The effect of rhizosphere soil on the flavonoid metabolism in the roots of Tetrastigma hemsleyanum

Background: Tetrastigma hemsleyanum (T. hemsleyanum) is a precious and rare traditional Chinese medicinal plant. Flavonoid is its main medicinal ingredient. Wild T. hemsleyanum (W-TH) growing in Zhejiang Province, China is recognized as a medicinal material of “San Ye Qing” Dao-di herbs. The different origins and thus the contents of medicinal ingredients are the key criteria used to determine whether the medicinal materials are authentic Dao-di herbs. However, it is less known how the eco-environments of its specific producing area, especially microbial community in rhizosphere soil, affect the content of medicinal ingredients in “San Ye Qing”. In the present study, we determined the content of total flavonoids and the enzymatic activity of phenylalanine ammonia-lyase (PAL) in the roots of W-TH and artificially cultivated T. hemsleyanum (C-TH), as well as the nutrient compositions and metagenome in rhizosphere soil. The effects of the rhizosphere soil on the flavonoid metabolism in T. hemsleyanum were evaluated. Results: The contents of total flavonoids and the PAL activity were higher in W-TH root than in C-TH root. The contents of both available phosphorus and available potassium were higher in the rhizosphere soil of W-TH than in that of C-TH, while the contents of nitrogen and organic matter in the rhizosphere root of C-TH were higher than that of W-TH. Compared with the rhizosphere of C-TH, the abundances of genera Enterobacter, Serratia, Raoultella, Kluyvera, Comamonas, Acinetobacte, and Arthrobacter, and the pathways related to nitrogen metabolism, inositol phosphate metabolism, phosphinate metabolism, and phosphotransferase system in the rhizosphere of W-TH were significantly different. There existed differences in phenylpropanoid biosynthesis and phenylalanine metabolism in the rhizospheres of metabolism, and phosphotransferase system (PTS) were different in the rhizospheres of W-TH and C-TH. These results suggest that the contents of nitrogen content and available phosphorus in the soil may change the diversity and abundance of the above-mentioned microorganisms, and the corresponding nitrogen and phosphorus metabolic pathways would be different, which will further affect the growth and development of T. hemsleyanum.

W-TH and C-TH.
Conclusions: The contents of nitrogen and available phosphorus in the rhizosphere soil affect diversity abundance, nitrogen metabolism, phosphorus metabolism, and phenylalanine (Phe) anabolism of rhizosphere microorganisms. They may further affect Phe content and PAL activity for the synthesis of flavonoids in the root of T. hemsleyanum.

