Functional Study of PgGRAS68-01 Gene Involved in the Regulation of Ginsenoside Biosynthesis in Panax ginseng

Ginseng (Panax ginseng C. A. Meyer) is a perennial herb from the genus Panax in the family Araliaceae. It is famous in China and abroad. The biosynthesis of ginsenosides is controlled by structural genes and regulated by transcription factors. GRAS transcription factors are widely found in plants. They can be used as tools to modify plant metabolic pathways by interacting with promoters or regulatory elements of target genes to regulate the expression of target genes, thereby activating the synergistic interaction of multiple genes in metabolic pathways and effectively improving the accumulation of secondary metabolites. However, there are no reports on the involvement of the GRAS gene family in ginsenoside biosynthesis. In this study, the GRAS gene family was located on chromosome 24 pairs in ginseng. Tandem replication and fragment replication also played a key role in the expansion of the GRAS gene family. The PgGRAS68-01 gene closely related to ginsenoside biosynthesis was screened out, and the sequence and expression pattern of the gene were analyzed. The results showed that the expression of PgGRAS68-01 gene was spatio-temporal specific. The full-length sequence of PgGRAS68-01 gene was cloned, and the overexpression vector pBI121-PgGRAS68-01 was constructed. The ginseng seedlings were transformed by Agrobacterium rhifaciens-mediated method. The saponin content in the single root of positive hair root was detected, and the inhibitory role of PgGRAS68-01 in ginsenoside synthesis is reported.


Introduction
Medicinal plants are inseparable from people's lives. They can produce secondary metabolites, which are used for health preservation, treatment, and prevention of diseases. Ginseng is a perennial herb of the genus Panax family Araliaceae. All the tissues and parts of Panax ginseng can be used as Chinese herbal medicine, and it is a very important medicinal material [1,2]. Ginsenosides are the main bioactive components of ginseng and secondary metabolites of ginseng cells, which have high health benefits and medicinal value. Oleane-type pentacyclic triterpenoid saponins and Dammarane-type tetracyclic triterpenoid saponins are the main types and bioactive components of ginsenosides. Dammarane-type tetracyclic triterpenoid saponins are divided into two groups: 20 s-protopanaxadiol (PPD), such as Ra1, Ra2, Rb, Rb2, and Rb3, and 20S-protopanaxatriol (PPT), such as Re, Rf, Rg1, Rg2, and Rh1 [3]. The biosynthetic pathways of triterpenoid saponins include mevalonate (MVA) and 2-C-methyl-D-erythritol 4-phosphate (MEP). The ginsenoside biosynthesis pathway consists of more than 20 consecutive enzymatic reaction steps. A series of key enzymes, such as 3-hydroxy-3-methylglutaryl CoA reductase (HMGR), farnesyl pyrophosphate synthase (FPS), squalene synthase (SS), and squalene epoxidase [4][5][6][7] play an important role in ginsenoside synthesis, and the study of these key enzymes and their regulatory genes has attracted much attention.
Transcription factors are important regulatory factors affecting the growth and development of higher plants, physiological processes, and network regulation, and are also essential in the regulation of gene expression. GRAS transcription factors appeared in land plants through lateral bacterial transfer and were transmitted in the ancestors of bryophytes, lycophytes, and higher plants [8].
GRAS proteins have highly conserved sequences in their carboxyl termini. However, the carboxyl termini of some GRAS proteins contain two conserved GRAS domains or one GRAS domain and one functional region. GRAS proteins contain several ordered motifs LHRI, VHIID, LHRII, PFYRE, and SAW at the carboxyl terminus [9,10]. Among these conserved sequences, the VHIID sequence (where V is valine, I is isoleucine, H histidine, and D is aspartic acid), which is essential for protein-protein interactions, is located midway between the leucine-rich LHR I and LHR II motifs. It is the most essential and central part of GRAS proteins [11,12]. The amino-terminal portion of GRAS proteins is variable except for the highly conserved DELLA protein and TVHYNP protein, both of which are involved in gibberellin signal transduction and can control plant growth and development [13]. This further demonstrates the functional specificity of GRAS proteins.
DELLA protein is a regulator of nuclear transcription, and DELLA protein mainly restricts plant growth by inhibiting GA signal transduction [14,15]. The DELLA domain is a domain unique to DELLA subfamily proteins in the GRAS transcription factor subfamily, and its function is related to GA response. Studies have shown that DELLA protein is a constitutive repressor of GA response and inhibits GA-mediated plant growth and development. GA promotes plant growth and development by reducing the repressor of DELLA proteins. Large-scale genetic screens and analyses in Arabidopsis and rice have identified several key components of GA signaling, including the soluble GA receptor GID1 and the growth repressor DELLA protein [16,17]. Liu et al. found that DELLA proteins in Arabidopsis could interact with JAZ1, a repressor of jasmonate [18]. In plants, JAZ1 inhibits the MYC2 transcription factor and thus affects the production of jasmonate. DELLA protein promotes MYC2 production and enhances its binding to downstream target genes by inhibiting JAZ1, thereby increasing jasmonate production [13]. According to the study by Zhang et al. in 2017, AtDELLA protein can promote anthocyanin synthesis. The anthocyanin synthesis pathway was inhibited in mutants such as gai-t6 and rgl1-1. DELLA protein increases anthocyanin synthesis by inhibiting PAP1 protein [19]. In medicinal plants, the overexpression of SmSCR1 in Salvia miltiorrhiza resulted in significantly higher accumulation of tanshinone than that in the control group [20]. In Dendrobium officinale, eight GRAS genes were upregulated in different tissues after heat and salt stress [21]. In garlic, GRAS gene has five protein-coding genes that have been identified as DELLA samples, three genes Asa2G00237.1, Asa2G00240.1, and Asa4G02090.1 have response to exogenous GA 3 treatment, and part of the GRAS genes regulating garlic bulb growth [22]. In a previous study, the PgGRAS gene expression was activated in ginseng hair-like roots treated with gibberellin acid (GAs), and qPCR analysis showed that PgGRAS genes of DELLA subfamily played an important role in response to GA treatment in ginseng hairy roots [23]. Therefore, we speculated that PgGRAS gene might also play an important role in the biosynthesis of ginsenosides.
In this study, PgGRAS68-01, a gene highly related to ginsenoside biosynthesis, was identified based on published PgGRAS genes transcription factor data [23]. The PCR cloning method was used to clone PgGRAS68-01 gene, and its protein sequence characteristics, phylogenetic relationship and expression pattern were analyzed. Then, Agrobacterium tumefaciens was used to overexpress PgGRAS68-01 gene in ginseng explants, and the role of PgGRAS68-01 gene in ginsenoside synthesis in ginseng was initially revealed. The results of this study can lay a theoretical foundation for further research on the function of PgGRAS and the industrial production of ginsenosides through genetic engineering.

