Relaxin-2 Peptide Research: RXFP1 Receptor Signaling and Antifibrotic Mechanisms

Relaxin-2 peptide research centres on a two-chain, insulin-superfamily hormone that signals through RXFP1, a G protein-coupled receptor with an unusual multi-domain architecture. Preclinical work has repeatedly shown that relaxin-2 suppresses TGF-β1-driven myofibroblast differentiation and upregulates matrix metalloproteinases through an ERK1/2 and nitric oxide-cGMP dependent pathway. The translational record is more complicated: a promising phase 2/3 signal in acute heart failure did not survive confirmation in a trial roughly five times larger.

Relaxin-2 is one of the more scientifically interesting peptides in circulation precisely because its story does not resolve neatly. The receptor biology is genuinely novel, the antifibrotic preclinical data are robust and reproducible across tissue types, and the large-scale clinical outcome data are null. For researchers building a working model of peptide pharmacology, that combination is more instructive than another compound with uniformly positive literature.

What Is Relaxin-2 and How Does It Differ From Other Peptides?

Human relaxin-2, often written H2 relaxin, belongs to the insulin/relaxin superfamily rather than to any of the more familiar neuropeptide or incretin families. Structurally it is not a single linear chain. It comprises an A chain and a B chain held together by two interchain disulfide bonds with a third intrachain disulfide inside the A chain, giving it an insulin-like fold. This matters for anyone handling the material analytically, because a two-chain disulfide-linked peptide has failure modes that a linear peptide does not, including chain scrambling and mismatched disulfide isoforms that co-elute closely with the correct species on reversed-phase methods.

The receptor-binding surface sits primarily on the B chain, where an Arg-X-X-X-Arg-X-X-Ile motif provides the critical contacts. Serelaxin, the compound used in the clinical programme discussed below, is recombinant human relaxin-2 and is structurally identical to the endogenous hormone rather than being an engineered analogue. That distinction is important when interpreting the trial results, because failure of serelaxin is a failure of the native ligand under a specific exposure regimen, not a failure of a modified molecule with altered receptor kinetics.

How Does the RXFP1 Receptor Work?

RXFP1 is where relaxin-2 research becomes structurally distinctive. Most peptide-binding GPCRs studied in this space, including the APJ receptor covered in our apelin-13 receptor research overview and the Mas receptor discussed in our angiotensin 1-7 mechanism article, present a relatively conventional orthosteric pocket. RXFP1 does not.

The receptor carries a large ectodomain built from an N-terminal low-density lipoprotein receptor class A module, referred to as the LDLa module, tethered to a leucine-rich repeat domain by a linker of roughly 32 residues. Relaxin-2 binds the leucine-rich repeats, but binding alone is not sufficient to activate signalling. The LDLa module is dispensable for ligand binding yet essential for activation, which places RXFP1 in a small category of receptors where the recognition event and the activation event are structurally separable.

The Autoinhibition Mechanism

Cryo-electron microscopy work on active-state human RXFP1 bound to a single-chain relaxin-2 construct and heterotrimeric Gs, deposited in the Protein Data Bank as structure 7TMW and published in Nature Chemical Biology in 2023, resolved a mechanism that had been inferred but not visualised. The receptor’s own extracellular loop 2 occupies the orthosteric site in the active state and functions as a tethered agonist. In the absence of relaxin-2, the ectodomain holds this arrangement in an autoinhibited configuration.

Relaxin-2 binding stabilises a helical conformation of the LDLa-to-LRR linker, which repositions residues from both the linker and the LDLa module so they can engage the transmembrane domain and relieve the autoinhibition. The practical implication for pharmacology is significant. A ligand that merely occupies the leucine-rich repeats will not necessarily activate the receptor, and screening assays built only on binding affinity will misclassify compounds. This is a recurring theme in peptide receptor work and a reason functional assays remain necessary alongside binding data.

What Are the Antifibrotic Mechanisms of Relaxin-2?

