Recent Progress in the Rational Design of Biothiol-Responsive Fluorescent Probes

Biothiols such as cysteine, homocysteine, and glutathione play significant roles in important biological activities, and their abnormal concentrations have been found to be closely associated with certain diseases, making their detection a critical task. To this end, fluorescent probes have become increasingly popular due to their numerous advantages, including easy handling, desirable spatiotemporal resolution, high sensitivity, fast response, and favorable biocompatibility. As a result, intensive research has been conducted to create fluorescent probes for the detection and imaging of biothiols. This brief review summarizes recent advances in the field of biothiol-responsive fluorescent probes, with an emphasis on rational probe design, including the reaction mechanism, discriminating detection, reversible detection, and specific detection. Furthermore, the challenges and prospects of fluorescence probes for biothiols are also outlined.


Introduction
Biological thiols, known as biothiols, are essential in a variety of important physiological and pathological processes in living cells and organisms [1,2]. Glutathione (GSH), cysteine (Cys), and homocysteine (Hcy) are the most relevant biothiols, and they have similar structures that can transform into each other within cells. GSH is the most abundant non-protein thiol in cells, ranging from 1 to 10 mM, and is closely related to cellular redox homeostasis, heterogeneous metabolism, signal transduction, and gene regulation. Abnormal expression of GSH is associated with various disorders, including Parkinson's disease, Alzheimer's disease, liver injury, and cancer [3][4][5]. Cys, with a concentration ranging from 30 µM to 200 µM in cells, is a precursor of GSH and a strong binding site for transition metals such as copper, lead, cadmium, and mercury. Cys plays important roles in protein synthesis, neuroprotection, and redox homeostasis. Overexpression of Cys would lead to cardiovascular and nervous system diseases, whereas Cys deficiency might result in liver and kidney injury, growth retardation, and other diseases [6][7][8]. Hcy has the lowest general concentration among these three biothiols (5-12 µM), and abnormal levels of Hcy could lead to neurological and cardiovascular diseases [9][10][11]. Therefore, developing reliable and effective methods for real-time monitoring of cellular biothiols is of great significance in order to provide crucial information for the investigation of their functions, thereby improving early diagnosis and therapeutic interventions for related diseases [12,13].
Capillary electrophoresis, electrochemical assays, high-performance liquid chromatography (HPLC), and mass spectrometry have all been described as detection methods for

Reaction Mechanisms of Biothiol-Responsive Fluorescent Probes
Fluorescence probes are typically designed by modifying fluorophores with reactive motifs that allow them to fluorescently respond to analytes of interest through several signal transduction mechanisms. These mechanisms include photoinduced electron transfer (PET), twisted intramolecular charge transfer (TICT), through bond energy transfer For example, taking advantage of 2,7-Naphthyridine derivatives with large Stokes shifts, high quantum yields, and good photo-stabilities, the She group used a 2,7-Naphthyridine derivative as a fluorophore to develop a thiolysis-based fluorescent probe 1 (AND-DNBS) for biothiol detection, with the dinitrobenzene sulfonyl (DNBS) moiety serving as the sensing site [37]. As depicted in Figure 2, probe 1 responded to biothiols in a turn-on manner: DNBS was broken off from probe 1 by a thiolysis reaction to remove the PET effect, resulting in the formation of AND-OH, with favorable performances including fast response time (45 s), desirable fluorescence enhancement (100 fold), and excellent Stokes shift (227 nm). Probe 1 showed a high quantum yield (0.36). With favorable membrane permeability and low cytotoxicity, probe 1 was successfully used to visualize endogenous and exogenous GSH in living A549 cells and zebrafish. For example, taking advantage of 2,7-Naphthyridine derivatives with large Stokes shifts, high quantum yields, and good photo-stabilities, the She group used a 2,7-Naphthyridine derivative as a fluorophore to develop a thiolysis-based fluorescent probe 1 (AND-DNBS) for biothiol detection, with the dinitrobenzene sulfonyl (DNBS) moiety serving as the sensing site [37]. As depicted in Figure 2, probe 1 responded to biothiols in a turn-on manner: DNBS was broken off from probe 1 by a thiolysis reaction to remove the PET effect, resulting in the formation of AND-OH, with favorable performances including fast response time (45 s), desirable fluorescence enhancement (100 fold), and excellent Stokes shift (227 nm). Probe 1 showed a high quantum yield (0.36). With favorable membrane permeability and low cytotoxicity, probe 1 was successfully used to visualize endogenous and exogenous GSH in living A549 cells and zebrafish.
serving as the sensing site [37]. As depicted in Figure 2, probe 1 responded to biothiols in a turn-on manner: DNBS was broken off from probe 1 by a thiolysis reaction to remove the PET effect, resulting in the formation of AND-OH, with favorable performances including fast response time (45 s), desirable fluorescence enhancement (100 fold), and excellent Stokes shift (227 nm). Probe 1 showed a high quantum yield (0.36). With favorable membrane permeability and low cytotoxicity, probe 1 was successfully used to visualize endogenous and exogenous GSH in living A549 cells and zebrafish.  Traditional organic dyes usually encounter aggregation-caused quenching (ACQ) in biological systems due to their poor water solubility, which hinders their applicability in fluorescence imaging. To address this issue, Tang proposed the concept of aggregationinduced emission (AIE) for the first time [38]. AIE dyes exhibit intensive fluorescence after aggregating in high-concentration aqueous solutions but display negligible emission in lowconcentration solutions. Furthermore, aggregation in a biological system can even avoid the dispersion of free dyes, improve the photostability of dyes, and promote the signalto-noise ratio and sensitivity of detection. Thus far, AIE-based fluorescent probes have drawn considerable research interest in the last 20 years [39,40]. The Wang group reported a thiol-responsive fluorescence probe 2 (DNBS-HCA) using an AIEgen hydroxychalcone (HCA) as the reporting unit [41]. The hydroxyl group of HCA was caged with the DNBS moiety to block the ESIPT process, thus quenching the fluorescence. After the thiolysis reaction of probe 2 by biothiols, the recovery HCA exhibited remarkable AIE and ESIPT characteristics with long-wavelength red fluorescence emission at 600 nm and a favorable Stokes shift of 140 nm (Scheme 2). Probe 2 was then applied to develop indicator papers for biothiol detection, and the study demonstrated that the indicator papers could afford desirable feasibility and sensitivity for the quantitative detection of biothiols (0.1-1 nM). Finally, probe 2 was used to visualize the variation of cellular biothiols, and the red emission got weaker after NEM treatment, indicating the ability to detect biothiols in vivo.
Traditional organic dyes usually encounter aggregation-caused quenching (ACQ) in biological systems due to their poor water solubility, which hinders their applicability in fluorescence imaging. To address this issue, Tang proposed the concept of aggregationinduced emission (AIE) for the first time [38]. AIE dyes exhibit intensive fluorescence after aggregating in high-concentration aqueous solutions but display negligible emission in low-concentration solutions. Furthermore, aggregation in a biological system can even avoid the dispersion of free dyes, improve the photostability of dyes, and promote the signal-to-noise ratio and sensitivity of detection. Thus far, AIE-based fluorescent probes have drawn considerable research interest in the last 20 years [39,40]. The Wang group reported a thiol-responsive fluorescence probe 2 (DNBS-HCA) using an AIEgen hydroxychalcone (HCA) as the reporting unit [41]. The hydroxyl group of HCA was caged with the DNBS moiety to block the ESIPT process, thus quenching the fluorescence. After the thiolysis reaction of probe 2 by biothiols, the recovery HCA exhibited remarkable AIE and ESIPT characteristics with long-wavelength red fluorescence emission at 600 nm and a favorable Stokes shift of 140 nm (Scheme 2). Probe 2 was then applied to develop indicator papers for biothiol detection, and the study demonstrated that the indicator papers could afford desirable feasibility and sensitivity for the quantitative detection of biothiols (0.1-1 nM). Finally, probe 2 was used to visualize the variation of cellular biothiols, and the red emission got weaker after NEM treatment, indicating the ability to detect biothiols in vivo.   [42]. Probe 3 exhibited almost no fluorescence due to the intramolecular twisted internal charge transfer (TICT) and shielding effect of the DNBS group [30]. After the reaction with sulfhydryl groups, probe 3 could release the quencher to generate the dicyanoisophorone dye M-OH-OH, resulting in elevated emission in the yellow channel (575 nm) under 400 nm excitation in the presence of GSH. Intriguingly, the thiolysis product M-OH-OH can further embed into the hydrophobic cavity of human serum albumin (HSA) to produce an augment of redshifted fluorescence in the red channel (660 nm) under excitation at 500 nm due to the blocked TICT ( Figure 3A,B). Because of its dualchannel spectral response capability, probe 3 was utilized to simultaneously monitor variations of drug-induced HSA and endogenous/exogenous GSH in HepG2 cells by fluorescent imaging.

Biothiol Fluorescent Probes Based on Michael Addition Reactions
It is well known that thiols can undergo Michael addition reactions with various unsaturated groups such as aldehydes, ketones, maleimides, squaraines, and malonitriles. A variety of outstanding Michael acceptors were involved in the construction of fluorogenic probes for biothiol sensing. The Chen group, for instance, reported a rhodol-derived probe 4 (FRB-T, Scheme 3) for detecting biothiols in living cells and labeling sulfhydryl-containing proteins [43]. A 2-cyclopentenone group was decorated onto rhodol-derived as a responsive site, allowing the 1,4-Michael addition between the α,β-unsaturated ketone and

Biothiol Fluorescent Probes Based on Michael Addition Reactions
It is well known that thiols can undergo Michael addition reactions with various unsaturated groups such as aldehydes, ketones, maleimides, squaraines, and malonitriles. A variety of outstanding Michael acceptors were involved in the construction of fluorogenic probes for biothiol sensing. The Chen group, for instance, reported a rhodol-derived probe 4 (FRB-T, Scheme 3) for detecting biothiols in living cells and labeling sulfhydrylcontaining proteins [43]. A 2-cyclopentenone group was decorated onto rhodol-derived as a responsive site, allowing the 1,4-Michael addition between the α,β-unsaturated ketone and free sulfhydryl, and the subsequent ring-opening reaction led to the formation of rhodol in ketone form, resulting in an enhanced fluorescence at 557 nm. In vitro tests revealed that probe 4 was extremely sensitive and selective for biothiols. Furthermore, biothiols in HeLa cells were visualized by probe 4, and fluorescence was enhanced after co-incubation with Cys or GSH, whereas attenuated fluorescence was observed with N-Ethylmaleimide (NEM) as a scavenger for biothiols. Finally, probe 4 was also utilized to label sulfhydryl proteins, giving different fluorescence signals between sulfhydryl proteins and non-sulfhydryl proteins. These results indicated that probe 4 could be a robust tool to track biothiols in vivo and label sulfhydryl-containing proteins.
Molecules 2023, 28, x FOR PEER REVIEW 7 of 41 Scheme 3. Chemical structure and reaction mechanism of probe 4 via the Michael addition reaction.
By connecting the chromene motif and methylene blue with a carbamate linker, the Yin group created a Michael addition-based fluorescent probe 5 (Scheme 4) [44]. Scheme 4 shows that methylene blue could be released in the presence of biothiols via a cascade reaction including 1,4-Michael addition on the α,β-unsaturated ketone, ring opening of pyran, and self-immolation of carbamate, resulting in a considerable increase in the fluorescence at 698 nm under illumination at 600 nm. Owing to its long-wavelength absorption and emission, probe 5 was expected to possess advantages such as deep tissue penetration, minimal phototoxicity, and a high signal-to-noise ratio (S/N). Probe 5 demonstrated a fast response (~150 s), long-wavelength emission, and high chemical stability under physiological circumstances. Probe 5 with minimal cytotoxicity was also successful in identifying biothiols in living cells and mouse models, with considerably higher levels of biothiols observed in drug-resistant cancer cells compared to the corresponding cancer cells, indicating that biothiols might facilitate chemotherapy resistance in cancer cells. Scheme 4. Chemical structure and reaction mechanism of probe 5 via the Michael addition reaction.

