Why BAX doesn't always finish what it starts at the membrane

Every cell carries the instructions for its own death. The protein that executes them is called BAX. When BAX activates, it punches holes in the outer membrane of the mitochondria (the cell's power generators), releasing signals that lock in the cell's fate. A new structural study on bioRxiv has captured, for the first time, exactly how BAX gets held in a "ready but not yet" state at that membrane, paused by a single checkpoint protein called VDAC2.

What BAX does and why it matters

BAX belongs to a family of proteins called BCL-2 proteins, whose job is to regulate whether a cell lives or dies. Some BCL-2 proteins promote survival; BAX promotes death. Under normal conditions, BAX floats mostly inactive in the cell's interior fluid. Under stress, it changes shape, migrates to the mitochondria, and assembles into pores in the outer membrane.

That event, mitochondrial outer membrane permeabilization (MOMP), is the point of no return. Once the membrane is breached, the cell releases pro-death signals into the cytoplasm, and the death program runs to completion.

BAX does not jump from inactive to pore in a single step. It moves through a series of shape changes. What controls those intermediate steps has been unclear, which is what this study set out to answer.

VDAC2 catches BAX mid-activation

Scientists have known that cells lacking VDAC2 are partially resistant to BAX-driven apoptosis (programmed cell death). But without a direct structural picture of the two proteins together, the mechanism stayed unresolved.

This study built that picture. The researchers reconstituted a stable VDAC2-BAX complex in the lab, meaning they produced both proteins, combined them under controlled conditions, and confirmed they physically bind. They then characterized the complex using biochemical and biophysical methods, plus structure predictions from AlphaFold3 (the AI protein-structure modeling tool from Google DeepMind), constrained by experimental data.

VDAC2 captures BAX at a specific point in its activation sequence. At this stage, BAX has already inserted its anchor helix (a segment called the α9 helix) into the mitochondrial outer membrane. Its BH3 domain, a short loop that other death-promoting proteins recognize and engage, is exposed. A region near the N-terminal end of BAX (one end of the protein chain) has also become more accessible. BAX at this stage looks activated, but it has not yet built a pore.

A separate finding: VDAC1, a protein closely related to VDAC2, did not form this complex. Same protein family, structurally similar, different function. This kind of isoform specificity (two nearly identical proteins doing different things) matters when designing experiments that target one but not the other.

What the structural model reveals

By combining AlphaFold3 predictions with experimental constraints, the team built a model of the complex. BAX is anchored to the membrane via its α9 helix, while the rest of its water-soluble domain drapes partially over the VDAC2 pore. VDAC2 is a channel protein that normally allows small molecules to cross the outer mitochondrial membrane. In this model, it also holds BAX in place at a specific intermediate.

The authors describe VDAC2 as a checkpoint: it holds BAX in what they call an "activation-competent intermediate." That phrase means BAX is primed and waiting. Whether it proceeds to form a pore depends on additional signals arriving. VDAC2 gates the pathway at this intermediate; the kill switch is downstream.

Earlier models suggested VDAC2 might be a structural component of the death pore. This study argues it is not. VDAC2 controls access to the final activation step from outside the pore structure.

Bench notes for researchers in this space

If you are running assays with BH3-domain peptides or other compounds that engage BCL-2-family proteins, this study gives you a sharper picture of where in the activation sequence your compounds are likely acting. A few practical points:

• Store BCL-2-family peptides cold and reconstitute fresh. Synthetic BH3 peptides can form unwanted secondary structures at room temperature, reducing their ability to engage the target groove. Freeze-thaw cycles erode activity over time.
• Diluent endotoxin levels affect every apoptosis assay. Endotoxin contamination in bacteriostatic water or other diluents can independently trigger apoptotic signaling in cell-based experiments, adding background signal that looks like compound activity. Use a sterile diluent with published endotoxin specifications.
• Purity grade matters for short peptides. BH3 peptides are short sequences with strong folding tendencies. Lower-purity lots carry truncated or oxidized variants that may fail to engage the BH3-recognition groove at expected concentrations.
• Use glass vessels for low-concentration work. Peptides at nanomolar concentrations adsorb to plastic surfaces, shifting your effective working concentration without any visible sign. Glass cartridges give your stock solution a more stable baseline.

