Nitrile hydratase activity of a recombinant nitrilase

Article abstract of DOI:10.1002/adsc.200600269

Appreciable amounts of amide are formed in the course of nitrile hydrolysis in the presence of recombinant nitrilase from Pseudomonas fluorescens EBC 191, depending on the α-substituent and the reaction conditions. The ratio of the nitrilase and nitrile hydratase activities of the enzyme is profoundly influenced by the electronic and steric properties of the reactant. In general, amide formation increased when the α-substituent was electron-deficient; 2-chloro-2-phenylacetonitrile, for example, afforded 89% amide. We found, moreover, that (R)-mandelo-nitrile was hydrolysed with 11% of amide formation whereas 55% amide was formed from the (S)-enantiomer; a similar effect was found for the O-acetyl derivatives. A mechanism that accomodates our results is proposed.

Full text of DOI:10.1002/adsc.200600269

FULL PAPERS
DOI: 10.1002/adsc.200600269
Nitrile Hydratase Activity of a Recombinant Nitrilase
Bruno C. M. Fernandes,a Cesar Mateo,a Christoph Kiziak,b,c Andrzej Chmura,a
Jan Wacker,b Fred van Rantwijk,a Andreas Stolz,b and Roger A. Sheldona,*
a
Laboratory of Biocatalysis and Organic Chemistry, Delft University of Technology, Julianalaan 136, 2628 BL Delft, The
Netherlands
Fax : (+31)-15-278-1415; e-mail: r.a.sheldon@tudelft.nl
Institut für Mikrobiologie, Universität Stuttgart, Allmandring 31, 70550 Stuttgart, Germany
Present address: Lonza AG, Biotechnology Research and Development, 3930 Visp, Switzerland
b
c
Received: June 8, 2006; Accepted: October 18, 2006
Abstract: Appreciable amounts of amide are formed 89% amide. We found, moreover, that (R)-mandelo-
in the course of nitrile hydrolysis in the presence of nitrile was hydrolysed with 11% of amide formation
recombinant nitrilase from Pseudomonas fluorescens whereas 55% amide was formed from the (S)-enan-
EBC 191, depending on the a-substituent and the re- tiomer; a similar effect was found for the O-acetyl
action conditions. The ratio of the nitrilase and ni- derivatives. A mechanism that accomodates our re-
trile hydratase activities of the enzyme is profoundly sults is proposed.
influenced by the electronic and steric properties of
the reactant. In general, amide formation increased Keywords: amide formation; enantiomer selectivity;
when the a-substituent was electron-deficient; 2- mechanism; nitrilase; Pseudomonas fluorescens EBC
chloro-2-phenylacetonitrile, for example, afforded 191; selectivity
Introduction
present in microbes and involves a hydration step in
the presence of a nitrile hydratase (NHase, E.C.
Nitriles are important intermediates in synthetic 4.2.1.84), which yields the corresponding amide, fol-
chemistry because they are simple to prepare and lowed by amidase-catalysed hydrolysis of the latter.
give access to a wide variety of interesting carboxylic
In contrast with common wisdom, which dictates
acid derivatives. Chemical hydrolysis of nitriles is well that NLases transform nitriles into the carboxylic
established but requires harsh conditions that may ne- acids exclusively, reports can be found, scattered in
cessitate the protection of any sensitive functional literature as far back as the 1960s, of modest amounts
groups and leads to the generation of large amounts of amides being formed in the presence of nitrilas-
of waste. A preferred option would be to conduct the es.[2–4] This side-activity, which is not accounted for by
hydrolysis of the nitriles enzymatically, at mild reac- the commonly accepted nitrilase mechanism,[3] has
tion conditions with regio- and/or enantioselectivity largely been neglected. Presumably, amide formation
as an additional advantage.
often was ascribed to the presence of a contaminating
It is generally accepted that Nature employs two nitrile hydratase. In many cases, amide formation may
enzymatic pathways for nitrile hydrolysis (see have gone unnoticed, either because a contaminating
Figure 1).[1] In the direct pathway a nitrilase (NLase, amidase hydrolysed the amide by-product into the
E.C. 3.5.5.1) catalyses the addition of two molecules acid or because the detection of the catalytic activity
of water to give the carboxylic acid. Nitrilases are or the selection of catalysts depended on ammonia re-
widely found in Nature and have been isolated from lease, which only occurs upon full hydrolysis into the
microbes and plants. The second pathway is mainly acid.
