Next Article in Journal
Emodin and the Anthraquinone Scaffold: Therapeutic Promise and Strategies to Overcome Translational Barriers
Next Article in Special Issue
Volatile Compound Profiling and Quality Assessment of Sweet Fermented High-Amylose Rice: A Comparative GC-MS Analysis with Traditional Glutinous Rice Fermentation
Previous Article in Journal
Biological Activity of Stilbenoids Against Fungal, Parasitic, and Viral Pathogens
Previous Article in Special Issue
Chitosan Electrospun Membranes Possess Natural pH-Responsive Indicators for Smart Real-Time Spoilage Detection of Shrimp
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Molecules
Volume 31
Issue 5
10.3390/molecules31050831
molecules-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Comparative Metabolomics and Transcriptomics Analysis of Rosa roxburghii Tratt and Rosa kweichonensis var. sterilis

by
Xiaohui Wu
1,2,†,
Yi Yuan
1,2,†,
Yu Yang
2,
Min Yan
1,2,
Qiong Yang
2,
Jun Zha
1,
Guoshun Pei
2,
Yali Wei
1,2,* and
Li Tang
1,2,3,*
1
Key Laboratory of Minority Medical Resources Development and Utilization in Guizhou Province of State Ethnic Affairs Commission, Guizhou Minzu University, Guiyang 550025, China
2
School of Chinese Ethnic Medicine, Guizhou Minzu University, Guiyang 550025, China
3
State Key Laboratory of Discovery and Utilization of Functional Components in Traditional Chinese Medicine, Guizhou Medical University, Guiyang 561113, China
*
Authors to whom correspondence should be addressed.
†
These authors contributed equally to this work.
Molecules 2026, 31(5), 831; https://doi.org/10.3390/molecules31050831
Submission received: 9 January 2026 / Revised: 17 February 2026 / Accepted: 23 February 2026 / Published: 2 March 2026
(This article belongs to the Special Issue 30th Anniversary of Molecules—Recent Advances in Food Chemistry)
Browse Figures
Versions Notes

Abstract

Rosa roxburghii Tratt (RR) and Rosa kweichonensis var. sterilis (RS) are both edible medicinal plants. However, they are often confused due to their similar phenotypic characteristics, which may limit their targeted development and utilization. Here, we integrated targeted metabolomics (UPLC-MS/MS) and transcriptomics (Illumina HiSeq) to systematically dissect the metabolic and transcriptional differences between the two species. Metabolomic profiling identified 558 differentially accumulated metabolites (DAMs), defined as metabolites with significantly different abundance between RS and RR, predominantly classified as flavonoids and phenolic acids. Among these, vitamin C (L-ascorbic acid) and argininosuccinic acid were prioritized as key DAMs based on their significant fold changes, high abundance, and functional relevance to bioactivity and stress tolerance. Transcriptomic analysis further revealed that vitamin C synthesis is primarily driven by the coordinated up-regulation of USP and GME genes in the L-ascorbic acid metabolic pathway, while argininosuccinic acid production, as an intermediate in the urea cycle, is mainly mediated by the up-regulated glnA gene. These findings not only clarify the molecular basis of metabolic divergence between RR and RS but also provide potential biomarkers for their identification, laying a solid foundation for their development as distinct functional foods.

1. Introduction

Rosa roxburghii Tratt (RR) is a deciduous shrub that is taxonomically classified within the Rosaceae family. It is found predominantly in southwestern Chinese regions, such as the Guizhou, Yunnan, and Sichuan provinces [1]. This makes it a key agricultural industry in the province and RR has been utilized for both medicinal and culinary purposes [2]. Rosa roxburghii Tratt (RR) is rich in diverse bioactive components, including flavonoids [3], organic acids and vitamin C [4], and polysaccharides [5]. Modern pharmacological studies have demonstrated that RR exhibits multiple bioactivities, such as anti-aging [6], anti-apoptotic, antioxidant [7,8], anti-tumor [9], hypoglycemic [10], and anti-rtpoisoning [11]. The fruits of RR and RS are both primary medicinal and edible organs, each possessing significant nutritional advantages.
Rosa kweichonensis var. sterilis (RS) is a wild, deciduous shrub belonging to the Rosaceae family, endemic to Guizhou Province, China. While it occurs naturally in the wild, it is also widely cultivated and propagated in large quantities in Guizhou, with an annual production exceeding 10,000 tons [12]. This species is a distinct variety, rather than a cultivar of RR. Despite its economic significance, there has been a paucity of research on the phytochemical composition of RS fruits [13]. Preliminary research indicates that RS shares bioactive constituents with RR and may possess a broader spectrum of nutritional compounds [14]. However, most products on the market are RR-related functional products, with few specifically designed for RS.
Rosa roxburghii Tratt (RR) and Rosa kweichonensis var. sterilis (RS) are well-recognized as medicinal and edible dual-purpose plants, possessing substantial nutritional value and medicinal potential. These fruits have long been consumed fresh or processed into functional foods (e.g., beverages, preserves) and contain abundant bioactive components, including high concentrations of vitamin C, superoxide dismutase, and phenolic compounds [15]. Collectively, these bioactive components possess powerful antioxidant, anti-inflammatory, and antitumor properties. Rosa kweichonensis var. sterilis (RS) exhibits a range of beneficial physiological effects [16], yet the metabolic differences between RR and RS—and their regulatory mechanisms—remain limited, representing a critical gap in current research [17].
Metabolomics is a powerful analytical method that allows for the detailed examination of an organism’s metabolic activities, involving the profiling of metabolites in biofluids, cells, and tissues, and is commonly used for identifying biomarkers [18]. High-throughput transcriptome sequencing serves as a powerful platform for identifying candidate genes implicated in primary and secondary metabolite biosynthesis. Recent advancements have led to the integration of metabolomic and RNA-sequencing (RNA-seq) analytical frameworks. These frameworks have been utilized to characterize dynamic compositional shifts in complex metabolic profiles across plants and fungi. These multi-omics approaches further elucidate differentially expressed genes (DEGs) governing metabolic pathway regulation [19].
In this study, an integrated targeted metabolomics and transcriptomics strategy was employed to systematically compare metabolite profiles and related gene expression patterns between RR and RS. Findings provide a molecular basis for understanding differences in bioactive components between these two Rosa species, and support further development and utilization of RR and RS as high-value-added functional and health-promoting products.

2. Materials and Methods

2.1. Samples and Reagents

Fresh fruits of RR and RS were collected for this study. A total of six fruit tissue samples were selected and divided into two groups (RR and RS) for metabolomic analysis, with three biological replicates per group. The three biological replicates for both RR and RS were collected from healthy fruits of plants grown under identical environmental conditions and at the same developmental stage. This experimental design minimizes environmental variability and ensures that the replicates capture the core metabolic differences between the two species. A sample size of three biological replicates is widely accepted in plant metabolomics studies to balance statistical power and practical feasibility [20,21]. After collection, the fresh fruits were stored at room temperature until freeze-drying.
All reagents and standards used were of chromatographic or analytical grade. Methanol and acetonitrile (HPLC grade) were purchased from Merck (Darmstadt, Germany). Authentic chemical standards were obtained from BioBioPha (Kunming, China) and Sigma-Aldrich (St. Louis, MO, USA).

2.2. Sample Preparation and Extraction

After vacuum freeze-drying in a lyophilizer (Scientz-100F, Scientz Biotechnology, Ningbo, China), the biological samples were ground into a fine powder using a Retsch MM 400 (Retsch GmbH, Haan, Germany) mixer mill with zirconia beads at 30 Hz for 1.5 min. The powder was stored at room temperature until metabolite extraction. A 100 mg portion of the powder (solid sample) was accurately weighed and dissolved in 1.0 mL of 70% aqueous methanol for metabolite extraction. The mixture was vortexed for 30 s, and this vortexing step was repeated every 30 min for a total of six times, followed by overnight storage in a 4 °C refrigerator. After centrifugation at 12,000 rpm for 10 min, the supernatant was carefully aspirated, filtered through a 0.22 μm pore size microporous membrane (SCAA-104; ANPEL, Shanghai, China), and the filtrate was stored in an injection vial for ultra performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) analysis using a system consisting of a UPLC module (Shim-pack UFLC SHIMADZU CBM A system, Shimadzu Corporation, Kyoto, Japan) and a mass spectrometry module (QTRAP® 4500+ System, SCIEX, Framingham, MA, USA) [22].

