Fluorescence sensing of caffeine in aqueous solution with carbazole-based probe and imaging application in live cells

Article abstract of DOI:10.1016/j.bmcl.2012.07.055

The host-guest interaction of caffeine in aqueous solutions has been achieved by using carbazole based imino-phenol receptors with oxopurines influences their fluorescence properties and can be exploited for 'turn-ON' fluorescence sensing of caffeine. Thi

Full text of DOI:10.1016/j.bmcl.2012.07.055

Bioorganic & Medicinal Chemistry Letters 22 (2012) 5379–5383
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Fluorescence sensing of caffeine in aqueous solution with carbazole-based
probe and imaging application in live cells
Ajit Kumar Mahapatra a, , Jagannath Roy a, Prithidipa Sahoo a, Subhra Kanti Mukhopadhyay b,
⇑
Anindita Roy Mukhopadhyay c, Debasish Mandal d
a Department of Chemistry, Bengal Engineering and Science University, Shibpur, Howrah 711103, India
b Department of Microbiology, The University of Burdwan, Burdwan, West Bengal, India
c Department of Microbiology, MUC Womens College, West Bengal, India
d Department of Spectroscopy, India Indian Association for The Cultivation of Science, Jadavpur Kolkata 700032, India
a r t i c l e i n f o
a b s t r a c t
Article history:
The host–guest interaction of caffeine in aqueous solutions has been achieved by using carbazole based
imino-phenol receptors with oxopurines influences their fluorescence properties and can be exploited for
‘turn-ON’ fluorescence sensing of caffeine. This new fluorescent probe with high sensitivity and selectiv-
ity for detection and first time imaging of caffeine in living cells was developed in aqueous media.
Ó 2012 Elsevier Ltd. All rights reserved.
Received 29 May 2012
Revised 28 June 2012
Accepted 13 July 2012
Available online 21 July 2012
Keywords:
Caffeine
Carbazole
‘Turn-ON’ fluorescent probe
Xanthine alkaloid
Candida albicans
Caffeine (3) is a widely consumed alkaloidal drug throughout
the world.1 The traditional significant sources of caffeine in our
daily lives are coffee, black tea, guarana paste, cola nuts, cocoa
beans etc. The alkaloid is also an ingredient of cola beverages
and energy drinks.2 While caffeine is regarded to be harmless for
adults in general, there are severe concerns about unfavorable
influences of caffeine on young children and pregnant women,
including the risk of fetal death.3 So the extensive and diverse con-
sumption of caffeine requires further analytical tools to be devel-
oped for enhanced product safety. For caffeine quantitation,
standard chromatographic methods are available such as (HP)TLC,
HPLC-UV, LC, MIPs and CE.4 However all the above methods re-
quire a long period of time, so methods requiring shorter time
for analysis is needed. This area is left to chemosensors5, which
represent a vivid field in supramolecular chemistry. Several re-
search groups6–9 including our group10–13 accomplished the syn-
thesis of artificial caffeine receptors. All these H-bond receptors
were only investigated in organic solvents; a significant selectivity
and high affinity for the target molecule in aqueous media will be
of particular interest for a promising application in caffeine
detection.
Caffeine recognition in water has been studied with zinc based
porphyrin peptide type receptor.14 The another receptor cucur-
bit[7]uril containing macrocyclic host is able to bind caffeine in
water using hydrophobic effects.15 As part of our ongoing research
toward the development of fluorophores for xanthine alkaloids,10–
13 here, we report that hydrogen bonding interactions between caf-
feine and carbazole based imino-phenol can be used for the fluori-
metric detection of caffeine in water. To the best of our knowledge,
this is the first report, where we explored the possibility of using
carbazole based chemosensor (1) for the in vivo recognition of
the caffeine in living cells using confocal imaging experiments.
It has long been known that caffeine interacts with polypheno-
lic molecules and these interactions influences the texture, the fla-
vor, and the physiological effects of caffeine-containing beverages.
