Introduction

Nitraria berry, possessing unique physiological characteristics, is a distinctive berry shrub in the Qinghai-Tibet Plateau. Nitraria berry is a traditional Chinese medicine, documented in “Medicinal Plants in Desert Areas of China” and is mainly used as folk medicine. It is called “Kami” in the “Dictionary of Traditional Chinese Medicine” which is harvested when ripe and then dried in the sun as medicine1. In the field of pharmacy, there are three species of plants that have received more attention: Nitraria tangutorum Bobr. (NT), Nitraria sibirica Pall. (NS), and Nitraria roborowskii Kom. (NR), whose berries have significant differences in shape and color. In addition, there are differences in plant distribution, with NT mostly growing in desert areas, NS mostly growing in saline-alkali lowlands or arid mountain slopes, and NR mainly distributed in deserts in Central Asia and northwest China, which may also cause differences in the accumulation of metabolites2. At present, NT pays more attention to the economy, while NR and NS are neglected, leading to the berries mainly ate by cows, sheep, and camels, resulting in the underutilization of the resource.

Nitraria, a bioactive compound-rich medicinal plant, is recorded in the “Chinese Materia Medica”. NS (the whole herb) is rich in alkaloids, including nitraramine, L-vasicinone, and sibirine. NT (the seeds) are rich in flavonoids, including quercetin, isorhamnetin, and kaempferol-7-O-α-L-rhamnoside. NR has no recorded characteristic bioactive compounds3. Residents frequently use dried ripe berries to treat illness of the stomach, spleen, dyspepsia and colds4. Phenolic acids, anthocyanins, flavonoids and alkaloids are attractive in medicinal chemical constituents. Pharmacological studies indicate that Nitraria berries have the effects of decreasing lipoidemia, improving visual function, inhibiting platelet aggregation and antioxidation5,6. In addition, studies on the fatty acid content in NT and NS berries showed good development prospects7,8.

The multiple origins of Chinese medicine are key features of traditional Chinese medicine. Various related plants or animals of the same or different family or genus are recognized as the source material for one Chinese medicine used to treat the same diseases. Ensuring the safety and efficacy of clinical drugs is paramount, mainly when dealing with multiple origins of Chinese medicine that vary significantly in composition across different varieties. Comparing the chemical constituents of Chinese medicines is crucial for maintaining consistent quality and therapeutic outcomes. NT, NS and NR have a long history of folk medicines. Often growing together, these three species are frequently mixed in traditional Chinese medicine prescriptions and unclear differences in their chemical composition, which may affect their therapeutic effects. In addition, there is limited systematic research on the chemical composition of Nitraria berry. Some literature has used LC-MS to identify 9 chemical substances in NT9, but there are few reports on the chemical constituents of NS and NR. Besides, the research on the chemical composition differences of the three species of berries has yet to be reported. Therefore, to clarify the fundamental differences in active substances among the three species of Nitraria berries and to facilitate their differentiated development and utilization as medicinal resources, it is essential to investigate the variations in active components at a holistic level and explore the corresponding metabolic mechanisms of these distinct components.

Plant metabolomics based on high-resolution LC-MS provides a favorable overall solution for basic research of medicinal plant chemicals10. It can more accurately reveal the correlation between plant phenotypic characteristics and biodiversity. RNA-Seq has been proven to be a powerful method for understand gene complexity and is widely used to study gene expression and regulatory networks during berry ripening11. Thus, this method can be used to analyze and identify the key genes related to important secondary metabolites in the development and maturation of berries. The joint analysis of transcriptomics and metabolomics can better discover the key metabolic pathways, key regulatory genes and molecular mechanisms to explain the accumulation of phytochemical components.

In this study, the UHPLC-QE-MS technical platform was applied to obtain the comprehensive metabolic profiles of Nitraria berries. Afterward, the non-targeted metabolomics method was adopted to screen the different components of three Nitraria berries comprehensively and analyze them from the perspective of medicinal value, providing theoretical support for developing drugs using Nitraria berries as raw materials. By combining analysis of the transcriptomics and metabolomics, key metabolic pathways and genes were found to explain the accumulation mechanism of different metabolites (DMs) between three species of Nitraria berries. This paper aims to provide a theoretical basis and development suggestions for the comprehensive utilization and in-depth development of Nitraria berries.

Materials and methods

Statement

All the plants used in this study were wild and all the methods were complied with the IUCN Policy Statement on Research Involving Species at Risk of Extinction and the Convention on the Trade in Endangered Species of Wild Fauna and Flora.