2.3. UPLC Conditions

A targeted metabolomics approach using multiple reaction monitoring (MRM) was employed. The analysis was based on an in-house spectral library, which was constructed using reference retention time and fragmentation information. For metabolite identification, we matched the retention times and m/z transitions of detected metabolites against this library. For quantification, we performed chromatographic peak integration and correction using MultiQuant software (version 3.0.3), where the peak area (Area) represents the relative content of each metabolite. This approach enabled reliable identification and quantification of a broad range of primary and secondary metabolites, including amino acids, organic acids, flavonoids, and vitamins.
The sample extracts were subjected to ultra-performance liquid chromatography-electrospray ionization-tandem mass spectrometry (UPLC-ESI-MS/MS) analysis utilizing a system consisting of a UPLC module (Shim-pack UFLC SHIMADZU CBM A system, Shimadzu Corporation, Kyoto, Japan, https://www.shimadzu.com/) and a mass spectrometry module (QTRAP® 4500+ System, SCIEX, Framingham, MA, USA, https://sciex.com/) [23]. An ACQUITY UPLC HSS T3 C18 column (2.1 mm× 100 mm, 1.8 µm; Waters) was employed. The analysis used a 2 µL injection volume, a 0.4 mL/min flow rate, and a column temperature of 40 °C [24]. The solvent system comprised water and acetonitrile (both supplemented with 0.1% formic acid), with the following gradient elution program: 95:5 (v/v) at 0 min, linear gradient to 5:95 (v/v) at 10.0 min, maintenance at 5:95 (v/v) until 11.0 min, rapid reversion to 95:5 (v/v) at 11.1 min, and hold at 95:5 (v/v) until 15.0 min [25,26].

2.4. ESI-Q TRAP-MS/MS

Compound identification was performed by matching the retention time, precursor ion mass (Q1), and product ion mass (Q3) of sample peaks with those of authentic standards in our in-house database. A self-constructed database was used for metabolite annotation, which integrated secondary mass spectrometry (MS/MS) information from public metabolomics databases (e.g., Metlin, HMDB) and our laboratory’s accumulated plant metabolite mass spectrometry data. During the annotation process, duplicate signals (isotopes, ion adducts, and fragment ions) were removed to improve accuracy. The confidence level of metabolite identification was based on MS/MS spectrum matching between the samples and the database, with a matching threshold of ≥70% to ensure reliability. Absolute quantification was achieved using external calibration curves constructed from serial dilutions of each standard, ensuring traceability and accuracy [27].
Metabolite quantification was performed using multiple reaction monitoring (MRM) mode on a triple quadrupole mass spectrometer. Representative MRM chromatograms in both positive and negative ionization modes (Figure S1) demonstrate the overall peak quality and separation of metabolites detected across all our samples. In MRM mode, the first quadrupole screens for precursor ions (parent ions) of target substances, excluding ions of other molecular weights to preliminarily eliminate interference. After collision-induced ionization in the collision cell, precursor ions fragment into numerous product ions. The third quadrupole then filters these to select characteristic fragment ions, excluding non-target ion interference for more precise quantification and improved reproducibility. After acquiring mass spectrometry data from all samples, peak area integration was performed for all detected peaks. For the same metabolite across different samples, peak integration was calibrated to account for variations in peak appearance. Mass spectrometric detection was carried out on an AB4500 Q TRAP system (AB Sciex, Framingham, MA, USA) interfaced with an ESI source [26]. Mass spectrometric detection was carried out on an AB4500 Q TRAP system (AB Sciex) interfaced with an ESI source. Operating in both polarities under Analyst 1.6.3 control, the source parameters were: turbo spray, 550 °C, IS voltage ±5500/4500 V, GSI 50 psi, GSII 60 psi, CUR 25.0 psi, CAD high [28]. Calibration utilized polypropylene glycol solutions (QQQ: 10 μmol/L; LIT: 100 μmol/L). MRM experiments (medium nitrogen collision gas) were performed with optimized DP and CE for each transition, and the monitored transitions were scheduled according to metabolite retention times [29,30].
To ensure the accuracy of metabolite quantification, integral correction was performed on the mass spectral peaks of all metabolites. Specifically, the integral intervals of chromatographic peaks for the same metabolite across different samples were aligned based on retention time and peak shape information, which eliminated quantitative errors caused by retention time drift and peak shape differences. As shown in Figure S2, the integral correction results of randomly selected metabolites visually validate the effectiveness of this procedure. This method has been widely applied and verified in plant metabolomics studies [31].

2.5. RNA Extraction, cDNA Library Construction and Transcriptome Analysis

The project utilizes Oligo (dT) magnetic bead enrichment as a method to isolate mRNA containing polyA structures from total RNA. Subsequently, the RNA is fragmented using ion fragmentation, yielding fragments with an approximate length of 300 base pairs (bp). The selection of fragments at 300 bp was determined by the fixed adapter length. Adequate fragment length is critical for sequencing library quality. Insufficient length causes a disproportionate increase in adapter content, lowering valid data output, while excessive length physically interferes with the cluster formation process [32]. Following first-strand cDNA synthesis from RNA primed by random hexamers and reverse transcriptase, the second strand was produced by employing the first strand as a template.
After the library was constructed, it was first amplified by PCR for enrichment, followed by size selection to obtain fragments of approximately 450 bp. After construction, library quality was assessed using an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). The total and effective concentrations of the library were then determined [33]. The effective concentration of the library and the required data volume are used to determine the proportion of libraries containing different index sequences. Each sample is assigned a unique index, and the sequencing data is later distinguished by the index. The pooled library is uniformly diluted to 2 nanomoles, undergoes alkali denaturation, and forms a single-stranded library.
Subsequent to the extraction, purification, and library preparation of the RNA samples, they were subjected to sequencing using Next-Generation Sequencing (NGS) based on the Illumina HiSeq sequencing platform. The libraries were sequenced using paired-end (PE) sequencing [33].
To obtain high-quality sequence for further analysis, the raw data were first filtered to remove sequences with adapters and low-quality reads. The parameters were set as follows: (1) Cutadapt was used to remove sequences with 3′ adapters; (2) Reads with an average quality score lower than Q20 were removed. Then, the Rosa chinensis genome was selected as the reference due to its high quality, comprehensive annotation, and close phylogenetic relationship with the studied Rosa species. The filtered clean data is aligned to the reference genome named Rosa_chinensis.RchiOBHm-V2.dna.toplevel.fa (http://ftp.ensemblgenomes.org/pub/plants/release-48/gff3/rosa_chinensis/) accessed on 10 November 2022 [34].

2.6. Analysis of DEGs

To facilitate cross-gene and cross-sample comparisons, expression levels were normalized using the FPKM (Fragments Per Kilobase per Million fragments) metric. FPKM is a metric that quantifies the number of fragments from a specific gene per thousand base pairs per million fragments. For pair-end sequencing, each fragment has two reads. Gene expression was quantified via FPKM, followed by differential analysis using DESeq software (version 1.38.3). Significantly DEGs were defined by |log2FoldChange| > 1 and p < 0.05, and their distribution was visualized in a volcano plot created with R/ggplot2, depicting fold-changes and statistical significance. Finally, GO and KEGG enrichment profiling of these DEGs was conducted, with a significance threshold of p < 0.05.

2.7. Comparative Analysis

A reference genome index was constructed using the software Bowtie2 (version 2.4.5), and then the filtered reads were mapped to the reference genome using Tophat2. In the course of the TopHat mapping process, reads and reference genome sequences exhibiting mismatches of up to 2 were designated as successfully mapped. The findings indicate that the sequence coverage ratio of the sequencing data generated in the experiment exceeds 70%. Consequently, the reference genome was meticulously selected, and the related experiments were found to be free of contamination.

3. Results and Discussion

3.1. Metabolic Profiling of RR and RS

In this study, differentially accumulated metabolites (DAMs) included both primary metabolites (e.g., amino acids, organic acids) and secondary metabolites (e.g., vitamins, flavonoids) that exhibited significant differential accumulation between RS and RR.
To ensure comprehensive and accurate metabolite comparison across RR and RS samples, a rigorous data correction procedure was implemented, involving the calibration of mass spectrometry peaks based on metabolite retention time and peak shape (Figure S2). This step guaranteed the reliability of both qualitative and quantitative analyses. Quality control (QC) samples, prepared by pooling sample extracts, were used to assess technical repeatability. Overlay analysis of total ion current (TIC) profiles from different QC samples showed high overlap (Figure S3), indicating consistent retention times, peak intensities, and stable mass spectrometer signals—critical for validating data robustness in metabolomic studies. This finding indicates that the mass spectrometer demonstrates adequate signal stability when detecting the same sample at varying times, thereby substantiating the reliability of the data.
Principal Component Analysis (PCA) is an unsupervised pattern recognition method that transforms correlated multidimensional data into linearly independent principal components [35]. For six fruit tissue samples (three biological replicates per group from RR and RS), PCA results showed that Principal Component 1 (PC1, 81.25%) and PC2 (8.46%) collectively explained 89.71% of the total variance, clearly separating RR and RS into discrete clusters (Figure 1A). Furthermore, PCA analysis of the three biological replicates showed high intra-group similarity (Pearson correlation coefficient > 0.95), confirming the reliability of the sample selection and the robustness of the results. Pearson correlation cluster analysis visualized as heatmaps further confirmed distinct grouping of the two plant types (Figure 1B). These findings demonstrate inherent metabolic disparities between RR and RS, which are not only reflective of metabolic network reorganization but also lay the foundation for their functional differences. Such clear metabolic separation is consistent with the role of metabolomics in distinguishing closely related plant varieties and predicting their nutritional and medicinal potential [36].
The clear metabolic separation between RR and RS suggests that these two plant types have evolved distinct metabolic strategies, which may be linked to their environmental adaptation and functional differentiation [37]. Metabolic profiling consistency, validated by QC sample TIC overlap, ensures that subsequent analyses of metabolic differences are biologically meaningful rather than technical artifacts. Such metabolic divergence implies potential differences in their nutritional and medicinal values, as primary and secondary metabolites are key determinants of plant functional properties. This finding underscores the necessity of investigating specific metabolite variations to unlock their unique applications in functional foods.

