A new (R)-hydroxynitrile lyase from Prunus mume: Asymmetric synthesis of cyanohydrins

Article abstract of DOI:10.1016/j.tet.2005.08.105

A new hydroxynitrile lyase (HNL) was isolated from the seed of Japanese apricot (Prunus mume). The enzyme has similar properties with HNL isolated from other Prunus species and is FAD containing enzyme. It accepts a large number of unnatural substrates (benzaldehyde and its variant) for the addition of HCN to produce the corresponding cyanohydrins in excellent optical and chemical yields. A new HPLC based enantioselective assay technique was developed for the enzyme, which promotes the addition of KCN to benzaldehyde in a buffered solution (pH=4.5).

Full text of DOI:10.1016/j.tet.2005.08.105

Tetrahedron 61 (2005) 10908–10916
A new (R)-hydroxynitrile lyase from Prunus mume:
asymmetric synthesis of cyanohydrins
*
Samik Nanda, Yasuo Kato and Yasuhisa Asano
Biotechnology Research Center, Toyama Prefectural University, 5180 Kurokawa, Kosugi, Toyama 939-0398, Japan
Received 8 August 2005; revised 31 August 2005; accepted 31 August 2005
Available online 23 September 2005
Abstract—A new hydroxynitrile lyase (HNL) was isolated from the seed of Japanese apricot (Prunus mume). The enzyme has similar
properties with HNL isolated from other Prunus species and is FAD containing enzyme. It accepts a large number of unnatural substrates
(benzaldehyde and its variant) for the addition of HCN to produce the corresponding cyanohydrins in excellent optical and chemical yields.
A new HPLC based enantioselective assay technique was developed for the enzyme, which promotes the addition of KCN to benzaldehyde in
a buffered solution (pHZ4.5).
q 2005 Elsevier Ltd. All rights reserved.
1. Introduction
has proved not sufficient simply to screen for the presence
of carbonyl compounds or HCN, since the breakdown of
cyanohydrins may be catalyzed by enzymes other than
HNLs: instead, the catalysis of cyanohydrin formation is a
more reliable indicator of the presence of a HNL.6
Hydroxynitrile lyases (HNLs) are widely distributed in
nature and believed mainly to be a plant enzyme. There are
four classes of HNL reported in the literature till now,1
among them mandelonitrile lyase (EC 4.1.2.10) has been
extensively studied in the area of enzymology and organic
synthesis. Hydroxynitrile lyases (HNLs) are one of the key
enzymes in cyanogenic plants,2 catalyzing the final step in
the biodegradation pathway of cyanogenic glycosides
releasing HCN and the corresponding carbonyl components.
HCN released by HNL can serve as a repellent factor to
predators or as susceptibility component of plants to fungal
attack when produced at high local concentrations.3
Mandelonitrile lyase, one of the most well studied HNLs,
are mainly found in Rosaceae species (Genus: Prunus),
catalyzes the formation of benzaldehyde and HCN from
(R)-mandelonitrile.4 Whereas formation of the cyanohydrin
from aldehyde and HCN should also be catalyzed by the
same enzyme. Pioneering work by Effenberger, Griengl and
others opened a new era in the area of asymmetric
cyanohydrin synthesis by HNLs.5 For example, crude
meals from the kernels of almond, apple, cherry, apricots,
and plums serve as source of HNLs. To detect new HNLs, it
Optically active cyanohydrins are excellent building blocks
for the total synthesis of several biologically active natural
products.7 Both the alcohol and the nitrile parts of the
cyanohydrin functionality can undergo transformation to a
range of groups.8 There are general methods that proceed
without racemization, so that the optical purity is retained.
In our previous communication we had reported a few
cyanogenic plant species capable of showing mandelonitrile
lyase activity9 from a rich biodiversity of plant kingdom
available in Japan. Japanese flowering apricot may be the
longest lived of the flowering fruit trees eventually forming
a picturesque 20-ft tall tree. It produces Ume fruit round in
shape and 1–3 inches in diameter. The ripened fruit is
yellowish red in color and have a fleshy covering. In this
article, we would like to report our research on a new HNL
from the seed of Japanese apricot.
2. Results and discussion
2.1. Enzyme assay
Keywords: Hydroxynitrile-lyase; Prunus mume; Cyanohydrins; Asym-
metric synthesis.
Abbreviations: HNL, hydroxynitrile lyase; PmHNL, Prunus mume
hydroxynitrile lyase; PaHNL, Prunus amygdalus hydroxynitrile lyase;
DIPE, diisopropyl ether; TBME, tert-butyl methyl ether; DBE, di-n-butyl
ether; DME, 1,2-dimethoxyethane.
The enzyme was isolated from the seeds of ripened Ume
fruit. Crude preparation of enzyme serves the purpose of our
study, but we have partially purified the enzyme by 30%
(NH4)2SO4 precipitation to obtain a homogeneous solution
*
Corresponding author. Tel.: C81 766 56 7500x530; fax: C81 766 56
2498; e-mail: asano@pu-toyama.ac.jp
0040–4020/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2005.08.105

S. Nanda et al. / Tetrahedron 61 (2005) 10908–10916
Table 1. Monosubstituted benzaldehydes screened with crude PmHNL
10909
Substrate
Conversion (%)a
ee (%)b
Configurationc
Benzaldehyde
2-Chlorobenzaldehyde
3-Chlorobenzaldehyde
4-Chlorobenzaldehyde
4-Bromobenzaldehyde
4-Fluorobenzaldehyde
2-Methylbenzaldehyde
3-Methylbenzaldehyde
4-Methylbenzaldehyde
2-Methoxybenzaldehyde
3-Methoxybenzaldehyde
4-Methoxybenzaldehyde
2-Trifluoromethylbenzaldehyde
3-Trifluoromethylbenzaldehyde
4-Trifluoromethylbenzaldehyde
3-Nitrobenzaldehyde
13
37
38
21
22
28
6.0
7.5
7.0
6.0
31
17
72
91
90
87
89
5.8
nd
nd
42
6.4
4.8
93
56
92
99
99
84
61
87
95
41
92
97
5
68
76
65
71
98
0
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
4-Nitrobenzaldehyde
4-Benzyloxybenzaldehyde
4-N,N-Dimethylaminobenzaldehyde
4-Hydroxybenzaldehyde
3-Phenoxybenzaldehyde
4-Allyloxybenzaldehyde
4-tert-Butyldimethylsilyloxybenzaldehyde
5
R
R
R
R
O99
98
97
a Conversion (%) was calculated after 30 min of reaction time.15
b Determined by chiral HPLC.
