Mitochondrial Mutations and Autism
August 18th, 2008
My piece on how “alternative” practitioners keep people from recognizing the futility of their treatments will have to wait a bit longer. Something has come up.
Since the Hannah Poling case was conceded late last year, the “autism community” has been abuzz with talk of mitochondrial disease. It is a measure of how “hip” mitochondrial diseases have become that they have acquired their own shorthand term: “mito” (as in, “My DAN! doctor wants to test my son for mito.”).
More recently, the same community has thrilled to what they have been told is a “bombshell” in the hypothetical link between autism and mitochondrial disease. I refer, of course, to the study published by Elliot et al in the 8 August 2008 edition of The American Journal of Human Genetics.
The Article:
In this study, the authors looked for ten of the most common pathological (i.e. associated with one of more diseases) mutations in the mitochondrial DNA (mtDNA) extracted from the cord blood samples from 3168 consecutive births in North Cumbria. They found 15 children with mutations in their mtDNA.
They then went back and tried to find blood samples from the mothers of these 15 children with mtDNA mutations. They found maternal samples for eight of the fifteen children (one was from the same mother, which suggests twins) and found that only three of the eight mutations were not found in the mother (i.e. five of the eight children - 63% - had the same mutations as their mothers).
Now, this worked out to about 0.47% of the children having mtDNA mutations that have been associated with mitochondrial diseases, or about 1 in 211. If you look only at the subjects where mtDNA sequencing was successful, it works out to 0.54%, or 1 in 185. This is much higher than the prevalence of mitochondrial disease – generally thought to be about 1 in 5000.
As a result, the authors felt that it was very important to let the scientific and medical communities know – via their article – that the prevalence of mitochondrial mutations is much higher than would be predicted by the prevalence of mitochondrial diseases. They made two comments in their conclusion that are relevant to this point:
The first, which has been taken up as a rallying cry by certain scientifically under-literate journalists, was:
“We have identified a massive reservoir of pathogenic mtDNA mutations in the general population, placing greater emphasis on developing techniques to prevent the transmission of pathogenic alleles that could segregate to high levels and thus cause mtDNA diseases in subsequent generations.”
[Note: the authors fail to explain why this is a problem that needs to be addressed, given that the "massive reservoir of pathogenic mtDNA mutations" has probably been present for most of recorded history, if not longer. They present no data or explanations to suggest that this "problem" is either new or increasing. While it is always better for prospective parents to have all possible information about potential genetic disease issues in their offspring, this "problem" is one that is essentially insoluble - the prospective mother-to-be who has a modest percentage of mutated mtDNA will have no options except to forgo having children.]
The second, which has largely been ignored, is perhaps even MORE relevant:
“Detecting heteroplasmic mtDNA mutations in >1 in 200 individuals of the background population has implications for studies reporting mtDNA mutations in specific disease groups. Our data show that putative disease associations, such as the reported high frequency of m.3243A/G in diabetes mellitus, could be a chance finding irrelevant to pathogenesis.” [emphasis added]
So, what Elliot et al have shown is that mtDNA mutations appear to be much more common than can be explained by the prevalence of mitochondrial diseases. This suggests that there is more to the story than we currently know.
It also suggests that people who have trumpeted that “1 in 200 people have mitochondrial disease” or even “1 in 200 people have mitochondrial mutations” as vindication of the “autism is a mitochondrial disease” need to read the paper. If having the mutation was enough to cause a mitochondrial disease, 1 in 200 people would have a mitochondrial disease, not just be able to transmit it to their children.
In fact, this study shows that mtDNA mutations found “in association” with certain diseases may be simply due to chance, not because the mutation causes the disease. This finding should cause people to be much more cautious about attributing a disease to the presence of a mutated gene and makes the “1 in 200″ number much more of a question mark.
If you were just interested in the study, you can stop reading here. For more information about mitochondria and their fascinating genetics, read on…
Mitochondrial Genetics 101:
Much of the confusion over the relationship between mitochondrial disease and autism lies in the rather confusing (to the average layperson) genetics of the mitchondria themselves.
