Abstract
The adenosine A2A receptor (A2AR) facilitates effects of calcitonin gene-related peptide and vasoactive intestinal peptide, two important neuropeptides in migraine pathophysiology, and is the molecular target of caffeine, which is used in migraine treatment. We therefore determined whether A2AR gene variation might influence migraine susceptibility Migraine patients (n = 265) with or without aura and migraine-free controls (n = 154) were assessed and genotyped for six genetic variants spanning the A2AR gene. A six-marker haplotype was more frequent in migraine patients with aura (P < 0.01) but not in patients without aura, compared with the control group. This indicates that A2AR gene variation may contribute to the pathogenesis of migraine with aura.
Introduction
Migraine is a common disorder with a major genetic component and a multifactorial, polygenic mode of inheritance. Genome-wide linkage studies have identified several loci on different chromosomes, but these results are inconsistent except for the rare forms of familial hemiplegic migraine (1, 2). For the more common forms of migraine, genes of the dopamine and serotonin neurotransmitter system have been investigated as candidate genes. In addition, genes of the adenosine system have to be considered as candidate genes, as adenosine receptors appear to be key receptors in the modulation of the action of migraine-related neuropeptides. Two highly selective adenosine A1 receptor agonists, GR-79236 and GR-190178, can inhibit trigeminovascular activation in the trigeminal nucleus and calcitonin gene-related peptide (CGRP) release in the cranial circulation (3). In contrast, the adenosine A2A receptor (A2AR) facilitates effects of CGRP, released by the terminals of the trigeminal nerve after stimulation of the trigeminal ganglion, vasoactive intestinal polypeptide (VIP), which is a marker for cranial parasympathetic nerve activation (4). VIP and CGRP are elevated in external jugular venous blood during a migraine attack and trigger the sterile neurogenic inflammation of the blood vessel walls. Further, A2AR is the molecular target of the receptor antagonist caffeine, which is effective in the treatment of migraine attacks in combination with analgesic drugs (5, 6).
Due to the central modulation of CGRP and VIP actions and the efficacy of caffeine in the treatment of migraine, we hypothesized that variation in the A2AR gene might contribute to susceptibility to migraine. To test this hypothesis, migraine patients with and without aura and migraine-free controls were compared for the frequency of six genetic variants spanning the A2AR gene.
Patients and methods
Subjects
We extended our previously described sample of 197 patients with migraine (7) to 265 patients (222 women, 43 men; age 43.7 ± 13.0 years) by recruitment from the Headache Clinics of the Universities of Würzburg and Homburg, after informed consent and approval by the local ethics committees. Patients were diagnosed according to the 2nd edition of the International Classification of Headache Disorders after completion of a standardized headache questionnaire and full neurological examination. A subgroup of 122 patients had migraine with aura (MA; 96 women, age 42.9 ± 12.7 and 26 men, age 48.2 ± 12.1 years), whereas the remaining 143 patients had migraine without aura (MoA; 126 women, age 43.3 ± 13.4 years and 17 men, age 45.0 ± 12.1 years). We did not further subdivide patients with MA into subgroups. The 154 control subjects (111 women, 43 men; age 64.7 ± 9.6 years) were negative for any history of migraine, i.e. they had never had migraine in the past or present. A complete neurological history was taken and a complete neurological physical examination of the controls was also performed. Moreover, persons with a positive family history of migraine were excluded from the control group. Finally, to exclude the possibility of any late-onset migraine, we chose to recruit control probands with a minimum age of 55 years. For all subjects depression and other comorbid psychiatric disorders were excluded by history and a score of <10 on the Beck Depression Index. All subjects were of German Caucasian origin (recorded by personal interviews).
Genotyping
DNA was isolated from venous blood samples using standard protocols. Genotyping of the validated exonic A2AR gene polymorphisms (rs5760405, rs5751876 and rs35320474; Table 1) was carried out by polymerase chain reaction (PCR)-based restriction fragment length polymorphism assays or single-strand conformation polymorphism analysis, as previously described (8, 9). Three additional polymorphisms covering the intronic and the 5′ and 3′ flanking regions (rs5751862, rs3761422 and rs4822492 from the Hapmap database rel#16, on NCBI B34 assembly, dbSNP b124, http://www.hapmap.org) were genotyped by TaqMan 5′-exonuclease assays using different coloured fluorophores for allele labelling and allelic discrimination by fluorescence signal detection (Applied Biosystems, Darmstadt, Germany). Preparation of PCR reaction mixtures was performed by Genesis Workstation RSP 150 (Tecan, Crailsheim, Germany). For PCR amplification and allelic discrimination the ABI Prism 7900 Sequence Detection System and SDS software version 2.1 (Applied Biosystems) were used. All genotypes were determined blind to the disease status.
A2AR allele and haplotype frequencies in MA and MoA patients and migraine-free controls
Values in parentheses are ratios. A2AR, adenosine A2A receptor; MA, migraine with aura; MoA, migraine without aura.
Significant after 10 000 cycles of permutation testing (P < 0.05).
Statistical analysis
Clinical symptoms and history data (frequency of nausea, vomiting, photophobia, phonophobia, unilateral pain, pulsating pain, mean intensity of pain, mean duration of one attack, age, mean age at onset and mean number of first-degree relatives with migraine) were compared between 122 MA patients and 143 MoA patients by Mann–Whitney U-test for non-parametric variables and t-test for parametric variables. Hardy–Weinberg equilibrium, linkage disequilibrium and allele-based association analyses of single markers and haplotypes were analysed by Haploview (10) version 3.2 (http://www.broad.mit.edu/mpg/haploview). Correction for multiple testing was done by the permutation method implemented in Haploview.
