Endogenous salicylic acid shows different correlation with baicalin and baicalein in the medicinal plant Scutellaria baicalensis Georgi subjected to stress and exogenous salicylic acid

Salicylic acid (SA) is synthesized via the phenylalanine lyase (PAL) and isochorismate synthase (ICS) pathways and can influence the stress response in plants by regulating certain secondary metabolites. However, the association between SA and particular secondary metabolites in the Chinese medicinal plant Scutellaria baicalensis Georgi is unclear. To elucidate the association between SA and the secondary metabolites baicalin and baicalein, which constitute the primary effective components of S. baicalensis, we subjected seedlings to drought and salt stress and exogenous SA treatment in a laboratory setting and tested the expression of PAL and ICS, as well as the content of free SA (FSA), total SA (TSA), baicalin, and baicalein. We also assessed the correlation of FSA and TSA with PAL and ICS, and with baicalin and baicalein accumulation, respectively. The results indicated that both FSA and TSA were positively correlated with PAL, ICS, and baicalin, but negatively correlated with baicalein. The findings of this study improve our understanding of the manner in which SA regulates secondary metabolites in S. baicalensis.


Introduction
Plants are frequently challenged by a variety of stresses associated with drought, salt, and temperature. Salicylic acid (SA) can regulate the physiological processes of plants under a variety of stresses, thereby altering their resistance to these challenges [1,2]. During this regulatory process, SA induces the expression of many defense genes [3,4], which results in alterations in the accumulation of some secondary metabolites. One of the major functions of these secondary metabolites is to improve the tolerance of plants to stress [5]. The accumulation of several of these metabolites has been well studied in plants treated with SA in response to pathogen infection, and some have been functionally identified as antimicrobial compounds. For instance, SA treatment induced the synthesis of phenol-2,4-bis (1,1-dimethylethyl)  72h before harvesting. The roots of the seedlings were harvested, rinsed with ultrapure water, immediately frozen in liquid nitrogen, and stored at −80˚C until further use.

Total RNA extraction and reverse transcription
RNA was extracted from 0.01 g of root material. The root samples were homogenized with 400 μL extraction buffer (4 M guanidine isothiocyanate, 20 mM MES, 20 mM pH 7.0 EDTA, 50 mM β-mercaptoethanol) and the homogenate was centrifuged at 13,362 g at 4˚C for 5 min.
The supernatant was then transferred to a new tube and mixed with 500 μL chloroform-isopentanol and vortexed for 1 min. The mixture was centrifuged at 13,362 g and 4˚C for 15 min, and the supernatant was transferred to a new tube and combined with 100 μL 3 M sodium acetate buffer and 350 μL ethanol. The mixture was maintained at −20˚C for 30 min and then centrifuged at 13,362 g at 4˚C for 20 min. The precipitate was re-suspended with 100 μL 3 M sodium acetate buffer and the suspension was centrifuged at 13,362 g and 4˚C for 15 min. The precipitate was then washed with 800 μL 70% ethanol, dried at room temperature for 5 min, and then the dried precipitate was dissolved in 50 μL DEPC water. In order to reduce DNA contamination, total RNA was treated with the DNA-Be-Gone-A kit (kit no. B641312, Sangon Biotech Co., Ltd. Shanghai, China) according to the kit manual, and the RNA quantity and quality were determined using a NanoDrop 2000 (Thermo Fisher Scientific Inc., Wilmington, USA).
Forty nanograms of total RNA was mixed with 1 μL of 10 μM oligo dT primer and 1 μL of 10 mM dNTPs, and diluted with RNA-free water to 12 μL. The mixture was kept at 65˚C for 5 min and then rapidly cooled by placing on ice. Then, 4.8 μL reverse buffer, 1 μL RTase, and 0.6 μL RNase inhibitor were added to the mixture, and RNA-free water was added to yield a final volume of 24 μL. The mixture was kept at 42˚C for 45 min, and then 95˚C for 5 min.

