A nanoelectronics-blood-based diagnostic biomarker for ME/CFS (2019) Esfandyarpour, Davis et al

[Kalliope]

IFLS has an article as well. They have 25 million Facebook followers.

 

[Snow Leopard]

Sodium potassium pumps are sexually dimorphic and if the controls are not matched then you would have an overrepresentation of females in the CFS population leading to altered impedance.

Is there any evidence for this in human cell culture or PBMCs? The studies showing differences are in animals where there are sex specific differences at the organ level (as a result of different hormone/endocrine signals), but not necessarily the cellular level.

It is still disappointing that the age/gender of the samples were only matched for 5 comparisons and the overall age and gender were not reported. Likewise, the lack of apparent blinding that you pointed out earlier is cause for concern.

I wonder why there is basically no discussion about them trying other cellular stress models other than salt/osmotic stress... The results as they stand just seem incomplete.

 

[FMMM1]

That is unknown, though “just a reaction to being circumstances” seems to be underplaying the problem. Certainly some cases seem to be caused by immune problems, and that’s an active area of investigation. But non biomedical problems can cause very serious depression.

I think we need to be careful about saying that something is non biomedical - after all ME is claimed to be non biomedical. Ron Davis's nano-needle test etc. are forcing some rethinks about ME.

Great posts by the way - thanks.

 

[FMMM1]

This is the key issue, particularly given the frequent claim that this shows ME/CFS is not "all in the mind" (I really dislike that phrase because it denigrates mental health problems).


Depression probably has a stronger immune signal than ME/CFS. Michael VanElzakker has pointed out that you get stronger signal of microglial activation in the brain for depression than for this illness.


So it is very important to show that there isn't a comparable signal from the Nanoneedle set up in illnesses such as depression and anxiety.

The nano-needle might give "a comparable signal --- in illnesses such as depression and anxiety" wow it might assist in those illnesses (or others) - great.

I think the nano-needle might just provide a cascade of diagnostic tests i.e. for ME. Once you have a well defined group of people, with this biomedical problem, then researchers will presumably test this group and identify other diagnostic tests for ME. I'm hoping that this test is a success for ME and for OMF (financially).

 

[Kitty]

Very interesting discussion! I also feel concerned about the level of hype over early-stage research, but I wonder if publishing now is a carefully calculated risk. We urgently need more funding and the involvement of more researchers; if publication in such a high-profile journal adds credibility to OMF's work and flags ME/CFS as an interesting field where a career might be built, then arguably we've made gains whatever the eventual outcome of this project.

 

[Snow Leopard]

Nanoelectronic impedance detection of target cells (Dec 2013) Esfandyarpour et al
https://onlinelibrary.wiley.com/doi/abs/10.1002/bit.25171

On that topic, I wanted to do a research thesis involving microfluidics, back in 2015/2016 and the overlap between Dr Davis' ME research and microfluidics was intriguing to me, including that paper. I majored in Chemistry, but microfluidics was really the only area I was really interested in towards the end of my degree and microfluidics will play a major future role in terms of both biochemical research and diagnostics.

I was interested in assays like this however:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3926080/

I realised (sadly), that I did not have the ability to do such research given my symptoms have slowly been getting worse and completing my undergraduate degree was hard enough as it was...

 

[Jonathan Edwards]

But also, I had assumed that some of the cells would wander off but that with 4000 Micro chambers and five samplings a second the device would capture an average-occupancy-based figure.

Where does it say 4000 micro chambers? That would seem to need 200ml of cell suspension, which is a crazy amount. I am afraid I am still lost with this.

 

[FMMM1]

In the discussion, the authors say


"According to our experimental results, ME/CFS blood cells display a unique characteristic in the impedance pattern…"


Yes, as they go on to say, it is significantly different from healthy controls. But no one has a problem distinguishing ME patients from healthy controls, it is the comparison with other diseases that is so important.


What they could've said is that they have achieved clear blue water between healthy controls and patients, with the pilot results achieving 100% specificity and 100% sensitivity. This is a remarkable result and does give hope of finding differences with other illnesses.


So the most important step now is to demonstrate its ability to discriminate from other illnesses — at least it is the priority if this is ever to become a biomarker. There I am hardly the first person to say that.

