New chelating ligands for Co(III)-based peptide-cleaving catalysts selective for pathogenic proteins of amyloidoses

Article abstract of DOI:10.1007/s00775-010-0750-y

The Co(III) complex of 1,4,7,10-tetraazacyclododecane has been employed as the catalytic center of target-selective peptide-cleaving catalysts in previous studies. As new chelating ligands for the Co(III) ion in the peptide-cleaving catalysts, 1-oxo-4,7,1

Full text of DOI:10.1007/s00775-010-0750-y

J Biol Inorg Chem (2011) 16:511–519
DOI 10.1007/s00775-010-0750-y
ORIGINAL PAPER
New chelating ligands for Co(III)-based peptide-cleaving catalysts
selective for pathogenic proteins of amyloidoses
•
Woo Suk Chei Heeyeon Ju Junghun Suh
•
Received: 30 July 2010 / Accepted: 7 December 2010 / Published online: 18 December 2010
Ó SBIC 2010
Abstract The Co(III) complex of 1,4,7,10-tetraazacy-
clododecane has been employed as the catalytic center of
target-selective peptide-cleaving catalysts in previous
studies. As new chelating ligands for the Co(III) ion in the
peptide-cleaving catalysts, 1-oxo-4,7,10-triazacyclodede-
cane, 1-aryl-1,4,7,10-tetraazacyclodecane, and 7-aryl-1-
oxo-4,7,10-triazacyclodecane were examined in the present
study. A chemical library comprising 612 derivatives of the
Co(III) complex of the new chelating ligands was con-
structed. The catalyst candidates were tested for their
activity to cleave the soluble oligomers of amyloidogenic
peptides amyloid b-42 and human islet amyloid polypep-
tide (h-IAPP), which are believed to be the pathogenic
species for Alzheimer’s disease and type 2 diabetes mel-
litus, respectively. One derivative of the Co(III) complex of
1-aryl-1,4,7,10-tetraazacyclodecane was found to cleave
the oligomers of h-IAPP. Cleavage products were identi-
fied and cleavage yields were measured at various catalyst
concentrations for the action of the new catalyst. The
present results reveal that effective catalytic drugs for
amyloidoses may be obtained by using Co(III) complexes
of various chelating ligands.
Abbreviations
Ab42
AmP
Amyloid b-42
Amyloidogenic peptide or protein
arylcyclen
1-Aryl-1,4,7,10-tetraazacyclodecane
aryloxacyclen 7-Aryl-1-oxo-4,7,10-triazacyclodecane
cyclen
DIEA
h-IAPP
HPLC
MALDI
MS
1,4,7,10-Tetraazacyclododecane
Diisopropylethylamine
Human islet amyloid polypeptide
High-performance liquid chromatography
Matrix-assisted laser desorption/ionization
Mass spectrometry
oxacyclen
T2DM
TFA
1-Oxo-4,7,10-triazacyclodedecane
Type 2 diabetes mellitus
Trifluoroacetic acid
TOF
Time of flight
Introduction
We have proposed that catalytic drugs based on target-
selective peptide-cleaving catalysts may become a new
paradigm in drug design [1–4]. A small synthetic molecule
that can recognize a disease-related protein or oligopeptide
and cleave the peptide backbone of the target selectively
can destroy the bioactivity of the target. This may cure the
disease related to the target. Conventional drugs targeting
disease-related proteins recognize and strongly bind to the
active sites and, therefore, cannot be applied to proteins
that do not have active sites. On the other hand, peptide-
cleaving catalysts can recognize any part of the target and,
therefore, can be used even for pathogenic proteins without
active sites.
Keywords Amyloid disease Á Catalytic drug Á
Co(III) complex Á Peptide-cleaving catalyst Á
1-Aryl-1,4,7,10-tetraazacyclodecane
W. S. Chei Á H. Ju Á J. Suh (&)
Department of Chemistry,
Seoul National University,
Seoul 151-747, Korea
The first peptide-cleaving catalysts selective for proteins
lacking active sites have been designed [5–7] by using the
e-mail: jhsuh@snu.ac.kr
123

512
J Biol Inorg Chem (2011) 16:511–519
Structure 1
NH
NH
III
Co
NH
NH
III
NH
III
Co
O
O
NH
NH
III
NH
NH
NH
Co
OH2
OH2
Co
OH2
OH2
OH2
N
NH
N
NH
OH2
OH2
OH2
Ar
Ar
Co(III)cyclen
Co(III)oxacyclen
Co(III)arylcyclen
Co(III)aryloxacyclen
Co(III) complexes of 1,4,7,10-tetraazacyclododecane (cy-
clen) derivatives to cleave the soluble oligomers of an
amyloidogenic peptide or protein (AmP) such as amyloid
b-42 (Ab42), human islet amyloid polypeptide (h-IAPP), or
a-synuclein. It is generally believed that the soluble olig-
omers of Ab42, h-IAPP, and a-synuclein are the pathogenic
species [8–10] of amyloidoses [8] such as Alzheimer’s
disease, type 2 diabetes mellitus (T2DM), and Parkinson’s
disease, respectively. The peptide-cleaving catalysts
selective for the soluble oligomers of AmP provided a new
therapeutic option for amyloid diseases [4–7]. Since con-
ventional drugs targeting active sites of disease-related
proteins cannot be designed for AmP, clinical testing and
therapeutic application of peptide-cleaving catalysts are
important in curing amyloidoses.
