AMPK activation, mitochondrial metabolism, energy expenditure — and what the human data actually show
A Protokol X intelligence brief on ATX-304, formerly O304, and the emerging attempt to pharmacologically influence cellular energy sensing without pretending a pill can replace exercise.
ATX-304, previously known as O304, is an investigational oral small molecule designed to influence the AMPK energy-sensing network and mitochondrial metabolism.
Those human signals make ATX-304 more than a preclinical curiosity. But they do not establish long-term weight-loss efficacy, cardiovascular benefit, durable body-composition effects, or an “exercise pill.”
AMP-activated protein kinase — AMPK — is a central cellular energy sensor. When cellular energy availability falls, AMPK signaling helps shift metabolism away from energy-consuming anabolic processes and toward pathways that generate or conserve usable energy.
That includes effects on glucose uptake, fatty-acid metabolism, mitochondrial activity, lipid synthesis, and other processes involved in metabolic homeostasis.
Exercise naturally engages parts of this network because working tissue creates an energy demand. That overlap is why compounds targeting AMPK are sometimes described with terms such as exercise mimetic.
“Exercise mimetic” describes overlap in selected molecular or metabolic pathways. It does not mean ATX-304 reproduces the full cardiovascular, musculoskeletal, neurological, functional, or adaptive effects of physical exercise.
The molecule first appeared in the scientific literature as O304, developed by Betagenon. The current clinical-development name is ATX-304, now being advanced by Amplifier Therapeutics within Cambrian Bio's pipeline.
The older O304 literature matters because it gives ATX-304 a longer human and mechanistic history than its newer name might suggest.
In the earlier TELLUS Phase IIa study, 65 adults with type 2 diabetes treated with metformin were randomized to O304 or placebo for 28 days. The study evaluated fasting plasma glucose, safety, and exploratory metabolic and vascular measures.
ATX-304 is described as increasing AMPK activity, including by suppressing dephosphorylation of the activating Thr172 site.
Experimental work also identifies increased cellular respiration and mitochondrial proton conductance as part of the compound's metabolic profile.
The strategic idea is to influence both sides of cellular energy balance: pathways governing energy supply and pathways governing energy demand.
This is not appetite pharmacology wearing a different label. ATX-304 is being developed around cellular energy sensing and metabolism.
The most important new evidence comes from a randomized, double-blind, placebo-controlled Phase 1b study in 23 adults with obesity and prediabetes. Participants received 400 mg ATX-304 once daily or placebo for eight weeks, with an optional open-label extension.
The reported data also showed a significant increase in adiponectin. Treatment-emergent adverse events were predominantly mild and occurred at a frequency similar to placebo in this small study.
Critically, weight loss was minimal at the exposure tested.
The 8% figure refers to measured resting metabolic rate. It is not an 8% reduction in body weight, an 8% increase in total daily calorie expenditure, or proof of meaningful long-term fat loss. The Phase 1b dataset was presented at the American Diabetes Association's 2026 Scientific Sessions; detailed peer-reviewed publication of the full trial remains an important next step.
A drug can produce measurable metabolic changes without producing large short-term changes on the scale.
The Phase 1b study was designed around safety, pharmacokinetics, pharmacodynamics, and human translation of metabolic effects. The exposure tested produced an RMR signal and changes in several metabolic measures, but little weight loss.
That makes the result more interesting scientifically, not less: it separates the question “Does the mechanism do anything in humans?” from the later question “Can higher or optimized exposure produce clinically meaningful weight loss?”
Mechanism first. Outcome second. Do not turn a metabolic signal into a weight-loss claim.
The O304 story did not begin in 2026. A 2018 peer-reviewed paper reported the 28-day TELLUS Phase IIa trial in adults with type 2 diabetes receiving metformin.
That trial randomized 65 participants to O304 or placebo and explored fasting plasma glucose along with safety and metabolic and vascular endpoints. The study supported continued investigation of the compound in glucose homeostasis and microvascular function.
The formulation and development strategy have evolved since then, so the older trial should not be treated as interchangeable with the current obesity program. But it establishes that the molecule has already crossed the animal-to-human boundary more than once.
A 2025 JCI Insight study examined ATX-304 in a mouse model of progressive metabolic dysfunction–associated steatotic liver disease, or MASLD.
