Synthesis of highly selective lysosomal markers by coupling 2-(2′-hydroxyphenyl)benzothiazole (HBT) with benzothiazolium cyanine (Cy): the impact of substituents on selectivity and optical properties

Article abstract of DOI:10.1039/c9tb01672d

HBT-Cy 1 has been previously reported as a highly selective fluorescent probe for lysosome visualization in live cells. To further investigate the role of the structural components of HBT-Cy in lysosome selectivity, cyanine based fluorescent probe series

Full text of DOI:10.1039/c9tb01672d

Journal of
Materials Chemistry B
View Article Online
View Journal | View Issue
PAPER
Synthesis of highly selective lysosomal markers by
coupling 2-(20-hydroxyphenyl)benzothiazole (HBT)
with benzothiazolium cyanine (Cy): the impact of
substituents on selectivity and optical properties†
a
Cite this: J. Mater. Chem. B, 2019,
7, 7502
Chathura S. Abeywickrama,
Keti A. Bertman,a Lucas J. Mcdonald,a
a
a
Nicolas Alexander,a Dipendra Dahal,
Hannah J. Baumann,a Carrie R. Salmon,
Chrys Wesdemiotis,a Michael Konopka,
Claire A. Tessiera and Yi Pang
*
a
ab
HBT-Cy 1 has been previously reported as a highly selective fluorescent probe for lysosome visualization
in live cells. To further investigate the role of the structural components of HBT-Cy in lysosome selectivity,
cyanine based fluorescent probe series (2–5) have been synthesized in good yields by connecting
benzothiazolium cyanine (Cy) with 2-hydroxyphenylbenzothiazole (HBT) via a meta phenylene ring.
Probes 2–5 exhibited exceptional photophysical properties including bright red-emission (lem E 630–650 nm),
a large Stokes shift (Dl 4 130 nm) and high fluorescence quantum yields (ffl E 0.1–0.5). Probes 2, 3,
and 5 exhibited exceptional selectivity towards cellular lysosomes in NHLF and MO3.13 cells. Our further
study revealed that the phenyl benzothiazolium cyanine component (6) was the lysosome directing
group in the HBT-Cy probe structure. The attachment of the hydroxyphenyl benzothiazole (HBT) com-
ponent to the HBT-Cy probe structure has significantly improved its photophysical properties. Lysosome
probes 2, 3 and 5 exhibited excellent biocompatibility, quick staining, bright red fluorescence, and
wash-free application for live cell imaging. These probes further exhibited excellent characteristics for
bioimaging experiments including a non-alkalinizing nature, high biocompatibility, high photostability
and long-term imaging ability (44 hours).
Received 7th August 2019,
Accepted 3rd October 2019
DOI: 10.1039/c9tb01672d
rsc.li/materials-b
developed to date for various imaging applications.10–16
Such fluorescent probes are capable of delivering detectable
1. Introduction
Exploring lysosome form and function is important to study fluorescence signals upon lysosomal localization. Lysosomes
severe disease conditions which occur due to lysosomal are membrane bound spherical organelles which contain an
malfunction.1–3 Lysosomal dysfunction can lead to severe disease acidic lumen (pH E 4–5).1,3,17,18 The acidic environment in the
conditions such as lysosomal storage diseases, metabolic disorders, lysosomal lumen is essential for the optimal activity of their
neurodegenerative disorders and cancer.1,2,4–8 Recent research degradative enzymes which are responsible for the biomacro-
has emphasized the correlation of lysosomal malfunction with molecule digestion process.6,11,18 Many lysosome-targeting
possible pro-oncogenic stages.2,8,9 Lysosome visualization in probes are using the acidic environment of the lysosomal
live or fixed cell samples provides important information such lumen in designing fluorescent probes.10,11,19 The most well-
as the lysosomal pH, morphology, distribution and abundance. known strategy is to employ basic functional groups (i.e. amine
Visualization of cellular lysosomes via fluorescence confocal groups) in the probe structure (i.e. LysoTrackers probes) which
microscopy is an advanced imaging technique currently act as a ‘‘targeting group’’ towards the acidic lysosomal
employed for lysosome studies. A vast number of lysosome lumen.10,20 The lysosome selectivity of such probes require an
selective organic small molecule fluorescent probes have been acidic pH environment to operate. However, fluorescent probes
with basic functional groups are well known to elevate the
lysosomal pH due to their characteristic alkalinizing effect
a Department of Chemistry, University of Akron, Akron, Ohio 44325, USA.
(i.e. LysoTrackers Green DND-26, LysoTrackers Red DND-99).
E-mail: yp5@uakron.edu
b Maurice Morton Institute of Polymer Science, University of Akron, Akron,
Therefore, the applications of such probes are often limited
to short time periods due to their inherent cytotoxicity. Alter-
Ohio 44325, USA
natively, some different approaches have been employed to
design spirolactam, flavonoid and peptide-based lysosome
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c9tb01672d
7502 | J. Mater. Chem. B, 2019, 7, 7502--7514
This journal is ©The Royal Society of Chemistry 2019

View Article Online
Paper
Journal of Materials Chemistry B
selective fluorescent probes which do not exhibit any alkaliniz-
ing effect.16,21–25 In an alternative approach, pyrene based
lysosome directing probes have been reported recently, which
are accumulated in the lysosomal membrane instead of the
lysosomal lumen.12 In addition, several ratiometric fluorescent
probes have been reported for lysosome visualization.26–28
However, the design and synthesis of pH independent lyso-
some probes is still in high demand due to the recent findings
of the lysosomal involvement in cancer.7,9,17,29–31
Many existing commercial probes such as LysoTrackers Green
DND-26 (lex B 504 nm and lem B 510 nm), LysoTrackers Red
DND-99 (lex B 577 nm and lem B 590 nm) and LysoTrackers
Blue DND-22 (lex B 380 nm and lem B 420 nm) suffer from a
small Stokes shift and a low signal to noise ratio (i.e., acridine
orange and neutral red) which limit their application in fluores-
cence microscopy.10,32 Therefore, an ideal candidate for lysosome
visualization should (1) be independent of the proton uptake,
(2) exhibit a large Stokes shift, (3) be highly selective, (4) provide
a high signal to noise ratio, and (5) be readily excitable with the
available lasers in the fluorescence microscope. In our previous
work, we reported a novel lysosome selective fluorescent probe 1
by coupling 2-(2-hydroxyphenyl)benzothiazole (HBT) with
benzothiazolium cyanine (Cy) via a meta phenylene ring
system.11 Probe 1 exhibited a remarkable selectivity towards
cellular lysosomes in live cell imaging experiments. New probe
1 provided an excellent platform for designing and synthesizing
a new class of lysosome selective HBT-Cy fluorescent probes. In
summary, HBT-Cy based lysosome probes have several advan-
tages over currently reported other probes including (1) struc-
tural simplicity and facile synthesis, (2) high biocompatibility,
(3) long-term imaging ability, (4) non-amino group containing
structure (no proton consumption within the cell), (5) lower
working concentration (i.e., 500 nM), (6) quick staining (7) high
specificity, and (8) wash-free staining. The HBT-Cy probe
skeleton also exhibits interesting photophysical characteristics
due to its structural features. Remarkably a large Stokes shift
(Dl 4 130 nm) is observed due to excited state intramolecular
proton transfer (ESIPT) coupling with intra-molecular charge
transfer (ICT). In addition, a large fluorescence quantum yield
is observed due to the co-planner rigid geometry of the probe
structure and the bright red to near infra-red (NIR) emission
makes these probes ideal candidates for bioimaging applica-
tions. One of the intriguing fundamental questions is how to
identify the effect of each component/substituent in a fluoro-
phore structure on its inherent lysosome selectivity and photo-
physical performance.
Scheme 1 Structures of the lysosome selective probes (2–6).
importance of the structural features in the observed lysosome
selectivity and optical properties by considering the role of each
components in the HBT-Cy structure.
2. Materials and methods
All solvents and reagents were used as received without further
purification unless specified otherwise. Reactions were per-
formed under standard atmospheric conditions in oven-dried
glassware. All molecular biology grade reagents for cell culture
and fluorescence confocal microscopy were purchased from
Fisher Scientific. UV-Vis spectra were obtained using a Hewlett
Packard-8453 diode array spectrophotometer at 25 1C. Fluores-
cence spectra were measured using a HORIBA Fluoromax-4
spectrofluorometer. 1H NMR spectra were obtained using a
Varian 300 MHz spectrometer in deuterated dimethyl sulfoxide
(d-DMSO) or chloroform (CDCl3) solvents. Fluorescence confocal
laser microscopy imaging was performed using a Nikon A1 system
with a 60ꢀ or 100ꢀ oil objective.
2.1 Synthesis
To further investigate the role of substituents in the HBT-Cy
structure, we have synthesized a library of isomeric HBT-
2.1.1 General procedure for synthesis of 2a, 2b and 7. In a
cyanine probes of 1 (probes 2–5) and a series of model com- flask, 1 mmol of the corresponding aldehyde (8–9) was dissolved in
pounds (6–7) with different substituents and connectivity 20 mL of methanol. Then, the corresponding 2-aminothiophenol
(Schemes 1 and 2). The isomeric probe versions 2–5 were (1.2 mmol) was added and the solution was stirred at room
synthesized by altering the substituents and connectivity across temperature for 5 minutes. Then I2 (10% by the aldehyde mole)
the HBT-Cy fluorophore skeleton. Probes 2, 3 and 5 exhibited was added and the resulting solution was stirred at room
exceptional lysosome selectivity in normal human lung fibro- temperature for 12 hours. After completion of the reaction,
blast (NHLF) and MO3.13 progenitor oligodendrocyte cell line the resulting product was collected by vacuum filtration
under wash-free conditions. In this work we further discuss the and washed with cold methanol (3 ꢀ 10 mL) 3 times.
This journal is ©The Royal Society of Chemistry 2019
J. Mater. Chem. B, 2019, 7, 7502--7514 | 7503

