Combined Metabolome and Transcriptome Analyses Reveal the Flavonoids Changes and Biosynthesis Mechanisms in Different Organs of Hibiseu manihot L.

Hibiseu manihot L. (Jinhuakui, JHK), also known as a garden landscape plant, is widely cultivated as a landscape plant having pharmacological effects due to its high flavonoids content. Although flavonoids were the main active pharmaceutical ingredients in JHK, little information was obtained about the content, composition, and accumulation pattern of flavonoids in different tissues. Most studies only identified a few kinds of flavonoids in JHK limited by separation and identification problems. Therefore, combined metabolome and transcriptome analysis was performed to explore the accumulation patterns and biosynthesis mechanisms of flavonoids in JHK. In this study, we identified 160 flavonoids in 15 samples of JHK (flower, leaf, root, stem, and seeds) by using LC-MS/MS. Consistent with the total flavonoid content determination, these flavonoids were significantly accumulated in flowers, followed by leaves, stems, roots, and seeds. Among them, certain flavonoids, with high content, were also identified for the first time in JHK, such as tricetin, catechin, hesperidin, ncyanidin-3-O-sambubioside, astragalin, procyanidin B2/B3/C1, apigenin-5-O-glucoside, etc. Different tissues underwent significantly reprogramming of their transcriptomes and metabolites changes in JHK, particularly in the flavonoid, flavone, and flavonol biosynthesis pathways. We conducted a correlation analysis between RNA-seq and LC-MS/MS to identify the key genes and related flavonoids compounds, rebuild the gene-metabolites regulatory subnetworks, and then identified 15 key genes highly related to flavonoids accumulation in JHK. These key genes might play a fine regulatory role in flavonoids biosynthesis by affecting the gene expression level in different organs of JHK. Our results could be helpful for the improvement of the market/industrial utilization value of different parts of JHK, to pave the way for the regulatory mechanism research of flavonoids biosynthesis, and provide insight for studying the production quality improvement of JHK.


INTRODUCTION
Hibiseu manihot L., one of the endangered plants from the family of Malvaceae and the genus of Abelmoschus, is recorded as "Jinhuakui" (JHK) in Chinese, and distributed in the Hebei province of China (Cao and Miao, 2016). The plant shape of JHK is like okra, with large flowers and golden corolla, which can be eaten directly, and has high ornamental value and nutritional value among more than 200 okra plants (Liu, 2008;Peng and Wu, 2008). Therefore, JHK can be used for the development of landscape plants, medicinal plants, and healthy food. Previous studies have shown that JHK has high medicinal value, and the whole plant, including flowers, leaves, roots, stems, and seeds, exhibit wide pharmacological activities (Peng and Wu, 2008;Lan et al., 2012) and can be used as medicine in China. The flowers, fruits, and seeds of JHK have multiple pharmacological effects, which are effective in relieving pain and reducing inflammation as well as applied it to treat traumatic injuries or sprains (Chen L.M. et al., 2016). Pharmacological research studies have shown that the extract of JHK has been used in clinical practice as a remedy for certain diseases, and it also possesses a good therapeutic effect in antioxidation, anticonvulsant, antiinflammation, and immune regulation (Lu and Jia, 2015). The extracts of JHK (flowers, leaves, stems, roots, and seeds) have potential biological active constituents that are responsible for inhibition of inflammation effects or wound healing bioactivity (Zhang et al., 2019). Therefore, studying the compounds (active ingredients or secondary metabolites) of JHK could serve as potential resources for the development of health products that possessed bioactive properties.
