Enzyme-promoted desymmetrisation of prochiral bis(cyanomethyl)phenylphosphine oxide

Article abstract of DOI:10.1016/j.tetasy.2007.08.027

Prochiral bis(cyanomethyl)phenylphosphine oxide has been successfully transformed into the corresponding optically active monoamide and monoacid with enantiomeric excesses ranging from low (15%) to very high (up to 99%) using a broad spectrum of nitrile-h

Full text of DOI:10.1016/j.tetasy.2007.08.027

Tetrahedron: Asymmetry 18 (2007) 2108–2112
Enzyme-promoted desymmetrisation of prochiral
bis(cyanomethyl)phenylphosphine oxide
a,
a
a
a
*
Piotr Kiełbasi n´ ski, Michał Rachwalski, Małgorzata Kwiatkowska, Marian Mikołajczyk,
b
c
b
d,
*
Wanda M. Wieczorek, Małgorzata Szyrej, Lesław Siero n´ and Floris P. J. T. Rutjes
a
Centre of Molecular and Macromolecular Studies, Polish Academy of Sciences, Department of Heteroorganic Chemistry,
Sienkiewicza 112, 90-363 Ł o´ d z´ , Poland
_
Institute of General and Ecological Chemistry, Technical University of Ł o´ d z´ , Zeromskiego 116, 90-924 Ł o´ d z´ , Poland
b
c
Jan Długosz University, Institute of Chemistry and Environmental Protection, al. Armii Krajowej 13/15, 42-201 Cz e˛ stochowa, Poland
d
Radboud University Nijmegen, Institute for Molecules and Materials, Toernooiveld 1, 6525 Nijmegen, The Netherlands
Received 27 July 2007; accepted 30 August 2007
Abstract—Prochiral bis(cyanomethyl)phenylphosphine oxide has been successfully transformed into the corresponding optically active
monoamide and monoacid with enantiomeric excesses ranging from low (15%) to very high (up to 99%) using a broad spectrum of
nitrile-hydrolysing enzymes.
ꢀ
2007 Elsevier Ltd. All rights reserved.
1. Introduction
under the action of amidases can be converted into the
appropriate acids (Scheme 1).
5
,6
Nitriles are important intermediates in synthetic chemistry
as they are simple to prepare and give access to a wide vari-
ety of interesting carboxylic acid derivatives. However, the
1
O
nitrilase
transformation of nitriles into the corresponding amides or
carboxylic acids usually requires rather harsh conditions,
for example, concentrated acids or bases, heavy metal salts
R
C
N
R
OH
2
nitrile
hydratase
(NHase)
or elevated temperatures. A solution for these problems
amidase
O
seems to be the enzymatic hydrolysis of nitriles. Within
the past decade, nitrile-transforming biocatalysts have
become very common both in academic laboratories and
R
NH2
3
in industry. These reactions can be conducted under mild
Scheme 1. Nitrile-hydrolysing enzymes.
conditions (aqueous buffers, neutral pH, ambient tempera-
ture) and have an additional advantage of being highly
4
The nitrilase-mediated selective hydrolysis of dinitriles to
cyanocarboxylic acids is of particular synthetic interest,
since a selective chemical hydrolysis of dinitriles is virtually
regio- and/or enantioselective. The enzyme-catalyzed
hydrolysis of nitriles can follow two different pathways.
According to the first one, nitrilases convert nitriles directly
into the corresponding acids without the amide as an
intermediate. In the second one nitrile hydratases (NHases)
transform nitriles into the corresponding amides, which
7
impossible.
Our previous work on the enzymatic desymmetrisation of
prochiral compounds and on the enzymatic resolution of
racemic substrates led to efficient approaches for the
synthesis of various classes of optically active com-
In particular, our results on the enzymatic
desymmetrisation of bis(cyanomethyl)sulfoxide repre-
sented the first example of a stereoselective hydrolysis of
2
,8–11
*
Corresponding authors. Tel.: +48 42 680 3234; fax: +48 42 6847126
P.K.); fax: +31 24 3653393 (F.P.J.T.R.); e-mail addresses:
pounds.
