Regulatory T cells (Treg) play a crucial role in controlling immune homeostasis. Several inflammatory diseases including multiple sclerosis and inflammatory bowel disease have been associated with dysfunctional and/or reduced numbers of Treg. While several mechanisms of action have been discovered by which Treg can exert their function, disease-specific Treg requirements remain unknown. The Treg pool consists of highly diverse subpopulations, indicating that there is a potential to optimise Treg-targeted therapies if disease-relevant mechanisms can be established. Microarray data from our lab suggests a marked upregulation of the integrin αv as well as the IL-33 receptor ST2 in Treg retrieved from the inflamed central nervous system (CNS) during experimental autoimmune encephalomyelitis compared to peripheral lymphoid organs. These two molecules were further investigated within this PhD project with the aim to understand their role in Treg function during chronic inflammatory disease. αvβ integrins have been reported to be needed for effector T cell migration to inflamed sites through binding of extracellular matrix components and are involved in TGF-β activation by a variety of cell types. Conditional knockout mice lacking the integrin αv specifically in Foxpγ+ Treg were generated to address the role of αv integrins on regulatory T cells in inflammatory disease. αv-/- Treg showed a deficiency in activating latent TGF-β, but were able to suppress responder T cell proliferation in vitro as well as in vivo. αv-/- Treg were also able to migrate to the inflamed CNS during EAE and resolve disease. However, αv-/- Treg were detected at significantly lower numbers and proportions in the inflamed gut during a curative T cell transfer model of colitis; this led to a quantitative impairment in the ability of αv-/- Treg to cure colitis when compared to wild-type (WT) Treg. Whether this is a deficit in migration, survival, proliferation, or Foxp3 stability, remains to be investigated. IL-33 acts as an alarmin and is best studied as a cytokine released upon tissue damage that induces a potent type 2 immune response by acting on a multitude of immune cells. Expression of the IL-33 receptor ST2 on Treg has recently been associated with positive metabolic parameters in visceral adipose tissue, protection from gut inflammation, and tissue-restorative function in other inflamed tissues such as injured muscle or lung. This project showed that in steady state, ST2+ Treg expressed high levels of several markers which have been associated with potent regulatory function. When stimulated in vitro, ST2+ Treg showed a better survival and expansion rate compared to their ST2- counterparts, even more so in the presence of IL-33. T-bet deficiency in Treg resulted in an increased ST2+ Treg pool, and T-bet-associated cytokine IFN-γ was found to antagonise IL-33-induced expansion of the ST2+ Treg pool in a T-bet-independent manner. When ST2+ and ST2- Treg were tested for their respective suppressive capacity in vivo, ST2+ Treg were able to suppress responder T cell expansion despite being found only at low numbers in secondary lymphoid organs compared to ST2- Treg. However, in a curative model of T cell transfer colitis, ST2+ Treg were less capable of controlling the ongoing immune response than ST2- Treg. A possible explanation for the superiority of ST2- Treg in this setting can be found in the fact that injected ST2- Treg acquired a distribution of ST2 expression reminiscent of WT Treg over the course of disease. On the other hand, an increased starting pool of ST2+ Treg as occurs in T-bet-/- Treg significantly enhanced the capacity of Treg to control colitis compared to WT Treg. In conclusion, both ST2- and ST2+ Treg are likely to have a distinct, non-redundant role in suppressing T cell activation in secondary lymphoid organs and controlling ongoing inflammation in peripheral tissue, respectively.

