Poly-γ-glutamic acid enhanced drought resistance of maize by improving photosynthesis and affecting rhizosphere microbial community

Abstract

Poly-γ-glutamic acid (γ-PGA) is a non-toxic, water-soluble, biodegradable and environment-friendly biopolymer, which is composed of D/L-glutamic acid monomers and fermented by Bacillus subtilis [22,23]. In line with its different molecular weights, γ-PGA could be used in many elds, such as food, medicine, cosmetics and agriculture [24]. γ-PGA has been paid more and more attention as an environmentally friendly fertilizer synergist because of its strong water solubility and retention, biodegradability and innocuity [25]. Recent studies have found that γ-PGA plays an important role in plant growth and regulation, and can be used as a water retaining agent and soil conditioner to improve crop productivity [26][27][28]. It has been reported that exogenous application of γ-PGA could signi cantly enhance the stress resistance of plants [26,[29][30][31]. Most of the previous studies focused on cold and salt stress of vegetables such as Brassica napus and cucumber. For example, it was found that γ-PGA could increase salt and cold tolerance of Brassica napus by activating the crosstalk between H 2 O 2 and Ca 2+ signals [32] and enhance drought resistance of Brassica napus by promoting ABA accumulation. However, only few studies assessed the effect of γ-PGA on drought resistance of plants, especially crops. And the regulation mechanism of γ-PGA on drought resistance of maize remains unclear. Maize is an important crop integrating grain, feed, energy and industrial raw materials, and plays an extremely important role in world food security and economic development [33]. Its yield is always severely affected by drought stress [34]. In this study, the effect of γ-PGA on the growth of maize seedlings under drought stress was assessed by adding γ-PGA to soil. In addition, RNAseq was performed to study the gene expression of maize leaves after drought stress, and the changes of rhizosphere microbial community after exogenous application of γ-PGA were also studied, so as to understand the mechanism of exogenous application of γ-PGA to change drought resistance of maize. So as to understand the mechanism of exogenous application of PGA to change drought resistance of maize.