3.2. Identification of DAMs in RR and RS

Orthogonal partial least squares-discriminant analysis (OPLS-DA) confirmed substantial discrepancies between RR and RS samples (Figure S4), with a Q2 value of 0.995 after 200 random permutations, indicating optimal model performance and high predictive capability. Using a UPLC-MS/MS platform with a self-built database, 851 metabolites were detected (Table S1). Applying criteria of VIP ≥ 1 and |log2FC| ≥ 1, 558 DAMs were identified, including 459 up-regulated and 99 down-regulated metabolites in RR relative to RS (Figure 1C). Among these 558 DAMs, 420 were identified with a confidence level of ≥70% based on MS/MS spectrum matching, while the remaining 138 were tentatively identified based on accurate molecular weight. Categorization showed phenolic acids (20.78%), flavonols (11.65%), amino acids and derivatives (11.47%), organic acids (6.27%), flavones (5.73%), and nucleotides and derivatives (5.37%) as major classes (Figure 1D).
Heatmap visualization of normalized metabolite content via hierarchical cluster analysis (HCA) showed that up-regulated DAMs with the highest FC values (e.g., Valoneoyl-glucose, FC = 92,900, this extremely high FC arises from its low abundance near the quantification limit in RR and high accumulation in RS) belonged to phenolic acids, flavones, and organic acids, with RR levels significantly exceeding the average (Figure 2A). Down-regulated DAMs (e.g., punicic acid, FC = 0.4979) were mainly phenolic acids, amino acids and derivatives, and alkaloids, with no significant content differences (Figure 2B). The most abundant DAM categories were phenolic acids, flavonoids, amino acids and derivatives, and organic acids, with more up-regulated than down-regulated DAMs (Figure 2C). KEGG enrichment analysis annotated DAMs to 87 pathways, with “Metabolic pathways” (136 DAMs, 80.47%), “Biosynthesis of secondary metabolites”, “ABC transporters”, “Biosynthesis of amino acids”, and “Phenylpropanoid biosynthesis” as top-enriched pathways (Figure 2D).
Secondary metabolites in plants form a unique pool of pharmaceutical substances and are also important for food additives, flavors, and industrial biochemicals. Plants often show increased accumulation when subjected to different stress conditions, such as exposure to elicitors or signaling molecules. In addition to their adaptive protective functions in plant physiology, secondary metabolites serve as unique sources of bioactive compounds that contribute to human health, such as dietary antioxidants and functional food ingredients [38].
The dominance of phenolic acids and flavonoids among DAMs highlights their critical role in distinguishing RR and RS. Phenolic compounds, with antioxidant and antibacterial properties, are increasingly used in the food industry to inhibit oxidative deterioration [39,40]. Based on the obtained results, it can be inferred that the difference in phenolic acid content between RR and RS may be one of the reasons for their varying antioxidant capacities. Existing research indicates that plant phenolic acids (benzoic acid derivatives and cinnamic acid derivatives) exert anticancer effects through multiple mechanisms, including inducing cancer cell apoptosis and inhibiting cell proliferation, while also mitigating the side effects of conventional chemotherapy [41,42]. Phenolic acids also exhibit anti-inflammatory properties and can help lower blood sugar levels in diabetes [43]. This indicates that RR and RS hold significant application potential in the development of functional foods. Epidemiological evidence links dietary flavonoid intake to reduced incidence of degenerative diseases [44], RR and RS also exhibited significant differences in flavonoid metabolite content. Flavonoids are multifunctional molecules that enable plants to withstand environmental stresses and ensure growth and reproduction [45]. Upon dietary intake, flavonoids exert limited direct antioxidant effects owing to concentration constraints and metabolic modifications. Their core biological significance lies in signaling regulation, acting as “natural signaling molecules” to maintain cellular homeostasis and health. As key components of health supplements, flavonoids possess diverse therapeutic potential, with core mechanisms including antioxidant activity, induction of cancer cell apoptosis, downregulation of TNF-α and IL-6, inhibition of microglial activation, and precise binding to the ATP-binding site of protein kinases to suppress critical signaling pathways via the protein-binding capacity of their B-ring dihydroxy structure [46]. This indicates that RR and RS may play roles in cell cycle regulation and cancer prevention, with RS being more suitable than RR for cancer prevention in most populations. Furthermore, multiple studies have reported the antidiabetic activity of flavonoids and phenolic acids [47,48].
Flavonoids and phenolic acids, as natural chemical constituents of plants, are key substances shaping the flavor characteristics and sensory experience of plants and their processed products. Flavonoids are an important component of plant flavor profiles. They can influence texture by interacting with plant cell wall components, softening fruit pulp and reducing harshness [49]. Although phenolic acids are not the core compounds responsible for dominant flavors, they can alter the tactile sensation of the oral mucosa through interactions with proteins and polysaccharides, thereby mitigating the sharp acidity of organic acids. Their antioxidant activity also reduces the accumulation of undesirable flavor compounds during plant processing and storage [50]. The synergistic interaction between these two factors jointly determines the sensory qualities of plants while also performing ecological functions in plant reproduction and survival. This provides a natural material foundation for flavor optimization and quality enhancement of RR and RS in food processing.
Amino acids and derivatives are among the primary nutrients and active components in both RR and RS, playing a crucial role in health. Although amino acid content differs between RR and RS, we found their amino acid compositions to be similar, consistent with previous research findings [51]. Research indicates that amino acids and their derivatives can prevent obesity by mitigating inflammatory reactions, increasing energy expenditure and alleviating oxidative stress [52]. Our study found that the amino acid and derivative content in RS was significantly higher than that in RR, suggesting that RS is more suitable for obese individuals compared to RR. Plant-based foods serve as significant sources of polyamines and amino acids for humans. Different plant-based foods exhibit distinct enrichment capacities for specific amino acids, enabling targeted supplementation of human nutritional needs. In our study, the content of multiple functional amino acids such as Gln, Arg, and Glu in RS was higher than that in RR (Table S2). Amino acids and their derivatives serve as key substrates for polyamine synthesis in the human body, forming an “amino acid-polyamine” synergistic regulatory mechanism. This mechanism can induce autophagy, delay aging, and holds significant importance for metabolic health in the elderly [53]. It can also further enhance the proliferation and repair capabilities of intestinal cells [54]. This suggests that in terms of these amino acids, RS may possess higher nutritional value and be more suitable for individuals with diarrhea, intestinal dysfunction, and cardiovascular disease risk. Both RR and RS can accumulate significant amounts of amino acids and their derivatives under natural conditions, providing a foundation for their application in the food industry (such as nutritional fortifiers and functional food ingredients) and dietary supplement sectors.

3.3. Transcriptomic Profiling of RR and RS

The Illumina HiSeq sequencing platform was used to generate 6 Gb of raw sequencing data from RR and RS samples. The sequencing platform’s built-in software then converted these files to generate raw data in FASTQ format, i.e., off-machine data. Analysis of the off-machine data showed that the Q30 scores for all samples ranged from 94.47% to 95.08% (Table S3), indicating high-quality sequencing data. A reference genome index was created with Bowtie2, and Tophat2 was then used to map the filtered reads to the reference genome. On average, 79.53% of clean reads were successfully mapped to the reference genome (Table S4), suggesting that the reference genome was appropriately selected and that the relevant experiments were not contaminated.
Six fruit tissue samples (three biological replicates each from RR and RS) were selected and divided into two groups for integrated metabolomic and transcriptomic analyses. Each group contained three biological replicates. The PCA results show that Principal Component 1 (PC1) (88.35%) and PC2 (11.65%) collectively explained the separation of samples into two distinct groups (Figure 3A). To visually represent the genetic differences between RR and RS types, hierarchical cluster analysis visualized the samples as a heatmap. As shown in Figure 3B, the six samples were clearly separated into RR and RS groups, indicating genetic distinctions between the two plant types.
High sequencing quality and mapping rates ensure reliable transcriptomic data, providing a robust basis for linking gene expression to metabolic phenotypes. The clear genetic separation between RR and RS, consistent with metabolic profiling, suggests that transcriptional regulation drives metabolic differences. This genetic divergence may reflect adaptive evolution to distinct ecological niches, leading to the development of unique metabolic networks. Integrating transcriptomic and metabolomic data is thus essential to unravel the molecular mechanisms underlying metabolite accumulation differences.

