Aggregation-Induced Emission-Based Fluorescence Probe for Fast and Sensitive Imaging of Formaldehyde in Living Cells.

Formaldehyde (FA), as a reactive carbonyl species and signaling molecule, plays an important role in living systems. Here, an FA-responsive probe with fast response and great selectivity is designed based on aggregation-induced emission. The probe is prepared by functionalizing tetraphenylethene (TPE) with two amine groups. FA is detected based on the solubility differences between the amine-functionalized TPE and the corresponding Schiff bases after reaction with FA. The probe exhibits a limit of detection of 40 nM and a response time of ∼90 s. Furthermore, its ability to detect both endogenous and exogenous FA is demonstrated in living cells with high specificity. Moreover, the probe is also introduced to image endogenous FA in real time with fast response. These results suggest that our probe holds great potential for tracking FA in living systems under various physiological conditions as well as related biomedical applications.

Introduction

Formaldehyde (FA) is a common hazardous molecule that can be generated from various natural and anthropogenic sources. As a reactive carbonyl species, it can also be endogenously generated via various biological processes, including one-carbon metabolism, metabolite oxidation, demethylation events, epigenetic modifications, and so forth.[1−4] FA is also an important signaling molecule that is involved in promoting proliferation and mediating memory formation. However, aberrant elevations of FA concentrations are associated with various pathological states such as cancer, diabetes, and neurodegenerative diseases.[5−7] Therefore, FA is delicately balanced between production and consumption in biological systems to maintain proper cellular functions. For instance, the intracellular concentration of FA was maintained in the range of 1.5–4.0 μM.[8] Because of its significance in physiological and pathological processes, it is essential to develop new methods to track FA concentrations in living cells.

Fluorescent probes are emerging as powerful tools for tracking trace small molecules in living systems.[9−12] Currently, there are mainly two types of reactivity-based probes for detecting intracellular FA, which were designed by harnessing the relatively strong electrophilicity of the carbonyl group. The first type was designed by masking aldehyde fluorophores with a homoallylamine moiety. FA was detected based on a 2-aza-Cope rearrangement which subsequently hydrolyzed to aldehyde fluorophores with activated fluorescence.[13−16] A self-immolative linker was further incorporated to improve design versatility.[17,18] However, the reaction kinetics of these probes toward FA is rather slow, and several hours (2–3 h) are generally required to obtain a decent signal-to-background ratio. Their long response time severely hindered real-time tracking of FA fluctuation in biological systems as the half-life of FA is approximately 90 s in organisms.[19] Alternatively, smart designs of FA probes with fast kinetics were developed based on the photoinduced electron-transfer mechanism by using the reactions between FA and hydrazine or amines.[20−22] Though displaying rapid responses in vitro within 10 s to 30 min, these probes suffered from less desirable selectivity and were susceptible to interfering reactions with other aldehydes such as acetaldehyde. Moreover, these probes displayed a rather sluggish response to FA in living cells. Because of the transient and reactive nature of FA, it is still highly desirable to develop fluorescent probes that can monitor FA fluctuations in living cells with fast response, high sensitivity, and great selectivity.

Herein, we develop a novel strategy for designing a fluorescence probe for FA based on the aggregation-induced emission (AIE) that features fast response, favorable selectivity, and high sensitivity. Molecules with typical AIE attributes, nonemissive in dilute solution but highly emissive in the aggregated status, have been widely introduced for bioimaging and biosensing.[23−25] We envision that the decreased aqueous solubility of Schiff bases as compared to that of amine groups can be utilized to design a fluorescent probe for FA based on the AIE phenomenon. On the basis of this rationale, we design the FA-responsive AIE probe, AIE-FA, by introducing two FA-reactive amine groups in tetraphenylethene (TPE), a well-known AIE luminogen (Scheme ). The two amine groups not only function as the reactive moieties for FA but also increase its aqueous solubility. AIE-FA is nonfluorescent in the dissolved state. Upon condensation with FA, the amine groups are converted to Schiff bases, resulting in poor solubility and the formation of aggregated products, which turns on the fluorescence signals because of the restriction of the intramolecular rotation-induced energy dissipation pathway.[26] To our knowledge, this is the first time that AIE is used in the design of fluorescence probes for detection and imaging of FA. It is demonstrated that AIE-FA exhibits ultrafast response kinetics, high sensitivity, and great selectivity toward FA in vitro. Live cell studies reveal that AIE-FA is capable of imaging endogenous and exogenous FA in living cells. Additionally, the ability of AIE-FA to image endogenous FA in real time is also demonstrated. These advantages may endow our AIE-FA probe with great potential for long-term tracking of FA concentrations in living systems under different physiological conditions.