In a recent characterisation study of a purified, re-
combinantly expressed NLase from Arabidopsis thali-
ana (AtNIT4), >60% amide was formed in the
course of the hydrolysis of 3-cyano-l-alanine, the pre-
sumed natural substrate.[5] A different recombinant
NLase from A. thaliana (AtNIT1) has been the sub-
ject of the most complete study up to date, which
Figure 1. Pathways of enzymatic nitrile hydrolysis.
showed the effects of electron-withdrawing substitu-
Adv. Synth. Catal. 2006, 348, 2597 – 2603
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2597

Bruno C. M. Fernandes et al.
FULL PAPERS
ents on amide formation.[6,7] Thus, hydrolysis of 2-cro- phenylacetamide (3a). From the time-course of the
tonitrile in the presence of AtNIT1 afforded 99% reaction (see Figure 3), it becomes apparent that 2a
crotonic acid, whereas 95% amide was formed from and 3a are formed concurrently. Temporary accumu-
the electron-deficient 3-nitroacrylonitrile.[7] It is worth
noting in this respect that the cysteine-dependent pro-
tease papain has been transformed, by changing one
amino acid residue, into a modestly active nitrile hy-
dratase.[8] We hypothesise, on the basis of the experi-
mental evidence, that the nitrilase and nitrile hydra-
tase activities in NLases are intimately related.
Nitrilase-catalysed amide formation deserves to be
studied in detail as it may have serious implications
regarding the yield and purity of the desired product
and complicates the downstream processing. Further-
more, highly selective NLase-mediated amide synthe-
Figure 3. Hydrolysis of 2-phenylacetonitrile (1a, 10 mM) in
the presence of PfNLase (35.5 mgmLÀ1) in 100 mM phos-
sis, assuming that this latter activity can be controlled,
would afford a synthetic route to enantiomerically
pure amides. As such, it could be a valuable addition
to the synthetic repertoire as NLases are often enan-
tioselective, whereas it is an exception with NHases.
~
*
phate buffer pH6 at 108C; acid ( ), amide ( ).
lation of amide and subsequent hydrolysis are not ob-
The synthetic relevance of amide formation in the served. Hence, 3a is not an intermediate in the hy-
presence of NLases prompted us to study this latter drolysis of 1a into 2a, as with nitrile hydratase/ami-
side-activity in more detail, to get a better under- dase systems. It is pertinent to note that PfNLase
standing of the factors that influence the amide for- does not hydrolyse 3a, neither does it mediate the for-
mation, such as steric requirements and electron den- mation of 3a from 2a and ammonia.
sity. For this purpose we employed a recombinantly
Summarising, amide and acid are both bona fide re-
produced nitrilase originating from Pseudomonas flu- action products that are formed directly from the ni-
orescens EBC 191 (PfNLase), which is known to be trile in a ratio of 23:1. This latter parameter, the acid/
prone to amide formation.[9] It is worth noting that amide ratio (Ac/Am), will be used throughout the
this latter enzyme is an arylacetonitrilase and toler- paper to characterise the selectivity of PfNLase.
ates quite bulky groups on the a-position, in contrast
We found that the reaction pH, when varied from 5
to the NLases from Arabidopsis thaliana. The use of to 9, influenced the acid/amide ratio only slightly
an enzyme expressed in E. coli ensures that a single (Figure 4a), with lower pH slightly reducing the
nitrile hydrolysing enzyme is present as the non-trans- amount of amide. The effect of the reaction tempera-
formed E. coli strain shows no nitrile hydrolysing ac-
tivity.
Results and Discussion
Hydrolysis of 2-Phenylacetonitrile
Hydrolysis of 2-phenylacetonitrile (1a, see Figure 2),
a simple model substrate, in the presence of PfNLase
afforded, besides the acid 2a, a minor amount of 2-
Figure 4. Acid/Amide ratio for PfNLase (35.5 mgmLÀ1) in
the hydrolysis of 1a (10 mM) in 100 mM phosphate buffer;
a: effect of pH at 258C; b: temperature effect at pH 6.
Figure 2. Hydrolysis of nitriles using a nitrilase.
