bis(2-quinolylmethyl)ethylenediaminediacetic acids (bqendas), tqen–edta hybrid molecules as...

DaltonTransactions

PAPER

Cite this: Dalton Trans., 2014, 43,10013

Received 25th January 2014,Accepted 10th April 2014

DOI: 10.1039/c4dt00261j

www.rsc.org/dalton

Bis(2-quinolylmethyl)ethylenediaminediaceticacids (BQENDAs), TQEN–EDTA hybrid moleculesas fluorescent zinc sensors†

Yuji Mikata,*a,b Saaya Takeuchi,b Hideo Konno,c Satoshi Iwatsuki,d Sakiko Akaji,e

Ikuko Hamagami,e Masato Aoyama,f Keiko Yasuda,f Satoshi Tamotsuf andShawn C. Burdetteg

Molecular hybrids of TQEN (N,N,N’,N’-tetrakis(2-quinolylmethyl)ethylenediamine) and EDTA (ethylenedi-

amine-N,N,N’,N’-tetraacetic acid) were examined as fluorescent Zn2+ sensors. Upon the addition of Zn2+,

N,N-BQENDA (N,N-bis(2-quinolylmethyl)ethylenediamine-N’,N’-diacetic acid, 1a) exhibits a 30-fold

emission enhancement at 456 nm (λex = 315 nm, ϕZn = 0.018) in buffer (HEPES, pH = 7.5, 100 mM KCl).

The fluorescence enhancement is Zn2+-specific as Cd2+ induces much smaller increases (ICd/I0 = 5

and ICd/IZn = 16%). These spectroscopic properties, as well as the excellent water-solubility, represent a

significant improvement compared to the parent TQEN sensor (ϕZn = 0.007, ICd/IZn = 64%). The isoquino-

line analog N,N-1-isoBQENDA (N,N-bis(1-isoquinolylmethyl)ethylenediamine-N’,N’-diacetic acid, 1b)

possesses a similar Zn2+ fluorescence response to the parent 1-isoTQEN (N,N,N’,N’-tetrakis(1-isoquinolyl-

methyl)ethylenediamine) sensor, but exhibits diminished fluorescence intensity. Apo 1a and 1b extract

more than 50% of the Zn2+ from an equimolar amount of [Zn(TPEN)]2+ (TPEN = N,N,N’,N’-tetrakis(2-pyri-

dylmethyl)ethylenediamine) or [Zn(EDTA)]2−, whereas TPEN and EDTA cannot effectively remove Zn2+

from [Zn(1a)] and [Zn(1b)]. The reduction of steric crowding in [Zn(TQEN)]2+ resulting from the substi-

tution of two quinolines with carboxylates enhances the interaction between the metal ion and the

remaining quinoline nitrogen atoms. The stronger bonding interaction results in enhanced emission

intensity, Zn2+ selectivity and metal ion affinity. This is in contrast to [Zn(1-isoTQEN)]2+ where the isoqui-

noline-carboxylate replacement does not relieve any coordination distortion, therefore no significant

changes in fluorescence or metal binding properties are observed.

Introduction

Zinc is involved in numerous biological processes includingenzymatic transformations, gene replication and transcription,immune responses and neurotransmission.1–5 The mobile

(free) zinc pool in cells regulates the free intracellular zinc con-centration and releases zinc in response to biological stimuli.Real-time, high resolution tracking of mobile zinc in livingcells has been accomplished with fluorescent probes.4–12 Snap-shots of Zn2+ distribution after introducing various stimuli canreveal biological roles and response mechanisms. Althoughseveral fluorescent probes have been successfully applied tocellular imaging, improvements in Zn2+ selectivity, protoninsensitivity, modulation of metal binding affinity, and con-trolling intracellular distributions remain topics of interest.

TSQ and Zinquin (Chart 1) are prototypical quinoline-basedfluorescent Zn2+ sensors. The synthetic accessibility andoptical properties of quinoline derivatives have made themattractive candidates for fluorescent sensors.13–19 Our grouprecently described TQEN (N,N,N′,N′-tetrakis(2-quinolylmethyl)-ethylenediamine), a TPEN-based (TPEN = N,N,N′,N′-tetrakis-(2-pyridylmethyl)ethylenediamine) fluorescent Zn2+ probe(Chart 2).20 Substituting quinoline with isoquinoline provides1-isoTQEN (N,N,N′,N′-tetrakis(1-isoquinolylmethyl)ethylenedi-amine, Chart 2), which exhibits enhanced emission intensity

†Electronic supplementary information (ESI) available: Experimental proceduresfor compound preparation and Fig. S1–S41. See DOI: 10.1039/c4dt00261j

aKYOUSEI Science Center, Nara Women’s University, Nara 630-8506, Japan.

E-mail: [email protected]; Fax: +81 742 20 3095; Tel: +81 742 20 3095bDepartment of Chemistry, Faculty of Science, Nara Women’s University, Nara

630-8506, JapancNational Institute of Advanced Industrial Science and Technology (AIST),

1-1-1 Higashi, Tsukuba, Ibaraki 305-8565, JapandDepartment of Chemistry of Functional Molecules, Faculty of Science and

Engineering, Konan University, 8-9-1 Okamoto, Higashinada, Kobe 658-8501, JapaneOptical Spectroscopy Team, Horiba, Ltd, 2 Miyanohigashi, Kisshoin, Minami-ku,

Kyoto 601-8510, JapanfDepartment of Biological Sciences, Faculty of Science, Nara Women’s University,

Nara 630-8506, JapangDepartment of Chemistry and Biochemistry, Worcester Polytechnic Institute,

Worcester, MA 01609-2280, USA

This journal is © The Royal Society of Chemistry 2014 Dalton Trans., 2014, 43, 10013–10022 | 10013

Publ

ishe

d on

10

Apr

il 20

14. D

ownl

oade

d by

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es o

n 25

/10/

2014

06:

00:2

9.

