Artificial metalloesterases constructed by site-directed attachment of oximinato metal centers to poly(ethylenimine) containing β-cyclodextrin

Article abstract of DOI:10.1006/bioo.1998.1090

In an effort to establish a methodology for construction of active sites of artificial enzymes, site-directed attachment of 2,6- diacetylpyridineketoxime (III) to poly(ethylenimine) (PEI) containing β- cyclodextrin (CD) was attempted. The site-directed functionalization exploited a t-butylphenyl ester (I) of a carboxylic acid containing a precursor of III. Anchoring of the t-butylphenyl group of I by the CD portion followed by transfer of III moiety to an amino group located in the vicinity of the CD moiety would lead to introduction of III in proximity to the CD moiety. By acylation in DMSO of CD-PEI with the phenyl ester (II), instead of the t-butylphenyl ester, III was introduced randomly to CD-PEI. In the presence of the Ni(II) or Zn(II) complex of the III-containing CD-PEI prepared by either the site-directed or the random functionalization method, ester hydrolysis of 4-(4'-acetoxy-phenylazo)-benzenesulfonate (IV) was considerably enhanced. Analysis of the kinetic data measured at various pHs revealed that k(cat) for the PEI derivative prepared by site-directed modification was three to six times greater than that by random modification. The results were taken to indicate that I transferred III to the vicinity of the CD moiety of CD-PEI, but that orientation of III and the CD cavity in the resulting PEI derivative was not very productive for deacylation of IV complexed by the CD cavity.

Full text of DOI:10.1006/bioo.1998.1090

BIOORGANIC CHEMISTRY 26, 103–117 (1998)
ARTICLE NO. BH981090
Artificial Metalloesterases Constructed by Site-Directed
Attachment of Oximinato Metal Centers to Poly(ethylenimine)
Containing ͱ-Cyclodextrin
Junghun Suh1 and Won Jong Kwon
Department of Chemistry and Center for Molecular Catalysis, Seoul National University,
Seoul 151-742, Korea
Received February 11, 1998
In an effort to establish a methodology for construction of active sites of artificial enzymes,
site-directed attachment of 2,6-diacetylpyridineketoxime (III) to poly(ethylenimine) (PEI)
containing ͱ-cyclodextrin (CD) was attempted. The site-directed functionalization exploited
a t-butylphenyl ester (I) of a carboxylic acid containing a precursor of III. Anchoring of the
t-butylphenyl group of I by the CD portion followed by transfer of III moiety to an amino
group located in the vicinity of the CD moiety would lead to introduction of III in proximity
to the CD moiety. By acylation in DMSO of CD-PEI with the phenyl ester (II), instead of
the t-butylphenyl ester, III was introduced randomly to CD-PEI. In the presence of the Ni(II)
or Zn(II) complex of the III-containing CD-PEI prepared by either the site-directed
or the random functionalization method, ester hydrolysis of 4-(4Ј-acetoxy-phenylazo)-
benzenesulfonate (IV) was considerably enhanced. Analysis of the kinetic data measured at
various pHs revealed that kcat for the PEI derivative prepared by site-directed modification
was three to six times greater than that by random modification. The results were taken to
indicate that I transferred III to the vicinity of the CD moiety of CD-PEI, but that orientation
of III and the CD cavity in the resulting PEI derivative was not very productive for deacylation
of IV complexed by the CD cavity.
1998 Academic Press
INTRODUCTION
In devising artificial enzymes, it is necessary to incorporate catalytic elements
exploited by enzymes in relatively small spaces on synthetic molecules. Furthermore,
in order to achieve cooperative catalytic participation of several functional groups
in the chemical transformation of the bound substrate, fine alignment of convergent
catalytic groups is needed. Creation of such active sites on molecular skeletons
provided by relatively small synthetic hosts would not be easy. For this reason,
branched poly(ethylenimine) (PEI) has been exploited as a molecular backbone in
an approach to the design of artificial enzymes (1).
PEI contains ethylamine as the repeating unit and, thus, is highly soluble in water.
The molecular weight of the PEI used in the construction of artificial enzymes is
ca. 60,000, corresponding to 1,400 monomer residues. Among the 1,400 nitrogens
present in PEI, ca. 25, 50, and 25% are primary, secondary, and tertiary amines,
1 To whom correspondence should be addressed.
103
0045-2068/98 $25.00
Copyright 1998 by Academic Press
All rights of reproduction in any form reserved.

