Search

Abstract Background Early evidence that patients with (multiple) pre-existing diseases are at highest risk for severe COVID-19 has been instrumental in the pandemic to allocate critical care resources and later vaccination schemes. However, systematic studies exploring the breadth of medical diagnoses are scarce but may help to understand severe COVID-19 among patients at supposedly low risk. Methods We systematically harmonized >12 million primary care and hospitalisation health records from ~500,000 UK Biobank participants into 1448 collated disease terms to systematically identify diseases predisposing to severe COVID-19 (requiring hospitalisation or death) and its post-acute sequalae, Long COVID. Results Here we identify 679 diseases associated with an increased risk for severe COVID-19 (n = 672) and/or Long COVID (n = 72) that span almost all clinical specialties and are strongly enriched in clusters of cardio-respiratory and endocrine-renal diseases. For 57 diseases, we establish consistent evidence to predispose to severe COVID-19 based on survival and genetic susceptibility analyses. This includes a possible role of symptoms of malaise and fatigue as a so far largely overlooked risk factor for severe COVID-19. We finally observe partially opposing risk estimates at known risk loci for severe COVID-19 for etiologically related diseases, such as post-inflammatory pulmonary fibrosis or rheumatoid arthritis, possibly indicating a segregation of disease mechanisms. Conclusions Our results provide a unique reference that demonstrates how 1) complex co-occurrence of multiple – including non-fatal – conditions predispose to increased COVID-19 severity and 2) how incorporating the whole breadth of medical diagnosis can guide the interpretation of genetic risk loci.

Abstract Background Metformin and sodium-glucose-cotransporter-2 inhibitors (SGLT2i) are cornerstone therapies for managing hyperglycemia in diabetes. However, their detailed impacts on metabolic processes, particularly within the citric acid (TCA) cycle and its anaplerotic pathways, remain unclear. This study investigates the tissue-specific metabolic effects of metformin, both as a monotherapy and in combination with SGLT2i, on the TCA cycle and associated anaplerotic reactions in both mice and humans. Methods Metformin-specific metabolic changes were initially identified by comparing metformin-treated diabetic mice (MET) with vehicle-treated db/db mice (VG). These findings were then assessed in two human cohorts (KORA and QBB) and a longitudinal KORA study of metformin-naïve patients with Type 2 Diabetes (T2D). We also compared MET with db/db mice on combination therapy (SGLT2i + MET). Metabolic profiling analyzed 716 metabolites from plasma, liver, and kidney tissues post-treatment, using linear regression and Bonferroni correction for statistical analysis, complemented by pathway analyses to explore the pathophysiological implications. Results Metformin monotherapy significantly upregulated TCA cycle intermediates such as malate, fumarate, and α-ketoglutarate (α-KG) in plasma, and anaplerotic substrates including hepatic glutamate and renal 2-hydroxyglutarate (2-HG) in diabetic mice. Downregulated hepatic taurine was also observed. The addition of SGLT2i, however, reversed these effects, such as downregulating circulating malate and α-KG, and hepatic glutamate and renal 2-HG, but upregulated hepatic taurine. In human T2D patients on metformin therapy, significant systemic alterations in metabolites were observed, including increased malate but decreased citrulline. The bidirectional modulation of TCA cycle intermediates in mice influenced key anaplerotic pathways linked to glutaminolysis, tumorigenesis, immune regulation, and antioxidative responses. Conclusion This study elucidates the specific metabolic consequences of metformin and SGLT2i on the TCA cycle, reflecting potential impacts on the immune system. Metformin shows promise for its anti-inflammatory properties, while the addition of SGLT2i may provide liver protection in conditions like metabolic dysfunction-associated steatotic liver disease (MASLD). These observations underscore the importance of personalized treatment strategies.

Summary: Metabolism oscillates between catabolic and anabolic states depending on food intake, exercise, or stresses that change a multitude of metabolic pathways simultaneously. We present the HuMet Repository for exploring dynamic metabolic responses to oral glucose/lipid loads, mixed meals, 36-h fasting, exercise, and cold stress in healthy subjects. Metabolomics data from blood, urine, and breath of 15 young, healthy men at up to 56 time points are integrated and embedded within an interactive web application, enabling researchers with and without computational expertise to search, visualize, analyze, and contextualize the dynamic metabolite profiles of 2,656 metabolites acquired on multiple platforms. With examples, we demonstrate the utility of the resource for research into the dynamics of human metabolism, highlighting differences and similarities in systemic metabolic responses across challenges and the complementarity of metabolomics platforms. The repository, providing a reference for healthy metabolite changes to six standardized physiological challenges, is freely accessible through a web portal.

Abstract In this study, we aimed to understand the potential role of the gut microbiome in the development of Alzheimer's disease (AD). We took a multi-faceted approach to investigate this relationship. Urine metabolomics were examined in individuals with AD and controls, revealing decreased formate and fumarate concentrations in AD. Additionally, we utilised whole-genome sequencing (WGS) data obtained from a separate group of individuals with AD and controls. This information allowed us to create and investigate host-microbiome personalised whole-body metabolic models. Notably, AD individuals displayed diminished formate microbial secretion in these models. Additionally, we identified specific reactions responsible for the production of formate in the host, and interestingly, these reactions were linked to genes that have correlations with AD. This study suggests formate as a possible early AD marker and highlights genetic and microbiome contributions to its production. The reduced formate secretion and its genetic associations point to a complex connection between gut microbiota and AD. This holistic understanding might pave the way for novel diagnostic and therapeutic avenues in AD management.

Abstract Background B vitamins such as folate (B9), B6, and B12 are key in one carbon metabolism, which generates methyl donors for DNA methylation. Several studies have linked differential methylation to self-reported intakes of folate and B12, but these estimates can be imprecise, while metabolomic biomarkers can offer an objective assessment of dietary intakes. We explored blood metabolomic biomarkers of folate and vitamins B6 and B12, to carry out epigenome-wide analyses across up to three European cohorts. Associations between self-reported habitual daily B vitamin intakes and 756 metabolites (Metabolon Inc.) were assessed in serum samples from 1064 UK participants from the TwinsUK cohort. The identified B vitamin metabolomic biomarkers were then used in epigenome-wide association tests with fasting blood DNA methylation levels at 430,768 sites from the Infinium HumanMethylation450 BeadChip in blood samples from 2182 European participants from the TwinsUK and KORA cohorts. Candidate signals were explored for metabolite associations with gene expression levels in a subset of the TwinsUK sample (n = 297). Metabolomic biomarker epigenetic associations were also compared with epigenetic associations of self-reported habitual B vitamin intakes in samples from 2294 European participants. Results Eighteen metabolites were associated with B vitamin intakes after correction for multiple testing (Bonferroni-adj. p

Complementary multi-omics approach, in silico metabolic flux, and neuroimaging analyses link the sphingolipid pathway to Alzheimer’s disease pathogenesis with therapeutic targets validated using behavioral assessments in amyloidogenic APP/PS1 mice.

