Role of caspases in CD95-induced biphasic activation of acid sphingomyelinase.

Acid sphingomyelinase (A-SMase) plays an important role in the initiation of CD95 signaling by forming ceramide-enriched membrane domains that enable clustering and activation of the death receptors. In TNF-R1 and TRAIL-R1/R2 signaling, A-SMase also contributes to the lysosomal apoptosis pathway triggered by receptor internalization. Here, we investigated the molecular mechanism of CD95-mediated A-SMase activation, demonstrating that A-SMase is located in internalized CD95-receptosomes and is activated by the CD95/CD95L complex in a biphasic manner.Since several caspases have been described to be involved in the activation of A-SMase, we evaluated expression levels of caspase-8, caspase-7 and caspase-3 in CD95-receptosomes. The occurrence of cleaved caspase-8 correlated with the first peak of A-SMase activity and translocation of the A-SMase to the cell surface which could be blocked by the caspase-8 inhibitor IETD.Inhibition of CD95-internalization selectively reduced the second phase of A-SMase activity, suggesting a fusion between internalized CD95-receptosomes and an intracellular vesicular pool of A-SMase. Further analysis demonstrated that caspase-7 activity correlates with the second phase of the A-SMase activity, whereas active caspase-3 is present at early and late internalization time points. Blocking caspases-7/ -3 by DEVD reduced the second phase of A-SMase activation in CD95-receptosomes suggesting the potential role of caspase-7 or -3 for late A-SMase activation.In summary, we describe a biphasic A-SMase activation in CD95-receptosomes indicating (I.) a caspase-8 dependent translocation of A-SMase to plasma membrane and (II.) a caspase-7 and/or -3 dependent fusion of internalized CD95-receptosomes with intracellular A-SMase-containing vesicles.

Nuclear death receptor TRAIL-R2 inhibits maturation of let-7 and promotes proliferation of pancreatic and other tumor cells.

BACKGROUND & AIMS: Tumor necrosis factor-related apoptosis inducing ligand (TRAIL-R1) (TNFRSF10A) and TRAIL-R2 (TNFRSF10B) on the plasma membrane bind ligands that activate apoptotic and other signaling pathways. Cancer cells also might have TRAIL-R2 in the cytoplasm or nucleus, although little is known about its activities in these locations. We investigated the functions of nuclear TRAIL-R2 in cancer cell lines. METHODS: Proteins that interact with TRAIL-R2 initially were identified in pancreatic cancer cells by immunoprecipitation, mass spectrometry, and immunofluorescence analyses. Findings were validated in colon, renal, lung, and breast cancer cells. Functions of TRAIL-R2 were determined from small interfering RNA knockdown, real-time polymerase chain reaction, Drosha-activity, microRNA array, proliferation, differentiation, and immunoblot experiments. We assessed the effects of TRAIL-R2 overexpression or knockdown in human pancreatic ductal adenocarcinoma (PDAC) cells and their ability to form tumors in mice. We also analyzed levels of TRAIL-R2 in sections of PDACs and non-neoplastic peritumoral ducts from patients. RESULTS: TRAIL-R2 was found to interact with the core microprocessor components Drosha and DGCR8 and the associated regulatory proteins p68, hnRNPA1, NF45, and NF90 in nuclei of PDAC and other tumor cells. Knockdown of TRAIL-R2 increased Drosha-mediated processing of the let-7 microRNA precursor primary let-7 (resulting in increased levels of mature let-7), reduced levels of the let-7 targets (LIN28B and HMGA2), and inhibited cell proliferation. PDAC tissues from patients had higher levels of nuclear TRAIL-R2 than non-neoplastic pancreatic tissue, which correlated with increased nuclear levels of HMGA2 and poor outcomes. Knockdown of TRAIL-R2 in PDAC cells slowed their growth as orthotopic tumors in mice. Reduced nuclear levels of TRAIL-R2 in cultured pancreatic epithelial cells promoted their differentiation. CONCLUSIONS: Nuclear TRAIL-R2 inhibits maturation of the microRNA let-7 in pancreatic cancer cell lines and increases their proliferation. Pancreatic tumor samples have increased levels of nuclear TRAIL-R2, which correlate with poor outcome of patients. These findings indicate that in the nucleus, death receptors can function as tumor promoters and might be therapeutic targets.

