Endoplasmic reticulum-targetable selenium-doped carbon nanodots with redox-responsive fluorescence for in situ free-radical scavenging in cells and mice

2.1 Reagents

Selenocystamine, 5,5-dimethyl-1-pyrroline N-oxide (DMPO), glutathione, riboflavin, methionine, nitro blue tetrazolium (NBT), FeCl₂, NaHCO₃, H₂O₂ were supplied by Aladdin Chemistry Co. Ltd. N-(2-aminoethyl)-4-methyl-benzenesulfonamide, phorbol 12-myristate 13-acetate (PMA), 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT), and LiOH were bought from Sigma-Aldrich. Methyl 3-chloro-3-oxopropanoate was a kind gift from Dr. Zhichao Liu. 2′,7′-Dichlorofluorescein diacetate (DCFH-DA) were purchased from Beyotime Chemical Reagent Co. Ltd. High glucose Dulbecco’s modified Eagle’s media (DMEM) were received from KeyGEN Biotech. Co. Ltd. All the mentioned reagents were of analytical grade and used as received.

2.2 Characterizations

Absorption spectra were acquired on a Shimadzu UV-2550 spectrophotometer. Fluorescence spectra were measured on a F-4600 fluorescence spectrophotometer (Hitachi). Transmission electron microscopy (TEM) experiments were performed on transmission electron microscope working at an acceleration voltage of 200 kV (JEOL JEM-2100F). Atomic force microscopic (AFM) characterizations were conducted in the ScanAsyst mode under ambient conditions (Bruker). X-ray photoelectron spectroscopy (XPS) data was collected using a thermoelectron instrument (Thermo Scientific ESCALAB 250). Fourier transform infrared spectroscopy (FTIR) spectrum of the samples were acquired by applying a Nicolet iS10 FTIR spectrometer. Electron spin resonance (ESR) experiments were carried out on a Bruker ELEXSYS E500 ESR spectrometer. ¹H and ¹³C nuclear magnetic resonance (NMR) spectra were obtained from a Bruker 500 MHz spectrometer. Confocal fluorescence and bright field images (512 × 512 pixels) were acquired with a Leica TCS-SP8 confocal laser scanning microscope. Cells images were captured using a 63 × objective lens.

2.3 Synthesis of Se-CDs

Se-CDs can be synthesized from selenocystamine via a benign, one-pot solvothermal carbonization approach. Briefly, 0.1 g selenocystamine was dissolved in 40 mL of deionized water under rapid stirring. Then, the aqueous dispersion was transferred into a Teflon-lined autoclave (50 mL) and heated for 2 h at a temperature of 200 ℃. The resulting solution with reddish brown color was filtrated with a 0.22 μm filter membrane, centrifugated at 12000 g for 10 min, and dialyzed through a dialysis tubing with a molecular weight cutoff of 1000 Da (changing the water every 8 h). After lyophilization of the purified solution, Se-CDs powder can thus be acquired.

2.4 Synthesis of 3-((2-((4-methylphenyl)sulfonamido)ethyl)amino)-3-oxopropanoic acid (abbreviated as MOA)

To a round-bottomed flask (100 mL), in which N-(2-aminoethyl)-4-methyl-benzenesulfonamide (1.0 g, 4.7 mmol) was dissolved into a mixed solution of tetrahydrofuran (20 mL) and saturated NaHCO₃ solution (5 mL), was introduced methyl 3-chloro-3-oxopropanoate (0.7 g, 5.1 mmol). After reaction for 1 h at 0 ℃, LiOH solution (2 M, 5 mL) was slowly injected, followed by another incubation at room temperature for ∼ 12 h. Afterwards, the solvent was discarded under reduced pressure and the obtained residue was purified by HPLC system using a gradient program (mobile phase: acetonitrile to water, from 5% to 95%, with 0.1% TFA) at 20 mL/min flow rate to afford desired product as pale yellow oily matter (0.24 g, yield: 17%). ¹H NMR (400 MHz, CDCl₃) δ (ppm): 7.92 (s, 1H), 7.77 (s, 1H), 7.67 (d, 2H), 7.22 (d, 2H), 6.31 (s, 1H), 3.37 (m, 4H), 3.00 (s, 2H), 2.34 (s, 3H). ¹³C NMR (400 MHz, CDCl₃) δ (ppm): 171.11, 169.13, 143.63, 136.36, 129.82, 126.99, 42.16, 40.36, 39.66, 21.46. HRMS-ESI (m/z): [M + Na]⁺ calcd for C₁₂H₁₆N₂NaO₅S: 323.0678; found: 323.0672.

