Does the cellular localization of antigens in or on apoptotic blebs influence the pathogenicity or benefit of cognate antibodies? Comment on the article by Dieudé et al

Authors


Does the cellular localization of antigens in or on apoptotic blebs influence the pathogenicity or benefit of cognate antibodies? Comment on the article by Dieudé et al

To the Editor:

Given the flood of attention directed toward understanding the affector arm in autoimmune diseases, studies that address the effector arm to study the pathogenicity of the perpetrator are of great interest. In certain circumstances the effects of an autoantibody may even be beneficial. This was in fact the situation being studied by Dieudé et al, who sought the mechanism by which autoantibodies to nuclear lamin B1 abrogate the strong prothrombotic risk associated with the lupus anticoagulant (1). The authors tested the hypothesis that circulating anti–lamin B1 antibodies block the procoagulant effect of apoptotic blebs by binding to lamin B1 displayed at the external surface of the blebs. However, by both biochemical and morphologic criteria, their extensive studies convincingly negated this notion. Lamin B1, in contrast to SSB/La, is not present on the surface of the apoptotic blebs but rather remains buried within the bleb, thereby being inaccessible to external anti–lamin B1 antibodies. Although these findings were considered negative in one context, they are indeed informative when considering another context.

As discussed by the authors, although apoptotic blebs serve as an important target for certain autoantibodies, it is now apparent that this is not generalizable to all autoantigens. With regard to the pathogenicity of anti-SSA/Ro and anti-SSB/La in the development of congenital heart block, unambiguous demonstration of antibody binding to the surface of apoptotic cardiomyocytes was a critical link that provided a plausible explanation of how antibodies were accessible to otherwise sequestered intracellular antigens (2). If intracellular trafficking of sequestered nuclear antigens to the membrane surface is not universally applicable to all such antigens, this may help explain the specificity of one antibody system versus another in the pathogenesis of disease, in particular tissue damage in the developing fetal heart.

In parallel with the studies presented by Dieudé et al (1), we have confirmed and extended the observation that lamin B1 is redistributed during apoptosis but, unlike SSA/Ro or SSB/La, is not bound by cognate antibodies. Induction of apoptosis included not only staurosporine, the protein kinase C inhibitor used by Dieudé et al, but also the redox cycling quinone, 2,3-dimethoxy-1,4-naphthoquinone (DMNQ). A third method of inducing apoptosis exploited culturing of the cells on poly-(2-hydroxyethyl methacrylate) (poly-HEMA), which results in apoptosis because cells are unable to adhere due to the loss of anchoring signals (3). The target cells were human cardiomyocytes, isolated and cultured from fetuses of gestational age 16–24 weeks. Cells were separately incubated with 0.5 μM staurosporine for 7 hours, 0.2 mM DMNQ for 7 hours, or plated on poly-HEMA for 12 hours. Cells were then incubated in the absence or presence of mouse anti–lamin B1 antibody (Chemicon, Temecula, CA) and a combination of affinity-purified human antibodies to 48-kd SSB/La, 52-kd SSA/Ro, and 60-kd SSA/Ro for 30 minutes at 37°C. The media were removed, and the cells were washed twice in phosphate buffered saline with calcium. Cells were fixed with 4% paraformaldehyde for 20 minutes at 22°C. In a second set of experiments the apoptotic cells were first permeabilized with 100% acetone for 60 minutes at 22°C prior to addition of the primary antibodies. The permeabilized and nonpermeabilized cells were subsequently incubated with fluorescein isothiocyanate–conjugated rabbit anti-mouse IgG (for localization of lamin B1) or CY3-conjugated goat anti-human IgG (for localization of SSA/Ro and SSB/La) for 30 minutes at 22°C. Matched isotype controls were negative.

In permeabilized nonapoptotic cardiomyocytes, lamin B1 and SSA/Ro–SSB/La antigens were localized to the nucleus with minor cytoplasmic staining, consistent with previous reports (1, 4, 5, and data not shown). With regard to the latter antigens, it is acknowledged that the topology observed cannot be ascribed uniquely to 52-kd or 60-kd SSA/Ro or to 48-kd SSB/La, given the affinity-purified antisera used for detection. After induction of apoptosis by all 3 methods, lamin B1 as well as SSA/Ro–SSB/La translocated to apoptotic blebs (Figure 1A and B). However, lamin B1, in contrast to SSA/Ro–SSB/La, was not detected at the cell surface (Figure 1D).

Figure 1.

Cellular topology of SSA/Ro–SSB/La and lamin B1 in permeabilized and nonpermeabilized apoptotic cultured human fetal cardiomyocytes. For all fields shown, cultured human fetal cardiomyocytes were incubated with staurosporine to induce apoptosis (round condensed nuclei; Hoechst stained [blue]). A–C, Cells were permeabilized prior to staining. A, Apoptotic blebs emerging from the surface were stained by a combination of affinity-purified human antibodies to 48-kd SSB/La, 52-kd SSA/Ro, or 60-kd SSA/Ro (red fluorescence). B, Similar results were observed in permeabilized apoptotic cells stained with anti–lamin B1 antibodies (green fluorescence). In this same field, anti–lamin B1 antibodies stained only the nucleus in the nonapoptotic cell. C, Neither the 3 apoptotic cells nor the single nonapoptotic cell stained with either normal human serum or mouse isotype control. D, Apoptotic cells were not permeabilized prior to staining. Strong red fluorescence indicates surface staining by the affinity-purified anti-SSA/Ro–anti-SSB/La antibodies. In contrast, there is no staining by the anti–lamin B1 antibody. Hoechst crosses intact membranes, accounting for the blue staining in these nonpermeabilized cells.

Taken together, our observations and those of Dieudé et al support discordance in the final cellular destination of translocated nuclear autoantigens during the process of apoptosis. In the case of lamin B1, physiologic noninflammatory clearance of apoptotic cells should proceed uneventfully even in the presence of circulating cognate antibodies. However, in congenital heart block, the maternal anti-SSA/Ro–anti-SSB/La antibodies result in opsonization and inflammatory/fibrotic sequelae. Even if it turns out that SSA/Ro–SSB/La are not unique in this regard, there may be other factors, such as complement binding of certain antigens or degradation of antigens, that facilitate clearing without further sequelae. Establishing the fact that at least one other nuclear autoantigen is not surface-bound during apoptosis of human fetal cardiomyocytes is a step forward.

Robert M. Clancy PhD*, Edward K. L. Chan PhD†, Sarayu Chandrashekhar BS*, Jill P. Buyon MD*, * Hospital for Joint Diseases and New York University School of Medicine New York, NY, † University of Florida Gainesville, FL.

Ancillary