A versatile fluorescent probe for hydrogen peroxide in serotonergic neurons of living brains of mice with depression

Abstract

Depression, a prevalent mental illness, is intricately linked with the neurotransmitters in the brain, while serotonin as a crucial regulator of mood, energy levels, and memory, has been implicated in depression. So, the release of serotonin by serotonergic neurons plays a significant role in the development of depression. Notably, the foremost marker of oxidative stress, hydrogen peroxide (H₂O₂), can interfere with the functioning of serotonergic neurons and potentially contribute to depression. Investigating the impact of H₂O₂ on serotonergic neurons could offer valuable insights into the mechanisms underlying depression. However, there have been no effective tools for selectively imaging H₂O₂ in these neurons so far. To address this gap, we created a small molecular fluorescent probe, PF-H₂O₂, designed specifically for imaging H₂O₂ in serotonergic neurons under oxidative stress. PF-H₂O₂ exerts excellent serotonergic neuron-targetability and notable selectivity for H₂O₂. Furthermore, we discovered increased H₂O₂ in serotonergic neurons of mice with depressive symptoms. Altogether, this endeavour unveils a pioneering tool for exploring pathophysiology linked to serotonergic neuronal dysfunction.

---

Introduction

Depression is a widespread and often recurrent mental health disorder, with a global prevalence rate of 4.4%.¹ Major depressive disorder (MDD) significantly affects various aspects of life, including educational attainment, interpersonal relationships, and occupational functioning.² Additionally, it may be associated with comorbid conditions such as obesity and cardiovascular diseases, as well as an increased risk of premature mortality, including suicide.³ Nevertheless, the etiology of depression is multifaceted, and elucidating its pathogenesis remains a significant challenge.

The brain exhibits a pronounced sensitivity to oxidative stress due to its high lipid content and extensive oxygen metabolism.⁴ Oxidative stress refers to a pathological imbalance resulting from the accumulation of reactive oxygen species (ROS) within the body, which surpasses the capacity of endogenous antioxidant defenses.^5,6 This imbalance can lead to cellular damage and functional impairments, with neurons being particularly vulnerable to such detrimental effects.⁷ Neuronal dysfunction is one of the important pathological mechanisms of brain diseases such as depression. Serotonergic neurons are responsible for the synthesis and release of serotonin (5-HT), which regulates mood, sleep, and memory, and their dysfunction is strongly linked to depression.^8,9 Hydrogen peroxide (H₂O₂) is an important intracellular redox signaling molecule that is normally involved in cell signaling and functional regulation.¹⁰ However, elevated H₂O₂ levels caused by oxidative stress can damage DNA, proteins, and lipids, ultimately leading to cell apoptosis.¹⁰ In the nervous system, excessive accumulation of H₂O₂ may lead to dysfunction of neurons such as serotonergic ones, which can induce or worsen depression. Therefore, understanding the changes in H₂O₂ levels in serotonergic neurons during the occurrence and development of depression is crucial for comprehending the pathological mechanism of depression.

Currently, several methods are available to detect H₂O₂ in the brain. For example, Kasai et al. designed an electrochemical sensor to monitor H₂O₂ levels in rat hippocampal slices,¹¹ while Michael et al. employed an electrical current sensor to track H₂O₂ concentrations in the environment surrounding brain cells.¹² However, these methods require in vitro testing or minimally invasive surgery on living brains and cannot simultaneously provide temporal and spatial information about hydrogen peroxide. In contrast, fluorescence imaging offers advantages such as high sensitivity, high resolution, non-invasiveness, and real-time in situ monitoring, making it a powerful tool for the real-time detection of H₂O₂.^13–18 For example, Chang et al. developed a borate fluorescent probe to detect changes in H₂O₂ concentration in living cells, including hippocampal neurons, using confocal microscopy and two-photon microscopy.¹⁹ Additionally, they created an H₂O₂ sensor that visualizes cellular redox signaling in microglia–neuron co-culture cell models.²⁰ However, effective fluorescent probes for the real-time and in situ detection of H₂O₂ in serotonergic neurons are still lacking.

