Eccentric medium spiny neuron (eMSN)

[Eddie]

@Eddie, I think you may be at cross purpose with Utsikt. Utsikt is saying that what cells deep in the brain experience is not just based on what is currently coming in through primary sensory channels like peripheral nerves from muscles. It is based on complex inferences derived from signals over time. Even hearing a "G" sound depends on such complex inference.

Some have argued that it seems unlikely that fatigue sensation is based on similar complex inference on primary sensory inputs over hours. To me such an inferential system makes sense since fatigue seems to be our index of how to pace and schedule activity over extended times.

That makes sense. I guess my point was that what ultimately drives the experience must be the state of the deep brain cells when the send the signal to the part that does the experiencing. All the prior steps are useful to know, but those thoughts are also signals being sent by other cells now.

 

[Eddie]

predicting how tired you'll be at the nth hour of a conference / how tiring it would be to carry something across edinburgh at different stages. Forming memories, associations, learning etc is something humans are so so good at in a way that my dog is able to do to a much more limited extent - and Audrey and I insist our dog is very intelligent.

What I am unsure about is why predicting is relevant. I can predict I will be tired if I do certain things, but that prediction doesn't produce the fatigue. There might be brain cells that collect information over time, but they aren't really predicting fatigue, just responding when enough sensory inputs have occurred.

 

[hotblack]

I’d go along with what I think @Eddie and @chillier have said. We don’t need a system like this. Okay, maybe @Jonathan Edwards is right and one exists. But equally the body simply responding to the signals it receives may be plenty without a higher level accounting system.

Its an interesting discussion but I also wonder if the difference matters. I (currently) have a feeling that it’s aberrant inputs or smoothing/management of those that are the problem not the totting up of those inputs. The prolactin response and eMSN/TIDA neuron connection seems to indicate that could be the case. And the way in which so many of us have outsized reposes to different inputs too (physical, mental or emotional activity but also changes in position or noise, light and sound, or infection or even food).

 

[chillier]

What I am unsure about is why predicting is relevant. I can predict I will be tired if I do certain things, but that prediction doesn't produce the fatigue. There might be brain cells that collect information over time, but they aren't really predicting fatigue, just responding when enough sensory inputs have occurred.

I think the model Jo is proposing it *would* generate the fatigue once you start doing the activity again, maybe prophylactically to protect you. Then maybe in ME the predictions are way off and when you start making a cup of tea it erroneously generates massive amounts of fatigue when there is no need. Maybe I've got this wrong do correct me if so.

But this would require a 'fatigue totter-upper over time and future prediction engine' somewhere in the brain which I an unconvinced of the existence of. The type of prediction you're talking about I think is what I think is more likely which is a rationalisation of how you think you'll feel based on past experiences.

 

[Evergreen]

What I am not sure about is the effort preference idea. I don't see how a disruption of the accounting/discounting system is a 'preference'.

Coming back to this.

No-one else really uses the term effort preference. They use effort sensitivity.

Madian defines effort preference as "the inter-individual variability" in effort discounting, and specifies that much of that is normal, with his example of one person going "Sure! I'll do push-ups for 20 bucks." and another person going "No thanks mate. Not worth it." [my paraphrasing of Madian's example]

They're saying that if effort discounting is "disrupted" or "damaged" by a medical condition, then that person's effort preference changes "radically".

I think of what they're proposing is people's effort preference changing such that their brains either weigh the rewards lower (as in Parkinson's. fronto-temporal dementia, lots of psychiatric illnesses) or perhaps in our case, weigh the punishments higher.

Or the energy costs. Am not clear on what exactly the energy costs and punishments refer to, now that I think about it. PEM certainly feels like punishment. There's also immediate punishment, sometimes. But when normal people are weighing up the "energy costs", they're calculating "If I do this, will I be able to do that", which sounds like what our brains are doing too.