I wonder if this technology will question current understanding of a range of diseases. I think Ron mentioned that there's fatigue in MS but of course the doctors say that's different. I'm not just thinking Lyme, or Fibro, here. Presumably there are a number of diseases which are mediated(?) by exosome signalling; as ME appears to be.

Anyone figure out why the don't run the test with filtered/un-filtered plasma/serum (with/without exosomes)? Presumably it's costs and the fact that it would then be a laboratory test, rather than a GP Surgery test.

 

[Snow Leopard]

Where does it say 4000 micro chambers? That would seem to need 200ml of cell suspension, which is a crazy amount. I am afraid I am still lost with this.

I don't know what Simon means by "micro chamber", but the schematics show multiple nanoneedle sensor tips perpendicular to the microfludic flow. Like most binding processes, cells forming an interface near the coating on the electrode is a dynamic process and the measured impedance reflects the equilibrium over the sampling period, rather than individual binding events which would be quite noisy...

These have already been posted, and there are other manuscripts which explain how nanoneedle impedance sensors work, so I would recommend having a read...

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3751968/
https://onlinelibrary.wiley.com/doi/abs/10.1002/bit.25171

 

[Barry]

Given that the impedance will vary with frequency I wonder if Ron Davis' team have considered looking at the impedance characteristics in an additional way. They have homed in on doing impedance-against-time graphs at 15kHz, but will doubtless have experimented at other frequencies also. Feels like it might be interesting to see what happens for a wider spectrum of frequencies - effectively as a 3rd axis off of the graph. You could then, on the one hand, choose various slices, each at a given frequency, to see how the various curves change - the current characteristics are effectively a single slice at 15kHz. On the other hand you could also look at the characteristics as impedance-against-frequency, any given slice being at a given time. I can't help wondering if this way of looking at the impedances might also exhibit differences for pwME.

 

[Jonathan Edwards]

This link has already been posted, and there are other manuscripts which explain how nanoneedle impedance sensors work, so I would recommend having a read...

The more I read these papers the more I get confused. I am now confused about 'microfluidic flow'. Where does the new paper talk about flow? The link you give is for solutions it seems.

When cells do stick to things they tend to do so in an active way a bit like slugs crawling up tiles. The membrane changes with contact.

I remain lost. I wish they had given us a clear diagram of the geometry.

 

[Snow Leopard]

Given that the impedance will vary with frequency I wonder if Ron Davis' team have considered looking at the impedance characteristics in an additional way.

Yes, it is a little disappointing that they didn't provide impedance vs frequency curves and the only impedance graph over time they provided were of the purely resistive component (Zre). The magnitude of change was much greater for the resistive component (Zre) than the reactive component (Zim):

However, the increase in impedance was followed by a marked excursion above the initial baseline value by 74.92% ± 0.69, 301.67% ± 3.55, and 64.73% ± 0.62, for |Z|, Zre, and Zim, respectively, figures that are significantly greater than the values observed for the healthy control.

Which makes you wonder whether there wasn't simply some factor with lower conductivity accumulating in the interfacial layer over the electrodes over time.

 

[Snow Leopard]

The more I read these papers the more I get confused. I am now confused about 'microfluidic flow'. Where does the new paper talk about flow? The link you give is for solutions it seems.

That was a mistake in my part, I was meant to say microfluidic channel. The flow is just during filling...

When cells do stick to things they tend to do so in an active way a bit like slugs crawling up tiles. The membrane changes with contact.

I remain lost. I wish they had given us a clear diagram of the geometry.

I agree, a clearer diagram of the geometry would be much more helpful.

The manuscript I linked however stated:

Various thicknesses and geometrical designs have been fabricated and tested. The sensor design used in this study consists of electrodes 100 nm thick and a middle oxide layer 30 nm thick. The top protective oxide layer thickness is 20 nm and the bottom oxide layer thickness is 250 nm. The width of the nanoneedle tip is 5 μm.

The other manuscript stated:

Various thicknesses and geometrical designs of nanoneedle biosensors have been fabricated and tested. The thickness of top oxide layer, top electrode, middle oxide layer, bottom electrode and the bottom oxide layer are 20, 100, 30, 100, and 250 nm, respectively, and the sensor width is 3 mm (Esfandyarpour et al., 2012, 2013a). The active sensing region of the device is the middle oxide layer.

Note the latter (bolded). I wish the width of the nanoneedle was stated in the current study...

