Ecnoglutide (HRS9531): once-weekly weight-loss & diabetes injection
An injectable drug in the same family as Ozempic that lowers blood sugar and curbs appetite; still experimental and not approved anywhere.
- Class
- Long-acting GLP-1 receptor agonist (cAMP-biased)
- Status
- Approved or pending approval in China (NMPA); not approved by FDA, EMA, MHRA, Health Canada, or PMDA
- Best-supported effect
- Approximately 13.2% mean body weight reduction at 40 weeks in overweight and obese Chinese adults (Phase 3 RCT); positive glycemic outcomes in a parallel Phase 3 program in type 2 diabetes in China
- Main caveat
- All published efficacy data are from Chinese populations; no Western-population trials, no head-to-head data against semaglutide or tirzepatide, and no cardiovascular outcomes trial have been reported
A researcher, an agent, or an algorithm wrote down the sequence and picked a target to hit.
An AI model like OpenFold3 or AlphaFold built a 3D structure and scored how well it fits the binding site.
A second contributor repeated the computation on their own hardware and the scores matched.
A chemistry service or a researcher ordered the sequence, it was manufactured, and mass spectrometry confirmed the right molecule was produced.
A binding or activity measurement confirmed that it actually does what the computer predicted — or didn't.
What this is
Ecnoglutide (development code XW003) is a once-weekly injectable peptide that mimics GLP-1, a gut hormone the body releases after eating that lowers blood sugar and reduces appetite. It belongs to the same drug class as semaglutide (Ozempic, Wegovy) and liraglutide (Victoza, Saxenda), but is engineered with a pharmacological twist: at the GLP-1 receptor, ecnoglutide preferentially activates one downstream signal (cAMP) while largely avoiding another (β-arrestin recruitment). It is being developed by Hangzhou Sciwind Biosciences for type 2 diabetes and chronic weight management, with a clinical development program centered in China. As of early 2026, ecnoglutide is approved or under review by China's National Medical Products Administration (NMPA); it is not FDA-approved, EMA-authorized, or approved by any other Western regulator.
History
Ecnoglutide originated at Hangzhou Sciwind Biosciences, a Chinese biopharmaceutical company focused on metabolic disease. The discovery program was published by Guo and colleagues (Molecular Metabolism, 2023), describing the screening of acylated GLP-1 analogs composed only of natural amino acids and the selection of a candidate engineered to preferentially activate cAMP signaling over β-arrestin recruitment at the GLP-1 receptor. The biased-agonism rationale was that reducing β-arrestin engagement might slow receptor desensitization and internalization, potentially improving durability or tolerability versus balanced GLP-1 agonists. Sciwind advanced the molecule through first-in-human Phase 1, Phase 2 dose-ranging in both overweight/obesity and type 2 diabetes, and pivotal Phase 3 programs in both indications in Chinese populations. The Phase 3 weight-loss trial was published as Ji and colleagues (Lancet Diabetes & Endocrinology, 2025), and the Phase 3 type 2 diabetes monotherapy trial (EECOH-1) and a dulaglutide-controlled non-inferiority trial (EECOH-2) were subsequently published in Nature Communications and Lancet Diabetes & Endocrinology in 2025.
What it does
In humans, ecnoglutide reduces appetite, slows gastric emptying, enhances glucose-dependent insulin secretion, and suppresses glucagon — the same physiological actions that make GLP-1 receptor agonists effective for type 2 diabetes and chronic weight management. The dominant patient-experienced effects in the published Phase 3 weight-loss trial were reduced food intake, body-weight reduction, and gastrointestinal symptoms (nausea, vomiting, diarrhea) that clustered at dose-escalation steps and attenuated with continued treatment (Ji and colleagues, Lancet Diabetes & Endocrinology 2025).
