To D or not to D?: On estimating the microenvironment polarity of biomolecular cavities

Article abstract of DOI:10.1039/b707719j

Knowledge of the local polarity of specific cavities in biopolymers can facilitate the design of selective low MW ligands that impact the structure and function of macromolecules. The most common tools for interrogating local polarity are fluorescent probes that are sensitive to their microenvironment. Researchers often evaluate and express this local polarity using dielectric constants, a parameter that reflects an inherent bulk property. A more appropriate expression should take into account solvent-solute interactions at the molecular level. Reevaluation of commonly used fluorophores illustrates the improved correlation between observed Stokes shift changes and E_T(30) values as compared to the corresponding dielectric constants. The Royal Society of Chemistry.

Full text of DOI:10.1039/b707719j

View Article Online / Journal Homepage / Table of Contents for this issue
ESSAY
www.rsc.org/obc | Organic & Biomolecular Chemistry
To D or not to D?
On estimating the microenvironment polarity of biomolecular cavities†
Renatus W. Sinkeldam and Yitzhak Tor*
Received 22nd May 2007, Accepted 29th June 2007
First published as an Advance Article on the web 10th July 2007
DOI: 10.1039/b707719j
Knowledge of the local polarity of specific cavities in biopolymers can facilitate the design of selective
low MW ligands that impact the structure and function of macromolecules. The most common tools
for interrogating local polarity are fluorescent probes that are sensitive to their microenvironment.
Researchers often evaluate and express this local polarity using dielectric constants, a parameter that
reflects an inherent bulk property. A more appropriate expression should take into account
solvent–solute interactions at the molecular level. Reevaluation of commonly used fluorophores
illustrates the improved correlation between observed Stokes shift changes and E (30) values as
T
compared to the corresponding dielectric constants.
Introduction
Expression of polarity
to develop reliable microscopic solvent
polarity scales evaluated the impact of
solvents on the chemical reactivity of
alkyl halides and fundamentally relied on
Experimentalists as well as theoreticians
have struggled, for a while, with the
proper description of the polarity within
confined spaces inside biopolymers. The
notion of microenvironment polarity is,
of course, of key importance, since intra-
and intermolecular forces (such as H-
bonding) are critically dependent on their
surroundings, and particularly, on the
attenuating (or strengthening) power of
Medium effects on the structure, confor-
mation and reactivity of small and large
molecules have been at the foundation of
physical chemistry. Early studies utilized
dielectric constants, or relative permittiv-
ity, as a measure of polarity, expressed
in units of Debye (D). Such values rep-
resent a molecule’s ability to attenuate an
electric field generated between macro-
scopically distant electrodes relative to
vacuum (e = 1 by definition). It can be
viewed as a measure of bulk polarizability,
the capability of molecules to respond
to the applied field and reorganize to
minimize the generated potential. Polar
solvents, such as water and low MW alco-
hols, are capable of effectively attenuating
the generated field, and therefore have
relatively high dielectric constants, while
apolar solvents, such as hydrocarbons,
with little or no ability to respond to an
applied field, are characterized by very low
dielectric constants. It is intuitively clear
that such values, representing bulk order
or disorder, do not faithfully represent the
first or second solvation spheres surround-
ing a solute molecule, or the environment
3
linear free energy relationships (LFER).
The Y scale, developed by Grunwald and
3
Weinstein, yielded a quantitative measure
ꢀ
for solvents ionization power, but was
limited in terms of its general applica-
bility, range of solvents and cumbersome
4
determination. The turning point came
when spectroscopic approaches were con-
ceived. The use of chromophoric charge
transfer complexes to energetically de-
fine solvent polarity was pioneered by
1
solvent molecules. In recent years, ele-
gant approaches utilizing environmentally
sensitive fluorescent probes, have been
employed in an attempt to quantify the
polarity of biological cavities, both in
proteins and nucleic acids. These attempts
are telling and informative, but suffer from
two major predicaments: (a) any probe
placed within the cavity to be assessed
inherently modifies the molecular archi-
tecture of the native environment, and
5
Kosower. The Z value solvent polarity
scale relied on a simple measurement of the
absorption maximum of a charge transfer
complex. A related scale, that has gained
popularity due to its ease of use and
coverage of a wide range of solvents and
solvent mixtures, was developed by Dim-
roth and Reichardt. An E
in kcal mol ) is similarly determined by
measuring a charge transfer band of a
T
(30) value (given
(
b) many studies have utilized dielectric
−1
constants as their gauge, a parameter that
defines bulk solvent property and not an
anisotropic medium. It is the latter aspect
that is discussed here.