104
SUH AND KWON
respectively. The tertiary amines represent branching points on the polymer back-
bone, and PEI is highly branched. The amino nitrogens of PEI can be easily modified
by alkylation, acylation, or imine-formation, allowing incorporation of various cata-
lytic elements to the backbone.
One of the basic obstacles to overcome in the design of artificial enzymes by
using synthetic polymers is the lack of specific binding sites on the macromolecular
backbone. As for binding sites for certain hydrophobic moieties, ͱ-cyclodextrin
(CD) has been attached to PEI to obtain CD-PEI2 (2). CD-PEI manifested high
affinity toward t-butylphenyl compounds and accelerated deacylation of esters con-
taining t-butylphenyl moieties. In a subsequent study, macrocyclic metal complexes
have been randomly built on the PEI backbone of CD-PEI (3), which acted as
effective catalytic groups for ester hydrolysis.
As a further step toward design of effective artificial enzymes based on CD-PEI,
it is needed to develop methodology for introduction of additional catalytic elements
in the vicinity of the CD cavity by site-directed functionalization (4). Functionaliza-
tion of CD-PEI can be also regarded as preparation of new derivatives (5) of CD.
In the present study, site-directed functionalization of CD-PEI was undertaken
for introduction of oximinato metal centers. Attachment of oximinato metal centers
in proximity to the CD cavity could lead to cooperation by metal centers and the
oxime group as catalytic groups as well as the CD cavity as the binding site. Previous
2 The abbreviated names of the PEI derivatives indicate nature of the pendants (CD, Ac, DAP,
DAPOx, or M(II)) attached to PEI and whether DAP is introduced by site-directed or random (SD
or Ran, respectively) functionalization. For example, [Zn(II)DAPOx-CD]SDAcPEI stands for the PEI
derivative containing CD and acetyl groups as well as the Zn(II) complex of DAPOx introduced by
site-directed functionalization.

CONSTRUCTION OF ARTIFICIAL METALLOESTERASES
105
studies (6) showed that attack of 2-pyridyl oxime groups at esters and breakdown
of the resulting carboxy oxime esters are catalyzed by the metal ion bound to the
oxime nitrogen. In this article, introduction of oximinato metal complexes to
CD-PEI by site-directed or random functionalization is described together with
kinetic data for ester deacylation by the resulting artificial metalloesterases.
EXPERIMENTAL PROCEDURES
Materials
4-t-Butylphenyl 4-(6-acetyl-pyridin-2-yl)-4-oxobutyrate (I). A benzene solution
(100 ml) of 2,6-diacetylpyridine (16 g, 0.10 mol), ethyleneglycol (6.2 g, 0.10 mol),
and p-toluenesulfonic acid (0.20 g, 0.20 mol) was refluxed for 1 h. The solvent was
evaporated under a reduced pressure and the resulting residue was purified by
column chromatography (SiO2 , 6 : 1 hexane-ethyl acetate) to obtain 1-[6-(2-methyl-
1
[1,3]dioxola-2-yl)-pyridin-2-yl]-ethanone (Ia) as a pale yellow oil: H NMR (200
MHz, CDCl3) ͳ 1.82 (s, 3H), 2.76 (s, 3H), 3.99 (m, 2H), 4.15 (m, 2H), 7.75 (dd,
1H), 7.85 (t, 1H), 7.89 (dd, 1H). To a solution of Ia (1.0 g, 4.8 mmol) in tetrahydrofu-
ran (THF) (20 ml) was added sodium bis(trimethylsilyl)amide (5.8 ml of 1.0 M
solution in THF, 5.8 mmol) and the mixture was stirred for 30 min at Ϫ78ЊC. After
addition of ethyl bromoacetate (0.53 ml, 4.8 mmol), the reaction mixture was stirred
for 1 h at Ϫ20ЊC. After quenching the reaction by adding aqueous 5% NaHCO3
(100 ml), the reaction mixture was extracted with CH2Cl2 and the residue obtained
by evaporation of the organic layer under a reduced pressure was purified by column
chromatography (SiO2 , 4 : 1 hexane-ethyl acetate) to obtain ethyl 4-[6-(2-methyl-
1
[1,3]dioxolan-2-yl)-pyridin-2-yl]-4-oxobutyrate (Ib) as a pale yellow oil: H NMR
(200 MHz, CDCl3) ͳ 1.82 (s, 3H) 2.75 (t, 2H), 3.62 (t, 2H), 3.97 (m, 2H), 4.14 (m,
2H), 7.75 (dd, 1H), 7.85 (t, 1H), 7.98 (dd, 1H). Ester Ib was hydrolyzed by refluxing
in 1 : 1 THF-3 N HCl (100 ml) for 3 h. The product was extracted with CH2Cl2 and the
residue, 4-(6-acetyl-pyridin-2-yl)-4-oxobutyric acid (Ic), obtained by evaporation of
1
the solvent was used in the next step without further purification: H NMR (200
MHz, CDCl3) ͳ 2.80 (s, 3H), 2.86 (t, 2H), 3.63 (t, 2H), 8.00 (t, 1H), 8.23 (m, 2H).
To a solution of Ic (1.0 g, 2.8 mmol) and t-butylphenol (0.87 g, 5.8 mmol) in
CH2Cl2 were added N,NЈ-dicyclohexylcarbodiimide (1.2 g, 5.8 mmol) and 4-N,N-
dimethylaminopyridine (0.1 g) and the mixture was stirred for 1 h at 0ЊC. After
evaporation of the solvent, the residue was dissolved in 1 : 1 ethyl acetate-hexane
(100 ml). After removal of undissolved solids, the solution was concentrated and
purified by column chromatography (SiO2 , 6 : 1 hexane-ethyl acetate) and recrystal-
lization from hexane-ether to obtain I as white crystals, mp 71–73ЊC. 1H NMR (200
MHz, CDCl3) ͳ 1.30 (s, 9H), 2.80 (s, 3H), 3.06 (t, 2H), 3.76 (t, 2H), 7.04 (m, 2H),
7.48 (m, 2H), 8.00 (t, 1H), 8.23 (m, 2H). FAB-MS 354 (MHϩ). Anal. Calcd. for
C21H23NO4 : C, 71.37; H, 6.56; N, 3.96. Found: C, 71.09; H, 6.72; N, 3.86.
Phenyl 4-(6-acetyl-pyridin-2-yl)-4-oxobutyrate (II). This compound was prepared
by the method used for preparation of I by using phenol instead of 4-t-butylphenol,
mp 67–68ЊC. 1H NMR (200 MHz, CDCl3) ͳ 2.80 (s, 3H), 3.06 (t, 2H), 3.76 (t, 2H),

