Annexin A5 controls VDAC1-dependent mitochondrial Ca²⁺ homeostasis and determines cellular susceptibility to apoptosis

The study addresses how Annexin A5 (AnxA5), a Ca²⁺-dependent phospholipid-binding protein implicated in intracellular Ca²⁺ signaling and apoptosis, regulates mitochondrial Ca²⁺ homeostasis. Prior reports showed AnxA5 binds cardiolipin in mitochondria and influences apoptosis, while VDAC1 in the outer mitochondrial membrane conducts Ca²⁺ released from the ER. The authors hypothesize that AnxA5 modulates mitochondrial Ca²⁺ signaling at the outer mitochondrial membrane by influencing VDAC1-dependent Ca²⁺ flux into the mitochondrial intermembrane space, thereby affecting downstream mitochondrial dynamics and cell susceptibility to apoptosis. The purpose is to define AnxA5’s localization, mechanism of action on trans-OMM Ca²⁺ transport, and its impact on mitochondrial structure/function and apoptotic sensitivity.

Background literature highlights: (1) AnxA5 binds negatively charged phospholipids in a Ca²⁺-dependent manner, can insert into membranes, and has been linked to Ca²⁺ influx in vesicles and cells under oxidative stress. (2) In mitochondria, AnxA5 binds cardiolipin and can immobilize cardiolipin microdomains, with roles in apoptosis; AnxA5 depletion can alter sensitivity to apoptosis and affect VDAC1 expression. (3) VDAC1 is a Ca²⁺-permeable channel in the OMM enabling ER-to-mitochondria Ca²⁺ transfer; its permeability is modulated by proteins like α-synuclein and Bcl-xL, and its oligomerization contributes to apoptosis. (4) Mitochondrial Ca²⁺ uptake across the IMM requires IMS Ca²⁺ thresholds via MCU complex, counterbalanced by NCLX. These reports frame the question of whether AnxA5 regulates OMM Ca²⁺ passage, potentially via VDAC1, especially during IP₃-mediated ER Ca²⁺ release and in apoptotic contexts.

• Cell models: CRISPR/Cas9-generated AnxA5 knockout (AnxA5-KO) in HeLa and EA.hy926 cells; perivascular cells isolated from WT and AnxA5-KO mice. Rescue by transient expression of WT or mutant AnxA5 constructs.
• Ca²⁺ imaging: Genetically encoded, targeted sensors: mitochondrial matrix (4mtD3cpv), intermembrane space (MICU1-140-GEMGECO1/IMS-GEM-GECO1), cristae lumen (ROMO-GEM-GECO1), MICU1-CFP/YFP FRET sensor for MICU1 rearrangements; cytosolic Ca²⁺ via Fura-2 AM or jGCaMP7c; ER Ca²⁺ via DIER. Stimuli: ATP or histamine (IP₃ agonists), BHQ (SERCA inhibitor), SOCE by Ca²⁺ readdition; NCLX inhibitor CGP37157.
• Mitochondrial membrane potential: TMRM with FCCP depolarization control.
• ER-mitochondria contacts and mitochondrial morphology: Confocal 3D imaging with Mitotracker and ER markers; SPLICS split-GFP MERC sensor; quantitative morphology (volume, surface, branching, elongation) analyses; structured illumination microscopy (SIM) for cristae dynamics in MERCs and whole mitochondria; MCU redistribution relative to MICU1 (IBM association index).
• Ultrastructure: Transmission electron microscopy (TEM) for cristae density and spatial distribution (PCM) analyses.
• Subcellular localization of AnxA5: Subcellular fractionation to cytosol, crude and pure mitochondria; proteinase K protection assays; immunogold TEM under basal and post-ER Ca²⁺ release (cryo-fixation) to map AnxA5 relative to OMM.
• Protein interactions/proximity: Co-immunoprecipitation for AnxA5–VDAC1; in situ proximity ligation assay (PLA) for sub-30 nm proximity between AnxA5 and VDAC1; controls with VDAC1 knockdown and VDAC1–IP₃R interaction.
• Genetic perturbations: siRNA knockdown of VDAC1; overexpression of VDAC2-FLAG or VDAC3-FLAG to test rescue; AnxA5 mutants disrupting Ca²⁺ binding (AnxA5-2Mt: D144N,E228Q; AnxA5-3Mt: D144N,E228Q,D303N) or self-assembly (AnxA5-5Mt: R18E,R25E,K29E,K58E,K193E).
• Electrophysiology: Patch clamp in mitochondria-attached configuration on intact isolated mitochondria to record OMM Ca²⁺-permeable channel activity; varied voltages, Ca²⁺ dependence, measurement of conductance and open probability (NPo); recombinant AnxA5 in pipette; effects of VDAC1 knockdown.
• Apoptosis and VDAC1 oligomerization assays: Treatments with cisplatin or selenite; co-treatment with VBIT-4 (VDAC1 oligomerization inhibitor). Outcomes: mitochondrial/cytosolic/IMS Ca²⁺, cell viability/apoptosis (Annexin V-FITC/PI by flow cytometry and live-cell imaging), VDAC1 dimerization by chemical cross-linking and immunoblot; VDAC1 clustering by VDAC1-TC labeling and confocal quantification.
• Statistics: Appropriate parametric/non-parametric tests (Student’s t-test, ANOVA with Tukey, Kruskal–Wallis, Kolmogorov–Smirnov), multiple biological replicates and single-cell n values reported.

