Metabolomics analysis has become a cornerstone technology for the modernization of Traditional Chinese Medicine (TCM), enabling comprehensive profiling of herbal chemical constituents, systematic discovery of quality markers, and mechanistic elucidation of multi-component, multi-target therapeutic actions. By integrating high-resolution mass spectrometry with advanced chemometrics and pathway analysis, metabolomics bridges the gap between empirical TCM practice and evidence-based pharmaceutical science. This article presents a technical overview of metabolomics platforms, sample types, testing items, analytical workflows, and practical applications in TCM research and quality control.
Technical Overview
Traditional Chinese Medicine is characterized by multi-component formulations, multi-target pharmacology, and holistic therapeutic principles, which present substantial challenges for conventional single-analyte quality control and mechanism-of-action studies. Metabolomics, defined as the systematic qualitative and quantitative analysis of small-molecule metabolites (typically 50–1,500 Da) within a biological system or herbal matrix, offers a systems-level solution to these challenges. When applied to TCM modernization, metabolomics enables global chemical fingerprinting of raw herbs and finished formulas, identification of quality markers correlated with clinical efficacy, and characterization of the in vivo metabolic fate of herbal constituents and their endogenous regulatory effects.
Modern TCM metabolomics relies on three complementary analytical strategies: untargeted profiling for unbiased biomarker discovery, widely targeted (pseudo-targeted) acquisition for balanced coverage and quantitative robustness, and targeted multiple reaction monitoring (MRM) for absolute quantification of predefined metabolite panels. High-resolution mass spectrometry platforms, including UHPLC coupled to quadrupole time-of-flight (Q-TOF) and Orbitrap analyzers, together with triple quadrupole LC-MS/MS and GC-MS systems, provide the sensitivity, mass accuracy, and dynamic range required to resolve the complex chemical space of herbal decoctions, extracts, and biospecimens from treated subjects. Downstream chemometric and bioinformatics pipelines convert raw spectral data into biologically interpretable results, including differential metabolites, disturbed metabolic pathways, and candidate quality markers.
Testing Samples
Herbal Materials and Extracts
Raw medicinal herbs, processed (Pao Zhi-treated) materials, powdered crude drugs, water decoctions, ethanol extracts, and intermediate products from extraction and purification processes are accepted. Samples should be homogenized to a fine powder, typically 50–200 mg per replicate, with at least 3–6 biological replicates per group. Freeze-dried materials are preferred to eliminate moisture-related variability in extraction yield.
Herbal Formulas and Finished Preparations
Multi-herb formulas, concentrated granules, tablets, capsules, oral liquids, and injections can be analyzed to characterize batch-to-batch consistency, quantify marker compounds, and evaluate compatibility-induced chemical transformations. Formula samples should be fully dissolved or extracted under validated conditions, and the complete formula composition must be documented to support data interpretation.
Biofluids
Serum, plasma, urine, cerebrospinal fluid, and saliva from animal models or clinical subjects receiving TCM interventions are used for pharmacodynamic metabolomics and pharmacokinetic studies. Recommended volumes are 100–200 µL for serum/plasma and 200–500 µL for urine. All biofluids must be collected using standardized protocols, aliquoted to avoid repeated freeze–thaw cycles, and stored at −80 °C with dry-ice shipment.
Tissues and Organs
Liver, kidney, spleen, brain, intestine, and tumor tissues from treated animal models support tissue-specific mechanism and toxicity studies. Fresh tissues should be snap-frozen in liquid nitrogen immediately after harvest; approximately 50–100 mg is required per sample, with dry-ice shipping and −80 °C storage throughout.
Cell and Microbial Culture Samples
Cell pellets (1 × 10⁶ to 1 × 10⁷ cells), culture media, and bacterial or fungal fermentation broths are applicable to in vitro efficacy screening, serum pharmacology, and herb–microbiota interaction studies. Rapid quenching of metabolism with pre-chilled solvents is essential to preserve the intracellular metabolite profile.
Fecal Samples
Fecal specimens (100–200 mg) from subjects or animals treated with TCM preparations enable investigation of gut microbiota–host co-metabolism, including bile acid remodeling, short-chain fatty acid production, and tryptophan pathway modulation, which are increasingly recognized as key mediators of TCM efficacy.
Testing Items
The metabolomics service portfolio for TCM modernization covers the following core testing items:
- Untargeted metabolomics: global detection and relative quantification of endogenous and xenobiotic metabolites for biomarker discovery and chemical profiling.
