Peripheral neurons, CRH, and sickness behavior

[ScoutB]

That Familial Episodic Pain Syndrome paper (apparently I only remember the most recent paper I've read) had an interesting example of how the ion channel works differently in different species:

A fascinating aspect of TRPA1 activation lies in interspecies discrepancies. Some electrophilic thioaminal-containing compounds are able to activate rat TRPA1 while others inhibit human TRPA1 (Chen et al., 2008). This difference is due to key amino acids situated in the S6 segment. A similar story applies to caffeine, which was shown to activate mouse TRPA1 but to inhibit the human channel (Nagatomo and Kubo, 2008), and menthol, which activates mammalian TRPA1, blocks nonmammalian TRPA1 (Xiao et al., 2008), and exerts a bimodal action on murine TRPA1 (Karashima et al., 2009). Therefore, similar compounds are able to exert opposite effects in different species and this suggests a complex gating mechanism for TRPA1 in which amino acid substitutions at key positions determine agonist binding and gating.

It reminded me of how human biology likes to reuse the same molecules in different tissues in different, and sometimes opposing, ways. And made me think that maybe this is not just convenience, but also because that molecule has "strong relationships" with other molecules in the relevant system (meaning high binding affinities or something, idk, I'm over my head), and then it only takes a tweak somewhere else to swap the actual direction of its effect. (?)

So all that to say, I wonder if it makes sense to view CRH's interactions with immune things in mice as clues that CRH might interact with similar parts of the human immune system, but that the exact problem it solves and actual direction of some of its effects in humans could be different or even opposite? This idea would probably be more useful if we knew exactly *how* CRH was mediating the sickness behaviour in mice.

Basically, I guess it seems like the answer to the below question could be "yes, involved, but possibly in a different way." ?

So I guess it comes down to whether CRH is involved in regular human "sickness behavior" or not.

I wonder if we know anything about CRH in primates..

 

[jnmaciuch]

That Familial Episodic Pain Syndrome paper (apparently I only remember the most recent paper I've read) had an interesting example of how the ion channel works differently in different species:


It reminded me of how human biology likes to reuse the same molecules in different tissues in different, and sometimes opposing, ways. And made me think that maybe this is not just convenience, but also because that molecule has "strong relationships" with other molecules in the relevant system (meaning high binding affinities or something, idk, I'm over my head), and then it only takes a tweak somewhere else to swap the actual direction of its effect. (?)

So all that to say, I wonder if it makes sense to view CRH's interactions with immune things in mice as clues that CRH might interact with similar parts of the human immune system, but that the exact problem it solves and actual direction of some of its effects in humans could be different or even opposite? This idea would probably be more useful if we knew exactly *how* CRH was mediating the sickness behaviour in mice.

Basically, I guess it seems like the answer to the below question could be "yes, involved, but possibly in a different way." ?


I wonder if we know anything about CRH in primates..

Could be! Though in the case of CRH, my instinct is that it’s more likely to be evolutionarily conserved. TRPA1 is one of many different chemosensitive proteins, where all the proteins that class have varied structures specifically so they can be attuned to different stimuli. There’s a lot of wiggle room for things to mutate without inducing a fitness cost for the animal.

In contrast, CRH is a pretty central linchpin in a lot of fundamental homeostatic processes. And in the case of its potential connection to sickness behavior, it would be a linchpin for a mechanism that spans multiple parts of the body, neural circuits, etc. Those types of fundamental multisystem processes tend to be conserved across species because small mutations are way more likely to throw a wrench in the whole system. Though it may have just happened to work out differently between human and mouse for reasons I couldn’t account for!

 

[ScoutB]

Thanks for the insights @jnmaciuch, that's all very helpful to know.

Just adding something here I'd like to remember to ponder in the future. And maybe others have thoughts. In the new presentation by the Dutch Autopsy study (aka the the first post in this thread) they mention that they are now looking at orexin/hypocretin neurons and that (so far) those look like they are decreased in ME/CFS patients too.

So, a super preliminary result, but maybe interesting to speculate about? Before, we were thinking orexin/hypocretin was normal in ME/CFS and so maybe ME/CFS could be some form of "narcolepsy minus something". But if orexin/hypocretin neurons do turn out to be affected, then could we imagine a scenario where ME/CFS is "narcolepsy plus something"? I've been reading posts on the narcolepsy subreddit about what it's like to resist a sleep attack and some of them mention feeling very ill.

In any case, if the lack of orexin pans out, the fact that pwME don't usually report feeling sleepy (and in fact report feeling unpleasantly wired+tired at times) seems like a really interesting apparent contradiction.

