Preparation of N-unsubstituted β-ketoamides by Rhodococcus rhodochrous-catalysed hydration of β-ketonitriles

Article abstract of DOI:10.1016/j.tet.2003.10.096

A varied set of N-unsubstituted β-ketoamides, hardly obtainable or non-accessible by non-enzymatic methods, have been synthesized, with good to excellent yields, by the generally fast hydration of the corresponding β-ketonitriles, catalysed by the bacterium Rhodococcus rhodochrous IFO 15564. This bacterium shows nitrile hydratase and amidase activities, the latter being inhibited during its growth phase with diethyl phosphoramidate (DEPA). Optimization of the processes and studies concerning large-scale biotransformations were also carried out. β-Ketoamides exist as keto-enol mixtures whose composition depends on their substituents and varies with solvent polarity.

Full text of DOI:10.1016/j.tet.2003.10.096

Tetrahedron 60 (2004) 607–618
Preparation of N-unsubstituted b-ketoamides by Rhodococcus
rhodochrous-catalysed hydration of b-ketonitriles
†
*
Vicente Gotor, Ramon Liz and Ana Mª Testera
´
¯
Departamento de Quımica Organica e Inorganica, Universidad de Oviedo, 33071 Oviedo, Spain
´
´
´
Received 20 June 2003; revised 25 July 2003; accepted 17 October 2003
Abstract—A varied set of N-unsubstituted b-ketoamides, hardly obtainable or non-accessible by non-enzymatic methods, have been
synthesized, with good to excellent yields, by the generally fast hydration of the corresponding b-ketonitriles, catalysed by the bacterium
Rhodococcus rhodochrous IFO 15564. This bacterium shows nitrile hydratase and amidase activities, the latter being inhibited during its
growth phase with diethyl phosphoramidate (DEPA). Optimization of the processes and studies concerning large-scale biotransformations
were also carried out. b-Ketoamides exist as keto–enol mixtures whose composition depends on their substituents and varies with solvent
polarity.
q 2003 Elsevier Ltd. All rights reserved.
1. Introduction
The enzymes involved in the desired biotransformation, i.e.
nitrile hydratases (EC 4.2.1.84), are not commercially
available, which forced us to use an appropriate whole
cell culture. However, the bacterial strains able to hydrolyze
nitriles can act according to one or both of the following
routes: (i) a nitrile hydratase converts the nitrile into an
amide, which is subsequently transformed into a carboxylic
acid by an amidase; (ii) a nitrilase directly converts the
nitrile into an acid. The presence of amidases and/or
nitrilases is a serious drawback for our purposes of stopping
the process in the amide stage, and it should be avoided.
b-Ketoamides are versatile intermediates for the synthesis
of heterocycles, for instance 3-acyltetramic acids1 (used in
the total synthesis of tirandamycin2 and other related natural
antibiotics), pyrans,3 alkaloids,4 lactams and spirolactams,5
and azetidin-2-ones,6 as well as several 3-hydroxyisothiazol
bioisosteres of glutamic acid and analogs of the AMPA
receptor agonist.7 Recently, some b-ketoamides have been
converted into g-ketoamides, a class of compound related
with a wide variety of biologically relevant systems.8
In connection with our current interest in the bioreduction of
b-ketoamides and their subsequent transformation into non-
racemic b- and g-aminoalcohols,9 we needed a set of
N-unsubstituted b-ketoamides as starting materials. This
type of b-ketoamide is hardly accessible by means of their
usual preparative methods,10 mostly directed towards the
N-substituted ones. For such a cause, we decided to prepare
them via the hydration of the corresponding b-ketonitriles, a
commercially or easily available kind of compound. After a
series of failed attempts using traditional, non-enzymatic
methodologies, we envisaged that enzymatic hydration of
b-ketonitriles would be a suitable, environmentally benign
way of reaching our objective.
A solution would be the use of some bacterial strains
(Rhodococcus sp. N774, Pseudomonas chlororaphis B23,
Rhodococcus rhodochrous J1) which develop nitrile
hydratases almost exclusively when growing under appro-
priate conditions. Consequently, they have been used for an
important industrial biotransformation (Nitto Chemical
Industry Co.) leading to acrylamide, as well as for the
preparation of other commodity chemicals.11 One of these
strains (P. chlororaphis B23) has also been employed for
the industrial conversion of adiponitrile to 5-cyanovaler-
amide (DuPont),12 as the first step in the manufacture of a
new herbicide. Unfortunately, these strains are commer-
cially unavailable.
Apart from these examples, the scarcely reported bacterial
transformations of nitriles into amides13 seem to be related
to an accidental, unfavourable steric interaction between the
amidase and the corresponding amide, which slows down its
further conversion to acid. Thus, several ortho-substituted
benzonitriles (but not their meta- or para-isomers) and five
membered hetarene-2-carbonitriles were transformed into
Keywords: Amido-ketones; Enzyme inhibitors; Microbial reactions;
Rhodococcus rhodochrous; Tautomerism.
*
Corresponding author. Tel./fax: þ34-985-103448;
e-mail address: vgs@sauron.quimica.uniovi.es
†
´
´
´
Present address: Departamento de Quımica Organica, Escuela Tecnica
Superior de Ingenieros Industriales, Universidad de Valladolid, 47071
Valladolid, Spain.
0040–4020/$ - see front matter q 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2003.10.096

608
V. Gotor et al. / Tetrahedron 60 (2004) 607–618
their corresponding amides after short reaction times with
R. rhodochrous AJ270,14 another inaccessible strain. In the
same way, the hydration carried out with R. rhodochrous
IFO 15564 of three cyanocarbohydrate derivatives to
amides,15 as well as that of 3-cyanoprop-1-enylboronic
acid,16 can probably be rationalised.
When applied on b-ketonitriles, it worked acceptably on 1d
again (entry 4), but not on the presumably acid-stable
pivaloylacetonitrile 1b (entry 5), and even less on the acid-
labile heterocyclic 1m (entry 6).
It can be predicted that hydration of b-ketonitriles under
basic conditions would not work, since the abstraction of the
acidic a-hydrogen will prevent a further nucleophilic attack
to the cyano group.21 Indeed, using 1d as a test substrate, we
observed no reaction after heating with potassium hydrox-
ide in tert-butyl alcohol (entry 7),22 and neither with
potassium trimethylsilanolate in tetrahydrofuran (entry 8).23
Last, the use of basic hydrogen peroxide under phase-
transfer catalysed conditions (entry 9)21 converted 1d into
benzoate anion, presumably via a retro-Claisen type
reaction, opposite to that used by us for preparing some of
the b-ketonitriles employed in this study (see Section 3.2).
