Angiotensin-(1-7)

An endogenous seven-amino-acid peptide of the renin-angiotensin system that activates the MAS receptor and counteracts angiotensin II, developed clinically as TXA127 and tested for two decades across oncology, COVID-19, and muscle-wasting indications without an approval anywhere or a replicated efficacy result.

Also referenced as: Ang-(1-7), Angiotensin 1-7, TXA127, TXA-127, Talfirastide

Also appears in: Tissue repair · Neuroprotection

COAPublic product evidenceSearch the public certificate ledger for this compoundNo exact compound records are currently indexed under this profile name.→

---

What is Angiotensin-(1-7)?

Angiotensin-(1-7) is a genuine endogenous peptide — a heptapeptide with the sequence Asp-Arg-Val-Tyr-Ile-His-Pro, molecular formula C41H62N12O11 and a molecular weight of about 899 Da. It is produced mostly when the enzyme ACE2 clips the C-terminal phenylalanine off angiotensin II, and it forms what is usually called the counter-regulatory arm of the renin-angiotensin system: where angiotensin II acting at the AT1 receptor is vasoconstrictive, pro-fibrotic, and pro-inflammatory, Ang-(1-7) acting at the MAS receptor generally pushes the other way.

It has a real pharmaceutical identity behind it. A synthetic formulation has been carried through human trials since the mid-2000s under the code TXA127 — first by US Biotest and Tarix Pharmaceuticals, then by the Tarix Orphan subsidiary, later by Constant Therapeutics — and the molecule carries the drug name talfirastide. That clinical pedigree, rather than any approved product, is most of why it appears in research-peptide catalogs.

Its visibility jumped in 2020 for a specific reason: ACE2, the enzyme that makes Ang-(1-7), is also the cell-entry receptor for SARS-CoV-2. The hypothesis that COVID-19 disrupted the balance between angiotensin II and Ang-(1-7) drove a wave of trials, and much of the attention around the compound still dates from that period — including the trials that went on to fail.

How it works

• MAS receptor agonism. The receptor encoded by the Mas proto-oncogene was identified as a functional Ang-(1-7) receptor when genetic deletion of Mas in mice abolished Ang-(1-7) binding in the kidney, eliminated its antidiuretic action after an acute water load, and removed the relaxation response in isolated aorta (Santos et al., Proceedings of the National Academy of Sciences, 2003;100(14):8258–8263). This is the anchor citation for essentially every downstream claim about the compound.
• Formation by ACE2, degradation by ACE. ACE2 is a carboxypeptidase that removes only the C-terminal residue of its substrates, and when screened against a panel of 126 biological peptides it processed angiotensin II with roughly 400-fold higher catalytic efficiency than angiotensin I (Vickers et al., Journal of Biological Chemistry, 2002;277(17):14838–14843). The reverse limb matters just as much: ACE itself degrades Ang-(1-7), so the peptide sits between two enzymes and its endogenous levels shift with ACE inhibitor therapy (Santos et al., Physiological Reviews, 2018;98(1):505–553).
• Endothelial nitric oxide release via Akt. In Mas-transfected CHO cells and in human aortic endothelial cells, Ang-(1-7) drove reciprocal phosphorylation of eNOS at Ser1177/Thr495, sustained Akt phosphorylation, and roughly doubled nitric oxide release; the effect was blocked by the MAS antagonist A-779 and suppressed by the PI3K inhibitor wortmannin (Sampaio et al., Hypertension, 2007;49(1):185–192). This is the mechanistic basis of the vasodilatory and endothelial-function claims.
• TGF-β/Smad inhibition and anti-fibrotic action in muscle. In mdx mice, the standard model for Duchenne muscular dystrophy, infused or orally administered Ang-(1-7) normalized skeletal muscle architecture, decreased local fibrosis, and improved muscle function, acting through inhibition of TGF-β Smad signaling with an accompanying drop in the pro-fibrotic microRNA miR-21; mdx mice given A-779 or lacking the MAS receptor deteriorated instead (Acuña et al., Human Molecular Genetics, 2014;23(5):1237–1249). Note carefully what this is — preservation of damaged muscle by suppressing fibrosis, not anabolic growth signaling.

