Pregnancy-associated changes in host permissiveness to metastasis

THESIS SUMMARY : Metastasis is a multistep process involving tumour cell-autonomous features, the host tissue stroma of the primary tumour, the blood or lymphatic system as well as a receptive target organ. Most studies on factors influencing metastasis have concentrated on the characteristics of the disseminating tumour cell and on early steps of metastasis including invasion and angiogenesis. Although these steps are necessary for tumour cells to disseminate, it is the challenges encountered in the later steps of metastasis -survival while in the circulation and engraftment and outgrowth in the target organ -that account for the inefficiency of circulating tumour cells in establishing secondary lesions. Full understanding of the metastatic process therefore requires elucidation of the mechanisms that regulate these late steps, and in particular that determine what makes any given tissue permissive for metastatic tumour growth. To address this issue, we assessed the mechanisms whereby a physiological situation -pregnancy -can alter host permissiveness toward metastasis. We show that pregnant NOD/SCID mice -injected intravenously with tumour cells -develop more metastases than their non-pregnant counterparts irrespective of the tumour cell type. There was no direct effect of pregnancy-related circulating factors on tumour cell proliferation, and subcutaneous tumour growth does not vary between pregnant and nonpregnant animals. However, decreased elimination of tumour cells from the lung microvasculature was observed in pregnant mice, prompting us to assess whether pregnancy-related adaptations in innate immunity could account for this differential clearing. We found that natural killer (NK) cell fractions are decreased in blood and spleen of pregnant mice and that NK cell cytotoxicity is impaired, as reported previously. The use of NK-deficient mice or tumour cell lines resistant to NK killing abrogates the difference in metastasis load between pregnant and virgin mice. CD11 b+ Gr-1+ myeloid-derived suppressor cells (MDSC) have previously been shown to accumulate in tumour-bearing mice and to down-modulate NK activity. Accordingly, we show an increase in MDSC in pregnant mouse blood, spleen, lungs and liver. Depletion of MDSC prior to tumour cell injection decreased metastasis load in pregnant NOD/SCID mice but had no effect on virgin mice. Similarly, adoptive transfer of MDSC extracted from pregnant mice into virgin mice lead to increased metastasis take. In parallel, we investigated whether the lung and liver microenvironments are modified during pregnancy thereby providing a more "permissive soil" for the establishment of metastases. A comparative analysis of microarray data of pregnant mouse lungs and liver with "premetastatic niche" gene expression profiles of these organs shows that similar mechanisms could mediate an increase in lung and liver metastasis in pregnant mice and in mice harbouring an aggressive primary tumour. Several commonly up-regulated genes point towards the recruitment of myeloid cells, consistent with the accumulation of MDSC observed in pregnant mice. MDSC have never been evoked in the context of pregnancy before. Although the role of MDSC in pregnancy requires further investigation we suggest that MDSC accumulation constitutes an important and hitherto unrecognised common denominator of maternal immune tolerance and cancer immune escape. RESUME DE THESE : La métastatisation est un processus en plusieurs étapes qui implique des compétences particulières chez les cellules tumorales, le stroma de la tumeur primaire, les vaisseaux sanguins ou lymphatiques ainsi qu'un organe cible' réceptif. Jusqu'alors, la recherche s'est principalement intéressée aux facteurs qui influencent les étapes précoces de la métastatisation donc aux caractéristiques de la cellule métastatique, et aux processus tels que l'invasion et l'angiogenèse, tandis que peu d'études traitent des étapes tardives tel que la survie dans la circulation sanguine et l'établissement d'une lésion dans l'organe cible. En particulier, l'élucidation des facteurs qui déterminent la permissivité d'un tissu à la greffe de cellules disséminantes est indispensable à la compréhension de ce processus complexe qu'est la métastatisation. Nous proposons ici un modèle de souris récapitulant les étapes tardives de la métastatisation dans un contexte d'une permissivité accrue aux métastases chez la souris gravide, et nous évaluons les mécanismes impliqués. Les souris gestantes développent plus de métastases après l'injection intraveineuse de cellules tumorales, indépendamment du type de tumeur d'origine. Les taux élevés d'hormones et de facteurs de croissance chez la souris gravide n'inflúencent pas la prolifération des cellules tumorales et fa croissance de tumeurs sous-cutanées n'est pas non plus accélérée par la gestation. En revanche, une fois injectées, les cellules tumorales sont éliminées ` moins rapidement des vaisseaux pulmonaires chez la souris gravide que chez les contrôles. Cette observation est compatible avec un effet de la gestation sur l'immunité innée et nous avons mis en évidence une diminution des proportions de cellules NK (natural killer) dans le sang et la rate en particulier, ainsi qu'une cytotoxicité moindre envers des cellules tumorales. En utilisant des souris déficientes en cellules NK ou en injectant des cellules résistantes à l'attaqué par des cellules NK, la différence entre souris gestantes et non-gestantes disparaît. Il a été démontré chez des souris porteuses de tumeurs, que l'accumulation de cellules immunosuppressives de la lignée myélo-monocytaire (ou MDSC pour myeloid-derived suppressor tells) pouvait être responsable d'une inhibition de l'activité de cellules NK. Des nombres augmentés de ces cellules, caractérisées par les marqueurs de surface CD11b et Gr-1, ont été trouvés dans le sang, la rate, les poumons et le foie de souris gravides. Leur rôle dans la métastatisation est démontré par le fait que leur dépletion diminue le nombre de lésions secondaires chez la souris gestante, tandis que leur transfert dans des souris non-gestantes augmente le taux de métastases. L'utilisation de puces à ADN sur les foies et poumons de souris gravides a permis de mettre en évidence des différences d'expression génique proches de celles observées dans l'établissement de niches pré-métastatiques. Ceci suggère que des mécanismes similaires pourraient être responsables d'une permissivité accrue aux métastases chez la souris gravide et chez la souris porteuse d'une tumeur primaire agressive, telle que, en particulier, l'accumulation de cellules immunosuppressives dans les organes cibles. C'est la première fois que l'accumulation de MDSC est évoquée chez la souris gravide et nous proposons ici que celles-ci jouent un rôle dans la tolérance immunitaire envers le foetus et sont responsables de l'échappement de cellules tumorales injectées à la surveillance immunitaire par des cellules NK.

