Mitochondria are the major source of reactive oxygen species (ROS). During mitochondrial respiration, nearly 0.1–4% of oxygen is reduced to the superoxide ion (O
2•−) due to electron leakage from the respiratory chain. This species is then transformed into other ROSs via enzymatic or non-enzymatic pathways
[25]. Meanwhile, endogenously produced H
2S is oxidized in the presence of mitochondrial ROS to form sulfane sulfurs, which can also be formed directly via enzymes such as 3MST. Thus, to better understand redox homeostasis, monitoring sulfane sulfurs via fluorescence imaging is useful. Mitochondria possess a unique double-layered membrane structure with a negative membrane potential (as high as −180 mV)
[26]. Hence, in most cases, mitochondria-targeted probes possess at least one lipophilic cation
[27]. Non-cationic probes can be functionalized by attaching triphenyl phosphonium
[20] or pyridinium
[28][29][30] as the anchor. However, functionalized cationic dyes are also known to target other organelles
[31][32][33][34][35]. Based on colocalization experiments with commercially available mitochondria-targeting dyes, some non-cationic dyes have been reported to selectively target the mitochondria due to their unique structures
[36].
In 2015, Chen and coworkers developed a reaction-based near-infrared (NIR) fluorescent probe (Mito-ss) for the detection of mitochondrial hydrogen polysulfides (H
2S
n, n > 1)
[37]. Mito-ss consists of (i) a NIR dye based on the azo-BODIPY chromophore, (ii) a lipophilic triphenyl phosphonium group, and (iii) an H
2S
n-reactive nitrofluorobenzoate moiety (
Scheme 1a). They chose a NIR fluorophore because NIR lights possess certain advantages, including deep tissue penetration, low cytotoxicity, and minimum background noise. Nitrofluorobenzoate is a commonly used functional group for the design of H
2S
n sensors
[38]. Nitrofluorobenzoate bears two electrophilic sites. H
2S
n first reacts with it via nucleophilic aromatic substitution (S
NAr) to replace the F atom and form a persulfide (-SSH) intermediate, which then undergoes a spontaneous intramolecular cyclization with the ester group to uncage the fluorophore. Due to its electron with-drawing nature, nitrofluorobenzoate quenches the fluorescence of the azo-BODIPY chromophore via a donor-excited photoinduced electron transfer (d-PET) process. As such, Mito-ss is a reaction-based ‘turn-on’ sensor for H
2S
n. Mito-ss reacts rapidly (~30 s) with H
2S
n and exhibits a 24-fold fluorescence increase at an emission of 730 nm. The probe was examined with various ROS, reactive nitrogen species (RNS), and other RSS and demonstrated no fluorescence turn-on. Biothiols such as glutathione (GSH), cysteine (Cys),
N-acetyl-L-cysteine, etc., could react with Mito-ss. However, as the reaction stopped at the S
NAr step, no fluorescence was observed. The limit of detection (LOD) for Mito-ss was calculated to be 25 nM. The probe was used for the real-time detection of exogenous and endogenous H
2S
n using six different cell lines. Mito-ss was also found to be suitable for the in vivo detection of exogenously injected H
2S
n in BALB/c mice.
Scheme 1. Structures and reactions of probes (a) Mito-ss, (b) Mito-NRT-HP, (c) HCy-FN, and (d) HCy-Mito.