On the other hand, it has recently been demonstrated that
diethyl phosphoramidate [H2N–P(O)(OEt)2; DEPA]
behaves as an amidase inhibitor for three nitrile hydrolyzing
bacterial strains when exposed to 2-arylpropanenitriles and
to 2-arylpropanamides. With Agrobacterium tumefaciens
strain d3, 2-arylpropanenitriles were enantioselectively
converted into 2-arylpropanamides,17 as the hydration of
the nitrile to amide was not inhibited by DEPA;17b similar
results were observed for the strains Rhodococcus equi
A418a and Rhodococcus sp. MP50.18b
The above observations prompted us to undertake the
biotransformations of a varied set of b-ketonitriles into their
corresponding b-ketoamides. This paper describes our
findings with the concourse of the bacterium R. rhodochrous
IFO 15564 in metabolic resting phase. Since the obtained
b-ketoamides are also accepted as substrates by this strain,
we have reached our goal by using DEPA as amidase
inhibitor. To the best of our knowledge, bacterial hydration
of b-ketonitriles has not been previously performed for
preparative purposes.
Two enzymatic reactions with an acid-stable (entry 10) and
an acid-labile b-ketonitrile (entry 11) were also included in
Table 1 for comparison. From their results, it can be
advanced that the bacterial biotransformation is the method
of choice for hydrating b-ketonitriles to b-ketoamides, as
will be plain in the sections to come.
2.3. Bacterial strain selection
After checking that no conventional procedure was able to
convert the majority of b-ketonitriles into b-ketoamides, we
decided to submit them to microbial hydration. As
previously stated, except for a few anecdotal cases,24 the
overwhelming majority of known nitrile-converting micro-
organisms are bacteria. After discarding those exhibiting
nitrilase activity, our attention was focused on four of them
available at some culture collection, and able to transform
nitriles via nitrile hydratase and amidase activities:
Brevibacterium imperiale B222 (CBS 498.78),13f Rhodo-
coccus sp. R312 (CBS 717.73),25,26 Rhodococcus erythro-
polis IFO 1253927 and R. rhodochrous IFO 15564.15,16,28
2. Results and discussion
2.1. Choice of the starting b-ketonitriles
Eighteen varied b-ketonitriles, R1-CO–CHR2–CN (1),
were selected for this study (see Table 3). Sixteen of them
(1a–r) show no substituent at a position (i.e., R2¼H). Three
of them have R1 of aliphatic nature (1a–c), although 1c
includes a hetarene ring in R1. The following eight (1d–k)
possess aromatic R1 groups, with or without electron-donor
or electron-withdrawing substituents. In five cases (1m–r),
R1 is heterocyclic, both p-excessive and p-deficient, as well
as mono- or bicyclic. Last, two starting materials (1s,t) are
a-substituted, showing their R1 and R2 groups having
interchanged aliphatic and aromatic nature. Some of them
(1c,m,n and, in lesser extent, 1r) are acid-labile. Such a
selection seems to be wide enough to test the versatility of
the biotransformation, as well as to compare with the
following non-enzymatic reactions.
Benzoylacetonitrile (1d) was chosen as a reference substrate
to evaluate the enzymatic activities of these four strains.
Bacteria were grown both in the presence and in the absence
of DEPA. Incubations with 1d were carried out with resting
phase bacteria and monitored by TLC. With B. imperiale
neither benzoylacetamide (2d) nor benzoylacetic acid (3d)
were detected (see Scheme 1). Amidase from Rhodococcus
sp. R312, after growing this bacterium in the absence of
DEPA, was much more active than the corresponding nitrile
hydratase as only 1d and 3d, but not the ketoamide 2d, were
observed during the biotransformation. The amidase
activity of the bacterium clearly decreased if growing in
the presence of DEPA, although not enough to prevent a
considerable fall in the ketoamide yield.
2.2. Non-enzymatic attempts
A variety of conventional reactions to transform nitriles into
their corresponding amides were tried in order to find a
general method for a wide range of b-ketonitriles 1 (see
Table 1). First, the classical hydration using hydrogen
chloride dissolved in concentrated hydrochloric acid was
employed,19 which led to acceptable results with the acid-
stable starting material 3-oxo-3-phenylpropanenitrile (ben-
zoylacetonitrile, 1d; entry 1), but not with acid-labile
heterocyclic substrates (entries 2 and 3). Another recent,
mild and acid mediated way of converting nitriles into
amides implies the use of chlorotrimethylsilane–water.20
On the contrary, R. rhodochrous has been described to show
higher nitrile hydratase than amidase activity. In fact, in our
hands both this strain and R. erythropolis gave the best
results in transforming 1d into 2d, especially after growing
them in the presence of DEPA. Under certain reaction
conditions, they allow the isolation of high yields of the
ketoamide 2d, the ketonitrile 1d and the ketoacid 3d being
observed only in very small or null amounts. A comparative

V. Gotor et al. / Tetrahedron 60 (2004) 607–618
609
Table 1. Non-enzymatic attempts at hydratinga R1-CO–CHR2–CN (1)b to R1-CO–CHR2–CONH2 (2)
Entry
Substratec
R1
Reagents
Reaction time (h)
2, Crude yield (%)/remarks
70
13/Notable polymerisation
0/Full polymerisation
71
36
5
No reaction
No reaction
Retro-condensation to PhCO2H
1
2
3
4
5
6
7c
8d
9
10e
11e
1d
1m
1c
Ph
2-Furyl
HCl(g)/HCl(aq)
13
13
13
23
39
39
2
22
2
0.67
1.2
HCl(g)/HCl(aq)
HCl(g)/HCl(aq)
Me3SiCl/H2O
Me3SiCl/H2O
Me3SiCl/H2O
KOH/ButOH
(1-Methylpy-rrol-2-yl)methyl
Ph
1d
1b
1m
1d
1d
1d
1d
1m
But
2-Furyl
Ph
Ph
Ph
Ph
2-Furyl
Me3SiO2Kþ/THF
H2O2/NaOH(aq)/Bu4NþCl2
R. rhodochrous/H2O
R. rhodochrous/H2O
96
90
a
The reactions were carried out at room temperature except for entries 7, 8, 10 and 11.
The four substrates selected for this table have R2¼H.
Reaction temperature, 83 8C.
Reaction temperature, 66 8C.
Enzymatic process included for the sake of comparison; reaction temperature, 28 8C.
b
c
d
e
analysis of the 1H NMR spectra of the crude products
allowed us to select R. rhodochrous IFO 15564 as the most
suitable bacterial strain for the enzymatic hydration of
b-ketonitriles to b-ketoamides.
values, ranging from 14% (first row) to 27% (sixth row) of
avoided b-ketoacid formation. However, DEPA also shows
an unexpected feature as promoter of the nitrile hydratase
activity, as deduced from the comparison of both 1d-
columns: the average reaction rate (rows 1 to 4) of the
ketonitrile to ketoamide step is ca. 18% higher when the
bacterium was grown in the presence of DEPA. Thus, the
actual amidase inhibitory power of DEPA is higher than that
merely inferred from the DS values since, with its
mediation, less b-ketoacid is formed just when more
b-ketoamide is present in the reaction medium.
