyze the function of ZmNAC20 in drought stress response, the coding sequence of ZmNAC20 was cloned and inserted into a modified pCambia3300 vector. ZmUbiquitin1 (ZmUbi) was used as the promoter to drive ZmNAC20 expression. The ZmNAC20-OE construct was transformed into the maize inbred B104. Seeds of the maize inbred line B104 and ZmNAC20-overexpressing transgenic lines (ZmNAC20-OE1 and ZmNAC20-OE2) with the same size were selected, and immersed in 75% ethanol solution for disinfection for 1 min. After disinfection, the maize seeds were washed with distilled water 5-6 times. Then, the seeds were immersed in the distilled water and cultivated in the dark at 26 • C for 24 h. The soaked seeds with embryo sides faced upward were placed evenly in the seedling tray with filter paper, adding a proper amount of distilled water (the seeds just touched the water surface). After that, the seedling tray was sealed with plastic wrap, and placed in darkness at 28 • C for 3 days. The germinated seeds were transplanted into the mixed nutrient soil (vermiculite: nutrient soil = 1:1) and cultivated in the 28 • C plant light incubator under a photoperiod of 16 h/8 h (day/night), 60% relative humidity, and 16,800 Lux light intensity. After three true leaves grew from the maize seedlings, the seedlings were subjected to the follow-up experiment.

Drought Treatment
When maize seedlings of B104, ZmNAC20-OE1, and ZmNAC20-OE2 developed three true leaves, water was withheld for drought stress treatment. After about 15 days of drought treatment, watering was resumed to allow the plants to recover. The survival rate was recorded after three days of rewatering. Seedlings with green leaves and hard stems were recorded as survivors. In each test, at least 20 plants for each line were involved. Phenotype and data analysis was based on three independent experiments.

Relative Water Content Assay
After about 7 days of drought treatment, the middle segments of fresh leaves were cut about 5 cm long. The weight of the middle segments of fresh leaves was measured immediately and recorded as the fresh weight. Then the above leaves were immersed in distilled water at room temperature for 12 h. After that, the saturated fresh weight was recorded. At last, the leaves were placed in a 70 • C oven for 12 h, and the dry weight was measured. We used the formula [relative water content= (fresh weight − dry weight)/(saturated fresh weight − dry weight) × 100%] to calculate the relative water content of leaves. Statistical analysis was based on three independent experiments.

Detached Leaves Water Loss Assay
The above-ground parts of B104 and ZmNAC20-OE1 maize seedlings with three true leaves were cut and spread in the plant light incubator, with a temperature of 28 • C, relative humidity of 30%, and light intensity of 16,800 Lux. The dry weight (DW) of detached leaves was measured at 15 min, 45 min, 60 min, 100 min, 120 min, 160 min, 180 min, 200 min, 300 min, and 400 min after dehydration stress treatment. The weight of detached leaves was measured as fresh weight (FW) at 0 min of dehydration stress treatment. We used the formula [relative water loss rate = (FW − DW)/FW × 100%] to calculate the relative water loss rate. Statistical analysis was based on three independent experiments.

Determination of Biomass
After drought stress treatment for about 7 days, the aerial parts of maize seedlings were cut off and immediately weighed. The weight was recorded as fresh weight. The aerial parts of maize seedlings were then put into a 70 • C oven for continuous drying for 24 h. After that, the weight of the above materials was measured as the dry weight. Biomass (g) = dry weight or fresh weight. There were three biological replicates in total, with every replicate containing four maize seedlings.

Stomatal Aperture Assay
The aerial parts of B104 and ZmNAC20-OE1 maize seedlings with three true leaves were cut off and immersed in the stomatal opening buffer (10 mM KCl, 50 mM CaCl 2 , and 10 mM MES/Tris, pH 5.6), and then incubated in dark for 2 hours and in light for 3 hours. After incubation in the stomatal opening buffer, the leaves were taken out and immediately put into the 0, 1, and 10 µM ABA solution (taking stomatal opening buffer as a solution and ABA as solute) prepared in advance respectively. At 1 h of ABA treatment, transparent nail polish was coated on the upper epidermis of maize leaves. When nail polish was completely dried, the upper epidermis was torn off with tweezers and made into a temporary slide. Ten stomata were randomly selected in the view of the microscope, and the length and width were measured with Image. Three biological replicated assays were performed.

