A high-throughput screening assay for hydroxynitrile lyase activity

Article abstract of DOI:10.1039/b607863j

A high-throughput screening assay for hydroxynitrile lyase activity accepting a wide range of HNL-substrates is presented, which is useful either for enzyme fingerprinting or screening of huge variant libraries generated in metagenome or directed evolution approaches. The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b607863j

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
A high-throughput screening assay for hydroxynitrile lyase activity
Jennifer Andexer, Jan-Karl Guterl, Martina Pohl and Thorsten Eggert*
Received (in Cambridge, UK) 2nd June 2006, Accepted 7th August 2006
First published as an Advance Article on the web 29th August 2006
DOI: 10.1039/b607863j
A high-throughput screening assay for hydroxynitrile lyase
activity accepting a wide range of HNL-substrates is presented,
which is useful either for enzyme fingerprinting or screening of
huge variant libraries generated in metagenome or directed
evolution approaches.
pipetting workstations. The broad applicability of the assay is
demonstrated using different aliphatic and aromatic cyanohydrins
(Fig. 1); two of them were achiral, four of them were used as a
racemic mixture and furthermore benzaldehyde cyanohydrin was
applied as (R)- and (S)-enantiomers to detect the enantioselectivity
of recombinant HNL from Manihot esculenta (MeHNL) which
was used as a model.
Hydroxynitrile lyases (HNLs) naturally occur in plants integrated
in microbial and herbivore defense mechanisms by HCN-release
due to cyanohydrin cleavage. Some HNLs are biochemically
characterized in detail concerning substrate specificity and
enantioselectivity. The HNLs from Prunus species, like
The assay consists of two parts, first the biotransformation step
yielding HCN by cyanohydrin cleavage; subsequently HCN is
detected spectrophotometrically at 600 nm wavelength in MTPs
(
Fig. 1). The crucial parameter in the cyanohydrin cleavage
reaction is the pH, because many cyanohydrins decompose at pH
6.0, whereas on the other hand the enzyme activity and stability
1
a
1b
P. amygdalus (bitter almond) and P. mume (Japanese apricot)
1c
as well as the HNL from Linum usitatissimum (flax) show (R)-
.
selectivity, whereas the other HNLs from Sorghum bicolor
1
d
1e
millet), Hevea brasiliensis (para rubber tree) and Manihot
(
1f
esculenta (cassava) show (S)-selectivity towards different cyano-
hydrins. Since the reverse reaction is also catalyzed by HNLs, these
enzymes are valuable catalysts for the synthesis of cyanohydrins,
which are versatile chiral building blocks in the pharmaceutical
1
and agrochemical industries. Apart from their availability, the
application of HNLs is often restricted by their substrate range
and low stability under technical conditions.
Novel or improved HNLs can be found by screening plants for
2
appropriate enzymes, by directed evolution, rational design, or by
3
metagenomic approaches. However, one major necessity for all
these strategies is a simple and powerful high-throughput screening
(HTS) assay to identify potential novel or better performing
HNLs. So far HNL-activity has usually been determined by GC-
2
a
or HPLC-analysis, which is not practical in high-throughput. In
addition, a spectrophotometric assay based on HCN-detection
4
,5
using the well-known K o¨ nig reaction
However, this assay was applied in a total reaction volume of
0 mL which is not suitable for high-throughput screening
has been described.
1
approaches. Furthermore, only acetone cyanohydrin has been
6
used so far as a substrate. Another spectrophotometric assay for
activity towards benzaldehyde cyanohydrin is available, detecting
the increase in absorption of the released benzaldehyde at 280 nm
7
wavelength. Although this assay is applicable for high-through-
put, it is restricted to aromatic substrates only and requires
microtiter plates (MTPs) which are transparent in the UV.
Furthermore, recently a colony assay, based on NADH-fluores-
cence has been developed to detect HNL-activity towards
Fig. 1 Schematic overview of the assay system. A: Biotransformation
step. The cyanohydrin 1 is enzymatically converted to a carbonyl
compound 2 and HCN. Six different cyanohydrins (3: acetaldehyde
cyanohydrin, 4: propionaldehyde cyanohydrin, 5: benzaldehyde cyanohy-
drin, 6: 3-phenoxybenzaldehyde cyanohydrin, 7: acetone cyanohydrin, 8:
cyclohexanone cyanohydrin) were tested with MeHNL. B: Cyanide-
8
benzaldehyde cyanohydrin.
