<p>The Immunity-Related GTPases (IRGs) are a family of dynamin-like proteins found in vertebrates that play critical roles in cell-autonomous resistance to bacteria and protozoa. The IRGs are divided into two subfamilies, with the GMS IRGs exerting a regulatory function over the GKS IRGs, affecting GKS IRG expression, localization, and ultimately function. The profound loss of host resistance seen in mice lacking the GMS protein Irgm1 suggests that GMS IRGs may additionally have broader functions beyond the regulation of GKS IRGs, though the nature of these functions remains poorly understood. In this dissertation, we address the regulatory functions of GMS IRGs in mouse cells.</p><p>We first addressed regulation of GKS IRGs (Irga6 and Irgb6) by GMS IRGs (Irgm1 and Irgm3). We found that in both fibroblasts and macrophages lacking these GMS IRGs, that the GKS IRGs relocalized to form punctate structures that were ubiquitin-, p62-, and LC3-positive. A biochemical analysis indicated that the GKS IRGs were directly ubiquitinated through K63 linkages. Collectively, these results suggested that GMS IRGs regulate aggregation of GKS IRGs and their transfer to autophagosomes through one of at least two possible mechanisms -- by the direct association of GMS IRGs with GKS IRGs to block their aggregation that subsequently leads to autophagic removal, and/or by directly promoting autophagic removal of spontaneously forming GKS aggregates. The latter hypothesis was addressed using a series of complementary assays, which ultimately showed that absence of Irgm1 has no effect on the maturation of autophagosomes in fibroblasts, and only a very small and statistically insignificant effect in macrophages. Thus, we conclude that the major mechanism through which GMS IRGs regulate GKS IRGs is by directly inhibiting their aggregation, rather than through general effects on autophagic initiation or maturation of GKS IRG-containing autophagosomes.</p><p>We also addressed the possibility of broad regulatory functions of GMS IRGs beyond the regulation of GKS IRGs by examining whether GMS IRGs can affect another family of dynamin-like GTPases, the guanylate-binding proteins (GBPs). Despite no previous evidence of interactions between these two protein families, we found that the absence of GMS IRGs had striking effects on the localization of the murine Gbp2, leading it to colocalize with GKS IRG aggregates formed as a consequence of GMS IRG deficiency. We further demonstrated that unlike the GKS IRGs, Gbp2 was not tagged with K63-linked ubiquitin chains, which might have targeted it for specific macroautophagy, implying that Gbp2 is not aggregating in the absence of Irgm1. We then showed both that Gbp2 forms puncta in the presence of generic protein aggregates, and that guanylate-binding proteins including Gbp2 promote the degradation of GKS IRG protein aggregates. These findings suggest that GMS IRGs do not exert direct control over GBPs, but rather that GBPs are involved in the macroautophagic degradation of protein aggregates as a primary function, and are thus influenced indirectly by GMS IRGs.</p><p>In total, our experiments contribute to the understanding of regulatory interactions among GMS IRGs, GKS IRGs, and GBPs. These results will be important in establishing the mechanisms through which these important families of proteins influence eradication of bacterial and protozoan pathogens through key innate immune mechanisms.</p> / Dissertation

Cardiovascular disease, and myocardial infarction (MI) in particular, remains a major burden in the developed world today. In fact, the remodeling process, which follows the initial ischemic episode of MI, is a major determinant of heart failure. Although several key mechanistic pathways involving cell growth and death have been identified, there is limited knowledge surrounding the role of the innate immune response as a positive or negative regulator of cardiac remodeling. Recent data strongly support a role for key regulatory components within the toll-like receptor (TLR) family as potent modulators of cardiac remodeling post-MI. It has been demonstrated that targeted gene knockdown of TLR4, as well as downstream adaptor proteins and kinases, significantly improve cardiac function and overall survival. While the well-known NF-κB transcriptional factor that is downstream to TLR4 signaling has been linked to remodeling, there has been no evidence thus far describing a role of the parallel interferon regulatory factor-3 (IRF3) signaling cascade in any facet of this process. Several key findings suggest that IRFs contribute to both cell growth and apoptosis, thus providing an appealing, and novel target for interrogation. In this thesis I describe how IRF3 contributes to maladaptive remodeling post-MI. In my first set of experiments, I demonstrate that IRF3 is acutely upregulated within the cardiomyocyte following MI and that this response contributes to excessive apoptosis post-MI. A targeted deletion of the IRF3 gene enhances cardiac function, decreases infarct size, and improves survival following MI. In the second set of experiments I demonstrate that IRF3 attenuates angiogenesis at the ischemic border zone by upregulating the expression of thrombospondins. I have shown that IRF3 deficiency, which liberates endogenous anti-angiogenic signals, promotes angiogenesis following ischemic injury. These data suggest that IRF3 is a potent regulator of cardiac remodeling and may be an effective therapeutic target to ameliorate maladaptive cardiac repair post-MI.