Results
Exogenous application of γ-PGA enhanced drought resistance of maize In order to investigate the effect of exogenous application of γ-PGA on maize under drought stress, the drought-resistant phenotype of maize with different concentrations of γ-PGA (0, 50, 70, 100mg/L) were examined (Additional le 1: Fig. S1). The results showed that the addition of γ-PGA could signi cantly enhance the drought resistance of maize, even at a lower concentration (50mg/L), and could regenerate maize rapidly after rewatering, while most of the control maize showed severe wilting and could not grow again after rewatering. 50mg/L γ-PGA treatment was used for the subsequent experiments.
Maize treated with 50mg/L γ-PGA exhibited a better phenotype after 7 days of drought stress (Fig. 1A).
The dry weight, content of ABA, soluble sugar, proline, chlorophyll and the photosynthetic parameters of maize seedlings after 5 days drought treatment were determined. As shown in Fig. 1B, under drought condition, the dry weight of maize treated with γ-PGA (0.96g) was signi cantly higher than that of control maize (0.39g), indicating that γ-PGA could alleviate the inhibition of drought stress on the growth of maize seedlings. In addition, compared with the control group, the contents of ABA, soluble sugar, proline and chlorophyll in γ-PGA treatment group were 27.46%, 43.61%, 108% and 51.51% higher, respectively (Fig. 1B). This indicated that γ-PGA could promote the accumulation of ABA, soluble sugar, proline and the chlorophyll in maize under drought stress. The photosynthetic parameters of the maize under drought for 5d were also measured, the results showed that both the net photosynthetic rate and stomatal conductance of the maize added γ-PGA were signi cantly higher than the control maize under drought stress (Fig. 1B).
In order to observe the effect of γ-PGA on maize growth under drought stress more directly, the simulated drought experiment with 18% PEG6000 solution was performed. It was found that the fresh weight of leaves and roots in γ-PGA treatment group was higher than that of the control group (Fig. 2), indicating that the drought resistance of leaves and roots in γ-PGA + PEG group was signi cantly higher than that of the control group.
γ-PGA signi cantly improved roots development, urease activity and ABA contents of maize rhizospheric soil under drought stress It was found that γ-PGA signi cantly improved the roots development both under the normal condition and drought stress ( Fig. 2A). Under normal growing conditions, the maize treated with γ-PGA had a better developed root system, and the fresh weight of roots was signi cantly increased than that of the control group. Under PEG simulate drought stress, the roots growth of the control group was signi cantly inhibited, however, the roots of the maize treated with γ-PGA were little affected by drought stressand the roots fresh weight was signi cantly higher than that of control group. Since maize rhizospheric soil was closely contacted with the roots, the urease activity (closely related to soil nitrogen transformation) and ABA contents (closely related to the drought resistance) of the maize rhizospheric soil under the severe drought stress were also detected. It was observed that the urease activity of rhizospheric soil of γ-PGA treatments was increased by 27.74%, while the ABA contents of γ-PGA treatments soil was also increased by 21.70% (Table 1). Values are means ± sd (n ≥ 3 repeats). Signi cant differences are indicated by asterisks (**, P ≤ 0.01).
Differentially expressed genes (DEGs) between maize with γ-PGA addition and control under drought stress In order to explain the mechanism of γ-PGA in improving the drought resistance of maize, the leaves of γ-PGA treatment and the control maize under drought condition for 5 days were used for RNA sequencing to identify the DEGs and pathways in response to drought stress. The total raw reads, clean reads, genome mapping ratio, and uniquely mapping ratio were listed in Additional le 10: Table S1. 16126 DEGs were identi ed and the distribution of the DEGs was illustrated in Fig. 3A (Fig. 3B). The results of GO annotation functions enrichment analysis also showed that GO terms such as photosynthesis and photosystem, response to abiotic stimulus, chlorophyll metabolic process, response to biotic stimulus, electron transport chain and so on were signi cantly enriched (Additional le 2: Fig. S2B). A more detailed classi cation of the terms of response to abiotic stimulus showed that these DEGs were mainly related to the response to stress (osmotic stress, salt, heat, cold, reactive oxygen species, and hydrogen peroxide), the response to hormone (ABA, JA, and SA), ABA biosynthetic process, chlorophyll metabolic process, proline biosynthetic process, protein folding, and so on (Additional le 2: Fig. S2B).
