Transcription Factor SiDi19-3 Enhances Salt Tolerance of Foxtail Millet and Arabidopsis

Salt stress is an important limiting factor of crop production. Foxtail millet (Setaria italica L.) is an important model crop for studying tolerance to various abiotic stressors. Therefore, examining the response of foxtail millet to salt stress at the molecular level is critical. Herein, we discovered that SiDi19-3 interacts with SiPLATZ12 to control salt tolerance in transgenic Arabidopsis and foxtail millet seedlings. SiDi19-3 overexpression increased the transcript levels of most Na+/H+ antiporter (NHX), salt overly sensitive (SOS), and calcineurin B-like protein (CBL) genes and improved the salt tolerance of foxtail millet and Arabidopsis. Six SiDi19 genes were isolated from foxtail millet. Compared with roots, stems, and leaves, panicles and seeds had higher transcript levels of SiDi19 genes. All of them responded to salt, alkaline, polyethylene glycol, and/or abscisic acid treatments with enhanced expression levels. These findings indicate that SiDi19-3 and other SiDi19 members regulate salt tolerance and other abiotic stress response in foxtail millet.


Introduction
Salt stress affects physiological factors such as plant growth, development, and yield [1][2][3]. Plants acquire various molecular, biochemical, and physiological responses to tolerate salt stress [1]. Re-establishing ionic equilibrium to enhance salt tolerance is a crucial process for plants [4]. Additionally, the salt overly sensitive (SOS) pathway plays an important role in the response to salt stress in plants. The plasma membrane-localized Na + /H + antiporter SOS1 and its regulatory proteins, including SOS2 (a protein kinase, also known as CIPK24), SOS3/calcineurin B-like protein 4 (CBL4), and other CBL proteins, maintain this process [5][6][7][8]. Under salt stress, SOS3 and CBL10 collaborate in the roots and shoots to detect the rise in cytoplasmic calcium and transmit this signal to the downstream serine/threonine protein kinase SOS2. Subsequently, SOS2 phosphorylates SOS1 to enhance Na + /H + exchange activity and improve plant salt tolerance [1,3]. Simultaneously, tonoplast-localized Na + /H + antiporters (NHXs) that sequester Na + within vacuoles also play vital roles in re-establishing ionic homeostasis under salt stress [9,10]. Loss of function of SOS and NHX genes results in hypersensitivity to salt stress owing to high Na + accumulation in the cytosol. In contrast, overexpression of the products of these genes improves salt tolerance owing to the low Na + accumulation in the cytosol.
The Cys2/His2-type zinc-finger protein motif was first identified in transcription factor IIIA from Xenopus laevis which can be induced by drought stress [11][12][13]. Proteins containing the Cys2/His2-type zinc-finger are one of the best-characterized DNA-binding transcription factors in eukaryotes [14]. Di19 (drought induced) protein was firstly isolated from drought treated Arabidopsis roots [15]. All seven of the AtDi19 proteins contain two atypical Cys2/His2-type zinc-finger motifs [13]. Every member of the AtDi19 family, except AtDi19-2, has putative nuclear localization signals (NLS), and all but two are localized in and SiDi19-3 fused to N-terminus of YFP (SiDi19-3-NYFP) were co-transformed into tobacco epidermis cells; however, no YFP signals were observed when SiPLATZ12-CYFP and NYFP, SiDi19-3-NYFP, and CYFP, or NYFP and CYFP were co-transformed ( Figure  1D). The YFP signal could also be merged with the DAPI dye. These findings indicate an interaction between SiDi19-3 and SiPLATZ12. The SiDi19-3 and SiPLATZ12 interaction using in vitro pull-down assays. SiPLATZ12-His was incubated with GST or GST-SiDi19-3 purified from E.coli and was washed to remove unbound proteins. The bound proteins were eluted and analyzed using immunoblotting with anti-GST and anti-His antibodies. (D) The SiDi19-3 and SiPLATZ12 interaction in N. benthamiana indicated in bimolecular fluorescence complementation (BiFC) assay. DAPI, nuclear dye, 4′,6-diamidino-2-phenylindole. The photographs were taken using the green channel (YFP fluorescence), blue channel (DAPI fluorescence), bright channel, and their combination under a confocal microscope. Scale bars = 20 μm.

