Metabolic phenotypes: Molecular bridges between health homeostasis and disease imbalance

Abstract

Metabolic phenotypes represent the overall characterization of an individual’s metabolites at a specific point in time. They precisely reflect the complex interactions among genetic background, environmental factors, lifestyle, and gut microbiome, thereby serving as a key molecular link between healthy homeostasis and disease-related metabolic disruption. In recent years, high-throughput metabolomics strategies have enabled the systematic analysis of small molecule metabolites in physiological and pathological processes. These metabolites not only serve as biomarkers for disease diagnosis, prognosis assessment, and treatment response prediction, but also elucidate novel mechanistic pathways in disease progression. The high-coverage, high-sensitivity detection of metabolites afforded by mass spectrometry and NMR-based metabolomics enables advances in precision medicine, facilitating biomarker discovery, pharmacokinetic studies, and the assessment of nutritional interventions. This review uses several common metabolic diseases, such as obesity, diabetes, cardiovascular diseases, and cancer, to explore the key role of metabolic phenotypes in disease risk stratification and precise prediction. Future phenotypic research will shift toward integrating artificial intelligence, big data mining, and multi-omics with the goal of revealing the complete network through which metabolic phenotypes regulate diseases. This research is expected to advance early diagnosis, precise prevention, and targeted treatment, contributing to a medical paradigm shift from disease treatment to health maintenance.

Graphical Abstract

Highlights

• •Metabolic phenotypes bridge health and disease by integrating genetic, environmental, and microbiome influences.
• •High-throughput metabolomics enables precise biomarker discovery for early disease diagnosis and prognosis.
• •Multi-omics integration reveals mechanistic links between metabolic dysregulation and complex diseases.
• •Metabolic reprogramming in cancer and diabetes highlights potential therapeutic targets and interventions.
• •AI and spatial metabolomics advance personalized medicine through dynamic metabolic profiling.

1. Introduction

The proposal of metabolomics at the end of the 20th century marked a new era for studying metabolic networks in organisms. [1] Since the metabolic phenotype was defined as the systemic metabolic description of an organism under specific physiological conditions, it has become a core bridge connecting genes with macroscopic health phenotypes and an important frontier in life sciences. [2] Metabolic phenotypes offer a brand-new perspective on health and disease. This perspective moves beyond the isolated examination of individual indicators to focus on explaining the dynamic biological interactions behind them. Consequently, research in this field not only provides a more accurate assessment of health and disease for individuals but also delves into underlying disease mechanisms. Furthermore, it guides drug development and safe medication practices, while also providing powerful tools for scientific research and clinical diagnosis.

The rapid evolution of metabolomics, driven by advances in high-coverage analytical methods, metabolic flux analysis, and bioinformatics databases, has significantly enhanced the efficiency and accuracy of metabolomic studies. These advancements, encompassing revolutionary breakthroughs like spatial metabolomics and in vivo monitoring, not only demonstrate multidimensional breakthroughs in the technology itself but also mark the dawn of a new era for metabolic phenotype research. [3], [4] Metabolomics technology has overcome key limitations of traditional diagnostic methods, such as insufficient sensitivity and a reliance on single markers, by providing comprehensive, dynamic metabolic profiling. This capability holds significant clinical potential for early disease detection and precise risk prediction. For example, N1-acetylspermidine is a potential blood biomarker for T lymphoblastic leukemia/lymphoma,and compounds such as succinate, uridine, and lactate have been implicated as biomarkers for the early diagnosis of gastric cancer. [5], [6] Similarly, Kanzonol Z, Xanthosine, Nervonyl carnitine and other markers in urinary extracellular vesicles were found, which can be used for early diagnosis of lung cancer. [7] Secondly, metabolic phenotype research is directly related to the effectiveness, safety, and individualized application of drugs, and is a crucial link in translational medicine. [8], [9] For example, customized nutrient intake based on urine metabolic profiles can effectively control the weight of obese patients. [10] Directing sunlight exposure regimens for vitamin D metabolic pathways can effectively reduce the risk of osteoporosis in people. [11] Notably, strategies targeting metabolic vulnerabilities in cancer, such as targeted restoration of hepatocellular carcinoma leucine metabolism to inhibit liver cancer progression. [12]The metabolic phenotype lies at the intersection of genetic, environmental, and other phenotypic factors, functioning as a crucial "bridge" for analyzing the mechanisms of complex diseases. Many physiological or pathological states can be directly interpreted by observing alterations in metabolites. By integrating metabolome data with clinical phenotypes, it is possible to pinpoint the key metabolites most relevant to diseases. For example, the accumulation of amyloid-beta peptides is a well-validated biomarker for Alzheimer's disease. [13], [14]

Metabolic phenotypes serve as molecular keys to deciphering the mechanisms of complex diseases. Unlike traditional single-target approaches, which often fail to fully explain disease processes involving multiple metabolic pathways, metabolic phenotypes provide comprehensive physiological fingerprints of an organism's functional state. They effectively reflect physiological and pathological conditions across various levels, from small molecules to the whole organism. This article examines both health homeostasis and metabolic imbalances in diseases such as obesity, diabetes, cardiovascular diseases, and cancer. This study aims to provide a basis for the early warning and precise prediction of diseases, as well as for guiding drug research and development and clinical medication, by systematically comparing the characteristic changes under different metabolic disease state.

