Megasatellites are polymorphic tandem repetitive sequences with repeat-units longer than or equal to 1000 base pairs. The novel algorithm Megasatfinder predicts megasatellites in the human genome. A structured method of analysing the algorithm is developed and conducted. The analysis method consists of six test scenarios. Scripts are created, which execute the algorithm using various parameter settings. Three nucleotide sequences are applied; a real sequence extracted from the human genome and two random sequences, generated using different base probabilities. Usability and accuracy are investigated, providing the user with confidence in the algorithm and its output. The results indicate that Megasatfinder is an excellent tool for the detection of megasatellites and that the generated results are highly reliable. The results of the complete analysis suggest alterations in the default parameter settings, presented as user guidelines, and state that artificially generated sequences are not applicable as models for real DNA in computational simulations.

L'objectif de cette thèse est de concevoir une chimiothèque à partir de synthons C-glycosidiques. Les composés obtenus sont destinés à l'identification de modulateurs de l'activité des protéines pour comprendre leur rôle dans les mécanismes cellulaires. Pour réaliser ce travail, nous nous sommes basés sur le concept de la Synthèse Orientée vers la Diversité (SOD) pour générer une collection de molécules par des réactions à plusieurs composants. A partir de glycals peracétylés, nous avons préparés les C-glycosides de départ qui portent une fonction aldéhyde sur l'aglycone. Nous avons montré qu'il était possible de synthétiser ces produits en irradiant le milieu réactionnel par les micro-ondes. Puis ces synthons ont été traité par une amine et un alcyne ce qui a permis de générer une série de propargylamines. Parmi les propargylamines synthétiseés nous avons pu montrer que la L-proline avait une activité auto catalytique et qu'elles conduisaient a une addition stéréosélective de l'alcyne sur l'iminium intermédiaire. Nous avons alors utilisé cette propriété pour combiner la réaction à 3 composants avec la chimie click pour obtenir en quatre étapes une banque de 40 C-glycosides complexes hautement fonctionnalisées. Dans la deuxième partie de ce travail nous avons mis à profit la présence d'un alcool allylique sur le cycle pyranne pour mettre au point une nouvelle réaction tandem, la réaction A3M. Ce nouveau procédé permet d'obtenir en une seul étape des pyridino pyrannes par une réaction à trois composants. Après oxydation de l'alcool en cétone, le C-glycoside est irradie par les micro ondes en présence d'une amine primaire, et d'une alcyne (couplage A3). Cette première réaction est suivie d'une addition de Michael intramoléculaire sur la double liaison activée pour donner stéréo sélectivement le composé bicyclique

Sphingolipids are found in essentially all animals, plants and fungi, and some prokaryotic organisms and viruses. Sphingolipids function as structural components of membranes, lipoproteins, and as cell signaling modulators and mediators. To complicate matters further, sphingolipids often vary in type in different regions of tissues, and even in single cells, the subcellular localization of sphingolipids and their metabolic enzymes, transport proteins and targets may influence their functions. It is important to study sphingolipids spatial distribution within living organisms to understand how sphingolipids are involved in complex biochemical processes. As part of this thesis, procedures were optimized for the use of matrix assisted laser desorption/ionization (MALDI) tissue mass spectrometry (TIMS) to visualize the location of several types of lipids including sulfatides (ST), gangliosides and phosphoglycerolipids in brains from a mouse model for Tay-Sachs/Sandhoff disease. MALDI-TIMS was next applied to human ovarian carcinoma tissue to detect sulfatide location and established that ST are associated specifically with the regions of the ovarian tissue that bear the carcinoma. Electrospray ionization tandem mass spectrometry (ESI-MS-MS) was also used to confirm that ST and galactosylceramide (GalCer) are elevated in ovarian cancer. Gene expression data using tumor cells collected using laser capture microdissection revealed greater expression of mRNAs for GalCer synthase, GalCer sulfotransferase (Gal3ST1) and other enzymes of ST biosynthesis in epithelial ovarian carcinoma cells. This is a unique combination of two complementary, profiling technologies--mass spectrometry (metabolomic approach) with analysis of gene expression to study complex cancer pathology. The next study focused on the subcellular location of sphingolipids. In comparison with wild type Hek293 cells, a Hek293 cell line stably overexpressing serine palmitoyltransferase (SPT1/2 cells) was found to have elevated amounts of all subspecies of ceramide (Cer), but produces disproportionately higher amounts of C18-Cer and GalCer. Since Cer is known to inhibit protein ER/Golgi trafficking, these studies found that the higher production of Cer caused impairment of ER/Golgi trafficking of Ceramide synthase 1 (CerS1), thus increased C18-Cer. In addition, since GalCer is only synthesized in the lumen of the ER, this impairement of ER/Golgi trafficking also gave GalCer synthase access to its substrate and increased GalCer biosynthesis. These studies illustrate the complexity of sphingolipid biology and the usefulness of multiple tools to understand sphingolipid complex biological processes.

