Guest: Dr. Lyn Griffiths

View the recording from our Migraine Clinician Masterclass, developed in partnership with IVPN Neuropsychiatry. In this webinar, we hear from Dr. Lyn Griffiths, who talks about Migraine Genetics. Please note that this video is intended for healthcare providers.

TRANSCRIPT

Dr. Lyn Griffiths: So, first of all, I just want to thank you very much for the opportunity to present to this Migraine Masterclass, and particularly with a focus on genetics of migraine, an area I’ve been interested in for many years. So I’m just going to start off with just a little bit of background. I’m then going to talk about what we know about migraine genetics and the ways that we can approach trying to identify the genes that play a role in migraine. But I’m also going to look at what we currently can do with diagnostics and think about some future things.

So as we’re all aware, migraine is a very common, painful disorder that affects a significant proportion of the population. There are no simple, laboratory-based diagnostics for the common forms of migraine, although there are some genetic tests available for rare and monogenic forms of migraine. As I mentioned, migraine is an extremely common disorder affecting a significant proportion of the population. And it doesn’t really matter what population you look at.

There’s been quite a number of studies across different populations around the world, and in just about all cases, you’ll find that migraine occurs much more often in females than males, and it’s usually about three times more common in females than males. You can actually see in terms of age of onset, that there are different ages when people can develop migraine. But you’ll find for males, it varies from quite a young age. But you’ll notice in females, this big peak that occurs around about puberty for the age of onset in females.

So, this clearly indicates that this is the increase when you start to see females being affected so much more by migraine. The other thing to note about migraine is that it does show comorbidity with a number of other conditions, including cardiovascular disorders such as stroke, epileptic disorders, diabetes, and also psychiatric disorders. So, it is possible that there could be shared genetic factors in there, and this could provide us a bit of a hint in terms of what we should be looking at.

In terms of migraine, we know what the symptoms are in terms of how we would diagnose it. And we also have been able to define the two types of migraine with and without aura now a lot more clearly. I think the thing about genetics though, which is particularly interesting, is that migraine does show strong familial aggregation, meaning that it runs through families, and about 90% of migraine sufferers have a close relative who also suffers from the disorder. Twin heritability estimates vary depending on which particular population you’re looking at, but it’s usually around about 60 to 65%. So meaning there’s a significant genetic component, but there are also some environmental factors and components that play a role here.

So migraine is a complex disorder. It does have environmental components as well as that significant genetic background. To make it complex too, there are the two main types of migraine, and it was thought originally that maybe there would be different genes for the different types. But it is fairly complex because you’ll find both migraine with and without aura run through families, the same families, and even the same individual can suffer from different types of migraine at different stages.

As I’ve mentioned, there’s also the comorbidity of migraine with other disorders. And in terms of genetics, when we first started working on migraine genetics – and I certainly started quite some time ago – we were kind of hoping that there might be just a number of key genes involved in the disorder. But we now know that there’s a very significant number of genes, and we certainly haven’t identified all of those. 

So at present, the number of genes is unknown. Rare and severe forms of migraine, though, and certainly for common migraine, there’s probably many genes. But for the more rare and severe forms of migraine, we do know that there are some monogenic forms, meaning that there’d be a single gene that plays a role in specific families and a causative role.

So how do you actually go about trying to identify the genes that play a role in a disorder? Well there’s two main ways to do that. So one of them is to look at case control populations. And the cases and the controls need to be carefully matched in terms of things like age, sex, but also ethnicity, so that you’re not picking up genetic variants that might be playing a role in ethnicity or that might be playing a role in the gender differences. So that’s called a case-controlled approach.

Second approach that’s often used to investigate genes for a particular disorder is to use a family-based approach, and migraine is a disorder that clearly runs through families. So this is one of our migraine families, and you can see that it’s cropping up in every generation. Occasionally, it might be missing a generation here. But in general, that could be because it may not be as severe in the person, so it may not be diagnosed in that person, or it could be really that they’re just not manifesting at that particular stage of life and may have manifested at other stages, perhaps not even diagnosed. So, it’s quite easy to actually get quite large families like these who suffer from migraine, and that way, you can then start to track genes that might be inherited or segregating with the disease from one generation to the next. 

