A comprehensive understanding of ROS (reactive oxygen species)-dependent cellular interaction requires the previously unmet ability to simultaneously monitor the intra- and extra-cellular environments. The present review assesses the potential of novel electrochemical and fluorescent-based nanosensor approaches to address the limitations of existing techniques for ROS analysis. Data generated by these new approaches have already contributed significantly to current understanding of the roles that these species play in various in vitro scenarios. However, integration of these novel approaches has the potential to offer, for the first time, the unparalleled ability to measure simultaneously and in real-time ROS flux in both the intra- and extra-cellular environments.
ROS (reactive oxygen species) such as superoxide (O2–) play a key role in cellular physiology. However, if their production exceeds the body's natural ability to deal with these species, damage to protein, DNA and lipid may occur, a process that is central to the pathological importance of ROS . For example, it is known that the O2– in particular reacts rapidly with endogenous NO (nitric oxide) to form a variety of species [e.g. peroxynitrite (ONOO–) and hydroxyl radicals (OH•)] that are highly damaging to biomolecules.
In order to gain a full understanding of the role that O2– and ROS in general play in pathology, it is essential to measure these species in a variety of in vitro and in vivo models. However, the high reactivity of most free radicals makes their specific detection difficult. Attempts have been made to study free radicals in a variety of ways [2–6]. Typically these are indirect end-point analytical techniques such as the measurement of lipid peroxides  or the formation of spin trap adducts for subsequent analysis by ESR spectroscopy . It would represent a significant analytical advance if it were possible to measure the production of these analytes in living cells and in real time. Critically, such a system must exhibit rapid ROS response times (μs) with minimal influence upon the cellular environment. The present review looks at emerging electrochemical and optical techniques which can address the limitations of traditional ROS detection methodologies. The importance of these new approaches is illustrated by summarizing key examples of how these technologies have already provided new insight into the role that ROS play as cell signalling molecules and as initiators of cellular pathology. The possible integration of these techniques into a platform for the simultaneous and real-time detection of intra- and extra-cellular ROS production is also discussed. Such a format has the potential to significantly progress our understanding of free radicals in biology and medicine.
Electrochemical O2– detection
In the late 1970s, it was realized that suitably modified (or functionalized) electrode surfaces could interact in a specific and non-degradative manner with proteins, to allow stable and reversible direct electrochemistry that was not affected by artefacts .
Direct electrochemistry of the redox protein cytochrome c was obtained from a functionalized gold electrode by McNeil et al. . In this work, the gold electrode was modified with N-acetylcysteine. Once covalently attached, the immobilized cytochrome c was used as an integral part of an amperometric O2– sensor. Superoxide generated by XOD (xanthine/xanthine oxidase) caused the one electron reduction of cytochrome c3+ to cytochrome c2+. The reduced protein was then reoxidized at the electrode surface (poised at +25 mV with respect to a silver/AgCl reference electrode). Current rates recorded were directly proportional to the rate of O2– production by XOD.
Manning et al.  designed a much simpler protocol for the fabrication of the system described by McNeil et al.  by covalently attaching cytochrome c to a gold working electrode through surface modification with DTSSP [3,3′-dithiobis(sulfosuccinimidylpropionate)]. This linking molecule possesses a disulfide group for covalent attachment to the gold surface. Following attachment, the molecule presented two carboxy groups that readily formed amide linkages with cytochrome c. The resulting electrode was poised at +100 mV (against silver/AgCl). Superoxide generated by xanthine/XOD caused the one-electron reduction of cytochrome c3+ to cytochrome c2+. The reduced protein was then reoxidized at the electrode surface. Additionally, this electrode has a rapid response time (<100 μs) and a calculated O2– detection limit of 10 nM . This electrode has successfully been used to detect O2– in a wide variety of in vitro biological applications [9–15]. The specificity of the O2– electrode has been extensively reported [9–15] and has been primarily attributed to a low operating potential of +100 mV (against silver/AgCl reference electrode).
Direct real-time monitoring of superoxide generation in isolated mitochondria
Work carried out in the research groups of Manning and McNeil recently used the cytochrome c-modified gold electrode to identify the role of specific respiratory-chain complexes in mitochondrial O2– production . ROS within cells are known to act as secondary messengers in intracellular signalling cascades, which induce and maintain the oncogenic phenotype of cancer cells. Redox imbalances have been found to be present in many cancer cells compared with normal cells . Thus redox imbalance may be related to oncogenic stimulation. Identification of the site of ROS production within mitochondria could lead to the development of new therapeutic targets.
