PACAP-38

A 38-amino-acid neuropeptide of the VIP-secretin-glucagon superfamily that signals through PAC1, VPAC1 and VPAC2 receptors; its preclinical record is neuroprotective, while every drug program built on it blocks the peptide rather than supplying it.

Also referenced as: PACAP38, PACAP 38, PACAP(1-38), PACAP 1-38, Pituitary Adenylate Cyclase-Activating Polypeptide 38, Pituitary Adenylate Cyclase Activating Polypeptide 38

Also appears in: Cognitive

COA Public product evidence No certificate records are bound to a tracked PACAP-38 product page A record counts here only when a vendor publishes a certificate against one of its own tracked PACAP-38 product pages. Other certificates naming this compound may still exist — search the wider certificate database. →

---

What is PACAP-38?

Pituitary adenylate cyclase-activating polypeptide is a 38-residue, C-terminally alpha-amidated peptide isolated from ovine hypothalamic extract in 1989, using stimulation of cAMP formation in rat anterior pituitary cell cultures as the assay that guided the purification (Miyata et al., Biochemical and Biophysical Research Communications, 1989, 164(1):567–574). It belongs to the vasoactive intestinal polypeptide–secretin–growth hormone-releasing hormone–glucagon superfamily, and its sequence has been conserved from protochordates to mammals (Vaudry et al., Pharmacological Reviews, 2009, 61(3):283–357). A shorter 27-residue form, PACAP-27, corresponding to the N-terminal 27 amino acids of PACAP-38 with an amidated C-terminus, was recovered from side fractions during the same purification; it shares considerable homology with vasoactive intestinal polypeptide yet stimulates adenylate cyclase roughly 1000 times more strongly (Miyata et al., Biochemical and Biophysical Research Communications, 1990, 170(2):643–648). Both forms arise from a single precursor, cloned from ovine and human cDNA shortly afterward (Kimura et al., Biochemical and Biophysical Research Communications, 1990, 166(1):81–89).

The reason PACAP-38 appears in research catalogs is a large animal literature on neuronal survival after ischemic and toxic injury. The reason it appears in clinical journals is almost the opposite: in humans it is infused to provoke headache, and the pharmaceutical programs built around it are antibodies and receptor blockers designed to neutralize the peptide, not to supply it.

How it works

• Three receptors, one of them PACAP-selective. PACAP binds two classes of site: type I sites with high affinity for PACAP and much lower affinity for vasoactive intestinal polypeptide, and type II sites that bind both about equally. Cloning resolved these into the PACAP-preferring PAC1 receptor, which couples to several transduction systems, and the PACAP/VIP-indifferent VPAC1 and VPAC2 receptors, which couple primarily to adenylyl cyclase. PAC1 is concentrated in brain, pituitary and adrenal gland, while the VPAC receptors predominate in lung, liver and testis (Vaudry et al., Pharmacological Reviews, 2009).
• The proximate signal is cAMP. The peptide was found by its ability to raise cAMP in pituitary cells, and receptor-driven adenylyl cyclase activation remains the canonical pathway; PAC1 splice variants add phospholipase C coupling, which is part of why a single ligand produces secretory, trophic and nociceptive effects in different tissues (Vaudry et al., Pharmacological Reviews, 2009).
• It is a potent and unusually long-acting vasodilator in humans. Infused into migraine patients, PACAP-38 widened the superficial temporal artery by 37.5% and lowered middle cerebral artery flow velocity by 16.1% at 20 minutes (Schytz et al., Brain, 2009, 132(Pt 1):16–25). In a head-to-head crossover against vasoactive intestinal polypeptide, 22 patients completing both infusion days, both peptides dilated extracranial but not intracranial arteries, yet the PACAP-38 dilatation persisted beyond 2 hours while the VIP response had normalized by then (Amin et al., Brain, 2014, 137(Pt 3):779–794).
• Mast cells are one proposed downstream effector. PACAP-38 degranulates dural mast cells, which has been advanced as a mechanism for the delayed headache phase that follows infusion (Baun et al., Cephalalgia, 2012, 32(4):337–345).
• In animal injury models the peptide is protective, and the effect is partly immune-mediated. In rats, an intravenous bolus of 20 nmol/kg followed by a 48-hour infusion begun 4 hours after a 2-hour middle cerebral artery occlusion cut infarct size by 50.9%, though starting at 8 or 12 hours did not reach significance (Reglodi et al., Stroke, 2000, 31(6):1411–1417). PACAP-38 failed to reduce the infarct in interleukin-6-null mice, implicating IL-6 induction in the protection (Ohtaki et al., Proceedings of the National Academy of Sciences, 2006, 103(19):7488–7493).

Research status

The human and animal literatures point in opposite directions, and that divergence runs through the whole record.

Animal neuroprotection is well replicated. Systemic and central PACAP-38 reduces infarct volume and neuronal loss across global and focal ischemia models, with benefit reported even when dosing begins hours after the insult, and reviews have collected this as a case for a wide therapeutic window (Reglodi et al., Stroke, 2000; Fang et al., Aging and Disease, 2020, 11(6):1496–1512). None of it has been tested as a therapy in people.

What human administration actually does is cause headache. In a randomized double-blind crossover study, 10 pmol/kg/min PACAP-38 infused over 20 minutes produced headache in all 12 healthy subjects and in 11 of 12 migraine patients without aura. Seven patients developed migraine-like attacks after PACAP-38 and none after placebo (P = 0.016), with 6 of the 7 attacks arriving in the 2-to-12-hour window after infusion, at a mean of 6 hours (Schytz et al., Brain, 2009). The later comparison study, a crossover in which each patient received both peptides on separate days, put the attack rate at 73% for PACAP-38 versus 18% for vasoactive intestinal polypeptide (P = 0.002) (Amin et al., Brain, 2014). This provocation model is now a standard pharmacodynamic tool in headache research.