View Article Online
Journal of Materials Chemistry B
Paper
Scheme 2 Synthesis of probes 2–7.
The desired product (2a, 2b and 7) was collected using a the resulting product was collected by vacuum filtration and
Buchner funnel (70–85%).
washed with cold ethyl acetate (3 ꢀ 10 mL) 3 times. The desired
product (2c) was collected using a Buchner funnel (78% yield).
(E)-2-(3-(Benzo[d]thiazol-2-yl)-4-hydroxystyryl)-3-ethylbenzo[d]-
thiazol-3-ium iodide (2a). 2a was obtained as a dark green
powder with 66% yield. H NMR (300 MHz in DMSO) d 11.18
3-Ethyl-2-((E)-4-hydroxy-3-((E)-((2-mercaptophenyl)imino)methyl)-
styryl)benzo[d]thiazol-3-ium iodide (2c). 2c was obtained as a dark
grey powder with 77% yield. 1H NMR (300 MHz in DMSO)
d 12.56 (s, 1H), 8.85 (d, J = 5.5 1H), 8.46 (m, 2H), 8.30 (s, 1H),
8.19 (t, J = 6.7 Hz, 1H), 8.11 (t, J = 6.8 Hz, 1H), 7.86 (m, 3H), 7.55
(dq, J = 29.6, 7.0 Hz, 2H), 7.23 (m, 1H), 7.01 (dq, J = 21.6, 6.7 Hz,
1H), 6.71 (t, J = 7.0 Hz, 1H), 6.41 (q, J = 7.1 Hz, 1H), 5.44 (s, 1H),
1
(s, 1H), 8.36 (d, J = 8.01 Hz, 1H), 8.22 (m, 2H), 8.02 (d, J = 11.1 Hz,
1H), 7.83 (t, J = 7.8 Hz, 2H), 7.73 (m, 1H), 6.96 (t, J = 8.5 Hz, 2H),
6.80 (m, 2H), 6.53 (d, J = 7.4 Hz, 2H), 4.90 (q, J = 6.6 Hz, 2H),
and 1.41 (t, J = 6.4 Hz, 3H). HRMS (ESI) found (m/z) for
+
[C24H19N2OS2 ] 415.0952, 416.1301 and 417.1438. Calculated
+
(m/z) for [C24H19N2OS2 ] 415.0939, 416.0939 and 417.0939.
4.95 (q, J = 6.8 Hz, 2H), and 1.47 (t, J = 6.7 Hz, 3H). HRMS (ESI)
+
(E)-2-(3-(Benzo[d]thiazol-2-yl)-4-methoxystyryl)-3-ethylbenzo[d]- found (m/z) for [C24H21N2OS2 ] 417.1051, 418.1350 and 419.1360.
+
thiazol-3-ium iodide (2b). 2b was obtained as a dark orange Calculated (m/z) for [C24H21N2OS2 ] 417.1095, 418.1095 and
powder with 70% yield. 1H NMR (300 MHz in DMSO) d 8.44 419.1095.
(d, J = 8.1 Hz, 1H), 8.26 (dt, J = 15.6, 8.1 Hz, 2H), 8.03 (s, 1H),
2.1.3 General procedure for synthesis of 3–5. In a flask,
7.80 (m, 3H), 7.29 (t, J = 10.1 Hz, 1H), 6.95 (m, 3H), 6.77 (m, 2H), 1.1 mmol of the corresponding aldehyde (8–9) was dissolved in
4.93 (q, J = 6.8 Hz, 2H), 3.97 (s, 3H), and 1.46 (t, J = 6.5 Hz, 3H). 25 mL of methanol. Then, the corresponding 2-methyl benzo-
+
HRMS (ESI) found (m/z) for [C25H21N2OS2 ] 429.1048, 430.1341 thiazolium salt 12 (R+Xꢁ; 1.0 mmol) was added and the solution
+
and 431.1341. Calculated (m/z) for [C25H21N2OS2 ] 429.1095, was stirred at room temperature for 10 minutes. Following the
430.1095 and 431.1095.
addition of pyridine (0.25 mL), the resulting solution was
2.1.2 Synthesis of 2c. In a flask, 1 mmol of the corres- heated up and stirred at 65 1C for 12 hours. After completion
ponding aldehyde (8) was dissolved in 20 mL of methanol. Then, of the reaction, the reaction mixture was cooled down to room
the corresponding 2-aminothiophenol (1.1 mmol) was added temperature and concentrated under vacuum. To the resulting
and the solution was refluxed for 3 hours. After completion of dark red crude, ethyl acetate (10 mL) was added, and the
the reaction, 10 mL of ethyl acetate was added to the mixture and product was precipitated as a dark colored solid in the bottom
7504 | J. Mater. Chem. B, 2019, 7, 7502--7514
This journal is ©The Royal Society of Chemistry 2019