B virus (Cercopithecine herpesvirus 1) is an alphaherpesvirus indigenous to macaque monkeys and is closely related to herpes simplex virus type 1 (HSV-1). Disease caused by B virus, which is often mild or asymptomatic in its natural host, the macaque monkey, is similar in infected macaques to HSV-1 infection in humans. When B virus zoonotically infects foreign hosts, e.g., humans, high morbidity and mortality are evidenced in > 80% of untreated cases. To explore the underlying reasons for differences in pathogenesis between B virus and HSV-1 infection in humans, human microarrays were used to comparatively examine global cellular gene expression patterns engaged as a result of infection of human foreskin fibroblasts (HFFs). Our results demonstrate that these closely related simplexvirus family members have divergent strategies to thwart host cell pathways related to innate defenses. In these studies, B virus did not induce detectable interferon, cytokine or chemokine genes, in sharp contrast to HSV-1, which induced innate immune responsive genes in infected cells. Although no innate immune response genes were found to be up-regulated by B virus infection, B virus induced I£eB£a, which was the only gene found to be involved in the NF-£eB signaling pathway within the innate immunity biological network. Quantification of NF-£eB p50 DNA binding activity in virus-infected nuclear extracts demonstrated that NF-£eB p50 DNA binding activity was lower in B virus-infected cells. Suppression of I£eB£a in B virus infected cells by siRNA restored NF-£eB-induced cytokine and chemokine expressions. Data presented here support the model that I£eB£a inhibits NF-£eB regulated immune responsive genes in B virus-infected HFF cells, and this response differs from that observed in HFF cells infected with HSV-1. The result is that B virus alters the NF-£eB regulated expression of cytokine and chemokine genes of HFF cells differently from HSV-1 early after infection. These differences in cytokine and chemokine expression may be associated with the delayed or reduced host responses observed in B virus infected humans and underlie the failure of adaptive responses in zoonotically infected humans.

Mannose binding lectin (MBL) belongs to the collectin family of soluble pattern recognition molecules that elicit diverse biologic activities. Via multiple carbohydrate-recognition domains (CRD), MBL binds to mannose and N-acetyl-glucosamine oligosaccharides present on the surface of bacteria, fungi and yeast. Following pathogen recognition, MBL activates the complement system via MBL associated serine proteases in a manner independent of antibody and C1 complex. Deficiency in function and level of MBL is found in 25% of otherwise apparently healthy individuals, representing the most prevalent innate immune deficiency. MBL deficiency is a risk factor for the development of infections in humans and mice. The role of MBL as a modulator of infection is complex. MBL deficiency may influence proinflammatory cytokine production, expression of leukocyte adhesion molecules, or vascular damage, during the course of infection. Given that dendritic cells (DC) are antigen presenting cells (APC) with potent capacity to respond to microbial stimulation, I hypothesized that MBL deficiency may be reflected in DC functions associated with microbial stimulation. Initially, I investigated the association of MBL with human immune cells and demonstrated that in both MBL-Sufficient (MBL-S) and MBL-Deficient (MBL-D) individuals, MBL was particularly associated with monocytes. RT-PCR analysis demonstrated MBL was not transcribed by monocytes or other immune cells investigated (T, B, and NK cells, CD11c+DC, immature monocyte derived DC [MoDC], LPS matured MoDC, and granulocytes), suggesting MBL association with the cell surface may be via an adapter or co-receptor. Magnetically separated monocytes but not MoDC bound exogenous purified human plasma MBL (hpMBL). Addition of hpMBL (5 -15 µg/mL) did not induce MoDC activation, and MBL added together with lipopolysaccharide (LPS) did not induce MoDC activation above the level induced by LPS only. In the second part of this study, I used the particulate MBL ligand zymosan (Zy) as a pathogenic stimulus in a whole blood model to gain a greater understanding of the consequences of MBL deficiency. I compared surface phenotype, inflammatory cytokine production and antigen presenting capacity of blood myeloid (M)DC of MBL-D and MBL-S individuals following stimulation with Zy and MBL opsonised Zy (MBL-Zy). Blood MDC in MBL-D individuals, unlike their counterpart in MBL-S individuals, displayed unique functional characteristics, including higher production of proinflammatory