B7-33: an anti-scarring relaxin fragment that has only been tested in rodents

B7-33 reduces fibrosis in the heart and lungs of rats and mice about as well as the full relaxin-2 hormone it was modelled on, and in one head-to-head test it matched the full hormone for reducing heart scarring from a molecule far cheaper to manufacture [1][6]. The same protective effect has held across heart-attack, drug-induced heart damage, lung, vessel and implant models. What the compound does not carry is a person: no human has received B7-33 in a published study, so there is no known human dose, side-effect profile or benefit.

What it is

Relaxin-2 is a hormone best known from pregnancy, when it helps loosen and remodel connective tissue and widens blood vessels. Outside pregnancy, researchers have spent two decades studying it as a possible treatment for fibrosis: the stiff, collagen-heavy scarring that follows organ injury in the heart, kidney, lung, liver and skin. A 2017 review from the lead laboratory summarised the appeal. In experimental work relaxin blocked fibroblasts from overproducing collagen when they were driven by pro-scarring signals, but not at rest, and it boosted the enzymes that break down excess collagen [2].

The full hormone is awkward to use as a drug. It has two chains held together by three disulfide bridges, which makes it hard and expensive to manufacture. In 2016 a team at the Florey Institute and Monash University reported B7-33, a linear peptide based on just the B-chain [1]. It binds relaxin's receptor, RXFP1, and preferentially switches on one signalling route (ERK phosphorylation) over another (cAMP). The authors argued the cAMP arm was likely behind tumour-promoting effects seen with full relaxin, and showed that, unlike the hormone, B7-33 didn't promote prostate tumour growth in mice [1].

Treat all of that as mechanism theory: it comes from receptor assays, cultured cells and rodents.

The human evidence

For B7-33 itself, no human study of any design exists: we searched PubMed and found no trial, no case series and no pharmacokinetic study.

What does exist is human data on the parent hormone, and it's sobering context.

• Scleroderma. A double-blind trial gave recombinant human relaxin (10 or 25 μg/kg/day by continuous infusion under the skin) or placebo for 24 weeks to people with diffuse systemic sclerosis, a disease defined by skin and organ scarring [3]. Skin thickness scores didn't improve compared with placebo, lung function fell in the relaxin groups, and seven people who had received relaxin had serious kidney events, mostly after stopping it, versus none on placebo [3].
• Acute heart failure. RELAX-AHF-2 randomised 6,545 patients to a 48-hour infusion of serelaxin (recombinant relaxin-2) or placebo. Cardiovascular death at 180 days was 8.7% versus 8.9% (hazard ratio 0.98, p = 0.77), and worsening heart failure at day 5 didn't differ either [4].

Neither trial tested B7-33, and neither was designed around long-term anti-fibrotic treatment in the way B7-33's developers imagine. Still, when the full hormone didn't reverse scarring in a scarring disease, a fragment of it needs its own human evidence before anyone assumes it will.

What the animal and lab work suggests

The animal record is small and consistent, and it comes mostly from a closely linked group of Australian laboratories.

The original paper, 2016

In three rodent models of heart or lung disease, B7-33 prevented or reversed organ fibrosis and dysfunction with potency similar to the full hormone [1]. In a rat heart-attack model it improved heart function over 12 weeks about as well as relaxin, and in a mouse model of drug-induced heart damage it reduced left-ventricle scarring by day 14 [1].

Heart attack in mice, 2020

A group in Virginia, collaborating with the peptide's developers, tested B7-33 after a 30-minute blockage of a coronary artery in mice.

In cultured mouse heart cells, B7-33 reduced a marker of endoplasmic-reticulum stress and improved survival after simulated injury [5]. The treatment timing, around the moment of injury in a young healthy mouse, is quite unlike a person arriving at hospital hours into a heart attack with years of coronary disease behind it.

Against a standard drug, 2023

In mice given isoprenaline to damage the heart, treatment from day 7 to day 14 with B7-33, full relaxin or the ACE inhibitor perindopril was compared [6]. B7-33 and relaxin reduced left-ventricle fibrosis and heart-muscle thickening equally; perindopril lowered blood pressure and inflammation but didn't reduce fibrosis within that one-week window [6]. That doesn't show B7-33 beats ACE inhibitors in people. ACE inhibitors are judged on years of outcomes, not one week of collagen staining in mice.

Other rodent work

In rats, a single injection of B7-33 improved one kind of blood-vessel relaxation in gut arteries three hours later, matching full relaxin; in mouse arteries in a dish, it prevented damage to the vessel lining caused by placental culture fluid [7]. Plastic implants coated to slowly release B7-33 developed a scar capsule 49.2% thinner after six weeks under the skin of mice [8]. And a 2025 study attached B7-33 to iron-oxide nanoparticles given by mouth to mice, reporting less heart fibrosis than perindopril over four weeks [10].

