none

Aim: To assess SELDI-TOF MS technology as a tool for biomarker discovery in the stool and serum of colorectal cancer patients. Materials and Methods: 1.Initially a technique of analysis was developed and optimised using tumour samples and matched normal mucosa, obtained from the Tayside Tissue Bank. These samples were then analysed using SELDI on a PBS II Protein Chip Mass Spectrometer to identify abundant proteins. 2. A technique of stool preparation and subsequent SELDI analysis was developed and then optimised (CM10 chip at pH4) to allow comparison of faecal samples from cancer and controls. Faecal samples were then collected from cancer patients and controls and analysed. In addition, FOB testing was carried out on all stool samples from cancer and controls and subgroup analysis of spectras was performed controlling for FOB status. 3. A test set of cancer and normal serum samples was used to optimise the method of analysis using 4 different chip surfaces at differing pH. Serum samples were collected from cancer patients and normal controls and were analysed on the H50 chip. Serum was then depleted of major proteins in an attempt to improve the detection of peaks. The mass spectra from each sample type were compared to identify any common protein peaks. Results: 1. Tumour analysis methods were optimised using an initial 4 samples of tumour and normal mucosa. Analysis of 8 further paired samples showed protein peaks at 2826, 3374, 3444, 3489 and 10854 Da which were abundant in tumour and reduced in the normal mucosa. 2. In serum analysis the initial experiment of 10 cancer versus 10 normal revealed 4 peaks on the H50 chip (3479, 3364, 3434, 3700 Da) that had significantly higher mass to charge ratios in cancer. The experiment was repeated on the H50 chip using 92 cancers and 92 controls and 5 different peaks were identified (7901, 8124, 8566, 8799 and 17 409Da) as being significant but these had higher mass to charge ratios in the controls. After depletion of the serum samples of albumin, transferrin, haptoglobin, anti-trypsin, IgG and IgA SELDI-TOF analysis showed a greatly reduced profile that yielded no meaningful spectra. 3. Stool analysis revealed 5 protein peaks (4633, 16511, 33423, 37087 and 47026 Da) in colorectal cancer patients, which were absent in stools from controls with a sensitivity of 83% when using all 5 peaks. Degradation of the spectra was observed after prolonged storage of stool samples. Conclusions: A method of stool analysis has been developed that yielded valid peaks differentiating between cancer and normal, which warrant further research through protein identification. Serum analysis was not reproducible across experiments and depletion of major proteins failed to reveal the sub-proteome raising doubts about whether discovery-based serum proteomics can accurately detect cancer. SELDI-TOF was not able to demonstrate that any of the peaks present in the tumour analysis were present in the stool or the serum samples.

Cystic fibrosis (CF) lung disease is characterised by early airways infection and inflammation, chronic suppuration, frequent infective exacerbations and an increased influx of acute, and chronic inflammatory cells. The inflammatory process involves activation of many cell types including neutrophils, macrophages and epithelial cells, and leads ultimately to the development of progressive respiratory failure and death. Accurate assessment of the inflammatory process is a crucial part of disease monitoring and should allow appropriate evaluation of therapeutic interventions so as to maximize control of the respiratory sequelae of the disorder. Lung function markers such as FEV1 are insensitive and indirect. Direct but invasive methods such as fibreoptic bronchoscopy and biopsy are limited in application, repeatability and safety. Non-invasive methods of assessment are, therefore, attractive. Exhaled Breath Gases, Exhaled Breath Condensate and Induced Sputum provide potential for such measures. These techniques are safe, simple, repeatable and could assess all airways and can be used in children as young as 6 years. We hypothesised that biomarkers of inflammation in Cystic Fibrosis Lung Disease are measurable in samples collected noninvasively, and can be developed into clinically useful assays. These assays would have the ability to reflect the level of inflammation in the CF lungs as well as holding the potential to act as surrogate markers of CFTR function. Methods Non-invasive markers of inflammation in Cystic Fibrosis lung disease Methods. Exhaled breath gases, exhaled breath condensate, bronchoalveolar lavage fluid and induced sputum were investigated using a number of analysis techniques to identify the markers which best discriminated CF from non CF subjects. Analysis techniques used were electrochemical cells, chemiluminescene, ELISA, EIA, ion selective probes and mass spectrometry. Results Markers found to discriminate CF from non CF subjects were EBC pH and ammonium, and 38 proteomic markers were found in induced sputum. 21 proteomic markers were found in bronchoalveolar lavage fluid. One biomarker has been identified with confidence, Calgranulin A. Discussion A large component of the work of this thesis was focussed on exhaled breath condensate. Two markers, pH and Ammonium were different between the CF and control groups. The measurement of EBC pH and ammonium as markers of inflammation should be used in future gene therapy trials as they are cheap, quick and simple to perform Using clean techniques free from contamination, no proteins are repeatedly detectable in EBC using highly sensitive SELDI techniques. This technique reflects the highest sensitivity of any available proteomics instrument and therefore until new technologies become available, it would be incorrect to assay any proteins in EBC. The induced sputum proteomics study identified 38 independent markers of CF lung inflammation Therefore, sampling by collection of induced sputum should be used in gene therapy trials. The endpoints should be assessed by a combination of SELDI as an endpoint and by ELISA where this is available. The marker Calgranulin is likely to report on neutrophil recruitment to the lung. It is anticipated that this will be a sensitive marker of inflammation in the lung and it also has the potential to report on successful of gene transfer as it is raised in heterozygote carriers as well as homozygotes with CF. Therefore, the non-invasive technique induced sputum coupled to proteomic analysis would have the ability to reflect the level of inflammation in CF subjects and may also report on CFTR function.

