Hydrolysis of nitriles using an immobilized nitrilase: Applications to the synthesis of methionine hydroxy analogue derivatives

Article abstract of DOI:10.1021/jf048827q

Mild and selective hydrolysis of a large range of nitriles leading to carboxylic acids was achieved under neutral conditions by an immobilized and genetically modified enzyme preparation from Alcaligenes faecalis ATCC8750. This immobilized nitrilase has been shown to be an effective catalyst for the stereoselective hydrolysis of mandelonitrile 1a to R-(-)-mandelic acid 1c. This method is particularly useful for the production of hydroxy analogues of methionine derivatives 2c-4c that could have an interest in cattle feeding and for the transformation of compounds containing other acid- or base-sensitive groups 3a-10a. A series of aliphatic dinitriles 11a-15a was hydrolyzed to the corresponding cyano acids. The suitability of the immobilized catalyst as a robust and versatile biocatalyst is discussed, and models to account for the stereoselectivity of the enzymic hydrolysis have been proposed.

Full text of DOI:10.1021/jf048827q

J. Agric. Food Chem. 2004, 52, 8155−8162 8155
Hydrolysis of Nitriles Using an Immobilized Nitrilase:
Applications to the Synthesis of Methionine Hydroxy Analogue
Derivatives
PATRICK REY,† JEAN-CHRISTOPHE ROSSI,*,§ JACQUES TAILLADES,§
‡
GEORGES GROS,‡ AND OLIVIER NORE
Organisation Mole´culaire-Evolution et Mate´riaux Fluore´s (UMR CNRS 5073) CC009/
CC017-Universite´ Montpellier II, Place Euge`ne Bataillon, 34095 Montpellier Cedex 5, France,
Archemis, 24 Avenue Jean Jaures, 69153 De´cines Charpieu, France, and Adisseo,
42 Avenue Aristide Briand, B.P. 100, 92164 Antony Cedex, France
Mild and selective hydrolysis of a large range of nitriles leading to carboxylic acids was achieved
under neutral conditions by an immobilized and genetically modified enzyme preparation from
Alcaligenes faecalis ATCC8750. This immobilized nitrilase has been shown to be an effective catalyst
for the stereoselective hydrolysis of mandelonitrile 1a to R-(-)-mandelic acid 1c. This method is
particularly useful for the production of hydroxy analogues of methionine derivatives 2c-4c that could
have an interest in cattle feeding and for the transformation of compounds containing other acid- or
base-sensitive groups 3a-10a. A series of aliphatic dinitriles 11a-15a was hydrolyzed to the
corresponding cyano acids. The suitability of the immobilized catalyst as a robust and versatile
biocatalyst is discussed, and models to account for the stereoselectivity of the enzymic hydrolysis
have been proposed.
KEYWORDS: Biotransformation; enzymes; hydrolases; methionine hydroxy analogue derivatives;
dinitriles
INTRODUCTION
the chemical hydrolysis of nitriles usually requires rather harsh
conditions. Such media are often incompatible with other
hydrolyzable or acid/base sensitive groups within the molecule,
and in these cases neutral hydrolytic conditions would be a clear
advantage. Chemical hydration under neutral conditions has been
conducted, albeit with varying degree of success. Often relatively
expensive and elaborate reagent systems are required, for
example, platinum (10), palladium (11), and cobalt complexes
(12). Furthermore, elevated reaction temperatures are usually
necessary. However, the preparations of hydrophilic cross-linked
polymers that are carrying carbonyl groups of the 4-piperidone
type catalyze the hydration of R-aminonitriles to R-amino acids
have been reported (13, 14). Moreover, borates are an efficient
method for the hydration of R-hydroxynitriles to R-hydroxyacids
(15). Biocatalysis plays an increasingly important role in
synthetic organic chemistry due to the very mild conditions
normally required and the usually high chemo-, regio-, and
stereoselectivity with which such transformations take place.
Moreover, the continuous development of genetic methodologies
has permitted the achievement of controlled and highly produc-
tive engineered microorganisms. In this respect, the ability of
enzymes to hydrolyze nitriles, in particular, employed as whole
microorganisms, is well-established (16-18), and a range of
both aromatic and aliphatic nitriles as well as dinitriles has been
successfully hydrolyzed (19-21). Optically active products have
also been obtained (22, 23). Indeed, in addition to operating
During the past 15-20 years, there has been an enormous
increase in the use of enzymes as catalysts for organic synthesis
(1). In particular, hydrolytic enzymes (2) (e.g., lipases, esterases,
and acylases) have proven to be particularly useful catalysts
for the preparation of homochiral molecules. There are two
major reasons for the widespread use of these catalysts by a
variety of laboratories: (i) the enzymes operate on wide range
of substrates with very high stereoselectivity and (ii) they are
robust and extremely easy to use. Furthermore, the replacement
of water by organic solvent (3) has expended the range of
reactions that can be affected (e.g., esterification and peptide
bond formation). In the same context, chemical transformation
of nitriles into carboxylic acids or amides is of synthetic
importance owing to the ease with which the desired organic
nitrile usually can be obtained. Thus, organonitriles can be
readily prepared by a number of methods including the addition
of cyanide ion to alkyl halides (4), the Strecker reaction (5),
the Sandmeyer reaction (6), the reaction of aryl halides with
copper cyanide (7), the reaction of ketone with tosylmethyl
isocyanide (8), and the dehydration of amides (9). However,
* To whom correspondence should be addressed. Tel: +330467143544.
Fax: +330467544757. E-mail: j-christ.rossi@univ-montp2.fr.
† Archemis.
‡ Adisseo.
§ Universite´ Montpellier II.
10.1021/jf048827q CCC: $27.50
© 2004 American Chemical Society
Published on Web 11/18/2004

8156 J. Agric. Food Chem., Vol. 52, No. 26, 2004
Rey et al.
Recently, the purified nitrilase from ATC8750 was genetically
modified and immobilized on alumina. This immobilized
catalyst that showed an excellent turnover was used in an
industrial process of production of the hydroxy analogue of
methionine (36).
The aim of this work is to demonstrate the suitability and
the robustness of this immobilized catalyst toward a wide range
of nitriles. First of all, we examined nitrile precursors of
mandelic acid 1c, O-acetyl mandelic acid 2c, a hydroxy analogue
of methionine (3c), and a hydroxy analogue of methionine
derivatives (4c-6c) in which the hydrophobic groups (O-acetyl
or O-alkyl) have cropped up in the alcohol function (Figure
1). These compounds should be interesting in the field of cattle
feeding.
Indeed, (()-2-hydroxy-4-methylthiobutanoic acid 1c (me-
thionine hydroxy analogue call HMB hereafter) is the only
known instance in which an amino acid has been successfully
replaced by an analogue in nutrition. This compound is produced
industrially in very high quantities because of its important
interest in poultry feeding. In cattle feeding, the effectiveness
of administering free HMB has been questioned in a study on
nutrition in lactating dairy cows (37). Rumen microbes either
altered or degraded 99% of the HMB since recovery of the
analogue in duodenal digesta is less than 1% of the amount fed
for both the acid form and the calcium salt (38). However, it
has been showed that the insertion of a hydrophobic group in
the acid function facilities stomach crossing (39, 40). Recently,
this concept was also envisaged by the use of oligopeptides of
HMB (41).
Second and to generalize the use of this catalyst, the
immobilized nitrilase was used for the hydrolysis of nitriles
bearing acido or baso labile groups (7a-12a) and dinitriles
(13a-17a) (Figure 1).
Moreover, this work is completed on a several gram scale of
substrates, thus authorizing an industrial extrapolation.
Figure 1. Reaction and structures of substrates tested with the immobilized
nitrilase from A. faecalis ATCC8750.
under mild conditions, enzyme-catalyzed hydrolysis offers the
potential for carrying out (chemo-, regio, stereo-) transformations
that are difficult to achieve nonenzymatically (24). The mech-
anism of these enzyme systems has been investigated (25-27),
and two distinct pathways are believed to be operating: first,
the stepwise hydrolysis of the nitrile (a) into an amide (b) via
a nitrile hydratase followed by transformation of the latter into
a carboxylic acid (c) by an amidase, and second, the direct
hydrolysis of the nitrile via a nitrilase (Figure 1).
