Following established guidelines and practice recommendations, manual annotation of neutrophil metabolites was performed using in house tools (tameNMR; https://github.com/PGB-LIV/tameNMR) and profiling software NMR Procflow v1.4 (https://nmrprocflow.org/) [DOI: 10.1007/s11306-017-1178-y], Metabolite peaks were then compared against several open-access metabolite databases, including HMDB (http://www.hmdb.ca/) [DOI: 10.1093/nar/gkab1062], Biological Magnetic Resonance dataBank (BMRB; bmrb.io) [DOI:10.1093/nar/gkac1050], Chenomx library (Chenomx NMR Suite v8.2), in-house reference library, and prior publications on intracellular neutrophil metabolomics. Then, annotation accuracy and consistency were ensured through guidance from an expert spectroscopist. Moreover, biological context was considered crucial when assigning identities to previously unidentified peaks.
Spectral binning process defining left and right boundaries for each peak multiplet was employed to select and integrate peaks and peak multiplets accommodating minimal chemical misalignments of the computed area under the peak in each aligned spectra. Where possible peaks were annotated to known metabolites or marked as unknown. A representative peak bin for each annotated metabolite was selected via correlation reliability score ‘CRS’ filtering (Grosman 2019) through comparison of Pearson correlation matrices using R v4.2. This approach yielded a single representative bucket for each annotated metabolite reducing the number of variables but also removing unknown NMR peaks. As such analysis of the fully integrated spectra (containing information on known and unknown metabolite peaks) as well as the filtered spectra containing only representative peaks from all annotated metabolites present were both compared in subsequent spectral analysis. All data acquired, processing steps, parameters and metabolite annotations and CRS representative peak selections are deposited in the European Bioinformatics Institute (EBI) open repository for metabolomics (www.ebi.ac.uk/metabolights) [DOI: 10.1093/nar/gkad1045].
"],"repository":["MetaboLights"],"study_status":["Public"],"ptm_modification":[""],"instrument_platform":["Nuclear Magnetic Resonance (NMR) -"],"publication":["An NMR Metabolomics Analysis Pipeline for Human Neutrophil Samples with Limited Source Material. 10.3390/metabo15090612. PMID:41002996"],"nmr_spectroscopy_protocol":["Samples were acquired on a Bruker 700MHz Avance IIIHD equipped with triple resonance TCI cryoprobe and chilled SampleJet autosampler. Spectrometer quality assurance was completed daily via temperature calibration to 25 °C (with margin error of 0.1 °C) by a methanol thermometer (cat number Z10627 99.8% Methanol-d4, 5 mm, Bruker, UK) [Findeisen [https://doi.org/10.1002/mrc.1941] and 3D shimming on Bruker standard 2 mM sucrose (cat number Z10902 2; mM Sucrose 0.5 mM DSS 2 mM NaN3 in H2O/D2O 90/10, 5 mm Bruker, UK) to ensure linewidth half height of DSS reference peak is within acceptance criteria (<1 Hz).
"],"submitter_affiliation":["University of Liverpool"],"submitter_name":["Marie Margaret Phelan"],"organism_part":["neutrophil"],"technology_type":["NMR spectroscopy"],"disease":[""],"extraction_protocol":["Neutrophil metabolites were extracted using an established method. A mixture of 50:50 v/v ice cold HPLC grade acetonitrile:water (ddH2O) was added to each sample (500 μL per cell pellet), followed by a 10 min incubation on ice. Samples were sonicated three times for 30 sec at 23 kHz and 10 μm amplitude using an exponential probe with 30 sec rest in between of sonication in an ice water bath. Sonicated samples were centrifuged at 12,000 g for 5 mins at 4 °C, followed by transfer of the supernatants to cryovials and flash frozen in liquid nitrogen prior to lyophilisation. All lyophilised samples were stored at -80 °C prior to analysis in spectrometer.
"],"organism":["Homo sapiens"],"data_transformation_protocol":["All acquired spectra were automatically phased referenced to 3-(trimethylsilyl) propionic-2,2,3,3-d4 acid sodium salt (deuterated trimethylsilyl propionate; TSP-d4) at 0 ppm, phasing and baseline corrected using automated macro (apk0.noe) provided within TopSpin v4.4.1 (Bruker Corporation) followed by compliance check to minimum quality criteria outlined by the Metabolomics Standards Initiative (MSI) to suppress the macromolecule signals, selectively accentuate small molecule metabolite signals, ensure consistent linewidths, baseline corrections, and water suppression. All spectra were thoroughly evaluated to ensure they met the current best practice of Metabolomics Society, ensuring a robust consistency in phasing, baseline correction, peak alignment, chemical shift reference to TSP-d4 at 0 ppm, line width, and water suppression conditions for neutrophil intracellular metabolomic reporting. Only technical replicates with the lowest line width at half height from each sample which passed quality control process including the flat baseline, consistent line widths, and water suppression consistency were selected. Any spectra that did not meet the requirements of recommended reporting standards were subsequently removed from further processing.