View Article OnlineView Journal | View Issue

www.rsc.org/dalton

http://dx.doi.org/10.1039/c4dt00261j

http://pubs.rsc.org/en/journals/journal/DT

http://pubs.rsc.org/en/journals/journal/DT?issueid=DT043026

and Zn2+ selectivity (ICd/IZn = 14%);21 however, the pooraqueous solubility of TQEN and 1-isoTQEN limits the potentialbiological applications for these probes.

To address the deficiencies of TQEN and 1-isoTQEN, wehave prepared and characterized N,N-BQENDA (N,N-bis(2-qui-nolylmethyl)ethylenediamine-N′,N′-diacetic acid, 1a, Chart 2)and N,N-1-isoBQENDA (N,N-bis(1-isoquinolylmethyl)-ethylene-diamine-N′,N′-diacetic acid, 1b, Chart 2), which substitute qui-noline groups with carboxylate ligands. The correspondingN,N′-isomers, 2a and 2b (Chart 2), also were investigated to

ascertain additional structure-dependent fluorescenceresponses to Zn2+ binding. The pyridine analogs of 1 and 2were investigated by the Delangle22,23 and Orvig24,25 groupsrespectively, to evaluate the differential coordination stabilityand/or separation of trivalent metal cations including acti-nides and lanthanides. In HEPES buffer solution, the (iso)qui-noline probes 1a and 1b exhibited a distinct Zn2+-specificspectroscopic response. The strong fluorescence intensity of 1aand 1b at longer wavelengths is attributed to the N,N-configur-ation of these systems, which might induce an excimer-likeinteraction between adjacent quinoline chromophores uponcomplex formation.21 The rapid, high affinity Zn2+ binding of1a and 1b was demonstrated using TPEN and EDTA as poten-tial competitors.

ExperimentalGeneral

All reagents and solvents used for synthesis were obtainedfrom commercial sources and used as received. TPEN was pre-pared according to a literature method.26 N,N-Dimethylform-

Chart 1

Chart 2

Paper Dalton Transactions

10014 | Dalton Trans., 2014, 43, 10013–10022 This journal is © The Royal Society of Chemistry 2014

Publ

ishe

d on

10

Apr

il 20

14. D

ownl

oade

d by

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es o

n 25

/10/

2014

06:

00:2

9.

View Article Online


amide (DMF, Dojin) was spectral grade (Spectrosol). Allaqueous solutions were prepared using Milli-Q water (Milli-pore). Zinc acetate was used as a Zn2+ source for the spectro-scopic measurements. 1H NMR (300 MHz) and 13C NMR(75.5 MHz) spectra were recorded on a Varian GEMINI 2000spectrometer and referenced to internal Si(CH3)4 or solventsignals. UV-vis and fluorescence spectra were measured on aJasco V-660 spectrophotometer and a Jasco FP-6300 spectro-fluorometer, respectively. Fluorescence quantum yields weremeasured on a HAMAMATSU photonics C9920-02 absolute PLquantum yield measurement system. Fluorescence lifetimeswere measured on a HORIBA FluoroCube 5000U system.

N,N-Bis(2-quinolylmethyl)ethylenediamine-N′,N′-diacetic acid(N,N-BQENDA, 1a)

Compound 5a (170 mg, 0.30 mmol) was dissolved in formicacid (10 mL) and stirred for 1.5 days at room temperature Afterevaporation of the solvent, the residue was recrystallized fromCH3CN to give N,N-BQENDA (1a) as a light yellow powder(84 mg, 0.18 mmol, 60%). 1H NMR (DMSO-d6): δ 8.30 (d, J =8.7 Hz, 2H), 7.96 (d, J = 8.2 Hz, 2H), 7.93 (d, J = 8.7 Hz, 2H),7.73 (d, J = 8.5 Hz, 2H), 7.72 (dd, J = 8.5, 7.3 Hz, 2H), 7.55 (dd,J = 7.9, 7.0 Hz, 2H), 4.00 (s, 4H), 3.38 (s, 4H), 2.93 (m, 2H), 2.70(m, 2H). 13C NMR (DMSO-d6): δ 172.0, 159.5, 146.5, 136.0,129.1, 128.2, 127.4, 126.6, 125.8, 120.9, 60.3, 54.9, 51.9, 51.2.Anal. calcd for C26H26.4N4O4.2 (1a·0.2H2O): H, 5.76; C, 67.58; N,12.12. Found: H, 5.72; C, 67.40; N, 12.16. HRMS (ESI-MS) m/z:[M + Na]+ calcd for C26H26N4O4Na 481.1851; found 481.1820.