cytokines IL-6 and TNF-, but poor capacity for allo-T cell effector cell induction. It appeared that stimulation with MBL-Zy reduced elevated production of IL-6 but not TNF- by blood MDC in MBL-D individuals. In the third part, expression microarray analysis was utilised to provide broad information on the genes and potential signalling pathways involved in the MDC responses in MBL-D and MBL-S individuals following stimulation with Zy and MBL-Zy. MBL-S individuals demonstrated greater capacity to induce T cell and NK cell signalling pathways than MBL-D individuals. Further, MBL acted as a regulator of important inflammatory molecules, namely T-cell receptor zeta (CD247), IFN-γ and perforin 1. The data presented in this study provides novel information on blood MDC function in MBL-S and MBL-D individuals in response to pathogen stimulation, and provided insight into mechanisms involved in the increased frequency of infection observed in MBL-D individuals.

Much of our current mechanistic understanding of the innate immune response in animals has grown out of empirical work in insect models, especially the fruit fly Drosophila melanogaster. The mainstream understanding of the fly immune response to bacteria has been that it exists in two parts; a cellular and a humoral response. Drosophila also harbor substantial genetic variation for antibacterial defense and investment in immunity is thought to involve a costly trade-off with life history traits, including development, life-span and reproduction. My first study (chapter 2) aimed to understand the way in which insects invest in fighting bacterial infection. We selected for survival following systemic infection with the opportunistic pathogen, Pseudomonas aeruginosa in wild-caught D. melanogaster over 10 generations. We then examined genome wide changes in expression in the selected flies relative to unselected controls, both of which had been infected with the pathogen to specifically identify the genetic basis of the evolved immune response. In response to selection, population level survivorship to infection increased from 15% to 70%. The evolved capacity for defense was costly as evidenced by reduced longevity and larval viability and a rapid loss of the trait once selection pressure was removed. Counter to expectation, we observed more rapid developmental rates in the selected flies. Selection associated changes in expression of genes with dual involvement in developmental and immune pathways suggest pleiotropy as a possible mechanism for the positive correlation. We also found that both the Toll and Imd pathways work synergistically to limit infectivity and that cellular immunity plays a more critical role in overcoming P. aeruginosa infection than previously reported. Females usually produce a more robust immune response and are often less susceptible to infection. This female bias has been documented in humans, mice and some birds and reptiles. The most common explanation is that males increase their mating success at the cost of immune investment whilst females invest in immunity to maximize life-time egg production. In insects, however, there is growing evidence of male-biased immune performance. Using fly survival data from my first study, I found that males exhibited higher post-infection survival than females. In my second study (chapter 3), I related these differences in survival rate to changes in gene transcription. Firstly, we examined the expression of a set of immunity genes in both sexes in the presence and absence of infection. We found that male-biased survival may be partially attributable to a higher baseline expression of immune genes in males. Contrary to previous published work, we found that immune gene expression was readily induced in flies upon exposure to P. aeruginosa and that the two sexes responded in a similar manner. Lastly, we found that selection altered the expression of genes in males alone and only in the presence of infection. Together our findings suggest a superior immune response in male Drosophila. Wolbachia pipientis is an obligate intracellular bacterium capable of spreading itself through populations by manipulating the reproduction of its hosts. The Wolbachia strain wMelPop, which reduces longevity in D. melanogaster, has been introduced into the Dengue virus mosquito vector, Aedes aegypti, as a strategy to reduce disease transmission. The infecting Wolbachia halve the lifespan of the mosquito and induce numerous behavioral and physiological abnormalities including heightened locomotor activity and an age dependent reduction in blood feeding success. In my third study (chapter 4), we aimed to understand the mechanism underpinning these changes and hence