Rational Design of a Dual-Reactivity Based Fluorescent Probe for Visualizing Intracellular HSNO

Abstract

Thionitrous acid (HSNO), the smallest S-nitrosothiol, is emerging as a potential key intermediate in cellular redox regulation linking two signaling molecules H₂S and NO. However, the chemical biology of HSNO remains poorly understood. A major hurdle is the lack of methods for selective detection of HSNO in biological systems. Here we report the rational design, synthesis, and evaluation of the first fluorescent probe TAP-1 for HSNO detection. TAP-1 showed high selectivity and sensitivity to HSNO in aqueous media and cells, providing a useful tool for understanding the functions of HSNO in biology.

Graphical Abstract

Light-up intracellular HSNO: A unique dual-reactivity based fluorescent probe for HSNO was developed. The probe exhibited high selectivity and sensitivity to HSNO in aqueous media and cells, providing a useful tool for understanding the functions of HSNO in biology.

S-Nitrosothiols (RSNO) belong to the reactive sulfur species (RSS) family, which includes biothiols, sulfenic acids (RSOH), hydrogen sulfide (H₂S), polysulfides, and persulfides (RSSH). RSNO play vital roles in the storage and transport of nitric oxide (NO) as well as in achieving its regulatory function via protein S-nitrosation.^[1–3] RSNO can be generated by the reaction of nitrosating agents with thiols, and then transported or diffused to the site of action to exert their regulatory function.^[1–3] S-Nitrosoglutathione (GSNO) and S-nitrosocysteine (CysNO) are often used to investigate the mechanism of cellular RSNO formation and transport. However, they require the assistance of a specific thiol or transporter.^[4,5] Recently, thionitrous acid (HSNO), the smallest RSNO, has received particular attention as a possible H₂S-derived signaling molecule.^[6–22] Evidences suggest that HSNO is a potential key intermediate in cellular redox regulation linking H₂S and NO. Unlike larger RSNO, HSNO is believed to be able to freely diffuse through membranes to reach its intracellular targets so to facilitate transnitrosation of proteins.^[6] The facile diffusion ability of HSNO could be conducive to understand the regulated mechanisms for RSNO formation and the process of signaling propagation. Meanwhile these studies raise interesting questions regarding the formation of HSNO. It has recently been proposed that HSNO is formed through the reaction H₂S with RSNO or NO.^[6–23] Thus the role of H₂S in modulating intracellular RSNO or NO may be related to HSNO generation. In addition, the formation of HSNO may further produce SSNO^-[18–22] and this topic remains controversial.

In order to better understand the contribution of HSNO in biological settings, it is important to study the fundamental chemistry/reactivity of HSNO and develop methods for its specific detection. Previously FTIR, ¹⁵N NMR and mass spectrum have been used to detect and characterize HSNO under physiologically relevant conditions.^[6] However, these methods are not applicable for real-time and visual detection in biological samples. In this regard fluorescent probes can be very useful due to their high sensitivity and spatiotemporal resolution capability. Unfortunately, there is no report on such fluorescent probes for HSNO so far. To this end, we have initiated a program to study new reactions of HSNO. Since 2009, our laboratory has developed a series of specific reactions which can convert unstable RSNO to stable and detectable species.^[23] While HSNO is highly reactive and unstable, we believe if a specific reaction recognizing HSNO can be developed, such a reaction should be useful for developing HSNO fluorescent probes. Herein, we report the rational design, synthesis, and evaluation of a fluorescent probe for HSNO detection.

It is known that nucleophiles can attack either the nitrogen or sulfur atom of RSNO.^[23] The pathway is usually dependent on the substrate. For instance, some studies showed that the nitrogen atom of RSNO was attacked by arylamines to form aryl diazonium.^[24] This indicates that RSNO may react with o-phenylenediamine to generate benzotriazole in a similar manner to NO under aerobic conditions.^[25] We envisioned that HSNO should have such a reaction due to its high reactivity. On the other hand, the sulfur atom of HSNO could be attacked by an appropriate nucleophile, like 2-mercaptobenzoate, to form an –SH containing intermediate and then the –SH residue would subsequently trigger a tandem reaction.^[26] In this process HSNO acts as sulfane sulfur. Such a dual-reactivity is unique for HSNO as other species like NO or other RSNO won’t have similar reactivity. Therefore, fluorescent probes based on this dual reactivity are possible.

With this idea in mind, we proposed a fluorescent probe (Thionitrous Acid Probe, TAP-1) for HSNO detection (Scheme 1). In this design, 2-mercaptobenzoate and o-phenylenediamine were used as reaction sites to modify a fluorescein-based fluorophore. The fluorescence of TAP-1 should be completely quenched via the protection of hydroxyl group of the fluorophore and the intramolecular spirocyclization effect.^[27] Therefore, TAP-1 should bear very low background fluorescence that is favorable for high sensitivity. With HSNO presented, the 2-mercaptobenzoate moiety should undergo nucleophilic reaction and intramolecular cyclization to release benzodithiolone and compound 2. Meanwhile, HSNO should also react with the o-phenylenediamine unit to form benzotriazole derivative 3. Then, the conjugation system of the fluorophore will be restored and 3 should show strong fluorescence. Of note, polysulfides or RSSH alone may only react with 2-mercaptobenzoate moiety to induce -OH deprotection of the fluorophore and the resulting product 2 should show no fluorescence or weak fluorescence due to intramolecular spirocyclization and the photo-induced electron transfer (PeT) process.^[25,27] As such, TAP-1 should be a specific sensor for HSNO.

