Crosslinking of imprinted proteases to maintain a tailor-made substrate selectivity in aqueous solutions

Article abstract of DOI:10.1016/S0968-0896(99)00156-X

A covalent method to keep imprinted properties of proteins stable in aqueous as well as in organic environment is described. To stabilize the ligand induced acceptance for d-configured substrates by α-chymotrypsin or subtilisin Carlsberg, each protein was first vinylated by acylation with itaconic anhydride. Then, the tailoring of the derivatized proteins by precipitation in the presence of N-acetyl-D-tryptophan from an aqueous medium with 1-propanol, and the subsequent crosslinking of the enzyme preparations with ethylene glycol dimethacrylate in cyclohexane was carried out. The crosslinked imprinted proteins (CLIPs) obtained catalyzed the hydrolysis of N-acetyl-D-tryptophan ethyl ester in phosphate buffer and the corresponding back reaction in cyclohexane, respectively. The repeated use of CLIP-α- chymotrypsin in D-ester hydrolysis was demonstrated. Furthermore, this particular CLIP-α-chymotrypsin showed no loss in activity when it subsequently was used in the synthesis of N-acetyl-D-tryptophan ethyl ester in cyclohexane again. In the case of D-ester hydrolysis the reaction rate acceleration (k(enz)/k(nonenz)) was in the same order of magnitude of about 10⁴ -10⁵ mM^-1 for the two CLIP-proteases. The results suggest that enzymes tailored by imprinting technique do not lose their induced 'new' property in the presence of water when they are prepared according to the described vinylation/crosslinking method (CLIP technique).

Full text of DOI:10.1016/S0968-0896(99)00156-X

Bioorganic & Medicinal Chemistry 7 (1999) 2231±2237
Crosslinking of Imprinted Proteases to Maintain a Tailor-Made
Substrate Selectivity in Aqueous Solutions
Fabian Peiûker and Lutz Fischer*
Institute of Biochemistry and Biotechnology, Spielmannstr. 7, D-38106 Braunschweig, Germany
Received 30 November 1998
AbstractÐA covalent method to keep imprinted properties of proteins stable in aqueous as well as in organic environment is
described. To stabilize the ligand induced acceptance for d-con®gured substrates by a-chymotrypsin or subtilisin Carlsberg, each
protein was ®rst vinylated by acylation with itaconic anhydride. Then, the tailoring of the derivatized proteins by precipitation in
the presence of N-acetyl-d-tryptophan from an aqueous medium with 1-propanol, and the subsequent crosslinking of the enzyme
preparations with ethylene glycol dimethacrylate in cyclohexane was carried out. The crosslinked imprinted proteins (CLIPs)
obtained catalyzed the hydrolysis of N-acetyl-d-tryptophan ethyl ester in phosphate bu€er and the corresponding back reaction in
cyclohexane, respectively. The repeated use of CLIP-a-chymotrypsin in d-ester hydrolysis was demonstrated. Furthermore, this
particular CLIP-a-chymotrypsin showed no loss in activity when it subsequently was used in the synthesis of N-acetyl-d-tryptophan
ethyl ester in cyclohexane again. In the case of d-ester hydrolysis the reaction rate acceleration (kenz/knonenz) was in the same order
4
5
� 1
of magnitude of about 10 � 10 mM for the two CLIP-proteases. The results suggest that enzymes tailored by imprinting tech-
nique do not lose their induced ``new'' property in the presence of water when they are prepared according to the described viny-
lation/crosslinking method (CLIP technique). # 1999 Elsevier Science Ltd. All rights reserved.
4
Introduction
substrate selectivity8 or the enantioselectivity of
9
enzymes and, the binding properties of proteins were
manipulated. Also a ``de novo'' catalytic ability in
existing mature proteins via imprinting with a transition
To tailor the properties of proteins in order to match
new desired abilities concerning stability, substrate
selectivity or enantioselectivity, rational1 or random2
mutagenesis have emerged as powerful methods that
modify the protein at its genetic level. However, for
practical reasons it would be desirable to have methods
available that allow a predictable alteration of protein
properties directly at the protein level. One step in this
direction is based on the ability of proteins to maintain
a structural memory when transferred from aqueous
10
state analogue was reported recently. Unfortunately,
all these examples of tailoring proteins by imprinting
techniques are limited by the fact that the new protein
properties are available in nearly anhydrous environ-
ment only since aqueous environment causes a rena-
3
turation of the protein and loss of the imprinting e€ect.
