Development of new double-stranded phenylalanyl chelators using η-χ diagrams and binding constants for chelators and lanthanide ions

Article abstract of DOI:10.1248/cpb.58.620

We examined the interaction between several ligands and alkaline (M ⁺) or lanthanide (Ln³⁺) ions using η-χ diagrams. η and χ represent absolute hardness and absolute electronegativity, respectively. Based on the diagrams, new double-stranded amino acid chelators (1) conjugated to Cat (catechol) were synthesized. They had two cavities coordinating hard metal ions, such as Li⁺ and K⁺, and soft metal ions, such as Lu³⁺ and Eu³⁺. 1 formed a solvated molecular complex, 1-Ln³⁺, with a molar ratio of 1:1 at a binding constant (K_b) of 10⁴-10⁶ with Y³⁺, La³⁺, Eu³⁺, and Lu³⁺, respectively. With a chemical hard ion, Li⁺, we found that the K_b of the 1-Eu³⁺ complex increased. The optimized geometries of 1-Lu ³⁺ and 1-La³⁺ complexes were obtained from B3LYP correlation functions using Stuttgart-Dresden-Bonn (SDD) Effective Core Potential (ECP) basis sets for Lu³⁺ and La³⁺ ions. The double-stranded chelator may prove useful for studying the selective separation of lanthanide ions, and the development of functional peptides.

Full text of DOI:10.1248/cpb.58.620

620
Regular Article
Chem. Pharm. Bull. 58(5) 620—627 (2010)
Vol. 58, No. 5
h c
Development of New Double-Stranded Phenylalanyl Chelators Using –
Diagrams and Binding Constants for Chelators and Lanthanide Ions
Shigeki KOBAYASHI,* Makoto WATANABE, and Toshiyuki CHIKUMA
Department of Analytical Chemistry of Medicines, Showa Pharmaceutical University; 3–3165 Higashi-tamagawagakuen,
Machida, Tokyo 194–8543, Japan.
Received October 2, 2009; accepted February 6, 2010; published online February 15, 2010
We examined the interaction between several ligands and alkaline (Mꢀ) or lanthanide (Ln3ꢀ) ions using h–c
diagrams. h and c represent absolute hardness and absolute electronegativity, respectively. Based on the dia-
grams, new double-stranded amino acid chelators (1) conjugated to Cat (catechol) were synthesized. They had
two cavities coordinating hard metal ions, such as Liꢀ and Kꢀ, and soft metal ions, such as Lu3ꢀ and Eu3ꢀ. 1
formed a solvated molecular complex, 1–Ln3ꢀ, with a molar ratio of 1 : 1 at a binding constant (Kb) of 104—106
with Y3ꢀ, La3ꢀ, Eu3ꢀ, and Lu3ꢀ, respectively. With a chemical hard ion, Liꢀ, we found that the Kb of the 1–Eu3ꢀ
complex increased. The optimized geometries of 1–Lu3ꢀ and 1–La3ꢀ complexes were obtained from B3LYP cor-
relation functions using Stuttgart-Dresden-Bonn (SDD) Effective Core Potential (ECP) basis sets for Lu3ꢀ and
La3ꢀ ions. The double-stranded chelator may prove useful for studying the selective separation of lanthanide
ions, and the development of functional peptides.
Key words chemical hardness; double-stranded amino acid; chelator; lanthanide ion; binding constant
Lanthanides (Ln) as heavy rare earth elements and Y are ated with the theory of chemical reactivity in molecules, are
attracting the interest of chemists since complexes prepar- being applied to the design of chelators and chemicals.16,17)
ed by the reaction of lanthanide ions (Ln3ꢀ) with chelators is the total electron number and E is total electron energy.
N
have a variety of applications, including probes for fluorome-
To develop chelators, we have focused on a convenient
try,1—4) catalysts,5) sensitizing chromophores,6—8) and nano- model using quantum chemical h–c diagrams for chelators
materials.9,10) Recently, lanthanide ions with unique photo- and lanthanide ions based on the chemical hardness concept,
physical and coordination properties have been used to de- h and c, as described above. The h–c diagram shows that
velop lanthanide-antenna sensors as luminescent sensors for phenol and Cat (catechol) are appropriate for the reaction of
cellular imaging,8) and protein and DNA assays. As chemical a ligand with Lnnꢀ ions since Lnnꢀ is chemically harder than
and physicochemical properties, lanthanides are chemically phenol and Cat.
similar to metal ions because of the shielding of progres-
Here, we report a new chelator (1) that conjugated Cat
sively filled inner 4f orbitals by the filled outer 5s and 5p to bis(tert-butoxycarbonyl (Boc)-L-Phe)-N,N-ethylene-
orbitals and the lanthanide contraction effect. The study of dioxy-bis(ethylamine), which should support a b-sheet-like
the stability of lanthanide-ligand complexes is interesting conformation.18,19) The chelator could expect flexible confor-
because the lanthanide contraction effect appears to vary di- mation due to a backbone conjugated with two amino acid
rectly with the ionic radius.
residues to the N-terminus of a spacer, –O–CH2–CH2–O–,
The development of chelator–Ln3ꢀ complexes has been of and its binding site is Cat, not an amido group. Chelator 1
interest to investigators of the regioselective lanthanide- provides a coordinate bond complex with Ln3ꢀ of the largest
induced hydrolysis of phosphate esters, including DNA and ionic radius in 10 mM N-(2-hydroxyethyl)piperazine-Nꢃ-(2-
RNA.11) Such compounds are potentially applicable to phar- ethanesulfonic acid) (HEPES) buffer in 70% MeCN (pH
macological fields12) as chemotherapeutic agents, such as the 4.8). Indeed, the binding constants of 1–Ln3ꢀ complexes are
platinum anticancer drug cisplatin. Generally, chelator–Lnnꢀ about 104—106 Mꢂ1. Therefore, we show that compound 1
complexes have low solubility. The design of chelators, analogs with ligand Cat are excellent chelators for Ln3ꢀ.
which have both hydrophobic parts (or regions) and hy- Complexes of chelator 1 with Ln3ꢀ may be applied in the
drophilic parts in the same molecule, is required to increase pharmacological field as chemotherapeutic agents and lumi-
the solubility of Lnnꢀ ions. We envisage the development nescent sensors,8) which are a subject of many recent studies.
of highly soluble chelator–Lnnꢀ complexes which express
In addition, the quantum chemical h–c diagram is a useful
biological effects through coordination with the binding sites tool for the design of chelators, and we will deal with the
of drugs, proteins, and DNA. It is necessary to understand synthesis and characterization of the complexes of the chela-
which ligands coordinate easily with Ln3ꢀ; however, there is tor with Ln3ꢀ. To rationalize the structure of chelator–Lnnꢀ
no convenient chemical method of examining the interaction complexes, we also discussed the structural properties of
between chelators and metal ions.
complexes of the chelator with La3ꢀ and Lu3ꢀ using the
In the past several years, the development of theoretical B3LYP level.