Plant material

Ripe fruits of three Nitraria species (NT, NS, and NR) were collected in September 2021 from sandy habitats in Wulan County, Qinghai Province (36°58′E, 98°16′N), at an altitude of approximately 3000 m. Nitraria is an annual fruit-bearing plant, with a flowering period from May to June and a fruit ripening period from July to September. The three species were identified based on leaf morphology and berry characteristics: NR exhibits a creeping growth habit, with both leaves and berries covered in hairs; NT has moderately hairy leaves and hairless berries; and NS is characterized by an upright growth habit, small hairless berries, and smooth leaves.

All the berries were immediately frozen in liquid nitrogen and stored at − 80 °C. A total of 3 groups of samples were collected, with 6 samples in each group for metabolomics analysis, and then 6 samples were randomly mixed into 3 repeats equally for transcriptomic determination.

Reagents and instruments

LC-MS grade acetonitrile, methanol, were attained from CNW Technologies (Duesseldorf, Germany). 2-Chloro-L-phenylalanine (internal standard in LC-MS analysis) was purchased from Hengbai Biotechnology Co., Ltd. (Shanghai, China). 7 types of standards: kaempferol-3-O-rutinoside, rutin, diosmin, vitexin, chlorogenic acid, quercetin, kaempferol, purchased from Sigma Aldrich Co. (St. Louis, MO, USA). The purity of all standard products is greater than or equal to 97%.

RNA purity was assessed using a NanoPhotometer spectrophotometer (IMPLEN, Germany), while RNA concentration was quantified using a Qubit 2.0 Fluorometer (Life Technologies, USA). RNA integrity was evaluated using an Agilent 2100 Bioanalyzer (Agilent, USA).

Metabolomic analysis

Samples preparation

50 mg of sample was weighed into an eppendorf tube and 50 mL of extraction solvent was added (acetonitrile/methanol/water = 2:2:1, containing internal standard), the samples were homogenized at 45 Hz for 4 min, ultrasonically extracted for 5 min in an ice-water bath and vortexed for 30s. The homogenate and sonicate steps were repeated 3 times, followed by incubation at -20°C for 1 h and centrifugation at 12,000 rpm and 4°C for 15 min. The resulting supernatants were transferred to LC-MS vials and stored at -80°C until the UHPLC-QE-MS analysis. The quality control (QC) samples were prepared by mixing an equal aliquot of the supernatants from all the samples.

Establishment of a database of chemical components in berries

To ensure the accuracy and completeness of peak identification, we reviewed berry-relevant literature, combining multiple databases, including Traditional Chinese Medicine System Pharmacology Database (https://old.tcmsp-e.com/tcmsp.php), Traditional Chinese Medicine Integrated Database (http://www.megabionet.org/tcmid), Chem Spider (http://www.chemspider), Chemical Book (http://www.chemicalbook.com). We compiled the names, molecular formula, molecular weight, structural formula, and other information of over 1000 compounds collected into an Excel file, established a database of secondary metabolic chemical components of berries.

Metabolomic analysis conditions

LC-MS/MS analysis was performed using a combination system of UltiMate 3000 and Q-Exactive Orbitrap MS (Thermo, MA, USA). The analytical parameters were configured as follows: UHPLC: Column: HSS T3 column (2.1 mm × 100 mm, 1.8 μm); Solvent System: The mobile phase A was water (0.1% formic acid) and the mobile phase B was acetonitrile/methanol (v/v: 15/85); Gradient Program: 0 min, 1% B; 10 min, 1% B; 25 min, 10% B; 30 min, 10% B; 40 min, 15% B; 60 min, 20% B; 70 min, 20% B; 90 min, 25% B; 110 min, 100% B; Flow Rate: 0.3 mL/min; Temperature: 40℃; Injection Volume: 2 µL.

ESI source conditions were set as follows: sheath gas flow rate as 13.5 L/min; aux gas flow rate as 4.5 L/min; capillary temperature 320°C; full MS resolution as 70,000; MS/MS resolution as 17,500; collision energy as 20/40/60 eV in NCE model; spray Voltage as 3.8 kV (positive mode) or -3.1 kV (negative mode), respectively.