Research status

Angiotensin-(1-7) has an unusually long human trial record, spanning roughly two decades and several disease areas. It is also a record with very few positive results and a great deal of attrition.

Hematology and oncology (2005–2016). The earliest human work treated the peptide as a hematopoietic agent for chemotherapy-induced cytopenias. A Phase I/II dose escalation in newly diagnosed breast cancer gave 15 patients five different dose levels with 5 patients receiving filgrastim as a comparator; no dose-limiting toxicity was observed, and the authors suggested a 100 µg/kg daily dose might attenuate multilineage cytopenias (Rodgers et al., Cancer Chemotherapy and Pharmacology, 2006;57(5):559–568). A separate Phase I in advanced solid tumors reframed the compound as an antiangiogenic hormone: 18 patients received subcutaneous injections once daily for 5 days on a 3-week cycle, dose-limiting toxicities at 700 µg/kg were a grade 4 stroke and a grade 3 reversible cranial neuropathy, one patient had a 19% tumor reduction, and the recommended Phase II dose was set at 400 µg/kg (Petty et al., Clinical Cancer Research, 2009;15(23):7398–7404). The follow-on Phase II in metastatic sarcoma dosed 20 mg daily subcutaneously and explicitly failed to confirm the placental growth factor biomarker effect that had motivated it; median progression-free survival was 2.7 months and median overall survival 10.2 months, with prolonged stabilization in two patients with vascular sarcomas (Savage et al., Sarcoma, 2016;2016:4592768).

The one controlled trial that reported positive endpoints. A randomized, double-blind, placebo-controlled Phase 2b enrolled 34 patients with ovarian, Fallopian tube, or peritoneal carcinoma and randomized them to 100 µg/kg, 300 µg/kg, or placebo alongside gemcitabine and platinum chemotherapy. It is the closest thing in the entire record to a positive controlled result, and it needs reading precisely. The primary endpoint — reduction in Grade 3–4 thrombocytopenia — did not reach significance (no Grade 4 events at 100 µg/kg versus 6% of chemotherapy cycles on placebo, p = 0.07), while secondary pharmacodynamic endpoints did: maximal percentage increase in platelet concentration from baseline (p = 0.02) and reduced nadir absolute neutrophil count (p = 0.04). The 300 µg/kg dose differed from placebo on nothing. The authors concluded the findings were consistent with stimulation of thrombogenesis in the marrow (Pham et al., Cancer Chemotherapy and Pharmacology, 2013;71(4):965–972; NCT00771810, results posted). No confirmatory trial ever followed it.

The commercial program stalled. Alongside those published studies, the TXA127 registry record from this era is largely abandoned work: trials in graft-versus-host disease prevention (NCT01882374, NCT01882387) and in pediatric transplant patients (NCT01554254) were withdrawn with zero patients enrolled; a Phase I in myelodysplastic syndrome (NCT01362036) was terminated after 2 patients for enrollment feasibility; a cord-blood engraftment study (NCT01300611) was terminated; a platelet-recovery study after autologous transplant (NCT01121120) was terminated after enrolling 75; and an early study in HIV-infected patients (NCT00757250) was terminated at 13 for recruitment difficulty. Two company studies from this period did complete — the ovarian Phase 2b above and a Phase 1 comparing TXA127 with filgrastim for peripheral blood stem cell mobilization (NCT01543971, 18 patients, no results posted).

COVID-19 — the decisive negative result. The best-powered test of the compound is the ACTIV-4 Host Tissue platform, which ran two parallel randomized trials against a shared placebo in adults hospitalized with COVID-19 and new-onset hypoxemia across 35 US sites. The TXA-127 arm enrolled 343 patients (170 drug, 173 placebo) on 0.5 mg/kg intravenously once daily for 5 days. It met prespecified early stopping criteria for a low probability of efficacy. Oxygen-free days at day 28 showed no difference (adjusted OR 0.88, 95% CrI 0.59–1.30), and 28-day all-cause mortality was 13.5% versus 13.3%. The authors’ stated conclusion is that the results do not support the hypothesis that increasing angiotensin (1-7) improves outcomes in severe COVID-19 (Self et al., JAMA, 2023;329(14):1170–1182; NCT04924660).