The third way of looking…first is case control, the second way is to look at families, but the third way is to look at what’s called founder effect populations. We’ve actually been working with one from an island off the coast of Australia called Norfolk Island. Norfolk Island is what’s called an isolated founder effect population. And in fact, the majority of the people on the island can trace their ancestry back to a kind of famous historical event, and that was the Mutiny on the Bounty. 

So there was a mutiny on a ship that was traveling a long time ago, in the 1700s, it’s called the Bounty, and it was traveling from England through the Polynesian Islands and was aiming originally to get to the West Indies. Stopped around Tahiti, there was a mutiny then, and a number of captain and people were put off the boat. But the long and short of it was that Fletcher Christian, who led the mutiny, and his supporters took the boat with a number of Tahitian women that they were happy to travel with. And all of this population then went across to a small island called Pitcairn Island and proceeded to be there for many, many years.

They had children and they grew quite comfortably in terms of size, rediscovered in the 1850s. Queen Victoria then pardoned the offspring and moved them all across to the island that I first mentioned, which was Norfolk Island. Norfolk was an uninhabited island, so basically people from the 70s and 80s, 12 maternal Tahitian women, 9 paternal bounty mutineers, were the founders of the majority of the current population on Norfolk Island.

So, we’ve been working with that population since the year 2000 to undertake genetic studies. We’ve put together a big pedigree, and this is just a small component of the pedigree. We actually have a pedigree with about 6,500 people in it that goes back 12 generations to the original founders. And we’ve collected samples over about five generations, and we’ve collected them longitudinally since the year 2000, 2010, and more recently now, in 2021.

We’ve undertaken a lot of DNA studies but also phenotype studies and investigated them for CVD risk traits, ocular phenotypes, but also migraine. Interestingly, when you look at that population, we found that there’s quite a significant prevalence of migraine. About 25.5% of the population suffer from migraine on the island, which is a lot higher than what you’d expect in Australia and New Zealand, which is thought to be round about probably 12 to 15%.

It also shows significant heritability, meaning that there’s a number of genes that are playing a role in that population, on Norfolk, and this is exactly the sort of thing that you would find in an isolated founder effect population. If there are genes for disorders, and it’s kept within a closed population, you will tend to find that amplified, so there would be more people affected as it’s kept in the population. This is the third way that we’ve currently been investigating disorders, and this is a way that a lot of other people have also been investigating trying to find genes in disorders.

If we actually start looking at how you can then start to identify genes involved in a disorder like migraine, there’s two types of variants that generally play a role and that tend to be investigated when you’re looking at genetics of a disorder. The first are called rare variants. These are less than 1% in terms of their minor allele frequency.

Many of these rare variants are benign. However, if they do occur within a disease gene, within a gene, and they change something in that gene, such as an amino acid, and change the function of the protein or the effect of it in terms of gene expression, then they’re no longer benign. They may then be pathogenic, and then they can be classified as a mutation. So rare variants can be mutations that are pathogenic and can cause disorders. 

The other type of variations that can be investigated are common variants. Common variants, in the opposite way, have a minor allele frequency that’s greater than 1%. Most of the common variations that tend to be used for genetic studies are what’s called single nucleotide polymorphisms. That’s a variation at one point within the genome, where one of the nucleotides is changed from one base to another base. There are many of these – there’s approximately 10 million in the human genome – and it’s believed that there’s about one SNP every 300 base pairs. These common variants can also influence traits and also influence disease risk, so both of these can be used for studies.

When we look at the sorts of studies that you can undertake, I mentioned before that people tend to use genetic association studies when they compare cases and controls. The sorts of markers that are often used for these are those SNPs, single nucleotide polymorphisms, because they can act as biological markers to help locate the genes that are playing a role in a disease or a trait. It could be that they play a direct role if it’s a gene – it’s a variation that occurs within the regulatory sequences of a gene – or it could be that they’re actually coinherited with a causative variant. They could mark the genomic region that’s associated with the trait. 

