Differences and Similarities in TRAIL- and Tumor Necrosis Factor-Mediated Necroptotic Signaling in Cancer Cells.

crossmark

Differences and Similarities in TRAIL- and Tumor Necrosis FactorMediated Necroptotic Signaling in Cancer Cells Justyna Sosna,a* Stephan Philipp,a* Johaiber Fuchslocher Chico,a Carina Saggau,a Jürgen Fritsch,a Alexandra Föll,a Johannes Plenge,a Christoph Arenz,b Thomas Pinkert,b Holger Kalthoff,c Anna Trauzold,c,d Ingo Schmitz,e Stefan Schütze,a Dieter Adama Institut für Immunologie, Christian-Albrechts-Universität zu Kiel, Kiel, Germanya; Institut für Chemie der Humboldt-Universität, Berlin, Germanyb; Institut für Experimentelle Tumorforschung, Christian-Albrechts-Universität zu Kiel, Kiel, Germanyc; Klinik für Allgemeine Chirurgie, Viszeral-, Thorax-, Transplantations- und Kinderchirurgie, Universitätsklinikum Schleswig-Holstein, Kiel, Germanyd; Systemorientierte Immunologie und Entzündungsforschung, Institut für Molekulare und Klinische Immunologie, Otto von Guericke Universität Magdeburg, Helmholtz Zentrum für Infektionsforschung, Braunschweig, Germanye

Recently, a type of regulated necrosis (RN) called necroptosis was identified to be involved in many pathophysiological processes and emerged as an alternative method to eliminate cancer cells. However, only a few studies have elucidated components of TRAIL-mediated necroptosis useful for anticancer therapy. Therefore, we have compared this type of cell death to tumor necrosis factor (TNF)-mediated necroptosis and found similar signaling through acid and neutral sphingomyelinases, the mitochondrial serine protease HtrA2/Omi, Atg5, and vacuolar Hⴙ-ATPase. Notably, executive mechanisms of both TRAIL- and TNF-mediated necroptosis are independent of poly(ADP-ribose) polymerase 1 (PARP-1), and depletion of p38␣ increases the levels of both types of cell death. Moreover, we found differences in signaling between TNF- and TRAIL-mediated necroptosis, e.g., a lack of involvement of ubiquitin carboxyl hydrolase L1 (UCH-L1) and Atg16L1 in executive mechanisms of TRAIL-mediated necroptosis. Furthermore, we discovered indications of an altered involvement of mitochondrial components, since overexpression of the mitochondrial protein Bcl-2 protected Jurkat cells from TRAIL- and TNF-mediated necroptosis, and overexpression of Bcl-XL diminished only TRAIL-induced necroptosis in Colo357 cells. Furthermore, TRAIL does not require receptor internalization and endosome-lysosome acidification to mediate necroptosis. Taken together, pathways described for TRAIL-mediated necroptosis and differences from those for TNF-mediated necroptosis might be unique targets to increase or modify necroptotic signaling and eliminate tumor cells more specifically in future anticancer approaches.

T

he molecular pathways of regulated cell death (RCD) consist of important coordinated processes that govern the functionality and viability of each cell. Moreover, the precise control of RCD is crucial to maintain cellular homeostasis, and dysfunction of this equilibrium can lead to cancer progression (1). RCD can be divided into three main groups: caspase-dependent cell death, caspase-independent cell death, and autophagic cell death (2). While caspase-dependent extrinsic apoptosis and intrinsic apoptosis (which may involve both caspase-dependent and caspaseindependent mechanisms) have been extensively studied over the last decades, signaling pathways of caspase-independent regulated necrosis (RN) remain poorly understood, and only a few studies are available in the literature. During inhibition of caspases [e.g., by benzyloxycarbonyl-Val-Ala-Asp(OMe)-fluoromethylketone (zVAD)], death receptors (e.g., tumor necrosis factor [TNF], TNF-related apoptosis-inducing ligand [TRAIL], or CD95) and Toll-like receptors (TLRs) (e.g., TLR-3 and TLR-4) have been shown to induce RN that is tightly orchestrated by a set of signal transduction pathways and catabolic mechanisms. The types of RN (i.e., mitotic catastrophe, parthanatos, netosis, entosis, and cornification) are characterized and named according to the activation of specific signaling molecules (3); e.g., caspase-independent RN, induced by death receptors and promoted by the phosphorylation of the serine/threonine kinases RIPK1 and RIPK3 and mixed-lineage kinase domain-like protein (MLKL), is called necroptosis (4). Recently, we have shown that TNF and TRAIL together with pharmacological caspase inhibition (i.e., by zVAD) are able to induce necroptosis in cancer cells (5). Furthermore, we were able to show that necroptosis induced by TRAIL, which is one of the most promising alternative molecules for cancer ther-

2626

mcb.asm.org

apy, can be used for cancer cell reduction (6). However, only necroptosis induced by TNF, which is known to cause strong systemic toxic side effects (7) and neurotoxicity (8), has been investigated in detail so far. Many executive mechanisms, such as the involvement of ceramide and the serine protease HtrA2/Omi, the subsequent ubiquitination of UCH-L1 (9, 10), or the lack of poly(ADP-ribose) (PAR) polymerase 1 (PARP-1) involvement (11), have been identified for TNF-induced necroptosis. Moreover, the increased expression of antiapoptotic proteins such as cFLIP, Bcl-XL, or XIAP and the diminished expression of proapoptotic molecules such as caspases or Bid allow cancer cells to evade apoptotic cell death (12). Thus, RN induced under caspase-compromised conditions, e.g., by naturally occurring substances (13) or drugs (14), arose as an alternative way to eliminate cancer cells. Although TRAIL-mediated necroptosis is emerging as the focus of

Received 15 October 2015 Returned for modification 1 January 2016 Accepted 14 July 2016 Accepted manuscript posted online 15 August 2016 Citation Sosna J, Philipp S, Fuchslocher Chico J, Saggau C, Fritsch J, Föll A, Plenge J, Arenz C, Pinkert T, Kalthoff H, Trauzold A, Schmitz I, Schütze S, Adam D. 2016. Differences and similarities in TRAIL- and tumor necrosis factor-mediated necroptotic signaling in cancer cells. Mol Cell Biol 36:2626 –2644. doi:10.1128/MCB.00941-15. Address correspondence to Justyna Sosna, [email protected]. * Present address: Justyna Sosna and Stephan Philipp, University of California— Irvine, Molecular Biology and Biochemistry School of Biological Sciences, Irvine, California, USA. Copyright © 2016, American Society for Microbiology. All Rights Reserved.

Molecular and Cellular Biology

October 2016 Volume 36 Number 20

Comparison of TRAIL- and TNF-Mediated Necroptosis

anticancer therapy, signaling pathways are still incompletely understood, and only a few studies, e.g., about the inhibitory action of TRAF2 (15) or the involvement of ceramide (16) during TRAIL-induced necroptosis, are available. Moreover, direct comparisons and analyses of differences in signaling pathways between TNF- and TRAIL-mediated necroptosis in the context of future anticancer therapies are still missing. In this study, we addressed this gap and investigated the roles of various signaling molecules (including acid sphingomyelinase [ASMase]/neutral sphingomyelinase [N-SMase], Atg5, HtrA2/Omi, PARP-1, p38␣, UCH-L1, and Atg16L1) in RIPK1-dependent TRAIL-mediated necroptosis in comparison to TNF-mediated necroptosis. Taken together, our data provide new insights into the signaling of TRAIL- and TNF-mediated necroptosis and show differences in both forms that might be critical for their application in future anticancer approaches. MATERIALS AND METHODS Reagents. Highly purified soluble human recombinant killerTRAIL, spiroepoxide, and 3-O-methyl-sphingomyelin (3-OMS) were obtained from Enzo Life Sciences (Lörrach, Germany). Highly purified human recombinant TNF (hrTNF) was provided by BASF Bioresearch (Ludwigshafen, Germany). zVAD, N-benzyloxycarbonyl-Asp-Glu-Val-Asp-O-methylfluoromethylketone (zDEVD), and N-benzyloxycarbonyl-Ile-Glu-ThrAsp-O-methyl-fluoromethylketone (zIETD) were purchased from Bachem (Bubendorf, Switzerland). Nec-1s (7-Cl-O-Nec-1) was obtained from BioVision (Milpitas, CA). 7-Aminoactinomycin D (7-AAD) was obtained from BD Biosciences (Heidelberg, Germany). Tricyclodecan-9-yl (D609) was purchased from Biomol (Hamburg, Germany). Fumonisin B1, 3-methyladenine (3-MA), 5-[5-(2-nitrophenyl)furfurylidine]-1,3-diphenyl-2-thiobarbituric acid (Ucf-101), and benzyloxycarbonylPhe-Phe-fluoromethylketone (zFF) were obtained from Merck Millipore (Schwalbach, Germany). Zoledronic acid was purchased from Novartis (Basel, Switzerland). ARC39 (1-aminodecan-1,1-biphospho acid) was described previously (17), and TP064/14e (1-O-dodecylsulfonyl-myoinositol-3,5-biphosphate) was described previously (18). Desipramine hydrochloride, imipramine hydrochloride, and GW4869 {N,N=-bis[4-(4,5dihydro-1H-imidazol-2-yl)phenyl]-3,3=-p-phenylene-bis-acrylamide dihydrochloride hydrate} were obtained from Sigma-Aldrich (Steinheim, Germany). Radicicol (RC) was purchased from Tebu-Bio (Offenbach, Germany). Geldanamycin (GA) and bafilomycin A1 (BafA1) were purchased from LC Laboratories (Woburn, MA). 3-Aminobenzamide (3-AB) was obtained from TCI Europe (Eschborn, Germany), N-(5,6dihydro-6-oxo-2-phenanthridinyl)-2-acetamide hydrochloride (PJ-34) was obtained from Bio-Techne (Wiesbaden-Nordenstad, Germany), and olaparib (AZD2281) was purchased from Axon (Groningen, Netherlands). Birinapant (TL32711) was provided by ChemieTek (Indianapolis, IN). Butylated hydroxyanisole (BHA), butylated hydroxytoluene (BHT), cycloheximide (CHX), (2S,3S)-trans-epoxysuccinyl-L-leucylamido(4-guanidino)butane (E-64), tert-butyl hydroperoxide (BuOOH), benzyloxycarbonyl-Phe-Ala-fluoromethylketone (zFA), tosyl phenylalanyl chloromethyl ketone (TPCK), chloroquine diphosphate salt (CQ), LDN57444, LDN91946, WIKI4, and XAV939 were purchased from Sigma-Aldrich. PitSop2 and dynasore were obtained from Abcam (Cambridge, UK). (L-3-trans-(Propylcarbamoyl)oxirane-2-carbonyl)-L-isoleucyl-L-proline methyl ester (Ca074 Me) was obtained from the Peptide Institute (Minoh-Shi, Osaka, Japan). CM-H2DCFDA (5= [and 6=]-chloromethyl-2=7=-dichlorodihydrofluorescein diacetate acetyl ester) and CMTMRos MitoTracker orange were purchased from Life Technologies (Darmstadt, Germany). Annexin V-Fluos was obtained from Roche (Mannheim, Germany). Chemical compounds not listed were obtained from Merck and Sigma-Aldrich and were of molecular-grade purity. Cell culture. Human Jurkat and HT-29 cell lines were obtained from the American Type Culture Collection (ATCC) (Manassas, VA). RIPK1-


deficient and Bcl-2-overexpressing Jurkat cells were originally obtained from Brian Seed (Massachusetts General Hospital, Boston, MA). Jurkat Fas-associated death domain protein (FADD)-deficient cells that were additionally stably transfected with TNF-R2, referred to as Jurkat I.42 cells (19), were a kind gift from Francis Ka-Ming Chan (Worcester, MA). The pancreatic carcinoma Colo357 cell line overexpressing Bcl-XL or transfected with a vector was described previously (20). A818-6 human pancreatic adenocarcinoma cells were described previously (21). Murine “genuine” L929ATCC cells were directly obtained from the ATCC. The L929Ts cell line is referred to as a “TRAIL-sensitive” subline, and the L929sA cell line is a TNF-hypersensitive subline of L929, both of which were described previously (16). NIH 3T3 cells naturally expressing RIPK3 and therefore sensitive to necroptosis were described previously (22). Immortalized HtrA2/Omi-deficient mouse embryonic fibroblasts (MEFs) and their wild-type counterparts were originally generated by Julian Downward (Cancer Research UK Signal Transduction Laboratory, London, UK) (23) and provided by Thomas Langer (University of Cologne Institute for Genetics). Immortalized PARP-1-deficient MEFs and wildtype cells were donated by Françoise Dantzer (École Supérieure de Biotechnologie de Strasbourg, Illkirch-Graffenstaden, France) (24). Immortalized Atg16L1-deficient MEFs and their wild-type controls were obtained from Philip Rosenstiel (Institute of Clinical Molecular Biology, Christian Albrechts University, University Hospital Schleswig-Holstein, Kiel, Germany) and Paul Saftig (Christian Albrechts University, Institute of Biochemistry, Unit of Molecular Cell Biology and Transgenic Research, Kiel, Germany) (25). Immortalized Atg5-deficient MEFs and their wildtype counterparts were generated and described previously by Keil et al. (26). MEFs stably expressing the wild-type Atg5 protein and the Atg5 T75A mutant were described previously (27). p38␣-deficient MEFs and their wild-type counterparts were obtained from Angel Nebreda (Institute for Research in Biomedicine, Barcelona, Spain) and were described previously (28). Cells were cultivated in Dulbecco’s modified Eagle’s medium (DMEM) (MEFs), McCoy’s medium (HT-29 and A818-6), RPMI 1640 (Colo357), or a mixture of Click’s medium and RPMI 1640 medium (all other cell lines) supplemented with 10% (vol/vol) fetal bovine serum and 2 mM L-glutamine and additionally supplemented with 1 mM sodium pyruvate (HT-29), 50 ␮M ␤-mercaptoethanol in 0.9% (wt/vol) NaCl (all other cell lines), and 100 ␮g/ml each of streptomycin and penicillin in a humidified incubator containing 5% (wt/ vol) CO2. Measurement of loss of plasma membrane integrity. Cells were seeded into 12-well plates at a density of 1 ⫻ 105 cells/well. After treatment, both detached and adherent cells were collected by centrifugation (5 min at 400 ⫻ g at 4°C). Cells were resuspended in phosphate-buffered saline (PBS)–5 mM EDTA and stained with 2 ␮g/ml propidium iodide (PI) solution. The red fluorescence of 10,000 cells was then measured by using a FACSCalibur analyzer (BD Biosciences, San Diego, CA). Measurement of intracellular ATP levels. The cellular ATP content as an indicator of cell viability was determined with a microplate reader (Tecan, Crailsheim, Germany) by using a bioluminescence assay (Cell Titer Glo assay kit; Promega, Mannheim, Germany) according to the manufacturer’s instructions. Measurement of reactive oxygen species levels. Following stimulation, cells were detached and centrifuged as described above; washed twice with PBS; and stained with 30 ␮M dihydroethidium (DHE) for 15 min, 10 ␮M CM-H2DCFDA for 30 min, or 2 ␮g/ml PI solution added 1 min before measurement. The green or red fluorescence of 10,000 cells was measured by flow cytometry. The level of reactive oxygen species (ROS) was measured by flow cytometry and is shown as the fold change in relative fluorescence units (RFUs) in comparison to those for untreated control cells, normalized to a value of 1. Flow cytometric analysis of ⌬⌿m. After treatment, cells were washed with PBS and resuspended in a total volume of 100 ␮l of 150 nM CMTMRos for 30 min at 37°C in the dark. Afterwards, cells were centrifuged (5 min at 400 ⫻ g at room temperature [RT]) and fixed in 4%


mcb.asm.org

2627

Sosna et al.

