Main

Cellular FLICE inhibitory proteins (c-FLIP) inhibit death receptor (DR)-mediated apoptosis, by preventing caspase-8 activation.1 Among the three identified c-FLIP splicing forms,2, 3 c-FLIPS,R were described as cytosolic, whereas c-FLIPL was also observed in the nucleus. A pool of membrane-bound c-FLIPL was also described4 suggesting that caspase-8/c-FLIPL could re-distribute on stimulation, leading to a more subtle regulation of caspase-8 activity depending on substrates localization.5 Furthermore, caspase-8 itself and Fas-Associated Death Domain adaptor protein (FADD) were found or were shown to re-loca5lize in local complexes on ER6, 7, 8 and mitochondria,9, 10 mediating the exchange of signals between the two organelles.11, 12, 13 Several molecular platforms containing both membrane-bound proteins and cytosolic apoptosis modulators have been identified at the ER-mitochondria interface (the so-called mitochondria-associated membranes or MAMs),14 controlling ER-mitochondria anchorage as well as lipid metabolism, Ca2+ signaling and apoptosis.15 MAMs have been recently described as lipid raft-like domains that orient proteins to promote the ER-mitochondria juxtaposition;16 consequently, alterations in their composition may profoundly affect the physical and functional inter-organelle crosstalk. Furthermore, as mitochondrial and ER membranes are continuously and concertedly remodeled,17 it is not surprising that membrane-shaping proteins can also exert a function in regulating the ER-mitochondria coupling.12, 18 Different families of ER-shaping proteins control the organization of peripheral ER, which consists of sheet-like cisternae and tubules connected by three-way junctions.19 Among these, Reticulons (RTN) and Deleted in Polyposis locus 1 (DP1) proteins cause the ER membrane to curve and tubulate,20, 21 whereas the GTPases Atlastins (ATL) promote the branching of ER tubules;22 finally, ER sheet-enriched proteins such as the 63-kDa cytoskeleton-linking membrane protein (CLIMP63) control the width of ER cisternae, anchoring the organelle to microtubules and maintaining its spatial distribution.23, 24 Along with other components of the extrinsic apoptosis, here we described for the first time the enrichment of c-FLIPL at ER and ER-mitochondria interface. Furthermore, we observed that ER structure and tethering to mitochondria are impaired in cells lacking c-FLIP. Given the importance of membrane-shaping proteins and MAM complexes in regulating organelles structure and ER-mitochondria juxtaposition, we focused on the mechanism underlying this phenotype and we found that c-FLIPL deficiency induces the caspase-mediated processing of RTN4, thus affecting organelle shape and coupling to mitochondria. We therefore concluded that c-FLIPL is a novel regulator of ER morphology and ER-mitochondria crosstalk.

Results

c-FLIPL localizes at ER and MAMs

We investigated c-FLIP subcellular localization by performing cell fractionation and indirect immunofluorescence. Percoll-purified fractions from the mouse liver, devoid of cross-contaminations as indicated by western blotting (WB) analysis of specific markers (Figure 1a), were assayed. In agreement with previous evidence,25, 26 c-FLIPS was found in the cytosol (Figure 1b). Conversely, c-FLIPL localized also at the ER+LM (Light Membranes) and MAMs fractions (Figure 1b). The pattern of distribution of calnexin (CNX) 27 and AKT,28 both previously identified at MAMs, further substantiated the specificity of c-FLIPL subcellular localization (Figures 1a and b). In homogenates from mouse embryonic fibroblasts (MEFs), c-FLIPL was similarly retrieved in both ER+LM and crude mitochondria still containing ER and MAMs as contaminants (Figure 1c). Confocal analysis corroborated c-FLIP distribution in MEFs. In the left panel of Figure 1d, yellow discrete areas represented the ER-mitochondria contact points. A similar pattern of co-localization was observed for c-FLIP and mitochondria (Figure 1d, central panel), whereas a wider overlapping was reported between the ER-protein calreticulin (CRT) and c-FLIP (Figure 1d, right panel), confirming c-FLIP distribution at the ER and ER-mitochondria contacts points. According to biochemical and morphological evidence, we concluded that ER and MAMs contain c-FLIPL, which is conversely absent from mitochondria devoid of the interacting ER.

Figure 1
figure 1

c-FLIPL is retrieved at ER and MAMs. (a, b) Proteins (20 μg) from Percoll-purified subcellular fractions of the mouse liver were separated by SDS–PAGE and immunoblotted with the indicated antibodies, specific for each fraction. The translocase of the mitochondrial outer membrane 20 (TOM20) was used as marker for mitochondria, the lactate dehydrogenase (LDH) as marker for cytosolic fraction. Calnexin (CNX) is considered as a marker of both ER and MAMs, whereas the long-chain fatty-acid CoA synthases (FACL4) is MAM-enriched. Distant parts from the same gel were joined at dashed lines. (c) Subcellular fractions and immunoblot analysis from WT MEFs. (d) Immunofluorescence and confocal analysis of endogenous c-FLIP subcellular localization. WT MEFs were transfected with pDsRed2-mito encoding for the mitochondrial-targeted fluorescent protein (mtRFP) to label mitochondria and immunostained for the endogenous ER marker calreticulin (CRT) and for c-FLIP. Left and central panels indicate co-localization of ER with mitochondria, and of mitochondria with c-FLIP. Right panel indicates co-localization of ER and c-FLIP. Boxes denote magnified areas. Scale bar, 20 μm

