pe
pep-10905 v1 CC-BY-SA-4.0

Petrelintide: weekly weight-loss injection (NN9213/ZP8396)

An experimental weekly injection for weight loss that works through the body's amylin system, a different pathway from Ozempic-type drugs; developed by Zealand Pharma and partnered with Roche in 2025. Experimental, not yet an approved drug.

statuscomputed targetCALCR length38 aa refs2
status 2 / 5
prediction metrics boltz-2 2.2.1
ipTM0.831
pTM0.777
avg pLDDT81.9
ranking score0.821
STRUCTURE · PEP-10905 × CALCR
ranking0.821
target interface 4.5Å peptide drag rotate · ctrl+scroll zoom · right-click pan
boltz-2 2.2.1 · mmCIF ↓ download
sequence38 aa
1510152025303538
KCNTATCATQRLANFLVHS SNNFGPILSSTNVGSNTYX
in the news 15 articles
overview readme

What this is

Petrelintide (also called ZP8396) is an investigational weekly injection for weight management that works through the body's amylin system — a pathway completely separate from the GLP-1 drugs (such as semaglutide and tirzepatide) that currently dominate the obesity market. It was developed by the Danish biotech Zealand Pharma and partnered with Roche in March 2025 in a deal valued at up to $5.3 billion, the largest single-asset pharma partnership of that year.

Unlike native amylin and earlier analogs such as pramlintide, petrelintide is heavily engineered — acylated and sequence-modified — to be chemically stable at neutral pH, which prevents the amyloid fibrillation that hampered earlier amylin drugs and allows co-formulation with other peptides. This stability, and a half-life of approximately 33.8 hours, is what makes once-weekly dosing possible. It is structurally distinct from cagrilintide, Zealand Pharma's earlier amylin analog being developed separately in combination with semaglutide under a Novo Nordisk partnership.


History

Petrelintide grew out of Zealand Pharma's amylin medicinal-chemistry program, which had previously produced cagrilintide. The key engineering challenge was reformulating an amylin analog to remain stable at neutral pH — avoiding the fibrillation that limited native amylin and required the acidic formulations of pramlintide — while extending half-life enough to support once-weekly dosing. The discovery and optimization chemistry is described in a 2025 Journal of Medicinal Chemistry publication (PMID 41217931).

Following Phase 1 single-ascending-dose (SAD) and Phase 1b multiple-ascending-dose (MAD) studies, Zealand reported topline data from the Phase 2b ZUPREME-1 trial (493 participants, 42 weeks) in March 2026, showing up to 10.7% mean weight loss versus 1.7% on placebo, with GI tolerability described as similar to placebo. The March 2025 Roche partnership reshaped the asset's development path: a Phase 3 monotherapy program is expected to begin in H2 2026, alongside a Phase 2 combination study pairing petrelintide with Roche's dual GLP-1/GIP agonist CT-388.


What it does

Petrelintide mimics amylin — a hormone the pancreas releases with insulin after meals — to tell the brain that enough food has been consumed. In clinical studies, this translated to meaningful weight loss over 42 weeks, with a side-effect profile that looked much closer to placebo than to GLP-1 receptor agonists. Withdrawal rates in the ZUPREME-1 trial were notably lower on the drug (8.4%) than on placebo (13.6%), suggesting that participants tolerated it well enough to stay on it.

In preclinical diet-induced obesity (DIO) rat studies, petrelintide produced preferential fat-mass loss while preserving relative lean mass compared with a liraglutide comparator — a potential differentiator from the GLP-1 class — though this has not yet been confirmed in published human body-composition data.