γ-PGA improved drought resistance of maize by affecting the expression of photosynthesis-related genes As known, drought could signi cantly reduce the photosynthesis capability of plants. However, KEGG analysis showed that under drought stress, compared with the control maize, the photosynthesis related genes of maize treated with γ-PGA were signi cantly enriched (Fig. 3B), with most of related genes were dramatically upregulated. As shown in Fig. 4 and Additional le 10: Table S1, most genes in DEGs of photosystem II complex were upregulated, except PsbA, PsbB, PsbC, PsbE, PsbF and 1 for PsbP, which were downregulated. In photosystem I complex, all of the DEGs were upregulated. In cytochrome b6/f complex, 7 genes encoding PetA, 2 genes encoding PetC and 1 genes encoding PetG were upregulated, while only 1 gene encoding PetD and 1 gene encoding PetA were downregulated. In photosynthetic electron transport, other 16 genes encoding PetE, PetF, PetH and PetJ were all up-regulated except 3 genes encoding PetF and 2 for PetH,. In F-type ATPase complex, except 1 gene encoding beta, 1 for gamma and 1 b which were downregulated, the other 14 genes encoding alpha, beta, gamma, delta, epsilon, a, b and c subunits respectively were upregulated. Additionally, all DEGs (67 genes) encoding antenna proteins were also up-regulated (Additional le 3: Fig. S3). To con rm the results, 14 genes with different transcript abundances were validated by real-time RT-PCR (Additional le 4: Fig. S4). The expression of these genes showed good consistency between the two detection methods. Meanwhile, the motifs in the promoter region of these genes were analyzed, higher percentage of drought, lowtemperature, salicylic stress and ABA response elements were found (Additional le 5: Fig. S5, Additional le 6: Fig. S6).
γ-PGA affected the bacterial community diversity and structure of rhizospheric soil In order to study the in uence of γ-PGA on bacterial community diversity under drought stress, the relative abundance and diversity of maize rhizospheric soil bacteria were analyzed by high-throughput sequencing of 16S rRNA. The species curve showed that the samples were representative enough to obtain a true bacterial community (Additional le 8: Fig. S8). NMDS (stress = 0.00422) of the weighted UniFrac distance ordinations were conducted (Fig. 5A), the results indicated that the bacterial community composition of the soil with γ-PGA application brought shifts compared with that of the soil without γ-PGA under the drought stress, the communities in maize rhizospheric soil with γ-PGA were grouped together and signi cantly separated from those in soil without γ-PGA under the drought stress. The obtained high-quality sequences were belonged to 36 phylum,among which the main phylum was Proteobacteria, followed by Actinobacteria, Chloro exi, Bacteroidetes, and Acidobacteria. Although the diversity of bacterial community changed after the addition of γ-PGA under drought stress the predominant phylum were similar. There was no difference in species composition among these samples, but the relative abundances of some species changed (Fig. 5B). Compared to the control, the relative abundance of Actinobacteria and Chloro exi were higher in soil added γ-PGA under drought stress. LEfSe analysis (LDA ≥ 3) showed the species with the most signi cant variation (Fig. 5C). Under drought stress, the application of γ-PGA could signi cantly enrich Actinobacteria, Chloro exi and Cyanobacteria at phylum level, while Alphaproteobacteria and Deltaproteobacteria were enriched at class level. At the genus level, bacteria such as Rhodobacter Sphingobium, Sphingomonas, Sphingopyxis, Haliangium, Methylibium, Lysobacter, Azoarcus and Arenimonas of Proteobacteria, Aeromicrobium, Lechevalieria and Streptomyces of Actinobacteria, Subgroup_10 of Acidobacteria, Clostridium and Pelotomaculum of Firmicutes, Chloronema, A4b and KD4-96 of Chloro exi were dominant in γ-PGA added rhizosphere soil under the persistent severe drought condition. The abundances of these genera in maize rhizospheric soil with γ-PGA addition were all higher than that of control (Additional le 9: Fig. S9), while Bacillus of Proteobacteria was dominated in control (Fig. 5C).