Full size image

Lack of c-FLIPL controls ER integrity and ER-mitochondria tethering

The subcellular distribution of c-FLIPL prompted us to investigate its involvement in the regulation of ER morphology and tethering to mitochondria. To this aim, WT and c-FLIP−/− MEFs29 transfected with mitochondria- and ER-targeted fluorescent proteins (respectively mtRFP and ERYFP) were used, and ER and mitochondrial morphology was investigated by comparing three-dimensional (3D) reconstructions of volume-rendered confocal stacks of mtRFP and ERYFP. c-FLIP depletion did not significantly alter mitochondrial morphology or distribution (Figure 2a). Conversely, whereas WT cells showed a well-organized network of interconnected tubules spanning the whole cytosol (Figure 2b, left panel), c-FLIP−/− cells displayed a deranged ER, characterized by fragmented and enlarged cisternae, seemingly to the detriment of reticular ER at the cell periphery (Figure 2b, central panel). Such defects were partially restored on c-FLIPL re-introduction in ablated MEFs (Figure 2b, right panel). To ensure that ER defects were not a result of transfections toxicity, immunofluorescence analysis of the endogenous ER-protein CRT (Figure 2c) and electron microscopy (EM) analysis were performed (Figure 2d), confirming the enlargement of ER cisternae in c-FLIP−/− cells with respect to the tubular ER in WT MEFs. Fluorescence recovery after photobleaching (FRAP) of ERYFP was finally used to quantitatively assess the ER integrity, as the speed with which fluorescence returns to the photobleached region is dependent on the degree of ER luminal contiguity.30 In accordance with morphological evidence, FRAP analysis further substantiated ER impairment and lack of integrity at the cell periphery of c-FLIP−/− cells (Figure 2e), which were partially restored by c-FLIPL re-introduction, hence corroborating c-FLIP role in this phenotype. Confirming the ER fragmentation mainly at the cell periphery, no major differences in ER luminal contiguity were observed in the perinuclear area (Figure 2f). In view of c-FLIPL localization at MAMs and of its involvement in ER shape, we investigated whether c-FLIP could have a role also in ER-mitochondria tethering. To address this point, we firstly measured the ability of crude mitochondria from WT and c-FLIP−/− MEFs to interact with microsomal components as CNX. As shown in Figure 2g, crude mitochondria from ablated MEFs were less contaminated by microsomes than those isolated from WT, indicating diminished ER-mitochondria interactions. Successively, we measured the ER-mitochondria distance by semi-quantitative confocal microscopy analysis of volume-rendered 3D reconstructions of z-axis stacks of confocal images of ERYFP and mtRFP.17, 18 Whereas WT MEFs showed several yellow areas of co-localization, representing the ER-mitochondria contact points, c-FLIP−/− cells revealed lower ER-mitochondria juxtaposition (Figure 2h), which was partially restored on c-FLIPL re-introduction. Quantitatively, Manders’ co-localization coefficient 31 confirmed these evidence, revealing ~15% reduction in the ER-to-mitochondria tethering in c-FLIP−/− MEFs, that was counterbalanced by c-FLIPL re-expression (Figure 2i). The occurrence of inter-organelle contacts in WT and c-FLIP−/− MEFs was also verified by EM analysis (Figure 2d).

Figure 2
figure 2

c-FLIPL depletion affects ER morphology and tethering to mitochondria. (a) Left: Representative 3D-reconstructions of mitochondria in MEFs of the indicated genotype transfected with mtRFP. Right: mean±S.E.M. (n=4, 25 cells per experiment) of data from a. (b) Representative 3D-reconstructions of ER in MEFs of the indicated genotype co-transfected with ERYFP and with plasmids encoding for c-FLIPL or for the empty vector. ‘c-FLIP’ indicates c-FLIPL re-introduction in depleted MEF. Scale bar, 25 μm. (c) Immunofluorescence analysis of the resident ER protein CRT in WT and c-FLIP−/− MEFs. (d) Representative EM images of ER and ER-mitochondria contact points in WT and c-FLIP−/− cells. Scale bar, 0.54 μm in the upper, right panel; 0.35 μm in other panels. Arrows indicate ER-mitochondria proximity. (e, f) Fluorescence recordings from real-time sequence of ERYFP FRAP in c-FLIP−/− and WT MEFs at the cell periphery (e) and in the perinuclear region (f). The histogram reported mean±S.E.M. (n=5, 10 cells per experiment) of data from e (**P≤0.001; ***P≤0.0001). In e, f, 16-μm2 regions were photobleached for a total of 3 s, using 100% laser power of the 488 nm line. The post-bleaching images were taken for a total of 120 s. ‘c-FLIP’ indicates c-FLIPL re-introduction in depleted MEF, ‘EV’ the empty vector. (g) WB of the indicated proteins in crude mitochondrial fractions (20 μg) from WT and c-FLIP−/− MEFs. (h) Representative 3D-reconstructions of ER and mitochondria in MEFs of the indicated genotype co-transfected with mtRFP, ERYFP and the reported plasmids. Yellow areas indicate that organelles are closer than 270 nm. Scale bar, 25 μm. (i) Mean±S.E.M. (n=5, 10 cells per experiment) of interaction data from h (***P≤0.0001)

