Mammalian apoptosis in a parasitic worm

Apoptosis, or programmed cell death, is a process in multicellular organisms for the elimination of damaged and unneeded cells that prevents pathologic conditions and maintains tissue homeostasis. Although aspects were first described a century before, it was not until 1972 that Kerr et al. morphologically characterized the phenomenon and named it apoptosis (1) and another two decades later that Hengartner et al. revealed the molecular underpinnings of programmed cell death in the roundworm (nematoda) Caenorhabditis elegans (2). The discovery that the proto-oncogene bcl-2 (B-cell CLL/lymphoma 2) was homologous with one of the genes identified in C. elegans revealed a parallel but more complicated mammalian apoptosis pathway involving mitochondria (3). Now, in PNAS, Lee and colleagues show that a similar mitochondrial apoptosis pathway is present in the flatworm (platyhelminthes) Schistosoma (4).

In mammalian cells, the Bcl-2 family proteins are central regulators of apoptosis (5, 6). These proteins fall into two classes, multidomain proteins that are either pro- or antiapoptotic and that share similarity in three or four regions (referred to as Bcl-2 homology domains, BH1–4), and the proapoptotic BH3-only proteins that are homologous in only a single region. Proapoptotic multidomain proteins (Bax and Bak) act at the mitochondrion and are essential for the cell death process. Upon activation, they homo-oligomerize to release factors such as cytochrome c that then interact with Apaf-1 (apoptotic protease activating factor 1) to initiate a proteolytic (caspase) cascade that executes cell death through cleavage of cellular proteins (Fig. 1). Antiapoptotic multidomain family members (e.g., Bcl-2) inactivate proapoptotic proteins by binding their amphipathic α-helical BH3 region via a hydrophobic cleft (7). The diverse BH3-only proteins act as sentinels for cellular stress and initiate apoptosis by either directly activating Bax and Bak (8) or by disrupting the binding between proapoptotic and antiapoptotic multidomain family members (9). In contrast to the more complicated mammalian system, C. elegans lacks a multidomain proapoptotic protein and the single BH3-only protein Egl-1 frees the Apaf-1 counterpart Ced-4 from the antiapoptotic Bcl-2 homolog Ced-9 to initiate caspase (Ced-3) activation without the requirement of cytochrome c release from mitochondria (10).

Fig. 1.

Apoptotic pathways. (A) In C. elegans, BH3-only protein Egl1 displaces Ced4 from antiapoptotic Ced9 to activate caspase Ced3. In mammals, BH3-only proteins directly activate multidomain proapoptotic proteins Bax and Bak, displace Bax/Bak-activating BH3-only proteins from antiapoptotic Bcl-2 proteins, or displace Bax/Bak from antiapoptotic Bcl-2 proteins leading to activation. Activated Bax/Bak induces cytochrome c release from the mitochondrion that interacts with Apaf-1 to initiate caspase activity. (B) Comparison of C. elegans, mammalian, and schistosome Bcl-2 family proteins. BH1–4, Bcl-2 homology domains.

By analyzing recently published schistosome genomes (Schistosoma japonicum and Schistosoma mansoni), Lee et al. identify proteins that appear to indicate an intact apoptotic pathway. These include proteins with regions homologous to all four BH domains (sjA, sjB, smA, and smB), proteins incorporating an apparent BH3 domain with no other similarity to Bcl-2 family proteins (sjC and smC), and homologs of Apaf-1 and caspases. The key question is whether this system recapitulates that found in C. elegans or whether one or more of the multidomain proteins is in fact proapoptotic, which would imply mitochondrial involvement and similarity to the mammalian system.

Interrogating cell death is complicated by the fact many cellular perturbations will induce stresses that ultimately impinge on mitochondria, resulting in cytochrome c release and apoptosis. Cells from mice deficient in both Bax and Bak (Bax^−/−Bak^−/−) serve as a model system to elucidate the function of Bcl-2 family proteins because they are resistant to signals from upstream in the apoptotic cascade (11). Lee et al. use this strategy and turn to mammalian cells to show that sjC kills normal, but not Bax^−/−Bak^−/− cells, and that this activity is dependent on the BH3 region. Thus, sjC is proapoptotic and must act upstream of Bax and Bak, characteristics that are consistent with those of a BH3-only Bcl-2 family protein. In contrast, overexpression of sjB kills both normal and Bax^−/−Bak^−/− cells and sensitizes Bax^−/−Bak^−/− cells to a mammalian BH3 protein, suggesting a multidomain Bax/Bak-like protein. That sjA represents the antiapoptotic counterpart is confirmed by its ability to abrogate sjC- and sjB-mediated cell death.

Lee and colleagues provide compelling evidence that a Bcl-2 family-regulated mitochondrial cell death pathway exists in schistosomes, placing them evolutionally much closer to humans than C. elegans. Final confirmation will require evidence of these activities in schistosome cells, perhaps through similar overexpression or knockdown strategies. This discovery raises some interesting questions. First, what is the role of the additional multi-BH domain proteins that were identified? Redundancy exists in the mammalian system, and these may represent additional pro- or antiapoptotic proteins. If so, do selective interactions between schistosome Bcl-2 family members modulate responses to different stresses and does their relative importance differ by cell type as it does in mammalian cells? In the final stages of cell death following caspase activation in mammalian cells, endonucleases are activated to degrade chromosomal DNA and phosphatidylserine externalization signals for phagocytic clearance of apoptotic bodies (12). Do these or related processes occur in schistosomes? Finally, a second, receptor-mediated extrinsic apoptosis pathway is present in mammalian cells that can circumvent mitochondria (12). Is this pathway also a characteristic of schistosomes?

Schistosomiasis is a snail-transmitted parasitic disease that is endemic to tropical and subtropical locations (13). Affecting as many as 200 million people worldwide, chronic infection impacts multiple organ systems and is estimated to lead to the loss of 1.5 million disability-adjusted life years. The elucidation of the molecular underpinnings of apoptosis quickly led to the proposal to modulate these pathways to target diverse diseases (14), and today inhibitors of antiapoptotic Bcl-2 proteins have advanced to clinical investigation as potential oncology chemotherapeutics (15). The discovery of a mitochondrial apoptosis pathway in schistosomes therefore revealed a unique approach to confront this parasite. Encouragingly, the known Bcl-2 inhibitor ABT-737 (16) binds with moderate affinity to sjA, suggesting this is a feasible strategy.

Bcl-2 proteins are involved in normal tissue homeostasis (17), and hence selectivity for the schistosome protein will likely be required for any therapeutic agent. The observation that sjA has a different pattern of binding to BH3 proteins than mammalian Bcl-2 family members indicates a binding site that is likely sufficiently different for this selectivity to be achievable. Because expression of additional Bcl-2 proteins can render tumor cells resistant to inhibitors selective to a subset of the family (18, 19), the prospect of additional antiapoptotic schistosome proteins raises the question of whether a single target will prove to be an Achilles’ heel or whether inhibition of multiple proteins will ultimately be necessary. This issue suggests an additional opportunity for therapeutic intervention. Up-regulation of Bcl-2 proteins is a mechanism of resistance to toxic insults, and chemoresistance in cancer cells appears to be associated primarily with a single member of the Bcl-2 family (20). If failure of current anthelmintics such as praziquantel is mediated by a Bcl-2 protein in schistosomes, then these findings may pave the way to combination regimens for treatment of resistant disease.