Bayesian independent component analysis recovers pathway signatures from blood metabolomics data

Interpreting the complex interplay of metabolites in heterogeneous biosamples still poses a challenging task. In this study, we propose independent component analysis (ICA) as a multivariate analysis tool for the interpretation of large-scale metabolomics data. In particular, we employ a Bayesian ICA method based on a mean-field approach, which allows us to statistically infer the number of independent components to be reconstructed. The advantage of ICA over correlation-based methods like principal component analysis (PCA) is the utilization of higher order statistical dependencies, which not only yield additional information but also allow a more meaningful representation of the data with fewer components. We performed the described ICA approach on a large-scale metabolomics data set of human serum samples, comprising a total of 1764 study probands with 218 measured metabolites. Inspecting the source matrix of statistically independent metabolite profiles using a weighted enrichment algorithm, we observe strong enrichment of specific metabolic pathways in all components. This includes signatures from amino acid metabolism, energy-related processes, carbohydrate metabolism, and lipid metabolism. Our results imply that the human blood metabolome is composed of a distinct set of overlaying, statistically independent signals. ICA furthermore produces a mixing matrix, describing the strength of each independent component for each of the study probands. Correlating these values with plasma high-density lipoprotein (HDL) levels, we establish a novel association between HDL plasma levels and the branched-chain amino acid pathway. We conclude that the Bayesian ICA methodology has the power and flexibility to replace many of the nowadays common PCA and clustering-based analyses common in the research field.     

Mining the Unknown: A Systems Approach to Metabolite Identification Combining Genetic and Metabolic Information

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.     

A Variant In the Abo Gene Explains the Variation in Soluble E-Selectin Levels—Results from Dense Genotyping in Two Independent Populations

Background

Elevated soluble (s) E-selectin levels have been associated with various cardiovascular diseases. Recently, genetic variants in the ABO blood group have been related to E-selectin levels in a small cohort of patients with type 1 diabetes. We evaluated whether this association is reproducible in two large samples of Caucasians.

Methodology/ Principal Findings

Data of the present study was drawn from the population-based MONICA/KORA Augsburg study (n = 1,482) and the patients-based LURIC study (n = 1,546). A high-density genotyping array (50K IBC Chip) containing single-nucleotide polymorphisms (SNPs) from E-selectin candidate genes selected on known biology of E-selectin metabolism, mouse genetic studies, and human genetic association studies, was used for genotyping. Linear regression analyses with adjustment for age and sex (and survey in KORA) were applied to assess associations between gene variants and sE-selectin concentrations. A number of 12 SNPs (in KORA) and 13 SNPs (in LURIC), all from the ABO blood group gene, were significantly associated with the log-transformed concentration of E-selectin. The strongest association was observed for rs651007 with a change of log-transformed sE-selectin per one copy of the minor allele of −0.37 ng/ml (p = 1.87×10⁻¹⁰³) in KORA and −0.35 ng/ml (p = 5.11×10⁻⁸⁴) in LURIC. Inclusion of rs651007 increased the explained sE-selectin variance by 0.256 in KORA and 0.213 in LURIC. All SNPs had minor allele frequencies above 20% showing a substantial gene variation.

Conclusions/ Significance

Our findings in two independent samples indicate that the genetic variants at the ABO locus affect sE-selectin levels. Since distinct genome-wide association studies linked the ABO gene with myocardial infarction (MI) in the presence of coronary atherosclerosis and with coronary artery disease, these findings may not only enhance our understanding of adhesion molecule biology, but may also provide a focus for several novel research avenues.     

Preservation of Metabolic Flexibility in Skeletal Muscle by a Combined Use of n-3 PUFA and Rosiglitazone in Dietary Obese Mice

Insulin resistance, the key defect in type 2 diabetes (T2D), is associated with a low capacity to adapt fuel oxidation to fuel availability, i.e., metabolic inflexibility. This, in turn, contributes to a further damage of insulin signaling. Effectiveness of T2D treatment depends in large part on the improvement of insulin sensitivity and metabolic adaptability of the muscle, the main site of whole-body glucose utilization. We have shown previously in mice fed an obesogenic high-fat diet that a combined use of n-3 long-chain polyunsaturated fatty acids (n-3 LC-PUFA) and thiazolidinediones (TZDs), anti-diabetic drugs, preserved metabolic health and synergistically improved muscle insulin sensitivity. We investigated here whether n-3 LC-PUFA could elicit additive beneficial effects on metabolic flexibility when combined with a TZD drug rosiglitazone. Adult male C57BL/6N mice were fed an obesogenic corn oil-based high-fat diet (cHF) for 8 weeks, or randomly assigned to various interventions: cHF with n-3 LC-PUFA concentrate replacing 15% of dietary lipids (cHF+F), cHF with 10 mg rosiglitazone/kg diet (cHF+ROSI), cHF+F+ROSI, or chow-fed. Indirect calorimetry demonstrated superior preservation of metabolic flexibility to carbohydrates in response to the combined intervention. Metabolomic and gene expression analyses in the muscle suggested distinct and complementary effects of the interventions, with n-3 LC-PUFA supporting complete oxidation of fatty acids in mitochondria and the combination with n-3 LC-PUFA and rosiglitazone augmenting insulin sensitivity by the modulation of branched-chain amino acid metabolism. These beneficial metabolic effects were associated with the activation of the switch between glycolytic and oxidative muscle fibers, especially in the cHF+F+ROSI mice. Our results further support the idea that the combined use of n-3 LC-PUFA and TZDs could improve the efficacy of the therapy of obese and diabetic patients.     