The two smaller COVID trials are weaker evidence than they are usually made to look. A single-center randomized, placebo-controlled, double-blinded pilot at Columbia randomized 22 patients (20 analyzed) to TXA-127 0.5 mg/kg intravenously daily for up to 10 days. It found no between-group difference in intubation, length of stay, mortality, or acute kidney injury — but it was designed as a safety and proof-of-concept study, its authors describe it as clearly underpowered to detect a difference in outcome, and it was stopped early once local admissions fell and the NIH multicenter trial was about to begin. Its stated conclusion is that TXA-127 was safe to administer, not that it was ineffective (Wagener et al., Critical Care, 2022;26(1):229; NCT04401423). A seamless Phase 1–2 trial infusing Ang-(1-7) at 10 µg/kg/day in ICU patients — run at two hospitals in Belo Horizonte, Brazil, under a Belgian sponsor — enrolled 28 patients in Phase 1 and 79 in Phase 2 before being stopped prematurely for a low recruitment rate. In the randomized Phase 2 intention-to-treat analysis, oxygen-free days did not differ between groups; a difference appeared only when the open-label Phase 1 patients were pooled in (19 versus 14 days, p = 0.04) — a pooled analysis, not the prespecified randomized comparison (Martins et al., Annals of Intensive Care, 2024;14(1):139; NCT04633772). A planned Stanford trial (NCT04570501) was withdrawn without enrolling anyone, and several further COVID-19 registrations — including the ATCO trial (NCT04332666) and NCT04605887 — still carry a registry status of “unknown” with no posted results, so the registered COVID record is larger than the published one.

Muscle wasting and current programs. The muscular dystrophy rationale rests on rodent work — the mdx studies above, plus the finding that ACE2 is augmented in dystrophic skeletal muscle and plays a role in decreasing associated fibrosis (Riquelme et al., PLoS One, 2014;9(4):e93449). On that preclinical basis TXA127 collected FDA Fast Track designation for Duchenne muscular dystrophy in October 2015, plus orphan designations spanning DMD, limb-girdle muscular dystrophy, laminin-deficient congenital muscular dystrophy (MDC1A), and recessive dystrophic epidermolysis bullosa. No completed human efficacy trial underlies any of them, and the multi-site Phase 2 in DMD patients announced alongside the Fast Track award does not appear in the trial registry. The two active company programs are small and their public records have gone quiet: an open-label Phase 2 with a planned 10 non-ambulant patients with DMD-associated cardiomyopathy (NCT06013839) and a Phase 2 with a planned 50 post-ischemic stroke patients (NCT06135103) both carry a registry status of “unknown,” were last updated in January 2024 and December 2023 respectively, and have no results posted — so even their actual enrollment is unverifiable. An investigator-initiated Phase 1/2 in moderate to severe traumatic brain injury (NCT06282965, 90 participants planned, University of Arizona) is recruiting.

Investigator-led studies have fared poorly too. Trials of Ang-(1-7) in essential hypertension (NCT02245230, terminated at 24 when enrollment and funding both ran short), obesity-related hypertension (NCT03604289, terminated at 8 when funding ended), peripheral arterial disease (NCT03240068, terminated at 6 when funding ended), cognitive function after coronary bypass surgery (NCT03252093, terminated at 6), cognitive function in heart failure (NCT03159988, suspended at 6 for slow enrollment), and primary autonomic failure (NCT02591173, terminated at 7) all closed early and well short of their targets. A handful of early-phase physiology studies remain active: cardiovascular effects in aging (NCT05301192), metabolic effects (NCT02646475), and obesity hypertension (NCT06482853) are recruiting, while an obesity energy-expenditure study (NCT03777215) is active but no longer recruiting.

No regulatory authority has approved angiotensin-(1-7) for any indication.

