Impact of fibre supplementation on microbiome and resilience in healthy participants: A randomized, placebo-controlled clinical trial.

BACKGROUND AND AIMS: The gut microbiome exerts important roles in health, e.g., functions in metabolism and immunology. These functions are often exerted via short-chain fatty acid (SCFA) production by gut bacteria. Studies demonstrating causal relationships between interventions targeting the microbiome and clinical outcomes are limited. This study aimed to show a causal relationship between microbiome modulation through fibre intervention and health. METHODS AND RESULTS: This randomized, double-blind, cross-over study included 65 healthy subjects, aged 45-70 years, with increased metabolic risk (i.e., body mass index [BMI] 25-30 kg/m2, low to moderate daily dietary fibre intake, <30g/day). Subjects took daily a fibre mixture of Acacia gum and carrot powder or placebo for 12 weeks, with an 8-week wash-out period. Faecal samples for measurement of SCFAs and microbiome analysis were collected every 4 weeks. Before and after each intervention period subjects underwent the mixed-meal PhenFlex challenge Test (PFT). Health effects were expressed as resilience to the stressors of the PFT and as fasting metabolic and inflammatory state. The fibre mixture exerted microbiome modulation, with an increase in ß-diversity (p < 0.001). α-diversity was lower during fibre mixture intake compared to placebo after 4, 8 and 12 weeks (p = 0.002; p = 0.012; p = 0.031). There was no effect observed on faecal SCFA concentrations, nor on any of the primary clinical outcomes (Inflammatory resilience: p = 0.605, Metabolic resilience: p = 0.485). CONCLUSION: Although the intervention exerted effects on gut microbiome composition, no effects on SCFA production, on resilience or fasting metabolic and inflammatory state were observed in this cohort. REGISTRATION NUMBER CLINICALTRIALS.GOV: NCT04829396.

Lipidome Analysis in Brain and Peripheral Plasma Following Milk Fat Globule Membrane Supplementation in Rodents.

SCOPE: Milk fat globule membrane (MFGM) is an essential component of milk. Bovine MFGM (bMFGM) has been shown to support cognitive development and increase relative concentrations of serum phospholipids. This study investigates bioavailability of bMFGM components after oral administration in two preclinical models to explore whether dietary bMFGM induces parallel changes to plasma and brain lipidomes. METHODS AND RESULTS: Transgenic APOE*3.Leiden mice (n = 18 per group) and Sprague-Dawley rats (n = 12 per group) are fed bMFGM-enriched (MFGM+) or Control diet, followed by phospholipid profile-determination in plasma, hippocampus, and prefrontal cortex tissue by targeted mass spectrometry. Multivariate analysis of lipidomic profiles demonstrates a separation between MFGM+ and Control plasma across rodents. In plasma, sphingomyelins contributed the most to the separation of lipid patterns among both models, where three sphingomyelins (d18:1/14:0, d18:1/23:0, d18:1/23:1[9Z]) are consistently higher in the circulation of MFGM+ groups. A similar trend is observed in rat prefrontal cortex, although no significant separation of the brain lipidome is demonstrated. CONCLUSION: bMFGM-enriched diet alters plasma phospholipid composition in rodents, predominantly increasing sphingomyelin levels in the systemic circulation with similar, but non-significant, trends in central brain regions. These changes may contribute to the beneficial effects of bMFGM on neurodevelopment during early life.

A Novel Personalized Systems Nutrition Program Improves Dietary Patterns, Lifestyle Behaviors and Health-Related Outcomes: Results from the Habit Study.

