Macrophages (mφ) are one the most numerous leukocytes present in the healthy gut and contribute to both harmful and beneficial immune reactions. In the colon, mφ are exposed continuously to large amounts of material from the environment, including harmful agents such as invasive bacteria, viruses and parasites, as well as harmless materials such as food proteins and the commensal bacteria which inhabit the healthy intestine. As a result, mφ play an important role in helping defend the intestine against harmful invaders. However if these cells make similar reactions to harmless food proteins or commensal bacteria, it would be both wasteful and detrimental, likely leading to inflammatory diseases such as coeliac disease and Crohn’s disease. Several genes, which underlie susceptibility to Crohn’s disease are involved in controlling how macrophages respond to the microbiota, with considerable evidence indicating that this reflects a loss of the normal unresponsiveness that characterises intestinal macrophages in the healthy intestine. One of the most significant aspects of the epidemiology of Crohn’s disease is a particularly rapid increase in its incidence in childhood, suggesting that the first encounters between the microbiota and intestinal macrophages may be of critical importance in determining disease susceptibility. Given this link, it is essential that we elucidate the processes controlling macrophage seeding and development in the intestine and this was an aim of this thesis.In the adult healthy colon, two main mφ subsets can be identified: A dominant and homogenous one, made up of mature mφ, which express high levels of F4/80, MHC II, CX3CR1, are CD11bint/+, highly phagocytic and produce high amounts of IL10. The second mφ group is relatively smaller and is much more heterogeneous. These cells express intermediate levels of F4/80 and CX3CR1, are CD11b+ and can be divided into 3 subsets based on their levels of Ly6C and MHC II. These subsets represent a maturation continuum towards the mature mφ phenotype. Recent reports have suggested that resident macrophages in healthy tissues may be derived from yolk-sac and/or foetal liver precursors that seed tissues during development and subsequently self-renew locally. In contrast, it is proposed that macrophages in inflammation are generated by recruitment of blood monocytes, raising the possibility that these different origins could be exploited in therapy. However none of these studies have examined macrophages in the intestine and recent work in our laboratory has suggested that monocytes may be the precursors of macrophages in both healthy and inflamed gut of adult mice. Therefore, the aims of this thesis were to investigate the development of murine colonic mφ from birth until adulthood, examining the relative roles of the yolk sac, foetal liver and bone marrow monocytes, exploring their functions and comparing them with the well-characterised adult mφ. In addition, I also examined how mφ phenotype and functions are influenced by the microbiota using broad-spectrum antibiotics and germ free mice. Lastly, I examined the role of fractalkine and its receptor CX3CR1 in defining the development and functions of intestinal macrophages.Development of macrophages in early lifeThe initial characterisation and comparison of colonic mφ subsets is included in Chapter 3. In this chapter, I describe a series of experiments adapting existing protocols and techniques used for examining the adult murine intestine in order to analyse the origin, phenotype and functions of murine colonic macrophages from late foetal life through to adulthood. These studies found that intestinal mφ are present before birth, with similar levels of phagocytic ability and IL10, TNFα and CD163 mRNA expression to the adult. However, the numbers and phenotype of mφ in the intestine do not reach the adult level until the 3rd week of postnatal life. This phenomenon appears to reflect the de novo recruitment of blood monocytes in a CCR2-dependent fashion at this time and throughout adult life, but not at early stages of life.In the colon of newborn mice, two macrophage populations can be observed and are clearly differentiated based on their F4/80 and CD11b expression: F4/80hi CD11bint/+ and F4/80lo CD11b+. Interestingly, unlike adult colonic F4/80hi mφ, the majority of F4/80hi neonatal cells do not express MHC II, however they gradually express this molecule as they age. In addition to acquiring MHC II expression, the two populations in the newborn colon gradually merge and from the 3rd week of life it is difficult to discriminate them reliably. My experiments show that both mφ subsets proliferate actively during the first 2 weeks of life, but this is later reduced and maintained at low levels indicating that there is no self-renewal of mature mφ. Moreover, fate-mapping analysis carried out in collaboration with Professor Frederic Geissmann, showed that yolk sac-derived precursors contribute only