Dual luminescent charge transfer probe for quantitative detection of serum albumin in aqueous samples

Article abstract of DOI:10.1016/j.saa.2020.118305

In diagnostic medicine serum albumin is considered as an important biomarker for assessment of cardiovascular functions and diagnosis of renal diseases. Herein, we report a novel donor-π-π-acceptor fluorophore for selective detection of serum albumin in u

Full text of DOI:10.1016/j.saa.2020.118305

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 235 (2020) 118305
Contents lists available at ScienceDirect
Spectrochimica Acta Part A: Molecular and Biomolecular
Spectroscopy
journal homepage: www.elsevier.com/locate/saa
Dual luminescent charge transfer probe for quantitative detection of
serum albumin in aqueous samples
⁎
Rajib Choudhury , Benjamin Quattlebaum, Charles Conkin, Siddhi Rajeshbhai Patel, Kallie Mendenhall
Department of Physical Sciences, Arkansas Tech University, Russellville, AR 72801, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 6 January 2020
Received in revised form 14 March 2020
Accepted 24 March 2020
Available online 24 March 2020
In diagnostic medicine serum albumin is considered as an important biomarker for assessment of cardiovascular
functions and diagnosis of renal diseases. Herein, we report a novel donor-π-π-acceptor fluorophore for selective
detection of serum albumin in urine samples. In our design, a phenolic donor was conjugated with a
tricyanofuran (TCF) acceptor through
a dimethine bridge via a simple condensation reaction. The
stereoelectronic effects of the incorporated methoxy (-OCH3) groups and the TCF moiety—in conjunction with
the extended π-electron conjugation—led to dual red and NIR-I absorption/emission in water. Moreover, due
to superior electron transfer between a phenolate donor and the TCF acceptor and the subsequent energy
decay from the charge transfer states, the fluorophore displayed negligible fluorescence emission in water and
other polar solvents. Consequently, we have been able to utilize the fluorophore for quantitative estimation of
serum albumin both in the red (b700 nm) and NIR-I (700–900 nm) regions of the electromagnetic spectrum
with excellent reproducibility. The fluorophore selectively recognized human serum albumin over other proteins
and enzymes with a limit of detection of 10 mg/L and 20 mg/L in simulated urine samples at red and NIR-I emis-
sion window of the spectrum, respectively. By molecular docking analysis and experimental displacement assays,
we have shown that the selective response of the fluorophore toward human serum albumin is due to tighter su-
pramolecular complexation between the fluorophore and the protein at subdomain IB, and the origin of the NIR-I
(780 nm) emission was attributed to a twisted conformer of phenolate-π-π-TCF system in aqueous solution.
These findings indicate that the fluorophore could be utilized for quantitative detection of human serum albumin
in urine samples for clinical diagnosis of albuminuria.
Keywords:
Dual emission probe
Selective protein recognition
Near infrared emission
Human serum albumin
Microalbuminuria
© 2020 Elsevier B.V. All rights reserved.
1. Introduction
significant importance for early diagnosis of renal damage and cardio-
vascular disease.
Serum albumin is a major protein component of blood plasma [1–3].
It maintains osmotic blood pressure and facilitates transport of drugs,
fatty acids, and other metabolites through blood stream [2]. Normally,
in healthy individuals serum albumin is excreted with urine below
20 mg/L. [2,4] However, this amount might increase due to complicated
health conditions. For instance, HSA level between 20 and 200 mg/L in
urine—a medical term is known as microalbuminuria—has been linked
to the risk of cardiovascular disease and early stages of kidney damage
[2,5,6]. Amount N200 mg/L is diagnosed as macroalbuminuria—a condi-
tion which indicates advanced stages of kidney damage [2,5,6]. In other
words, an elevated level of albumin in urine could indicate that the kid-
neys are no longer functioning properly. Therefore, quantitative detec-
tion of serum albumin in biological samples, such as in urine, is of
Quantitative estimation of serum albumin by bio-analytical tech-
niques has been reported in literature [7–14]. Some common examples
are kinetic method, immunoelectrophoresis, spectrophotometric tech-
nique, immunoturbidimetry, enzyme-linked immunosorbent assay
(ELISA), and radioimmunoassay [7–14]. Though some of these tech-
niques are currently in use, they are not free from intrinsic drawbacks,
such as, poor sensitivity and selectivity, extensive sample preparation
method, requirement for other biomolecules and enzymes for assay
preparation, and high cost associated with the specialized reagents
and apparatus [14,15]. Therefore, fluorescence based detection and
monitoring of serum albumin has attracted much attentions recently,
mainly because of high sensitivity and simple operation methodology
associated with a fluorimetric technique [16–22]. However, successful
application of fluorescence based technique is limited due to lack of
suitable fluorophores—a suitable fluorophore interacts and fluoresces
only in presence of a target analyte. Recently, a handful of small mole-
cule fluorophores have been reported for quantitative estimation of
⁎
Corresponding author.