2
(
piokiel@bilbo.cbmm.lodz.pl; f.rutjes@science.ru.nl
0957-4166/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2007.08.027

P. Kiełbasi n´ ski et al. / Tetrahedron: Asymmetry 18 (2007) 2108–2112
2109
a dinitrile containing a prostereogenic centre located on a
heteroatom, the sulfinyl sulfur. As such, we clearly proved
that nitrile-hydrolysing enzymes are capable of recognising
and stereoselectively transforming substrates possessing a
prochiral heteroatom. We then investigated whether this
approach can also be applied to substrates with other
prochiral heteroatoms. To this end, we chose a model
substrate that contains a prochiral phosphinyl moiety,
that is, bis(cyanomethyl)phenylphosphine oxide 2. To
underline the relevance of these investigations, it should
be mentioned that P-chiral phosphorus compounds play
an important role as ligands in asymmetric catalysis,
mechanistic studies and as biologically active species.
possible products 3–7 (Scheme 3). Three of them 3, 5 and
6 may be obtained in each enantiomeric form. The hydro-
lyses were performed in a potassium phosphate buffer solu-
tion of pH 7.2 using a broad spectrum of nitrile-converting
enzymes, namely a whole cell preparation from Rhodococ-
1
5
cus erythropolis NCIMB 11540 (freeze dried cells) and a
series of commercially available nitrilases. The reaction
times varied from 72 to 144 h. The results are collected in
Table 1. Inspection of Table 1 shows that out of the five
possible products, only two, cyanomethylphenylphos-
phinylacetamide 3 and cyanomethylphenylphosphinyl-
acetic acid 5, were formed in various proportions
and enantioselectivities. Both products were isolated by
column chromatography and analysed spectroscopically.
1
31
Their structures were established by H NMR, P NMR,
1
3
2
. Results and discussion
C NMR, MS (CI) and HRMS (CI).
2.1. Synthesis of prochiral bis(cyanomethyl)phosphine
oxide 2
It should be stressed that under the conditions applied, the
conversion never reached 100%. In each case, certain
amounts of the unreacted substrate were isolated and in
three cases (entries 4, 6 and 8) the substrate was the main
component of the product mixture. Our investigations
showed that the enzyme-promoted hydrolysis of prochiral
bis(cyanomethyl)phenylphosphine oxide was not always
chemoselective, since in several cases, more than one prod-
uct was formed. Interestingly, most nitrilases, apart from
their normal behaviour, also exhibited nitrile hydratase
activity leading both to the monoacid 5 and monoamide
The starting material, prochiral bis(cyanomethyl)phos-
phine oxide 2, was synthesised via a modification of a
method that we described previously.12 It started from
dichlorophenyl phosphine, which upon treatment with
bromoacetonitrile in the presence of zinc dust in boiling
THF, provided bis(cyanomethyl)phosphine 1 in quantita-
tive yield. The latter compound was (without isolation) oxi-
dised with PhIO to give the desired phosphine oxide 2 in
2
1
,2,7
4
0% yield after purification (Scheme 2). Alternative meth-
3. Similar findings have already been reported.
The
ods for the synthesis of 2, as previously described by Dahl
stereoselectivity of the enzymes towards the heteroatom
prostereogenic centre (phosphoryl group) varied from low
to very high, being generally lower as compared to
1
3,14
et al.,
appeared to be unsuccessful in our hands.