6,7
pyridinium betaine dye (1) (Fig. 1). Val-
ues for the dielectric constant and E (30)
T
values for selected solvents are depicted in
Table 1.
2
within a small molecular cavity.
Although a number of other ways to
7,8
Experimental support for such di-
chotomy has emerged in early physical
organic studies where, for example, rates
of solvolysis reactions were measured in
diverse media and showed no correlation
with dielectric constants. Early attempts
express polarity have been explored, this
article focuses on the comparison between
the commonly used dielectric constant and
Department of Chemistry and Biochemistry, Uni-
versity of California, San Diego, La Jolla, CA
9
2093, USA
E (30) values in the expression of polarity
T
†
Electronic supplementary information (ESI)
as derived from changes in the Stokes shift
of polarity sensitive probes.
available: Synthesis and spectroscopic details.
See DOI: 10.1039/b707719j
This journal is © The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 2523–2528 | 2523

View Article Online
a
Table 1 Polarity parameters of common solvents
−
1
Solvents
Hexane
e/D
n
Df
T
E (30)/kcal mol
1.889
2.219
20.190
20.800
25.290
33.520
80.180
1.375
1.422
1.377
1.384
1.360
1.329
1.333
0.000
0.022
0.277
0.275
0.290
0.309
0.320
31.0
36.0
48.4
50.7
51.9
55.4
63.1
1
2
1
,4-Dioxane
-Propanol
-Propanol
Ethanol
Methanol
Water
a
See ESI for references.
Fig. 1 Reichardt’s dye (1), polarity sensitive fluorophores (2, 5, 6, 10, 11, 12, 13), fluorophore modified nucleosides (3, 4, 7, 8, 14) and fluorophore
modified alanine (9).
the ground and excited state, with e and
n representing the dielectric constant and
the refractive index, respectively (eqn (1)).
If only general solvent effects play a
role, the Stokes shift is linearly dependent
on the polarizability (Df ) of the solvent.
Deviation from linearity implies the con-
tribution of solvent specific effects such
as hydrogen bonding, acid–base interac-
tions or charge transfer interactions for
which the approximations used in deriv-
How is the polarity of
biomolecular cavities probed?
Inherent limitations of the natural building
blocks found in nucleic acids and proteins
have triggered the synthesis and implemen-
tation of designer, environmentally sensi-
tive, fluorescent probes. Regardless of the
probe’s structure and its placement within
a biomolecule, probing the biopolymer’s
polarity under native conditions is always
referenced to values determined for the iso-
lated probe in solvents or solvent mixtures
of known polarity. This section outlines
a typical procedure for correlating ex-
perimental observables with an unknown
environmental polarity.
2
e − 1
n − 1
D f =
−
(1)
2
2e + 1 2n + 1
The orientational polarizability param-
eter is part of an equation that describes
the solvent’s influence on the energy dif-
ference between the ground and excited
ing the Lippert–Mataga equation cannot
12,13
account.
As a result, investigators tend
to utilize a limited window of linearly cor-
related orientational polarizability values,
which typically translates into a narrow
range of solvents or solvent mixtures. The
final step in creating a polarity reference
scale, is correlating the solvent’s or solvent
mixture’s orientational polarizability (Df )
to its dielectric constant (e).
To estimate the polarity of a bio-
molecular cavity, one then experimen-
tally determines the Stokes shift for the
biomolecule-probe conjugate. Using the
reference scale generated by the Lipper–
9
state, derived by Lippert and Mataga et al.