But, if the imprinting e€ect relies on any kinds of
structural manipulation of the proteins,11 it should be
possible to maintain them by crosslinking the ``rigid''
protein in organic solvent. Here, ®rst results of studies
are reported which suggest that the imprinting e€ect can
indeed be covalently stabilized by combining imprinting
with a two step immobilization method and, the imprin-
ted enzymes do permanently maintain their ``new''
property in aqueous as well as in organic environment.
3
into anhydrous environment. So, if a native protein is
4
5
precipitated or lyophilized from aqueous environment
in the presence of a de®ned ligand, the ligand can cause
a predictable alteration of the property of the protein in
organic solvents. Generally, this phenomenon is called
6
(
bio)imprinting. By this, the catalytic activity,7 the
Key words: Tailor-made enzymes; imprinting; bioimprinting; immo-
bilization.
Results
Abbreviations: CT, a-chymotrypsin; SUB, subtilisin carlsberg; NIP,
native imprinted protein; DIP, derivatized imprinted protein; CLIP,
crosslinked imprinted protein; l-ee, N-acetyl-l-tryptophan ethyl ester;
d-ee, N-acetyl-d-tryptophan ethyl ester; N-Ac-d-Trp, N-acetyl-d-tryp-
tophan.
Manipulation of the substrate selectivity of native
ꢀ
-chymotrypsin
4
*
3
Corresponding author. Tel.: +49-531-3915730; fax: +49-531-
915763; e-mail: l.®scher@tu-bs.de
As described recently, the substrate selectivity of native
a-chymotrypsin (CT) in organic solvents is altered when
0968-0896/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(99)00156-X

2
232
F. Peiûker, L. Fischer / Bioorg. Med. Chem. 7 (1999) 2231±2237
CT is ``imprinted'' with its competitive inhibitor N-ace-
tyl-d-tryptophan (N-Ac-d-Trp) in a bu€er solution. In
comparison to non-imprinted CT, imprinted CT was
then able to accept N-Ac-d-Trp as the substrate in ester
be 0.11, 0.162 and 0.174, respectively. In all further stu-
dies, an additive concentration of 20 mM N-Ac-d-Trp
was used. This value corresponds to an approx. 5-fold
concentration of the K of CT for N-Ac-d-Trp (4.16
i
4
13
synthesis. In order to investigate this ®nding in more
mM).
detail, solubilized CT was precipitated with 1-propanol
in the presence of N-Ac-d-Trp at pH 3.6, 5.4 and 7.8.
Then, the washed and dried protein precipitates were
used for N-Ac-d-Trp ethyl ester synthesis in cyclohex-
Aside from achieving the imprinting e€ect by precipita-
tion, lyophilization of the mixture containing CT and
N-Ac-d-Trp could also be used. After extensive wash-
ing of the lyophilizated powder with 1-propanol it was
tested for N-Ac-d-Trp ethyl ester synthesis in cyclohex-
ane and, approx. 83% of the speci®c activity measured
with the precipitated imprinted CT (see above) was
obtained.
