B7-33

What it is

Relaxin-2 is an endogenous hormone known from pregnancy: it widens blood vessels, increases kidney blood flow and remodels connective tissue. As a drug it is impractical, because it consists of 53 amino acids in two chains with three disulfide bridges.

B7-33 shortens this. It consists of only part of the B-chain: six residues at the start have been removed, and the cysteines at positions 11 and 23 replaced by serine so that the peptide does not clump. The result is linear, positively charged and water-soluble, whereas the unmodified B-chain is insoluble and non-functional. Not to be confused with serelaxin: that is the complete, recombinant relaxin-2 with large human studies behind it.

How it is supposed to work

The receptor RXFP1 triggers several signaling pathways, including cAMP and the kinases ERK1/2. B7-33 uses the same contact points as the hormone: if the key residues at positions B13, B17 and B20 are exchanged, it completely loses binding and activity.

The actual finding is the split. In cells that artificially carry the receptor in large numbers, B7-33 is weak: pKi 5.54 versus 8.96 for relaxin-2, cAMP potency pEC50 5.12 versus 10.49. In fibroblasts that naturally carry RXFP1, the picture reverses — there B7-33 activates ERK1/2 like the hormone and raises the collagen-degrading enzyme MMP-2 to the same extent. This makes it the first functionally selective agonist of this receptor: it engages the antifibrotic arm and leaves aside the cAMP arm, to which unwanted effects of relaxin-2 are linked.

What was measured in the heart in animal experiments

Three independent groups have taken measurements in the heart, and the direction is consistent. In the rat model after myocardial infarction, treatment began only eight weeks after the damage; afterwards, less collagen was detectable in the left ventricle and the end-diastolic pressure was lower. In the isoprenaline model in mice, B7-33 lowered interstitial collagen staining by about 51 percent, as much as relaxin-2 with about 57 percent, with unchanged blood pressure.

The second group switched models: 30 minutes of occlusion of a coronary artery in mice, then reopening. After 24 hours, the infarct size was smaller at 21.99 versus 45.32 percent and pump function was better. The most revealing comparison comes from 2023: in isoprenaline cardiomyopathy in mice, B7-33 and relaxin lowered fibrosis equally, while the ACE inhibitor perindopril lowered blood pressure and inflammation but not fibrosis.

Fibrosis outside the heart

The effect is not specific to the heart, and that fits the mechanism, because fibroblasts are found everywhere. In the ovalbumin model in mice, two weeks of intranasal administration normalized the thickened epithelial layer of the airways and the total collagen of the lungs; airway hyperresponsiveness, by contrast, improved only partially. In the kidney, there was less fibrosis after experimental ureteral ligation. A completely different application: if an implant is coated with a polymer that releases B7-33, the connective tissue capsule is 49.2 percent thinner after six weeks.

Why serelaxin is part of the picture

Anyone who wants to assess B7-33 has to talk about serelaxin. In RELAX-AHF with 1,161 patients with acute heart failure, one of two primary endpoints was met: shortness of breath improved by 448 mm × h on the visual analog scale compared with placebo. A secondary endpoint was striking, with 65 deaths under placebo versus 42 under serelaxin by day 180.

This figure carried the class until the confirmatory trial. RELAX-AHF-2 enrolled 6,545 patients and missed both primary endpoints: cardiovascular death at day 180 in 8.7 versus 8.9 percent, worsening heart failure at day 5 in 6.9 versus 7.7 percent. Two things follow from this. The receptor can be targeted in humans: in 71 patients, serelaxin lowered pulmonary capillary wedge pressure by 2.44 mmHg more than placebo. In acute heart failure, this changed nothing. The antifibrotic track that B7-33 targets was not tested there: it is open, not refuted.

What stands between peptide and use

The hardest practical limit is the residence time: in laboratory testing, the half-life of B7-33 in serum is about 6 minutes; with fatty acid conjugation it could be increased to 60 minutes. That is exactly where the field is working, with lipidated variants and nanoparticles as carriers. Accordingly, the animal studies use daily administration. The class is also being developed further with other molecules: long-acting relaxin analogs have since been studied in humans, and an RXFP1 agonist has made it into a completed phase IIb trial. B7-33 itself never took this step.

