Locations and reorientations of multi-ring-fused 2-pyridones in ganglioside GM1 micelles

Article abstract of DOI:10.1021/la104051z

Fluorescent multi-ring-fused 2-pyridones, with chemical resemblance to other biologically active 2-pyridone systems, were solubilized in spherical micelles formed by the gangloiside G_M1 and studied with respect to their spatial localization and

Full text of DOI:10.1021/la104051z

pubs.acs.org/Langmuir
© 2011 American Chemical Society
Locations and Reorientations of Multi-Ring-Fused 2-Pyridones in
Ganglioside GM1 Micelles
Radek Sachl,†,‡ Erik Rosenbaum,† Magnus Sellstedt,† Fredrik Almqvist,† and
ꢀ
,†
˚
Lennart B.-A. Johansson*
†
‡
Department of Chemistry, Umea University, SE-90 187 Umea, Sweden, and J. Heyrovsky Institute of Physical
˚
Chemistry of the Academy of Sciences of the Czech Republic, Dolejskova 3, 182 23 Prague 8, Czech Republic
˚
ꢁ
ꢀ
Received October 11, 2010. Revised Manuscript Received December 3, 2010
Fluorescent multi-ring-fused 2-pyridones, with chemical resemblance to other biologically active 2-pyridone systems,
were solubilized in spherical micelles formed by the gangloiside GM1 and studied with respect to their spatial localization
and rotational mobility. For this, electronic energy transfer between the multi-ring-fused 2-pyridone (donor) and
BODIPY-FL-labeled GM1 was determined, as well as their fluorescence depolarization. From the obtained efficiency of
energy transfer to the acceptor group (BODIPY-FL), either localized in the polar or in the nonpolar part of the
ganglioside, it has been possible to estimate the most likely localization of the multi-ring-fused 2-pyridones. The center
˚
of mass of the studied multi-ring-fused 2-pyridones are located at approximately 33 A from the micellar center of mass,
which corresponds to the internal hydrophobic-hydrophilic interfacial region. At this location, the reorienting rates of
the multi-ring-fused 2-pyridones are surprisingly slow with typical correlation times of 35-55 ns. No evidence was found
for the formation of ground and excited state dimers, even when two monomers were forced to be near each other via a
short covalent linker.
Introduction
pronounced when the core is in the gel state.7 Release from the
micellar corona on the other hand, which is quite flexible, may be
too rapid to monitor.
A recently synthesized series of multi-ring-fused 2-pyridones
has been spectroscopically characterized,1 and their use as cell-
staining fluorescent dyes demonstrated.1a Ring-fused 2-pyridones
are commonly found to have interesting biological activity.
Camptothecin2 and mappicine3 exemplify substances with anti-
tumor and antiviral properties, respectively. Bicyclic 2-pyridones
have been reported as antibacterial agents, targeting virulence
factors in pathogenic bacteria.4 Due to the biophysical properties
of 2-pyridones and their potential use in drug delivery vehicles, it
is of interest to ascertain the precise location of the compounds,
for example, within the micellar domain of a block copolymer
hydrogel. The use of various lipid phases as drug carriers have
been investigated for some time,5 and polymers that form
macromolecular structures have been investigated as drug deliv-
ery matrices.6 If the drug preferentially resides in the micellar core,
the loading efficiency is restricted by its small size relative to the
micelle and the release is then slower as compared to when the
drug is located in the hydrophilic region. This is especially
Detailed knowledge of the location of drug molecules inside a
delivery vehicle provides for an added level of understanding,
possibly pertaining to drug loading and release processes.7 In the
present study, the location of the multi-ring-fused 2-pyridones
was investigated when solubilized in GM1 ganglioside micelles.
GM1 ganglioside micelles were chosen8 for several reasons,
namely: (i) they exhibit a narrow size distribution and their
aggregation number is known,9 (ii) the thickness of the hydro-
philic headgroup region (which approximately equals that of the
hydrocarbon chain radius) offers a large radial variation in
physicochemical properties (e.g., the local dielectric constant,
solubility, etc.), and (iii) specifically labeled fluorescent ganglio-
sides exist, which form suitable donor-acceptor pairs with these
multi-ring-fused 2-pyridones.9b
Materials and Methods
Chemicals. GM1 ganglioside was isolated from bovine brain as
described by Svennerholm,10 BODIPY-FL-GM1 was synthesized
as reported in ref 11, while 581/591-BODIPY-GM1 was synthe-
sized as described elsewhere.9b Tris(hydroxymethyl)amino-
methane hydrochloride and chloroform (spectroscopic grade)
were bought from Sigma Aldrich.
*To whom correspondence should be addressed.
(1) (a) Sellstedt, M.; Nyberg, A.; Rosenbaum, E.; Engstrom, P.; Wickstrom, M.;
€
€
˚
€
Gullbo, J.; Bergstrom, S.; Johansson, L. B.-A.; Almqvist, F. Eur. J. Org. Chem.