ꢀ
ane at various temperatures (15±35 C). The highest
yields in d-ester synthesis were obtained when CT was
precipitated at pH 5.4 and the reaction was carried out
ꢀ
at 25 C. Another important parameter for ester synth-
1
2
eses in organic solvents is the free water activity. Thus,
supplementary water in the range from 0 to 1.7 M was
added in the beginning of the d-ester synthesis reaction
running in dry cyclohexane and, the yield of N-Ac-d-
Trp ethyl ester was measured after 48h. For this reac-
tion, the optimum of the water concentration turned out
to be 440 mM. However, this value deviated sig-
Derivatization and crosslinking of ꢀ-chymotrypsin
To maintain the manipulated properties of CT caused
by imprinting with N-Ac-d-Trp permanently, poly-
merizable vinyl groups were introduced on CT by
acylation with itaconic anhydride. When itaconic anhy-
dride is used to acylate proteins it mainly reacts with the
amino groups of lysines and to some extent also with
the hydroxyl groups of tyrosines and the thiol groups of
cysteines.14 A non-linear correlation between the degree
of acylation of CT and its resulting residual hydrolytic
activity towards benzoyl tyrosine ethyl ester (BTEE)
was found (Fig. 1). In order to have enough attachment
points per CT available for crosslinking but also an
acceptable residual activity, a derivatization degree of
about 60% was chosen empirically in further experi-
ments. This derivatized CT was imprinted with N-Ac-d-
Trp in bu€er solution, precipitated, dried and either
tested directly for N-Ac-d-Trp ethyl ester synthesis in
organic solvent or, it was ®rst crosslinked with ethylene
glycol dimethacrylate by radical polymerization and
then tested for d-ester synthesis. In Figure 2 the time
4
ni®cantly from the published data so far (1.2 mM). The
speci®c activities calculated in our experiments were
0.02 nkat/g imprinted CT when no supplementary water
was added to the reaction medium, 0.17 nkat/g at a
water concentration of 440 mM and no activity when
1.7 mM water was added. The latter probably is
explained by the renaturation of the CT in the presence
of too many water molecules and with that, CT's loss of
the imprinted property.
Also the e€ect of the concentration of the additive N-
Ac-d-Trp used in the bu€er prior to the precipitation
step was investigated. CT was co-solubilized with either
10 mM, 20 mM or 40 mM N-Ac-d-Trp. The yields in N-
Ac-d-Trp ethyl ester syntheses obtained with the result-
ing CTs were estimated by HPLC after 24 h. The spe-
ci®c activities (nkat/g imprinted CT) were calculated to
Figure 1. Activity of derivatized a-chymotrypsin dependent on the particular degree of enzyme acylation with itaconic anhydride.

F. Peiûker, L. Fischer / Bioorg. Med. Chem. 7 (1999) 2231±2237
2233
Figure 2. Synthesis of the N-acetyl-d-tryptophan ethyl ester in cyclohexane using various kinds of a-chymotrypsin [NIP (d-)-CT=native imprinted
protein (a-chymotrypsin) that was imprinted with N-acetyl-d-tryptophan, DIP (d-)-CT=derivatized imprinted protein (a-chymotrypsin) that was
imprinted with N-acetyl-d-tryptophan, CLIP (d-)-CT=crosslinked imprinted protein (a-chymotrypsin) that was imprinted with N-acetyl-d-trypto-
ꢀ
phan]. Reaction conditions: 10 mg of each biocatalyst preparation, N-acetyl-d-tryptophan=20 mM, total reaction volume=1.26 mL, 25 C; ther-
momixer 1200 rpm.
course of d-ester synthesis of the native, imprinted pro-
tein [NIP (D)-CT], the derivatized, imprinted protein
Based on the apparent kenz for the enzymes, a relative
4
� 1
reaction rate acceleration (kenz/knonenz) of 5.3 10 mM
5
� 1
[
[
DIP (D)-CT] and the crosslinked, imprinted protein
CLIP (D)-CT] is shown. Clearly, all three kinds of
and 1.1 10 mM for CLIP (D)-CT and CLIP (D)-
SUB were estimated. The ®rst time it was demonstrated
that an imprinted property of an enzymeÐin this case
the manipulated enantioselectivity of CT and SUBÐ
could be stabilized by covalent crosslinking and thus,
this ``tailor-made'' property was maintained and usable
in aqueous solution.
imprinted CT kept their imprinted property to catalyze
the N-Ac-d-Trp ethyl ester synthesis. After 96h the
yields of d-ester product produced by one gram of DIP
(
D)-CT and CLIP (D)-CT, respectively, were only
about 25% and 15% lower compared to NIP (D)-CT.
For the CLIP (D)-CT mass transfer e€ects were recog-
nizable. Nevertheless, its reaction rate was almost linear
up to 96h. The average particle size of the CLIP (D)-CT
was determined to be about 10 mm (Fig. 3) and the spe-
The operational stability of CLIP (D)-CT was tested in
4 consecutive cycles over a total reaction time of 408 h.