Therefore, in order to reveal the changes in H₂O₂ levels in brain serotonergic neurons, we have designed and synthesized a novel fluorescent probe (PF-H₂O₂) that specifically detects H₂O₂ in serotonergic neurons. The probe is divided into 3 parts: (1) since the serotonin transporter (SERT) is a transmembrane protein unique to serotonergic neurons,^21,22 we used paroxetine, a specific inhibitor of SERT,²³ as the specific targeting group of serotonergic neurons. (2) We used fluorescein with a high quantum yield as the fluorophore. (3) We employed phenylborate as the H₂O₂ recognition group (Fig. 1). The incorporation of phenylborate groups at the 3′ and 6′ positions of a xanthenone scaffold would compel this platform to assume a closed, colorless, and non-fluorescent lactone configuration. Upon exposure to H₂O₂, the hydrolytic deprotection of the phenylborate would result in the formation of an open, colored, and fluorescent fluorescein product (Fig. 1). Using PF-H₂O₂, we visualized changes in the levels of H₂O₂ in serotonergic neurons and found elevated levels of H₂O₂ in serotonergic neurons in depressed mouse brains through in vivo imaging.

Fig. 1 Structure and recognition mechanism of PF-H₂O₂.

Experimental section

Apparatus and reagents

Details regarding the apparatus and reagents used in this study can be found in the ESI.†

Synthesis of PF

5-Carboxyfluorescein (0.3763 g, 1 mmol) was dissolved in 6 ml N,N-dimethylformamide (DMF), with DMAP (0.122 g, 1 mmol) and DCC (0.26 g, 1 mmol) and stirred for 30 min. Paroxetine (0.3748 g, 1 mmol) was then added and stirred at room temperature for 24 h. The compound PF (0.42 g, 60% yield) was purified by thin layer chromatography. PF: ¹H NMR (400 MHz, DMSO-d₆) δ 8.03–7.87 (m, 1H), 7.47–7.24 (m, 2H), 7.13 (dt, J = 8.8, 4.1 Hz, 2H), 6.57 (dd, J = 9.3, 2.6 Hz, 1H), 6.12–5.98 (m, 2H), 5.98–5.85 (m, 3H), 5.82 (d, J = 7.9 Hz, 3H), 3.58 (s, 1H), 3.16 (s, 3H), 2.89 (s, 2H), 2.73 (s, 2H), 1.75–1.68 (m, 4H), 1.66–1.55 (m, 4H).¹³C NMR (101 MHz, DMSO-d₆) δ 181.20, 180.85, 162.78, 157.89, 157.87, 148.34, 141.66, 124.12, 124.12, 123.31, 115.78, 115.78, 115.56, 103.58, 102.97, 101.45, 52.74, 49.05, 47.97, 40.61, 40.40, 40.19, 39.98, 39.77, 39.56, 39.36, 36.25, 33.81, 31.75, 31.24, 31.24, 29.58, 29.48, 29.16, 29.04, 28.90, 27.02, 25.81, 25.71, 25.57, 25.39, 24.98, 24.94, 22.56, 14.42. HRMS (ESI), m/z calculated for C₄₀H₃₀FNO₉ [M]⁻ 686.1821, found: 686.1931.

