Aminopeptidase function of dinuclear zinc(II) complexes of phenol-based dinucleating ligands: Effect of p-substituents

Article abstract of DOI:10.1016/S0020-1693(03)00119-1

The aminopeptidase functions of newly obtained dinuclear zinc(II) complexes [Zn₂(bonp)(MeCO₂)₂]BPh₄ (1) and [Zn₂(bocp)(MeCO₂)₂]BPh₄ (2) were estimated using L-leucine-p-nitroanilide as a substrate [H(bonp): 2,6-bis[bis(2-methoxyethyl)aminomethyl]-4-nitrophenol; H(bocp): 4-chloro-2,6-bis[bis(2-methoxyethyl)aminomethyl]phenol]. Both complexes showed aminopeptidase activities, and a second-order rate equation was obtained as v=k[complex][substrate]. The second-order rate constants k were 5.9(5)×10^-1 for complex 1, and 2.7(1)×10^-2 dm³ mol^-1 s^-1 for complex 2. Compared to the previously reported complex, [Zn₂(bomp)(MeCO₂)₂]BPh₄ (3) [H(bomp): 2,6-bis[bis(2-methoxyethyl)aminomethyl]-4-methylphenol], the aminopeptidase activity was improved 250 times by the substitution of the p-methyl group into p-nitro group. The rate constants were found to be ordered as 1>2>3 (250:10:1), and this order is a decreasing order of the electron-withdrawing effect of the p-substituents of the dinucleating ligands 1(-NO₂)>2(-Cl)>3(-CH₃).

Full text of DOI:10.1016/S0020-1693(03)00119-1

Inorganica Chimica Acta 351 (2003) 256ꢀ260
/
www.elsevier.com/locate/ica
Aminopeptidase function of dinuclear zinc(II) complexes of phenol-
based dinucleating ligands: effect of p-substituents
Hiroshi Sakiyama *, Yasuhiro Igarashi, Yuuki Nakayama, Md. Jamil Hossain,
Kei Unoura, Yuzo Nishida
Department of Material and Biological Chemistry, Faculty of Science, Yamagata University, Kojirakawa, Yamagata 990-8560, Japan
Received 5 February 2003; accepted 22 February 2003
Abstract
The aminopeptidase functions of newly obtained dinuclear zinc(II) complexes [Zn2(bonp)(MeCO2)2]BPh4 (1) and [Zn2(bocp)(Me-
CO2)2]BPh4 (2) were estimated using
nitrophenol; H(bocp): 4-chloro-2,6-bis[bis(2-methoxyethyl)aminomethyl]phenol]. Both complexes showed aminopeptidase activ-
ities, and a second-order rate equation was obtained as vꢁk[complex][substrate]. The second-order rate constants k were 5.9(5)ꢂ
10ꢃ1 for complex 1, and 2.7(1)ꢂ10ꢃ2 dm3 molꢃ1 sꢃ1 for complex 2. Compared to the previously reported complex,
L
-leucineꢀ
/
p-nitroanilide as a substrate [H(bonp): 2,6-bis[bis(2-methoxyethyl)aminomethyl]-4-
/
/
/
[Zn2(bomp)(MeCO2)2]BPh4 (3) [H(bomp): 2,6-bis[bis(2-methoxyethyl)aminomethyl]-4-methylphenol], the aminopeptidase activity
was improved 250 times by the substitution of the p-methyl group into p-nitro group. The rate constants were found to be ordered
as 1ꢀ
/
2ꢀ/3 (250:10:1), and this order is a decreasing order of the electron-withdrawing effect of the p-substituents of the
dinucleating ligands 1(ꢀ
/
NO2)ꢀ2(ꢀCl)ꢀ3(ꢀCH3).
/
/
/
/
# 2003 Elsevier Science B.V. All rights reserved.
Keywords: Dinuclear zinc(II) complexes; Aminopeptidase function; Hydrolysis
1. Introduction
peptidase [7] (PAP, EC 3.4.11.9). LAP, AAP, and SGAP
contain dinuclear zinc(II) cores at their active sites;
MAP contains a dinuclear cobalt(II) core; and PAP
contains a dinuclear manganese(II) core.