Scheme 1

Illustration of Fluorescence Turn-On Mechanism for FA Detection and Imaging

Results and Discussion

In Vitro Characterizations

On the basis of the above rationale, two probes for FA were synthesized by functionalizing TPE with one amine moiety (TPE-NH2) and two amine moieties (AIE-FA) (Scheme S1). The intermediates and final compounds were confirmed with 1H NMR, 13C NMR, and mass spectroscopy (MS) (Figures S1–S8). After synthesizing these two probes, we began to examine that the fluorescence emission spectra of AIE-FA and TPE-NH2 were obtained in phosphate-buffered saline (PBS), with different volume fractions of dimethylsulfoxide (DMSO). AIE-FA displayed an evident fluorescence peak at 530 nm when the volume fraction of DMSO was smaller than 10% (Figure S9). In contrast, there appeared a distinct fluorescence emission peak at 490 nm when the volume fraction of DMSO was smaller than 30% for TPE-NH2 (Figure S10). The emergence of fluorescence emission peaks was due to the AIE phenomenon. These results indicated that AIE-FA was in the dissolved state when the volume fraction of DMSO was larger than 10%, while the volume fraction of DMSO should be larger than 30% for TPE-NH2 to be in its dissolved state. Then, their responses toward FA were investigated. AIE-FA was essentially nonfluorescent in PBS buffer, supplemented with 10% DMSO, attributed to the two hydrophilic amine groups which improved its aqueous solubility. However, upon addition of 15 μM FA, there was a ∼12-fold fluorescence enhancement at 530 nm (Figure a). This was probably because the amine groups were converted to less aqueous soluble Schiff bases upon a condensation reaction with FA, which activated fluorescence due to a typical AIE phenomenon. The AIE characteristics of the product were further investigated via measuring fluorescence emission in a PBS/DMSO binary system. There was gradual increase in fluorescence as the volume fractions of PBS increased, displaying a typical AIE phenomenon (Figure S11).[27] In addition, there was also a slight red shift and increase in absorbance after reaction with FA, shifting from 325 to 350 nm (Figure S12). This bathochromic shift was probably due to the formation of Schiff bases which extended the π-conjugated structure. These results indicated that AIE-FA could be a potential FA probe. By contrast, TPE-NH2 exhibited only a 1.7-fold increase in fluorescence in PBS, supplemented with 30% DMSO (Figure S13). Therefore, AIE-FA was chosen for further studies because of its better aqueous solubility and larger fluorescence enhancement.

Figure 1

(a) Fluorescence spectra of AIE-FA and AIE-FA + FA (15 μM). (b) Selectivity of AIE-FA toward different relevant species in PBS, supplemented with 10% DMSO: (1) CH3CHO (5 μM), (2) CHOCHO (5 μM), (3) CH3COCHO (15 μM), (4) H2O2 (100 μM), (5) cysteine (1.0 mM), (6) glutathione (10 mM), (7) NaCl (100 mM), (8) KCl (50 mM), (9) NaHSO3 (200 μM), and (10) FA (15 μM).

