Investigating the Metabolome of Schistosoma mansoni by High-Resolution Mass Spectrometry Imaging Cumulative Dissertation by Patrik Kadesch prepared at the Institute of Inorganic and Analytical Chemistry for the Degree of Doctor rerum naturalium (Dr. rer. nat.) Justus Liebig University Giessen, 2021 2 Table of contents List of Abbreviations 4 List of Publications 5 Statement in Lieu of an Oath 6 Abstract 7 Chapter I – Synopsis 8 Introduction and Motivation 8 Mass Spectrometry 9 Mass Spectrometry Imaging 12 Lipid Assignment and Identification 14 Schistosomiasis 17 Mass Spectrometric Investigations of Lipids in Adult Schistosoma mansoni 18 Atmospheric Pressure Matrix Assisted Laser Desorption/Ionization Mass 19 Spectrometry Imaging of Biological Tissues (Publication 1) Mass Spectrometry Imaging of Adult Schistosoma mansoni Parasites 20 Mass Spectrometry Imaging – Characterizing the Spatial Distribution of Lipids in 22 Adult Schistosoma mansoni Parasites (Publication 2) Multivariate Statistical Analysis 22 Characterizing the Lipids on the Tegumental Surface of Adult Schistosomes 25 Compared to Inner Worm Tissues Conclusions and Future Perspectives 27 References 28 Chapter II - Publication 1 37 Lipid Topography in Schistosoma mansoni Cryosections, Revealed by 38 Microembedding and High-Resolution Atmospheric-Pressure Matrix-Assisted Laser Desorption/Ionization (MALDI) Mass Spectrometry Imaging Chapter III - Publication 2 55 Tissue- and sex-specific lipidomic analysis of Schistosoma mansoni using high- 56 resolution atmospheric pressure scanning microprobe matrix-assisted laser desorption/ionization mass spectrometry imaging Curriculum vitae 106 Acknowledgements 108 3 List of Abbreviations Abbreviation Explanation (AP-S)MALDI (Atmospheric Pressure Scanning Microprobe) Matrix Assisted Laser Desorption/Ionization (D)ESI (Desorption) Electrospray Ionization (HR)MS (High Resolution) Mass Spectrometry (L)PC (Lyso) Phosphatidylcholine (L)PE (Lyso) Phosphatidylethanolamine (M)ANOVA (Multiple-Class) Analysis Of Variance (U)HPLC (Ultra) High Performance Liquid Chromatography Cer Ceramide CID Collision Induced Dissociation CMC Carboxymethylcellulose C-trap C-Shaped Ion Trap DDA Data-Dependent Acquisition DG Diglyceride FA Fatty Acid FDR False-Discovery-Rate FFPE Formalin Fixation and Paraffin Embedded FT-ICR Fourier-Transform Ion Cyclotron Resonance Mass Spectrometer GC Gas Chromatography HETE Hydroxyeicosatetraenoic Acid IL Interleukin IUPAC International Union of Pure and Applied Chemistry LC Liquid Chromatography LSD Light Scattering Detector m/z Mass-To-Charge Number MS Mass Spectrometry MS2 Tandem Mass Spectrometry MSI Mass Spectrometry Imaging NTDs Neglected Tropical Diseases OCT Optimal Cutting Temperature PAF Platelet Activation Factor PCA Principle Component Analysis PI Phosphatidylinositol PLA Phospholipase A PS Phosphatidylserine PUFA Polyunsaturated Fatty Acid PZQ Praziquantel QqQ Triple Quadrupole Mass Spectrometer RGB Red, Green and Blue RGB Red-Green-Blue ROI Region-Of-Interest ROS Reactive Oxygen Species S. mansoni Schistosoma mansoni SM Sphingomyelin TG Triglyceride TH2 T-Helper Cell Type 2 TIC Total Ion Current TLR Toll-Like Receptor ToF Time of Flight 4 List of Publications This thesis is based on the following publications in peer-reviewed journals Publication 1 Kadesch, Patrik; Quack, Thomas; Gerbig, Stefanie; Grevelding, Christoph G.; Spengler, Bernhard; Lipid Topography in Schistosoma mansoni Cryosections, Revealed by Microembedding and High-Resolution Atmospheric-Pressure Matrix-Assisted Laser Desorption/Ionization (MALDI) Mass Spectrometry Imaging, Analytical Chemistry 2019, 91 (7), pp 4520-4528. Publication 2 Kadesch, Patrik; Quack, Thomas; Gerbig, Stefanie; Grevelding, Christoph G.; Spengler, Bernhard; Tissue- and sex-specific lipidomic analysis of Schistosoma mansoni using high- resolution atmospheric pressure scanning microprobe matrix-assisted laser desorption/ionization mass spectrometry imaging, PLOS Neglected Tropical Diseases 2020, 14(5), e0008145 5 Statement in Lieu of an Oath I declare that I have completed this dissertation single-handedly without the unauthorized help of a second party and only with the assistance acknowledged therein. I have appropriately acknowledged and cited all text passages that are derived verbatim from or are based on the content of published work of others, and all information relating to verbal communications. I consent to the use of an anti-plagiarism software to check my thesis. I have abided by the principles of good scientific conduct laid down in the charter of the Justus Liebig University Giessen „Satzung der Justus-Liebig-Universität Gießen zur Sicherung guter wissenschaftlicher Praxis“ in carrying out the investigations described in the dissertation. ____________________________ Date, Signature First referee: Prof. Dr. Bernhard Spengler Second referee: Prof. Dr. Christoph G. Grevelding Day of oral exam: 6 Abstract Neglected tropical diseases (NTD) are a burden to one billion humans in the (sub-)tropics and poverty-related regions, worldwide. Schistosomiasis, caused by the parasitic flatworm Schistosoma mansoni, is one of those NTDs. The disease is currently spreading because of climate change and migration, exposing approximately 700 million people to the risk of infection. Therefore, novel strategies are required to prevent infection and to eliminate the worm burden. One promising drug target is the surface (tegument) of adult male and female schistosome worms. Schistosomes live in constant pairing contact, established via the teguments of male and female, as a prerequisite for egg production. Living in the blood stream, they are also in contact with the host’s immune system and therefore require immune evasion, moderated by the outer tegument. Lipids are one major class of constituents of the tegument, but limited information is available on its exact biochemical composition. The abundance and spatial distribution of lipids is therefore of high interest. MS is the technique of choice to answer this research question, as it allows multiplexed lipid detection in a nontargeted analysis approach. AP-SMALDI MSI is capable of delivering high spatial resolution in the micrometer range. Due to the small size of Schistosoma (approximately 500 µm in length), this high lateral resolution is essential to resolve detailed structures within the worms. In addition, this technique requires low sample quantities, and recent instrumental advances enable analysis of 3D-surfaces. We utilized MSI to investigate and characterize the spatial distribution of lipids on the surface of adult schistosome worms in comparison to their inner tissue. However, there are no suitable protocols available for production of the necessary cryosections and analysis by MSI. To overcome this limitation, different embedding protocols, like the classical embedding in cryomolds were tested and modified, e.g. by centrifugation steps for improved planarity. However, embedded worms were lacking planar orientation and sections were not intact after cutting. Also, the tissue was partially disrupted during this process, leading to poor section quality. Finally, a microembedding approach was developed, which uses small quantities of gelatin and represents a high-precision approach. This protocol allowed preparing consecutive high-quality cryosections of a mating worm couple. MS ion images of intact couples revealed differences between male and female in metabolites and lipids. Detailed structures observed from light microscopic images were retained in ion images at 10 µm and 5 µm spatial resolution. To further investigate putative isobaric interferences, on- tissue tandem mass spectrometry imaging (MS2I) was utilized to trace characteristic lipid fragments across the tissue and to demonstrate the high sensitivity of the setup even at a lateral resolution < 10 µm. This work was summarized in publication one and enabled the investigation of surface in comparison to the inner tissue of schistosomes. High-resolution MSI of male and female surfaces and couple sections was conducted. An LC-MS/MS-based data repository in combination with unsupervised ion image annotation using the “Metaspace” software was employed for lipid assignment. Multivariate statistical analysis of MSI data by hierarchical clustering revealed deviating signal intensities of lipids on surface vs inner tissue of the worm. PC and specific PE signals were enhanced inside the worm, while SM, PS, LPC and other PE lipids were more abundant on the surface. These findings were in accordance with literature, but enhanced the compositional information from lipid class level to lipid species level. In addition, for PEs, the number of carbon atoms in the fatty acyl chains was found to be decreased on the surface in comparison to the inner tissue. Differences between male and female surface compositions were observed as well. Several sex-specific TGs were found, which differed in numbers of fatty acid carbon atoms and double bonds. For the first time, differences in lipid composition were found between male 7 and female S. mansoni worms. Now a broad toolbox of preparative and data interpretation workflows is available to the scientific community, adaptable to a variety of research issues. Chapter I - Synopsis Introduction and Motivation Mass spectrometry (MS) has become a valuable tool in natural sciences over the past decades. For the nontargeted analysis of biomolecules, MS is currently the technique of choice applicable to the detection of a vast variety of molecules simultaneously as it allows identification, quantitation with high sample throughput. Since its invention in 1994,1 MALDI MS imaging (MSI) has become an emerging technique, allowing for the investigation of the spatial distribution of biomolecules in a wide variety of tissues which is of high interest for the life sciences.2-4 Matrix-assisted laser desorption/ionization (MALDI) MSI is the most promising technique for MS imaging, as it is capable of analyzing the distribution of unfragmented biomolecules with a high spatial resolution of typically 5-10 µm5-10 down to 1.4 µm,11 and recent advances in instrumentation enable 3D-surface analysis.7 However, protocols and applications for the most recent technical advances are very sparse yet. Neglected tropical diseases (NTD), endemic to (sub-)tropical and poverty related regions, are threatening millions of people worldwide. This brings the human pathogen Schistosoma mansoni, causing schistosomiasis, also known as bilharzia, into research focus.12 Schistosomes are parasitic flatworms (trematodes or blood flukes) which have developed two different sexes. Male worms are up to 1 cm long and several-hundred micrometer thick, while the female can be longer but thinner. Research should concentrate on understanding this parasitic disease on a fundamental, biological level in order to develop strategies to counteract spreading and severe impairments associated with this disease. It is known that the tegumental surface of schistosomes comprises a wide variety of lipids, which is expected to be of high biological importance,13 since the parasite resides in venules and is therefore in constant, direct contact with the host. Therefore, permanent interaction between the immune systems of both, host and pathogen, occurs. The schistosome surface comprises mostly lipids. Host-derived lipids are acquired and subsequently modified by the parasite to obtain specialized lipids, expected to be crucial for its survival.13 However, knowledge on the exact composition is limited and thus the picture is far from complete for understanding the function of individual key components. For nontargeted, spatial analysis of lipids in such small schistosomes, high-resolution 3D-surface atmospheric pressure scanning microprobe MALDI (AP-SMALDI) MSI is the technique of choice. One major advantage of the technique is that MS and MSI analyses require low sample quantities.7 Therefore, the biological model system S. mansoni is ideally suited for investigating the spatial distribution of lipids on the surface compared to the worm inner tissues by MSI. The anatomical structure of adult schistosome worms is complex as they comprise a number of organs. Amongst those, the sexual organs are of high interest as they are key for reproduction. The preparation of tissue sections is required to access these inner organs of the sub-millimeter-sized worms. However, there was no protocol available to obtain longitudinal sections, in which the shape and structure of adult worms is preserved and which is also compatible with MSI. Additionally, tissue sections are a pre-requisite for characterizing the tegumental surface in comparison to the inner structure of the worms. Therefore, the first step of our work was to develop a sectioning procedure which is compatible with MSI and which enables the preparation of longitudinal tissue sections, 8 allowing to allocate single organs and to investigate paired couples. This work was described in publication number one. To put this information into biological context, the surface of previously coupled male and female worms was compared to signals from the worms’ inner tissues in biological triplicates. After data acquisition, unsupervised annotation and subsequent MS2-based identification, a self-developed multivariate statistical analysis workflow was employed for data reduction and classification of ion signals into biological groups. Differentially abundant lipid signals were found for surface and inner worm tissues across several lipid classes. In addition, differences between male and female surfaces were found, mostly regarding triglyceride (TG) lipids. Our findings were in line with published literature but enhance the knowledge on lipid species level enabling tailored research on lipid drug targets. This work was described in