List of causation hypothesis for follow up after DecodeME

[SNT Gatchaman]

Re: MAIT T cells being found in the meninges in Mucosal-associated invariant T cells restrict reactive oxidative damage and preserve meningeal barrier integrity and cognitive function (2022)

It has nothing to do with trafficking

Yes the claim is that they're tissue-resident, as part of the barrier function of meninges. From the editorial on the Zhang paper Meningeal MAIT cells maintain meningeal and brain function (2022) —

the meninges, a trilayer structure composed of the parenchyma-adjacent pia, the highly vascular arachnoid, and the outermost dense dura mater, contain various resident lymphocyte subsets, including innate lymphoid cells and γδ T cells in the non-inflamed state. The presence of these immune cells may affect cognition and memory, although the mechanisms underlying their function in this context are incompletely understood. Zhang et al. now add MAIT cells to the lymphocyte meningeal repertoire, exploring their importance for the integrity of protective barriers and how their loss affects homeostatic mechanisms that limit neuroinflammation.

 

[chillier]

I’m not a neuroscience person so maybe I’m missing some important context—but if it had to do with something traveling down a neuron that wasn’t an action potential, wouldn’t you expect the progression of PEM to be more, well, progressive across the body based on shortest to longest axonal length? I definitely get whole-body muscle weakness, stiffness, and pain as part of PEM and sometimes symptoms have a temporal offset between them, but not in any way that would correspond to neuron length. For me, it’s more that pain/stiffness/weakness is the worst in the muscle groups that I was using during activity, and then everywhere else it builds at the same rate and around the same intensity.

For that particular idea, it wouldn't be that you would feel PEM in different places over time. It would be that signals from eg just used muscles would take a while to reach the neuron nuclei and cause a transcriptional response, but once it did the whole neuron could be eg sensitised at once leading then to PEM (which is a whole body feeling, at least for me). In other words the length of the axons would correspond to how soon PEM is felt, not where it is felt.

Anyway I think the scaling idea might be a bit better.

 

[jnmaciuch]

I just wonder whether it would be worth looking for links with migraine on the basis that that seems to be another situation where a brain loop generates signals that should not be there.

Could you share any relevant papers about this? I read up on migraine quite often since several people in my life suffer from them, and have not come across any papers showing that migraine occurs when the brain generates a signal that should not be there. Spreading depolarization is the opposite phenomenon and next to nothing is known about what causes SD.

 

[DMissa]

@DMissa not sure if this is the right place to post this, but if you’re doing any high throughput screeen automation I’m happy to donate my time. I’m an automation engineer with 7+ years of experience in biotech specializing in high throughput screening on liquid handlers: Hamilton, Beckman (Bravos & echo), Tecans, thermo-fisher work cells, and custom integration work cells.

I’ve ask OMF in the past but have learned it’s cheaper outside of biotech to just hire grad students to pipette instead of buying robotics. If anyone working on CFS needs automation consulting happy to donate my time.

Thank you, and noted, but yes as you might imagine the obstacle is capital

 

[Hutan]

Please discuss the utility of AI/large language models for generating hypotheses here:
Can Large Language Models (LLMs) like ChatGPT be used to produce useful information?

We've moved some posts on that topic to that thread.

 

[Kitty]

For that particular idea, it wouldn't be that you would feel PEM in different places over time. It would be that signals from eg just used muscles would take a while to reach the neuron nuclei and cause a transcriptional response, but once it did the whole neuron could be eg sensitised at once leading then to PEM (which is a whole body feeling, at least for me).

Could molecules such as prostaglandins interact with this idea?

I keep thinking about the 'immune' symptoms in PEM, and prostaglandins could potentially be behind head-cold and gut symptoms. (They may also be implicated in 'period 'flu', the symptoms of which are the same as day 1 PEM.)

One of the things that prompts release from cells is injury, and I've wondered if alert signals could somehow be triggered when there is no injury, just activity.

