Introduction

As a neurovascular condition, migraine is classified into two main subtypes based on the presence or absence of aura, defined as visual, sensory, or neurological symptoms occurring prior to the headache (Zhang et al. 2016). Migraine attacks have a number of clinical phases. The headache is associated with trigeminal activation. However, before this phase, patients might experience symptoms such as fatigue, mood alterations, photophobia, and other symptoms (Dodick 2018). Some affected persons also have an aura phase, characterized by optical, sensory, speech, and motor defects, in addition to disturbances in higher cortical activity shortly before or concurrently with the headache (Dodick 2018). Migraine has a complicated genetic background. According to family and twin studies, the heritability of migraine is expected to be 30–60% (Honkasalo et al. 1995; Mulder et al. 2003; Polderman et al. 2015). However, at least a monogenic form of migraine has been described which is caused by mutations in the CACNA1A, ATP1A2, and SCN1A genes and is associated with hemiplegia (Sutherland et al. 2019). Generally, abnormalities in glutamatergic neurotransmission and cortical excitability have been reported to be associated with aura. In addition, a number of genetic variations within genes coding ion channels and solute carriers, or those implicated in the modulation of neurotransmitters at synaptic regions, are associated with monogenic migraine and similar conditions (Sutherland et al. 2019). Apart from these monogenic forms, migraine has also been associated with several single-nucleotide polymorphisms (SNPs) and other types of genetic variants in different genes. A number of studies have also reported aberrant expression of mRNA or microRNA (miRNA) coding genes in patients with migraine. Here we review studies evaluating the role of genetic polymorphisms and mRNA/miRNA dysregulation in migraine.

Genetic Polymorphisms in Migraine

Palmirotta et al. genotyped the insertion/deletion (I/D) polymorphism within the angiotensin-converting enzyme (ACE) gene in a cohort of Caucasian migraineurs and healthy subjects. They reported a significant association between the I/I genotype and a lower rate of preventive medication use in migraineurs with aura and also in patients with chronic migraine. Furthermore, the I/I genotype was suggestively more frequent in migraineurs with aura who did not have a family history of migraine. Collectively, the authors stated that the I/D variation in the ACE gene is not a direct susceptibility parameter in migraine, but it might affect the clinical characteristics of this disorder (Palmirotta et al. 2014). Abedin-Do et al. investigated the association between rs4343 in the ACE gene and vulnerability to migraine in an Iranian cohort. They demonstrated an association between rs4343 and migraine. They also reported over-representation of the GG genotype in patients with aura relative to those without aura (Abedin-Do et al. 2017). Fernandez et al. genotyped two possibly functional SNPs within the dopamine beta-hydroxylase (DBH) gene in two cohorts of migraineurs. They demonstrated a substantial association between the promoter marker −1021C > T and migraine in both cohorts. However, the polymorphism located in exon 11 (+1603C > T) was not associated with risk of migraine in any set of patients (Fernandez et al. 2009). Notably, the SNP located in the promoter region was previously demonstrated to explain half of the DBH activity in plasma (Zabetian et al. 2001). The association between these two SNPs and an additional SNP within this gene, namely +444G > A (rs1108580), was also assessed in a Turkish cohort. The study showed a significant association between the +1603C > T SNP but no other SNPs and migraine (Sezer et al. 2016). DBH has another genetic variant, i.e. the 19-bp I/D polymorphism. This variation has not been associated with risk of migraine or the majority of clinical parameters. However, in a subgroup of chronic migraine, patients having the D allele were more susceptible to misuse of analgesics (Barbanti et al. 2019). The rs4680 (Val158Met) SNP within the COMT gene has not been associated with migraine in female subjects. However, female subjects with the Met/Met genotype had greater levels of migraine-associated debility compared with other genotypes. Moreover, within the subgroup of female subjects with chronic migraine, the Met/Met genotype was associated with higher levels of depression and anxiety (Fernández-de-Las-Peñas et al. 2019). The association between several SNPs within the GABAergic system and susceptibility to migraine has also been assessed. Among the assessed SNPs, the GABRE rs1139916 AA genotype has tended to have a protective role in the female gender, but the differences did not remain significant after correction for multiple comparisons. It was found that the GABRQ rs3810651 might affect the age of migraine onset. Moreover, GABRA4 rs2229940 and GABRQ rs3810651 tended to be associated with the impact of ethanol on the migraine attacks, although the differences were not significant after correction for multiple comparisons (García-Martín et al. 2018). Ling et al. reported an association between the T allele of the TRPM8 rs10166942 and chronic migraine. The T allele of this SNP has also been associated with allodynic symptoms in migraineurs (Tang et al. 2019). MTHFR C677T is another polymorphism whose contribution to the risk of migraine has been assessed by numerous researchers (Fig. 1).