The antifibrotic mechanism is the most reproducible part of the relaxin-2 literature. Relaxin signals through RXFP1 to phosphorylated ERK1/2 and then through a neuronal nitric oxide synthase, nitric oxide, soluble guanylyl cyclase, cGMP dependent pathway. The downstream consequence is disruption of the TGF-β1 and Smad2 axis that drives fibroblast to myofibroblast transition and collagen deposition.

Chow and colleagues, publishing in PLOS One in 2012, quantified the matrix metalloproteinase response directly. In TGF-β1-stimulated human dermal fibroblasts, H2 relaxin increased MMP-1 by approximately 50 percent, MMP-2 by approximately 80 percent, and MMP-9 by approximately 80 percent. In rat renal myofibroblasts the same treatment increased MMP-13 by approximately 90 percent, MMP-2 by approximately 130 percent, and MMP-9 by approximately 115 percent. Critically, the investigators demonstrated pathway dependence rather than mere correlation: the MMP upregulation was significantly blocked by selective inhibitors of nNOS, iNOS, and guanylyl cyclase.

That the effect reproduces across species and across two unrelated tissue types, dermal and renal, is what gives the mechanism credibility. The peptide is not simply suppressing collagen synthesis. It is shifting the balance of extracellular matrix turnover toward degradation, which is a mechanistically different proposition from the actin-cytoskeletal and cell-migration effects seen with tissue repair peptides such as those described in our TB-500 actin dynamics research summary.

Key Research Findings

• RXFP1 activation requires the LDLa module even though relaxin-2 binds the leucine-rich repeat domain, meaning binding affinity and functional potency are structurally decoupled at this receptor.
• Cryo-EM structure 7TMW resolved extracellular loop 2 acting as a tethered agonist occupying the orthosteric site, with the ectodomain enforcing autoinhibition in the unliganded state.
• H2 relaxin increased MMP-1 by roughly 50 percent, MMP-2 by roughly 80 percent, and MMP-9 by roughly 80 percent in TGF-β1-stimulated human dermal fibroblasts (Chow et al., PLOS One, 2012).
• In rat renal myofibroblasts the same study reported increases of roughly 90 percent for MMP-13, 130 percent for MMP-2, and 115 percent for MMP-9, with the effect abolished by nNOS, iNOS, and guanylyl cyclase inhibitors.
• RELAX-AHF (Teerlink et al., The Lancet, 2013) randomised 1,161 subjects, 581 to serelaxin and 580 to placebo, and met one of two co-primary endpoints with a dyspnoea VAS AUC difference of 448 mm x h (95% CI 120 to 775, p=0.007) while failing the Likert co-primary endpoint (26% placebo versus 27% serelaxin, p=0.70).
• RELAX-AHF-2 (Metra et al., New England Journal of Medicine, 2019) randomised 6,545 subjects and found 180-day cardiovascular death of 8.7% with serelaxin versus 8.9% with placebo (HR 0.98, 95% CI 0.83 to 1.15, p=0.77).

What Did the Large Clinical Studies Show?

The relaxin-2 clinical programme is a case study in why early efficacy signals require confirmation at scale. RELAX-AHF, reported by Teerlink and colleagues in The Lancet in 2013, enrolled 1,161 subjects hospitalised with acute heart failure, assigning 581 to serelaxin and 580 to placebo. The trial had two co-primary endpoints and split them. The visual analogue scale area under the curve for dyspnoea through day 5 favoured serelaxin by 448 mm x h with a 95 percent confidence interval of 120 to 775 and a p value of 0.007. The Likert scale co-primary endpoint showed essentially nothing, with 150 placebo subjects (26 percent) and 156 serelaxin subjects (27 percent) responding, p equal to 0.70. Secondary endpoints covering cardiovascular death, heart failure readmission, and renal failure through day 60 were not significant.

What generated enthusiasm was an exploratory observation of a roughly 37 percent reduction in all-cause and cardiovascular death through day 180. A mortality signal of that magnitude, appearing in a trial whose primary endpoints were symptomatic, was biologically surprising given that the intervention was delivered as a short intravenous infusion over 48 hours.