2010, published online Sept 24, http://dx.doi.org/10.1002/ejoc.201000796. (b) Rosenbaum,
E.; Sellstedt, M.; Almqvist, F.; Johansson, L. B.-Å. J. Fluoresc. 2010, http://dx.doi.org/
10.1007/s10895-010-0676-3.
(2) Wall, M. E.; Wani, M. C.; Cook, C. E.; Palmer, K. H.; McPhail, A. T.; Sim,
G. A. J. Am. Chem. Soc. 1966, 88, 3888.
(3) (a) Govindac, T. R.; Ravindra, K. R.; Viswanat, N. J. Chem. Soc., Perkin
Trans. 1974, 1, 1215. (b) Pendrak, I.; Barney, S.; Wittrock, R.; Lambert, D. M.; Ambert,
D. M.; Kingsbury, W. D. J. Org. Chem. 1994, 59, 2623.
(7) (a) Allen, C.; Maysinger, D.; Eisenberg, A. Colloids Surf., B 1999, 16, 3.
(b) Teng, Y.; Morrison, M. E.; Munk, P.; Webber, S. E.; Prochazka, K. Macromolecules
1998, 31, 3578.
(8) Brocca, P.; Corti, M.; Favero, E. D.; Raudino, E. A. Langmuir 2007, 23,
3067.
(4) (a) Cegelski, L.; Pinkner, J. S.; Hammer, N. D.; Cusumano, C. K.; Hung,
˚
C. S.; Chorell, E.; Aberg, V.; Walker, J. N.; Seed, P. C.; Almqvist, F.; Chapman,
M. R.; Hultgren, S. J. Nat. Chem. Biol. 2009, 5, 913. (b) Pinkner, J. S.; Remaut, H.;
€
Buelens, F.; Miller, E.; Åberg, V.; Pemberton, N.; Hedenstrom, M.; Larsson, A.; Seed, P.;
(9) (a) Sachl, R.; Stepanek, M.; Prochazka, K.; Humpolickova, J.; Hof, M.
Langmuir 2008, 24, 288. (b) Mikhalyov, I.; Gretskaya, N.; Johansson, L. B.-Å. Chem.
Phys. Lipids 2009, 159, 38.
Waksman, G.; Hultgren, S. J.; Almqvist, F. Proc. Natl. Acad. Sci. U.S.A. 2006, 103,
17897. (c) Åberg, V.; Almqvist, F. Org. Biomol. Chem. 2007, 5, 1827.
(5) Bender, J.; Simonsson, C.; Smedh, M.; Engstrom, S.; Ericson, M. B.
€
(10) Svennerholm, L. Methods Carbohydr. Chem. 1974, 6, 464.
˚
J. Controlled Release 2008, 129, 153.
(11) Marushchak, D.; Gretskaya, N.; Mikhalyov, I.; Johansson, L. B.-A. Mol.
(6) Torchilin, V. P. Pharm. Res. 2007, 24, 1.
Membr. Biol. 2007, 24, 102.
1662 DOI: 10.1021/la104051z
Published on Web 01/06/2011
Langmuir 2011, 27(5), 1662–1667

ꢀ
Sachl et al.
Article
Syntheses of 2-Pyridonebased Compounds. The com-
pounds D-Me, D-H, and D-F (cf. Figure 1) have been synthesized
previously.1a The bis-polyaromatic 2-pyridones, (D-H)C2 and
(D-H)C3, were synthesized as described in the following.
Dimerization of (D-H)OH
.
A total of 37 mg (0.099 mmol) of
(D-H)OH and 2 mg (0.016 mmol) of 4-dimethylaminopyridine
was taken up in 0.5 mL of tetrahydrofuran. Then 0.50 mL (0.050
mmol) of 0.10 M of the appropriate diol in tetrahydrofuran
was added followed by 29 mg (0.151 mmol) of N-(3-
dimethylaminopropyl)-N0-ethylcarbodiimide hydrochloride.
The slurry was stirred at room temperatureovernight. An amount
of 0.5 mL of CH2Cl2 was added, and the mixture was stirred
another 7 h and then concentrated under reduced pressure. The
residue was taken up in CH2Cl2 and washed with 1 M HCl (aq).
The aqueous phase was extracted with CH2Cl2, and the combined
organic phases were then dried over anhydrous Na2SO4, filtrated,
and concentrated under reduced pressure. The crude product was
purified with silica gel chromatography using heptane/ethyl
acetate 2:1 f 1:1 as a mobile phase.
All reactions are run under N2(g) with anhydrous solvents
unless otherwise stated. The 1H and 13C NMR spectra were
recorded at 298 K with a Bruker DRX-400 spectrometer and
calibrated using the residual peak of CHCl3 or dimethyl sulfoxide
(DMSO) as internal standard (CHCl3: δH 7.26 ppm, δC 77.16
ppm; DMSO: δH 2.50 ppm, δC 39.52 ppm). In cases where the
diastereomers give different chemical shifts, signals from major
and minor diasteromere are indicated with “maj” and “min”,
respectively. LRMS was conducted on a Micromass ZQ mass
spectrometer with ESþ ionization.