Until now, the previous research on the JHK was mainly focused on the optimization of the cultivation mode (Fan et al., 2020), exploring seed germination conditions (Yu, 2021), optimization of extraction method of total flavonoids in JHK (Cui et al., 2020), and a limited number of flavonoids and polyphenol identification (Tsumbu et al., 2012;Wang M. et al., 2021), such as hyperoside, caffeic acid, chlorogenic acid, rosmarinic acid, rutin, and tannins. Although there are certain studies on flavonoids in JHK, most of them only studied its extraction, detection, and preliminary activity analysis. Most of the studies are mainly on the phytochemical constituents and biological activities of the relative species of Abelmoschus manihot (Huangshukui in Chinese) (Lai et al., 2009;Pan X. et al., 2017). However, limited information on the metabolic compound (chemical constituents) and biological activities, especially the types of flavonoids in JHK still remained unknown. It is known that the vegetation stage of plants together with environmental conditions may affect changes and the formation of phenolic and flavonoids compounds (Sytar et al., 2013(Sytar et al., , 2014Li et al., 2020). Furthermore, the dynamic change and distribution differences of these flavonoids in the flowers, leaves, roots, stems, and seeds of JHK were also unclear.
In recent years, metabolomics combined with transcriptomics (RNA-seq) provides a powerful tool for the investigation and characterization of postgenomic processes and the molecular basis in many plants (Morgenthal et al., 2006). The metabolomics and RNA-seq profiles have been analyzed in many plants to expose the mechanism of metabolic compounds synthesis and dynamic changes in different plant species, including Abelmoschus manihot (Pan X. et al., 2017), tea plants , jujube , passion fruit (Li C. et al., 2021), Sophora alopecuroides (Zhu et al., 2021), longan , etc. Therefore, to systematically study the types, distribution, and dynamic changes of flavonoids in different tissues of JHK, in the present study, metabolomics combined with transcriptome analysis was performed in different tissues (flowers, leaves, stems, roots, and seeds) of JHK. The distribution and content ratio of flavonoids in different tissue of JHK were systematically studied and compared, and the mechanism of flavonoid synthesis and related key genes were also identified. This study will pave the way for the research on the active constituents and pharmacology of JHK, and provide the basis and reference for further study on the mechanism of flavonoids synthesis.

Plant Materials and Treatments
The flowers, leaves, stems, roots, and seeds of the JHK were used in this study (Figure 1). All flowers, leaves, stems, roots, and seeds samples were collected from the Shimen Town, Antu County, Yanbian Korean Autonomous Prefecture, Jilin Province (129.02 • E, 43.03 • N), which were planted at the end of May 2020 and sampling was done at the end of August 2020. The average annual temperature is around 3-5 • C, with the coldest January averaging minus 18 • C and the hottest July averaging around 20 • C. The flowers, leaves, stems, roots, and seeds of JHK were named F, L, St, R, and Se, respectively. Five plants with the same (or close) growth were selected as one repeat, with a total of three biological replicates in each group. The flowers, leaves, stems, roots, and seeds of the JHK were sampled separately, and the soil at the roots was rinsed repeatedly with distilled water. The surfaces of leaves, seeds, and flowers of JHK were also rinsed with distilled water. All these samples were quickly frozen in liquid nitrogen containers and brought back to the laboratory, and stored at −80 • C refrigerators for further experimental analysis.

Determination of Total Flavonoids Content
The tissues of JHK were ground in liquid nitrogen using a high-speed multifunction pulverizer (Zhejiang, China). The sample powder (5.0 g) was weighed and extracted by ultrasonic method (40 kHz) under the conditions of ethanol concentration of 80%, extraction time of 30 min, extraction temperature of 60 • C, and the solid-liquid ratio of 1:20 g/ml. The filtrate was then centrifuged and transferred to a 50 ml volumetric flask after rotary evaporation by RE 52-99 rotary evaporator (Shanghai, China). Then, 80% ethanol solution was used for constant volume, and the filtrate was shaken well for reserve use. The sampling solution was used as the blank control and the experimental group (flowers, leaves, stems, roots, and seeds) was taken in colorimetric tubes, and then 80% ethanol solution was added to 5.0 ml successively. Then 5.0 ml of 80% ethanol solution was added to the sample calibration tube to make the volume constant at 10 ml. Blank control and sample solution were added to 0.3 ml 5% Na 2 NO 2 solution, shaken well, and left for 6 min. Then 0.3 ml was added to 10% Al (NO 3 ) 3 solution, shaken well, and left for 6 min. Then 4.0 ml of 1 mol/L NaOH solution, and 0.4 ml deionized water were added, shaken well, and left to stand for 15 min. The absorbance (OD) was determined by spectrophotometer at 510 nm (7230G Visible Spectrophotometer, Shanghai, China), and the content of total flavonoids was calculated by standard curve. Rutin was used as a reference standard, and results were expressed as rutin equivalents in mg/kg extract.