In the present study, attempts were made to discover
new peptide-cleaving catalysts for the pathogenic proteins
for Alzheimer’s disease or T2DM by employing three new
chelating ligands for the catalytic center. Introduction of an
aryl moiety to a nitrogen atom of cyclen to obtain 1-aryl-
1,4,7,10-tetraazacyclodecane (arylcyclen) lowers the basi-
city of cyclen. Likewise, introduction of an aryl moiety to a
nitrogen atom of oxacyclen to obtain 7-aryl-1-oxo-4,7,
10-triazacyclodecane (aryloxacyclen) lowers the basicity
of oxacyclen. Discovery of new catalysts containing
the Co(III) complex of oxacyclen [Co(III)oxacyclen], the
Co(III) complex of arylcyclen [Co(III)arylcyclen], or the
Co(III) complex of aryloxacyclen [Co(III)aryloxacyclen]
(Structure 1) was undertaken in this study.
Both the Cu(II) complex [Cu(II)cyclen] of cyclen and
the Co(III) complex [Co(III)cyclen] of cyclen can be
exploited as the catalytic center in designing target-selec-
tive peptide-cleaving catalysts [11, 12]. Co(III)cyclen is,
however, more appropriate [5–7, 13] as the catalytic site in
view of the exchange-inertness [14] of Co(III) complexes.
The metal ions in exchange-labile complexes such as
Cu(II)cyclen may be easily extracted by metal-sequestering
agents present in biotic environments, leading to loss of the
catalytic site and release of free cyclen. Cyclen has very
high affinity for various metal ions [15] and may be toxic.
Moreover, the Co(III)cyclen-containing catalysts previ-
ously reported for cleavage of the soluble oligomers of
Ab42, h-IAPP, or a-synuclein have been found to penetrate
animal cells readily [16].
Results
To obtain peptide-cleaving catalysts selective for patho-
genic proteins for Alzheimer’s disease or T2DM based on
new chelating ligands, a chemical library was constructed
as summarized in Scheme 1. Here, derivatives of
Co(III)oxacyclen, Co(III)aryloxacyclen, and Co(III)aryl-
cyclen were synthesized by attaching various substituents
to 1,3,5-triazine. The substituents used in the construction
of the library are summarized in Figs. 1, 2, and 3. Three
compounds (1h–1j in Fig. 2) were used as component R2-
NRH to obtain candidates of peptide-cleaving catalysts
based on Co(III)oxacyclen. When 1k (Fig. 2) is used as
component R2-NRH, the secondary amine of 1k is attached
to the triazine ring, leading to the formation of candidates
of peptide-cleaving catalysts based on Co(III)aryloxacy-
clen. When 1l (Fig. 2) is used as component R2-NRH, the
secondary amine of 1l is attached to the triazine ring,
resulting in the formation of candidates of peptide-cleaving
catalysts based on Co(III)arylcyclen.
The search for a more effective chelating ligand for
Co(III) ion is important in the discovery of therapeuti-
cally safe catalytic drugs for amyloidoses based on the
Co(III) complexes. In a previous study, the proteolytic
activity of the Co(III) complex of 1-oxo-4,7,10-triazacy-
clodedecane (oxacyclen) was found to be higher than that
of Co(III)cyclen in the cleavage of albumin, c-globulin,
and myoglobin [17]. The improved proteolytic activity
may be related to the lower basicity of oxacyclen com-
pared with cyclen, leading to the greater Lewis acidity of
the metal center in the Co(III) complex. In addition, the
structural change accompanying the modification of the
chelating ligand might have some effects on the proteo-
lytic activity.
To obtain the derivatives of Co(III)oxacyclen, 1e–1g
(Fig. 1) were used as component R1-XH and group A and
group B amines (Fig. 3) were used as component R3-NRH.