Research studies on the mechanisms of the enzymatic
hydrolysis were essentially concentrated toward the nitriles
hydratases. Thus, several nitrile hydratases from Rhodococcus
(28), Corynebacterium sp. (29), BreVibacterium R312 (30), and
Pseudomonas chlororaphis B23 (31) were isolated and purified.
It results from the investigations carried out by Yamada et al.
that these enzymes would contain in their active site of the
enzyme a prosthetic group, the pyroloquinoleine quinone (PQQ)
(31), as well as an iron atom (32) or cobalt (28). In the case of
the nitrilases, the mechanism defined since 1964 per Hook et
al. (33) described an analogy with the traditional chemical
hydrolysis of the nitrile function in basic medium.
MATERIALS AND METHODS
Unless otherwise stated, all reagents were obtained from commercial
suppliers and used without further purification. NMR spectra were
recorded on a Brucker DRX 400 MHz spectrometer (400 MHz 1H and
100.6 MHz 13C). Chemical shift (δ) is expressed in ppm downfield
from tetramethylsilane and reported as δH, δC, number of equivalent
nuclei (by integration), and multiplicity (s ) singlet, d ) doublet, q )
quartet, m ) multiplet, br ) broad). IR spectra were determined on a
Brucker IFS 25 spectrophotometer. Mass spectra (m/z) were recorded
on JEOL JMS-DX 300. All HPLC were run using a Varian chromato-
graph with an UV detector set at 210 nm. HPLC was achieved using
a Nucleosil RP C18 (5 µm, 100 Å). Determination of ee was run using
an HP1050 chromatograph with an UV detector set at 210 nm.
Separation of enantiomers was achieved using a chiral column
CHIRALCEL OD (25 cm, Baker, Deventer, The Netherlands) eluted
at 0.9 mL/min with Heptan/2-propanol 90:10 (v/v), 0.1% (v/v) H3PO4.
The pH 7.3 phosphate buffer solution used for the biotransformation
was prepared by using a 120 mmol dm-3 solution of KH2PO4 and K2-
HPO4.
Preparation of (()-2-Acetoxymandelonitrile 2a. To a round-
bottomed flask was added 100 mL of dichloromethane, 15 g (0.113
mol) of commercial mandelonitrile 1a and 17.3 g (0.17 mol) of acetic
anhydride. The mixture was agitated a few minutes at ambient
temperature. Then, 2 g (0.025 mol) of acetyl chloride was added to
the reaction medium. The mixture was carried to backward flow (70
°C) during 20 h. At the ambient temperature, 100 mL of saturated Na2-
CO3 solution was added with precaution under agitation. The solution
was separated, and the organic phase was evaporated. A total of 100
mL of ether was then added to the oil. The ether phase was washed
with water until neutral pH and dried with Na2SO4. Yield of (()-2-
The nitrilase of A. faecalis ATCC8750 was purified from cell-
free extract 29.0-fold in a yield of 17.9% by Yamamoto et al.
(34, 35) who examined the ability of the enzyme to catalyze
the hydrolysis of various nitriles. The enzyme appears to have
strict substrate specificity for aromatic hydroxy nitriles derived
from mandelonitrile but showed no activity for other hydroxy
nitriles such as hydroxy acetonitrile, lactonitrile, and 2-hydroxy-
4-phenybutyronitrile.
1
acetoxymandelonitrile: 86%. H NMR (DMSO d6) δH ppm: 2.16 (s,

Synthesis of Methionine Hydroxy Analogue Derivatives
J. Agric. Food Chem., Vol. 52, No. 26, 2004 8157
3H, CH3CO), 6.69 (s, 1H, CH (OAc)CN), 7.52-7.61 (m, 5H, aromatic
CH). 13C NMR (DMSO d6) δC ppm: 21.03 (CH3CO), 63.69 (CH(OAc)-
CN), 117.9 (CN), 128.61 (C-2,6), 129.9 (C-3,5), 130.9 (C-4), 136.26
(C-1), 169.55 (CO2CH3). IR (CHCl3) cm-1: 2245, 1750. Mass FAB+
(NBA): m/e (relative intensity) 176 (20), 149 (5), 116 (70)0.
2-Hydroxy-4-methylthiobutanenitrile (3a) 97% w/w. 1H NMR
(DMSO d6) δH 1.67 (td, 2H, S-CH2-CH2), 1.79 (s, 3H, CH3-S), 2.26 (t,
2H, CH3-S-CH2), 4.31 (td, 1H, HO-CH-CN), 6.54 (d, 1H, OH); 13C
NMR (DMSO d6) δC 15.28 (CH3-S), 29.13 (CH3-S-CH2), 35.18 (S-
CH2-CH2), 59.40 (HO-CH-CN), 121.76 (CN); IR (CHCl3) cm-1 3300-
3500, 2175; mass FAB+ (GT), m/e (relative intensity) 132 (25), 131
(20), 115 (100), 105 (35), 61 (45).
To (()-2-methoxymethoxy-but-3-enenitrile previously prepared (0.20
g, 1.6 mmol) at -10 °C was added carefully 0.25 mL (2.5 mmol) of
methanethiol with a syringe through a septum. The solution was stirred
at temperature for 10 min under UV irradiation. Next, the temperature
was raised to 20 °C, and 2.0 mL (2 mmol) of the triethylborane solution
from Aldrich (1 M in hexane) was added slowly dropwise with a
syringe. The reaction was then stirred at 20 °C for 3 h under UV
irradiation. After being stirred for 3 h, the system was purged with
nitrogen, and the reaction mixture was washed with 20 mL of aqueous
solution. The aqueous solution was extracted with ether (3 × 20 mL).
The ether was dried over Na2SO4 and evaporated to dryness to give
compound 4a (0.26 g, 1.52 mmol): yield, 95%; 1H NMR (DMSO d6)
δH 2.05 (td, 2H, S-CH2-CH2), 2.11 (s, 3H, CH3-S-CH2), 2.59 (t, 2H,
S-CH2-CH2), 3.34 (s, 3H, CH3-O), 4.67 (td, 1H, O-CH-CN), 4.73 (s,
2H, O-CH2-O); 13C NMR (DMSO d6) δC 15.30 (CH3-S), 29.08 (S-
CH2-CH2), 33.34 (S-CH2-CH2), 56.46 (CH3-O), 64.49 (O-CH-CN),
96.50 (O-CH2-O), 119.74 (CN); IR (CHCl3) cm-1 2950, 2145;
mass FAB+ (NBA), m/e (relative intensity) 175 (absent), 149 (95), 45
(100).
Preparation and Immobilization of the Enzyme. The immobilized
enzyme was courteously provided by Adisseo. Preparation and im-
mobilization are described in the Adisseo U.S. Patent 6180359 (39).
The batch used in this study had a nitrilase activity of 3600 µmol/h
per mg of dry catalyst (substrate, 2-hydroxy-4methylthiobutanenitrile).
General Procedure for the Enzymatic Hydrolysis of Nitriles. The
substrates were suspended in potassium phosphate buffer (120 mmol
dm-3, pH 7.3). The immobilized enzyme system was added, and the
reaction was shaken at 200 rpm, 30 °C. The reaction was terminated
by filtration of the biocatalyst. The aqueous filtrate was acidified (pH
1.5 with HCl 10 N) and extracted with ethyl acetate or ether. The
combined extracts were dried (Na2SO4) and filtered, and the filtrate
was subjected to rotary evaporation to afford any unchanged nitrile
and/or acid product.
