Monday, 30 May 2016

Sense, Missense or Nonsense - Interpreting Genetic Research in Autism (TCF4, TSC2 , Shank3 and Wnt)

Some clever autism researchers pin their hopes on genetics, while some equally clever ones are not convinced.

One big problem is that genetic testing is still not very rigorous, it is fine if you know what you are looking for, like a specific single gene defect, but if it is a case of find any possible defect in any of the 700+ autism genes it can be hopeless.

Most of the single gene types of autism can be diagnosed based on known physical differences and then that specific gene can be analyzed to confirm the diagnosis.

Today’s post includes some recent examples from the research, and they highlight what is often lacking - some common sense.

There are numerous known single gene conditions that lead to a cascade of dysfunctions that can result in behaviors people associate with autism.  However in most of these single gene conditions, like Fragile X or Pitt-Hopkins, there is a wide spectrum, from mildly affected to severely affected.

There are various different ways in which a gene can be disturbed and so within a single gene condition there can be a variety of sub-dysfunctions.  A perfect example was recently forwarded to me, a study showing how a partial deletion of the Pitt Hopkins gene (TCF4) produced no physical features of the syndrome, but did unfortunately produce intellectual disability.

The study goes on to suggest that “screening for mutations in TCF4 could be considered in the investigation of NSID (non-syndromic intellectual disability)”

Partial deletion of TCF4 in three generation family with non-syndromic intellectual disability, without features of Pitt-Hopkins syndrome

This all matters because one day when therapies for Pitt Hopkins are available, they would very likely be effective on the cognitive impairment of those with undiagnosed partial-Pitt Hopkins.

Another reader sent me links to the studies showing:-

Rapamycin reverses impaired social interaction in mouse models of tuberous sclerosis complex.

Reversal of learning deficits in a Tsc2+/- mouse model of tuberous sclerosis.

But isn’t that Tuberous sclerosis (TSC) extremely rare? like Pitt Hopkins.  Is it really relevant?

Tuberous sclerosis (TSC)  is indeed a rare multisystem genetic disease that causes benign tumors to grow in the brain and on other vital organs such as the kidneys, heart, eyes, lungs, and skin. A combination of symptoms may include seizures, intellectual disability, developmental delay, behavioral problems, skin abnormalities, and lung and kidney disease. TSC is caused by a mutation of either of two genes, TSC1 and TSC2, 

About 60% of people with TSC have autism (biased to TSC2 mutations) and many have epilepsy.

How rare is TSC?  According to research between seven and 12 cases per 100,000, with more than half of these cases undetected.  

Call it 0.01%, rare indeed.

How rare is partial TSC?  What is partial TSC?  That is just my name for what happens when you have just a minor missense mutation, you have a mutation in TSC2 but have none of the characteristic traits of tuberous sclerosis, except autism.
In a recent study of children with autism 20% has a missense mutation of TSC2. 

Not so rare after all.

Mutations in tuberous sclerosis gene may be rife in autism

Mutations in TSC2, a gene typically associated with a syndrome called tuberous sclerosis, are found in many children with autism, suggests a genetic analysis presented yesterday at the 2016 International Meeting for Autism Research in Baltimore.
The findings support the theory that autism results from multiple ‘hits’ to the genome.
Tuberous sclerosis is characterized by benign tumors and skin growths called macules. Autism symptoms show up in about half of all people with tuberous sclerosis, perhaps due to abnormal wiring of neurons in the brain. Tuberous sclerosis is thought to result from mutations in either of two genes: TSC1 or TSC2.
The new analysis finds that mutations in TSC2 can also be silent, as far as symptoms of the syndrome go: Researchers found the missense mutations in 18 of 87 people with autism, none of whom have any of the characteristic traits of tuberous sclerosis.
“They had no macules, no seizure history,” says senior researcher Louisa Kalsner, assistant professor of pediatrics and neurology at the University of Connecticut School of Medicine in Farmington, who presented the results. “We were surprised.”
The researchers stumbled across the finding while searching for genetic variants that could account for signs of autism in children with no known cause of the condition. They performed genetic testing on blood samples from 87 children with autism.

