Description: Highlights • Fatty acids preserved in an Oligocene whale bone were analysed. • The fatty acid content of the fossil was in the permil range vs. a recent whale vertebra. • Ca. 80% of the n-C16 and n-C18 alkyl moieties were extractable, ca. 20% being bound to kerogen. • Endogenous fatty acids were largely of microbial origin (sulfate reducers, actinobacteria). Abstract The taphonomic and diagenetic processes by which organic substances are preserved in animal remains are not completely known and the originality of putative metazoan biomolecules in fossil samples is a matter of scientific discussion. Here we report on biomarker information preserved in a fossil whale bone from an Oligocene phosphatic limestone (El Cien Fm., Mexico), with a focus on fatty acyl compounds. Extracts were quantitatively analysed using gas chromatography–mass spectrometry (GC–MS) and, to identify macromolecular-linked remains, demineralised extraction residues were subjected to catalytic hydropyrolysis (HyPy). To better recognise potential authentic (i.e. animal-derived) lipids, the data from the ancient bone were compared with those obtained from (i) the adjacent host sediment of the fossil and (ii) a recent whale (Phocoena phocoena) vertebra. In addition, the spatial distribution of organic and inorganic species was observed at the μm level by imaging MS (time-of-flight-secondary ion mass spectrometry, ToF-SIMS). Our results revealed a rather even distribution of hydrocarbon-, O- and N-containing ions in the trabecular network of the ancient bone. A different, more patchy arrangement of organic compounds was evident in the former marrow cavities that were partly cemented by clotted micrites of putative microbial origin. The concentration of fatty acids (FAs) in the ancient bone was in the permil range of the amount extracted from the recent whale vertebra. Endogenous compounds, including monoenoic n-C16 and n-C18 as well as branched FAs, were identified in the fossil bone by comparison with the host sediment. Ca. 80% of the prevalent n-C16 and n-C18 moieties in the ancient bone were extractable as FAs, whereas ca. 20% were covalently bound in the non-saponifiable kerogen fraction. Ample pyrite precipitates, distinctive 10-methyl branched FAs and microbial microborings (“tunneling”) indicate that sulfate reducers and collagen-degrading actinomycetes were central players in the microbial decomposition of the bone. Similarities with reported microbial FA patterns suggest that the FAs in the fossil bone were largely contributed by these microbial “last eaters”. The results highlight some of the degradation and preservation mechanisms during marine FA diagenesis in the “natural laboratory” of bones, and therefore the processes that lead to either degradation, preservation, or introduction of these widespread biomolecules in the fossils of ancient marine animals.

Description: Domoic acid, a neurotoxin to vertebrates predominantly produced by the diatom Pseudo-nitzschia has been suggested to serve as an organic ligand. By binding iron and copper, it could increase their solubility and bioavailability. Domoic acid has to be released by the cells to serve this function and thus occur dissolved in sea water. Samples were pre-concentrated and desalted using solid-phase extraction, a procedure commonly applied for dissolved organic matter. Dissolved domoic acid was quantified in the East Atlantic, where it occurred ubiquitously, especially in the ocean surface. The maximum domoic acid concentration measured was 173 pmol L-1 and the average domoic acid carbon yield was 7.7 ppm. Both, carbon yield and dissolved domoic acid concentration, decreased with increasing water depth. Samples were taken during the cruise PS73 (ANT-XXV) on RV Polarstern. The extraction efficiency of domoic acid was 91%. The detection limit for solid-phase extractable domoic acid (DA-SPE) was 10 pmol L-1 and limit of quantification was 26 pmol L-1. Domoic acid concentrations below the limit of detection are marked as 〈LOD and concentrations below limit of quantification are marked as 〈LOQ in the data set.

Description: The widespread diatom Pseudo-nitzschia can produce domoic acid (DA). DA is a compound with well described neurotoxic effects on vertebrates including humans known as amnesic shellfish poisoning (ASP) syndrome. It has also been suggested to serve as an organic ligand that binds to iron and copper. By binding these trace elements, DA may increase their solubility and bioavailability. In order to serve this function, DA has to be excreted and reabsorbed by the cells. Only few records of dissolved domoic acid (dDA) concentrations in the ocean exist. To accomplish quantification by ultra performance liquid chromatography (UPLC), samples have to be pre-concentrated and desalted using solid-phase extraction, a procedure commonly applied for dissolved organic matter. Our major goals were to quantify dDA in a basin-wide assessment in the East Atlantic Ocean, to determine extraction efficiencies for complexed and uncomplexed dDA, and to assess whether domoic acid is represented by its molecular formula in direct-infusion high resolution mass spectrometry. Our results showed that dDA was extracted almost quantitatively and occurred ubiquitously in the ocean surface but also in deeper (and older) water, indicating surprisingly high stability in seawater. The maximum concentration measured was 173 pmol L−1 and the average molar dDA carbon yield was 7.7 ppm. Both carbon yield and dDA concentration decreased with increasing water depth. Providing quantification of dDA in the water column, we seek to improve our understanding of toxic bloom dynamics and the mechanistic understanding of DA production.

