That's my feeling too: that my ME is producing a barrier to doing activities. My body is capable of the activity, but this ME effect is similar to being hungry, overheated, wearing uncomfortable shoes, being swarmed by mosquitoes, etc. It shifts activities from "I want to get this done" to "I don't feel up to doing this right now". That's why I consider this neurological rather than physical.
Preprint Indistinguishable mitochondrial phenotypes after exposure of healthy myoblasts to myalgic encephalomyelitis or control serum, 2025, Ryback et al
- Thread starter chillier
- Start date
Why? Can't it be an inappropriate response to normal signals? Even small amounts of some food components will trigger serious worsening of my ME symptoms, but I doubt that my gut is sending seriously inappropriate signals; nothing that a lab test would flag as seriously abnormal.
Then there's cognitively-triggered PEM. Are the brain cells involved in that activity sending out seriously abnormal signals, or is it normal signals that a few other cells are hyperreacting to?
It also needn't be a single serious abnormality: in a feedback loop there could be multiple deviations that all add up in such a way as to cause a serious change in function. That's another possibility that won't readily show up on lab tests. For that situation, I think it would be necessary to have a theory first, to know which signals in which cells to monitor to the necessary degree.
Jonathan Edwards
Senior Member (Voting Rights)
Which is why Jo, Jackie and I have tried to construct such a theory - to explain how things can go so wrong without anything showing up on lab tests.
I think you could argue that ultimately nothing is wrong except the brain mis-responding to anything and everything but nobody likes that idea much and I personally don't think it is very plausible, based on the various considerations we go through in the Qeios paper.
If T cells are playing around with all the input signals without generating any inflammation I think one could voter all stages including very severe. The test would be to track down the misbehaving T cells in ways that would catch them doing that sort of thing - maybe short term co-cultures.
wigglethemouse
Senior Member (Voting Rights)
I was watching the recent Action for ME webinar where Audrey presented this work and the following slide was shown at (link to 11:33) and it shows individual variations between people's samples that were part of the replicates for each person.
@DMissa is it expected to have such variation between replicates for many of the individual sample sets on the Seahorse analyser in your work? When I see variations in data like this I often wonder whether the experiment is controlled to give reliable repeatable data. If not, either the Seahorse, or the experimental method could be inappropriate to show differences due to the variation that is embedded in the experiment. I guess what I'm asking is, are we looking at data that is just noise due to the experimental procedure for the machine used?
Note: The data is scaled for cell count.
Does anyone know what muscle cells alone would show in a test like this?
EDIT : This is not a knock on Audrey and team. I think it was superb to include so many replicate tests for each sample. I wish this was done more in Science.
@DMissa is it expected to have such variation between replicates for many of the individual sample sets on the Seahorse analyser in your work? When I see variations in data like this I often wonder whether the experiment is controlled to give reliable repeatable data. If not, either the Seahorse, or the experimental method could be inappropriate to show differences due to the variation that is embedded in the experiment. I guess what I'm asking is, are we looking at data that is just noise due to the experimental procedure for the machine used?
Note: The data is scaled for cell count.
Does anyone know what muscle cells alone would show in a test like this?
EDIT : This is not a knock on Audrey and team. I think it was superb to include so many replicate tests for each sample. I wish this was done more in Science.
Last edited:
chillier
Senior Member (Voting Rights)
The seahorse assay introduces a lot of technical variation in general, but there will be even more in our particular set up because the cells (which are seeded in their wells near confluence) have been incubating in serum for a whole week prior to running the actual assay.
We are not looking at data that is just noise. The key number is that we have a repeatability of 56%. That means 56% of all the variance in our data comes from the biological differences between the individuals, not from noise, plate differences, some other source of technical variation etc. It means we have consistency of the results for a given individual over all 15 replicates we did for that individual.
Thanks for the question and the answer.
That really underlines how important it is to use the Seahorse machine with all the checks and balances that this team put in place. Relying on a couple of replicates per individual in a study with a handful of people in each cohort will not be good enough.
That really underlines how important it is to use the Seahorse machine with all the checks and balances that this team put in place. Relying on a couple of replicates per individual in a study with a handful of people in each cohort will not be good enough.
DMissa
Senior Member (Voting Rights)
Yep, it is very variable. The important thing you can do is perform the assay with a great deal of care and consistency, many replicates within the same experiments (we aim for at least 5 or 6, this is standard) and do at least 3 runs of each sample (we'll often do many more).
The team here seems to have done all of this so I would not be concerned
it also depends on things such as cell type, organism, specifics of how things are being run (time courses, adherence matrix - if any, media formulation, the effects of any treatments, the consistency of whichever drugs are being injected, the instrument model, how you are normalising if at all, injection volume vs handling time to avoid evaporation... particularly on the 96 models which use smaller volumes... which means potentially the need for multi-person parallel workflows.... etc)..