So, SNPs are very commonly used now as genetic variants or markers to investigate disease.

There’s probably two ways that this can occur. A lot of the earlier studies focused on what was called candidate gene association studies. This means it’s investigating variations that occur within specific genes that could plausibly play a role in the disorder, that is candidate genes.

For migraine, there are many candidate gene studies focusing on things like neuronal genes, vascular pathway genes, or genes that play a role in hormonal pathways. Instead of investigating just specific SNPs within particular genes, though, a more comprehensive approach is to use a genome-wide association study. What this does is investigate SNPs, not just in one gene but across the whole genome.

That way, you can perform association tests for many markers across the genome. Because you’re testing many different markers, and many of the arrays that have been used for SNP testing now have up to about a million SNP markers on one of those arrays. Because you’re testing so many arrays, it’s really important to correct for multiple testing but also to have sufficient power to detect something significant, it’s important to have large sample sizes to investigate these for power.

In terms of studies that have been undertaken, both candidate gene and GWAS studies have been used to investigate migraine genes. There’s been quite a number of specific genes that have been implicated, and some of the ones listed up the top here, such as estrogen receptor 1 and 2, progesterone receptor, and some of the others, such as MTHFR, have been investigated as candidate genes, and shown association with either migraine with or without aura or with migraine, in general. 

There’s others that have been investigated using the genome-wide association approach. And these have involved much larger migraine populations and more recently have involved the International Headache Genetics Consortium, where a number of groups across the world have pooled their samples together to make sure that we have sufficient power to investigate appropriate numbers of cases and controls and detect significant results.

And these have implicated also a significant number of genes. Other approaches, as I mentioned, have utilized families. And one particular gene, the TRESK potassium channel gene, has been implicated using a large migraine with aura family, and this has shown complete segregation of a frameshift mutation through that family. And then finally, other approaches have investigated the more rare but more severe monogenic forms of migraine, and that has implicated a number of specific ion channel genes that I’ll talk a little bit more about shortly.

So, as I’ve mentioned, how to identify those has involved genome-wide association studies, and this is one of the earlier ones, also, large migraine family and DNA sequencing implicating a frameshift variant within the TRESK potassium channel gene. But also genome-wide association studies and these are the results of one, in 2016, that identified 38 different susceptibility gene loci for migraine. And the very recent one – that’s come out now in 2022 – again, it’s a collaboration of the International Headache Genetics Consortium has shown now 123 risk loci for migraine. So, certainly, as we increase the power and the number of cases and controls that we can investigate, the more we are detecting a number of genes that are involved in these common types of migraine.

And if you look at those genes – and there’s many, as I’ve mentioned – most of them tend to fit within either those that have neuronal or vascular function. So if you look at this diagram, this outlines some of the ones that have been implicated. Those outlined in black, such as these ones, are from genome-wide association studies of the common types of migraine. And those outlined in red tend to be using families investigating monogenic forms of migraine, where there are single causative markers within specific families.

If you look at those that were identified using association studies, such as GWAS studies, most of those associated SNPs have relatively small effect sizes, with odds ratios of about 0.9 to 1.1. However, you can start to pool some of those associated SNPs together to give a profile and what’s called a polygenic risk score. And this provides a good estimate of the effect of those genes on migraine. And what it does show is that we can identify a profile for migraine. But more importantly, what it does show is that it still doesn’t explain all of the heritability that plays a role in the common forms of migraine.

So we do need to start thinking about how we can take that forward to explain all of that heritability. Possibilities, again, are to increase the sample sizes to make sure we have well-characterized case and control populations. But perhaps we also need to be thinking about other types of genetic markers that could be playing a role in migraine.

So these could be epigenetic markers, which is not so much changes in the DNA but things that affect the expression, and I’ll come up to that a little bit later as well. But the other one is perhaps there are not just common variants but maybe rare variants that are also playing a role in the common types of migraine. So this is obviously an area that’s ongoing research in terms of migraine with and without aura, and it might take us a little while to actually start to increase those numbers. But as I’ve said, there’s now the most recent one has identified 123, and there will undoubtedly be more markers involved in migraine. 