Dysregulation of lipid metabolism is associated with coronary artery disease (CAD). Here, the authors perform GWAS of the serum lipidome to identify variants associated with lipid species that are putatively in the mechanistic pathway to CAD.

Abstract Background Periodontitis is among the most common chronic diseases worldwide, and it is one of the main reasons for tooth loss. Comprehensive profiling of the metabolite content of the saliva can enable the identification of novel pathways associated with periodontitis and highlight non-invasive markers to facilitate time and cost-effective screening efforts for the presence of periodontitis and the prediction of tooth loss. Methods We first investigated cross-sectional associations of 13 oral health variables with saliva levels of 562 metabolites, measured by untargeted mass spectrometry among a sub-sample (n = 938) of the Study of Health in Pomerania (SHIP-2) using linear regression models adjusting for common confounders. We took forward any candidate metabolite associated with at least two oral variables, to test for an association with a 5-year tooth loss over and above baseline oral health status using negative binomial regression models. Results We identified 84 saliva metabolites that were associated with at least one oral variable cross-sectionally, for a subset of which we observed robust replication in an independent study. Out of 34 metabolites associated with more than two oral variables, baseline saliva levels of nine metabolites were positively associated with a 5-year tooth loss. Across all analyses, the metabolites 2-pyrrolidineacetic acid and butyrylputrescine were the most consistent candidate metabolites, likely reflecting oral dysbiosis. Other candidate metabolites likely reflected tissue destruction and cell proliferation. Conclusions Untargeted metabolic profiling of saliva replicated metabolic signatures of periodontal status and revealed novel metabolites associated with periodontitis and future tooth loss.

Food intake triggers extensive changes in the blood metabolome. The kinetics of these changes depend on meal composition and on intrinsic, health-related characteristics of each individual, making the assessment of changes in the postprandial metabolome an opportunity to assess someone's metabolic status. To enable the usage of dietary challenges as diagnostic tools, profound knowledge about changes that occur in the postprandial period in healthy individuals is needed. In this study, we characterize the time-resolved changes in plasma levels of 634 metabolites in response to an oral glucose tolerance test (OGTT), an oral lipid tolerance test (OLTT), and a mixed meal (SLD) in healthy young males (n = 15). Metabolite levels for samples taken at different time points (20 per individual) during the challenges were available from targeted (132 metabolites) and non-targeted (502 metabolites) metabolomics. Almost half of the profiled metabolites (n = 308) showed a significant change in at least one challenge, thereof 111 metabolites responded exclusively to one particular challenge. Examples include azelate, which is linked to ω-oxidation and increased only in OLTT, and a fibrinogen cleavage peptide that has been linked to a higher risk of cardiovascular events in diabetes patients and increased only in OGTT, making its postprandial dynamics a potential target for risk management. A pool of 89 metabolites changed their plasma levels during all three challenges and represents the core postprandial response to food intake regardless of macronutrient composition. We used fuzzy c-means clustering to group these metabolites into eight clusters based on commonalities of their dynamic response patterns, with each cluster following one of four primary response patterns: (i) “decrease-increase” (valley-like) with fatty acids and acylcarnitines indicating the suppression of lipolysis, (ii) “increase-decrease” (mountain-like) including a cluster of conjugated bile acids and the glucose/insulin cluster, (iii) “steady decrease” with metabolites reflecting a carryover from meals prior to the study, and (iv) “mixed” decreasing after the glucose challenge and increasing otherwise. Despite the small number of subjects, the diversity of the challenges and the wealth of metabolomic data make this study an important step toward the characterization of postprandial responses and the identification of markers of metabolic processes regulated by food intake.

In recent years, bile acids (BA) have received great interest due to their pleiotropic biological activity and the presence of plasma membrane-bound and nuclear receptors. Moreover, BA in blood have been identified by metabolite screening approaches as biomarkers that are associated with various diseases and even with a human longevity phenotype. With the growing interest in the microbiota contribution to the health-disease trajectory, BA that undergo deconjugation and other modifications by bacteria in the large intestine have become a prime target as a microbiome diversity modifier. We here profiled BA by a quantitative and a semiquantitative approach in 15 healthy and phenotypically very similar young individuals for over a 36-h fasting period, an oral glucose tolerance test (OGTT), and an oral lipid tolerance test (OLTT). We demonstrate a remarkable heterogeneity of the responses and describe the different dynamics of the plasma changes that likely originate from different routes by which BA enters the peripheral blood, and that may represent a direct secretion from the liver into the blood and a route that reaches the blood as a spill-over after passing from the gallbladder through the intestine and the portal system. We discuss the finding that an individual transport process involved in the passage of BA could be a critical determinant in the kinetics of plasma appearance and the overall phenotypic variability found.

Finding effective treatments for COVID-19 depends upon understanding genetic regulation of proteins involved in SARS-CoV-2 infection and host response. Here, the authors identify genetic variants linked to expression of such proteins, data which could lead to the discovery of therapeutic targets.

Correlation network inference is typically based on the significance of the correlation coefficients, but this procedure is not guaranteed to capture biological mechanisms. Here, the authors develop a cutoff selection algorithm that maximizes the overlap between inferred networks and prior knowledge.

Sex and the APOE ε4 genotype are important risk factors for late-onset Alzheimer’s disease. In the current study, the authors investigate how sex and APOE ε4 genotype modify the association between Alzheimer’s disease biomarkers and metabolites in serum.

Objective: Technological advances have brought a steady increase in the availability of various types of omics data, from genomics to metabolomics. Integrating these multi-omics data is a chance and challenge for systems biology; yet, tools to fully tap their potential remain scarce. Methods: We present here a fully unsupervised and versatile correlation-based method – termed Correlation guided Network Integration (CoNI) – to integrate multi-omics data into a hypergraph structure that allows for the identification of effective modulators of metabolism. Our approach yields single transcripts of potential relevance that map to specific, densely connected, metabolic subgraphs or pathways. Results: By applying our method on transcriptomics and metabolomics data from murine livers under standard Chow or high-fat diet, we identified eleven genes with potential regulatory effects on hepatic metabolism. Five candidates, including the hepatokine INHBE, were validated in human liver biopsies to correlate with diabetes-related traits such as overweight, hepatic fat content, and insulin resistance (HOMA-IR). Conclusion: Our method's successful application to an independent omics dataset confirmed that the novel CoNI framework is a transferable, entirely data-driven, flexible, and versatile tool for multiple omics data integration and interpretation.

Biomarkers that predict mortality are of interest for clinical as well as research applications. Here, the authors analyze metabolomics data from 44,168 individuals and identify key metabolites independently associated with all-cause mortality risk.