Personalized nutrition may be more effective in changing lifestyle behaviors compared to population-based guidelines. This single-arm exploratory study evaluated the impact of a 10-week personalized systems nutrition (PSN) program on lifestyle behavior and health outcomes. Healthy men and women (n = 82) completed the trial. Individuals were grouped into seven diet types, for which phenotypic, genotypic and behavioral data were used to generate personalized recommendations. Behavior change guidance was also provided. The intervention reduced the intake of calories (-256.2 kcal; p < 0.0001), carbohydrates (-22.1 g; p < 0.0039), sugar (-13.0 g; p < 0.0001), total fat (-17.3 g; p < 0.0001), saturated fat (-5.9 g; p = 0.0003) and PUFA (-2.5 g; p = 0.0065). Additionally, BMI (-0.6 kg/m2; p < 0.0001), body fat (-1.2%; p = 0.0192) and hip circumference (-5.8 cm; p < 0.0001) were decreased after the intervention. In the subgroup with the lowest phenotypic flexibility, a measure of the body's ability to adapt to environmental stressors, LDL (-0.44 mmol/L; p = 0.002) and total cholesterol (-0.49 mmol/L; p < 0.0001) were reduced after the intervention. This study shows that a PSN program in a workforce improves lifestyle habits and reduces body weight, BMI and other health-related outcomes. Health improvement was most pronounced in the compromised phenotypic flexibility subgroup, which indicates that a PSN program may be effective in targeting behavior change in health-compromised target groups.

Allergen risk assessment: Food intake levels of the general population represent those of food allergic patients.

Unintentional intake of allergens through food products poses a daily risk for allergic patients. Models estimating the risk of reactions mostly use intake data from general population surveys. Our study evaluates the comparability of food intake levels in the general population to those in the food allergic population. Data were collected by a 24-h recall method on 2 non-consecutive days in 38 cow's milk and/or hen's egg and 35 peanut and/or tree nut allergic adult patients. All products were assigned to food groups previously developed for allergen risk assessment. Food intake distributions from the allergic populations and a matched sample from the Dutch National Food Consumption Survey were compared, and risk assessments were performed. Food intake data was obtained for 92% of the food groups. Comparison of the intake showed no statistically significant differences between either of the two allergic populations and the general population. Consequently, only small variations in estimated risks were found, that would not result in different risk management decisions. In conclusion, food intake data from the general population can be used for food allergen risk assessment and will not lead to a relevant under- or overestimation of the risk for the food allergic population.

The possibilities of the use of N-of-1 and do-it-yourself trials in nutritional research.

BACKGROUND: N-of-1 designs gain popularity in nutritional research because of the improving technological possibilities, practical applicability and promise of increased accuracy and sensitivity, especially in the field of personalized nutrition. This move asks for a search of applicable statistical methods. OBJECTIVE: To demonstrate the differences of three popular statistical methods in analyzing treatment effects of data obtained in N-of-1 designs. METHOD: We compare Individual-participant data meta-analysis, frequentist and Bayesian linear mixed effect models using a simulation experiment. Furthermore, we demonstrate the merits of the Bayesian model including prior information by analyzing data of an empirical study on weight loss. RESULTS: The linear mixed effect models are to be preferred over the meta-analysis method, since the individual effects are estimated more accurately as evidenced by the lower errors, especially with lower sample sizes. Differences between Bayesian and frequentist mixed models were found to be small, indicating that they will lead to the same results without including an informative prior. CONCLUSION: For empirical data, the Bayesian mixed model allows the inclusion of prior knowledge and gives potential for population based and personalized inference.

Evaluation of chitotriosidase as a biomarker for adipose tissue inflammation in overweight individuals and type 2 diabetic patients.

