ATX-304

What it is

ATX-304 is a small molecule developed by Betagenon AB (Sweden) as a potential therapeutic for metabolic disease. It’s not a peptide, not a hormone, not a SARM, not a GLP-1. It is a synthetic oral compound that works on one of the most fundamental energy sensors in human cells: AMPK (AMP-activated protein kinase).

The “ATX” naming convention is its commercial/development name; in the clinical literature it’s referenced as O304, and that’s the search term that finds the primary research.

---

How it works

AMPK is the cell’s fuel gauge. When cellular energy drops — because you’re exercising, fasting, or in caloric deficit — AMP:ATP ratios rise, and AMPK gets switched on. Once active, AMPK shifts the cell into energy-extraction mode: it promotes fat oxidation, increases glucose uptake, triggers mitochondrial biogenesis (via PGC-1α), and suppresses energy-storage pathways. This is a core part of why exercise improves metabolic health — AMPK is one of the primary mediators. [ESTABLISHED — Hardie 2011 *Nat Rev Mol Cell Biol*, PMID 21360254]

How most things activate AMPK: by promoting the phosphorylation of AMPK at its activation site (Thr172), either directly or indirectly. Metformin takes an indirect route via Complex I inhibition.

ATX-304’s distinct approach: it works downstream — blocking the phosphatase that removes the activating phosphate tag. Instead of pushing the ON-switch, it prevents the OFF-switch from firing. AMPK stays in its naturally-phosphorylated, active state for longer.

The practical question — whether blocking dephosphorylation produces better, worse, or equivalent outcomes vs. promoting phosphorylation directly — is one the research hasn’t fully answered. The mechanism distinction is real; the clinical significance remains to be established.

ATX-304’s second mechanism — mild mitochondrial uncoupling. Alongside AMPK activation, ATX-304 has been shown to produce mild mitochondrial uncoupling as a secondary feature. The basics: mitochondria make ATP by pumping protons across the inner membrane, building pressure that’s released through the ATP synthase channel to produce usable energy. Fuel burning is “coupled” to ATP production. Uncoupling creates a small proton leak so some pressure escapes without going through the ATP channel — the energy leaves as heat instead. The mitochondria then burn more fuel just to maintain steady energy output. One in-vitro kidney-cell study measured basal oxygen consumption rising approximately 38% under ATX-304.

For ATX-304 this uncoupling is mild and secondary — layered on top of the AMPK job, not the primary mechanism. This is a meaningful distinction: the dedicated uncoupler in this space is BAM15 (see below). ATX-304 is primarily an AMPK activator that also uncouples mildly; BAM15 does nothing but uncouple.

Comparison to MOTS-c (the most important cross-reference for OHM’s audience):

Feature	ATX-304	MOTS-c
Type	Small molecule (synthetic)	Peptide (mitochondrial-derived)
Route	Oral	Subcutaneous injection
AMPK activation path	Phosphatase inhibition (downstream)	Folate cycle disruption → AMPK (indirect); nuclear translocation also involved
Endogenous?	No	Yes — your mitochondria make MOTS-c, especially during exercise
Human trial data	TELLUS Phase IIa (T2DM, metabolic endpoints)	Observational human studies; interventional trials still emerging
OHM vendor available?	No	Yes — Alyve MOTS-C (10mg, $48)

Neither compound is established as superior to the other on current evidence — they hit the same target (AMPK) via different upstream mechanisms and likely have different tissue profiles. For OHM’s audience, the practical decision is simpler: MOTS-c is available from verified vendors; ATX-304 is research-stage with limited vendor landscape and an active mislabeling problem in the market.

---

What the research shows

The TELLUS Phase IIa trial:

• Population: Type 2 diabetics on metformin background therapy
• N: ~60 patients
• Duration: 28 days
• Endpoints: blood glucose and metabolic markers

Reported results: reduced fasting blood glucose, improved insulin resistance, lower blood pressure, improved small-vessel blood flow. Well tolerated; gastrointestinal complaints were the main mild side effect; no serious adverse events attributed to ATX-304.