Evidence
- Human: Phase 3 evidence in Chinese populations. The pivotal weight-loss trial (Ji and colleagues, Lancet Diabetes & Endocrinology 2025) randomized 664 adults with overweight or obesity to ecnoglutide 1.2, 1.8, or 2.4 mg weekly versus placebo for 40 weeks; the 2.4 mg arm produced a least-squares mean body-weight change of approximately −13.2%. In the type 2 diabetes program, EECOH-1 (Nature Communications 2025) demonstrated significant HbA1c reductions versus placebo, and EECOH-2 (Lancet Diabetes & Endocrinology 2025) showed non-inferiority versus dulaglutide on glycemic endpoints over 52 weeks. A first-in-human Phase 1 study established pharmacokinetics, pharmacodynamics, and safety in healthy volunteers and supported once-weekly dosing (Guo and colleagues, Molecular Metabolism 2023). No Western-population trial, no completed cardiovascular outcomes trial, and no head-to-head trial versus semaglutide or tirzepatide have been reported.
- Animal: Preclinical pharmacology in rodent models (Guo and colleagues, Molecular Metabolism 2023) characterized the biased-agonism signature and demonstrated dose-dependent glycemic and body-weight effects, with reported reductions in blood glucose and body weight in diet-induced obesity and diabetic models.
- In vitro: Receptor pharmacology showed potent cAMP induction with minimal GLP-1 receptor internalization or β-arrestin recruitment (Guo and colleagues, 2023), establishing the cAMP-biased signature that distinguishes ecnoglutide from balanced GLP-1 agonists.
Known effects
- Weight reduction — Phase 3 in Chinese adults with overweight or obesity (~13.2% mean reduction at 40 weeks on 2.4 mg weekly)
- Glycemic improvement in type 2 diabetes — Phase 3 in Chinese adults (EECOH-1 versus placebo; EECOH-2 versus dulaglutide)
- Appetite suppression and delayed gastric emptying — class-consistent GLP-1 effects, observed across Phase 2 and Phase 3 programs
- Cardiovascular risk reduction — not established; no SELECT- or LEADER-equivalent outcomes trial has been completed for ecnoglutide
Safety signals
Gastrointestinal symptoms are the dominant adverse events reported across Phase 2 and Phase 3 trials, consistent with the GLP-1 class:
- Nausea (most common), vomiting, diarrhea, and constipation — concentrated at dose-escalation steps and typically attenuating over 1–2 weeks at each step
- Decreased appetite — an expected pharmacological effect rather than an adverse event
- Injection-site reactions and headache — reported at low frequencies, class-consistent
Beyond ecnoglutide-specific data, GLP-1 class warnings inform the broader safety framing. Rodent C-cell tumor findings underlie a class-level contraindication in personal or family history of medullary thyroid carcinoma (MTC) and Multiple Endocrine Neoplasia type 2 (MEN2); rare post-marketing pancreatitis signals across the class motivate caution in patients with a history of pancreatitis; and class-level gastric-emptying delay can worsen severe gastroparesis. Whether the cAMP-biased pharmacology of ecnoglutide alters the magnitude of any of these signals relative to balanced GLP-1 agonists is unstudied. Long-term safety beyond the 40–52 week trial windows is not characterized, and real-world pharmacovigilance is orders of magnitude smaller than for semaglutide.
Regulatory status
- China (NMPA): Approved or under regulatory review following the Phase 3 program in overweight and obesity and the Phase 3 program in type 2 diabetes. Chinese approval does not constitute approval by any Western regulator.
- US (FDA): Not approved. No US regulatory filing has been reported as of early 2026.
- EU (EMA), UK (MHRA), Canada (Health Canada), Australia (TGA): Not reviewed.
- WADA: Not named on the Prohibited List. Outside China, ecnoglutide is not approved for human therapeutic use by any Western governmental regulatory health authority, which places it within the WADA S0 catch-all category covering substances not approved for human therapeutic use. Athletes subject to WADA, USADA, UKAD, or equivalent programs should assume the peptide is prohibited in and out of competition.
