Interaction preferences in protein - DNA complexes Dávid Jakubec Abstract Interactions of proteins with DNA lie at the basis of many fundamental bio- logical processes. Despite ongoing efforts, the rules governing the recognition of specific nucleic acid sequences have still not been universally elucidated. In this work, I attempt to explore the recognition process by splitting the intricate network of contacts at the protein - DNA interface into contribu- tions of individual amino acid - nucleotide pairs. These pairs are extracted from existing high-resolution structures of protein - DNA complexes and in- vestigated by bioinformatics and computational-chemistry based methods. Criteria of specificity based on the coupling of observed geometrical prefer- ences and the respective interaction energies are introduced. The application of these criteria is used to expand the library of amino acid - nucleotide pairs potentially significant for direct sequence recognition. Electrostatic poten- tial maps are calculated for individual nucleotides as well as for selected complexes to investigate the physical basis of the observed specificity. 1

O diabetes mellitus tipo 1 (DM1) tem etiologia autoimune, enquanto o diabetes mellitus tipo 2 (DM2) e o diabetes mellitus gestacional (DMG) são considerados como distúrbios metabólicos. Neste trabalho, foi realizada análise do transcriptoma das células mononucleares do sangue periférico (do inglês, peripheral mononuclear blood cells - PBMCs), obtidas de pacientes com DM1, DM2 e DMG, realizando análises por module maps a fim de comparar características patogênicas e aspectos gerais do tratamento com anotações disponíveis de genes modulados, tais como: a) análises disponíveis a partir de estudos de associação em larga escala (do inglês genome-wide association studies GWAS); b) genes associados ao diabetes em estudos clássicos de ligação disponíveis em bancos de dados públicos; c) perfis de expressão de células imunológicas fornecidos pelo grupo ImmGen (Immunological Project). Foram feitos microarrays do transcriptoma total da plataforma Agilent (Whole genome onecolor Agilent 4x44k) para 56 pacientes (19 DM1, 20 DM2 e 17 DMG). Para a compreensão dos resultados foram aplicados filtros não-informativos e as listas de genes diferencialmente expressos foram obtidas por análise de partição e análise estatística não-paramétrica (rank products), respectivamente. Posteriormente, análises de enriquecimento funcional foram feitas pelo DAVID e os module maps construídos usando a ferramenta Genomica. As análises funcionais contribuíram para discriminar os pacientes a partir de genes envolvidos na inflamação, em especial DM1 e DMG. Os module maps de genes diferencialmente expressos revelaram: a) genes modulados exibiram perfis de transcrição típicos de macrófagos e células dendríticas, b) genes modulados foram associados com genes previamente descritos como genes de complicação ao diabetes a partir de estudos de ligação e de meta-análises; c) a duração da doença, obesidade, número de gestações, níveis de glicose sérica e uso de medicações, tais como metformina, influenciaram a expressão gênica em pelo menos um tipo de diabetes. Esse é o primeiro estudo de module maps mostrando a influência de padrões epidemiológicos, clínicos, laboratoriais, imunopatogênicos e de tratamento na modulação dos perfis transcricionais em pacientes com os três tipos clássicos de diabetes: DM1, DM2 e DMG. / Type 1 diabetes (T1D) is an autoimmune disease while type 2 (T2D) and gestational diabetes (GDM) are considered as metabolic disturbances. We performed a transcriptome analysis of peripheral mononuclear blood cells obtained from T1D, T2D and GDM patients, and we took advantage of the module map approach to compare pathogenic and treatment features of our patient series with available annotation of modulated genes from i) genome-wide association studies; ii) genes provided by diabetes meta-analysis in public databases, iii) immune cell gene expression profiles provided by the ImmGen project. Whole genome one-color Agilent 4x44k microarray was performed for 56 (19 T1D, 20 T2D, 17 GDM) patients. Noninformative filtered and differentially expressed genes were obtained by partitioning and rank product analysis, respectively. Functional analyses were carried out using the DAVID software and module maps were constructed using the Genomica tool. Functional analyses contributed to discriminate patients on the basis of genes involved in inflammation, primarily for T1D and GDM. Module maps of differentially expressed genes revealed that: i) modulated genes exhibited transcription profiles typical of macrophage and dendritic cells, ii) modulated genes were associated with previously reported diabetes complication genes disclosed by association and meta-analysis studies, iii) disease duration, obesity, number of gestations, glucose serum levels and the use of medications, such as metformin, influenced gene expression profiles in at least one type of diabetes. This is the first module map study to show the influence of epidemiological, clinical, laboratory, immunopathogenic and treatment features on the modulation of the transcription profiles of T1D, T2D and GDM patients.