E-mail address: rchoudhury@atu.edu (R. Choudhury).
https://doi.org/10.1016/j.saa.2020.118305
1386-1425/© 2020 Elsevier B.V. All rights reserved.

2
R. Choudhury et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 235 (2020) 118305
Fig. 1. (a) Absorption spectra of probe 1 in various solvents. [1] = 1.0 × 10−5 M. Aqueous solution contains 1% DMSO. (b) Fluorescence spectra of probe 1 in various solvents. [1] =
1.0 × 10−5 M. Aqueous solution contains 1% DMSO. For emission spectra, all the samples were excited at the absorption maxima, except for in methanol solution (λex = 485 nm in
methanol).
serum albumin in buffer as well as urine samples, but they encounter
some limitations such as poor analyte selectivity, low aqueous solubil-
ity, low sensitivity, extensive preparation method, and high background
fluorescence [9,18,19,23–26].
at ~9.65 ppm in DMSO confirmed the presence of a phenolic –OH group
on 1 (Fig. S2). Selectivity to all-trans isomer for the condensation reac-
tion was very high, at least within the limits of the NMR timescale. Pres-
ence of trans hydrogens was confirmed from the large coupling
constants (J = 16 Hz) [30].
Herein, we have reported development of a new phenol based
donor-π-π-acceptor probe, which displayed selective fluorescence re-
sponse with human serum albumin in the red (665 nm) and NIR-I
(780 nm) region of the electromagnetic spectrum [27]. The fluorophore
exhibited sufficient aqueous solubility to hinder any aggregate forma-
tion in the test samples. Moreover, it remained ionized within a broad
pH range—including at neutral pH—due to stereo-electronic properties
of the incorporated methoxy (-OCH3) and tricyanofuran (TCF) groups
(Fig. 2c) [22,28]. In aqueous solution the ionized form populated the
charge transfer (CT) states considerably. Therefore, significant amount
of excited state energy was lost via competing non-radiative decays
[29]. As a result, the background fluorescence of the probe was very
low. The quenched emissions were then recovered upon complexation
of the fluorophore within the protein's subdomain, resulting 10 mg/L
and 20 mg/L limit of detection (LOD) in simulated urine solutions in
the red and NIR-I region of the spectrum, respectively. The fluorophore
selectively responded to human serum albumin (HSA) over homologue
bovine serum albumin (BSA) and other proteins and enzymes with
rapid enhancement of the fluorescence emission in both aqueous buffer
and simulated urine samples. By employing fluorescence titrations, dis-
placement assays, and molecular docking studies the underlying mech-
anism of the high selectivity was examined, and it was attributed to the
site-specific supramolecular complexation of the fluorophore with HSA.
Moreover, high sensitivity, superior selectivity, and dual emission win-
dow resulted quantitative detection of serum albumin in human urine
samples, both in the red and NIR-I widow.
The absorption and emission maxima of 1 displayed a noticeable de-
pendence on the solvent polarity. As shown in Fig. 1a, as the polarity of
the solvent increased, the λambasx bathochromically shifted. In polar protic
solvents, such as water and methanol, the absorption spectra were more
structured and extended in the NIR-I region. Similarly, the fluorescence
spectra showed positive solvent dependency; ~80 nm bathochromic
shift—along with reduced intensity— was recorded when solvent was
switched from ethyl acetate to dimethyl sulfoxide. In general, the fluo-
rescence intensities in polar solvents were lower than that of the non-
polar solvents, in agreement with the formation of the solvent stabilized
charge transfer states [29,31]. Also, the quantum yield of 1 was calcu-
lated in three different solvents with respect to a reference dye and it
was found to be 0.001 in DMSO, 0.003 in methanol, and 0.012 in
acetone.
