Jun 09 2011
The Genetics of Autism
The recent issue of Neuron has a series of articles providing more information about the genetics of autism spectrum disorder (ASD). ASD is characterized by decreased social ability as a core feature, with other clinical features being variable. It is also not a single disease or disorder, and not just because of the spectrum of clinical features. Like many clinical entities, there can be many underlying causes that result is similar-looking clinical effects.
While debate rages as to possible environmental triggers or even causes of ASD, researchers have been slowly building a picture of ASD as a complex genetic syndrome. Literally hundreds of genes have been potentially implicated. Many of the genes linked to ASD are involved with brain proteins or brain organization.
The recent studies in Neuron look specifically at families with a single child with ASD. In families with multiple children, or with a parent and child on the spectrum, the disorder is likely inherited. But what of families with only one child with ASD, and at least one child without, with no affected parent? Such a pattern can also be consistent with an inherited disorder, if it is recessive or X-linked. In a recessive disorder both parents can be unaffected carriers, and 25% of children will be affected – so having a single affected child is not unusual. However, ASD genes tend to be dominant, which means at least one parent should be affected along with at least 50% of children.
Another way for single individuals to appear with genetic disorders is called variable penetrance. The same genes do not always translate into identical clinical manifestations. One family member can have a very subtle manifestation of an inherited disorder, while another (the one who gets identified and diagnosed) can have a more severe presentation.
There is also the phenomenon known as amplification. Some genetic disorders are caused by trinucleotide repeats – repeated three letter segments of DNA. The greater the number of repeats, the more severe the clinical syndrome. Repeat numbers tend to increase with each affected generation, and so a very mild manifestation in a parent can amplify to a severe presentation in a child. Such disorders, however, also tend to be dominant.
Yet another way for a genetic (but not inherited) disorder to appear in a single individual in a family is for that individual to have a spontaneous mutation – a mutation unique to them, and not inherited from a parent. This is the specific hypothesis being tested in the current study.
Sanders et al studied 1124 families with one affected child with ASD and one unaffected child, with unaffected parents. They found that 6-8% of them could be explained by changes in copy number variants – how many copies of specific genes were present. Most intriguing, they found:
We find significant association of ASD with de novo duplications of 7q11.23, where the reciprocal deletion causes Williams-Beuren syndrome, characterized by a highly social personality.
In other words – at one gene location where a deletion causes a syndrome characterized by a highly social personality, they found duplications associated with a decrease in sociability. This strongly suggests that the gene in question has a strong influence on sociability, and can be either turned up or down depending on the genetic change.
In another study in the same issue researchers find that large networks of genes are responsible for the ASD phenotype. Further they provide evidence that could explain the fact that males are at higher risk of ASD than females. They find that a much greater perterbation in the gene network is required for ASD to manifest in girls than in boys.
Conclusion
ASD is a complex disorder but researcher are slowly unraveling the changes in the brain that correlate with and cause the features of ASD. Further, research into the genetics of ASD is progressing rapidly and producing highly significant results.
We are still far from fulling explaining the genetic or neuroanatomy and physiology of ASD – but the utility of a scientific theory is better understood by how well the research is progressing, rather than how far we have already come. The fact that the genetic theory of autism is producing such useful results speaks to the power of this theory. This, of course, does not rule out environmental effects playing a role. Genes evolve to respond to the environment, and increasingly scientists are identifying important epigenetic factors in disease. But the research to date suggests a highly dominant role for genetics in autism, and this recent research adds to the growing body of research supporting this.
16 Responses to “The Genetics of Autism”
Leave a Reply
You must be logged in to post a comment.
To Quote Shakespeare “Our remedies oft in ourselves do lie,
Which we ascribe to Heaven.
Personally, I would refrain from interpreting this research, and leave it up to people who have less bias in the matter.
One of the main points of AoA-people being that autism is caused (mainly or solely) by environmental factors, I would love to have that conclusively disproved. Which is why, if look at this kind of research, I’ll probably see what I want to see, whether it’s in the data or not.
But I’m looking forward to see how our understanding develops.
Can environmental factors can cause CNV mutations?
Also, does this mean that we can talk about two types of autism: this CNV mutation type and an inherited type?