models of metals and molecules based on electrophilicity
(w) and chemical hardness (h) indexes for the characteriza- Results and Discussion
tion of chemical reactions, biological activity, and atoms has
Design of an Active Site: Quantum Chemical h–c Dia-
progressed steadily.13) The concepts of chemical hardness, grams for Lanthanide Ions The relation between the ion-
absolute hardness [hꢁ1/2(∂2E/∂N2)] and absolute elec- ization potential (Ip) and electron affinity (Ea) of metal ions
tronegativity [cꢁꢂ(∂E/∂N)],14,15) which have been associ- (Mnꢀ) is shown in Eq. 1. The values of absolute hardness (h)
© 2010 Pharmaceutical Society of Japan
∗ To whom correspondence should be addressed. e-mail: kobayasi@ac.shoyaku.ac.jp

May 2010
621
and absolute electronegativity (c) were calculated using data pling and orbital moments of other Ln3ꢀ ions with 4f elec-
published previously.20)
trons are important to obtain estimates of magnetic moments
Coordinate r(c, h) as the electronic structure of Ln3ꢀ was and magnetic ordering temperatures.22) Therefore, heavy rare
calculated from Eqs. 1, 2, and 3.
earth elements and related elements, Sc3ꢀ, Y3ꢀ, La3ꢀ, and
Gd3ꢀ, which can be neglected in terms of the effect of rela-
tivity theory, are determined using the r(c, h) diagram. For
example, the h of Ce3ꢀ is 8.28 and that of Gd3ꢀ is 11.71.
Ce3ꢀ is chemically softer and Gd3ꢀ is harder than other Ln3ꢀ
ions with 4f electrons. The h of lanthanum ions, La3ꢀ, is
14.875 and La3ꢀ lack three electrons, (5d)1(6s)2, since the
Ln3ꢀ: Ln3ꢀ(Ea) → Ln4ꢀ(Ip) ꢀ eꢂ
(IpꢀEa)
(1)
(2)
ꢂ μ ꢁ
ꢁ χ
2
(IpꢂEa)
ηꢁ
(3)
2
Where m is chemical potential. Here, we found that the electron configuration of a neutral La is (4d)10(4f )0(5s)2(5p)6.
r(c, h) of lanthanide ions, Ln3ꢀ, having 4f electrons, has no Eu3ꢀ has the electron configuration (4d)10(4f )6(5s)2(5p)6.
linear relationship between absolute hardness and absolute
To achieve the design and synthesis of functional chela-
electronegativity. In the diagram of r(c, h), Sc3ꢀ (4fꢁ0), tors, which selectively bind with chemically soft Eu3ꢀ and
Y3ꢀ (4fꢁ0), La3ꢀ (4fꢁ0), and Gd3ꢀ (4fꢁ7) have a two- Eu3ꢀ ions, we discuss a method for the development of
order correlation between h and c, and we found that the chelators using r(c, h) diagrams of Ln3ꢀ ions and ligands
correlation curve was fitted with hꢁꢂ0.0308c2ꢀ3.2518cꢂ (L) (2—10). As a soft-based chelator is favorable for binding
61.171 (correlation coefficient (r): rꢁ0.9994); however, Ce3ꢀ soft-acid metal Mnꢀ ions according to the hardness concept,
(4fꢁ1)–Eu3ꢀ (4fꢁ6) and Dy3ꢀ (4fꢁ9)–Lu3ꢀ (4fꢁ14) our target is a chemically soft ligand. In addition, our aim is
groups did not show a linear relationship between h and c to design a L–Ln3ꢀ complex in which the binding force be-
(Fig. 1). The graph has no two-order or three-order correla- tween the ligand and Mnꢀ ion is weak; that is, ligand ex-
tions; however, the behavior of Gd3+ is similar to that of change is active. The interaction of Ln3ꢀ ions with the ligand
Sc3ꢀ, Y3ꢀ, and La3ꢀ although it has 4f electrons. This may be as a donor is a result of an overlap of orbitals or of electro-
explained by the relativity theory for heavy rare earth ele- static binding of two Ln3ꢀ ions and L (Eq. 4).
ments.21)
ꢀ Ln3ꢀ
L–Ln3ꢀ
(4)
→
L
(ligand)
←
Hughes et al. reported that although the effects of spin or-
bital coupling of Gd3ꢀ can be neglected, the spin orbital cou-
We show a r(c, h) diagram plotting the h and c of several
ligands calculated using the density functional theory B3LYP
with a 6-31G(d) basis set (Fig. 1). When Ln3ꢀ forms a com-
plex with L, electrons flow from L to Ln3ꢀ. The quantitative
charge transfer (DQ) and stabilization energy (DE) were cal-
culated using Eqs. 5 and 6.14)
(L–Ln3ꢀ
2(L–Ln3ꢀ
)
ΔQ ꢁ
(5)
(6)
)
(L–Ln3ꢀ
4(L–Ln3ꢀ
)
2
ΔE ꢁꢂ
)
According to Eqs. 5 and 6, the softer the target chelator,
the less favorably it binds with Ln3ꢀ ions. In the diagram in
Fig. 1a, the Cat (7) is a softer chelator than 18-crown-6-ether
(2), en (3), ethylenediaminetetraacetic acid (EDTA) (4), 12-
ane-4 (5), and calixarene (8), and the c of Cat is smaller than
that of 3, 4, or 5. In making up the electrostatic Cat–Eu3ꢀ
complex of Cat with Eu3ꢀ, the stabilization energy DE is
ꢂ20.48 (eV), as listed in Table 1. The value is larger than
that of 2—6 and 8—10. These values are expressed as the
binding force of the chelators with La3ꢀ and Eu3ꢀ ions.
Moreover, the DQ of the 7–Eu3ꢀ complex is ꢂ1.317. This in-
dicates that electrons easily withdraw Eu3ꢀ from 7. Calcu-
lated DQ and DE values indicate that the stabilization of
ligand–Eu3ꢀ complexes decreases in the following order:
7–Eu3ꢀꢄ8–Eu3ꢀꢄ10–Eu3ꢀꢄ9–Eu3ꢀꢄ6–Eu3ꢀꢄ3–Eu3ꢀꢄ4–
Eu3ꢀꢄ5–Eu3ꢀ and ꢄ2–Eu3ꢀ (Table 1). This indicates that
ligand 6 would form a stable solvated complex with Ln3ꢀ
ions in solution; therefore, we designed and synthesized a
novel chelator (1), in which a chemically soft ligand 7 was
conjugated to two terminals in the molecule. Although the
spacer –HN–CH2CH2–O–CH2CH2–O–CH2CH2–NH– of 1 is
as chemically hard as ligand 2 or 5, 1 joined to 7 is a chemi-
Fig. 1. Plot of r(c, h) as a Coordinate Diagram of Electronic Structure for
Alkaline (Mꢀ) and Lanthanide Ions (Ln3ꢀ) (a), and Structures of Several
Chelators (2—10) (b)
(a) Data of Ip and Ea for alkaline and lanthanide ions were from ref. 20. (b) Upper
square figure is enlarged the small square figure at the bottom left in Fig. 1a. Symbol
AA is amino acid residues.