Phenol is moderately acidic in water (pKa ≈ 10). However, its pKa
can be lowered by incorporating electron withdrawing group(s) on
the benzene ring [32]. Interestingly, 1 remains predominantly in the
abs
deprotonated form in pH 7.4 buffer solution (λ
= 630 nm; ε =
max
28,480 Lmol−1 cm−1, Fig. 2a). The signal at 480 nm was attributed to
the neutral form which gradually decreased upon increasing the pH
from 4.0–6.0–7.4–9.0, passing through an isobestic point at 540 nm. It
completely disappeared at ~pH 9.0, indicating presence of highest frac-
tion of deprotonated form (Fig. 2c, ionic form I); the extinction coeffi-
cient at λmax was greater at pH 9.0 and 10.0 than pH 7.4. Fluorescence
of 1 was also recorded in different buffer solutions (Fig. 2b). When 1
was excited at 480 nm, broad structured emission (range:
550–750 nm; Stokes' shift 135 nm) was observed. On the other hand,
2. Results and discussion
narrow, low-intensity signals were observed upon excitation (λex
=
630 nm) of the deprotonated forms, displaying a small Stokes' shift
(~25 nm).
Probe 1 was synthesized in two steps as shown in Scheme S1 (elec-
tronic supplementary information) [20]. First, to obtain the TCF acceptor
(C), 3‑hydroxy‑3‑methylbutan‑2‑one was refluxed with malononitrile
and sodium ethoxide under N2 atmosphere. TCF intermediate was
then coupled with 3‑(4‑hydroxy‑3,5‑dimethoxyphenyl)acrylaldehyde
via Knoevenagel condensation in presence of a weak base to yield the
target probe (1). The probe was purified by silica gel chromatography
and characterized by IR, NMR, and elemental analysis (Figs. S1 and S2,
electronic supplementary information). In the IR spectrum, presence
of a sharp signal at ~2200 cm−1 and a broad signal at ~3320 cm−1 indi-
cate nitrile and phenolic functional group, respectively (Fig. S2). In the
1H NMR spectrum, signal at 3.82 ppm confirmed presence of –OCH3
groups attached to a phenyl moiety [22]. Moreover, its similar integral
ratio with signal at 1.76 ppm on the TCF moiety suggests a successful
condensation reaction. Additionally, the broad downfield shifted signal
Probe 1 exhibited an absorption band at 740 nm (ε = 3790 Lmol−1-
cm−1) in pH 7.4 buffer solution (Fig. 2a). The intensity of the signal in-
creased at higher pH (ε = 19,960 Lmol−1 cm−1 at pH 9.0; ε = 20,470
Lmol−1 cm−1 at pH 10.0), indicating existence of another solvent stabi-
lized form of 1 in water, and its amount increases as the basicity of the
solution increases. Presumably, this is a twisted conformer, such as a
cisoid conformer, of 1 which originates from the rotations of the
bonds in the dimethine bridge. Moreover, the 110 nm bathochromic
shift appears only when there is a significant amount of deprotonated
form (ionic form I) present in the solution; no absorption beyond
630 nm was detected in pH 6.0 buffer solution, indicating strong
donor-acceptor charge transfer is required in the ground state for pop-
ulation of this twisted conformer [33]. In our previous studies, no such

R. Choudhury et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 235 (2020) 118305
3
Fig. 2. (a) Absorption spectra of 1 in water at different pH values. (b) Fluorescence spectra of 1 in water at different pH values. All the aqueous solutions contain 1% DMSO. For emission
spectra, all the samples were excited at the absorption maxima. (c) Protonated, deprotonated, and probable twisted conformer of 1 in water. [1] = 1.0 × 10−5 M. Green circle indicates the
isobestic point. Dotted ellipse shows absorption centered at 740 nm in higher pH solutions.
near infrared absorption was recorded with one alkenyl conjugated
phenolate-O(−)-π-acceptor fluorophore, suggesting that this is unique
for a phenolate-O(−)-dimethine-acceptor system [22].
both neutral and deprotonated form (ionic form I). As shown in
Fig. 3a, the π-electrons on the HOMO of neutral 1 was delocalized on
the π-conjugated framework, but the LUMO was positioned near the ac-
ceptor side. When the phenol-OH was deprotonated, the HOMO was
mainly centered at the phenolate site, and for the LUMO the larger
lobes were observed on the acceptor unit. In fact, it is the large destabi-
lization of the HOMO of the phenolate-O(−) reduces the overall HOMO/
Phenolate-O(−) is a better electron donating group than phenol-OH
[22,32]. To find a correlation between the donor strength and the optical
properties 1, semi-empirical calculations (ZINDO/1) were performed on
HyperChem suite in gas phase, and the HOMO/LUMO were plotted for
Fig. 3. (a) Molecular orbital plots (HOMO and LUMO) and the corresponding energies of the neutral form (Fig. 2c) in the ground state. (b) Molecular orbital plots (HOMO and LUMO) and
the corresponding energies of the ionic form I (Fig. 2c) in the ground state.