<p>Megasatellites are polymorphic tandem repetitive sequences with repeat-units longer than or equal to 1000 base pairs. The novel algorithm Megasatfinder predicts megasatellites in the human genome. A structured method of analysing the algorithm is developed and conducted. The analysis method consists of six test scenarios. Scripts are created, which execute the algorithm using various parameter settings. Three nucleotide sequences are applied; a real sequence extracted from the human genome and two random sequences, generated using different base probabilities. Usability and accuracy are investigated, providing the user with confidence in the algorithm and its output. The results indicate that Megasatfinder is an excellent tool for the detection of megasatellites and that the generated results are highly reliable. The results of the complete analysis suggest alterations in the default parameter settings, presented as user guidelines, and state that artificially generated sequences are not applicable as models for real DNA in computational simulations.</p>

Peptide a-amidation is known as a signature of bioactivity due to the fact that half of the bioactive peptides found in the nervous and endocrine systems are a-amidated and that most known a-amidated peptides are bioactive. a-Amidated peptides are produced by the oxidative cleavage of glycine-extended precursors. Peptidylglycine a-amidating monooxygenase (PAM) is the only known enzyme responsible for catalyzing this reaction and its sole physiological function is to convert glycine extended prohormones to their a-amidated forms. High levels of PAM are found in certain tissues with no corresponding level of amidated products suggesting the presence of undiscovered a-amidated peptide hormones. Liquid chromatography coupled tandem mass spectrometry (LC-MS/MS) has emerged as a powerful tool for peptide identification due to its advantages of speed, sensitivity and applicability to complex peptide mixtures. Normally, spectra are interpreted using database search engines. However, database searching is inefficient and ineffective for the identification of endogenous peptide with post-translational modifications (PTM) due to its low identification rate and high demand for computing power. There is a specific mass difference of 58.0055 units between an a-amidated peptide and its corresponding C-terminal glycine-extended precursor. The two peptides will have similar chromatographic retention time and MS/MS fragmentation patterns resulting from the identical amino acids sequences except for relatively the small differences at the C-termini. Based on this, a new LC-MS/MS based strategy for screening for a-amidated peptides was developed. This strategy depends on PAM inhibition and the mass accuracy of mass spectrometry (< 3 ppm). The coexistence of a-amidated peptides and their C-terminal glycine-extended precursors was insured by growing cells in the presence of a PAM inhibitor. After LC-MS/MS, masses and retention times of parent ions were extracted from raw data files and scanned by a script for peptide pairs with similar retention times and a mass difference around 58.0055. Resulting pairs were further validated by comparing their fragmentation patterns in MS/MS spectra. Only peptide pairs that met all three criteria were considered for further interpretation. This reduced the number of MS/MS spectra requiring interpretation by >99% and, thus, enable the manual inspection of MS/MS for the candidate peptide pairs. A total of 13 a-amidated peptides were successfully identified from cultured mouse pituitary AtT-20 cells using this method and a few of these newly identified a-amidated peptides exhibited bioactivity. The adaptability of this strategy to screening for other PTMs is also discussed. Peptidylglycine a-hydroxylating monooxygenase (PHM) is one of PAM domains which can be expressed separately. It is a copper dependent enzyme that catalyzes the first step of the two-step peptide amidation reaction. Removal of the copper ions results in the loss of enzyme catalytic activity. A PHM based a-amidated peptide enrichment method was developed. This method includes two steps. First, cells grown in culture were treated with a PAM inhibitor to effect the cellular accumulation of glycine-extended peptides. In the second step, copper-depleted PHM (apo-PHM) was used to selectively bind glycine-extended peptides present in the cell extract. All other unbound peptides were removed during wash runs. apo-PHM was then reinstated with copper to convert bound glycine-extended peptides to hydroxylated peptides and release them. Hydroxylated product can be converted to a-amidated peptide under basic conditions. Experiments carried out using model glycine extended peptides showed a 40 – 120-fold enrichment using HPLC-fluorometric assay or MALDI-TOF quantification. This method proved successful when working with complex samples like cell extracts. The relative intensity of a known a-amidated peptide mouse joining peptide (mJP) from an AtT-20 extract was dramatically increased after enrichment experiments.

