The nanoneedle salt stress test – too good a clue to leave abandoned on the lab bench?

What can the salt stress test tell us about the biology of ME?
The nanoneedle test itself has stalled, and only its inventor seems capable of taking it forward. However, perhaps more valuable than the nanoneedle diagnostic test are the biological changes induced by the salt stress test - and these can still be pursued.

The differences vs healthy controls were very large with total separation between patients are controls. When have we ever seen that before?

At rest, there was no difference between patients and controls. But the salt, stress test generated a big response from the PBMCs of people with ME, and almost no change from those of healthy controls. Such a big difference might well indicate that the changes brought about by the salt stress test, tell us something fundamental about the biology of ME.

In addition, there is evidence that there is a transferable plasma factor ("something in the blood") that can propagate this biological difference from ME cells to healthy ones. (The evidence for this is relatively weak.)

I can't help feeling that understanding what changes the salt stress test induces should be a priority for ME research.

Comments?
I'd welcome any thoughts on this idea, particularly:– does this idea make sense & important?

How could the research be pursued?
E.g., can we interest a cell biologist or an immunologist? (these are immune cells) in at least thinking about the problem?

Thanks

Simon
Some background information is in the next post.

I'm tagging @Jonathan Edwards, and I'd be grateful if others could tag anyone from the nanoneedle thread who might be able to contribute to this. I'm afraid I do not have the energy to go through such a long thread. And I had to dip out of the thread early on when the physics became too hard for me!

 

Background information
The original motivation for using the salt stress test was to as the cellular equivalent of provoking PEM. Something that would expose differences in ME cells that are missed under normal conditions. Or a cellular equivalent proxy of the two-day exercise test (that test is not only more extreme but also only finds far more subtle differences than those seen in the nanoneedle paper).

The PNAS paper (thread) states the authors shifted their focus from ATP originally to finding a diagnostic biomarker:
"We initially hypothesized that stressing PBMCs (in a high-salt environment) would result in extensive consumption and potential depletion of ATP, a high-energy metabolite.

However, ...the promising experimental findings produced by the use of our assay led us to the conclusion that ...[we had] identified constitutes a reliable method of differentiating CFS patients from healthy controls, which is the main focus of this work."

The salt stress test
Increasing the external sodium chloride concentration would be expected to increase the ATP demand on cells through mechanisms, including the sodium and potassium pump. This is because the cell would need to expel sodium to avoid being damaged.

However, increasing the salt concentration outside the cell - causing hyperosmotic stress - has broader effects where the osmotic pressure outside the cell is higher than inside: this draws fluid from the cell into the extracellular space, which could be lethal.

The cells respond with a Regulatory Volume Increase (RVI), a broad range of coordinated activity (including pumping sodium ions out via the Na+/K+ pump) that effectively increases the osmotic pressure inside the cell, preventing loss of fluid. More information at Science Direct (see 4.3).

The PNAS paper states the authors don't know what biological changes underpin the electrical differences they detect (though they speculate ). However, it does say that they used light microscopy and they observed no difference in cell sizes between patients and controls. Separately, they reported they didn't observe excessive cell death in ME cells. And they didn't report any other visual differences.

I wouldn't know where to start, but I'm guessing that a cell biologist would, and maybe some people here would as well (any cell biologists?)

 

I don't understand cell biology well but a problem with the sodium-potassium pump has been proposed by Scheibenbogen and it sounds like the kind of thing that can maybe explain the salt stress test results. Paper here: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9733289/

I have no idea how a transferable factor (other than an antibody) would lead to such a problem. But antibodies don't seem likely. If I remember right Ron has also said that the transferable factor is too large to be an antibody.

 

I remember discussing the findings with colleagues at UCL who are professional immunologists used to looking at immune cells in all sorts of ways (fluorescence sorting, cytokine production, ROS production, adhesion, etc etc). Nobody could make any real sense of these findings and the general view was that the cells were either dead or moribund. Nobody had any idea why they were looking at impedance at the cell surface.

 

My background is in molecular and computational biology, and not in electrical engineering but I did look into the nanoneedle a while back. The assay is super weird and not like something I've seen used in the field before.

I suppose it most closely resembles a coulter counter, where you pass a direct current across a channel and as a cell passes through it the resistance changes and you can read of the resultant change in current to count cells. In this case however they're using an alternating current at 15KHz (15,000 oscillations between positive and negative voltage per second). So now you are measuring in some way the dielectric properties of the cell - in otherwords, sort of how well can it store and release charge as this oscillating voltage is being passed over it.

