Quick Answer
All four raise AMPK activity, by different routes. Metformin acts indirectly, through mitochondrial complex I and, at low concentrations, a lysosomal pathway. AICAR becomes ZMP inside cells, which imitates AMP at the enzyme itself. The peptide MOTS-c disrupts the folate cycle so that the cell's own AICAR builds up. ATX-304, a small molecule rather than a peptide, was first described as protecting AMPK's activating phosphorylation and later shown to uncouple mitochondria too; which action dominates is debated. Only metformin is an approved medicine with long-term outcome data.
The label "AMPK activator" hides very different pharmacology. This comparison sets ATX-304 (previously O-304), a clinical-stage candidate, against three long-studied reference points: metformin, AICAR and the peptide MOTS-c. The ATX-304 research profile covers development history and status; this article concentrates on mechanism and on how far each compound's evidence reaches.
How AMPK is switched on
AMP-activated protein kinase (AMPK) has a catalytic alpha subunit and regulatory beta and gamma subunits. It responds to falling cellular energy, signalled by rising AMP to ATP and ADP to ATP ratios: AMP and ADP bind the gamma subunit, promote phosphorylation of threonine 172 on the alpha subunit (the main activating mark) and shield it from phosphatases, and AMP also activates the enzyme allosterically. Active AMPK favours ATP-generating processes such as glucose uptake and fatty acid oxidation over ATP-consuming ones such as lipid synthesis (Steinberg and Hardie, 2023).
Indirect activators create energy stress or engage an upstream pathway, and usually hit other targets too. Direct activators act on the AMPK complex itself, by imitating AMP or by binding a separate allosteric drug site; the pan-AMPK activator MK-8722 is the best-characterised synthetic example (Myers et al., 2017). The distinction matters because an indirect activator's effects cannot be credited to AMPK without genetic controls such as AMPK-deficient cells.
The four compounds side by side
| Compound | What it is | Route to AMPK | Strongest human evidence |
|---|---|---|---|
| ATX-304 (O-304) | Synthetic small molecule, not a peptide | Slows removal of the threonine 172 mark; also uncouples mitochondria | Two small, short, developer-run studies |
| Metformin | Biguanide; approved medicine | Indirect: complex I and energy state; lysosomal route at low concentrations | Decades of clinical use; long-term outcome trials such as UKPDS 34 |
| AICAR | Adenosine analogue | Converted to ZMP, an AMP mimic acting at the gamma subunit | One large surgical trial (as acadesine), stopped for futility |
| MOTS-c | 16-amino-acid mitochondrial-derived peptide | Indirect: folate cycle inhibition raises the cell's own AICAR | No published randomised trial of the peptide itself |
ATX-304: a mechanism still being argued
ATX-304 was discovered by the Swedish company Betagenon AB, working with Umeå University, and published as O304. Amplifier Therapeutics, launched by Cambrian Bio in 2023, acquired Betagenon and now develops it as ATX-304.
The original account was specific: with recombinant AMPK, O304 suppressed phosphatase removal of the threonine 172 mark; in cells it raised phosphorylated AMPK without lowering ATP; and it did not activate the enzyme allosterically (Steneberg et al., 2018). On that description it sits nearer the direct activators than metformin does.
In 2023 the originating group reported that O304 is also a mitochondrial uncoupler: protons leak back across the inner membrane, so respiration rises without matching ATP production (Norlin et al., 2023). An unaffiliated laboratory confirmed this in 2025, reporting reduced membrane potential in two cell lines (Li et al., 2025). Uncoupling is a recognised indirect route to AMPK, because lower membrane potential can mean lower ATP.
The developer disputes that uncoupling explains the AMPK effect. A company co-authored abstract reported proton leak in isolated mouse muscle mitochondria, but preserved membrane potential in intact cells, no fall in ATP in human hepatocytes, and similar activation of normal AMPK and an AMP-insensitive mutant (Schneider et al., 2026). In a cisplatin kidney-injury study, ATX-304 protected mice and kidney cells, and the protection disappeared in fibroblasts lacking AMPK (Katerelos et al., 2024). The fair summary: both actions are documented, groups disagree about whether the uncoupling lowers cellular energy in intact cells, and the developer's own label, a dual AMPK and mitochondrial activator, fits better than "direct activator".
What the ATX-304 data show
Rodent studies form the deepest layer. In diet-induced obese mice, O304 increased muscle glucose uptake and reduced markers of beta cell stress (Steneberg et al., 2018). In aged mice it prevented and reversed age-associated insulin resistance and improved cardiac function and treadmill capacity; heart weight was 15% higher, without fibrosis or glycogen build-up, which the authors read as exercise-like adaptation (Ericsson et al., 2021). Heart findings are watched closely because MK-8722 caused cardiac hypertrophy with raised cardiac glycogen in rodents and monkeys (Myers et al., 2017). In a mouse model of progressive fatty liver disease, ATX-304 reduced steatosis and fibrosis development (Holm et al., 2025).
Human data are limited. TELLUS was a 28-day, randomised, double-blind, placebo-controlled Phase 2a trial in 65 people with type 2 diabetes on background metformin. Its fasting glucose result came from a post hoc analysis restricted to a baseline glucose range, alongside reported improvements in insulin resistance, calf microvascular perfusion and blood pressure (Steneberg et al., 2018). A Phase 1b study in 23 adults with obesity and prediabetes, randomised two to one against placebo for eight weeks, reported higher adiponectin, lower liver and visceral fat and a higher resting metabolic rate, all measured against baseline rather than placebo (Thieroff-Ekerdt et al., 2026). It exists only as a conference abstract written largely by company employees, and it reports no weight-loss result.
ClinicalTrials.gov currently lists no study under either code, and the developer describes larger Phase 2 trials as planned.
Metformin: the indirect benchmark
Metformin, an approved and widely used medicine, is the benchmark. Early work placed its primary action at mitochondrial complex I (Owen et al., 2000) and showed AMPK activation in hepatocytes and rat skeletal muscle (Zhou et al., 2001). AMPK is not the whole story: in mice lacking liver AMPK, metformin still lowered blood glucose, through a fall in hepatic energy state (Foretz et al., 2010). Later, clinically relevant concentrations were reported to activate AMPK through a lysosomal pathway involving PEN2, without changing cellular AMP (Ma et al., 2022). What sets metformin apart is the human record, including the UKPDS 34 randomised trial in overweight people with newly diagnosed type 2 diabetes, followed for a median of more than ten years (UKPDS Group, 1998).
AICAR: the classic tool compound
AICAR is an adenosine analogue converted inside cells to ZMP, which mimics both activating effects of AMP on AMPK; in the original rat hepatocyte work it did so without disturbing ATP, ADP or AMP (Corton et al., 1995). It became the default research activator; in sedentary mice it induced muscle metabolic genes and increased running endurance (Narkar et al., 2008). Its weakness is selectivity: ZMP is also a purine synthesis intermediate, and a 2021 systematic review concluded that numerous effects once credited to AMPK are AMPK-independent (Višnjić et al., 2021). As acadesine, it was tested in humans for protection during coronary bypass surgery; the RED-CABG trial was stopped for futility (Newman et al., 2012).
MOTS-c: a peptide route to the same intermediate
MOTS-c is a 16-amino-acid peptide encoded within the mitochondrial 12S rRNA gene. Its mechanism links neatly to AICAR: in cell experiments it inhibited the folate cycle and the purine synthesis tethered to it, so that endogenous AICAR accumulated and AMPK was activated, and in mice it prevented diet-induced obesity and insulin resistance (Lee et al., 2015). The same group later reported that exercise raises endogenous MOTS-c in human muscle and blood, and that MOTS-c improved physical performance in young, middle-aged and old mice (Reynolds et al., 2021). Intervention data come from rodents and cells; human data are observational or exercise physiology, and no published randomised trial has tested MOTS-c itself. The MOTS-c research update covers this in more detail.
Reading the comparison honestly
Ranked by human evidence, metformin is far ahead of the other three. Ranked by how cleanly each isolates AMPK, none scores well: each has documented actions beyond the kinase, and for ATX-304 the balance between AMPK activation and uncoupling is an open question.
For laboratory work, that argues for the same controls whichever compound is used: AMPK-deficient cells or tissues, adenine nucleotide measurements alongside phosphorylated AMPK and its substrate acetyl-CoA carboxylase, and a structurally unrelated second activator before an effect is attributed to AMPK. ATX-304 is investigational and is not supplied by Peptx.
