Abstract Background Paris polyphylla Sm. is a precious medicinal plant rich in various active ingredients. In addition to the well-known saponins, the flavonoids it contains have unique pharmacological potential in antioxidant, neuroprotective, and metabolic regulation. However, the flavonoids in Paris polyphylla Sm. have not been fully researched and developed yet. In this work, we conducted a comprehensive metabolomics and transcriptomics analysis to reveal the metabolic differences and biosynthetic mechanisms of flavonoids in the leaves, stems, and roots of Paris polyphylla Sm. Results Non-targeted metabolomics analysis detected a total of 332 metabolites in Paris polyphylla Sm., among which flavonoids accounted for 19.49%. The diversity and abundance of flavonoids in leaves are the highest, followed by stems and roots. By comparing the metabolites of the roots, stems, and leaves in Paris polyphylla Sm., it was found that there were 45 differential metabolites (DMs) between the leaves and roots, of which flavonoids accounted for 35%. There are 38 DMs between leaves and stems, of which flavonoids account for 45.45%. And there are 52 DMs in stems and roots, among which flavonoids account for 25.53%. A total of 62,766 genes were detected by transcriptomics, and pairwise comparison showed that there were tens of thousands of differentially expressed genes (DEGs) between each group. Afterwards, we selected 39 flavonoids and related metabolites (e.g., kaempferol-3-O-glucoside, quercetin 3-β-D-glucoside, rutin) for targeted metabolomics validation and performed RT-qPCR validation on 29 key flavonoid synthesis genes (e.g., C4H, CHS, FLS, F3’H) to verify the reliability of non-targeted metabolomics and transcriptomics. Conclusions This work indicated that leaves are the main site for the biosynthesis of flavonoids in Paris polyphylla Sm. Among them, kaempferol-3-O-glucoside, quercetin 3-β-D-glucoside, rutin, and other flavonoids are present in higher contents in leaves (P < 0.05). Further research on its biosynthetic mechanism indicates that naringenin chalcone is converted to naringenin by chalcone isomerase (CHI). Among them, CHI may be the rate-limiting enzyme in the biosynthesis of flavonoids in Paris polyphylla Sm. The expression of FLS is higher in leaves (P < 0.05) and tends to promote the synthesis of flavonols. This work promotes the utilization of non-medicinal parts of Paris polyphylla Sm. and enhances the sustainable development of this precious traditional Chinese medicine resource. Supplementary Information The online version contains supplementary material available at 10.1186/s12870-025-07350-8. Keywords: Paris polyphylla sm., Flavonoids, Metabolomics, Transcriptomics, Biosynthesis Introduction Paris polyphylla Sm. is a plant belonging to the Melanthiaceae family. It has a long history of medicinal use in China and is also distributed in regions such as Bhutan, Vietnam, India, and Nepal. This species primarily grows in forests, grasslands, or on slopes [[46]1]. However, due to habitat destruction and illegal harvesting, the population of wild Paris polyphylla Sm. has significantly declined, leading to its classification as a “vulnerable” species on the IUCN Red List of Threatened Species [[47]2]. The 2020 edition of the Chinese Pharmacopoeia includes Paris polyphylla var. yunnanensis and Paris polyphylla var. chinensis, which are two of the main medicinal plants used in traditional formulations such as Yunnan Baiyao and Gongxuening Capsules [[48]3]. The root of Paris polyphylla Sm. is the primary medicinal part. Modern pharmacological studies have revealed that Paris polyphylla Sm. exhibits multiple biological activities, including anti-cancer, antibacterial, anti-inflammatory, antioxidant, and immunoregulatory effects. The plant primarily contains steroidal saponins, β-ecdysone, polysaccharides, flavonoids, amino acids, and other chemical components, underscoring its significant medicinal and economic value [[49]4, [50]5]. The saponin components in Paris polyphylla Sm. have rich pharmacological activities, such as anti-inflammatory, antibacterial, and immunoregulation [[51]6]. Saponins are the most extensively studied compounds in Paris polyphylla Sm., yet the flavonoids with rich pharmacological activities present in the plant have received insufficient attention. The research on flavonoids in Paris polyphylla Sm. is expected to fill the limitations of saponin research and further promote the development of Paris polyphylla Sm. resources. The basic skeleton of flavonoids consists of two benzene rings (A and B rings) and several phenolic hydroxyl groups, connected by three central