2.5 Preparation of ER-Se-CDs

The immobilization of MOA on Se-CD’s surface was realized under the activation of EDC/NHS. Typically, MOA solution (1.0 mL, 1.5 mM) was mixed with EDC/NHS (50 mg/50 mg) and activated at room temperature for 2 h. After then, Se-CDs solution (2.0 mL, 1.0 mg·mL⁻¹) was introduced, and further reacted for one night to afford MOA functionalized Se-CDs (ER-Se-CDs). The unreacted MOA was removed by dialysis. By measuring the absorption spectra of MOA in the above collected water and the stock solution, density of MOA on the Se-CD surface were determined to be 46.6 μmol·g⁻¹.

2.6 Cytotoxicity evaluations

HeLa cells were fostered at a density of ∼ 1 × 10⁴ cells per well in 96-well plates, and grown in DMEM containing 10% fetal bovine serum, 80 U·mL⁻¹ penicillin, and 80 μg·mL⁻¹ streptomycin in a humid incubator with 5% CO₂/95% air. After 12 h, the culture media were removed, and fresh one with serial dilutions of ER-Se-CDs (0–80 μg·mL⁻¹) was introduced, cultured for 48 h. As to each amount, five independent experiments were carried out. Thereafter, MTT solution (1.0 mg·mL⁻¹) with a volume of 20 μL was injected into each well, culturing for 4 h to allow the formation of formazan crystals. Subsequently, 150 μL DMSO was added into the wells. Absorbance (A) of the resultant hybrid at 570 nm was measured. Cellular viability values were counted according to the following equation: cellular viability (%) = A_test/A_control × 100%, where A_control refers to the absorbance recorded from the control group, and A_test refers to the absorbance obtained with the presence of ER-Se-CDs.

2.7 Cellular imaging

The subcellular localization of ER-Se-CDs was investigated by co-localization bioimaging experiment, in which ER-Se-CDs (20 μg·mL⁻¹) stained cells were again stained for 20 min with ER-Tracker Red (50 nM). After finishing the labeling experiments, cells were rinsed thoroughly with cool PBS. Fluorescence signal of ER-Tracker Red was obtained in 570–650 nm wavelength region (excited at 552 nm), as for ER-Se-CDs, an excitation wavelength of 405 nm was selected, and its blue fluorescence was collected in the wavelength region of 410–480 nm.

2.8 In vivo imaging

All animal experiments were approved by the Animal Ethics Committee of East China Normal University. Kunming male mice with a body weight of 15 g were selected for establishing an ear inflammation model, which was triggered by topically applying PMA solution (100 μg·mL⁻¹, 50 μL) on the right ear. After incubation for 6 h, the mice were subcutaneously administered with ER-Se-CDs (20 μg·mL⁻¹, 100 μL; 40 μg·mL⁻¹, 100 μL). 0.5 h later, DCFH-DA (1 mM, 50 μL), a non-fluorescent cell permeable dye that reacts with intracellular ROS and emits fluorescence at 520 nm was injected. After another 0.5 h, fluorescence image of the mice body was captured with a PerkinElmer in vivo imaging system (λ_ex = 488 nm; λ_em = 520 nm).