Evidence

  • Human: Phase 2b ZUPREME-1 (493 adults with overweight or obesity, 42 weeks) reported topline results in March 2026: up to 10.7% mean weight loss at the 9 mg dose versus 1.7% on placebo, with all five dose arms meeting the primary endpoint versus placebo at week 28. Diarrhea and constipation rates were described as single-digit and similar to placebo; nausea was mostly mild and largely resolved after participants reached the maintenance dose; 98% of top-cohort participants reached that dose. Earlier Phase 1b MAD work in 48 adults over 16 weeks showed dose-dependent weight loss of 4.8–8.6%, and Phase 1 SAD studies in 56 adults showed an early weight signal of up to 4.2% at one week sustained to six weeks. A Phase 2b trial in adults with type 2 diabetes (ZUPREME-2) is ongoing with topline results expected H2 2026. The ZUPREME-1 full trial publication has not yet appeared in the available literature; the evidence base rests on topline reporting.
  • Animal: In DIO rats, petrelintide produced preferential fat-mass loss with relative lean-mass preservation compared with liraglutide. The discovery and optimization chemistry, including preclinical receptor pharmacology, is described in the 2025 Journal of Medicinal Chemistry paper (PMID 41217931).
  • In vitro / mechanistic: Amylin and calcitonin receptor biology is well-established through decades of pramlintide and amylin research (reviewed by Hay and colleagues and others; see Madsbad and colleagues, Expert Opinion on Investigational Drugs, 2025 for pipeline context). Petrelintide's specific receptor pharmacology and neutral-pH engineering are characterized in the J Med Chem 2025 development paper.

Known effects

  • Weight loss — Phase 2b (topline): up to 10.7% at 42 weeks in adults with overweight or obesity
  • GI tolerability advantage vs placebo comparison — Phase 2b: diarrhea/constipation similar to placebo; nausea mostly mild and resolving at maintenance dose
  • Low trial discontinuation — Phase 2b: 8.4% withdrawal on drug vs 13.6% on placebo
  • Lean-mass preservation (vs liraglutide) — Preclinical only (DIO rat model); not confirmed in human data
  • Satiety / gastric emptying delay — Mechanistic (AMYR/CTR agonism); class-level evidence from pramlintide program
  • Postprandial glucagon suppression — Mechanistic (class-level); not individually extracted from petrelintide-specific human data
  • Leptin sensitivity restoration — Proposed preclinical mechanism; not established in human data

Safety signals

The safety profile reported in ZUPREME-1 (topline, 42 weeks) is notably favorable relative to GLP-1 receptor agonists: nausea was mostly mild and largely resolved after reaching the maintenance dose, and diarrhea and constipation rates were single-digit and similar to placebo. Injection-site reactions were described as mild and transient. The 98% maintenance-dose attainment rate in the top-dose cohort and the lower trial withdrawal rate on drug than on placebo (8.4% vs 13.6%) reflect this tolerability profile.

Important gaps remain. Long-term safety beyond 42 weeks has not been characterized; Phase 3 will be the first dataset for multi-year exposure. No cardiovascular outcomes trial has been conducted, so MACE effects are unestablished. Chronic calcitonin receptor agonism raises a theoretical question about bone metabolism that has not been studied in humans. No specific petrelintide pancreatitis signal was reported in available sources, but this has not been formally excluded in a long-term study.

Class-level considerations from pramlintide and the broader amylin literature include a potential for hypoglycemia risk when combined with insulin or insulin secretagogues, and slowed gastric emptying that can alter the absorption rate and peak concentration of orally co-administered medications with narrow therapeutic windows. These are class-level and precautionary inferences, not established petrelintide-specific label findings, given that petrelintide remains investigational.


Regulatory status

  • US (FDA): Not approved. Investigational drug; legally administered only to participants in Zealand- or Roche-sponsored clinical trials. Not available through compounding pharmacies, telehealth platforms, or research-chemical suppliers.
  • International: Not approved by any regulatory authority. No commercial access exists; trial sites operate under national investigational frameworks.
  • WADA: Not individually listed by name on the current Prohibited List. The S0 category (substances not approved by any government regulatory health authority) and S2 (peptide hormones, growth factors, related substances and mimetics) categories are relevant. Athletes subject to WADA or equivalent anti-doping codes should treat petrelintide as prohibited.

FDA approval is not anticipated before approximately 2029–2030, pending Phase 3 completion and regulatory review.