Chromosome Distribution and Gene Replication of PgGRAS Gene
We found that members of the PgGRAS gene family showed an uneven distribution on 24 chromosomes of ginseng ( Figure 1A). These genes are not found on chromosomes 8, 15, 16, 18, 20, and 24. We noted that PgGRAS gene had the highest gene density at different positions on chromosome 1, while there was only one PgGRAS gene on chromosome 3, 6, and 13. Interestingly, to investigate PgGRAS gene replication events, we found collinear segmented replication of the PgGRAS gene in ginseng. This result can prove the existence of PgGRAS gene replication in ginseng genome ( Figure 1B). phylogenetic relationship and expression pattern were analyzed. Then, Agrobacterium tumefaciens was used to overexpress PgGRAS68-01 gene in ginseng explants, and the role of PgGRAS68-01 gene in ginsenoside synthesis in ginseng was initially revealed. The results of this study can lay a theoretical foundation for further research on the function of PgGRAS and the industrial production of ginsenosides through genetic engineering.

Chromosome Distribution and Gene Replication of PgGRAS Gene
We found that members of the PgGRAS gene family showed an uneven distribution on 24 chromosomes of ginseng ( Figure 1A). These genes are not found on chromosomes 8, 15, 16, 18, 20, and 24. We noted that PgGRAS gene had the highest gene density at different positions on chromosome 1, while there was only one PgGRAS gene on chromosome 3, 6, and 13. Interestingly, to investigate PgGRAS gene replication events, we found collinear segmented replication of the PgGRAS gene in ginseng. This result can prove the existence of PgGRAS gene replication in ginseng genome ( Figure 1B).

GA Treatment Affects the Synthesis of Ginsenosides
The total saponins of hair roots of ginseng treated with gibberellin were measured. Compared with the untreated hairy root, the total glycosides content of the treated hairy root was higher than that of the untreated hairy root ( Figure 2). It can be proven that gibberellin plays a role in inhibiting ginsenoside synthesis in the hair root culture system of ginseng.

GA Treatment Affects the Synthesis of Ginsenosides
The total saponins of hair roots of ginseng treated with gibberellin were measured. Compared with the untreated hairy root, the total glycosides content of the treated hairy root was higher than that of the untreated hairy root ( Figure 2). It can be proven that gibberellin plays a role in inhibiting ginsenoside synthesis in the hair root culture system of ginseng.

Screening of GRAS Candidate Genes Involved in Ginsenoside Biosynthesis
Ginsenoside is the main active component of ginseng, but its content in ginseng is very low, so it is very important to study the synthesis pathway of ginsenoside. A total of 34 genes were found to be significantly correlated with ginsenoside content (Table S1), among which 18 PgGRAS genes were positively correlated with ginsenoside content, and 14 PgGRAS genes were negatively correlated with ginsenoside content. Two PgGRAS genes (PgGRAS62-03 and PgGRAS69-03) were positively correlated with saponin content, and also negatively correlated with saponin content.
Many key enzymes in ginsenoside synthesis have been cloned and verified. PgGRAS genes may be related to key enzyme genes in ginsenoside synthesis pathway. In order to explore the relationship between PgGRAS genes and key enzyme genes, SPSS version 23.0 software was used to calculate the correlation between the expression levels of key enzyme genes and PgGRAS genes (Table S2). A total of 34 genes were found to be significantly correlated with the expression of key enzyme genes. Among them, the expression of 19 PgGRAS genes was positively correlated with the expression of key enzyme genes. Among them, the expression of 9 PgGRAS genes was significantly negatively correlated with the expression of key enzyme genes.

