Spectroscopic study of a novel bis(heptamethine cyanine) dye and its interaction with human serum albumin

Article abstract of DOI:10.1366/0003702053945976

A newly synthesized near-infrared (NIR) bis(heptamethine cyanine) dye 7 was evaluated for its utility as a non-covalent label for proteins. This dye forms inter- and intramolecular H-aggregates in polar solvents, even at very low concentrations. The intra

Full text of DOI:10.1366/0003702053945976

Spectroscopic Study of a Novel Bis(heptamethine cyanine)
Dye and Its Interaction with Human Serum Albumin
GABOR PATONAY,* JUN SEOK KIM, RAVIKUMAR KODAGAHALLY, and
LUCJAN STREKOWSKI*
Department of Chemistry, University Plaza, Georgia State University, Atlanta, Georgia 30303
A newly synthesized near-infrared (NIR) bis(heptamethine cyanine)
dye 7 was evaluated for its utility as a non-covalent label for pro-
teins. This dye forms inter- and intramolecular H-aggregates in po-
lar solvents, even at very low concentrations. The intramolecular
dimeric form of the dye can be described as a clam-shell complex
with two interacting hydrophobic carbocyanine moieties. In this in-
tramolecular H-aggregate, the chromophore has a low extinction
coefficient and low fluorescence quantum yield. In aqueous solution,
in the absence of human serum albumin (HSA), dye 7 has charac-
teristic absorption bands at 792 and 435 nm, and its fluorescent
emission is significantly diminished in comparison to that in meth-
anol or when compared to its monomeric equivalent 5. Dye 7 seems
to be more advantageous than its monomeric counterpart 5 as a
non-covalent label for biomolecules. Upon addition of HSA, the H
and D bands are decreased and the monomeric band is increased,
with concomitant increase in fluorescence intensity, suggesting that
clam-shell H-aggregates open up in the complex with HSA. The
binding stoichiometry is 1:1. The main advantage of this dimeric
dye as a non-covalent label is that the free dye has negligible fluo-
rescence.
centrations or in organic solvents at higher concentra-
tions. The conformation of the aggregates is dependent
on several factors such as dye structure, concentration,
10
solvent polarity, pH, ionic strength, and temperature.
Carbocyanines have recently received significant atten-
tion as noncovalent labels for the detection of protein and
other biomolecules. Protein labeling with cyanine dyes is
based on the relatively low fluorescence of H-aggregates
in aqueous solutions. The fluorescence intensity of the
dyes is enhanced in the presence of protein or other large
biomolecules as the dye interaction with biomolecules
breaks up the aggregates. Examples of such applications
have been demonstrated by using human serum albumin
(HSA), transport blood plasma protein, and DNA, among
others. The main spectroscopic feature of this non-co-
valent interaction is the increase in the fluorescence quan-
1
1
tum yield (⌽), facilitating the detection of biomolecules.
Several non-covalent interactions have been studied in
1
2,13,20,21
Index Headings: Near infrared; NIR; Bis(heptamethine cyanine)
the past.
The main driving factor in studying non-
dye; H-aggregates; Human serum albumin; HSA; Stoichiometry.
covalent interactions is to replace the labor-intensive con-
1
2
ventional covalent label methods. In addition to the rel-
atively complicated chemistry of covalent labeling, often
there is either no proper functionality available for co-
valent labeling on the biomolecule or covalent labeling
may influence the proper function of the biomolecule. By
contrast, non-covalent labels can simply be mixed with
the biomolecule with no additional preparation steps.
However, there is difficulty in determining a non-covalent
label’s physical mechanisms, such as hydrophobic inter-
actions, electrostatic interactions, and hydrogen bonding.