Kernel-based Mapping of Meteorological Fields in Complex Orography

Due to the advances in sensor networks and remote sensing technologies, the acquisition and storage rates of meteorological and climatological data increases every day and ask for novel and efficient processing algorithms. A fundamental problem of data analysis and modeling is the spatial prediction of meteorological variables in complex orography, which serves among others to extended climatological analyses, for the assimilation of data into numerical weather prediction models, for preparing inputs to hydrological models and for real time monitoring and short-term forecasting of weather.In this thesis, a new framework for spatial estimation is proposed by taking advantage of a class of algorithms emerging from the statistical learning theory. Nonparametric kernel-based methods for nonlinear data classification, regression and target detection, known as support vector machines (SVM), are adapted for mapping of meteorological variables in complex orography.With the advent of high resolution digital elevation models, the field of spatial prediction met new horizons. In fact, by exploiting image processing tools along with physical heuristics, an incredible number of terrain features which account for the topographic conditions at multiple spatial scales can be extracted. Such features are highly relevant for the mapping of meteorological variables because they control a considerable part of the spatial variability of meteorological fields in the complex Alpine orography. For instance, patterns of orographic rainfall, wind speed and cold air pools are known to be correlated with particular terrain forms, e.g. convex/concave surfaces and upwind sides of mountain slopes.Kernel-based methods are employed to learn the nonlinear statistical dependence which links the multidimensional space of geographical and topographic explanatory variables to the variable of interest, that is the wind speed as measured at the weather stations or the occurrence of orographic rainfall patterns as extracted from sequences of radar images. Compared to low dimensional models integrating only the geographical coordinates, the proposed framework opens a way to regionalize meteorological variables which are multidimensional in nature and rarely show spatial auto-correlation in the original space making the use of classical geostatistics tangled.The challenges which are explored during the thesis are manifolds. First, the complexity of models is optimized to impose appropriate smoothness properties and reduce the impact of noisy measurements. Secondly, a multiple kernel extension of SVM is considered to select the multiscale features which explain most of the spatial variability of wind speed. Then, SVM target detection methods are implemented to describe the orographic conditions which cause persistent and stationary rainfall patterns. Finally, the optimal splitting of the data is studied to estimate realistic performances and confidence intervals characterizing the uncertainty of predictions.The resulting maps of average wind speeds find applications within renewable resources assessment and opens a route to decrease the temporal scale of analysis to meet hydrological requirements. Furthermore, the maps depicting the susceptibility to orographic rainfall enhancement can be used to improve current radar-based quantitative precipitation estimation and forecasting systems and to generate stochastic ensembles of precipitation fields conditioned upon the orography.