2598
www.asc.wiley-vch.de
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2006, 348, 2597 – 2603

Nitrile Hydratase Activity of a Recombinant Nitrilase
FULL PAPERS
ture was more pronounced (see Figure 4b) and Ac/ enantiomerically pure (S). This result strongly indi-
Am changed from 11 at 58C to 38 at 408C. Measure- cates that the absolute configuration of the a-carbon
ments of the initial rates of formation of 2a and 3a strongly influences the Ac/Am ratio but, unfortunate-
(data not shown) showed that acid formation is more ly, this effect could not be investigated in detail as 1c
temperature-dependent, which indicates a higher acti- racemised under the reaction conditions.
vation barrier for the formation of acid. These results
suggest that elevated temperature and low pH steer
the reaction towards acid, whereas a low temperature Effects of the a-Substituent Size and the Absolute
and increased pH bend the selectivity towards the Configuration
amide.
Enantiomerically pure mandelic acid (2d) is industri-
ally produced as a chiral intermediate in the synthesis
Electronic Effects of the a-Substituent
of fine chemicals and as a chiral auxiliary in industrial
resolutions. (R)-2d is produced[11] on a multiton scale
It has been shown that the Ac/Am ratio of a nitrilase via enantioselective hydrolysis of mandelonitrile (1d)
from Arabidopsis thaliana (AtNIT1) was highly sensi- in the presence of a nitrilase.[12] The pure enantiomers
tive to electronic effects.[7] We wished to investigate of 1d are readily accessible via hydrocyanation of
the behaviour of PfNLase in this respect by substitut- benzaldehyde in the presence of the appropriate oxy-
ing the a-position in 1a with similarly sized groups of nitrilase;[13] hence, they are, in principle, good model
opposing electronic character. Thus, we compared 2- compounds for investigating possible effects of the
phenylpropionitrile (1b, see Figure 2) and 2-chloro-2- absolute configuration of the reactant on the Ac/Am
phenylacetonitrile (1c) and found that the electron- ratio.
deficient 1c reacts faster by a factor of 2.4.[10] The
The modest stability of 1d in neutral aqueous
methyl-substituted reactant 1b was converted into medium is an obstacle that requires careful consider-
>99% acid (2b), whereas 90% amide was formed ation. Compound 1d decomposes into benzaldehyde
from 1c (see Table 1). Hence, it would seem that the and hydrogen cyanide; the process is reversible and is
the major racemisation pathway of enantiomerically
Table 1. Electronic effects on the acid/amide ratio.[a]
pure 1d. The nitrile becomes stable at pH <4–5 and
lower temperature but such conditions also have a
strong deleterious effect on nitrilase activity. We com-
promised by investigating the hydrolysis of enantio-
merically pure (R)- and (S)-1d in the presence of
PfNLase at pH 6 and 08C. Under these conditions the
racemisation of the nitrile is much slower than the en-
Reactant
1b
1c
Initial rate [mmol minÀ1 mgÀ1]
Acid [%]
ee [%]
Amide [%]
ee [%]
Ac/Am
0.81
>99
n.d.
<1
n.d.
>99
2.13
11
82 (R)
89
>94 (S) zymatic hydrolysis. Both enantiomers reacted with
0.1
comparable rates, but with widely diverging product
selectivity (see Table 2). While the hydrolysis of (R)-
1d yielded only 11% amide (Ac/Am=8.1), the amide
(S)-3d was the major product from the hydrolysis of
(S)-1d (Ac/Am=0.8). Hence, it appears that Ac/Am
[a]
Reaction conditions: 10 mM nitrile in 100 mM phosphate
buffer pH 6 at 258C.
extent of amide increases with increasing electronega- is influenced by electronic as well as by steric factors.
tivity of the a-substituent: CH3 <H<Cl. A compara- Here, the study of enantiopure compounds gives an
ble effect of a-fluoro substitution has been reported insight into the system that is never possible by the
for Arabidopsis nitrilases.[6]
use of racemates.