Resistance training promotes metabolic health and stimulates muscle hypertrophy, but the precise routes by which resistance exercise (RE) conveys these health benefits are largely unknown. Aim: To investigate how acute RE affects human skeletal muscle metabolism. Methods: We collected vastus lateralis biopsies from six healthy male untrained volunteers at rest, before the first of 13 RE training sessions, and 45 min after the first and last bouts of RE. Biopsies were analysed using untargeted mass spectrometry-based metabolomics. Results: We measured 617 metabolites covering a broad range of metabolic pathways. In the untrained state RE altered 33 metabolites, including increased 3-methylhistidine and N-lactoylvaline, suggesting increased protein breakdown, as well as metabolites linked to ATP (xanthosine) and NAD (N1-methyl-2-pyridone-5-carboxamide) metabolism; the bile acid chenodeoxycholate also increased in response to RE in muscle opposing previous findings in blood. Resistance training led to muscle hypertrophy, with slow type I and fast/intermediate type II muscle fibre diameter increasing by 10.7% and 10.4%, respectively. Comparison of post-exercise metabolite levels between trained and untrained state revealed alterations of 46 metabolites, including decreased N-acetylated ketogenic amino acids and increased beta-citrylglutamate which might support growth. Only five of the metabolites that changed after acute exercise in the untrained state were altered after chronic training, indicating that training induces multiple metabolic changes not directly related to the acute exercise response. Conclusion: The human skeletal muscle metabolome is sensitive towards acute RE in the trained and untrained states and reflects a broad range of adaptive processes in response to repeated stimulation.

Genome-wide association studies (GWAS) with non-targeted metabolomics have identified many genetic loci of biomedical interest. However, metabolites with a high degree of missingness, such as drug metabolites and xenobiotics, are often excluded from such studies due to a lack of statistical power and higher uncertainty in their quantification. Here we propose ratios between related drug metabolites as GWAS phenotypes that can drastically increase power to detect genetic associations between pairs of biochemically related molecules. As a proof-of-concept we conducted a GWAS with 520 individuals from the Qatar Biobank for who at least five of the nine available acetaminophen metabolites have been detected. We identified compelling evidence for genetic variance in acetaminophen glucuronidation and methylation by UGT2A15 and COMT, respectively. Based on the metabolite ratio association profiles of these two loci we hypothesized the chemical structure of one of their products or substrates as being 3-methoxyacetaminophen, which we then confirmed experimentally. Taken together, our study suggests a novel approach to analyze metabolites with a high degree of missingness in a GWAS setting with ratios, and it also demonstrates how pharmacological pathways can be mapped out using non-targeted metabolomics measurements in large population-based studies.

Abstract Background Most research into myocardial infarctions (MIs) have focused on preventative efforts. For survivors, the occurrence of an MI represents a major clinical event that can have long-lasting consequences. There has been little to no research into the molecular changes that can occur as a result of an incident MI. Here, we use three cohorts to identify epigenetic changes that are indicative of an incident MI and their association with gene expression and metabolomics. Results Using paired samples from the KORA cohort, we screened for DNA methylation loci (CpGs) whose change in methylation is potentially indicative of the occurrence of an incident MI between the baseline and follow-up exams. We used paired samples from the NAS cohort to identify 11 CpGs which were predictive in an independent cohort. After removing two CpGs associated with medication usage, we were left with an “epigenetic fingerprint” of MI composed of nine CpGs. We tested this fingerprint in the InCHIANTI cohort where it moderately discriminated incident MI occurrence (AUC = 0.61, P = 6.5 × 10−3). Returning to KORA, we associated the epigenetic fingerprint loci with cis-gene expression and integrated it into a gene expression-metabolomic network, which revealed links between the epigenetic fingerprint CpGs and branched chain amino acid (BCAA) metabolism. Conclusions There are significant changes in DNA methylation after an incident MI. Nine of these CpGs show consistent changes in multiple cohorts, significantly discriminate MI in independent cohorts, and were independent of medication usage. Integration with gene expression and metabolomics data indicates a link between MI-associated epigenetic changes and BCAA metabolism.

Abstract Background The N-terminal prohormone of brain natriuretic peptide (NT-proBNP) is an important biomarker for the diagnosis of heart failure. Apart from this and only recently recognized, NT-proBNP levels associate with higher HDL- and lower LDL-cholesterol levels comprising a favorable blood lipid profile. To further examine this observation, the lipoprotein profile in relation to NT-proBNP was examined in-depth by proton nuclear magnetic resonance spectroscopy (1H-NMR). We complemented this investigation with a state-of-the-art untargeted metabolomics approach. Methods Lipoprotein particles were determined by 1H-NMR spectroscopy in 872 subjects without self-reported diabetes from the population-based Study of Health in Pomerania (SHIP)-TREND with available NT-proBNP measurements. Comprehensive metabolomics data for plasma and urine samples were obtained. Linear regression models were performed to assess the associations between serum concentrations of NT-proBNP and the metabolites/lipoprotein particles measured in plasma or urine. Results An increase in serum NT-proBNP was associated with a benefical lipoprotein profile, including a decrease in VLDL, IDL and LDL-particles along with an increase in large HDL particles. These findings were replicated in a second independent cohort. Serum concentrations of NT-proBNP showed significant inverse associations with seven plasma metabolites while associations with 39 urinary metabolites, mostly comprising amino acids and related intermediates, were identified. Mediation analyses revealed adiponection as mediating factor for the associations observed with lipoproteins particles. Conclusions Most of the metabolic changes associated with NT-proBNP implicate significant influence on the blood lipid profile besides vasodilatory and the diuretic action of BNP signaling. Our data suggest that the more favorable lipoprotein profile as associated with elevated NT-proBNP concentrations in mainly cardiac healthy individuals might relate to adiponectin signaling indicating even indirect cardio-protective effects for NT-proBNP.

Abstract Background Inflammation occurs as an immediate protective response of the immune system to a harmful stimulus, whether locally confined or systemic. In contrast, a persisting, i.e., chronic, inflammatory state, even at a low-grade, is a well-known risk factor in the development of common diseases like diabetes or atherosclerosis. In clinical practice, laboratory markers like high-sensitivity C-reactive protein (hsCRP), white blood cell count (WBC), and fibrinogen, are used to reveal inflammatory processes. In order to gain a deeper insight regarding inflammation-related changes in metabolism, the present study assessed the metabolic patterns associated with alterations in inflammatory markers. Methods Based on mass spectrometry and nuclear magnetic resonance spectroscopy we determined a comprehensive panel of 613 plasma and 587 urine metabolites among 925 apparently healthy individuals. Associations between inflammatory markers, namely hsCRP, WBC, and fibrinogen, and metabolite levels were tested by linear regression analyses controlling for common confounders. Additionally, we tested for a discriminative signature of an advanced inflammatory state using random forest analysis. Results HsCRP, WBC, and fibrinogen were significantly associated with 71, 20, and 19 plasma and 22, 3, and 16 urine metabolites, respectively. Identified metabolites were related to the bradykinin system, involved in oxidative stress (e.g., glutamine or pipecolate) or linked to the urea cycle (e.g., ornithine or citrulline). In particular, urine 3’-sialyllactose was found as a novel metabolite related to inflammation. Prediction of an advanced inflammatory state based solely on 10 metabolites was well feasible (median AUC: 0.83). Conclusions Comprehensive metabolic profiling confirmed the far-reaching impact of inflammatory processes on human metabolism. The identified metabolites included not only those already described as immune-modulatory but also completely novel patterns. Moreover, the observed alterations provide molecular links to inflammation-associated diseases like diabetes or cardiovascular disorders.