c Confirmed by comparing the retention time with that of optically active standard compounds.
of enzyme. Previously reported HNL assay method10
mainly involved the decomposition of mandelonitrile in a
buffered solution and the formation of benzaldehyde was
measured spectrophotometrically. We have developed a
new HPLC based assay technique, in which the formation of
(R)-mandelonitrile from benzaldehyde in citrate buffer
(pHZ4.5) was monitored by chiral HPLC. At different
time intervals aliquots were withdrawn and analyzed by
chiral HPLC. It was assumed that at low pH, corresponding
chemical reaction was suppressed, so the overall mandelo-
nitrile production only comes from HNL catalyzed addition.
However, to obtain accurate results, a blank reaction was
performed side by side without enzyme, and the amount of
mandelonitrile obtained was deducted from the biocata-
lyzed reaction product. The HPLC based HNL assay
technique has several advantages over the other traditional
assay technique. Sensitivity and reproducibility are very
high in the HPLC based assay method. Using different
dilution of enzyme and varying substrate concentration
yields almost similar results in all case. Moreover,
cyanohydrins itself are well separated in a chiral column,
hence avoiding further derivatization method. Finally,
configurational assignment of the product cyanohydrins is
also possible with comparing standard compounds (synthe-
sized by chemical methods).
2.2. Reaction condition for substrate screening
Next, our focus is to test a large number of different
aldehydes and ketones for possible substrate of PmHNL
(Prunus mume hydroxynitrile lyase). Generally similar
reaction condition as the PmHNL assay was employed for
the substrate screening methods. In a typical screening
methods substrates were taken in aq buffer (pH 4.5)
followed by addition of KCN (1.25 equiv) and enzyme
solution (see Section 4 for detailed description). After
certain time (see Tables 1–6), reaction mixtures were taken
and the formation of products were analyzed with chiral
HPLC/GC. The main intention of the study was to test a
large number of substrates and find the substrates, which
provide better enantioselection. We have attempted an
enzymatic transcyanation with acetone cyanohydrin as the
cyanide source, but the slow reaction (decomposition of
acetone cyanohydrin is slow at lower pH e.g., 4–4.5) rate
Table 2. Heteroaromatic and polycyclic aldehydes screened with crude PmHNL
Substrate
Conversion (%)a
ee (%)b
Configurationc
2-Furancarboxaldehyde
2-Thiophenecarboxaldehyde
2-Pyridinecarboxaldehyde
3-Pyridinecarboxaldehyde
4-Pyridinecarboxaldehyde
2-Quinolinecarboxaldehyde
4-Quinolinecarboxaldehyde
1-Napthalenecarboxaldehyde
2-Napthalenecarboxaldehyde
9-Anthral
1.2
31
89
90
65
38
73
60
58
16
98
88
22
75
41
21
28
93
98
0
S
S
S
R
R
S
R
R
R
a Conversion (%) was calculated after 30 min of reaction time.15
b Determined by chiral HPLC.
c Confirmed by comparing the retention time with that of optically active standard compounds.

10910
S. Nanda et al. / Tetrahedron 61 (2005) 10908–10916
Table 3. Disubstituted benzaldehydes screened with crude PmHNL
Substrate
Conversion (%)a
ee (%)b
Configurationc
2,3-Dichlorobenzaldehyde
2,4-Dichlorobenzaldehyde
2,5-Dichlorobenzaldehyde
2,6-Dichlorobenzaldehyde
3,4-Dichlorobenzaldehyde
3,5-Dichlorobenzaldehyde
2,3-Dimethoxybenzaldehyde
2,4-Dimethoxybenzaldehyde
2,5-Dimethoxybenzaldehyde
2,6-Dimethoxybenzaldehyde
3,4-Dimethoxybenzaldehyde
3,5-Dimethoxybenzaldehyde
Piperonal
11
13
8.8
10
7.9
21
7.0
11
9.0
6.5
13
17
34
22
78
57
12
94
92
37
48
63
32
78
97
98
86
R
R
R
R
R
R
R
R
R
R
R
R
R
R
2,4-Dimethylbenzaldehyde
5.8
a Conversion (%) was calculated after 30 min of reaction time.15
b Determined by chiral HPLC.
c Confirmed by comparing the retention time with that of optically active standard compounds.
Table 4. Polysubstituted benzaldehydes screened with crude PmHNL
Substrate
Conversion (%)a
ee (%)b
Configurationc
2,3,4-Trimethoxybenzaldehyde
2,4,5-Trimethoxybenzaldehyde
3,4,5-Trimethoxybenzaldehyde
2,3,4,5-Tetrafluorobenzaldehyde
2,3,5,6-Tetrafluorobenzaldehyde
2,3,4,5,6-Pentafluorobenzaldehyde
2,3,4,5,6-Pentabromobenzaldehyde
14
16
24
26
21
31
18
11
28
31
23
12
0
R
R
R
R
R
0
a Conversion (%) was calculated after 1 h of reaction time.15
b Determined by chiral HPLC.
c Confirmed by comparing the retention time with that of optically active standard compounds.
