Self-Assembled Fluorescent Nanoprobe Based on Forster Resonance

May 29, 2018 - Carbon monoxide (CO) is recognized as a biologically essential gaseous neurotransmitter that modulates many physiological processes in ...
0 downloads 0 Views 2MB Size
Download PDF
Letter Cite This: Anal. Chem. XXXX, XXX, XXX−XXX

pubs.acs.org/ac

Self-Assembled Fluorescent Nanoprobe Based on Forster Resonance Energy Transfer for Carbon Monoxide in Living Cells and Animals via Ligand Exchange Ruizhen Jia,† Pengfei Song,† Jingjing Wang,‡ Hengtang Mai,† Sixian Li,† Yu Cheng,‡ and Song Wu*,† †

School of Pharmaceutical Sciences, Wuhan University, Wuhan, Hubei 430072, P. R. China Shanghai East Hospital, The Institute for Biomedical Engineering & Nano Science, Tongji University School of Medicine, Shanghai, 200120, P. R. China

‡

S Supporting Information *

ABSTRACT: Carbon monoxide (CO) is recognized as a biologically essential gaseous neurotransmitter that modulates many physiological processes in living subjects. Currently reported fluorescent probes for CO imaging in cells basically utilize palladium related chemistry which requires complicated synthetic work. Herein we provide a new strategy to construct a fluorescent nanoprobe, NanoCO-1, based on the Forster resonance energy transfer (FRET) mechanism by entrapping the existing dirhodium complex as the energy acceptor and the CO recognition part, and a commonly used nitrobenzoxadiazole (NBD) dye as energy donor into a micelle formed by self-assembly. The exchange of ligands in the dirhodium complex by CO in the nanoprobe disrupts the FRET and leads to the turn-on of fluorescence. The merits of NanoCO-1 including good biocompatibility, selectivity, photostability, and low cytotoxity, render this nanoprobe ability to track CO in living cells, zebrafish embryo, and larvae. Our straightforward approach can be extended to establish the CO fluorescent probes based on adsorption of CO on a variety of metal derivatives.

C

attached to a metallic ruthenium(II) compound and released by CO has been reported.21 However, all the probes require complicated synthetic procedures. Therefore, there is an urgent need for the development of biocompatible and easily accessible fluorescent probes suitable for CO imaging in living subjects. It is widely recognized that CO is reliable to be adsorbed by some metals and metal derivatives under ambient conditions, which is exemplified by the well-known fact that CO is preferentially adsorbed by Fe2+ in heme over oxygen in animals. ́ In particular, Martinez-Má ñez’s group has developed a series of chromogenic sensors based on bimetallic rhodium for CO detection in air.22,23 As illustrated in Scheme 1a, the two axial ligands of the representative dirhodium sensor cis[Rh2(C6H4PPh2)2(O2CCH3)2](CH3COOH)2 (1(CH3COOH)2 in Scheme 1a) acetic acids can be exchanged by CO in air stepwise at remarkable low ppm levels, with strong color changes that can be even differentiated by the naked eye. However, compound 1(CH3COOH)2 and its dirhodium analogues are insoluble in aqueous solution, thus not applicable for detection of CO under physiological conditions. An ideal

arbon monoxide (CO) is a well-known gas pollutant in air that endangers human health with long-time exposure under high concentrations. Its paradoxical role as an essential gaseous neurotransmitter has recently attracted tremendous interest. Mainly produced by breakdown of heme in body,1 CO modulates many physiological processes like cytoprotective signaling and anti-inflammation in animals.2,3 Damage of homeostasis of CO in the human body has been proved to be closely related to some pathological changes4 and eventually leads to a variety of diseases such as cardiovascular disease,5 lung disease,6 hypertension,7 and Alzheimer’s disease,8 etc. Due to potentially therapeutic applications of CO, a variety of CO releasing molecules (CORMs) have been developed for treatment of different types of diseases.9,10 Meanwhile, developing probes suitable for detection of CO in living organisms is also of great significance. CO typically exhibits inactive chemical properties under mild conditions, which makes the design of reaction-based sensor highly challenging. Since Chang’s pioneered work in 2012,11 several palladium-related fluorescent probes for CO imaging in living cells have been developed, either based on hydrolysis of palladium12−15 or palladium-mediated Tsuji-Trost reactions.16−18 Since palladium catalysts have been used for bioconjugation for cysteine and lysine in vitro,19,20 the aforementioned probes may interfere normal bioactivities in vivo. Very recently, a fluorescent probe with a fluorophore © XXXX American Chemical Society

