Berberine has built a reputation as the “natural metformin,” a plant compound popular for blood sugar and metabolic support. A new laboratory study has nothing to do with blood sugar at all. It points to dihydroberberine, the form the gut converts berberine into, as a molecule capable of manipulating how colorectal cancer cells die.
The study, published in Phytomedicine in September 2026, identifies a specific protein target, PANX2, and a self-reinforcing cell death loop that dihydroberberine appears to switch on in colorectal cancer cells.
Why dihydroberberine exists in the first place
Oral berberine has a well-documented problem: it barely absorbs. Human studies have repeatedly shown that less than 1% of an oral dose reaches systemic circulation intact, which is part of why supplement companies have marketed dihydroberberine as a workaround.
Gut bacteria reduce berberine into dihydroberberine, a form that crosses the intestinal wall far more efficiently. Once inside the body, it gets re-oxidized back into berberine, which is the version that actually does the metabolic work researchers have studied for blood sugar and lipid effects.
That conversion step has mostly been framed as a bioavailability curiosity. The new paper suggests dihydroberberine itself, before it even converts back, may have biological activity worth separate attention.
Two cell death pathways that feed each other
To understand what the researchers found, it helps to know what ferroptosis and autophagy are. Ferroptosis is a form of cell death driven by iron and runaway lipid oxidation, first formally named and characterized by Brent Stockwell’s lab at Columbia University in 2012.
Autophagy is the cell’s internal recycling system, normally framed as a survival mechanism that clears damaged components. In cancer biology, the two processes do not stay in their own lanes.
Autophagy can break down ferritin, the protein that stores iron safely inside a cell, in a selective process called ferritinophagy. That releases free iron, which fuels ferroptosis, and the oxidative stress from ferroptosis can in turn trigger more autophagy, creating a loop that amplifies itself once it starts.
Where PANX2 fits in
Pannexin 2, or PANX2, is a channel protein embedded in the cell membrane that helps cells release signaling molecules like ATP to the outside environment. It has turned up in prior cancer research with inconsistent roles, sometimes tumor suppressive, sometimes supportive of cancer cell survival, depending on the tissue type involved.
In this study, the authors identify PANX2 as the point of contact where dihydroberberine exerts its effect on colorectal cancer cells. Rather than acting broadly across the cell, the compound appears to engage this specific channel protein to adjust how the ferroptosis-autophagy loop behaves.
That kind of target specificity matters in cancer pharmacology. A compound that hits one identifiable protein is far easier to study, refine, and eventually test for selectivity than one with diffuse, unexplained effects across a cell.
What the researchers describe finding
According to the study, dihydroberberine binding to PANX2 shifted the balance of the feedback loop in colorectal cancer cells toward sustained ferroptosis, intensified by the accompanying autophagic activity. The combined effect pushed cancer cells further toward death than either process achieved on its own.
This is the kind of mechanistic detail that matters most to researchers designing follow-up work, not to anyone deciding what to take with breakfast. Positive feedback loops are notoriously hard to study because they can be difficult to interrupt once triggered, and difficult to initiate reliably if the trigger is subtle.
Identifying PANX2 as a tractable entry point gives future researchers a specific lever to test, block, or enhance in follow-up experiments.
Why this isn’t a reason to self-treat with berberine
This was laboratory research conducted in colorectal cancer cells, the kind of early-stage mechanistic study that sits many steps before any human application. It says nothing about whether oral berberine or dihydroberberine supplements, taken at commercially available doses, would replicate this effect inside an actual tumor in a living person.
Drug development from a cell culture finding like this one typically takes years, if it advances at all, and usually requires chemically modified or concentrated versions of a compound rather than the over-the-counter form. Using berberine supplements as an anti-cancer strategy based on this study would be an enormous and unsupported leap.
The existing human evidence for berberine remains centered on metabolic markers: several meta-analyses have found modest improvements in fasting glucose and lipid profiles, which is a separate and much better-established body of research from what this new paper describes.
The bigger picture
Colorectal cancer remains one of the most commonly diagnosed cancers worldwide, and resistance to ferroptosis-inducing therapy is an active area of investigation because some tumor cells adapt and evade this death pathway over time. Finding natural compounds that can tip that resistance back in the other direction, through a defined molecular target like PANX2, is valuable groundwork even at this early stage.
What makes this study notable is less the specific compound and more the pattern it reveals: a molecule already sitting in millions of supplement bottles for an entirely different reason turns out to have a second, unrelated biological identity once researchers look closely enough. That gap between what a compound is sold for and what it actually does inside a cell is where a surprising amount of future pharmacology tends to come from.