4
R. Choudhury et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 235 (2020) 118305
Fig. 4. (a) Fluorescence spectra of 1 (10 μM) in different ratios of buffer (pH = 7.4)-ethylene glycol mixtures, λex = 630 nm. (b) Fluorescence spectra of 1 (10 μM) in different ratios of
buffer (pH = 7.4)-glycerol mixtures, λex = 630 nm.
LUMO gap, most likely responsible for the large shift of λmax (150 nm) in
the absorption spectrum (Fig. 2a).
solution. Upon addition of one equivalent of HSA into a solution of 1
(10 μM; pH 7.4 buffer), the absorption maxima bathochromically
(~18 nm) and hyperchromically shifted, suggesting a ground state com-
plex formation (Fig. S3, electronic supplementary information) [35].
The fluorescence emission increased fifteen fold (Q.Y. of 1 in buffer
was 0.01 and in presence of one equivalent HSA was 0.10) when 1
was excited at 630 nm in presence of one equivalent of HSA (excitation
spectra Fig. S4a, electronic supplementary information). Similarly, upon
excitation at 740 nm, fifteen fold increase in emission at ~780 nm was
recorded under identical experimental conditions (Fig. 5b, excitation
spectra Fig. S4b, electronic supplementary information). A concentra-
tion dependent study in simulated urine samples revealed a linear rela-
In polar solvents—as well as in basic buffer solutions—1 exhibits low
fluorescence intensity, suggesting energy loss from the excited states
via non-radiative pathways (Figs. 1b and 2b). Fig. 4 shows effect of solu-
tion viscosity on the fluorescence emission of 1 (λex = 630 nm). Ethyl-
ene glycol and glycerol are viscous liquids (viscosity of ethylene glycol
35.7 and glycerol 950 cP at 25 °C, respectively) [29,31,34]. Upon increas-
ing the glycerol amount in a water-glycerol mixture of 1, a systematic
fluorescence enhancement was observed (Fig. 4b). Similar results
were obtained in ethylene glycol-water mixtures (Fig. 4a). Two fold
and four fold fluorescence enhancement was recorded in 9:1 water-
ethylene glycol and water-glycerol mixture, respectively, suggesting re-
striction of conformational freedom of the rotatable bonds [29].
Conformational change about the dimethine chain connecting the
phenolate donor and the TCF acceptor can be prevented in a highly vis-
cous microenvironment. Consequently, we investigated the optical
properties of 1 in presence of human serum albumin in aqueous buffer
tionship between the emission and the HSA up to 500 mg/L for the red
em
emission (λemmax~665 nm) and 300 mg/L (λ
~780 nm) for the NIR
max
emission. From these concentration dependent fluorescence data the
limit of detections (LOD) were calculated (LOD = 3σ/slope), and they
were found to be 10 mg/L and 20 mg/L, respectively (Fig. 5c for the
red emission and Fig. 5d for the NIR-I emission) [36]. (Table S1,
Fig. 5. (a) Plot of emission (λex = 630 nm) of probe 1 in the presence of HSA. Inset: Showing fluorescence of 1 in absence (i) and presence (ii) of one equivalent of HSA when excited with a
max
laser (source 630 nm). (b) Plot of emission (λex = 740 nm) of probe 1 in the presence of HSA. (c) Linear relationship between the emission maxima of 1 (λ
~665 nm) and the amount of
em
max
HSA in simulated urine. (d) Linear relationship between the emission maxima of 1 (λ
~780 nm) and the amount of HSA in simulated urine.
em

R. Choudhury et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 235 (2020) 118305
5
Fig. 6. (a) Fluorescence emission of 1 (10 μM) in presence of various ions and bio-analytes, λex = 630 nm. (b) Fluorescence emission of 1 (10 μM) in presence of HSA and other
biomacromolecules in phosphate buffer (pH = 7.4), λex = 630 nm.
electronic supplementary information) These wide detection ranges
make 1 a promising probe for estimation of HSA in normal and
microalbuminuria conditions. Moreover, as shown in Fig. S5 (electronic
supplementary information), the emission did not change under nor-
mal laboratory lights. The emission intensity almost immediately en-
hanced upon addition of HSA, and it was constant over an extended
period of time.