2. Biological basis of metabolic phenotype

Metabolic phenotypes arise from the interplay of genes, the environment, and microorganisms, and this dynamic interaction directly shapes the output of physiological functions and the expression of disease phenotypes in living organisms. Genome-wide association studies (GWAS) form the foundation for understanding genotype-phenotype relationships. Integrated multi-omics analysis builds upon this foundation by revealing the complex, multi-dimensional molecular regulatory networks within organisms. This integrated approach provides a comprehensive view of how genetic variants influence gene expression and metabolite levels, ultimately leading to phenotypic changes, thereby enabling an in-depth analysis of the genetic mechanisms underlying complex traits. Genetic polymorphisms play a critical role in driving metabolic variation. As an illustration, APOE genetic variants are well-established modulators of lipid metabolism. [15] Genetic polymorphism of CYP450 exogenous liver metabolic enzyme significantly affects drug metabolic efficiency and toxicity risk. [16], [17] Genes, as the initiators of metabolic phenotypes, undertake the important task of communicating with phenotypes from the biological framework of genome-transcriptome-proteome-metabolome.

The gut microbiota shapes the host's metabolic phenotype primarily through the synthesis of various metabolites. Acting as a crucial regulator, it influences metabolic processes, engages in co-metabolic activities, and contributes to inter-individual variations. One important regulatory mechanism involves short-chain fatty acids (SCFAs), which the gut microbiota exerts significant effects on energy absorption, insulin sensitivity, and inflammation. [18], [19] Similarly, the gut microbiota maintains the stability of the host’s microniche by maintaining host-flora co-metabolism. It participates in bile acid metabolism and vitamin synthesis, and directly regulates host lipid and glucose homeostasis. For example, gut microbiota may affect obesity and obesity susceptibility by affecting host lipid changes, histidine metabolism, linoleic acid metabolism, and other pathways. [20] Differences in microbiota composition are also associated with the risk of metabolic diseases such as obesity and diabetes. For example, research on obesity has demonstrated that dietary fiber helps alleviate the condition by regulating intestinal microbiota homeostasis, restoring intestinal barrier function, and improving obesity. [21], [22] Microorganisms affect the host's normal physiological, immune, and metabolic processes by producing metabolites, thereby promoting the emergence of novel metabolic phenotypes and equipping the host with additional physiological capabilities. [23], [24]

Environmental factors and xenobiotic exposure can significantly shape an individual's metabolic phenotype. As a key manifestation of these factors, diet plays a crucial role not only in establishing this phenotype but also in contributing to its long-term stability. Diseases such as metabolic diseases diabetes and obesity are indirectly associated with unhealthy dietary performance. For example, a high-fat diet induces lipid synthesis gene expression, alters fatty acid metabolism, and develops a tendency toward obesity. [25] Similarly, healthy eating behaviour is also conducive to the prevention and control of obesity. For example, restricting diet can reduce fat tissue weight and promote lipolysis, improving obesity symptoms. [26], [27] Various xenobiotics—including pharmaceuticals, personal care products, food additives, and environmental pollutants—can alter an individual's metabolic phenotype through multiple mechanisms. For example, naringin may maintain the lipid homeostasis of macrophages by downregulating lipid uptake genes, upregulating cholesterol metabolism genes, and reducing the release of proinflammatory factors to restore macrophage phenotype and function. [28] The use of cosmetic products can also impact metabolic phenotypes. For example, exposure to contaminants in these products has been linked togastrointestinal dysbiosis and renal metabolic dysfunction. [29] Similarly, prevalent food additives can adversely impact host health by compromising the gut microbiota, disrupting the intestinal barrier, and contributing to metabolic and neurobehavioral abnormalities. [30]

3. Characteristic map of healthy metabolic phenotypes

The characteristic map of healthy metabolic phenotypes is a multi-dimensional and dynamic evaluation system, which aims to comprehensively define the metabolic health status of the human body from the static and dynamic perspectives. Healthy metabolic phenotypes are coordinated by the dynamic balance of metabolism and static of biomarkers. [31] Dynamic balance refers to the host's capacity to restore metabolic homeostasis in response to external stimuli like diet or drugs, as observed in changes in blood sugar and lipid levels. In contrast to the static snapshot provided by biomarker levels, dynamic balance is continually shaped by factors such as physical activity, gut microbiota homeostasis, and circadian rhythms [32], [33].

Static biomarkers, as key indicators for assessing health phenotypic status, cover traditional clinical indicators and new molecular markers. Traditional clinical indicators such as fasting blood glucose, triglycerides, and blood pressure may reflect the body's basal metabolic state. Compared to traditional biomarkers, novel biomarkers offer significant advantages, as they can more accurately predict metabolic changes and are less susceptible to interference from confounding factors. For instance, serum glycosylated hemoglobin (HbA1c) levels reliably reflect average blood glucose concentrations over the preceding 2–3 months, while elevated fasting levels of branched-chain amino acids are a strong indicator of early insulin resistance. [34]

A healthy metabolic phenotype is characterized by robust circadian metabolic rhythms, where daily fluctuations in metabolic processes are synchronized with the body's physiological needs. For instance, insulin sensitivity peaks in the morning and declines throughout the day, while hepatic gluconeogenesis increases at night to maintain glucose homeostasis during fasting. Daytime food intake and physical activity optimally stimulate these anabolic processes. Conversely, disruptions to this temporal organization, such as nighttime eating, can have negative metabolic consequences by inhibiting fat oxidation, promoting lipid storage, and increasing the risk of obesity. [35] Similarly, disordered rhythms increase the risk of metabolic diseases, leading to a decrease in adiponectin and an increase in inflammatory factors, which promote the accumulation of white fat. Uncontrolled hepatic gluconeogenesis can lead to the risk of fasting hyperglycemia.