carbon atoms [[52]7]. Flavonoids are mainly divided into six categories, including chalcones, flavones, flavonols, flavandiols, anthocyanins, and condensed tannins (or proanthocyanidins) [[53]8]. These compounds are predominantly found in the roots and leaves of plants. Flavonoids exist in two primary forms: free aglycones and glycoside-bound forms. They exhibit a wide range of functional activities, including antioxidant, antibacterial, antiviral, anti-inflammatory, and anti-tumor effects, etc [[54]9, [55]10]. For example, hesperidin is a dihydroflavonoid glycoside that has antihypertensive, antioxidant, and hypoglycemic effects [[56]11]. Baicalin is a flavonoid extracted from Scutellaria baicalensis, which has significant anti-cancer activity against various malignant tumors [[57]12]. As of 2024, a total of 19 flavonoid-based drugs have been approved for market use, with natural flavonoids accounting for 52.6% of them. Additionally, there are 36 flavonoid-based clinical candidate drugs currently undergoing or at various stages of development, among which natural flavonoids make up 44.4% [[58]13]. Due to their rich biological activity, flavonoids remain the main choice in drug development and have important development value and potential. Although there is currently a lack of research on flavonoids in Paris polyphylla Sm., there are still a few studies indicating that flavonoids are one of the important pharmacological active substances in Paris polyphylla Sm. For example, Zhang et al. found that flavonoids are one of the main substances in the roots of Paris polyphylla Sm [[59]14]. In addition, Su et al. also found some flavonoid compounds in the leaves of Paris polyphylla Sm., such as the presence of kaempferol 3-Glucoside [[60]15]. However, these studies still mainly focus on the exploration of steroidal saponins in Paris polyphylla Sm. Currently, there is still a lack of comprehensive and in-depth research on flavonoids in Paris polyphylla Sm. In recent years, metabolomics has been widely used for high-throughput analysis of small molecule metabolites in plants [[61]16], while transcriptomics has been used to analyze gene expression profiles in plants [[62]17]. Metabolomics is often combined with transcriptomics to achieve a comprehensive physiological mechanism interpretation of plants from genes to metabolites [[63]18]. Metabolomics and RNA-seq profiles of many plants have been analyzed to reveal the mechanisms of compound synthesis and dynamic changes in different plant species, including chrysanthemums [[64]19], tobacco [[65]20], Atractylodes macrocephala Koidz [[66]21], Astragalus membranaceus [[67]22], etc. Of course, metabolomics and transcriptomics can be used to study the biosynthetic mechanisms of flavonoids. For example, Jia et al. combined metabolomics and transcriptomics to analyze the biosynthesis mechanism of flavonoids in the roots of Aster tataricus and found that genes such as cinnamate 4-hydroxylase (C4H) and flavonol synthase (FLS) are key genes involved in the synthesis of kaempferol [[68]23]. Therefore, in order to systematically study the types and distribution of flavonoids in different tissues (leaves, stems, and roots) of Paris Polyphylla Sm., metabolomics combined with transcriptomics was used for analysis of different tissues of Paris Polyphylla Sm. At present, the utilization of non-traditional medicinal parts in traditional Chinese medicine (TCM) is one of the trends in the development of Chinese medicine resources. For example, Jiang et al. discovered two components (henicosane-1 and decahydroisoquinoline-2) with significant anti-inflammatory effects from the aerial parts of Glycyrrhiza spp [[69]24]. As a precious TCM, studying the distribution, content, and biosynthetic mechanism of flavonoids in Paris Polyphylla Sm. will promote its development, especially the rational utilization of non-medicinal parts. Materials and methods Plant materials and sampling The Paris polyphylla Sm. samples collected in this work were cultivated by the Shiyan Longwangzhai Rare Chinese Herbal Medicine Planting and Breeding Farmer Professional Cooperative. In May 2024, 7-year-old Paris polyphylla Sm. plants in the fruiting stage were collected from well grown individuals in the lower layer of the forest, and identified by Professor Li Zhihao from the Department of Pharmacy, Sinopharm Dongfeng General Hospital, Hubei University of Medicine. A voucher specimen (voucher number: HIB0258879) has been deposited in the Herbarium of Wuhan Botanical Garden, Chinese Academy of Sciences (CAS). It was identified by Guangwan Hu. Paris polyphylla Sm. Plants with consistent growth status were