Combined risk:

To see whether silent TSC2 mutations are equally prevalent in the general population, the researchers scanned data from 53,599 people in the Exome Aggregation Consortium database. They found the mutation in 10 percent of the individuals.
The researchers looked more closely at the children with autism, comparing the 18 children who have the mutation with the 69 who do not.
Children with TSC2 mutations were diagnosed about 10 months earlier than those without a mutation, suggesting the TSC2 mutations increase the severity of autism features. But in her small sample, Kalsner says, the groups show no differences in autism severity or cognitive skills. The researchers also found that 6 of the 18 children with TSC2 mutations are girls, compared with 12 of 69 children who don’t have the mutation.
TSC2 variants may combine with other genetic variants to increase the risk of autism. “We don’t think TSC is the sole cause of autism in these kids, but there’s a significant chance that it increases their risk,” Kalsner says.

"hyperactivation of the mechanistic target of rapamycin complex 1 (mTORC1) is a consequence of tuberous sclerosis complex (TSC) 1/2 inactivation."

"the combination of rapamycin and resveratrol may be an effective clinical strategy for treatment of diseases with mTORC1 hyperactivation."

So for the 20% of autism with partial TSC, so-called Rapalogs and other mTOR inhibitors could be helpful, but Rapalogs all have side effects.

One interesting option that arose in my earlier post on Type 3 diabetes and intranasal insulin is Metformin. The common drug used for type 2 diabetes.


Metformin regulates mTORC1 signaling (but so does insulin).

'Metformin activates AMPK by inhibiting oxidative phosphorylation, which in turn negatively regulates mTORC1 signaling via activation of TSC2 and inhibitory phosphorylation of raptor. In parallel, metformin inhibits mTORC1 signaling by suppressing the activity of the Rag GTPases and upregulating REDD1."

Source:  Rapalogs and mTOR inhibitors as anti-aging therapeutics

Clearly you could also just use intranasal insulin.  It might be less potent but should have less side effects because it acting only within the CNS (Metfornin would be given orally).

The Shank protein and the Wnt protein family

Mutations in a gene called Shank3 occur in about 0.5 percent of people with autism.  
But what about partial Shank3 dysfunction?

Shank proteins also play a role in synapse formation and dendritic spine maturation.

Mutations in this gene are associated with autism spectrum disorder. This gene is often missing in patients with 22q13.3 deletion syndrome

Researchers at MIT have just shown, for the first time, that loss of Shank3 affects a well-known set of proteins that comprise the Wnt signaling pathway.  Without Shank3, Wnt signaling is impaired and the synapses do not fully mature.

“The finding raises the possibility of treating autism with drugs that promote Wnt signaling, if the same connection is found in humans”

I have news for MIT, people already do use drugs that promote Wnt signaling, FRAX486 and Ivermectin for example.  All without any genetic testing, most likely.

Reactivating Shank3, or just promote Wnt signaling

The study below showed that in mice, aspects of autism were reversible by reactivating the Shank3 gene.  You might expect that in humans with a partial Shank3 dysfunction you might jump forward to the Wnt signaling pathway and intervene there.

Mouse study offers promise of reversing autism symptoms

One reader of this blog finds FRAX486 very helpful and to be without harmful side effects.  FRAX 486 was recently acquired by Roche and is sitting over there on a shelf gathering dust.

Where from here?

I think we should continue to look at the single gene syndromes but realize that very many more people may be partially affected by them.

Today’s genetic testing gives many false negatives, unless people know what they are looking for; so many dysfunctions go unnoticed.

This area of science is far from mature and there may be many things undetected in the 97% of the genome that is usually ignored that affect expression of the 3% that is the exome.

So best not to expect all the answers, just yet, from genetic testing; maybe in another 50 years.

Understanding and treating multiple-hit-autism, which is the majority of all autism, will require more detailed consideration of which signaling pathways have been disturbed by these hits.  There are 700 autism genes but there a far fewer signaling pathways, so it is not a gargantuan task.  For now a few people are figuring this out at home.   Good for them.