- Widely targeted metabolomics: high-throughput acquisition covering 1,000–3,000 annotated metabolites using a curated MRM library, combining discovery-scale coverage with targeted-grade quantitative precision.
- Targeted metabolomics panels: absolute quantification of specific metabolite classes, including amino acids and derivatives, bile acids, organic acids, TCA cycle intermediates, fatty acids, acylcarnitines, phospholipids and sphingolipids, neurotransmitters, nucleotides, and steroid hormones.
- TCM-specific phytochemical panels: targeted profiling of flavonoids, alkaloids, terpenoids, phenolic acids, anthraquinones, coumarins, saponins, and glycosides in herbal materials and biospecimens.
- Quality marker (Q-marker) discovery: correlation of chemical variation across herb batches, geographic origins, harvest seasons, and processing methods with bioactivity data to nominate candidate Q-markers.
- Chemical fingerprinting and pattern recognition: establishment of reference fingerprints, similarity evaluation, and discrimination of authentic versus adulterant or counterfeit materials.
- Metabolic pathway analysis: mapping of differential metabolites onto KEGG, HMDB, and Reactome pathways to identify disturbed biological processes and mechanism hypotheses.
- Integrated multi-omics interface: coordination with transcriptomics, proteomics, and 16S rRNA sequencing data for network pharmacology and multi-omics mechanism studies.
Testing Methods
Sample Preparation
Herbal samples are extracted with methanol–water or acetonitrile–water systems, typically at ratios of 70–80% organic solvent, aided by ultrasonication or vortexing, followed by centrifugation and filtration. Biofluids undergo protein precipitation with pre-chilled methanol or acetonitrile (3–4 volumes) containing isotopically labeled internal standards. For GC-MS analysis, samples are derivatized with methoxyamine hydrochloride and N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA) to enhance volatility and chromatographic behavior of polar metabolites.
Chromatographic and Mass Spectrometric Acquisition
Reversed-phase UHPLC separation (C18 column, 1.7–2.1 µm) with water and acetonitrile mobile phases, each containing 0.1% formic acid, is employed for nonpolar and moderately polar metabolites; HILIC separation is used for highly polar compounds. High-resolution MS acquisition is performed in both positive and negative electrospray ionization modes using data-dependent or data-independent acquisition for untargeted workflows. Targeted quantification uses scheduled MRM transitions on triple quadrupole instruments, with external calibration curves (r² ≥ 0.99) and isotopically labeled standards ensuring quantitative accuracy. GC-MS with electron ionization at 70 eV supports volatile and derivatized metabolite analysis following NIST library matching.
Quality Control Procedures
Pooled quality control (QC) samples, prepared by combining equal aliquots from all study samples, are injected at regular intervals (every 8–10 analytical injections) to monitor system stability. Acceptance criteria include relative standard deviations (RSD) below 30% for QC metabolite intensities, stable retention time drift within 0.2 min, and mass accuracy within 5 ppm. Blank extractions are analyzed to remove background features, and randomized injection sequences are applied to minimize batch effects.
Data Processing and Statistical Analysis
Raw data are converted to open formats (mzXML or mzML) and processed using MS-DIAL, XCMS, or Compound Discoverer for peak detection, alignment, gap filling, and annotation. Metabolite identification follows MSI confidence levels: Level 1 requires matching retention time, accurate mass, and MS/MS spectra against authentic standards; Levels 2–4 rely on database matches against HMDB, METLIN, MassBank, KEGG, LipidBlast, and in-house TCM phytochemical libraries. Univariate statistics (Student's t-test or ANOVA with FDR correction, fold-change thresholds ≥ 1.5 or ≤ 0.67) are combined with multivariate modeling, including PCA for outlier detection, PLS-DA and OPLS-DA for group discrimination, with model validity confirmed by cross-validation (R², Q²), permutation testing (200 iterations), and VIP score thresholds greater than 1.0.
Deliverables
Standard deliverables include the raw data package, metabolite quantification matrix, multivariate analysis figures (PCA, PLS-DA/OPLS-DA score and permutation plots, volcano plots, heatmaps, VIP plots), pathway enrichment bubble charts, KEGG pathway maps, candidate biomarker receiver operating characteristic (ROC) curves, and a comprehensive technical report with methodology, results, and biological interpretation.
Applications
Quality Control and Authentication of Herbal Materials
Metabolomic fingerprints enable discrimination of herbs by geographic origin, harvest time, cultivation mode, and plant part, supporting authentication of high-value materials and detection of adulteration. Correlation of chemical variation with HPLC quantification and bioassay data facilitates nomination of evidence-based quality markers for pharmacopoeial standards.