4. Characteristics of metabolic phenotypes of disease

The disease metabolic phenotype refers to a state of systemic metabolic dysfunction caused by the interplay of genetic, environmental, and lifestyle factors, and it manifests common pathological features across many chronic diseases. A key hallmark shared by conditions ranging from cancer to metabolic disorders is impaired mitochondrial oxidative phosphorylation, which severely compromises cellular energy production. For example, tumor cells rely on aerobic glycolysis, meaning they preferentially utilize the glycolytic pathway for energy production even when oxygen is sufficient, resulting in lactic acid accumulation and acidification of the microenvironment. [36]In diabetes and obesity, mitochondrial function is partially inhibited, leading to a decrease in ATP synthesis and the accumulation of reactive oxygen species (ROS). [37]Second, chronic low-grade inflammation and oxidative stress are also common features of disease metabolic phenotypes. [38] Multiple metabolic network disorders are common features of disease metabolic phenotypes. For example, VLDL and LDL were significantly increased in liver cirrhosis, MASH, and cardiovascular diseases, reflecting impaired liver lipid output and clearance disorders. [39]Aromatic amino acids such as phenylalanine and tyrosine accumulate abnormally in salt-sensitive hypertension and diabetes, and are positively associated with cardiovascular risk. [40] In the following sections, we examine several common diseases for discussion, elaborating on the key positioning of metabolic phenotypes in different diseases.

4.1. Metabolic phenotypic characteristics of Metabolic obesity

Obesity, as a typical metabolic disease, has obvious metabolic characteristics. Energy metabolism disorder is the core of obesity. Obesity is mainly manifested as the imbalance of fat production and lipolysis, resulting in excessive accumulation of fat and metabolic disorders. By detecting markers of lipolysis and production, dynamic changes in lipid metabolism in obesity can be analyzed to guide the exploration of potential therapeutic strategies. [41], [42] Lipid metabolism represents a fundamental mechanism for maintaining homeostasis across multiple organs, including the kidneys, liver, and cardiovascular system. Environmental factors can disrupt lipid metabolism via influencing factors such as specific susceptibility windows, gender differences, and dietary habits. Such disruptions can promote obesity and subsequently alter the levels of key lipid molecules like ceramide and sphingomyelin in both plasma and adipocytes. [43], [44] Lipid changes caused by obesity are also key to exacerbating other complications. For example, lncRNAs mediating lipolysis by activating lipid transcription factors and positively regulating the transcription of genes involved in lipid metabolism. [45] Most obese patients exhibit impaired glucose metabolism, which is not only related to mitochondrial function, but also to levels of oxidative stress. Glucose metabolism occupies a central position in obesity research, mainly because dysregulation and impairment of glucose metabolism levels are fundamental factors in insulin resistance and metabolic dysfunction. [46]

In obese patients, alterations in lipid metabolism were particularly pronounced. Studies indicate that levels of certain phospholipids, such as phosphatidylcholine and a specific form of phosphatidylglycerol, were significantly reduced. In contrast, most species of phosphatidylethanolamine and phosphatidylinositol showed a significant increase. [47] In addition, a key pathophysiological feature of obesity is resistance to metabolic hormones - Leptin and Ghrelin. [48], [49]

4.2. Metabolic characteristics of diabetes

Diabetes mellitus is a chronic metabolic disease characterized by hyperglycemia caused by insulin deficiency or resistance. It is mainly affected by genetic and environmental factors, and is manifested as the change in blood sugar caused by insufficient insulin utilization or insufficient insulin sensitivity of the body. Epidemiological studies have found that lifestyle, excess nutrition and insufficient physical activity are the main causes of diabetes mellitus. Thus, adopting healthier lifestyle habits is fundamental to the prevention and management of the disease. Although the diagnosis of diabetes is based on blood glucose levels or the glycated hemoglobin (HbA1c) value, it is recognized as a multisystem disease. Diabetes poses a major global challenge. Its pathological mechanism involves abnormal accumulation or deletion of various metabolites, and early prevention and intervention are the key. For example, key metabolites such as glucokinase, glycoisoenzyme, and SGLT-2 in sugar metabolism can reduce sugar levels in the body by promoting glycolysis, gluconeogenesis, and sugar reabsorption. [50], [51] Changes in metabolites associated with lipid metabolism, such as free fatty acids and ketone bodies, can serve as key indicators for monitoring the progression of diabetes. Diabetes is a metabolic disorder primarily characterized by dysregulated glucose metabolism, as well as disturbances in lipid and protein metabolism. (Fig. 1) In addition, it is also found that bile acid metabolism, amino acid metabolism and cholesterol metabolism are abnormally disturbed. [52], [53] A primary manifestation of diabetes mellitus is reduced glucose uptake and utilization in skeletal muscle, fat, and liver. This occurs due to reduced insulin secretion or insulin resistance, resulting in elevated blood sugar. [54] Due to abnormal glucose metabolic pathways, hyperglycemia induces the activation of alternative glucose metabolic pathways, including the enhancement of oxidative stress and inflammatory pathways, which are also the main causes of diabetes complications. Secondly, changes in serine and branched-chain amino acids are the most common metabolic changes in diabetes and its complications, which are mainly due to the reduction of precursors caused by impaired glycolysis. [55] Insulin resistance, hyperglycemia and dyslipidemia are the main factors of diabetes complications, such as abnormal increases in triglycerides and significant cholesterol levels, which can easily lead to fat deposition, which is also the main reason why diabetic patients are prone to macrovascular and microvascular complications. [56], [57]The metabolic phenotype of diabetes profoundly reveals the essence of its pathogenesis and pathological changes. It is far from a simple ‘elevated blood sugar’, but rather a systemic disorder involving multiple organs and metabolic pathways throughout the body.