Extraction rate (%) = (flavonoid content in the extract/ weight of the sample used) × 100%

Sample Preparation for UPLC-MS/MS
The sample was placed in a lyophilized machine (SCIENTZ-100F) for vacuum freeze-drying. Then, the freeze-dried samples were ground to a powder using a mixer mill (MM 400, Retsch) with a zirconia bead for 1.5 min at 30 Hz. A total of 100 mg of powder was weighed and dissolved in 1.2 ml 70% aqueous methanol (v/v) for extraction. The samples were vortexed every 30 min for 30 s for 6 total vortexes and placed in a refrigerator overnight at 4 • C. Then, the extracted samples were centrifuged (10,000 g, 10 min), absorbing supernatant for filtration by Millipore filtration system (0.22 µm pore size, ANPEL, Shanghai, China) and stored in the injection bottle for UPLC-MS/MS analysis.
The mass spectrometry analysis was followed based on the method of Li S. et al. (2021) with slight modifications. LIT and triple quadpole (QQQ) scans were obtained on a triple quadpole linear ion trap mass spectrometer (Q Trap), AB4500 Q Trap UPLC/MS/MS system equipped with an ESI turbo ionspray interface. It can be controlled by analyst 1.6.3 software (AB SCIEX) to run both positive and negative ion modes. The ESI source operation parameters are as the follows: ion source turbo spray; source temperature 550 • C; ion spray voltage (IS) 5,500 V (Positive ion mode)/−4,500 V (negative ion mode); ion source gas I (GSI), gas II (GSII), and curtain gas (CUR) set at 50, 60, and 25.0 psi, respectively; and high collision gas (CAD). The instrument was tuned and calibrated with 10 and 100 µmol/L polypropylene glycol solution in QQQ and LIT modes, respectively. The QQQ scans were acquired as multiple reaction monitoring (MRM) experiments with the collision gas (nitrogen) set to medium. Through further optimization of DP and CE, the DP and CE of each MRM ion pair were completed. A specific set of MRM ion pairs was monitored in each period based on the metabolites eluted in each period. , and roots (R). The asterisks (** and *) above each bar represent a significant difference at p < 0.01 and p < 0.05, respectively. metabolites were characterized according to the second-order spectral information. During the analysis, isotope signals, repeated signals containing K+ ions, Na+ ions, NH4+ ions, and fragments of other substances with larger molecular weight were removed. MRM analysis of QQQ mass spectrometry was performed to quantitative analysis of metabolites. In the MRM mode, the quadrupole firstly screened the precursor ions (parent ions) of the target substance and excluded the related ions of other molecular weight metabolites to preliminarily eliminate the interference. After induced ionization in the collision chamber, the precursor ions were fractured to form a lot of fragments, and then the fragments were filtered through the triple four-bar filter to select a characteristic fragment ion needed to eliminate the interference of non-target ions so that the quantification is more accurate and the reproducibility is better. Finally, the mass spectrum peak area was used to determine the relative metabolite contents (Fraga et al., 2010).

Qualitative and Quantitative Analyses of Metabolites
The filtered metabolite data were analyzed by Analyst 1.6.1 software for orthogonal partial least squares-discriminant analysis (OPLS-DA) and unsupervised principal component analysis (PCA). The relative importance of each metabolite to the OPLS-DA model was checked using the parameter called variable importance in projection (VIP). Hierarchical clustering analysis of the metabolites between each sample was performed using R software. 12 The screening criteria, | log2(fold change) | ≥ 1 and VIP ≥ 1, was used to identify differentially accumulated metabolites (DAMs) in this study.

