Sequence-specific oxidative degradation of tripeptides by a cobalt(III) complex containing a terpyridine ligand

Article abstract of DOI:10.1002/ejic.200400261

A sequence-specific oxidative degradation of an oligopeptide consisting of aliphatic amino acids was performed with a ternary cobalt(III) complex. Under aerobic conditions the reaction of [Co(CO₃)(OH)(terpy)] and an equimolar amount of tripeptide at 40 °C at pH 8.5 gave two fraction bands (a and b) with water and then two bands (c and d) with NaCl solution, respectively, after column separation. In the case of gly-gly-leu (ggl, 1), [Co(gg'l)(terpy)] [1a, gg'l = glycyl(2-oxoglycyl)leucine], [Co(g'l)(terpy)] [1b, g'l = N-(2-oxoglycyl)leucine], [Co(ggl)(terpy)]⁺ (1c), and [Co(terpy)₂]³⁺ (1d) were obtained. The crystal structure of 1c, which was the main product, revealed that the starting ggl ligand was cleaved to give a bis-amide compound, g'l, which might be generated by loss of N-terminal glycine and carbonylation of the α-carbon of the N-2 glycine residue. Although the source of the oxygen of the newly generated carbonyl group is not clear, gg'l is supposed to be an intermediate for g'l on the basis of time-dependent HPLC analyses. NMR investigation of other tripeptides [βala-gly-leu (2), gly-gly-phe (3), and gly-gly-ala (4)] containing a bulky C-terminal aliphatic side chain indicated the formation of the corresponding oxidation complexes. On the other hand, peptides containing a C-terminal glycine residue [gly-gly-gly (5) and phe-gly-gly (6)] or an N-2 β-alanine residue [gly-βala-leu (7)] did not give such a transformation product, The crystal structure of [Co(fgg)-(terpy)] (6c) revealed that the tripeptide is coordinated to cobalt with a tridentate N-terminal N-N-N geometry. Such sequence specificity demonstrated in the reaction of the aliphatic tripeptide complexes is interpreted by the interligand interaction between the side chain of the C-terminal amino acid residue and terpy. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2004.

Full text of DOI:10.1002/ejic.200400261

FULL PAPER
Sequence-Specific Oxidative Degradation of Tripeptides by a Cobalt(III
Complex Containing a Terpyridine Ligand
)
Koichiro Jitsukawa,*[a] Hiroshi Takahashi,[a] Ryoji Hyuga,[a] Hidekazu Arii,[a] and
Hideki Masuda*[a]
Keywords: Cleavage reactions / Cobalt / N ligands / Oxidation / Peptides
A sequence-specific oxidative degradation of an oligopep-
tide consisting of aliphatic amino acids was performed with
a ternary cobalt(III) complex. Under aerobic conditions the
reaction of [Co(CO3)(OH)(terpy)] and an equimolar amount
of tripeptide at 40 °C at pH 8.5 gave two fraction bands (a
and b) with water and then two bands (c and d) with NaCl
solution, respectively, after column separation. In the case of
time-dependent HPLC analyses. NMR investigation of other
tripeptides [βala-gly-leu (2), gly-gly-phe (3), and gly-gly-ala
(4)] containing a bulky C-terminal aliphatic side chain indi-
cated the formation of the corresponding oxidation comple-
xes. On the other hand, peptides containing a C-terminal
glycine residue [gly-gly-gly (5) and phe-gly-gly (6)] or an N-
2 β-alanine residue [gly-βala-leu (7)] did not give such a
gly-gly-leu (ggl, 1), [Co(ggЈl)(terpy)] [1a, ggЈl = glycyl(2-oxo- transformation product. The crystal structure of [Co(fgg)-
glycyl)leucine], [Co(gЈl)(terpy)] [1b, gЈl = N-(2-oxoglycyl)leu- (terpy)] (6c) revealed that the tripeptide is coordinated to co-
cine], [Co(ggl)(terpy)]+ (1c), and [Co(terpy)2]3+ (1d) were ob-
tained. The crystal structure of 1c, which was the main pro-
duct, revealed that the starting ggl ligand was cleaved to
give a bis-amide compound, gЈl, which might be generated
balt with a tridentate N-terminal N-N-N geometry. Such se-
quence specificity demonstrated in the reaction of the ali-
phatic tripeptide complexes is interpreted by the interligand
interaction between the side chain of the C-terminal amino
by loss of N-terminal glycine and carbonylation of the α-car- acid residue and terpy.