Full size image

ER-shaping protein Reticulon-4 is reduced in c-FLIP ablated cells

Defective ER morphology indicated that c-FLIPL could have a role in maintaining the ER structure. To characterize the molecular mechanism involved, we first compared the expression levels of different ER proteins to discern whether ER expansion could be related to excessive ER membrane proliferation. However, we did not detect significant differences between WT and c-FLIP−/− cells in any of the tested ER-resident proteins (Figure 3a). We further compared MFN2 levels, as its involvement in ER-mitochondria tethering and morphology was previously reported.18 However, we did not notice major differences either in MFN2 levels between WT and c-FLIP−/− MEFs, or in c-FLIPL between WT and MFN2−/− cells, suggesting a MFN2-independent mechanism (Figure 3b). As aberrant ER enlargement observed in c-FLIP−/− cells was similar to ER defects associated with RTN4 depletion,20 or to overexpression of ATLs 32 or CLIMP63,23 we compared the expression levels of these ER-shaping proteins in microsomal fractions from WT and c-FLIP−/− MEFs (Figures 3c and e). Interestingly, c-FLIP-depleted cells displayed diminished expression of RTN-4A,B2,1 isoforms (Figure 3c), whereas no significant variations were observed in CLIMP63 or in ATL1,2 (Figures 3d and e). Only a slight, but not statistically significant, reduction in ATL3 was noticed (Figure 3e), indicating the involvement of RTN4 in ER alterations affecting c-FLIP−/− MEFs. Total lysates from the two cell lines, besides evidencing the differences in RTN-4A,B2,1, also revealed additional lower-molecular weight bands, compatible with cleaved fragments of this protein and visible predominantly in ablated MEFs (Figure 3f). Conversely, no extra bands of CLIMP63 were observed (Figure 3g); therefore, proposing that cleavage events could affect the RTN4 integrity in c-FLIP−/− MEFs, in turn the disrupted ER morphology.

Figure 3
figure 3

Evaluation of different ER markers in WT and c-FLIP-ablated cells. (a) Total lysates (50 μg) from WT and c-FLIP−/− MEFs were separated by SDS–PAGE and immunoblotted for different ER markers. (b) Upper panel: representative WB analysis of MFN2 level in total lysates from WT and c-FLIP−/− MEFs (n=4). Lower panel: representative WB analysis of c-FLIPL level in total lysates from WT and MFN2−/− MEFs (n=3). (c, d) Representative WB analysis of expression levels of RTN-4A and -B2,1 (c) and CLIMP63 (d) in ER fractions isolated from WT and c-FLIP−/− MEFs. The same amount of proteins was used and protein loading was controlled by evaluating the expression of the ER-resident protein BAP31. Histograms indicate the mean±S.E.M. of full length, ER+LM-associated RTNs densitometric analysis (n=3 for RTN-4A; n=3 for RTN-4B2,1). (e) WB analysis of expression levels of ATL1,2,3 in ER fractions. BAP31 was used as loading control. The densitometric analysis reported in the histogram, represented the mean±S.E.M. (n=3) of ATL3 level. (f, g) WB analysis of RTN-4A,B2,1 (f) and CLIMP63 (g) levels in total lysates from c-FLIP−/− and WT MEFs. Histograms indicate the mean±S.E.M. of full-length RTN densitometric analysis (n=3 for RTN-4A; n=6 for RTN-4B2,1) In c, d, f broken lines indicate where a lane was deleted

Full size image

TNFα induces RTN4 processing

Previous evidence indicated that caspases and their inhibitors could impinge on levels of endogenous ER morphogenic proteins. In particular, caspase-7 was reported to mediate the cleavage of RTN-4B.33 Owing to the ability of c-FLIP to control the apical caspase-8, which activates the caspase-7, we investigated whether increased caspase activation has occurred in cells lacking c-FLIP, in turn accounting for RTN4 processing. In agreement with this hypothesis, the basal caspase-8 activity was enhanced in c-FLIP−/− MEFs (Figure 4a). Successively, to better characterize the involvement of caspase-8/c-FLIP axis and extrinsic apoptotic response in RTN4 cleavage and ER alterations, we exposed c-FLIP−/− cells to increasing doses of TNFα, known to strongly induce the extrinsic pathway of apoptosis through caspase activation.29 As shown, we observed a rapid, but defective cleavage of RTN4 (Figure 4b, upper panel: RTN-4A; lower panel: RTN-4B2,1), decreasing at higher doses of TNFα despite the dose-dependent increase in caspase-8 activity (Figure 4c) and cleavage of caspase-3 and poly-(ADP-ribose) polymerase (PARP) (Figure 4d), which suggested an efficient activation of caspase-mediated pathways. As non-neural cells as MEFs abundantly express RTN-4B2,1 as compared with the little amount of RTN-4A, evaluation of RTN-4B2,1 was chosen for the following experiments. As reduction of RTN-4B2,1 cleavage at increasing TNFα doses (Figure 4b) or at longer time points (Figures 4e and f) occurred concomitantly to the formation of a higher-weight form of RTN-4B2,1 (indicated by the asterisk in the less-exposed lane) compatible with the phosphorylated form of the protein, we hypothesized that in our model RTN-4B2,1 processing could be prevented by an unidentified TNFα-induced phosphorylation event. Interestingly, cyclin-dependent kinases 1/2 (CDK1/2)-mediated phosphorylation of RTN-4B was reported to negatively control RTN4 processing.33 In accordance with our hypothesis, co-treatment of c-FLIP−/− MEFs with TNFα and the CDK1/2 pharmacological inhibitor JNJ-7706621 lowered the accumulation of the higher-weight form of RTN-4B2,1 and increased the formation of cleavage fragments with respect to c-FLIP−/− cells treated with TNFα alone (Figure 4g). We also observed mild RTN-4B2,1 cleavage in WT cells on TNFα treatment. Such cleavage increased when TNFα was combined with the de novo protein synthesis inhibitor cycloheximide (data not shown). Together our data highlight a role for TNFα-induced caspase activation in RTN4 cleavage, in c-FLIP-deficient cells as well as in their WT counterpart.