622
Vol. 58, No. 5
Table 1. Calculated Hardness (h), Electronegativity (c), Quantitative of Charge Transfer (DQ) and Stabilization Energy (DE) for Chelators and Chela-
tor–Ln3ꢀ Complexes
DQ
La3ꢀ b)
complex
(eV)
DE
La3ꢀ b)
complex
(eV)
DQ
Eu3ꢀ b)
complex
(eV)
DE
Eu3ꢀ b)
complex
(eV)
a)
a)
ehomo
elumo
ha)
(eV)
ca)
(eV)
Chelator
(eV)
(eV)
2
3
4
5
6
7
8
9
10
ꢂ6.973
ꢂ6.152
ꢂ6.609
ꢂ5.602
ꢂ6.357
ꢂ5.625
ꢂ5.848
ꢂ6.066
ꢂ6.899
0.8475
2.192
ꢂ0.4355
1.916
ꢂ1.318
0.2195
ꢂ0.2788
ꢂ0.4787
ꢂ0.1650
3.910
4.172
3.087
3.759
2.520
2.922
2.784
2.794
3.367
3.063
1.980
3.522
1.843
3.837
2.703
3.063
3.272
3.532
ꢂ0.839
ꢂ0.855
ꢂ0.864
ꢂ0.878
ꢂ0.883
ꢂ0.895
ꢂ0.892
ꢂ0.886
ꢂ0.851
ꢂ13.21
ꢂ13.94
ꢂ13.41
ꢂ14.37
ꢂ13.57
ꢂ14.26
ꢂ14.05
ꢂ13.86
ꢂ13.20
ꢂ1.201
ꢂ1.218
ꢂ1.264
ꢂ1.264
ꢂ1.314
ꢂ1.317
ꢂ1.317
ꢂ1.307
ꢂ1.235
ꢂ18.47
ꢂ19.39
ꢂ19.15
ꢂ20.20
ꢂ19.68
ꢂ20.48
ꢂ20.24
ꢂ19.95
ꢂ18.70
a) At the B3LYP/6-31G(d) level. b) Coordinate r(c, h) of electronic state of La3ꢀ and Eu3ꢀ is r(34.57, 14.88) and r(33.81, 8.89), respectively.
Fig. 2. Synthesis of Several Double-Stranded Amino Acid Chelators
Keys: (i) 90% TFA at 4 °C. (ii) Cat, DEPC, and TEA in dry DMF. (iii) 5% Pd/C–H2
in MeOH.
cally softer chelator than 7 or 8.
Fig. 3. Circular Dichroism Spectra of Double-Stranded Chelators
Synthesis and Chemical Properties The synthesis of
the target double-stranded amino acid chelators is shown in
Fig. 2. Boc-L-Phe-OH was conjugated to 1,8-(ethylene-
dioxy)bis(ethylamine) to yield a Boc-protected bis(Boc-L-
Circular dichroism spectra of (a) 14a, b and (b) 15a, b in 70% CH3CN (10 mM
HEPES buffer, pH 4.8) at 23ꢅ0.1 °C. Concentration of 14a, b and 15a, b:
1.50ꢆ10ꢂ4 M.
Phe)-N,N-(ethylenedioxy)bis(ethylamine) (11a) and 11a was conjugated with benzoic acid is 8.0 Hz, which is also an ac-
purified using silica gel chromatography with CHCl3 and 3% ceptable coupling constant for a b-sheet-like conformation.
MeOH/CHCl3. Removal of the Boc groups by trifluoroacetic Figures 3a and b show circular dichroism (CD) spectra of
acid (TFA) afforded the free compound (12a) in good 14a, b and 15a, b in 70% CH3CN (10 mM HEPES, pH 4.8) at
yield.16,19) Conjugated 2,3-bis(benzyloxy)benzoic acid was 23ꢅ0.1 °C, and support that these are pure optical isomers.
prepared from 2,3-dihydroxybenzoic acid,16) with the N-ter- The CD curve of 14a is obviously a mirror image of that of
minal of 12a afforded by treatment with DEPC (phosphoro- 14b, and the intensity of mole ellipticity ([q]) is equal to that
cyanidate) on dry N,N-dimethylformamide (DMF). Depro- of 14b (Fig. 3a). Two positive and negative bands of CD of
tection of the Bzl (benzyl) group in 13a using 5% Pd/C–H2 14b (or 14a) at 234 and 257 nm are observed. In contrast, the
gave a new peptide chelator (14a) in 65% yield. Both com- CD band at 257 nm was not observed in 15a (or 15b)-conju-
1
pounds, 14a and 14b, gave satisfactory results in IR, H- gated nonphenolic benzoic acid or 15a (or 15b) (Fig. 3b);
1
NMR, 13C-NMR, H–1H correlation spectroscopy (COSY), therefore, a second cotton effect of 14a (or 14b) at 257 nm
13C–1H COSY NMR, and FAB-MS analyses.
was assigned in conjugated catecholic OH groups.
The aH-1 of 14a observed at 4.82 ppm in CDCl3–DMSO-
Determination of Binding Constants To characterize
d6 (volume ratio; 0.5/0.2) is coupled with an amido proton the ability of synthesized chelators to form complexes with
3
(–NHbCO–) of 8.48 ppm and the observed Ja,NH is 7.9 Hz. lanthanides ions (Ln3ꢀ), the binding constant (Kb), molar
The value indicates that 12a has a b-sheet (7.0—10.0 Hz)- ratio (n : m) of the complex, and molar absorbance coefficient
like conformation. The observed 3Ja,NH value of aH-1 of 15a (e) were obtained from UV/Vis titration data in Benesi–

May 2010
623
Fig. 4. UV/Vis Spectral Changes (a—d) of 14a on the Addition of a Solution of Ln3ꢀ and Benesi–Hildebrand Plots (e) on Complexation of 14a with Lan-
thanide Ions Eu3ꢀ and Y3ꢀ
(a) 14a–Y3ꢀ, (b) 14a–La3ꢀ, (c) 14a–Eu3ꢀ, and (d) 14a–Lu3ꢀ complexes. Determination of the binding constant (KMY) of 14a–Eu3ꢀ (lmaxꢁ340 nm) and 14a–Y3ꢀ (lmaxꢁ350 nm)
complexes (e) with a Hildebrand plot.
Hildebrand plot analysis.23) Figures 4a—d show the spectral
change in UV/Vis titration studies on the binding of 14a with
Y3ꢀ, La3ꢀ, Eu3ꢀ, or Lu3ꢀ. UV/Vis spectra of the chelator 14a
were measured in 10 mM HEPES buffer in 70% MeCN (pH
4.8) and the first excited wavelength (lmax) was 313 nm.