I’m confused. Are you saying there are codon repeats, gene copy repeats or as written, nucleotide repeats of three, akin to telomers?
Thanks for the links! I am looking forward to reading the other articles in the issue.
I was fortunate enough to attend the IMFAR meeting last month, and one of the keynote speakers (Dr. Ricardo Dolmetsch) is studying families with syndromic forms of autism – specifically Phelan McDermid syndrome (PMS) and Timothy syndrome (TS). In PMS, the deletion disrupts SHANK3, an important glutamate synaptic protein. In TS, the mutation interferes with L type calcium channels. His research involves taking skin cells from affected individuals, de-differentiating them into stem cells, and then re-differentiating them into neurons. His lab is studying the changes in neurophysiology brought on by these genetic defects in cell culture, and already potential therapeutic agents have been suggested by their findings. Other synaptic genes like neurexin and neuroligin have been implicated in autism as well. Even though in sum these genetically characterized forms of autism account for only a small fraction of ASD cases, they shed invaluable light on the biochemical pathways likely involved in the uncharacterized cases. Isn’t science kick-ass?
http://www.ncbi.nlm.nih.gov/pubmed/19454329
http://www.ncbi.nlm.nih.gov/pubmed/19916019
http://www.ncbi.nlm.nih.gov/pubmed/21424692
http://www.ncbi.nlm.nih.gov/pubmed/19243448
I would also point out that some researchers presenting at the meeting did discuss the gender differences in ASD diagnosis. One proposed theory is that girls with ASD are under-recognized. A boy who quietly sits alone and lines up his toys is considered odd, whereas a girl exhibiting the same behavior might be considered well-behaved. I don’t know if there is published data on that particular theory, though.
“I’m confused. Are you saying there are codon repeats, gene copy repeats or as written, nucleotide repeats of three, akin to telomers?”
He is referring to trinucleotide repeats, as seen in Huntington’s disease (CAG repeats), Fragile X syndrome (CGG repeats), myotonic dystrophy (CTG repeats), and DRPLA (CAG repeats).
You can read about all of them in OMIM if you like:
http://www.ncbi.nlm.nih.gov/omim
draal:
I believe Dr. Novella was just giving an example of a type of mutation, called the trinucleotide expansion. Woody was spot on. However, it is not a trinucleotide expansion that is being discussed as the mechanism for ASD here.
7q11.23 is a region of DNA on the long arm of chromosome 7 that comprises 23-25 genes (depending on which source you look at). The article states an association between a duplication of all these chromosomes and contrasts that with a known disease process where those same genes are deleted and cause hyper-sociability (amongst other deleterious physical changes as well).
Also, telomeres in vertebrates are actually a repeat of TTAGGG not just 3 nucleotides as in Huntingtons or fragile X.
@mike:
I am not an expert on CNV mutations, but from my reading it seems that the mechanism by which they arise is not fully understood. It is likely an issue with DNA repair mechanisms (a la microsattelite instability) relating specifically to what is called microhomology-mediated break-induced replication. Essentially, there are short regions that are quite similar in the coding DNA (micro-homology) that then break apart during DNA replication and stick back together with a nearby matching (homologous) strand before the DNA polymeras/repair proteins have a chance to do anything about it. Then the replication will continue after the re-annealed portion and replicate that gene(s) again.
This is observed in monozygotic twins demonstrating that identical womb environments with identical genomes can lead to phenotypic differences of expression of such CNV genetic diseases (source)
To me, this would indicate that this happens sporadically and certain genotypes are simply more predisposed for it, but not necessarily 100% likely to happen. While that doesn’t fully exclude an environmental trigger being able to cause it, since there is discordance amongst MZ twins like this, it means that an environmental trigger is at least not necessary. With everything else taken into account it seems that at best, environmental trigger does not account for the bulk of CNV mutations.
As for two kinds of autism, I think the data doesn’t speak one way or another on it. It could well be that this is merely a sporadic form of the same genetic process that causes all autism. It could be that the sporadic form is a different mutation from the heritable form, but they are both still genetic diseases. Either way we would still lump them into the same disease (Burkitt’s lymphoma, for example has a sporadic and endemic form with different etiologies but still has the same clinical picture – cystic fibrosis is always inherited but there are thousands of different mutations that cause it).