A Fast Gradient Approximation for Nonlinear Blind Signal Processing

When dealing with nonlinear blind processing algorithms (deconvolution or post-nonlinear source separation), complex mathematical estimations must be done giving as a result very slow algorithms. This is the case, for example, in speech processing, spike signals deconvolution or microarray data analysis. In this paper, we propose a simple method to reduce computational time for the inversion of Wiener systems or the separation of post-nonlinear mixtures, by using a linear approximation in a minimum mutual information algorithm. Simulation results demonstrate that linear spline interpolation is fast and accurate, obtaining very good results (similar to those obtained without approximation) while computational time is dramatically decreased. On the other hand, cubic spline interpolation also obtains similar good results, but due to its intrinsic complexity, the global algorithm is much more slow and hence not useful for our purpose.

Gene expression patterns unveil a new level of molecular heterogeneity in colorectal cancer.

The recognition that colorectal cancer (CRC) is a heterogeneous disease in terms of clinical behaviour and response to therapy translates into an urgent need for robust molecular disease subclassifiers that can explain this heterogeneity beyond current parameters (MSI, KRAS, BRAF). Attempts to fill this gap are emerging. The Cancer Genome Atlas (TGCA) reported two main CRC groups, based on the incidence and spectrum of mutated genes, and another paper reported an EMT expression signature defined subgroup. We performed a prior free analysis of CRC heterogeneity on 1113 CRC gene expression profiles and confronted our findings to established molecular determinants and clinical, histopathological and survival data. Unsupervised clustering based on gene modules allowed us to distinguish at least five different gene expression CRC subtypes, which we call surface crypt-like, lower crypt-like, CIMP-H-like, mesenchymal and mixed. A gene set enrichment analysis combined with literature search of gene module members identified distinct biological motifs in different subtypes. The subtypes, which were not derived based on outcome, nonetheless showed differences in prognosis. Known gene copy number variations and mutations in key cancer-associated genes differed between subtypes, but the subtypes provided molecular information beyond that contained in these variables. Morphological features significantly differed between subtypes. The objective existence of the subtypes and their clinical and molecular characteristics were validated in an independent set of 720 CRC expression profiles. Our subtypes provide a novel perspective on the heterogeneity of CRC. The proposed subtypes should be further explored retrospectively on existing clinical trial datasets and, when sufficiently robust, be prospectively assessed for clinical relevance in terms of prognosis and treatment response predictive capacity. Original microarray data were uploaded to the ArrayExpress database (http://www.ebi.ac.uk/arrayexpress/) under Accession Nos E-MTAB-990 and E-MTAB-1026. © 2013 Swiss Institute of Bioinformatics. Journal of Pathology published by John Wiley & Sons Ltd on behalf of Pathological Society of Great Britain and Ireland.

Living with an imperfect cell wall: compensation of femAB inactivation in Staphylococcus aureus.

BACKGROUND: Synthesis of the Staphylococcus aureus peptidoglycan pentaglycine interpeptide bridge is catalyzed by the nonribosomal peptidyl transferases FemX, FemA and FemB. Inactivation of the femAB operon reduces the interpeptide to a monoglycine, leading to a poorly crosslinked peptidoglycan. femAB mutants show a reduced growth rate and are hypersusceptible to virtually all antibiotics, including methicillin, making FemAB a potential target to restore beta-lactam susceptibility in methicillin-resistant S. aureus (MRSA). Cis-complementation with wild type femAB only restores synthesis of the pentaglycine interpeptide and methicillin resistance, but the growth rate remains low. This study characterizes the adaptations that ensured survival of the cells after femAB inactivation. RESULTS: In addition to slow growth, the cis-complemented femAB mutant showed temperature sensitivity and a higher methicillin resistance than the wild type. Transcriptional profiling paired with reporter metabolite analysis revealed multiple changes in the global transcriptome. A number of transporters for sugars, glycerol, and glycine betaine, some of which could serve as osmoprotectants, were upregulated. Striking differences were found in the transcription of several genes involved in nitrogen metabolism and the arginine-deiminase pathway, an alternative for ATP production. In addition, microarray data indicated enhanced expression of virulence factors that correlated with premature expression of the global regulators sae, sarA, and agr. CONCLUSION: Survival under conditions preventing normal cell wall formation triggered complex adaptations that incurred a fitness cost, showing the remarkable flexibility of S. aureus to circumvent cell wall damage. Potential FemAB inhibitors would have to be used in combination with other antibiotics to prevent selection of resistant survivors.

