Novel 7-(dimethylamino)fluorene-based fluorescent probes and their binding to human serum albumin

Article abstract of DOI:10.1039/b911605b

A novel solvatochromic fluorescent molecule, 9,9-dibutyl-7-(dimethylamino)- 2-fluorenesulfonate 2 was synthesized from 2-nitrofluorene in moderate yield. The fluorescence spectra of 2 and 7-(dimethylamino)-2-fluorenesulfonate 1 shift to shorter wavelength

Full text of DOI:10.1039/b911605b

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Novel 7-(dimethylamino)fluorene-based fluorescent probes and their binding to
human serum albumin† ‡
a
b
c
Kwanghee Koh Park,* Joon Woo Park* and Andrew D. Hamilton
Received 16th June 2009, Accepted 16th July 2009
First published as an Advance Article on the web 14th August 2009
DOI: 10.1039/b911605b
A novel solvatochromic fluorescent molecule, 9,9-dibutyl-7-(dimethylamino)-2-fluorenesulfonate 2 was
synthesized from 2-nitrofluorene in moderate yield. The fluorescence spectra of 2 and 7-(dimethyl-
amino)-2-fluorenesulfonate 1 shift to shorter wavelengths as the polarity of the medium decreases. Both
1
and 2 bind to hydrophobic sites of human serum albumin (HSA). The apparent binding constants
6
-1
6
-1
were determined by fluorescence titration to be 0.37 ¥ 10 M for 1 and 2.2 ¥ 10 M for 2. The energy
of the Trp-214 fluorescence of HSA is transferred to the HSA-bound fluorophores with near 100%
efficiency. The covalent bonding of acrylodan (AC) to Cys-34 has little effect on the binding affinity of 2
to HSA or fluorescent behavior of HSA-bound 2. Bound 2 also has little effect on the fluorescence of
AC, but 2→AC and Trp-214→2→AC resonance energy transfers were observed. Competitive binding
between the fluorene compounds and other ligands such as 1-anilino-8-naphthalenesulfonate, aspirin,
S-(+)-ibuprofen and phenylbutazone were also studied fluorometrically. The results indicated that the
primary binding site of 2 to HSA is site II in domain IIIA, whereas 1 binds to site I in domain IIA, but
a different region from the phenylbutazone binding site. Because of its large molar absorptivity, strong
fluorescence, sensitivity to its environment, and high binding constant to HSA, 2 can be used
successfully in the study of proteins and their binding properties.
3
,5-11
Introduction
(site I) and IIIA (site II) (Fig. 1).
These binding sites are often
selective, however, binding to multiple sites on HSA has also been
Fluorescent molecules whose spectral properties are exquisitely
sensitive to their environments have been widely used as probes
for the study of membrane microenvironments and hydrophobic
domains on proteins, as well as the dynamics and conformational
changes of macromolecules. In addition to the environmental
sensitivity of the fluorescent properties, a large molar absorptivity
and fluorescence quantum yield are desirable characteristics for a
highly sensitive fluorescent probe. Moderate solubility in aqueous
media and high binding affinity for the target are also requisite for
applications with biological molecules.
5,6,8-16
observed.
The HSA-binding affinity of a compound is the key
factor for determining the concentration of its unbound form in
the plasma. Thus the binding affinity of a drug to HSA affects the
distribution, pharmacokinetics, toxicity and rate of excretion of
1,2
3,8
the drug. Therefore, information on the binding affinity of a drug
to HSA is particularly useful for the identification of potential in
11,17
vivo pharmaceutical problems.
Human serum albumin (HSA) is the most abundant protein in
the circulatory system, and its principal function is to transport
a variety of bioactive molecules such as fatty acids, steroid
hormones, vitamins and numerous pharmaceuticals. HSA also
3,4
contributes significantly to colloid osmotic pressure. HSA
contains 585-amino acids with three homologous a-helical do-
4-8
mains (I-III) and has a single tryptophan (Trp-214) residue.
Hydrophobic or amphiphilic ligands for HSA bind mostly to one
of the two principal binding sites, located in the subdomain IIA
a
Department of Chemistry, Chungnam National University, Daejeon, 305-
7
64, Korea. E-mail: khkoh@cnu.ac.kr
b
Department of Chemistry, Ewha Womans University, Seoul, 120-750, Korea.
E-mail: jwpark@ewha.ac.kr
c
Department of Chemistry, Yale University, New Haven, CT, 06520-8107,
USA
†
Part of this work was carried out during KKP and JP’s stay at Yale
University.
Electronic supplementary information (ESI) available: Absorption,
‡
Fig. 1 Crystal structure of human serum albumin and the location of
major binding sites. The positions of Trp-214 and Cys-34 are shown. The
structure was obtained from a protein data bank (1ha2).
fluorescence and excitation spectra of 1 and 1/HSA mixture; fluores-
cence spectra of 1/HSA mixture in the presence of aspirin. See DOI:
1
0.1039/b911605b
This journal is © The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4225–4232 | 4225

View Article Online
A variety of naphthalene-based fluorescent probes such as
Results and discussion
18
5
-(dimethylamino)-1-naphthalenesulfonate,
1-anilino-8-naph-
15,16,19,20
The ammonium salt of 1 is slightly soluble in water and in polar
organic solvents, but practically insoluble in non-polar solvents.
To increase its solubility in water and organic solvents and also
to modify its fluorescent properties and binding affinity to HSA,
the 9-position of 1 was doubly alkylated to obtain 2. Starting
from 2-nitrofluorene 3, compound 2 was synthesized in four steps
thalenesulfonate (ANS),
naphthalene (acrylodan),
6-acryloyl-2-dimethylamino-
6-propionyl-2-dimethylamino-
6-propionyl-2-methoxynaphthalene
39
2
1-26
2
2,26-28
naphthalene (prodan),
promen),
22,29
30
9,14,31-33
(
orange II, dansyl derivatives,
and 1-
3
4
hydroxy-2-acetonaphthone have been used for fluorescent
studies of HSA: acrylodan binds to Cys-34 covalently, while
others bind non-covalently to HSA. Other fluorescent molecules,
(Scheme 1) in moderate overall yield (see Experimental Section for
9,11,14,31,33,35,36
details). 2 was highly soluble in water as well as in organic solvents.