Matrix metalloproteinases (MMPs) network with other biological molecules to maintain the extracellular matrix (ECM) in normal physiology and perform different roles. Understanding and assigning specific role to each of 24 members of these endoproteinases is impeded because of lack of specific and efficient detection methods in biological samples. Moreover, MMP-based anti-cancer drug development has also been challenged because, currently, there is no robust methodology to distinguish the inactive pro-enzymes, active enzymes or those complexed with endogenous inhibitors in biological specimens. The objective of this project is to develop a chemical proteomics strategy based on Matrix assisted laser desorption ionization tandem mass spectrometry (MALDI-MS/MS) to help identify and discriminate the various MMP forms. Firstly, a triazine dye-based ligand immobilized on chromatography beads was utilized to assess whether it binds to recombinant human MMPs (rhMMPs). The results highlighted that the ligand interacts with latent forms of MMPs in agreement with the literature. Secondly, the potential of the ligand was assessed using MALDI-MS/MS based methodology in in vitro cancer models. Cell line culture supernatants were used in amounts to emulate the availability of tumour biopsies in clinical settings. The MS/MS spectral peaks specific to MMPs (MMP-2 and MMP- 14), and two endogenous inhibitors TIMP-1 and TIMP-2 were found in affinity chromatography eluates of cell culture supernatants with higher Mascot scores for the latter. While western blot detected MMP-2 in cell extracts, MALDI-MS/MS did not detect MMPs because of amounts below the limit of detection (LOD) of the instrument. Although the ligand was found to be interacting with MMPs and detergent-free salt elution buffers improved MALDI analysis, recovery of MMPs from biological samples was sub-optimal. The dye ligand was observed to bind other enzymes and despite various strategies to reduce non-specific binding of proteins or enable selective elution did not improve MMP enrichment. Further work using methodology described in this study is required after scaling up the MMP amounts in biological specimen and to resolve the issue of non-specific binding of proteins to the ligand by understanding its structure.

Managed aquifer recharge (MAR) is a low-energy subsurface water treatment system with the potential of being an important component of sustainable water reuse schemes. Alongside common wastewater contaminants, MAR systems have been shown to attenuate a range of trace organic chemicals (TOrCs). Despite several factors being possibly important for TOrC attenuation, many have not been investigated in depth. This research effort investigated three factors affecting attenuation of the moderately degradable TOrCs: primary substrate, adaptation of the microbial community to presence of TOrCs, and groundwater temperature. The overall goal was to optimize TOrC attenuation using different MAR configurations considering how these factors affect TOrC attenuation. The primary substrate composition and concentration significantly impacted attenuation of the moderately degradable TOrCs. Lower primary substrate concentrations and more refractory carbon generally resulted in better TOrC transformation, a more diverse microbial community in the infiltration zone and more diverse capabilities for TOrC degradation. The enzyme group cytochrome P450 may be important for TOrC transformation since its genes were more abundant under carbon-starving primary substrate conditions. Adaptation of the microbial community by pre-exposure to TOrCs was not required in order to degrade them. However, adaptation to the primary substrate was necessary for TOrC biotransformation due to its effect on the microbial community. Attenuation of most TOrCs was unaffected by changes in temperature. Some moderately degradable TOrCs, however, were better attenuated at higher temperatures likely due to increased microbial activity. Others were better degraded at lower temperatures likely due to favorable sorption conditions. In the context of applying MAR systems to potential water reuse schemes within Saudi Arabia, a reconnaissance study of TOrC occurrence in treated wastewater effluents was undertaken. Most of the TOrCs targeted were detected at similar concentrations to US effluents at comparable plants. One of the plants studied, however, displayed a significantly different TOrC footprint from the other treatment plants due to the large number of international visitors in its sewershed. Findings from this occurrence study as well from other tasks provided inputs to a risk assessment framework to compare the effectiveness of MAR systems as part of a multiple-barrier water reuse scheme.

