Skip to main content
NCBI home page
Molecular Genetics & Genomic Medicine logoLink to Molecular Genetics & Genomic Medicine
. 2024 Aug 14;12(8):e2503. doi: 10.1002/mgg3.2503

Association between MAPK and PI3K/Akt signaling pathway‐related gene polymorphisms and migraine

Mingxue Wang 1, Yujia Gu 2, Shuhan Meng 1, Lixin Kang 1, Jing Yang 3, Degang Sun 3, Yuxing Liu 4, Ze Wan 5, Yi Shan 6, Dongjie Xue 7, Chang Su 8, Shufen Li 9, RanYan 1, Yu Liu 2, Yonghui Pan 10,✉, Yashuang Zhao 1,✉
PMCID: PMC11323340  PMID: 39140707

Abstract

Background

The causes of migraine remain unclear. Evidence suggests that the MAPK and PI3K/Akt signaling pathways play a role in migraine pathogenesis. However, studies on genetic polymorphisms in the two pathways associated with migraine are still limited.

Methods

This study included 226 migraineurs and 452 age‐ and sex‐matched nonmigraine control individuals. Genotyping of 31 Single Nucleotide Polymorphisms (SNPs) in 21 genes was performed. The relationship between migraine and gene polymorphisms was analyzed by using logistic regression. SNP–SNP interactions were examined by a generalized multifactor dimension reduction (GMDR) approach. The possible role of SNPs was evaluated with gene expression data from the GTEx database.

Results

The RASGRP2‐rs2230414 GT genotype was associated with decreased migraine risk compared with the wild‐type GG genotype [ORadj (95% CI): 0.674(0.458–0.989)]. PIK3R1‐rs3730089 was associated with migraine in the recessive model [ORadj (95% CI): 1.446(1.004–2.083)]. The CACNA1H‐rs61734410 CT genotype was associated with migraine risk [ORadj (95% CI): 1.561(1.068–2.281)]. One significant two‐way SNP–SNP interaction was found (PRKCA rs2228945‐BDNF rs6265) (p = 0.0107). Significant eQTL and sQTL signals were observed for the SNP rs2230414.

Conclusions

This is the first study to systematically reveal significant associations between MAPK and PI3K/Akt signaling pathway‐related gene polymorphisms and migraine risk.

Keywords: MAPK, migraine, PI3K/Akt, signaling pathway, single nucleotide polymorphism

This is the first systematic study to assess the association between polymorphisms of MAPK and PI3K/Akt signaling pathway‐related genes and the risk of migraine susceptibility.

graphic file with name MGG3-12-e2503-g001.jpg

1. INTRODUCTION

Migraine is the second highest cause of disability worldwide and the leading cause among women under 50, according to the 2019 Global Burden of Disease (GBD) study (Steiner et al., 2020). It has been suggested that migraine is a complicated and polygenetic disorder with a heritability of approximately 40%–70% (Sutherland & Griffiths, 2017). Traditionally, research on the molecular mechanisms underlying migraine has focused on monogenic studies. However, advances in DNA sequencing and genome‐wide association studies (GWASs) have uncovered a broader array of genetic factors and pathways, offering new insights into migraine (Gormley et al., 2016; Rasmussen et al., 2020). Despite these advances, GWAS findings have shown limited replicability. The largest GWAS to date, involving 102,084 cases and 771,257 controls, identified 123 genomic loci associated with migraine, of which 86 were novel (Hautakangas et al., 2022). This suggests that relying solely on GWAS is insufficient for a comprehensive understanding of migraine genetics.

The MAPK and PI3K/Akt signaling pathways play a vital role in migraine pathogenesis (Lin et al., 2020; Liu et al., 2021; Long et al., 2020). These pathways are involved in a variety of cellular processes, including neuronal signaling, inflammation, and cellular stress responses. The MAPK pathway is particularly implicated in peripheral and central sensitization, pain hypersensitivity, neuronal plasticity, and inflammatory responses associated with migraine (Cheng et al., 2018; Impey et al., 1999; Iwashita et al., 2013; Nakaya et al., 2016). In animal models, inhibition of the MAPK pathway has been shown to reduce inflammation and neuropathic pain (Ji et al., 2009). Despite these findings, the role of MAPK pathway gene polymorphisms in migraine remains underexplored.

Similarly, the PI3K/Akt signaling pathway regulates crucial cellular functions such as cell proliferation, apoptosis, differentiation, and glucose transport (Liu et al., 2017). Recent studies have highlighted the importance of the Akt and ERK1/2 pathways in migraine pathogenesis and neural signaling coordination (Guan et al., 2015; Matsuzaki & Darcha, 2015; Srivastava et al., 2014). Inhibition of the upstream modulator of the PI3K/Akt pathway may suppress glycogen synthase kinase (GSK)‐3β expression (Liu et al., 2017). Additionally, differentially regulated circRNAs might influence migraine development by modulating inflammation through the PI3K/Akt pathway (Lin et al., 2020). Activation of the microglial glucagon‐like peptide‐1 receptor (GLP‐1R) has been shown to inhibit central sensitization in chronic migraine via the PI3K/Akt pathway (Jing et al., 2021). Furthermore, functional enrichment analyses have implicated the focal adhesion‐PI3K‐Akt–mTOR and MAPK signaling pathways in migraine (Yang et al., 2018). However, most of these studies are in vitro or animal‐based, with limited validation in humans. Moreover, there is a scarcity of research on polymorphisms in PI3K/Akt pathway‐related genes in migraine patients.

The integration of findings from both pathways suggests a synergistic role in migraine development and progression. However, the genetic polymorphisms in MAPK and PI3K/Akt pathway‐related genes and their association with migraine risk and subtypes remain inadequately studied. This gap in knowledge highlights the need for comprehensive research to evaluate these associations systematically. Therefore, this study aims to fill this gap by selecting 21 genes related to the MAPK and PI3K/Akt pathways, based on existing literature, and systematically investigating their polymorphisms' role in genetic susceptibility to migraine, including migraine with and without aura. By conducting a case control study in a Chinese population, we aim to provide new insights into the genetic factors contributing to migraine and potentially identify novel therapeutic targets.

2. MATERIALS AND METHODS

2.1. Subjects

In this case control study, we enrolled 251 migraine patients and 600 nonmigraine control individuals. All participants were enrolled from four centers (the First Affiliated Hospital of Harbin Medical University and three hospitals of Agricultural Reclamation, Heilongjiang Province: Hong Xinglong Administration Hospital, Bao Quanling Central Hospital, and Bao Quanling Farm Hospital) in northern China. Migraine patients were recruited from the neurology clinics of each hospital, and controls were from the medical examination centers of each hospital during the same period. Migraine patients were diagnosed by two professional neurologists in the Department of Neurology according to the ICHD‐3β criteria (Headache Classification Committee of the International Headache, 2013). The patients were matched at a 1:2 frequency based on sex and age (±2). We also performed sensitivity analysis on the nonmatching data, and detailed characteristics are shown in Table S6. All subjects signed informed consent forms. The Harbin Medical University Medical Science Ethics Committee approved this study.

2.2. Anxiety, depression, and sleep quality assessment

The Self‐Rating Anxiety Scale (SAS) is a 20‐item questionnaire that was used to assess anxiety levels (Zung, 1971). The total score was calculated by multiplying the raw score by 1.25. SAS scores under 50 indicated no anxiety, and scores over 50 indicated anxiety. Every elevation of 10 in scores indicated increasing levels of anxiety (mild, moderate, and severe anxiety). The Patient Health Questionnaire‐9 (PHQ‐9) was used to assess depression levels (Kroenke et al., 2001). PHQ‐9 scores are divided into four levels: 0–4, which indicates no depression; 5–9, which indicates mild depression; 10–14, which indicates moderate depression; and over 14, which indicates moderate or severe depression. The Pittsburgh Sleep Quality Index (PSQI) was used to assess sleep quality and is a 19‐item questionnaire composed of 7 parts (Buysse et al., 1989). Each part is weighted from zero to three scores; thus, the total score is less than 21. In this study, poor sleep quality was defined as PSQI scores ≥5.

2.3. Genomic DNA extraction

First, 2 mL of peripheral blood was collected from all subjects in tubes without anticoagulants and then stored at −80°C. This was followed by genomic DNA extraction (QIAamp® DNA Blood Mini Kit; Qiagen, Germany) as instructed. Later, DNA was measured using a Nanodrop 2000 (Thermo Scientific, USA).

2.4. SNPs selection and genotyping

SNPs of MAPK and PI3K/Akt signaling pathway‐related genes were selected in CHB and CHS (Han Chinese in Beijing and Southern) from the 1000 Genomes Project based on tag SNPs with a MAF >0.1 and R 2 > 0.8 using HaploView (version 4.2) software. Another 10 SNPs were selected from the literature, for a total of 31 SNPs in 21 genes (Table S1). We selected SNPs located in the untranslated region (3′‐UTR, 5′‐UTR, introns) and exonic regions of the gene and without LD (R 2 < 0.8) as candidate SNPs. 48‐plex SNPscan kit (Shanghai, China). We randomly selected 60 samples (two of which were contaminated and not included in the analysis) for repeat genotyping. A total of 16 SNPs had 100% concordance, and the other 15 SNPs were above 94.7% (Table S2).

2.5. Statistical and bioinformatic analyses

Analysis of the HWE (Hardy–Weinberg equilibrium) for every SNP in control individuals was performed using the SNPassoc package of R (version 4.0.3) (Gonzalez et al., 2007). Logistic regression was adopted in the association analysis of gene polymorphisms in two pathways and migraine. R 4.0.3 was used for statistical analyses. p < 0.05 was considered significant. Multitest calibration was performed using Bonferroni correction (0.05/31). Haploview 4.2 software was used for LD and haplotype analyses. SNP–SNP interaction effects were determined using GMDR instruments (generalized multifactor dimensionality reduction). Power analysis was performed using PASS 15 software.

