Fluorescence of cis-1-amino-2-(3-indolyl)cyclohexane-1-carboxylic acid: A single tryptophan χ1 rotamer model

Article abstract of DOI:10.1021/ja016542d

A constrained derivative, cis-1-amino-2-(3-indolyl)cyclohexane-1-carboxylic acid, cis-W3, was designed to test the rotamer model of tryptophan photophysics. The conformational constraint enforces a single χ₁ conformation, analogous to the χ

Full text of DOI:10.1021/ja016542d

Published on Web 10/12/2002
Fluorescence of
cis-1-Amino-2-(3-indolyl)cyclohexane-1-carboxylic Acid: A
Single Tryptophan ø1 Rotamer Model
Bo Liu,† Reema K. Thalji,‡ Paul D. Adams,† Frank R. Fronczek,‡
Mark L. McLaughlin,*,‡ and Mary D. Barkley*,†
Contribution from the Department of Chemistry, Case Western ReserVe UniVersity,
CleVeland, Ohio 44106, and Department of Chemistry, Louisiana State UniVersity,
Baton Rouge, Louisiana 70803
Received July 3, 2001
Abstract: A constrained derivative, cis-1-amino-2-(3-indolyl)cyclohexane-1-carboxylic acid, cis-W3, was
designed to test the rotamer model of tryptophan photophysics. The conformational constraint enforces a
single ø1 conformation, analogous to the ø1 ) 60° rotamer of tryptophan. The side-chain torsion angles in
the X-ray structure of cis-W3 were ø1 ) 58.5° and ø2 ) -88.7°. Molecular mechanics calculations suggested
two ø2 rotamers for cis-W3 in solution, -100° and 80°, analogous to the ø2 ) (90° rotamers of tryptophan.
The fluorescence decay of the cis-W3 zwitterion was biexponential with lifetimes of 3.1 and 0.3 ns at 25
°C. The relative amplitudes of the lifetime components match the ø2 rotamer populations predicted by
molecular mechanics. The longer lifetime represents the major ø2 ) -100° rotamer. The shorter lifetime
represents the minor ø2 ) 80° rotamer having the ammonium group closer to C4 of the indole ring (labeled
C5 in the cis-W3 X-ray structure). Intramolecular excited-state proton transfer occurs at indole C4 in the
tryptophan zwitterion (Saito, I.; Sugiyama, H.; Yamamoto, A.; Muramatsu, S.; Matsuura,T. J. Am. Chem.
Soc. 1984, 106, 4286-4287). Photochemical isotope exchange experiments showed that H-D exchange
occurs exclusively at C5 in the cis-W3 zwitterion, consistent with the presence of the ø2 ) 80° rotamer in
solution. The rates of two nonradiative processes, excited-state proton and electron transfer, were measured
for individual ø2 rotamers. The excited-state proton-transfer rate was determined from H-D exchange and
fluorescence lifetime data. The excited-state electron-transfer rate was determined from the temperature
dependence of the fluorescence lifetime. The major quenching process in the -100° rotamer is electron
transfer from the excited indole to carboxylate. Electron transfer also occurs in the 80° rotamer, but the
major quenching process is intramolecular proton transfer. Both quenching processes are suppressed by
deprotonation of the amino group. The results for cis-W3 provide compelling evidence that the complex
fluorescence decay of the tryptophan zwitterion originates in ground-state heterogeneity with the different
lifetimes primarily reflecting different intramolecular excited-state proton- and electron-transfer rates in various
rotamers.
The complex photophysics of tryptophan is both a blessing
and a curse. On one hand, the fluorescence quantum yield and
lifetime are highly sensitive to the environment. On the other
hand, this very sensitivity hampers the use of protein fluores-
cence as a structural tool. Despite two decades of work by many
investigators, the heterogeneous fluorescence decays of single
tryptophans in peptides and proteins remain largely un-
interpretable. Even the biexponential fluorescence decay of
tryptophan zwitterion is not fully understood. Over the years,
several explanations have been advanced to account for the
lifetime heterogeneity. Although the origin of the multi-
exponential decays of proteins remains elusive,1 the rotamer
model has prevailed for simple tryptophan derivatives in aqueous
solution.2,3 According to the rotamer model, multiple fluores-
cence lifetimes originate from ground-state rotamers about the
CR-Câ (ø1) and Câ-Cγ (ø2) bonds of the tryptophan side chain.
Presumably, differences in proximity of the indole ring to
substituents that quench the fluorescence cause lifetime differ-
ences among rotamers. Lifetime measurements in supersonic
jets have demonstrated that individual conformers have mono-
exponential fluorescence decays.4,5 Tryptophan has six preferred
rotamers: three ø1 rotamers, (60° and 180°, each with two ø2
rotamers, (90°. NMR6 and molecular modeling7,8 indicate that
all of them exist in solution. Inspection of the structures of
(2) Donzel, B.; Gauduchon, P.; Wahl, P. J. Am. Chem. Soc. 1974, 96, 801-
808.
* To whom correspondence should be addressed at CWRU or LSU.
CWRU Fax: 216-368-0604. E-mail: mdb4@po.cwru.edu. LSU Fax: 225-
388-3458. E-mail: mmclaugh@lsu.edu.
(3) Szabo, A. G.; Rayner, D. M. J. Am. Chem. Soc. 1980, 102, 554-563.
(4) Sipior, J.; Sulkes, M.; Auerbach, R.; Boivineau, M. J. Phys. Chem. 1987,
91, 2016-2018.
† Case Western Reserve University.
‡ Louisiana State University.
(1) Lakowicz, J. R. Photochem. Photobiol. 2000, 72, 421-437.
(5) Philips, L. A.; Webb, S. P.; Martinez, S. J., III; Fleming, G. R.; Levy, D.
H. J. Am. Chem. Soc. 1988, 110, 1352-1355.
9
10.1021/ja016542d CCC: $22.00 © 2002 American Chemical Society
J. AM. CHEM. SOC. 2002, 124, 13329-13338
13329

A R T I C L E S
Liu et al.
individual rotamers reveals differences in proximity of the indole
ring to the ammonium and carboxylate groups. Moreover, there
is compelling evidence for several excited-state reactions in the
tryptophan zwitterion, including proton transfer,9 electron
transfer,10,11 and rotamer interconversion.12 However, several
questions still await more straightforward answers. Why do the
six rotamers have only two fluorescence lifetimes? What are
the lifetimes of individual rotamers? Also, what quenching
mechanisms govern those lifetimes?
The indole chromophore has multiple nonradiative decay
pathways that compete with emission for deactivation of the
excited electronic state and thereby quench the fluorescence.
Our previous work on simple indole and constrained tryptophan
derivatives dissected the different nonradiative pathways by
measuring individual rates.13-17 The total nonradiative rate knr
is
indoles and thereby limit comparison with tryptophan. Here we
present a second generation constrained tryptophan, cis-1-amino-
2-(3-indolyl)cyclohexane-1-carboxylic acid, cis-W3, which closely
mimics the ø1 ) 60° rotamer of tryptophan. With a judiciously
located six-member ring, a single conformation of the CR-Câ
bond of tryptophan is dictated by equatorial placement of the
sterically bulky 3-indolyl substituent. Steric hindrance between
the amino group and protons on the benzene and cyclohexane
rings in the ø2 ) 90° rotamer (indicated by half-circles) strongly
favors the ø2 ) -90° rotamer. Ground-state conformation is
determined by X-ray crystallograpy and molecular modeling.
Excited-state properties are determined by steady-state and time-
resolved fluorescence techniques and photochemical isotope
exchange experiments. The implications of our results for
tryptophan photophysics are discussed.
Experimental Section
knr ) kisc + ksi + kpt + ket
(1)
Synthesis. 2-(3-Indolyl)cyclohexanone. Glacial acetic acid (50 mL)
and 2 N H3PO4 (13 mL) were heated to 135-140 °C, indole (7 g, 0.06
mol; 99%, Aldrich) and (()-2-hydroxycyclohexanone dimer (10 g, 0.8
mol; Aldrich) were added, and the mixture was refluxed for 30 min.