Figure 4
figure 4

Caspase activation impacts on RTN4 levels in c-FLIP−/− MEFs. (a) Basal caspase-8 activity was measured in c-FLIP−/− and WT MEFs. Mean±S.E.M. (n=7) (*P≤0.05). (b) Evaluation of RTN-4A and -B2,1 cleavage by WB analysis in c-FLIP−/− MEFs treated for 2 h with the indicated doses of TNFα. Histograms indicate the mean±S.E.M. of RTN4 fragments densitometric analysis (n=3 for RTN-4A in the upper panel; n=3 for RTN-4B2,1 in the lower panel). (c) Colorimetric evaluation of caspase-8 activity on treatment of c-FLIP−/− cells with the indicated doses of TNFα. (d) WB analysis of caspase-3 and PARP cleavage in c-FLIP−/− MEFs treated with the indicated doses of TNFα. (e, f) Representative WB analysis of RTN-4B2,1 cleavage in c-FLIP−/− MEFs treated, respectively, with TNFα 0,1 ng/ml (e) and 10 ng/ml (f) at the indicated time points. In b, d–f, β-actin was used to normalize. (g) Evaluation of RTN-4B2,1 cleavage by WB analysis in c-FLIP−/− MEFs treated with TNFα 10 ng/ml alone or in combination with JNJ-7706621 15 μM

Full size image

A caspase-mediated pathway is involved in RTN4 alterations and influences ER morphology and tethering to mitochondria

As already reported in other experimental models,34, 35 we verified that c-FLIP ablation in MEFs enhanced the basal caspase-8 activity (Figure 4a), which was restored by both the specific caspase-8 inhibitor zIETD and c-FLIPL re-introduction (Figure 5a). To confirm the involvement of caspases in TNFα-mediated RTN4 processing in c-FLIP−/− cells, zIETD and the pan-caspases inhibitors zVAD were used. Although both inhibitors were efficient in reducing the basal RTN4 proteolysis, even though to a different extent (Figure 5b), zVAD efficaciously diminished also the TNFα-induced RTN-4B2,1 processing (Figure 5c). This evidence further corroborated the role of caspases and components of DR-mediated pathway in controlling the RTN4 cleavage. In view of the essential role of reticulons in ER tubulogenesis, as well as of the involvement of caspases in RTN4 regulation, we wondered whether the altered ER morphology of c-FLIP−/− cells could be associated with the caspase-dependent RTN4 cleavage. To test this hypothesis, ER morphology and ER-mitochondria tethering of c-FLIP−/− MEFs pre-treated with zVAD were compared with those of WT and c-FLIP−/− cells exposed to the vehicle DMSO. Strengthening the proposed model, semi-quantitative confocal microscopy analysis of volume-rendered 3D reconstructions of z-axis stacks of confocal images of WT and c-FLIP−/− cells transfected with ERYFP revealed that caspase-inhibition was sufficient to rescue the ER morphology of c-FLIP−/− MEFs (Figure 5d). Moreover, as estimated by measuring the quantitative Manders’ co-localization coefficient, zVAD was also efficient in restoring the ER-mitochondria distance of c-FLIP−/− MEFs to levels comparable to those of WT (Figure 5e and f), hence underlying the potential role of caspases in regulating membrane-localized proteins and ER-mitochondria interactions.

Figure 5
figure 5

Caspase-inhibition rescues RTN4 processing and ameliorates ER morphology and tethering to mitochondria. (a) Caspase-8 activity was measured in c-FLIP−/− MEFs on caspase-8 inhibition with zIETD and c-FLIPL re-introduction. Mean±S.E.M. (n=3 experiments) (**P≤0.001; ***P≤0.0001). (b) WB evaluation of basal RTN4 cleavage, on pharmacological caspase inhibition with zIETD or zVAD. Broken lines indicate that we joined together distant parts from the same gel. The histogram represents the mean±S.E.M. of RTN-4B2,1 fragments densitometric analysis (n=3). (c) WB analysis of RTN4 fragments formation in c-FLIP−/− cells on treatment with the indicated dose of TNFα, alone or in combination with the inhibitor zVAD. (d) Representative 3D-reconstructions of ER in MEFs of the indicated genotype transfected with ERYFP, and treated with DMSO or zVAD (40 μM). Scale bar, 20 μm. (e) Representative 3D-reconstructions of ER and mitochondria in MEFs of the indicated genotype co-transfected with mtRFP, ERYFP and treated with DMSO or zVAD (40 μM). Scale bar, 20 μm. (f) Mean±S.E.M. (n=5, 10 cells per experiment) of interaction data from e (***P≤0.0001). In b, c β-actin was used to normalize

Full size image

c-FLIP ablation results in disturbed Ca2+ signaling

Integrity of ER and ER-mitochondria contact points have been demonstrated to be crucial for Ca2+ signaling.36 In view of the role of c-FLIPL at the ER and MAMs, we investigated its involvement in Ca2+ exchange between the two organelles. We firstly compared the increase in cytosolic Ca2+ concentration ([Ca2+]c) in Fura-2AM-loaded WT and c-FLIP−/− MEFs on treatment with the sarcoplasmic-endoplasmic reticulum Ca2+ ATPase (SERCA) inhibitor thapsigargin (TG), causing the passive ER discharge, and with ATP which induces ER Ca2+ release from inositol 1,4,5-trisphosphate (IP3)-sensitive stores. EGTA was added to all solutions to assess only intracellular calcium fluxes. c-FLIP−/− MEFs displayed severely blunted ER Ca2+ release in response to both TG and ATP (Figures 6a and b). In addition, as [Ca2+]c is buffered by mitochondria, along with the blunted ER Ca2+ release, in Rhod-2AM loaded c-FLIP−/− cells also the mitochondrial Ca2+ peak evoked by ATP was significantly lower than that in WT (Figure 6c). Nevertheless, such defects were not restored by c-FLIPL re-introduction, as we did not observe any amelioration in either ER Ca2+ release (Figure 6d) or mitochondrial Ca2+ entry (Figure 6e) in c-FLIP−/− cells re-expressing c-FLIPL with respect to c-FLIP−/− MEFs transfected with the EV. According to our data, c-FLIP−/− MEFs displayed reduced ER-mitochondria juxtaposition and diminished Ca2+ exchange. We therefore wondered whether the reduced ability of ER and mitochondria to exchange signals and calcium could impact on Ca2+-mediated apoptosis. In agreement with this hypothesis, cells lacking c-FLIP were more resistant than WT to hydrogen peroxide (H2O2), which is known to induce a Ca2+-dependent apoptotic response (Figure 6f).37 Conversely, c-FLIP−/− MEFs were as susceptible as WT to staurosporine (STS), which activates the mitochondrial pathway, but much more sensitive to DR-stimulation induced by TNFα, as previously reported (Figure 6g). Altogether our data demonstrate that c-FLIP ablation loosens ER-mitochondria juxtaposition, lowering Ca2+ transfer to mitochondria and impacting on sensitivity to Ca2+-dependent apoptosis.

