Exploiting long-lived molecular fluorescence

Article abstract of DOI:10.2533/000942903777679424

Fluorophores based on the azo chromophore 2,3-diazabicyclo[2.2.2]oct-2-ene, referred to as fluorazophores, display an exceedingly long fluorescence lifetime. Besides the use in time-resolved screening assays, where the long-lived fluorescence can be time-gated, thereby reproving the signal to background ratio, a distinct application of fluorazophores lies in the area of biopolymer dynamics. For this purpose, one chain end is labeled with a fluorazophore and the other one with an efficient fluorescence quencher. The fluorescence lifetime of the probe/quencher-labeled peptide then reflects the kinetics of intramolecular end-to-end collision. Applications to polypeptides are described and control experiments which establish the nature of the quenching mechanism as a diffusive process requiring intimate probe/quencher contact are described.

Full text of DOI:10.2533/000942903777679424

LAUREATES: AWARDS AND HONORS (SCS)
161
CHIMIA 2003, 57, No. 4
Chimia 57 (2003) 161–167
© Schweizerische Chemische Gesellschaft
ISSN 0009–4293
Exploiting Long-Lived
Molecular Fluorescence
Werner M. Nau*, Fang Huang, Xiaojuan Wang, Huseyin Bakirci, Gabriela Gramlich,
and Cesar Marquez
*Werner Prize Winner 2002
Abstract: Fluorophores based on the azo chromophore 2,3-diazabicyclo[2.2.2]oct-2-ene, referred to as
fluorazophores, display an exceedingly long fluorescence lifetime. Besides the use in time-resolved screening
assays, where the long-lived fluorescence can be time-gated, thereby improving the signal to background
ratio, a distinct application of fluorazophores lies in the area of biopolymer dynamics. For this purpose, one
chain end is labeled with a fluorazophore and the other one with an efficient fluorescence quencher. The
fluorescence lifetime of the probe/quencher-labeled peptide then reflects the kinetics of intramolecular end-
to-end collision. Applications to polypeptides are described and control experiments which establish the
nature of the quenching mechanism as a diffusive process requiring intimate probe/quencher contact are
described.
Keywords: Azoalkanes · Fluorescence · Kinetics · Photochemistry · Peptides
Werner Nau graduated with a M.Sc. in Chem- tional fellowship, and a SNF Profil fellowship.
istry 1992 from St. Francis Xavier Universi- His work has led to the award of several
ty, Canada. His thesis was supervised by prizes, including the 1999 International Gram-
D. Klapstein and dealt with molecular spec- maticakis-Neumann prize for photochemistry,
troscopy (UV, IR, photoelectron) of acyl awarded by the Swiss Group for Photochem-
iso(thio)cyanates. He got his Ph.D. together istry and Photobiology, and the 2000 ADUC-
with W. Adam in 1994 from the University of Prize, awarded for his habilitation thesis.
Würzburg on the EPR and transient absorption
spectroscopy of 1,3-cyclopentanediyl diradi-
cals. Thereafter, Werner Nau spent his post-
doctoral studies with J.C. Scaiano at the Uni-
Fluorescent probes and sensors are
well-established tools in analytical and bio-
logical chemistry, spanning such diverse
applications as calcium ion detection, cell
staining, and polarity sensing [1]. An inter-
esting sub-class of fluorescent probes com-
prises chromophores with a particularly
long fluorescence lifetime, e.g. more than
50 ns [2]. Perhaps the simplest yet very im-
portant application based on long-lived
fluorescence (or generally luminescence)
versity of Ottawa, where he worked on the
mechanistic photochemistry of n,π*-excited
states. In 1996, he joined J. Wirz at the Uni-
versity of Basel, where he has started up with
an independent research group and became
a SNF assistant professor in 2000. In the
same year he finished his habilitation. Since
the fall semester 2002 he has been appointed
as a professor of chemistry at the newly
founded International University Bremen.