Synthesis of PF-H₂O₂

0.3435 g (0.5 mM) of PF was dissolved in 5 mL of DMF, 0.2969 g (1 mM) of 4-bromomethylphenylboronic acid pinacol ester, and 0.65 g (2 mM) of cesium carbonate at 80 °C for 18 h, and purified by TLC to obtain PF-H₂O₂ (0.18 g, 0.18 g), yield 32%. PF-H₂O₂: ¹H NMR (400 MHz, DMSO-d₆) δ 8.08 (s, 1H), 7.96 (s, 3H), 7.80 (d, J = 7.8 Hz, 2H), 7.74 (d, J = 8.0 Hz, 4H), 7.63 (d, J = 8.0 Hz, 4H), 7.33 (d, J = 8.0 Hz, 6H), 7.27 (d, J = 8.0 Hz, 3H), 5.76 (s, 2H), 5.21 (dt, J = 32.9, 5.7 Hz, 4H), 4.59–4.46 (m, 6H), 3.17 (d, J = 5.2 Hz, 2H), 2.89 (s, 1H), 2.74 (s, 1H), 1.29 (s, 24H).¹³C NMR (101 MHz, DMSO-d₆) δ 162.79, 157.22, 148.27, 146.56, 146.56, 141.65, 134.73, 127.41, 126.13, 101.44, 83.97, 74.02, 63.17, 56.46, 47.93, 40.67, 40.62, 40.46, 40.41, 40.20, 39.99, 39.78, 39.57, 39.36, 36.26, 33.81, 31.24, 29.49, 29.16, 26.16, 25.82, 25.44, 25.16, 24.91, 19.03. HRMS (ESI), m/z calculated for C₆₅H₆₂B₂FNO₁₃ [M]⁺ 1120.4641, found: 1120.4799.

Results and discussion

To test the ability of PF-H₂O₂ to detect H₂O₂, we investigated the photophysical properties of the probe. First, we tested the UV absorption spectra and fluorescence emission spectra of PF-H₂O₂ under simulated physiological conditions (10 mM, pH 7.4, phosphate buffered saline). As shown in Fig. 2a, there is a weak absorption peak near 490 nm, and when 50 μM H₂O₂ is added, the absorption peak at 490 nm increases significantly (about 8.4-fold). In Fig. 2b, in the absence of H₂O₂, only weak fluorescence emission near 525 nm can be observed at 490 nm excitation. However, the addition of 50 μM H₂O₂ resulted in an enhancement of fluorescence at 525 nm (about 8-fold). The fluorescence intensity at 525 nm increases linearly with the increase in H₂O₂ concentration (0–50 μM), and the linear equation is represented as F = 33.60[H₂O₂] (μM) + 42.39 (R² = 0.998). The detection limit of PF-H₂O₂ for H₂O₂ was found to be 1.1 μM (Fig. 2c and d). The above results show that PF-H₂O₂ can sensitively detect H₂O₂. Subsequently, we tested the photostability of PF-H₂O₂ and fluorescence intensity vs. the reaction time of PF-H₂O₂. As shown in Fig. 2e and f, the fluorescence intensity of PF-H₂O₂ reached a plateau after reacting with H₂O₂ for 50 minutes and demonstrated excellent light stability.

Fig. 2 Optical properties of PF-H₂O₂. (a) Absorption spectra of PF-H₂O₂ before (black) and after (red) reaction with 50 μM H₂O₂. (b) Fluorescence excitation and emission spectra before (black) and after (red) reaction of PF-H₂O₂ with 50 μM H₂O₂. (c) Fluorescence spectra of PF-H₂O₂ after reaction with H₂O₂ (0–50 μM). (d) Linear relation of fluorescence intensity and H₂O₂ concentration. (e) Photostability experiment of PF-H₂O₂ (f) Plots of fluorescence intensity vs. the reaction time of the probe PF-H₂O₂ (25 μM) with 50 μM H₂O₂ (λ_ex/em = 490 nm/525 nm) in modified PBS buffer solution. (g) Fluorescence response of PF-H₂O₂ (25 μM) with ROS [1: 50 μM H₂O₂ 2: blank-only PF-H₂O₂ 3: 2 mM NO 4: 100 μM TBHP 5: 100 μM NaClO 6: 2 mM ˙OH 7: 1 mM GSH 8: 10 μM ¹O₂ 9: 2 μM O₂˙⁻]. (h) Fluorescent response of PF-H₂O₂ (25 μM) to amino acids and neurotransmitters. [1: 50 μM H₂O₂ 2: blank-only PF-H₂O₂ 3: 100 μM Thr 4: 100 μM Phe 5: 100 μM Gln 6: 100 μM lle 7: 100 μM Pro 8: 100 μM Met 9: 100 μM Ala 10: 100 μM Glu 11: 100 μM Val 12: 100 μM DA 13: 100 μM 5-HT 14: 100 μM NE]. (i) Fluorescence response of PF-H₂O₂ (25 μM) with metal ions. [1: 50 μM H₂O₂ 2: blank-only PF-H₂O₂ 3: 1 mM K⁺ 4: 1 mM Mg²⁺ 5: 1 mM Na⁺ 6: 1 mM Zn²⁺ 7: 1 mM Fe²⁺ 8: 1 mM Fe³⁺ 9: 1 mM Cu⁺ 10: 1 mM Cu²⁺ 11: 1 mM Ca²⁺]. All spectra were acquired in 40 mM PBS (pH 7.4) at λ_ex = 490 nm using an FLS-980 Edinburgh fluorescence spectrometer. Similar results were observed in five independent experiments. The data are expressed as the mean ± SD.