The absorbance of a colorless solution of 14a at 313 nm
decreased on the addition of Eu3ꢀ from 2.60 to 129.0ꢆ
10ꢂ5 M under a fixed concentration of 14a (8.067ꢆ10ꢂ5 M).
However, the absorbance of the new wavelength of the
14a–Eu3ꢀ complex at 340 nm (lmax) increased on the addi-
tion of Eu3ꢀ (Fig. 4c). As the 14a–Eu3ꢀ complex has a 1 : 1
molar ratio, the 1 : 1 binding constant (KMY) of 14 for Ln3ꢀ is
calculated from the Benesi–Hildebrand plot. A charge in ab-
sorbance (ODi) is only induced by the formation of a 1 : 1
Fig. 5. Mean Binding Constant (KMY) of Molecular Complexes Obtained
from Spectrophotometric Titrations of Chelator 14a with Lanthanide Ions
complex between 14 ([C0], M) and Ln3ꢀ ion ([Di], M), and the
OD intensity (ODi) at maximum wavelength (lmax, 340 nm
when 14a) can be expressed by the following equations:
Each bar is the mean of two independent determinations at ꢈꢅ3% of the absolute
errors.
the 14a–Eu3ꢀ complex, the binding constant K14b–Eu3ꢀ was
measured by similar method: however, the K14b–Eu3ꢀ of D-form
14b was almost the same value with L-form 14a.
[C0 ]
1
1
ꢁ
ꢀ
(7)
(8)
ODi
[Di ]⋅ KMY ⋅Δε
Δε
1
1
ꢁ y -intercept and
ꢁslope
The shift of the second absorption from 313 to 317 nm of
Δε
Δε ⋅ KMY
the complexes of 14a with La3ꢀ was much smaller than that
Figure 4e shows Benesi–Hildebrand plots of the ab- of complexes of 14a with Y3ꢀ and Lu3ꢀ (Figs. 4a, b, d). The
sorbance at 350 nm or 340 nm as a function of Y3ꢀ or Eu3ꢀ KMY of 14a–La3ꢀ is 3.43ꢆ106 Mꢂ1, the largest in the Ben-
concentrations at a fixed concentration of 14a according to esi–Hildebrand plot. Although the absorption shift of the
Eqs. 7 and 8. From the slope, the KMY of the 14a–Y3ꢀ and 14a–La3ꢀ complex was small, the KMY value was the largest
14a–Eu3ꢀ complexes was 2.12 and 4.37ꢆ104 Mꢂ1, respec- in this study (Fig. 5). The lmax of 14a–Eu3ꢀ (lmax 340 nm)
tively. The binding constants of other complexes are shown shifted to a larger value than that of chelator 14a (lmax
in Fig. 5. Stabilization increases in the following order: 313 nm). The shifts of 14a–Ln3ꢀ complexes decreased in
14a–Lu3ꢀꢇ14a–Eu2ꢀꢇ14a–Y3ꢀꢇ14a–Eu3ꢀꢇ14a–La3ꢀ
.
the following order: 14a–Lu3ꢀꢄ14a–Y3ꢀꢄ14a–Eu3ꢀꢄ14a–
The result indicates that the KMY value of the complex be- La3ꢀ.
tween 14a and Ln3ꢀ is about 104—106 Mꢂ1, similar to that of
The reason for the difference in KMY, for instance 14a–
Ln3ꢀ with calix[6]arene analogs.24) In order to compare the Eu3ꢀ and 14a–La3ꢀ complexes, is as follows. A large shift in
binding ability of the enantiomer 14b–Eu3ꢀ complex with lmax indicates that the complex is polarized in the excited

624
Vol. 58, No. 5
state and ionizied; however, a small shift means that the com- The continuous variation method gives the result for plots of
plex is greatly polarized as in the case of the 14a–La3ꢀ com- [14a]/[Eu3ꢀ] vs. absorbance. The molar ratio was the result
plex. In the case of 14a–Eu3ꢀ, the large shift of lmax in the for plots of [14a]/[Eu3ꢀ] vs. absorbance for equilibrium
excited state is associated with little polarization of the com- under a fixed concentration of Eu3ꢀ (1.17ꢆ10ꢂ5 M) at pH 4.8.
plex; therefore, 14a and related chelators would be applica- With the continuous variation method (Fig. 6a), the plots
ble to the separation of Ln3ꢀ ions in solution. In addition, gave two linear slopes in the ratio of concentration of
em
ex
14a emits fluorescence with a l
of about 425 nm at l
[14a]/[Eu3ꢀ]: 5.74ꢆ10ꢂ5/5.16ꢆ10ꢂ4—4.59ꢆ10ꢂ4/1.15ꢆ
max
340
nm in 10 mM HEPES buffer in 70% MeCN (pH 4.8) (data not 10ꢂ4. It was found that the 14a–Eu3ꢀ complex has a 1 : 1
shown). Here, 15a did not bind with the Ln3ꢀ ions used in ratio since the intersection of the two slopes of correlation
this study since the shift and increase in intensity at lmax of coefficient rꢁ0.974 and 0.999 give a value of n/mꢁ1 for
the 15a–Ln3ꢀ complex was not observed by UV/Vis titration. altitude. Similar results were provided by the molar ratio
This shows that catecholic ligands provide the function of 14, method (Fig. 6b).
not amide groups.
The molar ratio (14a : Eu3ꢀꢁ1 : 1) is useful to estimate the
Molar Ratio and Structure The molar ratio of the com- chemical structure of the 14a–Ln3ꢀ complex. Generally,
plex between Ln3ꢀ and the chelator was analyzed by continu- Ln3ꢀ ions yield complexes with coordination numbers of 8—
ous variation (a)25) and molar ratio (b) methods (Figs. 6a, b). 12. To clarify the optimized structure satisfying conditions
such as a 1 : 1 ratio and coordination number of 8, we esti-
mated the optimized structure of 14a (Fig. 7a) and the
14a–La3ꢀ complex (Fig. 7c) using the ab initio molecular or-
bital (MO) method. Figure 7 shows the results of B3LYP hy-
brid density functional methods with Stuttgart–Dresden–
Bonn (SDD) effective core potentials (ECP)26) basis set using
the Gaussian 03 program27); however, the SDD ECP basis set
was used for Ln3ꢀ ions. C, H, N, and Cl atoms were opti-
mized at the 6-31G(d) basis set level. The calculated mean
O(4)–La3ꢀ and HO(3)–La3ꢀ bond lengths in the 14a–La3ꢀ
complex are about 2.306 and 2.747 Å, respectively (Table 2),
and the O–La3ꢀ bond length is close to the observed length
of the O–La3ꢀ bond (2.443 Å) in TREN-1,2-HOIQO–La3ꢀ.28)
The O(4)–Lu3ꢀ bond length (2.202 Å) in the 14a–Lu3ꢀ com-
plex is also similar to the observed length of the O–Lu3ꢀ
bond (2.287 Å) in the TREN-1,2-HOIQO–Lu3ꢀ complex.28)
The results indicate that the structures of 14a–La3ꢀ and
Fig. 6. Determination of Compositional Ratio of the 14a–Eu3ꢀ Complex
with the Continuous Variation Method (a) and Molar Ratio Method (b)
(a) Ratio of [14a/Eu3ꢀ]: 0.11, 0.25, 0.43, 0.67, 1.0, 1.5, 2.33, and 4.0; in 70% MeCN
(10 mM HEPES, pH 4.8). (b) Ratio of [14a/Eu3ꢀ]: 0.1, 0.2, 0.3, 0.5, 0.8, 1.0, 1.5, 2.0,
3.0 and 5.0; at lꢁ340 nm; in 70% MeCN (10 mM HEPES, pH 4.8). Concentration of
14a and Eu3ꢀ: 2.34ꢆ10ꢂ4 M and 1.17ꢆ10ꢂ5 M.
