An ESIPT-induced NIR fluorescent probe to visualize mitochondrial sulfur dioxide during oxidative stress in vivo

Abstract

Based on the change in electron distribution of the benzopyrylium unit before and after sulfite addition, a 2-(2′-hydroxyphenyl)benzothiazole (HBT)-based fluorophore generated the excited state intramolecular proton transfer (ESIPT) process with a near-infrared enhanced emission at 836 nm and a large Stokes shift (286 nm). The probe was applied to image SO₂ derivatives in cells and mice. Our data will provide new ideas for the development of ESIPT-based fluorescence probe with longer wavelength emission.

---

An appropriate balance between oxidants and antioxidants is necessary for the survival of mammalian cells. An imbalance of redox homeostasis can cause metabolic syndrome, DNA damage, and even cancer.¹ Reactive oxygen species (ROS) are the signaling molecules for oxidative stress.² Cells can inhibit oxidative stress through endogenous antioxidants under an immune mechanism. Sulfur dioxide (SO₂) is an antioxidant. It can play an important part in maintaining redox homeostasis. SO₂ can be generated in mitochondria with sulfur-containing amino acids (RSS) oxidized by ROS.³ However, some studies have claimed that SO₂ has oxidative capability in complex biological environments. Thus, the role of SO₂ under oxidative stress must be reassessed.⁴ Mitochondria are the main sites of oxidation and phosphorylation. Excess oxidative stress can cause mitochondrial dysfunction. Therefore, exploring the level of SO₂ in mitochondria under oxidative stress will help to further understand the role of SO₂ in maintaining redox homeostasis.

Fluorescent probes are important tools for understanding the biological functions and mechanism of action of important biomolecules.⁵ Much effort has been devoted to developing fluorescent probes that target mitochondrial SO₂.⁶ We synthesized a series of probes based on the benzopyrylium group, which has a multifunctional unit⁷ with a built-in site for sulfite, one positive charge required for targeting mitochondria, and good solubility in water. These probes could detect sulfite with high sensitivity and rapidly, but along with a serious blue shift in the emission spectrum due to the destroyed conjugated system. That feature limited their practical applications in bioimaging because of interference from increased biological autofluorescence. Similar situations have been documented in most other probes that used “C=C” bonds as reaction sites for sulfite (Table S1, ESI†). Near-infrared (NIR) fluorescence imaging could be used due to the decreased autofluorescence background, deep penetration into tissue, and low light scattering.⁸ The induced enhanced emission was <700 nm for the reported SO₂ probes,⁹ which made them unsuitable for high-precision imaging. Hence, we transferred our interest to building an off–on long-wavelength NIR fluorescent probe for sulfite.

We considered the change in electron density of the benzopyrylium unit before and after sulfite addition, and the special conditions for a 2-(2′-hydroxyphenyl)benzothiazole (HBT)-based fluorophore to generate the excited state intramolecular proton transfer (ESIPT) process. An activated NIR-TS probe based on the ESIPT off–on process was designed under the guidance of hole–electron analyses based on time-dependent density-functional theory (TDDFT) calculations. As a response to SO₂, the NIR-TS probe was expected to generate NIR enhanced emission with a large Stokes shift, which was an improved signal quantification¹⁰ due to avoidance of autofluorescence inference and detrimental crosstalk¹¹ between excited and emitted light. The NIR-TS probe was synthesized, and characterized by ¹H nuclear magnetic resonance (NMR) and ¹³C NMR spectroscopy and high-resolution mass spectrometry (Scheme S1 and Fig. S1, ESI†). Initially, the spectral response of the NIR-TS probe to SO₂ was investigated using sodium sulfite (Na₂SO₃) as the supply source of SO₂. Absorption of the NIR-TS probe peaked at 313 nm and 568 nm, which was attributed to 2-(benzothiazole-2-yl)-4-methylphenol and benzopyrylium groups, respectively. In turn, the NIR-TS probe showed emission at 653 nm when excited by 550 nm light (Fig. 1). The red-light emission enabled visualization of mitochondria targeted by the NIR-TS probe. After addition of Na₂SO₃ to the NIR-TS probe, its absorption at 568 nm decreased with a change from purple to colorless, and the 836 nm emission in the NIR region emerged and increased simultaneously (Fig. 1). The emission reached a plateau upon addition of Na₂SO₃ (90 μM) to the NIR-TS probe. Furthermore, a good linear relationship between the emission intensity and Na₂SO₃ concentration was found at 0.5–40 μM (Fig. S2C, ESI†). The limit of detection (LoD) was calculated to be 0.067 μM by the general 3σ/k method. Some reports have documented sulfite levels of 0–10 μM in the serum of healthy people, and (0–4 μM) in mitochondria.¹² Hence, the NIR-TS probe was suitable for sensitive monitoring of Na₂SO₃ levels in a biological system.

