Photophysics of 7-hydroxytetrahydroisoquinoline-3-carboxylic acid and its derivatives

Article abstract of DOI:10.1021/ja960852s

The following derivatives of 7-hydroxytetrahydroisoquinoline-3-carboxylic acid {Tic(OH) [I]}, a conformationally restricted analogue of tyrosine, were synthesized for the purpose of photophysical studies and in order to elucidate the nature of tyrosine fluorescence and its decay: Ac-Tic(OH) [II], Ac-Tic(OH)-NHMe [III], Tic(OH)-NHMe [IV], Ala-Tic(OH) [V], Ac-Ala-Tic(OH) [VI], and Tic(OH)-Gly-NH₂ [VII]. For the simple Tic(OH) derivatives I-IV, the N-methylamide was found to be a more effective quencher than the acetyl group. For the peptidic derivatives V-VII the highest quenching of the fluorescence of the phenolic chromophore was observed in the case of Ala-Tic(OH). The simple Tic(OH) derivatives I-IV were also the subject of theoretical studies (MOPAC 93). The obtained thermodynamic parameters (MOPAC calculations) and the fluorescence components were discussed on the basis of the rotamer theory in order to explain the participation of an individual rotamer in the complex process of the fluorescence decay of tyrosine.

Full text of DOI:10.1021/ja960852s

8300
J. Am. Chem. Soc. 1996, 118, 8300-8307
Photophysics of 7-Hydroxytetrahydroisoquinoline-3-carboxylic
Acid and Its Derivatives
Wiesław Wiczk,* Krystyna Stachowiak, Piotr Skurski, Leszek Łankiewicz,
Alicja Michniewicz, and Anna R o´ j
Contribution from the Faculty of Chemistry, UniVersity of Gda n´ sk,
Sobieskiego 18, 80-952 Gda n´ sk, Poland
X
ReceiVed March 15, 1996
Abstract: The following derivatives of 7-hydroxytetrahydroisoquinoline-3-carboxylic acid {Tic(OH) [I]}, a con-
formationally restricted analogue of tyrosine, were synthesized for the purpose of photophysical studies and in order
to elucidate the nature of tyrosine fluorescence and its decay: Ac-Tic(OH) [II], Ac-Tic(OH)-NHMe [III], Tic(OH)-
NHMe [IV], Ala-Tic(OH) [V], Ac-Ala-Tic(OH) [VI], and Tic(OH)-Gly-NH2 [VII]. For the simple Tic(OH)
derivatives I-IV, the N-methylamide was found to be a more effective quencher than the acetyl group. For the
peptidic derivatives V-VII the highest quenching of the fluorescence of the phenolic chromophore was observed in
the case of Ala-Tic(OH). The simple Tic(OH) derivatives I-IV were also the subject of theoretical studies (MOPAC
93). The obtained thermodynamic parameters (MOPAC calculations) and the fluorescence components were discussed
on the basis of the rotamer theory in order to explain the participation of an individual rotamer in the complex
process of the fluorescence decay of tyrosine.
Introduction
of the presence of a number of ground-state rotamers, some of
which do not interconvert within the fluorescence time scale
The fluorescence of tyrosine, tyrosine derivatives, and tyrosine
residue in peptides and proteins is the subject of extensive
investigations. The tyrosine zwitterion and derivatives with an
ionized R-carboxyl group exhibit monoexponential decay kinet-
ics. Conversion of the R-carboxyl group to the corresponding
(
typically 3-5 ns). Individual rotamers are assumed to exhibit
monoexponential decay kinetics. This model, introduced by
Gauduchon and Wahl, suggests a charge transfer quenching
3
between the excited aromatic chromophore (indole or phenol
ring) as a donor and electrophilic units in the amino acid
amide or its protonation results in a complex fluorescence
1
5-17,20
backbone (carbonyl or protonated amino group
) as the
decay.1
-4
Several explanations for the complex fluorescence
1
,21
acceptor. As shown by Laws et al., the shorter fluorescence
decay lifetime was associated with the protonated carboxyl
group, while the longer lifetime was associated with ionized
carboxylate. The experimental basis of the rotamer theory was
the observation of Cowgill that the peptide carbonyl, or the
amide group is responsible for the quenching of tyrosine
kinetics of tryptophan and tyrosine have been forwarded,
1
1
5-7
including the involvement of the La and Lb states, an excited
state reaction,8
-10
and different lifetimes for the side-chain
R
â
3,11-22
rotamers around the C -C bond.
multiexponential decay kinetics are proposed to be the result
In the rotamer model,
2
3
X
Abstract published in AdVance ACS Abstracts, August 15, 1996.
24
fluorescence in proteins and the suggestion of Tournon et al.
(1) Laws, W. R.; Ross, J. B. A.; Wyssbrod, H. R.; Beechem, J. M.; Brand,
that the carbonyl groups can quench the fluorescence of aromatic
rings efficiently by a charge transfer mechanism. Each rotamer
has a different distance between the phenolic ring and the
quenching groups (amide and/or carboxyl), which explains the
differentiation in the photophysical behavior of different rota-
mers. Direct interaction between the peptide carbonyl or amide
group and the phenol ring is responsible for the quenching of
the tyrosine fluorescence in peptides and proteins as was
L.; Sutherland, J. C. Biochemistry 1986, 25, 599-607.
(2) Lakowicz, J. R.; Laczko, G.; Gryczy n´ ski, I. Biochemistry 1987, 26,
8
2-90.
(
3) Gauduchon, P.; Wahl, P. Biophys. Chem. 1987, 8, 87-104.
4) Ross, J. B. A.; Laws, W. R.; Rousslang, K. W.; Wyssbrod, H. R. In
(
Topics in Fluorescence Spectroscopy, Vol. 3, Biochemical Applications,
Lakowicz, J. R., Ed.; Plenum Press: New York, 1992, pp 1-63.
