1. Introduction
Oxidation is the process of producing free radicals, which often contribute to aging and the development of chronic diseases, and it can be slowed down or even inhibited by antioxidants [
1]. At present, natural antioxidants from plants have attracted significant interest in human health care as well as disease prevention and have found wide applications in food, drugs, additives, cosmetics, and other fields [
2,
3]. Citrus, belonging to the genus
Citrus L. of the family
Rutaceae, is an important fruit crop cultivated extensively worldwide with significant production [
4]. Citrus is enriched with natural antioxidants in addition to vitamin C in the flesh and abundant secondary metabolites distributed in other tissues, such as phenolic acids, flavonoids, coumarins, limonoids, alkaloids, carotenoids and essential oils, all of which possess notable antioxidant capacity [
5]. These secondary metabolites participate in important physiological processes such as stress resistance, disease resistance and signal transduction in plants [
6], and are also vital sources of dietary antioxidants.
Citrus have been used as traditional medicinal herbs in Asian countries [
5], with various parts, including peels, leaves, and seeds, being used for medicinal purposes. Currently, several studies have compared the antioxidant function of different tissues of citrus, finding that the peel and seed showed more prominent antioxidant capacity than the pulp [
7]. However, the antioxidant capacity of roots, stems, leaves, and other parts of the fruit have seldom been assessed. The antioxidant capacity of different tissues is intricately linked to the accumulation of bioactive components. Secondary metabolites, known for their diverse bioactive functions, demonstrate a spatially specific distribution in citrus. For example, polymethoxylated flavones (PMFs) have been widely detected in the flavedo and limonoids have been found in the segment membranes and seeds, while flavonoids, alkaloids, and coumarins have been detected in the flowers and fruits [
8]. However, a detailed illustration of the secondary metabolome profile in the citrus plant is still lacking. Due to processing or pruning every year, there are many by-products of citrus, such as peels, seeds, pomace, branches, and leaves [
9], all of which are believed to be rich sources of natural bioactive substance and antioxidants. Utilizing these by-products scientifically could reduce waste and further enhance the health care potential of citrus.
At present, the biosynthesis pathways of secondary metabolites in plants have been extensively studied, including flavonoids [
10], coumarins [
11], phenolic acids [
12], terpenoids [
13], and alkaloids [
14]. However, knowledge of the biosynthesis pathway of secondary metabolites in citrus is still incomplete, especially the transcriptional regulation mechanism, which remains unclear. The combination of metabolome and transcriptome analysis is recognized as an effective approach to compare the metabolic diversity between samples as well as to reveal related molecular mechanisms [
15]. Weighted gene co-expression network analysis (WGCNA) is an efficient tool for identifying cluster of genes with similar functions based on their correlation, thus facilitating the construction of predictive regulatory networks [
16]. WGCNA combined with metabolome data is a powerful tool for identifying the genetic basis of the specific accumulation of secondary metabolites in citrus.
In this study, we evaluated and compared the antioxidant capacity of 12 tissues from Citrus reticulata ‘Ponkan’ (hereinafter referred to as ‘Ponkan’), a widely recognized citrus variety in China, using four chemical assays. Through metabolome analysis, we outlined a comprehensive metabolic profile of citrus plant and analyzed possible secondary metabolites that are related to antioxidant functions in different tissues. Through combined transcriptomic and metabolomic analysis as well as WGCNA, we excavated genes related to the accumulation and regulation of citrus secondary metabolite synthesis. Our research will contribute to the full utilization of citrus resources and the development of the medicinal value of citrus plants.
2. Materials and Methods
2.1. Plant Materials
Three whole Ponkan plants were harvested in Quzhou, Zhejiang Province. As shown in
Figure 1A, the tissues including flavedo (FL), albedo (AL), tangerine pith (TP), segment membranes (SM), juice sacs (JS), seeds (SE), main stems (MS), side branches (SB), tender bines (TB), mature leaves (ML), young leaves (YL), and roots (RO) were isolated from various parts of each plant. Three plants served as three biological replicates for each tissue sample. The samples were rapidly frozen in liquid nitrogen and stored at −80 °C, then lyophilized by vacuum freeze-dryer (Scientz-100F, Scientz, Ningbo, China) and ground using a mixer mill (MM 400, Retsch, Haan, Germany) with a zirconia bead (1.5 min, 30 Hz). A 100 mg quantity of lyophilized powder was added to 1.2 mL 70% methanol solution, and then vortexed for 30 s per 30 min for 6 times in total, then finally placed at 4 °C overnight. The extracts were collected by high-speed centrifugation (12,000 rpm, 10 min) and filtrated through filtration membrane before UPLC-MS/MS analysis.