Each human cell (except mature red blood cells) has between dozens and thousands of mitochondria. They carry out a variety of metabolic processes, the most famous being the production of ATP, the energy “currency” of the cell, through oxidative phosphorylation. Mitochondria look and act like bacteria – in fact, they are distantly related to alpha-proteobacteria. They each have their own circular chromosome (like bacteria) which encodes for bacteria-like ribosomal RNA (rRNA), among other things.
Over the course of evolution (it is thought that the mitochondrial association with eukaryotic cells is about 2 billion years old), the mitochondrial DNA (mtDNA) has lost a lot of genes (compared to their free-living alpha-proteobacterial “cousins”) and transferred nearly all of the rest (about 1400) to the nuclear DNA (nDNA) of their host cell.
There are three important points to remember about the mitochondrial genome:
[1] The mtDNA replicates separately from the nuclear DNA (which, in mature differentiated cells may happen essentially never) and behaves like a bacterial chromosome. However, unlike bacteria, the mtDNA has only 37 genes. This is smaller than many viral genomes and is inadequate for the functions that the mitochondria perform. Most of the mitochondrial genes (about 1400) are located on the nuclear DNA.
[2] Genes on the mtDNA are present in one copy per genome (each mitochondria can have between 5 and 20 copies of its genome, which doesn’t change the issue – all the copies are essentially identical). There are no “homozygous/heterozygous” or “recessive/dominant” genes in mtDNA – what see is what you’ve got. In other words, you can’t have a “silent carrier” state in mtDNA mutations in the same way that you can have “silent carriers” in nuclear genes.
[3] Each cell has multiple mitochondria. Depending on when in development a mutation occurred - and the random “luck of the draw” in the mitochondria the original oocyte (egg cell) got from mother - the mitochondria can all have the mutation or just some of them can have it. This is referred to as homoplasmy (all mitochondria have identical genomes) or heteroplasmy (the mitochondria do not all have identical genomes).
At the time the oocytes (egg cells) form (very early embryonic development), the mitochondria in the oogonia are divided randomly between the resulting primary oocytes and, later, between the resulting secondary oocytes and polar bodies. If the cells are heteroplasmic, there is a chance that the resulting secondary oocyte will end up with more (or less) of one type of mitochondria than the other cells in the embryo.
At least one study has estimated that as few as ten mitochondria are passed from the oogonia to the resulting secondary oocyte, which would mean that there is a significant probability that a heteroplasmic embryo could end up with oocytes (”egg cells”) that have a very different percentage of mitochondrial genotypes than the rest of the embryo. That, of course, translates to the heteroplasmic mother passing down to her children a different percentage of mitochondrial genotypes than she has. In fact, a heteroplasmic mother’s children will most likely have a different distribution of mitochondrial genotypes, just as a function of chance.
Now, this could be bad (as Elliot et al discuss) if the children end up with a higher percentage of “bad” mitochondrial genotypes (mutations). However, depending on the percentages, it may be that the percentage of “bad” genotypes would be lower in the children than in the mother. Let’s “run the numbers:
If mother’s mitochondria are 50% “good” and 50% “bad”, and only 10 are randomly selected to form the egg’s mitochondrial population, the chance that the reulting egg will have mitochondria that are:
more than 50% “bad” = 37.7%
exactly 50% “bad” = 24.6%
less than 50% ”bad = 37.7%
In this scenario, the probabilities are equal the the offspring will end up with more or fewer ”bad” mitochondria. What if “mom” has 80% “bad mitochondria?:
more than 80% “bad” = 37.6%
exactly 80% ”bad” = 30.2%
less than 80% “bad” = 32.2%
Here we see that if “mom” has a high percentage of “bad” mitochondria, the trend will be toward getting more “bad” mitochondria. On the flip side, if “mom” has only 20% “bad” mitochondria:
more than 20% “bad” = 32.2%
exactly 20% “bad” = 30.2%
less than 20% ”bad” = 37.6%
Then the numbers “flip over” and the trend is toward a lower percentage of ”bad” mitochondria than “mom”.