bon of the N-2 glycine residue. Although the source of the
oxygen of the newly generated carbonyl group is not clear,
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
ggЈl is supposed to be an intermediate for gЈl on the basis of Germany, 2004)
Introduction
peptide (DP) mediated by a ternary cobalt complex with
the 2,2Ј:6Ј,2ЈЈ-terpyridine (terpy) ligand under aerobic con-
ditions, which is a model reaction for the peptidylglycine α-
hydroxylating monooxygenase (PHM) reaction.[7] The terpy
ligand is known to accelerate the oxidative transformation
of another ligand coordinated trans to it,[8] and a ternary
complex with the terpy ligand could therefore cause the oxi-
dative degradation of a peptide compound. Herein, we de-
scribe a sequence-specific oxidative degradation of tripep-
tides (TPs), such as GGL (1), BGL (2), and GGF (3),
whose reactivities are compared with the cases of GGA (4),
In oxidative transformation of proteins, which plays a
major role in many oxidation processes within cells, the rad-
icals generated damage both the peptide backbone and the
side chains.[1] In CϪN bond dissociation during the peptide
degradation, active oxygen species, such as HO·, and HOO·,
attack at the α-position of the amino-acid residues to gener-
ate a carbon-centered radical.[2,3] Once formed, this reacts
further with molecular oxygen to form peroxy radicals,
which expel superoxide species to give carbocations, and are
then hydrolyzed to give the corresponding alcohols or are
deprotonated to give the corresponding imines.[4,5] Some
iron edta complexes attached to cysteine residue of proteins
site-specifically cleaved the peptide backbone, in which the
iron-hydroperoxo species directly attacked the amide car-
bonyl moiety.[6] Recently, we have reported selective α-hy-
droxylation of the C-terminal amino-acid residue of a di-
[a]
Nagoya Institute of Technology,
Gokiso-cho, Showa-ku Nagoya 466-8555, Japan
Fax: ϩ81-52-7355240
E-mail: jitsukawa.koichiro@nitech.ac.jp
jitkk@ach.nitech.ac.jp
Scheme 1. Oxidative cleavage of tripeptides
4140
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/ejic.200400261
Eur. J. Inorg. Chem. 2004, 4140Ϫ4145

Cobalt(III)-Mediated Oxidative Degradation of Tripeptides
FULL PAPER
GGG (5), FGG (6), and GBL (7) (Scheme 1 and 2). Upon
Fortunately, a single crystal of 1b suitable for X-ray
the formation of the ternary cobalt() complexes, some tri- analysis was obtained from the column eluent by allowing
peptides containing bulky aliphatic amino acids at the C- it to stand at room temperature. The crystal structure of 1b
terminus are cleaved to give unprecedented dipeptide de- (Figure 1)[11] reveals an unprecedented formation of the bis-
rivatives on account of the assistance of the terpy ligand.
amide compound N-(2-oxoglycyl)leucine (GЈL), as expected
from the NMR spectroscopic data. The crystal data are
summarized in the Exp. Sect. In order to determine the
species substituted at the α-position of the N-2 residue in
GЈL, a positive-ion ESI mass spectrum of 1b was measured.