We used the GTEx data to study the tissue‐specific expression of quantitative trait loci (eQTL) and splice quantitative trait loci (sQTL) to further analyze the effect of significant SNPs on gene expression. The GTEx online database is available at https://gtexportal.org/home/, and it contains 54 human tissues from nearly 1000 individuals (24).

3. RESULTS

3.1. Characteristics of participants

A total of 226 migraine patients and 452 nonmigraine control groups were included in the study. All basic characteristics of migraine patients and control individuals from matched or unmatched data are shown in Table 1 and Table S6. Migraine and nonmigraine subjects were 42.394 ± 13.489 and 42.754 ± 13.476 years of age, respectively. No significance difference in either age or sex was found between migraines and control individuals, but the educational level, occurrence of depression and prevalence of sleep disorders between the two groups were significantly different (p < 0.001). Out of 226 migraineurs, 46 (20.354%) had migraine with aura, and 180 (79.646%) had migraine without aura. The mean ages were 41.696 ± 12.456 and 42.572 ± 13.768 years in the MA and MO groups, respectively (Table 1).

TABLE 1.

Basic demographic and clinical characteristics of the migraine and control groups.

Characteristics Migraine Control p MA MO p
N = 226 N = 452 N = 46 N = 180
Age, mean ± SD 42.394 ± 13.489 42.754 ± 13.476 0.743 41.696 ± 12.456 42.572 ± 13.768 0.678
Age (%)
≤35 99 (43.805) 198 (43.805) 0.942 21 (45.652) 78 (43.333) 0.844
35~ 54 (23.894) 106 (23.451) 10 (21.739) 44 (24.444)
50~ 47 (20.796) 89 (19.690) 11 (23.913) 36 (20.000)
60~ 26 (11.504) 59 (13.053) 4 (8.696) 22 (12.222)
Sex (%)
Male 46 (20.354) 92 (20.354) 1.000 9 (19.565) 37 (20.556) 1.000
Female 180 (79.646) 360 (79.646) 37 (80.435) 143 (79.444)
BMI (kg/m2), mean ± SD 23.153 ± 3.210 23.617 ± 3.377 0.082 22.711 ± 2.694 23.266 ± 3.327 0.240
Marital status (%)
Unmarried 78 (34.513) 179 (39.602) 0.229 15 (32.609) 63 (35.000) 0.896
Married 148 (65.487) 273 (60.398) 31 (67.391) 117 (65.000)
Nationality (%)
Han nationality 212 (93.805) 420 (92.920) 0.787 46 (100.000) 166 (92.222) 0.079
Minority 14 (6.195) 32 (7.080) 0 (0.000) 14 (7.778)
Educational level (%)
Elementary and below 24 (10.619) 18 (3.982) <0.001 6 (13.043) 18 (10.000) 0.218
Junior school 49 (21.681) 48 (10.619) 5 (10.870) 44 (24.444)
Senior school 29 (12.832) 81 (17.920) 7 (15.217) 22 (12.222)
College and above 124 (54.867) 305 (67.478) 28 (60.870) 96 (53.333)
Smoking (%)
Yes 39 (17.257) 64 (14.159) 0.344 7 (15.217) 32 (17.778) 0.848
No 187 (82.743) 388 (85.841) 39 (84.783) 148 (82.222)
Alcohol (%)
Yes 36 (15.929) 78 (17.257) 0.744 7 (15.217) 29 (16.111) 1.000
No 190 (84.071) 374 (82.743) 39 (84.783) 151 (83.889)
Exercise (%)
Yes 86 (38.053) 217 (48.009) 0.017 18 (39.130) 68 (37.778) 1.000
No 140 (61.947) 235 (51.991) 28 (60.870) 112 (62.222)
History of hypertension (%)
Yes 23 (10.177) 36 (7.965) 0.413 7 (15.217) 16 (8.889) 0.271
No 203 (89.823) 416 (92.035) 39 (84.783) 164 (91.111)
History of diabetes (%)
Yes 7 (3.097) 23 (5.088) 0.322 1 (2.174) 6 (3.333) 1.000
No 219 (96.903) 429 (94.912) 45 (97.826) 174 (96.667)
Family history of diabetes (%)
Yes 46 (20.354) 71 (15.708) 0.161 10 (21.739) 36 (20.000) 0.955
No 180 (79.646) 381 (84.292) 36 (78.261) 144 (80.000)
Family history of migraine (%)
Yes 77 (34.071) 90 (19.912) <0.001 16 (34.783) 61 (33.889) 1.000
No 149 (65.929) 362 (80.088) 30 (65.217) 119 (66.111)
SAS score, mean ± SD 36.537 ± 11.576 37.181 ± 9.691 0.472 39.120 ± 13.804 35.876 ± 10.880 0.144
SAS grade (%)
No anxiety 180 (79.646) 398 (88.053) 0.013 31 (67.391) 149 (82.778) 0.039
Mild anxiety 39 (17.257) 44 (9.735) 12 (26.087) 27 (15.000)
Moderate anxiety 7 (3.097) 10 (2.212) 3 (6.522) 4 (2.222)
PHQ‐9 score, mean ± SD 4.009 ± 4.212 2.916 ± 4.002 0.001 5.043 ± 4.376 3.744 ± 4.140 0.074
PHQ‐9 grade (%)
No depression 125 (55.310) 330 (73.009) <0.001 18 (39.130) 107 (59.444) 0.071
Mild depression 70 (30.973) 70 (15.487) 19 (41.304) 51 (28.333)
Moderate depression 21 (9.292) 39 (8.628) 6 (13.043) 15 (8.333)
Moderate to severe depression 10 (4.425) 13 (2.876) 3 (6.522) 7 (3.889)
PSQI score, mean ± SD 4.863 ± 2.838 3.916 ± 2.647 <0.001 5.652 ± 3.150 4.661 ± 2.726 0.055
PSQI grade (%)
Good sleep quality 112 (49.558) 295 (65.265) <0.001 19 (41.304) 93 (51.667) 0.276
Poor sleep quality 114 (50.442) 157 (34.735) 27 (58.696) 87 (48.333)
Open in a new tab

Note: Bold values indicate p‐value less than 0.05.

Abbreviations: MA, migraine with aura; MO, migraine without aura.

3.2. Association between gene polymorphisms and migraine susceptibility

All 31 SNPs were consistent with HWE (p > 0.05) (Table S1). The genotype distribution of rs2230414, rs61734410, rs3732033, and rs2049046 was significantly different between the migraine and control groups (Table 2). For RASGRP2‐rs2230414 (G>T), the GT genotype was associated with decreased susceptibility to migraine compared to the wild‐type GG genotype [OR (95% CI): 0.674 (0.458–0.989), p = 0.044]. rs2230414 was associated with migraine risk in recessive and additive models [ORadj (95% CI): 0.624 (0.436–0.889), p = 0.009; 1.603 (1.125–2.295), p = 0.009, respectively].

TABLE 2.

Associations between BDNF, RASGRP2, PRKCA, PIK3R1, G6PC2, and CACNA1H polymorphisms and the risk of migraine.

SNPs/alleles/genotypes/genetic models Control no. (%) Migraine no. (%) OR (95% CI) p ORadj (95% CI) p
BDNF‐rs2049046 T>A
Risk allele, A 387 (0.428) 223 (0.493) 1.301 (1.037–1.631) 0.023
TT 145 (32.080) 58 (25.664) ref ref
TA 227 (50.221) 113 (50.000) 1.244 (0.854–1.825) 0.258 1.226 (0.815–1.858) 0.331
AA 80 (17.699) 55 (24.336) 1.719 (1.086–2.724) 0.021 1.653 (1.003–2.728) 0.049
Dominant model 1.368 (0.960–1.966) 0.086 1.339 (0.911–1.985) 0.141
Recessive model 0.991 (0.720–1.364) 0.957 0.991 (0.700–1.404) 0.961
Additive model 1.009 (0.733–1.389) 0.957 1.009 (0.712–1.428) 0.961
RASGRP2‐rs2230414 G>T
Risk allele, T 322 (0.356) 171 (0.378) 1.100 (0.871–1.390) 0.425
GG 180 (39.823) 96 (42.478) ref ref
GT 222 (49.115) 89 (39.381) 0.752 (0.530–1.065) 0.109 0.674 (0.458–0.989) 0.044
TT 50 (11.062) 41 (18.142) 1.537 (0.948–2.488) 0.080 1.340 (0.786–2.275) 0.280
Dominant model 0.896 (0.648–1.240) 0.507 0.798 (0.558–1.141) 0.216
Recessive model 0.673 (0.486–0.930) 0.017 0.624 (0.436–0.889) 0.009
Additive model 1.486 (1.076–2.059) 0.017 1.603 (1.125–2.295) 0.009
PRKCA‐rs2286674 A>G
Risk allele, A 543 (0.601) 277 (0.613) 1.052 (0.835–1.326) 0.666
AA 165 (36.504) 75 (33.186) ref ref
AG 213 (47.124) 127 (56.195) 1.312 (0.926–1.867) 0.129 1.292 (0.886–1.892) 0.185
GG 74 (16.372) 24 (10.619) 0.714 (0.412–1.206) 0.216 0.815 (0.453–1.439) 0.488
Dominant model 1.648 (1.022–2.741) 0.046 1.434 (0.858–2.465) 0.179
Recessive model 1.439 (1.045–1.987) 0.026 1.367 (0.968–1.934) 0.077
Additive model 0.695 (0.503–0.957) 0.026 0.732 (0.517–1.034) 0.077
PIK3R1‐rs3730089 G>A
Risk allele, A 173 (0.191) 96 (0.212) 1.139 (0.861–1.507) 0.360
GG 297 (65.708) 135 (59.735) ref ref
GA 137 (30.310) 86 (38.053) 1.381 (0.984–1.935) 0.061 1.410 (0.976–2.036) 0.066
AA 18 (3.982) 5 (2.212) 0.611 (0.198–1.568) 0.340 0.553 (0.166–1.554) 0.291
Dominant model 1.292 (0.928–1.794) 0.128 1.305 (0.912–1.864) 0.144
Recessive model 1.412 (1.009–1.974) 0.043 1.446 (1.004–2.083) 0.047
Additive model 0.708 (0.507–0.991) 0.043 0.691 (0.480–0.996) 0.047
G6PC2‐rs3732033 T>C
Risk allele, C 78 (0.086) 51 (0.113) 1.347 (0.929–1.953) 0.116
TT 379 (83.850) 176 (77.876) ref ref
TC 68 (15.044) 49 (21.681) 1.552 (1.028–2.330) 0.035 1.370 (0.875–2.135) 0.166
CC 5 (1.106) 1 (0.442) 0.431 (0.022–2.696) 0.443 0.361 (0.017–2.597) 0.380
Dominant model 1.475 (0.983–2.200) 0.058 1.297 (0.834–2.004) 0.243
Recessive model 1.563 (1.036–2.347) 0.032 1.382 (0.883–2.152) 0.154
Additive model 0.640 (0.426–0.965) 0.032 0.724 (0.465–1.133) 0.154
CACNA1H‐rs61734410 T>C
Risk allele, C 148 (0.164) 90 (0.199) 1.270 (0.950–1.697) 0.106
TT 312 (69.027) 143 (63.274) ref ref
CT 132 (29.204) 76 (33.628) 1.256 (0.888–1.771) 0.195 1.561 (1.068–2.281) 0.021
CC 8 (1.770) 7 (3.097) 1.909 (0.658–5.418) 0.220 1.887 (0.571–6.064) 0.286
Dominant model 1.294 (0.923–1.808) 0.133 1.582 (1.093–2.290) 0.015
Recessive model 1.228 (0.870–1.728) 0.239 1.531 (1.050–2.233) 0.027
Additive model 0.814 (0.579–1.149) 0.239 0.653 (0.448–0.953) 0.027
Open in a new tab