Fig. 1.

Metabolic changes in diabetes are caused by reduced insulin sensitivity or insulin resistance. Including abnormal glucose metabolism, abnormal protein metabolism and abnormal lipid metabolism. Due to insufficient insulin secretion or resistance, the utilization of glucose is impaired, resulting in persistently elevated blood sugar levels, accompanied by typical symptoms of diabetes. At the same time, it shows insufficient energy supply, further promoting the breakdown of fat and protein, leading to weight loss and increasing the risk of ketoacidosis.

4.3. Cardiovascular disease metabolic remodeling

Cardiovascular disease has become the leading cause of death threatening public health and is particularly prevalent among middle-aged and elderly patients. Its pathogenesis is primarily due to atherosclerosis, hypertension, hyperlipidemia, and abnormalities in blood composition. WHF data show that cardiovascular diseases will account for about one-third of all global deaths in 2021 and remain the leading cause of death worldwide. [58] Modern studies have shown that various metabolic hormones and enzymes play crucial roles in regulating cardiovascular function. Key among them are insulin, glucagon, and glucocorticoids, all of which participate in the control of myocardial contractility and vascular tone. Increased glycolysis is considered one of the markers of metabolic remodeling in cardiovascular diseases. The glycolytic pathway can provide energy to cardiomyocytes during ischemia and reperfusion, and can also improve cardiovascular symptoms through the pentose phosphate pathway and the polyol pathway. [59], [60] Secondly, abnormal lipid metabolism is also a predisposing factor for cardiovascular diseases, such as atherosclerosis caused by abnormal metabolism of low-density lipoprotein and triglycerides. [61] Likewise, lipids are involved in the regulation of inflammation, and adipose tissue releases inflammatory factors such as IL-6 and TNF-α, which contribute to chronic inflammation of the blood vessel walls. Second, lipid disorders also lead to oxidative stress disorders. The accumulation of large amounts of reactive oxygen species also damages the cardiovascular system and forms an "inflammation-oxidation" environment. [62], [63] The metabolic pathways of amino acids are closely linked to the development of cardiovascular diseases. The underlying mechanisms primarily involve their roles in glucose metabolism, oxidative stress, and the regulation of inflammatory responses. For example, taurine metabolism may reduce the risk of non-ischemic cardiomyopathy by reducing oxidative stress; [64] Changes in serum linoleic acid metabolism in patients with coronary heart disease indirectly affect inflammation regulation. [65]

The occurrence of cardiovascular disease (CVD) is closely related to metabolic disorders, and a variety of key metabolites participate in pathological processes by affecting inflammation, oxidative stress, platelet activity, and energy metabolism. Studies have found that benzoyl glutamine, a secondary metabolite of intestinal flora, can enhance platelet activity, promote atherosclerosis, target inhibition or reduce thrombosis and reduce the risk of atherosclerosis. [66] Secondly, lipid molecules represented by long-chain monounsaturated fatty acids and phosphatidylcholine are also related to atherosclerosis, and their mechanism of action is mainly related to the regulation of fatty acid oxidation. [67] Succinic acid is involved in energy metabolism and inflammation regulation. For example, under hypoxic conditions, succinic acid accumulates and activates HIF-1α, promotes the release of inflammatory factors, and exacerbates myocardial ischemia. [68] In addition, bile acids and tryptophan metabolites have demonstrated therapeutic potential in cardiovascular diseases. For example, bile acids exert beneficial effects by regulating GLP-1 secretion and influencing hepatic glucose and lipid metabolism. [69] When viewed through the lens of their metabolic phenotypes, cardiovascular diseases emerge as complex metabolic conditions rather than merely hemodynamic disorders. This shift in perspective facilitates earlier diagnosis and more targeted therapies.