Looks like Madian is doing a study looking at the immediate effects of exertion Functional Neuroimaging to Detect the Neural Signatures of the Unpleasantness of Pain and Effort

Some of these conditions may also cause heightened responses to effort; this is an unpleasant sensation felt during physical and mental exertion. [my bolding]

These references are provided:

• Zald DH, Treadway MT. Reward Processing, Neuroeconomics, and Psychopathology. Annu Rev Clin Psychol. 2017 May 8;13:471-495. doi: 10.1146/annurev-clinpsy-032816-044957. Epub 2017 Mar 15.
• Chong TT, Bonnelle V, Manohar S, Veromann KR, Muhammed K, Tofaris GK, Hu M, Husain M. Dopamine enhances willingness to exert effort for reward in Parkinson's disease. Cortex. 2015 Aug;69:40-6. doi: 10.1016/j.cortex.2015.04.003. Epub 2015 Apr 20.
• Peters J, Buchel C. Overlapping and distinct neural systems code for subjective value during intertemporal and risky decision making. J Neurosci. 2009 Dec 16;29(50):15727-34. doi: 10.1523/JNEUROSCI.3489-09.2009.
• Levy DJ, Glimcher PW. The root of all value: a neural common currency for choice. Curr Opin Neurobiol. 2012 Dec;22(6):1027-38. doi: 10.1016/j.conb.2012.06.001. Epub 2012 Jul 3.
• Basten U, Biele G, Heekeren HR, Fiebach CJ. How the brain integrates costs and benefits during decision making. Proc Natl Acad Sci U S A. 2010 Dec 14;107(50):21767-72. doi: 10.1073/pnas.0908104107. Epub 2010 Nov 30.

 

[Jonathan Edwards]

It kind of seems like a specifically human ability to be able to do this

I suspect not. Even a runner bean plant is capable of integrating over time and space to set up optimal patterns f movement, as shown by Paco Calvo's videos. The tip of the stem will circle round in space until it senses a support - possibly even by its shadow rather than touch. It may then move to a diametrically opposite position and 'fling' itself back across to precisely where the support is, repeatedly, until it can catch hold. The only explanation for this I can think of is that it optimises the energy and structure costs of movement in terms of radius of curvature required.

Similarly, many mammals and birds show long term accounting behaviours. Jays and squirrels can hide hundreds of nuts and know which ones are most likely to be ready to eat and where they are at any time over a period of months. A puffin collecting fish for chicks almost certainly benefits from the same sort of accounting of time available, tides, distances, and so on. We tend not to think of these strategic issues for animals but they must be crucial fr survival.

What is different about humans is flexible language. The linguists have suggested that an important part of the language faculty is absence of hard-wired routines like the innate routine of the cuckoo call. I think we underestimate the 'intelligence' of dogs and other pets because they have to operate in an environment that does not suit their hard-wiring very well.

There are clearly some situations where use of activity accounting is much better in humans, like the invention of the bicycle, but I see activity accounting as likely to be very well developed in most vertebrates.

 

[Utsikt]

That makes sense. I guess my point was that what ultimately drives the experience must be the state of the deep brain cells when the send the signal to the part that does the experiencing. All the prior steps are useful to know, but those thoughts are also signals being sent by other cells now.

All cells do experiencing.

But this is also not completely the same as what I’m talking about. I’m saying that the cells that «experience» fatigue might also have changed due to previous input, affecting how future input is «interpreted», i.e. the output from the input.

In other words, brain cells are not static.

 

[Evergreen]

I don't see how a disruption of the accounting/discounting system is a 'preference'.

And coming back to this again, I agree. You explained preference to me years ago in terms of birds' eating habits. If a bird that prefers a particular seed develops a disease, and that disease makes the bird sick if it eats that seed, so the bird minimises its intake of that seed and favours other foods, is that bird really subconsciously preferring to avoid that seed?