Interestingly, the formation of acid (2c) from race-
A more stable derivative of 1d was required to
mic 1c was strongly biased towards the (R)-enantio- measure the effects of the pH on amide formation.
mer (see Table 1), whereas the amide (3c) was nearly The ester derivative, 2-acetoxy-2-phenylacetonitrile
Table 2. Hydrolysis 1d and 1e in the presence of PfNLase.[a]
(R)-1d
(S)-1d
(R)-1e
(S)-1e
Initial rate [mmol minÀ1 mgÀ1]
Acid [%]
Amide [%]
0.62
89
11
0.62
45
55
0.23
67
33
2
1.73
38
62
Ac/Am
8.1
0.8
0.6
[a]
Reaction conditions: 10 mM reactant, 100 mM phosphate buffer pH 6 at 08C (1d) or 10% methanol at 258C (1e).
Adv. Synth. Catal. 2006, 348, 2597 – 2603
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.asc.wiley-vch.de
2599

Bruno C. M. Fernandes et al.
FULL PAPERS
(1e),[14] does not suffer from HCN elimination but
racemisation still occurs, via deprotonation on the a-
position. Fortunately, this latter process is slow in
aqueous media at pH<7 (data not shown). Spontane-
ous hydrolysis of the ester bond in 1e (with formation
of 2d and 3d), which is significant at pH >7, re-
mained restricted to <5% but often much less, de-
pending on the pH.
Much to our surprise, the ester group profoundly
influenced the course of the enzymatic hydrolysis.
The reaction rates (see Table 2) should not be com-
pared directly as different temperatures were used,
but it is noteworthy that (R)-1e reacted three times
slower than (R)-1d, the 258C temperature difference
^
Figure 6. pH profile for the hydrolysis of (R)-1e ( ) and (S)-
^
1e ( ); reaction conditions: 10 mM reactant, 100 mM phos-
phate buffer, 258C.
notwithstanding. The enantiomeric kinetic bias shifted which leads us to the conclusion that the enantiomers
from pro-(R), with 1d, to pro-(S). Both enantiomers bind in very different ways.
produced more amide than the corresponding enan-
tiomers of the free cyanohydrin; the Ac/Am ratio of
the hydrolysis of (R)-1e was even four times lower Mechanism of Action
than that of (R)-1d. The progress curve for the sepa-
rate hydrolysis of (R)- and (S)-1e (Figure 5) illustrates Any discussion of the catalytic mechanism of nitrilas-
well that the amide is not a reaction intermediate and es is hampered by a lack of crystal structures. It is
that the two enantiomers are differently hydrolysed.
commonly accepted, nevertheless, that nitrilases
harbor a Cys-Glu-Lys triad in the active site[15] and
that the reaction takes place via a thioimidate inter-
mediate (I, see Scheme 1).[2,3,16] The active site triad
has been identified in the crystal structures of the
worm NitFhit fusion protein[17] and in a putative ni-
trile hydrolase from yeast.[18] This nitrile hydrolysis
mechanism is depicted in Scheme 1. A bridging water
molecule has been added because it has been identi-
fied in the yeast protein,[18] as well as in a structurally
closely related d-carbamoylase (N-carbamoyl-d-
amino acid amidohydrolase).[19]
Elimination of ammonia from the tetrahedral inter-
mediate (pathway A), with formation of the acyl-
enzyme intermediate III, requires a positive charge
on the nitrogen atom in the reactant, stabilised by the
Glu residue. If, in contrast, the positive charge is not
*
Figure 5. Hydrolysis profiles for (R)- and (S)-1e: (S)-2e ( ),
~
~
*
(S)-3e ( ), (R)-2e ( ), (R)-3e ( ) at pH 6.
We conclude, on the basis of these results, that on the reactant but, for example, on the Lys residue
enantiomers of chiral reactants should be treated as (tetrahedral intermediate IIb), formal thiol elimina-
separate entities, and that unambiguous conclusions tion is expected to prevail (pathway B). Such a charge
as regards selectivity issues cannot be based on race- distribution could be expected, for example, when an
mate conversions. Thus, for example, in the hydrolysis electron-demanding R group destabilises the positive
of (R,S)-1e, (S)-2e and (S)-3e are the major products charge on the reactant N. Otherwise, steric interac-
initially, with Ac/Am=0.7, whereas at total conver- tions could force the N atom away from the stabilising
sion (R)-2e and (S)-3e predominate (Ac/Am=1.1, Glu (see Scheme 1, pathway B). Summarising, NLase-
data not shown).
