Determination of the Enantioselectivity for Kinetically Controlled Condensations Catalysed by Amidases and Peptidases

Article abstract of DOI:10.1002/1615-4169(200212)344:10<1115::AID-ADSC1115>3.0.CO;2-D

For kinetically controlled synthetic reactions catalysed by amidases and peptidases, the enantio- or stereoselectivity determined from initial rates with racemic mixtures (E_{syn, rac}) was found to differ from the enantioselectivity determined from measurements with isolated enantiomeric nucleophiles (E_syn). This was observed for kinetically controlled condensation of R-phenyglycineamide with S-, R- and racemic Phe and S-, R- and racemic Leu catalysed by penicillin amidase from E. coli and for kinetically controlled condensation of N^α-acetyl-S-tyrosine ethyl ester with S-, R- and racemic AlaNH₂ catalysed by bovine α-chymotrypsin. It is shown that only E_{syn, rac} determined with racemic nucleophiles is an intrinsic enzyme property which should be used to study the influence of the primary structure, physicochemical parameters and immobilisation on biocatalyst enantioselectivity in kinetically controlled synthetic reactions catalysed by these enzymes.

Full text of DOI:10.1002/1615-4169(200212)344:10<1115::AID-ADSC1115>3.0.CO;2-D

FULL PAPERS
Determination of the Enantioselectivity for Kinetically
Controlled Condensations Catalysed by Amidases and Peptidases
Boris Galunsky, Volker Kasche*
Biotechnology II, Technical University Hamburg-Harburg, Denickestr. 15, 21071 Hamburg, Germany
Fax: ( ‡ 49)-40-42878-2127, e-mail: kasche@tuhh.de
Received: May 15, 2002; Accepted: August 2, 2002
Dedicated to Professor R. Sheldon on the occasion of his 60th birthday
Abstract: For kinetically controlled synthetic reac- catalysed by bovine a-chymotrypsin. It is shown that
tions catalysed by amidases and peptidases, the only Esyn, rac determined with racemic nucleophiles is
enantio- or stereoselectivity determined from initial an intrinsic enzyme property which should be used to
rates with racemic mixtures (Esyn, rac) was found to study the influence of the primary structure, phys-
differ from the enantioselectivity determined from icochemical parameters and immobilisation on bio-
measurements with isolated enantiomeric nucleo- catalyst enantioselectivity in kinetically controlled
philes (Esyn). This was observed for kinetically con- synthetic reactions catalysed by these enzymes.
trolled condensation of R-phenyglycineamide with
S-, R- and racemic Phe and S-, R- and racemic Leu
catalysed by penicillin amidase from E. coli and for Keywords: a-chymotrypsin; enantioselectivity; kineti-
kinetically controlled condensation of Na-acetyl-S- cally controlled synthesis; kinetic resolution; penicil-
tyrosine ethyl ester with S-, R- and racemic AlaNH2 lin amidase; stereoselectivity
Introduction
ignated S1 for the residue with the carboxyl group and S×1
for the residue with the amino group involved in the
Biocatalytic production of pure chiral substances is an formation of the amide (peptide) bond.[4] Thus, quanti-
expanding area in the pharmaceutical and fine chemical tative measures for the enantioselectivity of the differ-
industry.[1] A large number of these substances (b- ent binding sub-sites of an enzyme are required for the
lactam antibiotics, biologically active peptides, oligo- selection of the optimal enzyme for a specific racemate
saccharides, fatty acid esters, oligonucleotides) can be resolution. This has mainly been done for equilibrium
synthesised in kinetically and equilibrium controlled controlled enzyme-catalysed reactions such as hydrol-
reactions catalysed by hydrolytic enzymes. The max- ysis by hydrolases. For these it has been shown that Ehyd
imum yield of product is reached faster and is larger in equals the ratio of the specificity constants for the
the kinetically than in equilibrium controlled synthesis. enantiomeric substrates,[5]
Contrary to the latter, the yield in the kinetically
controlled synthesis depends also on the kinetic proper-
ties of the enzyme.[2] When these processes are used for
kinetic racemate resolutions the steric purity of the
ꢀ2†
products depends on the enantio- or stereoselectivity of
the used hydrolase. It is quantified by the ratio of the i.e., it reflects an intrinsic enzyme property. Ehyd can be
rate constants for the competing enzymatic transforma- determined with isolated enantiomers or racemic
tions of the enantiomeric substrates.[3]