2.4. Optimization of the growth and biotransformation
conditions
In order to determine the optimal conditions for obtaining
the best yield of benzoylacetamide (2d) from benzoylace-
tonitrile (1d), as well as to check the amidase inhibitory
power of DEPA, we carried out a series of biotransform-
ations (see Table 2). Half of these processes were carried out
with R. rhodochrous grown in the presence of DEPA, and
the other half in the absence of this amidase inhibitor. The
percentage values of compounds 1d and 2d were deduced
Apart from the obvious need of growing R. rhodochrous in
the presence of DEPA, the ideal reaction conditions for
preparing benzoylacetamide (2d) must include an adequate
balance between the nitrile hydratase concentration (i.e., the
absorbance or optical density of the bacterial suspension)
and the reaction time. Data of Table 2 show that an A650
(absorbance of the bacterial suspension at l¼650 nm) value
of 3.0 and a reaction time of 40 min lead to an excellent
yield of 2d. Data of row 6, if compared with those of row 3,
clearly prove how longer reaction times favours the
formation of b-ketoacid 3d, which is an evident disadvan-
tage in order to reach a good ketoamide yield.
1
from the H NMR spectra of the crude reaction products,
which were obtained from the cell-free supernatant reaction
medium by means of continuous extraction and further
solvent elimination. This work-up precludes the possibility
of isolating benzoylacetic acid, 3d (or other b-ketoacids 3
when other substrates were used), since the extractions were
carried out at pH 8. No efforts were made for detecting
b-ketoacids 3; however, their undesirable formation was
proved: (1) by TLC monitoring of the bioconversions;29 (2)
by the presence in some crude products of the ketones 4,
coming from the decarboxylation of the corresponding 3
during the biotransformation at 28 8C30 (see Scheme 1).
Among the standard reaction conditions for the other
substrates, we adopted the above A650 value of 3.0, the
biotransformations being stopped after TLC showed the
disappearance of the corresponding b-ketonitrile 1. The use
of more concentrated bacterial suspensions causes some
problems of entrapping reaction products. Nevertheless, this
method or other variations were applied when necessary, as
described in the next Sections 2.5 and 2.6.
From the above paragraph, it is clear that, in each process
contained in Table 2, the difference 1002S roughly
represents the molar percentage of b-ketoacid coming
from the amidase activity of the bacterium. Thus, by
comparing both columns of S values, the efficiency of
DEPA as amidase inhibitor can be estimated from the DS
It has been described that DEPA acts as an amidase inhibitor
Scheme 1. (i) R. rhodochrous IFO 15564 (nitrile hydratase activity), aq. phosphate buffer pH 8, 28 8C, 200 rpm; (ii) R. rhodochrous IFO 15564 previously
grown in the presence of DEPA, aq. phosphate buffer pH 8, 28 8C, 200 rpm; (iii) R. rhodochrous IFO 15564 (amidase activity) previously grown in the absence
of DEPA, aq. phosphate buffer pH 8, 28 8C, 200 rpm.

610
V. Gotor et al. / Tetrahedron 60 (2004) 607–618
Table 2. Molar percentages of recovered substances in R. rhodochrous-catalysed hydration of benzoylacetonitrile (1d)a
b
Row
A650
Reaction time
(min)
R. rhodochrous grown in the absence
of DEPA
R. rhodochrous grown in the presence
of 15 mM DEPA
1d
2d
Sc
1d
52
38
27
2d
Sc
1
2
3
4
5
6
1.0
1.0
1.0
2.0
3.0
1.0
10
20
30
30
40
240
60
48
38
27
4
22
30
39
53
71
42
82
78
77
80
75
52
44
57
66
79
96
79
96
95
94
95
96
79
16
n.o.d
n.o.d
10
a
1d (0.70 mmol, 102 mg) in EtOH (1.0 mL) was added to a suspension of R. rhodochrous in 0.10 M potassium phosphate buffer (100 mL, pH 8.0); rotary
shaking (200 rpm) at 28 8C.
b
c
d
Absorbance (optical density) of the bacterial suspension at 650 nm.
Sum of percentages of 1d and 2d.
Not observed by 1H NMR.
if added to the reaction mixture during the biotransform-
conditions and will be briefly discussed in the next
paragraphs. Since typical bacterial nitrile hydrolyses were
carried out in aqueous buffers containing ca. 0.1% [w(g)/
v(mL)] of substrate,13e we usually started from 100–130 mg
(0.60–0.80 mmol) of a b-ketonitrile 1 and a R. rhodochrous
suspension in 100 mL of aqueous 0.10 M potassium
phosphate buffer at pH 8.0 (A650¼3.0). The same results
are roughly obtained if the pH varies in the range 6–8. For
other details concerning the standard methodology, see
Sections 3.3 and 3.4. Other non-standard processes will be
discussed in Section 2.6.
ation,17,18 which has the drawback of leading to an impure
crude reaction product needing chromatographic purifi-
cation. Thus, our observation is noteworthy that DEPA also
works if added only during the growth of R. rhodochrous.
This was our standard methodology, since it usually allows
us to obtain crude b-ketoamides as NMR-pure compounds
ready to use, for instance, as starting materials for further
fungal bioreductions.
2.5. Preparative biotransformations
Scheme 1 shows the straightforward course of these
biotransformations which, for preparative purposes, have
to be stopped in the b-ketoamide 2 stage. Table 3
summarizes the preparation of b-ketoamides 2a–t. Stan-
dard reaction conditions already were outlined in Section
2.4. A few cases deviated somewhat from the standard
All b-ketoamides yields lie in the good to excellent range
71–94%, except for the ketoamide 2q (60%). Most of the
crude products are pure b-ketoamides as their H NMR
spectra reveal. However, trace amounts of the b-ketonitriles
1 or the ketones 4 are also present in a few cases.
Exceptionally, b-ketoamides 2i and 2q are contaminated
1
Table 3. b-Ketoamides 2 obtained by R. rhodochrousa-catalysed hydrationb of b-ketonitriles 1c
Compound
R1
R2
Reaction time (min)
Yield (%)d
Compound
R1
R2
Reaction time
Yield (%)d
2a
2b
2c
Pr
H
H
H
90
25
120
87
90
91
2j
4-Cl–C6H4
3,4-Cl2C6H3
2-Furyl
H
H
H
110 min
60 min
70 min
72
71
87
But
2ke
2m
2n
3-Furyl
2-Thienyl
3-Pyridyl
H
H
H
H
40 min
75 min
9.5 h
87
89
60
75
2p
2qf
2rg
2d
Ph
H
40
93
14 h
2e
2f
2g
2h
2ii
3-Me–C6H4
4-Me–C6H4
4-MeO–C6H4
3-F3C–C6H4
3-Cl–C6H4
H
H
H
H
H
60
60
70
215
45
93
88
94
89
85
2sh
2t
Me
Ph
Ph
Et
23.5 h
190 min
79
77
a
Grown in sterilized medium M (see Section 3.3) plus 15 mM DEPA, except for 2s.
b
c
d
e
f
In a bacterial suspension (A650¼3.0, unless otherwise specified) in aq. 0.10 M phosphate buffer (100 mL), pH 8.0 (28 8C, 200 rpm).