Ignore me if I'm talking rubbish again, I really don't know anything!

 

[Jonathan Edwards]

have not come across any papers showing that migraine occurs when the brain generates a signal that should not be there. Spreading depolarization is the opposite phenomenon and next to nothing is known about what causes SD.

Depolarisation is signalling (especially if it spreads) and if it is abnormal and associated with symptoms it should not be there!

 

[chillier]

Could molecules such as prostaglandins interact with this idea?

Ignore me if I'm talking rubbish again, I really don't know anything!

Good question I'm afraid I'm too ignorant to know! I don't think it's talking rubbish - it can only be talking rubbish if someone is pretending to know something they don't.

 

[chillier]

How about something like this as a model for how fragile synapses could produce ME/CFS like disease dynamics:



- Using here a hypothetical set up of two neurons that could explain the dynamics of PEM, and how you to get locked into a disease state that could be escapable in some situations. Obviously it would have to scale up to many neurons.

- assuming that the problem the genetics point to with synapses is that they are fragile and prone to downscaling, breakdown, and/or loss of connectivity.

- There would be a threshold you cross that leads to the disease state based on the ratio of rate of synapse breakdown:rate of synapse repair

- Neuron A is the presynaptic neuron that connects to the postsynaptic neuron: Neuron B

- When neuron A fires it triggers neuron B to fire, in the process the AB synapse would be slightly weakened because of local downscaling of the postsynapse. This weakening could be made more pronounced by genetic background (lack of energy supply, loss of shank complex, lack of receptor recycling etc)

- In steady state conditions the maintenance and repair of the synapse would balance the downscaling due to the firing of neuron A.

- If neuron A bombards synapse AB with signal, firing again and again, synapse repair would not keep up and the synapse loses a lot if not all of its ability to signal.

- The rate of synapse repair would not be constant, and would be much slower for a highly compromised synapse, and now even if neuron A returns to a normal or low level of firing, it is now still too much for synapse repair to keep up with, and we are now trapped in this state, unless neuron A massively reduces its firing rate to not interfere with synapse repair (aka, forcing the individual to rest).

- meanwhile, over the next 24-48 hours, neuron B realises it has lost signal from neuron A and is not firing as much as it was previously. To maintain homeostatic firing rates, neuron B would increase its sensitivity (via global synaptic upscaling of all the other synapses and other changes) and start amplifying and firing noise. In this way you could have a loss of signal that leads to things like brain fog, but also in other situations things like sensitivity to bright lights, that all occur on a PEM timescale.

- To knock someone into a disease state, you would either need excessing firing of a neuron(s) (achievable by acetylcholinesterase inhibitors or overtraining), or an increased fragility of the synapse (maybe inducible by inteferon??). Having both would help.

- Once you're in the disease state, you're in a trap where if your neurons fire much more, it pushes you further into the disease state, repeatedly inducing PEM as it does, tightening the trap each time.

 

[Utsikt]

Once you're in the disease state, you're in a trap where if your neurons fire much more, it pushes you further into the disease state, repeatedly inducing PEM as it does, tightening the trap each time.

What would this model predict to happen if someone with mild ME/CFS did extreme rest for weeks/months?

Would that allow the repair to catch up and reach a sustainable non-diseases state?

 

[EndME]

What would this model predict to happen if someone with mild ME/CFS did extreme rest for weeks/months?

Would that allow the repair to catch up and reach a sustainable non-diseases state?

Would it predict anything? How does lying in bed effect the firing rate of a local region?

Perhaps sleeping a lot might reduce overall firing rate, but perhaps the problem is precisely occuring in regions of neurons that also predominantly fire at night? What about an induced coma?

 

[EndME]

How about something like this as a model for how fragile synapses could produce ME/CFS like disease dynamics:



- Using here a hypothetical set up of two neurons that could explain the dynamics of PEM, and how you to get locked into a disease state that could be escapable in some situations. Obviously it would have to scale up to many neurons.

