D-Phenylglycine and D-4-hydroxyphenylglycine methyl esters via penicillin G acylase catalysed resolution in organic solvents

Article abstract of DOI:10.1016/S0957-4166(00)00129-4

Penicillin G acylase in organic solvents catalyses specifically the acylation of the L-enantiomers of methyl esters of phenylglycine and 4- hydroxyphenylglycine. Hydrolytic reactions are prevented by controlling the water activity of the system and no excess of acylating agent is required. The process leads to the facile isolation of the enantiomerically pure D- enantiomer, which is of practical use for the preparation of β-lactam antibiotics. Electrospray mass spectroscopy has been applied to the study of the enantioselectivity of the enzyme. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0957-4166(00)00129-4

Tetrahedron: Asymmetry 11 (2000) 1789±1796
d-Phenylglycine and d-4-hydroxyphenylglycine methyl
esters via penicillin G acylase catalysed resolution in
organic solvents
Alessandra Basso, Paolo Braiuca, Luigi De Martin, Cynthia Ebert, Lucia Gardossi*
and Paolo Linda
Dipartimento di Scienze Farmaceutiche, UniversitaÁ di Trieste, Piazzale Europa 1, 34127 Trieste, Italy
Received 18 February 2000; accepted 30 March 2000
Abstract
Penicillin G acylase in organic solvents catalyses speci®cally the acylation of the l-enantiomers of methyl
esters of phenylglycine and 4-hydroxyphenylglycine. Hydrolytic reactions are prevented by controlling the
water activity of the system and no excess of acylating agent is required. The process leads to the facile
isolation of the enantiomerically pure d-enantiomer, which is of practical use for the preparation of
b-lactam antibiotics. Electrospray mass spectroscopy has been applied to the study of the enantioselectivity
of the enzyme. # 2000 Elsevier Science Ltd. All rights reserved.
1
. Introduction
Optically active d-amino acids are widely used in the pharmaceutical industry as intermediates
for the production of semisynthetic antibiotics, pesticides and new drugs.1 Activated forms of
�5
d-4-hydroxyphenylglycine and d-phenylglycine are employed in the total enzymatic synthesis of
4,6,7
b-lactams following a kinetically controlled approach.
Penicillin G acylase (PGA) has been largely used in aqueous solutions to obtain optically active
compounds by means of phenylacetyl derivative hydrolysis of interesting compounds.8
�10
Moreover, Cole demonstrated, already in 1969, that the enzyme can be employed also in the
synthetic direction1
1a,b
and Zmijewski was the ®rst to use PGA for catalysing acylations in the
resolution of racemates.12 However, acylation in aqueous media requires excess of the acyl donor
component or extracting/precipitating the product2
synthesis. Recently Rosell et al. resolved racemic mixtures of phenylglycine using the synthetic
,12�15
to reverse the equilibrium towards
2
route in water by means of a stabilised derivative of PGA. The application of the synthetic route
to the resolution of phenylglycine methyl ester (PhGlyOMe) leads to the major advantage that
*
Corresponding author. Fax: +39-040-52572; e-mail: gardossi@univ.trieste.it
0957-4166/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(00)00129-4

1
790
A. Basso et al. / Tetrahedron: Asymmetry 11 (2000) 1789±1796
the useful d-enantiomer can be recovered unreacted, since PGA acylates preferentially the l-isomer.
On the contrary, the hydrolytic route in aqueous solution requires the preventive chemical deriv-
13,14
atisation of the racemate which, however, gives the l-enantiomer upon enzymatic hydrolysis.
In principle, partially replacing water with organic solvents should be bene®cial for acylation
reactions and also would translate into a higher solubility of phenylacetic derivatives.16
Unfortunately, water/solvent mixtures are not optimal media for PGA since high percentages of
17
co-solvents reduce the activity of the enzyme dramatically.
Water is essential for preserving the activity of PGA, nevertheless we have demonstrated
recently that native and immobilised PGA are active in di€erent pure organic solvents16,18,19 only
when the hydration of the enzyme is controlled. This can be achieved by maintaining the thermo-
dynamic parameter of water activity (aw)2 of the reaction system in the range 0.45±0.85.