Objective: Obesity-associated WAT inflammation is characterized by the accumulation and local activation of macrophages (MΦs), and recent data from mouse studies suggest that macrophages are modifiers of adipocyte energy metabolism and mitochondrial function. As mitochondrial dysfunction has been associated with obesity and the metabolic syndrome in humans, herein we aimed to delineate how human macrophages may affect energy metabolism of white adipocytes. Methods: Human adipose tissue gene expression analysis for markers of macrophage activation and tissue inflammation (CD11c, CD40, CD163, CD206, CD80, MCP1, TNFα) in relationship to mitochondrial complex I (NDUFB8) and complex III (UQCRC2) was performed on subcutaneous WAT of 24 women (BMI 20–61 kg/m2). Guided by these results, the impact of secreted factors of LPS/IFNγ- and IL10/TGFβ-activated human macrophages (THP1, primary blood-derived) on mitochondrial function in human subcutaneous white adipocytes (SGBS, primary) was determined by extracellular flux analysis (Seahorse technology) and gene/protein expression. Results: Stepwise regression analysis of human WAT gene expression data revealed that a linear combination of CD40 and CD163 was the strongest predictor for mitochondrial complex I (NDUFB8) and complex III (UQCRC2) levels, independent of BMI. IL10/TGFβ-activated MΦs displayed high CD163 and low CD40 expression and secreted factors that decreased UQCRC2 gene/protein expression and ATP-linked respiration in human white adipocytes. In contrast, LPS/IFNγ-activated MΦs showed high CD40 and low CD163 expression and secreted factors that enhanced adipocyte mitochondrial activity resulting in a total difference of 37% in ATP-linked respiration of white adipocytes (p = 0.0024) when comparing the effect of LPS/IFNγ- vs IL10/TGFβ-activated MΦs. Conclusion: Our data demonstrate that macrophages modulate human adipocyte energy metabolism via an activation-dependent paracrine mechanism.

Abstract Background Genome-wide association studies allow us to understand the genetics of complex diseases. Human metabolism provides information about the disease-causing mechanisms, so it is usual to investigate the associations between genetic variants and metabolite levels. However, only considering genetic variants and their effects on one trait ignores the possible interplay between different “omics” layers. Existing tools only consider single-nucleotide polymorphism (SNP)–SNP interactions, and no practical tool is available for large-scale investigations of the interactions between pairs of arbitrary quantitative variables. Results We developed an R package called pulver to compute p-values for the interaction term in a very large number of linear regression models. Comparisons based on simulated data showed that pulver is much faster than the existing tools. This is achieved by using the correlation coefficient to test the null-hypothesis, which avoids the costly computation of inversions. Additional tricks are a rearrangement of the order, when iterating through the different “omics” layers, and implementing this algorithm in the fast programming language C++. Furthermore, we applied our algorithm to data from the German KORA study to investigate a real-world problem involving the interplay among DNA methylation, genetic variants, and metabolite levels. Conclusions The pulver package is a convenient and rapid tool for screening huge numbers of linear regression models for significant interaction terms in arbitrary pairs of quantitative variables. pulver is written in R and C++, and can be downloaded freely from CRAN at https://cran.r-project.org/web/packages/pulver/ .

Genome sequencing alone fails to provide a genetic diagnosis for many Mendelian disorder patients. Here, the authors utilize RNA sequencing to complement genotyping of patients with a rare mitochondrial disease by detecting aberrant RNA expression, splicing and allele-specific expression.

Abstract The role of androgens in metabolism with respect to sex-specific disease associations is poorly understood. Therefore, we aimed to provide molecular signatures in plasma and urine of androgen action in a sex-specific manner using state-of-the-art metabolomics techniques. Our study population consisted of 430 men and 343 women, aged 20–80 years, who were recruited for the cross-sectional population-based Study of Health in Pomerania (SHIP-TREND), Germany. We used linear regression models to identify associations between testosterone, androstenedione and dehydroepiandrosterone-sulfate (DHEAS) as well as sex hormone-binding globulin and plasma or urine metabolites measured by mass spectrometry. The analyses revealed major sex-specific differences in androgen-associated metabolites, particularly for levels of urate, lipids and metabolic surrogates of lifestyle factors, like cotinine or piperine. In women, in particular in the postmenopausal state, androgens showed a greater impact on the metabolome than in men (especially DHEAS and lipids were highly related in women). We observed a novel association of androstenedione on the metabolism of biogenic amines and only a small sex-overlap of associations within steroid metabolism. The present study yields new insights in the interaction between androgens and metabolism, especially about their implication in female metabolism.

Individual genetic variation can affect the levels of protein in blood, but detailed data sets linking these two types of data are rare. Here, the authors carry out a genome-wide association study of levels of over a thousand different proteins, and describe many new SNP-protein interactions.

BACKGROUND:Genome-wide association studies of common diseases or metabolite quantitative traits often identify common variants of small effect size, which may contribute to phenotypes by modulation of gene expression. Thus, there is growing demand for cellular models enabling to assess the impact of gene regulatory variants with moderate effects on gene expression. Mitochondrial fatty acid oxidation is an important energy metabolism pathway. Common noncoding acyl-CoA dehydrogenase short chain (ACADS) gene variants are associated with plasma C4-acylcarnitine levels and allele-specific modulation of ACADS expression may contribute to the observed phenotype. METHODS AND FINDINGS:We assessed ACADS expression and intracellular acylcarnitine levels in human lymphoblastoid cell lines (LCL) genotyped for a common ACADS variant associated with plasma C4-acylcarnitine and found a significant genotype-dependent decrease of ACADS mRNA and protein. Next, we modelled gradual decrease of ACADS expression using a tetracycline-regulated shRNA-knockdown of ACADS in Huh7 hepatocytes, a cell line with high fatty acid oxidation-(FAO)-capacity. Assessing acylcarnitine flux in both models, we found increased C4-acylcarnitine levels with decreased ACADS expression levels. Moreover, assessing time-dependent changes of acylcarnitine levels in shRNA-hepatocytes with altered ACADS expression levels revealed an unexpected effect on long- and medium-chain fatty acid intermediates. CONCLUSIONS:Both, genotyped LCL and regulated shRNA-knockdown are valuable tools to model moderate, gradual gene-regulatory effects of common variants on cellular phenotypes. Decreasing ACADS expression levels modulate short and surprisingly also long/medium chain acylcarnitines, and may contribute to increased plasma acylcarnitine levels.