Metabolite profiling analysis

The collected RAW data were imported into Compound Discoverer 3.0 software and subjected to baseline correction, peak calibration, deconvolution analysis, etc. Set the molecular weight error to 5 ppm and the retention time error to ± 0.1. After importing the peak areas of primary characteristic ions, adjusted with internal standards, into SIMCA-P 14.1 (Umetrics, Sweden) software, the OPLS-DA model was analyzed to obtain variable importance in projection (VIP) values and calculate the p-values between datasets using the T-test. Compounds that result in VIP > 1.0 and p < 0.05 are used as DMs. The variability was assessed in differential accumulation metabolites by calculating the fold-change values. Analyze the differential accumulation metabolites corresponding to metabolic pathways using the Kyoto Encyclopedia of Genes and Genomes (KEGG, http://www.genome.jp/kegg) and MetaboAnalyst platform (https://www.metaboanalyst.ca/).

Quantitative analysis conditions

Detailed conditions are as follows: Solvent System: The mobile phase A was water (0.9% acetic acid) and the mobile phase B was methanol; Gradient Program: 0–9 min, 20-100% B; 9–10 min, 100% B; 10–11 min, 100%-20% B; 11–14 min, 20% B; Flow Rate: 0.3mL/min; Temperature: 30℃; Injection Volume: 1 µL. Mass spectrometry condition reference metabolomic analysis conditions.

Transcriptome analysis

RNA extraction, library construction and sequencing

The extraction of total RNA was performed with three biological replicates from each group. According to the manufacturer’s protocol, the total RNA of Nitraria berries was extracted using Trizol reagent kit. The RNA purity was checked using the NanoPhotometer spectrophotometer. A Qubit RNA Assay Kit in a Qubit 2.0 Fluorometer was used to measure the RNA concentration. RNA integrity was evaluated using an Agilent 2100 Bioanalyzer. The degradation and contamination were tested on a 1% agarose gel. The construction of the library and the RNA-Seq were performed on Illumina HiSeq™ 4000 by Gene Denovo Biotechnology Co., Ltd (Guangdong, China). Paired-end 150 bp sequencing technology was employed to sequence DNA fragments, with a library insert length of 300 bp.

Functional annotation and differential genes analysis

To get high-quality clean reads, reads were further filtered by fastp v0.18.012. Functional annotation was mainly based on the following databases: NCBI non-redundant protein database (Nr, http://www.ncbi.nlm.nih.gov), Swiss-Prot database (http://www.expasy.ch/sprot), the KEGG and the COG/KOG database (http://www.ncbi.nlm.nih.gov/COG). The identification of DEGs was performed by DESeq2. The genes with the parameters of false discovery rate (FDR) < 0.05 and | log2(FC) | ≥ 1 were considered as DEGs. All the DEGs were mapped to the KEGG pathway database. The formula of the hypergeometric test calculated the significance of DEGs enrichment.

Correlation analysis of the metabolomic and transcriptomic data

We calculated the Pearson correlation coefficients between DMs and DEGs. Building on this, we further integrated pathway localization analysis, thereby establishing methodological framework for the integrative analysis of genes and metabolites.

Statistical analysis

Using SPSS Statistics 25.0, the statistical significance was calculated using Student’s t-test. The significance level was set as p < 0.05 (significant) and p < 0.01 (extremely significant). Fold change was calculated to show the changes in metabolite levels and gene expression.

Results

Phenotype analysis of three kinds of Nitraria berries

The Nitraria berries of three high-altitude species are shown below: NT (Fig. 1A), NS (Fig. 1B) and NR (Fig. 1C), the images clearly shows the color differences in ripened berries: deep purplish-red, dark red and purple-black, respectively. The NR berry is larger than the other two species and is uniquely covered with fine white hairs on its peel. In terms of shape, the NT berry is more like oval, while the other two are more similar circles. These distinct phenotypic differences reflect varied metabolite and gene expression levels across these species of berries.

Fig. 1
figure 1

Morphological features of (A) N. tangutorum Bor. (NT), (B) N. sibirica. (NS), (C) N. roborowskii. (NR).