N,N-Bis(1-isoquinolylmethyl)ethylenediamine-N′,N′-diaceticacid (N,N-1-isoBQENDA, 1b)

Compound 5b (272 mg, 0.48 mmol) was dissolved in formicacid (10 mL) and stirred for 1.5 days at room temperature.After evaporation of the solvent, the residue was recrystallizedfrom CHCl3–ether (1 : 10) to give N,N-1-isoBQENDA (1b) as awhite powder (136 mg, 0.30 mmol, 63%). 1H NMR (DMSO-d6):δ 8.83 (d, J = 5.8 Hz, 2H), 8.10 (d, J = 8.2 Hz, 2H), 7.90 (d, J =8.2 Hz, 2H), 7.73 (d, J = 5.5 Hz, 2H), 7.68 (dd, J = 7.6, 7.3 Hz,2H), 7.30 (dd, J = 7.6, 7.3 Hz, 2H), 4.30 (s, 4H), 3.26 (s, 4H),2.86 (br., 2H), 2.68 (br., 2H). 13C NMR (DMSO-d6): δ 172.0,162.6, 157.3, 140.7, 135.4, 129.8, 126.5, 126.4, 126.1, 120.2,58.8, 54.2, 51.4, 49.8. Anal. calcd for C26H26.4N4O4.2

(1b·0.2H2O): H, 5.76; C, 67.58; N, 12.12. Found: H, 5.80; C,67.44; N, 12.07. HRMS (ESI-MS) m/z: [M + Na]+ calcd forC26H26N4O4Na 481.1851; found 481.1820.

N,N′-Bis(2-quinolylmethyl)ethylenediamine-N,N′-diacetic acid(N,N′-BQENDA, 2a)

Compound 7a (198 mg, 0.35 mmol) was dissolved in formicacid (11 mL) and stirred for 1.5 days at room temperature.After evaporation of the solvent, the residue was recrystallizedfrom EtOH–ether (2 : 15) to give N,N′-BQENDA (2a) as a whitepowder (117 mg, 0.26 mmol, 74%). 1H NMR (DMSO-d6): δ 8.20(d, J = 8.5 Hz, 2H), 7.91 (d, J = 9.5 Hz, 4 H), 7.71 (ddd, J = 8.5,7.0, 1.2 Hz, 2H), 7.59 (d, J = 8.5 Hz, 2H), 7.54 (m, 2H), 4.04 (s,4H), 3.38 (s, 4H), 2.85 (s, 4H). 13C NMR (DMSO-d6): δ 171.8,

159.7, 135.9, 129.1, 128.0, 127.4, 126.6, 125.8, 120.7, 60.1, 54.8,51.6. Anal. calcd for C26H27N4O4.5 (2a·0.5H2O): H, 5.82; C,66.80; N, 11.98. Found: H, 5.80; C, 66.94; N, 11.96. HRMS(ESI-MS) m/z: [M + Na]+ calcd for C26H26N4O4Na 481.1851;found 481.1820.

N,N′-Bis(1-isoquinolylmethyl)ethylenediamine-N,N′-diaceticacid (N,N′-1-isoBQENDA, 2b)

Compound 7b (200 mg, 0.35 mmol) was dissolved in formicacid (10 mL) and stirred for 1.5 days at room temperature.After evaporation of the solvent, the residue was recrystallizedfrom EtOH–ether (1 : 5) to give N,N′-1-isoBQENDA (2b) as acream colored powder (38 mg, 0.08 mmol, 23%). 1H NMR(DMSO-d6): δ 8.43 (d, J = 8.5 Hz, 2H), 8.30 (d, J = 6.0 Hz, 2H),7.90 (d, J = 7.9 Hz, 2H), 7.69 (d, J = 6.1 Hz, 4H), 7.69 (dd, J =8.9, 6.1 Hz, 4H), 7.46 (dd, J = 7.2, 7.2 Hz, 2H), 4.30 (s, 4H), 3.29(s, 4H), 2.79 (s, 4H). 13C NMR (DMSO-d6): δ 171.7, 157.4, 140.5,135.3, 129.9, 126.7, 126.6, 126.3, 125.7, 120.1, 57.9, 54.7, 50.9.Anal. calcd for C26H26.4N4O4.2 (2b·0.2H2O): H, 5.76; C, 67.58;N, 12.12. Found: H, 5.83; C, 67.34; N, 12.14. HRMS (ESI-MS)m/z: [M + Na]+ calcd for C26H26N4O4Na 481.1851; found481.1820.

Fluorescent microscopic analysis for detection of intracellularZn2+ concentration change

PC-12 rat adrenal pheochromocytoma cells were cultured inRPMI 1640 supplemented with 5% fetal bovine serum (FBS),10% horse serum (HS) and 1% penicillin–streptomycin (PS).All cells were maintained in a humidified incubator at 37 °Cand 5% CO2. The media was changed to those containing100 µM of ligands and 2% (v/v) DMSO, and incubated for 4 h.Cells were rinsed with FBS, soaked in the growth media, thenanalyzed with a fluorescence microscope. Zn–pyrithione(Tokyo Chemical Industry Co., Ltd; final concentration: 90 µM)in DMSO was added to the media during analysis and the fluo-rescence enhancement was monitored.

Results and discussionLigand synthesis

Ligands N,N-BQENDA (1a), N,N-1-isoBQENDA (1b), N,N′-BQENDA (2a) and N,N′-1-isoBQENDA (2b) were synthesized fromN-Boc-ethylenediamine or ethylenediamine by the route shownin Scheme 1. All new compounds were characterized by 1H and13C NMR, and the purity was verified by elemental analysis.

Zn2+-induced changes in the absorbance and fluorescence of1a and 2a

Spectral measurements were conducted under simulated phys-iological conditions (50 mM HEPES, 100 mM KCl; pH = 7.50)with 34 µM of sensor at 25 °C. Upon addition of Zn2+, theabsorbance of 1a increased at 303 nm and decreased between250–280 nm, with a well-defined isosbestic point at 285 nm(Fig. 1a). The absorbance maxima red shifted slightly from312 nm to 315 nm upon formation of the Zn2+ complex. No

Dalton Transactions Paper


Publ

ishe

d on

10

Apr

il 20

14. D

ownl

oade

d by

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es o

n 25

/10/

2014

06:

00:2

9.