Biothiol Fluorescent Probes Based on Reduction Reactions
The disulfide bond (-S-S-) could be reduced by thiols into two sulfhydryl groups. Thus, disulfide cleavage reactions were also utilized in the design of biothiol-responsive fluorescent probes. Wang et al. designed and synthesized a novel disulfide reductionbased ratiometric fluorescent probe 6 (RB-GSH) for biothiol imaging (Scheme 5) [45]. Probe 6 was constructed on a Förster resonance energy transfer (FRET) scaffold using oxanthrene as the acceptor and BODIPY as the donor, with the disulfide linker between Scheme 3. Chemical structure and reaction mechanism of probe 4 via the Michael addition reaction.
By connecting the chromene motif and methylene blue with a carbamate linker, the Yin group created a Michael addition-based fluorescent probe 5 (Scheme 4) [44]. Scheme 4 shows that methylene blue could be released in the presence of biothiols via a cascade reaction including 1,4-Michael addition on the α,β-unsaturated ketone, ring opening of pyran, and self-immolation of carbamate, resulting in a considerable increase in the fluorescence at 698 nm under illumination at 600 nm. Owing to its long-wavelength absorption and emission, probe 5 was expected to possess advantages such as deep tissue penetration, minimal phototoxicity, and a high signal-to-noise ratio (S/N). Probe 5 demonstrated a fast response (~150 s), long-wavelength emission, and high chemical stability under physiological circumstances. Probe 5 with minimal cytotoxicity was also successful in identifying biothiols in living cells and mouse models, with considerably higher levels of biothiols observed in drug-resistant cancer cells compared to the corresponding cancer cells, indicating that biothiols might facilitate chemotherapy resistance in cancer cells.
Molecules 2023, 28, x FOR PEER REVIEW 7 of 41 Scheme 3. Chemical structure and reaction mechanism of probe 4 via the Michael addition reaction.
By connecting the chromene motif and methylene blue with a carbamate linker, the Yin group created a Michael addition-based fluorescent probe 5 (Scheme 4) [44]. Scheme 4 shows that methylene blue could be released in the presence of biothiols via a cascade reaction including 1,4-Michael addition on the α,β-unsaturated ketone, ring opening of pyran, and self-immolation of carbamate, resulting in a considerable increase in the fluorescence at 698 nm under illumination at 600 nm. Owing to its long-wavelength absorption and emission, probe 5 was expected to possess advantages such as deep tissue penetration, minimal phototoxicity, and a high signal-to-noise ratio (S/N). Probe 5 demonstrated a fast response (~150 s), long-wavelength emission, and high chemical stability under physiological circumstances. Probe 5 with minimal cytotoxicity was also successful in identifying biothiols in living cells and mouse models, with considerably higher levels of biothiols observed in drug-resistant cancer cells compared to the corresponding cancer cells, indicating that biothiols might facilitate chemotherapy resistance in cancer cells. Scheme 4. Chemical structure and reaction mechanism of probe 5 via the Michael addition reaction.

Biothiol Fluorescent Probes Based on Reduction Reactions
The disulfide bond (-S-S-) could be reduced by thiols into two sulfhydryl groups. Thus, disulfide cleavage reactions were also utilized in the design of biothiol-responsive fluorescent probes. Wang et al. designed and synthesized a novel disulfide reductionbased ratiometric fluorescent probe 6 (RB-GSH) for biothiol imaging (Scheme 5) [45]. Probe 6 was constructed on a Förster resonance energy transfer (FRET) scaffold using oxanthrene as the acceptor and BODIPY as the donor, with the disulfide linker between the energy dyad serving as the reaction site. Upon excitation with the absorption of BOD-Scheme 4. Chemical structure and reaction mechanism of probe 5 via the Michael addition reaction.

Biothiol Fluorescent Probes Based on Reduction Reactions
The disulfide bond (-S-S-) could be reduced by thiols into two sulfhydryl groups. Thus, disulfide cleavage reactions were also utilized in the design of biothiol-responsive fluorescent probes. Wang et al. designed and synthesized a novel disulfide reduction-based ratiometric fluorescent probe 6 (RB-GSH) for biothiol imaging (Scheme 5) [45]. Probe 6 Molecules 2023, 28, 4252 7 of 38 was constructed on a Förster resonance energy transfer (FRET) scaffold using oxanthrene as the acceptor and BODIPY as the donor, with the disulfide linker between the energy dyad serving as the reaction site. Upon excitation with the absorption of BODIPY at 480 nm, a near-infrared (NIR) fluorescent emission from oxanthrene at 656 nm was identified in the fluorescence spectra of probe 6 due to the resonance energy transfer between donor and acceptor, whereas the fluorescence emission from the BODIPY donor at 511 nm was relatively faint. In the presence of reductive biothiols, the disulfide bond was cleaved by a reduction reaction to effectively inhibit the FRET process, thus leading to a gradual increase at 511 nm and a slow decrease at 656 nm. The fluorescence ratio (I 512nm /I 656nm ) displayed an excellent linear relationship with GSH concentration from 10 µM to 100 µM. Therefore, this fluorescent probe 6 could determine the biothiols based on the fluorescent signals from two different wavelengths. In contrast to turn-on fluorescent probes, which rely on changes in fluorescence intensity to determine biothiols, ratiometric fluorescent probes employ two signals that act as internal references to each other during the detection process. This approach can overcome potential disturbances caused by factors such as microenvironment (pH, polarity, viscosity, and so on), distribution and photobleaching of probes, and intensity of light sources, thereby enabling probe 6 to accurately detect intracellular GSH in a ratiometric manner. probes, which rely on changes in fluorescence intensity to determine biothiols, ratiometric fluorescent probes employ two signals that act as internal references to each other during the detection process. This approach can overcome potential disturbances caused by factors such as microenvironment (pH, polarity, viscosity, and so on), distribution and photobleaching of probes, and intensity of light sources, thereby enabling probe 6 to accurately detect intracellular GSH in a ratiometric manner.

Scheme 5.
Chemical structure and response mechanism of probe 6 via the disulfide reduction reaction.

Biothiol Fluorescent Probes Based on SNAr Reactions
In addition to thiolysis reactions, Michael addition reactions, and disul ide reduction reactions, aromatic nucleophilic substitution reactions (SNAr) were another powerful tool to develop biothiol probes based on their nucleophilicity [46]. Generally, the SNAr-based probes undergo two successive reactions in the response processes, including an intermolecular nucleophilic substitution and an intramolecular nucleophilic substitution, enabling the specific detection of GSH over Hcy/Cys, which we will discuss in the sections below.

Design Strategies for Specific Fluorescent Probes for Intracellular Biothiols
Traditional thiol-responsive fluorescent probes have difficulty distinguishing different biothiols from each other due to their structural similarities. So far, smart fluorescent probes capable of selectively detecting and imaging GSH or Cys/Hcy have attracted intensive research interest to obtain more detailed information for studies on physiological processes and disorders related to one specific biothiol. Despite the fact that it is a significant challenge, fortunately, some subtle structural differences amongst biothiols offer the possibility of achieving specific fluorescent sensing of biothiols. Several sophisticated complex reactions have been proposed to design fluorescent probes for real-time bioimaging of particular thiols in living cells. In this section, we will provide a brief introduction to common strategies for designing specialized fluorescent probes for intracellular biothiols. Scheme 5. Chemical structure and response mechanism of probe 6 via the disulfide reduction reaction.

Biothiol Fluorescent Probes Based on SNAr Reactions
In addition to thiolysis reactions, Michael addition reactions, and disulfide reduction reactions, aromatic nucleophilic substitution reactions (SNAr) were another powerful tool to develop biothiol probes based on their nucleophilicity [46]. Generally, the SNAr-based probes undergo two successive reactions in the response processes, including an intermolecular nucleophilic substitution and an intramolecular nucleophilic substitution, enabling the specific detection of GSH over Hcy/Cys, which we will discuss in the sections below.

Design Strategies for Specific Fluorescent Probes for Intracellular Biothiols
Traditional thiol-responsive fluorescent probes have difficulty distinguishing different biothiols from each other due to their structural similarities. So far, smart fluorescent probes capable of selectively detecting and imaging GSH or Cys/Hcy have attracted intensive research interest to obtain more detailed information for studies on physiological processes and disorders related to one specific biothiol. Despite the fact that it is a significant challenge, fortunately, some subtle structural differences amongst biothiols offer the possibility of achieving specific fluorescent sensing of biothiols. Several sophisticated complex reactions have been proposed to design fluorescent probes for real-time bioimaging of particular thiols in living cells. In this section, we will provide a brief introduction to common strategies for designing specialized fluorescent probes for intracellular biothiols.

Specific Fluorescent Probes for Intracellular Cys
Acrylates are widely used as reactive sites for the specific detection of cysteines. As shown in Scheme 6, an electron-deficient acrylate can quench the fluorescence of a free probe, and in the presence of GSH, a thioether can be obtained by Michael addition but without any emission. On the contrary, Cys and Hcy could induce further intramolecular nucleophilic substitution reactions between the primary amidogen and the carbonyl, leading to the release of light fluorophores. Furthermore, the formation of a seven-membered ring structure is kinetically more favored by far, allowing for specific detection of Cys. A large number of acrylate-based probes have been developed so far to monitor intracellular Cys based on various fluorogens in visible and near-infrared regions or with AIE and two-photon characteristics (Scheme 7).