2.2. In Vitro Chemical Antioxidant Capacity Evaluation
Ponkan extract preparation: For each tissue, 0.2 g of fresh sample powder was accurately weighed and combined with 2 mL methanol. Ultrasonication-assisted extraction was performed for 30 min followed by centrifugation (8000 rpm,10 min), repeated twice, and the supernatants were combined. Four chemical antioxidant capacity evaluation assays were employed to evaluate the in vitro antioxidant capacities of Ponkan extracts, including 2,2-diphenyl-1-picrylhydrazyl (DPPH) free radical scavenging assay, ferric ion-reducing antioxidant power (FRAP) assay, 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) radical scavenging assay and oxygen radical absorbance capacity (ORAC) assay, and the detailed operation method was referred to Wang et al. [
17]. The test of each sample repeated three times and all the results were expressed as mg Trolox equivalent antioxidant capacity (TEAC)/g FW (fresh weight).
2.3. Metabolite Profiling Using UPLC-MS/MS
The metabolome analysis was performed by Wuhan Metware Biotechnology Co., Ltd., Wuhan, China, following the methods created by Chen et al. [
18]. The sample extracts were analyzed using a UPLC-ESI-MS/MS system (UPLC, SHIMADZU Nexera X2, SHIMADZU, Kyoto, Japan; MS, Applied Biosystems 4500 Q TRAP, AB SCIEX, Framingham, MA, USA) with the following specifications: UPLC column, Agilent SB-C18 (1.8 µm, 2.1 mm × 100 mm); solvent A, water (0.1% formic acid); solvent B, acetonitrile (0.1% formic acid); gradient program, 0/95, 9/5, 10/5, 11.1/95 and 14/95. (min/A%); flow rate, 0.35 mL/min; injection volume, 4 μL. LIT and triple quadrupole (QQQ) scans were acquired on a triple quadrupole–linear ion trap mass spectrometer (Q TRAP), AB4500 Q TRAP UPLC/MS/MS System, equipped with an ESI Turbo Ion-Spray interface, operating in positive and negative ion mode and controlled by Analyst 1.6.3 software (ABSCIEX, Framingham, MA, USA). The reference database for metabolite identification was Metware Database (MWDB), created by Wuhan Metware Biotechnology Co., Ltd. The identification of metabolites was synthetically based on the precise mass of metabolites, the information of MS2 fragments, the isotopic distribution of MS2 fragments, and the retention time (RT). Through the intelligent matching method independently developed by Metware, the MS2 information and RT of metabolites in the sample were intelligently matched with the database one by one. The MS and MS2 tolerances were set to 20 ppm, and the RT tolerances were set to 0.2 min. The differentially accumulated metabolites (DAMs) were identified based on variable importance in projection (VIP) ≥ 1 and fold-change (FC) ≥ 2 or ≤0.5.
2.4. RNA-Seq Analysis
A total of 36 high-quality RNA samples were used to create sequencing libraries using the NEBNext UltraTM RNA Library PrepKit for Illumina (NEB, Beverly, MA, USA). Then the library sequencing was performed on an Illumina NovaSeq 6000 platform. After data filtering by fastp v0.19.3, the clean reads were aligned to the Citrus clementina v1.0 genome using HISAT v2.1.0. Gene expression level was evaluated by fragments per kilobase of transcript per million fragments mapped (FPKM), which was calculated using feature Counts v1.6.2. DESeq2 v1.22.1 was used to identify differentially expressed genes (DEGs) between different groups, and the Benjamin and Hochberg method was used to correct p-value. Genes with fold-change (FC) ≥ 2 or ≤0.5 and FDR < 0.05 were assigned as DEGs.
