Untargeted Metabolomics in Plant Research: Opportunities and Challenges

Untargeted metabolomics is a high-throughput technique designed to systematically identify all small-molecule metabolites in a biological sample without pre-setting targets. In plant research, it helps reveal the diversity of secondary metabolites and their functions in stress responses. This article briefly discusses its opportunities and challenges in plant research.

Core Technology: Capturing the Metabolic Fingerprint of Plants

Untargeted metabolomics primarily relies on technology platforms such as liquid chromatography-mass spectrometry (LC-MS), gas chromatography-mass spectrometry (GC-MS), and nuclear magnetic resonance (NMR) spectroscopy. LC-MS is well-suited for analyzing water-soluble and thermally unstable metabolites, offering high sensitivity; GC-MS excels at analyzing volatile and small-molecule metabolites, such as organic acids, fatty acids, amino acids, and sugars; NMR's advantage lies in its ability to provide non-destructive structural and quantitative information of metabolites without complex pretreatment.

A typical untargeted metabolomics workflow includes sample preparation, metabolite extraction and separation, data acquisition and quality control, data processing and analysis, and metabolite identification and annotation.

Each step requires careful optimization; for example, metabolite extraction typically uses solvents of different polarities, such as methanol and water, to cover a wider range of metabolites.

Unique Value: Discovering the Unknown Metabolic Landscape

The power of untargeted metabolomics lies in its comprehensiveness and unbiasedness. It does not pre-define target metabolites, enabling the discovery of potential biomarkers and novel metabolites across a broad spectrum, revealing previously unknown metabolic changes. For example, in studies of root exudates, non-targeted metabolomics methods can identify an average of approximately 874 metabolites per study, with terpenes, shikimates, phenylpropanoids, and carbohydrates being the main components.

This comprehensive perspective allows researchers to uncover the complex metabolic reprogramming of plants in response to stress. For instance, under drought stress, secondary metabolites can account for nearly three-quarters of plant root exudates, a finding that is significant for understanding plant-soil interactions.

Typical LC-MS-based metabolomics workflow (Allwood JW et al., 2021)

Opportunities of Non-Targeted Metabolomics in Plant Research

Non-targeted metabolomics opens up unprecedented possibilities for plant research, with key opportunities manifesting in the following areas:

Discovery of Novel Metabolites and Exploration of Chemical Diversity

• Opportunity Description: Plants are natural chemists, producing over 200,000 highly diverse secondary metabolites. Non-targeted metabolomics is a powerful tool for discovering entirely new natural products without prior assumptions. By comparing different species, tissues, or under specific treatments, previously unknown compounds can be discovered.
• Example: In studying a disease-resistant wild tomato relative, comparing it with cultivated varieties using non-targeted metabolomics might reveal a group of novel glycoalkaloids with antifungal activity that accumulate only in wild species (Szymański J et al., 2020).

Unveiling the Biochemical Basis of Complex Traits

• Many important agronomic traits (such as disease resistance, stress tolerance, flavor, and nutritional value) are ultimately determined by metabolomics profiles. Non-targeted metabolomics can directly link specific metabolites or metabolic networks to these phenotypes.
• Case Study: By comparing the metabolic profiles of susceptible and resistant tomatoes after pathogen infection, differential accumulation of key metabolites such as trigonelline and ACC was directly discovered, revealing a potential disease resistance mechanism (Muñoz Hoyos L et al., 2024).

Analyzing Metabolic Pathways and Gene Function

• Through integration with genomic and transcriptomic data (multi-omics integration), non-targeted metabolomics can help annotate gene function. For example, plants with a knockout or overexpression of a gene exhibit specific changes in their metabolic profile, suggesting that the gene is involved in the synthesis or regulation of a specific metabolite.
• Case Study: Overexpression of an R gene with an unknown function resulted in a significant accumulation of a certain class of flavonoids in the plant, suggesting that this disease resistance gene may function by activating flavonoid synthesis pathways (Hernandez JM et al., 2007).