Preparation of 2-Acetoxy-4-methylthiobutanenitrile (4a). To
2-acetoxy-but-3-enenitrile (Aldrich) (2.5 g, 20 mmol) at -10 °C was
added carefully methanethiol (1.6 mL, 28 mmol) with a syringe through
a septum. Then, the solution was stirred for 10 min under UV irradiation
while returning to 20 °C. Next, the temperature was raised to 20 °C,
and 4.0 mL (4 mmol) of the triethylborane solution (Aldrich) (1 M in
hexane) was added dropwise with a syringe. The reaction was then
stirred at 20 °C for 4 h under UV irradiation. After being stirred for 4
h, the system was purged with nitrogen, and the reaction mixture was
washed with 30 mL of aqueous solution. The aqueous solution was
extracted with ether (3 × 50 mL). The ether was dried over Na2SO4
and evaporated to dryness to give the compound 4a (3.35 g, 19.4 mmol);
1
yield, 97%; H NMR (DMSO d6) δH 2.03 (td, 2H, S-CH2-CH2) 2.07
(s, 3H, CH3-S), 2.12 (s, 3H, CH3-CO2), 2.58 (t, 2H, CH3-S-CH2), 5.48
(t, 1H, CH (OAc)CN); 13C NMR (DMSO d6) δC 15.25 (CH3-S), 20.93
(CH3-CO2), 28.83 (CH3-S-CH2), 31.76 (S-CH2-CH2), 60.86 (CH(OAc)-
CN), 118.1 (CN), 170.1 (CH3CO2); IR (CHCl3) cm-1 2145, 1745; mass
FAB+ (NBA), m/e (relative intensity) 174 (55), 149 (45), 131 (65),
105 (35), 61 (75).
Preparation of 2-(1-Ethoxyethoxy)-4-methylthiobutanenitrile (5a).
(()-2-(1-Ethoxyethoxy)-but-3-enenitrile was prepared as described by
The aqueous portion was then basified (pH 10, 2 mol dm-3 NaOH)
and extracted with ethyl acetate or ether. The combined extracts were
dried (Na2SO4), filtered, and subjected to rotary evaporation to afford
the amide. Amounts of substrate catalyst and buffer used as well as
reaction times and purification techniques are given in Table 4. Yields,
1H and 13C NMR, IR, and mass spectrometry data for each individual
reaction are given.
1
ref 47 yield, 88%; H NMR (CDCl3) δH 1.15 (td, 3H, CH3-CH2-O),
1.29 (dd, 3H, CH3-CH), 3.50 (qd, 2H, O-CH2-CH3), 4.92 (q, 1H, -O-
CH (CH3)-O), 4.82-4.97 (m, 1H, O-CH-CN), 5.35-5.55 (m, 2H,
CH2dCH), 5.72-5.81 (m, 1H, CH2)CH); 13C NMR (CDCl3) δC 15.34
(CH3-CH2-O), 20.03 (CH3-CH-), 58.50 (CH3-CH2-O), 63.27 (O-CH-
CN), 99.80 (-O-CH(CH3)-O), 117.24 (CN), 120.47 (CH2)CH), 131.06
(CH2)CH); IR (CHCl3) cm-1 2950, 2184, 1635; mass FAB+ (NBA),
m/e (relative intensity) 154 (35), 73 (100), 43 (44).
Enzymatic Hydrolysis of Mandelonitrile (1a). Substrate concentra-
tions were 8.0 mmol dm-3. White crystals were isolated and shown by
To (()-2-(1-ethoxyethoxy)-but-3-enenitrile (8) (3.2 g, 20 mmol) at
-10 °C was added carefully methanethiol (2.0 mL, 36 mmol) with a
syringe through a septum. Then, the solution was stirred during 10
min under UV irradiation while returning to 20 °C. Next, the
temperature was raised to 20 °C and 2.0 mL (2 mmol) of the
triethylborane solution Aldrich (1 M in hexane) was added slowly
dropwise with a syringe. The reaction was then stirred at 20 °C for 3
h under UV irradiation. After being stirred for 3 h, the system was
purged with nitrogen, and the reaction mixture was washed with 30
mL of aqueous solution. The aqueous solution was extracted with ether
(3 × 50 mL). The ether was dried over Na2SO4 and evaporated to
dryness to give the compound 5a (2.92 g, 14.4 mmol): yield, 72%; 1H
NMR (DMSO d6) δH 1.23 (td, 3H, CH3-CH2-O), 1.55 (dd, 3H, CH3-
CH), 2.07 (m, 2H, S-CH2-CH2), 2.11 (s, 3H, CH3-S), 2.65 (m, 2H,
S-CH2-CH2), 3.51 (m, 2H, O-CH2-CH3), 4.68 (m, 1H, O-CH-CN), 4.90
(m, 1H, O-CH(CH3)-O); 13C NMR (DMSO d6) δC 15,67 (CH3-S), 15.89
(CH3-CH2-O), 21.23 (CH3-CH), 27.40 (S-CH2-CH2), 33.72 (CH3-S-
CH2), 57.2 (CH3-CH2-O), 60.65 (O-CH-CN), 99.49 (O-CH(CH3)-O),
119.13 (CN); IR (CHCl3) cm-1 2950, 2145; mass FAB+ (NBA), m/e
(relative intensity) 204 (5), 188 (5), 158 (15), 149 (25), 73 (100), 61
(30).
Preparation of (()-2-Methoxymethoxy-4-methylthiobutanenitrile
(6a). (()-2-Methoxymethoxy-but-3-enenitrile was prepared as described
by ref 48: yield, 63%; 1H NMR (CDCl3) δH 3.36 (s, 3H, CH3-O), 4.71-
4.82 (ABq, 2H, O-CH2-O), 4.86 (td, 1H, O-CH-CN), 5.40-5.60 (dd,
2H, CH2dCH), 5.75-5.85 (m, 1H, CH2dCH); 13C NMR (CDCl3) δC
56.54 (CH3-O), 65.49 (O-CH-CN), 95.51 (O-CH2-O), 117.06 (CN),
121.28 (CH2dCH), 130.36 (CH2dCH); IR (CHCl3) cm-1 2985, 2174,
1665; mass FAB+ (NBA), m/e (relative intensity) 126 (absent), 57 (98),
55 (100), 43 (96).
1
1H NMR spectroscopy to be mandelic acid 1c: yield, 99%; H NMR
(DMSO d6) δH 5.06 (s, 1H, CH (OH)CO2H), 5.88 (s, 1H, OH), 7.25-
7.52 (m, 5H, CH aromatic), 12.54 (s, 1H, CO2H); 13C NMR (DMSO
d6) δC 73.26 (CH(OH)CN), 127.51 (C-2,6), 128.52 (C-4), 128.92 (C-
3,5), 141.07 (C-1), 175.01 (CO2H); IR (CHCl3) cm-1 2500-3500, 1743;
mass FAB- (NBA), m/e (relative intensity) 152 (100), 122 (20), 92
(5).
Enantiomeric excess (chiral HPLC): 61%. Retention time (S)-(+)-
mandelic acid 8.99 min, (R)-(-)-mandelic acid 10.95 min.
Enzymatic Hydrolysis of O-Acetyl Mandelonitrile (2a). Substrate
concentration of 8.0 mmol dm-3. White crystals were isolated and
shown by 1H NMR spectroscopy to be a mixture of O-acetyl mandelic
acid 6c (83%) and mandelic acid 5c (16%): 1H NMR (DMSO d6) δH
2.13 (s, 3H, CH3CO2), 5.83 (s, 1H, CH (OAc)CO2H), 7.38-7.51 (m,
5H, CH aromatic), 13.26 (s, 1H, CO2H); 13C NMR (DMSO d6) δC 21.32
(CH3CO2), 75.07 (CH(OAc)CO2H), 128.53 (C-2,6), 129.51 (C-3,5),
129.84 (C-4), 135.22 (C-1), 170.75 (CO2CH3), 170.75 (CO2H); IR
(CHCl3) cm-1 2500-3500, 1734, 1725; mass FAB- (NBA), m/e
(relative intensity) 193 (40), 153 (70), 149 (40); enantiomeric excess
(chiral HPLC), 0%; retention time (S)-(+)-mandelic acid, 8.99 min,
(R)-(-)-mandelic acid, 10.95 min, (S)-(+)-2 acetoxy-mandelic acid,
6.59 min, and (R)-(-)-2-acetoxymandelic acid 7.12 min.