To evaluate the specific and selective recognition of H₂O₂ by PF-H₂O₂, we investigated the fluorescence reaction of PF-H₂O₂ with other competing molecules, such as reactive oxygen species, metal ions, and amino acids. As shown in Fig. 2g–l, PF-H₂O₂ exhibits weak fluorescence in the presence of various reactive oxygen species, metal ions, amino acids, and neurotransmitters. Only after the addition of H₂O₂ was the fluorescence intensity significantly enhanced (about 8-fold), demonstrating its specific recognition of H₂O₂. These experiments indicate that PF-H₂O₂ can detect H₂O₂ with high sensitivity and selectivity in vitro and has potential applications in cells and diseased tissues.

In order to evaluate whether PF-H₂O₂ can be applied in cell experiments, we used cell-counting-kit-8 (CCK8) assay to evaluate its cytotoxicity (Fig. S2, ESI†) and obtained an IC₅₀ value of 2.913 × 10⁷ μM. These results indicate that PF-H₂O₂ has low biotoxicity and good biocompatibility, making it suitable for subsequent cell experiments. To know when the reaction between PF-H₂O₂ and intracellular H₂O₂ was complete, we tested the change in the fluorescence intensity of PF-H₂O₂ in cells over time. As illustrated in Fig. S3 (ESI†), the fluorescence intensity of PF-H₂O₂ reached a plateau within one hour. Therefore, for the subsequent cell experiment, the probe's incubation time was set to one hour.

In the brain, the serotonin transporter (SERT) is a transmembrane protein specific to serotonergic neurons.⁹ Therefore, we utilized SERT-expressing HEK293 cells to mimic serotonergic neurons. Initially, HEK293 cells were transfected with the hsert-pcDNA3 plasmid to induce SERT expression, resulting in hsert-HEK293 cells. Western blot analysis was conducted to extract proteins from both HEK293 cells and hsert-HEK293 cells. As depicted in Fig. S4 (ESI†), in comparison with untreated HEK293 cells, hsert-HEK293 cells exhibited significantly higher SERT expression. Research has demonstrated that elevated glutamate levels can lead to neuronal dysfunction, consequently triggering oxidative stress and causing an overproduction of ROS.²⁴ Hence, in our experiments, we employed high concentrations of glutamate (10 mM) to stimulate cells and simulate oxidative stress environments.

We initially investigated whether PF-H₂O₂ could detect changes in H₂O₂ levels in hsert-HEK293 cells. We stimulated the hsert-HEK293 cells with 10 mM glutamate and then incubated them with 25 μM PF-H₂O₂. As illustrated in Fig. 3a, the fluorescence intensity of hsert-HEK293 cells stimulated by glutamate (glutamate) was higher than (about 1.54-fold) that of hsert-HEK293 cells not stimulated by glutamate (control), indicating that PF-H₂O₂ could sensitively detect changes in H₂O₂ levels in hsert-HEK293 cells. To demonstrate that the fluorescence change was a result of the fluctuation in H₂O₂ levels within the cell, we employed N-acetylcysteine (NAC, a direct scavenger of H₂O₂) to clear H₂O₂.^25–27 In comparison with glutamate-stimulated hsert-HEK293 cells, the fluorescence intensity of hsert-HEK293 cells with the addition of NAC to clear H₂O₂ decreased (Fig. 3a, about 0.65-fold), further confirming that PF-H₂O₂ could effectively detect changes in H₂O₂ levels in hsert-HEK293 cells. Next, we investigated whether PF-H₂O₂ could specifically enter cells via SERT to image H₂O₂. Similar to the above treatment, we treated HEK293 and hsert-HEK293 cells with glutamate and PF-H₂O₂, respectively. The results showed (Fig. 3a) that compared with HEK293 cells that did not express SERT, the fluorescence intensity of hsert-HEK293 cells that could express SERT was increased (about 2.71-fold), indicating that PF-H₂O₂ could target SERT-expressing cells and effectively image H₂O₂ in cells.