His research interests lie in the general
area of physical organic chemistry and fo-
cus on photochemistry, radical chemistry,
supramolecular chemistry, and biomolecular
chemistry, including both synthetic-prepara- relies on the reliable differentiation of long
tive, mechanistic, kinetic, and spectroscopic
aspects. He has recently introduced a novel
lived luminescence components. This is of
class of fluorescent probes, referred to as
fluorescence lifetimes from any shorter-
*Correspondence: Prof. Dr. W.M. Nau
School of Engineering and Science
International University Bremen
Campus Ring 1
interest, in particular, for screening assays
where fluorescent probes are employed to
signal molecular events such as the inhibi-
tion of an enzyme by a library of potential
drugs. Short-lived emission is ubiquitous
and may stem from other additives, sample
impurities, biological components, scat-
tered light, the solvent, or sample container
materials of cuvettes and microplates.
fluorazophores, which are based on the azo
chromophore of 2,3-diazabicyclo[2.2.2]oct-
2-ene. Fluorazophores are applied as sensors
for antioxidants, versatile guest molecules in
supramolecular chemistry, kinetic probes for
biopolymer folding, and fluorophores for time-
resolved screening assays.
Werner Nau has held numerous fellow-
ships, among others a Kekulé and Liebig
fellowship of the Fonds of the Chemical In-
dustry, a NATO fellowship, a NSERC Interna-
D–28759 Bremen
Tel.: +49 421 200 3233
Fax: +49 421 200 3229
E.Mail: w.nau@iu-bremen.de
and
Department of Chemistry
University of Basel
Klingelbergstrasse 80
CH–4056 Basel

LAUREATES: AWARDS AND HONORS (SCS)
162
CHIMIA 2003, 57, No. 4
The shorter-lived components can be
eliminated from detection through a time-
gate, such that the emission from the long-
lived fluorescent label (which serves as the
signaling unit) can be selectively detected.
This reduces the background during the
measurement dramatically. An instructive
example is depicted in Fig. 1, which com-
pares the fluorescence decay of a long-lived
fluorescent probe (τ = 500 ns) with that
of a shorter-lived fluorescing component
(τ = 10 ns), with the latter one, however,
being much more intense (106 times larger
preexponential factor). If as usual the inte-
grated fluorescence intensity would be
compared through steady-state methods, a
‘signal-to-background’ of 0.00005:1 would
result, which would prevent any useful
information to be obtained. If one carries
out the experiment in lifetime mode with
a time-gate between 200–1000 ns and
integrates the areas under the curves, the
‘signal-to-background’ ratio becomes bet-
Fig. 1. Comparison of the fluorescence decay of a long-lived fluorescent
probe (τ = 500 ns) with that of a shorter-lived component (τ = 10 ns);
the shorter-lived component is 106 times more intense (relative preex-
ponential factors). A suitable time gate for use in a time-resolved assay
is shown at 200 ns.
Relatively few organic molecules dis- ent azoalkane and cucurbit[7]uril (CB7), a
ter than 10000:1, an impressive improve- play lifetimes in this long time regime, with barrel-like organic host molecule [5]. Over
ment by more than eight orders of magni- azoalkanes derived from 2,3-diazabicy- the past six years, we have investigated this
tude, which has its underlying reason in the clo[2.2.2]oct-2-ene (DBO) displaying the interesting chromophore in great detail and
exponential decay kinetics of the fluores- longest fluorescence lifetime in solution have prepared several DBO derivatives,
cence. This improvement renders the selec- [3]. The record for the longest fluorescence which we refer to as fluorazophores (‘fluo-
tive detection of long-lived fluorophores in lifetime of an organic chromophore in solu- rescent azo chromophores’). Some of the
so-called ‘time-resolved’ screening assays tion lies currently at 1.03 µs (in aerated investigated derivatives are shown in
generally an entirely instrumental problem. H2O!) [4] and is held by the supramolecu- Scheme 1.