Figure 6
figure 6

c-FLIP absence alters Ca2+ homeostasis and Ca2+-dependent apoptosis. (a) Fura-2AM recordings of cytosolic Ca2+ ([Ca2+]c) after passive discharge of ER Ca2+ stores induced by 1 μM TG in MEFs of the indicated genotype. Upper panel: maximum Ca2+ concentrations in bar charts. Mean±S.E.M. of n=70–100 cells. Lower panel: changes in Ca2+ concentrations shown as representative traces. (b) Fura-2AM recordings of cytosolic Ca2+ ([Ca2+]c) after active ER Ca2+ release induced by 100 mM ATP in MEFs of the indicated genotype. Upper panel: maximum Ca2+ concentrations in bar charts. Mean±S.E.M. of n=70–100 cells. Lower panel: changes in Ca2+ concentrations shown as representative traces. (c) Rhod-2AM recordings of mitochondrial Ca2+ entry (F1/F0) after active ER-to-cytosol Ca2+ release induced by 100 mM ATP in MEFs of the indicated genotype. Upper panel: maximum Ca2+ concentrations in bar charts. Mean±S.E.M. of n=70–100 cells. Lower panel: changes in Ca2+ concentrations. In a–c *P≤0.05; **P≤0.001; ***P≤0.0001. (d) Fura-2AM recordings of cytosolic Ca2+ ([Ca2+]c) after active ER Ca2+ release induced by 100 mM ATP in c-FLIP−/− MEFs transfected with the indicated plasmids. (e) Rhod-2AM recordings of mitochondrial Ca2+ entry (F1/F0) after active ER-to-cytosol Ca2+ release induced by 100 mM ATP in c-FLIP−/− MEFs transfected with the indicated plasmids. (f) Representative WB analysis of total lysates (40 μg) from MEFs of the indicated genotype, on treatment with H2O2 1 mM for 3 h. Apoptosis was investigated evaluating cleavage of PARP and caspase-3. (g) Representative WB analysis of total lysates (40 μg) from MEFs of the indicated genotype, on treatment with STS (2 μM) for 3h and TNFα 10 ng/ml for 2 h. Apoptosis was investigated evaluating cleavage of PARP and caspase-3. (n=5 for H2O2 treatment; n=3 for STS and TNFα treatments)

Full size image

Calcium chelation in WT MEFs induces ER changes

TG-induced ER Ca2+ release is an indirect measure of ER Ca2+ content at the steady state. According to our data suggesting diminished Ca2+ levels and Ca2+ exchange in cells lacking c-FLIP, we speculated that these differences could directly contribute to the observed ER defects. We therefore tested whether the chelator of free, cytosolic calcium 1,2-Bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid tetrakis(acetoxymethyl ester) (BAPTA-AM) could modify the overall ER structure in WT and c-FLIP−/− MEF. Unexpectedly, BAPTA-AM affected ER shape and RTN-4B1,2 integrity of the two cell lines. These effects were more evident in WT MEFs (Figure 7a, left panels), leading to ER enlargement and fragmentation and resembling ER defects of c-FLIP−/− MEFs, which were only marginally affected by Ca2+ chelation (Figure 7a, right panels). WB analysis further revealed that BAPTA-AM caused RTN-4B1,2 processing in WT cells, augmenting fragments formation in also c-FLIP−/− cells. Despite these effects, BAPTA-AM did not modulate c-FLIPL levels in WT cells, nor induced PARP cleavage (Figure 7b), suggesting a caspase-independent mechanism. Concordantly, whereas zVAD efficaciously diminished the basal RTN-4B1,2 cleavage in c-FLIP−/− MEFs (Figure 7c), it did not abrogate the BAPTA-AM-induced RTN4 processing in either of the two cell lines (respectively, in Figures 7c and d). These data suggest that, in c-FLIP−/− cells, ER effects and RTN4 processing could be also directly associated with disturbed calcium levels, besides to the caspase-dependent pathway.

Figure 7
figure 7

Chelation of cytosolic calcium in WT MEFs modifies ER shape and RTN4 integrity. (a) Immunofluorescence analysis and representative 2D-reconstructions of the resident ER protein CRT in WT and c-FLIP−/− MEFs treated with BAPTA-AM 5 μM or with the vehicle DMSO. Scale bar, 25 μm. (b) Evaluation of RTN-4B2,1 by WB analysis in WT and c-FLIP−/− MEFs treated for 3 h with BAPTA-AM 5 μM or with the vehicle DMSO as indicated. (c, d) WB analysis of RTN-4B2,1 fragments, respectively, in c-FLIP−/− (c) and WT (d) MEFs treated for 3 h with BAPTA-AM 5 μM alone or in combination with the pan-caspase inhibitor zVAD.