Common dosage forms

• Lyophilized powder in vials for reconstitution, the standard research-market presentation. The subcutaneous route this implies matches the oncology trials, which used daily subcutaneous injection.
• Sterile solution for intravenous infusion, the format used in every COVID-19 trial — 0.5 mg/kg daily in both ACTIV-4 and the Columbia pilot, 10 µg/kg/day continuous infusion in the Brazilian ICU study. This is a clinical-manufacturing format rather than a research-catalog one.
• Oral and inhaled preparations appear in academic formulation work — cyclodextrin-based oral delivery in rodent studies, and spray-dried powders developed for respiratory delivery. These are formulation research, not established products.
• Topical and cosmetic listings, where an angiotensin-family heptapeptide appears under the INCI name sh-Heptapeptide-13. Cosmetic ingredient databases describe that name inconsistently — several call it a recombinant equivalent of “angiotensin I,” which is a ten-residue peptide, while listing a seven-residue composition — so it should not be assumed to be identical material.
• Multi-ingredient blends, in which Ang-(1-7) is combined with other compounds and its individual content is often not stated separately.

This section describes formats only and is not dosing guidance.

Key considerations

• Designations are not approvals. Fast Track and orphan drug designations for TXA127 are frequently cited as if they were evidence of efficacy. They are administrative incentives awarded on the strength of preclinical rationale and unmet need, and in this case none of them followed a positive human efficacy trial. The compound has no marketing approval anywhere in the world.
• The largest and best-controlled human test was negative, and said so explicitly. ACTIV-4 Host Tissue randomized 343 patients to TXA-127 or placebo, stopped early for low probability of efficacy, and concluded against the underlying hypothesis. Any description that cites “Phase 2 COVID-19 trials” without that outcome is reporting how much activity there was, not what it showed. The honest counterweight is also small: one randomized placebo-controlled Phase 2b in ovarian cancer reported significant secondary pharmacodynamic endpoints at 100 µg/kg, but missed its primary endpoint and was never replicated.
• Underpowered is not the same as negative. The Columbia COVID pilot is routinely listed as a second failed trial. Its authors call it clearly underpowered and conclude only that the drug was safe. Counting it as efficacy evidence overstates the negative case in the same way that counting the pooled Brazilian analysis overstates the positive one.
• Documented safety signals exist, and the pharmacokinetics are short. The Phase I dose-limiting toxicities were a grade 4 stroke and a grade 3 reversible cranial neuropathy at 700 µg/kg (Petty et al., 2009), and in the open-label phase of the ICU trial one serious adverse event — bradycardia — was judged possibly related to the infusion and justified its discontinuation (Martins et al., 2024). Twenty-eight-day toxicology at 10 mg/kg/day subcutaneously found no detectable toxicity in rats or beagle dogs, with a plasma half-life of only 20–30 minutes in both species (Mordwinkin et al., Journal of Pharmaceutical Sciences, 2012;101(1):373–380). The record is neither alarming nor clean by default, and the short half-life means circulating levels do not persist between doses.
• The muscle rationale is rodent and anti-fibrotic, not anabolic. Every muscle claim traces back to mdx mice, where the mechanism is suppression of TGF-β-driven fibrosis in already-damaged tissue. The one registered human muscular dystrophy trial targets DMD-associated cardiomyopathy in a planned 10 non-ambulant patients — a cardiac endpoint in advanced disease, not skeletal muscle building in healthy people. Nothing in the human record supports a muscle-growth reading.
• The naming space around this compound is genuinely treacherous. Angiotensin II is a different molecule with the opposite hemodynamic effect and is an approved drug (a vasopressor for septic or other distributive shock) — confusing the two inverts the pharmacology. “TXA” is also the standard clinical abbreviation for tranexamic acid, so string searches on that code return a mix of two unrelated drugs. Aclerastide (DSC127, NorLeu3-Ang-(1-7)) is an analog, not the native peptide — a topical gel whose three Phase 3 diabetic foot ulcer trials were all terminated (NCT01830348, NCT01849965, NCT01840085). TRV-027, tested in the companion arm of the same JAMA report, is an AT1 receptor biased ligand and is also not Ang-(1-7). A citation should always be checked against the actual molecule.