Fig 2: Illustration of how ?-secretase controls abundance and signaling of Fn14 and release of sFn14 Under normal conditions, Fn14 is processed by ?-secretase, which generates sFn14 and degrades Fn14. Fn14 ligand TWEAK binds to remaining Fn14 on the surface, which activates downstream NF?B signaling.Upon inhibition of ?-secretase activity, Fn14 proteolysis is reduced. This ablates sFn14 release and increases Fn14 in the cells and on the cell surface, allowing enhanced Fn14-mediated signaling upon TWEAK ligand binding.

Fig 3: Inhibiting ?-secretase in MDA-MB-231 cells enhances Fn14 mediated NF?B signaling MDA-MB-231 cells were transfected with an siRNA pool against human Fn14 or nontargeting control (Ctrl) siRNA. A day after transfection, cells were treated with ?-secretase inhibitor DAPT (1 µM) or vehicle overnight. Either TWEAK (100 ng/ml) or positive control TNF (10 ng/ml) were applied for indicated time points. The cell lysate was blotted against pI?B and I?B to evaluate NF?B activation or against Fn14 to verify the effect of the DAPT and siFn14 treatment, or against calnexin as a loading control. Shown are representative blots from N = 4 experiments.MDA-MB-231 cells were treated with ?-secretase inhibitor DAPT (1 µM) or vehicle overnight. TWEAK (100 ng/ml) was applied for 10 min. The cell lysate was blotted against pP65 and P65 to evaluate NF?B activation or against ß-actin as a loading control. Shown are representative blots from N = 3 experiments. The dashed vertical line indicates that sample were run on the same blot but not directly next to each other.

Fig 4: Endogenous Fn14 processing in mouse cell lines by ?-secretase AMouse glioblastoma cell line GL261 showed cellular accumulation of Fn14 upon ?-secretase inhibition. Cells were treated overnight with ?-secretase inhibitor DAPT (1 µM) or vehicle. Lysates of biological replicates (Rep.) were blotted for Fn14 with an antibody that targets the C-terminal end of the protein, or against calnexin as loading control.BQuantification of blot in panel (A). Intensity values of Fn14 were normalized to the respective Calnexin loading control. The average of the control condition, where the cells were only treated with vehicle (DMSO), was consecutively normalized to 1.CConditioned media of the GL261 cells from panel (A) were collected, and sFn14 levels were measured by ELISA.DMouse breast cancer cell line 4T1 showed cellular accumulation of Fn14 upon ?-secretase inhibition. Cells were treated overnight with ?-secretase inhibitor DAPT (1 µM) or vehicle. Lysates were blotted for Fn14 with an antibody that targets the C-terminal end of the protein, or against calnexin as loading control.EQuantification of blot in panel (D). Intensity values of Fn14 were normalized to the respective calnexin loading control. The average of the control condition, where the cells were only treated with vehicle (DMSO), was consecutively normalized to 1.FConditioned media of the 4T1 cells from panel (D) were collected and sFn14 levels measured by ELISA.GMouse ovarian cancer cell line ID8 showed cellular accumulation of Fn14 upon ?-secretase inhibition. Cells were treated overnight with ?-secretase inhibitor DAPT (1 µM) or vehicle. Lysates were blotted for Fn14 with an antibody that targets the C-terminal end of the protein, or against calnexin as loading control.HQuantification of blot in panel (G). Intensity values of Fn14 were normalized to the respective calnexin loading control. The average of the control condition, where the cells were only treated with vehicle (DMSO), was consecutively normalized to 1.IConditioned media of the ID8 cells from panel (G) were collected and sFn14 levels measured by ELISA.J, KConditioned media of (J) MDA-MB-231 or (K) SKOV-3 cells were collected at indicated time points after DAPT (1 µM) or vehicle treatment and endogenous sFn14 levels were measured by ELISA. Even after 72 h DAPT still completely blocked ?-secretase, as evidenced by the lack of sFn14 secretion. Data information: All quantification data are shown as mean ± SEM. The P-values that are above 0.05 have not been included into the panels. For all the panels, three biological replicates are performed. For panels (B), (C), (E), (F), (H), and (I), two-tailed unpaired t-tests were used. For panel (J) and (K), no statistical analysis was performed.

Fig 5: Fn14 is processed by the protease ?-secretase Domain structure of Fn14. Fn14 has a short, compact N-terminal extracellular domain. It consists of 53 amino acids that contain three disulfide bridges (dashed lines) that form the cysteine-rich domain, CRD (red). Fn14 has a single transmembrane domain (blue) and a short C-terminal cytosolic tail with a TRAF binding site (green) important for its signaling function. Protein N- and C-termini (N, C) are indicated.Detection of sFn14 in conditioned media. HEK293E cells were transfected with either empty vector or a plasmid encoding human Fn14 that bears an N-terminal HA-tag and a C-terminal double FLAG-tag. Conditioned media and lysates of the transfected cells were collected and analyzed by immunoblotting with the indicated antibodies. Calnexin served as a loading control. Shown are representative blots from N = 3 experiments.Generation of sFn14 is sensitive to the ?-secretase inhibitor DAPT. HEK293E cells were transfected with either Fn14 or C99 (C-terminal fragment of APP), both containing a N-terminal HA-tag and C-terminal double FLAG-tag. One day after transfection, cells were treated with ?-secretase inhibitor DAPT (1 µM), broad-spectrum metalloprotease inhibitor TAPI-1 (50 µM), or the corresponding amount of vehicle DMSO as indicated. The conditioned media and the lysates were blotted with anti-HA antibody. Shown are representative blots from N = 4 experiments.Schematic representation of Fn14 shedding by the ?-secretase complex. ?-Secretase is a hetero-tetrameric complex consisting of the indicated subunits, with presenilin being the catalytic subunit and the ectodomain of nicastrin forming a lid-like structure on top of the ?-secretase complex. Proteolysis takes place at the catalytic core (indicated by black arrow head), within the lipid bilayer. Proteolysis results in one fragment released into the extracellular space (sFn14), and one fragment into the cytosol (Fn14 ICD, intracellular domain). In vitro ?-secretase cleavage assay. HEK293E cells were transfected with epitope-tagged Fn14 or C99. Cellular membranes were collected and incubated under indicated conditions for the ?-secretase activity assay. Reactions were terminated and ultracentrifuged. Supernatant (containing ?-secretase cleavage products) was used for detecting the ICD fragment while the pellet was used to blot for full-length proteins Fn14 or C99. Blotting is done by anti-FLAG antibody. Shown are representative blots from N = 3 experiments.sFn14 production requires the proteolytic presenilin subunit of ?-secretase. HEK293 cells stably transfected with APP carrying the Swedish double mutation (K595N/M596L) and with a CRISPR/CAS9-mediated knockout of presenilin 1 (PS1 KO) or of both PS1 and PS2 (PS1/2 dKO), were transiently transfected with epitope-tagged Fn14. Conditioned media and lysates of the transfected cells were blotted with the indicated antibodies. C83 is a small C-terminal fragment of APP (containing the C-terminal 83 amino acids), which is also subjected to ?-secretase processing. Upon longer exposure, the C99 (generated by BACE1 from APP) is also visible in the PS1/2 dKO condition. Shown are representative blots from N = 4 experiments.Quantification of results from panel (F). sFn14 levels in the conditioned media quantified and normalized to WT sFn14 signal.Data Information: Panel (G) shows mean ± SEM, along with P-values calculated by ordinary one-way ANOVA test with Tukey's multiple comparison test. Number of biological replicates performed is indicated in the corresponding panel legend.