Fig. 3 The targeting ability of PF-H₂O₂ to SERT. (a) Fluorescence imaging of HEK293 and hsert-HEK293 cells by PF-H₂O₂. Control group: 25 μM PF-H₂O₂ incubated for 1 h, glutamate group: 10 mM glutamate incubated for 12 h, 25 μM PF-H₂O₂ treated cells for 1 h, glutamate + NAC group: 10 mM glutamate incubated for 12 h, 1 mM NAC incubated for 1 h, 25 μM PF-H₂O₂ incubated for 1 h, and 480 nm stimulated images were obtained, scale = 50 μm. (b) Confocal fluorescence images of H₂O₂. Cells were incubated with 25 μM PF-H₂O₂ for 1 h (control). The cells were incubated with 10 mM glutamate for 12 h and 25 μM PF-H₂O₂ for 1 h (glutamate). Incubation with 10 mM glutamate for 12 h, incubation with 78 μg mL⁻¹ fluoxetine for 2 h, and incubation with 25 μM PF-H₂O₂ for 1 h (fluoxetine). The cells were incubated with 10 mM glutamate for 12 h, then with 60 μg mL⁻¹ paroxetine for 2 h, and with 25 μM PF-H₂O₂ for 1 h (paroxetine), scale = 100 μm. (c) Fluorescence imaging of hsert-HEK293 cells, scale = 50 μm. PF: the green channel is the fluorescent signal emitted by the dye PF. Images acquired by excitation at 480 nm. Fluorescence emission windows: λ_em = 510–540 nm. SERT: The red channel is the fluorescent signal emitted by the serotonin transporter antibodies. Images acquired by excitation at 560 nm. Fluorescence emission windows: λ_em = 570–590 nm. (d) Relative fluorescence intensity output of a. (e) Relative fluorescence intensity output of b. Similar results were observed in five independent experiments. **P < 0.01 and ***P < 0.001 is compared with the control group.

Next, we further verified the specificity of PF-H₂O₂ for SERT-expressing cells. Paroxetine and fluoxetine, which are SERT inhibitors, can inhibit the transport capacity of SERT.²⁸ Therefore, we incubated hsert-HEK293 cells with 60 ng ml⁻¹ paroxetine and 80 ng ml⁻¹ fluoxetine, respectively. The fluorescence intensity of hsert-HEK293 cells incubated with paroxetine and fluoxetine was reduced (about 0.48-fold) compared with the control group (Fig. 3b), indicating that the entry of PF-H₂O₂ into the cells after SERT inhibition was significantly reduced. These results indicate that PF-H₂O₂ can enter hsert-HEK293 cells through SERT, allowing for the in situ observation of changes in H₂O₂.

To further verify the targeting ability of PF-H₂O₂, we also conducted fluorescence co-localization experiments of probe PF-H₂O₂ and SERT protein. We immobilized hsert-HEK293 cells and incubated them with serotonin transporter antibodies for 24 hours. Subsequently, the fluorescent secondary antibody was incubated at room temperature for two hours, followed by a 2-hour incubation with PF-H₂O₂. Confocal microscopy imaging revealed significant co-localization between green fluorescence (PF-H₂O₂) and red fluorescence (SERT), with a Pearson coefficient of 0.81. This indicates that PF-H₂O₂ can target SERT-expressing cells (Fig. 3c) and has the potential to target serotonergic neurons.