2
2.2. Enzymatic hydrolysis of prochiral bis(cyanomethyl)-
phenylphosphine oxide 2
bis(cyanomethyl)sulfoxide and C-prochiral hydroxyglut-
4
aronitriles. The ee values of monoamide 3 and monoacid
5
were determined by chiral HPLC (Varian Pro Star 210,
The prochiral substrate 2 was subjected to a general enzy-
Chiralpak AS). In two instances, monoamide 3 was
matic procedure, which in principle, can lead to the five
obtained in excellent enantiomeric excess (96% and over
Ph
P
PhPCl2, Zn
NC
PhIO2
CHCl3
O
Ph
CN
NC
P
CN
Br
CN
THF
1
(100%)
2 (40%)
Scheme 2. Synthesis of bis(cyanomethyl)phenylphosphine oxide.
O
O
O
O
Ph
P
O
Ph
NC
P
NH2
2
H N
NH2
3
4
O
O
Ph
O
Ph
enzyme
buffer
NC
P
CN
NC
P
OH
5
2
O
O
O
O
O
Ph
O
Ph
P
P
H N
2
OH
HO
OH
6
7
Scheme 3. Possible products of enzymatic hydrolysis of 2.

2110
P. Kiełbasi n´ ski et al. / Tetrahedron: Asymmetry 18 (2007) 2108–2112
Table 1. Enzymatic hydrolysis of prochiral 2
Entry Enzyme Reaction Product, [a]
fourth one being only slightly different at a distant position
CH CO–N vs CH CO–O). The CD spectra are shown in
Figure 1.
a
(
D
ee [%] Abs.
conf.
2
2
time [h]
yield [%]
1
2
3
Nitrilase 101 144
2, 10.4
c
The monoamide (+)-(S)-3 exhibits two strong Cotton
effects—a positive one at 218 nm and a negative one at
223 nm. Obviously, the enantiomeric monoamide (À)-(R)-
3 exhibits the Cotton effects that are mirror images of the
above. In turn, the monoacid (À)-5 exhibits two Cotton
effects which are weaker than those of the amides. Since
the shapes of the curves and the signs of the Cotton effects
of (+)-(S)-3 and (À)-5 are similar, it is reasonable to
assume that the absolute configuration of the monoacid
is (À)-(S).
3, 4.6
À4.0
À1.5
56
53
(R)
(S)
c
5, 73.1
Nitrilase 102 144
2, 8.2
c
c
3, 9.6
À3.5
49
16
(R)
(S)
5, 82.2
À0.45
Nitrilase 103 144
2, 4.0
c
c
3, 8.6
À3.8
55
15
(R)
(S)
5, 85.3
À0.42
4
5
Nitrilase 105 144
Nitrilase 106 144
2, 80.0
, 7.4
2, 17.3
c
3
À2.6
36
(R)
b
b
3
, 10.8
, 51.0
+7.2
À2.0
>99
70
(S)
(S)
As far as the mechanism of the above desymmetrisation is
concerned, two courses of the reaction must be considered:
in the first one, monoamide 3 and monoacid 5 are formed
concurrently; in the second one the amide is formed first
and then subsequently hydrolysed to the acid. Identical
absolute configurations of both products obtained in the
same reaction (Table 1, entry 5) undoubtedly speak in
favour of the first assumption. The cases in which the
amide and the acid formed in the same reaction have oppo-
site absolute configurations (Table 1, remaining entries)
may suggest that the real stereorecognition is a result of
a kinetic resolution at the stage of the hydrolysis of the
initially formed racemic amide. However, in the latter case
the concurrent formation of both products cannot be
5
6
7
Nitrilase 107 144
2, 78.0
, 3.7
2, 16.0
c
3
À3.0
42
(R)
Nitrilase 108
72
c
c
3
, 18.7
, 70.7
À6.9
À1.7
96
60
(R)
(S)
5
d
8
Whole cells
144
2, 75.0
, 11.1
c
3
À2.8
39
(R)
a
b
c
In D
2
O + CD
3
COCD
3
(c 1).
Determined by chiral HPLC.
Enantiomeric excess, determined by comparison of [a]
Rhodococcus erythropolis NCIMB 11540 (freeze dried for the cases, in
D
values.
d
which ee’s were determined by chiral HPLC cells).
