Beyond necroptosis: alternative pathways in necrotic cell death

Regulation of the necrotic cell death is a highly debated issue and several questions are still open about mechanisms, genes and pathways involved. The aim of this PhD thesis was to explore the necrotic response by comparing two different stimuli: DMNQ, a generator of oxidative stress, and G5, an isopeptidases inhibitor and inducer of alterations in cell adhesion and of actin cytoskeleton. We took advantage of U87MG glioblastoma cells expressing Bcl-xL, as model of human cells that die through necrosis. Our results show that necrosis can be governed by completely different signaling pathways. DMNQ necrotic cell death relies on the RIP1 kinase activity, whereas G5-necrosis is RIP1-independent. Mitochondria play crucial role in energy metabolism and cell survival; furthermore alterations in mitochondrial bio-energetic activity and morphology have been observed during necrosis. Analysis of mitochondrial fragmentation in U87MG cells revealed that both necrotic insults triggered mitochondrial fragmentation. This fragmentation was not accompanied by MOMP, as verified by Smac release. G5 mitochondrial fragmentation occurred before the appearance of cytoplasmic vacuolization, thus representing an early marker of the necrotic response. To understand the role of mitochondrial fragmentation, we used Mdivi-1, an inhibitor of the mitochondrial fission protein Drp-1. Drp-1 inhibition reduced the necrotic death induced by DMNQ, but was unable to counteract the necrotic effect of G5. In fact, U87MG/Bcl-xL cells overexpressing the dominant negative mutant (K38A) of Drp1 showed that only DMNQ necrotic effect depends on Drp1 activity. Moreover, the inhibition of RIP1, using necrostatin-1, demonstrated that two distinct necrotic pathways were activated by DMNQ and G5. Indeed, Nec-1 efficiently rescued only DMNQ-induced cell death. To gain insight into necrotic effect elicited by G5, we analyzed different stress or pro-survival signaling pathways. p38, ERKs and Akt were transiently activated but only Akt dephosphorylation status was paired with the necrotic outcome. Next, we simultaneously evaluated PP2A, which can dephosphorylate and inactivate Akt. Interestingly, we showed an increase of its catalytic subunit amount and of its activity in response to G5. This led us to identify PP2A as a regulator of necrotic cell death. In fact, siRNA mediated down-regulation of PP2Ac reduced G5 cell death while increased DMNQ necrotic effect. Necrotic response induced by G5 provoked a re-localization of PP2Ac, with a reduced nuclear localization and increased cytoplasmatic shuttling. This was paired with the release of HMGB1 from the nucleus, a well-established necrotic marker. Finally, we investigated the link between the dramatic reorganization of the actin cytoskeleton induced by G5 and a possible role of PP2A in this context. For this purpose, we used U87MG/Bcl-xL cells overexpressing phospho mutants of cofilin, an actin de-polymerizing protein substrate of PP2A. Our results demonstrated that phosho-mimetic mutant (S3D) of cofilin slightly decreased G5 induced cell death, so defining a connection in this necrotic mechanism. In conclusion, these results describe alternative necrotic program beyond the actually well defined “programmed necrosis” (necroptoptic) pathways. DMNQ necrotic cell death is dependent on the canonical necroptotic elements while G5 relies on different players. We finally proved for the first time a connection between PP2A, actin cytoskeleton remodeling and necrotic outcome.