2.5. RT-qPCRAnalysis
Total RNA samples were extracted as described above, and cDNA was synthesized using a PrimeScript RT reagent kit with a gDNA remover (TaKaRa, Dalian, China). The qRT-PCR was performed by CFX96™ Real-Time System (C1000™ thermal cycler, Bio-Rad, Hercules, CA, USA). Primers were designed in NCBI Primer-BLAST (
https://www.ncbi.nlm.nih.gov/tools/primer-blast/, accessed on 1 January 2024). Citrus β-actin was selected as an internal standard, and for each tissue, three replicates were used to calculated the relative expression level (represented by ΔCt).
2.6. Weighted Gene Co-Expression Network Analysis
WGCNA v1.69 was used for weighted gene co-expression network analysis. The FPKM expression file was filtered using the varfilter function of the genefilter package in R before WGCNA. The soft threshold was set at 19, and a similar module merge threshold was set at 0.25. Genes with a high co-expression connection within the module were filtered, and Cytoscape was used to construct a co-expression network illustration.
2.7. Recombinant Protein Construction and Enzyme Assay
The coding sequences (CDSs) of OMT-2 from Ponkan were cloned into the pET32a expression vector with a histidine tag, followed by the recombinant vectors transferred into BL21(DE3) cells (Promega, Madison, WI, USA). The bacteria were cultured in LB medium containing 0.1g·L−1 ampicillin until the OD600 reached 0.8. The induced expression of recombinant proteins was carried out by the addition of isopropyl-β-D-thiogalactopyranoside (IPTG) for 24 h at 16 °C, then the bacteria cells were collected by centrifugation (4000 rpm, 4 °C, 10 min) and resuspended by 1× PBS buffer. After disruption by a sonicator, the cell debris was discarded, while the supernatant was collected by centrifugation (5000 rpm, 4 °C, 15 min). Recombinant proteins were purified using a HisTALON™ Gravity Columns Purification Kit (Takara, Dalian, China) and finally eluted in Tris-HCl buffer (50 mM, pH 8.0) containing 10% glycerol and 2 mM DTT. SDS-PAGE was used to confirm the target recombinant proteins. The in vitro enzyme assays were performed in a total volume of 200 μL consisting of 1 mM SAM, 200 μM substrates, and 25 μM of purified recombinant proteins in Tris-HCl buffer (50 mM, pH 8.0) at 37 °C for 3h, and finally stopped by adding an equal volume of methanol. After concentration by a rotary evaporator (Eppendorf, Hamburg, Germany), the precipitate was redissolved with 100 μL methanol, and the supernatant collected by high-speed centrifugation (12,000 rpm, 15 min) was ready for LC-MS detection.
HPLC conditions were as follows: detection system, Waters e2625 system equipped with a DAD detector coupled with an ODS C18 column (SunFire 5 μm, 4.6 × 250 mm, Waters, Milford, MA, USA); flow rate, 1 mL/min; solvent A, acetonitrile; solvent B, water (0.1% formic acid); gradient program, 0/20, 3/20, 5/30, 10/30, 15/40, 20/60, 23/80, 25/100, 27/20, 30/20. (min/A%); injection volume, 10 μL; UV detection wavelength, 350 nm. For MS analysis, an AB Triple TOF 5600 plus System (AB SCIEX, Framingham, MA, USA) was used and MS spectra were obtained in negative ion mode or positive ion mode (ESI). The products were identified by the retention time compared with the standards as well as the ion fragments.
2.8. Statistical Analysis
Statistical significance analysis for results 3.1 was conducted using one-way analysis of variance (ANOVA) using SPSS version 23.0 (IBM, Armonk, NY, USA). Principal component analysis (PCA), heatmaps, venn diagrams and volcano pots were conducted using online tools of Metware Cloud (
https://cloud.metware.cn, accessed on 1 January 2024).
