Amylin and the amylin receptors in migraine: Is there another pathway to target?

Abstract

Here, we review the available evidence implicating amylin in migraine, its relationship with calcitonin gene-related peptide (CGRP) and its potential utility as a therapeutic target. The pathophysiology of migraine is currently better understood and the role of CGRP is key. Treatments targeting this pathway have been successful and migraine is a complex disorder, with so many molecules being implicated. Amylin, as for CGRP, is part of the calcitonin/CGRP peptide family. Some therapies intended to block the CGRP pathway can also target amylin receptors. Similar to CGRP, amylin can trigger migraine attacks in a provocation study. Amylin plasma levels have been highlighted as a potential migraine biomarker in one study in migraine patients. Moreover, some preclinical studies in rodents have also discussed sex differences. Comprehending the distinct and overlapping mechanisms between amylin and CGRP signalling could develop further our understanding of migraine pathophysiology. In summary, this review reveals, through initial studies, that targeting the amylin pathway may have a potential role as a novel treatment option for those who may not respond to other treatments, or as a better alternative.

Introduction

The discovery of calcitonin-gene related peptide (CGRP), its role in migraine pathophysiology and the utility of treatments targeting its pathway is indubitably a paradigm of bench to bed-side medicine (1). The role of CGRP role provides important insights into migraine and offers an efficacious pharmacological target for many patients although not for all (2). While CGRP is important in migraine pathophysiology, it does not explain all the variability of this complex disorder (3), such that a lack of response in some patients is not surprising. There is sufficient evidence to suggest that some patients benefit from some treatments considered to target the CGRP pathway but not from others targeting that same pathway (4). Namely, some patients may benefit from only some CGRP-pathway monoclonal antibodies or CGRP receptor antagonists (5). Part of this variability may be explained by the fact that we are also targeting other pathways in which CGRP is implicated.

Calcitonin/CGRP family

The calcitonin peptide family includes calcitonin, α-CGRP, β-CGRP, adrenomedullin (AM), adrenomedullin2/intermedin (AM2/IMD) and amylin (6). All these molecules bind to a family of G protein-coupled receptors. The two main components of these receptors are either the calcitonin receptor (CTR) or calcitonin receptor-like receptor (CLR) and one of the three receptor activity-modifying proteins (RAMP1–3). The specific combination of these components significantly affects their affinity to the different neuropeptides. CGRP not only strongly binds to the canonical CGRP receptor which consists of the CLR + RAMP1 components, but also, somewhat less so, to the amylin receptor 1 (AMY₁), which contains CTR + RAMP1_. AMY₂ contains CTR + RAMP2 and AMY₃ CTR + RAMP3. ⁶ Of note, AM2/IMD can activate CTR and AMY receptors. Across species, the activity of AM2/IMD has been suggested to be greater at rat AMY₃(a) receptors compared to the human receptor (7,8). Similarly AM, may also have more activity at rodent compared to human AMY₃(a) receptors (8). Studies in rat trigeminal ganglion neurons, presumably with both AMY₁ and CGRP receptors, rat αCGRP stimulated production of cAMP and phosphorylation of p38 but not extracellular signal regulated kinase (ERK). Transfected cells with human AMY₁ and CGRP receptors exposed to human αCGRP stimulated production of cAMP and phosphorylation of ERK among others, but not p38 (6). These and other potential differences among species in amylin receptors should be taken into account when considering results of animal models and extrapolating these to human models.

Amylin is synthesized as an 89-residue preprohormone, the 67-residue proform is processed in the Golgi apparatus and in the insulin β-cell secretory granule to yield the mature hormone of 37 aminoacids. The active peptide has an amidated C-terminus and a disulfide bridge between Cys-2 and Cys-7 (9). Amylin is a very weak agonist of the human canonical CGRP. It is a high-affinity ligand for the AMY₁ and AMY₃ receptors, variable of AMY₂ (6,10)_.

