Systematic characterization of microbes in several tumors including colorectal cancer (CRC) and pancreatic ductal adenocarcinoma (PDAC) has revealed the presence of multiple species of intracellular bacteria within tumors. However, there is limited knowledge on how these bacteria colonize tumors, how they survive inside host cells, how they modulate host cell phenotypes, and if their elimination should complement cancer therapy. This is, in part, due to the lack of representative animal models, challenges in co-culture of host epithelial cells and bacteria, and limited resolution of available analytical techniques to study host-microbial interactions. I have addressed these challenges by harnessing multiple technologies from microbiology, genetic engineering, tissue engineering, and microfluidics, in order to investigate the role of an emerging oncomicrobe, Fusobacterium nucleatum, in the tumor microenvironment (TME). F. nucleatum is a Gram-negative, anaerobic bacterium that is normally found within the oral cavity. However, its selective enrichment in CRC and PDAC tumors is correlated with poor clinical outcomes. My work along with collaborators in the Verbridge, Slade, and Lu labs at Virginia Tech has revealed a multifactorial impact of F. nucleatum in influencing cancer progression. First, in CRC, we discovered that F. nucleatum infection of host cancer cells induced robust secretion of select cytokines that increased cancer cell migration, impacted cell seeding, and enhanced immune cell recruitment. In PDAC, we uncovered additional cytokines that were secreted from both normal and cancerous pancreatic cell lines upon infection with F. nucleatum that increased cancer cell proliferation and migration via paracrine and autocrine signaling, notably in the absence of immune cell participation. In order to examine the contribution of a hypoxic TME on infection dynamics, we used a multi-omics approach that combined RNA-seq and ChIP-seq of H3K27ac to determine epigenomic and transcriptomic alterations sustained within hypoxic CRC cells upon infection with F. nucleatum. Our findings revealed that F. nucleatum can subvert host cell recognition in hypoxia and can modulate the expression of multiple cancer-related genes to drive malignant transformation. Insights gained from this research will pave the way for future studies on the impact of the tumor microbiome in cancer and will identify novel targets for therapy and clinical intervention to control bacteria-induced exacerbation of cancer. / Doctor of Philosophy / Colorectal cancer (CRC) and pancreatic ductal adenocarcinoma (PDAC) are the second and third leading causes of cancer death in the United States, respectively. Recent systematic characterization of various tumor types revealed the presence of distinct bacteria within tumors. However, there is limited knowledge on how these bacteria colonize tumors, how they survive inside host cells, how they modulate host cell phenotypes, and if their elimination should complement cancer therapy. This is, in part, due to the lack of representative animal models, challenges in developing host cell-microbe co-culture models, and limited resolution of available analytical techniques to study host-microbial interactions. I have addressed these challenges by harnessing multiple technologies from microbiology, genetic engineering, tissue engineering, and microfluidics, in order to investigate the role of an emerging cancer-associated microbe, Fusobacterium nucleatum, in the tumor microenvironment (TME). F. nucleatum is a microbe commonly found within the oral cavity. However, clinical studies revealed that selective enrichment of F. nucleatum in CRC and PDAC tumors significantly correlated with poor prognosis. My work along with collaborators in the Verbridge, Slade, and Lu labs at Virginia Tech has revealed a multifactorial impact of F. nucleatum in influencing cancer progression. First, in CRC, we discovered that F. nucleatum invasion of host cancer cells induced the secretion of select proteins called cytokines that cells use to signal and communicate with each other. These cytokines directly stimulated the cell migration of host cancer cells which is usually associated with increased cancer aggressiveness. In PDAC, F. nucleatum infection induced the secretion of additional cytokines from both cancer cells and normal cells that, in addition to cell migration, impacted the proliferation of cancer cells, another feature of aggressive cancers. F. nucleatum usually thrives in a low oxygen environment that is prevalent in cancer tissue and hence, we examined how a low oxygen environment can influence infection dynamics using sequencing technologies that probe the genomic constitution within cells. Our findings revealed that F. nucleatum can escape recognition in low oxygen environments and can modulate the expression of multiple cancer-related programs within the cell to drive cancer progression. Insights gained from this research will pave the way for future studies on the impact of the tumor-associated microbes in cancer and will identify novel targets for therapy and clinical intervention to control bacteria-induced exacerbation of cancer.