I hope someone does trials of metformin and intranasal insulin in autism.  Intranasal insulin looks very interesting and I was surprised to see in those earlier posts is apparently without side effects.

The odd thing is that metformin is indeed being trialed in autism, but not for its effect on autism, but its possible effect in countering the obesity caused by the usual psychiatric drugs widely prescribed in the US to people with autism.

My suggestion would be to ban the use of drugs like Risperdal, Abilify, Seroquel, Zyprexa etc.

Vanderbilt enrolling children with autism in medication-related weight gain study

Here are details of the trial.

Metformin will be dispensed in a liquid suspension of 100 mg/mL. For children 6-9 years of age, metformin will be started at 250 mg at their evening meal for 1 week, followed by the addition of a 250 mg dose at breakfast for 1 week. At the Week 2 visit, if metformin is well-tolerated, the dose will be increased to 500 mg twice daily. For children from 10-17 years of age, metformin will be started at 250 mg at their evening meal for 1 week, followed by the addition of a 250 mg dose at breakfast for 1 week. At the Week 2 visit, if metformin is well-tolerated, the dose will be increased to 500 mg twice daily. At the Week 4 visit, if metformin is well-tolerated, the dose will be increased to 850 mg twice daily.


  1. Based on my own results,I am a big advocate of whole exome and whole genome sequencing for autism.I recently saw somewhere,that an estimated 60% of the genes involved in the different causes of autism have not yet been found.While there are a growing number of researchers who are performing such tests on autistic children,I do not believe it is as common as it should be.Many doctors are adverse to going on "fishing expeditions" looking for genetic mutations.I would like to see as close as possible to universal whole exome and genome sequencing on every individual with an autism diagnosis.Again based on my own results,I think you will find there are novel mutations of genes for neurological disorders not previously known to be associated with autism,as well as mutations for more than one genetic disorder in the same patient.In the more complex patients,you might find clear cut mutations for both neurological disorders,and for metabolic or immune disorders,as I have.A multiple hit picture of causation,that might make perfect sense given the family history,and clinical picture,of each patient.Personalized medicine needs to be brought more into autism.

    Universal testing would let us know much sooner,exactly which genes are involved,and how common the mutations in these genes are.

  2. Fascinating Peter. We see more and more children with "Pitt Hopkins" but a very modified form. Two in particular wouldn't even be qualified as ID... they are musical savants, verbal, have perfect pitch, one is hyperlexic, but also have intense anxiety, adhd, and what some dx as 'Atypical autism"

  3. Been doing resveratrol personally for over a decade now and it really does help a lot with keeping your cardiovascular system working well, even if you end up being sedentary for bouts at a time. At higher doses it does impair wound healing a little bit and your nails and hair grow a bit slower, but that is part of the idea for doing what you can to decelerate aging a bit (use interventions to repair your existing cells, cleanout the broken cells, and try not to unnecessarily grow new ones so you don't wear out your stem cells).

    With respect to autism, I have used it for quite a while now for different reasons. For one, it is an MTORC1 inhibitor (indirectly though) while rapamycin is an MTORC1 and MTORC2 inhibitor (MTORC2 is where the immunosuppressant side effects come from). So it has been shown for a while now that MTORC is hyperactive, especially in cases involving macrocephaly (my son's head size is in the 99% range as is the case with many people in my family tree).

    Resveratrol also increases the NAD+ ratio via upregulating NAD recycling via the NAMPT enzyme (having a good NAD+/NADH ratio is important when it comes to oxidative stress because it tends to get depleted fast). Nicotanimide Riboside helps boost NAD+ through another direct mechanism as well.

    Resveratrol is also a PDE4 inhibitor and PDE4 has been shown to be hyperactive in the brains of those with classic autism.

    Lots of other reasons too, but don't expect it to be a miracle cure, but rather just another tool in battling oxidative stress over the long-term which if not dealt with will likely increase the severity of other autism related symptoms that are exacerbated by excess oxidative stress.