RNA-Seq Analysis
The total RNA of JHK (F, L, St, R, and Se) was extracted using the FastPure plant total RNA isolation kit (Vazyme, RC401) according to the manufacturer's protocol. Fifteen sequencing libraries were generated using NEBNext Ultra RNA Library Prep Kit for Illumina (NEB, United States) following the manufacturer's recommendations and index codes were added to attribute sequences of each sample. Divalent cations under elevated temperature in NEBNext First Strand Synthesis Reaction Buffer (5×) were used for fragmentation. Random hexamer primer and M-MuLV reverse transcriptase were used to synthesize first-strand cDNA. Second-strand cDNA synthesis was subsequently carried out using DNA Polymerase I and RNase H. TTo select cDNA fragments of preferentially 240 bp in length, the library fragments were purified with the AMPure XP system (Beckman Coulter, Beverly, MA, United States). The clustering of the index-coded samples was performed on a cBot Cluster Generation System using TruSeq PE Cluster Kit v3-cBot-HS (Illumia) according to the manufacturer's instructions. After cluster generation, the library preparations were sequenced on an Illumina Hiseq 2000 platform, and the paired-end reads were generated. The sequences were further processed with a bioinformatic pipeline tool, BMKCloud 13 online platform. Raw data (raw reads) of fastq format were firstly processed through in-house perl scripts. After removing reads containing adapter, reads containing ploy-N and low-quality reads from raw data and clean data (clean reads) were obtained. Clean reads were assembled into expressed sequence tag clusters (contigs) and de novo assembled into the transcript by using Trinity. At the same time, Q20, Q30, GC-content, and sequence duplication levels were calculated in the clean data. All the downstream analyses were based on clean data with high quality. The function of the genes was annotated based on the databases, namely, NR (NCBI non-redundant protein sequences), KOG/eggNOG (clusters of orthologous groups of proteins), Pfam (protein family), SwissProt (a manually annotated and reviewed protein sequence database), GO (gene ontology), KEGG (Kyoto Encyclopedia of Genes and Genomes). Differentially expressed genes (DEGs) were identified using the DESeq functions estimate size factors and the nbinom test. A value of p < 0.05 and fold change > 2 or fold change < 0.5 was set as the threshold for significantly differential expression.

Quantitative Real-Time PCR Analysis
Twenty DEGs in the flavonoid synthesis pathway were selected for quantitative real-time PCR (qRT-PCR) analysis. The primers were designed using Primer 3 14 and they are listed in Supplementary Table 1. The RNA was extracted from JHK and was used to synthesize first-strand cDNA by using TransScript First-Strand cDNA Synthesis SuperMix (TransGene, AT301-02) following the manufacturer's instructions. qRT-PCR was performed using the SYBR Green PCR kit (Qiagen, 204054) according to the manufacturer's instructions. In this study, all the genes were repeated in three biological replicates (each biological replicate contains 3 technical replicates). The 2 − Ct method was used to calculate the mRNA expression level of genes. The relative gene expression level and FPKM were normalized by using log2 (fold change) measurements. The R software package version 3.1.3 15 was performed to analyze the correlation between RNA-seq and qRT-PCR data.

Determination of Total Flavonoids Content in Hibiseu manihot L.
Flavonoids are the main metabolites and medicinal ingredients in the JHK plant. Therefore, the total flavonoids content in different tissues (flowers, leaves, stems, and roots) of JHK was firstly detected in this experiment. The results showed that the content of total flavonoids in flowers was significantly higher than that in leaves, stems, and roots (p < 0.01). Among them, the total flavonoids content in flowers of JHK was 44.46% higher than that in leaves, while the total flavonoids content in roots and stems was very low (Figure 1F). Therefore, to explore the mechanism of the changes in flavonoid accumulation in different tissues of JHK, the flowers, leaves, stems, roots, and seeds of JHK were studied in the present experiment by using LC-MS/MS and RNA-seq.