106
SUH AND KWON
7.04 (m, 2H), 7.32 (m, 1H), 7.48 (m, 2H), 8.00 (t, 1H), 8.23 (m, 2H). FAB-MS 298
(MHϩ). Anal. Calcd. for C17H25NO4 : C, 68.69; H, 5.09; N 4.71. Found: C, 68.42;
H, 4.92; N, 4.63.
CD-PEI. By the procedure reported previously (2), CD was attached to PEI and
the content of CD in the purified CD-PEI was estimated as 1.4 residue mol %.
[DAP-CD]SDPEI. To a solution of 0.162 residue M CD-PEI in 650 ml water and
100 ml DMSO, I (0.598 g, 1.69 mmol) dissolved in 50 ml DMSO was added dropwise
over a period of 1.5 h at 40ЊC, and the solution was stirred at 40ЊC for 3 days. The
resulting PEI derivative, [DAP-CD]SDPEI, was purified by dialysis against 30%
(v/v) ethanol-water (thrice), saturated NaCl solutions (twice), and water (thrice).
[DAP-CD]RanPEI. An aqueous solution of 0.242 residue M CD-PEI (100 ml)
was evaporated under a reduced pressure below 40ЊC and the resulting dry polymer
was dissolved in 200 ml DMSO. After addition of II (0.101 g, 0.339 mmol), the
reaction mixture was stirred at 40ЊC for 3 days. The resulting PEI derivative,
[DAP-CD]RanPEI, was purified by dialysis as described above for [DAP-CD]SDPEI.
[DAPOx-CD]SDPEI and [DAPOx-CD]RanPEI. To a solution of [DAP-CD]SDPEI
or [DAP-CD]RanPEI (0.242 residue M, 250 ml, pH 7.0), hydroxylamine hydrochlo-
ride (14 g, 200 mmol) dissolved in 10 ml water was added and the resulting solution
was stirred for 6 days at 40ЊC. The resulting PEI derivatives, [DAPOx-CD]SDPEI
and [DAPOx-CD]RanPEI, were purified by dialysis against saturated NaCl solutions
(thrice) followed by water (four times).
[DAPOx-CD]SDAcPEI and [DAPOx-CD]RanAcPEI. Acetylation of [DAPOx-
CD]SDPEI or [DAPOx-CD]RanPEI was carried out with acetic anhydride as reported
previously (3, 4). The mixture was stirred at pH 12 for 12 h to hydrolyze acetyl oxime
that might have been formed during the acetylation stage. After neutralization to
pH 7, the resulting PEI derivatives, [DAPOx-CD]SDPEI and [DAPOx-CD]RanPEI,
were purified by dialysis against saturated NaCl solutions (thrice) followed by water
(four times).
2,6-Diacetylpyridine dioxime (III). To a solution of 2,6-diacetylpyridine (3.3 g,
20 mmol) in ethanol (50 ml), an aqueous solution (20 ml) of hydroxylamine hydro-
chloride (1.5 g, 22 mmol) was added and the mixture was refluxed for 4 h. The
residue obtained by evaporation of the solvent under a reduced pressure was
1
recrystallized from ethyl acetate-hexane to obtain III, mp 252–254ЊC (dec). H
NMR (80 MHz, DMSO-d6) ͳ 2.27 (s, 6H), 7.80 (t, 1H), 8.02 (d, 2H). FAB-MS
194(MHϩ). Anal. Calcd. for C9H11N3O2 : C, 55.95; H, 5.74; N, 21.75. Found: C, 55.84;
H, 5.68; N, 21.82.
4-(4Ј-Acetoxy-phenylazo)-benzenesulfonic acid (IV). An aqueous solution (10
ml) of sodium nitrite (0.75 g, 11 mmol) was poured into a concentrated HCl solution
(150 ml) containing sulfanilic acid (1.7 g, 10 mmol) at 0ЊC. The reaction mixture
was stirred for 30 min in an ice bath and then added dropwise into a solution of
phenol (1.0 g, 11 mmol) and sodium carbonate (100 g, 1.0 mol) in 4 : 1 water-
methanol (250 ml) at 0ЊC. The reaction mixture was stirred for 1 h and then
acidified with 1 N HCl (100 ml). The precipitates thus obtained were purified by
recrystallization from ethanol to obtain 4-(4Ј-hydroxy-phenylazo)-benzenesulfonic
acid (IVa), mp 282–284ЊC (dec). That the azo coupling reaction took place at the
1
para position of phenol was confirmed by 13C NMR data. H NMR (80 MHz,