I'm guessing he tried samples on any interesting technologies he had around and saw what stuck, rather than starting from biological principles and choosing this assay. Not sure it's necessarily measuring impedance at the cell surface but it could be any and all parts of the cell which I suppose have different dielectric properties. I think nobody knows what in terms of cell physiology this would correlate to but things like ion concentration seem like sensible candidates.

Not sure if it has been mentioned before but another possibility is calcium ion changes. Cytoplasmic calcium concentrations can fluctuate a lot as they are sequestered in the ER ready to be released en masse - and these specifically increase the rate of respiration which you would imagine might be necessary if you suddenly find yourself needing to deal with a stressor. Mitochondria and the ER are tightly coupled and have junctions that directly connect them. Could also be that the cells are just dead (I think cell death might also involve ca2+ release not sure, it's a ubiquitous ion). In any case if the cells are dying more readily with ME serum then a cell survival/growth assay would be a simple assay to carry out.

My partner and I looked into if there's any proprietary hardware that could sort of emulate Ron's nanoneedle. I remember we found a device used to count pollen, that acts as a coulter counter but lets you change to an alternating current and adjust the frequency. It would just be the one electrode though as opposed to Ron's thousands of electrodes, and presumably the geometry of the whole thing would be different. I think you're right that the assay itself is not the avenue worth taking but the stressor.

 

My understanding is that the needle has two electrodes very close together, neither of which is placed inside the cell, so I think they must have been measuring impedance at the membrane. Why that should be a good measure of any particular biological process is unclear to me. It was not at all clear me how the cells lay on the electrodes. I don't think there was a fluid gap between the electrodes that allowed cells to sit between.

 

Maybe in experiments where they missed an important factor? It is possible that some process or reagent is used in other experiments and is considered to have no effect on results, but in a completely new experimental technique, it does make a difference. Maybe ME serum (from subjects stressed by travelling to the lab) has some chemical difference that reacts with the nanoelectrodes, changing their properties rather than the cell's. I expect there are many stories about amazing discoveries that turned out to be unexpected equipment issues rather than real effects. The fact that despite the dramatic results, it hasn't been vigorously pursued, gives me the impression that the people involved might not be convinced that it's a real effect worth gambling on.

 

I'm a strong advocate for the nanoneedle. The results are so dramatic we need to keep developing it until it's either a validated test for ME or until we understand what failed. I would say:

• Develop a more efficient nanoneedle machine that can test many samples at once
• Validate it by doing additional studies of people with ME vs healthy controls, people with autoimmune disease, etc.
• Determine reference ranges based on factors that could affect the test: Age, sex, etc.
• Then start using it to probe ME pathology
  
• Get it approved by whatever process is required
• Start using it to diagnose ME

We won't learn anything about ME with a totally unvalidated test, as the results could well be noise. If the test works, we should learn more about it:

• If the patient is in PEM, does it affect the results?
• What chemical or mechanical changes drive the impedance change?
• Are the cells alive, dead, or dying? Does a sudden change in impedance reflect a change in its state?
• What is happening at other frequencies? What's happening to the resistive and reactive components of the impedance?
• How does the blood have to be handled? For all we know, a few hours could make a big difference.

Pertinent questions about ME would be:

• What drugs/chemicals make ME cells look healthy?
  
• What makes healthy people's cells look sick?
• We know that healthy cells in ME plasma act somewhat sick, and ME cells in healthy plasma act healthy. What chemical is doing that?
• Can editing genes make cells appear sick/healthy? Could we make a mouse model based on this?

 

I can't comment on the science side of things but I think it pointless hypothesising about it being a means of diagnosing ME until it has been tested against other conditions. AFAIK this never happened.

 

Right, looking again at the diagram it seems both electrode terminals are located in a single projection that pokes into a channel roughly the width of a cell. There's probably 1 or 2 microns between the anode and cathode. I wish they could have shown a picture of the cells sitting in their device - so we could get a sense of whether the cells are squeezed up against the electrodes or floating around nearby or what.

I'm still not confident you can say it is just measuring the cell membrane as this surely depends on the dielectric properties of the membrane and other nearby components (molecules sitting in the cytoplasm and plasma), the frequency that is used, and how far the charges propagate through the cell. Induced dipoles in atoms polarize orders of magnitude faster than the 15KHz I believe they use, and changes due to movement of ions/ polarised molecules flipping around and so on all have different relaxation times so their contribution will depend on the frequency of the voltage applied. As I say I'm no electrical engineer, I'm happy to be overruled by anyone who knows better.