Principal Component Analysis and Sample Correlation Analysis of Hibiseu manihot L.
Firstly, we carried out a quality control (QC) analysis on the metabolite detection results of these samples. The superposition diagram of the total ion chromatogram (TIC) detected by the QC sample essential spectrum is shown in Supplementary Figure 1. The results showed that the curves of the total ion flow detected by the metabolites had a high overlap in JHK, that is, the retention time and peak intensity were consistent, which indicated that the signal was stable when the same sample was detected at different times by mass spectrometry. The high stability of the instrument provides an important guarantee for the repeatability and reliability of the data, where N represents the negative ion mode and P represents the positive ion mode. Therefore, the results indicated that the metabolite detection in this study is reliable and can be used for subsequent analysis.
Principal component analysis showed that the first principal component (PC1) could explain 44.64% of the total variance and distinguish samples based on the different tissues/organs of JHK (Figure 2A). The flower (T2 or F) was separated from the leaf (L), stem (St), seed (Se), and root (R) of JHK. There was a significant difference between the flavonoids in flowers and other tissue parts, which was consistent with the results of total flavonoids detection in the front section ( Figure 1F). The second principal component (PC2) could explain 19.73% of the total variance and separate organs of L from others (F, St, Se, and R) in JHK (Figure 2A). It can be seen from PC2 that the flavonoids in the leaves of JHK were also significantly different from those in other tissue parts. The results of this experiment preliminarily concluded that the differences of flavonoids in flowers and leaves of JHK were significant compared with other tissue parts, while the differences of flavonoids in roots, stems, and seeds were not particularly significant.
The heat map of sample correlations (sample to sample clustering) showed that metabolites accumulation values among 15 samples of JHK were reproducible between the three biological replicates and batch effects were controlled ( Figure 2B). Pearson's correlation coefficient r was used as the evaluation index of biological repeat correlation. Pearson correlation coefficient is calculated by using the built-in cor function of R software. The r 2 is closer to 1, the stronger correlation between the two repeated samples. At the same time, the higher the correlation coefficient between the samples within groups and the samples in different groups, the more reliable the differential metabolites were obtained. The present results showed that the reproducibility of different tissue parts of JHK was good and the experimental results were reliable.

Analysis of Metabolites of Flavonoids in
Hibiseu manihot L.
A total of 160 flavonoids were detected by LC-MS/MS in JHK (Additional file 1). According to hierarchical cluster analysis (heat map Figure 3A), seed (Se) and root (R) were clustered into one cluster firstly, then cluster with stem (St) and leaf (L), while R, St, L, and Se were all clustered in different branches from the flower (F or T2) of JHK. The distribution and relative content of flavonoids in these tissues were also significantly different. It can be preliminarily seen from Figure 3B that the content of flavonoids in flowers is significantly higher than that in other tissues, followed by leaves and stems. In addition, the flowers, leaves, stems, seeds, and roots of JHK contained their own unique and significantly accumulated flavonoids (Table 1). Therefore, we conducted a detailed analysis of these 160 flavonoids identified in the present study. First of all, we classified the 160 flavonoids (Figure 3B), and these flavonoids mainly included anthocyanins (12.50%), chalcones (1.88%), dihydroflavone (2.50%), dihydroflavonol (2.50%), flavanols (4.38%), flavonoid (24.38%), flavonoid carbonoside (1.88%), flavonols (42.50%), isoflavones (0.63%), proanthocyanidins (2.50%), and tannin (4.38%). The results showed that these flavonoids may constitute the main pharmacological activity fraction of JHK. Among them, the content and types of flavonoids in flowers were the highest than other tissues.

Analysis of the Differential Accumulated Flavonoids in Organs of Hibiseu manihot L.