3.4. Differentially Expressed Genes in RR and RS

To this end, we employed the DESeq software to conduct a differential analysis of gene expression in RR and RS, and subsequently screened for DEGs meeting the following criteria: an expression difference multiple |log2Fold Change| > 1 and a significance p-value < 0.05. The results demonstrated that a total of 7621 DEGs were identified, of which 4278 were up-regulated and 3343 were down-regulated (Figure 3C).
As shown in Figure 3D, RR and RS are clearly separated into two distinct groups, indicating significant differences in the abundance of DEGs across categories within RR and RS.
GO enrichment analysis and KEGG enrichment analysis were performed on the DEGs in the sample to determine the main biological functions of the DEGs and classify their functions. The results showed that the DEGs between RR and RS were annotated to 119 metabolic pathways (Table S5), and all DEGs were successfully annotated to three GO functional categories: “Cellular Component”, “Biological Process”, and “Molecular Function”.
Based on GO enrichment results, the degree of enrichment is measured by the Rich factor, FDR value, and the number of DEGs enriched in this GO Term. The Rich factor refers to the ratio of the number of DEGs enriched in this GO Term to the total number of annotated DEGs. A higher Rich factor indicates greater enrichment. The FDR typically ranges from 0 to 1, with values closer to zero indicating more significant enrichment. The top 20 GO terms with the smallest FDR values—representing the most significantly enriched terms—are selected for display (Figure 3E). Combined with Figure S5, this result shows that the types of terms with the highest enrichment levels mostly belong to Biological Process, and the most pronounced enrichment is in “response to acid chemical”.
A subsequent analysis of the KEGG enrichment results indicated that the majority of the DEGs were classified into the Metabolism category. The KEGG enrichment results indicated that all DEGs were distributed across four categories: “Genetic Information Processing”, “Metabolism”, “Cell Processes”, and “Environmental Information Processing”. Within the Metabolism category, the top three pathways that exhibited significant enrichment were “Ubiquinone and other terpenoid-quinone biosynthesis”, “Circadian rhythm-plant”, and “Fructose and mannose metabolism” (Figure 3F).
As shown in Table S2, among these categories, punicic acid and gallocatechin exhibited the most significant fold changes. Punicic acid has shown strong antioxidant and anti-inflammatory effects, which contribute to its observed beneficial effects on various diseases [55]. Gallocatechin is a flavonoid compound that has a causal relationship with improved vascular function [56].
The large number of DEGs indicates extensive transcriptional reprogramming between RR and RS, which likely underpins their metabolic differences. Enrichment in the GO term “response to acid chemical” reflects plant adaptation to acidic environments, which may broadly regulate the accumulation of diverse metabolites (with organic acids as a representative example). This regulatory link is consistent with the well-documented role of organic acids in mediating plant tolerance to acidic stimuli and soil habitats, as organic acids can chelate toxic ions (e.g., Al3+, Mn2+) or maintain cellular homeostasis under low-pH conditions [57,58]. KEGG enrichment in metabolic pathways aligns with metabolomic findings, emphasizing the role of transcriptional regulation in shaping metabolic profiles. “Ubiquinone and other terpenoid-quinone biosynthesis” (linked to antioxidant and electron transport functions) and “Fructose and mannose metabolism” (central carbon metabolism) highlight key pathways driving functional differences, such as energy supply and stress resistance.

3.5. Screening of Key Differentially Accumulated Metabolites

We sorted the results according to fold change and selected the DAMs with a p-value < 0.05, including valoneoyl-glucose, vitamin C, argininosuccinic acid, quercetin-3-O-rutinoside (rutin), and UDP-N-acetylglucosamine.
Preliminary findings from metabolomics research indicate that the vitamin C content of RS is significantly higher compared to that of RR. Rosa roxburghii Tratt (RR) and Rosa kweichonensis var. sterilis (RS) are notable for their high vitamin C content, which is the most abundant antioxidant in this species. Its antioxidant capacity is significantly higher than that of common fruits and vegetables. It has been hypothesized that vitamin C plays a pivotal role in the resilience of RR and RS to environmental stresses, including ultraviolet radiation and drought [59].
Argininosuccinic acid, the immediate precursor of arginine and a key intermediate in the urea cycle [60], has received little research attention in plants. Although metabolomics research has revealed significant differences in argininosuccinic acid content between RR and RS, no in-depth analysis has been conducted.
Therefore, we selected vitamin C and arginine succinic acid as key metabolites for quantitative and comparative analysis to reveal differences in bioactive component levels and their functional implications between RS and RR.

3.6. Comprehensive Analysis Combining Metabolomics and Transcriptomics

The DEGs and DAMs from the two comparisons were co-mapped to the KEGG database to obtain their common pathways. As illustrated in Figure 4A, rank and analyze the 20 KEGG pathways jointly mapped to DEGs and DAMs based on their p-value magnitude. Among these pathways, “phenylalanine metabolism” exhibited a higher enrichment level in the comparison, with a statistically significant p-value of 0.0296 (p < 0.05). Studies have shown that phenylalanine metabolism acts as a critical bridge connecting primary and secondary metabolism, playing a vital role in maintaining cellular metabolic homeostasis and overall vitality. As a key aromatic amino acid, phenylalanine serves as the primary precursor for the biosynthesis of numerous bioactive metabolites, including phenolic acids, flavonoids, and lignin [61].
Vitamin C and argininosuccinic acid were selected as key DAMs based on their significant FC, high abundance, and functional relevance to bioactivity and stress adaptation. These metabolites were mapped to the KEGG database to identify their associated metabolic pathways. The results showed that they were co-mapped to the same metabolic pathway. Then selected the pathways “ascorbate and aldarate metabolism” and “arginine biosynthesis” from the metabolic pathways related to vitamin C and argininosuccinic acid.
A total of 8 up-regulated DEGs and 6 down-regulated DEGs were identified in ascorbate and aldarate metabolism. We selected pathway modules directly related to vitamin C for analysis (Figure 4B). As illustrated in the accompanying Figure 3 DEGs were found to be up-regulated, including GME (RchiOBHm_Chr4g0388911, RchiOBHm_Chr3g0451161), USP (RchiOBHm_Chr4g0418881), and 2 DEGs were found to be down-regulated, including UGDH (RchiOBHm_Chr7g0182711) and GalDH (RchiOBHm_Chr6g0246091). The differential expression of these genes is found upstream of vitamin C. Additionally, pearson correlation analysis was conducted between vitamin C levels and the expression of up-regulated genes to evaluate their association. As shown in Figure 4B, all up-regulated genes exhibited significant positive correlations with vitamin C, while no significant correlations were observed among the three genes themselves. This result suggests that the up-regulation of these genes may directly contribute to the accumulation of vitamin C in the fruit. This indicates that the RchiOBHm_Chr3g0451161, RchiOBHm_Chr4g0388911, and RchiOBHm_Chr4g0418881 genes within the up-regulated glnA gene family can directly regulate vitamin C content, leading to its elevation. Ascorbate has been observed to engage in multifaceted signaling pathways initiated by both ROS and reactive nitrogen species. The apoplastic ascorbate/dehydroascorbate (DHA) ratio, which is subject to regulation by ascorbate oxidase (AO), has been demonstrated to sculpt the apoplastic ROS signal. This signal subsequently regulates numerous physiological processes, such as polarized cell growth, defenses against biotic and abiotic factors, and communication between cells. Additionally, it has been observed to regulate the light-dependent regulation of photosynthesis [62].
Vitamin C is a principal antioxidant constituent of RR and RS, playing a crucial role in maintaining the balance of cellular oxidation and reduction. A comparison of vitamin C concentrations between RR and RS reveals that both species are rich in this bioactive compound with significant nutritional value. However, subsequent metabolomic analysis revealed significantly higher vitamin C content in RS than in RR. As a key antioxidant, vitamin C plays critical roles in scavenging ROS, enhancing stress tolerance, and supporting nutritional quality [62]. Transcriptomic analysis further showed that, within the L-ascorbic acid metabolic pathway, expression levels of USP and GME were higher in RS than in RR, whereas those of UGDH and GalDH were lower.
The GME gene is responsible for encoding a critical enzyme that plays a pivotal role in the biosynthesis pathway of vitamin C [63]. The USP enzyme can regulate the stability, subcellular localization, and activity of substrate proteins by removing ubiquitin modifications [64]. The overexpression of the GME gene significantly increased the content of vitamin C in plants, and USP optimized stress signal transduction by regulating protein homeostasis. Therefore, we speculate that GME and USP directly influence the efficiency of vitamin C synthesis. It is well known that vitamin C has powerful antioxidant properties. Moreover, vitamin C exhibits significant consequences for human health in view of its function in epigenetic programming.
A total of 2 up-regulated DEGs and 10 down-regulated DEGs were identified in arginine biosynthesis. A selection of pathway modules directly related to argininosuccinic acid was subjected to analysis (Figure 4C). As illustrated in Figure 4C, 2 DEGs have been found to be up-regulated, including glbA (RchiOBHm_Chr3g0491191, RchiOBHm_Chr4g0446601), and 7 down-regulated genes, including NAGS (RchiOBHm_Chr1g0384171, RchiOBHm_Chr2g0108261), argAB (RchiOBHm_Chr6g0312541), argG (RchiOBHm_Chr7g0190481, RchiOBHm_Chr7g0190471), argH (RchiOBHm_Chr6g0244231), and glnA (RchiOBHm_Chr7g0178021). We performed Pearson correlation analysis between argininosuccinic acid levels and the expression of up-regulated genes. As shown in Figure 4C, the RchiOBHm_Chr4g0446601 gene and RchiOBHm_Chr3g0491191 gene within the glnA family exhibited a significant positive correlation with argininosuccinic acid. with RchiOBHm_Chr4g0446601 exhibiting a stronger correlation than RchiOBHm_Chr3g0491191. Thus, we can conclude that the up-regulated glnA family genes RchiOBHm_Chr4g0446601 and RchiOBHm_Chr3g0491191 regulate argininosuccinic acid levels, leading to their elevation, thereby exhibiting antioxidant and stress-resistant properties.
Argininosuccinic acid is an intermediate in the urea cycle and a key molecule in stress regulation in plants. It integrates nitrogen metabolism and energy supply through the urea cycle, playing a central role in plant by regulating the synthesis of arginine, polyamines, and ornithine [65,66]. The results of the metabolomics study indicate that the content of argininosuccinic acid in RS is higher than that in RR. In conjunction with the findings of transcriptomics, glnA is identified as an up-regulated gene in arginine biosynthesis, while NAGS, argAB, argG and argH, among others, are found to be down-regulated. The glnA gene primarily encodes various substances with signal transduction functions in plants, thereby performing nitrogen fixation and influencing nitrogen metabolism efficiency and plant growth [67].
Studies have shown that argininosuccinic acid can effectively alleviate nitrogen poisoning and hyperazotemia [68]. Furthermore, increasing the content of arginine succinate in sheep milk can enhance its nutritional and therapeutic effects [69]. The administration of argininosuccinic acid has been shown to enhance the accumulation of arginine within the body. This, in turn, has been observed to improve microcirculation and promote protein synthesis. It has shown therapeutic potential in preventing and easing various health issues, such as cardiovascular disease, neurodegenerative disorders, metabolic conditions, immune function, and anti-aging effects [70,71]
Rosa roxburghii Tratt (RR) and Rosa kweichonensis var. sterilis (RS) are important medicinal and edible plants, which are widely used as edible ingredients for nutritional supplementation and as medicinal materials for their antioxidant, anti-inflammatory, and immunomodulatory effects. The differentiation of their metabolic profiles arises not only from environmental adaptation but also reflects the restructuring of metabolic networks during species evolution. Our research indicates that RR and RS exhibit significant differences in the content of various substances, suggesting that RR and RS may exert varying degrees of antioxidant and anti-inflammatory effects in functional food development. However, due to their similarities in phenotypic characteristics and applications, they are often regarded as functionally equivalent. Consequently, specialized deep-processed products targeting RS fruits are currently less common in the market, which may indicate that the nutritional and medicinal value of these two unique functional foods has not been fully exploited. This has led to a perceived scarcity of specialized deep-processed products targeting RS in the market, which, in turn, has resulted in the potential underutilization of their distinct nutritional and medicinal values. Importantly, as shown in Figure S6, the key metabolites identified in this study (e.g., vitamin C and argininosuccinic acid) could serve as potential biomarkers to distinguish RS from RR, providing a molecular basis for the targeted development and utilization of these two unique functional foods. Therefore, it is necessary to conduct in-depth research on the differences between RR and RS in terms of metabolites and gene regulation.