Fig. 7. Optimized Structure of 14a and the 14a–La3ꢀ Complex
Structure of (a) 14a and (c) the 14a–La3ꢀ complex shown as a ball and spoke model. At the B3LYP level using the SDD basis set for La, and an all-electron 6-31G(d) basis set
for the remaining atoms.
Table 2. Calculated Bond Lengths, Dipole Moment, and Energies in Optimized Complexes of Chelator 14a with Lanthanide Ions
Energy
(au)
Ln–O(2)
(Å)
Ln–O(4) Ln–O(3)H Ln–O(5)H
Ln–Cl
(Å)
Ln–O(6)H2 Ln–O(7)H2 Ln–O(8)H2 Dipole moment
(Å)
(Å)
(Å)
(Å)
(Å)
(Å)
(debye)
14aa)
ꢂ2444.671
ꢂ3568.838
ꢂ4369.489
—
2.286
2.144
—
2.306
2.202
—
2.747
2.433
—
2.721
2.471
—
2.831
2.670
—
2.597
2.355
—
2.965
2.513
—
2.642
2.418
1.433
3.523
2.634
14a–La3ꢀ b)
14a–Lu3ꢀ b)
a) At the B3LYP/6-31G(d ) level. b) At B3LYP/6-31G(d ), and SDD basis set was used for the optimization of La3ꢀ and Lu3ꢀ
.

May 2010
625
forms the random conformation (16) to the folding confor-
mation (17). Next, the chemically soft acid Eu3ꢀ interacts
with a chemically soft base, catecholic hydroxide of 14a, and
the complex (18) is stabilized by the valance, in both ionic
size and chemical hardness. Molecular orbital (MO) calcula-
tions using the B3LYP method provided the geometry of 14a
in which two catechols were opposite; however, binding
of Kꢀ (or Liꢀ) to the spacer –CH2CH2–O–CH2CH2–O–
CH2CH2– would fix the flexible conformation of 14a. As ex-
pected, the two catechols formed binding sites of 14a, which
are face to face and close. This provides a new functional
chelator for controlling the binding of 14a with Ln3ꢀ in the
presence of alkaline metal ions. Finally, this results indicates
that chelator 14 is active in displacing the ligand in the solu-
tion and a dynamic flexible→rigid equilibrium is produced in
Kꢀ (or Liꢀ) and Ln3ꢀ solutions.
Conclusion
We reported the synthesis and chemical properties of a
series of new double-stranded phenylalanine chelators 14
conjugated to catechol. The interactions between double-
stranded amino acid chelators (or other ligands) and lan-
thanide series ions (Ln3ꢀ) were discussed using h–c dia-
grams as coordinates of electron structure r(c, h) based on
the chemical hardness theory. According to chemical hard-
ness, 14 is a chemically soft base chelator containing a
chemically hard base –CH2CH2–O–CH2CH2–O–CH2CH2–
spacer. The diagram suggested that catechol ligand provides
stable chelators for Ln3ꢀ ions, but not Liꢀ and Kꢀ ions;
Fig. 8. Changes of UV Spectral Absorbance of (a) 14a on Addition of
Eu3ꢀ with Kꢀ and (b) Folding Model of 14a Bound with Kꢀ Ions
Arrow shows that hard Kꢀ moves into a hard base site from soft base Cat.
–Lu3ꢀ complexes are similar to the calculation results (Fig. therefore, 14–conjugated catechol produced a stable solvated
7c). In the 14a–Lu3ꢀ complex, the O(4)–Lu3ꢀ bond length 14–Ln3ꢀ complex. Indeed, binding constants of the 14–Ln3ꢀ
(2.202 Å) is about 0.104 Å shorter than that of 14a–La3ꢀ. complex were 104—106 Mꢂ1. Interestingly, the order of bind-
The difference in length suggests an effect of the lanthanide ing stability is identical to the order of the lanthanide con-
contraction of Lu3ꢀ. Spectrophotometric titration data show traction effect: Lu3ꢀꢇEu3ꢀꢇY3ꢀ. Spectrophotometric titra-
that the binding constant of the 14a–Ln3ꢀ complexes in- tion also provided evidence that the binding constant for
creases in the following order: 14a–Lu3ꢀꢇ14a–Eu2ꢀꢇ14a– Eu3ꢀ and 14a increased in the presence of a chemical hard
Y3ꢀꢇ14a–Eu3ꢀꢇ14a–La3ꢀ, in the buffer (pH 4.8). Interest- ion, Liꢀ (Naꢀ or Kꢀ) (Fig. 5). This result suggested that 14 is
ingly, this order may be identical to the order of the lan- ligand displacement active, and that the motion of the molec-
thanide contraction effect: Lu3ꢀꢇEu3ꢀꢇLa3ꢀ (Fig. 5).
ular skeleton is flexible in solution. Our results indicate the
The three mean H2O–La3ꢀ bond lengths in the 14a–La3ꢀ h–c diagram is a useful tool for developing functional chela-
complex are 2.734 Å. The H2O–La3ꢀ bond length of the tors. Further studies of the isolation of the 14–Ln3ꢀ complex
14a–La3ꢀ complex increased about 0.305 Å more than the and inhibitory effect on cell growth are in progress.
H2O–Lu3ꢀ bond of the 14a–Lu3ꢀ complex.