Compound (D-H)OH (cf. Scheme 1). A total of 1.05 mL of 0.5
M LiOH (aq) was slowly added to 194 mg (0.501 mmol) of D-H1a
in 5.0 mL of tetrahydrofuran and 5.0 mL of methanol. The
mixture was stirred at room temperature overnight and then
concentrated under reduced pressure. The residue was taken up
in 10 mL of acetic acid and 10 mL of chloroform, and approxi-
mately 3 mL of Amberlite IR-120 (Hþ) was added. After 3 h of
shaking, the resulting solution was filtrated and concentrated to
afford 173 mg (93%) of compound (D-H)OH. 1H NMR (DMSO-
d6): δ 13.54 (1H, bs), 8.93 (1H, s), 8.18 (1H, d, J = 8.0 Hz), 7.87
(1H, d, J = 8.1 Hz), 7.63-7.35 (8H, m), 5.70 (1H, d, J = 8.3 Hz),
Compound (D-H)C2. By following the procedure for dimer-
ization of (D-H)OH, using ethylene glycol as diol, 18 mg (47%) of
(D-H)C2 was obtained as a mixture of diasteromeres in an
approximately 6:4 ratio according to 1H NMR. 1H NMR
(CDCl3): δ 8.98 (2H maj, s), 8.97 (2H min, s), 8.02-7.95 (2H,
m), 7.71 (2H, d, J = 8.1 Hz), 7.60 (2H, s), 7.58-7.39 (14H, m),
5.78 (2H maj, dd, J = 8.0 Hz, 2.4 Hz), 5.72-5.67 (2H min, m),
4.63-4.40 (4H, m), 3.63-3.53 (2H, m), 3.50-3.42 (2H, m). 13
C
NMR (CDCl3): δ 168.4 (2C maj), 168.3 (2C min), 161.6 (2C),
137.7 (2C), 136.7 (2C), 135.8 (2C), 135.1 (2C, split), 131.0 (2C),
130.8 (2C), 130.4 (2C), 129.6 (2C), 129.5 (2C), 129.2 (2C, split),
129.0 (2C), 128.5 (2C), 128.4 (2C, split), 128.0 (2C), 126.0 (2C),
122.7 (2C), 121.9 (2C), 112.8 (2C), 63.7 (2C maj), 63.6 (2C min),
63.0(2C maj), 62.9(2C min), 31.4(2C, broad). LRMS(ESþ) calcd
for C46H33N2O6S2 [MþH], 773; found, 773.
3.83 (1H, dd, J = 11.6 Hz, 8.3 Hz), 3.52 (1H, d, J = 11.6 Hz). 13
C
NMR (DMSO-d6): δ 169.9, 160.1, 139.2, 136.3, 135.1, 133.6,
130.3 (3C), 129.2 (2C, splitted), 129.1, 128.8, 128.6, 128.2, 127.7,
125.9, 122.4, 120.7, 110.5, 62.7, 31.1. LRMS (ESþ) calcd for
C22H16NO3S [M þ H], 374; found, 374.
Compound (D-H)C3. By following the general procedure for
dimerization of (D-H)OH, using 1,3- propanediol as diol, 16 mg
(41%) of (D-H)C3 was obtained as a mixture of diasteromeres in
1
1
an approximately 7:4 ratio according to H NMR. H NMR
(CDCl3): δ 8.99 (2H, s), 8.05-7.95 (2H, m), 7.76-7.68 (2H, m),
7.63-7.35 (18H, m), 5.79-5.70 (2H, m), 4.44-4.21 (4H, m),
3.65-3.53 (2H, m), 3.44-3.38 (2H min, m), 3.32 (2H maj, dd,
J = 11.6 Hz, 2.2 Hz). 13C NMR (CDCl3): δ 168.5 (2C), 161.6 (2C),
137.6 (2C, split), 136.7 (2C), 135.8 (2C), 134.1 (2C), 131.0 (2C),
130.8 (2C), 130.4 (2C), 129.7 (2C, split), 129.5 (2C), 129.3 (2C),
129.0 (2C), 128.5 (2C), 128.4 (2C), 128.0 (2C), 126.0 (2C), 122.7
(2C, split), 121.9 (2C), 112.8 (2C, split), 63.0 (2C), 62.4 (2C) 31.4
(2C), 27.8 (2C maj) 27.7 (2C min). LRMS (ESþ) calcd for
C47H35N2O6S2 [MþH], 787; found, 787.
Preparation of GM1 Micelles. Appropriate amounts of GM1
were dissolved in a chloroform/methanol mixture (2:1, v/v), to
which the desired amount of fluorophore was added. After
evaporation of the organic solvents by a continuous flow of Ar(g),
the sample was dried under high vacuum for 3 h. The obtained
lipid film was hydrated to 0.16 mM by adding a TRIS-HCl buffer
(pH 7.4) containing 150 mM NaCl.