Note: Bold type indicates p < 0.05. ORadj: adjusted for age, sex, BMI, marital status, nationality, educational level, smoking, alcohol, exercise, history of hypertension, history of diabetes, family history of diabetes, family history of migraine, SAS grade, PHQ9 grade, PSQI grade.

PIK3R1‐rs3730089 (G>A) was associated with increased susceptibility to migraine in the recessive model [ORadj (95% CI): 1.446 (1.004–2.083), p = 0.047], and associated with decreased migraine susceptibility in the additive model [ORadj (95% CI): 0.691 (0.480–0.996), p = 0.047]. The CACNA1H‐rs61734410 genotype CT frequency was higher than the TT frequency in migraine patients compared to nonmigraine subjects [ORadj (95% CI): 1.561 (1.068–2.281), p = 0.021]. rs61734410 was also related to migraine in both the dominant and recessive models [ORadj (95% CI): 1.582 (1.093–2.290), p = 0.015; 1.531 (1.050–2.233), p = 0.027, respectively].

In addition, PRKCA‐rs2286674 (A>G) was associated with increased susceptibility to migraine in the dominant and recessive models [OR (95% CI): 1.648 (1.022–2.741), p = 0.046; 1.439 (1.045–1.987), p = 0.026]. In the additive model, PRKCA‐rs2286674 was associated with decreased susceptibility to migraine [OR (95% CI): 0.695 (0.503–0.957), p = 0.026]. Compared with control individuals, the frequency of the G6PC2‐rs3732033 TC was higher than that of TT in migraine patients [OR (95% CI): 1.552 (1.028–2.330), p = 0.035]. rs3732033 was also associated with migraine in the genetic models [the recessive model was CC vs. TC + TT: OR (95% CI): 1.563 (1.036–2.347), p = 0.032; the additive model was CC + TT vs. TC: OR (95% CI): 0.640 (0.426–0.965), p = 0.032]. However, the adjusted ORs of the two SNPs (rs2286674 and rs3732033) were not significant.

Moreover, the AA genotype of BDNF‐rs2049046 (T>A) was higher in migraine patients than in nonmigraine patients when taking TT as a reference [ORadj (95% CI): 1.653 (1.003–2.728), p = 0.049]. rs2049046 A was associated with a higher risk of migraine [OR (95% CI): 1.301 (1.037–1.631), p = 0.023]. However, none of these meaningful results passed the Bonferroni correction with p < 0.0016 (0.05/31), and all associations were no longer statistically significant. Nevertheless, these meaningful results cannot be denied. The results of the other 25 SNPs are summarized in Table S3. Five SNPs (rs2230414, rs61734410, rs2049046, rs2286674, rs3732033) were associated with migraine in the sensitivity analysis (Table S7).

3.3. Haplotype analysis

Two haplotype blocks were detected in SYK and G6PC2, and relatively strong LDs were observed in SYK rs2290887‐rs2306042 (D′ = 1.000, r 2 = 0.07) and G6PC2 rs492594‐rs3732033 (D′ = 1.000, r 2 = 0.07) (Figure S1). However, we did not find any haplotypes statistically associated with migraine (Table S5) or in the unmatched data analysis (Table S9).

3.4. Gene–gene interaction and risk alleles

GMDR analysis results are presented in Table 3. One significant two‐way interaction model (PRKCA rs2228945 × BDNF rs6265) contributed to the best model, which showed a CVC of 10/10 and a testing accuracy of 0.5877 (p = 0.0107) (Figure S2). In the sensitivity analysis of the matched data, we found that a significant five‐locus model (rs1480153 × rs2279574 × rs2286674 × rs492594 × rs7342847) contributed the highest CVC of 9/10 and p of 0.0107 (Table S9; Figure S3). In addition, we investigated the relationship between the risk alleles of PPP2R2B‐rs1480153 G, DUSP6‐rs2279574 A, PRKCA‐rs2286674 A, G6PC2‐rs492594 G, and PRKCA‐rs7342847 T and susceptibility to migraine. Figure S4a shows the percentage of migraine patients and nonmigraine control subjects in every group of risk SNP alleles. Compared with subjects with zero to four risk alleles, migraine patients with nine to 10 risk alleles had the highest migraine risk. After covariate adjustment, the OR (95% CI) for migraine was 1.604 (95% CI = 1.053–2.471, p = 0.029) (Figure S4b).

TABLE 3.

GMDR analysis for the best SNP–SNP interaction models.

Locus no. Best combination Training Testing CV Sign test (p)
Accuracy Accuracy Consistency
1 rs2230414 0.5550 0.4977 6/10 3 (0.9453)
2 rs2228945rs6265 0.5988 0.5877 10/10 9 (0.0107)
3 rs2228945rs492594rs6265 0.6314 0.5186 4/10 8 (0.0547)
4 rs1480153rs16016rs7342847rs786906 0.6950 0.5353 5/10 7 (0.1719)
5 rs1480153rs16016rs2049046rs7342847rs786906 0.7765 0.4792 4/10 4 (0.8281)
Open in a new tab

Note: Bold type indicates p < 0.05. Adjusted factors: age, sex, BMI, marital status, nationality, educational level, smoking, alcohol, exercise, history of hypertension, history of diabetes, family history of diabetes, family history of migraine, SAS grade, PHQ9 grade, PSQI grade.

Abbreviations: CV, cross validation; GMDR, generalized multifactor dimension reduction.

3.5. Subgroup analysis

In addition, we found that migraine patients with or without aura had different significant SNPs, as shown in Table 4. The frequency of the PPP2R2B‐rs1480153 AG genotype was lower in migraine patients than the wild‐type genotype GG in MA compared to control individuals (ORadj = 0.425, 95% CI = 0.188–0.943). The frequency of the EGFR‐rs2227983 GG genotype was significantly lower in MA than in nonmigraine control individuals when using the AA genotype as a reference (ORadj = 0.321, 95% CI = 0.100–0.891). The frequency of the rs2227983 A allele was higher in the migraine group [OR (95% CI): 1.746 (1.116–2.730)]. CAMK1D‐rs2895526 and NGF‐rs7523831 were associated with migraine with aura risk in a dominant model [ORadj (95% CI): 4.505 (1.366–22.161) and 3.626 (1.248–14.015), respectively]. The frequency of the PRKCA‐rs7342847 CC genotype was significantly lower in MA than in nonmigraine control individuals when taking the TT genotype as a reference (ORadj = 0.216, 95% CI = 0.046–0.715). The rs7342847 T allele was associated with increased MA risk [OR (95% CI): 1.762 (1.109–2.799)]. However, four other SNPs (SYK‐rs2290887, PIK3R1‐rs3730089, G6PC2‐rs3732033 and CACNA1H‐rs61734410) were significantly associated with migraine without aura rather than migraine with aura.

TABLE 4.

Associations between PPP2R2B, EGFR, RASGRP2, SYK, CAMK1D, PIK3R1, G6PC2, CACNA1H, PRKCA, and NGF gene polymorphisms and the risk of migraine by aura.

