Different Types of Excitatory/Inhibitory Imbalance in Autism, Fragile-X & Schizophrenia

There is much written in the complex scientific literature about the Excitatory/Inhibitory (E/I) imbalance between neurotransmitters in autism. 

Many clinical trials have already been carried out, particularly in Fragile-X.  These trials were generally ruled as failures, in spite of a significant minority who responded quite well in some of these trials.

As we saw in the recent post on the stage II trial of bumetanide in severe autism, there is so much “background noise” in the results from these trials and it is easy to ignore a small group who are responders.  I think if you have less than 40%, or so, of positive responders they likely will get lost in the data. 

You inevitably get a significant minority who appear to respond to the placebo, because people with autism usually have good and bad days and testing is very subjective.

There are numerous positive anecdotes from people who participated in these “failed” trials.  If you have a child who only ever speaks single words, but while on the trial drug starts speaking full sentences and then reverts to single words after the trial, you do have to take note. I doubt this is a coincidence.

Here are some of the trialed drugs, just in Fragile-X, that were supposed to target the E/I imbalance:-

Metabotropic glutamate receptor 5 (mGluR5) antagonist

·        Mavoglurant

·        Lithium

mGluR5 negative allosteric modulator

·        Fenobam

N-methyl-D-aspartic acid (NMDA) antagonist

·        Memantine

Glutamate re-uptake promoter

·        Riluzole

Suggested to have effects on NMDA & mGluR5 & GABA_A

·        Acamprosate

GABA_B agonist

·        Arbaclofen

Positive allosteric modulator (PAM) of GABA_A receptor

·        Ganaxolone

Best not to be too clever

Some things you might use to modify the E/I imbalance can appear to have the opposite effect, as was highlighted in the comments in the post below:-

NMDAR hypo-function causing E/I imbalance in Autism and Schizophrenia – Baclofen, Sodium benzoate and Cinnamon (again)

So whilst it is always a good idea to try and figure things out, you may end up getting things the wrong way around, mixing up hypo and hyper.

The MIT people who work on Fragile-X are really clever and they have not figured it all out.

Fragile-X and Idiopathic Autism

Fragile-X gets a great deal of attention, because its biological basis is understood.  It results in a failure to express the fragile X mental retardation protein (FMRP), which is required for normal neural development.

We saw in the recent post about eIF4E, that this could lead to an E/I imbalance and then autism.

eIF4E inhibitors for Autism – Why not Ribavirin?

Our reader AJ started looking at elF4E and moved on to EIF4E- binding protein number 1.

In the green and orange boxes below you can find elF4E and elF4E-BP2.

This has likely sent some readers to sleep, but for those whose child has Fragile-X, I suggest they read on, because it is exactly here that the lack of fragile X mental retardation protein (FMRP) causes a big problem.  The interaction between FMRP on the binding proteins of elF4E, cause the problem with neuroligins (NLGNs), which causes the E/I imbalance.  Look at the red oval shape labeled FMRP and green egg-shaped NLGNs.

In which case, while AJ might naturally think Ribavirin is a bit risky for idiopathic autism, it might indeed be very effective in some Fragile-X.  You would hope some researcher would investigate this.

Can you have more than one type of E/I imbalance?

Readers whose child responds well to bumetanide probably wonder if they have solved their E/I imbalance.

I think they have most likely improved just one dysfunction that fits under the umbrella term E/I imbalance.  There are likely other dysfunctions that if treated could further improve cognition and behavior.

On the side of GABA, it looks like turning up the volume on α3 sub-unit and turning down the volume on α5 may help. We await the (expensive) Down syndrome drug Basmisanil for the latter, given that the cheap 80 year old drug Cardiazol is no longer widely available. Turning up the volume on α3 sub-unit can be achieved extremely cheaply, and safely, using a tiny dose of Clonazepam.

It does appear that targeting glutamate is going to be rewarding for at least some of those who respond to bumetanide.

One agonist of NMDA receptors is aspartic acid. Our reader Tyler is a fan of L-Aspartic Acid, that is sold as a supplement that may boost athletic performance.  

Others include D-Cycloserine, already used in autism trials; also D-Serine and L-Serine.

D-Serine is synthesized in the brain from L-serine, its enantiomer, it serves as a neuromodulator by co-activating NMDA receptors, making them able to open if they then also bind glutamate. D-serine is a potent agonist at the glycine site of NMDA receptors. For the receptor to open, glutamate and either glycine or D-serine must bind to it; in addition a pore blocker must not be bound (e.g. Mg2+ or Pb2+).

D-Serine is being studied as a potential treatment for schizophrenia and L-serine is in FDA-approved human clinical trials as a possible treatment for ALS/Motor neuron disease.  

