Ratiometric fluorescence sensing and cellular imaging of Cu2+ by a new water soluble trehalose-naphthalimide based chemosensor

Article abstract of DOI:10.1039/c3ra43988g

A new turn-on Cu²⁺ fluorescent sensor (CST) having a trehalose moiety, which confers a relatively large solubility in water, has been synthesized. The chemosensor is therefore suitable for studies in aqueous solution. Full potentiometric and UV-vis characterization evidence that at physiological pH CST forms with Cu²⁺ a species with a 1 : 1 stoichiometry allowing for a straightforward correlation between CST response and copper(ii) concentration. The presence of the trehalose unit does not negatively affect the selectivity of CST for Cu²⁺ over a series of metal ions of interest as proven by fluorescence measurements. The novel chemosensor, tested in differentiated neuroblastoma SH-SY5Y cells, is able to detect Cu²⁺ in the extracellular region, as well as to track copper transfer processes upon cell stimulation induced by cellular depolarization.

Full text of DOI:10.1039/c3ra43988g

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Ratiometric fluorescence sensing and cellular imaging
2
+
of Cu by a new water soluble trehalose-naphthalimide
Cite this: RSC Adv., 2013, 3, 24288
based chemosensor†
a
b
a
b
Giuseppa Ida Grasso, Salvatore Gentile, Maria Laura Giuffrida, Cristina Satriano,
b
c
b
b
Carmelo Sgarlata, Massimo Sgarzi, Gaetano Tomaselli, Giuseppe Arena*
and Luca Prodi*c
2+
A new turn-on Cu fluorescent sensor (CST) having a trehalose moiety, which confers a relatively large
solubility in water, has been synthesized. The chemosensor is therefore suitable for studies in aqueous
solution. Full potentiometric and UV-vis characterization evidence that at physiological pH CST forms
2+
with Cu a species with a 1 : 1 stoichiometry allowing for a straightforward correlation between CST
response and copper(II) concentration. The presence of the trehalose unit does not negatively affect the
Received 29th July 2013
2+
selectivity of CST for Cu
over a series of metal ions of interest as proven by fluorescence
Accepted 9th October 2013
measurements. The novel chemosensor, tested in differentiated neuroblastoma SH-SY5Y cells, is able to
DOI: 10.1039/c3ra43988g
2+
detect Cu in the extracellular region, as well as to track copper transfer processes upon cell stimulation
www.rsc.org/advances
induced by cellular depolarization.
(
molecules formed by a binding and a signalling unit) based on
Introduction
ion-induced changes in the uorescence response have been
15
Metal ions play crucial roles in living systems and have an proposed as useful detecting agents and are particularly
extremely remarkable eco- and toxicological impact on both the attractive due to their high sensitivity, selectivity and quick
environment and human beings. As a consequence, the detection response. Most of the reported copper(II) uorescent chemo-
and quantitative determination of metal ions has long received sensors are based on the quenching effect caused by the para-
16
1,2
2+
considerable attention. Among transition metals, copper(II)
carries out key functions as, inter alia, a catalytic cofactor for a
variety of metalloenzymes. Moreover, alteration of the homeo-
magnetic nature of Cu and, thus, belong to the turn-off
class.
rescence enhancement) as a result of the metal-ion interaction
17–20
Fluoroionophores showing a turn-on response (uo-
3
static levels of copper(II) has been observed both in biological are obviously more attractive as proved by their rapidly
2
1–23