So, while there is a chance that the offspring will have a higher percentage of “bad” mitochondria, it depends a lot on the percentage of “bad” mitochondria.
In the Elliot et al study, most of the children with mutations were heteroplasmic – with between 89% and 0.5% of the mitochondria having the mutation in question. The median heteroplasmy was 12.9%. Three of the children were homoplasmic, with 100% of their mitochondria having the mutation. These three children came from two mothers (again, suggesting a twin birth or two births in fairly rapid succession) who were also homoplasmic for the same mutations.
The article is silent about whether the mothers of these three children had any signs of mitochondrial disease or had a history of mitochondrial disease in the family. Perhaps that will come out in a later article.
Homoplasmy/heteroplasmy in mitochondria plays much the same role as homozygosity/heterozygosity in the nuclear chromosomes, with a difference.
If a cell has a small percentage of its mitochondria that cannot function adequately, it will probably be able to function normally by relying on the rest. As the percentage of “dysfunctional” mitochondria increases, so does the likelihood that the cell will be unable to function normally. That’s the simplified version of mitochondrial genetics.
It gets more complicated when you look at the mitochondrial genes on the nuclear chromosomes. These genes code for (primarily) proteins which are then tagged and exported to the mitochondria. If the nuclear genes code for something that can substitute for the mutation on the mtDNA, then the mitochondria will work just fine. This is what causes the folks working on the genetics of mitochondrial disorders to have sleepless nights. It might take TWO mutations – one in the mtDNA and one in the nuclear DNA – to cause a mitochondrial disease.
Mitochondria and “Toxins”:
Much of the “alternative” press on mitochondrial diseases and autism has focused on the possibility that an “environmental factor” (care to guess which ones?) can cause mitochondrial damage, leading to a permanent disability. In the next few paragraphs, I’ll run through two ways that has been proposed to happen.
Mitochondrial damage:
Killing or damaging mitochondria is serious business. Some of the most lethal poisons (e.g. cyanide) work on the mitochondria. Damaged mitochondria can trigger apoptosis (programmed cell death). If this happens in the wrong place or at a developmentally sensitive time, the results can be catastrophic.
If the mitochondria are more “fragile” than normal – due to a mutation in either mtDNA or the nuclear genes encoding for mitochondrial components – they may be more sensitive ”stressors”. Heavy metals (e.g. mercury) have been shown to “stress” mitochondria, as have a number of other compounds. The fever caused by viral, bacterial or parasitic infections – as well as by vaccines – can potentially cause mitochondrial “stress” as well, although the vaccines typically cause less of a response than the disease they were developed to prevent. The idea of leaving a child with mitochondrial “dysfunction” unvaccinated begs the question: isn’t it better to prevent the “full blown” disease, even if you run the risk of “triggering” mitochondrial “stress”?
Another well-studied stressor for mitochondria is oxygen. Under “normal” oxygen concentrations, mitochondria produce a lot of reactive oxygen species (ROS), which cause a lot of damage to the mitochondrial proteins, membranes and mtDNA (which may be one reason that most of the mitochondrial genes “migrated” to the nucleus). Elevated oxygen concentrations (e.g. HBOT) increase the production of ROS, leading to more damage.
Damage to “fragile” mitochondria by “stress” (”toxins”, heat, oxygen, viral infection, etc.) at a developmentally critical point is perhaps the only plausible connection between autism and mitochondrial disease. Although it may have been fever caused by vaccination that caused the damage in Hannah Poling’s case, it could have just as easily been caused by any of thousands (if not millions) of viral and bacterial infections that children are susceptible to.
mtDNA mutations:
Mutating the mtDNA can also be a catastrophic event….for the individual mitochondrion. If it is severe enough to prevent the mitochondrion from functioning, it will die, leaving the other 99 to 1999 mitochondria to carry the load. Of course, the mitochondria periodically divide, so they will soon make good any losses. Even mitochondria in the brain cells divide about once a month, so any loss of mitochondria that is not acutely lethal will be made good in a short while.
