Expression Patterns and Molecular Mechanisms Regulating Drought Tolerance of Soybean [Glycine max (L.) Merr.] Conferred by Transcription Factor Gene GmNAC19

NAC transcription factors are commonly involved in the plant response to drought stress. A transcriptome analysis of root samples of the soybean variety ‘Jiyu47’ under drought stress revealed the evidently up-regulated expression of GmNAC19, consistent with the expression pattern revealed by quantitative real-time PCR analysis. The overexpression of GmNAC19 enhanced drought tolerance in Saccharomyces cerevisiae INVSc1. The seed germination percentage and root growth of transgenic Arabidopsis thaliana were improved in comparison with those of the wild type, while the transgenic soybean composite line showed improved chlorophyll content. The altered contents of physiological and biochemical indices (i.e., soluble protein, soluble sugar, proline, and malondialdehyde) related to drought stress and the activities of three antioxidant enzymes (i.e., superoxide dismutase, peroxidase, and catalase) revealed enhanced drought tolerance in both transgenic Arabidopsis and soybean. The expressions of three genes (i.e., P5CS, OAT, and P5CR) involved in proline synthesis were decreased in the transgenic soybean hairy roots, while the expression of ProDH involved in the breakdown of proline was increased. This study revealed the molecular mechanisms underlying drought tolerance enhanced by GmNAC19 via regulation of the contents of soluble protein and soluble sugar and the activities of antioxidant enzymes, providing a candidate gene for the molecular breeding of drought-tolerant crop plants.


Introduction
Soybean is a leguminous crop plant with significant economic and nutritional demand worldwide [1].Given its wide range of cultivation, soybean growth and development are commonly influenced by a variety of biological and abiotic factors [2].For example, drought is one of the most important factors influencing soybean yield [1,3,4].In particular, drought could affect photosynthesis, respiration, and transpiration [5], osmotic regulation [6,7], antioxidant enzyme activity [8], and growth in soybean, thereby altering the growth and development of soybeans and their symbiotic nitrogen fixation, ultimately causing a decrease in soybean yield [1,9].Hence, the development of drought-tolerant soybean varieties is crucial to meet the growing global demand for soybean [10].
It is well known that plants have evolved a variety of metabolic strategies to cope with abiotic stresses, e.g., drought, while transcription factors (TFs) are commonly involved in the regulation of molecular responses to drought in plants [11].Due to their regulatory effects on many genes involved in adverse conditions, TFs are generally considered one of the most appropriate genetic targets for the molecular breeding and development of stress-tolerant crop varieties, e.g., inducing drought response in plants [12,13].Studies have shown that a large number of TFs, e.g., NAC [14], AP2/ERF [15], bZIP [16], bHLH [17], MYB [18,19], and WRKY [20], are frequently involved in the plant response to water stress.In particular, NAC TFs are one of the largest families of plant-specific transcriptional regulators and are widely distributed in plants, including mosses, ferns, gymnosperms, and angiosperms [21,22].The NAC domain is named after the initial letters of no apical meristem (NAM) of Petunia, Arabidopsis transcription activation factors (ATAF1 and ATAF2), and cup-shaped cotyledon 2 (CUC2) [23].Due to their highly conserved N-terminal NAC domains and diverse C-terminal domains, which are involved in the regulation of transcription activities, NAC TFs have been widely studied [23,24].The N-terminal NAC domain, composed of about 160 amino acids, contains a rare TF folding structure consisting of several helically surrounded β-folds, and structural analysis has identified the functional dimers formed by the N-terminal NAC domain [25].These functional dimers contain nuclear localization signals and bind to the cis-elements of their target genes.The C-terminal region is highly variable, containing trans-activated regions, and the C-terminal of individual proteins could interact with other proteins to regulate the gene expression, while other domains of NAC TFs contain transmembrane motifs [26].To date, studies have shown that the NAC TFs of plants play an important role in various biological activities, including plant senescence, cell division, seed development, lateral root formation, and the response to both biotic and abiotic stresses [14,22,27].
Our previous studies identified the molecular functions of both GmNAC3 and GmNAC4 in the drought tolerance of plants [36,37].The results of transcriptome sequencing analysis of soybean roots showed that GmNAC19 was significantly up-regulated in response to the treatment of drought stress in soybean.Therefore, we further explored the expression pattern of GmNAC19 and a group of physiological and biochemical indices related to drought tolerance, as well as the activities of three antioxidant enzymes, to clarify the functions of GmNAC19 and the molecular mechanism underlying enhanced drought tolerance in soybean.The objectives of this study were to characterize the expression pattern of GmNAC19, to detect the variations in the physiological and biochemical factors related to drought tolerance, and to evaluate the activities of three antioxidant enzymes in soybean and Arabidopsis thaliana.Our study provided strong experimental evidence to support the function of soybean GmNAC19 and its molecular response to drought stress in plants, revealing a candidate gene for the potential molecular breeding and development of drought-resistant soybean varieties.

Gene Expression of GmNAC19 Based on Transcriptome Analysis of Drought-Treated Soybean Roots
The results of transcriptome sequencing (https://www.ncbi.nlm.nih.gov/sra/;accessed on 30 October 2023; BioProject accession PRJNA1033409) showed that a total of 49 genes in the NAC TF family were significantly up-regulated in soybean roots under drought stress simulated by 20% PEG6000.Specifically, due to its high expression level and the identification of the protein encoded by GmNAC19 in the NAC domain TF superfamily of proteins, GmNAC19 was further explored in this study to reveal its molecular functions in response to drought stress in plants.Specifically, the results of the gene expression of GmNAC19 in soybean roots treated with 20% PEG6000 revealed the highest expression level at 6 h, which was about 20 times higher than that at 0 h (i.e., without the drought treatment).Then, the relative expression level was increased to about 12 and 13.6 times higher than that at 0 h at 18 h and 24 h, respectively (Figure 1).The results of transcriptome sequencing (https://www.ncbi.nlm.nih.gov/sra/;accessed on 30 October 2023; BioProject accession PRJNA1033409) showed that a total of 49 genes in the NAC TF family were significantly up-regulated in soybean roots under drought stress simulated by 20% PEG6000.Specifically, due to its high expression level and the identification of the protein encoded by GmNAC19 in the NAC domain TF superfamily of proteins, GmNAC19 was further explored in this study to reveal its molecular functions in response to drought stress in plants.Specifically, the results of the gene expression of GmNAC19 in soybean roots treated with 20% PEG6000 revealed the highest expression level at 6 h, which was about 20 times higher than that at 0 h (i.e., without the drought treatment).Then, the relative expression level was increased to about 12 and 13.6 times higher than that at 0 h at 18 h and 24 h, respectively (Figure 1).

Figure 1.
Relative gene expression levels of GmNAC19 based on transcriptome sequencing analysis of root tips (3 cm in length) of 7-day-old soybean roots treated with 20% PEG6000 for 0, 6, 12, 18, and 24 h.FPKM, fragments per kilobase of transcript per million mapped reads.The statistical significance between two groups was determined by one-way ANOVA, with significant differences in the relative expression level of GmNAC19 indicated by different lowercase letters (a, b, or c) based on p < 0.05 (n = 3).The same letters (a, b, or c) indicate no significant difference.

Gene Expression of GmNAC19 in Soybean Treated with PEG6000 Based on Quantitative Real-Time PCR
Based on qRT-PCR, the results of the gene expression of GmNAC19 in soybean roots treated with 20% PEG6000 showed that the relative gene expression levels of GmNAC19 were first increased from 0 to 6 h, reaching the highest expression level at 6 h of about 7.9 times higher than that at 0 h, and then, decreased from 6 to 24 h (i.e., about 1.7 times higher than that at 0 h) (Figure 2).Based on qRT-PCR, the results of the gene expression of GmNAC19 in soybean roots treated with 20% PEG6000 showed that the relative gene expression levels of GmNAC19 were first increased from 0 to 6 h, reaching the highest expression level at 6 h of about 7.9 times higher than that at 0 h, and then, decreased from 6 to 24 h (i.e., about 1.7 times higher than that at 0 h) (Figure 2).

Gene Expression of GmNAC19 in Different Organs of Soybean at Three Different Developmental Stages
The relative expression levels of GmNAC19 were further detected in different organs (i.e., stems, roots, leaves, flowers, and pods) of soybean plants at three developmental stages, i.e., seedling, flowering, and podding (Figure 3).The results showed that among the different organs, the leaves were consistently revealed to have the highest relative expression levels of GmNAC19 during the three developmental stages, whereas the lowest relative expression levels were consistently detected in the stems.Specifically, the relative expression levels in leaves were 2.9, 56, and 55 times higher than those in the stems during the seeding, flowering, and podding stages, respectively.During the seedling stage (Figure 3A), the relative expression levels of GmNAC19 were significantly different between the stems and roots and between the stems and leaves, whereas no significant difference was detected between the roots and leaves.During the flowering stage (Figure 3B), no significant differences were detected among The relative expression levels of GmNAC19 were further detected in different organs (i.e., stems, roots, leaves, flowers, and pods) of soybean plants at three developmental stages, i.e., seedling, flowering, and podding (Figure 3).The results showed that among the different organs, the leaves were consistently revealed to have the highest relative expression levels of GmNAC19 during the three developmental stages, whereas the lowest relative expression levels were consistently detected in the stems.Specifically, the relative expression levels in leaves were 2.9, 56, and 55 times higher than those in the stems during the seeding, flowering, and podding stages, respectively.During the seedling stage (Figure 3A), the relative expression levels of GmNAC19 were significantly different between the stems and roots and between the stems and leaves, whereas no significant difference was detected between the roots and leaves.During the flowering stage (Figure 3B), no significant differences were detected among the stems, roots, and flowers, whereas during the podding stage (Figure 3C), significant differences were detected among all four organs examined.The relative expression levels of GmNAC19 were further detected in different organs (i.e., stems, roots, leaves, flowers, and pods) of soybean plants at three developmental stages, i.e., seedling, flowering, and podding (Figure 3).The results showed that among the different organs, the leaves were consistently revealed to have the highest relative expression levels of GmNAC19 during the three developmental stages, whereas the lowest relative expression levels were consistently detected in the stems.Specifically, the relative expression levels in leaves were 2.9, 56, and 55 times higher than those in the stems during the seeding, flowering, and podding stages, respectively.During the seedling stage (Figure 3A), the relative expression levels of GmNAC19 were significantly different between the stems and roots and between the stems and leaves, whereas no significant difference was detected between the roots and leaves.During the flowering stage (Figure 3B), no significant differences were detected among the stems, roots, and flowers, whereas during the podding stage (Figure 3C), significant differences were detected among all four organs examined.To construct the pYES2-GmNAC19 yeast expression vector, the amplified target PCR product was connected with a pYES2 yeast expression vector and transformed into competent cells of Escherichia coli DH5α.A single colony was selected for further culture and plasmid extraction.The successful construction of the pYES2-GmNAC19 yeast expression vector was verified using a bacterial solution and plasmid PCR analyses based on the target PCR-amplified products and sequences obtained (Figure 4).
To construct the pYES2-GmNAC19 yeast expression vector, the amplified target PCR product was connected with a pYES2 yeast expression vector and transformed into competent cells of Escherichia coli DH5α.A single colony was selected for further culture and plasmid extraction.The successful construction of the pYES2-GmNAC19 yeast expression vector was verified using a bacterial solution and plasmid PCR analyses based on the target PCR-amplified products and sequences obtained (Figure 4).The drought tolerance in transgenic yeast with GmNAC19 was further evaluated.The results showed that on YPD medium, the yeast in both the control and the transgenic groups could grow normally, showing no significant variations in their development (Figure 5A).With the increase in mannitol content in the YPD medium, the growth of yeast in both the control and the transgenic groups was decreased, whereas the transgenic group showed improved growth performance, as observed in the increased cell number (Figure 5B,C).These results indicated that transgenic yeast with GmNAC19 gained enhanced drought tolerance.The drought tolerance in transgenic yeast with GmNAC19 was further evaluated.The results showed that on YPD medium, the yeast in both the control and the transgenic groups could grow normally, showing no significant variations in their development (Figure 5A).With the increase in mannitol content in the YPD medium, the growth of yeast in both the control and the transgenic groups was decreased, whereas the transgenic group showed improved growth performance, as observed in the increased cell number (Figure 5B,C).These results indicated that transgenic yeast with GmNAC19 gained enhanced drought tolerance.) and transgenic (pYES2-GmNAC19) groups based on yeast inoculation dilution of 10⁻ 5 CFU/mL.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the number of yeast cells between the control without treatment of mannitol and controls with treatment of mannitol, and between transgenic group without treatment of mannitol and transgenic groups with treatment of mannitol; two-way ANOVA tests were performed to detect significant differences in the number of yeast cells between the control and transgenic groups cultured on the same type of medium (indicated by asterisk above the bracket) based on p < 0.05 (*) or p < 0.01 (**) (n = 3).