uids and tissues of patients affected by breast, prostate, lung increasing number.
However, only few chemosensors from
4–7
and gastrointestinal cancer as well as in some neurodegener- this class have good solubility in water, which is a key issue for
ative disorders, such as Parkinson's (PD) and Alzheimer's Disease applications in biological systems.
8,9
2+
(AD). Copper(II) has been involved in b-amyloid peptide (Ab)
Our goal has been the design and synthesis of a Cu che-
aggregation and neurotoxicity; in AD brain tissue it was observed
an excess of copper in the extracellular space and in the amyloid
mosensor that may work in the turn-on mode and may also have
a satisfactory solubility in water. The uorescent sensor repor-
10,11
24,25
plaques
compared to the control healthy brain tissue.
The need for molecular systems able to effectively recognize over, the presence in the molecular scaffold of amino groups as
as well as a decrease of intracellular copper if
ted by Xu et al.
(1, Fig. 1) showed an enhancement of the
11
uorescence signal resulting from metal complexation. More-
12–14
26
copper ions in solution
and that could also work in living the metal binding sites renders the compound suitable for
2
+
2+
organisms has increased in the last years. Chemosensors both endogenous (Cu and Zn ) and toxic metal ions as it can
form stable metal chelates in solution. Unfortunately, ligand 1
a
b
is poorly soluble in water. In order to increase its water solu-
Dipartimento di Scienze Chimiche, Universit a` di Catania, viale A. Doria 6, Catania, bility, 1 was functionalized with a sugar moiety (trehalose)
Istituto di Biostrutture e Bioimmagini, CNR, viale A. Doria 6, Catania, Italy
27
Italy. E-mail: garena@unict.it; Fax: +39 095 337678; Tel: +39 095 738 5071
(CST, Fig. 1).
c
Dipartimento di Chimica “G. Ciamician”, Universit a` di Bologna, via Selmi 2, Bologna,
The ability of a chemosensor to bind and detect a metal ion
depends on the characteristics of the solvent, the temperature
and pH. Amines can be easily protonated in aqueous solution
and protons may compete with the target metals for the
complexation of the ligand. Consequently, an accurate study of
Italy. E-mail: luca.prodi@unibo.it; Fax: +39 051 20 9 9456; Tel: +39 051 20 9 9481
†
Electronic supplementary information (ESI) available: NMR spectra, species
distribution diagram of the proton complexes of CST, UV-vis of CST, inuence
of the pH on the uorescence emission of CST and uorescence response of
CST to some selected anions. See DOI: 10.1039/c3ra43988g
2
4288 | RSC Adv., 2013, 3, 24288–24297
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
RSC Advances
corresponding oxides with the appropriate amount of nitric
acid and diluting them with pure water. Lead(II) and cobalt(II)
nitrate as well as nickel(II) perchlorate solutions were
prepared by dissolving the salts in water and by titrating the
resulting solution with EDTA using methylthymol blue,
2
9
xylenol orange and murexide, respectively; calcium(II) and
magnesium(II) solutions were titrated with EDTA using mur-
exide and eriochrome black T, respectively. Silver(I) nitrate
3
0
solutions were standardized according to Mohr's procedure.
Iron(II) solutions were standardized using a stock KMnO4
solution. Solutions of HNO and KOH were standardized by
3
titration with tris(hydroxymethyl)aminomethane (THAM) and
potassium hydrogen phthalate, respectively. High purity
water (Millipore, Milli-Q Element A 10 ultrapure water) was
used throughout.
0
Synthesis of 6-bromo-6-deoxy-a,a -trehalose (TH–Br)
Triphenylphosphine (PPh3) (16.6 g) was added to a solution of
Fig. 1 Sketches of the fluorescent sensor reported by Xu et al. (1) and its water- anhydrous trehalose (10.8 g) in DMF in a 2 : 1 molar ratio. The
24
soluble trehalose-functionalized derivative (CST).
mixture was cooled and N-bromosuccinimide (11.4 g, NBS/
PPh3 in 1 : 1 molar ratio) was slowly added under stirring. The
reaction mixture was kept at room temperature for 48 h. 100 ml
of methanol were added to decompose the excess reagent.
the acid-base features of the ligand is mandatory to gure out
the best pH window that allows for a useful application of the
sensor. In this paper, we report on the design and synthesis of a
water-soluble trehalose-derivatized amine-based chemo-
Methanol and DMF were later evaporated under vacuum at
ꢁ
3
5 C. The oily product was washed with ethanol and the white
crystals which formed were ltered off using a Millipore
2
+
sensors, its proton and Cu complexation properties and its
system. The ltered solution was evaporated under vacuum at
2
+
use as a Cu probe in living systems.
ꢁ
3
5 C and the solid was puried by chromatography on a RP8
Proof-of-working experiments of live cell imaging were per-
formed on differentiated neuroblastoma SH-SY5Y cells to
demonstrate the capability of CST sensor to track the presence of
copper(II) ions in a cellular environment. Moreover, cellular
depolarization experiments were performed to follow Cu
release upon potassium chloride treatment.
2
column, using a linear gradient H O/EtOH (0–25%). Yield:
7
5
0%. TLC: Rf
: 3 : 1 : 1).
¼
2 3
0.54 for TH–Br (PrOH/H O/AcEt/NH
2
+
0
Synthesis of 6-azido-6-deoxy-a,a -trehalose (TH–N )
3
TH–Br (2.4 g) was dissolved in 30 ml of water and sodium azide
(
3.86 g) was added to the solution in a 1 : 10 molar ratio. The
Experimental section
Chemicals
ꢁ
reaction mixture was stirred at 100 C for 4 h. The solvent was
removed under vacuum and the remaining solid was puried by
8
Commercially available reagents were used as received unless chromatography on a RP column, using a linear gradient of
otherwise noted. Trehalose (TH) (Fluka) was dried under vacuum
H
2
O/EtOH (0–30%). Yield: 90%. TLC: Rf ¼ 0.52 for TH–N
3
ꢀ
2
ꢁ
(
10 mm Hg) for 24 h at 80 C, using a P
O
2 5
trap. N-bromo- (PrOH/H
2
O/AcEt/NH
3
5 : 3 : 1 : 1).
succinimide was purchased from Fluka; triphenylphosphine,
sodium azide, 2-(aminomethyl)pyridine, 2-metoxyethanol and
DMF were purchased from Sigma-Aldrich. Potassium nitrate was
0
Synthesis of 6-amino-6-deoxy-a,a -trehalose (TH–NH )
2
purchased from Merck. Thin-layer chromatography (TLC) was TH–N (1.05 g) was dissolved in DMF and PPh (5.25 g) was
3
3
performed on silica gel (60F-254, 0.20 mm, Macherey-Nagel); a added to the solution in a 1 : 7 molar ratio. The reaction
ꢁ
5
% solution of p-anisaldehyde in ethanol (containing 5% of was carried out at 40 C under stirring for 1 h. The product
H
2
SO
4
) was used to reveal the compounds which were not was treated with a 30% ammonium hydroxide solution at
room temperature for about 24 h. White crystals formed and
detectable under a UV lamp.
Copper(II) nitrate solutions were prepared by dissolving the were ltered off using a Millipore system. The ltered solution
salt in water and adding a slight excess of nitric acid to was evaporated under vacuum and the product, 6-amino-6-
0
prevent hydrolysis; the concentration of the resulting solution deoxy-a,a -trehalose (TH–NH
), was puried by column
2
+
was determined by EDTA titrations using murexide as the chromatography (CM-Sephadex C-25, 20 ꢂ 600 mm, NH
4
indicator. The HNO excess in the metal stock solution was form), using a linear gradient of H O/NH HCO (0–0.3 M).
3
2
4
3
2
8
determined by Gran's method. Zinc(II), cadmium(II) and Yield: 61%. TLC: Rf ¼ 0.19 for TH–NH
2
2 3
(PrOH/H O/AcEt/NH
mercury(II) nitrate solutions were obtained by dissolving the 5 : 3 : 1 : 1).
This journal is ª The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 24288–24297 | 24289

View Article Online
RSC Advances
Paper
Synthesis of N-trehalose-4-bromo-5-nitro-1,8-naphthalimide
Thermogravimetric analysis
Thermogravimetric analysis of the CST ligand was carried out
(
TH-naphthalimide)
N-bromosuccinimide (18 g) was dissolved in 50 ml of DMF and using a Perkin-Elmer TGS-2 instrument. Typically, ligand
the resulting solution was added to a suspension of acenaph- samples (ca. 2 mg) were heated under a controlled atmosphere
ꢁ
ꢁ
ꢀ1
thene (15.4 g, 1 : 1 molar ratio) at room temperature. Since the (air) from 50 to 800 C (heating rate 10 C min ).
reaction was exothermic, the mixture was poured into water
aer ca. 2 h. The solution was ltered and 4-bromoacenaph-
thene was obtained with a yield of 79%. 4 g of 4-bromoace-
naphthene were dissolved in glacial acetic acid (33 ml) under
NMR spectroscopy
1
ꢁ
2
H NMR spectra were recorded at 25 C in D O with a Varian Unity
1
Plus 500 spectrometer at 499.883 MHz. The H NMR spectra were
recorded by using the standard pulse programs from the Varian
library. In all cases, the length of 90 pulse was ca. 7 ms. The two-
dimensional (2D) experiments were acquired using 1 K data
points, 256 increments, and a relaxation delay of 1.2 s.
ꢁ
stirring at 15 C; aer 30 min, a mixture of concentrated nitric
acid (2.5 ml) and glacial acetic acid (7.5 ml) was added at a
ꢁ
ꢁ
temperature lower than 15 C. The suspension was stirred at
ꢁ
1
0–15 C for 10 h, ltered and the yellow residue (3.0 g) was
crystallized from glacial acetic acid to obtain 4-bromo-5-
nitroacenaphthene (yellow needles). 4-bromo-5-nitro-
acenaphthene (2.3 g) was added to a solution of sodium
Potentiometric measurements
dichromate dihydrate (6 g) in 25 ml of glacial acetic acid under Potentiometric titrations were performed by means of two
stirring and the resulting solution was slowly heated up under home-assembled, fully automated apparatus (Metrohm E654
reux. The solution turned from bright orange to dark green pH-meter, Metrohm combined pH glass electrode 6.0234.100,
2
ꢀ
3+
owing to the reduction of Cr
2
O
7
to Cr . Aer reuxing for 3 Hamilton Microlab 500 series dispenser) controlled by the
h, the dark green liquor was diluted with cold water (10 ml), appropriate soware set up in our laboratory. The titration cell
ꢁ
cooled down and the yellow-orange residue washed with a (4 ml) was thermostated at 25.0 ꢃ 0.2 C and all solutions were
small amount of glacial acetic acid and then stirred into a 4% kept under an atmosphere of argon. The system was calibrated
+
aqueous solution of sodium hydroxide. Aer ltering, the on the pH ¼ ꢀlog [H ] scale by titrating HNO with CO -free
3
2
ꢁ
solution was neutralized with a 5% solution of hydrochloric base. The values of E , E , K and the Nernstian slope of the
j
w
acid to give an orange precipitate of 4-bromo-5-nitro-1,8- electrodic system were determined in separate experiments by
naphthalic anhydride (yield: 42%). 4-bromo-5-nitro-1,8-naph- titrating nitric acid with CO -free potassium hydroxide. Base
2
thalic anhydride (0.120 g) was added to a solution of TH-NH
2
solutions (ca. 0.1 M) were added through a Hamilton burette
(
0.130 g) in DMF (1 : 1 molar ratio). The reaction was carried equipped with a 1 ml syringe. In order to determine the
ꢁ
out at 60 C under stirring for 5 h. The solvent was removed protonation constants of CST, solutions of the ligand with
ꢀ4
under vacuum and the remaining solid was puried by chro- concentrations ranging from 8.5 to 9.7 ꢂ 10 M were titrated
matography on a RP column, using linear gradients of H O/ with 0.1 M potassium hydroxide. Three independent runs were
8
2
EtOH (0–30% and 30–60%). Yield: 69%. TLC: Rf ¼ 0.69 for collected, each run including 50–60 points. The initial pH was
N-trehalose-4-bromo-5-nitro-1,8-naphthalimide (PrOH/H O/ always adjusted to 2.4. To avoid systematic errors and verify the
2
1
AcEt/NH
3
5 : 3 : 1 : 1). H NMR (500 MHz, CDCl
3
) d (ppm) 8.59 reproducibility, the emf values of each experiment were taken
(
8
1H, d, J ¼ 8 Hz), 8.37 (1H, d, J ¼ 8 Hz), 8.22 (1H, d, J ¼ 8 Hz), at different time intervals. The ionic strength (I) was main-
.05 (1H, d, J ¼ 8 Hz), 4.55 (1H, d, J ¼ 4 Hz), 4.49 (1H, d, J ¼ 4 tained at 0.1 M (KNO ). Other experimental details have been
3
31
Hz), 4.39 (2H, d, J ¼ 6 Hz), 4.08 (1H, m), 3.71 (1H, t, J ¼ 9 Hz), reported previously.
3
3
.50-3.67 (5H, m), 3.41 (1H, t, J ¼ 9 Hz), 3.15 (1H, t, J ¼ 10 Hz),
.04 (1H, dd, J ¼ 3 Hz), 1.90 (1H, d, J ¼ 2 Hz).
Spectrophotometric measurements
UV-vis titrations were carried out with an Agilent 8453 diode-
array spectrophotometer interfaced with a home-assembled
potentiometric apparatus (Metrohm E654 pH-meter, Orion
Synthesis of N-trehalose-4,5-di[(2-aminomethyl)pyridine]-1,8-
naphthalimide (CST)
N-trehalose-4-bromo-5-nitro-1,8-naphthalimide (0.170 g) was 9103SC combined micro pH glass electrode, Hamilton Microlab
dissolved in 15 ml of 2-methoxyethanol. 2-(aminomethyl)pyri- 500 series dispenser) all controlled by a soware set up in our
dine (570 ml) was added to the solution in a 1 : 21 molar ratio. laboratory. The quartz cell (1 cm path length) was thermostated
ꢁ
ꢁ
The reaction mixture was stirred at 130 C for 6 h. The solvent at 25.0 ꢃ 0.2 C. In order to double check the proton complexes
ꢁ
was removed under vacuum at 35 C and the remaining solid constant values and to determine the stability constants for the
2
+
ꢀ5
was puried by a RP
gradients of H
O/EtOH (0–30% and 30–70%). Yield: 20%. TLC: M) or the ligand (from 2.5 to 3.0 ꢂ 10 M) and the metal (M/L
Rf ¼ 0.26 for CST (PrOH/AcEt/H
500 MHz, CD
OD) d (ppm) 8.35 (2H, d, J ¼ 4.5 Hz), 8.19 (2H, d, hydroxide solutions (0.052 M). Base solutions were added
8
column chromatography using linear Cu complexes, solutions of the ligand (from 2.0 to 5.0 ꢂ 10
ꢀ5
2
1
2
3
O/NH 5 : 3 : 1 : 1). H NMR ratios ranged from 0.5 to 0.8) were titrated with potassium
(
3
J ¼ 8.5 Hz), 7.72 (2H, t, J ¼ 7.5 Hz), 7.49 (2H, d, J ¼ 8 Hz), 7.27 through a Hamilton burette equipped with 1 ml syringe. The
(
(
2H, t, J ¼ 7 Hz), 6.71 (2H, d, J ¼ 9 Hz), 4.86 (1H, d, J ¼ 4 Hz), 4.64 ionic strength of all solutions was adjusted to 0.1 M (KNO ). Five
3
5H, s), 4.33–4.39 (2H, m), 4.29(2H, dd), 4.19–4.13 (1H, m, independent titrations were carried out, each including 70–80
broad), 3.54–3.68 (5H, m), 3.45 (1H, dd), 3.35 (1H, dd).
points. The initial pH was adjusted to 2.8 for all solutions.
2
4290 | RSC Adv., 2013, 3, 24288–24297
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
RSC Advances
Calculations
Argon laser (457 nm, 488 nm, 515 nm, total 30 mW), HeNe(G)
laser (543 nm, 1 mW), HeNe(R) laser (633 nm, 1 mW); oil
immersion objective (60xO PLAPO) and spectral ltering
system. Emitted light was detected in sequential mode. The
detector gain was xed at a constant value and images were
taken for all of the samples at random locations throughout the
area of the well.
Fluorescent probes addition was performed in situ, on the
sample mounted on the microscope stage. Cells were rinsed
with 10 mM MOPS buffer, then CST (0.5–2 mM) and Mitotracker
ꢁ
The slope of the electrodic system, E and K values were
determined by the computer program ACBA. Potentiometric
w
32
and spectrophotometric data were handled with the soware
33
Hyperquad which carries out a non-linear minimization of the
sum of the squared residuals U and also allows for the simul-
taneous renement of data from different titrations
P
2
U ¼ (Eexp ꢀ Ecalc
)
where Eexp and Ecalc are the experimental and the calculated Deep Red (50 nM in PBS, Invitrogen) were used to track cop-
electrode potentials, respectively. The species distribution was per(II) and mitochondria organelles, respectively. For cell
34
calculated by using the program Hyss.
depolarization experiments, the cells previously labeled with
Mitotracker and CST were rinsed with articial cerebro-spinal
ꢁ

uid (aCSF) buffer (20 mM, pH 7.4 at 25 C) and 80 mM KCl was
Fluorescence measurements
added.
Image analysis, including uorescence intensity proles
were carried out using the public domain, Java-based image
Fluorescence spectra were recorded from 460 to 700 nm in a
ꢁ
quartz cell (1 cm path length) at 25 ꢃ 1 C by using a Varian Cary
Eclipse Fluorescence Spectrophotometer. An excitation wave-
length of 455 nm was used in all experiments. Excitation and
emission bandwidths were set to 10 and 5 nm, respectively.
35
processing program Image J (version 1.46e).
The statistical analysis was performed with a one-way Anova
test by using Microcal Origin (version 8.6). Specically, aer
background subtraction, the mean of each eld of cells was
averaged over the n elds and the associated standard error was
calculated.
Cu(NO
Ni(ClO
3
)
)
2
, AgNO
, Pb(NO
3
, Cd(NO
3
)
2
, Co(NO
3
)
2
, Fe(NO , Hg(NO
3
)
2
3 2
) ,
4
2
3
)
2
, Zn(NO
3
)
2
, Ca(NO
3
)
2
, MgSO and NaCl
4
ꢀ
5
aqueous solutions with concentrations ranging from 1 ꢂ 10
ꢀ
5
M to 5 ꢂ 10 M were added to a buffered CST solution (8 ꢂ
ꢀ
7
ꢀ6
10
O 1.3 ꢂ 10 M) in order to explore a metal/ligand ratio up
to 10. Competition experiments were run for both metal ions
and some selected anions (phosphate, sulphate and chloride)
by adding up to ve and 1000 equivalents of the metal ion or the
Results and discussion
Synthesis of copper sensor trehalose (CST)
ꢀ
6
anion, respectively, to a 1 ꢂ 10 M CST solution containing
The synthesis of CST implied the steps illustrated in Fig. 2. The
bromination of trehalose (TH) was carried out by treating the
alcohol with N-bromosuccinimide and triphenylphosphine;
such procedure has been reported as an efficient route for the
selective replacement of the primary hydroxyl groups in carbo-
2
+
one equivalent of Cu . Solutions were maintained at pH 7.2
using 3 (N-morpholino)propanesulfonic acid (MOPS).
Cell cultures
36
Human neuroblastoma SH-SY5Y cells were propagated in hydrates. The ratio between the components of the haloge-
DMEM-F-12 (1 : 1) medium, supplemented with 10% fetal nation mixture (NBS and PPh ), along with the reaction time
3
bovine serum (FBS), 1% penicillin/streptomycin, 2 mM L-gluta- and temperature, resulted crucial for driving the mono- or
ꢁ
mine and maintained in a humidied incubator at 37 C in 5% dibromination at the primary positions of the disaccharide. The
CO
2
. Cells were plated into glass-bottom dishes (Willco Wells B. reaction likely proceeds by the initial formation of an alkoxy-
5
V., The Netherlands) at a density of 1 ꢂ 10 cells/dish and phosphonium ion which later reacts with bromide ions to give
medium were replaced every three days with freshly prepared the desired product.
DMEM-F-12 containing decreasing concentration of heat-inac-
The bromo-trehalose derivative (TH–Br) was aminated
tivated FBS. To induce differentiation of SH-SY5Y into neuronal- through a two steps reaction: the synthesis of the azido derivative
37
like cells, 5 mM of all-trans retinoic acid (RA) were added every (TH–N
3
) and the reduction of the azide to amino-group. TH–N
3
two days to medium containing 1% of FBS, for up to 7 days.
was rst obtained by reaction of TH–Br with NaN in aqueous
3
solution; the reduction of TH–N3 with triphenylphosphine in
DMF, followed by hydrolysis of the iminophosphorane interme-
MTT assays
4 2
diate with NH OH, yielded the amino derivative (TH–NH ).
Neuronal survival to CST was assessed on differentiated SH-
SY5Y cells by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazo-
lium bromide (MTT) reduction assay; the above cells had been
previously treated for 24 hours with 0.5–2 mM CST.
Acenaphthene was brominated with NBS in DMF and the
resulting 4-bromoacenaphthene was nitrated in acetic acid.
Oxidation of 4-bromo-5-nitro-acenaphthene led to 4-bromo-5-
38
nitro-1,8-naphthalic anhydride. The latter compound was
condensed with TH–NH to obtain N-trehalose-4-bromo-5-nitro-
,8-naphthalimide which was subsequently converted to the
2
Live-cell imaging experiments
1
39
Live cell imaging experiments were carried out by using an nal product (CST) by reaction with 2-aminomethylpyridine.
Olympus FV1000 confocal laser scanning microscope (LSM) Fig. S1–S5† show the NMR spectra of the precursor N-trehalose-
equipped UV/visible lasers: 405 nm (50 mW), 20 mW Multiline 4-bromo-5-nitronaphthalimide and the ligand CST.
This journal is ª The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 24288–24297 | 24291

View Article Online
RSC Advances
Paper
Table 1 Log K values for the protonation of CST at 25 ꢁC and I ¼ 0.1 M (KNO
)
a
3
Equilibrium
log K
+
+
CST + H % CSTH
4.35 (6)
3.67 (6)
+
+
2+
CSTH + H % CSTH
2
a
3s in parentheses.
solvent due to the low solubility of the ligand in pure water. The
replacement of the butyl chain of Xu's chemosensor with a
carbohydrate moiety increased the solubility in water of the
ꢀ3
resulting compound (CST) up to values (about 1 ꢂ 10 M) that
are appropriate for a thorough characterization in aqueous solu-
tions. In particular, the relatively large concentrations that may be
reached by dissolving the sugar-derivatized compound in water
allowed for the determination of the formation constants for both
the proton and the metal complexes of CST. This study intended
to ascertain whether or not the trehalose derivatized ligand still
2
+
maintained its complexing properties towards Cu .
The protonation equilibria of CST in aqueous solution are
given in eqn (1) and (2)
+
+
L + H % LH
(1)
(2)
+
+
2+
LH + H % LH
2
and the proton complexes constant values obtained by the
analysis of the potentiometric data are reported in Table 1.
The stability constant values indicate that the above equi-
libria likely deal with the protonation of the two pyridine
nitrogen atoms of the CST scaffold. The distribution diagram
reported in Fig. S6† shows that the ligand exists as a neutral
species at neutral pH; above this pH value this species is the
2
+
only existing in solution. The investigation of Cu complexes
formation was initially carried out through potentiometry
under the same experimental conditions (i.e., concentrations of
ꢀ
3
the order of 10
M) used for the determination of the
Fig. 2 Scheme of the synthesis of CST ligand.
protonation constants. Unfortunately, the formation of a bluish
2
+
Proton and Cu complexes
Since the accurate determination of the stability constants
requires concentrations to be precisely known, the effective
molecular weight of the bare CST was determined via ther-
mogravimetric analysis (TGA). The rst drop observed in the
ꢁ
TGA curve (up to 120 C) results from the adsorbed residual
solvent that amounts to ca. 10% of the total ligand weight. The
ꢁ
solid was fully decomposed at about 600 C, and the thermo-
gram remained practically at up to the maximum temperature
ꢁ
reached in our experiments (800 C). The above observations
indicate that the solid contains unbound water only, does not
contain inorganic compounds and is therefore suitable for
potentiometric analysis.
24
The chemosensor originally designed by Xu et al. is unfor-
tunately poorly soluble in water and, consequently, a full charac-
terization and investigation of its acid-base properties in aqueous
solution, where it is intended to be employed, was hampered.
ꢀ5
Fig. 3 Typical UV-vis titration for the copper(II)-CST (2.7 ꢂ 10 M) system in
ꢁ
Fluorescence studies had been reported in ethanol–water mixed water at 25 C and I ¼ 0.1 M (KNO
3
).
2
4292 | RSC Adv., 2013, 3, 24288–24297
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
RSC Advances
ꢁ
Table 2 Stability constant values for the copper(II) complexes of CSTat 25 C and
a
I ¼ 0.1 M (KNO
3
)
Equilibrium
log b
2
+
2+
Cu + CST % [CuCST]
7.49 (6)
ꢀ1.35 (9)
2+
+
+
Cu + CST % [CuCSTHꢀ1] + H
a
3s in parentheses.
precipitate even at relatively acidic conditions (pH < 4)
hampered a potentiometric investigation and, thus, we were
forced to resort to UV-vis spectrophotometric titrations that
require lower concentrations of both the ligand and the metal
ꢀ5
solutions. When 10 M metal and ligand solutions were used,
no precipitate was ever detected over the entire pH range
monitored. The protonation of the free ligand was also re-
investigated by UV-vis measurements and this allowed for a
cross-check of the protonation constant values determined by
potentiometry.
Fig. 5 Fluorescence emission spectra of a 1 ꢂ 10ꢀ M CST solution upon addi-
tion of increasing amounts of a copper(II) solution. pH was maintained at 7.2
using MOPS. Excitation wavelength was 455 nm. Inset: ratiometric plot
6
(
normalized I477/I535) as a function of the metal/ligand ratio.
Fig. S7† shows a typical spectrophotometric titration of a
CST aqueous solution with KOH from pH 3 to pH 10.
The absorption spectrum of free CST at pH 3 exhibits a band
centred at 447 nm and a shoulder at 427 nm which, upon
addition of increasing amounts of base, undergo a red shi to
Several models, and combinations thereof, having different
values for m, l, and h, were offered to the computer program.
However, the data were satisfactorily tted only when using the
model below:
456 and 434 nm, respectively, along with an hyperchromic
effect. These evidences indicate that a deprotonation event is
occurring in the ligand solution. The multi-wavelength and
multi-pH analysis of a series of spectrophotometric titration
data yielded protonation constant values that nicely reproduce
those obtained by potentiometry.
Fig. 3 shows a typical spectrophotometric titration for the
copper(II)-CST system carried out from pH 2.8 to pH 10.6. The
copper(II)-CST solution at pH 2.8 shows a shoulder at 427 nm
and a band centred at 447 nm which again undergo a red shi
upon addition of a KOH solution.
The UV-vis titration data were rened to obtain the best set of
complex formation constant values, bmlh, dened according to
the following equation:
m
]/[Cu] [L] [H]
l
h
b
mlh ¼ [Cu
m
L
H
l h
(3)
Fig. 6 Fluorescence response of CST to various metal ions in MOPS (10 mM, pH
7
.2) buffered solution. The response is normalized with respect to free CST. (A)
2
+
ꢀ6
One equivalent of Cu or other metal ions were added to a 1 ꢂ 10
solution of CST. (B) Five equivalents of different metal ions were added to a 1 ꢂ
M
ꢀ6
2+
1
0
M CST solution containing one equivalent of Cu . Excitation wavelength
Fig. 4 Species distribution diagram for the copper(II)-CST system.
This journal is ª The Royal Society of Chemistry 2013
was 455 nm.
RSC Adv., 2013, 3, 24288–24297 | 24293

View Article Online
RSC Advances
Paper
2
+
The existence of a main species ([Cu(CST)] ), that, what is
more, has a 1 : 1 stoichiometry, is rather encouraging as it
would allow for a straightforward correlation between the
response of the proposed sensor and copper(II) concentration.
Chemosensor selectivity
Ratiometric titrations, that imply the measurement of the
changes in the ratio of the intensities of absorption or emission
bands at two wavelengths, were carried out to monitor the
selectivity of the CST ligand towards the target metal ion. Flu-
oroionophores employed for metal ions detection in aqueous
solution are oen affected by the interaction with protons and,
accordingly, we started off by looking for the suitable pH range
Fig. 7 Cell viability by MTTassay following a 24 hour treatment with CST (0.5 mM
and 2 mM). Bars are means ꢃ S.E.M. of three independent determinations (n ¼ 3). where the interaction of protons with the ligand is minimal. A
solution of CST was titrated with a base and we found that the

uorescence intensity of CST remains basically unaltered over
2
+
2+
the pH range from 5 to 8, while it dramatically decreases at
lower pH values (Fig. S8†).
Fig. S6† clearly shows that protons do not affect the uo-
rescence response of the ligand in the physiological pH region
Cu + L % [CuL]
(4)
(5)
2
+
+
+
Cu + L % [CuLH ] + H
ꢀ
1
The stability constant values for the equilibria shown in eqn and this makes CST a promising chemosensor for the pH range
4) and (5) are reported in Table 2 while the species distribution it is designed for. Accordingly, uorescence titrations of CST
(
resulting from these values is shown in Fig. 4.
with different metal ions were carried out at pH 7.2 using MOPS
The diagram emphasizes the importance of a thorough buffer. The addition of increasing amounts of a copper(II)
investigation of the species really forming in solution. solution to CST solutions caused a marked decrease of the
2
+
[
Cu(CST)] is the main species existing over the physiological intensity of the emission band at 535 nm and the formation of a
+
pH range; [CuCSTH 1] amounts to less than 5% below pH 7.5. blue-shied emission band at 477 nm (Fig. 5).
ꢀ
Fig. 8 LSM micrographs of differentiated neuroblastoma cells in 10 mM MOPS buffer (pH 7.4): (a) untreated control, (b) 500 nM CST (5 min), (c) 1 mM CST (5 min), (d)
1
mM CuSO
4 4
(30 s), (e) same as d (5 min), (f) 2 mM CuSO (5 min), (g) 2 mM CST (30 s), (h) same as g (5 min). In each panel three images are shown; from left to right:
l
ex/lem ¼ 405 nm/460–500 nm (blue); lex/lem ¼ 488 nm/500–600 nm (green); bright field micrographs.
2
4294 | RSC Adv., 2013, 3, 24288–24297
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
RSC Advances
The plot of the ratio of the intensities of the bands at 477 nm indicating that CST is internalized by cells. A labile uorescence
and 535 nm (I477/I535) as a function of the metal/ligand ratio is enhancement is also observed immediately aer the addition of
reported in Fig. 5 (inset). In line with the results obtained from 2 mM CST, likely due to the intrinsic uorescence of the probe in
UV-vis experiments (Fig. 4), the plot obtained at pH 7.2 shows a this specic spectral region. However, no changes are detected
break point at M/L values of about 1. This indicates that in the intracellular space by exogenous copper addition
2
+
[
Cu(CST)] species prevails at neutral pH although in associa- (monitored 5 minutes aer copper addition).
+
tion with minor amounts of [CuCSTHꢀ1] . However, the
On the other hand, in the extracellular space (Fig. 9b), the
simultaneous presence of minor species does not affect the addition of CST corresponds to an increase of uorescence,
ratiometric response of the proposed sensor in aqueous which is quenched by copper addition; this nicely parallels what
solution.
Similar experiments were carried out with some other metal
observed in the emission spectra of CST solution (see Fig. 5).
The imaging of intracellular copper by uorescent reporters
+
2+
2+
2+
2+
2+
2+
2+
2+
ions, such as Ag , Cd , Co , Fe , Hg , Ni , Pb , Zn , Ca , is a very complex issue, as it does not relate to the sensitivity and
2
+
+
Mg , Na , to ascertain whether the sugar-derivatized chemo- specicity of the probe only, but critically depends on the cell
2
+
sensor maintains the selectivity towards Cu already reported type used. Despite some reports on chemosensors able to detect
24
for the poorly water-soluble analogous. It was observed that copper(II) ions in live cells, including a pyrene-based sensor by
40
2+
41
the addition of the above metal ions caused, in all cases, a Wu et al., a coumarin-derived Cu uorescent sensor,
a
change of the uorescence spectra of CST though the change naphthalimide-based uorescent sensor for highly sensitive
2
+
2+
2+ 42
due to Cu is by far the most pronounced one (Fig. 6A). and selective detection of Cd and Cu , the redox differen-
+
2+
Competition experiments were carried out on a CST solution in tiation between Cu and Cu inside cells is still matter of
the presence of 1 equiv. of Cu ion along with 5 equiv. of the debate.
2
+
43
+
2+
2+
2+
2+
2+
2+
2+
other metal ions (Ag , Cd , Co , Fe , Hg , Ni , Pb , Zn ,
2
+
2+
+
Ca , Mg , Na , respectively). No signicant change of the
uorescence response of the CST-copper(II) solution was

observed when competing metal ions were added (Fig. 6B) thus
conrming that these metals do not interfere with the detec-
2
+
tion of Cu by the chemosensor. Noteworthy, the uorescence
response of the CST-copper(II) solution remains unaffected
even in the presence of fairly large anion concentrations (i.e.
anion/Cu : CST ratios ꢄ1000) (Fig. S9†). The above experiments
highlight that CST is selective for copper(II) in aqueous
solution.
LSM live-cell imaging
The differentiated cells, SH-SY5Y, do not show any sign of
toxicity following incubation with CST-copper concentrations
ranging from 0.5 to 2 mM (Fig. 7).
Representative results from live-cell imaging experiments
monitored with confocal microscopy are shown in Fig. 8 and 9.
Before starting any addition, the culture medium was rinsed off
and the adhered cells maintained in 10 mM MOPS buffer. The
experiments were run with increasing concentrations of CST/
4
CuSO as detailed in Fig. 8b–h. Both blue and green regions of