1
excluded, as has been shown by Sheldon et al., who
9
9%) albeit in yields of 18.7% and 10.8%, respectively
proposed a bidirectional mechanism of action of nitrilases
leading to both types of products. Taking into account the
fact that amides are hardly accepted as substrates by
nitrilases, the bidirectional mechanism seems more
reasonable.
(
entries 5 and 7). Since compound 3 proved to be crystal-
line, we were able to perform X-ray analysis and determine
its absolute configuration as (+)-(S) (Fig. 2). The highest ee
value of monoacid 5 was 70% (entry 5). Since 5 appeared to
be an oil and, moreover, was never obtained in a fully
enantiopure form, it was more difficult to determine its
absolute configuration. Therefore, we decided to measure
the CD spectra of amides (+)-(S)-3 and (À)-(R)-3 and of
acid (À)-5, assuming that comparison of the Cotton effects
exhibited by each sample would allow us to determine the
absolute configuration of the acid. This assumption seemed
justifiable because the stereogenic phosphorus atom is in
both products bound to three identical substituents, the
3. X-ray analysis
The crystal structure of (+)-3 contains two crystallograph-
ically independent molecules (A and B) in the asymmetry
unit (Fig. 2).
Figure 2. The molecular structure of two independent molecules A and B
of (+)-3, showing 50% probability displacement ellipsoids and the atom-
numbering scheme.
The angles about the P atom range from 103.06(8)ꢁ to
103.27(8)ꢁ (C1–P1–C3) and from 115.92(8)ꢁ to 115.92(8)
Figure 1. The CD spectra of compounds 3 and 5.

P. Kiełbasi n´ ski et al. / Tetrahedron: Asymmetry 18 (2007) 2108–2112
2111
1
(
O1–P1–C1) in molecules A and B, respectively, indicating
136–137 ꢁC;
H
NMR (CD COCD ): d = 3.83 (d,
3 3
31
a distorted tetrahedral environment. The absolute configu-
ration at the atom P is (S) in both crystallographically
independent molecules. In both molecules a significant
shortening of the P1–C1 bond lengths is observed:
J = 14.2 Hz, 4H), 7.96–8.03 (m, 5H);
(CD COCD ): d = 24.4; C NMR (CD COCD ): 20.8,
22.1, 114.41 (d), 130.8 (d), 132.6 (d), 135.3. MS (CI): m/z
205 (M+H). HRMS (CI) calcd for C H N OP:
P
NMR
1
3
3
3
3
3
1
0
9
2
˚
˚
1
.7877(17) A for molecule A and 1.7939(16) A for molecule
204.0454; found 204.0458.
˚
B; a typical P–C single bond length is 1.816(2) A.
5.3. Enzymatic hydrolysis of 2—general procedure
4
. Conclusions
Compound 2 (0.050 g, 0.245 mmol) was dissolved in a buf-
fer solution and the enzyme (5 mg) was added. The reac-
tion was shaken at 30 ꢁC for 72–144 h (see Table 1) and
monitored by TLC. Water was then evaporated and the
residue was separated using column chromatography
Enzymatic hydrolysis of prochiral bis(cyanomethyl)phenyl-
phosphine oxide was achieved for the first time using a
broad spectrum of nitrile-converting enzymes under mild
conditions (buffer solution of pH 7.2, 30 ꢁC). Two products
were formed, cyanomethylphenyl-phosphinylacetamide
and cyanomethylphenyl-phosphinylacetic acid, in different
proportions and various enantioselectivity ranging from
(chloroform/methanol in gradient) to yield the correspond-
ing products 3 and 5. The yields and specific rotations are
shown in Table 1.
1
5% to 99% ee. The absolute configurations of the products
5
.3.1. Cyanomethylphenylphosphinylacetamide 3. White
have also determined.
