Search

Metalloprotease-processed CD95L (cl-CD95L) is a soluble cytokine that implements a PI3K/Ca(2+) signaling pathway in triple-negative breast cancer (TNBC) cells. Accordingly, high levels of cl-CD95L in TNBC women correlate with poor prognosis, and administration of this ligand in an orthotopic xenograft mouse model accelerates the metastatic dissemination of TNBC cells. The molecular mechanism underlying CD95-mediated cell migration remains unknown. Here, we present genetic and pharmacologic evidence that the anti-apoptotic molecules BclxL and Bcl-2 and the pro-apoptotic factors BAD and BID cooperate to promote migration of TNBC cells stimulated with cl-CD95L. BclxL was distributed in both endoplasmic reticulum (ER) and mitochondrion membranes. The mitochondrion-localized isoform promoted cell migration by interacting with voltage-dependent anion channel 1 to orchestrate Ca(2+) transfer from the ER to mitochondria in a BH3-dependent manner. Mitochondrial Ca(2+) uniporter contributed to this flux, which favored ATP production and cell migration. In conclusion, this study reveals a novel molecular mechanism controlled by BclxL to promote cancer cell migration and supports the use of BH3 mimetics as therapeutic options not only to kill tumor cells but also to prevent metastatic dissemination in TNBCs.

CD95 is a death receptor whose stimulation by either the physiologic ligand CD95L or the agonistic antibodies leads to the formation of a multi-molecular complex termed DISC (death-inducing signaling complex) and the subsequent induction of a caspase-driven apoptotic signal. According to the magnitude of the DISC formation, two types of cells have been identified. Although type I cells generate an important DISC, the complex is barely found in type II cells. Analyzing the early stages preceding the DISC formation, we found that unlike CD95L, the commonly used agonistic antibody APO1-3 internalized the death receptor. Using inhibitors of actin polymerization, we showed that the remodeling of the actin cytoskeleton did not alter the capping of the CD95 receptor or its partitioning into the lipid rafts. In addition, whereas the disruption of F-actin prevented the internalization of CD95, the DISC formation and the apoptotic signal induced by the agonistic antibody APO1-3 in type I cells, it did not affect the signal triggered by the soluble and membrane-bound CD95L, regardless of the type of cells. In conclusion, the addition of APO1-3 on type I cells triggers an actin-dependent apoptotic signal, which is absent or marginal in cells (both types I and II) treated with CD95L.

FasL and TRAIL are apoptotic ligands of the TNF-like cytokines family, acting via activation of the transmembrane death domain containing receptors Fas for FasL, and DR4 or DR5 for TRAIL. A glycosylphosphatidylinositol-linked TRAIL receptor called DcR1 behaves as a decoy receptor inhibiting TRAIL-mediated cell death in several cellular systems. We engineered and stably expressed a chimeric GPI-linked Fas receptor (Fas-GPI) in T-lymphocyte cell lines constitutively expressing functional transmembrane Fas. Surprisingly, despite lacking the death domain region of functional Fas, Fas-GPI was able to significantly increase Fas-mediated cell death triggered by membrane bound or soluble FasL, whereas engagement of Fas-GPI alone did not trigger apoptosis. This potentiating effect, but not transmembrane Fas activation, was selectively inhibited by protein kinase C activation with phorbol esters, demonstrating that Fas-GPI activated a specific synergistic signal transduction pathway. Fas-GPI and transmembrane Fas were localized in distinct membrane compartments, since Fas-GPI, but not transmembrane Fas, was found in the glycolipid-rich membrane microdomains. These results suggest that apoptosis induced by members of this ligand/receptors family may be differentially modulated through other and parallel signalling pathways.