Screening of GRAS Candidate Genes Involved in Ginsenoside Biosynthesis
Ginsenoside is the main active component of ginseng, but its content in ginseng is very low, so it is very important to study the synthesis pathway of ginsenoside. A total of 34 genes were found to be significantly correlated with ginsenoside content (Table S1), among which 18 PgGRAS genes were positively correlated with ginsenoside content, and 14 PgGRAS genes were negatively correlated with ginsenoside content. Two PgGRAS genes (PgGRAS62-03 and PgGRAS69-03) were positively correlated with saponin content, and also negatively correlated with saponin content.
Many key enzymes in ginsenoside synthesis have been cloned and verified. PgGRAS genes may be related to key enzyme genes in ginsenoside synthesis pathway. In order to explore the relationship between PgGRAS genes and key enzyme genes, SPSS version 23.0 software was used to calculate the correlation between the expression levels of key enzyme genes and PgGRAS genes (Table S2). A total of 34 genes were found to be significantly correlated with the expression of key enzyme genes. Among them, the expression of 19 PgGRAS genes was positively correlated with the expression of key enzyme genes. Among them, the expression of 9 PgGRAS genes was significantly negatively correlated with the expression of key enzyme genes.
Since the key enzyme genes of ginseng participate in the ginsenoside synthesis pathway and are an important part of it, it is particularly important to study the correlation between the expression of key enzyme genes and ginsenoside content. Therefore, the comparison and analysis of the results calculated by SPSS version 23.0 software found a total of 10 genes (PgGRAS44-04, PgGRAS45-01, PgGRAS51-01, PgGRAS51-02, PgGRAS56-02, PgGRAS62-06, PgGRAS63-01, PgGRAS63-02, PgGRAS65-03, and PgGRAS68-01) were significantly correlated with both key enzyme genes and ginsenoside content of ginseng. However, the PgGRAS68-01 gene was found to be related to more key enzyme genes through network (Figure 3), so PgGRAS68-01 gene was finally selected for subsequent functional verification. Since the key enzyme genes of ginseng participate in the ginsenoside synthesis pathway and are an important part of it, it is particularly important to study the correlation between the expression of key enzyme genes and ginsenoside content. Therefore, the comparison and analysis of the results calculated by SPSS version 23.0 software found a total of 10 genes (PgGRAS44-04, PgGRAS45-01, PgGRAS51-01, PgGRAS51-02, PgGRAS56-02, PgGRAS62-06, PgGRAS63-01, PgGRAS63-02, PgGRAS65-03, and PgGRAS68-01) were significantly correlated with both key enzyme genes and ginsenoside content of ginseng. However, the PgGRAS68-01 gene was found to be related to more key enzyme genes through network (Figure 3), so PgGRAS68-01 gene was finally selected for subsequent functional verification.

Sequence Analysis of PgGRAS68-01 Gene
The total length of PgGRAS68-01 gene is 2041 bp, encoding 540 amino acids with molecular weight of 59,906.01 kDa and isoelectric point (PI) of 5.03. The secondary structure of PgGRAS68-01 protein contained 257 alpha helix (47.59%), 27 beta turn (5.00%), 202 random coil (37.40%), and 54 extended strand (10.00%) ( Figure 4A). Tertiary structural modeling showed that PgGRAS68-01 was composed of alpha helix, beta turn, and random coil ( Figure 4B). In order to reveal the evolutionary relationship between GRAS genes in different species, the protein sequences of three GRAS family members of Arabidopsis, rice, and tomato were downloaded from NCBI (Table S3). Phylogenetic trees were constructed using three PgGRAS protein sequences containing PgGRAS68-01 and nine GRAS protein sequences from three other species ( Figure 4C). PgGRAS68-01 has the closest evolutionary relationship with Arabidopsis AtGRAS3 and AtGRAS10. At the protein level, PgGRAS68-01 showed a high degree of similarity to its orthologs in different plant species ( Figure 4D).

Expression Pattern Analysis of PgGRAS68-01 Gene in Ginseng
In order to understand the expression pattern of PgGRAS68-01 gene in ginseng, we retrieved the PgGRAS68-01 gene expression data from 4 different aged stages of ginseng roots, 14 different tissues of 4-year-old ginseng, and 42 farmer's cultivars of 4-year-old ginseng roots, and plotted the bar graph ( Figure 5). In order to more intuitively reflect the expression level of PgGRAS68-01 gene, we drew the gene expression heat map. Among the 4 different aged stages of ginseng roots, the expression level of PgGRAS68-01 gene was the highest in 18-year-old ginseng roots and the lowest in 12-year-old ginseng roots ( Figure 6A). Among the 14 different tissues of 4-year-old ginseng, PgGRAS68-01 was expressed in all the tissues, with high expression in fruit flesh and rhizome and the least expression in leaf blade ( Figure 6B). Among the 42 farmer's cultivars of 4-year-old ginseng roots, PgGRAS68-01 gene expression was significantly different in most of the farmer's cultivars, with higher expression in S21 and S23 and lower expression in S7 and S36 ( Figure 6C).