In particular, the real background may be lar complex (CB7•DBO) between the par-
dominated by detector noise rather than
contributions from short-lived emission.
Scheme 1

LAUREATES: AWARDS AND HONORS (SCS)
163
CHIMIA 2003, 57, No. 4
A representative synthetic sequence to ob-
tain Fmoc-Fluorazophore is shown in
Scheme 2, which has been scaled up to af-
ford typically 3 g of Fmoc-DBO in an over-
all yield of 10%, sufficient for the commer-
cial synthesis of up to 30 polypeptides
(10 mg scale).
The intriguing photophysics of fluo-
razophores has been worked out in great de-
tail [6–11] and the fascinating quenching
pathways have been investigated through a
combination of experimental and theoreti-
cal methods [12–14]. On the more applied
side, fluorazophores have proven to be use-
ful as sensors for antioxidant activity, both
in solution [15][16] as well as in mem-
brane-mimetic systems [17], as probes for
measuring the kinetics of association with
supramolecular systems [18][19], as tools
to investigate the geometries of cyclodex-
trin inclusion complexes by means of cir-
cular dichroism [20][21], and as probes for
the polarizability inside molecular contain-
er molecules [22].
Most recently, we have employed fluo-
razophores to measure the kinetics of end-
to-end collision in biopolymers (kcoll in
Scheme 3), including peptides [23][24],
Scheme 2
and oligonucleotides [25]. For this purpose,
the fluorazophore (P) is attached to one end
of the biopolymer and an efficient (nearly
diffusion-controlled) fluorescence quencher
(Q) is attached to the other side; in peptides
this quencher moiety is usually tryptophane
and in oligonucleotides guanine. The kine-
tics of fluorescence quenching can then be
equated to the rate of end-to-end collision,
which has proven difficult to obtain accu-
rately by alternative techniques.
Keeping in mind that alternative me-
thods have other advantages [26][27], the
fluorazophore approach presents, arguably,
the most sensitive and most accurate tool
for measuring end-to-end contact in
biopolymers known to date. Note that this
application is made possible by the exceed-
ingly long fluorescence lifetime, which al-
lows the fluorazophore to ‘wait’ sufficient-
ly long until it is being approached (and im-
Scheme 3
mediately quenched) by the other end; this [29], and for understanding protein domain and intermolecular encounter complex for-
diffusive approach of the chain ends takes motions [30]. The size of biopolymers and mation between two fragments.
10 ns to 1 µs in aqueous solution. The life- the effects of solvation, including salt ef-
Consider Scheme 4a, the intermolecular
time of fluorazophores (ca. 505 ns in D2O fects, still present a major challenge to com- case. We assume an ideal solution with no
under air) is therefore ideally suited for in- putational chemistry, which demands ex- enthalpic interactions between the probe
vestigations in this time regime.
Being able to measure the absolute rates provide invaluable benchmark values for concentration of the encounter complex,
for the motions within biopolymers or at the calibration of theoretical models [31]. CPQ, can be obtained according to Eqn. (1)
least knowing the time scale for these A simple problem already arises if one with Ve being the volume of the encounter
perimental data. The latter, in turn, may (P) and the quencher (Q) molecules. The
processes is of fundamental importance for attempts to predict the time scale of diffu- complex with radius a (spherical approxi-
understanding the mechanism of protein sion-controlled end-to-end collision in mation). CP0 and CQ0 are the total concen-
folding [28], for predicting the kinetics of (bio)polymers from the available rates for trations of P and Q. The ‘equilibrium con-
intramolecular reactions in biopolymers diffusion-controlled intermolecular reac- stant’ for encounter complex formation is
(formation of hydrogen bonds, cystine tions. To allow a comparison, it is useful to then given by Eqn. (2), assuming that the
bridges, proton transfer, electron transfer) compare the probability for intramolecular concentration of molecules in contact is

LAUREATES: AWARDS AND HONORS (SCS)
164
CHIMIA 2003, 57, No. 4
ly function of the chain for the intramolec-
ular case is to limit the distance between
probe and quencher.