The mixture was cooled in an ice bath and then poured into 150 mL of
ice-cold concentrated ammonium hydroxide with vigorous stirring. The
yellow precipitate was collected by filtration and extracted with ethyl
acetate. The organic phase was washed, dried, and evaporated to
dryness. The gooey residue was dissolved in a small volume of CH2Cl2,
applied to a silica gel column, and eluted with CH2Cl2/ethyl acetate
(19:1 vol/vol). After removal of solvent, a light brown solid was
Intersystem crossing to the triplet state is independent of
temperature and environment in aqueous solution. The inter-
system crossing rate kisc is determined from triplet yield and
fluorescence lifetime measurements.16 Solvent quenching refers
to the major temperature-dependent nonradiative process that
occurs in all indoles in protic solvent and causes a solvent
isotope effect on the fluorescence quantum yield and lifetime
of simple indoles. The solvent quenching rate ksi is obtained
from the temperature dependence of the fluorescence life-
time.13,15 Excited-state exchange of aromatic protons at indole
C2, C4, and C7 occurs at neutral pH in the presence of a good
proton donor, such as ammonium.9,14 The proton-transfer rate
kpt is measured in photochemical isotope exchange experiments
monitored by NMR or mass spectrometry.17,18 Electron transfer
from excited indole to electrophilic acceptors is probably the
most important quenching mechanism in proteins. The electron-
transfer rate ket can be extracted from the temperature depend-
ence of the fluorescence lifetime.16,17
1
obtained (4.3 g, 48% yield). H NMR (200 MHz, CDCl3): δ 8.11 (br
s, 1H), 7.46 (d, J ) 6.0 Hz, 1H), 7.35 (d, J ) 6.0 Hz, 1H), 7.21-7.08
(3H), 3.97-3.92 (dd, J ) 12, 6 Hz, 1H), 2.65-1.83 (8H). EI-MS: calcd
for C14H15NO 213.1154, found 213.1152.
2-(3-Indolyl)cyclohexanespiro-5′-hydantoin. 2-(3-Indolyl)cyclo-
hexanone (2.8 g, 13 mmol) was dissolved in 30 mL of warm methanol,
and (NH4)2CO3 (3.0 g, 31 mmol) was added. NaCN (3.0 g, 61 mmol;
99.99%, Aldrich) dissolved in 35 mL of water was added dropwise to
the mixture in 30 min. Another 3.0 g of (NH4)2CO3 was added, and
the suspension was stirred at 67 °C for 48 h. The final mixture
containing a light yellow precipitate was poured into 250 mL of water,
and the pH was adjusted to 7.0 with 6 N HCl. The solution was cooled
to 0 °C. The white crystalline precipitate was filtered, washed with a
small amount of water, and dried at 80 °C for 3 h (4.32 g, 104% yield;
96% yield after correction for waters of hydration19): mp 159-160
1
°C. IR νmax(KBr): 1724, 1759 (CdO). H NMR (250 MHz, DMSO-
d6): δ 10.88 (s, 1H), 10.01 (s, 1H), 8.50 (s, 1H), 7.52 (d, J ) 7.5 Hz,
1H), 7.27 (d, J ) 7.5 Hz, 1H), 7.09 (s, 1H), 7.00 (t, J ) 7.5 Hz, 1H),
6.90 (t, J ) 7.5 Hz, 1H), 3.34 (dd, J ) 12, 4 Hz, 1H), 1.99-1.42
(8H). 13C NMR (62.5 MHz, DMSO-d6): δ 177.72, 156.92, 135.33,
(9) Saito, I.; Sugiyama, H.; Yamamoto, A.; Muramatsu, S.; Matsuura, T. J.
Am. Chem. Soc. 1984, 106, 4286-4287.
(10) Chang, M. C.; Petrich, J. W.; McDonald, D. B.; Fleming, G. R. J. Am.
Chem. Soc. 1983, 105, 3819-3824.
(11) Petrich, J. W.; Chang, M. C.; McDonald, D. B.; Fleming, G. R. J. Am.
Chem. Soc. 1983, 105, 3824-3832.
(12) Willis, K. J.; Szabo, A. G.; Kracjarski, D. T. Chem. Phys. Lett. 1991, 182,
614-616.
(13) McMahon, L. P.; Colucci, W. J.; McLaughlin, M. L.; Barkley, M. D. J.
Am. Chem. Soc. 1992, 114, 8442-8448.
Our strategy for testing the rotamer model is to reduce the
number of conformational degrees of freedom. Previously, we
studied tetrahydrocarbazoles, which contain 2,3-disubstituted
(14) Yu, H.-T.; Colucci, W. J.; McLaughlin, M. L.; Barkley, M. D. J. Am. Chem.
Soc. 1992, 114, 8449-8454.
(15) Yu, H.-T.; Vela, M. A.; Fronczek, F. R.; McLaughlin, M. L.; Barkley, M.
D. J. Am. Chem. Soc. 1995, 117, 348-357.
(16) Chen, Y.; Liu, B.; Yu, H.-T.; Barkley, M. D. J. Am. Chem. Soc. 1996,
118, 9271-9278.
(17) Chen, Y.; Barkley, M. D. Biochemistry 1998, 37, 9976-9982.
(18) Shizuka, J.; Serizawa, M.; Kobayashi, J.; Kameta, K.; Sugiyama, H.;
Matsuura, T.; Saito, I. J. Am. Chem. Soc. 1988, 110, 1726-1732.
(19) Gauthier, T. J.; Yokum, T. S.; Morales, G. A.; McLaughlin, M. L.; Liu,
Y.-H.; Fronczek, F. R. Acta Crystallogr. 1997, C53, 1659-1662.
(6) Dezube, B.; Dobson, C. M.; Teague, C. E. J. Chem. Soc., Perkins Trans.
2 1981, 730-735.
(7) Engh, R. A.; Chen, L. X. Q.; Fleming, G. R. Chem. Phys. Lett. 1986, 126,
365-372.
(8) Gordon, H. L.; Jarrell, H. C.; Szabo, A. G.; Willis, K. J.; Somorjai, R. L.
J. Phys. Chem. 1992, 96, 1915-1921.
9
13330 J. AM. CHEM. SOC. VOL. 124, NO. 44, 2002

Tryptophan Rotamer Model
A R T I C L E S
126.66, 122.84, 120.59, 118.89, 118.02, 113.78, 110.98, 66.11, 38.71,
34.22, 29.39, 25.70, 20.71. EI-MS: calcd for C16H17N3O2 283.1321,
found 283.1321.
215.1311, found 215.1314. trans-Isomer (0.3 g) was crystallized twice
from benzene/petroleum ether (0.15 g, 13% yield): mp ) 156-157
°C. H NMR (300 MHz, CDCl3): δ 8.10 (br s, 1H), 7.75 (d, J ) 7.5
1
Hz, 1H), 7.40 (d, J ) 8.0 Hz, 1H), 7.23 (t, J ) 7.5 Hz, 1H), 7.14 (t,
J ) 7.5 Hz, 1H), 7.11 (s, 1H), 3.80-3.77 (1H), 2.82-2.75 (1H), 2.19-
1.37 (9H). EI-MS: calcd for C14H17NO 215.1311, found 215.1319.
X-ray Diffraction. X-ray diffraction data were collected on an Enraf-
Nonius CAD4 diffractometer using Cu KR radiation (λ ) 1.54184 Å)
and a graphite monochromator. Crystal data: C15H20N2O2‚H2O, FW
) 276.3; orthorhombic space group P212121; a ) 10.1293(6), b )
10.428(1), c ) 13.195(1) Å; V ) 1393.8(3) Å3; Z ) 4; Dc ) 1.317 mg
m-3; T ) 299 K. Data reduction included corrections for background,
Lorentz, polarization, and absorption. Absorption corrections were based
on ψ scans. The structure was solved by direct methods and refined
using the MolEN programs.21 Refinement was by full-matrix least
squares, with neutral-atom scattering factors and anomalous dispersion
cis-1-Amino-2-(3-indolyl)-cyclohexane-1-carboxylic Acid, cis-W3.