mediated nitrile hydrolysis into the carboxylic acid
A pH profile of the hydrolysis of 1e (Figure 6) re- and water addition to give the amide are two branch-
vealed widely diverging effects on the propensity of es of the nitrilase mechanism, as originally suggested
the two enantiomers to produce amide. Amide forma- by Hook and Robinson.[2] We propose that the charge
tion from (R)-1e was found to increase when the distribution in the tetrahedral intermediate, depend-
medium was turned acidic, whereas the Ac/Am ratio ing, in turn, on the stereochemical and electronic
of the (S)-enantiomer passed through a minimum at properties of the R group acts as a mechanistic
pH 6. Both patterns are completely different from switch.
what is seen with 1a. Changes in charge distribution
We note that amide elimination from the tetrahe-
obviously affect (R)- and (S)-1e in different ways, dral intermediate is amply supported in the chemical
2600
www.asc.wiley-vch.de
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2006, 348, 2597 – 2603

Nitrile Hydratase Activity of a Recombinant Nitrilase
FULL PAPERS
Scheme 1. Proposed nitrilase mechanism for the formation of acid (A) and amide (B).
literature. It is the major pathway in the hydrolysis of Conclusions
thioimidate esters[20] and mercaptoethanol-catalysed
nitrile hydrolysis[21–23] at neutral pH. In both cases the We have shown that Pseudomonas fluorescens nitri-
product selectivity shifted towards acid at pH <3. We lase converts nitriles into the carboxylic acid as well
found a similar pH effect, although under near-neu- as the amide. The relative formation of acid and
tral conditions and of smaller magnitude, in the enzy- amide is subject to the pH and the temperature. Elec-
matic hydrolysis of 1a.
tron-withdrawing substituents at the a-position favour
The active site does not contain bound amide at amide formation.
any time, according to our mechanism and amide,
We have shown, for the first time, that the absolute
once formed, is not hydrolysed any further. It is com- configuration at the a-position in the reactant exerts
monly observed that NLases do not hydrolyse amides, a dramatic influence on the extent of amide forma-
although there are strong indications that NLases, ali- tion.
phatic amidases and d-carbamoylases have very simi-
lar active site architectures.[24] A very modest amidase
activity (6000 times less than the NLase activity),
mainly due to a reduced Vmax, has been measured
Experimental Section
with the nitrilase from Rhodococcus rhodochrous
J1.[25] If this factor also applies to PfNLase, hydrolysis
of 3e would be in the order of 0.01% over 24 h.
Why amides cannot react via IIa into the thiol ester
intermediate (III) is not clear but an obvious explana-
tion is that IIa is too high in energy, relative to the
amide. As regards the amidase from Pseudomonas
aeruginosa, it has been suggested that only one amide
Chemicals
Soluble, His6-tagged Pseudomonas fluorescens EBC 191 ni-
trilase (PfNLase) was recombinantly produced in E. coli
JM109
(pIK9).[9] The stock solution contained approx.
A
48 mgmLÀ1 of protein, according to a Bradford test. Its ac-
tivity in the hydrolysis of racemic mandelonitrile was 3.1
mmol (mg protein)À1 minÀ1 at 08C and pH 6. 2-Phenylpropio-
nitrile (1b) from Acros was purified by distillation under re-
molecule can be bound per hexameric amidase mole- duced pressure in a Kugelrohr apparatus (Büchi) to remove
cule,[26] possibly because conformational changes in
an unknown contaminant that inhibited PfNLase. Unless
otherwise stated, the chemicals were obtained from normal
the entire hexamer are required to smooth the ener-
commercial sources and used without further purification.
(R)- and (S)-mandelonitrile [(R)- and (S)-1d] were syn-
getic pathway towards the acyl-enzyme intermediate.
As regards nitrilase catalysis, it would seem that the
enzyme conserves the chemical energy in the nitrile
triple bond, maintaining IIa and IIb well above
ground-state level, to transform the reactant smoothly
into acid and amide.
thesised via enzymatic hydrocyanation of benzaldehyde in
the presence of the appropriate oxynitrilase.[27,28] Enantio-
meric purities were verified by chiral HPLC to be 99% (R)
and 98% (S).
(R)-2-Acetoxy-2-phenylacetonitrile [(R)-1e] was synthes-
ised by acylation of (R)-1d with acetic anhydride and pyri-
dine in 1,4-dioxane at room temperature. The mixture was
Adv. Synth. Catal. 2006, 348, 2597 – 2603
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.asc.wiley-vch.de
2601

Bruno C. M. Fernandes et al.