So right across the genome, I will continue to find markers, but I want to focus a little bit more on those monogenic forms of migraine at the moment. So when we think about those, there are three well-known genes that play a role in what’s termed familial hemiplegic migraine. So there are the three corresponding types of this, familial hemiplegic migraine types 1, 2, and 3. This is a subtype of migraine with aura. When it runs through families, there are some sporadic forms, but the majority do tend to be familial. It shows autosomal dominant inheritance and also a relatively early age of onset.

So, in addition to sensory disturbances, FHM manifests with some more severe symptoms, including things like hemiparesis, deafness, and stigmas, also effects on the retina, retinal degeneration, and in some cases, even coma. It has been linked causally to mutations within three specific genes. So, the CACNA1A gene, ATP1A2, and the SCN1A gene.

So if we think about the first gene – and this one was the one that was first identified a number of years ago by Ophoff et al. – CACNA1A, we now know, has quite a number of mutations across the whole gene. And it’s different mutations in different families that play a causative role in FHM1. So these are gain-of-function mutations. They’re inherited in an autosomal dominant fashion, and they result in individuals quite often, as I’ve said, from a young age suffering from FHM1.

Interestingly, most of these are missense mutations, so single-point mutations, a change of a single base, and as I’ve mentioned, they’re usually gain-of-function. However interestingly, there’s also additional phenotypes that are due to mutations within the same gene. So episodic ataxia type 2 is also a genetic disorder.

This one’s primarily characterized by episodes of poor balance and coordination, and it usually involves not so much gain-of-function but loss-of-function mutations in the same gene. Episodes can be precipitated by some specific factors, such as trauma, stress, or particular chemical triggers, and additional symptoms include partial paralysis, altered vision, slurred speaking, etc. There’s also often an effective treatment in the form of acetazolamide. So, here’s two different disorders coming from the same gene with different types of mutations. 

And then another third phenotype, again from the same gene, is spinocerebellar ataxia type 6, or SCA6. Sufferers from this show poor coordination, speech problems, some vibration and other movement problems, and sometimes cerebellar atrophy as well. This is due to a totally different type of mutation, and the most common type of mutation is a CAG triplet repeat expansion that can lead to an extra-long polyglutamate region in the 3′ end of the gene. So, normally it’s between 4 and 18, and those affected, it goes up to 20 to 33. So, that was FHM1 gene, the CACNA1A.

So similarly, the second FHM gene, FHM2, this was identified originally a number of years ago in a number of Italian families that showed inheritance or segregation of specific mutations within the ATP1A2 and complete segregation with FHM within those families. So this one has been termed the FHM2 gene. So, as I mentioned, it’s causative of this particular type of FHM, but it’s also been associated with another phenotype, this one with alternating hemiplegia in childhood.

This typically develops at a really very young age, of 18 months of age, and includes episodes of paralysis, maybe hemiplegic, or it can affect different combinations of limbs, and episodes can last hours to days and sometimes can resolve after sleep. Also, walking balance problems can occur as well, and it can sometimes be accompanied by seizures and/or developmental delay. So again, two phenotypes with the same specific gene but with different mutation points.

And then the third FHM gene is the sodium channel gene, SCN1A, and this was first identified as causing FHM3. But interestingly, again, there’s another phenotype that potentially plays a role with this one, and this relates to epilepsy. So, as mentioned before, there’s some comorbidity between migraine and epilepsy with epilepsy patients with two or more first-degree affected relatives having a twofold increase in the risk of having migraine with aura. There’s also a syndrome of migraine with aura immediately followed by an epileptic seizure. But most importantly, and going back to the specific gene SCN1A, we know that specific mutations in this gene can cause FHM, and this one’s classified as type 3. But also mutations in the gene can be causative of Dravet syndrome as well, which is a common pediatric epilepsy.