Researchers reported reduced body fat mass, lower blood cholesterol, reduced liver steatosis, mitigation of fibrosis, and a liver metabolic shift toward increased fatty-acid oxidation and reduced lipid synthesis.
The human Phase 1b report later showed a statistically significant reduction in liver fat measured by MRI-PDFF. That does not prove treatment of MASLD, but it creates a translational thread worth following.
Preclinical research has also tested ATX-304 with semaglutide in diet-induced obese mice.
Acts through GLP-1 receptor signaling and can reduce food intake through appetite and satiety pathways.
Targets the AMPK network and mitochondrial metabolism without relying primarily on appetite suppression.
In the reported mouse experiments, ATX-304 increased energy expenditure and was studied alone, in combination with semaglutide, and after semaglutide withdrawal. Researchers reported additional weight-loss effects and reduced weight regain in those animal models.
No human trial has established that adding ATX-304 to semaglutide improves weight loss, preserves lean mass, prevents weight regain, or permits lower incretin dosing. Those remain research questions.
No — at least not in the ordinary meaning of that phrase.
Exercise changes far more than resting metabolic rate. It produces mechanical loading, cardiovascular conditioning, neuromuscular adaptation, changes in bone, connective tissue, muscle strength, aerobic capacity, cognition, and numerous endocrine and immune effects.
ATX-304 is interesting because it targets a network that participates in the cellular response to energy demand. That makes the exercise comparison useful for explaining the biology, but dangerous if turned into a replacement claim.
| Claim | Evidence Level | What the Research Actually Shows |
|---|---|---|
| ATX-304 was previously called O304 | ESTABLISHED | The scientific and development literature identifies ATX-304 as formerly O304. |
| ATX-304 targets AMPK signaling | MECHANISTIC + CLINICAL-STAGE | Published research and current development materials describe AMPK-network activation as a central mechanism. |
| ATX-304 also affects mitochondrial metabolism | MECHANISTIC | Experimental studies report increased respiration and mitochondrial proton conductance/uncoupling-related activity. |
| ATX-304 increased resting metabolic rate by 8% | PHASE 1b HUMAN DATA | Reported in 23 adults with obesity and prediabetes at ADA 2026; p<0.01. |
| ATX-304 reduced liver fat | PHASE 1b HUMAN SIGNAL | MRI-PDFF liver fat decreased significantly in the small Phase 1b study; larger confirmatory trials are needed. |
| ATX-304 reduced visceral adipose tissue | PHASE 1b HUMAN SIGNAL | A statistically significant decrease was reported in the Phase 1b study. |
| ATX-304 caused major human weight loss | NOT SHOWN | Weight loss was minimal at the Phase 1b exposure tested. |
| ATX-304 works with semaglutide | ANIMAL COMBINATION DATA | Combination and withdrawal experiments are preclinical; human combination efficacy is unestablished. |
| ATX-304 preserves muscle during human weight loss | UNESTABLISHED | No human evidence currently demonstrates this outcome. |
| ATX-304 replaces exercise | FALSE / MISLEADING | Pathway overlap does not reproduce the full physiological effects of exercise. |
| ATX-304 is approved for obesity | FALSE | It remains investigational and is entering/planned for Phase 2 development. |
Cambrian's current pipeline identifies Phase 2 studies REWIRE-1 and REWIRE-2 as planned. Those studies should begin answering the questions that Phase 1b could not.
Those are the milestones that will determine whether ATX-304 becomes a meaningful metabolic therapy or simply an interesting AMPK experiment.
Measure the signal. Respect the boundary.
ATX-304 deserves attention because the story has moved beyond mice.
The molecule — formerly O304 — has a human research history in type 2 diabetes and now a newer Phase 1b signal in adults with obesity and prediabetes.
The headline result is an 8% increase in resting metabolic rate at the tested exposure, accompanied by significant changes in liver fat, visceral adipose tissue, triglycerides, and adiponectin.
But the most important counterweight is equally clear: weight loss was minimal.
That means ATX-304 has demonstrated evidence of human metabolic activity. It has not yet demonstrated that it can become a major weight-loss therapy, replace exercise, preserve muscle during weight loss, or improve incretin outcomes in humans.
Phase 2 is where those claims begin to face the test.