We first compare the absorption and fluorescence spectroscopic
characteristics of 1 and 2 in various solvent media, and then
describe the binding characteristics of 1 and 2 to HSA. Energy
transfer from excited Trp-214 to HSA-bound 2, and from HSA-
bound 2 to Cys-34-bound acrylodan (AC) in the HSA complex
are followed. Finally, competitive binding studies of 1 and 2 with
other ligands (shown below) of pharmaceutical interest to HSA
are presented.
e.g. warfarin
and 2-(6-diethylaminobenzo[b]furan-2-
36
yl)-3-hydroxychromone have also been used for this role. Using
these probes, the characteristics of ligand binding and binding
sites of HSA, competitive binding of other ligands, and the
spatial relationship between Trp-214 and the probe-binding sites
have been investigated. Also the fluorescent probes have been
20,21,26,28,32
utilized for the investigation of unfolding processes
and
25,30
dynamics
of HSA.
It has been reported that fluorescent properties of aminofluo-
renes are sensitive to solvent environments, in a similar way to
3
7-39
aminonaphthalenes.
This, together with the large molar ab-
38-40
sorptivity, high fluorescence quantum yield,
and high binding
41
affinity of 2-(dimethylamino)fluorenes to b-amyloid, suggested
that aminofluorene derivatives might be novel and sensitive fluo-
rescent probes for proteins and biomimetic systems. In a previous
paper, we reported a novel water-soluble aminofluorene derivative,
7
-(dimethylamino)-2-fluorenesulfonate 1 and the effects of solvent
39
polarity on its absorption and fluorescence behavior. Here we
report the synthesis of another water-soluble fluorene derivative,
9
,9-dibutyl-7-(dimethylamino)-2-fluorenesulfonate 2, its solvent-
dependent fluorescent properties, and a comparative HSA binding
study of 1 and 2. We demonstrate that the fluorene compounds
can be used as highly sensitive fluorescent probes for studies of
biological and bio-mimetic systems.
Solvent dependence of UV-visible absorption and fluorescence
spectra of 1 and 2
Compounds 1 and 2 show absorption bands in the 260–380 nm
region with that of 2 appearing at a slightly longer wavelength
than 1. Though the shape and position of the absorption bands
vary somewhat with solvent, no systematic dependence on solvent
parameters could be deduced (see ref. 39 for variation of the
spectrum of 1 on solvent). Close examination of the spectra
reveals that it is a sum of at least two transitions, presumably
Scheme 1 Synthetic pathway to 2.
4
226 | Org. Biomol. Chem., 2009, 7, 4225–4232
This journal is © The Royal Society of Chemistry 2009

View Article Online
a
Table 1 Comparison of absorption and fluorescence characteristics of 1 and 2
Absorption
Fluorescence
1
2
1
2
b
b
Medium
lmax/nm
log emax
lmax/nm
log emax
lmax/nm
I
max
lmax/nm
I
max
Water (pH 7.0)
319
321
318
314
314c
319
—
4.38
4.39
4.40
4.42
4.41c
4.44
325
337
335
318
320
344
317
335
4.33
4.38
4.36
4.34
4.36
4.38
4.39
4.36
432
397
385
383
376
375
1.00
1.11
1.12
1.06
1.13
1.15
427
395
385
383
375
374
380
373
1.00
1.16
1.23
1.25
1.31
0.99
0.90
1.04
CH
3
OH
OH
CN
-PrOH
C
2
H
5
CH
2
3
d
p-dioxane
DMF
THF
—
a
b
Data for 1 are taken from ref. 39. Relative intensity at emission maxima, normalized to light absorption by dividing the measured intensities by
-Abs
c
d
(
1–10 ): Abs is absorbance at lex
.
In 96% p-dioxane–4% water. In 98% p-dioxane–2% water (lmax,em = 378 nm).
39
to different vibrational states, and the lower energy transition is
more preferred in non-polar media. This is more pronounced for 2
than for 1. The absorption maxima and log emax values in various
solvents are summarized in Table 1.
In a previous report, we showed that the fluorescence sensitivity
(
(
e
max ¥ U ) of 1 is about 15–25 times greater than that of
F
39
dimethylamino)naphthalenes. The fluorescence sensitivity of
2
was similar to that of 1. In all solvent media studied, the
fluorescence spectra of 1 and 2 show a single band which shows
blue shift with low solvent polarity. The normalized fluorescence
spectra of 2 are shown in Fig. 2 (the spectra of 1 can be found
in ref. 39). The emission maxima and relative emission intensities
are included in Table 1. The observed dependence of fluorescence
spectra on solvent polarity is typical of that of emission from an
1
excited intramolecular charge transfer (ICT) state, and indicates
that 1 and 2 can be used as fluorescent probes.
Fig. 3 Absorption spectra of HSA, 2, HSA-bound 2, ANS, and acrylodan
(
(
AC). The spectrum of HSA-bound 2 is assumed to be that of an equimolar
2.0 ¥ 10 M each) mixture of 2 and HSA taken against HSA blank.
-5
Solvent media are aqueous 0.01 M phosphate buffer (pH 7.0, I = 0.1 M
with NaCl), except ANS and AC which were taken in methanol.
absorption maximum of 1 shifted from 319 nm in water to 335 nm
upon addition of HSA (ESI, Fig. S1‡).
The dependence of the fluorescence spectra of 1 and 2 on the
concentrations of HSA are shown in Fig. 4. As the concentration
of HSA increases, the emission at longer wavelength, which is from
free fluorophores, decreases and a new emission band at shorter
wavelength, which is from HSA-bound fluorophore, grows. This
clearly reflects the binding of 1 and 2 to hydrophobic sites of HSA.