Mechanism

Petrelintide is a 36-amino-acid acylated human amylin analog engineered for chemical stability at neutral pH. The acylation and sequence modifications are not represented in the stored raw sequence. Native amylin is a 37-residue peptide co-secreted by pancreatic beta cells with insulin in response to meals; its biological role is to reinforce satiety, slow gastric emptying, and suppress postprandial glucagon. Petrelintide reproduces this biology with extended duration by acting as a balanced potent agonist at both the amylin receptor (AMYR — a heterodimer of the calcitonin receptor, CTR, and receptor activity-modifying proteins, particularly RAMP1–3) and at the calcitonin receptor (CTR) directly.

Central signaling occurs primarily in hindbrain structures — the area postrema and nucleus tractus solitarius — producing satiety signals, delaying gastric emptying, suppressing postprandial glucagon, and in preclinical models restoring leptin sensitivity and increasing energy expenditure. Because these pathways are distinct from the GLP-1 receptor system, the tolerability profile differs substantially from GLP-1 receptor agonists.

The neutral-pH engineering solves the fibrillation problem that limited native amylin and pramlintide: early amylin analogs required acidic formulations incompatible with co-formulation. Petrelintide's neutral-pH stability is central to the Roche strategy, enabling potential fixed-dose co-formulation with CT-388. The half-life of approximately 33.8 hours supports once-weekly dosing at steady state. The receptor pharmacology and formulation chemistry are described in the J Med Chem 2025 development paper (PMID 41217931).


Open questions

  • Full Phase 2b publication: ZUPREME-1 topline results were reported in March 2026; the full peer-reviewed trial publication with detailed safety tables, body-composition subgroup analyses, and dose-response characterization has not yet appeared in the available literature.
  • Human body-composition data: Preclinical DIO rat data showed preferential fat-mass loss with relative lean-mass preservation versus liraglutide, but whether this advantage translates to humans has not been confirmed in published data.
  • Cardiovascular outcomes: No cardiovascular outcomes trial has been conducted. Whether petrelintide reduces MACE events as demonstrated with approved GLP-1 receptor agonists is not established.
  • Long-term safety: Phase 3 will be the first characterization of multi-year exposure, including potential signals related to chronic CTR agonism on bone metabolism and pancreatic function.
  • Type 2 diabetes: ZUPREME-2 is ongoing; dedicated T2DM efficacy and safety data are not yet available.
  • Combination with CT-388: The Phase 2 combination study is expected to begin in H1 2026 and will be the first test of whether petrelintide plus a GLP-1/GIP incretin can match or exceed CagriSema-class outcomes with improved tolerability.
  • Real-world adherence: Trial adherence was notably favorable, but performance outside controlled trial conditions has not been characterized.

Related peptides

  • Cagrilintide — Zealand Pharma's earlier long-acting amylin analog, licensed to Novo Nordisk; being developed primarily in the CagriSema combination with semaglutide rather than as a monotherapy
  • Pramlintide — the first approved amylin analog (Symlin, FDA 2005), used as an adjunct to insulin in T1D and T2D; short-acting (multiple daily injections), acidic formulation, the clinical predecessor to the long-acting amylin class
  • Semaglutide — GLP-1 receptor agonist (Ozempic/Wegovy); the leading comparator and prospective combination partner context via CT-388
Hypotheses5 directions▾ collapse

Research directions for this peptide, selected from the current sources — hypotheses you can explore and model. None of it is proven yet; tap any one to see the full thinking.

openupdated 2026-06-05

Does petrelintide mainly suppress appetite by signaling through a specific brain region that is directly accessible from the bloodstream?

If true, it would explain variation in how well the drug works across individuals and could help doctors predict who will benefit most, while also guiding the design of drug combinations that target complementary appetite-suppression pathways.