The obtained prominent peaks at m/z ϭ 492.2 and 514.2
corresponding to the [{Co(gЈl)(terpy)}
ϩ
H]ϩ and
[{Co(gЈl)(terpy)} ϩ Na]ϩ ions, respectively, indicate that the
newly substituted atom is oxygen, although the oxygen
source is not yet clear. Formation of a [Co(gЈl)(terpy)] com-
plex is also supported by the elemental analysis. Moreover,
˚
the relatively short bond length [1.22(2) A] between the α-
carbon [C(16)] and the newly generated atom [O(1)] and the
sp2 character around C(16) [O(1)ϪC(16)ϪN(4) ϭ 128(1)°
and O(1)ϪC(16)ϪC(17) ϭ 119(1)°] in the crystal structure
of 1b indicate the formation of a CϭO group at the α-posi-
tion of the N-2 residue. The cleavage of the amide bond
between the N-1 and N-2 residues in the GGL ligand took
place in 1b, but the α-protons of N-1 and N-3 moieties still
remain in 1a and 1c.
In the case of GGA (4), complete column separation of
4a and 4b was not accomplished, although their 1H NMR
spectra indicated formation of the corresponding cleavage
product. Due to the smaller size of the C-terminal side
chain in 4 in comparison with the those in 1Ϫ3, it is hard
to separate complexes 4a and 4b. In the cases of some
tripeptide ligands, for example GGG (5), FGG (6), and
GBL (7), which have either an N-terminal glycine or N-2
β-alanine moiety, the main product was the
[Co(tp)(terpy)]ϩ complex (5c, 6c, or 7c), accompanied by
a small amount of [Co(terpy)2]3ϩ (d).[9] Transformation
Scheme 2. Tripeptides employed: gly-gly-leu (GGL; 1), βala-gly-leu
(BGL; 2), gly-gly-phe (GGF; 3), gly-gly-ala (GGA; 4), gly-gly-gly
(GGG; 5), phe-gly-gly (FGG; 6), gly-βala-leu (GBL; 7)
Results and Discussion
Characterization of the Ternary Complexes
The reaction of the cobalt carbonyl complex and an equi- compounds corresponding to a or b were not obtained.
1
molar amount of tripeptides 1Ϫ7 was carried out at pH 8.5 The H NMR spectral patterns of the N1-, N2- and N3-
and 40 °C in an aqueous solution for 10 h under aerobic α-proton moieties in 5c, 6c, and 7c are very similar to that
conditions. After passing the reaction mixture through a of 1c. On the basis of the spectral analogy among the
cation-exchange column, four fractional bands containing complexes 1cϪ7c, the coordination mode of the tripeptide
cobalt() complexes (aϪd) were obtained (see Exp. Sect.). to cobalt is an N-terminal N-N-N form (N-1 terminal am-
In order to clarify the structure of the cobalt() complexes ine nitrogen, N-2 amide nitrogen, N-3 amide nitrogen),
isolated from the column separation a H NMR spectro- which agrees well with the empirical rule.[10,11] The triden-
1
scopic investigation was carried out. Transformation of the tate coordination of FGG in 6c is indicated by the upfield
TP ligand was observed in the ternary complexes prepared shift of the α-protons of the C-terminal glycine moiety
from 1Ϫ3 (Table 1). Complex 1a shows α-proton peaks of (δ ϭ 2.52 and 2.62 ppm) in comparison with the tetraden-
N-1 (N-terminal) at δ ϭ 2.60 and 2.90 ppm, and that of the tate coordination of FGG in [Co(NH3)2(fgg)], where the
C-terminal (N-3) leucine at δ ϭ 5.07 ppm, but no peaks corresponding protons are observed at δ ϭ 4.00 ppm.[12]
corresponding to those of the N-2 (central) residue. Neither A complex-formation shift was not observed in the C-ter-
the α-proton peaks corresponding to N-1 nor N-2 glycine minal α-position of 1cϪ7c. The single-crystal structure of
residues were detected in the spectrum of complex 1b, while 6c, shown in Figure 2, reveals that the cobalt ion is coor-
that of N-3 was detected at δ ϭ 5.07 ppm. In complex 1c, dinated by two tridentate ligands, terpy and FGG, in
all α-proton peaks of GGL were detected. The spectro- which the coordination mode of the FGG to cobalt is an
scopic and column behavior of the cobalt complexes pre- N-terminal N-N-N geometry, as expected from the NMR
pared from BGL (2) and GGF (3) was similar to that of spectroscopic data. Moreover, the benzene ring of the N-
GGL (1), indicating that transformation of the glycine moi- terminal amino-acid side chain is very close to the terpy
˚
eties takes place in the peptide backbone of 1a, 1b, 2a, 2b, ring, with a distance of about 3.45 A (the closest in-
3a, and 3b.
teratomic carbon-carbon distance).