Table 5. Aliphatic methylketones screened with crude PmHNL
Substrate
Conversion (%)a
ee (%)b
Configurationc
2-Butanoned
2-Pentanone
48
46
39
48
40
28
39
30
22
20
18
21
14
58
62
72
81
42
80
88
38
74
76
67
65
52
31
0
R
R
R
R
R
R
R
R
R
R
R
R
3-Methyl-2-butanone
2-Hexanoned
4-Methyl-2-pentanoned
3,3-Dimethyl-2-butanone
2-Heptanone
5-Methyl-2-hexanone
2-Octanone
2-Nonanone
2-Decanone
2-Undecanone
2-Dodecanone
Cyclopropylmethylketone
Trimethylsilylmethylketone
0
72
R
a Conversion (%) was calculated after 3 h by analyzing 1H NMR of the crude reaction mixture.
b Determined by chiral GC (b-Dex120).
c Confirmed by comparing the retention time with that of optically active standard compounds.
d These cyanohydrins are not resolved in chiral GC column, hence ee was determined by preparing the corresponding TMS–ether derivative.
Table 6. Aliphatic aldehydes screened with crude PmHNL
Substrate
Conversion (%)a
ee (%)b
Configurationc
Propanal
Butanald
48
51
43
29
36
38
51
54
78
84
88
92
85
81
91
94
R
R
R
R
R
R
R
R
Isobutyraldehyde
Pivaladehyde
Pentanald
Hexanald
Cyclopentanecarboxaldehyde
Cyclohexanecarboxaldehyde
a Conversion (%) was calculated after 3 h by analyzing 1H NMR of the crude reaction mixture.
b Determined by chiral GC (b-Dex325).
c Confirmed by comparing the retention time with that of optically active standard compounds.
d These cyanohydrins are not resolved in chiral GC column, hence ee was determined by preparing the corresponding TMS–ether derivative.

S. Nanda et al. / Tetrahedron 61 (2005) 10908–10916
10911
makes the process not applicable for screening purpose,
whereas the use of HCN was avoided due to difficulties in
handling. The pH profile of the reaction was investigated
and it was found that the enzyme is active mainly at lower
pH. Increasing the pH (pH of the final reaction mixture)
higher than 5.5 led to yielding products of low optical purity
due to the corresponding chemical reaction. At pH 6.5
activities was totally lost, and the product was completely
racemic (Fig. 1).
(DBE). With all the three solvents more than 95%
enantioselectivity was observed. However, the rate of the
reaction was very slow and it takes 24–36 h to obtain
appreciable amount of conversion. HCN in organic solvent
and acetonecyanohydrin was used as the cyanating agent
(Fig. 3).
Figure 3. Solvent effect on ee of PmHNL catalyzed (R)-mandelonitrile
synthesis.
Figure 1. pH effect on ee of PmHNL catalyzed (R)-mandelonitrile
formation.
2.3. Substrate specificity
A large number of substrates including aliphatic, aromatic,
and heteroaromatic aldehydes as well as aliphatic ketones
were tested for PmHNL activity.
The enzyme works nicely in a wide temperature range
varying from K10 to 40 8C. The best temperature for the
enzymic reaction was found to be 15–25 8C, using
benzaldehyde as a substrate. Above 40 8C, the enzymic
activity decreased significantly and it produces racemic
cyanohydrin (Fig. 2).
2.3.1. Aromatic aldehydes. Aromatic aldehydes (mono, di,
and polysubstituted), heteroaromatic aldehydes as well as
polycyclic aromatic aldehydes were employed as PmHNL
substrates. The results were summarized in Tables 1–5.
It was observed that substrates having ortho substituent
(irrespective of its electronic nature) are poor substrates in
terms of enantioselectivity of the resulting cyanohydrins.
We have taken a series of substituted chlorobenzaldehydes
and compared their enantioselectivity with PmHNL (Fig. 4).
Figure 2. Temperature dependence of ee for (R)-mandelonitrile synthesis
by PmHNL three sets of data were measured after 10, 20, and 30 min
reaction time. Temperature of the reaction was K10 8C (1), 0 8C (2), 5 8C
(3), 10–12 8C (4), 18–20 8C (5), 25 8C, (6), 40 8C (7), 50 8C (8), 70 8C (9).
Next, we have screened different organic solvents for the
PmHNL activity. It has been reported by several researchers
that, defatted powdered HNL meal (almond, apple, cherry)
are active in polar but aprotic organic solvents and produces
cyanohydrin with high enantioselectivity.11 Whereas immo-
bilization of HNLs on some support also allowed its use in
organic solvent.12 We have tested PmHNL’s activity in
several organic solvents by taking the partially purified
homogeneous enzyme solution. A biphasic reaction con-
dition (organic solvent/buffer, 10:1) allows us to detect the
product by HPLC. The best solvent was diisopropylether
(DIPE), tert-butylmethylether (TBME) and di-n-butylether
Figure 4. ee comparison of chlorosubstituted benzaldehydes (showing an
ortho effect).
2.3.2. Aliphatic ketones. Mainly methyl ketones were
tested for PmHNL activity. In a typical reaction ketones
were taken in citrate buffer (pH 4.0, reaction pH 4.5),

10912
S. Nanda et al. / Tetrahedron 61 (2005) 10908–10916
followed by addition of enzyme and aq KCN solution.
Generally it takes 3 h for appreciable amount of conversion
(detected by TLC). Longer reaction time allows complete
conversion but also leads to product having poor optical
purity due to the corresponding chemical cyanation. The
chemical yield of the obtained cyanohydrins was deter-
Reaction was monitored by TLC and chemical yield was
determined with the help of 1H NMR. It seems that aliphatic
aldehydes are better substrates than their methyl ketone
counterparts. Their reactivities are high probably due to the
high reactivity of aldehyde functionality than those of the
ketones. It is important to mention that all the aliphatic
aldehydes tested so far, provides excellent enantioselection.
