An Imidazo[1,5-a]pyridine Benzopyrylium-Based NIR Fluorescent Probe with Ultra-Large Stokes Shifts for Monitoring SO2

A mitochondria-targeted NIR probe based on the FRET mechanism was developed. It shows ultra-large Stokes shifts (460 nm) and emission shifts (285 nm). Furthermore, we also realized the imaging of SO2 in living SKOV-3 cells, zebrafish and living mice which may be useful for understanding the biological roles of SO2 in mitochondria and in vivo.


Introduction
Sulfur dioxide, a well-known atmospheric pollutant, has been regarded as a new possible gas transmitter following NO, CO and H 2 S [1][2][3][4]. It plays important roles in many physiological processes. SO 2 can dissolve easily in water to form its derivatives bisulfite (HSO 3 − ) and sulfite (SO 3 2− ), so the physiological functions of SO 2 can be attributed to its derivatives (HSO 3 − /SO 3 2− ). However, a high level of endogenous SO 2 , generated by the oxidation of H 2 S and thiol-containing amino acids in mitochondria, may bring about neurological disorders, cancers and other diseases [5][6][7][8]. Hence, it is greatly important to establish sensitive and rapid methods for SO 2 detection to further gain insight into its functions in biological systems, especially in mitochondria.
As classic fluorophores, hemicyanines have drawn increasing attention because of their simple synthesis and excellent response to SO 2 [26]. Their derivatives were selected as acceptors to construct FRET probes [27,28]. However, the emission of the hemicyanines is around 600 nm, which seriously limits their application in vivo. Therefore, it is of significance to search for new fluorophores, especially with NIR emission, as acceptors.
On the other hand, to build an effective FRET platform, the development of new fluorophores as donors whose emission overlaps well with the absorption of acceptors is essential. Owing to the good optical properties [29], imidazole[1,5-a]pyridines were selected as the donor to construct the FRET platform [30]. In addition, we chose benzopyran salt as the acceptor because of its NIR emission. Meanwhile, the benzopyran moiety could not only be used as a reactive site for the Michael addition reaction with SO 2 to achieve detection purposes, but it could also target mitochondria due to positive electricity. Therefore, the Molecules 2023, 28, 515 2 of 9 designed probe IPB-RL-1 could successfully achieve its imaging of SO 2 in mitochondria in SKOV-3 cells.

Synthesis of IPB-RL-1
The probe IPB-RL-1 was easily prepared using a classic organic reaction, as shown in Scheme 1. The structure was confirmed by NMR and HRMS.
Molecules 2023, 28, x FOR PEER REVIEW 2 of 9 Therefore, the designed probe IPB-RL-1 could successfully achieve its imaging of SO2 in mitochondria in SKOV-3 cells.

Synthesis of IPB-RL-1
The probe IPB-RL-1 was easily prepared using a classic organic reaction, as shown in Scheme 1. The structure was confirmed by NMR and HRMS. Scheme 1. Synthetic route of IPB-RL-1.

Optical Properties of IPB-RL-1
To examine the optical properties of IPB-RL-1, we first examined its selectivity. As shown in Figure 1a,b, there were no obvious changes in absorption and emission after the probe reacted with various ions (Br − , CH3COO − , Cl − , ClO4 − , ClO − , F − , H2PO4 − , HCO3 − , HPO4 2− , HS − , I − , NO2 − , NO3 − , S2O8 2− , SO4 2− , GSH, and Cys). However, when SO3 2− was added, it was clearly observed by the naked eye that the probe solution changed from blue to colorless, and the fluorescence intensity was quenched at 760 nm, indicating that IPB-RL-1 showed good selectivity for the detection of SO3 2− . The anti-interference experiment (Figure S1) demonstrated that IPB-RL-1 had good anti-interference performance and could specifically detect SO3 2− even in the presence of other ions.