SiDi19-3 Enhances Salt Tolerance of Foxtail Millet
To analyze the role of SiDi9-3 under salt stress, we overexpressed SiDi19-3 in the hairy roots of 'Yugu1′ seedlings using K599 agrobacterium-mediated transformation. Notably, the SiDi19-3 transcript levels in transgenic foxtail millet seedlings were higher than those in the control (Em, empty vector transgenic seedlings) ( Figure 2A). The 35S::SiDi19-3 and Em transgenic seedlings were grown in the same way under typical conditions. In 35S::SiDi19-3 transgenic seedlings under salt stress, primary root length was significantly greater and root fresh weight was higher than those in control seedlings ( Figure 2B-D), demonstrating that SiDi19-3 positively controls the salt tolerance of foxtail millet seedlings. Scale bar = 20 µm. (B) The SiDi19-3 and SiPLATZ12 interaction tested using yeast two-hybrid assay. DDO, Synthetic Dropout/-Leu-Trp; QDO, Synthetic Dropout/-Leu-Trp-His-Ade. (C) The SiDi19-3 and SiPLATZ12 interaction using in vitro pull-down assays. SiPLATZ12-His was incubated with GST or GST-SiDi19-3 purified from E.coli and was washed to remove unbound proteins. The bound proteins were eluted and analyzed using immunoblotting with anti-GST and anti-His antibodies. (D) The SiDi19-3 and SiPLATZ12 interaction in N. benthamiana indicated in bimolecular fluorescence complementation (BiFC) assay. DAPI, nuclear dye, 4 ,6-diamidino-2-phenylindole. The photographs were taken using the green channel (YFP fluorescence), blue channel (DAPI fluorescence), bright channel, and their combination under a confocal microscope. Scale bars = 20 µm.

SiDi19-3 Enhances Salt Tolerance of Foxtail Millet
To analyze the role of SiDi9-3 under salt stress, we overexpressed SiDi19-3 in the hairy roots of 'Yugu1' seedlings using K599 agrobacterium-mediated transformation. Notably, the SiDi19-3 transcript levels in transgenic foxtail millet seedlings were higher than those in the control (Em, empty vector transgenic seedlings) ( Figure 2A). The 35S::SiDi19-3 and Em transgenic seedlings were grown in the same way under typical conditions. In 35S::SiDi19-3 transgenic seedlings under salt stress, primary root length was significantly greater and root fresh weight was higher than those in control seedlings ( Figure 2B-D), demonstrating that SiDi19-3 positively controls the salt tolerance of foxtail millet seedlings.

SiDi19-3 Increases the Expression of SiSOS, SiCBL, and SiNHX Genes
To better understand how SiDi19-3 controls the salt stress response in foxtail millet seedlings, we evaluated the transcript levels of salt tolerance related genes, including SiSOSs, SiCBLs, and SiNHXs ( Figure 3). Under salt stress, compared to the Em transgenic foxtail millet seedlings, SiDi19-3-overexpression seedlings raised the transcript levels of all SiCBLs, except SiCBL5 and SiCBL6; all SiNHXs, except SiNHX3; SiSOS1; and SiSOS2. However, under normal conditions, SiDi19-3 overexpression did not alter the transcript levels of them. These findings implied that the majority of the analyzed NHX, SOS, and CBL genes were upregulated by SiDi19-3 under salt stress.