4. Conclusions

A combined multivariate analysis strategy, including metabolomics and transcriptomics, was applied to RR and RS. This approach identified a total of 558 DAMs and 7621 DEGs, revealing significant differences in metabolic profiles and gene expression patterns between the two species.
Metabolomics analysis revealed that the presence of flavonols, phenolic acids, organic acids, amino acids, and their derivatives serves as biomarkers distinguishing RR from RS. Both RR and RS are high-quality food resources with significant development and utilization potential, capable of exerting exceptional nutritional and medicinal value through the synergistic effects of multiple bioactive components. The subsequent research results filled the gap in the study of the differences in the molecular mechanisms of RR and RS metabolism and the regulation of metabolic diversity, providing a new perspective for exploring the intrinsic mechanisms of functional foods.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/molecules31050831/s1, Figure S1: Representative MRM chromatograms in positive and negative ionization modes; Figure S2: Integration Calibration Chart for Metabolite Quantitative Analysis; Figure S3: TIC overlap diagram for QC sample mass spectrometry detection; Figure S4: RR vs RS OPLS-DA score map; Figure S5: GO enrichment analysis bar chart; Figure S6: Comparison of key metabolite concentrations (peak area) between RR and RS; Table S1: Detailed identification information and quantitative results for all metabolites; Table S2: Summary of all DAMs with up-regulated and down-regulated expression; Table S3: Summary of RR and RS transcriptome sequencing data; Table S4: RNASeq Map Statistics. Table S5: Transcriptome KEGG pathway summary.

Author Contributions

X.W.: Investigation, Validation, Data curation, Formal analysis, Software, Writing—original draft, Writing—review and editing. Y.Y. (Yi Yuan): Investigation, Validation, Data curation, Writing—original draft, Writing—review and editing. Y.Y. (Yu Yang): Writing—original draft, Formal analysis, Investigation. M.Y.: Data curation, Visualization, Validation. Q.Y.: Conceptualization, Project administration. J.Z.: Conceptualization, Methodology. G.P.: Writing—review and editing, Data curation. Y.W.: Writing—original draft, Methodology, Investigation. L.T.: Conceptualization, Funding acquisition, Writing—review and editing, Formal analysis, Resources. All authors have read and agreed to the published version of the manuscript.

Funding

This work was financially supported by the National Natural Science Foundation of China (82405014); Guizhou Science and Technology Department (GMULH (2025) 019; Qiankehe Foundation-[2024] Youth 205; Qiankehe Foundation MS [2025] 222); Qiankehe jichu ZK [2022] Yiban201; Xinzhiqianyan Leading Project of Guizhou Provincial Association for Science and Technology (2025XZQYXM-01-09).

Institutional Review Board Statement

This study does not involve human participants or vertebrate animals. Therefore, ethical approval and informed consent are not required.

Informed Consent Statement

Not applicable.

Data Availability Statement

The original contributions presented in the study are included in the article/Supplementary Materials, further inquiries can be directed to the corresponding author.

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript:
RRRosa roxburghii Tratt
RSRosa kweichonensis var. sterilis
DAMsDifferentially accumulated metabolites
DEGsDifferentially expressed genes
RNA-seqRNA-sequencing
FCFold Change
MRMMultiple reaction monitoring
GMEGDP-D-Mannose Epimerase
USPUniversal stress protein
TNF-αTumor necrosis factor-alpha
IL-6Interleukin 6
QCQuality control
HCAHierarchical cluster analysis
TICTotal ion current
PCAPrincipal Component Analysis
NGSNext-Generation Sequencing
OPLS-DAOrthogonal partial least squares-discriminant analysis
FPKMFragments Per Kilobase per Million fragments
GluGlutamine
ArgArginine
ROSReactive oxygen species
UGDHUDP-glucose 6-dehydrogenase
GalDHGalactose dehydrogenase
glnAglutamine synthetase
NAGSN-acetylglutamate synthase
argABamino-acid N-acetyltransferase
argGargininosuccinate synthase
argHargininosuccinate lyase