Eur. J. Inorg. Chem. 2004, 4140Ϫ4145
www.eurjic.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4141

K. Jitsukawa, H. Takahashi, R. Hyuga, H. Arii, H. Masuda
FULL PAPER
Table 1. NMR spectroscopic data for the ternary complexes
Tripeptide
Column
Band
N-terminal
β-H
N-2 (center)
C-terminal
β-H
α-H
α-H
β-H
α-H
ggl
(1)
a
b
2.60 (d)
2.90 (d)
not detected
Ϫ
not detected
not detected
5.07 (q)
5.07 (q)
2.12 (m)
2.38 (m)
2.10 (m)
2.40 (m)
1.34 (m)
2.08 (m)
c
a
3.42 (s)
2.30 (m)
Ϫ
5.28 (s)
2.13 (d)
Ϫ
Ϫ
4.00 (q)
4.95 (m)
bgl
(2)
1.75 (m)
1.85 (m)
b
not detected
not detected
5.01 (q)
2.00 (m)
2.51 (m)
c
hard to identify
2.31 (d)
2.97 (d)
ggf
(3)
a
Ϫ
Ϫ
not detected
not detected
5.17 (m)
5.22 (m)
2.31 (dd)
3.63 (dd)
4.43 (dd)
3.67 (dd)
4.45 (dd)
2.59 (dd)
2.95 (dd)
b
c
not detected
3.30 (dd)
3.38 (dd)
4.57 (dd)
4.69 (dd)
gga
(4)
a ϩ b
mixture of complexes (hard to identify)
c
3.50 (dd)
Ϫ
4.75 (d)
4.81 (d)
Ϫ
2.50 (b)
2.85 (s)
0.41 (d)
ggg
(5)
a
b
c
a
b
c
not isolated
not isolated
3.49 (s)
Ϫ
4.68 (s)
Ϫ
Ϫ
Ϫ
fgg
(6)
not isolated
not isolated
3.84 (t)
2.78 (dd)
3.39 (dd)
4.52 (d)
4.72 (d)
2.52 (d)
2.62 (d)
gbl
(7)
a
b
c
not isolated
not isolated
3.23 (d)
Ϫ
2.65 (t), 4.16 (t)
3.53 (m)
1.17 (m)
Figure 1. Molecular structure of the transformation compound
Figure 2. Molecular structure of the tripeptide compound,
[Co(gЈl)(terpy)] (1b); hydrogen atoms are omitted for clarity; selec-
˚
[Co(fgg)(terpy)] (6c); hydrogen atoms are omitted for clarity; selec-
ted bond lengths (A) and angles (°): CoϪN(1) 1.87(1), CoϪN(2)
˚
ted bond lengths (A) and angles (°): CoϪN(1) 1.850(8), CoϪN(2)
1.99(1), CoϪN(3) 1.93(1), CoϪO(3) 1.92(1), CoϪN(4) 1.91(1),
CoϪN(5) 1.88(2); O(3)ϪCoϪN(4) 168.7(5), O(3)ϪCoϪN(5)
84.0(6), N(1)ϪCoϪN(5) 178.6(6), N(2)ϪCoϪN(3) 162.0(7)
1.974(8), CoϪN(3) 1.934(9), CoϪN(4) 1.961(9), CoϪN(5)
1.887(8),
N(4)ϪCoϪN(5)
N(2)ϪCoϪN(3) 164.9(3)
CoϪN(6)
1.942(9);
N(4)ϪCoϪN(6)
N(1)ϪCoϪN(5)
166.5(3),
180.0(4),
83.5(4),
Oxidative Degradation Pathway of the Peptide Bond
In this reaction, 1c is considered as the starting com- product distributions found by HPLC were observed in the
pound for 1a and 1b, because the yield of 1c decreases after reaction with BGL (2) and GGF (3). The reaction pathway
a long reaction time. HPLC analysis of the reaction mixture from 1c to 1b is obviously different from the cobalt()-
also revealed that it is the primary product of the complex promoted hydrolysis, which generally gives both amine and
formation with Co-terpy and GGL. Similar time-dependent carboxylic compounds.[11,13,14] The GЈL ligand in 1b is gen-
4142
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2004, 4140Ϫ4145

Cobalt(III)-Mediated Oxidative Degradation of Tripeptides
FULL PAPER
erated by the loss of N-terminal glycine and oxygenation of gen to give a peroxy radical, which decomposes by the
the α-position of the N-2 glycine residue of the starting HaberϪWeiss mechanism to give the α-carbonyl compound
GGL. The structure of complex 1a, a plausible intermediate 1a because the transition metal-terpy system acts as a redox
for 1b, is suggested by the positive ion ESI mass spectrum: center for the electron-transfer reaction (Scheme 3).[8] Such
The prominent peaks at m/z ϭ 549.2 and 571.0 indicate an oxidative transformation of a methylene moiety adjacent
the formation of a tripeptide complex containing glycyl(2- to an amide group has also been reported in the case of
oxoglycyl)leucine (GGЈL), which is consistent with the iron() and cobalt() complexes with an N-(picolyl)picolin-
NMR spectroscopic data (no signal of N-2 protons in amide ligand under aerobic alkaline conditions, although
Table 1). Such an unprecedented cleavage reaction takes the reaction mechanism is also unclear.[16] In aqueous con-