For me, the preference for activity is very much still there, and has to be consciously overridden all day every day, like @chillier described above:

Most of the time what is stopping me from exerting myself is not the feeling of fatigue but the knowledge of the consequences and suffering if I do.

 

[Snow Leopard]

Some have argued that it seems unlikely that fatigue sensation is based on similar complex inference on primary sensory inputs over hours. To me such an inferential system makes sense since fatigue seems to be our index of how to pace and schedule activity over extended times.

The problem is that it is largely falsified already -the sense disappears when the signal is blocked, and the brain does not demonstrate an overriding system. The current evidence is the brain senses peripheral fatigue and modulates motor cortex excitability accordingly (and this seems to happen more for larger limbs like the legs than say, our fingers - the brain does recognise different energetic demands based on where the limb is sensed), and sense of effort is upstream from this - so if the overall motor pathway excitability decreases, sense of effort will necessarily increase which is what we see in people suffering fatiguing disorders, including peripheral neuropathies, muscular dystrophies etc - the primary effect in all is "central fatigue", that can be blocked with a rapid response if the sensory afferents are blocked.

 

[chillier]

@Snow Leopard Interesting, thank you - do we have a picture for cognitive fatigue? eg a hard days work sitting writing grants reading papers and doing meetings. Motor pathway excitability reduction is probably not instrumental in producing fatigue here right? Must be excitability changes of many other brain pathways involving interpreting sights and sounds, fetching memories and integration into ideas.

 

[chillier]

If the fatigue in mecfs is coming from a misread of the muscle status in the periphery, the obvious question is: are the neurons responsible the sensory afferents or the neurons that process the information from the sensory neurons in the brain (I might hypothesize it's the brain - given the topic of this thread and also that ME fatigue also comes from cognitive exertion).

We could work that out if we had a way of measuring the firing rates of sensory neurons (maybe checked against EMG results in a muscle usage assay) but does such a technique exist? A while ago there was a thread about using microneurography to measure firing rates in I think long covid in something like autonomic neurons. I have no idea about this limitations of these techniques but I'm curious if it's possible.

 

[Jonathan Edwards]

Hi @Snow Leopard. I couldn't quote your post, which seems corrupted.

But there are all sorts of situations where the brain overrides. I think you must be referring to experiments deliberately designed to study one leg of a very complex system. The brain overrides at 4.00 in the morning, when you have flu, when a driver nods off after beer for lunch, in cataplexy, in boredom and fear etc. All of these are different levels and slants on what is a hugely complex control system.

We have agreed that 'fatigue' is not a very good name for ME/CFS symptoms anyway. If we are focusing on central mechanisms that does not necessarily translate to research on 'central fatigue' as traditionally sconceived.

 

[AliceLily]

The only thing I took at very severe ME/CFS that gave a similar effect to how I felt with one component of fatigue was Prozac. I was given 20mg dose, it felt like I had been given a lead coat to wear and I was already feeling a crushed exhaustion at that particular time of my illness.

I just want to add more to this post about another experience I had with Aropax. I don't know how similar the function of this tablet is to Prozac but with Aropax I had to stop after a week or two from memory because one night I had a terrible experience where suddenly it felt like all the synaptic junctions across my chest caught fire. I felt the burning heat at what felt like points all across my chest. It was so scary I stopped the tablet.

I don't know what could have happened, was there a chemical burn off? I wonder if I had too much of something and the Aropax caused a burn off of something. I thought I should mention in case anyone here understands more of the chemical makeup of tablets and what could have caused the burn off. In case there could be clues? Maybe not, but just posting anyway. This was in my very severe year that this happened.