selected, washed 3 times with ultrapure water, and then dissected into leaf, stem, and root tissues. These tissues were immediately frozen in liquid nitrogen and stored at −80 ℃ for subsequent analysis [[70]25]. Non-targeted metabolomics In the HPLC-MS procedures, after the sample was slowly thawed at 4 °C, an appropriate amount of sample (50 to 100 mg) was added to 1mL of a mixture (composed of water, acetonitrile, and isopropanol in a volume ratio of 1:1:1), vortexed for 60 s, extracted using frozen ultrasonication for 30 min, and then centrifuged at 4 °C and 12,000 rpm for 10 min [[71]26]. Then, the supernatant was collected and left to stand at −20 °C for 1 h to precipitate proteins. After centrifugation at 4 °C and 12,000 rpm for 10 min, the resulting solution was dried under vacuum. Finally, 200 µL of 30% acetonitrile solution was added for redissolution, centrifuged at 14,000 rpm and 4 °C for 15 min, and the supernatant was collected for instrumental analysis. The key components of the data collection apparatus were an ultra-high performance liquid chromatography system (Vanquish, UPLC, Thermo, USA) and a high-resolution mass spectrometer (Q Exactive HFX, Thermo, USA), equipped with a Waters HSS T3 chromatographic column (100 × 2.1 mm, 1.8 μm). The mobile phase consisted of an ultrapure water mixture (containing 0.1% formic acid) and an acetonitrile mixture (with 0.1% formic acid), flowing at a rate of 0.3 mL/min, and the column temperature was maintained at 40 °C. The injection volume was 2 µL. The elution gradient was as follows: at 0 min, the A/B phase ratio was (100:0, v/v); at 1 min, the A/B phase ratio was (100:0, v/v); at 12 min, the A/B phase ratio was (5:95, v/v); at 13 min, the A/B phase ratio was (5:95, v/v); at 13.1 min, the A/B phase ratio became (100:0, v/v); and at 17 min, the A/B phase ratio remained (100:0, v/v). Throughout the entire analysis process, the samples were stored in an autosampler at 4 °C. To minimize the impact of changes in the instrument’s detection signal, the samples were continuously analyzed randomly. During the sample analysis process, equal amounts of each test sample were uniformly mixed to prepare Quality Control samples (QC samples). The purpose of these QC samples was to assess the stability of the system and the reliability of the experimental results. The Thermo Q Exactive HFX high-resolution mass spectrometry system was used to collect both primary and secondary spectra. It was equipped with an electrospray ionization (ESI) source, with 40 arb of sheath gas, 10 arb of auxiliary gas, an ion spray voltage of + 3000 V/−2800 V, a temperature setting of 350 °C, and an ion transfer tube temperature of 320 °C. The scanning was performed in Full-ms-ddMS^2 mode, which could be either in positive or negative ion mode. The range of primary mass spectrometry scans was between 70 and 1050 Da, with a primary resolution of 70,000 and a secondary resolution of 17,500. Transcriptomics sequencing and analysis RNA Extraction and Library Preparation: RNA Isolation: Tissues were homogenized in TRIzol reagent (Invitrogen, USA). Total RNA was extracted via chloroform/isoamyl alcohol phase separation, precipitated with sodium acetate, washed with 75% ethanol, and dissolved in DEPC-treated water. mRNA Enrichment: Poly-A mRNA was isolated using magnetic beads (Truseq™ RNA Sample Prep Kit, Illumina, USA). cDNA Synthesis: Double-stranded cDNA was synthesized, amplified, and size-selected (200–500 bp) via 2% agarose gel electrophoresis (Bio-Rad, USA). Sequencing and Data Processing: Platform: Illumina HiSeq X Ten (2 × 150 bp paired-end reads). Raw Data Processing: Adapters and low-quality reads (Q ≤ 10 in > 50% bases, N content > 10%) were removed to generate clean data. Clean reads were aligned to the Paris polyphylla Sm. reference genome using HISAT2. Transcript assembly and quantification were performed with StringTie. Targeted metabolomics validation Due to the inability of non-targeted metabolomics to accurately measure the content of metabolites, We conducted targeted metabolomics validation on flavonoid metabolites. Information on standard substances is shown in Table [72]S1. An appropriate amount of sample was added to 0.5mL of 80% methanol aqueous solution (containing 0.2% vitamin C) and vortex mixed. Ultrasonic extraction was performed at room temperature for 30 min, followed by centrifugation at 12,000 rpm for 10 min. After collecting the supernatant, repeat the extraction twice and finally perform instrument analysis. Real-time quantitative PCR (RT-qPCR) analysis We conducted RT-qPCR to validate 29 genes involved in the biosynthesis of flavonoids. And the experiment