The VDAC2-BAX finding does what good structural biology does: it converts a long-standing genetic observation into a physical picture with a mechanism attached. For anyone designing experiments around this pathway, that picture is now sharper.

Frequently asked questions

What does VDAC2 do to BAX during apoptosis?

VDAC2 binds BAX after it inserts its anchor helix into the mitochondrial outer membrane but before it forms a death pore, holding it in a primed intermediate state. It acts as a checkpoint, not part of the pore itself.

Why does VDAC1 not form the same complex with BAX as VDAC2?

The two proteins are closely related but structurally distinct. Only VDAC2 can capture and stabilize the primed BAX conformation. This isoform specificity means functional behavior does not always follow from structural similarity.

What is the BH3 domain and why does its exposure matter?

The BH3 domain is a short loop on BAX that becomes exposed during activation. In the VDAC2-BAX complex, it is accessible, a hallmark of the primed intermediate. BH3-mimetic peptides are designed to engage exactly this exposed region.

Prompted by this coverage at bioRxiv →

Related from our lab: the pen · cartridges · bacteriostatic water · reconstitution calculators

Shared by PreppinPeppers for research, educational, and demonstration awareness only. We link to third-party coverage; we do not endorse it, and nothing here is medical advice or a recommendation to use any substance in humans or animals. Our products are sold for laboratory research use only.

What the research community gets wrong about BAX

BAX shows up in a lot of assay designs, and a few old assumptions can quietly skew what you see at the bench. Here are the ones worth checking.

• Treating BAX as a simple on/off switch. BAX does not jump straight from inactive to a finished pore. It passes through several shape changes, and it can sit in a primed intermediate at the membrane without ever forming a pore. If your readout only scores the final pore, you may miss where a compound is actually acting.
• Assuming VDAC1 and VDAC2 are interchangeable. They are close relatives and look structurally similar, but they do not behave the same here. In the reconstituted work, VDAC2 captured BAX and VDAC1 did not. When a reagent or knockout targets one isoform, confirm which one you are actually hitting.
• Thinking VDAC2 is part of the death pore. The newer structural picture places VDAC2 outside the pore, holding BAX at an intermediate and gating access to the final step. That makes it a checkpoint, not a building block of the pore itself.
• Handling BH3 peptides as if they were stable stock. These are short sequences that fold on their own, oxidize, and lose activity across freeze-thaw cycles. At low concentrations they also stick to plastic, so your effective concentration in the vial can drift below what the label says.
• Ignoring the diluent. Endotoxin in bacteriostatic water or other diluents can trigger apoptotic signaling on its own. That background can look like compound activity, so a diluent with published endotoxin specs keeps the comparison clean.

From our bench: If you reconstitute a BH3-domain peptide and split the same stock between a plastic tube and a glass vessel, tell us what you see. Measure the working concentration from each after a set hold time at your usual bench temperature, and note any difference in recovery. Send your paired numbers, the peptide length and purity grade, and your hold time, and we will add real observations (not estimates) to this page.

Sources

1. Bacteriostatic Water for Injection, USP , FDA/DailyMed label (0.9% benzyl alcohol)
2. Duerkop et al., Biotechnol J 2018 , Impact of Cavitation, High Shear Stress and Air/Liquid Interfaces on Protein Aggregation
3. Sigma-Aldrich (Merck) , Handling and Storage Guidelines for Peptides and Proteins
4. UniProt Q07812 , Apoptosis regulator BAX (human), function and mitochondrial permeabilization
5. UniProt P45880 , Voltage-dependent anion-selective channel protein 2 (VDAC2), human
6. NCBI Gene 581 , BAX, BCL2 associated X apoptosis regulator