References

  1. Archana, J.; Surendra, S.; Qihai, G.; Qin, W.; Jingshan, S. Chemical diversity, traditional uses, and bioactivities of Rosa roxburghii Tratt: A comprehensive review. Pharmacol. Ther. 2024, 259, 108657. [Google Scholar] [CrossRef] [PubMed]
  2. Yin, C.; Zhang, Y.; Zhang, L.; Tian, Y.; Zhong, X.; Fang, X.; Yang, Y.; Tao, A. Exploring Rosa roxburghii Tratt polysaccharides: From extraction to application potential in functional products—An in-depth review. Int. J. Biol. Macromol. 2024, 280, 135543. [Google Scholar] [CrossRef] [PubMed]
  3. Ming-Hua, H.; Fan, Z.; Xing-Xia, L.; Fan, Z.; Li-Juan, W.; Sai-Juan, X.; Jin-Hua, Z.; Hong-Long, J.; Ping, X. Qualitative and quantitative analysis of catechin and quercetin in flavonoids extracted from Rosa roxburghii Tratt. Trop. J. Pharm. Res. TJPR 2018, 17, 71–76. [Google Scholar]
  4. Wang, L.; Wei, T.; Zheng, L.; Jiang, F.; Ma, W.; Lu, M.; Wu, X.; An, H. Recent Advances on Main Active Ingredients, Pharmacological Activities of Rosa roxbughii and Its Development and Utilization. Foods 2023, 12, 1051. [Google Scholar] [CrossRef]
  5. Tang, Z.; Chen, Y.; Cai, C.; Ma, S.; Tan, Z. CO2-responsive ionic liquid for the sustainable extraction and separation of Rosa roxburghii polysaccharides. Int. J. Biol. Macromol. 2025, 308, 142708. [Google Scholar] [CrossRef]
  6. Liu, C.; Chan, L.-P.; Liang, C.-H. The Anti-Aging Activities against Oxidative Damages of Rosa roxburghii and Multi-Fruit Concentrate Drink. J. Food Nutr. Res. 2019, 7, 845–850. [Google Scholar] [CrossRef]
  7. Zikela, L.; Yu, Z.; Han, J.; Zhu, H.; Wang, D.; Wang, X.; Li, S.; Han, Q. Protective effects of fermented Rosa roxburghii Tratt juice against ethanol-induced hepatocyte injury by regulating the NRF2-AMPK signaling pathway in AML-12 cells. Mol. Med. Rep. 2024, 30, 174. [Google Scholar] [CrossRef]
  8. Ping, X.; Xingxia, L.; Xiwen, X.; Wenbo, Z.; Xinhua, C.; Peiyong, Q.; Minghua, H.; Lijuan, W.; Dandan, L.; Xiuhua, Z.; et al. Flavonoids of Rosa roxburghii Tratt Exhibit Anti-Apoptosis Properties by Regulating PARP-1/AIF. J. Cell. Biochem. 2017, 118, 3943–3952. [Google Scholar]
  9. Liu, W.; Li, N.; Hou, J.; Cao, R.; Jia, L.; Guo, Y.; Xu, J. Structure and antitumor activity of a polysaccharide from Rosa roxburghii. Int. J. Biol. Macromol. 2024, 273, 132807. [Google Scholar] [CrossRef]
  10. Shen, C.; Wang, Y.; Zhang, H.; Li, W.; Chen, W.; Kuang, M.; Song, Y.; Zhong, Z. Exploring the active components and potential mechanisms of Rosa roxburghii Tratt in treating type 2 diabetes mellitus based on UPLC-Q-exactive Orbitrap/MS and network pharmacology. Chin. Med. 2023, 18, 12. [Google Scholar] [CrossRef]
  11. Yang, L.; Zhu, K.; Wang, W.; Sun, B.; Luo, P.; Zhang, A. Rosa roxburghii Tratt juice ameliorates metabolic disorder in arsenicosis patients based on the analysis of untargeted plasma metabolomics. Front. Pharmacol. 2025, 16, 1488113. [Google Scholar] [CrossRef] [PubMed]
  12. Luan, L. Study on Gradient Preparation, Functional Evaluation and Application of Rosa sterlis S.D. Shi Bioactive Ingredients. Master’s Thesis, Zhejiang University, Hangzhou, China, 2024. [Google Scholar]
  13. Wang, L.; Li, Y.; Xia, R.; Zheng, X.; Li, X.; Wu, S.; Zhang, Q.; Li, S.; Deng, Y.; Yao, Y.; et al. Component analysis and anti-pulmonary fibrosis effects of Rosa sterilis juice. Food Funct. 2022, 13, 12915–12924. [Google Scholar] [CrossRef] [PubMed]
  14. Liu, M.H.; Zhang, Q.; Zhang, Y.H.; Lu, X.Y.; Fu, W.M.; He, J.Y. Chemical Analysis of Dietary Constituents in Rosa roxburghii and Rosa sterilis Fruits. Molecules 2016, 21, 1204. [Google Scholar] [CrossRef] [PubMed]
  15. Deng, Y.; Kan, H.; Li, Y.; Liu, Y.; Qiu, X. Analysis of Volatile Components in Rosa roxburghii Tratt. and Rosa sterilis Using Headspace–Solid-Phase Microextraction–Gas Chromatography–Mass Spectrometry. Molecules 2023, 28, 7879. [Google Scholar] [CrossRef]
  16. Luan, L.; Pan, H.; Chen, Y.; Ye, X.; Hou, Z.; Chen, S. Role of extracted phytochemicals from Rosa sterilis S. D. Shi in DSS-induced colitis mice: Potential amelioration of UC. Food Funct. 2023, 14, 8369–8382. [Google Scholar] [CrossRef]
  17. Chen, K.; Zhang, Q.; Yang, S.; Zhang, S.; Chen, G. Comparative Study on the Impact of Different Extraction Technologies on Structural Characteristics, Physicochemical Properties, and Biological Activities of Polysaccharides from Seedless Chestnut Rose (Rosa sterilis) Fruit. Foods 2024, 13, 772. [Google Scholar] [CrossRef]
  18. Johnson, C.H.; Ivanisevic, J.; Siuzdak, G. Metabolomics: Beyond biomarkers and towards mechanisms. Nat. Rev. Mol. Cell Biol. 2016, 17, 451–459. [Google Scholar] [CrossRef]
  19. Zhang, J.; Wang, N.; Chen, W.; Zhang, W.; Zhang, H.; Yu, H.; Yi, Y. Integrated metabolomics and transcriptomics reveal metabolites difference between wild and cultivated Ophiocordyceps sinensis. Food Res. Int. 2023, 163, 112275. [Google Scholar] [CrossRef]
  20. Shi, T.; Zhu, A.; Jia, J.; Hu, X.; Chen, J.; Liu, W.; Ren, X.; Sun, D.; Fernie, A.R.; Cui, F.; et al. Metabolomics analysis and metabolite-agronomic trait associations using kernels of wheat (Triticum aestivum) recombinant inbred lines. Plant J. 2020, 103, 279–292. [Google Scholar] [CrossRef]
  21. Salem, M.A.; Perez de Souza, L.; Serag, A.; Fernie, A.R.; Farag, M.A.; Ezzat, S.M.; Alseekh, S. Metabolomics in the Context of Plant Natural Products Research: From Sample Preparation to Metabolite Analysis. Metabolites 2020, 10, 37. [Google Scholar] [CrossRef]
  22. Chen, W.; Gong, L.; Guo, Z.; Wang, W.; Zhang, H.; Liu, X.; Yu, S.; Xiong, L.; Luo, J. A Novel Integrated Method for Large-Scale Detection, Identification, and Quantification of Widely Targeted Metabolites: Application in the Study of Rice Metabolomics. Mol. Plant 2013, 6, 1769–1780. [Google Scholar] [CrossRef] [PubMed]
  23. Chen, L.; Ma, Y.; He, T.; Chen, T.; Pan, Y.; Zhou, D.; Li, X.; Lu, Y.; Wu, Q.; Wang, L. Integrated transcriptome and metabolome analysis unveil the response mechanism in wild rice (Zizania latifolia griseb.) against sheath rot infection. Front. Genet. 2023, 14, 1163464. [Google Scholar] [CrossRef] [PubMed]
  24. Zhang, Q.; Zhang, A.; Wu, F.; Wang, X. UPLC-G2Si-HDMS Untargeted Metabolomics for Identification of Yunnan Baiyao’s Metabolic Target in Promoting Blood Circulation and Removing Blood Stasis. Molecules 2022, 27, 3208. [Google Scholar] [CrossRef] [PubMed]
  25. Matić, J.; Alva, C.; Witschnigg, A.; Eder, S.; Reusch, K.; Paudel, A.; Khinast, J. Towards predicting the product quality in hot-melt extrusion: Small scale extrusion. Int. J. Pharm. X 2020, 2, 100062. [Google Scholar] [CrossRef]
  26. Fraga, C.G.; Clowers, B.H.; Moore, R.J.; Zink, E.M. Signature-Discovery Approach for Sample Matching of a Nerve-Agent Precursor Using Liquid Chromatography−Mass Spectrometry, XCMS, and Chemometrics. Anal. Chem. 2010, 82, 4165–4173. [Google Scholar] [CrossRef]
  27. Caixach, J.; Flores, C.; Spoof, L.; Meriluoto, J.; Schmidt, W.; Mazur-Marzec, H.; Hiskia, A.; Kaloudis, T.; Furey, A. Liquid Chromatography–Mass Spectrometry. In Handbook of Cyanobacterial Monitoring and Cyanotoxin Analysis; John Wiley & Sons: Hoboken, NJ, USA, 2016; pp. 218–257. [Google Scholar]