But yes, generally, seahorse is one of those assays where you need to be very rigorous and use many replicates
There are a lot of things that can go wrong or introduce variation. It is something that experience and attention to every detail really benefits. One has to be very fastidious to produce consistency
wigglethemouse
Senior Member (Voting Rights)
Thank you both @chillier and @DMissa for the explanation. I've learned that using the Seahorse is not plug and play and needs proper knowledge and experimental design. I didn't realise different Seahorse models had different well sizes that needed to be taken into account. I love when I can learn something new even with severe ME. Thank you again.
DMissa
Senior Member (Voting Rights)
I have a methodological comment relating to the original 2016 seahorse design (the current study is a replication effort so it is not a knock against the current study)
in the seahorse done originally in the 2016 paper, the injection were glucose -> oligomycin -> fccp ->rotenone+antimycin A. I think this may impose some limitations upon particular, very specific pieces information that might otherwise be extracted from a similar assay. I will explain why. (whether these particular things matter or not is another question - it depends on whether the bits of information that are affected are relevant to a disease process)
Combining Rotenone (Complex 1 inhibitor) and antimycin A in one injection means that you cannot pull out the complex 1 contributions directly. The advantage is that it frees up an injection port for glucose, and this is useful if you want to incorporate some consideration of glycolysis into the assay, for sure (or to do something else with the free port). The issue is that pyruvate is already provided in the medium according to the methods, and so you probably won't expect to see a full effect when injecting glucose. Which from what I can see at a glance is seen in both papers, looks like a small effect from injecting glucose (this makes sense).
If glycolysis is intended to be examined, it would be good to see glycolysis-tailored assays without pyruvate in the medium and with glycolytic inhibitors injected, and in separate mitochondrially-tailored assays it would be useful not to overlook Complex 1 if there is a mitochondrial hypothesis in play.
This is particularly important because much of the conversation is around maximum respiratory capacity and this is primarily driven by complex 1
If one were to entertain ideas about "oxidative stress" (I have no strong opinions either way) Complex 1 would also be a major suspect
These are just examples of where complex 1 might be relevant to people's ideas.
As to whether any of this is relevant to an underlying disease process is another matter - this is just from a methodology perspective because I see these things overlooked more than not (not just in ME/CFS research, everywhere, even by dedicated mitochondrial labs).
There is an obvious economical/practical consideration here because this approach (dedicated glycolysis assays and dedicated mitochondrial assays) would double the workload, but if these particular aspects are pursued in detail in other work or considered relevant to whichever ideas were held in mind, it would be good to see them dissected more directly
Again, these things might not matter, they may not have been practical to look at in more detail, but this is potentially useful method context for people who are interested in this stuff but without close knowledge of the technique
in the seahorse done originally in the 2016 paper, the injection were glucose -> oligomycin -> fccp ->rotenone+antimycin A. I think this may impose some limitations upon particular, very specific pieces information that might otherwise be extracted from a similar assay. I will explain why. (whether these particular things matter or not is another question - it depends on whether the bits of information that are affected are relevant to a disease process)
Combining Rotenone (Complex 1 inhibitor) and antimycin A in one injection means that you cannot pull out the complex 1 contributions directly. The advantage is that it frees up an injection port for glucose, and this is useful if you want to incorporate some consideration of glycolysis into the assay, for sure (or to do something else with the free port). The issue is that pyruvate is already provided in the medium according to the methods, and so you probably won't expect to see a full effect when injecting glucose. Which from what I can see at a glance is seen in both papers, looks like a small effect from injecting glucose (this makes sense).
If glycolysis is intended to be examined, it would be good to see glycolysis-tailored assays without pyruvate in the medium and with glycolytic inhibitors injected, and in separate mitochondrially-tailored assays it would be useful not to overlook Complex 1 if there is a mitochondrial hypothesis in play.
This is particularly important because much of the conversation is around maximum respiratory capacity and this is primarily driven by complex 1
If one were to entertain ideas about "oxidative stress" (I have no strong opinions either way) Complex 1 would also be a major suspect
These are just examples of where complex 1 might be relevant to people's ideas.
As to whether any of this is relevant to an underlying disease process is another matter - this is just from a methodology perspective because I see these things overlooked more than not (not just in ME/CFS research, everywhere, even by dedicated mitochondrial labs).
There is an obvious economical/practical consideration here because this approach (dedicated glycolysis assays and dedicated mitochondrial assays) would double the workload, but if these particular aspects are pursued in detail in other work or considered relevant to whichever ideas were held in mind, it would be good to see them dissected more directly
Again, these things might not matter, they may not have been practical to look at in more detail, but this is potentially useful method context for people who are interested in this stuff but without close knowledge of the technique
Last edited:
hotblack
Senior Member (Voting Rights)
Thanks for all the great insight @DMissa and @chillier we’re lucky to have you all and this knowledge on this forum.
And great question to prompt it @wigglethemouse
Maybe one day some of us will recover enough to if not make great scientists then great lab technicians.
And great question to prompt it @wigglethemouse
Maybe one day some of us will recover enough to if not make great scientists then great lab technicians.