The selectivity of AIE-FA toward FA was then investigated. AIE-FA was treated with various possible interfering substances in physiological conditions including acetaldehyde, glyoxal, methylglyoxal, hydrogen peroxide, cysteine, glutathione, sodium chloride, and potassium chloride. The fluorescence spectra were then obtained. There was negligible fluorescence increase at 530 nm for all of the tested substances, except for acetaldehyde, which exhibited an approximately 2.2-fold fluorescence enhancement at concentrations of 5 μM (Figure b). Particularly, the negligible fluorescence enhancements at 530 nm upon incubation with physiological relevant concentrations of KCl and NaCl implied that AIE-FA was stable without appreciable aggregation. In contrast, treatment with FA at 15 μM resulted in a ∼12-fold fluorescence enhancement. These results indicated that AIE-FA exhibited high selectivity, holding great potential in detecting FA in physiological conditions.

Then, pH effects on AIE-FA and its response to FA were also investigated. AIE-FA was essentially nonfluorescent in pH ranging from 6.8 to 9.0. After reacting with FA, a distinct fluorescent enhancement was observed in the pH range of 4.0–9.0 (Figure a). The slight decrease in fluorescence for pH smaller than 6.8 might be due to the slow hydrolysis of the Schiff base at acidic conditions.[28,29] These results indicated that our probe could be introduced to detect FA in physiological pH. The reversibility of AIE-FA toward FA was also investigated by adding NaHSO3 to the reaction product of AIE-FA and FA. The fluorescence intensity decreased by ∼90%, suggesting that the reaction between AIE-FA and FA was reversible (Figure S14 in the Supporting Information). Afterward, the response times of AIE-FA toward various concentrations of FA (0, 0.5, 5.0, and 15 μM) were then studied by recording the fluorescence signals at 530 nm in real time. The fluorescence emission was rapidly increased upon the addition of FA, reaching a plateau within 90 s for 15 μM FA (Figure b). This rapid response might be attributed to fast reaction kinetics between amine moieties and FA.[15] This rapid reaction rate would permit our probe to track transient FA in biological systems.

Figure 2

(a) Fluorescence intensity at 530 nm for AIE-FA (10 μM) with or without FA (10 μM) in buffers with different pH values (pH 4.0, 5.0, 6.0, 6.4, 6.8, 7.2, 7.6, 8.0, and 9.0), supplemented with 10% DMSO. (b) Real-time fluorescence responses of AIE-FA (10 μM) to different concentrations of FA (0, 0.5, 5.0, and 15 μM).

After demonstrating that AIE-FA afforded a fast response and excellent selectivity toward FA, its ability to quantify FA concentrations was then investigated. As shown Figure a, the fluorescence signals exhibited dynamic increase in response to increasing concentrations of FA in the range of 100 nM to 15 μM. A signal-to-background ratio as high as ∼12 was achieved across this concentration range. This dynamic range spanned over the intracellular FA concentrations of 1.5–4.0 μM, indicating its potential in imaging intracellular FA.[8] Moreover, a linear correlation between the fluorescence signals to the FA concentration was obtained in the range of 100 nM to 1 μM with a detection limit estimated to be 40 nM (Figure b). Such a low detection limit implied that the AIE-FA afforded the highest sensitivity for FA detection among existing fluorescence probes.[13−16]

Figure 3

(a) Fluorescence spectra of AIE-FA toward different concentrations FA (0, 0.1, 0.2, 0.4, 0.6, 0.8, 1.0, 2.0, 5.0, 10, and 15 μM) in PBS, supplemented with 10% DMSO. (b) Fluorescence intensity at 530 nm as a function of FA concentrations and linear fit between the fluorescence intensity and the concentration of FA (inset).

Reaction Mechanism Validation

To confirm the reaction mechanism, the reaction product of AIE-FA with FA was first verified with 1H NMR. The signals related to the protons on the amine groups of AIE-FA (chemical shift = 5 ppm) disappeared after reaction with FA (Figure a,b). Meanwhile, downfield was also observed for the ortho- and metaprotons on the amine-substituted aromatic rings of AIE-FA after reaction. These results indicated that the amine groups reacted with FA (Figure a,b). The reaction products of AIE-FA and FA were further confirmed by ESI analysis. There appeared a new peak upon reaction with FA, corresponding to the Schiff base derivatives, (calcd for C28H22N2 [M + H]+m/z, 387.2; found, 387.2). These results provided clear evidence that AIE-FA reacted with FA via a typical condensation reaction (Figure S15).