Stroke is the 5th leading cause of death and the leading cause of disability in the United States, but there are only two available therapies, tissue plasminogen activator and endovascular thrombectomy. As both therapies focus on removal of the clot, the subsequent pathologic processes, i.e. inflammation, cerebrovascular breakdown, ATP depletion, etc. are left untreated, contributing to worsened patient outcome. Many clinical trials have unsuccessfully attempted to address these mechanisms. The blood-brain barrier (BBB), a system of non-fenestrated endothelial cells, extracellular matrix, and astrocytic endfeet, is significantly impacted after ischemic stroke in its role of preventing the free movement of proteins from the blood into the brain. In fact, BBB dysfunction is viewed as one of the major facilitators of damage following ischemic stroke, leading to increased infarct volumes and worsened patient outcomes. Interestingly, a family of endothelial integrins, the b1 integrins, have been shown to regulate tight junction proteins preventing the free movement of molecules. When expression of the tight junctions are decreased, this results in increased BBB permeability. To test this concept, our laboratory has previously shown the knockout of the particular β1 integrin, α5β1, is neuroprotective following ischemic stroke through BBB stabilization. To determine if therapeutically targeting integrin a5b1 was feasible, we first determined if brain integrin a5b1 expression increases after experimental mouse ischemic stroke model, specifically tandem/transient common carotid artery/middle cerebral artery occlusion. We found that integrin a5b1 does increase acutely, by post-stroke day (PSD)2, and continued in an exponential fashion through PSD4. Next, we determined if integrin a5b1 was therapeutically accessible by systemic treatment (i.e. intraperitoneal or intravenous) by being located on the inside (luminal surface) of vasculature. We found that location of integrin a5b1 was dependent on the area relative to the stroke injury. The core, or area of direct impact, demonstrated expression of integrin a5b1 on the outside vasculature (abluminal surface), while per-infarct expression was localized to the lumen. Lastly, to determine the activity of integrin a5b1 following ischemic stroke, we showed that the potential ligands (binding partners), plasma fibronectin, fibrinogen, and amyloid-b, do not bind integrin a5b1 after ischemic stroke. Next, we determined the therapeutic potential of targeting integrin a5b1 with the small peptide, ATN-161. ATN-161 has undergone clinical trials in solid tumors, with limited side effects reported. First, we determined that intraperitoneal (IP) injection of ATN-161 was safe after ischemic stroke, showing no changes in heart rate, pulse distention (blood pressure), or body temperature. Next, we found that IP administration of ATN-161 after experimental ischemic stroke reduced infarct volumes, edema, and functional deficit. Furthermore, these results were due to reduction of BBB permeability and anti-inflammatory effects. Interestingly, ATN-161 reduced cytokine production, prevented leukocyte infiltration, and leukocyte recruitment. Collectively, these results suggest that targeting integrin a5b1 with ATN-161 is 1) feasible, 2) safe and 3) effective, suggesting that ATN-161 may be a novel therapeutic treatment for ischemic stroke.

Le glioblastome (GBM) est la tumeur cérébrale la plus agressive. L’échec thérapeutique s’explique par son caractère hétérogène. Nous étudions l’intégrine α5β1, le récepteur de la fibronectine, dont les travaux du laboratoire ont montré l'implication dans l’agressivité du GBM. Il a été observé que l’expression de l’intégrine α5 est hétérogène entre patients et au sein d’une même tumeur. Nous avons émis l’hypothèse que cette hétérogénéité peut être expliquée par l’existence de différentes cellules souches de GBM (CSGs). Nos travaux montrent que certaines CSGs expriment l’intégrine après différentiation, démontrant une hétérogénéité intertumorale. Une expression hétérogène a également été mise en évidence au sein d’une même lignée, indiquant une hétérogénéité intratumorale. Nos résultats démontrent que les CSGs programmées à exprimer l’intégrine forment des cellules différenciées agressives. Nos données montrent que l’intégrine α5 est impliquée dans l’expansion tumorale mais aussi dans la survie et la migration en présence de fibronectine. Ce travail permet de mieux comprendre l’origine de foyers plus agressifs et souligne l’intérêt de l’intégrine α5β1 comme cible thérapeutique pour une sous-population de patients. / Glioblastoma (GBM) is the most aggressive primary brain tumor. Treatment failure is explained by inter and intratumoral heterogeneity. Our previous results showed that the α5β1 integrin, the fibronectin receptor, is implicated in GBM aggressiveness. We observed that its expression was heterogeneous between patients' tumors but also between different areas in a given tumor. We hypothesized that this heterogeneity may be linked to different glioma stem cells (GSC). We confirmed that α5 expression is induced after differentiation in about half of the cell lines supporting the notion of inter-tumoral heterogeneity. We also observed with single cell-derived clone analysis that intra-tumoral GICs heterogeneity exists. We noticed that when GICs are programmed to express α5 integrin, differentiated cells become more aggressive. Our data support that differentiated cells expressing the integrin have an increased capacity of proliferation and acquire a fibronectin-dependent motility and a resistance phenotype. This work provides a better understanding of the origin of more aggressive foci and highlights the interest of α5β1 integrin as a therapeutic target for a subpopulation of patients.