Dinuclear zinc(II) cores are often seen in biological
systems, such as phosphatases [1,2] and aminopeptidases
[3ꢀ7]. Phosphatases are enzymes that hydrolyse phos-
/
The mechanism of aminopeptidase function has been
well studied for LAP [8], AAP [9], and MAP [10], and
the ‘two-metal ion mechanism’ has been proposed for
the three enzymes. The proposed mechanisms for the
enzymes have the following features in common: (1) an
N-terminal carbonyl oxygen of a substrate will bind to
one of the metal ion in the active center; and (2) a
hydroxo ion bridging the two metal ions will act as a
nucleophile to attack the N-terminal carbonyl carbon.
The former feature is useful to activate the substrate by
reducing the electron density of the target carbonyl
carbon in the hydrolysis. The latter feature is useful to
phomonoesters, and aminopeptidases are enzymes that
remove the N-terminal amino acid from a protein or
peptide. It is interesting that all of the well-characterised
aminopeptidases contain dinuclear metal cores at their
active sites, while carboxypeptidases, which remove the
C-terminal amino acid, do not have dinuclear cores. At
present, five types of aminopeptidases are structurally
characterised: leucine aminopeptidase [3] (LAP, EC
3.4.11.1), methionine aminopeptidase [4] (MAP, EC
3.4.11.18), aminopeptidase from Aeromonas proteolytica
[5] (AAP, EC 3.4.11.10), Streptomyces griseus amino-
generate the nucleophilic Mꢀ
/
OH (Mꢁ
MnII, CoII, ZnII)
species even under the pH condition that enough free
/
peptidase [6] (SGAP, EC 3.4.11.ꢀ), and proline amino-
/
hydroxo ions do not exist.
Previously, we reported [11] a dinuclear zinc(II)
complex [Zn(bomp)(MeCO2)2]BPh4 (3) [H(bomp): 2,6-
* Corresponding author. Tel.: ꢄ
4591.
E-mail address: saki@sci.kj.yamagata-u.ac.jp (H. Sakiyama).
/
81-23-628-4601; fax: ꢄ81-23-628-
/
0020-1693/03/$ - see front matter # 2003 Elsevier Science B.V. All rights reserved.
doi:10.1016/S0020-1693(03)00119-1

H. Sakiyama et al. / Inorganica Chimica Acta 351 (2003) 256ꢀ
/
260
257
bis[bis(2-methoxyethyl)aminomethyl]-4-methylphenol]
as a first functional aminopeptidase model. Complex 3
contains a m-phenoxo-bis(m-acetato)dizinc(II) structure,
which resembles the active site structure of the native
aminopeptidases. The aminopeptidase function of 3 was
2.3. Preparations
2.3.1. Na(bonp)
A formaldehyde solution (37%, 7.8 ml, 100 mmol)
and ethanol (20 ml) were added to an aqueous solution
(50 ml) containing p-nitrophenol (6.95 g, 50 mmol),
NaOH (2.00 g, 50 mmol) and bis(2-methoxyethyl)amine
(13.32 g, 100 mmol), and the resulting solution was
refluxed for 5 days. After evaporation to dryness the
resulting oily products were extracted with ether. Ether
was removed by evaporation to give Na(bonp) as
colourless oil. Yield 5.25 g (23%). Selected IR data [n/
cmꢃ1] using NaCl plates: 3000ꢀ
1420, 1370, 1290, 1190, 1115, 1015, 840. H NMR [d/
ppm] in CDCl3: 2.89 (t, 8H), 3.21 (s, 12H), 3.36 (t, 8H),
3.51 (s, 4H), 8.13 (s, 2H).
estimated using
and the rate of the hydrolysis could be expressed by a
second-order rate equation, i.e., vꢁk[complex][sub-
strate] with kꢁ2.3(1)ꢂ
10ꢃ3 dm3 molꢃ1 sꢃ1. We
L
-leucineꢀ/p-nitroanilide as a substrate,
/
/
/
were successful in preparing the first functional amino-
peptidse model; however, the activity was less than 1/
10 000 times that for native aminopeptidases. Therefore,
our next purpose is to improve the aminopeptidse
activity of the functional model. In the present study,
two new dinucleating ligands, bonpꢃ and bocpꢃ
[H(bonp): 2,6-bis[bis(2-methoxyethyl)aminomethyl]-4-
nitrophenol; H(bocp): 4-chloro-2,6-bis[bis(2-methox-
yethyl)aminomethyl]phenol], in which the p-methyl
group of bompꢃ is substituted, have been synthesised,
and the aminopeptidase activities of the dinuclear
zinc(II) complexes with the ligands [Zn2(bonp)(Me-
CO2)2]BPh4 (1) and [Zn2(bocp)(MeCO2)2]BPh4 (2)
have been examined.