chose to explore how Wolbachia may be interacting with the insect’s nervous and muscle tissue. Because wMelPop over-replicates in Drosophila, first we measured the bacterial density in A. aegypti. We found that there was a relationship between some of the feeding associated defects in the mosquito and the density of Wolbachia in the nervous and muscle tissue. Next, we carried out a series whole genome profiling experiments based on the head and muscle tissues to identify mosquito pathways affected by the microbe. Key findings that may relate to the phenotypes of interest include increased expression of genes relating to muscle contraction and synthesis of the neurotransmitter dopamine. Other novel findings that may not relate directly to the phenotypes of interest include evidence of a strong local tissue based immune response and widespread changes in expression of mosquito methylation and acetylation associated genes. We then used then amplification of inter-methylated sites (AIMS test) to obtain DNA fingerprints representative of the methylome of A. aegypti infected and uninfected with wMelPop. The presence of wMelPop caused hypermethylation in loci where they were not methylated in uninfected mosquitoes.

γδ T cells belong to the T cell lineage however they possess some innate like properties. γδ T cells recognize non-peptidic microbial and stress induced self antigens in a non-MHC restricted manner and are proposed to bridge the gap between innate and adaptive immunity. Dendritic epidermal T cells are a prototypic population of intraepithelial γδ T cells in murine skin. Found in the basal layer of epidermis in close contact with Langerhans cells and keratinocytes DETC facilitate vital immunological and physiological processes e.g. wound healing, homeostasis, tumour surveillance and regulation of inflammation. The purpose of this thesis was to elucidate whether γδ T cells and in particular DETC play a role in generation of adaptive immune responses to foreign cutaneous antigen (OVA) in the context of skin grafts. Skin grafting has long been established as a means to test cutaneous and epithelial immunity. To answer this question, γδ T cell knock-out mice (TCRδ-/-), transgenic K5mOva mice and a skin grafting model were used. It is shown in this study that in the absence of γδ in the skin and not in the circulation there is a lower rejection rate of OVA expressing skin grafts. This phenomenon is observed in both freshly placed and well healed grafts. To understand which part of the immune response is affected by the absence of γδ T cells the priming and effector phases of the immune response was examined in TCRδ-/- mice. The priming phase was studied using two approaches: the first approach was to test priming to maximal doses of subcutaneous antigen in conjunction with an adjuvant and the second approach involved testing priming to an antigen in the context of skin grafts (graft priming). Using ELISPOTs and CFSE proliferation assays we found that while administration of OVA in conjunction with an adjuvant (QuilA) via the subcutaneous route results in sufficient priming in γδ T cell knockout mice, cross priming to OVA in the context of - freshly placed and well healed skin grafts is impaired in TCRδ-/- mice. By immunizing TCRδ-/- mice prior to skin grafting or by transferring in vitro primed OT-I cells to RAG-/- mice grafted with K5mOVA or TCRδ-/-OVA skin it was shown that 100% of all OVA grafts are rejected regardless of presence or absence of γδ T cells, concluding that effector phase of the immune response is not affected in this model. The inability of DETC to perform the role of cross presentation leads to the hypothesis that DETC indirectly enhance this process by affecting professional antigen presenting cells (APC) of the skin. Based on the contribution of DETC to wound healing it was hypothesized that the migration of dendritic cells (DCs) from the skin grafts to the lymph nodes may be affected. When this hypothesis was tested using hapten sensitization and congenically marked skin grafts it was shown that migration of DCs from skin grafts is not affected by the absence of DETC. In another hypothesis the co-stimulatory markers CD40 and CD86 were examined on migrating DCs found in the skin draining lymph nodes of grafted mice and it was shown that expression levels of those molecules were lower on DCs from TCRδ-/- grafted mice compared to C57BL/6 control mice. In addition using cytometric bead array, we show that the cytokine milieu in TCRδ-/- skin and skin draining lymph nodes is different from that of wildtype C57 skin and this disparate cytokine profile may be contributing to the less efficient cross priming and graft rejection in TCRδ-/- mice.