3.1 Characterization of Se-CDs

In our experiment, Se-CDs can be conveniently prepared by one-step hydrothermal treating of selenocystamine solution (Fig. 1). The size and morphology of the obtained Se-CDs were characterized by TEM and AFM. TEM image in Fig. 2A shows that Se-CDs are distributed randomly and uniformly with a narrow size distribution in 2.4 ∼ 3.2 nm range (Fig. 2B). This size is comparable to other heteroatom doped CDs (Gong et al., 2016; Liu et al., 2017). The morphology of Se-CDs was further examined by AFM (Fig. 2C). From which, we can see that Se-CDs were monodispersed. The line scan of three individual CDs reveals that their topographic heights range from 2.6 to 3.3 nm (Fig. 2D), which agrees well with the TEM characterization.

Close

Fig. 2

(A) TEM characterization of Se-CDs. (B) Size distribution of Se-CDs. (C) Typical AFM image of Se-CDs. (D) Height distribution of Se-CDs along the black line in panel C.

Export to PPT

Then, functional groups of the Se-CDs were studied by XPS experiments. As depicted in the XPS survey spectrum (Fig. 3A), four main peaks were found at 530.8 eV (O_1s), 399.8 eV (N_1s), 284.7 eV (C_1s), along with 55.1 eV (Se_3d), demonstrating that Se-CDs are dominantly composed of oxygen (atomic ratio: 9.57%), nitrogen (18.62%), carbon (64.91%), and selenium (6.9%). Deconvoluted XPS spectra for C_1s (Fig. 3B) show that the C_1s band contains three distinct peaks at 284.7, 286.1, and 287.9 eV, attributed to C–C, C-N/C-O/C-Se, and C = O, respectively (Li et al., 2017; Liu et al., 2020; Liu et al., 2023). Deconvoluted XPS spectra for N_1s spectra show two peaks at 399.8 and 401.1 eV, ascribed to C-N-C and amino N, respectively (Fig. 3C) (Lu et al., 2017; Zhao et al., 2021). The high-resolution XPS spectrum of Se3d (55.1 eV) reveals -C-Se group existed on the surface of Se-CDs (Fig. 3D) (Huang et al., 2020). Besides, the absorption peaks appeared in the FTIR spectrum at 3429, 1578, and 1221 cm⁻¹, respectively corresponding to N–H, C = N, and C-N, which suggested the existence of amine groups on the surface of the Se-CDs (Fig. S1, curve a) (Gong et al., 2019).

Close

Fig. 3

(A) XPS survey spectrum of Se-CDs. (B) High-resolution C1s spectrum of Se-CDs. (C) High-resolution N1s spectrum of Se-CDs. (D) High-resolution Se3d spectrum of Se-CDs.

Export to PPT

Subsequently, the optical features of Se-CDs were monitored. The absorption spectrum of Se-CDs presents two peaks at 241 and 299 nm, as indicated in Fig. S2, which respectively attributes to the π-π* transition of C = C band and n-π* electron transition of C = O band (Lu et al., 2018; Wang et al., 2016; Li et al., 2018). Under the excitation with a maximum wavelength of 369 nm (Fig. 4A, curve a), a strong emission peak centered at 445 nm is observable (Fig. 4A, curve b). Using quinine sulfate in 0.10 M H₂SO₄ solution as the reference, fluorescence quantum yield of Se-CDs was calculated to be 8.3%, which is comparable to Se-CDs reported by Xu and coworkers (Li et al., 2017). However, unlike previously reported Se-CDs, which show excitation-energy dependent fluorescence, the fluorescence peak of Se-CDs did not change under excitation in the wavelength range from 360 to 410 nm (Fig. 4B). The excitation-energy independent fluorescence of our Se-CDs probably result from its uniform size (Wang and Hu, 2014; Li et al., 2010). Remarkably, Se-CDs had robust photostability. Illuminating Se-CDs with a 405 nm light for 1 h witnessed truly little variation of its fluorescence (Fig. 4C). Besides, there were no changes in fluorescence intensity at high ionic strengths (Fig. 4D), suggesting the great potency of Se-CDs to be used in physical salt concentrations.

Close

Fig. 4

(A) Fluorescence spectra of Se-CDs (a: excitation spectrum; b: emission spectrum). (B) Fluorescence spectra of Se-CDs gained at varied excitation from 360 to 410 nm. (C) Change of fluorescence intensity under uninterrupted irradiation by a 90 W Xe lamp (λ_ex = 405 nm). (D). Variation of fluorescence intensity at 445 nm in NaCl solution.