Mechanistic insights into cytosolic molecular chaperones in protein unfolding and disaggregation

Under optimal non-physiological conditions of low concentrations and low temperatures, proteins may spontaneously fold to the native state, as all the information for folding lies in the amino acid sequence of the polypeptide. However, under conditions of stress or high protein crowding as inside cells, a polypeptide may misfold and enter an aggregation pathway resulting in the formation of misfolded conformers and fibrils, which can be toxic and lead to neurodegenerative illnesses, such as Alzheimer's, Parkinson's or Huntington's diseases and aging in general. To avert and revert protein misfolding and aggregation, cells have evolved a set of proteins called molecular chaperones. Here, I focussed on the human cytosolic chaperones Hsp70 (DnaK) and HspllO, and co-chaperone Hsp40 (DnaJ), and the chaperonin CCT (GroEL). The cytosolic molecular chaperones Hsp70s/Hspll0s and the chaperonins are highly upregulated in bacterial and human cells under different stresses and are involved both in the prevention and the reversion of protein misfolding and aggregation. Hsp70 works in collaboration with Hsp40 to reactivate misfolded or aggregated proteins in a strict ATP dependent manner. Chaperonins (CCT and GroEL) also unfold and reactivate stably misfolded proteins but we found that it needed to use the energy of ATP hydrolysis in order to evict over- sticky misfolded intermediates that inhibited the unfoldase catalytic sites. Ill In this study, we initially characterized a particular type of inactive misfolded monomeric luciferase and rhodanese species that were obtained by repeated cycles of freeze-thawing (FT). These stable misfolded monomeric conformers (FT-luciferase and FT-rhodanese) had exposed hydrophobic residues and were enriched with wrong ß-sheet structures (Chapter 2). Using FT-luciferase as substrate, we found that the Hsp70 orthologs, called HspllO (Sse in yeast), acted similarly to Hsp70 as were bona fide ATP- fuelled polypeptide unfoldases and was much more than a mere nucleotide exchange factor, as generally thought. Moreover, we found that HspllO collaborated with Hsp70 in the disaggregation of stable protein aggregates in which Hsp70 and HspllO acted as equal partners that synergistically combined their individual ATP-consuming polypeptide unfoldase activities to reactivate the misfolded/aggregated proteins (Chapter 3). Using FT-rhodanese as substrate, we found that chaperonins (GroEL and CCT) could catalytically reactivate misfolded rhodanese monomers in the absence of ATP. Also, our results suggested that encaging of an unfolding polypeptide inside the GroEL cavity under a GroES cap was not an obligatory step as generally thought (Chapter 4). Further, we investigated the role of Hsp40, a J-protein co-chaperone of Hsp70, in targeting misfolded polypeptides substrates onto Hsp70 for unfolding. We found that even a large excess of monomeric unfolded a-synuclein did not inhibit DnaJ, whereas, in contrast, stable misfolded a-synuclein oligomers strongly inhibited the DnaK-mediated chaperone reaction by way of sequestering the DnaJ co-chaperone. This work revealed that DnaJ could specifically distinguish, and bind potentially toxic stably aggregated species, such as soluble a-synuclein oligomers involved in Parkinson's disease, and with the help of DnaK and ATP convert them into from harmless natively unfolded a-synuclein monomers (chapter 5). Finally, our meta-analysis of microarray data of plant and animal tissues treated with various chemicals and abiotic stresses, revealed possible co-expressions between core chaperone machineries and their co-chaperone regulators. It clearly showed that protein misfolding in the cytosol elicits a different response, consisting of upregulating the synthesis mainly of cytosolic chaperones, from protein misfolding in the endoplasmic reticulum (ER) that elicited a typical unfolded protein response (UPR), consisting of upregulating the synthesis mainly of ER chaperones. We proposed that drugs that best mimicked heat or UPR stress at increasing the chaperone load in the cytoplasm or ER respectively, may prove effective at combating protein misfolding diseases and aging (Chapter 6).  - Dans les conditions optimales de basse concentration et de basse température, les protéines vont spontanément adopter un repliement natif car toutes les informations nécessaires se trouvent dans la séquence des acides aminés du polypeptide. En revanche, dans des conditions de stress ou de forte concentration des protéines comme à l'intérieur d'une cellule, un polypeptide peu mal se replier et entrer dans un processus d'agrégation conduisant à la formation de conformères et de fibrilles qui peuvent être toxiques et causer des maladies neurodégénératives comme la maladie d'Alzheimer, la maladie de Parkinson ou la chorée de Huntington. Afin d'empêcher ou de rectifier le mauvais repliement des protéines, les cellules ont développé des protéines appelées chaperonnes. Dans ce travail, je me suis intéressé aux chaperonnes cytosoliques Hsp70 (DnaK) et HspllO, la co-chaperones Hsp40 (DnaJ), le complexe CCT/TRiC et GroEL. Chez les bactéries et les humains, les chaperonnes cytosoliques Hsp70s/Hspl 10s et les « chaperonines» sont fortement activées par différentes conditions de stress et sont toutes impliquées dans la prévention et la correction du mauvais repliement des protéines et de leur agrégation. Hsp70 collabore avec Hsp40 pour réactiver les protéines agrégées ou mal repliées et leur action nécessite de 1ATP. Les chaperonines (GroEL) déplient et réactivent aussi les protéines mal repliées de façon stable mais nous avons trouvé qu'elles utilisent l'ATP pour libérer les intermédiaires collant et mal repliés du site catalytique de dépliage. Nous avons initialement caractérisé un type particulier de formes stables de luciférase et de rhodanese monomériques mal repliées obtenues après plusieurs cycles de congélation / décongélation répétés (FT). Ces monomères exposaient des résidus hydrophobiques et étaient plus riches en feuillets ß anormaux. Ils pouvaient cependant être réactivés par les chaperonnes Hsp70+Hsp40 (DnaK+DnaJ) et de l'ATP, ou par Hsp60 (GroEL) sans ATP (Chapitre 2). En utilisant la FT-Luciferase comme substrat nous avons trouvé que HspllO (un orthologue de Hsp70) était une authentique dépliase, dépendante strictement de l'ATP. De plus, nous avons trouvé que HspllO collaborait avec Hsp70 dans la désagrégation d'agrégats stables de protéines en combinant leurs activités dépliase consommatrice d'ATP (Chapitre 3). En utilisant la FT-rhodanese, nous avons trouvé que les chaperonines (GroEL et CCT) pouvaient réactiver catalytiquement des monomères mal repliés en absence d'ATP. Nos résultats suggérèrent également que la capture d'un polypeptide en cours de dépliement dans la cavité de GroEL et sous un couvercle du complexe GroES ne serait pas une étape obligatoire du mécanisme, comme il est communément accepté dans la littérature (Chapitre 4). De plus, nous avons étudié le rôle de Hsp40, une co-chaperones de Hsp70, dans l'adressage de substrats polypeptidiques mal repliés vers Hsp70. Ce travail a révélé que DnaJ pouvait différencier et lier des polypeptide mal repliés (toxiques), comme des oligomères d'a-synucléine dans la maladie de Parkinson, et clairement les différencier des monomères inoffensifs d'a-synucléine (Chapitre 5). Finalement une méta-analyse de données de microarrays de tissus végétaux et animaux traités avec différents stress chimiques et abiotiques a révélé une possible co-expression de la machinerie des chaperonnes et des régulateurs de co- chaperonne. Cette meta-analyse montre aussi clairement que le mauvais repliement des protéines dans le cytosol entraîne la synthèse de chaperonnes principalement cytosoliques alors que le mauvais repliement de protéines dans le réticulum endoplasmique (ER) entraine une réponse typique de dépliement (UPR) qui consiste principalement en la synthèse de chaperonnes localisées dans l'ER. Nous émettons l'hypothèse que les drogues qui reproduisent le mieux les stress de chaleur ou les stress UPR pourraient se montrer efficaces dans la lutte contre le mauvais repliement des protéines et le vieillissement (Chapitre 6).