SiDi19-3 Increases the Expression of SiSOS, SiCBL, and SiNHX Genes
To better understand how SiDi19-3 controls the salt stress response in foxtail millet seedlings, we evaluated the transcript levels of salt tolerance related genes, including SiSOSs, SiCBLs, and SiNHXs ( Figure 3). Under salt stress, compared to the Em transgenic foxtail millet seedlings, SiDi19-3-overexpression seedlings raised the transcript levels of all SiCBLs, except SiCBL5 and SiCBL6; all SiNHXs, except SiNHX3; SiSOS1; and SiSOS2. However, under normal conditions, SiDi19-3 overexpression did not alter the transcript levels of them. These findings implied that the majority of the analyzed NHX, SOS, and CBL genes were upregulated by SiDi19-3 under salt stress.   shown are means of three biological replicates. Data represent the mean ± SEM of three biological repeats. Student s t-test showed the significance at * p < 0.05, ** p < 0.01, and *** p < 0.001 levels. ns, no significance.

Heterologous Expression of SiDi19-3 Increases Salt Tolerance of Arabidopsis
To confirm its function, SiDi19-3 was heterologously expressed in Arabidopsis thaliana Col-0 plants. Transgenic Arabidopsis lines #1, #3, and #4 indicated higher SiDi19-3 transcript levels ( Figure 4A). Under salt stress, the SiDi19-3 transgenic Arabidopsis exhibited faster germination rate and growth with higher fresh weight and longer primary root length than those of the wild type (WT; Figure 4B-F). Furthermore, expression of SiDi19-3 increased the expression of AtCBL1-10 and AtSOS1-2 but not AtCBL5 in Arabidopsis seedlings under salt stress ( Figure 5). Under normal conditions, no change was observed in germination rate, fresh weight, and primary root length nor in the expression of NHXs, SOSs, and CBLs in Arabidopsis. These results demonstrated that SiDi19-3 also acts as a positive regulator of salt stress in Arabidopsis.    Student′s t-test indicated the significance at * p < 0.05, ** p < 0.01, and *** p < 0.001 levels. ns, no significance.

Identification of SiDi19 Gene Family in Foxtail Millet
To determine whether other SiDi19 members in foxtail millet exhibited a similar response to salt stress and other abiotic stressors, we searched the 'Yugu1′ genome assembly containing Di19 sequences from foxtail millet (Supplemental Table S1). We found that SiDi19 genes were unevenly distributed among nine chromosomes (chr) in foxtail millet. In comparison, chr3 contained three SiDi19 genes, chr5 had two, and chr1 had only one SiDi19 gene. Based on their chromosomal placement, the SiDi19 genes were named SiDi19-1-SiDi19-6 ( Figure 6A). A bioinformatical neighbor-joining tree analysis divided Di19 genes from Setaria italica, Oryza sativa, Sorghum bicolor, Zea mays, Glycine max, and Arabidopsis into three subfamilies that were similar to each other ( Figure 6B), indicating that Di19 probably plays conserved roles among plant species. According to gene structural analysis, all SiDi19 genes had the same structure, consisting of five variable-length exons and four introns ( Figure 6C). It is interesting that all Di19 proteins have two conserved motifs except SiDi19-6, which indicate that they may have similar biological functions and that SiDi19-6 protein might play a distinct function in plants ( Figure 6D). AtGAPDH and AtUBQ10 were used as two normalizing expression values within the calculation leading to the relative expression values. Data shown are means of three biological replicates. Data represent the mean ± SEM of three biological repeats. Student's t-test indicated the significance at * p < 0.05, ** p < 0.01, and *** p < 0.001 levels. ns, no significance.