What this trial does and doesn’t tell us:

• ✅ ATX-304 produces the metabolic effects expected from AMPK activation in humans at 28 days
• ✅ The safety profile over 28 days was acceptable
• ❌ No body-composition data (muscle gain, fat loss) in any population
• ❌ No exercise-performance data
• ❌ No data in healthy, non-diabetic individuals
• ❌ No long-term safety data

The honest bottom line: ATX-304 is a legitimately interesting compound with a real human trial, not just animal and in-vitro data. But the trial was built for the therapeutic drug-development use case (T2DM, metabolic endpoints) — extrapolating from it to the health-optimization / exercise-mimetic / longevity use case means reasoning past the available evidence. That gap may close as the compound develops further; it hasn’t closed yet.

Body composition data — animal only. Body-composition work exists only in preclinical models. In obese mice, reported results include reduced fat mass, lower blood cholesterol, improved fatty liver markers, and increased energy expenditure. One notable finding: ATX-304-treated mice consumed more food than controls and still maintained reduced fat mass. The proposed explanation is the mild mitochondrial uncoupling described in the mechanism section — extra intake was dissipated as heat rather than stored. One mouse study; hold it as exactly that.

---

BAM15 — the dedicated mitochondrial uncoupler

BAM15 is the compound ATX-304 is most often compared to, and the distinction matters: where ATX-304 is primarily an AMPK activator that also causes mild uncoupling as a secondary feature, BAM15 is a pure, dedicated mitochondrial uncoupler — that is essentially all it does.

How BAM15 works: BAM15 goes directly to the inner mitochondrial membrane and makes it leaky to protons. It shuttles protons across, bypassing the ATP synthase gate entirely. The result is the same proton-leak → heat → increased fuel oxidation loop described in the mechanism section above, but as the primary intended effect rather than a mild side feature. Fatty acids are oxidized to meet the increased energy demand.

Preclinical results in obese mice:

• Reduced fat mass
• No reduction in food intake
• No measured loss of lean mass
• No rise in core body temperature

The DNP comparison — why “no temperature rise” matters. DNP (2,4-dinitrophenol) is an older uncoupler on the same general principle. It is genuinely dangerous: it uncouples aggressively and system-wide, and can produce runaway internal overheating. People have died. BAM15 was specifically designed to achieve the uncoupling benefit without that runaway heat signature, and in the animal work it achieves that on the measured parameters. The critical qualifier: in the animal work.

The honest caveat — no human trials. BAM15 has no human trials. None. Every claim above comes from mice and cell cultures. Its safety, real dosing, and effects in an actual human are entirely unknown. Vendors and forum posts may describe it with confidence; that confidence has no human data behind it.

---

ATX-304 vs BAM15 — comparison

Feature	ATX-304	BAM15
Primary mechanism	AMPK activation (phosphatase inhibition)	Mitochondrial uncoupling (direct proton leak)
Secondary mechanism	Mild mitochondrial uncoupling	None noted
Route	Oral, small molecule	Research chemical; human route not established
Human trials	Yes — TELLUS Phase IIa (~60 T2DM, 28 days)	None
Body composition	Animal only — reduced fat mass, improved metabolic markers	Animal only — reduced fat mass, no lean loss, no food intake reduction
Human safety	28-day TELLUS profile (GI complaints; no serious AEs)	Entirely unknown
vs DNP	Not primarily an uncoupler	Designed as safer alternative (in animals)
OHM vendor	None	None
Evidence tier	for metabolic; for body comp	only

Signal vs. physics — the key distinction. ATX-304 is primarily a signal: it amplifies the cell’s own AMPK activation by blocking the off-switch, then uncouples mildly as a secondary bonus. BAM15 is primarily a physical effect: it makes the inner membrane leaky directly, bypassing any cellular signaling. ATX-304 changes what the cell decides to do. BAM15 changes the physics of the power plant regardless of what the cell wants.