(
5) Rayner, D. M.; Szabo, A. G. Can. J. Chem. 1978, 56, 743-745.
(6) Ruggiero, A. J.; Todd, D. C.; Fleming, G. R. J. Am. Chem. Soc.
1
990, 112, 1003-1014.
7) Hansen, J. E.; Rosental, S. J.; Fleming, G. R. J. Phys. Chem. 1992,
6, 3034-3040.
8) Saito, I.; Sugiyama, H.; Yamamoto, A.; Muramatsu, S.; Matsuura,
T. J. Am. Chem. Soc. 1984, 106, 4286-4287.
9) Gudgin, E.; Lopez-Delgado, R.; Ware, W. R. J. Phys. Chem. 1983,
7, 1559-1565.
10) Vekshin, N.; Vincent, M.; Gallay, J. Chem. Phys. Lett. 1992, 199,
2
3,25,26
27,28
(
previously suggested by Cowgill
and Feitelson.
By
9
(
(17) Ross, J. B. A.; Wyssbrod, H. R.; Porter, R. A.; Schwartz, G.;
Michaels, C. A.; Laws, W. R. Biochemistry 1992, 31, 1585-1594.
(18) Tilstra, L.; Sattler, M. C.; Cherry, W. R.; Barkley, M. D. J. Am.
Chem. Soc. 1990, 112, 9176-9182.
(
8
4
8
(
(19) Colucci, W. J.; Tilstra, L.; Sattler, M. C.; Fronczek, F. R.; Barkley,
M. D. J. Am. Chem. Soc. 1990, 112, 9182-9190.
59-463.
(11) Donzel, B.; Gauduchon, P.; Whal, Ph. J. Am. Chem. Soc. 1974, 96,
(20) Ross, J. B. A; Laws, W. R.; Buku, A.; Sutherland, J. C.; Wyssbrod,
H. R. Biochemistry 1986, 25, 607-612.
01-808.
(12) Robbins, R. J.; Fleming, G. R.; Beddard, G. S.; Robinson, G. W.;
(21) Ross, J. B. A.; Laws, W. R.; Sutherland, J. C.; Buku, A.; Panayotis,
P. G.; Schwartz, G.; Wyssbrod, H. R. Photochem. Photobiol. 1986, 44,
365-370.
Thistlethweit, P. J.; Woolf, G. J. J. Am. Chem. Soc. 1980, 102, 6271-
6
5
2
(22) Contino, P. B.; Laws, W. R. J. Fluorescence 1991, 1, 5-13.
(23) Cowgill, R. W. Biochim. Biophys. Acta 1967, 133, 6-18.
(24) Tournon, J. E.; Kuntz, E.; Bayoumi, M. A. Photochem. Photobiol.
1972, 16, 425-433.
(25) Cowgill, R. W. Biochim. Biophys. Acta 1963, 75, 272-273.
(26) Cowgill, R. W. Arch. Biochem. Biophys. 1963, 100, 36-44.
(27) Feitston, J. Phys. Chem. 1964, 68, 391-397.
Am. Chem. Soc. 1983, 105, 3819-3823.
16) Petrich, J. W.; Chang, M. C.; McDonald, B. D.; Fleming, G. R. J.
Am. Chem. Soc. 1983, 105, 3819-3824.
(
(28) Feitston, J. Photochem. Photobiol. 1969, 9, 401-410.
S0002-7863(96)00852-9 CCC: $12.00 © 1996 American Chemical Society

7-Hydroxytetrahydroisoquinoline-3-carboxylic Acid
J. Am. Chem. Soc., Vol. 118, No. 35, 1996 8301
the use of the global and linked-function analysis Laws et al.
were able to show that the rotamer model can explain the
1
,21
complex fluorescence decay of tyrosine analogues,
ty-
rosine,2
0,21
and tryptophan residue in small peptides. The
17
rotamer model showed in the case of tyrosine analogues and
peptides containing tyrosine or tryptophan residue that the
rotamer in which the phenol or indole ring can come into closest
contact with the carbonyl group has the shortest fluorescence
lifetime. The analysis of these results makes it clear that the
R
â
rotamer interconversion around the C -C bond in tyrosine is
slower than the lifetime of the excited state. Unfortunately for
tyrosine components exhibiting a single-exponential decay, Laws
2
et al. were unable to establish whether (i) the slow-exchange
rotamer model is the accurate description, but the three rotamers
have similar unresolved fluorescence lifetimes, or (ii) the
rotamer interconversion is fast enough to average the emission.
The rotamer model has also been used to explain the acrylamide
quenching of tyrosine amide.22 The charge-transfer reaction
between the electrophilic unit in the amino acid backbone and
the excited aromatic phenol subunit leads to a biexponential
fluorescence decay of tyrosine in an acidic aqueous solution. It
is also reasonable that quenching is more efficient with
protonated carboxyl groups since ionization would diminish the
electron-accepting capability of the carbonyl function. This
phenomenon was investigated by Kungl.29 Based on the
dynamics of tyrosine and the peptide Gly-Tyr-Gly in Vacuo,
and in water, calculated with classical molecular dynamics and
with stochastic computer simulations, he concluded that, since
the rotamers frequently interconvert within the fluorescence
lifetime of tyrosine, their contribution to the non-exponential
fluorescence decay should be negligible.
Molecular dynamics simulations of the conformational dy-
Figure 1. Synthetic scheme.