No decrease in speci®c activity was observed during that
time. Furthermore, when this particular preparation of
CLIP (D)-CT was reused for d-ester synthesis in cyclo-
hexane afterwards (after extensively washing and drying
under vacuum) no loss in speci®c activity was observed
after 96h in comparison to a CLIP (D)-CT that was not
incubated in aqueous solution before.
2
ci®c surface area was about 5.5 m /g. Figure 4 presents
a picture of CLIP (D)-CT.
In addition, active site titration of the di€erent CT spe-
cies was done with N-trans-cinnamoyl imidazole. The
value obtained for native CT was set to 100%. In com-
parison, NIP (D)-CT and CLIP (D)-CT resulted in 85
and 75% of accessible catalytically active sites. Thus,
the derivatization and crosslinking procedures seemed
to be quite gentle.
Kinetic data of native CT, imprinted native CT and
imprinted crosslinked CT
For comparison, imprinting of native CT and deriva-
tized CT was done with N-Ac-d-Trp or N-Ac-l-Trp
prior to precipitation. The derivatized CTs were cross-
linked in the same manner as described above. Then, the
various CT preparations were subjected to N-Ac-d-Trp
ester syntheses (Table 1a) or hydrolyses (Table 1b),
respectively. Concerning the ``natural'' l-ester synthesis
Maintenance of the imprinted property of proteases in
aqueous solution
Subtilisin Carlsberg (SUB) was treated in the same
manner as CT and the two CLIP D-proteases were able
to hydrolyze N-Ac-d-Trp ethyl ester in bu€er solution
(
Fig. 5). Neither the native enzymes nor the NIP (D)-
reaction (Table 1a), the highest V
was measured with
max
CT/SUB nor the DIP (D)-CT/SUB enhanced the slow
spontaneous hydrolysis of the substrate, that was cal-
culated to be 1.9 10 min at pH 7.8 (see Fig. 5).
the precipitated, non-imprinted CT. Imprinting of CT
with N-Ac-l-Trp increased the anity towards the sub-
�
5
� 1
strate [K 8.4 mM of NIP (L)-CT)] but decreased V
M
.
max

2
234
F. Peiûker, L. Fischer / Bioorg. Med. Chem. 7 (1999) 2231±2237
Figure 3. Distribution curve resulting from particle size analysis of CLIP (d)-CT.
any other preparation of CT because none of them
accepts the d-ester as a substrate. For that reason, the
and k values are mentioned here and the ratio
k
enz
nonenz
4
� 1
.
kenz/knonenz results in 5.3 10 mM
Discussion
The main interest of this work was to stabilize imprinted
protein properties, which were up to now stable only in
3
organic solvents, also in aqueous surroundings. The
results of the experiments demonstrated clearly that
crosslinking the acylated and imprinted enzymes with
ethylene glycol dimethacrylate in an organic solvent
prevented the two proteases from losing their ligand
induced d-amino acid acceptance in the aqueous
medium.
Figure 4. SEM picture of CLIP (magni®cation 20,000x).
Since lyophilization itself causes unwanted structural
changes on proteins,1
1,15
4
alcohol precipitation seems to
The CLIP (L)-CT had the lowest KM (3.7 mM) and
Vmax. The latter may be rationalized by a large mass
transfer impact of the particles that is also recognized
by comparing the Vmax values of NIP (D)-CT (100%)
and CLIP (D)-CT (20%) for d-ester synthesis.
be the more advantageous method to dehydrate pro-
teins for imprinting purposes. The vinylation of the
proteases under controlled bu€er conditions was inten-
ded to introduce selective groups for subsequently
crosslinking in organic solvents and thus, not further
disturbing the intramolecular interactions of the protein
keeping its manipulated property. By using itaconic
In the case of the l-ester hydrolysis (Table 1b), native
non-imprinted CT had a high anity towards the sub-
strate (KM 0.08 mM). Crosslinked non-imprinted CT as
well as CLIP (L)-CT possessed an about 100-fold lower
anity and astonishingly, the two were identical
according to KM and Vmax. However, only CLIP (D)-
CT was able to hydrolyze N-Ac-d-Trp ethyl ester and,
1
4
anhydride for this tight amide bonds were obtained,
14
and in the case of the b-glucosidase from almonds
nearly 100% of activity was retained after a complete
derivatization of the enzyme as measured by TNBS
method. However, also other vinylization agents, for
1
6
instance acryloyl chloride, could be used in principle.