Background
Tetrastigma hemsleyanum Diels & Gilg ex Diels (T. hemsleyanum)is a medical plant belonging to the family Vitaceae andthe genus Tetrastigma. It is mainly distributed in the south of Yangtze River in China. It has strict demands for illumination and temperature in its habitat. Due to the excessive collection of the wild resources of T. hemsleyanum and the destruction of its suitable habitat, its population has been rapidly reduced. Thus, it has been listed as one of the endangered plants by both Zhejiang and Jiangxi Provinces in China [1][2][3].
T. hemsleyanum is rooted in Chinese medicine, and it, called "San Ye Qing" in Chinese, is one of the most important herbs in traditional Chinese medicine [4]. It has multiple medical values, including heat-clearing and toxin-detoxifying, phlegmremoving, blood circulation-promoting and pain-relieving, immunity-regulating functions, as well as anti-inflammatory and anti-virus effects [5,6]. Especially in recent years, T. hemsleyanum has been found to have anti-tumor effects [7][8][9][10].
Wild T. hemsleyanum (abbreviation as W-TH) growing in Zhejiang Province is recognized as a medicinal material of "San Ye Qing" Dao-di herbs [4], which has been a clinically preferred Chinese medicine for long time and is produced in a specific region with a specific production process. Dao-di herbs has higher quality and better curative effect compared with those of other medicinal materials [11,12] and its formation is a complex evolution process, involving multiple factors. The different origins and thus the content of medicinal ingredients are the key criteria used for determining whether medicinal materials are authentic Dao-di herbs [13][14][15]. Yuan et al. (2015) reported that in addition to traditional genetics, epigenetics also plays an important role in formation of Dao-di herbs. They proposed epigenetic mechanism in the study of Dao-di herbs formation from specific phenotype and regional analysis [16].  deemed that Dao-di herbs had been recognized as "quality models" with a high status. The advancement of various omics technologies has provided new methods for the analysis of complex biological systems, which are also suitable for studying the quality formation in Dao-di herbs as well. With achievements of omics in the study of Dao-di herbs from the genetics to phenotyping, the use of these new methods of quality evaluation can be investigated in the biosynthetic pathways of secondary metabolites and the interaction with human body [17]. Hao et al. (2019) reported that as a treasure of traditional Chinese medicine, Dao-di herbs are famous for their high quality and good effect. However, traditional characterization of Dao-di herbs and their producing areas is mostly confined to qualitative description but lacks the objective evaluation indicators [18]. Specially, Fu et al. (2019) investigated the effects of different nitrogen levels on the growth of T. hemsleyanum and the content of phytochemicals and antioxidant activity in its stems and leaves. They found that a certain amount of nitrogen had positive effects on most of the biological traits but excessive dose of nitrogen went against growth of T. hemsleyanum. With the increase in nitrogen levels, the polysaccharide contents in stems and leaves were not significantly changed, while the contents of total flavonoids and phenolic components, and antioxidant activities were increased steadily. Antioxidant activities and contents of total flavonoid and phenolic components showed a significant and positive correlation [19].
W-TH growing in Zhejiang Province is the authentic Dao-di herbs of "San Ye Qing". However, it is less known how the eco-environments in its specific producing area, especially microbial community in the rhizosphere soil affect the content of medicinal ingredients in "San Ye Qing". In the present study, the metagenome in the rhizosphere soil of W-TH and the artificially cultivated T. hemsleyanum The measurements of flavonoid content, PAL activity and nutrient content showed that (1) the content of total flavonoids in W-TH roots was more than two times that of C-TH ( Fig. 1), (2) the PAL activity of W-TH roots was higher than that of C-TH ( Fig.   2) and (3) the contents of available phosphorus and available potassium in root soil of W-TH were higher than that of C-TH, while the contents of nitrogen and organic matter were higher in C-TH than in W-TH (Table 1). There were significant differences in root flavonoid content, PAL activity in root and contents of soil nutrient between W-TH and C-TH.
Sequencing results of rhizosphere soil metagenome, relative abundance and analysis of microbial organisms at genus level The originally down-loaded data of sequencing and the assembled scafigs were provided in additional files 1 and 2: Tables S1 and S2. The sequencing results meet the requirements for analysis. For the assembled scafigs, the fragments with fewer than 500 bp were filtered out, and MetaGeneMark was used for Open Reading Frame (ORF) prediction, which was obtained for gene catalogue (Unigenes). The Unigenes were compared with the bacterial, fungal, archaea, and viral sequences extracted from the NCBI nr database to obtain species annotation information for W-TH and C-TH (additional files 3: Table S3). At the genus level, the abundances of the genera including Burkholderia, Enterobacter, Bradyrhizobium, Paraburkholderia, Acinetobacter, Arthrobacter, etc were determined. There were significant differences in community diversity distribution between W-TH and C-TH. A histogram showing the relative abundances of the top 10 genera were drawn (Fig.   3). Analysis of variance on W-TH vs. C-TH was performed using DEGseq software.
The screening results were corrected using the Benjamini and Hochberg method (BH). Compared with those in C-TH, the genera Enterobacter, Serratia, Raoultella, Kluyvera, Comamonas, Acinetobacte, and Arthrobacter in W-TH were more abundant.
The significantly up-regulated profiles in the top 10 genera were drawn as a bar graph (Fig. 4). The color is displayed according to the logFC size, and the volcano of the differential genera on W-TH vs. C-TH was shown in additional files 4: Fig. S1.
According to the consistency of the genera shown in Figs Analysis of microbial metabolism pathways Three levels are set for differential pathway levels: level 1 is the biological metabolic pathway, level 2 is the sub-function of the biological metabolic pathway, and level 3 is the detailed metabolic pathway of the sub-function. The compression Fisher test between W-TH and C-TH revealed that there were 3 significantly differential pathways at level 1 (additional files 5: Table S4), 19 significantly differential pathways at level 2 (additional files 6: Table S5), and 103 significantly differential pathways at level 3 ( Table 2 and additional files 7: Table S6). The differential pathways in level 1 between W-TH vs C-TH are mainly related to genetic information processing and environmental information processing. A large number of sub-pathways related to metabolism pathways at level 2, including glycan biosynthesis and metabolism, folding, sorting and degradation, metabolism of cofactors and vitamins, nucleotide metabolism; amino acid metabolism; and lipid metabolism etc. Moreover, the specific metabolic pathways with specific functions were further explored. At level 3, the significantly differential pathways included the following: inositol phosphate metabolism (phosphonate and phosphinate metabolism), tropane, piperidine and pyridine alkaloid biosynthesis, phenylpropanoid biosynthesis, nitrogen metabolism, phenylalanine metabolism, and phosphotransferase system (PTS), etc.