3
0
namics of tryptophan were performed by Gordon et al. They
have obtained different results depending on the model used
for hydrogen representation in the simulation. The rotamer
model for the tryptophan zwitterion can be supported using the
CHARM intramolecular potential on condition that hydrogen
atoms are explicitly included in the model of tryptophan.
are due to ground-state rotamers. Accordinglly, we decided to
synthesize a tyrosine derivatives7-hydroxytetrahydro-isoquino-
line-3-carboxylic acid (Tic(OH))sin which rotations about the
R
â
â
γ
C -C and C -C bonds are restricted by incorporating the
amino group into the six-membered isoquinoline ring. This
constraint limits the number of conformers available in the
molecule greatly affecting the electronic structure of the
chromophore. In this paper we present the fluorescence decay
results for derivatives of the modified tyrosinesTic(OH).
3
0
Gordon et al. have also shown that the predicted relative
populations of tryptophan rotamers are not consistent with the
experimental data. On the other hand, the rotamer model is
not supported by the results obtained by James and Ware for
homo-tryptophan.31 In homo-tryptophan (an analogue of tryp-
tophan with an additional methylene group separating the indole
ring from the R-carbon) the R-â conformers play a less
significant role by virtue of the â-γ and γ-δ conformers.
Homo-tryptophan, with a much larger number of conformations,
exhibits the same qualitative behavior as tryptophan, but with
Experimental Methods
Synthetic Methods. The following derivatives of 7-hydroxytet-
rahydroisoquinoline-3-carboxylic acid (H-Tic(OH)-OH) were ob-
tained: Ac-Tic(OH)-OH [II], Ac-Tic(OH)-NHMe [III], H-Tic(OH)-
NHMe [IV], H-Ala-Tic(OH)-OH [V], Ac-Ala-Tic(OH)-OH [VI],
quenching lifetimes somewhat larger. On the other hand,
H-Tic(OH)-GlyNH [VII]. The methodology of Tic(OH) and prepara-
2
Barkley et al.,1
8,19
by a global analysis of time-resolved
tion of its derivatives are presented in the scheme in Figure 1. Here
we describe only briefly the major steps. 1,2,3,4-Tetrahydro-7-
hydroxyisoquinoline-3-carboxylic acid [H-Tic(OH)-OH (I)] was syn-
thesized by a Pictet-Spengler reaction of 3,5-diiodo-L-tyrosine with
formaldehyde, followed by catalytic hydrogenation (removal of io-
dine).32 3,5-Diiodo-L-tyrosine was purchased from Fluka AG. The
parent compound [I] was N-acylated by di-tert-butyl dicarbonate, N-tert-
butyloxycarbonylalanine N-hydroxysuccinimide ester, or acetic anhy-
dride giving Boc-Tic(OH)-OH, Boc-Ala-Tic-(OH)-OH, or Ac-Tic(OH)-
OH [II], respectively. Boc-Tic(OH)-OH was used as a substrate for
fluorescence data of a rotationally constrained tryptophan
derivative, 3-carboxy-1,2,3,4-tetrahydro-2-carboline (Tcc), re-
vealed biexponential decay for a zwitterion and an anion. The
relative amplitudes match the relative populations of the two
conformers. Consequently, one lifetime was assigned to each
conformer. The shorter lifetime component was associated with
the conformer having the carboxylate closest to the indole ring.
The ethyl ester of Tcc, the derivative with better electron
acceptor properties, has both lifetimes shorter.18,19 This suggests
that the intramolecular electron transfer may be an important
mode of fluorescence quenching.
the synthesis of H-Tic(OH)-NHMe [IV] and H-Tic(OH)-Gly-NH [VII].
For this purpose Boc-Tic(OH)-OH was coupled with methylamine (33%
CH NH in EtOH) in the presence of isobutyl chloroformate or with
3 2
glycine amide using the 1-ethyl-3-(3-(dimethylamino)propyl)carbodim-
2
The question remains unanswered as to whether the two
fluorescence lifetimes of simple tyrosine derivatives in a solution
ide hydrochloride/N-hydroxybenzotriazole [EDC/HOBT] method fol-
(
29) Kungl, A. J. Biophys. Chem. 1992, 45, 41-50.
(31) James, D. J.; Ware, W. R. Chem. Phys. Lett. 1985, 120, 450-454.
(32) Verschueren, K.; Toth, G.; Tourwe, D.; Lebl, M.; Van Binst, G.;
Hruby, V. J. Synthesis 1992, 458-461.
(30) Gordon, H. L.; Jarrell, H. C.; Szabo, A. G.; Somorjai, R. L. J. Phys.
Chem. 1992, 96, 1915-1921.

8302 J. Am. Chem. Soc., Vol. 118, No. 35, 1996
Wiczk et al.
lowed by acidolitic removal of Boc protection. Compounds V [H-Ala-
Tic-(OH)-OH] and VI [Ac-Ala-Tic(OH)-OH] were obtained from Boc-
Ala-Tic-(OH)-OH after deprotection of the tert-butyloxycarbonyl group
and N-acetylation of the alanyl residue. The tert-butyloxycarbonyl
group as a temporary protection was removed in each of the cases
mentioned by the action of a solution of 4 N HCl in dioxane. Ac-
Tic(OH)-NHMe [III] was obtained by the reaction of Ac-Tic(OH)-
OH [II] with methylamine using EDC/HOBT as the coupling reagent.
After purification by means of column chromatography, the chemical
homogeneity of synthesized compounds was checked by 1H-NMR
spectra, analytical reversed HPLC (a linear 60 min gradient from 0 to
8
3 2
0% CH CN in 0.1% TFA in H O at a flow rate of 1 mL/min on a
Kromasil C-8 column, 4.6 × 250 mm), 5 µm and mass spectrometry
(FDMS, FAB MS).
Spectroscopic Measurements. Fluorescence decays were collected
by the time-resolved single photon counting technique on an Edinburgh
Analytical Instrument type CD-900 fluorimeter interfaced with an IBM
PC AT. The excitation source was a flash lamp filled with 0.5 atm of
hydrogen operated at 40 kHz with about 6.5 kV across a 1 mm electrode
gap. The half width of the instrument response was 1.2 ns. The
excitation (270 nm) and emission wavelength (310 nm) were selected
by means of monochromators (about 10 nm band width).