Mechanism
Ecnoglutide is an acylated GLP-1 analog composed entirely of natural amino acids; the GLP-1(7-37) backbone is extended at the C-terminus and modified with a fatty-acid linker that enables albumin binding and supports once-weekly dosing — the stored 34-letter sequence reflects the unmodified backbone, and the fatty-acid moiety that gives the molecule its multi-day half-life is not represented in the raw letters. Reported steady-state half-life in the multi-day range supports once-weekly administration (Guo and colleagues, Molecular Metabolism 2023).
At the receptor, ecnoglutide binds the GLP-1 receptor, a Class B G-protein-coupled receptor expressed on pancreatic beta cells, hypothalamic and brainstem neurons, and cardiovascular and gastrointestinal tissue. GLP-1 receptor activation couples to two principal downstream arms: the Gαs–adenylyl cyclase–cAMP axis, which drives glucose-dependent insulin secretion, glucagon suppression, slowed gastric emptying, and central appetite suppression via arcuate-nucleus and area-postrema circuits; and β-arrestin recruitment, which terminates Gs signaling and drives receptor internalization and desensitization. Ecnoglutide's distinguishing feature is preferential activation of the cAMP arm with minimal β-arrestin recruitment in vitro — Guo and colleagues (2023) reported potent cAMP induction alongside minimal receptor internalization in cell-based assays, defining the biased-agonism signature. The theoretical clinical implication is that reduced β-arrestin engagement might sustain receptor responsiveness over chronic dosing, potentially improving durability or tolerability. However, no head-to-head trial has demonstrated that this in vitro signature translates into measurable patient-level differentiation versus semaglutide or tirzepatide: Phase 3 GI tolerability has been class-consistent, and the historical track record of converting biased-agonism theory into clinically superior outcomes across receptor systems is mixed.
Myths and misconceptions
- "Ecnoglutide is a better semaglutide because it's biased." Biased agonism at the GLP-1 receptor is a pharmacologically interesting design choice with a plausible theoretical advantage. But clinical differentiation versus balanced GLP-1 agonists has not been demonstrated in head-to-head human trials. Ecnoglutide's Phase 3 weight loss (~13.2% at 40 weeks on 2.4 mg) sits in a similar range to semaglutide's STEP-1 result, but cross-trial comparisons across different populations do not support a "better" claim.
- "Approval in China means it is safe to source from research-chemical suppliers." Chinese approval (or pending approval) applies to a GMP-manufactured clinical product distributed through regulated pharmacy channels. Research-chemical product marketed as "XW003" is not the same — identity, purity, potency, and sterility are not verifiable, and cross-jurisdiction regulatory approval does not legitimize unregulated supply.
- "The biased-agonism mechanism guarantees fewer side effects." The biased-agonism rationale suggests that reduced β-arrestin recruitment could improve tolerability. Whether that translates to fewer reports of nausea, vomiting, or diarrhea versus semaglutide or tirzepatide in head-to-head comparison is not established; reported Phase 3 GI tolerability has been class-consistent.
Open questions
- Head-to-head versus semaglutide and tirzepatide. No completed trial directly compares ecnoglutide against the market-leading GLP-1 therapies for weight loss; EECOH-2 provided dulaglutide-comparator data in type 2 diabetes but not against semaglutide or tirzepatide.
- Generalizability beyond Chinese populations. All published Phase 1, 2, and 3 efficacy and safety data come from Chinese adults; pharmacokinetics, tolerability, and efficacy in Western populations have not been characterized in dedicated trials.
- Cardiovascular outcomes. No trial equivalent to semaglutide's SELECT or liraglutide's LEADER has been completed. Class extrapolation is not a substitute for indication-specific outcome data.
- Long-term safety. The longest published follow-up is in the 40–52 week range. Multi-year real-world pharmacovigilance, as exists for semaglutide, is not yet available for ecnoglutide.
- Post-discontinuation weight trajectory. No dedicated ecnoglutide discontinuation study has been published. Class-level data from semaglutide's STEP-1 extension predict meaningful weight regain after stopping; whether biased-agonism pharmacology alters this is unstudied.