1
crystals
(MeCN),
mp = 183–185 ꢁC;
H
NMR
(D O + CD COCD ): d = 3.63 (AB, 2H), 3.86 (AB, 2H),
2
3
3
3
1
7
.75–7.84 (m, 5H);
P NMR (D O + CD COCD ):
2 3 3
5. Experimental
1
3
d = 33.9; C NMR (D O + CD COCD ): 18.03, 38.80,
2
3
3
1
17, 133.4, 169.06. MS (CI): m/z 223 (M+H). HRMS
5
.1. General
(CI) calcd for C H N O P: 222.0556; found 222.0553.
1
0
11
2
2
The enzymes were purchased from BioCatalytics Europe
GmbH, Grambach, Austria. NMR spectra were recorded
5
.3.2. Cyanomethylphenylphosphinylacetic acid 5. Yellow
1
on Bruker instruments at 200 MHz with D O and
2
oil; H NMR (D O + CD COCD ): d = 2.93 (AB, 2H),
2 3 3
31
CD COCD as solvents. Optical rotations were measured
3
3
3.06 (AB, 2H), 7.47–7.82 (m, 5H); P NMR (D O + CD -
2 3
1
3
on a Perkin–Elmer 241 MC polarimeter (c = 1). Column
chromatography was carried out using Merck 60 silica
gel. TLC was performed on Merck 60 F254 silica gel plates.
The enantiomeric excess (ee) values were determined by
chiral HPLC (Varian Pro Star 210, Chiralpak AS).
COCD ): d = 17.8; C NMR (D O + CD COCD ): 21.46,
3
(M+H), 180 (MÀCO2).
3
2
3
3
1.80, 118.45, 129.3, 133.8, 179.4. MS (CI): m/z 224
6
. Crystallographic data
5
.2. Synthesis of prochiral 2
Molecular formula: C H N O P. Formula weight:
1
0
11
2
2
THF (30 mL) and zinc dust (1.31 g, 0.02 mol) were placed
in a 100 mL three-neck round-bottom flask equipped with
a mechanical stirrer, dropping funnel and reflux condenser
M = 222.18. Crystallographic system: monoclinic. Space
r
˚
˚
group: P2 ; a = 13.457(3) A, b = 5.7298(11) A, c =
3.966(3) A, a = 90.00ꢁ, b = 91.49(3)ꢁ, V = 1076.5(4) A ,
Z = 4, D = 1.371 Mg m . Mo Ka radiation. Cell para-
1
3
˚
˚
1
with a CaCl tube. A mixture of bromoacetonitrile (2.40 g,
2
À3
x
0
.02 mol) and dichlorophenyl phosphine (1.79 g, 0.01 mol)
meters from 5646 reflections: h = 2.9–31.4ꢁ, l = 0.24
in THF (5 mL) was slowly added to the mixture under
À1
mm , T = 295(2) K.
gentle reflux. During the addition, a crystal of I was added
2
to initiate the reaction (the brown colour of iodine
disappeared almost immediately). After the addition, the
mixture was very vigorously stirred and refluxed for 1 h.
Crystallographic data for (+)-(S)-3 have been deposited
with the Cambridge Crystallographic Data Center No.
CCDC 654477 and can be obtained free of charge on appli-
cation to CCDC, 12 Union Road, Cambridge CB2 1EZ,
UK [fax: (internat.) +44-1223/336-033; e-mail: mailto:
deposit@ccdc.cam.ac.uk].
3
1
The reaction was monitored by P NMR and by TLC.
3
1
After 1 h, the P NMR spectrum showed one signal
(
d = À29.4) and the TLC plate revealed only one spot.
After cooling to room temperature, water was added and
the mixture was extracted with chloroform (3 · 20 mL).
The combined organic layers were dried over anhydrous
MgSO . After evaporating off CHCl phosphine 1 was
Acknowledgements
4
3
obtained in quantitative yield. This compound was
dissolved in chloroform and PhIO2 (1.18 g, 0.005 mol)
was added to the solution. The reaction was followed by
The work was carried out under the Polish–Dutch Joint
Research Project: ‘Dynamic kinetic resolution and desym-
metrisation of sulfur and phosphorus compounds using
enzymes’. Financial support from the Ministry of Science
and Information Society Technologies (Poland), Grant
No. 3 T09A 166 27 for P.K., is gratefully acknowledged.