What the studies show

Hossain et al., Chem Sci 2016 — the paper that brought B7-33 into existence

Design, synthesis and characterization. In cells with overexpressed human RXFP1, B7-33 bound weakly (pKi 5.54 versus 8.96) and stimulated cAMP weakly (pEC50 5.12 versus 10.49); in fibroblasts with natural receptor expression, it activated ERK1/2 like the hormone. In addition, three animal models with 7 to 15 animals per group: rat heart after infarction, mouse heart under isoprenaline, mouse lung in the ovalbumin model.

Devarakonda et al., J Am Heart Assoc 2020 — ischemia and reopening

Adult male CD1 mice, 30 minutes of occlusion of the left coronary artery, followed by 24 hours or 7 days of reperfusion. Infarct size 21.99 versus 45.32 percent under vehicle, fractional shortening 29 versus 23 percent after 24 hours. In isolated heart muscle cells, better survival, ERK1/2-dependent.

Alam et al., Biomed Pharmacother 2023 — B7-33 against an ACE inhibitor

Male mice with isoprenaline cardiomyopathy, treatment from day 7 to 14. B7-33 and relaxin lowered fibrosis of the left ventricle equally, normalized inflammation and cardiomyocyte hypertrophy, and restored vessel density and aortic contractility. Perindopril lowered blood pressure and inflammation, but not fibrosis and hypertrophy. The endpoint is tissue, not survival.

Metra et al., N Engl J Med 2019 — RELAX-AHF-2 as a warning

Not B7-33, but the complete hormone: 6,545 patients with acute heart failure, 48-hour infusion against placebo. Cardiovascular death at day 180 in 8.7 versus 8.9 percent, worsening heart failure at day 5 in 6.9 versus 7.7 percent: both primary endpoints missed.

Where the data stop

For B7-33 itself, no study in humans exists. A query of the US trial registry returns zero entries, and of the twelve papers that mention the peptide in the abstract, none has human participants. There is no pharmacokinetics, no toxicology, no case reports.

The receptor finding is also incomplete, and the developer group says so openly: poor affinity and potency at the overexpressed receptor, full effect only in fibroblasts. That the design pattern does not work on its own is shown by an independent study on short-chain analogs of the related relaxin-1 B-chain: no affinity for RXFP1, no relaxin-like effect. The longest published treatment duration is four weeks in a mouse model. Everything beyond fibrosis in animal tissue — exercise capacity, symptoms, disease course, life expectancy — has not been measured.

Status, approval and legal

B7-33 is not an approved medicine or a legally marketable dietary supplement in Germany, the EU or the US; it is in preclinical research and is sold as a research peptide not intended for human use. The parent molecule has no approval either: the EMA refused Reasanz, with the active ingredient serelaxin, on May 23, 2014, because, among other things, the studies showed no benefit for short-term relief of shortness of breath; the FDA issued a Complete Response Letter in 2014. In sport, B7-33 is prohibited at all times: relaxin and its analogs are not named in the WADA Prohibited List 2026, so section S0 applies for substances without approval by a governmental health authority.

Safety

Nothing is known about the safety of B7-33 in humans: no phase 1 study, no pharmacovigilance, no case reports. Caution regarding blood pressure can be inferred, because RXFP1 activation widens blood vessels; in the large studies, serelaxin was used only at a systolic blood pressure of at least 125 mmHg. There are no quality data on commercial products; the purity of 97 percent in the original paper is a laboratory figure for study material. There are no data for pregnancy and breastfeeding, children, cancer, or kidney and liver failure, and interactions have never been studied.

Sources

• Hossain et al., Chem Sci 2016 — design of B7-33, functional selectivity at RXFP1, three animal models
• Devarakonda et al., J Am Heart Assoc 2020 — B7-33 in ischemia-reperfusion in mice
• Alam et al., Biomed Pharmacother 2023 — B7-33 versus perindopril in a cardiomyopathy model
• Marshall et al., Eur J Pharmacol 2017 — vascular effects of B7-33 compared with serelaxin
• Handley et al., Int J Mol Sci 2023 — affinity and potency of B7-33, state of further development
• Praveen et al., Int J Mol Sci 2023 — half-life of B7-33 in serum and lipidation
• Welch et al., ACS Appl Mater Interfaces 2019 — B7-33-releasing coating and capsule formation
• Bhuiyan et al., FASEB J 2021 — kidney fibrosis under relaxin, B7-33 and perindopril
• Metra et al., N Engl J Med 2019 — RELAX-AHF-2, serelaxin in acute heart failure
• EMA — questions and answers on the refusal of the marketing authorization for Reasanz (serelaxin) on May 23, 2014