CONSTRUCTION OF ARTIFICIAL METALLOESTERASES
107
SCHEME 1
DMSO-d6) ͳ 6.98 (d, 2H), 7.80 (m, 6H), 10.32 (s, 1H). 13C NMR (80 MHz, DMSO-
d6) ͳ 116.01, 121.63, 126.69, 143.24, 145.27, 149.24, 152.03, 161.12. Anal. Calcd. for
C12H10N2O4S: C, 51.79; H, 3.62; N, 10.07; S, 11.52. Found: C, 51.53; H, 3.77; N, 9.97;
S, 11.43. The solution of IVa (2.8 g, 10 mmol) in 1 : 6 DMF-acetic anhydride (350
ml) was stirred for 4 h at 50ЊC and then cooled in an ice bath. The precipitates
were purified by recrystallization from ethanol to obtain IV, mp 162–163ЊC (dec).
1H NMR (80 MHz, DMSO-d6) ͳ 2.31 (s, 3H), 6.94 (d, 2H), 7.80 (m, 6H). FAB-MS
321(MHϩ). Anal. Calcd. for C14H12N2O5S: C, 52.49; H, 3.78; N, 8.75; S, 10.10. Found:
C, 52.31; H, 3.84; N, 8.63; S; 9.89.
Measurements
Kinetic studies for deacylation of IV were carried out spectrophotometrically
with a Beckman DU 650 UV-Vis spectrophotometer. pH was adjusted with a
Dongwoo Medical System pH/ion meter DP 880. Buffers (0.05 M) used were
sodium acetate (pH 4.8), 2-[N-morpholino]ethanesulfonate (pH 5.8–6.5) and N-[2-
hydroxyethyl]piperazine-NЈ-[2-ethanesulfonic acid] (pH 7–8).
RESULTS
The t-butyl phenyl ester (I) and the phenyl ester (II) of 4-(6-acetyl-pyridin-2-yl)-
4-oxobutyric acid were prepared according to the synthetic route summarized in
Scheme 1. As summarized in Scheme 2, CD-PEI was coupled with I in 13–19%
(v/v) DMSO-water to obtain [DAP-CD]SDPEI. By treatment with NH2OH, [DAP-
CD]SDPEI was converted into [DAPOx-CD]SDPEI, the corresponding oxime. Ace-
tylation of primary and secondary amino groups of the PEI backbone with acetic

108
SUH AND KWON
SCHEME 2
anhydride transformed [DAPOx-CD]SDPEI into [DAPOx-CD]SDAcPEI. By cou-
pling CD-PEI with II in DMSO, [DAP-CD]RanPEI was obtained, which was con-
verted into [DAPOx-CD]RanPEI and [DAPOx-CD]RanAcPEI by further treatment
with NH2OH followed by acetic anhydride. It is known that treatment of PEI
derivatives with acetic anhydride effectively acetylates primary and secondary amino
groups of the PEI backbone (3, 4).
The content of CD in CD-PEI was estimated as 1.4 residue mol% and, therefore,
each CD-PEI molecule contained 20 CD moieties on the average. In the preparation
of [DAP-CD]SDPEI or [DAP-CD]RanPEI, the molar amount of I or II used was
equivalent to that of the CD moieties. If the acylation of CD-PEI with I or II was
quantitative, the content of the pyridyl moieties in the PEI derivatives subsequently
prepared would be also 1.4 residue mol%. Elemental analysis of the PEI derivatives
did not give reliable data for estimation of the content of the pyridyl moieties.
Instead, the content of the oxime was estimated by comparing absorbance (305
nm) of [DAPOx-CD]SDAcPEI or [DAPOx-CD]RanAcPEI with that of III in 3 M
NaOH solutions. The content of III moiety thus estimated was 0.88 residue mol%
for [DAPOx-CD]SDAcPEI and 1.2 residue mol% for [DAPOx-CD]RanAcPEI. The
number of III moiety for each CD moiety is 0.60 and 0.82, respectively, for [DAPOx-
CD]SDAcPEI and [DAPOx-CD]RanAcPEI.
Anionic ester IV was used as the substrate instead of neutral esters. This is
because of the much greater solubility of the ionic ester that allows kinetic studies
under the conditions of S0 (initially added substrate concentration) ӷ C0 (initially
added catalyst concentration). With neutral substrates, a large amount of an organic
cosolvent is needed to raise the substrate concentration to a sufficiently high level