Enhanced Drought Tolerance in Transgenic Arabidopsis thaliana with GmNAC19
• Drought tolerance of transgenic Arabidopsis thaliana with GmNAC19 under drought stress simulated by PEG6000 (1) Seed germination of transgenic Arabidopsis thaliana with GmNAC19 The seed germination percentages of both the wild-type (WT) and one line (OE-6) of T3 generation transgenic A. thaliana with GmNAC19 were evaluated based on seeds sown on MS medium with PEG6000 at 0%, 3%, 6%, and 9% in 15 d (Figure 6).The results showed ) and transgenic (pYES2-GmNAC19) groups based on yeast inoculation dilution of 10 −5 CFU/mL.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the number of yeast cells between the control without treatment of mannitol and controls with treatment of mannitol, and between transgenic group without treatment of mannitol and transgenic groups with treatment of mannitol; two-way ANOVA tests were performed to detect significant differences in the number of yeast cells between the control and transgenic groups cultured on the same type of medium (indicated by asterisk above the bracket) based on p < 0.05 (*) or p < 0.01 (**) (n = 3).

•
Drought tolerance of transgenic Arabidopsis thaliana with GmNAC19 under drought stress simulated by PEG6000 (1) Seed germination of transgenic Arabidopsis thaliana with GmNAC19 The seed germination percentages of both the wild-type (WT) and one line (OE-6) of T3 generation transgenic A. thaliana with GmNAC19 were evaluated based on seeds sown on MS medium with PEG6000 at 0%, 3%, 6%, and 9% in 15 d (Figure 6).The results showed that the average germination percentages of WT A. thaliana were revealed to have a largely decreasing pattern, reaching 97.4%, 92.02%, 86.98%, and 75.52% as the concentrations of PEG6000 were increased, i.e., 0%, 3%, 6%, and 9%, respectively.Under the treatments of PEG6000 at 6% and 9%, the seed germination percentages of transgenic A. thaliana with GmNAC19 (95.31% and 93.23%) were significantly higher than those of WT (86.98% and 75.52%, respectively).(OE-6) treated with PEG6000 at 0%, 3%, 6%, and 9% for 15 d.Each treatment group contains 3 replicates (64 seeds in each replicate).One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in seed germination percentage between WT without treatment of PEG6000 and WT with treatment of PEG6000, and between transgenic group without treatment of PEG6000 and transgenic groups with treatment of PEG6000; two-way ANOVA tests were performed to detect the significant differences in seed germination percentage between WT and transgenic groups in the same treatment group (indicated by asterisk above the bracket) based on p < 0.01 (**) and p < 0.001 (***) (n = 3).
(2) Growth of roots in transgenic Arabidopsis thaliana with GmNAC19 Seeds of both WT and one line (OE-6) of T3 generation transgenic A. thaliana of GmNAC19 were sown on MS medium with PEG6000 at 0%, 3%, 6%, and 9%, with the root length measured in a total of 15 plants (three replicates each of five plants) of each group in 15 d (Figure 7).The results showed that the average root lengths of WT A. thaliana were 3.8 cm, 3.19 cm, 1.53 cm, and 0.9 cm on MS medium containing PEG6000 at 0%, 3%, 6%, and 9%, respectively, which were significantly shorter than those of transgenic A. thaliana with GmNAC19, i.e., 5.23 cm, 4.36 cm, 3.13 cm, and 1.58 cm, respectively.These results showed that the root elongation in A. thaliana was inhibited by an increased concentration of PEG6000, whereas the root growth in transgenic A. thaliana with GmNAC19 grown in MS medium with different concentrations of PEG6000 was significantly improved in comparison with that of WT.
To summarize, the results of the seed germination and root growth of transgenic A. thaliana with GmNAC19 treated with PEG6000 at various concentrations showed that as the level of drought stress was increased, the germination percentage of transgenic A. thaliana remained above 93%, whereas the germination percentage of WT was significantly decreased to 86.98% (at 6% PEG6000) and 75.52% (at 9% PEG6000).The largest increase of 1.6 cm was detected from the average root lengths of WT to transgenic A. thaliana with GmNAC19 grown under the treatment of 6% PEG6000.Therefore, the treatment of 6% PEG6000 with a high seed germination percentage maintained was selected for the subsequent evaluation of physiological and biochemical indices related to drought tolerance.6) treated with PEG6000 at 0%, 3%, 6%, and 9% for 15 d.Each treatment group contains 3 replicates (64 seeds in each replicate).One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in seed germination percentage between WT without treatment of PEG6000 and WT with treatment of PEG6000, and between transgenic group without treatment of PEG6000 and transgenic groups with treatment of PEG6000; two-way ANOVA tests were performed to detect the significant differences in seed germination percentage between WT and transgenic groups in the same treatment group (indicated by asterisk above the bracket) based on p < 0.01 (**) and p < 0.001 (***) (n = 3).
(2) Growth of roots in transgenic Arabidopsis thaliana with GmNAC19 Seeds of both WT and one line (OE-6) of T3 generation transgenic A. thaliana of GmNAC19 were sown on MS medium with PEG6000 at 0%, 3%, 6%, and 9%, with the root length measured in a total of 15 plants (three replicates each of five plants) of each group in 15 d (Figure 7).The results showed that the average root lengths of WT A. thaliana were 3.8 cm, 3.19 cm, 1.53 cm, and 0.9 cm on MS medium containing PEG6000 at 0%, 3%, 6%, and 9%, respectively, which were significantly shorter than those of transgenic A. thaliana with GmNAC19, i.e., 5.23 cm, 4.36 cm, 3.13 cm, and 1.58 cm, respectively.These results showed that the root elongation in A. thaliana was inhibited by an increased concentration of PEG6000, whereas the root growth in transgenic A. thaliana with GmNAC19 grown in MS medium with different concentrations of PEG6000 was significantly improved in comparison with that of WT.  6) treated with PEG6000 at 0%, 3%, 6%, and 9%, for 15 d.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in root length and root diameter between WT without treatment of PEG6000 and WT with treatment of PEG6000, and between transgenic group without treatment of PEG6000 and transgenic groups with treatment of PEG6000; twoway ANOVA tests were performed to detect the significant differences in root length and root diameter between WT and transgenic groups in the same treatment group (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3 with each replicate of 5 plants).
The results of phenotypic variations revealed no significant difference in the growth of WT and transgenic A. thaliana under the normal growth conditions, whereas the wilting degree of the WT was significantly higher than that of the transgenic plants under drought stress (i.e., treatment of 6% PEG6000), and the recovery ability of the transgenic plants was stronger than that of the WT after re-watering (Figure S1).Furthermore, the results of the 100-seed weight showed that the 100-seed weights of the transgenic plants were significantly higher than those of the WT (Table 1).
Table 1.The 100-seed weight (mg) in both wild-type (WT) and transgenic Arabidopsis thaliana with GmNAC19.Data are presented as mean ± standard deviation.The statistical difference between the WT and transgenic lines was determined by Student's t test based on p < 0.05 (*) and p < 0.01 (**).treated with PEG6000 at 0%, 3%, 6%, and 9%, for 15 d.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in root length and root diameter between WT without treatment of PEG6000 and WT with treatment of PEG6000, and between transgenic group without treatment of PEG6000 and transgenic groups with treatment of PEG6000; two-way ANOVA tests were performed to detect the significant differences in root length and root diameter between WT and transgenic groups in the same treatment group (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3 with each replicate of 5 plants).
To summarize, the results of the seed germination and root growth of transgenic A. thaliana with GmNAC19 treated with PEG6000 at various concentrations showed that as the level of drought stress was increased, the germination percentage of transgenic A. thaliana remained above 93%, whereas the germination percentage of WT was significantly decreased to 86.98% (at 6% PEG6000) and 75.52% (at 9% PEG6000).The largest increase of 1.6 cm was detected from the average root lengths of WT to transgenic A. thaliana with GmNAC19 grown under the treatment of 6% PEG6000.Therefore, the treatment of 6% PEG6000 with a high seed germination percentage maintained was selected for the subsequent evaluation of physiological and biochemical indices related to drought tolerance.
The results of phenotypic variations revealed no significant difference in the growth of WT and transgenic A. thaliana under the normal growth conditions, whereas the wilting degree of the WT was significantly higher than that of the transgenic plants under drought stress (i.e., treatment of 6% PEG6000), and the recovery ability of the transgenic plants was stronger than that of the WT after re-watering (Figure S1).Furthermore, the results of the 100-seed weight showed that the 100-seed weights of the transgenic plants were significantly higher than those of the WT (Table 1).(3) Contents of physiological and biochemical indices related to drought stress in transgenic Arabidopsis thaliana with GmNAC19 under drought stress simulated by PEG6000 On MS medium without PEG6000, no significant differences in the contents of soluble protein, soluble sugar, proline, and malondialdehyde (MDA) were detected between the WT and transgenic A. thaliana with GmNAC19 (Figure 8).On MS medium containing 6% PEG6000, the contents of soluble protein, proline, and soluble sugar were significantly higher in the transgenic A. thaliana with GmNAC19 than those of the WT (Figure 8A-C), while a significant decrease was detected in the content of MDA between the WT and transgenic Arabidopsis thaliana with GmNAC19 (Figure 8D).On MS medium without PEG6000, no significant differences in the contents of ble protein, soluble sugar, proline, and malondialdehyde (MDA) were detected be the WT and transgenic A. thaliana with GmNAC19 (Figure 8).On MS medium conta 6% PEG6000, the contents of soluble protein, proline, and soluble sugar were signific higher in the transgenic A. thaliana with GmNAC19 than those of the WT (Figure 8 while a significant decrease was detected in the content of MDA between the WT transgenic Arabidopsis thaliana with GmNAC19 (Figure 8D).-6) under drought stress simulated by 6% PEG6000 for 15 d.On ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differen the content of each index between WT without treatment of PEG6000 and WT with treatm PEG6000, and between transgenic group without treatment of PEG6000 and transgenic grou treatment of PEG6000; two-way ANOVA tests were performed to detect the significant diffe in the content of each index between WT and transgenic groups in the same treatment group on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).
(4) Activities of antioxidant enzymes in transgenic Arabidopsis thaliana with GmN under drought stress simulated by PEG6000 There was no significant difference in the activities of the three antioxidant enz on MS medium without PEG6000 between the WT and transgenic A. thaliana GmNAC19 (Figure 9).On MS medium with PEG6000, superoxide dismutase ( showed higher activity in the transgenic group than that on MS medium without PEG  -6) under drought stress simulated by 6% PEG6000 for 15 d.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the content of each index between WT without treatment of PEG6000 and WT with treatment of PEG6000, and between transgenic group without treatment of PEG6000 and transgenic group with treatment of PEG6000; two-way ANOVA tests were performed to detect the significant differences in the content of each index between WT and transgenic groups in the same treatment group based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).
(4) Activities of antioxidant enzymes in transgenic Arabidopsis thaliana with GmNAC19 under drought stress simulated by PEG6000 There was no significant difference in the activities of the three antioxidant enzymes on MS medium without PEG6000 between the WT and transgenic A. thaliana with GmNAC19 (Figure 9).On MS medium with PEG6000, superoxide dismutase (SOD) showed higher activity in the transgenic group than that on MS medium without PEG6000 (Figure 9A).On MS medium containing 6% PEG6000, the enzymatic activity of peroxidase (POD) was higher in the transgenic A. thaliana with GmNAC19 than that in the WT (Figure 9B), while higher activities of both POD and catalase (CAT) were observed in the WT and transgenic groups than those on MS medium without PEG6000 (Figure 9B,C).It was noted that a significant increase was detected in the enzymatic activities of SOD, POD, and CAT in transgenic A. thaliana on MS medium containing PEG6000 compared with the WT on MS without PEG6000.
Int. J. Mol.Sci.2024, 25, x FOR PEER REVIEW 10 of 30 9B), while higher activities of both POD and catalase (CAT) were observed in the WT and transgenic groups than those on MS medium without PEG6000 (Figure 9B,C).It was noted that a significant increase was detected in the enzymatic activities of SOD, POD, and CAT in transgenic A. thaliana on MS medium containing PEG6000 compared with the WT on MS without PEG6000.under drought stress simulated by 6% PEG6000 for 15 d.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the enzymatic activities between WT without treatment of PEG6000 and WT with treatment of PEG6000, and between transgenic group without treatment of PEG6000 and transgenic group with treatment of PEG6000; twoway ANOVA tests were performed to detect the significant differences in the enzymatic activities between WT and transgenic groups in the same treatment group (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).Note: A significant increase is detected in the enzymatic activities of SOD, POD, and CAT in transgenic A. thaliana on MS medium containing PEG6000 compared with WT on MS without PEG6000 based on Student's t test (p < 0.05).
• Drought tolerance of transgenic Arabidopsis thaliana with GmNAC19 under drought stress induced by water loss (1) Contents of physiological and biochemical indices related to drought stress of transgenic Arabidopsis thaliana with GmNAC19 under drought stress induced by water loss Significant variations in the content of soluble protein were observed between WT and transgenic A. thaliana under drought stress; in 15 d, the content of soluble protein was decreased in the WT, with an average of 72.1 mg g −1 , which was significantly lower than that of the transgenic group of A. thaliana, with an average of 75.9 mg g −1 (Figure 10A).In 15 d, the contents of soluble sugar in the WT and transgenic groups were significantly increased by 0.70 mg g −1 and 0.87 mg g −1 , respectively, in comparison to the WT without drought stress (Figure 10B), and it was noted that a significant increase in the content of soluble sugar was detected in transgenic A. thaliana under drought stress compared with the WT without drought stress.Under drought stress, a significant increase was detected in the proline content of transgenic A. thaliana compared with the WT (Figure 10C); these results were consistent with the above results based on the treatment of drought stress simulated by PEG6000.The contents of MDA were found to exhibit an increasing pattern in both the WT and transgenic groups, with the contents of MDA in the transgenic group significantly lower than those of the WT at 10 and 15 d, respectively, under drought stress (Figure 10D).-6) under drought stress simulated by 6% PEG6000 for 15 d.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the enzymatic activities between WT without treatment of PEG6000 and WT with treatment of PEG6000, and between transgenic group without treatment of PEG6000 and transgenic group with treatment of PEG6000; two-way ANOVA tests were performed to detect the significant differences in the enzymatic activities between WT and transgenic groups in the same treatment group (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).Note: A significant increase is detected in the enzymatic activities of SOD, POD, and CAT in transgenic A. thaliana on MS medium containing PEG6000 compared with WT on MS without PEG6000 based on Student's t test (p < 0.05).