Received: March 29, 2018 Accepted: May 23, 2018

A

DOI: 10.1021/acs.analchem.8b01411 Anal. Chem. XXXX, XXX, XXX−XXX

Letter

Analytical Chemistry

amphiphilic copolymer mPEG-DSPE (1,2-dimyristoyl-s-glycero-3-phospho-ethanolamine-N-(methoxy(poly ethylene glycol)-2000) through self-assembly. Herein, we would like to report our effort on the fluorescent nanoprobe based on FRET by self-assembly via ligand exchange which can track CO in living cells, zebrafish embryos, and larvae. To the best of our knowledge, such a fluorescent nanoprobe for CO detection has not been reported before. We first synthesized dirhodium complex 1(CH3COOH)2 by simply refluxing rhodium(II) acetate dimer and triphenylphosphine in acetic acid/toluene. The two ligands of acetic acids are not very stable and can be readily replaced by two water molecules by reacting with the base of sodium carbonate in acetone.23 Considering that such replacement could occur in various solvent systems, we thus evaluated the optical performances of 1(CH3COOH)2 in the solvents like ethanol and colsovent (ethanol/water) in which the following encapsulation process used. 1(CH3COOH)2 is soluble in ethanol, with the absorption at 550 nm. As speculated, both CO gas and CORM-2 (a CO donor) led to the disappearance of this absorption peak, with color change from purple to yellow (Figure S1). The similar phenomena happened in the cosolvent of ethanol/water at different ratios (Figure S2). Although it is unlikely to determine the accurate replacement of acetic acids by water or/and ethanol, our observations proved the above hypothesis and suggested that CO has a stronger affinity to the metal rhodium than those ligands. As for the corresponding energy acceptor, a fluorophore of NBD 2 with a lipophilic tail of hexyl group intended to enhance its hydrophobicity for the following encapsulation was selected and simply synthesized via one-step amination using 4-chloro7-nitrobenzofurazan and n-hexylamine as the starting materials.24 The ethanol solution of compound 2 showed a strong broad emission around 540 nm upon excitation at 470 nm, which matches the absorption of compound 1(CH3COOH)2 (Figure S3). The good spectral overlap between the energy donor and acceptor enables the occurrence of FRET if they are close enough. Nanosized micelles formed by self-assembly have been widely used in the area of bioimaging and drug delivery for their ability to simply encapsulate hydrophobic chromophores or drugs.25−28 In our case, encapsulation was first performed by mixing the ethanol solution of 1(CH3COOH)2 or 2 in mPEGDSPE aqueous solution and following centrifugation. Under this condition, the lipophilic compound 1(CH3COOH)2 or 2 was entrapped into the hydrophobic interior of a micelle formed by self-assembly. As anticipated, micelles entrapped with compound 1(CH3COOH)2 (NanoP) showed a strong absorption peak at 550 nm but nonfluorescent, while micelles containing compound 2 (NanoF) exhibited intensive emission around 550 nm upon excitation at 470 nm (Figure 1a,b). The observations unambiguously proved the successful encapsulation of the corresponding compounds. In addition, while adding CO to the aqueous solution of NanoP, the absorption peak at 550 nm disappeared (Figure S4). This result is consistent with that of compound 1(CH3COOH)2 in ethanol/ water solutions, indicating that encapsulation process did not apparently affect CO-ligand exchange. A series of micelles entrapped with different molar ratios from 1:1 to 8:1 between compound 1(CH3COOH)2 and 2 were prepared to test their efficiency in energy transfer (represented by the quenching degree of the emission of compound 2). Compared to NanoF, the emission intensities of

