Key takeaways
- Oxidative stress is an imbalance, not a toxin: reactive oxygen species outpacing the systems that manage them.
- ROS are produced by normal metabolism; they are not an invader and they are not entirely harmful.
- More antioxidants is not automatically better. Several large trials found harm at high doses.
- The comet assay images DNA fragmentation cell by cell, which is why it is used as a reference method.
The basic mechanism
Every cell that uses oxygen to make energy produces reactive oxygen species (ROS) as a by-product. This happens in the mitochondria, during the electron transport chain, and it is not a malfunction. It is the unavoidable cost of aerobic metabolism.
ROS are chemically reactive because they carry unpaired electrons, which makes them prone to grabbing electrons from whatever is nearby: proteins, membrane lipids, DNA. That grabbing is what "oxidative damage" describes.
Cells are not defenceless. They run enzymatic systems (superoxide dismutase, catalase, glutathione peroxidase) whose job is to neutralise ROS continuously. Oxidative stress is the state where production outpaces neutralisation, sustained over time. It is a balance problem, not a poison.
Why "antioxidant" stopped meaning anything
In the 1990s the free-radical theory of aging produced an obvious-looking conclusion: if ROS cause damage, flood the system with antioxidants. An entire industry followed.
Then the large trials reported. High-dose beta-carotene supplementation increased lung cancer incidence in smokers. High-dose vitamin E showed no benefit and signals of harm in some populations. Meta-analyses of antioxidant supplementation found no mortality benefit overall.
The reason, in hindsight, is that ROS are not purely destructive. They function as signalling molecules. Exercise adaptation depends partly on an ROS signal; immune cells use ROS deliberately. Suppressing them indiscriminately suppresses signalling the body relies on. This is why "more antioxidants" is not a strategy, and why the word on a label tells you nothing.
What actually gets damaged
Three targets matter most, and they fail differently.
- Lipids. Membranes are built from oxidation-prone fats. Oxidised membrane lipids change how the membrane behaves, which changes what enters and leaves the cell.
- Proteins. Oxidised proteins can misfold or lose function, and the clearance systems that remove them are themselves subject to the same environment.
- DNA. The most consequential target, because DNA is the template. Oxidative lesions can produce strand breaks, and while cells have extensive repair machinery, repair is not perfect and its efficiency changes with age.
How it is measured
Indirect markers exist: malondialdehyde for lipid peroxidation, 8-OHdG for oxidative DNA lesions, glutathione ratios for redox state. They are useful, but they are population averages inferred from a sample.
The comet assay, formally single-cell gel electrophoresis, takes a different approach and measures DNA integrity directly, one cell at a time. Cells are embedded in a gel, lysed until only DNA remains, then subjected to an electric field. Intact DNA is too large to migrate; fragmented DNA moves toward the anode and forms a tail. Tail length and intensity quantify damage in that specific cell.
Its advantage is the distribution. Instead of one averaged number, you see how many cells are intact, how many are badly damaged, and how wide the spread is, which is often more informative than the mean.
The standard way to provoke the damage under test conditions is hydrogen peroxide (H₂O₂), applied to cells in vitro so that the resulting fragmentation can be scored against an unchallenged control. It is a blunt, well-characterised oxidative challenge, which is precisely why it is used as a reference insult rather than as a model of everyday physiology.
That is the method behind the published work on this strain. A 2024 paper in Advanced Gut & Microbiome Research (Brouchkov et al.) reported that cell lysates of Bacillus sp. F, the permafrost isolate recovered in 2009, reduced hydrogen-peroxide-induced DNA damage in murine blood cells, ex vivo (doi 10.1155/2024/4559054). Those three qualifiers matter: mouse cells, outside a living animal, under a chemical challenge. It is preclinical mechanism work, not a human outcome, and it is published as such.
The comet-assay panels we publish use the same reference method. The imaging and the full methodology sit on the Science page rather than being summarised here, because a claim about evidence should link to the evidence.
| Criterion | What it reads | What you get |
|---|---|---|
| Malondialdehyde | Lipid peroxidation | One averaged number from a sample |
| 8-OHdG | Oxidative DNA lesions | One averaged number from a sample |
| Glutathione ratio | Redox state | One averaged number from a sample |
| Comet assay | DNA integrity, one cell at a time | A distribution: how many cells are intact, how many damaged, how wide the spread |
What to actually do about it
The unexciting answers dominate. Sleep, because repair processes run during it. Exercise, which raises ROS acutely and improves endogenous antioxidant capacity chronically, a hormetic effect that mega-dose supplementation can blunt. Not smoking, which is the largest single oxidative-load intervention available to most people. Diet built on whole foods rather than isolated high-dose antioxidant compounds.
Supplements sit at the margins of that list, not at the centre. Anything positioned in this space, including the preparation we sell, is a small input relative to sleep and training, and any brand implying otherwise is overselling.
An oxidative stress supplement: what is claimed and what we do not
Our preparation is positioned on exactly one thing: the cell’s resilience under oxidative conditions, imaged by the comet assay described above. The Bacillus F drops are the metabolite fraction of a strain isolated from permafrost in 2009 and documented in patent WO2012060729. The evidence is preclinical, murine cells ex vivo under a hydrogen-peroxide challenge, and the Science page shows the panels rather than describing them.
We do not claim it raises NAD+, induces autophagy or moves an epigenetic clock. If those are the levers you want, the NMN comparison and the spermidine comparison explain which compound sits where. The protocol is 2 ml a day for 30 days; the strain page has the deposit record and the publications for anyone who wants to read the primary material first. That answers what is oxidative stress, and it marks where the preparation sits.