place only in tripeptides (1Ϫ4) with an alkyl side-chain at ditions the amide complex 1a might be hydrolyzed to give
the C-terminal amino-acid residue, and not in the C-ter- 1b. Under anaerobic conditions, however, the transform-
minal glycine (5 and 6) and N-2 β-alanyl compounds (7). ation product 1b was not detected in the HPLC analysis.
These results indicate that the oxidative degradation of the
tripeptide ligand mediated by the ternary cobalt() terpy features in electron-transfer, photochemical, and related re-
complex occurs sequence-specifically.
actions[8] and activate the trans-positioned ligand through
The terpy ligand is known to demonstrate very attractive
Although the detailed mechanism for this cleavage reac- the central metal,[17,18] which is attributed to the vacant
tion is not yet clear, the downfield shift of the α-protons of low-lying π* orbitals that attract electron density from the
the N-2 glycine moiety in 1c relative to the metal-free GGL metal.[19] Moreover, the assistance of terpy for hydrolysis of
(δ ϭ 4.01 and 4.03 ppm), might be essential for the reaction. diribonucleoside monophosphate diesters has been demon-
Since the GGL ligand in 1c strongly coordinates to the me- strated with a manganese() complex.[20] The interligand in-
tal center in an N-terminal N-N-N manner, a nucleophile teraction caused by terpy is also known to regulate the α-
in aqueous solution (e. g. hydroxide ion) prefers to attack hydroxylation of dipeptides.[7] From these results, we con-
an electro-deficient state at the N-2 moiety. The α-proton sider that the terpy ligand in this ternary cobalt() complex
of the glycine moiety is known to be easily abstracted to plays an important role in the oxidative transformation of
give a glycyl radical, whose stability is explained by the cap- the tripeptide. For sequence specificity, the interligand in-
todative effect.[15] In the oxidative transformation of the di- teraction between the C-terminal side chain and terpy in 1c
peptide ligand previously reported, the downfield shift of is important to abstract the α-proton or to stabilize the rad-
the α-proton attached to the N-2 residue of the dipeptide is ical intermediate that gives 1a under aerobic conditions. In
important for the abstraction of the proton.[7] The radical the ternary complex 1c, the CH-π interaction between the
intermediate generated at this position is attacked by dioxy- peptide and terpy ligands, as shown from the upfield shifts
Scheme 3. Plausible oxidation pathway of the cleavage reaction
Eur. J. Inorg. Chem. 2004, 4140Ϫ4145
www.eurjic.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4143

K. Jitsukawa, H. Takahashi, R. Hyuga, H. Arii, H. Masuda
FULL PAPER
of γ-H (δ ϭ 0.82 ppm) and the δ-H’s (δ ϭ 0.23 and
0.58 ppm) of the leucine moiety, seems to fix the C-terminal
Experimental Section
side chain of GGL on the terpy ring. On the other hand, General: Reagents used for synthesis were of the highest grade
available and were employed without further purification. The
starting cobalt() complex, [Co(CO3)(OH)(terpy)], was prepared
according to the reported method.[22] All solvents for spectroscopic
measurements were purified by distillation before use. Electronic
absorption spectra were recorded on a JASCO UVIDEC-660 spec-
trometer. 1H NMR spectra were measured on a Varian VXR-300S
or JEOL Lambda-500 spectrometer with TMS as an internal stand-
ard. Crystal-structure analyses were performed with an
EnrafϪNonius CAD4-EXPRESS four-circle diffractometer. Single
crystals of the complexes suitable for the X-ray diffraction measure-
in the ternary complexes 5c or 6c, where the C-terminal
residues are glycine, such upfield shifts are not observed.