publication number two. Mass Spectrometry Mass spectrometry is a technique applied to ions, determining the mass-to-charge-number (m/z) ratio. In a mass spectrum, the signal intensity is plotted against the m/z ratio, resulting in peaks. Besides the qualitative information, the signal intensity is proportional to the concentration, thus simultaneously enabling quantification. Based on the accurate mass, compound structures can be assigned using databases, based on ppm-range mass accuracy. One of the most important values in MS is the calculation of mass resolution (R), determining the peak width of a signal (m) for example at half-maximum (Δm50%), according to Equation 1. This number is a measure to describe the separability of two peaks by a mass spectrometer. Equation 1: Calculation of mass resolution from an MS peak. Mass resolution (R), mass-to-charge-number 14 ratio (m/z) and half-maximum peak width (Δm/z50%). Another important parameter, especially to estimate the probability of true-positive database assignment, is mass accuracy (Δm/m). Equation 2 describes the deviation of observed mass (mobs.) from theoretical mass (mth.) relative to the theoretical mass. Mass accuracy is typically reported in parts-per-million (ppm). Equation 2: Calculation of mass accuracy from empirical and theoretical values. Mass accuracy (Δm/m), 15 observed (mobs.) and theoretical mass (mth.). Besides solely relying on exact mass determination and database assignments, identification can be conducted by performing fragmentation experiments and comparing fragmentation spectra to those of authentic standards. The most common fragmentation mechanism is collision-induced dissociation (CID). An ion package of a small m/z range, typically belonging to only one substance, is isolated as the precursor. Collisions are induced by introduction of an inert gas at elevated pressure and subsequent acceleration of the ions, ultimately leading to dissociation into product ions. Depending on substance class, characteristic fragment ions occur, which serve for structure elucidation and thus identification. A wide variety of mass spectrometers is available for mass analysis. The most common mass analyzers are (triple-)quadrupole (QqQ), time of flight (ToF), fourier-transform ion cyclotron resonance (FT-ICR) and orbital trapping instruments. The relative numbers of 9 publications reporting the use of these four mass spectrometer types are shown Figure 1. To date, predominantly ToF-analyzers are used. However, market share decreases because of the newly developed orbitrap instruments. In the early 2000s, the use of QqQ increased and since 2011 remained constant. FT-ICR mass analyzers, as first commercial and widespread high-resolution MS instruments, were extensively used in the 90s. However, the market share decreased in the 2000s and is relatively constant since 2010. The latest MS technology are the orbitrap instruments. After introduction of commercial instruments in 2005, the market share increased each year. Most likely, many researchers replaced ToF and FT-ICR instruments by orbitrap instruments as they offer high mass resolving power and accuracy at adequate scan speeds and require fewer resources than FT-ICR instruments. Figure 1: Mass analyzers used in scientific literature per webofknowledge.com-database search published between 1995 and April 2019. MALDI for analysis of large biomolecules by use of an organic matrix was described by Karas and Hillenkamp in 1988.16 An organic matrix, nicotinic acid, was used for controlled energy uptake and soft desorption of biomolecules, allowing desorption and ionization of proteins above 10 kDa by an ultraviolet laser system.16 Because of the “soft” ionization character of MALDI, intact quasi-molecular ions can be observed, similar to electrospray ionization (ESI).16-18 Nowadays, organic matrices are widely used by the MALDI community with the most prominent matrix being 2,5-dihydroxybenzoic acid.19 In MALDI, the matrix serves to dilute and spatially separate the analytes.20 Laser-generated photons are absorbed by the matrix, leading to transition into the gas-phase of both, matrix and analyte.20 To date, the most prominent model for MALDI is the so-called “lucky-survivor model”.20 Upon laser irradiation, negatively and positively charged ions in different charge states and neutral particles are obtained.20 In the MALDI-plume, emitted with high initial velocity and velocity spread, neutralization occurs upon rapid recombination of particles, proportional to charge state.20 Singly charged ion species are the “lucky survivors” of this process.20 Fenn et al. published the concept of ESI for analysis of (large) biomolecules in 1989.21 Nowadays, ESI is one of the most commonly used ionization techniques in bioanalysis worldwide.22 ESI is a spray ionization method in which a substance is transferred from liquid- neutral to gaseous-ionized state.17,21 The “soft” ionization character of ESI leads to intact 10 quasi-molecular ions, compared to more “hard” ionization techniques prone to analyte fragmentation.21 The benefit over classical soft ionization techniques is that ESI can be operated ex vacuo, under ambient conditions.21 This enables in-line coupling of liquid chromatography (LC) to MS, as is commonly used in bioanalysis.17,22 In ESI, a liquid analyte-containing solution is pumped through a stainless steel capillary at microliter-per-minute flow rates. An electrostatic field of several kV/mm is applied between capillary and MS-inlet.17,21 Excess charge at the needle tip leads to formation of fine, charged droplets because of Coulomb forces induced by the electric field.21 At elevated temperatures in the MS interface region, solvent evaporates from the fine droplets until the surface charge density exceeds the Rayleigh limit and Coulomb explosion/repulsion leads to smaller daughter droplets.21 A series of this cascades yields quasi-molecular ions suitable for detection by MS.21 In 2000, Makarov published the concept of an electrostatic axially-harmonic orbital trapping mass analyzer, also known as orbitrap.23 Coupling to an ESI ion source was published in 2003, which opened the field for commercial success and widespread use by the scientific community.24-26 The orbitrap consists of a spindle-like inner electrode and a split, barrel-like outer electrode.23 Ions are trapped in an electric field induced between inner and outer electrode, forcing ions to orbit on stable trajectories around the central electrode.23 Oscillation in axial direction induces an electric current at the split, outer electrode.23 A frequency spectrum is obtained by time-domain transient and subsequent Fourier- transformation.23,25 The mass-to-charge dependency of the axial oscillation frequency (ω) is described in Equation 3.23 Obtained mass spectra show characteristics of a high mass resolving power above 100,000 at m/z 200 in combination with a high mass accuracy in the lower ppm range.25 Equation 3: Description of ion motion in an orbitrap. Axial oscillation frequency (ω), axial restoring force 23,27 (k), charge (q) and mass of an ion (m). Well-controlled ion injection into the orbitrap is key to obtain high mass resolving power and accuracy.25 First versions used electrostatic lenses to guide ions into the orbitrap, demanding discontinuous ion supply.23,25 Implementation of a linear quadrupole ion trap or C-trap, a C- shaped linear ion trap, allowed collection of dense ion packages prior to HRMS analysis.25 Ejection of the ion package in nanosecond pulses forces the ions onto more stable trajectories.25 The accumulation of ion packages does not only increase sensitivity but also leads to an increase in mass resolving power and higher dynamic range, because of fewer axial dephasing.25 Additionally, by implementation of a “lock mass” function, recalibration of mass spectra to ubiquitous contaminants on the fly, sub-ppm mass accuracy can be obtained.28 Solely based on MS1 spectra, database assignments are a valuable tool for assigning molecular structures to m/z values. However, only MS2 experiments enable reliable compound identification. This feature was not available for orbitrap mass spectrometers. To overcome this limitation, an octapole collision cell was implemented at the back end of the C- trap.29 Fragmentation experiments enable the aforementioned reliable identification of biomolecular compounds, especially because of FT orbitrap performance characteristics available in MS2.29 To date, the setup described here is state-of-the-art and globally used by the bioanalytical community.26 11 Mass Spectrometry Imaging Classical biological techniques such as (immuno-)staining are either very unspecific or are tailored for recognition of one particular analyte or class.2,30,31 Knowledge of the sample composition concerning target-molecules is required prior to investigation.2,4,31,32 In addition, production of customized, novel antibodies for immunostaining is time consuming and typically requires animal experiments.33 In contrast, MSI is able to detect a large variety of biomolecules in a complex sample, label-free and multiplexed.2 To investigate the spatial distribution of unknown substances, a wide variety of MSI1 tools are available.30,32,34 The most common MSI ionization techniques are MALDI and desorption electrospray ionization (DESI).4 The ionization process in DESI is taking advantage of electrospray ionization.35 An electrospray, consisting of primary, charged solvent droplets is directed towards the sample surface, where analyte molecules are dissolved and desorbed.35 Secondary charged and analyte-containing droplets are attracted by the MS inlet because of an electric field.35 The benefit of DESI is that multiple charging is obtained and no matrix has to be added.35 The spatial resolution is less than one millimeter,36 with a minimum of 35 µm.37 MALDI on the other hand has a spatial resolution of 5-10 µm,2,3 but only singly charged ions are obtained.20 By combination with ToF MS, analysis of large biomolecules such as proteins or glycans becomes feasible.2,3 However, the extended mass range of ToF instruments comes at the cost of reduced mass resolving power and accuracy, indispensable when investigating complex small molecules such as lipids. This limitation can be overcome by coupling the ionization source to either FT ICR or orbitrap mass spectrometers.5,31 Ion packages are collected for each pixel prior to HRMS analysis, requiring ion trapping. The C- shaped linear ion trap is the most efficient commercialized ion trap for that purpose.25,38 To date, the C-trap is only available for orbitrap mass spectrometers. The ion yield obtained by MALDI from very small spots is rather low. Therefore, at high spatial resolution, orbitrap mass analyzers are the instruments of choice. Figure 2: MALDI MSI operating principle. The sample is scanned in a pixel-by-pixel workflow where matrix/analyte co-crystals are desorbed/ionized by a pulsed UV laser beam, generating one mass spectrum per pixel. MS ion images are formed, displaying semi-quantitative analyte distributions in regions of interest. 12 For MSI, the sample area of interest is divided into pixels. The surface is probed by recording one mass spectrum per pixel in a pixel-by-pixel-type workflow. Thereafter, images are constructed by combining spatial x,y- and mass spectrometric m/z-information. The signal intensity is usually depicted by a color-scale or brightness gradient to visualize the spatial abundance of one m/z-signal. Upon overlay of the three native color channels, red, green and blue (RGB), overlays can be constructed losslessly. The workflow of MSI is shown for MALDI in Figure 2. Depending on the biological question, tissue, e.g. organs, are cut into cryosections with thicknesses in the lower micrometer-range.31 For sectioning of fragile samples, several embedding methods are available.31 After drying of the sample in a desiccator, the sample is coated with matrix by either spraying or sublimation.2,3,11,31 The surface is analyzed by the MALDI probe pixel by pixel.1 Images are then generated for different analytes and subsequently interpreted.1-3,31 The pixel size in MALDI MSI is typically in the range of 20-50 µm.2,3,6 Recent technical advances enabled to reach laser spot sizes down to 5-10 µm,5-10 and even 1 µm.11,39 The focal depth of the laser is inversely proportional to the laser focus, which is especially relevant at high spatial resolution. Sample topography can thus lead to changes in spot size, affecting fluence and therefore changes in the MALDI desorption and ionization process and signal intensity.7,20 Topography-related measurement artifacts can be reduced by use of a laser based, triangulation autofocusing system, enabling analysis of non-flat surfaces.7 However, the MALDI process in most instruments is taking place under vacuum conditions, therefore demanding appropriate sample preparation according to volatility of analyte and matrix.16,20 This limitation can be overcome by use of an atmospheric-pressure MALDI source (AP-SMALDI).2,5,31 AP-SMALDI is well suited for analysis of small molecules such as lipids, saccharides, peptides and drug molecules.2,5,7,31 Coupled to an orbitrap mass spectrometer, this system allows for studying biological specimen with high resolution and high accuracy in mass and space. Data evaluation is one of the key steps defining the experimental outcome. Most experimental settings use MSI at MS1 level to determine the distribution of compounds.2-4,31 Signal annotation then solely relies on database assignments. For large datasets and non- targeted analysis, this process requires extensive human resources, because annotations need to be verified according to factors such as spatial distribution, possible adducts, mass accuracy or isotope ratios. Metaspace, a recently published online repository, allows for