Cloning full length of ICS cDNA
The full length ICS cDNA was cloned according to the partial ICS sequence in S. baicalensis [16] using RACE. 3'-and 5'-RACE was conducted using the primer sequences listed in Table 1. The annealing temperature was set to 68˚C. 5'-RACE was conducted with a kit according to the manual (Invitrogen, kit no. 18374-058).

Determination of gene expression levels
The relative expression levels of the PAL and ICS genes were determined by Real-Time (q) PCR (IQ2, Biorad Co., Ltd.) using the SYBR kit (kit No. 208056, Qiagen co., Ltd.). The primers CGAGTAGTGATTGGGTGATA and AGCCCTTGTTGCTGTATG were designed for PAL according to the sequence available on GenBank (HM062775.1). The primers TTGAAGGAAGTTCGA TGATG and ATGAGTCTGGTGAAGCATAA were designed for ICS according to the full-length ICS cDNA. The data were analyzed using the 2 −ΔΔCt method [17].

Enzyme extraction and activity determination
PAL activity was determined using High Performance Liquid Chromatography (HPLC) according to the modified method of Kovacik and Klejdus [18]. The samples were homogenized in borate buffer (pH 8.5) on ice, and the homogenates were immediately centrifuged at 20,879 g and 4˚C for 20 min. The reaction mixture consisting of 200 μL supernatant and 200 μL 20 mM L-phenylalanine was incubated at 40˚C for 1 h, and then 10 μL 6 M HCl was added. Parallel controls with HCl added prior to incubation were analyzed to determine the endogenous trans-cinnamic acid (CA) content. The HPLC assay was performed on an Agilent 1100 system equipped with a UV detector (G1314, Agilent, Santa Clara, CA, USA) and a 5 μm, 15 cm × 4.6 mm SUPELCOSIL™ LC-ABZ column (Sigma-Aldrich Co., St. Louis, MO, USA). PAL activity was expressed as CA content. The flow rate and elution gradient were determined as documented in Kovacik and Klejdus [18]. ICS activity was determined using a modified fluorescence analysis based on the method of Young and Gibson [19]. For the estimation of ICS activity, the samples were homogenized in extraction buffer (100 mM Tris-HCl at pH 7.

Determination of SA
SA content was determined with HPLC according to the method of Dewdney et al. [20]. Briefly, root tissue samples (0.5 g) were homogenized with 3 mL extraction solution (90% MeOH, 0.55 mM o-anisic acid), vortexed, sonicated for 20 min, and then centrifuged at 9,279 g and 4˚C for 20 min. The pellet was re-extracted with 2 mL 90% MeOH. The supernatants were divided into two equal-volume portions (for total SA and free SA measurements) and vacuum dried. For total SA samples, 500 μL β-glucosidase (80 U mL −1 in 100 mM sodium acetate pH 5.2, Sigma-Aldrich) was added to each sample. The samples were sonicated for 5 min, vortexed, covered with foil, and incubated for 90 min at 37˚C. For both the total and free SA samples, 2.5 mL 5% trichloroacetic acid was added, and the samples were vortexed, sonicated for 5 min, and centrifuged at 9,279 g for 15 min. The supernatant was extracted twice with 2.5 mL extraction solution (ethyl acetate: cyclopentane, 1: 1). The organic phases were vacuum dried. Just prior to loading the samples on the HPLC, each was re-suspended in 250 μL of 20% MeOH and sonicated for 5 min.
HPLC was performed on an Agilent 1100 system equipped with a fluorescence detector (G1321A, Agilent, Santa Clara, CA, USA) and a 5 μm, 15 cm × 4.6 mm SUPELCOSIL™ LC-ABZ Plus Column (Sigma-Aldrich Co., St. Louis, MO, USA) preceded by a LC-ABZ Plus guard column. The flow rate and elution gradient were as documented in Dewdney [20]. SA and o-anisic acid were quantified using the fluorescence detector programmed to 302 nm excitation and 412 nm emission wavelengths.