All these factors are important during the complex for-
mation with the biomolecule. One of the most important
variables that must be measured during this binding pro-
cess is the binding constant and the binding stoichiome-
try. The dye–protein complex can be characterized by
simply measuring the emission of the complex relative to
INTRODUCTION
Dye aggregates have been paid considerable attention
because they play an important role in several areas such
as laser technologies, photodynamic therapy of cancer,
1
2,3
4
energy transfer in photosynthesis, and light-harvesting
5
,6
arrays in artificial photosynthetic systems. Cyanine
dyes generally have two nitrogen heterocyclic systems
conjugated through a polymethine bridge. Delocalization
of positive charge generates strong intermolecular van
der Waals interactions that promote dye aggregation or
self-association. This aggregation results in formation of
different structures depending on how the carbocyanine
moieties interact with each other and is usually described
as H-aggregation (hypsochromic) and J-aggregation
1
3
7
that of free dye.
(
bathochromic) depending on the angle of slippage ␣.
Fluorescent dyes have been used to study biomolecules
due to decreased background noise relative to absorbance
Representative models of dye aggregates in solution are
shown in Fig. 1. H-aggregates are referred to as ‘‘card-
pack’’ arrangement, whereas J-aggregates are known as
1
4
and improved sensitivity. Autofluorescence from bioan-
alytical and other complex samples in the visible region
can be a problem, but this background fluorescence is
significantly decreased in the NIR range. Thus, the ad-
‘
‘brickwork’’ arrangement. Large molecular slippage (
Ͻ ϳ32 ) results in a bathochromic shift, while small slip-
page (␣ Ͼ ϳ32
␣
Њ
8
,9
Њ
) results in a hypsochromic shift. Con-
1
5–17
vantages of NIR fluorophores have been recognized.
sequently, the characteristic absorption spectrum of the
aggregate is different from that of the monomeric state,
i.e., di-, tri-, tetramers, and so forth have different elec-
tronic states. It is well known that carbocyanine dyes can
easily aggregate in aqueous solutions even at low con-
In the NIR region (670–1100 nm), very few biomole-
cules possess intrinsic fluorescence. Generic NIR dyes
have considerable advantageous photophysical proper-
ties, with high molar absorptivities (
␧
, 150000–250 000)
and fluorescence quantum yields (
⌽
, 0.05–0.5). More-
over, quantum yields of dyes typically increase upon im-
mobilization with proteins or other biomolecules.
Received 15 November 2004; accepted 2 February 2005.
Authors to whom correspondence should be sent.
1
8,19
*
0
᭧
003-7028/05/5905-0682$2.00 /0
2005 Society for Applied Spectroscopy
682
Volume 59, Number 5, 2005
APPLIED SPECTROSCOPY

state at any concentration. As part of these studies, a con-
formationally flexible dimeric dye 7 (Scheme 1) has been
synthesized. The dimeric molecule exists in an intramo-
lecular aggregate when free in solution, and the aggregate
is characterized by low fluorescence. The aim of this
work is to study the comprehensive photophysical spec-
tral properties of a newly synthesized dimeric heptame-
thine cyanine dye 7 in the absence and presence of HSA
using absorption and fluorescence spectroscopy.
The dimeric dye 7 contains two cyanine units joined
by a decamethylene chain. The choice of this dimer for
synthesis was made after inspecting molecular models of
several dye candidates with internal bridges ranging from
four to ten carbon atoms. The analysis strongly suggested
that unfavorable steric interactions within the polymeth-
ylene bridge were the lowest for the molecule of 7 in a
number of conformations mimicking intramolecular ag-
gregation.
F
IG. 1. Schematic representations of cyanine dye aggregates in solu-
tion: Dye aggregates with (a) brickwork, (b) ladder, and (c) staircase
molecular arrays;
after Ref. 9.
␣ is the angle of slippage. Aggregate models drawn
Consequently, NIR dyes are well suited for bioanalytical
applications. Reported analytical applications for NIR
dyes include enzymatic assays, hydrophobicity, metal ion
and pH studies, immunoassays, drug displacement, and
EXPERIMENTAL
1
2,16–21
DNA sequencing applications.
Synthesis of Cyanine 5 and its Dimeric Analog 7.