2-(3-Indolyl)cyclohexanespiro-5′-hydantoin (250 mg, 0.88 mmol) and
Ba(OH)2‚8H2O (16 g) were dissolved in 125 mL of water by heating
for 2 h at 80 °C in a 500 mL Parr high-pressure bomb. The solution
was purged for 15 min with argon, and the bomb was sealed and placed
in a sand bath at 230 °C for 4 days. The temperature was lowered to
80 °C, and (NH4)2CO3 (8.0 g) was added to the solution with stirring
for 1 h. The solution was clarified by filtration and concentrated to
about 10 mL under reduced pressure at 70 °C. The light yellow
crystalline product was filtered, washed with 2 mL of water, and dried
under vacuum (106 mg, 58% yield): mp 305-307 °C. cis-W3 was
2
2
corrections. Weights were w ) 4Fo [σ2(I) + (0.02Fo )2]-1. All non-
hydrogen atoms were refined anisotropically. Hydrogen atoms were
refined isotropically. At convergence, R ) 0.030 for 262 parameters
and 2733 observed data.
Molecular Mechanics. Calculations were done on a Silicon Graphics
INDY workstation using the Discover module of InsightII (Molecular
Simulations, Inc.) with the CVFF force field, cross and Morse potential
terms, and the dielectric constant of water at 25 °C, ꢀ ) 78.3. Starting
structures for the calculations were based on the X-ray structure of
cis-W3. The carboxylate was rotated to different positions by changing
the ψ torsion angle N2-C14-C15-O2 from 0 to 360° at 15° intervals.
For each ψ angle, torsional barriers for ø2 rotamers were determined
by driving the ø2 torsion angle C14-C9-C7-C6 at 5° intervals and
constraining it during energy minimization. cis-N-Acetyl-W3 was
constructed from the cis-W3 zwitterion in the Builder module by
removing two H’s from N2, adding an aldehyde fragment to N2 and a
methyl fragment to the aldehyde, and defining the N2-C(carbonyl)
bond as a trans peptide bond. The N-acetyl group was rotated to
different positions by changing the φ torsion angle C(carbonyl)-N2-
C14-C15 from 0 to 360° at 15° intervals. For each φ angle, ø2 torsional
barriers were determined as described above. Rotamer populations were
calculated from a Boltzmann distribution of the relative enthalpy at
298 K.
Nuclear Magnetic Resonance. NOESY experiment was performed
on a Varian Unity-plus 600 MHz NMR spectrometer at ambient
temperature. Ten milligrams of cis-W3 was dissolved in 0.6 mL of
D2O. The pD was adjusted to 11.0 with NaOD (99+% D, Aldrich).
The HDO signal was suppressed by rf saturation. Acquisition parameters
were 6000 Hz (10 ppm) spectral width, 2048 and 256 points in D1
and D2 dimensions (eight scans each), 600 ms mixing time. NMR data
were processed using Felix 95.0 (Molecular Simulations, Inc.). Chemical
shifts were referenced to the HDO signal taken as 4.80 ppm. A sinebell
squared apodization function was used in both D1 and D2 dimensions.
Absorbance and Steady-State Fluorescence. Absorbance was
measured on a Cary 3E UV-vis spectrophotometer. The pK of cis-
W3 was determined from the absorbance change at 270 nm.22 Sample
absorbance was adjusted to <0.1 at 280 nm for steady-state fluorescence
measurements and <0.2 at 290 nm for time-resolved fluorescence
measurements. cis-W3 samples were prepared in 0.01 M phosphate
buffer at the indicated pH(D).
1
recrystallized twice from water. H NMR (300 MHz, DMSO-d6): δ
10.91 (s, 1H), 7.71 (d, J ) 6.0 Hz, 1H), 7.25 (d, J ) 6.0 Hz, 1H), 7.14
(s, 1H), 6.99 (t, J ) 6.0 Hz, 1H), 6.87 (t, J ) 6.0 Hz, 1H), 7.50-7.2
(br s, 3H), 3.53 (dd, J ) 13.5, 3.5 Hz, 1H), 2.15-1.90 (2H), 1.90-
1.79 (2H), 1.70-1.52 (2H), 1.52-1.35 (2H). 13C NMR (75 MHz,
DMSO-d6): δ 172.68, 135.72, 127.22, 123.04, 120.58, 120.27, 117.87,
114.73, 110.65, 63.06, 38.18, 33.61, 28.58, 26.05, 20.67. FAB-MS:
calcd for C15H19N2O2 (MH+) 259.1447, found 259.1447.
+
cis-N-Acetyl-1-amino-2-(3-indolyl)-cyclohexane-1-carboxylic Acid,
cis-N-Acetyl-W3. cis-W3 (70 mg) was dissolved in 1 mL of 2 N NaOH
in an ice bath, and 0.25 mL of acetic anhydride and 4.5 mL of 2 N
NaOH were added dropwise in 15 min. The resulting solution was
stirred at room temperature for 30 min, and the pH was adjusted to 5
by addition of acetic anhydride. The white precipitate was filtered and
washed with 10 mL of water (39 mg, 55% yield): mp 237-238 °C.
cis-N-Acetyl-W3 was crystallized from methanol/water. 1H NMR (300
MHz, DMSO-d6): δ 11.8 (br s, 1H), 10.9 (s, 1H), 7.51 (s, 1H), 7.44
(d, J ) 7.8 Hz, 1H), 7.29 (d, J ) 8.1 Hz, 1H), 7.02 (t, J ) 7.1 Hz,
1H), 6.91 (t, J ) 7.5 Hz, 1H), 6.81 (s, 1H), 2.7 (d, J ) 12 Hz, 1H),
2.2-1.2 (12H). 13C NMR (75 MHz, DMSO-d6): δ 173.9, 169.0, 135.4,
127.4, 123.0, 120.6, 118.4, 118.1, 114.6, 111.1, 61.9, 30.4, 28.6, 25.4,
23.0, 20.6. EI-MS: calcd for C17H20N2O3 300.1473, found 300.1480.
cis- and trans-2-(3-Indolyl)cyclohexanol, cis- and trans-3-InChOH,
were synthesized as described previously.20 The product was a mixture
of cis- and trans-isomers of 2-(3-indolyl)cyclohexanol (0.7 g, 87%
yield). The isomeric mixture was dissolved in a small volume of
chloroform, loaded onto a silica gel column, and eluted with 1%
methanol in chloroform. Fractions were collected and monitored by
TLC. The Rf values of the cis- and trans-isomers are 0.24 and 0.19.
Fractions were pooled and evaporated to dryness. cis-Isomer (0.4 g)
was crystallized from benzene/petroleum ether (0.2 g, 17% yield): mp
) 109-110 °C. 1H NMR (300 MHz, CDCl3): δ 8.08 (br s, 1H), 7.65
(d, J ) 9.0 Hz, 1H), 7.40 (d, J ) 9.0 Hz, 1H), 7.23 (t, J ) 7.5 Hz,
1H), 7.13 (t, J ) 8.0 Hz, 1H), 7.10 (s, 1H), 4.21 (s, 1H), 3.16 (dd, J
) 12, 4 Hz, 1H), 2.06-1.42 (9H). EI-MS: calcd for C14H17NO
Steady-state fluorescence was measured on an SLM 8000 spectro-
fluorometer with single excitation and emission monochromators as
described elsewhere.13 Fluorescence quantum yields ΦF were deter-
mined at 280 nm excitation wavelength, 25 °C, using a value of 0.14
for tryptophan (Sigma, recrystallized four times from 70% ethanol) in
water.23
(21) Fair, C. K. MolEN, An InteractiVe Structure Solution Procedure; Delft:
The Netherlands, 1990.
(22) Hermans, J.; Donovan, J. W.; Scheraga, H. A. J. Biol. Chem. 1960, 235,
91-93.
(20) Freter, K. Liebigs Ann. Chem. 1978, 1357-1364.
(23) Chen, R. F. Anal. Lett. 1967, 1, 35-42.
9
J. AM. CHEM. SOC. VOL. 124, NO. 44, 2002 13331

A R T I C L E S
Liu et al.
Time-Resolved Fluorescence. Fluorescence decays were measured
by time-correlated single photon counting as described elsewhere.16,17
Fluorescence decays were acquired at 288 or 290 nm excitation
wavelength in a PCA3 multichannel analyzer (Oxford Instruments) with
sample alternation. A solution of coffee creamer in water was used for
the instrumental response. Decay data were acquired in 1024 channels
of 5, 9, 25, and 46 ps/channel for cis-W3, and 55 and 64 ps/channel
for the other indoles. Lifetimes of monoexponential standards were
1.06 ns for p-terphenyl (99+%, Eastman Kodak) in ethanol and 2.85-
2.90 ns for N-acetyltryptophanamide (NATA; 99+%, Sigma) in water
at 25 °C.
Decay curves were deconvolved assuming a multiexponential decay
I(t) )
iRi exp(-t/τi)
(2)
∑
where Ri and τi are the amplitude and lifetime of component i. The
goodness of fit was judged by the value of reduced chi-square ør2 and
the shape of the autocorrelation function of the weighted residuals. The
radiative kr and total nonradiative knr rates were calculated from
fluorescence quantum yield and lifetime data.
Figure 1. X-ray structure of cis-W3 with 40% ellipsoids.
τ-1 ) kr + knr
kr ) ΦF/τ
(3)
(4)
to an unirradiated sample. The yield of deuteron transfer ΦD was
calculated from18,27
ΦD ) ΦR[(kH + kD)/kH]
(9)
For multiexponential decays, the average lifetime jτ was used in eqs 3
and 4.
where kH(D) is the rate constant for C-H(D) bond cleavage.
Results and Discussion
τj )
iRiτi/ iRi
∑ ∑
(5)
Synthesis. Racemic cis-W3 is synthesized via a highly
diastereoselective Bucherer-Bergs reaction.28-30 The hydantoin
intermediate in the synthesis of cis-W3 crystallized as the pure
diastereomer.19 Both cis-2-(3-indolyl)cyclohexane-5′-hydantoin
and cis-W3 give a single spot on TLC and no extraneous peaks
in the 1H and 13C NMR spectra. The other diastereomer, trans-
W3, has a smaller Rf value and different NMR spectra from
cis-W3.31
Fluorescence decay curves acquired at various temperatures in the
range 5-60 °C were deconvolved in a global analysis with the
Arrhenius parameters linked24
I(t) ) R exp[-{k0 + A1 exp(-E1*/RT) + A2 exp(-E2*/RT)}t ] (6)
where k0 is the temperature-independent rate, Ai and Ei* are the
frequency factor and activation energy of nonradiative process i, R is
the gas constant, and T is absolute temperature. For indoles16
X-ray Structure. The racemic sample crystallized as a
conglomerate of enantiomerically pure crystals. The absolute
structure chosen for X-ray analysis was arbitrary. cis-W3 exists
as the zwitterion in the monohydrate crystal. The amine N2 is
protonated, and the carboxyl group is deprotonated, as evidenced
by the C-O distances of 1.242(2) and 1.245(2) Å. The
configuration of the zwitterion and the atomic labeling are
illustrated in Figure 1. This numbering scheme is used through-
out the paper. Both carboxylate O atoms accept intermolecular
hydrogen bonds, O1 from the water molecule (O‚‚‚O 2.6414(14)
Å), and O2 from ammonium N2 (N‚‚‚O 2.7238(14) Å). The
cyclohexyl moiety has the chair conformation, with endocyclic
torsion angle magnitudes in the range 51.5(1)-56.6(2)°. It is
twisted with respect to the indole such that the ø2 torsion angle
C6-C7-C9-C14 is -88.7(1)°. The carboxylate is nearly
eclipsed with the ammonium substituent, forming a N2-C14-
C15-O2 torsion angle of 8.7(1)°. The ø1 torsion angle C7-
C9-C14-C15 of 58.5(1)° in cis-W3 mimics the ø1 ) 60°
rotamer of tryptophan.
k0 ) kr + kisc
(7)
Photochemical Isotope Exchange. A solution of cis-W3 in D2O (1
mg/mL) was prepared by sonication, placed in a 1-cm quartz cell, and
irradiated with a 200 W xenon lamp mounted in a LH 150 Spectral
Energy housing. A Corning 7-54 filter was used to remove light <260
nm. H-D exchange at aromatic carbons was monitored on a Varian
300 MHz NMR spectrometer. The HDO resonance was suppressed by
rf saturation. Assignment of indole protons was taken from the
literature.9
The reaction quantum yield ΦR of H-D exchange was determined
at 280 nm excitation (8-nm band-pass) by irradiation with the 400 W
xenon lamp of the SLM fluorometer. A 1-cm quartz cell was placed in
the filter holder at the exit of the excitation monochromator and masked
to control illumination volume.
ΦR ) CVfNA/I0
(8)
C is the concentration of cis-W3 in D2O, V is the volume of solution,
f is the fraction of H-D exchange measured by NMR, NA is Avogadro’s
number, and I0 is the incident light intensity measured by ferrioxalate
actinometry.25,26 Irradiation times of 1.5 h were used to give <10%
H-D exchange. The fraction of H-D exchange was calculated relative
Solution Conformation. The conformation of cis-W3 in
solution was examined by molecular mechanics calculations and
(27) Shizuka, H.; Serizawa, M.; Shimo, T.; Saito, I.; Matsuura, T. J. Am. Chem.
Soc. 1988, 110, 1930-1934.
(28) Bucherer, H. T.; Steiner, W. J. Prakt. Chem. 1934, 140, 291-316.
(29) Bucherer, H. T.; Lieb, V. A. J. Prakt. Chem. 1934, 141, 5-43.
(30) Lopez, C. A.; Trigo, G. G. AdV. Heterocycl. Chem. 1985, 38, 177-228.
(31) Liu, B. Ph.D. Thesis, Case Western Reserve University, 1999.
(24) Beechem, J. M. Chem. Phys. Lipids 1989, 50, 237-251.
(25) Parker, G. A. Proc. R. Soc. London 1953, A220, 104-116.
(26) Hatchard, C. G.; Hatchard, C. A. Proc. R. Soc. London 1956, A235, 518-
536.
9
13332 J. AM. CHEM. SOC. VOL. 124, NO. 44, 2002

Tryptophan Rotamer Model
A R T I C L E S
Figure 2. Torsional barriers for ø2 rotamer interconversion calculated by
molecular mechanics.
NMR. Minimum enthalpy conformations of the cyclohexane
ring were explored from a variety of starting structures. The
chair is the lowest enthalpy conformation, with the boat and
twist boat being at least 3 kcal/mol higher. The ø2 potential
surface was obtained by driving the torsion angle C6-C7-
C9-C14 (Figure 2). Two minimum enthalpy rotamers at about
ø2 ) -105 to -100° and ø2 ) 75 to 80° were found for all
orientations of carboxylate. The -100° rotamer with ψ ) 4.8°
represents the global minimum, differing slightly from the
crystal structure. The relative enthalpy of the ø2 ) 80° rotamer
is 1.1 kcal/mol. The calculated populations are 0.87 and 0.13
for the -100° and 80° rotamers at 25 °C. The lowest torsional
barrier between these two rotamers is about 7.6 kcal/mol. The
plateau at about ø2 ) 10 to 15° represents a rotamer with the
indole ring perpendicular to the plane of ammonium, carbox-
ylate, and R-carbon C14. This metastable rotamer is about 5.1
kcal/mol above the global minimum.
Figure 3. Spectra of the cis-W3 (-) zwitterion and (‚‚‚) anion at 25 °C.
(a) Absorption spectra of 2.9 × 10-5 M cis-W3 at (-) pH 6.5 and (‚‚‚) pH
10.6. Inset is a plot of the absorbance change at 270 nm versus pH: ∆A )
A(pH) - A(pH 7.4); ∆A(max) ) A(pH 11.5) - A(pH 7.4). (b) Fluorescence
emission spectra at 280 nm excitation wavelength at (-) pH 7.0 and (‚‚‚)
pH 10.4.
H9 than H8-H9 is consistent with ø2 ) -100° being the major
conformation. The weak H8-H9 NOE could arise from one or
both ø2 rotamers. However, the antiparallel orientation of H8
and H9 protons in the ø2 ) -100° rotamer may give a reduced
or even negative NOE.34 This together with the parallel
orientation and 1 Å shorter distance in the ø2 ) 80° rotamer
supports the presence of a minor ø2 ) 80° conformation.