In total, 387 compounds were prepared as the derivatives of
Co(III)oxacyclen. To obtain the derivatives of Co(III)ary-
loxacyclen, 1e–1g (Fig. 1) were employed as component
R1-XH and group A and group B amines (Fig. 3) were
employed as component R3-NRH. In total, 129 compounds
123

J Biol Inorg Chem (2011) 16:511–519
513
Cl
Cl
Cl
N
R1-XH
N
N
N
N
R2-NRH
DIEA
N
(X = O or NH)
RN
R2
N
X R1
Cl
N
X R1
DIEA
Cl
N
Cl
1c
1b
1a
R3
RN
R3-NRH
DIEA
N
N
removal of
CoCl2, O2
a Co(III) complex
protecting
groups
RN
R2
N
X R1
1d
Scheme 1 Synthetic route for construction of the chemical library
of candidates for peptide-cleaving catalysts containing Co(III)
complexes of 1-oxo-4,7,10-triazacyclodedecane [Co(III)oxacyclen],
7-aryl-1-oxo-4,7,10-triazacyclodecane [Co(III)aryloxacyclen], and
1-aryl-1,4,7,10-tetraazacyclodecane
diisopropylethylamine
[Co(III)arylcyclen].
DIEA
H2N
O
N
N
N
S
HO
N
H2N
N
S
1e
1g
1f
Fig. 1 Compounds used as component R1-XH in construction of the library; the primary amino or the hydroxyl group is attached to the triazine
ring
the library showing activity to cleave Ab42 or h-IAPP were
boc
N
boc
N
boc
N
synthesized on a larger scale, thoroughly purified, and
tested again for the proteolytic activity to confirm the
activity. Only one compound (A; Structure 2) was found to
have the activity necessary to cleave h-IAPP and no com-
pound was able to cleave Ab42. Among the 612 compounds,
only compound A appears to contain the organic pendant
leading to recognition and peptide cleavage of the oligo-
mers of h-IAPP.
H
NH2
O
O
N
N
N
O
N
NH2
NH2
N
boc
N
boc
N
boc
O
1j
1h
1i
boc
N
N
boc
N
H
boc
N
O
N
H
N
N
boc
boc
1l
1k
The MALDI–TOF mass spectrum obtained after incu-
bation of h-IAPP with A is illustrated in Fig. 4. The
molecular weights corresponding to the peaks shown by the
spectrum correspond to fragments h-IAPP1–13 (1,378),
h-IAPP20–37 (1,845), h-IAPP1–19 (2,076), and h-IAPP14–37
(2,542). Here, the h-IAPP fragments are named according
to the amino acid sequence of h-IAPP. The structures of h-
IAPP20–37, h-IAPP1–19, and h-IAPP14–37 were positively
identified by MALDI–TOF/TOF LIFT MS [18].
Fig. 2 Compounds used as component R2-NRH in construction of
the library; the primary amines of 1h–1j and the secondary amines of
1k and 1l are attached to the triazine ring. boc tert-butyloxycarbonyl
were prepared as the derivatives of Co(III)oxacyclen. To
obtain the derivatives of Co(III)arylcyclen, 1e–1g (Fig. 1)
were employed as component R1-XH and group A and
group C amines (Fig. 3) were employed as component R3-
NRH. In total, 96 compounds were prepared as the deriv-
atives of Co(III)oxacyclen.
After the stock solution of h-IAPP had been added to the
buffer solution containing A, the mixture was incubated for
36 h at 37 °C and pH 7.50. The product solution thus
obtained by the reaction of h-IAPP (4.0 lM) with A was
filtered through a membrane with a molecular weight cut-
off of 10,000 to remove aggregates of h-IAPP. Then, the
resulting solution was subjected to high performance liquid
chromatography (HPLC) separation to isolate the oligo-
peptide fragments formed by the cleavage of h-IAPP and to
ensure the exclusion of nucleated h-IAPP. The oligopeptide
fragments were hydrolyzed under alkaline conditions, and
Compounds of 1b and 1c were purified by column
chromatography. The 612 compounds of 1d were partially
purified by using short columns and the purity of the
Co(III) complexes contained in the library is estimated as
70–90%. The 612 Co(III) complexes thus prepared were
screened by matrix-assisted laser desorption/ionization
(MALDI) time-of-flight (TOF) mass spectrometry (MS) for
their activity to cleave Ab42 (4.0 lM) or h-IAPP (4.0 lM)
at pH 7.50 and 37 °C for 48 h. Some Co(III) complexes of
123

514
J Biol Inorg Chem (2011) 16:511–519
Fig. 3 Compounds used as
component R3-NRH in
construction of the library; the
primary or the secondary
amines are attached to the
triazine ring
(1) group A amines
NH2
NH2
NH2
N
NH2
NH2
H
O
Cl
O
H
N
NH2
NH2
NH2
NH2
NH2
NH2
N
H
H
N
N
NH2
NH2
NH2
NH2
N
H
O
(2) group B amines
H2N
N
H2N
H
OH
HO
H2N
OH
NH
O
H2N
NH2
N
H2N
H2N
F
N
H2N
N
OH
N
H2N
H2N
Cl
H
N
O
N
O
H2N
N
N
NH2
N
N
NH
H2N
N
O
OH
H
OH
OH
N
HO
HO
NH2
HO
OH
H2N
NH2
H2N
H2N
NH2
OH
OH
(3) group C amines
NH2
NH2
NH2
N
NH2
NH2
H
F
F
H
NH2
CF3
NH2
N
NH2
N
NH2
N
O
N
N
H
the resulting amino acids were quantified with fluoresca-
mine as described previously [6]. The amounts of oligo-
peptide fragments obtained by the action of A were
estimated as the mole percentage of the initially added
amount of h-IAPP and were defined as the cleavage yields.