Low energy (tens of eV) polyatomic cations were used as probes for characterization of monolayers of spontaneously chemisorbed thiols on gold. Characteristics including chemical composition, surface order and orientation of the self-assembled monolayers (SAMs) can be derived by monitoring the products of projectile ion neutralization, surface-induced dissociation (SID), and ion-surface reactions.To study the influence of the terminal group chemical structures and orientations of the SAMs on ion-surface interactions, a series of semi-fluorinated alkane thiols with difluoromethylenes buried underneath hydrocarbon terminal groups were examined (CH3CF2CH2− and CH3CH2CF2−). Compared to terminally fluorinated SAMs, they showed more projectile ion neutralization and less internal to vibrational energy deposition into precursor ions. Projectile ion-hydrocarbon reactions decreased significantly when difluoromethylenes are one or two bonds away from the terminal group. Furthermore, ion-surface reaction results on surfaces with odd and even chain lengths suggested that they have similar terminal methyl orientations to their hydrocarbon counterparts.Mixed monolayers of CF3CF2(CH2)14SH (F-SAMs) and CH3(CH2)15SH (H-SAMs) with systematically changing electron transfer, energy deposition and ion-surface reaction were prepared using mixed thiols solution and micro-contact printing (μ-CP). The solution mixture system showed linear variations in electron transfer and energy deposition with different F-SAM surface concentrations, while non-linear changes occur for ion-surface reaction suggesting strong lateral interactions between the two components. These interactions are minimized in the μ-CP system containing domains of each thiol. Energy deposition on the patterned surfaces varies non-linearly with changing F-SAM concentration which differs from the homogenously mixed system.To explore SID with a 90 collision angle, eV SID of a series of protonated peptide ions were performed in an in-line sector Time-Of-Flight (TOF) mass spectrometer. The results were compared to keV collision-induced dissociation (CID) data collected with the same instrument. Fragmentation efficiency for SID was higher than CID for those peptides. In addition to the excellent control over laboratory collision energies with SID, different amount of energy deposition can be achieved when varying surface composition, e.g. using mixed F-SAM/H-SAM.Reactive ion scattering spectrometry (RISS) results provided more in-depth knowledge of low energy ion-surface interactions that will promote usage of RISS as a novel surface characterization technique.

Certains médicaments en se métabolisant dans le corps forment des métabolites dits « réactifs » (MR) qui peuvent entraîner une toxicité en se liant de manière covalente à des protéines pour former des adduits. Notre étude se concentre sur l'analyse d'adduits formés à partir d'une protéine sanguine : la sérum albumine humaine. En particulier, les adduits formés au niveau du site actif de l'albumine situé sur la cystéine en position 34. L'objectif principal de ce travail était de mettre en place une méthode d'analyse en chromatographie liquide haute performance (HPLC) couplée à la spectrométrie de masse en tandem (LC-MS/MS) permettant la détection et la caractérisation des adduits peptidiques issus de la sérum albumine modifiée. Le travail a consisté tout d'abord à valider par des réactions in vitro la formation d'adduit issus de médicaments connus pour former des MR avec un antioxydant naturel : le glutathion (GSH). En parallèle, en mimant le site actif de la sérum albumine humaine avec un peptide synthétique, le QQCPF, le profil de fragmentation en LC-MS/MS des adduits formés avec le QQCPF ont été comparé avec ceux obtenus par le GSH et ont permis de le valider comme agent de trappage. Pour de futures réactions in vivo, l'identification des adduits d'albumines en LC-MS implique une étape de digestion de l'albumine en peptides, le protocole de digestion de l'albumine en comparant plusieurs enzymes a été optimisé pour l'analyse en LC-MS des peptides digérés comportant la cystéine 34. Par ailleurs, une méthode de quantification de ce peptide modifié a été mise en place par l'introduction d'un étalon interne provenant d'une autre espèce d'albumine, l'albumine de rat. Le but ultime est que la sérum albumine soit utilisée comme biomarqueur sanguin de toxicité en utilisant cette méthodologie qui permettra avec un prélèvement sanguin d'identifier à partir de la sérum albumine humaine modifiée, la toxicité du médicament ingéré et ainsi améliorer la prévention de risque du médicament. ______________________________________________________________________________ MOTS-CLÉS DE L’AUTEUR : Sérum albumine, métabolite réactif, adduits, liaison covalente, spectrométrie de masse, fragmentation, séquençage, chromatographie liquide.

In quantitative proteomics, many of the LC-MS based approaches employ stable isotopic labelling to provide relative quantitation of the proteome in different cell states. In a typical approach, peptides are first detected and identified by tandem MS scans prior to quantifying proteins. This provides the researcher with a large amount of data that are not useful for quantitation. It is desirable to improve the throughput of current approaches to make proteomics a more routine experiment with an enhanced capacity to detect differentially expressed proteins. This thesis reports the developments towards this goal, including an assessment of the viability of stable dimethyl labelling for comparative proteomic measurements and the evaluation of a dynamic algorithm called Parallel Isotopic Tag Screening (PITS) for the detection of isotopically labelled peptides for quantitative proteomics without the use of tandem MS scans.