Sequence Analysis of PgGRAS68-01 Gene
The total length of PgGRAS68-01 gene is 2041 bp, encoding 540 amino acids with molecular weight of 59906.01 kDa and isoelectric point (PI) of 5.03. The secondary structure of PgGRAS68-01 protein contained 257 alpha helix (47.59%), 27 beta turn (5.00%), 202 random coil (37.40%), and 54 extended strand (10.00%) ( Figure 4A). Tertiary structural modeling showed that PgGRAS68-01 was composed of alpha helix, beta turn, and random coil ( Figure  4B). In order to reveal the evolutionary relationship between GRAS genes in different species, the protein sequences of three GRAS family members of Arabidopsis, rice, and tomato were downloaded from NCBI (Table S3). Phylogenetic trees were constructed using three PgGRAS protein sequences containing PgGRAS68-01 and nine GRAS protein sequences from three other species ( Figure 4C). PgGRAS68-01 has the closest evolutionary relationship with Arabidopsis AtGRAS3 and AtGRAS10. At the protein level, PgGRAS68-01 showed a high degree of similarity to its orthologs in different plant species ( Figure 4D).

Cloning and Vector Construction of PgGRAS68-01 Gene
The total RNA of Jilin ginseng was extracted and reverse transcribed into cDNA, which was used as a template to clone PgGRAS68-01 gene. As shown in Figure 7A, a clear band at 1656bp was observed. The cloned target gene fragment was ligated with T vector and transformed into Escherichia coli and screened by blue and white spots. The selected single colonies were amplified and PCR positive clones were verified. The cloned vector was sequenced and the results were identical with the target gene. Plasmids were recovered and double digested with Xba I and Sma I ( Figure 7B). Agarose gel electrophoresis was used for separation and recovery. After the target band was recovered, T4 ligase was used to connect it to the pBI121 expression vector to construct the recombinant plant expression vector plasmid PBI121-PGGras68-01 ( Figure 7C), which was transformed and verified by bacterial liquid PCR ( Figure 7D). These results indicated that the expression vector pBI121-PgGRAS68-01 was successfully constructed.

Genetic Transformation of Ginseng by Recombinant Plasmid
The recombinant plasmid was transformed into Agrobacterium A4 and PCR positive cloning detection was performed; the bacterial solution with clear and bright bands was selected for preservation. The petioles of ginseng sterile seedlings were infected with the transformed Agrobacterium A4 and cultured in 1/2MS medium containing 200 mg·L −1 Cef under dark conditions for 5 weeks. Hairy roots were observed on the explants ( Figure 8A-D). When the hairy roots grew to about 2 cm, they were transferred to the medium containing 200 mg·L −1 Cef for propagation. After obtaining hair-like roots, PCR method was used to detect positive hair-like roots. According to the recombinant plasmid, four primers were designed ( Figure 8E), and the genomic DNA of hair-like roots was extracted for PCR detection. The size of PgGRAS68-01 gene band is shown in Figure 8F and the band is bright and clear. PgGRAS68-01 gene integration in the hair root genome can be preliminarily identified. coil (37.40%), and 54 extended strand (10.00%) ( Figure 4A). Tertiary structural modeling showed that PgGRAS68-01 was composed of alpha helix, beta turn, and random coil ( Figure  4B). In order to reveal the evolutionary relationship between GRAS genes in different species, the protein sequences of three GRAS family members of Arabidopsis, rice, and tomato were downloaded from NCBI (Table S3). Phylogenetic trees were constructed using three PgGRAS protein sequences containing PgGRAS68-01 and nine GRAS protein sequences from three other species ( Figure 4C). PgGRAS68-01 has the closest evolutionary relationship with Arabidopsis AtGRAS3 and AtGRAS10. At the protein level, PgGRAS68-01 showed a high degree of similarity to its orthologs in different plant species ( Figure 4D).

Expression Pattern Analysis of PgGRAS68-01 Gene in Ginseng
In order to understand the expression pattern of PgGRAS68-01 gene in ginseng, we retrieved the PgGRAS68-01 gene expression data from 4 different aged stages of ginseng roots, 14 different tissues of 4-year-old ginseng, and 42 farmer's cultivars of 4-year-old ginseng roots, and plotted the bar graph ( Figure 5). In order to more intuitively reflect the expression level of PgGRAS68-01 gene, we drew the gene expression heat map. Among the 4 different aged stages of ginseng roots, the expression level of PgGRAS68-01 gene was the highest in 18-year-old ginseng roots and the lowest in 12-year-old ginseng roots ( Figure 6A). Among the 14 different tissues of 4-year-old ginseng, PgGRAS68-01 was expressed in all the tissues, with high expression in fruit flesh and rhizome and the least expression in leaf blade ( Figure 6B). Among the 42 farmer's cultivars of 4-year-old ginseng roots, PgGRAS68-01 gene expression was significantly different in most of the farmer's cultivars, with higher expression in S21 and S23 and lower expression in S7 and S36 ( Figure 6C).