So, our design of a water-soluble carbazole fluorosensor for caf-
feine possessing two imino-phenol moieties that have an affinity
towards the ring oxygen and most basic imidazole ‘N’ of the caf-
feine. We chose carbazole as the signal transduction unit because
of its chemical stability, large Stokes shift and high fluorescence
quantum yield.16
Carbazole and its derivatives are an important type of nitrogen
containing aromatic heterocyclic compounds, and it is present in a
variety of naturally occurring medicinally active substances.17
These characteristics result in the extensive potential applications
of carbazole-based derivatives in the field of medicinal chemistry
⇑
Corresponding author. Tel.: +91 33 2668 4561; fax: +91 33 2668 4564.
E-mail address: mahapatra574@gmail.com (A.K. Mahapatra).
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2012.07.055

5380
A. K. Mahapatra et al. / Bioorg. Med. Chem. Lett. 22 (2012) 5379–5383
(antitumor, antimicrobial, antihistaminic, antioxidative, anti-
inflammatory, psychotropic agents etc.).18 These characteristics
prompted us to investigate carbazole-based imino-phenol as a po-
tential probe for the detection of caffeine in living cell.
Compound 2 was synthesized from carbazole in two steps
according to our previous procedure.19,20 Compound 1 was synthe-
sized (Scheme 1) by condensation of N-butyl- carbazole-dicarbal-
dehyde 2 and ortho-aminophenol in high yield, and its structure
has been proved by various spectroscopic characterizations can
be found in Supplementary data Figures S1–S5.
The interaction of probe 1 with an aqueous solution of caffeine
was investigated by spectrophotometric and spectrofluorimetric
titrations in buffered aqueous solution (pH 7.2, 25 mM HEPES buf-
fer) as shown in Figure 1.
During the photometric titration of caffeine to probe 1, a
absorption band at 275 nm was gradually enhanced, while the lit-
tle intensity of absorption at 343 nm was decreased correspond-
ingly. A clear isosbestic point was observed at 299 nm that
indicate that one complex structure is formed. The data of the pho-
tometric titration were employed to determine the association
1 was designed in such way that in which both the optical reporter
carbazole and the caffeine binding imino-phenol part were water
soluble, biocompatible and stable in buffer solution. So the sensi-
tivity of probe 1 for caffeine was verified in living cells, indicating
its potential for application in intracellular imaging of caffeine.
To understand the nature of interaction between probe 1 and
caffeine, the fluorescence variation of probe 1 was measured upon
the addition of caffeine with buffers adjusted different pH medium.
Furthermore, to examine whether the acidic OH groups are plays
crucial role during binding event of caffeine, we studied the inter-
action of caffeine with salt of probe 1. Fluorescence titration exper-
iments suggest that the emission intensity was affected (Fig. S10).
In order to understand the binding potencies and sensing selectiv-
ities of probe 1, parallel investigations were carried out with a vari-
ety of biologically relevant analytes such as the physiologically
important xanthine derivatives theophylline and theobromine.
Figure 3 displays the changes in fluorescence of probe 1 upon
addition various xanthine derivatives. As in the case of caffeine, a
marked change of fluorescent emission of probe 1 was observed,
but other xanthine derivatives showed negligible effect (however,
addition of theobromine leads to a slight increasing). Accordingly,
somewhat lower binding constants were obtained from fluores-
cence titration experiments (Ka = 148 6 MÀ1for theobromine
and Ka = 113 2 MÀ1 for theophylline). None of the other tested
analytes gave a pronounced fluorescence response (Fig. 3).This re-
sult demonstrated that probe 1 exhibited high selectivity towards
caffeine in water.
constant for the complex of probe
1
with caffeine
(1.75 0.36) Â 103 MÀ1.The interaction of probe 1 with caffeine
was also studied by fluorescence spectroscopy. The emission spec-
tra are shown in Figure 2. When caffeine was added to a buffered
aqueous solution (pH 7.2, 25 mM HEPES buffer) of probe 1
(5 lM), a pronounced increase in fluorescence intensity was ob-
served (kmax = 431 nm, Fig. 2, inset). The inset shows a linear rela-
tionship, which is always important for easy and accurate analysis.