At high concentration of HSA, the emission maximum of 1 is
Fig. 2 Normalized fluorescence spectra of 2. Solvents are H
2
O (pH 7),
CH OH, C OH, DMF, 2-PrOH and THF (from right to left of the
3
2 5
H
-6
spectra). [2] = 1.9 ¥ 10 M. lex = 334 nm. (For relative intensity, see
Table 1).
4
4
10 nm which is similar to that observed in a 60 (v/v)% methanol–
0% water mixture, whereas that of 2 is 377 nm, similar to that in
Binding of 1 and 2 to HSA
39
Addition of HSA to the solutions of 1 or 2 (pH 7, phosphate buffer)
results in large spectroscopic changes. Fig. 3 shows the absorption
spectra of HSA, 2, and HSA-bound 2. A red-shift of absorption
spectrum of 2 upon binding to HSA, i.e. from 325 nm in water
to 340 nm in the presence of HSA, is clearly seen. Similarly, the
2-propanol. This suggests that 1 and 2 might bind to hydrophobic
sites of HSA of different polarity and/or the exposed extent of the
bound fluorophores may be different.
Fluorescence titrations of 1 and 2 with HSA show iso-emissive
points only at [HSA]/[F] ꢀꢀ 1 (Fig. 4): F denotes the fluorophore.
This journal is © The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4225–4232 | 4227

View Article Online
Fig. 5 Changes of fluorescence intensity (DI
F
) of 1 at 397 nm (᭹) and 2
at 377 nm (᭺) with [HSA]. The contributions from HSA to the measured
intensities were corrected. The DI scale was adjusted to show similar DIF,•
for both systems. Inset shows the plots of DI /[HSA] against DI
. [1] =
F
F
F
-7
[
2] = 1.0 ¥ 10 M.
substituent of 2 increases the binding affinity to HSA. The K value
of 2 is significantly larger than those of widely used fluorescent
6
-1 9,14,31,32
probes for HSA, dansyl derivatives (0.02–0.5 ¥ 10 M ),
6
-1 31,35
6
-1 27,28
warfarin (ca. 0.4 ¥ 10 M ),
prodan (0.1–0.4 ¥ 10 M ),
6
-1 29
6
-1 9,15
promen (0.2 ¥ 10 M ) and 1,8-ANS (0.9 ¥ 10 M ). From
DIF,• values, the fluorescence intensities at emission maxima of the
HSA-bound 1 and 2 were found to be 1.8 and 3.0 times greater
than the respective values in buffer. The enhancement of emission
intensities upon binding to HSA are much greater than those
in homogeneous media caused by solvent change (Table 1). As
we excited the solutions at isosbestic points, the enhancement
of emission intensity upon binding to HSA is approximately
proportional to the increase in the emission quantum yields of the
fluorophores. This may arise from protection of the fluorophores
from quenching and/or from decrease in the non-radiative decay.
The high binding constant, and large fluorescence quantum
yield and molar absorptivity, together with the large fluorescence
spectral change upon binding to HSA give promise that 2 could be
utilized as a sensitive fluorescent probe even at low concentrations
of the probe and HSA.
-
7
Fig. 4 Fluorescence spectra of 1.0 ¥ 10 M 1 (A) and 2 (B) in the presence
of various concentrations of HSA. [HSA] are given in the figure. The
excitation wavelengths were 330 nm for 1 and 320 nm for 2.
A plausible explanation for this is that the fluorophores bind to
multiple sites of HSA, and the binding is non-cooperative. At
a high [HSA]/[F] ratio, there would be little chance of multiple
binding, and thus a microscopic binding equilibrium to site i can
be written as eqn (1) and the fraction (f ) of F bound to any site of
HSA is given by eqn (2).
F + HSA = (F-HSA)
i
; K
i
= [(F-HSA)
i
]/{[F][HSA]}
(1)
(2)
f = R [(F-HSA) ]/[F] = K[HSA]/(1 + K[HSA]); K = R
i
i
o
i i
K
Eqn (2) is analogous to that derived for 1:1 binding, but the
apparent binding constant K equals the sum of the microscopic
binding constants. As the observed fluorescence intensity is the
sum of the intensities from contributing species, the intensity
Resonance energy transfer from Trp-214 to HSA-bound 1 or 2
HSA has a single tryptophan residue (Trp-214) which gives a
fluorescence emission in the 300–400 nm region. The emission
band of Trp-214 overlaps with absorption bands of 1 and 2.
Thus, F o¨ rster resonance energy transfer from the excited Trp-
214 to the fluorophores is possible within the complexes. Fig. 6
shows the emission spectra of HSA in the presence of various
concentrations of 2, excited at 282 nm where the tryptophan
residue absorbs. At a given concentration of HSA, the tryptophan
fluorescence (~340 nm) decreases and the fluorescence from
HSA-bound 2 (~380 nm) increases as the concentration of 2 is
higher. No appreciable tryptophan fluorescence is observed at high
concentration of 2. Similar results are observed in the presence
of 1 (ESI, Fig. S2‡). These indicate near 100% efficiency in the
resonance energy transfer from the excited Trp-214 to HSA-bound
change (DI
F
) caused by the presence of HSA is derived as;
DI
F
/[HSA] = DIF,• K - DI
where DIF,• is the intensity change expected when all of F is bound
F
K
(3)
to HSA. Under the condition of a large excess of HSA, [HSA] can
be replaced by total HSA concentration, [HSA]
In Fig. 5, DI values of 1 and 2 at wavelengths showing the largest
emission intensity change as functions of [HSA] are plotted. The
plots of DI /[HSA] versus DI according to eqn (3) are given in
the inset of the figure. The plots gave good linearity with the
coefficient of correlation >0.98. The K values of 0.37 ¥ 10 M for
and 2.2 ¥ 10 M for 2 were obtained from the plots. The larger
o
.
F
o
F
F
6
-1
6
-1
1
binding constant of 2, compared to 1, suggests that the dibutyl
4
228 | Org. Biomol. Chem., 2009, 7, 4225–4232
This journal is © The Royal Society of Chemistry 2009

View Article Online
of the reaction mixture was varied from 0.1–10 mM suggesting
nearly quantitative covalent bonding of the reacting pair.