We conducted further analysis of the differential accumulated flavonoids (DAFs) in different comparison groups (Supplementary Figure 2). A total of 144 DAFs were found and identified in JHK. Among them, 20 DAFs were downregulated and 82 DAFs were significantly accumulated in the L vs. F comparison group. In the R vs. F comparison group, 5 DAFs were downregulated and 114 DAFs were significantly accumulated. In the Se vs. F comparison group, only 1 DAF was downregulated and 121 DAFs were significantly accumulated. In the St vs. F comparison group, 6 DAFs were downregulated and 92 DAFs were significantly accumulated. The results showed that most of the DAFs were significantly accumulated in flowers of JHK, which was consistent with the results of PCA and total flavonoids detection.
We then performed a Venn analysis for the accumulation of DAFs in these comparison groups of JHK. The results showed that there was a total of 60 DAFs in different comparison groups (Figure 4A), which included flavonols (29), flavonoids (11), anthocyanins (7), and tannin (4). Figure 4B shows the content and accumulation of 60 common DAFs in different tissue parts of JHK. The results showed that most of these 60 DAFs were significantly accumulated in the flowers and partly in the stems as well. Compared with other tissue, two proanthocyanidins (procyanidin B2 and procyanidin B3) were only found significantly accumulated in the roots, whereas no specific significant accumulation DAFs were found in the seeds. Then we analyzed the specific (with high content) of DAFs in each component in different parts of JHK. Thirty-three flavonoids were identified with high concentrations only in the flowers, while these DAFs were not or at extremely low concentrations in leaves, roots, stems, and seeds ( Table 1). These 33 DAFs were mainly including flavonols, flavonoid, and anthocyanins, such as hesperetin 5-O-glucoside, keracyanin, tamarixin, laricitrin, quercetagetin, etc. were only present in flowers of HJK. There were four unique accumulation DAFs in roots and leaves, respectively. Pelargonidin-3-O-rutinoside in roots, cyanidin-3-O-arabinoside, apigenin-5-O-glucoside, kaempferol-3-O-rutinoside-7-O-glucoside in leaves. And only one specific accumulation of DAFs in seeds [quercetin-3-O-(6 -p-coumaroyl) galactoside]. However, no specific accumulation of DAFs was found in stems ( Table 1). These results indicated that the specific accumulation and distribution pattern of these DAFs was the main reason for the difference of flavonoids in different tissue parts of JHK. Therefore, to further study the mechanism of changes in flavonoids at different tissue parts of JHK, we conducted transcriptome sequencing (RNA-seq) on these groups to further explore the possible pathways/mechanism that causes flavonoid accumulation differences in JHK.

Transcriptome Sequencing (RNA-Seq) Analysis
Transcriptome sequencing was performed on 15 samples from different tissue parts of JHK. A total of 97.64 Gb of clean data was obtained through sequencing QC, and the clean data of each sample reached 5.81 Gb. The GC content was 42.98-44.84%, and the percentage of Q30 bases was >92.73% ( Table 2). The results showed that the RNA-Seq profile was reliable. A total of 52,990 unigenes were obtained after assembling clean reads, among which 17,417 unigenes with lengths over 1 kb were obtained. Then we annotated the functions of these unigenes, including comparisons with NR, Swiss-Prot, KEGG, COG, KOG, GO, and PFAM databases, and obtained 39,651 unigenes annotation results. The annotated unigenes were analyzed for significant enrichment among different comparison groups, and the correlation analysis was carried out in combination with metabolomics (LC-MS/MS).

Combined Transcriptome and
Metabolome Analysis of Hibiseu manihot L.  Figure 3). Then we furtherly analyzed the genes and related metabolites significantly enriched in these pathways.