Biological exploration of early biomarkers for chronic kidney disease (CKD) in (pre)diabetic individuals is crucial for personalized management of diabetes. Here, we evaluated two candidate biomarkers of incident CKD (sphingomyelin (SM) C18:1 and phosphatidylcholine diacyl (PC aa) C38:0) concerning kidney function in hyperglycemic participants of the Cooperative Health Research in the Region of Augsburg (KORA) cohort, and in two biofluids and six organs of leptin receptor-deficient (db/db) mice and wild type controls. Higher serum concentrations of SM C18:1 and PC aa C38:0 in hyperglycemic individuals were found to be associated with lower estimated glomerular filtration rate (eGFR) and higher odds of CKD. In db/db mice, both metabolites had a significantly lower concentration in urine and adipose tissue, but higher in the lungs. Additionally, db/db mice had significantly higher SM C18:1 levels in plasma and liver, and PC aa C38:0 in adrenal glands. This cross-sectional human study confirms that SM C18:1 and PC aa C38:0 associate with kidney dysfunction in pre(diabetic) individuals, and the animal study suggests a potential implication of liver, lungs, adrenal glands, and visceral fat in their systemic regulation. Our results support further validation of the two phospholipids as early biomarkers of renal disease in patients with (pre)diabetes.

A Correction to this paper has been published: https://doi.org/10.1038/s41467-021-21370-6

Shared metabolomic patterns at delivery have been suggested to underlie the mother-to-child transmission of adverse metabolic health. This study aimed to investigate whether mothers with gestational diabetes mellitus (GDM) and their offspring show similar metabolomic patterns several years postpartum. Targeted metabolomics (including 137 metabolites) was performed in plasma samples obtained during an oral glucose tolerance test from 48 mothers with GDM and their offspring at a cross-sectional study visit 8 years after delivery. Partial Pearson’s correlations between the area under the curve (AUC) of maternal and offspring metabolites were calculated, yielding so-called Gaussian graphical models. Spearman’s correlations were applied to investigate correlations of body mass index (BMI), Matsuda insulin sensitivity index (ISI-M), dietary intake, and physical activity between generations, and correlations of metabolite AUCs with lifestyle variables. This study revealed that BMI, ISI-M, and the AUC of six metabolites (carnitine, taurine, proline, SM(-OH) C14:1, creatinine, and PC ae C34:3) were significantly correlated between mothers and offspring several years postpartum. Intergenerational metabolite correlations were independent of shared BMI, ISI-M, age, sex, and all other metabolites. Furthermore, creatinine was correlated with physical activity in mothers. This study suggests that there is long-term metabolic programming in the offspring of mothers with GDM and informs us about targets that could be addressed by future intervention studies.

Metabolomics studies have seen a steady growth due to the development and implementation of affordable and high-quality metabolomics platforms. In large metabolite panels, measurement values are frequently missing and, if neglected or sub-optimally imputed, can cause biased study results. We provided a publicly available, user-friendly R script to streamline the imputation of missing endogenous, unannotated, and xenobiotic metabolites. We evaluated the multivariate imputation by chained equations (MICE) and k-nearest neighbors (kNN) analyses implemented in our script by simulations using measured metabolites data from the Netherlands Epidemiology of Obesity (NEO) study (n = 599). We simulated missing values in four unique metabolites from different pathways with different correlation structures in three sample sizes (599, 150, 50) with three missing percentages (15%, 30%, 60%), and using two missing mechanisms (completely at random and not at random). Based on the simulations, we found that for MICE, larger sample size was the primary factor decreasing bias and error. For kNN, the primary factor reducing bias and error was the metabolite correlation with its predictor metabolites. MICE provided consistently higher performance measures particularly for larger datasets (n > 50). In conclusion, we presented an imputation workflow in a publicly available R script to impute untargeted metabolomics data. Our simulations provided insight into the effects of sample size, percentage missing, and correlation structure on the accuracy of the two imputation methods.

Metabolomics can be a tool to identify dietary biomarkers. However, reported food-metabolite associations have been inconsistent, and there is a need to explore further associations. Our aims were to confirm previously reported food-metabolite associations and to identify novel food-metabolite associations. We conducted a cross-sectional analysis of data from 849 participants (57% men) of the PopGen cohort. Dietary intake was obtained using FFQ and serum metabolites were profiled by an untargeted metabolomics approach. We conducted a systematic literature search to identify previously reported food-metabolite associations and analyzed these associations using linear regression. To identify potential novel food-metabolite associations, datasets were split into training and test datasets and linear regression models were fitted to the training datasets. Significant food-metabolite associations were evaluated in the test datasets. Models were adjusted for covariates. In the literature, we identified 82 food-metabolite associations. Of these, 44 associations were testable in our data and confirmed associations of coffee with 12 metabolites, of fish with five, of chocolate with two, of alcohol with four, and of butter, poultry and wine with one metabolite each. We did not identify novel food-metabolite associations; however, some associations were sex-specific. Potential use of some metabolites as biomarkers should consider sex differences in metabolism.

A metabolome-wide genome-wide association study (mGWAS) aims to discover the effects of genetic variants on metabolome phenotypes. Most mGWASes use as phenotypes concentrations of limited sets of metabolites that can be identified and quantified from spectral information. In contrast, in an untargeted mGWAS both identification and quantification are forgone and, instead, all measured metabolome features are tested for association with genetic variants. While the untargeted approach does not discard data that may have eluded identification, the interpretation of associated features remains a challenge. To address this issue, we developed metabomatching to identify the metabolites underlying significant associations observed in untargeted mGWASes on proton NMR metabolome data. Metabomatching capitalizes on genetic spiking, the concept that because metabolome features associated with a genetic variant tend to correspond to the peaks of the NMR spectrum of the underlying metabolite, genetic association can allow for identification. Applied to the untargeted mGWASes in the SHIP and CoLaus cohorts and using 180 reference NMR spectra of the urine metabolome database, metabomatching successfully identified the underlying metabolite in 14 of 19, and 8 of 9 associations, respectively. The accuracy and efficiency of our method make it a strong contender for facilitating or complementing metabolomics analyses in large cohorts, where the availability of genetic, or other data, enables our approach, but targeted quantification is limited.

Recent research suggested a metabolic implication of osteocalcin (OCN) in e.g. insulin sensitivity or steroid production. We used an untargeted metabolomics approach by analyzing plasma and urine samples of 931 participants using mass spectrometry to reveal further metabolic actions of OCN. Several detected relations between OCN and metabolites were strongly linked to renal function, however, a number of associations remained significant after adjustment for renal function. Intermediates of proline catabolism were associated with OCN reflecting the implication in bone metabolism. The association to kynurenine points towards a pro-inflammatory state with increasing OCN. Inverse relations with intermediates of branch-chained amino acid metabolism suggest a link to energy metabolism. Finally, urinary surrogate markers of smoking highlight its adverse effect on OCN metabolism. In conclusion, the present study provides a read-out of metabolic actions of OCN. However, most of the associations were weak arguing for a limited role of OCN in whole-body metabolism.

