Mitochondrial-targeted fluorescent probes for reactive oxygen species

Abstract

As the primary consumers of oxygen within all aerobic organisms, mitochondria are a major source of cellular reactive oxygen species (ROS) derived from the in vivo chemistry of oxygen metabolism. Mitochondrial ROS have been traditionally implicated in aging and in a variety of pathologies, including cancer, neurodegeneration, and diabetes, but recent studies also link controlled mitochondrial ROS fluxes to cell regulation and signaling events. Progress in the development of mitochondrial-targeted fluorescent small-molecule indicators that detect specific ROS with high selectivity offers a promising approach for interrogating mitochondrial ROS production, trafficking, and downstream biological effects.

Introduction

Reactive oxygen species (ROS) have emerged as prevalent and important components of both physiological and pathological states of living organisms [1–9]. Classically, the presence of ROS in biological systems has been associated predominantly with disease and form the underpinning for the “free-radical theory of aging” (FRTA) [10]. Indeed, oxidative stress is connected to many diseases where age is a risk factor [11], including cancer [12, 13] and neurodegenerative disorders such as Alzheimer’s, Parkinson’s, and Huntington’s diseases [14, 15]. On the other hand, the controlled release and compartmentalization of ROS is also critical to maintaining normal physiology. For example, macrophages engulf invading pathogens into phagocytic vesicles and then produce a variety of ROS inside the vesicles to help neutralize the threat. Moreover, the production of certain ROS, such as hydrogen peroxide (H₂O₂), has been detected in a variety of tissues, and mounting evidence suggests this ROS is utilized as a signaling molecule for a wide range of healthy physiological events [16–22].

The complex biology of ROS is patently dictated by the chemical properties of each specific oxygen metabolite as well as the localization and trafficking of that metabolite at the cellular level. In this context, mitochondria are the major consumers of cellular oxygen and hence play a central role in ROS biology. Harman’s FRTA focused on mitochondria as the “biologic clocks” of the cell that possess the primary generators and targets of ROS [23]. Figure 1 outlines various potential sources of mitochondrial ROS [24–26], and extensive reviews have been written on this topic [25] [27, 28]. The balance between the production and destruction of ROS in the mitochondria has critical in vivo implications [29], as the overexpression of either superoxide dismutase (SOD) [30] or catalase [31] enzymes that consume O₂⁻ and H₂O₂, respectively, can increase the lifespan of various model organisms. Excess ROS can clearly have aberrant consequences, but the fact that aerobic organisms have conserved the ability to generate a basal level of mitochondrial ROS suggests that a minimum threshold of these oxygen metabolites is beneficial and provides some form of evolutionary fitness.

Figure 1.

A simplified scheme for mitochondrial ROS metabolism. The primary source of ROS in the mitochondria is derived from the one-electron reduction of molecular oxygen (O₂) by the electron transport chain (ETC) to form superoxide (O₂⁻). O₂⁻ can then be converted to hydrogen peroxide (H₂O₂) either spontaneously or catalyzed by superoxide dismutases (SOD). H₂O₂ can also be produced by monoamine oxidases (MAO), which catalyze the oxidative deamination of dietary and neurotransmitter amines. Mitochondrial nitric oxide synthases (NOS) produce nitric oxide (NO) that can potentially react with superoxide to produce peroxynitrite (ONOO⁻). H₂O₂ can either be destroyed by peroxiredoxins (PRX), glutathione peroxidases (GPx), or catalases (Cat), converted to a hydroxyl radical (•OH) by iron or copper-mediated Fenton chemistry, or transformed into hypochlorous acid (HOCl) by myeloperoxidase (MPO) catalysis.

The complex interplay between mitochondrial ROS and health, aging, and disease provides motivation for developing new tools to study the chemistry and biology of ROS in living systems with high spatial and temporal resolution. Fluorescence imaging with mitochondrial-targeted reporters for monitoring ROS offers a potentially powerful methodology to achieving this goal, and this review will summarize selected recent examples of small-molecule probes that simultaneously localize to the mitochondria and respond to various ROS molecules. Although outside the scope of this review, we note that protein-based fluorescent reporters provide a complementary approach to targeted mitochondrial ROS imaging and several elegant examples have been published [32–37].