Absorption and Fluorescence Spectra. The absorption
spectra were recorded on a Varian Cary 5000 UV-vis spectrom-
eter. The fluorescence spectra were transcribed by means of a
Figure 1. Structures of the studied multi-ring-fused 2-pyridones.
Scheme 1. Synthesis of the Multi-Ring-Fused 2-Pyridone (D-H) Dimersa
a EDC = N-(3-dimethylaminopropyl)-N0-ethylcarbodiimide hydrochloride; 4-DMAP= 4-dimethylaminopyridine. The diastereomeric ratio (dr) was
estimated from 1H-NMR spectra.
Langmuir 2011, 27(5), 1662–1667
DOI: 10.1021/la104051z 1663

ꢀ
Sachl et al.
Article
Figure 2. Schematic of a spherical micelle; also shown is a space
fillingmodelofthe GM1 ganglioside. EachGM1 lipidisillustratedin
a hypothetical conformation, rather than being in the true liquid
state. The vector end points indicate the center of mass of a donor
molecule (RBD) and an acceptor (RBA) group. The donor position is
expressedinsphericalcoordinatesRBD = RD(cosRsinβ, sinRsinβ,
cos β).
Figure 3. Corrected fluorescence spectra of the D-F variant (with
unknown position within a micelle) (solid), the acceptor FL-NP
(dotted), the acceptor 581/591-NP (densely dotted) (acceptor’s
positions are known), and D-F in the presence of FL-NP acceptor
(dashed). On the very left, the absorption spectrum of D-F is
displayed (solid) followed by the absorption spectrum of the
acceptor FL-NP (solid). The arrows pointing downward indicate
the wavelength selected for monitoring the fluorescence intensity.
Fluorolog -3 spectrometer (Jobin Yvon) equipped with Glan-
Thompson polarizers. The bandwidth of the excitation and
emission light was 3 and 2 nm, respectively. All fluorescence
spectra were corrected. For the study of the donors’ location
inside the micelles, the emission intensity was monitored at either
470 nm (FL-BODIPY-GM1 acceptor) or at 505 nm (581/591-
BODIPY-GM1 acceptor) (cf. Figure 3).
center of mass at the different effective radii RD and RA,
respectively. (cf. Figure 2). For such a D-A arrangement, Uhlik
et al.18 have derived an equation for the mean energy transfer rate
which has been used for the analyses of the steady-state fluores-
cence data obtained here. The average rate of energy transfer (ω)
depends on the distribution of distances |RB| = |RBD - RBA|, which
is random, due to the spherical symmetry of the micelle. The
transfer rate also depends on the number of acceptors NA. The
transfer rate can be written as
Fluorescence Lifetime Measurements. The time-resolved
fluorescence decays were measured by means of the time-
correlated single photon-counting technique12 by using a
Fluorolog-TCSPC (Horiba) spectrometer. The fluorescence de-
cays were collected over 2048 channels, with a resolution of 100
ps/ch, with at least 8000 photons in the peak maximum for the
lifetime experiments, which were performed with the emission
polarizer set to magic angle (54.7ꢀ) with respect to the excitation
polarizer. For the time-dependent anisotropy experiments, the
fluorescence decays were collected with a 10000-count difference
in peak maxima between the decays collected with parallel and
perpendicular polarizer settings. For the pulsed excitation, a
NanoLED 280 nm and a 404 nm diode laser were used in
combination with an interference filter centered at 280 and 404
nm (HBW 10 = nm), respectively. The emission monochromator
(IBH system, U.K.) was used in combination with an interference
filter centered at 470 nm (HBW 10 = nm). The fluorescence
lifetime was calculated by means of a deconvolution method,13
which is based on the Levenberg-Marquardt algorithm.14
Z
R06NA
PðRÞ dR
Rmax
ω ¼
ð1Þ
τD
R6
Rmin
In eq 1, the donor-acceptor distance (R) depends on the spherical
2
2
polar orientation angle, β, according to R = (RD þ RA
-
2RDRA cos β). Here P(R) is the normalized number density of
acceptors at the distance R from a donor (= probability that an
acceptor is found at the distance R from a donor) and is given by
P(R) = R/(2RARD).18 Then the total energy transfer rate (ωt)
which is given by FD(RD)ω is proportional to the following
expression:
2
2
RA þ RD
Theoretical Prerequisites
ωt ꢁ F ðRDÞNA
ð2Þ
ꢀ
ꢁ
D
4
RA2 - RD
2
Important theoretical work on donor-acceptor energy trans-
fer in restricted geometries has been published by several research
groups (Levitz and co-workers15). Based on the formalism devel-
oped by the Blumen and Klafter groups,15a,16 Yekta et al.17 have
derived a general equation for energy transfer from donors to
acceptors which exhibits different concentration profiles in
spherically symmetric systems. The general equation derived by
Yekta et al. is rather complex, but it can be simplified for certain
concentration profiles. In this paper, we assume that donors (D)
and acceptors (A) are randomly distributed about the micellar
Since the distribution of acceptors around any donor is identical,
the fluorescence decays monoexponentially with a rate constant =
1/τD þ ω. The randomly distributed acceptors surrounding the
micellar center of mass can be considered as one effective acceptor
which quenches donor fluorescence at a transfer rate = ωt. The
ratio between the fluorescence intensity of the donor in the
absence (F0D) and in the presence (FD) of acceptors can then be
related to the total energy transfer rate in a similar way as is done
for a D-A pair:19
(12) Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 3rd ed.; Springer:
Singapore, 2006.