Thus, it is quite evident that aliphatic aldehydes are better
substrates for PmHNL than the corresponding methyl
ketones. From the previous discussion we have also noticed,
with the branching at a-position in the methyl ketone leads
with poor enantioselection (t-butylmethyl ketone and
isopropylmethyl ketone), whereas the corresponding alde-
hyde partners, for example, pivaladehyde and isobutyralde-
hyde both are excellent substrates for PmHNL (Table 6).
Thus, presence of methyl moiety in the methyl ketones must
play a decisive role for poor enantioselectivity. The optical
purity of the product cyanohydrins was determined directly
by chiral GC cyclodextrin columns and in some cases
cyanohydrins were derivatized to the corresponding
trimethylsilylether (OTMS).
1
mined from H NMR of crude samples. By comparing the
integral values of Me-signals (singlet, at d 2.2 ppm in the
ketone and d 1.5–1.6 ppm in the cyanohydrin) the relative
ratios hence the chemical yield was determined.
The optical purity of the ketone cyanohydrins were
measured efficiently by chiral GC (b-cyclodextrin stationary
phase). Only three of the synthesized cyanhohydrins need to
be derivatized (ethylmethylketone, n-butylmethylketone
and isobutylmethylketone, as their TMS ether) in order to
determine the ee. Remaining all cyanohydrins is well
separated by b-cyclodextrin stationary phase, hence
eliminating the derivatization protocol as often used by
many researchers. The optical purity of the synthesized
cyanohydrins is lower when compared with similar HNL
from almonds. Methyl ketones having no branching at
a-position are good substrates, for example, ethyl, n-propyl,
n-butyl, and n-amyl methyl ketones provide cyanohydrin
with 72–88% ee. Whereas substrates having branching at
a-position give poor ee, for example, i-propyl and t-butyl
methylketone yields cyanohydrin with 42 and 38% ee only.
Increasing the chain length has a mild but definite effect on
optical purity of the synthesized cyanohydrins. Undecanone
(having C9 unit at one end) is the limiting substrate in the
series accepted by PmHNL, although ee is low (31%,
Table 5), where as dodecanone yields racemic cyanohydrin.
2.3.4. Preparative scale asymmetric synthesis of cyano-
hydrins by PmHNL. Owing to their broad synthetic
potential, cyanohydrins have attracted attention as starting
materials for the preparation of several important classes of
compounds, and it is probable that the use of enantio-
merically pure cyanohydrins as building blocks for the
production of chiral industrial chemicals will continue to
grow. The pioneering work of Becker and Pfeil demon-
strated for the first time that the HNL isolated from almond
can be employed successfully for the production of
(R)-cyanohydrins from aldehyde and HCN on a kilogram
scale. One should consider four major points in a large scale
HNL catalyzed synthesis of cyanohydrins, for example,
reaction media, cyanating source, physical state of enzyme
2.3.3. Aliphatic aldehydes. Aliphatic aldehydes were
screened for PmHNL activity by similar methods as
described earlier for corresponding methyl ketones.
Table 7a. Preparative scale cyanohydrin a synthesis by PmHNL with selected aromatic aldehydes
Substratea
Yield (%)
Time (h)b
ee (%)c
Benzaldehyde
para-Tolualdehyde
para-Anisaldehyde
4-Fluorobenzaldehyde
4-Chlorobenzaldehyde
4-Bromobenzaldehyde
4-Nitrobenzaldehyde
65
68
78
60
78
82
93
96
88
65
82
80
78
82
88
68
52
58
65
48
60
70
78
24
42
20
16
16
26
6
95
96
98
86
99
99
71
64
72
96
88
65
14
21
97
98
94
95
92
93
98
90
96
3-Nitrobenzaldehyde
9
8
4-Trifluoromethylbenzaldehyde
2-Furan carboxaldehyde
2-Thiophene carboxaldehyde
3-Pyridine carboxaldehyde
2-Quinlonine carboxaldehyde
4-Quinlonine carboxaldehyde
Piperonal
3-Phenoxybenzaldehyde
3,4-Dichlorobenzaldehyde
3,5-Dichlorobenzaldehyde
3,4-Dimethoxybenzaldehyde
3,5-Dimethoxybenzaldehyde
4-Benzyloxybenzaldehyde
Napthalene-1-carboxaldehyde
Napthalene-2-carboxaldehyde
78
24
12
14
15
30
36
52
60
48
50
72
32
40
a All reactions were performed with 5 g of substrate, 1.5 equiv of acetonecyanohydrin in biphasic reaction media (total reaction volume is 50 mL, DIPE/aq
buffer, 10:1).
b The reaction was monitored by TLC.
c ee’s were determined by chiral HPLC analysis at 254 nm (by applying same condition as described earlier).

S. Nanda et al. / Tetrahedron 61 (2005) 10908–10916
10913
and external conditions (temperature and pH). We generally
apply a vigorously stirred biphasic (citrate buffer, pH 4.0,
and organic solvent) reaction media using transcyanation
reaction with acetonecyanohydrin and PmHNL as the
biocatalyst. We generally employed DIPE, TBME, and
DBE as the organic media for transcyanation reaction. By
applying the above reaction condition we are able to
synthesize a large number of cyanohydrins with excellent
enantioselection from several aromatic aldehydes
(Table 7a). It was observed that substrates, which showed
poor enantioselectivity in aq media, (e.g., 3-NO2, 4-NO2,
3-pyr, 4-CF3, 2 and 4 quinolinecarboxaldehyde) do not
markedly change their activity with change in the reaction
media (Table 7a). Moderate to low enantioselectivity was
observed in the synthesized cyanohydrins from those highly
reactive aldehydes. As all those aldehydes are highly
activated due to their structural feature, chemical yield of
the corresponding cyanohydrins is very high. The possible
reason for the low enantioselectivity can be attributed due to
the corresponding chemical cyanation, which occurs side by
side with the enzymatic cyanation. Though it was thought
that by changing the reaction media from aq to organic
solvent, the corresponding chemical reaction may be
suppressed.
allows us to overcome all the associated problems as
discussed earlier.
