Die later with ESCRT!

The consequences of necroptosis depend on immunomodulatory molecules, the expression of which requires time before the burst of a cell. Gong et al. now provide evidence for ESCRT-III-mediated plasma membrane repair to extend the time to death during necroptosis. Regulated cell death is not restricted to apoptosis, but includes several forms of regulated necrosis. The best characterized signaling pathway of regulated necrosis is necroptosis. Diverse signaling pathways, such as TNFR1-signaling, TLR-signaling and others may result in receptor-interacting protein kinase 3 (RIPK3)dependent phosphorylation and oligomerization of the pseudokinase mixed-lineage kinase domain-like (MLKL). It has been proposed that pMLKL may form pores in the plasma membrane, but the actual processes following plasma membrane translocation remain unclear [1]. In a recent report published in Cell, Gong et al. discovered the ESCRT-III-machinery as an active counterpart of pMLKLassociated membrane damage [2]. As necroptosis actively shapes the immune response [3], ESCRT-III indirectly controls the immunogenicity of necroptotically dying cells. Applying a system to artificially dimerize RIPK3 or MLKL, Gong et al. detected cells that rapidly expose phosphatidylserine (PS) at the outer leaflet of the plasma membrane, a cell death feature that had been associated with apoptosis for the last decades. Single cell analysis revealed the shedding of PS-positive “bubbles”. Unlike apoptotic bodies, these “bubbles” did not contain cytosolic remnants but consisted of broken membranes, as they are permeable to 10 kDa dextran-NH2. These bubbles formed at sites of pMLKL-accumulation, quite similar to what had been previously reported about viral budding in dependence of the ESCRT complex machinery [4]. Indeed, the ESCRT-III-protein CHMP4B co-localized with MLKL at the basis of “bubbles”. A previous study demonstrated ESCRT-mediated repair of laser-damaged membranes [4], but a link to necroptosis was not provided. In line with this, silencing of ESCRT-III-proteins CHMP2A or CHMP4B resulted in spontaneous necroptosis which was prevented by silencing of RIPK3 or MLKL, or by addition of RIPK1-inhibitor Nec-1s. Conversely, silencing of ESCRT-III-proteins CHMP4B, VPS4B, CHMP2A or ESCRT-I-proteins TSG101 or VPS37B sensitized cells to TNF-induced necroptosis even with active caspase-8 which appears to prevent necroptosis independently of this mechanism. Therefore, this study demonstrates the complex interplay between the ESCRT-III machinery and necroptosis execution, prompting the hypothesis of ESCRT-III to delay the time to plasma membrane rupture and release of damage associated molecular patterns (DAMPs). Indeed, even hardly detectable levels of pMLKL were sufficient to induce necroptosis if ESCRT-III was compromised. Along similar lines, active ESCRT-III delayed membrane breakdown resulting in less chemokine production and less efficient cross-priming of T cells as a consequence of a delay in DAMP release. MLKL was recently demonstrated to trigger processing and release of both anti-inflammatory cytokines, such as IL-33 and CXCL1, as well as the proinflammatory cytokine IL-1β [5]. It will be of interest to precisely unravel the proand anti-inflammatory components released by necroptotically dying cells, especially in comparison with cellular cytokines released during distinct pathways of regulated necrosis, such Editorial: Autophagy and Cell Death