4.4. Cancer Metabolic Reprogramming

Warburg effect and mitochondrial functional changes are key to cancer metabolic reprogramming. Cancer cells generate energy primarily through glycolysis, leading to high glucose consumption and lactic acid production. [70], [71] Although this process is less efficient in terms of energy yield per glucose molecule, its rate is much faster than mitochondrial oxidative phosphorylation. This rapid glycolytic flux provides both energy and biosynthetic intermediates to support rapid tumor growth. [72] Second, mitochondrial function remains active in some cancers, where energy is derived from OXPHOS or glutamine breakdown. A key adaptation is seen under nutrient stress: tumors activate the GCN2-ATF4-DDIT3 signaling axis to inhibit the expression of mitochondrial proteins. This mechanism concurrently reduces ROS production and enhances glycolytic flux, ensuring a continuous ATP supply. [73] Currently, inhibitors targeting glutamine metabolism represent a promising area in cancer research. These drugs aim to cut off the abnormal "nutrient supply" that tumor cells depend on, thereby inhibiting their growth. However, no such inhibitors have yet been approved for clinical use. This lack of translation is largely attributed to challenges such as tumor heterogeneity and drug resistance. Furthermore, a significant challenge lies in achieving precise targeting of tumor cells by these inhibitors without impairing anti-tumor immunity.

For a start, hyperanabolism, glutamine deprivation, and lipid metabolism reprogramming are another metabolic feature of cancer. Glycolytic intermediates affect serine and glycine synthesis through phosphoglycerate dehydrogenase. As an essential metabolic fuel glutamine is utilized by many rapidly proliferating cancer cells to meet their heightened energy demands. [74](Fig. 2). Similarly, lipid metabolic reprogramming supports cancer cell proliferation by enhancing de novo fatty acid synthesis and upregulating the expression of key enzymes expression, thereby facilitating membrane phospholipid synthesis. [75]

Fig. 2.

The characteristics of glutamine metabolism in cancer - Glutamine replenishment drives the TAC. Tumor cells use a large amount of nitrogen atoms of glutamine for biosynthesis, rather than mainly for energy production as some normal cells do. Glutamine enters the tricarboxylic acid cycle through a ‘replenishment reaction’ and is converted into intermediate substances such as α -ketoglutaric acid, playing the role of ‘supplementary fuel’. This enables the tricarboxylic acid cycle to continue operating, ensuring the production of energy (ATP) while providing the necessary carbon framework for other biosynthetic pathways. The figures created by BioRender.

Microenvironmental remodeling and metabolic interactions jointly affect cancer cells. Lactate is secreted into the microenvironment through the monocarboxylic acid transporter (MCT-1), reducing pH, inhibiting T cell function and promoting M2 macrophage polarization, forming an immunosuppressive microenvironment. [76] Cancer cells consume glucose, leading to glucose deficiency in the microenvironment, inhibiting T cell metabolism and activity. Cancer cells in the same hypoxic zone break down their own components through autophagy, releasing lactic acid for use by neighboring cancer cells.

A hallmark of cancer is metabolic reprogramming, wherein cancer cells rewire their metabolic pathways to secure energy and biosynthetic building blocks for unchecked growth. This rewiring encompasses glucose, lipid, and glutamine metabolism, ensuring supplies for elevated biosynthetic demands. However, metabolic biomarkers in cancer, such as ceramides and ketone bodies, demonstrate lower robustness and greater heterogeneity compared to those in other metabolic diseases. Their levels often mirror the overall metabolic dysregulation and are complicated by tumor heterogeneity and microenvironmental factors, making their link to specific cancer types intricate. Consequently, despite advances, tumor metabolism research is primarily exploratory and has not been integrated as a cornerstone in standard clinical practice.

5. Metabolic phenotypic research technology system

The metabolic phenotyping technology system functions as an integrated framework that encompasses sample preparation, data acquisition, bioinformatics, and multi-omics validation. Its power stems not from any individual component, but from the synergistic integration of these technologies, which empowers researchers to holistically and dynamically quantify the metabolic underpinnings of health and disease.

5.1. Metabolic Phenotyping Technology Platform

In clinical metabolic phenotype analysis, the key to success lies not only in advanced instruments but also in a powerful bioinformatics platform and algorithms. These tools transform complex raw data into interpretable biological insights. As a key technology for metabolomic analysis, mass spectrometry is increasingly used for the monitoring and detection of metabolites due to its high throughput and coverage. For example, non-targeted metabolomic analysis of urine from patients with urinary tract infections using mass spectrometry technology has screened two key metabolites, agmatine and N6-methyladenine, which can be used as potential biomarkers for identifying urinary tract infections. [77] Similarly, novel ionization technologies such as DESI and LDI-MS simplify the sample preparation process and lower the detection limit, thereby enabling the screening of lung cancer biomarkers. [78] The non-destructive nature of NMR technology makes it particularly valuable for metabolome analysis, as it allows for dynamic monitoring and preservation of the native metabolic state in intact tissue samples by avoiding extraction-induced damage. For example, based on H NMR analysis, the correlation between 168 metabolites in serum and heart failure was investigated. [79] Similarly, the multi-dimensional spectrum detection technology can specifically detect phosphorylated metabolites and analyze the state of energy metabolism. [80]