(13) Demas, J. N. Excited state lifetime measurements; Academic Press: New
York, 1983.
FD0
¼ 1 þ ωtτD
ð3Þ
FD
(14) Levenberg, K. Q. Appl. Math. 1944, 2, 164.
(15) (a) Levitz, P.; Drake, J. M.; Klafter, J. J. Chem. Phys. 1988, 89, 5224. (b)
Drake, J. M.; Klafter, J.; Levitz, P. Science 1991, 251, 1574. (c) Duhamel, J.; Yekta, A.;
Ni, S.; Khaykin, Y.; Winnik, M. Macromolecules 1993, 26, 6255. (d) Farinha, J. P. S.;
Thus, the energy transfer causes a decrease of fluorescence
intensity of a donor. Since the donor and acceptor molecules
ꢁ
Schillen, K.; Winnik, M. A. J. Phys. Chem. B 1999, 103, 2487. (e) Baumann, J.; Fayer,
M. D. J. Chem. Phys. 1986, 85, 4087.
(16) Blumen, A.; Klafter, J.; Zumofen, G. J. Chem. Phys. 1986, 84, 1397.
(17) Yekta, A.; Winnik, M. A.; Farinha, J. P. S.; Martinho, J. M. G. J. Phys.
Chem. A 1997, 101, 1787.
(18) Uhlik, F.; Limpouchova, Z.; Matejicek, P.; Prochazka, K.; Tuzar, Z.;
Webber, S. E. Macromolecules 2002, 35, 9497.
(19) Valeur, B. Top. Fluoresc. Spectrosc. 1994, 4, 21.
1664 DOI: 10.1021/la104051z
Langmuir 2011, 27(5), 1662–1667

ꢀ
Sachl et al.
Article
Table 1. Radial Distances between Various 2-Pyridiones and the
Micellar Centera
experiments performed with same acceptor group, localized at
different positions in the GM1 ganglioside. By forming the ratio
between the corrected experimental ratios (cf. eq 2) obtained
with the acceptor pairs FL-P and FL-NP or 581/591-P and 581/
591-NP, the explicit value of FD(RD) is not needed. Thereby, the
question of determining possible values of RD distances is defined
by finding the minima of the following expression:
b
˚
˚
donor
acceptor
Ænæ P0 [(FD(NP)/FD(P)]exp RA (A) RD (A)
D-F FL-NPc
FL-Pc
0.75 0.47
0.77 0.46
1.15
1.15
22.7
41.6
33.5
33.5
D-H FL-NPc
FL-Pc
0.71 0.49
0.76 0.47
1.06
1.06
22.7
41.6
32.8
32.8
"
#
2
ꢂ
ꢃ
FD0 ðNPÞ - FDðNPÞ FDðPÞ
ωtðNPÞ
ωtðPÞ
D-H 581/591-NPd 1.27 0.28
1.03
22.7
31.0
-
¼ 0
ð5Þ
FD0 ðPÞ - FDðPÞ
FDðNPÞ
D-Me 581/591-Pd 1.36 0.26
1.07
1.07
1.07
41.6
22.7
41.6
32.8
32.8
32.8
FL-NPc
FL-Pc
0.73 0.48
0.77 0.46
a The concentration of GM1 ganglioside was 0.16 mM, for which the
micellar aggregation number is 16821. P0 denotes the fraction of micelles
which is not containing any acceptors. The average number of acceptor
In eq 5, P and NP refer to the acceptor groups linked to the
polar and nonpolar region of the GM1 lipid, respectively. The
values of ωt depend on RD according to eq 2. An iterative
molecules per micelle Ænæ = NA/Nmic
.
b Corresponds to a fully extended
˚
˚
calculation method yields a single solution for RD ∈ Æ0;54æA, that
21
d
GM1 molecule. c R0 = 49.5 A. R0 = 38.4 A.
˚
is, for a physically relevant value of RD (Table 1), which must not
˚
exceed the known value of the micellar radius (= 54 A). The
are distributed among the micelles, one needs to account for the
probability that micelles do not contain any acceptor. By assum-
ing a Poisson distribution,20 this fraction of micelles (P0) can be
calculated from the relation P0 = exp(-ÆNA/Nmicæ). The ratio
within the angled brackets represents the average occupation
number of acceptors per micelle, which can be calculated from the
known values of the number of acceptor molecules (NA) and
micelles (Nmic) in the studied system. For calculating the latter
value, the previously determined value of the micellar aggregation
number of GM1 gangliosides has been used (Nagg = 168).21
Hence, it is possible to correct the experimental donor fluores-
cence intensities (F 0D,exp and F eDxp) with respect to micelles
lacking acceptors according to the relations F 0D = F D0,exp(1-P0)
and FD = F eDxp-F D0,expP0.