Kit-based assays, such as AbsoluteIDQTM p150, are widely used in large cohort studies and provide a standardized method to quantify blood concentrations of phosphatidylcholines (PCs). Many disease-relevant associations of PCs were reported using this method. However, their interpretation is hampered by lack of functionally-relevant information on the detailed fatty acid side-chain compositions as only the total number of carbon atoms and double bonds is identified by the kit. To enable more substantiated interpretations, we characterized these PC sums using the side-chain resolving LipidyzerTM platform, analyzing 223 samples in parallel to the AbsoluteIDQTM. Combining these datasets, we estimated the quantitative composition of PC sums and subsequently tested their replication in an independent cohort. We identified major constituents of 28 PC sums, revealing also various unexpected compositions. As an example, PC 16:0_22:5 accounted for more than 50% of the PC sum with in total 38 carbon atoms and 5 double bonds (PC aa 38:5). For 13 PC sums, we found relatively high abundances of odd-chain fatty acids. In conclusion, our study provides insights in PC compositions in human plasma, facilitating interpretation of existing epidemiological data sets and potentially enabling imputation of PC compositions for future meta-analyses of lipidomics data.

Determination of metabolomic signatures of pulmonary function and chronic obstructive pulmonary disease (COPD) in the general population could aid in identification and understanding of early disease processes. Metabolome measurements were performed on serum from 4742 individuals (2354 African-Americans and 1529 European-Americans from the Atherosclerosis Risk in Communities study and 859 Europeans from the Cooperative Health Research in the Region of Augsburg study). We examined 368 metabolites in relation to cross-sectional measures of forced expiratory volume in 1 s (FEV1), forced vital capacity (FVC), their ratio (FEV1/FVC) and COPD using multivariable regression followed by meta-analysis. At a false discovery rate of 0.05, 95 metabolites were associated with FEV1 and 100 with FVC (73 overlapping), including inverse associations with branched-chain amino acids and positive associations with glutamine. Ten metabolites were associated with FEV1/FVC and seventeen with COPD (393 cases). Enriched pathways of amino acid metabolism were identified. Associations with FEV1 and FVC were not driven by individuals with COPD. We identified novel metabolic signatures of pulmonary function and COPD in African and European ancestry populations. These may allow development of biomarkers in the general population of early disease pathogenesis, before pulmonary function has decreased to levels diagnostic for COPD.

Iron deficiency is the most frequent deficiency disease and parameters of iron metabolism appear to be linked to major metabolic and cardiovascular diseases. We screened a large set of small molecules in plasma for associations with iron status among apparently healthy subjects to elucidate subclinical profiles which may provide a link between iron status and onset of diseases. Based on mass spectrometry and nuclear magnetic resonance spectroscopy we determined 613 plasma metabolites and lipoprotein subfractions among 820 apparently healthy individuals. Associations between ferritin, transferrin, haemoglobin and myoglobin and metabolite levels were tested by sex-specific linear regression analyses controlling for common confounders. Far more significant associations in women (82 out of 102) compared to men became obvious. The majority of the metabolites associated with serum ferritin and haemoglobin in women comprising fatty acid species, branched-chain amino acid catabolites and catabolites of heme. The latter was also obvious among men. Positive associations between serum transferrin and VLDL and IDL particle measures seen in women were observed in men with respect to serum ferritin. We observed a sexual-dimorphic fingerprint of surrogates of iron metabolism which may provide a link for the associations between those parameters and major metabolic and cardiovascular disease.

Metabolomic profiling is a powerful approach to characterize human metabolism and help understand common disease risk. Although multiple high-throughput technologies have been developed to assay the human metabolome, no technique is capable of capturing the entire human metabolism. Large-scale metabolomics data are being generated in multiple cohorts, but the datasets are typically profiled using different metabolomics platforms. Here, we compared analyses across two of the most frequently used metabolomic platforms, Biocrates and Metabolon, with the aim of assessing how complimentary metabolite profiles are across platforms. We profiled serum samples from 1,001 twins using both targeted (Biocrates, n = 160 metabolites) and non-targeted (Metabolon, n = 488 metabolites) mass spectrometry platforms. We compared metabolite distributions and performed genome-wide association analyses to identify shared genetic influences on metabolites across platforms. Comparison of 43 metabolites named for the same compound on both platforms indicated strong positive correlations, with few exceptions. Genome-wide association scans with high-throughput metabolic profiles were performed for each dataset and identified genetic variants at 7 loci associated with 16 unique metabolites on both platforms. The 16 metabolites showed consistent genetic associations and appear to be robustly measured across platforms. These included both metabolites named for the same compound across platforms as well as unique metabolites, of which 2 (nonanoylcarnitine (C9) [Biocrates]/Unknown metabolite X-13431 [Metabolon] and PC aa C28:1 [Biocrates]/1-stearoylglycerol [Metabolon]) are likely to represent the same or related biochemical entities. The results demonstrate the complementary nature of both platforms, and can be informative for future studies of comparative and integrative metabolomics analyses in samples profiled on different platforms.

Serum metabolite profiling can be used to identify pathways involved in the pathogenesis of and potential biomarkers for a given disease. Both restless legs syndrome (RLS) and Parkinson`s disease (PD) represent movement disorders for which currently no blood-based biomarkers are available and whose pathogenesis has not been uncovered conclusively. We performed unbiased serum metabolite profiling in search of signature metabolic changes for both diseases.456 metabolites were quantified in serum samples of 1272 general population controls belonging to the KORA cohort, 82 PD cases and 95 RLS cases by liquid-phase chromatography and gas chromatography separation coupled with tandem mass spectrometry. Genetically determined metabotypes were calculated using genome-wide genotyping data for the 1272 general population controls.After stringent quality control, we identified decreased levels of long-chain (polyunsaturated) fatty acids of individuals with PD compared to both RLS (PD vs. RLS: p = 0.0001 to 5.80x10-9) and general population controls (PD vs. KORA: p = 6.09x10-5 to 3.45x10-32). In RLS, inositol metabolites were increased specifically (RLS vs. KORA: p = 1.35x10-6 to 3.96x10-7). The impact of dopaminergic drugs was reflected in changes in the phenylalanine/tyrosine/dopamine metabolism observed in both individuals with RLS and PD.A first discovery approach using serum metabolite profiling in two dopamine-related movement disorders compared to a large general population sample identified significant alterations in the polyunsaturated fatty acid metabolism in PD and implicated the inositol metabolism in RLS. These results provide a starting point for further studies investigating new perspectives on factors involved in the pathogenesis of the two diseases as well as possible points of therapeutic intervention.