Le transport de composés à travers les membranes biologiques est crucial pour la physiologie des cellules eucaryotes. Cependant la fonction de nombreux transporteurs putatifs reste inconnue. C’est notamment le cas de nombreux transporteurs intracellulaires exportant les catabolites du lysosome. Le transporteur lysosomal de cystine, baptisé cystinosine, se caractérise par la présence d’un motif dupliqué appelé " boucle PQ ". Sa dysfonction entraîne une maladie lysosomale, la cystinose, caractérisée par l'accumulation de cystine dans les lysosomes. Les protéines possédant un motif PQ sont retrouvées plus souvent dans les cellules eucaryotes et, à l'exception de la cystinosine, leur fonction reste inconnue. Dans cette thèse, nous démontrons qu'une autre protéine à motif PQ, PQLC2 est le transporteur responsable de l'efflux lysosomal des acides aminés cationiques et qu'il est impliqué dans le traitement de la cystinose.L'hypothèse de départ était basée, d'une part, par sur des prédictions par analyse protéomique de la localisation lysosomale de PQLC2 et, d'autre part, sur des résultats chez S.cerevisiae impliquant les orthologues putatifs de PQLC2, situés à la membrane de la vacuole, dans l'homéostasie des acides aminés cationiques. En utilisant une approche consistant à délocaliser PQLC2 à la membrane plasmique et à acidifier le pH extracellulaire pour mimer la lumière acide du lysosome, nous avons pu, par mesure d'accumulation intracellulaire de composés radiomarqués et par mesure électrophysiologique sur cellule entière, faire la preuve du transport sélectif, actif à bas pH et de faible affinité des acides aminés cationiques par PQLC2. Dans une seconde partie, nous avons mis en évidence l'implication de ce transporteur dans l'efflux lysosomal du produit de réaction entre la cystine accumulée dans les lysosomes de cellules de patients cystinotiques et le principe actif (cystéamine) du traitement pharmacologique de la cystinose.Enfin, dans une dernière partie, nous avons effectué une analyse bioinformatique préliminaire des protéines à motif PQ qui exploitait la pseudo-symétrie de ces protéines pour identifier des résidus potentiellement impliqués dans l'activité de transport. / Transport of solutes across biological membranes is crucial to eukaryotic cell physiology. However, the function of many putative transporters remains unknown, such as the proteins responsible for lysosomal export of metabolites. Cystinosin, the lysosomal cystine exporter defective in cystinosis, is characterized by a duplicated motif termed the PQ loop. PQ-loop proteins are more frequent in eukaryotes than in prokaryotes, and, except for cystinosin, their molecular function remains unknown. Here we show that another PQ-loop protein, PQLC2, is a lysosomal transporter for cationic amino acids and that it is required for the treatment of cystinosis. The hypothesis that PQLC2 is a lysosomal metabolite transporter was based on a proteomic study predicting that PQLC2 is located at the lysosomal membrane and on a genetic study that linked putative yeast orthologues with cationic amino acid homeostasis. Using an approach that consisted in misrouting PQLC2 to the plasma membrane of frog oocytes and in acidifying the extracellular medium to mimic the acidic lysosomal lumen, we showed an accumulation of radiolabelled cationic amino acids into mRNA-injected oocytes and an electrogenic, inward current due to a selective, pH-dependent, low-affinity transport of cationic amino acids by PQLC2. Moreoever, we showed that PQLC2 exports a key chemical intermediate (cysteamine-cysteine mixed disulfide) from cystinotic lysosomes treated with the aminothiol drug cysteamine, thus explaining the mechanism underlying the current drug therapy of cystinosis. Finally, in a last chapter, we performed a preliminary bioinformatic study of the family of PQ-loop proteins that took advantage of the pseudo-symmetric structure of these proteins to identify residues potentially important for the transport activity.

