Talk:Comparison of respirometric methods

From Bioblast

Tony Hickey: Calibration is my chief concern

Anthony Hickey
At the University of Auckland we recently had a demonstration of the Seahorse XFe96.
This system, as most are aware of, has putative uses for high throughput. I had mounting concerns as the demonstration progressed. It may be high throughput, but what does it all mean. I had concerns.

No Zero and calibration media

The system is calibrated to 100% air using a special proprietary buffer. How this relates to different media is an issue as different species mitochondria, or cells have evolved to exist in fluids of different osmolarities, amphibians (e.g. Xenopus) are extremely low, freshwater aquatic fish (including zebrafish also a biomedical model) are lower than mammals, marine teleosts are higher than mammals, marine invertebrates very high (~1000 mosmol!). This affects oxygen solubility. Moreover different media have different O2 saturation profiles (as a function of temperature).
Negative oxygen (red arrow) - a Seahorse presentation. See Rogers (2011).
A critical issue is a lack of zeroing of the plate. There is no Zero calibration. This is not good practice. This is an obvious problem for the system as calibration to two points is essential (assuming if the system is linear). It likely accounts for the erroneous data, seen below in trace B, from a presentation on quality control. Note it illustrates negative oxygen pressures. How oxygen pressure/concentration is determined is a mystery, and given this is not the Hadron Collider (although $/kg it is similar) I doubt we are observing antimatter. On questioning how the zero was determined I was informed it was done so by โ€œa very complicated algorithmโ€ (a recurrent theme).
There are perhaps several reasons why there is no zero. Those I immediately thought of were;
  • You cannot zero prior to the experiment as it will like have to be done chemically (i.e. Dithionite), which will contaminate the experiment.
  • How about at the end of the assay? This will take up one of the limited four titration choices.
  • How about leaving the cells/mitochondria to draw down all the O2.
  • But it may also point to another issue. Back flux presents an issue. Perhaps a zero is impossible to achieve? The chambers do not seal and they are made of plastic, which likely holds considerable oxygen.

Back flux

This was a primary question I raised, how was this accounted for?
I noticed that the data used to derive O2 flux at specific checkpoints were non-linear (however those for ECAR, or delta pH were linear). So there was something specific to O2 flux. Moreover, low fluxes, such is in inhibited states (oligomycin or rotenone) showed less linear plots, they curved more.
I posed the question how is the flux derived if it is non-linear? The response was โ€œthere are complex algorithmsโ€. A semi log plot may have been a better response. But what was causing the curved response in flux? My conclusion is back flux. This will have a greater influence at low O2 fluxes than high fluxes. This means back flux has disproportionate effects on high and low flux states. So what do RCRs etc. in this system mean?

Large volumes

Large volumes of injection - 10 ul? Most oxygraphy systems attempt to titrate the smallest volumes possible. The reaction cell of the Seahorse XFE96 is ~200 ul, and smaller still (~7ul) for the actual assay volume in the โ€œclosedโ€ state. We aim to add 1-10 ul of anything into to a 2 ml chamber of the OROBOROS Oxygraph-2k. This contributes 0.1-1% of the volume. 10 ul into 200ul is 5%.


I had other concerns re temperature control, atmospheric pressure changes. But fundamentally, as the very expensive XFe system is supposed to measure oxygen, calibration is my chief concern.
Dr Anthony Hickey
Applied Surgery and Metabolism
School of Biological Sciences
University of Auckland, New Zealand
[email protected]
MiPNet Reference Lab: NZ_Auckland_Hickey_AJ

Erich Gnaiger: Bioenergetics made simple?

Misleading simple interpretations: Coupling control in intact cells can be simply studied in a sequence of ROUTINE respiration (R), oligomycin-induced LEAK respiration (L), and evaluation of Electron transfer system capacity (E) (Hรผtter et al 2004). The R/E flux control ratio provides information on ROUTINE respiratory activity relative to ET capacity (Gnaiger 2008, Gnaiger 2009). In commercially oriented (XFe) discussions, David Nicholls interprets the equivalent of the R/E flux control ratio as an index of reserve capacity. This is not supported by modern concepts on coupling control and is refuted by experimental evidence in human cell models and many tissue preparations (Gnaiger 2012 MitoPathways). This unjustified simplification (Bioenergetics made simple) ignores along the lines of outdated terminology (State 3 and State 3u) the fact that OXPHOS capacity (P) cannot in general be determined by noncoupled states of respiration (E). But E needs to be corrected for the P/E ratio if any apparent reserve capacity of ROUTINE respiration should be determined. This may still be considered as a very simple concept, but apparently not suffiently simple for XFe advertising. - Gnaiger Erich