Discussion
Among all abiotic stresses, drought has the greatest impact on soil organisms and plants [40]. Drought could adversely affect the important physiological and biochemical processes of plants,resulting in serious loss of crop yield worldwide [41]. It is critical to improve the plant tolerance to drought stress. As a natural and environment friendly biopolymer, γ-PGA has been widely used in agricultural production [42].
However, there are few reports about the effect and mechanism of γ-PGA on drought resistance of plants, especially crops. In this study, the effect of γ-PGA on maize drought resistance and its comprehensive mechanism by RNAseq and rhizosphere soil bacterial community diversity analysis were rstly reported.
The effects of exogenous application of γ-PGA on dry weight, the contents of ABA, soluble sugar, proline and chlorophyll of maize leaf under severe drought stress were characterized. These physiological indexes have been often used to evaluate the drought resistance of plants. As the osmoprotectants, proline and soluble sugar could provide osmotic adjustments in plants under drought stress [43]. Proline has strong hydration ability, which can protect cell structure and enzymes, reduce cell acidity and regulate redox potential under stress. ABA is considered to be the most critical hormone regulating tolerance to drought stress. Drought stress could trigger a huge increase in ABA biosynthesis. As a key chemical messenger of drought signal, ABA could activate a series of signal transduction reactions to regulate stomatal closure, calcium signal and the expression of some ABA-responsive genes to resist the drought stress. Drought stress can signi cantly decline the chlorophyll content of leaves [44,45]. Plants with higher chlorophyll contents under drought stress could use light energy more e ciently and have better drought resistance. In this study, we found that under drought stress, γ-PGA could promote the accumulation of ABA, soluble sugar, proline and chlorophyll, the drought resistance of maize was signi cantly enhanced by adding γ-PGA. In addition, γ-PGA could increase the dry weight of maize under drought stress, indicating that maize added with γ-PGA could still maintain a certain growth compared with that of control. In order to observe the root morphology under drought stress more directly, PEG6000 was used to simulate the drought treatment in the solution culture process. The results showed that, under PEG treatment, maize added with γ-PGA had more developed roots than that of control, which could make maize absorb deeper and more water of soil during drought stress.
To explore the molecular mechanism of enhanced drought resistance by exogenous application of γ-PGA, the differentially expressed genes (DEGs) of the leaves were evaluated by RNAseq analysis. KEGG analysis showed that photosynthesis related genes were signi cantly enriched which was consistent with the increase of photosynthetic rate in the maize treatment with γ-PGA under the drought stress. Most of the photosynthesis related genes, including 20 genes in photosystem I, 28 genes involved in photosystem II, 16 genes in photosynthetic electron transport, 10 genes in cytochrome b6/f complex, 14 genes in ATPase complex, and 31 genes encoding antenna proteins (9 genes encoding LHCI complex, 22 genes encoding LHCII complex), were dramatically upregulated in γ-PGA treatment maize compared with that of control. Photosynthesis is one of the main processes affected by drought [46]. However, under severe drought stress, the photosynthesis related genes in maize added with γ-PGA still maintained a high expression level than that of control, which may be the main reason for the higher drought resistance of maize treated with γ-PGA, while the reduced chlorophyll contents under drought in control leaded to the inactivation of photosynthesis.
ABA is considered to be the most critical hormone involved in the adaptive responses of plants to drought stress. DEGs related with carotenoid biosynthesis pathway which contains ABA biosynthesis pathway were also found to be signi cantly enriched in this study. In ABA biosynthesis, β-carotene is converted to zeaxanthin by CHY2 enzyme rstly, the epoxidation of zeaxanthin and antheraxanthin to violaxanthin was catalyzed by zeaxanthin epoxidase (ZEP/ABA1) afterwards [35]. Violaxanthin is converted to 9-cisviolaxanthin after a series of structural modi cations. The next step is also a rate-limiting step, that is, 9cis-violaxanthin is converted to xanthoxin under the catalysis of 9-cis-epoxycarotenoid dioxygenase (NCED) [36]. Subsequently, xanthoxin is converted to abscisic aldehyde, and then ABA is produced by two-step reaction via ABA-aldehyde. The enzyme (alcohol dehydrogenase/reductase) encoded by ABA2 catalyzes the rst step of this reaction and generates ABA aldehyde [37], and abscisic aldehyde oxidase encoded by AAO3 catalyzes the last step of ABA synthesis [38]. In this study, it was found that the ABA biosynthesis related genes including 2 genes encoding CHY2, 7 genes encoding ABA1, 3 genes encoding NCED, 2 genes encoding ABA2, and 1 genes encoding AAO3 were signi cantly upregulated, while 2 genes encoding 8'-hydroxyase which played important role in the catabolism of ABA were downregulated in maize with the application of γ-PGA. The expression level of these DEGs led to the increase of ABA level in γ-PGA treated maize under drought stress. The results indicated that γ-PGA could promote ABA accumulation under drought condition, and the accumulation of ABA can activate the core ABA signaling pathway including PYR/PYL/RCAR receptor, PP2C proteins, SnRK2 family members, AREB/ABF transcription factors and downstream regulatory genes, as well as ABA-activated signaling pathway to resist drought stress [47]. In addition, many reports have shown that among the promoters of the stressresponsive genes, there was a major cis-acting element (ABRE) which was regarded to be necessary for ABA response [48]. We found that ABRE element were present in the promoters of these upregulated photosynthesis related genes, suggesting that these genes may also be regulated by ABA. In addition, it was also found that many stress-responsive genes, including the DEGs response to abiotic stimulus, were signi cantly enriched (Fig. 6).
Many reports have shown that drought stress has a great impact on soil microbial communitieswhich play an important role in regulating plant response to drought stress [49]. Drought stress could lead to a signi cant reduction of microbial biomass [50][51][52] and change the composition of plant rhizosphere microbial. The drought tolerance of plants is related to the change of relative abundance of speci c bacterial groups [29][30][31][32]40]. Although our understanding of the interaction between plants and soil microbial in drought responses is advancing, most of the knowledge comes from non-crop plants. The results in this study showed that the application of γ-PGA under the drought stress did not affect the species of dominant bacteria, but change the bacterial community diversity. Under drought stress, Actinobacteria and Chloro exi were signi cantly enriched in soil supplemented with γ-PGA (Fig. 5C). Actinobacteria and Chloro exi were reported to be the most prominent phylum of drought enrichment [53]. Actinobacteria was previously found to promote the decomposition or formation of humus, making it easier to be absorbed [54,55], and it was also reported to have the important role in plant defense and growth promotion [56][57][58]. In this study, it was also found that Alphaproteobacteria and Deltaproteobacteria were enriched at class level after addition of γ-PGA. Most members of Proteobacteria were reported to play important roles in nitrogen xation [59,60]. Exogenous application of γ-PGA under drought stress could also enrich Sphingobium, Sphingomonas, Sphingopyxis, Haliangium of Proteobacteria. In addition, Subgroup_10 of Acidobacteria, Clostridium and Pelotomaculum of Firmicutes, which was previously reported to promote plant growth through nitrogen xation, phosphate solubilization and production of plant hormone [54], were also found to be signi cantly enriched in the γ-PGA added soil in this study.
It is worth noting that γ-PGA increased the urease activity of rhizosphere soils of maize under the severe drought stress ( Table 2). The activities of soil urease play an important role in soil nitrogen transformation, which produces NH 3, NH 4 + and CO 3 2− in the process of urea hydrolysis and provides nutrition for plants. The results implied that exogenous application of γ-PGA could contribute to improve the soil biochemical reaction and plant growth under the drought stress condition. In addition, interestingly, we also detected a signi cant increase in ABA content in the rhizosphere soil after γ-PGA application, which will also play an important role in drought resistance of maize. The mechanism of the increase of urease activity and ABA content in soil by exogenous application of γ-PGA needs further study. Our results showed that exogenous application of PGA not only affected the physiological and biochemical indexes and gene expression related to drought resistance of plants, but also profoundly affected the microbial community and physiological and biochemical properties of rhizosphere soil.