Le potentiel régulateur des lymphocytes B (LB), largement associé avec la production d’interleukine-10 (IL-10), a été mis en évidence dans des modèles murins de pathologies spécifiques d’Ag. Les cellules B transitionnelles (Tr.) CD24fortes CD38fortes ont été décrites comme régulatrices, au travers de la production d’IL-10, de l’inhibition de la prolifération T, ainsi que de la suppression de la réponse inflammatoire des cellules T. Les LB transitionnels représentent un stade de développement central dans la maturation des cellules B, en faisant le lien entre les cellules immatures de la moelle osseuse et celles matures situées dans les organes lymphoïdes secondaires. Dans une première étude, nous montrons que cette population est hétérogène, et composée de LB Tr. de type 1 (T1), T2, T3 et Tr. CD27+. Les LB T3 anergiques semblent jouer un rôle dans la tolérance périphérique en limitant la prolifération des lymphocytes T (LT) CD4+, tandis que les LB Tr. CD27+ IL-10+ nouvellement décrits inhibent la différenciation des LT CD4+ en cellules productrices d’IFN-γ et de TNF-α. Notons que les LB T1 et Tr. CD27+ se différencient rapidement en cellules productrices d’Ac suite à la reconnaissance de signaux de l’immunité innée. La production d’IL-10 est en partie dépendante des signaux perçus, provenant du microenvironnement. Nous avons décrit dans un second travail que les LB s’adaptent aux cellules avec lesquelles ils sont cultivés. En effet, les cellules B régulent spécifiquement les LT CD4+ mémoires (et non naïfs), en limitant leur prolifération avant d’induire une mort cellulaire. Ces caractéristiques fonctionnelles pourraient être associées avec une modification du programme transcriptionnel, permise par la plasticité des cellules B, qui se polarisent en LB régulateurs (Breg) de façon ciblée. L’expression des gènes PRDM1 et IL10 serait associée avec une signature Breg spécifique en culture mixte autologue, en opposition avec celle des gènes NFκB1 et BCL6. La transplantation rénale est un excellent modèle physiopathologique, pour étudier l’importance de certaines populations de LB dans la tolérance immunologique. L’étude BHL (B lymphocytes in humoral rejection and alloimmunisation) nous a permis de confirmer que les LB Tr. ont probablement un rôle important dans cette tolérance du greffon. La présence d’anticorps spécifiques du donneur (DSA) semble limiter l’émergence des LB Tr., même si le pourcentage de cellules B CD24fortes CD38fortes n’est a priori pas associé avec la capacité du compartiment lymphocytaire B à réguler la prolifération des cellules T des patients alloimmunisés. / Regulatory B cells (Breg) were first reported to be interleukine-10 (IL-10) producing B cells in mice. The almost concurrent discovery of Breg cells drew interest toward potential links with transitional B cells because of phenotypic and functional similarities. In addition with IL-10 production, CD24high CD38high transitional B cells limit the proliferation of T cells and the polarization of CD4+ T cells into Th1 cells. Transitional B cells represent a central developmental stage in B-cell maturation, linking generation in the bone marrow with differentiation in periphery. In a first study, we reveal for the first time that human transitional B cells encompass not only transitional type 1 and type 2 B cells, but also distinct anergic type 3 B cells, as well as IL-10-producing CD27+ transitional B cells. Interestingly, the latter two subsets differentially regulate CD4+ T-cell proliferation and polarization toward Th1 effector cells. Additional experiments showed that type 1 and CD27+ transitional B cells are capable to differentiate into antibody secreting cells after toll-like receptor 9 engagement. In a second work, we wanted to explore the ability of B cells to target T-cell populations. We demonstrate that B cells can be suppressive cells. B cells are capable to target CD4+ memory T-cell, limiting the proliferation and inducing the death of this T-cell population. At the opposite, B cells seem to be effector of CD4+ naïve T-cell functions. These properties are probably associated with a specific transcriptional program. Thus, we observed that suppressive B cells overexpress PRDM1 and IL10, whereas effector B cells preferentially express BCL6 and NFκB1 in in vitro mixed culture. In the last part, we worked on B-cell phenotype and functions in transplanted patients. BHL (B lymphocytes in humoral rejection and alloimmunisation) is a clinical study that aims to better understand the role of B cells in the alloimmunisation and the chronic rejection occurring after renal transplantation. Donor specific antibodies (DSA) seem to limit the expansion of transitional B cells, which are probably not associated with the ability of B cells to regulate T-cell proliferation in DSA+ patients.