1
0,11
(
eqn (2)).
In this equation, mabs − mem is
the difference in absorption maximum and
the emission maximum in cm (the Stokes
−
1
shift), h represents Planck’s constant and c
is the speed of light. The l
E
and l param-
G
ꢀ
eters represent the dipole moment of the
excited and ground state, respectively, and
a is the radius of the cavity occupied by the
fluorophore that is assumed to reside in a
continuum of unified dielectric constant.
The solvent molecules ability to ac-
commodate a photoinduced dipole has a
pronounced influence on the fluorescence
properties of polarity sensitive fluoro-
phores. Typically, the observed changes in
emission are correlated to the orientational
polarizability, Df , of a solvent mixture.
This parameter expresses the electronic
and motional polarizability of solvent
molecules surrounding the fluorophore in
ꢀ
ꢁ
2
2
e − 1
n − 1
Mataga equation, one can correlate mabs
−
mabs − mem
=
−
2
hc 2e + 1 2n + 1
mem to a specific Df value (Fig. 2 path a).
2
(l
E
− l
G
)
This Df value is then subsequently
correlated to a corresponding e using
×
+ const
(2)
3
a
2
524 | Org. Biomol. Chem., 2007, 5, 2523–2528
This journal is © The Royal Society of Chemistry 2007

View Article Online
1
7
18
polarity and its sequence dependency,
structurally related to dansyl. Both fluo-
rophores share the naphthalene core sub-
stituted with a dimethylamine electron
donating functionality in conjugation with
an electron withdrawing moiety leading
to strong charge transfer bands. With
DAN connected to a natural nucleobase
is the dansyl [1-(dimethylamino)-naphth-
alene-5-sulfonyl] fluorophore (2). It has
been connected to either the ribose unit
3) or the nucleobase moiety of a
nucleoside
15
(
17,18
(4). Despite its popularity,
spectroscopic data on minimally modified
sulfonamide chromophores are lacking.
The N-methyl sulfonamide derivative 5
was therefore synthesized and its absorp-
tion and emission spectra were taken in
various solvents. Dansylamide 5 is char-
acterized by minor changes in the absorp-
tion spectra but large wavelength shifts and
intensity variations in the emission spectra
see ESI†).
To examine the correlation between
solvent polarity and the changes in ab-
sorption and emission maxima of 5, the
observed Stokes shifts are plotted against
dielectric constants, the orientational po-
22
20,21
via either a rigid (7) or flexible (8)
amide linker, the polarity of the minor
and major groove of B-DNA as well as
2
1
22
A-DNA, Z-DNA and the interior of a
19
23
DNA-binding protein–protein complex
have been estimated.
In the past, polarity sensitive fluoro-
phores attached to long hydrocarbons
have been used to probe biological mem-
20
(
Fig. 2 A schematic illustration showing how to
correlate an experimentally determined Stokes
shift to the corresponding dielectric constant, e,
via the orientational polarizability, Df .
24
branes. A related approach, where amino
acids are substituted with chromophoric
residues, is typically undertaken in
25
exploring local polarity in proteins. One
of the most elegant and recent examples
describes the synthesis and incorporation
of Aladan, a chromophoric amino acid
obtained by conjugating DAN to alanine
(9, Fig. 1). Incorporation of this unnatu-
ral building block into a polypeptide, faci-
litates the estimation of local polarity by
means of steady state and time-resolved
a second correlation (Fig. 2 paths b
and c).
larizability and E (30) values (Fig. 3).
Plotting the Stokes shifts against dielectric
T
Despite the fact that the Lippert–
Mataga theory does not take solvent spe-
cific interactions into account, it does
provide an easy tool to correlate changes
in Stokes shift with polarity. To illustrate
the features and caveats associated with
the parameters discussed above and their
applicability for estimating polarity, three
different fluorescent molecules are ana-
lyzed below.
constants yields a scattered relationship
(Fig. 3a). A more exponential relation is
observed between the Stokes shift and
Df (Fig. 3b). In contrast, a respectable
linear relationship is obtained between the
26,27
Stokes shift and the E (30) indicating that
T
fluorescence.