2.4. Structure–activity relationships
Though hydroxynitrile lyase research was widely explored
by the researchers, no suitable model except one rough
model by Ognyanov et al.13 was known from which one can
predict structure–activity relationship of a given sets of
substrate. The model reported by Ognyanov and co-workers
is based on computer assisted modeling, and one can have a
rough idea of predicting the steric limits of modified
substrate for PaHNL (Prunus amygdalus Batsch. Syn:
Prunus dulcis) catalyzed cyanohydrin formation. Though
one can explain the minimum steric requirements needed to
be a perfect HNL substrate by using this model, but one
cannot explain the minimum electronic parameters needed
to be an ideal HNL substrate. The following structure–
activity relationship can be predicted from the above
discussion as well as previous results reported by other
researchers, although the prediction is purely hypothetical,
as no theoretical calculations were performed. The
prediction is entirely based on the results obtained by us
and several previous researchers. When we compare our
enzyme PmHNL with the most widely known (R)-HNL,
PaHNL (almond); we found that PmHNL is better enzyme
than PaHNL with regard to the aromatic substrates. With
aliphatic aldehydes PmHNL gives similar results as PaHNL.
However, PaHNL is superior when aliphaticmethyl ketones
were used as substrate. The following points are predicted
based on our research on PmHNL and previous results with
PaHNL.
By employing the above reaction condition, aliphatic
methylketones and aliphatic aldehydes were subjected to
PmHNL catalyzed cyanohydrin synthesis. The detailed
results are shown in Table 7b. In case of aliphatic carbonyl
compounds, those compounds were tested only which
provides good enantioselection in the earlier screening
method (Tables 5 and 6).
(a) presence of substituents in ortho position (irrespective
of its electronic nature) always lead to sluggish
reaction (poor ee) compared to meta and para
substituents.
(b) electron donating groups such as KMe, reduced the
reactivity of the aldehyde, thereby affording lower
yields compared to that of the parent compound
benzaldehyde. Whereas presence of group having KI
(inductive) effect and CR (resonance) effect, for
example, halogen, OMe are more reactive than the
parent benzaldehyde.
It is clear from the above discussion that by applying a
vigorously stirred biphasic reaction media (aq buffer/polar
aprotic solvent like DIPE), the yield and optical purity of the
synthesized cyanohydrins can be enhanced markedly. This
method is also applicable for obtaining a large quantity of
chiral cyanohydrins for industrial purposes. Further focus
will be on the reuse of this new enzyme PmHNL by
immobilization methods as reported by other researchers
with different HNLs. And currently we are working on those
aspects to find out an effective PmHNL system, which
Table 7b. Preparative scale cyanohydrin synthesis by PmHNL with selected aliphatic aldehydes and methyl ketones
Substratea
Yield (%)
Time (h)b
ee (%)c
Propanal
Butanal
Isobutanal
Pivalaldehyde
Pentanal
Hexanal
Cyclopentanecarboxaldehyde
Cyclohexanecarboxaldehyde
2-Pentanone
68
58
62
52
57
60
70
72
60
65
56
49
52
11
14
10
6
16
22
14
12
42
58
48
60
18
94
90
94
96
88
90
94
93
72
83
88
65
78
2-Hexanone
4-Methyl-2-pentanone
5-Methyl-2-hexanone
Trimethylsilylmethylketone
a All reactions were performed with 5 g of substrate, 1.5 equiv of acetonecyanohydrin in biphasic reaction media (total reaction volume is 50 mL, DIPE/aq
buffer, 10:1).
b The reaction was monitored by TLC.
c ee’s were determined by chiral GC analysis (by applying same condition as described earlier).

10914
S. Nanda et al. / Tetrahedron 61 (2005) 10908–10916
(c) presence of strong electron withdrawing substituents
such as KNO2, CF3 activated the aldehyde so much,
that almost in all cases complete conversion was
observed within a short time. High reactivity of these
aldehydes also caused the low ee due to non-enzymatic
cyanation reaction. It was reported by Han et al.5e that
para-trifluoromethylbenzaldehyde when treated with
PaHNL, produced cyanohydrin with no enantioselec-
tivity; but the same substrate provides 75% ee when
reacted with PmHNL. Where as meta-trifluoromethyl
benzaldehyde provides 70–80% ee with both the
enzyme. The reason for para-selectivity of PmHNL
for this particular substrate is unclear.
(d) 4-hydroxybenzaldehyde acted as poor substrate for
PmHNL. But when the hydroxyl group was protected
as ether functionality (–OBn, –Oallyl, –OTBS), the
respective aldehydes are excellent substrates for
PmHNL in terms of enantioselection.
(e) N-containing heteroaryl carboxaldehydes turned out to
be poor substrates from an enantioselective point of
view, as observed in the case of pyridine and quinoline
series aldehydes. Whereas O and S-containing hetero-
cyclic carboxaldehydes are excellent substrates for
PmHNL as well as PaHNL.
(f) in the polycyclic aromatic aldehyde series, poor
reactivity of 9-anthral suggested that maximum two
rings can be accommodated in the PmHNL active site
(both of the substrates 1 and 2-napthal provides good
selectivity with PmHNL compared to PaHNL as
reported by Riva et al.5f).
(g) increasing the number of substituents in the aromatic
ring often lead to poor enantioselection. Tri, tetra, and
penta substituted aromatic aldehydes always provides
poor or no enantioselectivity.
(h) both PmHNL and PaHNL accept a broad range of
aliphatic aldehydes (acyclic and cyclic) as their
substrates, and overall good selectivity was observed.