Export to PPT

Moreover, we found Se-CDs exhibited redox-responsive fluorescence. As shown in Fig. 5A and Fig. 5B (curve a and curve b), addition of H₂O₂ to Se-CD solution could induce an apparent increasement of its fluorescence. At the same time, selenic acid was generated in the presence of H₂O₂ (Fig. 5C). However, when reduced glutathione was introduced into the above mixed solution of Se-CDs and H₂O₂, the fluorescence intensity was nearly reduced back to its initial state (Fig. 5B, curve c). Concurrently, -C-Se was generated from the reduction of selenic acid (Fig. 5D). Overall, these analyses suggested that the redox state of Se atom had a prominent influence on the fluorescence of Se-CDs.

Close

Fig. 5

(A) Fluorescence response of Se-CDs (20 μg·mL⁻¹) to various amounts of H₂O₂ (From bottom to up: 0, 20, 40, 60, 80, and 100 μM). (B) Redox-responsive fluorescence of Se-CDs: (a) Se-CDs (20 μg mL⁻¹); (b) Se-CDs (20 μg mL⁻¹) + H₂O₂ (100 μM); (c) Se-CDs (20 μg mL⁻¹) + H₂O₂ (100 μM) + glutathione (1.0 mM). (C) High-resolution XPS spectrum of Se_3d in Se-CDs reacted with H₂O₂. (D) High-resolution XPS spectrum of Se_3d in Se-CDs reacted with H₂O₂ and glutathione.

Export to PPT

3.2 Synthesis of ER-Se-CDs

We then synthesized an ER targeting molecule, MOA, whose chemical structure was verified by ¹H NMR, ¹³C NMR, and HRMS (Fig. S3–S6). MOA was rationally designed with two functional moieties. Moiety one is a methyl sulphonamide group, which is commonly used for targeting molecules and/or nanomaterials into ER region (Xiao et al., 2018; Huang et al., 2022), and moiety two is a carboxyl group that can be used for further conjugation with amine group. The fixation of MOA on Se-CDs was affirmed by FTIR spectroscopy. As shown in Fig. S1 (curve c), an absorption band at 1153 cm⁻¹, assigned to the C-S group in MOA, was found in the FTIR spectrum of ER-Se-CDs. Additionally, a larger hydrodynamic size was obtained for ER-Se-CDs, as compared with that of Se-CDs, which further testified the successful preparation of ER-Se-CDs (Fig. S7). It is worthy to note that the attachment of MOA on Se-CDs had no obvious influence on the size and fluorescence of Se-CDs (Fig. S8-S9).

3.3 ROS scavenging efficiency of ER-Se-CDs

The existence of redox-responsive unit -C-Se on the surface of ER-Se-CDs implied its possibility to be used for eliminating •OH and O₂^•-. For investigating its elimination efficiency toward •OH, ESR experiments, a direct and reliable method to detect ROS, were carried out, in which Fe²⁺/H₂O₂ was chosen as the source to provide •OH and 5,5-dimethyl-1-pyrroline N-oxide (DMPO) as the spin trap agent (Huang et al., 2019). Typically, an intense four-line ESR signal with a height ratio of 1:2:2:1, characteristic of DMPO-OH adduct, could be detected once Fe²⁺/H₂O₂ was injected to DMPO solution, evidencing the high level of •OH (Fig. 5A, curve a). Upon ER-Se-CDs treatment, the ESR intensity of the above mixture significantly weakened in a dose-dependent manner (Fig. 6A, curve b-f). The ESR signal intensity decreased down to 28.1% (elimination efficiency: 71.9%) with the involvement of ER-Se-CDs at a concentration of 50 μg·mL⁻¹ (Fig. 6B), indicating ER-Se-CDs possessed an excellent removal capacity toward •OH. Subsequently, O₂^•- removal capacity was assessed through determining the inhibition ratio of NBT reduction by O₂^•-. As seen in Fig. 6C (curve a), a strong absorption signal around 560 nm appeared after photoirradiation (intensity: 3500 lx; irradiation time: 15 min) of a mixture containing methionine, riboflavin, and NBT, reflecting the generation of O₂^•- in this process (Zhang et al., 2021). Once ER-Se-CDs was added into the aforesaid mixture, the absorption signal significantly decreased (Fig. 6C, curve b-f), as expected. More than 70% O₂^•- could be removed from the solution with ER-Se-CDs treatment at the concentration of 50 μg·mL⁻¹ (Fig. 6D), suggesting the forceful O₂^•- removing ability of ER-Se-CDs. These results demonstrated that ER-Se-CDs can act as efficient scavengers for both •OH and O₂^•-.