/
2820, 1675, 1590, 1460ꢀ
/
1
2.3.2. Na(bocp)
This compound was prepared as oil by a method
similar to that of Na(bonp) using p-chlorophenol
instead of p-nitrophenol. Yield 6.35 g (29%). Selected
IR data [n/cmꢃ1] using NaCl plates: 2980ꢀ
1590, 1490ꢀ
NMR [d/ppm] in CDCl3: 2.90 (t, 8H), 3.33 (s, 12H), 3.57
/
2820, 1660,
1
/
1440, 1385, 1270, 1200, 1110, 1005, 830. H
(t, 8H), 3.91 (s, 4H), 7.16 (s, 2H).
2.3.3. [Zn2(bonp)(MeCO2)2]BPh4 (1)
Zinc(II) acetate dihydrate (0.44 g, 2.0 mmol) was
added to a methanolic solution (15 ml) of Na(bonp)
(0.45 g, 1.0 mmol), and the resulting solution was
refluxed for 1 h to give a yellow solution. The addition
of sodium tetraphenylborate (0.34 g, 1.0 mmol) resulted
in the precipitation of white microcrystals. Yield 0.61 g
(61%). Anal. Found: C, 57.65; H, 6.05; N, 4.20; Zn, 12.9.
Calc. for C48H60BN3O11Zn2: C, 57.85; H, 6.00; N, 4.20;
Zn, 13.1%. Selected IR data [n/cmꢃ1] using KBr disk:
2. Experimental
3050ꢀ2920, 1600, 1505, 1430, 1315, 1195, 1090, 1010,
/
1
910, 850, 730, 700. H NMR [d/ppm] in CDCl3: 2.01 (s,
6H), 2.86 (t, 8H), 3.21 (s, 12H), 3.36 (t, 8H), 3.57 (s, 4H),
6.89 (t, 4H), 7.02 (t, 8H), 7.38 (m, 8H), 8.13 (s, 2H).
2.1. Measurements
Elemental analyses were obtained at the Elemental
Analysis Service Centre of Kyushu University. Infrared
(IR) spectra were recorded on a Hitachi 270-50 spectro-
2.3.4. [Zn2(bocp)(MeCO2)2]BPh4 (2)
1
meter. H NMR spectra (400 MHz) were recorded on a
This was prepared as white microcrystals by a method
similar to that of 1 using Na(bocp) instead of Na(bonp).
Yield 0.35 g (35%). Anal. Found: C, 58.50; H, 6.15; N,
2.85; Zn, 13.4. Calc. for C48H60BClN2O9Zn2: C, 58.45;
JEOL JNM-a400 spectrometer in CDCl3 using SiMe4 as
the internal standard. Electronic spectra were recorded
on a Shimadzu UV-240 spectrophotometer.
H, 6.15; N, 2.85; Zn, 13.3%. Selected IR data [n/cmꢃ1
]
using KBr disk: 3050ꢀ2920, 1600, 1425, 1325, 1250,
/
1
2.2. Materials
1095, 1010, 850, 730, 700. H NMR [d/ppm] in CDCl3:
1.99 (s, 6H), 2.90 (t, 8H), 3.33 (s, 12H), 3.57 (t, 8H), 3.91
(s, 4H), 6.90 (t, 4H), 7.04 (t, 8H), 7.16 (s, 2H), 7.39 (m,
8H).
All chemicals were commercial products and were
used as supplied.