Full size image

Discussion

Components of DR signaling as caspase-8 have been identified on ER and mitochondria, suggesting that target specificity may depend on cellular compartments.5 In this study, we report that a fraction of the caspase-8 inhibitor c-FLIPL is enriched at the ER and MAMs. Along with other MAM-proteins, whose ablation affects ER and mitochondrial morphology or their coupling,18, 37, 38 here we demonstrate that c-FLIP deficiency in MEFs impacts on ER organization, causing enlargement of ER sheets to the detriment of tubular structures. The relative amount of the ER-shaping proteins RTN/DP1 and CLIMP63 was described to influence the cell-type-specific ratio of sheets to tubules in mammalian cells.39 Jozsef et al.40 demonstrated that RTN4 deficiency in MEFs is sufficient to convert ER tubules to cisternal structures, without affecting other ER-shaping factors or ER-resident proteins.40 Interestingly, we found decreased expression of RTN-4A,B2,1 in c-FLIP−/− MEFs, whereas no major differences were observed in other ER-shaping proteins or in ER proteins, confirming that ER defects of c-FLIP−/− MEFs were not related to variations in the overall ER content or to excessive CLIMP63 expression or ATL-mediated fusion.32 As caspase-7 was formerly reported to process the RTN-4B on its cytoplasmic tail on apoptotic induction,33 we hypothesized that ER defects observed in c-FLIP−/− cells could be associated with a deregulated, caspase-dependent processing of RTN4 occurring in cells lacking the inhibitor of caspase-8. In agreement with our hypothesis, here we demonstrate that c-FLIP−/− MEFs show an increased basal RTN4 processing, and that TNFα is sufficient to accelerate its cleavage. Interestingly, despite TNFα promptly induced a dose-dependent activation of caspase-8 and caspase-3, we observed only an initial cleavage of RTN4, decreasing at higher doses of TNFα or at longer time points, concomitantly to the accumulation of an higher-weight form of RTN-4B2,1. As Cyclin-dependent kinases 1/2 (CDK1/2) phosphorylate RTN4, abrogating its cleavage and inducing the accumulation of an higher-weight phosphorylated form, we hypothesized that the incomplete RTN4 processing could be associated with a CDK-mediated phosphorylation event. In accordance with previous evidence, c-FLIP−/− MEFs exposed to TNFα in combination with the CDK1/2 inhibitor JNJ-7706621 showed increased RTN-4B fragments formation and disappearance of the higher-weight form. Consistent with the reported role of caspases in RTN4 processing, we confirmed that, RTN4 cleavage was caspase-dependent, as in c-FLIP−/− MEFs the pan-caspase inhibitor zVAD effectively decreased basal and TNFα-induced RTN4 proteolysis. Interestingly, along with blocking basal RTN4 processing, zVAD ameliorated ER morphology of c-FLIP−/− MEFs, strengthening the nexus between caspases and mechanisms of ER shaping. As MAM complexes control ER-mitochondria juxtaposition either by linking the two organelles,18 or by affecting mitochondrial and ER shape,12 we successively investigated whether caspases and c-FLIP could exert a role in the maintenance of ER mitochondria distance in c-FLIP−/− cells. In agreement with our hypothesis, we demonstrate that in, c-FLIP−/− MEFs, rescue of a proper ER organization induced by both c-FLIPL re-introduction and zVAD also restores the ER-mitochondria juxtaposition to levels comparable to those of WT. Impaired ER and mitochondrial morphology and tethering ultimately affect the ER-to-mitochondria Ca2+ transfer. We therefore focused also on c-FLIP involvement in Ca2+ homeostasis and found decreased Ca2+ exchange and deregulated sensitivity to Ca2+ dependent apoptosis in c-FLIP−/− cells. As already demonstrated, we confirmed that c-FLIP depletion considerably sensitized c-FLIP-deficient cells to extrinsic apoptosis induced by TNFα, but only marginally influenced the apoptotic response to STS, which stimulates the mitochondrial pathway.29 Here we also demonstrate that c-FLIP deficiency protects against H2O2-mediated cell death. As H2O2 requires ER-mitochondria Ca2+ transfer,37 this result correlates with the impaired inter-organellar Ca2+ exchange. Despite c-FLIP−/− MEFs showed lower IP3- and TG-induced ER-to-cytosol Ca2+ release and mitochondrial Ca2+ entry than WT counterpart, such defects were not rescued by c-FLIPL re-introduction, suggesting possible additional mechanisms not directly relying on c-FLIP expression. To address this point, we evaluated whether ER derangement observed in c-FLIP−/− MEFs could also be associated with the deregulated Ca2+ signaling in a c-FLIP-independent way. Ferrari et al.41 previously demonstrated that on apoptosis induction, ER remodelling was not consequent to caspases activation, but controlled by chelation of Ca2+. Surprisingly, in our experimental model, the Ca2+ chelator BAPTA-AM induced ER enlargement, concomitantly to RTN-4B2,1 cleavage in WT cells, and worsened RTN-4B2,1 fragmentation in c-FLIP−/− MEFs. These effects were more pronounced in WT than in c-FLIP−/− cells, probably as a consequence of the different Ca2+ content of the two cell lines. To better investigate whether modulation of Ca2+ levels could modify the ER organization independently of c-FLIP and caspases, we evaluated c-FLIPL levels and PARP cleavage as a measure of caspase activation. Interestingly, BAPTA-AM neither modulated c-FLIPL in WT MEFs, nor induced RTN-4B2,1 cleavage via caspases. Nevertheless, although caspase activation was not described as necessary to ER remodeling during apoptosis,41 we cannot exclude that the increased caspase-8 activity observed in c-FLIP−/− cells at basal conditions could impinge on RTN4 cleavage and ER integrity, in turn emphasizing the role of caspase-mediated processing of RTN4. To summarize, in addition to showing c-FLIPL at the ER and MAMs, in this paper we demonstrate its relevance in modulating ER morphology and tethering to mitochondria. We also provide evidence of a new role for caspases in ER structure modulation via processing of RTN4. Although the importance of RTNs transmembrane domains in ER tubulation was firstly described,20, 21 modifications in the intrinsically unstructured, cytosolic regions of RTN4 42 have been reported to influence the variable functions of RTNs or specific interactions.20, 40 We can therefore speculate that caspases, acting on exposed regions of RTN4, might disturb the localization43 or the ability of ER-shaping proteins to form complexes and interact with cytoskeletal components or organelles.44 Concluding, the present study highlights additional non-apoptotic roles for the death-receptor pathway, a frame in which c-FLIPL and caspases emerge as new factors in the maintenance of ER homeostasis.