Consequently, the corresponding oxidative-degradation
compounds were not generated in the cases of 5 and 6. It
is clear that the interligand interaction caused by the bulki-
ness of the C-terminal side-chain controls the sequence
specificity of this reaction. The reason why the cleavage re-
action does not take place in GBL (7), which contains a C-
terminal bulky side-chain might be explained by the diffi-
culty of abstraction of the protons attached on the N-2 β- ment were mounted on a glass capillary. The diffraction data were
collected with graphite-monochromated Mo-Kα radiation with the
ω-2θ scan technique at room temperature. Crystal data and exper-
imental details are listed in Table 2, in which weighted refinement
values (Rw) are obtained as follows: Rw ϭ [Σw(F2o Ϫ Fc2)2/Σw(F2o)2]1/
2. At the appropriate time of the reaction with GGL, the yields of
[Co(ggЈl)(terpy)] (1a), [Co(gЈl)(terpy)] (1b), and [Co(ggl)(terpy)] (1c)
were determined by HPLC analysis (column, ODS CrestPak C18s.
alanine residue.
Conclusion
A new type of oxidative transformation of peptide com-
pounds is carried out nonenzymatically in the presence of eluent, MeOH/water gradient method (5:95 Ǟ 35:65). UV-detector,
wavelength ϭ 513 nm). The ratio of these complexes varied during
the reaction.
transition metal complexes containing a terpy ligand. The
sequence-specific cleavage of peptides consisting of C-ter-
minal aliphatic amino acids occurs because of the interli-
Complexation and Characterization of Ternary Cobalt(III) Com-
gand interactions. Previously, the regioselective hydrolysis plexes: Under aerobic conditions, the reaction of [Co(CO3)(OH)(t-
erpy)] (0.37 g, 2 mmol) and an equimolar amount of tripeptide
[GGL (1), BGL (2), GGF (3), GGA (4), GGG (5), FGG (6), or
GBL (7)] was carried out at pH 8.5 and 40 °C in an aqueous solu-
tion (50 mL). After 10 h, separation of the reaction mixture on a
SP Sephadex C-25 column gave three separation patterns. In the
cases of 1Ϫ3 as the first group separation pattern, two fraction
bands (a and b) containing a neutral complex were eluted with
water, one (c) containing a positively charged complex was eluted
with 0.1  NaCl solution, and one (d) was eluted with 0.2  solu-
tion. In the case of 4 as the second group, one fraction band (a ϩ
b) was eluted with water, in which obvious separation of 4a and 4b
of peptides containing histidine or methionine residues has
been reported with palladium() complexes,[21] in which co-
ordination of the functional group of the peptide side chain
to the palladium center was the key step for generating the
selectivity. As compared with the previous reactions, our
finding that the bulkiness of the aliphatic side chain — a
nonfunctional group of the amino-acid residue — controls
the specificity is the first report of the oxidative degradation
of a peptide due to intramolecular noncovalent interactions
with a ternary cobalt() complex.