mixtures from initial rate measurements.[6] When
determined with isolated enantiomers in initial rate
(1) studies Ehyd gives fundamental information on how
binding and activation energies are utilised by the
E ˆ kS /kR
To obtain high steric purity of the products E should be enzyme to discriminate between the enantiomers and
> 100 for an S-specific enzyme or <0.01for an R- allows one to analyse the influence of primary structure,
specific enzyme. An enzyme generally has different temperature and pH on biocatalyst stereoselectivity.[7]
binding sub-sites adjacent to the bond changed during These studies also showed that the stereoselectivity of
the enzyme-catalysed reaction. For hydrolases such as the different binding sub-sites of amidases and pepti-
amidases and peptidases they are conventionally des- dases may differ considerably. Penicillin amidases have
Adv. Synth. Catal. 2002, 344, No. 10
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 1615-4150/02/34410-1115 1119 $ 17.50+.50/0
1115

Boris Galunsky, Volker Kasche
FULL PAPERS
an R-specific S1- and an S-specific S×1-binding sub-site.
Whether Esyn or Esyn, rac are intrinsic properties of the
Ehyd determined from initial rates cannot, however, enzyme can be analysed based on the following
generally be used for practical racemate resolutions. In mechanism for the kinetically controlled synthesis that
racemic mixtures each enantiomer is a competitive has been shown to apply for amidases, peptidases and
inhibitor for the other.[8] For determinations of the glucosidases with covalent acyl- or glycosyl-enzyme
enantioselectivity with racemic mixtures mainly prog- intermediates.[2]
ress curves have been analysed using integrated rate
equations.[6,9] These studies indicated that enzyme
enantioselectivity can be modulated by product inhib-
ition.[9]
For kinetically controlled synthesis with amidases and
peptidases the apparent ratios of the transferase to
hydrolase rate constants, used to quantitate the nucle-
ophile specificity, determined for the S- and R-enantio-
meric nucleophiles have been used to evaluate the
stereoselectivity of the S×1-binding sub-site.[10] It has not
yet been analysed whether this is a suitable quantitative
Scheme 1. Mechanism for the hydrolase-catalysed synthesis
of a condensation product AN from an activated substrate
AB and a racemic nucleophile NH. The mechanism implies a
covalent acyl-enzyme intermediate and nucleophile binding
to the acyl-enzyme before its deacylation.[2] The mechanism
is simplified neglecting nucleophile binding to EH and
EH ¥¥¥ AB.
measure for the enantioselectivity in kinetically con-
trolled synthesis. This is done here, where enantioselec-
tivities were determined for both isolated enantiomers
and racemic mixtures.
Results and Discussion
Assuming equilibrium in NH binding and negligible
AN hydrolysis, (kT/kH)app for an isolated enantiomer
determined from the initial rates of formation of con-
Enantioselectivity Measures in Kinetically Controlled
Synthesis
To quantify the S1×- or nucleophile enantioselectivity in densation and hydrolysis products is given by the relation[2]
kinetically controlled synthesis catalysed by amidases or
peptidases the following ratio has been used[10]
ꢀ6†
ꢀ3†
(which also covers the case where the acyl-enzyme can
The ratios of apparent second-order rate constants
for acyl transfer to nucleophile and water were deter-
mined from initial rate measurements with isolated
enantiomers
be formed from the enzyme-substrate complex with
bound nucleophile) where KN is the equilibrium
constant for nucleophile binding to the acyl-enzyme, kt
is the deacylation rate constant of the acyl-enzyme-
nucleophile complex, kh is the deacylation rate constant
of the acyl-enzyme by water, and kh,N is the deacylation
rate constant of the acyl-enzyme-nucleophile complex
by water. Then E is
ꢀ4†
where vT and vH are the initial rates for formation of the
condensation and hydrolysis products respectively.[2]
From the definition (Equation 1) and considering equal
initial concentrations of the enantiomeric nucleophiles,
E for a racemic mixture
ꢀ7†
i.e., it is not an intrinsic property of the enzyme as it
depends on the nucleophile concentration. When using
racemic nucleophiles for determination of E both
nucleophiles compete for the acyl-enzyme and from
Scheme 1can be derived
ꢀ5†
Thus, only when Esyn ˆ Esyn, rac can the enantioselectivity
be determined from measurements with isolated enan-
tiomers.