(5)
Use of the respective expressions for the
equilibrium constants for intermolecular
and intramolecular encounter complex for-
mation then affords Eqn. (6). If one further
expresses the intermolecular collision rate
constant through the relationship between
the intermolecular diffusion rate constant
and the diffusion coefficient (
),
one obtains Eqn. (7), with D being the mu-
tual intermolecular diffusion coefficient.
Eqn. (7) provides the ideal relationship
Scheme 4
small (dilute solution, i.e. CP ≈ CP0 and CQ which assumes for the case of a very long between the unknown rate constant for in-
≈ CQ0). It follows that the equilibrium con- chain (N >> 1) that the concentration of tramolecular end-to-end collision and the
stant for encounter complex formation (in molecules in contact is small (CA ≈ CA + known (diffusion-controlled) rate constant
units of m3mol–1) equals the volume of the CA'). Again, this relates directly to the loss for an intermolecular probe-quencher pair.
encounter complex multiplied with the of entropy associated with the formation of Accordingly, the rate of end-to-end colli-
Avogadro constant (Na). This in turn corre- an end-to-end encounter complex within an sion in a biopolymer increases linearly with
sponds, due to the absence of enthalpic ef- ideal chain, i.e. ∆collS0 = RlnKintra
fects, to the loss of entropy associated with
.
the diffusion coefficient and size of the en-
the formation of the encounter complex, i.e.
∆
collS0 = Rln(1000KinterC0), where the fac-
(3)
tors 1000 and C0, the standard concentra-
tion in M, are added to give a dimensionless
equilibrium constant.
(1)
(2)
(4)
The intramolecular case, Scheme 4b,
describes the pertinent situation for en-
The kinetics of end-to-end contact for- counter complex (radius a), it depends in-
counter complex formation within
a
mation for the intramolecular reaction can versely on the cubed chain segment length
biopolymer chain labeled with a probe (P) be related to that of the intermolecular re- (b), and it decreases with increasing chain
and a quencher (Q) at opposite ends. For action (Eqn. (5)) if one reduces the equilib- length (N) with the characteristic exponent
simplicity, we assume that the chain does rium constants to ratios of microscopic rate of –3/2. Eqn. (7) has been derived in a diffe-
not introduce additional interactions be- constants and considers further that the ele- rent context by Szabo, Schulten, and Schul-
tween probe and quencher except to restrict mentary rate of dissociation of the en- ten through an analysis of the first passage
the distance by which they can diffuse apart counter complex must be identical for both time of end-attached groups based on a
(ideal chain, Gaussian chain). Like for the species within the approximations made modified Smoluchowski equation [32][33].