its K versus the d-enantiomer (7.3 mM) was lower
The amount of crosslinker used was high, taking into
account the experiences in the ®eld of synthetic imprin-
ted polymers.17 This caused mass transfer limitations
M
compared to the CLIP (L)-CT versus the l-enantiomer
(
12.8 mM). The Vmax cannot be directly compared to

F. Peiûker, L. Fischer / Bioorg. Med. Chem. 7 (1999) 2231±2237
2235
Figure 5. Hydrolysis of N-acetyl-d-tryptophan ethyl ester catalyzed by CLIP (d-) proteases [CLIP (d-)-CT/-SUB=crosslinked imprinted proteins
(
absence as well as in the presence of the polymer containing the non-imprinted acylated proteases. Reaction conditions: 10 mg of each biocatalyst
a-chymotrypsin/subtilisin) that were imprinted with N-acetyl-d-tryptophan] in comparison to the nonenzymatic hydrolysis of the d-ester in the
ꢀ
preparation, N-acetyl-d-tryptophan 0.03 M, 0.05 M phosphate bu€er pH 7.8, 25 C; thermomixer 1200 rpm.
Table 1a. Kinetic data of the synthesis of N-acetyl-l-tryptophan
ethyl ester (l-ee) and N-acetyl-d-tryptophan ethyl ester (d-ee) cata-
lyzed by precipitated a-chymotrypsin (CT), NIP-CT (native, imprinted
protein [a-chymotrypsin] and CLIP-CT (crosslinked imprinted protein
Table 1b. Kinetic data of the hydrolysis of N-acetyl-l-tryptophan
ethyl ester (l-ee) and N-Acetyl-d-tryptophan ethyl ester (d-ee) cata-
lyzed by CT (a-chymotrypsin) and CLIP-CT (crosslinked imprinted
5
4
protein [a-chymotrypsin] [100% activity=9.3Â10 nkat/g (5.6Â10
[
a-chymotrypsin]
units/g)]
a
a
Reaction
Biocatalyst preparation
K
M
[mM]
V
max [%]
100
Reaction
Biocatalyst
preparation
K
M
[mM]
V
max [%]
l-ee synthesis
Native a-CT,
precipitated
19.5
l-ee hydrolysis
Native
Crosslinked,
0.08
12.8
100
12.6
b
NIP (l-) -CT
8.4
3.7
63
21
b
CLIP (l-) -CT
non-imprinted CT
b
CLIP (L-) -CT
12.8
7.3
12.6
b
d-ee synthesis
NIP (d-) -CT
16.1
25.1
100
20
b
b
� 1
� 1
CLIP (d-) -CT
d-ee hydrolysis
CLIP (D-) -CT
1.02 mM min
(
k
enz
� 5
)
a
� 1
K
M
of the corresponding enantiomer of N-acetyl-Trp.
The appendix (l-) means that the biocatalyst was imprinted with
N-acetyl-l-tryptophan and (d-) means the imprinting with the corre-
sponding d-enantiomer.
Spontaneous
non-enzymatic
hydrolysis
1.9 10 min
b
(knonenz
)
a,b
See Table 1a.
and the rather low speci®c surface area of the polymer.
Whether this high concentration of crosslinker is really
necessary to maintain an imprinted property of proteins
has to be investigated in the future. Also a comparison
of the catalytic eciency of optimized CLIPs with
``natural'' l-ester hydrolysis, the latter was about 1000-
fold higher. However, this has to be judged carefully
since the CLIP (D)-CT preparation is just a very ®rst
example for the stabilization of an imprinted property
by crosslinking. The fact that the ``new'' substrate
acceptance of two evolutionary convergent proteases
CLIP (D)-CT and CLIP (D)-SUB lead to kenz/knonenz
values for the enzymatic d-ester hydrolysis in the range
18,19
the glutaraldehyde crosslinked enzyme crystals
CLECs) would be interesting since both methods make
(
use of crosslinking enzymes in their conformational
stable forms (CLIPs: enzymes are ``rigid'' due to non-
aqueous environment; CLECs: enzymes are ``rigid'' due
to crystallization).