Long term conservation of human metabolic phenotypes and link to heritability

Changes in an individual’s human metabolic phenotype (metabotype) over time can be indicative of disorder-related modifications. Studies covering several months to a few years have shown that metabolic profiles are often specific for an individual. This “metabolic individuality” and detected changes may contribute to personalized approaches in human health care. However, it is not clear whether such individual metabotypes persist over longer time periods. Here we investigate the conservation of metabotypes characterized by 212 different metabolites of 818 participants from the Cooperative Health Research in the Region of Augsburg; Germany population, taken within a 7-year time interval. For replication, we used paired samples from 83 non-related individuals from the TwinsUK study. Results indicated that over 40 % of all study participants could be uniquely identified after 7 years based on their metabolic profiles alone. Moreover, 95 % of the study participants showed a high degree of metabotype conservation (>70 %) whereas the remaining 5 % displayed major changes in their metabolic profiles over time. These latter individuals were likely to have undergone important biochemical changes between the two time points. We further show that metabolite conservation was positively associated with heritability (rank correlation 0.74), although there were some notable exceptions. Our results suggest that monitoring changes in metabotypes over several years can trace changes in health status and may provide indications for disease onset. Moreover, our study findings provide a general reference for metabotype conservation over longer time periods that can be used in biomarker discovery studies.     

Associations between thyroid hormones and serum metabolite profiles in an euthyroid population

Metabolomics - The aim was to characterise associations between circulating thyroid hormones—free thyroxine (FT4) and thyrotropin (TSH)—and the metabolite profiles in serum samples from...     

Gonadal function in young adults after surgical treatment of cryptorchidism.

In a follow-up study of 48 young men who had been surgically treated for cryptorchidism before puberty testicular function was assessed by examining the genitalia, testicular volume, secondary sex characteristics, semen, plasma luteinising hormone (LH) and follicle-stimulating hormone (FSH) concentrations after luteinising hormone-releasing hormone stimulation, and plasma testosterone concentrations. Clinical androgen effects were normal. The mean testicular volume of both testes was in the low normal range in those who had had unilateral cryptorchidism and below normal in those who had had bilateral cryptorchidism. Of 37 patients whose sperm counts were recorded (14 bilateral) six showed azoospermia (all bilateral), five had severe oligospermia (four bilateral), and 10 had moderate oligospermia (one bilateral). In nearly all those who had had bilateral cryptorchidism and most of those who had had unilateral cryptorchidism plasma gonadotrophin levels were increased. Four cases of possible partial LH deficiency were identified. Plasma testosterone concentrations were normal in all except two patients.     

Regulation of phenol sulfotransferase expression in cultured bovine bronchial epithelial cells by hydrocortisone

One conjugative pathway for the inactivation of endogenous and exogenous hydroxylated aromatic compounds is catalyzed by phenol (aryl) sulfotransferases (PSTs), which esterify phenolic acceptors with sulfate. The tracheobronchial epithelium is commonly exposed to phenolic drugs and pollutants, and metabolic sulfation and PST activity in this tissue have been previously demonstrated. To determine what factors may control PST expression, extracts of serum-free, growth factor-supplemented cultures of bovine bronchial epithelial cells were assayed for PST activity and PST antigen. The most significant finding was dose-dependent, apparent stimulated expression by hydrocortisone (EC₅₀ = 4 nM, maximal stimulation at 20 nM). Time-course experiments, however, revealed progressive loss of PST in the absence of corticosteroid. After decay of extant PST in steroid-free medium, hydrocortisone reinduced the expression of PST three to fivefold. Western blots using mouse anti-bovine PST revealed corresponding increases in 32 kDa PST protein levels in response to hydrocortisone. Steady state kinetic analyses indicated apparent K_m values of 1—3 μM for 2-naphthol regardless of culture conditions. These results suggest that detoxification of phenolic compounds by sulfation may be regulated by corticosteroids.     

Gaussian graphical modeling reconstructs pathway reactions from high-throughput metabolomics data

Background  

With the advent of high-throughput targeted metabolic profiling techniques, the question of how to interpret and analyze the resulting vast amount of data becomes more and more important. In this work we address the reconstruction of metabolic reactions from cross-sectional metabolomics data, that is without the requirement for time-resolved measurements or specific system perturbations. Previous studies in this area mainly focused on Pearson correlation coefficients, which however are generally incapable of distinguishing between direct and indirect metabolic interactions.