Comparison of molecular signatures from multiple skin diseases identifies mechanisms of immunopathogenesis.

The ability to obtain gene expression profiles from human disease specimens provides an opportunity to identify relevant gene pathways, but is limited by the absence of data sets spanning a broad range of conditions. Here, we analyzed publicly available microarray data from 16 diverse skin conditions in order to gain insight into disease pathogenesis. Unsupervised hierarchical clustering separated samples by disease as well as common cellular and molecular pathways. Disease-specific signatures were leveraged to build a multi-disease classifier, which predicted the diagnosis of publicly and prospectively collected expression profiles with 93% accuracy. In one sample, the molecular classifier differed from the initial clinical diagnosis and correctly predicted the eventual diagnosis as the clinical presentation evolved. Finally, integration of IFN-regulated gene programs with the skin database revealed a significant inverse correlation between IFN-β and IFN-γ programs across all conditions. Our study provides an integrative approach to the study of gene signatures from multiple skin conditions, elucidating mechanisms of disease pathogenesis. In addition, these studies provide a framework for developing tools for personalized medicine toward the precise prediction, prevention, and treatment of disease on an individual level.

Highlight Removal in Spectral Images

Monissasovelluksissa on hyvin tärkeää vähentää valolähteen vaikutusta kohteen oikean värin havainnoimiseksi. Tämä on tarpeen mm. virtuaalisissa museoissa, telelääketieteessä, verkkokaupassa ja verkkorahassa. Tässä tutkielmassa on kehitetty tekniikkaa kirkkaiden heijastusten poistoon spektrikuvista. Työ sisältää katsauksen yleisen värillisen kuvan ymmärtämiseen, mihin perustuen analysoitiin erilaisia kirkkaiden heijastusten poistO'tekniikoita. Työssä kehitettiin uusi kirkkaiden heijastusten poistO'menetelmä, joka perustuu dikromaattiseen heijastus-malliin, joka kuvaa spektrisen datan objektin omaan väriin ja valaisevan valon väriin perustuen. Ehdotettu kirkkaiden heijastusten poistO'menetelmä hyödyntää erilaisia olemassaolevia menetelmiä, kuten pääkomponenttimenetelmää ja tiedon luokittelu-menetelmää. Yritys kehittää nopeasti toimiva algoritmi, joka myös suoriutuu tehtävästä hyvin, on onnistunut. Kokeet toteutettiin ehdotetun menetelmän mukaisesti ja toimivalla algoritmilla saatiin halutut lopputulokset. Edelleentyö sisältää ehdotuksia esitetyn algoritmin parantamiseksi.

Gene filtering with optimal threshold selection

Gene filtering is a useful preprocessing technique often applied to microarray datasets. However, it is no common practice because clear guidelines are lacking and it bears the risk of excluding some potentially relevant genes. In this work, we propose to model microarray data as a mixture of two Gaussian distributions that will allow us to obtain an optimal filter threshold in terms of the gene expression level.

Many valued algebraic structures as measures of comparison

This thesis studies the properties and usability of operators called t-norms, t-conorms, uninorms, as well as many valued implications and equivalences. Into these operators, weights and a generalized mean are embedded for aggregation, and they are used for comparison tasks and for this reason they are referred to as comparison measures. The thesis illustrates how these operators can be weighted with a differential evolution and aggregated with a generalized mean, and the kinds of measures of comparison that can be achieved from this procedure. New operators suitable for comparison measures are suggested. These operators are combination measures based on the use of t-norms and t-conorms, the generalized 3_-uninorm and pseudo equivalence measures based on S-type implications. The empirical part of this thesis demonstrates how these new comparison measures work in the field of classification, for example, in the classification of medical data. The second application area is from the field of sports medicine and it represents an expert system for defining an athlete's aerobic and anaerobic thresholds. The core of this thesis offers definitions for comparison measures and illustrates that there is no actual difference in the results achieved in comparison tasks, by the use of comparison measures based on distance, versus comparison measures based on many valued logical structures. The approach has been highly practical in this thesis and all usage of the measures has been validated mainly by practical testing. In general, many different types of operators suitable for comparison tasks have been presented in fuzzy logic literature and there has been little or no experimental work with these operators.

High-Throughput Screening for Novel Prostate Cancer Drug Targets –Getting Personal