One source I've found suggests that you start measuring membrane capacitance at frequency ranges of 500Khz to 6Mhz, and cytoplasmic conductance 6-20Mhz (other sources seem to differ). If that's the case then in the nanoneedle the membrane would just be blocking current like in a coulter counter - maybe there is a difference in how tautly the membrane is held over the electrode and resisting current flow that way, due to a subtle difference in cell deformability.

Another thought to do with membranes is a difference in composition of cholesterol and sphingolipids - which form thickened areas of the membrane called lipid rafts which I believe form hubs for cell signaling.

 

Thanks for all the comments, though I'm afraid the discussions of the nanoneedle itself are over my head.

My working assumption is that whatever changes the nanoneedle detected, they are likely to be significant given the total separation between cases and controls. Plus, I think this is the first time that a molecular stressor has been tested.

It's been pointed out that the results might be due to the cells being dead or moribund:

Assuming this doesn't mean cells being accidentally incapacitated or killed through experimental error, wouldn't it still be pretty significant if ME cells were dying as a result of hyposmotic stress?

The researchers increased the plasma salt concentration by 65 mmol/ml to 200 mmol/ml, which they characterised as a middling level of hyperosmotic stress.

I know very little about this area, would you expect a relatively modest increase like this to kill or incapacitate cells?

They also used live microscopy, presumably to identify if any of the cells were shrinking as a result of being unable to meet the challenge of the hyperosmotic stress. They didn't see any differences in size.

Is there some other factor, apart from the hyperosmotic stress, that might account for the very large differences between case and control cells?

 

As I replied earlier, it could be the difference is in how the electrodes react with the samples, with the cells themselves having no differences. The differences could also be with other aspects of the experimental apparatus. Unlikely in this case, but I can imagine dramatic experimental results that are due to the writing on the vials being in different inks for subjects and controls. I think science is filled with such unexpected effects. I vaguely recall polywater being an exciting discovery ... which turned out to be an error due to how the glassware was cleaned.

 

Were the samples from the patients and controls collected and processed in exactly the same way?
If they were collected at different times and places by different people and hence stored differently then that might explain the difference.

Remember the XMRV fiasco when differences were found between patients and controls and it turned out that the healthy controls were a fresh draw being compared with the patients samples that were old and had been in and out of a freezer many times, a freezer that had been contaminated with the mouse virus.

The clear separation between patients and controls looks very suspicious to me.

 

"too-good-a-clue-to-leave-abandoned" or "too good to be true"? If someone is convinced it's too good to be abandoned, and thus a potential career boost, I expect someone will continue that research. It doesn't sound all that expensive (doesn't require a billion dollar synchrotron or 35 Tesla MRI machine), and cellular impedance can probably be measured by other means (reflected signal, signal loss, changed absorption of photons of other wavelengths), all which can be done on bulk samples. Actually, it should be tried by other methods, because exact duplication could duplicate equipment or procedural errors.

 

I know nothing about biology and a bit about electronics. So I think you are saying that that by passing an a.c. signal through some part of the cell, by measuring its impedance (which in this case is primarily a combination of resistance and capacitive reactance), it is then possible to derive its capacitance - which as you say is related to its dielectric properties. Given that reactance is frequency dependant, by measuring impedance at different a.c. frequencies, they could then deduce the capacitive reactance at those frequencies, and therefore the actual capacitance.

A capacitor, of any kind, stores electrical charge when a voltage is applied across it, and the energy thereby stored is a function of the stored charge and the voltage required to hold it there. So once you know a capacitor's value, you then know it's energy storage capability, within its voltage limits. So I think you are pondering if this might also be an indicator of a cell's energy storage capability?

I suppose if there were a significant difference between this and healthy controls, you would get an indication of how the cells of pwME differed in their ability to store electrical energy.

BUT, my lack of biological knowledge means I have no idea if this storage of electrical energy bears any relation to how a cell actually stores the energy it normally needs to in order to function as a cell.

 

The details of mechanisms being described here are way over my head.

To me it is very clear that the nanoneedle has flunked. They published on this potential biomarker 4 years ago. Nowadays, the OMF is funding projects that are looking for biomarkers in ME/CFS and Long-Covid. One doesn't waste extremely useful resources and especially time on this if one already has a biomarker sitting in ones drawer. Of course there's the whole story of Rahim needing grants for tenure etc., but if you truely believe you have a biomarker it is extremely irresponsible to not put all efforts into that. A biomarker would change everything. I think it's fair to assume that things, for whatever reason, didn't quite hold up.

 

That surprises me. It seems that impedance measurements of cells is a standard method for cell analysis. Low-frequency (less than 1 kHz) and high-frequency (a few tens of kHz) impedance typically differ and give different information about the cell characteristics. Am I missing something here? https://www.axionbiosystems.com/technology/impedance-general