After validating the effectiveness of PF-H₂O₂ in the hsert-HEK293 cell model, we then applied it to mouse primary neurons in vitro. Firstly, we evaluated the targeting effect of PF-H₂O₂ on serotonergic neurons by conducting fluorescence co-localization experiments of PF-H₂O₂ and SERT proteins in neurons. The results (Fig. 4a) showed good co-localization between green fluorescence (PF-H₂O₂) and red fluorescence (SERT), with a Pearson coefficient of 0.89, confirming that PF-H₂O₂ could effectively target serotonergic neurons. We then examined whether PF-H₂O₂ could image H₂O₂ in serotonergic neurons. Neurons were incubated with 10 mM glutamate to induce oxidative stress and imaged using 25 μM PF-H₂O₂. The findings presented in Fig. 4b and c demonstrate that the fluorescence intensity of glutamate-stimulated neurons (glutamate) was significantly higher than that of the control group (control). Furthermore, the introduction of NAC to eliminate H₂O₂ resulted in a reduced fluorescence intensity in the neurons compared to those stimulated by glutamate. These results indicate that PF-H₂O₂ can effectively detect changes in H₂O₂ levels in neurons. In summary, our findings suggest that PF-H₂O₂ can target primary serotonergic neurons in mice and effectively detect changes in H₂O₂ levels within them.

Fig. 4 Fluorescent imaging of primary mouse neurons. (a) Fluorescence imaging of serotonergic neurons, scale = 50 μm. PF-H₂O₂: the green channel is the fluorescent signal emitted by the probe PF-H₂O₂. Images acquired by excitation at 480 nm. Fluorescence emission windows: λ_em = 510–540 nm. SERT: the red channel is the fluorescent signal emitted by the serotonin transporter antibodies. Images acquired by excitation at 560 nm. Fluorescence emission windows: λ_em = 570–590 nm. (b) Fluorescence imaging of serotonergic neurons by PF-H₂O₂. Control group: 25 μM PF-H₂O₂ incubated for 1 h, glutamate group: 10 mM glutamate incubated for 12 h, 25 μM PF-H₂O₂ treated neurons for 1 h, glutamate + NAC group: 10 mM glutamate incubated for 12 h, 1 mM NAC incubated for 1 h, 25 μM PF-H₂O₂ incubated for 1 h, and 480 nm stimulated images were obtained, scale = 250 μm. (c) Relative fluorescence intensity output of b. Similar results were observed in five independent experiments. **P < 0.01 and ***P < 0.001 is compared with control group.

To evaluate the ability of PF-H₂O₂ to target serotonergic neurons within biological tissues, we performed immunofluorescence imaging colocalization experiments on brain tissue sections. After perfusing the mouse heart, brain tissue was extracted, dehydrated, and sliced into 50 μm sections. These sections were then incubated with serotonin transporter antibodies for 24 hours and subsequently exposed to fluorescent secondary antibodies at room temperature for two hours. Subsequently, they were incubated with PF (the dye that we synthesized to target SERT) for two hours. The results obtained are shown in Fig. 5b. There is clear fluorescence co-localization between the green fluorescence (PF) and the red fluorescence (SERT), with a Pearson coefficient of 0.78, indicating that PF-H₂O₂ can effectively target serotonergic neurons in tissue sections.

Fig. 5 Mouse brain imaging in vivo and immunofluorescence experiment of mouse brain tissue section. (a) Mouse processing flowchart. (b) Depression-like behavior test in mice. Sucrose preference test in mice with and without CORT (left). Immobility time for forced swim test and tail suspension test in behavioral tests (right). (c) Fluorescence imaging of the brains of mice in stress group (left) and control group (right). Control group: mice without drug administration. Stress group: mice that received continuous CORT administration for 21 days. (d) Relative fluorescence intensity output of c. (e) Confocal fluorescence images of mouse brain tissue section. PF-H₂O₂: the green channel is the fluorescent signal emitted by the dye PF. Images acquired by excitation at 480 nm. Fluorescence emission windows: λ_em = 510–540 nm. SERT: the red channel is the fluorescent signal emitted by the serotonin transporter antibodies. Images acquired by excitation at 560 nm. Fluorescence emission windows: λ_em = 570–590 nm. Scale bar = 50 μm. Similar results were observed in five independent experiments. *P <0.05 and ***P <0.001 is compared with the control group.