How to read the rodent results

It helps to look at what these models actually are. Isoprenaline is a heart-stimulating drug; in the 2023 study mice received 25 mg/kg a day under the skin until the heart muscle was injured and began to scar, and treatment started a week later, in otherwise healthy adult animals [6]. The heart-attack model ties off a coronary artery for exactly 30 minutes and then releases it, under anaesthesia, with the drug given on a fixed schedule [5]. Both are good for asking whether a molecule touches the scarring process at all. Neither resembles the way most human fibrosis develops: slowly, over years, in older people with high blood pressure, diabetes, kidney disease or repeated small injuries, and usually on several other medicines.

Both mouse heart studies described above used only male animals [5,6], while the developers' own review notes that relaxin's anti-fibrotic effects are influenced by sex [2]. And scar measurements in these studies are mostly collagen stains on tissue slices at one time point. They show less collagen, which is the right direction, but not whether an animal lives longer or functions better over its lifetime.

None of that makes the findings wrong. It means they answer an early question (does this peptide reduce collagen build-up in a rodent under controlled injury?) and leave the questions a patient would ask untouched.

The practical problem: it disappears fast

A 2023 paper from the developers measured how long B7-33 survives in serum in the lab: its half-life was about six minutes [9]. Attaching a fatty-acid chain stretched that to around 60 minutes. The nanoparticle work above is another attempt to keep it active for longer [10]. These are signs of a research programme working on delivery, which is sensible. They also mean plain B7-33, as a free peptide, is unlikely to stay active in the body for long, and nobody has measured what happens to it in a human.

Safety and side effects

Unknown in people. No human safety data exist. Rodent studies weren't designed to catalogue side effects, and short treatment windows of one to six weeks in young animals can't reveal slow harms. The 2016 finding that B7-33 didn't feed prostate tumours in mice is reassuring but narrow [1].

Two warnings from the parent hormone are worth keeping in mind, with the caveat that B7-33 may behave differently. In scleroderma, stopping relaxin was followed by serious kidney problems and high blood pressure in seven treated patients [3]. And relaxin is a vasoactive hormone; the developers' own rat data show B7-33 also acts on blood vessels [7]. Anyone with heart, kidney or blood-pressure problems would be in the group most likely to be affected, and also the group most likely to be drawn to an anti-fibrotic.

Legal and regulatory status

B7-33 isn't approved as a medicine anywhere. It has never entered a registered clinical trial that we could find, so it has no UK or US regulatory status beyond being an unapproved research chemical.

What is still open

• Whether B7-33 does anything in people, at any dose.
• How long it lasts in human blood, and whether a practical dosing schedule exists.
• Whether rodent anti-fibrotic effects over one to twelve weeks translate to human fibrosis that builds over years.
• Whether it avoids the kidney and blood-pressure problems seen after stopping full relaxin.
• Whether any laboratory outside the developers' network can reproduce the core findings.

What this means

B7-33's scientific case is genuine, and it rests on a claim that is easy to state and hard to prove: that switching on relaxin's receptor through ERK signalling while mostly avoiding the cAMP route keeps the anti-scarring effect and drops the tumour-promoting one. In mice the developers showed exactly that, and the anti-fibrotic results have since held across heart-attack, drug-induced heart damage, lung, vessel and implant models [1][5][6][7][8]. That mechanism is worth testing properly, and it is the reason B7-33 is a more interesting molecule than a generic relaxin fragment.

But a fragment does not inherit its parent's promise without also inheriting its questions. The full hormone went into two large, well-conducted human trials — 6,545 patients in acute heart failure and a 24-week study in diffuse scleroderma — and in neither did it reverse the scarring the field was hoping for; in the scleroderma study, stopping treatment was followed by kidney and blood-pressure problems in seven patients [3][4]. B7-33 was designed to be selective, not inert. It still works through RXFP1, and it still acts on blood vessels. The two things that make it attractive — a cleaner signalling profile and a simpler molecule — have not been shown to change any of that in a person.

Our position is that B7-33 is a research compound, not a product, and it should be treated as one until a human study changes the picture. If you meet it sold as a relaxin mimetic for recovery or scarring, the useful questions are not about dose — nobody knows one — but about what is actually in the vial. Demand independently verified identity, purity, sterility and endotoxin and heavy-metal testing, tied to a certificate of analysis for the batch you receive. A peptide with a six-minute serum half-life and no human safety data behind it fails the first of those checks more often than sellers admit.

The direction we would back is narrow and specific: not another rodent fibrosis model, but a formulation that keeps the peptide active in human blood long enough to matter, followed by ordinary first-in-human dosing and safety work in healthy volunteers. The developers' own lipidation and nanoparticle papers read as an attempt to solve exactly that problem [9][10]. Until it is solved, the anti-fibrotic question cannot be asked in people at all, and every claim built on the mouse data remains a claim about mice.

Educational content only — not medical advice. Many peptides discussed on HPR are not approved for human use. Talk to a qualified clinician before making any decision about your health.