Fig. 1
figure 1

Homozygosity for the T allele of MTHFR C677T results in 70% reduction in the activity of this enzyme and a significant increase in homocysteine levels. An increase in this factor leads to vasodilation, enhances NMDAR activity, and increases oxidative stress, all of which increase susceptibility to migraine (Cacciapuoti 2017; Orsini et al. 2018)

Full size image

Table 1 shows the results of investigations assessing the association between genetic polymorphisms and migraine.

Table 1 The role of genetic polymorphisms in migraine
Full size table

HLA Alleles and Risk of Migraine

The association between HLA alleles and different types of migraine has been assessed by a number of researchers. For instance, Coelho et al. demonstrated no significant difference in HLA-DQB1*0602 allele frequency between migraineurs with aura and those without aura (Coelho et al. 2007). Martelletti et al. reported a similar distribution of HLA-A, HLA-B, and HLA-C antigens in migraine patients compared to healthy subjects. However, the HLA-DR2 antigen was less frequent in patients with aura than in those without aura and in healthy controls (Martelletti et al. 1999). Another study of HLA-DRB1 allele frequencies among Italian individuals revealed a lower frequency of the DRB1*12 allele in migraineurs, in spite of the higher frequency of the DRB1*16 allele in these patients as compared with healthy controls. Notably, the HLA-DRB1**16 allele frequency was higher only in the subgroup of migraineurs without aura (Rainero et al. 2005). O’Neill et al. compared the distribution of HLA-A and HLA-B alleles among different subtypes of migraineurs (classical/common migraine and migraineurs with/without a family history of this disorder). They demonstrated no significant association between these alleles and risk of migraine (O'Neill et al. 1979). Finally, Huang et al. reported a remarkable association between HLA-B and HLA-C alleles and clinic-based migraine. HLA-B*39:01, HLA-B*51:01, HLA-B*58:01, and HLA-C*03:02 were identified as risk factors for migraine. Clinic-based migraineurs who had HLA-B*58:01 or HLA-C*03:02 were more prone to chronic migraine with drug-abuse headache compared to episodic migraine. However, no HLA allele was associated with self-reported headache or migraine in the community (Huang et al. 2020). Table 2 reviews the studies assessing the contribution of HLA alleles in the development of migraine.

Table 2 The role of HLA alleles in conferring risk of migraine
Full size table

Mitochondrial DNA Changes and Polymorphisms in Migraine

The role of mitochondrial genome alterations in the development of migraine has been assessed in a few studies. Fachal et al. investigated 15 SNPs in the mitochondrial genome in patients with diverse disorders, namely Alzheimer disease, Parkinson disease, and migraine, versus controls. The authors detected no substantial association for any SNPs or haplogroup in Alzheimer disease and Parkinson patients. However, T4216C, G13708A, and haplogroup J were associated with risk of migraine, although this association was not verified in another cohort of patients (Fachal et al. 2015). Wang et al. reported a higher frequency of homoplasmic sequence variants in the nt 16040–16188 region among migraineurs without aura compared with controls (Wang et al. 2004). Guo et al. showed a higher prevalence of migraine in individuals having the 3243A > G mutation compared with controls. This observation was confirmed in both genders and both subtypes of migraine (with/without aura) (Guo et al. 2016). Zaki et al. demonstrated an association between the 16519C > T polymorphism and migraine, and the 3010G > A polymorphism was also associated with risk of migraine in persons with 16519 T (Zaki et al. 2009). A summary of studies assessing the role of mitochondrial DNA changes in migraine is presented in Table 3.