Amylin: History

Amylin was first described in 1987 and its function has mainly been studied in relation to its function to the gastrointestinal system and glucose homeostasis (11). Within the pancreatic β-cells, amylin is released from the same granules from which insulin is secreted. It is thought that, while insulin regulates glucose disposition, amylin regulates glucose inflow. The longstanding notion of glucose homeostasis being explained solely by insulin and glucagon was disrupted by adding amylin to the mix of substances that regulate this functional equilibrium (12). One hundred years ago, in 1923, the discovery of insulin was awarded with the Nobel prize to Banting and Macleod from the University of Toronto (13). Since then, our understanding of the properties and functions of insulin has evolved. For example, its ability to modulate nociception has been shown in a recent study using a rat model showing how insulin itself could activate or sensitize the transient receptor potential vanilloid 1 (TRPV1) in the trigeminovascular system leading to the release of CGRP and increasing the susceptibility of the receptor to capsaicin (14). Amylin restrains gastric emptying, which is reversed by hypoglycaemia (15), and also suppresses postprandial glucagon secretion, impeding the production of endogenous sugar by the liver when this is not necessary (16). Indeed, diabetes is no longer understood exclusively as a deficit of insulin because patients may also benefit from amylin analogues (17,18).

Due to the role of amylin in glucose homeostasis, deficiency of amylin is one component of the pathophysiology of diabetes (19,20). The accumulation of amylin aggregates is assumed to be cytotoxic and is suspected of playing a role in the impairment of pancreatic β-cell function in diabetes (21). Amylin also acts by stimulating the satiety centre in the brain (22). Indeed, amylin is, to a lesser extent, produced by brain cells, specifically at the level of the hypothalamus in the rostral medial preoptic area as demonstrated in animal models (23). Interestingly, amylin mRNA expression in this area was increased by 24-fold in lactating female rats compared to pup-deprived and nulliparous rodents with no expression in male rats (24). This particular difference should be considered in migraine models involving rodents because it may explain a difference in response between sexes to both amylin agonists and antagonists.

There is no evidence to date supporting a direct relationship between diabetes and migraine. Futhermore, in the initial stages of type 2 diabetes, the primary issue is insulin resistance. As the disease progresses, insulin secretion becomes impaired. Amylin levels are not necessarily reduced in type 2 diabetes and the ratio of insulin to amylin production is likely to be altered (25). However, there is evidence showing that both basal and stimulated amylin pancreatic secretion are significantly higher in patients with obesity with and without glucose intolerance (26). It is possible that shared genetic or metabolic pathways contribute to both migraine and obesity with glucose intolerance, and amylin may be a key mediator given that there is also literature suggesting co-morbidity between migraine and glucose-related traits (27).

Amylin and migraine

Knowledge of the function of the CGRP receptor family and the interaction between the different receptors has increased in recent years (28). Experimental and clinical evidence demonstrates that elevation of plasma CGRP levels is found in migraine (29) and activation of the canonical CGRP receptor is able to trigger migraine-like attacks (30,31), while its blockade either through a receptor antagonist or a monoclonal antibody is able to abort migraine attacks and reduce attack susceptibility thereby reducing attack frequency (32,33). Taken together, these data raise the question as to whether other receptors of the same receptor family would share similar properties (2).

This hypothesis was addressed in a recent Danish study which showed that the infusion of the amylin receptor agonist pramlintide, comprising a US Food and Administration Drug approved drug for the treatment of diabetes (34–36), is able to trigger migraine-like attacks in a subgroup of patients previously diagnosed with migraine without aura (37). While of the thirty-four enrolled patients, 30 (88%) developed an immediate unspecific headache after pramlintide infusion, compared to 33 (97%) after CGRP infusion (p = 0.375), the crucial finding was that 14 patients (41%) developed migraine-like attacks after pramlintide infusion, compared to 19 patients (56%) after CGRP infusion (p = 0.180). Moreover, the headache intensity score was significantly higher for CGRP infused patients compared to pramlintide infused patients (p < 0.001). The timeline of migraine-like attacks triggered by pramlintide followed the same pattern observed with CGRP-triggered migraine-like attacks. These findings are supported by preclinical studies which suggest that amylin may also be of relevance in mediating cutaneous allodynia and photophobia even if the effect appears to be slightly weaker compared to CGRP (37,38). These studies therefore suggest that amylin may play a similar role to CGRP in the susceptibility and triggering of migraine attacks. Of note, the observed vascular effects of pramlintide were much less pronounced compared to those observed with CGRP (37), suggesting a direct neuronal effect.

