Amylin analogues explained: amylin receptors and the peptides that target them

The biology of amylin analogues is older than most people assume: amylin was isolated in 1987, its receptors were worked out between 1998 and 2022, and the first analogue reached the US market in 2005. What changed recently is chemistry. Fatty-acid lipidation turned a hormone that lasts minutes into molecules that last a week or more, and that brought a wave of long-acting candidates into phase 2 and phase 3 trials.

This guide explains amylin biology, the three amylin receptors and why “selective” and “non-selective” matter, then compares the analogues by structure, receptor profile, half-life and published results. Each investigational molecule has its own overview: cagrilintide, eloralintide and amycretin. Nothing here is medical or usage guidance.

What is amylin?

Amylin, also called islet amyloid polypeptide (IAPP), is “a 37-amino-acid peptide” made by pancreatic beta cells [1]. It was first purified in 1987 as the major protein component of islet amyloid from the pancreases of three people with type 2 diabetes. The researchers named it diabetes-associated peptide, calculated its mass at 3905 Da, and noted that it was 46% identical to calcitonin gene-related peptide (CGRP) [2].

UniProt records human amylin as residues 34 to 70 of an 89-residue precursor, with a disulfide bond between the second and seventh residues of the mature chain and a C-terminal tyrosine amide [3]. The mature sequence is KCNTATCATQRLANFLVHSSNNFGAILSSTNVGSNTY-NH2 [3].

What amylin does

Amylin “is colocated with insulin in secretory granules and cosecreted with insulin by pancreatic beta cells in response to food intake” [4]. From nonclinical studies, the pramlintide label lists three actions [4]:

• it slows gastric emptying without altering overall nutrient absorption;
• it suppresses glucagon secretion, which reduces glucose output from the liver;
• it regulates food intake through centrally mediated effects on appetite.

A 2015 review describes amylin’s major role as glucoregulatory, notes that it regulates energy metabolism, and states that it “acts principally in the circumventricular organs of the central nervous system” [1]. In people with type 1 and type 2 diabetes, secretion of both insulin and amylin after meals is reduced [4].

Why amylin itself is not a drug

Human amylin has a “high propensity toward the formation of amyloid fibrils, which makes it a challenging drug design effort” [5]. That property, the same one that made it identifiable in diabetic islet amyloid [2], makes it hard to formulate. It is also short-lived. Every analogue below is an answer to one or both problems.

Amylin receptors: AMY1, AMY2 and AMY3

Amylin receptors are unusual: they are built from two proteins. The story starts with receptor activity-modifying proteins (RAMPs). In 1998, McLatchie and colleagues showed that the calcitonin-receptor-like receptor (CRLR, now CLR) could act as either a CGRP receptor or an adrenomedullin receptor depending on which RAMP was co-expressed, and that RAMPs were required to transport it to the cell surface [6].

In 1999, Christopoulos and colleagues showed that RAMPs do the same for the calcitonin receptor (CTR). Co-expressing RAMP1 or RAMP3 with the human calcitonin receptor induced specific amylin binding and produced two distinct amylin receptor phenotypes; RAMP2 did not cause a significant increase in amylin binding in that system [7]. Current nomenclature, as used in the 2022 cryo-EM work, describes amylin receptors as “heterodimers of the calcitonin (CT) receptor (CTR) and one of three receptor activity-modifying proteins (RAMPs), AMY1R, AMY2R, and AMY3R” [8].

Receptor	Core receptor	Partner	Main ligands	Ref
CTR	Calcitonin receptor	None	Calcitonin	[8]
AMY1R	Calcitonin receptor	RAMP1	Amylin; also high affinity for salmon calcitonin and CGRP	[7] [8]
AMY2R	Calcitonin receptor	RAMP2	Amylin	[8]
AMY3R	Calcitonin receptor	RAMP3	Amylin	[7] [8]
CGRP receptor	CLR	RAMP1	CGRP	[6]
AM1 / AM2 receptors	CLR	RAMP2 / RAMP3	Adrenomedullin	[6] [7]

The 2022 cryo-EM structures showed that amylin bound all three amylin receptors in a similar conformation, constrained by the RAMP and by an ordered mid-peptide “bypass motif”, while calcitonin-based peptides bound amylin receptors in a distinct way that overlapped their binding to CTR [8]. The authors concluded that activation by calcitonin-based and amylin-based peptides is distinct, with “important implications for the development of AMYR therapeutics” [8].