So, it’s interesting when you start looking at FHM genes and the symptoms associated with them, so you can look at the CACNA gene and outline of the symptoms here. But mutations within this gene can cause not only FHM but also episodic ataxia type 2 and spinocerebellar ataxia type 6. ATP1A2, if you look at those symptoms, that can occur in those seizures, loss of consciousness, and mutations also occurring in alternating hemiplegia of childhood. And as I’ve mentioned, in SCN1A, number of symptoms here, and mutations can relate to FHM3 but also Dravet syndrome.

So despite these associations with the genes and with the symptoms, there’s actually considerable variability between and within families. And also in terms of diagnostics and doing things like sequencing, sometimes it can be difficult based on symptoms to determine which particular gene might be causatively involved. 

And to make matters even worse, there’s another disorder called CADASIL. This is also a dominantly inherited disorder, but it’s one of small vessels of the brain. This one occurs a bit later, usually fourth or fifth decade, can lead to cognitive impairments and also dementia. And it is the most common form of small vessel disorders. So some of the symptoms associated with this would be migraine, stroke, but also white matter lesions. And it’s usually migraine that is the first presentation of the disorder, and this can occur much younger than the cognitive symptoms.

So, it’s a fairly complex aura. It can be combined with early stroke episodes. And as I mentioned, some of this tends to overlap some of the FHM symptoms and vice versa. Interestingly enough, the gene that plays a big role in CADASIL has been identified. This is the NOTCH3 gene. And NOTCH3 plays a causative role with single mutations within this gene being inherited in that autosomal dominant manner through families. 

So if we start looking at FHM, it’s a disorder that has a prevalence of somewhere between 1 and 10,000. As I’ve mentioned, the symptoms associated with it can be variable. So, the first one is the most common ones, but there’s also 15% of people show alternating FHM and migraine with aura attacks. Quite a significant number show prolonged aura attacks. 25% show bilateral sensor motor disturbances, 20% of them show episodic, and also some of them show some associated seizures. There are a number of treatments for FHM, as well, that can relate to some of those and be helpful for individuals. 

If we start thinking about some of the symptoms associated with episodic ataxia – so again, that same gene – you can look at that, and you can see that there’s definitely overlapping symptoms between FHM1 and episodic ataxia. Also, if you start looking at CADASIL, again some of those symptoms are overlapping. But all of these have different genes involved and also potentially different treatments, particularly for things like CADASIL.

So why is that important? Why is it important to be looking at monogenic forms of migraine? Well certainly, it’s important for diagnostic purposes so that you can determine which particular disorder you’re looking at. Also, as I mentioned, there may be different treatments for different types. And also, it just makes it easier, I suppose, to differentiate whether you’re looking at migraine or other disorders as well.

So genetic testing is now available for FHM and related disorders. And in fact, our lab’s been involved in that for many years. We’ve been an accredited diagnostics lab since 1999, started off looking at the CACNA1A gene and extended diagnostics to include other genes as they’ve been identified. And we currently have been able to undertake tests and do tests for all of the disorders that I’ve mentioned here. 

In Australia, the accreditation system is what’s called NARTA accreditation, and we’ve had that accreditation since 1999. And we grow that from our research interest originally. And we also change our technology as things tend to change. So, originally we used a method called Sanger sequencing, which has always been in the past the gold standard for DNA sequencing. It’s very reliable but slow and expensive. And in 2013, we switched over to a newer method called next-generation sequencing. We were, in fact, the first to set up a next-generation sequencing accredited panel for these target disorders. And that’s been accredited, was set up in 2012, but accredited in 2013.

So just thinking about that, since we do all of the diagnostic tests for Australia and New Zealand, we have had quite a number of samples over the years sent to us through hospitals, through pathology labs, and through neurologists to investigate potential patients with these disorders. So when we did the Sanger sequencing method, we had quite low diagnostics rates. So just looking at the 700 patients that have been sent for us for diagnosis, we were able to detect only about 8% with identifiable mutations.