0
16
The use of PGA in organic media at controlled a allows the acylation to be carried out while
w
avoiding competing hydrolytic reactions and improving the solubility of hydrophobic sub-
strates.16
Now we present the ®rst study concerning the enantioselectivity of PGA in pure organic
solvent. Two di€erent preparations of PGA were employed for the resolution of PhGlyOMe and
4
-hydroxyphenylglycine methyl ester (4-HO-PhGlyOMe) using the synthetic route in organic
19
media, avoiding any excess of acylating agent. This allows the facile production and isolation of
the enantiomerically pure d-enantiomer. Electrospray mass spectroscopy (ES-MS) has been used
to con®rm the enantioselectivity of the method.
2
. Results
Two preparations of PGA were demonstrated to be enantioselective towards the l-enantiomer
of PhGlyOMe. A racemic mixture of d,l-PhGlyOMe was ®rstly resolved by acylation catalysed
by a covalently immobilised form of penicillin G acylase (PGA-450 from Boehringer). PGA-450
(original water content=63% w/w) was previously partially dehydrated by suspending the enzyme
in petroleum ether and adding dry Celite R-640 rods. After this treatment the immobilised
1
18
catalyst suspended in toluene gave a values between 0.73 and 0.77.
w
Two equivalents of d,l-PhGlyOMe 1 reacted with an equivalent of methyl 4-hydroxy-
phenylacetate 2 giving, within 24 h, exclusively the corresponding l-amide 4 which precipitated
upon formation. At the end of the reaction the amide and the enzyme were removed by ®ltration
and the unreacted d-PhGlyOMe 3 was easily recovered from the organic solution by converting it
to the corresponding hydrochloride salt 5, that precipitated in the organic solvent (Scheme 1).
The enantioselectivity of the enzymatic process was determinated by polarimetric analysis of the
unreacted d-PhGlyOMe as hydrochloride 5 (ee%>98).
1
21
d,l-PhGlyOMe was also resolved using the native PGA adsorbed on Celite R-640 rods,
working in dichloromethane at a =0.85 and employing the same acylating agent 2. The unreacted
w
d-PhGlyOMe was recovered by precipitating the corresponding hydrochloride salt 5. The use of
the pure hydrophobic organic solvent instead of aqueous mixtures makes the leakage of the
enzyme negligible so that the biocatalyst is suitable to be recycled.
Amide 4 is more soluble in dichloromethane than in toluene and does not precipitate during
the process. However, dichloromethane led to unsatisfactory results in reactions catalysed by
PGA-450, since the catalyst ¯oats on the surface of the solvent, thus limiting the mass transfer. On
1
the other hand, toluene proved to be unsuitable for carrying out reactions catalysed by PGA/Celite ,

A. Basso et al. / Tetrahedron: Asymmetry 11 (2000) 1789±1796
1791
Scheme 1.
1
because the amide formed during the reaction precipitated and adsorbed on the Celite rods,
causing severe di€usion limitations, which translated into a dramatic decrease of the reaction rate
and eventually into a loss of selectivity of the process (ee%=75 by polarimetric analysis).
The enantioselectivity of the two enzymatic processes described here was con®rmed by
exploiting the high sensitivity of ES-MS. Namely, the d -methyl ester of the d-enantiomer was
3
synthesised and the enzymatic acylation was carried out on an equimolar mixture of l-PhGlyOMe
and d-PhGlyOCD (Scheme 2).
3
Scheme 2.
As shown in Fig. 1a, there is a signal corresponding to the amide 4 with an m/z=300 (M+1)+
whereas there is no signal corresponding to the deuterated amide (m/z=303). It must be noted
that we have veri®ed that ES-MS allows to detect the deuterated amide at concentrations lower
than 1%, and therefore it represents a very sensitive technique for analysing samples having high
enantiomeric excess.
ES-MS was also used to study the enantioselectivity of the PGA-450 catalysed resolution of
-HO-PhGlyOMe performed in toluene. Also in this case, ES-MS analysis of the crude reaction
4
mixture con®rmed the enantioselectivity of PGA towards the l-enantiomer (Fig. 1b). Despite the
fact that the poor solubility of substrates has often seemed as a restriction for enzymatic reactions
in organic solvents, we have been able to carry out the synthesis in hydrophobic solvent
employing a poorly soluble amine which was present in the reaction medium mainly as a
suspension.