258
H. Sakiyama et al. / Inorganica Chimica Acta 351 (2003) 256ꢀ260
/
2.4. Aminopeptidase activity
Aminopeptidase activity of the complexes was esti-
mated using -leucineꢀp-nitroanilide as a substrate at
room temperature. Kinetic measurements were made as
follows. -Leucineꢀp-nitroanilide was dissolved in 1.5
L
/
L
/
ml of Tricine buffer solution (pH 8), and to this was
added 1.0 ml of DMF solution containing complex 1 at
room temperature. Hydrolysis of
L
-leucineꢀp-nitroani-
/
lide into l ꢀleucine and p-nitroaniline was monitored by
/
detecting the formation of p-nitroaniline using spectro-
meter at 405 nm. In the measurement, spontaneous
hydrolysis of the substrate was subtracted as a back-
ground. This measurement was examined at various
Fig. 1. An absorbance change at 405 nm. Conditions: [complex 1]ꢁ
3.29ꢂ -leucineꢀp-nitroanilide]ꢁ2.08ꢂ
10ꢃ5 mol dmꢃ3, [ 10ꢃ4 mol
dmꢃ3, DMF:H2O (pH 8.0)ꢁ
2:3. Solid line was drawn by a method in
the text using parameters kꢁ5.90ꢂ
10ꢃ1 dm3 molꢃ1 sꢃ1 and k?ꢁ
35.1 sꢃ1
/
10ꢃ5 mol dmꢃ3
/
L
/
/
/
complex concentrations from 0 to 4.5ꢂ
for 1 and from 0 to 8.25ꢂ
10ꢃ4 mol dmꢃ3 for 2, and at
various substrate concentrations from 0 to 5.75ꢂ
10ꢃ4
mol dmꢃ3 for 1 and from 0 to 7.52ꢂ10ꢃ4 mol dmꢃ3
/
/
/
/
/
/
/
.
/
for 2. In the case of 1, a spontaneous decomposition of
the complex cation causes an increase in absorbance.
Since this rate is first order to the complex concentra-
tion, this increment is subtracted to obtaine the initial
rate.
measurement, spontaneous hydrolysis of the substrate
was subtracted as a background although the sponta-
neous hydrolysis was much slower. An example of the
absorbance change is shown in Fig. 1. This kind of
measurement was carried out at various substrate
concentrations and at various complex concentrations,
and the initial rate v was obtained for the each
experiment. When the initial rate over the substrate
concentration v/[substrate] is plotted against the com-
plex concentration [complex], a good linearity is ob-
served for both complexes 1 and 2 (Figs. 2 and 3), which
indicates that the initial rate can be written as a second-
order rate equation as follows:
3. Results and discussion
3.1. General properties of complexes 1 and 2
The IR spectra of complexes 1, 2 and 3 show the
antisymmetric and symmetric n(CO2) vibrations of the
acetate group at 1600ꢀ
/
1610 and 1425ꢀ
/
1445 cmꢃ1
,
respectively. A Dn value [Dnꢁ
/
nas(CO2)ꢃns(CO2)] smal-
/
vꢁk [substrate] [complex]
(1)
ler than 200 cmꢃ1 is typical of the bridging carboxylate
group [12]. The crystal structure of complex 3 was
determined previously, and two zinc(II) ions were found
to be bridged by one phenolic oxygen and two acetate
groups, forming a m-phenoxo-bis(m-acetato)dizinc(II)
core. Judging from the similarity of the IR spectra, it
seems reasonable to assume the m-phenoxo-bis(m-acet-
ato)dizinc(II) structures for both 1 and 2. Complexes 1
and 2 show no appreciable absorption in the region
above 350 nm in DMF solution, and neither does 3, in
accord with the d10 electronic configuration of the
zinc(II) ion.
where k is the second-order rate constant. The k value
was calculated as 5.9(5)ꢂ
10ꢃ1 dm3 molꢃ1 sꢃ1 for 1
and 2.7(1)ꢂ
10ꢃ2 dm3 molꢃ1 sꢃ1 for 2. The previously
10ꢃ3 dm3
/
/
obtained k value for complex 3 was 2.3(1)ꢂ
/
mo1ꢃ1 sꢃ1 under the same conditions [11]. Thus, the
rate for 1 is about 250 times that for 3 and the rate for 2
is about ten times that for 3. When the log k values were
3.2. Aminopeptidase function
3.2.1. Aminopeptidase activity by an initial-rate method
First, aminopeptidase activity of complexes 1 and 2
was estimated by an initial-rate method using
p-nitroanilide as a substrate in a mixture of 40% DMF
L
-leucineꢀ
/
and 60% aqueous solution at pH 8.0. The hydrolysis of
the substrate into L-leucine and p-nitroaniline was
monitored by the formation of p-nitroaniline, which
has an absorption band in the near-UV region. In the
Fig. 2. A v/[substrate] vs. [complex] plot in the hydrolysis of
leucineꢀp-nitroanilide by 1 at pH 8.0.
