Differential Metabolomics

Metabolomics is the global profiling of small molecular weight compounds of less than 1,500 daltons called metabolites. Metabolites within the metabolome represent the end point of biological metabolic activity starting from the genome and extending through protein function.¹ Differential metabolomics has emerged as an approach for designing a set of experimental questions, hypotheses, and protocols aimed at comparing control samples with experimental conditions and extracting  semi-quantitative assessment of differential of metabolites—metabolites that show dissimilar associations between different physiological states. Studying the effects of perturbations or treatment provides insights into potential next steps toward diagnoses or treatments.

Metabolite levels can be influenced by drugs, genetic mutations, disease, or dietary intake, and many researchers are focused on investigating the metabolome in the context of control vs. treated samples. Differential metabolomics compares samples to a control to understand these influences.^2-4

Applications of Differential Metabolomics

The advancement in understanding physiology and disease biology has been greatly facilitated by the utilization of differential metabolomics. Since metabolites and their associated pathways have widespread roles across mammalian physiology, researchers and clinicians are now able to select the metabolite systems that are relevant to their research questions. Differential metabolomics analysis finds application across many molecules and systems, spanning from amino and fatty acids to lipidomic analysis and cannabinoids.

A pivotal aspect of differential metabolomics is its requirement for measuring metabolites under diverse conditions (e.g., disease or other perturbations). This method has been heavily applied to investigate many diseases including cardiovascular and metabolic disorders, neurodegenerative diseases, various cancers, and drug discovery. Additionally, differential metabolomics is also utilized in nutrition, food safety, and agriculture for quality control, yield, and crop optimization.

Insights Gained from Differential Metabolomics

A wealth of research aimed at understanding changes in the metabolome has advanced our understanding of physiology, drug discovery, biomarkers, and disease monitoring. For instance, one group of researchers examined sex differences in metabolic profiles using differential metabolomics. To achieve this, scientists analyzed plasma metabolites from males and females using a differential correlation network analysis approach. Their results identified key sex-specific interconnection differences in metabolites associated with gut microbiota composition, suggesting sex-specific distinctions in metabolic profiles that could provide potential therapeutic targets.⁵

Others have utilized differential metabolomics to identify new biomarkers. In a study examining liver health, researchers investigated the impact of acetaminophen on metabolite levels in mouse liver and serum.⁶ By comparing two mouse models of acetaminophen-induced hepatotoxicity using coupling capillary electrophoresis with electrospray ionization time-of-flight mass spectrometry (CE-TOFMS), their findings revealed differences in metabolite levels in the two mouse models and identified phenylpropionic acid as a modulator of liver injury.

Differential metabolomics can also facilitate comparisons of metabolite profiles among diverse genotypes. For example, one report examined differences in metabolic fingerprints in treatment response to non-small cell lung cancer and colon cancer, with a particular focus on KRAS, the most frequently mutated form of the RAS oncogene in cancer, and its relationship to ixazomib, a promising antitumor drug.⁷ Interestingly, their findings showed a correlation between the KRAS genotype and ixazomib sensitivity where KRAS-G13D tumors had higher oxidative stress and glucose utilization and did not show the same metabolic regulation by ixazomic that was seen in the wildtype KRAS. Together, this finding is an excellent example of how differential metabolomics can advance the understanding of cancer pathology and help guide treatment strategy.

Similarly, differential metabolomics has been integral in understanding treatment responses. In the context of cardiovascular diseases, one study compared pre- and post-treatment of two commonly used antiplatelet drugs (i.e., P2Y12 inhibitors), ticagrelor, and clopidogrel.⁸ Their findings revealed distinct metabolic fingerprints that are capable of categorizing patients based on their responses to these P2Y12 inhibitors. These signatures implicated polyunsaturated fatty acids and omega-3 fatty acids as sensitive metabolites that can inform treatment responses.

Challenges Faced in Differential Metabolomics

Despite the many important contributions made possible by differential metabolomics studies, there are some critical challenges to performing high-quality studies that yield actionable results.

Generally, researchers use multivariate and univariate statistics in metabolome differential analysis—some of the more common protocols include principal component analysis (PCA) or partial least square discriminant analysis (PLS) to allow for binary comparisons and differentiation.^9,10 A significant challenge to this approach, however, is the lack of a consistent standard for statistical comparisons among researchers. Due to this inconsistency, interpreting causal factors for metabolite fluctuations can be difficult. 