Fig. 1 Spectroscopic responses of the NIR-TS probe to Na₂SO₃. (A) Responses of UV-vis absorption spectroscopy of the NIR-TS probe (30 μM) towards Na₂SO₃ (100 μM); (B) fluorescent responses of the NIR-TS probe (10 μM) towards Na₂SO₃ (5 μM and 90 μM).

The NIR-TS probe had excellent photostability at pH 5–10 and had a relatively strong response to Na₂SO₃ at pH 8.0 (Fig. S2D, ESI†). The pH in mitochondria is ∼8, so the NIR-TS probe could detect SO₂ (and its derivatives) in mitochondria effectively. We also studied the time-dependent spectral response of the NIR-TS probe towards Na₂SO₃ (Fig. S3, ESI†). The fluorescent signal at 836 nm leveled off within 10 s, and reached 30-times that of the NIR-TS probe itself. This ultrafast and sensitive manner enabled the NIR-TS probe to monitor the levels of SO₂ (and its derivatives) in cells.

Subsequently, the selectivity of the NIR-TS probe towards Na₂SO₃ was evaluated by treatment with various biologically relevant species: sulfur-containing species, reducing agents, ROS, and other inorganic salts (Fig. S4, ESI†). The changes in fluorescence intensity caused by these competing species were negligible. Hence, the NIR-TS probe could monitor levels of SO₂ (and its derivatives) selectively in a complex physiological condition. To confirm further the reaction mechanism of the NIR-TS probe towards Na₂SO₃, mass spectrometry was conducted (Fig. S5, ESI†). Addition of Na₂SO₃ to a solution of NIR-TS revealed a prominent peak of [M − H]⁻ at m/z 587.1684, which corresponded to the adduct for NIR-TS towards Na₂SO₃ (calcd: 587.1680).

We also conducted quantum chemical calculations to rationalize the spectroscopic properties of NIR-TS. Before the reaction with SO₂, the first excited state of NIR-TS was dominated by the benzopyrylium moiety with red emissions (highlighted in red; Scheme 1a). This denoted the distribution of the corresponding electrons and holes residing in the benzopyrylium fragment (Scheme 1b and Fig. S6, ESI†). However, upon the reaction with SO₂, the π-conjugation of the benzopyrylium moiety was broken, along with a blue shift in the UV-vis absorption spectrum. Consequently, the distribution of electrons and holes of the first excited state of the newly formed adduct (E) shifted to the hydroxybenzothiazole and neighboring unit (highlighted in green; Scheme 1), thereby activating ESIPT. Indeed, further calculations showed that ESIPT was energetically favorable for adduct E, and ESIPT led to significantly red-shifted emissions from the ketone form K (Scheme 1 and Fig. S6, ESI†).

Scheme 1 (a) Design of the NIR-TS probe for subcellular SO₂ detection. (b) Calculated electron–hole distributions during photoexcitation of the NIR-TS probe in water before and after reactions with SO₂. The geometry of NIR-TS and K is based on excited-state optimized structures, whereas that of E is based on the ground-state optimized structure.