Selective versus non-selective agonists

Because every amylin receptor contains the calcitonin receptor, a drug can activate amylin receptors only, or amylin receptors plus the calcitonin receptor on its own. The eloralintide developers frame it this way: “amylin receptor agonists can be selective or non-selective, depending on their affinities for amylin and calcitonin receptors” [9]. Agonists that activate both are often called DACRAs, dual amylin and calcitonin receptor agonists [10]. Whether selectivity changes efficacy or tolerability in people has not been tested in a head-to-head trial; the rationale so far comes from cell and rodent studies [9].

Pramlintide: the first amylin analogue

Pramlintide is the only amylin analogue with FDA approval. Its original application, NDA 021332, was approved on 16 March 2005 [11]. It is indicated as an adjunct for people with type 1 or type 2 diabetes who use mealtime insulin and have not reached glucose targets despite optimal insulin therapy [4].

Structurally it is a minimal change. Pramlintide is a “synthetic 37-amino acid polypeptide, which differs in amino acid sequence from human amylin by replacement with proline at positions 25 (alanine), 28 (serine), and 29 (serine)” [4]. It keeps the Cys2–Cys7 disulfide and the C-terminal tyrosine amide [12]. The label gives a molecular weight of 3949.4 and formula C171H267N51O53S2 for the peptide [4].

Its pharmacokinetics show why the field moved on. Absolute subcutaneous bioavailability is about 30% to 40%, about 40% of the drug is unbound in plasma, it does not extensively bind albumin, and its half-life in healthy individuals is approximately 48 minutes [4]. The cagrilintide developers note that pramlintide “requires three daily injections due to its short half-life” [5]. The label carries a boxed warning: use with insulin increases the risk of severe hypoglycaemia, particularly in type 1 diabetes [4].

The long-acting amylin analogues

The current generation keeps pramlintide’s anti-aggregation logic and adds a fatty di-acid that binds albumin, the same half-life strategy used in semaglutide and tirzepatide. A 2026 review lists cagrilintide, eloralintide, petrelintide and three earlier-stage candidates (MET-233i, ABBV-295 and AZD6234) among amylin-based therapies in human development for obesity, plus combination approaches [13]. Only the molecules with peer-reviewed structural and clinical data are compared below.

Cagrilintide (non-selective, weekly)

Cagrilintide is “the stable, lipidated long-acting amylin analogue” its developers selected for obesity [5]. It is a 37-residue peptide (PubChem: C194H312N54O59S2, about 4409 g/mol, monoisotopic mass about 4406.25 Da) [14]. Its sequence, KCNTATCATQRLAEFLRHSSNNFGPILPPTNVGSNTP-NH2 with a Cys2–Cys7 disulfide and an N-terminal γGlu-linked C20 fatty di-acid [15], differs from human amylin at positions 14, 17, 25, 28, 29 and 37 [3] [15], as detailed in the cagrilintide overview.

Two independent cryo-EM studies characterised its receptor binding. One found a similar “bypass” binding mode at AMY1R and CTR, with Phe23 anchoring the peptide, a Glu14–Arg17 intramolecular salt bridge stabilising the helix, and non-selective Gs activation through both receptors [10]. The other determined structures with all three amylin receptors and CTR and found an amylin-like binding mode but distinct conformational dynamics at the receptors [16]. In a phase 1b study its half-life was 159–195 hours [17]. In a 26-week phase 2 trial in 706 people on cagrilintide, mean weight reductions were 6.0% to 10.8% versus 3.0% with placebo [18].

Petrelintide (human amylin analogue, weekly)

Petrelintide is described by its developers as “a potent, stable, long-acting analogue of human amylin that can be formulated at about neutral pH”, achieved by optimising chemical stability while retaining potency, with a view to co-formulation with other peptides [19]. Its full sequence is reported in the discovery paper; it is not reproduced here.

In single and multiple ascending dose phase 1 trials, petrelintide was slowly absorbed, had a half-life of about 10 days and showed steady-state dose proportionality; body weight fell by up to 8.6% after 16 weeks [20]. Nausea occurred in 16.7% to 33.3% of participants on petrelintide versus 16.7% on placebo in the second multiple-dose part [20]. The phase 2 ZUPREME 1 trial is summarised in the trial table below [21].