View Article Online
Paper
Journal of Materials Chemistry B
of the flask. After the solution was allowed to settle for 15 minutes, 7.51 (m, 2H), 7.21 (d, J = 7.5 Hz, 1H), 4.98 (q, J = 6.2 Hz, 2H), and
the resulting solid was collected by vacuum filtration and washed 1.50 (t, J = 6.0 Hz, 3H). HRMS (ESI) found (m/z) for
+
with ethyl acetate (3 ꢀ 10 mL) 3 times. The desired product (3–5) [C24H19N2OS2 ] 415.0969, 416.1243 and 417.1304. Calculated
+
was collected using a Buchner funnel (65–85% yield).
(m/z) for [C24H19N2OS2 ] 415.0939, 416.0939 and 417.0939.
(E)-2-(5-(Benzo[d]thiazol-2-yl)-2-methoxystyryl)-3-ethylbenzo[d]-
thiazol-3-ium iodide (5b). 5b was obtained as a dark yellow
powder with 70% yield. 1H NMR (300 MHz in DMSO) d 8.82
(d, J = 7.6 Hz, 1H), 8.47 (d, J = 7.4 Hz, 1H), 8.33 (m, 2H), 8.23
(d, J = 7.3 Hz, 1H), 8.10 (m, 3H), 7.83 (dd, J = 17.4, 4.8 Hz, 2H), 7.15
(d, J = 7.5 Hz, 2H), 5.00 (q, J = 6.7 Hz, 2H), 3.86 (s, 3H), and 1.50
(E)-2-(3-(Benzo[d]thiazol-2-yl)-4-hydroxy-5-methylstyryl)-3-methyl-
benzo[d]thiazol-3-ium iodide (3a). 3a was obtained as a dark red
1
powder with 69% yield. H NMR (300 MHz in DMSO) d 13.16
(s, 1H) 8.43 (m, 1H), 8.33 (d, J = 4.2 Hz, 1H), 8.22 (m, 3H), 8.11
(dd, J = 8.1, 4.2 Hz, 1H), 7.94 (d, J = 4.2 Hz, 1H), 7.88 (m, 1H), 7.86
(dd, J = 7.7, 4.1 Hz 1H), 7.56 (m, 1H), 7.54 (dd, J = 7.8, 4.1 Hz, 1H),
4.35 (s, 3H), and 2.36 (s, 3H). HRMS (ESI) found (m/z) for
+
(t, J = 6.5 Hz, 3H). HRMS (ESI) found (m/z) for [C25H21N2OS2 ]
+
429.1093, 430.1233 and 431.1234. Calculated (m/z) for
[C24H19N2OS2 ] 415.0940, 416.1288 and 417.1425. Calculated
+
+
[C25H21N2OS2 ] 429.1095, 430.1095 and 431.1095.
(m/z) for [C24H19N2OS2 ] 415.0939, 416.0939 and 417.0939.
2.1.4 General procedure for synthesis of 6. In a flask,
1.1 mmol of the corresponding aldehyde (11) was dissolved in
20 mL of methanol. Then, the corresponding 2-methyl benzo-
thiazolium salt 12 (R+Xꢁ; 1.0 mmol) was added and the solution
was stirred at room temperature for 10 minutes. Following the
addition of pyridine (0.5 mL), the resulting solution was heated
up and stirred at 65 1C for 18 hours. After completion of
the reaction, the reaction mixture was cooled down to room
temperature and concentrated under vacuum. To the resulting
dark red crude, ethyl acetate (10 mL) was added, and the product
was precipitated as a colored solid in the bottom of the flask.
After the solution was allowed to settle for 15 minutes, the
resulting solid was collected by vacuum filtration and washed
with ethyl acetate (3 ꢀ 10 mL) 3 times. The desired product (6)
was collected using a Buchner funnel (62–69% yield).
(E)-2-(3-(Benzo[d]thiazol-2-yl)-4-hydroxy-5-methylstyryl)-3-ethyl-
benzo[d]thiazol-3-ium iodide (3b). 3b was obtained as a dark
brown powder with 70% yield. 1H NMR (300 MHz in DMSO)
d 13.17 (s, 1H) 8.44 (m, 2H), 8.39 (q, J = 9.0, 8.0 Hz, 5H), 8.12
(d, J = 8.0 Hz, 1H), 7.90 (m, 2H), 7.79 (d, J = 7.6 Hz, 1H), 7.60 (d,
J = 7.6 Hz, 1H), 7.52 (d, J = 7.6 Hz, 1H), 4.98 (q, J = 6.9 Hz, 2H),
2.39 (s, 3H), and 1.47 (t, J = 6.5 Hz, 3H). HRMS (ESI) found (m/z)
+
for [C25H21N2OS2 ] 429.1088, 430.1457 and 431.1533. Calculated
+
(m/z) for [C25H21N2OS2 ] 429.1095, 430.1095 and 431.1095.
(E)-2-(3-(Benzo[d]thiazol-2-yl)-4-hydroxy-5-methylstyryl)-3-propyl-
benzo[d]thiazol-3-ium iodide (3c). 3c was obtained as a dark green
1
powder with 67% yield. H NMR (300 MHz in DMSO) d 13.16
(s, 1H) 8.42 (m, 3H), 8.30 (m, 3H), 8.12 (d, J = 7.4 Hz, 1H), 7.83
(m, 3H), 7.53 (m, 2H) 4.92 (t, J = 6.4 Hz, 2H), 2.39 (s, 3H), 1.91
(m, 2H) and 1.00 (t, J = 6.2 Hz, 3H). HRMS (ESI) found (m/z)
(E)-3-Ethyl-2-styrylbenzo[d]thiazol-3-ium iodide (6a). 6a was
+
1
for [C26H23N2OS2 ] 443.1246, 444.1473 and 445.1711. Calculated
obtained as a yellow powder with 62% yield. H NMR (300 MHz
+
(m/z) for [C26H23N2OS2 ] 443.1252, 444.1252 and 445.1252.
in DMSO) d 8.44 (d, J = 7.0 Hz, 1H), 8.22 (dd, J = 17.9, 4.7 Hz, 2H),
8.07 (m, 3H), 7.86 (m, 2H), 7.56 (m, 3H), 4.97 (q, J = 6.8 Hz, 2H),
and 1.47 (t, J = 6.5 Hz, 3H). HRMS (ESI) found (m/z) for
[C17H16NS+] 266.1003, 267.1215 and 268.1316. Calculated (m/z)
for [C17H16NS+] 266.1003, 267.1003 and 268.1003.
(E)-2-(3-(Benzo[d]thiazol-2-yl)-4-hydroxy-5-methylstyryl)-3-benzyl-
benzo[d]thiazol-3-ium bromide (3d). 3d was obtained as a dark
brown powder with 68% yield. 1H NMR (300 MHz in DMSO)
d 8.44 (m, 3H), 8.20 (m, 6H), 8.22 (d, J = 7.7 Hz, 3H), 7.75
(d, J = 7.8 Hz, 2H), 7.55 (m, 3H), 6.25 (s, 2H), and 2.37 (s, 3H).
(E)-2-(4-Hydroxystyryl)-3-methylbenzo[d]thiazol-3-ium iodide (6b).
+
1
HRMS (ESI) found (m/z) for [C30H23N2OS2 ] 491.1244, 492.1439
6b was obtained as a dark red powder with 65% yield. H NMR
+
and 493.1726. Calculated (m/z) for [C30H23N2OS2 ] 491.1252,
(300 MHz in DMSO) d 10.56 (s, 1H), 8.38 (d, J = 7.3 Hz, 1H), 8.17
(m, 2H), 7.92 (d, J = 7.1 Hz, 2H), 7.81 (d, J = 7.4 Hz, 1H), 7.76
(m, 2H), 6.94 (d, J = 6.8 Hz, 2H), and 4.26 (s, 3H). HRMS (ESI) found
(m/z) for [C16H14NOS+] 268.0782, 269.1091 and 270.1238. Calculated
(m/z) for [C16H14NOS+] 268.0796, 269.0796 and 270.0796.
492.1252 and 493.1252.
(E)-2-(5-(Benzo[d]thiazol-2-yl)-2,4-dihydroxy-3-methylstyryl)-3-
ethylbenzo[d]thiazol-3-ium iodide (4a). 4a was obtained as a
dark grey powder with 65% yield. 1H NMR (300 MHz in DMSO)
d 8.41 (d, J = 7.1 Hz, 1H), 8.36 (d, J = 7.4 Hz, 2H), 8.19 (m, 2H),
8.06 (d, J = 7.6 Hz, 1H), 7.99 (dd, J = 14.9, 4.8 Hz, 1H), 7.50
(dd, J = 19.7, 4.9 Hz, 2H), 4.92 (q, J = 6.9 Hz, 2H), 2.22 (s, 3H),
and 1.51 (t, J = 6.4 Hz, 3H). HRMS (ESI) found (m/z) for
[C25H21N2O2S2+] 445.1091, 446.1263 and 447.1369. Calculated
(m/z) for [C25H21N2O2S2+] 445.1045, 446.1045 and 447.1045.
(E)-3-Ethyl-2-(4-hydroxystyryl)benzo[d]thiazol-3-ium iodide (6c).
6c was obtained as dark yellow powder with 69% yield. 1H NMR
(300 MHz in DMSO) d 10.58 (s, 1H), d 8.37 (d, J = 7.4 Hz, 1H), 8.27
(d, J = 7.2 Hz, 1H), 8.12 (d, J = 7.4 Hz, 1H), 7.95 (d, J = 7.1 Hz, 2H),
7.81 (d, J = 6.9 Hz, 1H), 7.77 (m, 3H), 6.94 (d, J = 7.7 Hz, 2H), 4.89
(q, J = 6.6 Hz, 2H), and 1.45 (t, J = 6.7 Hz, 3H). HRMS (ESI) found
(m/z) for [C17H16NOS+] 282.1014, 283.1280 and 284.1192. Calcu-
lated (m/z) for [C17H16NOS+] 282.0953, 283.0953 and 284.0953.
(E)-2-(5-(Benzo[d]thiazol-2-yl)-2-hydroxystyryl)-3-ethylbenzo[d]-
thiazol-3-ium iodide (5a). 5a was obtained as a dark brown
1
powder with 81% yield. H NMR (300 MHz in DMSO) d 11.76
(E)-3-Benzyl-2-(4-hydroxystyryl)benzo[d]thiazol-3-ium bromide (6d).
(s, 1H) 8.67 (d, J = 7.4 Hz, 1H), 8.44 (d, J = 7.6 Hz, 2H), 8.34 6d was obtained as an orange powder with 67% yield. 1H NMR
(d, J = 7.6 Hz, 1H), 8.16 (m, 4H), 7.82 (dd, J = 16.3, 4.2 Hz, 2H), (300 MHz in DMSO) d 10.66 (s, 1H), 8.41 (d, J = 7.6 Hz, 1H),
This journal is ©The Royal Society of Chemistry 2019
J. Mater. Chem. B, 2019, 7, 7502--7514 | 7505

