Home arrow Info arrow Using fluorescent probes to detect ROS in cells



Remember me
Password Reminder

| Home | How-to | Info | Tools |

Using fluorescent probes to detect ROS in cells Print E-mail

Using fluorescent probes to detect ROS in cells.

    If you are thinking about using H2DCF, cmDCF, MitoSox, or DHE in your study to detect ROS in cells, perhaps this discussion can be enlightening and useful to you. These excerpts are from MITOCHONDRIA email list, lightly edited by Dr. Starkov for clarity. Enjoy!

Question: On 12/9/2011 9:33 AM, a question (Q) was posted to MITOCHONDRIA mail list:

“Hello, We are looking for a good probe to measure ROS in vitro in a microplate reader.”

Answers (A):

    A1: The journal Free Radic. Biol. Med. recently published a "position paper" on this topic which I would recommend highly to anyone getting into this topic... http://www.ncbi.nlm.nih.gov/pubmed/22027063 I would advise strongly against the use of any fluorescent probe for ROS in a plate reader setting. The only thing you will be able to say is that the cells got "more oxidized/reduced". You cannot make any claims as to the chemical identity of the ROS. Some specific problems are as follows:
1) DCF is useless, it is a very good reporter of free iron levels in the cell, not much else.
2) DHE is no better - its chief product upon specific reaction with superoxide is not ethidium, but hydroxyethidium. These are not distinguishable with poor optics such as in a plate reader - you either need a very good fluorimeter with dial monochromators, or separate the products by HPLC. MitoSOX is just DHE with a mitochondria-targeting Note#1 tag on it.
In very controlled experimental systems, such as isolated mitochondria or purified enzymes, where you can use narrow-bandwidth optics for detection, maybe you can use the above. BUT, if you just want to throw these dyes on cells and use filter-cube based detectors, don't waste your money.

Paul S. Brookes
Associate Professor,
Dept. Anesthesiology,
Univ. Rochester.
Med. Ctr.601 Elmwood Avenue,
Rochester, NY 14642, USA.

     A2:   I disagree with Dr. Brookes. We have used DCF (dichlorofluorescein) and its chloromethyl version (cmDCF) successfully in something like 20 papers. Like any other probe, DCF and cmDCF must be used correctly:
a) H2DCF (dihydrodiclorofluorescein) and cmH2DCF do not react with or measure H2O2
b) DCF and cmDCF fluorescence does not measure iron. Calcein is a good indicator of chelatable iron, specifically reduced, ferrous iron (Fe2+). When hepatocytes are exposed to bafilomycin to block the vacuolar proton-pumping ATPase and alkalinize lysosomes, lysosomal Fe2+ is released into the cytosol and then goes into mitochondria via the electrogenic mitochondrial calcium uniporter (MCU). However, this iron release does not cause an increase of cmDCF fluorescence above baseline. Similarly, exposure to a moderate concentration of tert-butyl hydroperoxide (t-BOOH) does not increase cmDCF fluorescence. By contrast, the combination of bafilomycin to release iron and t-BOOH to cause mild mitochondrial oxidative stress produces a marked increase of cmDCF fluorescence, indicative of an intramitochondrial oxidative stress that then leads to the mitochondrial permeability transition (MPT) and cell death (Uchiyama et al. Hepatology 2008; 48, 1644-1654.). Desferal, starch desferal (which enters lysosomes but not the cytosol by endocytosis) and Ru360 (MCU inhibitor) block these events.
c) The conversion of non-fluorescent dihydrodichlorofluorescein to its fluorescent form likely involves reactions with activated lipid hydroperoxides formed downstream of Fenton chemistry and lipid peroxidation (Cathcart R et al. Anal Biochem 1983;134:111-116). Thus, a number of chemical species may be causing the fluorescent conversion, but DCF is still a useful indicator of ROS formation since species that react with DCF are the result of ROS generation.
d) H2DCF is great for fluorescent plate readers. However, one cannot load H2DCF-DA in advance, because H2DCF is so reactive that it will be exhausted due to room light, etc., by the time you make your first measurement. Rather, we incubate with H2DCF-DA (or H2-cm-DCF) as we make measurements to replenish H2DCF continuously as it is exhausted by conversion to DCF (Nieminen, A.-L et al. Am. J. Physiol. 1997; 272, C1286-C1294).
e) H2DCF and H2cmDCF both work well for confocal microscopy, but H2cmDCF is better because it forms covalent adducts with intracellular proteins and is not released after plasmalemmal or mitochondrial membrane permeabilization. H2cmDCF, especially, allows determination of the subcellular sources of ROS formation. The trick is to keep laser excitation to the minimum possible. Too much laser light with convert H2DCF to DCF directly. H2cmDCF shows very nicely mitochondrial ROS formation in adult cardiac myocytes after ischemia/reperfusion leading to the MPT and cell death, events that are blocked by desferal and antioxidants like diphenylphenylenediamine (DPPD, a blocker of peroxidation chain reactions) and 2-mercaptopropionyl glycine (2-MPG, a GSH analog) (Kim J.-S. et al. J. Physiol. 2006; 290, H2024-H2034).