 

[ScoutB]

(Reviewing other threads where these ideas were discussed and found this)

it seems that IFN-a has been found to actively repress both hypocretin production and cAMP signaling in neurons (some evidence in supplementals). So I'm not sure it provides answers if we're trying to find something that represses CRH + vasopressin through cAMP but not hypocretin.

This might be a point in favour of INF-a then?

 

[jnmaciuch]

So, a super preliminary result, but maybe interesting to speculate about? Before, we were thinking orexin/hypocretin was normal in ME/CFS and so maybe ME/CFS could be some form of "narcolepsy minus something". But if orexin/hypocretin neurons do turn out to be affected, then could we imagine a scenario where ME/CFS is "narcolepsy plus something"? I've been reading posts on the narcolepsy subreddit about what it's like to resist a sleep attack and some of them mention feeling very ill.

That’s really interesting, thank you for sharing! I did not know they also looked at orexin. I’m almost positive I or someone else linked a study on CSF from ME/CFS which found no difference in orexin levels from controls, but I don’t remember which thread it was on. I’ll have to look for the study again to confirm.

If I am remembering it correctly, then I think it’s really vital to find out if all the people in the autopsy study were very severe. We can pretty much assume anyone consenting to travel to a hospital and do a lumbar puncture is on the milder end. So I’m wondering if that might explain the discrepancy. In which case neither orexin nor CRH would probably be the disease mechanism on their own since they would have to be abnormal even in mild ME/CFS to explain disease, but they could be suppressed as a side effect of some other abnormal signaling and only become substantially different from controls when that abnormal signaling is very strong. In which case, yes, neuronal IFN-a is one possibility.

Alternatively, we just have contradictory results on our hands, which would be disappointing.

 

[ScoutB]

Yeah, I also felt like some previous study suggested normal orexin, but couldn't immediately find it. I belatedly remembered that cataplexy is supposed to be caused by a lack of orexin -- I haven't heard of pwME reporting cataplexy, so that would also support the missing orexin being downstream of the disease mechanism, if it is a real effect.

 

[jnmaciuch]

Yeah, I also felt like some previous study suggested normal orexin, but couldn't immediately find it.

Couldn't find anything after a quick search--I may have been thinking of an -omics study done on CSF where orexin, if it was strongly differential, should have been detectable. For all the complaints with the Walitt study I think it's a safe assumption that it would have shown up there if it was a strong hallmark of the disease

I belatedly remembered that cataplexy is supposed to be caused by a lack of orexin -- I haven't heard of pwME reporting cataplexy, so that would also support the missing orexin being downstream of the disease mechanism, if it is a real effect.

Maybe I'm reading too much into it but the presenter in that video said "decrease in orexin neurons" rather than "depletion" or any more dramatic words, which makes me think it's not nearly as dramatic as the CRH depletion. I'm pretty sure decreased (but not substanitally depleted) orexin has been found in other conditions without cataplexy necessarily. All of which is to say that yeah, if these findings are real, a mild downstream effect on orexin seems more likely to me--maybe strong enough to induce some level of sleepiness in the more severe but not strong enough to induce cataplexy

 

[jnmaciuch]

Found another mouse study:

Central neurons encode interleukin-1β signals and mediate stress-induced inflammation

The brain encodes and stores information about peripheral inflammation and can directly recapitulate prior inflammatory responses. However, whether individual cytokines activate specific neural circuits to produce distinct physiological responses remains unknown. To address this fundamental question, we mapped brain-wide responses to IL-1β and found prominent engagement of the bed nucleus of the stria terminalis (BNST). Using targeted recombination in active populations, snRNA sequencing, and circuit tracing, we discovered that corticotropin-releasing hormone-expressing BNST neurons encode IL-1β signals. Chemogenetic reactivation of these neurons precisely recapitulates the physiological signatures of IL-1β exposure with increased circulating IL-6 and corticosterone and tachycardia. These responses require a defined BNST→paraventricular nucleus→rostral ventrolateral medulla→β receptor adrenergic signaling pathway. Critically, restraint stress also activates these BNST IL-1β–encoding neurons to generate the same physiological responses. Our findings establish how a single inflammatory mediator uses a precise neural circuit to activate systemic responses and provide mechanistic insight into the neuroimmune interactions underlying stress-related psychiatric and inflammatory diseases.

Usual concerns apply about applicability of a mouse model to humans (particularly the "restraint stress" piece and the claims of relevance to psychiatric disorders). But I do think it's interesting that multiple studies from different labs have found that a cluster of CRH+ neurons seem to be the hub for receiving cytokine signals from the periphery. At least in mice.

