Lothar Willmitzer

Background: Type 2 diabetes mellitus (T2DM) is associated with dysregulated lipid metabolism, but the serum lipid species associated with established T2DM may vary across populations and analytical platforms. We conducted an exploratory cross-sectional untargeted lipidomic analysis to characterize serum lipid features differing between adults with established T2DM and control participants. Methods: Serum samples from 36 patients with T2DM and 33 controls were analyzed by ultra-performance liquid chromatography-Orbitrap mass spectrometry. Lipidomic variation was assessed by principal component analysis and hierarchical clustering, followed by univariate analysis of putatively annotated lipid species, covariate-adjusted models, correlation with glycemic and lipid parameters in patients, and internally cross-validated receiver operating characteristic analysis. Results: Several phosphatidylcholine species (PC 38:4, PC 38:5, and PC 40:8) and PE 38:1 were higher in T2DM, whereas PI 36:1, TAG 54:6, and PC 32:2 were lower. These associations persisted after adjustment for age, sex, and body mass index and in participants with comparable reported fasting duration (>8 h but <10 h), but none of the principal species correlated significantly with HbA1c, plasma glucose, triglycerides, non-HDL cholesterol, or the triglyceride-glucose index among patients. A sex-by-disease interaction was supported only for PE 36:1, which was elevated predominantly in female patients. Conclusions: In this exploratory cross-sectional cohort, established T2DM was associated with selective differences in putatively annotated phospholipid and glycerolipid features. These findings do not establish causality, specificity for diabetes relative to broader metabolic dysfunction, or diagnostic utility and require future confirmation in larger longitudinal cohorts with targeted quantification and independent validation.

As a comprehensive analysis of all metabolites in a biological system, metabolomics is being widely applied in various clinical/health areas for disease prediction, diagnosis, and prognosis. However, challenges remain in dealing with the metabolomic complexity, massive data, metabolite identification, intra- and inter-individual variation, and reproducibility, which largely limit its widespread implementation. This study provided a comprehensive workflow for clinical metabolomics, including sample collection and preparation, mass spectrometry (MS) data acquisition, and data processing and analysis. Sample collection from multiple clinical sites was strictly carried out with standardized operation procedures (SOP). During data acquisition, three types of quality control (QC) samples were set for respective MS platforms (GC-MS, LC-MS polar, and LC-MS lipid) to assess the MS performance, facilitate metabolite identification, and eliminate contamination. Compounds annotation and identification were implemented with commercial software and in-house-developed PAppLine^TM and Ulib^MS library. The batch effects were removed using a deep learning model method (NormAE). Potential biomarkers identification was performed with tree-based modeling algorithms including random forest, AdaBoost, and XGBoost. The modeling performance was evaluated using the F1 score based on a 10-times repeated trial for each. Finally, a sub-cohort case study validated the reliability of the entire workflow.

Co-fractionation mass spectrometry (CF-MS)-based approaches enable cell-wide identification of protein-protein and protein-metabolite complexes present in the cellular lysate. CF-MS combines biochemical separation of molecular complexes with an untargeted mass-spectrometry-based proteomics and/or metabolomics analysis of the obtained fractions, and is used to delineate putative interactors. CF-MS data are a treasure trove for biological discovery. To facilitate analysis and visualization of original or publically available CF-MS datasets, we designed PROMISed, a user-friendly tool available online via https://myshiny.mpimp-golm.mpg.de/PDP1/ or as a repository via https://github.com/DennisSchlossarek/PROMISed. Specifically, starting with raw fractionation profiles, PROMISed (i) contains activities for data pre-processing and normalization, (ii) deconvolutes complex fractionation profiles into single, distinct peaks, (iii) identifies co-eluting protein–protein or protein–metabolite pairs using user-defined correlation methods, and (iv) performs co-fractionation network analysis. Given multiple CF-MS datasets, for instance representing different environmental condition, PROMISed allows to select for proteins and metabolites that differ in their elution profile, which may indicate change in the interaction status. But it also enables the identification of protein–protein and protein–metabolite pairs that co-elute together across multiple datasets. PROMISed enables users to (i) easily adjust parameters at each step of the analysis, (ii) download partial and final results, and (iii) select among different data-visualization options. PROMISed renders CF-MS data accessible to a broad scientific audience, allowing users with no computational or statistical background to look for novel protein–protein and protein–metabolite complexes for further experimental validation.

In contrast to climacteric fruits such as tomato, the knowledge on key regulatory genes controlling the ripening of strawberry, a nonclimacteric fruit, is still limited. NAC transcription factors (TFs) mediate different developmental processes in plants. Here, we identified and characterized Ripening Inducing Factor (FaRIF), a NAC TF that is highly expressed and induced in strawberry receptacles during ripening. Functional analyses based on stable transgenic lines aimed at silencing FaRIF by RNA interference, either from a constitutive promoter or the ripe receptacle-specific EXP2 promoter, as well as overexpression lines showed that FaRIF controls critical ripening-related processes such as fruit softening and pigment and sugar accumulation. Physiological, metabolome, and transcriptome analyses of receptacles of FaRIF-silenced and overexpression lines point to FaRIF as a key regulator of strawberry fruit ripening from early developmental stages, controlling abscisic acid biosynthesis and signaling, cell-wall degradation, and modification, the phenylpropanoid pathway, volatiles production, and the balance of the aerobic/anaerobic metabolism. FaRIF is therefore a target to be modified/edited to control the quality of strawberry fruits.