Figure 2. Absorption spectra of Tyr, Ac-Tic(OH)-NHMe, and H-Tic
(OH)-NHMe measured in water at room temperature, at pH 7.0.
Fluorescence decays from sample and the reference (Ludox,
4
observation wavelength 310 nm) were measured to 4 × 10 counts in
the peak. The counting rate did not exceed 2% of the lamp repetition
rate. The decay curves were stored in 1024 channels at 0.054
ns/channel. Fluorescence decay data were fitted by iterative convolution
to the sum of exponents:
structures were true minima. The geometries and energies of the
molecules in the transition states were optimized in two consecutive
steps. First, the SADDLE calculations4 led to the approximation of
the saddle point structure. Next, the gradient minimization by the TS
procedure39 for this structure was carried out. After the transition state
was optimized, the corresponding energy Hessian always had one and
only one negative eigenvalue that indicated a first order transition state.
In order to study the solvent effect, the Conductor-Like Screening
Model (COSMO)42 incorporated into the MOPAC program package
was applied both for minimum structures and saddle point configura-
tions. The dielectric constant of water was taken as 78.4 (at 298 K).
It should be noted that such formalism provides only electrostatic
contribution to the free enthalpy of solvation. One should note that
the same ideal gas expressions for the transitional and rotational terms
in thermal correction for both gaseous and aqueous phase were used.
0,41
I(t) )
R exp(-t/τ )
(1)
∑i
i
i
where τ
pre-exponential factor.
i
is the decay time of the ith component and R
i
is its
The adequacy of the exponential decay fitting was judged by
inspection of the plots of weighted residuals, the statistical parameters
2
ø , and the shape of the autocorrelation function of the weighted
residuals and serial variance ratio (SVR).
The steady-state spectra were obtained on a Perkin-Elmer LS-50B
spectrofluorimeter with a 2.5 nm band width for excitation and emission.
The excitation wavelength was 270 nm. The quantum yields were
measured relative to the value of 0.14 for tyrosine in water at room
4
3
In the past the approximation turned out to be sufficient. The only
difference is in the vibrational term which results from different
frequencies for gaseous and aqueous systems.
3
3
-5
temperature. A sample concentration was about 5 × 10 M in the
All calculations were carried out using a Hewett-Packard 735 Apollo
workstation.
-
4
steady-state measurement and 1 × 10 M in the time-resolved
experiments. All measurements were made in double-deionized water,
pH 7.0, at room temperature. The pH of the solutions was adjusted to
Results and Discussion
7
.0, directly in cuvette, with sodium hydroxide solution (Suprapur,
Merck) using combination pH electrode (ultra-thin, Aldrich) and pHM-
2 apparatus (Radiometer) just before measurements. Then the cuvette
was closed with a cap to avoid diffusion of CO
to the solution. We
Absorbance. The absorption spectra of Tic(OH) and its
derivatives are similar to the spectra of the appropiate derivatives
of tyrosine by means of the shape and position of characteristic
bands (Figure 2). The respective spectra of the derivatives of
Tic(OH) are only somewhat broader and slightly red-shifted.
The maximum of the longwave absorption spectrum of Tyr in
water is around 275 nm, whereas the respective absorption
spectrum of the Tic(OH) derivative with an unblocked amino
group has its maximum around 279 nm. Moreover the
absorption spectra of Tic(OH) derivatives with the blocked
amino group displayed an additional red-shift (1 nm) of the
maximum. The modification of the carboxylic group did not
influence the position of the maximum of the absorption spectra
of derivatives of the cyclic analogue of tyrosine or the shape of
the band. In the case of Tic(OH) and its derivatives less
structured absorption spectra than for tyrosine were observed.
The red-shift observed for the absorption spectra of N-acetylated
5
2
did not use any buffer because of a possibility of an additional
fluorescence quenching related to a hydrogen bond formation between
3
4-36
the phenolic hydroxyl group and a buffer.
Absorption spectra were measured on a Beckman model DU-600
spectrophotometer.
Computational Details. Quantum mechanics calculations were
3
7
carried out at the semiempirical PM3 level employing the MOPAC
3
8
9
3
package.
Unconstrained geometry optimization of the molecules studied was
3
9
performed using the EF optimization procedure. The final norm of
the energy gradient was always less than 0.1 kcal/mol. After each
optimization was completed, energy Hessian was calculated and
checked for positive definiteness, in order to assess whether the
(
(
(
33) Chen, R. F. Anal. Lett. 1964, 1, 35-42.
34) Willis, K. J.; Szabo, A. G. J. Phys. Chem. 1991, 95, 1585-1589.
35) Lee, J. K.; Ross, R. T.; Thampi, S.; Leurgans, S. J. Phys. Chem.
1
992, 96, 9158-9162.
(40) Dewar, M. J. S.; Healy, E. F.; Stewart, J. J. P. J. Chem. Soc. Faraday
Trans. 2 1984, 3, 227-233.
(
36) Pal, H.; Palit, D. K.; Mukherjee, T.; Mittal, J. P. J. Photochem.
Photobiol. A: Chem. 1990, 52, 391-409.
(41) Dewar, M. J. S. Int. J. Quant. Chem. Symp. 1988, 22, 557-565.
(42) Klamt, A.; Sch u¨ u¨ rmann, G. J. Chem. Soc., Perkin Trans. 2 1993,
799-804.
(
37) Steward, J. J. P. J. Comput. Chem. 1989, 10, 209-210.
(38) Steward, J. J. P. MOPAC 93, Copyright Fujitsu: Tokyo, Japan,
1
993.
39) Baker, J. J. Comput. Chem. 1986, 7, 385-393.