1
792
A. Basso et al. / Tetrahedron: Asymmetry 11 (2000) 1789±1796
Figure 1. ES-MS of the amides enzymatically synthesised using ester 2 as acyl donor. (a) Product obtained starting
+
+
from l-PhGlyOMe and d-PhGlyOCD : m/z=300 (M+1) ; m/z=322 (M+Na) . (b) Product obtained from l-4-HO-
3
+
PhGlyOMe and d-4-HO-PhGlyOCD : m/z=316 (M+1) ; m/z=338 (M+Na)
3
A further proof of the selectivity of the enzyme came from the kinetic study of the acylation of
each of the enantiomer of the amines taken separately. As shown in Table 1, l-PhGlyOMe was
acylated with an initial rate 18 times higher than the d-enantiomer, whereas, in the case of the
4
-HO-PhGlyOMe, no reaction at all was observed after 48 h using the d-enantiomer.
3
. Conclusions
PGA in toluene and in dichloromethane catalyses the speci®c acylation of the l-enantiomers of
the methyl esters of phenylglycine and 4-hydroxyphenylglycine, leading to the facile recovery of
the d-enantiomer which is used in the preparation of b-lactam antibiotics. Electrospray mass
spectroscopy has proved to be a valuable tool to con®rm the enantioselectivity of the reactions.
This study demonstrates for the ®rst time, to the best of our knowledge, the enantioselectivity
of PGA in pure organic media and opens new perspectives for practical applications of this
enzyme to asymmetric synthesis of valuable compounds.

A. Basso et al. / Tetrahedron: Asymmetry 11 (2000) 1789±1796
1793
Table 1
0
Initial rates (v ) of acylation catalysed by PGA-450 in toluene
4
. Experimental
4
.1. Materials
PGA-450 was a generous gift of Boehringer-Mannheim. It consists of penicillin G acylase from
E. coli covalently immobilised on a polymer the chemical nature of which is not disclosed by the
producer. PGA-450 has a water content of 62.3% and it was partially dehydrated before use with
1
the aid of Celite R-640 according to the following procedure: 2 g of enzyme were added to 100
17
1
mL of petroleum ether containing 4 g of dry Celite R-640. After 15 days at 4 C the Celite was
ꢀ
1
ꢀ
removed and PGA was stored in the same solvent at 4 C. The ®nal water content of the enzyme
was 27% (determined by Karl Fischer titration±Mettler) and its activity was 401 U/g. Batches
were prepared on a gram scale and there was no decrease in activity over at least four months. As
required, enzymatic samples were withdrawn and the volatile organic solvent used for the storage
was removed from the enzymatic sample at room temperature and atmospheric pressure without
causing any detrimental e€ect to the catalyst.
Penicillin G acylase from E. coli (EC 3.5.1.11) was purchased from Fluka as an aqueous
suspension that was stored as lyophilised powder (12.3 U/mg).
1
Celite R-640 was from Fluka. Toluene and dichloromethane used in enzymatic reactions were
Ê
dried over molecular sieves (4 A).
Methyl 4-hydroxyphenylacetate, d- and l-phenylglycine methyl ester hydrochloride and cinnamyl
alcohol were from Aldrich. d- and l-4-Hydroxyphenylglycine (4-HO-PhGlyOH) and d-phenyl-
glycine were from Fluka. Deuterated methanol (CD OD) was from Trimital.
3
4
.1.1. Preparation of d- and l-PhGlyOMe free bases
Hydrochloride salt (20 mmol) was suspended in 20 mL of dry dichloromethane together with
.
0 mmol of Na CO 10H O. After 15 min of magnetic stirring the organic solution was ®ltered
2 3 2
3
and dried over anhydrous sodium sulphate. Finally, the solvent was removed under vacuum to
1
obtain the free amines (yields: 71 and 73%, respectively) that were characterised by H NMR.
1
d- and l-PhGlyOMe: H NMR (200 MHz, CDCl ): 2.05 (s, 2H, NH ); 3.51 (s, 3H, CH O);
.58 (s, 1H, CHPh); 7.31 (m, 5H, Ph).
3
2
3
4

1
794
A. Basso et al. / Tetrahedron: Asymmetry 11 (2000) 1789±1796
4
.1.2. Synthesis of d- and l-4-HO-PhGlyOMe, d-4-HO-PhGlyOCD and d-PhGlyOCD3
3
1
0 mmol of d- or l-4-HO-PhGlyOH, 100 mmol of dry methanol and 0.5 mL of concentrated
sulphuric acid were re¯uxed for 3 h until the suspension converted to a clear solution that was
concentrated under vacuum to remove the methanol in excess. The residue was dissolved in 10
mL of water and the pH was adjusted to 8 using a NaOH solution (0.1N), obtaining the ester as a
1
13
white precipitate that was ®ltered, rinsed with dichloromethane and characterised by H and
NMR.
