Plasma cells (PCs) are the terminally differentiated effector cells of the humoral immune system. The majority of PCs are short-lived and undergo programmed cell death in the form of apoptosis after a few days of intensive immunoglobulin secretion. Despite potentially wide-ranging implications for infection control, auto-immunity, and PC dyscrasias, the mechanisms that govern the initiation and execution of PC apoptosis are poorly understood. We used two well-established murine systems of PC differentiation and immunohistochemistry of human lymphoid tissue sections to study the regulation of PC apoptosis. IgM-secreting post-mitotic CD138+B220 PCs were differentiated in vitro from primary mouse splenic B cells using cytokines and LPS and purified by magnetic selection. Murine I.29mu+ B lymphoma cells were induced to undergo plasmacytic differentiation by stimulation with LPS. In both systems, terminal PC differentiation is followed by spontaneous apoptosis of half of the PCs within 48h, similar to PC apoptosis in vivo. We found that a sharp increase in endoplasmatic reticulum (ER) stress, which is caused by an imbalance between secretory load and capacity in the ER, occurs in PCs that have completed differentiation and begin to undergo apoptosis. In parallel, susceptibility specifically to ER stress-induced apoptosis but not to other apoptotic stimuli increases substantially in differentiated PCs, despite an ongoing ER stress response and expansion of the secretory machinery. Caspase-12, which has been linked specifically to ER stress-induced apoptosis, is activated and processed during programmed PC death. Using the specific inhibitor of caspase-12, zATADfmk, we found that caspase-12 mediates apoptotic DNA fragmentation and chromatin condensation in PCs undergoing apoptosis but not in B cells undergoing tunicamycin-induced apoptosis. In contrast, the major apoptotic effector caspases (caspase-9, caspase-3, caspase-7) downstream of the mitochondria become resistant to activation by apoptotic ER stress during terminal PC differentiation and are not activated during PC apoptosis. We observed that he pan-caspase inhibitor, zVADfmk, completely blocks tunicamycin-induced apoptosis in B cells but does not inhibit PC apoptosis or tunicamycin-induced cell death in PCs. Using the small molecule PAC-1, which specifically activates caspase-3 by targeting a “safety-catch” amino acid sequence that keeps caspase-3 inactive, we found that caspase-3 is stabilized in its inactive form in PCs and human myeloma cell lines, but not in B cells. Immunohistochemistry of human lymphoid tissue sections demonstrated that most primary reactive PCs and extramedullary myeloma cells undergo spontaneous apoptosis in vivo without activation of caspase-3. Thus, ER stress plays a major role in limiting the life span of short-lived PCs and activates caspase-12, which mediates nuclear apoptosis specifically in PCs. The major apoptotic effector caspases, however, become resistant to activation during terminal PC differentiation, and PC apoptosis is largely independent of caspases downstream of the mitochondria. These observations lead us to propose that developmentally regulated inhibition of key apoptotic caspases, which rapidly execute apoptosis in most cells, has evolved in PCs as a means to delay apoptosis under conditions of increasing ER stress linked to immunoglobulin secretion. Overwhelming ER stress ultimately limits the life span of short-lived PCs by inducing apoptosis using alternative mechanism involving caspase-12, which is redundant for the execution of ER stress-induced apoptosis in cells that can activate the classical effector caspases.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author