Human cytomegalovirus (HCMV) is a highly prevalent opportunistic infection and a major pathogen in immune-compromised patients. The virus exhibits a wide cell tropism and is able to lytically infect virtually any cell type, with detectable gene expression and release of new virions, but not the neutrophil. This cell is the first immune cell to engage most pathogens, engulfing and killing them before undergoing apoptosis and clearance by macrophages. However certain viruses and bacteria are able to evade host defences and use the neutrophil as a “Trojan horse” for replication and dissemination. In this context, enhanced neutrophil survival may promote infection. This work describes a profound neutrophil survival phenotype elicited by contact with live or UV-inactivated HCMV, in the absence of lytic viral gene expression. The effect does not involve signalling through candidate Toll-like receptors, but is dependent on activation of the ERK MAPK and NFκB signalling pathways, and is viral strain-dependent, restricted to clinical strains of the virus. Furthermore, HCMV triggers the secretion of a bioactive secretome that induces a similar paracrine anti-apoptotic effect in fresh neutrophils, and stimulates monocyte chemotaxis and differentiation to a phenotype that is permissive for HCMV infection. This “transferrable” effect is not due to residual virus or the presence of well-known neutrophil survival factors such as IL-6 or IL-8, but is mediated by a heat-stable protein or lipid, secreted late in culture. These results are supported by data in neutrophils isolated from patients with CMV viraemia and pneumonitis which show increased survival ex vivo, and will be further investigated using plasma membrane profiling by amino-oxybiotinylation and tandem mass tag mass spectrometry. This technique, used for the first time here in a primary cell type, allows quantitative proteomics to be performed for the first time in the neutrophil. This work demonstrates that the technique provides a comprehensive readout of all neutrophil plasma membrane proteins in a sample, with high plasma membrane purity and minimal neutrophil activation and necrosis, validated by flow cytometry. Furthermore, this has been applied to generate plasma membrane profiles for the resting, inflammatory and apoptotic neutrophil, revealing a number of neutrophil cell surface molecules not found by previous membrane proteomic methods. This technique has the potential to analyse the effect of HCMV and other pathogens on the expression profile of the neutrophil surface membrane and to examine how neutrophil signalling and function is modulated. These data shed light on the role of neutrophil apoptosis as a potential promoter of HCMV infection, and have the potential to increase our understanding of both the neutrophil’s response to pathogen invasion and to generate future approaches to combating HCMV dissemination and pathogenesis.

In the first part of the theses, I studied therapeutical effect of intratumoral application of various types of bacteria on melanoma B16-F10 bearing mice alone or in combination with another immunostimulatory compounds. Tumour size, metastasis and survival were monitored. The second part was focused on study of cytotoxic effect of neutrophils on melanoma cells and the role of neutrophil extracellular traps formation.

Saliva of Ixodid ticks contains a whole array of pharmacologically active molecules with vasodilatory, antihemostatic, and immunomodulatory activities. This thesis focuses on two types of salivary proteins, serpins and cystatins, and their role in immunomodulation. These protease inhibitors are known to affect many biological functions. To better understand their role in tick saliva we examined their effect on dendritic cells and their ability to modulate the immune response after pathogen infection. As model pathogens, Borrelia spirochetes and tick-borne encephalitis virus were used.