O diabetes mellitus tipo 1 (DM1) tem etiologia autoimune, enquanto o diabetes mellitus tipo 2 (DM2) e o diabetes mellitus gestacional (DMG) são considerados como distúrbios metabólicos. Neste trabalho, foi realizada análise do transcriptoma das células mononucleares do sangue periférico (do inglês, peripheral mononuclear blood cells - PBMCs), obtidas de pacientes com DM1, DM2 e DMG, realizando análises por module maps a fim de comparar características patogênicas e aspectos gerais do tratamento com anotações disponíveis de genes modulados, tais como: a) análises disponíveis a partir de estudos de associação em larga escala (do inglês genome-wide association studies GWAS); b) genes associados ao diabetes em estudos clássicos de ligação disponíveis em bancos de dados públicos; c) perfis de expressão de células imunológicas fornecidos pelo grupo ImmGen (Immunological Project). Foram feitos microarrays do transcriptoma total da plataforma Agilent (Whole genome onecolor Agilent 4x44k) para 56 pacientes (19 DM1, 20 DM2 e 17 DMG). Para a compreensão dos resultados foram aplicados filtros não-informativos e as listas de genes diferencialmente expressos foram obtidas por análise de partição e análise estatística não-paramétrica (rank products), respectivamente. Posteriormente, análises de enriquecimento funcional foram feitas pelo DAVID e os module maps construídos usando a ferramenta Genomica. As análises funcionais contribuíram para discriminar os pacientes a partir de genes envolvidos na inflamação, em especial DM1 e DMG. Os module maps de genes diferencialmente expressos revelaram: a) genes modulados exibiram perfis de transcrição típicos de macrófagos e células dendríticas, b) genes modulados foram associados com genes previamente descritos como genes de complicação ao diabetes a partir de estudos de ligação e de meta-análises; c) a duração da doença, obesidade, número de gestações, níveis de glicose sérica e uso de medicações, tais como metformina, influenciaram a expressão gênica em pelo menos um tipo de diabetes. Esse é o primeiro estudo de module maps mostrando a influência de padrões epidemiológicos, clínicos, laboratoriais, imunopatogênicos e de tratamento na modulação dos perfis transcricionais em pacientes com os três tipos clássicos de diabetes: DM1, DM2 e DMG. / Type 1 diabetes (T1D) is an autoimmune disease while type 2 (T2D) and gestational diabetes (GDM) are considered as metabolic disturbances. We performed a transcriptome analysis of peripheral mononuclear blood cells obtained from T1D, T2D and GDM patients, and we took advantage of the module map approach to compare pathogenic and treatment features of our patient series with available annotation of modulated genes from i) genome-wide association studies; ii) genes provided by diabetes meta-analysis in public databases, iii) immune cell gene expression profiles provided by the ImmGen project. Whole genome one-color Agilent 4x44k microarray was performed for 56 (19 T1D, 20 T2D, 17 GDM) patients. Noninformative filtered and differentially expressed genes were obtained by partitioning and rank product analysis, respectively. Functional analyses were carried out using the DAVID software and module maps were constructed using the Genomica tool. Functional analyses contributed to discriminate patients on the basis of genes involved in inflammation, primarily for T1D and GDM. Module maps of differentially expressed genes revealed that: i) modulated genes exhibited transcription profiles typical of macrophage and dendritic cells, ii) modulated genes were associated with previously reported diabetes complication genes disclosed by association and meta-analysis studies, iii) disease duration, obesity, number of gestations, glucose serum levels and the use of medications, such as metformin, influenced gene expression profiles in at least one type of diabetes. This is the first module map study to show the influence of epidemiological, clinical, laboratory, immunopathogenic and treatment features on the modulation of the transcription profiles of T1D, T2D and GDM patients.