Cloning and Vector Construction of PgGRAS68-01 Gene
The total RNA of Jilin ginseng was extracted and reverse transcribed into cDNA, which was used as a template to clone PgGRAS68-01 gene. As shown in Figure 7A, a clear band at 1656bp was observed. The cloned target gene fragment was ligated with T vector and transformed into Escherichia coli and screened by blue and white spots. The selected single colonies were amplified and PCR positive clones were verified. The cloned vector was sequenced and the results were identical with the target gene. Plasmids were recovered and double digested with Xba I and Sma I ( Figure 7B). Agarose gel electrophoresis was used for separation and recovery. After the target band was recovered, T4 ligase was used to connect it to the pBI121 expression vector to construct the recombinant plant expression vector plasmid PBI121-PGGras68-01 ( Figure 7C), which was transformed and verified by bacterial liquid PCR ( Figure 7D). These results indicated that the expression vector pBI121-PgGRAS68-01 was successfully constructed.

Genetic Transformation of Ginseng by Recombinant Plasmid
The recombinant plasmid was transformed into Agrobacterium A4 and PCR positive cloning detection was performed; the bacterial solution with clear and bright bands was selected for preservation. The petioles of ginseng sterile seedlings were infected with the transformed Agrobacterium A4 and cultured in 1/2MS medium containing 200 mg·L −1 Cef under dark conditions for 5 weeks. Hairy roots were observed on the explants ( Figure 8A-

Detection of Ginsenoside Content in Positive Ginseng Hairy Roots
The single root system of ginseng hair-like root with positive test was propagated in large numbers ( Figure 9A), and the fresh ginseng hair-like root ( Figure 9B) was dried to obtain the dry ginseng hair-like root ( Figure 9C). The methanol extract of positive hair-like root was obtained by grinding the dry ginseng hair-like root and extracting with Soxhlet extractor. The content of ginsenoside in positive hair-like roots was detected by HPLC, and the results showed that the sum content of ginsenoside (Re + Rf + Rb1 + Rc + Rb2 + Rd + Rg3 + Rh2 + PPT + PPD) in the single root of positive hair-like roots of transgenic line-18, transgenic line-19, and transgenic line-30 was lower than that of the control group ( Figure 10A), indicating that the transfer of PgGRAS68-01 gene had an inhibitory effect on ginsenoside. This was also consistent with the previous results that PgGRAS68-01 gene was significantly negatively correlated with the content of saponins. As shown in Figure 10B-G, analysis of monomeric saponins showed that compared with the control group, the contents of monomeric saponins Rg1, Rb1 and Rg3 were significantly reduced, PPD and PPT were also significantly reduced, but the content of monomeric saponins Re was increased. These results demonstrate that PgGRAS68-01 regulates ginsenoside synthesis.

Detection of Ginsenoside Content in Positive Ginseng Hairy Roots
The single root system of ginseng hair-like root with positive test was propagated in large numbers ( Figure 9A), and the fresh ginseng hair-like root ( Figure 9B) was dried to obtain the dry ginseng hair-like root ( Figure 9C). The methanol extract of positive hair-like root was obtained by grinding the dry ginseng hair-like root and extracting with Soxhlet extractor. The content of ginsenoside in positive hair-like roots was detected by HPLC, and the results showed that the sum content of ginsenoside (Re + Rf + Rb1 + Rc + Rb2 + Rd + Rg3 + Rh2 + PPT + PPD) in the single root of positive hair-like roots of transgenic line-18, transgenic line-19, and transgenic line-30 was lower than that of the control group ( Figure 10A), indicating that the transfer of PgGRAS68-01 gene had an inhibitory effect on ginsenoside. This was also consistent with the previous results that PgGRAS68-01 gene was significantly negatively correlated with the content of saponins. As shown in Figure 10B-G, analysis of monomeric saponins showed that compared with the control group, the contents of monomeric saponins Rg1, Rb1 and Rg3 were significantly reduced, PPD and PPT were also significantly reduced, but the content of monomeric saponins Re was increased. These results demonstrate that PgGRAS68-01 regulates ginsenoside synthesis.

Discussion
To date, GRAS transcription factors have been studied in many species and have been identified in plants such as Arabidopsis thaliana (33) [12] (Lee et al., 2008), rose (59) [24], rice (57) [25], tomato (53) [26], and grape (52) [27]. The GRAS gene family of Arabidopsis thaliana, a model plant, has been well studied, and the functions of several family members have been verified. We found that most of the ginseng GRAS gene family was distributed on 24 pairs of chromosomes, and only a few chromosomes did not have PgGRAS gene. PgGRAS gene duplication occurred in 24 pairs of ginseng chromosomes, and a total of 11 duplicate events were identified (Figure 1). Meanwhile, GRAS gene duplications

Discussion
To date, GRAS transcription factors have been studied in many species and have been identified in plants such as Arabidopsis thaliana (33) [12] (Lee et al., 2008), rose (59) [24], rice (57) [25], tomato (53) [26], and grape (52) [27]. The GRAS gene family of Arabidopsis thaliana, a model plant, has been well studied, and the functions of several family members have been verified. We found that most of the ginseng GRAS gene family was distributed on 24 pairs of chromosomes, and only a few chromosomes did not have PgGRAS gene. PgGRAS gene duplication occurred in 24 pairs of ginseng chromosomes, and a total of 11 duplicate events were identified (Figure 1). Meanwhile, GRAS gene duplications