Designing small-molecule probes that target to mitochondria

Figure 2 outlines various approaches to deliver small-molecule probes to cellular mitochondria. The most common way to target molecules to the mitochondria of living cells is through the use of lipophilic cations, which are attracted to the negative potential caused by the proton gradient across the inner mitochondrial membrane. For example, rhodamine and the MitoTracker dyes possess an overall positive charge that is delocalized through resonance, allowing for passage through plasma membranes and subsequent accumulation within mitochondria [38]. Through the same general mechanism, Murphy and colleagues have championed the use of triphenylphosphonium (TPP) head groups to deliver a variety of cargoes to the mitochondria, including antioxidants such as vitamin E [39], coenzyme Q [40], S-nitrosothiols [41], as well as SOD and peroxidase mimics [42]. A newer approach developed by Kelley and co-workers utilizes mitochondria penetrating peptides (MMPs). By balancing charge and liophilicity through a combination of natural and synthetic amino acids, various cargoes can be delivered to the mitochondria, including a singlet-oxygen generating fluorophore [43] that allows the localized generation of this ROS. We refer the reader to more comprehensive reviews for TPP [44] and general mitochondrial targeting [45] that have appeared in the recent literature, and the remainder of this review will present a brief survey of probes for various mitochondrial ROS molecules.

Figure 2.

General methods for delivering molecular cargo to the mitochondria. Small molecules can be delivered to the mitochondria through the use of lipophilic cations such as rhodamine dyes (a) and triphenylphosphonium moieties (b), which take advantage of the proton gradient and subsequent electrochemical potential generated within the matrix of mitochondria. Recently, mitochondria-targeted peptides (c) that can contain both natural and unnatural amino acids have been rationally designed and screened as mitochondrial delivery vehicles [45].

A fluorescent probe for mitochondrial superoxide

Derived from the one-electron reduction of oxygen, superoxide (O₂⁻) is a marker for early ROS generation in primary or secondary chemical reaction cascades and a major ROS leaking from the respiratory chain. The most widely used fluorescent indicator for O₂⁻ is hydroethidium (HE), a two-electron reduced form of the nucleic acid stain ethidium. HE has been employed to detect ROS production during phagocytic respiratory bursts [46] and intracellular oxidative stress [47]. Appending a TPP targeting group to HE yields MitoSOX; the TPP successfully directs the probe away from the nucleus and to the mitochondria of living cells (Figure 3a). A major experimental complication for applying HE probes to monitor O₂⁻ specifically is that HE can be oxidized by multiple ROS to yield multiple fluorescent species; however, only O₂⁻-dependent oxidation can generate a hydroxylated product (2-OH-E) that has a unique excitation at 396 nm in addition to the typical ethidium excitation at 510 nm [48]. Results obtained using only ethidium excitation at 510 nm must be interpreted with caution as they are likely to overestimate the contributions of O₂⁻ while underestimating the roles of other ROS. Careful control experiments and/or direct excitation of the 2-OH-E product are necessary to distinguish whether O₂⁻ is the dominant participating ROS in a given situation [49].

Figure 3.

Selected mitochondrial-targeted probes for detection of reactive oxygen species (ROS). (a) MitoSOX is a dihydroethidium-based probe bearing a triphenylphosphonium (TPP) targeting moiety. This probe reacts with several ROS, but the product from superoxide oxidation can be distinguished from other potentially formed oxidized products by selective excitation of the 2-hydroxylethidium product. (b) Two examples of mitochondrial-targeted nitrone spin-trap probes that utilize TPP localization groups. (c) MitoAR and MitoHR are rhodamine-like probes that react with highly reactive oxygen species (hROS) including hydroxyl radical (•OH), hypochlorous acid (HOCl), and peroxynitrite (ONOO⁻). (d) MitoPY1 is a boronate-based hybrid rhodamine/fluorescein probe with a TPP targeting group for chemospecific detection of hydrogen peroxide (H₂O₂).

Mitochondrial-targeted probes for free radicals

Oxygen-derived free radicals are frequently associated with aging and disease, and mitochondria are a major source and target of these damaging molecular oxidants. Electron paramagnetic resonance (EPR) spectroscopy is a useful way to detect free radical ROS and other molecules possessing unpaired electrons [50]. In particular, spin trapping with nitrones like TEMPO can generate, upon reaction with ROS, persistent nitroxide adducts that are detectable by EPR. Several laboratories have described the use of mitochondrial-targeted nitrone traps for EPR detection or antioxidant treatments using TPP (Figure 3b, left) [51] or hemigramicidin fragments of the membrane-active antibiotic Gramicidin S for cargo delivery [52, 53]. A trifunctional reporter combining a TPP for mitochondrial targeting, a TEMPO derivative for spin trapping, and a fluorescein moiety for optical tracking has been reported (Figure 3b, right) [54]. Confocal microscopy experiments show that this probe accumulates within the mitochondria of live RAW264.7 macrophages, but more studies are needed to validate this potential free radical imaging strategy.