Conclusions
Our study demonstrated that exogenous application of γ-PGA could signi cantly enhance the drought resistance of maize under severe drought stress. γ-PGA can regulate the expression of ABA biosynthesis, ABA signal transduction related genes, photosynthesis-related genes and other stress-responsive genes. At the same time, γ-PGA could enrich the plant-promoting bacteria such as Actinobacteria, Chloro exi, Firmicutes, Alphaproteobacteria and Deltaproteobacteria. This study highlighted the possibility of using γ-PGA to improve crop drought resistance and soil environment under drought condition.

Plant materials and drought treatments
Maize (inbred line KN5585) seeds (provied by Weimi Biotechnology (Jiangsu) Co., Ltd (Changzhou, China)) were sown in a soil box (10cm*10cm*10cm). When seeds germinated, the seedlings were watered with different concentrations (0, 50, 70, 100 mg/L) of γ-PGA (10KD) solution and grown in greenhouse at 28±2℃ under nature light and 25±2℃ at night. All seedlings at three-leaf stage were exposed to a drought stress treatment by stopping watering to select the most suitable treatment concentration of γ-PGA. After drought for 7 days (the soil water content decreased to 4.9 %, and the control plants wilted seriously), the seedlings were rewatered. After rewatering for 1 day, the recovered maize added with γ-PGA were recorded and compared with the control maize. 50 mg/L γ-PGA was selected for the subsequent experiment according to the results of drought lethal test. The physiological parameters including photosynthetic parameters (net photosynthetic CO 2 assimilation rate, stomatal conductance), soluble sugar, proline, chlorophyll and ABA contents of the maize added with γ-PGA and the control maize were measured after 5 days treatment (soil water content decreased to 9.8 %). Soil water content was monitored by using Soil Moisture Content Meter (TZS, TOP instrument, China). Each experiment had at least three biological repetitions, and the determination of photosynthetic parameters was repeated at least ve times. The leaves were taken for RNA sequencing. Finally, the dry weights of plants under drought conditions were measured.
In the experiment of using PEG to simulate drought stress, maize (inbred line KN5585) seeds were surface sterilized using 75% alcohol and germinated on moist lter paper in sterile petri dishes (diameter:12.5 cm) in the dark at 28°C. After 4 days, the germinated seeds were transferred to the culture asks (height: 15cm, diameter: 7cm) with Hoagland Solution, and grown at 28℃/25℃ (16h light/8h dark) until maize reached to two-leaf stage. Then the maize seedlings were divided into four groups and cultured as follows: group1, cultured with Hoagland Solution only; group2, cultured with Hoagland Solution supplemented with 18% (m/v) PEG6000 (−0.77 MPa) solution; group3, cultured with Hoagland Solution supplemented with γ-PGA (10 kDa, 50 mg L −1 ); group 4, cultured with Hoagland Solution supplemented with γ-PGA (10 kDa, 50 mg L −1 ) + 18% (m/v) PEG6000. The nutrient solution was renewed every 2 days, aerated with a mini air pump and supplemented with fresh solution. The phenotypes of plants were examined, the leaf and root fresh weight were measured.