The cleavage yields measured after the reaction with var-
ious initial concentrations (C0) of A are plotted against
log(C0/M) in Fig. 5. Since the isolation of the oligopeptide
fragments inevitably involves some losses, the cleavage
yields are somewhat underestimated.
pH 7.50 and then adding A (final concentration 1.0 lM).
After incubation for a further 36 h, the amount of the
cleavage products was measured. The cleavage yields
were 11.1 1.1 and 1.1 0.6 mol%, respectively, when
h-IAPP was preincubated for 3 and 24 h, respectively, prior
to the reaction with A.
To demonstrate that A does not cleave nontarget proteins,
A (10 lM) was incubated with proteins (4–40 lM) such as
bovine serum albumin, bovine serum c-globulin, chicken
egg white lysozyme, and horse heart myoglobin at pH 7.50 or
9.50 for 3 days at 37 °C. For those nontarget proteins, no
protein cleavage was observed by MALDI–TOF MS.
The cleavage yields were also measured by incubating
the solution of h-IAPP (4.0 lM) for 3 or 24 h at 37 °C and
123

J Biol Inorg Chem (2011) 16:511–519
515
16
14
12
10
8
NH
III
Co
OH2
OH2
NH
NH
N
A
N
N
6
O
N
HN
4
N
2
S
0
N
-9
-8
-7
-6
-5
-4
log(C0/M)
Structure 2
Fig. 5 Cleavage yield against log(C0/M) for the cleavage of h-IAPP
(4.0 lM) with A for 48 h at pH 7.50 and 37 °C. The cleavage yields
reported in this study are the average values resulting from four to six
measurements carried out with different reaction mixtures
300
200
cleavage
products
¥²
.
(R)-(LCo
¥²
)
(AmP)ol-m
(R)-(LCo
)
d
h-IAPP
.............
.............
c
(AmP)ol-2
(AmP)ol-1
AmP
b
100
a
(AmP)ol-z
.............
(AmP)ol-m
fibrils
protofibrils
0
1500
2000
2500
3000
3500
4000
Fig. 6 Formation of various assemblies of amyloidogenic peptide or
protein (AmP) and reduction of the assemblies by using the cleavage
agents
m/z
Fig. 4 Matrix-assisted laser desorption/ionization time-of-flight
mass spectrum of the reaction mixture obtained after incubation of
human islet amyloid polypeptide (h-IAPP) (4.0 lM) with A (Struc-
ture 2; 1.0 lM) for 48 h at pH 7.50 and 37 °C. Peaks a, b, c,
and d corresponds to h-IAPP1–13, h-IAPP20–37, h-IAPP1–19, and
h-IAPP14–37
The mechanism [19] proposed for peptide hydrolysis
catalyzed by Co(III)cyclen is summarized in Fig. 7. Here,
the catalysis involves polarization of the carbonyl group
of the amide bond by Co(III) ion and the nucleophilic
attack by the Co(III)-bound hydroxide ion at the carbonyl
carbon atom leading to formation of a tetrahedral inter-
mediate containing a four-membered ring. Similar mech-
anisms may be operative for peptide hydrolysis by
Co(III)arylcyclen, Co(III)oxacyclen, and Co(III)aryloxa-
cyclen. The catalytic rate would depend on the stability of
the rate-determining transition state involving the four-
membered ring. The Lewis acidity of Co(III) ion and the
geometry of the complex formed between Co(III) ion and
the chelating ligand would affect the stability of the
transition state.
Discussion
Reduction of the level of pathogenic oligomers of AmP
through peptide cleavage may cure the corresponding dis-
ease [4–7]. The peptide cleavage is schematically sum-
marized in Fig. 6, where the cleavage agent is indicated as
(R)–(LCoIII) and various soluble oligomers are shown in
the rectangle [5–7]. As illustrated in Fig. 6, cleavage of the
AmP included in a target oligomer reduces the concentra-
tion of the target oligomer, leading to decreases in the
concentrations of other oligomers which readily transform
into the target oligomer. Although the nature of the path-
ogenic oligomers of AmP is not known, cleavage of the
target oligomers would be accompanied by a decrease in
the concentration of the pathogenic oligomer.