(wt/vol) paraformaldehyde for 15 min at 37°C in the dark, and 10,000 cells were analyzed by flow cytometry. Flow cytometric analysis of membrane integrity and phosphatidylserine externalization. Following treatment, cells were washed with PBS and stained with a mixture of 2% (vol/vol) annexin V-Fluos and 5% (vol/vol) 7-AAD in staining buffer (10 mM HEPES [pH 7.4], 140 mM NaCl, 5 mM CaCl2) for 15 min at 37°C in the dark, and 10,000 cells were measured by flow cytometry. RNA interference. Silencer Select small interfering RNAs (siRNAs) specific for murine HtrA2/Omi(1) (catalogue number s82292), HtrA2/ Omi(2) (catalogue number s82293), murine UCH-L1 (catalogue number s75710), and murine (catalogue number s62053) and human (catalogue number s1097) PARP-1 were obtained from Life Technologies. A total of 1 ⫻ 106 cells were transfected with 150 pmol siRNA by using Cell Line Nucleofector kit V (Amaxa, Cologne, Germany) and the T-20 (L929Ts) or X-001 (Jurkat) program. Each transfection experiment was validated by using a negative control consisting of a scrambled siRNA that does not target any gene (siNT) (catalogue number AM4611). The efficiency of downregulation after siRNA transfection was analyzed by Western blotting for the proteins of interest. Immunological reagents. Cells were harvested after treatment and lysed at 4°C in TNE buffer (50 mM Tris [pH 8.0], 1% [vol/vol] NP-40, 150 mM NaCl, 3 mM EDTA, 1 mM sodium orthovanadate [Na3VO4], 5 mM sodium fluoride [NaF], and a complete protease inhibitor cocktail [Roche]). Identical amounts of protein per lane were resolved by electrophoresis on SDS-polyacrylamide gels and transferred onto nitrocellulose membranes by Western blotting. Reactive proteins were detected by using antibodies specific for Atg16L1 (D5D6, catalogue number 8089; Cell Signaling), ␤-actin (catalogue number A1978; Sigma), Bcl-XL (2H12, catalogue number 551020; BD), Bcl-2 (catalogue number sc-509; Santa Cruz), I␬B␣ (Ser32) (14D4, catalogue number 2859; Cell Signaling), I␬B␣ (C-21, catalogue number sc-371G; Santa Cruz), p-p65 (Ser536) (93H1, catalogue number 3033; Cell Signaling), p65 (C22B4, catalogue number 4764; Cell Signaling), HtrA2/Omi (catalogue number ab32092; Abcam), LC3 (catalogue number 0231-100/LC3-5F10; nanoTools), PAR (catalogue number 551813, component 51-8114KC; BD Pharmingen), PARP-1 (catalogue number 9542; Cell Signaling), p38␣ (5F11, catalogue number 9217; Cell Signaling), RIPK1 (catalogue number 610459; BD Biosciences), murine RIPK3 (catalogue number PRS2283; Sigma), human RIPK3 (catalogue number PAB0287; Abnova), UCH-L1 (polyclonal antibody [PAb] CL95101; Cedarlane), and UCH-L1 (monoclonal antibody [MAb] described previously [9]) and the LumiGLO chemiluminescent substrate (Cell Signaling, Danvers, MA) and captured on Amersham Hyperfilm ECL (GE Healthcare, Munich, Germany). Equal loading as well as efficiency of transfer were routinely verified for all Western blots by reprobing the membranes for ␤-actin. Transient expression of Atg16L1 constructs. The Atg16L1 construct was provided by Philip Rosenstiel (Institute of Clinical Molecular Biology, Christian Albrechts University, University Hospital Schleswig-Holstein, Kiel, Germany). Briefly, for each transfection point, 1 ⫻ 106 MEFs were electroporated by the use of a basic primary fibroblast Nucleofector kit (catalogue number VAPI-1002; Lonza, Cologne, Germany) with the A-023 program for 48 h, followed by cytotoxicity assays and measurement of the loss of plasma membrane integrity by fluorescence-activated cell sorter (FACS) analysis. Internalization study by using ImageStream. Briefly, for internalization studies, cells were detached by using ice-cold PBS-EDTA, cooled on ice for 15 min, and labeled with either Fc-tagged TNF (purified from 293T cell supernatants) or Fc-tagged TRAIL (kindly provided by Henning Walczak, UCL, Cancer Institute, London, UK) coupled to protein G-Alexa Fluor 488 (catalogue number P-11065; Thermo Fisher, Darmstadt, Germany), which was added and incubated for 30 min on ice. zVAD was added for the final 20 min of incubation. After washing of the cells with cooled PBS, synchronous receptor internalization was allowed for 30 min by adding 50 ␮l prewarmed medium without fetal calf serum (FCS) and

2628

mcb.asm.org

incubating the cells in the dark at 37°C. Internalization was stopped by the addition of cold PBS to the cells. Plasma membranes were stained with CellMask Deep Red plasma membrane stain (catalogue number C10046; Thermo Fisher, Darmstadt, Germany) for 5 min on ice. Subsequently, cells were centrifuged and fixed in 2% paraformaldehyde prior to measurement. Image acquisition was performed by using an ImageStreamX Mark II imaging flow cytometer (Amnis Corporation, Seattle, WA) equipped with INSPIRE software. For each assay, 10,000 images were acquired at a ⫻60 magnification. Data analysis was performed by using IDEAS software (Amnis Corporation). Microscopic observations. Cells undergoing RCD were visualized with an Axiovert 10 light microscope with achrostigmat objectives with a resolution (magnification/aperture) of 32⫻/0.40 or 20⫻/0.30, and images were digitalized with a Nikon DS-5M-L1 Digital Sight camera system. Statistical analysis. P values were calculated by using Student’s t test. The statistical significance of increased or decreased values is indicated in the figure legends.

RESULTS

TRAIL triggers RIPK1-dependent necroptosis after caspase-8 inhibition. As we have shown previously (5, 6, 16), TRAIL in combination with the pancaspase inhibitor zVAD (Fig. 1A), and, for some human cancer cells, additionally after sensitization with a protein synthesis inhibitor (i.e., CHX), is able to induce RIPK1dependent necroptosis (Fig. 1B). The role of RIPK1 during TRAIL-mediated necroptosis was confirmed by genetic ablation (Fig. 1B) or the application of two inhibitors of the 90-kDa heat shock protein (HSP90) (29), GA and RC (Fig. 1C), which are known to downregulate the protein level of RIPK1 (30). Furthermore, the main role of RIPK1 during necroptosis was confirmed for all analyzed cell lines (L929Ts, NIH 3T3, Jurkat, and HT-29), as Nec-1s (analogue of Nec-1, a specific and stable RIPK1 inhibitor, but lacking the targeting effect on the immunomodulatory enzyme indoleamine 2,3-dioxygenase [IDO] [31]) prevented the execution of TRAIL-mediated necroptosis (Fig. 1D). Moreover, TRAIL-mediated necroptosis, compared to caspase-dependent apoptosis, is characterized by the late-stage externalization of phosphatidylserine and a loss of mitochondrial membrane potential (Fig. 1E and F). Furthermore, we started to compare a broad range of signaling events during TRAIL- and TNF-induced necroptosis. One of the previously described differences between TRAIL- and TNF-mediated necroptosis is the dependence on the presence of FADD (Fasassociated protein with a death domain), one of the adaptor molecules building the death-inducing signaling complex (DISC). As it was previously shown that necroptosis is induced by TNF alone or by the combination of TNF and zVAD in Jurkat cells lacking FADD (32), TRAIL- or TRAIL-zVAD-mediated necroptosis is not executed in cells lacking FADD (Fig. 1G). Since there are already indications of differences in the signaling of TRAIL- and TNFinduced necroptosis, we analyzed and compared further signaling events to describe the diversity and similarities in TRAIL- and TNF-mediated necroptosis. TRAIL-mediated necroptosis is inhibited by Ca-074 Me-dependent mechanisms. It was shown previously that commonly used caspase inhibitors also interfere with other proteases at higher concentrations, including cathepsins and calpains (33). Thus, we investigated if, similarly to the inhibition of caspases (by zVAD) (Fig. 1A), the inhibition of cathepsins and calpains may lead to the initiation of TRAIL-mediated cell death. Interestingly, zFA, a broad-range inhibitor of cathepsins B, L, and S, and zFF, a




FIG 1 (A) TRAIL-induced necroptosis is enhanced by inhibition of caspase-8. To further characterize TRAIL-mediated necroptosis, we evaluated the necessity of caspase inhibition for the induction of this type of cell death. In the time-dependent manner, we compared the effects of the inhibition of effector caspase-3 (by zDEVD) and initiator caspase-8 (by zIETD). Only inhibition of caspase-8 (by zIETD or the pancaspase inhibitor zVAD) enhanced cell death mediated by TRAIL and strongly sensitized cells to TRAIL. Cells were left untreated or stimulated with 30 ng/ml killerTRAIL and 20 ␮M zDEVD, zIETD, or zVAD for 14 or 24 h. (B to D) TRAIL-induced necroptosis is mediated through RIPK1. (B) Wild-type (WT) (RIPK1⫹/⫹) and RIPK1-deficient (RIPK1⫺/⫺) Jurkat cells were left untreated or stimulated with 50 ng/ml killerTRAIL, 50 ␮M zVAD, and 2 ␮g/ml CHX for 20 h. Insets show control Western blots for endogenous RIPK1 and ␤-actin as a loading control. (C) Cells were left untreated or preincubated with 1 ␮g/ml GA or 5 ␮g/ml RC (L929Ts), 0.5 ␮g/ml GA or 5 ␮g/ml RC (NIH 3T3), or 1 ␮g/ml GA or 0.5 ␮g/ml RC (Jurkat) for 24 h and stimulated afterwards with 30 ng/ml killerTRAIL and 20 ␮M zVAD (L929Ts); 100 ng/ml killerTRAIL and 20 ␮M zVAD (NIH 3T3); or 50 ␮g/ml killerTRAIL, 50 ␮M zVAD, and 2 ␮g/ml CHX (Jurkat) for 24 h. The fraction of surviving cells is displayed relative to cells that were treated as described above, except that no TRAIL-zVAD (TRAIL-zVAD-CHX) was added. (D) Cells were pretreated for 2 h with 50 ␮M Nec-1s and subsequently treated with either 30 ng/ml (14 h for L929Ts cells and 16 h for HT-29 cells), 50 ng/ml (20 h for Jurkat cells), or 100 ng/ml (NIH 3T3) killerTRAIL in combination with 20 ␮M (L929Ts, NIH 3T3, and HT-29) or 50 ␮M (Jurkat) zVAD and additionally with 2 ␮g/ml (Jurkat) or 5 ␮g/ml (HT-29) CHX. Shown are means ⫾ standard deviations (n ⱖ 9), with differences being considered significant at a P value of ⬍0.001 (***) (as determined by a t test). The morphology of untreated cells undergoing TRAIL-mediated necroptosis and TRAIL-induced necroptosis inhibited by Nec-1s is shown. Bar, 100 ␮m. (E) Necroptotic cell death induced by TRAIL is not accompanied by early phosphatidylserine exposition. An involvement of apoptosis in TRAIL-induced necroptosis was excluded, as phosphatidylserine, detected by using annexin V, was exposed to the outer membrane only during the late phase of cell death and was accompanied by a simultaneous loss of membrane integrity, which was measured by the uptake of 7-AAD. Cells were treated with 20 ␮M zVAD and 30 ng/ml (L929Ts) or 100 ng/ml (NIH 3T3) killerTRAIL or with 50 ng/ml killerTRAIL, 50 ␮M zVAD, and 2 ␮g/ml CHX (Jurkat). (F) Loss of ⌬⌿m is a late event during TRAIL-induced necroptosis. Cells were treated as described above for panel E. Mitochondrial membrane potential was measured after the indicated times. Shown are means ⫾ standard deviations (n ⫽ 4 [A], n ⫽ 3 [B], n ⫽ 4 [C], n ⱖ 3 [D], n ⫽ 3 [E], and n ⫽ 3 [F]), with differences being considered significant at P values of ⬍0.05 (*), ⬍0.01 (**), and ⬍0.001 (***) (as determined by a t test). (G) Deficiency in FADD inhibits TRAIL-mediated necroptosis. Jurkat I.42 cells (FADD deficient and TNF-R2 positive) were pretreated or not with 50 ␮M zVAD and treated with 50 ng/ml killerTRAIL for 14 to 24 h (two independent experiments; n ⫽ 5). n.s., nonsignificant.



mcb.asm.org

2629

Sosna et al.