Full size image

Metabolic profiling

The non-targeted metabolomics was performed using the LC-MS method to analyze the DMs of three Nitraria berries. The identification process for unsuspected compounds is as follows: the collected primary mass spectrometry RAW data was imported into Thermo Scientific Compound Discoverer 3.0 software, by automatic screening for the established self-built berry database, it was found that compounds with the matching molecular formula are potential candidate compounds. A list of compounds was manually created with a matching score greater than 85.00, including their retention times and accurate precursor ion masses. Based on the list, a targeted MS/MS detection was carried out. The retention time, accurate mass with an error of less than 5 ppm and reasonable fragmentation pathways were used to identify the candidate compounds, and verified by available standards. Moreover, using the online database Mzcloud, we further supplement and validate the search results of the self-built database. Then, we plotted the extraction ion flow chromatography of these compounds (Fig. S1). There were 103 compounds, including 22 anthocyanins4,13,14,15, 30 flavonoids13,16,17,18, 25 phenolic acids13,17,18, 9 organic acids19, 13 amino acids16,18 and 4 sugars. The detailed identification result and secondary fragment information is (Table S1) arranged.

The metabolite composition data sets were analyzed using the principal component analysis (PCA) (Fig. 2A). The sample of NS, NT and NR showed clear separation and the sample of QC showed close together, the stability and reliability of instrumental analysis were ensured. Then, VIP ≥ 1 and p-value < 0.05 was used to determine significantly differential metabolites between groups (Fig. 2B). In the comparison of group NT vs. NS, there were 12 up-regulated metabolites and 14 down-regulated metabolites. In the comparison of group NS vs. NR, there were 29 up-regulated metabolites and 8 down-regulated metabolites. In the comparison of group NT vs. NR, there were 24 up-regulated metabolites and 13 down-regulated metabolites. The screening results of DMs (Table 1) of three species of Nitraria berries are arranged.

Fig. 2
figure 2

The PCA score plot compars metabolic profiles of NR, NT and NS (A), the number of up-regulated and down-regulated DMs (B), Metabolic pathway enrichment analysis of DMs across three Nitraria species berries (C). (a) Anthocyanin biosynthesis, (b) Flavone and flavonol biosynthesis, (c) Fatty acid biosynthesis.

Full size image
Fig. 3
figure 3

Content of partial differential metabolite in NR, NT and NS (*p < 0.05, **p < 0.01).

Full size image
Table 1 Some of key metabolites differentially accumulated between three species Nitraria berries (NT, NS, NR).
Full size table

Pathway enrichment of DMs

Pathway enrichment analysis was performed using the pathway analysis module of the MetaboAnalyst platform (Fig. 2C). The circles, size and color of each circle correspond to the metabolic pathways, pathway’s impact value and p-value, respectively. The pathways with an impact value > 0.1, and p < 0.05 were selected as potential differential metabolic pathways. Most of the pathways were concentrated on anthocyanins, flavonoids and fatty acid metabolism. Anthocyanin and flavonoid biosynthesis are the most significantly different metabolic pathways among the three species of Nitraria berries, mainly involving 4 cyanidins, 2 pelargonidins, 4 petunidins, 3 delphinidins, vitexin, 3 quercetins, homoorientins, 3 kaempferols, 2 isorhamnetins and diosmetin. In addition, multiple fatty acids exhibit significant differences in the metabolism of three species of Nitraria berries, including oleic acid, palmitoleic acid, palmitic acid, linoleic acid and tetradecanoic acid.

Metabolomics validation

To further validate the metabolomics analysis results, selected DMs in Nitraria berries were quantified. The quantitative analysis revealed that kaempferol-3-O-rutinoside and quercetin exhibited the highest content in NS and the lowest in NR. Similarly, rutin, diosmin, and vitexin showed the highest content in NT and the lowest in NR. In contrast, chlorogenic acid demonstrated the highest content in NR and the lowest in NT. These findings (Fig. 3) are consistent with the metabolomics results based on log2(FC) values.

Transcriptomics analysis

After sequencing the cDNA libraries, high-quality clean reads were assembled using Trinity software20. The number of clean reads per library ranged from 38,026,402 to 54,132,504 after filtering out the low-quality data, a total of 65,398 unigenes across three groups. These unigenes had a GC content of 43.61%, an N50 of 17.12%, with an N50 length of 1,535 bp and an average length of 874 bp (Table S2). The length distribution of de novo assembled unigenes (Fig. S2A) is accumulated. The annotation results (Table S3) for all unigenes are shown, with 38,642 genes annotated out of 65,398, representing 59.09% of the total. Venn plot illustrates the annotation results (Fig. S2B), revealing that 19,362 unigenes (50.11% of those annotated) are present in all four databases, while only 3,976 unigenes (10.29%) are exclusive to one database.