A fluorescence titration experiment with caffeine concentrations
between 0 and 50 mM was then performed. Judging from the titra-
tions, continuous variation method was used to determine the stoi-
chiometric ratios of the host and the specific guest, which was
found to be a 1:1 probe (1)-to-caffeine complexation (Fig. S6).
For a complex of 1:1 stoichiometry, the association constant
Ka = (1.56 0.31) Â 103 MÀ1 could be determined by non-linear
fitting analyses of the titration curves (Figs. S7–S8) according to
literature report.21
In order to know more about interactions between caffeine and
probe 1, 1H NMR titration was also performed in CDCl3 and the re-
sults are shown in Figure 4. The phenolic proton and N@CH (imine)
proton displayed down field shift with increasing addition of caf-
feine by 0.55 and 0.15 ppm respectively, indicating the presence
of hydrogen bond interaction between acidic OH together with
CH and caffeine. In contrast, other hydrogens in the carbazole ring
were affected very small. These observations suggested that only
the phenolic OH proton of probe 1 participated to complex with
caffeine.
The probe 1 presents some excellent advantages compared to
known fluorescent sensor for caffeine22 due to ‘turn-ON’ fluores-
cence response. The ‘turn-ON’ aqueous sensors has a number of
advantages such as (a) it reduces the chance of a false positive, ob-
served in some turn-off probes, (b) allows for the use of multiple
probes, selective for different analytes, and (c) is applicable in
the analysis of both, aqueous environmental and biological sam-
ples. The probe 1 itself is weak emissive because of rapid C@N
isomerisation. The C@N isomerisation is the predominant decay
process23 of the excited state. The significant enhancement in fluo-
rescence (10 times at 431 nm) is probably caused by the formation
of a1:1 complex of probe 1 with caffeine in which the rotation of
acyclic C@N isomerization is prevented upon caffeine binding
and hence fluorescence enhancement occurs (Fig. S9). The probe
Binding of caffeine by probe 1 has also been investigated by
quantum chemical calculations at the TDDFT level. The energy
minimized structure of the caffeine–probe 1 complex and their
HOMOÀLUMO energy gap were computed by using Gaussian
2003(B3LYP/6-31G(d,p)).24,25 The result indicates that the most
favourable geometry is found for caffeine–probe 1 complex in Fig-
ure 5a, which consists of two hydrogen bonds with bond distances
1.89 and 1.84 Å respectively (see Supplementary data). The
trons on the HOMO of caffeine–probe 1 complex is mainly located
on the whole -conjugated carbazole framework (excluding the
p elec-
p
butyl group), but the LUMO is mostly positioned at the center of
the guest caffeine. Moreover, the HOMOÀLUMO energy gap of
complex become smaller relative to that of probe 1 in Figure 5b.
The energy gaps between HOMO and LUMO in the probe 1 and caf-
feine complex were 8.6832 eV and 8.3792 eV respectively (see
Supplementary data).
To further explored the ability of the probe 1 to image caf-
feine in living cells (Fig. 6), we carried out experiments in live
Candida albicans cells. The cells were pre-treated with probe 1
OH
HO
OHC
CHO
2-aminophenol
EtOH, reflux, 5h
N
N
N
R
(10
l
M in 0.01 M phosphate buffer, pH 7.4) at 37 °C for 30 min
M) for
and then were incubated with caffeine (initially
5
l
N
R
2
10 min the cells displayed intracellular fluorescence (Fig. 6b),
indicating the ability of the probe can penetrate the cell
membrane. The cells also exhibited more intense fluorescence
when more caffeine was introduced onto the cells externally,
and fluorescence responses increase with the increase in caffeine
concentration, which could be evident from the cellular imaging
(Fig. 6c–f). Moreover, the cells treated with various concentra-
tions of probe 1 for up to 2 h, the result showed no significant
R = C4H9
1
O
CH3
N
H3C
O
N
N
N
CH3
3
Scheme 1. Synthesis of Probe 1 and structure of Caffeine 3.

