We thank Clancy et al for their interest in our work and for providing us with an opportunity to speculate regarding why some autoantigens are confined within apoptotic blebs while others are expressed at the bleb surface. In a large, multicenter study, we demonstrated the association between the presence of autoantibodies to nuclear lamin B1 in patients with systemic lupus erythematosus (SLE) and protection against thrombosis (thromboprotection). We then aimed to elucidate the mechanism by which autoantibodies to nuclear lamin B1 cause thromboprotection in vivo. Because a number of autoantigens in SLE have been localized specifically to the external surface of apoptotic blebs (1–3), we hypothesized that circulating autoantibodies to nuclear lamin B1 may block the procoagulant effect of apoptotic blebs by binding to lamin B1 displayed at the external bleb surface. Therefore, with biochemical and morphologic studies, we determined the localization of lamin B1 in apoptotic cells and blebs. We used Jurkat cells, a well-studied cell type model for apoptosis, and human umbilical vein endothelial cells. Apoptosis was induced using 2 agents: staurosporine (to induce the mitochondrial apoptotic pathway) and anti-Fas antibody (to trigger the cell death receptor apoptosis pathway). In all cases, lamin B1 was shown to be translocated into surface blebs during apoptosis but was entirely enclosed within the apoptotic bleb membrane.
Clancy et al, with microscopy studies, now extend our results to another cell type, cardiomyocytes, which were induced into apoptosis using DMNQ and poly-HEMA. In the figure provided by Clancy et al, corresponding phase-contrast images and scale bars would have been helpful to better identify morphologic structures, especially for the abnormal nuclear morphology observed in Figure 1D, which appears quite different from the nuclear morphology observed in Figures 1A–C. Nevertheless, Clancy et al confirm our results by showing that during apoptosis lamin B1 is relocalized to the blebs but is not recognized by anti–lamin B1 antibodies under nonpermeabilized conditions.
Our finding that lamin B1 is not expressed at the surface of blebs is in striking contrast with the finding that other autoantigens, such as Ro and La, are expressed at the surface of blebs. As discussed by Clancy et al, the demonstration of antibody binding to the surface of apoptotic cells is a potentially critical link to the physiopathologic mechanisms applicable to anti-Ro and anti-La, but obviously this is not generalizable to all autoantigens such as lamin B1.
These data raise the important question of the predictability of autoantigen expression at the surface versus the interior of apoptotic blebs (i.e., why are certain autoantigens expressed at the surface while others are not?). Autoantigens such as Ro and actin (4) that are translocated at the external surface of the blebs are cytoplasmic. Even the La nuclear antigen is known to shuttle between nucleus and cytoplasm, playing a role in the biogenesis of RNA polymerase III transcripts and translation (5). Indeed, La is predominantly, but not exclusively, immunolocalized to the nucleus in nonapoptotic cells (6). Furthermore, the signal for La to reenter the nucleus, located in the La C-terminus, is cleaved during early apoptosis, causing La accumulation in the cytoplasm (7).
In contrast, nuclear lamin B1 is an inner nuclear membrane–anchored protein, a member of the intermediate filament family of proteins, and one of the major components of the nuclear lamina (for review, see ref. 8). In nonapoptotic cells, lamin B1 immunodetection is strictly nuclear, in contrast with the autoantigens mentioned above. During mitosis, the nuclear lamina filament network is disassembled and reassembled, and lamin B1 tends to remain associated with the nuclear membrane (9). During apoptosis, lamin B1 is cleaved by caspase 6 but remains associated with the inner nuclear membrane (10). In apoptotic blebs, lamin B1 is consistently colocalized with a nuclear-like membrane structure. Thus, we conclude that the only autoantigen sequestered within apoptotic blebs is the only one with a strict nuclear origin.
Therefore, we hypothesize that soluble autoantigens with a cytoplasmic or nonexclusive nuclear origin have the potential to be translocated to the apoptotic bleb surface. Thus, many autoantigens would be potentially accessible to their cognate circulating autoantibodies, and this interaction may contribute to pathogenesis. However, we further hypothesize that this mechanism is not universally applicable to all autoantigens, particularly insoluble autoantigens from a strict nuclear origin such as lamin B1.