The hypothesis
Petrelintide's satiety effect is mediated primarily through calcitonin receptor (CalcR) signaling on area postrema neurons rather than on hypothalamic neurons, because the area postrema lacks a blood-brain barrier and would be the first CNS site accessible to a peripherally injected acylated peptide of this size, and CalcR density in the area postrema predicts inter-individual variation in petrelintide response more strongly than hypothalamic CalcR expression.
Why it’s plausible
The boltz-2 ipTM=0.83 with pLDDT=81.9 for petrelintide/CalcR represents the most confident complex prediction in this peptide set, indicating high-quality structural docking. The CalcR is abundantly expressed in the area postrema (AP), a circumventricular organ with fenestrated capillaries, making it the primary CNS entry point for large peripherally dosed peptides. Native amylin's central anorectic effects are known to involve AP CalcR, but petrelintide's acylation and molecular weight may favor AP over downstream hypothalamic sites. Inter-individual variation in AP CalcR density, documented for other neuropeptide receptors, could explain response heterogeneity in Phase 1 data.
Why it matters
AP-centric signaling would predict that petrelintide's efficacy is sensitive to conditions affecting AP integrity (alcohol use, chemotherapy, aging) and would guide combination with GLP-1R agonists whose AP versus hypothalamic signal contributions differ from amylin, rationalizing the biological basis of the Roche combination strategy.
Plausibility.70
Novelty.50
Impact.65
Basis · grounding1 paper · 2 computed/notes
[1]
structureipTM=0.830, pLDDT=81.9 for petrelintide/CalcR complex, highest confidence interface in this set, consistent with strong CalcR engagement
[2]
sourceCalcR structure-activity and selectivity relationships reviewed; receptor distribution across CNS subregions is pharmacologically relevant
[3]
paper
Petrelintide described as novel long-acting amylin analog in early clinical trials; mechanism of central action not yet fully resolved
doi: 10.1007/s13300-025-01733-8
openupdated 2026-06-05

When petrelintide and a GLP-1 drug are combined, do they produce more hunger suppression together than you would predict by simply adding their separate effects?

If true, it would scientifically justify the billions invested in developing this combination and could mean that patients who plateau on GLP-1 drugs alone could achieve substantially greater weight loss by adding petrelintide, addressing one of obesity medicine's most pressing clinical unmet needs.

The hypothesis
Co-formulation of petrelintide with a GLP-1R agonist will produce synergistic rather than merely additive satiety because the two pathways converge on distinct but complementary neuronal populations in the nucleus tractus solitarius (NTS): GLP-1R agonists primarily activate GLP-1R-expressing NTS neurons (which relay gastric distension signals) while amylin/petrelintide activates CalcR-expressing AP neurons projecting to NTS, and co-activation of both input streams produces a supraadditive reduction in meal size through lateral inhibition circuits in the NTS.
Why it’s plausible
The Roche partnership described in the readme is explicitly built around combining petrelintide with other agents, likely GLP-1R agonists given Roche's pipeline. Native amylin and GLP-1 are known to have synergistic anorectic effects in rodent models (10.1038/s42255-025-01324-8 discusses metabolic adaptation in this context). The NTS integrates peripheral satiety signals from multiple receptor systems; CalcR and GLP-1R are expressed on distinct NTS subpopulations, and convergent activation of parallel inhibitory circuits is a known mechanism for signal synergy in neural circuits. If petrelintide activates AP-to-NTS CalcR projections while GLP-1R agonists activate gastric vagal NTS inputs, co-dosing would simultaneously activate two independent NTS silencing pathways.
Why it matters
Mechanistic evidence for synergy would justify the high-value Roche partnership economically and would predict that the combination should outperform either component in clinical trials by a margin greater than either drug's monotherapy effect added together, providing a testable differentiating claim for regulatory review.
Plausibility.65
Novelty.50
Impact.70
Basis · grounding2 papers · 1 computed/note
[1]
paper
Metabolic adaptation and circulating factors in obesity context, relevant to the sustained satiety signals that must be maintained for long-term weight management
doi: 10.1038/s42255-025-01324-8
[2]
paper
GLP-1R agonist pipeline review including combination strategies; petrelintide's amylin pathway is complementary to GLP-1R axis
doi: 10.1080/13543784.2025.2472408
[3]
noteRoche partnership valued at up to $5.3 billion specifically for combination development potential with other obesity agents
openupdated 2026-06-05

Is the proline substitution in petrelintide's core sequence the key change that stops the peptide from forming the toxic protein clumps seen with older amylin drugs?

If true, it would mean future amylin-based drugs could be stabilized with minimal sequence changes, making them easier to manufacture and less likely to trigger immune reactions than drugs requiring extensive modification throughout their structure.