These results encouraged us to apply PF-H₂O₂ at the mouse level. We first performed in vivo toxicity assessment in mice. By injecting 0.7 mg kg⁻¹ of PF-H₂O₂ intraperitoneally for 7 consecutive days, the body weight of mice in both groups did not change significantly (Fig. S5, ESI†), and the hair color of mice in both groups did not change significantly upon observation with the naked eye, indicating that PF-H₂O₂ could be applied in mouse experiments. We established a depression model by dosing corticosterone (CORT, 100 μg mL⁻¹) for 21 consecutive days and validated the model through depression-like behavioral tests (forced swimming test, tail suspension test, sugar water preference test) (Fig. 5c).²⁹ We imaged the changes in H₂O₂ levels in serotonergic neurons of depressed mice and normal mice. 0.5 mg kg⁻¹ of PF-H₂O₂ was injected intraperitoneally, and in vivo imaging was performed 1 hour later. As shown in Fig. 5d and e, the level of H₂O₂ in the brains of depressed mice was higher than that in normal mice. Next, we removed the brain tissue and cut it into 50 μm slices using frozen section technology. Confocal imaging showed that the fluorescence intensity of depressed mice was also higher than that of normal mice (Fig. S6, ESI†). These results suggest that PF-H₂O₂ can effectively detect H₂O₂ levels in serotonergic neurons in mice. Additionally, these experiments also revealed that H₂O₂ levels in serotonergic neurons of depressed mice were higher than those in normal mice, indicating that oxidative stress in serotonergic neurons is an important factor affecting the development of depression.

Conclusions

In summary, we developed a novel small molecule fluorescent probe (PF-H₂O₂) designed to specifically target serotonergic neurons and visualize dynamic changes in H₂O₂ levels. PF-H₂O₂ incorporates paroxetine as the targeting moiety, luciferin as the fluorophore, and phenylborate for H₂O₂ recognition. At the cellular level, we found that H₂O₂ levels increased in SERT-expressing hsert-HEK293 cells under oxidative stress. Similarly, primary serotonergic neurons cultured in vitro showed elevated H₂O₂ levels in response to high glutamate stimulation. Additionally, in vivo imaging revealed increased H₂O₂ levels in serotonergic neurons of depressed mice. This study offers an effective approach to exploring redox signaling pathways in serotonergic neurons and suggests that our imaging tool could be a valuable resource for investigating serotonergic neuron-related diseases.

Author contributions

Feida Che: conceptualization, methodology, investigation, and writing – original draft. Xiaoming Zhao: visualization and investigation. Qi Ding: investigation. Xiwei Li: investigation. Wen Zhang: supervision and reviewing. Xin Wang: supervision, reviewing and editing. Ping Li: supervision, reviewing and editing. Bo Tang: supervision and reviewing.

Data availability

The data supporting this article have been included as part of the ESI.†

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (22134004, 22074083, 22304107, and 22377070) and the Natural Science Foundation of Shandong Province of China (ZR2023ZD31, ZR2023YQ016, and ZR2021QB042).