˚
obtained values of RD ≈ 33 A for all the studied multi-ring-fused
2-pyridones imply that these compounds are localized in the
interfacial region of GM1 micelles. A similar localization was
found, for example, for the aromatic probe PRODAN when
solubilized in PCL-PEO vesicles.24 PRODAN is frequently used
for monitoring solvent/environmental relaxation caused by an
instantaneous electronic perturbation (vide infra). However, the
majority of PRODAN molecules are easily released from the
vesicles into the bulk phase by the addition of small amounts of
THF (approximately 10% by volume). Similarly, the BODIPY
and NBD probes preferentially reside at the interface of lipid
membranes,24 while other probes, such as 2,5,8,11-tetra-tert-
butylperylene, prefer to solubilizein the interior of lipid bilayers.25
For the compounds studied here, however, the substitution of
hydrogen by fluorine or by a methyl group has no influence on
their effective distance from the micellar center of mass (cf.
Table 1). This is in agreement with a recent study where the
substitution of four hydrogen atoms by four methyl groups in the
aromatic core of BODIPY connected to phospholipid did not
prevent the chromophore from residing close to the lipid-water
interface.26
From fluorescencedepolarisationexperiments, the steady-state
and the time-resolved anisotropy were calculated. In the analyses,
the time-resolved anisotropy was modeled according to
X
rðtÞ ¼
rkð0Þ expð - t=φkÞ
ð4Þ
k
In eq 4, φk denotes rotational correlation times and r(0) = Σkrk(0)
e r0, where r0 stands for the fundamental anisotropy.22
Considering the chemical structure of the acceptors, one might
question the obtained donor positions, which are calculated
within the assumption of a fully extended GM1 molecule. These
results (Table 1) are based on the assumption that the acceptor
groups reside effectively at the same distance from the micellar
center of mass in the nonpolar (FL-NP) as well as in the polar
(FL-P) positions of a fully extended lipid (cf. Figure 2). Conse-
quently, one could also question the above estimations of donor
depths (RD). There is a certain degree of spatial freedom of the
BODIPY group, since it is attached to the lipid chain via a linker,
which allows for displacements toward the micellar center of mass
as well as the bulk phase. This positional uncertainty influences
Results
Depths of Fluorescent Molecules Solubilized in GM1
Micelles. For positioning the studied multi-ring-fused 2-pyri-
€
dones within GM1 micelles, the Forster mechanism of electronic
energy transfer23 has been applied. The studied donor molecules
are polyaromatic 2-pyridones, which are hereafter referred to as
D-Me, D-H, and D-F (cf. Figure 1). Four different BODIPY
modified GM1 molecules constitute the acceptors: FL-BODIPY-
GM1, in which the acceptor group is specifically attached in either
the nonpolar (FL-NP) or in the polar (FL-P) region of the GM1
micelle; also 581/591-BODIPY-GM1 was used for labeling the
polar (581/591-P) and nonpolar (581/591-NP) region. The posi-
tions of the acceptor groups in GM1 micelles refer to the fully
extended lipid molecule (cf. Table 1). The surface density values of
the donors appearing in eqs 1 and 2 are not known. This can be
circumvented if the donor concentration is identical in each of two
˚
the absolute values of RD, which range between 19 A (for FL-NP
˚
and FL-P pointing toward the center) and 35 A (for FL-NP and
FL-P pointing toward the bulk phase). However, the analysis
yields a relative position with respect to the acceptors, which is
given by [RD(P) - RA(NP)]/[RA(P) - RA(NP)] ranging between
0.57 and 0.60. Thus, this relatively small range means that the donor
molecules reside at the nonpolar/polar interface at approximately
(20) Recherches sur la probabiliteꢁ des jugements en matieꢂre criminelle et en
matieꢂre civile; Poisson, S. D., Ed.; Bechelier: Paris, 1837.
(21) Sachl, R.; Mikhalyov, I.; Hof, M.; Johansson, L. B.-A. Phys. Chem. Chem.
(24) Menger, F. M.; Keiper, J. S.; Caran, K. L. J. Am. Chem. Soc. 2002, 124,
˚
11842.
˚
Phys. 2009, 11, 4335.
(25) Kalman, B.; Clarke, N.; Johansson, L. B.-A. J. Phys. Chem. 1989, 93, 4608.
(26) Sachl, R.; Boldyrev, I.; Johansson, L. B.-A. Phys. Chem. Chem. Phys. 2010,
˚
(22) Jablonski, A. Acta Phys. Pol. 1950, 10, 33.
€
(23) Forster, T. Ann. Phys. 1948, 2, 55.
12, 6027.
Langmuir 2011, 27(5), 1662–1667
DOI: 10.1021/la104051z 1665

ꢀ
Sachl et al.