 

[jnmaciuch]

Not sure if this is the right place to put this, but I wanted to consolidate a couple thoughts that span multiple threads. I was looking into the genes driving eMSN enrichment in Paolo's analysis. The leading edge genes from the top hit were, in order: DNMT3A / PPP2R2A / SYNGAP1 / DLGAP4 / CACNA2D3 / ZC3H13 / CAMK2A / PTPN11 / PDYN / CDR2 / BAIAP2

I should note that all of these hits were a very weak signal in DecodeME. But with the exception of BAIAP2, all were in the core list of Zhang et al. hits (DNMT3A being the #1 hit and several others being top 20)

Considering that Zhang et al. is a rare gene analysis which focused on predicted loss of function genes, I did some more digging on what effects loss of function in those genes might have (focusing on effects that aren't necessarily developmental and can manifest later in life).

I found this paper for DMNT3A:

https://www.cell.com/immunity/fulltext/S1074-7613(22)00295-3

Deleterious somatic mutations in DNA methyltransferase 3 alpha (DNMT3A) and TET mehtylcytosine dioxygenase 2 (TET2) are associated with clonal expansion of hematopoietic cells and higher risk of cardiovascular disease (CVD). Here, we investigated roles of DNMT3A and TET2 in normal human monocyte-derived macrophages (MDM), in MDM isolated from individuals with DNMT3A or TET2 mutations, and in macrophages isolated from human atherosclerotic plaques. We found that loss of function of DNMT3A or TET2 resulted in a type I interferon response due to impaired mitochondrial DNA integrity and activation of cGAS signaling. DNMT3A and TET2 normally maintained mitochondrial DNA integrity by regulating the expression of transcription factor A mitochondria (TFAM) dependent on their interactions with RBPJ and ZNF143 at regulatory regions of the TFAM gene. These findings suggest that targeting the cGAS-type I IFN pathway may have therapeutic value in reducing risk of CVD in patients with DNMT3A or TET2 mutations.

Obviously mtDNA/interferon is the favorite pet theory of your local graduate student, so I assumed my attention was just being caught by what I was already looking for.

But interestingly, a second gene from this list (PPP2R2A) was associated with the exact same thing:

PPP2R2A insufficiency enhances PD-L1 immune checkpoint blockade efficacy in lung cancer through cGAS-STING activation - PMC

PP2A B55α, a regulatory subunit of protein phosphatase 2 (PP2A), is underexpressed in greater than 40% of non–small cell lung cancer (NSCLC) cases due to loss of heterozygosity of PPP2R2A, the gene encoding this protein. Given that low PPP2R2A ...

pmc.ncbi.nlm.nih.gov

PP2A B55α, a regulatory subunit of protein phosphatase 2 (PP2A), is underexpressed in greater than 40% of non–small cell lung cancer (NSCLC) cases due to loss of heterozygosity of PPP2R2A, the gene encoding this protein. Given that low PPP2R2A expression correlates with poor prognosis, treating PPP2R2A-deficient NSCLC represents an unmet medical need. Here, we show that PPP2R2A knockdown or its heterozygosity (PPP2R2A+/–) increases cytosolic DNA, leading to cGAS-STING-type I IFN pathway activation. PPP2R2A deficiency results in elevated expression of immune checkpoint protein PD-L1 via GSK-3β- and STING-dependent mechanisms. PPP2R2A+/– cancer cells have enhanced sensitivity to PD-L1 blockade in a mouse model of lung cancer due to modulation of the tumor immune microenvironment, resulting in increased NK cells and reduced infiltration and function of Tregs. Consequently, PD-L1 antibody treatment increases CD8+ T infiltration and activity, especially in tumors with PPP2R2A heterozygosity. Furthermore, systemic or Treg-specific IFNAR1 blockade reduces the efficacy of PD-L1 blockade in PPP2R2A+/– tumors. Patients with NSCLC with a low PPP2R2A/PD-L1 ratio respond better to immune checkpoint blockade (ICB). These findings underscore the therapeutic potential of ICB in treating PPP2R2A-deficient NSCLC and suggest that PPP2R2A deficiency could serve as a biomarker for guiding ICB-based therapies.