The synthesis is depicted in Scheme 1. Briefly, in the
preparation of 5, indolenine 1 was quaternized by the
reaction with butyl iodide and the resultant indolium de-
rivative 2 was allowed to react with Vilsmeier–Haack
reagent 3 (0.5 equiv). The use of 1 equiv of 3 for con-
densation with 2 produced compound 4 as a major prod-
uct. The condensation of 4 with bis-indolium salt 6, ob-
tained by quaternization of 1 with 1,10-dibromodecane,
The major problem with non-covalent interactions is
that the free dye that is not bound to the biomolecule
exhibits significant fluorescence. Although NIR dyes tend
to form non-fluorescent aggregates when in aqueous so-
lutions, aggregate formation is not significant at low con-
centrations. The use of higher concentrations is imprac-
tical in most non-covalent labeling situations. In the ideal
case, the free dye would be in a non-fluorescent aggregate
Scheme 1. Synthesis of cyanine 5 and its dimeric analog 7.
APPLIED SPECTROSCOPY
683

furnished the desired dimeric dye 7. Detailed procedures
for 5 and 7 are given below. All reactions were conducted
under a nitrogen atmosphere.
sodium phosphate dibasic (monohydrate) were purchased
from Fisher Scientific (Fair Lawn, NJ). Water was Nan-
opure grade (Barnstead model D4751 ultrapure water
system). Methanol was obtained from the Aldrich Chem-
ical Company (Milwaukee, WI) in HPLC grade.
Instrumentation. Absorption measurements were ac-
quired on a Perkin-Elmer Lambda UV/VIS/NIR (Lambda
50) spectrophotometer (Norwalk, CT). Fluorescence
emission spectra were taken with a K2 spectrofluorometer
(ISS, Champaign, IL) equipped with an R928 Hamama-
tsu photomultiplier tube (Bridgewater, NJ). Commercial
GaAlAs laser diodes (Laser Max, Rochester, NY) were
used as the excitation source at 690 nm. The spectral
bandpass was 16 nm and the integration time was 3 s.
All absorption and fluorescence measurements were tak-
en in a 1 cm cuvette.
2
-[7
chloro-3
-yl]-1-butyl-3,3-dimethyl-3H-indol-1-ium iodide (5). A
؅
-(1
؆
-Butyl-3
؆
,3
؆
-dimethylindolin-2
؆
-ylidene)-4
؅-
؅
,5
؅
-(propane-1
ٟ
,3 -diyl)-1 ,3 ,5
ٟ
؅ ؅ ؅-heptatrien-
1
؅
solution of 2,3,3-trimethylindolenine (1, 0.48 g, 3 mmol)
and butyl iodide (1.7 mL, 15 mmol) in anhydrous ace-
tonitrile (10 mL) was heated under reflux for 15 h. The
resultant precipitate of 1-butyl-2,3,3-trimethyl-3H-indol-
ium iodide (2) was crystallized from methanol; yield 0.98
g (95%); mp
Ͼ 250 ЊC (dec.). A solution of 2 (0.34 g, 1
2
2
mmol), Vilsmeier–Haack reagent 3, (0.18 g, 0.5 mmol),
and anhydrous sodium acetate (0.25 g, 3 mmol) in an-
hydrous ethanol (25 mL) was heated under reflux for 4
h. After cooling, the mixture was filtered, and the solution
containing 5 was concentrated under reduced pressure.
Dye 5 was crystallized by the addition of ether to a so-
Methods. Stock solutions of dye 5 (2 mM) and dye 7
(1 mM) in methanol were stored in the dark at 4 ЊC when
lution in methanol; yield 0.63 g (90%); mp
Ͼ
150
Њ
C
7
not in use. The 0.1 mM stock solution of HSA was pre-
pared in 20 mM dibasic phosphate buffer solution and
adjusted to pH 7.2 with phosphoric acid. Stock solution
of HSA was prepared fresh daily.
The dye–HSA mixture was vortexed for 30 s to allow
for equilibration. Various dye–HSA ratios were investi-
gated to determine the utility of dye 7 as a noncovalent
label for serum albumin. For all non-covalent labeling,
the stock solution of the dimeric cyanine in methanol was
diluted with phosphate buffer because the dye has rela-
tively poor solubility in aqueous solutions. All working
solutions contained only 1% (v/v) methanol to facilitate
dye dissolution while avoiding denaturation of HSA. All
measurements were performed at room temperature.
1
(
dec.); H NMR (DMSO-d , 300 MHz)
␦
0.87 (t, J ϭ
6
Hz, 6H), 1.44 (m, 4H), 1.90 (m, 2H), 2.74 (m, 4H), 4.24
m, 4H), 6.38 (d, J 14 Hz, 2H), 7.31 (m, 2H), 7.46 (m,
H), 7.64 (d, J 8 Hz, 2H), 8.29 (d, J 14 Hz, 2H);
TOF MS (9.64 eV) m/z 567 (M –I , for Cl), 569 (M –
(
4
ϭ
ϭ
ϭ
ϩ
Ϫ
35
ϩ
Ϫ
37
I , for Cl). Analysis. Calcd. For C H ClIN : C, 65.66;
3
8
48
2
H, 6.96; N, 4.03. Found: C, 65.50; H, 6.93; N, 3.94.
Decamethylene[bis[2-[7 -(1 -butyl-3 ,3 -dimethylin-
dolin-2 ylidene)-4 -chloro-3 ,5 -(propane-1 ,3 -diyl)-
,3 ,5 -heptatrien-1 -yl]-3,3-dimethyl-3H-indol-1-ium]
؅
؆
؆ ؆
؆
؅
؅
؅
ٟ ٟ
1
؅
؅
؅
؅
2
2
diiodide (7). A solution of 2 (0.34 g, 1 mmol), 3, (0.36
g, 1 mmol), and anhydrous sodium acetate (0.17 g, 2
mmol) in anhydrous ethanol (25 mL) was heated to 50
ЊC for 2 h. After concentration under reduced pressure
the resultant mixture of 4 and 5 was separated by chro-
RESULTS AND DISCUSSION
matography on silica gel eluting with ethyl acetate/hex-
MeOH
anes (1:4). Compound 4: yield 0.34 g (60%);
␭
ϭ
Photophysical Properties of Dimeric Dye 7. In this
study, photophysical properties of a novel dimeric dye 7
in organic solvent and aqueous buffer were characterized.
All spectra were normalized for easy comparison. In gen-
eral, monomeric carbocyanines form aggregates with in-
max
4
78 nm (blue in solution); TOF MS (9.64 eV) m/z 445
ϩ Ϫ 35 ϩ Ϫ
37
(M –I , for Cl), 447 (M –I , for Cl).
In the synthesis of decamethylene[bis(2,3,3-trimethyl-
H-indolium)] dibromide (6), a solution of 1 (0.48 g, 3
3
3
–5
mmol) and 1,10-dibromodecane (0.3 g, 1 mmol) in xy-
lenes (5 mL) was heated under reflux for 10 h. Product
creasing dye concentration even in organic solvent.
Figure 2 shows the normalized spectra of various con-
centrations of dimeric dye 7 in methanol. No significant
spectral changes can be observed. It is clear that the spec-
tra of 7 are less dependent on dye concentration in meth-
anol compared to those of monomeric carbocyanine dyes.
The spectra of monomeric dye 5 were also obtained in
phosphate buffer. When comparing normalized spectra of
6
was isolated by silica gel chromatography eluting with
methanol/ethyl acetate (1:5); yield 0.19 g (30%); mp
1
ϩ
Ϫ
66–168
ЊC; TOF MS m/z 458 (M –2Br ).
The dimeric dye 7 was synthesized by heating under
reflux for 24 h a solution of 4 (0.29 g, 0.5 mmol), 6 (0.15
g, 0.25 mmol), and anhydrous sodium acetate (0.08 g, 1
mmol) in anhydrous ethanol (20 mL). After standard
workup, product 7 was isolated by silica gel chromatog-
the monomeric dye 5 (10 ␮M) and the dimer 7 (10 ␮M)
in aqueous buffer solution (Fig. 3), two significant dif-
ferences can be observed at 435 nm and 710 nm. Ac-
cording to the literature, these features are characteristic
of H aggregation. As can be seen, the H bands for mo-
nomeric dye 5 are much weaker than those for 7. It is
likely that both intra and intermolecular aggregation take
place in the case of dimeric dye 7.
Dye–HSA Interaction. Absorption Spectra. The ab-
sorption spectrum of the dimeric dye 7 in methanol has
a characteristic band at 781 nm. This band intensity is
diminished in phosphate buffer (Fig. 4). The absorption
maximum of the dye in phosphate buffer is slightly red
raphy eluting with ethyl acetate/dichloromethane (1:4);
1
yield 0.06 g (17%); H NMR (DMSO-d , 300 MHz)
␦
6
0
1
6
7
1
.94 (t, J ϭ 7 Hz, 6H), 1.0–1.6 (m, 24H), 1.68 (s, 24H),
.85 (m, 4H), 2.70 (m, 8H), 4.20 (m, 8H), 6.31 (m, 4H),
.60 (m, 2H), 6.68 (m, 2H), 7.04 (m, 2H), 7.12 (m, 2H),
.32 (m, 2H), 7.45 (m, 4H), 7.65 (m, 2H), 8.27 (d, J
ϭ
ϩ
Ϫ
4 Hz, 4H); MALDI MS m/z 1160.9 (M –2I , for 2
3
5
ϩ
Ϫ
37
ϩ
Ϫ
Cl), 1162.9 (M –2I , for 35Cl Cl), 1164.9 (M –2I ,
3
7
for 2 Cl). Analysis Calcd. For C H Cl I N : C, 66.14;
7
8
98
2 2
2
H, 6.97; N, 3.96. Found: C, 66.16; H, 6.84; N, 3.92.
Other Materials and Solvents. Fatty acid free HSA
(Ն
96% purity) was obtained from Sigma (St. Louis,
shifted (␭max
ϭ 792 nm) compared to its maximum in
MO). Sodium phosphate monobasic (monohydrate) and
methanol. It is likely that this dye forms mostly intra-
684
Volume 59, Number 5, 2005

FIG. 2. Concentration effect on aggregation of dimeric dye 7 in methanol. Normalized absorption spectra were taken at three different concentra-
tions: 0.1 M (dotted line), 1 M (dashed line), and 10 M (solid line).
␮
␮
␮
molecular aggregation in aqueous solution, but some in-
termolecular aggregation is also likely. In the presence of
HSA the spectra become similar to that in methanol (␭max
in the presence and absence of HSA are given in Fig. 5.
Two characteristic bands (␭max at 792 nm and 435 nm)
can be distinguished in the absorption spectra in buffer
solution compared to dye solution in methanol at the
same concentration. The long wavelength absorption
ϭ
787 nm). This shift is most likely caused by the de-
crease in intramolecular van der Waals interactions as the
clam-shell opens up. Similar shifts have been observed
for monomeric carbocyanine dye aggregates upon the
formation of the dye–HSA complex.
band (␭max
ϭ 792 nm) is S0–S1 transition of the strong
clam-shell intramolecular dimer dye and weak intermo-
lecular aggregates in aqueous buffer solution. The spec-
trum is red-shifted about 11 nm in comparison to the
Representative absorption spectra of the dimeric dye 7
FIG. 3. Normalized absorption spectra of monomeric dye 5 (dotted line) and dimeric dye 7 (solid line) in phosphate buffer at a dye concentration
of 10 M.
␮
APPLIED SPECTROSCOPY
685

FIG. 4. Absorption spectra of dye 7 in (top) methanol and (bottom) methanol–phosphate buffer mixture at room temperature.
similar transition in methanol due to intramolecular in-
teractions between the two carbocyanine moieties and in-
termolecular interaction between the two or more dimeric
molecules. Upon addition of HSA the inter- and intra-
molecular aggregates break up and concomitant changes
in the H, D, and M bands are observed. The absorption
spectra of the dimeric dye at constant concentration of
depicted in Fig. 5. Upon increase in the HSA concentra-
tion, up to CHSA 50 M, the characteristic features of
the dimeric dye absorption spectra change sharply. The
M-band relative intensity increases strongly, while the D-
band and H-band dramatically decrease in intensity. Up
ϭ
␮
to CHSA
in the H-band intensity. However, at CHSA
is negligible change in the intensities of H-, D-, and M-
ϭ
50
␮
M, a continuous decrease can be observed
ϭ
70 M, there
␮
Ϫ5
1
0
M with different amounts of HSA added are also
FIG. 5. Absorption of dye 7 (10
␮M) in the presence of HSA: (a) 100, (b) 70, (c) 50, (d) 30, (e) 10, (f) 5, (g) 1, (h) 0.5, (i) 0.1, and (j) 0 ␮M.