ꢀ8†
1116
Adv. Synth. Catal. 2002, 344, 1115 1119

Determination of the Enantioselectivity for Kinetically Controlled Condensations
FULL PAPERS
that it is an intrinsic property of the enzyme. From For the PA-catalysed transformations the observed
Equations (7) and (8) it follows that Esyn ˆ Esyn, rac only difference is about an order of magnitude. When
when kh,N,R ˆ kh,N,S ˆ 0, i.e., the acyl-enzyme cannot be racemic Phe was used as a nucleophile only one
deacylated by water when it has bound a nucleophile. diastereomeric product (R-S) was observed. These
When this is not the case, only the enantioselectivity results show that Esyn=Esyn, rac and that the acyl-
E
syn, rac determined with a racemic mixture is an intrinsic enzyme-nucleophile complex can be deacylated by
molecular property, i.e., independent on the nucleophile water. Therefore, only Esyn, rac can be used as a measure
concentration (Equation 8).
for enzyme S1×-enantioselectivity in kinetically control-
led synthesis. The Esyn, rac-values apply for initial rate
measurements. Of practical importance, however, is the
Experimental Determination of the Enantioselectivity enantiomeric or diastereomeric excess (ee, de) at the
in Kinetically Controlled Synthesis
kinetically controlled maximal concentration of the
desired product. Analytical expressions or numerical
The S1×-enantioselectivity in kinetically controlled solutions for this concentration have not yet been
synthetic reactions was determined for two hydrolytic developed. It was, however, found that the observed
enzymes: E. coli penicillin amidase (PA) and bovine a- diastereomeric excess (de) with racemic nucleophiles at
chymotrypsin (a-CT). The enantiomeric ratios were the kinetically controlled maximal concentration of the
determined in reactions with isolated enantiomeric and product with the more specific enantiomeric nucleo-
racemic nucleophiles. PA was used to catalyse the phile [ANmax R-S
corresponds to the enantioselectivity[11]
]
condensation of R-phenyglycineamide (R-PhgNH2) Esyn, rac derived from the initial rates vT,S and vT,R of the
with S-, R- and racemic Phe and with S-, R- and racemic same reaction. Thus, Esyn, rac is relevant to estimate the
Leu. a-CT was used as a catalyst for the kinetically enantiomeric or diastereomeric excess at the maximal
controlled condensation of Na-acetyl-S-tyrosine ethyl product concentration. For very high enantiomeric
ester (ATEE) and S-, R- and racemic AlaNH2. In both ratios, however, the difference between Esyn and Esyn, rac
cases the use of chiral activated substrate allowed becomes less important in the calculation of ee or de.
resolution of the diastereomeric products without
chiral chromatography.
The results obtained with the two enzymes are Conclusions
presented in Table 1. For all studied reactions the
enantioselectivity determined with isolated enantio- For the S1×-enantioselectivity in kinetically controlled
meric nucleophiles was considerably lower than the one condensation reactions catalysed by amidases and
determined with racemic nucleophiles. For the reactions peptidases only the enantiomeric ratio Esyn, deter-
rac
catalysed by a-chymotrypsin they differ by about 6-fold. mined with racemic mixtures of the nucleophile is an
Table 1. The S×1- enantioselectivities Esyn (Equation 7) and Esyn, rac (Equation 8) for penicillin amidase- and a-chymotrypsin-
catalysed kinetically controlled syntheses determined with enantiomeric and racemic nucleophiles.