Article
the fluorescence relaxation and reorienting rate. This is investi-
gated through studying the fluorescence anisotropy of D-H,
(D-H)C2, and (D-H)C3 when solubilized in GM1 micelles. The
micelles’ rotational correlation time is estimated to be approxi-
mately 200 ns, which implies a negligible influence on the
fluorescence anisotropy. Interestingly, the average fluorescence
lifetimes (∼18-19 ns) for all compounds are very similar, and so
are the rotational correlation times; that is, all compounds reveal
both a short (∼1-3 ns) and a much longer (35-55 ns) correlation
time. For (D-H)C2 and (D-H)C3, the reorientations are very
similar; that is, the fast and slow correlation times are ∼2.7 and
∼55 ns, respectively. The corresponding correlation times for
D-H are ∼1.6 and ∼35 ns, respectively. The overall faster
reorienting motions obtained for D-H can be expected, since its
molecular volume is close to one-half of that for a dimer. A likely
explanation for the lower reorientation rate stems from slow local
reorienting motions of the GM1 molecules in the region between
the nonpolar and polar parts inside of the micelle. Thus, an
influence similar to solvent relaxation can be expected.
In a previous study, the obtained fundamental fluorescence
anisotropy22 of D-H was r0 = 0.28 ( 0.01. For D-H solubilized in
GM1 micelles, the initial value of the time-resolved fluorescence
anisotropy r(0) = 0.287 (cf. eq 4). Thus, unresolved fast reorient-
ing motions of D-H molecules in micelles are not present. The
corresponding initial anisotropy of (D-H)C2 and (D-H)C3 is
r(0) = 0.252 and r(0) = 0.257, respectively. A significant lowering
of r(0) < r0 for the dimers due to reorienting motions is not
expected, since the molecular volume is approximately twice as
large. A reasonable explanation for the apparent lowering,
however, could be intramolecular energy migration, whereby
r(0) would decrease. The only exception to this is the instance
when the transition dipoles within a dimer happen to be colinearly
oriented, which implies that r(0) = r0. The lowered r(0) values for
(D-H)C2 and (D-H)C3 are ascribed to very fast depolarizations
caused by fast intramolecular electronic energy migration within
the bis-fluorophoric molecules. Since the distance between the
D-H groups within a dimer is short, this process will occur on a
time scale beyond the time resolution of the equipment used. By
using the following relation 1/2(3cos2 θ - 1) = r(0)/r0, the lowered
r(0) value for the dimers allows for an estimation of the intramo-
lecular angle (θ) between the S0 T S1 transition dipoles of the two
D-H groups. For (D-H)C2 and (D-H)C3, the possible solutions
obtained are θC2 = 16.6ꢀ or 163.4ꢀand θC3 = 15.3ꢀ or 164.7ꢀ,
respectively. For aromatic molecules, the electronic transition
dipoles between singlet states are in-plane polarized with the
lowest transitions frequently directed along the molecular long
axis.28 Thus, the molecular long axes of the two monomers either
tend to be mutually parallel or antiparallel. The latter implies an
extended molecule, as is indicated in Figure 1. But what solutions
are the most probable? By considering the molecular structure of
the D-H molecule, a nonvanishing permanent dipole moment is
expected. Furthermore, the structure of (D-H)C2 and (D-H)C3
infers that these dipoles tend to repel each other, so that an
extended conformation resembling the structure shown in
Figure 1 would be energetically favorable. The extended structure
appears to be favorable in vacuum, as is evident from results of
computer simulations of the dimer. These were performed based
on using an energy minimization program (CS Chem3D Pro
version 5.0). In the excited state, the possibility of intramolecular
interactions causing excimer formation might exist. However, no
Figure 4. Time-resolved fluorescence decay for D-H in GM1 mi-
celles occurring on the nanosecond time scale. Emission has been
recorded at 470 and 550 nm (upper curve).
half the distance between P-acceptor and NP-acceptor, while still
somewhat shifted toward to the P-acceptor.
Fluorescence Lifetime and Depolarization Studies of
Multi-Ring-Fused 2-Pyridones in GM1 Micelles. Previous
fluorescence lifetime experiments performed on the monomeric
multi-ring-fused 2-pyridones dissolved in CH2Cl2 revealed a
single-exponential relaxation.1b The fluorescence lifetimes of the
bis-forms (D-H)C2 and (D-H)C3 in CHCl3 are 13.1 and 13.2 ns,
respectively. These values are very similar to 13.4 ns, as was
previously obtained for D-H dissolved in CH2Cl2,1b which
suggests negligible intramolecular quenching. Interestingly, the
obtained relative fluorescence quantum yield (Φf) of D-H and (D-
H)C2, that is, Φf,(D-H)2/Φf,D-H, is 0.85. This reveals an influence of
static quenching. Moreover, the molar absorptivity of the bis-
form is twice that of the monomer. Furthermore, this is supported
by the almost identical fluorescence and absorption spectra as
compared to the monomeric form. Sterically however, the linker
groups (-CH2CH2- and -CH2CH2CH2-) could very likely
allow for the formation of intramolecular ground state dimers, as
well as excited state dimers. When residing in GM1 micelles, the
fluorescence relaxation of (D-H)C2 and (D-H)C3 are more com-
plex than in CHCl3, and can be adequately described by a sum of
two or three exponential functions. The corresponding average
lifetimes are 18.9 and 18.2 ns, respectively. These values are
similar to the monoexponential fluorescence (τf = 17 ns) pre-
viously reported for DH in glycerol.1b The formation of excited
state dimers is unlikely, since no additional emission bands were
observed. Also the presence of intramolecular ground state
dimers is negligible, since the shapes of the absorption spectra
were practically identical to that measured for D-H dissolved in
GM1 micelles. Neither absorption nor emission spectra differ
between D-Me, D-H, and D-F when solubilized in GM1 micelles.
