Stanford Community Symposium 2018: Phair, Metabolic traps, Tryptophan trap

[RDP]

2.2 Would the model predicts that for most cell types in the body where IDO 2 is not expressed behave as if IDO 2 was damaged? If so, does that mean that almost all cell types would be affected? And would everyone would be at risk of these cell types falling into the metabolic trap, regardless of their ID0 2 status?

If you have seen single cell RNAseq data for cell types where only IDO1 is expressed and not IDO2 or TDO, please point them out to me. One place to look is the Broad Institute at MIT where they are accumulating scRNAseq data on gene expression and have built a public portal to access the data.

Among the cells where I've seen scRNAseq data, I always see IDO1 and IDO2 expressed together. In dendritic cells, for example, all five cell types show this pattern and none of them express TDO.

Nevertheless, you are correct that any cell type expressing ONLY IDO1 is at risk for falling into the trap. Substrate inhibition is a very well-established feature of IDO1 enzyme kinetics which has been of considerable interest to hard core enzyme kineticists interested in the molecular mechanism of substrate inhibition in dioxygenases.

 

[RDP]

3.1 All the controls matched for high levels of damage to copies of IDO 2, so that the evidence indicates the problem is due to the trap being srung, not simply due to damaged IDO 2?

Good question. For the moment, we are relying on the high probability that any normal control has a 40% chance of being at least heterozygous for a broken copy of IDO2. Since our controls are sometimes unaffected relatives, our chance of getting heterozygotes is probably even greater. If the biochemical data support the trap theory, we will definitely genotype the controls as well as the patients. Peidong Shen (in Ron's group at Stanford) is setting up a new targeted genotyping assay for the key genes in the trap theory.

 

[RDP]

I dont recall Phair mentioning the TDO enzyme. Was it just implied?

The IDO metabolic trap cannot occur in any cell type where TDO2 is is expressed. The TDO2 gene product does not undergo substrate inhibition, so TDO2 is a backup for substrate-inhibited IDO1. If we or anyone else finds that TDO is expressed in midbrain serotonergic neurons, then the IDO metabolic trap can have no consequences for the serotonergic diffuse modulatory system of the CNS. There is some evidence for TDO2 expression in BRAIN, but it's the individual cell types that are critical.

If anyone finds literature data on IDO1, IDO2, or TDO2 expression in serotonergic neurons or in Raphe nuclei, I want to hear about it. Thanks.

 

[RDP]

I've just read Cort's overview of the hypothesis (I'm still to watch the OMF videos - are they broken down on YouTube yet?) and other than giving me a shudder of horror, recalling systems engineering lectures from almost two decades back, my main question is this: wouldn't an inhibition in IDO-mediated tryptophan catabolism, causing increased serotonin, lead to symptoms of serotonin syndrome?

Yes, depending on which isoform of TPH (the first enzyme in the Trp-to-serotonin pathway) is expressed in serotonergic neurons. One isoform is, itself, substrate inhibited so high intracellular Trp could cause low serotonin synthesis. The other (I'm guessing this is the one in serotonergic nerves) is normal Michaelis-Menten, so would lead to high serotonin and something like serotonin syndrome (although caused by high input rather than low removal).

 

[RDP]

3.2 Will the experiments also be measuring serotonin and NAD, if appropriate, to explain the symptoms of ME/CFS?

We are already measuring cell serotonin in our mixed population of PBMCs, but it's likely that only the dendritic cells are in the trap, so it may be hard to make the required measurements without first isolating the cell types we think are in the trap.

Since the kynurenine pathway (KP) is not the only way to make NAD+, we aim to measure KP metabolites down to the immediate precursor of NAD+, but as I've said in response to other posts, I don't think synthesis of NAD+ is the principal reason that inhibiting the KP is bad for a person. If you have neither the KP working nor niacin in your diet, then you have pellagra, a disease that was endemic in the southern US in the early 1900s, characterized by the infamous 3 D's, diarrhea, dementia, and death.

 

[RDP]

4 Statistics
Would it be appropriate to apply a statistical correction for the multiple comparisons made when looking at multiple individual predicted-damaging mutations?