Cardiovascular tolerability bounded the dose early. The first study of intravenous PACAP-38 on cerebral hemodynamics in 12 healthy volunteers found no major direct effect on regional cerebral blood flow once end-tidal CO2 was accounted for, but heart rate rose 61.9 ± 22.4% (P < 0.0001 versus placebo) and end-tidal CO2 fell 8.9 ± 3.8%. The authors named the tachycardia and the CO2-driven flow reduction as dose-limiting factors for any future clinical study (Birk et al., Regulatory Peptides, 2007, 140(3):185–191). That paper opened by describing PACAP-38 as a candidate for human stroke studies; nearly two decades later, those studies have not been run.

Drug development has consistently moved toward blocking the peptide. AMG 301, a PAC1 receptor antibody, was tested in 343 randomized patients and produced no separation from placebo on any efficacy measure (Ashina et al., Cephalalgia, 2021, 41(1):33–44; NCT03238781). Neutralizing the ligand worked better: the anti-PACAP antibody Lu AG09222 blunted PACAP38-induced vasodilation, heart-rate increase and headache in healthy volunteers (Rasmussen et al., The Journal of Headache and Pain, 2023, 24(1):60; NCT04976309), and in the HOPE phase 2 trial a single 750 mg intravenous infusion reduced monthly migraine days by 6.2 versus 4.2 on placebo over weeks 1 through 4 in 237 adults whose previous preventives had failed (difference −2.0 days; 95% CI −3.8 to −0.3; P = 0.02) (Ashina et al., New England Journal of Medicine, 2024, 391(9):800–809; NCT05133323). A competing anti-PACAP antibody, LY3451838, missed in a 38-patient phase 2 study (Johnson et al., Cephalalgia, 2025, 45(8); NCT04498910). Lu AG09222 has since advanced under the name bocunebart: in the phase 2b PROCEED trial (NCT06323928) the subcutaneous part stopped for futility at a prespecified interim analysis, and the sponsor reported on February 12 and June 4, 2026 that the intravenous part, with 431 patients, reduced monthly migraine days by 4.24 against 2.86 on placebo over weeks 1 through 12 (difference 1.38 days; P = 0.0178). No PACAP-targeting medicine of any kind has reached approval.

Other lines of PACAP research are genetic or observational rather than interventional. A genome-wide study linked PACAP and a common variant in the PAC1 receptor gene to post-traumatic stress disorder diagnosis and symptoms in women (Ressler et al., Nature, 2011, 470(7335):492–497), a finding about endogenous signaling that says nothing about administering the peptide. Across ClinicalTrials.gov, the registered studies that administer PACAP-38 or PACAP-27 at all are mechanistic: headache and flushing induction (NCT00380263, NCT03881644, NCT03471039, NCT03560024), imaging during induced attacks (NCT01471990), induction in cluster headache (NCT03814226) and post-traumatic headache (NCT05378061), challenge studies testing whether erenumab, eptinezumab or glibenclamide blunt the response (NCT02542605, NCT05635604, NCT04960657), and the newest registration, posted January 15, 2026 by the Danish Headache Center, which proposes to use PACAP-38 to induce migraine attacks with aura (NCT07342296).

Common dosage forms

• Research-grade lyophilized peptide. Human PACAP-38, usually supplied as the trifluoroacetate or acetate salt in microgram to low-milligram quantities for receptor binding, cell-culture and tissue-bath work. Ovine and rat sequences are stocked as separate catalog items, and species mismatch matters in assay work.
• Related peptides sold as distinct items. PACAP-27 is a separate product with its own pharmacology, and the PACAP(6-38) fragment is a receptor antagonist rather than an agonist. Biotinylated and radiolabeled derivatives exist for binding studies and are not interchangeable with the native peptide.
• Research-protocol intravenous infusion. Every clinical administration in the published record has been an intravenous infusion prepared for a research protocol and delivered by study staff in a hospital setting, not a fixed-strength vial, pen or spray.

Key considerations

• Regulatory status. The Research Market status reflects the absence of any approved product containing PACAP-38 and of any trial testing it as a therapy, alongside its sale by research-reagent suppliers. The registered human studies that administer it are provocation and imaging protocols, and the only PACAP programs in active clinical development are antibodies that neutralize it.
• The strongest human data describe a harm, not a benefit. Infusion reliably produces headache in healthy volunteers and migraine-like attacks in most migraine patients, with attacks appearing hours after the infusion, and the tachycardia it causes was flagged as dose-limiting in the first hemodynamic study.
• Naming is a live source of confusion. PACAP-38 and PACAP-27 are different peptides from the same precursor with different potencies; PACAP(6-38) is an antagonist fragment used as a laboratory tool; vasoactive intestinal polypeptide is a related superfamily member with markedly weaker cAMP stimulation and a shorter vascular effect; PAC1, VPAC1 and VPAC2 are three distinct receptors; and ELISA or radioimmunoassay kits that measure PACAP are a different product class from the peptide itself. The gene symbols ADCYAP1 and ADCYAP1R1 refer to the precursor and the PAC1 receptor, not to the peptide as supplied.
• Receptor versus ligand. A PAC1 receptor antibody failed while two ligand-neutralizing antibodies produced a small positive signal and a nonsignificant one, a pattern the trial investigators have attributed to PACAP acting through more than one receptor in headache; the clinical case for this target rests on removing the peptide.
• No translational step was taken. A quarter-century after delayed systemic dosing halved infarct volume in rats, there is no published pharmacokinetic profile supporting a practical dosing route and no safety database beyond short mechanistic infusions in small groups of volunteers and headache patients.