Les algorithmes génétiques sont des méthodes d'optimisation destinées à des problèmes complexes. Ils peuvent jouer un rôle intéressant dans le cadre de la protéomique. Cette discipline est assez récente, elle étudie le patrimoine en protéines des individus. Elle produit des données de grande dimension. <br />La première partie aborde l'histoire, le fonctionnement des algorithmes génétiques et certains résultats théoriques. La partie suivante détaille la mise au point d'un tel algorithme pour la sélection de biomarqueurs en spectrométrie de masse et l'alignement de gels d'électrophorèse 2D. Cette partie met en évidence la difficulté de construction du critère à optimiser. La dernière partie aborde des résultats théoriques. La convergence des algorithmes génétiques avec élitisme est démontrée dans le cas non homogène et de mutations dirigées. Nous avons ensuite construit un critère de convergence alliant fondements théoriques et applicabilité, basé sur les occurrences de la solution localement optimale. Enfin, l'efficacité de l'introduction d'événements catastrophiques dans la résolution pratique de certains problèmes de convergence est montrée.

Le remodelage ventriculaire gauche (VG) en post-infarctus du myocarde (IDM) est associé à une augmentation du risque d’insuffisance cardiaque et de décès, mais il demeure difficile à prédire en pratique clinique.L’objectif principal de ma thèse était la recherche de biomarqueurs circulants du remodelage VG par l’approche protéine candidate et par protéomique différentielle dans la population REVE-2.Par l’approche protéine candidate, nous avons confirmé que le peptide natriurétique detype B (BNP) était un puissant facteur prédictif du remodelage VG en post-IDM. La métalloprotéase matricielle-8 (MMP-8), la MMP-9, l’hepatocyte growth factor (HGF), la Créactive protéine (CRP), la troponine I ont également fait la preuve de leur association.Par l’approche protéomique différentielle, en électrophorèse 2D différentielle en fluorescence (2D-DIGE), la clusterine a été identifiée comme biomarqueur potentiel,positivement associée au remodelage VG, nécessitant toutefois des travaux de confirmation.Par SELDI TOF MS, nous avons sélectionné 26 pics définis par leur rapport m/z, commebiomarqueurs potentiels du remodelage VG, dont 12 ont pu être identifiés et devrontdésormais être validés : le pic de m/z 2777 a été identifié comme le peptide N-terminal issu del’albumine après clivage par la pepsine. Les autres pics correspondraient à des fragments protéolytiques de protéines que sont le fibrinogène, le complément C3, C4 et C1q.La découverte de nouveaux biomarqueurs du remodelage VG devrait permettre d’améliorer la stratification du risque en post-IDM afin d’identifier les patients devant bénéficier d’un suivi plus rapproché et peut-être d’une prise en charge thérapeutique plus agressive / Left ventricular (LV) remodelling after myocardial infarction (MI) indicates a high risk of heart failure and death but remains difficult to predict in clinical practice. Biomarkers may help to refine risk stratification. The main purpose was to find circulating biomarkers of LV remodelling after MI, using two strategies : candidate protein approach and differential proteomic approach, working on a population with a clearly defined phenotype, the REVE-2 study, a prospective multicenter study including 246 patients with a first anterior Q-wave MI. Blood samples were obtained at hospital discharge, at 1 month, 3 months and 1 year. An echocardiography was performed at the same time except for the 1st month to assess LVR.By candidate protein approach, we confirmed that B-type natriuretic peptide (BNP) was a powerful predictor of LV remodelling after MI. Additional biomarkers, such as matrix metalloproteinase-8 (MMP-8), MMP-9, hepatocyte growth factor (HGF), C-reactive protein (CRP) and cardiac troponin I were found to be associated with LV remodelling, highlighting several pathways implicated in pathophysiology of LV remodelling. We have also shown that biomarkers in association (BNP and cardiac troponin I, BNP and MMP-8, BNP and MMP-9) could improve risk stratification in post-MI by selecting groups of patients at higher