The arrows indicate the spectral changes of the H-, D-, and M-bands.
686
Volume 59, Number 5, 2005

FIG. 6. Fluorescence emission spectra of monomeric dye 5 (dotted line) and dimeric dye 7 (solid line) in phosphate buffer at a dye concentration
of 10 M.
␮
bands and absorption of HSA becomes more pronounced.
Unfortunately, the overlap of the HSA absorption in this
range prevents the observation of additional changes.
Clearly, as the HSA concentration increases, the concen-
tration of the inter- and intramolecular H-aggregates de-
creases and the fraction of the dimeric dye molecules
bound to HSA in an open form increases. This indicates
that the interaction between the HSA and the open form
dimeric cyanine dye is stronger than the intramolecular
interaction between the two carbocyanine moieties in a
dye free in solution. This is in good agreement with sim-
ilar studies utilizing monomeric cyanines and HSA.
Clearly, binding of 7 to HSA is more favorable than
forming an intramolecular H-aggregate. It can be ex-
pected that intermolecular interactions between bis-car-
bocyanine molecules are even weaker, especially at low
concentrations.
Fluorescence Spectra. The novel bis-cyanine 7 was
found to interact with HSA, resulting in significantly en-
hanced fluorescence. The fluorescence emission of the
dye is very weak in aqueous solutions relative to pure
methanol. This is very advantageous for the use of bis-
cyanine 7 as a non-covalent label because the unbound
form of the dye has much lower quantum yield (Fig. 6).
It should be noted that one of the major drawbacks of
using monomeric carbocyanines as non-covalent labels is
the existence of a relatively strong fluorescence of the
unbound free form of the non-covalent label.
In free dye solutions, a fluorescence band at ␭max
10 nm becomes clearly apparent at the dye concentra-
tion of 10 M in methanol. The emission band at ␭max
794 nm can be interpreted as a band of inherent dimeric
ϭ
8
␮
ϭ
ϳ
dye fluorescence in buffer solution that is blue-shifted
over 16 nm in comparison to the emission in methanol.
This indicates that the bis-cyanine dye when bound to
HSA is in an environment that is more hydrophobic than
aqueous buffer but less hydrophobic than methanol (Fig.
7
). The data of Fig. 7 also indicate that the microhydro-
phobicity of the binding site is about equivalent to that
of the hydrophobicity of 50% methanol in water.
The emission spectra of the bis-cyanine dye at constant
concentration of 10
added are given in Fig. 8. Upon an increase in the HSA
concentration, up to 70 M, a strong fluorescence en-
hancement can be observed. These results can be ex-
plained in terms of a decrease in concentration of weakly
fluorescent intra- and intermolecular dye aggregates free
in solution as the concentration of a highly fluorescent
␮M with different amounts of HSA
␮
F
IG. 7. Fluorescence emission spectra of dye 7 (10
solvent systems, at room temperature. (a) 25% (v/v) methanol in water;
b) 25% (v/v) methanol in PB; (c) 50% (v/v) methanol in water; (d) 1:
ratio, 20 M total conc., of dye 7-HSA complex in 1% (v/v) methanol
in PB; (e) 50% (v/v) methanol in PB; and (f) 100% methanol.
␮M) in various
(
1
␮
APPLIED SPECTROSCOPY
687

FIG. 8. Fluorescence emission spectra of dye 7 (10 ␮M) in the presence of HSA: (a) 70, (b) 50, (c) 30, (d) 10, (e) 1, and (f) 0.1 ␮M.
complex of the open linear form of the dye with HSA
increases. The fluorescence emission spectra clearly show
a difference between the fluorescent intensity of the dye
in the absence (Fig. 7) and presence (Fig. 8) of HSA.