Fig. 7 shows three sets of fluorescence spectra (A–C) of a HSA–
AC conjugate solution in the presence of various concentrations
of 2 taken at different excitation wavelengths. The behavior of the
tryptophan fluorescence (Set A: 300–360 nm) taken with lex
90 nm is similar to that observed with free HSA solution in
the presence of 2 (Fig. 6). It was reported that 40% of Trp-214
=
2
22
fluorescence of the HSA–AC conjugate is transferred to AC. The
remaining Trp-214 fluorescence is further quenched by 2 indicating
binding of 2 to the HSA–AC conjugate. The binding is clearly seen
in the set B spectra (350–480 nm) taken with lex = 320 nm where
mostly 2 absorbs. When the [2]/[HSA-AC] ratio is less than 1,
the fluorescence spectra of this set are similar to that of AC-free
HSA-bound 2. From analysis of the fluorescence titration data
of HSA–AC with 2 using linear regression methods described in
42
a previous report, the binding constant for 2 to the HSA–AC
Fig. 6 Fluorescence spectra of HSA/2 solutions excited at 282 nm.
6
-1
conjugate was estimated to be approximately 1 ¥ 10 M . This
is similar to the binding constant of 2 to the native HSA. When
the solution was excited at 400 nm where only AC absorbs, a
slight decrease in AC fluorescence intensity, without noticeable
change in spectral shape, was observed (set C). This implies that
the bound AC does not alter the environment of the binding site
of 2 significantly. Also it appears that the binding of 2 to HSA–AC
does not affect the environment of AC.
-6
-6
[
0
6
HSA] = 1.7 ¥ 10 M. The concentrations of 2 are (1), 0.00 ¥ 10 ; (2),
-6
-6
-6
-6
.47 ¥ 10 ; (3), 0.94 ¥ 10 ; (4), 1.9 ¥ 10 ; (5), 3.8 ¥ 10 M. The spectrum
6
-
is from 0.94 ¥ 10 M 2 solution without HSA.
1
and 2. The efficiency is much greater than the corresponding
22,28
22
energy transfer efficiency to prodan
or promen, which is 38–
5
0%, suggesting that 2 is a better resonance energy acceptor from
excited Trp-214.
To estimate the contribution of the Trp-214 sensitized emission
to the fluorescence of HSA-bound species, we compared the
spectra 3 and 6 of Fig. 6, which were taken with the same
concentration of 2 in the presence and in the absence of HSA,
respectively. The maximum emission intensity (at 377 nm) of
spectrum 3 is 2.4 times greater than that of spectrum 6 (at 427 nm).
Unbound 2 contributes little to emission intensity at 377 nm (see
Fig. 4B). The fraction of 2 bound to HSA is calculated to be 0.70
from the binding constant of 2 with HSA and the concentrations
of 2 and HSA in the solution used for spectrum 3. Thus, it appears
that HSA-bound 2 gives 3.4 (= 2.4/0.7) times greater emission
intensity than free 2 at their emission maximum when lex = 282 nm.
The molar absorptivities of HSA-bound 2 and free 2 are estimated
as 5300 M- cm and 8300 M cm , respectively, at 282 nm
1
-1
-1
-1
(
Fig. 3). From these values, it is estimated that the HSA-bound 2
gives 5.3 times greater emission intensity than the free 2 at their
respective emission maxima. This is much greater than the ratio
Fig. 7 Fluorescence spectra of 2/HSA–acrylodan solutions. [HSA–
-6
-6
acrylodan] = 1.0 ¥ 10 M. The concentrations of 2 are (1), 0 ¥ 10 ;
-6 -6 -6 -6
3
.0 found in the previous section when the solutions are excited
at the isosbestic point of HSA-bound 2 and free 2. Thus, about
3% of 377 nm emission of HSA-bound 2 can be attributed to the
(2), 0.3 ¥ 10 ; (3), 0.6 ¥ 10 , (4) 1.0 ¥ 10 ; (5), 1.5 ¥ 10 M. The excitation
wavelengths were 290 nm for group A, 320 nm for group B, and 400 nm
for group C.
4
Trp-214 sensitized emission. This estimation matches well with
the similarity of molar absorptivities of the HSA-bound 2 and
-
1
-1
2
tryptophan (5600 M cm ), and the near 100% energy transfer
efficiency.
Competitive binding with other ligands
To obtain information on the binding site(s) of 1 and 2 in
HSA, competitive binding between the fluorene compounds and
other ligands such as 1-anilino-8-naphthalene sulfonate (ANS),
aspirin, S-(+)-ibuprofen (IBU) and phenylbutazone (PBZ) was
studied. Because of binding to multiple sites, ambiguity of binding
Binding to the HSA–acrylodan conjugate
Acrylodan (AC) was covalently attached to the Cys-34 site of
23,25,26
HSA.
Upon binding to HSA, the emission maximum of AC is
7-10,14-16,33
blue-shifted to ca. 480 nm from 525 nm in aqueous buffer, revealing
a hydrophobic character of the microenvironment surrounding
Cys-34 in HSA. We prepared the HSA–AC conjugate by reacting
HSA with AC in a 1:1 molar ratio. No appreciable change in
the fluorescence spectrum was observed when the concentration
constants,
the ligands bound to different sites,
and possible allosteric interactions between
9,44
it is difficult to analyze
the competitive binding data quantitatively. However, qualitative
analysis could be carried out using reported binding data for
competitors.
This journal is © The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4225–4232 | 4229

View Article Online
8
,9,43
The right panel of Fig. 8 shows the effect of 2 on the fluorescence
subdomains IIA and IIIA of HSA:
the binding constant to
-
6
-6
4
-1 9,15,43
spectra of 5.0 ¥ 10 M ANS in 1.2 ¥ 10 M HSA solution. As ANS
is highly fluorescent in non-polar media or in the hydrophobic
pockets of proteins, but very weakly fluorescent in homogeneous
aqueous media, the decrease in the ANS fluorescence intensity
upon the addition of 2 reflects the displacement of the bound
ANS by 2. It was reported that there are at least two types
the high affinity site was reported to be about 2 ¥ 10 M ,
3
-1 9,43
whereas that to low affinity site was 1.8 ¥ 10 M . Under our
-
7
-6
experimental conditions of 2.0 ¥ 10 M 2 and 1.0 ¥ 10 HSA, the
fluorescence intensity at 377 nm is decreased to about half by the
-
4
presence of 5 ¥ 10 M aspirin. As the free 2 does not contribute to
the fluorescence intensity at 377 nm, the results imply that about
half of HSA-bound 2 is displaced by aspirin. If 2 binds to multiple
sites on HSA, the extent of displacement from the sites for which it
is in competition with aspirin would be greater than this. Ignoring
the binding of 2, the fraction of HSA of which any given site is
not occupied by aspirin is given by 1/(1 + Kaspirin[aspirin]). Using
the reported Kaspirin values given above, the fractions for the high
affinity site and low affinity site are calculated as 0.09 and 0.5,
respectively, at [aspirin] = 5 ¥ 10 M: as aspirin is in large excess
of HSA, the total aspirin concentration is used for [aspirin]. If 2
and aspirin compete for the low affinity aspirin binding site, the
effect of 5 ¥ 10 M aspirin on the fluorescence of 2.0 ¥ 10 M 2 is
approximately the same as that by decreasing the concentration of
HSA to 0.5 ¥ 10 M from 1.0 ¥ 10 M in the absence of aspirin. As
the fraction of 2 bound to HSA (f HSA) is given by KHSA[HSA]/(1+
KHSA[HSA]), the f HSA value changes to about 0.48 from 0.65 by the
addition of 5 ¥ 10 M aspirin. This estimated change of f HSA is
much smaller than the observed displacement result. Thus we can
conclude that 2 is displaced from the high affinity aspirin binding
site. As 2 binds mostly to site II, as revealed from displacement of
ANS by 2, we can conclude that the high affinity site for aspirin
is also site II. This supports the conclusion from displacement
9,15,16
of fluorescence-active ANS binding sites in HSA.
The high
affinity site accommodates one ANS molecule with a binding
6
-1
constant (K) of ca. 0.9 ¥ 10 M , and the low affinity site can
6
-1 9,15
bind 2–3 molecules with K values of 0.08–0.13 ¥ 10 M .
A
recent paper reported about 6 times greater binding constants for
16
the respective sites, but these values are too large to account for
our results. Competitive binding studies with specific site markers
suggested that the primary binding site for ANS is the site II
-
4
9,15
in the subdomain IIIA. Bagatolli et al. showed that the ANS
molecules bound to the low affinity sites contribute about 20%
of ANS fluorescence from ANS/HSA solutions under similar
-
4
-7
15
conditions employed in this work. The 20% contribution of
ANS fluorescence from low affinity sites is much smaller than
the observed decrease in the ANS fluorescence intensity upon
-
6
-6
-
6
addition of 2 (43% decrease in the presence of 10 ¥ 10 M 2). This
implies that ANS and 2 compete for the site II of HSA, which is
the primary binding site of ANS.
-
4
15
experiments of ANS by aspirin.
We also studied the displacement of HSA-bound 1 by aspirin. A
clear displacement was shown, but the extent of the displacement
was much smaller than that of 2 (ESI, Fig. S4‡). This implies
that the primary binding site for 1 may be different from that
of 2.
More conclusive evidence for the difference in the primary
binding site for 1 and 2 was obtained from competitive binding
studies using site specific markers, S-(+)-ibuprofen (IBU) and
phenylbutazone (PBZ). The primary binding site of IBU is
5,31,33
5,8,10,14,31
reported to be site II,
whereas that of PBZ is site I.
Fig. 9 shows the relative fluorescence intensities of 1 and 2 in
HSA solutions, measured at emission maxima of the respective
HSA complexes, as a function of IBU or PBZ concentration. The
emission from HSA-bound 1 was increased by the addition of
IBU and PBZ. The emission maximum was shifted to shorter
wavelength by 2 nm in the presence of 40 mM PBZ. This indicates
that 1 is not displaced from HSA by either IBU or PBZ. The
enhancement of fluorescence of a probe by the addition of
drugs was reported for the case of ANS with iophenoxic acid
Fig. 8 (Right) Fluorescence spectra of 1,8-ANS/HSA solutions in the
-6
presence of various concentrations of 2. [1,8-ANS] = 5.0 ¥ 10 M; [HSA] =
-6
-6
1
.2 ¥ 10 M. The concentrations of 2 are 0, 1, 2, 3, 5, and 10 ¥ 10 M (from
top to bottom). The spectrum D is the difference spectrum between the
9
9
spectra 1 and 6. lex = 400 nm. (Left) Effect of aspirin on the fluorescence
and furemide, and for dansyl compounds with warfarin and
-
7
-6
spectra of 2/HSA solution. [2] = 2.0 ¥ 10 M; [HSA] = 1.0 ¥ 10 M;
31
other drugs. This was attributed to sensitivity of the probe to
-4
[
aspirin] = 0, 1.0, 2.0, 3.5, and 5.0 ¥ 10 M (from top to bottom). The
conformational changes in the protein occurring upon binding of
contributions of aspirin/HSA to the spectra were subtracted from each
31
the drugs. The blue shift in the emission maximum of 1 caused by
2
/HSA/aspirin spectrum. lex = 350 nm.
PBZ suggests that the PBZ-induced conformational change makes
the environment of bound 1 more hydrophobic.
The left panel of Fig. 8 shows the displacement of HSA-
The addition of IBU decreased the fluorescence from HSA-
bound 2 and increased the emission from free 2, whereas PBZ had
little effect on 2. This confirms that the primary binding site of 2
is site II, which is consistent with the conclusion obtained from
displacement studies of ANS by 2, and 2 by aspirin (vide supra).
bound 2 by aspirin. The addition of aspirin to 2/HSA solution
clearly decreases the emission from HSA-bound 2 (~377 nm)
and increases the emission from free 2 (~430 nm). It was shown
that aspirin is known to bind to at least two sites located in the
4
230 | Org. Biomol. Chem., 2009, 7, 4225–4232
This journal is © The Royal Society of Chemistry 2009

View Article Online
Experimental
Synthesis of 9,9-dibutyl-7-(dimethylamino)-2-fluorenesulfonic
acid 2
To a stirred reaction mixture of 3 (1.055 g, 5 mmol) and caesium
carbonate (4.885 g, 15 mmol) in DMF (25 mL) was added
n-butyl bromide (2.054 g, 1.62 mL, 15 mmol) under a nitrogen
atmosphere and stirring was continued at room temperature for
1
6 h. Water (30 mL) was added to the reaction mixture which was
then extracted with dichloromethane (3 ¥ 30 mL). The organic
layers were dried, concentrated, and then purified by column
chromatography (eluent: dichloromethane: hexane = 10:3) to give
◦
1
4
as light yellow solid (1.48 g, 91%). 4: Mp, 72–73 C; H NMR
(
CD
2
Cl
2
, 500 MHz): d 8.23 (dd, J = 8.5, 2.0 Hz, 1H), 8.20 (d, J =
2
4
0
1
1
.0 Hz, 1H), 7.82-7.79 (m, 2H), 7.44-7.37 (m, 3H), 2.09-1.99 (m,
Fig. 9 Effect of S-(+)-ibuprofen (᭹, ꢀ) and phenylbutazone (᭺, ꢀ) on
the 1/HSA (ꢀ, ꢀ) and 2/HSA (᭹, ᭺) solutions. 1/HSA: [1] = 5.0 ¥
H), 1.06 (sextet, J = 7.4 Hz, 4H), 0.64 (t, 6H, J = 7.4 Hz),
13
.57-0.49 (m, 4H); NMR (CD
2
2
Cl , 125 MHz): d 152.46, 152.08,
-
7
6
-6
-7
1
1
0
0
M, [HSA] = 5.0 ¥ 10 M; 2/HSA: [2] = 5.0 ¥ 10 M, [HSA] = 1.0 ¥
47.79, 147.34, 138.94, 129.30, 127.43, 123.39, 123.21, 121.23,
19.88, 118.39, 55.74, 39.89, 26.04, 23.01, 13.62.
-
M. The fluorescence intensities are measured at emission maxima of
the HSA-bound fluorophores, 397 nm for 1 and 376 nm for 2.
A reaction mixture of 4 (1.30 g, 4.02 mmol) and concentrated
sulfuric acid (8.0 mL) was stirred at room temperature for 24 h and
then poured into cold water (30 mL) to give precipitates, which
were filtered, washed with water and dried. More products were
recovered from the filtrate by extraction with ethyl acetate (3 ¥
Conclusive information on the binding site of compound 1 could
be given by X-ray crystallographic analysis, but this is beyond the
scope of this work. It is clear that 1 does not bind to site II, which is
6
smaller than site I: otherwise it would be displaced by IBU which
2
5 mL). The residue from ethyl acetate layers and the precipitates
has the same negative charge. One possibility is that 1 binds to a
site other than sites I and II. A third high affinity site located in
domains other than IIA and IIIA was reported for fatty acids and
were combined and purified by column chromatography (eluent:
ethyl acetate and then ethyl acetate:methanol = 8:1) to give 5 as
◦
1
light yellow solid (1.50 g, 92%). 5: Mp > 200 C (dec); H NMR
6,9,10
other drugs.
Another possibility is that 1 binds to site I. This
(
DMSO-d
6
, 400 MHz): d 8.36 (d, J = 2.0 Hz, 1H), 8.27 (dd, J =
is supported by the weak displacement of 1 by aspirin. The cavity
8
1
1
.0, 2.0 Hz, 1H), 8.10 (d, J = 8.0 Hz, 1H), 7.98 (d, J = 8.0 Hz,
6,14
of site I is large and has three regions for ligand binding. Due
to their relatively small sizes, 1 and PBZ may cooperatively bind
to site I. The effect of IBU on the binding of 1 can be attributed to
cooperative allosteric interaction of the ligands. Such interaction
between ligands bound to sites I and II has been reported for other
H), 7.74 (s, 1H), 7.71 (d, J = 8.0 Hz, 1H), 2.20–2.00 (m, 4H),
.05 (sextet, J = 7.4 Hz, 4H), 0.63 (t, J = 7.4 Hz, 6H), 0.51–0.44
(
m, 4H).
Compound 5 (0.66 g, 1.64 mmol) was reacted with 1 atmo-
spheric hydrogen gas in the presence of 5% Pd/C (0.15 g) at room
temperature for 5 h. Filtration of the reaction mixture followed
by concentration of the filtrate afforded 6 (0.60 g, 98%). 6: Mp >
44
ligand pairs.
◦
1
2
50 C (dec); H NMR (DMSO-d , 400 MHz): d 7.51–7.41 (m,
6
Conclusions
4
H), 6.58 (d, J = 2.0 Hz, 1H), 6.53 (dd, J = 8.0, 2.0 Hz, 1H), 5.26
A
novel water-soluble fluorescent molecule, 9,9-dibutyl-7-
(broad s, 2H), 1.93–1.75 (m, 4H), 1.06 (sextet, J = 7.4 Hz, 4H),
0.65 (t, J = 7.4 Hz, 6H), 0.57–0.45 (m, 4H).