Fluctuations can vary depending on physiological state or perturbations, and researchers need to consider several factors to appropriately interpret their findings. These can range from localized intervention of biological pathways (e.g., over-expression or knockout of enzyme-coding genes) to systemic perturbations (e.g., responses to stress or time-series measurements) to intrinsic variability not attributed to disease-related alterations but rather stemming from underlying variability in cellular metabolism.¹¹ Taken together, these factors often introduce several confounding variables. For instance, studying metabolite fluctuations due to circadian rhythms can uncover correlations between metabolites, but may not offer insights into causes beyond the time series. Additionally, transient disturbances can lead to minimal or no significant changes, yet metabolites inevitably correlate with each other.

Furthermore, as metabolomics has emerged as a valuable tool for examining cellular processes, there has been a lack of emphasis on system-level biological processes and regulatory systems. This issue is observed across omics studies, including differential metabolomics.¹² 

To translate metabolomic findings to clinical applications, experimental questions, hypotheses, and protocols must be adapted to incorporate the networks governing changes in physiological state and disease. Coupled with systematic statistical analyses, the intricate changes in metabolite levels can be understood in the context of a specific disorder, subsequently enabling clearer associations between metabolites and disease.

References

1. Beebe K, Kennedy AD. Sharpening Precision Medicine by a Thorough Interrogation of Metabolic Individuality. Comput Struct Biotechnol J. 2016;14:97-105. Published 2016 Jan 21. doi:10.1016/j.csbj.2016.01.001
2. Zhou W, Yao Y, Scott AJ, et al. Purine metabolism regulates DNA repair and therapy resistance in glioblastoma. Nat Commun. 2020;11(1):3811. Published 2020 Jul 30. doi:10.1038/s41467-020-17512-x
3. Mikaeloff F, Svensson Akusjärvi S, Ikomey GM, et al. Trans cohort metabolic reprogramming towards glutaminolysis in long-term successfully treated HIV-infection. Commun Biol. 2022;5(1):27. Published 2022 Jan 11. doi:10.1038/s42003-021-02985-
4. Schwerdtfeger LA, Nealon NJ, Ryan EP, Tobet SA. Human colon function ex vivo: Dependence on oxygen and sensitivity to antibiotic. PLoS One. 2019;14(5):e0217170. Published 2019 May 16. doi:10.1371/journal.pone.0217170
5. Verhaar BJH, Mosterd CM, Collard D, et al. Sex differences in associations of plasma metabolites with blood pressure and heart rate variability: The HELIUS study [published online ahead of print, 2023 May 28]. Atherosclerosis. 2023;S0021-9150(23)00210-1. doi:10.1016/j.atherosclerosis.2023.05.016
6. Cho S, Yang X, Won KJ, et al. Phenylpropionic acid produced by gut microbiota alleviates acetaminophen-induced hepatotoxicity. Gut Microbes. 2023;15(1):2231590. doi:10.1080/19490976.2023.2231590
7. Chattopadhyay N, Berger AJ, Koenig E, et al. KRAS Genotype Correlates with Proteasome Inhibitor Ixazomib Activity in Preclinical In Vivo Models of Colon and Non-Small Cell Lung Cancer: Potential Role of Tumor Metabolism. PLoS One. 2015;10(12):e0144825. Published 2015 Dec 28. doi:10.1371/journal.pone.0144825
8. Samman, KN, Mehanna, P, Takla, E, et al. Differential modulation of polyunsaturated fatty acids in patients with myocardial infarction treated with ticagrelor or clopidogrel. Cell Rep Med 2021;(2):100299.
9. Pinto, RC. Chemometrics Methods and Strategies in Metabolomics. Adv Exp Med Biol 2017;(965):163-190.
10. Worley, B, and Powers, R. Multivariate Analysis in Metabolomics. Curr Metabolomics 2013;(1):92-107.
11. Steuer, R. Review: on the analysis and interpretation of correlations in metabolomic data. Brief Bioinform 2006;(7):151-158.
12. Schadt, EE. Molecular networks as sensors and drivers of common human diseases. Nature 2009;(461):218-223.