Disruption of auto-inhibition underlies conformational signaling of ASIC1a to induce neuronal necroptosis

Figures

• 

  Fig. 1. Peptide NT1–20 protects mouse cortical neurons against acidosis-induced necroptosis. a Schematics of cell death assays for (b, c). CTB, cell titer blue; LDH, lactate dehydrogenase. b CTB assay for viability of neurons treated with indicated peptides (all at 10 μM). CP-1, CP-1-1, CP-1-2 and CP-1-3 reduced viability. Nec-1 (20 μM) attenuated CP-1-2 induced death. Compared to CP-1-2, CP-1-3 induced much less and the membrane impermeable CP-1-2S induced no death. n = 4–6, *** p < 0.001 vs. control peptide (Ctrl, TAT alone), ### p < 0.001 vs. CP-1-2, by ANOVA. c LDH release assay for viability of neurons treated with CP-1-2 (10 µM) without or with NT peptides (10 µM). The toxicity by CP-1-2 was significantly reduced by NT1–20, but not NT11–30 and NT21–41. n = 4–5, *** p < 0.001 vs. Ctrl, ## p < 0.01, ### p < 0.001 vs. CP-1-2, by ANOVA. d Schematics of cell death assay for (e–h). PI, propidium iodide (PI); SS, standard extracellular solution. e LDH release from neurons exposed to pH 7.4 and pH 6.0 SS. pH 6.0-induced death was reduced by treatment of 10 µM NT1–20, but not NT1–10 and NT11–20. n = 5–12, *** p < 0.001 vs. corresponding pH 7.4, ### p < 0.001 vs. Ctrl in pH 6.0, by ANOVA. f…

• 

  Fig. 2. Distal N-terminal region of ASIC1a is crucial for inhibiting CP-1 death motif. a De novo Rosetta modeling of full-length ASIC1a. Gray color represents closed state base on PDB structure 5wku, while blue color depicts open state based on PDB structure 4ntw. b Acidosis-induced dissociation of ASIC1a-CT from its NT, as measured by FRET: YFP/CFP emission ratio (F525/F482) with excitation at 405 nm. CHO cells expressing CFP-ASIC1a-YFP (WT) or CFP-ASIC1a-E235C/Y389C-YFP (E235C/Y389C) were untreated or treated with PcTX1 (100 nM). Bath solution pH changed from 7.4 to 6.0 as indicated. Data points are averages of n = 10-21 cells. Dashed lines are fits by exponential function. c Summary of % ratio changes at 5 min of pH 6.0 treatment for data in (b). ** p < 0.01, *** p < 0.001 vs. WT, by ANOVA. d Spectra FRET for energy transfer between CFP and YFP of WT, CFP-ASIC1a-HIF-YFP (HIF), and CFP-YFP at pH 7.4 and 6.0. n = 6–12, ** p < 0.01, *** p < 0.001 vs. corresponding pH 7.4, by ANOVA. e, f Representative images (e) and summary (f) of PI staining of CHO cells transfected with WT ASIC1a and its HIF and NT truncation mutants at pH 7.4. Scale bar, 20 µm. Nec-1 (20 µM), peptide NT1–20 (10 µM),…

• 

  Fig. 3. NSF is recruited to ASIC1a in ischemic brain to form NSF-ASIC1a-RIPK1 complex. a Image of Coomassie blue staining of SDS-PAGE of immunoprecipitants by a polyclonal ASIC1a antibody from contralateral (Contra) and ischemic (MCAO) hemispheres of a mouse subject to MCAO. Red box indicates the band of interest (~85 kD) reproducibly identified in three independent experiments. b NSF was identified from the ~85 kD band by MALDI-TOF-MS analysis, in which 18 peptide sequences (out of 39) were found to match NSF (GenBank accession no. 18195) with the protein sequence coverage of 22% (Supplementary Data 1). c Confirmation of increased RIPK1-ASIC1a-NSF association in MCAO hemispheres as compared to Contra hemispheres by co-IP. In two examples, both the anti-ASIC1a and anti-RIPK1 antibodies pulled down more NSF from MCAO than Contra groups, along with the markedly increased serine/threonine phosphorylation of RIPK1. Total levels of NSF were not changed after MCAO. d Summary data for (c). n = 3, * p < 0.05 vs. Contra by ANOVA. e Asic1a gene deletion prevented the ischemia-induced increase in RIPK1-NSF association. Anti-NSF antibody pulled down more RIPK1 from MCAO than Contra brain samples…

