Bipolar Role for Myelo-Monocytic Cells in Autoimmune Diseases and Psychiatric Disorders

Abstract

The immune system is a complex system of tissue with cells and messenger molecules interacting to protect an organism against pathogens. Autoimmunity is the failure of the immune system to recognize its own constituent parts as harmless self and therefore it leads to an immune response against its own cells and tissues. Diseases that are a results of autoimmunity are called autoimmune diseases. Autoimmune diseases discussed in this thesis are autoimmune thyroid disease (AITD), where thyrocytes of the thyroid gland are the target of the immune system and type 1 diabetes mellitus (T1DM), characterized by an autoimmune destruction of the insulin-producing β cells of the Islets of Langerhans in the pancreas. Various epidemiological studies have shown the association of psychiatric disease with AITD and T1D (and other autoimmune diseases), not only in patients, but also and independently in not affected family members of patients. This suggests a common underlying abnormality responsible for both the autoimmune endocrine diseases and mental afflictions. We hypothesized that dysfunctional cells of the myelo-monocytic cell lineage (monocytes, macrophages (MØ) and dendritic cells (DCs)) are this common underlying abnormality. The overall aim of this thesis was therefore to investigate the role of myelo-monocytic cells in the onset and pathogenesis of endocrine autoimmune diseases and psychiatric disease. We used several animal models of depressive and psychotic-like behavior and of T1DM to study the pre-stages of the disorders in chapter 2 to 4. We thereafter tested the hypothesis in chapter 5 and 6 that serum factors related to myelo-monocytic cell dysfunction can be used as tools to predict AITD and to detect the presence of schizophrenia. In Chapter 2 of this thesis, we studied the consequences of dysfunction of the MØ/DCs of the brain (the microglia) on the formation of the corpus callosum. We show in this chapter that microglia display a neurite-growth promoting function and are genuine actors of corpus callosum development. The chapter further shows that microglial inflammatory activation negatively impinges on this function. Microglial inflammatory activation was introduced by a loss-of-function of DAP12, or in the fetus by maternal inflammation via peritoneal injection of lipopolysaccharides (LPS) at embryonic day 15.5. These procedures resulted in fetal microglial activation and the defasciculation of dorsal callosal axons in the fetal brain, thereby revealing that prenatal inflammation can impair neuronal development. In Chapter 3 of this thesis, we used the Non-Obese Diabetic (NOD) mouse model and studied their putative microglial inflammatory activation. NOD mice are extensively used as a model of autoimmune T1DM and AITD. NOD mice are, however, also highly anxiogenic at steady state and show an exaggerated depressive-like behavioral response to LPS as compared to their parental strain, the CD1 mice. At steady state, we found that microglia of the NOD mouse displayed an altered gene profile characterized by a differential expression of genes involved in neuronal growth compared to CD1 mice. In addition NOD microglia showed an interferon (IFN) type 1 skewed inflammatory machinery in steady state, and an alternative IFN-driven activation pattern in response to LPS. We concluded that the differential expression in steady state of genes involved in neuronal growth supports a view of an altered development in NOD mice of brain regions critical for mood regulation, while the alternative IFN type 1 driven inflammatory reaction of microglia of LPS stimulated NOD mice supports a view in which IFN type 1 plays a critical role in LPS driven depressive-like behavior. We not only studied in the NOD mouse model the myelo-monocytic cells of the brain, but also those of the pancreas. In the pancreas DCs and MØ are the first cells to accumulate around the islets initiating the lymphocytic insulitis ultimately destroying the pancreatic β cells leading to hypo-insulinemia in the NOD mouse model of T1DM. In chapter 4 we analyzed the proliferation potential and gene expression profiles of DCs (DCs) isolated from the pancreas of pre-insulitis and pre-diabetic NOD mice. The pancreatic NOD DCs showed a reduced proliferation potential and a reduced expression of several gene networks important for the prime functions of the cell, i.e. for cell renewal, immune tolerance induction, migration and for the provision of growth factors including those for β cell regeneration. Despite these deficiencies NOD pancreatic DCs showed a hyper reactive response to LPS, which resulted in an enhanced pro-inflammatory state characterized by a molecular profile of an enhanced expression of a number of classical inflammatory cytokines. Our observations support a view in which the myelo-monocytic cells of the pancreas show a poor proliferation and differentiation in steady state, leading to early architectural islet disturbances (previously described in the thesis of Rosmalen, 2000) and a poor tolerance induction. Under conditions of Toll-like receptor (TLR) stress (e.g. local infections, a high apoptosis of β cells,) the “deficient” DC would over react with a strong inflammatory response tipping the balance over to islet autoimmunity. Collectively the findings in these three chapters and of previous theses of our group point to a crucial role of myelo‐monocytic cells in tolerance induction and in tissue homeostasis and development, most notably in the organogenesis of the brain and the islets. In steady state the cells provide various growth factors and growth mechanisms for a proper development of important brain structures, for β cells and for thyrocytes. The cells also travel, in steady state, to the draining lymph nodes carrying important endocrine auto-antigens along (thyroglobulin, insulin) and induce tolerance in the draining lymph nodes to these antigens via expansion of antigen specific natural T regulator cells. However, when the local myelo-monocytic cells are inflammatory activated by a danger signal (through the TLR for example) the provision of growth factors is hampered, leading to an abnormal development of brain structures, islets and thyroid tissue. Also tolerance induction shifts over to immunization against the endocrine auto-antigens. In the discussion a model is presented (Figure 1A) that summarizes and integrates our animal findings with those of others. This hypothetical model shows the abnormalities in proliferation, cell renewal and differentiation of the precursors in the myelo‐monocytic cell lineage (occurring at the level of the brain, bone marrow and endocrine tissues) as the key element underlying the pathogenesis of major psychiatric disorders, thyroid autoimmunity and autoimmune diabetes. The differentiation abnormality (either genetically or environmentally induced, or both) leads to a progeny of aberrant (“primed”) microglia, DC and MØ with a reduced growth support potential for the surrounding parenchymal cells, leading to architectural changes in the organs, e.g. in the brain to defasciculation and in the islets to earlier described morphological changes such as a high fibronectin (FN) content, irregularly shaped islets and mega‐islets. These morphological and growth abnormalities characterize the pre‐stages of the disorders which only become clinically evident after second hits. The differentiation abnormality of the local precursors also leads to a reduced number of local tolerogenic DC (suggested in this thesis and by earlier work at the level of the islets in the NOD) as well as an abnormal and often excessive inflammatory response of the local progeny (the microglia, DC and MØ) to TLR stimulation. This will further contribute to the loss of tolerance induction after microbial or necrotic hits and together with the innate T regulatory cell defects in the NOD mouse to the development of autoreactive T and B cells, ultimately leading to endocrine autoimmune disease.