Optical Properties of IPB-RL-1
To examine the optical properties of IPB-RL-1, we first examined its selectivity. As shown in Figure 1a . However, when SO 3 2− was added, it was clearly observed by the naked eye that the probe solution changed from blue to colorless, and the fluorescence intensity was quenched at 760 nm, indicating that IPB-RL-1 showed good selectivity for the detection of SO 3 2− . The antiinterference experiment ( Figure S1) demonstrated that IPB-RL-1 had good anti-interference performance and could specifically detect SO 3 2− even in the presence of other ions. For a better application in living systems, UV-vis and fluorescence titration experiments were also carried out. As shown in Figure 2a, IPB-RL-1 has a strong UV-vis absorption peak at 620 nm in the solution of DMSO/PBS (V/V = 3/7). Yet, with the continuous addition of SO 3 2− , the absorption peak at 620 nm decreased and the absorption peak at 310 nm increased. Meanwhile, the naked eye captured a rapid color change of the probe solution from blue to colorless. The near-infrared fluorescence emission peak at 760 nm decreased with the increase of SO 3 2− while the emission peak at 470 nm increased (Figure 2b), which further confirmed that the FRET was turned off. In addition, an excellent linear correlation between the ratio F 470 /F 760 and SO 3 2− concentration was observed. The detection limit was calculated to be 0.98 µM using the linear regression curve ( Figure S2) and LOD formula (LOD = 3 σ/k, σ is the standard deviation of the blank measurement, and k is the slope of the fluorescence emission ratio (I 475 /I 760 ) and SO 3 2− concentration). In the process of monitoring the reaction time between IPB-RL-1 and SO 3 2− , the fluorescence For a better application in living systems, UV-vis and fluorescence titration experiments were also carried out. As shown in Figure 2a, IPB-RL-1 has a strong UV-vis absorption peak at 620 nm in the solution of DMSO/PBS (V/V = 3/7). Yet, with the continuous addition of SO3 2-, the absorption peak at 620 nm decreased and the absorption peak at 310 nm increased. Meanwhile, the naked eye captured a rapid color change of the probe solution from blue to colorless. The near-infrared fluorescence emission peak at 760 nm decreased with the increase of SO3 2-while the emission peak at 470 nm increased (Figure 2b), which further confirmed that the FRET was turned off. In addition, an excellent linear correlation between the ratio F470/F760 and SO3 2-concentration was observed. The detection limit was calculated to be 0.98 μM using the linear regression curve (Fig. S2) and LOD formula (LOD = 3 σ/k, σ is the standard deviation of the blank measurement, and k is the slope of the fluorescence emission ratio (I475/I760) and SO3 2− concentration). In the process of monitoring the reaction time between IPB-RL-1 and SO3 2− , the fluorescence intensity reached equilibrium (Fig.S3) in a very short time (less than 10 s). These results indicated that IPB-RL-1 was suitable for further application in imaging in cells and in vivo.   For a better application in living systems, UV-vis and fluorescence titration experiments were also carried out. As shown in Figure 2a, IPB-RL-1 has a strong UV-vis absorption peak at 620 nm in the solution of DMSO/PBS (V/V = 3/7). Yet, with the continuous addition of SO3 2-, the absorption peak at 620 nm decreased and the absorption peak at 310 nm increased. Meanwhile, the naked eye captured a rapid color change of the probe solution from blue to colorless. The near-infrared fluorescence emission peak at 760 nm decreased with the increase of SO3 2-while the emission peak at 470 nm increased (Figure 2b), which further confirmed that the FRET was turned off. In addition, an excellent linear correlation between the ratio F470/F760 and SO3 2-concentration was observed. The detection limit was calculated to be 0.98 μM using the linear regression curve (Fig. S2) and LOD formula (LOD = 3 σ/k, σ is the standard deviation of the blank measurement, and k is the slope of the fluorescence emission ratio (I475/I760) and SO3 2− concentration). In the process of monitoring the reaction time between IPB-RL-1 and SO3 2− , the fluorescence intensity reached equilibrium (Fig.S3) in a very short time (less than 10 s). These results indicated that IPB-RL-1 was suitable for further application in imaging in cells and in vivo.  The results of the MTT (Methyl Thiazolyl Tetrazolium) experiment ( Figure S4) showed that IPB-RL-1 had a lower cytotoxicity to SKOV-3 cells and could be used for further cell imaging experiments. In Figure 3, fluorescence in the red and blue channels were observed after SKOV-3 cells were incubated with the probe for 1 h. However, when the cells were incubated with the probe for 1 h and then incubated with SO 3 2− for 20 min, the fluorescence in the blue channel was enhanced and the fluorescence in the red channel was significantly weakened, which suggested that probe IPB-RL-1 could be used to detect SO 3 2− in SKOV-3 cells.
Next, since the benzopyran part of IPB-RL-1 is positively charged, the mitochondriatargeted experiment was tested. As shown in Figure 4, the red fluorescence of MitoTracker Red and the blue fluorescence of probe IPB-RL-1 overlap well (coefficient = 0.91).