FIG 2 (A) TRAIL-induced apoptosis and necroptosis are enhanced by zFA and zFF. L929Ts cells were incubated with 30 ng/ml killerTRAIL (for apoptosis) or 20 ␮M zVAD (prestimulation for 30 min) and subsequently incubated with 30 ng/ml killerTRAIL (for necroptosis) in the absence or presence of 20 ␮M zFA, zFF, or E-64 for 16 h. (B to E) TRAIL-mediated necroptosis is inhibited by Ca-074 Me. L929Ts (B), L929sA (C), Jurkat (D), and HT-29 (E) cells were pretreated with Ca-074 Me for 2 h and stimulated with either 30 ng/ml (14 h for L929Ts cells and 16 h for HT-29 cells) or 50 ng/ml (20 h for Jurkat cells) killerTRAIL or 100 ng/ml (5 h for L929Ts cells, 16 h for HT-29 cells, and 20 h for Jurkat cells) hrTNF in combination with 20 ␮M (L929Ts and HT-29) or 50 ␮M (Jurkat) zVAD and additionally with 2 ␮g/ml (Jurkat) or 5 ␮g/ml (HT-29) CHX. Shown are means ⫾ standard deviations (n ⫽ 4 [A] and n ⫽ 9 [B to E]), with differences being considered significant at P values of ⬍0.05 (*) and ⬍0.001 (***) (as determined by a t test). n.s., nonsignificant.

cathepsin inhibitor which inhibits only cathepsins B and L (34), slightly enhanced the cytotoxicity of TRAIL alone (Fig. 2A). However, zFA and zFF did not enhance TRAIL-induced cell death to the level mediated by the combination of TRAIL and zVAD. Similarly to TRAIL-mediated apoptosis, both the inhibitors zFA and zFF slightly increased the level of TRAIL-zVAD-mediated necroptosis. However, these effects might rather be inhibitor-specific effects due to lysosomal dysfunction leading to increased cell death,

2630

mcb.asm.org

which is not necessarily associated with necroptosis (35). Therefore, the addition of the general cysteine protease inhibitor E-64 did not influence TRAIL-induced apoptosis or necroptosis in L929Ts cells (Fig. 2A). Taken together, the role of cysteine proteases in the initiation of TRAIL-mediated necroptosis or indications of the involvement of calpains or cathepsins in the execution of TRAIL-zVAD-induced necroptosis were not found. On the other hand, treatment with Ca-074 Me, an inhibitor of cathepsins B and L (36) that is believed to affect cathepsin-independent mechanisms upstream of lysosomal breakdown (37), diminished the execution of TRAIL-zVAD-induced necroptosis in murine L929Ts and L929sA cells. Surprisingly, no inhibitory effect of Ca-074 Me was observed for TNF-zVAD-induced necroptosis in murine cells (Fig. 2B and C), pointing to differences in the executions of TRAIL- and TNF-induced necroptosis. Intriguingly, this same inhibitor protected human Jurkat and HT-29 cell lines from both TRAIL- and TNF-mediated necroptosis (Fig. 2D and E), indicating that differences in the execution of TRAIL- or TNF-induced necroptosis might be cell type and/or species specific. Activities of A-SMase and N-SMase trigger TRAIL-mediated necroptosis. TRAIL-induced apoptosis is accompanied by signaling through ceramide (38). Moreover, recent studies identified ceramide as being pivotal for TNF-mediated necroptosis (10); therefore, we investigated whether ceramide is similarly involved in signaling during TRAIL-mediated necroptosis. Ceramide might be generated from diverse pathways, including the de novo pathway, synthesis from sphingosine by ceramide synthase, and the hydrolysis pathway executed by A-SMase or N-SMase. Thus, we verified which metabolic pathway led to the generation of ceramide during TRAIL-induced necroptosis using inhibitors of A-SMase (ARC39, zoledronic acid, TP064/14e, D609, desipramine, and imipramine), N-SMase (3-OMS, spiroepoxide, and GW4869), and ceramide synthase (fumonisin B1). In murine L929Ts and NIH 3T3 cells, the A-SMase inhibitors ARC39, zoledronic acid (not for NIH 3T3 cells), and D609 reduced TRAIL-mediated necroptosis (Fig. 3A and C). This is similar to TNF-induced necroptosis, during which the A-SMase inhibitors ARC39 and zoledronic acid (not for Jurkat I.42 cells) protected cells from death (Fig. 3B to E). Moreover, the role of A-SMase (inhibited by ARC39) during TRAIL-mediated necroptosis was confirmed for human A818-6 cells as well (Fig. 3F). These data corroborated previously reported observations, proving the pivotal role of A-SMase in necroptosis (10). Furthermore, for the first time, we have shown that two inhibitors of N-SMase, 3-OMS and spiroepoxide, reduced the rate of TRAIL-mediated (Fig. 3A and C) and TNF-mediated (Fig. 3B to E) necroptosis. These data suggest an as-yet-unexplored role of N-SMase in the execution of necroptosis. In addition, fumonisin B1, an inhibitor of ceramide synthase, was not able to protect cells from either TRAIL-mediated (Fig. 3A and C) or TNF-mediated (Fig. 3B and D) necroptosis. This finding is in consensus with data from previous studies proving that the inhibition of ceramide synthase, e.g., through the sphingosine analogue FTY720 (39), did not protect cells from but rather enhanced necroptosis (40). Participation of mitochondria in TRAIL-mediated necroptosis. Mitochondria have been implicated as an important factor in the regulation of necroptosis (41), and increased production of ROS during TNF-mediated necroptosis was described




FIG 3 TRAIL- and TNF-induced necroptosis are mediated by ceramide that is generated by A-SMase and N-SMase. (A to D) Cells were pretreated for 2 h with the indicated concentrations of inhibitors of A-SMase (ARC39, zoledronic acid, TP064/14e, D609, desipramine, and imipramine), N-SMase (3-OMS, spiroepoxide, and GW4869), and ceramide synthase (fumonisin B1), with the subsequent addition of 30 ng/ml killerTRAIL and 20 ␮M zVAD for 14 h (L929Ts) (A), 100 ng/ml hrTNF in combination with 20 ␮M zVAD for 5 h (L929Ts) (B), 100 ng/ml killerTRAIL and 20 ␮M zVAD for 16 h (NIH 3T3) (C), and 100 ng/ml hrTNF and 20 ␮M zVAD for 16 h (NIH 3T3) (D). (E) Inhibitors of A-SMase and N-SMase protect human Jurkat I.42 cells (FADD deficient and TNF-R2 positive) from TNF-mediated necroptosis. Cells were treated with the indicated concentrations of inhibitors and stimulated afterwards with 100 ng/ml hrTNF in combination with 50 ␮M zVAD for 6 h. (F) Inhibitors of A-SMase, HtrA2/Omi, UCH-L1, and vacuolar H⫹-ATPase protect the human pancreas adenocarcinoma cell line A818-6 from TRAIL-mediated necroptosis. Cells were pretreated for 2 h (or 3 h for LDN57444) with the indicated concentrations of inhibitors with the subsequent addition of 100 ng/ml killerTRAIL in combination with 50 ␮M zVAD for 24 h. Morphological changes of human A818-6 pancreas adenocarcinoma cells after induction of TRAIL-mediated necroptosis in combination with inhibitors of A-SMase, HtrA2/Omi, and vacuolar H⫹-ATPase were observed. Cells were pretreated for 2 h with 10 ␮M ARC39, 25 ␮M Ucf-101, or 10 ␮M BafA1 with the subsequent addition of 100 ng/ml killerTRAIL in combination with 50 ␮M zVAD for 24 h. Arrowheads in micrographs show typical necroptotic morphologies. Bar, 100 ␮m. Shown are means ⫾ standard deviations (n ⫽ 3 [A to D] and n ⫽ 9 [E and F]), with differences being considered significant at P values of ⬍0.05 (*), ⬍0.01 (**), and ⬍0.001 (***) (as determined by a t test).

previously (42). Similarly, we observed an accumulation of ROS during TRAIL-mediated necroptosis in all investigated cell lines (Fig. 4A). Moreover, the radical scavenger BHA (but not BHT), which was found to protect cells from hydroxyl radicals generated by BuOOH (Fig. 4A), also protected cells from the TRAIL-induced generation of ROS (Fig. 4A) and simultaneously abolished TRAIL-induced necroptosis in the murine L929Ts and NIH 3T3 cell lines (Fig. 4B). However, in human Jurkat cells, ROS production was not observed, and there-


fore, the radical scavenger BHA did not protect cells from TRAILmediated necroptosis. This suggests that ROS are not uniformly and essentially involved in TRAIL-induced necroptosis. Similar data for a cell type-specific involvement of ROS in TRAIL-induced necroptosis were described recently for human pancreatic MiaPaCa-2 and BxPC-3 cells (43). Differences in the effectiveness of the radical scavengers BHA and BHT in protection from TRAIL-mediated necroptosis might be explained by the additional role of BHT in blocking the formation of advanced glyca-


mcb.asm.org

2631

Sosna et al.

tion end (AGE) products generated due to enhanced glycolysis (44). As ROS are described to trigger the stress-activated protein kinase p38␣ (45), we investigated its role in necroptosis. Here, the absence of p38␣ augmented both TRAIL- and TNF-mediated necroptosis (Fig. 4C). This might be caused by an increased activation of NF-␬B, which, under certain conditions, may mediate cell death (46). The canonical NF-␬B dimer consists of p50 and p65 and is activated by the inducible phosphorylation, ubiquitylation, and degradation of I␬B␣, leading to the nuclear translocation of p50 and p65 (46). Here, we observed that although the levels of I␬B␣ in both cell lines were comparable, MEFs lacking p38␣ in the unstimulated state and upon stimulation with cytokines in combination with zVAD or CHX showed higher levels of phosphorylated I␬B␣ (p-I␬B␣) (Fig. 4D), which signals NF-␬B for proteasomal degradation. Furthermore, we observed phosphorylation of p65 at S536, which is independent of I␬B␣ regulation (47) and is required for the enhanced transcriptional activity of NF-␬B (48). In line with this, p65 was more strongly phosphorylated at S536, and its overall level was slightly higher in p38␣deficient cells (Fig. 4E), which altogether implicate a stronger transcriptional activation potential of NF-␬B in p38␣-deficient cells and consequently may lead to a faster execution of cell death. TNF-mediated phosphorylation of NF-␬B p65 on S536 was described previously by Sakurai and coworkers (49), and similarly, we observed this for TRAIL- and TNF-mediated necroptosis and apoptosis, especially in cells lacking p38␣. Although the application of CHX has no negative effect on the execution of cell death (as shown by FACS analyses) (Fig. 4C), the negative effect of CHX on the NF-␬B pathway is clearly presented in Fig. 4D and E. In both cell lines (p38␣⫹/⫹ and p38␣⫺/⫺), the administration of CHX diminished the intensity of the corresponding bands of p-I␬B␣ and p-p65 compared to cell death induced without the administration of CHX. As the level of p-I␬B␣ was per se higher in cells lacking p38␣ and phosphorylation of p65 was stronger in cells lacking p38␣ than in wildtype cells, we assumed that this pathway was one of the possible pathways to increase the execution of cell death; however, other mechanisms and multiple effects may account for the observed phenotype. To further investigate the role of mitochondrial pathways in necroptosis, we analyzed the involvement of the mitochondrial

FIG 4 (A and B) ROS are not uniformly the executioners of TRAIL-mediated

necroptosis. (A) All tested cell lines were pretreated or not with 150 ␮M BHA or BHT for 1 h, with the subsequent addition of 30 ng/ml (14 h for L929Ts cells), 100 ng/ml (16 h for NIH 3T3 cells), or 50 ng/ml (20 h for Jurkat cells) killerTRAIL; 20 ␮M zVAD (L929Ts and NIH 3T3); or 50 ␮M zVAD and 2 ␮g/ml CHX (Jurkat). Each cell line was treated with 1 mM BuOOH for 24 h as a positive control for ROS production. (B) Cells were treated as described above for panel A, and loss of membrane integrity was measured. (C) Lack of p38␣ increases execution of TRAIL- and TNF-mediated necroptosis. p38␣deficient immortalized MEFs and their wild-type counterparts were prestimulated for 30 min with or without 20 ␮M zVAD and with or without 1 ␮g/ml

2632

mcb.asm.org

CHX with the subsequent addition of 100 ng/ml killerTRAIL or 100 ng/ml hrTNF for 24 h. (D and E) Deficiency in p38␣ increases phosphorylation of I␬B␣ and phosphorylation of p65 during TRAIL- and TNF-induced cell death. Cells were stimulated as described above for panel C for 1 h, followed by total lysis and Western blot analyses. Shown are Western blots for total I␬B␣, phosphorylated (Ser32) I␬B␣ (p-I␬B␣), total p65 (*, unspecific band), and phosphorylated (Ser536) p65 (p-p65); p38␣ served as a control for deficiency, and ␤-actin served as a loading control. (F and G) Overexpression of Bcl-XL and Bcl-2 protects cells from TRAIL-mediated necroptosis. (F) Wild-type and Bcl2-overexpressing Jurkat cells were stimulated with 50 ng/ml killerTRAIL or 100 ng/ml hrTNF, 50 ␮M zVAD, and 5 ␮g/ml CHX for 20 h. (G) Colo357 cells stably overexpressing Bcl-XL or an empty vector were stimulated with 100 ng/ml killerTRAIL or 100 ng/ml hrTNF alone or in combination with 5 ␮g/ml CHX, 50 ␮M zVAD, and 1 ␮M the Smac mimetic birinapant (B) for 24 h. Shown are means ⫾ standard deviations (panels A and B show data from one representative experiment for one repetition out of four) (n ⫽ 9 [C, F, and G]), with differences being considered significant at a P value of ⬍0.001 (***) (as determined by a t test). n.s., nonsignificant. Insets show control Western blots for endogenous p38␣, Bcl-2, Bcl-XL, and ␤-actin, which served as a loading control.