Specific Fluorescent Probes for Intracellular Cys
Acrylates are widely used as reactive sites for the specific detection of cysteines. As shown in Scheme 6, an electron-deficient acrylate can quench the fluorescence of a free probe, and in the presence of GSH, a thioether can be obtained by Michael addition but without any emission. On the contrary, Cys and Hcy could induce further intramolecular nucleophilic substitution reactions between the primary amidogen and the carbonyl, leading to the release of light fluorophores. Furthermore, the formation of a seven-membered ring structure is kinetically more favored by far, allowing for specific detection of Cys. A large number of acrylate-based probes have been developed so far to monitor intracellular Cys based on various fluorogens in visible and near-infrared regions or with AIE and twophoton characteristics (Scheme 7). Scheme 6. Mechanism of acrylate-based probes specifically responding to Cys. The Chen group reported an acrylate-based fluorescence probe 7 (DPAS-Cys) to detect Cys in lipid droplets using an AIEgen DPAS [47]. Probe 7 fluoresced at 560 nm when stimulated at 360 nm in the presence of Cys, with a low detection limit of 2.4 µM. Due to its high sensitivity and selectivity for Cys, as well as its superior biocompatibility, probe 7 was effectively employed for fluorescence imaging of endogenous and exogenous Cys in living cells. Further colocalization experiments proved that probe 7 was capable of tracking the Cys in lipid droplets, which was attributable to its high hydrophobicity. The Liu group designed and synthesized a novel fluorescent probe 8 by incorporating two acrylate groups into the flavonoid moiety [48]. Most interestingly, probe 8 exhibited distinct responses to different concentration ranges of Cys. The emission at 509 nm increased within a low concentration range of Cys (0~10 mM, 0~1 equiv) and reached a plateau at a high concentration (>10 mM), while another increased fluorescence at 449 nm was observed in the high Cys concentration range (0~40 mM, 0~4 equiv). The intensity ratio (I449/I509) increased with the enhancement of Cys concentration with good linearity, indicating a potential to detect Cys by ratiometry. The probe with low cytotoxicity was finally used to differentially visualize Cys at different concentrations in living cells. The Liu group synthesized an acrylate-based ratiometric fluorescent probe 9 (TIFC) by utilizing two identical isophorone-malononitrile structures as the fluorophore [49], whose ratio of fluorescence intensity (I568/I540) was linearly related to the Cys concentrations in the range 0-300 µM, covering the range of the physiological Cys level, with a detection limit of 105.6 nM. With desirable stability and biocompatibility under physiological conditions, probe 9 was successfully applied to measure the endogenous and exogenous Cys in osteoblasts by ratiometric fluorescence imaging. However, the relatively short excitation and emission bands of the aforementioned probes are always implicated in poor penetration depth and large background noise, which are not favorable for tissue and in vivo bioimaging. Accordingly, an acrylate-decorated anthocyanidin derivative served as a red-emitting fluorescent probe 10 (HAS) to selectively detect Cys [50]. Probe 10 responded to Cys in a turn-on manner with an emission at 615 nm when excited at 580 nm, and the fluorescence intensity displayed a remarkable linear connection with Cys concentration. Due to its high membrane permeability and low cytotoxicity, probe 10 was successfully used for fluorescence imaging of Cys in live cells. An electron-withdrawing malononitrile group The Chen group reported an acrylate-based fluorescence probe 7 (DPAS-Cys) to detect Cys in lipid droplets using an AIEgen DPAS [47]. Probe 7 fluoresced at 560 nm when stimu-lated at 360 nm in the presence of Cys, with a low detection limit of 2.4 µM. Due to its high sensitivity and selectivity for Cys, as well as its superior biocompatibility, probe 7 was effectively employed for fluorescence imaging of endogenous and exogenous Cys in living cells. Further colocalization experiments proved that probe 7 was capable of tracking the Cys in lipid droplets, which was attributable to its high hydrophobicity. The Liu group designed and synthesized a novel fluorescent probe 8 by incorporating two acrylate groups into the flavonoid moiety [48]. Most interestingly, probe 8 exhibited distinct responses to different concentration ranges of Cys. The emission at 509 nm increased within a low concentration range of Cys (0~10 mM, 0~1 equiv) and reached a plateau at a high concentration (>10 mM), while another increased fluorescence at 449 nm was observed in the high Cys concentration range (0~40 mM, 0~4 equiv). The intensity ratio (I 449 /I 509 ) increased with the enhancement of Cys concentration with good linearity, indicating a potential to detect Cys by ratiometry. The probe with low cytotoxicity was finally used to differentially visualize Cys at different concentrations in living cells. The Liu group synthesized an acrylate-based ratiometric fluorescent probe 9 (TIFC) by utilizing two identical isophorone-malononitrile structures as the fluorophore [49], whose ratio of fluorescence intensity (I 568 /I 540 ) was linearly related to the Cys concentrations in the range 0-300 µM, covering the range of the physiological Cys level, with a detection limit of 105.6 nM. With desirable stability and biocompatibility under physiological conditions, probe 9 was successfully applied to measure the endogenous and exogenous Cys in osteoblasts by ratiometric fluorescence imaging. However, the relatively short excitation and emission bands of the aforementioned probes are always implicated in poor penetration depth and large background noise, which are not favorable for tissue and in vivo bioimaging.
Accordingly, an acrylate-decorated anthocyanidin derivative served as a red-emitting fluorescent probe 10 (HAS) to selectively detect Cys [50]. Probe 10 responded to Cys in a turn-on manner with an emission at 615 nm when excited at 580 nm, and the fluorescence intensity displayed a remarkable linear connection with Cys concentration. Due to its high membrane permeability and low cytotoxicity, probe 10 was successfully used for fluorescence imaging of Cys in live cells. An electron-withdrawing malononitrile group might chromatically shift excitation/emission wavelengths of dye into the red-emitting region by extending the conjugation and boosting the ICT effect of the D-A structure. Inspired by this, the Jiang group designed a deep red-emitting fluorescent probe 11 (DRP-Cys) based on a malononitrile-modified xanthene derivative with acrylate as the Cys recognition site [51]. After being excited at 580 nm, probe 11 displayed a turn-on response to Cys with a 30-fold increase in the fluorescence at 645 nm. Probe 11 obtained a fast response time of about 1 min and a low limit of detection of 10 nM in the Cys range of 0-200 µM, suggesting favorable sensitivity for Cys. Because of its high selectivity, biocompatibility, and membrane permeability, probe 11 was successfully used for fluorescent imaging of Cys in living cells and mouse models. Another strong electron-withdrawing group, 2-dicyanomethylene-3-cyano-4,5,5-trimethyl-2,5-dihydrofuran (TCF), had been reported to produce red-emitting D-A-type fluorophores. Therefore, the Zhu group developed a TCF-based long-range measuring probe 12 (TCFQ-Cys) for Cys detection, with acrylate serving as the response group [52]. The free probe 12 exhibited a turn-on response to Cys with a red emission at 633 nm; more excitingly, the emission intensity was linear to Cys concentrations throughout the wide range of 0-300 µM, covering the range of the physiological Cys level (30-200 µM), and the limit of detection was determined at 0.133 µM. Due to its high sensitivity and selectivity for Cys and favorable compatibility, probe 12 was successfully engaged in fluorescence imaging for endogenous and exogenous Cys in cells and Caenorhabditis elegans.
Because of its low background fluorescence, minimal tissue absorption, and deep tissue penetration, near-infrared (NIR) fluorescence with a wavelength in the range of 650-1000 nm is widely used in imaging in vivo. The Qiu group created a novel acrylatebased NIR fluorescent probe 13 (Cys-TCF) for selective in vivo bioimaging of Cys using a TCF-tailed NIR fluorophore [53]. Excited at 570 nm, the fluorescence intensity (607 nm) of probe 13 increased along with the enhancement of Cys concentration (0-200 µM) in favorable linearity, and the limit of detection was estimated at 0.04 µM. Owing to its NIRemitting characteristics, probe 13 proved successful in fluorescence imaging of Cys in live cells and zebrafish. By utilizing a (2-(2 -Hydroxyphenyl) benzothiazo (HBT) derivative integrated with a strong electron-withdrawing isophorone, the Yin group developed another acrylate-based NIR fluorescent probe 14 (SYP) for the real-time imaging of Cys in vivo [54]. Attributed to the ESIPT characteristic of HBT, probe 14 exhibited an excellent large Stokes shift of 263 nm with enhanced fluorescence at 686 nm in the presence of Cys, which was beneficial to avoiding background interference. Further optical experiments in buffer revealed that probe 14 was stable and appropriate for the selective detection of Cys at pH ranging from 5.0 to 9.0, with detection sensitivity at pH 7.4 being more sensitive and rapid than at pH 8.0. Due to its favorable sensitivity and biocompatibility, probe 14 was applied to monitor Cys in living cells and tumor mice by fluorescence imaging. Similarly, the Li group created a novel NIR fluorophore by hybridizing xanthene and isophorone and then caged it with an acrylate group to obtain a NIR fluorescent probe 15 (IX) for Cys [55]. In the presence of Cys, probe 15 displayed a remarkable increase in fluorescence at 770 nm with a large Stokes shift (180 nm), which is advantageous for biosensing and bioimaging. Probe 15 was successfully applied to the fluorescent imaging of Cys in living cells and HCT116xenograft tumor mice. Zeng et al. developed a novel near-infrared fluorescent probe 16 (BHcy-A) to visualize Cys via heteroatom substitution on the semi-heptamethine dye [56]. Probe 16 may preferentially respond to Cys with near-infrared (NIR) emission (742 nm), which is desirable for bioimaging due to its excellent penetration and high signal-to-noise ratio (S/N). The bioimaging results indicated that probe 16 could specifically accumulate in the lysosomes, revealing its potential to study the function of lysosomes and lysosomal Cys in living cells.
Two-photon fluorescent probes are also a research hotspot in bioimaging, benefiting from strong penetration, minimal cytotoxicity, reduced photobleaching, and high resolution. The Lin group reported a two-photon ratiometric fluorescent probe 17 (Nap-Cys) employing naphthalimide as the reporter [57]. Probe 17 specifically responded to Cys over Hcy and GSH; the emission intensity ratio between 550 nm and 440 nm (I 550 /I 440 ) is in good linearity with the concentration of Cys, with a moderate limit of detection at 1.80 µM in the range of 10-80 µM. The p-toluenesulfonamide serving as an endoplasmic reticulum (ER)-specific group allows probe 17 to detect endogenous and exogenous Cys levels in the ER by two-photon imaging for the first time. Given the advantages of long wavelengths, two-photon fluorescent probes emitting near-infrared light will be more suitable for biological in vivo imaging. The Zhang group developed a red-emitting probe 18 (ACP) to achieve the detection of Cys by two-photon fluorescence imaging [58]. Probe 18 demonstrated a fast response, significant intensity enhancement, and high selectivity for Cys over Hcy and GSH. Due to its long-wavelength emission (610 nm), two-photon fluorescence imaging of Cys was obtained in deep tissue up to 120 µm depth, and an increasing Cys was observed in the inflamed mouse model. The Zhou group reported a NIR-emitting two-photon fluorescent probe 19 (SDP-A) with an acrylate fraction as the ESIPT/ICT blocker and Cys response site [59]. In the presence of Cys, there was an increase in fluorescence intensity at 713 nm, with a noticeable Stokes shift of 302 nm due to the combined ESIPT and ICT. Probe 19 demonstrated desirable sensitivity and selectivity for Cys over Hcy and GSH, with a limit of detection of 102 nM. The fluorescent scaffold, most crucially, displayed a favorable two-photon absorption cross section (213.5 GM at 820 nm) and high quantum yields (1.52−18.17%), making it appropriate for two-photon fluorescence imaging of biological samples. By two-photon fluorescence imaging with probe 19, three-dimensional perspective images of the mouse brain were achieved, and finally, the abdominal cavity down to a depth of more than 200 µm was also observed.
To further improve the specificity for Cys, the Li group introduced furan and thiophene into the acrylate sensing group to build new Cys-responsive probes 20 and 21 (Scheme 8) [60]. The probes selectively respond to Cys by increasing their fluorescence at 523 nm. Both probes had exceptional fluorescence stability at physiological pH levels, and MTT assays confirmed their low cytotoxicity. Probes 20 and 21 were then applied for fluorescence imaging of endogenous and exogenous Cys in living cells. Finally, the Cys in aged rats chronically treated with APAP (N-acetyl-p-aminophenol, which consumes Cys by APAP metabolism) was visualized by fluorescence imaging using probe 21, and the results confirmed the permanent loss of Cys caused by the metabolism of APAP in the liver. It was also proven that Cys loss might aggravate sarcopenia and that the proper supplementation of Cys might be beneficial to the elderly under APAP treatment.
quantum yields (1.52−18.17%), making it appropriate for two-photon fluorescence imaging of biological samples. By two-photon fluorescence imaging with probe 19, three-dimensional perspective images of the mouse brain were achieved, and finally, the abdominal cavity down to a depth of more than 200 µm was also observed.
To further improve the specificity for Cys, the Li group introduced furan and thiophene into the acrylate sensing group to build new Cys-responsive probes 20 and 21 (Scheme 8) [60]. The probes selectively respond to Cys by increasing their fluorescence at 523 nm. Both probes had exceptional fluorescence stability at physiological pH levels, and MTT assays confirmed their low cytotoxicity. Probes 20 and 21 were then applied for fluorescence imaging of endogenous and exogenous Cys in living cells. Finally, the Cys in aged rats chronically treated with APAP (N-acetyl-p-aminophenol, which consumes Cys by APAP metabolism) was visualized by fluorescence imaging using probe 21, and the results confirmed the permanent loss of Cys caused by the metabolism of APAP in the liver. It was also proven that Cys loss might aggravate sarcopenia and that the proper supplementation of Cys might be beneficial to the elderly under APAP treatment. Mustafa Emrullahoğlu's group utilized propargylate as the reaction site to develop a novel fluorescent probe 22 (FL-PRP) [61]. The Cys could react with the propargylate by a cascade reaction involving nucleophilic addition and substitution to uncage the fluorescein dye, affording an increased emission band at 515 nm under excitation at 460 nm (Scheme 9). According to the titration experiment, the detection limit of probe 22 was calculated to be 182 nM. With favorable biocompatibility, probe 22 was successfully used for fluorescence imaging of endogenous and exogenous Cys in living cells. Cys/Hcy but not GSH can undergo a selective addition-cyclization reaction with aldehyde groups, leading to the formation of thiazolidines. Based on this reaction, many aldehyde-based specific fluorescent probes for Cys/Hcy have been developed. The Li group reported an unexpectedly specific dual-channel fluorescent probe 23 for Cys (Scheme 10) [62]. Within 15 min of adding Cys, an increased intensity was recorded at 440 nm, and the limit of detection was estimated at 63 nM. The detection of Cys was also shown to be stable and reliable due to its lack of interference from any other amino acids, including Hcy and GSH. However, it is worth noting that DMF, an organic solvent, was used in the detection system with a proportion as high as 50% in the solution, which is unsuitable for bioimaging. Finally, probe 23 was successfully applied to determine the Cys in human serum. Mustafa Emrullahoglu's group utilized propargylate as the reaction site to develop a novel fluorescent probe 22 (FL-PRP) [61]. The Cys could react with the propargylate by a cascade reaction involving nucleophilic addition and substitution to uncage the fluorescein dye, affording an increased emission band at 515 nm under excitation at 460 nm (Scheme 9). According to the titration experiment, the detection limit of probe 22 was calculated to be 182 nM. With favorable biocompatibility, probe 22 was successfully used for fluorescence imaging of endogenous and exogenous Cys in living cells.
quantum yields (1.52−18.17%), making it appropriate for two-photon fluorescence imaging of biological samples. By two-photon fluorescence imaging with probe 19, three-dimensional perspective images of the mouse brain were achieved, and finally, the abdominal cavity down to a depth of more than 200 µm was also observed.
To further improve the specificity for Cys, the Li group introduced furan and thiophene into the acrylate sensing group to build new Cys-responsive probes 20 and 21 (Scheme 8) [60]. The probes selectively respond to Cys by increasing their fluorescence at 523 nm. Both probes had exceptional fluorescence stability at physiological pH levels, and MTT assays confirmed their low cytotoxicity. Probes 20 and 21 were then applied for fluorescence imaging of endogenous and exogenous Cys in living cells. Finally, the Cys in aged rats chronically treated with APAP (N-acetyl-p-aminophenol, which consumes Cys by APAP metabolism) was visualized by fluorescence imaging using probe 21, and the results confirmed the permanent loss of Cys caused by the metabolism of APAP in the liver. It was also proven that Cys loss might aggravate sarcopenia and that the proper supplementation of Cys might be beneficial to the elderly under APAP treatment. Mustafa Emrullahoğlu's group utilized propargylate as the reaction site to develop a novel fluorescent probe 22 (FL-PRP) [61]. The Cys could react with the propargylate by a cascade reaction involving nucleophilic addition and substitution to uncage the fluorescein dye, affording an increased emission band at 515 nm under excitation at 460 nm (Scheme 9). According to the titration experiment, the detection limit of probe 22 was calculated to be 182 nM. With favorable biocompatibility, probe 22 was successfully used for fluorescence imaging of endogenous and exogenous Cys in living cells. Cys/Hcy but not GSH can undergo a selective addition-cyclization reaction with aldehyde groups, leading to the formation of thiazolidines. Based on this reaction, many aldehyde-based specific fluorescent probes for Cys/Hcy have been developed. The Li group reported an unexpectedly specific dual-channel fluorescent probe 23 for Cys (Scheme 10) [62]. Within 15 min of adding Cys, an increased intensity was recorded at 440 nm, and the limit of detection was estimated at 63 nM. The detection of Cys was also shown to be stable and reliable due to its lack of interference from any other amino acids, including Hcy and GSH. However, it is worth noting that DMF, an organic solvent, was used in the detection system with a proportion as high as 50% in the solution, which is unsuitable for bioimaging. Finally, probe 23 was successfully applied to determine the Cys in human serum. Scheme 9. Chemical structure and reaction mechanism of probe 22 with Cys.
Cys/Hcy but not GSH can undergo a selective addition-cyclization reaction with aldehyde groups, leading to the formation of thiazolidines. Based on this reaction, many aldehyde-based specific fluorescent probes for Cys/Hcy have been developed. The Li group reported an unexpectedly specific dual-channel fluorescent probe 23 for Cys (Scheme 10) [62]. Within 15 min of adding Cys, an increased intensity was recorded at 440 nm, and the limit of detection was estimated at 63 nM. The detection of Cys was also shown to be stable and reliable due to its lack of interference from any other amino acids, including Hcy and GSH. However, it is worth noting that DMF, an organic solvent, was used in the detection system with a proportion as high as 50% in the solution, which is unsuitable for bioimaging. Finally, probe 23 was successfully applied to determine the Cys in human serum.
Recently, the Yuan group developed a fluorescent probe 24 (SPI) based on a phenothiazinecyanine scaffold with a sulfamide moiety as a Cys-responsive group [63]. As depicted in Scheme 11, probe 24 can respond to Cys via thiolytic reaction to fluoresce at 538 nm with an excitation of 426 nm. Interestingly, the sulfur atom on the phenothiazine can be efficiently oxidized by ·OH into sulfoxide, forming a larger π-conjugation to exhibit red fluorescence (E x /E m = 485/608 nm). This dual-site probe 24 demonstrated desirable sensitivity and selectivity towards Cys/·OH over the other species, enabling quantitative detection and real-time imaging of ·OH and Cys in their own fluorescence channels. Finally, the Cys/·OH-regulated redox balance in mice and zebrafish was also monitored by fluorescent imaging with probe 24, indicating its potential application in the study of physiological and metabolic processes. Recently, the Yuan group developed a fluorescent probe 24 (SPI) based on a phenothiazine−cyanine scaffold with a sulfamide moiety as a Cys-responsive group [63]. As depicted in Scheme 11, probe 24 can respond to Cys via thiolytic reaction to fluoresce at 538 nm with an excitation of 426 nm. Interestingly, the sulfur atom on the phenothiazine can be efficiently oxidized by ·OH into sulfoxide, forming a larger π-conjugation to exhibit red fluorescence (Ex/Em = 485/608 nm). This dual-site probe 24 demonstrated desirable sensitivity and selectivity towards Cys/·OH over the other species, enabling quantitative detection and real-time imaging of ·OH and Cys in their own fluorescence channels. Finally, the Cys/·OH-regulated redox balance in mice and zebrafish was also monitored by fluorescent imaging with probe 24, indicating its potential application in the study of physiological and metabolic processes. Scheme 11. Chemical structure and reaction mechanism of probe 24 with Cys and ·OH.