  4. It may be hopeless trying to identify defects in more than 700 autism genes, but wouldn't it be wise to check possible gene mutations according to individual pathological profile?
    Metabolic syndromes, at least in my country, are rarely suspected unless they manifest themselves with severe dysfunctions which lead to rapid organ failure.
    However, congenital metabolic syndroms can also run different courses leading to unknown cascade of events.
    After an abdominal and renal ultrasonic testing my son has been found to have fatty liver disease of medium degree, with some kind of 2,5cm cyst or atypical hemangioma or anything, together with enlarged spleen. We need to investigate further with magnetic testing.
    Peter, I know that you are not a doctor; I have my regular doctor and also my researcher doctor who are responsible for his treatment. This information is for your research and just in case you can see any possible connection with his autism.

    1. Petra NAFLD (Non-alcoholic fatty liver disease) is related to insulin resistance. It would also explain his odd cholesterol metabolism. So it may well be that there are connections between your son's type of Asperger's and his liver disease. If he has central insulin resistance (recall a post on "Type 3 diabetes),
      some of the drug therapies for NAFLD, may also help his Asperger's.

      Insulin sensitizers (metformin and thiazolidinediones) are commonly used for insulin resistance in those with NAFLD.[27] Improvements in liver biochemistry and histology in patients with NAFLD through treatment with statins have been observed in numerous cases, although these studies were carried out on a relatively small sample of patients.[28] Statins have also been recommended for use in treating dyslipidemia for patients with NAFLD. Treatment with pentoxifylline has demonstrated improvements in the histological appearance of fatty liver tissue under the microscope in many small trials"

      So Metformin, as in this above post, may give a double benefit. Ask your researcher doctor about intranasal insulin. Statins appear yet again.

    2. A lot of things can go wrong in development and cascade into other problems from a poor or at least challenging environment that have little to do with genetics and don't always leave epigenetic signatures that researchers can work backwards from.

      For example, today (could not find doi for the paper yet) there was an interesting press release here:

      where the research group believes they know how maternal inflammation (whether it be from infection, autoimmune disorders, or even obesity and the associated metabolic syndrome which is now seen by many to be at its core an inflammatory disease) may impair serotonin signaling in offspring and serotonin problems are one of the oldest and most studied areas of research with respect to autism. What seems to go awry is that the inflammatory response causes the placenta to turbocharge the production of an enzyme that converts tryptophan to 5-HTP. This causes a massive surge of serotonin into the body and brain of the fetus which in response the brain downregulates serotonin receptors and serotonin producing cells (need to read the actual paper to be 100% clear on this but that is what I surmise). So what you get is a brain that develops in utero thinking it has to downregulate the entire serotonin system because it is being hyperstimulated from the placental serotonin and then once the baby is born, you now have a brain no longer in a steady-state of supplemental serotonin from the placenta that may now have massive withdrawal like complications that causes further developmental problems for the baby and an impaired serotonin system for life.

      The point is that this was not the result of genetics (at least for the baby), not even epigenetics, but the result of a poor environment in utero that provided too much of a necessary neurotransmitter that ended up having many developmental side-effects that downstream could and probably will have epigenetic effects, but unless you can go back in time and see where things first went wrong (as this research group cleverly did), a naive researcher looking at this from the genetic or epigenetic angle is going to be grasping at straws.

      Anyways, with respect to NAFLD I know that the sugar "trehalose" has been found to be protective of NAFLD because it binds to the glucose receptor (one of the GLUTS and I don't remember which one) in the liver, thereby preventing glucose from signaling the cell should be in a growth state. This causes a side-effect of causing the liver cells to think that the body is in a state of fasting and causing the liver cells to undergo autophagy (trehalose btw has been shown to stimulate autophagy in other tissues as well including the brain, probably from a similar process). Autophagy is the cellular house cleaning process that cleans out the cellular junk and also helps the organism live in thrifty times and MTOR is one of its master regulators. Trehalose however seems to have an MTOR independent method of autophagy based on the process mentioned above.