Table 3 Mitochondrial DNA changes and polymorphisms in migraine
Full size table

Altered Expression of Genes in Migraine

A few high-throughput sequencing experiments have assessed the transcriptome of migraineurs. Greco et al. measured expression levels of endocannabinoid system elements in the peripheral blood mononuclear cells of patients with episodic migraine, those with chronic migraine and drug overuse, and healthy subjects. Expression of cannabinoid receptor 1 and cannabinoid receptor 2 were higher in both groups of migraineurs compared with controls. On the other hand, expression of fatty acid amide hydrolase was downregulated in migraineurs. There were other dysregulated genes in the endocannabinoid system among migraineurs as well (Greco et al. 2020). Vgontzas et al. used single-cell RNA sequencing data obtained from different parts of the nervous system to investigate the signature of possible migraine-related genes in various cell types in these tissues. Their experiments revealed broad expression of most of migraine-associated genes. However, they also identified numerous cell-type-specific migraine-related genes as well (Vgontzas and Renthal 2020). Through a comprehensive transcriptome analysis, Kogelman et al. identified the differential expression of NMNAT2 and RETN genes between migraineurs with aura and healthy subjects; however, they could not verify these results in an independent set of patients (Kogelman et al. 2019). Table 4 reviews the results of high-throughput studies assessing RNA signatures in migraine.

Table 4 Altered expression of genes in migraine
Full size table

Expression of miRNAs in Migraine

miRNAs are small non-coding RNAs which are created via a multi-step procedure in two distinct cellular compartments, i.e. cytoplasmic and nuclear spaces. They principally attach to the 3′-UTR of complementary targets via their seeding region and degrade the transcript or suppress its translation (O'Brien et al. 2018). miRNA profiles have also been assessed in the peripheral blood of migraineurs. Andersen et al. measured serum miRNA signatures in these patients throughout migraine episodes and pain-free intervals compared with healthy subjects. The authors reported differential expression of 32 miRNAs and demonstrated a significant increase in miR-34a-5p and miR-382-5p levels during headache attacks. The former miRNA was also suggested as a biomarker for migraine based on the significant difference in its expression in the pain-free interval compared with healthy status (Andersen et al. 2016). Cheng et al. reported upregulation of miR-155, miR-126, and let-7 g in migraineurs compared with controls. Notably, expression levels of miR-155 and miR-126 in migraineurs were correlated with syncope rate in the previous year (Cheng et al. 2018). Tafuri et al. demonstrated upregulation of miR-27b and downregulation of miR-181a, let-7b, and miR-22 in migraineurs without aura compared with controls. The MiRNA signature accurately distinguished migraineurs from controls. Notably, these miRNAs were functionally related to atherosclerosis and stroke (Tafuri et al. 2015). Gallelli et al. quantified expression levels of hsa-miR-34a-5p and hsa-miR-375 in the serum and saliva of migraineurs without aura versus controls. They reported upregulation of hsa-miR-34a-5p and hsa-miR-375 in saliva samples of untreated migraineurs compared with controls. In addition, levels of these miRNAs were lower in treated migraineurs than in untreated persons (Gallelli et al. 2019). Finally, peripheral transcript levels of miR-30a were decreased in the peripheral blood of migraineurs compared with controls in association with hypermethylation of its promoter region. In addition, its expression was downregulated in subjects with bilateral seizures, insistent pain, and high pain index. miR-30a has been shown to target CALCA. Taken together, these results indicate that reduced expression of miR-30a in migraineurs could relieve migraine via suppression of CALCA (Zhai and Zhu 2018). Table 5 reviews the altered expression of miRNAs in migraine.