Scheme 1. Illustration of the Ligand-CO Exchange Processes of the Dirhodium Complex 1(CH3COOH)2 (a) and Design Principle of the FRET Nanoprobe Formed by Self-Assembly for CO Sensing (b)

strategy is to transfer the color change during the process of the ligand exchange to the corresponding detectable fluorescence signals under physiological conditions. It is notable that compound 1(CH3COOH) 2 has a strong characteristic absorption peak around 550 nm, while it disappears in its both mono- and disubstituted CO products (1(CH3COOH)(CO) and 1(CO)2 in Scheme 1a, respectively) formed depending on CO concentrations.22 These observations give us a hint of possibility to design a CO fluorescent probe based on absorption change via CO-ligand exchange. We postulated that replacement of possibly instable axial ligands of acetic acids by other solvents like water might alter neither the absorption of the rhodium complex 1(X)(Y) (X, Y refer to axial ligand such as acetic acid, water, etc.) significantly nor the exchange with CO. Furthermore, 1(X)(Y) can serve both as recognition moiety and energy acceptor, plus appropriate fluorophores as complementary energy donors, to construct a Forster resonance energy transfer (FRET) system for CO detection. In this system, a commonly used dye nitrobenzoxadiazole (NBD) 2 was selected as the energy acceptor owing to its emission around 550 nm, which is compatible to the absorption of compound 1(CH3COOH)2 for FRET. As briefly depicted in Scheme 1b, the fluorescence of compound 2 is expected to be quenched by dirhodium complex 1(X)(Y) owing to the FRET. However, after the ligand exchange by CO, along with the mono- or/and disubstituted dirhodium complex produced with the strong absorption change, the FRET process is disrupted due to the mismatch emission of compound 2, thus leading to the turn-on of fluorescence. Apart from spectral overlap, a short distance within 10 nm between compound 2 and complex 1(X)(Y) is also an essential factor for the efficiency of the FRET. Considering that routine covalent linkage of the two counterparts for the energy transfer needs complicated and tedious synthetic work, we considered to encapsulate the two elements into a confined space like a nanosized material which can be conveniently produced by an commercially available B

DOI: 10.1021/acs.analchem.8b01411 Anal. Chem. XXXX, XXX, XXX−XXX

Letter

Analytical Chemistry

in a dose-dependent manner and reached the enhancement up to maximum of 3.6-fold in the presence of 50 μM of CORM-2. Addition of more CO, e.g., 100 μM, did not significantly increase the fluorescence (Figure 2a). The fluorescence

Figure 2. (a) Concentration-dependent fluorescence enhancement of NanoCO-1 (10 μM) in the presence of CORM-2 (0, 2, 4, 5, 6, 8, 10, 50, 100 μM) for 30 min in PBS (pH = 7.4) at 37.5 °C. (b) Relative fluorescence intensity of 10.0 μM NanoCO-1 with varying amount of CORM-2 (0−10 μM).

Figure 1. (a) Absorption and (b) the emission spectra of micelles NanoP (entrapped with 50 μM 1(CH3COOH)2), NanoF (entrapped with 10 μM 2), and NanoCO-1 (entrapped with the both) in PBS (pH = 7.4) at 37.5 °C. (c) Average hydrodynamic size of NanoCO-1 measured by DLC and (d) TEM image of NanoCO-1. Scale bar: 20 nm.