Except for 1i,k,r, 100–130 mg (0.60–0.80 mmol) of starting material 1 in ethanol (1.0 mL) were added to the bacterial suspension.
After flash column chromatography purification.
Seventy four milligrams (0.35 mmol) of 1k and 0.74 mL of ethanol were employed.
The biotransformation was carried out with a bacterial absorbance of 6.7.
Fifty milligrams (0.25 mmol) of 1r and 0.50 mL of ethanol were employed; the biotransformation was carried out with a bacterial absorbance of 14.0.
Bacteria were grown in sterilized medium N (see Section 3.3) plus 15 mM DEPA; the biotransformation was carried out with a bacterial absorbance of 11.6.
60 mg (0.33 mmol) of 1i and 0.60 mL of ethanol were employed; the biotransformation was carried out in the presence of 5 mM DEPA.
g
h
i

V. Gotor et al. / Tetrahedron 60 (2004) 607–618
611
by remarkable amounts of DEPA and ketone 4q, respect-
ively, due to their special reaction conditions (see below).
Nevertheless, purification of all these compounds is easily
and efficiently accomplished by flash column chromato-
graphy (see Section 3.4). For instance, from the crude
product obtained after the biotransformation of 1q (85:15
mixture 2q/4q; crude yield, 63%), 60% yield of pure
nicotinoylacetamide (2q) was reached.
of some unfavourable steric and/or electronic interaction
between the amide 2s and the amidase.32
Finally, a b-ketonitrile not included in Table 3 also deserves
some comments. 3-(2-Chlorophenyl)-3-oxopropanenitrile
(1, R1¼2-Cl–C6H4, R2¼H) is an ortho-substituted aromatic
substrate whose solubility in aqueous media is very low.
Nonetheless, it was accepted by R. rhodochrous under
standard conditions, TLC analysis showing a little amount
of b-ketoamide, but not b-ketoacid, formed after 2 h. As the
time elapses (10 h), most of the ketonitrile remains, a clear
spot of ketoacid emerges, but only a trace of ketoamide is
seen. Several variations were established in order to
overcome the low solubility and/or to stop the process at
the ketoamide stage (use of half amount of starting material,
increase of the bacterial absorbance, presence of DEPA
during biotransformation, change of the culture medium),
some of them even simultaneously. However, the best yield
in ketoamide was only 27%, this product being always
accompanied by ketonitrile and 2-chloroacetophenone.
Thus, we must conclude that this ortho-substitution
negatively affects the nitrile hydratase activity of
R. rhodochrous.
All chlorinated b-ketonitriles (1i–k) and, especially, 3-(1-
methylindol-3-yl)-3-oxopropanenitrile (1r) are only slightly
soluble in the aqueous phosphate buffer, so that their
bacterial hydration reactions are very slow. For instance,
using 1r as substrate under standard conditions, only ca.
17% of the corresponding amide 2r was obtained after 5
days, together with ca. 40% of the remaining substrate
(entrapped in the bacterial cake, from which its recovery is
difficult) and ca. 40% of 3-acetyl-1-methylindole (4r).31
Since much substrate was wasted, we decided to diminish its
amount (only 50 mg, 0.25 mmol), and moreover, to increase
the bacterial optical density to an A650 value of 14.0, in order
to accelerate the biotransformation. Both experimental
changes led to 80% crude yield of 2r (75% after
chromatographic purification), in spite of the long time
(14 h) required for the disappearance of the substrate.
2.6. Biotransformations on a larger scale
A good 71% yield of 2,3-dichlorobenzoylacetamide (2k)
was obtained by merely starting from ca. half of the usual
substrate amount. In the case of 3-chlorobenzoylacetamide
(2i), the best yield (85%) was reached not only by using a
reduced amount of 1i, but also in the presence of 5 mM
DEPA during the biotransformation. Under standard
conditions, only 54 and 64% of 2k and 2i, respectively,
could be obtained. In the case of 4-chlorobenzoylacetamide
(2j), however, the standard methodology led to a good 74%
yield, and no relevant improvements were observed by
changing it.
Apart from the standard reactions used for synthetic
purposes, it would be interesting to know to what extent
the catalytic potential of the R. rhodochrous cells could be
exploited. Thus, choosing benzoylacetonitrile (1d) as a
reference substrate again, we initially carried out several
experiments with the bacterium previously grown in the
presence of DEPA and suspended (with a reference
absorbance A650¼1.0) in 100 mL of the usual phosphate
buffer. The cells contained in this suspension (resuspended
in distilled water) gave ca. 90 mg of dry residue (cell dry
weight, CDW) after evaporation at 120 8C.
The bacterial hydration of nicotinoylacetonitrile, 1q, is also
very slow. With the standard bacterial absorbance
(A650¼3.0), only ca. 10% of nicotinoylacetamide, 2q, was
obtained after 6 h; most of the crude product was 1q,
together with a significant amount of gummy material. With
the purpose of speeding up the bacterial hydration and
decreasing the formation of impurities, we decided to
increase the bacterial absorbance to an A650 value of 6.7.
Thus, although some gummy material still appeared, we
were able to isolate 60% of 2q after 9.5 h.
Starting from 100 mg (0.69 mmol) of 1d, a very good yield
of 2d was obtained after 2 h, just after 1d disappeared
(Table 4, entry 1). Doubling the amount of substrate, it was
necessary to wait for 8 h until disappearance of the substrate
(entry 2); moreover, the ketoamide yield strongly fell due to
the residual amidase activity. Serious problems arose
starting from 250 mg (1.72 mmol) of 1d (entry 3). TLC
monitoring showed that, at the most, only half of the
substrate was transformed, approximately. A remarkable
amount of b-ketoamide 2d was formed in the beginning, but
it was hydrolyzed to the corresponding acid and disappeared
after 2 days. Thus, it seems that an excess of 1d inhibits the
Of a-substituted b-ketonitriles 1s and 1t, the latter offered
no special problems, but the former failed as substrate under
the standard conditions. In fact, after 24 h, TLC does not
show noteworthy transformation of 1s; after several days,
some 2s appeared (1H NMR), but approximately half of the
mass was lost and, moreover, the crude product was a
complex mixture. After other failed attempts, we observed
that the use of a somewhat more complex medium for
growing the bacterium (medium N28c instead of the standard
medium M, see Section 3.3) led to 79% yield of 2s after a
23.5 h biotransformation. Despite this long reaction time, no
b-ketoacid 3s was observed. Moreover, the yield of 2s was
very similar (76%) after growing the bacterium in the
absence of DEPA. These facts seem to prove the existence
Table 4. Biotransformations of variable amounts of benzoylacetonitrile
(1d) with R. rhodochrousa (A650¼1.0)
Entry Mass of 1db (mg (mmol)) Reaction time (h) Yieldc of 2d (%)
1
2
3
100 (0.69)
200 (1.38)
250 (1.72)
1.5
8
48d
91
62
—
d
a
Grown in the presence of 15 mM DEPA and suspended in 100 mL of
0.10 M potassium phosphate buffer, pH 8.0.
10 mL EtOH/mg 1d were added to the reaction mixture.