PCA results showed that the NR was significantly different compared to the NT and NS, while the NT and NS shown slight difference (Fig. 4A). DEGs in all the paired comparison groups were identified through using DESeq2, with FDR < 0.05 and | log2(FC) | > 1. There were 20,367 DEGs (12,149 up-regulated and 8,218 down-regulated) in NR vs. NS, which was the maximum. Additionally, 5,535 DEGs (4,001 up-regulated and 1,534 down-regulated) were found in NT vs. NS, and 9,545 DEGs (3,135 up-regulated and 6,410 down-regulated) were found in NR vs. NT (Fig. 4B).

KEGG pathway enrichment analysis of DEGs

To determine the biological metabolic pathways and functions of DEGs among the different species of Nitraria we performed functional enrichment analysis. The distribution of DEGs annotated in the primary metabolic pathways of different comparison groups of Nitraria berries is similar, belonging to five categories (Fig. 4C). Among them, the “Metabolism” category contained the most DEGs and also had the most secondary pathways. To further screen the related regulatory genes of DMs, this study analyzed the secondary pathways under metabolism, including carbohydrate metabolism, amino acid metabolism, lipid metabolism, co-factor and vitamin metabolism and biosynthesis of other secondary metabolites. The DEGs involved in the biosynthetic pathways of amino acid and lipid metabolism and other metabolites, are closely related to the regulation of major DMs.

Flavonoids (including anthocyanins) and fatty acid synthesis pathways are the main metabolic pathways corresponding to the significant DMs of the three species of Nitraria berries, involving the biosynthesis of other metabolites and lipid metabolism pathways. Therefore, a pairwise comparison between different species of NR, NS and NT showed that DEGs are mainly concentrated in phenylpropanoid biosynthesis, flavonoid biosynthesis, isoflavonoid biosynthesis and fatty acid biosynthesis. The number of DEGs in each pathway is revealed (Fig. 4C).

Fig. 4
figure 4

The PCA score plot of the transcriptome of NR, NT and NS (A), Volcano plots to filter the DEGs of berries in different species groups (B), The KEGG enrichment analysis and functional classification of DEGs in different species groups (C). KEGG primary classification results of DEGs (a), secondary classification results of DEGs enriched into “metabolic” category (b), tertiary classification results of DEGs enriched into pathways (c).

Full size image

Transcription factors analysis

As a molecular switch, transcription factor (TF) controls various growth and development processes of plants. TFs can activate the expression of multiple genes in the secondary metabolic synthesis pathway, thereby effectively initiating the secondary metabolic pathway. The different types and amounts of TFs in the three comparison groups of different species are shown (Fig. S3). The NR vs. NS had the most significant number of different TFs, including C2H2, ERF, bHLH, MYB, bZIP, NAC and WRKY. The NR vs. NT, TFs included ERF, MYB, bHLH, NAC, C2H2, bZIP and HD-ZIP. The NT vs. NS had the minimum number of different TFs, including MYB, bHLH, ERF, C2H2, bZIP, B3 and Dof. Although the differences of TF involved in the three control groups were not entirely the same, significant differences were found in 5 TF families, including C2H2, ERF, bHLH, MYB and ZIP. This indicates that these 5 TF families may play a crucial role in regulating the accumulation of metabolites in Nitraria berries.

Joint analysis of metabolomics and transcriptomics

This study utilized non-targeted metabolomics to comprehensively screen for differentially accumulated metabolites in different species of Nitraria berries. The results show that flavonoids (anthocyanins) are the most significant DMs. Transcriptomic analysis identified DEGs, which were functionally classified and enriched KEGG pathway analysis. This approach revealed multiple genes key enzymes encoding flavonoids (including anthocyanins) were screened, including CHS, F3H, DFR, ANS, FLS. On this basis, a metabolic network diagram (Fig. 5) of flavonoids and corresponding structural genes in Nitraria berries was constructed, and the main differential accumulation metabolites and structural genes conducted relevant analysis.

Naringin is a precursor of flavonols, anthocyanins, proanthocyanidins, flavonoids and isoflavones. During flavonoid biosynthesis, it is hydroxylated to dihydrokaempferol by naringin 3-dioxygenase (F3H), and then a double bond is introduced to form kaempferol by flavonol synthase (FLS). Kaempferol is transformed into quercetin by flavonoid 3’-monooxygenase (F3’H). Quercetin then becomes quercetin-3-O-glucoside through glucosyltransferase (E2.4.1.91), which is finally converted to rutin by flavonol-3-O-glucoside L-rhamnosyltransferase (FG2). In the NS vs. NR and NT vs. NR groups, genes encoding key enzymes in the flavonoid biosynthesis pathway include: 5 PAL (Unigene0062152, Unigene0063772, Unigene0063773, Unigene0000017, Unigene0020003); 2 4CL (Unigene0034164, Unigene0065010); CHI (Unigene0016893), 2 F3’H (Unigene0034447, Unigene0052438). All of these are significantly upregulated, which may be the main reason for the higher flavonoid content in NT and NS compared to NR.