response also displayed a good linear relationship among 0− 10 μM (R = 0.99044) with a detection limit calculated to be 0.1792 μM (S/N = 3, n = 11) (Figure 2b). The low detection limit renders NanoCO-1 possibility to sense CO in vivo. Meanwhile, the fluorescence response showed a time-dependent manner, in which the enhancement reached a plateau within 40 min (Figure S8). The photostability of NanoCO-1 was investigated by measuring the fluorescence intensity of the probe excessively after irradiation of UV light (365 nm, 16 W) at various time points. As shown in Figure S9, after irradiation for 10 min to 2 h, minimal fluorescence variations were observed, suggesting the excellent photostability of NanoCO-1. NanoCO-1, as well as its response to CO, also displayed good stability toward pH value ranging from 5 to 10 as proved by the slight fluorescence variation (Figure S10), indicative of its ability to sense CO under different biological environments. NanoCO-1 also showed good selectivity to CO against physiologically related ions and biomolecules, as illustrated in Figure S11. The cytotoxicity of NanoCO-1 was then evaluated by the methyl thiazolyl tetrazolium (MTT) assay. As shown in Figure S12, NanoCO-1 showed negligible cytotoxicity toward Hela cells after treatment with various concentrations for 12 h and limited cytotoxicity at high concentrations for 24 h (cell viability, 85% for 50 μM and 80% for 100 μM, respectively), which is applicable for bioimagings. Encouraged by the positive results in vitro obtained above, we next tested imaging abilities of NanoCO-1 in living cells. A549 cells were cultured with NanoCO-1 for 40 min and then treated with different concentrations of CO for 40 min. As shown in Figure 3b, the uptake of NanoCO-1 could be observed by the weak fluorescence in cells, suggesting the good biocompatibility of this nanoprobe. Furthermore, Addition of CO led to the dose-dependent fluorescence enhancement in cells (Figure 3c,d). The similar test was performed in zebrafish embryos. Compared with the embryo treated with NanoCO-1 alone for 1 h, the fluorescence intensity of embryos remarkably increased when treated with 50 μM CORM-2 for another 1 h (Figure 4). In order to expand the bioapplication of the nanoprobe, we finally tried to track CO in zebrafish larvae by adding CO into 9-day old larvae pretreated with NanoCO-1 for 40 min. It is

micelles encapsulated with various ratios of the mixture of compound 1(CH3COOH)2 and 2 all showed different quenching levels. At the ratio of 1:1, only a slight quenching degree of 19.8% was observed. However, with the increasing ratios at 5:1 and 8:1 between 1(CH3COOH)2 and 2, the quenching degrees dramatically increased up to 87.7% and 89.5%, respectively (Figure S5). Our observations clearly demonstrated that the FRET process occurs between the compound 1(X)(Y) and compound 2 in the micelles, and its efficiency depends on the ratio of these two counterparts. In order to find the optimal nanoprobe for CO, we then compared the reactions between CO and the micelles encapsulated with 1(CH3COOH)2 and 2 at the ratios of 5:1 and 8:1 with higher FRET efficiency. As shown in Figure S5, the emission intensities of the micelles enhanced 58.4% and 45.6%, respectively. Given that higher loading of compound 1(X)(Y) may potentially bring about higher cytotoxicity, and plus the comparable fluorescence performance, we thus chose the micelle entrapped with the ratio of 5:1 for the further investigations and name it as NanoCO-1. The comparison of the absorption and emission among NanoP, NanoF, and NanoCO-1 is shown in Figure 1a,b. NanoCO-1 was characterized by dynamic light scattering (DLS) and transmission electron microscopy (TEM). The average size of NanoCO-1 was measured to be ∼9 nm and also confirmed by TEM (Figure 1c,d). The results demonstrated the formation of nanosized micelles via self-assembly and also explained the occurrence of FRET in the even smaller hydrophobic interior which ensures the distance close enough for energy transfer from the donor to the acceptor. NanoCO-1 also showed excellent stability, which was verified by the small variations of the average size and fluorescence, even after immersed in PBS buffer over 50 h (Figure S6). The lifetime of NanoCO-1 was calculated to be τ1 = 32.6 ns and τ2 = 1.8 ns, respectively, which is shorter than that of 41.0 and 1.9 ns in the presence of CO (Figure S7), indicating that the original FRET system in the nanoprobe was disrupted by CO. With NanoCO-1 in hand, we began to evaluate its response toward CO. The emission intensities of NanoCO-1 increased C

DOI: 10.1021/acs.analchem.8b01411 Anal. Chem. XXXX, XXX, XXX−XXX

Letter

Analytical Chemistry

based on FRET mechanism. This approach is fairly straightforward by simply entrapping the existing dirhodium complex as energy acceptor and the commonly used NBD dye as energy donor in a nanosized micelle formed by self-assembly. Furthermore, since it is very common that CO is adsorbed by various metal derivatives with optical or energy changes, our strategy can be extended to construct CO fluorescence probes with higher sensitivity and tunable emissions. We believe our studies will benefit the physiological and pathological studies of CO.