•
Drought tolerance of transgenic Arabidopsis thaliana with GmNAC19 under drought stress induced by water loss (1) Contents of physiological and biochemical indices related to drought stress of transgenic Arabidopsis thaliana with GmNAC19 under drought stress induced by water loss Significant variations in the content of soluble protein were observed between WT and transgenic A. thaliana under drought stress; in 15 d, the content of soluble protein was decreased in the WT, with an average of 72.1 mg g −1 , which was significantly lower than that of the transgenic group of A. thaliana, with an average of 75.9 mg g −1 (Figure 10A).In 15 d, the contents of soluble sugar in the WT and transgenic groups were significantly increased by 0.70 mg g −1 and 0.87 mg g −1 , respectively, in comparison to the WT without drought stress (Figure 10B), and it was noted that a significant increase in the content of soluble sugar was detected in transgenic A. thaliana under drought stress compared with the WT without drought stress.Under drought stress, a significant increase was detected in the proline content of transgenic A. thaliana compared with the WT (Figure 10C); these results were consistent with the above results based on the treatment of drought stress simulated by PEG6000.The contents of MDA were found to exhibit an increasing pattern in both the WT and transgenic groups, with the contents of MDA in the transgenic group significantly lower than those of the WT at 10 and 15 d, respectively, under drought stress (Figure 10D).(2) Activities of antioxidant enzymes of transgenic Arabidopsis thaliana with GmNAC19 under drought stress induced by water loss With the extension of drought stress treatment time, the enzymatic activities of SOD, POD, and CAT in both the WT and transgenic groups of A. thaliana were generally increased, reaching the highest levels at 15 d (Figure 11).At 15 d, the levels of enzymatic activities of SOD, POD, and CAT in the transgenic group were increased by 1847.7,2902.3, and 95.8 U g −1 , respectively, which were higher than those in the WT, i.e., 1104.4,2095.3, and 52.4 U g −1 , respectively.It was noted that at 15 d, the enzymatic activities of SOD, POD, and CAT were significantly increased in the transgenic groups compared with the WT without drought stress.(2) Activities of antioxidant enzymes of transgenic Arabidopsis thaliana with GmNAC19 under drought stress induced by water loss With the extension of drought stress treatment time, the enzymatic activities of SOD, POD, and CAT in both the WT and transgenic groups of A. thaliana were generally increased, reaching the highest levels at 15 d (Figure 11).At 15 d, the levels of enzymatic activities

Enhanced Drought Tolerance in Transgenic Soybean Composite Lines with GmNAC19
• Content of chlorophyll in transgenic soybean composite lines with GmNAC19 under drought stress simulated by 20% PEG6000 As the treatment time of drought stress was increased from 0 to 9 h, the chlorophyll contents of both the control group and the transgenic soybean composite lines with GmNAC19 showed no significant variation (Figure 12).At 12 h, the chlorophyll contents of both control group and transgenic soybean composite lines with GmNAC19 were significantly lower than those without drought stress (0 d), respectively.The transgenic group was revealed to have significantly higher chlorophyll content than the control group at all sampling time points.

Figure 12.
Chlorophyll content in the control and transgenic soybean composite lines with GmNAC19 (OE) under drought stress simulated by 20% PEG6000.The control group contains plants transformed with Agrobacterium rhizogenes K599 without GmNAC19.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the chlorophyll content between control at 0 h and controls at 3, 6, 9, and 12 h, and between transgenic group at 0 h and transgenic groups at 3, 6, 9, and 12 h; two-way ANOVA tests were performed to detect the significant differences in the chlorophyll content between control and transgenic groups at the same sampling time point (indicated by asterisk above the bracket) based on p < 0.05 (*) and p < 0.01 (**) (n = 3).

Enhanced Drought Tolerance in Transgenic Soybean Composite Lines with GmNAC19
• Content of chlorophyll in transgenic soybean composite lines with GmNAC19 under drought stress simulated by 20% PEG6000 As the treatment time of drought stress was increased from 0 to 9 h, the chlorophyll contents of both the control group and the transgenic soybean composite lines with GmNAC19 showed no significant variation (Figure 12).At 12 h, the chlorophyll contents of both control group and transgenic soybean composite lines with GmNAC19 were significantly lower than those without drought stress (0 d), respectively.The transgenic group was revealed to have significantly higher chlorophyll content than the control group at all sampling time points.

•
Contents of physiological and biochemical indicators related to drought stress in transgenic soybean composite lines with GmNAC19 under drought stress simulated by 20% PEG6000 No significant differences were detected in the contents of soluble protein in hairy roots in the control group and the transgenic soybean composite lines with GmNAC19, while as the treatment time was increased, significant variations in the contents of soluble protein were observed in the stems and leaves of the control and the transgenic soybean composite lines with GmNAC19 (Figure 13A-C).At 12 h, the content of soluble protein in the leaves of the transgenic group was significantly higher than that of the control.
contents of both the control group and the transgenic soybean composite lines with GmNAC19 showed no significant variation (Figure 12).At 12 h, the chlorophyll contents of both control group and transgenic soybean composite lines with GmNAC19 were significantly lower than those without drought stress (0 d), respectively.The transgenic group was revealed to have significantly higher chlorophyll content than the control group at all sampling time points.