Angiotensin-I-converting enzyme (ACE) inhibitors are an important class of antihypertensives whose action on the human organism is still not fully understood. Although it is known that ACE especially cleaves COOH-terminal dipeptides from active polypeptides, the whole range of substrates and products is still unknown. When analyzing the action of ACE inhibitors, effects of genetic variation on metabolism need to be considered since genetic variance in the ACE gene locus was found to be associated with ACE-concentration in blood as well as with changes in the metabolic profiles of a general population. To investigate the interactions between genetic variance at the ACE-locus and the influence of ACE-therapy on the metabolic status we analyzed 517 metabolites in 1,361 participants from the KORA F4 study. We replicated our results in 1,964 individuals from TwinsUK. We observed differences in the concentration of five dipeptides and three ratios of di- and oligopeptides between ACE inhibitor users and non-users that were genotype dependent. Such changes in the concentration affected major homozygotes, and to a lesser extent heterozygotes, while minor homozygotes showed no or only small changes in the metabolite status. Two of these resulting dipeptides, namely aspartylphenylalanine and phenylalanylserine, showed significant associations with blood pressure which qualifies them-and perhaps also the other dipeptides-as readouts of ACE-activity. Since so far ACE activity measurement is substrate specific due to the usage of only one oligopeptide, taking several dipeptides as potential products of ACE into account may provide a broader picture of the ACE activity.

Genome-wide association studies with metabolic traits (mGWAS) uncovered many genetic variants that influence human metabolism. These genetically influenced metabotypes (GIMs) contribute to our metabolic individuality, our capacity to respond to environmental challenges, and our susceptibility to specific diseases. While metabolic homeostasis in blood is a well investigated topic in large mGWAS with over 150 known loci, metabolic detoxification through urinary excretion has only been addressed by few small mGWAS with only 11 associated loci so far. Here we report the largest mGWAS to date, combining targeted and non-targeted 1H NMR analysis of urine samples from 3,861 participants of the SHIP-0 cohort and 1,691 subjects of the KORA F4 cohort. We identified and replicated 22 loci with significant associations with urinary traits, 15 of which are new (HIBCH, CPS1, AGXT, XYLB, TKT, ETNPPL, SLC6A19, DMGDH, SLC36A2, GLDC, SLC6A13, ACSM3, SLC5A11, PNMT, SLC13A3). Two-thirds of the urinary loci also have a metabolite association in blood. For all but one of the 6 loci where significant associations target the same metabolite in blood and urine, the genetic effects have the same direction in both fluids. In contrast, for the SLC5A11 locus, we found increased levels of myo-inositol in urine whereas mGWAS in blood reported decreased levels for the same genetic variant. This might indicate less effective re-absorption of myo-inositol in the kidneys of carriers. In summary, our study more than doubles the number of known loci that influence urinary phenotypes. It thus allows novel insights into the relationship between blood homeostasis and its regulation through excretion. The newly discovered loci also include variants previously linked to chronic kidney disease (CPS1, SLC6A13), pulmonary hypertension (CPS1), and ischemic stroke (XYLB). By establishing connections from gene to disease via metabolic traits our results provide novel hypotheses about molecular mechanisms involved in the etiology of diseases.

Biological systems consist of multiple organizational levels all densely interacting with each other to ensure function and flexibility of the system. Simultaneous analysis of cross-sectional multi-omics data from large population studies is a powerful tool to comprehensively characterize the underlying molecular mechanisms on a physiological scale. In this study, we systematically analyzed the relationship between fasting serum metabolomics and whole blood transcriptomics data from 712 individuals of the German KORA F4 cohort. Correlation-based analysis identified 1,109 significant associations between 522 transcripts and 114 metabolites summarized in an integrated network, the 'human blood metabolome-transcriptome interface' (BMTI). Bidirectional causality analysis using Mendelian randomization did not yield any statistically significant causal associations between transcripts and metabolites. A knowledge-based interpretation and integration with a genome-scale human metabolic reconstruction revealed systematic signatures of signaling, transport and metabolic processes, i.e. metabolic reactions mainly belonging to lipid, energy and amino acid metabolism. Moreover, the construction of a network based on functional categories illustrated the cross-talk between the biological layers at a pathway level. Using a transcription factor binding site enrichment analysis, this pathway cross-talk was further confirmed at a regulatory level. Finally, we demonstrated how the constructed networks can be used to gain novel insights into molecular mechanisms associated to intermediate clinical traits. Overall, our results demonstrate the utility of a multi-omics integrative approach to understand the molecular mechanisms underlying both normal physiology and disease.

Advances in the "omics" field bring about the need for a high number of good quality samples. Many omics studies take advantage of biobanked samples to meet this need. Most of the laboratory errors occur in the pre-analytical phase. Therefore evidence-based standard operating procedures for the pre-analytical phase as well as markers to distinguish between 'good' and 'bad' quality samples taking into account the desired downstream analysis are urgently needed. We studied concentration changes of metabolites in serum samples due to pre-storage handling conditions as well as due to repeated freeze-thaw cycles. We collected fasting serum samples and subjected aliquots to up to four freeze-thaw cycles and to pre-storage handling delays of 12, 24 and 36 hours at room temperature (RT) and on wet and dry ice. For each treated aliquot, we quantified 127 metabolites through a targeted metabolomics approach. We found a clear signature of degradation in samples kept at RT. Storage on wet ice led to less pronounced concentration changes. 24 metabolites showed significant concentration changes at RT. In 22 of these, changes were already visible after only 12 hours of storage delay. Especially pronounced were increases in lysophosphatidylcholines and decreases in phosphatidylcholines. We showed that the ratio between the concentrations of these molecule classes could serve as a measure to distinguish between 'good' and 'bad' quality samples in our study. In contrast, we found quite stable metabolite concentrations during up to four freeze-thaw cycles. We concluded that pre-analytical RT handling of serum samples should be strictly avoided and serum samples should always be handled on wet ice or in cooling devices after centrifugation. Moreover, serum samples should be frozen at or below -80°C as soon as possible after centrifugation.

Nature Communications 8: Article number: 14357 (2017); Published 27 February 2017; Updated 11 April 2017 The original version of the Supplementary Information attached to this Article did not include Supplementary Note 1 The HTML has now been updated to include a corrected version of the Supplementary Information.

HSC-Explorer (http://mips.helmholtz-muenchen.de/HSC/) is a publicly available, integrative database containing detailed information about the early steps of hematopoiesis. The resource aims at providing fast and easy access to relevant information, in particular to the complex network of interacting cell types and molecules, from the wealth of publications in the field through visualization interfaces. It provides structured information on more than 7000 experimentally validated interactions between molecules, bioprocesses and environmental factors. Information is manually derived by critical reading of the scientific literature from expert annotators. Hematopoiesis-relevant interactions are accompanied with context information such as model organisms and experimental methods for enabling assessment of reliability and relevance of experimental results. Usage of established vocabularies facilitates downstream bioinformatics applications and to convert the results into complex networks. Several predefined datasets (Selected topics) offer insights into stem cell behavior, the stem cell niche and signaling processes supporting hematopoietic stem cell maintenance. HSC-Explorer provides a versatile web-based resource for scientists entering the field of hematopoiesis enabling users to inspect the associated biological processes through interactive graphical presentation.