Good critical question. When I first encountered the 5 mutations in IDO2, I was looking at genes one by one, manually. I started with all the mitochondrial proteins in the electron transport chain, then moved on to central energy metabolism. IDO2 was the 86th gene I looked at in the SIPS study. So absolutely, to avoid false positives, you could say I should do a Bonferroni correction and insist on p < 0.05/86. But consider this. None of the previous 85 genes had even a single damaging mutation that was common to even 3 of the SIPS patients. In IDO2, I found two pretty common predicted damaging mutations and 3 rare damaging mutations and every SIPS patient had at least one. Moreover, even SIPS patients who did not have the most common damaging mutation had at least one and sometimes three other damaging mutations in the same gene. These did not seem like clues to ignore, no matter what Bonferroni says.

And when I realized that IDO1 was substrate inhibited, the idea of the metabolic trap lept out at me. It was exactly the kind of thing I was hunting for - a bistable system.

So, for me, statistics is the last resort. This process of looking for common damaging mutations began with a conviction based on the existence of ME/CFS epidemics. It's just logic. But I'm a classic Popperian hypothesis tester. Now that the hypothesis is formulated it makes predictions I can test. Testing hypotheses. This is where science gets done. It's also the hardest part. Nature does not make it easy to test hypotheses well.

 

[RDP]

Despite dietary niacin, my vitamin B3 levels were below 15% of the RDA. So it could be possible for some of us, but perhaps it's a gut issue preventing absorption?

I'm assuming your "below 15% of the RDA" means below 15% of the normal reference range for plasma vit B3. You're right there could be a limitation in niacin transport, but some is clearly getting in. And fortunately, NAD+ is not consumed in energy production, just cyclically converted from NAD+ to NADH during fuel oxidation, and then back again as it delivers electrons to the mitochondrial electron transport chain. That's why it's called a cofactor.

As I mentioned in other posts, if you had no source of NAD+, you'd have pellagra.

 

[alex3619]

Exactly. Though 'simultaneous' may not be required; rather:

• Trigger
• Second trigger before recovery from first trigger

....which may not require the triggers to occur simultaneously per se.

Which is what I have long called the two trigger hypothesis.

ps, now I recall I called it the two hit hypothesis back in the XMRV hype days.

PPS I have started calling this metabolic trap the tryp-trap.

 

[alex3619]

There is some evidence for TDO2 expression in BRAIN, but it's the individual cell types that are critical.

This seems very clear to me. In an intracellular problem the specific cell types are important ... for example its not clear what IDO2 expression is in different cell lines when you look up brain tissue expression. I am very glad some are looking at individual gene expression, and that stem cell experimentation (presumably differentiating the cells into the different cell lines) is being looked at.

 

[mariovitali]

@RDP Thank you posting here and for opening this channel of communication. The more i look at kynurenine / indoleamine, the more exciting it gets.


I would like to take the opportunity to give another perspective regarding Niacin and NADH. In the document (the first version was circulated in October 2017) i sent to Dr Davis and Robert Phair -among others- we find the following on page 29 :









We see how Network Analysis identifies NAD as a critical element of ME/CFS (Hypothesis) and Tocotrienol. Given the potential importance of these topics i decided to try them on myself early on. The first one to try was NADH :

https://www.webmd.com/vitamins/ai/ingredientmono-1016/nadh


The result is that i crashed. I also tried Tocotrienol and Niacin. My body was not happy with them, at all. In the document i also hypothesize on why Tocotrienol supplementation may not be a good idea, the reason being is that it inhibits Hmg-CoA just like Statins.

An interesting question would be if statins are well tolerated by ME/CFS patients but it appears that statins may be responsible for causing Fatigue and tiredness on non-ME-CFS patients :

https://universityhealthnews.com/daily/heart-health/new-study-ties-statins-to-excessive-tiredness/


Here is one study regarding Niacin and Liver Damage :

Niacin induced coagulopathy as a manifestation of occult liver injury.
Ali EH1, McJunkin B, Jubelirer S, Hood W.
Author information

Abstract
Niacin is an effective lipid-lowering agent which occasionally may cause hepatic failure. Liver enzymes are periodically tested during niacintherapy to assess for hepatic injury. We report a case of suppressed synthesis of hepatically derived coagulation factors and other liverproteins in a patient on niacin with no elevation of hepatic aminotransferases. The protein abnormalities reversed rapidly on discontinuation of niacin. It appears that niacin can cause occult liver injury without frank aminotransferase elevations, and may portend severe hepatotoxicity. Periodic assessment of prothrombin time should be considered in addition to aminotransferase levels to screen for liver injury. We believe this is the first reported case of occult hepatic injury due to extended release niacin, presenting as coagulopathy.