Enzymatic Hydrolysis of 2-Hydroxy-4-methylthiobutanenitrile
(3a). HPLC analysis had indicated that all the starting material had
been converted into a single more polar product (water/acetonitrile
80:20 (v/v)/0.05% (v/v) trifluoroacetic acid; λ ) 210 nm; flow ) 1
mL/min; column: Nucleosil RP C18 (5 µm, 100 Å). After recovery, a
liquid was isolated (11.1 g, 74 mmol) and shown by 1H NMR
spectroscopy to be 2-hydroxy-4-methylthiobutanoic acid 1c: yield,

8158 J. Agric. Food Chem., Vol. 52, No. 26, 2004
Rey et al.
1
99%; H NMR (DMSO d6) δH 1.78 (s, 3H, CH3-S), 1.56-1.68 (m,
Enzymatic Hydrolysis of (()-2-(1-Ethoxyethoxy)-but-3-enenitrile
10a. Yield after extraction ) 86% of (()-2-(1-ethoxyethoxy)-but-3-
enoic acid 9c: 1H NMR (DMSO d6) δH 1.06 (t, 3H, CH3-CH2-O), 1.19
(d, 3H, CH3-CH), 3.40-3.55 (m, 2H, CH3-CH2-O-), 4.32 (m, 1H, CH2d
CHsCH), 4.75-4.66 (qd, 1H, O-CH-O), 4.95-5.2 (m, 2H, CH2dCHs),
5.78-5.91 (m, 1H, CH2dCH), 12.32 (s, 1H, CO2H); 13C NMR (DMSO
d6) δC 15.94 and 16.06 (CH3-CH2-O-), 20.75 and 21.15 (CH3-CH),
60.24 and 61.10 (CH3-CH2-O-), 78.61 and 78.92 (CH2dCHsCH),
98.31 and 98.87 (O-CH-O), 113.86 and 115.0 (CH2dCHs), 138.25
and 138.56 (CH2dCH), 174.56 (CO2H); IR (CHCl3) cm-1 2500-3500,
1725; mass FAB- (GT), m/e (relative intensity) 173 (100), 43 (35).
Enzymatic Hydrolysis of (()-2-Methoxymethoxy-but-3-enenitrile
11a. Yield after extraction ) 80% in (()-2-methoxymethoxy-but-3-
enoic acid 11c: 1H NMR (DMSO d6) δH 3.30-3.95 (m, 3H, CH3-O-),
4.37 (m, 1H, CH2dCHsCH), 4.62-4.71 (dd, O-CH2-O), 5.07-5.23
(m, 2H, CH2dCHs), 5.91-5.98 (m, 1H, CH2dCH), 11.47 (s, 1H,
CO2H); 13C NMR (DMSO d6) δC 55.58 (CH3-O-), 78.87 (CH2dCHs
CH), 94.62 (O-CH2-O), 114.38 (CH2dCHs), 138.11 (CH2dCH),
173.67 (s, 1H, CO2H); IR (CHCl3) cm-1 2500-3500, 1725; mass FAB-
(GT), m/e (relative intensity) 193 (100).
2H, S-CH2-CH2), 2.26 (t, 2H, S-CH2-CH2), 3.79 (CH (OH)CO2H), 11.74
(s, 1H, CO2H); 13C NMR (DMSO d6) δC 15.38 (CH3-S), 30.08 (CH3-
S-CH2), 34.32 (S-CH2-CH2), 69.34 (CH(OH)CO2H), 176.40 (CO2H);
IR (CHCl3) cm-1 2500-3500, 1725; mass FAB- (NBA), m/e (relative
intensity) 149 (100), 122 (5), 46 (15).
Enzymatic Hydrolysis of 2-Acetoxy-4-methylthiobutanenitrile
(4a). HPLC analysis had indicated that all the starting material had
been converted into two more polar products (water/acetonitrile 80:20
(v/v)/0.05% (v/v) trifluoroacetic acid; λ ) 210 nm; flow ) 1 mL/min;
column: Nucleosil RP C18 (5 µm, 100 Å). A liquid was isolated and
shown by 1H NMR spectroscopy to be a mixture of 2-acetoxy-4-
methylthiobutanoic acid 4c (82%) and 2-hydroxy-4-methylthiobutanoic
acid 3c (6%): 1H NMR (DMSO d6) δH 1.75 (m, 2H, S-CH2-CH2),
1.79 (s, 3H, CH3CO2), 1.84 (s, 3H, CH3-S), 2.28 (t, 2H, CH3-S-CH2),
4.69 (td, 1H, CH (OAc)CN), 11.9 (s, 1H, CO2H); 13C NMR (DMSO
d6) δC 15.24 (CH3-S), 21.17 (CH3-CO), 29.73 (CH3-S-CH2), 30.96 (S-
CH2-CH2), 71.35 (CH(OAc)CN), 170.77 (CH3-CO2), 171.86 (CO2H);
IR (CHCl3) cm-1 2500-3500, 1745, 1720.; mass FAB- (NBA), m/e
(relative intensity) 191 (100), 149 (70), 59 (25).
Enzymatic Hydrolysis of 2-Cyan-O-ethyl Acetate 12a. Yield after
extraction ) 94% of 2-carbetoxyacetic acid 12c: 1H NMR (DMSO
d6) δH 1.19 (t, 3H, CH3-CH2-O), 3.33 (s, 2H, HO2C-CH2-CO2Et), 4.12
(q, 3H, CH3-CH2-O), 10.71 (s, 1H, CO2H); 13C NMR (DMSO d6) δC
14.66 (CH3-CH2-O-), 42.34 (HO2C-CH2-CO2Et), 64.47 (CH3-CH2-O),
167.7 (-CO2-CH2-CH3), 168.9 (CO2H); IR (CHCl3) cm-1 2500-3500,
1745, 1725; mass FAB- (GT), m/e (relative intensity) 131 (65), 104
(40), 87 (100), 41 (25).
Enzymatic Hydrolysis of Malononitrile 13a. Yield after extraction
) 87% of 2-cyanoacetic acid 13c: 1H NMR (DMSO d6) δH 3.57 (s,
2H, -CH2-), 11.57 (s, 1H, CO2H); 13C NMR (DMSO d6) δC 25.46
(CH2-CN), 115.4 (CCN), 165.6 (CO2H); IR (CHCl3) cm-1 2500-
3500, 2240, 1725; mass FAB- (GT), m/e (relative intensity) 84 (100),
46 (60).
Enzymatic Hydrolysis of Adiponitrile 14a. Yield after extraction
) 90% of 5-cyanopentanoic acid 14c: 1H NMR (DMSO d6) δH 1.60
(m, 2H, HO2C-CH2-CH2-), 1.62 (m, 2H, NC-CH2-CH2-), 2.23 (t, 2H,
-CH2-CO2H), 2.51 (t, 2H, -CH2-CN), 12.41 (s, 1H, CO2H); 13C NMR
(DMSO d6) δC 16.92 (NC-CH2-CH2), 24.92 (NC-CH2-CH2), 25.31
(CH2-CH2-CO2H), 34.95 (CH2-CO2H), 121.38 (CN), 176.41 (s, 1H,
CO2H); IR (CHCl3) cm-1 2500-3500, 2245, 1730; mass FAB- (GT):
m/e (relative intensity) 126 (100).
Enzymatic Hydrolysis of Fumaronitrile 15a. Yield after extraction
) 72% in 3-cyanoacrylic acid 15c and 28% of 3-cyanoacrylamide 15b.