A. K. Mahapatra et al. / Bioorg. Med. Chem. Lett. 22 (2012) 5379–5383
5381
Figure 1. (a) UV–vis spectra of probe 1 (c = 1.0 Â 10À5 M) (5
lM) with the increasing concentrations of Caffeine (0–2 equiv) in pH 7.2 HEPES buffer (25 mM , pH 7.2,
containing 1.0% methanol). (b) Enhancement of UV–Vis spectra between 300 and 450 nm.
Figure 2. (a) Fluorescence spectra of probe 1 (5
l
M) (c = 1.0 Â 10À5 M) with the increasing concentrations of Caffeine (0–50 mM) (c = 5.0 Â 10À4 M) in pH 7.2 HEPES buffer
(25 mM, pH 7.2, containing 1.0% methanol). (b) The fluorescence intensity changes at 431 nm of probe 1 (5
l
M) (c = 1.0 Â 10À5 M) with the amount of Caffeine
(c = 5.0 Â 10À4 M).
Figure 3. Bar diagram showing interaction of probe 1 (5
l
M) (c = 1.0 Â 10À5 M)
with the increasing concentrations of various xanthine derivatives (0–50 mM)
(c = 5.0 Â 10À4 M) in pH 7.2 HEPES buffer (25 mM, with tested pH 7.2, containing
1.0% methanol) at 431 nm.
Figure 4. 1H NMR (300 MHz) spectra of Compound 1 (a), 1 + .5 equiv Caffeine (b),
1 + 1 equiv Caffeine (c), and 1 + 1.2 equiv Caffeine in CDCl3(d).
cell death in 2 h incubation, indicating 1 was of low toxicity or
non-toxic26 to cultured cells under the experimental conditions.
The results demonstrate for the first time that probe 1 is poten-
tially useful for monitoring caffeine in living organisms.
In conclusion, we present the carbazole based imino-phenol 1
as a promising fluorescent probe for xanthine alkaloids caffeine
detection. Probe 1 exhibits a highly sensitive and selective fluores-
cence enhancement toward the caffeine over other xanthine
related alkaloids in aqueous buffer (pH 7.2). It represents one of
the few fluorescent probes that allow a selective ‘turn-ON’ detection
of caffeine in water. Probe 1 was successfully expressed in cells,
which demonstrates its potential usefulness as a molecular probe
in biological systems.

5382
A. K. Mahapatra et al. / Bioorg. Med. Chem. Lett. 22 (2012) 5379–5383
Figure 5. (a) Energy-minimized structure of the Probe 1-Caffeine complex (b) HOMO and LUMO of Probe 1 and Probe1-Caffeine Complex.
Figure 6. Fluorescence and brightfield images of living cells (a) images of Candida albicans cells + probe 1 (10
lM) (b) images of cells + probe 1 + caffeine (5.0
lM), and
fluorescence images (c–f) of Candida albicans cells incubated with 10 lM, 20 lM, 30 lM, 40 lM of probe 1 for 25 min, respectively.
4. Hurst, W. J.; Martin, R. A., Jr.; Tarka, S. M., Jr. In Caffeine; Spiller, G. A., Ed.; CRC
Press: Boca Raton, 1998; p 13.
Acknowledgment
5. Czarnik, A. W. In Advances in Supramolecular Chemistry; Gokel, G. W., Ed.; JAI
Press: Greenwich, Conneticut, 1993; Vol. 3, p 131.
6. Siering, C.; Kerschbaumer, H.; Nieger, M.; Waldvogel, S. R. Org. Lett. 2006, 8,
1471.