Effect of Hard Acid Liꢀ and Kꢀ Ions Chemical soft
Experimental
General Methods The general approach to the synthesis of the com-
and hard binding regions exist in 14. Liꢀ or Kꢀ ions are
pounds (11, 12) was described in our previous papers.17,18) The peptide and
harder than Eu3ꢀ, as shown in the h–c diagram (Fig. 1a). In
related ligands were detected on TLC (thin-layer chromatography) plates
using iodine vapor or UV (ultraviolet spectrum) absorption. Silica gel col-
umn chromatography was performed on silica gel 60N (100 mesh, neutral;
order to elucidate the effect of Kꢀ and Liꢀ ions, UV/Vis
titrations of 14a with Eu3ꢀ were performed in the presence
Kanto Chemical Co., Japan). Solvent systems were as follows, A: CHCl3–
of Kꢀ or Liꢀ (7.70ꢆ10ꢂ5 M). Titration was performed at a
MeOH (20 : 1), and B: CHCl3–MeOH (10 : 1). UV/vis (ultraviolet/visible)
spectra were measured with a JASCO V-530 spectrophotometer. pH was
molar ratio of 1 : 1 of [Kꢀ]/[14a] in 10 mM HEPES buffer in
70% MeCN (pH 4.8) (Fig. 8a). The binding constant of the
14a–Eu3ꢀ complex with Kꢀ is 5.94ꢆ104 Mꢂ1. The binding
constant increased about 40% without Kꢀ. In the presence of
chemically harder Liꢀ, the binding constant of the 14a–Eu3ꢀ
complex (6.71ꢆ104 Mꢂ1) increased about 53.5% more than
with Kꢀ.
measured with a TOA pH instrument (Model HM-60G). Specific rotations
were measured with a JASCO DIP-140 digital polarimeter. NMR (nuclear
1
magnetic resonance, as 13C-NMR, H-NMR) spectra were obtained with a
JEOL a-500 or a Bruker AV300 spectrometer, and NMR samples were dis-
solved in DMSO-d6/CDCl3 (volume ratio: 0.5 ml/0.2 ml) with TMS (tetra-
methylsilane) as an internal reference. FAB-MS (fast atom bombardment
mass spectrometry) spectral data were obtained on a JEOL LMS-HX110
spectrometer, and relevant data were tabulated as m/z.
Synthesis of Bis(Cat-L-Phe)-N,N-(ethylenedioxy)bis(ethylamine) (14a).
Bis(L-Phe)-N,N-(ethylenedioxy)bis(ethylamine) (12a) CDI (1,1ꢃ-car-
bonyldiimidazole) (4.45 g, 26 mmol) was added to a solution of Boc-L-Phe-
OH (5.08 g, 19.2 mmol) in dry CHCl3 (40 ml). The solution was stirred
The increase in K13a–Eu3ꢀ may be explained by the illustra-
tion (Fig. 8b). First, equilibrium moves to the right with the
coordination of Kꢀ (or Liꢀ) to a chemically hard base
–O–CH2CH2–O– spacer of 14a, which provides the folding
conformation. The Kꢀ–(–O–CH2CH2–O–) complex trans- for 1 h at room temperature, 2,2ꢃ-(ethylenedioxy)bis(ethylamine) (1.26 g,

626
Vol. 58, No. 5
8.5 mmol) was added, and the mixture was stirred overnight. This solution
surements were made along a 1 cm path length in 10 mM HEPES (pH 4.8) in
was evaporated dry. After the addition of aq. MeOH, the sediment was col- 70% MeCN, with a circular quartz cell at 23ꢅ0.1 °C. The CD spectra are
lected, washed with aq. MeOH, 5% citric acid, 5% NaHCO3, and water, and shown in Fig. 3 with molar ellipticity, [q] (deg·cm2·dmlꢂ1), versus wave-
dried in vacuo. The resulting product, bis(Boc-L-Phe)-N,N-(ethylenedioxy)-
bis(ethylamine) (11a), a colorless solid, was obtained in 80% yield. The
length, l (nm).
Determination of Binding Constants The binding constants (Kb) were
crude product was purified by chromatography on a silica gel column (41 g) obtained by the UV/Vis titration method. Stock solutions of 14a and lan-
and eluted in CHCl3 and 3% MeOH/CHCl3 (stepwise elution); mp 92—
thanide ions (Ln3ꢀ) were prepared in the range of 8.0—8.1ꢆ10ꢂ4 M and
94 °C (from MeCN). Rf(A)ꢁ0.37. FAB-MS (nitrobenzylalchohol, NBA) 2.5—2.6ꢆ10ꢂ3 M in 10 mM HEPES (70% MeCN, pH 4.8), respectively and
m/z: 643 (MꢀH)ꢀ.
were mixed just before recording the spectra. The concentration of 14a was
To 11a (4.57 g, 5.04 mmol) was added 90% TFA (26 ml). The mixture was fixed to 14a–Y3+: [14a]: 8.12ꢆ10ꢂ5 M, [Y3ꢀ]: 0.0, 1.5, 3.0, 6.1, 9.1, 15.2,
stirred for 1 h in an ice bath. After a 5% NaHCO3 and 1 M NaOH workup,
24.3, 30.4, 45.6, 60.8, 91.2, 121.6, and 152.0ꢆ10ꢂ5 M. 14a–La3ꢀ
:
product 12a was extracted with CHCl3, the organic layer was washed with [14a]: 1.02ꢆ10ꢂ4 M, [La3ꢀ]: 0.0, 1.3, 2.5, 5.1, 7.6, 12.7, 20.2, 25.3, 38.0,
water and dried with dry Na2SO4, and 12a was obtained in 75.0% yield; 50.6, 75.9, 101.2, and 126.5ꢆ10ꢂ5 M. 14a–Eu3ꢀ: [14a]: 8.07ꢆ10ꢂ5 mol/l,
Rf(B)ꢁ0.16. Oily; 1H-NMR (CDCl3/DMSO-d6ꢁ0.5/0.2) d: 3.33 (1H, m,
[Eu3ꢀ]: 0.0, 2.6, 5.2, 7.7, 12.9, 20.6, 25.8, 38.7, 51.6, 77.4, 103.2, and
–CH2–N–), 3.47—3.50 (1H, m, –O–CH2–), 3.52—3.55 (1H, m, –CH2–O–),
129.0ꢆ10ꢂ5 M. 14a–Lu3ꢀ: [14a]: 8.05ꢆ10ꢂ4 M, [Lu3ꢀ]: 0.0, 1.3, 2.6, 5.2,
2.69 (1H, dd, Jꢁ6.9, 8.8 Hz, bH, –C6H5–CH–), 3.11 (1H, dd, Jꢁ5.5, 7.7, 12.9, 20.6, 25.8, 38.7, 51.6, 77.4, 103.2, and 129.0ꢆ10ꢂ5 M. 14a–Eu2ꢀ
:
6.9 Hz, bH, –C6H5–CH–), 3.47—3.55 (1H, m, –CaH–), 7.19—7.21 (5H, [14a]: 8.05ꢆ10ꢂ5 M, [Eu2ꢀ]: 0.0, 1.3, 2.6, 5.2, 7.7, 12.9, 20.6, 25.8, 38.7,
m, aromatic protons), 7.77 (1H, t, Jꢁ5.5 Hz, –NHaCO–). 13C-NMR
(CDCl3/DMSO-d6ꢁ0.5/0.2) d: 38.63, 41.06, 56.30, 69.48, 69.92, 126.34,
128.28, 129.19, 138.15, and 174.34. FAB-MS m/z: 443 (MꢀH)ꢀ.
51.6, 77.4, 103.2, and 129.0ꢆ10ꢂ5 M.