Specific Fluorescent Probes for GSH
Sulfone and sulfoxide, which can be nucleophilically substituted by biothiols via the SNAr reaction mechanism, were commonly employed as reactive sites in the construction of fluorescence probes for GSH with high selectivity. For example, the Fang group recently developed a sulfuryl-based fluorescent probe 25 (R13) with a naphthalimide skeleton to quantify GSH concentration in living cells and tissues ( Figure 4A) [64]. Probe 25 exhibited negligible fluorescence because of the inhibited ICT effect. After adding GSH, a SNAr reaction led to the formation of Nap-SG, which exhibited increased fluorescence in the green channel (498 nm) ( Figure 4A,B). With excellent sensitivity and moderate selectivity toward GSH in physiological conditions, probe 25 was used to determine GSH levels in mouse livers after X-ray irradiation, revealing that irradiation-induced oxidative stress leads to the depletion of GSH. Encouraged by the results, which were in good agreement with those obtained by HPLC, quantification of GSH content in Parkinson's mouse brains was achieved by a straightforward fluorometric assay with probe 25. Recently, the Yuan group developed a fluorescent probe 24 (SPI) based on a phenothiazine−cyanine scaffold with a sulfamide moiety as a Cys-responsive group [63]. As depicted in Scheme 11, probe 24 can respond to Cys via thiolytic reaction to fluoresce at 538 nm with an excitation of 426 nm. Interestingly, the sulfur atom on the phenothiazine can be efficiently oxidized by ·OH into sulfoxide, forming a larger π-conjugation to exhibit red fluorescence (Ex/Em = 485/608 nm). This dual-site probe 24 demonstrated desirable sensitivity and selectivity towards Cys/·OH over the other species, enabling quantitative detection and real-time imaging of ·OH and Cys in their own fluorescence channels. Finally, the Cys/·OH-regulated redox balance in mice and zebrafish was also monitored by fluorescent imaging with probe 24, indicating its potential application in the study of physiological and metabolic processes. Scheme 11. Chemical structure and reaction mechanism of probe 24 with Cys and ·OH.

Specific Fluorescent Probes for GSH
Sulfone and sulfoxide, which can be nucleophilically substituted by biothiols via the SNAr reaction mechanism, were commonly employed as reactive sites in the construction of fluorescence probes for GSH with high selectivity. For example, the Fang group recently developed a sulfuryl-based fluorescent probe 25 (R13) with a naphthalimide skeleton to quantify GSH concentration in living cells and tissues ( Figure 4A) [64]. Probe 25 exhibited negligible fluorescence because of the inhibited ICT effect. After adding GSH, a SNAr reaction led to the formation of Nap-SG, which exhibited increased fluorescence in the green channel (498 nm) ( Figure 4A,B). With excellent sensitivity and moderate selectivity toward GSH in physiological conditions, probe 25 was used to determine GSH levels in mouse livers after X-ray irradiation, revealing that irradiation-induced oxidative stress leads to the depletion of GSH. Encouraged by the results, which were in good agreement with those obtained by HPLC, quantification of GSH content in Parkinson's mouse brains was achieved by a straightforward fluorometric assay with probe 25. Scheme 11. Chemical structure and reaction mechanism of probe 24 with Cys and ·OH.