was conducted using an ABI Stepone Plus Real-Time PCR System. After extraction, the total RNA was detected by RT-qPCR, with a total reaction volume of 20 µL, including 2 × One Step SYBR Green Mix (10 µL), One Step SYBR Green Enzyme Mix (1 µL), 50 × ROX Reference Dye 1 (0.4 µL), Gene Specific Primer Forward (10 µM, 0.3 µL), Gene Specific RT Primer (10 µM, 0.3 µL), URP Primer (0.3 µL), template RNA, and Supplemented to 20 µL with RNase free ddH[2]O. The primer sequence is supplemented in Table [73]S2. The reverse transcription reaction program is 50 ℃ for 10 min, and 95 ℃ for 1 min. The amplification program consists of 40 cycles at 95 ℃ for 5 min, 60 ℃ for 30 s, and 72 ℃ for 30 s. The final program consists of one cycle at 95 ℃ for 15 s, 55 ℃ for 60 s, and 95 ℃ for 15 s. The detection data was analyzed using the 2- ^△ △ CT method. Results Non-targeted metabolomics analysis of Paris polyphylla Sm. A total of 332 metabolic compounds were detected in the samples derived from root, stem, and leaf tissues of Paris polyphylla Sm. The metabolite identification reached Level 2 or above, ensuring the reliability of the qualitative results [[74]27]. The relative proportions of these compounds mainly include flavonoids (19.49%), carboxylic acids and derivatives (18.41%), fatty acyls (7.94%), prenol lipids (6.86%), organooxygen compounds (5.78%), as shown in Fig. [75]1A. Notably, flavonoids represent the largest metabolite category. Fig. 1. [76]Fig. 1 [77]Open in a new tab Non-targeted metabolomics profiling and quality control of different tissues in Paris polyphylla Sm. (A) Distribution of metabolite types in Paris polyphylla Sm.; (B–D) OPLS-DA score plots of the Leaves group vs. Roots group, Leaves group vs. Stems group, and Stems group vs. Roots group, respectively(n = 6) Principal Component Analysis (PCA), an unsupervised multivariate statistical method, was employed to evaluate the overall differences between sample groups and intra-group variability. The PCA results revealed distinct differences among the three tissues of Paris polyphylla Sm., with each biological replicate showing strong intra-group correlation (Fig. [78]S1A). QC samples exhibited high aggregation consistency, with 95.72% of metabolic features falling within a 70% reproducibility threshold (Fig. [79]S1B). Additionally, more than 90% of the metabolites had relative standard deviations (RSD) below 10% [[80]28]. These results confirm the stability of the analytical system and the reliability of the dataset. PCA score plots revealed clear separation among samples from different parts of Paris polyphylla Sm., indicating substantial inter-group variability. The cumulative interpretation rates (R^2X) for the first three principal components were 55%, 50.1%, and 52.6%, respectively, all exceeding the threshold of 0.5, which confirms the robustness of the model (Fig. [81]S1C, E, and G). To further refine the analysis, supervised Orthogonal Partial Least Squares-Discriminant Analysis (OPLS-DA) was applied to filter out non-relevant variations and identify differential metabolites (DMs). The validity of the OPLS-DA model was assessed through permutation testing. OPLS-DA score plots demonstrated distinct clustering among the three comparison groups (Roots, Stems, and Leaves), further supporting significant differences between sample groups (Fig. [82]1B, C, and D). Permutation test results showed that the blue regression line of the Q2 parameter intersected the vertical axis (left) at or below zero, confirming the absence of overfitting and the reliability of the model (Fig. [83]S1D, F, and H). The volcano maps compared pairwise were plotted based on P-value < 0.05 and fold change (FC) > 1.5 or FC < 0.67, as shown in Fig. [84]2A, C, and E. Generally, variable importance in projection (VIP) values > 1 are considered important variables. Therefore, the screening criteria for DMs are P-value < 0.05 and VIP > 1. Comparing the Leaves group with the Roots group, a total of 45 DMs were screened, of which 26 were up-regulated and 19 were down-regulated. Figure [85]2B shows the classification of these DMs, with flavonoids metabolites having the highest proportion, accounting for 35%. Comparing the Leaves group with the Stems group, a total of 38 DMs were screened, of which 20 were up-regulated and 18 were down-regulated. Figure [86]2D shows the classification of these DMs, with flavonoids metabolites having the highest proportion, accounting for 45.45%. Comparing the Stems group with the Roots group, a total of 52 DMs were screened, of which 33 were up-regulated and 19 were down-regulated. As shown in Fig. [87]2F, flavonoid metabolites still have the