Using the same nitrofluorobenzoate reaction site, Han et al. developed a ratiometric fluorescent probe, Mito-NRT-HP, for the detection of mitochondrial H
2S
n in 2019
[39]. The structure of Mito-NRT-HP is similar to Mito-ss, though with a two-photon responsive naphthalimide fluorophore instead of a single-photon responsive fluorophore. The naphthalimide fluorophore has the advantage of easily tunable photophysical properties by blocking and/or unblocking the internal charge transfer (ICT) process. It is highly photostable, resistant to pH interference, and possesses a large two-photon absorption cross-section. Importantly, 1,8-naphthalimide can be easily functionalized by simple synthetic tailoring
[40]. The main advantages of two-photon excitation over single-photon excitation include deep tissue penetration, lesser damage, poor scattering, etc. In the case of two-photon excitation, a femto second pulsed laser is used, and the molecule can be excited only at the focal point of the laser. Three-dimensional imaging can be obtained
[41]. Upon titrating with different concentrations of Na
2S
2, it was found that Mito-NRT-HP gave ratiometric responses with a changing fluorescence color from blue to green. When the solution of Mito-NRT-HP was treated with H
2S
n, the initial emission maximum at 478 nm decreased gradually, with a concomitant peak increase at 546 nm. The detection limit was 10 nM which suggests that Mito-NRT-HP could have the relevant sensitivity needed for the quantitative detection of H
2S
n under physiological conditions. The two-photon absorption cross-section values (δ) of Mito-NRT-HP and its fluorophore Mito-NRT (
Scheme 1b) were recorded in a buffer using a pulsed laser, and fluorescein was used as the reference molecule. Their δ values were measured over a range of wavelengths starting from 750 nm to 825 nm. The highest δ was 290 GM [1 GM (Goeppert-Mayer) = 10
−50 cm
4 s photon
−1] for Mito-NRT-HP and ~190 GM for Mito-NRT at 810 nm. Mito-NRT-HP was found to exhibit good cell permeability and weak cytotoxicity, which was suitable for the ratiometric imaging of endogenous H
2S
2 in cells. Mito-NRT-HP was colocalized with MitoTracker Red (MTR) and LysoTracker Red (LTR), and the colocalization coefficients were found to be 0.94 and 0.42, respectively, indicating that Mito-NRT-HP was specifically localized in the mitochondria. Using two-photon microscopy, images of the tissue slices from mice with lipopolysaccharide (LPS)-induced acute organ injury were taken and compared with the control tissues. The enhanced fluorescence in the former case was observed. In 2021, Han et al. reported a similar probe for the detection of mitochondrial H
2S
n during H
2O
2-induced redox imbalance
[42]. The structure of this probe (Mito-RHP) only differed from Mito-NRT-HP in the linker between naphthalimide and the triphenylphosphonium unit. Upon the addition of Na
2S
2 to the solution of Mito-RHP, the initial emission spectra of Mito-RHP at 485 nm gradually decreased, and a continuous increase in the new peak to 550 nm was observed, along with a change in fluorescence color from blue to yellowish green. In this case, the Stokes shift was 109 nm, which was higher than that of Mito-NRT-HP. The detection limit was calculated to be 20 nM. Other properties, such as photostability, solubility, permeability, and cytotoxicity, were similar. However, the mitochondria-targeting ability of the new probe (overlap coefficient = 0.836) was not as good as that of Mito-NRT-HP (overlap coefficient = 0.94). The in vivo imaging of exogenous H
2S
n (using Na
2S
2) was performed in zebrafish using Mito-RHP.
An interesting single-component multi analyte responsive NIR fluorescent probe was reported by Chen and coworkers in 2015 for the detection of the superoxide ion (O
2•−) and H
2S
n to understand redox homeostasis in the mitochondria
[43]. Both O
2•− and H
2S
n are short-lived reactive species, and their concentrations change quickly. To solve this problem, they developed a cyanine-based NIR probe, HCy-FN. This probe consists of two different reaction sites: one for the abstraction of hydrogen to detect O
2•− and the other for the detection of H
2S
n using nitrofluorobenzoate (
Scheme 1c). Both sensing steps were monitored by two different channels. Upon reacting with O
2•−, HCy-FN was oxidized to Cy-FN, and this transformation was monitored by an increase in the emission intensity from channel 1 at 794 nm (λ
ex = 750 nm). Next, the nitrofluorobenzoate part of Cy-FN reacted with H
2S
n to result in a decrease in the emission intensity of channel 1 followed by an increase in the emission intensity in channel 2 at 625 nm (λ
ex = 535 nm) due to the formation of Keto-Cy. They examined different ROS with HCy-FN and found that only O
2•− was able to oxidize the probe. Similarly, the reactivity of other RSS towards Cy-FN was also evaluated, and no changes in emission spectra were noted. HCy-FN was used for the detection of exogenous and endogenous H
2S
n with the macrophage cell line RAW264.7 to monitor both sensing steps by dual channel emission. It was found that the Pearson correlation coefficient (R
r) of Cy-FN and mitochondria-localizing Rhodamine 123 was 0.98, confirming that Cy-FN was localized in the mitochondria. Moreover, HCy-FN could detect endogenously produced O
2•−/H
2S
n in BALB/c mice. This work represents an interesting way to detect O
2•−/H
2S
n in the biological system. However, the claim that the probe is capable of monitoring mitochondrial O
2•−/H
2S
n may not be accurate. The authors only provided the R
r value for the intermediate compound Cy-FN and not for the actual probe. The structure of HCy-FN suggests that it may not be a suitable candidate to target the mitochondria because of the lack of a lipophilic cationic moiety.