Subcellular compartmentalization of TNF receptor-1 and CD95 signaling pathways.

Receptors that belong to the family of death-receptors including TNF receptor-1 (TNF-R1), CD95 (Fas, APO-1) and TRAIL receptors (TRAIL-R1, TRAIL R2/DR4/DR5) transduce signals resulting in entirely different biological outcomes: They promote cell death via apoptosis but are also capable of inducing anti-apoptotic signals through the transcription factor nuclear factor NF-κB or activation of the proliferative MAPK/ERK protein kinase cascade resulting in cell protection and tissue regeneration. Recent findings revealed a regulatory role of receptor internalization and its intracellular trafficking in selectively transmitting signals that lead either to apoptosis or to the survival of the cell, providing a clue to the understanding of these contradictory biological phenomena. In this chapter we review our data obtained during the Collaborative Research Center 415 (CRC 415) focusing on the compartmentalization of TNF-R1 and CD95 pro and anti-apoptotic signaling. We will address the role of internalization in determining the fate of the receptors. We suggest that fusion of internalized TNF-receptosomes with trans-Golgi vesicles is a novel mechanism to transduce death signals along the endosomal trafficking route. The roles of acid sphingomyelinase, the lipid second messenger ceramide, and the aspartate-protease cathepsin D as novel players in the cell death scenario is also highlighted. We report on the regulation of NF-κB signaling by recruitment of the endosomal E3-ubiquitin ligases CARP-2 and CARP-1 during TNF-receptosome trafficking. The biological significance of TNF receptor-1 compartmentalization is demonstrated by the strategy of adenoviruses to impede TNF-R1 internalization and by this preventing host cell apoptosis.

Caspase-8 and caspase-7 sequentially mediate proteolytic activation of acid sphingomyelinase in TNF-R1 receptosomes.

We previously demonstrated that tumour necrosis factor (TNF)-induced ceramide production by endosomal acid sphingomyelinase (A-SMase) couples to apoptosis signalling via activation of cathepsin D and cleavage of Bid, resulting in caspase-9 and caspase-3 activation. The mechanism of TNF-mediated A-SMase activation within the endolysosomal compartment is poorly defined. Here, we show that TNF-induced A-SMase activation depends on functional caspase-8 and caspase-7 expression. The active forms of all three enzymes, caspase-8, caspase-7 and A-SMase, but not caspase-3, colocalize in internalized TNF receptosomes. While caspase-8 and caspase-3 are unable to induce activation of purified pro-A-SMase, we found that caspase-7 mediates A-SMase activation by direct interaction resulting in proteolytic cleavage of the 72-kDa pro-A-SMase zymogen at the non-canonical cleavage site after aspartate 253, generating an active 57 kDa A-SMase molecule. Caspase-7 down modulation revealed the functional link between caspase-7 and A-SMase, confirming proteolytic cleavage as one further mode of A-SMase activation. Our data suggest a signalling cascade within TNF receptosomes involving sequential activation of caspase-8 and caspase-7 for induction of A-SMase activation by proteolytic cleavage of pro-A-SMase.

TRAIL signaling is mediated by DR4 in pancreatic tumor cells despite the expression of functional DR5.