[d]
Enzyme
Condensation reaction
(kT/kH )app
de (R-S),
AB ‡ NH
%
E. coli PA
R-PhgNH2 ‡ S-Phe[a]
10000
5
2000
200
R-PhgNH2 ‡ R-Phe[a]
R-PhgNH2 ‡ rac Phe[a]
R-PhgNH2 ‡ S-Leu[a]
>10000[e]
n.o.[e]
2900
>99.9
4000
20
R-PhgNH2 ‡ R-Leu[a]
R-PhgNH2 ‡ rac Leu[b]
3000
99.9
[a]
Bovine a-CT
ATEE ‡ S-Ala NH2
1550
60
26
[a]
ATEE ‡ R-Ala NH2
[c]
ATEE ‡ rac Ala NH2
160
163
99.0
[a]
10 mM AB, 100 mM NH.
20 mM AB, 100 mM NH.
10 mM AB, 200 mM NH.
[b]
[c]
[d]
[e]
Determined from initial rate measurements at less than 10% conversion of AB. Standard deviation less than 20%.
Only one diastereomer (R-S) was observed.
Adv. Synth. Catal. 2002, 344, 1115 1119
1117

Boris Galunsky, Volker Kasche
FULL PAPERS
 10À5 for PA and 1  10À6 2  10À6 for a-CT. Periodically
aliquots were withdrawn, diluted (1:10) and immediately
analysed by HPLC. Each experiment was paralleled by a
reference experiment without enzyme for a reading of the
spontaneous hydrolysis of the activated substrate which was
used to correct the input for calculation of kT/kH. The
transferase to hydrolase ratio was calculated from the initial
rates (less than 10% activated substrate exhausting) of
condensation (vT) and hydrolysis (vH) product formation
using Equation 4.[7c] The enantiomeric ratio for isolated
intrinsic molecular property (Equation 8). It should be
used as a measure:
for the stereoselectivity of different enzymes used as
biocatalysts for the same process, and
to study the influence of temperature, pressure, pH
and immobilisation on enantioselectivity in kinetically
controlled condensations with the same enzyme.[7]
The difference between Esyn, rac and Esyn (determined
with isolated enantiomers) implies that the acyl-enzyme
with bound nucleophile can be deacylated by water enantiomeric nucleophiles was calculated using Equation 3,
for racemic nucleophiles Equation 5 was used, respectively.
The reactions with racemic nucleophiles were followed up to
the maximal concentration of the specific diastereomeric
product.
(Scheme 1). This should also apply for other hydrolases
(such as glucosidases) that have been shown to catalyse
kinetically controlled synthetic reactions according to
Scheme 1.
HPLC Analysis
Experimental Section
Products and reactants were identified and analysed by HPLC
using a Pharmacia LKB 2249 solvent delivery system and LKB
Bromma 2151 variable wavelength detector with two sequen-
tially connected analytical columns. For the PA-catalysed
condensation between R-PhgNH2 and Phe the substances were
quantified using two RP-8, 5 mm, 10 cm (Merck) columns
thermostatted at 50 8C and calibration curves for the peak
areas at 225 nm obtained with reference compounds of known
concentrations. For all measurements the peak integration was
done with Chrom Star software 1994 (Bruker). The response
factor for the diastereomeric R-Phg-S-Phe and R-Phg-R-Phe
was inferred from the response factors for R-PhgNH2, R-Phg
and assuming a constant sum of the concentrations of the
activated substrate, condensation and hydrolysis products
during a reaction. The elution system for resolution of R-Phg,
R-PhgNH2, S- or R-Phe, and the diastereomeric R-Phg-S-Phe
and R-Phg-R-Phe was: 67 mM KH2PO4 (pH 4.5) for 3 minutes
followed by a step gradient methanol/67 mM KH2PO4 (pH 4.5)
(35:65 v/v) and elution for further 5 minutes. At flow rate
2 mL/min the retention times for R-Phg, R-PhgNH2, S- or R-
Phe, R-Phg-S-Phe and R-Phg-R-Phe were 0.9, 1.8, 2.4, 5.3 and
6.5, respectively, at good base-line resolution. For determina-
tion of kT/kH only the peaks for the hydrolysis product (R-Phg)
and the condensation product were integrated. For determi-
nation of vT,S/vT,R only the two diastereomeric products were
considered. For the PA-catalysed condensation between R-
PhgNH2 and Leu the substances were quantified using two RP-