Figure 4

(a) 1H NMR spectrum of the reaction product of FA and AIE-FA. (b) 1H NMR spectrum of AIE-FA.

FA Imaging in Living Cells

Having demonstrated that AIE-FA could detect FA with fast response, high sensitivity, and high selectivity in vitro, its ability to track endogenous FA in living cells was then investigated, using HeLa cells as a model cell line. First, the cytotoxicity of the AIE-FA was studied with a WST-1 assay. The cells had over 90% viability for concentrations of AIE-FA up to 30 μM for 24 h, indicating that AIE-FA is highly biocompatible (Figure S16). Then, AIE-FA was introduced to detect endogenous and exogenous FA in HeLa cells. AIE-FA (10 μM) was incubated with HeLa cells for 1 h, and fluorescent images were then acquired with confocal microscopy after washing twice with PBS. There was evident fluorescence emission from the HeLa cells, indicating that AIE-FA could detect endogenous FA (Figure a1–a3 and S17). Afterward, the ability of AIE-FA to detect exogenous FA was demonstrated by adding 5.0 μM of FA to AIE-FA incubated HeLa cells. There was an approximately twofold enhancement in fluorescence as compared to that of the cells without the addition of FA (Figures b1–b3 and S18). These results indicated that AIE-FA could detect both endogenous and exogenous FA in living cells. Moreover, the inhibitory experiment was also performed to test probe specificity. Sodium bisulfide (NaHSO3) was chosen as the inhibitor because it could react with carbonyl of FA.[19,20] There was a dramatic decrease in the fluorescence signal for HeLa cells pretreated with NaHSO3 (200 μM) (Figure c1–c3). This was due to the consumption of intracellular FA by NaHSO3. In addition, the colocalization assay using Lyso-Tracker red revealed that AIE-FA did not specifically localize in lysosomes (Figure S19 in the Supporting Information). All together, these results indicated that AIE-FA could track FA in living cells with high sensitivity and high specificity, holding great potential for interrogation of the roles of FA in biology.

Figure 5

Confocal microscopy images. (a1–a3) HeLa cells incubated with AIE-FA (10 μM) for 1 h. (b1–b3) HeLa cells incubated with AIE-FA (10 μM) for 1 h and then treated with 5.0 μM FA for 0.5 h. (c1–c3) HeLa cells pretreated with NaHSO3 (200 μM) and then AIE-FA (10 μM) for 1 h. λex = 405 nm; λem = 480–570 nm. Scale bar = 20 μm.

To investigate whether AIE-FA could image endogenous FA in real time, confocal fluorescence images were obtained at different time points after the addition of AIE-FA (Figure ). There was no fluorescence signal before the addition of AIE-FA. With the addition of AIE-FA, there was evident fluorescence enhancement within 30 s. The fluorescence signals were gradually increased and became stable in ∼5 min (Figure S20 in the Supporting Information). This rapid increase in fluorescence indicated that AIE-FA could readily diffuse into cells and reacted with FA, generating fluorescence signals. The saturated fluorescence signal within 5 min indicated that our probe could rapidly reflect the endogenous FA concentrations. These results demonstrated that our probe could monitor FA concentrations in living cells with fast response and high sensitivity.

Figure 6

Confocal fluorescence images of HeLa cells with the addition of AIE-FA recorded at 0, 30, 90, 180, 270, and 330 s (a–f). λex = 405 nm; λem = 480–570 nm. Scale bar = 20 μm.

Conclusions

In summary, we have developed an AIE-based fluorescence sensing strategy for ultrasensitive and ultrafast detection of FA. By utilizing the aqueous solubility differences between amine-functionalized AIE and the Schiff base-modified AIE after reaction with FA, a highly sensitive and rapid assay for FA was readily achieved. The probe was demonstrated to possess rapid response, great selectivity, and high sensitivity toward FA in vitro. The ability of AIE-FA to image both endogenous and exogenous FA with high sensitivity and specificity was also demonstrated in living cells. Moreover, AIE-FA exhibited a fast response toward FA in real time in living cells. We believe that AIE-FA would hold great potential in sensing FA concentrations in living systems under various physiological conditions.