To explore the capabilities of the NIR-TS probe to image SO₂ (Na₂SO₃) in mitochondria, a cytotoxic assay of the NIR-TS probe was undertaken in HeLa cells (Fig. S7, ESI†). We discovered that >85% of cells were alive upon treatment with the NIR-TS probe (5 μM) for 10 h, which indicated the low cytotoxicity of the NIR-TS probe. Similar to other mitochondria-targetable probes with positive charges, enrichment of the NIR-TS probe in mitochondria was possible.¹³ To examine this possibility, colocalization experiments were carried out (Fig. S8, ESI†). Cells treated with the NIR-TS probe and the MitoTracker™ Green (mitochondrial stain) exhibited strong emissions in red and green channels. The merged image (Fig. S8c and f, ESI†) indicated that both channels overlaid very well. Pearson's colocalization coefficient was calculated to be 0.82, demonstrating that the NIR-TS probe could accumulate specifically in the mitochondria of living cells.

Next, the ability of the NIR-TS probe to image Na₂SO₃ in living cells was evaluated (Fig. S9, ESI†). For control cells incubated only with the NIR-TS probe, no fluorescence signals were captured in the NIR channel. For cells incubated sequentially with the NIR-TS probe and Na₂SO₃, concentration-dependent fluorescence signals were noted in the NIR channel. Hence, the NIR-TS probe could detect exogenous Na₂SO₃ with excellent sensitivity in living cells. Na₂S₂O₃ can be combined with glutathione via thiosulfate sulfurtransferase (TST) to produce endogenous SO₂ derivatives in mammals.¹⁴ To evaluate the ability of the NIR-TS probe to detect endogenous SO₂ derivatives, HeLa cells were imaged after treatment with NIR-TS and Na₂S₂O₃ (500 μM) (Fig. 2c). NIR fluorescence signals were activated due to the GSH-induced generation of SO₂. For confirmation, control cells were prepared by successive treatment of N-ethylmaleimide (NEM; thiol inhibitor), the NIR-TS probe and Na₂S₂O₃ for imaging. Control cells did not produce an obvious NIR fluorescence signal (Fig. 2d). Hence, the NIR-TS probe could work as an endogenous SO₂ sensor. Furthermore, lipopolysaccharide (LPS) can also produce low-level sulfite endogenously by inducing an inflammatory response in cells.¹⁵ NIR signals were observed when the NIR-TS probe was used to image HeLa cells pretreated with LPS (1 μg mL⁻¹) (Fig. 2e). However, the signals were lost for cells treated with LPS and formaldehyde (FA) due to the inhibition of sulfite by FA (Fig. 2f). Therefore, the NIR-TS probe could track the behavior of low-content SO₂ (and its derivatives) in cells.

Fig. 2 NIR fluorescence imaging of SO₂ (Na₂SO₃) in HeLa cells. (a) HeLa cells were incubated with the NIR-TS probe (10 μM) for 15 min. (b) HeLa cells were pre-treated with the NIR-TS probe and then incubated with exogenous Na₂SO₃ (50 μM) for 10 min. (c) HeLa cells were pretreated with the NIR-TS probe (10 μM), then incubated with Na₂S₂O₃ (500 μM) for 30 min. (d) The NIR-TS probe-stained HeLa cells were incubated with 500 μM NEM for 30 min, followed by Na₂S₂O₃ for another 30 min. (e) HeLa cells were treated with the NIR-TS probe, then incubated with 1 μg mL⁻¹ lipopolysaccharide. (f) HeLa cells were incubated with the NIR-TS probe, then FA (200 μM) and lipopolysaccharide (1 μg mL⁻¹) were added. Scale bar: 20 μm.

We wished to investigate the effect and levels of SO₂ during oxidative stress in mitochondria.¹⁶ HeLa cells were treated with LPS, and NIR fluorescent signals documented. However, when LPS-pretreated cells were incubated with 200 μM N-acetylcysteine (NAC; antioxidant) and imaged by the NIR-TS probe, NIR fluorescent signals were decreased obviously (Fig. 3). These results illustrated that SO₂ could have an anti-oxidative-stress effect in mitochondria, and that its levels increased during oxidative stress.