You may be thinking, my kid has autism, what has this got to do with ALS/Motor neuron disease (from the ice bucket challenge)? Well one of the Fragile-X trial drugs at the beginning of this post is Riluzole, a drug developed for specially for ALS.  Although it does not help that much in ALS, it does something potentially very useful for some autism, ADHD and schizophrenia; it clears away excess glutamate.

Fragile-X is likely quite different to many other types of autism

I suspect that within Fragile-X there are many variations in the downstream biological dysfunctions and so that even within this definable group, there may be no universal therapies.  So for some people an mGluR5 antagonist may be appropriate, but not for others.

Even within this discrete group, we come back to the need for personalized medicine.

I do not think Fragile-X is a good model for broader autism.

Glutamate Therapies

There are not so many glutamate therapies, so while the guys at MIT might disapprove, it would not be hard to apply some thoughtful trial and error.

You have:

mGluR5

     ·        mGluR5 agonists (only research compounds)

·        mGluR5 positive allosteric modulators (only research compounds)

·        mGluR5 antagonists (Mavoglurant, Lithium)

·        mGluR5 negative allosteric modulators (Fenobam, Pu-erh tea decreases mGluR5 expression )

Today you can only really treat too much mGluR5 activity.  It there is too little activity, the required drugs are not yet available.  I wonder how many people with Fragile-X are drinking Pu-erh tea, it is widely available.

NMDA agonists

D-Cycloserine an antibiotic with similar structure to D-Alanine (D-Cycloserine was trialed in autism and schizophrenia)

ɑ-amino acids:

·         Aspartic acid (trialed and used  by Tyler, suggested for schizophrenia)

·         D-Serine (trialed in schizophrenia)

•    L-Serine

• D-Alanine (trialed in schizophrenia)

• L-Alanine

NMDA antagonists

·        Memantine (widely used off-label in autism, but failed in clinical trials)

·        Ketamine (trialed intra-nasal in autism)

Glutamate re-uptake promoters via GLT-1

·        Riluzole

·        Bromocriptine

·        Beta-lactam antibiotics

Synergistic Benefit of Low Dose Dopamine (Greek Coffee) and Diuretics (Bumetanide/Furosemide); better than Bromocriptine?

I did think of highlighting this post to the Bumetanide researchers in France, but I do not think they would take it seriously.

Another one to mention would be this new study, funded by Rodakis, to look at why some antibiotics improve some autism.  Dr Luna at Baylor College is running the study.  Its basic assumption is that the effect must be to do with bacteria, but as our reader Agnieszka has highlighted, common penicillin type antibiotics increase expression of the gene GLT-1 which then reduces glutamate in the brain.  It has nothing to do with bacteria.  Maybe for other antibiotics the effect does relate to bacteria.

But if you tell Dr Luna about GLT-1, quite likely she will not be interested.  

Study to investigate connection between antibiotic use and autism symptoms

Researchers will compare the gut microbiome (bacteria, yeasts and fungi found in the gut) and metabolome (small biological molecules produced by the microbes) of those who experience a change in symptoms during antibiotic use to those who do not. The study may provide valuable insight into when and why these changes occur and how this information can be harnessed for future interventions.  

There is even a case study very well documented here:-

Beta-Lactam Antibiotics as A Possible Novel Therapy for Managing Epilepsy and Autism, A Case Report and Review of Literature

Petra, our regular reader from Greece, has pointed out that Bumetanide has a greater effect in her adult son, with Asperger’s, when taken with Greek coffee and suggested why this might be. 

Her reference is this article:- 

Added to Diuretics, Low-Dose Dopamine Renoprotective in Small Acute-HF Trial

It shows that the diuretic effect of low dose furosemide, with dopamine, is greater than the effect of high dose furosemide.

The diuretic effect of Furosemide is via the transporter NKCC2, which is the same affected by Bumetanide. 

NKCC2 is found in your kidneys, while the very similar NKCC1 is found in your brain.  Furosemide and Bumetanide affect both NKCC1 and NKCC2.

The caffeine in coffee is known to indirectly produce dopamine in your body.

Greek coffee is nothing like your instant coffee or watery Starbucks coffee, it contains a serious amount of caffeine. 

The question is how does dopamine interact with furosemide/bumetanide and will the effect in the kidney (NKCC2) also affect the brain (NKCC1). 

By more effectively blocking NKCC1 in neurons you would further lower chloride levels and potentially further improve cognitive functioning.  

This would further validate Petra’s observation. 

Then we would consider if there is an alternative to Greek coffee, or just accept that caffeine is the simplest and safest method to enhance Bumetanide.    

In the then end my conclusion is that coffee, or just the caffeine, is a better option than a selective Dopamine D2 receptor agonist.  But there is an interesting drug called Bromocriptine that may be better in some cases. 