risk.As the ideal biomarker was still not identified, we applied a differential proteomic approach, with no a priori hypothesis, in order to characterize proteomic signature of LV remodelling. The use of a protein enrichment kit, consisting of a library of combinatorial hexapeptide ligands, compressed the protein concentration range of plasma and serum, through the simultaneous onestep dilution of high-abundance and concentration of lowabundance proteins. Protein enrichment kit prior to two-dimensional (2D) electrophoresis or SELDI TOF MS (surface-enhanced laser desorption–ionization time of flight) analysis enabled the detection of proteins that were not detected in native blood sample and the accessibility to proteolytic fragments obtained from major proteins. Clusterin (apolipoprotein J) was identified as a potential biomarker of LV remodelling by 2D-DIfferential Gel Electrophoresis (2D-DIGE). Clusterin was quantified by Western blot and ELISA and was found to be positively associated with LV remodelling. However, this association was not found with all LV remodelling parameters nor at each time during the year following MI, requiring further analysis. Differential proteomic approach by SELDI TOF MS selected 26 m/z peaks, as potential biomarkers of LV remodelling. Of them, 12 were identified by mass spectrometry. The 2777 m/z peak was identified directly from the ProteinChip array as being the N-terminal peptide (24–48 aa) generated from albumin by pepsin cleavage. Other peaks were identified after purification using chromatographic columns or liquid-phase isoelectric focusing : most of them were found to be proteolytic fragments of proteins like fibrinogen, C3, C4 and C1q complement. Identifications have now to be validated with specific techniques, usually by immmunoprecipitation and Western blot analysis.Finding new biomarkers of LV remodelling could help refine risk stratification and identify patients in whom more aggressive therapy and/or more frequent follow-up could be needed.

No description available.

<p>Type 2 diabetes mellitus (T2DM) is a heterogeneous disease affecting millions of people worldwide. Both genetic and environmental factors contribute to the pathogenesis. The disease is characterized by alterations in many genes and their products. Historically, genomic alterations have mainly been studied at the transcriptional level in diabetes research. However, transcriptional changes do not always lead to altered translation, which makes it important to measure changes at the protein level. Proteomic techniques offer the possibility of measuring multiple protein alterations simultaneously.</p><p>In this thesis, the proteomic technique surface enhanced laser desorption ionization time-of-flight mass spectrometry (SELDI-TOF MS) has been applied and evaluated in the context of T2DM research. Protocols for pancreatic islet and serum/plasma protein profiling and identification have been developed. In addition, the technique was used to analyze the influence of genetic background versus diabetic environment by determining serum protein profiles of individuals with normal glucose tolerance (NGT) and T2DM with or without family history of diabetes. In total thirteen serum proteins displayed different levels in serum from persons with NGT versus patients with T2DM. Among these proteins, apolipoprotein CIII, albumin and one yet unidentified protein could be classified as being changed because of different genetic backgrounds. On the other hand, ten proteins for instance transthyretin, differed as a result of the diabetic environment.</p><p>When plasma protein patterns of NGT and T2DM individuals characterized by differences in early insulin responses (EIR) were compared, nine proteins were found to be varying between the two groups. Of these proteins five were identified, namely two forms of transthyretin, hemoglobin α-chain, hemoglobin β-chain and apolipoprotein H. However no individual protein alone could explain the differences in EIR. In conclusion, SELDI-TOF MS has been successfully used in the context of T2DM research to identify proteins associated with family history of diabetes and β-bell function. </p>