Table 2. Crystal data for the complexes 1b and 6c
Complex
{[Co(gЈl)(terpy)]}2·3.5H2O (1b)
Co2C46H51N10O11.5
1045.84
triclinic
P1
[Co(fgg)(terpy)]·2H2O (6c)
Formula
Fw
Crystal system
Space group
CoC28H29N6O6
604.51
triclinic
P1
˚
a (A)
11.336(6)
14.668(6)
8.304(4)
96.50(3)
100.06(4)
96.92(4)
1336(1)
1
1.299
6.85
0.083
0.275
8.884(1)
11.694(2)
7.594(2)
96.75(2)
105.75(2)
72.55(1)
723.8(3)
1
1.387
6.44
0.067
0.190
˚
b (A)
˚
c (A)
α (°)
β (°)
γ (°)
3
˚
V (A )
Z
ρcalcd. (g cmϪ3
)
µ (cmϪ1
)
[a]
R1
Rw[b]
˚
λ. A
0.71070
3805
0.71070
2500
No. of reflections used [I Ͼ 2σ (I)]
No. of variables
631
371
[a]
[b]
2
2 2
2
R1 ϭ Σ||Fo| Ϫ |Fc||/Σ|Fo|. Rw ϭ [Σw(Fo Ϫ Fc ) /Σw(Fo2)2]1/2; w ϭ 4Fo /σ2(Fo)2.
4144  2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.eurjic.org
Eur. J. Inorg. Chem. 2004, 4140Ϫ4145

Cobalt(III)-Mediated Oxidative Degradation of Tripeptides
FULL PAPER
from the column was not accomplished, and two bands, containing
(4c) and (4d), were eluted with 0.1  NaCl, and 0.2  NaCl solu-
tion, respectively. In the case of the reaction with 5Ϫ7, no frac-
tional band corresponding to a or b was eluted with water; one
band (c) was eluted with 0.1  NaCl solution and another (d) with
0.2  NaCl solution. These column patterns for the fractional
bands are typical for the sequence of tripeptides employed here. In
all cases, the complex contained in the last band (d) was identified
[1]
C. L. Hawkins, M. J. Davies, Biochem. Biophys. Acta 2001,
1504, 196Ϫ219.
W. M. Garrison, Chem. Rev. 1987, 87, 381Ϫ398.
E. R. Stadtman, Methods. Enzymol. 1995, 258, 379Ϫ393.
M. Jonsson, H.-B. Kaartz, J. Chem. Soc., Perkin Trans. 2
1997, 2673Ϫ2676.
[2]
[3]
[4]
[5]
R. T. Dean, S. Fu, R. Stocker, M. J. Davies, Biochem. J. 1997,
324, 1Ϫ18.
as [Co(terpy)2]3ϩ [9]
In the case of GGL (1), the amounts of 1b
.
[6] [6a]
T. M. Rana, C. F. Meares, J. Am. Chem. Soc. 1990, 112,
[6b]
increased with increasing reaction time. The structure of 1b was
confirmed by elemental analysis as well as by X-ray single crystal
and 1H NMR analysis. Elemental analysis of 1b:
[Co(gЈl)(terpy)]·4H2O; C23H30CoN5O8 (563.45): calcd. C 49.03, H
5.37, N 12.43; found C 49.23, H 5.30, N 12.13.
2457Ϫ2458.
T. M. Rana, C. F. Meares, Proc. Natl. Acad.
Sci. USA 1991, 88, 10578Ϫ10582.
K. Jitsukawa, T. Irisa, H. Einaga, H. Masuda, Chem. Lett.
2001, 30Ϫ31.
E. C. Constable, Tetrahedron 1992, 48, 10013Ϫ10059, and ref-
erences cited therein.
H. Elsbernd, J. K. Beattie, J. Inorg. Nucl. Chem. 1972, 34,
771Ϫ774.
J. K. Moran, C. F. Mears, Dictionary of Inorganic Compounds,
Chapman and Hall, London, 1992, p. 3075.