Fig. 3 NIR fluorescence images of HeLa cells. Cells were treated with the NIR-TS probe (10 μM); cells were treated with the NIR-TS probe (10 μM) and LPS; cells were treated with the NIR-TS probe, LPS and NAC (200 μM).

The concentrations used in the experiments mentioned above for Na₂S₂O₃ (500 μM) or LPS (1 μg mL⁻¹) were based on the literature.¹⁷ A certain amount of sulfite could enter cells in a short time and “lighten” the cells, so Na₂SO₃ (50 μM) was selected.

Furthermore, we investigated the capability of the NIR-TS probe for imaging SO₂ (and its derivatives) in a mouse model (Fig. 4). For mice injected with the NIR-TS probe alone, no NIR fluorescence signals emerged. NIR fluorescence signals appeared for NIR-TS probe-pretreated mice injected in situ with Na₂SO₃ (50 μM), Na₂S₂O₃ (500 μM) or LPS (1 μg mL⁻¹) (Fig. 4b, c and e). However, for control mice pretreated with NEM or FA, the fluorescence signals were reduced significantly (Fig. 4d and f). The results in mice models were consistent with those in cell experiments.

Fig. 4 NIR fluorescence imaging of Na₂SO₃ in mice. (a) NIR-TS probe (10 μM)-treated mice. (b), (c) and (e) NIR-TS ptobe (10 μM)-treated mice were also injected with Na₂SO₃ (50 μM), Na₂S₂O₃ and LPS, respectively. (d) As a control of (c), mice were treated with NEM, then Na₂S₂O₃ and the NIR-TS probe. (f) As a control of (e), mice were treated with LPS, then FA and the NIR-TS probe.

The SO₂ level was increased for LPS-treated mice. However, after mice had been treated with the antioxidant NAC, the SO₂ level decreased (Fig. S10, ESI†). Results showed that SO₂ had an anti-oxidative-stress effect in mice.

Results also showed that the NIR-TS probe provided a reliable approach for non-invasive visualization of SO₂ (and its derivatives) in a mouse model during oxidative stress.

In summary, we developed an NIR fluorescent probe which displayed excellent mitochondrial-targeting ability and which could be applied to SO₂ visualization during oxidative stress in HeLa cells as well as in a mouse model. The strategy for constructing the NIR-TS probe was different to the traditional design for NIR fluorophores. Also, we reassessed the unique merits of ESIPT mechanism in designing this NIR-TS probe.

This work was supported by the Natural Science Foundation of China (21775096, 21878180), One Hundred People Plan of Shanxi Province, Shanxi Province “1331 project” Key Innovation Team Construction Plan Cultivation Team (2018-CT-1), 2018 Xiangyuan County Solid Waste Comprehensive Utilization Science and Technology Project (2018XYSDJS-05), Shanxi Province Foundation for Returnees (2017-026), Shanxi Collaborative Innovation Center of High Value-added Utilization of Coal-related Wastes (2015-10-B3), Shanxi Province 2019 Annual Science and Technology Activities for Overseas Students Selected Funding Projects, Innovative Talents of Higher Education Institutions of Shanxi, Scientific and Technological Innovation Programs of Higher Education Institutions in Shanxi (2020L0541), Key R&D Program of Shanxi Province (201903D421069), Key R&D and Transformation Plan of Qinghai Province (2020-GX-101) and Scientific Instrument Center of Shanxi University (201512).

Conflicts of interest

None.