SNPs/alleles/genotypes/genetic models Control MA MO
No. (%) OR (95% CI) p ORadj (95% CI) p No. (%) OR (95% CI) p ORadj (95% CI) p
PPP2R2B‐rs1480153 G>A
Risk allele, G 464 (0.513) 54 (0.587) 1.348 (0.873–2.079) 0.178 196 (0.544) 1.133 (0.887–1.448) 0.317
GG 124 (27.434) 19 (41.304) ref ref 54 (30.000) ref ref
AG 216 (47.788) 16 (34.783) 0.483 (0.237–0.974) 0.042 0.425 (0.188–0.943) 0.036 88 (48.889) 0.936 (0.626–1.406) 0.747 0.991 (0.641–1.540) 0.968
AA 112 (24.779) 11 (23.913) 0.641 (0.284–1.385) 0.267 0.741 (0.295–1.797) 0.512 38 (21.111) 0.779 (0.476–1.265) 0.315 0.825 (0.485–1.395) 0.475
Dominant model 1.048 (0.531–2.227) 0.897 0.829 (0.383–1.901) 0.644 1.231 (0.817–1.885) 0.329 1.205 (0.774–1.904) 0.416
Recessive model 0.583 (0.302–1.085) 0.095 0.482 (0.230–0.969) 0.045 1.045 (0.739–1.477) 0.803 1.079 (0.743–1.567) 0.689
Additive model 1.716 (0.922–3.306) 0.095 2.074 (1.032–4.350) 0.045 0.957 (0.677–1.353) 0.803 0.927 (0.638–1.345) 0.689
EGFR‐rs2227983 A>G
Risk allele, A 479 (0.530) 61 (0.663) 1.746 (1.116–2.730) 0.015 191 (0.531) 1.003 (0.785–1.281) 0.982
AA 125 (27.655) 21 (45.652) ref ref 55 (30.556) ref ref
GA 229 (50.664) 19 (41.304) 0.494 (0.254–0.954) 0.036 0.611 (0.290–1.289) 0.193 81 (45.000) 0.804 (0.536–1.210) 0.292 0.900 (0.581–1.404) 0.641
GG 98 (21.681) 6 (13.043) 0.364 (0.130–0.887) 0.036 0.321 (0.100–0.891) 0.039 44 (24.444) 1.020 (0.632–1.642) 0.934 1.089 (0.653–1.813) 0.744
Dominant model 1.846 (0.816–4.960) 0.176 2.377 (0.935–7.219) 0.092 0.856 (0.572–1.293) 0.453 0.861 (0.561–1.335) 0.498
Recessive model 0.685 (0.366–1.261) 0.229 0.878 (0.437–1.748) 0.711 0.797 (0.562–1.126) 0.199 0.866 (0.595–1.259) 0.452
Additive model 1.459 (0.793–2.735) 0.229 1.139 (0.572–2.288) 0.711 1.255 (0.888–1.778) 0.199 1.154 (0.794–1.680) 0.452
RASGRP2‐rs2230414 G>T
Risk allele, T 322 (0.356) 38 (0.413) 1.272 (0.822–1.967) 0.280 133 (0.369) 1.059 (0.822–1.365) 0.658
GG 180 (39.823) 18 (39.130) ref ref 78 (43.333) ref ref
GT 222 (49.115) 18 (39.130) 0.811 (0.408–1.612) 0.547 0.540 (0.244–1.173) 0.121 71 (39.444) 0.738 (0.506–1.075) 0.114 0.680 (0.450–1.024) 0.066
TT 50 (11.062) 10 (21.739) 2.000 (0.841–4.535) 0.103 1.470 (0.530–3.878) 0.444 31 (17.222) 1.431 (0.845–2.401) 0.178 1.182 (0.660–2.091) 0.569
Dominant model 1.029 (0.557–1.946) 0.927 0.696 (0.339–1.431) 0.320 0.865 (0.610–1.230) 0.418 0.775 (0.527–1.138) 0.192
Recessive model 0.666 (0.353–1.229) 0.199 0.479 (0.230–0.963) 0.043 0.675 (0.474–0.957) 0.028 0.651 (0.443–0.952) 0.028
Additive model 1.501 (0.814–2.836) 0.199 2.087 (1.039–4.346) 0.043 1.482 (1.045–2.111) 0.028 1.536 (1.051–2.257) 0.028
SYK‐rs2290887 G>A
Risk allele, G 724 (0.801) 70 (0.761) 0.791 (1.311–0.477) 0.363 296 (0.822) 1.150 (0.839–1.576) 0.386
GG 285 (63.053) 26 (56.522) ref ref 123 (68.333) ref ref
GA 154 (34.071) 18 (39.130) 1.281 (0.672–2.397) 0.442 1.361 (0.651–2.815) 0.407 50 (27.778) 0.752 (0.510–1.098) 0.145 0.650 (0.426–0.980) 0.042
AA 13 (2.876) 2 (4.348) 1.686 (0.255–6.551) 0.507 1.661 (0.228–7.576) 0.553 7 (3.889) 1.248 (0.459–3.123) 0.646 1.300 (0.460–3.400) 0.602
Dominant model 0.651 (0.173–4.252) 0.581 0.689 (0.158–4.903) 0.657 0.732 (0.295–1.976) 0.513 0.663 (0.256–1.853) 0.408
Recessive model 1.244 (0.657–2.303) 0.492 1.309 (0.634–2.663) 0.459 0.744 (0.506–1.083) 0.127 0.640 (0.422–0.961) 0.034
Additive model 0.804 (0.434–1.522) 0.492 0.764 (0.375–1.577) 0.459 1.344 (0.923–1.976) 0.127 1.562 (1.040–2.372) 0.034
CAMK1D‐rs2895526 T>A
Risk allele, T 497 (0.550) 56 (0.609) 1.274 (0.822–1.974) 0.279 208 (0.578) 1.121 (0.876–1.434) 0.366
TT 137 (30.310) 13 (28.261) ref ref 60 (33.333) ref ref
TA 223 (49.336) 30 (65.217) 1.418 (0.729–2.900) 0.318 1.655 (0.778–3.706) 0.203 88 (48.889) 0.901 (0.610–1.335) 0.602 0.949 (0.623–1.450) 0.806
AA 92 (20.354) 3 (6.522) 0.344 (0.077–1.102) 0.103 0.306 (0.058–1.154) 0.111 32 (17.778) 0.794 (0.476–1.308) 0.370 0.880 (0.508–1.508) 0.644
Dominant model 3.663 (1.298–15.345) 0.033 4.505 (1.366–22.161) 0.030 1.182 (0.764–1.867) 0.462 1.100 (0.687–1.794) 0.695
Recessive model 1.925 (1.035–3.710) 0.043 2.264 (1.120–4.768) 0.026 0.982 (0.695–1.388) 0.919 0.995 (0.686–1.443) 0.978
Additive model 0.519 (0.270–0.967) 0.043 0.442 (0.210–0.893) 0.026 1.018 (0.721–1.439) 0.919 1.005 (0.693–1.459) 0.978
PIK3R1‐rs3730089 G > A
Risk allele, A 173 (0.191) 17 (0.185) 0.958 (1.663–0.552) 0.878 79 (0.219) 1.188 (0.881–1.603) 0.260
GG 297 (65.708) 30 (65.217) ref ref 105 (58.333) ref ref
GA 137 (30.310) 15 (32.609) 1.084 (0.551–2.050) 0.809 1.188 (0.569–2.410) 0.638 71 (39.444) 1.466 (1.018–2.105) 0.039 1.495 (1.008–2.215) 0.045
AA 18 (3.982) 1 (2.174) 0.550 (0.030–2.813) 0.567 0.262 (0.010–2.118) 0.302 4 (2.222) 0.629 (0.179–1.730) 0.411 0.640 (0.173–1.884) 0.454
Dominant model 1.022 (0.529–1.908) 0.947 1.048 (0.511–2.091) 0.895 1.369 (0.959–1.949) 0.083 1.392 (0.948–2.040) 0.090
Recessive model 1.113 (0.568–2.095) 0.747 1.241 (0.596–2.509) 0.553 1.498 (1.043–2.145) 0.028 1.526 (1.031–2.253) 0.034
Additive model 0.899 (0.477–1.761) 0.747 0.806 (0.399–1.677) 0.553 0.668 (0.466–0.959) 0.028 0.655 (0.444–0.970) 0.034
G6PC2‐rs3732033 T>C
Risk allele, C 78 (0.086) 7 (0.076) 0.872 (1.949–0.390) 0.739 44 (0.122) 1.475 (0.999–2.177) 0.051
TT 379 (83.850) 39 (84.783) ref ref 137 (76.111) ref ref
TC 68 (15.044) 7 (15.217) 1.000 (0.396–2.201) 0.999 0.749 (0.272–1.814) 0.545 42 (23.333) 1.709 (1.104–2.622) 0.015 1.533 (0.956–2.438) 0.073
CC 5 (1.106) 0 (0.000) — — — — 1 (0.556) 0.553 (0.029–3.470) 0.591 0.410 (0.019–3.075) 0.452
Dominant model 0.932 (0.370–2.045) 0.870 0.721 (0.263–1.742) 0.493 1.630 (1.061–2.483) 0.024 1.448 (0.910–2.282) 0.114
Recessive model 1.014 (0.402–2.230) 0.975 0.758 (0.276–1.836) 0.563 1.719 (1.111–2.637) 0.014 1.544 (0.963–2.455) 0.068
Additive model 0.987 (0.448–2.490) 0.975 1.319 (0.545–3.627) 0.563 0.582 (0.379–0.900) 0.014 0.648 (0.407–1.038) 0.068
CACNA1H‐rs61734410 T>C
Risk allele, C 148 (0.164) 17 (0.185) 1.158 (0.665–2.016) 0.605 73 (0.203) 1.299 (0.952–1.773) 0.099
TT 312 (69.027) 31 (67.391) ref ref 112 (62.222) ref ref
CT 132 (29.204) 13 (28.261) 0.991 (0.487–1.914) 0.980 1.214 (0.541–2.615) 0.627 63 (35.000) 1.330 (0.916–1.921) 0.131 1.586 (1.058–2.374) 0.025
CC 8 (1.770) 2 (4.348) 2.516 (0.369–10.592) 0.256 2.228 (0.217–14.549) 0.441 5 (2.778) 1.741 (0.517–5.329) 0.340 2.193 (0.597–7.410) 0.213
Dominant model 1.078 (0.550–2.030) 0.820 1.290 (0.600–2.690) 0.504 1.353 (0.941–1.939) 0.101 1.621 (1.093–2.404) 0.016
Recessive model 0.955 (0.472–1.832) 0.893 1.192 (0.533–2.557) 0.659 1.305 (0.901–1.881) 0.155 1.548 (1.036–2.311) 0.032
Additive model 1.047 (0.546–2.121) 0.893 0.839 (0.391–1.878) 0.659 0.766 (0.532–1.109) 0.155 0.646 (0.433–0.966) 0.032
PRKCA‐rs7342847 T>C
Risk allele, T 522 (0.577) 65 (0.707) 1.762 (1.109–2.799) 0.016 217 (0.603) 1.110 (0.866–1.424) 0.409
TT 155 (34.292) 22 (47.826) ref ref 72 (40.000) ref ref
CT 212 (46.903) 21 (45.652) 0.698 (0.369–1.317) 0.265 0.480 (0.226–0.996) 0.051 73 (40.556) 0.741 (0.504–1.091) 0.128 0.725 (0.478–1.097) 0.128
CC 85 (18.805) 3 (6.522) 0.249 (0.058–0.744) 0.027 0.216 (0.046–0.715) 0.024 35 (19.444) 0.886 (0.543–1.430) 0.625 0.886 (0.523–1.484) 0.648
Dominant model 3.320 (1.174–13.920) 0.049 3.116 (1.011–13.954) 0.080 0.960 (0.624–1.501) 0.853 0.953 (0.596–1.545) 0.841
Recessive model 0.951 (0.513–1.747) 0.871 0.681 (0.333–1.358) 0.282 0.772 (0.543–1.095) 0.148 0.754 (0.515–1.100) 0.144
Additive model 1.052 (0.573–1.949) 0.871 1.468 (0.736–3.006) 0.282 1.295 (0.914–1.842) 0.148 1.326 (0.909–1.941) 0.144
NGF‐rs7523831 C>G
Risk allele, C 483 (0.534) 57 (0.620) 1.420 (0.915–2.202) 0.118 199 (0.553) 1.077 (0.843–1.377) 0.552
CC 130 (28.761) 15 (32.609) ref ref 53 (29.444) ref ref
GC 223 (49.336) 27 (58.696) 1.049 (0.545–2.090) 0.888 1.081 (0.514–2.340) 0.840 93 (51.667) 1.023 (0.687–1.533) 0.912 1.086 (0.706–1.681) 0.707
GG 99 (21.903) 4 (8.696) 0.350 (0.098–0.999) 0.070 0.289 (0.070–0.942) 0.056 34 (18.889) 0.842 (0.506–1.389) 0.505 0.788 (0.453–1.356) 0.393
Dominant model 2.945 (1.156–9.967) 0.044 3.626 (1.248–14.015) 0.033 1.204 (0.786–1.880) 0.402 1.336 (0.840–2.170) 0.230
Recessive model 1.459 (0.793–2.735) 0.229 1.565 (0.785–3.184) 0.207 1.098 (0.777–1.552) 0.597 1.197 (0.823–1.744) 0.348
Additive model 0.685 (0.366–1.261) 0.229 0.639 (0.314–1.274) 0.207 0.911 (0.644–1.287) 0.597 0.836 (0.574–1.215) 0.348
Open in a new tab