Experiment Section

Reagents and Materials

Zinc powder and titanium tetrachloride were bought from the commercial suppliers of J&K Chemical Co. Ltd. 4-Nitrophenyl phenyl ketone was purchased from Energy Chemical. Organic solvents were of analytical grade and obtained from Sinopharm Chemical Reagent Co. Ltd (Shanghai, China). All of the organic solvents were dried over 4 Å molecular sieves before use. Thin-layer chromatography was carried out using silica gel 60 F254 (Qingdao Ocean Chemicals, Qingdao, China). Silica gel (200–300 mesh) was used as the solid phases for column chromatography, and it was also purchased from Qingdao Ocean Chemicals (Qingdao, China). HeLa cells (cervical cancer cell lines) were obtained from the cell bank of Central Laboratory at Xiangya Hospital (Changsha, China). Ultrapure water with an electric resistance larger than 18.3 MΩ was obtained through a Millipore Milli-Q water purification system (Billerica, MA, USA).1H and 13C NMR spectra were recorded on a Bruker DRX-400 spectrometer. Electrospray ionization mass spectrometry (ESI–MS) was determined using Finnigan LCQ Advantage MAX (Thermo Finnigan). Fluorescence spectra were recorded at room temperature on F-7000 (Hitachi, Japan). Ultraviolet absorption spectra were measured on Shimadzu UV2450 (Japan). Confocal images of HeLa cells were obtained using an inverted fluorescence microscope (OLYMPUSFV-1000 MPE).

Synthetic Procedure

4-Benzoylbenzenamine was synthesized according to the reported protocol with slight modifications.[30]1H NMR (400 MHz, DMSO-d6): δ (ppm) 7.62–7.50 (m, 7H), 6.63 (d, J = 8.0 Hz, 2H); 6.18 (s, 2H). 13C NMR (100 MHz, DMSO-d6): 193.90, 154.27, 139.51, 133.07, 131.49, 129.24, 128.66, 124.20, 113.04.

AIE-FA was synthesized according to previous procedures with slight modifications.[31] Briefly, zinc powder (19 mmol) was added to 4-benzoylbenzenamine (10 mmol) in absolute tetrahydrofuran (THF, 20 mL). The mixture was cooled down to −30 °C. Then, titanium tetrachloride (1.0 mL) was gradually added. The mixture was further refluxed for 10 h. After the reaction, the mixture was extracted with ethyl acetate (EA; 40 mL × 3). The combined organic phase was washed with brine (40 mL × 2) and dried over Na2SO4. The mixture was then filtered and the filtrate was concentrated under reduced pressure. The desired residual was purified with a silica-gel column (petroleum ether/EA = 2:1) to yield AIE-FA as a yellow solid (2.52 g, yield: 70%). 1H NMR (400 MHz, DMSO-d6): δ (ppm) 7.62–7.50 (m, 7H), 6.63 (d, J = 8.0 Hz, 2H); 6.18 (s, 2H). 13C NMR (100 MHz, DMSO-d6): 147.30, 145.16, 139.04, 132.00, 131.93, 131.47, 127.92, 126.20, 113.71. MS (ESI) calcd for C26H21N [M + H]+m/z, 348.1; found, 348.1.