Potential mechanisms of amylin-induced migraine attacks

The mechanisms underlying the attack-triggering effect of amylin remain to be elucidated and several modes of action have been hypothesised.

First, the effect of pramlintide could be mediated via the release of CGRP. However, the fact that only 35% of the patients included in the described cross-over study developed migraine-like attacks after infusion with pramlintide, as well as after the infusion of CGRP, makes this less likely (37). In addition, the infusion of pramlintide did not induce an increase of plasma CGRP. The data does strongly suggest that the attack-triggering abilities of CGRP and pramlintide are to some extent independent and mediated through different pathways.

Second, in contrast to amylin (39,40), CGRP may not readily cross the blood–brain barrier (BBB). More research in warranted given the observation that in some patients, migraine-like attacks may be triggered by pramlintide, in some by CGRP and in some by both peptides. To what extent differing brain penetration could play a role in these results has not been explored. However, the ability of amylin to cross the BBB is relatively weak (39,40).

Third, increasing evidence shows that, at physiological concentrations, CGRP is able to bind the canonical CGRP receptor, as well as the AMY₁ receptor, and the same applies to CGRP receptor antagonists (41,42). As previously discussed, there is some weak agonist action of amylin at the CGRP receptor. The attack-triggering ability of pramlintide may be mediated through an independent action at the AMY₁ receptor, both through AMY₁ and CGRP receptors or, less likely, through an exclusively direct action on the canonical CGRP receptor. In addition, the canonical CGRP receptor and the AMY₁ receptor differ in that after binding of its ligand, the CGRP receptor is internalised. However, the functional and even more so the clinical significance of this mechanistic difference remains unknown (43,44). Given the AMY₁ receptor is only scarcely internalised, this could be in line with more accessible acute therapies (45). AMY₁ receptor may be expressed in the trigeminal fibres (37), CTR is expressed in the trigeminal ganglion (46) and amylin expression has been reported in the dorsal root ganglia (47), as well as the trigeminal ganglion (48). Even if it is possible that the antibody used to detect amylin is detecting CGRP given the similarity between these neuropeptides (49), the latest studies have used an anti-amylin antibody with limited CGRP cross-reactivity (47).

Interestingly, signalling of amylin in the brain seems to be very dependent on sex hormones because it is reduced in the context of ovariectomy and recovered with oestradiol replacement (50). The main binding site of peripheral amylin in the brain is the area postrema, which is outside of the BBB and transmits to the nucleus tractus solitarius (NTS) and lateral parabrachial nucleus, and from there propagates to the central amygdala and bed nucleus of the stria terminalis (22). The area postrema plays a key role in the mediation of nausea (51) and the NTS and the amygdala have been implicated in migraine pathophysiology (3).

Independent of the area postrema, peripheral amylin can also bind to the arcuate nucleus of the hypothalamus where it activates pro-opiomelanocortin (POMC) and neuropeptide Y (NPY) neurons (22). Likewise, with respect to this hypothalamic nucleus, POMC and NPY have been considered as potentially implicated in the pathophysiology of migraine. Specifically, impaired NPY signalling could be implicated in stress-induced migraine (3).

No obvious relationship has been established between migraine and diabetes, which requires consideration given some of the implications of amylin with respect to both conditions. Diabetes type 2 is frequently associated with metabolic syndrome (52). Obesity has been shown to be a risk factor for progression to chronic migraine in episodic migraine patients (53) and the association between obesity and migraine has been observed more clearly in females aged younger than 55 years (54).