The methods are also, as I mentioned, time-consuming and expensive. Most people would have picked a specific gene. We tend to focus on the exons with most of the mutations that are known to be there. We then might go out to other exons. We might then move on to if that gene was negative, going to the next gene. That made things slow, but it also made things expensive too, because the method was expensive.

So what we did was we set up a method allowing us to investigate all of those genes at once, and that’s using a next-generation method. Doing that, we covered about 92% of all of those genes, and we were able to then test that on the samples that we had so we could verify and validate that it was a correct method and get it accredited. But we could also determine whether it actually helped in enabling diagnosis.

So as I mentioned, we had about 8% method when we used the Sanger older methods. But when we switched to next-generation sequencing, we got it up to 32% of FHM patients and about 20% of CADASIL patients. So overall, there was an increase of about 7.7 to 26%, so about a fourth of an increase in mutation rate. Interestingly, some of the patients that were sent to us, here’s an example of one with a potential episodic ataxia.

So we were asked to test for mutations in the CACNA gene. But because we’re now using a panel that covered all those genes, we actually detected a mutation in the ATP1A2 gene, that is FHM2 gene. Another couple of examples for case two, this is one that was referred for NOTCH3 screening for CADASIL, but it actually was a mutation in the CACNA1A gene, so doesn’t have CADASIL, actually has FHM. And then the next one here were two brothers referred for FHM screening, but they actually had mutations in the NOTCH3 gene, which are causative of CADASIL.

So, interestingly here, you can find that using a panel approach, it costs a lot less, it’s about a quarter of the previous costs, no need to do iterative testing, it’s much faster, and there’s also no need to define which specific gene that you’re interested in. Even so though, we’re still only getting about one-third of the cases diagnosed, so clearly there are mutations that we’re missing. We’ve now recently developed another panel with 15 genes. We’ve brought in some other genes that have been implicated in FHM, such as the PRRT2 gene, which has now got some accumulated multiple reports of FHM symptoms in mutation carriers.

A couple of other genes that have been causative in a number of reports for ataxias, ATP1A3, which can also cause alternating hemiplegia of childhood. And we’ve also introduced some other genes that play a role in small vessel disorders, such as HTRA1 for CARASIL, COL4A1 and 4A2, which has also played a role in similar stroke-related disorders. A couple of others that also play a role in similar stroke-related phenotypes as well. So we now have a panel that covers 15 genes. It’s got a coverage of 99.19%, and it’s also been accredited as a diagnostic service. 

So that’s a good way of if you think that you’re looking at FHM, or you’re looking potentially at CADASIL or one of the ataxias, that’s a good way of targeting and looking at those genes. But what if it’s a more complex phenotype, such as epilepsy, or you get a negative result out of the panel? Well we’ve also recently accredited a whole exon. That means that you test all different exons across the whole genome. And so, using that approach, you actually pick an awful lot more. When you start looking at all genes across the genome, you’re going to get a lot more variants that you have to investigate.

So as an example, when we look at our panel, we currently get about 60 or so variants per person. That’s not too difficult for us to investigate and see whether they have good coverage, whether they’re an artifact, whether it’s just a common polymorphism, or whether it’s a mutation. But if you want to do the whole exon – so doing all the genes – you pick up over 35,000 variants per person.

So you get a lot more things that you have to investigate, and it’s quite difficult. So what we’ve done is also set up a bioinformatic pipeline that allows us to prioritize which genes that we’re investigating and allows us to filter them and to identify which could be potentially playing a role. So we prioritize this based on how common it is. If it’s a causative mutation in a family, it’s unlikely to be a common polymorphism. So how common it is is important for us, and that’s one of our ways of calling a variant. 

The second thing is when we do curation and whether we potentially define a potential variant, we have to undertake analysis according to the American College of Medical Genetics guidelines. We look at the literature to see whether a particular variant’s been examined previously. We look at all the known databases to see if there’s been anything that mentions this particular novel or whether it might be a known variant. We also use in silico tools to determine whether the variant we’ve identified is potentially mutagenic or not, so is it likely to be pathogenic and affect function? Then we curate those based on those guidelines and provide a diagnostic result that goes out.