18, 5 mm, 10 cm (Merck) columns thermostatted at 60 8C and
calibration curves for the peak areas at 225 nm obtained as
described above. Theelution system for resolution of R-Phg, R-
PhgNH2, S- or R-Leu and the diastereomeric R-Phg-S-Leu and
R-Phg-R-Leu was: methanol/67 mM KH2PO4 (pH 6.5) (10:90
v/v) for 2.5 minutes followed by a step gradient methanol/67
mM KH2PO4 pH 6.5 (35:65 v/v) and elution for further
7 minutes. At flow rate 2.2 mL/min the retention times for R-
Phg, S- or R-Leu, R-PhgNH2, R-Phg-R-Leu and R-Phg-S-Leu
were 1.2, 1.7, 2.9, 5.6 and 7.5, respectively, at good base-line
resolution. For the a-CT-catalysed condensation of ATEE and
AlaNH2 the substances were quantified using two RP-18, 5 mm,
10 cm (Merck) columns thermostatted at 60 8C and calibration
curves for the peak areas at 280 nm obtained with reference
compounds of known concentrations. For quantification of the
diastereomeric dipeptides, the calibration curve for ATEE was
General Remarks
R-PhgNH2 was kindly provided by Rˆhm (Darmstadt,
Germany). ATEE, R-phenyglycine (R-Phg), Na-acetyl-S-
tyrosine (AT), racemic and enantiomeric pure S- and R-Phe,
Leu and enantiomeric pure S- and R-AlaNH2 and R-Phg-Leu
were from Sigma and Bachem. Racemic AlaNH2 was prepared
by mixing equimolar amounts of S- and R-enantiomers. The
chiral purity of the enantiomeric substrates given by the
manufacturer was > 99%. All other chemicals were analytical
grade. Penicillin amidase (EC 3.5.1.11) from E. coli was
purchased from Sigma and purified by ion-exchange chroma-
tography using a Mono Q column HR10/10 (Pharmacia,
Sweden) as previously described.[12] Only the proteolytically
processed form with isoelectric point 7.0 was used. The purity
and the homogeneity of the enzyme was analysed by activity
and protein stains on isoelectric focusing gels.[12] The molar
concentration of PA was determined by active site titration
with phenymethanesulfonyl fluoride[13] using 6-nitro-3-phenyl-
acetamidobenzoic acid as a substrate. Bovine a-chymotrypsin
(EC 3.4.21.1) was purchased from Sigma and was purified by
affinity chromatography using soybean trypsin inhibitor
(Sigma) immobilised[14] on Eupergit C250-L (Rˆhm, Darm-
stadt). The purity of the enzyme was determined by isoelectric
focusing. The purified a-chymotrypsin was homogeneous and
its concentration was determined by the absorbance at 280 nm,
using an e280 value of 4.9 Â 104 MÀ1 cmÀ1.
Kinetically Controlled Condensations
In a typical experiment for enzyme-catalysed kinetically
controlled synthesis of diastereomeric dipeptides, 2 mL
reaction mixture containing 10 or 20 mM activated substrate
(R-PhgNH2 or ATEE) and 100 or 200 mM enantiomeric or
racemic nucleophile (Phe, Leu, AlaNH2; each concentration
ˆ
specified in Table 1) in bicarbonate buffer I 0.2 with pH
adjusted to 9.0 was preincubated at 25 8C. The reaction was
initiated by addition of 10 mL enzyme solution preincubated at
the same temperature in the same buffer. The final enzyme
concentration in the reaction mixture was (depending on the
nucleophile and the concentrations) in the range 1 Â 10À5
2
1118
Adv. Synth. Catal. 2002, 344, 1115 1119

Determination of the Enantioselectivity for Kinetically Controlled Condensations
FULL PAPERS
used since for the absorption at 280 nm the Tyr moiety is
dominant (the difference of the peak areas for 1mM ATEE
and 1mM ATis <1%). The elution system for resolution of AT,
ATEE and the diastereomeric Na-acetyl-S-Tyr-S-AlaNH2, and
Na-acetyl-S-Tyr-R-AlaNH2 (AlaNH2 was not detectable at
280 nm) was: methanol/5 mM KH2PO4 (pH 4.5) (6:94 v/v) for
5 minutes, followed by a step gradient methanol/5 mM
KH2PO4 (pH 4.5) (35:65 v/v) and elution for further
8 minutes. At flow rate 2 mL/min the retention times of AT,
Na-acetyl-S-Tyr-S-AlaNH2, Na-acetyl-S-Tyr-R-AlaNH2 and
ATEE were 1.8, 4.7, 5.8 and 11.9, respectively.