FIG 5 TRAIL-mediated necroptosis is executed through the chymotrypsin-like serine protease HtrA2/Omi. (A) Cells were stimulated with 20 ␮M zVAD and 30

ng/ml (14 h for L929Ts cells) or 100 ng/ml (16 h for NIH 3T3 cells) killerTRAIL; 50 ␮M zVAD, 2 ␮g/ml CHX, and 50 ng/ml killerTRAIL (20 h for Jurkat cells); or 20 ␮M zVAD, 5 ␮g/ml CHX, and 30 ng/ml killerTRAIL (16 h for HT-29 cells) in the presence of the indicated concentrations of TPCK. (B) Cells were pretreated for 2 h with Ucf-101, with the subsequent addition of 30 ng/ml killerTRAIL, 20 ␮M zVAD (14 h for L929Ts cells), and additionally 5 ␮g/ml CHX (16 h for HT-29 cells). (C) L929Ts cells (top) and HT-29 cells (bottom) were stimulated as described above for panel B, and necroptotic changes in morphology (arrowheads) were investigated by microscopic observation. Bar, 100 ␮m. (D) Cells were transfected with two different siRNAs specific for murine HtrA2/Omi (siRNAs 1 and 2) or a nontargeting control siRNA (NT). At 72 h posttransfection, cells were stimulated with 30 ng/ml killerTRAIL and 20 ␮M zVAD for 14 h. Measurement of intracellular ATP levels served as an indicator of cell viability. Control Western blots show downregulation of endogenous murine HtrA2/Omi in transfected but untreated cells and ␤-actin as a loading control. (E) Wild-type (WT) or HtrA2/Omi-deficient (HtrA2/Omi⫺/⫺) MEFs were incubated with 30 ng/ml killerTRAIL, 20 ␮M zVAD, and 1 ␮g/ml CHX for 16 h. Insets show control Western blots for endogenous HtrA2/Omi and the RIPK1 and RIPK3 kinases pivotal for signaling of necroptosis; ␤-actin served as a loading control. (F) Impact of HtrA2/Omi deficiency on morphology of MEFs undergoing TRAIL-mediated necroptosis. Cells were stimulated as described above for panel E. Arrowheads indicate cell morphology typical of necroptosis. Bar, 100 ␮m. Shown are means ⫾ standard deviations (n ⫽ 4 [L929Ts and NIH 3T3], n ⫽ 6 [Jurkat and HT-29] [A and B], and n ⫽ 9 [E]; panel D shows data from one representative experiment in two repetitions out of two), with differences being considered significant at P values of ⬍0.05 (*), ⬍0.01 (**), and ⬍0.001 (***) (as determined by a t test).

protein Bcl-2, which is a downstream target of p38␣ and is considered to be phosphorylated by p38␣, leading to the attenuation of its prosurvival properties (50). We examined whether the presence of Bcl-2 might have prosurvival properties during necroptosis. In line with this, the overexpression of Bcl-2 in Jurkat cells increased the viability of these cells during TNF-mediated necroptosis and similarly advanced the survival of cells during TRAILmediated necroptosis (Fig. 4F). Correspondingly, the role of another mitochondrial prosurvival protein, Bcl-XL, during TNFand TRAIL-mediated necroptosis and apoptosis was analyzed. The overexpression of Bcl-XL in Colo357 cells strongly inhibited the execution of both TRAIL- and TNF-mediated apoptosis, even though TRAIL- and TNF-mediated apoptosis were induced in combination with the protein synthesis inhibitor CHX, proving both the prosurvival and antiapoptotic potentials of each protein after overexpression (Fig. 4F and G). Surprisingly, the overexpression of Bcl-XL revealed differences in signaling between TRAILand TNF-induced necroptosis. Whereas TRAIL-mediated necroptosis in Colo357 cells was diminished after overexpression


of Bcl-XL, this protein did not influence TNF-mediated necroptosis (Fig. 4G). We confirmed this effect independently by inducing TRAIL- and TNF-mediated necroptosis in combination with the Smac mimetic birinapant (Fig. 4G). These results support the hypothesis that TRAIL and TNF may induce necroptosis through distinct mechanisms. Serine protease HtrA2/Omi but not ubiquitin protein hydrolase UCH-L1 executes TRAIL-induced necroptosis. Previously, we showed that the serine protease HtrA2/Omi executes TNFmediated necroptosis (9). To find out whether TRAIL-mediated necroptosis is also executed by serine proteases, the chymotrypsin-like serine protease inhibitor TPCK was used. The addition of TPCK protected all tested cell lines from TRAIL-mediated necroptosis (Fig. 5A), indicating a putative role of serine proteases during this type of cell death. Furthermore, pharmacological inhibition of the mitochondrial serine protease HtrA2/Omi by the specific inhibitor Ucf-101 reduced the levels of TRAIL-mediated necroptosis in murine and human cells (Fig. 3F and 5B and C). On the other hand, the downregulation of HtrA2/Omi did not rescue


mcb.asm.org

2633

Sosna et al.

murine cells from TRAIL-mediated necroptosis (Fig. 5D), probably due to an insufficient reduction of HtrA2/Omi expression (and, thus, activity). However, genetic ablation of HtrA2/Omi nearly completely protected cells from TRAIL-mediated necroptosis (Fig. 5E and F). Moreover, we previously showed that TNFmediated necroptosis is accompanied by HtrA2/Omi-dependent monoubiquitination of the ubiquitin protein hydrolase UCH-L1 (9), which probably leads to its activation (51) and to the execution of TNF-mediated necroptosis. Therefore, we investigated the role of UCH-L1 during TRAIL-mediated necroptosis. Measurements of the loss of membrane integrity after inhibition of UCH-L1 with the specific inhibitors LDN57444 and LDN91946 (Fig. 3F and 6A and B) revealed that only TNF-induced and not TRAIL-induced necroptosis is mediated by UCH-L1 activity. In line with this finding, the downregulation of UCH-L1 did not rescue cancer cells from TRAIL-mediated necroptosis (Fig. 6C). Instead, an enhancement of TRAIL-mediated necroptosis was observed after the inhibition of UCH-L1 in certain cell lines (Fig. 3F and 6A and B). In line with previously reported data, the HtrA2/ Omi-dependent disappearance of the main form of UCH-L1 (25 kDa) and the appearance of a monoubiquitinated, activated form of UCH-L1 (ca. 35 kDa [9]) were observed only during TNFmediated necroptosis and not during TRAIL-mediated necroptosis (Fig. 6D). Similarly, time course analyses confirmed the lack of HtrA2/Omi-dependent monoubiquitination of UCH-L1, which might execute TRAIL-mediated necroptosis (Fig. 6E). These data corroborate the concept that TRAIL-mediated necroptosis is regulated by signaling pathways that are different from those of TNFmediated necroptosis. Vacuolar Hⴙ-ATPase but not lysosomal flux executes TRAIL-mediated necroptosis. Since we confirmed the role of serine proteases in proteolytic processes during necroptosis, we were interested in whether other forms of cellular decomposition, e.g., “autophagic degradation,” might contribute to TRAIL- and TNFmediated necroptosis. In contrast to L929Ts cells, which do not induce cell death in response to the caspase inhibitor zVAD (Fig. 7A), in L929ATCC cells, the pancaspase inhibitor zVAD alone induces autophagy (52). This was verified in L929ATCC cells by the inhibition of zVAD-induced autophagy by the addition of the autophagy inhibitor 3-MA (Fig. 7A and B). On the contrary, zVAD-induced autophagy was not inhibited with CQ or BafA1, which are not strict inhibitors of autophagy but rather inhibit lysosomal acidification and activation of vacuolar H⫹-ATPase (VATPase), respectively. Importantly, in L929Ts cells, BafA1 was able to reduce the level of TRAIL-mediated necroptosis, whereas TNF-mediated necroptosis was diminished by both the inhibitors CQ and BafA1 (Fig. 7A and B). Further investigation of human cell lines revealed that similarly to murine cells, TNF-mediated necroptosis is decreased by inhibition with CQ and BafA1 in JurkatFADD⫺⁄⫺⁄TNFR2⫹⁄⫹ cells (Fig. 7A), but TRAIL-mediated necroptosis is diminished only with BafA1 and might even be increased when inhibitors of autophagy (3-MA) or lysosomal flux (CQ) are applied (Fig. 3F). These data suggest clear differences in the executions of TRAIL- and TNF-mediated necroptosis that might have future clinical consequences. Atg5 but not Atg16L1 or receptor internalization promotes TRAIL-mediated necroptosis. We have demonstrated that TRAIL-induced necroptosis is a signaling pathway distinct from autophagy; however, some studies suggest that necrosome assem-

2634

mcb.asm.org

bly occurs on autophagosomal membranes, e.g., after treatment with obatoclax, an inhibitor of the antiapoptotic protein Bcl-2 (53), or with the multityrosine kinase inhibitor sorafenib (54). An involvement of the autophagic machinery during necroptosis is still controversial, as necroptosis was shown to be independent of autophagy in murine T cells lacking caspase-8 and FADD (55), but murine T cells deficient for caspase-8 and expressing dominant negative FADD undergo necroptosis via the activation of autophagy by RIPK1 (56). An interaction of Atg5 with caspase-8, FADD, and RIPK1 was proposed previously for T cell receptormediated T cell activation (57) but was not shown to play a role in TRAIL- or TNF-mediated necroptosis. Here, we confirm the pivotal role of Atg5 for both types of necroptosis (Fig. 7C and D), as Atg5-deficient cells were protected from TRAIL-zVAD-CHX-, TNF-zVAD-, and TNF-zVAD-CHX-induced necroptosis. These effects were reversed in Atg5-deficient cells by the reexpression of wild-type Atg5 (Atg5RE) (Fig. 7C and D), probably due to the obligatory role of Atg5 in the formation of necrosome membranes that aggregate RIPK1 and RIPK3 (53). Moreover, we investigated whether there is further cross talk between p38␣ and Atg5, as it was described previously that p38␣ phosphorylates Atg5 at threonine 75 to inhibit autophagy. Here, we showed that preventing the phosphorylation of T75 on Atg5, which leads to enhanced basal autophagy (26), did not enhance the level of TRAIL- or TNF-mediated necroptosis (Fig. 7C and D), confirming that p38␣-mediated phosphorylation of Atg5 does not execute TRAIL- and TNF-induced necroptosis but that solely the presence of Atg5 is pivotal for necroptosis. Furthermore, considering Atg5 to associate with RIPK1 and Atg16L (56), the putative role of Atg16L1 during TRAIL- and TNF-mediated necroptosis was addressed. Surprisingly, only TNF-mediated and not TRAIL-mediated necroptosis was abrogated by the lack of Atg16L1 (Fig. 7E and F). However, the protective effects of Atg16L1 were completely abolished in the presence of CHX, suggesting that Atg16L1, in contrast to Atg5, has to first be synthesized in order to be involved in TNF-mediated necroptosis. Of note, TRAIL-mediated necroptosis (irrespective of protein synthesis inhibition by CHX) was rather boosted by the lack of Atg16L1 (Fig. 7E and F), as was all TRAIL-, TNF-, TRAIL-CHX-, and TNF-CHX-mediated apoptosis. All effects were only poorly reversed by the reexpression of wild-type Atg16L1 (Atg16L1RE) (Fig. 7E), as the transfection procedure for MEFs lacking Atg16L1 was toxic and presented low efficacy. In MEFs possessing or not possessing Atg5 or Atg16L1, TRAIL- and TNF-induced necroptosis was inhibited by the addition of the RIPK1 inhibitor Nec-1s, in contrast to TRAIL- and TNF-induced apoptosis, which were unaffected by the absence of Atg5 or Atg16L1 or the addition of Nec-1s (Fig. 7C and E). In addition to the distinct roles of Atg5 and Atg16L1 in the formation of signaling membranes and indications of the internalization of TNF-R1 during apoptosis (58), the role of receptor internalization during the initiation of TRAIL-mediated necroptosis in comparison to TNF-mediated necroptosis was analyzed. For this purpose, PitStop2, an inhibitor of clathrin-independent endocytosis (59), and dynasore, an inhibitor of dynamin essential for clathrin-dependent coated-vesicle formation (60), were applied. Neither PitStop2 nor dynasore inhibited TRAIL-mediated necroptosis, but both inhibitors abrogated TNF-mediated necroptosis (Fig. 7G). This finding suggests that clathrin-dependent as well as clathrin-independent events and TNF internalization are responsible for TNF-mediated necroptosis. On the con-



FIG 6 Ubiquitinated, active UCH-L1 is not involved in the execution of TRAIL-mediated necroptosis. (A) Cells were prestimulated for 3 h with the indicated concentrations of the UCH-L1 inhibitor LDN57444 (top) or LDN91946 (bottom), with the subsequent addition of either 30 ng/ml killerTRAIL and 20 ␮M zVAD for 14 h; 100 ng/ml hrTNF and 20 ␮M zVAD for 5 h (L929Ts); 50 ng/ml killerTRAIL, 50 ␮M zVAD, and 2 ␮g/ml CHX for 20 h (Jurkat); or 30 ng/ml killerTRAIL, 20 ␮M zVAD, and 5 ␮g/ml CHX for 16 h (HT-29). (B) L929Ts cells were treated as described above for panel A, and afterwards, necroptotic changes in morphology (arrowheads) were observed by microscopy. Bar, 100 ␮m. (C) Cells were transfected with siRNAs specific for murine UCH-L1 (siRNA) or a nontargeting control siRNA (NT). At 24 h posttransfection, cells were stimulated with 30 ng/ml killerTRAIL and 20 ␮M zVAD for 14 h. Measurement of intracellular ATP levels served as an indicator of cell viability. Shown are means ⫾ standard deviations (n ⫽ 4 [A] and n ⫽ 3 [C], each with five repetitions, relative to untreated cells), with statistical significance at P values of ⬍0.05 (*), ⬍0.01 (**), and ⬍0.001 (***) (as determined by a t test). n.s., nonsignificant. Control Western blots show downregulation of endogenous murine UCH-L1 (PAb) and ␤-actin as a loading control. (D) MEFs deficient for HtrA2/Omi and their wild-type counterparts were left untreated or treated with 100 ng/ml hrTNF or 30 ng/ml killerTRAIL, 20 ␮M zVAD, and 1 ␮g/ml CHX for 16 h. * and *** indicate the disappearance of the main form of UCH-L1 (MAb and PAb, respectively). ** indicates the appearance of an active, ubiquitinated form of UCH-L1 (PAb). (E) Cells were treated as described above for panel D, omitting hrTNF, for the indicated times. Further PAbs for UCH-L1 were used to analyze the appearance of the active form of UCH-L1. WB, Western blot.



mcb.asm.org

2635

Sosna et al.