(43) Wong, M. W.; Wiberg, K. B.; Frisch, M. J. J. Chem. Phys. 1991,
95, 8991-8998.
(

7-Hydroxytetrahydroisoquinoline-3-carboxylic Acid
J. Am. Chem. Soc., Vol. 118, No. 35, 1996 8303
Table 1. Fluorescence Quantum Yields (φ) and Lifetimes (τ) of
7
-Hydroxytetrahydroisoquinoline-3-carboxylic Acid
(H-Tic(OH)-OH) Derivatives in Water (pH 7.0 at Room
Temperature)
no.a
compd
φ
τ [ns]
ø2
b
R
I
II
III
IV
V
VI
VI
H-Tic(OH)-OH
Ac-Tic(OH)-OH
0.096
0.095
0.091
0.077
0.023
0.060
0.065
0.14
2.92
2.88
2.82
2.23
0.82
1.87
2.10
3.42
1.04
1.03
0.99
1.12
1.10
1.11
1.10
0.99
Ac-Tic(OH)-NHMe
H-Tic(OH)-NHMe
H-Ala-Tic(OH)-OH
Ac-Ala-Tic(OH)-OH
H-Tic(OH)-Gly-NH
H-Tyr-OH
2
a
Compound number in the reaction scheme. b Goodness of fit.
quantum yield for Tic(OH)-Gly-NH2, φ ) 0.065), but the effect
of this transformation is not as strong as observed for Tic(OH)-
NHMe. This change in the quenching efficiency is distinctly
an effect of the distance between the chromophores (the phenol
and the amide)sthe separation of the chromophores by the one
additional -CH2- group (as in Tic(OH)-Gly-NH2) decreased
the quenching efficiency. A similar dependence of the fluo-
rescence quenching from the distance between chromophores
was observed by Cowgill for derivatives of tyrosine (Tyr-
Gly and Tyr-Gly-Gly).
A significant influence on the fluorescence quenching was
observed for the N-acylation of the cyclic analogue of tyrosine
Figure 3. Emission and fluorescence excitation spectra of Tyr, Ac-
Tic(OH)-NHMe, and H-Tic (OH)-NHMe measured in water at room
temperature, at pH 7.0.
44
derivatives of Tic(OH) in comparison with that of the non-
acetylated derivatives indicates a certain interaction of the acetyl
group with the phenolic chromophore (hyperresonance), despite
the separation of these two moieties by one methylene group
(Tic(OH)) by alanine and acetylalanine . The quantum yield
for Ac-Ala-Tic(OH)-OH, the compound containing two amide
groups, φ ) 0.060, is comparable with the data for Tic(OH)-
Gly-NH2 (φ ) 0.065)sthe compound containing also two amide
moieties. The quenching efficiency of the fluorescence of Ala-
Tic(OH)-OH is about 75% of that observed for the N-acylated
derivative (Ac-Ala-Tic(OH)-OH). The influence of the proto-
nation of the amino group on the quenching efficiency of the
fluorescence of the phenolic chromophore by carbonyl group
can be analyzed by comparing the fluorescence data (quantum
yield) for Tic(OH)-NHMe (φ ) 0.077) and Ac-Tic(OH)-NHMe
(-CH2-). The shift of the maximum of the absorption spectra
of the cyclic analogue of tyrosine is probably due to a change
in the conformation of a semirigid six-membered ring which is
the effect of a different characteristic of isoquinoline nitrogen
(isoquinoline contains a secondary amino group and its acety-
lation results in a more rigid flat conformation of the ring).
Similar changes and relations were observed for fluorescence
excitation spectra of Tic(OH) derivatives (Figure 3).
Steady-State Fluorescence. The steady-state emission spec-
tra of the zwitterion of Tic(OH) and its derivatives are broad
and featureless (Figure 3). The emission maximum occurs at
(
φ ) 0.091). These parameters explicitly indicate the important
role of the amino group in the neighborhood of the quencher
carbonyl) in the quenching process of the fluorescence of the
(
aromatic amino acids. The same direction in the effect of the
acetylation of the amino group was observed for tyrosine amide,
e.g. the longer fluorescence lifetime for Ac-Tyr-NH2 than for
3
07 nm for Tic(OH) and all other studied derivatives. All
emission spectra of Tic(OH) derivatives are red-shifted as
compared with the tyrosine spectra (λmax ) 303 nm) and
possesses a greater half-width.
2
Tyr-NH2 (studies performed by Laws et al. and by Gauduchon
4
and Wahl ).
Contrary to the increase of the quantum yield observed for
The quenching efficiency of the ammonium group in tryp-
tophan and tyrosine has been discussed in terms of the
electrostatic effect on the quenching efficiency of the carbonyl
group and not attributed directly to a straight quenching
the rigid analogue of Trp (3-carboxy-1,2,3,4-tetrahydro-2-
carboline),1
8,19
the fluorescence quantum yield of the Tic(OH)
zwitterion in water at pH 7.0 is slightly lower (φ ) 0.096) than
that of tyrosine in water (φ ) 0.14). The fluorescence quantum
yields of Tic(OH) derivatives are in the range of 0.095 for Ac-
Tic(OH)-OH to 0.023 for the most quenched derivative, H-Ala-
Tic(OH)-OH (Table 1).
3
,5,23,44
process.