At least, that is my take on it. I’m happy if anyone can add to or refine it.
The tendency for trinucleotide repeat expansion disorders to worsen in successive generations has always been referred to as “anticipation” in every genetics course I’ve taken. I’ve never heard it referred to as “amplification” before.
Is there a subtlety here that I am missing? Like “amplification” referring to a worsening of symptoms whereas “anticipation” refers to an earlier manifestation? Does anybody know?
@hyperlalia:
My understanding is that anticipation refers to the phenotypic expression of the trinucleotide repeats and amplification refers to the actual number of repeats.
In other words the amplification of CAG in the Huntingtin gene leads to the disease presenting earlier in life, which is the anticipation. The disease is the same regardless of how many repeats there are – it doesn’t “worsen” but it manifests earlier.
Hope that makes sense.
nybrus-
I’ve read that sometimes repeats in the DNA will slow the cells ability to produce certain proteins.
That is the cell can still make the normal protein, but it does so more slowly than a normal cell would.
(Sorry- I lost my reference for this, but I’m hoping it’s something you know about.)
Anyway, it seems that the more slowly a cell could make a protein, the sooner a lack of that protein would manifest as disease.
That is to say- the more repeats the sooner the disease shows up.
You like?
“In other words the amplification of CAG in the Huntingtin gene leads to the disease presenting earlier in life, which is the anticipation. The disease is the same regardless of how many repeats there are – it doesn’t “worsen” but it manifests earlier.”
I have a real life example.
There was a family down the road from where we used to live. The father developed Huntington’s late in life and died of a heart attack in his mid seventies while still living at home. At the time of his death, his son was already in a nursing home as a result of his disability.
@sonic:
I did a small amount of research in my spare time and thought over your question. I don’t think that the slow protein hypothesis is a particularly compelling one.
Specifically for Huntington’s disease, the mechanism of manifestation is a post-translational modification. After the protein is synthesized it is chopped up and folded. The CAG repeats create a whole bunch of extra glutamines. These are cut out and left behind. The more you have, the more little chunks of glutamine oligopeptide residues are left and they actually end up congealing and clogging up your cells. This is toxic to your neurons and Bob’s your uncle.
For fragile X it is because the repeats are in a non-coding region and the longer they are the more likely the chromosome is to literally snap at that point, creating instability in the growing cells and causing all kinds of problems.
That doesn’t address all possible trinucleotide repeat diseases, so I suppose there is one that could work that way. But in my understanding, when you have a gene that codes for a protein it gets translated just fine regardless of how long it is. Obviously the longer the code is the longer it would take to churn out an individual protein, but I really don’t think that would become manifest because our cells have the capacity to upregulate ribosomes and churn out proteins like mad (think of say, Bence-Jones proteins in multiple myeloma).
As far as I know, invariably the problem arises from protein folding issues or early termination. So if the mutation codes for the ribosome to stop making the protein early (like DF506 in cystic fibrosis) or if the mutation codes for the wrong amino acid(s) and makes it so the protein can’t fold right and then it gets degraded instead of used.
But if you can find that source you are referring to I’d be keen to look at it and see what it has to say – there’s always more to learn about molecular genetics!
nybgrus-
I can’t find the reference. It was probably a one of study. I don’t know enough about this to make a judgement, but from my looking it doesn’t seem this phenomena is well known– perhaps it isn’t real.
It doesn’t seem likely to be a factor in Huntington’s- the glutamine residue sounds right.
Probably just a false alarm.
I have become used to them.
sonic: it happens.
It could have been related to a very specific disease as well – that is why I was curious about it. Proteins can interact with DNA/RNA and if a protein takes longer to translate it can potentially have funky and complicated downstream effects. But that would be a specific example and, at least from I have learned to date, not the general rule for protein errors.
@steve12 -
“Can environmental factors can cause CNV mutations?”
Yes. My MZ twin daughters have different, non-shared CNVs. Twin A has two copy losses (1q44, Xp22) while twin B has one (6p25).
Since they both came from identical genetic material, the only way that the CNVs could be different is if something caused them after they split from each other, which in their case was about 1 to 3 days after conception.
However, they both have the same severity of autism which makes me wonder whether CNVs like these are the cause of autism or just a side effect of whatever the actual cause is.