acase@tulane.edu

Recombinant adeno-associated virus (rAAV) is a widely used gene therapy vector. Although a wide range of rAAV serotypes can effectively enter most cell types, their transduction efficiencies (i.e., transgene expression) can vary widely depending on the target cell type. Integrins play important roles as co-receptors for rAAV infection, however, it remains unclear how integrin-dependent and -independent mechanisms of rAAV endocytosis influence the efficiency of intracellular virus processing and ultimately transgene expression. In this thesis, I examined the contribution of integrin-mediated endocytosis to transduction of fibroblasts by rAAV2. I found that promoting AAV2/integrin binding with Mn++ greatly enhanced (~17-fold) rAAV2 transduction independently of cell binding and endocytosis. Subcellular localization studies of rAAV2 demonstrated that integrin activation by Mn++ promoted AAV2 aggregation on alpha5 and beta1 integrins and recruitment of the cytosolic integrin effector protein vinculin. Focal adhesion kinase (FAK), a down stream effector of integrin signals, was essential for AAV/integrin complex endocytosis and transduction, but not AAV2 recruitment to integrins. Recruitment of FAK to AAV2/integrin complexes was increased by transiently trapping the endocytic event at the plasma membrane by pharmacologic inhibition of dynasore. This also increased the size of AAV2 clusters found beneath the cell at FAK/integrin complexes resembling immature filopodia and caused a large, FAK-dependent (75-fold) increase in AAV2 transduction. These findings support a model whereby integrin activation at the cell surface can redirect rAAV2 toward a FAK-dependent entry pathway that is more productive for cellular transduction. This pathway appears to be conserved for other rAAV serotypes that contain a capsid integrin-binding domain (AAV1 and 6).

This dissertation is devoted to the study of the molecular biology of major tumor suppressors, defined as those that prevent the cellular processes identified as the hallmarks of cancer. Specifically, the major tumor suppressors pRb and STK11 are explored in the context of osteosarcoma and lung cancer, respectively. RB1 was the first tumor suppressor gene discovered. Over four decades of work have revealed that the Rb protein (pRb) is a master regulator of biological pathways influencing virtually every aspect of intrinsic cell fate including cell growth, cell-cycle checkpoints, differentiation, senescence, self-renewal, replication, genomic stability and apoptosis. While these many processes may account for a significant portion of RB1's potency as a tumor suppressor, a small, but growing stream of evidence suggests that RB1 also significantly influences how a cell interacts with its environment, including cell-to-cell and cell-to-extracellular matrix interactions. Chapter 2 highlights pRb's role in the control of cell adhesion and how alterations in the adhesive properties of tumor cells may drive the deadly process of metastasis. Chapter 3 defines a role for pRb as a suppressor of the progression to metastasis by upregulating integrin α10. Transcription of this integrin subunit is herein found to be pRb-dependent in mouse osteoblasts. Classic pRb partners in cell cycle control, E2F1 and E2F3, do not repress transcription of integrin α10 and phosphorylation of pRb is not necessary for activation of the integrin α10 promoter. Promoter deletion revealed a pRb responsive region between -108bp to -55bp upstream of the start of the site of transcription. pRb activation of transcription also leads to increased levels of integrin α10 protein and a greater concentration of the integrin α10 protein at the cell membrane of mouse osteoblasts. These higher levels of integrin α10 correspond to increased binding to collagen substrate. Consistent with our findings in mouse osteoblasts, we found that integrin α10 is significantly underexpressed in multiple solid tumors that have frequent inactivation of the pRb pathway. Bioinformatically, we identified data consistent with an 'integrin switch' that occurs in multiple solid tumors consisting of underexpression of integrins α7, α8, and α10 with concurrent overexpression of integrin β4. pRb promotes cell adhesion by inducing expression of integrins necessary for cell adhesion to a substrate. We propose that pRb loss in solid tumors exacerbates aggressiveness by debilitating cellular adhesion, which in turn facilitates tumor cell detachment and metastasis. Lung cancer is the leading cause of cancer-related death in the U.S. and additional targeted therapies are desperately needed to treat these patients. STK11 is the third most frequently mutated gene in lung adenocarcinoma following only KRAS and TP53, yet its mutational status is not currently clinically evaluated and no therapies have been approved to specifically target its pathway. A deep understanding of the complex pathways controlled by STK11 and their alterations in cancer are required to develop effective therapies for patients with loss-of-function mutations. In Chapter 4 we present the current understanding of STK11, focusing on its molecular biology and therapeutic implications, including a compilation of studies evaluating STK11 somatic mutations in human lung cancer tissue and how the frequency of these mutations varies across histological subtypes and patient populations. Finally, we review the strategies being used to target STK11-deficient cancers at the clinical trial, pre-clinical, and basic science levels as well as proposing potential new therapies that might benefit this patient population. STK11 is a tumor-suppressor commonly mutated in lung adenocarcinoma (LuAd). There are a number of agents that may selectively target the deregulated pathways in STK11 mutated tumors, and thus, identifying the subset of adenocarcinomas that harbor these mutations could have significant clinical benefit. In Chapter 5, we characterized a cohort of 442 adenocarcinoma patients with respect to STK11 mutation status and subset of this cohort using immunochemistry, gene expression, and western blotting. We found that measuring STK11 mutation status is complicated by the fact that many STK11 mutations lead to expression of a stable protein that is indistinguishable from wild type (WT) via immunohistochemistry. To circumvent this, we used published cell line mutation and gene expression data to derive a signature correlating with STK11 mutation status. This signature was validated in the cohort of 442 lung adenocarcinomas and strongly correlates with mutation status (ROC curve AUC = 85.29). These data suggest that STK11; mutation status may be best assessed by measuring the downstream targets included in our signature.