BACKGROUND: Overweight and obesity can lead to adipose tissue inflammation, which causes insulin resistance and on the long-term type 2 diabetes mellitus (T2D). The inflammatory changes of obese-adipose tissue are characterized by macrophage infiltration and activation, but validated circulating biomarkers for adipose tissue inflammation for clinical use are still lacking. One of the most secreted enzymes by activated macrophages is chitotriosidase (CHIT1). OBJECTIVE: To test whether circulating CHIT1 enzymatic activity levels reflect adipose tissue inflammation. METHODS: Plasma and adipose tissue samples of 105 subjects (35 lean, 37 overweight, and 33 T2D patients) were investigated. CHIT1 mRNA levels were determined in adipose tissue-resident innate immune cells. CHIT1 mRNA levels, protein abundance, and plasma enzymatic activity were subsequently measured in adipose tissue biopsies and plasma of control subjects with varying levels of obesity and adipose tissue inflammation as well as in T2D patients. RESULTS: In adipose tissue, CHIT1 mRNA levels were higher in stromal vascular cells compared to adipocytes, and higher in adipose tissue-residing macrophages compared to circulating monocytes (p < 0.001). CHIT1 mRNA levels in adipose tissue were enhanced in overweightcompared to lean subjects and even more in T2D patients (p < 0.05). In contrast, plasma CHIT1 enzymatic activity did not differ between lean, overweight subjects and T2D patients. A mutation of the CHIT1 gene decreases plasma CHIT1 activity. CONCLUSIONS: CHIT1 is expressed by adipose tissue macrophages and expression is higher in overweight subjects and T2D patients, indicating its potential as tissue biomarker for adipose tissue inflammation. However, these differences do not translate into different plasma CHIT1 activity levels. Moreover, a common CHIT1 gene mutation causing loss of plasma CHIT1 activity interferes with its use as a biomarker of adipose tissue inflammation. These results indicate that plasma CHIT1 activity is of limited value as a circulating biomarker for adipose tissue inflammation in human subjects.

Metabolic Profiling Reveals Differences in Plasma Concentrations of Arabinose and Xylose after Consumption of Fiber-Rich Pasta and Wheat Bread with Differential Rates of Systemic Appearance of Exogenous Glucose in Healthy Men.

BACKGROUND: The consumption of products rich in cereal fiber and with a low glycemic index is implicated in a lower risk of metabolic diseases. Previously, we showed that the consumption of fiber-rich pasta compared with bread resulted in a lower rate of appearance of exogenous glucose and a lower glucose clearance rate quantified with a dual-isotope technique, which was in accordance with a lower insulin and glucose-dependent insulinotropic polypeptide response. OBJECTIVE: To gain more insight into the acute metabolic consequences of the consumption of products resulting in differential glucose kinetics, postprandial metabolic profiles were determined. METHODS: In a crossover study, 9 healthy men [mean ± SEM age: 21 ± 0.5 y; mean ± SEM body mass index (kg/m2): 22 ± 0.5] consumed wheat bread (132 g) and fresh pasta (119 g uncooked) enriched with wheat bran (10%) meals. A total of 134 different metabolites in postprandial plasma samples (at -5, 30, 60, 90, 120, and 180 min) were quantified by using a gas chromatography-mass spectrometry-based metabolomics approach (secondary outcomes). Two-factor ANOVA and advanced multivariate statistical analysis (partial least squares) were applied to detect differences between both food products. RESULTS: Forty-two different postprandial metabolite profiles were identified, primarily representing pathways related to protein and energy metabolism, which were on average 8% and 7% lower after the men consumed pasta rather than bread, whereas concentrations of arabinose and xylose were 58% and 53% higher, respectively. Arabinose and xylose are derived from arabinoxylans, which are important components of wheat bran. The higher bioavailability of arabinose and xylose after pasta intake coincided with a lower rate of appearance of glucose and amino acids. We speculate that this higher bioavailability is due to higher degradation of arabinoxylans by small intestinal microbiota, facilitated by the higher viscosity of arabinoxylans after pasta intake than after bread intake. CONCLUSION: This study suggests that wheat bran, depending on the method of processing, can increase the viscosity of the meal bolus in the small intestine and interfere with macronutrient absorption in healthy men, thereby influencing postprandial glucose and insulin responses. This trial was registered at www.controlled-trials.com as ISRCTN42106325.

Quantifying phenotypic flexibility as the response to a high-fat challenge test in different states of metabolic health.