  28. Chen, D.; Xiao, Y.; Zheng, X.; Sun, H.; Zhang, C.; Zhu, J.; Xue, T. Seasonal dynamics and molecular regulation of flavonoid biosynthesis in Cyclocarya paliurus (Batal.) Iljinsk. Front. Plant Sci. 2025, 16, 1525226. [Google Scholar] [CrossRef]
  29. Cao, K.; Chen, J.; Huang, R.; Lu, R.; Zhou, X.; Bu, Y.; Li, L.; Yao, C. Metabolomics analysis reveals the differences between Abrus cantoniensis Hance and Abrus mollis Hance. BMC Plant Biol. 2023, 23, 375. [Google Scholar] [CrossRef]
  30. Thévenot, E.A.; Roux, A.; Xu, Y.; Ezan, E.; Junot, C. Analysis of the Human Adult Urinary Metabolome Variations with Age, Body Mass Index, and Gender by Implementing a Comprehensive Workflow for Univariate and OPLS Statistical Analyses. J. Proteome Res. 2015, 14, 3322–3335. [Google Scholar] [CrossRef]
  31. Meyer, S.; Riedo, A.; Neuland, M.B.; Tulej, M.; Wurz, P. Fully automatic and precise data analysis developed for time-of-flight mass spectrometry. J. Mass Spectrom. 2017, 52, 580–590. [Google Scholar] [CrossRef]
  32. Yu, G.; Wang, Z.; Wei, M.; Jia, L.; Qu, Y.; Jiang, Y. Exogenous Sugar Alleviates Salt Stress in Cucumber Seedlings by Regulating the Antioxidant System and Hormone Signaling. Curr. Issues Mol. Biol. 2025, 47, 754. [Google Scholar] [CrossRef]
  33. Li, L.; Li, Y.; Ding, G. Response mechanism of carbon metabolism of Pinus massoniana to gradient high temperature and drought stress. BMC Genom. 2024, 25, 166. [Google Scholar] [CrossRef] [PubMed]
  34. Ma, W.-T.; Lu, M.; An, H.-M.; Yi, Y. Comparative transcriptomic analysis of Rosa sterilis inflorescence branches with different trichome types reveals an R3-MYB transcription factor that negatively regulates trichome formation. J. Integr. Agric. 2022, 21, 2926–2942. [Google Scholar] [CrossRef]
  35. Jolliffe, I.T.; Cadima, J. Principal component analysis: A review and recent developments. Philos. Trans. Ser. A Math. Phys. Eng. Sci. 2016, 374, 20150202. [Google Scholar] [CrossRef] [PubMed]
  36. Salam, U.; Ullah, S.; Tang, Z.-H.; Elateeq, A.A.; Khan, Y.; Khan, J.; Khan, A.; Ali, S. Plant Metabolomics: An Overview of the Role of Primary and Secondary Metabolites against Different Environmental Stress Factors. Life 2023, 13, 706. [Google Scholar] [CrossRef]
  37. Ravenscroft, C.H.; Fridley, J.D.; Grime, J.P. Intraspecific functional differentiation suggests local adaptation to long-term climate change in a calcareous grassland. J. Ecol. 2014, 102, 65–73. [Google Scholar] [CrossRef]
  38. Akula, R.; Ravishankar, G.A. Influence of abiotic stress signals on secondary metabolites in plants. Plant Signal. Behav. 2011, 6, 1720–1731. [Google Scholar] [CrossRef]
  39. Kiokias, S.; Proestos, C.; Oreopoulou, V. Phenolic Acids of Plant Origin—A Review on Their Antioxidant Activity In Vitro (O/W Emulsion Systems) Along with Their in Vivo Health Biochemical Properties. Foods 2020, 9, 534. [Google Scholar] [CrossRef]
  40. Vaquero, M.J.R.; Alberto, M.R.; de Nadra, M.C.M. Antibacterial effect of phenolic compounds from different wines. Food Control 2007, 18, 93–101. [Google Scholar] [CrossRef]
  41. Abotaleb, M.; Liskova, A.; Kubatka, P.; Büsselberg, D. Therapeutic Potential of Plant Phenolic Acids in the Treatment of Cancer. Biomolecules 2020, 10, 221. [Google Scholar] [CrossRef]
  42. Al Jitan, S.; Alkhoori, S.A.; Yousef, L.F. Chapter 13—Phenolic Acids from Plants: Extraction and Application to Human Health. In Studies in Natural Products Chemistry; Atta-ur-Rahman, B., Ed.; Elsevier: Amsterdam, The Netherlands, 2018; Volume 58, pp. 389–417. [Google Scholar]
  43. Tatipamula, V.B.; Kukavica, B. Phenolic compounds as antidiabetic, anti-inflammatory, and anticancer agents and improvement of their bioavailability by liposomes. Cell Biochem. Funct. 2021, 39, 926–944. [Google Scholar] [CrossRef]
  44. Heiss, C.; Keen, C.L.; Kelm, M. Flavanols and cardiovascular disease prevention. Eur. Heart J. 2010, 31, 2583–2592. [Google Scholar] [CrossRef]
  45. Brunetti, C.; Di Ferdinando, M.; Fini, A.; Pollastri, S.; Tattini, M. Flavonoids as Antioxidants and Developmental Regulators: Relative Significance in Plants and Humans. Int. J. Mol. Sci. 2013, 14, 3540–3555. [Google Scholar] [CrossRef]
  46. Okoduwa, S.I.R.; Abdulwaliyu, I.; Igiri, B.E.; Arekemase, S.O.; Okoduwa, U.J.; Itiat, J.F.; Egbule, M.N.; Mustapha, R.A. Multi-therapeutic potential of flavonoids as an essential component in nutraceuticals for the treatment and management of human diseases. Phytomedicine Plus 2024, 4, 100558. [Google Scholar] [CrossRef]
  47. Ali Asgar, M. Anti-Diabetic Potential of Phenolic Compounds: A Review. Int. J. Food Prop. 2013, 16, 91–103. [Google Scholar] [CrossRef]
  48. Sarian, M.N.; Ahmed, Q.U.; Mat So’ad, S.Z.; Alhassan, A.M.; Murugesu, S.; Perumal, V.; Syed Mohamad, S.N.A.; Khatib, A.; Latip, J. Antioxidant and Antidiabetic Effects of Flavonoids: A Structure-Activity Relationship Based Study. BioMed Res. Int. 2017, 2017, 8386065. [Google Scholar] [CrossRef] [PubMed]
  49. Yue, M.; Jiang, L.; Zhang, N.; Zhang, L.; Liu, Y.; Wang, Y.; Li, M.; Lin, Y.; Zhang, Y.; Zhang, Y.; et al. Importance of FaWRKY71 in Strawberry (Fragaria × ananassa) Fruit Ripening. Int. J. Mol. Sci. 2022, 23, 12483. [Google Scholar] [CrossRef]
  50. Ghasemzadeh, A.; Ghasemzadeh, N. Flavonoids and phenolic acids: Role and biochemical activity in plants and human. J. Med. Plants Res. 2011, 5, 6697–6703. [Google Scholar] [CrossRef]
  51. He, J.-Y.; Zhang, Y.-H.; Ma, N.; Zhang, X.-L.; Liu, M.-H.; Fu, W.-M. Comparative analysis of multiple ingredients in Rosa roxburghii and R. sterilis fruits and their antioxidant activities. J. Funct. Foods 2016, 27, 29–41. [Google Scholar] [CrossRef]
  52. Zheng, J.; Xiao, H.; Duan, Y.; Song, B.; Zheng, C.; Guo, Q.; Li, F.; Li, T. Roles of amino acid derivatives in the regulation of obesity. Food Funct. 2021, 12, 6214–6225. [Google Scholar] [CrossRef]
  53. Hou, Y.; He, W.; Hu, S.; Wu, G. Composition of polyamines and amino acids in plant-source foods for human consumption. Amino Acids 2019, 51, 1153–1165. [Google Scholar] [CrossRef]
  54. Kong, S.; Zhang, Y.H.; Zhang, W. Regulation of Intestinal Epithelial Cells Properties and Functions by Amino Acids. BioMed Res. Int. 2018, 2018, 2819154. [Google Scholar] [CrossRef] [PubMed]
  55. Guerra-Vázquez, C.M.; Martínez-Ávila, M.; Guajardo-Flores, D.; Antunes-Ricardo, M. Punicic Acid and Its Role in the Prevention of Neurological Disorders: A Review. Foods 2022, 11, 252. [Google Scholar] [CrossRef] [PubMed]
  56. Bondonno, C.P.; Croft, K.D.; Ward, N.; Considine, M.J.; Hodgson, J.M. Dietary flavonoids and nitrate: Effects on nitric oxide and vascular function. Nutr. Rev. 2015, 73, 216–235. [Google Scholar] [CrossRef] [PubMed]
  57. José, G.V.; Sonia, O. Chapter 10—Organic Acids. In Postharvest Physiology and Biochemistry of Fruits and Vegetables; Woodhead Publishing: Cambridge, UK, 2019; pp. 207–224. [Google Scholar]
  58. Jing, C.H.E.; Xue Qiang, Z.; Shen, R.F. Molecular mechanisms of plant adaptation to acid soils: A review. Pedosphere 2023, 33, 14–22. [Google Scholar]