Scheme 1.

Design of dual-reactivity based probe TAP-1.

The synthesis of the probe TAP-1 is shown in Scheme 2. Briefly, compound 4 was coupled with N-Boc-1,2-phenylenediamine and then treated with LiOH for deacylation to form 5. The subsequent amide reduction with borane followed by chloranil oxidation provided 6. Esterification of 6 with 2-(2-pyridinyldithio)-benzoic acid gave 7, which was finally subjected to Boc deprotection and disulfide reduction to furnish the desired probe TAP-1.

Scheme 2.

Synthesis of TAP-1. Reagents and conditions: (i) SOCl₂, Et₃N, N-Boc-1,2-phenylenediamine, rt, 1 h; (ii) LiOH, THF, H₂O, rt, 4h; (iii) BH₃/THF, reflux, 4 h; (iv) chloranil, CH₂Cl₂, rt, 2 h; (v) 2-(2-pyridinyldithio)benzoic acid, EDC, DMAP, CH₂Cl₂, rt, 3 h; (vi) HCl gas, CH₂Cl₂, rt, 2 h; (vii) PPh₃, THF/H₂O, rt, 1 h.

With TAP-1 in hand, we tested its fluorescence properties and responses to HSNO. In these experiments freshly prepared mixture of 1 mM GSNO and 0.3 mM Na₂S in 50 mM pH 7.4 PBS buffer (hereinafter referred to as “HSNO solution”) was used as an easy-to-handle HSNO source.^[6] As expected, TAP-1 showed almost no fluorescence emission (Φ = 0.01) in PBS. In the presence of 6-fold diluted HSNO solution (50 μM), the probe gave dramatic fluorescence turn-on response at 521 nm. The maximum emission intensity of TAP-1 reached a plateau within 8 min (Figure 1A). The pseudo-first-order rate constant for the reaction was determined as 0.66 min⁻¹. To further validate the selectivity of TAP-1 for HSNO, TAP-1 was separately treated with a series of other reactive species including glutathione (GSH), cysteine (Cys), homocysteine (Hcy), Na₂S, Na₂SO₃, MCPD (a persulfide donor),^[28] Na₂S₂, Na₂S₃, GSNO, pyrrolidine-NONOate (a NO donor), Angeli’s salt (a HNO donor), peroxynitrite (ONOO-), tert-butyl nitrite (t-BuONO), NaNO₂, KNO₃. As shown in Figure 1B, these molecules did not trigger significant fluorescence increase. Recent studies suggested that HSNO could come from the interaction of H₂S with NO. We then applied TAP-1 in detecting in situ generated HSNO from H₂S and NO. When Na₂S was premixed with pyrrolidine-NONOate in PBS solution, addition of TAP-1 produced obvious fluorescence signals, suggesting the possible formation of HSNO in this system, and its formation efficiency may be dependent on the concentration of NO and H₂S. Moreover, we have developed a fluoride-triggered O-silyl-mercaptan-based HSNO donor,^[29] TAP-1 also showed obvious fluorescence response to HSNO generated from this system (Figure S1). It should be noted that in theory TAP-1 can be turned on if NO and sulfane sulfide (like H₂S₂) co-exist. We have tested this scenario. As shown in Figure S2, when TAP-1 was treated with a mixed solution of NO and H₂S₂ (both at 50 μM), we indeed observed some fluorescence increase, but at a much lower level compared to HSNO. These results suggest that NO and H₂S₂ can rapidly react with each other (much faster than their separated reactions with TAP-1). The reaction between NO and persulfides like H₂S₂ is rather complicated. It can produce many possible products including HSNO, HSSNO (or dithionitrate/dithionitric acid), H₂S, HNO, S₈, etc,^[16] and therefore, decreasing the turn-on efficiency of TAP-1. Taken together, these results demonstrated TAP-1 could be used for selective and sensitive detection of HSNO in aqueous media.

Figure 1.

(a) Turn-on fluorescence response of 10 μM TAP-1 to 6-fold diluted HSNO solution at different reaction time (0, 1, 2, 4, 6, 8, 10 min for curves 1–7, respectively); (b) Fluorescence intensity at 521 nm of 10 μM TAP-1 in the presence of various reactive species: 1) blank; 2) 10 mM GSH; 3) 1 mM Cys; 4) 1 mM Hcy; 5) 200 μM Na₂S; 6) 100 μM Na₂SO₃; 7) 100 μM MCPD; 8) 100 μM Na₂S₂; 9) 100 μM Na₂S₃; 10) 1 mM GSNO; 11) 200 μM pyrrolidine-NONOate; 12) 100 μM Angeli’s salt; 13) 100 μM ONOO⁻; 14) 100 μM t-BuONO; 15) 1 mM NaNO₂; 16) 1 mM KNO₃; 17) 50 μM Na₂S + 50 μM pyrrolidine-NONOate; 18) HSNO solution (50 μM).