Metabolism maintains homeostasis at chronic hypercaloric conditions, activating postprandial response mechanisms, which come at the cost of adaptation processes such as energy storage, eventually with negative health consequences. This study quantified the metabolic adaptation capacity by studying challenge response curves. After a high-fat challenge, the 8 h response curves of 61 biomarkers related to adipose tissue mass and function, systemic stress, metabolic flexibility, vascular health, and glucose metabolism was compared between 3 metabolic health stages: 10 healthy men, before and after 4 wk of high-fat, high-calorie diet (1300 kcal/d extra), and 9 men with metabolic syndrome (MetS). The MetS subjects had increased fasting concentrations of biomarkers representing the 3 core processes, glucose, TG, and inflammation control, and the challenge response curves of most biomarkers were altered. After the 4 wk hypercaloric dietary intervention, these 3 processes were not changed, as compared with the preintervention state in the healthy subjects, whereas the challenge response curves of almost all endocrine, metabolic, and inflammatory processes regulating these core processes were altered, demonstrating major molecular physiologic efforts to maintain homeostasis. This study thus demonstrates that change in challenge response is a more sensitive biomarker of metabolic resilience than are changes in fasting concentrations.

Toxicity assessment of aggregated/agglomerated cerium oxide nanoparticles in an in vitro 3D airway model: the influence of mucociliary clearance.

We investigated the toxicity of aggregated nanoparticles of cerium oxide (CeO2) using an in vitro 3D human bronchial epithelial model that included a mucociliary apparatus (MucilAir™). CeO2 was dispersed in saline and applied to the apical surface of the model. CeO2 did not induce distinct effects in the model, whereas it did in BEAS-2B and A549 cell cultures. The absence of effects of CeO2 was not because of the model's insensitivity. Nanoparticles of zinc oxide (ZnO) elicited positive responses in the toxicological assays. Respiratory mucus (0.1% and 1%) added to dispersions increased aggregation/agglomeration to such an extent that most CeO2 sedimented within a few minutes. Also, the mucociliary apparatus of the model removed CeO2 from the central part of the apical surface to the borders. This 'clearance' may have prevented the majority of CeO2 from reaching the epithelial cells. Chemical analysis of cerium in the basal tissue culture medium showed only minimal translocation of cerium across the 3D barrier. In conclusion, mucociliary defence appeared to prevent CeO2 reaching the respiratory epithelial cells in this 3D in vitro model. This model and approach can be used to study compounds of specific toxicological concern in airway defence mechanisms in vitro.

Raman spectroscopy as a promising tool for noninvasive point-of-care glucose monitoring.

Self-monitoring of glucose is important for managing diabetes. Noninvasive glucose monitors are not yet available, but patients would benefit highly from such a device. We present results that may lead to a novel, point-of-care noninvasive system to measure blood glucose based on Raman spectroscopy. A hospitalized cohort of 111 subjects was measured using a custom-made Raman spectrometer system. Blood glucose reference samples were used to correlate Raman data to glucose levels, using advanced preprocessing and analysis algorithms. A correlation coefficient (R (2)) of .83 was found correlating independent Raman-based predictions on reference blood glucose for the full cohort. Stratification of the cohort in gender-specific groups raised correlation levels to .88 (females) and .94 (males). Glucose could be measured noninvasively with average errors as low as 0.9 mM. We conclude that this novel system shows promising results for the advance of noninvasive, point-of-care glucose monitoring.

Identification of biomarkers for intake of protein from meat, dairy products and grains: a controlled dietary intervention study.

In the present controlled, randomised, multiple cross-over dietary intervention study, we aimed to identify potential biomarkers for dietary protein from dairy products, meat and grain, which could be useful to estimate intake of these protein types in epidemiological studies. After 9 d run-in, thirty men and seventeen women (22 (SD 4) years) received three high-protein diets (aimed at approximately 18% of energy (en%)) in random order for 1 week each, with approximately 14 en% originating from either meat, dairy products or grain. We used a two-step approach to identify biomarkers in urine and plasma. With principal component discriminant analysis, we identified amino acids (AA) from the plasma or urinary AA profile that were distinctive between diets. Subsequently, after pooling total study data, we applied mixed models to estimate the predictive value of those AA for intake of protein types. A very good prediction could be made for the intake of meat protein by a regression model that included urinary carnosine, 1-methylhistidine and 3-methylhistidine (98% of variation in intake explained). Furthermore, for dietary grain protein, a model that included seven AA (plasma lysine, valine, threonine, α-aminobutyric acid, proline, ornithine and arginine) made a good prediction (75% of variation explained). We could not identify biomarkers for dairy protein intake. In conclusion, specific combinations of urinary and plasma AA may be potentially useful biomarkers for meat and grain protein intake, respectively. These findings need to be cross-validated in other dietary intervention studies.