Specific Fluorescent Probes for GSH
Sulfone and sulfoxide, which can be nucleophilically substituted by biothiols via the SNAr reaction mechanism, were commonly employed as reactive sites in the construction of fluorescence probes for GSH with high selectivity. For example, the Fang group recently developed a sulfuryl-based fluorescent probe 25 (R13) with a naphthalimide skeleton to quantify GSH concentration in living cells and tissues ( Figure 4A) [64]. Probe 25 exhibited negligible fluorescence because of the inhibited ICT effect. After adding GSH, a SNAr reaction led to the formation of Nap-SG, which exhibited increased fluorescence in the green channel (498 nm) ( Figure 4A,B). With excellent sensitivity and moderate selectivity toward GSH in physiological conditions, probe 25 was used to determine GSH levels in mouse livers after X-ray irradiation, revealing that irradiation-induced oxidative stress leads to the depletion of GSH. Encouraged by the results, which were in good agreement with those obtained by HPLC, quantification of GSH content in Parkinson's mouse brains was achieved by a straightforward fluorometric assay with probe 25.
The Liu group designed and synthesized a dual-response fluorescence probe 26 (Mito-NA-BP) for GSH and its metabolic product SO 2 based on a FRET scaffold, using chromene as the acceptor and naphthalimide as the donor ( Figure 5A) [65]. With 4-fluoro sulfoxide and chromene as the reactive units towards GSH and SO 2− 3 , Probe 26 could respond to GSH and SO 2− 3 in distinct patterns via two emission channels at 496 nm (green) and 638 nm (red) by modulating the FRET process ( Figure 5B,C). Owing to its favorable performance, including high sensitivity, selectivity, and low cytotoxicity, probe 26 was applied to track exogenous SO 2− 3 and endogenous Cys in real-time, monitor the metabolic process of GSH-SO 2 , and visualize the effect of high-dose SO 2 on GSH in living cells in real-time.   Figure 6A) [70]. Free probe 27 could not react with GSH and showed negligible fluorescence due to a PET process from diamine to BODIPY. In the presence of NO, the diamine group of BNS was transformed into a triazole derivative with enhanced fluorescence at 565 nm (yellow) by the inhibited PET process [29]. Furthermore, the electron-deficient triazole made the three-position in BODIPY electrophilic enough to react with GSH to form a thiol-BODIPY, resulting in a red-shifted fluorescence at 595 nm (red) ( Figure 6B). The yellow and red fluorescence intensities showed a good linear relationship with their corresponding analyte concentrations, with  Figure 6A) [70]. Free probe 27 could not react with GSH and showed negligible fluorescence due to a PET process from diamine to BODIPY. In the presence of NO, the diamine group of BNS was transformed into a triazole derivative with enhanced fluorescence at 565 nm (yellow) by the inhibited PET process [29]. Furthermore, the electron-deficient triazole made the three-position in BODIPY electrophilic enough to react with GSH to form a thiol-BODIPY, resulting in a red-shifted fluorescence at 595 nm (red) ( Figure 6B). The yellow and red fluorescence intensities showed a good linear relationship with their corresponding analyte concentrations, with favorable selectivity and sensitivity, respectively. Imaging of NO and GSH was originally performed in human umbilical vein endothelial cells (HUVECs), with variations in fluorescence intensities indicating that the production of GSH could be induced by exogenous NO in the living cells ( Figure 6C). The further imaging results consistently showed that either pravastatin or Vitamin C (VC) could increase the level of intracellular NO, which in turn induced the activation of γ-Glutamylcysteine synthetase (γ-GCS) and eventually led to the upregulation of intracellular GSH.
Boron-dipyrromethene (BODIPY) derivatives with reactive groups, such as DNBS, chlorine, 2,4-dinitrobenzenyl (DNB), and selenium, can also respond to biothiols by SNAr reactions [66][67][68][69]. Chen et al. used a chlorinated BODIPY to create a sequentially activated fluorescent probe 27 (BNS) for visualizing nitric oxide (NO)-induced GSH upregulation in living cells and in vivo ( Figure 6A) [70]. Free probe 27 could not react with GSH and showed negligible fluorescence due to a PET process from diamine to BODIPY. In the presence of NO, the diamine group of BNS was transformed into a triazole derivative with enhanced fluorescence at 565 nm (yellow) by the inhibited PET process [29]. Furthermore, the electron-deficient triazole made the three-position in BODIPY electrophilic enough to react with GSH to form a thiol-BODIPY, resulting in a red-shifted fluorescence at 595 nm (red) ( Figure 6B). The yellow and red fluorescence intensities showed a good linear relationship with their corresponding analyte concentrations, with favorable selectivity and sensitivity, respectively. Imaging of NO and GSH was originally performed in human umbilical vein endothelial cells (HUVECs), with variations in fluorescence intensities indicating that the production of GSH could be induced by exogenous NO in the living cells ( Figure 6C). The further imaging results consistently showed that either pravastatin or Vitamin C (VC) could increase the level of intracellular NO, which in turn induced the activation of γ-Glutamylcysteine synthetase (γ-GCS) and eventually led to the upregulation of intracellular GSH.   Figure 7A) [71]. Probe 28 exhibited absorption at 509 nm and emission at 650 nm (red channel). In the presence of GSH, DNB was removed from the probe, leading to an enhancement of fluorescence at 535 nm (yellow channel), which increased gradually with the smoothly decreased fluorescence at 650 nm ( Figure 7A-C). Whereas the increased viscosity could prevent the free rotation of the vinyl group, resulting in an elevated fluorescence at 627 nm when excited at 510 nm. Probe 28 was capable of not just tracking GSH and viscosity in living cells but also visualizing the production of bleb vesicles after nystatin stimulation. Apoptosis induced by cisplatin was finally investigated by fluorescence imaging using probe 28, with enhancements in both red and yellow channels indicating the upregulation of GSH and viscosity induced by apoptosis.
GSH and viscosity, with 2,4-dinitrobenzenyl (DNB) as the reactive site ( Figure 7A) [71]. Probe 28 exhibited absorption at 509 nm and emission at 650 nm (red channel). In the presence of GSH, DNB was removed from the probe, leading to an enhancement of fluorescence at 535 nm (yellow channel), which increased gradually with the smoothly decreased fluorescence at 650 nm ( Figure 7A-C). Whereas the increased viscosity could prevent the free rotation of the vinyl group, resulting in an elevated fluorescence at 627 nm when excited at 510 nm. Probe 28 was capable of not just tracking GSH and viscosity in living cells but also visualizing the production of bleb vesicles after nystatin stimulation. Apoptosis induced by cisplatin was finally investigated by fluorescence imaging using probe 28, with enhancements in both red and yellow channels indicating the upregulation of GSH and viscosity induced by apoptosis. The Yin group developed a fluorescent probe 29 (HBT-COU) to detect GSH in living cells [72], using a hybrid of coumarin and 2-(2-Hydroxyphenyl)benzothiazole featuring ESIPT as the fluorophore. Probe 29 exhibited a fast ratiometric response to biothiols, with a green emission at 519 nm owing to the ICT process of the enol form and another red emission at 621 nm attributed to the ESIPT process of the keto form (Scheme 12). With high selectivity towards GSH over Hcy and Cys, probe 29 was then successfully utilized for ratiometric fluorescence imaging of endogenous and exogenous GSH in living cells; declining GSH levels were visualized in apoptotic cell lines compared to the healthy lines.
In addition to the specific fluorescent probes for Cys and GSH, a few selective probes for Hcy have also been reported [73,74], of which we will not give detailed examples here. Readers interested in these can read the corresponding references provided by us. The Yin group developed a fluorescent probe 29 (HBT-COU) to detect GSH in living cells [72], using a hybrid of coumarin and 2-(2-Hydroxyphenyl)benzothiazole featuring ESIPT as the fluorophore. Probe 29 exhibited a fast ratiometric response to biothiols, with a green emission at 519 nm owing to the ICT process of the enol form and another red emission at 621 nm attributed to the ESIPT process of the keto form (Scheme 12). With high selectivity towards GSH over Hcy and Cys, probe 29 was then successfully utilized for ratiometric fluorescence imaging of endogenous and exogenous GSH in living cells; declining GSH levels were visualized in apoptotic cell lines compared to the healthy lines.
In addition to the specific fluorescent probes for Cys and GSH, a few selective probes for Hcy have also been reported [73,74], of which we will not give detailed examples here. Readers interested in these can read the corresponding references provided by us.

Mitochondria-Specific Fluorescent Probes for Biothiols
Mitochondria are the main sites of cellular metabolism, and their malfunction frequently causes cellular redox imbalance, which leads to cell death. Biothiols are important reducing agents in mitochondria that may remove oxidative free radicals generated by mitochondria and maintain intracellular redox homeostasis. Therefore, the development of mitochondria-specific probes for biothiols is of practical significance for studying the physiological and pathological functions of biothiols at the subcellular level. Molecules 2023, 28, x FOR PEER REVIEW 17 of 41 Scheme 12. Chemical structure and reaction mechanism of probe 29 with GSH.

Mitochondria-Specific Fluorescent Probes for Biothiols
Mitochondria are the main sites of cellular metabolism, and their malfunction frequently causes cellular redox imbalance, which leads to cell death. Biothiols are important reducing agents in mitochondria that may remove oxidative free radicals generated by mitochondria and maintain intracellular redox homeostasis. Therefore, the development of mitochondria-specific probes for biothiols is of practical significance for studying the physiological and pathological functions of biothiols at the subcellular level.
The proton pumps in the inner mitochondrial membrane transport protons into the mitochondrial intermembranous region during mitochondrial respiration, resulting in the creation of a strongly negative mitochondrial transmembrane potential. Inspired by this process, many delocalized lipophilic cations exhibit mitochondrion targeting capabilities. One typical targeting ligand is triphenylphosphonium (TPP). Utilizing it as the mitochondria targeting group, Zhang et al. reported a dual-channel fluorescent probe 30 (NTG) for simultaneous detection of GSH and ONOOin mitochondria [75]. Probe 30 could respond to GSH and ONOOin two channels; an enhanced fluorescence was recorded in the red channel (670 nm) in the presence of GSH, while another increased emission was found in the green channel (530 nm) in the simultaneous presence of GSH and ONOO − , whereas ONOO − could afford only weak fluorescence in the green channel ( Figure 8A,B). Probe 30 displayed desirable selectivity, sensitivity, stability, and mitochondria targeting ( Figure  8C), which was then applied to track endogenous and exogenous GSH and ONOOin normal, inflammatory, and tumor cells of different stages. Imaging experiments were also carried out to assess the capability of discriminating different zebrafish models. Significantly enhanced green fluorescence and sharply decreased red fluorescence were found in inflammatory zebrafish compared with the normal ones, while increased green fluorescence and unchanged red fluorescence were recorded in cancerous ones ( Figure 8D).