FULL PAPERS
diluted with ether and consecutively washed with water, 1M References
hydrochloric acid, saturated sodium carbonate and brine,
dried over MgSO4 and concentrated under vacuum. The
product was isolated as a colourless liquid with ee >99%
according to chiral GC. 1H NMR (300 MHz, CDCl3): d=
2.16 (s, 3H, CH3), 6.41 (s, 1H, CH), 7.26–7.55 (m, 5H, aro-
matic); 13C NMR (75 MHz, CDCl3): d=20.4 (CH3), 62.8
(CH), 11.6 (CN), 127.8 (C-2,6), 129.2 (C-3,5), 130.3 (C-4),
131.7 (C-1), 168.9 (C=O).
(S)-2-Acetoxy-2-phenylacetonitrile [(S)-1e] was obtained
using a published procedure[29] with enantiomeric purity
93.7% according to chiral GC.
[1] T. Nagasawa, H. Yamada, Trends Biotechnol. 1989, 7,
153–159.
[2] R. H. Hook, W. G. Robinson, J. Biol. Chem. 1964, 239,
4263–4267.
[3] A. Goldlust, Z. Bohak, Biotechnol. Appl. Biochem.
1989, 11, 581–601.
[4] D. E. Stevenson, R. Feng, F. Dumas, D. Groleau, A.
Mihoc, A. C. Storer, Biotechnol. Appl. Biochem. 1992,
15, 283–302.
[5] M. Piotrowski, S. Schçnfelder, E. W. Weiler, J. Biol.
Chem. 2001, 276, 2616–2621.
[6] F. Effenberger, S. Oßwald, Tetrahedron:Asymmetry
Analysis
2001, 12, 279–285.
[7] S. Osswald, H. Wajant, F. Effenberger, Eur. J. Biochem.
2002, 269, 680–687.
[8] ꢁ. Dufour, A. C. Storer, R. Mꢂnard, Biochemistry 1995,
34, 16382–16388.
[9] C. Kiziak, D. Conradt, A. Stolz, R. Mattes, J. Klein, Mi-
crobiology 2005, 151, 3639–3648.
[10] Electron-deficiency in esters is known to increase their
reactivity towards lipases, which likewise act via gener-
al base catalysis.
[11] M. Breuer, K. Ditrich, T. Habicher, B. Hauer, M. Keß-
eler, R. Stürmer, T. Zelinski, Angew. Chem. Int. Ed.
2004, 43, 788–824.
[12] K. Yamamoto, K. Oishi, I. Fujimatsu, K.-I. Komatsu,
Appl. Environm. Microbiol. 1991, 57, 3028–3032.
[13] M. Schmidt, H Griengl, Top. Curr. Chem. 1999, 200,
193–226.
[14] a) N. Layh, A. Stolz, S. Fçrster, F. Effenberger, H.-J.
Knackmuss, Arch. Microbiol. 1992, 158, 405–411; b) U.
Heinemann, C. Kiziak, S. Zibek, N. Layh, M. Schmidt,
H. Griengl, A. Stolz, Appl. Microbiol Biotechnol. 2003,
63, 274–281.
[15] C. Brenner, Curr. Opin. Struct. Biol. 2002, 12, 775–782.
[16] D. B. Harper, Biochem. J. 1977, 165, 309–319.
[17] H. C. Pace, S. C. Hodawadekar, A. Draganescu, J.
Huang, P. Bieganowski, Y. Perkarsky, C. M. Croce, C.
Brenner, Curr. Biol. 2000, 10, 907–917.
[18] D. Kumaran, S. Eswaramoorthy, S. E. Gerchman, H.
Kycia, F. W. Studier, S. Swaminathan, Proteins 2003, 52,
283–291.
[19] a) T. Nakai, T. Hasegawa, E. Yamashita, M. Yamamo-
to, T. Kumasaka, T. Ueki, H. Nanba, Y. Ikenaka, S. Ta-
kahashi, M. Sato, T, Tsukihara, Structure 2000, 8,
729–739; b) W.-C. Wang, W.-H. Hsu, F.-T. Chien, C-Y.
Chen, J. Mol. Biol. 2001, 306, 251–261.
