ZFP36 stabilizes RIP1 via degradation of XIAP and cIAP2 thereby promoting ripoptosome assembly

Abstract

Background: ZFP36 is an mRNA binding protein that exerts anti-tumor activity in glioblastoma by triggering cell death, associated to an increase in the stability of the kinase RIP1.

Methods: We used cell death assays, size exclusion chromatography, Co-Immunoprecipitation, shRNA lentivectors and glioma neural stem cells to determine the effects of ZFP36 on the assembly of a death complex containing RIP1 and on the induction of necroptosis.

Results: Here we demonstrate that ZFP36 promotes the assembly of the death complex called Ripoptosome and induces RIP1-dependent death. This involves the depletion of the ubiquitine ligases cIAP2 and XIAP and leads to the association of RIP1 to caspase-8 and FADD. Moreover, we show that ZFP36 controls RIP1 levels in glioma neural stem cell lines.

Conclusions: We provide a molecular mechanism for the tumor suppressor role of ZFP36, and the first evidence for Ripoptosome assembly following ZFP36 expression. These findings suggest that ZFP36 plays an important role in RIP1-dependent cell death in conditions where IAPs are depleted.

Figure 1

ZFP36 controls RIP1 stability by depleting IAPs and promotes RIP1-dependent cell death. A). Transfection of HEK293T cells with a ZFP36 construct (ZFP36) correlates with the depletion of cIAP2 and with an increase in the stability of RIP1 at 48 hours. HEK293T cells transfected with a XIAP construct depleted of its own 3′UTR (Δ3′UTR XIAP) show a reduction in the levels of RIP1, even in the presence of ZFP36 (ZFP36 + Δ3′UTR XIAP). B) Image J software was used to normalize the levels of RIP1 over beta-ACTIN. C) Real Time PCR showing RIP1 mRNA levels following ZFP36 and Δ3′UTR XIAP over-expression. D) Both the transfection of ZFP36 and the treatment with Etoposide induce the formation of MLKL homo-oligomers and a decrease of MLKL monomers. E.V. cells were incubated with DMSO as this is the vehicle for Nec-1 administration. Non-reducing conditions were applied in order to preserve intact oligomers of MLKL. E) HEK293T transfected with the ZFP36 construct (ZFP36) show increased positivity to Propidium Iodide staining (P.I.), which is reduced in the presence of the Δ3′UTR XIAP construct (ZFP36+ Δ3′UTR XIAP). F) Percentage of P.I. positive cells obtained from 5 fields for each condition from experiment 1C. G-H) 48 hours treatment with 50 μM Necrostatin-1 inhibits ZFP36-dependent cell death (ZFP36 Nec-1) if compared to DMSO treated cells (ZFP36 vehicle), as assessed by cytofluorimetric analysis of Annexin V / P.I. or positivity to Trypan blue. The graphs in this figure represent the mean and SEM calculated on a set of three to four independent experiments (*p < 0.05; **p < 0.01).

Figure 2

ZFP36 promotes RIP1 assembly in a high molecular weight complex containing caspase8 and FADD. A, B) Following 24 hours of empty vector/ZFP36 transfection, HEK293T cells were further treated for 24 hours in the presence of 20 μM Z-VAD-FMK, in order to block caspases’ activity and facilitate the recovery of the Ripoptosome, as described in Tenev et al. [8]. A positive control for Ripoptosome assembly was added in the co-immunoprecipitation experiments depicted in figure B, where HEK293T were treated for 24 hours with 100 μM Etoposide and 20 μM Z-VAD-FMK. A) 2 milligrams of lysates from either empty vector (E.V.) or ZFP36 transfected cells (ZFP36) were loaded on a Sepharose column and fractions were eluted at defined timing and then loaded on polyacrylamide gel to perform Western blotting. In the presence of ZFP36, RIP1 was detected in higher molecular weight fractions (54–71), while this was less evident in E.V. samples. A positive control for RIP1 was loaded for each sample series (C+). Molecular weight markers were eluted through the Sepharose column in order to identify the 2MDa fractions. B) Ripoptosome recovery was carried out by immunoprecipitating caspase8, while immunoprecipitation with nonspecific mouse IgG was performed on the Etoposide (Eto.) sample as a technical control for the procedure. Both in the presence of ZFP36 or Etoposide (Eto.) RIP1 was associated with caspase8, while this was less evident in the empty vector population (E.V.). Higher levels of FADD were associated to caspase8 in the empty vector population, as opposite to the ZFP36 and Etoposide (Eto.) samples. Input controls are represented on the right panel. C, D) Image J software provided a relative quantification of the levels of RIP1 and FADD associated to caspase8.

Figure 3

ZFP36 controls RIP1 stability in glioma neural stem cells. Human G144 glioma neural stem cells were grown in adhesion in proliferation medium and transduced with a short hairpin lentivirus directed against ZFP36 (shZFP36) or with a scrambled short hairpin sequence (pGIPZ) and sorted based on eGFP-positivity. Western blot analysis suggests that RIP1 stability is reduced in the absence of ZFP36 (shZFP36) when compared to G144 that retain ZFP36 expression (pGIPZ).