- Western regulatory trajectory. Whether Sciwind pursues independent FDA or EMA filings, partners the asset, or remains focused on Asian markets will determine ecnoglutide's global accessibility over the next several years.
- Clinical translation of the biased-agonism signature. The in vitro cAMP-bias profile is well characterized; converting biased-agonism theory into clinically superior outcomes has historically proven difficult across multiple GPCR systems, and ecnoglutide's contribution to that translational question will depend on head-to-head data that have not yet been generated.
Related peptides
- Semaglutide (/card/pep-00016) — balanced once-weekly GLP-1 receptor agonist; the standard-of-care benchmark for any new GLP-1 agonist and the natural comparator for the biased-agonism differentiation hypothesis.
- Liraglutide (/card/pep-10868) — once-daily GLP-1 receptor agonist; the first-generation acylated GLP-1 analog that established the albumin-binding fatty-acid conjugation strategy used across the long-acting class.
- Tirzepatide — dual GIP/GLP-1 receptor agonist; the non-monoselective benchmark for weight loss (SURMOUNT-1 reported ~20% mean weight reduction in a Western population).
- Retatrutide — investigational triple GLP-1/GIP/glucagon receptor agonist; a distinct receptor profile within the broader incretin-mimetic landscape.
Research directions for this peptide, selected from the current sources — hypotheses you can explore and model. None of it is proven yet; tap any one to see the full thinking.
By avoiding the signal that causes GLP-1 receptors to be pulled inside cells and switched off, could ecnoglutide maintain appetite control better over a week?
If this works as designed, patients might experience more consistent appetite suppression throughout the week rather than the drug 'wearing off' before the next injection, potentially improving weight loss outcomes.
Could avoiding one particular receptor signal reduce the nausea that makes GLP-1 drugs hard for some patients to tolerate?
If the hypothesis holds, ecnoglutide could allow patients to reach effective doses without the nausea that causes many people to stop taking GLP-1 drugs, improving long-term adherence and weight loss outcomes.
If ecnoglutide keeps the beneficial cellular signal in heart muscle cells active longer than other GLP-1 drugs, could it reduce heart attack risk more effectively?
GLP-1 drugs already reduce heart disease risk, but if ecnoglutide's design amplifies that effect in heart muscle, it could become the preferred choice for patients with obesity and heart disease.
Because ecnoglutide uses only natural protein building blocks, could it be grown in bacteria rather than synthesized chemically, making it far cheaper to produce?
If ecnoglutide can be manufactured by fermentation, it could become affordable in countries where current GLP-1 drugs are inaccessible due to cost, potentially treating millions more people with obesity and diabetes worldwide.
Does the short flexible extension at the end of ecnoglutide keep its fatty acid anchor away from the working end of the molecule, making its receptor signal more selective?
If confirmed, drug designers could use this structural rule to build better-tolerated obesity drugs by deliberately positioning fatty acid tails to avoid interfering with receptor binding geometry.
▸full evidence table2 metrics
| metric | value | tool |
|---|---|---|
| ipTM | 0.7775827050209045 | openfold3-mlx |
| ranking score | 0.8394559025764465 | openfold3-mlx |
▸structural qualityopenfold3
| metric | value | note |
|---|---|---|
| gpde | 0.741 | global PDE — lower = better |
| disorder | 0.146 | fraction disordered |
| chain pair ipTM (A, B) | 0.778 | interface quality |
▸3-letter notation
▸recipeopenfold3-mlx 0.3.1
| parameter | value |
|---|---|
| model | openfold3-mlx 0.3.1 |
| weights | — |
| hardware | — |
| mlx version | — |
| python | — |
| random seed | — |
| msa strategy | — |
| diffusion samples | 1 |
| runtime | 556s |
| predicted by | mlx@peptide |
| predicted at | 2026-05-03 |
▸citationbibtex
@peptide{pep10903,
sequence = {HAEGTFTSDVSSYLEGQAAKEFIAWLVKGRGGPX},
target = {gipr},
author = {peptidemodel},
year = {2026},
status = {computed}
}