Figure 12.
Chlorophyll content in the control and transgenic soybean composite lines with GmNAC19 (OE) under drought stress simulated by 20% PEG6000.The control group contains plants transformed with Agrobacterium rhizogenes K599 without GmNAC19.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the chlorophyll content between control at 0 h and controls at 3, 6, 9, and 12 h, and between transgenic group at 0 h and transgenic groups at 3, 6, 9, and 12 h; two-way ANOVA tests were performed to detect the significant differences in the chlorophyll content between control and transgenic groups at the same sampling time point (indicated by asterisk above the bracket) based on p < 0.05 (*) and p < 0.01 (**) (n = 3).
Figure 12.Chlorophyll content in the control and transgenic soybean composite lines with GmNAC19 (OE) under drought stress simulated by 20% PEG6000.The control group contains plants transformed with Agrobacterium rhizogenes K599 without GmNAC19.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the chlorophyll content between control at 0 h and controls at 3, 6, 9, and 12 h, and between transgenic group at 0 h and transgenic groups at 3, 6, 9, and 12 h; two-way ANOVA tests were performed to detect the significant differences in the chlorophyll content between control and transgenic groups at the same sampling time point (indicated by asterisk above the bracket) based on p < 0.05 (*) and p < 0.01 (**) (n = 3).Agrobacterium rhizogenes K599 without GmNAC19.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the content of each index between control at 0 h and controls at 3, 6, 9, and 12 h, and between transgenic group at 0 h and transgenic groups at 3, 6, 9, and 12 h; two-way ANOVA tests were performed to detect the significant differences in the content of each index between control and transgenic groups at the same sampling time point (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).Note: At 12 h, a significant difference is detected between the transgenic plant and the control without drought stress in the content of soluble protein in the leaves, in the content of soluble sugar in all three organs, in the proline content in the stems and leaves, and in the content of MDA in both the roots and leaves based on Student's t test (p < 0.05).
No significant variations were revealed in the contents of soluble sugar in the hairy roots and stems in the control group, while a significant difference was observed in stems and leaves of control group at 9 and 12 h, respectively (Figure 13D-F).As the treatment time of drought stress was increased to 9 and 12 h, the soluble sugar content in the hairy roots, stems, and leaves of the transgenic groups were significantly higher than that at 0 h.At 9 and 12 h, the contents of soluble sugar in the stems and leaves of the transgenic group were significantly higher than those of the control group.
No significant differences were detected in the contents of proline in the hairy roots, stems, and leaves of the control group, while the transgenic group showed significant variations in the stems (at 12 h) and leaves (at 9 and 12 h) (Figure 13G-I).Significant variations were revealed in the stems (at 9 h) and leaves (at 9 and 12 h) between the control and the transgenic groups.
The contents of MDA were generally not significantly different in the hairy roots, stems, and leaves of the control and the transgenic groups, except for the hairy roots (at 12 h) in the transgenic group and the stems (at 9 h) in the control group (Figure 13J-L).Significant variations were revealed in the hairy roots (at 12 h), stems (at 6 and 9 h), and leaves (at 3 h) between the control and the transgenic groups.
• Antioxidant enzyme activities of transgenic soybean composite lines with GmNAC19 under drought stress simulated by 20% PEG6000 As the treatment time of drought stress was increased, the SOD activities in the hairy roots, stems, and leaves of both the control and the transgenic groups showed an increasing pattern, showing significantly higher SOD activities in the hairy roots (at 6, 9, and 12 h), stems (at 3, 6, and 12 h), and leaves (at 12 h) of the transgenic group than those in the control group (Figure 14A-C).At 12 h, the SOD activities in the hairy roots, stems, and leaves were increased to 4.3, 3.1, and 1.9 times higher than those without the treatment of drought stress in the control group, and 7.7, 4.3, and 2.9 times higher than those without the treatment of drought stress in the transgenic group, respectively.
The highest levels of enzymatic activities of POD were detected in the hairy roots, followed by the leaves and stems (Figure 14D-F).In the hairy roots, the enzymatic activities of POD were first increased (reaching the highest levels at 3 and 6 h, which was 1.78 times higher than that at 0 h), and then, decreased (at 9 and 12 h) in the control group, whereas the transgenic group showed an increasing pattern; significant differences were revealed in the hairy roots between the control and transgenic groups at 3, 9, and 12 h.In both the stems and leaves, with an increase in the PEG6000 treatment time, the POD activities were significantly increased in the stems (at 6, 9, and 12 h in the control and at 9 and 12 h in the transgenic group) and leaves (at 12 h in the control and at 9 and 12 h in the transgenic group).At 12 h, the POD activities in the hairy roots, stems, and leaves of the transgenic group were 1.5, 1.3, and 1.4 times higher than those in the control group.
With an increase in the PEG6000 treatment time, the enzymatic activities of CAT in the hairy roots, stems, and leaves were generally increased in both the control and transgenic groups (Figure 14G-I).A significant difference in the enzymatic activity of CAT was observed in the leaves between the control and the transgenic groups at 9 and 12 h.At 12 h, the CAT activities were increased by 56.2, 55.4, and 58.6 U g −1 in the hairy roots, stems, and leaves, respectively, in the control group, and by 64.2, 83.9, and 86.3 U g −1 , respectively, in the transgenic group.It was noted that at 12 d, the enzymatic activities of SOD, POD, and CAT in all three organs of the transgenic plants were significantly higher than those of the controls without drought stress (0 d).

•
Gene expression of GmNAC19 in soybean hairy roots under drought stress simulated by 20% PEG6000 Under the treatment of drought stress, the expression levels of GmNAC19 in the control group were first increased to the highest level at 6 h, and then, decreased (at 12 h) (Figure 15), which was consistent with the expression pattern of GmNAC19 revealed by the transcriptome sequencing and qRT-PCR analyses (Figures 1 and 2), showing that the gene expression levels of GmNAC19 in soybean roots treated with 20% PEG6000 were first increased from 0 to 6 h, reaching the highest expression level at 6 h, and then, decreased from 6 to 24 h (Figure 2).In the transgenic group, the gene expression of GmNAC19 was gradually increased and reached the highest level at 12 h, which was 53 and 47 times higher than that of the control and the transgenic groups without the treatment of drought stress (at 0 h), respectively.G-I) in the hairy roots, stems, and leaves of transgenic soybean composite lines with GmNAC19 (OE) under drought stress simulated by 20% PEG6000.The control groups contain the plants transformed with Agrobacterium rhizogenes K599 without GmNAC19.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the enzymatic activity between control at 0 h and controls at 3, 6, 9, and 12 h, and between transgenic group at 0 h and transgenic groups at 3, 6, 9, and 12 h; two-way ANOVA tests were performed to detect the significant differences in the enzymatic activity between control and transgenic groups at the same sampling time point (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).Note: At 12 h, the enzymatic activities of SOD, POD, and CAT in all three organs of transgenic plants are significantly higher than those of the controls without drought stress (0 d) based on Student's t test (p < 0.05).

•
Gene expression of GmNAC19 in soybean hairy roots under drought stress simulated by 20% PEG6000 Under the treatment of drought stress, the expression levels of GmNAC19 in the control group were first increased to the highest level at 6 h, and then, decreased (at 12 h) (Figure 15), which was consistent with the expression pattern of GmNAC19 revealed by  G-I) in the hairy roots, stems, and leaves of transgenic soybean composite lines with GmNAC19 (OE) under drought stress simulated by 20% PEG6000.The control groups contain the plants transformed with Agrobacterium rhizogenes K599 without GmNAC19.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the enzymatic activity between control at 0 h and controls at 3, 6, 9, and 12 h, and between transgenic group at 0 h and transgenic groups at 3, 6, 9, and 12 h; two-way ANOVA tests were performed to detect the significant differences in the enzymatic activity between control and transgenic groups at the same sampling time point (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).Note: At 12 h, the enzymatic activities of SOD, POD, and CAT in all three organs of transgenic plants are significantly higher than those of the controls without drought stress (0 d) based on Student's t test (p < 0.05).
gene expression levels of GmNAC19 in soybean roots treated with 20% PEG6000 were first increased from 0 to 6 h, reaching the highest expression level at 6 h, and then, decreased from 6 to 24 h (Figure 2).In the transgenic group, the gene expression of GmNAC19 was gradually increased and reached the highest level at 12 h, which was 53 and 47 times higher than that of the control and the transgenic groups without the treatment of drought stress (at 0 h), respectively.

Figure 15.
Relative expression of GmNAC19 in transgenic soybean hairy roots with GmNAC19 (OE) under drought stress simulated by 20% PEG6000.The control group contains the soybean hairy roots transformed with Agrobacterium rhizogenes K599 without GmNAC19.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the relative expression of GmNAC19 between control at 0 h and controls at 3, 6, 9, and 12 h, and between transgenic group at 0 h and transgenic groups at 3, 6, 9, and 12 h; two-way ANOVA tests were performed to detect the significant differences in the relative expression of GmNAC19 between control and transgenic groups at the same sampling time point (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).

•
Expression of four key genes involved in proline metabolic pathway in transgenic soybean hairy roots with GmNAC19 under drought stress simulated by 20% PEG6000 Varied expression patterns were observed in the four key genes, i.e., P5CS, OAT, P5CR, and ProDH, involved in the proline metabolic pathway (Figure 16).The expression levels of P5CS in the control group were first decreased (at 3 h), and then, increased to the highest levels (at 6 and 9 h), and finally, decreased (at 12 h), whereas significant variation was only detected at 12 h in the transgenic group (Figure 16A); the expression levels of P5CS showed significant differences between the control and transgenic groups at 3, 6, and 9 h.The same expression pattern of OAT was revealed in both the control and transgenic groups, i.e., first increased (at 6 and 9 h in the control group and at 6 h in the transgenic group), and then, decreased (at 12 h in the control group and at 9 and 12 h in the transgenic group) (Figure 16B); the expression levels of OAT showed significant differences between the control and transgenic groups at 6, 9, and 12 h.The expression patterns of P5CR were largely the same as those of OAT, showing that the expression levels of P5CR were first increased (at 6 and 9 h), and then, decreased (at 12 h) in both the control and transgenic groups (Figure 16C).The expression levels of ProDH were first increased (at 3 h, reaching the highest level, which was 3.95 times higher than that of the control group at 0 h), and then, decreased (at 6, 9, and 12 h) in the control group, whereas in the transgenic group, the expression levels of ProDH were first increased (at 3 h), and then, Figure 15.Relative expression of GmNAC19 in transgenic soybean hairy roots with GmNAC19 (OE) under drought stress simulated by 20% PEG6000.The control group contains the soybean hairy roots transformed with Agrobacterium rhizogenes K599 without GmNAC19.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the relative expression of GmNAC19 between control at 0 h and controls at 3, 6, 9, and 12 h, and between transgenic group at 0 h and transgenic groups at 3, 6, 9, and 12 h; two-way ANOVA tests were performed to detect the significant differences in the relative expression of GmNAC19 between control and transgenic groups at the same sampling time point (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).

•
Expression of four key genes involved in proline metabolic pathway in transgenic soybean hairy roots with GmNAC19 under drought stress simulated by 20% PEG6000 Varied expression patterns were observed in the four key genes, i.e., P5CS, OAT, P5CR, and ProDH, involved in the proline metabolic pathway (Figure 16).The expression levels of P5CS in the control group were first decreased (at 3 h), and then, increased to the highest levels (at 6 and 9 h), and finally, decreased (at 12 h), whereas significant variation was only detected at 12 h in the transgenic group (Figure 16A); the expression levels of P5CS showed significant differences between the control and transgenic groups at 3, 6, and 9 h.The same expression pattern of OAT was revealed in both the control and transgenic groups, i.e., first increased (at 6 and 9 h in the control group and at 6 h in the transgenic group), and then, decreased (at 12 h in the control group and at 9 and 12 h in the transgenic group) (Figure 16B); the expression levels of OAT showed significant differences between the control and transgenic groups at 6, 9, and 12 h.The expression patterns of P5CR were largely the same as those of OAT, showing that the expression levels of P5CR were first increased (at 6 and 9 h), and then, decreased (at 12 h) in both the control and transgenic groups (Figure 16C).The expression levels of ProDH were first increased (at 3 h, reaching the highest level, which was 3.95 times higher than that of the control group at 0 h), and then, decreased (at 6, 9, and 12 h) in the control group, whereas in the transgenic group, the expression levels of ProDH were first increased (at 3 h), and then, decreased at 6 h, and finally, increased (at 12 h, reaching the highest level, which was 8.46 times higher than that of the control group at 0 h) (Figure 16D); the expression levels of ProDH showed a significant difference between the control and transgenic groups at 9 and 12 h.decreased at 6 h, and finally, increased (at 12 h, reaching the highest level, which was 8.46 times higher than that of the control group at 0 h) (Figure 16D); the expression levels of ProDH showed a significant difference between the control and transgenic groups at 9 and 12 h.Agrobacterium rhizogenes K599 without GmNAC19.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the relative expression of each gene between control at 0 h and controls at 3, 6, 9, and 12 h, and between transgenic group at 0 h and transgenic groups at 3, 6, 9, and 12 h; two-way ANOVA tests were performed to detect the significant differences in the relative expression of each gene between control and transgenic groups at the same sampling time point (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).

•
Correlation analysis between gene expression levels of GmNAC19 and four key genes involved in proline metabolic pathway Pearson correlation analyses between the gene expression level of GmNAC19 (Figure 15) and the expression levels of the four key genes (Figure 16) involved in proline metabolism in soybean hairy roots were performed in both the control and the transgenic groups (Table 2).The results showed that in the control group, the gene expression level of GmNAC19 was significantly positively correlated with that of the three genes (i.e., P5CS, OAT, and P5CR) involved in the synthesis of proline and was not significantly correlated with that of the gene ProDH involved in the breakdown of proline.In the transgenic group, the gene expression level of GmNAC19 was significantly positively and negatively correlated with that of ProDH and P5CS, respectively.Agrobacterium rhizogenes K599 without GmNAC19.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the relative expression of each gene between control at 0 h and controls at 3, 6, 9, and 12 h, and between transgenic group at 0 h and transgenic groups at 3, 6, 9, and 12 h; two-way ANOVA tests were performed to detect the significant differences in the relative expression of each gene between control and transgenic groups at the same sampling time point (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).