Evidence suggested that most of the O2– generated by the mETC (mitochondrial electron-transport chain) originated from Complexes I and III . Our work set out to demonstrate the relative contribution of these complexes to O2– production . In this study, mitochondrial electron-transport Complexes I and III were selectively inhibited with rotanone and antimycin A respectively and resulting O2– production monitored electrochemically. This method allowed, for the first time, the dynamics of mitochondrial O2– release following the specific inhibition of complexes I and III of the mETC to be examined comparatively and in real time .
Inhibition of Complex III with antimycin A produced a large and immediate increase in O2– production (Figure 1A). Complex III is known to asymmetrically generate O2– that passes both into the matrix and into the intermembrane space of the mitochondrion. However, O2–, once in its anionic form, it is too strongly charged to readily cross the mitochondrial inner membrane [19,20]. Thus O2– production exhibits a distinct membrane sidedness or ‘topology’. Furthermore, it has been known for some time that partial turnover of the cytochrome bc1 components within Complex III in the presence of antimycin A, a Qi site inhibitor, results in accumulation of a semiquinone at the Qo site, which can result in O2– production on the outer aspect of the inner membrane adjacent to the intermembrane space . As a result of these combined features, we proposed that the O2– detected by the electrode was that released outside the inner membrane . Superoxide generation at Complex I was attributed to electron leakage at the quinine-binding site  and in the presence of a Q-site inhibitor (such as rotenone) the rate of O2– production can increase 10–30-fold . Rotenone was believed to block electron transfer from the N2 iron–sulfur cluster to ubiquinone . However, as O2– generated at Complex I was released exclusively into the mitochondrial matrix , O2– should not be detected by the electrochemical method. Consistent with this, inhibition of Complex I by rotenone (Figure 1B) produced a much smaller change in current than that observed during Complex III inhibition (Figure 1A). However, the low permeability of the mitochondrial membrane to O2– would suggest that no O2– would be detected by the electrode following the addition of rotenone, yet the change in current observed demonstrates that some O2– was produced (Figure 1B). We suggested that O2– detected following the addition of rotenone may have been generated from upstream sites in Complex I , as the site of action for rotenone is thought to be in the distal position of the complex . The in vitro specificity of the electrode for O2– has been reported previously [9–15] and was confirmed in our previous study through the addition of SOD (superoxide dismutase) during and prior to O2– generation (Figures 1C and 1D) . The re-stabilization of the current response at a level below the pre-inhibition baseline observed in Figure 1(C) was accounted for by the presence of O2– in the sample prior to antimycin A addition. It was noted that the current response observed following the addition of both rotenone and antimycin A to isolated mitochondria was much larger than would be expected in whole-cell samples . This was not surprising considering the fact that the antioxidant mechanisms that exist in cellular systems were not present in the mitochondrial fractions. Superoxide is rapidly converted into H2O2 by MnSOD in the mitochondrial matrix  or by Cu/ZnSOD in the intermembrane space and cytosol .
Superoxide production following selective mitochondrial respiratory complex inhibition
The data presented in our previous study  clearly demonstrated the analytical capabilities of amperometric sensing for the direct, real-time analysis of qualitative O2– generation from isolated mitochondria following the chemical modulation of electron-transport complexes. We concluded that the sensing technology had clear potential to greatly improve current understanding of O2– flux in vitro.
Intracellular fluorescence-based analysis
Direct cell loading of fluorescent dyes continues to be the method of choice for quantification of intracellular analytes. However, such ‘free dyes’ can themselves be cytotoxic and are prone to bind non-specifically to proteins and cell organelles. These limitations were initially addressed by Sasaki et al. , who were the first to describe the encapsulation of fluorescein in a porous polyacrylamide nanoparticle and applied this technology to pH-sensing. In 1998, the PEBBLE (probes encapsulated by biologically localized embedding) concept was introduced by Kopelman and co-workers , specifically using fluorescence-based optical nanosensors to make, for the first time, intracellular measurements. Nanosensor fabrication was based on microemulsion polymerization techniques to obtain nanoparticles of an even size distribution (typically 50 nm in diameter based on AFFF (asymmetric field flow fractionation) characterization studies ). PEBBLE nanosensors for calcium , zinc , glucose  and pH  have all been reported. Numerous methods for introducing them into the intracellular environment have been reported including linking nanosensors to the cell penetrating TAT peptide , physical insertion via a gene gun  and utilizing innate cellular phagocytosis . All of these methods have their limitations . However, a recent innovation that delivered nanosensors via lipid transfection  has been shown to offer a reliable method that was suitable for both immortalized cell lines and primary cells. The size of the devices enables insertion into cells with minimal physical perturbation. Additionally, PEBBLEs allow ratiometric analysis. The inclusion of a non-responsive ‘control’ flurophore together with the sensing dye means that each nanosensor incorporates its own internal control signal. Work carried out by Henderson et al.  has recently described the development and application of an ROS-responsive nanosensor.