4
5
� 1
of 10 ±10 mM fold remains impressive. The data are
already in the same order of magnitude as with the
majority of catalytic antibodies.20
If the d-ester hydrolysis activity of CLIP (D)-CT is com-
pared with the activity of the free native CT catalyzing the
Though this was only a ®rst step to make the use of
imprinted proteins available and stable in aqueous

2
236
F. Peiûker, L. Fischer / Bioorg. Med. Chem. 7 (1999) 2231±2237
Enzyme vinylization. For functional group acylation
with itaconic anhydride of about 60%, 50 mg of a-chy-
motrypsin/subtilisin were dissolved in 10 mL of potas-
sium phosphate bu€er (0.01 M; pH 7.8). 120 mg of
itaconic anhydride were added slowly in small portions
to the stirred enzyme solution. The pH value was
maintained by titration with a NaOH solution (5 M).
After all of the anhydride had been added the stirring
was prolonged for 30 min. Gel®ltration using PD-10
columns (Pharmacia) with H O dist. as the eluent was
2
applied in order to cut o€ low-molecular compounds
from the enzyme solution. The eluents were combined
and ®nally lyophilized. The extent of enzyme acylation
was determined with 2,4,6,trinitrobenzenesulfonic acid
(
TNBS) as described elsewhere.14 Since the TNBS assay
is not selective for a particular protein amino acid
composition, it was assumed that the obtained data
correspond to the statistical average of all proteins in
the sample and thus for CT or SUB.
Enzyme imprinting. Typically, 30 mg of a-chymotrypsin/
subtilisin were dissolved in 1 mL potassium phosphate
bu€er (0.01 M, pH 7.8) containing N-acetyl-d-trypto-
phan or N-acetyl-l-tryptophan (0.02 M). The solution
ꢀ
was cooled to 0 C for 30 min, 4 mL of 1-propanol were
ꢀ
added at a temperature of � 20 C and the mixture was
mixed vigorously. After an additional 30 min the
resulting precipitate was centrifuged and the super-
natant was decanted. The precipitate was washed 3
times with each 3 mL of 1-propanol and ®nally dried
under vacuum.
Scheme 1. Broadening the substrate selectivity through the combina-
tion of bioimprinting and a covalent immobilization technique.
phases, the principle of the method seems to allow the
tailoring of biocatalysts by biochemical means (Scheme
Synthesis of N-acetyl-D-/N-acetyl-L-tryptophan ethyl es-
ter. In general 10 mg of each enzyme preparation (NIP,
DIP, CLIP) were suspended in 1 mL dry cyclohexane
puriss. in 1.5 mL organic solvent resistant tubes (Soer-
ensen, Bioscience Inc.). After the suspension was soni-
®ed with a Sonorex DK 514 B for 5 min the reaction
was started upon addition of 250 mL dry ethanol puriss.
containing N-acetyl-d-tryptophan or the corresponding
2
1
1). Its degree of freedom concerning the amounts and
kinds of vinylization agents and crosslinking agents, the
conditions for derivatization and crosslinking (e.g. add-
ing stabilizing additives in each case) is promising to
achieve higher speci®c activities towards creating
imprinted ``new'' enzyme properties in the future.
l-enantiomer in a concentration of 0.1 mol/L and 10 mL
ꢀ
(Eppendorf 5436) at 1200 rpm.
H O dist. at 25 C with the help of a thermomixer
2
Experimental
Chemicals. All chemicals used were commercially avail-
able and purchased in their highest grades. N-Acetyl-d-
tryptophan was purchased from Bachem (Bubendorf,
Switzerland, no. F-1090, lot 502198)
Ester analysis. The rate of esteri®cation was measured
by determining the amount of N-acetyl-d-tryptophan-
or N-acetyl-l-tryptophan ethyl ester formed during the
reaction at certain intervals. A 50 mL aliquot of the
reaction mixture, from which the enzyme had been
removed by centrifugation, was diluted 1/40 and injected
in a HPLC RP-18 column (LiChroCart 125-4, Merck).