<p>Integrins are heterodimeric cell surface receptors, composed of an α and a β subunit, mainly binding for extracellular matrix proteins. lntegrin subunit β1 can combine with at least 12 a subunits and thus form the biggest subfamily within the integrin family. In this thesis, functional properties of the splice variant β1Β, and the effects of several mutations in the cytoplasmic tail of integrin subunit β1Α were studied. In addition, the border between the transmembrane and cytoplasmic domains of several integrin subunits was determined.</p><p>The β1Β splice variant has been reported to have a dominant negative effect on functions of β1Α integrins. In this study, it was studied if the expression of β1Β had similar negative effects on the αvβ3 integrin functions since the β3 subunit is structurally similar to β1Α. The β1Β subunit was expressed in an integrin β1-deficient cell line and it was found that the presence of β1Β does not interfere with adhesion or signaling of endogenous αvβ3</p><p>The border between the cytoplasmic domain and the C-terminal end of the transmembrane domain of integrin α and β subunits has been unclear. This question was experimentally addressed for integrin subunits β1, β2, α2 and α5. It was found that integrin subunits contain a positively charged lysine, which is embedded in the membrane in the absence of interacting proteins.</p><p>The functional importance of the lysine in integrin transmembrane domains was investigated by mutating this amino acid to leucine in β1Α. The mutation affected cell spreading and tyrosine phosphorylation of the adapter protein CAS. The activation of focal adhesion kinase and tyrosine phosphorylation of paxillin was not affected. Furthermore, the mutation of two tyrosines to phenylalanines in the β1Α cytoplasmic tail was found to reduce the capability of β1Α integrins to mediate cell spreading and migration. Activation of focal adhesion kinase in response to the later β1Α mutant was shown to be impaired as well as tyrosine phosphorylation of adapter proteins paxillin and tensin whereas overall tyrosine phosphorylation of CAS was unaffected. These data suggests the presence of focal adhesion kinase-dependent and -independent pathways for tyrosine phosphorylation of CAS after integrin β1Α-mediated adhesion. </p>