View Article Online
Journal of Materials Chemistry B
Paper
8.25 (d, J = 7.3 Hz, 1H), 8.11 (d, J = 7.6 Hz, 1H), 7.92 (m, 3H), Then the solid deposited on the bottom of the flask was
7.79 (m, 2H), 7.33 (m, 5H), 6.92 (d, J = 7.9 Hz, 2H), and 6.22 collected by vacuum filtration and washed with ice cold ethyl
(s, 2H). HRMS (ESI) found (m/z) for [C22H18NOS+] 344.1088, acetate (15 mL). The desired product (8) was collected using a
345.1330 and 346.1451. Calculated (m/z) for [C22H18NOS+] Buchner funnel (72% 85% yield).
344.1109, 345.1109 and 346.1109.
(E)-3-Ethyl-2-(3-formyl-4-hydroxystyryl)benzo[d]thiazol-3-ium iodide
(8a). 8a was obtained as a red powder with 85% yield. 1H NMR
(E)-3-Ethyl-2-(3-hydroxystyryl)benzo[d]thiazol-3-ium iodide (6e).
6e was obtained as a yellow-green powder with 68% yield.
1H NMR (300 MHz in DMSO) d 9.80 (s, 1H), 8.43 (d, J = 7.4 Hz,
1H), 8.32 (d, J = 7.1 Hz, 1H), 8.13 (d, J = 7.6 Hz, 1H), 7.86 (m, 3H),
7.44 (m, 3H), 7.01 (d, J = 7.2 Hz, 2H), 4.96 (q, J = 6.9 Hz, 2H), and
1.47 (t, 6.3 Hz, 3H). HRMS (ESI) found (m/z) for [C17H16NOS+]
282.0912, 283.1301 and 284.1337. Calculated (m/z) for [C17H16NOS+]
282.0953, 283.0953 and 284.0953.
(300 MHz in DMSO) d 11.89 (s, 1H), 10.35 (s, 1H), 8.41 (d, J =
8.0 Hz, 1H), 8.25 (m, 3H), 7.96 (s, 1H), 7.83 (m, 2H), 7.16 (d, J =
7.6 Hz, 1H), 4.96 (q, J = 6.6 Hz, 2H), and 1.43 (t, J = 6.4 Hz, 3H).
HRMS (ESI) found (m/z) for [C18H16NO2S+] 310.0891, 311.1134
and 312.1134. Calculated (m/z) for [C18H16NO2S+] 310.0902,
311.0902 and 312.0902.
(E)-3-Ethyl-2-(3-formyl-4-methoxystyryl)benzo[d]thiazol-3-ium
iodide (8b). 8b was obtained as an orange powder. 1H NMR
(300 MHz in DMSO) d 10.41 (s, 1H), 8.42 (d, J = 7.8 Hz, 2H), 8.30
(t, J = 7.7 Hz, 2H), 8.01 (d, J = 7.4 Hz, 1H), 7.86 (m, 3H), 7.42
(d, J = 7.4 Hz, 2H), 4.98 (q, J = 6.6 Hz, 2H), 4.06 (s, 3H), and 1.45
(t, J = 6.3 Hz, 3H). HRMS (ESI) found (m/z) for [C19H18NO2S+]
324.1090, 325.1364 and 326.1388. Calculated (m/z) for [C19H18NO2S+]
324.1058, 325.1058 and 326.1058.
(E)-3-Ethyl-2-(2-hydroxystyryl)benzo[d]thiazol-3-ium iodide (6f).
6f was obtained as a green powder with 63% yield. 1H NMR
(300 MHz in DMSO) d 10.85 (s, 1H), 8.40 (d, J = 7.7 Hz, 1H), 8.28
(d, J = 7.3 Hz, 2H), 8.01 (dd, J = 17.5, 4.5 Hz, 2H), 7.80 (m, 2H),
7.39 (m, 1H), 6.98 (m, 2H), 4.92 (q, J = 6.4 Hz, 2H), and 1.45
(t, J = 6.1 Hz, 3H). HRMS (ESI) found (m/z) for [C17H16NOS+]
282.0912, 283.1239 and 284.1275. Calculated (m/z) for
[C17H16NOS+] 282.0953, 283.0953 and 284.0953.
2.2 Fluorescence quantum yield calculation
(E)-3-Ethyl-2-(4-methoxystyryl)benzo[d]thiazol-3-ium iodide (6g).
6g was obtained as a yellow powder with 67% yield. H NMR
The fluorescence quantum yields (ffl) for compounds were calcu-
lated by using quinine sulfate or Rhodamine 6G (Sigma) as the
standard. The following equation was used for the calculation:
1
(300 MHz in DMSO) d 8.42 (d, J = 7.5 Hz, 1H), 8.24 (d, J = 7.2 Hz,
1H), 8.19 (d, J = 7.5 Hz, 1H), 8.06 (d, J = 7.6 Hz, 2H), 7.86 (m, 2H),
7.77 (d, J = 7.3 Hz, 1H), 7.12 (d, J = 7.4 Hz, 2H), 4.95 (q, J = 7.9 Hz,
2H), 3.88 (s, 3H), and 1.45 (t, J = 6.4 Hz, 3H). HRMS (ESI) found
(m/z) for [C18H18NOS+] 296.1098, 297.1170 and 298.1134. Calcu-
lated (m/z) for [C18H18NOS+] 296.1109, 297.1109 and 298.1109.
2.1.5 Synthesis of 7. In a flask, 1 mmol of the corres-
ponding aldehyde (9) was dissolved in 20 mL of methanol.
Then, the corresponding 2-aminothiophenol (1.2 mmol) was
added and the solution was stirred at room temperature for
5 minutes. Then I2 (0.1 mmol) was added and the resulting
solution was stirred at room temperature for 10 hours. After
completion of the reaction, the resulting product was deposited
on the bottom of the flask. Then the product was collected by
vacuum filtration and washed with cold methanol (3 ꢀ 10 mL) 3
times. The desired product (9) was collected using a Buchner
funnel (85% yield).
(ffl) sample = fref ꢀ (Absref/Abssample) ꢀ (Isample/Iref) ꢀ (Zsample)2/(Zref)2
where Abs is the absorbance of the sample, I is the integrated
fluorescence emission intensity and Z is the refractive index of
the solvent.
2.3 Live cell imaging
Fluorescence confocal microscopy imaging was performed
using a Nikon A1 confocal system with a 60ꢀ and 100ꢀ oil
objective with a numerical aperture of 1.45 and a refractive
index of 1.5. The imaging temperature was maintained at 37 1C
at all time. The excitation wavelengths used for our dyes were
405 nm and 561 nm. Standard DAPI, FITC, mCherry and Cy5
emission filters were used. All imaging was done using an
Okolab Bold Cage Incubator or at 37 1C, and the images were
processed using NIS Elements or ImageJ Pro (NIH) imaging
2,4-Bis(benzo[d]thiazol-2-yl)phenol (7). 7 was obtained as a software. The overlap coefficients were calculated by using the
light-yellow powder with 81% yield. 1H NMR (300 MHz in ImageJ (NIH) software (n 4 30). Commercial LysoTrackers Green
DMSO) d 12.16 (s, 1H), 9.05 (d, J = 7.8 Hz, 1H), 8.15 (m, 3H), DND-26 and MitoTrackers Green-FM probes were excited using
8.08 (d, J = 7.6 Hz, 2H), 7.54 (d, J = 7.7 Hz, 2H), 7.47 (d, J = a 488 nm laser line and emission was collected in the range of
7.6 Hz, 2H), and 7.24 (d, J = 7.4 Hz, 1H). HRMS (ESI) found (m/z) 495–550 nm. Commercial LysoTrackers Red DND-99 was
+
for [C20H13N2OS2 ] 361.0474, 362.0616 and 363.0632. Calcu- excited with a 561 nm laser line and emission was collected
+
lated (m/z) for [C20H13N2OS2 ] 361.0469, 362.0469 and 363.0469. in the 570 nm to 700 nm range. The probes were excited with
2.1.6 General procedure for synthesis of 8. In a flask, a 405 or 561 nm laser line and emissions were collected
1 mmol of the corresponding benzothiazolium salt (6) and according to the characteristic emission.
HMTA (1.2 mmol) were dissolved in 10 mL of trifluoroacetic
Note: Co-localization experiments were conducted with laser
acid. Then, the mixture was refluxed overnight (B12 hours). lines in multiple tracks cautiously to avoid any possible back-
After completion of the reaction, the resulting reaction mixture ground interferences. No background fluorescence signal was
was cooled down to 0 1C and neutralized with dilute NaOH. obtained for LysoTrackers Green with a 405 nm laser line.
7506 | J. Mater. Chem. B, 2019, 7, 7502--7514
This journal is ©The Royal Society of Chemistry 2019