intermolecular case, we assume no en-
thalpic interactions between the probe and
the quencher residues. The fraction of
chains with the two ends in contact, (CA'), can
be obtained in this case from the distribu-
tion function (g) in Eqn. (3), which has ana-
lytical solutions for a very long chain (N >> 1,
with N the number of chain segments)
and for the shortest chain (N = 2). r is taken
as the distance between the chain ends and
b equals the length of an individual chain
segment. The ‘equilibrium constant’for end-
to-end encounters is then given by Eqn. (4),
(
). Recall, in particular that the on-
(6)
(7)

LAUREATES: AWARDS AND HONORS (SCS)
165
CHIMIA 2003, 57, No. 4
To apply Eqn. (7), one may use com- with different length are listed in the Table
The dependence of the collision rates on
monly accepted intermolecular diffusion and range from 10–75 ns. Subject to the as- the chain length as derived in Eqn. (7) pre-
coefficients of 10–5 cm2s–1 in water (the sumption of diffusion-controlled quench- dicts a linear increase of the logarithmic
mutual diffusion coefficient in Eqn. (7) cor- ing, the collision rates can be directly ob- collision rates with the logarithm of the
responds to twice this value), a van-der- tained through a correction for the inherent number of chain segments (N). The corre-
Waals reaction radius of 5 Å, and a chain fluorescence lifetime (τ0) according to Eqn. sponding plot (Fig. 3) for the experimental
segment length of 5 Å (typical for one pep- (8). The data for the collision rate constants data shows that this linear relationship is
tide unit). The resulting estimates for the in- demonstrate that end-to-end contact forma- not observed. Instead, one obtains a plot
tramolecular collision times (1/kcoll) are tion in short polypeptides may occur as fast with a strong negative curvature. Moreover,
0.2 ns for N = 4 and 2.7 ns for N = 20. This as 10 ns in water, significantly faster than the theoretical slope [34][35] of –3/2 is on-
means that end-to-end contact formation in previous estimates of rates for peptides ly reached for the longer chains. These vari-
biopolymers in solution may occur as fast with the same length, but also substantially ances indicate deviations from the ideal
as several ns according to the simplest the- smaller than expected from the ideal-chain behavior, which are presumably related to
oretical framework. These theoretical rates behavior according to Eqn. (7) (see above). effects of chain stiffness, which impose an
can now be compared with the experimen- Presumably, the diffusion coefficients of increased internal friction for end-to-end
tal results obtained from the fluorazophore the peptide chain ends are much smaller collision in the short chains [36].
probe/quencher technique in synthetic than those of the free probe and quencher;
The interpretation of the fluorescence
polypeptides. the use of smaller diffusion coefficients in lifetimes in terms of end-to-end collision
Our initial experimental study focused Eqn. (7) than the intermolecular ones would rate constants requires a collision-induced
on the length dependence of the end-to-end bring the theoretical data much closer to the fluorescence quenching, i.e. probe and
collision rates in peptides with the general experimental ones [36].
structure Trp-(Gly-Ser)n-DBO-NH2 (struc-
ture below); these peptides are water solu-
quencher must come into van-der-Waals
contact (2–3 Å distance) for quenching to
occur. This is naturally fulfilled for quench-
ing by hydrogen atom transfer or exciplex-
induced quenching, which are the two
prototypal quenching mechanisms of the
DBO chromophore [3][6][8][13][14][37–39].
However, quenching by electron transfer,
which presents an alternative quenching
mechanism, could operate through bond
(superexchange mechanism); it could also
occur over a considerable distance through
space or through the solvent (up to 5–8 Å)
and must therefore be excluded [40]. The
same applies for Dexter-type triplet energy
transfer, which has been employed in other
intramolecular probe/quencher pairs to as-
sess end-to-end contact formation [26]
[41][42]. Quenching over larger distances
than van-der-Waals contact would result in
a continuum of distance-dependent rate
constants, which could not be analyzed in
terms of a diffusion-controlled collision
ble and presumed to be ‘structureless’ ac-
cording to a previous study [26]. This is an
important requirement to apply theories
based on Gaussian chain behavior [34][35].
kcoll = 1/τ – 1/ τ0
(8)
Table. Fluorescence lifetimes and end-to-
end collision rate constants for Trp–(Gly–
Ser)n–DBO–NH2 polypeptides
n
0
Na
2
τ / nsb
23.3
14.3
19.5
30.5
45.6
74.4
kcoll / 107 s–1c
4.1
6.8
4.9
3.1
2.0
1.1
1
4
2
6
4
10
14
22
6
10
The fluorescence decays of all peptides
as recorded with the time-correlated single
photon counting technique were strictly
monoexponential (Fig. 2). The resulting
a
Number of peptide units between probe and
b
quencher. Fluorescence lifetime in aerated
D2O at 23 °C measured by time-correlated
single photon counting. Obtained from Eqn. 8
c
fluorescence lifetimes (τ) for the peptides with τ0 = 500 ns.