Two of these genes (1 C3 H and 1 HCT) were downregulated in the F of JHK as compared to R and St, respectively. The expression level of 6 genes (LAR, DFR, ANR, CYP73A, and E5.5.1.6) did not shown any difference in L and Se. As a result, two coding genes of LAR were identified with no significant difference in expression between L and Se, while DFR, ANR, CYP73A and E5.5.1.6 were identified with only one coding gene with no significant difference between L and Se ( Figure 5). However, 23 key genes were significantly upregulated in F as compared with L, St, R, and Se of JHK, respectively. These genes were involved in flavonoids synthesis in JHK. In addition, three phenylalanine ammonia-lyase (PAL) genes (c96791.graph_c0, c103354.graph_c0, and c79103.graph_c1) were found in our transcriptome profile, and they were all significantly upregulated in F. In JHK, PAL converts phenylalanine to cinnamoyl-CoA, which is then converted to p-coumaroyl-CoA by CYP73A enzyme catalysis. After some enzymes (CHS, E5.5.1.6, DFR, ANR, HCT, etc.) catalyzed the final synthesis of flavonoid  compounds (Figure 5), the results of differentially accumulated flavonoids were consistent with the associated gene expression results ( Figure 6A). All the 17 DAFs significantly enriched in flavonoids synthesis pathway were accumulated in F compared with L, Se, R, and St of JHK. Among them, seven flavonoids were significantly associated with key enzyme genes of flavonoid synthesis, and their contents were regulated by the expression of these enzyme genes. Therefore, the above key genes have been used to co-express the network and character the genes that regulate metabolites (flavonoids) compounds. The genes, including LAR, ANR, DFR, F3H, CHS, E5.5.1.6, F3 H, CYP73A, HCT, C3 H, and FLS, have the highest degree of connectivity with gallocatechin, naringenin, dihydrofisetin (fustin), catechin, tricetin, epicatechin, and hesperetin-7-O-neohesperidoside (neohesperidin) biosynthesis ( Figure 6B). Two hub metabolites (gallocatechin and naringenin) were significantly associated with key enzyme genes in flavonoid synthesis. The expression level of HCT, C3 H, DFR, F3H, FLS, CHS, and ANR genes were significantly related to the synthesis of flavonols (5), dihydroflavonol (4), flavonoid (3), flavanols (3), and flavonoid carbonoside (2) in JHK. Therefore, these results suggested that these structure genes might be a key regulator of flavonoids biosynthesis in JHK.  in Hibiseu manihot L. In the above heat map, the red color indicates flavonoids accumulation significantly, and the blue color indicates flavonoids content significant reduction. In the network, the green circles represent the DEGs and the red circles represent the corresponding metabolites. In each figure, F, L, St, R, and Se represent the flower, leaves, stems, roots, and seeds of Hibiseu manihot L., respectively.

DISCUSSION
Jinhuakui is directly edible, and used in medicinal and health care functions for developing functional foods/products to improve humans health. It can also be used as garden plants having ornamental value. Flavonoids, an active medicinal substance, are the highest in JHK than other known natural products species Li et al., 2012). Studies found that the extractive compounds from the JHK have been shown to decompose or scavenge DPPH, ABTS radical, hydroxyl, and superoxide radicals, inhibiting LDL oxidation (Yang et al., 2006) and chemo-preventive activities, and inhibiting xanthine-oxidase and lipid peroxidation (Wright et al., 2007). The flowers and calyces of JHK can be used to treat heart and nerve diseases, antitumor, diuretic, anti-scorbutic, sedative, colorectal, and intestinal antiseptic (Yang et al., 2006;Wright et al., 2007). The leaves and seeds of JHK were also used to treat conjunctivitis, ringworms, tumor, and abscesses or to alleviate headache, rheumatism, and hemorrhoid (Yang et al., 2006;Wright et al., 2007). JHK has high medicinal value and application prospects. Considering that JHK has so many pharmacological activities, it is of great significance to study the composition and synthesis mechanism of secondary metabolites in JHK.