John J. Lemasters, MD, PhD
Professor and GlaxoSmithKline Distinguished Endowed Chair Director,
Center for Cell Death, Injury & Regeneration
Departments of Pharmaceutical & Biomedical Sciences and Biochemistry & Molecular Biology
Medical University of South Carolina
DD504 Drug Discovery Building 70
President Street, MSC 140 Charleston, SC 29425
Email: JJLemasters @ musc.edu

A3:  My point is to bring awareness to the potential pitfalls. Please see Chem. Res. Toxicol. 5:227-231, 1992. and Arch. Biochem. Biophys. 302:348-355, 1993". These articles showed for the first time that oxidation of DCFH is not by hydrogen peroxide but it requires a stronger oxidant that is either metal catalyzed of peroxidase catalyzed. There are two fundamental issues with these dyes:

1) Second order rate constants are sufficiently low that unless one can load mitochondrial with mM concentrations then it unlikely they can compete with Mn SOD for superoxide (second order rate constant >10^9 M-1 s-1).

2) All these dyes will undergo one electron oxidation yielding a radical intermediate e.g., DCFH will give DCFdot (the one electron oxidation product). The radical intermediate with then reduce oxygen to superoxide. The second order rate of this reduction (one-electron oxidation of DCFH with oxygen to form superoxide is rather fast 10^8 M-1s-1!! Essentially all one needs is one electron oxidation to kick start the process which can then self-propagate. Since it is difficult to know the rate of self-propagation vs. the rate of oxidation by oxidants it is difficult to compare either rates or final product and determine the true oxidation rate. See Wardman P. Free Radic Biol Med. 2006 Aug 15;41(4):657-67 and Ron Mason Free Radic. Biol. Med. 40:968-975; 2006.

Harry Ischiropoulos     ischirop @ MAIL.MED.UPENN.EDU

A4: Some words of caution about the use of MitoSOX.

a) Regardless of the actual ROS species (let's say superoxide for simplicity) one measures the rate of increase in fluorescence in mitochondrial regions; this is proportional to the level of ROS times the concentration of MitoSOX in the matrix, with a loading proviso (see below).

b) The equilibrium matrix concentration is a function of the external MitoSOX concentration and the plasma and mitochondrial membrane potentials, according to the Nernst equilibria. If either potential changes, or differs between samples, then corrections must be applied. We run a non-quench TMRM experiment in parallel and assume that the MitoSOX concentration in the matrix parallels that of TMRM.

c) We find that MitoSOX takes about 30min to load to equilibrium in neurons. Therefore the first 30 min of fluorescence increase is non-linear and should be ignored.

d) As with hydroethidine, the fluorescence of the product is greatly enhanced by intercalation into DNA (in this case mtDNA). it is therefore important that the binding capacity of mtDNA is not exceeded (otherwise overspill onto nDNA occurs). Look at our paper: Johnson-Cadwell et al. (2007) J. Neurochem. 101, 1619-1631

PS: despite worries about the nature of the oxidation product, I haven't seen the distinctive emission of hydroxyethidium (rather than ethidium) Taking Harry's first point, the inefficient competition with SOD of Het and MitoSOX can actually be an advantage: you don't measure the amount of ROS but rather the steady-state level, which is as relevant. If you only trap about 1% of the superoxide it means that you don't disturb the system significantly. If your project is to look at the effects of ROS it would be no good removing it all with the assay.

David G. Nicholls
Professor of Mitochondrial Physiology.


Last Updated ( Saturday, 18 February 2012 )


    Copyright 2004 - 2005  OXPHOS  International Pty Ltd. All rights reserved.   Vizitors: