The study of a curcumin-resembling molecular probe for the pH-responsive fluorometric assay and application in cell imaging

Talanta ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Contents lists available at ScienceDirect

Talanta journal homepage: www.elsevier.com/locate/talanta

The study of a curcumin-resembling molecular probe for the pH-responsive fluorometric assay and application in cell imaging Decheng Xiang a, Qinghua Meng a,n, Heng Liu a, Minbo Lan b, Gang Wei c a

School of Chemistry & Chemical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China Shanghai Key Laboratory of Functional Materials Chemistry, East China University of Science and Technology, Shanghai 200237, China c CSIRO Materials Science and Engineering, P.O. Box 218, Linfield, NSW 2070, Australia b

art ic l e i nf o

a b s t r a c t

Article history: Received 22 March 2015 Received in revised form 18 May 2015 Accepted 25 May 2015

A molecular probe of DibOH (2,6-bis(4-hydroxybenzylidene)cyclohexanone) designed as the ameliorant to the curcumin was prepared in which one enol unit was removed to avoid the intramolecular hydrogen bond and thus more rigidity and better coplanarity were achieved. Deprotonation of the phenolic hydroxyl group led to the negative charge and the intramolecular charge transfer (ICT) functioned accordingly. The DibOH probe in basic conditions exhibited the absorption peak at 468 nm, which indicated a red shift of 183 nm relative to that in acid conditions and was visible as a brown color to naked eyes. The ratio of fluorescence intensity at 612 nm to that at 520 nm (I612/I520) was calculated and a clear correlation with the pH value was obtained. The density functional theory (DFT) was performed on theoretical investigation of the acidic and basic forms of DibOH. The DibOH probe was evaluated in fluorescent imaging of human breast cancer cells (MCF-7) wherein the heterogeneous distribution of the intracellular pH microenvironment was observed & 2015 Elsevier B.V. All rights reserved.

Keywords: pH-responsive Molecular probe Fluorescence Cell imaging

1. Introduction The intracellular pH (pHi) is an important parameter for studying physiological activity including phagocytosis [1], apoptosis [2], ion transportation [3], muscle contraction [4] etc. Abnormal pHi values may cause anormogenesis of cells and thus be associated with some cancers [5] and the Alzheimer's disease [6]. Compared with the other pHi measurement methods such as microelectrodes, NMR, and absorbance spectroscopy, the fluorescence measurement of the pHi value by the pH-dependence fluorescent probe has the advantages of the high sensitivity, high spatial resolution, synchronicity, nondestructive inspection, simple operation and low cost [7]. Best et al. designed and investigated series of rhodamine-based fluorescent probes for optical measurements of pH, in which probes exhibited a spirocyclic quenching of the fluorophore [8]. Yuan et al. developed rhodamine amide derivatives that systematically tuned the pKa values and employed them to detect acidic pH variations in living Hela cells with a turn-on signal [9]. In their study, the ratiometric fluorescence analysis was applied as an optimized method for the n

Corresponding author. þ Fax: 86 2154741297. E-mail address: [email protected] (Q. Meng).

quantitative determination of biological events occurring under complex conditions by simultaneously recording fluorescence intensities at two wavelengths and calculating their ratios. Zhou et al. designed a ratiometric chemosensor for pH that was constructed by introducing a pH-insensitive coumarin fluorophore as a fluorescence resonance energy transfer (FRET) donor into a pHsensitive amino-naphthalimide derivative as the FRET acceptor [10]. Tang et al. prepared a complex probe in which the fluorescence lifetime of CdTeSe/ZnS quantum dots was modulated by NIR organic chromophores for generating reproducible pH-sensing nanomaterials. The resultant hybrid construct transfered the pH sensitivity of photolabile NIR cyanine dyes to long-lifetime pHinsensitive QDs and induced the fluorescence lifetime from 29 ns at pH 47 to 12 ns at pH o5 [11]. In our previous work, a carbocyanine type pH-responsive probe was investigated in which the mechanism came from the coplanarity of the configurational structure controlled by the pH-sensing intramolecular hydrogen bond. It indicated a good anti-photobleaching performance due to the chlorine atoms substituted in the structure and applied in fluorescence imaging of PANC-1 cells [12]. In view of the physiological safety to live cells, the fluorophor units of the probes were proposed preferentially to be derivated from natural substrates like curcumin analogs with the fluorescence property [13], physiological antioxidation activity and

http://dx.doi.org/10.1016/j.talanta.2015.05.066 0039-9140/& 2015 Elsevier B.V. All rights reserved.