Of note, to investigate some of the sex differences in migraine in rodent models’ nitroglycerine (NTG)-induced neuronal activation was established more easily in females compared to males. Interestingly, in both castrated females and males, NTG activation was reduced in the nucleus trigeminalis caudalis and, only in castrated females, different brain areas, including the central amygdala, the NTS and the area postrema, showed reduced activation (55). As previously discussed, these areas are activation sites for amylin. In rodents, hypothalamic amylin expression in the hypothalamus is higher in wild-type females compared to males, although these differences reversed when the animals were exposed to a 60% high-fat diet (56). To objectively quantify pain, an automated assay looking at squint response in mice focusing on the interpalpebral fissure area was tested by administering CGRP at different doses as well as amylin. CGRP induced squint in female mice at lower doses than in males, and amylin-only caused significant squint in females, again highlighting potential sex differences with respect to pain pathways and amylin in rodents (57).

Current evidence through studies

A recent study conducted in female rats during different phases of their oestrous cycle recorded the trigeminocervical complex response to dural stimuli, which was increased with pramlintide and decreased with the amylin antagonist AC187. Pre-treatment with AC187 during the oestrogen fall phase prevented dural-evoked and pramlintide-induced responses (58).

Although a mediation by amylin was not investigated, excess abdominal fat, measured with air displacement plethysmography and ViScan, was observed in migraine patients with increased cutaneous allodynia (59). CGRP and pituitary adenylyl cyclase-activating peptide were able to provoke mechanical periorbital allodynia in mice in contrast to amylin and adrenomedullin (60).

Of note, the positive provocation study of amylin (37) would appear to be in contrast to amylin analogues being useful managing weight gain and obesity (61).

Finally, an aspect that has been largely neglected is the potential relevance of splice variants of the different receptors of the calcitonin/CGRP family because these may very well explain different affinities of the receptors to the different ligands. The CGRP receptor is not known to have any splice variants, unlike CTR with 6 and with 12.

Amylin receptors are expressed in migraine-relevant sites including the trigeminocervical complex and the hypothalamus (62).

Further studies will need to dissect the pathophysiological and clinical significance of the actions of CGRP and amylin on the different receptors of the CGRP family aiming to further understand the current clinical observations. However, it is conceivable that these aspects could be of key importance in developing personalised treatment options for migraine patients.

Amylin as a biomarker?

A case–control study conducted by Irimia et al. (63) investigated interictal levels of amylin and CGRP. The interictal data obtained suggest that amylin also plays a role in attack susceptibility because the interictal increase may lower the threshold of the activation of migraine-relevant nociceptive pathways. There is also evidence for amylin detection in the trigeminovascular system (64), which has a pivotal role in migraine pathophysiology (3).

Amylin or its receptors are speculated to have a causative role in migraine and Irimia et al. (63) looked at the levels of amylin in migraineurs compared to controls in the interictal phase. To avoid any possible discrepancies, patients who may have increased levels of amylin otherwise, such as individuals with inflammatory conditions, or those who were immunosuppressed or recently infected, were excluded. It was observed that the mean plasma levels of amylin were significantly higher in patients with chronic migraine compared to those with episodic migraine. The results indicated that amylin may act as a diagnostic marker for migraine compared to CGRP (the levels of CGRP also increased in patients with chronic migraine compared to healthy controls). The results further indicated that mean plasma levels of amylin were higher in chronic patients compared to episodic patients and again compared to healthy controls. A demonstration of the mean plasma levels of amylin reducing, as has been shown in migraine patients after treatment with humanised monoclonal CGRP antibodies (65), would be an important outcome. This would be vital to investigate with respect to amylin levels when aiming to verify the findings reported by Irimia et al. (63). Given that amylin and CGRP are related because they belong to the same family (66), an understanding of this particular peptide, as well as determining whether amylin can follow development patterns as in CGRP, would be promising.