View Article Online


spectral changes were observed above the addition of 1 equiv.of Zn2+, indicating the formation of a 1 : 1 complex with 1a(Fig. S1a†).

Fig. 1b shows the Zn2+-induced fluorescence emissionchanges of 1a. Upon excitation at 315 nm, the apo-ligand 1aexhibits negligible fluorescence. After the addition of 1 equiv.of Zn2+, the emission at 456 nm increases 32-fold. The fluo-rescence signal becomes saturated with the addition of1 equiv. of Zn2+, which suggests the formation of a 1 : 1complex (Fig. S1b†).

The fluorescence spectra of [Zn(1a)] and [Zn(TQEN)]2+ inDMF–water (1 : 1) at the same concentration are shown inFig. 1c. Since only the short-wavelength region changesslightly, the solvent effects with 1a appear to be minimal

(Fig. S1c†). Each spectrum in Fig. 1c consists of two emissionbands centred in the ranges 350–400 and 400–480 nm. Therelative intensity of these two emission bands is altered by thereplacement of two quinolines with carboxylates. The long-wavelength emission comes from an excimer-like interactionof quinoline rings, especially in the N,N-bis(2-quinolylmethyl)configuration.21 The reduced steric hindrance in [Zn(1a)] com-pared to [Zn(TQEN)]2+ promotes a quinoline–quinoline orbitalinteraction that exhibits strong excimer-like emission at longwavelengths. No concentration-dependent fluorescence spec-tral changes of [Zn(1a)] were observed in the 1–50 µM range.As a result, the fluorescence quantum yield of [Zn(1a)] inDMF–water (ϕ = 0.019) is 2.5-fold greater than that of[Zn(TQEN)]2+ (ϕ = 0.007)20 in the same solvent (Table 1).

Scheme 1

Fig. 1 (a) UV-vis absorption and (b) fluorescence (λex = 315 nm) spectra of 34 µM N,N-BQENDA (1a) in HEPES buffer (pH = 7.5, 100 mM KCl) at25 °C in the presence of various concentrations of Zn2+ between 0 and 68 µM. (c) Comparison of fluorescence spectra between 34 µM 1a (red, λex =315 nm (Abs = 0.21)) and TQEN (blue, λex = 317 nm (Abs = 0.46)) in DMF–H2O (1 : 1) at 25 °C in the presence of 1 equiv. of Zn2+.



Publ

ishe

d on

10

Apr

il 20

14. D

ownl

oade

d by

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es o

n 25

/10/

2014

06:

00:2

9.

View Article Online


In contrast to N,N-BQENDA (1a), N,N′-BQENDA (2a) has adifferent substitution pattern of quinolines and carboxylates,which impairs the Zn2+-sensing properties (Fig. 2). Althoughthe Zn2+-induced absorbance changes of 2a are similar to 1aand the titrations also indicate the formation of a 1 : 1 complex(Fig. S2†), the emission enhancement upon Zn2+ addition isvery small. The emission profile of [Zn(2a)] only consists ofone emission band a at short wavelength (λem = 377 nm) dueto the absence of a N,N-bis(2-quinolylmethyl) moiety. Theintensity of the short-wavelength band is slightly weaker thanthe corresponding [Zn(TQEN)]2+ emission, even consideringthe different number of quinoline rings and possible solventeffects.

Zn2+-induced changes in the absorbance and fluorescence of1b and 2b

Replacement of the 2-quinolylmethyl moieties in TQEN with1-isoquinolylmethyl substituents afforded 1-isoTQEN thatexhibited (i) an increase in fluorescence intensity, (ii) shiftingof the excitation/emission to longer wavelengths, (iii) anenhanced Zn2+ selectivity and (iv) a stronger metal bindingaffinity.21 With the observed broad improvements in spectraland metal binding properties, integrating isoquinoline fluoro-phores into a chelator framework with higher water solubilitycould provide optimized fluorescent sensors.

The Zn2+-induced absorbance changes of N,N-1-isoBQENDA(1b) are very small (Fig. 3a). The absorbance peak at 311 nmincreased slightly on the addition of Zn2+ and a distinct isos-bestic point was observed at 300 nm. The spectral changesalso become saturated after the addition of 1 equiv. of Zn2+,indicating the formation of a 1 : 1 complex (Fig. S3a†).

Fig. 3b shows the fluorescence changes for the titration of1b with Zn2+ (λex = 324 nm). A significant emission increasewas observed at 469 nm, similar to the analogous tetrakis(iso-quinoline), 1-isoTQEN. The quantum yield of [Zn(1b)] (ϕ =0.023) only increases 1.5-fold compared to 1b (ϕ = 0.017),however, because of a significant decrease in emission at351 nm.

The effect of the carboxylates on the spectroscopic pro-perties of 1b was assessed by comparison with 1-isoTQEN(Fig. 3c). The emission spectra of [Zn(1b)] in HEPES buffer and

Table 1 Fluorescence properties for the Zn2+ complexes of 1, TQENand 1-isoTQEN in DMF–water (1 : 1)

N,N-BQENDA(1a) TQEN

N,N-1-isoBQENDA(1b) 1-isoTQEN

λex (nm) 317 317 328 328λem (nm) 456 383 351, 469 357, 477Ilong/Ishort — — 1.32 1.66IZn/I0 32a 23 17 (at 469 nm)a 12 (at

477 nm)ICd/I0 5.0a 14 1.4a 1.6ICd/IZn 0.16a 0.64 0.08a 0.14ϕZn 0.019 (0.018)a 0.007 0.025 (0.023)a 0.034

a In HEPES buffer.