  59. Mellidou, I.; Koukounaras, A.; Kostas, S.; Patelou, E.; Kanellis, A.K. Regulation of Vitamin C Accumulation for Improved Tomato Fruit Quality and Alleviation of Abiotic Stress. Genes 2021, 12, 694. [Google Scholar] [CrossRef]
  60. Ratner, S.; Petrack, B.; Rochovansky, O. Biosynthesis of Urea: V. Isolation and Properties of Argininosuccinic Acid. J. Biol. Chem. 1953, 204, 95–113. [Google Scholar] [CrossRef]
  61. Adams, Z.P.; Ehlting, J.; Edwards, R. The regulatory role of shikimate in plant phenylalanine metabolism. J. Theor. Biol. 2019, 462, 158–170. [Google Scholar] [CrossRef]
  62. Foyer, C.H.; Kyndt, T.; Hancock, R.D. Vitamin C in Plants: Novel Concepts, New Perspectives, and Outstanding Issues. Antioxid. Redox Signal. 2019, 32, 463–485. [Google Scholar] [CrossRef]
  63. Tao, J.; Wu, H.; Li, Z.; Huang, C.; Xu, X. Molecular Evolution of GDP-D-Mannose Epimerase (GME), a Key Gene in Plant Ascorbic Acid Biosynthesis. Front. Plant Sci. 2018, 9, 1293. [Google Scholar] [CrossRef]
  64. Zhang, S.; Hu, N.; Yu, F. Insights into a functional model of key deubiquitinases UBP12/13 in plants. New Phytol. 2024, 242, 424–430. [Google Scholar] [CrossRef]
  65. Zhao, J.; Wang, X.; Pan, X.; Jiang, Q.; Xi, Z. Exogenous Putrescine Alleviates Drought Stress by Altering Reactive Oxygen Species Scavenging and Biosynthesis of Polyamines in the Seedlings of Cabernet Sauvignon. Front. Plant Sci. 2021, 12, 767992. [Google Scholar] [CrossRef]
  66. Tischner, R.; Galli, M.; Heimer, Y.M.; Bielefeld, S.; Okamoto, M.; Mack, A.; Crawford, N.M. Interference with the citrulline-based nitric oxide synthase assay by argininosuccinate lyase activity in Arabidopsis extracts. FEBS J. 2007, 274, 4238–4245. [Google Scholar] [CrossRef]
  67. Liao, H.-S.; Lee, K.-T.; Chung, Y.-H.; Chen, S.-Z.; Hung, Y.-J.; Hsieh, M.-H. Glutamine induces lateral root initiation, stress responses, and disease resistance in Arabidopsis. Plant Physiol. 2024, 195, 2289–2308. [Google Scholar] [CrossRef]
  68. Wang, S.; Li, X.; Zhang, M.; Li, M. Cetobacterium somerae-derived argininosuccinic acid promotes intestinal and liver ureagenesis to alleviate ammonia intoxication. Microbiome 2025, 13, 163. [Google Scholar] [CrossRef]
  69. Kahraman, M.; Yurtseven, S.; Sakar, E.; Daş, A.; Yalçın, H.; Güngören, G.; Boyraz, M.; Koyuncu, İ. Pistachio, Pomegranate and Olive Byproducts Added to Sheep Rations Change the Biofunctional Properties of Milk through the Milk Amino Acid Profile. Food Sci. Anim. Resour. 2023, 43, 124–138. [Google Scholar] [CrossRef]
  70. Luiking, Y.C.; Poeze, M.; Ramsay, G.; Deutz, N.E.P. The Role of Arginine in Infection and Sepsis. J. Parenter. Enter. Nutr. 2005, 29, S70–S74. [Google Scholar] [CrossRef]
  71. Oyovwi, M.O.; Atere, A.D. Exploring the medicinal significance of l-Arginine mediated nitric oxide in preventing health disorders. Eur. J. Med. Chem. Rep. 2024, 12, 100175. [Google Scholar] [CrossRef]
Molecules 31 00831 g001
Figure 1. Metabolomics profiles of RR and RS: including (A) Metabolomics PCA score plots of RR and RS (B) Correlation Analysis Heatmap of RR and RS (C) Volcano plot of DAMs between RR and RS (D) Pie chart showing classification of differentially accumulated metabolites.
Figure 1. Metabolomics profiles of RR and RS: including (A) Metabolomics PCA score plots of RR and RS (B) Correlation Analysis Heatmap of RR and RS (C) Volcano plot of DAMs between RR and RS (D) Pie chart showing classification of differentially accumulated metabolites.
Molecules 31 00831 g001
Molecules 31 00831 g002
Figure 2. Metabolomics profiles of RR and RS: including (A) Heatmap of upregulated DAMs in RR and RS (B) Heatmap of downregulated DAMs in RR and RS (C) The category with the most distinct compounds between RR and RS, along with their quantities (D) KEGG pathway enrichment analysis of differentially accumulated metabolites in RR and RS.
Figure 2. Metabolomics profiles of RR and RS: including (A) Heatmap of upregulated DAMs in RR and RS (B) Heatmap of downregulated DAMs in RR and RS (C) The category with the most distinct compounds between RR and RS, along with their quantities (D) KEGG pathway enrichment analysis of differentially accumulated metabolites in RR and RS.
Molecules 31 00831 g002
Molecules 31 00831 g003
Figure 3. Transcriptomics profiles of RR and RS: including (A) Transcriptome PCA score plots of RR and RS (B) Pearson correlation analysis between three biological replicates in transcriptomics. (C) Volcano plot of DEGs between RR and RS (D) Heatmap visualization (E) GO enrichment analysis of DEGs in RR and RS (F) KEGG pathway enrichment analysis of DEGs in RR and RS.
Figure 3. Transcriptomics profiles of RR and RS: including (A) Transcriptome PCA score plots of RR and RS (B) Pearson correlation analysis between three biological replicates in transcriptomics. (C) Volcano plot of DEGs between RR and RS (D) Heatmap visualization (E) GO enrichment analysis of DEGs in RR and RS (F) KEGG pathway enrichment analysis of DEGs in RR and RS.
Molecules 31 00831 g003
Molecules 31 00831 g004
Figure 4. Combined analysis of DAMs and DEGs: including (A) KEGG enrichment analysis of co-mapped DAMs and DEGs in contrast RR vs. RS. Comparative analysis of DAMs and DEGs in the vitamin C-related metabolic pathways (B) and argininosuccinic acid-related metabolic pathways (C) based on the KEGG pathway database. The red colors represent upregulation in RR and RS, while green colors represent downregulation.
Figure 4. Combined analysis of DAMs and DEGs: including (A) KEGG enrichment analysis of co-mapped DAMs and DEGs in contrast RR vs. RS. Comparative analysis of DAMs and DEGs in the vitamin C-related metabolic pathways (B) and argininosuccinic acid-related metabolic pathways (C) based on the KEGG pathway database. The red colors represent upregulation in RR and RS, while green colors represent downregulation.
Molecules 31 00831 g004
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Wu, X.; Yuan, Y.; Yang, Y.; Yan, M.; Yang, Q.; Zha, J.; Pei, G.; Wei, Y.; Tang, L. Comparative Metabolomics and Transcriptomics Analysis of Rosa roxburghii Tratt and Rosa kweichonensis var. sterilis. Molecules 2026, 31, 831. https://doi.org/10.3390/molecules31050831

AMA Style

Wu X, Yuan Y, Yang Y, Yan M, Yang Q, Zha J, Pei G, Wei Y, Tang L. Comparative Metabolomics and Transcriptomics Analysis of Rosa roxburghii Tratt and Rosa kweichonensis var. sterilis. Molecules. 2026; 31(5):831. https://doi.org/10.3390/molecules31050831

Chicago/Turabian Style

Wu, Xiaohui, Yi Yuan, Yu Yang, Min Yan, Qiong Yang, Jun Zha, Guoshun Pei, Yali Wei, and Li Tang. 2026. "Comparative Metabolomics and Transcriptomics Analysis of Rosa roxburghii Tratt and Rosa kweichonensis var. sterilis" Molecules 31, no. 5: 831. https://doi.org/10.3390/molecules31050831

APA Style

Wu, X., Yuan, Y., Yang, Y., Yan, M., Yang, Q., Zha, J., Pei, G., Wei, Y., & Tang, L. (2026). Comparative Metabolomics and Transcriptomics Analysis of Rosa roxburghii Tratt and Rosa kweichonensis var. sterilis. Molecules, 31(5), 831. https://doi.org/10.3390/molecules31050831

Article Metrics

Back to TopTop