Fish oil and inflammatory status alter the n-3 to n-6 balance of the endocannabinoid and oxylipin metabolomes in mouse plasma and tissues.

It is well established that dietary intake of n-3 fatty acids is associated with anti-inflammatory effects, and this has been linked to modulation of the oxylipin and endocannabinoid metabolomes. However, the amount of data on specific tissue effects is limited, and it is not known how inflammation affects this relation. In the present study we systematically explored the combined effects of n-3 fatty acid diets and inflammation on the in vivo endocannabinoid and oxylipin metabolomes using a multicompartment, detailed targeted lipidomics approach. Male C57BL/6 mice received diets containing 0, 1, or 3 % w/w fish oil (FO) for 6 weeks, after which 2 mg/kg LPS or saline was administered i.p. Levels of endocannabinoids/N-acylethanolamines (NAEs) and oxylipins, covering n-3 and n-6 fatty acid derived compounds, were determined in plasma, liver, ileum and adipose tissue using LC-MS/MS. FO generally increased 'n-3' NAEs and oxylipins at the expense of compounds derived from other fatty acids, affecting all branches of the oxylipin metabolome. LPS generally increased levels of endocannabinoids/NAEs and oxylipins, with opposing effects across plasma and tissues. Multivariate data analysis revealed that separation between diet groups in the saline treated groups was primarily explained by decreases in other than n-3 derived compounds. In the LPS treated groups, the separation was primarily explained by increases in n-3 derived compounds. In conclusion, FO caused marked changes in the n-3 to n-6 balance of the endocannabinoid and oxylipin metabolomes, with specific effects depending on inflammatory status. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1007/s11306-012-0421-9) contains supplementary material, which is available to authorized users.

Time-dependent effect of in vivo inflammation on eicosanoid and endocannabinoid levels in plasma, liver, ileum and adipose tissue in C57BL/6 mice fed a fish-oil diet.

Eicosanoids and endocannabinoids/N-acylethanolamines (NAEs) are fatty acid derived compounds with a regulatory role in inflammation. Considering their complex metabolism, it is likely that inflammation affects multiple compounds at the same time, but how lipid profiles change in plasma and other tissues after an inflammatory stimulus has not been described in detail. In addition, dietary fish oil increases levels of several n-3 fatty acid derived eicosanoids and endocannabinoids, and this may lead to a broader change in the profiles of bioactive lipids. In the present study mice were fed a diet containing 3% w/w fish oil for 6 weeks before receiving i.p. saline or 3 mg/kg lipopolysaccharide (LPS) to induce an inflammatory response. Eicosanoid and endocannabinoid/NAE levels (in total 61 metabolites) in plasma, liver, ileum, and adipose tissue were quantified using targeted lipidomics after 2, 4, 8, and 24 h, respectively. Tissue- and time-dependent effects of LPS on bioactive lipid profiles were observed. For example, levels of CYP derived eicosanoids in the ileum were markedly affected by LPS, whereas this was less pronounced in the plasma and adipose tissue. For some compounds, such as 9,10-DiHOME, opposing effects of LPS were seen in the plasma compared to the other tissues, suggesting differential regulation of bioactive lipid levels after an inflammatory stimulus. Taken together, our results show that plasma levels do not always correlate with the effects found in the tissues, which underlines the need to measure profiles and pathways of mediators involved in inflammation, including endocannabinoid-like structures, in both plasma and tissues.

Visualization and identification of health space, based on personalized molecular phenotype and treatment response to relevant underlying biological processes.