(
dimethylamino)-2-fluorenesulfonate 2 has been successfully pre-
pared from 2-nitrofluorene in moderate overall yield. The flu-
orescence spectrum of 2, as well as that of 7-(dimethylamino)-
To a mixture of 6 (0.366 g, 0.975 mmol) and paraformaldehyde
(0.360 g, 12.0 mmol) in glacial acetic acid (10 mL) was added
sodium cyanoborohydride (0.315 g, 5.0 mmol). After stirring at
room temperature under a nitrogen atmosphere for 18 h, the
reaction mixture was poured into ice water (10 mL) to obtain
precipitates. The precipitates were filtered, washed with water, and
put in hot water (10 mL) and then basified with 30% ammonia
solution to obtain a clear solution. The solution was acidified
to ca. pH 3 with 1 N HCl to produce precipitates, which were
filtered, washed with water and then vacuum-dried to give the
product 2 (97 mg). More products were recovered from the filtrate
by extraction with ethyl acetate (5 ¥ 15 mL). The residue from
the ethyl acetate layers was purified by reverse-phase medium
3
2
-fluorenesulfonate 1, is blue-shifted with decreasing polarity of
medium or with binding to human serum albumin (HSA). The ap-
6
-1
parent binding constants to HSA are 0.37 ¥ 10 M for 1 and 2.2 ¥
6
-1
1
0 M for 2. The primary binding site on HSA is site II in domain
IIIA for 2, whereas that of 1 is believed to be site I in domain IIA,
but involving a different region from the phenylbutazone binding
site. The aminofluorene-based fluorescent molecules, particularly
2, exhibit many desirable characteristics for sensitive fluorescent
probes, such as high molar absorptivity, strong fluorescence, and
a high sensitivity of the fluorescence to environmental changes.
This, together with the large binding constant and site specificity to
HSA will make them useful as fluorescent probes for proteins and
other biological molecules, potentially at one-order of magnitude
lower concentrations than the widely used aminonaphthalene-
based probes.
pressure liquid chromatography (flow rate 20 mL/min, CH CN–
1% TFA/H O–1% TFA gradient) to give the product (0.152 g).
2
The combined yield of 2 from 6 was 0.249 g (64%). Mp, 164–
◦
1
166 C; H NMR (CD OD, 500 MHz, residual solvent peak
3
This journal is © The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4225–4232 | 4231

View Article Online
d = 4.840, 3.309): d 7.96 (d, J = 8.0 Hz, 1H), 7.91 (s, 1H), 7.89–7.83
8 X. M. He and D. C. Carter, Nature, 1992, 358, 209.
9
G. Sudlow, D. J. Birkett and D. N. Wade, Mol. Pharmacol., 1975, 11,
(
m, 2H), 7.71 (s, 1H), 7.58 (d, J = 7.5 Hz, 1H), 3.36 (s, 6H), 2.17–
8
24; G. Sudlow, D. J. Birkett and D. N. Wade, Mol. Pharmacol., 1976,
2, 1052.
2
0
.05 (m, 4H), 1.07 (sextet, J = 7.4 Hz, 4H), 0.65 (t, J = 7.4 Hz, 6H),
1
1
3
.62–0.45 (m, 4H); C NMR (CD OD, 125 MHz, residual solvent
3
10 I. Sj o¨ holm, B. Ekman, A. Kober, I. Ljungstedt-P a˚ hlman, B. Seiving
peak d = 49.05): d 155.39, 152.38, 146.44, 143.68, 143.62, 142.48,
and T. Sj o¨ din, Mol. Pharmacol., 1979, 16, 767.
1
1
1 U. Kragh-Hansen, Mol. Pharmacol., 1988, 34, 160.
1
26.70, 123.17, 121.93, 121.51, 120.91, 116.64, 57.37, 47.72, 40.78,
7.24, 23.94, 14.19. Anal. Calcd for C23 O: C, 65.84;
S·H
2 F. P. Nicoletti, B. D. Howes, M. Fittipaldi, G. Fanali, M. Fasano, P.
Ascenzi and G. Smulevich, J. Am. Chem. Soc., 2008, 130, 11677.
2
H
31NO
3
2
H, 7.93; N, 3.34; S, 7.64. Found: C, 65.60; H, 7.76; N, 3.66; S,
13 A. A. Bhattacharya, T. Gr u¨ ne and S. Curry, J. Mol. Biol., 2000, 303,
21.
7
7
.54%.
1
1
1
1
1
4 K. Yamasaki, T. Maruyama, U. Kragh-Hansen and M. Otagiri,
Biochim. Biophys. Acta, 1996, 1295, 147.
5 L. A. Bagatolli, S. C. Kivatinitz, F. Aguilar, M. A. Soto, P. Sotomayer
and G. D. Fidelio, J. Fluoresc., 1996, 6, 33.
6 K. Takehara, K. Yuki, M. Shirasawa, S. Yamasaki and S. Yamada,
Anal. Sci., 2009, 25, 115.
7 G. Colmenarejo, A. Alvarez-Pedraglio and J.-L. Lavandera, J. Med.
Chem., 2001, 44, 4370.
The ammonium salt of 2 was obtained by adding 30% ammonia
solution to an aqueous suspension of 2, followed by evaporation
39
of the solvent.
Other materials and methods
HSA (essentially fatty acid free) and acrylodan (AC) were obtained
from Sigma and Molecular Probes, respectively. Other chemicals
were of reagent grade quality, obtained from commercial sources
and were used without further purification. The solvents used
for the spectroscopic measurements were spectroscopic grade.
The aqueous solutions were prepared with a phosphate buffer
8 M. C. Doody, A. M. Gotto, Jr. and L. C. Smith, Biochemistry, 1982,
21, 28.
19 L. A. Bagatolli, S. C. Kivatinitz and G. D. Fidelio, J. Pharm. Sci., 1996,
5, 1131; D. Shcharbin, B. Klajnert, V. Mazhul and M. Bryszewska,
J. Fluoresc., 2005, 15, 21; N. Seedher and S. Bhatia, Pharmacol. Res.,
8
2
006, 54, 77; B. K. Sahoo, K. S. Ghosh and S. Dasgupta, Biopolymers,
2009, 91, 108.
2
0 B. Ahmad, M. Z. Ahmed, S. K. Haq and R. H. Khan, Biochim. Biophys.
Acta, 2005, 1750, 93.