• AnxA5 is essential for mitochondrial Ca²⁺ uptake upon IP₃-mediated ER Ca²⁺ release: Mitochondrial matrix [Ca²⁺] responses to ATP/histamine were markedly reduced in AnxA5-KO HeLa, EA.hy926, and mouse perivascular cells, with normal basal levels; rescue by re-expression of AnxA5 restored signals.
• Cytosolic and ER Ca²⁺ handling are largely intact: Cytosolic Ca²⁺ transients and ER Ca²⁺ depletion/refilling (histamine, BHQ) were comparable between WT and AnxA5-KO (with slightly enhanced cytosolic Ca²⁺ in EA.hy926 KO), and mitochondrial Ca²⁺ extrusion via NCLX was not responsible (CGP37157 had no effect on maximal mitochondrial Ca²⁺ rise in KO).
• MERCs are preserved but mitochondrial morphology is altered: MERC abundance (Pearson’s R; SPLICS) unchanged in AnxA5-KO. KO cells showed increased mitochondrial volume and branching, with unchanged overall cristae amount/density but a central increase in cristae membrane density (PCM) distribution.
• AnxA5 localizes at and within mitochondria: Fractionation and proteinase K assays indicate a substantial mitochondrial pool with a majority accessible on the cytosolic leaflet of the OMM; immunogold TEM detected AnxA5 in cytosol, OMM, and inside mitochondria. Upon ER Ca²⁺ release, AnxA5 accumulates within ~20 nm of the OMM on both sides.
• AnxA5 specifically facilitates IMS Ca²⁺ signaling from ER sources: IMS [Ca²⁺] increases to histamine were strongly reduced in AnxA5-KO; BHQ-induced leak and SOCE-induced IMS Ca²⁺ rises were similar to WT, indicating specificity for IP₃-driven ER release. Dose–response showed an EC50 shift for IMS Ca²⁺ from 1.3 µM (WT) to 4.8 µM (KO), while cytosolic Ca²⁺ EC50 remained ~2 µM in both.
• AnxA5’s Ca²⁺ binding is required: KO rescue with WT AnxA5 or the self-assembly mutant (AnxA5-5Mt) restored IMS Ca²⁺ signaling; Ca²⁺-binding mutants (AnxA5-2Mt, -3Mt) failed to rescue.
• AnxA5 shapes IMM dynamics via IMS Ca²⁺: MICU1-FRET decrease upon ER Ca²⁺ release was reduced in KO, indicating less MICU1 de-oligomerization. Cristae membrane dynamics in MERCs slowed with ER Ca²⁺ release in WT but not KO; cristae [Ca²⁺] rises were reduced in KO. Histamine-induced MCU translocation toward the inner boundary membrane (IBM) occurred in WT but was impaired in KO; associated mitochondrial area/aspect ratio remodeling was blunted in KO.
• AnxA5 is proximal to VDAC1 and required for VDAC1/VDAC2 Ca²⁺ permeability: PLA showed sub-30 nm proximity between AnxA5 and VDAC1 (lost with AnxA5 KO and reduced with VDAC1 KD); co-IP did not detect a stable complex. VDAC1 KD reduced mitochondrial Ca²⁺ signals in WT but not further in KO, indicating interdependence. Overexpressed VDAC2 (but not VDAC3) rescued VDAC1 KD in WT; this rescue required AnxA5, implicating AnxA5 in enabling Ca²⁺-permeable states of VDAC1/VDAC2.
• Electrophysiology: Identified a Ca²⁺-dependent 35 pS OMM channel with voltage-dependent NPo (−60 to −120 mV). Channel occurrence and NPo were reduced in AnxA5-KO mitochondria; addition of recombinant AnxA5 restored both. VDAC1 KD reduced channel occurrence/NPo; AnxA5 KO did not further reduce NPo in VDAC1 KD cells. No channel activity without Ca²⁺ in the pipette.
• Apoptosis context: After 12 h cisplatin (5–10 µM), mitochondrial matrix and IMS Ca²⁺ elevations were reduced in KO vs WT, while cytosolic Ca²⁺ was similar. Despite lower mitochondrial Ca²⁺ signals, AnxA5-KO cells had higher apoptosis and cell death at 24–48 h. VBIT-4 co-treatment reduced apoptosis in both WT and KO.
• VDAC1 oligomerization: Cisplatin increased VDAC1 dimerization more in KO (e.g., at 24 h: ~3.2-fold WT vs ~5.2-fold KO; at 48 h: ~8.4-fold WT vs ~9.9-fold KO relative to DMSO). VBIT-4 reduced dimerization. Selenite similarly induced greater dimerization and apoptosis in KO, both reduced by VBIT-4. VDAC1 clustering (VDAC1-TC imaging) increased with cisplatin, more so in KO; VBIT-4 reduced cluster size.