Prostate cancers form a heterogeneous group of diseases and there is a need for novel biomarkers, and for more efficient and targeted methods of treatment. In this thesis, the potential of microarray data, RNA interference (RNAi) and compound screens were utilized in order to identify novel biomarkers, drug targets and drugs for future personalized prostate cancer therapeutics. First, a bioinformatic mRNA expression analysis covering 9873 human tissue and cell samples, including 349 prostate cancer and 147 normal prostate samples, was used to distinguish in silico prevalidated putative prostate cancer biomarkers and drug targets. Second, RNAi based high-throughput (HT) functional profiling of 295 prostate and prostate cancer tissue specific genes was performed in cultured prostate cancer cells. Third, a HT compound screen approach using a library of 4910 drugs and drug-like molecules was exploited to identify potential drugs inhibiting prostate cancer cell growth. Nine candidate drug targets, with biomarker potential, and one cancer selective compound were validated in vitro and in vivo. In addition to androgen receptor (AR) signaling, endoplasmic reticulum (ER) function, arachidonic acid (AA) pathway, redox homeostasis and mitosis were identified as vital processes in prostate cancer cells. ERG oncogene positive cancer cells exhibited sensitivity to induction of oxidative and ER stress, whereas advanced and castrate-resistant prostate cancer (CRPC) could be potentially targeted through AR signaling and mitosis. In conclusion, this thesis illustrates the power of systems biological data analysis in the discovery of potential vulnerabilities present in prostate cancer cells, as well as novel options for personalized cancer management.

Algorithmic Techniques in Gene Expression Processing. From Imputation to Visualization

The amount of biological data has grown exponentially in recent decades. Modern biotechnologies, such as microarrays and next-generation sequencing, are capable to produce massive amounts of biomedical data in a single experiment. As the amount of the data is rapidly growing there is an urgent need for reliable computational methods for analyzing and visualizing it. This thesis addresses this need by studying how to efficiently and reliably analyze and visualize high-dimensional data, especially that obtained from gene expression microarray experiments. First, we will study the ways to improve the quality of microarray data by replacing (imputing) the missing data entries with the estimated values for these entries. Missing value imputation is a method which is commonly used to make the original incomplete data complete, thus making it easier to be analyzed with statistical and computational methods. Our novel approach was to use curated external biological information as a guide for the missing value imputation. Secondly, we studied the effect of missing value imputation on the downstream data analysis methods like clustering. We compared multiple recent imputation algorithms against 8 publicly available microarray data sets. It was observed that the missing value imputation indeed is a rational way to improve the quality of biological data. The research revealed differences between the clustering results obtained with different imputation methods. On most data sets, the simple and fast k-NN imputation was good enough, but there were also needs for more advanced imputation methods, such as Bayesian Principal Component Algorithm (BPCA). Finally, we studied the visualization of biological network data. Biological interaction networks are examples of the outcome of multiple biological experiments such as using the gene microarray techniques. Such networks are typically very large and highly connected, thus there is a need for fast algorithms for producing visually pleasant layouts. A computationally efficient way to produce layouts of large biological interaction networks was developed. The algorithm uses multilevel optimization within the regular force directed graph layout algorithm.

Modélisation des bi-grappes et sélection des variables pour des données de grande dimension : application aux données d’expression génétique

Les simulations ont été implémentées avec le programme Java.

Mécanismes moléculaires de la réplication préférentielle du VIH-1 dans les cellules à polarisation Th1Th17 versus Th1 : rôle de PPARG dans la régulation négative de la réplication virale