View Article Online
Paper
Journal of Materials Chemistry B
2.4 Cell culturing and staining
according to the MTT manufacturer’s instruction. From this
point, the plates were protected from light and placed on a
shaker for 15 minutes. The absorbance values were read at
570 nm with a reference at 630 nm. The percent viability was
determined using the following equation (treatment absorbance –
blank)/(average vehicle control absorbance – blank) ꢀ 100. The
dose response curves and statistics were generated using the
GraphPad Prism 5 software.
For imaging experiments, probe stock solutions were prepared
in DMSO and the total DMSO% in imaging solutions was
maintained below 0.1% (v/v). NHLFs (normal human lung
fibroblasts) and MO3.13 cells (progenitor oligodendrocytes)
were plated on a Mat Tek 35 mm dish with glass bottom at a
density of 1 ꢀ 105 cells per plate in DMEM enriched with 10%
fetal bovine serum (Gibco) and 1% penicillin/streptomycin. The
cells were incubated overnight at 37 1C and in 5% CO2.
Subsequent to incubation, the cells were washed with 1ꢀ PBS
and treated with 0.5 mM probe (2–6) in Invitrogent Molecular
Probest Live Cell Imaging Solution for 30 min. After treatment,
the cells were washed once for 1 minute with 1 ꢀ PBS.
Invitrogent Molecular Probest Live Cell Imaging Solution
was added to the cells for imaging. Lysotrackers stock solution
was made with DMSO and the total DMSO levels were main-
tained in the media were below 0.25% (v/v%). All cells were
maintained in a 5% CO2 humidified atmosphere at 37 1C.
3. Results and discussion
3.1 Synthesis
The desired products 2–6 were synthesized according to Scheme 2.
Products 2–6 were further purified by precipitation in absolute
ethyl acetate. All products were obtained in good yields (65–90%)
1
and characterized by H NMR spectroscopy and high-resolution
mass spectrometry (ESI,† Fig. S2 and S3).
3.2 Spectroscopic properties
2.5 Cell viability assay
The spectroscopic properties of probes 2–6 were studied in
The viability of NHLF cells following treatment with probes different solvents and reported in Table 1 and the ESI,† Fig. S4.
(2–5) was evaluated using Biotium MTT assay. NHLFs were Probes 6a to 6g exhibited a significantly weaker emission in all
plated in a 96 well plate at a density of 7.3 ꢀ 104 cells per mL in solvents (Table 1) in comparison with probes 2–5. The observed
DMEM with 10% FBS and 1% penicillin/streptomycin and low fluorescence quantum yield in 6 (ffl E 0.001–0.007) can be
allowed to adhere overnight at 37 1C with 5% CO2. Probes, explained by considering the possible high degree of non-
0.14–70 mM, or vehicle control, 0.5% DMSO, were dissolved in radiative relaxation in skeleton 6. In comparison to probe 6a
DMEM with 10% FBS and 1% penicillin/streptomycin. Each (lem E 385 nm and lem E 475 nm in DCM) the attachment of a
treatment concentration was applied in triplicate, with the moderate donor group (i.e., OH) via the para position of the
exception of the vehicle control being applied to 12 wells and skeleton (6b–6d) could significantly enhance the ICT process in
incubated for 24 hours. After 24 hours, the media and the the molecule which leads to a noticeable bathochromic shift
treated probes were aspirated and replaced with DMEM with in the optical spectra of the probes (lex E 450 nm and lem
E
10% FBS and 1% penicillin/streptomycin containing 0.1 mg mLꢁ1 530 nm in DCM). In addition, the attachment of the donor
MTT and incubated for 3.5 hours at 37 1C. The plates were group (–OH) via the ortho position (6f) also exhibited a similar
centrifuged at 1000 ꢀ g for 10 minutes, after which the medium effect. Probe 6e with meta hydroxy connectivity to the benzo-
was carefully aspirated and replaced with MTT solvent, prepared thiazolium moiety did not exhibit a noticeable bathochromic
Table 1 Spectroscopic properties of probes 2–7
Spectroscopic properties in different solvents
Benzene
DCM
DMSO
EtOH
Water
labs
lem (nm)
labs
lem
labs
lem
labs
(nm) (ffl)
lem (nm)
labs
(nm) (ffl)
lem (nm)
e
Dl (cmꢁ1
)
(DCM)
(DCM)
Entry (nm) (ffl)
(nm) (nm) (ffl)
(nm) (nm) (ffl)
(Mꢁ1 cmꢁ1
)
2a
2b
3a
3b
3c
3d
4a
5a
5b
6a
6b
6c
6d
6e
6f
415
423
420
421
422
428
449
442
439
380
428
450
451
391
421
431
438
605 (0.0016) 440
608 (0.17)
526 (0.06)
618 (0.14)
614 (0.13)
614 (0.13)
626 (0.15)
607 (0.19)
612 (0.110)
595 (0.071)
475 (0.0021) 377
527 (0.0038) 424
532 (0.0051) 426
536 (0.0061) 431
510 (0.0012) 385
551 (0.0031) 418
515 (0.0021) 414
418
407
413
414
415
420
450
428
423
609 (0.41)
528 (0.12)
626 (0.45)
625 (0.47)
627 (0.42)
634 (0.52)
641 (0.62)
641 (0.310)
600 (0.085)
471 (0.0040) 376
526 (0.0058) 429
528 (0.0071) 430
531 (0.0141) 438
514 (0.0010) 384
575 (0.0041) 420
519 (0.0032) 418
433
414
426
427
429
435
458
431
430
614 (0.24)
405
402
401
403
405
397
430
424
408
591 (0.0071) 33 624
532 (0.0032) 22 176
6280
4298
6041
5988
5985
5956
4802
5638
5123
4787
3397
3327
3083
5455
4999
3677
6211
495 (0.0020) 429
605 (0.0014) 450
608 (0.0024) 449
612 (0.0022) 449
610 (0.0018) 456
603 (0.0018) 470
588 (0.0015) 455
581 (0.0013) 456
438 (0.0001) 387
505 (0.0021) 447
523 (0.0002) 452
504 (0.0003) 460
500 (0.0006) 396
508 (0.0001) 432
494 (0.0002) 433
636 (0.0027) 445
515 (0.08)
623 (0.28)
620 (0.34)
619 (0.31)
631 (0.41)
635 (0.27)
617 (0.244)
588 (0.036)
464 (0.0015) 370
511 (0.0032) 406
515 (0.0042) 405
518 (0.0052) 415
523 (0.0008) 377
539 (0.0028) 398
507 (0.0011) 405
590 (0.0015) 40 346
599 (0.0027) 42 269
605 (0.0022) 40 952
600 (0.0031) 45 613
620 (0.0032) 43 438
615 (0.0030) 36 193
612 (0.0036) 34 416
452 (0.0006) 32 276
515 (0.0003) 45 486
514 (0.0002) 42 188
520 (0.0004) 54 886
505 (0.0001) 22 141
535 (0.0003) 24 412
510 (0.0001) 39 254
592 (0.0031) 31 424
6g
7
615 (0.11)
424
609 (0.087)
432
608 (0.081)
411
This journal is ©The Royal Society of Chemistry 2019
J. Mater. Chem. B, 2019, 7, 7502--7514 | 7507

View Article Online
Journal of Materials Chemistry B
Paper
shift (lex E 390 nm and lem E 500 nm in DCM) in the optical
spectra as conjugation does not permit donor–acceptor inter-
action (i.e., meta linkage). It is important to notice that probe 6d
(R3 = benzyl substituent) exhibited a relatively higher quantum
yield in all solvents when compared to 6b (R3 = methyl
substituent) which indicated the importance of the substituent
(R3) in the cyanine skeleton with regard to photophysical
characteristics.12 This trend was further consistent with probe
series 3 where higher fluorescence quantum yields were exhibited
by 3d (R3 = benzyl substituent) in comparison to 3a–3c.
One possible explanation for the observed large fluorescence
quantum yield in the presence of the benzyl substituent is the
relatively high stability and less susceptibility towards photo-
fading of the probe.12,33–35
The attachment of the HBT fragment to the phenyl
benzothiazolium (Cy) structure (6a) has significantly improved
the photophysical properties of the probes 2–5. Probes 2–5
exhibited significantly high fluorescence quantum yields
(ffl E 0.1–0.4) in comparison with 6. It is important to notice
that probes 2–3 exhibited a significant bathochromic shift in
the fluorescence emission (lem E 600–620 nm in DCM) due to
ESIPT (HBT fragment) coupled with ICT (cyanine fragment)
(see Scheme 3). Also, it is important to notice that the absorp-
tion spectra of 2–3 did not show a significant bathochromic
shift compared to those of probe 6. This can be explained by
considering the connectivity of the HBT moiety to the cyanine
fragment via the meta-phenylene ring system which does not
permit the extension of the p-conjugation system of the probe.
3.3 Low temperature fluorescence
The large bathochromic shift observed in the emission of 3
(lem B 620–640 nm) in comparison to 6 (lem B 510–530 nm)
could be attributed to the effect of excited state intramolecular
proton transfer (ESIPT) coupled with intramolecular charge
transfer (ICT) (Scheme 3). In order to further evaluate the
content of the ESIPT effect, we decided to freeze 3b in liquid
nitrogen to limit the molecular motion and bond changes that
are associated with the ESIPT process. Thus, the solution of 3b
Fig. 1 (a) Absorption and emission of probe 3b (1 ꢀ 10ꢁ5 M) in different
solvents at room temperature. 3b was excited at 420 nm. (b) Excitation
and emission spectra of 2b and 3b (1 ꢀ 10ꢁ6 M in ethanol) at different
temperatures.
in a quartz tube was quickly cooled down by immersing the spectra were acquired. At an extremely low temperature
sample into liquid nitrogen in a quartz Dewar and optical (at ꢁ189 1C), the molecule was frozen in the solvent matrix,
which showed the emission peak at about 550 nm (Fig. 1b).
When further raising the temperature to room temperature,
a bathochromic shift in the emission spectra was observed. The
observed spectral shift was quite large lem E 72 nm (from
550 nm to B622 nm) in response to the temperature rise. The
excitation spectrum of 3b obtained at ꢁ189 1C was nearly
identical to that at 20 1C (Fig. 1b), showing the excitation
maxima lex = 417 nm. The observed fluorescence spectral shift
at the low temperature was, therefore, related to the rigid
molecular environment in the frozen state, which restricts the
relaxation of the excited state. In other words, ESIPT and ICT
were not fully realized in the frozen solvent matrix, as the bond
reorganization and rotation were restricted. The emission spectra
of 2b at room temperature were found to be lem E 518 nm which
was significantly blue-shifted from the emission spectra of 3b at
Scheme 3 ESIPT coupling with ICT in HBT-Cy 2–4.
ꢁ189 1C (lem E 520 nm). Since the ESIPT process is entirely
7508 | J. Mater. Chem. B, 2019, 7, 7502--7514
This journal is ©The Royal Society of Chemistry 2019