CONSTRUCTION OF ARTIFICIAL METALLOESTERASES
109
FIG. 1. Plot of k0 against C0 measured for the deacylation of IV in the presence of Ni(II) ion and
[DAPOx-CD]SDAcPEI at pH 5.88 and 25ЊC under the conditions of C0 ӷ S0 . The amount of Ni(II)
ion added is 1.2 (᭡), 1.0 (᭺), 0.8 (ᮀ), or 0 (᭹) equivalent of the oxime moiety of the PEI derivative.
The curve is obtained by analysis with Eq. [2].
and the presence of such a cosolvent would interfere with the recognition of the
hydrophobic substrate by the CD moiety of the catalyst.
Rates for deacylation of IV was studied in the presence of [DAPOx-CD]SDAcPEI
and [DAPOx-CD]RanAcPEI. When Ni(II) or Zn(II) was added to the reaction
mixture, the rate was enhanced greatly. In Fig. 1, a sample of kinetic data obtained
for deacylation of IV by the PEI derivatives in the presence of the bivalent metal
ion added in various ratios relative to the oxime moieties is illustrated. When neither
Ni(II) ion nor Zn(II) ion was added, the deacylation rate was much slower than
the rate measured in the presence of the metal ions. This excludes the possibility
of nucleophilic attack by the hydroxyl groups of the CD cavities or the amino
groups of the PEI backbone. When tested at various pHs, the rate data indicated
that addition of Ni(II) or Zn(II) ion in a 1 : 1 molar ratio to the oxime moieties of
the PEI derivatives was sufficient to achieve maximal rates.
Kinetic data obtained with reactions catalyzed by PEI derivatives conform to the
Michaelis–Menten scheme (Eq. [1]) (1–4). Under the conditions of C0 ӷ S0 ,3 Eq.
[2] is derived from Eq. [1] as the expression of the pseudo-first-order rate constant
(k0). From dependence of k0 on C0 (expressed in terms of the concentration of the
CD moieties) measured under the conditions of C0 ӷ S0 , values of kcat and Km
were estimated. Values of the kinetic parameters estimated at various pHs for the
Ni(II) or Zn(II) complexes of [DAPOx-CD]SDAcPEI and [DAPOx-CD]RanAcPEI
3 More accurately, C0 ȁ [C] ӷ [CS].

110
SUH AND KWON
FIG. 2. pH profiles of kcat for deacylation of IV in the presence of [Ni(II)DAPOx-CD]SDAcPEI (a,
᭹), [Zn(II)DAPOx-CD]SDAcPEI (b, ᭺), [Ni(II)DAPOx-CD]RanAcPEI (c, ᭿), and [Zn(II)DAPOx-
CD]RanAcPEI (d, ᮀ) at 25ЊC. When the data were analyzed according to the scheme of Eq. [3], the
following parameter values were estimated, which were used for construction of the theoretical curves:
lim
3
1
lim
3
1
k
ϭ (3.16 Ϯ 0.55) ϫ 10Ϫ sϪ and pK ϭ 6.59 Ϯ 0.42 for curve a, k ϭ (3.94 Ϯ 0.65) ϫ 10Ϫ sϪ and
cat
cat
lim
3
1
pK ϭ 7.04 Ϯ 0.27 for curve b, k ϭ (1.60 Ϯ 0.26) ϫ 10Ϫ sϪ and pK ϭ 6.82 Ϯ 0.33 for curve c, and
cat
lim
3
1
k
ϭ (1.55 Ϯ 0.28) ϫ 10Ϫ sϪ and pK ϭ 7.03 Ϯ 0.30 for curve d.
cat
are summarized in Figs. 2 and 3. The pH profiles of kcat illustrated in Fig. 2 were
analyzed in terms of Eq. [3] by analogy with the method employed in enzyme
kinetics (7).
Km
kcat
C ϩ S
CS —Ǟ C ϩ Pi
[1]
[2]
k0 ϭ kcat C0/(Km ϩ C0)
lim
K
k
cat
(CS)Hϩ
CS —Ǟ C ϩ Pi
[3]
When the deacylation of IV was carried out in the presence of [Ni(II)DAPOx-
5
3
CD]SDAcPEI under the conditions of C0 (2.32 ϫ 10Ϫ M) Ӷ S0 (1.93 ϫ 10Ϫ M),
the concentration of III consumed during the period of 3 h at pH 6.96 and 25ЊC
4
was 2.8 ϫ 10Ϫ M. Since a low concentration of the catalyst was used, the rate of
the catalyzed reaction was slow but was still much faster than the spontaneous
reaction. The amount of the substrate consumed under the reaction conditions is
about 12 times greater than C0 .
The pKa values for the two oxime groups of Ni(II) or Zn(II) complex of III were
estimated by spectral titration. The absorbance (278.5 nm for Ni(II) complex and
4
341.3 nm for Zn(II) complex) of buffer solutions containing III (1.0 ϫ 10Ϫ M for