MIG-List: Oxygraph-2k for HRR versus XF multi-well method

Von: Mitochondria Interest Group [1] Im Auftrag von Erich Gnaiger. Gesendet: 24 September 2011 23:47
An: [email protected]
Betreff: [MITOCHONDRIA-L] Oxygraph-2k for HRR versus XF multi-well method
Content of this Email is solely the Responsibility of the Author. Please see Disclaimer at the bottom.--------
Dear all, dear Pablo:
In a discussion on high-resolution respirometry in relation to the publication by Rogers et al 2011, I raised the question: "High throughput without high output?" with the following topics:
1.1 Instrumental specifications of the XF multi-well system are missing
1.2 Uncoupled flux does not reflect electron transfer system capacity
1.3 Lack of validation with high-resolution respirometry
1.4 Lack of quality control of isolated mitochondria - lessons in


I would be glad to receive your scientific input to this discussion, which has been triggered by two messages circulated to the MIG list under the heading "functional mitochondria isolation from primary cortical cell cultures".
(1) Dr. Anne Murphy attached a pdf-reprint that does not deal with mt-isolation procedures from cell cultures but describes the 'validation' of the XF multi-well method for isolated mitochondria.
(2) The comment by Prof. David Nicholls: "Look to see whether a colleague has a Seahorse" (good science practice would require a disclaimer statement to be added, in line with good practice on the MIG list).
The original question raised by Dr. Pablo R. Castello ("functional mitochondria isolation from primary cortical cell cultures") stimulated a useful discussion, even if the subject changed from 'mt-isolation procedure' to the platform of measurement. I would like to learn more from experiences comparing the Oxygraph-2k for HRR and the XF multi-well method.
Following the suggestion by Prof. Lech Wojtczak, I would also recommend to use a validated method for permeabilization of cells with fully functional mitochondria, and application of advanced diagnostic protocols with multiple substrate-uncoupler-inhibitor titrations (Pesta and Gnaiger 2012).
Best wishes
DISCLAIMER / CONFLICT OF INTEREST: Dr Erich Gnaiger (Medical University of Innsbruck, Austria) is Chairman of the Mitochondrial Physiology Society He is also founder and managing director of OROBOROS INSTRUMENTS

-----Ursprรผngliche Nachricht-----
Von: Mitochondria Interest Group [2] Im Auftrag von Murphy, Anne

Gesendet: 19 September 2011 18:53

An: [email protected]
Betreff: Re: [MITOCHONDRIA-L] functional mitochondria isolation from primary cortical cell cultures
Content of this Email is solely the Responsibility of the Author. Please see Disclaimer at the bottom.--------
Since you have a Seahorse, you may be interested in making the small yield of mitochondria from the cultured neurons go a lot further experimentally.
We recently published a paper on methodology for using isolated mitochondria in the Seahorse. We did the validation work with the XF24, but it works in the XF96, as well. Attached is the pdf.
Best of luck,

-----Ursprรผngliche Nachricht-----
On Sep 18, 2011, at 2:40 PM, Pablo R Castello wrote:
Content of this Email is solely the Responsibility of the Author. Please see Disclaimer at the bottom.--------
Dear David, Anne, Gloria, Tibor and Charles, Thank you so much for you advice. David, I already ran almost 80 plates on both a XF 24 and a XF 96 only this year using primary cortex cell cutures. I found a very interesting model of mitochondria damage that is allowing us to test potential neuroprotective small molecules. The idea is to validate my findings using a different analytical method. We have a polarographic instrument, and I thought that It would be a good idea to isolate mitochondria from our stressed cultures and analyze them with this method. From the advice of you all, It seems difficult (because the high number of cells) but not impossible. What do you think? I strongly appreciate this discussion.
Best Regards,
Medical advice and information that could identify an individual is discouraged on this list. List members are reminded of their responsibility to critically evaluate the content of the postings. The information, opinions, data, and statements contained herein are not necessarily those of the U. S. Government, the National Institutes of Health (NIH), or the Mitochondria Interest Group (MIG) and should not be interpreted, acted on or represented as such. Please see the MIG website:

Dan Kane: Complimentary

  • ... regarding the Seahorse ... they were adament that their instrument was useful for quantitative measurements, even though the vast majority of the data was expressed as fold-change, or % control. In the end, they said that their system was more complimentary to existing oxygraphs, and even suggested that the Seahorse be used for optimization experiments to determine the best conditions for O2k experiments. - Daniel A. Kane, East Carolina University, Greenville, NC, USA (2009-07-30)
Cookies help us deliver our services. By using our services, you agree to our use of cookies.