Results
No differences were detected between the 122 MA patients and the 143 MoA patients regarding the clinical symptoms and history data (all P-values >0.1), except for the presence or absence of aura.
All six polymorphic variants (SNPs) of the A2AR gene were in Hardy–Weinberg equilibrium (P > 0.1) and in high linkage disequlibrium (LD) (Fig. 1). This led to one haplotype block spanning all six markers and containing five different haplotype alleles with frequencies >1.0%. Of these alleles, haplotype G-C-C-C-T-G was more frequent in the MA patients (12%) than in the control group (6%, P < 0.01). In contrast, the MoA patients showed no significant difference compared with the control group. No significant difference between the three groups existed at the single-marker level (Table 1).
The six adenosine A2A receptor (A2AR) gene markers represent one block of high linkage disequlibrium (LD) (block assignment based on confidence intervals) with pairwise standardized LD coefficient r 2 > 0.16 and D' >0.93; D' values are given as numbers in boxes, except for D' = 1.0 (= empty boxes), shades of grey indicate extent of LD with darker grey = higher LD. The intermarker distances and the marker positions are indicated relative to the A2AR gene (Vertebrate Genome Annotation database, Vega v16: OTTHUMT00000030247), which contains five exons with the translated region spanning exons 4–5 (given as black bars).
Discussion
The major finding of the current study is that variation in the A2AR gene modulates susceptibility to MA. A particular haplotype is significantly associated with MA but not MoA when compared with the migraine-free control group. This finding is consistent with epidemiological twin studies indicating a higher heritability for MA compared with MoA (11). The fact that our patients were selected from specialized headache clinics, however, may limit the implications of our study. Assuming that patients with the most severe forms of migraine are mostly referred to a headache clinic, our results may not necessarily be transferrable to the general population of patients with MA.
The A2AR modulates the effect of CGRP and VIP in the brain. Thus, activation of the receptor leads to excitatory actions of CGRP on synaptic transmission in the hippocampal CA1 area (12). Second, A2AR agonists enhance the excitatory action of VIP on GABA release (4). Therefore an influence of A2AR activation may be related to migraine pathophysiology by enhancing CGRP and VIP actions after their release. Furthermore, N-methyl D-aspartate (NMDA) receptor-mediated currents are inhibited by adenosine A1 receptor activation. Activation of A2AR attenuates the inhibitory capacity of A1 receptors (3), thus facilitating NMDA receptor-mediated transmission, which is supposed to play a role in the generation of migraine aura (13). Therefore, the A2AR may also influence the genesis of the putative aura equivalent, spreading depression, by influencing signalling of ion channels. There are very few data concerning the relation between migraine aura and VIP. VIP also potentiates the effect of glutamate-mediated neurotransmission in cortical neurons, astrocytes and intraparenchymal microvessels (14). Additionally, an interesting case report discussed a new manifestation of MA due to a VIP producing left atrial myxoma and the resolution of the migraine aura after resection of the tumour (15).
Caffeine, on the other hand, is an antagonist of A2AR, and combinations of caffeine with acetylsalicylic acid and paracetamol are more effective than the combination without caffeine in the acute treatment of migraine attacks (6). It would be interesting to study whether MA patients with the G-C-C-C-T-G haplotype of the A2AR gene repond to the antimigraine effects of caffeine more favourably than others. This may form the basis of future studies on the pharmacogenomics of caffeine and migraine.
Polymorphisms of the A2AR gene have previously been investigated in certain forms of anxiety. For example, the exonic rs5751876 polymorphism of the A2AR gene was shown to be associated with panic disorder (8, 16). In contrast to patients with migraine, however, patients with panic disorder, as well as a subgroup of healthy controls, respond adversely to caffeine (9). This contrasting effect of caffeine suggests that opposite allelic variants of the A2AR gene should modify the risk of developing MA and panic disorder, respectively. Indeed, this is born out by the observation that in panic disorder and in caffeine-induced anxiety in healthy controls the observed association was with the T-allele of rs5751876 (1976C/T) (8, 9), whereas in MA an association is found with a haplotype containing the C-allele at this position. Thus, opposite alleles at A2AR gene position rs5751876 modulate the risk for MA and anxiety, which suggests a functional role for this variant.
Our results are further supported by findings in mice lacking the A2AR. Loss of A2AR function is associated with hypoalgesia, but an increase in anxiety (17). This is consistent with the notion that activation of A2AR exerts opposite effects on pain and anxiety, thus underscoring our findings in patients. As the situation in polygenic diseases is very complex and multiple genetic and psychosocial factors contribute to the pathogenesis of anxiety disorders and migraine, this opposite effect of A2AR variants relates to only one aspect of the pathophysiology of both disorders and does not suggest a dichotomy in their aetiopathogenesis, a notion which is supported by a report of increased comorbidity of anxiety and migraine (18).
While the present results are suggestive of a role of A2AR in MA, replication studies in independent, preferably family-based samples are warranted. Similarly, in vitro cell culture and in vivo animal studies are needed to define the functional consequences of the investigated A2AR gene polymorphisms. In summary, this is the first study showing that variation in the gene of the A2AR has a role in the pathophysiology of migraine with aura.
Acknowledgements
This study was supported by the Deutsche Forschungsgemeinschaft (SFB 581) and the IZKF Würzburg (N-2, 01KS9603). We gratefully acknowledge the expert technical assistance of K. Schwarte.