Background and aims : Crohns disease (CD) and ulcerative colitis (UC) known as inflammatory bowel diseases (IBD) are chronic immuno-inflammatory pathologies of the gastrointestinal tract. These diseases are multifactorial, polygenic and of unknown etiology. Clinical presentation is non specific and diagnosis is based on clinical, endoscopic, radiological and histological criteria. Novel biomarkers are needed to improve early diagnosis and classification of these pathologies as well as to monitor or predict the effects of therapy. Methods We performed a study with 120 serum samples collected from patients classified in 4 groups: 30 CD, 30 UC, 30 inflammatory controls (IC) and 30 healthy controls (HC), according to accredited criteria. We compared protein sera profiles obtained with a Surface Enhanced Laser Desorption Ionization-Time of Flight-Mass Spectrometer (SELDI-TOF-MS). Data analysis with an original multivariate statistical method based on multiple decision trees algorithms allowed us to select new potential biomarkers. Eight of them were purified and identified by mass spectrometry and antibody based methods. Moreover, a similar analysis was applied on sera taken from patients before and after treatment with Infliximab. We obtained multivariates models based on biomarkers able to predict the response and monitor changes in protein profiles before and after therapy. One biomarker was identified by an antibody based method and is able to predict clinical response to anti-TNF therapy. Results : Multivariate analysis generated models that could classify samples with good sensitivity and specificity (minimum 80%) discriminating IBD versus HC and IC or CD versus UC. The discrimination of these multivariate models and those of some identified combined biomarkers were compared to ASCA and ANCA tests and showed better or equivalent power of discrimination. Eight biomarkers were purified and identified: Platelet aggregation Factor 4 (PF4), the Myeloid Related Protein 8 (MRP8), the Fibrinopeptide A (FIBA), the Haptoglobin α2 subunit (Hpα2). They were detected in sera by classical methods, when available. Unique decision tree was built with these biomarkers and correlations with characteristics of patients and between biomarkers were calculated. Finally, we also adressed with a similar strategy predictor of response to Infliximab therapy. Conclusions : SELDI-TOF-MS technology combined with the use of the multiple decision trees method, as robust statistical tool, led to the selection of protein biomarker patterns and of specific biomarkers which could be helpful for diagnosis of inflammatory bowel diseases and prediction of therapy effects as well as understanding of these diseases pathophysiology. Buts : La maladie de Crohn (CD) et la rectocolite ulcéro-hémorrhagique (UC), désignées conjointement sous le nom « Inflammatory Bowel Disease » (IBD), sont deux maladies inflammatoires chroniques de lintestin. Ces pathologies nont pas une étiologie formellement identifiée et semblent être multifactorielles et polygéniques. Leurs présentations cliniques sont aspécifiques et leur diagnostic est actuellement basé sur les données cliniques, endoscopiques, radiologiques et histologiques. Ainsi, de nouveaux marqueurs spécifiques et sensibles de ces pathologies sont nécessaires afin daméliorer le diagnostic précoce et la classification de ces pathologies, ainsi que pour suivre, voire prédire la réponse aux traitements. Méthodes : Une étude protéomique a été réalisée sur 120 échantillons sériques provenant de patients classés en 4 groupes (30 CD, 30 UC, 30 IC et 30 HC), suivant les critères accrédités. Nous avons comparé les profils protéiques obtenus par « Surface Enhanced Laser Desorption Ionization-Time of Flight-Mass Spectrometer » (SELDI-TOF-MS). Lanalyse des données avec une méthode statistique multivariée originale, basée sur la construction darbres de décisions multiples, nous a permis de sélectionner certains biomarqueurs potentiels. Huit dentre eux ont été identifiés par spectrométrie de masse et des techniques de reconnaissance spécifique par anticorps. De plus, une analyse similaire a été effectuée sur des échantillons provenant de patients avant et après traitement par Infliximab. Nous avons obtenu à partir des profils protéiques, des modèles de classification statistiques multivariés et finalement certains biomarqueurs potentiels de prédiction de la réponse au traitement. De plus, un biomarqueur a été identifié par reconnaissance via un anticorps spécifique. Résultats : Lanalyse multivariée a généré des modèles de classification présentant de bonnes sensibilités et spécificités (minimum 80%) pour la discrimination des patients IBD versus contrôles ou encore, CD versus UC. Les résultats ont été comparés avec ceux des tests ASCA et ANCA et montrent une efficacité équivalente ou supérieure. Huit biomarqueurs ont été purifiés et identifiés : le facteur dagrégation plaquettaire 4 (PF4), la protéine MRP8 ou «Myeloid Related Protein 8», le fibrinopeptide A (FIBA), la sous-unité α2 de lhaptoglobine (Hpα2). Ces protéines et peptides ont été détectés, lorsque possible, dans le sérum par les méthodes classiques. Un arbre de décision multiple a été construit sur base de ces biomarqueurs. Les corrélations entre ceux-ci et avec les caractéristiques des patients ont été également évaluées. Conclusions : La technique SELDI-TOF-MS combinée à lutilisation des arbres de décisions multiples en temps que méthode danalyse statistique robuste, a mené à lobtention de signatures protéiques caractéristiques et à la sélection de biomarqueurs spécifiques qui peuvent se révéler utiles au diagnostic des pathologies IBD, au suivi, comme à la prédiction des effets des thérapies et à la compréhension de la physiopathologie de ces maladies