CONSTRUCTION OF ARTIFICIAL METALLOESTERASES
111
FIG. 3. pH profiles of Km for deacylation of IV in the presence of [Ni(II)DAPOx-CD]SDAcPEI (ᮀ),
[Zn(II)DAPOx-CD]SDAcPEI (᭺), [Ni(II)DAPOx-CD]RanAcPEI (᭜), and [Zn(II)DAPOx-CD]RanAc-
PEI (᭹) at 25ЊC.
the Ni(II) complex and 2.6 ϫ 10Ϫ4 M for the Zn(II) complex) and equimolar amount
of Ni(II) or Zn(II) was measured at pH 5–9 (25ЊC, 0.1 M NaCl, 2% (v/v) DMSO
for Ni(II), 5% (v/v) DMSO for Zn(II)). Analysis of the data revealed pKa1 ϭ 5.87 Ϯ
0.23 and pKa2 ϭ 7.62 Ϯ 0.27 for the Ni(II) complex of III and pKa1 ϭ 6.78 Ϯ 0.09
and pKa2 Ͼ 9.5 for the Zn(II) complex. As checked with the changes in absorbance
caused by pH changes, the oxime groups in the Ni(II) or Zn(II) complexes of
[DAPOx-CD]SDAcPEI or [DAPOx-CD]RanAcPEI appeared to ionize in the range
of pH 6–7.5, although the pKa values were not accurately measured.
Rates for the deacylation of IV were also measured in the presence of Ni(II) or
Zn(II) complex of III under the conditions of S0 Ӷ C0 (0.6–1.1 mM) at pH 6.96.
The second-order rate constants (k0/C0) measured for Ni(II)III and Zn(II)III were
2
1
1
2
1
1
(3.8 Ϯ 0.2) ϫ 10Ϫ sϪ MϪ and (4.7 Ϯ 0.2) ϫ 10Ϫ sϪ MϪ , respectively. These
values are 27 and 18 times, respectively, smaller than the values4 of kcat/Km (based
on C0 calculated as the concentration of the oxime moiety instead of CD moiety)
measured for [Ni(II)DAPOx-CD]SDAcPEI or [Zn(II)DAPOx-CD]SDAcPEI at the
same pH.
DISCUSSION
Procedures for the site-directed or random functionalization of CD-PEI to pre-
pare [DAPOx-CD]SDAcPEI or [DAPOx-CD]RanAcPEI, respectively, are summa-
rized in Scheme 2. Site-directed functionalization of CD-PEI takes advantage of
4 When Km ӷ C0 , kcat/Km ϭ k0/C0 .

112
SUH AND KWON
SCHEME 3
the recognition of t-butylphenyl groups by CD. As illustrated in V, if t-butylphenyl
moiety of I is anchored by CD-PEI, I would selectively transfer the diacetylpyridyl
moiety to an amino group located in vicinity to the CD cavity. Due to the insolubility
of I in water, it was inevitable to add an organic solvent to the reaction mixture.
Since CD recognizes hydrophobic molecules even in 60% (v/v) DMSO-water (5b),
site-directed functionalization of CD-PEI with I was attempted in 13–19% (v/v)
DMSO-water.
It is well established that CD forms inclusion complexes with t-butylphenyl deriva-
tives more strongly than the corresponding phenyl derivatives. This has been also
confirmed by a kinetic study on ester deacylation promoted by CD-PEI (2). The
recognition of the t-butylphenyl compounds by CD is due to hydrophobic interaction
between the guest and the host. The hydrophobic effect would not be significant
in nonaqueous media such as DMSO. Functionalization of CD-PEI with I was
conducted in water containing 13–19% (v/v) DMSO, whereas that with II was
carried out in 100% DMSO. Considering the structural differences between I and
II as well as the solvent employed, it is quite reasonable to assume the site-directed
and random functionalization of CD-PEI by I and II, respectively.
As summarized in Scheme 3, nucleophilic attack by the oximate anion of 2-
pyridylcarboxaldoxime or 2-acylpyridineketoxime is accelerated by complexation
of the oximes to metal ions due to the enhanced ionization of the oximes (6). In

CONSTRUCTION OF ARTIFICIAL METALLOESTERASES
113
addition, hydrolysis of the resulting acyl oxime esters is catalyzed by metal ions
(6). Since both formation and breakdown of the acyl-oxime intermediate is acceler-
ated, the metal complexes of 2-pyridylcarboxaldoxime or 2-acylpyridineketoxime
act as effective catalysts in hydrolysis of phenyl esters.
In view of the catalytic ability of the metal complexes of 2-acylpyridineketoximes,
III was attached to CD-PEI. The results illustrated in Fig. 1 revealed that the
maximum activity of [DAPOx-CD]SDAcPEI is attained when Ni(II) or Zn(II) ion
was added in a 1 : 1 molar ratio to the oxime moieties attached to the polymer. This
is consistent with the strong and selective metal binding to the oxime moieties,
although the formation constants for the metal complexes were not measured. The
pH dependence of kcat measured with the metal complexes of [DAPOx-CD]SD
AcPEI or [DAPOx-CD]RanAcPEI agrees with the ionization of the oxime groups.
The pK values calculated from analysis of pH profiles of kcat (Fig. 2) are consistent
with the pKa values of oxime groups estimated with the metal complexes of either
the PEI derivatives or III. The possibility that the increased kcat at higher pHs is
due to amino groups of PEI backbone or hydroxyl groups of CD moieties is excluded
on the ground of the lack of activity of the PEI derivatives in the absence of Ni(II)
or Zn(II) ion (Fig. 1). The data summarized in Fig. 1 show that the tertiary amino
groups as well as the primary or secondary amino groups that might have not been
blocked by acetylation do not contribute significantly to deacylation of IV.
Control experiments were carried out with the Ni(II) or Zn(II) complex of III
which is not covalently attached to CD-PEI. They manifested much lower activity
for deacylation of IV compared with the oximinato complexes attached covalently
to CD-PEI. Kinetic data obtained with the metal complexes of [DAPOx-CD]SD
AcPEI or [DAPOx-CD]RanAcPEI as well as the control compounds demonstrate
that both the metal-oxime complex and CD moiety attached to PEI are responsible
for the effective deacylation of IV in the presence of the metal complexes of
[DAPOx-CD]SDAcPEI or [DAPOx-CD]RanAcPEI.
If the acyl-oxime intermediate formed by the attack at IV by the oxime group
attached to the PEI derivatives resists hydrolysis, the oxime-containing PEIs do
not catalyze the hydrolysis of IV. To clarify that the acyl-oxime intermediates indeed
turns over to regenerate the catalytic oxime group, hydrolysis of IV by the Ni(II)
complex of [DAPOx-CD]SDAcPEI was carried out under the conditions of S0 ӷ
C0 . The result revealed that the oxime catalyst was regenerated and, thus, is an
effective catalyst for the hydrolysis IV.
The kinetic data measured under the conditions of S0 Ӷ C0 provide information
on attack of the oxime catalysts at IV leading to the formation of the acyl-catalyst
intermediate even when the breakdown of the acyl-catalyst intermediate is the rate-
controlling step. Although such kinetic data do not necessarily produce quantitative
information on overall hydrolysis of IV, they lead to clues to conformation of
catalytic elements positioned on the PEI derivatives.
If the III moieties take productive positions in vicinity to the CD cavities of CD-
PEI, an effective artificial metalloesterase would be obtained (VI). When the oxime
group is attached to CD-PEI by random functionalization, attack by the oxime
group at the bound ester should be preceded by a large change in the conformation
of the polymer backbone (VII).