Discussion
To date, GRAS transcription factors have been studied in many species and have been identified in plants such as Arabidopsis thaliana (33) [12] (Lee et al., 2008), rose (59) [24], rice (57) [25], tomato (53) [26], and grape (52) [27]. The GRAS gene family of Arabidopsis thaliana, a model plant, has been well studied, and the functions of several family members have been verified. We found that most of the ginseng GRAS gene family was distributed on 24 pairs of chromosomes, and only a few chromosomes did not have PgGRAS gene. PgGRAS gene duplication occurred in 24 pairs of ginseng chromosomes, and a total of 11 duplicate events were identified (Figure 1). Meanwhile, GRAS gene duplications have been found in many species, including quinoa [28], switchgrass [29], millet [30], sorghum [31] and alfalfa [32]. In Arabidopsis, DELLA proteins can inhibit plant growth by inhibiting the GA pathway [17]. Ginsenosides are the main active components of ginseng and have high medicinal value. Previous studies showed that the dry weight and fresh weight of hair roots of ginseng decreased after exogenous addition of GA, indicating that the addition of GA inhibited its growth [23]. In order to explore the effect of exogenous GA on ginsenosides content, different concentrations of GA were used to treat ginseng hair-like roots. The results showed that the total glycosides content of all treated ginseng hair-like roots was lower than that of untreated ginseng hair-like roots (Figure 2). The results showed that exogenous GA could inhibit the growth and development of ginseng and the synthesis of ginsenoside.
GRAS genes are involved in the regulation of anthocyanins and other secondary metabolites [19], but studies on GRAS genes involved in the biosynthesis of ginsenosides have not been reported. In the PgGRAS gene family, PgGRAS68-01, which is highly correlated with ginsenoside content, belongs to the DELLA subfamily. Sequence alignment and evolutionary analysis with other species revealed that PgGRAS68-01 had the closest evolutionary relationship with Arabidopsis AtGRAS10 and rice OsGRAS25, and both had DELLA domains at the N-terminus, which was consistent with our previous study. Interestingly, after analyzing the expression pattern of PgGRAS68-01 gene, we found that PgGRAS68-01 was expressed in all the 42 farmer's cultivars, indicating that the expression of PgGRAS68-01 was widespread. However, PgGRAS68-01 also had its specificity and was highly expressed in S21 and S23 farm cultivars. The expression of PgGRAS68-01 gene in 14 different tissues of 4-year-old ginseng was also specific, and the highest expression level was found in fruit flesh. The expression level of PgGRAS68-01 was the highest in the roots of 4 different aged stages of ginseng roots, which indicated that the expression of PgGRAS68-01 gene in ginseng was spatio-temporal specific.
Then, we cloned the full-length PgGRAS68-01 gene and constructed a recombinant plant overexpression vector. After genetic transformation of ginseng explants, positive hairy roots were finally obtained. Detection of ginsenoside content in single root of positive hairlike root showed that PgGRAS68-01 gene inhibited ginsenoside production, which also proved that PgGRAS68-01 gene regulated ginsenoside synthesis. Subsequently, some monomeric saponins were analyzed, and it was found that the content of monomeric saponins such as Rg1, Rb1, and Rg3 decreased, while the content of monomeric saponins Re increased ( Figure 10). Through the study of ginsenoside synthesis pathway, it was found that monomeric saponin Re was generated from monomeric saponin Rg1 in the original ginsentriol saponin synthesis pathway [3]. However, the content of monomeric saponin Re increased while the content of monomeric saponin Rg1 decreased in the single root of positive hair roots. Therefore, it is speculated that PgGRAS68-01 can promote the conversion of a large amount of Rg1 to Re, so the content of Rg1 decreases while the content of Re increases.
In conclusion, in this paper we believe that under the action of gibberellin, the expression of PgGRAS68-01 gene increases, thus inhibiting ginsenoside synthesis ( Figure 11). Previous studies have found that the generation of monomeric saponins is regulated by key enzyme genes in the ginsenoside synthesis pathway [4,5], and previous studies have shown that transcription factors generally affect metabolic pathways through the regulation of structural genes. Therefore, it is speculated that PgGRAS68-01 gene affects the content of ginsenoside by acting on key enzyme genes in the ginsenoside synthesis pathway, but this part remains to be studied.

Data and Materials
The data sources used in this study were all from the Jilin Ginseng Unigenes Database constructed by our laboratory [33]. The database was obtained by RNA extraction and sequencing of 14 tissues of four-year-old Jilin ginseng cultivar Damaya. Then, with 248,993 Unigenes constructed as a reference, Trinity software (version 2013-02-25) was used to determine the expression levels of 25-, 18-, 12-, and 5-year-old roots of ginseng, 14 tissues of ginseng, and 42 farmer's cultivars of ginseng. Expression was calculated using Transcripts Per Million (TPM). In addition, the genome derived from ginseng [34] (Wang et al., 2022) was also used in this study. The ginseng receptor material used in this experiment is sterile ginseng seedlings generated from ginseng seeds kept in the laboratory. All strains and vectors used were kept in the laboratory.