Please cite this article as: D. Xiang, et al., Talanta (2015), http://dx.doi.org/10.1016/j.talanta.2015.05.066i

D. Xiang et al. / Talanta ∎ (∎∎∎∎) ∎∎∎–∎∎∎

2

O

O

O

O O

HO

H

O

O

OH

O

HO

OH

Scheme 1. The keto–enol tautomerization in the structure of curcumin.

tyrosinase inhibition [14]. However, the fluorescence efficiency would be greatly reduced by the keto–enol tautomerization facilitated by the intramolecular hydrogen bond (Scheme 1) which typically existed in the natural curcumin analogs [15], and the metal chelation of the enol unit [16] was another disadvantage for its application as a fluorescence probe. In this paper, the molecular probe of DibOH (2,6-bis(4-hydroxybenzylidene)cyclohexanone) designed as the ameliorant to the curcumin was prepared in which one enol unit was removed to avoid formation of the intramolecular hydrogen bond (Scheme 2). In this way better fluorescence efficiency might be achieved which was ratiocinatively attributed to the better rigidity and coplanarity of the molecular backbone [17]. Two phenolic hydroxyl groups were at the opposite sides of the DibOH probe and deprotonation of the hydroxyl group would lead to the negative charge and have p–π conjugation with the main body of the molecular probe. Thus the pH-responsive fluorescent probe based on the intramolecular charge transfer (ICT) functioned accordingly.

2. Materials and methods 2.1. General Chemicals were commercially available and used directly without further purification. NMR spectra were recorded on an AVANCE200 NMR Spectrometer (Bruker) in CDCl3 using (CH3)4Si as an internal standard, and J values were given in Hz. LC–MS was performed on an HP 1100 HPLC (Agilent), interfaced with a triplequadruple mass spectrometer equipped with an ESI ion source and an in-line diode-array ultraviolet–visible detector. High resolution electrospray ionization mass spectrometry (HR-ESIMS) were recorded on an Acquity UPLC-QTof-MS Premier instrument (Waters) operating in the positive electrospray ionization mode. The FT-IR spectra were recorded on an Avatar360 FT-IR spectrophotometer (Nicolet) in a water-purged environment. All spectra were obtained from compressed KBr pellets in which the samples were evenly dispersed. 32 scans were used to record each FT-IR spectrum with 2 cm  1 spectral resolution at room temperature. The probe DibOH was initially dissolved in DMSO and diluted by the phosphate buffered solutions of different pH values. A PHS25 pH-meter (Shanghai Leici Instrument Factory, China) was used to measure/calibrate the pH value of each sample. Deionized water was used. The salts were NaCl, KCl, CaCl2, MgCl2, Al(NO3)3  9H2O, LiCl, (NH4)2Fe(SO4)2  6H2O and ZnCl2. 2.2. Synthesis procedure Synthesis of 2,6-Bis(4-hydroxybenzylidene)cyclohexanone (DibOH) was involved by the Claisen–Schmidt condensation. 1.12 g

O

(10 mmol) of 4-hydroxybenzaldehyde and 0.49 g (5 mmol of cyclohexanone were dissolved in 30 mL tetrahydrofuran with 0.5 mL 98% sulfuric acid added dropwisely. The solution was stirred for 20 h at room temperature and was concentrated to about 5 mL in vacuum. The product was obtained by elution from silica using 1/9 (V/V) ethyl acetate/hexane as the brown oil, 1.48 g, 97% yield. 1H NMR (400 MHz, DMSO-d6) δ: 10.58 (s, 2H, –OH), 9.77 (s, 2H, – CH ¼), 7.74 (d, J ¼ 8.7 Hz, 4H, Ar-H), 6.82 (d, J ¼ 8.6 Hz, 4H, Ar-H), 2.83 (m, 4H, –CH2–), 1.68 (m, 2H, –CH2–). 13C NMR (100 MHz, DMSO-d6) δ: 191.59, 158.99, 136.47, 133.96, 129.08, 127.12, 116.50, 28.65, 11.03. ESIMS (ESI þ): 307.15(M þH). HR-ESIMS (ESI þ): 307.1346 (calcd. for C20H19O3, [MþH]: 307.1334). 2.3. The spectral measurement and theoretical calculation The UV–vis spectra were measured in standard quartz cuvettes on a TU1901 UV–vis spectrophotometer (PG Instruments Limited). The fluorescence spectra were measured in standard quartz cuvettes on a LS 50B spctrofluorimeter (Perkin Elmer). The relative fluorescence quantum yield (Φf) was calculated from the integrated area under the corrected emission curve with Rhodamine B as a reference [18]. The transient fluorescence spectrum of the DibOH molecule (20 μM, pH 7.4, Ex ¼385 nm, Em ¼520 nm) was measured in standard quartz cuvettes on a QM/TM/NIR system (PTI). The fluorescence lifetime was then calculated from the slope of the simulated decay curve according to Lakowicz's method [19]. All the measurements were carried out at 20 °C. The geometry optimizations were performed using density functional theory (DFT) with Becke's three-parameter hybrid exchange function, Lee–Yang–Parr gradient-corrected correlation functional (B3-LYP) and 6–31G basis set in the Gaussian 03 package [20–21]. No constraints to bonds/angles/dihedral angles were applied in calculations. 2.4. The photobleaching test The photobleaching test for the DibOH probe was conducted in a 20 μM solution (pH ¼7.4) irradiated by a UV source power (8 W, 365 nm) and evaluated by the variation of fluorescence intensity at 520 nm. The experiment was done at 20 °C. 2.5. Imaging of the human breast cancer cells The human breast cancer cells (MCF-7) seeded in a flat-bottomed culture bottle were maintained at 37 °C in Dulbecco's modified Eagle's medium (DMEM)/10% fetal bovine serum. The next day, cells were incubated with 20 μM DibOH probe (pH ¼7.4) for 30 min at 37 °C. The fluorescent images of cells were then obtained on an Olympus IX71 inverted fluorescence microscope (Olympus Optical Co., Tokyo, Japan) equipped with a 100 W mercury lamp for epi-illumination at 20 °C. The emission from the sample was collected and separated from Rayleigh and Ramanshifted light by a combination of filters, an excitation filter 380DF50 (Omega Optical), a dichroic mirror 475DCLP (Omega Optical), and an emission filter 520DF50 (Omega Optical) and then focused into a cooled Evolve 512 EM CCD camera (Roper). Image acquisition and processing were performed using Micro-Manager 1.4 imaging Software.

3. Results and discussion

HO

OH

DibOH Scheme 2. Molecular structure of the DibOH probe.

3.1. Synthesis and characterization of DibOH probe Synthesis of the DibOH probe (2,6-bis(4-hydroxybenzylidene) cyclohexanone) involved a typical acid catalyzed Claisen–Schmidt

Please cite this article as: D. Xiang, et al., Talanta (2015), http://dx.doi.org/10.1016/j.talanta.2015.05.066i

D. Xiang et al. / Talanta ∎ (∎∎∎∎) ∎∎∎–∎∎∎ O

O CHO

H+

+

2 HO

3

900

HO

OH

600

Scheme 3. Synthesis of the probe DibOH. 300

3.2. The pH-responsive spectra The effects of pH conditions on the UV–vis absorption of the DibOH probe are shown in Fig. 3. Two absorption peaks at 385 and 468 nm were found and they behaved in the opposite manners when the pH value increased from 2.0 to 12.0 (Fig. 3a). The DibOH probe in acidic and neutral conditions exhibited the absorption peak at 385 nm, which was in the near ultraviolet region and invisible to naked eyes. The probe in basic conditions, on the other hand, exhibited the absorption peak at 468 nm, which indicated a red shift of 183 nm and was visible as a brown color to naked eyes

50

T (%)

40

O

30

O

H

O

20

1900

1800

1700

1600

1500

1400

-1

Wavenumber (cm ) Fig. 1. The IR spectra of the DibOH (solid line) and curcumin (dotted line).

150

155 160 Time (ns)

165

80 0.12 60 0.08 40 0.04

0.00 300

20

Fluorescence intensity (a.u.)