Tumor necrosis factor related apoptosis-inducing ligand (TRAIL) and agonistic anti-DR4/TRAIL-R1 and anti-DR5/TRAIL-R2 antibodies are currently under clinical investigation for treatment of different malignancies. TRAIL activates DR4 and DR5 and thereby triggers apoptotic and non-apoptotic signaling pathways, but possible different roles of DR4 or DR5 in these responses has poorly been addressed so far. In the present work, we analyzed cell viability, DISC formation as well as IL-8 and NF-kappaB activation side by side in responses to TRAIL and agonistic antibodies against DR4 (mapatumumab) and against DR5 (lexatumumab) in pancreatic ductal adenocarcinoma cells. We found that all three reagents are able to activate cell death and pro-inflammatory signaling. Death-inducing signaling complex (DISC) analysis revealed that mapatumumab and lexatumumab induce formation of homocomplexes of either DR4 or DR5, whereas TRAIL additionally stimulated the formation of heterocomplexes of both receptors. Notably, blocking of receptors using DR4- and DR5-specific Fab fragments indicated that TRAIL exerted its function predominantly via DR4. Interestingly, inhibition of PKC by Goe6983 enabled DR5 to trigger apoptotic signaling in response to TRAIL and also strongly enhanced lexatumumab-mediated cell death. Our results suggest the existence of mechanisms that silence DR5 for TRAIL- but not for agonistic-antibody treatment.

Different structural stability and toxicity of PrP(ARR) and PrP(ARQ) sheep prion protein variants.

The polymorphisms at amino acid residues 136, 154, and 171 in ovine prion protein (PrP) have been associated with different susceptibility to scrapie: animals expressing PrP(ARQ) [PrP(Ala136/Arg154/Gln171)] show vulnerability, whereas those that express PrP(ARR) [PrP(Ala136/Arg154/Arg171)] are resistant to scrapie. The aim of this study was to evaluate the in vitro toxic effects of PrP(ARR) and PrP(ARQ) variants in relation with their structural characteristics. We show that both peptides cause cell death inducing apoptosis but, unexpectedly, the scrapie resistant PrP(ARR) form was more toxic than the scrapie susceptible PrP(ARQ) variant. Moreover, the alpha-helical conformation of PrP(ARR) was less stable than that of PrP(ARQ) and the structural determinants responsible of these different conformational stabilities were characterized by spectroscopic analysis. We observed that PrP toxicity was inversely related to protein structural stability, being the unfolded conformation more toxic than the native one. However, the PrP(ARQ) variant displays a higher propensity to form large aggregates than PrP(ARR). Interestingly, in the presence of small amounts of PrP(ARR), PrP(ARQ) aggregability was reduced to levels similar to that of PrP(ARR). Thus, in contrast to PrP(ARR) toxicity, scrapie transmissibility seems to reside in the more stable conformation of PrP(ARQ) that allows the formation of large amyloid fibrils.

Identification of the actin-binding domain of Ins(1,4,5)P3 3-kinase isoform B (IP3K-B).

Dewaste et al. [Dewaste, Moreau, De Smedt, Bex, De Smedt, Wuytaack, Missiaen and Erneux (2003) Biochem. J. 374, 41-49] showed that over-expressed EGFP (enhanced green fluorescent protein) fused to Ins(1,4,5)P3 3-kinase B (IP3K-B) co-localizes with the cytoskeleton, as well as with the endoplasmic reticulum and the plasma membrane. The domains responsible for these subcellular localizations are not yet identified. For the endogenous enzyme, we confirmed both actin and endoplasmic reticulum localization by employing a high affinity antibody against IP3K-B. F-actin targeting is exclusively dependent on the non-catalytic N-terminal region of IP3K-B. By expressing fragments of this N-terminal domain as EGFP-fusion proteins and inspecting transfected cells by confocal microscopy, we characterized a distinct 63-amino-acid domain comprising amino acids 108-170 of the enzyme which is responsible for F-actin targeting. A truncation of this fragment from both sides revealed that the full size of this segment is essential for this function. Deletion of this segment in a full-length over-expressed IP3K-B-EGFP-fusion protein completely abolished F-actin interaction. Direct interaction of this actin-binding segment with only F-actin, but not with G-actin, was observed in vitro using a bacterially expressed, affinity-purified GST (glutathione S-transferase)-Rattus norvegicus IP3K (aa 108-170) fusion protein. Helix-breaking mutations within this isolated segment abolished the F-actin binding properties both in vitro and when over-expressed in cells, indicating that an intact secondary structure is essential for actin targeting. The segment shows sequence similarities to the actin-binding region in IP3K-A, but no similarity to other actin-binding domains.