3-Cyanoacrylic acid 15c: 1H NMR (DMSO d6) δH 5.84 (d, 1H, HO2Cs
CHdCH-CN), 6.54 (d, 1H, HO2CsCHdCHsCN), 13.10 (s, 1H,
CO2H); 13C NMR (DMSO d6) δC 115.03 (HO2CsCHdCHsCN),
119.09 (CN), 153.22 (HO2CsCHdCHsCN), 167.13 (CO2H); IR
(CHCl3) cm-1 2500-3500, 2225, 1695; mass FAB- (GT), m/e (relative
intensity) 96 (100), 52 (10). 3-Cyanoacrylamide 15b: 1H NMR (DMSO
d6) δH 6.56 (d, 1H, H2NOCsCHdCHsCN), 7.04 (d, 1H, H2NOCs
CHdCHsCN), 7.72-7.99 (d, 2H, CONH2); 13C NMR (DMSO d6) δC
109.54 (H2NOCsCHdCHsCN), 117.95 (CN), 144.95 (H2NOCsCHd
CHsCN), 164.01 (CONH2); IR (CHCl3) cm-1 3100-3500, 2235, 1695;
mass FAB- (GT), m/e (relative intensity) 95 (10).
Enzymatic Hydrolysis of 1,4-Dicyanobut-2-ene 16a. Yield after
extraction ) 82% of 5-cyanopent-3-enoic acid 16c: 1H NMR (DMSO
d6) δH 3.04 (dd, 2H, HO2C-CH2), 3.31 (dd, 2H, CH2-CN), 5.51 (td,
1H, NCsCH2sCHd), 5.82 (td, 1H, H2OCsCH2sCHd), 10.51 (s,
1H, CO2H); 13C NMR (DMSO d6) δC 20.34 (CH2-CN), 38.14 (H2OC-
CH2), 119.44 (CN), 122.21 (NCsCH2sCHd), 128.75 (H2OCsCH2s
CHd), 173.44 (CO2H); IR (CHCl3) cm-1 2500-3500, 2255, 1725; mass
FAB- (GT), m/e (relative intensity) 124 (100), 80 (25), 46 (30).
Enzymatic Hydrolysis of 2-Methylglutaronitrile 17a. Yield after
extraction ) 91% of 4-cyanopentanoic acid 17c: 1H NMR (DMSO
d6) δH 1.23 (d, 3H, CH3-CH), 1.79 (td, 2H, NC-CH(CH3)-CH2-), 2.35
(td, 2H, -CH2-CO2H), 2.84 (m, 2H, -CH-CN), 10.89 (s, 1H, CO2H);
13C NMR (DMSO d6) δC 18.08 (CH3-CH), 24.94 (NC-CH-), 29.33
(CH2-CH2-CO2H), 31.89 (CH2-CO2H), 123.78 (CN), 174.25 (s, 1H,
CO2H); IR (CHCl3) cm-1 2500-3500, 2240, 1735; mass FAB- (GT):
m/e (relative intensity) 126 (100), 54 (5).
Enzymatic Hydrolysis of 2-(1-Ethoxyethoxy)-4-methylthio-
butanenitrile (5a). HPLC analysis had indicated that all the starting
material had been converted into a single more polar product (water/
acetonitrile 95:5 (v/v)/0.05% (v/v) trifluoroacetic acid; λ ) 210 nm;
flow ) 1 mL/min; column: Nucleosil RP C18 (5 µm, 100 Å). A liquid
was isolated (4.54 g, 20.4 mmol) and shown by 1H NMR spectroscopy
to be 2-(1-ethoxyethoxy)-4-methylthiobutanoic acid 5c: yield, 93%;
1H NMR (DMSO d6) δH 1.09 (td, 3H, CH3-CH2-O), 1.18 (dd, 3H,
CH3-CH), 1.72-1.84 (m, 2H, S-CH2-CH2), 2.10 (s, 3H, CH3-S), 2.52
(t, 2H, S-CH2-CH2), 3.51 (m, 2H, O-CH2-CH3), 3.84 (m, 1H, O-CH-
CO2H), 4.72 (qd, 1H, O-CH (CH3)-O), 11.54 (s, 1H, CO2H); 13C NMR
(DMSO d6) δC 15.42 (CH3-S), 16.1 (CH3--CH2-O), 21.32 (CH3-CH),
30.07 (S-CH2-CH2), 33.89 (S-CH2-CH2), 62.00 (CH3-CH2-O), 77.06
(CH(CO2H)-O), 99.49 (O-CH(CH3)-O), 176.42 (CO2H); IR (CHCl3)
cm-1 2500-3500, 1725; mass FAB- (NBA), m/e (relative intensity)
221 (100), 152 (40), 122 (30), 46 (65).
Enzymatic Hydrolysis of 2-Methoxymethoxy-4-methylthio-
butanenitrile (6a). HPLC analysis indicated that all the starting material
had been converted into a single more polar product (water/acetonitrile
80:20 (v/v)/0.05% (v/v) trifluoroacetic acid; λ ) 210 nm; flow ) 1
mL/min; column: Nucleosil RP C18 (5 µm, 100 Å). A liquid was
1
isolated (2.98 g, 15.5 mmol) and shown by H NMR spectroscopy to
1
be 2-methoxymethoxy-4-methylthiobutanoic acid 6c: yield, 70%; H
NMR (DMSO d6) δH 1.96 (m, 2H, S-CH2-CH2), 2.05 (s, 3H, CH3-S-
CH2), 2.59 (t, 2H, S-CH2-CH2), 3.31 (s, 3H, CH3-O), 3.91 (m, 1H,
O-CH-CO2H), 4.60-4.72 (dd, 2H, O-CH2-O), 11.32 (CO2H); 13C NMR
(DMSO d6) δC 15.37 (CH3-S), 30.80 (CH3-S-CH2), 33.60 (S-CH2-CH2)
55.74 (CH3-O), 76.55 (O-CH-CO2H), 95.44 (O-CH2-O), 176.13 (CO2H);
IR (CHCl3) cm-1 2500-3500, 1720; mass FAB- (NBA), m/e (relative
intensity) 193 (70), 149 (55), 115 (20), 46 (100).
Enzymatic Hydrolysis of Isobutanenitrile 7a. Yield in isobutanoic
acid ) 71% 7c: 1H NMR (DMSO d6) δH 1.05 (d, 6H, (CH3)2-CH),
2.40 (h, 1H, (CH3)2-CH), 11.74 (s, 1H, CO2H); 13C NMR (DMSO d6)
δC 19.61 ((CH3)2-CH), 33.90 (CH3)2-CH), 178.7 (s, 1H, CO2H); IR
(CHCl3) cm-1 2500-3500, 1715; mass FAB- (GT), m/e (relative
intensity) 87 (100), 79 (20).
Enzymatic Hydrolysis of Acrylonitrile (8a). Yield in acrylic acid
) 93% 8c: 1H NMR (DMSO d6) δH 5.90-6.13 (m, 2H, CH2dCH),
6.32 (dd, 1H, CH2dCH), 12.43 (s, 1H, CO2H); 13C RMN (DMSO d6)
δC 130.7 (CH2dCH), 131.45 (CH2dCH), 171.16 (CO2H); IR (CHCl3)
cm-1 2500-3500, 1695; mass FAB- (GT), m/e (relative intensity) 71
(70), 53 (20). NMR analysis revealed 0.8% of acrylamide.
Enzymatic Hydrolysis of (()-2-Acetoxy-3-butenenitrile 9a. Yield
after extraction ) 92% of (()-2-acetoxybut-3-enoic acid 9c: NMR
analysis showed 6% of (()-2-hydroxy-but-3-enoic acid; 1H NMR
(DMSO d6) δH 2.01 (s, 3H, CH3-CO2), 4.87-5.25 (m, 3H, CH2dCHs
CH), 5.93-6.01 (m, 1H, CH2dCHsCH), 12.33 (CO2H); 13C NMR
(DMSO d6) δC 21.77 (CH3-CO2), 76.69 (CH2dCHsCH), 115.12
(CH2dCH), 135.59 (CH2dCH), 171.24 (CH3-CO2), 176.86 (CO2H);
IR (CHCl3) cm-1 2500-3500, 1740, 1715; mass FAB- (GT), m/e
(relative intensity) 143 (100), 59 (30).