<p>This thesis aimed at studying mechanisms involved in control of tissue fluid homeostasis during inflammation.</p><p>The interstitial fluid pressure (P_IF) is of importance for control of tissue fluid balance. A lowering of P_IF <i>in vivo</i> will result in a transport of fluid from the circulation into the tissue, leading to edema. Loose connective tissues that surround blood vessels have an intrinsic ability to take up fluid and swell. The connective tissue cells exert a tension on the fibrous network of the tissues, thereby preventing the tissues from swelling. Under normal homeostasis, the interactions between the cells and the fibrous network are mediated by β1 integrins. Connective tissue cells are in this way actively controlling P_IF.</p><p>Here we show a previously unrecognized function for the integrin αVβ3, namely in the control of P_IF. During inflammation the β1 integrin function is disturbed and the connective tissue cells release their tension on the fibrous network resulting in a lowering of P_IF. Such a lowering can be restored by platelet-derived growth factor (PDGF) -BB. We demonstrated that PDGF-BB restored P_IF through a mechanism that was dependent on integrin αVβ3. This was shown by the inability of PDGF-BB to restore a lowered P_IF in the presence of anti-integrin β3 IgG or a peptide inhibitor of integrin αVβ3. PDGF-BB was in addition unable to normalize a lowered P_IF in β3 null mice. Furthermore, we demonstrated that extracellular proteins from <i>Streptococcus equi</i> modulated αVβ3-mediated collagen gel contraction. Because of the established concordance between collagen gel contraction <i>in vitro</i> and control of P_IF <i>in vivo</i>, a potential role for these proteins in control of tissue fluid homeostasis during inflammation could be assumed. Sepsis and septic shock are severe, and sometimes lethal, conditions. Knowledge of how bacterial components influence P_IF and the mechanisms for tissue fluid control during inflammatory reactions is likely to be of clinical importance in treating sepsis and septic shock.</p>

The chemokine, Monocyte Chemoattractant Protein (MCP-1), enhances integrin mediated monocyte adhesion to the vascular endothelium during inflammation. In this study, we demonstrate that MCP-1 promotes rapid sub-second adhesion of THP-1 cells to Vascular Cell Adhesion Molecule-1 (VCAM-1), but not to Intercellular Cell Adhesion Molecule-1 (ICAM-1). MCP-1 activates membrane tethered Very Late Antigen 4 (VLA-4, α4β1), but not necessarily cytoskeleton anchored VLA-4. Activated tethered VLA-4 bonds tremendously increased the period of time monocytes remain bound from hundreds of milliseconds to several seconds and also increased the distance over which immunologic surveillance occurs from several microns up to 20 microns along the endothelium. Lastly at the single molecule level, MCP-1 stimulated tethered VLA-4 bonds exhibit increased resistance to pulling force. In conclusion MCP-1 increased tethered VLA-4 bond resistance to force providing a mechanism for monocyte recruitment to the endothelium.

Integrins are heterodimeric cell surface receptors, composed of an α and a β subunit, mainly binding for extracellular matrix proteins. lntegrin subunit β1 can combine with at least 12 a subunits and thus form the biggest subfamily within the integrin family. In this thesis, functional properties of the splice variant β1Β, and the effects of several mutations in the cytoplasmic tail of integrin subunit β1Α were studied. In addition, the border between the transmembrane and cytoplasmic domains of several integrin subunits was determined. The β1Β splice variant has been reported to have a dominant negative effect on functions of β1Α integrins. In this study, it was studied if the expression of β1Β had similar negative effects on the αvβ3 integrin functions since the β3 subunit is structurally similar to β1Α. The β1Β subunit was expressed in an integrin β1-deficient cell line and it was found that the presence of β1Β does not interfere with adhesion or signaling of endogenous αvβ3 The border between the cytoplasmic domain and the C-terminal end of the transmembrane domain of integrin α and β subunits has been unclear. This question was experimentally addressed for integrin subunits β1, β2, α2 and α5. It was found that integrin subunits contain a positively charged lysine, which is embedded in the membrane in the absence of interacting proteins. The functional importance of the lysine in integrin transmembrane domains was investigated by mutating this amino acid to leucine in β1Α. The mutation affected cell spreading and tyrosine phosphorylation of the adapter protein CAS. The activation of focal adhesion kinase and tyrosine phosphorylation of paxillin was not affected. Furthermore, the mutation of two tyrosines to phenylalanines in the β1Α cytoplasmic tail was found to reduce the capability of β1Α integrins to mediate cell spreading and migration. Activation of focal adhesion kinase in response to the later β1Α mutant was shown to be impaired as well as tyrosine phosphorylation of adapter proteins paxillin and tensin whereas overall tyrosine phosphorylation of CAS was unaffected. These data suggests the presence of focal adhesion kinase-dependent and -independent pathways for tyrosine phosphorylation of CAS after integrin β1Α-mediated adhesion.