unsupervised database annotations of high-resolution MSI data (metaspace2020.eu).40 Metabolite annotation is controlled by false-discovery-rate (FDR) based bioinformatics, scoring each signal to the measures ρchaos, rating the randomness of an ion image signal distribution, ρspectral, for matching spectral isotope ratios, and ρspatial, rating the co-localization of putative isotopologues.40 In sum, these signals represent the metabolite-signal match (MSM) score.40 After scoring against an in silico decoy database, containing unexpected/not meaningful adducts, such as heavy metals, an FDR-score value is obtained to rate the probability of false-positive assignment.40 Commonly used metabolite and lipid-containing databases are Lipid Maps (Lipidomics Gateway, lipidmaps.org),41,42 SwissLipids (a knowledge resource for lipids and their biology, swisslipids.org)43 and the Human Metabolome Database (HMDB, hmdb.ca).44-47 However, these databases are often limited to known metabolites. This limitation can be overcome by generating a home-built database from MS2 experiments, allowing reliable identification. In case of MALDI, in particular at high spatial resolution, however, the resulting ion population is relatively low. This is especially problematic in MS2-experiments, where, again, many ions are lost during transfer and 13 fragmentation. MS2 imaging (MS2I) experiments may help to overcome this issue. MS2I is not feasible for samples that are analyzed with highest lateral resolution or for experiments with very low amount of sample material. Research has been conducted towards smart methods for data-dependent acquisition (DDA) in MSI.48 However, commercial solutions are not available to the MSI community yet. Currently, this limitation is partially overcome by building up home-built, custom databases, e.g. by performing extraction and LC-MS2 experiments complementarily. Lipid Assignment and Identification According to International Union of Pure and Applied Chemistry (IUPAC), lipids are defined as “substances of biological origin [...] soluble in nonpolar solvents”.50 Biochemically, lipids can be defined as “a chemically diverse group of compounds, the common and defining feature of which is their insolubility in water”.51 This non-stringent IUPAC definition has been described more precisely in a classification system, which is referred to as Lipid Maps terminology.49,52,53 Structural features are the basis for a hierarchical classification system.43,49 An example for hierarchical classification of one phosphatidylcholine is shown in Table 1. In Figure 3, a putative phosphatidylcholine ether head group is shown for correlation with lipid hierarchy. At category level, the only information on the lipid is that phosphate is bonded, by definition, to a glycerol backbone in sn-3 position, depicted in orange and blue in Figure 3. The head group is defined further on main class level as glycerophosphocholine, corresponding to orange, blue and green structures, shown in Figure 3. The substituents at sn-1 and sn-2 position (see Figure 3) are defined at sub-class level, in this case as esters (here, in Figure 3, an ester in sn-1 position and an ether in sn-2 position) and thus phosphatidylcholine (PC). At species level, the fatty acid sum composition is known as number of carbon atoms in the fatty acyl chains. The fatty acid composition is depicted as number of carbon atoms and number of double bonds separated by a colon. Therefore, PC (34:1) represents a phosphatidylcholine with the fatty acyl substituents comprising 34 carbon atoms and one double bond. Information on fatty acid chain lengths is obtained at molecular subspecies level e.g. PC (16:0_18:1). Sn-positions of fatty acyl substituents are defined at structural subspecies level as sn-1/sn-2 e.g. PC (16:0/18:1). Information on the double bond is added on isomeric subspecies level. The double bond position and configuration for aforementioned example could be PC (16:0/18:1(9Z)), which means that the double bond is located at carbon number 9 starting from acid group and has cis (Z) configuration. Most standard lipids can be categorized using this classification system. 49 Table 1: Hierarchical classification system of lipids (example taken from SwissLipids.org). Hierarchy level Example Category Glycerophospholipid Main class Glycerophosphocholine Sub class Phosphatidylcholine (PC) Species PC (34:1) Molecular subspecies PC (16:0_18:1) Structural subspecies PC (16:0/18:1) Isomeric subspecies PC (16:0/18:1(9Z)) 14 Despite lipids comprising only a very limited elemental complexity (consisting only of C, H, N, O, P and S), isobaric variants are challenging to the field. Instead of fatty acids, ethers can occur at sn-1/-2 positions. A PC plasmalogen at sn-2 is depicted in Figure 3, as one representative of ether lipids, with the double bond being at Δ1 position. The MS toolbox for identification of sn-isomers or even double bond elucidation in lipids, to date is relatively sparse and reliable lipid assignment/identification is complicated. Lipids, down to isomeric subspecies level, however, are important for a variety of endogenous functions, e.g. influencing membrane packaging/arrangement54 or arrangement of lipid-protein complexes.55 Therefore, improved MS methods are required in the future to further investigate lipid isomers. Figure 3: Generic phosphatidylcholine ether. Glycerol (red), phosphate group (blue) and tertiary ethanol amine (green) forming the phosphatidylcholine backbone. Alkyl ester in sn-1 position and alkenyl ether 1 2 in sn-2 position. Substituents R and R with variable alkyl chain length. MSI typically relies on differentiation of compounds solely based on MS1 data. However, the lipid adducts lyso-phosphatidylethanolamine (LPE) (18:1) as sodium adduct (m/z 502.2904) and the protonated molecule of LPE (20:4) (m/z 502.2928) require a mass resolving power of >210,000 for signal separation and a mass accuracy better than ± 2 ppm for correct assignment. Commercial orbitrap instruments are not able to deliver this resolution in this m/z-range. Database assignments are therefore prone to misinterpretation.56 This limitation can partially be overcome by using more reliable databases. Such databases can be derived from e.g. top-down lipidomic techniques, where a precursor, an intact lipid ion, is selected and fragmented, yielding characteristic fragments for identification of head group and nonpolar substituents. Fragmentation experiments are essential for reliable identification. Common mass spectrometric approaches use either direct infusion techniques, referred to as shotgun approach,57 or separation by chromatography prior to MS analysis.58,59 Both approaches require extraction of lipids by use of (non-)polar organic solvents. The most common methods for extraction use a combination of methanol and water and either chloroform60 or methyl-tert-butyl ether.61 The underlying assumption is that all analytes are extracted quantitatively. This is an intrinsic problem, because model systems for a more detailed investigation of extraction processes are lacking. It has been observed, that, after tedious method optimization, lipids can be extracted almost quantitatively.61 Non-quantitative extraction may result in extraction bias throughout different analyte classes, leading to over- /underestimation of abundance. This issue was addressed by the microbial-research community regarding extraction of e.g. DNA,62 lipids63,64 and proteins,65 and subsequent high- throughput analysis. Non-quantitative extraction may lead to unintentional falsification of results, putatively altering experimental outcome and therefore affecting biological interpretation. To shed light on the severity of such effects further systematic investigation 15 are required in the future to help to develop strategies for estimating and counteracting extraction bias. Mass spectrometry is widely accepted as a suitable method for comprehensive analysis of lipids. For absolute quantitation, typically a set of stable isotope-labelled internal standards is used.66 Lipids are defined by a wide variety of physico-chemical properties such as polarity and solubility. Therefore, differences in ionization efficiency occur which become increasingly challenging throughout lipid hierarchy and require even more elaborate experimental designs for enabling comprehensive analysis.66 This is already complex in shotgun-lipidomics and becomes even more difficult when separation techniques such as LC are used.67 For quantification, multiple standards are needed to overcome limitations derived from differences in retention time, otherwise hindering absolute quantitation.67,68 However, lipid studies are often solely limited to descriptive findings, whereas biochemical, mechanistic studies of e.g. individual lipids are relatively sparse and often specific to one particular organism. Another general problem is the sheer amount of data generated throughout all ‘omics disciplines. One promising attempt to cope with the data is to use extensive bioinformatic strategies to combine ‘omics data and to apply known biological pathways as a template.69,70 By combining information e.g. from transcriptomic, proteomic and metabolomic datasets using a dimensionality-reduction approach, it is possible to deduce putative mechanistic linkages, which then require further investigation and verification in vivo.70 Figure 4: Fragment spectrum of precursor m/z 804.564 acquired in negative-ion polarity by higher- 71,72 collisional dissociation (HCD). Data interpretation was conducted by Lipid Data Analyzer. The signal was identified as PC (16:0_18:1) based on fatty acyl (m/z 255.2327, Δm/m= 1.0 ppm and m/z 281.2484, Δm/m= 0.7 ppm) and characteristic PC head group fragments (at m/z 168.0420, Δm/m= 73,74 6.6 ppm and m/z 224.0690, Δm/m= 1.5 ppm). One classical way, as was done in this work, is to acquire data for nontargeted analysis by data-dependent-acquisition (DDA), where precursors are selected and fragmented according to pre-defined criteria in dependency of analytes of interest. The subsequent data interpretation utilizes fragmentation-rule-based algorithms for reliable identification in silico. An example for a fragment spectrum acquired in negative-ion polarity is shown in Figure 4. Experimentally, the precursor ion at m/z 804.564 was isolated in an isolation window of ± 0.5 Da and fragmented by HCD. In silico, Lipid Data Analyzer (LDA) identified the compound as PC (16:0_18:1). Fragments of the fatty acyl (FA) substituent can be observed in the spectrum (black numbering). The FA (16:0) is shown at m/z 255.2327, while FA (18:1) 16 appears at m/z 281.2484. The head group can be identified according to characteristic fragments at m/z 168.0420 and m/z 224.0690.73,74 All characteristic fragments, required for identification, were detected. Schistosomiasis The three most prominent Schistosoma species are S. japonicum, S. mansoni, and S. haematobium, which inhabit the mesenteric veins of gut and bladder of their hosts, respectively.12,75 In the venous system the adult male and female worms reside in a mated state, producing approximately 300 eggs per day.12 Maturation of the female requires constant direct pairing contact with the male, resulting in fertility and egg production.12 Therefore, intervention in the male-female interaction is ideally suited for disrupting the life cycle chemotherapeutically. Approximately 50% of the eggs reach the gut lumen and are excreted via feces. Residing eggs may reach the liver or spleen where they are trapped.12 On the other hand, parasitic miracidia hedge from excreted eggs upon contact with water to close the life-cycle.12 The sex of the worm is already defined at that stage.76,77 Specific intermediate sweet water snail hosts of the genus Biomphalaria are infected and production of multiple generations of cercaria occurs.12 Vertebrate-infective cercaria are released upon light exposure and fresh-water contact.12 This vertebrate-infective state penetrates the skin of e.g. homo sapiens and loses the tail, which was previously required for motility in water.12 The parasites utilize the blood stream to reach the host lungs, where adulteration of schistosomula to mature flatworms occurs.12 Pairing contact between sexually immature male and female is established between dorsal tegument of the female and the ventral tegument of the male, also called the gynaecophoric canal.12 Sexual adulteration of the female is initiated and maintained during permanent pairing contact with the male and is a reversible process.12 The pairing status of a female can be assessed visually by phenotype, because its size increases as a result of pairing-stimulated vitelline cell production.12 The mated couple utilizes passive transport to reach the mesenteric veins of the hosts gut.12 Also, oral and ventral suckers can be utilized for active movement. For successful reproduction, spermatozoids are produced in the males’ testes and released towards the female. Taken up by the female, spermatozoids can be stored to later fertilize oocytes. For composite egg production, an oocyte is surrounded with multiple vitelline cells in the ootype. The female releases composite eggs to close the life cycle. Schistosomes occur in (sub-)tropical areas and are endemic to Africa, the Middle East, South America and Southeast Asia, but first cases have been reported in Europe already.12,75 Approximately 220 million people suffer from schistosomiasis worldwide, 50% of which are treated annually.75 Prevalence and endemic spread are expected in this century due to global warming.78,79 The major problem of schistosomiasis is that the host immune system typically reacts to trapped eggs by granuloma formation and ultimately calcification.12 Depending on the worm burden and immune competence, schistosomiasis can lead to lethal liver cirrhosis if not treated in time.12 The only chemotherapeutic treatment available is Praziquantel (PZQ), commonly used in mass drug administration for temporarily eliminating the worm burden of human communities in endemic areas.75,80 Patients often live in poverty with limited access to sanitation and clean water and are thus prone to contemporary reinfection.12,75,80 The World Health Organization classified schistosomiasis as a “neglected tropical disease”, to raise awareness to this devastating disease and for focusing research to develop novel treatments and countermeasures.75 One prerequisite for developing novel strategies is knowledge of fundamental biology e.g. to identify key metabolites, essential for the parasite’s survival. 