Not only is it a dopamine D2 receptor agonist, but Bromocriptine also “inhibits the release of glutamate, by reversing the GLT-1 (EAAT2) transporter”. 

We came across the GLT-1 (EAAT2) transporter when we found why some people with autism improve when on beta-lactam antibiotics (that include the penicillin ones).   

GLT-1/ EAAT2 is the principal transporter that clears the excitatory neurotransmitter glutamate from the extracellular space at synapses in the central nervous system. Glutamate clearance is necessary for proper synaptic activation and to prevent neuronal damage from excessive activation of glutamate receptors. EAAT2 is responsible for over 90% of glutamate reuptake within the brain 

We saw that the drug riluzole approved for the treatment of ALS (Amyotrophic Lateral Sclerosis) upregulates EAAT2/GLT-1.

I suggested that people with autism who improve on penicillin types antibiotics should get a similar effect from riluzole.  But riluzole is one of those monstrously expensive drugs.  

Based on my logic, we would then think that bromocriptine should help treat ALS (Amyotrophic Lateral Sclerosis).  What did I find when I looked it up:- 

Bromocriptine Mesylate Attenuates Amyotrophic Lateral Sclerosis: A Phase 2a, Randomized, Double-Blind, Placebo-Controlled Research in Japanese Patients. 

So then how much does Bromocriptine cost?  It is a cheap generic.  So a cost effective potential drug for ALS. 

Bromocriptine has two potentially useful functions (Dopamine D2 and GLT-1),but it has numerous other effects:- 

Bromocriptine blocks the release of a hormone called prolactin, but this should not be an issue for males. 

Risperidone, one of only two drugs approved for side effects of autism, can boost levels of prolactin.  Elevated prolactin levels are linked to a range of side effects, including gynecomastia, or growth of breasts, in men and boys.  This did not stop the drug being approved.

Bromocriptine agonizes the following monoamine receptors

• Dopamine D₁ family
• D₁ (K_i=682 nM)
• D₅ (K_i=496 nM)
• Dopamine D₂ family
• D₂ (K_i=2.96 nM)
• D₃ (K_i=5.42 nM)
• D₄ (K_i=328 nM)
• Serotonin 5-HT
• 5-HT_1A (K_i=12.9 nM)
• 5-HT_1B (K_i=355 nM)
• 5-HT_1D (K_i=10.7 nM)
• 5-HT_2A (K_i=107 nM)
• 5-HT_2B (K_i=56.2 nM)
• 5-HT_2C (K_i=741 nM)
• 5-HT₆ (K_i=33 nM)
• Adrenergic α family
• α_1A (K_i=4.17 nM)
• α_1B (K_i=1.38 nM)
• α_1D (K_i=1.12 nM)
• α_2A (K_i=11.0 nM)
• α_2B (K_i=34.7 nM)
• α_2C (K_i=28.2 nM)
• Adrenergic β family
• β₁ (K_i=589 nM)
• β₂ (K_i=741 nM)

  

This is why drugs have side effects. 

But for people with ALS who cannot afford riluzole, the cheap generic bromocriptine might be a good choice.

How about bromocriptine for autism? 

Well there was a trial in Italy a long time ago on girls with Rett syndrome 

Bromocriptine in the Rett syndrome

Twelve typical cases of the Rett syndrome and one forme fruste were treated with bromocriptine for six months and then had a washout for two months followed by resumption of the bromocriptine treatment. During the first bromocriptine treatment there were improvements in communication and relaxation in some of the girls: a more regular sleep pattern was observed in 4 and a more varied facial expression in 8, and 4 girls began to utter a few words. The bouts of hyperpnea disappeared in 5 and grinding of the teeth in 3. There was also a reduction in stereotypic hand activities in 5 girls and signs of improved motor abilities in 3. The washout caused a general decrease in the positive effects of the previously administered bromocriptine and resumption of the treatment with this drug led to less marked improvement. Metoclopramide was tested in all the girls before the treatment, and it was noted that, while endorphins were hyporesponsive, prolactin was hyperresponsive. This test was repeated two months after the bromocriptine treatment had been performed and, while beta-lipotropin remained unchanged, beta-endorphin showed increased responsiveness.

Current use of Dopamine with Lower Dose Diuretics 

There is extensive knowledge of the effect of taking dopamine with a bumetanide type diuretic. 

Bumetanide by itself has a plateau above which a higher dose causes no further diuresis, but when combined with dopamine there is more diuresis.  Alternatively you can use a lower dose of bumetanide and get the same amount of diuresis by adding dopamine. 

Of interest to people with autism, it is found that you can reduce the amount of potassium lost for the same amount of diuresis.