Eloralintide (selective, weekly)

Eloralintide is a 37-residue C-terminally amidated amylin analogue in which the native disulfide is replaced by a methylene thioacetal bridge “which enhances chemical stability”, and Lys26 carries a C20 fatty di-acid through a linker for once-weekly administration [9]. In cAMP assays it was about 12-fold more potent at human AMY1R (EC50 23.9 pM) than at human CTR (291.0 pM) and about 11-fold more potent than at AMY3R (253.8 pM) [9]. Its terminal half-life in a single-dose phase 1 study was 310–366 hours [9]. The developers contrast it with cagrilintide, which “is a non-selective amylin receptor agonist” [9].

Amycretin (GLP-1 and amylin receptor agonist in one molecule)

Amycretin, now named zenagamtide, is a different design: “a unimolecular entity, with GLP-1R and AMYR agonist moieties connected by a linker consisting of four glycine residues and one glutamic acid residue”, acylated with a C18 di-acid side chain on the GLP-1 part [22]. In cAMP assays its mean EC50 values were 1.30 pM at AMY1(a)R, 13.3 pM at AMY2(a)R, 17.2 pM at AMY3(a)R and 1.89 pM at the calcitonin receptor CTR(a) [22]. It is therefore non-selective within the amylin family and adds GLP-1 receptor agonism. Its 68-residue structure is set out in the amycretin overview.

Amylin analogues compared

Feature	Pramlintide	Cagrilintide	Petrelintide	Eloralintide	Amycretin
Length	37 [4]	37 [14]	Human amylin analogue [19]	37 [9]	68 [22]
Ring	Cys2–Cys7 disulfide [12]	Cys2–Cys7 disulfide [15]	Not summarised here	Methylene thioacetal [9]	Amylin-type domain [22]
Half-life extension	None [4]	N-terminal C20 di-acid [5]	Long-acting design [19]	C20 di-acid on Lys26 [9]	C18 di-acid on GLP-1 domain [22]
Receptor profile	Amylin analogue [4]	Non-selective (AMYR and CTR) [10]	Not summarised here	AMY1R-selective [9]	Non-selective plus GLP-1R [22]
Human half-life	About 48 min [4]	159–195 h [17]	About 10 days [20]	310–366 h [9]	Not reported in human abstracts
Status (October 2026)	FDA-approved 2005 [11]	Phase 3, mainly with semaglutide [23]	Phase 2 completed [21]	Phase 3	Phase 3

Phase 3 status for eloralintide and amycretin is documented, with trial registrations, in the respective overviews.

Key clinical findings

Dose levels are omitted; ranges span the active arms. The trials differ in duration, population and estimand, and none compared these analogues with each other.

Molecule	Trial	Phase / n	Duration	Weight change vs placebo	Most common adverse event	Ref
Cagrilintide	Dose-finding	2; 706 on cagrilintide, 99 liraglutide, 101 placebo	26 weeks	−6.0% to −10.8% vs −3.0%	Gastrointestinal events and administration-site reactions	[18]
Cagrilintide with semaglutide	REDEFINE 1	3a; 3,417	68 weeks	−20.4% vs −3.0%	Gastrointestinal events (79.6% vs 39.9%)	[23]
Petrelintide	ZUPREME 1	2; 485 dosed	28 weeks (primary), 42 weeks total	−7.9% to −9.8% vs −1.7%; up to −10.7% by week 42	Nausea (20% vs 6%)	[21]
Eloralintide	Phase 2	2; 263	48 weeks	−9% to −20% vs −0.4%	Nausea (11% to 64% vs 14%) and fatigue (0% to 46% vs 12%)	[24]
Amycretin	Phase 1b/2a	1b/2a; 125	20 to 36 weeks	Up to −24.3% vs −1.1% (week 36, highest-dose part)	Gastrointestinal, mostly mild to moderate	[25]

Cagrilintide

In the phase 2 dose-finding trial, the highest cagrilintide arm also exceeded daily liraglutide (10.8% versus 9.0%, estimated treatment difference 1.8%, p = 0.03) [18]. In REDEFINE 1, the cagrilintide-semaglutide combination produced −20.4% mean weight change at 68 weeks versus −3.0% with placebo (difference −17.3 percentage points) [23]. Most late-stage cagrilintide data concern this combination rather than cagrilintide alone.