The development and in vitro characterization of an intracellular nanosensor responsive to ROS
This work described the synthesis and in vitro application of a novel ROS-responsive nanosensor, based on PEBBLE technology. The ROS-sensitive fluorescent probe DHR 123 (dihydrorhodamine 123; λex 500 nm and λem 536 nm) was employed as the sensing element of the PEBBLE through entrapment within a porous, bio-inert polyacrylamide nanostructure, enabling passive monitoring of free radical flux within the intracellular environment. Alexa Fluor® 568 (λex 578 nm and λem 603 nm) was used as the non-responsive reference dye. Successful delivery of the nanosensors into NR8383 rat alveolar macrophage cells via phagocytosis was achieved. Concentration-dependent stimulation of PEBBLE loaded NR8383 cells with PMA enabled real-time monitoring of ROS generation within the cell without affecting cellular viability (Figure 2A). The nanosensors also allowed, for the first time, ROS generation within the macrophages to be followed in real time (Figure 2B). These data showed that PEBBLE nanosensors could offer real-time ratiometric analysis of ROS flux within a cell which could not be reliably achieved with the existing technology.
Fluorescent intracellular ROS detection
Advancement towards an integrated ROS detection platform
The simultaneous and real-time measurement of multiple ROS in both the intracellular and extracellular environments will provide unparalleled information about ROS as cell signalling species and as initiators of pathology. To this end, several groups have begun to develop miniaturized electrode arrays which will selectively sense the production of multiple ROS. Electrode arrays, in which each individual electrode is specifically tailored to selectively detect a particular analyte by functionalizing it with the appropriate sensing chemistry, have provided a development platform for the simultaneous detection of multiple species. Such arrays have been used to investigate the response of cells to drug treatment [35,36]. Further, it has been possible to map cellular responses to locally delivered chemical stimulus by collecting dynamic information about the desired specific analyte. McNeil and other groups have previously shown the feasibility of arrayed planar disc electrodes in a 24-well cell culture plate format for the simultaneous detection of NO and glutamate  and NO and O2– . In the latter example, O2– sensing was carried out at gold electrodes functionalized with cytochrome c as previously described . NO detection was carried out at gold electrodes coated with carbon. NiTSPc (nickel tetrasulfonated phthalocyanine) was then electrodeposited on to these carbon disc electrodes (1 mm outer diameter) followed by coating in Nafion® . This platform also provided a high degree of biocompatibility which would allow cells to be cultured directly on the arrayed electrodes (thus minimizing their distance from the sensing surface). The work reported by McNeil et al.  demonstrated the potential of the electrode platform as a potential drug-screening system by using A172 glioblastoma cells. These macrophage-like cells were cultured directly on the electrode arrays and chemically stimulated to produce ROS with the protein kinase C agonist PMA. Simultaneous and direct monitoring of NO and O2– production were recorded (Figure 3A). The results obtained demonstrated that it would be possible to envisage a drug-screening platform for compounds designed to be inhibitors of NOS (nitric oxide synthase) or to have an inhibitory effect on O2– production.
Electrochemical extracellular ROS detection
Conclusions and future developments
A key and as yet unrealized goal in the area of ROS quantification is the creation of a unified platform that will enable simultaneous intra- and extra-cellular analysis of free radical and ROS production directly and in real time. However, the ability of amperometric electrochemical sensing devices to offer rapid, selective and sensitive free radical measurements in real-time has been well established. Equally the limitations of traditional fluorescent probes for intracellular analysis have been addressed through the development of polyacrylamide nanosensors. The next step must be to combine these proven technologies into an integrated platform capable of simultaneously measuring intracellular ROS-generating events and subsequently following their externalization and influence on adjacent cellular populations. Such a technology would, for the first time, enable ROS-based signalling events to be comprehensively analysed. This in turn would significantly advance current understanding of the role of ROS in the biology of disease.
Analysis of Free Radicals, Radical Modifications and Redox Signalling: A Biochemical Society Focused Meeting held at Aston University, Birmingham, U.K., 18–19 April 2011. Organized and Edited by Helen Griffiths (Aston University, U.K.), Corinne Spickett (Aston University, U.K.) and Paul Winyard (Exeter, U.K.).
We acknowledge financial contributions from both the Biotechnology and Biological Sciences Research Council and the Royal Society of Chemistry that supported this work.