Next, the interfering effect of ionic species and bio-analytes—com-
monly found in urine— was investigated. No significant change in fluo-
rescence emission (λex = 630 nm) of 1 (10 μM) was observed in
presence of external species, up to a concentration as high as 200
micro-molars (Fig. 6a). Additionally, the selective response property
of serum albumin were calculated from the linear relationship between
fluorescence intensity and the concentration of albumin. The calculated
amount obtained from the standard curve was found to be very similar
to the added amount of HSA (Table S2, electronic supplementary infor-
mation). Moreover, ~16-fold fluorescence enhancement was recorded
in human blood serum in presence of 10 μM of 1.
To probe the high selectivity and turn-on sensing mechanism of 1, li-
gand competitive experiments with three site specific compounds
(warfarin, domain IIA marker; salicylic acid, site marker for subdomains
IB and IIA; ibuprofen, site IIIA marker) were performed [37,38]. The
emission at ~665 nm (λex = 630 nm) was monitored for all the exper-
iments. Addition of incremental amount of warfarin (10, 20, 30, 40 and
50 μM) in the mixtures of 1@HSA resulted increase in fluorescence
emission (Fig. 7a). An overall ~50% increase in fluorescence was re-
corded. Whereas, addition of salicylic acid (Ka = 6.0 × 104 M−1) de-
creased the fluorescence up to ~70% (Fig. 7a), and ibuprofen did not
change the fluorescence of 1@HSA (Fig. S7a, electronic supplementary
information). Addition of even excess (12 equiv.) did not show any ef-
fect on the fluorescence of 1@HSA complex. More importantly, myristic
acid (a site IB binder) reduced the fluorescence emission of 1 by ~85% of
the original value (Fig. S7b, electronic supplementary information).
Separate control experiments with all the site specific markers have
confirmed absence of any quenching for 1 in buffer solutions.
was studied by measuring the fluorescence emission of 1 (λem
=
665 nm, λex = 630 nm) in the presence of several proteins and enzymes
such as BSA (a homologue HSA), γ-globulin (antibody isotype), chicken
egg albumin, insulin, lysozyme, hemoglobin, pepsin, and trypsin
(Fig. 6b). Our probe selectively recognized HSA. Though slight fluores-
cence enhancement (~3 fold) was observed in presence of BSA; its pres-
ence is unlikely in the urine samples. Other biomolecules, including γ-
globulin, had no effect on the fluorescence of 1 (Fig. 6b). Similar selec-
tive recognition was also observed with the NIR-I emission wavelength
at ~780 nm (Fig. S6, electronic supplementary information, λex
=
740 nm). To test the stability and performance of 1, three real human
urine samples were spiked with serum albumin and fluorescence inten-
sities were measured at both emission wavelengths. The measured fluo-
rescence intensities were fitted to the standard curve and the amounts
As probe 1 is non-fluorescent in aqueous buffer, an increase in signal
intensity upon addition of warfarin might suggest two events: (1) war-
farin displaces 1 from the binding pocket, and (2) 1 translates into other
Fig. 7. (a) The change in fluorescence intensity at 665 nm (λex = 630 nm) upon addition of increasing concentrations of site specific compounds warfarin and salicylic acid to the
complexes of 1 (5 μM) and HSA (333 mg/L). (b) Docking conformation of 1@HSA complex showing predominant bindings at site IB. (c) The change in fluorescence intensity at
665 nm (λex = 630 nm) upon addition of increasing concentrations of site specific compounds warfarin and salicylic acid to the complexes of 1 (5 μM) and BSA (333 mg/L).

6
R. Choudhury et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 235 (2020) 118305
binding site(s) of the protein when a competitive binder (warfarin,
Ka = 3.4 × 104 M−1) is present in the solution. However, with salicylic
acid an opposite effect was observed. Upon addition of increasing
amount of salicylic acid (10, 20, 30, 40 and 50 μM), the fluorescence in-
tensity gradually decreased to an overall 70% less than the original, indi-
cating effective displacement of 1 from the HSA. Salicylic acid binds at
both subdomain IB and IIA [39]. Therefore, the effective displacement
by salicylic acid indicates that 1 binds predominantly at IB or between
IB and IIA; in the latter scenario, it would be displaced and migrated
into the subdomain IB in presence of a site IIA binder—such as warfarin.