"],"study_factor":["Number of cells","Method","Number of scans"],"metabolights_link":["https://www.ebi.ac.uk/metabolights/MTBLS6220"],"submitter_email":["mphelan@liv.ac.uk"],"sample_collection_protocol":["Human peripheral blood samples were obtained by venipuncture into lithium-heparin vacutainers. Neutrophils were isolated immediately in order to maintain viability. Two different neutrophil isolation protocols were compared.
For the first 'Ficoll-Paque' method, whole blood was mixed with HetaSep (STEMCELL Technologies) at a dilution ratio of 5:1 and incubated at 37 °C for 30 minutes to allow sedimentation of erythrocytes. The nucleated cells were then layered onto Ficoll Paque Plus solution (Merck) in 1:1 ratio and centrifuged for 30 min at 500g. Contaminating erythrocytes were removed with hypotonic ammonium chloride lysis buffer and cells were resuspended in RPMI-1640 media without supplementation of Phenol Red buffer and HEPES. Cellular viability was assessed with Trypan blue (Merck) exclusion method in Neubauer haemocytometer. Neutrophil purity was confirmed with Wright Giemsa staining and morphology. Only samples with ≥95% viability and purity were eligible for further experiments.
In the second 'Beads' method, ultrapure neutrophils (>99.9% purity) were obtained using a negative selection kit for neutrophil isolation, employing magnetic bead separation (STEMCELL Technologies). For this method, whole blood was mixed with HetaSep (STEMCELL Technologies) at a dilution ratio of 5:1, and incubated at 37 °C for 30 minutes to allow sedimentation of erythrocytes. The nucleated cells were then mixed with an equal volume of ice-cold isolation buffer (PBS, 2% BSA, 100 μL 0.2 M EDTA). The suspension was centrifuged at 400g for 5 minutes, and the pellet was resuspended in 2 mL of isolation buffer. An antibody cocktail (25 μL) containing tetrameric antibody complexes recognising cell markers of unwanted leukocytes and erythrocytes and magnetic particles was added to the suspension and incubated on ice for 5 minutes. Following this, 50 μL of magnetic beads solution was added for 5 minutes, before placing in a magnetic chamber for 3 minutes. The unlabelled neutrophils were decanted into a clean tube and resuspended in HEPES-free and Phenol Red-free RPMI 1640 media. Cellular viability was assessed with Trypan blue (Merck) exclusion method in Neubauer haemocytometer. Neutrophil purity was confirmed at >99% with Wright Giemsa staining and morphology.
Neutrophils were pelleted by centrifugation at 1,000 g at 25 °C for 2 min and supernatant was aspirated. Cell pellets were heated at 100 °C for 1 min prior to being snap-frozen using liquid nitrogen and stored at -80 °C for further analysis.
"],"nmr_assay_protocol":["Spectra was acquired using standard (vendor supplied) 1D 1H Carr-Purcell-Meiboom-Gill (CPMG; cpmgpr1d) spectra with 256, 512, 1024 or 2048 transients as specified, a 15-ppm spectral width, 32K points, 9.6 ms echo time, 3.1 s acquisition time and 4 s interscan delay. Additionally, 1D 1H NOESY pre-saturation experiments (noesypr1d) were performed to verify shimming quality and ensure effective water suppression for each sample during data acquisition. This step was essential to maintain spectral clarity and consistency across samples.
Lyophilised metabolite pellets were resuspended in 200 μL of 100 mM deuterated sodium phosphate buffer pH 7.4 with 100 μM of TSP-d4 and 0.05% NaN3. All samples were vortexed for 20 sec and centrifuged at 12,000 g for 1 min at 20 °C, prior transferring into 3 mm (outer diameter) NMR tubes for acquisition using 700 MHz Avance IIIHD Bruker NMR spectrometer equipped with TCI cryoprobe and chilled SampleJetTM autosampler.