In cis-N-acetyl-W3, molecular mechanics calculations identi-
fied four minimum enthalpy conformations, the -100° and 80°
ø2 rotamers of two backbone torsion angles, φ ) 60° and φ )
-45°. The φ ) 60°, ø2 ) -100° conformer represents the global
minimum. The relative enthalpies of the other three conformers
are about 0.7 kcal/mol. The calculated populations are 0.52 for
the φ ) 60°, ø2 ) -100° conformer and 0.16 for each of the
others. The torsional barriers for ø2 rotamer interconversion are
higher in the N-acetyl derivative (9.1 and 11 kcal/mol in the φ
) 60° and -45° conformers) than in cis-W3. The torsional
barriers for interconversion between the two φ conformers are
4.9 kcal/mol in both ø2 rotamers.
Absorbance and Steady-State Fluorescence. Figure 3a
shows the absorption spectra of the zwitterion and anion of cis-
W3, which are identical to the spectra of tryptophan. The
absorption spectrum shifts about 2 nm to the red upon
deprotonation of the ammonium. A pK of 9.5 was determined
for the amino group of cis-W3 from the absorbance change at
270 nm (inset of Figure 3a), in good agreement with the value
of 9.4 for tryptophan.22 These results indicate that the cyclo-
hexane ring in cis-W3 does not perturb the electronic energy
levels of the indole ring.
The 1H NMR spectrum of cis-W3 in D2O shows the
cyclohexane ring in the chair conformation with the indole in
the equatorial position. The cyclohexane H9 couples with the
adjacent diastereotopic CH2-10 methylene and appears as a
resolved double of doublet with coupling constants J of 12.9
and 4.0 Hz. In cyclohexane rings, coupling constants between
axial protons are about 8-14 Hz, whereas coupling constants
between axial and equatorial protons are about 2-5 Hz.32 The
calculated torsion angles H9-C9-C10-H10e and H9-C9-
C10-H10a are 64° and 180°,33 indicating that the cyclohexane
1
ring is quite rigid. The H and 13C NMR spectra at 5 °C show
a single set of resonances. If two ø2 rotamers are present in
solution, either the two rotamers interconvert rapidly on the
NMR time scale or both proton and carbon chemical shifts are
insensitive to environmental differences between rotamers.
A semiquantitative analysis of local NOE intensities suggests
that the predominant rotamer in solution is ø2 ) -100°. The
NOESY spectrum of cis-W3 in D2O shows a strong NOE
between indole H5 and cyclohexane H9 and a weak NOE
between indole H8 and H9 (data not shown). According to
molecular modeling, the H5 and H9 protons are parallel and
2.6 Å apart in the ø2 ) -100° rotamer, as compared to
perpendicular and 4.1 Å apart in the ø2 ) 80° rotamer. The H8
and H9 protons are antiparallel and 3.8 Å apart in the ø2 )
-100° rotamer, but parallel and only 2.8 Å apart in the ø2 )
80° rotamer. The about 8-fold higher NOE intensity for H5-
(32) Bovey, F. A.; Jelinski, L.; Mirau, P. A. Nuclear Magnetic Resonance
Spectroscopy, 2nd ed.; Academic Press: San Diego, 1988.
(33) Colucci, W. J.; Jungk, S. J.; Gandour, R. D. J. Am. Chem. Soc. 1985, 107,
335-343.
(34) Derome, A. E. Modern NMR Techniques for Chemistry Research; Pergamon
Press: Oxford, 1987; pp 122-124.
9
J. AM. CHEM. SOC. VOL. 124, NO. 44, 2002 13333

A R T I C L E S
Liu et al.
Table 1. Quantum Yield and Lifetime Data in H2O and D2O, 25 °Ca
λmax, nm
ΦF
τb (ns)
k (107 s-1
)
knr (107 s-1
)
r
cis-W3
pH 6.5
pD 7.0
pH 11.2
pD 12.2
346
346
359
359
0.14 ( 0.01
0.20 ( 0.01
0.32 ( 0.02
0.46 ( 0.02
2.7 ( 0.1
3.6 ( 0.1
6.5 ( 0.1
10.0 ( 0.1
5.1 ( 0.4
5.6 ( 0.3
4.9 ( 0.3
4.6 ( 0.3
31 ( 1
22 ( 1
10.5 ( 0.4
5.4 ( 0.4
cis-N-acetyl-W3
pH 6.0
pD 6.5
362
362
0.31 ( 0.02
0.46 ( 0.02
7.8 ( 0.1
11.2 ( 0.1
4.0 ( 0.3
4.1 ( 0.2
8.8 ( 0.3
4.8 ( 0.3
tryptophan
pH 7c
352
352
359
0.14
0.29d
0.31
2.6
5.2c
7.6
9.0
5.4
5.6
4.1
33
14
9.1
pD 7
pH 10.5c
pD 9.9e
N-acetyltryptophan
4.8c
5.8d
pH 7
pD 7
363f
0.21c
0.30d
4.4
5.2
16
12
cis-2-(3-indolyl)cyclohexanol
0.32 ( 0.01 8.4 ( 0.1
0.46 ( 0.02 11.8 ( 0.1
pH 6.5
pD 6.5
360
360
3.8 ( 0.1
3.9 ( 0.2
8.1 ( 0.2
4.6 ( 0.2
trans-2-(3-indolyl)cyclohexanol
pH 7.0
pD 7.0
360
360
0.35 ( 0.02
0.48 ( 0.02
8.8 ( 0.1
11.8 ( 0.1
4.0 ( 0.2
4.1 ( 0.2
7.4 ( 0.2
4.4 ( 0.2
3-methylindoleg
pH 7.0
pD 7.0
364
364
0.34
0.50
8.2
12.0
4.1
4.2
8.1
4.1
a Values with standard deviation are from three to five experiments. b Average lifetime for biexponential decays. c Reference 3 at 20 °C. d Reference 37.
e Reference 51. f Reference 40 at 20 °C. g Reference 14.
proton transfer.13,14 The deuterium isotope effect at intermediate
pH is smaller in cis-W3 than in tryptophan. The ratio of quantum
yields in D2O to H2O is only 1.4 for the cis-W3 zwitterion as
compared to 2.1 for the tryptophan zwitterion (Table 1).
Blocking the amino group of cis-W3 by acetylation increases
the quantum yield in H2O and D2O over 2-fold to values
approaching those of cis-W3 anion and 3-methylindole (Table
1). The effect of N-acetylation on quantum yield is less
pronounced in tryptophan.3,37 Replacing the amino acid func-
tional groups at the R-carbon C14 by a hydroxyl group cis or
trans to the indole ring at the â-carbon C9 has similar effects.
The quantum yields of cis- and trans-2-(3-indolyl)cyclohexanol
in H2O and D2O are very close to those of 3-methylindole.
Time-Resolved Fluorescence. Fluorescence decay data were
collected for the constrained tryptophan derivatives in H2O and
D2O at 25 °C (Table 2). Time-resolved emission spectral data
were acquired for the cis-W3 zwitterion and cis-N-acetyl-W3
in H2O. Decay curves at different emission wavelengths were
deconvolved in single curve analyses and in multiple curve
analyses with lifetimes but not amplitudes constrained to be
the same in all of the curves. The data for the W3 zwitterion in
H2O and D2O and cis-N-acetyl-W3 in H2O were best fit by
double exponential functions. The data for cis-W3 anion as well
as cis- and trans-2-(3-indolyl)cyclohexanol gave adequate fits
to single exponential functions (ør2 ) 1.2-1.6).
Figure 4. pH dependence of fluorescence quantum yield of cis-W3 at 25
°C, 280 nm excitation wavelength: (b) H2O and (O) D2O.