C
d-4-HO-PhGlyOCD and d-PhGlyOCD were synthesised according to the same procedure
3
3
but employing deuterated methanol. In the case of d-PhGlyOCD the product was extracted from
3
the basic solution (pH=8) with dichloromethane, dehydrated with Na SO and concentrated to
obtain a yellow oil.
2
4
1
d- and l-4-HO-PhGlyOMe: H NMR (200 MHz, DMSO): 2.17 (s, 2H, NH ); 3.52 (s, 3H,
2
CH O); 4.35 (s, 1H, CHPh); 6.65 (d, 2H, H H -Ph, J=6.0 Hz,); 7.10 (d, 2H, H H -Ph, J=8.0
3
3
5
2
6
Hz). 13C NMR (200 MHz, DMSO): 52.40; 58.36; 115.71; 128.56; 132.00; 157.35; 175.51.
+
d-4-HO-PhGlyOCD : m/z=185 (M+1) .
3
1
d-PhGlyOCD : H NMR (200 MHz, CDCl ): 2.05 (s, 2H, NH ); 4.58 (s, 1H, CHPh); 7.31 (m,
3
3
2
5
H, Ph). 13C NMR (200 MHz, CDCl ): 58.80; 126.86; 128.08; 128.83; 140.20; 174.39.
3
4
.2. Polarimetry
A JASCO DIP-1000 polarimeter was used. The speci®c rotation of a chemical standard of
.
21
commercial (Aldrich) d-PhGlyOMe HCl was measured: [ꢀ] =^117 (c=1, H O) (lit. ^118).
D
2
Speci®c rotation results are reported in Sections 4.5.1 and 4.5.2.
4
.3. Determination of water activity
Water activity was measured using a hygrometer DARAI-Trieste, Italy. Measurements were
carried out by sealing the sensor into the open end of 5 mL glass vials, thermostatted, until constant
reading.19 Samples were equilibrated for at least 24 h in an air-bath type thermostatted orbital
shaker (DARAI-Trieste, Italy).
4
.4. Assay of PGA activity in water
Enzymatic activity of the immobilised PGA was assayed in phosphate bu€er by automated
titration (TTT80 Radiometer, Denmark) of the phenylacetic acid formed during the hydrolysis of
benzylpenicillin (Aldrich). One enzymatic unit corresponds to the amount of enzyme that
ꢀ
19
hydrolyses 1 mmol of benzylpenicillin for 1 min at pH 8.0 at 37 C.
4
.5. Enzymatic resolution of 1
4
.5.1. Reaction catalysed by PGA-450
Reaction was carried out taking samples of the enzyme stored in petroleum ether. After the fast
evaporation of the solvent 1 g of PGA-450 was added to 10 mL of dry toluene and the system was
equilibrated for 24 h (a =0.73±0.77). Reaction was started by adding 2 mmol of d,l-PhGlyOMe
w
1 and 1 mmol of methyl 4-hydroxyphenylacetate 2 and stopped when all the ester 2 was
consumed. The PGA-450 was removed by ®ltration and the organic solution was cooled down in

A. Basso et al. / Tetrahedron: Asymmetry 11 (2000) 1789±1796
1795
an ice-bath and a ¯ow of gaseous HCl was bubbled into the solution leading to the precipitation
of the d-PhGlyOMe hydrochloride salt 5. The white solid was ®ltered and rinsed with dichloro-
methane (60% isolated yield).
ꢀ
1
.
d-PhGlyOMe HCl: [ꢀ] =^115 (c=1, H O). Enantiomeric excess=98.3%. Mp=204 C. H
D
2
NMR (200 MHz, D O): 3.51 (s, 3H, CH O); 5.08 (s, 1H, CHPh); 7.31 (m, 5H, Ph).
