Role of VIP/PACAP in primary headaches

In this issue of Cephalalgia Tuka and colleagues examine and compare levels of pituitary adenylate cyclase-activating polypeptide-38 (PACAP-38) with calcitonin gene-related peptide (CGRP) in blood from the cubital vein of 87 migraineurs (ictal and interictal) and of 40 healthy controls. The authors report that both peptides are elevated in the peripheral circulation during migraine attacks as compared to their levels between attacks. There was no association with gender, age, attack frequency, allodynia, visual analogue scale score or migraine type (1). PACAP-38 was lower in interictal blood of the patients than in healthy controls. The reason behind this is not clear but we should realize that any analysis of peptides in the peripheral circulation reflects PACAP-38 levels that may come from the entire body, and therefore it is difficult to link it directly to the brain or to cranial structures. Since both PACAP-38 and CGRP are elevated during migraine attacks, the authors suggest a role in migraine. Another interesting finding is the observation of significant release of PACAP-38 during migraine attacks (i) in nonmenstrual migraine and (ii) in subjects without chronic back pain. The implication is so far not clear but opens the way for more research.

In support of this study, the Copenhagen group has revealed that intravenous administration of PACAP-38 results in headache in healthy individuals and in migraine-like effects (after a mean of six hours) in migraine without aura patients (2). This group has in addition reported that PACAP-38 results in a related increase in the diameter of cranial arteries (2,3). They propose that PACAP-38 caused vasodilation, which resulted in the headache, providing support for the vascular theory of migraine (4). Because the closely related peptide vasoactive intestinal peptide (VIP) elicits the same degree of vasodilatation via the same type of receptors but lacks the migraine-inducing response (5), we are awaiting the head-to-head comparison study and an explanation of this interesting observation.

Where do we find PACAP-38 and its receptors?

The intracranial blood vessels are innervated by sympathetic, parasympathetic and sensory neurons, each of which contains multiple neuronal messenger molecules (6,7). The role of the individual neuronal messenger, however, still remains unclear. The two related peptides VIP and PACAP have inspired renewed interest in recent years, in part because of the above studies.

VIP was first isolated from porcine intestines by Said and Mutt in 1970 (8) while PACAP was discovered and isolated in 1989 by Miyata et al., who also found it to be a potent stimulant of cyclic adenosine monophosphate (cAMP) production by adenylate cyclase (AC) in cultured anterior pituitary cells (9). Both peptides share sequence homology at the N-terminal end and are considered members of the glucagon/secretin superfamily of peptides (10,11).

PACAP exists in a 38 amino acid form (PACAP-38) and a C-truncated 27 amino acid form (PACAP-27), of which the former is the most predominant. They are both derived from the same 175 amino acid precursor. VIP, PACAP-38 and PACAP-27 all bind to the VIP/PACAP receptors VPAC₁ and VPAC₂ receptors with equal affinity. The PAC₁ receptor has a high affinity for both forms of PACAP, but has 100- to 1000-fold less affinity for VIP. The VPAC₁, VPAC₂ and PAC₁ receptors belong to a large group of seven-transmembrane G-protein-coupled receptors (12). The principal effect of VPAC₁/VPAC₂ or PAC₁ receptor activation is an increase in cAMP through AC activation. Activation of other second messenger systems including phospholipase C and phospholipase D has been implied and may be present in concordance with AC activation.

VIP and PACAP co-exist in most of the parasympathetic fibers

In particular the parasympathetic cell bodies distributing perivascular nerves in the cranial circulation are localized in the otic and sphenopalatine ganglia (6). These neurons contain VIP/PACAP, nitric oxide synthase and a minor population of choline acetyltransferase, forming acetylcholine. In a few migraine patients and in cluster headache there is release of VIP during the attacks, and this release seems to correlate with facial reddening (13,14). Recently, Csati et al. provided evidence that there are sensory CGRP-containing nerve fibers (probably from the trigeminal ganglion) that project to the cranial parasympathetic ganglia and that there are CGRP receptor components (CLR, RAMP1) on satellite glial cells in the sphenopalatine ganglion (15). This provides evidence for an interaction between the parasympathetic and sensory systems.

To my mind it is likely that most of the cranial-released VIP and PACAP would originate in the parasympathetic nerves and ganglia. In support, tracing studies in intracranial vessels have revealed that VIP-positive nerves originate in the sphenopalatine and otic ganglia (16), while the PACAP-positive neurons are found both in the above parasympathetic ganglia and to a lesser degree in the trigeminal ganglion (17).

Interestingly, there is a subpopulation of PACAP-containing neurons in the human trigeminal ganglion, but no VIP immunoreactivity (18). The role of PACAP seems to involve both these pathways. The localization of PACAP to both the parasympathetic and the sensory systems has attracted a discussion that this peptide has a multifaceted role, possibly involving a vasomotor function and a role in sensory mechanisms, the latter since it, unlike VIP, elicited a migraine-like attack in migraineurs (2,5).

It is worth noting that immunohistochemistry has shown PACAP immunoreactivity in the feline trigeminal nucleus caudalis in the Rexed lamina I/II, where also CGRP can be seen (19). In general there exist neurons and fibers within the central nervous system (CNS) that store both VIP (20) and PACAP (10,11); however, it is unlikely that these are involved in the effects of systemic PACAP or in PACAP measured in the peripheral circulation because the peptides are large molecules that only with difficulty can pass the blood-brain barrier (21).

Where does circulating PAPAP-38 originate?

PACAP-38 measured in the cited study likely could have two sites of origin: (i) The peptide can be found in many organs in the body and hence the basal level is likely to reflect this; and (ii) the elevated level seen in conjunction with a migraine attack probably originates from the activation of cranial parasympathetic and sensory ganglia. There is an association with the PACAP elevation in the ictal period relative to the attack-free period in migraineurs that supports this assumption (1). However, it would have been interesting if the authors had included treatment with a triptan to substantiate this and the relation to the clinical characteristics.

We might also speculate as to the role of PACAP-38 in the migraine attack. It is well known that VIP and PACAP are potent vasodilations of cerebral and dural (middle meningeal) arteries (21–23). Work on human cranial arteries suggests that all three receptors for VIP/PACAP are present on cranial arteries and may induce vasodilatation (21,24,25), provided that the peptides can pass the blood-brain barrier. Other functions of PACAP have been suggested: Baun and colleagues (26) have proposed that mast cell degranulation in the dura mater could be involved in the PACAP-induced migraine-like attack that can be induced by systemic administration of the peptide. Another way could be via interaction between the parasympathetic cranial ganglia and the sensory trigeminal ganglion in humans (15). The results obtained by Tuka et al. (1) are interesting and indicate novel aspects of migraine pathophysiology that need to be analyzed in future research.