Conversions and yields were determined by reversed phase
HPLC analysis, using a Waters 590 pump, a 50”4.6 mm
Merck Chromolith SpeedROD RP-18e column and
a
Waters 486 UV detector. Mobile phase ACN-H2O-TFA
10:90:0.1 (1a–3a), 20:80:0.1 (1b–e–3b–e) at 1 mLminÀ1.
Enantiomeric purities were determined by chiral HPLC,
using a Waters 510 pump, a 4.6”250 mm Chiracel AD-H
column and a Waters 486 UV detector. Mobile phase
hexane-isopropyl alcohol-TFA=95:5:0.1 (2c, 3c) or 92:8:0.1
(2e, 3e).
Chiral GC of 1c and 1e was performed with a Shimadzu
GC-17 A instrument equipped with a Shimadzu Auto Injec-
tor AOC-20i, a 25 m”0.25 mm Varian CP-Chirasil-Dex CB
column and an FID detector. Column temperature: 1408C
(1c), 1458C (1e).
Enzymatic Hydrolysis
Enzymatic hydrolysis of nitriles (1a–e) was carried out at 1–
2 mL scale in magnetically stirred Eppendorf tubes im-
mersed in a water-bath at the desired temperature. Stock
solutions of the substrate and internal standard (in metha-
nol), PfNLase and buffer (in water) were used. Final con-
centrations were 10 mM of substrate, 1 mM internal stan-
dard, 10% methanol and 100 mM phosphate buffer at the
desired pH.
The reaction was started by the addition of the enzyme
solution (all enzyme weights refer to the total protein con-
tent). Samples were taken periodically, diluted with acid (to
stop the reaction) and filtered over a Microcon YM-10 cen-
trifugal filter device prior to analysis. Samples for chiral
analysis were further extracted with ether and dried over
MgSO4. The solvent was evaporated and residue was redis-
solved in the appropriate solvent.
Reported Ac/Am ratios were measured when these had
stabilised and stayed constant until full conversion.
[20] R. K. Chaturvedi, A. E. MacMahon, G. L. Schmir, J.
Am. Chem. Soc. 1967, 89, 6984–6993.
[21] C. Zervos, E. H. Cordes, J. Org. Chem. 1971, 36, 1661–
1667.
Acknowledgements
[22] M. Winkler, A. Glieder, N. Klempier, Chem. Commun.
2006, 1298–1300.
Thanks are due to Prof. N. Klempier (TU Graz) and Dr. G.
Huisman (Codexis, Inc.) for discussions. This work was sup-
ported by the Spanish CICYT (project PPQ 2002–01231), by
a Ramon y Cajal Contract from M.E. y C. (Spain) and by
COST D25/0002/02. Financial support to B.C.M. Fernandes
by Codexis, Inc. (Redwood City, CA) is gratefully acknowl-
edged.
[23] Dithiothreitol-catalyzed amide formation[22] may have
increased the confusion in the past. The enzyme prepa-
ration that we used did not contain dithiothreitol.
[24] H. C. Pace, C. Brenner, Genome Biology 2001, 2, 1–9;
[online journal, URL: http://genomebiology.com/2001/
2/1/reviews/0001].
2602
www.asc.wiley-vch.de
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2006, 348, 2597 – 2603

Nitrile Hydratase Activity of a Recombinant Nitrilase
FULL PAPERS
[25] M. Kobayashi, M. Goda, S. Shimizu, Biochem. Biophys.
Res. Commun. 1998, 253, 662–666.
[26] C. Novo, S. Farnaud, R. Tata, A. Clemente, P. R.
[28] S. Fçrster, J. Roos, F. Effenberger, H. Wajant, A. Spra-
uer, Angew. Chem. Int. Ed. Engl. 1996, 35, 437–439.
[29] L. Veum, L. T. Kanerva, P. J. Halling, T. Maschmeyer,
U. Hanefeld, Adv. Synth. Catal. 2005, 347, 1015–1021.
Brown, Biochem. J. 2002, 365, 731–738.
[27] L. M. van Langen, R. P. Selassa, F. van Rantwijk, R. A.
Sheldon, Org. Lett. 2005, 7, 327–329.
Adv. Synth. Catal. 2006, 348, 2597 – 2603
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.asc.wiley-vch.de
2603