Dihydrokaempferol is converted to dihydromyricetin by flavonoid 3’,5’-hydroxylase (F3’5’H), which is then turned into leucodelphindin by dihydroflavonol 4-reductase (DFR). Leucodelphindin is transformed into delphinidin by the anthocyanin synthesis (ANS) enzyme, which then undergoes methylation to become petunidin. Subsequently, petunidin is converted into its glycosides by UDP-glucose flavonoid 3-O-glucosyltransferase (UFGT) and other enzymes. In the compersion of NT vs. NR and NS vs. NR groups, 2 C4H (Unigene0051686, Unigene0051683); F3’5’H (Unigene0038667); 2 DFR (Unigene001763, Unigene0051889); ANS (Unigene0034286) and 2 UFGT (Unigene0034286, Unigene0053979), showed positive correlations with petunidins expression. However, these structural genes were lowly expressed in both NS and NT conditions. It is speculated that under the combined influence of these factors, the content of petunidins in NR is higher than that in NT and NS.

Dihydrokaempferol is converted to taxifolin by F3’H enzymes, which is then transformed into leucodelphindin by DFR. Finally, leucodelphindin becomes cyanidin under the influence of ANS. Cyanidin is converted into its glycosides by glycosyltransferases like UFGT. In the NS vs. NR and NT vs. NR groups, includes 5 PAL (Unigene0062152, Unigene0063772, Unigene0063773, Unigene0000017, Unigene0033089) and 2 4CL (Unigene0034164, Unigene0065010), which are positively correlated with the expression of the substance and these structural genes are highly expressed in NT and NS. In addition, the content of cyanidins in NT is higher than that in NS, with C4H (Unigene0051686), 3 CHS (Unigene0054891, Unigene0054892, Unigene0008745) and ANS (Unigene0034286) structural gene expression was upregulated, consistent with the accumulation of this compound.

Fig. 5
figure 5

Flavonoid biosynthetic structural genes of DEGs and DMs in different species groups. According to log10(FC): red means up, blue means down.

Full size image

Analysis of the interaction between flavonoid synthesis genes and TFs in different species of Nitraria berries shows that the NS vs. NR group had the highest number of screened TFs, including 41 C2H2, 36 ERF, 27 bHLH, 27 MYB, 25 bZIP, 18 NAC and 13 WRKY. The protein-protein interaction of TFs and synthetic genes (Fig. S4) revealed that 2 WRKY (WRKY24, WRKY75), 7 MYB (MYB114, MYB4, MYB7, MYB5, MYB3, MYB90, MYB58), bHLH (bHLH121), 2 NAC (NAC055, NAC029) and 4 ERF (ERF113, ERF114, ERF1B, ERF106) are directly or indirectly involved in regulating structural genes CHI, FLS, F3’5’H, ANS, 4CL, UFGT and CHS. From the NT vs. NR group TFs, 29 ERF, 25 MYB, 22 bHLH, 19 NAC, 16 C2H2 and 10 bZIP were screened. The protein-protein interaction of TFs and synthetic genes revealed that 7 MYB (MYB113, MYB4, MYB305, MYB5, MYB90, MYB114, MYB61) and WRKY 75 regulate flavonoid synthesis structural genes FLS, CHS, CHI, PAL and UFGT. NS vs. NT group had the least differential TFs, including 14 MYB, 13 bHLH, 12 ERF, 9 C2H2 and 7 bZIP. The protein-protein interaction of TFs and synthetic genes revealed that 2 WRKY (WRKY44, WRKY46), 6 MYB (MYB44, MYB305, MYB61, MYB5, MYB3, MYB308), and bHLH co-regulate FLS, ANS, CHS, CHI, F3’H and other flavonoid synthesis structural genes.