Crop Improvement and Precision Breeding

• Crop germplasm resources with ideal metabolic profiles (e.g., high nutrient content, good flavor, low resistance to nutrients) can be screened, and metabolites can be used as molecular markers to assist breeding (metabolite-assisted breeding), accelerating the selection of superior varieties.
• By conducting large-scale screening of the fruit metabolomes of different tomato varieties, key metabolites related to sugar-acid ratio and aroma components are identified, serving as biochemical indicators for breeders to screen for high-quality offspring.

Plant-Environment Interaction Research

• This research comprehensively assesses how plants respond to biotic stresses (pathogens, insects) and abiotic stresses (drought, salinity, high temperature), as well as the interaction mechanisms between plants and microorganisms (mycorrhizal fungi, rhizobia), revealing the essence of their chemical dialogue.

Key Applications of Non-Targeted Metabolomics in Plant Research

Untargeted metabolomics is similar to "casting a net," aiming to detect as many small molecule metabolites as possible in a sample without bias. The following examples illustrate this characteristic:

• Untargeted metabolomics successfully identified trigonelline as a key defense metabolite against A. alternata in tomatoes, and validated its function in vitro. A total of 4496 quality characteristics were detected, of which 2499 were filtered and used for analysis. The PLS-DA model showed significant separation between A. alternata-inoculated and chitin-treated samples and the control (water-treated), while A. solani-treated samples showed less metabolic change.
  • Analysis revealed that ACC (1-aminocyclopropane-1-carboxylic acid), as an ethylene precursor, significantly accumulated after A. alternata and chitin treatment, suggesting that the ethylene signaling pathway may be involved in the resistance response.
  • Trigonelline, an alkaloid, was significantly upregulated after A. alternata and chitin treatment, but did not accumulate in A. solani treatment. Trigonelline biosynthesis is associated with the nicotinic acid/NAD cycle and may be involved in defense regulation.
  • Antifungal activity assays showed that trigonelline significantly inhibited the radial growth and spore germination of both fungi at concentrations ≥0.328 mg/mL; while its direct precursor, nicotinic acid, showed no complete inhibitory effect at the same concentration. This indicates that trigonelline possesses specific antifungal activity and may contribute to the resistance of tomato to A. alternata (Muñoz Hoyos L et al., 2024).

Antifungal activity of trigonelline on Alternaria strains (Muñoz Hoyos L et al., 2024)

• Liu XH et al. employed non-targeted metabolomics to comprehensively analyze metabolites throughout the entire pathogenic development process of *Strombus rice* (especially during appressorium formation). The goal was to identify key metabolites and metabolic pathways closely related to pathogenicity without bias.
  • Through non-targeted analysis, the study obtained a high-resolution map of dynamic metabolite changes. The most crucial finding was that the content changes of a sphingolipid called ceramide during appressorium development were highly similar to those of glycerol, a compatible solute essential for the pathogen to generate infection pressure. This indicates that ceramide is not only a structural component of the cell membrane but may also participate in key signal transduction and metabolic regulation. The non-targeted analysis revealed that the entire sphingolipid metabolic pathway is significantly activated during pathogenicity. This pathway, centered on ceramide, can synthesize more complex sphingolipids downstream, such as glucosylceramide (GlcCer) and inositol phosphoceramide (IPC).
  • This study demonstrates that the sphingolipid biosynthesis pathway is an ideal target for antifungal drugs. Targeting this pathway with small-molecule inhibitors can effectively control rice blast, providing a direct basis and feasible strategy for developing novel, green, targeted fungicides.

Metabolome profiling during appressorial development in M. oryzae (Liu XH et al., 2019)

• Silva E et al. used ultra-high performance UHPLC-MS/MS for non-targeted metabolomics analysis. Comparing the leaf metabolite profiles of rust-resistant soybean genotypes inoculated with rust fungus with the uninoculated control group, they found:
  • Upregulation of defense substances: Flavonoids (such as daidzein) and isoflavones with direct antibacterial activity were induced in large quantities as phytoalexins; simultaneously, the content of coumarins, which have unique photoactivated antibacterial capabilities, was also significantly increased.
  • Activation of signaling pathways: The enrichment of precursors of the jasmonic acid (JA) biosynthesis pathway indicated that the JA signaling pathway was activated to coordinate defense. The phenylpropanoid metabolic pathway from phenylalanine to downstream products was activated as a whole, becoming the core factory for producing defense compounds.
  • Metabolic reprogramming: The content of some terpenoid compounds decreased, suggesting that the plant may have preferentially allocated resources to more direct and effective defense pathways (such as the phenylpropanoid pathway), resulting in a precise reallocation of energy and resources.