Dye–HSA Saturation and Stoichiometry. The satu-
rable binding curve for the complex of dye 7 with HSA
is shown in Fig. 9. The fluorescence intensity change at
This plot indicates that a strong binding exists between
the dimeric dye and HSA.
The predominant stoichiometry of the dye–HSA com-
plex was determined using Job’s method (Fig. 10). Fluo-
rescence versus mol fraction of HSA was determined
Ϫ
5
with a constant total concentration of 2.0
ϫ 10 M. The
Job’s plot indicates a predominant stoichiometry of 1:1.
Absorbance spectra are less suitable to obtain this stoi-
chiometry due to a severe overlap of the spectra of the
free and bound species present in the mixture. However,
the practically nonexistent fluorescence of the clam-shell
form allows this calculation from the fluorescence data.
8
00 nm is plotted at a fixed concentration of the dye. The
HSA concentrations of over 100 M cause no further
␮
increase in fluorescence, indicating that the dye is fully
complexed with HSA and, as a result, the addition of
HSA has no effect on the fluorescent intensity measured.
FIG. 9. Saturable binding curve as a function of increasing HSA concentration. The concentration of dye 7 was constant at 10 ␮M, and the
fluorescence was measured at 800 nm.
688
Volume 59, Number 5, 2005

IG. 10. Job’s plot to determine the stoichiometry of the dye–HSA complex at 800 nm. The total concentration of the complex was 2 ϫ
10 5 M.
Ϫ
F
Association Constant. The maximum absorption
CONCLUSION
wavelength of the free dye 7 is not different enough from
that of the bound dye to reliably calculate the association
constant. In addition, there are no isosbestic points be-
tween free and bound dye. Nevertheless, it was possible
to determine the association constant utilizing the quan-
tum yield change of the complex because, as shown in
Fig. 9, the fluorescence upon the addition of HSA in-
creases and free dye has low fluorescence. The associa-
In polar solvents, the novel bis-cyanine dye 7 is self-
associated to form an intramolecular H-aggregate due to
strong intramolecular van der Waals interaction between
the carbocyanine moieties. H-aggregates are generally
weakly fluorescent and often have lower extinction co-
efficients compared to their monomeric counterparts.
Deaggregation of dye aggregates in the presence of HSA
or in hydrophobic solvents results in a greatly increased
quantum yield and the appearance of the monomeric ab-
sorption band. In the case of 7, the intramolecular clam-
shell like dimer opens up and the concentration of the
clam-shell aggregate formed by the intramolecular inter-
action decreases with concomitant changes in the spectral
features. According to the data shown in Fig. 5, addition
of HSA leads to opening of the intramolecular H-aggre-
gate. It can be suggested that this novel bis-heptamethine
cyanine dye associates with hydrophobic cavities of
HSA. Alternatively, it may interact with charged regions
on the HSA surface since the charge of HSA (pI 4.8)
under pH 7.2 is negative, down to 4.8.
The newly synthesized NIR dye 7 has considerable
advantages over the commonly used visible dyes. First,
spectral interferences generated by biomolecules, which
generally have autofluorescence in the UV and visible
regions, are virtually eliminated with 7. Second, its strong
intramolecular H-aggregation results in low fluorescent
signal in aqueous solution. Importantly, this feature is
concentration independent. In summary, the novel NIR
dimeric heptamethine cyanine dye 7 should be of interest
for the bioanalytical community.
tion constant K was determined by using a method sim-
a
2
3
ilar to that reported previously by Tarazi, et al. Accord-
ing to this method, the interaction between dye D and
protein P can be represented by Eq. 1:
D
ϩ
P
⇔
DP
(1)
The association constant for the dye–protein complex for-
mation can be expressed by Eq. 2:
[
DP]
Ka
ϭ
(2)
[D][P]
where [DP], [D], and [P] are molar concentrations at
equilibrium. Subsequently, the association constant K of
a
the dye–protein complex was calculated from the data for
the fluorimetric titration of the dye with HSA by using
2
4
Eq. 3, where k is a constant dependent upon the instru-
mentation and quantum efficiency of the process and is
the measured fluorescence emission intensity of the dye–
protein mixture.