Identification of SiDi19 Gene Family in Foxtail Millet
To determine whether other SiDi19 members in foxtail millet exhibited a similar response to salt stress and other abiotic stressors, we searched the 'Yugu1' genome assembly containing Di19 sequences from foxtail millet (Supplemental Table S1). We found that SiDi19 genes were unevenly distributed among nine chromosomes (chr) in foxtail millet. In comparison, chr3 contained three SiDi19 genes, chr5 had two, and chr1 had only one SiDi19 gene. Based on their chromosomal placement, the SiDi19 genes were named SiDi19-1-SiDi19-6 ( Figure 6A). A bioinformatical neighbor-joining tree analysis divided Di19 genes from Setaria italica, Oryza sativa, Sorghum bicolor, Zea mays, Glycine max, and Arabidopsis into three subfamilies that were similar to each other ( Figure 6B), indicating that Di19 probably plays conserved roles among plant species. According to gene structural analysis, all SiDi19 genes had the same structure, consisting of five variable-length exons and four introns ( Figure 6C). It is interesting that all Di19 proteins have two conserved motifs except SiDi19-6, which indicate that they may have similar biological functions and that SiDi19-6 protein might play a distinct function in plants ( Figure 6D).

Expression Patterns of SiDi19 Genes
Next, using quantitative reverse transcription polymerase chain reaction (RT-qPCR) analysis, we examined the transcriptional levels of SiDi19 genes in various organs. SiDi19-1, SiDi19-4, and SiDi19-6 indicated higher transcript levels in stems than other members of the gene family, whereas SiDi19-2, SiDi19-3, and SiDi19-4 were preferentially expressed in seeds ( Figure 7A). Furthermore, SiDi19-1, SiDi19-3, and SiDi19-4 indicated higher transcript levels in the panicles than other plant parts. All six SiDi19 genes had low transcript levels in the roots and leaves. These gene expression patterns in different tissues suggested that some members may function redundantly at different developmental stages.