Description: The second Data Science Symposium at AWI gathered several data science related talks from AWI, GEOMAR and HZG.

Description: Rational: High-resolution mass spectrometry (HRMS) with high sample throughput has become an important analytical tool for the analysis of highly complex samples and data processing has become a major challenge for the user community. Evaluating direct-infusion HRMS data without automated tools for batch processing can be a time consuming step in the analytical pipeline. Therefore, we developed a new browser-based software tool for processing HRMS data. Methods: The software named UltraMassExplorer (UME) was written in the R programming language using the shiny library to build the graphical user interface. The performance of the integrated formula library search algorithm was tested using HRMS data derived from analyses of up to 50 extracts of marine dissolved organic matter. Results: The software supports the processing of lists of calibrated masses of neutral, protonated, or deprotonated molecules, respectively, with masses of up to 700 Da and a mass accuracy 〈 3 ppm. In the performance test, the number of assigned peaks per second increased with number of submitted peaks and reached a maximum rate 4,745 assigned peaks per second. Conclusions: UME offers a complete data evaluation pipeline comprising a fast molecular formula assignment algorithm allowing for the swift reanalysis of complete datasets, advanced filter functions, and the export of data, metadata, and publication-quality graphics. Unique to UME is a fast and interactive connection between data and its visual representation. UME provides a new platform enabling an increased transparency, customization, documentation and comparability of datasets.

Description: In the evaluation of high-resolution mass spectrometric data a considerable amount of time and computational power can be spent on matching molecular formulas to the neutral mass of measured ions. During the evaluation of multiple samples using the classical combinatory approach based on molecular building blocks and nested loops, the time consuming step of calculating the molecular mass may be repeated for the same molecular formula multiple times. To avoid repetitive calculations, we implemented a formula library based search approach into our data evaluation pipeline. In our approach, the step of calculating molecular formulas and corresponding masses is limited to the process of building a library. The library calculation requires an a priori definition of the maximum molecular mass and the isotopes contained, e.g. formulas in the mass range of ≤ 650 Da consisting of 12C, 1H, 14N, 16O, 31P, 32S, 13C, and 34S. The subsequent matching process is based on scrolling through a mass-sorted formula library and comparison with a mass-sorted list of measured peaks. The time required for processing is primarily a function of the size of the formula library. Consequently, at constant library size, the matching algorithm becomes more efficient with increasing number of supplied peaks (up to 4700 formula assignments s-1 on a standard workstation) and is thus particularly suited for processing large datasets. We implemented the matching algorithm into our R Shiny based interactive, evaluation software UltraMassExplorer (UME). In combination with the graphical user interface of UME, our algorithm provides the basis for fast and reproducible (re-)analysis of complete sample sets with currently up to 400,000 peaks in a user friendly, integrated environment. The code of our open-source algorithm is available through the UME website [1]. References [1] www.awi.de/en/ume

Description: Rationale High-resolution mass spectrometry (HRMS) with high sample throughput has become an important analytical tool for the analysis of highly complex samples and data processing has become a major challenge for the user community. Evaluating direct-infusion HRMS data without automated tools for batch processing can be a time-consuming step in the analytical pipeline. Therefore, we developed a new browser-based software tool for processing HRMS data. Methods The software, named UltraMassExplorer (UME), was written in the R programming language using the shiny library to build the graphical user interface. The performance of the integrated formula library search algorithm was tested using HRMS data derived from analyses of up to 50 extracts of marine dissolved organic matter. Results The software supports the processing of lists of calibrated masses of neutral, protonated or deprotonated molecules, with masses of up to 700 Da and a mass accuracy 〈3 ppm. In the performance test, the number of assigned peaks per second increased with the number of submitted peaks and reached a maximum rate of 4745 assigned peaks per second. Conclusions UME offers a complete data evaluation pipeline comprising a fast molecular formula assignment algorithm allowing for the swift reanalysis of complete datasets, advanced filter functions and the export of data, metadata and publication-quality graphics. Unique to UME is a fast and interactive connection between data and their visual representation. UME provides a new platform enabling an increased transparency, customization, documentation and comparability of datasets.