2
3
1
4
.5.2. Reaction catalysed by PGA/Celite
1
Native PGA (50 mg) dissolved in 350 mL of water were added to 1 g of Celite R-640 rods in 10
mL of dry toluene contained in a 20 mL glass vial.22 Due to the presence of the hydrophobic
1
solvent, a uniform coating of the aqueous phase formed around the Celite rods and the enzymatic
1
solution was adsorbed on the Celite within 24 h (30 C in an orbital shaker). The toluene was
ꢀ
1
removed from the vessel and the Celite rods rinsed with 2 mL of dry dichloromethane. Dry
dichloromethane (10 mL) was then added together with the reactants (2 mmol of d,l-PhGlyOMe
1
0
and 1 mmol of methyl 4-hydroxyphenylacetate 2). During the course of the reaction a was
w
1
.85. The reaction was stopped when all the ester was consumed. The PGA/Celite was removed by
®ltration and the d-PhGlyOMe hydrochloride salt 5 was precipitated and puri®ed as previously
described (isolated yield=65%).
.
ꢀ
1
d-PhGlyOMe HCl: [ꢀ] =^88 (c=1, H O). Enantiomeric excess=75%. Mp=204 C. H NMR
D
2
(
200 MHz, D O): 3.51 (s, 3H, CH O); 5.08 (s, 1H, CHPh); 7.31 (m, 5H, Ph).
2 3
4
.6. Electrospray mass spectroscopy
An API I (Perkin±Elmer SCIEX) electrospray mass spectrometer was used in a positive mode
analysis, 5500 V. Positive ion detection was selected since low molecular weight polar compounds
form single charged ions by gaining a proton, thus leading to a peak at m/z=M+1.
Enzymatic reactions analysed by mass spectrometry were carried out in 1 mL of toluene using
1
either 62.5 mg of PGA-450 or 5 mg of native PGA adsorbed on 100 mg of Celite R-640 rods and
employing 100 mmol of ester 2 and 100 mmol each of the deuterated and not deuterated forms.
It was veri®ed that there is no appreciable variation between the rates of acylation of d-PhGlyOMe
and d-PhGlyOCD3.
4
.7. Determination of the initial rates of the acylation of the single enantiomers
0 mmol of each amine in the pure enantiomeric form and 100 mmol of the ester 2 were added
to 1 mL of dry toluene together with 62.5 mg of PGA-450 and incubated in a thermostatted
8
ꢀ
orbital shaker at 30 C. After complete conversion amides were isolated simply by evaporating the
solvents, since no side product is formed during the reactions. The reaction course was followed
withdrawing samples from the supernatant phase and analysing them by RP HPLC (Pharmacia)
using a C-18 Chrompack column. Isocratic elution: acetonitrile:water=70:30 and 0.025% of
TFA, ¯ow=1 mL/min. Detector: UV±vis spectrophotometer at a lmax=260 nm.
Conversions were evaluated on the basis of the ratio of the areas of the ester 2 and an internal
standard (cinnamyl alcohol). First 15% conversion data were utilised to determine initial rates as
usual in enzymatic reactions.23
1
13
1
Amides were characterised by H and C NMR. Compound 4: H NMR (200 MHz, CDCl ):
3
3
NH, J =6.2 Hz); 6.73 (d, 2H, H H C H OH); 7.07 (d, 2H, H H C H OH); 7.20±7.35 (m, 5H,
.52 (s, 2H, PhCH CO); 3.67 (s, 3H, CH O); 5.54 (d, 1H, CHPhGly, J =7.0 Hz); 6.48 (d, 1H,
2
3
1
1
3
5
6
4
2
6
6
4

1
796
A. Basso et al. / Tetrahedron: Asymmetry 11 (2000) 1789±1796
Ph). 13C NMR (200 MHz, CDCl ): 42.64; 53.01; 56.68; 116.00; 125.86; 127.17; 128.64; 129.03;
3
1
30.63; 136.62; 155.35; 170.54; 171.17.
In the case of 4-HO-PhGlyOMe only the l-enantiomer was acylated giving the corresponding
1
13
1
amide that was characterised by H NMR and C NMR: H NMR (200 MHz, DMSO): 3.34 (s,
H, PhCH CO); 3.54 (s, 3H CH O); 5.16 (d, 1H, CHPh, J=6.0 Hz); 6.62 (d, 2H, H H
6
2
2
3
2
HOPhGly, J=8.0 Hz); 6.72 (d, 2H, H H C H OH, J=8.0 Hz); 7.00 (d, 2H, H H HOPhGly,
5
2
6
6
4
3
13
J=8.0 Hz); 7.13 (d, 2H, H H C H OH, J=8.0 Hz); 8.68 (d, 1H, NH, J=6.0 Hz). C NMR
3
5
6
4
(
1
200 MHz, DMSO): 41.38; 52.74; 56.64; 115.62; 116.09; 126.77; 126.90; 129.68; 130.59; 156.56;
58.23; 171.30; 172.16.