The fluorescence enhancement in presence of warfarin explains that
phenomenon, as discussed earlier. In summary, based on these dis-
placement experiments, we conclude that 1 predominantly binds
within the IB subdomain of HSA.
To obtain a more realistic view of the binding site within HSA, mo-
lecular dockings were performed with AutoDock Vina program [40].
Total three blind dockings were performed, and in each docking the
search space included the entire surface of the protein. The binding en-
ergy for each pose was negative, suggesting spontaneous, energetically
favorable binding process. As shown in Fig. 7b, 1 was mainly (93% of the
total 27 conformers) bound at site IB, thus corroborating with the
experiments.
On the other hand, the site specific binders exhibited minimal dis-
placement when titrated against 1@BSA complex (Fig. 7c), as revealed
by the minimal change of fluorescence emission. For warfarin 4% in-
crease and for salicylic acid 21% decrease was recorded, respectively
(Fig. 7c). As anticipated, insignificant change in intensity was observed
when a 1:1 complex of 1@BSA was titrated with ibuprofen. All these dis-
placement assays suggest that 1 does not bind in the common drug
binding sites of BSA. Therefore, to obtain a visual perspective of the com-
plexation site, three blind molecular dockings were performed with
BSA. As shown in Fig. S8 (electronic supplementary information), 1 pre-
dominantly binds between site IB and IIIA (60% of the total 27 con-
formers), which is in accordance with the displacement experiments,
and only 11% of the conformers were located at site IB.
Furthermore, in highly viscous glycerol, the emission intensity at
780 nm increases dramatically. An eight fold enhancement was re-
corded with 90% glycerol in phosphate buffer (Fig. S12, electronic sup-
plementary information), suggesting a significant conformational
restriction must be responsible for the reduced decay of the excited
state energy of the conformer.
3. Conclusion
We have designed and developed a novel long-wavelength emitting
charge transfer probe. Our design is based on a donor-π-π-acceptor
fluorophoric framework that can undergo deprotonation at physiologi-
cal pH to produce a new fluorochrome.
The new fluorochrome displayed turn-on response toward HSA,
both in the red and NIR-I region of the spectrum. It selectively
responded to HSA over other common proteins and enzymes, including
homologue BSA. Based on concentration dependent titration experi-
ments the limit of detections were calculated and found to be 10 mg/L
and 20 mg/L in the red and NIR-I window, respectively. The fluores-
cence enhancement of the probe was attributed to the minimization
of the torsional modes of deactivation of the dimethine-bridge. The se-
lective sensing mechanism was investigated by the displacement assays
and molecular docking, and it was ascribed to a strong supramolecular
complex of the probe at a specific site of HSA.
The dual emission window and the selective turn-on response to-
ward HSA make 1 a practically viable probe for biological applications
[43]. Moreover, the strategy of generating a stronger phenolate donor
from its latent form provides a convenient route to produce background
free turn-on probes for various biological applications. The introduction
of various substituents on the benzene ring to manipulate the intramo-
lecular charge transfer to extend the absorption and emission wave-
lengths appears to be a viable approach. Such fluorophore should be
accessible synthetically from condensation reaction between an alde-
hyde precursor and a reactive acceptor.
From the experimental and molecular modelling results the selec-
tive response of 1 toward HSA can be attributed to its site-specific com-
plexation with HSA. The calculated average binding energy obtained
from the molecular dockings (−8.8 kcal/mol)—as well as the experi-
mentally determined association constant (Benesi-Hildebrand plot;
Ka = 1.4 × 105 M−1; Fig. S9, electronic supplementary information)—
suggest a strong supramolecular complex between 1 and HSA [18,41]
Site I in HSA comprises four helix bundles located in subdomain IB.
This site is water accessible, however, it is more hydrophobic than the
site I of BSA [42]. The hydrophobic residue Phe 122 and two other hy-
drophobic residues res 120 and res 126 are located on a small surface
exposed helix. At site IB, the hydrophobic as well as other non-
covalent interactions, such as polar interaction with lysine 137, rigidify
the fluorophore, which could make it more emissive in the bound state
(Fig. S10a, b, and d in electronic supplementary information) [18–22].