Discussion
Flavonoids are one group of the main medicinal ingredients of herbs "San Ye Qing". In the study, we observed that the contents of nitrogen and organic matter were higher in C-TH root soil than in W-TH root soil, while the contents of available phosphorus and available potassium were higher in W-TH root soil than in C-TH root soil. Furthermore, the results of metagenomic analysis showed that the abundances of genera Enterobacter, Serratia, Raoultella, Kluyvera, Comamonas, Acinetobacte, and Arthrobacter in the root soil of W-TH were different from those in C-TH. The pathways related to nitrogen and phosphorus metabolism were also different, i.e. nitrogen metabolism, inositol phosphate metabolism, phosphonate and phosphinate metabolism, and phosphotransferase system (PTS) were different in the rhizospheres of W-TH and C-TH. These results suggest that the contents of nitrogen content and available phosphorus in the soil may change the diversity and abundance of the above-mentioned microorganisms, and the corresponding nitrogen and phosphorus metabolic pathways would be different, which will further affect the growth and development of T. hemsleyanum.
It is worth noting that comparing W-TH with C-TH, there exist differences in the bacterial genera and secondary metabolism, including tropane, piperidine and pyridine alkaloid biosynthesis, phenylpropanoid biosynthesis, phenylalanine metabolism, especially the phenylalanine metabolism. It is known that phenylalanine (Phe) is the starting point of phytoflavonoid biosynthesis, and PAL is the first key enzyme of flavonoid synthesis. PAL catalyzes the formation of cinnamic acid and coumaric acid from Phe and is the key to the connection of phenylpropane compounds and primary metabolism. This enzyme plays an important role in regulating the biosynthesis of flavonoids [24]. It has been shown that the exogenous addition of appropriate amount of Phe can promote the plant to synthesize more secondary metabolite anthocyanins, and also restore the mutant phenotype with low contents of anthocyanin. Phe plays an important role in the secondary metabolism of anthocyanin synthesis [25]. The PAL activity in root of W-TH is higher than that in C-TH, while the "mother" plant of C-TH is derived from W-TH, which has the same genetic background of W-TH. Considering that the physiological features of C-TH and W-TH growth are similar, only the root soils for their growth are different, but the flavonoid content in W-TH is significantly higher than that in C-TH. This suggests that the contents of nitrogen and phosphorus in the soil affect the microbial abundance, the metabolisms of nitrogen and phosphorus, and the anabolism of phenylalanine. These factors further affect the Phe content and PAL activity in the root of T. hemsleyanum, which, in turn, affects the anabolism of flavonoids. However, how do the bacterial genera present in root soil affect the Phe content and PAL activity of T. hemsleyanum still needs to be explored. In addition, it is unclear whether soil pH and available potassium affect the anabolism of flavonoids in T. hemsleyanum and this aspect needs to be addressed as well.

Metagenome analysis
The raw data obtained by sequencing were filtered to obtain clean data, which were then assembled with scaftigs [28][29][30]. Gene prediction was performed using MetaGeneMark [31,32] to construct gene catalogue (Unigenes). From the gene catalogue, the clean data of each sample were combined to obtain the information about abundance of the gene catalogue in each sample. Unigenes was compared to microNR database in NCBI using DIAMOND software [33], and species annotation information for this sequence was further determined using the LCA algorithm [34].
From the LCA annotated results and the gene abundance table, the abundance information of each sample at each classification level (e.g. genus and species) was obtained. Abundance histogram display, principal component analysis (PCA) and significant difference analysis were performed. To determine whether there was a significant difference in the genus level between the samples, DEGseq [35] was used for differential analysis, and corrected by Benjamini and Hochberg method (BH) (q≤ 0.001, logFC≥ 1), with a focus on up-regulated bacterial group (at genus level). From the gene catalogue, KEGG, eggNOG and CAZy analysis were performed [36][37][38], and a Fisher test (p< = 0.05) between samples for differential functional abundance was performed. Cluster analysis and metabolic pathway analysis were also performed.

Statistical analysis
The total flavonoids in root of T. hemsleyanum and the nutrient composition in root soil were measured three times, and statistical analysis was performed using SPSS software (version 19.0). The difference between groups with p<0.05 was regarded statistically significant. The experimental data were expressed as the mean value + standard deviation (SD) of the results of repeated experiments. Three rhizosphere soil samples of W-TH and C-TH were collected respectively. After being mixed well, the soil samples were used to determine the respective metagenomic groups of W-TH and C-TH.

Consent for publication
Not applicable.

Availability of data and material
All the data generated or analyzed during this study are included in this published article and its supplementary information files.

Competing interests
The authors declare that they have no competing interests.

Funding
The project was supported by grants from the National Natural Science Foundation of China (Grant No. 31872181). The funding body did not play any roles in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript.    Figure 1 Contents of total flavonoids in roots of W-TH and C-TH.

Figure 1
Contents of total flavonoids in roots of W-TH and C-TH.   Relative abundance distribution profiles of the top 10 genera between W-TH and C-TH.

Figure 3
Relative abundance distribution profiles of the top 10 genera between W-TH and C-TH.

Figure 4
The significantly up-regulated profiles of the top 10 genera by W-TH vs. C-TH.