The useful properties of
this polarity scale describes the solvent-
such chromophores have inspired the deve-
lopment of related fluorophores, including
6DMN (6-N,N-dimethylamino-2,3-naph-
induced changes in Stokes shifts more
accurately than the dielectric constant or
the orientational polarizability.
The example discussed above illus-
trates the superior performance of micro-
scopic solvent polarity scales, such as
28
thalimide) (10) and the smaller 4-DMAP
(4-(N,N-dimethylamino)-phthalimide)
Probing polarity in biomolecules
using dansyl and DAN fluorophores
29
(11), by Imperiali and coworkers (Fig. 1).
The remarkable sensitivity of DAN for
its environmental polarity has been estab-
lished in 1979 with Prodan (6-propionyl-
2-(dimethylamino)-naphthalene, 12), by
The probing of DNA groove polarity was
pioneered by Breslauer and Jin using non-
covalently bound bisbenzimide (Hoechst
Reichardts’s E (30), in portraying the
T
immediate environment of a chromophore.
In studying DNA groove polarity, Majima
et al. implicitly addressed this issue by re-
porting values not only in Debye, but also
14
30
3
3258). More recent approaches have
involved the use of covalently-linked
fluorophores. popular probe, used
to locate base mismatches in DNA,
identify adenosine–adenosine deaminase
Weber and Farris. The Stokes shift data
for Prodan, as reported by the authors, are
plotted against the corresponding dielec-
tric constants (Fig. 4a), the orientational
polarizability (as in the original paper,
21
A
in the more appropriate E (30) value.
T
15
Their studies were conducted with DAN
(6-dimethylamino-2-acylnaphthalene) (6),
another popular fluorophore, which is
16
complexes, and to estimate DNA groove
Fig. 4b), and E (30) values (Fig. 4c) of the
T
Fig. 3 Plots of the stokes shifts of dansyl 5 in various solvents against (a) e, (b) Df and (c) E
T
(30).
This journal is © The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 2523–2528 | 2525

View Article Online
30
Fig. 4 Plot of the Stokes shift of Prodan (12) against (a) e, (b) Df , additional frame displays the more linear part of the curve used for polarity
determination in biocavities, and (c) E (30).
T
3
1
solvents used. The lack of correlation
between the Stokes shifts and dielectric
constants suggests that a solvent bulk
polarity parameter does not adequately
describe the immediate microenvironment
of the chromophore. Although not linear,
a seemingly exponential trend is revealed
upon correlating the Stokes shifts with ori-
entational polarizability similar to dansyl
solvents and solvent mixtures, regard-
less of their solvent specific interactions.
For this reason, cyclohexane and water,
two solvents with obviously very different
characteristics, are found on the same
linear fit (Fig. 4c and ESI). Thus, where
the dansyl probe only indicates the more
Spectroscopic characterization of 13 in
common solvents reveals almost identical
absorption spectra, but displays a remark-
able drop in fluorescence intensity with
concomitant red shift of the emission max-
20
imum (see ESI†). This finding prompted
us to study 13 in a binary mixture of
dioxane–water (Fig. 5b). With this binary
system, in which the two components are
miscible in all ratios, a wide polarity range
appropriate use of the E (30) value, the re-
T
sults obtained with the DAN fluorophore
(
5). This better correlation is intriguing
strongly support the use of the E (30) scale
T
−
1
since this polarity scale still relies on
dielectric constants and refractive indexes,
both bulk solvent parameters (eqn (1)).
Note that to overcome this challenge,
researchers often use a narrow range of this
correlation where a semi-linear correlation
can be deduced (see Fig. 4b).