Scheme 12.
Chemical structure and reaction mechanism of probe 29 with GSH.
The proton pumps in the inner mitochondrial membrane transport protons into the mitochondrial intermembranous region during mitochondrial respiration, resulting in the creation of a strongly negative mitochondrial transmembrane potential. Inspired by this process, many delocalized lipophilic cations exhibit mitochondrion targeting capabilities. One typical targeting ligand is triphenylphosphonium (TPP). Utilizing it as the mitochondria targeting group, Zhang et al. reported a dual-channel fluorescent probe 30 (NTG) for simultaneous detection of GSH and ONOOin mitochondria [75]. Probe 30 could respond to GSH and ONOOin two channels; an enhanced fluorescence was recorded in the red channel (670 nm) in the presence of GSH, while another increased emission was found in the green channel (530 nm) in the simultaneous presence of GSH and ONOO − , whereas ONOO − could afford only weak fluorescence in the green channel ( Figure 8A,B). Probe 30 displayed desirable selectivity, sensitivity, stability, and mitochondria targeting ( Figure 8C), which was then applied to track endogenous and exogenous GSH and ONOOin normal, inflammatory, and tumor cells of different stages. Imaging experiments were also carried out to assess the capability of discriminating different zebrafish models. Significantly enhanced green fluorescence and sharply decreased red fluorescence were found in inflammatory zebrafish compared with the normal ones, while increased green fluorescence and unchanged red fluorescence were recorded in cancerous ones ( Figure 8D).
On the basis of a positively charged cyanine fluorophore, Zhu et al. developed a novel mitochondria-targetable probe 31 for biothiols (Cy-DNBS), with the DNBS moiety serving as the reactive site (Scheme 13) [76]. Probe 31 showed weak fluorescence under excitation at 545 nm; after adding biothiols, a tremendous increase was found at 604 nm with a moderate Stokes shift of 59 nm, and the intensity at 604 nm was used to estimate the level of biothiols. Further experiments demonstrated the probe's favorable properties, including higher sensitivity and selectivity to biothiols over other biological species, excellent photostability, low cytotoxicity, and mitochondria targeting. Probe 31 was then used for fluorescent imaging of endogenous biothiols in living cells, and attenuation of biothiols was observed in HeLa cells pretreated with hydrogen peroxide, demonstrating its potential to track intracellular biothiol changes under oxidative stress. On the basis of a positively charged cyanine fluorophore, Zhu et al. developed a novel mitochondria-targetable probe 31 for biothiols (Cy-DNBS), with the DNBS moiety serving as the reactive site (Scheme 13) [76]. Probe 31 showed weak fluorescence under excitation at 545 nm; after adding biothiols, a tremendous increase was found at 604 nm with a moderate Stokes shift of 59 nm, and the intensity at 604 nm was used to estimate the level of biothiols. Further experiments demonstrated the probe's favorable properties, including higher sensitivity and selectivity to biothiols over other biological species, excellent photostability, low cytotoxicity, and mitochondria targeting. Probe 31 was then used for fluorescent imaging of endogenous biothiols in living cells, and attenuation of biothiols was observed in HeLa cells pretreated with hydrogen peroxide, demonstrating its potential to track intracellular biothiol changes under oxidative stress.
Lin reported another cyanine-based NIR fluorescent probe, 32 (GalCys), to monitor levels of Cys in mitochondria (Scheme 11) [77]. The probe exhibited significantly enhanced fluorescence emission at 668 nm after responding to Cys (0-200 µM). With outstanding sensitivity and selectivity to Cys, desirable NIR emission, and favorable biocompatibility, GalCys was applied to detect changes in Cys in mitochondria of living cells, proving for the first time that the development of diabetes led to oxidative stress in mitochondria. A dramatic decrement of Cys was observed in further imaging experiments on mouse models exposed to excess particulate matter from environmental pollution, indicating serious destruction from airborne particulate matter pollution to the redox balance of organisms. In recent years, multi-targeting fluorescent probes have demonstrated great potential to monitor the interplay between different organelles and the imaging of invisible biophysical parameters. The Song group designed and synthesized a Cys-responsive fluorescent probe 33 (BEB-A), which can label mitochondria due to the benzoindolium cationic group serving as a targeting group [78]. In the presence of Cys, probe 33 exhibited an elevated emission at 616 nm under excitation at 550 nm, with a limit of detection of 0.027 µM. Most importantly, after responding to the Cys, probe 33 could enter the nucleus and Lin reported another cyanine-based NIR fluorescent probe, 32 (GalCys), to monitor levels of Cys in mitochondria (Scheme 11) [77]. The probe exhibited significantly enhanced fluorescence emission at 668 nm after responding to Cys (0-200 µM). With outstanding sensitivity and selectivity to Cys, desirable NIR emission, and favorable biocompatibility, GalCys was applied to detect changes in Cys in mitochondria of living cells, proving for the first time that the development of diabetes led to oxidative stress in mitochondria. A dramatic decrement of Cys was observed in further imaging experiments on mouse models exposed to excess particulate matter from environmental pollution, indicating serious destruction from airborne particulate matter pollution to the redox balance of organisms.
In recent years, multi-targeting fluorescent probes have demonstrated great potential to monitor the interplay between different organelles and the imaging of invisible biophysical parameters. The Song group designed and synthesized a Cys-responsive fluorescent probe 33 (BEB-A), which can label mitochondria due to the benzoindolium cationic group serving as a targeting group [78]. In the presence of Cys, probe 33 exhibited an elevated emission at 616 nm under excitation at 550 nm, with a limit of detection of 0.027 µM. Most importantly, after responding to the Cys, probe 33 could enter the nucleus and bind to the RNA, leading to an enhancement in red fluorescence and providing a new approach for the study of mitochondria and nucleolus. Similarly, the He group developed another Cysresponsive mitochondrion-targetable probe 34 (TSQC) using acrylate-decorated cationic AIEgen [79]. Free probe 34 emitted at 750 nm, whereas after adding Cys, the probe could spontaneously target lipid drops with an enhanced emission at 650 nm, indicating its ability to monitor LDs and mitochondria in dual channels. Using Probe 34, apoptosis induced by LPS, H 2 O 2 , or UV light exposure treatments was tracked by dual-channel visualization. Finally, apoptosis associated with acute and chronic epilepsies was also monitored in mouse models, providing excellent applicability for Cys-activatable imaging of LDs and mitochondria in two channels for the first time.

Design Strategies for Reversible Fluorescent Probes for Biothiols
Biothiols, as essential regulators in living systems, play critical roles in various physiological processes such as redox balance maintenance, signal transduction, gene expression, protein synthesis, and so on. Real-time monitoring of dynamic changes in biothiols is of great significance to reveal their physiological and pathological functions and provide a foundation for early treatment and diagnosis of diseases. To this end, many reversible fluorescent probes have been reported to monitor the dynamics of GSH in living specimens recently, and these reversible fluorescent probes are frequently designed on the basis of the reversible Michael addition reactions (Scheme 14) [80,81]. Optical experiments indicated that the intensity ratio of I 505 /I 587 was competent to detect GSH ranging from 0.1 to 10 mM, while the intensity of I 505 was more suitable for GSH with a larger concentration range (0.1 to 30/50 nM). Considering the level of intracellular GSH expression, the former could be used to detect intracellular GSH. Most importantly, probe 36 exhibited a reversible and extremely fast response toward GSH (half-time 3 s), which was necessary for real-time imaging. Due to its favorable biocompatibility, probe 36 was successfully used for ratiometric fluorescent imaging in living cells, and an image processing program was developed in MATLAB (a commercial mathematics software) to facilitate the real-time imaging of GSH.
Shen et al. reported a reversible fluorescent probe 37 (B-GSH) based on a BODIPY derivative with exocyclic extended conjugations (Scheme 14) [84]. The addition of GSH afforded an increased absorption at 527 nm with enhanced fluorescence at 544 nm (green channel) and a decreased absorption at 594 nm with reduced fluorescence at 603 nm (red channel). Additionally, reversible behaviors were verified after the elimination of GSH. Finally, intracellular GSH dynamics were monitored by fluorescent imaging using probe 37. Pretreated with cisplatin, GSH concentration first increased to the maximum, then fell back to the initial concentration without significant apoptosis in cisplatin-resistant A549 cells, while GSH levels increased until cell death in cisplatin-sensitive cell lines. The results effectively proved the potential of probe 37 to detect intracellular GSH dynamics in response to therapeutics.
The Cui group developed a novel reversible NIR ratiometric fluorescent probe 38 (EpSiP) by molecular engineering, enabling in vivo quantification of GSH in intact tissues and animals (Scheme 14) [85]. Probe 38 was constructed on a FRET scaffold, where Si-rhodamine and phospha-rhodamine were used as the FRET acceptor and donor, respectively. An effective FRET process was recorded in the NIR region, and the absorption and emission spectra of the two fluorophores were well resolved (∆λ abs ≈ ∆λ em = 61 nm), which was conducive for bioimaging. Free probe 38 exhibited a strong emission in the red channel (736 nm) belonging to the phospha-rhodamine and a weak emission in the green channel (675 nm) belonging to the Si-rhodamine under an excitation of 640 nm. In the presence of GSH, the intensity in the red channel decreased with an enhancement in the green channel, and the spectra recovered immediately after adding H 2 O 2 , indicating a reversible response to GSH. Probe 38 exhibited favorable reaction kinetics and thermodynamics to GSH (K d = 4.9 mM, k = 81 M −1 s −1 , t 1/2 = 0.57, [GSH] = 10 mM), and was then applied to quantitatively detect the GSH levels in different biospecimens, including living cells, xenografted tumors on the mouse, chronic renal failure, and liver fibrosis in time.
The Yin group designed a reversible mitochondria-targeted fluorescent probe 39 (Mito-1, Scheme 3) based on a coumarin derivative to detect changes in mitochondrial Cys (Scheme 14) [86]. Michael addition shortened the conjugation system of probe 39, which led to a blue shift in the UV-vis absorption and an enhancement in fluorescence intensity at 498 nm. Due to its advantages, including high sensitivity, fast response, excellent reversibility, and low cytotoxicity, probe 39 was successfully used for the fluorescent detection of mitochondrial Cys in living cells. Encouraged by these results, another lysosome-targeting probe named probe 40 was also designed. The two probes were used to study the activation of the cellular inherent antioxidation system by resveratrol, and the enhancement in fluorescence indicated the Cys upregulation induced by the resveratrol.

Design Strategies of Fluorescent Probes for Discriminated Detection of Biothiols
In the above sections, we have introduced specific and reversible fluorescent probes. Most of these probes have been applied to monitor biothiols in cells and in vivo, showing potential applications in biomedical and disease research. However, they always fail to provide researchers with detailed information on diverse biothiols, as they cannot simultaneously detect multiple biothiols. Of course, this dilemma can be overcome by using multiple probes, but such multi-probe strategies often result in distorted detection results due to the differences in biothiol distribution, metabolism, and membrane penetration. Therefore, it is of great significance to construct discriminating fluorescent probes that can simultaneously detect and discriminate among multiple biothiols [87,88].