The electrostatic field produced by the am-
monium group in the derivatives of tyrosine with protected
carboxylate can change the polarization of the surrounding water
molecules, which will lead to higher polarization of the carbonyl
group and increasing electron-acceptor property of the amide
group or can strongly increase hydration of the amide bond,
facilitating electron transfer. Low quantum yield of the
fluorescence of tyrosine with ionized hydroxyl and its appear-
Analyzing the data collected in Table 1, it is clear that
N-acetylation did not change the fluorescence of parent com-
pounds significantly. The same phenomenon was observed for
tyrosine and acetyltyrosine. The decay times (τ) obtained by
Laws et al.2 for Ac-Tyr and Tyr were 3.6 and 3.76 ns,
4
5
ance at a longer wavelength (λmax ) 345 nm) can cause an
apparent quenching effect of tyrosine fluorescence. The data
of the fluorescence quenching obtained for Tic(OH)-NHMe and
Ala-Tic(OH)-NHMe and Tic(OH) derivatives with the acety-
4
respectively, whereas Gauduchon and Wahl obtained shorter
ones valuess3.2 ns for Ac-Tyr and 3.38 ns for Tyr. The
fluorescence quenching increases a little when N-acetylation is
combined with modification of the carboxylate (amidation). On
the other hand, introduction of the carboxyamide to the
compound without N-protection gave a high, about 20%,
increase of the quenching. The amidation of the peptide Tic-
(44) Cogwill, R. W In Biochemical Fluorescence: Concepts; Chen, R.
F., Edelhoch, Eds., Marcel Dekker, Inc.: New York and Basel, 1976,
Chapter 9.
(45) Rayner, D. M.; Krajcarski, D. T.; Szabo, A. G. Can. J. Chem. 1978,
(OH)-Gly-OH also improved the fluorescence quenching (the
56, 1238-1245.

8304 J. Am. Chem. Soc., Vol. 118, No. 35, 1996
Wiczk et al.
Figure 4. Measured (doted line) and fitted decay curve (solid line) vs
time (left set of pointsslamp profile) for H-Tic(OH)-NHMe. The
weighted residuals are plotted at the lower part of the figure.
Figure 5. Measured (doted line) and fitted decay curve (solid line) vs
time (left set of pointsslamp profile) for Ac-Tic(OH)-NHMe. The
weighted residuals are plotted at the lower part of the figure.
lated amino group proved the strong influence of the protonated
amino group on the quenching properties of the amide group.
Fluorescence Decays. Fluorescence decays of all derivatives
in Tables 2 and 3. Although only two minimum energy
structures could be found, Figures 6 and 7 illustrate the two
conformers viewed down the plane of the phenol ring. For
compounds with a free amino group, these structures correspond
to the well-characterized half-chair forms of cyclohexane.
4
of Tic(OH) were measured up to 4 × 10 counts per peak to
obtain the highest possible accuracy of the parameters. All
measured decays are monoexponential (Table 1, Figures 4 and
2
5
) with the statistical parameter ø R in the range 0.99-1.12 and
did not improve even when biexponential decay fitting was
applied. The values of weighted residuals and an autocorrelation
function do not display any systematic deviations and they are
around zero. There is a direct correlation between the fluores-
cence decay times and the quantum yields for all compounds
studied (Table 1)sthe derivatives with shorter decay times
possess lower quantum yields. Fluorescence rate constants (kf)
calculated using the equation kf ) φ/τ are in the range of from
1
8,19
Similar structures were found by Barkley et al.
for the
rotationally constrained tryptophan derivatives. For the com-
pounds with blocked amino groups the lowest energy conforma-
tion differs from the characteristic half-chair form because of
the planarity of the amide bond formed after acetylation (see
Figure 7).
The T type conformation, named according to the paper by
1
8
Barkley et al., always has a lower gaseous-phase energy than
the T′ type, with the exception of the cation for which the T′
type is the lowest energy conformation. The heat of formation
of both conformers and the transition state for all studied
derivatives of Tic(OH) are also presented in Tables 2 and 3.
These data were used to calculate the population of the particular
conformers based on the stationary-state theory using the
following equations:
7
-1
7 -1
3
.3 × 10 s for Ac-Tic(OH)-OH to 2.8 × 10 s for H-Ala-
Tic(OH)-OH, and these values are close to the rate constant
7
-1 2,32
calculated for tyrosine (3.8 × 10 s ).
The incorporation
of the substituents of the amino group and carboxylate does
not change the fluorescence rate constants in the cyclic tyrosine
derivatives significantly and the lower values of the quantum
yields and the shorter decay times are obviously the result of
the fluorescence quenching caused by the substituents.
Theoretical Calculations. Theoretical calculations of the
lowest energy conformers of Tic(OH) and derivatives of Tic-
∆
G ) ∆H - T∆S
(2)
(3)
(OH) with the modified amino group and/or carboxylate were
∆G ) -RT ln K
performed using the PM3 method (the derivatives containing
the peptide bond were not subjects of these calculations). The
calculations were carried out in (i) a gaseous phase and (ii)
including water as a solvent. The calculated thermodynamic
quantities of the different conformations of Tic(OH) derivatives
and the transition states between the conformations are given
and the assumption that the sum of all conformers is normalized
to 1 (pT + pT′ ) 1).
The differences in the heat formation for both conformers of
all studied Tic(OH) derivatives possessing a charge are small
(about 1 kcal/mol) except for the disubstituted derivative (Ac-

7-Hydroxytetrahydroisoquinoline-3-carboxylic Acid
J. Am. Chem. Soc., Vol. 118, No. 35, 1996 8305
Table 2. Thermodynamic Parameters and Equilibrium Constants between T and T′ conformers for Tic(OH) Derivatives in the Gas Phase
thermodynamic parameter
conformer
+H
2
TicNHMe
AcTicNHMe
AcTicCOO-
+H TicCOO-
2
a
HOF (kcal/mol)
T
T′
TS
T
T′
TS
T
-90.377
-88.094
96.276
166.098
166.028
118.202
0.020
-105.828
-105.459
-99.293
136.115
133.509
135.632
0.876
-186.255
-185.325
-180.304
139.959
140.104
123.433
0.184
-82.097
-77.783
-71.348
107.018
109.402
107.991
0.998
c
Sb (cal/(mol‚K))
population
T′
0.980
0.021
0.124
6.91
0.816
0.225
0.002
435.30
K
T
/
T′
a
HOF ) heat of formation. b S ) entropy. Transition state.