Defensinas são peptídeos antimicrobianos produzidos na mucosa intestinal e fazem parte da imunidade inata, agindo sobre vários microrganismos luminais. Deficiência na expressão de defensinas tem sido relatada em doenças inflamatórias intestinais (DII), no entanto a contribuição de cada tipo de defensina, num cenário de polimorfismo genético, mantém alguma controversa. Βeta-defensinas humanas (HBDs) têm atividade antimicrobiana contra uma ampla variedade de fungos, bactérias e vírus e têm também, um papel na ligação entre a imunidade inata e adaptativa atuando como quimiotáticos. O gene DEFB1 (8p23), codificando a beta-defensina humana 1 (HBD-1), é expresso normalmente por células epiteliais de uma série de tecidos, mas sua expressão pode variar entre indivíduos e pode ser modificada durante processo inflamatório. Produção deficiente de defensinas parece contribuir para a patogênese de DII, e uma diminuição na expressão de HBD-1 tem sido relatada na mucosa de pacientes com doença de Crohn (DC) e retocolite ulcerativa (RCU). Nós avaliamos a possível associação de três polimorfismos do gene DEFB1 com a suscetibilidade a DII, RCU e DC, em 149 pacientes, 79 com DC e 70 com RCU; e 200 controles saudáveis do sul do Brasil. No nosso estudo não se observou diferença estatisticamente significativa entre a distribuição das frequências alélicas para DEFB1 SNPs -52G>A. -44C>G e -20G>A entre o total de pacientes com DII e controles. Porém, quando pacientes com DC foram estratificados de acordo com a localização anatômica, o alelo -20G>A foi mais frequente em pacientes com DC colônica do que em controles (65 % VS 44 %, p=0,048). De forma similar, o genótipo A/A foi mais frequente em pacientes com DC colônica do que em controles (36 % VS 16 %), mas neste caso, a diferença não foi estatisticamente significativa (p=0,07). Embora não se achou uma clara e forte associação entre os SNPs 5’-UTR DEFB1 e suscetibilidade/proteção à doença inflamatória intestinal, nossos resultados sugerem possível envolvimento do gene DEFB1 nestas enfermidades, especialmente com a localização colônica da doença de Crohn. Estudos com amostras maiores e populações diversas serão úteis para avaliar a tendência observada no nosso grupo. / Defensins are antimicrobial peptides produced by the intestinal mucosa and are part of the innate immune system, playing a protective role against various intestinal microorganisms. Deficiency in the expression of defensins has been reported in inflammatory bowel diseases (IBD), however there is some controversy over the contribution of each type of defensine, in a setting of great genetic polymorphism. Beta-defensins (HBDs) have an antimicrobial activity against a great variety of fungi, bacteria and viruses, and also have a role in connecting the innate and the adaptive immunity, acting as a chemostatic agent. Deficient production of defensins appears to contribute to the pathogenesis of IBD, and the lower expression of HBD-1 has been reported on the mucosa of Ulcerative colitis (UC) and Crohn’s disease (CD) patients. We evaluated a possible association of three polymorphisms of gene DEFB1 with susceptibility to develop IBD, UC and CD in 149 patients, 79 with CD and 70 with UC; and 200 healthy controls from the south of Brazil. The gene DEFB1 (8p23), which codifies human beta-defensin 1 (HBD-1), is constitutivelly expressed by epithelial cells of several tissues, but its expression may vary among different individuals and may be modified by inflammation. In our study we did not find a statistically significant difference between the distribution of the allelic frequencies for DEFB1 SNPs -52G>A, -44C.G and -20G>A between the total number of patients and controls. However, when patients were stratified according to the anatomic location, the allele -20G>A was more frequent in patients with colonic CD than in contros (65% VS 44%, p=0,048). Similarly, the genotype A/A was more frequent in patients with colonic CD than in controls (36% vs 16%), however, in this case, the difference wasn’t statistically significant (p=0,07). Although we did not find a clear and strong association between the 5’-UTR DEFB1 SNP and susceptibility to IBD, our results suggest a possible involvement of the DEFB1 gene and these diseases, particularlly colonic CD. Further studies with larger samples and diverse populations will be usefull to evaluate the trend observed by our group.