uorescence were monitored in real time aer CST and/or
copper supplementation. At the same time, the corresponding
bright eld images were acquired to assess whether the
different treatments used would affect cell morphology. The
optical micrographs in Fig. 8 conrm the MTT assays nding
that cell viability is not perturbed by any of the treatments, thus
demonstrating that CST is not toxic to the cells. As to the uo-
rescence micrographs, the slight intracellular auto uorescence
detected in the control cells (Fig. 8a) is darkened by CST addi- Fig. 9 Mean intensities measured over regions of interest (n ¼ 10) randomly
chosen in intracellular (a) and extracellular (b) regions of differentiated neuro-
tion (Fig. 8b–c, g and h), which, instead, brightens extracellular
regions; incidentally, such an effect is much more evident in the
green than in the blue channel.
blastoma cells kept in 10 mM MOPS buffer (pH ¼ 7.4). Data bars correspond to
the same sequence (a–h) detailed in Fig. 8. Error bars ¼ S.E.M. One-way Anova
test: * ¼ p < 0.01 w.r.t. control cells (ctrl); ** ¼ p < 0.01 w.r.t. CST (1 mM)-treated
Fig. 9a shows that the intracellular emission in the green
cells; *** ¼ p < 0.01 w.r.t. 2 mM copper-supplemented cells; **** ¼ p < 0.01 w.r.t.
region is signicantly affected by the addition of 1 mM CST, thus 2 mM CST-treated cells (evaluated after 30 seconds).
This journal is ª The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 24288–24297 | 24295

View Article Online
RSC Advances
Paper
Fig. 10 LSM micrographs of differentiated neuroblastoma cells, stained with Mitotracker (Red) and 2 mM CST, in 20 mM aCSF buffer (pH 7.4), before (a) and after
addition of 80 mM KCl (b). From left to right: lex/lem ¼ 405 nm/440–500 nm (blue); lex/lem ¼ 488 nm/540–600 nm (green); lex/lem ¼ 633 nm/640–700 nm (red);
bright field micrographs. In the insets: a magnified area in Hi-Lo look up table (LUT) setting, where blue pixels measure zero intensity and red pixels measure saturation.
Our results agree well with the broadly accepted view that
divalent copper concentration within cells is either null or
Concluding remarks
2
+
negligible; within cells the metal ion exists as Cu(I) and is bound A new Cu sensor has been synthesized and characterized in
to several proteins, chaperones and low-molecular weight aqueous solution. The novel chemosensor belongs to the turn-
ligands in a complex pathway mechanism which regulates/ on class and has the advantage over the uorescent chemo-
44
24,25
depends on copper homeostasis/dishomeostasis.
sensor synthesized by Xu et al.
of being water soluble while
2
+
2+
To address the capability of our CST to detect Cu in maintaining the selectivity for Cu shown by Xu's sensor and is
physiological conditions for the specic cell line used in the thus amenable to studies in aqueous solution. Fluorescence
2
+
present work, we focused on Cu release induced by cell experiments performed on neuroblastoma cells show that the
2
+
2+
depolarization. Indeed, Cu is known to be released from turn-on chemosensor is able to detect Cu present in the
synaptic vesicles following neuronal depolarization induced by extracellular medium. Moreover, CST readily detects in real-
2
+
4
5
KCl. The concentration of copper in the synaptic cle is tip- time Cu release induced by treatment of cells with KCl, indi-
2
+
ically in the range of ꢄ15 mM and may reach concentrations of cating that it may appropriately follow dynamic processes Cu
46
up to 100 mM.
is involved in, such as those relating to neuronal excitability
2
+
In this context, release of copper in the differentiated wherein Cu may reach concentrations as large as 100 mM.
neuroblastoma SH-SY5Y cells was followed by imaging CST-
treated cells following exposure to potassium ions.
Acknowledgements
Time-lapsed confocal micrographs (Fig. 10) clearly show a

uorescence change resulting from 80 mM KCl addition
We thank MIUR (PRIN 2009Z9ASCA, PRIN 2010MZJARJ, FIR-
B_RINAME_RBAP114AMK and FIRB_MERIT_RBN-E08HWLZ)
and University of Catania (Progetto d'Ateneo) for nancial
support.
(Fig. 10b). In particular, the blue emission signicantly
increases in the extracellular space while a marked blue uo-
rescence distribution along the peripheral processes is also
readily detected. On the other hand, the green emission shows
only a minor emission change compared to the blue uores-
cence enhancement (see insets in Fig. 10). Finally, no signi-
cant changes are observed for either the mitochondria
organelles (in red) or the bright eld images (in grey). These
Notes and references
1 I. Leray and B. Valeur, Eur. J. Inorg. Chem., 2009, 24, 3525.
2 G. Arena, F. Attanasio, D. Zhang, Y. Yang, R. A. Bartsch and
C. Sgarlata, Anal. Bioanal. Chem., 2013, 405, 1133.
3 H. Tapiero, D. M. Townsend and K. D. Tew, Copper Biomed.
Pharmacother., 2003, 57, 386.