Materials and Methods

Cell cultures and reagents

Wild type (WT) and c-FLIP−/− MEFs were a generous gift of Tak W. Mak (Amgen Institute, Toronto, Canada). MEFs were cultured in DMEM, enriched with 10% fetal bovine serum, glutamine (2 mmol/l), sodium pyruvate (1 mmol/l) and non-essential amino acids, in the presence of penicillin (100 U/ml) and streptomycin (100 μg/ml). The Insolution Caspase-8 inhibitor II, indicated as Z-IE(OMe)TD(OMe)-FMK (zIETD), was purchased from Calbiochem (San Diego, CA, USA). Cells were treated for 24 h with zIETD 40 μM or with the vehicle DMSO. The N-Benzyloxycarbonyl-Val-Ala-Asp(O-Me) fluoromethyl ketone (zVAD-FMK) was purchased from Sigma-Aldrich (St Louis, MO, USA). Cells were treated for 24 h with zVAD 40 μM or with the vehicle DMSO. Where indicated, JNJ-7706621 was added to cell cultures. After 1 h pre-treatment, TNFα was added. JNJ-7706621 was synthesized by the Cancer Therapeutics Research Team (Johnson & Johnson Pharmaceutical Research and Development, LLC). Transfections with pDsRed2-Mito (Clontech laboratories INC—Palo Alto, CA, USA), pEGFP-ER (Clontech) and pBluescript or c-FLIPL plasmid (kindly provided by M. Peter (The Ben May Institute for Cancer Research, University of Chicago, Chicago, IL USA) were performed using Transfectin reagent (BioRad, Hercules, CA, USA) at 2 μg/μl, or Lipofectamin Reagent (Invitrogen, San Giuliano Milanese, Italy) at 4 μg/μl, according to the manufacturer’s instructions. Where indicated, BAPTA-AM (Sigma-Aldrich) 5 μM or the vehicle DMSO was added to cell cultures for 3 h at 37 °C and ER morphology and WB analysis of RTN4 were performed.

Subcellular fractionation

Whole mouse liver from adult CD1 mice or WT and c-FLIP−/− MEFs obtained from four, confluent, 500 cm2 petri dishes was washed twice with PBS, re-suspended in isolation buffer (IB, 200 mM sucrose, 1 mM EGTA-Tris, and 10 mM Tris-MOPS, pH 7.4), and then broken up by dounce homogenization. Purified fractions were obtained according to Frezza et al.45 Protein concentration was determined by micro BCA assay (Pierce Rockford, USA) and 20 μg of protein extract was separated by SDS–PAGE and immunoblotted for the indicated antibodies. All animal procedures were conducted in compliance with the Italian Ministry for Health guidelines.

Immunoblotting

Cells were collected and lysed in lysis buffer (Cell Signaling Technology, Danvers, MA, USA) in the presence of complete protease-inhibitor mixture (Sigma-Aldrich). Protein concentration was determined by micro BCA assay (Pierce, Rockfod, IL, USA) and the indicated proteins were separated by SDS–PAGE and transferred onto polyvinylidene difluoride (PVDF, BioRad) or nitrocellulose membranes (Amersham Bioscience, Piscataway, NJ, USA). Membranes were probed using the following antibodies: anti-β-actin (1:10 000, Sigma-Aldrich); anti-calnexin (1:2000, Abcam, Cambridge, UK); anti-Tom20 (1:10000, Santa Cruz Biotechnology, Santa Cruz, CA, USA); anti- lactate dehydrogenase (LDH) (1:2000, Rockland); anti-long-chain fatty-acid CoA synthases (FACL4) (1:500, Santa Cruz Biotechnology); anti-calreticulin (CRT) (1:1000, Abcam); anti-c-FLIP Dave II clone (1:800, Alexis, Lausen, Switzerland); anti-RTN4 (1:1000, Abcam); anti-CLIMP63 (1:1000, Proteintech Group, Chicago, IL, USA); anti-cleaved caspase-3 (1:500, Cell Signaling Technology) and anti-PARP (1:1000; Cell Signaling Technology).