Synthesis of Methionine Hydroxy Analogue Derivatives
J. Agric. Food Chem., Vol. 52, No. 26, 2004 8159
Figure 2. Dissociation equilibrium of cyanohydrins.
RESULTS AND DISCUSSION
As reported elsewhere (34), the ATCC8750 nitrilase is
capable of converting racemic mandelonitrile to R-(-)-mandelic
acid without the concomitant production of mandelamide
indicating that this nitrilase is highly enantioselective for
R-mandelonitrile. This means that the S-mandelonitrile remain-
ing in the reaction was spontaneously racemized by the chemical
equilibrium and supplied as substrate. This finding suggested
the possibility of establishing a new method for enantiomerical
pure R-hydroxycarboxylic acids.
In the present work, and to evaluate the efficiency of our
immobilized biocatalyst, we first tested racemic mandelonitrile
1a, O-acetyl mandelonitrile 2a, and 2-hydroxy-4-methylthio-
butanenitrile 3a for the production of mandelic acid 1c, O-acetyl
mandelic acid 2c, and HMB 3c. The O-acetyl mandelonitrile
was used to avoid the retro conversion of the cyanohydrins.
Indeed, studies showed that this nitrilase was inhibited by HCN
and the aldehyde resulting from the decomposition of cyano-
hydrins in aqueous solution (Figure 2). The latter are, respec-
tively, regarded as irreversible and competitive inhibitors of the
nitrilase from Alcaligenes faecalis ATCC8750 (34).
Figure 3. Time course in minutes of the 2-hydroxy-4-methylthiobutanoic
acid-producing reaction by the immobilized nitrilase from A. faecalis
ATCC8750. A reaction mixture (750 mL) containing 75 mmol of 3a and
120
with shaking for 47 h, after which the metabolites were analyzed as
described in the Materials and Methods (symbols 3a, 3c). Initial
concentration 0.1 M, ionic strength 0.01 M, mass ratio of substrate/dry
catalyst 0.55.
µmol of potassium phosphate buffer (pH 7.3) was incubated (30 °C)
O
b
)
Table 1. Enzymatic Hydrolysis of Nitriles 3a−6a
substrate
incubation time (h)
product(s)
yielda (%)
We compared enantiomerical results of these acids obtained
in the same experimental conditions (see Materials and Meth-
ods). We found that the immobilized nitrilase produced 1c with
high optical specificity. This study denoted that pH conditions
were an important parameter for the optical purity of R-(-)-
mandelic acid 1c. When mandelonitrile was incubated at various
pHs from 7 to 8.5, optical purity between 60 and 99% ee was
obtained. Dissociation of mandelonitrile to the corresponding
aldehyde and hydrogen cyanide showed pH dependence, and
the racemic reaction itself was faster under alkaline conditions,
although the mandelic acid production by the responsible
enzyme was accelerated under acidic conditions. As a result,
R-(-)-mandelic acid with highest optical purity was obtained
around pH 8.5 thus confirming the synergic effect of racem-
ization on enantioselectivity. We next examined the production
of 2c and 3c in the same conditions at pH 7.3 to minimize
racemization. When 2a and 3a were used as substrates, 2c and
3c were produced with 0 versus 60% ee for 1c. In conclusion,
our results for 2a and 3a less subject to racemization than 1a
suggested that the aromatic ring was essential for enantio-
selectivity with this class of substrate. Moreover, it may be
suggested that enantioselectivity mechanism is more complicated
and that a second interaction by hydrogen bonding between the
substrate and the residues at the active enzyme site are also
necessary.
Insofar as the absence of enantioselectivity for the conversion
of 3a to 3c was not a critical point for the access to methionine
hydroxy analogues derivatives that are bioavailable in their two
forms L an D, we optimized laboratory hydrolysis conditions
with 3a as the reference substrate (see Materials and Methods),
and we transposed these conditions for the hydrolysis R-ac-
etoxynitrile 4a and R-alkoxynitriles 5a-6a.
The enzyme efficiently attacked cyanohydrin 3a as shown,
for example, in Figure 3, representing the time course of the
2-hydroxy-4-methylthiobutanoic acid-producing reaction by the
immobilized nitrilase from A. faecalis ATCC8750. In these
conditions, the total bioconversion at a 0.1 M scale of 1a to 1c
was realized in 47 h.
3a
4a
5a
6a
47
33
115
120
3c
4c, 2c
5c
99
82, 6
93
6c
70
a Yield obtained after filtration of the biocatalyst, acidification (pH 1.5 with HCl
10 N), extraction with ethyl acetate or ether, and re-extraction of aqueous portion
before it was made basic (pH 10 with 2 M NaOH). See Materials and Methods.
Treatment of raw data ([substrate]/initial rate vs [substrate])
revealed Michaelis-Menten kinetics. In these experimental
conditions, the nitrilase appears to have a strict specificity for
3a estimated to 3600 µmol/h/mg of dry immobilized catalyst.
This value takes into account that the retrogradation of the
substrate at pH 7.3 is in conformity with the previous work
(36).
Second, the experimental protocol was transposed to hy-
drolysis of nitriles 3a-6a. Both reactions were carried out under
the same conditions with 0.l M concentration of substrate. The
results from substrates 3a-6a are summarized in Table 1.
Similarly, (R,S)-2-acetoxy-4-methylthiobutyronitrile 4a was
hydrolyzed, albeit more slowly, under the same reaction
conditions and gave the corresponding (R,S)-2-acetoxy-4-
methylthiobutanoic acid 4c (yield 82%), and HPLC analysis
revealed the presence of 6% acid 3c. The formation of this
product is thought to be due to the hydrolysis ester function by
an esterase. This result confirms that ester hydrolysis is due to
the action of a lipase initially present in the host microorganism.
In a kinetic point of view, the bioconversion rate of 4a is
distinctly lower than 3a (Figure 4). The ratio of initial rates of
bioconversion is close to 5. But this value is certainly
underestimated because of the inhibition of HCN and the
aldehyde, which decreases nitrilase activity during 3a hydrolysis.
Upon incubation of 5a and 6a with the catalyst, under the
usual conditions, the corresponding acids 5c and 6c were also
obtained. In addition, these reactions have to be run more slowly
than 3a and 4a. We did not detect the hydrolysis of the ether
bond implying that the biocatalyst has no etherase activity.

8160 J. Agric. Food Chem., Vol. 52, No. 26, 2004
Rey et al.
Table 2. Enzymatic Hydrolysis of Nitriles 7a−12a
substrate
reaction duration (h)
yielda (%)
7a
8a
9a
10a
11a
12a
58
9
27
25
47
7
71
93
92
86
80
94
a Yield in extracted acid after total bioconversion of the nitrile (except for the
nitrile 7a) and recovery of the acid. See Materials and Methods.
Table 3. Enzymatic Hydrolysis of Nitriles 13a−16a
substrate
reaction duration (h)
conversion (%)
yielda (%)
13a
14a
15a
16a
17a
51
53
25
25
96
89
91
100
98
87
90
72
82
91
Figure 4. Evolution of nitrile concentrations
experimental conditions ([nitrile] 0.1 M, pH 7.3, 30
per g of substrate).
O
3a,
b4a in the similarly
°C, 0.55 g of dry catalyst
92
a Yield isolated. Yield in extracted acid after total bioconversion of the nitrile
and recovery. See Materials and Methods.
Table 4. Experimental Conditions: Amounts of Substrate, Catalyst,
and Buffer Used, as well as Reaction Times and Recovery Procedure
Figure 5. Et3B induced radical addition of thiol on alkene.
amount of
sub-
strate
substrate
(g/mmol)
amount of
catalyst (g) buffer (mL)
amount of
reaction
time (h)
recovery
procedure
Furthermore, no amide was detected at any time during the
reaction, indicating the absence of nitrile hydratase.