1
1
1
1
ϭ
ϩ
(3)
΂
΃
F
k[D]
k[D]K [P]
a
Therefore, a plot of 1/F, inverse fluorescence, versus
/[P] produces a straight line, with K equal to the inter-
1
ACKNOWLEDGMENT
a
cept divided by the slope of the line. The association
We thank the donors of the Petroleum Research Fund, administered
by the American Chemical Society, for support of this research.
5
Ϫ1
constant K of 3
ϫ 10 M was obtained. However, it
a
should be noted that Eq. 3 is valuable only for dye–pro-
tein complexes formed with a 1:1 stoichiometry, which
appears to be the prevalent stoichiometry in this case.
1. M. Maeda, Lasers Dyes (Academic Press, Tokyo, 1984).
APPLIED SPECTROSCOPY
689

2
. T. J. Dougherty, J. E. Kaufman, A. Goldfarb, K. R. Weishaupt, D.
G. Boyle, and M. Mittelman, Cancer Res. 38, 2628 (1978).
. T. J. Dougherty and R. E. Thoma, ‘‘Lasers’’, in Photomedicine and
Photobiology, R. Pratesi and C. A. Sachi, Eds. (Springer-Verlag,
Berlin, 1980).
14. J. P. Landers, Handbook of Capillary Electrophoresis (CRC Press,
Boca Raton, FL, 1997), 2nd ed., p. 379.
15. G. Patonay, M. D. Antoine, and A. E. Boyer, Proc. SPIE-Int. Soc.
Opt. Eng. 1435, 52 (1991).
16. M. B. Baars and G. Patonay, Anal. Chem. 71, 667 (1999).
17. B. L. Legendre, Jr., D. L. Moberg, D. C. Williams, and S. A. Soper,
J. Chromatogr., A 779, 185 (1997).
3
4
. Bioenergetics of Photosynthesis, Govindjee, Ed. (Academic Press,
New York, 1975).
5
. M. Minuro, T. Nozawa, N. Tamai, K. Shimada, I. Yamazaki, S. Lin,
R. S. Knox, B. P. Wittmershaus, D. C. Brune, and R. E. Blanken-
ship, J. Phys Chem. 93, 7503 (1989).
18. D. Andrews-Wilberforce and G. Patonay, Spectrochim. Acta, Part
A 46, 1153 (1990).
19. M. D. Antoine, S. Devanathan, and G. Patonay, Spectrochim. Acta,
Part A 47, 501 (1991).
20. J. Sowell, J. C. Mason, L. Strekowski, and G. Patonay, Eectropho-
resis 22, 2512 (2001).
6
7
8
. Y. Wang, Chem. Phys. Lett. 126, 209 (1986).
. E. E. Jelley, Nature (London) 139, 631 (1937).
. W. J. Harrison, D. L. Mateer, and G. J. Tiddy, Jr., J. Phys. Chem.
1
00, 2310 (1996).
. V. Czikkely, H. D. Forsterling, and H. Kuhn, Chem. Phys. Lett. 6,
07 (1970).
21. J. Sowell, R. Parihar, and G. Patonay, J. Chromatogr., B 757, 1
(2001).
22. S. M. Makin, L. I. Boiko, and O. A. Shavrggina, Zh. Org. Khim.
13, 1189 (1977).
23. L. Tarazi, N. Narayanan, and G. Patonay, Microchem. J. 64, 247
(2000).
24. G. A. Casay, D. B. Shealy, and G. Patonay, Fluorescence Spec-
troscopy, J. Lakowicz, Ed. (Plenum Press, New York, 1994).
9
2
1
1
1
1
0. A. H. Herz, Adv. Colloid. Interface Sci. 8, 237 (1977).
1. A. S. Tatikolov and S. M. B. Costa, Biophys. Chem. 107, 33 (2004).
2. C. Colyer, Cell. Biochem. Biophys. 33, 323 (2000).
3. F. Welder, B. Paul, H. Nakazumi, S. Yagi, and C. Colyer, J. Chro-
matogr., B 793, 93 (2003).
690
Volume 59, Number 5, 2005