highest proportion among DMs, accounting for 25.53%. Given their prominent biological activities, flavonoids were prioritized for further analysis. Upset plot analysis identified 9 metabolites with significant differences in expression levels among roots, stems, and leaves (Fig. [88]2G). Among them, there are 5 flavonoid metabolites, including: 6’-O-L-arabinopyranosylastragalin, vincetoxicoside A, luteolin-3’,7-di-O-glucoside, luteolin 4’-glucoside, and fisetin. Fig. 2. [89]Fig. 2 [90]Open in a new tab DMs analysis of Paris polyphylla Sm. (A, C,and E) The volcano plot of Leaves group vs. Roots group, Leaves group vs. Stems group, and Stems group vs. Roots group in sequence. (B, D,and F) The pie chart of DMs classification for Leaves group vs. Roots group, Leaves group vs. Stems group, and Stems group vs. Roots group, respectively. (G) The upset plot of DMs compared pairwise All DMs identified in the comparison groups were mapped to the KEGG database to annotate their associated metabolic pathways. A statistical analysis of the top 20 enriched pathways (ranked by the number of DMs) was visualized via bar charts. As shown in Fig. [91]S2A, the DMs in the Leaves group and Roots group are related to the biosynthesis pathway of flavonoids, including (-) - fustin, phlorizin, kaempferol, and luteolin. As shown in Fig. [92]S2B, the DMs in the Leaves group and Stems group are related to the biosynthesis pathway of flavonoids, including (-) - fustin, phlorizin, kaempferol, and luteolin. And Fig. [93]S2C shows that the DMs related to the biosynthesis pathway of flavonoids in the Roots group and Stems group are quercetin. Non-targeted metabolomics analysis showed that flavonoids were the most abundant metabolic products (accounting for about 34.92%) in Paris polyphylla Sm., with the highest content in the leaves, accounting for about 56.86%. It is worth noting that metabolites related to the synthesis of flavonoids are enriched in leaf tissues, indicating that leaves play a central role in the production of flavonoids. Transcriptomics analysis of Paris polyphylla Sm. After removing low-quality reads, a total of 1,026,924,376 clean reads were obtained, with Q30 percentages ranging from 93.76 to 97.16% and GC content between 45.91% and 47.57%, indicating high sequencing quality. PCA of the samples demonstrated strong intra-group aggregation and clear inter-group separation (Fig. S3A), further validating the robustness of the sequencing data and significant biological differences between groups. Differentially expressed genes (DEGs) were identified using thresholds of |FC| ≥ 2 and false discovery rate (FDR) < 0.05. The DEGs results of pairwise comparisons between groups are shown in Table [94]1. The gene expression levels and FC were visualized through Mean-Average (MA) plots (Fig. S3B-D) and volcano plots (Fig. [95]3A, D, and G). KEGG pathway enrichment analysis of DEGs showed significant enrichment in flavonol biosynthesis across all comparison groups (Fig. [96]3B, E, and H). Table 1. Results of differential gene expression in Paris polyphylla Sm. DEGs Set DEGs Number up-regulated down-regulated Leaf vs. Root 32,519 15,932 16,587 Leaf vs. Stem 18,210 9941 8269 Stem vs. Root 31,097 14,956 16,141 [97]Open in a new tab Fig. 3. [98]Fig. 3 [99]Open in a new tab The volcano plot of DEGs, the KEGG pathway enrichment map of DEGs, and the heat map of genes related to flavonoids biosynthesis in each comparison group. (A, D,and G) The volcano plot of DEGs for Leaves group vs. Roots group, Leaves group vs. Stems group, and Stems group vs. Roots group, respectively. (B, E,and H) The KEGG pathway enrichment map of DEGs for Leaves group vs. Roots group, Leaves group vs. Stems group, and Stems group vs. Roots group, respectively. (C, F,and I). The heat map of genes related to flavonoids biosynthesis for Leaves group vs. Roots group, Leaves group vs. Stems group, and Stems group vs. Roots group, respectively (n = 4) The analysis of genes related to flavonol biosynthesis showed that hydroxycinnamyl transferase (HCT), chalcone isomerase (CHI), FLS, chalcone synthase (CHS) and flavanone 3-hydroxylase (F3H) were highly expressed in leaves, followed by stems. The expression of leucoanthocyanidin dioxygenase (LDOX), cytochrome P450 98A2 (CYP98A2), naringenin 7-O-methyltransferase (NB40MT), cinnamate 4-hydroxylase (CYP73A) were the highest in stems, followed by roots. The dihydroflavonol 4-reductase (DFR) was the highest in leaves and stem compared to roots (Fig. [100]3C, F, and I). Transcriptomics profiling demonstrated significant up-regulation of flavonoid biosynthetic genes in leaves and stems, suggesting that these tissues are