L-
/

H. Sakiyama et al. / Inorganica Chimica Acta 351 (2003) 256ꢀ
/
260
259
Fig. 3. A v/[substrate] vs. [complex] plot in the hydrolysis of
leucineꢀp-nitroanilide by 2 at pH 8.0.
L-
/
Fig. 5. The k vs. nominal pH plot for 2 (k) and 3 (^).
plotted against the Hammet’s p-substituent constants sp
of the p-substituents [13] in the dinucleating ligands, a
good linearity was seen (Fig. 4). This indicates that the
the parameters kdepro
pKaꢁ9.1 for 2 and kdepro
and pKaꢁ9.4 for 3, respectively. This result indicates
the formation of the ZnꢀOH species as a nucleophile in
ꢁ
/
5.8ꢂ
/
10ꢃ1 dm3 molꢃ1 sꢃ1 and
10ꢃ2 dm3 molꢃ1 sꢃ1
/
ꢁ
/
4.1ꢂ
/
hydrolysis is accelerated [k(3)B
electron-withdrawing effect of the dinucleating ligand
became strong [sp(3)Bsp(2)Bsp(1)].
/
k(2)Bk(1)] when the
/
/
/
/
/
the hydrolysis. The pKa value of the bound water
molecule for 2 is slightly smaller than that for 3, and
this can be explained by the stronger Lewis acidity
caused by the stronger electron-withdrawing p-substi-
tuent in 2. The difference in pKa value is not so much for
the complexes, however, the k value differs significantly.
This indicates that the improvement in activity is mainly
caused by the reduction of the electron density on the
target carbonyl carbon through coordination bond.
3.2.2. Effect of pH for aminopeptidase function
The above procedure was also carried out at pH’s
varying from 7.0 to 10.0, and the second-order rate
equation was always found to be valid. Hereafter a term
‘nominal pH’ will be used because the experiments were
carried out in a mixture of DMF and water. The k
versus nominal pH plots for 2 and 3 are shown in Fig. 5.
The both plots are sigmoidal around the nominal pH 9,
and this behavior is typical for the reaction promoted by
the deprotonated form of the catalyst [14]. In this case a
zinc-bound water molecule is thought to be deproto-
3.2.3. Some considerations about the mechanism
According to the crystal structure of 3, coordination
geometry around each zinc(II) ion is saturated octahe-
dral, and the two zinc(II) ions are bridged by two acetate
ions. However, dissociation of the acetate ions occurs
rather easily in an aqueous solution, affording vacant
coordination sites for substrate incorporation. Ideally
for LAP and MAP, the carbonyl oxygen and the amine
nitrogen of the substrate will be bounded to each
metal(II) ion [8,10] (Fig. 6), but, in the present case,
this mode is thought to be difficult because of the steric
repulsion between the methoxy groups of the dinucleat-
ing ligand and the p-nitrophenyl group of the substrate.
At the moment we are not sure about the intermediate
structure, but from the above discussion, the cabonyl
carbon of the substrate must be bound to one of the
nated to generate a ZnꢀOH species, which can act as a
nucleophile in the hydrolysis [14]. In this study pH-curve
fitting was also made based on Eq. (2).
/
Ka
Kaꢄ[Hꢄ
kꢁkdepro
(2)
]
where kdepro is the rate constant for the deprotonated
form. The data for the complexes were well fitted using
Fig. 4. Correlation between the log k and sp.
Fig. 6. An ideal intermediate for LAP and MAP.