"],"metabolite_name":["L-Arginine","Taurine","Propylene glycol","Indolelactic acid","L-Phenylalanine","Diethylhexyl adipate","ADP","L-Isoleucine","L-Asparagine","unknown","L-Homoserine","Myoinositol","L-Threonine","Formic acid","L-Cysteine","L-Alanine","Pyroglutamic acid","Dimethylamine","Saccharopine","NADP","L-Histidine","Adenosine triphosphate","Imidazolepropionic acid","Glutathione","L-Glutamic acid","L-Leucine","Propionic acid","Glycerol","Phosphorylcholine","Acetamide","2-Hydroxy-3-methylpentanoic acid","L-Proline","NAD","Isopropyl alcohol","L-Methionine","Guanosine triphosphate","L-Glutamine","Acetaminophen","Choline","L-Serine","L-Aspartic acid","D-Lactic acid","L-Lysine","Acetic acid","L-Valine","Acetone"],"pubmed_abstract":["Background/Objectives: Untargeted 1H NMR metabolomics is a robust and reproducible approach used to study the metabolism in biological samples, providing unprecedented insight into altered cellular processes associated with human diseases. Metabolomics is increasingly used alongside other techniques to detect an instantaneous altered cellular function, for example, the role of neutrophils in the inflammatory response. However, in some clinical settings, blood samples may be limited, restricting the amount of cellular material available for a metabolomic analysis. In this study, we wanted to establish an optimal 1D 1H NMR metabolomic pipeline for use with human neutrophil samples with low amounts of input material. Methods: We compared the effect of different neutrophil isolation protocols on metabolite profiles. We also compared the effect of the absolute cell counts (100,000 to 5,000,000) on the identities of metabolites that were detected with an increasing number of scans (NS) from 256 to 2048. Results/Conclusions: The variance in the neutrophil profile was equivalent between the isolation methods, and the choice of isolation method did not significantly alter the metabolite profile. The minimum number of cells required for the detection of neutrophil metabolites was 400,000 at an NS of 256 for the spectra acquired with a cryoprobe (700 MHz). Increasing the NS to 2048 increased metabolite detection at the very lowest cell counts (<400,000 neutrophils); however, this was associated with a significant increase in the analysis time, which would be rate-limiting for large studies. The application of a correlation-reliability-score-filtering method to the spectral bins preserved the essential discriminatory features of the PLS-DA models whilst improving the dataset robustness and analytical precision."],"pubmed_title":["An NMR Metabolomics Analysis Pipeline for Human Neutrophil Samples with Limited Source Material."],"pubmed_authors":["Filbertine Grace G, Abdullah Genna A GA, Gill Lucy L, Grosman Rudi R, Phelan Marie M MM, Chiewchengchol Direkrit D, Hirankarn Nattiya N, Wright Helen L HL"],"additional_accession":[]},"is_claimable":false,"name":"An NMR metabolomics analysis pipeline for human neutrophil samples with limited source material","description":"Untargeted 1H NMR metabolomics is a robust and reproducible approach to study metabolism in biological samples, providing unprecedented insight into altered cellular processes associated with human diseases. Metabolomics is increasingly used alongside other -omics techniques, such as proteomics and transcriptomics, to detect instantaneous altered cellular function, for example the role of blood neutrophils in the inflammatory response. However, in some clinical settings e.g. pediatrics research, blood samples may be limited, restricting the amount of cellular material available for metabolomics analysis. In this study we wanted to establish the optimal 1D 1H NMR metabolomics pipeline for use with human neutrophil samples with low input material. We compared the effect of neutrophil isolation protocols using Ficoll-Paque or negative selection using antibody-labelled magnetic beads on neutrophil metabolite profiles. We also compared the effect of absolute cell counts ranging from 100,000 to 5 million neutrophils in total on the identities of metabolites that can be detected with increasing number of scans (NS) from 256 to 2,048. We found that variance in neutrophil profile was higher with N isolation method event though choice of isolation method did not significantly alter the metabolite profile of human neutrophils measured by 1D 1H NMR spectroscopy. The minimum number of cells required for detection of human metabolites was 400,000 at NS 256 for spectra acquired with a cryoprobe equipped 700MHz. Increasing the NS to 2,048 increased metabolite detection at the very lowest cell counts (<400,000 neutrophils), however this was associated with a significant increase in analysis time which would be rate-limiting for large studies. Application of a correlation reliability score (CRS) filtering method to the spectral bins preserved the essential discriminatory features of the PLS-DA models whilst improving the dataset robustness and analytical precision.","dates":{"publication":"2025-11-06","submission":"2025-08-12"},"accession":"MTBLS6220","cross_references":{"HMDB":["HMDB0000317","HMDB0000172","HMDB0000687","HMDB0000883","HMDB0000237","HMDB0001881","HMDB0000863","HMDB0001311","HMDB0000167","HMDB0000161","HMDB0040270","HMDB0000182","HMDB0000042","HMDB0031645","HMDB0000719","HMDB0000148","HMDB0000641","HMDB0001659","HMDB0000267","HMDB0000125","HMDB0000696","HMDB0000191","HMDB0000087","HMDB0000168","HMDB0000279","HMDB0000159","HMDB0000097","HMDB0001565","HMDB0000517","HMDB0000211","HMDB0000251","HMDB0000131","HMDB0000187","HMDB0000574","HMDB0000162","HMDB0001273","HMDB0000538","HMDB0002271","HMDB0001859","HMDB0000177","HMDB0000671","HMDB0000217","HMDB0000902","HMDB0001341","HMDB0000142"],"pubmed":["41002996"]}}