minimally to the pool of colonic mφ, even at early life stages. Conversely, additional fate mapping studies suggested that most intestinal macrophages are derived from Flt3+ progenitors. Taken together, the results in this chapter demonstrate that blood monocytes are vital in replenishing the intestinal macrophage pool in the steady state, setting them apart from other tissue macrophages, which derive from primitive progenitors. Investigating the effect of the microbiota on intestinal macrophage subsetsIn Chapter 4, I assessed the effects of the commensal microbiota on intestinal mφ, using two different approaches: First, I assessed the function and gene expression of colonic macrophages following administration of broad-spectrum antibiotics. My results showed that this did not alter the numbers, phenotype, intracellular cytokine production or mRNA expression by macrophages. Several reasons may account for this, including dose/nature of antibiotics, length of administration or lifespan of macrophages. To overcome these issues, I compared the phenotype of colonic mφ in germ free (GF) and conventionally (CNV) reared mice of different ages in collaboration with Dr David Artis. Absolute absence of microbiota in GF mice severely impacted Ly6Chi monocyte recruitment to the colon, suggesting that constant recruitment of monocytes to the gut is at least in part due to the microbial burden. The biggest differences between GF and CNV mice were evident at 3 weeks of age, when GF mice had a much lower number and frequency of monocyte-derived cells than their CNV counterparts. By 12 weeks of age, Ly6Chi mφ populations from GF mice were partially restored, although the expression of MHC II by F4/80hi mφ remained reduced. Additionally, I FACS-purified F4/80hi cells from GF and CNV adults and sent RNA for microarray analysis, the results of which we are waiting to receive. This data will provide further information regarding how GF intestinal mφ differ from those found in conventional animals.Role of the CX3CL1-CX3CR1 axis in mφ development and functionAs mature colonic mφ express high levels of the chemokine receptor CX3CR1 (fractalkine), finally, in Chapter 5 I went on to investigate the role of CX3CL1-CX3CR1 axis in colonic lamina propria. In addition to the high expression of CX3CR1 by colonic mφ, its ligand, CX3CL1 has been reported to be expressed at high levels by the intestinal epithelium. Furthermore, as there is strong evidence that the CX3CL1-CX3CR1 axis may be involved in inflammation in several tissues, we hypothesised this axis might play a role in mφ function in the gut. To this end, I examined mφ phenotype, activation status and survival following in vitro co-culture of WT or CX3CR1-deficient bone marrow-derived mφ with an epithelial cell line modified to express either the soluble or membrane-bound forms of CX3CL1. I also examined the development of chemically induced colitis in CX3CR1-deficient mice. Finally, since it has been reported by the lab of Oliver Pabst, that the lack of CX3CR1 results in reduced IL10 production by intestinal mφ, I compared the ability of WT and CX3CR1-deficient mice to prime T cells after being fed with ovalbumin together with an adjuvant. The results from this chapter failed to show any definitive role of the CX3CL1-CX3CR1 axis in mφ function in either the steady state or in the setting of inflammation.My in vitro studies did not show any significant difference between WT and CX3CR1 deficient intestinal mφ in terms of survival, or co-stimulatory molecule expression, nor did bone marrow mφ (BMM) from CX3CR1 KO mice show differences in co-stimulatory molecules and pro-inflammatory cytokine production with or without stimulation by LPS. Moreover, the responses of wild type BMM were not altered by exposure to exogenous CX3CL1 either in soluble form, or when expressed as a transmembrane form by epithelial cells.The in vivo assessment of CX3CR1 during inflammation, Ly6Chi CX3CR1int cells increased after 4 days on DSS, however, the lack of CX3CR1 failed to confer protection from colitis in a consistent manner, suggesting that there may be more factors responsible for colonic inflammation apart from the CX3CL1-CX3CR1 axis. Taken together, the results of this thesis highlight that important cellular changes take place during the development of mφ in the intestine. In addition, the presence or absence of microbiota plays a crucial role in this development with acquisition of MHC II depending at least in part on the presence of microbes. Microarray data obtained from purified F4/80hi mφ populations of GF and CNV mice may reveal interesting differences and suggest how mφ phenotype and function may be regulated by the microbiota. Finally, I have shown that the CX3CL1-CX3CR1 axis plays a redundant role in the regulation of intestinal mφ phenotype and function with mφ from CX3CR1-deficient animals appearing to function normally in both health and disease.
PDF
download
878987797
028-85220240
