Variability in fasting human venous plasma metabolomic profiles with tourniquet-induced hemostasis
Study design and inclusion criteria
Apparently healthy female subjects (aged 18-60) volunteering to participate in the study were recruited from the Division of Nutrition, St. John’s Research Institute, St. John’s Medical College and Hospital, Bangalore. The exclusion criteria were: age outside the range of 18 to 60 years, refusal to participate in the study, participation in other studies, positive test for hepatitis (HBsAg) or HIV, need for chronic or daily medical treatment (connective tissue diseases, inflammatory bowel disease, active tuberculosis, symptomatic heart disease), had serious pre-existing clinical conditions.
Ethical approval and informed consent
Ethical approval for the study was sought and obtained from the Institutional Ethics Committee of St. John’s Medical College and Hospital, Bangalore. The study protocol was explained in the language understood by the participants. The informed and signed consent of the participants was obtained during recruitment. All relevant guidelines and regulations were followed when performing the study protocol.
Sociodemographic data and medical history
Well-structured questionnaires were provided to participants and explained by study staff trained to obtain detailed socio-demographic data. Information on recent medical history (3 months prior to recruitment) of participants and any medications and nutritional supplements currently taken by them was obtained and duly recorded.
Anthropometric measurements and vital signs
A digital scale, with an accuracy of 0.1 kg, was used to weigh the subjects, in minimal clothing, while the subjects’ height was recorded to the nearest 0.1 cm. Blood pressure and pulse measurement and recording were appropriately performed by trained study personnel.
Sample collection and clinical chemistry
An overnight fasting blood sample (11-12 hrs fasting) (10 mL) was drawn on EDTA (BD Vacutainer®, Becton, Dickinson and Company, Franklin Lakes, NJ) tubes between ?? 8:30 am and 9:30 am by venipuncture of the arm (antecubital vein). Four hours of collection [T1, T2, T4: 1, 2 and 4 min respectively, after tying the tourniquet, and NT (no tourniquet): collection 5 min after removal of the tourniquet] were included for each study participant.
The samples were transferred immediately and stored in a cooler (
High Resolution Accurate Mass Data Analysis (HRAM)
Plasma samples were processed following a protocol similar to that described previously.25. Briefly, plasma samples (100 L) were spiked with an internal standard (IS) of a 2Mixture of H-labeled amino acids (20 L, 1 ng / mL; U-2Mixture of H-labeled amino acids> 97% purity; Cambridge Isotope Laboratories, Massachusetts, USA) and deproteinized using a refrigerated organic solvent (8: 1: 1, acetonitrile: methanol: acetone). Samples were vortexed and incubated at 4 ° C for 30 min before centrifugation at 20,000 rcf for 20 min in a refrigerated centrifuge (5810 R, Eppendorf, Eppendorf AG, Hamburg, Germany). The supernatants were dried at 40 ° C in a vacuum concentrator (Labconco, USA) and the dried extracts were reconstituted in acetonitrile / water (1: 1). Non-targeted metabolomics analysis was performed on a High Resolution Accurate Mass Platform (HRAM) consisting of an Ultra High Pressure Liquid Chromatograph (UHPLC, Thermo Scientific, Vanquish Flex Binary, Waltham, MA, USA) coupled to an orbitrap-based mass spectrometer (Q Exactive, Thermo Scientific, San José, USA). The mass spectrometer was calibrated using a positive ion calibration solution (Pierce LTQ Velos ESI Positive Ion Calibration Solution, ThermoFisher Scientific, Waltham, MA, USA) on a daily basis before starting a sequence. analytical consisting of solvent blanks, pooled quality control (QC). samples, which included six technical replicates of an aliquot pool derived from the study plasma samples26 and studying plasma samples. Separation of metabolites was performed using a Zorbax Eclipse plus-C18 (150 × 2.1 × 1.8 micron, Agilent Technologies, Santa Clara, CA, USA) column maintained at 40 ° C. The other method parameters related to LC-MS / MS are similar to those previously described.25.
Raw data files acquired using Xcalibur software27 (version 4.1, ThermoFisher Scientific, MA, USA) were initially processed using the Compound Discoverer software28 (version 188.8.131.525, ThermoFisher Scientific, Waltham, MA, USA) for positive polarity with untargeted metabolomic workflow as previously described25 to find and identify differences between samples. The workflow used the adaptive curve model with a maximum offset of 2 min, a mass tolerance of 5 ppm, and a threshold of 3 S / N (signal / noise) for the retention time alignment. Peak detection required a mass error of less than 5 ppm for ion chromatograms extracted with a minimum peak intensity of 1,000,000. [M + H]+1 was defined as the base ion taking into account the other adducts. Peaks should have a width at half height of less than 0.5 min and a minimum of 5 scans. The maximum number of elements for modeling the isotopic model was C90H190Brother3Cl4K2NOTtenN / A2oh15P3S5. All compounds detected were pooled in the samples with a mass error of 5 ppm and a retention time offset of 0.2 min. Missing peaks (not initially detected) in a given sample were determined using the Fill Gaps node algorithm with a mass error of 5 ppm and a S / N threshold of 1.5 with actual peak detection . The Fill Gaps node calculates the area of the missing chromatographic peaks as follows: match the detected ions based on m / z and retention time regardless of adduct assignment, re-detecting peaks at lower thresholds, simulating peaks as a function of expectations m / z, and imputation of the spectral noise based on the detection limit values. Additionally, a QC-based area correction is applied for instrument drift using the cubic spline regression model. Each compound had to be detected in at least 50% of QC tests with a relative standard deviation (RSD) of less than 30%. Identification of compounds was performed using mzCloud (ddMS2) and ChemSpider (exact formula or mass) and similarity searches for all compounds with ddMS2 data were performed using mzCloud and the mzLogic algorithm applied to classify the compounds. ChemSpider results. The preprocessed data was exported to .xlsx files for further statistical analysis.
Anthropometric and metabolic profile data were presented as mean ± SD and median with interquartile range (IQR). Metabolite abundance data were scaled using the Pareto scaling method29 and the variance of the metabolites were decomposed into inter- and intra-individual variabilities by a standard linear random effects model. The proportion of intra-individual variability over the total was assessed by intra-class correlation (ICC) for each metabolite. A chi-square test was performed to test whether the temporal variability is other than zero in the estimation algorithm of the random-effects model. We considered an acceptable false positive rate of at most 5%. Additionally, using the raw unscaled abundance (peak intensities) of metabolites, the interindividual variability of each metabolite at four time points of collection was derived by the change in coefficient (CV) calculated as the ratio of the median absolute deviation (MAD) between the median and the median. the crude abundance of the respective metabolites. R statistical software version 4.0.2 was used for the analysis30 (R Core Team, 2020, Vienna, Austria).