Table 5 Altered expression of miRNAs in migraine
Full size table

In Vivo Studies

Based on the acknowledged role of inflammatory substances and vascular inflammatory agents in the pathogenesis of migraine, Abdollahi et al. assessed IL-6 and CRP levels in migraineurs following administration of curcumin and ω-3 fatty acids. They demonstrated downregulation of both substances following treatment with ω-3 and nano-curcumin. Thus, they suggested ω-3 fatty acids and curcumin supplementation as a therapeutic modality for prevention of migraine (Abdolahi et al. 2019). These two substances had synergic effects in decreasing expression of COX-2/iNOS transcripts in the serum samples of migraineurs. In addition, the combined administration of these agents significantly decreased the rate, severity, and length of headaches (Abdolahi et al. 2018). An animal study demonstrated the effects of long-term and recurrent administration of systemic nitroglycerin in inducing the expression of CGRP in central areas and its possible role in the process of pain perception and its association with the GABAergic system (Greco et al. 2018). Another animal study revealed the role of microglial NLRP3 inflammasome stimulation in the mediation of IL-1β production and central sensitization in an animal model of migraine (He et al. 2019). In a mouse model of nitroglycerin-induced migraine, 109 genes showed nitroglycerin treatment-by-region interaction. The solute carrier family 32 member 1 (Slc32a1) and preproenkephalin (Penk) were two of these genes with reversal of expression profiles between the nitroglycerin and control groups. Erbb4 and Slc1a2 displayed constant differential expression between treatments. Notably, numerous transcription factors were found to be among the nitroglycerin-disturbed target genes (Jeong et al. 2018). Table 6 summarizes the results of in vivo studies in migraine which reported altered gene expression.

Table 6 The results of in vivo studies in migraine
Full size table

In Vitro Studies

CGRP is a neuropeptide with strong vasoactive properties and is a biomarker of trigeminal inflammation. This molecule has essential roles in diverse kinds of migraine including the pure menstrual one. Ansari et al. assessed the anti-inflammatory impact of melatonin on CGRP transcript levels, levels of inducible nitric oxide synthase (iNOS), NO, and IL-1β production in an in vitro assay. They reported the role of melatonin treatment in reducing CGRP and NO synthesis as well as iNOS activity in the peripheral blood mononuclear cells of migraineurs (Ansari et al. 2017). Another in vitro study demonstrated the role of NO in the stimulation of trigeminal ganglion neurons for production of CGRP and other migraine-associated substances, possibly through induction of GSK-3β (Yao et al. 2020). Table 7 summarizes the results of in vitro studies in this regard.

Table 7 In vitro studies in migraine
Full size table

Discussion

Migraine is a disorder with a complex genetic background. Numerous genome-wide association studies (GWAS) and single-gene association studies have reported an association between migraine and variants within neuronal and vascular related genes (Sutherland et al. 2019). Genetic variants might also influence the age of migraine onset or other clinical parameters. Although the frequency of alleles of some of these polymorphisms were not different among subgroups of migraineurs, a number of polymorphisms such as HLA-DR2 antigens were differently distributed among different subtypes of migraineurs. Therefore, these genes may signify different pathogenic events among migraine subtypes.

Although the numbers of HLA genotyping studies in migraine are few and the results are not consistent, they point to the presence of a putative immunologic background for migraine. Replicated studies in different ethnic groups are needed to verify the contribution of HLA loci in the pathophysiology of migraine.

Expression profiles of several mRNAs and miRNAs have been found to differ between migraineurs and healthy subjects. Notably, a specific miRNA signature was able to differentiate migraineurs in pain-free intervals from healthy controls, implying its role as a biomarker for migraine. Other miRNAs were also correlated with the occurrence of certain comorbid conditions in the migraineurs. There was a functional overlap between migraine-associated miRNAs and those related to atherosclerotic events, further supporting the role of vascular events in the pathogenesis of migraine.

The data presented above support dysregulation of several types of transcripts in the peripheral blood of migraineurs. However, few studies have explored the functional annotation of these genes and pathway-based analysis. A simultaneous analysis of mRNA and miRNA profiles would help in determining the functional links between dysregulated members of these transcripts and better identification of the molecular events in the pathologic course of migraine. Moreover, assessment of the functional link between migraine-associated SNPs and genes with altered expression would further facilitate the understanding of the pathology of migraine. For instance, CALCA is regarded as a risk locus for migraine, while this gene has been shown to be targeted by miR-30a, a miRNA which is dysregulated in migraineurs through epigenetic mechanisms.

Collectively, migraine is a disorder with several genetic loci, each explaining a small portion of migraine heritability. Epigenetic factors such as DNA marks can also change expression patterns of migraine-associated genes, thus participating in the pathogenesis of this condition. Future studies should focus on identifying the interplay between environmental factors and genetic/epigenetic factors in the context of migraine. The signature of all classes of non-coding RNAs should be recognized during the migraine attack and in response to environmental risk factors.