I would therefore like to express my concerns in using Niacin, especially to ME/CFS patients that are severely ill. Normal Liver function tests do not rule out Liver disease. I also hypothesise that the crashes by Whey protein that @RDP mentioned can be attributed to impaired Liver function (there is tons of info regarding the perils of protein shakes and associated Liver / Kidney problems)

EDIT FOR CLARIFICATION : I did not try the extended release version of Niacin that is mostly related to Liver damage but the non-extended release version of this supplement.

 

[Hoopoe]

Out of curiosity, does the metabolic trap hypothesis fit with gradual illness onset? About 2/3 of patients report an infection that triggered the illness, and most of the remaining patients report a more gradual onset. The proportion of gradual onset cases may be even higher because acute onset is more easily recognized.

PS: I understand that the body responds to infection by increasing production of kynurenine. If there is no infection in gradual onset cases how do they end up in the metabolic trap? If they end up in it and don't have a different kind of problem. Are all the patients tested so far infectious onset cases?

 

[Simon M]

@RDP Let me echo @Trish's thanks to you for joining the forum and the discussion (it's great to have researchers here) and add my own thanks for the detailed responses to my many questions.

I'd like to respond to a few of your points, grouped below into a few themes:

Focusing on individual cell types

This would seem to be an important issue:

What we need is single cell RNAseq (scRNAseq) and many labs are now doing this for individual cell types.

I agree that scRNAseq is the best form of evidence.

If you have seen single cell RNAseq data for cell types where only IDO1 is expressed and not IDO2 or TDO, please point them out to me. One place to look is the Broad Institute at MIT where they are accumulating scRNAseq data on gene expression and have built a public portal to access the data.

I haven't seen any scRNA data, but the fact that tissue expression levels are much higher for IDO1 than for IDO2 means it would be a surprise if all cells express both.

The IDO metabolic trap cannot occur in any cell type where TDO2 is is expressed. The TDO2 gene product does not undergo substrate inhibition, so TDO2 is a backup for substrate-inhibited IDO1. If we or anyone else finds that TDO is expressed in midbrain serotonergic neurons, then the IDO metabolic trap can have no consequences for the serotonergic diffuse modulatory system of the CNS. There is some evidence for TDO2 expression in BRAIN, but it's the individual cell types that are critical.

Interesting - so that’s three enzymes for each cell type that are important in evaluating the metabolic trip.

I know your theory focuses on the explanatory power of changes to the Raphe nuclei in the brain, but is it also worth looking in the gut? I understand that most serotonergic cells lie there, both neurons and mucosal cells.

One place to look is the Broad Institute at MIT where they are accumulating scRNAseq data on gene expression and have built a public portal to access the data.

As the single cell RNA work seems to be so important to establish which cell types would actually affected, I wondered if patients can help with this work? I'm afraid I'm too ill to do so, but perhaps if you think this is worth pursuing you can say so and see if there are any volunteers (the power of the crowd?)


Statistics

So, for me, statistics is the last resort.

I think we will have to disagree on the importance of stats, but I do agree with you that, ultimately, the science comes down to putting the (any) hypothesis to the test.

And I'm impressed by the studies you have in hand especially as they 1) already include measuing serotonin, 2) that if the Trap checks out you will be genotyping controls to check they have comparable levels of broken IDO levels (see next para); and 3) especially the plans for deriving and testing sertonergic neurones from adult stem cells.

Peidong Shen (in Ron's group at Stanford) is setting up a new targeted genotyping assay for the key genes in the trap theory.

Excellent.


Miscellaneous

Kynurenine (Kyn) is important as an immunoregulator. Kyn or Kyn metabolites control the function of Treg cells and thus also control the expansion of effector T cells. If an APC [antigen presenting cell, such as a dendritic cell, needed to kick off B cell and T cell differentiation] is in the trap and not making Kyn, then T cell expansion takes off and you get autoimmune symptoms.

