ATX-304

An orally dosed small molecule — not a peptide — that activates all AMPK complexes by protecting the kinase from dephosphorylation, currently in sponsor-run clinical development for obesity and cardiometabolic disease and sold as capsules alongside research peptides.

Also referenced as: O304, O-304, ATX304, 4-chloro-N-[2-[(4-chlorophenyl)methyl]-3-oxo-1,2,4-thiadiazol-5-yl]benzamide

Also appears in: Cardiovascular · Longevity

COAPublic product evidence10 certificate records mentioning this compound or product name1 provider-linked · 2 exact product matches→

---

What is ATX-304?

ATX-304 is a synthetic small molecule — 4-chloro-N-[2-[(4-chlorophenyl)methyl]-3-oxo-1,2,4-thiadiazol-5-yl]benzamide, CAS 1261289-04-6, formula C16H11Cl2N3O2S, molecular weight 380.2. It is a chlorinated thiadiazolone benzamide with no amino acid residues and therefore no peptide bonds. It does contain amide-type carbonyls — the benzamide that ties a 4-chlorobenzoyl group to the ring, and the ring’s own 3-oxo group sitting in an N–C(=O)–N motif — but a peptide bond is specifically the link between one amino acid and the next, and neither of these is that. Calling it a peptide is a category error, exactly as it is for 5-Amino-1MQ, and it appears in this database for the same reason: the vendors that stock research peptides also stock it, and it is marketed to the same audience.

The compound came out of a cell-based screen at Betagenon AB in Umeå, Sweden, and was published under the code O304 (Steneberg et al., JCI Insight, 2018). In 2023 Cambrian Bio built a pipeline company, Amplifier Therapeutics, around the asset and redesignated it ATX-304; Betagenon became a wholly-owned subsidiary and remains the sponsor of record on the European trial registrations. O304, O-304, and ATX-304 are one molecule under three sponsor labels.

How it works

AMPK is the cell’s energy sensor: it switches on when AMP and ADP rise relative to ATP, as they do during exercise and caloric restriction, and turns on catabolic pathways while shutting anabolic ones down. Most AMPK drugs either raise AMP indirectly (metformin) or bind allosterically at a site present on only some subunit combinations. ATX-304 does neither.

• It blocks dephosphorylation rather than activating the kinase allosterically. In vitro it suppressed protein phosphatase 2C–mediated removal of the Thr172 phosphate from the AMPK alpha subunit without inhibiting PP2C activity itself, raising phosphorylated AMPK across alpha, beta, and gamma trimer combinations — the basis for the “pan-AMPK” description. It did so without lowering cellular ATP, and only in cells expressing the upstream kinase LKB1, meaning it amplifies AMPK signaling already running rather than starting it from nothing (Steneberg et al., JCI Insight, 2018;3(12):e99114). Note that some of the secondary literature compresses this to “PP2C inhibition,” which is not what the original paper reported.
• It also acts as a mitochondrial uncoupler. The originating group later reported that O304 generates metabolic demand by uncoupling, the mechanism they proposed to explain sustained glucose uptake and utilization (Norlin et al., Communications Biology, 2023;6(1):877). An independent group in China confirmed it, showing dose-dependent increases in oxygen consumption after ATP synthase inhibition and reduced mitochondrial membrane potential in vascular smooth muscle and hepatoma cells (Li et al., Chemico-Biological Interactions, 2025;421:111788).
• In muscle it drives insulin-independent glucose uptake and appears to rest beta cells. In rodents O304 stimulated skeletal muscle and cardiac glucose uptake without insulin — “glucose effectiveness” — while preventing glycogen accumulation, and in db/db mice it preserved insulin secretion, beta-cell mass, and pancreatic insulin content (Norlin et al., 2023, above).
• In the liver it shifts the metabolic program. In a progressive fatty-liver mouse model, ATX-304 reduced body fat mass and blood cholesterol and mitigated steatosis and fibrosis development, alongside increased fatty acid oxidation and reduced lipid synthesis — though with pronounced local heterogeneity, including variation in lipid distribution between liver lobes (Holm et al., JCI Insight, 2025;10(7):e179990).

Research status

ATX-304 has more human exposure behind it than most compounds on the research market, and less published human evidence than the headline numbers suggest.

One published human trial. TELLUS was an exploratory, randomized, parallel-group, double-blind, placebo-controlled 28-day Phase IIa study of O304 at 1,000 mg/day in 65 type 2 diabetes patients on metformin for at least three months. It was reported inside the 2018 JCI Insight paper rather than as a standalone clinical publication, and three design points matter when reading it:

• Enrollment was on HbA1c (≥6.5% and ≤9.0% at screening), not fasting plasma glucose. The widely quoted glucose result — a mean fall of –0.60 mM on O304 versus –0.10 mM on placebo (P = 0.010) — comes from a post hoc subgroup of patients whose day-1 FPG fell between 7 and 13.3 mmol/L. The authors label it as post hoc in the paper.
• The HOMA-IR improvement (P = 0.0097 absolute) was a within-group change from day 1 to day 28 in the O304 arm, not a between-arm comparison.
• Blood pressure fell by a mean of 5.8 mmHg systolic and 3.8 mmHg diastolic on O304 against small increases on placebo (+1.2 and +0.9 mmHg), with the within-group changes reaching significance. The microvascular perfusion endpoint came with the authors’ own caveat that the cohort’s peripheral circulation was largely not depressed at baseline.

O304 was reported as well tolerated in TELLUS, with no ECG abnormalities and no effect on left ventricular mass.

