Autoimmunity and apoptosis

I n this issue of the Journal, Casciola-Rosen et al. (1) call . attention to a topic that has been of recent interest, the possible association of apoptosis with autoimmunity. Apoptosis, sometimes called programmed cell death, was described by Kerr et al. (2)and Wyllie et al. (3) as a form of cell death characterized by cell shrinkage, nuclear condensation, and surface blebbing not accompanied by inflammatory cell infiltration. In tissues, it affects scattered single cells rather than tracts of contiguous cells, and fragments of apoptotic cells are phagocytosed and digested by resident cells; in epithelial tissue, they are taken up by adjacent epithelial cells. A generally accepted biochemical correlate of apoptosis is the cleavage of chromatin at the intemucleosomal linker regions to produce nucleosomal fragments of different lengths which, in agarose gel dectrophoresis, appear as DNA ladders (4). In systemic lupus erythematosus (lupus), which is often regarded as a prototypic systemic autoimmune disease, there is a characteristic autoantibody response directed against selected intracellular antigens. Casciola-Rosen et al. show that UV irradiation of cultured human keratinocytes induces changes consistent with apoptosis and that certain clusters of appropriate antigens are grouped together in different apoptotic blebs in these damaged cells. The skin has long been recognized as an organ that plays an important role in lupus. This is related to the fact that skin rashes are a dominant feature of lupus. Characteristic deposits of immunoglobulin and complement at the dermalepidermal junction imply the participation of antibodies in its pathogenesis (5). In addition, the disease can manifest clinically de novo after extended exposure to sunlight and ongoing disease might be exacerbated after such exposure. For these reasons, the skin and keratinocytes have been objects of investigation by several groups (6-8). The novelty of the Casciola-Rosen study is the recognition of UV-induced injury to keratinocytes as consistent with apoptosis and the identification of clusters of nuclear and cytoplasmic antigens in different translocated sites. A population of smaller apoptotic blebs contained fragmented endoplasmic reticulum, ribosomes, and a class of ribonucleoproteins called SS-A/Ro, and larger apoptotic blebs contained nucleosomal DNA in addition to ribonucleoproteins. The prevailing dogma as to how the human organism protects itself from autoimmune reactions comes primarily from studies in the mouse showing that immature self-reactive T cells are deleted or become anergic in the thymus when confronted with self-antigens on thymic presenting cells, induding bone marrow-derived dendritic cells, a process that has been called negative selection (9, 10, 11). It appears that one mechanism by which this type of clonal T cell deletion takes place is apoptosis (12, 13). The possible rdevance of these observations to autoimmunity was heightened by the report that the MRL/Mp-Ipr/Ipr mouse, which develops a disease analogous to human lupus, has an abnormality in the Fas gene that mediates apoptosis (14). This abnormality has been determined to result from integration of the retrotransposon, Etn, into the second intron of the Fas gene resulting in diminished Fas expression and presumably leading to diminished or absent Fas-mediated apoptosis (15-17). However, several investigators have reported that Ipr mice do have the potential to delete autoreactive T cells (18-20). It has been suggested that in spite of the Fas gene abnormality observed, there is the possibility that low levels of Fas protein might be expressed (15, 17). At present, the importance of the Fas gene in autoimmune murine lupus appears to be unresolved. If it does have a role, it is not a universal feature since abnormalities of Fas mRNA expression have not been detected in NZB/W thymus or spleen (17), another murine modal of lupus which is perhaps closer in clinical and serological features to human lupus than the Igr model. Furthermore, the MRL/Mp-+/+ mouse, which is genetically identical to MRL/Ipr except for the lpr abnormality, also develops lupus, although later in life. A striking feature that has emerged from studies to learn more about the structure and function of intranuclear autoantigens is that many of these antigens are involved in important or key biosynthetic functions. Some of these are shown in Table 1. In diseases like lupus, mixed connective tissue disease, scleroderma, and dermatomyositis, autoantigens are involved in fundamental cell functions such as transcription, precursor messenger KNA splicing, DNA replication, cell division, and ribosomal protein synthesis (for a review see reference 21). Why cellular components participating in these important biosynthetic functions are targeted for autoimmune reactions is entirely unclear at the present time. The major autoantigens in these diseases have now been identified (21), and it appears that there might be a preferential autoantibody response to biosynthetic nuclear and cytoplasmic components and a paucity of antibody responses to structural and membrane components. Many investigators have been interested in defining the antigenic determinants or epitopes that are recognized by autoantibodies. An emerging consensus from these studies is that many autoepitopes are not continuous primary sequence peptides but are conformation dependent and made up of