The hypothesis
The acylation and sequence modifications in petrelintide that prevent amyloid fibrillation at neutral pH do so by disrupting the cross-beta spine formed by residues 20-29 of the native amylin sequence (SNNFGAILS in native; corresponding to SNNFGPILSS in petrelintide), where the Pro26 substitution inserts a helix-breaking residue that sterically prevents the intermolecular beta-sheet hydrogen-bonding required for fibril nucleation, while the remaining sequence modifications tune receptor affinity.
Why it’s plausible
Petrelintide sequence KCNTATCATQRLANFLVHSSNNFGPILSSTNVGSNTYX contains SNNFGPILSS at positions 22-31. Native amylin contains SNNFGAILS at equivalent positions, and residues 20-29 of native amylin are the fibrillogenic core. Comparing SNNFGAILS (native) to SNNFGPILSS (petrelintide), the substitution of A25 to P25 (proline) would break the beta-strand at a critical position in the fibril core. Proline is uniquely incompatible with beta-sheet because its cyclic pyrrolidine ring eliminates the backbone NH required for intermolecular hydrogen bonding. This structural rationale, based on direct sequence comparison, provides a mechanistic explanation for the neutral-pH stability described in the readme without invoking any literature claims.
Why it matters
Identifying Pro26 as the primary anti-fibrillation substitution would provide a minimal engineering principle for stabilizing amylin analogs: a single proline insertion at the fibrillogenic core, rather than wholesale sequence modification, potentially reducing immunogenicity risk from extensive sequence changes.
Plausibility.65
Novelty.55
Impact.60
Basis · grounding3 computed/notes
[1]
sequenceSNNFGPILSS in petrelintide vs SNNFGAILS in native amylin: proline at position 26 (P26 relative to sequence start) within the known fibrillogenic 20-29 core
[2]
notePetrelintide is engineered to be chemically stable at neutral pH, preventing amyloid fibrillation that hampered earlier amylin drugs
[3]
sourceProteolytic stability enhancement through amino acid substitution is a validated design strategy, and proline substitutions are among the most established backbone-stabilizing changes
openupdated 2026-06-05

Does petrelintide's N-terminal loop structure prevent it from activating the receptor responsible for blood vessel dilation and flushing side effects?

If true, it would provide structural reassurance that petrelintide can be safely dosed weekly for years without the vascular side effects that derailed some earlier amylin analogs, strengthening the case for its long-term use in obesity treatment.

The hypothesis
Petrelintide's pharmacological selectivity for CalcR over the closely related calcitonin gene-related peptide receptor (CGRP-R, which shares RAMPs but uses CLR rather than CalcR as its core subunit) is maintained by the N-terminal ring structure (Cys2-Cys7 disulfide implied by KCNTATCATQ), which positions the N-terminal pharmacophore to engage the CalcR ECD in a geometry that is sterically incompatible with the CLR ECD, thereby avoiding CGRP-R-mediated vasodilation.
Why it’s plausible
Petrelintide sequence KCNTATCATQRLANFLVHSSNNFGPILSSTNVGSNTYX contains cysteines at positions 2 and 9 (C2 and C9, i.e., KCNTATCATQ), consistent with a disulfide-constrained ring of 8 residues analogous to the native amylin/calcitonin N-terminal ring. CalcR and CLR differ in their extracellular domain topology, and the disulfide-constrained N-terminal loop is the primary specificity determinant for CalcR vs CGRP-R selectivity in the calcitonin family. Petrelintide's extensive sequence modification preserves this ring; if the ring geometry is optimized for CalcR over CLR, CGRP-R-mediated side effects (flushing, vasodilation, headache) would be absent, which is clinically important for a chronically dosed obesity drug.
Why it matters
Confirming CalcR/CGRP-R selectivity via the N-terminal ring would predict an absence of vasodilatory side effects that limited some amylin analog development, and would provide a structure-activity principle for all future CalcR-targeting amylin analogs in metabolic disease.
Plausibility.60
Novelty.50
Impact.60
Basis · grounding1 paper · 2 computed/notes
[1]
sequenceC2 and C9 in KCNTATCATQRL: disulfide ring of 8 residues analogous to native amylin C2-C7 ring, which is the canonical CalcR/CLR selectivity determinant
[2]
paper
CalcR and CLR RAMP complex pharmacology reviewed; ECL1 and ECD structures determine ligand binding selectivity between receptor subtypes
doi: 10.1111/j.1476-5381.2011.01525.x
[3]
structureipTM=0.830 for petrelintide/CalcR, high confidence, consistent with well-matched ring geometry for CalcR ECD
openupdated 2026-06-05

Could petrelintide also slow bone thinning in older women, addressing two major postmenopausal health problems with one weekly injection?