17 Mass Spectrometric Investigations of Lipids in Adult Schistosoma mansoni A number of studies have been conducted describing the abundance of lipids in adult S. mansoni worms using mass spectrometry. In 1998, a first study used high performance liquid chromatography (HPLC), coupled to a light scattering detector (LSD) to investigate phospholipids from tegument and worm carcasses of S. mansoni worms.13 Fractions of PC and PE were collected and quantified by the LSD.13 Subsequently, lipids were identified according to characteristic head group fragments and digestion by phospholipase A2 (PLA2) for structure elucidation of sn-positional isomers.13 Extracts of whole worm versus tegumental fragments prepared by a protocol by Dyer and Blight revealed enrichment of saturated and unsaturated PC in the tegument. In addition, ether-linked species of PC and PE were described in adult S. mansoni for the first time. PC-O were discussed to be putative precursors of platelet activation factor (PAF), known for the ability to modulate immune host cells and to act pro-inflammatoryly.13 However, PE-O was found to be highly enriched in the tegument.13 Also, parasite-initiated elongation of host-derived lipids was observed.13 The biological importance of unsaturated PC and plasmalogen species has been discussed with regards to increased resistance against reactive oxygen species (ROS), as sequestered by neutrophils and macrophages, and resistance against the host immune system.13 Finally, the authors propose fatty acid (FA) (20:1) as possible second messenger for signal transduction from tegument to worm body, because of its high abundance throughout different lipid species in the tegument.13 In 2008, a study based on HPLC-MS was conducted, extensively describing PC, PE, PS and PI in S. mansoni, compared to hamster blood.81 Overall, 400 phospholipids were quantified in both, positive- and negative-ion mode.81 Very long-chain FA substituents, containing more than 20 carbon atoms, were observed for PC, PE and PS in the worms, but were absent in the host.81 Thereby it was confirmed that S. mansoni is able to elongate host-derived FAs.81 The authors proposed that schistosomes either comprise specific trans-acylases for elongation of fatty acids in phospholipids or a distinct DG-pool for phospholipid synthesis.81 In 2014, rapid identification and quantitation of lipids was enabled using an Orbitrap instrument (Thermo Fisher Scientific, Bremen), which is able to acquire high-resolution mass spectra. The Orbitrap significantly increased throughput and sensitivity. Thereby detailed lipidomic investigations of S. mansoni became feasible. One study combined MALDI MSI and ESI-HRMS to distinguish different Brazilian strains, and males and females.82 Glycerophospholipids, DG and TG were investigated on whole worm couples.82 The authors claim to have localized structural organs such as tegument, oral and ventral sucker, digestive system and reproductive organs.82 However, according to anatomy of adult schistosomes, the locations of the reproductive system shown in either MS ion image or microscopic image, match the authors’ assignments,82 but could also be attributed to topography-related artifacts. More abundant signals were mostly detected in males,82 because the female resides in the gynaecophoric canal of the male and thus cannot be reached solely by MALDI. From principle component analysis (PCA), markers for different strains were obtained and most belonged to the classes of TG and PC.82 The class of TG was found exclusively in males.82 The literature suggests that free FA are stored as TG, because catabolism through β-oxidation cannot occur in schistosomes.82 Only one signal (m/z 825) could be attributed to females.82 Another study by the same group in 2015 used MALDI MSI to distinguish males and females, determining the distribution of lipids and investigate the effect of PZQ on males compared to females.83 Differences between males and females were found from raw data.83 After PZQ treatment in mice, the phospholipid profile of male and female S. mansoni worms was altered.83 The authors conclude that molecular pathways are affected differently in both sexes.83 The location of PS was attributed 18 to oral and ventral sucker and the reproductive system.83 Another HPLC-MS study investigated PC, PS, PE and PI as lyso and diacyl species in whole worms, tegument and hamster blood.84 PC and PS were found to be different in the tegument compared to whole worm.83 The findings show PS to be more abundant on the tegumental surface84 and are in contradiction to studies by Ferreira et al from 2015.83 Additionally, the tegument was found to be enriched in lyso species, which the authors thought to be involved in host-parasite interaction.84 However, to test this hypothesis, the authors attempted to detect lyso species from extracts of incubation medium in vitro and from blood in vivo, but were not able to detect such lyso species.84 For the first time, a peculiar double-bond position at Δ5 was described in PC (34:1) for schistosomes, not detectable in the host.84 No differences in PE and PI were observed between surface and worm body.84 The authors suggest that schistosomes comprise molecular mechanisms to selectively enrich their tegument in certain glycerophospholipids.84 Recently, a comprehensive lipidome analysis, covering the whole life cycle and excretory products, was conducted using LC-MS and gas chromatography (GC) coupled with MS.85 A database containing 350 lipids and bioactive molecules was established and deposited for further use.85 The three predominant species, present throughout the whole life cycle were PC (34:1), PC (36:1) and PC (36:2).85 The lipid profile of whole worms and eggs was found to be similar.85 Immunomodulatoryly relevant compounds, such as polyunsaturated fatty acids were found, but in both stages, with lower abundance in adult worms.85 To cover functional aspects, S. mansoni-derived PS-fractions were added to dendritic cells, inducing anti-inflammatory TH2 and interleukin 10 (IL-10)-producing T-cells by activation of toll-like receptor 2 (TLR2).85 Arachidonic acid-derived 15- hydroxyeicosatetraenoic acid (15-HETE) was hypothesized to play a role in host-parasite interaction.85 Altogether, the resulting database serves as a starting point for further identification of immunomodulatory lipids.85 Lipids are speculated to play a key role in protection of the parasite from the host’s immune system and extensive modulation thereof. However, further studies are required to confirm findings of this study and to prove the proposed interaction hypotheses. Atmospheric Pressure Matrix Assisted Laser Desorption/Ionization Mass Spectrometry Imaging of Biological Tissues (Publication 1) MALDI MSI is commonly applied to a large variety of biological samples.2 For investigation of e.g. animal organs, it is necessary to prepare micrometer-thin tissue sections prior to matrix application.2,5 Typically this is achieved by cutting fresh-frozen tissue in a cryotome and subsequent mounting on a glass slide.5 Thereby, the inner substructure of organs becomes accessible by MSI and additionally a planar surface is obtained allowing to record MS images at high spatial resolution of ≤ 20 µm.11 For smaller specimens, however, direct sectioning often is not possible. This limitation can be overcome by use of embedding material prior to cutting. One commonly used embedding agent in histology is ‘optimal cutting temperature’ (OCT) compound. The use of OCT is not recommended in mass spectrometry because contained glycols and resins cause severe ion suppression effects in MALDI.31 Another common procedure utilizes fixation by formalin (methanol and water containing formaldehyde solution) and subsequent embedding in paraffin (FFPE) to preserve the sample, enabling storage under ambient conditions over a long period of time.86 This FFPE procedure is commonly used in clinical routine after surgical removal of tissues to enable assessments by pathologists and for generation of large tissue libraries e.g. in oncology. This protocol is not compatible with MALDI, because embedding and deparaffinization require xylene, leading to delocalization and removal of many lipid classes.87,88 Nevertheless, there 19 are few substances available for embedding, compatible with MSI. Carboxymethylcellulose (CMC) can be used as 2 %-5 % aqueous solution and has been successfully applied to specimens as small as Anopheles stephensi mosquitoes.89,90 Gelatin allowed cryosectioning of the cerebral ganglia of the freshwater snail Lymnaea stagnalis and subsequent MALDI MSI was conducted.34 To investigate the effect of an eye drop preservative in a rabbit model system, eyes were embedded in aqueous tragacanth gum solution before cryosectioning.91 However, embedding agents need to be selected, formulated and optimized, depending on sample and analytical question. For small, micrometer-sized tissue samples, a variety of specialized protocols is available. To investigate the whole-body of Caenorhabditis elegans, a freeze-cracking method was applied to enable studies by MALDI MSI.92 Additionally, the signal intensity in the mass range of phospholipids increased, compared to direct analysis of the surface.92 For investigation of the fruit fly, Drosophila melanogaster was embedded in CMC.90 The protocol was later adapted to the use of gelatin solution,93 allowing to cut 20 µm thick tissue sections for molecular investigations by MALDI MSI.90,93 Modifying this protocol by adding an ethanol series for water removal prior to embedding in either CMC or gelatin also successfully gave longitudinal cryosections and additionally allowed preparation of D. melanogaster brain sections.94 Besides increasing the tissue stability by an ethanol series, a variety of chemical fixation techniques is available such as formalin fixation without paraffin embedding. Another fixative is glutaraldehyde (1,5-pentanedial) which reacts with amino groups, leading to cross- linking of proteins and thus increasing solidity. Glutaraldehyde has been useful for preparing sections of bovine eye balls and analysis by MSI.95 A severe decrease in sphingo- and phospholipids was observed, likely because of high glutaraldehyde concentration and long incubation time, ultimately leading to quantitative protein cross-links, hindering extraction and ionization by MALDI.95 However, when decreasing the amount of fixative and reducing the time of exposure, even single cells can be fixed, allowing detection of carbohydrates, nucleic acids and lipids.96 Therefore, glutaraldehyde is a promising candidate for sample preservation, compatible with MSI. Mass Spectrometry Imaging of Adult Schistosoma mansoni Parasites To access inner organs and investigate male and female in copula, artifact-free tissue sections of paired couples were required. S. mansoni worms are approximately 8-11 mm long and 500 µm thick. Therefore, direct cutting is not possible and application of embedding techniques is required. Gelatin was chosen for embedding, which is most fluidic at elevated temperatures and still flexible under ambient conditions. The content of gelatin was varied to find the best compromise between fluidity and solidity at –20 °C and –30 °C. The optimum was found at an aqueous solution of 8 %. Classical cryomold embedding gave longitudinal sections, but severe artifacts and fissures of tissue were observed. The yield of tissue sections was solely based on chance. To increase the probability to obtain sections with desired orientation, a centrifugation step was added prior to freezing the sample. Thereby, section quality was improved and worm orientation was visually assessable. Nevertheless, tissue fissures were observed and anatomical structures were not determined. This limitation was overcome by using a miniaturized sample holder and microliter amounts of embedding agent. Two consecutive sections of a S. mansoni couple were obtained and scanned with MALDI MSI with high spatial resolution of 10 µm and 5 µm, respectively. Differences in lipid composition were visible between male and female. In addition, structural features were 20 preserved and organs such as the gut were visible in the ion images. Structures were more detailed at 5 µm pixel size. However, doubling the lateral resolution quadruples the number of pixels and therefore recording time. Differences in lipid composition between both sexes are visualized in Figure 5. MSI was conducted with high resolution in mass and space of 240,000 and 5 µm, respectively. The MS ion signal shown in Figure 5A at m/z 782.5674 is one representative example for equally distributed signals between male and female. Based on accurate mass, this signal can be assigned to isobaric ions of either PE(37:1) or PC(34:1) as sodium adduct when compared to the human metabolome database (HMDB).44-47 A signal at m/z 810.5988, which is more abundant in the male, is shown in green in Figure 5B. Assignment according to HMDB44-47 led to PE(39:1) or PC(36:1) as sodiated species. The red ion channel shown in Figure 5C corresponds to either PE(39:2) or PC(36:2) as sodiated molecule. This signal shows an