On the other hand, poor interactions (average binding energy =
−7.6 kcal/mol obtained from the molecular dockings) of 1 with BSA at
a non-specific site (between domain I and III, Fig. S8) most likely results
minimal fluorescence enhancement (Figs. S10c, d, and S11, electronic
supplementary information). In case of BSA, 1 was predominately
found at the intersection between IB and III (Figs. S8 and S10d in elec-
tronic supplementary information). This intersection is solvent accessi-
ble and the amino acid sequence differs from HSA in several analogous
helix regions [42]. For example, between HSA and BSA at analogous res-
idues 189 and 190, HSA is very hydrophilic whereas BSA is strongly hy-
drophobic. The composition of amino acid sequence between 455 and
457 at site III in BSA also differs from HSA. But, both are considered as
very hydrophobic. Moreover, at the binding site of BSA no polar interac-
tions between 1 and residues were found from the molecular docking
analysis. These docking results demonstrate that 1 selectively binds at
site IB which results in the difference in fluorescence response from BSA.
CRediT authorship contribution statement
Rajib Choudhury:Conceptualization, Methodology, Resources,
Writing - original draft, Supervision, Project administration, Funding ac-
quisition, Investigation, Validation.Benjamin Quattlebaum:Investiga-
tion, Formal analysis.Charles Conkin:Investigation, Formal analysis.
Siddhi Rajeshbhai Patel:Investigation, Visualization.Kallie Menden-
hall:Investigation.
Declaration of competing interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Acknowledgements
This publication was made possible by the Arkansas INBRE program,
supported by a grant from the National Institute of General Medical Sci-
ences (NIGMS), and P20 GM103429 from the National Institutes of
Health. We wish to thank Dr. Anindya Ghosh at University of Arkansas,
Little Rock, for collecting the 1H and 13C NMR spectra.
Appendix A. Supplementary data
Details of the synthesis of the probes, 1H and 13C NMR spectra,
photostability study under visible light, details of molecular dockings, ti-
tration experiments, and additional spectra. This material is available
online via www.elsevier.com. Supplementary data to this article can
be found online at doi: 10.1016/j.saa.2020.118305.

R. Choudhury et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 235 (2020) 118305
[21] R. Choudhury, S.R. Patel, H.R. Lykins, The Chemist 91 (2018) 27–32.
7
References
[22] R. Choudhury, S.R. Patel, A. Ghosh, J. Photochem. Photobiol. A 376 (2019) 100–107.
[23] G. Chakraborty, A.K. Ray, P.K. Singh, H. Pal, Chem. Commun. 54 (2018) 8383–8386.
[24] Q. Jin, L. Feng, S. Zhang, D. Wang, F. Wang, Y. Zhang, J. Cui, W. Guo, G. Ge, L. Yang,
Anal. Chem. 89 (2017) 9884–9891.
[1] G.J. Quinlan, G.S. Martin, T.W. Evans, Hepatology 41 (2005) 1211–1219.
[2] T. Peters, All About Albumin: Biochemistry, Genetics, and Medical Applications, Ac-
ademic Press Inc, San Diego, USA, 1996.
[3] B.T. Doumas, T. Peters, Clin. Chim. Acta 258 (1997) 3–20.
[4] S. Arques, P. Ambrosi, J. Card. Failure 17 (2011) 451–458.
[5] K. Hoogenberg, W.J. Sluiter, R.P. Dullaart, Acta Endocrinol. 129 (1993) 151–157.
[6] R. Amin, B. Widmer, A.T. Prevost, P. Schwarze, J. Coope, J. Edge, L. Marcovecchio, A.
Neil, R.N. Dalton, D.B. Dunger, Br. Med. J. 336 (2008) 697–701.
[7] H. Thakkar, D.J. Newman, F. Holownia, C.L. Davey, C.C. Wang, J. Lloyd, A.R. Craig, C.P.
Price, Clin. Chem. 43 (1997) 109–113.
[8] Z.Z. Huang, H.N. Wang, W.S. Yang, Appl. Mater. Interface 7 (2015) 8990–8999.
[9] H.X. Sun, J.F. Xiang, X.F. Zhang, H.B. Chen, Q.F. Yang, Q. Li, A.J. Guan, Q. Shang, Y.L.
Tang, G.Z. Xu, Analyst 139 (2014) 581–584.
[10] R.E. Wang, L. Tian, Y. Chang, J. Pharm. Biomed. Anal. 63 (2012) 165–169.