Imaging

For confocal imaging and analysis of the morphology and tethering of ER and mitochondria, cells seeded on 24-mm round glass coverslips and incubated in Hanks balanced salt solution (HBSS), supplemented with Hepes 10 mM, were placed on the stage of a Nikon Eclipse TE300 inverted microscope equipped with a spinning-disk PerkinElmer Ultraview LCI confocal system, a piezoelectric z-axis motorized stage (Pifoc, Physik Instrumente, Germany) and an Orca ER 12-bit charge-coupled device camera (Hamamatsu Photonics, Japan). Cells expressing ERYFP and/or mtRFP were excited using the 488 nm or the 543 nm line of the HeNe laser (PerkinElmer, Waltham, MA, USA) by using a × 60 1.4 NA Plan Apo objective (Nikon Corporation, Japan). For confocal z-axis stacks, images separated by 0.2 μm along the z-axis were acquired. To evaluate ER morphology and ER-mitochondria tethering on zVAD or BAPTA treatment a Leica confocal microscope (Laser Scanning TCS SP2, Leica Microsystems Heidelberg GmbH, Mannheim, Germany) equipped with Ar/ArKr and He/Ne lasers and with a × 40 oil 1.40 plan Apo immersion objective was utilized. Analysis of the interactions between ER and mitochondria was performed with the software ImageJ. Deconvolution, 2D- and 3D-reconstruction and surface rendering were realized using the VolumeJ plugin of ImageJ. Analysis of the interaction between ER and mitochondria was performed using Manders’ colocalization coefficient.31

Fluorescence recovery after photobleaching

Cells were plated and transfected with ERYFP as previously described. After 48 h they were incubated in HBSS and placed on the stage of a laser scanning microscope (TCS SP5, Leica). Using the LasAF software (Leica), 16-μm2 regions were manually defined to be bleached. To bleach the YFP fluorescence, one z-plane was bleached for a total of 3 s using 100% laser power of the 514 nm laser line with a × 63, 1.4NA objective. The post-bleaching images were taken at 0.937 s intervals for a total of 120 s. Intensities of the photobleached regions were measured and normalized to the intensities of the same region before photobleaching.

Ratiometric measurement of cytosolic Ca2+ concentration

A total of 15x104 WT and c-FLIP−/− MEFs were seeded on 35 mm Petri dishes. Twenty-four hours after seeding, cells were incubated for 45 min at 37 °C in culture medium containing Fura-2AM 3.5 μM (Invitrogen), which is a ratiometric fluorescent dye binding to free intracellular calcium. Cells were then washed free of Fura2-AM and Hanks balanced solution (HBSS, 130 mM NaCl, 5.2 mM KCl, 1 mM MgCl2, 10 mM glucose, 10 mM HEPES, pH 7.4) was added. EGTA was added to all solutions. Dishes were placed into a culture chamber at 37 °C on the stage of an inverted fluorescence microscope (Nikon TE2000E) and connected to a cooled CCD camera (512B Cascade; Roper Scientific, Princeton Instruments, Trenton, NJ, USA). Samples were illuminated alternately at 340 and 380 nm using a random access monochromator (Photon Technology International, Birmingham, NJ, USA) and emission was detected using a 510 nm emission filter. Images were acquired (1-ratio image/s) using Metafluor software (Universal Imaging Corporation, Downingtown, PA, USA). Calibration was obtained at the end of each experiment by maximally increasing intracellular Ca2+-dependent Fura-2AM fluorescence with ionomycin salt solution (Sigma-Aldrich) 5 μM, followed by recording minimal fluorescence in a Ca2+-free medium. Ca2+ concentration was calculated according to the previously described formulas.46 Where indicated, c-FLIP−/− cells were transiently transfected with pBluescript or c-FLIPL plasmids and ER-to-cytosol Ca2+-fluxes were measured in Fura-2AM-loaded cells.

Measurement of mitochondrial Ca2+concentration

A total of 15 × 104 WT and c-FLIP−/− MEFs were seeded on 35 mm, high, glass bottom Petri dishes. Twenty-four hours after seeding, cells were incubated for 1 h at 37 °C in culture medium containing Rhod2-AM 6 μM (Invitrogen). The ester forms of this calcium indicator are cationic, resulting in potential-driven uptake into mitochondria. Cells were then washed with standard medium and EGTA-containing HBSS at room temperature. Single-cell fluorescence was excited at 545 nm using a random access monochromator (Photon Technology International) and emission was detected using a 580 nm emission filter. Images of the emitted fluorescence, obtained with a × 60 oil objective, were collected using Metafluor software (Universal Imaging Corporation). Where indicated, c-FLIP−/− cells were transiently transfected with pBluescript or c-FLIPL plasmids and mitochondrial Ca2+ entry were measured in Rhod-2AM-loaded cells.

Evaluation of caspase-8 activity

Caspase-8 activitiy was measured by the previously tested47 colorimetric FLICE/Caspase-8 Assay kit (BioVision/Vinci-Biochem, Firenze, Italy) according to the manufacturer’s protocol. Levels of caspase-8 activation were normalized by total protein concentration.

Transmission electron microscopy

For ultrastructural observation, cell monolayers were fixed in 2.5% glutaraldehyde (Electron Microscopy Science, Pennsylvania, USA) and in 0.1% cacodylate buffer (Electron Microscopy Science), postfixed in 1% osmium tetroxide (Electron Microscopy Science) and then treated with 1% tannic acid (Acros Organics, New Jersey, NJ, USA). Dehydrated in ethanol (Carlo Erba), cells were detached from the plastic dish by a brief treatment in 1% toluene (Carlo Erba). The pelleted monolayer was then incubated in a solution of 50% toluene and 50% EPON 812 (Electron Microscopy Science) and finally processed for conventional EPON 812 embedding. Ultra-thin sections were cut in a Leica Ultracut R ultramicrotome, contrasted for 10 min in 1% acid tannic and successively in 1% lead hydroxide (Società Italiana Chimici, Rome, Italy) and then viewed in a Hitachi 7000 transmission electron microscope.

Statistical analysis

Values are expressed as mean±S.E. Statistical evaluation was performed by Student's t-test analysis and a value of P≤0.05 was considered statistically significant.