CAM2KA is part of the CAMKII complex, which mediates an important calcium-responsive signaling cascade. I couldn't link loss of function of CAM2KA to mtDNA release, though post-translational modifications on another protein that CAM2KA interacts with (in the CAMKII complex) have been associated with it in a heart disease context:

Maternal hyperglycemia-induced O-GlcNAcylation of CaMKIIδ promotes mtDNA release and cardiac remodeling in offspring - Nature Communications

The study demonstrates that maternal hyperglycemia activates CaMKIIδ via O-GlcNAcylation, leading to mitochondrial DNA release. Targeting this pathway alleviates cardiac remodeling and identifies circulating mtDNA as a potential biomarker of fetal cardiac injury.

www.nature.com

So several genes that are upregulated in eMSNs and picked up in genetic analyses have direct links to the same biological process.

These examples are in heart tissue and immune cells, so the connection to eMSNs wasn't entirely clear. But given that MSNs appear to have different calcium handling than other neurons and mtDNA/RNA release is a calcium-responsive phenomenon, I wanted to see if there was any link to this phenomenon happening in neurons. Turns out MSNs, the cell type most affected in Huntington's, showed distinct mtRNA and DNA release in both mouse models and human samples:

https://www.cell.com/neuron/fulltext/S0896-6273(20)30475-X

The mechanisms by which mutant huntingtin (mHTT) leads to neuronal cell death in Huntington’s disease (HD) are not fully understood. To gain new molecular insights, we used single nuclear RNA sequencing (snRNA-seq) and translating ribosome affinity purification (TRAP) to conduct transcriptomic analyses of caudate/putamen (striatal) cell type-specific gene expression changes in human HD and mouse models of HD. In striatal spiny projection neurons, the most vulnerable cell type in HD, we observe a release of mitochondrial RNA (mtRNA) (a potent mitochondrial-derived innate immunogen) and a concomitant upregulation of innate immune signaling in spiny projection neurons. Further, we observe that the released mtRNAs can directly bind to the innate immune sensor protein kinase R (PKR). We highlight the importance of studying cell type-specific gene expression dysregulation in HD pathogenesis and reveal that the activation of innate immune signaling in the most vulnerable HD neurons provides a novel framework to understand the basis of mHTT toxicity and raises new therapeutic opportunities.

We've talked about huntington's a few times on the forum, and one thing that always struck me as interesting about it was the late in life onset. It makes me wonder whether there's something that happens in this neuron population starting in midlife. Probably an extra reason to make sure age of onset is catalogued in any future rare variant analyses in ME/CFS.

Also, this study did find a population of FOXP2+ neurons that are probably eMSNs, and surprisingly they were the only cell population that actually showed less mtRNA release during active huntington's disease compared to controls (the 2 other MSN populations had the strongest upregulation). It was a small sample size and I'm not sure from the supplemental material if the same finding was corroborated in their mouse model. But if it is real, it makes me curious whether there is something specific about eMSNs that normally protects them from mtRNA/DNA release, and whether the genetic hits in DecodeME might point to that mechanism being bypassed.

I should also note that the Huntington's paper didn't establish mtRNA/DNA release was directly the cause of cell death in that disease. It might just be a contributing factor or something that happens downstream when HTT accumulation is already stressing the cell.

 

[Utsikt]

We've talked about huntington's a few times on the forum, and one thing that always struck me as interesting about it was the late in life onset. It makes me wonder whether there's something that happens in this neuron population starting in midlife. Probably an extra reason to make sure age of onset is catalogued in any future rare variant analyses in ME/CFS.

Typical age of onset of symptoms is often listed as 30-50 or 35-45, but it can be as early as adolescence in some cases.

It would be interesting to know if symptom onset is due to reaching a critical mass of disruption of the normal functioning of neurons from damage, or if it’s due to e.g. epigenetic silencing or other less irreversible factors. There is obviously damage later on, but it might not be the only mechanism at play.