Amylin acts similarly to CGRP given that they share receptors, and some patients who do not experience relief from CGRP antibodies may obtain some relief if the levels of amylin are reduced, indicating that some patients have possible varying mechanistic pathways. Although the findings from this particular study have been extremely significant, a future suggestion would be to recruit patients from a wider database, such as through other headache centres, university recruitment centres, etc., to allow for the sample size to better represent the affected population (63).

Future studies

The breakthrough of CGRP in migraine research opened an avenue of exploration for other areas, with one being amylin. Biomarkers could be useful in the migraine field due to a delay in diagnosis. Currently, in our practice, we cannot demonstrate solely based on an objective biomarker that patients have migraine, and so it would be useful to have a biomarker and also to verify this to the patients through some evidence to assure them of the diagnosis. The studies conducted thus far seem to follow the successful pattern of CGRP, indicating that it is important to establish that there are elevated levels of amylin interictally, as shown in the study by Irimia et al. (63). Further studies must establish this to aggregate the veracity of these findings. Moving on, the amylin analogue provoked migraine attacks (37), similarly to that shown for CGRP (30). Finally, a stimulating avenue is the targeting of amylin receptors as a novel therapeutic (49).

At present, there is no treatment option designed to selectively block the AMY₁ receptor. However, it is worth noting that medications such as erenumab and rimegepant, which target the CGRP receptor, have demonstrated an ability to also inhibit the AMY₁ receptor to a certain extent, albeit with a lower affinity compared to the CGRP receptor (41,67). Both erenumab and gepants have reported constipation as a potential side effect (68–72). It is possible that this side effect is related to the inhibition of the AMY₁ receptor; however, this side effect has also been reported for antibodies targeting CGRP ligand, fremanezumab and galcanezumab (73), and amylin receptor agonism slows gastric emptying, which is associated with constipation (74).

To establish an understanding of possible therapies targeting the amylin pathway, it is important that clinical trials are developed to potentially explore the tolerability, efficacy and safety of novel amylin-based therapies. Moreover, developing cross-over trials comparing established migraine preventives and amylin analogues may allow us to both better understand the pathophysiology of migraine and develop novel therapeutic targets. Another possible avenue to explore is the investigation of possible sex differences in the levels of amylin because it is evident that migraine is more prevalent in women (75). It would therefore be interesting to investigate how possible short- and long-lasting hormone fluctuations in women, including those occurring during menstruation, as well as after menopause, affect the levels of amylin and thus any patterns in migraine attacks. Based on the outcome, amylin targeting therapies can then be developed. Finally, a method of patient stratification could be considered. It is known that amylin plays a focal role in other disorders with comorbid migraine, and it would be useful to investigate those individuals with pre-existing amylin focused conditions such as diabetes with comorbid migraine.

Conclusions

Considering the literature, our clinical experience and the evolving field, it can be stated that amylin and amylin receptors may have an involvement in migraine pathophysiology. The existing evidence holds value in developing our understanding of the role of amylin in migraine biology. However, as a field, we are in the infancy of our understanding of this particular complex question. First and foremost, a thorough assessment of the mechanisms involved and phenotypic variability is needed. Moreover, the commonality of amylin in both migraine and diabetes can present an interesting yet complex comorbidity with respect to understanding the role of amylin alone. Collaborating with investigators in the field of endocrinology, specifically for diabetes, could give rise to clinical trials studying hormone fluctuations, as well as disorder presentation, and even the development of novel therapies for migraine. Of further interest, studying the role of amylin in migraine may also contribute to understanding more about amylin in diabetes. Importantly, there is currently a single paper detailing amylin provocation (37), which is a very clear limitation and places constraints on our understanding. However, this further emphasises the necessity for further research in thus direction. Overall, the developments made in this region are stimulating and aid progression of disorder understanding. However, the role of amylin in migraine is yet to be fully elucidated and warrants further exploration. Clinical implications

Amylin, as CGRP, is part of the calcitonin peptide family.

Amylin can trigger migraine attacks in a provocation study.

Amylin plasma levels have been suggested as a potential migraine biomarker in one study.

Some CGRP pathway treatments also target amylin receptors.