Fluorescent mitochondrial probes for highly reactive oxygen species

Highly reactive oxygen species (hROS), including hydroxyl radical (•OH), peroxynitrite (ONOO⁻), and hypochlorous acid (HOCl), are potent oxidants that are capable of directly oxidizing nucleic acids, proteins, and lipids. hROS are generated in secondary cellular ROS reaction cascades involving O₂⁻ and H₂O₂. For example, H₂O₂ can trigger uncontrolled Fenton chemistry in the presence of iron or copper centers that are prevalent in the mitochondria to make the exceedingly reactive and damaging •OH species. Likewise, nitric oxide (NO) generated by plasma membrane or mitochondrial nitric oxide synthases can combine with O₂⁻ to make the nitrating oxidant ONOO⁻. In addition, myeloperoxidases (MPOs) catalyze the transformation of H₂O₂ to HOCl, which can then potentially diffuse into and damage the mitochondria. To monitor mitochondrial hROS, Nagano and co-workers have exploited rhodamine-like fluorophores capped with either a 4-aminophenyl aryl ether (MitoAR) or a 4-hydroxy aryl ether group (MitoHR) (Figure 3c). The ether capping groups quench the fluorophore emission by photoinduced electron transfer (PET). Reaction with hROS cleaves off the PET quenching moiety, resulting in the highly fluorescent rhodamine type reporter [55]. MitoAR responds mainly to •OH and HOCl, whereas MitoHR is most sensitive to •OH. Both probes react with ONOO⁻, but at a slower rate. Fluorescence imaging experiments establish that MitoAR can accumulate selectively within the mitochondria of live HeLa cells and respond with increased fluorescence to exogenously added HOCl. Importantly, treatment of MitoAR-loaded HeLa cells, which are lacking MPO, with H₂O₂ does not result in a fluorescence turn-on. However, addition of H₂O₂ to MitoAR-loaded HL-60 cells, which do possess MPO, causes a turn-on fluorescence increase.

A targetable fluorescent probe for mitochondrial hydrogen peroxide

Major chemical pathways to cellular H₂O₂ production include the incomplete reduction of O₂ to H₂O during mitochondrial electron transport, the two-electron, two-proton reduction of O₂ by oxidase activity, or the reduction of O₂⁻ by spontaneous or SOD-catalyzed processes. Aberrant H₂O₂ fluxes within mitochondria can trigger apoptosis, but controlled bursts of H₂O₂ elicited by specific ligand-receptor interactions are beneficial to metabolic function. To help disentangle the diverse contributions of H₂O₂ to mitochondrial biology, we sought to create new chemical tools for monitoring H₂O₂ levels within this subcellular locale. Previous work from our laboratory established that the H₂O₂-mediated conversion of aryl boronates to phenols can provide a chemospecific, biologically compatible reaction method for detecting endogenous H₂O₂ production [56–60]. Appending a TPP moiety onto a boronate-masked hybrid fluorescein/rhodamine reporter gives MitoPY1 (Figure 3d), which shows good localization to the mitochondria of a variety of mammalian cell types, including HeLa, HEK293, Cos-7, and CHO.K1 [61]. MitoPY1 can respond selectively to rises in H₂O₂ levels by a turn-on fluorescence increase as determined by both confocal imaging and flow cytometry experiments (Figure 4). Furthermore, MitoPY1 can detect mitochondrially-derived H₂O₂ in a neurodegenerative oxidative stress model; treatment of HeLa cells with paraquat, an uncoupler of the mitochondrial electron transport chain and small-molecule model for Parkinson’s disease, causes a robust, mitochondrial-localized fluorescence enhancement due to elevated generation of H₂O₂.

Figure 4.

Confocal fluorescence images of live Cos-7 cells with varying levels of mitochondrial H₂O₂ as visualized using MitoPY1. Cos-7 cells incubated with 5 µM MitoPY1 for 60 minutes at 37 °C and imaged with either MitoPY1 (a), MitoTracker Deep Red and Hoechst (overlay, b), MitoPY1 with MitoTracker Deep Red (overlay, c), or in brightfield mode (d). Cos-7 cells incubated with 5 µM MitoPY1 with 300 µM H₂O₂ added for the final 40 minutes and imaged with either MitoPY1 (e), MitoTracker Deep Red and Hoechst (overlay, f), MitoPY1 with MitoTracker Deep Red (overlay, g), or in brightfield mode (h) with a 20 µm scale bar.

Conclusions and outlook

The foregoing examples have highlighted a select but growing number of targeted small-molecule chemical tools for studying the biology of mitochondrial ROS. These indicators, along with emerging approaches toward detecting peroxynitrite [62], superoxide [63], nitric oxide [64, 65], hypochlorous acid [66, 67], ozone [68], and reversible redox changes [69], presages the possibility that one can help elucidate the dynamic cascades involving mitochondrial ROS metabolism using a family of targeted reagents that detect different ROS molecules. In addition, one can also envision using a set of probes with selectivity for a specific ROS to construct a multicolor calibration scale for the mitochondrial redox environment. Combining parallel advances in the delivery of molecular cargoes to the mitochondria with the development of highly specific and selective molecular switches would be of potential broad utility in further deepening our understanding of mitochondria redox biology, as well as delineating their contributions to the more complex oxidation biology of living cells, tissue, and organisms.