Figure 3b shows the emission spectra of the zwitterion and
anion of cis-W3. The emission spectrum shifts about 13 nm to
the red upon deprotonation. The zwitterion emission spectrum
is blue shifted about 6 nm in cis-W3 as compared to that of
tryptophan. The anion emission spectra are identical in cis-W3
and tryptophan. The pH dependence of the fluorescence quantum
yield of cis-W3 in H2O and D2O at 25 °C is shown in Figure
4. The pH profiles are similar to those of tryptophan,35 with
large increases in quantum yield upon deprotonation of the
ammonium and a deuterium isotope effect throughout the entire
pH range. The quantum yield drops at the extremes of pH are
due to two excited-state proton-transfer reactions: acid-catalyzed
protonation of the indole ring with pK* ≈ 3 and base-catalyzed
deprotonation of the indole nitrogen with pK* ≈ 12.36 In
tryptophan, the quenching process at intermediate pH is due to
a regioselective intramolecular excited-state proton transfer
catalyzed by the ammonium.9 The deuterium isotope effect
comprises contributions from water quenching and excited-state
The cis-W3 zwitterion in H2O at 25 °C has lifetimes of 3.1
and 0.33 ns, very close to the lifetimes of tryptophan at 20 °C.3
Unlike tryptophan, the amplitudes of the two decays are
independent of emission wavelength from 330 to 400 nm,
indicating that the two lifetime components have the same
emission spectra. The amplitudes of the longer (0.87 and 0.9)
and shorter (0.13 and 0.10) lifetime components of the cis-W3
(35) Lehrer, S. S. J. Am. Chem. Soc. 1970, 92, 3459-3462.
(36) Vander Donckt, E. Bull. Soc. Chim. Belg. 1969, 78, 69-75.
(37) Kirby, E. P.; Steiner, R. F. J. Phys. Chem. 1970, 74, 4480-4490.
9
13334 J. AM. CHEM. SOC. VOL. 124, NO. 44, 2002

Tryptophan Rotamer Model
A R T I C L E S
Table 2. Time-Resolved Fluorescence Data in H2O and D2O, 25
°Ca
compound
cis-W3
solvent
R1 (λ, nm)
τ1 (ns)
τ2 (ns)
ør2
pH 6.5b 1.0
3.1 ( 0.1
3.6
0.87 ( 0.02 3.1 ( 0.1 0.33 ( 0.02 1.4
pD 7.0b 1.0
0.90 ( 0.02 3.9 ( 0.1 0.40 ( 0.02 1.5
pH 11.2c 1.0
6.5 1.6
0.94 6.6 ( 0.1 1.3 ( 0.1 1.2
pD 12.2c 1.0
cis-N-acetyl-W3 pH 6.0d 1.0
0.67 (330)
3.9 ( 0.1
2.3
10.0 ( 0.1
8.3 ( 0.3
9.7
1.4
4.3
1.6
5.0
0.65 (350)
0.55 (370)
0.54 (390)
pD 6.5c 1.0
11.2 ( 0.1
1.6
a Values with standard deviations are from three to five experiments.
∑iRi ) 1. b Emission wavelength 350 nm. c Emission wavelength 370 nm.
d Global analysis of data acquired at 330-390 nm emission wavelength,
20-nm increments.
zwitterion in H2O and D2O correlate well with the ground-state
populations of 0.87 and 0.13 of the two ø2 rotamers of cis-W3
determined by molecular modeling. This assigns the longer
lifetime component to the major ø2 ) -100° rotamer and the
shorter lifetime component to the minor ø2 ) 80° rotamer.
cis-N-Acetyl-W3 has lifetimes of 9.7 and 5.0 ns with
amplitudes of about 0.6 and 0.4 in H2O at 25 °C. The amplitudes
of the biexponential decay depend slightly on emission wave-
length. The 9.7-ns lifetime of the major component is close to
the value for 3-methylindole, whereas the 5.0-ns lifetime of the
minor component is close to the value of N-acetyltryptophan
(Table 1). The amplitudes of the longer (0.67-0.54) and shorter
(0.33-0.46) lifetime components correlate roughly with the
ground-state populations of the two φ conformers (0.68 for φ
) 60° and 0.32 for φ ) -45°) predicted by molecular modeling.
The fluorescence decay in D2O collapses to a single exponential,
presumably because the two lifetimes become too close to
resolve. Both mono- and biexponential fluorescence decays have
been reported for N-acetyltryptophan.3,38-40
The radiative kr and total nonradiative knr rates were calculated
from quantum yield and lifetime data according to eqs 3-5.
The radiative rates of the cis-W3 zwitterion and anion and of
cis-N-acetyl-W3 are independent of solvent isotope within error.
The kr values of cis-W3 and tryptophan appear to be about the
same. In contrast, the nonradiative rates depend on solvent
isotope. As noted above for quantum yields, the deuterium
isotope effect on knr is less in the cis-W3 zwitterion than in the
tryptophan zwitterion.
Figure 5. Deuterium incorporation upon irradiation of cis-W3 in D2O.
Partial 1H NMR spectra after (top) 0 h, (middle) 1.5 h, and (bottom) 4.5 h
irradiation with a 200 W xenon lamp.
potential proton-transfer sites on the indole ring. The ammonium
N2 can come 1.8 Å closer to C5 in the minor 80° rotamer (N‚
‚‚C 3.5 Å) than in the major -100° rotamer (N‚‚‚C 5.3 Å), and
1.4 Å closer to C8 in the major -100° rotamer (N‚‚‚C 3.5 Å)
than in the minor 80° rotamer (N‚‚‚C 4.9 Å). Therefore, we
expect to observe H-D exchange at both C5 of the minor 80°
rotamer and C8 of the major -100° rotamer.
Photochemical H-D exchange experiments were performed
on cis-W3 in D2O at pD 6.5 at room temperature. Deuterium
incorporation at C5 on the indole ring is apparent in the NMR
spectra of irradiated samples (Figure 5). H-D exchange on other
aromatic carbons was negligible in samples with up to 67%
exchange at C5. The observation of H-D exchange at C5
confirms the presence of the ø2 ) 80° rotamer in solution, as
predicted by molecular modeling. The absence of exchange at
C8 indicates that excited-state proton transfer occurs only in
the 80° rotamer. The reaction quantum yield ΦR ) 0.077 (
0.006 for H-D exchange in D2O was determined by ferrioxalate
actinometry using the total concentration of cis-W3 in eq 8.
The reaction yield of 80° rotamer is ΦR2 ) 0.59 ( 0.04 on the
basis of the mole fraction of this rotamer estimated from
molecular mechanics calculations. The isotope exchange reaction
is proposed to proceed by electrophilic attack of deuteron on
aromatic carbon followed by cleavage of the C-H or C-D
bond.42 Loss of a deuteron is not detectable. For photochemical
isotope exchange reactions with aryl cation intermediates, kH/
kD ) 1.7 ( 0.3 in 20% acetonitrile solvent. The quantum yield
of deuteron transfer in the 80° rotamer is calculated from eq 9.
The resulting ΦD2 ) 0.94 ( 0.07 is used to calculate the
deuteron transfer rate kptD of the 80° rotamer from the lifetime
in D2O (Table 2)
Excited-State Proton Transfer. Intermolecular excited-state
proton transfer catalyzed by good proton donors, such as the
R-ammonium group of amino acid zwitterions, occurs at the
C2, C4, and C7 positions of the indole ring in 3-methyl-
indole.14,41 The corresponding positions in cis-W3 are C8, C5,
and C2. The sites(s) of intramolecular proton transfer in
tryptophan derivatives are governed by stereochemical consid-
erations.18 Inspection of molecular models shows that the two
ø2 rotamers of cis-W3 differ in proximity of the ammonium to
D
kptD ) ΦD2/τ2
(10)
(38) Ross, J. B. A.; Rousslang, K. W.; Brand, L. Biochemistry 1981, 20, 4361-
4369.
(39) Privat, J. P.; Wahl, P.; Brochon, J. C. Biochimie 1985, 67, 949-958.
(40) Eftink, M. R.; Jia, Y.; Hu, D.; Ghiron, C. A. J. Phys. Chem. 1995, 99,
giving kpt ) (2.3 ( 0.2) × 109 s-1
.
D
5713-5723.
(41) Chen, Y.; Liu, B.; Barkley, M. D. J. Am. Chem. Soc. 1995, 117, 5608-
5609.
(42) Shizuka, H.; Tobita, S. J. Am. Chem. Soc. 1982, 104, 6919-6927.
9
J. AM. CHEM. SOC. VOL. 124, NO. 44, 2002 13335

A R T I C L E S
Liu et al.