Expression Patterns of SiDi19 Genes
Next, using quantitative reverse transcription polymerase chain reaction (RT-qPCR) analysis, we examined the transcriptional levels of SiDi19 genes in various organs. SiDi19-1, SiDi19-4, and SiDi19-6 indicated higher transcript levels in stems than other members of the gene family, whereas SiDi19-2, SiDi19-3, and SiDi19-4 were preferentially expressed in seeds ( Figure 7A). Furthermore, SiDi19-1, SiDi19-3, and SiDi19-4 indicated higher transcript levels in the panicles than other plant parts. All six SiDi19 genes had low transcript levels in the roots and leaves. These gene expression patterns in different tissues suggested that some members may function redundantly at different developmental stages. Quantitative reverse transcription polymerase chain reaction (RT-qPCR) analysis of SiDi19 genes in roots, stems, leaves, panicles, and seeds. Root, stem, and leaf samples were collected from ten uniformly developed 14day-old seedlings. Five approximately 5-cm-long, uniformly developed panicles from two-monthold Yugu1 seedlings were collected. Thirty mature seeds were collected. Each tissue sample was pooled and ground, and, respectively, three 100 mg tissue powder were weighed to extract RNA for RT-qPCR. All the tissues were collected three times from different plants for the three biological replicates. (B) RT-qPCR analysis of SiDi19 genes in 7-day-old Yugu1 seedlings after treatment with 200 mM NaCl, 50 mM NaHCO3, 100 μΜ ABA, or 20% PEG6000 for 0, 3, and 24 h. SiACTIN7 and 18S rRNA were used as two normalizing expression values within the calculation leading to the relative expression values. Data shown are means of three biological replicates. Data represent the mean ± SEM of three biological repeats. Student′s t-test indicated the significance at * p < 0.05, ** p < 0.01, and *** p < 0.001 levels.
We initially examined the cis-elements in the promoters of SiDi19 genes to determine whether any SiDi19 genes other than SiDi19-3 operated in response to abiotic stresses (Supplemental Table S2). Two or more ABA-responsive elements (ABREs) were predicted for all SiDi19 promoters. An auxin-responsive element (TGA) was predicted in the promoters of SiDi19-1, -3 and -5. Two or three ethylene response elements (ERFs) were predicted in the promoters of SiDi19-2 and -3. More than one drought responsive element (DRE/MBS) was predicted in the promoters of SiDi19-1, -2, -4, and -6. One or two lowtemperature responsive elements (LTRs) were predicted in the promoters of SiDi19-2 to -5. One to three salicylic acid-responsive elements (TCA/SARE) were predicted in the promoters of SiDi19-1 and -3, respectively. More than one wound responsive element (WRE) was predicted in the promoters of all SiDi19 genes except SiDi19-2. These results suggest the potential roles of SiDi19 genes in hormone-mediated abiotic stress responses.
Furthermore, we detected the transcript levels of the SiDi19 genes at 3 and 24 h after appropriate treatments. As depicted in Figure 7B, some treatments clearly induced the expression of several SiDi19 genes. For example, all SiDi19 genes were induced by NaCl, NaHCO3, and ABA treatments at 3 or 24 h, and all SiDi19 genes, except SiDi19-6, were induced by PEG. Furthermore, SiDi19-3 and -4 indicated the highest inducible levels in response to NaCl and ABA treatments. These results indicate the redundant roles of SiDi19 genes in response to multiple ABA-mediated abiotic stresses in foxtail millet. Quantitative reverse transcription polymerase chain reaction (RT-qPCR) analysis of SiDi19 genes in roots, stems, leaves, panicles, and seeds. Root, stem, and leaf samples were collected from ten uniformly developed 14-day-old seedlings. Five approximately 5-cm-long, uniformly developed panicles from two-monthold Yugu1 seedlings were collected. Thirty mature seeds were collected. Each tissue sample was pooled and ground, and, respectively, three 100 mg tissue powder were weighed to extract RNA for RT-qPCR. All the tissues were collected three times from different plants for the three biological replicates. (B) RT-qPCR analysis of SiDi19 genes in 7-day-old Yugu1 seedlings after treatment with 200 mM NaCl, 50 mM NaHCO 3 , 100 µM ABA, or 20% PEG6000 for 0, 3, and 24 h. SiACTIN7 and 18S rRNA were used as two normalizing expression values within the calculation leading to the relative expression values. Data shown are means of three biological replicates. Data represent the mean ± SEM of three biological repeats. Student's t-test indicated the significance at * p < 0.05, ** p < 0.01, and *** p < 0.001 levels.
We initially examined the cis-elements in the promoters of SiDi19 genes to determine whether any SiDi19 genes other than SiDi19-3 operated in response to abiotic stresses (Supplemental Table S2). Two or more ABA-responsive elements (ABREs) were predicted for all SiDi19 promoters. An auxin-responsive element (TGA) was predicted in the promoters of SiDi19-1, -3 and -5. Two or three ethylene response elements (ERFs) were predicted in the promoters of SiDi19-2 and -3. More than one drought responsive element (DRE/MBS) was predicted in the promoters of SiDi19-1, -2, -4, and -6. One or two low-temperature responsive elements (LTRs) were predicted in the promoters of SiDi19-2 to -5. One to three salicylic acid-responsive elements (TCA/SARE) were predicted in the promoters of SiDi19-1 and -3, respectively. More than one wound responsive element (WRE) was predicted in the promoters of all SiDi19 genes except SiDi19-2. These results suggest the potential roles of SiDi19 genes in hormone-mediated abiotic stress responses.
Furthermore, we detected the transcript levels of the SiDi19 genes at 3 and 24 h after appropriate treatments. As depicted in Figure 7B, some treatments clearly induced the expression of several SiDi19 genes. For example, all SiDi19 genes were induced by NaCl, NaHCO 3 , and ABA treatments at 3 or 24 h, and all SiDi19 genes, except SiDi19-6, were induced by PEG. Furthermore, SiDi19-3 and -4 indicated the highest inducible levels in response to NaCl and ABA treatments. These results indicate the redundant roles of SiDi19 genes in response to multiple ABA-mediated abiotic stresses in foxtail millet.
SiDi19-3 can control cytoplasmic Na + homeostasis. SOSs, CBLs, and NHXs are important for cytoplasmic Na + homeostasis. In Arabidopsis, SOS2-SOS3 (CBL4)/SCaBP8 (CBL10) complexes activate the transporter activities of SOS1 and NHXs [5,32,33]. The results indicated that SiDi19-3 increased the expression of the majority of the analyzed SOS, CBL, and NHX genes (Figures 3 and 5), indicating the involvement of SiDi19-3 in regulating cytoplasmic Na + homeostasis. SiCBL4 and SiCIPK24 interact to attract SiCIPK24 to the plasma membrane. Compared with WT plants, SiCIPK24 transgenic Arabidopsis plants are more resilient to salt stress [34]. SiCBL5-overexpressing foxtail millet plants demonstrate increased tolerance to salt stress. Furthermore, SiCBL5-SiCIPK24 affects SiSOS1 function in yeast cells [35]. We previously demonstrated that overexpression of SiNHX2, SiCBL4, and SiCBL7 in the hairy roots of foxtail millet seedlings improved salt tolerance [28]. The salt tolerance of foxtail millet correlates with the expression of SiNHX1 to SiNHX4 [28]. Therefore, SiDi19-3 mediates salt tolerance by maintaining cytoplasmic Na + homeostasis.
The development of foxtail millet might be influenced by SiDi19-3. Additionally, NHXs play roles in Arabidopsis stomatal control [36], plant growth [9,10,37], silique and seed development [38,39], and vacuolar K + and pH homeostasis [9,10]. The up-regulation of SOS, CBL, and NHX gene expression by SiDi19-3 under salt stress (Figures 3 and 5) indicates that SiDi19-3 may be involved in controlling foxtail millet growth and development under salt stress. The SiDi19-3 transcript levels were higher in the panicles and in response to drought and ABA treatments (Figure 7), also indicating the possible participation of SiDi19-3 in foxtail millet panicle or seed development and drought response. These roles of SiDi19-3 and the underlying mechanisms merit additional studies.