Fig. 3. Double-logarithmic plot of the end-to-end collision rate con-
Fig. 2. Fluorescence decays (time-correlated single photon counting, stants (kcoll) of Trp–(Gly–Ser)n–DBO–NH2 polypeptides versus the pep-
normalized intensity) of Trp–(Gly–Ser)n–DBO–NH2 polypeptides (n = 1, tide length, taken as the number of intervening peptide units (N). The
2, 4, and 6). The lowest trace corresponds to n = 1, the uppermost one dashed line has a slope of –3/2 and is shown to illustrate the deviation
to n = 6.
from the theoretical behavior (Eqn. 7).

LAUREATES: AWARDS AND HONORS (SCS)
166
CHIMIA 2003, 57, No. 4
process. In this case, quenching presents no
longer an elementary reaction to which a
single rate constant can be assigned. In fact,
fluorescence resonance energy transfer,
which operates over even larger distances,
has proven inapplicable to obtain the perti-
nent elementary rate constants [26][43].
In view of the possible complications
due to distance-dependent quenching rates,
it appeared compulsory to establish experi-
mentally that quenching through bond or
through the solvent do in fact not apply for
the DBO/Trp probe/quencher pair. We have
therefore performed a series of carefully
designed control experiments (Schemes
5–7). Control experiments of this type are
strongly recommended to establish alter-
native methods for assessing end-to-end
contact formation, in particular if triplet
energy transfer or electron transfer (both
Scheme 5
of which are candidates for distance-de-
pendent quenching rates) are the postulated
quenching mechanisms.
In the first experiment (Scheme 5), we
have compared the fluorescence lifetime of
the shortest peptide, in which probe and
quencher are directly attached, with the
longer ones. The lifetime of the shortest
peptide is in fact longer than for the next
longer one (Table), which speaks strongly
against a through-bond quenching mecha-
nism, but can be understood in terms of
internal friction (steric hindrance effects)
[36]. We encounter this effect in daily life:
It is more difficult to make a knot in a very
short rope than in a longer one.
In the second experiment (Scheme 6),
we have exchanged the presumably flexible
amino acids glycine and serine in the back-
bone of the peptide by rigid cyclic proline
spacers. The proline peptide has a much
longer lifetime than the glycine-serine one,
close to the lifetime in the absence of
quencher (505 ns in D2O). This suggests
that quenching through bond is unlikely
since the number of bonds remains identi-
cal in both species. The effect of rigidifying
the backbone provides also strong evidence
that it is the diffusion between the chain
ends which is decisive for the quenching
process. Incidentally, it should also be men-
tioned that any increase in the solvent vis-
cosity, as it can be achieved, for example,
through the addition of denaturants like
urea (5 M) or guanidinium chloride (6 M)
also decreases the end-to-end collision
rates of flexible DBO/Trp polypeptides,
consistent with a diffusive process.
Scheme 6
In the third experiment (Scheme 7), we
have left the peptide backbone unchanged,
but have added cucurbit[7]uril (CB7) to the
aqueous solution of the peptide. As demon-
strated by NMR experiments, CB7 com-
plexes selectively and quantitatively the
Scheme 7

LAUREATES: AWARDS AND HONORS (SCS)
167
CHIMIA 2003, 57, No. 4
DBO chromophore (K = 4 × 105 M–1) and pressing background fluorescence through
thereby provides a ‘protective shield’ time-resolved detection (see Fig. 1).
around the chromophore.[22] This shield
[24] F. Huang, W.M. Nau, Angew. Chem. Int.
Ed. 2003, 42, in press.
[25] X. Wang, W.M. Nau. Unpublished results.
[26] O. Bieri, J. Wirz, B. Hellrung, M.
Schutkowski, M. Drewello, T. Kiefhaber,
Proc. Natl. Acad. Sci. USA 1999, 96,
9597–9601.