PgGRAS Gene Duplication and Chromosome Localization
Our laboratory previously identified PgGRAS genes in Jilin Ginseng Unigenes Database [23]. We used Blastn to compare the PgGRAS genes identified with ginseng genome [34] to determine their distribution in the ginseng genome and the variation of PgGRAS gene family size among different genotypes. With identity ≥ 99% and cover length ≥ 300 bp as the comparison standard, the distribution of genes of this gene family in ginseng genome was compared with each genome combination. The R-Package Circlize structure of this gene was used to determine the pan-and core transcripts of this gene family in the ginseng genotypes. The position of the transcript on the chromosome was visualized using the MG2C online tool (http://www.mg2c.iask.in/mg2c_v2.1/index.html, accessed on 1 November 2021).

Data and Materials
The data sources used in this study were all from the Jilin Ginseng Unigenes Database constructed by our laboratory [33]. The database was obtained by RNA extraction and sequencing of 14 tissues of four-year-old Jilin ginseng cultivar Damaya. Then, with 248,993 Unigenes constructed as a reference, Trinity software (version 2013-02-25) was used to determine the expression levels of 25-, 18-, 12-, and 5-year-old roots of ginseng, 14 tissues of ginseng, and 42 farmer's cultivars of ginseng. Expression was calculated using Transcripts Per Million (TPM). In addition, the genome derived from ginseng [34] ( Wang et al., 2022) was also used in this study. The ginseng receptor material used in this experiment is sterile ginseng seedlings generated from ginseng seeds kept in the laboratory. All strains and vectors used were kept in the laboratory.

PgGRAS Gene Duplication and Chromosome Localization
Our laboratory previously identified PgGRAS genes in Jilin Ginseng Unigenes Database [23]. We used Blastn to compare the PgGRAS genes identified with ginseng genome [34] to determine their distribution in the ginseng genome and the variation of PgGRAS gene family size among different genotypes. With identity ≥ 99% and cover length ≥ 300 bp as the comparison standard, the distribution of genes of this gene family in ginseng genome was compared with each genome combination. The R-Package Circlize structure of this gene was used to determine the pan-and core transcripts of this gene family in the ginseng genotypes. The position of the transcript on the chromosome was visualized using the MG2C online tool (http://www.mg2c.iask.in/mg2c_v2.1/index.html, accessed on 1 November 2021).

Culture of Ginseng Hair Roots Induced by Gibberellin and Analysis of Saponin Content Change
Ginseng hairy roots were treated with gibberellin according to the method in Wang et al.
at 2020 year. The hair roots of ginseng treated with gibberellin were dried in an oven at 37 • C, and then the dried hair roots were ground to powder, wrapped in filter paper, and soaked in methanol overnight. Ultrasound-assisted method was used for 30 min at 60 • C. The saponins were extracted in a Soxhlet extractor, and the methanol was evaporated in a water bath at 60 • C using a rotary evaporator and redissolved in 20 mL distilled water. The aqueous solution was placed in a separate funnel and extracted three times with an equal volume of ethyl acetate, and the ester phase was removed. The water layer was extracted three times with an equal volume of saturated n-butanol. The n-butanol phase was recovered to a round-bottom flask and evaporated again at 90 • C using a rotary evaporator, Dissolve in 10 mL methanol and set aside. The 40 µL sample solution was dried in a water bath, and the freshly prepared 0.2 µL vanillin (50 mg/mL)-glacial acetic acid solution and 800 µL perchloric acid were added. The solution was heated in a constant temperature water bath at 60 • C for 15 min, and then cooled by running water for 5 min. Then, 5 mL glacial acetic acid was added and the solution was shaken well immediately. The accompanying reagent was used as a blank control, and the absorbance was measured at 544 nm with a microplate reader.

Identification of PgGRAS Genes Related to Ginsenoside Synthesis
The expression data of PgGRAS genes in 42 farmer's cultivars of Jilin ginseng and the saponin content data in 42 farmer's cultivars of Jilin ginseng were sorted out, and then Pearson correlation coefficient was analyzed by SPSS software version 23.0. SPSS Statistics version 23.0 software was used to calculate the correlation between ginsenoside content and PgGRAS genes expression in 42 farmer's cultivars (Table S1). The expression data of PgGRAS genes in 42 farmer's cultivars of Jilin ginseng were extracted and placed in the same table with the known expression levels of key enzyme genes for ginsenoside synthesis. Pearson correlation coefficient was also analyzed by SPSS version 23.0 software. Spearman's correlation coefficients were calculated using R programming language version 3.5.0 and software (http://www.rproject.org/, accessed on 1 December 2021). BioLayout Express 3D version 3.2 software was used to construct the co-expression network of PgGRAS gene and key enzyme genes of ginsenoside biosynthesis and complete the visualization of gene interaction network.

Characteristic Analysis of PgGRAS68-01 Gene
Online software Expasy ProtParam (https://web.expasy.org/protparam/, accessed on 10 December 2021) was used to predict PgGRAS68-01 the basic physical and chemical properties of protein, including the theory of relative molecular mass (KDa) and isoelectric point (PI). Using the SOPMA (https://npsa-prabi.ibcp.fr/cgi-bin/npsa_automat.pl?page= npsa_sopma.html, accessed on 14 December 2021) [35] and the SWISS-MODEL (https: //www.swissmodel.expasy.org/, accessed on 15 December 2021) [36], the secondary and tertiary structure of the protein based on the amino acid sequence of PgGRAS68-01 were analyzed. Phylogenetic trees were constructed with the neighbor joining method of MEGA-X [37] using the protein sequences of PgGRAS68-01 and those of other species, and the guidance repeats were set to 1000. The protein sequences of PgGRAS68-01 and the other three species were aligned using DNAMAN version 5.2.2 software.