Plant Materials and Growth Conditions
For investigation of gene expression and function, the foxtail millet (S. italica) cultivar 'Yugu1' and Arabidopsis Col-0 were employed. Foxtail millet seeds underwent sodium hypochlorite sterilization, were washed three times, and then were germinated for three days at 25 • C. Seven-day-old 'Yugu1' seedlings were exposed to Hoagland solution containing 200 mM NaCl, 50 mM NaHCO 3 , 100 µM ABA, or 20% (w/v) polyethylene glycol 6000 (PEG6000) for 3 and 24 h, respectively, for inducible expression analyses. The seedlings were collected and frozen in liquid nitrogen as soon as possible. Roots, stems, and leaves from 14-day-old seedlings, panicles from 2-month-old seedlings, and seeds of 'Yugu1' were collected and immediately frozen in liquid nitrogen for tissue expression analyses. The previously described techniques were used to grow foxtail millet seedlings with transgenic hairy roots [40].
The A. thaliana Col-0 was used as WT. The previously described techniques were used to grown Arabidopsis plants [41]. During the seed germination stage, seeds were exposed to 150 mM NaCl in 1/2 MS medium for 7 d. In a span of 7 d, the germinated seeds with 2 cm roots and 1 cm green shoots were counted daily. To induce salt stress, three-day-old seedlings germinated from seeds that were not salt treated of each genotype were transferred to 1/2 MS media containing 150 or 175 mM NaCl for an additional 14 d, and the root lengths and fresh weight of whole seedlings were measured. Three biological replicates of each experiment were performed.

Yeast Two-Hybrid (Y2H) Assays
The interaction between SiDi9-3 and SiPLATZ12 was tested using the Matchmaker GAL4 two-hybrid system (Clontech) as previously described [42]. Full-length coding sequences (CDSs) of SiDi9-3 and SiPLATZ12 were cloned into the pGBKT7 (BD) and pGADT7 (AD) vectors, respectively. The yeast strain AH109 was co-transformed with those constructs. Leu-Trp-His-Ade (QDO) was used for interaction selection, while double dropout (DDO) medium supplemented without Leu or Trp was used to grow the transformed cells. Primers used in this experiment are listed in Supplemental Table S3.