View Article Online
Paper
Journal of Materials Chemistry B
hindered for compound 2b, the result clearly indicates that (lmax = 558 nm). However, probe 3b become strongly fluores-
the impact of the substituent is OH 4 OCH3. Therefore, the cent in aqueous acidic solutions (pH 1–6) due to enhanced
presence of the HBT component in the HBT-Cy skeleton has ESIPT. The calculated pKa for probe 3b was found to be 6.01
significantly improved the probe’s photophysical properties.
(ESI,† Fig. S7). It is also imperative to notice that the emission
intensities of probe 3b remain nearly unchanged in the pH
range of 4–5 which is close to lysosomal inner pH (E4.5).6
3.4 pH sensitivity
The optical spectra of 3b were examined in different aqueous According to the calculated pKa (6.01) for 3b, the probe will not
pH solutions to evaluate the probe’s response (Fig. 2a and ESI,† have significant effect on the pH of the lysosomal lumen (i.e.,
Fig. S5) under different pH conditions. Probe 3b was nearly alkalinizing effect) upon internalization. To further evaluate
non-fluorescent in basic pH solutions (pH 4 9). This can be the role of the phenolic group in pH sensing, 4a was synthe-
attributed to the possible deprotonation of the phenolic proton sized as a model compound with two phenolic groups and
(Ph–OH - PhOꢁ + H+) which eventually hindered ESIPT and spectroscopic studies were conducted (Table 1) in comparison
resulted in the quenching of fluorescence.36,37 Another possible to 3b. In sharp contrast, probe 4a did not show any significant
mechanism is the nucleophilic attack (i.e., oxyanion species) pH response similar to probe 3b (ESI,† Fig. S6) under different
that occurs on the cyanine (CQN+) bond which disrupts the pH conditions. Also, the emission spectra of 4a were signifi-
probe’s conjugation and the ICT process. Under basic pH cantly weaker in all aqueous pH solutions, which was in sharp
conditions a new peak was observed for 3b due to the phen- contrast to 3b. The calculated average pKa for 4a was 6.91 (ESI,†
oxide formation (Ph–OH - PhOꢁ + H+) in aqueous solution Fig. S7). The observed significant difference in dissociation
constants [Ka (4a)/Ka (3b)] E 8 further explains the decreased
acidity of probe 4a. Therefore, the presence of an additional
phenolic group (R4 = OH) in 4a electronically activates the
aromatic system which shows that the dissociation of the
phenol is partially unfavorable. It is important to notice that
the calculated pKa values for probes 3 and 4 are significantly
lower in comparison with basic functional groups such as
trialkyl amines (pKa E 10.75) which are often employed in
commercial LysoTrackers probes. Therefore, HBT-Cy based
probes may not exhibit any alkalinizing effect to elevate cellular
pH upon internalization. Computational calculations were per-
formed by using the density functional theory method (DFT)
by utilizing the B3LYP hybrid function with the 6311G+(2d,p)
basis set to calculate the HOMO–LUMO energy levels for
probes 3 and 5 under solvent-free, acidic (protonated) and
basic (de-protonated) conditions (ESI,† Fig. S25–S27). Probe
3b exhibited relatively low-lying HOMO energies under all
conditions in comparison with 5a. In addition, the LUMOs of
3b and 5a exhibited a higher electron density in the thiazolium
fragment (Cy) indicating that strong electronic interaction
occurs in the systems involving ICT (ESI,† Fig. S24–S26).
In another experiment, 3b was dissolved in acidic water
(pH = 4.5) and spectroscopically titrated against 10% BSA (in
water). A significant fluorescence enhancement (E200%) was
observed in 3b emission upon addition of BSA (Fig. 2b). The pH
of the resulting aqueous acidic solution was measured with
sequential addition of the BSA to confirm that no significant
change in pH occurs during the addition of BSA. The absorp-
tion spectra of the probe 3b showed a significant increase of
the peak intensity at E535 nm which was attributed to the
formation of the phenoxide anion (Ph–OH - PhOꢁ) (ESI†
Fig. S8). However, it is important to rationalize that the acidic
pH environment in the solution (pH E 4.5) may not permit the
partial deprotonation (Ph–OH - PhOꢁ + H+) of the phenolic
group (also see the ESI,† Fig. S5). Therefore, it is possible that
Fig. 2 (a) Emission spectra recorded for probe 3b (1 ꢀ 10ꢁ5 M) in different
aqueous pH solutions at room temperature. Probe 3b was excited at
3b may be internalized into the hydrophobic binding pockets in
420 nm. (b) The emission enhancement observed for 3b (1 ꢀ 10ꢁ5 M) in
BSA where partial deprotonation is plausible. Therefore, the
observed large fluorescence enhancement could be due to the
aqueous acidic (pH 4.5) solution upon sequential addition of 10% BSA at
room temperature. Probe 3b was excited at 420 nm.
This journal is ©The Royal Society of Chemistry 2019
J. Mater. Chem. B, 2019, 7, 7502--7514 | 7509

View Article Online
Journal of Materials Chemistry B
Paper
stabilization of the probe in hydrophobic environments (i.e.,
the hydrophobic pockets in BSA).12,16
3.5 Biological cell studies
The observed interesting photophysical characteristics led us to
investigate the staining ability of probes 2–3 in normal human
lung fibroblast (NHLF) and MO3.13 progenitor oligodendrocyte
cells. Due to the observed significant fluorescence quantum
yield difference in organic and aqueous environments we
assumed that probes 2–3 will be possible candidates for cell
staining under wash-free conditions.11,12,16,38 NHLF cells were
initially incubated with probes at a concentration of 500 nM for
30 minutes and visualized by fluorescence confocal microscopy
under wash-free conditions (no post staining washing step was
carried out prior to imaging). The initial confocal microscopy
images obtained for 2–3 showed a non-uniform staining
pattern throughout the cell with no noticeable background
interference (ESI,† Fig. S9). Further colocalization experiments
were performed in the presence of commercial LysoTrackers
Green DND-26 and LysoTrackers Red DND-99 and probes 2–3
showed exceptional colocalization patterns in both NHLF and
MO3.13 cell lines (Fig. 3 and ESI,† Fig. S10 and S11). The
calculated Mander’s overlap coefficient (40.9 in NHLF in the
presence of commercial LysoTrackers probes) further con-
firmed their exceptional selectivity to lysosomes. These results
further supported our initial hypothesis that isomeric HBT-Cy
versions 2–3 will be excellent markers for cellular lysosome
detection similar to previously reported probe 1. In addition,
probe 2b (R1 = OCH3) produced relatively weak fluorescence
images under the 500 nM concentration range and therefore,
relatively higher concentrations of 2b (1–2 mM) were used for
imaging experiments. This observation can be further explained
by considering relatively lower fluorescence quantum yields
obtained for 2b (in comparison to 2a). Therefore, the ESIPT
enabled HBT-Cy structure (the presence of –OH instead of
–OCH3) showed significant advantages in fluorescence confocal
microscopy. However, it is important to notice that the excellent
colocalization images obtained for probe 2b (calculated Mander’s
Fig. 3 Fluorescence confocal microscopy images obtained for NHLF cells
incubated for 30 minutes with probes 2–3 (500 nM) and LysoTrackers
Green (70 nm). The probes were excited using a 405 nm laser and
LysoTrackers Green was excited using a 488 nm laser line. Note: Lyso-
Trackers Red (70 nM, 561 nm laser) was used for colocalization of probe
2b. The images from right to left show the staining of the commercial
LysoTrackers, the staining of probes 2–3, the overlapped image, and the
bright field image.
overlap coefficient E 0.87) further revealed that the presence of post-staining washing steps to minimize background inter-
the phenolic group was not mandatory for the ‘‘lysosome ferences during imaging. Consequently, the employment of the
selectivity’’ of the probe.
phenolic group in the HBT-Cy probe structure plays an important
To further investigate the validity of this argument, probes role in minimizing background fluorescence. In summary, the
5a and 5b were synthesized by interchanging fluorophore units observed lysosome selectivity in the HBT-Cy probe (2–3) is inde-
across the meta phenylene ring of 2a (Scheme 2). Probes 5a and pendent of the connectivity of its fluorophore fragments benzothia-
5b were further studied in NHLF cells in the presence of zolium cyanine (Cy) and 2-hydroxybenzothiazole (HBT) via a meta
commercial LysoTrackers Green (Fig. 4). Surprisingly, probes phenylene ring system. Yet, the ESIPT enables the skeleton (i.e., 2a)
5a (R1 = OH) and 5b (R1 = OCH3) exhibited excellent colocaliza- to exhibit enhanced photophysical properties (i.e., wash-free
tion (the Mander’s overlap coefficient calculated was 40.9) staining) in comparison to ESIPT disabled 5a (see Table 1).
with Lysotrackers Green in NHLF cells. Therefore, it was
To further investigate the effect of the phenolic groups in the
certain that the presence of the phenolic group would enhance HBT-Cy structure on lysosome selectivity we synthesized probe
the photophysical properties of the probe, however, was not 4a with multiple phenolic components (Scheme 1 and Table 1).
necessary for the lysosome selectivity. Despite the differences in Surprisingly, the presence of multiple phenolic groups (i.e.,
the acidity of the phenolic groups (i.e., 2a vs. 5a), the lysosome probe 4a where R1, R4 = OH) diminish the lysosome selectivity
selectivity of the probe is independent of the possible phenol/ of the probe. The obtained fluorescence confocal microscopy
phenoxide (Ph–OH 2 Ph–Oꢁ + H+) generation. It is also images of 4a in NHLF cells exhibited a bright non-specific
important to notice that probes 2b and 5b (R1 = OCH3) required staining pattern throughout the entire cell (see Fig. 4(a)–(d)).
7510 | J. Mater. Chem. B, 2019, 7, 7502--7514
This journal is ©The Royal Society of Chemistry 2019