c
Table 3. Thermodynamic Parameters and Equilibrium Constants between T and T′ Conformers for Tic(OH) Derivatives with Solvation
thermodynamic parameter
conformer
+H
2
TicNHMe
AcTicNHMe
AcTicCOO-
+H TicCOO-
2
a
HOF (kcal/mol)
T
T′
TS
T
T′
TS
T
20.468
20.016
23.469
100.514
102.895
102.010
0.124
-130.750
-127.221
-125.282
109.032
111.397
114.748
0.9915
-282.896
-281.083
-278.013
100.274
104.844
104.913
0.681
-136.609
-137.320
-133.109
95.254
98.964
97.84
c
Sb (cal/(mol‚K))
population
0.0446
T′
0.876
0.141
0.0085
116.96
0.319
2.13
0.9554
0.0467
K
T
/
T′
a
HOF ) heat of formation. b S ) entropy. c Transition state.
Figure 6. T and T′ conformers of the protonated form of Tic(OH)-NHMe.
Tic(OH)-NHMe) for which the difference is higher: 4.3 kcal/
mol. We also found small differences between entropies of the
individual conformerssabout 0.5 kcal/mol. Again, the higher
difference was found in the case of Ac-Tic(OH)-NHMesabout
than 2%)sit is the T′ type for Tic(OH)-NHMe and the T type
for Tic(OH) (zwitterion). On the other hand, the derivatives
with the acetylated amino group exist predominantly in the T
type conformation (more than 80% of the whole population of
conformers).
2
kcal/mol. The value of the energy barrier of the conformers’
conversion, calculated as the difference between the formation
heat of the saddle point and the energy of the lower energy
A different situation occurs when the solvation is included
in our calculations. The lowest energy conformers for the
derivatives with the free amino group are the T′ type regardless
of the status of the carboxylate (free or amidated). On the
contrary, the lowest energy conformers found for the acetylated
derivatives exist in the T type conformation, but similar to the
derivatives with the free amino group the state of the carboxylic
group does not influence the ring conformation. The enthalpies
of the formation calculated for individual conformers (solvation
included) differ significantly less than the ones obtained in the
gaseous phase, especially for the compounds possessing charges.
The compounds with the free amino group (positively charged)
have a difference in the enthalpies of around 0.7 kcal/mol,
conformer, is about 6 kcal/mol which is comparable with the
data obtained by Barkley et al.1
8,19,46
for the rotationally
constrained tryptophan derivatives and that found for the
inversion of 1,4-disubstituted cyclohexanes.47 The parent
compound, Tic(OH), was exceptional, in terms of the energy
barrier of the conformers’ interconversion. The energy barrier
was about 11 kcal/mol. The equilibrium constants and popula-
tions of the particular conformers collected in Table 2 (calculated
on the basis of the obtained thermodynamical parameters)
revealed that the derivatives with the free amino group exist as
practically one conformer (contribution of the second is less
(
46) Yu, H-T.; Vela, M. A.; Fronczek, F. R.; McLaughlin, M. L.; Barkley,
(47) Jesen, F. R.; Bushweller, C. H. J. Amer. Chem. Soc. 1969, 91, 5774-
5782.
M. D. J. Amer. Chem. Soc. 1995, 117, 348-357.

8306 J. Am. Chem. Soc., Vol. 118, No. 35, 1996
Wiczk et al.
Figure 7. T and T′ conformers of Ac-Tic(OH)-NHMe.
Table 4. Ring-Inversion Rate Constants of the T / T′
whereas for the compounds with free carboxylate (negatively
charged) the difference in the entalpies is slightly higher (1.2
kcal/mol). A more significant difference in the enthalpies of
conformer formation than in the gaseous phase (almost 1 order
of magnitude) was observed for Ac-Tic(OH)-NHMe (∆HOFpg
interconversion in Water Calculated Based on the Data of Tables 2
and 3
rate constant
(
lifetime)
+H2TicNHMe AcTicNHMe AcTicCOO- +H2TicCOO-
9
1
(
0- kTfT′ [s-1
]
53.9
(0.19)
11.5
4.44
(2.25)
506
(0.02)
6.80
(1.47)
14.5
25.0
(0.40)
1.16
)
0.37 kcal/mol, ∆HOFsol ) 3.2 kcal/mol). A more significant
1
0
10 τ [s])
difference was also found for the entropy of both conformers
when the solvation was taken into account during the calcula-
tions of all Tic(OH) derivatives (2.5-4.5 kcal/(mol K)). On
the other hand, the significant decrease in the enthalpy of
activation of the ring inversion was discovered for the deriva-
-9
kT′fT [s-1]
10
10
(
10 τ [s])
(0.87)
(0.69)
(8.57)
Tic(OH) is almost 2 times bigger than that found by Yu et al.46
for 3-carboxy-1,2,3,4-tetrahydro-2-carboline, whereas the rate
constants for all other studied Tic(OH) derivatives are even
biggersmore than 1 order of magnitude. According to Yu et
#
tives with the free amino group (∆H ∼3 kcal/mol), while for
the compounds with the acetylated amino group the decrease
#
was smaller (∆H ∼5.5 kcal/mol).
46
al. the energy barrier of the ring inversion in the excited state
The decrease of the enthalpy of activation of the ring
inversion and a higher difference in entropy of the particular
conformers and the transition state obtained in the experiments,
taking into account the solvation effect, do not permit neglect
of the entropy component in the calculations of the equilibrium
constant and rate of ring inversion possible in the gaseous phase
does not change very muchsit increases by 0.4-0.9 kcal/mol.