[27] L.J. Lapidus, W.A. Eaton, J. Hofrichter,
Proc. Natl. Acad. Sci. USA 2000, 97,
7220–7225.
[28] A.R. Fersht, Curr. Opin. Struct. Biol.
1997, 7, 3–9.
[29] D. Pogocki, E. Ghezzo-Schöneich, C.
Schöneich, J. Phys. Chem. B 2001, 105,
1250–1259.
[30] S. Hayward, Proteins 1999, 36, 425–435.
[31] I.-C. Yeh, G. Hummer, J. Am. Chem. Soc.
2002, 124, 6563–6568.
[32] A. Szabo, K. Schulten, Z. Schulten, J.
Chem. Phys. 1980, 72, 4350–4357.
[33] R.W. Pastor, R. Zwanzig, A. Szabo, J.
Chem. Phys. 1996, 105, 3878–3882.
[34] U.W. Suter, M. Mutter, P.J. Flory, J. Am.
Chem. Soc. 1976, 98, 5740–5745.
[35] M. Mutter, U.W. Suter, P.J. Flory, J. Am.
Chem. Soc. 1976, 98, 5745–5748.
[36] X. Wang, E.N. Bodunov, W.M. Nau, sub-
mitted for publication.
[37] W.M. Nau, U. Pischel, Angew. Chem. Int.
Ed. 1999, 38, 2885–2888.
[38] W.M. Nau, G. Greiner, H. Rau, M. Olivuc-
ci, M.A. Robb, Ber. Bunsen-Ges. Phys.
Chem. 1998, 102, 486–492.
[39] W.M. Nau, G. Greiner, H. Rau, J. Wall, M.
Olivucci, J.C. Scaiano, J. Phys. Chem. A
1999, 103, 1579–1584.
[40] A.K. Mishra, R. Chandrasekar, M. Farag-
gi, M.H. Klapper, J. Am. Chem. Soc. 1994,
116, 1414–1422.
[41] W.G. McGimpsey, L. Chen, R. Carraway,
W.N. Samaniego, J. Phys. Chem. A 1999,
103, 6082–6090.
[42] P.J. Wagner, P. Klán, J. Am. Chem. Soc.
1998, 121, 9626–9635.
[43] E. Haas, E. Katchalski-Katzir, I.Z. Stein-
berg, Biopolymers 1978, 17, 11–31.
prevents van-der-Waals contact with the Acknowledgements
This work was generously supported
through the Swiss National Science Foundation
(grant No. 620–58000). The studies were also
performed within the Swiss National Research
Program “Supramolecular Functional Materi-
als” NFP 47 (grant No. 4047–057552).
quencher, which is still free to diffuse since
solvent and peptide backbone have re-
mained unchanged. The resulting lifetime
of the complexed peptide was found to be
1.03 µs, suggesting that the quencher is not
able to quench the excited probe at all (the
longer lifetime than in D2O results from the
exclusion of oxygen and the solvent from
the cavity). This result provides strong evi-
dence for the view that quenching requires
intimate contact. If quenching would occur
through space or through the solvent it
should have also been mediated through the
supramolecular wall.
It follows from the control experiment
in Scheme 6, that the fluorescence lifetimes
of DBO/Trp end-labeled peptides are a
quantitative measure of the flexibility or
rigidity of the peptide backbone. We have
therefore most recently synthesized a series
of random-coil peptides, in which Trp and
DBO are separated by a sequence of identi-
cal amino acids (see structure below) [24].
Each peptide has a characteristic fluores-
cence lifetime, which can be interpreted in
terms of the conformational flexibility,
which a particular amino acid imposes on
the backbone. This allows one to define a
conformational flexibility scale for amino
acids in peptides. The following order of
flexibility applies, where glycine gives rise
to the most flexible peptide and proline pro-
duces the most rigid one:
Received: February 25, 2003
[1] B. Valeur, ‘Molecular Fluorescence’,
Wiley-VCH, Weinheim, 2002.