Fig. 2 (a) UV-vis absorption and (b) fluorescence (λex = 315 nm) spectra of 34 µM N,N’-BQENDA (2a) in HEPES buffer (pH = 7.5, 100 mM KCl) at25 °C in the presence of various concentrations of Zn2+ between 0 and 68 µM. (c) Comparison of fluorescence spectra between 34 µM 2a (red, λex =315 nm (Abs = 0.21) and 1a (blue, λex = 315 nm (Abs = 0.21)) in HEPES buffer (pH = 7.5, 100 mM KCl) at 25 °C in the presence of 1 equiv. of Zn2+.

Fig. 3 (a) UV-vis absorption and (b) fluorescence (λex = 324 nm) spectra of 34 µM N,N-1-isoBQENDA (1b) in HEPES buffer (pH = 7.5, 100 mM KCl)at 25 °C in the presence of various concentrations of Zn2+ between 0 and 68 µM. (c) Comparison of fluorescence spectra between 34 µM 1b (red,λex = 324 nm (Abs = 0.22)) and 1-isoTQEN (blue, λex = 328 nm (Abs = 0.42)) in DMF–H2O (1 : 1) at 25 °C in the presence of 1 equiv. of Zn2+.



Publ

ishe

d on

10

Apr

il 20

14. D

ownl

oade

d by

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es o

n 25

/10/

2014

06:

00:2

9.

View Article Online


DMF–water (1 : 1) exhibit negligible differences (Fig. S3c†). Thefluorescence intensity of [Zn(1b)] is approximately half of thatof [Zn(1-isoTQEN)]2+, which is consistent with the reducednumber of chromophores. In DMF–water (1 : 1) solution, theemission intensity ratio between the bands at 466 and 352 nm(Ilong/Ishort) and quantum yields of [Zn(1b)] and [Zn-(1-isoTQEN)] are similar (Table 1). The composite results suggestthat the introduction of carboxylates into the 1-isoTQEN frame-work is an effective strategy to enhance water solubility whileretaining the superior spectroscopic properties.

To further elucidate the emission characteristics, the fluo-rescence lifetime of [Zn(1b)] at 351 and 469 nm was investi-gated (Fig. S4 and S5†). The excitation spectra for these twoemission features were identical (data not shown) but the fluo-rescence lifetime at 469 nm (τ = 6.4 ns) is significantly longerthan that at 351 nm (τ = 0.1 ns). The fluorescence lifetime at469 nm exhibits similar characteristics to the 456 nm emissionof [Zn(1a)] (τ = 6.6 ns, Fig. S6†). The lifetimes are consistentwith fluorescence from the singlet excited states.

The Zn2+-induced UV-vis and fluorescence spectral changesof N,N′-1-isoBQENDA (2b) were similar to N,N′-BQENDA (2a)(Fig. 4 and S7†). Analogous to 2a and in contrast to 1b, theemission enhancement of 2b was only observed at short wave-lengths (Fig. 4b) because the arrangement of isoquinolinegroups prevents the formation of excimer-like interactions(Fig. 4c).

Metal ion selectivity of the fluorescence response

Fig. 5 shows the metal ion selectivity of the fluorescentresponse of N,N-BQENDA (1a) and N,N-1-isoBQENDA (1b) inbuffer. The fluorescence enhancement was specific to Zn2+

although Cd2+ also induces a slight increase. For 1b, Ca2+

induces fluorescence enhancement, but only at short wave-lengths (Fig. 5b and S8†). Since many Zn2+ probes respond toCd2+, a general strategy to suppress Cd2+-induced fluorescenceis of significant interest. Replacement of two (iso)quinoneswith carboxylates reduces the Cd2+ response relative to Zn2+

(ICd/IZn) and the free ligand (ICd/I0) compared with parent tetra-kis((iso)quinoline)-based sensors (Table 1). The diminishedfluorescence response to Cd2+ is probably due to the reducedformation of the quinoline–quinoline excimer-like featuresowing to the larger Cd2+ radius that accommodates less con-gested ligand arrangement.

The metal ion selectivity of N,N′-BQENDA (2a) at 377 nm alsowas examined, even though the emission is weak (Fig. S9a†). In2a, the Zn2+ versus Cd2+ selectivity is reversed (ICd/IZn = 200%),but the short emission wavelength and weak fluorescence inten-sity limit its potential as a Cd2+ sensor. In N,N′-1-isoBQENDA(2b), Zn2+ induces the largest emission at short wavelength, but asmaller yet significant response also was observed with Cd2+ andCa2+ (Fig. S9b†). Due to the lack of selectivity and spectroscopicdeficiencies, no additional studies on 2a and 2b were warranted.

Fig. 4 (a) UV-vis absorption and (b) fluorescence (λex = 324 nm) spectra of 34 µM N,N’-1-isoBQENDA (2b) in HEPES buffer (pH = 7.5, 100 mM KCl)at 25 °C in the presence of various concentrations of Zn2+ between 0 and 68 µM. (c) Comparison of fluorescence spectra between 34 µM 2b (red,λex = 324 nm (Abs = 0.20)) and 1b (blue, λex = 324 nm (Abs = 0.23)) in HEPES buffer in the presence of 1 equiv. of Zn2+.