Determination of baicalin and baicalein
Baicalin and baicalein content were determined with HPLC according to the method of Su et al. [16]. Samples were diluted twice for injection and baicalin and baicalein were extracted from 0.1 g tissue. The roots were ground into a fine powder in liquid nitrogen and extracted with 500 μL ethanol. The homogenate was vortexed for 2 min and centrifuged at 13,362 g and 4˚C for 15 min, and the supernatant assayed by HPLC. The HPLC assay was performed on an Agilent 1100 system equipped with a UV detector (G1314, Agilent, Santa Clara, CA, USA) and a 5 μm, 15 cm × 4.6 mm SUPELCOSIL™ LC-ABZ Plus Column (Sigma-Aldrich Co., St. Louis, MO, USA). The wavelength was set to 270 nm, the column was maintained at 35˚C, and the injection volume was 5 μL. The flow rate and elution gradient were as documented in Su et al. [16]. The column was eluted with 27% acetonitrile for 3 min prior to the next injection.

Statistics
Results from representative experiments were expressed as mean ± SE of nine independent repeats. Data from the experiments were analyzed by t-tests for simple comparisons between each treatment and its control. The correlations were analyzed using Spearman's analysis. The assumptions of the analysis of variance were considered to be statistically significant at P < 0.05. All analyses were conducted in SPSS version 17.0.

Cloning of full-length ICS cDNA
The full-length ICS cDNA in S. baicalensis was 1,801 bp and the coding sequence was 1,398 bp. Nucleotide blast indicated that the sequence had 76% similarity with its orthologs in Solanum lycopersicum (GenBank: NM_001247865.2), Nicotiana benthamiana (GenBank: EU257505.1), and Capsicum annuum (GenBank: NM_001325078.1), and 75% similarity with Catharanthus roseus (GenBank: AJ006065.1). Blastp based on the non-redundant protein database revealed that the predicted ICS protein in S. baicalensis had high amino acid sequence identity to its orthologs and possessed a conserved chorismate binding site (Fig 2) belonging to the supergene family of chorismate binding proteins.

Relative expression levels of PAL and ICS
PAL gene expression differed between the drought and salt stress treatments. The expression was down-regulated by 23% under drought stress, but up-regulated by 12% under salt stress; though this latter finding was not statistically significant (Fig 3A). Drought stress exhibited a more significant effect on PAL gene expression than salt stress. The expression of ICS was down-regulated significantly by 67% under drought stress, but significantly up-regulated by 38% under salt stress (Fig 3A).
Treatment with 140 μM SA resulted in the up-regulated expression of PAL, while the 70 and 280 μM treatments were associated with down-regulated expression at 24 and 48 h, followed by up-regulation at 72 h. The lowest PAL expression levels were detected at 48 h, where a 46% decrease was observed at 70 μM and a 68% decrease was observed at 280 μM. A non-significant expressional increase of 16% was detected in the 140 μM treatment. PAL gene expression peaked at 72 h, increasing 2.5-fold in the 70 μM treatment, 2.98-fold in the 140 μM treatment (Fig 3B).
ICS gene expression was up-regulated after 140 μM SA treatment, but was down-regulated following treatment with 70 or 280 μM SA. Treatment with 140 μM SA resulted in 1.81-, 1.04-, and 2.41-fold increased expression at 24, 48, and 72 h, respectively. The lowest expression level in the 70 μM treatment was detected at 24 h and corresponded to a decrease of 99.4%, while the lowest level in the 280 μM was observed at 72 h and corresponded to a 66% decrease. A non-significant 1.04-fold increase at 48 h constituted the lowest expression level in the 140 μM treatment (Fig 3C).