Unsupervised chemometric modeling (UHPLC-ESI-(+)-MS/MS data) (Silva E et al., 2021)

To learn more about the applications of metabolomics in agriculture, please refer to "Metabolomics in Agriculture: Transforming Sustainability and Crop Quality".

Challenges of Untargeted Metabolomics in Plant Research

Despite its promising prospects, untargeted metabolomics faces a series of significant challenges in practical applications, primarily at the technical and methodological levels:

Metabolite Identification: The Biggest Bottleneck

Challenge Description: Of the thousands of detected "peaks," the proportion that can be clearly identified is usually very low (often less than 10%). The challenges lie in:

• Incomplete Databases: Public mass spectrometry databases (such as GNPS and MassBank) have limited coverage, and many plant-specific metabolites lack standard spectra.
• High Structural Diversity: Isomers (with the same mass but different structures) are ubiquitous and difficult to distinguish using only primary mass spectrometry and retention time. More advanced mass spectrometry techniques (such as ion mobility) or nuclear magnetic resonance are needed for confirmation, but this is usually time-consuming and expensive.

Limitations in Chemical Coverage and Dynamic Range

Challenge Description: No single extraction method and analytical technique can cover all types of metabolites. The coexistence of hydrophilic and hydrophobic, acidic and basic, and high- and low-abundance metabolites leads to:

• Extraction bias: A single extraction method may miss certain types of metabolites.
• Detection bias: Commonly used liquid chromatography-mass spectrometry (LC-MS) is sensitive to moderately polar compounds, but performs poorly in detecting volatile, highly polar, or very large molecular weight compounds, requiring the combination of gas chromatography-mass spectrometry (GC-MS) or capillary electrophoresis-mass spectrometry (CES-MS).

Data complexity and analytical challenges

Challenge description: Massive, high-dimensional data brings enormous analytical pressure.

• Complex data preprocessing: Steps such as peak detection, alignment, and retention time correction have a significant impact on the final results; different software algorithms may produce different results.
• Statistical analysis pitfalls: High-dimensional data is prone to overfitting, requiring rigorous cross-validation and permutation tests. Biological reproducibility is crucial, but the high variability of plant samples themselves increases noise.
• Difficulty in biological interpretation: Even if significantly different metabolites are found, interpreting their biological significance requires deep knowledge of plant biochemistry.

Standardization and Reproducibility

Challenge Description: From sample collection, storage, and extraction to instrumental analysis, even minor differences in any step can introduce variation, affecting the reproducibility of results and interlaboratory comparability. Establishing standard operating procedures and using quality control samples are crucial but difficult to implement.

The Functional Validation Gap

• Challenge Description: Non-targeted metabolomics excels at generating hypotheses ("This metabolite may be related to a trait"), but cannot directly demonstrate function. Moving from "correlation" to "causation" requires subsequent, cumbersome functional validation experiments, such as:
• In vitro bioassays (e.g., the antifungal assay for trigonelline in the text).
• Gene editing or chemical interventions to alter in vivo metabolite levels and observe phenotypic changes.
• Isotope labeling to trace metabolic fluxes.

For more detailed information on plant primary metabolite analysis techniques, please refer to "Analytical Techniques for Plant Primary Metabolite Profiling".

The Next Frontier in Plant Metabolomics: From Data to Discovery

Untargeted metabolomics continues to transform how we understand plant chemical systems. This powerful approach reveals complex biochemical networks with unprecedented depth. Future advances will unlock even greater potential for crop improvement and drug discovery.

Key Development Priorities

Several critical areas require focused attention to maximize metabolomics impact:

• Advanced bioinformatics tools for processing complex chemical datasets
• Standardized protocols from sample processing to data analysis
• Expanded metabolite databases for accurate compound identification
• Integrated multi-omics approaches combining metabolic and genetic insights

Connecting Chemistry to Biological Function

The true power of metabolomics emerges when we contextualize chemical profiles within biological systems. Integrating transcriptomic data with metabolic profiles, for instance, helps identify key defense compounds in disease-resistant plants. This approach allows researchers to focus on the most biologically relevant chemical signatures.