The pK values for the Co(III)-bound water molecules
revealed the higher Lewis acidity of the Co(III) center
(pKa1 = 3.33, pKa2 = 6.96) in Co(III)oxacyclen compared
with that (pKa1 = 5.66, pKa2 = 8.14) in Co(III)cyclen
[17]. This is because the oxygen of oxacylen is much less
123

516
J Biol Inorg Chem (2011) 16:511–519
Fig. 7 The mechanism
proposed for the hydrolysis of
peptide bonds catalyzed by the
Co(III) of 1,4,7,10-
tetraazacyclododecane
[Co(III)cyclen]
O
R'
R
N
H
NH
III
NH
III
NH
III
R
OH2
NH
NH
OH2
OH2
NH
NH
O
NH
NH
Co
Co
-
Co
N H
O
-
O
H
NH
R'
H
NH
NH
COO-
NH3+
NH
III
R
NH
III
-
O
R
+ H
N H
OH2
NH
NH
NH
NH
Co
+
-
Co
R'
O
O
-
H
NH
R'
NH
basic than the nitrogen of cyclen. Amines become con-
siderably less basic when an aromatic ring is attached to the
nitrogen atom as in arylcyclen or aryloxacyclen [20]. The
CNH bond angle of an alkyl amine is nearly tetrahedral
[21] but becomes nearly planar when the alkyl group is
replaced by an aryl amine [22]. Variations in the basicity
and the geometry of the chelating ligands of the Co(III)
center may affect the catalytic efficiency of peptide-
cleaving catalysts toward the pathogenic proteins of
amyloidoses. This is supported by the proteolytic activity
of Co(III)oxacyclen, which is 4–14 times higher than that
of Co(III)cyclen toward albumin, c-globulin, and myoglo-
bin [17].
value is comparable to the values (10–22 mol%) for the
cleavage of soluble oligomers of h-IAPP by the eight
derivatives of Co(III)cyclen reported previously [6]. As
summarized in Fig. 5, A shows some activity even at
concentrations of 10–100 nM as observed with the eight
derivatives of Co(III)cyclen [6].
The catalytic efficiency of Co(III)arylcyclen derivative
A is, therefore, comparable to that of the derivatives of
Co(III)cyclen reported previously [6]. The cyclen moiety
is directly attached to the triazine ring in A, whereas
various linkers are inserted between the cyclen moiety and
the triazine ring in the eight derivatives of Co(III)cyclen.
Consequently, the molecular weight of A is considerably
smaller than the molecular weight of the corresponding
derivatives of Co(III)cyclen. The reduction in molecu-
lar weight is a significant advantage in application as
drugs.
Our previous study indicated that h-IAPP exists as the
monomer and small oligomers immediately after transfer to
the reaction buffer (pH 7.50) from the pH 10.5 medium [6].
After incubation for 3 h, about half of the initially added
h-IAPP becomes aggregates that cannot pass through a
10 kDa molecular mass cutoff filter. A small portion of
h-IAPP, however, remains as the monomer and small
oligomers even after incubation for 36 h.
In the previous studies, 888 derivatives of Co(III)cyclen
were synthesized, among which eight compounds were
discovered to cleave soluble oligomers of Ab42, h-IAPP,
and/or a-synuclein [5–7]. In this study, 612 derivatives of
Co(III)oxacyclen, Co(III)arylcyclen, or Co(III)aryloxacy-
clen were obtained. Among the 612 compounds, only
A has activity to cleave h-IAPP. In addition, A was not able
to cleave Ab42. The chemical libraries constructed for
Co(III)oxacyclen, Co(III)arylcyclen, or Co(III)aryloxacy-
clen as well as the chemical library constructed for
Co(III)cyclen in a previous study [6] are based on triazine
derivatives containing 1e, 1f, or 1g. In the present study we
could not discover catalytically active derivatives of
Co(III)oxacyclen or Co(III)aryloxacyclen in cleavage of
Ab42 or h-IAPP and derivatives of Co(III)arylcyclen for
cleavage of Ab42. It is still possible, however, that some
active catalysts based on Co(III)oxacyclen, Co(III)arylcy-
clen, or Co(III)aryloxacyclen may be obtained by using
various other types of organic pendants. It is also possible
that the new catalytic centers will have improved catalytic
efficiency in comparison with Co(III)cyclen.