■

Figure 3. Confocal fluorescence images for CO of NanoCO-1 in A549 cells. Cells were incubated with NanoCO-1 (50 μM) for 40 min and then costained with (a) blank, (b) 0 μM CORM-2, (c) 5 μM CORM2, (d) 50 μM CORM-2 for 40 min. (e) Quantification of fluorescence intensities based on imaging data for parts b−d. Scale bar: 50 μm.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.8b01411. Experimental details, absorption and emission spectra, and ratio optimization between 1(CH3COOH)2 and 2 for NanoCO-1 (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Song Wu: 0000-0002-6876-730X Notes

The authors declare no competing financial interest.

■ ■

Figure 4. Fluorescent imaging of CO using NanoCO-1 in zebrafish embryos: (a) blank, (b) zebrafish embryos were treated with 5 μM NanoCO-1 for 1 h, (c) zebrafish embryos were pretreated with 5 μM NanoCO-1 for 1 h, and then cultured with 50 μM CORM-2 for another 1 h. (d) Quantification of imaging data. Scale bar: 500 μm.

ACKNOWLEDGMENTS This work is financially supported by the National Natural Science Foundation of China (Grant No. 21572174).

interesting to find that zebrafishes treated NanoCO-1 show discernible fluorescence in digest system, suggesting that NanoCO-1 is readily absorbed by zebrafish. When cultured with 50 μM CORM-2 for another 1 h, this part exhibited brighter fluorescence (Figure 5). Taken together, our nanoprobe NanoCO-1 can track CO in living cells, zebrafish embryo, and zebrafish larvae as well. In summary, we proposed a new strategy to construct a nanoprobe with good photostability, biocompatibility, and low cytotoxicity to realize CO imaging in living cells and animals

REFERENCES

(1) Motterlini, R.; Otterbein, L. E. Nat. Rev. Drug Discovery 2010, 9, 728−743. (2) Heinemann, S. H.; Hoshi, T.; Westerhausen, M.; Schiller, A. Chem. Commun. 2014, 50, 3644−3660. (3) Levitt, D. G.; Levitt, M. D. Clin. Pharmacol.: Adv. Appl. 2015, 7, 37−56. (4) Szabo, C. Nat. Rev. Drug Discovery 2016, 15, 185−203. (5) Nakao, A.; Kaczorowski, D. J.; Sugimoto, R.; Billiar, T. R.; McCurry, K. R. J. Clin. Biochem. Nutr. 2008, 42, 78−88. (6) Ryter, S. W.; Choi, A. K. Transl. Res. 2016, 167, 7−34. (7) Schroeder, R. A.; Ewing, C. A.; Sitzmann, J. V.; Kuo, P. C. Dig. Dis. Sci. 2000, 45, 2405−2410. (8) Premkumar, D. R.; Smith, M. A.; Richey, P. L.; Petersen, R. B.; Castellani, R.; Kutty, R. K.; Wiggert, B.; Perry, G.; Kalaria, R. N. J. Neurochem. 1995, 65, 1399−1402. (9) Ji, X.; Zhou, C.; Ji, K.; Aghoghovbia, R. E.; Pan, Z.; Wang, B. Angew. Chem., Int. Ed. 2016, 55, 15846−15851. (10) Ji, X.; Ji, K.; Chittavong, V.; Yu, B.; Pan, Z.; Wang, B. Chem. Commun. 2017, 53, 8296−8299. (11) Michel, B. W.; Lippert, A. R.; Chang, C. J. J. Am. Chem. Soc. 2012, 134, 15668−15671. (12) (a) Zheng, K.; Lin, W.; Tan, L.; Chen, H.; Cui, H. Chem. Sci. 2014, 5, 3439−3448. (13) Cao, Y.; Li, D.-W.; Zhao, L.-J.; Liu, X.-Y.; Cao, X.-M.; Long, Y.T. Anal. Chem. 2015, 87, 9696−9701. (14) Li, Y.; Wang, X.; Xie, X.; Li, M.; Niu, J.; Tong, L.; Tang, B. Anal. Chem. 2016, 88, 11154−11159. (15) Liu, K.; Kong, X.; Ma, Y.; Lin, W. Angew. Chem. 2017, 129, 13674−13677. (16) (a) Pal, S.; Mukherjee, M.; Sen, B.; Mandal, S. K.; Lohar, S.; Chattopadhyay, P.; Dhara, K. Chem. Commun. 2015, 51, 4410−4413.