•
Correlation analysis between gene expression levels of GmNAC19 and four key genes involved in proline metabolic pathway Pearson correlation analyses between the gene expression level of GmNAC19 (Figure 15) and the expression levels of the four key genes (Figure 16) involved in proline metabolism in soybean hairy roots were performed in both the control and the transgenic groups (Table 2).The results showed that in the control group, the gene expression level of GmNAC19 was significantly positively correlated with that of the three genes (i.e., P5CS, OAT, and P5CR) involved in the synthesis of proline and was not significantly correlated with that of the gene ProDH involved in the breakdown of proline.In the transgenic group, the gene expression level of GmNAC19 was significantly positively and negatively correlated with that of ProDH and P5CS, respectively.Table 2. Pearson correlation analysis of gene expression of GmNAC19 with the expressions of 4 key genes (i.e., P5CS, OAT, P5CR, and ProDH) involved in proline metabolism in the control and transgenic soybean hairy roots with GmNAC19 based on the gene expression data presented in Figures 15 and 16, respectively.The control group contains soybean hairy roots transformed with Agrobacterium rhizogenes K599 without GmNAC19.Symbols "*" and "**" indicate significant differences in the correlation coefficients determined with p < 0.05 and p < 0.01, respectively.

Gene Expression Patterns of Soybean Transcription Factor GmNAC19
In response to drought stress, drought tolerance genes in plants are expressed and have been revealed to have varied expression levels and activities involved in many metabolic pathways to enhance drought tolerance in plants [13,[39][40][41].TFs are important and indispensable regulatory proteins involved in numerous biological activities in plants, including drought tolerance.In particular, NAC TFs play a key role in the molecular mechanism underlying drought tolerance in plants in response to drought stress by regulating the expression of related genes at the transcriptional level [28][29][30][31][32][33][34][35][36][37][38].
To date, many NAC genes have been revealed to exhibit enhanced expression under drought stress in many model plants, including rice, Arabidopsis, and soybean [36][37][38]42].In our study, the results of transcriptome sequencing analysis of the soybean variety 'Jiyu47' under drought stress revealed significant up-regulation in the expression of the TF gene GmNAC19, reaching the highest level (about 20 times higher than that without the treatment of drought stress) at 6 h under the treatment of drought stress, and then, increased to 12 times higher than that without the treatment of drought stress at 18 h, and finally, increased to 13.6 times higher than that without the treatment of drought stress at 24 h.These results were consistent with those derived from the transgenic soybean hairy roots under drought stress simulated by PEG6000.It is worth noting that the different NAC genes located in the nucleus showed varied molecular responses to drought stress (Table 3).For example, both GmNAC3 and GmNAC4 responded to drought stress mainly by increasing the enzymatic activities of SOD and CAT, as well as increasing the proline content, while GmNAC8 responded to drought stress by increasing the enzymatic activity of POD and the proline content [36][37][38].Our results showed that GmNAC19 responded to drought stress by not only significantly increasing the enzymatic activities of POD, SOD, and CAT and the chlorophyll content of soybeans, but also regulating chlorophyll content to maintain photosynthesis, respiration, and material accumulation, further influencing plant growth and development, and ultimately enhancing drought resistance in soybean plants.
Table 3.Molecular characteristics and responses to drought stress conferred by four NAC genes.Symbols "+" and "−" indicate a significant response and no significant response to drought stress, respectively."n/a" indicates that data are not available.POD, peroxidase; CAT, catalase; SOD, superoxide dismutase; MDA, malondialdehyde.Various expression patterns have been reported in many genes in response to drought stress simulated by PEG6000.In our study, the gene expression of GmNAC19 detected in hairy roots prior to the treatment of 20% PEG6000 was first increased, and then, decreased, reaching the highest expression level at 6 h, which was 8 times higher than that at 0 h.These results were consistent with those of the expression patterns of many TFs under drought stress.For example, the expressions of JrMYB44 in walnut [43], five PgbZIP genes in ginseng seedlings [44], and OsbZIP62 in rice [45] were first increased, and then, decreased as the treatment time of PEG6000 was increased.Similarly, the expression levels of GmNAC3 and GmNAC4 in soybean roots were first increased, and then, decreased under the treatment of PEG6000 [36,37].
Studies have shown that different TFs generally function at different sites of multiple metabolic pathways.For example, two TF genes of Arabidopsis thaliana, ANAC087 and ANAC046, were expressed in large quantities in the roots and involved in apoptosis [46], while TaRNAC1 was highly expressed in wheat roots to improve the growth of the root system and enhance drought tolerance [47].Furthermore, studies have shown that the expression levels of NAC TF genes such as GmNAC3 and GmNAC4 are generally high in soybean roots [36,37].However, our results showed that the highest gene expression levels of GmNAC19 were revealed in leaves during the seedling, flowering, and podding stages.Similarly, the highest expression level of HaDREBA5 was detected in the leaves of sunflower, with the overexpression of HaDREBA5 enhancing tolerance to low temperature, drought, and salt in tobacco plants [48].
Moreover, our results for the gene expression of GmNAC19 evaluated in transgenic soybean hairy roots showed that the gene expression of GmNAC19 and its expression pattern in the control group were consistent with those derived from the transcriptome sequencing analysis, and the gene expression level of GmNAC19 in the transgenic group showed an increasing pattern, reaching the highest level at 12 h.These results indicated that GmNAC19 significantly responded to the treatment of drought stress.Similarly, previous studies showed that under drought stress, multiple genes were expressed in plants to regulate the downstream products to alleviate the detrimental effects caused by drought [13]; both ANAC087 and ANAC046 were highly expressed in the roots of A. thaliana [46], and TaRNAC1 was largely expressed in wheat roots [47], to improve the drought tolerance of plants.

Enhanced Drought Tolerance Conferred by Soybean Transcription Factor Gene GmNAC19
Our results revealed enhanced drought tolerance in transgenic Saccharomyces cerevisiae INVSc1 with GmNAC19, as observed in its improved growth performance based on the number of transgenic yeast cells grown on the YPD medium containing mannitol.These results indicated the alleviating effect of GmNAC19 on drought stress, which was consistent with the results previously reported, showing the significantly improved drought resistance of yeast induced by the overexpression of walnut JreIF1A [49].Furthermore, our results revealed improvements in seed germination percentage, root growth, drought tolerance, and recovery ability after rehydration in the transgenic A. thaliana with GmNAC19 in comparison to those of WT plants under drought stress either simulated by PEG6000 or caused by water loss, indicating that GmNAC19 could enhance drought tolerance in A. thaliana.These results were consistent with those previously reported, showing that the overexpression of CaNAC46 and CarNAC4 improved the drought tolerance of Arabidopsis [28,34].Furthermore, previous studies showed that the plant response to adverse environmental conditions is indicated by maintaining a high seed germination rate because seed germination is generally inhibited by abiotic stresses [50], suggesting that soybean yield and quality could be affected in response to environmental stresses, e.g., drought.Moreover, studies indicated that the up-regulation of the transcription factor MYB68 improved the salt tolerance, the number of seeds per pod, and the weight per 100 seeds of soybean, while high levels of soluble sugar and proline were important indicators in response to environmental stresses [51].In addition, our results showed that under the drought stress treatment with 20% PEG6000, the chlorophyll content of leaves in the transgenic soybean composite lines with GmNAC19 was higher than that in the control group.These results were consistent with those previously reported, showing that under drought conditions, high chlorophyll contents were revealed in the leaves of soybean lines with overexpression of GmNAC3 and GmNAC4 [36,37], while the overexpression of MdNAC1 promoted photosynthesis in apple [30], ultimately improving the drought tolerance of the plants.

Variations in Physiological and Biochemical Indices Related to Drought Tolerance Conferred by the Gene GmNAC19
Studies have shown that under drought stress, the level of reactive oxygen species (ROS) harmful to plants is increased [36], while antioxidant enzymes play an important role in reducing the level of ROS and resisting oxidative stress in plants [8,52].Under drought conditions, the antioxidant and osmotic regulatory systems are generally activated, as indicated by the enhancement in antioxidant oxidase activity and the accumulation of soluble sugar, soluble protein, and proline.Therefore, variations in drought tolerance could be evaluated based on the contents of osmoregulatory substances and the activities of antioxidant enzymes (e.g., SOD, CAT, and POD) involved in the elimination of ROS [6].Furthermore, studies have revealed the cytotoxic effects of MDA, including the inhibition of gene expression and the promotion of cell death, while under drought stress, plants produce a large quantity of MDA, ultimately causing damage in plants [53].It is well known that numerous TFs are involved in improved resistance to adverse conditions by increasing the enzymatic activities of SOD, POD, and CAT, increasing the contents of soluble protein, soluble sugar, and proline, and decreasing the content of MDA.For example, previous studies showed that the accumulation of proline in transgenic Arabidopsis with CarNAC4 was increased [34], the contents of proline and soluble sugar in rice with overexpression of OsNAC066 were increased [35], the content of proline in transgenic A. thaliana with GmWRKY16 was increased [54], the enzymatic activities of SOD, POD, and CAT and the content of proline were increased, and the content of MDA was decreased in transgenic Arabidopsis with overexpression of VvNAC17 [33].These results were consistent with the findings revealed in our study, showing enhanced drought tolerance, as observed in the increased contents of soluble protein and soluble sugar and the enzymatic activities of SOD, POD, and CAT, and the decreased content of MDA, in transgenic A. thaliana with GmNAC19 treated with drought stress either simulated by PEG6000 or caused by water loss.Furthermore, our results showed that the proline content in the transgenic A. thaliana was significantly increased as the treatment time of drought stress was increased, suggesting a higher level of adaptation to drought stress than the control group.These results were consistent with those previously reported, with one study showing an increased concentration of proline in transgenic lemon with FcWRKY40 [55].
We further evaluated the contents of osmoregulatory substances, antioxidant enzyme activities, and the content of MDA in the hairy roots, stems, and leaves of soybean.Our results revealed increased contents of soluble protein, soluble sugar, and proline, enhanced enzymatic activities of SOD, POD, and CAT, and decreased content of MDA in the transgenic soybean with GmNAC19 compared to those in the control group.These results were consistent with those based on the transgenic A. thaliana with GmNAC19 and in accordance with those previously reported.For example, increased enzymatic activities of SOD and POD and decreased content of MDA were revealed in birch with overexpression of BpMYB123 under drought stress [56], while decreased enzymatic activities of SOD, POD, and CAT and increased content of MDA were detected in rice with knockout of OsNAC006 under drought stress [57].Moreover, three soybean NAC genes were revealed to have largely the same alleviating effect on the enhancement of drought tolerance, i.e., the enzymatic activities of POD and CAT and the content of proline were increased in soybean lines with overexpression of GmNAC3 and GmNAC4, the enzymatic activity of SOD and the content of proline were increased in soybean lines with overexpression of GmNAC8, and the enzymatic activity of SOD and the content of proline were decreased in soybean lines with defective GmNAC8 [36][37][38].Our results showed that in the hairy roots, stems, and leaves, the enzymatic activity of POD and the chlorophyll content of transgenic soybean were higher than those of the control group, indicating that the transgenic soybean with GmNAC19 could regulate POD activity and chlorophyll content to maintain the photosynthesis, respiration, material accumulation, and growth and development of plants, ultimately enhancing their drought tolerance.These results were consistent with those previously reported in various crop and model plants [36,37,56].For example, the overexpression of cherry ChNAC1 in Arabidopsis caused an increase in the contents of chlorophyll, water, proline, and protein, as well as the enzymatic activities of POD and SOD [58].Furthermore, the overexpression of ThNAC4 in both Tamarix and Arabidopsis enhanced the activities of antioxidant enzymes (e.g., SOD and POD) and the contents of osmoprotectants (e.g., proline) under stress conditions [59].Moreover, the overexpression of SlNAC6 in tomato caused a significant delay in growth with enhanced tolerance to PEG stress, decreased water loss and oxidative damage, and increased levels of proline content and antioxidant enzyme activity [60].