Listeria monocytogenes est un pathogène intracellulaire facultatif responsable d’une infection sévère d’origine alimentaire, la listériose. L’étude du processus d’infection cellulaire de cette bactérie a permis d’élucider divers mécanismes impliqués dans les interactions hôte-pathogène et dans le fonctionnement de la cellule eucaryote. En particulier, L. monocytogenes a été l’un des modèles pionniers dans la découverte du ciblage de la chromatine et de régulateurs nucléaires par des microbes. L’étude d’un facteur de virulence de L. monocytogenes, LntA, a permis l’identification d’un de ces régulateurs : BAHD1. En recrutant des protéines impliquées dans la formation de l’hétérochromatine, telles HDAC1/2 et HP1, BAHD1 stimule la formation d’une chromatine compacte à effet répressif. Lors d’une infection de cellules épithéliales par L. monocytogenes, BAHD1 réprime la réponse immunitaire stimulée par les interférons, une fonction inhibée par LntA. BAHD1 demeurant peu étudiée, mon doctorat a eu pour premier objectif de poursuivre la caractérisation de ce régulateur épigénétique. Par ailleurs, des données préliminaires suggéraient qu’un facteur de virulence de Listeria récemment découvert, InlP, avait la potentialité d’être, comme LntA, une nucléomoduline. Mon deuxième objectif a été d’explorer cette hypothèse.Les résultats de mon premier axe montrent que BAHD1 interagit avec MIER1 et que cette interaction est cruciale pour l’association de BAHD1 aux HDAC1/2. Nous reportons également que BAHD1 modifie la chromatine en changement la méthylation et l’acétylation des histones, ainsi que la méthylation de l’ADN, au niveau d’un gène cible, ESR1. Ces résultats nous permettent de proposer que BAHD1 forme, avec MIER1, un échafaudage assemblant un nouveau complexe de remodelage de la chromatine associé aux HDAC1/2 : le complexe BAHD1. Nous avons ensuite étudié un rôle de BAHD1 dans un organe ciblé par la Listeria, le cerveau. Nos résultats indiquent qu’une déficience totale en BAHD1 altère le transcriptome global de cet organe chez la souris. Les gènes majoritairement surexprimés sont impliqués dans des fonctions du système nerveux, le métabolisme et des troubles neurologiques. Les gènes majoritairement sous-exprimés sont impliqués dans des voies de l’immunité innée, dont des gènes de réponses aux interférons. Par ailleurs, une haplo-déficience en Bahd1 provoque des troubles comportementaux. En comparaison des souris Bahd1+/+, les souris Bahd1+/- souffrent d’une anxiété accrue et d’altérations du réflex de sursaut acoustique. Ces résultats suggèrent qu’une dérégulation de BAHD1, par des stimuli de l’environnement ou par des stimuli infectieux, pourrait avoir des effets neuro-pathologiques.Le second axe de ma thèse concernait l’étude des interactions d’InlP avec des protéines nucléaires de l’hôte, identifiées par un crible double-hybride. Nous montrons d’abord qu’InlP est une internaline atypique, avec des répétitions riches en leucine caractérisées par un motif LPX2. Nous identifions, ensuite, deux protéines nucléaires ciblées par InlP : le facteur d’épissage et suppresseur de tumeur RBM5 et le corépresseur RERE. Quand InlP est produite de façon ectopique dans les cellules humaines, elle se localise dans le noyau, où elle altère la formation de corps nucléaires enrichis en RERE. Dans des cellules sur-exprimant RBM5, InlP inhibe l’effet pro-apoptotique de RBM5 et stimule la formation de corps nucléaires denses associés à RBM5. Ces résultats suggèrent qu’InlP est une nucléomoduline agissant sur la l’assemblage et le désassemblage de compartiments de stockage de protéines cibles impliquées dans la synthèse et l’épissage d’ARNs de l’hôte.Ce travail ouvrent des perspectives dans la compréhension des interactions hôte-pathogène et dans une meilleure connaissance des mécanismes patho-épigénétiques, ainsi qu’en biologie cellulaire, dans la compréhension de la dynamique des organites nucléaires sans membrane. / Listeria monocytogenes is an optional intracellular pathogen responsible for a severe foodborne infection called listeriosis. The study of the cellular infection process of this bacterium has shed light on various mechanisms involved in host-pathogen interactions and in the functioning of the eukaryotic cell. In particular, L. monocytogenes has emerged as one of the pioneering models in the discovery of microbial targeting of chromatin and nuclear regulators. The study of a virulence factor of L. monocytogenes, LntA, allowed the identification of one of these regulators : BAHD1. By recruiting proteins involved in the formation of heterochromatin, such as HDAC1/2 and HP1, BAHD1 stimulates the formation of a compact chromatin with a repressive effect. When epithelial cells are infected with L. monocytogenes, BAHD1 suppresses the immune response stimulated by interferons, a function inhibited by LntA. Since BAHD1 is still under-researched, the first objective for my thesis was to further characterize this epigenetic regulator. In addition, preliminary data suggested that a recently discovered virulence factor of Listeria, InlP, had the potential to be, like LntA, a nucleomodulin. My second objective was to explore this hypothesis.The results of my first axis show that BAHD1 interacts with MIER1 and that this interaction is crucial for the association of BAHD1 with HDAC1/2. We also report that BAHD1 modifies chromatin by changing histone methylation and acetylation, as well as DNA methylation, at a target gene, ESR1. These results allow us to propose that BAHD1 form, with MIER1, a scaffold assembling a new chromatin remodeling complex associated with HDAC1/2 : the BAHD1 complex. We then studied the role of BAHD1 in an organ targeted by Listeria, the brain. Our results indicate that a total deficiency in BAHD1 alters the overall transcriptome of this organ in mice. Most of the overexpressed genes are involved in nervous system functions, metabolism and neurological disorders. The predominantly downregulated genes are involved in innate immunity pathways, including interferon response genes. In addition, a haplodeficiency in Bahd1 causes behavioral problems. Compared to Bahd1+/+ mice, Bahd1+/- mice suffer from increased anxiety and changes in acoustic startle reflex. These results suggest that deregulation of BAHD1, through environmental or infectious stimuli, may have neuro-pathological effects.The second axis of my thesis focused on the study of InlP interactions with host nuclear proteins, identified by a double-hybrid screen. First, we show that InlP is an atypical internalin, with leucine-rich repeats characterized by an LPX2 motif. We then identify two nuclear proteins targeted by InlP: the splicing factor and tumor suppressor RBM5 and the corepressor RERE. When InlP is produced ectopically in human cells, it is localized in the nucleus, where it alters the formation of nuclear bodies enriched in RERE. In RBM5-overexpressing cells, InlP inhibits the pro-apoptotic effect of RBM5 and stimulates the formation of dense nuclear bodies associated with RBM5. These results suggest that InlP is a nucleomodulin acting on the assembly and disassembly of target protein storage compartments involved in the synthesis and splicing of host RNAs.This work opens perspectives in the understanding of host-pathogen interactions and in a better knowledge of patho-epigenetic mechanisms, as well as in cell biology and the understanding of membraneless nuclear organelles dynamics.