Here, we describe an HNL assay in MTP format (200 mL),
which allows screening in high-throughput using automated
12
determination step (modified according to Markley et al. ): cyanide
anions are oxidized by N-chlorosuccinimide 9 (stabilized with succinimide
Institute of Molecular Enzyme Technology, Heinrich-Heine University
D u¨ sseldorf, Forschungszentrum J u¨ lich, 52426, J u¨ lich, Germany.
E-mail: t.eggert@fz-juelich.de; Fax: +49 (2461) 612490;
Tel: +49 (2461) 612939
10) to cyanide cations, which react with isonicotinic acid 11 forming a
dialdehyde 12, which is coupled to two molecules of barbituric acid 13 to
form the dye 14 which is measured spectrophotometrically at 600 nm.
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 4201–4203 | 4201

View Article Online
are significantly impaired at pH , 5.0. Therefore, assaying
6,7
cyanohydrin cleavage at pH 5.0 to 5.5 is a good compromise.
The assay is sufficiently sensitive to screen HNL-libraries using
crude cell extracts. For the purpose of library screening, first the
enzymatic reaction using E. coli crude cell extracts containing over-
expressed MeHNL is performed, thereby a certain amount of
cyanide is liberated from the cyanohydrin substrate. Therefore,
140 mL citrate–phosphate buffer pH 5.0, 10 mL of HNL containing
crude cell extracts and 10 mL cyanohydrin solution (final
concentration 15 mM) are mixed and incubated at room
temperature for 5 min. By addition of 10 mL of mix I
(
N-chlorosuccinimide 9/succinimide 10) the biotransformation step
9
2
+
is stopped, thereby oxidizing the liberated CN to CN . After
min the colorimetric detection step is started by adding 30 mL of
2
10
mix II (isonicotinic acid 11/barbituric acid 13). Subsequently, the
rate of color formation is measured spectrophotometrically over
20 min at 600 nm using a microtiter plate reader. The dye 14 is
stable for at least 2 hours. Barbituric acid 13 is applied instead of
the alternatively used dye compound 3-methyl-1-phenyl-5-pyrazo-
11,12
lone in HCN-detection,
because the latter is unsuitable at pH-
13
values below 7. Isonicotinic acid is a well suited alternative to the
6
widely used pyridine.
Instead of measuring the spectrophotometric properties of the
resulting aldehyde or ketone, the major advantage of this assay is
the possibility of analyzing HNL-activity towards virtually any
cyanohydrin by detecting the liberated HCN. This makes the assay
suitable for a vast substrate screening as well as for detection of
new or improved activities in enzyme libraries obtained by rational
design or directed evolution. We have used our HTS-assay to
determine MeHNL-activity towards six different aromatic and
aliphatic cyanohydrins. All substrates were converted by MeHNL
with 5 and 7 being the best substrates. The high selectivity of
MeHNL towards the (S)-enantiomer of benzaldehyde cyanohy-
drin 5 is obvious when (R)-5 and (S)-5 are used separately in the
assay (Fig. 2).
Fig. 3 Spectrophotometric detection of acetone cyanohydrin 7 cleavage
using different amounts of purified MeHNL (–%– 50 ng, –n– 250 ng). A:
The increase in absorbance at 600 nm over 20 min is shown. The amount
of liberated cyanide is calculated from the linear part of the curve.
Autolysis of the respective cyanohydrin is detected in a control without
enzyme (–e–) and subtracted from the slope values of the samples. B:
Hyperbolic cyanide calibration curve and linearization by double
reciprocal presentation.
cyanide standard curve (K [Zn(CN) ]) (Fig. 3A) by correlating the
2
4
Furthermore, the assay allows calculation of specific enzymatic
activity, since color development in the HCN-detection step is
proportional to the amount of cyanide in the solution. Time
dependent cyanohydrin conversion was calculated based on a
rate of color formation at 600 nm with the cyanide concentration.