1a
2a
3a
(0.54/4)
(0.78/4)
(9.86/75)
2.2
3.18
5.4
500
500
750
24
24
48
see general
see general
HPLC analysis
before recovery
HPLC analysis
before recovery
HPLC analysis
before recovery
HPLC analysis
before recovery
see general
see general
see general
see general
see general
see general
see general
see general
see general
see general
see general
The nitriles 5a and 6a exhibit anomalous behavior when we
compared the rates of bioconversion of 3a and 4a. This is all
the more remarkable since 3a and 5a-6a differ only by the
presence of the O-alkyl group. In an attempt to rationalize the
results obtained from all the R-hydroxynitriles substrates, we
propose the two following explanations. First, we rationalize
these results on the basis of an increasing steric hindrance close
to the nitrile function. Second, we suggest that the nitrilase
exhibit high substrate specificity for R-hydroxynitriles confirm-
ing that the R-hydroxyl group is involved in the activation
process.
Thus, these results open unlimited prospects for the synthesis
of the hydroxy-analogue of methionine 3c, which find their
applications in agrochemical (42) or pharmaceutical (43) fields
or in what concerns us directly nutritionally (36). Moreover,
cyanohydrin precursors can be obtained quantitatively in an
elegant way by radical addition of thiols (aliphatic or aromatic)
on allylic cyanohydrins (Figure 5) (44).
4a
5a
6a
(13/75)
7.15
2.45
2.10
750
220
220
48
115
120
(4.45/22)
(3.80/22)
7a
8a
9a
(5.21/75)
(4.00/75)
(9.40/75)
(7.80/50)
(3.81/30)
(8.50/75)
(5.00/75)
(8.02/75)
(5.90/75)
(5.30/75)
(8.27/75)
2.86
2.30
5.20
4.50
4.5
4.70
4.50
4.50
3.25
3.00
4.56
750
750
750
500
300
500
500
500
750
500
750
58
9
9
10a
11a
12a
13a
14a
15a
16a
17a
26
47
7
51
53
25
25
96
In front of these performances and to generalize the use of
the immobilized catalyst, we continued the study of the
chemioselectivity of the biocatalyst by testing the hydrolysis
of nitriles (7a-12a). This choice of substrates although non-
exhaustive allows nevertheless a sufficiently broad functional
sweeping. It can be divided into three classes: (i) aliphatic 7a
and vinyl nitrile 8a; (ii) allylic nitrile-ester 9a and nitrile-
ether 10a-11a; and (iii) nitrile-ester 12a.
Nitriles 7a-12a were hydrolyzed under operating conditions
identical to the hydrolysis of cyanohydrins 3a to free us from
the influence of the experimental conditions on the bioconver-
sions and thus to be able to compare directly the results (Table
2).
biotransformation of nitrile 9a leads to the privileged formation
of the acetylated allylic acid 9c with a small amounts of the
allylic hydroxyacide (<6% molar).
These results testify to the important chemioselectivity of this
immobilized biocatalyst. To complete the range of the substrates
of the catalyst, we tested symmetrical 13a-16a and dissym-
metrical 17a dinitriles (Table 3). The products resulting from
the monohydrolysis thus offer an access to the monocyanoacids.
The latest one is particularly appreciated in polymer chemistry
(48).
1
NMR H and 13C analyses revealed on the totality of the
undertaken experiments the absence of diacid corresponding to
the total hydrolysis of the dinitrile. This finding suggests that
the monoacid is not a substrate for the nitrilase. This is certainly
due to the predominance of the carboxylate form at pH 7.3 that
is unfavorable for the access to the active site of the enzyme.
Despite a different length of aliphatic chain, times of biotrans-
The hydrolysis of the substrates 7a-12a are, except for nitrile
7a, quasi-quantitative in correct residence times. The much
slower hydrolysis of the isobutanenitrile joined the results of
Yamamoto et al. (34) concerning the difficulty of the nitrilase
to transform the aliphatic nitriles. As previously reported, the

Synthesis of Methionine Hydroxy Analogue Derivatives
J. Agric. Food Chem., Vol. 52, No. 26, 2004 8161
(10) Bennett, M. A.; Yoshida, T. Homogeneously catalyzed hydration
of nitriles to carboxamides. J. Am. Chem. Soc. 1973, 95, 3030-
1.
(11) Paraskewas, S. Synthesis of carbamide by partial hydrolysis of
the palladium chloride-nitrile complex. Synthesis 1974, 574-
5.
(12) Chin, J.; Kim, J. H. Catalytic hydrolysis of acrylonitrile to
acrylamide under mild conditions. Angew. Chem. 1990, 102(5),
580-2.
formation of nitriles 13a and 14a are relatively close. The
modest yield obtained at the time for the hydrolysis of the
fumaronitrile 15a is justified by the partial chemical hydration
of the nitrile function into amide in the aqueous medium of
synthesis selected (pH 7.3; 30 °C). Times of bioconversion are
definitely higher than previously thought because of the low
solubility of dinitriles in aqueous medium. Among the hydrolysis
of the dinitriles 13a-16a, most remarkable and especially richest
of application is that obtained at the time of the hydrolysis of
the 2-methylglutaronitrile 17a. Indeed, its hydrolysis leads
exclusively to the cyanopentano¨ıc acid (NC-CH(CH3)-CH2-CH2-
CO2H). This highly regioselective reaction confirms the dis-
criminatory effect of the steric hindrance near the nitrile
function.
Conclusion. In summary, we have shown that the im-
mobilized and genetically modified nitrilase system from A.
faecalis ATCC8750 is a useful catalyst for the chemioselective
preparation of R-hydroxy and R-hydroxy-substituted carboxylic
acids under mild conditions. These carboxylic acids should have
an important interest in cattle feeding.
We showed that in addition to its use in synthesis derived
from HMB, it can also be of a great help for the preparation of
monocyanoacids. The results obtained testify to an exemplary
chemoselectivity. Concerning the stereoselectivity, several stud-
ies are currently underway in our laboratory to confirm the
assumptions announced previously. Thus, we try to rationalize
the enantioselective exit of the hydrolysis of substrate. Conse-
quently, the number growing of published works on the field
testifies in an obvious way of this awakening to the potential
applications related to the use of biocatalysts with nitrilasic
activity. The advent of new methodologies, based on the use
of the genetic engineering to the production of enzymes, should
make the enzymatic tool even more powerful. It is perfectly
conceivable that research of perfectibility, which can result in
particular in a better stability of the catalytic species, will allow
the nearest use of this type of catalyst in the industrial processes.
(13) Taillades, J.; Sola, R.; Brugidou, J.; Previero, A.; Commeyras,
A. Supported ketonic compounds as catalysts for the dehydration
of R-aminonitriles. Tetrahedron Lett. 1983, 24, 1501-4.
(14) Taillades, J.; Pascal, R.; Commeyras, A. Strecker and related
systems. X, Decomposition, and hydration of secondary R-ami-
nonitriles in aqueous basic media. Autocatalytic hydration
process and catalysis by acetone. Tetrahedron 1978, 34, 2275-
81.
(15) Jammot, J.; Pascal, R.; Commeyras, A. Process for the catalytic
hydration of a cyanohydrin. Fr. Patent 2 631 338, 1988.
(16) Arthaud, I.; Chatel, S.; Chauvin, A. S.; Bonnet, D.; Kopf, M.
A.; Leduc, P. Nitrile hydratase and related non-heme iron sulfur
complexes. Coord. Chem. ReV. 1999, 190-192, 577-86.
(17) Thompson, L. A.; Knowles, C. J.; Linton, E. A.; Wyatt, J. M.
Microbial transformations of nitriles. Chem. Brit. 1988, 24(9),
900-2, 912.
(18) Nagasawa, T.; Yamada, H. In Biocatalysis; Abramowicz, D. A.,
Ed.; Van Nostrand Reinhold: New York, 1990; p 277.
(19) Nagasawa, T.; Yamada, H. Microbial transformations of nitriles.
Trends Biotechnol. 1989, 7(6), 153-8.