Regarding the effect of Kꢀ ions on the formation of a molecular complex
of Eu3ꢀ with 14a, the binding constant of 14a–Eu3ꢀ was obtained using a
prepared concentration of [14a]: 7.14ꢆ10ꢂ5 M, [Kꢀ]: 7.7ꢆ10ꢂ5 M, and
Bis(Cat-L-Phe)-N,N-(ethylenedioxy)bis(ethylamine) (14a) DEPC
(phosphorocyanidate) (0.83 g, 5.1 mmol)29) was added to a solution of 2,3- [Eu3ꢀ]: 0.0, 1.2, 2.3, 4.6, 7.0, 11.6, 18.6, 23.2, 34.8, 46.4, 69.6, 92.8 and
bis(benzyloxy)benzoic acid (1.21 g, 3.63 mmol), 12a (0.697 g, 1.58 mmol), 116.0ꢆ10ꢂ5 M.
and TEA (triethylamine, 0.4 ml) at 4 °C in dry DMF (20 ml). The mixture
was stirred first at 4 °C for 1 h, and then at room temperature overnight. Ice and 14a (5.74ꢆ10ꢂ4 M) were prepared in 10 mM HEPES (70% MeCN, pH
water was added and the solution extracted several times with CHCl3. The
4.8). Before the measurement, the Eu3ꢀ and 14a solutions were mixed and a
combined organic extracts were dried over anhydrous Na2SO4 and evapo- standard solution was prepared with a [14a/Eu3ꢀ] ratio of 0.11, 0.25, 0.43,
rated dry. The crude residue was chromatographed on silica gel (41 g) with
0.67, 1.00, 1.50, 2.33, or 4.00. In all the spectra of 14a–Eu3ꢀ complex in
Continuous Variation Method Stock solutions of Eu3ꢀ (5.74ꢆ10ꢂ4 M)
CHCl3 and 3% MeOH/CHCl3 (stepwise elution) as eluents. 13a was ob- each [14a/Eu3ꢀ] ratio, absorbance (OD) was measured at maximum bands.
tained as a colorless solid in 83.0% yield; Rf(A)ꢁ0.66; HR-FAB-MS (NBA)
Molar Ratio Method Stock solutions of Eu3ꢀ (2.34ꢆ10ꢂ4 M) and 14a
(2.34ꢆ10ꢂ4 M) were prepared in 10 mM HEPES (70% MeCN, pH 4.8). To a
solution of Eu3ꢀ (1.17ꢆ10ꢂ5 M) was added a solution of 14a (2.34ꢆ10ꢂ4 M)
m/z: 1075.4863 (Calcd for C66H67N4O10 (MꢀH)ꢀ: 1075.4857).
To a solution of 13a (0.92 g, 0.85 mmol) in MeOH (60 ml) was added 5%
Pd–C (0.276 g). The mixture was shaken under a flow of H2. After the reac- to prepare [14a/Eu3ꢀ] with a molar ratio of 0.1, 0.2, 0.3, 0.5, 0.8, 1.0, 1.5,
tion was completed, the catalyst was removed with a glass filter (size G3-4). 2.0, 3.0, and 5.0. In all spectra of the 15a–Eu3ꢀ complex in each [14a/Eu3ꢀ
The filtrate was suspended in ice water and extracted several times with ratio, absorbance (OD) was measured at 340 nm.
CHCl3. The organic layer was dried with dry Na2SO4, filtered, and evapo-
]
rated. The residue was dried under vacuum to provide 14a as a pale yellow
Acknowledgments We thank Ms. Tamiko Kiyotani and Mr. Youichi
crystal in 75.3% yield; Rf(B)ꢁ0.48; mp 89—92 °C. [a]D23 ꢂ12.0° Takase (Showa Pharmaceutical University, Machida, Tokyo, Japan) for
(cꢁ0.00241, EtOH). 1H-NMR (CDCl3/DMSO-d6ꢁ0.5/0.2) d: 3.40—3.42 NMR and FAB-MS measurements.
(1H, m, –CH2–N–), 3.44—3.49 (1H, m, –O–CH2–), 3.55—3.60 (1H, m,
–CH2–O–), 3.11 (1H, dd, Jꢁ8.9, 13.7 Hz, bH, –C6H5–CH–), 3.20 (1H, dd, References and Notes
Jꢁ5.5, 7.0 Hz, bH, –C6H5–CH–), 4.82 (1H, m, –CaH–), 6.63—6.72 and
6.92—7.29 (3H, m, catechol protons), 7.16—7.29 (5H, m, aromatic pro-
tons), 7.80 (1H, t, Jꢁ5.3 Hz, –NHaCO–) and 8.48 (1H, m, Jꢁ7.9 Hz,
–NHbCO–). 13C-NMR (CDCl3/DMSO-d6ꢁ0.5/0.2) d: 37.63, 39.08, 54.74,
69.28, 69.94, 117.70 (catecholic C), 118.05 (catecholic C), 118.91 (cate-
cholic C), 126.38, 128.13, 129.19, 137.48, 146.06, 149.26, 169.44, and
171.08. HR-FAB-MS (NBA) m/z: 715.2968 (Calcd for C38H43N4O10
(MꢀH)ꢀ: 715.2984). FAB-MS (NBA) m/z: 715 (MꢀH)ꢀ.
Bis(Cat-D-Phe)-N,N-(ethylenedioxy)bis(ethylamine) (14b) 14b was
prepared in a similar manner to 14a and obtained as a pale yellow solid in
55% yield; Rf(B)ꢁ0.49. mp 89—92 °C. [a]D23 ꢀ13.0° (cꢁ0.00249, EtOH).
HR-FAB-MS (NBA) m/z: 715.297 (Calcd for C38H43N4O10 (MꢀH)ꢀ:
715.2981). FAB-MS (NBA) m/z: 715 (MꢀH)ꢀ.
1) Anderson C. J., Welch M. J., Chem. Rev., 99, 2219—2234 (1999).
2) Wu F. B., Han S. Q., Zhang C., He Y. F., Anal. Chem., 74, 5882—5889
(2002).
3) Bunzli J.-C. G., Piguet C., Chem. Soc. Rev., 34, 1048—1077 (2005).
4) Heather L. H., Gillies R. J., Life Sci., 77, 361—371 (2005).
5) Shibazaki M., Sasaki H., Arai T., Angew. Chem., Int. Ed. Engl., 33,
773—775 (1997).
6) Martínez-Máñez R., Sancenón F., Chem. Rev., 103, 4419—4476
(2003).
7) Harte A. J., Jensen P., Plush S. E., Kruger P. E., Gunnlaugsson T.,
Inorg. Chem., 45, 9465—9474 (2006).
8) Thibon A., Pierre V. C., Anal. Bioanal. Chem., 394, 107—120 (2009).
9) Steed J. W., Turner D. R., Wallace K. J., “Core Concepts in Supramole-
cular Chemistry and Nanochemistry,” John Wiley & Sons, Ltd., Hobo-
ken, 2007.