4. Discussion
Over the past few decades, interest in identifying natural molecules for use as food supplements has grown, with these plant-derived bioactive substances playing a pivotal role in health care and disease prevention [
2]. Citrus, a popular type of fruit, boasts numerous tissues demonstrating antioxidant potential, making it an ideal source of natural antioxidants. In this study, all the four chemical assays have been widely applied to evaluate the antioxidant capacities in natural product [
21]. Comparative studies of citrus tissue antioxidant capacities have primarily focused on peel and pulp, with the peel generally exhibiting higher levels [
22,
23,
24]. However, our study confirms that leaf tissues, particularly YL, possess the strongest antioxidant activity. Compared to fruits, citrus leaves offer advantages like broader availability and fewer seasonal limitations, indicating significant potential for development. The antioxidant activity of RO was medium to high, which we believe to be closely related to coumarin accumulation. Coumarins, as important phenolic substances, are also known for their outstanding antioxidant capacity [
25]. A gradual increase in antioxidant activity was observed from MS to SB to TB, and from ML to YL, potentially due to the increased accumulation of flavonoids in younger tissues. The antioxidant capacity of the edible parts of citrus plants is relatively weak, but there exist differences between different parts. Abeysinghe et al. found that the antioxidant capacity in segment membranes is higher than that of juice sacs and segments due to the high content of total phenolics [
26], aligning with our findings. In addition, we found that TP is the most antioxidant active part of the edible part, which also explains why it can be used as a traditional medicine. Contrary to previous studies, the seed did not exhibit significant antioxidant capacity in our analysis, despite being reported as rich in tannins and phenols [
7]. This discrepancy may be influenced by factors such as the choice of cultivar and the stage of development.
The biological activities of citrus depend on its abundant metabolites, especially secondary metabolites [
5]. Currently, both targeted and non-targeted metabolomic techniques are extensively used to detect secondary metabolites in different plant varieties [
27], tissues [
28], and development stages [
29], and also have certain applications in citrus research [
8,
30]. In this study, 512 secondary metabolites from 12 Ponkan tissues were identified by widely targeted metabolomics, and the categories of metabolites detected were consistent with previous studies. The number of secondary metabolites detected in this study is much higher than previous studies, which greatly enriched the citrus secondary metabolite pool. In addition, PCA and heatmap analysis showed that the secondary metabolite profile of different tissues of citrus showed its own characteristics, which has a guiding role for the development and utilization of the medicinal value of different tissues.
Flavonoids, characterized by their high abundance and multiple biological activities, are among the most important secondary metabolites in citrus [
31]. The main site for flavonoid synthesis and accumulation are oil glands, which are mainly distributed in the stems, leaves, flowers, and flavedo of fruits [
32]. Therefore, leaves and stems are important sources of flavonoids. A variety of flavonoids have been identified from the leaves and stems of different citrus species [
33,
34]. Our research shows that leaves and tender stems are flavonoid-rich tissues, with the content YL > ML TB > SB > MS, which conforms to the conclusion that a large number of flavonoids accumulate in young and fast-growing tissues [
31,
35]. Flavonoid synthesis is mainly carried out in the early stages of organ and tissue development. Therefore, the utilization of citrus tender tissue appears to be more important. Coumarin is a secondary metabolite that plays an important part in the communication between plant and microbiome [
36,
37]. Our research shows that compared with MS, coumarins accumulate more in RO, which contributes to the improvement of rhizosphere microbial community and plant resistance. It should be emphasized that in order to increase yield and resistance, most currently planted citrus cultivars are grafted. Therefore, the root tissue used in this study is from the rootstock of trifoliate orange (
Poncirus), so the metabolite profile of root tissue does not fully represent the characteristics of
Citrus reticulata. In future studies, we aim to conduct further exploration using seed-derived seedlings.
At present, the research on the secondary metabolites of citrus fruits is mainly focused on flavonoids. The citrus flavedo constitutes an important source of flavonoids, comprising specific polymethoxylated flavones (PMFs) such as nobiletin, tangeretin, and sinensetin [
38]. Our study shows that the number of flavonoids enriched in FL is much higher than that in other fruit tissues, which agrees with previous research [
39]. In addition, the accumulation of alkaloids in the FL was also higher than that in other fruit tissues. As the FL is the tissue directly exposed to the environment of citrus fruit, the accumulation of flavonoids and alkaloids in FL may improve the fruit’s resistance to biological and abiotic stresses [
40,
41]. Phenolic acid is also rich in citrus peel. Previous studies have shown that in some citrus varieties, the content of phenolic acid in flavedo is much higher than that in albedo [
42], but our research indicated that AL is the tissue with the richest phenolic acid accumulation. We believe that it may be caused by varietal differences. It has been reported that limonoid accumulates in segment membrane [
39], but there are few studies on the secondary metabolites in tangerine pith. We found that coumarins also accumulate in TP in addition to limonoids. We believe that the detailed analysis of metabolome profile in citrus fruit will provide a valuable reference for our further utilization of specific tissues.