For this purpose the hyperbolic standard curve was linearized in a
14
double reciprocal diagram (Fig. 3B). For purified MeHNL the
calculated specific activity for acetone cyanohydrin 7 was 130 ¡
3
0 U/mg, which is consistent with data from the literature, giving
15
values between 92 U/mg and 260 U/mg depending on the assay
conditions. Comparison of the specific activities towards different
substrates in Table 1 clearly demonstrates the highest catalytic
activity of MeHNL towards the natural substrate acetone
cyanohydrin. However, it must be taken into account that
substrates 3–6 were applied as racemic mixtures containing 50%
of the non-favored enantiomer, whereas substrates 7 and 8 are
achiral.
Table 1 Results of cyanohydrin cleavage catalyzed by MeHNL.
Substrates 3–8 (15 mM, see Fig. 1) were incubated with purified
MeHNL
14
Fig. 2 Spectrophotometric detection of hydroxynitrile lyase activity.
Microtiter plate with different substrates 3–8 (see Fig. 1). Control:
autolysis of the respective cyanohydrin (without enzyme). For reactions
Substrate
Specific activity [U/mg]
3
4
5
6
7
8
1.3 (¡0.4)
0.4 (¡0.2)
19.1 (¡4.9)
0.1 (¡0.04)
130.0 (¡30.0)
1.0 (¡0.4)
10 mL of E. coli crude cell extracts containing over-expressed MeHNL
were used. Faint blue to purple color represents an increasing amount of
cyanide. The application of enantiomerically pure substrates can be used
to estimate the enantioselectivity of the biocatalyst as demonstrated in the
case of enantiomerically pure (R)- and (S)-benzaldehyde cyanohydrin 5.
4
202 | Chem. Commun., 2006, 4201–4203
This journal is ß The Royal Society of Chemistry 2006

View Article Online
S. F o¨ rster, J. Roos and H. Wajant, ChemBioChem, 2003, 4, 211–216; (g)
H. Griengl, H. Schwab and M. Fechter, Trends Biotechnol., 2000, 18,
In summary, a novel HCN-based high-throughput screening
assay for HNL activity was developed. The assay is useful to detect
activity and enantioselectivity of HNLs theoretically towards any
cyanohydrin substrate. Limitations might occur in the case of
hydrophobic substrates due to poor water solubility. This problem
can be overcome by the use of emulsifying agents like gum arabic.
As tested, the increased turbidity has no influence on the formation
and spectrophotometric detection of the dye (data not shown).
Therefore, the assay is useful for both preparing enzyme
fingerprints and screening large variant libraries generated in
metagenome or directed evolution approaches. The assay is highly
sensitive; at least 5 ng of purified MeHNL representing 1 mU of
enzyme activity was reliably detectable in the assay. Furthermore,
the assay is robust and easy to handle without the necessity of
expensive equipment; however, it is possible to automate the test
by using pipetting robots in order to increase the sample
throughput.
252–256; (h) M. Sharma, N. N. Sharma and T. C. Bhalla, Enzyme
Microb. Technol., 2005, 37, 279–294; (i) F. Effenberger, S. F o¨ rster and
H. Wajant, Curr. Opin. Biotechnol., 2000, 11, 532–539.
(a) Y. Asano, K. Tamura, N. Doi, T. Ueatrongchit, A. H.-Kittikun and
T. Ohmiya, Biosci., Biotechnol., Biochem., 2005, 69, 2349–2357; (b)
A. Hickel, G. Heinrich, H. Schwab and H. Griengl, Biotechnol. Tech.,
1997, 11, 55–58.
3 (a) K.-E. Jaeger, T. Eggert, A. Eipper and M. T. Reetz, Appl. Microbiol.
Biotechnol., 2001, 55, 519–530; (b) P. Berglund and S. Park, Curr. Org.
Chem., 2005, 9, 325–336; (c) P. Lorenz and J. Eck, Nat. Rev. Microbiol.,
2
2005, 3, 510–516.