Proline Metabolism in Transgenic Soybean and Arabidopsis thaliana with GmNAC19
Numerous studies have shown that the accumulation of proline in plants is positively correlated with the improvement in stress resistance [61].It is well known that proline is accumulated in plant cells to maintain cell turgor [62], improve protein stability and protect membrane integrity by binding with hydrogen bonds [63], and protect cells by increasing water absorption potential and promoting enzyme activation [64].Proline is also a powerful antioxidant defense molecule, metal-chelating agent, protein stabilizer, ROS scavenger, and inhibitor of apoptosis [62].However, studies have shown that excessive accumulation of proline could lead to impaired cellular and physiological functions [65].In our study, the results showed that under drought stress either simulated by PEG6000 or caused by water loss, the contents of proline in A. thaliana were revealed to have an increasing pattern as the treatment time of drought stress was increased, while the content of proline in the transgenic A. thaliana was lower than that in the control group.These results were consistent with those previously reported, showing that under drought stress, most transgenic ryegrass with CBF1 was revealed to have lower content of proline than that of the control group [66].
In order to further explore the relationships between the gene expression level of GmNAC19 and the content of proline, the variations in the relative expression levels of four key genes involved in the proline metabolic pathway were evaluated in the transgenic soybean composite lines, i.e., three genes (P5CS, P5CR, and OAT) related to the synthesis of proline and one gene ProDH related to the breakdown of proline [67].The results showed that the contents of proline in the hairy roots, stems, and leaves of the transgenic soybean composite lines were consistently higher than those in the control group.Furthermore, the results of the correlation analysis showed that in the control group, the gene expression level of GmNAC19 was significantly positively correlated with that of P5CS, OAT, and P5CR, whereas in the transgenic group, the gene expression level of GmNAC19 was significantly positively and negatively correlated with that of ProDH and P5CS, respectively (Figure 17).These results indicated that GmNAC19 was involved in the regulation of proline content by regulating the expression of genes involved in the proline metabolic pathway, ultimately maintaining the dynamic balance between the content of proline and the level of drought stress to improve drought tolerance in soybean.These results were consistent with those previously reported [36,37,68], showing that the overexpression of many TF genes was positively correlated with the expression of key genes involved in the proline metabolic pathway, ultimately increasing the content of proline and adaptation to adverse conditions.For example, the TF gene DFR1 maintained the functional activities of proline and improved drought tolerance in Arabidopsis through the down-regulation of glutamate synthesis based on guanidinosuccinic acid, ultimately promoting the synthesis of proline and drought tolerance [65].Furthermore, our results showed that under drought stress, both the transgenic A. thaliana and soybean composite lines were revealed to have enhanced levels of soluble protein compared with the control group.However, these increases were not comparable with those of soluble sugars, whereas similar variations were observed between the contents of soluble protein and proline, suggesting that proline was not only involved in the regulation of the plant response to drought stress, but also used as a material for the synthesis of soluble protein.These results were consistent with those previously reported, showing that plants responded to adverse conditions by regulating the content of proline, e.g., the synthesis level of proline was decreased in A. thaliana with knockout of P5CS, while the expression of P5CS was up-regulated in rice with overexpression of OsWRKY50 under salt stress [68].In summary, the expression levels of these key genes were crucial for the accumulation of proline under drought stress.the synthesis of proline and drought tolerance [65].Furthermore, our results showed that under drought stress, both the transgenic A. thaliana and soybean composite lines were revealed to have enhanced levels of soluble protein compared with the control group.However, these increases were not comparable with those of soluble sugars, whereas similar variations were observed between the contents of soluble protein and proline, suggesting that proline was not only involved in the regulation of the plant response to drought stress, but also used as a material for the synthesis of soluble protein.These results were consistent with those previously reported, showing that plants responded to adverse conditions by regulating the content of proline, e.g., the synthesis level of proline was decreased in A. thaliana with knockout of P5CS, while the expression of P5CS was upregulated in rice with overexpression of OsWRKY50 under salt stress [68].In summary, the expression levels of these key genes were crucial for the accumulation of proline under drought stress.2).Symbols "*" and "**" indicate significant differences in the correlation coefficients determined with p < 0.05 and p < 0.01, respectively (Table 2).Symbols "↑" and "↓" indicate up-regulation and down-regulation, respectively.OAT, ornithine aminotransferase; P5CS, delta-1-pyrroline5-carboxylate synthase; P5CDH, delta-1-pyrroline-5-carboxylate dehydrogenase; P5CR, pyrroline-5-carboxylate reductase; ProDH, proline dehydrogenase.

Plant and Microbial Materials
Both soybean variety 'Jiyu47' and Arabidopsis thaliana Colombian (Col-0) WT were planted in a growth chamber with varied temperatures and photoperiod cycles for different experiments (below).Bacterial strains (i.e., competent cells of Agrobacterium rhizogenes K599, A. tumefaciens AGL0, and Escherichia coli DH5α), plasmid pCAMBIA3301-GFP, and yeast expression vector pYES2 were kept in stock at the Crop Germplasm Innovation Laboratory, Jilin Agricultural University.Competent cells of Saccharomyces cerevisiae INVSc1 were obtained from Coolaber (Beijing, China), with the cloning vector pMD18-T purchased from Takara Co., Ltd.(Beijing, China).

Cultivation and Treatment of Soybean Materials
To perform the transcriptome sequencing analysis, soybean seeds were sterilized in 2% sodium hypochlorite for 10 min, washed 5 times with sterile water, and sown in a mixture of nutrient soil and vermiculite (3:1; v/v).The seedlings were transferred to 1 L black plastic pots containing Hogland nutrient solution, which was replaced every 3 d.The 7-day-old soybean plants were selected and treated with 20% PEG6000 (Solarbio Science and Technology Co., Ltd., Beijing, China) to simulate drought stress for 24 h.Root tips of 3 cm in length were collected at 0, 6, 12, 18, and 24 h and stored in liquid nitrogen for subsequent transcriptome sequencing analysis.The raw data of transcriptome sequencing analysis were deposited in the National Center for Biotechnology Information database (https://www.ncbi.nlm.nih.gov/sra/;accessed on 30 October 2023; BioProject accession PRJNA1033409).
To evaluate the gene expression of GmNAC19, soybean seeds were sown at the experimental base of Jilin Agricultural University.Different organs (i.e., roots, stems, leaves, flowers, and pods) were collected at three developmental stages (seedling, flowering, and podding), quickly frozen in liquid nitrogen, and stored in a refrigerator at −80 • C for subsequent gene expression analysis of GmNAC19.
To establish the transgenic soybean composite lines with GmNAC19, soybean seeds were first sterilized, then sown with 1-2 cm intervals in the mixture of nutrient soil and vermiculite (1:1; v/v), and grown for 6-8 d in the dark.Then, the cotyledon hypocotyls were infected with Agrobacterium rhizogenes K599, followed by hydroponic culture with Hogland nutrient solution.At 15-20 d, the primary roots were removed to allow the recovery growth of hairy roots.Soybean plants infected with A. rhizogenes K599 without GmNAC19 were used as the control group and infected with A. rhizogenes K599 transformed with GmNAC19 as the transgenic group.The transgenic soybean composite lines were selected as previously reported [69,70], with slight modifications.The fluorescence exciter (LUYOR-3415RG Hand-Held Blue Green Lamp, LUYOR Instrument Co., Ltd.; Shanghai, China) was first used to verify the successful transformation of soybean hairy roots with soybean GmNAC19, i.e., the positive hairy roots were identified by fluorescent light.Then, the identification of positive hairy roots was further confirmed by PCR analysis, i.e., the genomic DNA was extracted and used to PCR-amplify both the 35S promoter and the bar gene, and then, the PCR products were evaluated using electrophoresis.The transformation efficiency in this experiment was 93%, i.e., in a total of 50 soybean sprouts infected by transgenic Agrobacterium rhizogenes K599 with GmNAC19, 45 survived with the development of hairy roots, with a total of 42 positive plants identified.Briefly, the soybean seeds were sown at the same time, and later, the infection by Agrobacterium rhizogenes K599 transformed with GmNAC19 simultaneously using the soybean sprouts of the similar developmental stages.Then, the infected plants were grown at the same time, with fast-growing or slow-growing plants removed and the plants of largely the same developmental stages chosen for further drought stress experiments.The soybean seedlings of the same developmental stages were grown in a nutrient solution containing 20% PEG6000 in a growth chamber (22 • C with relative humidity of 50% and a photoperiod cycle of 12 h dark and 12 h light with light intensity of 360 µmol m −2 s −1 ) and sampled at 0, 1, 3, 6, 12, and 24 h.Root tips of 3 cm in length were collected and quickly frozen in liquid nitrogen and stored in a refrigerator at −80 • C for later use.

Cultivation and Treatment of Arabidopsis thaliana
Seeds of A. thaliana were first vernalized 4 • C overnight, and then, sown in nutrient pots with the mixture of nutrient soil and vermiculite (1:1; v/v).The transgenic experiment of GmNAC19 using the expression vector pCAMBIA3301-GFP was completed via the flower-dipping method using Agrobacterium tumefaciens AGL0.The transgenic plants of the T3 generation of A. thaliana with GmNAC19 were used as the experimental group, while the control group contained the non-transgenic plants.The seeds of both WT and transgenic A. thaliana plants were evenly sown on the MS medium containing PEG6000 at 0%, 3%, 6%, and 9%, with the seed germination percentage calculated and the root length and diameter measured at 15 d.The root diameter was measured around the thickest place immediately below the junction of the stem and root under an inverted biological microscope (DM IL LED, Leica Microsystems Trading Co., Ltd., Shanghai, China).The phenotypic variations were observed in both the WT and transgenic A. thaliana plants with GmNAC19 treated with 6% PEG6000 for 15 d, and then, rehydrated for 5 d.The 100-seed weights of both the WT and transgenic A. thaliana plants with GmNAC19 were measured based on 100 seeds (n = 3) of five lines of both the WT and transgenic plants of the T3 generation using a one ten-thousandth electronic analytical balance (AS 220.X2, Suzhou Peike Laboratory Instrument Technology Co., Ltd., Suzhou, China).Then, the WT and one line (OE-6) of T3 generation transgenic plants of A. thaliana were grown on MS medium containing 0% and 6% PEG6000, respectively.Samples were collected at 15 d to determine the physiological and biochemical indices related to drought tolerance, including the contents of soluble protein, soluble sugar, proline, and MDA, and enzymatic activities of SOD, POD, and CAT.Plants at the seedling stage were selected and treated with drought stress induced by water loss for 15 d, with samples collected and stored at 0, 5, 10, and 15 d for the subsequent determination of physiological and biochemical indices.The seedlings were cultured in a growth chamber at 22 • C with a relative humidity of 60% and a photoperiod cycle of 8 h dark and 16 h light (light intensity 310 µmol m −2 s −1 ).

Fluorescence-Based Quantitative Real-Time PCR
Total RNA of plant samples was extracted using TIANGEN's RNAiso Plus reagent (Tiangen Company, Beijing, China).TransScript ® Uni All-in-one First-strand cDNA Synthesis SuperMix for qPCR kit (TransGen Biotech Co., Ltd., Beijing, China) was used for both reverse transcription and qRT-PCR experiments using GmEF1A as an internal reference gene.Each qPCR experiment was repeated with three biological replicates, with the relative gene expression calculated according to the method of 2 −∆∆Ct .The primers used in the qPCR experiments were synthesized by Shanghai Biotech Bioengineering Co., Ltd.(Shanghai, China; Table 4).