Note: Bold type indicates p < 0.05. ORadj: adjusted for age, sex, BMI, marital status, nationality, educational level, smoking, alcohol, exercise, history of hypertension, history of diabetes, family history of diabetes, family history of migraine, SAS grade, PHQ9 grade, PSQI grade.

Abbreviations: MA, migraine with aura; MO, migraine without aura.

In the sensitivity analysis, we obtained comparable findings, as shown in Table S8. We found that four SNPs (PPP2R2B‐rs1480153, BDNF‐rs2049046, EGFR‐rs2227983, CAMK1D‐rs2895526) were significantly associated with migraine with aura, while three other SNPs (G6PC2‐rs3732033, PRKCA‐rs2286674, CACNA1H‐rs61734410) were associated with migraine without aura. Importantly, RASGRP2‐rs2230414 was associated with both MA and MO in matched and unmatched data analysis.

3.6. Significant eQTL and sQTL signals

Significant eQTL signals were identified for the rs2230414, rs61734410, rs1342847, and rs1480153 SNPs (Tables S11–S14). The threshold of p values was 0.001 (0.05/49 or 0.05/44). Multiple significant signals (31/49) were identified for the rs61734410 SNP and gene expression levels of PRKCA (Figure 1b,c; Table S13). Four significant signals for rs61734410 were found. As shown in Figure 1a,d, the rs2230414 T allele was correlated with increased RASGRP2 expression levels in neuro‐tibial (p = 8.60 × 10−6), and the rs1480153 G allele increased PPP2R2B gene expression in the brain‐spinal cord (p = 1.80 × 10−6). Twelve sQTL signals for rs2230414 on RASGRP2 were identified, and the rs2230414 T allele was correlated with reduced RASGRP2 levels of variable splicing in human tissues such as Nerve—Tibial (Table 5).

FIGURE 1.

FIGURE 1

Open in a new tab

Violin plots. (a) Violin plot for genotypes of SNP rs2230414 on RASGRP2 gene expression levels in nerve‐tibial. (b) Violin plots for genotypes of SNP rs7342847 on PRKCA gene expression levels in cell‐cultured fibroblasts. (c) Violin plots for genotypes of SNP rs7342847 on PRKCA gene expression levels in nerve‐tibial. (d) Violin plot for genotypes of SNP rs1480153 on PPP2R2B gene expression levels in brain‐spinal cord (cervical c‐1).

TABLE 5.

Significant sQTL signals of SNP rs2230414 on RASGRP2.

Gene Ref_Alt alleles SNP p‐value NES Tissue
RASGRP2 G_T rs2230414 2.20E‐21 −0.67 Whole Blood
RASGRP2 G_T rs2230414 1.70E‐16 −0.68 Artery—Tibial
RASGRP2 G_T rs2230414 1.30E‐08 −0.52 Lung
RASGRP2 G_T rs2230414 1.90E‐08 −0.6 Artery—Aorta
RASGRP2 G_T rs2230414 2.30E‐08 0.53 Heart—Atrial Appendage
RASGRP2 G_T rs2230414 1.10E‐07 −0.47 Esophagus—Muscularis
RASGRP2 G_T rs2230414 1.40E‐07 −0.48 Adipose—Visceral (Omentum)
RASGRP2 G_T rs2230414 7.10E‐07 −0.43 Thyroid
RASGRP2 G_T rs2230414 0.0000017 −0.6 Spleen
RASGRP2 G_T rs2230414 0.000002 −0.38 Nerve—Tibial
RASGRP2 G_T rs2230414 0.0000037 −0.5 Esophagus—Gastroesophageal Junction
RASGRP2 G_T rs2230414 0.0000039 −0.52 Artery—Coronary
Open in a new tab

Abbreviation: Ref_Alt Alleles, reference and alternative alleles.

3.7. GRS analysis

Based on the 12 SNPs associated with migraine, MA, or MO, we performed GRS analysis. The GRS scores in the migraine groups were significantly higher than those in the control individuals (11.407 ± 2.140 vs.10.664 ± 2.281, p <0.001). Compared with Q1 (4–9) group, the GRS scores in the Q2 (10 ~ 11) and Q4 (13 ~ 18) groups [ORadj (95% CI): 1.765 (1.098–2.867) and 2.433 (1.478–4.049), respectively] were significantly different, and the p for trend was 0.001 (Table 6).

TABLE 6.

Association between GRS scores and migraine.

GRS Control (N = 452) Migraine (N = 226) T or ORadj (95% CI) p
Mean ± SD 10.664 ± 2.281 11.407 ± 2.140 −4.1699 <0.001
Groups, N (%)
Q1 (4–9) 138 (0.305) 40 (0.177) ref
Q2 (10–11) 152 (0.336) 79 (0.350) 1.765 (1.098–2.867) 0.020
Q3 (12–12) 70 (0.155) 37 (0.164) 1.742 (0.982–3.090) 0.057
Q4 (13–18) 92 (0.204) 70 (0.310) 2.433 (1.478–4.049) 0.001
p for trend 0.001
Open in a new tab

Note: Total of 12 SNPs (BDNF‐rs2049046, RASGRP2‐rs2230414, PRKCA‐rs2286674, PIK3R1‐rs3730089, G6PC2‐rs3732033, CACNA1H‐rs61734410, PPP2R2B‐rs1480153, EGFR‐rs2227983, SYK‐rs2290887, CAMK1D‐rs2895526, PRKCA‐rs7342847 and NGF‐rs7523831) were included in the calculation of GRS. Bold type indicates p < 0.05. ORadj: Adjusted for age, sex, BMI, marital status, nationality, educational level, smoking, alcohol, exercise, history of hypertension, history of diabetes, family history of diabetes, family history of migraine, SAS grade, PHQ9 grade, PSQI grade.

Abbreviation: GRS, genetic risk scores.

3.8. Power analysis

We calculated the power of the study for 31 SNPs using odds ratios (ORs) of 2, 3, and 4 to assess whether the sample size of this study was sufficient to be biologically meaningful. Except for the rs1006737 and rs35737760 SNPs, the power analysis indicated that our sample size achieved sufficient power (>0.874) to conclude with OR 2.0 or higher (Table S15).

4. DISCUSSION

This is the first study to systematically assess the relationship between polymorphisms in MAPK and PI3K/Akt signaling pathway‐related genes and the risk of migraine susceptibility. We investigated 31 SNPs of 21 genes related to the MAPK and PI3K/Akt signaling pathways and attempted to reveal a possible association between the polymorphisms of genes related to the two pathways and migraine patients in China. Our findings identified novel associations of RASGRP2‐rs2230414, PIK3R1‐rs3730089, CACNA1H‐rs61734410, PRKCA‐rs2286674, and G6PC2‐rs3732033 with migraine susceptibility. These pathways are crucial for neuronal signaling, inflammation, and cellular stress responses, which are central to the pathophysiology of migraine. The MAPK pathway contributes to central sensitization, neuronal plasticity, and inflammatory responses in migraine (Cheng et al., 2018; Impey et al., 1999; Iwashita et al., 2013), while the PI3K/Akt pathway regulates neural signal transmission and promotes inflammation, driving migraine development (Guan et al., 2015; Lin et al., 2020; Matsuzaki & Darcha, 2015; Srivastava et al., 2014). Dysregulation in these pathways may lead to increased neuronal excitability and abnormal pain processing, contributing to migraine onset. By highlighting these genetic variations, our findings provide new insights into the molecular mechanisms that may underlie migraine susceptibility, thus informing potential new targets for therapeutic intervention (Table 7).