[6] A. J. J. Straathof, J. A. Jongejan, Enzyme Microb. Tech-
nol. 1997, 21, 559 571.
[7] a) R. S. Phillips, Trends Biotechnol. 1996, 13 16; b) V.
Kasche, B. Galunsky, A. Nurk, E. Piotraschke, A. Rieks,
Biotechnol. Lett. 1996, 18, 455 460; c) B. Galunsky, S.
Ignatova, V. Kasche, Biochim. Biophys. Acta 1997, 1343,
130 138; d) P. L. A. Overbeeke, J. Ottosson, K. Hult,
J. A. Jongejan, J. A. Duine, Biocatal. Biotransform. 1999,
17, 61 79; e) L. Lummer, A. Rieks, B. Galunsky, V.
Kasche, Biochim. Biophys. Acta 1999, 1433, 327 334.
[8] a) D. A. Mironenko, E. V. Kozlova, V. K. Svedas, V. A.
Solodenko, T. N. Kasheva, V. P. Kukhar, Biokhimia 1990,
55, 1124 1131; b) Z.-W. Guo, C. J. Sih, J. Am. Chem.
Soc. 1989, 111, 6836 6841.
References
[9] J. L. L. Rakels, B. Romein, A. J. J. Straathof, J. J. Heij-
nen, Biotechnol. Bioeng. 1994, 43, 411 422.
[1] a) R. A. Sheldon, Chirotechnology. Industrial Synthesis
of Optically Active Compounds, Marcel Dekker, New
York, 1993; b) A. L. Margolin, Enzyme Microb. Technol.
1993, 15, 266 280; c) S. C. Stinson, Chem. Eng. News
1994, 38 72; d) M. J. Van der Werf, W. J. J. van den
Tweel, J. Kamphuis, S. Hartmans, J. A. M. de Bont,
TIBTECH 1995, 12, 95 103.
[2] a) V. Kasche, Enzyme Microb. Technol. 1986, 8, 4 1 6;
b) V. Kasche, in Proteases a Practical Approach, (Eds.: J.
Bond, R. Beynon), 2nd edn., IRL-Press, Oxford, 2001,
pp. 265 292.
[3] C.-S. Chen, Y. Fujimoto, G. Girdaukas, C. J. Sih, J. Am.
Chem. Soc. 1982, 104, 7294 7299.
[4] I. Schechter, A. Berger, Biochem. Biophys. Res. Com-
mun. 1967, 27, 157 162.
[5] a) V. K. Svedas, I. Y. Galaev, Biochemistry (A translation
of Biokhimia) 1986, 51, 1283 1287; b) C.-S. Chen, C. J.
Sih, Angew. Chem. Int. Ed. Engl. 1989, 28, 695 707.
[10] a) H.-D. Jakubke, P. Kuhl, A. Kˆnnecke, Angew. Chem.
Int. Ed. Engl. 1985, 24, 85 93; b) L. Riechmann, V.
Kasche, Biochem. Biophys. Acta 1985, 830, 164 172;
c) V. Kasche, G. Michaelis, B. Galunsky, Biotechnol. Lett.
1991, 13, 75 80.
[11] Y. F. Wang, C.-S. Chen, G. Girdaukas, C. S. Sih, Enzymes
in Organic Synthesis, (Eds.: R. Porter, S. Klark), Pitman,
London, 1985, pp. 128 145.
[12] V. Kasche, U. Haufler, D. Markowski, A. Zeich, B.
Galunsky, Ann. N. Y. Acad. Sci. 1987, 501, 97 102.
[13] V. K. Svedas, A. L. Margolin, C. F. Sherestyuk, A. A.
Klesov, I. V. Berezin, Bioorg. Khim. 1977, 3, 545 553.
[14] N. Gottschlich, S. Weidgen, V. Kasche, J. Chromatogr. A
1996, 719, 267 274.
Adv. Synth. Catal. 2002, 344, 1115 1119
1119