Processing (Pao Zhi) Mechanism Studies
Comparative metabolomics of raw versus processed herbs reveals processing-induced chemical transformations, such as glycoside hydrolysis, Maillard reaction products, and toxicity attenuation, providing scientific evidence for traditional processing rationale and parameter optimization.
Mechanism of Action Elucidation
In vivo metabolomics of disease models treated with herbal extracts or single compounds identifies perturbed endogenous pathways, including amino acid metabolism, lipid metabolism, bile acid signaling, and energy metabolism, generating mechanistic hypotheses that integrate with network pharmacology predictions and verified biomarker panels.
Safety and Toxicity Evaluation
Metabolomics-based toxicology detects early metabolic perturbations associated with hepatotoxicity, nephrotoxicity, or cardiotoxicity of certain herbal constituents, enabling identification of sensitive toxicity biomarkers, dose–response characterization, and assessment of toxicity attenuation by processing or formula compatibility.
Formula Compatibility and Synergy Studies
Comparative profiling of individual herbs versus complete formulas quantifies compatibility-induced chemical changes and identifies synergistic or antagonistic metabolite interactions, providing modern scientific support for classical compatibility principles such as sovereign–minister–assistant–envoy (Jun-Chen-Zuo-Shi) arrangements.
Gut Microbiota–Host Interaction Research
Integrated fecal and plasma metabolomics with 16S rRNA sequencing reveals how TCM interventions remodel gut microbial communities and their metabolites, including short-chain fatty acids, secondary bile acids, and indole derivatives, linking microbial ecology to host therapeutic outcomes.
Biomarker Discovery and Translational Research
Clinical metabolomics of TCM-treated patient cohorts supports discovery of efficacy biomarkers, patient stratification markers, and pharmacodynamic endpoints applicable to modern clinical trials and TCM-based drug development programs.
FAQ
What is the difference between untargeted and targeted metabolomics for TCM studies?
Untargeted metabolomics provides unbiased, comprehensive detection of all measurable metabolites for discovery applications but yields relative quantification and lower annotation coverage. Targeted metabolomics quantifies a predefined panel of metabolites with high sensitivity, specificity, and absolute quantification using authentic standards and MRM transitions. Widely targeted metabolomics offers an intermediate solution, covering thousands of metabolites with targeted-grade quantitative performance, and is frequently recommended for TCM studies balancing discovery scope with data reliability.
How many biological replicates are required?
For animal studies, a minimum of 6–10 biological replicates per group is recommended to achieve statistical power in multivariate modeling. For cell experiments, at least 5–6 replicates per group are advised. Clinical studies should include at least 20–30 subjects per group, with careful matching of confounding factors such as age, sex, diet, and disease stage.
How are metabolites identified and what confidence levels are reported?
Metabolite identifications are assigned according to the Metabolomics Standards Initiative confidence levels. Level 1 identifications are confirmed by matching retention time, precursor mass, and MS/MS fragmentation against authentic standards analyzed under identical conditions. Level 2 annotations are based on accurate mass and MS/MS spectral matching against public or in-house libraries, and lower levels represent putatively characterized compound classes. TCM phytochemical libraries are additionally consulted for herbal-origin metabolite annotation.
Can complex multi-herb formulas be analyzed?
Yes. Multi-herb formulas are fully compatible with the workflow. Complete documentation of the formula composition, herb ratios, and preparation method enables assignment of detected constituents to their source herbs and supports interpretation of compatibility-induced chemical changes and in vivo exposure analysis.
What are the sample shipping and storage requirements?
All biological samples should be snap-frozen in liquid nitrogen or on dry ice immediately after collection, stored at −80 °C, and shipped on dry ice with temperature monitoring. Repeated freeze–thaw cycles must be avoided. Dried herbal powders can be shipped at ambient temperature with desiccant protection.
What is the typical project turnaround time?
Standard untargeted or widely targeted metabolomics projects, from sample receipt to delivery of the complete data package and report, typically require 4–6 weeks depending on sample quantity and analysis complexity. Targeted panel projects can often be completed in 2–4 weeks, while studies requiring Level 1 standard confirmation, method validation, or integrated multi-omics analysis may require extended timelines.
What bioinformatics support is included?
All projects include standard bioinformatics deliverables: data matrix generation, univariate and multivariate statistical analysis, differential metabolite screening, KEGG pathway enrichment and mapping, correlation network analysis, and biomarker ROC evaluation. Customized analyses, including integration with transcriptomic, proteomic, or microbiome datasets and O2PLS multi-omics modeling, are available upon request.