TABLE 7.

Power of the study with different odds ratios (OR) in an allelic model.

Gene SNP MAF OR = 2 OR = 3 OR = 4
BDNF rs2049046 0.449 0.989 1 1
BDNF rs6265 0.471 0.988 1 1
CACNA1A rs16016 0.326 0.986 1 1
CACNA1C rs1006737 0.053 0.625 0.966 0.998
CACNA1C rs216008 0.399 0.989 1 1
CACNA1C rs215976 0.325 0.986 1 1
CACNA1E rs35737760 0.032 0.456 0.865 0.979
CACNA1H rs1054645 0.116 0.874 0.999 1
CACNA1H rs61734410 0.179 0.950 1 1
DUSP6 rs2279574 0.382 0.989 1 1
EGFR rs2227983 0.466 0.988 1 1
ERBB4 rs11895168 0.276 0.982 1 1
FGF6 rs1024905 0.282 0.982 1 1
INSR rs1052371 0.44 0.989 1 1
NGF rs7523831 0.463 0.988 1 1
NGFR rs9908234 0.267 0.980 1 1
PRKCA rs2286674 0.404 0.989 1 1
PRKCA rs7342847 0.402 0.989 1 1
PRKCA rs2228945 0.215 0.968 1 1
RASGRP2 rs2230414 0.363 0.988 1 1
Open in a new tab

Abbreviation: MAF, minor allele frequency.

RASGRP2 is located on chromosome 11 and contains 22 exons. It activates Rap1 and R/Ras and suppresses apoptosis through the activation of the R/Ras/PI3K/Akt pathway (Takino et al., 2019). RASGRP2 is not only a protective factor for the maintenance of vascular health in the vascular endothelium but also an important signal transduction junction for αIIbβ3 integrin initiation in platelets (Canault & Alessi, 2020; Takino et al., 2021). Furthermore, bioinformatics evidence also indicates that the rs2230414 T allele is correlated with higher RASGRP2 gene expression levels and lower levels of selective splicing of the RASGRP2 gene. However, the expression level of rs2230414 and the pathogenesis of migraine are not clear. The PIK3R1 gene encodes the p85α subunit, which is located on chromosome 5. One previous study found that rs3730089 was associated with type 2 diabetes (T2DM) and schizophrenia (SCZ) risk (Huang et al., 2020; Karadoğan et al., 2018). Although there is no causal relationship between T2DM and migraine in the latest study, the two share 33 genes (Islam et al., 2022). Additional studies should be conducted on the RASGRP2 and PIK3R1 genes to unravel the underlying pathological mechanisms.

This study found that the CACNA1H‐rs61734410 polymorphism was associated with migraine susceptibility. Previous research has shown that the CACNA1E‐rs35737760 polymorphism is also associated with migraine (Ambrosini et al., 2017). Both CACNA1H and CACNA1E are important genes within the calcium channel family. Studies have found that somatic mutations in CACNA1H can genetically influence the development of aldosterone‐producing adenomas (Ambrosini et al., 2017). Given the crucial role of calcium channels in migraine, especially in familial hemiplegic migraine, we hypothesize that CACNA1H gene polymorphisms may reduce migraine risk by affecting calcium channel function (Kowalska et al., 2021). However, further research is needed to fully understand this mechanism.

PRKCA may be associated with the protective effect of EGCG against lipopolysaccharide (LPS)‐induced inflammation through the inhibition of pro‐inflammatory cytokine release by the MAPK signal pathways (Wang et al., 2021). G6PC2 is mainly expressed in pancreatic islet β‐cells. It is able to generate an ineffective substrate cycle to counteract the action of glucokinase (Hutton & O'Brien, 2009). Recently, an animal study showed that islet‐specific G6PC2 activity affects fasting glucose by modulating GSIS (glucose‐stimulated insulin secretion) sensitivity to glucose (Bosma, Rahim, Singh, et al., 2020). In addition, it has been reported that partial inhibition of G6PC2 is important in avoiding the risk of hypoglycemia (Bosma, Rahim, Oeser, et al., 2020). The relationship between blood glucose and migraine has been long established, with studies finding increased glucose uptake in the posterior white matter areas of the brain and cerebellum in migraine patients (Kassab et al., 2009). Moreover, glucose levels affect the frequency of cortical spreading depression, which consequently influences migraine development (Hoffmann et al., 2013). Therefore, PRKCA and G6PC2 gene polymorphisms are very likely to contribute to migraine by affecting inflammatory factor release or blood glucose levels.

In addition, our results indicated a statistically significant involvement of BDNF‐rs2049046 in migraine, which has been demonstrated previously (Cai et al., 2017; Sutherland et al., 2014). As pain signals, calcitonin gene‐related peptide (CGRP) and BDNF are elevated during migraine (Fischer et al., 2012). Moreover, CGRP stimulates BDNF release in neurons, which induces neuroplasticity (Buldyrev et al., 2006). The most studied polymorphisms in the BDNF gene are Val66Met (rs6265) and rs2049046. However, we did not find any association between rs6265 and migraine.

Migraines with or without aura (MA or MO) are the two major types of migraine. However, twin and genetic studies have shown that the two forms of migraine have different genetic susceptibility profiles (Pisanu et al., 2017). Our study also supports this finding. We showed that PPP2R2B‐rs1480153, PRKCA‐rs7342847, EGFR‐rs2227983, CAMK1D‐rs2895526, and NGF‐rs7523831 were associated with MA, of which the latter two SNPs (CAMK1D‐rs2895526 and NGF‐rs7523831) were previously demonstrated to be related to migraine (Li et al., 2017; Moreno‐Mayordomo et al., 2019). RASGRP2‐rs2230414 was associated with migraines both with and without aura. We first identified that SYK‐rs2290887, PIK3R1‐rs3730089, G6PC2‐rs3732033, and CACNA1H‐rs61734410 were associated with susceptibility to MO rather than MA. Furthermore, the sensitivity analysis we conducted yielded nearly consistent conclusions.

Moreover, the SNP rs1480153 G allele was associated with decreased susceptibility to MA and an increased PPP2R2B expression level. The PPP2R2B gene encodes B55β. Hypermethylation of specific cytosines of the PPP2R2B regulatory element leads to impaired upregulation of B55β (Madera‐Salcedo et al., 2019). Polymorphisms in PPP2R2B are also involved in neuropsychiatric disorders such as Alzheimer's disease and idiopathic tremor (Chen et al., 2009).

In our study, the frequency of the GG genotype of EGFR‐rs2227983 was related to a lower risk of MA. As the driver of tumorigenesis, EGFR is considered to be a biomarker of tumor drug resistance with a physiological function to regulate epithelial tissue development and homeostasis (Sigismund et al., 2018). Two studies found that the rs2227983 polymorphism was associated with several kinds of prognostic malignant cancers, such as metachronous colorectal liver metastases and breast cancer (Bekampytė et al., 2021; Zheng et al., 2018). The biological function of rs2227983 requires further investigation.

Migraine is a polygenic, multifactorial disease with a combination of genetic and environmental factors, and gene–gene interactions are becoming increasingly important. GMDR analysis showed a significant two‐way interactive model (PRKCA rs2228945 × BNDF rs6265) contributing to the best model. We found a five‐locus interaction model between rs1480153, rs2279574, rs2286674, rs492594, and rs7342847 in nonmatching data. Furthermore, we found that migraine risk increased with an increasing number of risk SNPs, indicating an additive effect of genetic polymorphisms on migraine risks. In addition, we explored the potential expression of the rs2230414, rs61734410, rs342847, and rs1480153 SNPs using eQTL and sQTL data from GTEx. Nevertheless, these data should be interpreted with caution because GTEx data are from healthy individuals. In addition, GRS analysis showed that the GRS scores were associated with a high risk of migraine susceptibility. The discovery of these gene polymorphisms enriches migraine genetic studies and may provide a reference for the personalized treatment of migraine patients.

However, all previously obtained significant SNPs failed to reach statistical significance after strict correction criteria (Bonferroni correction). Nevertheless, it cannot be denied that our findings have positive implications for the exploration of migraine. Such results are inseparably related to the insufficient sample size. This is also the major limitation of our study. In particular, the sample of the subgroup analysis was limited, and the findings need to be further validated by a large sample study. Second, the genetic polymorphisms have significant ethnic differences, and we have only studied them in Asian populations. Due to the unique lifestyle habits of people in northern China, these genetic polymorphisms may be associated with migraine exclusively within this population and have not yet been systematically investigated in other ethnic groups.

5. CONCLUSIONS

In conclusion, we identified one known SNP and found five novel SNPs associated with migraine risk in the Chinese population. Additionally, we replicated four SNPs and found two novel SNPs associated with susceptibility to migraine with aura. However, only one SNP was associated with both MA and MO, and the other four newly identified SNPs were associated with migraine without aura. Further functional and genetic association studies with larger multicenter and multiethnic samples are needed to clarify the role of these SNPs in migraine.

Further research into the role of the MAPK and PI3K/AKT pathways in the pathogenesis of migraine could provide direction for the development of new targeted therapies, offering hope for patients unresponsive to traditional treatments. Additionally, genetic screening to identify high‐risk individuals could facilitate preventive interventions, such as lifestyle modifications and prophylactic medications, thereby reducing the frequency and severity of migraine attacks and improving patients’ quality of life.

AUTHOR CONTRIBUTIONS

YZ and MW conceived the study and carried out the genetic studies. MW performed the statistical analysis and wrote this manuscript. YG coordinated and facilitated the study implementation. All authors contributed to clinical data and specimen collection and approved the final manuscript for publication.

FUNDING INFORMATION

This work was supported by the National Natural Science Foundation of China (No. 81773526).

CONFLICT OF INTEREST STATEMENT

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

ETHICS STATEMENT

Our study concerning human subjects was evaluated and approved by the Ethics Committee of Harbin Medical University (No. HMUIRB2022004) and all participants recruited signed informed consent.