PAL and ICS activity
PAL activity decreased by 14% under drought stress, though this was not statistically significant. Conversely, PAL activity increased significantly by 15% under salt stress (Fig 4A). ICS activity decreased significantly under drought stress (Fig 4A), while a significant increase of 53% was observed under salt stress (Fig 4A).
PAL activity was consistent with the gene expression data. Increased activity was observed after 140 μM SA treatment, with 6.67-, 1.83-, and 3.69-fold increases noted at 24 h, 48 h, and 72 h. As noted in the gene expression data, PAL activity decreased after 24 h and 48 h in the 70 μM and 280 μM treatments. The lowest activity was detected at 48 h, and corresponded to an 8.86% decrease at 70 μM and a 13.71% decrease at 280 μM, while the 1.83-fold increase observed in the 140 μM treatment was significant (Fig 4B).
ICS activity was broadly consistent with the gene expression patterns. Increased activity was associated with the 140 μM SA treatment, though the increase was only significant at 72 h. Conversely, the activity decreased after 70 μM or 280 μM treatment. The lowest activity in the 70 μM treatment was detected at 48 h and corresponded to a non-significant decrease of 3.57%, while a 94.81% decrease at 72 h was observed at 280 μM SA. A non-significant increase of 0.54% at 48 h was noted for the 140 μM treatment. Surprisingly, the ICS activity of the 70 μM treatment at 24 h did not corroborate the gene expression data (Fig 4C).

SA content
FSA decreased significantly to 86 ngÁg −1 under drought stress, but increased significantly to 245.88 ngÁg −1 under salt stress (Fig 5A). TSA content exhibited a similar pattern to FSA. The content decreased significantly by 43% to 228.23 ngÁg −1 under drought stress, but increased significantly by 55% to 549.6 ngÁg −1 under salt stress (Fig 5A).
FSA decreased in all the treatments, with the exception of a slight increase observed in the 140 μM treatment at 24 and 72 h. The lowest content in the 70 μM treatment was detected at 72 h and corresponded to a non-significant decrease of 14.28% to 266.67 ng/g, while a significant decrease of 26.78% to 227.78 ng/g was observed at 280 μM (Fig 5B).
TSA content decreased in most of the treatments, with the exception of the 70 μM treatment at 24 and 72 h; though this increase was non-significant. The lowest TSA content in the 70 μM treatment was detected at 48 h and corresponded to a 27.25% decrease to 305.56 ng/g. A decrease of 33.33% to 280 ng/g was observed at 48 h in the 140 μM treatment, while the lowest TSA content in the 280 μM treatment was noted at 72 h and corresponded to a 39.02% decrease to 189.03 ng/g (Fig 5C).

Baicalin and baicalein content
The baicalin content exhibited a sharp and significant decrease of 76.16% under drought stress to 31.28 μgÁg −1 . In contrast, under salt stress the content increased significantly by 23.77% to 156.36 μgÁg −1 (Fig 6A). Conversely, baicalein content increased sharply by 185% under drought stress to 2 μgÁg −1 . A statistically significant decrease by 31.81% to 0.45 μgÁg −1 was observed under salt stress (Fig 6A).
A sharp and significant increase in baicalin content was observed after 140 μM treatment, peaking at 1.01 mg/g at 24 h. A 20.05% increase to 261.31 μg/g baicalin occurred only at 72 h in the 70 μM treatment, while a 4.54% to 230.14 μg/g increase was observed at 24 h in the 280 μM treatment (Fig 6B).  (Fig 6C).

Correlation of SA with PAL, ICS, and metabolites
Under stress, both FSA and TSA were positively correlated with baicalin, though FSA exhibited a stronger and more significant correlation with baicalin in comparison to TSA. Both were significantly negatively correlated with baicalein ( Table 2).
Following exogenous SA treatment, FSA was significantly correlated with PAL. They were also positively correlated with baicalin and negatively correlated with baicalein, and FSA exhibited a stronger and more significant correlation with baicalin in comparison to TSA ( Table 3). The correlations of the exogenous SA treatment were broadly consistent with that of the stress treatment.