Grapes, one of the oldest agricultural crops, are cultivated to produce table fruits, dried fruits, juice, and wine. Grapevine variety is composed of clones that share common morphological traits. However, they can differ in minor genetic mutations which often result in not only notorious morphological changes but also in other non-visible sensorial distinctive attributes. In the present work, we identified three Vitis vinifera cv. Pinot noir clones grown under identical field conditions that showed different grape cluster types. Here, sensorial analysis together with non-targeted metabolite profiles by Ultra High performance Liquid Chromatography (UPLC) couples to Ultra High Resolution Mass Spectrometry (FT-ICR-MS) of wines elaborated from the three different grape cluster types was studied with the aim of (i) finding sensorial differences among these three types of wines, and, if there were, (ii) determining the molecular features (metabolites) associated with these sensorial attributes by a multivariate statistical approach.

Introduction: Insulin resistance (IR), a pathological state of low sensitivity to insulin in humans and animlas, is closely associated with type 2 diabetes mellitus, obesity and cardiovasculare dieseas. IR can be quantified using detailed protocols, such as the euglycemic-hyperinsulinemic clamp (EC) technique and the intravenous glucose tolerance test, or based on indices derived from the oral glucose tolerance test. Although these indices showed greater association with the incidence of diabetes, they allow no personalized estimation of the individual risk and cannot be used for monitoring of the individual changes in the insulin resistance. The modification of nutritional pattern is one of the first steps of prevention the IR and associated diseases. Diets with increased intake of branched chain amino acid lead to increase in IR in animals and human, possibly via disruption of insulin signaling in skeletal muscle. Here we aimed to investigate the changes in the plasma metabolome during constant insulin infusion. Additionally, the correlations between baseline concentrations of metabolites and changes of IR, which was measured in the EC-experiments, after high-protein and control diet were studied. Methods: In the first study (NCT00774488), middle-aged healthy obese subjects (n=14) underwent saline infusion and/or EC at a steady-state capillary plasma glucose concentration of 4.4 mmol/l. Plasma metabolites were measured using time-of-flight gas chromatography-mass spectrometry (GC-TOF/MS) technique. In the second study (NCT00579657), a randomized, controlled nutritional intervention (18-weeks) was conducted in 72 non-diabetic participants (overweight/obese: 29/43) with at least one further risk factor of metabolic syndrome. Participants were group-matched and allocated to 4 isoenergetic supplemented diets: control; high cereal-fiber; high-protein; or moderately increased cereal-fiber and protein (MIX). Whole-body IR was measured using EC. Plasma metabolome and IR were studied at 0, 6 and 18-weeks. Results: Eight metabolites: altrose, asparagine, glycerol-2-phosphate, gulose, heptadecanoic acid, phenylalanine, pyroglutamate and talose correlated with high-protein diet-induced changes in IR and significantly changed during EC. Conclusions: Multimarker strategy with use of plasma metabolic profiling appears to be a useful tool for both the assessment of IR and the ‘metabolic signature of insulin effects’ (i.e. doping control of elite athletes) in humans.

Key message Complementing genomic data with other “omics” predictors can increase the probability of success for predicting the best hybrid combinations using complex agronomic traits. Abstract Accurate prediction of traits with complex genetic architecture is crucial for selecting superior candidates in animal and plant breeding and for guiding decisions in personalized medicine. Whole-genome prediction has revolutionized these areas but has inherent limitations in incorporating intricate epistatic interactions. Downstream “omics” data are expected to integrate interactions within and between different biological strata and provide the opportunity to improve trait prediction. Yet, predicting traits from parents to progeny has not been addressed by a combination of “omics” data. Here, we evaluate several “omics” predictors—genomic, transcriptomic and metabolic data—measured on parent lines at early developmental stages and demonstrate that the integration of transcriptomic with genomic data leads to higher success rates in the correct prediction of untested hybrid combinations in maize. Despite the high predictive ability of genomic data, transcriptomic data alone outperformed them and other predictors for the most complex heterotic trait, dry matter yield. An eQTL analysis revealed that transcriptomic data integrate genomic information from both, adjacent and distant sites relative to the expressed genes. Together, these findings suggest that downstream predictors capture physiological epistasis that is transmitted from parents to their hybrid offspring. We conclude that the use of downstream “omics” data in prediction can exploit important information beyond structural genomics for leveraging the efficiency of hybrid breeding.

Interactions between metabolites and proteins play an integral role in all cellular functions. Here we describe an affinity purification (AP) approach in combination with LC/MS-based metabolomics and proteomics that allows, to our knowledge for the first time, analysis of protein–metabolite and protein–protein interactions simultaneously in plant systems. More specifically, we examined protein and small-molecule partners of the three (of five) nucleoside diphosphate kinases present in the Arabidopsis genome (NDPK1–NDPK3). The bona fide role of NDPKs is the exchange of terminal phosphate groups between nucleoside diphosphates (NDPs) and triphosphates (NTPs). However, other functions have been reported, which probably depend on both the proteins and small molecules specifically interacting with the NDPK. Using our approach we identified 23, 17, and 8 novel protein partners of NDPK1, NDPK2, and NDPK3, respectively, with nucleotide-dependent proteins such as actin and adenosine kinase 2 being enriched. Particularly interesting, however, was the co-elution of glutathione S-transferases (GSTs) and reduced glutathione (GSH) with the affinity-purified NDPK1 complexes. Following up on this finding, we could demonstrate that NDPK1 undergoes glutathionylation, opening a new paradigm of NDPK regulation in plants. The described results extend our knowledge of NDPKs, the key enzymes regulating NDP/NTP homeostasis.