Supporting information

Data S1:

MGG3-12-e2503-s001.docx (663.9KB, docx)

ACKNOWLEDGMENTS

We are grateful to all subjects who participated in the study and the clinicians and staff of four hospitals for their cooperation and assistance during the survey and sampling process. This work was supported by the National Natural Science Foundation of China (No. 81773526).

Wang, M. , Gu, Y. , Meng, S. , Kang, L. , Yang, J. , Sun, D. , Liu, Y. , Wan, Z. , Shan, Y. , Xue, D. , Su, C. , Li, S. , RanYan, Liu, Y. , Pan, Y. , & Zhao, Y. (2024). Association between MAPK and PI3K/Akt signaling pathway‐related gene polymorphisms and migraine. Molecular Genetics & Genomic Medicine, 12, e2503. 10.1002/mgg3.2503

Contributor Information

Yonghui Pan, Email: aigui1993@126.com.

Yashuang Zhao, Email: zhao_yashuang@263.net.

DATA AVAILABILITY STATEMENT

The data are available from the corresponding author upon reasonable request.

REFERENCES

  1. Ambrosini, A. , D'Onofrio, M. , Buzzi, M. G. , Arisi, I. , Grieco, G. S. , Pierelli, F. , Santorelli, F. M. , & Schoenen, J. (2017). Possible involvement of the CACNA1E gene in migraine: A search for single nucleotide polymorphism in different clinical phenotypes. Headache The Journal of Head and Face Pain, 57(7), 1136–1144. 10.1111/head.13107 [DOI] [PubMed] [Google Scholar]
  2. Bekampytė, J. , Bartnykaitė, A. , Savukaitytė, A. , Ugenskienė, R. , Korobeinikova, E. , Gudaitienė, J. , & Juozaitytė, E. (2021). The investigation of associations between rs1042522, rs2032809, rs9344, rs2227983 polymorphisms and breast cancer phenotype and prognosis. Diagnostics (Basel, Switzerland), 11, 1419. 10.3390/diagnostics11081419 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bosma, K. J. , Rahim, M. , Oeser, J. K. , McGuinness, O. P. , Young, J. D. , & O'Brien, R. M. (2020). G6PC2 confers protection against hypoglycemia upon ketogenic diet feeding and prolonged fasting. Molecular Metabolism, 41, 101043. 10.1016/j.molmet.2020.101043 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bosma, K. J. , Rahim, M. , Singh, K. , Goleva, S. B. , Wall, M. L. , Xia, J. , Syring, K. E. , Oeser, J. K. , Poffenberger, G. , McGuinness, O. P. , & Means, A. L. (2020). Pancreatic islet beta cell‐specific deletion of G6pc2 reduces fasting blood glucose. Journal of Molecular Endocrinology, 64(4), 235–248. 10.1530/JME-20-0031 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Buldyrev, I. , Tanner, N. M. , Hsieh, H.‐Y. , Dodd, E. G. , Nguyen, L. T. , & Balkowiec, A. (2006). Calcitonin gene‐related peptide enhances release of native brain‐derived neurotrophic factor from trigeminal ganglion neurons. Journal of Neurochemistry, 99(5), 1338–1350. https://pubmed.ncbi.nlm.nih.gov/17064360 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Buysse, D. J. , Reynolds, C. F., 3rd , Monk, T. H. , Berman, S. R. , & Kupfer, D. J. (1989). The Pittsburgh sleep quality index: A new instrument for psychiatric practice and research. Psychiatry Research, 28(2), 193–213. 10.1016/0165-1781(89)90047-4 [DOI] [PubMed] [Google Scholar]
  7. Cai, X. , Shi, X. , Zhang, X. , Zhang, A. , Zheng, M. , & Fang, Y. (2017). The association between brain‐derived neurotrophic factor gene polymorphism and migraine: A meta‐analysis. The Journal of Headache and Pain, 18(1), 13. 10.1186/s10194-017-0725-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Canault, M. , & Alessi, M.‐C. (2020). RasGRP2 structure, function and genetic variants in platelet pathophysiology. International Journal of Molecular Sciences, 21(3), 1075. 10.3390/ijms21031075 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chen, C.‐M. , Hou, Y.‐T. , Liu, J.‐Y. , Wu, Y.‐R. , Lin, C.‐H. , Fung, H.‐C. , Hsu, W. C. , Hsu, Y. , Lee, S. H. , Hsieh‐Li, H. M. , Su, M. T. , Chen, S. T. , Lane, H. Y. , & Lee‐Chen, G.‐J. (2009). PPP2R2B CAG repeat length in the Han Chinese in Taiwan: Association analyses in neurological and psychiatric disorders and potential functional implications. American Journal of Medical Genetics. Part B, Neuropsychiatric Genetics, 150B(1), 124–129. 10.1002/ajmg.b.30785 [DOI] [PubMed] [Google Scholar]
  10. Cheng, H. , Zhang, Y. , Lu, W. , Gao, X. , Xu, C. , & Bao, H. (2018). Caffeic acid phenethyl ester attenuates neuropathic pain by suppressing the p38/NF‐κB signal pathway in microglia. Journal of Pain Research, 11, 2709–2719. 10.2147/JPR.S166274 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Fischer, M. , Wille, G. , Klien, S. , Shanib, H. , Holle, D. , Gaul, C. , & Broessner, G. (2012). Brain‐derived neurotrophic factor in primary headaches. The Journal of Headache and Pain, 13(6), 469–475. 10.1007/s10194-012-0454-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Gonzalez, J. R. , Armengol, L. , Sole, X. , Guino, E. , Mercader, J. M. , Estivill, X. , & Moreno, V. (2007). SNPassoc: An R package to perform whole genome association studies. Bioinformatics, 23(5), 644–645. 10.1093/bioinformatics/btm025 [DOI] [PubMed] [Google Scholar]
  13. Gormley, P. , Anttila, V. , Winsvold, B. S. , Palta, P. , Esko, T. , Pers, T. H. , Farh, K. H. , Cuenca‐Leon, E. , Muona, M. , Furlotte, N. A. , Kurth, T. , Ingason, A. , McMahon, G. , Ligthart, L. , Terwindt, G. M. , Kallela, M. , Freilinger, T. M. , Ran, C. , Gordon, S. G. , … Palotie, A. (2016). Meta‐analysis of 375,000 individuals identifies 38 susceptibility loci for migraine. Nature Genetics, 48(8), 856–866. 10.1038/ng.3598 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Guan, X.‐H. , Fu, Q.‐C. , Shi, D. , Bu, H.‐L. , Song, Z.‐P. , Xiong, B.‐R. , Xu, B. , Manyande, A. , Cao, F. , & Tian, Y.‐K. (2015). Activation of spinal chemokine receptor CXCR3 mediates bone cancer pain through an Akt‐ERK crosstalk pathway in rats. Experimental Neurology, 263, 39–49. 10.1016/j.expneurol.2014.09.019 [DOI] [PubMed] [Google Scholar]
  15. Hautakangas, H. , Winsvold, B. S. , Ruotsalainen, S. E. , Bjornsdottir, G. , Harder, A. V. E. , Kogelman, L. J. A. , Thomas, L. F. , Noordam, R. , Benner, C. , Gormley, P. , Artto, V. , Banasik, K. , Bjornsdottir, A. , Boomsma, D. I. , Brumpton, B. M. , Burgdorf, K. S. , Buring, J. E. , Chalmer, M. A. , de Boer, I. , … Pirinen, M. (2022). Genome‐wide analysis of 102,084 migraine cases identifies 123 risk loci and subtype‐specific risk alleles. Nature Genetics, 54(2), 152–160. 10.1038/s41588-021-00990-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Headache Classification Committee of the International Headache Society (IHS) . (2013). The international classification of headache disorders, 3rd edition (beta version). Cephalalgia: An International Journal of Headache, 33(9), 629–808. 10.1177/0333102413485658 [DOI] [PubMed] [Google Scholar]
  17. Hoffmann, U. , Sukhotinsky, I. , Eikermann‐Haerter, K. , & Ayata, C. (2013). Glucose modulation of spreading depression susceptibility. Journal of Cerebral Blood Flow and Metabolism, 33(2), 191–195. 10.1038/jcbfm.2012.132 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Huang, J. , Chen, Z. , Zhu, L. , Wu, X. , Guo, X. , Yang, J. , Long, J. , & Su, L. (2020). Phosphoinositide‐3‐kinase regulatory subunit 1 gene polymorphisms are associated with schizophrenia and bipolar disorder in the Han Chinese population. Metabolic Brain Disease, 35(5), 785–792. 10.1007/s11011-020-00552-z [DOI] [PubMed] [Google Scholar]
  19. Hutton, J. C. , & O'Brien, R. M. (2009). Glucose‐6‐phosphatase catalytic subunit gene family. The Journal of Biological Chemistry, 284(43), 29,241–29,245. 10.1074/jbc.R109.025544 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Impey, S. , Obrietan, K. , & Storm, D. R. (1999). Making new connections: Role of ERK/MAP kinase signaling in neuronal plasticity. Neuron, 23(1), 11–14. https://pubmed.ncbi.nlm.nih.gov/10402188 [DOI] [PubMed] [Google Scholar]
  21. Islam, M. R. , The International Headache Genetics Consortium, I , & Nyholt, D. R. (2022). Genetic overlap analysis identifies a shared etiology between migraine and headache with type 2 diabetes. Genes (Basel), 13(10), 1845. 10.3390/genes13101845 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Iwashita, T. , Shimizu, T. , Shibata, M. , Toriumi, H. , Ebine, T. , Funakubo, M. , & Suzuki, N. (2013). Activation of extracellular signal‐regulated kinase in the trigeminal ganglion following both treatment of the dura mater with capsaicin and cortical spreading depression. Neuroscience Research, 77(1–2), 110–119. 10.1016/j.neures.2013.08.001 [DOI] [PubMed] [Google Scholar]