 

[Barry]

Where does it say 4000 micro chambers? That would seem to need 200ml of cell suspension, which is a crazy amount. I am afraid I am still lost with this.

"For each experiment, 50 μL of the prepared sample (SI Appendix) was injected into the microfluidic wells" ... I interpreted that to mean a total of 50μL into all the wells, not into each individual one.

 

[Barry]

I wish they had given us a clear diagram of the geometry.

Yes.

 

[wigglethemouse]

They have homed in on doing impedance-against-time graphs at 15kHz, but will doubtless have experimented at other frequencies also.

Some of their earlier papers provides analysis at different frequencies. In this paper they comment on the component of real and imaginery components of the impedance and their effects and that the real or in-phase component has the biggest effect. This means that lower frequencies are probably good.

Similarly, in-phase impedance showed the greatest separability (P = 7.27E-9), while Zim and jZj signals were also significantly separable (P =5.06E-5 and P = 2.67E-5, respectively)

I looked at impedance spectroscopy papers and commercial equipment and 15kHz seems to be in the range others use for experiments.

ETA from 2013 paper:

The optimum region for sensor operation occurs between 1–100 kHz given that the largest difference in response between water and cell solution occurs. As mentioned, in a low salt concentration electrolyte the double layer is 10-nm thick, thus the impedance resulting from the double layer capacitance (Cdl) can be calculated to be 1Gohm at 15 kHz.

By measuring the impedance at 1Hz and calculating Cdl at this frequency, the faradaic impedance (due to tunneling of electrons from the electrodes to the electrolyte is determined. The total impedance at 1Hz is 0.5Gohms, meaning that Rf in series with Rb (both frequency independent) can be no less than 0.5Gohms
across the whole spectrum. This means that the equivalent impedance of Cdl and Rf in parallel with each other and in series with Rb has to also be greater than 0.5GV. Comparing this to the total impedance of the sensor at 15 kHz which is 3.6Mohms, we are able to assume that the loop containing Cdl, Rf and Rb is essentially an open circuit allowing us to simplify our model significantly as shown in Figure 2B.

 

[Snow Leopard]

"For each experiment, 50 μL of the prepared sample (SI Appendix) was injected into the microfluidic wells" ... I interpreted that to mean a total of 50μL into all the wells, not into each individual one.

Which based on the reported concentration of 200 cells per microlitre means 10,000 cells...

 

[Barry]

Yes, it is a little disappointing that they didn't provide impedance vs frequency curves and the only impedance graph over time they provided were of the purely resistive component (Zre). The magnitude of change was much greater for the resistive component (Zre) than the reactive component (Zim).

Yes, I missed that. But the imaginary component was still non-trivial - else there would have been no point using an oscillatory signal, and could have just measured resistance.

Which makes you wonder whether there wasn't simply some factor with lower conductivity accumulating in the interfacial layer over the electrodes over time.

But in which case the healthy controls would have been subject to the same measurement artefacts. The fact there is a big difference suggests something else is going on. Even if it were due to what you suggest, something must be going on to give rise to that difference.

 

[Barry]

Some of their earlier papers provides analysis at different frequencies. In this paper they comment on the component of real and imaginery components of the impedance and their effects and that the real or in-phase component has the biggest effect. This means that lower frequencies are probably good.

I looked at impedance spectroscopy papers and commercial equipment and 15kHz seems to be in the range others use for experiments.

ETA from 2013 paper:

I guess the point is that by looking at the impedance, they are gathering signatures for both the resistive and capacitive (I'm guessing inductance is not an issue here) characteristics, which are presumable (inevitably?) due to different facets of the biology, so better chance of not missing something. Although the resistive component (i.e. real, in-phase) is dominant, the capacitive component (i.e. reactive, imaginary, out of phase) is non-trivial, so can still contribute to the picture.

 

[Adrian]

Isn’t the hint from both the PNAS paper and Fluge and Mella’s paper that there doesn’t seem to be any difference between ME/CFS cells and healthy ones? My limited understanding is that these studies hint that the difference seems to be in the plasma or serum, and that cells in ME/CFS plasma or serum act differently to cells put in healthy plasma or serum when they are made to work harder.

Apologies if I’ve misunderstood your point.

Yes I think I just made my point badly. I was just trying to say that the experiment and what it suggests is interesting rather than whether it is a biomarker.