Fig. 5 Fluorescence spectra of 34 µM (a) N,N-BQENDA (1a) (λex = 315 nm) and (b) N,N-1-isoBQENDA (1b) (λex = 324 nm) in HEPES buffer (pH = 7.5,100 mM KCl) at 25 °C in the presence of 1 equiv. of various metal ions. (c) The relative fluorescence intensity of N,N-BQENDA (1a) at 456 nm (redbars) and N,N-1-isoBQENDA (1b) at 469 nm (blue bars) upon the addition of various metal ions. I0 is the emission intensity of free ligand.



Publ

ishe

d on

10

Apr

il 20

14. D

ownl

oade

d by

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es o

n 25

/10/

2014

06:

00:2

9.

View Article Online


pH-induced fluorescence changes and protonation constantsof 1a and 1b

Fluorescent Zn2+ probes often exhibit proton-induced fluo-rescence and/or inhibition of the Zn2+-induced signal in solu-tions of different pH. To investigate the effect of pH on theemission of N,N-BQENDA (1a) and N,N-1-isoBQENDA (1b),emission changes were monitored in aqueous HCl or NaOHsolution in the presence/absence of 1 equiv. of Zn2+. Fig. 6 dis-plays the pH-dependent fluorescence profiles for 1a and 1b aswell as the corresponding Zn2+ complexes. In the presence of1 equiv. of Zn2+, a fluorescence enhancement was observedbetween pH 4–12, indicating a wide effective working rangewith these probes. Protonation of the ligand at lower pHvalues, and the formation of Zn(OH)2 at high pH prevents Zn2+

binding to the sensors.From the pH-dependent changes in the absorption spectra

in aqueous solution (100 mM KCl at 25 °C), the protonationconstants for 1a and 1b (LH2), defined as the correspondingacid dissociation constants Kai = [LHi−1][H

+]/[LHi], were deter-mined. Sensor 1a exhibits protonation events with pKa1 =9.96(3), pKa2 = 5.63(1) and pKa3 = 3.09(1). Sensor 1b exhibits asimilar profile with pKa1 = 9.30(7), pKa2 = 6.59(1) and pKa3 =3.48(1) (Fig. S10–S15†). These values are in good agreementwith those for the corresponding pyridine analog (pKa1 =9.60(6), pKa2 = 5.46(8) and pKa3 = 3.4(1)) derived from potentio-metric titration (100 mM KNO3 at 25 °C).22

The first protonation to L2− (pKa1) should occur at the ali-phatic nitrogen atom adjacent to the carboxylates because theabsorption spectrum of HL− calculated from the titration dataresembles that of L2− with characteristic (iso)quinoline π–π*

and n–π* absorbance features (Fig. S12 and S15†). The sameassignment was reported for the pyridine analog based on vari-able-pH 1H NMR spectroscopy.22,24 The second protonation(pKa2) generates the neutral species H2L as a bis(zwitterionic)structure, but the protonation site for this species is ambigu-ous.24 While aliphatic amines are generally stronger basesthan aromatic amines, the calculated absorption spectrum ofH2L (Fig. S12 and S15†) indicates at least a partial protonationof the aromatic nitrogen atoms;22,24 therefore, we propose abridged structure where the proton is shared between aromaticand aliphatic nitrogen atoms rather than local protonation at asingle nitrogen atom. The third protonation (pKa3) corres-ponds to the protonation of H2L at an aromatic nitrogen atom.Further protonation to occupy all four nitrogen sites occursoutside of the measurement range (pKa4 < 2). The pH-depen-dent distribution profiles of 1a and 1b (Fig. 7) demonstratethat the dominant species at neutral pH are HL− and H2L,which contain deprotonated carboxylates and partially proto-nated quinolines.

Competitive Zn2+ binding of 1a, 1b and 1-isoTQEN versusTPEN and EDTA

The metal ion binding affinity of N,N-BQENDA (1a) and N,N-1-isoBQENDA (1b) is too high to enable quantification of theformation constants by direct titration. Alternatively, competi-tive titration with a high affinity chelator like TPEN or EDTAcan be used to assess the binding strength (20 °C, I = 100 mM,log K = 18.0 for TPEN;27 log K = 16.50 for EDTA28). An immedi-ate fluorescence decrease upon addition of TPEN usually indi-cates Zn2+ extraction from the sensor complex; however,

Fig. 6 Effect of pH on fluorescence intensity of 34 µM (a) N,N-BQENDA (1a) at 456 nm and (b) N,N-1-isoBQENDA (1b) at 469 nm in the absence(blue points) and presence (red points) of 1 equiv. of Zn2+ (100 mM KCl at 25 °C).

Fig. 7 Distribution profiles calculated for (a) N,N-BQENDA (1a) and (b) N,N-1-isoBQENDA (1b) as a function of pH in water (100 mM KCl) at 25 °C.



Publ

ishe

d on

10

Apr

il 20

14. D

ownl

oade

d by

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es o

n 25

/10/

2014

06:

00:2

9.

View Article Online


neither [Zn(1a)] or [Zn(1b)] complexes exhibited a loss in emis-sion intensity when exposed to 1 equiv. of TPEN for 7 days. Inaddition to the lack of metal ion extraction, apo-1a and apo-1bextract Zn2+ from [Zn(TPEN)]2+ and [Zn(EDTA)]2− complexes(Fig. 8, circles and triangles). Formation of [Zn(1a)] and [Zn(1b)] was also confirmed by ESI-MS (data not shown). As wereported previously, the [Zn(1-isoTQEN)]2+ complex did notlose Zn2+ to TPEN/EDTA, but also could not extract Zn2+ from[Zn(TPEN)]2+ and [Zn(EDTA)]2− (Fig. 8, squares).21 Based onthe extraction of Zn2+ from [Zn(TPEN)]2+, there is a consider-able difference in the metal ion binding affinity of 1b and 1-isoTQEN. In contrast, the addition of TPEN to [Zn(TQEN)]2+

immediately results in complete emission loss.20 The differ-ence in metal binding affinity between 1a and TQEN is alsosignificant. These findings may provide an efficient strategy to

prepare a set of fluorescent probes with variable metal bindingaffinity that use the same chromophore.