Notes and references

1. S. Marwaha, E. Palmer, T. Suppes, E. Cons, A. H. Young and R. Upthegrove, Lancet, 2023, 401, 141–153.
2. C. Otte, S. M. Gold, B. W. Penninx, C. M. Pariante, A. Etkin, M. Fava, D. C. Mohr and A. F. Schatzberg, Nat. Rev. Dis. Primers, 2016, 2, 16065.
3. S. Law and P. Liu, Curr. Psychiatry Rep., 2008, 10, 80–86.
4. N. G. Rummel and D. A. Butterfield, Antioxid. Redox Signaling, 2021, 36, 1289–1305.
5. H. Peng, T. Wang, G. Li, J. Huang and Q. Yuan, Anal. Chem., 2022, 94, 1070–1075.
6. H. Sies, C. Berndt and D. P. Jones, Annu. Rev. Biochem., 2017, 86, 715–748.
7. H. Xiao, P. Li, X. Hu, X. Shi, W. Zhang and B. Tang, Chem. Sci., 2016, 7, 6153–6159.
8. L. C. Daws and G. G. Gould, Pharmacol. Ther., 2011, 131, 61–79.
9. A. Sørensen, H. G. Ruhé and K. Munkholm, Mol. Psychiatry, 2022, 27, 192–201.
10. S. G. Rhee, Science, 2006, 312, 1882–1883.
11. N. Kasai, C. Han and K. Torimitsu, Sens. Actuators, B, 2005, 108, 746–750.
12. N. V. Kulagina and A. C. Michael, Anal. Chem., 2003, 75, 4875–4881.
13. E. W. Miller, O. Tulyathan, E. Y. Isacoff and C. J. Chang, Nat. Chem. Biol., 2007, 3, 263–267.
14. J. Zhang, C. Wickizer, W. Ding, R. Van, L. Yang, B. Zhu, J. Yang, Y. Wang, Y. Wang, Y. Xu, C. Zhang, S. Shen, C. Wang, Y. Shao and C. Ran, Proc. Natl. Acad. Sci. U. S. A., 2023, 120, 4099–4100.
15. L. Zhu, Y. Pan, Z. Hua, Y. Liu and X. Zhang, J. Am. Chem. Soc., 2024, 146, 14307–14317.
16. S. Zeng, Y. Wang, C. Chen, H. Kim, X. Liu, M. Jiang, Y. Yu, Y. S. Kafuti, Q. Chen, J. Wang, X. Peng, H. Li and J. Yoon, Angew. Chem., Int. Ed., 2024, 63, e202316487.
17. W. Zhang, T. Liu, F. Huo, P. Ning, X. Meng and C. Yin, Anal. Chem., 2017, 89, 8079–8083.
18. X. X. Zhang, F. Yang, X. Zhao, Q. Wu, L. He, Z. Li, Z. Zhou, T. B. Ren, X. B. Zhang and L. Yuan, Angew. Chem., Int. Ed., 2024, e202410666.
19. E. W. Miller, A. E. Albers, A. Pralle, E. Y. Isacoff and C. J. Chang, J. Am. Chem. Soc., 2005, 127, 16652–16659.
20. H. Iwashita, E. Castillo, M. S. Messina, R. A. Swanson and C. J. Chang, Proc. Natl. Acad. Sci. U. S. A., 2021, 118, e2018513118.
21. J. H. Meyer, A. A. Wilson, S. Sagrati, D. Hussey, A. Carella, W. Z. Potter, N. Ginovart, E. P. Spencer, A. Cheok and S. Houle, Am. J. Psychiat., 2004, 161, 826–835.
22. D. Yang and E. Gouaux, Sci. Adv., 2021, 7, eabl3857.
23. S. Tavoulari, L. R. Forrest and G. Rudnick, J. Neurosci., 2009, 29, 9635–9643.
24. J. T. Coyle and P. Puttfarcken, Science, 1993, 262, 689–695.
25. B. Kalyanaraman, Redox Biol., 2022, 57, 102497.
26. Y. Chen, X. Shi, Z. Lu, X. Wang and Z. Wang, Anal. Chem., 2017, 89, 5278–5284.
27. M. Berk, G. S. Malhi, L. J. Gray and O. M. Dean, Trends Pharmacol. Sci., 2013, 34, 167–177.
28. R. E. Presby, R. A. Rotolo, E. M. Hurley, S. M. Ferrigno, C. E. Murphy, H. P. McMullen, P. A. Desai, E. M. Zorda, F. B. Kuperwasser, C. C. Ros, M. Correa and J. D. Salamone, Pharmacol., Biochem. Behav., 2021, 202, 173115.
29. H. Ding, X. Y. Cui, S. Y. Cui, H. Ye, X. Hu, H. L. Zhao, Y. T. Liu and Y. H. Zhang, Neurosci. Lett., 2018, 687, 202–206.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4tb01828a

‡ Those authors contributed equally.

---