References
- Steinberg GR, Hardie DG. 2023. "New insights into activation and function of the AMPK." Nature Reviews Molecular Cell Biology 24(4):255-272. PMID 36316383
- Myers RW, et al. 2017. "Systemic pan-AMPK activator MK-8722 improves glucose homeostasis but induces cardiac hypertrophy." Science 357(6350):507-511. PMID 28705990
- Steneberg P, et al. 2018. "PAN-AMPK activator O304 improves glucose homeostasis and microvascular perfusion in mice and type 2 diabetes patients." JCI Insight 3(12):e99114. PMID 29925691
- Norlin S, et al. 2023. "O304 ameliorates hyperglycemia in mice by dually promoting muscle glucose effectiveness and preserving β-cell function." Communications Biology 6(1):877. PMID 37626210
- Li WF, et al. 2025. "O304 is a mitochondrial uncoupler which extends C. elegans lifespan and induces vasorelaxation of rat mesenteric arteries." Chemico-Biological Interactions 421:111788. PMID 41130350
- Schneider EJ, et al. 2026. "1788-P: AMPK Activation by ATX-304 Is Not Secondary to Changes in ATP Levels." Diabetes 75(Supplement 1), conference abstract. doi:10.2337/db26-1788-p
- Katerelos M, et al. 2024. "The AMPK activator ATX-304 alters cellular metabolism to protect against cisplatin-induced acute kidney injury." Biomedicine and Pharmacotherapy 175:116730. PMID 38749175
- Ericsson M, et al. 2021. "AMPK activator O304 improves metabolic and cardiac function, and exercise capacity in aged mice." Communications Biology 4(1):1306. PMID 34795407
- Holm E, et al. 2025. "AMPK activator ATX-304 reduces oxidative stress and improves MASLD via metabolic switching." JCI Insight 10(7):e179990. PMID 40197369
- Thieroff-Ekerdt RI, et al. 2026. "1782-P: Phase 1b Study of AMPK/Mitochondrial Activator ATX-304 in Prediabetic Obese Participants." Diabetes 75(Supplement 1), conference abstract. doi:10.2337/db26-1782-p
- Owen MR, et al. 2000. "Evidence that metformin exerts its anti-diabetic effects through inhibition of complex 1 of the mitochondrial respiratory chain." Biochemical Journal 348(Pt 3):607-614. PMID 10839993
- Zhou G, et al. 2001. "Role of AMP-activated protein kinase in mechanism of metformin action." Journal of Clinical Investigation 108(8):1167-1174. PMID 11602624
- Foretz M, et al. 2010. "Metformin inhibits hepatic gluconeogenesis in mice independently of the LKB1/AMPK pathway via a decrease in hepatic energy state." Journal of Clinical Investigation 120(7):2355-2369. PMID 20577053
- Ma T, et al. 2022. "Low-dose metformin targets the lysosomal AMPK pathway through PEN2." Nature 603(7899):159-165. PMID 35197629
- UKPDS Group. 1998. "Effect of intensive blood-glucose control with metformin on complications in overweight patients with type 2 diabetes (UKPDS 34). UK Prospective Diabetes Study (UKPDS) Group." Lancet 352(9131):854-865. PMID 9742977
- Corton JM, et al. 1995. "5-aminoimidazole-4-carboxamide ribonucleoside. A specific method for activating AMP-activated protein kinase in intact cells?" European Journal of Biochemistry 229(2):558-565. PMID 7744080
- Narkar VA, et al. 2008. "AMPK and PPARdelta agonists are exercise mimetics." Cell 134(3):405-415. PMID 18674809
- Višnjić D, et al. 2021. "AICAr, a Widely Used AMPK Activator with Important AMPK-Independent Effects: A Systematic Review." Cells 10(5):1095. PMID 34064363
- Newman MF, et al. 2012. "Effect of adenosine-regulating agent acadesine on morbidity and mortality associated with coronary artery bypass grafting: the RED-CABG randomized controlled trial." JAMA 308(2):157-164. PMID 22782417
- Lee C, et al. 2015. "The mitochondrial-derived peptide MOTS-c promotes metabolic homeostasis and reduces obesity and insulin resistance." Cell Metabolism 21(3):443-454. PMID 25738459
- Reynolds JC, et al. 2021. "MOTS-c is an exercise-induced mitochondrial-encoded regulator of age-dependent physical decline and muscle homeostasis." Nature Communications 12(1):470. PMID 33473109