<div><div>In the forensic science community, there is a vast need for tools to help assist investigations when standard DNA profiling methods are uninformative. Methods such as Forensic DNA Phenotyping (FDP) and proteomics aims to help this problem and provide aid in investigations when other methods have been exhausted. FDP is useful by providing physical appearance information, while proteomics allows for the examination of difficult samples, such as hair, to infer human identity and ancestry. To create a “biological eye witness” or develop informative probability of identity match statistics through proteomically inferred genetic profiles, it is necessary to constantly strive to improve these methods. </div><div><br></div><div>Currently, two developmentally validated FDP prediction assays, ‘HIrisPlex’ and ‘HIrisplex-S’, are used on the capillary electrophoresis to develop a phenotypic prediction for eye, hair, and skin color based on 41 variants. Although highly useful, these assays are limited in their ability when used on the CE due to a 25 variant per assay cap. To overcome these limitations and expand the capacities of FDP, we successfully designed and validated a massive parallel sequencing (MPS) assay for use on both the ThermoFisher Scientific Ion Torrent and Illumina MiSeq systems that incorporates all HIrisPlex-S variants into one sensitive assay. With the migration of this assay to an MPS platform, we were able to create a semi-automated pipeline to extract SNP-specific sequencing data that can then be easily uploaded to the freely accessible online phenotypic prediction tool (found at https://hirisplex.erasmusmc.nl) and a mixture deconvolution tool with built-in read count thresholds. Based on sequencing reads counts, this tool can be used to assist in the separation of difficult two-person mixture samples and outline the confidence in each genotype call.<br></div><div><br></div><div>In addition to FDP, proteomic methods, specifically in hair protein analysis, opens doors and possibilities for forensic investigations when standard DNA profiling methods come up short. Here, we analyzed 233 genetically variant peptides (GVPs) within hair-associated proteins and genes for 66 individuals. We assessed the proteomic methods ability to accurately infer and detect genotypes at each of the 233 SNPs and generated statistics for the probability of identity (PID). Of these markers, 32 passed all quality control and population genetics criteria and displayed an average PID of 3.58 x 10-4. A population genetics assessment was also conducted to identify any SNP that could be used to infer ancestry and/or identity. Providing this information is valuable for the future use of this set of markers for human identification in forensic science settings. </div></div><div><br></div>

Les plantes, par leurs racines, offrent une myriade de niches écologiques pour les microorganismes du sol, et ceux-ci la protègent contre les attaques parasitaires et les stress abiotiques, et favorisent son approvisionnement en nutriments et en eau. Cependant, dans le sol, la plante joue aussi un rôle important lorsqu’elle émet depuis ses racines des composés qui influencent la composition des communautés microbiennes dudit sol, ce combiné à un changement du pH du sol par la plante et son apport en matière organique ainsi qu’en oxygène. Ces composés influencent les membres du microbiote souterrain de la plante et donc indirectement la plante elle-même. Plus on a une diversité du couvert végétal, plus la diversité des microorganismes du sol va être élevée et inversement, plus un sol sera divers en matière de microbes plus les plantes qui y poussent tendent à être en bonne santé. Pour une plante en particulier, il n’est pas inhabituel de développer des relations spécifiques avec des microorganismes eux aussi spécifiques qui vont améliorer sa survie. Cependant, une plante peut vivre dans différents environnements et les sols sont divers, donc les plantes doivent s’adapter aux microbes qu’elles trouvent à proximité en sélectionnant les microbes les plus bénéfiques pour elles. Du coup, il est possible que quel que soit l’environnement dans lequel la plante pousse, quelques microbes soit si importants pour sa survie et son développement qu’on les retrouve toujours en association avec ladite plante. Ces microbes toujours en association avec une plante donnée constituent une unité théorique nommée core microbiote dans la littérature scientifique. La gestion du microbiote des plantes cultivées pourrait améliorer la résistance au stress et la productivité des plantes cultivées et il est donc important d’en comprendre le fonctionnement. A ce jour, le microbiote souterrain des plantes demeure largement une « boîte noire » en raison de son incroyable complexité due à la diversité faramineuse des microorganismes qui le constituent. Au cours de ma recherche doctorale, j’ai voulu participer à ouvrir encore un peu plus cette « boite noire » pour augmenter la connaissance du fonctionnement et de la structure du microbiote souterrain des plantes. Pour ce faire, j’ai utilisé le canola (B. napus) comme plante modèle. J’ai étudié le microbiote racinaire, tel qu’influencé par le niveau de diversification du système cultural, à l’aide d’un dispositif expérimental établi par Agriculture et Agroalimentaire Canada à cinq emplacements dans la prairie canadienne en 2008. Le canola, B. napus est une Brassicaceae économiquement importante, mais aussi intéressante en tant que plante modèle, car le canola est associé à des communautés microbiennes racinaire moins complexes que bien d’autres plantes, à cause de sa production de composés antimicrobiens. J’ai utilisé le séquençage d’amplicons, des analyses statistiques multivariées et l’analyse de réseau pour approcher cette complexité