We’re not the only lab that has done this and does this. So as I mentioned, there are guidelines for how this is done. I’m just giving you a bit of an overview of the process that’s involved in this. So that’s the current method that we use for investigating diagnostically. But what happens if you actually have looked at all the potential known genes and you haven’t been able to pick anything from a panel or from known genes that could be potentially identified through whole exome sequencing? Well what we’ve done – and others have been doing – is undertake whole exome sequencing in those unsolved cases. Recently, we looked at 170 clinically diagnosed cases that were not positive, so they had no mutations in the known genes.

We also looked at family members of these, and we did whole exome sequencing on over 200 cases or individuals. And from that, we were able to identify a diverse range of potentially pathogenic rare variants in these particular families. Even though we didn’t pick up what was considered to be a recurrent mutation in the PRRT2 gene, we did pick up some other pathogenic mutations in this gene.

We also picked up some mutations in some other novel genes for hemiplegic migraine, including all the ones here. And basically, we added about another 10% in terms of our diagnostic hit rate into these. Most of the mutations detected in the genes are private mutations, meaning they’re a particular mutation that occurs in a specific family.

And it’s not necessarily common to a range of different probands for FHM. One particular area we’ve been focusing on recently in terms of the analysis of this data is to be looking at calcium channel genes and one, in particular, has identified rare variants in the CACNA1I gene. And we’ve shown that burden testing within this showed a statistically significant increase in variants within this gene in FHM versus controls.

And we looked at controls from gnomAD and also the UK Biobank with very significant control populations there. Since dysfunction in T-type calcium channels has been implicated in neurological conditions, we decided to then investigate these. So we did some functional studies investigating patch clamp electrophysiology in cells there.

And we found that numerous functional alterations, including reduced current density, voltage changes, and also current kinetics as well. So we’ve, in fact, quite recently published this in Frontiers in Molecular Neuroscience. But at this stage, it looks like there’s variations in this gene that could be playing a role in FHM. Whether it’s a causative role or whether it’s a modifying role, we’re not absolutely sure yet, but there’s certainly some results there that seem to imply that it’s not just the CACNA1a gene, but there may be other calcium channel genes that are playing a role in FHM. 

Our continuing studies here and other continuing studies that others are doing around the planet is to focus on diagnostic testing and extend this as we identify our variants and genes from our research and put that to full diagnostic use. We’re continuing to analyze our whole exome sequencing data to try and identify genes that play a role in FHM. We’re also undertaking similar studies in episodic ataxia, CADASIL, and some epilepsy-related disorders as well. We’re currently expanding functional analysis, but also familial segregation so testing these in families.

And I think the other thing that we really need to be thinking about in terms of genetics of migraine is that not everything will necessarily be single point mutations. We need to be looking at other methods of detecting other particular types of mutations, such as those that are not just in exons but maybe within promoters, maybe in enhancers, noncoding areas, or even things like copy number variants, structural variants. And so we’re currently looking at those areas as well.

Just towards the end here, I want to mention a couple of more things. But probably another area if we want to think about the genetics of migraine is to think about post-traumatic migraine genetics as well. And in particular, this has been sparked by information that’s made it clear that individuals with a specific CACNA1A mutation have really high sensitivity to bad outcomes from head impacts. So even minor head trauma can lead to seizures, cerebral edema, coma, and sometimes even death.

And we’ve published on this a number of years ago, but many others have as well. So this is rare, but individuals with these mutations, as I’ve mentioned, can have some quite significant outcomes from minor head trauma. So this led us to start looking at other ion channel genes in relation to things like susceptibility to concussion.

So of note, we started looking at other FHM-related genes to see if there was a potential role for this. And we found that there were notable characteristics in patients who had mutations within the ATP1A2 gene, that is the FHM2 gene. So when we looked at 172 of our patients, we found that 68% of those were due to mutations within ATP1A2.