further cell imaging experiments. In Figure 3, fluorescence in the red and blue channels were observed after SKOV-3 cells were incubated with the probe for 1 h. However, when the cells were incubated with the probe for 1 h and then incubated with SO3 2-for 20 min, the fluorescence in the blue channel was enhanced and the fluorescence in the red channel was significantly weakened, which suggested that probe IPB-RL-1 could be used to detect SO3 2− in SKOV-3 cells. Next, since the benzopyran part of IPB-RL-1 is positively charged, the mitochondriatargeted experiment was tested. As shown in Figure 4, the red fluorescence of MitoTracker Red and the blue fluorescence of probe IPB-RL-1 overlap well (coefficient = 0.91).
Owning to the excellent properties of IPB-RL-1 in cell imaging, its capability for the visualization of SO3 2− in zebrafish was examined. As depicted in Figure 5, weak blue and red fluorescent signals were observed in the control group. When the zebrafish were incubated with IPB-RL-1 for 1 h, the fluorescent signals became obviously strong both in the blue channel and the red channel. Yet, when the zebrafish were incubated with IPB-RL-1 for 1 h and then Na2SO3 for 30 min, the fluorescent signals in the red channel became obviously weak while there was no significant change in the blue channel. Therefore, we believe that IPB-RL-1 can effectively image in vivo. Hence, imaging in mice was conducted to further explore its application advantages. As NIR fluorescence emission is required for the experiments in vivo, only fluorescence changes in the 698-766 nm range were used. As shown in Figure 6b, obvious signals were observed after the probe was injected into mice for 5 min. However, with the increase in Na2SO3 concentration, the fluorescence signals gradually weakened (Figure 6c,d). As the response time is less than 5 min, it is very suitable for the real-time monitoring of SO3 2− in mice.   Owning to the excellent properties of IPB-RL-1 in cell imaging, its capability for the visualization of SO 3 2− in zebrafish was examined. As depicted in Figure 5, weak blue and red fluorescent signals were observed in the control group. When the zebrafish were incubated with IPB-RL-1 for 1 h, the fluorescent signals became obviously strong both in the blue channel and the red channel. Yet, when the zebrafish were incubated with IPB-RL-1 for 1 h and then Na 2 SO 3 for 30 min, the fluorescent signals in the red channel became obviously weak while there was no significant change in the blue channel. Therefore, we believe that IPB-RL-1 can effectively image in vivo. Hence, imaging in mice was conducted to further explore its application advantages. As NIR fluorescence emission is required for the experiments in vivo, only fluorescence changes in the 698-766 nm range were used. As shown in Figure 6b, obvious signals were observed after the probe was injected into mice for 5 min. However, with the increase in Na 2 SO 3 concentration, the fluorescence signals gradually weakened (Figure 6c,d). As the response time is less than 5 min, it is very suitable for the real-time monitoring of SO 3 2− in mice.   Based on the above results, we envisioned the mechanism of detection as follows (Scheme 2). At the excitation wavelength of 380 nm, the donor (imidazo[1,5-a]pyridine) transfers energy to the acceptor (benzopyran) and NIR fluorescence emission at 760 nm was observed. However, after the addition of SO3 2− , the reaction between SO3 2− and benzopyran breaks the π conjugate of benzopyran, resulting in the destruction of FRET, and thus, the energy of imidazo[1,5-a] pyridine cannot be transferred to the benzopyran. Therefore, the fluorescence emission at 760 nm disappeared and the emission at 475 nm increased. This is also confirmed by 1 H NMR ( Figure S5). Based on the above results, we envisioned the mechanism of detection as follows (Scheme 2). At the excitation wavelength of 380 nm, the donor (imidazo[1,5-a]pyridine) transfers energy to the acceptor (benzopyran) and NIR fluorescence emission at 760 nm was observed. However, after the addition of SO 3 2− , the reaction between SO 3 2− and benzopyran breaks the π conjugate of benzopyran, resulting in the destruction of FRET, and thus, the energy of imidazo[1,5-a] pyridine cannot be transferred to the benzopyran. Therefore, the fluorescence emission at 760 nm disappeared and the emission at 475 nm increased. This is also confirmed by 1 H NMR ( Figure S5).
(Scheme 2). At the excitation wavelength of 380 nm, the donor (imidazo[1,5-a]pyridine) transfers energy to the acceptor (benzopyran) and NIR fluorescence emission at 760 nm was observed. However, after the addition of SO3 2− , the reaction between SO3 2− and benzopyran breaks the π conjugate of benzopyran, resulting in the destruction of FRET, and thus, the energy of imidazo [1,5-a] pyridine cannot be transferred to the benzopyran. Therefore, the fluorescence emission at 760 nm disappeared and the emission at 475 nm increased. This is also confirmed by 1 H NMR ( Figure S5). Scheme 2. Proposed mechanism.