In contrast to the questionable correla-
tions illustrated above, a very good linear
relationship between the Stokes shifts and
over the use of dielectric constants.
from 2.2 D (E (30) = 36.0 kcal mol ) to
T
−
1
80.2 D (E (30) = 63.1 kcal mol ) was
T
obtained. The stark decrease in emission
An extremely water sensitive
fluorophore
with concomitant red-shift upon increas-
ing water content established the acute
sensitivity of the probe for water or better,
hydrogen bond donating solvents.
To unveil the apparent relationship be-
tween solvent polarity and the minimal
changes in the absorption characteristics
but large shifts in the emission maxi-
mum, the Stokes shifts are plotted against
three solvent polarity parameters: e, Df
Certain fluorophores display considerable
sensitivity to the presence of H-bonding
solvents. In certain cases, such solvents are
excluded from the evaluation due to their
“specific solvent–solute interactions”, as
the approximated Lippert–Mataga corre-
lation does not handle such interactions
E (30) values is observed (Fig. 4c). It is
T
reassuring that, in spite of the large vari-
ation of the solvents used, the best corre-
lation is indeed observed when the Stokes
shifts are correlated to a parameter that
inherently reflects specific solvent–solute
interactions. This observation underlines
the wide applicability of Reichardt’s dye
32
properly. To illustrate an extreme case of
such a sensitive fluorophore, we describe
the photophysical features of 2-phenyl-
and E (30), for both pure solvents (filled
T
circles) and water–dioxane mixtures (open
33
ethynylfluorenone (13). This emissive
fluorenone derivative represents the chro-
mophoric portion of a modified nucleoside
(14) we have recently synthesized and
examined.
circles) (Fig. 5). Notably, the relation be-
tween the e and the Stokes shift seems
to be of a different nature for the pure
solvents than the water–dioxane mixtures.
In the case of the pure solvents, there
(
1) and the resulting polarity scale that
allows for parameterization of virtually all
Fig. 5 Plots of the Stokes shifts of 2-phenylethynylfluorenone (13) in pure solvents (filled circles) and water–dioxane mixtures (open circles) against:
a) dielectric constant, (b) orientational polarizability, and (c) E (30) value.
(
T
2
526 | Org. Biomol. Chem., 2007, 5, 2523–2528
This journal is © The Royal Society of Chemistry 2007

View Article Online
is no clear correlation, while the solvent
mixture displays an exponential correla-
tion. A similar behavior is seen for the
Conclusions, contemplations and
prospects
size, shape and its intrinsic polarizability
profile will inevitably evoke changes within
the biomolecular cavity and its molecular
constitution (e.g., water and ions) that
will taint the readout. The significance
of the problem, coupled to the intriguing
multi-faceted challenges exemplified here,
are likely to accelerate the design and
implementation of smaller and possibly
less intrusive fluorescent probes.
34
correlation of the Stokes shifts with Df .
By using E (30) values, however, both
Fluorescence spectroscopy with cus-
tomized polarity sensitive probes could
be a powerful technique in estimating
the local polarity of biomolecular cavi-
ties. The three fluorophores surveyed here
demonstrate that the use of dielectric
constants, a bulk solvent property, while
not necessarily unacceptable, is signifi-
cantly inferior to employing microscopic,
spectroscopy-based, solvent parameters,
T
pure solvents and solvent mixtures show
a similar linear trend, illustrating again
the benefit of using a microscopic solvent
polarity scale that inherently accounts for
solvent–solutes interactions.
An improved Stokes shift–polarity
correlation
Acknowledgements
such as E (30). The experimentally deter-
T
We thank the National Institutes of Health
(GM069773) for support.
mined Stokes shifts are likely to continue
and serve as the observable of choice in
probing biomolecular cavities, since this
quantity represents a simple measurable
entity that probes both the ground and
excited states. Correlating Stokes shifts to
Df , as shown, suffers from some serious
deficiencies, particularly due to the limited
linearity of this relationship, inherently
resulting from the use of bulk solvent
parameters in calculating Df . The commu-
nity is encouraged to consider using the
modified Lippert–Mataga equation that
incorporates microscopic solvent polarity
parameters. In addition, probing the polar-
ity of a biomolecular cavity by referencing
it to a binary solvent mixture remains
questionable. Is such a limited reference
scale capable of mimicking the plethora of
interactions a probe experiences inside a
confined cavity?