      So you can fight sugar with sugar so to speak. Trehalose runs about 10 USA dollars per pound and you can get it from Swanson vitamins (at least that is where I have sourced it for the last 5 years).

    3. Hi Tyler
      My son has a PTEN mutation and macrocephaly. I'm targeting MTOR as well. It seems that I following a similar protocol as you. In currently giving my son: liposomal curcumin and resveratrol, trehalose, carnosine and ALA.
      As per Peter's advice I'm trying statins to increase PTEN.
      May I ask what doses and what other supplements are you giving your son?
      What kind of results have you seen?
      Thank you!

    4. Short answer Laura is quite a few, but specific to MTOR, aside from the strong drugs which have strong side effects (like rapamycin) resveratrol and spermidine are synergistic in downregulating MTOR. Spermidine in the USA cannot be sold in supplement form for some dumb reason the FDA gave, but it is in highest quantities in wheat germ and green peas. My son is very picky about food, but for a while I would mix with peanut butter the contents of resveratrol capsules, add a substantial amount of trehalose, and wheat germ (Bob's Red Mill is best I think). Then serve it up as a peanut butter sandwich. The trehalose helps deal with some of the sour taste of wheat germ.

      Trehalose helps promote autophagy but has nothing to do with MTOR as best I understand. Resveratrol and Spermidine inhibit separate enzymes that are both involved in MTOR expression (can't remember them both off the top of my head but I could dig out the paper for you if you wanted me to) that inhibit the same metabolic pathway to reduces MTOR expression. So you could say resveratrol and spermidine use different methods to achieve the same result so you are downregulating MTOR through two different mechanisms which is why they are synergistic (you can't just megadose resveratrol and get MTOR to go way way down as it has a floor and so does spermidine, but together they can both lower that floor).

      Other than those supplements, there are tons of other supplements/substances that very weakly have an effect on MTOR, but the most potent method of downregulating MTOR is fasting and caloric restriction which obviously makes more sense for grown adults, but may do more harm than good for children (the science is not clear on this yet).

      Another substance that may help with MTOR and works synergistically with resveratrol is nicotanimide riboside which I have mentioned here. It is pricey though. Just google "Niagen" if you want (all of the NR supplements just buy from the company that makes Niagen since they have the patents for efficiently manufacturing it).

      There are other interventions I employ for other reasons, but this is what I currently do for downregulating MTOR specifically.

  5. Peter, Metformin, intranasal insulin and statins, all look good options for a trial.
    My researcher doctor is recovering in hospital and we can't talk about my new findings. I know that he is doing research on liver encephalopathies and I am sure he is very familiar with NAFLD and diabets. Last time we met he told me to trial a probiotic called VLS 3 which I now see it was trialled with good results. I'll start with that.
    My regular doctor wants to have magnetic results first and then we could trial low dose metformin.
    Fatigue, brain on fire/ focus and sleeping disorder are symptoms I have to find possible solutions for.

    1. Petra, fatigue is a common symptom of NAFLD and established in research, but with no biological explanation.

      It would make sense that reduced insulin sensitivity would produce fatigue. As you apply therapies to improve insulin sensitivity you may well solve the fatigue.

      ALT, AST and ALT/AST ratio are common tests of liver function. They are often unusual in autism and some US doctors link this to mitochondrial dysfunction, I am not so sure. Did you test ALT and AST?

      Did you check thyroid function ? Lack of T3 would cause fatigue and this seems more common in people with autism.

      You can easily increase NO (nitic oxide) through diet and this might reduce fatigue. Athletes do this to increase exercise endurance. Either eat a lot of spinach or drink 150 ml of beetroot juice a day. It reduces vascular resistance so blood circulates more freely, and blood pressure is measurably lower.

  6. Correction VSL 3. Peter, if you wish to have a look, there are many articles on Pubmed about it. There is even one which says:
    "Modulation of intenstinal microbiota by probiotic VSL 3 resets brain gene expression and ameliorates the age-related deficits in LTP."


Post a comment