114
SUH AND KWON
The catalytic efficiency of [M(II)DAPOx-CD]SDAcPEI was examined by measur-
ing kinetic data for hydrolysis of IV by [M(II)DAPOx-CD]SDAcPEI and [M(II)DA-
POx-CD]RanAcPEI. The number of III moiety for each CD cavity is 0.60 and 0.82,
respectively, in [M(II)DAPOx-CD]SDAcPEI and [M(II)DAPOx-CD]RanAcPEI. To
reflect this difference, a statistical correction is made with kcat for [M(II)DAPOx-
CD]SDAcPEI by multiplying with 82/60, and the kinetic parameters for [M(II)DA-
POx-CD]SDAcPEI and in [M(II)DAPOx-CD]RanAcPEI are compared in Fig. 4.
Little difference is observed for Km between [Ni(II)DAPOx-CD]SDAcPEI and
[Ni(II)DAPOx-CD]RanAcPEI], whereas Km is larger by 1.3–3.3 times for [Zn(II)-
DAPOx-CD]SDAcPEI than for [Zn(II)DAPOx-CD]RanAcPEI (Fig. 5). On the other
hand, three- to fivefold greater kcat is observed for [M(II)DAPOx-CD]SDAcPEI
compared with [M(II)DAPOx-CD]RanAcPEI. If a much greater difference had
been observed for kcat between [M(II)DAPOx-CD]SDAcPEI and [M(II)DAPOx-
CD]RanAcPEI, successful site-direct functionalization of CD-PEI with I would have
been clearly demonstrated.
The moderate difference in kcat may be taken to indicate low efficiency in the site-
directed functionalization. Even if the site-directed functionalization is successful,
however, it can be associated with less effective acceleration. If III is indeed incorpo-
rated into [M(II)DAPOx-CD]SDAcPEI by site-directed functionalization, about
60% of the CD cavities of [M(II)DAPOx-CD]SDAcPEI would have nearby III
moieties. On the other hand, the remaining CD cavities of [M(II)DAPOx-CD]SDAc-

CONSTRUCTION OF ARTIFICIAL METALLOESTERASES
115
FIG. 4. Ratio (lighter bars, right side) of kcat measured for [Zn(II)DAPOx-CD]SDAcPEI and [Zn(II)-
DAPOx-CD]RanAcPEI and that (darker bars, left side) of kcat measured for [Ni(II)DAPOx-CD]SDAcPEI
and [Ni(II)DAPOx-CD]RanAcPEI at various pHs. See the text for statistical corrections made to reflect
the differences in the content of DAPOx moieties.
FIG. 5. Ratio (darker bars, right side) of Km measured for [Zn(II)DAPOx-CD]SDAcPEI and [Zn(II)-
DAPOx-CD]RanAcPEI and that (lighter bars, left side) of Km measured for [Ni(II)DAPOx-CD]SDAcPEI
and [Ni(II)DAPOx-CD]RanAcPEI at various pHs.

