Type 1 diabetes is characterized by a T cell mediated destruction of the insulin-producing ß cells in the islets of Langerhans that are situated in the pancreas. Prior to the infiltration of lymphocytes into the pancreas, an accumulation of macrophages (mf) and dendritic cells (DC) is observed. It is generally thought that these mf and DC originate from blood monocytes that have entered the pancreas and have differentiated into mf or DC. On the basis of their normal physiological role, these cells presumably take up self-antigens and process these into peptides, which they present, after a so-called “steady-state” or “homeostatic” trafficking of the DC to the draining lymph nodes, to T lymphocytes in the para-cortical area. Normally this leads to tolerance induction and T regulatory cells are induced. However in the case of diabetes development, not regulatory T cells, but erroneously effector T lymphocytes become activated and islet autoimmunity is induced. These effector T cells that were primed in the lymph node, become re-activated after re-circulation upon recognition of the self-antigens in the pancreas and consequently – together with macrophages - initiate inflammation and mediate ß cell destruction, which is a hallmark of type 1 diabetes. In this thesis I have studied the processes involved in the early accumulation of mf and DC in the pancreas prior to the infiltration of lymphocytes. The extravasation of monocytes from the circulation into the pancreas is a complex process. Amongst other factors, adhesion molecules, chemokines and myeloid related proteins (MRPs) play an important role in the adhesive and migratory responses of the monocytes that enable effective extravasation. I studied the adhesive and migratory behaviour of human monocytes of type 1 diabetic patients and compared these functions with those of monocytes of type 2 diabetic patients and healthy control subjects. First of all, I was not able to detect any differences between patients and control subjects regarding the subdivision of circulating monocytes in mature and immature cells based on the expression of CD14 and CD16 (chapter 2.1). Secondly, monocytes of patients with type 1 diabetes displayed an intrinsically increased surface expression of the pro-inflammatory molecule MRP8/14 and an increased serum level of MRP8/14. When monocytes were allowed to adhere to the extra cellular matrix component fibronectin the cells showed an enhanced expression and production of MRP8/14. The monocytes of type 1 diabetes patients showed the strongest expression and production of MRP8/14 which was significantly increased over that of healthy control monocytes (chapter 2.2). Furthermore, such activated type 1 diabetic monocytes showed an even stronger adhesion to fibronectin, which was indeed found to be the effect of exposition of the monocytes to MRP8/14 (chapter 2.1). My findings suggest a positive feedback mechanism regarding the adhesive capacity of monocytes in type 1 diabetes: circulating monocytes express and secrete higher levels of MRP8/14 as compared to healthy control subjects, resulting in increased MRP8/14 in the serum. The increased serum MRP8/14 induces an increased adhesive capacity to fibronectin of the monocytes, which leads to an even larger secretion of MRP8/14 compared to healthy controls. In this thesis I also describe that the increased MRP8/14 in the serum induced an increased expression of CD11b/CD18 on the monocytes that is likely involved in the increased adhesion of the type 1 diabetic monocytes to endothelial cells that I observed (chapter 2.2). After the adhesion studies I investigated the migratory behaviour of monocytes of type 1 diabetes patients and observed a remarkably decreased response towards the pro-inflammatory chemokines CCL2 and CCL3. Both the transendothelial migration (measured in a Transwell system) and the chemotaxis (measured in the classical Boyden assay) towards these pro-inflammatory chemokines were decreased for monocytes in type 1 diabetes. In contrast, the chemotaxis of monocytes of type 1 diabetes patients to the constitutively in lymphoid-tissue expressed CCL19 was increased in comparison to healthy controls (chapter 2.2). Since in human subjects it is not feasible to study the diabetes development in the pre-diabetes stage, I also studied the adhesive and migratory capacity of monocytes of the NOD mouse, a widely used spontaneous animal model for type 1 diabetes, before the actual development of lymphocytic insulitis. Unlike patients, NOD mice showed increased numbers of mature (in the mouse Ly-6Clow) monocytes in the circulation and a preferential differentiation of monocytes towards mf (chapter 3.1). In contrast to control mice, NOD mice did not down regulate the fibronectin adhesive capacity of their monocytes upon maturation and the NOD Ly-6Clow monocytes displayed an increased adhesion to fibronectin as compared to the NOD Ly-6Clow