What Is Apelin? The Heart Peptide With Two Ligands and No Drug

Peptide library

Apelin is a short cardiovascular peptide that dilates arteries, makes the heart contract harder, and opposes both angiotensin II and vasopressin. It has been infused into humans repeatedly and it does what the biology predicts. It has also, after nearly thirty years, produced no approved drug — and the two industry programmes that got furthest both stopped. That gap between clean physiology and empty pharmacy is the most useful thing about apelin, and it is worth understanding why.

What apelin is

Apelin was isolated in 1998 from bovine stomach extracts by Tatemoto and colleagues, who were hunting for whatever activated a receptor called APJ. APJ had been cloned five years earlier by O’Dowd’s group purely because it resembled the angiotensin II type 1 receptor, and it had sat unmatched ever since. Apelin was the answer, and the paper named it for that fact: the APJ endogenous ligand.

The human gene, APLN, sits at Xq26.1 and encodes a 77-residue precursor (UniProt Q9ULZ1). Cleavage produces a nested family rather than one product — apelin-36 (residues 42–77), apelin-31, apelin-28 and apelin-13 (residues 65–77, sequence QRPRLSHKGPMPF) all share the same C-terminus. Apelin-17 is the other commonly studied form. The C-terminal end is the business end; the N-terminal extensions tune how the peptide behaves rather than whether it works at all.

The correction most summaries need

Three things get garbled routinely. First, APLN is on the X chromosome; it is APLNR, the receptor, that sits on chromosome 11. The famous 1993 paper’s title says “chromosome 11” and it is about the receptor.

Second, APJ does not bind angiotensin. The IUPHAR review states it plainly: despite sharing roughly 40–50% identity with AT1 across the transmembrane domains, “the apelin receptor does not bind angiotensin.” The apelin–angiotensin antagonism is real but it happens one level up. Apelin binding drives APJ and AT1 into a heterodimer that forces AT1 into a low-affinity state — and critically, that only happens when apelin is present, not when angiotensin II is. Apelin is a brake on the renin-angiotensin system that works by rearranging its receptor, not by competing for it.

Third, the circulating peptide is not plain apelin-13. The N-terminal glutamine of apelin-13 spontaneously cyclises to pyroglutamate, and [Pyr¹]apelin-13 is the dominant species found in human plasma — in one study apelin-36, apelin-17 and unmodified apelin-13 were not detected at all. Pyroglutamation is endogenous chemistry, not a stabilising modification someone added in a lab.

ELABELA: the second ligand nobody expected

For fifteen years apelin was assumed to be APJ’s only endogenous ligand. Then in 2013 and 2014 two groups independently found another, working on zebrafish embryos. Chng and colleagues called it ELABELA and reported that losing it produced embryos “with a rudimentary heart or no heart at all, surprisingly phenocopying the loss of the apelin receptor.” Pauli and colleagues, arriving at the same molecule, called it Toddler and described it as a motogen — “neither an attractant nor a repellent” but a global driver of cell movement.

ELABELA is a genuinely separate hormone: gene APELA at 4q32.3, a 54-residue precursor (UniProt P0DMC3), a 32-residue mature peptide, and essentially no sequence homology to apelin. It was also found in a stretch of genome that had been annotated as non-coding. In 2019 IUPHAR formally recommended recognising it as a second endogenous ligand for the receptor. Anyone describing ELABELA as “another name for apelin” or “an apelin fragment” has it wrong.

What apelin does in people

The human physiology is unusually well characterised for a peptide with no product. In healthy volunteers, apelin-36 and [Pyr¹]apelin-13 produced reproducible vasodilation in forearm resistance vessels, and the effect was attenuated by nitric oxide blockade but unaffected by aspirin — so it is NO-dependent and not prostaglandin-mediated. Notably, neither isoform changed dorsal hand vein diameter, which means the common claim that apelin is a venodilator is contradicted in humans by the study most often cited for it.

In a later study spanning heart failure patients, angiography patients and healthy volunteers, an intracoronary bolus of apelin-36 increased coronary blood flow and the maximum rate of rise in left ventricular pressure while reducing filling pressures, and systemic [Pyr¹]apelin-13 raised cardiac index while lowering mean arterial pressure and peripheral vascular resistance. A six-hour infusion sustained the increase in cardiac index in chronic heart failure. Inotropy without an increase in afterload is exactly the profile heart failure pharmacology has been chasing for decades.