Subcellular Localization Assay
To determine the subcellular localization of ZmNAC20, the coding sequence without the stop code of ZmNAC20 was cloned and inserted into a 35S:GFP vector, obtaining 35S:NAC20-GFP. 35S:H2B-mCherry is a nuclear localization marker. The constructs of 35S:ZmNAC20-GFP and 35S:H2B-mCherry were transformed into the Agrobacterium tumefaciens strain GV3101, then used for transient expression in leaves of tobacco (Nicotiana benthamiana). The constructs 35S:GFP and 35S:H2B-mCherry were used to act as the control. The A. tumefaciens GV3101 harboring 35S:ZmNAC20-GFP and 35S:H2B-mCherry or 35S:GFP and 35S:H2B-mCherry, were cultured overnight at 28 • C and centrifuged for 10 min at 5000 rpm. The cultures were resuspended using the buffer solution with 10 mM MES (pH 5.5), and 10 mM MgCl 2 ·6H 2 O. The cultures were then infiltrated into the leaves of about one-month tobacco seedlings. After being cultured for 3 days, the infiltrated leaves were imaged using a Nikon A1 HD25 confocal microscope. The merged signals of GFP and RFP showed a yellow color.

qRT-PCR Assay
Total RNA from leaves was extracted using RNA-Solv ® Reagent of Omega Bio-tek (No. R6830). The RNA was reverse-transcribed to synthesize cDNA by TIANGEN's FastKing RT Kit (With gDNase) (No. KR116). The resultant cDNA was used as a template for the detection of relative gene expression, with the mazie ZmUbiquitin gene as the reference gene. qRT-PCR was performed with Hieff ® qPCR SYBR ® Green Master Mix of YEASEN (11201ES08). Each gene had three technical repeats and three biological repeats. Sequences of qRT-PCR primers were given in Table S3.

RNA-seq and Data Analysis
Three biological repeats of the wild type (WT-R1, WT-R2, and WT-R3) and the ZmNAC20-overexpressing transgenic plants ZmNAC20-OE1 (OE-R1, OE-R2, and OE-R3) were used for RNA sequencing (RNA-seq). Total RNAs were isolated from the detached leaves which were treated with dehydration for 3 h using TRI reagent (Sigma; catalog no. T9424, Sigma-Aldrich, Burlington, MA, USA). RNA was sequenced using Illumina 2500 instrument in Berry Genomics, Beijing, China. The clean reads were aligned to the genome sequences of maize (Zm-B73-REFERENCE-NAM-5.0) that was downloaded from the website (plants.ensembl.org); HISAT2 software (version 2.0.5) was used for the clean reads [56]. The gene expression levels were measured using StringTie software (version 1.3.6) [56]. Differential gene expression (DEG) analysis was conducted using DESeq2 software (version 1.34.0) [57]. DEG was screened by using the cutoff of the log 2 |fold change| ≥ 1 and q-value (adjusted p-value) ≤ 0.05 between the ZmNAC20-OE1 and the wild-type B104. The analysis of Gene Ontology (GO) enrichment was performed using AgriGO v2.0, a web-based tool, and a database for GO analyses [58]. The pathway KEGG enrichment for the DEG was performed using the OmicShare tools, a free online platform for data analysis (www.omicshare.com/tools, accessed on 7 February 2023). The RNA-seq data have been deposited into the sequence read archive (SRA) database of NCBI under accession number PRJNA893876.

Conclusions
In this study, we found a nucleus-localized NAC transcription factor ZmNAC20 that improved drought resistance in maize. ZmNAC20 was rapidly induced by drought and ABA. Overexpression of ZmNAC20 promoted stomatal closure and prevented water loss. ZmNAC20 regulated multiple pathways by modulation of many stress-responsive genes expression. Our results indicated that ZmNAC20 functioned as a positive regulator to improve drought resistance in maize. ZmNAC20 contributed to improving drought resistance by modulation of stomatal closure and activation expression of stress-responsive genes.