Les cellules T CD4+ humaines sont hétérogènes du point de vue de la permissivité à l’infection par le virus de l’immunodéficience humaine de type 1 (VIH-1). Notre laboratoire a préalablement démontré que les cellules Th1 à phénotype CXCR3+CCR6- sont relativement résistantes à l’infection par le VIH-1 alors que les cellules Th1Th17 à phénotype CXCR3+CCR6+ y sont hautement permissives. La réplication du VIH dépend de plusieurs facteurs cellulaires de restriction ou de permissivité agissant à différentes étapes du cycle viral. Toutefois, malgré plusieurs avancées, la compréhension des voies de signalisation cellulaire impliquées dans la régulation de la réplication du VIH est encore limitée. L’objectif majeur de ce projet de maîtrise est de caractériser les mécanismes moléculaires de la permissivité et de la résistance au VIH respectivement dans les cellules Th1Th17 et Th1. Ce mémoire est divisé en quatre parties qui visent: (i) l’identification des voies canoniques et des fonctions biologiques différemment régulées dans les cellules Th1Th17 versus Th1 par l’analyse de leur transcriptome au niveau du génome entier; (ii) la validation de l’expression différentielle des gènes d’intérêt identifiés par biopuces au niveau des transcrits et des protéines; (iii) la caractérisation du rôle fonctionnel de certains de ces facteurs (i.e., PPARG, AhR) sur la réplication du VIH dans les cellules Th1Th17 versus Th1; et (iv) l’identification du niveau auquel ces facteurs interfèrent avec le cycle de réplication du VIH. Nos résultats d’analyse du transcriptome du génome entier par Gene Set Enrichment Analysis et Ingenuity Pathway Analysis indiquent que les cellules à profil Th1Th17 sont plus susceptibles à l’activation cellulaire et à l’apoptose, favorisent plus l’inflammation et expriment moins fortement les gènes liés à la dégradation protéosomale comparé aux cellules à profil Th1. Ces différences dans la régulation de diverses voies et fonctions biologiques permettent en partie d’expliquer la susceptibilité à l’infection par le VIH dans ces cellules. Nous avons ensuite confirmé l’expression différentielle de certains gènes d’intérêt dans les cellules Th1Th17 (CXCR6, PPARG, ARNTL, CTSH, PTPN13, MAP3K4) versus Th1 (SERPINB6, PTK2) au niveau de l’ARNm et des protéines. Finalement, nous avons démontré le rôle des facteurs de transcription PPARG et AhR dans la régulation de la réplication du VIH. L’activation de la voie PPARG par la rosiglitazone induit la diminution importante de la réplication du VIH dans les cellules T CD4+, alors que l’activation de la voie AhR par les ligands exogènes TCDD et FICZ augmente de façon significative la réplication virale. Nous proposons que la voie PPARG agit comme un régulateur négatif de la réplication du VIH dans ces cellules, en interférant avec la polarisation Th17 et probablement en inhibant l’activité transcriptionnelle du facteur NF-kB. Les rôles des formes nucléaires versus cytoplasmiques du récepteur Ahr semblent être diamétralement opposés, dans la mesure où l’interférence ARN contre AhR s’associe également à l’augmentation de la réplication virale. Il est ainsi possible que la forme cytoplasmique d’AhR, connue par son activité E3 ligase, participe à la dégradation protéosomale des particules virales. Le mécanisme par lequel le AhR nucléaire versus cytoplasmique interfère avec la réplication virale est en cours d’étude au laboratoire. Cette étude représente la première caractérisation de l’expression différentielle de gènes au niveau du génome entier de sous-populations T CD4+ permissives versus résistantes à l’infection par le VIH. Nos résultats identifient de nouvelles cibles moléculaires pour de nouvelles stratégies thérapeutiques visant à limiter la réplication du VIH dans les lymphocytes T CD4+ primaires.

Actinobacillus pleuropneumoniae induces SJPL cell cycle arrest in G2/M-phase and inhibits porcine reproductive and respiratory syndrome virus replication

Background: Porcine reproductive and respiratory syndrome virus (PRRSV) is one of the most important pathogens in the swine industry and causes important economic losses. No effective antiviral drugs against it are commercially available. We recently reported that the culture supernatant of Actinobacillus pleuropneumoniae, the porcine pleuropneumonia causative agent, has an antiviral activity in vitro against PRRSV in SJPL cells. Objectives of this study were (i) to identify the mechanism behind the antiviral activity displayed by A. pleuropneumoniae and (ii) to characterize the active molecules present in the bacterial culture supernatant. Methods: Antibody microarray analysis was used in order to point out cellular pathways modulated by the A. pleuropneumoniae supernatant. Subsequent, flow cytometry analysis and cell cycle inhibitors were used to confirm antibody microarray data and to link them to the antiviral activity of the A. pleuropneumoniae supernatant. Finally, A. pleuropneumoniae supernatant characterization was partially achieved using mass spectrometry. Results: Using antibody microarray, we observed modulations in G2/M-phase cell cycle regulation pathway when SJPL cells were treated with A. pleuropneumoniae culture supernatant. These modulations were confirmed by a cell cycle arrest at the G2/M-phase when cells were treated with the A. pleuropneumoniae culture supernatant. Furthermore, two G2/M-phase cell cycle inhibitors demonstrated the ability to inhibit PRRSV infection, indicating a potential key role for PRRSV infection. Finally, mass spectrometry lead to identify two molecules (m/z 515.2 and m/z 663.6) present only in the culture supernatant. Conclusions: We demonstrated for the first time that A. pleuropneumoniae is able to disrupt SJPL cell cycle resulting in inhibitory activity against PRRSV. Furthermore, two putative molecules were identified from the culture supernatant. This study highlighted the cell cycle importance for PRRSV and will allow the development of new prophylactic or therapeutic approaches against PRRSV.