Combined transcriptomic and metabolomic analysis explains the specific metabolite distribution among different Ponkan tissues. PAL, 4CL, and C4H are key enzymes and rate-limiting enzymes in the phenylpropane biosynthesis pathway, closely related to the synthesis of lignins. In Arabidopsis, two
PALs are highly expressed in the roots, inflorescences, and stems [
43], and
C4H is also highly expressed in tissues with a high lignin content [
44]; two citrus genes,
Cit4CL2 and
Cit4CL3, are highly expressed in young stems and roots, respectively [
45]. Our research showed these three genes were expressed highly in the root, which may lead to the accumulation of lignin in it.
The key enzymes of the flavonoid synthesis pathway have been extensively studied. Zhao et al. compared the relative expression levelof
CHSs and
CHIs in four tissues from
Citrus reticulata ‘Ougan’, finding that they were expressed highly in young leaves and flowers, related to flavonoid content [
46]. Our research shows that
CHS-2,
CHS-3, and
CHI-2 have the highest expression level in the FL, while
CHI-1 had the highest expression level in ML, YL, and TB, also closely related to flavonoid content. The two
FNSs, one
F3′H, and two
F6/8Hs selected in our study presented specific high expression in flavonoid-rich tissues, indicating their possible catalytic functions, and some of these have been reported and confirmed [
47,
48]. Currently, there is limited research on the synthesis of coumarins in citrus. In
Ruta graveolens, RgC2′H has been proved to have the function of catalyzing the backbone of coumarin [
11]. MS have only been reported on in fig trees [
49]. For citrus, only the catalytic activity of CitPTs [
50,
51] and CitX5H [
52] have been reported. In this research, the coumarin synthesis-related genes highly expressed in RO, such as
C2′Hs,
PT-2 and
MSs, can be considered as candidate genes for the coumarin synthesis pathway, which may enrich the study of coumarin synthesis in citrus. OMTs have various catalytic substrates and sites, so the diversity of their expression patterns among tissues may be related to their different catalyzing functions. For example,
OMT-2 has been confirmed to have the function of catalyzing flavonoid substrates [
19], conforming to its high gene expression level in FL, leaves, and stems. Meanwhile, it also had a high expression level in RO. Our in vitro enzyme reactions prove that
OMT-2 does participate in the methoxylation of esculetin, confirming its important role in the synthesis of coumarin or lignin pathway. The results show a good correlation between our transcriptome and metabolome data, as differences in metabolites between different tissues can be reasonably explained by the gene expression levels.
Numerous studies have been conducted on transcription factors related to flavonoid synthesis, primarily focusing on flavonols and anthocyanins [
46]. Xu et al. found that the synthesis of anthocyanins and procyanidins in Arabidopsis is regulated by the MWB complex, composed of R2R3-MYB, basic helix loop helix (bHLH) and WD repeat proteins (WDR) [
53]. Zhao et al. found that CitRAV1 forms a transcription complex with CitERF33, which enhances the transcriptional activation of
CitCHIL1 promoter and increases the content of flavonoids in citrus [
46]. However, we still lack studies on the transcriptional regulation of coumarin synthesis. In this study, we constructed a co-expression network of genes highly related to flavonoid and coumarin content and excavated transcription factors strongly related to several structural genes. At present, there have been many studies on the transcriptional regulation of lignin synthesis, involving
PAL,
C4H,
4CL and other genes of the phenylpropane synthesis pathway [
54,
55], which are also reflected in our co-expression network. For
OMT-2 and
OMT-5 that specifically expressed in roots, we screened transcription factors that may participate in methoxylation regulation according to co-expression patterns, including AR2/ERF, bZIP, bHLH, etc., which will provide new ideas for transcriptional regulation research regarding coumarin.