Synthesis of the Probe IPB-RL-1
As demonstrated in Scheme 1, compounds 1-4 were synthesized according to the reported procedure [9,27].
Compound 3 (0.10 g, 0.24 mmol), compound 4 (0.10 g, 0.28 mmol) and CH3COOH (8 mL) were added to a 25 mL round-bottom flask. The mixture was heated to reflux for 3 h and then poured into water (100 mL). After being extracted with DCM (20 mL) three times, the combined organic solvent was removed under reduced pressure. The pure product was obtained by column chromatography (CH2Cl2:MeOH = 200:1). Black solid, 1

Synthesis of the Probe IPB-RL-1
As demonstrated in Scheme 1, compounds 1-4 were synthesized according to the reported procedure [9,27].
Compound 3 (0.10 g, 0.24 mmol), compound 4 (0.10 g, 0.28 mmol) and CH 3 COOH (8 mL) were added to a 25 mL round-bottom flask. The mixture was heated to reflux for 3 h and then poured into water (100 mL). After being extracted with DCM (20 mL) three times, the combined organic solvent was removed under reduced pressure. The pure product was obtained by column chromatography (CH 2 Cl 2 :MeOH = 200:1). Black solid, 1

Conclusions
In summary, a NIR ratiometric fluorescent probe IPB-RL-1 with an ultra-large Stokes shift (460 nm) that is superior to most reported probes has been developed. IPB-RL-1 shows high sensitivity and selectivity. Detection of SO 2 in mitochondria in living SKOV-3 cells was also realized. Moreover, the probe was successfully used to detect SO 2 in zebrafish which may be useful for the understanding of biological roles of SO 2 in mitochondria and in vivo. However, due to the small overlap between donor emission and acceptor absorption of the probe IPB-RL-1, the fluorescence transfer efficiency is only 51%, which implies that in order to obtain a high fluorescence transfer efficiency, the overlap effect between donor emission and acceptor absorption, in addition to the distance between donor and acceptor, should be carefully considered for the FRET-based probe design in the future.
Supplementary Materials: Supplementary data associated with this article can be found in the online version. The following supporting information can be downloaded at: https://www.mdpi.com/ article/10.3390/molecules28020515/s1, Figure S1: Ratiometric fluorescence responses F 475 /F 760 of IPB-RL-1 upon the addition of 10 equiv. SO 3 2− in the presence of 100 eq. background ions (1, probe;  Figure S5: Normalized emission spectra of donor (compound 3) and normalized absorption spectra of IPB-RL-1; Figure S6: The emission spectrum of probe IPB-RL-1 and donor; Figures S7-S13: 1 H NMR, 13 C NMR, HRMS of related compounds; Table S1: Comparison with other probes . Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data that support the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of Interest:
The authors declare no conflict of interest.