Intensity tradeoff. ATX-304’s uncoupling is mild and secondary; BAM15’s is the main event and more forceful. BAM15 is potentially more potent at raw energy expenditure. ATX-304 is potentially gentler and more regulated because it mostly amplifies a natural mechanism rather than forcing a structural leak.

Extra effects via AMPK. Because ATX-304 reaches its thermogenic effect through AMPK, it carries the full AMPK signaling package — improved insulin sensitivity, blood sugar control, blood pressure, vascular effects — as documented in the TELLUS trial. BAM15 shows similar metabolic improvements in animals, but they flow from uncoupling rather than from a broad signaling switch.

The fatigue angle. GLP-1 fatigue comes from under-eating. Both ATX-304 and BAM15 work in the opposite direction — they raise energy expenditure rather than cutting intake, and animal data shows food consumption maintained or increased. But aggressive mitochondrial uncoupling wastes energy as heat that muscles and brain would otherwise use; pushed hard enough, that is its own kind of fatigue. ATX-304 has a plausible case for being gentler here: the uncoupling is mild and secondary, and AMPK activation tends to support mitochondrial health over time (energy-positive direction). BAM15 carries a larger unknown — nobody has mapped where the comfortable dose sits in a human, because no human has been given it.

The practical summary. ATX-304 is the metabolic signal amplifier with a mild thermogenic bonus and actual human safety data. BAM15 is the pure thermogenic play — more direct and potentially more potent at raw energy output, but resting entirely on animal data. These are not rivals; they are two tools that overlap on the thermogenic endpoint through genuinely different mechanisms. For any research use, ATX-304 is the more evidence-grounded choice. BAM15 is genuinely interesting mechanistically but needs human trials — which don’t exist yet — before the animal findings translate into anything concrete.

Neither compound is available from OHM’s vetted vendors. Both are research-stage with active supply-chain quality concerns. No OHM affiliate hook applies to either.

---

Safety notes

From the TELLUS trial data: well-tolerated at 28 days; gastrointestinal complaints as the main mild side effect. No serious adverse events.

Beyond the 28-day trial, no long-term safety data. The theoretical additive concern with other AMPK activators (MOTS-c, metformin) and with glucose-lowering agents applies — stacking two AMPK activators in a glucose-lowering context warrants caution about hypoglycemia, though this has not been studied directly.

---

Regulatory status

Not FDA-approved for any use. In drug development (Betagenon AB) for metabolic disease indications. Not on WADA prohibited list as of mid-2026. Sold as a research chemical in various markets; the mislabeling-as-peptide problem noted above is a specific supply-chain red flag for this compound — always verify identity and CAS number against the listed compound before purchasing.

---

Commercial note

None of OHM’s three vendors (Alyve, US Pure Peptides, BioLongevity) currently carry ATX-304. This article is pure education — no OHM affiliate link or coupon applies.

If a reader asks “where to buy”: the research-chemical market for this compound has an active mislabeling problem (vendors listing ATX-304 as a peptide, or listing BAM15 under false categories). The three questions that don’t change regardless of which vendor you’re looking at: (1) What is it actually — is the compound named correctly and specifically? (2) Is it identified correctly — does the CAS number match? (3) What is the purity — is there a third-party COA? A compound sold under the wrong category on the label has already failed question one. OHM does not currently have a vetted vendor to recommend for either ATX-304 or BAM15.

---

Sources

• primary source (Derek/MPMD Substack, complete digest as of 2026-07-11)
• AMPK mechanism anchor: Hardie DG. “AMP-activated protein kinase — an energy sensor that regulates all aspects of cell function.” Nat Rev Mol Cell Biol. 2011;12(8):496-507. PMID 21360254. [ESTABLISHED]
• TELLUS trial:
• ATX-304 uncoupling (kidney-cell ~38% O2 rise):
• ATX-304 mouse body comp:
• BAM15 preclinical data (obese mice — fat mass, lean mass, temperature):

Related: MOTS-c · Mitochondrial Health Foundations — the anatomy, the dysfunction, the fix · 5-Amino-1MQ · SS-31 (Elamipretide) · NAD+