et: i) vérifier l’impact de la diversification du système de rotation cultural sur les communautés microbiennes souterraines du canola, ii) établir si un core microbiote fongique et bactérien existait bel et bien dans la rhizosphère du canola et le plein sol en culture de canola, iii) identifier de façon claire des espèces clef de voute interagissant intensivement dans les communautés fongiques, bactériennes, et mixtes, et finalement iv) évaluer la persistance des champignons mycorhiziens à arbuscules dans la rhizosphère du canola et le plein sol adjacent cette plante non-hôte, en systèmes culturaux basés sur le canola. Mes résultats confirment que les communautés fongiques de la rhizosphère du canola et de son sol étaient influencées par la diversification des rotations de cultures, mais démontrent que les communautés bactériennes ne l’étaient pas. La rhizosphère du canola avait un core microbiote fongique variant avec les années, tandis que chez les bactéries, seulement des core espèces ont été identifiées. J’ai aussi relevé des interactions potentielles entre microbiote fongique et microbiote bactérien du canola et identifié des espèces clef de voute. Les fluctuations de l’abondance de ces espèces pourraient alors faire varier celles de beaucoup d’autres microbes. Bradyrhizobium a été l’une de ces espèces. Mes résultats montrent aussi un maintien d’une communauté des champignons mycorhiziens à arbuscules chez le canola même après 10 ans de monoculture. En résumé, ma recherche apporte une lumière nouvelle dans l’étude du fonctionnement, de la structure et des dynamiques écologiques au sein du microbiote souterrain du canola et sur l’écologie microbienne théorique des plantes notamment en ce qui a trait à ses composantes invariantes telles que le core microbiote et les taxons clef de voûte. Des études en conditions contrôlées sont nécessaires pour vérifier la capacité des microbes clef de voute rapportés ici à influencer les communautés microbiennes du sol et les plantes qui y vivent. / Plants and soil microbes are closely linked. Plants provides myriads of ecological niches in and on its roots for microbes to thrive. In turn, microbes can protect host plants against pathogen attacks, abiotic stresses, and improve nutrient and water availability. In the distant soil, plant produce volatile compounds shaping microbial communities, with feedback on root-associated communities. The more diversity there is in the plant cover, the higher the diversity of soil microorganisms will be and conversely, the more diverse a soil will be in terms of microbes, the more de plants that grow there trend to be in good health. Certain plants can develop specific relationships with certain microbes improving the fitness of the plant. However, a plant can grow in different environments and soils are diverse, thus plant will have to adapt to the different microbes depending on the environment it is growing in while attracting the ones necessary for its growth. Certain microbes could be so important for a plant’s health and development that they are always associated with the plant. Such important microbes form a theoretical group called core microbiota that could be extremely important for plant health and a determinant of the composition of plant-associated microbial communities. The plant subterranean microbiota is often labelled as a “black box” due to the tremendous diversity and interactivity of the microbial communities plants host. In my thesis research I aimed to “crack the black box” a little further to enhance our understanding of plant subterranean microbial community dynamics and structure. To do so, I used a field experiment established in 2008 by Agriculture and Agri-Food Canada (AAFC) at five different sites in the Canadian Prairies under different crop rotations and canola as model plant. Canola (B. napus) is a crop plants of the Brassicaceae family that produces antimicrobial compounds and has “simpler” microbial community in its roots, and rhizosphere. To do so, I used amplicon sequencing, multivariate analysis, and network analysis. My objectives were i) to verify the impacts of plant cover diversification on canola microbial subterranean community, ii) to verify if a core microbiota of fungi and bacteria could exist in canola rhizosphere and bulk soil and if so, to describe this core, iii) to identify keystone bacteria and fungi, i.e. highly interacting components, in the bacterial and fungal communities associated with canola, and finally, iv) to investigate the persistence of arbuscular mycorrhizal fungi in the rhizosphere and bulk soil of canola, a non-host plant, in canola-based cropping systems. I found that the diversification of cropping systems influenced the structure of the fungal communities of canola rhizosphere and bulk soil, but diversification had no significant influence on bacterial community structure. A fungal core microbiota varying through years was found in canola rhizosphere, but no bacterial core-microbiota was found. However, we were able to identify a core-specie. Interactions among the fungal and bacterial microbiota in canola rhizosphere and bulk soil were found and Bradyrhizobium was among several potentially important keystone taxa. My results also show the maintenance of arbuscular mycorrhizal fungi in canola even after 10 years of monoculture despite this plant is not a host for AMF. Overall, my PhD research brings a new level of knowledge on the microbial structure and dynamics of canola subterranean microbiota, and also on the theoretical ecology of plant microbiota, particularly regarding its invariable components such as core microbiota and hub-taxa. Further investigations are needed to better understand how keystone species and core species influence the plants and their microbiome.