Discussion
SA is synthesized via the PAL and ICS pathways in plants and the levels are closely associated with the activity of these two enzymes under stress. PAL-and ICS-silenced S. baicalensis seedlings were constructed in a previous study, and the effect of gene silencing on SA synthesis and the influence of endogenous SA on baicalin and baicalein accumulation were assessed. The study found that PAL silencing significantly influenced FSA content, while ICS affected the TSA content, and furthermore, FSA significantly influenced baicalin content, but not baicalein content [16]. However, these associations required further investigation, prompting us to investigate the correlation of PAL and ICS with SA, and the correlation of SA with baicalin and baicalein. Interestingly, the correlations were broadly consistent across both the stress and exogenous SA treatments. Both FSA and TSA exhibited a strong positive correlation with PAL and ICS, respectively, and FSA was significantly correlated with baicalin (Table 4).
Both free and bound SA were significantly reduced in whole plants of A. thaliana treated with the PAL inhibitor 2-aminoindan-2-phosphonic acid (AIP) under pathogenic fungal stress [21]. Four isoforms of PAL exist in the Arabidopsis genome, and SA levels were found to decrease sharply in quadruple mutants of PAL [22]. Although the inhibition of PAL or mutation of PAL isoforms results in a reduction in SA, considerable amounts of SA still accumulated under stress in A. thaliana [21,22] and Pueraria thomsonii Benth. [23], implying that the ICS pathway is involved in SA synthesis. In comparison to the PAL quadruple mutant of A. thaliana, the ICS1 mutant (sid2) accumulated far less total SA in response to pathogen infection [22]. The ICS pathway was found to be the major contributor to SA in A. thaliana under O 3 exposure [24], and furthermore, was required for SA synthesis in plant defense [12]. Conversely, the PAL-and ICS-derived pathways contributed equally to pathogen-induced SA accumulation in soybean [25]. Our results indicated that both the PAL and ICS pathways are involved in SA synthesis in S. baicalensis, and SA was more strongly correlated with ICS than PAL under both drought and salt stress. Furthermore, drought stress resulted in a decrease in PAL and ICS activity accompanied by a decrease in SA, but under salt stress this was reversed, indicating that SA responds differently to different stresses.
As endogenous SA will be supplemented by the absorption of exogenous SA into the cell or organism, exogenous SA was applied to investigate its influence on the associated plant physiological processes. Interestingly, in a previous study, young pea seedlings grown from seed were soaked in 3 H-labeled SA solution prior to sowing, and the results suggested that the increased endogenous SA observed in the seedlings, including free and bound SA, was synthesized de novo rather than taken up by the plant [26]. Our results indicated that FSA and TSA content decreased at most sampling time points after exogenous SA treatment, which was accompanied by fluctuations in PAL or ICS activity. Additionally, FSA was strongly positively correlated with PAL. Exogenous SA treatment can thus influence PAL activity. In another study, PAL activity was found to be dependent on exogenous SA content in Taxus cells, and increased following treatment with SA [27]. Additionally, the expression of ICS increased in the pea seedlings in which SA was de novo synthesized [26]. In combination, these results suggest that the fluctuations in PAL or ICS under exogenous SA treatment regulated the de novo synthesis of endogenous SA, as opposed to the absorption of exogenous SA.
Secondary metabolites play an important role in the interaction of plants with various stressors, and endogenous SA is intimately linked with the accumulation of plant secondary metabolites. Camalexin, an indolic secondary metabolite in Arabidopsis, was significantly reduced in NahG plants upon pathogen infection, in which the content of SA was reduced due to the expression of the NahG gene encoding bacterial salicylate hydroxylase [28]. The majority of phenylpropanoids were found to be negatively correlated with SA in transgenic Populus, in which endogenous SA levels were constitutively elevated [11]. Our results demonstrated that SA was positively correlated with baicalin, but negatively correlated with baicalein. The correlations observed between the exogenous SA treatment and the gene silencing results strengthened when the seedlings were subjected to drought and salt stress, which is expected given that most secondary metabolites function in improving stress resistance in plants. Furthermore, the regulation of secondary metabolites by SA is necessary for the efficient provision of intermediates in plants to maintain physiological homeostasis.
In conclusion, PAL and ICS function differently in the synthesis of FSA and TSA, FSA was strongly positively correlated with PAL. Both FSA and TSA positively correlated with baicalin, but negatively with baicalein (Fig 7). The findings of this study improve our understanding of the manner in which SA regulates secondary metabolites in S. baicalensis. In addition, the results of SA regulation in baicalin and baicalein production may be useful for exploring new strategies to improve the production in S. baicalensis. Thus, it provides an example of secondary metabolism regulation engineering in herbal plants.