Notes and references

1. N. Sang, Y. Yun, H. Li, L. Hou, M. Han and G. K. Li, Toxicol. Sci., 2010, 114, 226–236.
2. Y. Sun, J. Liu, J. Zhang, T. Yang and W. Guo, Chem. Commun., 2013, 49, 2637–2639.
3. W. Chen, Q. Fang, D. Yang, H. Zhang, X. Song and J. Foley, Anal. Chem., 2015, 87, 609–616.
4. (a) Z. Meng, Inhalation Toxicol., 2003, 15, 181–195; (b) A. Kubo, H. Saji, K. Tanaka and N. Kondo, Plant Mol. Biol., 1995, 29, 479–489.
5. (a) R. Weissleder and M. J. Pittet, Nature, 2008, 452, 580–589; (b) A. Razgulin, N. Ma and J. Rao, Chem. Soc. Rev., 2011, 40, 4186–4216; (c) L. You, D. Zha and E. V. Anslyn, Chem. Rev., 2015, 115, 7840–7892; (d) Y. Wen, F. Huo and C. Yin, Chin. Chem. Lett., 2019, 30, 1834–1842.
6. (a) X. Yang, Y. Zhou, X. Zhang, S. Yang, Y. Chen, J. Guo and X. Li, Chem. Commun., 2016, 52, 10289–10292; (b) X. Yang, W. Liu, J. Tang, P. Li, H. Weng, Y. Ye and M. Xian, Chem. Commun., 2018, 54, 11387–11390; (c) D. Li, Z. Wang, X. Cao, J. Cui, X. Wang, H. Z. Cui, J. Miao and B. X. Zhao, Chem. Commun., 2016, 52, 2760–2763.
7. W. Zhang, F. Huo, Y. Zhang and C. Yin, J. Mater. Chem. B, 2019, 7, 1945–1950.
8. (a) M. Cui, M. Ono, H. Watanabe, H. Kimura, B. Liu and H. Saji, J. Am. Chem. Soc., 2014, 136, 3388–3394; (b) X. Wang, Z. Guo, S. Zhu, H. Tian and W. H. Zhu, Chem. Commun., 2014, 50, 13525–13528; (c) W. Chen, S. Xu, J. J. Day, D. Wang and M. Xian, Angew. Chem., 2017, 129, 16838–16842; (d) V. R. Mishra, C. W. Ghanavatkar and N. Sekar, ChemistrySelect, 2020, 5, 2103–2113.
9. V. R. Mishra, C. W. Ghanavatkar and N. Sekar, ChemistrySelect, 2020, 5, 2103–2113.
10. (a) H. Wen, Q. Huang, X. F. Yang and H. Li, Chem. Commun., 2013, 49, 4956–4958; (b) X. F. Yang, Q. Huang, Y. Zhong, Z. Li, H. Li, M. Lowry, J. O. Escobedo and R. M. Strongin, Chem. Sci., 2014, 5, 2177–2183; (c) Q. Wang, L. Zhou, L. Qiu, D. Lu, Y. Wu and X. B. Zhang, Analyst, 2015, 140, 5563–5569; (d) V. S. Patil, V. S. Padalkar, A. B. Tathe and N. Sekar, Dyes Pigm., 2013, 98, 507–517.
11. (a) Y. Liu, K. Li, K. Xie, L. Li, K. Yu, X. Wang and X. Q. Yu, Chem. Commun., 2016, 52, 3430–3433; (b) E. Karakuş, M. Üçüncü and M. Emrullahoğlu, Anal. Chem., 2016, 88, 1039–1043.
12. (a) Y. Liu, K. Li, K. X. Xie, L. L. Li, K. K. Yu, X. Wang and X. Q. Yu, Chem. Commun., 2016, 52, 3430–3433; (b) A. J. Ji, S. R. Savon and D. W. Jacobsen, Clin. Chem., 1995, 41, 897–903.
13. Y. Ma, Y. Tang, Y. Zhao, S. Gao and W. Lin, Anal. Chem., 2017, 89, 9388–9393.
14. H. Mitsuhashi, S. Yamashita, H. Ikeuchi and T. Kuroiwa, Shock, 2005, 24, 529–534.
15. J. J. Zhang, Y. X. Fu, H. H. Han, Y. Zang, J. Li, X. P. He, B. L. Feringa and H. Tian, Nat. Commun., 2017, 8, 987–996.
16. J. Z. Li, Y. H. Sun, C. Y. Wang, Z. Q. Guo, Y. J. Shen and W. H. Zhu, Anal. Chem., 2019, 91, 11946–11951.
17. W. Chen, Q. Fang, D. Yang, H. Zhang, X. Song and J. Foley, Anal. Chem., 2016, 88, 4426–4431.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/d0cc07398a

---