The preferred use of E (30) values over di-
T
electric constants to express changes in po-
larity has been demonstrated by Radhakr-
ishnan and Samanta et al. in determining
excited state dipole moments of coumarine
References
1 A. Warshel and S. T. Russel, Q. Rev. Biophys.,
1984, 17, 283–422; C. N. Schutz and A.
Warshel, Proteins, 2001, 44, 400–417.
35
dyes. In a modified Lippert equation
eqn (3)), the orientational polarizability
term (eqn (1)) has been substituted by the
(
2
E. M. Kosower, An Introduction to Physical
Organic Chemistry, John Wiley & Sons, Inc.,
New York, 1968, pp. 259–264.
N
T
7
unitless E scale, a normalized E (30)
scale.
T
36
3 E. Grunwald and S. J. Weinstein, J. Am.
Chem. Soc., 1948, 70, 846–854.
4
M. H. Abraham, R. M. Doherty, M. J. Kam-
let, J. M. Harris and R. W. Taft, J. Chem.
Soc., Perkin Trans. 2, 1987, 1097–1101.
E. M. Kosower, J. Am. Chem. Soc., 1958, 80,
3253–3260.
m
abs − mem
=
ꢂ
ꢅ
ꢀ
ꢁ
2
ꢃa
ꢄ
3
Dl
Dl
D
N
T
5
6
7
11307.6
E + C (3)
D
a
K. Dimroth and C. Reichardt, Liebigs Ann.
Eqn (3) illustrates the linearity between
Chem., 1969, 727, 93–105.
C. Reichardt, Chem. Rev., 1994, 94, 2319–
changes in the Stokes shift (mabs − mem
)
2
358.
N
T
and the solvent polarity parameter E .
8 H. Langhals, Angew. Chem., Int. Ed. Engl.,
The Dl
D
and a represent the changes of
D
1982, 21, 724–733.
9
E. Lippert, Z. Elektrochem., 1957, 61, 962–
75.
the dipole moment and Onsager radius
9
˚
of Reichardt’s dye (1), 9 D and 6.2 A,
1
0 N. Mataga, Y. Kaifu and M. Koizumi, Bull.
Chem. Soc. Jpn., 1956, 29, 465–470.
37
respectively. The Dl and a are the corre-
sponding quantities of the molecule under
study and c represents a constant. The
In light of the predicaments described
in the introduction, as well as the probe-
dependent correlations and the ambigu-
ity introduced when pure solvents are
compared to solvent mixtures of “identi-
cal” polarity illustrated above, a critical
question surfaces: do the polarity values
previously estimated faithfully reflect the
microenvironments probed? This is, of
course, impossible to universally answer.
It is intriguing to mention, however, that
values determined for the polarity of the
major groove of B-DNA using the general
techniques described here range from 55 D
11 A. Kawski, Acta Phys. Pol., 1966, 507–
18.
5
1
2 J. R. Lakowicz, Principles of Fluorescence
Spectroscopy, Kluwer Academic/Plenum
Publishers, New York, 2nd edn, 1999.
N
ET value for any given solvent or solvent
combination can be calculated based on its
7
E
T
(30) value according eqn (4).
13 S. Y. Fung, J. Duhamel and P. Chen, J. Phys.
Chem. A, 2006, 110, 11446–11454.
E
E
T
(30)solvent − E
T
(30)TMS
14 R. Jin and K. J. Breslauer, Proc. Natl. Acad.
Sci. U. S. A., 1988, 85, 8939–8942.
N
ET
=
(4)
T
(30)water − E
T
(30)TMS
1
1
1
5 K. Yamana, Y. Ohashi, K. Nunota and
The better linear correlation to the
Stokes shift can be attributed to the sub-
H. Nakano, Tetrahedron, 1997, 53, 4265–
4
270.
6 G. Skorka, P. Shuker, D. Gill, J. Zabicky and
A. H. Parola, Biochemistry, 1981, 20, 3103–
3109.