The integration of spatial metabolomics and single-cell technology is currently a cutting-edge direction in the field of life sciences, aiming to analyze the spatial distribution of metabolites in biological tissues and cell type-specific metabolic states, thus revealing the molecular mechanisms of complex biological processes. MALDI-MSI, DESI-MSI and SIMS techniques can combine metabolic and spatial information to describe their spatial variation information. For example, analysis of metabolic trajectories during human kidney development revealed specific changes in energy and lipid metabolism during kidney development. [81] Single-cell metabolomics uses in situ ionization and fluorescent probe technology to capture individual cell metabolic profiles to reveal cellular heterogeneity. For example, the SCLIMS cross-modal analysis platform was developed to integrate single-cell metabolomics with phenotypic data, which enables the revelation of metabolic heterogeneity in processes like oxidative stress and aging. [82]

In recent years, many algorithms and platforms have been developed in clinical metabolic phenotype analysis, and these tools have become the cornerstone for transforming complex data into clinical insights. MetaboAnalyst, as a one-stop online analysis platform, features advantages such as no programming required and integration of multiple statistical and functional analysis modules. It is often used in biomarker discovery, disease diagnosis model construction, and pathway analysis. [83] Similarly, raw data processing and integration platforms represented by XCMS and AntDAS-Profiler, due to their high efficiency and sensitivity, are often used for preprocessing non-targeted metabolomics data in large-scale cohort studies. [84], [85]The success of these algorithms and platforms is reflected in the fact that they together form a complete closed loop from raw data to clinical interpretation. However, in the future, it is still necessary to build a cloud-based collaboration platform through an automated metabolite annotation pipeline to share data and results and accelerate the discovery and validation of metabolic markers.

5.2. Bioinformatics strategies for metabolic phenotype data

Bioinformatics analysis of metabolic phenotype data requires an integrated strategy, progressing from data processing to biological information analysis and ultimately leading to the uncovering of association between metabolites and physiological/pathological states.(Fig. 3) First, the original data source needs to be converted into a general format for analysis through data conversion and peak extraction. Then, peak detection and alignment are effective ways to identify metabolite information and eliminate metabolic data retention time drift. [86] Secondly, data cleaning and correction is the key to ensuring the intensity of metabolite information, and technical errors are eliminated through denoising, missing value processing and normalization. [87], [88]

Fig. 3.

A schematic diagram of the basic flow of metabolomics analysis. Metabolomics analysis mainly consists of the following steps: sample preparation, data acquisition, and data processing, analysis, and interpretation of biological significance. The figures created by BioRender.

The screening of differential metabolites is often inseparable from statistical analysis, and univariate analysis and multivariate analysis techniques can be selected according to the variable form. Univariate analysis represented by t-test and ANOVA combined with error detection rate correction screened for significant differences. Secondly, multivariate analysis often begins with unsupervised PCA to reveal sample clustering or outliers, followed by supervised PLS-DA/OPLS-DA to construct classification models, and finally employs appropriate thresholds to screen for key metabolites. Metabolite annotation and pathway analysis are often the key to focusing on metabolic pathways. Metabolites are identified by precise mass number (m/z) matching database, MS/MS fragment ion comparison, and finally standard comparison. Finally, enrichment pathways are identified by KEGG, Reactome, etc. to analyze the synergistic changes of metabolite collections in pathways. [89]

5.3. Multi-omics verification facilitates the study of metabolic phenotypes

Although metabolomics can identify differential metabolites, it is difficult to clarify the regulatory mechanisms behind them. Multi-omics integration analysis provides an indispensable upstream biological background, thus becoming the core link connecting metabolic phenotypic phenomena with intrinsic biological mechanisms. The integrated analysis of metabolomics and genomics/transcriptomics is an effective strategy for clarifying the regulatory mechanisms of metabolic pathways. Specifically, when metabolome data suggest that a certain pathway is disrupted, transcriptome data can be used to verify whether the key enzymes regulating this pathway have undergone synchronous changes at the gene expression levels, thereby linking the changes in metabolic phenotypes to specific gene regulatory events. [90]For instance, in the research of Alzheimer's disease, abnormal phenomena of lipid metabolism and glucose metabolism were discovered. Further transcriptomic analysis indicated that these metabolic disorders were directly related to the function of the TREM2 gene in terms of mechanism, thereby revealing a new mechanism of this disease. [91]

As the end point of biochemical reactions, the levels of metabolites are directly regulated by the activity of their upstream catalytic enzymes (proteins). Therefore, the integration of metabolomics and proteomics can not only provide direct functional evidence for metabolite alterations through changes in enzyme abundance or activity, but also jointly construct a complete regulatory network from genes to metabolites. [92], [93] Multi-omics validation significantly enhances the reliability of biomarker or mechanism discovery. For drug development, multi-omics validation helps distinguish the core pathogenic drivers (upstream gene/protein targets) from downstream effector molecules (metabolites), thereby guiding more effective target selection.

6. The application value of metabolic phenotype research

The value of metabolic phenotype research stems from its ability to bridge macroscopic physiological and pathological states with microscopic molecular changes, thereby acting as a chemical bridge between genotype and phenotype. Its applications are expanding beyond basic biomarker discovery to include guiding clinical diagnosis, driving drug discovery, and enabling personalized treatment. Technological advances are poised to make metabolic phenotyping a cornerstone of future precision medicine.