Type 2 diabetes mellitus (T2DM) is a heterogeneous disease affecting millions of people worldwide. Both genetic and environmental factors contribute to the pathogenesis. The disease is characterized by alterations in many genes and their products. Historically, genomic alterations have mainly been studied at the transcriptional level in diabetes research. However, transcriptional changes do not always lead to altered translation, which makes it important to measure changes at the protein level. Proteomic techniques offer the possibility of measuring multiple protein alterations simultaneously. In this thesis, the proteomic technique surface enhanced laser desorption ionization time-of-flight mass spectrometry (SELDI-TOF MS) has been applied and evaluated in the context of T2DM research. Protocols for pancreatic islet and serum/plasma protein profiling and identification have been developed. In addition, the technique was used to analyze the influence of genetic background versus diabetic environment by determining serum protein profiles of individuals with normal glucose tolerance (NGT) and T2DM with or without family history of diabetes. In total thirteen serum proteins displayed different levels in serum from persons with NGT versus patients with T2DM. Among these proteins, apolipoprotein CIII, albumin and one yet unidentified protein could be classified as being changed because of different genetic backgrounds. On the other hand, ten proteins for instance transthyretin, differed as a result of the diabetic environment. When plasma protein patterns of NGT and T2DM individuals characterized by differences in early insulin responses (EIR) were compared, nine proteins were found to be varying between the two groups. Of these proteins five were identified, namely two forms of transthyretin, hemoglobin α-chain, hemoglobin β-chain and apolipoprotein H. However no individual protein alone could explain the differences in EIR. In conclusion, SELDI-TOF MS has been successfully used in the context of T2DM research to identify proteins associated with family history of diabetes and β-bell function.

Surface-enhanced laser desorption/ionization time-of-flight mass spectrometry (SELDI)is one of several proteomics technologies that can be used in biomarker discovery studies. Such studies often have the goal of finding protein markers that predict early onset of cancers such as cervical cancer. The reproducibility of SELDI has been shown to be an issue in the literature. There are numerous sources of error in a SELDI experiment starting with sample collection from patients to the signal processing steps used to estimate the protein mass and abundance values present in a sample. This dissertation is concerned with all aspects of signal processing related to SELDI's use in biomarker discovery projects. In chapter 2, we perform a comprehensive study of the most popular preprocessing algorithms available. Next, in chapter 3, we study the basic statistics of SELDI data acquisition. From here, we propose a quadratic variance measurement model for buffer+matrix only spectra. This model leads us to develop a modified Antoniadis-Sapatinas wavelet denoising algorithm that demonstrates superior performance when compared to MassSpecWavelet, one of the leading techniques for preprocessing SELDI data. In chapter 4, we show that the quadratic variance model 1) extends to real pooled cervical mucus QC data from a clinical study, 2) predicts behavior and reproducibility of peak heights, and 3) finds four times as many reproducible peaks as the vendor-supplied preprocessing programs. The quadratic variance measurement model for SELDI data is fundamental and promises to lead to improved techniques for analyzing the data from clinical studies using this instrument.