Discriminated Probes with a Single Reactive Site for Biothiols
7-nitrobenzofurazan (NBD) is a well-known reactive group involved in the design of specific and discriminating fluorescent probes for biothiols. Due to its strong electron deficiency, NBD quenches the fluorescence of the probe efficiently by the PET process. Reactions with biothiols result in the formation of free fluorophores with enhanced emission; specifically, reactions between NBD-probe and GSH lead to the formation of thiol-NBD, while amino-NBD derivatives are generated with Cys/Hcy, exhibiting distinct optical performance from thiol-NBD to amino-NBD. Therefore, a systematic approach can be achieved for distinguishing detection towards GSH and Cys/Hcys (Scheme 16). In 2021, Wang et al. reported a probe 41 (MZ-NBD) for the detection of GSH in living cells, with NBD as a fluorescent response group (Scheme 16) [89]. Hcy and Cys reacted with probe 41 by cascade nucleophilic substitution, leading to the formation of amino-NBD with fluorescence at 560 nm, whereas GSH and probe 41 afforded thiol-NBD by a one-step nucleophilic substitution with slightly red-shift fluorescence at 565 nm. Noteworthy, probe 41 could detect GSH with fluorescence fusion at 530 nm, while two distinct emission peaks were found in the detection of Hcy and Cys. Owing to its high sensitivity, favorable selectivity, fast response, and moderate biocompatibility, probe 41 was used to track the endogenous and exogenous GSH in MCF-7 cells, and its ability to avoid interference from Cys was also verified in the imaging experiments. The Feng group developed a unimolecular fluorescent probe 42 (ZED) based on a coumarin-rhodamine FRET platform with NBD and hydrazide as the Cys/Hcy and HClO sensing sites, respectively (Scheme 16) [90]. With imidazolium used as the targeting group for mitochondria, probe 42 was applied to track the Cys/Hcy, HOCl, ∆ψm, and opening of the MPT pore by four channels without any crosstalk.
dogenous and exogenous GSH in MCF-7 cells, and its ability to avoid interference from Cys was also verified in the imaging experiments. The Feng group developed a unimolecular fluorescent probe 42 (ZED) based on a coumarin-rhodamine FRET platform with NBD and hydrazide as the Cys/Hcy and HClO sensing sites, respectively (Scheme 16) [90]. With imidazolium used as the targeting group for mitochondria, probe 42 was applied to track the Cys/Hcy, HOCl, Δψm, and opening of the MPT pore by four channels without any crosstalk. Although chemotherapy is an effective method to treat cancer, it remains controversial whether drug therapy can lead to abnormal expression of reactive oxygen species (ROS) and reactive sulfur species (RSS) by increasing intracellular oxidative stress. The Ye group developed a multifunctional probe 43 (RSS-HClO) to track biothiols and HClO levels during chemotherapy (Scheme 17) [91]. Probe 43 was constructed with a coumarinhemicyanine moiety and NBD as the reporter and reactive site, allowing it to distinguish Cys/Hcy, GSH, and SO2 using three channels simultaneously. Despite almost no fluorescence due to the PET effect, probe 43 demonstrated a substantial increase in the red channel when exposed to GSH. On the other hand, either the red or green channel recorded significantly enhanced fluorescence after Cys/Hcy addition, while HClO only caused an increase in the fluorescence blue channel (Scheme 17). Attributed to its high sensitivity, selectivity, and favorable biocompatibility, probe 43 was successfully utilized to discriminatively visualize Cys/Hcy, GSH, and HClO, and its mitochondria targeting ability was verified by co-localization experiments. Further, an upregulation of cellular biothiols and HClO in tumor cells during chemotherapy was observed, indicating potential applications in studying pharmacologic actions. Although chemotherapy is an effective method to treat cancer, it remains controversial whether drug therapy can lead to abnormal expression of reactive oxygen species (ROS) and reactive sulfur species (RSS) by increasing intracellular oxidative stress. The Ye group developed a multifunctional probe 43 (RSS-HClO) to track biothiols and HClO levels during chemotherapy (Scheme 17) [91]. Probe 43 was constructed with a coumarin-hemicyanine moiety and NBD as the reporter and reactive site, allowing it to distinguish Cys/Hcy, GSH, and SO 2 using three channels simultaneously. Despite almost no fluorescence due to the PET effect, probe 43 demonstrated a substantial increase in the red channel when exposed to GSH. On the other hand, either the red or green channel recorded significantly enhanced fluorescence after Cys/Hcy addition, while HClO only caused an increase in the fluorescence blue channel (Scheme 17). Attributed to its high sensitivity, selectivity, and favorable biocompatibility, probe 43 was successfully utilized to discriminatively visualize Cys/Hcy, GSH, and HClO, and its mitochondria targeting ability was verified by co-localization experiments. Further, an upregulation of cellular biothiols and HClO in tumor cells during chemotherapy was observed, indicating potential applications in studying pharmacologic actions. The Tang group has designed and synthesized a fluorescent probe 44 (TAT-probe, Scheme 18) for discriminating imaging of biothiols based on a native chemical ligation (NCL) reaction, using the penetrating peptide TAT (RRQRRKKRG) as a mitochondrial targeting unit [92]. Probe 44 was constructed in a FRET form, with naphthalimide serving as Scheme 17. Illustration of probe 43 and its distinguishing response to biothiols.
The Tang group has designed and synthesized a fluorescent probe 44 (TAT-probe, Scheme 18) for discriminating imaging of biothiols based on a native chemical ligation (NCL) reaction, using the penetrating peptide TAT (RRQRRKKRG) as a mitochondrial targeting unit [92]. Probe 44 was constructed in a FRET form, with naphthalimide serving as the donor, rhodamine as the acceptor, and thiophenol ester as the sensing site and linker between the donor and acceptor. When excited at 404 nm, probe 44 exhibited a red fluorescence peaking at 585 nm in the absence of biothiols due to the FRET process. However, upon the addition of biothiols, the red fluorescence fell with an enhanced green emission at 520 nm, indicating the blocked FRET process, and the fluorescence intensities ratio (I 520 /I 585 ) showed good linearity to the concentration of biothiols, making this ratiometric fluorescence probe less prone to interference from the concentration and distribution of the probe. More excitingly, probe 44 exhibited almost unchanged red fluorescence towards GSH under excitation at 545 nm, while a significant decrease in red fluorescence was observed in Cys/Hcy titration, enabling probe 44 to sensitively distinguish GSH from these three biothiols. The probe was further utilized for fluorescent imaging of biothiols in living cells, with favorable compatibility and mitochondrial targeting from the TAT peptide. Most importantly, GSH discrimination was achieved in NEM-pretreated cells.
Scheme 17. Illustration of probe 43 and its distinguishing response to biothiols.
The Tang group has designed and synthesized a fluorescent probe 44 (TAT-probe, Scheme 18) for discriminating imaging of biothiols based on a native chemical ligation (NCL) reaction, using the penetrating peptide TAT (RRQRRKKRG) as a mitochondrial targeting unit [92]. Probe 44 was constructed in a FRET form, with naphthalimide serving as the donor, rhodamine as the acceptor, and thiophenol ester as the sensing site and linker between the donor and acceptor. When excited at 404 nm, probe 44 exhibited a red fluorescence peaking at 585 nm in the absence of biothiols due to the FRET process. However, upon the addition of biothiols, the red fluorescence fell with an enhanced green emission at 520 nm, indicating the blocked FRET process, and the fluorescence intensities ratio (I520/I585) showed good linearity to the concentration of biothiols, making this ratiometric fluorescence probe less prone to interference from the concentration and distribution of the probe. More excitingly, probe 44 exhibited almost unchanged red fluorescence towards GSH under excitation at 545 nm, while a significant decrease in red fluorescence was observed in Cys/Hcy titration, enabling probe 44 to sensitively distinguish GSH from these three biothiols. The probe was further utilized for fluorescent imaging of biothiols in living cells, with favorable compatibility and mitochondrial targeting from the TAT peptide. Most importantly, GSH discrimination was achieved in NEM-pretreated cells. The Song group reported a multi-response site probe 45 (KC) to discriminatively visualize thiols in living cells and zebrafish (Scheme 19) [93]. The probe utilized an azide moiety as a reactive site for H 2 S, causing a pronounced increase in the red channel. The thiocoumarin was applied to distinguish thiols based on different reaction mechanisms. GSH affords a sharp enhancement in the blue channel via the SNAr reaction, while Cys undergoes a significant promotion in the green channel by a cascade reaction that includes SNAr reaction and Smiles rearrangement. Hcy follows a similar process to Cys, but its slower rearrangement results in a dual increase in blue and green channels. Probe 45 displays moderate sensitivity, favorable selectivity, and low cytotoxicity, making it suitable for detecting thiols in living cells and three day old zebrafish models.
The Wang group has developed a novel fluorescence probe 46 (CySI) based on cyanines and thioesters capable of distinguishing Cys and Hcy through two distinct emission channels under single excitation (Scheme 20) [94]. When excited at 550 nm, the probe displayed enhanced fluorescence at 625 nm and 740 nm in reaction with Cys and increased fluorescence at 740 nm for Hcy, without any fluorescence response to GSH. This feature enables the simultaneous identification and determination of Cys and Hcy through two separate channels. Further, the probe was used to visualize exogenous Cys and Hcy in living cells, demonstrating its capability to effectively target mitochondria and its potential to monitor endogenous Cys and Hcy fluctuations in mitochondria through red emission channels. thiocoumarin was applied to distinguish thiols based on different reaction mechanisms. GSH affords a sharp enhancement in the blue channel via the SNAr reaction, while Cys undergoes a significant promotion in the green channel by a cascade reaction that includes SNAr reaction and Smiles rearrangement. Hcy follows a similar process to Cys, but its slower rearrangement results in a dual increase in blue and green channels. Probe 45 displays moderate sensitivity,favorable selectivity, and low cytotoxicity, making it suitable for detecting thiols in living cells and three day old zebrafish models. Scheme 19. Structure of probe 45 and its response mechanism towards thiols.
The Wang group has developed a novel fluorescence probe 46 (CySI) based on cyanines and thioesters capable of distinguishing Cys and Hcy through two distinct emission channels under single excitation (Scheme 20) [94]. When excited at 550 nm, the probe displayed enhanced fluorescence at 625 nm and 740 nm in reaction with Cys and increased fluorescence at 740 nm for Hcy, without any fluorescence response to GSH. This feature enables the simultaneous identification and determination of Cys and Hcy through two separate channels. Further, the probe was used to visualize exogenous Cys and Hcy in living cells, demonstrating its capability to effectively target mitochondria and its potential to monitor endogenous Cys and Hcy fluctuations in mitochondria through red emission channels. Scheme 20. Structure of probe 46 and its response mechanism towards thiols.

Discriminated with Multiple Reactive Sites
The Song group developed a ratiometric probe 47 (CP) based on the conjugate addition-cyclization reaction (Scheme 21) [95]. When the biothiols were added, probe 47 exhibited two distinct fluorescence bands upon a single excitation at 380 nm. Interestingly, Cys caused a single fluorescence in the blue channel (485 nm), whereas dual emissions in the blue (608 nm) and red channels (485 nm) were recorded in the detection of Hcy. Due to its high sensitivity, high selectivity, low cytotoxicity, and outstanding photophysical properties, probe 47 was successfully applied to ratiometric fluorescence imaging of Cys and Hcy in HeLa cells and zebrafish, with effective discrimination between Cys and Hcy. Scheme 20. Structure of probe 46 and its response mechanism towards thiols.