FIG 7 Continued on following page

2636

mcb.asm.org






mcb.asm.org

2637

Sosna et al.

trary, TRAIL-mediated necroptosis is independent of receptor internalization and signaled distinctly from TNF-mediated necroptosis. Data from the internalization study were quantified by calculating the percentages of the ligand-receptor complexes that were not associated with cell surface staining (Fig. 7H). Simultaneously, we observed that plasma membrane (Fig. 7H, red)localized TNF receptor-ligand complexes (green) were clearly internalized alone or in combination with the caspase inhibitor zVAD. On the contrary, TRAIL receptor-ligand complexes (Fig. 7I, green) stayed on the cell surface (red) irrespective of the induction of TRAIL-mediated apoptosis or TRAIL-zVAD-mediated necroptosis. The data described above suggest that autophagy per se does not play a role in necroptosis but rather that certain elements of the autophagic machinery might interact with components of the necroptotic pathway and are differently involved in the execution of TRAIL- or TNF-mediated necroptosis. Furthermore, TNF receptor but not TRAIL is internalized during both caspase-dependent and -independent cell death, indicating distinct mechanisms of cell death induction for both cytokines. PARP-1 is not involved in TRAIL-mediated necroptosis. Necroptosis mediated by TNF was previously described to be independent of and distinct from PARP-1 activation (11). On the contrary, TRAIL was previously described to induce necroptosis dependent on PARP-1 under acidic-pH conditions (31). Three PARP-1 inhibitors, PJ-34, 3-AB, and olaparib, rescued cells from the formation of TRAIL-zVAD-mediated PAR chains in a concentration-dependent manner (Fig. 8A). Furthermore, although the PARP-1 inhibitor PJ-34 protected murine cells from TRAIL-mediated necroptosis, other potent inhibitors of PARP-1, such as 3-AB or olaparib, did not rescue cells from this type of cell death (Fig. 8B). This was additionally demonstrated by the downregulation of PARP-1, which did not protect cells from TRAIL-mediated necroptosis (Fig. 8C). In line with these data, knockout of

PARP-1 did not abolish TRAIL-mediated necroptosis and even elevated the level of cell death (Fig. 8D), comparably to the downregulation of PARP-1 (Fig. 8C). Indeed, the inhibition of PARP-1 in some cell lines (such as HT-29) increased the level of TRAILmediated necroptosis (Fig. 8B), suggesting that PARP-1 inhibitors might be used for combinatory therapy in order to amplify TRAIL-mediated necroptosis during the elimination of cancer cells. Further analyses confirmed that the activation of PARP-1 by the DNA-damaging agent N-methyl-N=-nitro-N-nitrosoguanidine (MNNG) follows a rapid (30 min after induction of cell death) loss of intracellular ATP (Fig. 8E), in contrast to TRAILmediated necroptosis, which was characterized by a slower reduction of ATP together with an increased level of cell death. Similarly, the loss of intracellular NAD⫹ was not an early event during TRAIL-mediated necroptosis and was accompanied by an increasing loss of membrane integrity (Fig. 8F). Taken together, these data suggest that under conditions of physiological pH, PARP-1 does not execute TRAIL-mediated necroptosis, but targeting members of the poly(ADP-ribose) polymerase family such as PARP-1 might be useful for increasing the level of TRAIL-mediated necroptosis. Accordingly, we investigated whether the inhibition of further members of the poly(ADP-ribose) polymerase family, such as tankyrases 1 and 2, by XAV939 and WIKI4, respectively, similar to the inhibition of PARP-1, might increase the level of TRAIL-mediated necroptosis. Consistent with the above-described results, both inhibitors of tankyrases enhanced the level of TRAIL-mediated necroptosis (Fig. 8G), suggesting an as-yet-unexplored role of the Wnt/␤-catenin signaling pathway in TRAILmediated necroptosis, which might be targeted for future combinatory therapy. DISCUSSION

TRAIL is known to have no effect on healthy tissue, while it selectively induces apoptosis in cancer cells in vitro and in vivo. Because

FIG 7 (A and B) Autophagy does not play a crucial role in TNF- or TRAIL-mediated necroptosis in L929Ts or Jurkat I.42 cells. (A) L929ATCC cells were prestimulated for 2 h with the indicated concentrations of inhibitors of autophagy (3-MA), lysosome formation (CQ), or vesicle acidification (BafA1) and stimulated afterwards for 24 h with 20 ␮M zVAD. L929Ts cells were pretreated for 2 h with 3-MA, CQ, or BafA1 and stimulated with 20 ␮M zVAD, 30 ng/ml killerTRAIL for 14 h, or 100 ng/ml hrTNF for 5 h. Jurkat I.42 (FADD-deficient and TNF-R2-positive) cells were pretreated for 2 h with the indicated concentrations of 3-MA, CQ, or BafA1 before the addition of 50 ␮M zVAD in combination with 100 ng/ml hrTNF for 6 h. (B) Morphological analyses of the influence of 3-MA, CQ, or BafA1 on zVAD-induced autophagy in L929ATCC cells (top) and TRAIL-mediated (middle) and TNF-mediated (bottom) necroptosis in L929Ts cells. Cells were stimulated as described above for panel A. Bar, 100 ␮m. (C) Atg5 is required for both TRAIL- and TNF-mediated RIPK1dependent necroptosis. Atg5-deficient immortalized MEFs, their wild-type counterparts, and Atg5-deficient cells retransfected with wild-type Atg5 (Atg5RE) or mutated Atg5 (Atg5T75A) were prestimulated or not with 50 ␮M Nec-1s for 2 h, followed by the addition of 30 ng/ml killerTRAIL or 100 ng/ml hrTNF, 20 ␮M zVAD, and/or 1 ␮g/ml CHX to induce necroptosis or 100 ng/ml hrTNF or 30 ng/ml killerTRAIL together with 1 ␮g/ml CHX to induce apoptosis for 18 h. In order to verify their authenticity, all cell lines were left untreated in full medium or treated for 2 h with Earle’s balanced salt solution (EB SS) to induce autophagy by starvation accompanied by the appearance of LC3-II as an indicator of autophagy. Insets show Western blots for LC3 under normal and starvation conditions (Earle’s balanced salt solution), with ␤-actin serving as a loading control. (D) Morphological analyses of necroptosis in Atg5-deficient cells retransfected with wild-type Atg5 (Atg5RE), mutated Atg5 (Atg5T75A), and their wild-type counterparts. Cells were stimulated as described above for panel C. The bar in microphotographs is 100 ␮m. (E) Lack of Atg16L1 protects cells from TNF-mediated but not TRAIL-mediated necroptosis. Atg16L1-deficient immortalized MEFs, their wild-type counterparts, and Atg16L1-deficient cells transiently retransfected for 48 h with wild-type Atg16L1 (Atg16L1RE) were prestimulated or not with 50 ␮M Nec-1s for 2 h, followed by stimulation with 30 ng/ml killerTRAIL or 100 ng/ml hrTNF together with 20 ␮M zVAD and/or 1 ␮g/ml CHX to induce necroptosis or with 30 ng/ml killerTRAIL or 100 ng/ml hrTNF together with 1 ␮g/ml CHX to induce apoptosis for 24 h. The authenticity of all cell lines and efficiency of transfection with Atg16L1 were verified by Western blotting, with ␤-actin serving as a loading control. (F) Morphological analyses of necroptosis in Atg16L1-deficient and wild-type MEFs and MEFs retransfected with wild-type Atg16L1 (Atg16L1RE). Cells were stimulated as described above for panel E. Arrowheads indicate typical necroptotic morphology. Bar, 100 ␮m. (G) Receptor internalization is necessary for TNF-mediated but not for TRAIL-mediated necroptosis. L929Ts cells were prestimulated for 1 h with the indicated concentrations of PitStop2 or dynasore and stimulated with 30 ng/ml killerTRAIL and 20 ␮M zVAD or with 100 ng/ml hrTNF and 20 ␮M zVAD. Shown are means ⫾ standard deviations (n ⫽ 6 for L929Ts and n ⫽ 4 for Jurkat I.42 cells [A], n ⫽ 3 [C], and n ⫽ 6 [A]), and data were considered significant at a P value of ⬍0.001 (***) (as determined by a t test). (H) TNF but not TRAIL receptor is internalized during necroptosis in L929Ts cells. Cell surface-bound receptors were coupled to Fc-TNF or Fc-TRAIL ligands labeled with protein G-Alexa Fluor 488 alone or in combination with 20 ␮M zVAD at 4°C, followed by a temperature shift to 37°C for 30 min. A minimum of 10,000 cells were analyzed for each sample. Internalization Wizard was used to measure the ratio of the intensity inside the cell to the intensity of the entire cell (percent). (I) Representative images of cells (bright field) with a stained plasma membrane and TNF or TRAIL ligand-receptor complexes. Colocalization of ligand-receptor complexes with the plasma membrane (yellow) and internalization of TNF ligand-receptor complexes but not TRAIL ligand-receptor complexes were observed after 30 min of stimulation.

2638

mcb.asm.org




FIG 8 (A to F) TRAIL-mediated necroptosis is not mediated by PARP-1. (A) L929Ts cells were pretreated for 2 h with the indicated concentrations of PARP-1 inhibitors and treated with 30 ng/ml killerTRAIL and 20 ␮M zVAD for 14 h. Furthermore, total lysates were prepared, and the quantitative presence of PAR chains was analyzed by Western blotting, with ␤-actin serving as a loading control. (B) Cells were pretreated as described above for panel A and stimulated with 30 ng/ml (14 h for L929Ts cells and 16 h for HT-29), 50 ng/ml (20 h for Jurkat cells), or 100 ng/ml (16 h for NIH 3T3 cells) killerTRAIL in combination with 20 ␮M (L929Ts, NIH 3T3, and HT-29) or 50 ng/ml (Jurkat) zVAD and additionally with 2 ␮g/ml (Jurkat) or 5 ␮g/ml (HT-29) of CHX. (C) L929Ts or Jurkat cells were transfected using nucleofection (nucleofected) with siRNA that does not target any RNA sequence (NT) or with siRNA specific for murine or human PARP-1 (siRNA). Seventy-two hours after transfection, the cells were stimulated with 30 ng/ml killerTRAIL and 20 ␮M zVAD (5 h for L929Ts cells) or 50 ng/ml killerTRAIL, 50 ␮M zVAD, and 2 ␮g/ml CHX (24 h for Jurkat cells). (D) Wild-type (WT) and PARP-1-deficient MEFs were stimulated with 50 ng/ml killerTRAIL, 20 ␮M zVAD, and 1 ␮g/ml CHX for 24 h. Insets show Western blots for PARP-1, with ␤-actin serving as a loading control. (E and F) L929Ts, NIH 3T3, Jurkat, and HT-29 cells were stimulated with 0.5 mM MNNG for 15 min and left in fresh medium without MNNG as a control for the indicated periods of time or were stimulated with 30 ng/ml (L929Ts and HT-29), 50 ng/ml (Jurkat), or 100 ng/ml (NIH 3T3) killerTRAIL in combination with 20 ␮M (L929Ts, NIH 3T3, and HT-29) or 50 ␮M (Jurkat) zVAD and additionally with 2 ␮g/ml (Jurkat) or 5 ␮g/ml (HT-29) of CHX for the indicated periods of time before their intracellular ATP content (E) and intracellular NAD⫹ content and loss of membrane integrity as determined by PI staining (F) were measured. (G) TRAILmediated necroptosis is increased by WIKI4 and XAV939. HT-29 cells were prestimulated for 2 h with the indicated concentrations of inhibitors and stimulated with 100 ng/ml killerTRAIL, 50 ␮M zVAD, and 5 ␮g/ml CHX for 24 h. Shown are means ⫾ standard deviations (n ⫽ 3 [B], n ⫽ 6 [C], n ⫽ 9 [D], and n ⫽ 2 [E and F]), with differences being considered significant at P values of ⬍0.05 (*), ⬍0.01 (**), and ⬍0.001 (***) (as determined by a t test). n.s., nonsignificant.

of the low efficacy of TRAIL-based antitumor monotherapy and increasing resistance of TRAIL-induced apoptotic pathways during tumor evolution (61), alternative TRAIL-based but apoptosisindependent therapies are needed (reviewed in reference 62). Therefore, TRAIL-induced necroptosis became a promising pathway for the elimination of a broad range of cancer cells (5). However, TRAIL-induced necroptosis might be activated by synergistic stimulation, i.e., after TRAIL receptor ligation during exposition of cancer cells to the acidic extracellular milieu (31), the combination of TRAIL with the chemotherapeutic drug sirtinol (histone deacetylases inhibitor) (63), or the sensitization of tumor cells with various protein synthesis inhibitors (64) (e.g., the antileukemia drug homoharringtonine) under caspase-compromised conditions in combination with TRAIL (6). Nonetheless, only TNF-induced necroptosis is intensively studied, although the execution of TNF-mediated necroptosis is associated with systemic inflammation (65) and cannot be used as an alternative in anticancer therapy. Therefore, the signaling mechanisms that may be used for the regulation of TRAIL-induced necroptosis in cancer cells require further investigation. Here, we present novel differ-


ences in signaling between TRAIL- and TNF-mediated necroptosis (Fig. 9), which may be used as new molecular targets for the development of potentiated TRAIL-specific, innovative, experimental therapies. Of note, current clinical trials involve a rather “weak” recombinant form of soluble TRAIL or agonistic antibodies that specifically target TRAIL receptors, which have only a modest agonistic capacity to render effective TRAIL-based therapy (reviewed in references 66–68). In our study, we used soluble TRAIL possessing a linker peptide that promotes trimerization to form a more stable oligomer and mimics the natural clustering of TRAIL receptors on the cell surface. Therefore, it is important to design potential new TRAIL-based treatments with TRAIL of increased stability and improved in vivo performance (see reference 66). For the first time, we show that death receptor-induced necroptosis, depending on which death receptors are triggered, may be initiated through distinct mechanisms. Internalization of the receptor is crucial only for TNF-mediated and not for TRAILmediated necroptosis, as a disturbance of clathrin-dependent and -independent signaling did not affect TRAIL-mediated necropto-


mcb.asm.org

2639

Sosna et al.