P. A. Sutton, D. A. Buckingham, Acc. Chem. Res. 1987, 20,
357Ϫ364.
[7]
[8]
After 24 h, the isolated yield of 1b was over 70% based on the
starting cobalt complex, while the total yields of 1a and 1c were
under 10%. Reverse-phase HPLC analyses of the reaction mixture
indicated that the primary product of the complex formation with
Co-terpy and GGL at the initial stage of the reaction was 1c, which
was oxidized to 1a under aerobic conditions. Under anaerobic con-
ditions, however, 1a was not detected in the HPLC analyses. Similar
time-dependent product distributions were observed in the reaction
with BGL (2) and GGF (3). In the case of GGG (5), FGG (6),
and GBL (7), the positively charged complexes, 5c (75%), 6c (65%),
and 7c (60%), eluted with 0.1  NaCl solution, were the main prod-
ucts, respectively; neutral products, corresponding to 5a, 5b, 6a, 6b,
7a, and 7b, were not detected even after 24 h. NMR measurement
of all complexes isolated from the eluting solution was carried out
in D2O solution. These data are listed in Table 1. The X-ray analy-
ses of the single crystals of 1b and 6c were performed according to
the previous methods.[7] Crystal data are summarised in Table 2.
During the crystallization of 6c at room temperature, the positively
charged complex eluted from the column changed to a neutral one
(presumably, carbonate ions generated from carbon dioxide in the
atmosphere might act as a reagent for charge neutralization), pre-
sumably because a neutral compound crystallizes more easily than
a positive one.
[9]
[10]
[11]
[12]
[13]
E. J. Evans, J. E. Grice, C. J. Hawkins, M. R. Heard, Inorg.
Chem. 1980, 19, 3496Ϫ3502.
D. A. Buckingham, C. E. Davis, D. M. Foster, A. M. Sargeson,
J. Am. Chem. Soc. 1970, 92, 5571Ϫ5579.
J. Chin, Acc. Chem. Res. 1991, 24, 145Ϫ152.
F. Himo, P. E. M. Siegbahn, Chem. Rev. 2003, 103, 2412Ϫ2456.
J. M. Rowland, M. M. Olmstead, P. K. Mascharak, Inorg.
Chem. 2002, 41, 2754Ϫ2760.
K. Jitsukawa, T. Hata, T. Yamamoto, H. Masuda, H. Einaga,
Chem. Lett. 1994, 1169Ϫ1172.
M. Bakir, P. S. White, A. Dovletoglou, T. J. Meyer, Inorg.
Chem. 1991, 30, 2835Ϫ2836.
[14]
[15]
[16]
[17]
[18]
[19] [19a]
F. Barigelletti, L. Flamigni, V. Balzani, J.-P. Collin, J.-P.
Sauvage, A. Sour, E. C. Constable, A. M. W. C. Thompson, J.
[19b]
Am. Chem. Soc. 1994, 116, 7692Ϫ7699.
B. J. Coe, D. W.
Thompson, C. T. Culbertson, J. R. Schoonover, T. J. Meyer,
Inorg. Chem. 1995, 34, 3385Ϫ3395.
M. Yashiro, M. Higuchi, M. Komiyama, Y. Ishii, Bull. Chem.
Soc. Jpn. 2003, 76, 1813Ϫ1817.
[20]
[21] [21a]
T. J. Parac, G. M. Ullmann, N. M. Kostic, J. Am. Chem.
[21b]
Soc. 1999, 121, 3127Ϫ3135.
N. M. Milovic, N. M. Kostic,
J. Am. Chem. Soc. 2002, 124, 4759Ϫ4769.
E. S. Kucharski, B. W. Skelton, A. H. White, Aust. J. Chem.
1978, 31, 47Ϫ51.
[22]
Acknowledgments
This research was supported in part by a Grant-in-Aid for Scien-
tific Research from the Ministry of Education, Science, Sports
and Culture.
Received March 31, 2004
Early View Article
Published Online August 12, 2004
Eur. J. Inorg. Chem. 2004, 4140Ϫ4145
www.eurjic.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4145