On fluid balance, apelin runs opposite to vasopressin. Water deprivation lowers plasma apelin while causing it to accumulate in the hypothalamus, and apelin inhibits vasopressin neuron activity and vasopressin release. Together with its effect on AT1, the system reads as a general counterweight to the pressor and antidiuretic arms — closer in spirit to the natriuretic peptides than to anything in the renin cascade.

Why it is a hard target

Two problems, and the second is worse than the first.

It disappears in minutes. ACE2 — the same carboxypeptidase that clips angiotensin II — cleaves apelin-13 with a catalytic efficiency of 2.1 × 10⁶ M⁻¹s⁻¹, marginally higher than its efficiency on angiotensin II itself. It removes a single C-terminal residue, converting [Pyr¹]apelin-13 to [Pyr¹]apelin-13(1-12) and apelin-17 to apelin-16. Apelin-17’s measured plasma half-life is 4.6 minutes. That is a formulation problem of the kind half-life extension chemistry exists to solve, and engineered analogues have reached beyond 24 hours in plasma.

Worth resisting the neat story that ACE2 is therefore the villain. The ACE2 product retains meaningful activity, and the authors who characterised it concluded that raised ACE2 activity in cardiovascular disease “should not significantly compromise the beneficial effects of apelin based therapies.” ACE2 is also not the most complete inactivator: neprilysin degrades apelin peptides into fragments that no longer bind the receptor at all, and plasma kallikrein clips the N-terminal three residues off apelin-17.

The isoforms signal differently. This is the deeper issue. Measured at the human receptor, apelin-17 and [Pyr¹]apelin-13 differ about four-fold in potency for inhibiting cAMP but roughly seventy-fold in β-arrestin recruitment. That asymmetry means “an apelin receptor agonist” is an incomplete specification — the biased agonism question has to be answered deliberately. A purpose-built biased agonist, MM07, showed a bias of roughly 350- to 1300-fold toward the G protein pathway and produced double the peak forearm dilatation of [Pyr¹]apelin-13 in volunteers.

The clinical record, stated plainly

A registry search returns 68 studies mentioning apelin, but the great majority measure it as a biomarker rather than administering it. The studies that actually gave an apelin-pathway agent to people are the ones in the figure above: a cluster of academic infusion studies enrolling between 9 and 63 subjects, plus two industry programmes.

Amgen’s AMG 986, an oral small-molecule APJ agonist, ran a 182-subject phase 1 in healthy volunteers and heart failure patients from 2016 to 2019 and was terminated — the registry entry records the reason as a sponsor decision and states explicitly that it was “not terminated due to a safety reason.” BioAge’s azelaprag reached a 204-participant phase 2 combining it with tirzepatide in older adults with obesity; dosing was discontinued after “observation of liver transaminitis without clinically significant symptoms in some subjects receiving azelaprag.”

As of August 2026, searches of DailyMed and the FDA drug databases return no records for any apelin-pathway drug, and UniProt lists exactly one DrugBank entry against this receptor — azelaprag, investigational. Nothing here should be read as a global regulatory statement; other agencies were not searched. But the honest summary is that this pathway has never produced a marketed medicine, and the sample sizes behind everything we know about it in humans are small, which is common for peptide programmes and a reason to hold conclusions loosely.

Frequently asked questions

Is apelin the same as angiotensin?

No, and their receptors do not share ligands. APJ was discovered because it looks like the angiotensin AT1 receptor in its transmembrane regions, but it does not bind angiotensin. Their interaction is at the level of receptor heterodimerisation, driven by apelin.

Which apelin isoform is the “real” one?

It depends on the question. [Pyr¹]apelin-13 is the dominant form detected in human plasma. Apelin-17 is the most potent at the receptor in binding and in both signalling assays. They are not interchangeable, especially on the β-arrestin arm.

Does ACE2 destroy apelin?

ACE2 removes one C-terminal residue, and the resulting fragment is still biologically active — it has been detected endogenously in human heart and lung. Complete inactivation comes from other proteases, notably neprilysin.

Is there an approved apelin drug?

US labelling and approval databases returned no records as of August 2026, and the only two clinical-stage APJ agonists identifiable in ClinicalTrials.gov both had their lead studies terminated. Claims that apelin is a proven cardiovascular or longevity therapy are not supported by the registry.

Informational only — not medical advice · 21+