FIG 9 Overview of similarities and differences in signaling pathways of TRAIL- and TNF-mediated necroptosis in cancer cells. Death receptors such as TRAIL-R1/2 or TNF-R1 under caspase-compromised conditions (and, for some cancer cell lines, after sensitization with protein synthesis inhibitors) are able to mediate necroptosis. During necroptosis death receptors may be (for TNF-R1) internalized in a clathrin-dependent or clathrin-independent manner followed by receptosome formation or (for TRAIL-R1/R2) signaled without receptor internalization directly through formation of a necrosomal initiation complex. In the course of the first steps of necroptosis, membranes and membrane-bound molecules (e.g., Atg5 for TNF-R1 and TRAIL-R1/R2 and Atg16L1 for TNF-R1) are recruited to the necrosomal initiation complex consisting of, e.g., FADD (indispensable for TRAIL-R1/2 but not for TNF-R1) and, crucial for the execution of both TNF- and TRAIL-mediated necroptosis, proteins RIPK1 and RIPK3 (not shown), which are assembled in a filamentous fibril-like manner. Deficiency in some membrane-bound proteins such as Atg5 may inhibit the execution of TRAIL- and TNF-induced necroptosis. However, deficiency in other proteins such as Atg16L1 or inhibition of clathrin-dependent and -independent events reduces the level of TNF-induced necroptosis but enhances or has no influence, respectively, on TRAIL-induced necroptosis. Further phosphorylation events for RIPK1, RIPK3, and its downstream effector MLKL are necessary to execute necroptosis (not shown). The lack of the adaptor protein FADD within the necrosomal initiation complex abrogates the execution of TRAIL-induced necroptosis (Fig. 1G), but it potentiates the execution of TNF-induced necroptosis (88). As a consequence of necroptosis initiation, secondary messengers (i.e., ceramide) are produced by A-SMase and N-SMase as intracellular signals to promote necroptosis. As a result, a myriad of executive mechanisms in various cellular compartments is promoted to accomplish the execution of necroptosis. Inhibition of prosurvival pathways such as PARP-1, tankyrases (analyzed here only for TRAIL), and p38␣ leads to enhancement of TRAIL- and TNF-induced necroptosis. The executive, necroptotic pathway is built up by some common mechanisms, shared by both TRAIL- and TNF-induced necroptosis, that result in the same outcome, while they have been modulated (e.g., through inhibition, deficiency, or overexpression). However, among those executive mechanisms, some differences existed in TRAIL- or TNF-induced necroptosis. While for TNF-induced necroptosis inhibition of certain executive components led to a decrease in the level of cell death, e.g., through inhibition of lysosomal acidification and inhibition of UCH-L1, for TRAIL-induced necroptosis, on the contrary, modulation of those components led to an increase in or had no influence on the level of necroptosis. Along the way, some unique executive mechanisms were identified to play a role exclusively in TNF-induced necroptosis, such as monoubiquitination of UCH-L1. Moreover, overexpression of Bcl-XL did not influence TNF-induced necroptosis, but it reduced TRAIL-mediated necroptosis. The involvement of particular signaling molecules in the promotion or inhibition of TRAIL- and TNF-mediated necroptosis is described in detail in Discussion.

sis. Furthermore, the lack of Atg16L1 attenuated the execution of only TNF-induced necroptosis, as Atg16L1 may associate with clathrin heavy chains to form early autophagosome precursors (69) and internalize (independently from autophagy) clathrinbound vesicle structures together with the TNF receptor. On the

2640

mcb.asm.org

contrary, deficiencies in Atg16L1-regulated endocytosis and further exocytosis pathways (25) potentiated TRAIL-induced necroptosis, probably through sustained cellular stress. On the other hand, the role of Atg5 was common to both TRAIL- and TNF-mediated necroptosis, probably due to the fact that Atg5




might control further downstream signaling events, e.g., through its interaction with FADD (70), recruitment to caspase-8 (56), and control of membrane formation and complete membrane closure (27). Further differences in signaling between TRAIL- and TNFmediated necroptosis that might be used for modification of TRAIL-based anticancer therapies were shown for the involvement of lysosomal and vacuolar compartments in the execution of necroptosis. Vacuolar H⫹-ATPase played an executive role in both types of necroptosis, as a role for Atp6v1g2 (ATPase, H⫹transporting, lysosomal V1 subunit G2), a multisubunit enzyme that mediates the acidification of intracellular compartments, in TNF-mediated necroptosis was described previously (71). Vacuolar H⫹-ATPase may contribute to necroptosis by the maturation or acidification of cargo vesicles, promoting endogenous protein degradation (72) or receptor endocytosis (73). However, the suppression of the function of lysosomal enzymes by the lysosomotropic agent CQ inhibited only TNF-induced necroptosis and either did not influence or rather enhanced the execution of TRAIL-induced necroptosis in a cell type-specific manner. This may be due to a dependence of TRAIL-induced necroptosis on executive mechanisms other than lysosomal acidification, as CQ was previously identified to mediate mitochondrial protein degradation (74). For TRAIL-induced necroptosis, we identified the serine protease HtrA2/Omi as a common executioner of necroptosis, similarly to caspase-independent necrosis induced by imatinib (75) and TNF (9, 76). Although the mechanism of how Omi/HtrA2 mediates TRAIL-mediated necroptosis requires further investigation, we demonstrated that HtrA2/Omi-dependent monoubiquitination of UCH-L1 was present during TNF- but not TRAILinduced necroptosis. Importantly, the application of inhibitors specific for UCH-L1 exacerbates rather than ameliorates TRAILmediated necroptosis. Thus, influencing UCH-L1 during TRAILmediated necroptosis might rather be a potentiation factor and suggests that this unique point might be used to develop rational combination therapies using this signal transduction modulator. There are several similar uniformly regulatory mechanisms for TRAIL- and TNF-mediated necroptosis (Fig. 9). Both TRAIL- and TNF-mediated necroptosis might be pharmacologically modulated by radicicol and geldanamycin, as they are already applied for treatment of neuronal injury (77) or sepsis (78). In addition, bisphosphonate-based A-SMase inhibitors or N-SMase inhibitors may be used for the therapeutic control of necroptosis-based pathophysiological events or TRAIL- and TNF-mediated necroptosis. Further therapeutic regulation of TRAIL- and TNF-mediated necroptosis may be achieved by the application of Ca-074 Me, similarly as it protected U937 cells from staurosporine-induced necroptotic conditions (14). Ca-074 Me may decrease the execution of necroptosis by decreasing the accumulation of amyloid fibers (79), which have been described to promote the oligomerization of the pivotal signaling molecules RIPK1 and RIPK3 during necroptosis (80). Both TRAIL- and TNF-mediated necroptosis can be attenuated by the overexpression of Bcl-2. However, differences in signaling between TRAIL- and TNF-mediated necroptosis were found for the overexpression of another Bcl-2 family member, Bcl-XL. We were able to show that Bcl-XL overexpression decreased the levels of TRAIL-induced necroptosis, but it did not influence TNF-mediated necroptosis. The detailed mechanism of these differences requires further studies. Surprisingly, another group presented data showing that the over-


expression of Bcl-XL might likewise prevent the execution of TNF-mediated necroptosis (81). We consider that differences between our results and those reported previously by Irrinki and coworkers might be caused by different levels of Bcl-XL expression in completely different cell lines (MEFs in the study by Irrinki et al. and human pancreatic adenosquamous carcinoma Colo357 cells in our study). Of note, the influence of Bcl-XL on the execution of TRAIL-induced necroptosis is reasonable, since PGAM5, which promotes necroptosis (82), was identified as an interaction partner of Bcl-XL (83). Taken together, Bcl-2 and Bcl-XL levels in cancer cells might be prognostic markers for the application of TRAIL-mediated necroptotic anticancer therapy, as increased levels of these proteins imply a rather ineffective outcome of therapy. In order to reach an advantageous enhancement of TRAILinduced necroptosis during anticancer therapy, inhibition of prosurvival pathways should be applied (Fig. 9). Here, the lack of p38␣ augmented the necroptotic response to both TRAIL and TNF, probably due to the activation of cell death pathways by NF-␬B, as its prodeath activities under certain conditions and in certain cell types were described previously (reviewed in references 84 and 85). Although Ye and coworkers previously showed that suppression of the p38 –NF-␬B survival signaling pathway promoted TNF-induced necroptosis in L929Ts cells (86), we confirmed only that the suppression of p38␣ increased necroptosis. On the contrary, we have found that a deficiency of p38␣ caused the activation of NF-␬B, accompanied by an increased level of necroptosis. These discrepancies may be due to analyses of NF-␬B degradation performed by Ye and coworkers at only late time points, without direct analyses of early degradation of I␬B␣ or activation of p65 during cell death in p38␣-deficient cells. Therefore, high levels of p38␣ in highly proliferating tumor cells (87) might be disadvantageous, and rather, decreasing the level of p38␣ might enhance TRAIL-based antitumor therapy. To confer the prosurvival effect of p38␣ during necroptosis and enhance the elimination of cancer cells, a combinatory therapy of TRAIL-mediated necroptosis and p38␣ inhibitors in phase II/III clinical trials should be considered. Similarly, we showed that inhibition of PARP-1 or associated components, e.g., tankyrases, may be a therapeutic target to potentiate the execution of TRAIL-induced necroptosis in cancer cells. Importantly, our results demonstrate that TRAIL-mediated necroptosis, similarly to TNF-induced necroptosis (11), is not executed through the PARP-1 pathway in the physiological milieu. This may have important clinical implications, as the inhibition of PARP-1 does not abolish TRAIL-induced necroptosis but, as shown here, rather enhances caspaseindependent cell death in some cancer cell lines. Taken together, the results of our study for the first time reveal details of various signaling pathways of TRAIL-mediated necroptosis that were not studied previously. Additionally, we found differences in signaling between TRAIL- and TNF-mediated necroptosis that might be used for the potentiation of anticancer approaches based on TRAIL-mediated necroptosis. ACKNOWLEDGMENTS We thank Sabine Mathieu-Grützmacher and Parvin Davarnia for excellent technical assistance. We declare that we have no competing interests. J.S., S.P., J.F., C.A., T.P., H.K., A.T., I.S., and D.A. designed research; J.S., S.P., J.F.C., C.S., J.F., A.F., and J.P. performed research; J.S., S.P., J.F.C., C.S., J.F., A.F., J.P., I.S., H.K., A.T., S.S., and D.A. analyzed data;


mcb.asm.org

2641

Sosna et al.

and J.S. and D.A. wrote the manuscript. All authors read and approved the final manuscript. This work was supported by Deutsche Forschungsgemeinschaft SFB 877, project B2, Cluster of Excellence Inflammation at Interfaces EXC306PMTP1 and EXC306-PWTP2 to D.A.; SFB 877, project B1, Cluster of Excellence Inflammation at Interfaces EXC306-WTP4 to S.S.; DFG grant SCHU733/14-1 to S.S. and J.F.; Deutsche Forschungsgemeinschaft grant SCHM1586/3-1 to I.S.; fellowship A/08/79433 from the German Academic Exchange Service DAAD and Klara und Werner Kreitz Stiftung to J.S.; CAU Forschungsförderung 2014-Junior to S.P.; and a Deutsche Krebshilfe grant to D.A. and H.K. (110055).

FUNDING INFORMATION This work, including the efforts of Dieter Adam, was funded by Cluster of Excellence Inflammation at Interfaces (EXC306-PMTP1). This work, including the efforts of Justyna Sosna, was funded by German Academic Exchange Service (A/08/79433). This work, including the efforts of Justyna Sosna, was funded by Werner und Klara Kreitz Stiftung. This work, including the efforts of Stephan Philipp, was funded by CAU Forschungsfoerderung (Junior). This work, including the efforts of Dieter Adam, was funded by Cluster of Excellence Inflammation at Interfaces (EXC306-PWTP2). This work, including the efforts of Stefan Schütze, was funded by Cluster of Excellence Inflammation at Interfaces (EXC306WTP4). This work, including the efforts of Dieter Adam, was funded by Deutsche Forschungsgemeinschaft (DFG) (SFB 877/B2). This work, including the efforts of Stefan Schütze, was funded by Deutsche Forschungsgemeinschaft (DFG) (SFB877/B1). This work, including the efforts of Stefan Schütze, was funded by Deutsche Forschungsgemeinschaft (DFG) (SCHU733/14-1). This work, including the efforts of Jürgen Fritsch, was funded by Deutsche Forschungsgemeinschaft (DFG) (SCHU733/14-1). This work, including the efforts of Ingo Schmitz, was funded by Deutsche Forschungsgemeinschaft (DFG) (SCHM1586/3-1). This work, including the efforts of Dieter Adam, was funded by Deutsche Krebshilfe (German Cancer Aid) (110055). This work, including the efforts of Holger Kalthoff, was funded by Deutsche Krebshilfe (German Cancer Aid) (110055).

REFERENCES 1. Jin Z, El-Deiry WS. 2005. Overview of cell death signaling pathways. Cancer Biol Ther 4:139 –163. 2. Kroemer G, Galluzzi L, Vandenabeele P, Abrams J, Alnemri ES, Baehrecke EH, Blagosklonny MV, El-Deiry WS, Golstein P, Green DR, Hengartner M, Knight RA, Kumar S, Lipton SA, Malorni W, Nunez G, Peter ME, Tschopp J, Yuan J, Piacentini M, Zhivotovsky B, Melino G, Nomenclature Committee on Cell Death 2009. 2009. Classification of cell death: recommendations of the Nomenclature Committee on Cell Death 2009. Cell Death Differ 16:3–11. http://dx.doi.org/10.1038/cdd .2008.150. 3. Galluzzi L, Vitale I, Abrams JM, Alnemri ES, Baehrecke EH, Blagosklonny MV, Dawson TM, Dawson VL, El-Deiry WS, Fulda S, Gottlieb E, Green DR, Hengartner MO, Kepp O, Knight RA, Kumar S, Lipton SA, Lu X, Madeo F, Malorni W, Mehlen P, Nunez G, Peter ME, Piacentini M, Rubinsztein DC, Shi Y, Simon HU, Vandenabeele P, White E, Yuan J, Zhivotovsky B, Melino G, Kroemer G. 2012. Molecular definitions of cell death subroutines: recommendations of the Nomenclature Committee on Cell Death 2012. Cell Death Differ 19:107–120. http://dx.doi.org/10 .1038/cdd.2011.96. 4. Kaczmarek A, Vandenabeele P, Krysko DV. 2013. Necroptosis: the release of damage-associated molecular patterns and its physiological relevance. Immunity 38:209 –223. http://dx.doi.org/10.1016/j.immuni.2013 .02.003. 5. Voigt S, Philipp S, Davarnia P, Winoto-Morbach S, Roder C, Arenz C, Trauzold A, Kabelitz D, Schutze S, Kalthoff H, Adam D. 2014. TRAILinduced programmed necrosis as a novel approach to eliminate tumor cells. BMC Cancer 14:74. http://dx.doi.org/10.1186/1471-2407-14-74. 6. Philipp S, Sosna J, Plenge J, Kalthoff H, Adam D. 2015. Homoharringtonine, a clinically approved anti-leukemia drug, sensitizes tumor cells for TRAIL-induced necroptosis. Cell Commun Signal 13:25. http: //dx.doi.org/10.1186/s12964-015-0103-0.