Determined by 1H NMR of the crude reaction product.
See text.
b
c
d

612
V. Gotor et al. / Tetrahedron 60 (2004) 607–618
nitrile hydratase. Consequently, it can be estimated that the
bacterial mass equivalent to ca. 100 mg of CDW does not
accept more than about 150 mg (ca. 1 mmol) of 1d. Our
standard preparative substrate/bacterial mass ratio is notably
lower in order to attain the best yields in the shortest
reaction times.
(4.12 mmol, 79%) of 2d, which was accompanied in the
crude reaction product by 0.27 mmol (5%) of unchanged
1d.33 From these data, a bioconversion efficiency of
1.16 mmol of 1d/100 mg of CDW can be deduced. As a
corollary of these experiments, the nitrile hydratase and the
amidase do not seem to be appreciably inhibited by a
moderate excess of b-ketoamide 2d.
With the data of the previous paragraph in mind, we
resolved to carry out a series of larger scale experiments, all
of them in such conditions that the inhibition by substrate of
the nitrile hydratase was avoided. First we decided to
recycle the cells in our standard conditions. After each
reaction, cells were harvested by centrifugation, resus-
pended in fresh potassium phosphate buffer, and then
exposed to a new amount of substrate 1d. DEPA was
present, as usual, during the growth of the bacterium, but not
during the biotransformation. Using this methodology,
yields of 2d progressively diminished with the number of
recyclings, so that in the fourth reaction only a very little
amount of the substrate was transformed.
From the point of view of the bioconversion efficiency, the
recycling method (in the presence of DEPA) is better than
that of feeding in portions, but also is more tedious,
consumes more DEPA and requires more expensive column
chromatography. Therefore, the selection of one or another
will depend on the circumstances of each case.
Finally, a further non-standard experiment is noteworthy.
Using a 50 mL suspension of R. rhodochrous (A650¼11) in
the usual potassium phosphate buffer plus 5 mM DEPA,
454 mg (3.13 mmol) of 1d were converted in 485 mg
(2.97 mmol, 92% yield) of b-ketoamide 2d after 2 h. This
result is not surprising, since the substrate/bacterial mass
ratio is ca. one third lower than its previously estimated
maximum value (1 mmol of substrate/100 mg CDW).
However, this experiment represents an alternative, fast
and efficient way of obtaining notable amounts of
b-ketoamide from concentrated suspensions of the bacterium.
In view of the previous unsatisfactory results, we repeated
the recycling experiments introducing a change, namely,
adding 5 mM DEPA during the phase of reaction. The
results are collected in Table 5. The four first ca. 1 h-
reactions gave an excellent average yield of 93% of 2d as
the sole product. The fifth reaction is slightly worse, but,
even so, it produces a very good yield. Altogether, almost
750 mg of 1d were transformed by the cellular mass
equivalent to 265 mg of CDW, i.e. ca. 1.95 mmol of 1d/
100 mg of CDW, which roughly doubles the previously
estimated bioconversion efficiency for simple processes.
2.7. Keto–enol tautomerism in b-ketoamides 2a–t
Keto–enol tautomerism in b-ketoesters and b-diketones is a
topic extensively studied from several points of view and by
means of a variety of experimental methods.34,35 This
phenomenon, however, has been much less studied for
b-ketoamides34a,36 (Scheme 2). With a few exceptions,37 it
is usual to describe them only as keto forms,10 despite some
of them, in our hands, have shown to exist as a tautomeric
mixture where the enol form was the major tautomer.38 The
apparent carelessness in this matter contrasts with the recent
recognition of the b-ketoamides’ enol forms as important
structural elements of tetracyclines and piroxicams.39
Another obvious method of avoiding the substrate inhibition
would consist of the addition of the substrate in successive
portions on a same bacterial suspension. To such an aim, we
established three subroutines, according to the presence of
DEPA only during the growth phase, only during the
reaction phase, or during both phases. As expected, since
this methodology implies a long time contact between the
b-ketoamide 2d and the amidase, the third subroutine was
the best by far. To a bacterial suspension (A650¼4.0) in the
usual phosphate buffer (100 mL) plus 5 mM DEPA, seven
average 108 mg portions of 1d were added in regular
intervals of time during 8 h. The overall 755 mg
(5.21 mmol) of 1d were transformed into 672 mg
In view of this situation, we show in Table 6 the percentages
of enol forms in the keto–enol equilibria of b-ketoamides
2a–t (Scheme 2). Data were obtained by 1H NMR
spectroscopy [in CDCl3 for all of them and in CD3OD
and/or (CD3)2CO for some of them], at 0.01–0.20 molar
concentrations, after several days, in order to ensure that the
equilibria were reached. As a general rule, however, no time
dependence was observed in CDCl3, whereas an almost
imperceptible and a strong drop of the enol content in
(CD3)2CO and in CD3OD, respectively, were noticed as the
time elapses.
Table 5. Biotransformationsa (in the presence of DEPA) of benzoylace-
tonitrile (1d) with R. rhodochrous’b recyclingc (A650¼3.0)
Reaction no.
Reaction time (min)
Yieldd of 2d (%)
1
2
3
4
5
55
50
55
60
93
93
94
92
85
For b-ketoesters and b-diketones it is well established that
the presence of electron-withdrawing substituents raises the
270
a
150 mg (1.03 mmol) of 1d in 1.5 mL of EtOH were employed in each
experiment.
b
c
Grown in the presence of 15 mM DEPA and suspended in 100 mL of
0.10 M potassium phosphate buffer (pH 8.0) plus 5 mM DEPA.
After each reaction, the bacterial cake was resuspended in 100 mL of
fresh 0.10 M potassium phosphate buffer (pH 8.0) plus 5 mM DEPA, and
more 1d/EtOH was added.
d
Determined by 1H NMR of the crude reaction product.
Scheme 2. Keto–enol tautomerism in b-ketoamides 2.