 

[Jonathan Edwards]

Yes, the sensory pathway is intriguing. The claims of impact on disease look pretty shaky.

 

[jnmaciuch]

Wanted to examine the chemogenetic manipulation part of the paper, since the authors claimed they could recapitulate the effects of peripheral IL-1B administration by activating a specific population of neurons in the bed nucleus of the stria terminalis.

From Fig S1, the effects of administering IL-1B:



Fig 1:
These are the results from using the TRAP method, where the neurons that were previously activated by IL-1B are induced to express a synthetic drug receptor so that the neuron can be re-activated by a different drug (CNO). Induction of IL-6 and corticosterone was a fraction (though still an impressive fraction) of what IL-1B could induce directly. I would say that's probably a result of the chemogenetic manipulation technique just not being as robust as it should be, but the two autonomic responses (heart rate and body temp) were recapitulated very well.



Fig 2:
Here they were probing whether specific neuron populations are responsible for the result they get when reactivating all IL-1B responsive neurons. This is just using a Cre-ER model, so any cell that expresses Crh (or one of the other markers they tested) can now be induced to fire with CNO.


They're claiming that CRH is the relevant population here because it was the only one that significantly induced 3 IL-1B related changes (though quite a lot weaker than the original induction with IL-1B). They didn't show plots with body temperature change. I've seen other papers implicating other IL-1B responsive neuron subpopulations in temperature regulation, so my guess is that these markers just don't pick out the neurons that control body temp.

I think the story about CRH being the main driver here becomes a little less convincing when you see that the other neuron markers also induce changes, just not a significant change in all 3 responses they picked to measure. Unfortunately the authors didn't measure anything like total activity that might give us some hints about cytokine-induced lethargy. And with this genetic model we don't know for sure whether it's one cluster of CRH+ neurons or a bunch of CRH+ neurons around the brain causing these results. I should also note that this is a different CRH+ neuron population than the one in the Granton et al. biofilm paper I posted earlier in the thread.

 

[jnmaciuch]

Thread 'Brainstem ADCYAP1+ neurons control multiple aspects of sickness behaviour, 2022, Ilanges et al'

Sep 8, 2022

Abstract

Infections induce a set of pleiotropic responses in animals, including anorexia, adipsia, lethargy and changes in temperature, collectively termed sickness behaviours1. Although these responses have been shown to be adaptive, the underlying neural mechanisms have not been elucidated2,3,4.

Here we use of a set of unbiased methodologies to show that a specific subpopulation of neurons in the brainstem can control the diverse responses to a bacterial endotoxin (lipopolysaccharide (LPS)) that potently induces sickness behaviour...

• Andy
• area postrema nucleus of the solitary tract sickness behavior
• Replies: 13
• Forum: Other health news and research

• 

This paper used similar methods of chemogenetic activation and was able to show differences in activity level when triggering 1L-1B responsive neuron populations in the brainstem. And it doesn't seem to be the case that the brainstem is "upstream" of the CRH neurons and that ties everything together--strong activation of these brainstem neurons did induce strong behavioral changes (Fig. 5e-i), but inhibiting the neurons during LPS administration only somewhat lessened the behavioral changes (Fig. 6a-d). So things must have some level of redundancy, with various CRH neuron populations being one of the options. The question for ME/CFS is whether it's actually biologically possible for any of these circuits to get "stuck" on long enough or strong enough to explain the disease without immune signaling.

 

[SugarSquared]

@jnmaciuch I find your ideas very interesting, but I’ve seen you say on other places that your PhD does not focus on ME/CFS. Do you anticipate being able to do more research into our disease once your PhD is done?

 

[jnmaciuch]

@jnmaciuch I find your ideas very interesting, but I’ve seen you say on other places that your PhD does not focus on ME/CFS. Do you anticipate being able to do more research into our disease once your PhD is done?

That’s the hope

 

[SugarSquared]

That’s the hope

I’ll add my hope to yours! How many years do you think you have left in your PhD if you don’t mind sharing?

 

[Hutan]

Oh sure, it’s just spitballing to find a way to reconcile the mouse studies and the human autopsy study. If CRH upregulation is necessary for people to feel sick during infection, and pwME feel sick in similar ways, you would expect that there’s more CRH activity in ME/CFS.

But we have one study showing the opposite in ME/CFS, a stark depletion. The study presents this as evidence of neuron death but I and @ME/CFS Science Blog were discussing the likelihood that it’s actually just transcriptional silencing of the CRH gene rather than death of neurons.

I was wondering if there are CRH neurons in various places, so that, if some CRH neurons outside of the hypothalamus had a stuck 'on' switch, maybe the hypothalamus CRH neurons would have been put in 'off' mode in order to keep the levels of CRH at reasonable levels.