Close

Fig. 6

(A) ESR spectra of DMPO (50 mM) with the co-existence of •OH (10 μM) and different amounts of ER-Se-CDs (a: 0 μg·mL⁻¹; b: 10 μg·mL⁻¹;c: 20 μg·mL⁻¹; d: 30 μg·mL⁻¹; e: 40 μg·mL⁻¹; f: 50 μg·mL⁻¹). (B) The corresponding intensity of ESR signal and elimination efficiency of ER-Se-CDs toward •OH. (C) Absorption spectra of a system comprised of riboflavin, methionine, and NBT with the existence of different concentrations of ER-Se-CDs (a: 0 μg·mL⁻¹; b: 10 μg·mL⁻¹;c: 20 μg·mL⁻¹; d: 30 μg·mL⁻¹; e: 40 μg·mL⁻¹; f: 50 μg·mL⁻¹). (D) The corresponding absorbance and elimination efficiency of ER-Se-CDs toward O₂^•-.

Export to PPT

3.4 Scavenging activity of ER-Se-CDs in cells

After successfully verifying ER-Se-CDs’ antioxidant capacity, its potency as scavengers against intracellular ROS was explored. Prior to this investigation, cytotoxicity of ER-Se-CDs was studied by standard MTT assay with various dosages of ER-Se-CDs and incubated for a period of 48 h. The findings, as shown in Fig. 7A, revealed that even incubated with a high dosage (up to 80 μg·mL⁻¹) of ER-Se-CDs, the cells still showed a high viability rate over 90%. As such, we conclude that ER-Se-CDs is almost nontoxic to live cells.

Close

Fig. 7

(A) Viabilities of HeLa cells incubated with different amounts of ER-Se-CDs. (B) Co-localization investigations in HeLa cells that were co-labeled with ER-Se-CDs (20 μg·mL⁻¹) and ER-Tracker Red (50 nM): (a) fluorescence image from ER-Se-CDs (λ_em = 410–480 nm, λ_ex = 405 nm); (b) fluorescence image from ER-Tracker Red (λ_em = 570–650 nm, λ_ex = 552 nm); (c) the merged image of (a) and (b); (d) the bright field image; (e) the corresponding intensity correlation plot of the blue channel with red channel; (f) intensity profiles of ER-Se-CDs and ER-Tracker Red along the line across the cell. Scale bar: 10 μm. (C) Protective capabilities of ER-Se-CDs to HeLa cells from oxidative damage initiated by H₂O₂ (150 μΜ). Data showed are mean values and standard deviations from five independent experiments. (D) HeLa cells were co-incubated with DCFH-DA (10 µM) and PMA (3 μg·mL⁻¹) with or without ER-Se-CDs and analyzed by flow cytometry: (a) DCFH-DA + PMA; (b) DCFH-DA + PMA + ER-Se-CDs (20·μg mL⁻¹); (c) DCFH-DA + PMA + ER-Se-CDs (40 μg·mL⁻¹).