Because of the dynamic equilibrium in the excited state, the
ring inversion can occur several times during the fluorescence
lifetime (around 3 ns) of the excited state of the Tic(OH)
derivatives. For these reasons, the measured fluorescence decay
is the averaged decays of both conformers.
(especially at room or higher temperature). The free energy of
activation of the ring inversion, calculated with the solvation
Conclusions
effect included, for all studied derivatives was in the range 3.2-
In the rotamer model, three chemically distinct environments
3
.8 kcal/mol except for the parent compound, Tic(OH), for
R
â
exist for the phenol ring about the C -C bond and each rotamer
has its own associated decay constant, and the relative weighing
of the amplitudes is set by the ground-state rotamer populations.
The rotamers having different lifetimes have also been suggested
as the cause of multiexponential decay of tryptophan,
rotationally constrained tryptophan
protein crystals. The described data of the fluorescence decay
measurements for Tic(OH) derivatives have confirmed only
partially the rotamer model of quenching. Theoretical calcula-
tions showed that Tic(OH) can adopt two half-chair conforma-
tions in solution. The lowest energy conformation for Tic(OH)
derivatives with the free amino group is the T′ type, whereas
for the derivatives with the acetylated amino group it is the T
type. Including hydration (within the framework of the COSMO
approach) this results in lowering the barrier to ring inversion
to about 3.5 kcal/mol. The contribution of the second conformer
which the energy is higher (4.5 kcal/mol). These values are
generally lower than the energy calculated in the gaseous phase.
The rate constants of the conversion through the energy barrier
during the ring inversion are calculated on the basis of the
transition-state theory using the equation
1
1-16
,
18,19,46
and tryptophan in
48
#
kT
h
∆G
kTfT ) κ
exp -
(4)
( )
(
)
RT
where k is Boltzmann constant, T is the absolute temperature,
#
h is Planck’s constant, and G are the free energies of activation,
these being collected in Table 4. It is difficult to assign a value
to the transmission coefficient, κ, which incorporates all
correction factors and uncertainties. We chose, as Barkley et
1
9
al. did, κ ) 0.4, and for this value the calculated average
lifetimes of T and T′ conformers are tens of picoseconds. The
T type conformer of Ac-Tic(OH)-NHMe and the T′ type of Tic-
(the T type for the derivatives with the free amino group and
(OH) have the longest lifetimes (200 and 860 ps, respectively).
(48) Dahms, T. E. S.; Wilis, K. J.; Szabo, A. G. J. Amer. Chem. Soc.
The rate constant of the ring conversion obtained for the parent
1995, 117, 2321-2326.

7-Hydroxytetrahydroisoquinoline-3-carboxylic Acid
J. Am. Chem. Soc., Vol. 118, No. 35, 1996 8307
T′ for the acetylated group) in the whole population of rotamers
is lower than 20%, except Ac-Tic(OH)-NHMe for which the
contribution of the lower populated rotamer is less than 1%.
The obtained rate constant of the ring conversion for the parent
Tic(OH) is almost two times bigger than that found by Barkley
influence of the NH3+ group on the quenching of tyrosine
fluorescence. This group is not directly involved in the
quenching process, which is consistent with previous sugges-
3
,44
tions,
but it influences the quenching ability of the amide
group through its close space location, Via enhancement of
hydration of the carbonyl group, and has strict spatial require-
ments. On the basis of the above-mentioned statement about
the role of the amino group in the quenching, one can easily
explain the more efficient quenching of the fluorescence of the
amides: Tic(OH)-NHMe and Ala-Tic(OH)-NHMe, compounds
containing the quencher and the free amino group (see Table
1). However, on the basis of these data no satisfactory
explanation for the mechanism of the interaction between the
1
9,46
et al.
for 3-carboxy-1,2,3,4-tetrahydro-2-carboline and by
Yu et al.46 for 1,2,3,4-tetrahydro-2-carboline, whereas the rate
constants for all other studied Tic(OH) derivatives are even
greatersmore than one order of magnitude. So, the average
fluorescence lifetimes of all conformers in the solution are in
the range of hundred of picoseconds. The apparent monoex-
ponential decays imply that either the conformers interconvert
in the excited state rapidly, compared to the fluorescence time
scale (about 3 ns), or the lifetimes of the two conformers are
similar.
+
NH3 group and the amide bond could be found. The effect of
+
the charge field of the NH3 group and related to this better
hydration of the amide bond suggested by Cowgill44 seems to
be a reasonably good explanation for the mechanism of the
The results of the fluorescence decays obtained for Tic(OH)
derivatives provide new suggestions regarding the quenching
mechanism of the fluorescence of tyrosine analogues. The
fluorescence quenching by N-terminal or C-terminal protecting
groups depends on the distance between the chromophore
+
NH3 group action. A full explanation of the fluorescence decay
mechanism for tyrosine analogues requires further study,
especially studies of model compounds with different confor-
mational freedom of the chromophore and protecting groups at
N- and C-termini.
(phenolic ring) and a quenching moiety (acetyl, N-methylamide)
and their direct surroundings. A thorough analysis of the
dependence of the effectiveness of the quenching groups from
a distance between the chromophore and the quencher led us
to the conclusion that the previously suggested charge-transfer
mechanism3 is highly probable (the weak quenching of
tyrosine fluorescence by the acetyl in Ac-Tyr and the strong
Acknowledgment. This work was supported by the Polish
State Committee for Scientific Research (KBN) under grant PB
0
582/P3/03/04.
,24
JA960852S
4
9
quenching by the acetyl in the case of Ac-âTyr(Me) ). The
(49) Wiczk, W.; Łankiewicz, L.; Czaplewski, C.; Ołdziej, S.; Stachowiak,
results obtained in our investigations partially explained the
K.; Michniewicz, A.; Liwo, A. J. Photochem. Photobiol. A: Chem. In press.