the primary sites for the biosynthesis of most classified flavonoids in Paris polyphylla Sm. Combined metabolomic and transcriptomics analysis The PCA analysis combining transcriptomics and non-targeted metabolomics data showed significant inter group separation of samples, while intra group similarity was high, supporting their compatibility in joint analysis (Fig. S4). Cross analysis of KEGG pathways rich in DMs and DEGs showed that there were 36 shared pathways between Leaves group vs. Roots group, as well as between Leaves group vs. Stems group, and 44 shared pathways between Stems group vs. Roots group (Fig. [101]4A, C and E). The enrichment pathways in each pairwise comparison group were further visualized through bar charts (Fig. [102]4B, D, and F). Ko00941 is a biosynthetic signaling pathway for flavonoids, in which the Leaves group vs. Roots group mainly involves 40 DEGs and 4 DMs, the Leaves group vs. Stems group mainly involves 48 DEGs and 4 DMs, and the Stems group vs. Roots group does not involve. These results indicate that the leaves of Paris polyphylla Sm. are the core of flavonoid biosynthesis. Fig. 4. [103]Fig. 4 [104]Open in a new tab The Venn diagrams and enriched pathways of DMs and DEGs in pairwise comparison groups. (A, C, and E) The Venn diagrams of DMs and DEGs in Leaves group vs. Roots group, Leaves group vs. Stems group, and Stems group vs. Roots group, respectively. (B, D, and F) The enriched pathways of DMs and DEGs in Leaves group vs. Roots group, Leaves group vs. Stems group, and Stems group vs. Roots group, respectively The integration of DEGs and DMs reveals key regulatory mechanisms in the biosynthesis of flavonoid metabolites. As shown in Fig. S5A-C, in the biosynthetic pathway of flavonoid metabolites, DMs mainly involve phlorizin, luteolin, kaempferol, dihydrofisetin, and quercetin. In the comparative group, the up-regulation of gene expression encoding CHS ([105]K00660) was associated with an increase in phlorizin. The coordinated up-regulation of genes encoding CHS ([106]K00660), CHI ([107]K01859), and F3H ([108]K00475) likely enhanced the biosynthesis of dihydrofisetin. In the comparison between the Leaves group and the Roots group, CHS ([109]K00660), CHI ([110]K01859), F3H ([111]K00475), and FLS (K05278) were up-regulated, leading to the conversion of p-coumaroyl-CoA to kaempferol. In contrast, the expression of F3H ([112]K00475) remained unchanged between the Leaves group and the Stems group. In the comparison between the Stems group and the Roots group, the flavonoid biosynthesis related pathways were up-regulated. Targeted validation of key metabolites In order to further validate the flavonoid DMs screened by non-targeted metabolomics, we conducted targeted metabolomics validation on flavonoid DMs and their related metabolites, involving a total of 39 metabolites, of which 25 metabolites were successfully detected, as shown in Fig. [113]5A-I and Fig. S6A-P. We found that among flavonoids and their biosynthetic precursors, metabolites such as rutin, quercetin 3-β-D-glucoside, and kaempferol-3-O-glucoside had higher contents in leaves (P < 0.05). Specifically, the content of rutin, quercetin 3-β-D-glucoside, and kaempferol-3-O-glucoside in the leaves accounted for 3.126%, 0.255%, and 6.423%, respectively. The content of rutin, quercetin 3-β-D-glucoside, and kaempferol-3-O-glucoside in the stem accounts for 0.245%, 0.142%, and 0.475%, respectively. Among them, the contents of rutin and kaempferol-3-O-glucoside in leaves exceeded 1%, indicating their potential for industrial extraction. Fig. 5. [114]Fig. 5 [115]Open in a new tab Targeted metabolomics content bar chart. (R = Roots, S = Stems, L = Leaves, n = 4, ns: non-significant, *. P < 0.05; **. P < 0.01; ***. P < 0.001) RT-qPCR validation of key genes According to transcriptomics results, 29 genes (which encode 14 flavonoid biosynthesis related enzymes) were selected and validated by RT-qPCR, among which 5 genes showed no significant difference between groups (p > 0.05). The results are shown in Fig. [116]6A-L and Fig. S7 A-L, specifically involving C4H, flavonoid 3’-hydroxylase (F3’H), F3H, FLS, 4-coumarate-CoA ligase (4CL), HCT, caffeoyl-CoA O-methyltransferase (CCoAOMT), CHS, DFR, phenylalanine ammonia-lyase (PAL), LDOX, CHI, CYP98A2, flavone synthase (FNS). F3’H, FLS, PAL and other flavonoid biosynthesis related enzymes have the highest expression levels in leaves (P < 0.05). RT-qPCR detection results for these genes are consistent with transcriptomic data. Fig. 6. [117]Fig. 6 [118]Open in a new tab (A-L) RT-qPCR validation of key genes. (R = Roots, S = Stems, L = Leaves, n = 3, ns: non-significant, *. P < 0.05; **. P < 0.01; ***. P < 0.001) Biological synthesis process of flavonoids metabolites in Paris polyphylla Sm. Based on the above research results, we have sorted out the biosynthesis of flavonoid metabolites in Paris polyphylla Sm. (Fig. [119]7). The biosynthetic pathway of flavonoids is based on phenylalanine synthesized through the shikimic acid pathway, which is then converted into p-coumaroyl-CoA via key enzymes (PAL, C4H, 4CL) in the phenylpropanoid pathway. This is followed by the production of naringenin chalcone through CHS. Naringenin chalcone is modified with CHI to produce naringenin. RT-qPCR validation showed that CHI had a low overall expression level, suggesting that it is the rate-limiting enzyme in flavonoid biosynthesis. F3’H causes flavonoid biosynthesis in Paris polyphylla Sm. to lean towards flavonol metabolites. FLS is up-regulated in the leaves, leading to a greater flow of dihydroquercetin and dihydrokaempferol towards the production of quercetin and kaempferol, which is consistent with targeted metabolomics validation of higher content of quercetin and kaempferol derivatives. Fig. 7. [120]Fig. 7 [121]Open in a new tab Biological synthesis process diagram of flavonoids metabolites in Paris polyphylla Sm. Discussion Metabolite differences in different parts of the Paris polyphylla Sm. As a traditional Chinese medicine, Paris polyphylla Sm. has attracted widespread attention due to its rich pharmacological activities, including anti-tumor [[122]29], anti-inflammatory [[123]30], anti-aging [[124]31], etc. Paris polyphylla Sm. is rich in various active ingredients, such as steroidal saponins, flavonoids, polysaccharides, etc., which are the basis for its rich pharmacological activity [[125]32]– [[126]33]. In our work, we first analyzed the metabolites of the roots, stems, and leaves of Paris polyphylla Sm. using non-targeted metabolomics to clarify the differences in metabolite composition among different parts of the plant. 332 metabolites were detected in Paris polyphylla Sm., among which flavonoids accounted for 19.49% and had the highest content in the leaves, making them the most important metabolites. Flavonoids are one of the important active ingredients present in plants, with rich pharmacological activities such as anti-tumor [[127]12], anti-inflammatory [[128]34], anti diabetes [[129]35] and neuroprotective [[130]36]. And flavonoids in Paris polyphylla Sm. have great research and development potential. Therefore, in our subsequent research, we will mainly focus on exploring the flavonoid metabolites in Paris polyphylla Sm. To further validate the expression levels of DMs, targeted metabolomics was used for quantitative determination of differential flavonoid metabolites and their related metabolites. In the quantitative detection of 39 metabolites from different parts of Paris polyphylla Sm., 25 metabolites were successfully detected, among which rutin, quercetin 3-β-D-glucoside, and kaempferol-3-O-glucoside were highly expressed in leaves. Rutin is one of the important flavonoids with rich pharmacological activity. Wang et al. found that rutin reduces psoriasis inflammation by down-regulating the AGE-RAGE signaling pathway [[131]37]. And Ye et al. found that rutin improves inflammatory pain by inhibiting mast cell P2 X 7 receptors [[132]38]. quercetin 3-β-D-glucoside is a derivative of quercetin, which has been found to have anti-breast cancer cell proliferation effect [[133]39]. kaempferol-3-O-glucoside is a derivative of kaempferol that has been found to have a protective effect against age-related cognitive impairment [[134]40]. The extract of Erica multiflora, rich in quercetin-3-O-glucoside and kaempferol-3-O-glucoside, can alleviate high-fat and fructose diet induced fatty liver [[135]41]. The leaves of Paris polyphylla Sm. are often wasted, and its rich Substances such as rutin, quercetin 3-β-D-glucoside, and kaempferol-3-O-glucoside have the potential for development and utilization. However, unfortunately, this work did not investigate the pharmacological activity of these metabolites. In the future, we will conduct research on this to promote the utilization of non-medicinal parts of Paris polyphylla Sm. Genes differences in different parts of the Paris polyphylla Sm. Transcriptomics has been used for gene detection in different parts of Paris polyphylla Sm., and through pairwise comparisons, it has been found that there are tens of thousands of DEGs between each group. We screened the genes involved in the biosynthesis of flavonoids and quantitatively detected the key genes through RT-qPCR. We found that F3’H, CHI, CHS, and FLS had the highest