(20) Kobayashi, M.; Yanaka, N.; Nagasawa, T.; Yamada, H. Mono-
hydrolysis of an aliphatic dinitrile compound by nitrilase from
Rhodococcus rhodochrous K22. Tetrahedron 1990, 46, 5587-
90.
(21) Bengis-Garber, C.; Gutman, A. L. Selective hydrolysis of
dinitriles into cyano-carboxylic acids by Rhodococcus rhodo-
chrous N.C.I.B. 11216. Appl. Microbiol. Biotechnol. 1989, 32,
11-6.
(22) Kakeya, M.; Sakai, N.; Sugai, T.; Ohta, H. Microbial hydrolysis
as a potent method for the preparation of optically active nitriles,
amides, and carboxylic acids. Tetrahedron Lett. 1991, 32, 1343-
6.
ACKNOWLEDGMENT
(23) Arnaud, A.; Galzy, P.; Jallageas, J. C. Production of stereospecific
R-amino acids by biological hydrolysis of racemic R-amino
nitriles. Bull. Soc. Chim. Fr. 1980, 87-90.
(24) Tadros, Z.; Lagrifoul, P. H.; Mion, L.; Taillades, J.; Commeyras,
A. Enantioselective hydration of R-aminonitriles using chiral
carbonyl catalysts. J. Chem. Soc. Chem. Commun. 1991, 19,
1373-5.
(25) Linton, E. A.; Knowles, C. J. Utilization of aliphatic amides and
nitriles by Nocardia rhodochrous LL100-21. J. Gen. Microbiol.
1986, 132, 1493-501.
(26) Jallageas, J. C.; Arnaud, A.; Galzy, P. Bioconversions of nitriles
and their applications. AdV. Biochem. Eng. 1980, 14, 1-32.
(27) Asano, Y.; Fujishiro, K.; Tani, Y.; Yamada, H.; Microbial
degradation of nitrile compounds. Part V. Aliphatic nitrile
hydratase from Arthrobacter sp. J-1. Purification and character-
ization. Agric. Biol. Chem. 1982, 46, 1165-74.
(28) Hjort, C. M.; Godtfrendsen, S. E.; Emborg, C.; Isolation and
characterization of a nitrile hydratase from a Rhodococcus sp.
J. Chem. Technol. Biotechnol. 1990, 48(2), 2617-48.
(29) Nakajima, Y.; Doi, T.; Satoh, Y.; Fujiwara, A.; Watanabe, I.
Photoactivation of nitrile hydratase in Corynebacterium sp.
N-774. Chem. Lett. 1987, 9, 1767-70.
(30) Nagasawa, T.; Nanba, H.; Yamada, H. Nitrile hydratase of
Brevibacterium R312-purification and characterization. Bio-
chem. Biophys. Res. Commun. 1986, 139(3), 1305-12.
(31) Nagasawa, T.; Nanba, H.; Ryuno, K.; Takenchi, K.; Yamada,
H. Nitrile hydratase of Pseudomonas chlororaphis B23. Purifica-
tion and characterization. Eur. J. Biochem. 1987, 162(3), 691-
8.
We thank Adisseo for the generous donation of the biocatalyst.
LITERATURE CITED
(1) Davies, H. G.; Green, R. H.; Kelly, D. R.; and Roberts, S. M.
Biotransformation in preparatiVe Organic Chemistry; Academic
Press, London, 1989. Faber, K. Biotransformations in Organic
Chemistry; Springer-Verlag: New York, 1992.
(2) Santaniello, E.; Ferraboschi, P.; Grisenti, P.; Manzocchi, A. The
biocatalytic approach to the preparation of enantiomerically pure
chiral building blocks. Chem. ReV. 1992, 92(5), 1071-140.
(3) Klibanov, A. M. Asymmetric transformations catalyzed by
enzymes in organic solvents. Acc. Chem. Res. 1990, 23(4), 114-
20.
(4) The Chemistry of the Cyano Group; Wiley-Interscience: New
York, 1970; pp 77-86.
(5) Strecker, A. Liebigs Ann. Chem. 1850, 75, 27.
(6) Clarke, H. T.; Read, R. R. Modification of the Sandmeyer
synthesis of nitriles. J. Am. Chem. Soc. 1924, 46, 1001.
(7) Sandler, S. R.; Kano, W. Organic Functional Group Preparation;
Academic Press: New York, 1968; Ch. 17, p 453.
(8) Oldeniel, O. H.; van Leusen, D.; van Leusen, A. M. Chemistry
of sulfonylmethyl isocyanides. 13. A general one-step synthesis
of nitriles from ketones using tosylmethyl isocyanide. Introduc-
tion of a one-carbon unit. J. Org. Chem. 1977, 42, 3114-8.
(9) Harrisson, C. R.; Hodge, P.; Rogers, W. J. Conversion of
carboxamides and oximes to nitriles or imidoyl chlorides using
a polymer-supported phosphine and carbon tetrachloride. Syn-
thesis 1977, 41-3.

8162 J. Agric. Food Chem., Vol. 52, No. 26, 2004
Rey et al.
(32) Sugiura, Y.; Kuwahara, J.; Nagasawa, T.; Yamada, H. Nitrile
hydratase. The first non-heme iron enzyme with a typical low-
spin iron(III)-active center. J. Am. Chem. Soc. 1987, 109(19),
5848-50.
(33) Hook, R. H.; Robinson, W. G. Ricinine nitrilase. I. Reaction
product and substrate specificity. J. Biol. Chem. 1964, 239(12),
4257-62.
(41) Rajesh, M.; Kapela, S.; Nam, P.; Forciniti, D.; Lorbert, S.;
Schasteen, C.; Enzymatic synthesis and characterisation of
L-methionine and 2-hydroxy-4-methylthiobutanoic acid co-
oligomers. J. Agric. Food. Chem. 2003, 51, 2461-7.
(42) Monsanto, U.S. Patent 2 938 053, 1960.
(43) Baker, D. H.; Fahnenstich, R.; Kleeman, A.; Marten, J. Fr. Patent
79 32086, 1979.
(34) Yamamoto, K.; Fujimatsu, I.; and Komatsu, K. Purification and
characterization of the nitrilase from Alcaligenes faecalis ATCC
8750 responsible for enantioselective hydrolysis of mandeloni-
trile. J. Ferm. Bioeng. 1992, 73(6), 425-30.
(35) Yamamoto, K.; Fukimatsu, I.; Oishi, K.; Komatsu, K.; Production
of R-(-)-mandelic acid from mandelonitrile by Alcaligenes
faecalis ATCC 8750. Appl. EnViron. Microbiol. 1991, 57(10),
3028-38.
(36) Aventis Animal Nutrition, U.S. Patent 6 180 359.
(37) Pas. Bernard, J. K. Supplemental animal marine preotein blend
and rumen protected methionine hydroxy analogue for lactating
dairy cows. Prof. Anim. Sci. 1997, 13, 149-54.
(38) Prange, R. W.; Satter, L. D.; Jones, D. A. Degradation of
methionine hydroxy analogue in the rumen of lactating cows. J.
Dairy. Sci. 1988, 71, 525-9.
(44) Taillades, J.; Rey, P.; Garrait, M.; Gros, G. Fr. Patent FR2 824 823,
2002.
(45) Taillades, J.; Rey, P.; Garrait, M.; Gros, G. Eur. Patent EP1
260 500, 2002.
(46) Dickey, J. B.; Duennebier, F. C. U.S. Patent 2 439 081, 1948.
(47) Jacobson, R. M.; Lahm, G. P.; Clader, J. W. J. Org. Chem. 1980,
45, 395-405.
(48) Mander, L. N.; Turner, J. V. Tetrahedron Lett. 1981, 22, 3683-
6.
Received for review July 15, 2004. Revised manuscript received
September 23, 2004. Accepted October 5, 2004. We thank the Centre
National de la Recherche Scientifique and Adisseo for financial support
of this work.
(39) Aventis Animal Nutrition, U.S. Patent 6 372 788.
(40) Moncoulon, R.; European Patent Nï 0 628 257 A1, 1994.
JF048827Q