[11] D. Caballero, E. Martinez, J. Bausells, A. Errachid, J. Samitier, Anal. Chim. Acta 720
(2012) 43–48.
[25] Q. Sun, W. Wang, Z. Chen, Y. Yao, W. Zhang, L. Duan, J. Qian, Chem. Commun. 53
(2017) 6432–6435.
[26] S. Sakamoto, T. Komatsu, T. Ueno, K. Hanaoka, Y. Urano, Bioorg. Med. Chem. Lett. 27
(2017) 3464–3467.
[27] C. Staudinger, S.M. Borisov, Methods Appl. Fluoresc. (2015) 42005.
[28] W.E. Moerner, R.J. Tweig, D.W. Kline, M. He, U.S. Patent 20070134737A1, 2007.
[29] Z.R. Grabowski, K. Rotkiewicz, Chem. Rev. 103 (2003) 3899–4031.
[30] P.D.E. Pretsch, T. Clerc, J. Seibl, W. Simon, Spectral Data for Structure Determination
of Organic Compounds, Springer-Verlag, Heidelber, 1989.
[31] F.A.S. Chipen, A. Mishra, G. Krishnamoorthy, Phys. Chem. Chem. Phys. 14 (2012)
8775–8790.
[32] M. Ipuy, C. Billon, G. Micouin, J. Samarut, C. Andraud, Y. Bretonniere, Org. Biomol.
Chem. 12 (2014) 3641–3648.
[33] A. Maliakal, G. Lem, N.J. Turro, R. Ravichandran, J.C. Suhadolnik, A.D. DeBellis, M.G.
[12] M.C. Tu, Y.Z. Chang, Y.T. Kang, H.Y. Chang, P. Chang, T.R. Yew, Biosens. Bioelectron.
Wood, J. Lau, J. Phys. Chem. A 106 (2002) 7680–7689.
34 (2012) 286–290.
[34] B. Wandelt, P. Turkewitsh, B.R. Stranix, G.D. Darling, Faraday Trans. 91 (1995)
4199–4205.
[13] Z. Luo, B. Liu, K. Zhu, Y. Huang, C. Pan, B. Wang, L. Wang, Dyes Pigm 152 (2018)
60–66.
[14] R. Singh, F.W. Crow, N. Babic, W.H. Lutz, J.C. Lieske, T.S. Larson, R.A. Kumar, Clin.
[35] A.S. Tatikolov, S.M.B. Costa, Biophys. Chem. 107 (2004) 33–49.
[36] G.L. Long, J.D. Winefordner, Anal. Chem. 55 (1983) 712–724.
[37] N.E. Basken, G.J. Mathias, M.A. Green, J. Pharm. Sci. 98 (2009) 2170–2179.
[38] I. Petitpas, A.A. Bhattacharya, S. Twine, M. East, S. Curry, J. Biol. Chem. 276 (2001)
22804–22809.
[39] L. Karami, E. Tazikeh-Lemeski, A.A. Saboury, Phys. Chem. Res. 5 (2017) 483–496.
[40] O. Trott, A.J. Olson, J. Comput. Chem. 31 (2009) 455–461.
[41] H.A. Benesi, J.H. Hildebrand, J. Am. Chem. Soc. 71 (1949) 2703–2707.
[42] Y. Akdogan, J. Reichenwallner, D. Hinderberger, PLoS One 7 (e45681) (2012) 1–9.
[43] R. Tian, et al., Sci. Adv. 5 (2019) 1–11.
Chem. 53 (2007) 540–542.
[15] Y. Ermolenko, A. Anshakova, N. Osipova, M. Kamentsev, O. Maksimenko, V.
Balabanyan, S. Gelperina, J. Pharmacol. Toxicol. Methods 85 (2017) 55–60.
[16] X.D. Zeng, M.S. Ma, B.C. Zhu, L.A. Zhu, Anal. Sci. 32 (2016) 1291–1294.
[17] K. Rajasekhar, J.A. Chaithra, T. Govindaraju, Org. Biomol. Chem. 15 (2017)
1584–1588.
[18] S. Samanta, S. Halder, G. Das, Anal. Chem. 90 (2018) 7561–7568.
[19] S.I. Reja, I.A. Khan, V. Bhalla, M. Kumar, Chem. Commun. 52 (2016) 1182–1185.
[20] R. Choudhury, H.E. Parker, C.M. Cendejas, K.L. Mendenhall, Tet. Lett. 59 (2018)
3020–3025.