    

Added to Diuretics, Low-Dose Dopamine Renoprotective in SmallAcute-HF Trial

Synergetic Effects of Dopamine and High－Dose Bumetanide in Patients with oliguria

The effects of a combination of dopamine and bumetanide were studied in eight patients with oliguria not responsive to conventional treatment. Dopamine was infused at a rate of 3 чg/kg/min and bumetanide was given as a 0.05-0.1 mg/kg bolus every 2 hours intravenously. Administration continued for 3 to 15 days. Urine output, blood urea nitrogen, serum creatinine, the ratio of urine to plasma osmolarity, free water clearance, and serum electrolytes were measured before, during, and after the administration period. Six of the eight patients responded with an increase in urine output and improvement of the other variables ; the other two did not. We conclude that the combination of dopamine and high-dose bumetanide is effective in increasing diuresis in critically ill patients in the early stages of oliguria

How does dopamine interact with NKCC1/2?

This is a very logical question, but there is something in the literature on this subject.  It does come from frogs, but it was all I could find.

Functional expression of the murine D₂, D₃ and D₄ dopamine receptors in Xenopus laevis oocytes

The different murine D2-type dopamine receptors (D_2L, D_2S, D_3L, D_3S, and D₄) were expressed in Xenopus laevis oocytes. The D2-type receptors were all similarly and efficiently expressed in Xenopus oocytes and were shown to bind the D2 antagonist [¹²⁵I]sulpride. They were all shown to activate Cl⁻ influx upon agonist stimulation. Using the diagnostic inhibitor bumetanide, we were able to separate the Na⁺/K⁺/2Cl⁻ cotransporter component of the Cl⁻ influx from the total unidirectional Cl⁻ influx. The D_3L subtype was found to operate exclusively through the bumetanide-insensitive Cl⁻ influx whereas the other D2-type receptors acted on the Na⁺/K⁺/2Cl⁻ cotransporter as well. The pertussis toxin sensitivity of the receptor-activated chloride influx via the Na⁺/K⁺/2Cl⁻ cotransporter varied between the various D2-type receptors showing that they may couple to different G proteins, and activate different second messenger systems.

In contrast to the D2 and D3 receptor subtypes, D4 receptor activity was not significantly altered by the presence of PTX, suggesting that in Xenopus oocytes it may couple with one or more PTX-insensitive G proteins to cause changes in Cl3 influx. By contrast, in the case of the D2 receptor, PTX reduced the total Cl3 influx mediated by the D2S isoform by approximately 67%, and that mediated by the D2L isoform by approximately 40% (Fig. 2A). However, the activities of the two components of this ion influx, namely the bumetanide sensitive Na/K/2Cl- cotransporter and the bumetanide-insensitive Cl- influx, differed between these two isoforms. While the bumetanide-insensitive Cl3 influx was reduced by approximately 60% by PTX for the D2L isoform, it was only slightly reduced for the D2S isoform (Fig. 2C). Thus, the majority of the inhibitory effect of PTX on the D2S-induced influx was caused by uncoupling from the signalling cascade that activates the Na/K/2Cl- cotransporter. On the other hand, the signal transduction pathway that activates the cotransporter after stimulation of the D2L receptor remained relatively unaffected by PTX (Fig. 2B), indicating that D2S and D2L couple to different G proteins when expressed in Xenopus oocytes. For the D3 receptor, both long and short isoforms showed a reduction (50^60%) in the presence of PTX, at the bumetanide-insensitive Cl- influx (Fig. 2C), whereas for both D3 receptor isoforms, PTX had little or no effect on the Na/K/2Cl- cotransporter, indicated by the bumetanide-sensitive component of the Cl3 influx (Fig. 2B).  

PTX = pertussis toxin

  

Caffeine among its many effects is effectively a dopamine D2/3 receptor agonist.

Caffeine mimics the effect of a dopamine D2/3 receptor agonist onthe expression of immediate early genes in globus pallidus.

Conclusion

As I understand from the large scale trial use of bumetanide use in autism, there is indeed an issue with hypokalemia (loss of potassium).  

I would think that this should be solvable using a supplement and dietary potassium.  Agnieszka pointed out that kiwis have the advantage of potassium with little carbohydrate, as do avocados. Bananas and orange juice are the traditional potassium-rich foods for people on diuretics. 

This is a case where the care giver has to play an active role, it is not just about the doctor prescribing a pill.  The care giver has to manage the process to minimize the side effects.  So potassium needs to be managed, as does fluid intake. 

For people who struggle with hypokalemia, the idea of a lower dose of bumetanide, but with dopamine, could be interesting.  The other method is to add a potassium sparing diuretic like spironolactone. 

For my son, the dietary option, plus 250mg of potassium twice a day, is very effective.  Now I just have to persuade him to take a Greek coffee with his breakfast. 

For people whose autism responds to penicillin type antibiotics and who take bumetanide then Bromocriptine might be interesting as a caffeine alternative.