V. Gotor et al. / Tetrahedron 60 (2004) 607–618
613
Table 6. Percentagesa of enol tautomers for b-ketoamides 2b in three deuterated solventsc
Compound
R1
CDCl3
CD3OD
(CD3)2CO
Compound
R1
CDCl3
CD3OD
23
(CD3)2CO
2a
2b
2c
2d
2e
2f
2g
2h
2i
Pr
,3
5
11
12
9
2j
2k
2m
2n
2p
2q
2r
2s
4-Cl–C6H4
3,4-Cl2C6H3
2-Furyl
3-Furyl
2-Thienyl
3-Pyridyl
1-Methyl-indol-3-yl
Me (R2¼Ph)
Ph (R2¼Et)
17
30
,3
,4
,2
35
0
60
68
d
But
,40
1-(Methylpyrrol-2-yl)methyl
Ph
55
51
46
7
28
16
67
3-Me–C6H4
4-Me–C6H4
4-MeO–C6H4
3-F3C–C6H4
3-Cl–C6H4
8
13
6
42
28
,2
30
23
68
66
79
0
65
0
2t
a
Determined at room temperature by 1H NMR spectrometry after the equilibria were reached.
Except for 2s,t, R2¼H.
b
c
d
Solute concentrations: 0.01–0.20 mol/L.
This value clearly lies between 5 and 40%, although signal overlapping precludes its exact determination.
enol percentage,34g,h since electron withdrawal from
the enol ring enhances the electrostatic contribution to
the strength of the intramolecular hydrogen bond. Con-
versely, electron-releasing substituents decrease the enol
content.
a,g-Disubstituted 3-oxopropanamides 2s,t should be com-
mented independently. Alkyl substitution in the methylene
group of b-diketones highly favours the keto tautomer, as
shown by comparing diketones 5 and 6 and, especially 7 and
8 (Fig. 1). Although steric effects have been invoked to
explain this fact,34g the matter is far from being clear, since
similar steric hindrance are to be expected in the enol forms
of b-diketones 8 and 9, whose enol percentages differ as
dramatically as shown in Figure 1. In any case, a parallelism
between similarly substituted b-diketones 7 and 8 and
b-ketoamides 2d and 2t accounts for the exclusive existence
of the latter ketoamide as keto tautomer.42 On the contrary,
a-phenyl substituted b-ketoamide 2s preponderantly exists
as enol tautomer (79%), as it was to be expected after
knowing the case of 3-phenylpentane-2,4-dione 9.43
Such a tendency is clearly reflected in the enol percentages
of monosubstituted 3-oxopropanamides 2a–r when dis-
solved in deuteriochloroform (see Table 6). Since it is
usually assumed that phenyl group is a weakly withdrawing
one, the enol content of 2d is somewhat higher than those of
the alkyl substituted, weakly releasing 2a,b.40 If one
considers the aryl substituted 2d–k as a whole, it is evident
that those bearing one or two further electron-withdrawing
groups (2h–k) exhibit a higher enol percentage than the
‘unsubstituted’ 2d, and this percentage increases with
the withdrawing strength of the substituents (2h,k).41 The
opposite tendency is just observed for the aryl substituted
2e–g bearing meta- and para-substituents that release
electronic charge. The case of heterocyclic systems 2m–r
is also very illustrative. When the heterocycle is
p-excessive, i.e. releases electronic charge (2m–p,r), the
enol content is very low, especially for 2r, whose indole ring
is linked at its highly electron rich C-3 position. On the
contrary, the presence of a p-deficient, highly electron-
withdrawing pyridine ring (2q) dramatically increases the
percentage of enol tautomer.
Finally, let us consider the influence of solvent polarity on
the keto–enol equilibria of b-ketoamides 2. For b-diketones
and b-ketoesters, it is generally accepted that their
tautomeric equilibria are shifted towards their less polar
enol forms as the solvent polarity decreases.35 Our data for
b-ketoamides 2 (Table 6) concern three deuterated solvents,
whose relative polarities depend on the method employed
for evaluating them. For normal, non-deuterated solvents,
polarities decrease in the series MeOH.Me2CO.CHCl3
when measured from spectroscopic data,44 whereas they
differ as MeOH.CHCl3.Me2CO if deduced from equili-
bria data45 and from the solvent’s anion-solvating and
cation-solvating tendencies.46
Literature and Table 6 data show that the relationship
between solvent polarity and keto–enol equilibria is a very
complex problem. In fact, enol percentages deduced from
1H NMR for pentane-2,4-dione increase as in the former
series, i.e. CD3OD,(CD3)2CO,CDCl3,47 whereas for
ethyl acetoacetate they increase by the latter, namely,
CD3OD,CDCl3,(CD3)2CO.34e For their part, b-keto-
amides 2 also behaves in their own way. Thus, all
monosubstituted 3-oxopropanamides for which all three
data are collected in Table 6 (2b,f,j,n,q) show increasing
enol percentages in the series CDCl3,CD3ODp(CD3)2CO,
whereas the disubstituted 2s behaves as pentane-2,4-dione.
Thus, it seems that several factors, among which the high
number of possible hydrogen bonds in N-unsubstituted
b-ketoamides will probably play an important role, must be
considered in depth in order to clarify this matter, but this
study is beyond of the goals of this paper.
Figure 1. Enol percentages for several selected b-diketones (1H NMR,
CDCl3). (Compounds 5–8: Ref. 34g; Compound 9: Ref. 43).

614
V. Gotor et al. / Tetrahedron 60 (2004) 607–618
3. Experimental
same, fresh potassium phosphate buffer, adjusting the
absorbance of the bacterial suspension to the value of 3.0,
except for the preparation of 2q (A650¼6.7), 2r (A650¼14.0)
and 2s (A650¼11.6).
3.1. General
Thin-layer chromatography was performed on precoated
TLC plates of Merck silica gel 60F254, using potassium
permanganate as developing reagent. For column chroma-
tography, Merck silica gel 60 (particle size, 40–63 mm) was
used. Melting points were taken using a Gallenkamp or a
Stuart SMP3 apparatus and are uncorrected. Mass spectra
were recorded on a Hewlett–Packard 5897 A (electron
impact, 70 eV) and a Hewlett–Packard 1100 HPLC/MS
(electrospray) instruments. The C, H, N analyses were
performed on a Perkin–Elmer 2400 analyzer. Absorbances
(optical densities) were measured at 650 nm on a Perkin–
Elmer UV–Vis Lambda20 spectrometer. 1H and 13C NMR
spectra were obtained with a Bruker AC-200 spectrometer
3.4. Standard preparations of b-ketoamides 2
To the appropriate bacterial suspension (100 mL) as
described above, 0.60–0.80 mmol of the corresponding
b-ketonitrile 1 (0.25–0.35 mmol in the cases of 1i,k,r; vide
infra) and ca. 1 mL of ethanol were added. In the case of 1i,
5 mM DEPA (from a 1.0 M reserve solution in methanol)
was also incorporated. Incubation (rotary shaker, 200 rpm,
28 8C) was carried out until the substrate disappeared (TLC
monitoring, ethyl acetate as eluent; aliquots of 0.5 mL plus
one drop of aqueous 1 M HCl, extracted with ethyl acetate),
and then the cells were discarded by centrifugation
(5000 rpm, 3 min). The supernatant, a slightly basic liquid,
was continuously extracted with dichloromethane (8 h).