In chapter 5 the idea was explored to use serum compounds related to growth and extracellular matrix disturbances of the endocrine tissues, as found in the pre-stages of endocrine autoimmunity in the animal models, to detect pre-stages of AITD in the human. In addition we explored for this purpose also serum compounds related to the deficiencies and the hyper reactivity of myelo-monocytic cells, as found in the pre-stages of the animal models. We measured tissue growth/remodeling factors, adhesion molecules, chemokines and cytokines in 64 TPO-antibody negative Euthyroid female relatives with at least one 1st or 2nd degree relative with documented autoimmune hyper- or hypothyroidism, 32 of whom did and 32 did not seroconvert to TPO-Abs positivity in 5 year follow-up. The relatives were compared to 32 healthy controls. We found that both sero-converting and non-sero-converting family members showed an up regulation of FN and a down regulation of PDGF-BB and of the adhesion and migration factors CCL2, CCL4, sVCAM-1, TIE-2 and MMP-13. The sero-converters differed from the non-sero-converters by up regulation of the pro-inflammatory compounds IL-1β, IL-6 and CCL3. The results provide proof of principle that pre-sero-conversion stages and sero-conversion to AITD might be predicted using serum analytes related to growth/connective tissue abnormalities and migration/accumulation abnormalities of MØ and DCs, though clearly more explorative work needs to be done. In chapter 6 we measured the serum levels of the myelo-monocytic cell related cytokines/chemokines/adipokines in 144 schizophrenia patients and compared levels to those found in age and gender matched healthy controls. We found that many of the myelo-monocytic factors were increased in the serum of schizophrenia patients. Levels of many factors were found to be linked to (components of) the metabolic syndrome, which is prevalent in schizophrenia. However, the most dominant linkage was found with the disease schizophrenia itself. The study further supported our idea that an increased myelo-monocytic cell activation as a key component for understanding the pathogenesis of schizophrenia. In conclusion, outcomes of chapter 5 and 6 (the human studies) indicate that the various (pre‐disease) stages of autoimmune endocrine diseases (and probably also of the psychiatric diseases) can in principle be predicted by detecting the abnormalities in the growth and differentiation of the myelo‐monocytic cell lineage, the neuroendocrine cells and the connective tissue compartments underlying the pathogenesis of the these disorders. This might be done by finding abnormalities in the serum levels of compounds reflecting these abnormalities, such as: 1. Growth and differentiation factors reflecting the neuro‐endocrine growth abnormalities, e.g. in PDGF‐BB. 2. Immune compounds reflecting the poor development of DC and MØ, such as reduced chemokines (this thesis) and cytokines (this thesis), 3. Factors reflecting the abnormal connective tissue composition, e.g. FN and MMPs (this thesis).