6.1. Exploration of disease mechanisms and disease classification

Metabolic phenotype is a functional index reflecting the metabolic state of organisms, and its clinical transformation needs to solve core issues such as technical standardization, marker verification and multi-dimensional integration. Early diagnosis and screening of diseases is a key link to improving cure rates, reducing mortality and improving patients' quality of life. More studies have shown that small molecule metabolites have great potential as potential markers, such as neurodegenerative diseases, diabetes, cancer and other diagnostics. [94], [95], [96] For example, serum metabolomic analysis of Parkinson's patients and normal people found that serum PC (40:7), PC (40:6p), aspartic acid, etc. were significantly correlated with Parkinson's disease. Based on these findings, a predictive model was developed that achieved an accuracy of over 80 %. [97] Early diagnosis and treatment can significantly improve the quality of life of Parkinson's patients, effectively control symptoms, and delay the progression and prognosis of the disease.

Another important clinical role is disease typing. By classifying clinical features, pathological findings, and other relevant factors, clinicians can more effectively utilize insights into pathogenesis, pathology, and clinical presentation. Biomarkers for disease classification have been successfully identified and are being applied to the complex diseases. [98], [99] The classification of renal cell carcinoma into three subtypes through urine metabolomic data will effectively assist in the early diagnosis and treatment of renal cell carcinoma. [100] The above examples demonstrate the great potential of metabolic typing in exploring disease mechanisms and disease classification, which also highlights the importance of exploring therapeutic regimens targeting characteristic pathological changes to achieve precise intervention.

6.2. Assessment and prognosis of treatment response

In the current therapeutic landscape, drug resistance remains a major challenge. Biomarkers for early detection and monitoring- including nucleic acids (DNA/RNA), proteins, and metabolites- are increasingly being identified and applied in the management of various cancers and neurological diseases [101], [102] Non-coding RNAs can promote communication between tumor cells and microenvironmental cells, serving as potential biomarkers for tumor diagnosis and treatment. For example, miRNA sequencing of plasma exosomes from gastric cancer patients revealed that the miRNA expression profiles were significantly different between patients with different organ metastases. This study identified seven miRNA markers that could serve as biomarkers for predicting organ metastasis in gastric cancer. [103]

Therapeutic strategies guided by metabolic phenotypes are central to precision medicine. Their value lies in assessing metabolic capacity through functional testing, thereby enabling the adjustment of drug types and dosages based on the individual's specific metabolic profile. [104], [105] Clinical needs dynamic evaluation of enzyme activity, combining static genotyping with dynamic functional detection, to achieve personalized medication.

6.3. Metabolic targets in drug development

In modern drug development, ‘metabolic targets’ and ‘metabolic phenotypes’ are deeply intertwined, defining a new paradigm in the era of precision medicine. Metabolic phenotypes serve as maps to discover and validate potential metabolic targets. The ultimate goal of targeting a metabolic pathway with a drug is to reshape abnormal metabolic phenotypes, thereby treating diseases. Metabolic phenotypic characteristics have become the focus of modern drug development. One strategy involves developing drugs that target specific metabolic enzymes to disrupt disease-related pathways. For instance, inhibitors of the IDO1 enzyme, which is involved in tryptophan metabolism, are being investigated as immunotherapies for cancer. [106], [107] Second, s, targeting key metabolic pathways has significant benefits for disease-specific interventions. For example, inhibitors targeting lactate dehydrogenase (LDH) a key enzyme in the Warburg effect of tumors – can alleviate the acidic tumor microenvironment by reducing lactic acid production. [108], [109] Similarly, by inhibiting glutamine breakdown, the drug Telaglenastat cuts off the nitrogen supply to tumor cells, thereby inducing their apoptosis. [110] It has also shown superiority in other metabolic diseases, for example, by targeting the FXR pathway to regulate bile acid metabolism in cholestasis.

The strategy of targeting metabolites and their transporters has opened up new therapeutic avenues. A classic example is the use of xanthine oxidase inhibitors, which lower uric acid levels by blocking the production of this purine metabolism end-product, for the treatment of gout. [111], [112] Viewed as ‘cellular gatekeepers,’ transporters have become important drug targets, shifting the focus from single-target inhibition to dynamic network regulation. [113] For instance, drugs that target solute carrier (SLC) transporters illustrate this approach play crucial roles in nutrient absorption and metabolite balance by facilitating the uptake of various substances. In a more established example, SGLT2, a key protein responsible for glucose reabsorption in the kidney, has become a core therapeutic target for type 2 diabetes. [114], [115] Empagliflozin, a drug targeting SGLT2, acts by competitively binding to the glucose binding site on this transporter. This inhibits glucose reabsorption, leading to urinary glucose excretion and consequently lowering blood glucose levels. [116] To advance towards personalized medicine, future drug development must leverage the powerful synergy between dynamic metabolic phenotyping and multi-omics integration, thereby paving the way for rational pharmacotherapy tailored to the individual.

7. Current challenges and limitations

Metabolic phenotypic research holds great potential in revealing the chemical nature of life processes, but its transformation from the laboratory to clinical practice does indeed face multiple challenges from technical methods, data analysis, clinical translation, and biological heterogeneity.