Acknowledgements
We are grateful to Dr. Fabio Hollan for ES-MS analysis and to Boehringer-Mannheim for the
generous gift of PGA-450. Thanks are due to C.N.R. and M.U.R.S.T. (Roma) for ®nancial
support to P.L.
References
1
2
3
4
5
6
7
. Kim, M. G.; Lee, S. B. J. Mol. Catal. B: Enzymatic 1996, 1, 71±80.
. Rosell, C. M.; Ferna
. Rosell, C. M.; Ferna
Â
Â
ndez-Lafuente, R.; Guisa
ndez-Lafuente, R.; Guisa
Â
Â
n, J. M. J. Mol. Catal. 1993, 84, 365±371.
n, J. M. Ann. N. Y. Acad. Sci. 1995, 750, 425±428.
. van Langen, L. M.; de Vroom, E.; van Rantwijk, F.; Sheldon, R. FEBS Lett. 1999, 456, 89±92.
. Bossi, A.; Cretich, M.; Righetti, P. G. Biotechnol. Bioeng. 1998, 60, 454±461.
. Boccu
. Justiz, O. H.; Terreni, M.; Pagani, G.; Garcõ
Technol. 1999, 25, 336±343.
. Cardillo, G.; Tolomelli, A.; Tomasini, C. J. Org. Chem. 1996, 61, 8651±8654.
Á
, E.; Ebert, C.; Gardossi, L.; Gianferrara, T.; Zacchigna, M.; Linda, P. Il Farmaco 1991, 46, 565±577.
Â
Â
a, J. L.; Guisan, J. M.; Fernandez-Lafuente, R. Enzyme Microb.
Â
Â
8
9
. Baldaro, E.; D'Arrigo, P.; Pedrocchi-Fantoni, G.; Rosell, C. M.; Servi, S.; Tagliani, A.; Terreni, M. Tetrahedron:
Asymmetry 1993, 4, 1031±1034.
1
1
1
1
1
1
1
1
1
1
2
2
0. Fuganti, C.; Grasselli, P.; Servi, S.; Lazzarini, A.; Casati, P. Tetrahedron 1988, 9, 2575±2582.
1. (a) Cole, M. Biochem. J. 1969, 115, 747±756. (b) Cole, M. Biochem. J. 1969, 115, 757±764.
2. Zmijewski, M.; Briggs, B. S.; Thompson, A. R.; Wright, I. G. Tetrahedron Lett. 1991, 32, 1621±1622.
3. Kim, M. G.; Lee, S. B. J. Mol. Catal. B: Enzymatic 1996, 1, 181±190.
Ï
4. DidzÏ
5. Pessina, A.; Lu
iapetris, R.; Drabnig, B.; Schellenberger, V.; Jakubke, H.-D.; Svedas, V. FEBS 1991, 287, 31±33.
thi, P.; Luisi, P. L. Helv. Chim. Acta 1988, 71, 631±641.
È
6. Basso, A.; De Martin, L.; Ebert, C.; Gardossi, L; Linda, P. J. Mol. Catal. B: Enzymatic, in press.
7. Pohl, T.; Waldmann, H. Tetrahedron Lett. 1995, 36, 2963±2966.
8. Basso, A.; De Martin, L.; Ebert, C.; Gardossi, L.; Linda, P.; Zlatev, V. J. Mol. Catal. B: Enzymatic, in press.
9. De Martin, L.; Ebert, C.; Garau, G.; Gardossi, L.; Linda, P. J. Mol. Catal. B: Enzymatic 1999, 6, 437±445.
0. Halling, P. J. Enzyme Microb. Technol. 1994, 16, 178±206.
1. Clark, J. C.; Phillipps, G. H.; Steer, M. R.; Cooksey, A. R.; Stephenson, L. J. Chem. Soc., Perkin Trans. 1 1976,
4
71±474.
2
2
2. Basso, A.; De Martin, L.; Ebert, C.; Gardossi, L.; Linda, P. Chem. Commun. 2000, 467±468.
3. Ebert, C.; Gardossi, L.; Linda, P. Tetrahedron Lett. 1996, 37, 9377±9380.