Evolutionary Context Enhances Discovery

Placing metabolic data within evolutionary frameworks reveals why plants produce specific compounds. Understanding the ecological roles of these chemicals helps predict their functions in new contexts. This perspective accelerates the identification of valuable metabolites for pharmaceutical and agricultural applications.

As technology advances and standardization improves, metabolomics will continue to reveal nature's chemical complexity. These insights will drive the next generation of plant-based innovations across multiple industries.

People Also Ask

What are the advantages of untargeted metabolomics?

One primary advantage of untargeted metabolomics is its ability to uncover novel metabolites and pathways without bias. While targeted approaches focus on specific compounds, potentially overlooking significant metabolic changes, untargeted analysis captures a broader spectrum of the metabolome.

What is the difference between targeted metabolomics and untargeted metabolomics?

Untargeted metabolomics is a global and comprehensive analysis, encompassing the measurement of all metabolites in a sample, including unknown targets. In contrast, targeted metabolomics is the measurement of a defined set of characterized and biochemically annotated analytes.

Why are untargeted metabolomics-based experiments usually considered to take a longer time than similar targeted analyses?

The data obtained from untargeted approaches are typically more complex and require extensive preprocessing, data filtering, and statistical analysis. The process of extracting meaningful information from such complex data sets can be time-consuming.

How does untargeted metabolomics work?

Untargeted Metabolomics is a "hypothesis-generating discovery strategy" that compares groups of samples (e.g., cases vs controls); identifies the metabolome and establishes (early signs of) perturbations.

References

1. Tugizimana F, Mhlongo MI, Piater LA, Dubery IA. Metabolomics in Plant Priming Research: The Way Forward? Int J Mol Sci. 2018 Jun 13;19(6):1759.
2. Allwood JW, Williams A, Uthe H, van Dam NM, Mur LAJ, Grant MR, Pétriacq P. Unravelling Plant Responses to Stress-The Importance of Targeted and Untargeted Metabolomics. Metabolites. 2021 Aug 22;11(8):558.
3. Muñoz Hoyos L, Anisha WP, Meng C, Kleigrewe K, Dawid C, Hückelhoven R, Stam R. Untargeted metabolomics reveals PTI-associated metabolites. Plant Cell Environ. 2024 Apr;47(4):1224-1237.
4. Liu XH, Liang S, Wei YY, Zhu XM, Li L, Liu PP, Zheng QX, Zhou HN, Zhang Y, Mao LJ, Fernandes CM, Del Poeta M, Naqvi NI, Lin FC. Metabolomics Analysis Identifies Sphingolipids as Key Signaling Moieties in Appressorium Morphogenesis and Function in Magnaporthe oryzae. mBio. 2019 Aug 20;10(4):e01467-19.
5. Silva E, Perez da Graça J, Porto C, Martin do Prado R, Nunes E, Corrêa Marcelino-Guimarães F, Conrado Meyer M, Jorge Pilau E. Untargeted Metabolomics Analysis by UHPLC-MS/MS of Soybean Plant in a Compatible Response to Phakopsora pachyrhizi Infection. Metabolites. 2021 Mar 19;11(3):179.
6. Szymański J, Bocobza S, Panda S, Sonawane P, Cárdenas PD, Lashbrooke J, Kamble A, Shahaf N, Meir S, Bovy A, Beekwilder J, Tikunov Y, Romero de la Fuente I, Zamir D, Rogachev I, Aharoni A. Analysis of wild tomato introgression lines elucidates the genetic basis of transcriptome and metabolome variation underlying fruit traits and pathogen response. Nat Genet. 2020 Oct;52(10):1111-1121.
7. Hernandez JM, Feller A, Morohashi K, Frame K, Grotewold E. The basic helix loop helix domain of maize R links transcriptional regulation and histone modifications by recruitment of an EMSY-related factor. Proc Natl Acad Sci U S A. 2007 Oct 23;104(43):17222-7.