<p dir="ltr">Regulation of complex biological processes aims to achieve goals essential for an organism's survival or to exhibit specific phenotypes in response to stimuli. This regulation can occur at several levels, such as cellular metabolism, signaling pathways, gene transcription, mRNA translation into proteins, and post-translational modifications. Systems biology approaches can facilitate integrating mechanistic knowledge and high-throughput omics data to develop quantitative models that can help improve our understanding of regulations at various levels. However, computational modeling of biological processes is challenging due to the vast details of various processes with unknown mechanisms. The cybernetic modeling approach accounts for unknown control mechanisms by defining a biological goal that the system aims to optimize and subsequently mathematically formulates the cybernetic goal.</p><p dir="ltr">This thesis aims to develop a mathematical framework that integrates a cybernetic model with novel information-theoretic methods to study the inflammatory response in mammalian macrophage cells. The inflammatory response of the body is a protective mechanism that fights off infecting pathogens by inducing the production of immune signaling proteins called cytokines and chemokines, as well as specific lipids known as eicosanoids. However, excessive levels of cytokines and eicosanoids may result in chronic inflammatory diseases such as hyper-inflammation syndrome, COVID-19, and asthma. Only a few studies have focused on quantitative modeling of the role of lipid metabolism in inflammation. One key lipid is Arachidonic acid (AA), which during inflammation, gets converted into inflammatory lipids called eicosanoids. Previous models utilize Michaelis-Menten kinetics or assume the linear form and can, at best, include control at the gene expression level only. The distinguishing feature of a cybernetic model is that by defining a cybernetic objective, it can account for control at multiple levels, including transcriptional, translational, and post-translational modifications.</p><p dir="ltr">The following paragraphs address a specific research problem, outline the approaches to investigate it, and summarize the key findings.</p><p dir="ltr">First, we studied the cellular response to inflammatory stimuli that produce eicosanoids—prostanoids (PRs) and leukotrienes (LTs)—and signaling molecules—cytokines and chemokines—by macrophages. A few studies suggest that targeting eicosanoid metabolism could be a promising new approach to regulating cytokine storm in COVID-19 infection. We developed a cybernetic model combined with novel information-theoretic approaches to study the integrated system of eicosanoids and cytokines. Our cybernetic model formulates a cybernetic goal, which requires the causal relationship between the eicosanoid and cytokine secretion processes; however, this causal relationship is unknown due to insufficient mechanistic information. We developed novel information-theoretic approaches (discussed later in detail) to understand the causality between eicosanoids and cytokines. The causality result from information theory suggests that Arachidonic acid (AA) may be the cause for initiating the secretion of cytokine TNF. The model captured the data for all experimental conditions, including control, treatment with Adenosine triphosphate (ATP), (3-deoxy-d-manno-octulosonic acid) 2-lipid A (Kdo2-Lipid A, abbreviated as KLA), and a combined treatment of ATP and KLA in mouse bone marrow-derived macrophages (BMDM). The model explains the dynamics of metabolites for all experimental conditions, validating the hypothesis. It also enhanced our understanding of enzyme dynamics by predicting their profiles. The results indicated that the dominant metabolites are PGD2 (a PR) and LTB4 (an LT), aligning with their corresponding known prominent biological roles during inflammation. Based on the causality and cybernetic model result and using heuristic arguments, we also infer that AA overproduction can lead to increased secretion of cytokines/chemokines. Consequently, a potential clinical implication of this study is that modulating eicosanoid levels could lower TNFα expression, suggesting eicosanoids could be a viable strategy for managing hyperinflammation.