The correlation analysis of flavonoid TFs in different species of Nitraria berries is revealed (Fig. S5). In the NS vs. NR group, Cyanidins, delphinidins, quercetins, rutins and kaempferols were positively correlated with bHLH, ERF, WRKY, MYB and NAC. Peonidins and petunidins were positively correlated with bHLH, ERF, MYB, NAC and WRKY. In the NT vs. NR group, a similar pattern exists, cyanidins, delphinidins, quercetins, rutins and kaempferols were also positively correlated with MYB and WRKY. Peonidins and petunidins were positively correlated with MYB. In the NT vs. NS group, petunidins were positively correlated with WRKY and MYB.

Discussion

Nitraria berry is a traditional Chinese medicine. The research showed that NT contains β-carboline alkaloids for the treatment of arrhythmia, spasms, and neuropathic pain21. Zhang et al.4. found that anthocyanins in NT can protect H9c2 cardiomyocytes damaged by doxorubicin by directly clearing ROS. We first evaluated the lipid-lowering effect of phenolic compounds in NR in vivo, and the results showed that polyphenols mainly exert therapeutic effects by regulating glycerophospholipids and glycerol metabolism22. At present, the research on the chemical composition differences of the three species of berries has yet to be reported. Therefore, this paper uses the method of metabolomics and transcriptomics to reveal the difference in chemical composition. The results show that flavonoids (anthocyanins) are key compounds to the metabolic variations in Nitraria berries.

Anthocyanins are water-soluble pigments belonging to the flavonoid class and are a type of polyphenol in plants. According to pH, chemical changes and metal ions, they show a wide range of colors from light pink to dark purple in plant leaves, flowers, berries and seeds23. In this study, we observed that the anthocyanin composition of the three Nitraria berries is similar, encompassing petunidin, cyanidin, pelargonidin, and delphinidin. However, significant differences in their content were noted: petunidin anthocyanins were prominently accumulated in NR, whereas cyanidin, pelargonidin, and delphinidin were predominantly enriched in NT and NS. Among these compounds, the variations in petunidins and cyanidins content were the most pronounced. The petunidins mainly involving two compounds: petunidin 3-O-rutinoside glucose and petunidin-3-O-(6”-O-coumaroyl) glucose. Petunidins exhibit significant biological activity. Tang et al. research has shown that petunidins have strong free radical scavenging ability, which has a potential protective effect on intracellular reactive oxygen species damage induced by hydrogen peroxide in rat neuroid pheochromocytoma cell lines24. The research results of Zhang et al. show that petunidin-3-glucoside has an anti-inflammatory effect on acute gout foot swelling25. The study results of Nagaoka et al. showed that oral petunidin improved sRANKL induced bone loss in mice by increasing osteoid formation, accelerating osteoblast formation and inhibiting bone resorption26. NR contains abundant petunidin components. However, there currently needs to be more research reports on the biological activity of NR, which seriously limits the development and utilization. Furthermore, the accumulation of anthocyanins in NT and NS is similar, cyanidins showed the most significant fold changes, revealing that these compounds are the most characteristic active component of NT and NS. There are many reports on the activity of cyanidins, and the mechanism research is also relatively in-depth. The compound cyanidin-glucoside is widely distributed in berries, and studies have shown that it has significant antioxidant23, anticancer27, anti-inflammatory and cell protective effects28,29, as well as cardiovascular protective effects30,31. In vivo, Cyanidin-glucoside exerts a cellular protective effect to counteract oxidative stress-induced DNA damage29. Cyanidin-3-O-galactoside is also one of the most widely distributed anthocyanins in fruits, with various biological activities, including antioxidant32, anti-inflammatory33, and anticancer34. In recent years, the significant effect of this compound on cognitive ability has aroused strong interest among researchers35. Based on the above reasons, NT and NS are important raw materials for cyanidins active ingredients, with broad application prospects.