Determination of physiological parameters
The leaves and roots disk from the plants were excised, and the fresh weights (FWs) were recorded immediately. The dry weights (DWs) of leaves were obtained after drying in an oven at 80℃. The photosynthetic parameters (net photosynthetic CO 2 assimilation rate and stomatal conductance) were measured at 28°C PAR 1000 µmol/m 2 /s by a portable infrared gas analyser-based photosynthesis system (Yaxin-1105, China). Total soluble sugars of leaves (approximately 100 mg) were extracted in boiling water for 30 min and determined by anthrone reagent using glucose as the standard according to the methods described by Yemm and Willis [61]. Proline was detected using the protocol described by Bates et al [62]. Approximately 200 mg of the maize leaves was excised to measure chlorophyll content following the method described by Arnon [63]. The ABA content was measured with ELISA kit (code JM-01148P2, Jingmei Bio Inc,. Jiangsu, China) according to the manufacturer's protocol. The urease activity was determined according to the described method by Guan [64]. In this study, at least three biological repeats were sampled for one treatment, each replicate contained tissues from four plants, and the determination of photosynthetic parameters was repeated at least ve times.  Table S2. RNA sequencing and analysis RNA sequencing and primary bioinformatics analysis were performed by BGI Tech Solutions Co., Ltd. (She nzhen, China). Each treatment was made three biological replicates. Primary sequencing data (raw read) were produced by Illumina HiSeq™ 2000. After QC, raw reads were ltered into clean reads which will be aligned to the reference sequences. The alignment data was utilized to calculate distribution of reads on reference genes and mapping ratio. Gene expression was measured as fragments per kilobase of transcript per million fragments mapped (FPKM) using Cu inks. Differentially expressed genes (DEGs) were determined using DEseq2. The false discovery rate was used to adjust the P-values. Genes with signi cant differences in expression, |log2Fold Change|≥1, and adjusted P-value <0.05 were considered as DEGs. GO analysis and pathway enrichment analysis of all DEGs (Q value≤0.05) were performed by AgriGO (http:// bioinfo.cau.edu.cn/agriGO/) and KEGG (http://www.genome.jp/kegg/). The promoter motif analysis was conducted using PlantCARE (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/).

Bacterial Community Analysis of maize rhizosphere soil
The two-leaf stage maize seedlings watered with γ-PGA (0, 50mg/L) were treated under drought stress and kept the soil moisture content at 8.0% by replenishment. After 30 days, the tightly bound soils of roots (served as rhizosphere soils) were taken to analyze the microbial community, and three biological replicates were performed. Ampli cation and High-throughput sequencing of 16s rRNA from maize rhizosphere soil bacterial were performed as described by Wang et al.[65]. The primers of V4 region of bacterial 16S rRNA were 338F (5'-ACTCCTACGGGAGGCAGCA-3') and 806R (5'-GGACTACHVGGGTWTCTAAT-3'). High-throughput sequencing was conducted by Illumina Hiseq 2000 (Illumina Inc., San Diego, USA). Nonmetric multidimensional scaling (NMDS) was performed on distance matrices and the coordinates were used to draw 2D graphical outputs. Taxa abundances at the phylum, class, order, family and genus levels were statistically compared among samples or groups by Metastats. The LEfSe analysis (LDA≥3) was carried out to obtain the important indicator taxa with signi cant changes in relative abundance.

Statistical Analysis
All data have at least three biological replicates. The data were presented as the mean ± standard deviation (SD). The statistical analysis between the maize with and without γ-PGA was performed using T-test and Duncan's tests of one-way ANOVAs in SPSS (version 22.0.0.0). Signi cant differences were indicated by asterisks, *p < 0.05; **p < 0.01.

Declarations Ethics approval and consent to participate
There is no ethics approval and consent to participate in this manuscript.

Consent to publication
Not applicable.

Competing interests
The authors declare no con ict of interest.

Availability of data and materials
All datasets generated for this study are included in the article/Supplementary Materials.      Proposed model for the role of γ-PGA on maize under long-term drought. γ-PGA can improve the drought resistance of maize by regulating the expression of ABA biosynthesis, ABA signal transduction related genes, photosynthesis-related genes and other stress-responsive genes (osmatic protection, stress response and protein folding genes) and enriching the plant-promoting bacteria such as Actinobacteria, Chloro exi, Firmicutes, Alphaproteobacteria and Deltaproteobacteria. In addition, ABA contents and urease activity in maize rizosphere soil were also increased.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.