Table 3. Arrhenius Parameters
We can also estimate proton-transfer rates kpt by assuming
that the lifetime difference between rotamers in Table 2 is due
entirely to excited-state proton transfer.
a
k0
A
E *
ksia
(kcal/mol) (107 s-1
A
E *
ket
1
1
2
2
(107 s-1
)
(s-1
)
)
(s-1
)
(kcal/mol) (107 s-1
)
ør2
cis-W3 zwitterionb
-1
kpt ) τ2-1 - τ1
(11)
14
1.0 × 1016 10.6
17
4.0
3.3
9.2 1.6 × 1017 12.9
5.5 5.8 × 1012
6.1
20
17
3.9 × 1013c
7.3c
D
The lifetime data in D2O give a deuteron transfer rate kpt
)
cis-W3 anionb
6.5
D
(2.2 ( 0.1) × 109 s-1. The kpt values obtained from H-D
exchange and fluorescence lifetime data are essentially the same.
Estimating the proton transfer rate from lifetime data in H2O
8.1 5.1 × 1017 13.5
7.0 1.1 × 1017 12.7
2.0
2.8 1.7
5.4 1.9 × 1012
6.6
3.2
NATAd
H
gives kpt ) (2.7 ( 0.2) × 109 s-1. The kinetic isotope effect
9.8 6.0 × 1016 12.6
3.4 4.5 × 1010
20
on the regioselective intramolecular exchange reaction catalyzed
by the ammonium in cis-W3 appears to be only kptH/kptD ) 1.2.
This value is lower than the 2- to 4-fold isotope effects on
intermolecular exchange in 3-methylindole catalyzed by the
R-ammonium of glycine, the ꢀ-ammonium of lysine, and the
phenolic proton of tyrosine.17
cis-2-(3-indolyl) cyclcohexaneolb
6.8 1.2 × 1017 12.8
5.0
1.7
3-methylindoled
5.9
6.0 1.2 × 1017 12.7
a 25 °C. b Emission wavelength 370 nm. c Determined by Procedure I.
d Reference 16.
Excited-State Electron Transfer. The total nonradiative rate
knr is 3- to 4-fold higher in the zwitterions as compared to
3-methylindole and other compounds in Table 1. In the case of
cis-W3, the 80° rotamer represents about 10% of the population.
Proton transfer from the ammonium to C5 in the 80° rotamer
lowers the average lifetime about 10% relative to the lifetime
of the -100° rotamer and therefore contributes only about 10%
to the nonradiative rate of the cis-W3 zwitterion calculated from
eqs 3 and 5. As noted for the tryptophan zwitterion, there must
be another major fluorescence quenching mechanism besides
proton transfer.27 We propose that electron transfer from excited
indole to carboxylate is the other major quenching mechanism
in the cis-W3 zwitterion.17 Carboxylate does not quench
3-methylindole fluorescence.16 However, protonated carboxyl
is a moderate electron-transfer quencher.17 Presumably, the
carboxylate in cis-W3 and the tryptophan zwitterion is activated
by the proximal ammonium and becomes a better electron
acceptor.10
We can use eq 1 to estimate the electron transfer rate in the
-100° rotamer, because kpt ) 0 in this rotamer. Assuming that
the intersystem crossing and water quenching rates are roughly
the same as they are in tryptophan and NATA16 and using the
values of the radiative rate and τ1 for the cis-W3 zwitterion in
Tables 1 and 2, eqs 1 and 3 give ket ) 1.9 × 108 s-1 for the
-100° rotamer at 25 °C in H2O.
In favorable cases, the electron-transfer rate can be measured
directly from the temperature dependence of the fluorescence
lifetime.16
-1
functions as in eq 2, and an Arrhenius plot of ln(τ1 - k0 -
ksi) versus 1/T was constructed. The k0 value of 1.0 × 108 s-1
was estimated from eq 7 using the radiative rate kr of cis-W3
in Table 1 and the intersystem crossing rate of tryptophan kisc
) 5 × 107 s-1 16
.
The water quenching rate ksi of NATA with
A1 ) 6 × 1016 s-1 and E1* ) 12.6 kcal/mol was also used.16
Values of the frequency factor A2 ) 3.9 × 1013 s-1 and
activation energy E2* ) 7.3 kcal/mol for excited-state electron
transfer are determined from the intercept and slope of the
straight line, giving an electron transfer rate at 25 °C of ket )
1.7 × 108 s-1
.
Procedure II is a global analysis of the temperature data set.
The short lifetime component of the cis-W3 zwitterion represents
<2% of the steady-state intensity, so the decay curves were
fitted to a single-exponential function as in eq 6. Decay curves
at different temperatures were deconvolved in a multiple curve
analysis with k0, Ai, and Ei* constrained to be the same in all
of the curves. The fit assuming a single Arrhenius term gave
ør2 ) 4.0. The fit assuming two Arrhenius terms improved
somewhat to ør2 ) 3.3 with k0 ) 9.2 × 107 s-1, A1 ) 1.6 ×
1017 s-1, E1* ) 12.9 kcal/mol, A2 ) 5.8 × 1012 s-1, and E2* )
6.1 kcal/mol. The electron transfer rate at 25 °C calculated from
A2 and E2* is ket ) 2.0 × 108 s-1. The Arrhenius parameters A1
and E1* determined for water quenching in cis-W3 are slightly
higher than the values in NATA (Table 3). The Arrhenius
parameters for excited-state electron transfer A2 and E2* obtained
by Procedures I and II agree reasonably well. The activation
energies in cis-W3 are about twice the value in NATA, though
the electron-transfer rates at 25 °C for cis-W3 and NATA are
the same.
τ-1 ) k0 + A1 exp(-E1*/RT) + A2 exp(-E2*/RT)
(12)
The first Arrhenius term is the water quenching rate ksi ) A1
exp(-E1*/RT). If electron transfer is the only other temperature-
dependent nonradiative process, the second Arrhenius term gives
the electron transfer rate ket ) A2 exp(-E2*/RT). If electron
transfer is temperature independent, then the temperature-
independent nonradiative rate k0 is larger than that expected
from eq 7, and ket ) ∆k0. There are three nonradiative processes
in the -100° rotamer: intersystem crossing, which is temper-
ature independent, water quenching, and electron transfer.
Fluorescence decays of the cis-W3 zwitterion in H2O were
measured at nine temperatures from 5 to 40 °C. The temperature
data set was analyzed by two methods. Procedure I is a graphical
method. Individual decay curves were fitted to biexponential
Deprotonation of the ammonium reduces the electron transfer
rate in the cis-W3 anion to barely detectable levels. Using eqs
1 and 3 to estimate ket as above gives ket ) 2.1 × 107 s-1. The
temperature dependence of the fluorescence lifetime of the cis-
W3 anion in H2O was measured at seven temperatures from 5
to 55 °C. Global analysis of the temperature data set according
to eq 6 gave ør2 ) 2.0 for a single Arrhenius term, which
improved slightly to ør2 ) 1.7 for two Arrhenius terms. The
electron transfer rate at 25 °C calculated from A2 and E2* is ket
) 2.8 × 107 s-1. The slower electron-transfer rate in the cis-
W3 anion as compared to the zwitterion is consistent with the
proposition that the electrophilicity of carboxyl is enhanced by
the adjacent positively charged amine.10 The water quenching
9
13336 J. AM. CHEM. SOC. VOL. 124, NO. 44, 2002

Tryptophan Rotamer Model
A R T I C L E S
rate at 25 °C calculated from A1 and E1* is ksi ) 5.4 × 107 s-1
,
the tryptophan zwitterion by the adjacent ammonium.10,11 Our
results support their proposed mechanism and provide a value
of about 2 × 108 s-1 at 25 °C for ket in the ø2 ) -100° rotamer
of the cis-W3 zwitterion. Electron transfer rates depend on the
distance between donor and acceptor.48 In the two ø2 rotamers,
the through-bond distances between indole and carboxylate are
identical, and the through-space distances are nearly so. Thus,
the ket values should be about the same for both rotamers,
whether electron transfer occurs through-bond or through-space.
The dominant fluorescence quenching mechanism in the major
-100° rotamer is excited-state electron transfer to carboxylate.