• 

  Fig. 4. NSF binds to the distal N-terminus of ASIC1a. a Glutathione-S-transferase (GST) pull-down assay showing the pulldown of NSF by GST-NT, but not GST-CT. Purified GST, GST-ASIC1a C-terminus (GST-CT) or GST-ASIC1a N-terminus (GST-NT) were incubated with mouse brain extracts, immobilized on glutathione-Sepharose 4B beads, and immunoblotted for NSF (top) and GST (bottom). GST alone served as a negative control. b Amino acid sequences of NT deletion mutants of ASIC1a. c Co-IP experiments showing that deletion of aa1-41 (Δ1-41) or aa1-20 (Δ1-20) decreased the acid-induced association between ASIC1a and NSF in CHO cells. d Summary data of (c). n = 3, ** p < 0.01 vs. corresponding pH 7.4, ### p < 0.001 vs. WT, by ANOVA. e Co-IP experiments showing that NT1–20 (10 μM, 0.5 h pretreatment) reduced acid-induced formation of NSF-ASIC1a-RIPK1 complex in cultured mouse cortical neurons. NT1–20 also decreased acid-induced degradation of RIPK1 protein (lower panel). f Summary data of (e). n = 3, * p < 0.05, ** p < 0.01 vs. corresponding pH 7.4, # p < 0.05 vs. WT, by ANOVA. Error bars are SEM for all summary panels.

• 

  Fig. 5. NSF knockdown protects against acidotoxicity in cultured neurons. a, b Knockdown efficiency of NSF shRNA in cultured neurons determined by western blotting. Shown are representative blots (a) and summary data (b). n = 3, *** p < 0.001 vs. Scrm (scrambled shRNA) by paired t test. c shRNA knockdown of NSF attenuated acid-induced associations of ASIC1a with NSF and ASIC1a with RIPK1. d Summary data of NSF pulled down by the ASIC1a antibody. The data are normalized to NSF/ASIC1a of neurons treated with Scrm at pH 7.4. n = 3, *** p < 0.05 vs. corresponding pH 7.4, ## p < 0.01 vs. Scrm in pH 6.0, by ANOVA. e NSF knockdown attenuated acid-induced decrease of CFP-ASIC1a-YFP FRET. Similar to Fig. 2b, but the cells were transfected with Scrm or NSF shRNA as indicated. n = 11 cells for each. f Summary of ratio changes at 5 min of the pH 6.0 treatment for data in (e). *** p < 0.001 vs. CFP-ASIC1a-YFP with Scrm by unpaired t test. g PI staining of cultured cortical neurons prepared from WT and ASIC1a KO mice transfected with NSF-shRNA or Scrm shRNA. Scale bar, 50 µm. The neurons were treated at either pH 7.4 or pH 6.0 for 1 h and then returned to the normal culture medium for 24 h before PI…

• 

  Fig. 6. E6EEE9 at distal N-terminus of ASIC1a are critical for intracellular auto-inhibition. a Rosetta model of closed state ASIC1a suggests that negatively charged glutamate residues (E6EEE9) are juxtaposed to positively charged lysines (K468, K471, and K474) in the CT death motif, forming electrostatic interactions (upper). The distance between beta carbons of E7 and K468 is ~4.7 Å. With E → A substitutions at E6EEE9 (lower), ASIC1a-NT moves apart from the lysines at CT death motif. (b) The E/A mutant exhibited lower FRET than WT ASIC1a at pH 7.4. CFP and YFP were tagged at NT and CT, respectively. The YFP/CFP emission ratios were acquired under the same conditions using identical settings. n = 12–14, * p < 0.05 vs. WT by unpaired t test. c, d Expression of E/A mutant in CHO cells resulted in more cell death than WT ASIC1a at pH 7.4. Representative images of PI-stained cells (c) and quantification of PI-positive CHO cells (d). n > 200 transfected neurons (green) counted for each. Scale bar, 20 µm. * p < 0.05 vs. WT by unpaired t test. e Co-IP experiments of transfected CHO cells showing that E/A mutant exhibited increased association with NSF and RIPK1 at pH 7.4. f, g Summary data for…

• 

  Fig. 7. Schematic diagram of proposed mechanism underlying conformational signaling of ASIC1a in acidotoxicity. Under normal conditions, the CP-1 death motif of ASIC1a-CT is masked by interaction with ASIC1a-NT (left). Acidosis activates ASIC1a and causes conformational transition that disrupts the preformed N- to C-terminal interaction within ASIC1a to unmask the CT death motif. This transition is facilitated and stabilized by NSF binding to ASIC1a-NT. The freed ASIC1a-CT recruits RIPK1 and activates it by phosphorylation, resulting in neuronal necroptosis (middle). Peptide NT1–20 binds to ASIC1a-CT to prevent recruitment and thereby activation of RIPK1 to achieve neuroprotection (right). NT1–20 may also bind to NSF, decreasing its availability for ASIC1a-NT, similar to NSF shRNA knockdown or inhibition by NEM.