Validation of Drought Resistance in Saccharomyces cerevisiae
The recombinant plasmid pYES2-GmNAC19 was transformed into competent cells of Saccharomyces cerevisiae INVSc1 (Coolaber, Beijing, China) by following the manufacturer's instructions.The yeast solution was coated on SD/-Umedium single-deficient medium and cultured in the inverted mode at 29 • C for 2-3 d.Single clones were picked for yeast solution and plasmid PCR detection.Ex-Taq, plasmid extraction, and gel recovery kits were purchased from Takara Co., Ltd.(Dalian, China).Then, the yeast solution was further cultured until the OD 600 reached 0.5; after the gradient dilution, the yeast solution was sampled and cultured on YPD solid medium containing mannitol of two different concentrations (0.25 and 0.5 M) in the inverted mode in an incubator at 29 • C. DNA sequencing was performed by Shanghai Biotech Bioengineering Co., Ltd.(Shanghai, China).

Determination of Physiological and Biochemical Indicators Related to Drought Tolerance
The content of chlorophyll was measured using a chlorophyll meter SPAD-502 Plus (Konica Minolta, Tokyo, Japan).The content of soluble protein was determined using the Kaumas Brilliant Blue G-250 method.The content of soluble sugar was detected by the anthrone method.Bovine serum albumin (BSA) and glucose of analytical grade were purchased from Solarbio Science and Technology Co., Ltd.(Beijing, China).The content of MDA was determined by the thiobarbituric acid method.The content of proline was measured using the sulfosalicylic acid method.The enzymatic activities of SOD, POD, and CAT were evaluated using the nitrogen blue tetrazolium method, the guaiacol method, and the H 2 O 2 UV-absorption method [37,71,72], respectively.

Statistical Analysis
Each experiment was performed with three or more biological replicates.SPSS Statistics 26 was used to analyze the data, which were presented as mean ± standard deviation.Statistical significance was determined by Student's t test and either one-way or two-way analysis of variance (ANOVA) followed by a post hoc Tukey's test.GraphPad Prism 9 was used to generate graphs to present the results of the statistical analyses.Pearson correlation analysis was performed to investigate the relationships between the gene expression of GmNAC19 and the expressions of 4 key genes (i.e., P5CS, OAT, P5CR, and ProDH) involved in proline metabolism in the control and transgenic soybean hairy roots with GmNAC19.

Conclusions
Based on a comprehensive analysis of the gene expression patterns of GmNAC19 in transgenic soybean composite lines and Arabidopsis thaliana under drought stress, our investigations explored the molecular response of the soybean TF gene GmNAC19 to drought stress either simulated by PEG6000 or caused by water loss.The germination rate of transgenic A. thaliana seeds was improved under drought conditions with the promotion of rooting and improved recovery ability after rehydration.The variations in a group of physiological and biochemical indices involved in drought tolerance suggested that the beneficial effects of GmNAC19 on drought tolerance in both soybean and Arabidopsis were achieved by regulating the contents of soluble protein, proline, and soluble sugar, as well as the activities of antioxidant enzymes.However, it is realized that although some of the experiments showed evident effects of this gene on the drought-related biochemical and physiological indices, a stabilized pattern of these effects was not observed due to the lack of an extended period of treatment time.It is also noted that our study of this drought-resistance gene in soybean was performed under a drought condition generated by treatment with PEG6000, and further investigations under actual drought (i.e., reduced water supply over time) are necessary to verify the findings of the drought-resistance effects of this gene in soybean, e.g., the evaluation of the yield and quality of beans as well as soybean plant growth parameters.Our study provided a candidate gene for the molecular breeding and development of drought-tolerant crop plants.

2. 1 . 1 .
Gene Expression of GmNAC19 Based on Transcriptome Analysis of Drought-Treated Soybean Roots

Figure 1 .
Figure 1.Relative gene expression levels of GmNAC19 based on transcriptome sequencing analysis of root tips (3 cm in length) of 7-day-old soybean roots treated with 20% PEG6000 for 0, 6, 12, 18, and 24 h.FPKM, fragments per kilobase of transcript per million mapped reads.The statistical significance between two groups was determined by one-way ANOVA, with significant differences in the relative expression level of GmNAC19 indicated by different lowercase letters (a, b, or c) based on p < 0.05 (n = 3).The same letters (a, b, or c) indicate no significant difference.2.1.2.Gene Expression of GmNAC19 in Soybean Treated with PEG6000 Based on Quantitative Real-Time PCR

Figure 2 .
Figure 2. Relative gene expression levels of GmNAC19 in soybean roots treated with 20% PEG6000 based on qRT-PCR analysis using GmEF1A as the reference gene.The statistical significance between two groups was determined by one-way ANOVA, with the significant differences in the relative expression level of GmNAC19 represented by different lowercase letters (a, b, c, or d) based on p < 0.05 (n = 3).The same letters (a, b, c, or d) indicate no significant difference.2.1.3.Gene Expression of GmNAC19 in Different Organs of Soybean at Three Different Developmental Stages

Figure 3 .
Figure 3. Relative gene expression levels of GmNAC19 in different organs (i.e., stems, roots, leaves, flowers, and pods) of soybean plants at three developmental stages, i.e., seedling (A), flowering (B), and podding (C) based on qRT-PCR analysis using GmEF1A as the reference gene.The statistical significance between two groups was determined by one-way ANOVA, with the significant

Figure 2 .
Figure 2. Relative gene expression levels of GmNAC19 in soybean roots treated with 20% PEG6000 based on qRT-PCR analysis using GmEF1A as the reference gene.The statistical significance between two groups was determined by one-way ANOVA, with the significant differences in the relative expression level of GmNAC19 represented by different lowercase letters (a, b, c, or d) based on p < 0.05 (n = 3).The same letters (a, b, c, or d) indicate no significant difference.

Figure 2 .
Figure 2. Relative gene expression levels of GmNAC19 in soybean roots treated with 20% PEG6000 based on qRT-PCR analysis using GmEF1A as the reference gene.The statistical significance between two groups was determined by one-way ANOVA, with the significant differences in the relative expression level of GmNAC19 represented by different lowercase letters (a, b, c, or d) based on p < 0.05 (n = 3).The same letters (a, b, c, or d) indicate no significant difference.2.1.3.Gene Expression of GmNAC19 in Different Organs of Soybean at Three Different Developmental Stages

Figure 3 .
Figure 3. Relative gene expression levels of GmNAC19 in different organs (i.e., stems, roots, leaves, flowers, and pods) of soybean plants at three developmental stages, i.e., seedling (A), flowering (B), and podding (C) based on qRT-PCR analysis using GmEF1A as the reference gene.The statistical significance between two groups was determined by one-way ANOVA, with the significant

Figure 3 .
Figure 3. Relative gene expression levels of GmNAC19 in different organs (i.e., stems, roots, leaves, flowers, and pods) of soybean plants at three developmental stages, i.e., seedling (A), flowering (B), and podding (C) based on qRT-PCR analysis using GmEF1A as the reference gene.The statistical significance between two groups was determined by one-way ANOVA, with the significant differences in the relative expression level of GmNAC19 represented by different lowercase letters (a, b, c, or d) based on p < 0.05.The same letters (a, b, c, or d) indicate no significant difference.

Figure 5 .
Figure 5. Drought tolerance in control (pYES2) and transgenic (pYES2-GmNAC19) Saccharomyces cerevisiae with soybean GmNAC19 under the treatment of mannitol with two concentrations (0.25 M and 0.5 M) at 48 h.(A) Morphological observations of yeast growth based on various yeast inoculation dilutions.Scale bar = 1 cm.(B) Number of yeast cells in the control (pYES2) and transgenic (pYES2-GmNAC19) groups based on yeast inoculation dilution of 10⁻ 4 CFU/mL.(C) Number of yeast cells in the control (pYES2) and transgenic (pYES2-GmNAC19) groups based on yeast inoculation dilution of 10⁻5 CFU/mL.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the number of yeast cells between the control without treatment of mannitol and controls with treatment of mannitol, and between transgenic group without treatment of mannitol and transgenic groups with treatment of mannitol; two-way ANOVA tests were performed to detect significant differences in the number of yeast cells between the control and transgenic groups cultured on the same type of medium (indicated by asterisk above the bracket) based on p < 0.05 (*) or p < 0.01 (**) (n = 3).

Figure 5 .
Figure 5. Drought tolerance in control (pYES2) and transgenic (pYES2-GmNAC19) Saccharomyces cerevisiae with soybean GmNAC19 under the treatment of mannitol with two concentrations (0.25 M and 0.5 M) at 48 h.(A) Morphological observations of yeast growth based on various yeast inoculation dilutions.Scale bar = 1 cm.(B) Number of yeast cells in the control (pYES2) and transgenic (pYES2-GmNAC19) groups based on yeast inoculation dilution of 10 −4 CFU/mL.(C) Number of yeast cells in the control (pYES2) and transgenic (pYES2-GmNAC19) groups based on yeast inoculation dilution of 10 −5 CFU/mL.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the number of yeast cells between the control without treatment of mannitol and controls with treatment of mannitol, and between transgenic group without treatment of mannitol and transgenic groups with treatment of mannitol; two-way ANOVA tests were performed to detect significant differences in the number of yeast cells between the control and transgenic groups cultured on the same type of medium (indicated by asterisk above the bracket) based on p < 0.05 (*) or p < 0.01 (**) (n = 3).

30 Figure 6 .
Figure 6.Seed germination of wild-type (WT) and transgenic Arabidopsis thaliana with GmNAC19 (OE-6) treated with PEG6000 at 0%, 3%, 6%, and 9% for 15 d.Each treatment group contains 3 replicates (64 seeds in each replicate).One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in seed germination percentage between WT without treatment of PEG6000 and WT with treatment of PEG6000, and between transgenic group without treatment of PEG6000 and transgenic groups with treatment of PEG6000; two-way ANOVA tests were performed to detect the significant differences in seed germination percentage between WT and transgenic groups in the same treatment group (indicated by asterisk above the bracket) based on p < 0.01 (**) and p < 0.001 (***) (n = 3).

Figure 6 .
Figure 6.Seed germination of wild-type (WT) and transgenic Arabidopsis thaliana with GmNAC19 (OE-6) treated with PEG6000 at 0%, 3%, 6%, and 9% for 15 d.Each treatment group contains 3 replicates (64 seeds in each replicate).One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in seed germination percentage between WT without treatment of PEG6000 and WT with treatment of PEG6000, and between transgenic group without treatment of PEG6000 and transgenic groups with treatment of PEG6000; two-way ANOVA tests were performed to detect the significant differences in seed germination percentage between WT and transgenic groups in the same treatment group (indicated by asterisk above the bracket) based on p < 0.01 (**) and p < 0.001 (***) (n = 3).

Figure 7 .
Figure 7. Growth of roots of wild-type (WT) and transgenic Arabidopsis thaliana of GmNAC19 (OE-6) treated with PEG6000 at 0%, 3%, 6%, and 9%, for 15 d.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in root length and root diameter between WT without treatment of PEG6000 and WT with treatment of PEG6000, and between transgenic group without treatment of PEG6000 and transgenic groups with treatment of PEG6000; twoway ANOVA tests were performed to detect the significant differences in root length and root diameter between WT and transgenic groups in the same treatment group (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3 with each replicate of 5 plants).