BACKGROUND: Being able to visualize multivariate biological treatment effects can be insightful. However the axes in visualizations are often solely defined by variation and thus have no biological meaning. This makes the effects of treatment difficult to interpret. METHODS: A statistical visualization method is presented, which analyses and visualizes the effects of treatment in individual subjects. The visualization is based on predefined biological processes as determined by systems-biological datasets (metabolomics proteomics and transcriptomics). This allows one to evaluate biological effects depending on shifts of either groups or subjects in the space predefined by the axes, which illustrate specific biological processes. We built validated multivariate models for each axis to represent several biological processes. In this space each subject has his or her own score on each axis/process, indicating to which extent the treatment affects the related process. RESULTS: The health space model was applied to visualize the effects of a nutritional intervention, with the goal of applying diet to improve health. The model was therefore named the 'health space' model. The 36 study subjects received a 5-week dietary intervention containing several anti-inflammatory ingredients. Plasma concentrations of 79 proteins and 145 metabolites were quantified prior to and after treatment. The principal processes modulated by the intervention were oxidative stress, inflammation, and metabolism. These processes formed the axes of the 'health space'. The approach distinguished the treated and untreated groups, as well as two different response subgroups. One subgroup reacted mainly by modulating its metabolic stress profile, while a second subgroup showed a specific inflammatory and oxidative response to treatment. CONCLUSIONS: The 'health space' model allows visualization of multiple results and to interpret them. The model presents treatment group effects, subgroups and individual responses.

Metabolomics as a tool for target identification in strain improvement: the influence of phenotype definition.

For the optimization of microbial production processes, the choice of the quantitative phenotype to be optimized is crucial. For instance, for the optimization of product formation, either product concentration or productivity can be pursued, potentially resulting in different targets for strain improvement. The choice of a quantitative phenotype is highly relevant for classical improvement approaches, and even more so for modern systems biology approaches. In this study, the information content of a metabolomics dataset was determined with respect to different quantitative phenotypes related to the formation of specific products. To this end, the production of two industrially relevant products by Aspergillus niger was evaluated: (i) the enzyme glucoamylase, and (ii) the more complex product group of secreted proteases, consisting of multiple enzymes. For both products, six quantitative phenotypes associated with activity and productivity were defined, also taking into account different time points of sampling during the fermentation. Both linear and nonlinear relationships between the metabolome data and the different quantitative phenotypes were considered. The multivariate data analysis tool partial least-squares (PLS) was used to evaluate the information content of the datasets for all the different quantitative phenotypes defined. Depending on the product studied, different quantitative phenotypes were found to have the highest information content in specific metabolomics datasets. A detailed analysis of the metabolites that showed strong correlation with these quantitative phenotypes revealed that various sugar derivatives correlated with glucoamylase activity. For the reduction of protease activity, mainly as-yet-unidentified compounds correlated.

Analyzing longitudinal microbial metabolomics data.

A longitudinal experimental design in combination with metabolomics and multiway data analysis is a powerful approach in the identification of metabolites whose correlation with bioproduct formation shows a shift in time. In this paper, a strategy is presented for the analysis of longitudinal microbial metabolomics data, which was performed in order to identify metabolites that are likely inducers of phenylalanine production by Escherichia coli. The variation in phenylalanine production as a function of differences in metabolism induced by the different environmental conditions in time was described by a validated multiway statistical model. Notably, most of the metabolites showing the strongest relations with phenylalanine production seemed to hardly change in time. Apparently, potential bottlenecks in phenylalanine seem to hardly change in the course of a batch fermentation. The approach described in this study is not limited to longitudinal microbial studies but can also be applied to other (biological) studies in which similar longitudinal data need to be analyzed.

Large-scale human metabolomics studies: a strategy for data (pre-) processing and validation.