  23. Ji, R.‐R. , Gereau, R. W. , Malcangio, M. , & Strichartz, G. R. (2009). MAP kinase and pain. Brain Research Reviews, 60(1), 135–148. 10.1016/j.brainresrev.2008.12.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Jing, F. , Zou, Q. , Wang, Y. , Cai, Z. , & Tang, Y. (2021). Activation of microglial GLP‐1R in the trigeminal nucleus caudalis suppresses central sensitization of chronic migraine after recurrent nitroglycerin stimulation. The Journal of Headache and Pain, 22(1), 86. 10.1186/s10194-021-01302-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Karadoğan, A. H. , Arikoglu, H. , Göktürk, F. , İşçioğlu, F. , & İpekçi, S. H. (2018). PIK3R1 gene polymorphisms are associated with type 2 diabetes and related features in the Turkish population. Advances in Clinical and Experimental Medicine: Official Organ Wroclaw Medical University, 27(7), 921–927. 10.17219/acem/68985 [DOI] [PubMed] [Google Scholar]
  26. Kassab, M. , Bakhtar, O. , Wack, D. , & Bednarczyk, E. (2009). Resting brain glucose uptake in headache‐free migraineurs. Headache, 49(1), 90–97. 10.1111/j.1526-4610.2008.01206.x [DOI] [PubMed] [Google Scholar]
  27. Kowalska, M. , Prendecki, M. , Piekut, T. , Kozubski, W. , & Dorszewska, J. (2021). Migraine: Calcium channels and glia. International Journal of Molecular Sciences, 22(5), 2688. 10.3390/ijms22052688 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Kroenke, K. , Spitzer, R. L. , & Williams, J. B. (2001). The PHQ‐9: Validity of a brief depression severity measure. Journal of General Internal Medicine, 16(9), 606–613. 10.1046/j.1525-1497.2001.016009606.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Li, Z. , Chen, J. , Zhao, Y. , Wang, Y. , Xu, J. , Ji, J. , Shen, J. , Zhang, W. , Chen, Z. , Sun, Q. , Mao, L. , Cheng, S. , Yang, B. , Zhang, D. , Xu, Y. , Zhao, Y. , Liu, D. , Shen, Y. , Zhang, W. , … Shi, Y. (2017). Common variants in ZMIZ1 and near NGF confer risk for primary dysmenorrhoea. Nature Communications, 8(14), 900. 10.1038/ncomms14900 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Lin, J. , Shi, S. , Chen, Q. , & Pan, Y. (2020). Differential expression and Bioinformatic analysis of the circRNA expression in migraine patients. BioMed Research International, 2020, 4710780. 10.1155/2020/4710780 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Liu, M. , Fan, G. , Zhang, D. , Zhu, M. , & Zhang, H. (2021). Study on mechanism of Jiawei Chaiqin Wendan decoction in treatment of vestibular migraine based on network pharmacology and molecular docking technology. Evidence‐based Complementary and Alternative Medicine: Ecam, 2021, 5528403. 10.1155/2021/5528403 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Liu, Y.‐Y. , Jiao, Z.‐Y. , Li, W. , & Tian, Q. (2017). PI3K/AKT signaling pathway activation in a rat model of migraine. Molecular Medicine Reports, 16(4), 4849–4854. 10.3892/mmr.2017.7191 [DOI] [PubMed] [Google Scholar]
  33. Long, T. , He, W. , Pan, Q. , Zhang, S. , Zhang, D. , Qin, G. , Chen, L. , & Zhou, J. (2020). Microglia P2X4R‐BDNF signalling contributes to central sensitization in a recurrent nitroglycerin‐induced chronic migraine model. The Journal of Headache and Pain, 21(1), 4. 10.1186/s10194-019-1070-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Madera‐Salcedo, I. K. , Sánchez‐Hernández, B. E. , Svyryd, Y. , Esquivel‐Velázquez, M. , Rodríguez‐Rodríguez, N. , Trejo‐Zambrano, M. I. , García‐González, H. B. , Hernández‐Molina, G. , Mutchinick, O. M. , Alcocer‐Varela, J. , Rosetti, F. , & Crispín, J. C. (2019). PPP2R2B hypermethylation causes acquired apoptosis deficiency in systemic autoimmune diseases. JCI Insight, 5, e126457. 10.1172/jci.insight.126457 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Matsuzaki, S. , & Darcha, C. (2015). Co‐operation between the AKT and ERK signaling pathways may support growth of deep endometriosis in a fibrotic microenvironment in vitro. Human Reproduction (Oxford, England), 30(7), 1606–1616. 10.1093/humrep/dev108 [DOI] [PubMed] [Google Scholar]
  36. Moreno‐Mayordomo, R. , Ruiz, M. , Pascual, J. , Gallego de la Sacristana, M. , Vidriales, I. , Sobrado, M. , Cernuda‐Morollon, E. , Gago‐Veiga, A. B. , Garcia‐Azorin, D. , Telleria, J. J. , & Guerrero, A. L. (2019). CALCA and TRPV1 genes polymorphisms are related to a good outcome in female chronic migraine patients treated with OnabotulinumtoxinA. The Journal of Headache and Pain, 20(1), 39. 10.1186/s10194-019-0989-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Nakaya, Y. , Tsuboi, Y. , Okada‐Ogawa, A. , Shinoda, M. , Kubo, A. , Chen, J. Y. , Noma, N. , Batbold, D. , Imamura, Y. , Sessle, B. J. , & Iwata, K. (2016). ERK‐GluR1 phosphorylation in trigeminal spinal subnucleus caudalis neurons is involved in pain associated with dry tongue. Molecular Pain, 12, 1744806916641680. 10.1177/1744806916641680 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Pisanu, C. , Preisig, M. , Castelao, E. , Glaus, J. , Pistis, G. , Squassina, A. , Del Zompo, M. , Merikangas, K. R. , Waeber, G. , Vollenweider, P. , Mwinyi, J. , & Schiöth, H. B. (2017). A genetic risk score is differentially associated with migraine with and without aura. Human Genetics, 136(8), 999–1008. 10.1007/s00439-017-1816-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Rasmussen, A. H. , Kogelman, L. J. A. , Kristensen, D. M. , Chalmer, M. A. , Olesen, J. , & Hansen, T. F. (2020). Functional gene networks reveal distinct mechanisms segregating in migraine families. Brain, 143(10), 2945–2956. 10.1093/brain/awaa242 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Sigismund, S. , Avanzato, D. , & Lanzetti, L. (2018). Emerging functions of the EGFR in cancer. Molecular Oncology, 12(1), 3–20. 10.1002/1878-0261.12155 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Srivastava, S. , Sharma, K. , Kumar, N. , & Roy, P. (2014). Bradykinin regulates osteoblast differentiation by Akt/ERK/NFκB signaling axis. Journal of Cellular Physiology, 229(12), 2088–2105. 10.1002/jcp.24668 [DOI] [PubMed] [Google Scholar]
  42. Steiner, T. J. , Stovner, L. J. , Jensen, R. , Uluduz, D. , & Katsarava, Z. (2020). Migraine remains second among the world's causes of disability, and first among young women: Findings from GBD2019. The Journal of Headache and Pain, 21(1), 137. 10.1186/s10194-020-01208-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Sutherland, H. G. , & Griffiths, L. R. (2017). Genetics of migraine: Insights into the molecular basis of migraine disorders. Headache, 57(4), 537–569. 10.1111/head.13053 [DOI] [PubMed] [Google Scholar]
  44. Sutherland, H. G. , Maher, B. H. , Rodriguez‐Acevedo, A. J. , Haupt, L. M. , & Griffiths, L. R. (2014). Investigation of brain‐derived neurotrophic factor (BDNF) gene variants in migraine. Headache, 54(7), 1184–1193. 10.1111/head.12351 [DOI] [PubMed] [Google Scholar]
  45. Takino, J.‐I. , Miyazaki, S. , Nagamine, K. , & Hori, T. (2021). The role of RASGRP2 in vascular endothelial cells‐a mini review. International Journal of Molecular Sciences, 22(20), 11129. 10.3390/ijms222011129 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Takino, J.‐I. , Sato, T. , Nagamine, K. , & Hori, T. (2019). The inhibition of Bax activation‐induced apoptosis by RasGRP2 via R‐Ras‐PI3K‐Akt signaling pathway in the endothelial cells. Scientific Reports, 9(1), 16717. 10.1038/s41598-019-53419-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Wang, M. , Zhong, H. , Zhang, X. , Huang, X. , Wang, J. , Li, Z. , Chen, M. , & Xiao, Z. (2021). EGCG promotes PRKCA expression to alleviate LPS‐induced acute lung injury and inflammatory response. Scientific Reports, 11(1), 11014. 10.1038/s41598-021-90398-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Yang, Y. , Zhao, H. , Boomsma, D. I. , Ligthart, L. , Belin, A. C. , Smith, G. D. , Esko, T. , Freilinger, T. M. , Hansen, T. F. , Ikram, M. A. , Kallela, M. , Kubisch, C. , Paraskevi, C. , Strachan, D. P. , Wessman, M. , International Headache Genetics Consortium , van den Maagdenberg, A. M. J. M. , Terwindt, G. M. , & Nyholt, D. R. (2018). Molecular genetic overlap between migraine and major depressive disorder. European Journal of Human Genetics, 26(8), 1202–1216. 10.1038/s41431-018-0150-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Zheng, P. , Ren, L. , Feng, Q. , Zhu, D. , Chang, W. , He, G. , Ji, M. , Jian, M. , Lin, Q. , Yi, T. , Wei, Y. , & Xu, J. (2018). Differences in clinical characteristics and mutational pattern between synchronous and metachronous colorectal liver metastases. Cancer Management and Research, 10, 2871–2881. 10.2147/CMAR.S161392 [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Zung, W. W. (1971). A rating instrument for anxiety disorders. Psychosomatics, 12(6), 371–379. 10.1016/S0033-3182(71)71479-0 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Data S1:

MGG3-12-e2503-s001.docx (663.9KB, docx)

Data Availability Statement

The data are available from the corresponding author upon reasonable request.

Articles from Molecular Genetics & Genomic Medicine are provided here courtesy of Blackwell Publishing

RESOURCES