100

0

0.16

Abs

condensation between one equivalent of cyclohexanone with two equivalents of 4-hydroxybenzaldehyde (Scheme 3). The reaction proceeded completely with a yield of nearly 100% and no intermediate of 1:1 mono condensation was detected by LC–MS method. The chemical structure of the product was confirmed using the method of 1H NMR, 13C NMR, IR, ESIMS and HR-ESIMS. In the IR spectrum, DibOH existed almost entirely in the keto form confirmed by the vibrating peak at 1670 cm  1, and by contrast the curcumin had the vibrating peak at 1627 cm  1 which indicated the typical enol structure (Fig. 1). The UV–vis absorption and fluorescence spectra of the DibOH probe excited by the light of 385 nm are shown in Fig. 2. The broad fluorescence peak centering at 520 nm was shown and a shoulder peak at 612 nm was found as well. The large stokes shift of 138 nm was worth mentioning, for it was able to avoid the interference from the resonance fluorescence or Rayleigh scattering of the excitation light. The fluorescence quantum yield was measured as 0.183 (pH ¼7.4) and indicated obvious enhancement in contrast with that of curcumin (0.011– 0.104 [22]), thus supporting our design concept as well. As the indepth understanding of the photochemistry on the mirror asymmetry between absorption and fluorescence spectra, a complex energy relaxation might work in its electronic exited state and lead to the deviation from the typical Frank–Condon fluorescence transition. The energy relaxation was further supported by the fluorescent life decay of 1.42 ns (Fig. 2), which indicated a remarkable improvement over the previous pH-responsive probe of the carbocyanine type (0.48 ns) prepared by our group [12].

0 350

400

450

500

550

600

650

700

Wavelength (nm) Fig. 2. The UV–vis (left) and fluorescent (right) spectra of the DibOH molecule (20 μM, pH¼ 7.4, Ex ¼385 nm). Inset: the fluorescent life decay curve ( : experimental data, : simulation line).

(Fig. 3b). Thus the probe could be regarded as a visual pH-responsive molecular probe. As the DibOH probe in pH neutral condition showed the absorption peak at 385 nm, we examined the pH-responsive fluorometric analysis excited by the light of this wavelength. The fluorescent spectra of probe DibOH in different pH conditions are presented in Fig. 4a. The fluorescence peak at 520 nm remained as a prominent peak before pH 7.5, but decreased sharply in the pH region of 8.0–9.5. A small fluorescence peak at 612 nm appeared as a shoulder peak and the intensity changed little with the pH values. As a precise technique for quantitative determination under complex conditions, the ratiometric analysis could be used to evaluate the pH-responsive fluorometry of the DibOH probe. The ratio of fluorescence intensity at 612 nm to that at 520 nm (I612/I520) was then calculated and a clear correlation with pH values was obtained (Fig. 4b). In consideration of both pH-responsive UV–vis absorption and fluorescent spectra, two structural states (DibOH-A and DibOH-B) possessing different conjugated systems were suggested in acid and basic conditions respectively (Fig. 5). The equilibrium constant (pKa) of the reaction was determined by Miltsov's method [23] as 9.47 using the data of fluorometric titration (Fig. 4). To penetrate in depth into the pH-responsive mechanism of the DibOH probe, the theoretical investigation of two forms based on density functional theory (DFT) was performed and the calculation results are shown in Table 1. The acid form (DibOH-A) was a neutral molecule and the benzene rings on both sides were noncoplanar, which indicated a dihedral angle of 41.3º in the optimized geometry. Thus two small isolated D–π–A conjugated systems were found to be responsible for the relative short wavelength (385 nm) of the maximum absorption. However, fluorescence quantum yield was measured as 0.183 (λem ¼520 nm), which behaved in a manner of the typical intramolecular charge transfer (ICT) fluorescence process. The basic form (DibOH-A) which was an anion molecule and the benzene rings on both sides were coplanar which indicated a dihedral angle of only 4.7º in the optimized geometry. Thus a large integrated D–π–A–π–D conjugated system was responsible for the long wavelength (468 nm) of the maximum absorption. However, fluorescence quantum yield (Φf) was measured as low as 0.052 (λem ¼612 nm), which might be due to the internal energy relaxation between resonance forms and electron interference in the ICT process (Fig. 5).