[2] W.M. Nau, X. Wang, ChemPhysChem
2002, 3, 393–398.
[3] W.M. Nau, EPA Newsl. 2000, 70, 6–29.
[4] C. Marquez, W.M. Nau. Unpublished re-
sult.
[5] a) W.L. Mock, in ‘Comprehensive Supra-
molecular Chemistry’, vol. 2, Ed.:
F. Vögtle, Elsevier, Oxford, 1996, pp.
477–493; b) J. Kim, I.S. Jung, S.Y. Kim,
E. Lee, J.K. Kang, S. Sakamoto,
K. Yamaguchi, K. Kim, J. Am. Chem. Soc.
2000, 122, 540–541.
[6] U. Pischel, X. Zhang, B. Hellrung, E.
Haselbach, P.-A. Muller, W.M. Nau, J.
Am. Chem. Soc. 2000, 122, 2027–2034.
[7] U. Pischel, W.M. Nau, J. Phys. Org.
Chem. 2000, 13, 640–647.
[8] U. Pischel, X. Allonas, W.M. Nau, J. Inf.
Recording 2000, 25, 311–321.
[9] U. Pischel, W.M. Nau, J. Am. Chem. Soc.
2001, 123, 9727–9737.
[10] U. Pischel, W.M. Nau, Photochem. Photo-
biol. Sci. 2002, 1, 141–147.
[11] D. Klapstein, U. Pischel, W.M. Nau, J.
Am. Chem. Soc. 2002, 124, 11349–11357.
[12] W.M. Nau, G. Greiner, J. Wall, H. Rau, M.
Olivucci, M.A. Robb, Angew. Chem. Int.
Ed. 1998, 37, 98–101.
[13] A. Sinicropi, U. Pischel, R. Basosi, W.M.
Nau, M. Olivucci, Angew. Chem. Int. Ed.
2000, 39, 4582–4586.
Gly > Ser > Asp, Asn, Ala > Thr, Leu >
Phe, Glu, Gln > His, Arg > Lys > Val > Ile
> Pro
[14] A. Sinicropi, R. Pogni, R. Basosi, M.A.
Robb, G. Gramlich, W.M. Nau, M.
Olivucci, Angew. Chem. Int. Ed. 2001, 40,
4185–4189.
[15] W.M. Nau, J. Am. Chem. Soc. 1998, 120,
12614–12618.
[16] X. Zhang, C. Erb, J. Flammer, W.M. Nau,
Photochem. Photobiol. 2000, 71, 524–533.
[17] G. Gramlich, J. Zhang, W.M. Nau, J. Am.
Chem. Soc. 2002, 124, 11252–11253.
[18] W.M. Nau, X. Zhang, J. Am. Chem. Soc.
1999, 121, 8022–8032.
[19] X. Zhang, G. Gramlich, X. Wang, W.M.
Nau, J. Am. Chem. Soc. 2002, 124,
254–263.
[20] X. Zhang, W.M. Nau, Angew. Chem. Int.
Ed. 2000, 39, 544–547.
[21] B. Mayer, X. Zhang, W.M. Nau, G. Mar-
coni, J. Am. Chem. Soc. 2001, 123,
5240–5248.
[22] C. Marquez, W.M. Nau, Angew. Chem.
Int. Ed. 2001, 40, 4387–4390.
In summary, the intramolecular fluores-
cence quenching of fluorazophores pro-
vides a distinct tool for investigations in the
area of biopolymer dynamics. Future stud-
ies will involve oligonucleotides, larger,
structured peptides, mutation effects, the
determination of activation energies, and
the transfer of the kinetic results to applica-
tions in high-throughput screening technol-
ogy, where the long fluorescence lifetime
provides the additional advantage of sup-
[23] R.R. Hudgins, F. Huang, G. Gramlich,
W.M. Nau, J. Am. Chem. Soc. 2002, 124,
556–564.