Metabolomics Circle 2017 — Abstracts

Medical Research Journal
2017 Medical Research Journal  
The implementation of scientific knowledge, as a part of applied science, should be aimed at developing practical applications for many fields of natural world (e.g. medicine, pharmacy, cosmetology, food industry etc.). Consequently, in the case of separation techniques the evaluation of separation medium comprises an essential part of analytical methods improvement. Significant efforts have been made over the last few years to achieve stationary phases imitated natural matter (e.g. biological
more » ... embrane) as well as endogenous compounds (e.g. amino acids). Specificity of chemically bonded ligands in the case of new materials enables receiving a so-called dedicated stationary phases. Moreover, structural similarity of the immobilized ligands with the desired group of analytes determine the high specificity and selectivity of prepared stationary phases. Therefore, the investigations in accordance with the "3S" assumption -similarity, selectivity, and specificity -allow the development of a new generation of separation materials. The chemical immobilization of amino acids and peptides -one of the most essential compounds in life science allow the preparation of materials that exhibit unique interactions with amino acids and peptides as a analytes and beyond. The amino acids-and peptides-silica stationary phases show high selectivity for other groups of biologically significant compounds, i.e. carbohydrates, nucleosides, flavonoids, optical active compounds. Furthermore, stationary phases prepared in compliance with "3S" assumption exhibit wide range of applicability in separation techniques (RP HPLC, HILIC, IC). This approach also includes the preparation of stationary phases containing in the structure specific functional groups characteristics for biological membrane. The development of this type of materials enables the modeling of the transport of substances through the biomembrane, e.g. blood/brain barrier. On the other side, stationary phases imitating membrane lipids provide unique selectivity according to lipids separation, especially phospholipids. As a consequence, the natural biological systems provide a significant center of inspiration for development of chromatographic methods. Based on the principles that governed natural world, it is possible to obtain desired similarity, selectivity, and specificity that determine the resolution of separation methods. Metabolic phenotyping provides insight into biochemical pathways associated with disease pathology and thus offers great potential for diagnostic and prognostic applications. The challenge remains in translating metabolic phenotyping approaches into tangible clinical applications. The success of this translation largely depends upon the capacity of analytical platforms to produce accurate and reproducible data in a high-throughput manner. This presentation will describe an NMR-based pipeline developed and optimised at the MRC-NIHR National Phenome Centre (NPC), in Imperial College London (UK) for the analysis of large cohorts of human urine and blood plasma and serum samples [1]. The robustness and reproducibility of the pipeline has been tested through integration and combination of more than 8000 urine samples collected from 7 independent studies acquired over 4 years. I will also describe recently developed approaches for recovery of quantitative lipoprotein data from 1 H NMR profiles of plasma and serum samples [2] that have been assessed in a multi-laboratory, multi-spectrometer ring test trial. These results show perfect compliance with the National Cholesterol Educational Program, NCEP, criteria for lipid analysis indicating great potential for implementation of lipoprotein analysis by NMR in clinical settings. Finally, application of these approaches to heterogeneous multifactorial diseases including prediction of pneumonia in critical care ventilated patients [3] and treatment of lower urinary tract symptoms in women [4] will be presented. References 1. The development in the end of the XX-th century of new proteomic methodology, allowing to analyze complex protein sets, brought hope for its fast application for the search for new biomarkers of a variety of diseases, including cancer. Hopes were the highest in case of protein content of human body fluids and the application of newly detected biomarkers for diagnosis, prognosis and therapy monitoring in cancer. Initial, very optimistic results of the search for new proteomics-based biomarkers in oncology, were however verified negatively quite fast, and it was pointed out that the published results were not reproducible and the reasons for that sparked a more general discussion [1]. The necessity for further improvements and more systematic basic studies preceding the application in clinics was indicated. Guidelines were formulated for plasma/serum preparative procedures as well as for proteomic procedures especially in their quantitative aspects. The most important breakthrough was the development of a new measurement method for targeted quantitative measurements of proteins, named Multiple Reaction Monitoring (MRM). It