Potentially that could explain the Mark Davis finding of T cell clonal expansion, a personal interest of mine.

Re my Q as to why it takes 10 days to activate the trap and 8 weeks to make it irreversible.

In biological terms, what happens is that (after the point of no return) IDO1 is now inhibited by its own substrate, Trp. Importantly this can only happen in cells where the kynurenine pathway is the major degradation route for Trp and where TDO2 is not expressed. It becomes irreversible because it is no longer self-correcting. Instead, as Trp increases, IDO flux decreases, and as IDO flux decreases, Trp increases, and so on.

So if I've understood you correctly, tryptophan accumulation is pretty slow, and the long time to trigger the trap reflects this.

----​

Thanks again for joining the forum and for all your answers. Patients would benefit from more researchers being as willing to engage as you have been.

 

[Barry]

Great to have you join us here @RDP .

My wife's ME seemed to have been triggered when she came down with a flu bug just a day after she had a thyroid operation. I'm not a scientist, but could this double-hit be the kind of thing that might flip someone into this metabolic trap?

 

[Chris Ponting]

Good critical question. When I first encountered the 5 mutations in IDO2, I was looking at genes one by one, manually. I started with all the mitochondrial proteins in the electron transport chain, then moved on to central energy metabolism. IDO2 was the 86th gene I looked at in the SIPS study. So absolutely, to avoid false positives, you could say I should do a Bonferroni correction and insist on p < 0.05/86. But consider this. None of the previous 85 genes had even a single damaging mutation that was common to even 3 of the SIPS patients. In IDO2, I found two pretty common predicted damaging mutations and 3 rare damaging mutations and every SIPS patient had at least one. Moreover, even SIPS patients who did not have the most common damaging mutation had at least one and sometimes three other damaging mutations in the same gene. These did not seem like clues to ignore, no matter what Bonferroni says.

And when I realized that IDO1 was substrate inhibited, the idea of the metabolic trap lept out at me. It was exactly the kind of thing I was hunting for - a bistable system.

So, for me, statistics is the last resort. This process of looking for common damaging mutations began with a conviction based on the existence of ME/CFS epidemics. It's just logic. But I'm a classic Popperian hypothesis tester. Now that the hypothesis is formulated it makes predictions I can test. Testing hypotheses. This is where science gets done. It's also the hardest part. Nature does not make it easy to test hypotheses well.

Interesting. There are 174 different missense and loss-of-function alleles for IDO2 listed in EXAC, so the multiple testing correction really should be for 174 different tests. Quite a large burden of testing.

From UK Biobank there is no signal for (self-reported CFS/ME): http://geneatlas.roslin.ed.ac.uk/re...5&minregion=39741&chrom=8&representation=plot

 

[Amw66]

Interesting. There are 174 different missense and loss-of-function alleles for IDO2 listed in EXAC, so the multiple testing correction really should be for 174 different tests. Quite a large burden of testing.

From UK Biobank there is no signal for (self-reported CFS/ME): http://geneatlas.roslin.ed.ac.uk/re...5&minregion=39741&chrom=8&representation=plot

would love to know what this meant?

 

[Susan K]

Good critical question. When I first encountered the 5 mutations in IDO2, I was looking at genes one by one, manually. I started with all the mitochondrial proteins in the electron transport chain, then moved on to central energy metabolism. IDO2 was the 86th gene I looked at in the SIPS study. So absolutely, to avoid false positives, you could say I should do a Bonferroni correction and insist on p < 0.05/86. But consider this. None of the previous 85 genes had even a single damaging mutation that was common to even 3 of the SIPS patients. In IDO2, I found two pretty common predicted damaging mutations and 3 rare damaging mutations and every SIPS patient had at least one. Moreover, even SIPS patients who did not have the most common damaging mutation had at least one and sometimes three other damaging mutations in the same gene. These did not seem like clues to ignore, no matter what Bonferroni says.

And when I realized that IDO1 was substrate inhibited, the idea of the metabolic trap lept out at me. It was exactly the kind of thing I was hunting for - a bistable system.

So, for me, statistics is the last resort. This process of looking for common damaging mutations began with a conviction based on the existence of ME/CFS epidemics. It's just logic. But I'm a classic Popperian hypothesis tester. Now that the hypothesis is formulated it makes predictions I can test. Testing hypotheses. This is where science gets done. It's also the hardest part. Nature does not make it easy to test hypotheses well.