[When concentrated bacterial suspensions were used (in the
preparation of 2q,r,s), the bacterial cake was washed with
distilled water (20 mL), centrifuged again, and the super-
natant liquid combined with the previous one before
continuous extraction]. After drying the organic phase
with anhydrous Na2SO4, low pressure elimination of solvent
yielded the corresponding, essentially pure (except 2i,q)
b-ketoamide 2. Further purification was accomplished by
silica gel flash column chromatography [hexane–ethyl
acetate as eluent, except for 2q (see below)].
(200.13 MHz for H and 50.3 MHz for 13C), using the d
scale (ppm) for chemical shifts; calibration was made on the
corresponding solvent signal; 13C NMR spectra were edited
using DEPT techniques.
1
3.2. Starting materials 1
Substrates 1b,d–m,p,s are commercially available
(Aldrich, Avocado or Lancaster). Substrates 1a,c,n,q,r,t
were prepared by means of a modified previous method,48
namely, by reacting acetonitrile (propionitrile for 2t) with
sodium amide (ca. 230 8C), and then with the respective
R1-CO2Et (same temperature) until the ester disappeared.
In this way, the following b-ketoamides 2 were obtained. To
our knowledge, 2c,e–h,j,k,n,p,r are new compounds.
Complete NMR spectral data are included for all keto
tautomers, whereas only selected resonances are shown for
most enol tautomers. For signal assignments, Greek letters
are used for the aliphatic chains, and numerical locants for
the rings.
3.3. Cultures of R. rhodochrous IFO 15564
R. rhodochrous IFO 15564 was maintained at 4 8C on Petri
plates containing tryptone (10 g/L), yeast extract (5 g/L),
NaCl (10 g/L) and agar (15 g/L), with pH adjusted at 7.0.
Sub-culturing was performed every 3 months.
Except for the incubation of 1s, the bacterium was grown
in a culture medium M,13e whose components were
heat-sterilized (115 8C, 20 min) in two separate groups.
Group 1: glucose (15.0 g), yeast extract (1.0 g), KH2PO4
(0.50 g), K2HPO4 (0.50 g) and distilled water (950 mL),
with pH adjusted at 7.2; group 2: 1-caprolactam (5.0 g),
FeSO4·7H2O (0.50 g) and distilled water (50 mL). For the
incubation of 1s, the following medium N28c was employed:
heat-sterilized (115 8C, 20 min): glucose (15.0 g), peptone
(5.0 g), yeast extract (1.0 g), KH2PO4 (0.40 g), K2HPO4
(1.20 g), MgSO4·7H2O (0.50 g) and distilled water
(950 mL), with pH adjusted at 7.2; filter-sterilized: 1-capro-
lactam (1.0 g), FeSO4·7H2O (0.30 g) and distilled water
(50 mL).
3.4.1. 3-Oxohexanamide (2a). From 90 mg (0.80 mmol) of
1a/1.0 mL EtOH, 91 mg (0.69 mmol, 87%) of 2a were
1
obtained. Mp 80–81 8C (lit.,49 79–81 8C). H NMR (keto,
CDCl3): 0.87 (t, 3H1, J¼7.4 Hz), 1.56 (sext., 2Hd,
J¼7.4 Hz), 2.49 (t, 2Hg, J¼7.4 Hz), 3.37 (s, 2Ha), 6.37
(broad s, 1NH), 7.06 (broad s, 1NH). 13C NMR (keto,
CDCl3): 13.3 (C1), 16.6 (Cd), 45.4 (Cg), 48.7 (Ca), 168.5
(amide C), 206.3 (Cb). MS, m/z (%): 129 (Mþ, 35), 114 (29),
112 (21), 101 (60), 86 (81), 71 (88), 59 (100).
3.4.2. 4,4-Dimethyl-3-oxopentanamide (2b). From
100 mg (0.80 mmol) of 1b/1.0 mL EtOH, 103 mg
(0.72 mmol, 90%) of 2b were obtained. Mp 93–95 8C
(lit., 9550 and 83–847 8C). 1H NMR (keto, CDCl3): 1.15 (s,
9Hd), 3.48 (s, 2Ha), 5.59 (broad s, 1NH), 7.06 (broad s,
1NH); [enol, (CD3)2CO]: 5.12 (s, 1Ha), 14.7 (s, 1H, OH).
13C NMR (keto, CDCl3): 25.7 (Cd), 43.3 (Ca), 45.0 (Cg),
168.8 (amide C), 211.5 (Cb); (enol, CDCl3): 27.3 (Cd), 36.2
(Cg), 85.3 (Ca), 174.7 (amide C), 184.8 (Cb). MS, m/z (%):
143 (Mþ, 13), 115 (30), 111 (11), 86 (79), 59 (100), 57 (51).
A loop of solid culture of R. rhodochrous IFO 15564, from
an agar plate, was sowed on 200–500 mL of sterilized
medium M (medium N for the synthesis of 2s), sup-
plemented with 15 mM DEPA (from a 1.0 M reserve
solution in methanol). After growing (rotary shaker,
200 rpm, 28 8C) for 36–42 h (medium M; final A650¼2–
2.5) or for 60 h (medium N; final A650ø12.5), cells were
harvested by centrifugation (5000 rpm, 3 min), washed with
0.10 M KH2PO4–K2HPO4 buffer pH 8.0 (20 mL), again
collected by centrifugation and then resuspended in the
3.4.3. 4-(1-Methylpyrrol-2-yl)methyl-3-oxobutanamide
(2c). From 102 mg (0.63 mmol) of 1c/1.0 mL EtOH,
1
103 mg (0.57 mmol, 91%) of 2c (oil) were obtained. H
NMR (keto, CDCl3): 3.42 (s, 2Ha), 3.47 (s, 3H, N–Me),

V. Gotor et al. / Tetrahedron 60 (2004) 607–618
615
3.78 (s, 2Hg), 6.0–6.15 (m, 2H, H-3 and H-4), 6.25 (broad s,
1NH), 6.60 (t, 1H, H-5, J¼1.9 Hz), 6.76 (broad s, 1NH);
(enol, CDCl3): 4.60 (s, 1Ha), 13.9 (s, 1H, OH). 13C NMR
(keto, CDCl3): 33.7 (N–Me), 41.8 (Cg), 47.4 (Ca), 107.2
and 108.7 (C-3 and C-4), 122.9 (C-5), 123.7 (C-2), 168.1
(amide C), 202.8 (Cb); (enol, CDCl3): 89.5 (Ca), 174.0 and
175.3 (amide C and Cb). MS, m/z (%): 180 (Mþ, 58), 163
(7), 121 (20), 94 (100). Anal. (%) calcd for C9H12N2O2
(180.20): C, 59.99; H, 6.71; N, 15.55; found: C, 60.15; H,
6.58; N, 15.51.