260
H. Sakiyama et al. / Inorganica Chimica Acta 351 (2003) 256ꢀ260
/
Table 1
Rate constants for the hydrolysis by the complexes
4. Conclusion
Complex k (dm3 molꢃ1 sꢃ1
)
k? (sꢃ1
)
Reference
This work
In this study, the aminopeptidase activity of dinuclear
zinc(II) complexes has been improved by introducing
the stronger electron-withdrawing p-substituents,
although the stability of the dinuclear structure has
been decreased. From the kinetic studies, the incorpora-
tion of the substrate by the dinuclear center and the
1
2
3
5.9(5)ꢂ
5.90(3)ꢂ
/
10ꢃ1
/
10ꢃ1 a
1.02(5)ꢂ
9(5)ꢂ
10ꢃ3 a
3.3(8)ꢂ
10ꢃ4 a
/
10ꢃ2 a
2.7(1)ꢂ
3.0(6)ꢂ
2.3(1)ꢂ
2.5(4)ꢂ
/
10ꢃ2
This work
[11]
/
10ꢃ2 a
10ꢃ3
/
/
/
10ꢃ3 a
/
formation of nucleophilic ZnꢀOH species have been
/
proved in the present model system as found in the
native enzymes.
a
These values were obtained by the time-course method.
zinc(II) ions and Znꢀ/OH species must work as the
nucleophile.
5. Supplementary material
Routine kinetic results have been deposited as sup-
plementary material.
3.2.4. Aminopeptidase activity by a time-course method
In addition to the above discussion, which was based
on an initial-rate method, the data were also analyzed by
a time-course method, as we had previously done for
complex 3 [11]. In this treatment, two parallel reactions,
hydrolysis (Eq. (3)) and spontaneous complex decom-
position (Eq. (4)), were considered since the observed
time-course curves were lower than the curves expected
by Eq. (3) alone.
References
[1] C.G. Dealwis, L. Chen, C. Brennan, W. Mandecki, C. Abad-
Zapatero, Protein Eng. 8 (1995) 865.
[2] Y. Zhang, J.Y. Liang, S. Huang, H. Ke, W.N. Lipscomb,
Biochemistry 32 (1993) 1844.
[3] S.K. Burley, P.R. David, A. Taylor, W.N. Lipscomb, Proc. Natl.
Acad. Sci. USA 87 (1990) 6878.
k
[4] S.L. Roderick, B.W. Matthews, Biochem. 32 (1993) 3907.
[5] B. Chevrier, C. Schalk, H. D’orchymont, J.M. Rondeau, D.
Moras, C. Tarnus, Structure 2 (1994) 283.
(Zn2 complex)ꢄ(substrate) 0(deactivated complex)
ꢄ(hydrolysis product)
(3)
(4)
[6] H.M. Green Blatt, O. Almog, B. Maras, A. Spungin-Bialik, D.
Barra, S. Blumberg, G. Shoham, J. Mol. Biol. 165 (1997) 620.
[7] M.C. Wilce, C.S. Bond, N.E. Dixon, H.C. Freeman, J.M. Guss,
P.E. Lilley, J.A. Wilce, Proc. Natl. Acad. Sci. USA 95 (1998)
3472.
k?
(Zn2 complex) 0(decomposed species)
An example of the simulated curve is shown in Fig. 1.
The k and k? values obtained for the complexes are
summarised in Table 1. The k values obtained by the
time-course method are in good agreement with the
values by the initial-rate method. The k? values,
representing the spontaneous decomposition of the
[8] N. Strater, W.N. Lipscomb, Biochem. 34 (1995) 14792.
¨
[9] G. Chen, T. Edwards, V.M. D’souza, R.C. Holz, Biochem. 36
(1997) 4278.
[10] W.T. Lowther, B.W. Matthews, Biochim. Biophys. Acta 1477
(2000) 157.
[11] H. Sakiyama, R. Mochizuki, A. Sugawara, M. Sakamoto, Y.
Nishida, M. Yamasaki, J. Chem. Soc., Dalton Trans. (1999) 997.
[12] G.B. Deacon, R.J. Phillips, Coord. Chem. Rev. 32 (1980) 227.
[13] H.H. Jaffe´, Chem. Rev. 53 (1953) 191.
complex cation, are in the order of 1ꢁ
/
2ꢀ3. This result
/
is reasonable since the ligation ability of the dinucleating
ligands including strong electron-withdrawing p-substi-
tuents is poor.
[14] E. Kimura, T. Shiota, T. Koike, M. Shiro, M. Kodama, J. Am.
Chem. Soc. 112 (1990) 5805.