To clarify the fluorescence turn-on mechanism of TAP-1 by HSNO, the reaction between TAP-1 and in situ generated HSNO from GSNO and Na₂S was studied. As expected, ben-zodithiolone and benzotriazole derivative 3 were isolated with moderate yields (40%). The photophysical properties of 3 also were tested in 50 mM pH 7.4 PBS buffer solution (Figure S3). It showed strong fluorescence (Φ = 0.79) with a maximal emission at 521 nm and the maximal absorption at 499 nm. These results confirmed the HSNO-mediated fluorescence turn-on mechanism proposed in Scheme 1.

Encouraged by the above results, we wondered if TAP-1 could be used for imaging HSNO in cells. First, cytotoxity of TAP-1 was tested by CCK-8 assays (Figure S4), which demon-strated that TAP-1 exhibited low cytotoxicity and good bio-compatibility. The ability of TAP-1 to visualize intracellular HSNO was then evaluated. As shown Figure 2, HeLa cells were incubated with TAP-1 (10 μM) for 30 min and then extracellular probe was washed off. No significant fluorescence was observed in cells. However, when cells were subsequently treated with the mixture of Na₂S/GSNO for 25 min, strong fluorescence was observed (Figure 2d). When cells were treated with the mixture solution of Na₂S/pyrrolidine-NONOate, similar effects were noted (Figure 2e). Neither Na₂S nor GSNO alone gave noticeable fluorescence. Intriguingly, a modest fluorescence response was observed if cells were treated with pyrrolidine-NONOate (Figure 2f). This may be due to the reaction of endogenous H₂S with NO to generate HSNO. However, when HeLa cells were pretreated with a H₂S biosynthesis inhibitor propargylglycine (PAG), an obviously attenuated fluorescence was observed (Figure 2g). These results indicated TAP-1 had good cell permeability and could be used for the detection of intracellular HSNO.

Figure 2.

Fluorescence images of TAP-1 in HeLa cells. (a-f) Cells were incubated with 10 μM TAP-1 for 30 min, washed and subjected to different treatments for 25 min. a) Controls (only FBS-free media); b) 150 μM Na₂S; c) 1 mM GSNO; d) 100 μM HSNO solution; e) the mixture solution of 150 μM pyrrolidine-NONOate and 150 μM Na₂S; f) 150 μM pyrrolidine-NONOate. (g) Cells were pre-treated with 1 mM PAG for 1.5 h, washed and subjected to same treatment with f). (h) Mean fluorescence intensities of images a-g. Scale bar represents 50 μm for all images.

To further confirm the efficiency of TAP-1 in visualizing intracellular HSNO, we evaluated the fluorescence imaging of HSNO endogenously generated in human embryonic kidney (HEK293T) cells. Recently, we (T. Akaike) found the in vivo production of reactive sulfide species such as cysteine persulfide was mediated by cysteinyl-tRNA synthetase (CARS), which was eventually functioning as cysteine persulfide synthase (CPERS).^[30] The mitochondrial form of CARS (CARS2) is the major source of persulfides in cells. It should be noted that H₂S and persulfides always co-exist in equilibrium in the presence of cellular thiols (such as Cys and GSH).^[31] As such, CARS is contributing to the endogenous H₂S production. In HEK293T cells, this endogenously formed H₂S was reacted with NOC7, a NO donor, to form HSNO in situ. As shown in Figure 3, much stronger fluorescence response of TAP-1 was observed in wild type (WT) HEK293T cells treated with NOC7 than in CARS2 knockout (KO) cells, and this response was NOC7 concentration-dependent. These results suggest that the TAP-1 can react effectively with endogenous HSNO, which is produced through the reaction of CARS2/CPERS-derived H₂S with NO.

Figure 3.

(a) Fluorescence images of TAP-1 for endogenous HSNO in HEK293T cells. WT and CARS2 KO HEK293T cells were incubated with 10 μM of TAP-1 for 30 min, washed, and treated with 10, 100 and 300 μM of NOC7. (b) Mean fluorescence intensities in panel a. Scale bar represents 50 μm for all images. Values are means ± s.d. (n = 3). ***p < 0.001 (vs. WT HEK293T), two-way ANOVA followed by Tukey-Kramer test.

In summary, we reported in this study a new approach for HSNO detection through a dual-reactivity based fluorescent probe TAP-1. This probe showed high selectivity and sensitivity to HSNO in aqueous media and cells. It is expected to serve as a useful tool in understanding HSNO functions in biological systems.