It was observed that, aliphatic cyclic aldehydes, for
example, cyclopentane and cyclohecxane carboxalde-
hyde are excellent substrate compared to their linear
counterpart, as both of the aldehydes react fast and
provides the corresponding cyanohydrin in good
chemical and optical yield.
(i) aliphatic methylketones are also accepted as substrate,
but compared to their aldehyde counterpart less
enantioselectivity was observed. It is also important
to mention that PaHNL is little superior than PmHNL,
when aliphatic methyl ketones are used as substrates
(in terms of ee).
progress. We are also investigating the enzymological
properties of PmHNL in detail.
4. Experimental
4.1. General
Unless otherwise stated, materials were obtained from
commercial suppliers and used without further purification.
Ripened Ume fruit (P. mume) was obtained from local fruit
market and stored at 4 8C. All aldehydes used in the
experiment are freshly distilled or washed with aq NaHCO3
solution to minimize the amount of free acid, which are
supposed to inhibit HNL activity. Mandelonitrile and
acetone cyanohydrins were freshly distilled prior to use.
Reactions were monitored by TLC, carried out on 0.25 mm
silica gel plates (Merck) with UV light, ethanolic vanillin,
and phosphomolybdic acid/heat as developing agents. Silica
gel 100–200 mesh was used for column chromatography.
Yields refer to chromatographically and spectroscopically
homogeneous materials unless otherwise stated. NMR
spectra were recorded on 400 MHz spectrometer at 25 8C
in CDCl3 using TMS as the internal standard. Chemical
shifts are shown in d. 13C NMR spectra were recorded with a
complete proton decoupling environment. The chemical
shift value is listed as dH and dC for 1H and 13C, respectively.
Chiral HPLC was performed using chiral OJ-H column
(0.46!25 cm, Daicel industries) with water 717 auto-
sampler and UV–vis detector (254 nm). Eluting solvent
used was different ratio of hexane and 2-propanol. Chiral
GC analysis was performed in a Schimadzu autosampler
with cyclodextrins columns as chiral stationary phase
(Fused silica capillary column, 30 m!0.25 mm!0.25 mm
thickness, b-Dex-120 and b-Dex-325 from SUPELCO,
USA) using He as a carrier gas (Detector temperature:
230 8C and injection temperature 220 8C).
4.2. Enzyme extraction from P. mume
Ripened Ume fruit was taken and the fleshy cover was
removed to obtain the seeds. The upper layer of the seeds
was cracked with hammer to give the soft kernels inside.
Those kernels were collected and homogenized in a
homogenizer at 4 8C, with aq potassium phosphate buffer
(10 mM, pHZ6.0), to give a milky suspension. The
suspension was filtered through four layers of cheese cloth
to remove the insoluble part. After that it was centrifuged
(18,800 g, 30 min), removal of the residue gives a crude
preparation of PmHNL. The crude preparation was
fractionated with (NH4)2SO4. Proteins precipitating with
30% saturation were collected by centrifugation (18,800 g,
20 min), dissolved in minimum volume of phosphate buffer
and dialyzed against the same buffer with three changes.
After that the dialyzed solution was centrifuged and the
supernatant was stored at 4 8C and assayed for HNL activity.
3. Conclusion
In conclusion, we have found a new (R) hydroxynitrile lyase
from Japanese apricot (P. mume). The new enzyme accepts
a broad array of substrates ranging from aromatic,
heteroaromatic, bicyclic as well as aliphatic carbonyl
compounds and yields the corresponding cyanohydrins
with excellent enantioselection. Studies directed towards
finding new substrates for PmHNL and predicting an
effective model, which can explain the minimum require-
ments needed to be an efficient HNL substrate are in
4.3. PmHNL assay
In a typical assay reaction 1.0 M of benzaldehyde solution
(in DMSO, 40 mL) was dissolved in 400 mM citrate buffer
(760 mL, pHZ4.0), followed by addition of 100 mL of
enzyme solution and 100 mL of 1.0 M KCN solution (total

S. Nanda et al. / Tetrahedron 61 (2005) 10908–10916
10915
reaction volume 1 mL). After 5 min, the 100 mL of the
reaction mixture was taken out and extracted with 900 mL
hexane–2-propanol (9/1), the organic layer was analyzed
with chiral HPLC for the formation of (R)-mandelonitrile.
A blank reaction was also performed without enzyme, and
the amount of mandelonitrile obtained was deducted from
the biocatalyzed reaction product. One unit of the enzyme is
defined as the amount of the enzyme that produces 1 mmol
of (R)-mandelonitrile under reaction condition in 1 min.
pHZ4.0. The mixture was stirred for 20 min and then
extracted with an appropriate organic solvent. The HCN
solution was stored in a dark bottle at 0 8C, with little
addition of citrate buffer (pHZ4.0). The HCN concen-
tration of the obtained solution was determined as reported
by Sheldon et al.12a
4.6. Preparative scale biphasic synthesis of cyanohydrins
using PmHNL
The protein content in PmHNL was measured by Bradford
method using Bio-Rad protein assay kit using BSA as the
standard.14 The protein content in a crude extract of
PmHNL was found to roughly 10 mg/mL and activity was
120 U/mL as determined by the above method. The enzyme
is a flavoprotein (FAD containing), as evident by the yellow
color. Absorption spectroscopy also reveals the fact.
Carbonyl compounds (5 g) were taken in appropriate
solvent (DIPE, TBME or DBE, 50 mL) saturated with
5 mL of citrate buffer (pHZ4.0). Solution of PmHNL
(50 U/mmol of substrate) followed by acetonecyanohydrin
(1.5 equiv) was added to the reaction mixture. The reaction
was vigorously stirred until the desired conversion was
achieved (by TLC). Usual extractive work-up afforded the
cyanohydrins.