And so I searched to see where the CRH neurons are, and it turns out they are in various places, including in the brainstem. (wikipedia)

So, what if the CRH neurons in the brainstem were going gang busters, not responding to any feedback loop, because they are getting signals that there is an infection somewhere, probably false signals. That makes us have the sickness response. And the levels of CRH are enough that the CRH neurons in the hypothalamus are getting signals to not produce any.

So, I was wondering that. And then there is the paper linked above about brainstem neurons.

This paper used similar methods of chemogenetic activation and was able to show differences in activity level when triggering 1L-1B responsive neuron populations in the brainstem.

And it looks as though you jnmaciuch are wondering something similar.

 

[Jonathan Edwards]

I am having a hard time getting to read any of these papers in detail at the moment @jnmaciuch. Thanks for sifting through this one.

I am a bit puzzled by what they are trying to prove on the efferent side. They mention an increase in serum IL-6 and corticosterone but seem to imply that these are evidence of driving inflammation?

Corticoids are of course part of the systemic anti-inflammatory response that accompanies local inflammation (maybe as one of the many pathology-focusing aspects of the process). And IL-6 in the blood stream is not inflammatory. IL-6 does contribute to cellular interactions out in tissue but circulating IL-6 is not going to. It may raise CRP and acute phase protein levels but these may also be as much as anything part of the systemic anti-inflammatory aspect.

Sickness behaviour itself is presumably directly due to brain events - sensations and changes in activity mediated by events within neurons.

I may be getting this muddled but I cannot see their data indicating that CRH neurons actually generate inflammatory disease elsewhere.

There are some very peculiar situations where central nervous events do seem to induce local peripheral vascular changes - as in Sudeck's atrophy - but those are not actually inflammatory (in Sudeck's, the part is stony cold, even if swollen).

 

[jnmaciuch]

I’ll add my hope to yours! How many years do you think you have left in your PhD if you don’t mind sharing?

I’m starting my 3rd year, 6 years total is standard in the US. I will be trying to get out one or two small ME/CFS studies during my PhD, though.

 

[jnmaciuch]

I was wondering if there are CRH neurons in various places, so that, if some CRH neurons outside of the hypothalamus had a stuck 'on' switch, maybe the hypothalamus CRH neurons would have been put in 'off' mode in order to keep the levels of CRH at reasonable levels.

It is possible there’s some regulatory mechanism like that. Though I think it becomes a bit tricky to theorize in the brain because CRH is largely acting between synapses from a CRH producing cell to a CRH responsive cell. The amount that leaks out into the parenchymal fluid and circulates around is going to be quite small compared to the concentration at synapses—and if there was a weirdly large amount of CRH in the cerebrospinal fluid, we would have caught it already. There also might be some anatomical considerations with the brain stem that make a theory of neuropeptide “leakage”between it and other parts of the brain less feasible.

What seems more likely here is that evolution just favored using CRH+ neurons in infection-response pathways across different parts of the brain, possibly because of CRH’s other role in corticosteroid release or something else I’m not thinking of. Once certain genomic loci were essentially “tied together” by evolution so you get both CRH machinery and cytokine-signal-response machinery coming online simultaneously in a cell, it makes sense we would keep seeing that same combination across the brain.

 

[jnmaciuch]

I am a bit puzzled by what they are trying to prove on the efferent side. They mention an increase in serum IL-6 and corticosterone but seem to imply that these are evidence of driving inflammation?

They cited other papers showing that efferent neurons feed back into lymphoid tissues and induce local immune cells to produce cytokines via neuropeptide stimulation. So I think all theyre trying to show is that several infection responses known to be mediated by the nervous system sensing peripheral cytokines—autonomic changes, corticosterone release, and that lymphoid tissue-based peripheral cytokine increase—all originate from this specific CRH+“hub” of IL-1B-responsive neurons. Their data seems a bit too messy to prove that it is actually all coming from the same BNST CRH+ neuron population, though.

 

[Jonathan Edwards]

They cited other papers showing that efferent neurons feed back into lymphoid tissues and induce local immune cells to produce cytokines via neuropeptide stimulation.

OK, but that would not be inflammatory. It might facilitate cytokine based interactions in follicles and such like, involving maybe TNF or interferons, but if those cytokines are generated within lymphoid tissue they are not going to induce inflammation at other sites and the cellular interactions in lymphoid tissue are not inflammatory as such. OK, if T cells are activated they may go off and engage in inflammatory lesions but the way it was put in the abstract seemed a bit inappropriate.

And yes, caution on the specificity to these cells seems warranted.