TLC and after its completion CHCl was evaporated and
3
the residue was separated via column chromatography
(
CH Cl /acetone in gradient) to obtain the desired
2 2
phosphine oxide 2 (8.16 g, 40%) as a white powder, mp

2112
P. Kiełbasi n´ ski et al. / Tetrahedron: Asymmetry 18 (2007) 2108–2112
_
References
8. Kiełbasi n´ ski, P.; Zurawi n´ ski, R.; Albrycht, M.; Mikołajczyk,
M. Tetrahedron: Asymmetry 2003, 14, 3379–3384.
9
. Kiełbasi n´ ski, P.; Mikołajczyk, M. In Enzymes in Action:
Green Solutions for Chemical Problems; Zwanenburg, B.,
Mikołajczyk, M., Kiełbasi n´ ski, P., Eds.; Kluwer: Dordrecht,
1
2
3
. Fernandes, B. C. M.; Mateo, C.; Kiziak, C.; Chmura, A.;
Wacker, J.; van Rantwijk, F.; Stolz, A.; Sheldon, R. A. Adv.
Synth. Catal. 2006, 348, 2597–2603.
. Kiełbasi n´ ski, P.; Rachwalski, M.; Mikołajczyk, M.; Szyrej,
M.; Wieczorek, M. W.; Wijtmans, R.; Rutjes, F. P. J. T. Adv.
Synth. Catal. 2007, 349, 1387–1392.
. Kubac, D.; Cejkova, A.; Masak, J.; Jirku, V.; Lemaire, M.;
Gallienne, E.; Bolte, J.; Stloukal, R.; Martinkova, L. J. Mol.
Catal. B: Enzym. 2006, 39, 59–61.
2000; pp 161–191.
0. Mikołajczyk, M.; Kiełbasi n´ ski, P.; Zurawi n´ ski, R.; Wieczor-
ek, M. W.; Błaszczyk, J. Synlett 1994, 127–129.
_
1
1
1. Kiełbasi n´ ski, P.; Rachwalski, M.; Mikołajczyk, M.; Moe-
lands, M. A. H.; Zwanenburg, B.; Rutjes, F. P. J. T.
Tetrahedron: Asymmetry 2005, 16, 2157–2160.
1
2. Kiełbasi n´ ski, P.; Mikołajczyk, M. Synthesis 1995, 144–
4
5
6
7
. Wang, M.-X.; Liu, C.-S.; Li, J.-S. Tetrahedron: Asymmetry
1
46.
2
001, 12, 3367–3373.
1
1
3. Dahl, O. Acta Chem. Scand. B 1976, 30, 799–800.
4. Dahl, O.; Henriksen, U.; Trebbien, C. Acta Chem. Scand. B
. Enzyme Catalysis in Organic Synthesis; Drauz, K., Waldmann,
H., Eds.; Wiley-VCH: Weinheim, 2002; Vol. 2, pp 699–716.
. Faber, K. Biotransformations in Organic Chemistry, 4th ed.;
Springer: Berlin, 2004, pp 149–159.
1
983, 37, 639–642.
1
5. Vink, M. K. S.; Wijtmans, R.; Reisinger, C.; van den
Berg, R. J. F.; Schortinghuis, C. A.; Schwab, H. A.;
Schoemaker, H. E.; Rutjes, F. P. J. T. Biotechnol. J. 2006,
. (a) Effenberger, F.; Osswald, S. Synthesis 2001, 1866–1872;
(
b) Zhu, D.; Mukherjee, C.; Biehl, E. R.; Huaa, L. Adv.
1, 569–573.
Synth. Catal. 2007, 349, 1667–1670.