Petrelintide

ZUPREME 1 randomised 493 adults without type 2 diabetes 5:1 to petrelintide or placebo at 32 sites in Poland, Romania and the USA; 485 received at least one dose [21]. Mean weight change at week 28 (efficacy estimand) ranged from −7.9% to −9.8% across five arms versus −1.7% with placebo, with estimated treatment differences of −6.2 to −8.1 percentage points [21]. Vomiting was infrequent (3% versus 6% with placebo), and diarrhoea and constipation rates were similar to placebo [21]. The investigators reported no deaths [21].

Eloralintide

The 48-week phase 2 trial enrolled 263 adults without type 2 diabetes at 46 US centres; 78% were female and mean BMI was 39.1 [24]. Mean weight change ranged from −9% to −20% across six regimens versus −0.4% with placebo [24]. Fatigue was reported in up to 46% of participants in one arm versus 12% with placebo [24].

Amycretin

The subcutaneous phase 1b/2a study was a single-site trial of 125 participants with several escalation schemes, so its headline −24.3% at week 36 comes from a small group [25]. A 2026 network meta-analysis of six trials (4,642 participants) ranked high-dose subcutaneous amycretin first for placebo-adjusted weight change (−23.95%) and high-dose eloralintide second (−18.01%), but its authors call the data “sparse, low-certainty” and the findings “preliminary” [26].

Safety findings across the class

Gastrointestinal events dominate. Nausea was the most common adverse event with petrelintide (20% versus 6% with placebo) [21] and with eloralintide (11% to 64% versus 14%) [24]. With cagrilintide, gastrointestinal disorders and administration-site reactions were the most frequent events in phase 2 [18]. The 2026 network meta-analysis found gastrointestinal events, nausea, vomiting and constipation more common with high-dose amylin-based therapies, especially oral amycretin and the cagrilintide-semaglutide combination [26]. Pramlintide’s specific risk is severe hypoglycaemia when combined with insulin, the subject of its boxed warning [4]. Long-term safety for the long-acting analogues will come from the phase 3 programmes.

Amylin analogues in laboratory research

• Receptor panels. Selectivity is measured by cAMP or binding assays in cells expressing CTR alone or CTR with each RAMP, with CGRP and adrenomedullin receptors as counter-screens [7] [9] [22]. Published panels state the CTR variant used [7] and, for lipidated peptides, repeat assays with albumin present [9].
• Structural biology. Cryo-EM of Gs-coupled amylin receptor complexes is now established for amylin, calcitonins and cagrilintide [8] [10] [16].
• Aggregation testing. Because amylin-derived sequences form fibrils [5], stability work typically includes aggregation-sensitive methods alongside reversed-phase HPLC; petrelintide’s developers used a thioflavin assay at neutral pH [19].
• Identity. LC-MS intact mass and peptide mapping confirm the lipidation site and the ring (disulfide versus thioacetal) for each analogue.

Synthesis, impurity and storage considerations for these lipidated peptides are covered in SPPS vs LPPS, peptide API impurities and specifications and lyophilised peptide stability and storage. For the incretin side of combination research, see GLP-1 agonists explained.

Development history

Year	Milestone	Ref
1987	Amylin purified from islet amyloid	[2]
1998	RAMPs discovered as partners that set CLR ligand specificity	[6]
1999	RAMP1 and RAMP3 shown to create amylin receptors from the calcitonin receptor	[7]
2005	Pramlintide approved by FDA	[11]
2021	Cagrilintide (Novo Nordisk) discovery paper and phase 2 results	[5] [18]
2022	Cryo-EM structures explain amylin receptor phenotype	[8]
2025	REDEFINE 1, eloralintide (Eli Lilly) phase 2, petrelintide (Zealand Pharma) discovery paper, amycretin (Novo Nordisk) phase 1b/2a published	[23] [24] [19] [25]
2026	Petrelintide phase 1 and ZUPREME 1 phase 2 published	[20] [21]

Developers are named from the cited papers’ affiliations and funding statements. The 1987 work came from the Department of Biochemistry, University of Oxford [2].

How Winhycare Pharma supplies amylin analogues

Winhycare Pharma supplies cagrilintide as research-use-only material for laboratories and, separately, as API to licensed pharmaceutical manufacturers, and supplies amycretin and eloralintide as research-use-only material. All are supplied at ≥99% purity, as stated in our brochure, with a certificate of analysis for each batch, and custom synthesis of amylin analogues and fragments is available. Product details are on the cagrilintide, amycretin and eloralintide pages; enquiries go through the cagrilintide API form or the research peptide form.

For laboratory research use only. Not for human or veterinary use.