No ClinicalTrials.gov record. No study registered on ClinicalTrials.gov lists ATX-304, O304, Betagenon, or Amplifier Therapeutics — the program’s registrations live in the EU CTIS system instead, and none of them covers TELLUS. The two identifiers printed in the 2018 paper’s study-approval statement resolve to unrelated industry trials by other sponsors — NCT00508287 is a Bristol-Myers Squibb study of BMS-686117, and NCT01167881 is a Boehringer Ingelheim study of empagliflozin. The published human data therefore cannot be cross-checked against a registered protocol.

Phase 1b, completed but unpublished. EU CTIS 2023-505967-36, sponsored by Betagenon AB, was a translational two-part Phase 1b of ATX-304 sodium salt tablets, registered against the condition “obesity with prediabetes or type 2 diabetes mellitus” and run in the Netherlands and Germany from September 2023 to May 2025. It is listed as ended, and no results have been posted to the registry. In June 2026 Cambrian Bio announced poster data at the American Diabetes Association’s 86th Scientific Sessions describing 23 adults given 400 mg once daily or placebo for eight weeks plus an eight-week open-label extension, reporting an 8% rise in resting metabolic rate, higher adiponectin, lower triglycerides, and reductions in liver fat and visceral adipose tissue, with minimal weight loss at that exposure. That account is a sponsor communication about a conference poster, not a peer-reviewed report. A separate multiple-ascending-dose Phase 1 in healthy overweight and obese volunteers (EU CTIS 2024-514744-91, Germany) is authorised with an EU start date of August 2026, and two announced Phase 2 studies — REWIRE-1 on muscle function and lipid metabolism, REWIRE-2 a weight-loss proof of concept — have not reported.

Preclinical breadth. The animal literature is genuinely wide for a compound this early: exercise-mimetic effects on metabolism, cardiac function, and exercise capacity in aged mice (Ericsson et al., Communications Biology, 2021;4(1):1306); prevention of obesity-associated gene expression changes in islets (López-Pérez et al., Scientific Reports, 2021;11(1):24410); protection against cisplatin-induced acute kidney injury (Katerelos et al., Biomedicine & Pharmacotherapy, 2024;175:116730); protection against kidney aging (Zhu et al., Frontiers in Pharmacology, 2022;13:836496); reduced abdominal aortic aneurysm formation (Sun and Du, Frontiers in Pharmacology, 2024;15:1457817); reduced mechanical allodynia after lumbar disk puncture in mice (Das et al., Anesthesia and Analgesia, 2022;135(6):1293–1303); and lifespan extension in C. elegans (Li et al., 2025, above). Several of these papers carry a disclosure that a senior author co-founded and holds shares in the sponsor.

No regulatory authority has approved ATX-304 for any indication, and no Phase 3 trial exists.

Common dosage forms

• Oral capsules and tablets in 60-count bottles, most often 50 mg or 100 mg per unit. This is the dominant presentation across vendors carrying it, and it matches the clinical route, which has always been oral.
• Some listings give only a milligram figure without naming the format, so the format has to be read off the product page rather than the product name.
• Two chemical forms circulate. The clinical tablets are the sodium salt, as stated in the registered Phase 1b title. The free compound (CAS 1261289-04-6, MW 380.2) and the sodium salt (CAS 2639972-26-0, MW 402.2) are separately registered substances, and because the salt is heavier, a milligram of it carries slightly less of the active moiety. A capsule labeled only in milligrams frequently does not state which form it contains.
• No injectable or lyophilized vial presentation is characteristic of this compound, which is unusual for this market and consistent with an orally dosed small molecule.

This section describes formats only and is not dosing guidance.

Key considerations

• It is not a peptide. ATX-304 is a chlorinated thiadiazolone benzamide. Nothing about its handling, stability, purity testing, or pharmacology follows from peptide conventions, and a certificate of analysis for it should be a small-molecule assay, not a peptide purity run.
• The human evidence is thin and sponsor-controlled. One 28-day exploratory Phase IIa whose headline glucose result is an acknowledged post hoc subgroup, plus one completed Phase 1b known only from a sponsor announcement of a conference poster. There is no independent replication of any human finding, no registered protocol to check the published trial against, no Phase 3, and no approval anywhere.
• Mitochondrial uncoupling is the open safety question. The compound’s own literature, including work from the originating group, identifies it as a mitochondrial uncoupler — a mechanism class with a lethal history in unsupervised use (2,4-dinitrophenol). The sponsor specifically reported no adverse events indicative of mitochondrial failure at 400 mg/day, including no rise in continuously monitored core body temperature or 24-hour heart rate. That is reassuring at that exposure, and it is precisely the signal that would need watching at higher ones, which is where the announced Phase 2 program is headed.
• Naming collisions are easy to fall into. Searching only “O304” misses papers indexed under “ATX-304,” and vice versa — the cisplatin kidney-injury paper above is indexed under ATX-304 only, and the mechanism papers under O304 only. Some vendor listings pair it with the label “OS-01,” which belongs to an unrelated topical cosmetic peptide from a different developer; the two share no chemistry. The “ATX” prefix is also reused across unrelated programs, so a prefix-based literature search returns a large amount of irrelevant material.
• “First-in-class” describes the mechanism, not the outcomes. The sponsor’s framing is that ATX-304 is the first pan-AMPK activator — one that raises activity across all trimer combinations — to reach clinical study. The qualifier is doing the work: subunit-selective direct AMPK activators reached Phase 1 before it. Either way the phrase says nothing about efficacy having been established.