Figure 5. Fluorescence imaging of CO in 9-day old zebrafish larvae (a) blank, (b) larvae were cultured in 10 μM NanoCO-1 for 40 min, (c) larvae were cultured in 10 μM NanoCO-1 for 40 min, and then transferred to fresh water containing 50 μM CORM-2 for 1 h. Scale bar: 200 μm. D

DOI: 10.1021/acs.analchem.8b01411 Anal. Chem. XXXX, XXX, XXX−XXX

Letter

Analytical Chemistry (17) Feng, W.; Liu, D.; Feng, S.; Feng, G. Anal. Chem. 2016, 88, 10648−10653. (18) Feng, S.; Liu, D.; Feng, W.; Feng, G. Anal. Chem. 2017, 89, 3754−3760. (19) Vinogradova, E. V.; Zhang, C.; Spokoyny, A. M.; Pentelute, B.; Buchwald, S. L. Nature 2015, 526, 687−691. (20) Lee, H. G.; Lautrette, G.; Pentelute, B. L.; Buchwald, S. L. Angew. Chem., Int. Ed. 2017, 56, 3177−3181. (21) Torre, C. D.; Toscani, A.; Marín-Hernández, C.; Robson, J. A.; Terencio, M. A.; White, J. A.; Alcaraz, M. J.; Wilton-Ely, J. D.; Martínez-Máñez, R.; Sancenón, F. J. Am. Chem. Soc. 2017, 139, 18484−18487. (22) Esteban, J.; Ros-Lis, J. V.; Martínez-Máñez, R.; Marcos, M. D.; Moragues, M. Z.; Soto, J.; Sancenón, F. Angew. Chem., Int. Ed. 2010, 49, 4934−4937. (23) Moragues, M. E.; Esteban, J.; Ros-Lis, J. V.; Martínez-Máñez, R.; Marcos, M. D.; Martínez, M.; Soto, J.; Sancenón, F. J. Am. Chem. Soc. 2011, 133, 15762−15772. (24) Quarti, C.; Monteoleone, F. V.; Botta, C.; Daita, V.; Perdicchia, D.; Buttero, P. D.; Zoppo, M. D. Chem. Phys. Lett. 2014, 610, 357− 362. (25) Li, K.; Liu, B. Chem. Soc. Rev. 2014, 43, 6570−6597. (26) Zhao, C.; Zhang, X.; Li, K.; Zhu, S.; Guo, Z.; Zhang, L.; Wang, F.; Fei, Q.; Luo, S.; Shi, P.; Tian, H.; Zhu, W.-H. J. Am. Chem. Soc. 2015, 137, 8490−8498. (27) Swaminathan, S.; Garcia-Amoros, J.; Fraix, A.; Kandoth, N.; Sortino, S.; Raymo, F. M. Chem. Soc. Rev. 2014, 43, 4167−4178. (28) Swaminathan, S.; Fowley, C.; McCaughan, B.; Cusido, J.; Callan, J. F.; Raymo, F. M. J. Am. Chem. Soc. 2014, 136, 7907−7913.

E

DOI: 10.1021/acs.analchem.8b01411 Anal. Chem. XXXX, XXX, XXX−XXX