Why the Confirmatory Trial Diverged

RELAX-AHF-2, published by Metra and colleagues in the New England Journal of Medicine in August 2019, was built to test that mortality signal properly. It randomised 6,545 subjects, 3,274 to serelaxin and 3,271 to placebo, with a mean age of 73.1 years and 39.6 percent women. The result was flatly null. Cardiovascular death at 180 days occurred in 8.7 percent of the serelaxin group and 8.9 percent of the placebo group, giving a hazard ratio of 0.98 with a 95 percent confidence interval of 0.83 to 1.15 and a p value of 0.77. Worsening heart failure at day 5, the other primary endpoint, occurred in 6.9 percent versus 7.7 percent, hazard ratio 0.89, 95 percent confidence interval 0.75 to 1.07, p equal to 0.19. Adverse event rates were comparable between groups. Serelaxin did produce a significantly greater blood pressure reduction than placebo, confirming that the compound was pharmacologically active and that the null result was not an exposure failure at the level of basic hemodynamics.

Several interpretations remain live in the literature. The most mechanistically coherent is a mismatch between exposure duration and biological timescale. Antifibrotic remodelling through MMP-mediated matrix turnover is a process measured in weeks, while the trial delivered a 48-hour exposure window and then measured outcomes at 180 days. A second interpretation is that the original mortality signal was a chance finding in an exploratory analysis, which is the statistically parsimonious reading. The two are not mutually exclusive.

Biased Allosteric Agonism at RXFP1

The structural work explains why the field moved toward allosteric approaches. ML290 is a small molecule that acts as a biased allosteric agonist at RXFP1, binding outside the relaxin-2 recognition site and producing a signalling profile that differs from the native peptide. Because RXFP1 activation depends on relieving an ectodomain-enforced autoinhibition rather than on simple orthosteric occupancy, allosteric ligands can in principle engage the transmembrane domain directly and bypass the LDLa requirement altogether. For researchers, the practical value of biased ligands is as pharmacological tools that separate signalling branches, allowing the ERK-dependent and cGMP-dependent arms to be interrogated independently rather than as a bundled response.

Analytical Considerations for Two-Chain Peptides

Relaxin-2 illustrates why purity documentation for disulfide-linked, two-chain peptides deserves closer reading than for simple linear sequences. Reversed-phase HPLC purity, the figure most commonly reported on a certificate of analysis, resolves many process-related impurities but can under-resolve disulfide isomers and scrambled-chain species that share the same mass and similar hydrophobicity. Orthogonal identity confirmation by mass spectrometry establishes molecular weight but does not by itself distinguish correctly folded material from a mis-paired isomer of identical mass.

For this class of peptide the informative additions are non-reducing versus reducing analysis to confirm interchain linkage, and where warranted a secondary structure method to confirm the expected fold. Researchers evaluating supplier documentation for any multi-chain or cyclised peptide should read the certificate for what method was actually run rather than for the headline purity number alone. Our approach to independent batch verification is described on our certificates of analysis page, and the broader research catalogue is listed under research peptides.

Where Relaxin-2 Research Stands

The honest summary is that RXFP1 remains a well-validated antifibrotic target with an unusually well-characterised activation mechanism, while relaxin-2 itself has not demonstrated outcome benefit in a large acute heart failure population under short-infusion conditions. Those two statements are compatible. Current preclinical interest has shifted toward chronic fibrotic models in renal, hepatic, and pulmonary tissue, where the exposure timescale matches the biology, and toward engineered RXFP1 agonists with longer half-lives than the native two-chain peptide, which clears rapidly.

For anyone building a literature model of this target, the RELAX-AHF to RELAX-AHF-2 sequence is worth studying on its own terms. A 1,161-subject trial produced an exploratory mortality signal that a 6,545-subject trial could not reproduce. That is the expected behaviour of a well-run research programme, not an anomaly, and it is a useful calibration against over-reading early-phase peptide data in any compound class.

For research purposes only. Not for human consumption. Not for diagnostic or therapeutic use. All data referenced in this article are drawn from published peer-reviewed literature and are presented for scientific education. Maple Research Labs supplies materials for laboratory research use only.