increased signal intensity in some parts of the female. The red-green-blue (RGB) overlay in Figure 5D was obtained by combining Figure 5A-C. Differences between male and female and heterogeneities within each individual become even more obvious. Some structural features can be assigned to organs when compared to the digital light microscopic image in Figure 5E. For instance the gut of male and female can be recognized and shows the characteristic, meander-like shape (see white arrows in Figure 5D). Putative isobaric interferences may lead to wrong annotations when relying solely on MS1- data. Fragmentation experiments directly from tissue are required for reliable identification by MALDI, to eliminate hypothetical isobaric bias. In addition, these MS2I experiments were conducted as a proof-of-concept study, to demonstrate adequate sensitivity of the method. This allows to identify substances on tissue according to characteristic fragment mass. For Figure 5: MS ion images of a S. mansoni paired couple. A - blue ion channel, equally distributed between male and female. B - green ion channel more abundant in the male. C - red ion channel with increased signal intensity in the inner part of the worm. D - red-green-blue overlay. E - digital light microscopic image. All images are to scale. The scale bars are 100 µm. 21 instance, the ion at m/z 808.5828 (in Figure 5C) was identified by MS2I on-tissue. All fragments were exclusive to PC and no fragments indicated the presence of PE. Additionally, such fragments can be traced across the whole tissue, giving the opportunity to discover differential distributions of otherwise indiscriminable, isobaric fragments. A sample preparation method has been successfully developed, allowing reproducible production of tissue sections, subsequent data acquisition by high resolution MSI and finally detailed evaluation of the data. Thereby, this first publication sets the methodical fundamentals, enabling studies in a biological context by comparing surface versus inner worm tissues as shown in the second publication. Mass Spectrometry Imaging - Characterizing the Spatial Distribution of Lipids in Adult Schistosoma mansoni Parasites (Publication 2) The description of a biological system requires extensive planning, determining the later experimental outcome. In case of S. mansoni, where the tegumental surface is in constant, direct contact with the host, the unique lipid composition has been reported previously.84,97,98 Some investigations of the tegument were based on immunohistochemistry, failing to give information on the lipid-species level. More detailed studies of lipid species, however, were based on (LC-)ESI-MS, lacking spatial information and requiring extraction of lipids. This procedure is especially error prone, because the tegument is only nanometer-thick in contrast to a few hundred micrometer-thick worm carcass, therefore containing several orders of magnitude more lipid material.13,84,97,99,In another study, the surface was analyzed by MSI, but no tissue sections were prepared.82 However, reliable tegument studies by MSI were enabled by our group only recently, by implementation of a laser triangulation autofocusing system, especially important at high lateral resolution,7 and a method allowing to prepare longitudinal cryosections of schistosome worms.100 A strategy for data evaluation and interpretation is of utmost importance, because of the large data file size obtained per specimen, especially for high-resolution MSI. Our study attempted to describe the differences between tegumental surface and worm body tissue. Additionally, it was aimed to describe differences in surface lipids between male and female. To solve this issue efficiently, tissue sections of paired S. mansoni worm couples were prepared, harboring metabolic information of both, male and female in the paired state. This enabled comparison to the surface of paired, but freshly separated S. mansoni worms. Separation just prior to chemical fixation was desired to yield worms with known pairing history and thus reduce putative, but expected, differences derived from the unpaired state. The distribution of lipids on the surface of male and female worms was determined by MSI. Analysis of each class, tissue sections of couples on the one hand and male and female surfaces on the other hand, in biological triplicates has set the fundamentals for multivariate statistical analysis. After unsupervised MALDI MSI signal annotation by Metaspace and subsequent comparison to a homebuilt LC-MS2 database, one region of interest (ROI) was defined per biological sample. ROIs were defined based on one representative ion image, displaying the tissue, brought to superposition with a digital light microscopic image of the tissue. The afore-created mass list of all identified lipids was used for exporting mean signal intensity, to obtain one intensity value for each m/z-value per ROI. Multivariate Statistical Analysis The data evaluation process in MSI typically includes visual impressions by the operator. This is an error prone procedure, as it is based on subjective criteria and requires a lot of 22 time to compare and select images of interest. To overcome this limitation, multivariate statistical analysis is a powerful tool to speed-up this process as it allows investigation of all signals simultaneously and according to objective measures, ultimately representing a more robust signal classification approach. Gaussian distribution of signal intensities during one experiment is assumed for all signals, which is a prerequisite for usage of most statistic models. In MSI, statistical analysis is a tool towards more objective data interpretation. In perspective, statistical analysis can be used to decrease the time required for data analysis and interpretation. Nevertheless, the visual impression after statistics gives ultimate proof over the success of this model, categorized into true-positive, true-negative, false-positive and false-negative. The putative normal distribution of all signals obtained by MS is visualized in Figure 6 in blue. Normalization of each signal intensity, in MSI to the total ion current (TIC), leads to an increased probability density, assimilating the signal intensities of all m/z values (see Figure 6). However, the mean remains constant. Also, the area under the curve is still normalized to 1, representing the sum of all possible outcomes. The Z-score (zi) is a statistical transformation system. Within one measurement, the deviation of signal intensity (xi) of each observed m/z value from the mean signal intensity ( ) is calculated relative to the standard deviation (σ), assuming normal distribution. The Z-score can be calculated according to Equation 4. However, outliers or highly abundant signals may lead to falsification of the Z-score when applied to a mass spectrum. This issue can be partially overcome by using the median ( ), which is thought to be less error prone. Equation 4: Calculation of the Z-score. Z-score (zi), observed value (xi), 101 mean ( ) and standard deviation (σ). In MSI, signal intensities are highly heterogeneous across all signals. Based on abundance and ionization efficiency this becomes especially problematic for lower-abundant signals such as DG. The inequality in signal intensity of distinct m/z values is adjusted by applying the Z-score to one measurement. First, signal intensities are now normalized to the same scale and are thus comparable within and between measurements. Second, this brings values statistically slightly closer to each other within one measurement. Obtained scores Figure 6: Gaussian distributions. Standard deviation (σ) and probability density (f). A (TIC) normalization increases the density of the normal distribution. The mean is constant. The area under each curve is 1. 23 can be further processed for analysis of variance (ANOVA). Variations of ANOVA model systems are usually based on variance and test parameters to unravel (linear) correlations between observed values, here signal intensity of biological groups. For comparison of biological groups, multiple-class ANOVA (MANOVA) can be used for finding differences, e.g. based on signal intensity. The means, calculated from distinct biological classes, for instance, are tested for inequality. The means differ from each other if the test turns positive with a yes-type reply. However, it has been determined that there is a difference, but not how it is manifested, e.g. if a signal is more/less abundant. MANOVA enables testing of multiple dependent variables. For MSI data, this could be transferred to different isotopologues, which give multiple signals and always occur in the same ratio e.g. 12C, 13C, 13C2 etc. Overall, MANOVA is a useful tool to answer multiple research questions simultaneously,102 e.g. differences in signal intensities for acquired replicates, and is therefore well suited for evaluation in untargeted analyses of data such as those obtained by MSI. Post hoc analysis comprises multiple statistical testing algorithms. The aforementioned ANOVA is the basis for every post hoc test to discover differences between biological groups. For MS data this allows to rate a signal as either equally, more or less abundant. The significance of a finding is usually FDR controlled and based on p-value, the probability that the observed value can be explained by the null hypothesis. However, for MSI data, post hoc tests are reasonable to calculate differences more objectively and thus narrow-down the signals to be assessed visually for data evaluation. Using significantly different signals for performing MANOVA prior to post hoc testing may serve as an additional quality control procedure. Hierarchical clustering is a powerful tool for the classification of signals based on similarity. Similarity is described mathematically according to e.g. Euclidean distance, as used by the “Perseus” program. The Euclidean distance can be calculated by Equation 5 and expresses the distance between two points, e.g. in signal intensity. In hierarchical clustering, similarity is also quantifiable by a dendrogram in which the similarity and order in which clusters were Equation 5: Euclidean distance calculated from Cartesian coordinates. Euclidean distance (d), two 103 coordinates (p and q) and number of data points observed (n). built, can be visualized. Multiple restarts serve to compensate the starting point problem, because clustering starts from the first value in a matrix and during statistical processing the first value can be random, otherwise leading to irreproducible clustering. The dendrogram displays quantitative differences between signals according to the lengths of the branches. Signal-intensity-based categorization of lipids has been proven to simplify data analysis and speed up the evaluation process for large data file sizes. However, one of the limitations to be overcome is finding the ideal FDR threshold value when performing MANOVA and post hoc tests or hierarchical clustering. This setting was observed to heavily influence the amount of true/false-positive/negative results. After reaching a certain threshold, the classification system yields many more signals to be significantly different but many of these signals are false-positives, observed from MS images. Further optimization and more adequate mathematical procedures may help to balance true/false-positives/negatives. 24 Figure 7: Structures of lipids, analyzed by LC-MS. Phosphatidyl-glyceride (PG), -ethanolamine (PE), - choline (PC), -serine (PS) and -inositole (PI). Sphingolipids of sphingomyelin (SM) and ceramide (CE). R1/R2/R3: fatty acid substituent in sn-1/sn-2-position, and sn-3-position in case of triglyceride (TG). Characterizing the Lipids on the Tegumental Surface of Adult Schistosomes Compared to Inner Worm Tissues We attempted to characterize the surface in comparison to the inner tissue of adult S. mansoni worms by high resolution MSI. Biological triplicates of freshly separated intact males and females and sections of paired couples were subjected to MSI analyses. Signals obtained by MSI were automatically assigned, based on Metaspace online repository using SwissLipids database.40,43 In parallel, a lipid database was generated based on LC-MS of whole worm extracts. Comparing annotations by Metaspace40 to a homebuilt MS2-based database gave more confidence in assigned lipid structures. The head groups of all lipid classes covered in this study are shown in Figure 7. Lipid signals were categorized in an unsupervised fashion into either tissue or surface, or male/female surface-specific, as well as unspecific signals using a self-established multivariate statistical analysis approach. MS images of such categorized signals were assessed visually for correct categorization. Examples for visually correct category assignment are shown in Figure 8, showing the digital light microscopic images for reference in Figure 8A. Males (M) are marked blue and females (F) in red, with direct surface analysis on the left (Figure 8A) and tissue sections on the right (of Figure 8A). An overlay of two ions in a red-green (RG) superposition is shown in Figure 8B. The green signal at m/z 753.5881 was classified to be more abundant on the surface and represents SM (d36:1) as the sodiated molecular ion species. The outlines of the worms in cross sections are visible as well, but are absent in the worm body. The MS ion at m/z 832.5827 is shown in red and was identified as PC (40:7) as the protonated molecule. The red ion channel is more abundant in the inner tissues of the worm and cannot be seen on the surface. Therefore, the multivariate statistical data analysis workflow appears to be a valid procedure for unsupervised signal classification and was verified by visual assessment. Differences in lipid composition were investigated regarding