And I think more importantly, there was an over-representation of patients in the ATP-positive group compared to the negative- group who had noted mild TBI or concussion in their clinical notes and clinical background. And this was associated with specific mutations within that gene. And so, as I’ve mentioned, looking at those particular mutations within ATP1A2, they’re associated with mild TBI-induced concussion in 33% of ATP1A2-positive patients.

So the results of that – but also the results from the CACNA gene, which was identified quite some time ago – has led us to start thinking more about whether ion channel genes could be potentially playing a role in post-traumatic migraine as well. And so we’ve currently been undertaking whole exome sequencing to see if there are rare variants in other ion channel genes that could also be playing a role. And those studies are continuing.

So towards the last minute, I would just like to mention probably two other areas that, I believe, are important to the genetics of migraine. One of these is epigenetics, as I mentioned a little bit earlier. So epigenetics involves heritable changes in gene expression. And these occur in other mechanisms other than actual DNA sequence changes. So it could be modifications within the histone proteins that are important for packing the DNA shape into chromatin. It could be variations within the histone genes, could be microRNA, or it could also be DNA methylation, which has been known to play a role in gene silencing.

So epigenetic factors are heritable. They’re changes that can affect gene function and they do not involve changes in the DNA. So probably the most commonly investigated one is DNA methylation. And we know that it’s modifiable, and it can actually change across the lifespan. And it also is inherited. So it’s also been associated with a number of disease traits.

There’s been a few studies looking at methylation patterns and epigenetics in migraine, but not a great deal. So some of those have involved quite small numbers. And there’s one here by Gerring et al., a couple of others as well. And they’ve implicated potentially some genes, but no significant results really outstanding at this stage, although there’s a few implicated potential disease genes there that may be playing a role. So even though the migraine epigenetic studies, to date, have been quite limited, this is an area that’s certainly worthy of investigating more in relation to things like migraine development, potentially the subtypes, and even into treatment response. 

The last area I’ll mention that I think we should be exploring in relation to migraine is pharmacogenomics. So this is the study of differential response to medications due to different genetic variation backgrounds. The goal of pharmacogenetics is to develop individualized treatment strategies that are based on genetic makeup. Now since not all patients respond well to different migraine treatments, perhaps pharmacogenomics studies can help us define the most appropriate treatments.

There’s been a number of papers that have outlined the potential importance of this, of migraine pharmacogenomics, but there’s been very few studies so far. So the treatment of monogenic forms of migraine, as I’ve mentioned, and related disorders such as FHM and PRRT2 already utilize genetic mutation results to aid in treatment choice. So this could be further developed and applied to common types of migraine to help understand things like variability and response to treatments, and even to develop accurate predictive panels for acute and preventative treatment response. And if we could do that, that would really help in individualizing migraine treatment based on genetic information. 

So in conclusion, sequencing has become a cost-effective approach, not just for migraine research but also for diagnostics to identify the genes involved in the various types of migraine. Many genes and genetic variants have been implicated, particularly in the monogenic forms of migraine, and there’s ongoing studies to improve diagnostic success rates.

For the more common types of migraine, also many genes and genetic variants have been identified, but this doesn’t yet account for all the hereditary factors involved in the more common types of migraine. So there is a need for future research to do that, potentially investigating rare variants, structural variations such as copy number variants, also investigating epigenetic factors may also help account for this messy heritability in migraine. And finally, genetics also holds the promise of identifying predictive pharmacogenomic profiles, which could aid in developing more individualized treatment choices for migraine management. 

I want to acknowledge all the people that work with me in my lab, both the genotyping, bioinformatics, and diagnostic staff, also our clinical collaborators and contacts, and also the support that we’ve had from a variety of organizations for migraine research. And I just want to thank you for your time.

*The contents of this video are intended for general informational purposes only and does not constitute professional medical advice, diagnosis, or treatment. Always seek the advice of a physician or other qualified health provider with any questions you may have regarding a medical condition. AMD and the speaker do not recommend or endorse any specific course of treatment, products, procedures, opinions, or other information that may be mentioned. Reliance on any information provided by this content is solely at your own risk.