2642

mcb.asm.org

7. Almasan A, Ashkenazi A. 2003. Apo2L/TRAIL: apoptosis signaling, biology, and potential for cancer therapy. Cytokine Growth Factor Rev 14: 337–348. 8. Liu S, Wang X, Li Y, Xu L, Yu X, Ge L, Li J, Zhu Y, He S. 2014. Necroptosis mediates TNF-induced toxicity of hippocampal neurons. Biomed Res Int 2014:290182. http://dx.doi.org/10.1155/2014/290182. 9. Sosna J, Voigt S, Mathieu S, Kabelitz D, Trad A, Janssen O, MeyerSchwesinger C, Schutze S, Adam D. 2013. The proteases HtrA2/Omi and UCH-L1 regulate TNF-induced necroptosis. Cell Commun Signal 11:76. http://dx.doi.org/10.1186/1478-811X-11-76. 10. Thon L, Mohlig H, Mathieu S, Lange A, Bulanova E, Winoto-Morbach S, Schutze S, Bulfone-Paus S, Adam D. 2005. Ceramide mediates caspase-independent programmed cell death. FASEB J 19:1945–1956. http://dx.doi.org/10.1096/fj.05-3726com. 11. Sosna J, Voigt S, Mathieu S, Lange A, Thon L, Davarnia P, Herdegen T, Linkermann A, Rittger A, Chan FK, Kabelitz D, Schutze S, Adam D. 2014. TNF-induced necroptosis and PARP-1-mediated necrosis represent distinct routes to programmed necrotic cell death. Cell Mol Life Sci 71: 331–348. http://dx.doi.org/10.1007/s00018-013-1381-6. 12. Fernald K, Kurokawa M. 2013. Evading apoptosis in cancer. Trends Cell Biol 23:620 – 633. http://dx.doi.org/10.1016/j.tcb.2013.07.006. 13. Han W, Li L, Qiu S, Lu Q, Pan Q, Gu Y, Luo J, Hu X. 2007. Shikonin circumvents cancer drug resistance by induction of a necroptotic death. Mol Cancer Ther 6:1641–1649. http://dx.doi.org/10.1158/1535-7163 .MCT-06-0511. 14. Dunai ZA, Imre G, Barna G, Korcsmaros T, Petak I, Bauer PI, Mihalik R. 2012. Staurosporine induces necroptotic cell death under caspasecompromised conditions in U937 cells. PLoS One 7:e41945. http://dx.doi .org/10.1371/journal.pone.0041945. 15. Karl I, Jossberger-Werner M, Schmidt N, Horn S, Goebeler M, Leverkus M, Wajant H, Giner T. 2014. TRAF2 inhibits TRAIL- and CD95Linduced apoptosis and necroptosis. Cell Death Dis 5:e1444. http://dx.doi .org/10.1038/cddis.2014.404. 16. Thon L, Mathieu S, Kabelitz D, Adam D. 2006. The murine TRAIL receptor signals caspase-independent cell death through ceramide. Exp Cell Res 312:3808 –3821. http://dx.doi.org/10.1016/j.yexcr.2006.08.017. 17. Arenz C. 2010. Small molecule inhibitors of acid sphingomyelinase. Cell Physiol Biochem 26:1– 8. http://dx.doi.org/10.1159/000315100. 18. Roth AG, Redmer S, Arenz C. 2009. Potent inhibition of acid sphingomyelinase by phosphoinositide analogues. Chembiochem 10:2367–2374. http://dx.doi.org/10.1002/cbic.200900281. 19. Chan FK, Shisler J, Bixby JG, Felices M, Zheng L, Appel M, Orenstein J, Moss B, Lenardo MJ. 2003. A role for tumor necrosis factor receptor-2 and receptor-interacting protein in programmed necrosis and antiviral responses. J Biol Chem 278:51613–51621. http://dx.doi.org/10.1074/jbc .M305633200. 20. Hinz S, Trauzold A, Boenicke L, Sandberg C, Beckmann S, Bayer E, Walczak H, Kalthoff H, Ungefroren H. 2000. Bcl-XL protects pancreatic adenocarcinoma cells against CD95- and TRAIL-receptormediated apoptosis. Oncogene 19:5477–5486. http://dx.doi.org/10 .1038/sj.onc.1203936. 21. Lehnert L, Lerch MM, Hirai Y, Kruse ML, Schmiegel W, Kalthoff H. 2001. Autocrine stimulation of human pancreatic duct-like development by soluble isoforms of epimorphin in vitro. J Cell Biol 152:911–922. http: //dx.doi.org/10.1083/jcb.152.5.911. 22. Zhang DW, Zheng M, Zhao J, Li YY, Huang Z, Li Z, Han J. 2011. Multiple death pathways in TNF-treated fibroblasts: RIP3- and RIP1dependent and independent routes. Cell Res 21:368 –371. http://dx.doi .org/10.1038/cr.2011.3. 23. Martins LM, Morrison A, Klupsch K, Fedele V, Moisoi N, Teismann P, Abuin A, Grau E, Geppert M, Livi GP, Creasy CL, Martin A, Hargreaves I, Heales SJ, Okada H, Brandner S, Schulz JB, Mak T, Downward J. 2004. Neuroprotective role of the Reaper-related serine protease HtrA2/ Omi revealed by targeted deletion in mice. Mol Cell Biol 24:9848 –9862. http://dx.doi.org/10.1128/MCB.24.22.9848-9862.2004. 24. Munoz-Gamez JA, Rodriguez-Vargas JM, Quiles-Perez R, AguilarQuesada R, Martin-Oliva D, de Murcia G, Menissier de Murcia J, Almendros A, Ruiz de Almodovar M, Oliver FJ. 2009. PARP-1 is involved in autophagy induced by DNA damage. Autophagy 5:61–74. http: //dx.doi.org/10.4161/auto.5.1.7272. 25. Cadwell K, Liu JY, Brown SL, Miyoshi H, Loh J, Lennerz JK, Kishi C, Kc W, Carrero JA, Hunt S, Stone CD, Brunt EM, Xavier RJ, Sleckman BP, Li E, Mizushima N, Stappenbeck TS, Virgin HW, IV. 2008. A key




26.

27.

28.

29. 30.

31.

32.

33.

34.

35.

36. 37.

38.

39.

40.

role for autophagy and the autophagy gene Atg16l1 in mouse and human intestinal Paneth cells. Nature 456:259 –263. http://dx.doi.org/10.1038 /nature07416. Keil E, Hocker R, Schuster M, Essmann F, Ueffing N, Hoffman B, Liebermann DA, Pfeffer K, Schulze-Osthoff K, Schmitz I. 2013. Phosphorylation of Atg5 by the Gadd45beta-MEKK4-p38 pathway inhibits autophagy. Cell Death Differ 20:321–332. http://dx.doi.org/10.1038/cdd .2012.129. Mizushima N, Yamamoto A, Hatano M, Kobayashi Y, Kabeya Y, Suzuki K, Tokuhisa T, Ohsumi Y, Yoshimori T. 2001. Dissection of autophagosome formation using Apg5-deficient mouse embryonic stem cells. J Cell Biol 152:657– 668. http://dx.doi.org/10.1083/jcb.152.4.657. Porras A, Zuluaga S, Black E, Valladares A, Alvarez AM, Ambrosino C, Benito M, Nebreda AR. 2004. P38 alpha mitogen-activated protein kinase sensitizes cells to apoptosis induced by different stimuli. Mol Biol Cell 15:922–933. Kitson RR, Moody CJ. 2013. Learning from nature: advances in geldanamycin- and radicicol-based inhibitors of Hsp90. J Org Chem 78:5117– 5141. http://dx.doi.org/10.1021/jo4002849. Festjens N, Vanden Berghe T, Vandenabeele P. 2006. Necrosis, a wellorchestrated form of cell demise: signalling cascades, important mediators and concomitant immune response. Biochim Biophys Acta 1757:1371– 1387. http://dx.doi.org/10.1016/j.bbabio.2006.06.014. Jouan-Lanhouet S, Arshad MI, Piquet-Pellorce C, Martin-Chouly C, Le Moigne-Muller G, Van Herreweghe F, Takahashi N, Sergent O, Lagadic-Gossmann D, Vandenabeele P, Samson M, Dimanche-Boitrel MT. 2012. TRAIL induces necroptosis involving RIPK1/RIPK3-dependent PARP-1 activation. Cell Death Differ 19:2003–2014. http://dx.doi.org/10 .1038/cdd.2012.90. Holler N, Zaru R, Micheau O, Thome M, Attinger A, Valitutti S, Bodmer JL, Schneider P, Seed B, Tschopp J. 2000. Fas triggers an alternative, caspase-8-independent cell death pathway using the kinase RIP as effector molecule. Nat Immunol 1:489 – 495. http://dx.doi.org/10 .1038/82732. Rozman-Pungercar J, Kopitar-Jerala N, Bogyo M, Turk D, Vasiljeva O, Stefe I, Vandenabeele P, Bromme D, Puizdar V, Fonovic M, TrstenjakPrebanda M, Dolenc I, Turk V, Turk B. 2003. Inhibition of papain-like cysteine proteases and legumain by caspase-specific inhibitors: when reaction mechanism is more important than specificity. Cell Death Differ 10:881– 888. http://dx.doi.org/10.1038/sj.cdd.4401247. Tichy ED, Stephan ZA, Osterburg A, Noel G, Stambrook PJ. 2013. Mouse embryonic stem cells undergo charontosis, a novel programmed cell death pathway dependent upon cathepsins, p53, and EndoG, in response to etoposide treatment. Stem Cell Res 10:428 – 441. http://dx.doi .org/10.1016/j.scr.2013.01.010. Jung M, Lee J, Seo HY, Lim JS, Kim EK. 2015. Cathepsin inhibitioninduced lysosomal dysfunction enhances pancreatic beta-cell apoptosis in high glucose. PLoS One 10:e0116972. http://dx.doi.org/10.1371/journal .pone.0116972. Dietmar S. 2011. The cathepsin B-selective inhibitors CA-074 and CA074Me inactivate cathepsin L under reducing conditions. Open Enzym Inhib J 4:11–16. http://dx.doi.org/10.2174/1874940201104010011. Mihalik R, Imre G, Petak I, Szende B, Kopper L. 2004. Cathepsin B-independent abrogation of cell death by CA-074-OMe upstream of lysosomal breakdown. Cell Death Differ 11:1357–1360. http://dx.doi.org/10 .1038/sj.cdd.4401493. Dumitru CA, Carpinteiro A, Trarbach T, Hengge UR, Gulbins E. 2007. Doxorubicin enhances TRAIL-induced cell death via ceramide-enriched membrane platforms. Apoptosis 12:1533–1541. http://dx.doi.org/10.1007 /s10495-007-0081-9. Berdyshev EV, Gorshkova I, Skobeleva A, Bittman R, Lu X, Dudek SM, Mirzapoiazova T, Garcia JG, Natarajan V. 2009. FTY720 inhibits ceramide synthases and up-regulates dihydrosphingosine 1-phosphate formation in human lung endothelial cells. J Biol Chem 284:5467–5477. http: //dx.doi.org/10.1074/jbc.M805186200. Saddoughi SA, Gencer S, Peterson YK, Ward KE, Mukhopadhyay A, Oaks J, Bielawski J, Szulc ZM, Thomas RJ, Selvam SP, Senkal CE, Garrett-Mayer E, De Palma RM, Fedarovich D, Liu A, Habib AA, Stahelin RV, Perrotti D, Ogretmen B. 2013. Sphingosine analogue drug FTY720 targets I2PP2A/SET and mediates lung tumour suppression via activation of PP2A-RIPK1-dependent necroptosis. EMBO Mol Med 5:105–121. http://dx.doi.org/10.1002/emmm.201201283.


41. Marshall KD, Baines CP. 2014. Necroptosis: is there a role for mitochondria? Front Physiol 5:323. http://dx.doi.org/10.3389/fphys.2014.00323. 42. Vercammen D, Brouckaert G, Denecker G, Van de Craen M, Declercq W, Fiers W, Vandenabeele P. 1998. Dual signaling of the Fas receptor: initiation of both apoptotic and necrotic cell death pathways. J Exp Med 188:919 –930. http://dx.doi.org/10.1084/jem.188.5.919. 43. Zhang M, Harashima N, Moritani T, Huang W, Harada M. 2015. The roles of ROS and caspases in TRAIL-induced apoptosis and necroptosis in human pancreatic cancer cells. PLoS One 10:e0127386. http://dx.doi.org /10.1371/journal.pone.0127386. 44. Choi HS, Kang JW, Lee SM. 2015. Melatonin attenuates carbon tetrachloride-induced liver fibrosis via inhibition of necroptosis. Transl Res 166:292–303. http://dx.doi.org/10.1016/j.trsl.2015.04.002. 45. Emerling BM, Platanias LC, Black E, Nebreda AR, Davis RJ, Chandel NS. 2005. Mitochondrial reactive oxygen species activation of p38 mitogen-activated protein kinase is required for hypoxia signaling. Mol Cell Biol 25:4853– 4862. http://dx.doi.org/10.1128/MCB.25.12 .4853-4862.2005. 46. Luo JL, Kamata H, Karin M. 2005. IKK/NF-kappaB signaling: balancing life and death—a new approach to cancer therapy. J Clin Invest 115:2625– 2632. http://dx.doi.org/10.1172/JCI26322. 47. Sasaki CY, Barberi TJ, Ghosh P, Longo DL. 2005. Phosphorylation of RelA/p65 on serine 536 defines an IkappaBalpha-independent NF-kappaB pathway. J Biol Chem 280:34538 –34547. http://dx.doi.org/10.1074/jbc .M504943200. 48. Huang B, Yang XD, Lamb A, Chen LF. 2010. Posttranslational modifications of NF-kappaB: another layer of regulation for NF-kappaB signaling pathway. Cell Signal 22:1282–1290. http://dx.doi.org/10.1016/j.cellsig .2010.03.017. 49. Sakurai H, Suzuki S, Kawasaki N, Nakano H, Okazaki T, Chino A, Doi T, Saiki I. 2003. Tumor necrosis factor-alpha-induced IKK phosphorylation of NF-kappaB p65 on serine 536 is mediated through the TRAF2, TRAF5, and TAK1 signaling pathway. J Biol Chem 278:36916 –36923. http://dx.doi.org/10.1074/jbc.M301598200. 50. De Chiara G, Marcocci ME, Torcia M, Lucibello M, Rosini P, Bonini P, Higashimoto Y, Damonte G, Armirotti A, Amodei S, Palamara AT, Russo T, Garaci E, Cozzolino F. 2006. Bcl-2 phosphorylation by p38 MAPK: identification of target sites and biologic consequences. J Biol Chem 281:21353–21361. http://dx.doi.org/10.1074/jbc.M511052200. 51. Cartier AE, Djakovic SN, Salehi A, Wilson SM, Masliah E, Patrick GN. 2009. Regulation of synaptic structure by ubiquitin C-terminal hydrolase L1. J Neurosci 29:7857–7868. http://dx.doi.org/10.1523 /JNEUROSCI.1817-09.2009. 52. Ch’en IL, Tsau JS, Molkentin JD, Komatsu M, Hedrick SM. 2011. Mechanisms of necroptosis in T cells. J Exp Med 208:633– 641. http://dx .doi.org/10.1084/jem.20110251. 53. Basit F, Cristofanon S, Fulda S. 2013. Obatoclax (GX15-070) triggers necroptosis by promoting the assembly of the necrosome on autophagosomal membranes. Cell Death Differ 20:1161–1173. http://dx.doi.org/10 .1038/cdd.2013.45. 54. Kharaziha P, Chioureas D, Baltatzis G, Fonseca P, Rodriguez P, Gogvadze V, Lennartsson L, Bjorklund AC, Zhivotovsky B, Grander D, Egevad L, Nilsson S, Panaretakis T. 2015. Sorafenib-induced defective autophagy promotes cell death by necroptosis. Oncotarget 6:37066 – 37082. http://dx.doi.org/10.18632/oncotarget.5797. 55. Osborn SL, Diehl G, Han SJ, Xue L, Kurd N, Hsieh K, Cado D, Robey EA, Winoto A. 2010. Fas-associated death domain (FADD) is a negative regulator of T-cell receptor-mediated necroptosis. Proc Natl Acad Sci U S A 107:13034 –13039. http://dx.doi.org/10.1073/pnas.1005997107. 56. Bell BD, Leverrier S, Weist BM, Newton RH, Arechiga AF, Luhrs KA, Morrissette NS, Walsh CM. 2008. FADD and caspase-8 control the outcome of autophagic signaling in proliferating T cells. Proc Natl Acad Sci U S A 105:16677–16682. http://dx.doi.org/10.1073/pnas.0808597105. 57. Ch’en IL, Beisner DR, Degterev A, Lynch C, Yuan J, Hoffmann A, Hedrick SM. 2008. Antigen-mediated T cell expansion regulated by parallel pathways of death. Proc Natl Acad Sci U S A 105:17463–17468. http: //dx.doi.org/10.1073/pnas.0808043105. 58. Schneider-Brachert W, Tchikov V, Neumeyer J, Jakob M, WinotoMorbach S, Held-Feindt J, Heinrich M, Merkel O, Ehrenschwender M, Adam D, Mentlein R, Kabelitz D, Schutze S. 2004. Compartmentalization of TNF receptor 1 signaling: internalized TNF receptosomes as death signaling vesicles. Immunity 21:415– 428. http://dx.doi.org/10 .1016/S1074-7613(04)00234-1.