The cleavage yields measured in this study were
12.1 2.1, 11.1 1.1, and 1.1 0.6 mol% when h-IAPP
(4.0 lM) was incubated for 0, 3, and 24 h, respectively,
before addition of A (1.0 lM). This indicates that the
soluble oligomers of h-IAPP are the major species cleaved
by A. The cleavage yield would decrease considerably
upon preincubation of h-IAPP for 3 h if the monomer of
h-IAPP were the major species undergoing cleavage,
whereas the cleavage yield would increase considerably
upon preincubation of h-IAPP for 24 h if the polymer of
h-IAPP were the major species.
Presently, the identities of the oligomers of h-IAPP
cleaved by A are not known. The MALDI–TOF mass
spectrum in Fig. 4, however, revealed that Ala13-Asn14
and Ser19-Ser20 are included in the cleavage sites. The
plateau value of the cleavage yield is 12 mol% under the
experimental conditions as revealed by Fig. 5. This plateau
123

J Biol Inorg Chem (2011) 16:511–519
Materials and methods
Materials
517
cleavage agents, which was used without further purifica-
tion. To the MeOH (500 lL) solution of the TFA salt was
added LiOH (approximately 4.5 equiv) followed by
CoCl2Á6H2O (approximately 2 equiv), and the resulting
solution was shaken overnight under air [23]. Oxidation of
Co(II) to Co(III) was accompanied by the appearance of
deep-violet color. The solvent was evaporated completely
with a centrifugal vacuum concentrator (Hanil model
3180C) and water (500 lL) was added. Formation of the
Co(III) complexes included in the library was confirmed by
MALDI–TOF MS. The aqueous solution of the resulting
Co(III) complex was kept at room temperature for several
days or weeks before its ability to cleave the substrate was
tested.
Construction of the library
The chemical library of candidates for the peptide-cleaving
catalysts was built according to the route in Scheme 1.
Cyanuric chloride (1a) (0.20 g 1.1 mmol), R1-XH (0.20 g
0.90 mmol), and diisopropylethylamine (DIEA) (0.38 mL,
2.7 mmol) were mixed together in tetrahydrofuran (50 mL)
and the mixture was stirred for 4–8 h in an ice bath. The
residue obtained by evaporation of the mixture was purified
by column chromatography to obtain 1b. Here, R1-XH was
an amine or an alcohol. To an acetonitrile solution (25 mL)
of 1b (0.85 mmol) was added R2-NRH (0.85 mmol),
followed by DIEA (240 lL, 1.74 mmol). The mixture was
stirred at room temperature overnight. The residue obtained
by evaporation of the mixture was purified by column
chromatography to obtain 1c. To the suspension of 1c in
acetonitrile (25 mL) was added amine R3-NRH (7.4 lmol).
The mixture in the reaction tube was stirred at 80 °C for
8 h. The solvent was evaporated off, and short column
chromatography afforded 1d. Crude 1d (5 mg) was treated
with 50% trifluoroacetic acid (TFA) in methylene chloride
(50 lL) for 5 h and diethyl ether (1 mL) was added to the
TFA solution to remove the tert-butyloxycarbonyl pro-
tecting groups attached to the oxacyclen or cyclen moiety.
The precipitate was separated by centrifugation, washed
with diethyl ether several times, and dried under nitrogen
gas to obtain the TFA salt of the ligand of the candidate for
Synthesis of compound A
Cleavage agent A was synthesized according to the route
described in Scheme 2.
(4-Benzothiazol-2-yl-phenyl)-[2-(4,6-dichloro-[1,3,5]
triazin-2-yloxy)-ethyl]-methylamine (2a) [5] (110 mg,
0.30 mmol), 1,4,7,10-tetraazacyclododecane-1,4,7-tricar-
boxylic acid tri-tert-butyl ester (2b) [13] (185 mg,
0.35 mmol), and DIEA (0.10 mL, 0.74 mmol) were mixed
together in acetonitrile (50 mL) and the mixture was stirred
for 6 h in room temperature. The residue obtained by
evaporation of the mixture was purified by column chro-
matography to obtain 10-(4-{2-[(4-benzothiazol-2-yl-phe-
nyl)-methyl-amino]-ethoxy}-6-chloro-[1,3,5]triazin-2-yl)-1,
4,7,10-tetraazacyclododecane-1,4,7-tricarboxylic acid tri-
tert-butyl ester (2c). Rf 0.3 (ethyl acetate/hexane 1:2);
boc