To examine the kinetic competition for Zn2+ between 1a/1band TPEN/EDTA, Zn2+ was added in portions to an equimolarmixture of free 1a (or 1b) and TPEN (or EDTA) in DMF–water(1 : 1) (or HEPES buffer). An extremely slow metal exchangebetween 1 and TPEN/EDTA as demonstrated above is a prere-quisite to conduct this study. The results shown in Fig. 9clearly indicate that 1a and 1b bind Zn2+ as quickly as TPENand EDTA. Such rapid binding kinetics for ethylenediaminederivatives are contrary to the slow (but strong) metal binding of1-isoTQDACH29 and 6-MeOTQTACN,30 in which aliphatic aminenitrogen atoms are conformationally restricted by the 1,2-di-aminocyclohexane (DACH) and 1,4,7-triazacyclononane (TACN)backbones, respectively. The carboxylate ligands in 1 afford

Fig. 8 Thermodynamic competitive fluorescence intensity changes of N,N-BQENDA (1a) at 456 nm (circles), N,N-1-isoBQENDA (1b) at 469 nm (tri-angles) and 1-isoTQEN at 475 nm (squares) with (a) TPEN in DMF–water (1 : 1) and (b) EDTA in HEPES buffer (pH = 7.5, 100 mM KCl) at 25 °C. Redpoints and lines, [Zn(1a)] complex + apo-TPEN/EDTA; blue, TPEN/EDTA-Zn complex + apo-1a; green, [Zn(1b)] complex + apo-TPEN/EDTA; orange,TPEN/EDTA-Zn complex + apo-1b; pink, 1-isoTQEN-Zn complex + apo-TPEN/EDTA; cyan, TPEN/EDTA-Zn complex + apo-1-isoTQEN.

Fig. 9 Kinetic competitive fluorescence intensity change of (a, c) N,N-BQENDA (1a) at 456 nm and (b, d) N,N-1-isoBQENDA (1b) at 469 nm in thepresence (circles) and absence (triangles) of one equivalent of (a, b) TPEN in DMF–H2O (1 : 1) and (c, d) EDTA in HEPES buffer (pH = 7.5, 100 mMKCl) with increasing amount of zinc at 25 °C.



Publ

ishe

d on

10

Apr

il 20

14. D

ownl

oade

d by

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es o

n 25

/10/

2014

06:

00:2

9.

View Article Online


strong metal binding affinity without losing the fast complexa-tion kinetics of ethylenediamine-based fluorescent Zn2+

probes.

Fluorescent microscopic analysis for the detection ofintracellular Zn2+ concentration change

PC-12 rat adrenal pheochromocytoma cells were utilized toevaluate Zn2+ imaging with 1a and 1b in vitro. Accounting forthe expected membrane impermeability of receptors contain-ing free acids, the corresponding ethyl esters (N,N-BQENDA-Etand N,N-1-isoBQENDA-Et) were prepared to conduct the assay.The cells were incubated in growth media containing 100 µM1a, 1b, N,N-BQENDA-Et and N,N-1-isoBQENDA-Et and 2% (v/v)DMSO for 4 h and analyzed with a fluorescence microscopeafter cell washing and replacing the media. Zn–pyrithione(90 µM) was added during microscopic analysis and thefluorescence change was monitored.

As shown in Fig. 10, a bright Zn2+-induced fluorescenceappeared in the presence of N,N-1-isoBQENDA-Et. A small fluo-rescence response with N,N-BQENDA-Et was also detected.When 1a and 1b were evaluated, only minimal fluorescenceresponses were measured, which is consistent with the pre-dicted membrane impermeability of the free acid ligands. Incontrast, intracellular esterases release 1a and 1b after themore lipophilic esters cross the cell membrane and enter thecytosol. No cytotoxicity of these compounds was observed withup to 100 µM concentrations of sensors added to the growthmedia. The cell-permeability and intracellular Zn2+-response ofthe esters, especially N,N-1-isoBQENDA-Et, make the BQENDAfamily of sensors attractive candidates for imaging studies.

Conclusions

In this study, we examined the introduction of two carboxy-lates in place of two (iso)quinolines into tetrakis((iso)quino-line) Zn2+ probes with an ethylenediamine backbone. Theresulting hybrid (iso)TQEN-EDTA chelators show excellentwater-solubility and improved photophysical and physico-chemical properties relevant to fluorescent Zn2+ sensors. TheN,N-isomers 1a/1b exhibited higher fluorescence quantumyields when complexed with Zn2+, improved Zn2+ selectivityover Cd2+ and strong metal ion binding affinity. The N,N′-isomers 2a/2b are less amenable to Zn2+ sensing because theN,N-bis((iso)quinolylmethyl)amine moiety that promotes anexcimer-like (iso)quinoline–(iso)quinoline interaction uponZn2+ binding is absent. To the best of our knowledge, 1a and1b are the first fluorescence probes capable of extracting Zn2+

from [Zn(TPEN)]2+ and [Zn(EDTA)]2−. The superior photo-physical and physicochemical properties of 1a/1b, includinghigh water solubility and high metal ion binding affinity,provide an important roadmap for designing quinoline-basedfluorescent probes.