I love this reply. And the statistics issue is consistent with my experience using multiple test correction. Bonferroni is overly strict, and one loses good (true) biological data when it is used.

 

[adambeyoncelowe]

@RDP I wonder how the metabolic trap fits (if at all) with brain imaging studies?

I've seen recent Japanese PET research which links microglial activation with increased neuronal serotonin: https://www.s4me.info/threads/glial...-fatigue-syndrome-noda-et-al-japan-2018.6520/

There was also the study which suggested reduced signal in the brainstem was being compensated for by increased signal (and myelin) in the mid-brain: https://www.s4me.info/threads/hyper...m-abnormality-in-cfs-2018-barnden-et-al.5032/ VanElzakker expands upon this and thinks the brainstem abnormalities actually might be a vagus nerve problem (which is located right next to where this study found problems).

Both seem to tie into previous findings (the brainstem is a common one, and PET studies into microglia and astrocytes are more recent ones). Could a metabolic trap fit with any of these findings?

 

[Chris Ponting]

Interesting. There are 174 different missense and loss-of-function alleles for IDO2 listed in EXAC, so the multiple testing correction really should be for 174 different tests. Quite a large burden of testing.

From UK Biobank there is no signal for (self-reported CFS/ME): http://geneatlas.roslin.ed.ac.uk/re...5&minregion=39741&chrom=8&representation=plot

would love to know what this meant?

Sorry for not explaining. I think the question is: "Is the DNA variation in IDO2 unusual?" To me here 'unusual' means compared to all 20,000 genes across the entire human genome. The UK Biobank weblink I provided shows that IDO2 DNA variation is *not* unusual. Scroll down in the linked website and you'll see a cluster of dots hugging the X-axis. Each of these dots is a DNA change and the Y-axis shows the probability that this DNA change predicts whether someone has ME/CFS or not. The higher up on the plot the greater this DNA change can predict ME/CFS status. The critical thing on the plot is the horizontal red dashed line: this indicates the level of at which genome-wide significance is achieved. A blue dot (a DNA change) has to be *above* this line to be significant across the entire human genome and across all 20,000 genes. In other words, IDO2 is *not* unusual in how its DNA variation predicts ME/CFS over control status. There will be tens of thousands of other DNA variants *elsewhere* in the genome (near to other genes) that will better predict ME/CFS over control status. And even these are not statistically significant. For significance, we need to find blue dots that go above the dashed red line. Summary: UK Biobank data does not support the idea that IDO2 DNA variation is causal of ME/CFS. Hope this helps.

 

[ScottTriGuy]

No doubt this has been raised before, so please forgive my faulty memory - given that the ME community has a wide spectrum of severity, and a wide spectrum of symptoms, it may be that it is more than one gene is at play in ME, it may be a cluster of genes that determine severity and symptoms.

Does the Biobank data analyze for significance of cluster of genes for ME? Does IDO2 in combination with one or more other genes, indicate ME?

 

[Amw66]

Sorry for not explaining. I think the question is: "Is the DNA variation in IDO2 unusual?" To me here 'unusual' means compared to all 20,000 genes across the entire human genome. The UK Biobank weblink I provided shows that IDO2 DNA variation is *not* unusual. Scroll down in the linked website and you'll see a cluster of dots hugging the X-axis. Each of these dots is a DNA change and the Y-axis shows the probability that this DNA change predicts whether someone has ME/CFS or not. The higher up on the plot the greater this DNA change can predict ME/CFS status. The critical thing on the plot is the horizontal red dashed line: this indicates the level of at which genome-wide significance is achieved. A blue dot (a DNA change) has to be *above* this line to be significant across the entire human genome and across all 20,000 genes. In other words, IDO2 is *not* unusual in how its DNA variation predicts ME/CFS over control status. There will be tens of thousands of other DNA variants *elsewhere* in the genome (near to other genes) that will better predict ME/CFS over control status. And even these are not statistically significant. For significance, we need to find blue dots that go above the dashed red line. Summary: UK Biobank data does not support the idea that IDO2 DNA variation is causal of ME/CFS. Hope this helps.

Thank you.