In 2016, Chen and coworkers developed a probe (HCy-Mito) for the selective detection of superoxide anion (O
2•−) and H
2S
n in the mitochondria
[44]. The reaction sites for O
2•− and H
2S
n were the reduced cyanine dye (similar to HCy-FN) and
m-nitrophenyl ether (
Scheme 1d). In the presence of O
2•−, Hcy-Mito was oxidized to form a cyanine derivative, and the reduced nature of H
2S
n converted the nitro group to -NH
2, which terminated the d-PET process and resulted in an increase in emission intensity to 780 nm. The detection limits for O
2•− and H
2S
n by HCy-Mito were found to be 0.1 μM and 0.2 μM, respectively. In vitro experiments with RAW264.7 cells by HCy-Mito suggest that it could image exogenous and endogenous O
2•−/H
2S
n and localize specifically in the mitochondria (R
r = 0.93). This probe was further utilized for the in vivo detection of O
2•− (generated from phorbol myristate acetate (PMA) and H
2S
n (via injected Na
2S
4) in BALB/c mice.
In 2019, Meng et al., used a different reaction site based on the 2-(acylthio)benzoate for the design of a mitochondria-targeted probe for H
2S
n [45]. This template utilized both the nucleophilic and electrophilic nature of H
2S
n for its recognition (
Scheme 2a)
[46]. Briefly, the thioester exchange between the H
2S
n and 2-(acylthio)benzoate produced a thiophenol derivative, which, in turn, reacted with H
2S
n to form an -SSH intermediate. This intermediate underwent an intramolecular cyclization to release the fluorophore. This template was attached to a red-emitting fluorophore to develop the probe, HQO-PSP. HQO-PSP itself was non-fluorescent but, upon sensing H
2S
n, exhibited a fluorescence turn-on (86-fold) at an emission of 633 nm due to the formation of the keto derivative HQO. The probe was found to be relatively fast (7 min) and highly selective to H
2S
n, with a detection limit of 95.2 nM. In vitro studies with A549 cells revealed that HQO-PSP could specifically localize within the mitochondria (R
r = 0.98) and selectively image exogenously added H
2S
n in the live cells.
Scheme 2. Structures and reactions of probes (a) HQO-PSP, (b) SPS-M1, and (c) H1.
In the same year, Choi et al. reported that the ratiometric probe SPS-M1 for mitochondrial H
2S
n detection was based on a two-photon excitable naphthalene fluorophore
[47]. The reaction site was the same as that of HQO-PSP except for an additional self-immolating carbamate linker (S
cheme 2b). The probe exhibited a blue fluorescence (λ
em = 429 nm) but produced the deprotected yellow fluorescent dye M1 (λ
em = 506 nm) upon sensing H
2S
n. Interestingly, the two-photon absorption (TPA) cross-section (δ) of SPS-M1 and M1 was found to be 11 and 108 GM, respectively, at 750 nm. The large TPA cross-section resulted from the strong ICT process in M1. SPS-M1 was found to be suitable for the quantification of H
2S
n in live cells, and the in vitro detection limit was 1 μM. The two-photon microscopic imaging with SPS-M1 for endogenous H
2S
n using the wild-type and Parkinson’s disease (PD) model neurons and brain tissues of mice revealed that H
2S
n concentrations were higher in the PD model.
Another interesting approach for the spatiotemporal detection of mitochondrial H
2S
n was reported by Han et al. in 2018
[48]. Probe H1 consisted of a fluorescein dye attached to a triphenylphosphonium group and a nitrobenzyl photoactivable protecting group (
Scheme 2c). Upon irradiation with UV light (365 nm), the nitrobenzyl part produced an aldehyde derivative, which served as the H
2S
n recognition site. H
2S
n attacks the aldehyde group to form a persulfide intermediate, which then should undergo cyclization to liberate the fluorophore and generate a side product (4-hydroxybenzo[
d][1,2]dithiin-1(4
H)-one). H1 showed a turn-on of fluorescence at 525 nm only when it was irradiated with UV light along with H
2S
2 in the solutions. The detection limit was calculated to be 150 nM. The targeting ability of H1 was confirmed by counterstaining with MitoTracker Green (MTG) (R
r = 0.72). Although this photo-triggered probe was interesting, the authors did not provide experimental support for the proposed detection mechanism. This aldehyde-based intermediate may also possess some problems as 2-formyl carboxylate is a well-known H
2S recognition site
[49][50], and the aldehyde group has a high reactivity towards free cysteine
[51][52][53].