their abundance on the lipid class level. In total, the signals found to be more abundant on the surface comprised 27 SM, 10 PS, 9 PE, 3 LPC and 1 PC species, while signals that are more abundant in tissue sections were 20 PE, 20 PC and 1 LPC species. It has been reported that SM are exclusive to the surface97 and that surface-associated PS are involved in host-pathogen interaction.84 Also, differences in PC were previously found between both tissues.84 Therefore, our findings are well in accordance with literature. The class of PE was further compared on the lipid species level by plotting the number of carbon atoms in the fatty acyl chains versus the number of double bonds. The mean number 25 of carbon atoms decreased from n = 40 in the inner worm tissues to n = 37 on the surface. These findings are in contrast to literature, where no differences in PE were detected using outdated low-resolution, triple quadrupole mass spectrometry.84 However, the experimental conditions in MSI are quasi in situ because they do not require lipid extraction. In addition, instrumental improvements were made in the past years, boosting sensitivity and accuracy of analytical methods. Therefore, our findings are reasonable and expected to be authentic true-positive findings. Applying computational comparison of male versus female surface based on the same dataset revealed sex-specific maker signals. From the MS ion images, signals appeared to be almost exclusive to either sex, thus being highly specific. Analysis on the lipid class level revealed marker signals for females comprising 7 LPC, 6 PC, 4 TG, 3 PE, 3 SM and 1 Cer species, while 18 TG, 1 PS, 1 SM and 1 PC species were more abundant on the tegument of males. The compound classes of LPC, SM and Cer are known to play a role in signal transduction and may thus also be involved in male-female signal transduction.104,105 Looking further into the number of double bonds and number of carbon atoms in the fatty acyl chains of TG gave mean compositions of (50:1) and (58:5) in female and male respectively. This is assumed to affect a decrease in membrane fluidity of the male and thus enhance flexibility. It can be hypothesized that TGs serve as repositories for precursors to prepare schistosome- specific lipids such as phospholipids.106 Differences between worm surfaces and the inner tissues were found by MALDI MSI.#### Computational data analysis comprising Z-scoring, post hoc testing and hierarchical clustering was optimized successfully to reveal higher/lower abundant signals for the different biological groups. One major advantage over classical LC-MS-based methods is that the MSI workflow is significantly closer to in situ conditions. In sum, our findings based Figure 8: Multimodal imaging of S. mansoni worms of surfaces of males (M) and females (F) as well as cryosections of mated couples. A - digital light microscopic image. Black arrows indicate the anterior end. B - RG overlay of two MS ion images categorized upregulated on the surface (in green at + m/z 753.5881 ±3 ppm assigned to SM (d36:1) as [M+Na] with Δm/m ≤ 1 ppm) and worm tissue section + (in red at m/z 832.5827 ±3 ppm assigned to PC (40:7) as [M+H] with Δm/m ≤ 1 ppm). All images to scale. 26 on MSI are supported by literature, and conclusive findings were previously obtained by classical lipidomic techniques. The developed workflow has the potential to be adapted to a variety of other research questions. More advanced algorithms may lead to a more reliable classification as false-positive/negative findings need to be balanced manually by adjusting the FDR. Conclusions and Future Perspectives High-resolution 3D-surface AP-MALDI MSI was applied to the detection of a variety of small biomolecules in sections of S. mansoni parasites. A method was developed to prepare longitudinal cryosections of tiny, adult worms which is compatible with MSI and MS2I experiments at high spatial resolution of 5 µm and 10 µm. MSI in combination with multivariate statistical analysis enabled the assessment of differentially abundant lipids on the surface and in inner worm tissue, and on surfaces of males and females, respectively. The results are in line with literature but enhance the knowledge on surface composition at the lipid species level. For the first time, male and female were distinguished on the metabolic level by MSI. In sum, the developed tools can be further adapted to a large variety of other bioorganisms and analytical problems. Now, that the toolboxes are available to our workgroup and to the scientific community, a variety of follow-up studies are possible. Cutting of small objects by microembedding is now used extensively by the group e.g. for cutting Drosophila melanogaster embryos or even D. melanogaster brains. To characterize the inner tissues and organs of schistosomes, an organ isolation protocol can be used to harvest reproductive organs and built up a metabolic organ atlas for S. mansoni.107 In perspective, this could help to interpret complex MSI data obtained from sections and gain in-depth knowledge about biological processes when combined with other ‘omics data. To come one step closer to the goal of developing anti- schistosomal drugs, the distribution of a variety of active pharmaceutical ingredients can be tested in vitro. Investigations by MSI gather the potential to unravel mechanisms of uptake, metabolization and drug targets, possible sites of molecular interaction, e.g. starting with the gold-standard PZQ and going on to promising candidates such as Imatinib.12,108 MSI analysis on the metabolic level can be a valuable tool to get hints on key molecular pathways involved in maturation from virgin to pairing-experienced worms.77 Hindering the maturation and thus egg production can be a major goal for developing novel therapeutics to prevent devastating pathology. However, typically eggs get trapped in venules of the liver where they cause severe inflammation, because the hosts’ immune system is unable to degrade eggs timely. In addition to granuloma formation, schistosomiasis bears greater risk to develop hepatocellular carcinoma.109 MSI could be used here to visualize lipids and metabolites in hepatocytes, linked to tumor genesis. In sum, we initiated all these steps, potentially leading to many follow-up projects. Now that the surface of S. mansoni has been described, additional genera of schistosomes could be investigated such as S. japonicum, S. haematobium or S. mekongi, for instance. This is especially important because these species are endemic to different geographical and host-anatomical areas. S. japonicum for example resides in the venous complex of the bladder. In addition, the severity of pathology is different across strains. In the context of understanding biochemical pathways or interactions on the molecular level, fundamental knowledge is lacking on role and function of individual lipids and specific lipid classes, especially in S. mansoni parasites. Further studies should be conducted to shed light on molecular mechanisms. Especially enzymes seem to play an important role in 27 schistosomes, because de novo synthesis of certain fatty acids is not possible. Nevertheless, schistosomes developed an extraordinary lipid biology, for instance very-long chain fatty acids, odd-chain fatty acids or uncommon double-bond positional isomers.13,81,110,111 Substrates and enzymes catalyzing formation of such lipids are usually unknown. Since MS investigations can locate the presence of such unusual lipids, first hints towards their formation can be obtained using the methods developed and applied during this thesis. References (1) Spengler, B.; Hubert, M.; Kaufmann, R. MALDI Ion Imaging and Biological Ion Imaging with a new Scanning UV-Laser Microprobe. 42nd Annual Conf. On Mass Spectrom. and Allied Topics 1994, 1041. (2) Spengler, B. Mass Spectrometry Imaging of Biomolecular Information. Analytical Chemistry 2015, 87, 64-82. (3) Paine, M. R. L.; Kooijman, P. C.; Fisher, G. L.; Heeren, R. M. A.; Fernandez, F. M.; Ellis, S. R. Visualizing molecular distributions for biomaterials applications with mass spectrometry imaging: a review. Journal of Materials Chemistry B 2017, 5, 7444-7460. (4) Amstalden van Hove, E. R.; Smith, D. F.; Heeren, R. M. A. A concise review of mass spectrometry imaging. Journal of Chromatography A 2010, 1217, 3946-3954. (5) Koestler, M.; Kirsch, D.; Hester, A.; Leisner, A.; Guenther, S.; Spengler, B. A high- resolution scanning microprobe matrix-assisted laser desorption/ionization ion source for imaging analysis on an ion trap/Fourier transform ion cyclotron resonance mass spectrometer. Rapid communications in mass spectrometry : RCM 2008, 22, 3275-3285. (6) Ogrinc Potočnik, N.; Porta, T.; Becker, M.; Heeren, R. M. A.; Ellis, S. R. Use of advantageous, volatile matrices enabled by next-generation high-speed matrix-assisted laser desorption/ionization time-of-flight imaging employing a scanning laser beam. Rapid Communications in Mass Spectrometry 2015, 29, 2195-2203. (7) Kompauer, M.; Heiles, S.; Spengler, B. Autofocusing MALDI mass spectrometry imaging of tissue sections and 3D chemical topography of nonflat surfaces. Nature Methods 2017, 14, 1156-1158. (8) Soltwisch, J.; Kettling, H.; Vens-Cappell, S.; Wiegelmann, M.; Müthing, J.; Dreisewerd, K. Mass spectrometry imaging with laser-induced postionization. Science (New York, N.Y.) 2015, 348, 211-215. (9) Kettling, H.; Vens-Cappell, S.; Soltwisch, J.; Pirkl, A.; Haier, J.; Müthing, J.; Dreisewerd, K. MALDI Mass Spectrometry Imaging of Bioactive Lipids in Mouse Brain with a Synapt G2-S Mass Spectrometer Operated at Elevated Pressure: Improving the Analytical Sensitivity and the Lateral Resolution to Ten Micrometers. Analytical Chemistry 2014, 86, 7798-7805. (10) Korte, A. R.; Yandeau-Nelson, M. D.; Nikolau, B. J.; Lee, Y. J. Subcellular-level resolution MALDI-MS imaging of maize leaf metabolites by MALDI-linear ion trap-Orbitrap mass spectrometer. Analytical and bioanalytical chemistry 2015, 407, 2301-2309. (11) Kompauer, M.; Heiles, S.; Spengler, B. Atmospheric pressure MALDI mass spectrometry imaging of tissues and cells at 1.4-μm lateral resolution. Nature Methods 2017, 14, 90-96. 28 (12) McManus, D. P.; Dunne, D. W.; Sacko, M.; Utzinger, J.; Vennervald, B. J.; Zhou, X.-N. Schistosomiasis. Nature Reviews Disease Primers 2018, 4, 13. (13) Brouwers, J. F. H. M.; Gadella, B. M.; van Golde, L. M. G.; Tielens, A. G. M. Quantitative analysis of phosphatidylcholine molecular species using HPLC and light scattering detection. Journal of Lipid Research 1998, 39, 344-353. (14) Marshall, A. G.; Hendrickson, C. L.; Shi, S. D.-H. Peer reviewed: scaling ms plateaus with high-resolution FT-ICRMS. Analytical Chemistry 2002, 74, 252-259. (15) Brenton, A. G.; Godfrey, A. R. Accurate Mass Measurement: Terminology and Treatment of Data. Journal of the American Society for Mass Spectrometry 2010, 21, 1821- 1835. (16) Karas, M.; Hillenkamp, F. Laser desorption ionization of proteins with molecular masses exceeding 10,000 daltons. Analytical Chemistry 1988, 60, 2299-2301. (17) Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse, C. M. Electrospray ionization–principles and practice. Mass Spectrometry Reviews 1990, 9, 37-70. (18) El-Aneed, A.; Cohen, A.; Banoub, J. Mass Spectrometry, Review of the Basics: Electrospray, MALDI, and Commonly Used Mass Analyzers. Applied Spectroscopy Reviews 2009, 44, 210-230. (19) Strupat, K.; Karas, M.; Hillenkamp, F. 2,5-Dihydroxybenzoic acid: a new matrix for laser desorption—ionization mass spectrometry. International Journal of Mass Spectrometry and Ion Processes 1991, 111, 89-102. (20) Karas, M.; Glückmann, M.; Schäfer, J. Ionization in matrix-assisted laser desorption/ionization: singly charged molecular ions are the lucky survivors. Journal of Mass Spectrometry 2000, 35, 1-12. (21) Fenn, J.; Mann, M.; Meng, C.; Wong, S.; Whitehouse, C. Electrospray ionization for mass spectrometry of large biomolecules. Science (New York, N.Y.) 1989, 246, 64-71. (22) Jemal, M. High-throughput quantitative bioanalysis by LC/MS/MS. Biomedical Chromatography 2000, 14, 422-429. (23) Makarov, A. Electrostatic Axially Harmonic Orbital Trapping:  A High-Performance Technique of Mass Analysis. Analytical Chemistry 2000, 72, 1156-1162. (24) Hardman, M.; Makarov, A. A. Interfacing the Orbitrap Mass Analyzer to an Electrospray Ion Source. Analytical Chemistry 2003, 75, 1699-1705. (25) Perry, R. H.; Cooks, R. G.; Noll, R. J. Orbitrap mass spectrometry: Instrumentation, ion motion and applications. Mass Spectrometry Reviews 2008, 27, 661-699. (26) Eliuk, S.; Makarov, A. Evolution of Orbitrap Mass Spectrometry Instrumentation. Annual Review of Analytical Chemistry 2015, 8, 61-80. (27) Hu, Q.; Noll, R. J.; Li, H.; Makarov, A.; Hardman, M.; Graham Cooks, R. The Orbitrap: a new mass spectrometer. Journal of Mass Spectrometry 2005, 40, 430-443. (28) Olsen, J. V.; de Godoy, L. M. F.; Li, G.; Macek, B.; Mortensen, P.; Pesch, R.; Makarov, A.; Lange, O.; Horning, S.; Mann, M. Parts per Million Mass Accuracy on an Orbitrap Mass Spectrometer via Lock Mass Injection into a C-trap. Mol Cell Proteomics 2005, 4, 2010-2021. 