monocytes of control mice. Hence a shared feature of monocytes of human type 1 diabetic subjects and NOD mice is an increased adhesive capacity to fibronectin (chapters 2.1 and 3.1). With regard to the migratory potential of monocytes in NOD mice, I studied the in vivo recruitment of monocytes in response to inflammation using two models. In the peritonitis model, in which a sterile inflammation is induced in the peritoneum by injection of thioglycollate, the recruitment of monocytes was severely decreased as compared to control mice to various pro-inflammatory stimuli (chapter 3.2). Also in the air pouch model (a model in which sterile air is injected subcutaneously on the back of the mouse, creating a body cavity in which chemokines can be injected) the monocyte recruitment to the pro-inflammatory chemokines CCL2 and CCL3 was severely hampered (chapter 3.2). Like for adhesion these findings show a parallel between NOD mice and type 1 diabetes patients: in both situations monocytes show a poor reaction to pro-inflammatory chemokines. I also studied the expression of chemokines in the pancreas of NOD mice during the development of diabetes. An increased expression of the pro-inflammatory chemokine CCL2 could never be detected as compared to control mouse strains. The expression of the pro-inflammatory chemokines CCL5 and CXCL10 was found only to be expressed in the NOD pancreas at a higher level at a later stage of the insulitis, i.e. at the time of the infiltration of T and B lymphocytes, hence after the early accumulation of the mf and DC (chapter 4.1). At the time of the early mf and DC accumulation an increased expression of the lymphoid tissue-related chemokines CCL19 and CCL21 was observed. Interestingly DC of NOD mice showed in vitro a decreased migration to the lymphoid tissue-related chemokine CCL19 (chapter 4.1), but an increased adhesion to fibronectin. NOD monocytes were not tested for their chemotactic response towards this chemokine. It is likely that the increased adhesion to endothelial cells and fibronectin and the altered response of monocytes (and DC) to pro-inflammatory and constitutively expressed chemokines are variables determining the enhanced accumulation of mf and DC in the pancreas in the early stages of the diabetic process. My findings suggest that particularly the increased adhesion of the cells to fibronectin may lead to a retention of mf and DC in the pancreas. There the mf and DC may contribute to an inflammatory environment by the production of e.g. TNF-a. Such local inflammatory environment induces the expression of pro-inflammatory chemokines by the ß cells and likely also by the mf and DC. Indeed we observed an increased expression of pro-inflammatory chemokines in the NOD pancreas at stages in which already an accumulation of mf and DC was observed. The infiltration of T and B lymphocytes correlated with the expression of these pro-inflammatory chemokines. Our data hence suggest that the inflammatory environment in the pancreas of NOD mice is rather the consequence of the accumulating mf and DC than the cause. Both human and mouse monocytes showed a decreased migratory response towards pro-inflammatory chemokines, suggesting that an inflammatory-driven influx of monocytes as the origin of the early mf and DC accumulation in the pancreas is not likely. It could be that the monocytes use the lymphoid tissue-related chemokines CCL19 and CCL21 as an alternative route to enter the pancreas. It could also be that these chemokines are involved in the formation of secondary lymphoid tissue as have been previously been described in the NOD mice and the RIP-LCMV model for diabetes. The precise role of CCL19 and CCL21 in the development of diabetes is not clear, but it provides an interesting subject for further investigation. In conclusion, in this thesis I show an interesting parallel between the adhesive and migratory behaviour of human and mouse monocytes in type 1 diabetes, providing evidence for a relation between the aberrant adhesive and migratory capacity of monocytes and DC and the development of type 1 diabetes. The mechanisms whereby an increased fibronectin adhesion and altered migration of the monocytes and DC triggers the autoimmune process need to be investigated further, providing a new and interesting challenge.

H.A. Drexhage (Hemmo) , M.A. Versnel (Marjan)
Erasmus University Rotterdam
The studies were financially supported by the European Commission (QLRT-1999-00276-MONODIAB) and partly by the “Trustfonds” Erasmus University Rotterdam. The printing of the thesis was supported by Dr. Ir. Van de Laar Stichting, Sanquin-Reagents, J.E. Jurriaanse Stichting and by Biosource International.
Erasmus MC: University Medical Center Rotterdam

Bouma, G. (2005, March 16). Adhesion and Migration of Monocytes and Dendritic Cells in Type 1 Diabetes. Erasmus University Rotterdam. Retrieved from http://hdl.handle.net/1765/6740