Research on metabolic phenotypes faces several technical bottlenecks. First, the annotation rate of metabolites is low. The number of spectral peaks detected by analytical instruments far exceeds the number of compounds that can be confidently identified, meaning a significant portion of the data remains unidentified. Second, insufficient resolution of isomeric compounds poses a challenge. It is difficult to distinguish between molecules that share the same molecular formula but have different structures, which can lead to the misidentification of biomarkers. Additionally, LC-MS technology itself has inherent limitations, including batch effects, a limited dynamic range, and matrix interference, all of which compromise data quality and reproducibility. Secondly, extracting core biological information from vast amounts of raw data poses a significant challenge. Metabolism research generates high-dimensional datasets, and processing, integrating, and interpreting this complex data requires advanced professional analytical tools. Moreover, the intrinsic complexity of metabolic networks presents another layer of difficulty. As metabolites represent the endpoints and regulatory nodes of biochemical pathways, a change in a single metabolite's level can stem from alterations in multiple upstream pathways. Consequently, simple correlation analysis is often insufficient to reveal the precise underlying biological mechanism. Thirdly, while the ultimate aim of metabolic phenotype research is the clinical application of its findings, this path is fraught with challenges. A key barrier is the lack of uniform standards governing the entire process from sample handling to data analysis, making it difficult to compare and combine results across research centers. Another formidable barrier is the requirement for large-scale, multi-center prospective studies to prove the robustness needed to meet the high bar set by regulators for diagnostic tool. Finally, the inherent complexity of life systems poses significant challenges to metabolic phenotype research. A key issue is the high background noise generated by the numerous factors influencing an individual's metabolic phenotype, which often obscures the weaker signals emanating from a specific disease. Compounding this problem, the heterogeneity and adaptive nature of diseases thereby give rise to a diverse range of metabolic manifestations, making it difficult to establish consistent biomarkers.

8. Future directions and prospects

Leveraging advances in single-cell sequencing and spatial metabolomics, future research should shift from macroscopic analyses to precise microscopic measurements. Dynamic monitoring and non-invasive detection technologies will be essential for metabolic phenotyping research moving forward. First, multi-omics integration not only allows for the simultaneous analysis of data from the same sample but also helps elucidate the gene-protein-metabolism interaction network under various physiological and pathological conditions. Second, deep learning models should be developed to predict the metabolic outcomes of dietary or drug interventions by leveraging these integrated multi-omics datasets. Third, AI-driven analysis of metabolic information offers significant advantages, including rapid processing capabilities and the generation of highly accurate metabolic insights. Fourth, establishing unified standards for metabolomics analysis will be crucial for minimizing technical variations and ensuring the comparability of results obtained across different laboratories and platforms. Finally, the clinical application of metabolic phenotyping must be advanced to achieve early disease warning, precise classification, and personalized metabolic intervention. As our understanding of individual metabolic characteristics deepens, future treatment and health management strategies will become increasingly personalized.

9. Conclusion

This article aims to highlight the critical importance of metabolic phenotypes in life science and medical research through a systematic exposition of their essence, analytical techniques, and dynamic changes in physiological and pathological processes. Metabolic phenotypes offer a dynamic perspective for understanding health and disease: Health, from this viewpoint, is a finely balanced state (homeostasis) maintained by metabolic networks, whereas disease represents an imbalance stemming from the disruption of this equilibrium. The changes in metabolic phenotypes are the most sensitive "warning signals" in the early stage of diseases. Innovations in research technologies, such as high-throughput metabolomics, spatial metabolomics, and multi-omics integration, enable us to precisely analyze metabolic phenotypes, delving from identifying differential metabolites to exploring metabolic flow, organelle function, and dysregulation of interorgan dialogue, making it a functional framework for explaining disease mechanisms.

In the future, research on metabolic phenotypes will shift from description to precise intervention. Key directions include: (1) achieving early disease warning by constructing risk models combined with AI; (2) performing metabolic typing of complex diseases such as tumors to guide personalized treatment strategies; (3) utilizing metabolic profiles to guide nutritional and pharmaceutical interventions and dynamically monitor therapeutic efficacy; and (4) accelerating drug development by identifying metabolic vulnerability targets. Overall, metabolic phenotypes actively contribute to physiological and pathological processes. A deeper understanding of these phenotypes will drive medicine into a new era of precision health centered on metabolic homeostasis.

Funding

The authors are grateful for the generous support from the Program of Natural Science Foundation of State (Grant No. 81973745, 82104733), Research Fund Project of Heilongjiang University of Traditional Chinese Medicine grant number (201809), Hainan Province ‘Nanhai New Star’ Science and Technology Innovation Talent Platform Project by Hainan Provincial Department of Science and Technology (NHXXRCXM202317), Academic Enhancement Support Program by Hainan Medical University (XSTS2025079), Natural Science Foundation of Heilongjiang Province (YQ2019H030).

CRediT authorship contribution statement

Yang qiang: Writing – original draft, Resources. Shi Qiu: Writing – review & editing, Funding acquisition, Data curation. Aihua Zhang: Writing – review & editing, Funding acquisition, Conceptualization. Ying Cai: Writing – original draft, Data curation, Conceptualization. Yu Guan: Validation, Data curation, Conceptualization. Zhibo Wang: Visualization, Conceptualization. Sifan Guo: Data curation, Conceptualization.

Declaration of Competing Interest

The authors declare that they have no competing interests