7 D. A. Barawkar and K. Ganesh, Biochem.
Biophys. Res. Commun., 1994, 203, 53–58;
D. A. Barawkar and K. Ganesh, Nucleic
Acids Res., 1995, 23, 159–164.
stitution of bulk solvent parameters e and
N
T
n by E , a microscopic solvent polarity
parameter. Moreover, this approach allows
one to use virtually any solvent, including
solvents for which the e and n param-
eter are outdated, unreliable or simply
17,23
to 70 D.
Although seemingly straightforward,
the experimentally measured polarity, re-
gardless of the type of biomolecular cav-
ity probed, is a reflection of the inher-
ent alterations of the environment under
study, and probably, to a lesser extent,
the polarity of the immaculate biomolec-
ular cavity. Infinitesimally small probes
are obviously non-existent. The probe’s
18 V. R. Jadhav, D. A. Barawkar and K. N.
Ganesh, J. Phys. Chem. B, 1999, 103, 7383–
unknown, since the needed E (30) value
T
7
385.
can be easily determined experimentally.
This advantage is of particular interest if
solvent mixtures are used. It is somewhat
surprising that this improved correlation,
introduced more than a decade ago, has
not been more widely adopted.
1
9 The synthesis and all obtained spectroscopic
data as well as the graphs are added to the
ESI.
2
2
0 All spectroscopic data as well as the graphs
are added to the ESI.
1 T. Kimura, K. Kawai and T. Majima, Org.
Lett., 2005, 5829–5832.
This journal is © The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 2523–2528 | 2527

View Article Online
2
2
2
2 T. Kimura, K. Kawai and T. Majima, Chem.
Commun., 2006, 1542–1544.
29 M. E. V a´ zquez, D. M. Rothman and B.
Imperiali, Org. Biomol. Chem., 2004, 2,
1965–1966.
from: T. M. Aminabhavi and B. Gopala-
krishna, J. Chem. Eng. Data, 1995, 40, 856–
861. A curve fit of both data sets with
Microcal Origin was performed to inter-
polate intermediate values.
3 A. Okamoto, K. Tainaka and I. Saito,
Bioconjugate Chem., 2005, 16, 1105–1111.
4 A. S. Waggoner and L. Stryer, Proc. Natl.
Acad. Sci. U. S. A., 1970, 67, 579–589.
5 L. Stryer, Science, 1968, 162, 526–533.
6 B. E. Cohen, T. B. McAnaney, E. S. Park,
Y. N. Jan, S. G. Boxer and L. Y. Jan, Science,
30 G. Weber and F. J. Farris, Biochemistry, 1979,
18, 3075–3078.
31 All spectroscopic data is tabulated and added
to the ESI.
35 M. Ravi, T. Soujanya, A. Samanta and
T. P. Radhakrishnan, J. Chem. Soc., Faraday
Trans., 1995, 91, 2739–2742.
36 M. Ravi, A. Samanta and T. P. Radha-
krishnan, J. Phys. Chem., 1994, 98, 9133–
9136.
37 C. Reichardt, Solvents and solvent ef-
fects in organic chemistry, VCH, Weinheim,
1988.
2
2
32 M. Mizuta, K. Seio, K. Miyata and M.
Sekine, J. Org. Chem., 2007, ASAP.
33 The synthesis is added to the ESI.
34 Df values were calculated according to Lip-
pert (eqn (2)). Values for e were obtained
from: J. B. Hasted, Water: Compr. Treatise,
1973, 405, and values for n were obtained
2
001, 296, 1700–1703.
2
2
7 M. Sundd and A. D. Robertson, Nat. Struct.
Biol., 2002, 7, 500–501.
8 M. E. V a´ zquez, J. B. Blanco and B. Imperiali,
J. Am. Chem. Soc., 2005, 127, 1300–1306.
2
528 | Org. Biomol. Chem., 2007, 5, 2523–2528
This journal is © The Royal Society of Chemistry 2007