Response to letter to the Editor: Assessing migraine therapeutics

We appreciate the interest and discussion related to our recently published article “Migraine therapeutics differentially modulate the CGRP pathway” (1) by our colleagues, Drs. Xu and Bussiere (2). Firstly, we would like to clarify that we do not claim, but rather theorize based on the multiple marked differences reported in our study that “the diverse mechanisms of action of CGRP ligand versus receptor agents may differentially affect efficacy, safety, and/or tolerability in migraine patients” (1). We agree that there are potential limitations with in vitro studies that can complicate translation to the clinic. Preclinical investigations involving cell-based pharmacological assessments, however, are an integral part of drug discovery. Although direct translation to the clinic can be challenging, in vitro studies such as ours (1) and others (3,4), especially for a complex family like the calcitonin/CGRP family of peptides-receptors, may offer important insights into the mechanistic and pharmacological actions of migraine therapeutics.

One of the points raised by Xu and Bussiere (2) is regarding the higher agonist potency observed in the stable HEK293S-AMY₁ receptor cells used in our study compared to other reports. All the human AMY₁ receptor potency literature, which includes the references in Table 1 in Xu and Bussiere (2) and our study (1), are based on receptor component overexpression cell culture models, thus the identical caveats of cell-based systems apply. However, in contrast to other studies in which human AMY₁ receptor was transiently expressed, our study used both transient expression of human AMY₁ receptor and a stable HEK293S-AMY₁ receptor cell line. The stable cell line provided the advantage of consistent receptor levels for parallel concentration-response testing of drugs to investigate receptor binding, trafficking and signaling. A transient receptor expression system was initially used to unbiasedly test erenumab binding to the seven human calcitonin/CGRP family receptors (1). The data revealed that erenumab bound two receptors – the CGRP receptor (1; Figure 1(a)) and the AMY₁ receptor (1; Figure 5(a)). Transient transfection of the CTR/RAMP1 constructs used in Figure 5(a) resulted in pEC₅₀ values of 8.608 ± 0.04 for amylin and 9.355 ± 0.07 for calcitonin (unpublished data), which is in the range of agonist potencies at transiently expressed human AMY₁ receptor in the literature (5,6).

Although the reason for the higher agonist potency at the stable HEK293S-AMY₁ receptor cell line needs further study, we report several pieces of evidence that suggest that these cells express functional human AMY₁ receptor with similar traits, as observed in the literature, including: i) same rank order of agonist potency (5,6), ii) equivalent potencies of CGRP and amylin in stimulating cAMP responses (5) and iii) surface retention of the receptor in response to CGRP (7). Additionally, our observation of antagonism of amylin-induced signaling by telcagepant suggests that the gepant is binding to an appropriately folded and configured CTR/RAMP1 heterodimer interface of the human AMY₁ receptor (5). Thus, our study provides strong evidence that erenumab binds to both transiently and stably expressed human AMY₁ receptors. Investigations of AMY₁ receptor binding and/or antagonism in native cells/tissues may offer greater translational insights compared to transfected cell systems; however, these studies are extremely challenging for many reasons, including the lack of tools to distinguish between the CGRP and AMY₁ receptors (5), promiscuous ligand signaling (5), and the primate-favored selectivity of erenumab (8).

For comprehensive pharmacological comparison of CGRP receptor antagonists (antibody or gepants) versus CGRP ligand antibodies at the AMY₁ receptor, we agree that examining CGRP signaling will be informative. Although this was not experimentally assessed in our study, we highlight that “it is likely that CGRP ligand antibodies target CGRP activity at CGRP-family receptors that are responsive to CGRP, such as the AMY₁ receptor” (1). This unsurprising likelihood has been noted by others (3). All members of the CGRP family of peptides are agonists at the AMY₁ receptor (1,6); hence, future comparisons between CGRP/AMY₁ receptor antagonists versus CGRP ligand antibodies should also extend to calcitonin, adrenomedullin and intermedin signaling at this receptor.

There exists no convincing report that provides unequivocal evidence that erenumab does not bind to the human AMY₁ receptor. The main counter point offered by Xu and Bussiere (2) is that normalizing in vitro findings to physiological conditions should be done with caution and we are in full agreement with that point. However, the claimed and presumed selectivity of erenumab at the canonical CGRP receptor appears to be mainly based on in vitro studies involving endogenous and stably expressed CGRP receptor (8). Furthermore, it is unclear whether Xu and Bussiere (2) are of the opinion that erenumab may bind to both the human CGRP and AMY₁ receptors with potential cross-receptor pharmacology and unfortunately they do not offer any clarification. Interestingly, Xu and Bussiere emphasize that we “imply that blocking the effects of amylin on AMY1 has the potential for metabolism-related side effects, but this was not observed in clinical trials with erenumab (Aimovig® package insert)” (2). Amylin, a gut/pancreatic hormone with roles in gastric emptying and satiety, is a diabetes therapeutic target and its analogs are promising anti-obesity drug candidates (9). We highlight that “the biological relevance of the AMY₁ receptor in CGRP and/or amylin biology, or with respect to migraine and/or metabolism, is unclear” (1). However, based on the beneficial metabolic effects of pramlintide (an amylin analogue) in diabetes treatment and the unknown identity of the receptor subtype(s) for each of amylin’s physiological functions, it is reasonable to theorize that “the potential for possible metabolism-related side effects caused by the chronic blockade of the AMY₁ receptor exists” (1).

In a recent small clinical study, pramlintide induced migraine-like attacks in patients (10). The identity of the amylin receptor(s) involved is unknown, and based on in vitro studies there are four known receptors (one of which is the AMY₁ receptor) that may mediate these clinical observations (10). It has been suggested that the location of the AMY₁ receptor likely determines which hormone/neuropeptide (amylin or CGRP) may activate it (11). For example, it has been postulated that the AMY₁ receptor may be a candidate for CGRP/amylin signaling in the trigeminal ganglia in migraine pathophysiology (10,11) whereas, when expressed in other tissues/organs, the AMY₁ receptor may mediate different biological functions of amylin (11). The investigation of individual receptor responses enabled by in vitro studies has uncovered that binding to the human AMY₁ receptor is a commonality among all the current clinically approved CGRP receptor antagonists: erenumab (1), rimegepant (3) and ubrogepant (4). These findings emphasize the need to understand the physiological functions of the AMY₁ receptor and future research in this area is warranted.

The combination of in vitro pharmacological studies of CGRP pathway therapeutics like ours (1) and others (3,4), along with clinical data, may provide important insights related to the CGRP axis in migraine pathophysiology and lead to a deeper understanding of the mechanism of migraine therapeutics.