mcb.asm.org

2643

Sosna et al.

59. Dutta D, Williamson CD, Cole NB, Donaldson JG. 2012. Pitstop 2 is a potent inhibitor of clathrin-independent endocytosis. PLoS One 7:e45799. http://dx.doi.org/10.1371/journal.pone.0045799. 60. Macia E, Ehrlich M, Massol R, Boucrot E, Brunner C, Kirchhausen T. 2006. Dynasore, a cell-permeable inhibitor of dynamin. Dev Cell 10:839 – 850. http://dx.doi.org/10.1016/j.devcel.2006.04.002. 61. Stuckey DW, Shah K. 2013. TRAIL on trial: preclinical advances in cancer therapy. Trends Mol Med 19:685– 694. http://dx.doi.org/10.1016/j .molmed.2013.08.007. 62. Philipp S, Sosna J, Adam D. 2016. Cancer and necroptosis: friend or foe? Cell Mol Life Sci 73:2183–2193. http://dx.doi.org/10.1007/s00018-016 -2193-2. 63. Pal S, Shankar BS, Sainis KB. 2013. Cytokines from the tumor microenvironment modulate sirtinol cytotoxicity in A549 lung carcinoma cells. Cytokine 64:196 –207. http://dx.doi.org/10.1016/j.cyto.2013.07.029. 64. Chan J, Khan SN, Harvey I, Merrick W, Pelletier J. 2004. Eukaryotic protein synthesis inhibitors identified by comparison of cytotoxicity profiles. RNA 10:528 –543. http://dx.doi.org/10.1261/rna.5200204. 65. Zhou W, Yuan J. 2014. Necroptosis in health and diseases. Semin Cell Dev Biol 35:14 –23. http://dx.doi.org/10.1016/j.semcdb.2014.07.013. 66. de Miguel D, Lemke J, Anel A, Walczak H, Martinez-Lostao L. 2016. Onto better TRAILs for cancer treatment. Cell Death Differ 23:733–747. http://dx.doi.org/10.1038/cdd.2015.174. 67. Lemke J, von Karstedt S, Zinngrebe J, Walczak H. 2014. Getting TRAIL back on track for cancer therapy. Cell Death Differ 21:1350 –1364. http: //dx.doi.org/10.1038/cdd.2014.81. 68. Wajant H. 2015. Principles of antibody-mediated TNF receptor activation. Cell Death Differ 22:1727–1741. http://dx.doi.org/10.1038/cdd.2015 .109. 69. Puri C, Renna M, Bento CF, Moreau K, Rubinsztein DC. 2013. Diverse autophagosome membrane sources coalesce in recycling endosomes. Cell 154:1285–1299. http://dx.doi.org/10.1016/j.cell.2013.08.044. 70. Pyo JO, Jang MH, Kwon YK, Lee HJ, Jun JI, Woo HN, Cho DH, Choi B, Lee H, Kim JH, Mizushima N, Oshumi Y, Jung YK. 2005. Essential roles of Atg5 and FADD in autophagic cell death: dissection of autophagic cell death into vacuole formation and cell death. J Biol Chem 280:20722– 20729. http://dx.doi.org/10.1074/jbc.M413934200. 71. Hitomi J, Christofferson DE, Ng A, Yao J, Degterev A, Xavier RJ, Yuan J. 2008. Identification of a molecular signaling network that regulates a cellular necrotic cell death pathway. Cell 135:1311–1323. http://dx.doi.org /10.1016/j.cell.2008.10.044. 72. Yamamoto A, Tagawa Y, Yoshimori T, Moriyama Y, Masaki R, Tashiro Y. 1998. Bafilomycin A1 prevents maturation of autophagic vacuoles by inhibiting fusion between autophagosomes and lysosomes in rat hepatoma cell line, H-4-II-E cells. Cell Struct Funct 23:33– 42. 73. Jefferies KC, Cipriano DJ, Forgac M. 2008. Function, structure and regulation of the vacuolar (H⫹)-ATPases. Arch Biochem Biophys 476:33– 42. http://dx.doi.org/10.1016/j.abb.2008.03.025. 74. Webster BR, Scott I, Han K, Li JH, Lu Z, Stevens MV, Malide D, Chen Y, Samsel L, Connelly PS, Daniels MP, McCoy JP, Jr, Combs CA, Gucek M, Sack MN. 2013. Restricted mitochondrial protein acetylation initiates mitochondrial autophagy. J Cell Sci 126:4843– 4849. http://dx.doi.org/10 .1242/jcs.131300. 75. Okada M, Adachi S, Imai T, Watanabe K, Toyokuni SY, Ueno M, Zervos AS, Kroemer G, Nakahata T. 2004. A novel mechanism for imatinib mesylate-induced cell death of BCR-ABL-positive human leuke-

2644

mcb.asm.org

76.

77.

78.

79.

80.

81.

82.

83.

84.

85. 86.

87. 88.

mic cells: caspase-independent, necrosis-like programmed cell death mediated by serine protease activity. Blood 103:2299 –2307. http://dx.doi.org /10.1182/blood-2003-05-1605. Blink E, Maianski NA, Alnemri ES, Zervos AS, Roos D, Kuijpers TW. 2004. Intramitochondrial serine protease activity of Omi/HtrA2 is required for caspase-independent cell death of human neutrophils. Cell Death Differ 11:937–939. http://dx.doi.org/10.1038/sj.cdd.4401409. Chen WW, Yu H, Fan HB, Zhang CC, Zhang M, Zhang C, Cheng Y, Kong J, Liu CF, Geng D, Xu X. 2012. RIP1 mediates the protection of geldanamycin on neuronal injury induced by oxygen-glucose deprivation combined with zVAD in primary cortical neurons. J Neurochem 120:70 – 77. http://dx.doi.org/10.1111/j.1471-4159.2011.07526.x. Manaenko A, Lekic T, Ma Q, Ostrowski RP, Zhang JH, Tang J. 2011. Hydrogen inhalation is neuroprotective and improves functional outcomes in mice after intracerebral hemorrhage. Acta Neurochir Suppl 111: 179 –183. http://dx.doi.org/10.1007/978-3-7091-0693-8_30. Cho K, Yoon SY, Choi JE, Kang HJ, Jang HY, Kim DH. 2013. CA074Me, a cathepsin B inhibitor, decreases APP accumulation and protects primary rat cortical neurons treated with okadaic acid. Neurosci Lett 548: 222–227. http://dx.doi.org/10.1016/j.neulet.2013.05.056. Wu XN, Yang ZH, Wang XK, Zhang Y, Wan H, Song Y, Chen X, Shao J, Han J. 2014. Distinct roles of RIP1-RIP3 hetero- and RIP3-RIP3 homointeraction in mediating necroptosis. Cell Death Differ 21:1709 –1720. http://dx.doi.org/10.1038/cdd.2014.77. Irrinki KM, Mallilankaraman K, Thapa RJ, Chandramoorthy HC, Smith FJ, Jog NR, Gandhirajan RK, Kelsen SG, Houser SR, May MJ, Balachandran S, Madesh M. 2011. Requirement of FADD, NEMO, and BAX/BAK for aberrant mitochondrial function in tumor necrosis factor alpha-induced necrosis. Mol Cell Biol 31:3745–3758. http://dx.doi.org/10 .1128/MCB.05303-11. Wang Z, Jiang H, Chen S, Du F, Wang X. 2012. The mitochondrial phosphatase PGAM5 functions at the convergence point of multiple necrotic death pathways. Cell 148:228 –243. http://dx.doi.org/10.1016/j.cell .2011.11.030. Lo SC, Hannink M. 2006. PGAM5, a Bcl-XL-interacting protein, is a novel substrate for the redox-regulated Keap1-dependent ubiquitin ligase complex. J Biol Chem 281:37893–37903. http://dx.doi.org/10.1074/jbc .M606539200. Fan Y, Dutta J, Gupta N, Fan G, Gelinas C. 2008. Regulation of programmed cell death by NF-kappaB and its role in tumorigenesis and therapy. Adv Exp Med Biol 615:223–250. http://dx.doi.org/10.1007/978-1 -4020-6554-5_11. Ch’en IL, Hedrick SM, Hoffmann A. 2008. NF-kappaB as a determinant of distinct cell death pathways. Methods Enzymol 446:175–187. http://dx .doi.org/10.1016/S0076-6879(08)01610-8. Ye YC, Yu L, Wang HJ, Tashiro S, Onodera S, Ikejima T. 2011. TNFalpha-induced necroptosis and autophagy via supression [sic] of the p38-NF-kappaB survival pathway in L929 cells. J Pharmacol Sci 117:160 – 169. http://dx.doi.org/10.1254/jphs.11105FP. Faust D, Dolado I, Cuadrado A, Oesch F, Weiss C, Nebreda AR, Dietrich C. 2005. p38alpha MAPK is required for contact inhibition. Oncogene 24:7941–7945. http://dx.doi.org/10.1038/sj.onc.1208948. Christofferson DE, Yuan J. 2010. Necroptosis as an alternative form of programmed cell death. Curr Opin Cell Biol 22:263–268. http://dx.doi.org /10.1016/j.ceb.2009.12.003.



Necroptotic signaling in adaptive and innate immunity.

Restoring TRAIL mediated signaling in ovarian cancer cells.

Effect of interleukin-1 on lipid metabolism in the rat. Similarities to and differences from tumor necrosis factor.

Tumor necrosis factor, tumor necrosis factor inhibition, and cancer risk.

CD4 and Tumor Necrosis Factor-Related Apoptosis Ligand (TRAIL) localization in HIV-stimulated plasmacytoid dendritic cells by three-dimensional microscopy.

Sensitization of tumor necrosis factor-related apoptosis-inducing ligand (TRAIL)-resistant primary prostate cancer cells by isoegomaketone from Perilla frutescens.

Gli1 signaling in human breast cancer cells.

TRAIL-induced programmed necrosis as a novel approach to eliminate tumor cells.

Mechanism of dimethylnitrosamine and carbon tetrachloride-induced liver necrosis: similarities and differences.

Regulatory differences and similarities: systems in Japan.

Similarities and differences between the multidrug resistance phenotype of mammalian tumor cells and chloroquine resistance in Plasmodium falciparum.

Prenatal arsenic exposure and shifts in the newborn proteome: interindividual differences in tumor necrosis factor (TNF)-responsive signaling.

The statins: similarities and differences.

Carbon tetrachloride induced ultrastructural alterations in pancreatic acinar cells and in the hepatocytes. Similarities and differences.

Inhibition of NF-κB Pathway and Modulation of MAPK Signaling Pathways in Glioblastoma and Implications for Lovastatin and Tumor Necrosis Factor-Related Apoptosis Inducing Ligand (TRAIL) Combination Therapy.

Notch4 Signaling Confers Susceptibility to TRAIL-Induced Apoptosis in Breast Cancer Cells.

Activation of KIT modulates the function of tumor necrosis factor-related apoptosis-inducing ligand receptor (TRAIL-R) in mast cells.

Dominant negative effects of tumor necrosis factor (TNF)-related apoptosis-inducing ligand (TRAIL) receptor 4 on TRAIL receptor 1 signaling by formation of heteromeric complexes.

Chromosomal imbalances in human bladder urothelial carcinoma: similarities and differences between biopsy samples and cancer stem-like cells.

tTregs, pTregs, and iTregs: similarities and differences.

Depolarization Controls TRAIL-Sensitization and Tumor-Selective Killing of Cancer Cells: Crosstalk with ROS.

Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) signaling and cell death in the immature central nervous system after hypoxia-ischemia and inflammation.

Aβ42 and Aβ40: similarities and differences.

TRAIL and TRAIL receptors in patients with laryngeal cancer.