29 (29) Olsen, J. V.; Macek, B.; Lange, O.; Makarov, A.; Horning, S.; Mann, M. Higher-energy C-trap dissociation for peptide modification analysis. Nature Methods 2007, 4, 709-712. (30) Swales, J. G.; Hamm, G.; Clench, M. R.; Goodwin, R. J. Mass spectrometry imaging and its application in pharmaceutical research and development: a concise review. International Journal of Mass Spectrometry 2019, 437, 99-112. (31) Römpp, A.; Spengler, B. Mass spectrometry imaging with high resolution in mass and space. Histochemistry and Cell Biology 2013, 139, 759-783. (32) Bodzon-Kulakowska, A.; Suder, P. Imaging mass spectrometry: Instrumentation, applications, and combination with other visualization techniques. Mass Spectrometry Reviews 2016, 35, 147-169. (33) Bordeaux, J.; Welsh, A. W.; Agarwal, S.; Killiam, E.; Baquero, M. T.; Hanna, J. A.; Anagnostou, V. K.; Rimm, D. L. Antibody validation. BioTechniques 2010, 48, 197-209. (34) Altelaar, A. F.; van Minnen, J.; Jimenez, C. R.; Heeren, R. M.; Piersma, S. R. Direct molecular imaging of Lymnaea stagnalis nervous tissue at subcellular spatial resolution by mass spectrometry. Anal Chem 2005, 77, 735-741. (35) Takats, Z.; Wiseman, J. M.; Gologan, B.; Cooks, R. G. Mass spectrometry sampling under ambient conditions with desorption electrospray ionization. Science (New York, N.Y.) 2004, 306, 471-473. (36) Wiseman, J. M.; Puolitaival, S. M.; Takáts, Z.; Cooks, R. G.; Caprioli, R. M. Mass Spectrometric Profiling of Intact Biological Tissue by Using Desorption Electrospray Ionization. Angewandte Chemie International Edition 2005, 44, 7094-7097. (37) Campbell, D. I.; Ferreira, C. R.; Eberlin, L. S.; Cooks, R. G. Improved spatial resolution in the imaging of biological tissue using desorption electrospray ionization. Analytical and bioanalytical chemistry 2012, 404, 389-398. (38) Makarov, A.; Denisov, E.; Kholomeev, A.; Balschun, W.; Lange, O.; Strupat, K.; Horning, S. Performance evaluation of a hybrid linear ion trap/orbitrap mass spectrometer. Analytical chemistry 2006, 78, 2113-2120. (39) Zavalin, A.; Yang, J.; Hayden, K.; Vestal, M.; Caprioli, R. M. Tissue protein imaging at 1 mum laser spot diameter for high spatial resolution and high imaging speed using transmission geometry MALDI TOF MS. Analytical and bioanalytical chemistry 2015, 407, 2337-2342. (40) Palmer, A.; Phapale, P.; Chernyavsky, I.; Lavigne, R.; Fay, D.; Tarasov, A.; Kovalev, V.; Fuchser, J.; Nikolenko, S.; Pineau, C.; Becker, M.; Alexandrov, T. FDR-controlled metabolite annotation for high-resolution imaging mass spectrometry. Nature Methods 2016, 14, 57-61. (41) Fahy, E.; Sud, M.; Cotter, D.; Subramaniam, S. LIPID MAPS online tools for lipid research. Nucleic Acids Research 2007, 35, W606-W612. (42) Sud, M.; Fahy, E.; Cotter, D.; Brown, A.; Dennis, E. A.; Glass, C. K.; Merrill, A. H., Jr.; Murphy, R. C.; Raetz, C. R.; Russell, D. W.; Subramaniam, S. LMSD: LIPID MAPS structure database. Nucleic Acids Research 2007, 35, D527-532. (43) Bridge, A.; Gleizes, A.; Niknejad, A.; Aimo, L.; Bougueleret, L.; Hyka-Nouspikel, N.; Xenarios, I.; Kuznetsov, D.; Götz, L.; Liechti, R.; David, F. P. A.; van der Goot, F. G.; 30 Riezman, H. The SwissLipids knowledgebase for lipid biology. Bioinformatics 2015, 31, 2860-2866. (44) Wishart, D. S.; Tzur, D.; Knox, C.; Eisner, R.; Guo, A. C.; Young, N.; Cheng, D.; Jewell, K.; Arndt, D.; Sawhney, S.; Fung, C.; Nikolai, L.; Lewis, M.; Coutouly, M.-A.; Forsythe, I.; Tang, P.; Shrivastava, S.; Jeroncic, K.; Stothard, P.; Amegbey, G., et al. HMDB: the Human Metabolome Database. Nucleic Acids Research 2007, 35, D521-526. (45) Wishart, D. S.; Knox, C.; Guo, A. C.; Eisner, R.; Young, N.; Gautam, B.; Hau, D. D.; Psychogios, N.; Dong, E.; Bouatra, S.; Mandal, R.; Sinelnikov, I.; Xia, J.; Jia, L.; Cruz, J. A.; Lim, E.; Sobsey, C. A.; Shrivastava, S.; Huang, P.; Liu, P., et al. HMDB: a knowledgebase for the human metabolome. Nucleic Acids Research 2009, 37, D603-610. (46) Wishart, D. S.; Jewison, T.; Guo, A. C.; Wilson, M.; Knox, C.; Liu, Y.; Djoumbou, Y.; Mandal, R.; Aziat, F.; Dong, E.; Bouatra, S.; Sinelnikov, I.; Arndt, D.; Xia, J.; Liu, P.; Yallou, F.; Bjorndahl, T.; Perez-Pineiro, R.; Eisner, R.; Allen, F., et al. HMDB 3.0--The Human Metabolome Database in 2013. Nucleic Acids Research 2013, 41, D801-807. (47) Wishart, D. S.; Feunang, Y. D.; Marcu, A.; Guo, A. C.; Liang, K.; Vazquez-Fresno, R.; Sajed, T.; Johnson, D.; Li, C.; Karu, N.; Sayeeda, Z.; Lo, E.; Assempour, N.; Berjanskii, M.; Singhal, S.; Arndt, D.; Liang, Y.; Badran, H.; Grant, J.; Serra-Cayuela, A., et al. HMDB 4.0: the human metabolome database for 2018. Nucleic Acids Research 2018, 46, D608-617. (48) Ellis, S. R.; Paine, M. R. L.; Eijkel, G. B.; Pauling, J. K.; Husen, P.; Jervelund, M. W.; Hermansson, M.; Ejsing, C. S.; Heeren, R. M. A. Automated, parallel mass spectrometry imaging and structural identification of lipids. Nature Methods 2018, 15, 515-518. (49) Liebisch, G.; Vizcaino, J. A.; Kofeler, H.; Trotzmuller, M.; Griffiths, W. J.; Schmitz, G.; Spener, F.; Wakelam, M. J. Shorthand notation for lipid structures derived from mass spectrometry. Journal of Lipid Research 2013, 54, 1523-1530. (50) Moss, G. P.; Smith, P. A. S.; Tavernier, D. In Pure and Applied Chemistry; DeGruyter, 1995, pp 1307-1375. (51) Nelson, D. L.; Cox, M. M.; W.H. Freeman and Company: New York, 2004, pp 343-368. (52) Fahy, E.; Subramaniam, S.; Brown, H. A.; Glass, C. K.; Merrill, A. H., Jr.; Murphy, R. C.; Raetz, C. R.; Russell, D. W.; Seyama, Y.; Shaw, W.; Shimizu, T.; Spener, F.; van Meer, G.; VanNieuwenhze, M. S.; White, S. H.; Witztum, J. L.; Dennis, E. A. A comprehensive classification system for lipids. Journal of Lipid Research 2005, 46, 839-861. (53) Fahy, E.; Subramaniam, S.; Murphy, R. C.; Nishijima, M.; Raetz, C. R.; Shimizu, T.; Spener, F.; van Meer, G.; Wakelam, M. J.; Dennis, E. A. Update of the LIPID MAPS comprehensive classification system for lipids. J Lipid Res 2009, 50 Suppl, S9-14. (54) Martinez-Seara, H.; Róg, T.; Pasenkiewicz-Gierula, M.; Vattulainen, I.; Karttunen, M.; Reigada, R. Effect of Double Bond Position on Lipid Bilayer Properties:  Insight through Atomistic Simulations. The Journal of Physical Chemistry B 2007, 111, 11162-11168. (55) Bolla, J. R.; Agasid, M. T.; Mehmood, S.; Robinson, C. V. Membrane Protein–Lipid Interactions Probed Using Mass Spectrometry. Annual Review of Biochemistry 2019, 88, 85- 111. 31 (56) Fuchs, B.; Bischoff, A.; Süß, R.; Teuber, K.; Schürenberg, M.; Suckau, D.; Schiller, J. Phosphatidylcholines and -ethanolamines can be easily mistaken in phospholipid mixtures: a negative ion MALDI-TOF MS study with 9-aminoacridine as matrix and egg yolk as selected example. Analytical and bioanalytical chemistry 2009, 395, 2479-2487. (57) Han, X.; Gross, R. W. Shotgun lipidomics: electrospray ionization mass spectrometric analysis and quantitation of cellular lipidomes directly from crude extracts of biological samples. Mass spectrometry reviews 2005, 24, 367-412. (58) Peterson, B. L.; Cummings, B. S. A review of chromatographic methods for the assessment of phospholipids in biological samples. Biomedical Chromatography 2006, 20, 227-243. (59) Zhang, A.; Sun, H.; Wang, P.; Han, Y.; Wang, X. Modern analytical techniques in metabolomics analysis. Analyst 2012, 137, 293-300. (60) Bligh, E. G.; Dyer, W. J. A RAPID METHOD OF TOTAL LIPID EXTRACTION AND PURIFICATION. Canadian Journal of Biochemistry and Physiology 1959, 37, 911-917. (61) Matyash, V.; Liebisch, G.; Kurzchalia, T. V.; Shevchenko, A.; Schwudke, D. Lipid extraction by methyl-tert-butyl ether for high-throughput lipidomics. Journal of lipid research 2008, 49, 1137-1146. (62) Feinstein, L. M.; Sul, W. J.; Blackwood, C. B. Assessment of Bias Associated with Incomplete Extraction of Microbial DNA from Soil. Applied and Environmental Microbiology 2009, 75, 5428-5433. (63) Lesage, V.; Morin, Y.; Rioux, È.; Pomerleau, C.; Ferguson, S. H.; Pelletier, É. Stable isotopes and trace elements as indicators of diet and habitat use in cetaceans: predicting errors related to preservation, lipid extraction, and lipid normalization. Marine Ecology Progress Series 2010, 419, 249-265. (64) Mintenbeck, K.; Brey, T.; Jacob, U.; Knust, R.; Struck, U. How to account for the lipid effect on carbon stable-isotope ratio (δ13C): sample treatment effects and model bias. Journal of Fish Biology 2008, 72, 815-830. (65) Leary, D. H.; Hervey, W. J.; Deschamps, J. R.; Kusterbeck, A. W.; Vora, G. J. Which metaproteome? The impact of protein extraction bias on metaproteomic analyses. Molecular and Cellular Probes 2013, 27, 193-199. (66) Ejsing, C. S.; Sampaio, J. L.; Surendranath, V.; Duchoslav, E.; Ekroos, K.; Klemm, R. W.; Simons, K.; Shevchenko, A. Global analysis of the yeast lipidome by quantitative shotgun mass spectrometry. Proceedings of the National Academy of Sciences 2009, 106, 2136-2141. (67) Rampler, E.; Coman, C.; Hermann, G.; Sickmann, A.; Ahrends, R.; Koellensperger, G. LILY-lipidome isotope labeling of yeast: in vivo synthesis of 13C labeled reference lipids for quantification by mass spectrometry. Analyst 2017, 142, 1891-1899. (68) Rampler, E.; Criscuolo, A.; Zeller, M.; El Abiead, Y.; Schoeny, H.; Hermann, G.; Sokol, E.; Cook, K.; Peake, D. A.; Delanghe, B.; Koellensperger, G. A Novel Lipidomics Workflow for Improved Human Plasma Identification and Quantification Using RPLC-MSn Methods and Isotope Dilution Strategies. Analytical Chemistry 2018, 90, 6494-6501. 32 (69) Pedersen, H. K.; Gudmundsdottir, V.; Nielsen, H. B.; Hyotylainen, T.; Nielsen, T.; Jensen, B. A.; Forslund, K.; Hildebrand, F.; Prifti, E.; Falony, G. Human gut microbes impact host serum metabolome and insulin sensitivity. Nature 2016, 535, 376-381. (70) Pedersen, H. K.; Forslund, S. K.; Gudmundsdottir, V.; Petersen, A. Ø.; Hildebrand, F.; Hyötyläinen, T.; Nielsen, T.; Hansen, T.; Bork, P.; Ehrlich, S. D.; Brunak, S.; Oresic, M.; Pedersen, O.; Nielsen, H. B. A computational framework to integrate high-throughput ‘-omics’ datasets for the identification of potential mechanistic links. Nature Protocols 2018, 13, 2781- 2800. (71) Hartler, J.; Triebl, A.; Ziegl, A.; Trotzmuller, M.; Rechberger, G. N.; Zeleznik, O. A.; Zierler, K. A.; Torta, F.; Cazenave-Gassiot, A.; Wenk, M. R.; Fauland, A.; Wheelock, C. E.; Armando, A. M.; Quehenberger, O.; Zhang, Q.; Wakelam, M. J. O.; Haemmerle, G.; Spener, F.; Kofeler, H. C.; Thallinger, G. G. Deciphering lipid structures based on platform- independent decision rules. Nature Methods 2017, 14, 1171-1174. (72) Hartler, J.; Trotzmuller, M.; Chitraju, C.; Spener, F.; Kofeler, H. C.; Thallinger, G. G. Lipid Data Analyzer: unattended identification and quantitation of lipids in LC-MS data. Bioinformatics 2011, 27, 572-577. (73) Pi, J.; Wu, X.; Feng, Y. Fragmentation patterns of five types of phospholipids by ultra- high-performance liquid chromatography electrospray ionization quadrupole time-of-flight tandem mass spectrometry. Analytical Methods 2016, 8, 1319-1332. (74) Hsu, F.-F.; Turk, J. Electrospray ionization with low-energy collisionally activated dissociation tandem mass spectrometry of glycerophospholipids: mechanisms of fragmentation and structural characterization. J Chromatogr B Analyt Technol Biomed Life Sci. 2009, 877, 2673-2695. (75) diseases, W. H. O.-D. o. c. o. n. t., Integrating Neglected Tropical Diseases into Global Health and Development: Fourth WHO report on neglected tropical diseases: Geneva, 2017. (76) Grevelding, C. G. The female-specific W1 sequence of the Puerto Rican strain of Schistosoma mansoni occurs in both genders of a Liberian strain. Molecular and Biochemical Parasitology 1995, 71, 269-272. (77) Lu, Z.; Sessler, F.; Holroyd, N.; Hahnel, S.; Quack, T.; Berriman, M.; Grevelding, C. G. Schistosome sex matters: a deep view into gonad-specific and pairing-dependent transcriptomes reveals a complex gender interplay. Scientific Reports 2016, 6, 31150. (78) Yang, G.-J.; Bergquist, R. Potential Impact of Climate Change on Schistosomiasis: A Global Assessment Attempt. Tropical medicine and infectious disease 2018, 3, 117. (79) McCreesh, N.; Nikulin, G.; Booth, M. Predicting the effects of climate change on Schistosoma mansoni transmission in eastern Africa. Parasites & Vectors 2015, 8, 4. (80) Bergquist, R.; Zhou, X.-N.; Rollinson, D.; Reinhard-Rupp, J.; Klohe, K. Elimination of schistosomiasis: the tools required. Infectious Diseases of Poverty 2017, 6, 158. (81) Retra, K.; Bleijerveld, O. B.; van Gestel, R. A.; Tielens, A. G.; van Hellemond, J. J.; Brouwers, J. F. A simple and universal method for the separation and identification of phospholipid molecular species. Rapid Communications in Mass Spectrometry 2008, 22, 1853-1862. 33 (82) Ferreira, M. S.; Oliveira, D. N.; Oliveira, R. N.; Allegretti, S. M.; Vercesi, A. E.; Catharino, R. R. Mass spectrometry imaging: a new vision in differentiating Schistosoma mansoni strains. Journal of Mass Spectrometry 2014, 49, 86-92. (83) Ferreira, M. S.; de Oliveira, R. N.; de Oliveira, D. N.; Esteves, C. Z.; Allegretti, S. M.; Catharino, R. R. Revealing praziquantel molecular targets using mass spectrometry imaging: an expeditious approach applied to Schistosoma mansoni. International journal for parasitology 2015, 45, 385-391. (84) Retra, K.; deWalick, S.; Schmitz, M.; Yazdanbakhsh, M.; Tielens, A. G. M.; Brouwers, J. F. H. M.; van Hellemond, J. J. The tegumental surface membranes of Schistosoma mansoni are enriched in parasite-specific phospholipid species. International Journal for Parasitology 2015, 45, 629-636. (85) Giera, M.; Kaisar, M. M.; Derks, R. J.; Steenvoorden, E.; Kruize, Y. C.; Hokke, C. H.; Yazdanbakhsh, M.; Everts, B. The Schistosoma mansoni lipidome: Leads for immunomodulation. Analytica chimica acta 2018, 1037, 107-118. (86) Fox, C. H.; Johnson, F. B.; Whiting, J.; Roller, P. P. Formaldehyde fixation. Journal of Histochemistry & Cytochemistry 1985, 33, 845-853. (87) Ly, A.; Buck, A.; Balluff, B.; Sun, N.; Gorzolka, K.; Feuchtinger, A.; Janssen, K.