7. Bomkamp, M.; Siering, C.; Landrock, K.; Stephan, H.; Fröhlich, R.; Waldvogel, S.
R. Chem. -Eur. J. 2007, 13, 3724.
8. Siering, C.; Beermann, B.; Waldvogel, S. R. Supramol. Chem. 2006, 18, 23.
9. Siering, C.; Grimme, S.; Waldvogel, S. R. Chem. -Eur. J. 1877, 2005, 11.
10. Goswami, S.; Mahapatra, A. K.; Mukherjee, R. J. Chem. Soc., Perkin Trans. 2001, 1,
2717.
11. Mahapatra, A. K.; Sahoo, P.; Goswami, S.; Fun, H.-K.; Yeap, C. S. J. Inclusion
Phenom. Macrocyclic Chem. 2010, 67, 99.
12. Mahapatra, A. K.; Sahoo, P.; Goswami, S.; Chantrapromma, S.; Fun, H.-K.
Tetrahedron Lett. 2009, 50, 89.
13. Mahapatra, A. K.; Sahoo, P.; Goswami, S.; Fun, H.-K. J. Mol. Struct. 2010, 693, 63.
14. Fiammengo, R.; Crego-Calama, M.; Timmerman, P.; Reinhoudt, D. N. Chem. -
Eur. J. 2003, 9, 784.
We thank CSIR [Project No. 01(2460)/11/EMR-II] for financial
support. We also acknowledge DST, AICTE, UGC-SAP and MHRD
for funding instrumental facilities in our department. We are also
thankful to the Annesur Rahaman Center for High Performance
Computing, IACS, for providing us with computational time.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2012.
07.055.
References and notes
15. Wu, J.; Isaacs, L. Chem. -Eur. J. 2009, 15, 11675.
16. Jeon, S. O.; Lee, J. Y. Tetrahedron 2010, 66, 7295.
17. Zhang, F.-F.; Gan, L.-L.; Zho, C.-He. Bioorg. Med. Chem. Lett. 1881, 2010, 20.
18. Knolker, H.-J.; Reddy, K. R. Chem Rev. 2002, 102, 4303.
19. Mahapatra, A. K.; Roy, J.; Sahoo, P.; Mukhopadhyay, S. K.; Chattopadhyay, A.
Org. Biomol. Chem. 2012, 10, 2231.
1. Waldvogel, S. R. Angew. Chem., Int. Ed. 2003, 42, 604.
2. Parliment, T. H.; Ho, C.-T.; Schieberle, P. Caffeinated Beverages; American
Chemical Society: Washington, 2000.
3. Bech, B. H.; Nohr, E. A.; Vaeth, M.; Henriksen, T. B.; Olsen, J. Am. J. Epidem. 2005,
162, 983.
20. Mahapatra, A. K.; Roy, J.; Sahoo, P. Tetrahedron Lett. 2011, 52, 2965.

A. K. Mahapatra et al. / Bioorg. Med. Chem. Lett. 22 (2012) 5379–5383
21. Bourson, J.; Pouget, J.; Valeur, B. J. Phys. Chem. 1993, 97, 4552. 24. Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785.
5383
22. Rochat, S.; Steinmann, S. N.; Corminboeuf, C.; Severin, K. Chem. Commun. 2011,
47, 10584.
23. Wu, J.; Liu, W.; Ge, J.; Zhang, H.; Wang, P. Chem. Soc. Rev. 2011, 40, 3483.
25. Miehlich, B.; Savin, A.; Stoll, H.; Preuss, H. Chem. Phys. Lett. 1989, 157, 200.
26. Ratha, J.; Majumdar, K. A.; Mandal, S. K.; Bera, R.; Sarkar, C.; Saha, B.; Mandal,
C.; Saha, K. D.; Bhadra, R. Mol. Cell. Biochem. 2006, 290, 113.