4.4. Substrate screening condition using PmHNL
In an Eppendrof tube, 1.0 M solution (40 mL, in DMSO) of
the respective carbonyl compound was taken. Citrate buffer
(400 mM, pHZ4.0), 760 mL was added to it, followed by
addition of 100 mL PmHNL and 1.25 M, solution of KCN
(dissolved in citrate buffer, pHZ4.0). The reaction was
monitored by taking a small aliquot of reaction mixture
(every 5 min interval, 25 mL). The sample was extracted
with ethyl acetate and the organic layer was analyzed with
HPLC. The course of the reaction was followed until
appreciable amount of conversion was achieved (30 min to
1 h, depending on structural features of the aromatic
aldehydes). For the case of aliphatic aldehydes and ketones,
the above procedure was little modified, as substantial more
amount of sample is needed to analyze the product
formation by GC. Aliphatic carbonyl compounds (100 mg)
were dissolved in 5 mL of citrate buffer, followed by
addition of PmHNL solution (200 U/mmol of substrate) and
1.0 M of KCN solution (2 equiv). For those cyanohydrins,
which are well resolved in chiral GC column (comparing
from standard synthesized compounds), the course of the
reaction was followed by extraction and analyzing the
product formation by GC. In other cases, the reaction was
followed by TLC, and after appreciable amount of
conversation was achieved, the cyanohydrins are deriva-
tized as the corresponding silyl ether (OTMS) and analyzed
by chiral GC. It was assumed that in all the cases
corresponding chemical cyanation caused by KCN is
minimum due to low pH of the reaction medium. Using
KCN as a cyanating source, within 1–3 h (depending on the
structural features of the carbonyl compounds), appreciable
amount of conversion was achieved, hence large number of
substrates (approx. 20) can be analyzed in a single day
operation.
Acknowledgements
JSPS (Japan Society for Promotion of Science) post doctoral
fellowship to S. N. is gratefully acknowledged, this research
was partly supported by a Grant-in-Aid for Scientific
Research from the Ministry of Education, Culture, Sports,
Science, and Technology of Japan.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.tet.2005.08.
105. Spectral (1H and 13C NMR) data for selected
compounds and GC/HPLC retention time are available.
References and notes
1. (a) Smitskamp-Wilms, E.; Brussee, J.; Van der Gen, A.; van
Scharrenburg, G. J. M.; Sloothaak, J. B. Recl. Trav. Chim. Pay-
Bas. 1991, 110, 209–215. (b) Gerstner, E.; Kiel, U. Z. Physiol.
Chem. 1975, 356, 1853–1857. (c) Xu, L.-L.; Singh, B. K.;
Conn, E. E. Arch. Biochem. Biophys. 1986, 250, 322–328.
(d) Yemm, R. S.; Poulton, J. E. Arch. Biochim. Biophys. 1986,
247, 440–445. (e) Albrecht, J.; Jansen, I.; Kula, M.-R.
Biotechnol. Appl. Biochem. 1993, 17, 191–203. (f) Wajant,
¨
H.; Forster, S.; Selmar, D.; Effenberger, F.; Pfizenmaier, K.
Plant Physiol. 1995, 109, 1231–1238. (g) Wajant, H.; Mundry,
K.-W. Plant Sci. 1993, 89, 127–133. (h) Bove, C.; Conn, E. E.
J. Biol. Chem. 1961, 236, 207–210. (i) Kuroki, G. W.; Conn,
E. E. Proc. Natl. Acad. Sci. U.S.A. 1989, 86, 6978–6981.
4.5. Generation of HCN solution in organic solvent
¨
(j) Wajant, H.; Forster, S.; Bottinger, H.; Effenberger, F.;
All reaction equipment in which cyanides are used or
produced was placed in a well ventilated hood. Proper
gloves were worn in when handling cyanides, splash proof
goggles and proper mask were also used when dealing with
HCN. The solution of HCN in organic solvent (DIPE,
TBME, DBE) can be generated as follows. In an aq solution
of NaCN/KCN, concd HCl (35%) was dropped at 0 8C, until
Pfizenmaier, K. Plant Sci. 1995, 108, 1–11. (k) Hughes, J.;
Lakey, J. H.; Hughes, M. A. Arch. Biochim. Biophys. 1994,
311, 496–502.
2. (a) Hickel, A.; Hasslacher, M.; Griengl, H. Physiol. Plant
1996, 98, 891–898. (b) Wajant, H.; Effenberger, F. Biol. Chem.
1996, 377, 611–617. (c) Seigler, D. S. Cyanide and

10916
S. Nanda et al. / Tetrahedron 61 (2005) 10908–10916
Cyanogenic Glycosides. In The Chemical Participants, 2nd
6. Hickel, A.; Heinrich, G.; Schwab, H.; Griengl, H. Biotechnol.
Tech. 1997, 11, 55–58.
ed.; Rosenthal, G. A., Berenbaum, M. R., Eds.; Herbivores:
Their Interactions with Secondary Metabolites; Academic:
New York; Vol. 1, pp 35–77. (d) Conn, E. E. Ann. Rev. Plant.
Physiol. 1980, 31, 433–451.
7. Kruse, C. G. Chiral Cyanohydrins — Their Manufacture and
Utility as Chiral Building Blocks. In The Commercial
Manufacture and Applications of Optically Active Com-
pounds; Collins, A. N., Sheldrake, G. N., Crosby, J., Eds.;
Chirality in industry; Wiley: Chichester, UK, 1992; Chapter
14, pp 279–299.
3. (a) Nahrstedt, A. Plant Syst. Evol. 1985, 150, 35–47.
(b) Lieberei, R.; Selmar, D.; Biehl, B. Plant Syst. Evol. 1985,
150, 49–63. (c) Okolie, P. N.; Obasi, B. N. Phytochemistry
1993, 33, 775–778. (d) Kakes, P.; Hakvoort, H. Phyto-
chemistry 1992, 31, 1501–1505. (e) Wajant, H.; Riedel, D.;
Benz, S.; Mundry, K.-W. Plant Sci. 1994, 103, 145–154.