Plant flavonoids have been illustrated to be antioxidants and anti-inflammatory against cerebral ischemia injury . They also have antiradical properties to prevent the associated diseases (Rice-Evans et al., 1996;Pan F. et al., 2017) by activating related gene expression, by interacting with MAP-kinase and PI3-kinase, and having an enzyme activity that can mediate cellular signaling transduction and other pathways (Seifried et al., 2007;Vauzour et al., 2010). Flavonoids were one of the main active metabolites in the JHK, and the accumulation of flavonoids in the flower can be up to 5.6% of the dry weight (Yang, 2013), which is ten times higher than that of ginkgo biloba and soybean (commonly used in industry to extract flavonoids) (Zhou et al., 2021). To systematically study the distribution and accumulation pattern of flavonoids in different organs of JHK, we combined metabolome (LC-MS/MS) with transcriptome sequencing to explore the flavonoids compound biosynthesis mechanism in different tissues of JHK. In the present study, we determined the total flavonoids content in JHK, and the results showed that the total flavonoids content in flowers was significantly higher than that in leaves, stems, and roots. Numerous studies have indicated a positive correlation between the contents of flavonoids and phenolic compounds and the antioxidant capacities in plant extracts (Aryal et al., 2019;Muflihah et al., 2021). All of the flavonoids in JHK had an efficient inhibitory effect on ROS production (Tsumbu et al., 2012) to treat certain diseases caused by the accumulation of ROS. Therefore, to make better use of the JHK plant, we recommend using the flower of this plant as a raw material to extract total flavonoids, which can improve the efficiency of flavonoids' industrial production.
Although JHK has a high content of flavonoids, little information is available about the types and accumulation patterns of flavonoids in different tissues. In this study, a total of 160 flavonoids were identified in JHK by LC-MS/MS, and these flavonoids were mainly divided into 11 categories, of which the largest proportion was flavonols, followed by flavonoids and flavanols. At the same time, we found that the results of this experiment were similar to the results of the classification of flavonoids in Abelmoschus Manihot (relative species of JHK) (Pan X. et al., 2017), while there might be significant differences due to the content of these metabolites. Previous studies reported and identified eight major flavonoids in the flower of JHK Wu et al., 2019), which included vitexin rhamnoside, apigenin-8-C-glucoside (vitexin), rutin, hyperoside, quercetin, quercetin 3-O-robinobioside, quercetin-3-O-glucoside (isoquercitrin), and myricetin. Our results were consistent with these 8 kinds of flavonoids detected and identified by predecessors, and these DAFs have neuroprotective (Magalingam et al., 2014), antitumor (Huang et al., 2014;Lü, 2016), antioxidant, and antiinflammatory effects (Ahn and Lee, 2016) by suppressing the activation of nuclear factor-κB in mouse peritoneal macrophages (Kim et al., 2011). In addition, the results of this study showed that spiraeoside, quercetin-3-O-glucuronide, quercetin-3-O-(2 -O-glucosyl) glucuronide, quercetin-3-O-(6 -galloyl) glucoside, and quercetin-3-O-(2 -galloyl) glucoside were only determined in the flowers of JHK with high content, which have been illustrated with effects of antibacteria, anticonvulsant, antitumor, and anti-inflammatory activities (Chua, 2013) to inhibit partially the exocytosis of elastase and which might regulate the activation certain pathways (Selloum et al., 2003). Pelargonidin-3-Orutinoside, apigenin-5-O-glucoside, and quercetin-3-O-(6 -pcoumaroyl) galactoside were unique in the roots, leaves, and seeds of JHK, respectively.
In JHK, the accumulation of flavonoids was significantly associated with the flavonoid biosynthesis and the flavonol biosynthesis pathways, which directly target hundreds of flavonoid biosynthesis genes. There were significant differences in the accumulation patterns of flavonoids in different tissues/organs of JHK, and the related key genes that were significantly enriched in the pathway were also differentially expressed. Therefore, a significant cascade of transcriptional reprogramming and metabolite synthesis flow of flavonoids biosynthesis were studied in JHK. In the present study, we found that PAL catalyzes the conversion of phenylalanine to cinnamoyl-CoA, and then the CYP73A enzyme catalyzes the isomerization of cinnamoyl-CoA to p-coumaroyl-CoA. Compared with L, R, St, and Se, the genes coding PAL and CYP73A were significantly upregulated in F of JHK. The expression of these genes and upstream metabolic compounds accumulation pave the way for the synthesis of downstream flavonoids in JHK. CHI catalyzes the isomerization of naringenin