-P.; Kuppen, P. J. K.; van de Velde, C. J. H.; Weirich, G.; Erlmeier, F.; Langer, R.; Aubele, M.; Zitzelsberger, H.; McDonnell, L.; Aichler, M.; Walch, A. High-mass-resolution MALDI mass spectrometry imaging of metabolites from formalin-fixed paraffin-embedded tissue. Nature Protocols 2016, 11, 1428-1443. (88) Pietrowska, M.; Gawin, M.; Polańska, J.; Widłak, P. Tissue fixed with formalin and processed without paraffin embedding is suitable for imaging of both peptides and lipids by MALDI‐IMS. Proteomics 2016, 16, 1670-1677. (89) Kawamoto, T. Use of a new adhesive film for the preparation of multi-purpose fresh- frozen sections from hard tissues, whole-animals, insects and plants. Archives of histology and cytology 2003, 66, 123-143. (90) Khalil, S. M.; Rompp, A.; Pretzel, J.; Becker, K.; Spengler, B. Phospholipid Topography of Whole-Body Sections of the Anopheles stephensi Mosquito, Characterized by High- Resolution Atmospheric-Pressure Scanning Microprobe Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry Imaging. Analytical Chemistry 2015, 87, 11309- 11316. (91) Brignole-Baudouin, F.; Desbenoit, N.; Hamm, G.; Liang, H.; Both, J. P.; Brunelle, A.; Fournier, I.; Guerineau, V.; Legouffe, R.; Stauber, J.; Touboul, D.; Wisztorski, M.; Salzet, M.; Laprevote, O.; Baudouin, C. A new safety concern for glaucoma treatment demonstrated by mass spectrometry imaging of benzalkonium chloride distribution in the eye, an experimental study in rabbits. PloS one 2012, 7, e50180. (92) Hameed, S.; Ikegami, K.; Sugiyama, E.; Matsushita, S.; Kimura, Y.; Hayasaka, T.; Sugiura, Y.; Masaki, N.; Waki, M.; Ohta, I.; Hossen, M. A.; Setou, M. Direct profiling of the phospholipid composition of adult Caenorhabditis elegans using whole-body imaging mass spectrometry. Analytical and bioanalytical chemistry 2015, 407, 7589-7602. 34 (93) Niehoff, A.-C.; Schulz, J.; Soltwisch, J.; Meyer, S.; Kettling, H.; Sperling, M.; Jeibmann, A.; Dreisewerd, K.; Francesconi, K. A.; Schwerdtle, T.; Karst, U. Imaging by Elemental and Molecular Mass Spectrometry Reveals the Uptake of an Arsenolipid in the Brain of Drosophila melanogaster. Analytical Chemistry 2016, 88, 5258-5263. (94) Khalil, S. M.; Pretzel, J.; Becker, K.; Spengler, B. High-resolution AP-SMALDI mass spectrometry imaging of Drosophila melanogaster. International Journal of Mass Spectrometry 2017, 416, 1-19. (95) Rujoi, M.; Estrada, R.; Yappert, M. C. In situ MALDI-TOF MS regional analysis of neutral phospholipids in lens tissue. Analytical Chemistry 2004, 76, 1657-1663. ( 6) Schober, .; Guenther, S.; Spengler, B.; R mpp, A. Single cell matrix-assisted laser desorption/ionization mass spectrometry imaging. Analytical Chemistry 2012, 84, 6293-6297. (97) Redman, C. A.; Kennington, S.; Spathopoulou, T.; Kusel, J. R. Interconversion of sphingomyelin and ceramide in adult Schistosoma mansoni. Molecular and Biochemical Parasitology 1997, 90, 145-153. (98) Van Hellemond, J. J.; Retra, K.; Brouwers, J. F. H. M.; van Balkom, B. W. M.; Yazdanbakhsh, M.; Shoemaker, C. B.; Tielens, A. G. M. Functions of the tegument of schistosomes: Clues from the proteome and lipidome. International Journal for Parasitology 2006, 36, 691-699. (99) McLaren, D. J.; Hockley, D. J. Blood flukes have a double outer membrane. Nature 1977, 269, 147-149. (100) Kadesch, P.; Quack, T.; Gerbig, S.; Grevelding, C. G.; Spengler, B. Lipid Topography in Schistosoma mansoni Cryosections, Revealed by Microembedding and High-Resolution Atmospheric-Pressure Matrix-Assisted Laser Desorption/Ionization (MALDI) Mass Spectrometry Imaging. Analytical Chemistry 2019, 91, 4520-4528. (101) Jäger, B. P.; Rudolph, P. E. In Statistik : Verstehen durch Experimente mit SAS; De Gruyter: Berlin, 2016, p 398. (102) Huberty, C. J.; Petoskey, M. D. In Handbook of Applied Multivariate Statistics and Mathematical Modeling, Tinsley, H. E. A.; Brown, S. D., Eds.; Academic Press: San Diego, 2000, pp 183-208. (103) Anton, H. In Elementary Linear Algebra; Wiley: New York, 2013, p 145. (104) Meyer zu Heringdorf, D.; Jakobs, K. H. Lysophospholipid receptors: Signalling, pharmacology and regulation by lysophospholipid metabolism. Biochimica et Biophysica Acta (BBA) - Biomembranes 2007, 1768, 923-940. (105) Spiegel, S.; Foster, D.; Kolesnick, R. Signal transduction through lipid second messengers. Current Opinion in Cell Biology 1996, 8, 159-167. (106) Bexkens, M. L.; Mebius, M. M.; Houweling, M.; Brouwers, J. F.; Tielens, A. G. M.; van Hellemond, J. J. Schistosoma mansoni does not and cannot oxidise fatty acids, but these are used for biosynthetic purposes instead. International Journal for Parasitology 2019, 49, 647- 656. 35 (107) Hahnel, S.; Lu, Z.; Wilson, R. A.; Grevelding, C. G.; Quack, T. Whole-Organ Isolation Approach as a Basis for Tissue-Specific Analyses in Schistosoma mansoni. PLOS Neglected Tropical Diseases 2013, 7, e2336. (108) Beckmann, S.; Grevelding, C. Imatinib has a fatal impact on morphology, pairing stability and survival of adult Schistosoma mansoni in vitro. International journal for parasitology 2010, 40, 521-526. (109) Roderfeld, M.; Padem, S.; Lichtenberger, J.; Quack, T.; Weiskirchen, R.; Longerich, T.; Schramm, G.; Churin, Y.; Irungbam, K.; Tschuschner, A. Schistosoma mansoni Egg– Secreted Antigens Activate Hepatocellular Carcinoma–Associated Transcription Factors c‐Jun and STAT3 in Hamster and Human Hepatocytes. Hepatology 2018. (110) Brouwers, J. F.; Versluis, C.; van Golde, L. M.; Tielens, A. G. 5-Octadecenoic acid: evidence for a novel type of fatty acid modification in schistosomes. The Biochemical journal 1998, 334 ( Pt 2), 315-319. (111) Esch, P.; Heiles, S. Charging and Charge Switching of Unsaturated Lipids and Apolar Compounds Using Paternò-Büchi Reactions. Journal of The American Society for Mass Spectrometry 2018, 29, 1971-1980. 36 Chapter II - Publication 1 Lipid Topography in Schistosoma mansoni Cryosections, Revealed by Microembedding and High-Resolution Atmospheric-Pressure Matrix-Assisted Laser Desorption/Ionization (MALDI) Mass Spectrometry Imaging Kadesch, Patrik*; Quack, Thomas#; Gerbig, Stefanie*; Grevelding, Christoph G.#; Spengler, Bernhard* * Institute of Inorganic and Analytical Chemistry, Justus Liebig University Giessen, Giessen, Germany # Institute of Parasitology, Justus Liebig University Giessen, Giessen, Germany Analytical Chemistry 2019, 91 (7), pp 4520-4528 DOI: 10.1021/acs.analchem.8b05440 37 Reprinted with permission from [Anal. Chem. 2019, 91, 4520–4528]. Copyright [2019] American Chemical Society. 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 Chapter III - Publication 2 Tissue- and sex-specific lipidomic analysis of Schistosoma mansoni using high-resolution atmospheric pressure scanning microprobe matrix-assisted laser desorption/ionization mass spectrometry imaging Kadesch, Patrik*; Quack, Thomas#; Gerbig, Stefanie*; Grevelding, Christoph G.#; Spengler, Bernhard* * Institute of Inorganic and Analytical Chemistry, Justus Liebig University Giessen, Giessen, Germany # Institute of Parasitology, Justus Liebig University Giessen, Giessen, Germany PLOS Neglected Tropical Diseases 2020, 14(5), e0008145 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 S1 Table. Specifications of chemicals used in experimental section. https://doi.org/10.1371/journal.pntd.0008145.s001 73 S2 Table. UHPLC-MS method for identification of lipids in S. mansoni. Injection volume was 50 μL. https://doi.org/10.1371/journal.pntd.0008145.s002 74 S3 Table. Detailed lipid annotations of classified lipids from comparison of tegument and inner worm tissue. https://doi.org/10.1371/journal.pntd.0008145.s003 75 76 77 78 79 80 81 82 83 84 85 86 S4 Table. Detailed lipid annotations of classified lipids from comparison of male and female tegument. https://doi.org/10.1371/journal.pntd.0008145.s004 87 88 89 90 91 92 93 94 95 96 S1 Fig. Graphical illustration of LC-MS2 data analysis workflow. https://doi.org/10.1371/journal.pntd.0008145.s005 97 S2 Fig. Graphical illustration of the data analysis workflow for MS imaging data. Statistical evaluation work flow was adapted from literature. The statistical analysis comprised five key steps: 1. normalization of one signal to the sum of all signals per measurement, 2. z-score (using median), 3. multiple-class analysis of variance (ANOVA, permutation based false-discovery-rate, FDR, set to 5%, 250 restarts), 4. post-hoc test (5% FDR) and 5. hierarchical clustering (Euclidean distance using average linkage, preprocessing with k-means, maximum 10 iterations, 10 restarts). https://doi.org/10.1371/journal.pntd.0008145.s006 98 S3 Fig. Graphical illustration (from Perseus[41]) of multivariate statistical analysis and categorization of differentially abundant signals from MALDI experiments by hierarchical clustering. https://doi.org/10.1371/journal.pntd.0008145.s007 99 S4 Fig. Digital light microscopic images of male (M) surfaces (left), female (F) surfaces (middle) and cryosections of couples (right). The black arrows indicate the anterior end. https://doi.org/10.1371/journal.pntd.0008145.s008 100 S5 Fig. Distribution of previously reported most abundant lipid species in S. mansoni as protonated and sodiated ion species[20] PC (34:1) has been determined in the past to be differentially abundant in whole worm and tegument. [10] However, MS imaging data did not show significant differences based on HC. For PC (36:1) and PC (36:2), however, our findings are well in accordance with previous publications which found higher abundances inside the worm.[10] The same trend is suggested by unsupervised MS imaging data evaluation presented here. https://doi.org/10.1371/journal.pntd.0008145.s009 101 S6 Fig. Number of carbon atoms in fatty acyl chains vs the number of double bonds detected in phosphatidylethanolamines (PE). Isobaric PE/PC interferences were excluded for surface and section data. A–Comparison of worm-tissue (blue cross) and surface/tegument specific signals vs ions (orange +) with unspecific distribution (green square). Overlapping indicators are attributed to the presence of several adducts corresponding to one lipid species. B–Arithmetic mean fatty acyl and double bond composition for section/inner tissue (blue), surface/tegument (orange) and unspecific signals (green). Error bars show the standard deviation across one location. https://doi.org/10.1371/journal.pntd.0008145.s010 102 S7 Fig. Example for LC-MS/MS based identification of PE (37:4) with MS1 overview spectra and data-dependent MS2 spectra. A–MS1 overview spectrum. B–virtual magnification of mass range m/z 700–800 (from A) showing the mass of PE (37:4) as deprotonated species (C42H75NO8P). C–MS2 spectrum of precursor m/z 752.52 ±0.5 u showing characteristic fragments of PE head group (around m/z 140 and m/z 196), FA (17:0) and FA (20:4). The precursor is not visible in the spectrum and assumedly fragmented quantitatively at NCE = 30. D–virtual magnification of m/z 750– 755 (from A) showing mass and isotope ratio of PE 37:4 as 12C, 13C and 13C2 isotopologues. https://doi.org/10.1371/journal.pntd.0008145.s011 103 S8 Fig. Example for mass accuracy and resolution obtained in MSI experiments. A–MS ion image of nearly isobaric PE-adduct species [PE (39:5) + H]+ and [PE (37:2) + Na]+ (Δm = 3.1 ppm) at m/z 780.5537 ± 5 ppm. B–Signal intensity (abundance in NL; normalized level) vs mass deviation in ppm. A double peak can be observed shifted by approximately 1.5 ppm and 2.8 ppm. C–MS ion signal at m/z 780.55254 ± 0.2 ppm showing an increased signal intensity on the worm surface assigned to protonated PE (39:5). D–MS ion at m/z 780.55151 ± 0.2 ppm assigned to PE (37:2) as sodium adduct. By hierarchical clustering, the signal at m/z 780.5537 was determined to be more abundant in the worm body compared to tegumental surface (see Fig 3). The signal was assigned to PE (37:2) as sodiated molecule. The protonated species of PE (39:5), however, was classified as unspecific. The fluctuating signal intensity of the surface measurements putatively led to unspecific classification. This example thus verifies the accuracy and correctness of HC- based classification. https://doi.org/10.1371/journal.pntd.0008145.s012 104 S9 Fig. Putative adducts of PE (39:4) with different distributions. A—Distribution of m/z 782.5694 assigned to [PE (39:4) + H]+. B–distribution of m/z 804.5514 assigned to [PE (39:4) + Na]+. This difference in distribution could be explained by different concentrations of salt in tegument and inner tissue or by isobaric interferences that were not contained in the LC-MS/MS-database. https://doi.org/10.1371/journal.pntd.0008145.s013 105 Curriculum vitae The curriculum vitae was removed from the electronic version of the paper. 106 The curriculum vitae was removed from the electronic version of the paper. 107 Acknowledgements First, I would like to thank Prof. Dr. Bernhard Spengler for giving me the opportunity to work in his research group and for the proposal of such an interesting research project. I enjoyed working with you a lot. Thank you also for hosting such an international research group in Giessen and for your constant support travelling to national and international conferences and giving me the opportunity to represent our group in front of other, internationally recognized scientists. Thank you for your trust. I would like to thank Dr. Stefanie Gerbig for the supervision of my scientific work and especially for guiding me through the projects. Thank you for always being there. You have taught me how to become a better scientist. I am very grateful to Prof. Dr. Christoph C. Grevelding and Dr. Thomas Quack for their ideas, fruitful discussions and constant scientific input. Thank you for being the enthusiastic parasitologists and researchers you are. Thank you to all members of the analytical chemistry department. I really enjoyed the time at and off work, being it scientific conferences or yearly excursions to the Kleinwalsertal, or only brief chats in the coffee kitchen. It was a pleasure to work with you. I would also like to thank the groups of Prof. Grevelding and Prof. Taubert for their valuable scientific input and for helping me out in the labs. This project would not have been possible without your efforts. I would like to thank my family for their unconditional love and support throughout my studies. I especially thank Johanna for everything she did to support me as well as my dogs Frido and Logan. Finally, I would like to thank everyone who contributed to the success of my doctoral thesis and, in favor of your valuable life time, did not mention individually. 108