Deciphering the role of genetics and circadian rhythm in cluster headache

Sammanfattning: Cluster headache (CH) is a complex neurovascular disorder with a distinct circadian attack pattern. Although many aspects of the disease’s pathophysiology remain to be elucidated, it is likely caused by a combination of different genetic and environmental risk factors. Making use of an extensive CH biobank established by our lab, genetic material from patients and controls were screened for several single nucleotide polymorphisms (SNPs) in different candidate genes. In addition, gene expression was analyzed in fibroblast cell lines from patients and healthy controls. Using a hypothesis-free approach, a genome-wide association study (GWAS) was performed on the Swedish material as well as in a combined analysis with a CH cohort from the UK. To characterize the Swedish CH population in terms of clinical patterns and sex differences, two observational studies were conducted based on questionnaire data from CH patients. In study I, we could demonstrate a clear diurnal attack pattern for a majority of patients and that tobacco consumption delays the onset of CH. Pronounced gender differences were detected in study II, where we showed that a significantly higher proportion of female patients suffered from the chronic form of CH, had a positive family history for the disorder, and reported diurnal rhythmicity of their attacks to a larger extent than male patients. Because of evident circadian attack patterns in CH, study III-V focused on circadian rhythm genes. We found a link between one SNP in the hypocretin receptor 2 (HCRTR2) gene and the disorder, but could not confirm previously reported associations of other HCRTR2 SNPs with CH. However, a SNP in the core clock gene circadian locomotor output cycles kaput (CLOCK) was associated with CH and led to increased CLOCK gene expression. Another core clock gene, cryptochrome circadian regulator 1 (CRY1), included a variant that was less common in patients, and was more highly expressed in patients compared to controls. Alcohol, nitric oxide (NO), and calcitonin gene-related peptide (CGRP) are all vasodilators which may induce CH attacks, therefore genes connected to these molecules have been of interest in genetic studies of CH. The alcohol dehydrogenase 4 (ADH4) gene was previously linked to CH in smaller case-control studies, however in our much larger study VI, we could not confirm this association with ADH4. In study VII, we investigated SNPs in the different NO synthase (NOS) genes but could not identify a clear role for these variants in the disorder. In study VIII, we demonstrated a link between CH and a SNP in the receptor activity modifying protein 1 (RAMP1) gene, encoding a CGRP receptor component, as well as increased RAMP1 gene expression in CH patients compared to controls. The first-line prophylactic treatment for CH is verapamil, a calcium-channel blocker and vasodilator. The anoctamin 3 (ANO3) gene encodes for a calcium-activated ion channel, and in study IX we found an association between an ANO3 gene variant and CH. Previous GWAS on migraine have yielded two interesting SNPs in the Swedish migraine population. In study X, we reported that the variant in the metadherin (MTDH) gene was also associated with CH, while the variant in the PR/SET domain 16 (PRDM16) gene was migraine-specific. The first GWAS on CH was performed on a very small Italian cohort, andin study XI, we could not confirm the findings for PACAP receptor 1 (ADCYAP1R1), membrane metalloendopeptidase (MME), and a 14q21 variant. When performing a GWAS on our Swedish CH material in study XII, we detected two significant loci near the genes MER proto-oncogene, tyrosine kinase (MERTK) and special AT-rich sequence-binding protein 2 (SATB2), which could be consolidated in a UK CH cohort. These studies demonstrate an involvement of the circadian rhythm in the pathophysiology of CH, and revealed some possibly dysregulated pathways in relation to treatment of CH. The GWAS findings underline that there is a genetic component to CH which needs to be investigated further.