Figure 7 .
Figure 7. Growth of roots of wild-type (WT) and transgenic Arabidopsis thaliana of GmNAC19 (OE-6)treated with PEG6000 at 0%, 3%, 6%, and 9%, for 15 d.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in root length and root diameter between WT without treatment of PEG6000 and WT with treatment of PEG6000, and between transgenic group without treatment of PEG6000 and transgenic groups with treatment of PEG6000; two-way ANOVA tests were performed to detect the significant differences in root length and root diameter between WT and transgenic groups in the same treatment group (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3 with each replicate of 5 plants).

( 3 )
Int. J. Mol.Sci.2024, 25, x FOR PEER REVIEW Contents of physiological and biochemical indices related to drought stress in genic Arabidopsis thaliana with GmNAC19 under drought stress simulate PEG6000

Figure 8 .
Figure 8. Contents of physiological and biochemical indices, i.e., soluble protein (A), soluble (B), proline (C), and malondialdehyde (MDA) (D), in wild-type (WT) and transgenic Arab thaliana with GmNAC19 (OE-6) under drought stress simulated by 6% PEG6000 for 15 d.On ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differen the content of each index between WT without treatment of PEG6000 and WT with treatm PEG6000, and between transgenic group without treatment of PEG6000 and transgenic grou treatment of PEG6000; two-way ANOVA tests were performed to detect the significant diffe in the content of each index between WT and transgenic groups in the same treatment group on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).

Figure 8 .
Figure 8. Contents of physiological and biochemical indices, i.e., soluble protein (A), soluble sugar (B), proline (C), and malondialdehyde (MDA) (D), in wild-type (WT) and transgenic Arabidopsis thaliana with GmNAC19 (OE-6) under drought stress simulated by 6% PEG6000 for 15 d.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the content of each index between WT without treatment of PEG6000 and WT with treatment of PEG6000, and between transgenic group without treatment of PEG6000 and transgenic group with treatment of PEG6000; two-way ANOVA tests were performed to detect the significant differences in the content of each index between WT and transgenic groups in the same treatment group based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).

Figure 9 .
Figure 9. Antioxidant enzyme activities of superoxide dismutase (SOD) (A), peroxidase (POD) (B), and catalase (CAT) (C), in wild-type (WT) and transgenic Arabidopsis thaliana of GmNAC19 (OE-6) under drought stress simulated by 6% PEG6000 for 15 d.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the enzymatic activities between WT without treatment of PEG6000 and WT with treatment of PEG6000, and between transgenic group without treatment of PEG6000 and transgenic group with treatment of PEG6000; twoway ANOVA tests were performed to detect the significant differences in the enzymatic activities between WT and transgenic groups in the same treatment group (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).Note: A significant increase is detected in the enzymatic activities of SOD, POD, and CAT in transgenic A. thaliana on MS medium containing PEG6000 compared with WT on MS without PEG6000 based on Student's t test (p < 0.05).

Figure 9 .
Figure 9. Antioxidant enzyme activities of superoxide dismutase (SOD) (A), peroxidase (POD) (B), and catalase (CAT) (C), in wild-type (WT) and transgenic Arabidopsis thaliana of GmNAC19 (OE-6) under drought stress simulated by 6% PEG6000 for 15 d.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the enzymatic activities between WT without treatment of PEG6000 and WT with treatment of PEG6000, and between transgenic group without treatment of PEG6000 and transgenic group with treatment of PEG6000; two-way ANOVA tests were performed to detect the significant differences in the enzymatic activities between WT and transgenic groups in the same treatment group (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).Note: A significant increase is detected in the enzymatic activities of SOD, POD, and CAT in transgenic A. thaliana on MS medium containing PEG6000 compared with WT on MS without PEG6000 based on Student's t test (p < 0.05).

Figure 10 .
Figure 10.Contents of physiological and biochemical indices, i.e., soluble protein (A), soluble sugar (B), proline (C), and malondialdehyde (MDA) (D), related to drought stress in wild-type (WT) and transgenic Arabidopsis thaliana with GmNAC19 (OE-6) under drought stress induced by water loss.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the content of each index between WT at 0 d and WT at 5, 10, and 15 d, and between transgenic group at 0 d and transgenic groups at 5, 10, and 15 d; two-way ANOVA tests were performed to detect the significant differences in the content of each index between WT and transgenic groups at the same sampling time point (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).Note: In 15 d, a significant increase in the content of soluble sugar (B) is detected in transgenic A. thaliana under drought stress compared with WT without drought stress (0 d) based on Student's t test (p < 0.05).

Figure 10 .
Figure 10.Contents of physiological and biochemical indices, i.e., soluble protein (A), soluble sugar (B), proline (C), and malondialdehyde (MDA) (D), related to drought stress in wild-type (WT) and transgenic Arabidopsis thaliana with GmNAC19 (OE-6) under drought stress induced by water loss.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the content of each index between WT at 0 d and WT at 5, 10, and 15 d, and between transgenic group at 0 d and transgenic groups at 5, 10, and 15 d; two-way ANOVA tests were performed to detect the significant differences in the content of each index between WT and transgenic groups at the same sampling time point (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).Note: In 15 d, a significant increase in the content of soluble sugar (B) is detected in transgenic A. thaliana under drought stress compared with WT without drought stress (0 d) based on Student's t test (p < 0.05).

Figure 11 .
Figure 11.Antioxidant enzyme activities of superoxide dismutase (SOD) (A), peroxidase (POD) (B), and catalase (CAT) (C) in wild-type (WT) and transgenic Arabidopsis thaliana with GmNAC19 (OE-6) under drought stress induced by water loss.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the enzymatic activity between WT at 0 d and WT at 5, 10, and 15 d, and between transgenic group at 0 d and transgenic groups at 5, 10, and 15 d; two-way ANOVA tests were performed to detect the significant differences in the enzymatic activity between WT and transgenic groups at the same sampling time point (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).Note: In 15 d, the enzymatic activities of SOD, POD, and CAT are significantly increased in the transgenic groups compared with the WT without drought stress (0 d) based on Student's t test (p < 0.05).

Figure 11 .
Figure 11.Antioxidant enzyme activities of superoxide dismutase (SOD) (A), peroxidase (POD) (B), and catalase (CAT) (C) in wild-type (WT) and transgenic Arabidopsis thaliana with GmNAC19 (OE-6) under drought stress induced by water loss.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the enzymatic activity between WT at 0 d and WT at 5, 10, and 15 d, and between transgenic group at 0 d and transgenic groups at 5, 10, and 15 d; two-way ANOVA tests were performed to detect the significant differences in the enzymatic activity between WT and transgenic groups at the same sampling time point (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).Note: In 15 d, the enzymatic activities of SOD, POD, and CAT are significantly increased in the transgenic groups compared with the WT without drought stress (0 d) based on Student's t test (p < 0.05).

Figure 13 .
Figure 13.Contents of physiological and biochemical indicators, i.e., soluble protein (A-C), soluble sugar (D-F), proline (G-I), and malondialdehyde (MDA) (J-L) related to drought stress in hairy roots, stems, and leaves of the control and the transgenic soybean composite lines with GmNAC19 (OE) under drought stress simulated by 20% PEG6000.The control groups contain plants

Figure 13 .
Figure 13.Contents of physiological and biochemical indicators, i.e., soluble protein (A-C), soluble sugar (D-F), proline (G-I), and malondialdehyde (MDA) (J-L) related to drought stress in hairy roots, stems, and leaves of the control and the transgenic soybean composite lines with GmNAC19 (OE) under drought stress simulated by 20% PEG6000.The control groups contain plants transformed withAgrobacterium rhizogenes K599 without GmNAC19.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the content of each index between control at 0 h and controls at 3, 6, 9, and 12 h, and between transgenic group at 0 h and transgenic groups at 3, 6, 9, and 12 h; two-way ANOVA tests were performed to detect the significant differences in the content of each index between control and transgenic groups at the same sampling time point (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).Note: At 12 h, a significant difference is detected between the transgenic plant and the control without drought stress in the content of soluble protein in the leaves, in the content of soluble sugar in all three organs, in the proline content in the stems and leaves, and in the content of MDA in both the roots and leaves based on Student's t test (p < 0.05).

Figure 14 .
Figure 14.Antioxidant enzyme activities of superoxide dismutase (SOD) (A-C), peroxidase (POD) (D-F), and catalase (CAT) (G-I) in the hairy roots, stems, and leaves of transgenic soybean composite lines with GmNAC19 (OE) under drought stress simulated by 20% PEG6000.The control groups contain the plants transformed with Agrobacterium rhizogenes K599 without GmNAC19.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the enzymatic activity between control at 0 h and controls at 3, 6, 9, and 12 h, and between transgenic group at 0 h and transgenic groups at 3, 6, 9, and 12 h; two-way ANOVA tests were performed to detect the significant differences in the enzymatic activity between control and transgenic groups at the same sampling time point (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).Note: At 12 h, the enzymatic activities of SOD, POD, and CAT in all three organs of transgenic plants are significantly higher than those of the controls without drought stress (0 d) based on Student's t test (p < 0.05).

Figure 14 .
Figure 14.Antioxidant enzyme activities of superoxide dismutase (SOD) (A-C), peroxidase (POD) (D-F), and catalase (CAT) (G-I) in the hairy roots, stems, and leaves of transgenic soybean composite lines with GmNAC19 (OE) under drought stress simulated by 20% PEG6000.The control groups contain the plants transformed with Agrobacterium rhizogenes K599 without GmNAC19.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the enzymatic activity between control at 0 h and controls at 3, 6, 9, and 12 h, and between transgenic group at 0 h and transgenic groups at 3, 6, 9, and 12 h; two-way ANOVA tests were performed to detect the significant differences in the enzymatic activity between control and transgenic groups at the same sampling time point (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).Note: At 12 h, the enzymatic activities of SOD, POD, and CAT in all three organs of transgenic plants are significantly higher than those of the controls without drought stress (0 d) based on Student's t test (p < 0.05).

Figure 16 .
Figure16.Relative expression levels of genes, i.e., P5CS (A), OAT (B), P5CR (C), and ProDH (D), involved in proline metabolic pathway in soybean hairy roots with GmNAC19 (OE) under drought stress simulated by 20% PEG6000.The control groups contain soybean hairy roots transformed with Agrobacterium rhizogenes K599 without GmNAC19.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the relative expression of each gene between control at 0 h and controls at 3, 6, 9, and 12 h, and between transgenic group at 0 h and transgenic groups at 3, 6, 9, and 12 h; two-way ANOVA tests were performed to detect the significant differences in the relative expression of each gene between control and transgenic groups at the same sampling time point (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).

Figure 16 .
Figure16.Relative expression levels of genes, i.e., P5CS (A), OAT (B), P5CR (C), and ProDH (D), involved in proline metabolic pathway in soybean hairy roots with GmNAC19 (OE) under drought stress simulated by 20% PEG6000.The control groups contain soybean hairy roots transformed with Agrobacterium rhizogenes K599 without GmNAC19.One-way ANOVA followed by post hoc Tukey's test was performed to evaluate the significant differences in the relative expression of each gene between control at 0 h and controls at 3, 6, 9, and 12 h, and between transgenic group at 0 h and transgenic groups at 3, 6, 9, and 12 h; two-way ANOVA tests were performed to detect the significant differences in the relative expression of each gene between control and transgenic groups at the same sampling time point (indicated by asterisk above the bracket) based on p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) (n = 3).

Table 4 .
Primers and their sequences used in qPCR."-F" and "-R" indicate forward and reverse primers, respectively.

Table 1 .
The 100-seed weight (mg) in both wild-type (WT) and transgenic Arabidopsis thaliana with GmNAC19.Data are presented as mean ± standard deviation.The statistical difference between the WT and transgenic lines was determined by Student's t test based on p < 0.05 (*) and p < 0.01 (**).