Targeting enkephalins and pituitary adenylate cyclase-activating polypeptide (PACAP) in migraine

Migraine is a ubiquitous neurological disease that affects more than one billion people worldwide.¹ Its neurobiological underpinnings involve activation of primary afferents whose cell bodies are located in the trigeminal ganglia and upper cervical ganglia.¹ The mechanisms by which the primary afferents become activated and sensitized remain incompletely understood, but there is ample evidence to support the role of specific peptides in modulating nociceptive transmission of cephalic pain.¹ Indeed, several drugs targeting calcitonin gene-related peptide (CGRP) or its receptor have recently proven efficacious for migraine.¹ However, a considerable proportion of people with migraine do not respond to these new medications.¹ The need for additional drug discovery is thus clear, and two important animal studies that were published in Brain this year shed light on the therapeutic promise of targeting enkephalins and pituitary adenylate cyclase-activating polypeptide (PACAP).

In 1979, Sicuteri was the first to hypothesize dysfunction of the opioid system in migraine.² β-Endorphin is an endogenous opioid neuropeptide. Early studies reported that decreased levels of β-endorphin correlated with the severity of the disease in the CSF.³ Elevated levels of plasma and platelets methionine-enkephalin (MET) were reported during migraine attacks.⁴ In this issue of Brain, Descheemaeker et al.⁵ report the results of an animal study in which they hypothesized that protecting enkephalins from their degrading enzymes could be a possible drug target for migraine. They used an animal model to investigate the analgesic effect of the Dual ENKephalinase Inhibitor (DENKI) PL37 on cutaneous mechanical allodynia. PL37 is a small molecule that inhibits metalloproteases and thereby protects enkephalins from their rapid degradation. Recurrent administration of the NO donor, isosorbide dinitrate (ISDN), was used to induce cutaneous mechanical allodynia and sensitization of the trigeminocervical complex (TCC). The allodynia was used as a surrogate for headache. The major finding was that single oral PL37 inhibited ISDN-induced acute cephalic mechanical allodynia and single intravenous, but not oral, PL37 administration inhibited chronic cephalic mechanical allodynia. In addition, daily oral administration of PL37 prevented cephalic mechanical allodynia and decreased touch-induced c-Fos expression in TCC. In interpreting these data, two questions arise: (i) whether DENKIs or delta-opioid receptor agonists would possess a potential for abuse or cause safety and tolerability issues in humans; and (ii) whether preclinical models using NO donors to induce mechanical allodynia as a surrogate for headache are useful for drug testing. The former is not completely clarified and should be investigated in phase 1 studies before therapeutic use. To further address the role of endogenous opioids in migraine, more studies using validated assays on CSF and blood-based biomarkers are needed. Possible changes in endogenous opioids in response to treatment will be of interest in this context. With regard to the second question, it is important to address the concept of using the NO model in animals as a surrogate model for headache or migraine in humans. In humans, NO models have been used to test the efficacy of anti-migraine drugs. These studies showed that propranolol, zolmitriptan and olcegepant (a small molecule CGRP receptor antagonist) failed to prevent NO-induced headache and migraine. In addition, sumatriptan failed to reduce headache induced by orally administered NO donor isosorbide-5-mononitrate (5-ISMN) in healthy volunteers. The ineffectiveness of several anti-migraine drugs in the NO model limits its translatability, and findings should be interpreted with cautious optimism.⁶

Since its first publication in Brain as a migraine inducing substance in 2009,⁷ PACAP has emerged as a key molecule in primary headaches. PACAP and its receptors have been suggested as a potential novel target for migraine treatment.¹ PACAP exerts its effects at four different receptors—VPAC1, VPAC2, PAC1, and the orphan MrgB3 (in rodents) or MrgX2 (in humans) receptor. Experimental studies have implicated the CGRP/PACAP-cyclic adenosine monophosphate (cAMP)–mediated pathway in the genesis of migraine attacks and suggested the opening of ATP-sensitive potassium (K_ATP) channels as a possible unifying mechanism.¹

In a paper that was published earlier this year in Brain, Ernstsen et al.⁸ investigated whether PACAP-induced hypersensitivity is independent of CGRP. Both peptides are potent dilators of extracerebral arteries and known to induce migraine attacks.¹ An intriguing question is whether anti-PACAP drugs might be effective in individuals who do not respond to CGRP-signalling targeting therapies. This would suggest that multiple signalling pathways are involved in individuals with migraine. CGRP mediates its effects through calcitonin receptor-like receptor (CRLR) and a receptor activity-modifying protein (RAMP1). Ernstsen et al.⁸ used mouse models of provoked migraine-like pain by PACAP, CGRP and NO donor glyceryl trinitrate (GTN) and measurement of tactile sensitivity. The study included genetically modified mice lacking either functional CGRP receptors (RAMP1 knockout) or TRPA1 channels (TRPA1 knockout). In wild-type mice, PACAP-38 induced hypersensitivity both in the hind paws and in the cephalic area. The most important outcome was no differences in the development of PACAP-38-induced hypersensitivity between the RAMP1 deficient and wild-type mice. Furthermore, the humanized monoclonal CGRP antibody (ALD405) did not prevent PACAP-38-induced hypersensitivity in wild-type mice. In addition, the study found that the TRPA1 channel was not involved in PACAP-38-induced hypersensitivity. Interestingly, GTN hypersensitivity is dependent on RAMP1 and TRPA1. Finally, the study showed that a non-selective K_ATP channel inhibitor glibenclamide partially inhibited PACAP-38-induced hypersensitivity. These results contrast with human data showing that glibenclamide did not attenuate PACAP-38-induced headache and haemodynamic changes in healthy volunteers,⁹ but the potency of glibenclamide on the relevant subtype K_ATP channel is probably too small in humans.

Together, both studies provide new data supporting the possible use of delta opioid agonists and anti-PACAP drugs in migraine treatment. More studies in humans are needed to assess the safety and risk of addiction to delta opioids. Monoclonal PAC₁ receptor antibody failed in a proof-of-concept study for migraine prevention,¹⁰ and two ongoing randomized clinical trials on the efficacy, safety and tolerability of monoclonal antibodies designed to target PACAP (ClinicalTrials.gov Identifier: NCT04498910; ClinicalTrials.gov Identifier: NCT05133323) will reveal whether targeting PACAP-signalling is effective for migraine prevention.