Please cite this article as: D. Xiang, et al., Talanta (2015), http://dx.doi.org/10.1016/j.talanta.2015.05.066i

D. Xiang et al. / Talanta ∎ (∎∎∎∎) ∎∎∎–∎∎∎

4

0.20

0.25

0.20 0.15

pH 12 pH=12.0

0.10

Abs

Abs

0.15

0.10

2

0.05

0.05 pH=7.5

0.00

pH=2.0

375

400

425

450

475

0.00

500

525

550

575

600

2

4

6

Wavelength (nm)

8

10

12

pH

Fig. 3. The absorption spectra (a) of the DibOH probe in different pH conditions (20 μM) and the pH dependence of the colorimetry (b) of the DibOH probe (20 μM, : the changes of absorption values at 385 nm; : the changes of absorption values at 468 nm). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

2.0

I612 / I520

1.6

150

100

1.2 0.8

Fluorescence intensity (a.u.)

Fluorescence intensity (a.u.)

125

pH=2.0

100

75

pH=7.5

50

25

0.0 60

475

500

525

550

575

600

625

650

2

4

6

8

10

12

pH

40

20

pH=12.0

0

0.4

80

0

675

2

4

6

8

10

12

pH

wavelength (nm)

Fig. 4. The fluorescent spectra (a) of probe DibOH in different pH conditions (20 μM, λex ¼ 385 nm) and the pH dependence of the fluorescence intensities, (b) of the DibOH probe (20 μM, : the changes of fluorescence intensities at 520 nm; : the changes of fluorescence intensities at 612 nm). Inset: The ratio of fluorescence intensity at 612 nm to that at 520 nm.

O

HO

DibOH-A

OH H + OH

O-

O

-

O

OH

O

OH

DibOH-B Fig. 5. The mechanism of DibOH probe's fluorescent response to pH.

Please cite this article as: D. Xiang, et al., Talanta (2015), http://dx.doi.org/10.1016/j.talanta.2015.05.066i

D. Xiang et al. / Talanta ∎ (∎∎∎∎) ∎∎∎–∎∎∎

5

Table 1 The geometry optimization and fluorescence properties of two forms. Optimized geometryn

Form

Φf

Dihedral angle (deg)

520 nm

612 nm

DibOH-A

41.3

0.183

0.051

DibOH-B

4.7

0.036

0.052

n

Density functional theory (DFT) method at B3LYP/6-31G level.

100

80

Fluorescence intensity (a.u.)

Fluorescence Intensity (a.u.)

100

60

40

20

0

80

60

40

20

0

none

Na(I)

K(I)

Ca(II)

Mg(II)

Al(III)

DibOH + metal ions

Li(I)

Fe(II)

Zn(II)

0

10

20

30

40

50

60

Time (min)

Fig. 6. (a) The fluorescence responses of DibOH probe (20 μM) to different metal ions (40 μM), pH¼7.4, λex ¼385 nm, λem ¼520 nm) and (b) the photobleaching experiment, UV source power 8 W, 365 nm.

Fig. 7. The living cell imaging of human breast cancer cells (MCF-7) stained by the DibOH probe solution (20 μM, pH 7.4, 60  objective lens), (a) the fluorescence image excited by 385 nm light, (b) the bright-field transmission image.

Please cite this article as: D. Xiang, et al., Talanta (2015), http://dx.doi.org/10.1016/j.talanta.2015.05.066i

D. Xiang et al. / Talanta ∎ (∎∎∎∎) ∎∎∎–∎∎∎

6

3.3. The stability As the body fluids usually contained a number of metal ions that participated in life activities individually, the fluorescence responses of the DibOH probe to different metal ions including Na (I), K(I), Ca(II), Mg(II), Al(III), Li(I), Fe(II) and Zn(II) were investigated and the results were presented in Fig. 6a. No significant variation was observed in fluorescence intensity, which confirmed that our molecular design was able to avoid the chelation reaction by the enol structure in the curcumin [16]. The photostability of the DibOH probe was carried out as well through the photobleaching experiment by a UV source power (8 W, 365 nm) and the fluorescence intensity remained stable in a limited time span for tens of minutes (Fig. 6b), which afford enough experiment time for the fluorescence measurement in body fluids or cell imaging.

distribution of the intracellular pH microenvironment was observed. The pH sensor could be applied for not only measuring cellular pH, but also for visualizing stimulus-responsive changes of intracellular pH values.

Acknowledgments This study was supported by the Science and Technology Commission of Shanghai Municipality (No. 0752nm016). The authors thank Dr. Wenjuan Yu (Instrumental Analysis Center, Shanghai Jiao Tong University) for assistance the mass spectrometry service.