The data demonstrate that AnxA5 is a key regulator of mitochondrial intermembrane space Ca²⁺ signaling during ER IP₃-driven Ca²⁺ release by promoting a Ca²⁺-permeable state of VDAC1 (and VDAC2). AnxA5 rapidly accumulates at the OMM in high-Ca²⁺ microdomains (MERCs), where its Ca²⁺-binding capacity enables modulation of the lipid microenvironment surrounding VDAC1, thereby enhancing OMM Ca²⁺ flux into the IMS. This selective facilitation of ER-to-mitochondria Ca²⁺ transfer tunes MICU1 gating, cristae junction dynamics, MCU redistribution, and mitochondrial morphology during stimulation, linking IMS Ca²⁺ to structural remodeling. In apoptosis-inducing conditions (cisplatin/selenite), AnxA5’s proximity to VDAC1 constrains VDAC1 oligomerization, limiting apoptotic signaling and protecting cell viability; pharmacological inhibition of VDAC1 oligomerization (VBIT-4) phenocopies this protection. Together, findings position AnxA5 as an essential component of the VDAC1 microenvironment that integrates Ca²⁺ signaling and mitochondrial structural-functional responses, with implications for cell survival under stress.

This study identifies Annexin A5 as a Ca²⁺-binding regulator of VDAC1/VDAC2-mediated Ca²⁺ permeability at the outer mitochondrial membrane, essential for efficient IMS and matrix Ca²⁺ signaling upon ER IP₃-induced Ca²⁺ release. By enabling IMS Ca²⁺ microdomains, AnxA5 modulates MICU1, cristae dynamics, MCU localization, and mitochondrial remodeling. AnxA5 further guards against apoptosis by limiting cisplatin/selenite-induced VDAC1 oligomerization, with VBIT-4 confirming the mechanistic link. Future work should define the specific OMM phospholipids engaging AnxA5 in vivo, delineate the structural basis for AnxA5’s modulation of VDAC states, test tissue-specific physiological relevance in animal models, and explore therapeutic targeting of the AnxA5–VDAC1 axis in diseases involving mitochondrial Ca²⁺ dysregulation and apoptosis.

• The 35 pS OMM channel is linked to VDAC1 functionally (via KD) but not unequivocally identified as VDAC1 due to modest KD efficiency and lack of direct biophysical attribution; high-conductance VDAC states seen in artificial bilayers were not observed in intact mitochondria, highlighting model-dependent differences.
• Co-immunoprecipitation failed to show a stable AnxA5–VDAC1 complex, suggesting transient or indirect interactions; precise molecular mechanism (e.g., lipid mediation) remains unresolved.
• The exact OMM phospholipid species and microdomains that recruit AnxA5 are unknown.
• Most experiments were performed in cultured cell lines; in vivo validation of functional outcomes is limited.
• While SOCE effects on IMS Ca²⁺ appeared unaffected, broader channel/pathway specificity beyond ER-driven signals was not extensively characterized.