- assuming that the problem the genetics point to with synapses is that they are fragile and prone to downscaling, breakdown, and/or loss of connectivity.

- There would be a threshold you cross that leads to the disease state based on the ratio of rate of synapse breakdown:rate of synapse repair

- Neuron A is the presynaptic neuron that connects to the postsynaptic neuron: Neuron B

- When neuron A fires it triggers neuron B to fire, in the process the AB synapse would be slightly weakened because of local downscaling of the postsynapse. This weakening could be made more pronounced by genetic background (lack of energy supply, loss of shank complex, lack of receptor recycling etc)

- In steady state conditions the maintenance and repair of the synapse would balance the downscaling due to the firing of neuron A.

- If neuron A bombards synapse AB with signal, firing again and again, synapse repair would not keep up and the synapse loses a lot if not all of its ability to signal.

- The rate of synapse repair would not be constant, and would be much slower for a highly compromised synapse, and now even if neuron A returns to a normal or low level of firing, it is now still too much for synapse repair to keep up with, and we are now trapped in this state, unless neuron A massively reduces its firing rate to not interfere with synapse repair (aka, forcing the individual to rest).

- meanwhile, over the next 24-48 hours, neuron B realises it has lost signal from neuron A and is not firing as much as it was previously. To maintain homeostatic firing rates, neuron B would increase its sensitivity (via global synaptic upscaling of all the other synapses and other changes) and start amplifying and firing noise. In this way you could have a loss of signal that leads to things like brain fog, but also in other situations things like sensitivity to bright lights, that all occur on a PEM timescale.

- To knock someone into a disease state, you would either need excessing firing of a neuron(s) (achievable by acetylcholinesterase inhibitors or overtraining), or an increased fragility of the synapse (maybe inducible by inteferon??). Having both would help.

- Once you're in the disease state, you're in a trap where if your neurons fire much more, it pushes you further into the disease state, repeatedly inducing PEM as it does, tightening the trap each time.

So the end state of ME/CFS would be: Specific neurons firing too much resulting in symptoms.

Now my understanding is that such illnesses exist. My layman understanding is: In epilepsy neurons fire abnormally easily, in channelopathies (but you can also have the opposite problem of hypoexcitable neurons) you probably have a whole host of conditions that lower the firing action potential threshold (maybe something quite similar occurs in migraines with aura) and potentially something similar might be happening in certain cases of neuropathic pain.

So why does ME/CFS look different? Clearly it's not like epilepsy, so you can't have dynamics that allow for a sudden, synchronized, high-frequency firing, which means in the above model you'd need something that stops such a positive feedback loop from happening. Some sort of dynamic that allows for "a bit more firing but not arbitrary much and generally still some sort of steady-state rather than something progressive"?

 

[chillier]

So the end state of ME/CFS would be: Specific neurons firing too much resulting in symptoms.

Now my understanding is that such illnesses exist. My layman understanding is: In epilepsy neurons fire abnormally easily, in channelopathies (but you can also have the opposite problem of hypoexcitable neurons) you probably have a whole host of conditions that lower the firing action potential threshold (maybe something quite similar occurs in migraines with aura) and potentially something similar might be happening in certain cases of neuropathic pain.

So why does ME/CFS look different? Clearly it's not like epilepsy, so you can't have dynamics that allow for a sudden, synchronized, high-frequency firing, which means in the above model you'd need something that stops such a positive feedback loop from happening. Some sort of dynamic that allows for "a bit more firing but not arbitrary much and generally still some sort of steady-state rather than something progressive"?

I think it's maybe not quite that. One thing I believe that stops a positive feedback loop from happening is the global scaling of the firing rate of a neuron, and in this model that is happening: Neuron B loses its input from Neuron A so it globally upscales its firing rate to compensate - its overall firing rate would be more or less the same, but now it doesn't transmit a coherent signal, but just noise and maybe at inopportune times making for a jarring, confusing experience.