Pull-Down Assays
For these assays, the CDSs of SiDi19-3 and SiPLATZ12 were cloned into pGEX-4T-1 and pCold-TF vectors to generate SiDi19-3-GST and SiPLATZ12-His constructs, respectively, as previously described [41]. Subsequently, the constructs were transformed into competent Rosetta E. coli cells. SiDi19-3-GST and SiPLATZ12-His proteins were induced by IPTG and were then used for protein purification using the His-Tagged Protein Purification Kit (CWBIO, Beijing, China) and BeyoGold GST-tag Purification Resin (BeyoGold, Shanghai, China). Next, SiDi19-3-GST or GST proteins were incubated with SiPLATZ12-His protein in 300 µL of binding buffer (P2262, Beyotime, Shanghai, China) for 1 h at 4 • C in a continuously rotator. The proteins were detected using immunoblotting with anti-GST or anti-His antibodies after elution from the beads. Signals were detected using a Chemiluminescence Imaging System (K4000, KCRX Biotechnology, Beijing, China). The primers used in this experiment are listed in Supplemental Table S3.

Subcellular Localization Analysis
The transgenic hairy-roots harboring pSiDi19-3:SiDi19-3-GFP were obtained using the previously described techniques [39] to analyze the subcellular location of SiDi19-3. LSM880 high-resolution laser confocal microscope (Zeiss, Germany) was utilized to view GFP fluorescence in the root tips of the transgenic hairy roots. The DAPI was used to detect the nuclear dye.

Generation of Transgenic Foxtail Millet and Arabidopsis
Under the control of the 35S promoter, the full-length CDS of SiDi19-3 was amplified and cloned into PROKII vector. Based on an earlier study [39], the recombinant plasmid was transformed into A. tumefaciens K599. Positive transformants were inoculated into 20 mL of liquid solution and cultured with continuous shaking for 12 h. The culture was then centrifuged at 6000× g for 5 min. Subsequently, the cells were resuspended in 1/2 MS liquid medium to a final concentration with OD 600 = 1.0. Three-day-old shoot tips were cut and incubated with A. tumefaciens containing 35S::SiDi19-3 or the empty vector at 28 • C for 10-20 min with continuous shaking. The transformed shoot tips were transferred onto 1/2 MS solid medium supplemented with 100 mg/L timentin overnight to induce root formation [39]. Approximately 7 d later, transgenic hairy roots were confirmed using RT-qPCR. Twenty transgenic seedlings were transferred into 1/2 MS liquid medium with or without 150 mM NaCl for an additional 7 d. The phenotype was photographed, and root length was measured. Three biological replicates of each experiment were performed.
Transgenic Arabidopsis containing 35S::SiDi19-3 was simultaneously generated using the floral-dipping method as previously described [43]. T 3 transgenic Arabidopsis plants were selected using 1/2 MS medium supplemented with 50 mg/L kanamycin and confirmed by RT-qPCR using the primers listed in Supplemental Table S3.

RNA Extraction and RT-qPCR
Total RNAs was extracted from the indicated tissues and seedlings treated with the indicated stimuli using RNAiso Plus (Takara, Ohtsu, Japan). Two micrograms of total RNAs were used to synthesize cDNAs using the PrimeScriptTM RT reagent kit (TaKaRa, Ohtsu, Japan) according to the manufacturer's instructions. Further, RT-qPCR was performed using ChamQ Universal SYBR qPCR Master Mix (Vazyme, Biotech Co., Ltd.) in a three-step program on a CFX96TM Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA). SiACTIN7 and 18S rRNA were used as two normalizing expression values within the calculation leading to the relative expression values in foxtail millet [40,44]. AtGAPDH and AtUBQ10 were used as two normalizing expression values within the calculation leading to the relative expression values in Arabidopsis [41,42]. Three biological replicates were performed. The Primers used are listed in Supplementary Table S3.