Immunopathogenetic features characterize the autoimmune inflammatory myopathies - polymyositis, dermatomyositis, and related diseases: lymphocytic destruction of muscle cells, and humoral autoimmunity distinguished by a striking set of disease-specific autoantibodies. Although the muscle cell destruction is mediated by lymphocytes, the autoantibodies, particularly those directed against the family of functionally related but structurally diverse aminoacyl-tRNA synthetases, seem to offer a useful window on the disease and have been the focus of much of this group's research for a number of years. The question of repertoire selection has driven a considerable body of immunological research for many decades, with most attention being focused on the possibility that molecular mimicry between an inciting microbe and a self protein was responsible. The major alternative though not strictly competing idea is that autoantibodies are a consequence of dysregulation of the immune system. There are powerful reasons to doubt the ability of either of these ideas to explain the facts on the ground - molecular mimicry failing because of the paucity of convincing examples in naturally occurring circumstances despite intense searches; and dysregulation failing because of its inability to account for the quite limited repertoire of autoantibodies observed. This repertoire is approximately the same in all human populations that have been examined, and is the same in mice that develop autoimmune disease either spontaneously or as a result of immune system manipulation through knockout or transgenic technology. This strongly suggests that factors intrinsic to the target autoantigens are important, and observations by several groups have provided new ideas to support that view. For example, the frequent presence of long charge-rich regions, particularly coiled-coils; the presence of cleavage sites for the enzyme granzyme B; and the localization of autoantigens on or near the surface of cells undergoing apoptosis; and alterations in the structure such as phosphorylation, dephosphorylation, deglycosylation, or proteolytic or nucleolytic digestion. I considered the possibility that a direct interaction of an autoantigen with the immune system through a receptor concerned with immune responses might contribute to autoantibody synthesis. The immune response might entirely resemble an ordinary immune response to a foreign antigen - as was apparently the case for HRS - if the route to the response was the same as that a foreign antigen took. I asked Joost Oppenheim and Zack Howard of the Laboratory of Molecular Immunoregulation of the NCI Center for Cancer Research at Frederick if they would collaborate in a project to test available autoantigens for chemokine activity. The major test system has been a standard chemokine assay of the migration of appropriately purified human cells obtained from peripheral blood through a Millipore membrane. The results of the initial round of experiments with five aaRS have recently been publishe. In brief, the experiments established that HRS and another myositis specific autoantigen, AsnRS, have chemoattractant activity for immature dendritic cells (iDC), via CCR5 and CCR3 respectively; that SerRS, an occasional autoantigen in lupus, has chemoattractant activity for CCR3-transfected cells, but not for iDC; and that LysRS and AspRS lacked chemoattractant activity. The accumulated evidence is that HRS is not acting as a classical chemokine. It does not cause an immediate change in calcium flux; it does not block the binding of a chemokine, RANTES, to CCL5 even though it de-sensitizes the receptor for RANTES stimulation; it requires for its effects only a portion of the extracellular part of CCR5, as shown by experiments with chimeric constructs and with blocking antibodies; it appears not to lead to the up-regulation of CD80 or CD86 on cultured iDC. In the past year we have concentrated on following the events that occur when HRS interacts with iDC. There is reason to suppose that the interaction with iDC, and possibly with the other cells that are chemoattracted to HRS, is connected to the selection of HRS as a target of autoantibodies, but exactly how is not clear. In collaboration with Susan Pierce of NIAID, we are attempting to follow the fate within the cells using confocal microscopy. So far, there is evidence that HRS can enter human iDC and it appears to enter an MHC Class II pathway, but in these experiments, the receptor used has not been rigorously demonstrated. A number of experiments on the signaling and secretory events that follow the binding of HRS to iDC have so far been frustratingly negative. Neither phosphorylation of signaling pathway proteins nor MAP kinase activation has yet been demonstrated, though these experiments are being repeated using more sensitive detection. Experiments looking at phosphoinositol pathway events are planned. Neither Blys nor April synthesis was detected. Tests for the secretion several cytokines, carried out in Dr. Howard?s and Oppenheim?s lab have been negative. An important part of the project is to determine whether and in what form HRS or fragments of HRS get expressed on the surface of DCs. Peripheral lymphocyte lines grown with HRS stimulation, from three patients with anti-HRS autoantibodies have been made with the plan to consider trying to get iDC from the same patients for obtaining evidence of stimulation. In experiments with Dr. Howard, we have continued extending the chemoattractant observations to a much larger universe of autoantigens and non-autoantigens. A number of control recombinant proteins and more recombinant autoantigens have been tested, but the data collection is not complete. A project to determine whether CCR5 is actually necessary in order to develop anti-HRS autoantibodies was attempted in mice, but we were unable to reproduce the experimental system developed by a group in Australia to elicit myositis and anti-HRS antibodies via the injection of cDNA for human HRS, and the breeding experiments to bring the transgenes necessary to re-create our model myopathy in CCR5-/- mice were unsuccessful. However, in experiments in the Howard lab using a variety of retinal autoantigens, it has been shown that a number of them have definite chemoattractant properties using several other chemokine receptors found on iDC. In a three-way collaboration between our group, the Oppenheim group in NCI, and the Pierce group in NIAID, we will develop a test system to study the chemoattractant properties of autoantigens in mice and the signaling pathways in mouse iDC. Although the mouse iDC are harder to work with than human iDC, the ability to genetically, immunologically, and biochemically manipulate the system is far greater.
Showing the most recent 10 out of 11 publications