Export to PPT

As ER-Se-CDs was functionalized by methyl sulphonamide group, which possesses ER-targeting capability, we then examined its potential to localize into ER of cells. In this respect, co-localization experiment was conducted via treating cells with ER-Se-CDs (20 μg·mL⁻¹) and commercial ER-Tracker Red (50 nM) in sequence. As displayed in Fig. 7B, the bright blue fluorescence from ER-Se-CDs (Fig. 7Ba) showed a large extent of overlap with the red fluorescence from ER-Tracker Red (Fig. 7Bb), as observed from the intense pink signal in the merged image (Fig. 7Bc). The corresponding Pearson's coefficient is 0.87 (Fig. 7Be). Moreover, the variations in the intensity profiles of the lines across the cells are coinstantaneous in the two separate channels (Fig. 7Bf). On the other hand, a high Pearson’s coefficient value of 0.85 was also obtained when co-localization experiments were conducted in living astrocytes with ER-Se-CDs and ER-Tracker Red (Fig. S10). These results inferred that with the anchor of methyl sulphonamide group, ER-Se-CDs, as expected, can selectively reside in ER region.

Oxidative stress occurs as the amount of ROS exceeds the antioxidant competence of cellular antioxidants, giving rise to severe harm to nucleic acids, proteins, and lipids. Since ER-Se-CDs presented high elimination efficiency toward •OH and O₂^•-, low cytotoxicity, and ER targetability, we envisioned that ER-Se-CDs could exert an efficient protection against ROS-induced damage within ER. Then, H₂O₂ was served as the damage initiator to assess the protection ability of ER-Se-CDs against oxidative stress. The corresponding data are displayed in Fig. 7C. It was obvious that ER-Se-CDs, the first ER targetable nanoantioxidant, could dose-dependently restrain ROS-caused decline of cell viabilities. Moreover, to further monitor the cellular ROS level after treatment with different amounts of ER-Se-CDs, DCFH-DA assay was performed by flow cytometry. DCFH-DA itself showed no fluorescence. Once it enters into cells, it would be hydrolyzed by cellular enzyme to produce DCFH, which could emit strong fluorescence after being oxidized by ROS (Wei et al., 2019). As seen in Fig. 7D, a lower fluorescence intensity was observed for the cells pre-incubated with ER-Se-CDs than the cells treated with H₂O₂ alone, suggesting a lower level of ROS in the cells pre-treated with ER-Se-CDs. Together, these data demonstrate that ER-Se-CDs is competent to consume ROS in ER region.

3.5 In vivo anti-inflammation

The in vivo anti-inflammation performance of ER-Se-CDs was also studied. In the experiments, to establish an inflammation model of the right ear, PMA, an apoptotic drug that can initiate tissue inflammation was regionally applied. As shown, no visible fluorescence signal was found from the right ear of the mouse solely treated with DCFH-DA, manifesting a low level of ROS (Fig. 8A). In sharp contrast, a strong fluorescence was clearly perceptible from DCFH-DA and PMA co-treated ear, demonstrating the elevation of ROS level in this process (Fig. 8B). After subcutaneously treated with ER-Se-CDs, the fluorescence in the inflamed region strongly quenched (Fig. 8C and 8D). Additionally, as shown in Fig. S11, hematoxylin and eosin (H&E) stained histological images of heart, liver, spleen, lung, and kidney demonstrated negligible toxicity of ER-Se-CDs toward major organs. These results indicated that ER-Se-CDs possessed an efficient ROS scavenging ability against ear inflammation in live mice with great biosafety.

Close

Fig. 8

(A) In vivo fluorescence imaging of mouse solely treated with DCFH-DA (1 mM, 50 μL) in the right ear region. (B) In vivo fluorescence imaging of mouse sequentially treatment with PMA (100 μg·mL⁻¹, 50 μL) and DCFH-DA (1 mM, 50 μL). (C) In vivo fluorescence imaging of mouse orderly treated with PMA (100 μg·mL⁻¹, 50 μL), ER-Se-CDs (20 μg·mL⁻¹, 100 μL), and DCFH-DA (1 mM, 50 μL). (D) In vivo fluorescence imaging of mice sequentially treated with PMA (100 μg·mL⁻¹, 50 μL), ER-Se-CDs (40 μg·mL⁻¹, 100 μL), and DCFH-DA (1 mM, 50 μL). Each group contains five mice. (E) Relative fluorescence intensity in panel A–D. All the fluorescence intensities were compared to panel B, which was normalized to 1.0.

Export to PPT