ndings conrm the expected transfer of intracellular copper
pools upon KCl-mediated stimulation, thus pointing to the CST
capability to detect changes of labile Cu pools in living
differentiated SH-SY5Y neuroblastoma cells.
2
+
2
4296 | RSC Adv., 2013, 3, 24288–24297
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
RSC Advances
4
5
6
R. B. Franklin and L. C. Costello, Arch. Biochem. Biophys., 26 C. Bazzicalupi, A. Bencini, S. Bigini, E. Faggi, G. Farruggia,
2
007, 463, 211.
G. Andreani, P. Gratteri, L. Prodi, A. Spepi and
B. Valtancoli, Dalton Trans., 2010, 39, 7080.
27 G. I. Grasso, G. Arena, F. Bellia, G. Maccarrone,
M. Parrinello, A. Pietropaolo, G. Vecchio and E. Rizzarelli,
Chem.– Eur. J., 2011, 17, 9448.
A. Gonzalez, M. Peters, J. W. Lampe and E. White, Nutr.
Cancer, 2009, 61, 206.
Q. Pasha, S. A. Malik, J. Iqbal, N. Shaheen and M. H. Shah,
Biol. Trace Elem. Res., 2008, 125, 30.
7
8
9
M. Yaman, Curr. Med. Chem., 2006, 13, 2513.
A. I. Bush, Curr. Opin. Chem. Biol., 2000, 4, 184.
K. J. Barnham and A. I. Bush, Curr. Opin. Chem. Biol., 2008,
28 G. Gran, Analyst, 1952, 77, 661.
29 H. A. Flaschka, in EDTA Titrations, Pergamon Press, London,
UK, 1959.
1
2, 222.
30 I. M. Kolthoff, E. B. Sandell, E. J. Meehan and
S. Bruckenstein, in Quantitative Chemical Analysis, The
Macmillan Company, New York, 1969, vol. 2.
1
1
1
1
1
0 P. J. Crouch, K. J. Barnham, A. I. Bush and A. R. White, Drug
News Perspect., 2006, 19, 469.
1 G. Filiz, K. A. Price, A. Caragounis, T. Du, P. J. Crouch and 31 R. P. Bonomo, R. Cal `ı , V. Cucinotta, G. Impellizzeri and
A. R. White, Eur. Biophys. J., 2008, 37, 315. E. Rizzarelli, Inorg. Chem., 1986, 25, 1641.
2 G. Arena, A. Contino, G. Maccarrone, D. Sciotto and 32 G. Arena, E. Rizzarelli, S. Sammartano and C. Rigano,
C. Sgarlata, Tetrahedron Lett., 2007, 48, 8274. Talanta, 1979, 26, 1.
3 G. Arena, R. P. Bonomo, A. Contino, C. Sgarlata, G. Spoto and 33 P. Gans, A. Sabatini and A. Vacca, Talanta, 1996, 43, 1739.
G. Tabb `ı , Dalton Trans., 2004, 3205.
34 L. Alderighi, P. Gans, A. Ienco, D. Peters, A. Sabatini and
4 C. Satriano, G. Trusso Sfrazzetto, M. E. Amato,
A. Vacca, Coord. Chem. Rev., 1999, 184, 3111.
F. P. Ballistreri, A. Copani, M. L. Giuffrida, G. Grasso, 35 W. S. Rasband, Image J, National Institutes of Health,
A. Pappalardo, E. Rizzarelli, G. A. Tomaselli and
Bethesda, Maryland, USA, 1997–2012.
R. M. Toscano, Chem. Commun., 2013, 49, 5565.
36 S. Hannessian and P. Lavall `e e, Carbohydr. Res., 1973, 28, 303.
15 E. Rampazzo, S. Bonacchi, D. Genovese, R. Juris, M. Sgarzi, 37 B. Di Blasio, S. Galdiero, M. Saviano, C. Pedone, E. Benedetti,
M. Montalti, L. Prodi, N. Zaccheroni, G. Tomaselli,
S. Gentile, C. Satriano and E. Rizzarelli, Chem.–Eur. J.,
E. Rizzarelli, S. Pedotti, G. Vecchio and W. Gibbons,
Carbohydr. Res., 1996, 282, 41.
2
011, 17, 13429.
38 A. T. Peters and Y. S. S. Behesti, J. Soc. Dyers Colour., 1989,
105, 29.
1
1
6 R. Kramer, Angew. Chem., Int. Ed., 1998, 37, 772.
7 G. C. Yu, Z. B. Zhang, C. Y. Han, M. Xue, Q. Z. Zhou and 39 Z. Xu, X. Qian, J. Cui and R. Zhang, Tetrahedron, 2006, 62,
F. H. Huang, Chem. Commun., 2012, 48, 2958. 10117.
18 R. Pandey, P. Kumar, A. K. Singh, M. Shahid, P. Z. Li, 40 H. F. Wang and S. P. Wu, Sens. Actuators, B, 2013, 181, 743.
S. K. Singh, Q. Xu, A. Misra and D. S. Pandey, Inorg. Chem., 41 H. S. Jung, P. S. Kwon, J. W. Lee, J. I. Kim, C. S. Hong,
2
011, 50, 3189.
9 M. M. Zhang, K. L. Zhu and F. H. Huang, Chem. Commun.,
010, 46, 8131.
J. W. Kim, S. Yan, J. Y. Lee, J. H. Lee, T. Joo and J. S. Kim,
J. Am. Chem. Soc., 2009, 131, 2008.
42 W. Wang, Q. Wen, Y. Zhang, X. Fei, Y. Li, Q. Yang and X. Xu,
Dalton Trans., 2013, 42, 1827.
43 M. J. Pushie, L. Zhang, I. J. Pickering and G. N. George,
Biochim. Biophys. Acta, 2012, 1817, 938.
1
2
2
0 S. Yin, W. Yuan, J. Huang, D. Xie, B. Liu, K. Jiang and H. Qiu,
Spectrochim. Acta, Part A, 2012, 96, 82.
1 F. A. Abebe and E. Sinn, Tetrahedron Lett., 2011, 52, 5234.
2
2
2 E. L. Que, E. Gianolio, S. L. Baker, A. P. Wong, S. Aime and 44 G. Pappalardo, D. Milardi, E. Rizzarelli and I. Sovago, in
C. J. Chang, J. Am. Chem. Soc., 2009, 131, 8527.
3 G. K. Li, Z. X. Xu, C. F. Chen and Z. T. Huang, Chem.
Commun., 2008, 1774.
Neurodegeneration: Metallostasis and Proteostasis, D. Milardi
and E. Rizzarelli, ed. RSC Publishing, Cambridge, 2011.
45 A. Hopt, S. Korte, H. Fink, U. Panne, R. Niessner, R. Jahn,
H. Kretzsschmar and J. Herms, J. Neurosci. Methods, 2003,
128, 159.
2
2
2
4 Z. Xu, Y. Xiao, X. Qian, J. Cui and D. Cui, Org. Lett., 2005, 7,
889.
5 J. Huang, Y. Xu and X. Qian, Org. Biomol. Chem., 2009, 7, 46 J. Kardos, I. Kovacs, F. Hajos, M. Kalman and M. Simonyi,
299. Neurosci. Lett., 1989, 103, 139.
1
This journal is ª The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 24288–24297 | 24297