View Article Online
Paper
Journal of Materials Chemistry B
However, probe 4a produced a bright red emission observed by the ‘‘key component’’ for the observed lysosome specificity. The
confocal microscopy imaging in NHLF cells with 500 nM fluorescence confocal microscopy images showed a low signal
concentration for 30 minutes. Therefore, further colocalization to noise ratio for probe 6 and it is required to stain it with
images were obtained to investigate the possibility of the higher probe concentrations (1–2 mM range) with post-staining
selective staining of 4a under lower concentrations in NHLF washing steps to minimize the background interference.
cells. However, lowering the 4a concentration up to 100 nM did However, the fluorescence confocal microscopy imaging results
not lead to any selective visualization in live cells. However, the obtained for probe 3 (where the HBT component is connected
presence of multiple phenolic groups improved the fluores- to probe 6) showed exceptional lysosome selectivity and
cence quantum yield of the probe (see Table 1).
excellent photophysical properties. Therefore, the attachment
To further investigate the effect of the benzothiazolium of the 2-hydroxyphenylbenzothiazole (HBT) component to the
component on lysosome selectivity, we synthesized benzothia- probe structure significantly enhanced the photophysical prop-
zolium probes 6a–6g for studying in biological cells. In biolo- erties of the probe to exhibit (1) a large Stokes shift, (2) a high
gical cell studies, probes 6a–6g exhibited a good colocalization fluorescence quantum yield, (3) red-near infra-red emission,
(Mander’s overlap coefficient 40.8–0.7) in the presence of and (4) wash-free application.
commercial Lysotrackers probes (Fig. 5). Nevertheless, the
The lysosome selectivity of the HBT-Cy probe 3 is indepen-
observed colocalization coefficient for probe 6 was noticeably dent of the substituent group (R3) on the cyanine nitrogen atom
lower than those of probes 2–3 due to the observed significant in the probe structure. These results were further consistent with
background fluorescence. This result further revealed that the our previous findings.12 In addition, the lysosome selectivity of
lysosome selectivity of the probe structure was potentially the probes was also independent of the presence of the phenolic
attributed to the ‘‘phenyl benzothiazolium moiety’’ in the group (2a vs. 2b). Due to the positively charged nature of these
HBT-Cy structure. To further validate this hypothesis and to probes, it is also possible for one to assume that the probes may
identify the importance of the benzothiazole fragment with also show some affinity towards cellular mitochondria where
respect to lysosome selectivity, we synthesized and studied 2c many cationic fluorescent probes have been reported for mito-
(see Scheme 1 and Table 1) in live cell experiments in the presence chondrial selectivity.10,38,39 Nevertheless, HBT-Cy probes 2, 3 and
of LysoTrackers Red (ESI,† Fig. S14). Interestingly 2c (Schiff base 5 did not show any mitochondrial localization in colocalization
of 2a) exhibited a remarkable lysosome selectivity confirming the experiments (ESI,† Fig. S12). Further colocalization experiments
lysosome selectivity arising from the ‘‘benzothiazolium (Cy)’’ were conducted for probe 3 in the presence of commercial
fragment and not from the benzothiazole (HBT) component. ER-TrackerTM Red (ESI,† Fig. S23) in NHLF cells and the probe
As a control experiment, probe 7 was synthesized and tested in did not show any colocalization.
live cells to investigate its lysosome visualization ability. In
MTT cell proliferation viability assay was performed to
sharp contrast, probe 7 did to show any lysosome specificity or investigate the cytotoxicity of probes 2, 3 and 5 in NHLF cells
internalization into live cells. Therefore, certainly the phenyl (ESI,† Fig. S16). The probes exhibited an excellent biocompat-
benzothiazolium moiety (Cy) in the HBT-Cy structure acts as ibility where the calculated LC50 was 4 25 mM. In comparison
to the working concentration of these probes (i.e., 0.5 mM), the
calculated lethal concentration of the probes was significantly
higher. Thus, 2, 3 and 5 will be highly biocompatible and
selective fluorescent probes for visualizing cellular lysosomes
without perturbing cellular activities. Probe 3d was further
utilized to investigate the possibility of detecting the pH changes
in cellular environments. NHLF cells were incubated with different
pH buffer solutions (4.5, 7.4 and 10.1) in the presence of probe
3d (500 nM) and fluorescence confocal microscopy images were
obtained (ESI,† Fig. S21). The fluorescence microscopy images
were statistically analyzed to investigate their ability to detect
pH changes quantitatively. Interestingly, 3d showed the highest
fluorescence emission under physiological pH conditions
(pH = 7.4) in comparison to relatively acidic (pH = 4.5) or basic
(pH = 10.1) pH conditions. This result indicated the possibility
of utilizing probe 3 as a potential fluorescent marker to identify
pH fluctuations in biological environments. Probe 3d was
further investigated in NHLF cells to assess its ability to be used
as a stable fluorescent marker to visualize cells under stress
conditions. NHLF cells were incubated with probe 3d (500 nM)
under normal (cells media + FBS) and stress (cell media with no
FBS) conditions for 2 hours and fluorescence microscopy images
were obtained (ESI,† Fig. S22). However, probe 3d did not show
Fig. 4 Fluorescence confocal microscopy images obtained for NHLF cells
incubated for 30 minutes with probes 4–5 (500 nM) and LysoTrackers
Green (70 nm). The probes were excited using a 405/561 nm laser and
LysoTrackers Green was excited with a 488 nm laser line. The images from
right to left show the staining of the commercial LysoTrackers, staining of
probes 4–5, the overlapped image, and the bright field image.
This journal is ©The Royal Society of Chemistry 2019
J. Mater. Chem. B, 2019, 7, 7502--7514 | 7511

View Article Online
Journal of Materials Chemistry B
Paper
conditions was further confirmed by performing colocalization
experiments in the presence of LysoTrackers Green DND-26
(ESI,† Fig. S24).
3.6 Long-term imaging ability and photostability evaluation
Probes 2–3 were identified as the most promising lysosome
selective fluorescent probes in the reported series. Further
experiments were performed to evaluate the long-term imaging
ability and the photostability of the probes. NHLF cells were
incubated with probe 3d (500 nM) for 30 minutes and fluores-
cence confocal microscopy images were analyzed in 30 minute
time intervals up to 4 hours (Fig. 6a and ESI,† Fig. S18).
Interestingly, probe 3d generated stable localization patterns
within the cell up to 4 hours without generating any noticeable
background interferences or morphological changes. The
normalized fluorescence intensity of the probe 3d emission in
live cells was plotted as a function of imaging time (Fig. 6b).
These results further illustrated the inherent biocompatibility
and high photostability of the probe in biological environments
though the probe is excitable with a relatively high energy
(405 nm) laser line. To further investigate the photostability
and possible phototoxicity of these probes, NHLF cells were
Fig. 5 Fluorescence confocal microscopy images obtained for NHLF cells
incubated for 30 minutes with probes 6a–6g (1 mM) and LysoTrackers Red
(70 nm). The probes were excited with a 405 nm laser and LysoTrackers
Red was excited with a 561 nm laser line. The images from right to
left show the staining of the commercial LysoTrackers, staining of probes
6a–6g, the overlapped image, and the bright field image.
any noticeable changes in the staining pattern or the calculated
fluorescence intensity which indicates its possibility to stain
cellular lysosomes under different stress conditions. To further
evaluate the probe’s intrinsic properties to stain cellular lyso-
somes under different stress conditions and fluctuations,
NHLF cells were initially incubated with 1 mM carbonyl
cyanide-p-trifluoromethoxyphenylhydrazone (FCCP) for 1 hour.
Then the cells were further incubated with probe 3d for
30 minutes and analyzed via fluorescence confocal microscopy.
FCCP acts as a mitochondrial membrane potential inhibitor
which disrupts ATP synthesis by acting as an uncoupler. Inter-
estingly, probe 3d generated excellent confocal microscopy
images in the presence of FCCP which further confirms the
ability of probe 3 to be used as a potential fluorescent marker to
detect cellular lysosomes under different stress conditions
(ESI,† Fig. S23). The probe’s lysosome selectivity under stress
Fig. 6 (a) Long-term imaging ability assessment for probe 3d in NHLF
cells. The cells were incubated with 3d (500 nM) for 30 minutes and
confocal microscopy images were obtained at different time intervals
under consistent parameters of the microscope. Probe 3d was excited
using a 405 nm laser and the emissions were collected in the 570 nm to
700 nm range. (b) The averaged normalized fluorescence intensity
obtained for probe 3d at different time intervals.
7512 | J. Mater. Chem. B, 2019, 7, 7502--7514
This journal is ©The Royal Society of Chemistry 2019

View Article Online
Paper
Journal of Materials Chemistry B
4. Conclusions
In summary, a series of HBT-Cy fluorescent probes were
synthesized in high yields. HBT-Cy probes 2a and 3 exhibited
excellent fluorescence quantum yields (ffl E 0.1–0.5) and a
large Stokes shift (Dl 4130 nm) due to excited state intra-
molecular proton transfer coupled with intramolecular charge
transfer. Compared to the simple benzothiazolium version (Cy)
6, probes 2–3 exhibited excellent photophysical characteristics
due to the attachment of the 2-(20-hydroxyphenyl)benzothiazole
(HBT) fragment to the benzothiazolium (Cy) moiety. The
substituent on the cyanine nitrogen atom (R3) affected the
fluorescence quantum yield of the compounds and higher
fluorescence quantum yields were obtained for probes with
bulky groups (R3 = PhCH2) on cyanine nitrogen. The low
temperature results proved that the HBT group plays an impor-
tant role in the large Stokes shift due to ESIPT. Probe 3
exhibited significant sensitivity towards pH changes in aqueous
solutions. Thus, the calculated pKa values (pKa o 7) further
confirmed the non-alkalinizing properties of the probes. Probes
2, 3 and 5 showed excellent lysosome selectivity in live cell
imaging. The lysosome selectivity of probes 2–3 was indepen-
dent of the (1) substituent (R3) on the cyanine nitrogen atom,
(2) connectivity order of the cyanine and HBT fragments via the
meta phenylene ring, and (3) presence/absence of the phenolic
group. However, the attachment of a single phenolic group
significantly improved the photophysical properties of the
probe (i.e., 2a vs. 2b). The attachment of multiple phenol
groups in the probe structure diminished the lysosome selec-
tivity of the probe (i.e., probe 4a). The experimental results
indicated that the phenyl benzothiazolium (Cy) moiety 6 is the
key component for the observed lysosome selectivity. Probes 2a,
3, and 5a produced excellent fluorescence confocal images of
the cellular lysosomes under wash-free conditions with a high
signal to noise ratio. Probes 2, 3 and 5 exhibited high biocom-
patibility (LC 50 4 25 mM) for live cell applications. Probe 3
exhibited excellent photostability and long-term imaging ability
during the fluorescence confocal microscopy. In addition probe
3 was successfully utilized to visualize cellular lysosomes under
different stress conditions. The attachment of the 2-hydroxy-
phenylbenzothiazole (HBT) fragment to the probe structure
significantly enhanced the photophysical characteristics of
the probes (2–3) including (1) wash free application, (2) bright
fluorescence, (3) high stability, (4) large Stokes shift, (5) high
photostability and (6) high biocompatibility. The HBT skeleton
7 did not exhibit any lysosome specificity in the absence of the
benzothiazolium (Cy) component. Therefore, this new HBT-Cy
based lysosome probe design would be a superior replacement
for existing commercial LysoTrackers probes for cellular appli-
cations under non-alkalinizing conditions.
Fig. 7 (a) Fluorescence confocal microscopy images obtained for probe
3d (500 nM) in NHLF cells upon continuous irradiation with a 405 nm laser
line (laser power percentage 3.0; digital zoom = 1; pinhole = 1AU; master
gain = 150; digital offset = 0). Fluorescence confocal microscopy images
were obtained at 20 s time intervals. (b) The plot of recovered fluorescence
intensity versus irradiation time calculated for 3b (500 nM) and 3d (500 nM).
incubated with 3b and 3d (500 nM) for 30 minutes and
continuous irradiation was obtained using a 405 nm laser
for 4 minutes. Fluorescence confocal microscopy images were
obtained at 20 s time intervals during the experiment (Fig. 7 and
ESI,† Fig. S19, S20). The recovered fluorescence intensity (%) was
plotted as a function of irradiation time (Fig. 7b). The recovered
fluorescence intensity analysis further revealed that the possible
photobleaching of the probe upon continuous irradiation is
minimal under experimental conditions. Therefore, these probes
will be reliable candidates for imaging cellular lysosomes via
fluorescence confocal microscopy. However, it is also important
to notice that probe 3d exhibited relatively more stability towards
photobleaching in comparison to 3b. This result also suggested
that the substituent on the benzothiazolium segment had a large
impact on the photostability of the probe.12 The observed
relatively lower photostability of 3b was due to the possible
photodegradation of the probe upon interaction with molecular
oxygen singlet oxygen (1O2) or the superoxide anion (O2ꢁ) pre-
Author contributions
sent in the biological environments.33,34 These results further C. S. A. planned, conducted, and summarized the synthesis and
suggested that the relative electron density on the cyanine N characterization of probes. C. S. A., K. A. B. and L. J. M. planned
atom can affect its photostability.12,34
and performed cell imaging experiments. C. S. A. and D. D.
This journal is ©The Royal Society of Chemistry 2019
J. Mater. Chem. B, 2019, 7, 7502--7514 | 7513