Scheme 2 Synthetic route for
A. boc tert-butyloxycarbonyl,
TFA trifluoroacetic acid
boc
boc
N
N
N
N
N
N
boc
boc
N
Cl
HN
N
2b
boc
N
N
N
Cl
O
N
DIEA
S
O
N
Cl
N
2a
2c
N
S
N
N
boc
N
N
H
N
boc
boc
N
HN
N
NH
N
N
N
N
N
CoCl2, O2
O
N
HN
TFA
butylamine
DIEA
O
N
HN
A
N
N
2e
S
2d
S
N
N
123

518
J Biol Inorg Chem (2011) 16:511–519
1H NMR (CDCl3): d 7.96 (q, 3H), 7.83 (d, 1H), 7.42
(t, 1H), 7.29 (t, 1H), 6.77 (d, 2H), 4.56 (t, 2H), 3.82 (t, 2H),
3.66 (br, 4H), 3.43 (br, 12H), 3.11 (br, 3H), 1.43 (br, 27H);
13C NMR (CDCl3): d 170.60, 170.10, 169.36, 168.48,
166.53, 165.88, 157.48, 156.17, 154.33, 151.51, 150.62,
134.51, 129.03, 128.92, 126.01, 124.23, 122.24, 121.85,
121.36, 111.76, 111.64, 80.49, 80.35, 80.18, 80.15, 64.69,
60.37, 59.87, 54.59, 51.51, 50.89, 50.47, 49.92, 49.79,
49.42, 39.31, 29.68, 28.91, 28.46, 28.37; MS (MALDI–
TOF) m/z 867.42 (M?H)?, calcd for C42H58ClN9O7S
867.39. Compound 2c (173 mg, 0.20 mmol), n-butylamine
(36.57 mg, 0.50 mmol), and DIEA (0.10 mL, 0.74 mmol)
were mixed together in acetonitrile (50 mL) and the mix-
ture was stirred for 6 h at 80 °C. The residue obtained
by evaporation of the mixture was purified by column
chromatography to obtain 10-(4-{2-[(4-benzothiazol-2-yl-
phenyl)-methyl-amino]-ethoxy}-6-butylamino-[1,3,5]triazin-
2-yl)-1,4,7,10-tetraazacyclododecane-1,4,7-tricarboxylic
acid tri-tert-butyl ester (2d). Rf 0.2 (ethyl acetate/hexane
appearance of a deep-violet color. The resulting Co(III)
complex was purified by HPLC (Hewlett-Packard 1050
series) by using a gradient of two solvent systems con-
sisting of 0.1% TFA in water and 0.1% TFA in HPLC-
grade acetonitrile, a 10-lm C18 column (300 mm 9 3.9
mm, purchased from Phenomenex), and a UV (545 nm)
detector. The portion containing Co(III) complexes was
evaporated, and kept at room temperature for several days
to obtain the stock solution of A. The concentration of the
stock solution was determined through quantification of
cobalt by inductively coupled plasma atomic emission
spectroscopy (Shimadzu ICPS-1000IV) or inductively
coupled plasma MS (PerkinElmer ELAN 6100).
Other materials
Recombinant Ab42 was purchased from r-Peptide and
synthetic h-IAPP was purchased from Ana-Spec. Bovine
serum albumin, bovine serum c-globulin, chicken egg
white lysozyme, and horse heart myoglobin were obtained
from Sigma.
1
1:2); H NMR (CDCl3): d 7.96 (t, 3H), 7.83 (d, 1H), 7.43
(t, 1H), 7.30 (t,1H), 6.77 (d, 2H), 4.47 (br, 2H), 3.77 (br,
2H), 3.65 (br, 4H), 3.43 (br, 13H), 3.11 (br, 3H), 1.45 (br,
29H), 1.26 (br, 2H), 0.90 (t, 3H); 13C NMR (CDCl3): d
169.67, 168.59, 166.43, 156.28, 154.39, 150.86, 134.52,
128.98, 125.98, 124.21, 122.28, 121.35, 121.34, 111.55,
80.16, 79.81, 63.06, 51.23, 50.28, 40.51, 39.26, 31.87,
29.69, 28.49, 28.42, 19.99, 13.86; MS (MALDI–TOF) m/
z 904.92 (M?H)?, calcd for C46H68N10O7S 904.50.
Compound 2d (5 mg) was treated with 50% TFA in
methylene chloride (50 lL) for 5 h and diethyl ether
(1 mL) was added to the TFA solution. The precipitate was
separated by centrifugation, washed with diethyl ether
several times, and dried under nitrogen gas to obtain the TFA
salt of [4-{2-[(4-benzothiazol-2-yl-phenyl)-methylamino]
ethoxy}-6-(1,4,7,10-tetraazacyclododec-1-yl)-[1,3,5]tria-
zin-2-yl]-butylamine (2e). The TFA salt of 2e was used for
NMR and MS characterization; 1H NMR (CD3OD): d 7.94
(t, 3 H), 7.82 (d, 1 H), 7.42 (t, 1H), 7.28 (t, 1H), 6.76
(d, 2H), 4.58–4.40 (br, 2H), 3.82–3.67 (br, 6H), 3.40–3.30
(br, 2H), 3.18–3.05 (br, 3H), 3.00–2.82 (br, 4H), 2.80–2.70
(br, 4H), 2.68–2.52 (br, 4H), 1.52 (q, 2H), 1.34 (q, 2H),
0.91 (t, 3H); 13C NMR (CDCl3): d 169.71, 168.61, 168.55,
166.51, 156.38, 154.34, 150.84, 134.44, 128.92, 125.98,
124.21, 122.20, 121.43, 121.33, 111.51, 80.18, 79.79,
63.12, 51.13, 50.08, 47.95, 40.41, 39.25, 31.75, 29.66,
20.03, 13.82; High-resolution MS m/z 605.3508 (M?H)?,
calcd for C31H45N10OS 605.3499. The overall yield for
synthesis of 2e from starting materials 2a and 2b was 78%.