However, the proton transfer rate kpt in the minor 80° rotamer
is an order of magnitude faster than ket. Electron transfer is also
a quenching mechanism in the 80° rotamer, but the dominant
quenching mechanism is regioselective excited-state proton
transfer at C5. Intramolecular proton transfer does not occur at
C8 in the major -100° rotamer of cis-W3, despite similar N2‚
‚‚C5 and N2‚‚‚C8 distances in the 80° and -100° rotamers.
The criteria for proton transfer at a given position are not
apparent from photochemical H-D exchange data or ab initio
calculations on indole derivatives. Intramolecular proton transfer
was likewise observed at the C4 position, but not at the C2
position in the tryptophan zwitterion.9 However, glycine-
catalyzed intermolecular proton transfer in 2- and 3-methyl-
indoles was more extensive at the unsubstituted position of the
pyrrole ring than at the C4 and C7 positions of the benzene
ring.14 The C3 loses π-electron density in the 1La state of indole,
while C2 gains less density than C4 and C7.49 Perhaps, stringent
geometric requirements for the formation of the proposed
tetrahedral transition state in proton exchange preclude intra-
molecular proton transfer in the pyrrole ring of cis-W3 and
tryptophan.
The two fluorescence lifetimes of the cis-W3 zwitterion are
very close to those of tryptophan. It is tempting to speculate,
by analogy with cis-W3, that excited-state electron transfer is
responsible for the longer lifetime of tryptophan. The shorter
lifetime includes both excited-state proton and electron transfer
with regioselective proton transfer at C4 being the dominant
quenching mechanism. Excited-state proton transfer is unlikely
to occur in the two ø1 ) 180° rotamers, because the amino group
is far from the indole ring. However, the ø1 ) (60° rotamers
bring the amino group close to the indole ring, and one ø2
rotamer of each ø1 rotamer has the amino group close to C4.
The two fluorescence lifetimes of cis-N-acetyl-W3 cannot be
so neatly explained by the rotamer model. Nevertheless, we
argue that the observed lifetime heterogeneity is probably rooted
in ground-state heterogeneity as well. First, the low calculated
enthalpy barriers between φ ) 60° and -45° conformers predict
rapid interconversion of φ conformers at rates of 0.4-1.6 ×
109 s-1. This confounds simple interpretation of fluorescence
decay parameters in terms of populations and lifetimes of
ground-state conformers. Conformer interconversion on the
fluorescence time scale may produce lifetime heterogeneity even
in the absence of a lifetime difference among conformers.50
Second, the two intramolecular quenching mechanisms in cis-
W3, excited-state electron transfer to carboxylate and proton
the same as in the zwitterion. The temperature dependence of
the lifetime of cis-2-(3-indolyl)cyclohexanol in H2O was
measured at eight temperatures from 5 to 60 °C. Global analysis
gave an acceptable ør2 value for a single Arrhenius term. The
water quenching rate is ksi ) 5.0 × 107 s-1. The presence of
the cyclohexyl ring does not appear to affect the water quenching
rate. The ksi values for cis-W3, cis-2-(3-indolyl)cyclohexanol,
and 3-methylindole are about the same, although larger than
the value for NATA.
Conclusions
In this paper, we design a constrained tryptophan that provides
a definitive test of the rotamer model for tryptophan photo-
physics in an aqueous environment. We also identify the
intramolecular quenching mechanisms and measure the rates
of excited-state proton and electron transfer in individual ground-
state rotamers. The conformational constraint in cis-W3 elimi-
nates rotation about the CR-Câ bond of the tryptophan side
chain and puts the amino and carboxyl groups at precise
positions that mimic the ø1 ) 60° rotamer of tryptophan. The
electronic properties of the indole ring are preserved, as the
absorption spectra of cis-W3 and tryptophan are identical. The
simulated and experimental time constants for relaxation of
water around excited tryptophan are 1-2 ps,43,44 so the
biexponential fluorescence decay of cis-W3 cannot be due to
solvent relaxation. X-ray crystallography shows a single ø2 )
-90° rotamer of cis-W3. Molecular mechanics calculations
support two ø2 rotamers in solution, -100° and 80°. The
calculated ø2 rotamer populations match the relative amplitudes
of the two fluorescence lifetime components, assigning the major
component with the longer lifetime to the ø2 ) -100° rotamer.
Fluorescence decay times represent lifetimes of individual
ground-state conformations only if the conformers do not
interconvert in the excited state. We may estimate the rotamer
interconversion rate krot1frot2 by transition-state theory
krot1frot2 ) κ(kT/h) exp(-∆Gq/RT)
(13)
where κ is the transmission coefficient, k is the Boltzmann
constant, and h is the Planck constant. Assuming κ ) 1 and
∆Gq ) ∆Hq and using the calculated enthalpy barrier of 7.6
kcal/mol from ø2 ) 80° to ø2 ) -100°, k80°f-100° ) 1.7 × 107
s-1, which is slower than any of the decay rates. The conversion
rate from ø2 ) -100° to ø2 ) 80° is even slower, k-100°f80°
)
2.6 × 106 s-1. Because these estimated rates are likely upper
limits,45 we assume that ø2 rotamer interconversion has a
negligible effect on the fluorescence decay kinetics. However,
these interconversion rates are fast as compared to the NMR
time scale, so the two conformers in fast exchange have only a
single set of resonances.32
The carboxyl group does not quench 3-methylindole fluo-
rescence intermolecularly16,46 or intramolecularly in indole-3-
alkanoic acids47 unless it is protonated. Fleming and co-workers
proposed that the electrophilicity of carboxylate is enhanced in
(43) Simonson, T.; Wong, C. F.; Bru¨nger, A. T. J. Phys. Chem. A 1997, 101,
1935-1945.
(48) Marcus, R. A.; Sutin, N. Biochim. Biophys. Acta 1985, 811, 265-322.
(49) Hahn, D. K.; Callis, P. R. J. Phys. Chem. A 1997, 101, 2686-2691.
(50) McMahon, L. P.; Yu, H.-T.; Shui, L.; Fronczek, F. R.; McLaughlin, M.
L.; Barkley, M. D. J. Phys. Chem. B 1997, 101, 3269-3280.
(51) Gudgin, E.; Lopez-Delgado, R.; Ware, W. R. J. Phys. Chem. 1983, 87,
1559-1565.
(44) Shen, X.; Knutson, J. R. J. Phys. Chem. B 2001, 105, 6260-6265.
(45) Colucci, W. J.; Tilstra, L.; Sattler, M. C.; Fronczek, F. R.; Barkley, M. D.
J. Am. Chem. Soc. 1990, 112, 9182-9190.
(46) Ricci, R. W.; Nesta, J. M. J. Phys. Chem. 1976, 80, 974-980.
(47) James, D. R.; Ware, W. R. J. Phys. Chem. 1985, 89, 5450-5458.
9
J. AM. CHEM. SOC. VOL. 124, NO. 44, 2002 13337

A R T I C L E S
Liu et al.
transfer at C5, are eliminated by N-acetylation of the amino
group. However, electron transfer to the amide bond is expected
to be a quenching mechanism in cis-N-acetyl-W3.16 The
through-bond distances between indole and the amide group of
cis-N-acetyl-W3 are the same in all conformations. On the other
hand, the through-space distances are shorter, and the π-orbital
overlap of the indole ring and the amide bond is better in both
of the φ ) -45° conformers than in the φ ) 60° conformers.
Through-space electron transfer would be faster in the φ ) -45°
conformers, shortening the fluorescence lifetime of the φ )
-45° conformers relative to the φ ) 60° conformers. The stable
φ ) 60° conformers of cis-N-acetyl-W3 are caused by the
cyclohexane ring, as they are not found in the ø1 ) 60° rotamer
of N-acetyltryptophan. This conformational constraint may
account for the long fluorescence lifetime component observed
in cis-N-acetyl-W3.
Acknowledgment. This work was supported by NIH grant
GM42101.
Supporting Information Available: Table of crystal data,
structure solution and refinement, atomic coordinates, bond
lengths and angles, and anisotropic thermal parameters in CIF
format. This material is available free of charge via the Internet
at http://pubs.acs.org.
JA016542D
9
13338 J. AM. CHEM. SOC. VOL. 124, NO. 44, 2002