Synthesis of TPE-NH2

Zinc powder (38.5 mmol) was added to a mixture of 4-benzoylbenzenamine (10 mmol) and diphenylmethanone (10 mmol) in absolute THF (10 mL). The mixture was cooled down to −40 °C. Then, titanium tetrachloride (2.0 mL) was gradually added. Afterward, the mixture was slowly warmed up to room temperature and refluxed for 8 h. After reaction, the mixture was extracted with EA (50 mL × 3). The combined organic phase was washed with brine (50 mL × 2) and dried over Na2SO4. Then, the mixture was filtered and the filtrate was concentrated under reduced pressure. The desired residual was purified with a silica-gel column (petroleum ether/EA = 2:1) to yield TPE-NH2 as a yellow solid (2.02 g, 45% yield). 1H NMR (400 MHz, DMSO-d6): δ (ppm) 7.174–7.048 (m, 9H), 7.009 (d, J = 12.4 Hz, 4H); 6.935 (d, J = 7.2 Hz, 2H), 6.614 (d, J = 7.6 Hz, 2H), 6.315 (d, J = 7.6 Hz, 2H). 13C NMR (100 MHz, DMSO-d6): 147.77, 144.68, 144.46, 144.43, 141.68, 138.36, 132.12, 131.38, 131.30, 131.22, 130.82, 128.25, 128.14, 128.05, 126.73, 126.49, 126.47, 113.62. MS (ESI) calcd for C26H22N2 [M + H]+m/z, 363.18; found, 363.1.

In Vitro Assay and Cytotoxicity Study

A stock solution of AIE-FA (100 μM) was prepared with PBS (10 mM, pH 7.4), supplemented with 10% DMSO. A stock solution of TPE-NH2 (100 μM) was prepared with PBS (10 mM, pH 7.4) TPE-NH2 and AIE-FA with final concentrations of 10 μM were used throughout the in vitro experiments. A final concentration of 10 μM was used for FA, unless otherwise indicated. PBS (10 mM, pH 7.4) was used for all of the dilutions. Fluorescence spectra were recorded in the range of 390–710 nm, with an excitation wavelength of 370 nm. The excitation and emission slit widths were both 5 nm.

The cytotoxicity of AIE-FA against HeLa cells was studied using a WST-1 cell proliferation and cytotoxicity assay following the kit protocol. Briefly, cells were incubated with various concentrations of AIE-FA (0–30 μM) for 24 h. The cells were then washed with PBS and treated with WST-1 in PBS for 4 h. Cell viability was determined by measuring the absorbance at 450 nm with a Microplate Reader.

Fluorescence Imaging of Living Cells

HeLa cells were grown in RPMI-1640 supplemented with 10% fetal bovine serum, streptomycin (100 U/mL), and penicillin (100 U/mL) in an atmosphere of 5% CO2 at 37 °C. Cells were cultured on a 35 mm Petri dish with a 10 mm bottom well in a folate-free RPMI-1640 medium for 24 h, and then the dishes were washed with PBS three times.

For imaging endogenous FA, AIE-FA (10 μM) was added to HeLa cells and incubated at 37 °C for 1 h. Images were obtained after washing twice with PBS. For imaging exogenous FA, AIE-FA (10 μM) was first added to HeLa cells and incubated at 37 °C for 1 h. Subsequently, the cells were washed with PBS and incubated with 5.0 μM FA for another 0.5 h at 37 °C before imaging acquisition. To perform the inhibition experiment, NaHSO3 (200 μM) was first added to HeLa cells. Then, the cells were incubated with AIE-FA (10 μM) for 1 h before imaging acquisition. The imaging experiments were performed with three passages of cells, and three fields were imaged for each sample. For imaging endogenous FA in real time, one fluorescence image was obtained before the addition of AIE-FA. Then, fluorescence images were obtained at various time points (0, 30, 90, 180, 270, and 330 s) after the addition of AIE-FA. For colocalization assay, the prepared cells were incubated in 1 mL cell growth medium supplemented with Lyso-Tracker red for 20 min. After washing with PBS three times, AIE-FA was incubated with HeLa cells for 10 min, and then confocal fluorescence images of Lyso-Tracker red was obtained with an excitation wavelength of 559 nm and a collection channel of 590–640 nm. All of the fluorescence images were acquired using an oil objective lens of 100×, on an inverted confocal laser scanning fluorescence microscope equipped with an Olympus FV1000 confocal scanning system (Olympus IX81). An Ar+ laser (405 nm) was used as the excitation source, and a band pass filter (480–570 nm) was used for acquiring fluorescence images of cells incubated with AIE-FA. Maximum intensity projections were acquired for all of the imaging experiments. For imaging analysis, three ROIs of designated sizes were analyzed with ImageJ software.