Acknowledgements

The authors thank Prof. Ishida of Kitasato University forhelpful discussion about fluorescence lifetime measurements.This work was supported by the Research for Promoting Tech-nological Seeds, JST, Adaptable and Seamless TechnologyTransfer Program through Target-driven R&D, JST, Grant-inAid for Scientific Research from the MEXT, Japan and the NaraWomen’s University Intramural Grant for Project Research.

Notes and references

1 B. L. Vallee and K. H. Falchuk, Physiol. Rev., 1993, 73,79–118.

2 B. L. Vallee and D. S. Auld, Acc. Chem. Res., 1993, 26, 543–551.

3 C. J. Frederickson, S. W. Suh, D. Silva, C. J. Fredericksonand R. B. Thompson, J. Nutr., 2000, 130, 1471S–1483S.

4 E. L. Que, D. W. Domaille and C. J. Chang, Chem. Rev.,2008, 108, 1517–1549.

5 Z. Dai and J. W. Canary, New J. Chem., 2007, 31, 1708–1718.6 Z. Liu, W. He and Z. Guo, Chem. Soc. Rev., 2013, 42,

1568–1600.7 L. M. Hyman and K. J. Franz, Coord. Chem. Rev., 2012, 256,

2333–2356.8 Y. Jeong and J. Yoon, Inorg. Chim. Acta, 2012, 381, 2–14.9 J. F. Callan, A. P. de Silva and D. C. Magri, Tetrahedron,

2005, 61, 8551–8588.10 D. Buccella, J. A. Horowitz and S. J. Lippard, J. Am. Chem.

Soc., 2011, 133, 4101–4114.11 K. Hanaoka, Y. Muramatsu, Y. Urano, T. Terai and

T. Nagano, Chem. – Eur. J., 2010, 16, 568–572.

Fig. 10 Differential interference contrast (DIC) and fluorescence (FL)micrographs of cultured PC-12 rat adrenal cells incubated in the pres-ence of ligands (100 µM) for 4 h. Zn–pyrithione (90 µM) was addedduring microscopic analysis (photograph taken after ∼4 min).



Publ

ishe

d on

10

Apr

il 20

14. D

ownl

oade

d by

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es o

n 25

/10/

2014

06:

00:2

9.

View Article Online


12 Z. Xu, K.-H. Baek, H. N. Kim, J. Cui, X. Qian, D. R. Spring,I. Shin and J. Yoon, J. Am. Chem. Soc., 2010, 132, 601–610.

13 L. Xue, H.-H. Wang, X.-J. Wang and H. Jiang, Inorg. Chem.,2008, 47, 4310–4318.

14 H.-H. Wang, Q. Gan, X.-J. Wang, L. Xue, S.-H. Liu andH. Jiang, Org. Lett., 2007, 9, 4995–4998.

15 Y. Zhang, X. Guo, W. Si, L. Jia and X. Qian, Org. Lett., 2008,10, 473–476.

16 C. Ichimura, Y. Shiraishi and T. Hirai, Tetrahedron, 2010,66, 5594–5601.

17 M. Royzen, A. Durandin, V. G. Young, Jr., N. E. Geacintovand J. W. Canary, J. Am. Chem. Soc., 2006, 128, 3854–3855.

18 N. J. Williams, W. Gan, J. H. Reibenspies andR. D. Hancock, Inorg. Chem., 2009, 48, 1407–1415.

19 H. G. Lee, J. H. Lee, S. P. Jang, I. H. Hwang, S.-J. Kim,Y. Kim, C. Kim and R. G. Harrison, Inorg. Chim. Acta, 2013,394, 542–551.

20 Y. Mikata, M. Wakamatsu and S. Yano, Dalton Trans., 2005,545–550.

21 Y. Mikata, A. Yamanaka, A. Yamashita and S. Yano, Inorg.Chem., 2008, 47, 7295–7301.

22 M. Heitzmann, F. Bravard, C. Gateau, N. Boubals,C. Berthon, J. Pécaut, M.-C. Charbonnel and P. Delangle,Inorg. Chem., 2009, 48, 246–256.

23 M. Heitzmann, C. Gateau, L. Chareyre, M. Miguirditchian,M.-C. Charbonnel and P. Delangle, New J. Chem., 2010, 34,108.

24 P. Caravan, S. J. Rettig and C. Orvig, Inorg. Chem., 1997, 36,1306–1315.

25 P. Caravan, P. Mehrkhodavandi and C. Orvig, Inorg. Chem.,1997, 36, 1316–1321.

26 H. Tang, N. Arulsamy, M. Radosz, Y. Shen, N. V. Tsarevsky,W. A. Braunecker, W. Tang and K. Matyjaszewski, J. Am.Chem. Soc., 2006, 128, 16277–16285.

27 R. M. Smith and A. E. Martell, Critical stability constants,Plenum Press, New York, 1975.

28 A. E. Martell and R. M. Smith, Critical stability constants,Plenum Press, New York, 1974.

29 Y. Mikata, Y. Sato, S. Takeuchi, Y. Kuroda, H. Konno andS. Iwatsuki, Dalton Trans., 2013, 42, 9688–9698.

30 Y. Mikata, Y. Nodomi, A. Kizu and H. Konno, Dalton Trans.,2014, 43, 1684–1690.



Publ

ishe

d on

10

Apr

il 20

14. D

ownl

oade

d by

Was

hing

ton

Stat

e U

nive

rsity

Lib

rari

es o

n 25

/10/

2014

06:

00:2

9.

View Article Online


bis(2-quinolylmethyl)ethylenediaminediacetic acids (bqendas), tqen–edta hybrid molecules as...

Documents