expression levels in the leaves of Paris polyphylla Sm. They are key enzymes involved in the biosynthesis of flavonoid metabolites in plants. For example, Zhang et al. found that CHS is involved in the biosynthesis of flavonoids in Torreya grandis kernels [[136]42]. Zhu et al. found that CHI is a key enzyme in the biosynthesis pathway of flavonoids in Dracaena cambadiana [[137]43]. Similarly, F3’H and FLS are key enzymes involved in flavonoid biosynthesis in plants [[138]44]– [[139]45]. These enzymes are key enzymes in the biosynthesis of flavonoids, but their regulatory mechanisms are complex and precise. Flavonoids originate from the phenylpropane pathway, which involves the catalytic action of PAL, C4H, and 4CL to generate p-coumaroyl-CoA [[140]46]. P-coumaroyl-CoA is catalyzed by CHS and CHI to produce flavanones, which are key intermediate products for the subsequent generation of flavonoids [[141]47]. Flavonoids are generated from flavanones under the action of FNS, and flavonols (such as quercetin, kaempferol, rutin, and their derivatives) are generated under the action of F3’H and FLS [[142]48]– [[143]49]. The high expression of these key enzymes involved in flavonoid biosynthesis in leaves regulates the high content of flavonoid metabolites. The biosynthetic pathway of flavonoids in Paris polyphylla Sm. Based on the above data, we have summarized the mechanism of flavonoid biosynthesis Paris polyphylla Sm. Phenylalanine is converted to cinnamic acid by PAL, then to P-coumaric acid by C4H, and finally to p-coumaroyl-CoA by 4CL. In transcriptomics and RT-qPCR results, PAL, C4H, and 4CL were found to have high expression in the leaves. Under the catalysis of CHS, p-coumaroyl-CoA condenses with malonyl-CoA to form naringenin chalcone. And naringenin chalcone is isomerized to naringenin under the action of CHI. Compared to roots and stems, CHS and CHI have the highest expression in leaves, but the overall expression level of CHI is relatively low, suggesting that it may be the rate limiting enzyme in flavonoid biosynthesis. Under the catalysis of F3H, naringenin is converted to dihydrokaempferol, which is further transformed into kaempferol by the action of FLS. Finally, kaempferol is glycosylated to form kaempferol-3-O-glucoside. Our work revealed that F3H, FLS, as well as their metabolic products kaempferol and kaempferol-3-O-glucoside, are all highly expressed in the leaves, indicating that this pathway in leaves is significantly activated. Under the catalysis of F3’H, naringenin and dihydrokaempferol are converted to dihydroquercetin, which is then transformed into quercetin by FLS. Subsequently, quercetin serves as the precursor for the biosynthesis of various quercetin derivatives (such as quercetin 3-β-D-glucoside and rutin). Overall, we combined non-targeted metabolomics, targeted metabolomics, transcriptomics, and RT-qPCR to analyze and validate key metabolites and biosynthesis related genes of flavonoids. We investigated the high expression of flavonoid metabolites such as rutin, quercetin 3-β-D-glucoside, and kaempferol-3-O-glucoside in the leaves of Paris polyphylla Sm., and explored the biosynthesis pathways of flavonoids involving CHI, CHS, and FLS. Studying various stress factors such as chemical, physical, and biological factors on plant biological development and substance metabolism is currently one of the hot topics in research [[144]50]. The biosynthetic pathway of flavonoids in Paris polyphylla Sm. explored in this work provides a theoretical basis for further research on stress factors related to it. Certainly, our work has several limitations. For instance, due to database constraints, the non-targeted metabolomics approach failed to capture important components specific to Paris polyphylla Sm. (such as saponins) [[145]51]. Additionally, pharmacological investigations on the key DMs were lacking. These aspects require further refinement in future studies. Conclusions This work found that flavonoids are present in high concentrations in the leaves of Paris polyphylla Sm., among which rutin, quercetin 3-β-D-glucoside, and kaempferol-3-O-glucoside have the potential for industrial extraction. In the study of flavonoid biosynthesis in Paris polyphylla Sm., it was found that CHI, CHS, and FLS play important roles in it. This work provides insights into the potential utilization of non medicinal parts of Paris polyphylla Sm., especially leaves, in medicine or nutritional supplements, and promotes their resource development and utilization. Supplementary Information [146]Supplementary Material 1.^ (2.2MB, docx) [147]Supplementary Material 2.^ (12.4MB, zip) Acknowledgements