(
0.01 M phosphate, pH 7.0, I = 0.1 M with NaCl). HSA was
dissolved in the buffer and the concentration was calculated
2
2
1 J. Gonz a´ lez-Jim e´ nez and M. Cortijo, J. Protein Chem., 2002, 21, 75.
2 J. Gonz a´ lez-Jim e´ nez and M. Cortijo, Protein J., 2004, 23, 351.
-
1
-1 22,26
from absorbance data using e280 = 36,600 M cm .
The
concentrations of others were based on the weights. HSA/AC
conjugate was prepared by using Kamal’s method with minor
23 R. Narazaki, T. Maruyama and M. Otagiri, Biochim. Biophys. Acta,
1
997, 1338, 275.
2
4 F. Moreno, M. Cortijo and J. Gonz a´ lez-Jim e´ nez, Photochem. Photo-
26
-6
modifications: HSA and AC (5 ¥ 10 M each) in phosphate
buffer were reacted for 24 hrs at room temperature. Stock solutions
of 1, 2, AC, ANS, aspirin, ibuprofen and phenylbutazone were
prepared in methanol. To make a desired solution for spectral
measurement, a predetermined amount of a stock solution was
taken, methanol was evaporated off, and then an appropriate
solvent, buffer or HSA solution was added.
biol., 1999, 70, 695.
25 A. Buz a´ dy, J. Savolainen, J. Erosty a´ k, P. Myllyperki o¨ , B. Somogyi and
J. Korppi-Tommola, J. Phys. Chem. B, 2003, 107, 1208.
2
2
6 J. K. A. Kamal, L. Zhao and A. H. Zewail, Proc. Natl. Acad. Sci. USA,
2
004, 101, 13411.
7 F. Moreno, M. Cortijo and J. Gonz a´ lez-Jim e´ nez, Photochem. Photo-
biol., 1999, 69, 8.
2
2
3
8 S. S. Krishnakumar and D. Panda, Biochemistry, 2002, 41, 7443.
9 F. Moreno and J. Gonz a´ lez-Jim e´ nez, Chem-Biol. Inter., 1999, 121, 237.
0 A. Douhal, M. Sanz and L. Tormo, Proc. Natl. Acad. Sci. USA, 2005,
102, 18807.
Absorption spectra were taken with an Agilent 8453 spec-
trophotometer. Fluorescence spectra were obtained with a Hitachi
F4500 spectrofluorimeter. The solutions containing HSA were
equilibrated for at least 10 hrs, prior to spectral measurements.
31 D. E. Epps, T. J. Raub and F. J. K e´ zdy, Anal. Biochem., 1995, 227, 342.
3
2 K. Yamasaki, T. Maruyama, K. Yoshimoto, Y. Tsutsumi, R. Narazaki,
A. Fukuhara, U. Kragh-Hansen and M. Otagiri, Biochim. Biophys.
Acta, 1999, 1432, 313.
◦
All spectra were taken at room temperature, 22 ± 2 C.
3
3
3 C. Dufour and O. Dangles, Biochim. Biophys. Acta, 2005, 1721, 164.
4 J. A. Organero, C. Martin, B. Cohen and A. Douhal, Langmuir, 2008,
Acknowledgements
2
4, 10352.
3
5 V. Maes, Y. Engelborghs, J. Hoebeke, Y. Maras and A. Vercruysse, Mol.
Pharmacol., 1982, 21, 100.
This study was financially supported by research funds from
Chungnam National University in 2008.
36 S. Ercelen, A. S. Kylmchenko, Y. M e´ ly and A. P. Demchenko, Int. J. Bio.
Macromol., 2005, 35, 231.
3
7 S. Mazumdar, R. Manoharan and S. K. Dogra, J. Photochem.
Photobiol. A: Chem., 1989, 46, 301; V. Kartmakaran, T. Senyushkina,
G. Saroja, J. Liebscher and N. P. Emsting, J. Phys. Chem. A, 2007, 111,
10944.
References
1
2
3
4
5
6
7
C. Reichardt, Solvents and Solvent Effects in Organic Chemistry, Wiley,
New York, 3rd edn., 2003, pp. 329-358.
J. R. Lakowicz, Principles of Fluorescence Spectroscopy, Springer, New
York, 3rd edn., 2006.
T. Peters, Jr. All about Albumin. Biochemistry, Genetics and Medical
Applications, Academic Press, San Diego, 1996.
P. Ascenzi, A. Bocedi, S. Notari, G. Fanali, R. Fesce and M. Fasano,
Mini-Rev. Med. Chem., 2006, 6, 483.
38 R. Manoharan and S. K. Dogra, Can. J. Chem., 1987, 65, 2013; S. K.
Saha and S. K. Dogra, J. Mol. Struc., 1998, 470, 301.
39 K. K. Park, J. W. Park and A. D. Hamilton, J. Fluoresc., 2007, 17, 361.
40 S. K. Saha, S. Santra and S. K. Dogra, J. Mol. Struc., 1999, 478, 199.
41 C.-W. Lee, M.-P. Kung, C. Hou and H. F. Kung, Nuclear Med. Biol.,
2003, 30, 573.
42 J. W. Park, S. Y. Lee and S. M. Kim, J. Photochem. Photobiol. A: Chem.,
2005, 173, 271.
43 B. Bojko, A. Sulkowska, M. Maciazek, J. Rownicka, F. Njau and W. W.
Sulkowski, Int. J. Bio. Macromol., 2008, 42, 314.
44 I. Fitos, J. Visy, M. Simonyi and J. Hermansson, Chirality’s, 1999, 11,
115; J. Chen and D. S. Hage, Nature Biotechnol., 2004, 22, 1445.
J. Ghuman, P. A. Zunszain, I. Petitpas, A. A. Bhattacharya, M. Otagiri
and S. Curry, J. Mol. Biol., 2005, 353, 38.
J. R. Simard, P. A. Zunszain, J. A. Hamilton and S. Curry, J. Mol. Biol.,
2
006, 361, 336.
D. C. Carter and J. X. Ho, Adv. Protein Chem., 1994, 45, 153.
4
232 | Org. Biomol. Chem., 2009, 7, 4225–4232
This journal is © The Royal Society of Chemistry 2009