(f) Poulton, J. E. Enzymology of Cyanogenesis in
Rosaceous Stone Fruits. In b-Glucosidases: Biochemistry
and Molecular Biology; Esen, A., Ed.; ACS Symposium
Series 533; Oxford University Press: New York, 1993; pp
170–190. (g) Wu, H. C.; Poulton, J. E. Plant Physiol.
1991, 96, 1329–1337.
8. (a) Schmidt, M.; Herve, S.; Klempier, N.; Griengl, H.
Tetrahedron 1996, 52, 7833–7840. (b) Brussee, J.; Loos,
W. T.; Kruse, C. G.; van der Gen, A. Tetrahedron. 1990, 46,
979–986. (c) Menendez, E.; Brieva, R.; Rebolledo, F.; Gotor,
V. J. Chem. Soc., Chem. Commun. 1995, 989–990.(d) A list of
reactions on both the hydroxyl (protected and unprotected) and
nitrile functionality and corresponding literature references
can be found at Ref. 4b
9. (a) Asano, Y.; Tamura, K.; Doi, N.; Ohmiya, T., Biosci.
Biotech. Biochem. 2005, in press. (b) Tamura, K.; Asano, Y.
Studies on mandelonitrile lyases from plants. Masters thesis,
Toyama Prefectural University, 2003.
4. (a) Garcia-Urdiales, E.; Alfonso, I.; Gotor, V. Chem. Rev.
2005, 105, 313–354. (b) Gregory, R. J. H. Chem. Rev. 1999,
99, 3649–3682. (c) North, M. Tetrahedron: Asymmetry 2003,
14, 147–176. (d) Griengl, H.; Schwab, H.; Fechter, M. Trends
Biotechnol. 2000, 18, 252–256. (e) Effenberger, F. Chimia
1999, 53, 3–10. (f) Johnson, D. V.; Griengl, H. Adv. Biochem.
Eng./Biotechnol. 1999, 63, 32–55.
10. Hanefeld, U.; Straathof, A. J. J.; Heijnen, J. J. Biochim.
Biophys. Acta 1999, 1432, 185–193.
11. (a) Zandergen, P.; Van der Linden, J.; Brussee, J.; Van der
Gen, A. Synth. Commun. 1991, 21, 1387–1391. (b) Lin, G.;
Han, S.; Li, Z. Tetrahedron 1999, 55, 3531–3540. (c) Kiljunen,
E.; Kanerva, L. T. Tetrahedron: Asymmetry 1997, 8,
1225–1234.
5. Few recent references on HNL catalyzed cyanohydrin
synthesis and screening for new HNL. A detailed referencing
can be found in the previous reviews (4a–4e). (a) Kobler, C.;
Bohrer, A.; Effenberger, F. Tetrahedron 2004, 60,
12. (a) van Langen, L. M.; Selassa, R.; Rantwijk, F.; Sheldon,
R. A. Org. Lett. 2005, 7, 327–329. (b) Veum, L.;
Hanefeld, U.; Pierre, A. Tetrahedron 2004, 60,
´ ´
10397–10410. (b) Solıs, A.; Luna, H.; Manjarrez, N.; Perez,
H. I. Tetrahedron 2004, 60, 10427–10431. (c) Avi, M.;
¨
Fechter, M. H.; Belaj, F.; Pochlauer, P.; Griengl, H.
¨
10419–10425. (c) Effenberger, F.; Ziegler, T.; Forster, S.
Tetrahedron 2004, 60, 10411–10418. (d) Kobler, C.;
Effenberger, F. Tetrahedron: Asymmetry 2004, 15,
3731–3742. (e) Han, S.; Chen, P.; Lin, G.; Huang, H.; Li, Z.
Tetrahedron: Asymmetry 2001, 12, 843–846. (f) Cruz Silva,
M. M.; Melo, M. L.; Parolin, M.; Tessaro, D.; Riva, S.;
Danieli, B. Tetrahedron: Asymmetry 2004, 15, 21–27. (g) Li,
N.; Zong, M.-H.; Peng, H.-S.; Wu, H.-C.; Liu, C. J. Mol. Catal.
B: Enzym. 2003, 22, 7–12. (h) Li, N.; Zong, M.-H.; Liu, C.;
Peng, H.-S.; Wu, H.-C. Biotechnol. Lett. 2003, 25, 219–222.
Degussa. Eur. Patent 276,375, 1987. (d) Kula, M, R.;
Stu¨rtz, I, M.; Wandrey, C.; Kragl, U. Forschungszentrum
Ju¨lich. Eur. Patent 446,826, 1991. (e) Becker, W.; Pfeil,
E. Boehringer. Ger. Patent 1,593,260, 1966. (f) Wehtje,
E.; Adlercreutz, P.; Mattiason, B. Appl. Microbiol. Biotechnol.
1988, 29, 419. (g) Westman, T, L. Monsanto, U.S. Patent
3,649,457, 1968.
13. Ognyanov, V. I.; Datcheva, V. K.; Kyler, K. S. J. Am. Chem.
Soc. 1991, 113, 6992–6995.
14. Bradford, M. Anal. Biochem. 1976, 72, 248–254.
¨
(i) Frohlich, R. F. G.; Zabelinskaja-Mackova, A. A.; Fechter,
M. H.; Griengl, H. Tetrahedron: Asymmetry 2003, 14,
¨
15. Gerrits, P. J.; Zumbragel, F.; Marcus, J. Tetrahedron 2001, 57,
8691–8698.
´ ´
355–362. (j) Hernandez, L.; Luna, H.; Ruız-Teran, F.;
´
Vazquez, A. J. Mol. Catal. B: Enzym. 2004, 30, 105–108.
´