Expression Analysis of PgGRAS68-01 Gene
In order to further analyze the expression pattern of PgGRAS68-01 in ginseng, we selected the expression level of PgGRAS68-01 in 4 different aged stages (5, 12, 18, 25 years old) of ginseng roots, 14 different tissues of 4-year-old ginseng, and 42 farmer's cultivars of 4-year-old ginseng roots from the previous research data of the laboratory [23]. The expression was visualized using the TBtools version 1.096 software [38].

Cloning of PgGRAS68-01 Gene
Total RNA of ginseng was extracted by TRIZOL method and reverse transcribed into cDNA. The most appropriate primers were designed according to the basic principles of primer design, and the appropriate restriction sites were screened out in the target genes according to the vector map of pBI121. XbaI and protective bases were added at the upstream 5 end, and SmaI restriction sites were added at the downstream 5 end of the primers. Primers as follows, F: 5 -TGCTCTAGAATGTGGGAGGAAACTGAACAAG, R: 5 -TCCCCCGGGGAGGTCCACCACCAACTGAGT. The cDNA obtained by reverse transcription was used as the template for PCR amplification. The reaction system was as follows: 5 µL 2 × Master Mix, 0.5 µL Sense primer, 0.5 µL Anti primer, 1 µL cDNA, 3 µL ddH 2 O, for a total of 10 µL. The target gene fragment was obtained for use.

Construction of Vector and Genetic Transformation of Ginseng
Using TIANGEN PGM-T cloning ligation kit, the target gene fragment was ligated with the cloning vector and transferred into the competent Escherichia coli DH5α cells for culture. The sequencing results of the positive single colony were compared with the nucleotide sequence of the target gene. If the two were identical, the cloning vector PgGRAS68-01 was successfully constructed. Remember the pGM-T-PgGRAS68-01. The PgGRAS68-01 plasmid and pBI121 plasmid were double digested with XbaI and SmaI restriction enzymes, and the target gene of PgGRAS68-01 was ligated with the expression vector pBI121 by TAKALA T4 DNA ligase, and the double digestion was verified. The candidate gene overexpression vector pBI121-PgGRAS68-01 was successfully constructed.
To prepare genetically engineered bacteria, the pBI121-PgGRAS68-01 overexpression vector was constructed to transform competent Agrobacterium A4 cells. The petioles and roots of ginseng seedlings were precultured on MS solid medium containing 2, 4-D + 6-BA. After pre-feeding, cut into small pieces and place them in the resuspended A4 bacterial solution for infection. The infected explants were placed on solid MS medium containing acetosyringone (AS). After 3 days of co-culture at 22 • C in the dark, the explants were transferred to 1/2MS solid medium containing Cef and cultured at 22 • C in the dark, and the growth was observed and recorded. Primers were designed for the expression vectors successfully connected to the target gene, and these primers were used for PCR positive verification. The single root system of hair root which was positive was propagated.

Ginsenoside Content of Positive Hair Roots
According to the above method of extracting saponins, extract the positive hair root ginsenoside for use. The content of ginsenoside in positive hair roots was detected by HPLC with Waters e2695 HPLC on a Waters C18 column (4.6 × 250 mm, 5 µM) at 35 • C with an injection volume of 20.0 µL. Mobile phase: A. water, B. acetonitrile, gradient elution (Table 1), the flow rate of mobile phase was 1.0 mL/min, and the detection wavelength was 203 nm. Preparation of standard solution: ginsenoside standards Rb1, Rb2, Rb3, Rc, Rd, Re, Rf, Rg1, Rg2, Rg3, PPD, and PPT were accurately weighed, and an appropriate amount of methanol was added for later use.

Conclusions
In this study, most PgGRAS genes were located on 24 pairs of chromosomes in ginseng, and gene duplication was observed. The correlation analysis between the expression of PgGRAS genes and the content of saponins and the expression of key enzyme genes showed that 34 PgGRAS were significantly correlated with the content of monomeric saponins and 34 PgGRAS were significantly correlated with the expression of key enzyme genes. There were 10 genes that were significantly correlated with both saponin content and key enzyme gene expression. The PgGRAS68-01 gene was found to be associated with more key enzyme genes by network, indicating that PgGRAS genes were directly or indirectly involved in the regulation of ginsenoside synthesis. By analyzing the sequence of PgGRAS68-01 gene, we found that it has similar structure to GRAS gene of other species. After analyzing the expression pattern of PgGRAS68-01, we found that PgGRAS68-01 had spatiotemporal specificity. Then, we cloned PgGRAS68-01 gene, constructed plant expression vector, and transformed ginseng explants by Agrobacterium transformation method. After preliminary PCR identification, positive hairy roots were obtained. Compared with the control group, the contents of monomeric saponins Rg1, Rb1, Rg3, PPD, and PPT were significantly reduced, but the content of monomeric saponin Re was increased, indicating that PgGRAS68-01 gene regulates ginsenoside synthesis.