Flavonoids are a large class of secondary metabolites in plants, playing an irreplaceable role in the plant metabolic cycle. They are also an essential active component in Nitraria berries. Flavonoids have strong antioxidant properties, having good effects on improving cancer, atherosclerosis and Alzheimer’s disease36. In this study, we found that the flavonoids of three Nitraria berries are similar in composition, but have significant differences in content, with the isorhamnetin was significantly accumulated in NR, whereas other flavonoids (vitexin, homorientin, rutin, quercetin, kaempferol, and diosmetin) were highly accumulated in NT and NS compared to NR. Among this compounds, the changes in the content of isorhamnetins and kaempferols are the most significant. Isorhamnetins, including isorhamnetin 3-O-neohesperidoside and isorhamnetin-3-O-galactide, this type of substance is commonly found in the leaves, flowers and fruits of Hippophae rhamnoides L., ginkgo and other plants37, and is the first reported in NR. The pharmacological effects and mechanisms of these compounds are currently a hot research topic. Studies have shown that isorhamnetins have extensive pharmacological effects on cardiovascular diseases37 and various tumors38 and have the potential to prevent neurodegenerative diseases such as Alzheimer’s disease39. As an important and new source of isorhamnetin, NR has broad development prospects. In addition, research results indicate that the accumulation of flavonoids in NT and NS is similar, but there is a significant difference compared to flavonoids in NR. The most important changes were observed in the multiples of kaempferols, including three compounds: kaempferol, kaempferol-3-O-rutinoside and kaempferol-3-O-rhamnosylactioside-7-O-glucoside. Kaempferol is a type of flavonol widely present in tea, cauliflower, apples, strawberries and legumes40. Many studies have described the beneficial effects of dietary kaempferol in reducing the risk of chronic diseases, especially cancer. Epidemiological studies show that there is a negative correlation between kaempferol intake and the incidence rate of cancer41,42,43. It is worth noting that kaempferol not only inhibits the growth and angiogenesis of cancer cells and induces apoptosis, but also has a protective effect on normal cells, maintaining cellular viability, which is significance for cancer treatment44. NT and NS, as important sources of kaempferol, can be further developed and utilized.

Metabolomics analysis in this study showed that the contents of most flavonoids and anthocyanins in NS and NT were higher than those in NR. 38 flavonoid synthesis structural genes encoding 12 key enzymes were screened from DEGs. The expression patterns of 10 structural genes were up-regulated in NS vs. NR and NT vs. NR groups, consistent with the change of flavonoid content. The 10 genes include 5 PAL, 2 4CL, 1 CHI and 2 F3’H, most of them are upstream regulatory genes of flavonoids synthesis.Their high expression leads to significant accumulation of flavonoids, so that the content of most flavonoids in NT and NS is higher than in NR. Among them, the number of PAL genes encoding the synthesis of phenylalanine aminolyase was the largest, and a total of 5 homologous genes were significantly up-regulated, which may be the main reason causing the accumulation of flavonoids. PAL is the first key enzyme in the synthesis of flavonoids and plays a vital role in the synthesis of a variety of plant flavonoids45. Studies on PAL in different varieties (e.g., soybean, wheat, rice, potato, tomato and Arabidopsis) found significant differences in PAL expression among different varieties46, further supporting our speculation. In addition, DFR (Unigene0017632), ANS (Unigene0034286 and UFGT (Unigene0064801) were highly expressed in NR, and the changes in the multiple of UFGT were the most significant, which was consistent with the accumulation of petunidins in NR. Therefore, we speculate that UFGT is the leading cause of the significant accumulation of petunidins in NR. NT and NS had similar flavonoid composition but higher overall NT content. Compared with NT vs. NS, most structural genes of flavonoids were highly expressed, consistent with the accumulation of flavonoids.

Analysis results showed that MYB, WRKY and bHLH were also involved in regulating the structural genes of flavonoids in different varieties of Nitraria berries. These TFs are regulated by activating or inhibiting structural genes of flavonoid synthesis. In summary, for the first time, this study conducted a comprehensive comparative analysis of three kinds of Nitraria berries from two levels of transcriptomics and metabolomics. It provided necessary theoretical support for understanding the differences in chemical components of three Nitraria berries and a reference for distinguishing and rational development and utilization of three Nitraria berries.

Conclusion

In this study, the integration of metabolomics and transcriptomics data reveals that DMs and DEGs are mainly enriched in the biosynthetic pathways of anthocyanins and flavonoids. The metabolomics results show that the petunidins and isorhamnetins were significantly accumulated in the NR, while cyanidins and kaempferols were significantly accumulated in the NT and NS. The results of transcriptomics show that multiple homologous genes of structural genes PAL, 4CL, CHI, and F3’H were the main reasons for the differences in anthocyanins and flavonoids metabolism in different species of Nitraria berries. Based on this, a network metabolic map for flavonoids was constructed, and the acorrelation analysis was performed between major DMs and DEGs. In summary, for the first time, this study conducted a comprehensive comparative analysis of three kinds of Nitraria berries from two levels of transcriptomics and metabolomics. It provided necessary theoretical support for understanding the differences in chemical components of three Nitraria berries and a reference for distinguishing and rational development and utilization of three Nitraria berries.