Aims/hypothesisPolymorphisms in the transcription factor 7-like 2 (TCF7L2) gene have been shown to display a powerful association with type 2 diabetes. The aim of the present study was to evaluate metabolic alterations in carriers of a common TCF7L2 risk variant.MethodsSeventeen non-diabetic subjects carrying the T risk allele at the rs7903146 TCF7L2 locus and 24 subjects carrying no risk allele were submitted to intravenous glucose tolerance test and euglycemic-hyperinsulinemic clamp. Plasma samples were analysed for concentrations of 163 metabolites through targeted mass spectrometry.ResultsTCF7L2 risk allele carriers had a reduced first-phase insulin response and normal insulin sensitivity. Under fasting conditions, carriers of TCF7L2 rs7903146 exhibited a non-significant increase of plasma sphingomyelins (SMs), phosphatidylcholines (PCs) and lysophosphatidylcholines (lysoPCs) species. A significant genotype effect was detected in response to challenge tests in 6 SMs (C16:0, C16:1, C18:0, C18:1, C24:0, C24:1), 5 hydroxy-SMs (C14:1, C16:1, C22:1, C22:2, C24:1), 4 lysoPCs (C14:0, C16:0, C16:1, C17:0), 3 diacyl-PCs (C28:1, C36:6, C40:4) and 4 long-chain acyl-alkyl-PCs (C40:2, C40:5, C44:5, C44:6).DiscussionPlasma metabolomic profiling identified alterations of phospholipid metabolism in response to challenge tests in subjects with TCF7L2 rs7903146 genotype. This may reflect a genotype-mediated link to early metabolic abnormalities prior to the development of disturbed glucose tolerance.

ObjectiveTo characterise the influence of the fat free mass on the metabolite profile in serum samples from participants of the population-based KORA (Cooperative Health Research in the Region of Augsburg) S4 study.Subjects and methodsAnalyses were based on metabolite profile from 965 participants of the S4 and 890 weight-stable subjects of its seven-year follow-up study (KORA F4). 190 different serum metabolites were quantified in a targeted approach including amino acids, acylcarnitines, phosphatidylcholines (PCs), sphingomyelins and hexose. Associations between metabolite concentrations and the fat free mass index (FFMI) were analysed using adjusted linear regression models. To draw conclusions on enzymatic reactions, intra-metabolite class ratios were explored. Pairwise relationships among metabolites were investigated and illustrated by means of Gaussian graphical models (GGMs).ResultsWe found 339 significant associations between FFMI and various metabolites in KORA S4. Among the most prominent associations (p-values 4.75 × 10(-16)-8.95 × 10(-06)) with higher FFMI were increasing concentrations of the branched chained amino acids (BCAAs), ratios of BCAAs to glucogenic amino acids, and carnitine concentrations. For various PCs, a decrease in chain length or in saturation of the fatty acid moieties could be observed with increasing FFMI, as well as an overall shift from acyl-alkyl PCs to diacyl PCs. These findings were reproduced in KORA F4. The established GGMs supported the regression results and provided a comprehensive picture of the relationships between metabolites. In a sub-analysis, most of the discovered associations did not exist in obese subjects in contrast to non-obese subjects, possibly indicating derangements in skeletal muscle metabolism.ConclusionA set of serum metabolites strongly associated with FFMI was identified and a network explaining the relationships among metabolites was established. These results offer a novel and more complete picture of the FFMI effects on serum metabolites in a data-driven network.

Recent genome-wide association studies (GWAS) with metabolomics data linked genetic variation in the human genome to differences in individual metabolite levels. A strong relevance of this metabolic individuality for biomedical and pharmaceutical research has been reported. However, a considerable amount of the molecules currently quantified by modern metabolomics techniques are chemically unidentified. The identification of these "unknown metabolites" is still a demanding and intricate task, limiting their usability as functional markers of metabolic processes. As a consequence, previous GWAS largely ignored unknown metabolites as metabolic traits for the analysis. Here we present a systems-level approach that combines genome-wide association analysis and Gaussian graphical modeling with metabolomics to predict the identity of the unknown metabolites. We apply our method to original data of 517 metabolic traits, of which 225 are unknowns, and genotyping information on 655,658 genetic variants, measured in 1,768 human blood samples. We report previously undescribed genotype-metabotype associations for six distinct gene loci (SLC22A2, COMT, CYP3A5, CYP2C18, GBA3, UGT3A1) and one locus not related to any known gene (rs12413935). Overlaying the inferred genetic associations, metabolic networks, and knowledge-based pathway information, we derive testable hypotheses on the biochemical identities of 106 unknown metabolites. As a proof of principle, we experimentally confirm nine concrete predictions. We demonstrate the benefit of our method for the functional interpretation of previous metabolomics biomarker studies on liver detoxification, hypertension, and insulin resistance. Our approach is generic in nature and can be directly transferred to metabolomics data from different experimental platforms.

BackgroundHuman plasma and serum are widely used matrices in clinical and biological studies. However, different collecting procedures and the coagulation cascade influence concentrations of both proteins and metabolites in these matrices. The effects on metabolite concentration profiles have not been fully characterized.Methodology/principal findingsWe analyzed the concentrations of 163 metabolites in plasma and serum samples collected simultaneously from 377 fasting individuals. To ensure data quality, 41 metabolites with low measurement stability were excluded from further analysis. In addition, plasma and corresponding serum samples from 83 individuals were re-measured in the same plates and mean correlation coefficients (r) of all metabolites between the duplicates were 0.83 and 0.80 in plasma and serum, respectively, indicating significantly better stability of plasma compared to serum (p = 0.01). Metabolite profiles from plasma and serum were clearly distinct with 104 metabolites showing significantly higher concentrations in serum. In particular, 9 metabolites showed relative concentration differences larger than 20%. Despite differences in absolute concentration between the two matrices, for most metabolites the overall correlation was high (mean r = 0.81±0.10), which reflects a proportional change in concentration. Furthermore, when two groups of individuals with different phenotypes were compared with each other using both matrices, more metabolites with significantly different concentrations could be identified in serum than in plasma. For example, when 51 type 2 diabetes (T2D) patients were compared with 326 non-T2D individuals, 15 more significantly different metabolites were found in serum, in addition to the 25 common to both matrices.Conclusions/significanceOur study shows that reproducibility was good in both plasma and serum, and better in plasma. Furthermore, as long as the same blood preparation procedure is used, either matrix should generate similar results in clinical and biological studies. The higher metabolite concentrations in serum, however, make it possible to provide more sensitive results in biomarker detection.