The elution was done with water:acetonitrile:acetic acid
(65:30:5 v/v/v) at a ¯owrate of 1 mL/min. The d-, and
respectively, the l-con®gurated ester were detected at
275 nm. The calibration was done with the pure N-acetyl
amino acids or their ethyl esters as external standards.
All analyses were run at least in duplicates. The resulting
HPLC integrals deviated less than 8% from each other.
The optical purity of the product was con®rmed by
measuring the speci®c rotation with a polarimeter.
Enzymes. a-Chymotrypsin type II from bovine pancreas
(
EC 3.4.21.1) was purchased from SIGMA Chemical
Co. (St. Louis, MO, no. C 4129) as a dialyzed, essen-
tially salt-free, lyophilized powder. The enzyme pre-
paration was 3Â crystallized from 4Â crystallized
chymotrypsinogen (lot 91H7195). Subtilisin Carlsberg
type VIII from Bacillus licheniformis (EC 3.4.21.62) was
also ordered from SIGMA (No. P 5380, lot 46H0141) as
a crystallized and lyophilized preparation.
ꢀ-chymotrypsin/subtilisin assay. a-Chymotrypsin/sub-
tilisin activity was determined following the hydro-
lysis of benzoyl-l-tyrosin ethyl ester as described
elsewhere.
Crosslinking of DIP. Ten milligrams of DIP were
suspended in 0.5 mL cyclohexane and soni®ed until
2
2,23

F. Peiûker, L. Fischer / Bioorg. Med. Chem. 7 (1999) 2231±2237
2237
homogeneity. To this suspension 0.23 mL of ethylene
glycol dimethacrylate (EDMA) and 4 mg of AIBN were
added. The radical copolymerization was initiated by
Measurement of the CLIP's speci®c surface area. The
speci®c surface area was determined according to the
BET1
(Mikromeritics, Georgia, USA).
3,25
method using a Mikromeritics DeSorb 2300 A
ꢀ
UV irradiation at 366 nm at 4 C. After 5 h the resulting
polymer underwent three consecutive washing steps
each with 0.5 mL of cyclohexane. Finally the cross-
linked imprinted biocatalyst was dried under vacuum.
Determination of the CLIP's average particle size (APS).
The average particle size was measured with a Sympatec
HELOS particle analyzer (Sympatec, Germany).
Hydrolysis of N-acetyl-D-tryptophan ethyl ester. Ten
milligrams of CLIP was suspended into 1 mL of potas-
sium phosphate bu€er (0.01 M; pH 7.8), containing
N-acetyl-d-tryptophan ethyl ester (0.03 M). The reaction
Scanning electron microscopy (SEM) of CLIP. SEM
pictures of CLIP (d)-CT were taken with a Zeiss DSM
982 Gemini after the samples had been sputtered with
Au.
ꢀ
mixture was incubated at 25 C at 1200 rpm with the
help of a thermomixer. At certain intervals 50 mL ali-
quots were taken, centrifuged and the supernatant was
diluted 1/40 prior to HPLC analysis (s. ester analysis).
To exclude any nonspeci®c, non-enzymatical hydrolysis,
this reaction was conducted with the polymer alone, the
substrate alone and the polymer, containing non-
imprinted enzyme.
Acknowledgement
L. Fischer thanks the FCI (Fonds der Chemischen
Industrie) for partial ®nancial support of his research.
References
Operational stability of CLIP (D-)-ꢀ-CT. After the
initial 120 h of CLIP (d-)-a-CT catalyzed d-ester
hydrolysis, CLIP (D-)-a-CT was centrifuged and wash-
ing steps using potassium phosphate bu€er were applied
until there was no longer any substrate/product detect-
able in the washing solutions. After drying the CLIP
1
2
5
3
4
. Perona, J. J.; Craik, C. S. Protein Sci. 1995, 4, 337.
. Shao, P.; Arnold, F. H. Curr. Opin. Struct. Biol. 1996, 6,
13.
. Klibanov, A. M. Nature 1995, 374, 596.
Ê Ê
. Stahl, M.; Jeppsson-Wistrand, U.; Mansson, M.-O.;
Mosbach, K. J. Am. Chem. Soc. 1991, 113, 9366.