Bis(benzoyl-L-Phe)-N,N-(ethylenedioxy)bis(ethylamine) (15a) 15a
was prepared in a similar manner to 13a and obtained as a colorless solid in
78% yield; Rf(A)ꢁ0.43; mp 185—187 °C. [a]D23 ꢂ6.7° (cꢁ0.00161, EtOH). 10) Anderegg G., Arnaud-Neu F., Delgado R., Felcman J., Popov K., Pure
1H-NMR (CDCl3/DMSO-d6ꢁ0.5/0.2) d: 3.31—3.35 (1H, m, –CH2–N–),
3.38—3.41 (1H, m, –O–CH2–), 3.43—3.50 (1H, m, –CH2–O–), 3.08 (1H, 11) Wang C., Seema Choudhary S., Vink C. B., Secord E. A., Morrow J.
dd, Jꢁ6.9, 8.8 Hz, bH, C6H5–CH–), 3.20 (1H, dd, Jꢁ5.8, 7.0 Hz, bH, R., Chem. Commun., 2000, 2509—2510 (2000).
C6H5–CH–), 4.86 (1H, m, –CaH–), 7.13—7.30 (5H, m, –CO–C6H5), 7.35— 12) Fricker S. P., Chem. Soc. Rev., 35, 524—533 (2006).
Appl. Chem., 77, 1445—1495 (2005).
7.76 (5H, m, C6H5–CH2–), 7.81 (1H, t, Jꢁ5.3 Hz, –NHaCO–) and 8.16 (1H,
d, Jꢁ8.0 Hz, –NHbCO–). 13C-NMR (CDCl3/DMSO-d6ꢁ0.5/0.2) d: 37.80,
39.03, 54.91 (aC), 69.30, 69.98, 126.30, 127.31, 128.07, 129.21, 131.22,
134.13, 137.69, 166.86, and 171.49. HR-FAB-MS (NBA) m/z: 651.3182
(Calcd for C38H43N4O6 (MꢀH)ꢀ: 651.3183). FAB-MS (NBA) m/z: 651
(MꢀH)ꢀ.
Bis(benzoyl-D-Phe)-N,N-(ethylenedioxy)bis(ethylamine) (15b) Com-
pound 15b was prepared in a similar manner to 15a and obtained as a color-
less solid in 75% yield; Rf(A)ꢁ0.43. mp 185—187 °C. [a]D23 ꢀ7.0°
(cꢁ0.00298, EtOH). HR-FAB-MS (NBA) m/z: 651.3165 (Calcd for
C38H43N4O6 (MꢀH)ꢀ: 651.3193).
13) Parr R. G., Pearson R. G., J. Am. Chem. Soc., 105, 7512—7516
(1983).
14) Pearson R. G., J. Am. Chem. Soc., 107, 6801—6806 (1985).
15) Parr R. G., Szentpaly L. V., Liu S., J. Am. Chem. Soc., 12, 1922—1924
(1999).
16) Kobayashi S., Mizushima A., Sasuga A., Watanabe M., Chem. Pharm.
Bull., 54, 761—763 (2006).
17) Kobayashi S., Shinohara H., Tabata K., Yamamoto N., Miyai M.,
Chem. Pharm. Bull., 54, 1633—1638 (2006).
18) Kobayashi S., Kobayashi H., Yamaguchi T, Nishida M., Yamaguchi Y.,
Kurihara M., Miyata N., Tanaka A., Chem. Pharm. Bull., 48, 920—
934 (2000).
Circular Dichroism (CD) Measurements The CD spectra of 14a, 14b,
15a, and 15b were measured on a JASCO J-600 spectrophotometer. Mea- 19) Kobayashi S., Atuchi N., Wakamatsu H., Hattori M., Kawada A.,

May 2010
Asano K., Chem. Pharm. Bull., 55, 1585—1592 (2007).
20) The data of lanthanoide elements from lanthanum (La) to lutetium
(Lu) were taken from the www site: <http://en.wikipedia.org/wiki/
Lanthanoid>, from Wikipedia, the free encyclopedia, 29 January 2010.
21) Cao X, Li Q, Moritz A, Xie Z, Dolg M, Chen X, Fang W., Inorg.
Chem., 45, 3444—3451 (2006).
22) Hughes I. D., Dane M, Ernst A., Hergert W., Luders M., Poulter J.,
Staunton J. B., Svane A., Szotek Z., Temmerman W. M., Nature (Lon-
don), 446, 650—653 (2007).
23) Benesi H. A., Hildebrand J. H., J. Am. Chem. Soc., 71, 2703—2707
(1949).
627
Vreven Jr., T., Kudin K. N., Burant J. C., Millam J. M., Iyengar S. S.,
Tomasi J. Barone V., Mennucci B., Cossi M., Scalmani G., Rega N.,
Petersson G. A., Nakatsuji H., Hada M., Ehara M., Toyota K., Fukuda
R., Hasegawa J., Ishida M., Nakajima T., Honda Y., Kitao O., Nakai
H., Klene M., Li X., Knox J. E., Hratchian H. P., Cross J. B., Bakken
V., Adamo C., Jaramillo J., Gomperts R., Stratmann R. E., Yazyev O.,
Austin A. J., Cammi R., Pomelli C., Ochterski J. W., Ayala P. Y., Moro-
kuma K., Voth G. A., Salvador P., Dannenberg J. J., Zakrzewski V. G.,
Dapprich S., Daniels A. D., Strain M. C., Farkas O., Malick D. K.,
Rabuck A. D., Raghavachari K., Foresman J. B., Ortiz J. V., Cui Q.,
Baboul A. G., Clifford S., Cioslowski J., Stefanov B. B., Liu G.,
Liashenko A., Piskorz P., Komaromi I., Martin R. L., Fox D. J., Keith
T., Al-Laham M. A., Peng C. Y., Nanayakkara A., Challacombe M.,
Gill P. M. W., Johnson B., Chen W., Wong M. W., Gonzalez C., Pople
J. A., Gaussian, Inc., Wallingford CT, 2004.
24) Liu J.-M., Chen C.-F., Zheng Q.-Y., Huang Z.-T., Tetrahedron Lett., 45,
6071—6074 (2004).
25) Yoe J. H., Jones A. L., Ind. Eng. Chem. Anal. Ed., 16, 111—115
(1944).
26) Lindsay E. R., Jakubikova E., Guthrie M. G., Batista E. R., J. Phys. 28) Seitz M., Oliver A. G., Raymond N., J. Am. Chem. Soc., 129, 11153—
Chem. A, 113, 6745–6750 (2009). 11160 (2007).
27) Gaussian 03, Revision D.02, Frisch M. J., Trucks G. W., Schlegel H. 29) Shioiri T., Yokoyama Y., Kasai Y., Yamada S., Tetrahedron, 32,
B., Scuseria G. E., Robb M. A., Cheeseman J. R., Montgomery J. A., 2211—2217 (1976).