To the MeOH (500 lL) solution of the TFA salt of 2e
(3.0 mg, 0.0050 mmol) was added LiOH (1.0 mg, 0.025
mmol), followed by CoCl2Á6H2O (1.2 mg, 0.0050 mmol),
and the solution was shaken overnight under air [23].
Oxidation of Co(II) to Co(III) was accompanied by the
Measurement
Preparation of the alkaline stock solutions of Ab42 and h-
IAPP and measurement of cleavage yields were carried out
as described previously [5, 6]. MALDI–TOF MS mea-
surements were carried out with a Bruker Daltonics
Autoflex II MALDI–TOF/TOF mass spectrometer.
Acknowledgments This work was supported by the National
Research Foundation of Korea (NRF) (grant no. 2010-0016023)
funded by the Korean government (MEST).
References
1. Suh J (2003) Acc Chem Res 36:562–570
2. Chei WS, Suh J (2007) Prog Inorg Chem 55:79–142
3. Suh J, Chei WS (2008) Curr Opin Chem Biol 12:207–213
4. Lee TY, Suh J (2009) Chem Soc Rev 38:1949–1957
5. Suh J, Yoo SH, Kim MG, Jeong K, Ahn JY, Kim M, Chae PS,
Lee TY, Lee J, Lee J, Jang YA, Ko EH (2007) Angew Chem Int
Ed 46:7064–7067
6. Suh J, Chei WS, Lee TY, Kim MG, Yoo SH, Jeong K, Ahn JY
(2008) J Biol Inorg Chem 13:693–701
7. Lee J, Yoo SH, Jeong K, Lee TY, Ahn JY, Suh J (2008) Bull
Korean Chem Soc 29:882–884
8. Bitan G, Fradinger E, Spring S, Teplow D (2005) Amyloid
12:88–95
´
9. Lesne S, Koh MT, Kotilinek L, Kayed R, Glabe CG, Yang A,
Gallagher M, Ashe KH (2006) Nature 440:352–357
10. Lin CY, Gurlo T, Kayed R, Butler AE, Haataja L, Glabe CG,
Butler PC (2007) Diabetes 56:1324–1332
11. Jeon JW, Son SJ, Yoo CE, Hong IS, Song JB, Suh J (2002) Org
Lett 4:4155–4158
12. Jeon JW, Son SJ, Yoo CE, Hong IS, Suh J (2003) Bioorg Med
Chem 11:2901–2910
123

J Biol Inorg Chem (2011) 16:511–519
519
13. Chae PS, Kim M, Jeung CS, Lee SD, Park H, Lee S, Suh J (2005)
J Am Chem Soc 127:2396–2397
19. Takasaki MK, Kim JH, Rubin E, Chin J (1993) J Am Chem Soc
115:1157–1159
14. Greenwood NN, Earnshaw A (1997) Chemistry of the elements,
2nd edn. Butterworth-Heinemann, Oxford
15. Izatt RM, Pawlak K, Bradshaw JS (1991) Chem Rev
91:1721–2085
20. Anslyn EV, Dougherty DA (2004) Modern physical organic
chemistry. University Science Books, Sausalito, pp 281–283
21. Lee SJ, Mhin BJ, Cho SJ, Lee JY, Kim KS (1994) J Phys Chem
98:1129–1134
16. Chei WS, Lee J-W, Kim JB, Suh J (2010) Bioorg Med Chem
18:5248–5253
17. Kim HM, Jang B, Cheon YE, Suh MP, Suh J (2009) J Biol Inorg
Chem 14:151–157
22. Nguyen MT (2007) In: Rapporport Z (ed) The chemistry of
anilines, part 1. Wiley, Chichester, pp 75–166
23. Castillo-Blum SE, Sosa-Torres ME (1995) Polyhedron
14:223–229
18. Suckau D, Resemann A, Schuerenberg M, Hufnagel P, Franzen J,
Holle A (2003) Anal Bioanal Chem 376:952–965
123