If true, it would distinguish petrelintide from GLP-1 drugs, which may actually worsen bone density through rapid weight loss, and could offer postmenopausal women a weight-loss drug that simultaneously guards against fractures, a combination not currently available.

The hypothesis
Petrelintide would suppress bone resorption in postmenopausal osteoporosis through CalcR agonism on osteoclasts, additive with its central appetite-suppressing and GLP-1R combination weight-loss effects, because CalcR is the canonical calcitonin receptor whose activation directly inhibits osteoclast activity, and petrelintide's structural similarity to calcitonin receptor-activating amylin could confer direct anti-resorptive bone effects.
Why it’s plausible
The calcitonin receptor (CalcR) is petrelintide's annotated target (readme) and is the same receptor through which calcitonin (a distinct peptide hormone) suppresses osteoclast activity in osteoporosis. Osteoclasts express CalcR at high density. Amylin analogs have been shown to have weak bone-anabolic effects in rodent models through CalcR. If petrelintide's engineered high CalcR affinity (ipTM=0.83 against CalcR) extends to osteoclast CalcR expressed in bone, a bone-protective effect would be expected in the postmenopausal women who are also the target demographic for obesity treatment, creating a potentially valuable pleiotropy.
Why it matters
A single weekly injectable combining weight loss and bone protection would address two of the most prevalent postmenopausal conditions simultaneously, reducing polypharmacy and adding commercial differentiation from GLP-1-only agents, which may paradoxically reduce bone density through rapid weight loss.
Plausibility.40
Novelty.55
Impact.60
Basis · grounding1 paper · 2 computed/notes
[1]
structureipTM=0.830 for petrelintide/CalcR complex, high confidence interface consistent with potent CalcR engagement
[2]
sourceCalcR structure-activity relationships reviewed including selectivity determinants relevant to osteoclast versus neuronal CalcR subtypes
[3]
paper
Petrelintide described as amylin analog acting through CalcR; amylin receptors include CalcR/RAMP complexes expressed in osteoclasts
doi: 10.1007/s13300-025-01733-8
details expand to inspect
full evidence table2 metrics
metricvaluetool
ipTM 0.8308536410331726 boltz-2
ranking score 0.8210369944572449 boltz-2
3-letter notation
Lys-Cys-Asn-Thr-Ala-Thr-Cys-Ala-Thr-Gln-Arg-Leu-Ala-Asn-Phe-Leu-Val-His-Ser-Ser-Asn-Asn-Phe-Gly-Pro-Ile-Leu-Ser-Ser-Thr-Asn-Val-Gly-Ser-Asn-Thr-Tyr-X
recipeboltz-2 2.2.1
parametervalue
modelboltz-2 2.2.1
weights
hardwarevast_v100_32gb
mlx version
python
random seed1
msa strategycolabfold_local
runtime
predicted by
predicted at2026-05-22
citationbibtex
peptidemodel (2026). Petrelintide: weekly weight-loss injection (NN9213/ZP8396) (pep-10905, v1). PeptideModel. https://peptidemodel.com/card/pep-10905
@peptide{pep10905,
  sequence = {KCNTATCATQRLANFLVHSSNNFGPILSSTNVGSNTYX},
  target   = {calcr},
  author   = {peptidemodel},
  year     = {2026},
  status   = {computed}
}
related peptides 5 by signal overlap
clinical trials 4 on ct.gov · checked 2026-05-09
ct.gov trials 4
by phase
2phase 12phase 2
by status
3completed1active
references 2 papers
discussion no comments
sign in to comment
peptidemodel.com CC-BY-SA-4.0 research only · not for human use