For these solubilized compounds a complex fluorescence relaxa-
tion was observed, which is similar to that observed for (D-H)C2
and (D-H)C3. A slight lifetime dependence on the selected
wavelength region of emission was also found. The decay of
DH is biexponential with an average fluorescence lifetime of 19.6
ns at 470 nm; the decay is reasonably well described by a single
lifetime τf = 20.5 ns at 550 nm (cf. Figure 4). The wavelength
dependency is compatible with a weak solvent relaxation effect,
which is expected for polar molecules whose permanent dipole
moment changes upon electronic excitation, but its impact on
lifetime and fluorescence spectra is more strongly pronounced in
electronic charge transfer processes.27 A slow relaxation rate of
the nearby molecular surrounding of an excited fluorophore
means that the process occurs on a time scale comparable to
(27) Jurkiewicz, P.; Sykora, J.; Olzynska, A.; Humpolickova, J.; Hof, M.
(28) Michl, J.; Thulstrup, E. W. Spectroscopy with Polarized Light; VCH
J. Fluoresc. 2005, 15, 883.
Publishers, Inc.: New York, 1986.
1666 DOI: 10.1021/la104051z
Langmuir 2011, 27(5), 1662–1667

ꢀ
Sachl et al.
Article
evidence for this occurrence was found neither in GM1 micelles
nor in low viscous solvents, for example, chloroform.
linked covalently to nonpolar headgroup
of GM1
581/591-P N-(BODIPY-581/591-pentanoyl)-neuraminosyl-
)
ganglioside (GM = 581/591-BODIPY-GM1
Concluding Remarks
linked covalently to polar headgroup of GM1
581/591-NP N-(BODIPY-581/591-pentanoyl)-ganglioside
(GM1 = 581/591-BODIPY-C5-nonpolar-GM1
581/591-BODIPY-GM1 linked covalently to
nonpolar headgroup of GM1
)
Taken together, apart from having unusual fluorescence spec-
troscopic properties,1b the studied multi-ring-fused 2-pyridones
are located in the hydrophobic-hydrophilic micellar interface
where they undergo slow and restricted tumbling. Furthermore,
experimental evidence reveals no tendency for neither ground
state nor excited state dimer formation at locally high monomer
concentrations, nor indeed even when covalently linked together.
=
)
D-F
Methyl 8-Fluoro-2,3-dihydro-12-phenyl-benzo[g]-
thiazolo[3,2-b]isoquinoline-5-one-3-carboxylate
Methyl 2,3-Dihydro-12-phenyl-benzo[g]thiazolo-
[3,2-b]isoquinoline-5-one-3-carboxylate
Methyl 2,3-Dihydro-11-methyl-12-phenyl-
benzo[g]thiazolo[3,2-b]isoquinoline-5-one-3-
carboxylate
1,2-Di(2,3-dihydro-12-phenyl-benzo[g]thiazolo-
[3,2-b]isoquinoline-5-one-3-carboxylate) ethane
1,3-Di(2,3-dihydro-12-phenyl-benzo[g]thiazolo-
[3,2-b]isoquinoline-5-one-3-carboxylate) propane
Ganglioside GM1
D-H
D-Me
Acknowledgment. We are grateful to M.Sc. Oleg Opanasyuk
for preparing Figure 2. Also, this work was financially supported
by the Swedish Research Council, the Kempe Foundations, and
the Grant Agency of the Czech Republic (P208/10/1090).
(D-H)C2
(D-H)C3
List of Abbreviations
GM1
r0
r(0)
R0
τf
Fundamental fluorescence anisotropy
Initial fluorescence anisotropy
€
Forster radius
Fluorescence lifetime
FL-P
N-(BODIPY-FL-pentanoyl)-neuraminosyl-
ganglioside (GM1 = FL-BODIPY-GM1 linked
covalently to polar headgroup of GM1
FL-BODIPY-GM1 = N-(BODIPY-FL-
pentanoyl)-ganglioside (GM1=FL-BODIPY-GM1
)
FL-NP
φ
Rotational correlation time
Langmuir 2011, 27(5), 1662–1667
DOI: 10.1021/la104051z 1667