4 W. K o¨ nig, J. Prakt. Chem., 1904, 69, 105–137.
5 A. J. A. Essers, Acta Hortic., 1994, 375, 97–104.
6
D. Selmar, F. J. Carvalho and E. E. Conn, Anal. Biochem., 1987, 166,
08–211.
M. Bauer, H. Griengl and W. Steiner, Biotechnol. Bioeng., 1999, 62,
2
7
20–29.
8 C. Reisinger, F. van Assema, M. Schuermann, Z. Hussain,
P. Remler and H. Schwab, J. Mol. Catal. B: Enzym., 2006, 39,
149–155.
The work was partly supported by the Bundesministerium f u¨ r
Bildung und Forschung (BMBF) in the project ‘‘Biokatalytische
Hydrocyanierung & Hydroformylierung (BioHydroForm)’’ and
the DFG Graduiertenkolleg 1166 ‘‘BioNoCo’’. The authors thank
Julich Chiral Solutions GmbH for providing the hydroxynitrile
lyase from Manihot esculenta, Clariant GmbH for supplying
cyanohydrins and Prof. Dr U. Kragl (University of Rostock) for
synthesis of enantiomerically pure benzaldehyde cyanohydrin.
9
(a) M. A. Lischwe and M. T. Sung, J. Biol. Chem., 1977, 252,
4976–4980; (b) Incubation of MeHNL with 5mM N-chlorosuccinimide
shows complete inactivation after 10 min.
10 Assay solutions: citrate–phosphate buffer: 24.3 mL 0.1 M citric acid,
25.7 mL 0.2 M K HPO and 100 mL H O; Substrate solution: 300 mM
cyanohydrin in 0.1 M citric acid. Insoluble substrates were emulsified by
2
4
2
adding 20 mg/mL gum arabic; Mix I: 100 mM N-chlorosuccinimide with
1
1
0-fold excess of succinimide (w/w); Mix II: 65 mM isonicotinic acid,
25 mM barbituric acid in 0.2 M NaOH.
1 J. Epstein, Anal. Chem., 1947, 19, 272–274.
2 B. Markley, C. E. Meloan, J. L. Lambert and Y. C. Chiang, Anal. Lett.,
987, 20, 1225–1236.
3 J. H. Bradbury, M. G. Bradbury and S. V. Egan, Acta Hortic., 1994,
1
1
1
Notes and references
1
1
(a) A. Glieder, R. Weis, W. Skranc, P. Poechlauer, I. Dreveny, S. Majer,
M. Wubbolts, H. Schwab and K. Gruber, Angew. Chem., Int. Ed., 2003,
375, 87–96.
14 Enriched MeHNL solution (provided by Julich Chiral Solutions
GmbH) was further purified by anion exchange chromatography on a
Q-sepharose FF column at pH 5.7 using a modified protocol from
S. F o¨ rster, J. Roos, F. Effenberger, H. Wajant and A. Sprauer, Angew.
Chem., Int. Ed. Engl., 1996, 35, 437–438.
4
6
2, 4815–4818; (b) S. Nanda, Y. Kato and Y. Asano, Tetrahedron, 2005,
1, 10908–10916; (c) H. Breithaupt, M. Pohl, W. B o¨ nigk, P. Heim,
K. L. Schimz and M. R. Kula, J. Mol. Catal. B: Enzym., 1999, 6,
15–332; (d) H. Lauble, B. Miehlich, S. F o¨ rster, H. Wajant and
3
F. Effenberger, Biochemistry, 2002, 41, 12043–12050; (e) M. Hasslacher,
C. Kratky, H. Griengl, H. Schwab and S. D. Kohlwein, Proteins:
Struct., Funct., Genet., 1997, 27, 438–449; (f) H. B u¨ hler, F. Effenberger,
15 (a) H. Wajant, S. F o¨ rster, H. B o¨ ttinger, F. Effenberger and
K. Pfizenmaier, Plant Sci., 1995, 108, 1–11; (b) J. Hughes, J. H. Lakey
and M. A. Hughes, Biotechnol. Bioeng., 1997, 53, 332–338.
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 4201–4203 | 4203