116
SUH AND KWON
PEI as well as most of the CD cavities of [M(II)DAPOx-CD]RanAcPEI would
hardly have a nearby CD cavity. Effective catalysis in hydrolysis of IV would be
achieved only when IV is complexed into a CD cavity of [M(II)DAPOx-CD]SDAc-
PEI with a nearby III moiety and when the oximate anion is properly oriented
toward the ester linkage of the complexed IV.
The kinetic data were obtained under the conditions of C0 ӷ S0 . Thus, the CD
cavities in [M(II)DAPOx-CD]SDAcPEI and in [M(II)DAPOx-CD]RanAcPEI are
only partially (Ͻ5%) filled with IV. In the CS complex formed with IV and [M(II)-
DAPOx-CD]SDAcPEI, IV would occupy a CD cavity with or without a III moiety
present in proximity. If IV is complexed preferentially by CD cavities without nearby
III moieties due to steric hindrance, hydrolysis of IV would not be accelerated
considerably by [M(II)DAPOx-CD]SDAcPEI. Even if IV is complexed by CD cavi-
ties with nearby III moieties,5 effective catalysis is possible only when the orientation
of III and IV in the CS complex is productive.
Even if the diacetylpyridyl moiety is introduced by site-directed functionalization,
whether the conformation of the polymer is frozen to maintain the productive
conformation imposes another barrier to achieve high degree of acceleration. For
design of artificial enzymes by site-directed functionalization of the macromolecular
skeleton, therefore, it is desirable to develop a methodology to suppress the confor-
mational freedom of the polymer backbones.
The degree of rate acceleration achieved by [M(II)DAPOx-CD]SDAcPEI is only
moderately greater than that by [M(II)DAPOx-CD]RanAcPEI. The results, how-
ever, indicate that effective artificial enzymes can be designed with synthetic macro-
molecules by further improvement in the methodology of site-directed functionali-
zation.
ACKNOWLEDGMENTS
This work was supported by grants from Lotte Fellowship (1997) and from Korea Science and
Engineering Foundation through Center for Molecular Catalysis, Seoul National University.
REFERENCES
1. (a) Klotz, I. M. (1987) in Enzyme Mechanisms (Page, M. I., and Williams, A., Eds.), Chap. 2, R.
Soc. Chem., London. (b) Suh, J., Cho, Y., and Lee, K. J. (1991) J. Am. Chem. Soc. 113, 4198. (c)
Suh, J. (1992) Acc. Chem. Res. 25, 273. (d) Suh, J. (1996) in Polymeric Materials Encyclopedia
(Salamone, J. C. Ed.), 4210, CRC Press, Boca Raton.
2. Suh, J., Lee, S. H., and Zoh, K. D. (1992) J. Am. Chem. Soc. 114, 7916.
3. Zoh, K. D., Lee, S. H., and Suh, J. (1994) Bioorg. Chem. 22, 242.
5 Since no stable intermediate would accumulate prior to the release of the phenol leaving group
during deacylation of IV, 1/Km may be taken as the formation constant for the most stable complex
formed between the catalyst and the substrate. Smaller 1/Km of IV observed with [Zn(II)DAPOx-
CD]SDAcPEI compared with [Zn(II)DAPOx-CD]RanAcPEI suggests that steric hindrance may be in-
volved in complexation of IV to [Zn(II)DAPOx-CD]SDAcPEI.

CONSTRUCTION OF ARTIFICIAL METALLOESTERASES
4. Kim, S. M., Hong, I. S., and Suh, J. Bioorg. Chem., in press.
117
5. (a) Tabushi, I., Shimizu, N., Sugimoto, T., Shiozuka, M., and Yamamura, K. (1977) J. Am. Chem.
Soc. 99, 7100. (b) Breslow, R., and Czarniecke, M. F. (1978) J. Am. Chem. Soc. 100, 7771. (c) Breslow,
R., Bovy, P., and Hersh, C. L. (1980) J. Am. Chem. Soc. 102, 215. (d) Akkaya, E. U., and Czarnik,
A. W. (1988) J. Am. Chem. Soc. 110, 8553. (e) Nakamura, M., Ikeda, T., Nakamura, A., Ikeda, H.,
Ueno, A., and Toda, F. (1995) Chem. Lett. 343. (f) Breslow, R., and Schmuck, C. (1996) J. Am.
Chem. Soc. 118, 6601. (g) Breslow, R., and Zhang, X. (1996) J. Am. Chem. Soc. 118, 11678. (h)
Breslow, R., and Zhang, B. (1997) J. Am. Chem. Soc. 119, 1676.
6. (a) Breslow, R., and Chipman, D. (1965) J. Am. Chem. Soc. 105, 135. (b) Suh, J., Lee, E., Jang, E. S.
(1981) Inorg. Chem. 20, 1932. (c) Suh, J., Cheong, M., and Suh, M. P. (1982) J. Am. Chem. Soc. 108,
1654. (d) Suh, J., and Han, H. (1984) Bioorg. Chem. 12, 177. (e) Suh, J., Cheong, M., and Han, H.
(1984) Bioorg. Chem. 12, 188. (f) Suh, J., Suh, M. P., and Lee, J. D. (1985) Inorg. Chem. 24, 3088.
(g) Suh, J., Han, O., and Chang, B. (1986) J. Am. Chem. Soc. 108, 3057. (h) Suh, J., Kwon, B. N.,
Chang, S. H., and Lee, W. Y. (1987) Inorg. Chem. 26, 805. (i) Suh, J., and Chang, B. (1987) Bioorg.
Chem. 15, 167. (j) Suh, J., and Koh, D. (1987) J. Org. Chem. 52, 3446.
7. Zeffren, E., and Hall, P. L. (1973) in The Study of Enzyme Mechanism, pp. 79–86, Wiley, New York.