A large metabolomics study was performed on 600 plasma samples taken at four time points before and after a single intake of a high fat test meal by obese and lean subjects. All samples were analyzed by a liquid chromatography-mass spectrometry (LC-MS) lipidomic method for metabolic profiling. A pragmatic approach combining several well-established statistical methods was developed for processing this large data set in order to detect small differences in metabolic profiles in combination with a large biological variation. Such metabolomics studies require a careful analytical and statistical protocol. The strategy included data preprocessing, data analysis, and validation of statistical models. After several data preprocessing steps, partial least-squares discriminant analysis (PLS-DA) was used for finding biomarkers. To validate the found biomarkers statistically, the PLS-DA models were validated by means of a permutation test, biomarker models, and noninformative models. Univariate plots of potential biomarkers were used to obtain insight in up- or downregulation. The strategy proposed proved to be applicable for dealing with large-scale human metabolomics studies.

Assessing the performance of statistical validation tools for megavariate metabolomics data.

Statistical model validation tools such as cross-validation, jack-knifing model parameters and permutation tests are meant to obtain an objective assessment of the performance and stability of a statistical model. However, little is known about the performance of these tools for megavariate data sets, having, for instance, a number of variables larger than 10 times the number of subjects. The performance is assessed for megavariate metabolomics data, but the conclusions also carry over to proteomics, transcriptomics and many other research areas. Partial least squares discriminant analyses models were built for several LC-MS lipidomic training data sets of various numbers of lean and obese subjects. The training data sets were compared on their modelling performance and their predictability using a 10-fold cross-validation, a permutation test, and test data sets. A wide range of cross-validation error rates was found (from 7.5% to 16.3% for the largest trainings set and from 0% to 60% for the smallest training set) and the error rate increased when the number of subjects decreased. The test error rates varied from 5% to 50%. The smaller the number of subjects compared to the number of variables, the less the outcome of validation tools such as cross-validation, jack-knifing model parameters and permutation tests can be trusted. The result depends crucially on the specific sample of subjects that is used for modelling. The validation tools cannot be used as warning mechanism for problems due to sample size or to representativity of the sampling.

In search of secreted protein biomarkers for the anti-inflammatory effect of beta2-adrenergic receptor agonists: application of DIGE technology in combination with multivariate and univariate data analysis tools.

Two-dimensional difference gel electrophoresis (DIGE) in combination with univariate (Student's t-test) and multivariate data analysis, principal component analysis (PCA) and partial least squares discriminant analysis (PLS-DA) were used to study the anti-inflammatory effects of the beta(2)-adrenergic receptor (beta(2)-AR) agonist zilpaterol. U937 macrophages were exposed to the endotoxin lipopolysaccharide (LPS) to induce an inflammatory reaction, which was inhibited by the addition of zilpaterol (LZ). This inhibition was counteracted by addition of the beta(2)-AR antagonist propranolol (LZP). The extracellular proteome of the U937 cells induced by the three treatments were examined by DIGE. PCA was used as an explorative tool to investigate the clustering of the proteome dataset. Using this tool, the dataset obtained from cells treated with LPS and LZP were separated from those obtained from LZ treated cells. PLS-DA, a multivariate data analysis tool that also takes correlations between protein spots and class assignment into account, correctly classified the different extracellular proteomes and showed that many proteins were differentially expressed between the proteome of inflamed cells (LPS and LZP) and cells in which the inflammatory response was inhibited (LZ). The Student's t-test revealed 8 potential protein biomarkers, each of which was expressed at a similar level in the LPS and LZP treated cells, but differently expressed in the LZ treated cells. Two of the identified proteins, macrophage inflammatory protein-1beta (MIP-1beta) and macrophage inflammatory protein-1alpha (MIP-1alpha) are known secreted proteins. The inhibition of MIP-1beta by zilpaterol and the involvement of the beta(2)-AR and cAMP were confirmed using a specific immunoassay.

Fusion of mass spectrometry-based metabolomics data.

A general method is presented for combining mass spectrometry-based metabolomics data. Such data are becoming more and more abundant, and proper tools for fusing these types of data sets are needed. Fusion of metabolomics data leads to a comprehensive view on the metabolome of an organism or biological system. The ideas presented draw upon established techniques in data analysis. Hence, they are also widely applicable to other types of X-omics data provided there is a proper pretreatment of the data. These issues are discussed using a real-life metabolomics data set from a microbial fermentation process.