allows for the measurement of absolute values of protein concentration in the background of extremely complex protein mixtures (e.g. blood plasma/serum) using a coupled liquid chromatography -mass spectrometry (LC-MS) analysis. MRM method was named the Method of the Year 2012 by Nature Methods [2]. Moreover, in this method, a panel of 10-100 proteins, represented by hundreds of peptides, their proteolytic fragments, can be subjected to a parallel analysis in the same experiment. The foundations of the MRM method will be presented illustrated by a few applications. The applications include the assessment of the value of MRM-based cytokeratin peptide panels for diagnosis of the etiology of pleural fluid. The metabolomic data collected using high performance liquid chromatography coupled with mass spectrometry (LC-MS/MS or LC/TOF-MS) can provide signal intensities for a large number of compounds (peaks) present in set of samples (i.e. control and experimental). However, such large metabolomic data sets are usually difficult to interpret, due to the small sample sizes; search for effects (e.g. differences between control and experimental group) that are small; and considering a large number of hypotheses without paying enough attention to the problem of multiple comparisons. It might lead and often leads to conclusions (e.g. biomarker identification) that are exaggerated or have a wrong sign! The hierarchical modeling and Bayesian inference method can largely eliminate such problems and consequently limit the number of false positive results. Such a statistical models can be build using the existing theories, such as pharmacokinetics, and prior information available for the particular problem. The inferences are based on the Markov Chain Monte Carlo methods (JAGS/STAN) and can be run using many popular computing environments like R/Matlab. The lecture will discuss the analysis of metabolic data obtained from urine samples collected in two groups of rats (control and tumor-induced) at four time points. Low-dose computed tomography (LD-CT) screening in a high-risk group is a potential strategy for early detection of lung cancer. Pre-selection of candidates for LD-CT by using the blood-based biomarkers would lower the overall costs and benefit the patients by increasing the specificity of diagnosis. In the following study, we searched for a molecular signature based on a serum lipidome and metabolome profile distinguishing individuals with early lung cancer from healthy participants of the lung cancer screening program. Classification/discrimination models have found many attractive applications in different fields of science, including diagnostic based on metabolomic profiles. Both type of models differ with respect to assignment rules, but using physico-chemical description of a sample they aim to recognize group label of a given sample. The construction of logic classification/discrimination rules, enabling correct recognition of group labels, requires considering a few key aspects. The first one is concerned with the identification of a representative samples (a model set) used to construct a model. It should include samples that represent all present and expected sources of variability. The second aspect is related to the selection of the optimal number of components or logic rules. To simplify model and facilitate its further interpretation, elimination of irrelevant variables is carried out. Regardless Insulin is an anabolic peptide hormone with a systemic effect. The most important stimulus for its production is the postprandial glucose uptake in the blood. Influencing the effector cells (myocytes, adipocytes, hepatocytes) increases the glucose transport to the inside of the body, reducing its levels in the blood. In addition, it controls a number of mechanisms, mainly hormonal. Our ancient genes are not prepared for nutritional changes that have occurred in the last hundreds of years, and therefore the conditions of insulin resistance and hyperinsulinemia result in non-diabetic hormone abnormalities. Hyperinsulinemia is associated with many hormonal disturbances which destructively affect homeostasis of the body. The consequence of this condition is increased lipogenesis with blockade of lipolysis, resulting in excessive accumulation of visceral adipose tissue acting pro-inflammatory. Increased cortisol levels, the same adrenergic activity and the maintenance of hyperinsulinemia by incereasing of blood sugar. Lipid disorders, hyperlipidemia, are consequence of excessive carbohydrate intake due to recurrent hypoglycaemia caused by excess insulin. Insulin interferes with the daily rhythm by blocking the melatonin, thus insufficient somatostatin discharge impairs the body's regenerative capacity. It has a bearing on the chronic feeling of fatigue. DHEA secretion also causes blockade of sex hormones (estrogens, progesterone) and fertility disorders. Hyperinsulinemia is associated with thromboembolic complications. Given the multidirectional and comprehensive nature of insulin not only for carbohydrate, it is important to keep its normal values, to regulate the whole body's hormones.
doi:10.5603/mrj.2017.0009 fatcat:decttydbpjeg3k2bss4szlx3pi