Discriminated with Multiple Reactive Sites
The Song group developed a ratiometric probe 47 (CP) based on the conjugate additioncyclization reaction (Scheme 21) [95]. When the biothiols were added, probe 47 exhibited two distinct fluorescence bands upon a single excitation at 380 nm. Interestingly, Cys caused a single fluorescence in the blue channel (485 nm), whereas dual emissions in the blue (608 nm) and red channels (485 nm) were recorded in the detection of Hcy. Due to its high sensitivity, high selectivity, low cytotoxicity, and outstanding photophysical properties, probe 47 was successfully applied to ratiometric fluorescence imaging of Cys and Hcy in HeLa cells and zebrafish, with effective discrimination between Cys and Hcy. tion-cyclization reaction (Scheme 21) [95]. When the biothiols were added, probe 47 exhibited two distinct fluorescence bands upon a single excitation at 380 nm. Interestingly, Cys caused a single fluorescence in the blue channel (485 nm), whereas dual emissions in the blue (608 nm) and red channels (485 nm) were recorded in the detection of Hcy. Due to its high sensitivity, high selectivity, low cytotoxicity, and outstanding photophysical properties, probe 47 was successfully applied to ratiometric fluorescence imaging of Cys and Hcy in HeLa cells and zebrafish, with effective discrimination between Cys and Hcy. Scheme 21. Structure of probe 47 and its response mechanism towards thiols. Scheme 21. Structure of probe 47 and its response mechanism towards thiols.
The Yin group designed and synthesized a fluorescent probe 48 (CBB) for visualization of endogenous biothiols using the 4-chlorine moiety (Scheme 22) [96]. Probe 48 responded to biothiols with different mechanisms, affording simultaneous discrimination between Cys, Hcy, and GSH via three distinct emission channels. Owing to its favorable sensitivity, selectivity, and biocompatibility, probe 48 was employed for discriminative imaging of endogenous GSH, Hcy, and Cys and their transformation in living cells. The Yin group designed and synthesized a fluorescent probe 48 (CBB) for visualization of endogenous biothiols using the 4-chlorine moiety (Scheme 22) [96]. Probe 48 responded to biothiols with different mechanisms, affording simultaneous discrimination between Cys, Hcy, and GSH via three distinct emission channels. Owing to its favorable sensitivity, selectivity, and biocompatibility, probe 48 was employed for discriminative imaging of endogenous GSH, Hcy, and Cys and their transformation in living cells. The Liu group reported a fluorescent probe 49 (YF) for distinguishing the visualization of endogenous biothiols using a coumarin-hemicyanine fluorophore (Scheme 23) [97]. Probe 49 was equipped with three reactive sites and has multiple optical channels for fluorescence detection. Different sensing processes were recorded in the optical spectra when adding biothiols. Probe 49 exhibited similar trends in fluorescence to the three biothiols, but less amplitude was observed in GSH addition. In the time-dependent emission spectra, two-phase change processes were recorded for GSH, while three phases were found for Cys/Hcy, which was consistent with their varied response mechanisms. Therefore, it is possible for probe 49 to simultaneously discriminate GSH from Cys/Hcy. Due to its favorable selectivity, permeability, and low cytotoxicity, probe 49 was successfully employed to discriminate Cys/Hcy and GSH in A375 cells simultaneously based on different signaling patterns, and biothiols in fetal bovine serum samples were also measured by probe 49 with results consistent with those obtained by the Ellman method. The Liu group reported a fluorescent probe 49 (YF) for distinguishing the visualization of endogenous biothiols using a coumarin-hemicyanine fluorophore (Scheme 23) [97]. Probe 49 was equipped with three reactive sites and has multiple optical channels for fluorescence detection. Different sensing processes were recorded in the optical spectra when adding biothiols. Probe 49 exhibited similar trends in fluorescence to the three biothiols, but less amplitude was observed in GSH addition. In the time-dependent emission spectra, twophase change processes were recorded for GSH, while three phases were found for Cys/Hcy, which was consistent with their varied response mechanisms. Therefore, it is possible for probe 49 to simultaneously discriminate GSH from Cys/Hcy. Due to its favorable selectivity, permeability, and low cytotoxicity, probe 49 was successfully employed to discriminate Cys/Hcy and GSH in A375 cells simultaneously based on different signaling patterns, and biothiols in fetal bovine serum samples were also measured by probe 49 with results consistent with those obtained by the Ellman method.
Probe 49 was equipped with three reactive sites and has multiple optical channels for fluorescence detection. Different sensing processes were recorded in the optical spectra when adding biothiols. Probe 49 exhibited similar trends in fluorescence to the three biothiols, but less amplitude was observed in GSH addition. In the time-dependent emission spectra, two-phase change processes were recorded for GSH, while three phases were found for Cys/Hcy, which was consistent with their varied response mechanisms. Therefore, it is possible for probe 49 to simultaneously discriminate GSH from Cys/Hcy. Due to its favorable selectivity, permeability, and low cytotoxicity, probe 49 was successfully employed to discriminate Cys/Hcy and GSH in A375 cells simultaneously based on different signaling patterns, and biothiols in fetal bovine serum samples were also measured by probe 49 with results consistent with those obtained by the Ellman method. Scheme 23. The chemical structure and reaction mechanism of 49 towards biothiols. Scheme 23. The chemical structure and reaction mechanism of 49 towards biothiols.
As important biological mercaptans, Cys, Hcy, and GSH transform each other and participate in many critical physiological processes such as signal transduction, metabolism, and the maintenance of redox homeostasis. Studies have indicated that abnormal levels of sulfur dioxide (SO 2 ) are closely related to the occurrence and development of cancer. Since intracellular SO 2 is primarily generated through two distinct pathways, GSH and Cys, it is desirable to develop distinguishable fluorescent probes to trace these metabolic pathways and improve disease treatment. To concurrently follow these metabolic pathways, the Yin group developed probe 50 with multiple responding sites, including a constructed site in the reaction process (Scheme 24) [98]. Probe 50 could respond sequentially to both biothiols and SO 2 and afford separate fluorescence signals to Cys/Hcy and GSH, allowing them to differentiate the two metabolic processes. In the Cys/Hcy pathway, two enhanced fluorescence bands were found in the blue and red channels in the presence of Cys/Hcy, and the red channel fell with an increase in the blue channel after adding SO 2 , while in the GSH pathway, two elevated fluorescence bands were observed in the green and red channels in the presence of GSH, and both channels receded after adding SO 2 . Probe 50 was successfully applied to fluorescence imaging of Cys/GSH metabolism in living cells and HeLa tumor nude mice. Highly expressed thiols were observed in tumors compared to healthy tissues, and more excitingly, the probes performed efficient tumor targeting due to a benzopyrylium intermediate in the GSH pathway.
The Chen and Song groups described a novel fluorescent probe 51 (RC) using the combination of resorufin and coumarin moieties (Scheme 25) [99]. In the presence of thiols, a S N Ar substitution process cleaved off the resorufinyl group from the coumarin with a red emission, leading to the formation of thiol-coumarins. Subsequently, these thiol-coumarins underwent possible Smiles rearrangement and cyclization reactions, yielding diverse products and unique fluorescence combinations via three emission channels. Showing excellent selectivity, sensitivity, and low cytotoxicity, probe 51 was then applied to confocal imaging in living cells, indicating its robust potential to distinguish different cellular thiols.
of Cys/Hcy, and the red channel fell with an increase in the blue channel after adding SO2, while in the GSH pathway, two elevated fluorescence bands were observed in the green and red channels in the presence of GSH, and both channels receded after adding SO2. Probe 50 was successfully applied to fluorescence imaging of Cys/GSH metabolism in living cells and HeLa tumor nude mice. Highly expressed thiols were observed in tumors compared to healthy tissues, and more excitingly, the probes performed efficient tumor targeting due to a benzopyrylium intermediate in the GSH pathway. Scheme 24. Structure of probe 50 and its reaction mechanism towards biothiols and SO2.
The Chen and Song groups described a novel fluorescent probe 51 (RC) using the combination of resorufin and coumarin moieties (Scheme 25) [99]. In the presence of thiols, a SNAr substitution process cleaved off the resorufinyl group from the coumarin with a red emission, leading to the formation of thiol-coumarins. Subsequently, these thiolcoumarins underwent possible Smiles rearrangement and cyclization reactions, yielding diverse products and unique fluorescence combinations via three emission channels. Showing excellent selectivity, sensitivity, and low cytotoxicity, probe 51 was then applied to confocal imaging in living cells, indicating its robust potential to distinguish different cellular thiols. Scheme 24. Structure of probe 50 and its reaction mechanism towards biothiols and SO 2 .
Molecules 2023, 28, x FOR PEER REVIEW 28 of 41 Scheme 25. Structure of probe 51 and its reaction mechanism towards thiols.

Conclusions and Outlook
In conclusion, we reviewed the recent progress of biothiol fluorescent probes over the past four years and briefly introduced the strategies for biothiol probe design from four perspectives: response mechanism, specific imaging, reversible detection, and discriminated detection (Table 1). Over the past decades, fluorescent probes have been intensively studied for the detection of biothiols. Numerous fluorophores featuring red emission, a large Stokes shift, AIE characteristics, and high fluorescent quantum yields have been ingeniously developed and employed in the design of biothiol fluorescent probes [25,26,100]. Diverse imaging modalities, such as turn-on, two-photon, and ratiometric imaging, have also been applied to the biosensing of biothiols. Fluorescent imaging for biothiols has been successfully achieved on a spectrum of specimens, including food, serum, living cells, tissues, organs, and even animal models, which indicates the great potential of probes in pathological research, biomedicine, and disease diagnosis.
However, there are still challenges to studying novel biothiol probes. (1) Although many specific biothiol fluorescent probes have been developed, their selectivity is still insufficient for bioimaging considering the differences in concentration of diverse biothiols in the body. Moreover, compared with specific fluorescent probes for GSH and Cys, few specific fluorescent probes could be used for HCy detection. Therefore, it is vital to explore response mechanisms with efficient selectivity for biothiols. (2) Reversible and discrimi-Scheme 25. Structure of probe 51 and its reaction mechanism towards thiols.

Conclusions and Outlook
In conclusion, we reviewed the recent progress of biothiol fluorescent probes over the past four years and briefly introduced the strategies for biothiol probe design from four perspectives: response mechanism, specific imaging, reversible detection, and discriminated detection (Table 1). Over the past decades, fluorescent probes have been intensively studied for the detection of biothiols. Numerous fluorophores featuring red emission, a large Stokes shift, AIE characteristics, and high fluorescent quantum yields have been ingeniously developed and employed in the design of biothiol fluorescent probes [25,26,100]. Diverse imaging modalities, such as turn-on, two-photon, and ratiometric imaging, have also been applied to the biosensing of biothiols. Fluorescent imaging for biothiols has been successfully achieved on a spectrum of specimens, including food, serum, living cells, tissues, organs, and even animal models, which indicates the great potential of probes in pathological research, biomedicine, and disease diagnosis. However, there are still challenges to studying novel biothiol probes. (1) Although many specific biothiol fluorescent probes have been developed, their selectivity is still insufficient for bioimaging considering the differences in concentration of diverse biothiols in the body. Moreover, compared with specific fluorescent probes for GSH and Cys, few specific fluorescent probes could be used for HCy detection. Therefore, it is vital to explore response mechanisms with efficient selectivity for biothiols. (2) Reversible and discriminant fluorescent probes have been applied to the detection of biothiols in living cells and animal models. However, due to the influence of poor stability, photobleaching, and fast metabolism, it is still difficult to carry out long-term dynamic detection for biothiols in vivo. These drawbacks may be addressed by developing fluorescent dyes or nanoprobes with improved performance. (3) Fluorescence probes with NIR-I/NIR-II emission and two-photon imaging have been applied to detect biothiols in deep tissues in vivo, but the imaging depth is still limited. Therefore, there is still an urgent need to develop multi-modality fluorescent probes combined with NMR, PET, and photoacoustic imaging to provide holistic information on biothiols in vivo for physiological and pathological studies [101,102]. (4) Due to the ubiquitous presence of biothiols in living systems, it is necessary to develop fluorescent probes with controllable responsiveness to achieve high spatiotemporal resolution imaging of biothiols in vivo, such as photocontrollable fluorogenic probes [103,104]. (5) Fluorescence-guided surgery is a robust method to selectively remove tumor sites with the assistance of fluorescence imaging. However, the existing fluorescent probes may have serious adverse effects on the human body. Therefore, developing non-toxic biothiol fluorescent probes would be beneficial for fluorescence-guided surgery [105].
Author Contributions: Conceptualization, K.Z. and S.Y.; writing-original draft preparation, W.X., J.J. and K.Z.; writing-review and editing, W.X., S.Y. and D.S.; supervision, K.Z. and Y.Z.; funding acquisition, K.Z. and S.Y. All authors have read and agreed to the published version of the manuscript.