Spatially and temporally resolved metabolome of the oral human cavity, positive ionization mode, batch 1
Description
BACKGROUD: Saliva samples (#159) from 20 healthy volunteers, including ten males and ten females was analyzed using a semi-targeted LC-MS/MS platform. SAMPLE NAMES: Structure: [polarity]_[batch]_[sample type]_[subject id]_[time of collection]_[oral location]_[injection number]. explanation: Polarity (neg = negative, pos = positive), batch (1, 2), sample type (blank, QC, sample), time (M = morning, A = afternoon, E = evening), oral location (AT = above the tongue, BT = below the tongue, CK = cheek) Note: When subject id, time of collection, and oral location for blank and QC sample type are equal to 0. RESULTS: Many the metabolites that exhibit heterogeneous distribution, several important metabolites are linked to oral health, such as N-acetylated metabolites, L-ornithine, saccharides, and guanosine monophosphate. Next to the spatial distribution of metabolites, we identified known and novel time dependencies. Among the novel diurnal metabolites, N-acetyl methionine, N,N,N-trimethyl lysine, and N-acetyl tryptophan revealed diurnal patterns in all oral locations. Where the former two behave similarly to cortisol, and the latter has its nadir in the M. Moreover, a pool of amino acids and mono-phosphorylated nucleotides manifested a significant diurnal rhythm exclusively in BT.
Files
Steps to reproduce
Spectra deconvolution, peak alignment, and compound identification were performed using MS-DIAL version 4.90, see MSDIAL settings. Metabolite identification was achieved by matching exact precursor mass, retention time, adduct formation, and fragmentation spectra with an in-house LC-MS/MS spectral library of authentic compounds analyzed under identical experimental conditions. Data filtering involved the exclusion of unknown features, missing value filtering by removing metabolites with signals lower than the blank average in more than 66% of the samples, and RSD filtering by removing any compound with RSD higher than 30% as obtained for QCpool samples. Subsequently, all adducts belonging to the same metabolite were summed up, and for metabolites identified in both ionization modes, the ion form with the higher RSD was excluded.