. Gonzales-Navarro, H.; Braco, L. Biotechnol. Bioengin.
1998, 59, 122.
6. Mosbach, K.; Ramstro
7. Rich, J. O.; Dordick, J. S. J. Am. Chem. Soc. 1997, 119,
(d-)-a-CT under vacuum ``fresh'' substrate solution was
5
Â
added and again the hydrolysis was conducted for
another 96 h. This procedure was done 3 times, so that
the total reaction time amounted to 408 h (17 days).
È
m, O. Bio/Technology 1996, 14, 163.
3
8
245.
. Johansson, A.; Mosbach, K.; Ma
Ê
nsson, M.-O. Eur. J. Bio-
Reuse of CLIP (D-)-ꢀ-CT and CLIP (D-)-SUB. After
the d-ester hydrolysis took place, the substrate and the
product were removed with dry cyclohexane as the
washing solution until these compounds were no longer
detectable by HPLC analysis. Then CLIP (d-)-a-CT/
CLIP (d)-SUB was dried under vacuum and the synth-
esis of N-acetyl-d-tryptophan ethyl ester was conducted
again (see above).
chem. 1995, 227, 551.
9
1
1
2
1
. Dabulis, K.; Klibanov, A. M. Biotech. Bioeng. 1992, 39,
76.
0. Slade, C. J.; Vulfson, E. N. Biotechnol. Bioengin. 1998, 57,
11.
1. Mishra, P.; Griebenow, K.; Klibanov, A. M. Biotechnol.
Bioengin 1996, 52, 609.
12. Halling, P. J. Enzyme Microb. Technol. 1994, 16, 178.
1
1
4
1
È
3. Haul, R.; Dumbgen, G. Chem.-Ing.-Tech. 1960, 32, 349.
4. Fischer, L.; Peiûker, F. Appl. Microbiol. Biotechnol. 1998,
9, 129.
5. Prestrelski, S. J.; Arakawa, T.; Carpenter, J. F. Biophys. J.
Determination of the kinetic parameters Vmax and KM.
The kinetic constants were determined following the
initial velocities of the hydrolysis/synthesis reactions at
ꢀ
1993, 65, 661.
6. Wang, P.; Sergeeva, M. V.; Lim, L.; Dordick, J. S. Nature
Biotechnol. 1997, 15, 789.
7. Fischer, L.; Muller, R.; Ekbert, B.; Mosbach, K. J. Am.
25 C with increasing substrate concentrations up to
1
substrate saturation. The resulting hyperbolic satura-
tion curves were linearized according to Hanes. All
related transformations were conducted independently
of each other in triplicates.
1
È
Chem. Soc. 1991, 113, 9358.
18. Margolin, A. L. Trends Biotechnol. 1996, 14, 223.
19. Lalonde, J. J.; Govardhan, C.; Khalaf, N.; Martinez, A.
G.; Visuri, K.; Margolin, A. L. J. Am. Chem. Soc. 1995, 117,
Calculation of reaction acceleration. The reaction accel-
eration was calculated by comparing two second-order
rate constants, knonenz for the non-enzymatic reaction
6
2
2
845.
0. Kirby, A. J. Acta Chem. Scand. 1996, 50, 203.
1. Fischer, L.; Peiûker, F. 1996, Patent, DE19627162 Cl.
{
matic reaction {venz=(kcat
vnonenz=knonenz [S] [Solvent] } and k
for the enzy-
enz
�
K ) [S] [E] }.
M
1
22. Hummel, B. C. W. Can. J. Biochem. Physiol. 1959, 37,
.
1
2
393.
3. Worthington Enzyme Manual; Worthington Biochem.
Active-site titration. The accessibility of low-molecular
compounds to the active-site of a-chymotrypsin was
determined by titrating the enzyme with N-trans-cin-
Corp.; Freehold: New Jersey, 1972; pp 129±131.
4. Schonbaum, G. R.; Zerner, B.; Bender, M. L. J. Biol.
Chem. 1961, 236, 2930.
25. Haul, R.; Dumbgen, G. Chem. Ing. Tech. 1963, 35, 349.
2
2
4
nyamoylimidazol as described elsewhere.
È