Non-lethal mtDNA mutations are passed down to all of the “daughters” of the mutated mitochondrion. Again, since the other non-mutated mitochondria are also dividing, this affects only a small percentage of the mitochondria present in the cell. In fact, if the mutation makes the mitochondrion slower to grow and divide, it will eventually be “diluted out” by the daughter cells of the non-mutated mitochondria, which are dividing faster.
Mitochondrial mutations that occur after a point in early embryonic development cannot be passed on to offspring (of the larger, multicellular organism “hosting” the mitochondria, such as a human) unless the mutation occurs in the mitochondria of an egg cell (which are all formed during embryonic development). And since mammals get all of their mitochondria from their mothers, paternal mtDNA mutations - even in the sperm - don’t go anywhere (sometimes literally - sperm with dysfunctional mitochondria can’t swim).
So, the idea of a “toxin” or “virus” causing mtDNA mutations is not implausible. What is implausible is the idea of the same mutation occurring in a significant number of mitochondria. To give you an idea of the magnitude of the problem, let’s “run the numbers”.
The human mtDNA has 16,571 base pairs (bp). Each of these can be mutated to one of three bases (apart from the one “correct” base). Even if we assume that mutating the base to any one of the other three will cause a “dysfunctional” mutation, the odds of the same mutation happening by chance in one other mitochondrion is 1 in 16,571.
If we also assume that there are only 100 mitochondria in the cell and that mutating as little as 50% of them will cause “dysfunction” (both are low numbers – there are usually more mitochondria and it usually takes more than 50% to manifest disease), then the odds of that happening are:
1 in 1.2 X 10^182
Rounding down, that’s a 1 followed by 182 zeroes. Pretty long odds, even for a lottery. And that’s only for a single cell. If you want to imagine the same “toxic” exposure causing the same level of mutation in more than one cell, the numbers go up pretty fast.
Conclusion:
So, the Elliot et al study didn’t “prove” that mitochondrial disease is more prevalent than previously thought - they didn’t even look at the presence of disease at all. The study did show that mtDNA mutations are more common than the prevalence of mitochondrial disease would indicate. This suggests that there is more to “mitochondrial dysfunction” than simply genetics.
Does this mean that autism can’t be a “mitochondrial dysfunction”, at least in some cases? No. But is also no data to support that it can. When a study is done comparing the prevalence of mitochondrial disease or “dysfunction” in autistic and non-autistic matched controls, that question will begin to be answered. Currently, there is only speculation.
Postscript:
As I predicted earlier, practitioners have rushed to promote their own unique treatments for “mitochondrial autism”, including the usual suspects: supplements, vitamins, minerals and chelation. Some have even touted HBOT as a “treatment” (rather than a cause) of “mitochondrial dysfunction”. No doubt they all have carefully reasoned explanations of how their particular “treatment” will reverse whatever mitochondrial “dysfunction” is present.
Testing has lagged behind, although some local practitioners say that the “routine biomedical testing” will pick up “mitochondrial autism”. They fail to mention why this “routine biomedical testing” failed to detect mitochondrial “dysfunction” previously.
Clearly, there is money to be made in “mitochondrial autism” and the advertisements and postings on the Internet indicate that there is no shortage of people trying to cash in. If you are concerned that your child has a mitochondrial disorder, please see a real doctor and get real testing done. Mitochondrial testing is far from routine and requires a lab that is both meticulous and experienced. Mail-order labs are not going to be able to do it right. The sample preparation and analysis are too complex for most hospital labs - let alone direct-to-consumer mail-order labs.
Caveat emptor
Prometheus
Filed under: Autism Science, Autism Treatments, Critical Thinking | 32 Comments »