</p><p dir="ltr">Second, we studied the dynamics of the anti-inflammatory lipid mediators from eicosapentaenoic acid (EPA) metabolism, which can be beneficial in reducing the severity of diseases such as cancer and cardiovascular effects and promoting visual and neurological development. This study employed a cybernetic model to study the enzyme competition between AA and EPA metabolism in murine macrophages. The cybernetic model adequately captured the experimental data for control non-supplemented and EPA-supplemented conditions in RAW 264.7 macrophages. The cybernetic variables provide insights into the competition between AA and EPA for the COX enzyme. Predictions from our model suggest that the system undergoes a switch from a predominantly pro-inflammatory state in control to an anti-inflammatory state with EPA supplementation. A potential application of this study is utilizing the model estimation of the ratio of concentrations required for the switch to occur as 2.2, which aligns with the experimental observations and falls within the recommended range of 1-5 needed to promote anti-inflammatory response.</p><p dir="ltr">Third, we focused on predicting novel causal connections between AA and cytokines using time series analysis as mechanistic information connecting AA and cytokines is unknown. In this work, we developed Time delay Renyi Symbolic Transfer Entropy (TDRSTE), a novel model-free information-theoretic metric. We computed it from high-throughput omics datasets for bivariate non-stationary time series to quantify causal time delays. The TDRSTE method adequately estimated time delay for the synthetic dataset, captured causality for the real-world biological dataset of the AA metabolic network with a prediction accuracy of 80.6%, where it correctly identified 25 out of 31 connections, and detected novel connections between non-stationary lipidomics and transcriptomics profiles for eicosanoids and cytokines, respectively. The results indicate that AA may initiate the secretion of cytokines like TNFα, IL1α, IL18, and IL10. Conversely, cytokines such as IL6 and IL1β may have an early causal impact on AA. These findings suggest a potential causal link between AA and cytokines, paving the way for further exploration with more extensive experimental data in future investigations.</p><p dir="ltr">This thesis develops a theoretical framework that integrates the cybernetic modeling technique with novel information-theoretic approaches to study the inflammatory response in mouse macrophages. As described in previous paragraphs, the success of the cybernetic framework in capturing the dynamic behavior of multiple processes serves to validate the idea that regulation is driven toward achieving cellular goals. The cybernetic framework can be applied to better understand the mechanisms underlying the normal and diseased states and to predict the behavior of the system given a perturbation.</p>

<p dir="ltr">Open-access bioinformatics data enables accessible public health research for a variety of stakeholders, including teachers and low-resourced researchers. This project outlines two case studies utilizing open-access bioinformatics data sets and analysis software as proofs of concept for the types of research projects that can be adapted for workforce development purposes. The first case study is a spatial temporal analysis of Lyme disease rates in the United States from 2008 to 2020 using freely available data from the United States Department of Agriculture and Centers for Disease Control and Prevention to determine how urbanization and other changes in land use have impacted Lyme disease rates over time. The second case study conducts a pangenome analysis using bacteriophage data from the Actinobacteriophage Database to determine conserved gene regions related to host specificity.</p>

Next generation sequencing is a modern method applied in plant virology for sensitive detection of previously characterized and novel pathogens without any preceding knowledge of them. In this study three novel and two already described viruses were detected by de novo assembly of Illumina single-end reads ( Hi-Seq 2500 system) from total poly(A) enriched RNA of diseased red clover (Trifolium pratense) and indicator plant (Nicotiana occidentalis 37B). The complete genomic sequence of novel Red clover carlavirus A (RCCA) was determined from Illumina reads, 5´, 3´ RACE, cloning, RT-PCR and Sanger sequencing. The presence of RCCV was also confirmed in mechanically inoculated tobacco plant.