References 3.4. The cell imaging We then evaluated bio-imaging applications of the DibOH probe in cell systems. The enhanced fluorescent images of human breast cancer cells (MCF-7) incubated with 20 μM DibOH probe were obtained on an Olympus IX71 inverted fluorescence microscope excited by the light of 385 nm (Fig. 7). Observation of the cellular morphology showed that the DibOH probe was transported across the membrane smoothly without interference with the cell morphology. The distribution of the fluorescence intensity indicated the heterogeneity of the intracellular pH microenvironment. The basophilic organelles like the Golgi apparatus indicated much more luminous than the cell sap background. Thus the DibOH probe might have potential application to distinguish cell organelles by means of the pH-responsive fluorometric assay, which would be studied in depth in the future work.

4. Conclusions In conclusion, we designed, synthesized, and characterized a pH-responsive molecular probe for the colorimetric and fluorometric assay. The ratio of fluorescence intensity at 612 nm to that at 520 nm (I612/I520) was calculated and a clear correlation with pH values was obtained. The density functional theory (DFT) was performed on theoretical investigation of the acidic and basic forms. The DibOH probe was evaluated in the fluorescent images of human breast cancer cells (MCF-7) and the heterogeneous

[1] M. Miksa, H. Komura, R. Wu, K.G. Shah, P. Wang, J. Immunol. Methods 342 (2009) 71–77. [2] D. Perez-Sala, D. Collado-Escobar, F. Mollinedo, J. Biol. Chem. 270 (1995) 6235–6242. [3] E. Liang, P. Liu, S. Dinh, Int. J. Pharm. 338 (2007) 104–109. [4] E.R. Chin, D.G. Allen, J. Physiol. 512 (1998) 831–840. [5] H. Izumi, T. Torigoe, H. Ishiguchi, H. Uramoto, Y. Yoshida, M. Tanabe, T. Ise, T. Murakami, Cancer Treatment Rev. 29 (2003) 541–549. [6] T.A. Davies, R.E. Fine, R.J. Johnson, et al., Biochem. Biophys. Res. Commun. 194 (1993) 537–543. [7] J. Han, K. Burgess, Chem. Rev. 110 (2010) 2709–2728. [8] Q.A. Best, R. Xu, M.E. McCarroll, L. Wang, D.J. Dyer, Org. Lett. 12 (2010) 3219–3221. [9] L. Yuan, W. Lin, Y. Feng, Org. Biomol. Chem. 9 (2011) 1723–1726. [10] X. Zhou, F. Su, H. Lu, et al., Biomaterials 33 (2012) 171–180. [11] R. Tang, H. Lee, S. Achilefu, J. Am. Chem. Soc. 134 (2012) 4545–4548. [12] D. Xiang, H. Liu, Q. Meng, M. Lan, G. Wei, Acta Chim. Sin. 71 (2013) 1435–1440. [13] S.M. Khopde, K.I. Priyadarsini, D.K. Palit, T. Mukherjee, Photochem. Photobiol. 72 (2000) 625–631. [14] Z.Y. Du, Y.F. Jiang, et al., Biosci. Biotechnol. Biochem. 75 (2011) 2351–2358. [15] S. Arimori, L.I. Bosch, et al., Tetrahedron Lett. 42 (2001) 4553–4555. [16] M. Borsari, E. Ferrari, R. Grandi, et al., Inorg. Chim. Acta 328 (2002) 61–68. [17] Q. Meng, H. Liu, S. Cheng, et al., Talanta 99 (2012) 464–470. [18] R.A. Velapoldi, H.H. Tønnesen, J. Fluoresc. 14 (2004) 465–472. [19] J.R. Lakowicz, Principles of Fluorescence Spectroscopy, third ed, Springer, New York, 2009. [20] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785–789. [21] M. Frisch, G. Trucks, H. Schlegh, et al., Gaussian 03, Gaussian, Inc., Pittsburgh, 2003. [22] C.F. Chignell, P. Bilskj, K.J. Reszka, et al., Photochem. Photobiol. 59 (1994) 295–302. [23] S. Miltsov, C. Encinas, J. Alonso, Tetrahedron Lett. 40 (1999) 4067–4068.

Please cite this article as: D. Xiang, et al., Talanta (2015), http://dx.doi.org/10.1016/j.talanta.2015.05.066i