Like if you had a really dim image taken at night and couldn't see what was on it, so you raise the contrast to the maximum to create a greyish white image with a lot of static and not much detail. The average pixel brightness might still be the same as a crystal clear photo taken in the day.

 

[chillier]

What would this model predict to happen if someone with mild ME/CFS did extreme rest for weeks/months?

Would that allow the repair to catch up and reach a sustainable non-diseases state?

I think that would give you the best chance, that's my feeling about the way it works on a practical level too. The more rest the better. The problem is the neurons don't stop firing and life is random and tricky and you can just relapse anyway. That's just my experience, but the model would fit with that.

 

[EndME]

I think it's maybe not quite that. One thing I believe that stops a positive feedback loop from happening is the global scaling of the firing rate of a neuron, and in this model that is happening: Neuron B loses it's input from Neuron A so it globally upscales its firing rate to compensate - its overall firing rate would be more or less the same, but now it doesn't transmit a coherent signal, but just noise and maybe at inopportune times making for a jarring, confusing experience.

Like if you had a really dim image taken at night and couldn't see what was on it, so you raise the contrast to the maximum to create a greyish white image with a lot of static and not much detail. The average pixel brightness might be still be the same as a crystal clear photo taken in the day.

Thanks, I think that makes things clearer to me. So if I'm understanding correctly: In ME/CFS synapses become broken more easily or stop functioning as they should. This doesn't change network firing rates due to synaptic scaling (for example if Neuron B looses connection to Neuron A it now strengthens the connections to the other Neurons it's connected with), if this happens to sufficiently many neurons/synapses they fire less "optimally" (I guess this is a bit similar to alzheimers but instead of a "permanent damage model" leading to a network "without signal propagation" this is more of a "constant reorganisation model depending on which synapses currently can be repaired sufficiently" leads to "slightly more chaotic signalling").

I'm guessing you still have to rule out certain effects, maybe something "like synchronisation of noisy signal when many synapses are broken leading to epileptic like scenarios or loss of too synapses leading to complete loss of signal" but maybe that's quite possible if there is "always enough synapses still sufficiently functioning/enough network balance to stop synchronisation".

I'm guessing the model works quite well for excitatory and inhibitory synapses (in the one case you upscale your remaining connections in the other you downscale the network, both leading to more chaotic signalling)? In my head this model is a bit more similar to Alzheimers. But I guess one could one perhaps also propose the opposite model: Too many functioning synapses leading to an overorganised network that however avoids the problems occuring in epilepsy due to enough reorganisation?

I suppose there's also a slightly different option related to synapses (that is admittdely is extremely handwavy): Impaired homeostatic plasticity. Synaptic problems cause an impaired homeostatic plasticity (instead of a thermostat keeping the temperature constant you now have local peaks here and there, however you don't end up with a freezing bathroom and a blazing hot kitchen because there's still "sufficient plasticity" as the doors separting the two are never closed). In some areas you have too little scaling up of networks causing short term memory loss and brain fog and in other areas you have too much scaling up of networks causing sensory hypersensitivity. You swing out of balance locally, but not into extremes. Activities related to neural processing in certain areas cause further shifts in dysregulation.

 

[Kitty]

But I guess one could one perhaps also propose the opposite model: Too many functioning synapses leading to an overorganised network that however avoids the problems occuring in epilepsy due to enough reorganisation?

Yes, that's an interesting thought. Presumably it could just as easily work either way.

 

[Sean]

Using here a hypothetical set up of two neurons that could explain the dynamics of PEM, and how you to get locked into a disease state that could be escapable in some situations. Obviously it would have to scale up to many neurons.

With the extra possible/likely complication that a pair of neurons interacting may be a different phenomena to a few million interacting. Scaling up may introduce additional factors, that might also feedback into how single pairs interact.