View Article Online
Journal of Materials Chemistry B
Paper
15 N. B. Yapici, Y. Bi, P. Li, X. Chen, X. Yan, S. R. Mandalapu,
M. Faucett, S. Jockusch, J. Ju, K. M. Gibson, W. J. Pavan and
L. Bi, Sci. Rep., 2015, 5, 8576.
16 K. A. Bertman, C. S. Abeywickrama, H. J. Baumann,
N. Alexander, L. McDonald, L. P. Shriver, M. Konopka and
Y. Pang, J. Mater. Chem. B, 2018, 6, 5050–5058.
17 J. P. Luzio, B. A. Rous, N. A. Bright, P. R. Pryor, B. M. Mullock
and R. C. Piper, J. Cell Sci., 2000, 113(Pt 9), 1515–1524.
18 R. G. W. Anderson and L. Orci, J. Cell Biol., 1988, 106,
539–543.
performed the photophysical characterization of the probes.
N. A. conducted and summarized the mass spectra for the
probes. H. J. B. conducted the cell viability experiments. C. R. S.
performed computational calculations. Y. P. supervised the
project. C. S. A and Y. P. wrote the manuscript.
Conflicts of interest
There are no conflicts to declare.
19 W. Zhu, X. Zheng, Y. Huang, Z. Lu and H. Ai, Sci. China:
Chem., 2018, 1–7.
Acknowledgements
20 R. P. Haugland, The handbook: a guide to fluorescent probes
and labeling technologies, Molecular probes, 2005.
21 M. Fang, R. Adhikari, J. Bi, W. Mazi, N. Dorh, J. Wang,
N. Conner, J. Ainsley, T. G. Karabencheva-Christova, F.-T. Luo,
A. Tiwari and H. Liu, J. Mater. Chem. B, 2017, 5, 9579–9590.
22 Y. Han, M. Li, F. Qiu, M. Zhang and Y.-H. Zhang, Nat.
Commun., 2017, 8, 1307.
The NHLF cell line was a kind gift from Dr Sailaja Parachuri
at the University of Akron. We thank Dr Leah Shriver at the
University of Akron for providing the MO.3.13 cell line and cell
culture facility. We also thank Prof. Robert Stahelin (Purdue
University) and Dr Kaveesha Wijesinghe (University of Notre
Dame) for valuable discussions and guidance.
23 K. A. Bertman, C. S. Abeywickrama, A. Ingle, L. P. Shriver,
M. Konopka and Y. Pang, J. Fluoresc., 2019, 1–9.
24 J. Ou-Yang, W.-L. Jiang, K.-Y. Tan, H.-W. Liu, S.-J. Li, J. Liu,
Y.-F. Li and C.-Y. Li, Sens. Actuators, B, 2018, 260, 264–273.
25 J. Ou-Yang, Y.-F. Li, P. Wu, W.-L. Jiang, H.-W. Liu and
C.-Y. Li, ACS Sens., 2018, 3, 1354–1361.
References
1 C. de Duve and R. Wattiaux, Annu. Rev. Physiol., 1966, 28,
435–492.
2 S. Piao and R. K. Amaravadi, Ann. N. Y. Acad. Sci., 2016,
1371, 45–54.
26 X. Zhang, B. Wang, C. Wang, L. Chen and Y. Xiao, Anal.
Chem., 2015, 87, 8292–8300.
3 B. Alberts, D. Bray, K. Hopkin, A. Johnson, J. Lewis, M. Raff,
K. Roberts and P. Walter, Essential cell biology, Garland 27 Q. Wan, S. Chen, W. Shi, L. Li and H. Ma, Angew. Chem., Int.
Science, 2013. Ed., 2014, 53, 10916–10920.
4 J. P. Luzio, Y. Hackmann, N. M. G. Dieckmann and 28 H. Lim, H. Park and H. P. Kim, Biochem. Pharmacol., 2015,
G. M. Griffiths, Cold Spring Harbor Perspect. Biol., 2014,
6(9), a016840.
96, 337–348.
¨¨
¨
29 L. Groth-Pedersen and M. Jaattela, Cancer Lett., 2013, 332,
5 U. Repnik, V. Stoka, V. Turk and B. Turk, Biochim. Biophys.
Acta, Proteins Proteomics, 1824, 2012, 22–33.
265–274.
¨¨
¨
30 G. Kroemer and M. Jaattela, Nat. Rev. Cancer, 2005, 5, 886.
6 J. P. Luzio, P. R. Pryor and N. A. Bright, Nat. Rev. Mol. Cell 31 P. Jiang and N. Mizushima, Cell Res., 2014, 24, 69–79.
´
Biol., 2007, 8, 622.
32 A. Pierzynska-Mach, P. A. Janowski and J. W. Dobrucki,
7 A. Thorburn, D. H. Thamm and D. L. Gustafson, Mol.
Pharmacol., 2014, 85, 830–838.
Cytometry, Part A, 2014, 85, 729–737.
33 C. Chen, B. Zhou, D. Lu and G. Xu, J. Photochem. Photobiol.,
A, 1995, 89, 25–29.
¨¨
¨
8 T. Kirkegaard and M. Jaattela, Biochim. Biophys. Acta, Mol.
Cell Res., 1793, 2009, 746–754.
9 N. Fehrenbacher, N. Fehrenbacher, M. Jaattela and
34 X. Chen, X. Peng, A. Cui, B. Wang, L. Wang and R. Zhang,
J. Photochem. Photobiol., A, 2006, 181, 79–85.
35 S. J. Formosinho and L. G. Arnaut, J. Photochem. Photobiol.,
A, 1993, 75, 21–48.
¨¨
¨
¨¨
¨
M. Jaattela, Cancer Res., 2005, 65, 2993–2995.
10 G. Y. Wiederschain, Biochemistry, 2011, 76, 1276.
11 D. Dahal, L. McDonald, X. Bi, C. Abeywickrama, F. Gombedza, 36 C. S. Abeywickrama and Y. Pang, Tetrahedron Lett., 2017, 58,
M. Konopka, S. Paruchuri and Y. Pang, Chem. Commun., 2017,
53, 3697–3700.
12 C. S. Abeywickrama, K. J. Wijesinghe, R. V. Stahelin and
Y. Pang, Chem. Commun., 2019, 55(24), 3469–3472.
13 B. Wang, S. Yu, X. Chai, T. Li, Q. Wu and T. Wang,
Chem. – Eur. J., 2016, 22, 5649–5656.
1627–1632.
37 C. S. Abeywickrama and Y. Pang, Tetrahedron Lett., 2016, 57,
3518–3522.
38 C. S. Abeywickrama, K. J. Wijesinghe, R. V. Stahelin and
Y. Pang, Sens. Actuators, B, 2019, 285, 76–83.
39 C. S. Abeywickrama, H. J. Baumann, N. Alexander, L. P. Shriver,
M. Konopka and Y. Pang, Org. Biomol. Chem., 2018, 16,
3382–3388.
14 X. Zhang, C. Wang, Z. Han and Y. Xiao, ACS Appl. Mater.
Interfaces, 2014, 6, 21669–21676.
7514 | J. Mater. Chem. B, 2019, 7, 7502--7514
This journal is ©The Royal Society of Chemistry 2019