3.4.8. 3-Oxo-3-[(3-trifluoromethyl)phenyl]propanamide
(2h). From 128 mg (0.60 mmol) of 1h/1.0 mL EtOH,
123 mg (0.53 mmol, 88%) of 2h were obtained. Mp 69–
71 8C. 1H NMR (keto, CDCl3): 4.00 (s, 2Ha), 5.93 (broad s,
1NH), 6.98 (broad s, 1NH), 7.65–7.8 (m, 1H, H-5), 7.85–
7.95 (m, 1H, H-4), 8.2–8.4 (m, 2H, H-2 and H-6); (enol,
CDCl3): 5.61 (s, 1Ha), 14.3 (broad s, 1H, OH). 13C NMR
(keto, CDCl3): 45.3 (Ca), 122.7–136.5 (several aromatic
and F3C peaks), 167.6 (amide C), 194.1 (Cb); (enol,
CDCl3): 88.2 (Ca), 169.4 (amide C), 173.9 (Cb). MS, m/z
(%): 254 (MþNa)þ, 232 (MþHþ). Anal. (%) calcd for
C10H8F3NO2 (231.17): C, 51.96; H, 3.49; N, 6.06; found: C,
51.70; H, 3.47; N, 5.78.
3.4.4. 3-Oxo-3-phenylpropanamide (2d). From 101 mg
(0.70 mmol) of 1d/1.0 mL EtOH, 106 mg (0.65 mmol,
93%) of 2d were obtained. Mp 111.5–113.5 8C (lit.,10h
1
110–111 8C). H NMR (keto, CDCl3): 3.98 (s, 2Ha), 5.56
3.4.9. 3-(3-Chlorophenyl)-3-oxopropanamide (2i). From
60 mg (0.33 mmol) of 1i/0.60 mL EtOH, 56 mg
(0.28 mmol, 85%) of 2i were obtained. Mp 136.5–138 8C
(lit.,10h 135–137 8C). 1H NMR (keto, CDCl3): 3.96 (s, 2Ha),
5.65 (broad s, 1NH), 7.02 (broad s, 1NH), 7.4–7.5 (m, 1H,
H-5), 7.55–7.65 (m, 1H, H-4), 7.85–7.92 (m, 1H, H-6),
7.96–8.01 (m, 1H, H-2); (enol, CDCl3): 5.55 (s, 1Ha), 14.2
(s, 1H, OH). 13C NMR (keto, CDCl3): 45.1 (Ca), 126.7
(C-6), 128.5 (C-2), 130.2 (C-5), 134.0 (C-4), 135.3 (C-3),
137.5 (C-1), 167.4 (amide C), 194.3 (Cb); [enol,
(CD3)2CO]: 89.9 (Ca), 169.2 (amide C), 175.7 (Cb). MS,
m/z (%): 222 (Mþ2þNa)þ, 220 (MþNa)þ. Anal. (%) calcd
for C9H8ClNO2 (197.62): C, 54.70; H, 4.08; N, 7.09; found:
C, 54.73; H, 3.95; N, 6.93.
(broad s, 1NH), 7.13 (broad s, 1NH), 7.4–7.7 (m, 3H, H-3,
H-4 and H-5), 7.95–8.1 (m, 2H, H-2 and H-6); (enol,
CDCl3): 5.54 (s, 1Ha), 14.2 (s, 1H, OH). 13C NMR (keto,
CDCl3): 45.8 (Ca), 129.2 and 129.5 (C-2, C-3, C-5 and C-6),
134.8 (C-4), 136.7 (C-1), 169.1 (amide C), 196.2 (Cb);
(enol, CDCl3): 88.1 (Ca), 171.7 (amide C), 175.2 (Cb). MS,
m/z (%): 163 (Mþ, 13), 146 (9), 105 (100), 77 (46).
3.4.5. 3-(3-Methylphenyl)-3-oxopropanamide (2e). From
112 mg (0.70 mmol) of 1e/1.0 mL EtOH, 116 mg
(0.66 mmol, 93%) of 2e were obtained. Mp 98.5–99.5 8C.
1H NMR (keto, CDCl3): 2.42 (s, 3H, Me), 3.96 (s, 2Ha),
5.78 (broad s, 1NH), 7.16 (broad s, 1NH), 7.25–7.5 (m, 2H,
H-4 and H-5), 7.75–7.9 (m, 2H, H-2 and H-6); (enol,
CDCl3): 5.55 (s, 1Ha), 14.2 (s, 1H, OH). 13C NMR (keto,
CDCl3): 21.2 (Me), 45.0 (Ca), 125.7 (C-6), 128.6 and 128.9
(C-2 and C-5), 134.8 (C-4), 136.0 (C-1), 138.6 (C-3), 168.4
(amide C), 195.7 (Cb); (enol, CDCl3): 87.3 (Ca). MS, m/z
(%): 177 (Mþ, 57), 160 (14), 120 (16), 119 (100), 91 (75).
Anal. (%) calcd for C10H11NO2 (177.20): C, 67.78; H, 6.26;
N, 7.90; found: C, 67.90; H, 6.10; N, 7.69.
3.4.10. 3-(4-Chlorophenyl)-3-oxopropanamide (2j). From
128 mg (0.71 mmol) of 1j/1.0 mL EtOH, 101 mg
(0.51 mmol, 72%) of 2j were obtained. Mp 143–145 8C
(dec.). 1H NMR (keto, CDCl3): 3.94 (s, 2Ha), 5.67 (broad s,
1NH), 6.98 (broad s, 1NH), 7.48 (d, 2H, H-3 and H-5,
J¼8.6 Hz), 7.94 (d, 2H, H-2 and H-6, J¼8.6 Hz); (enol,
CDCl3): 5.53 (s, 1Ha), 7.38 (d, 2H, H-3 and H-5, J¼8.6 Hz),
7.69 (d, 2H, H-2 and H-6, J¼8.6 Hz), 14.3 (broad s, 1H,
OH). 13C NMR (keto, CDCl3): 45.0 (Ca), 129.2 (C-3 and
C-5), 129.9 (C-2 and C-6), 134.3 (C-1), 140.8 (C-4), 167.6
(amide C), 194.3 (Cb); [enol, (CD3)2CO]: 88.6 (Ca), 168.9
(amide C), 175.0 (Cb). MS, m/z (%): 199 [(Mþ2)þ, 4], 197
(Mþ, 14), 141 (31), 139 (100), 111 (37). Anal. (%) calcd for
C9H8ClNO2 (197.62): C, 54.70; H, 4.08; N, 7.09; found:
C,54.45; H, 4.17; N, 6.87.
3.4.6. 3-(4-Methylphenyl)-3-oxopropanamide (2f). From
111 mg (0.70 mmol) of 1f/1.0 mL EtOH, 109 mg
(0.61 mmol, 88%) of 2f were obtained. Mp 117–118 8C.
1H NMR (keto, CDCl3): 2.42 (s, 3H, Me), 3.94 (s, 2Ha),
5.75 (broad s, 1NH), 7.16 (broad s, 1NH), 7.29 (d, 2H, H-3
and H-5, J¼8.0 Hz), 7.89 (d, 2H, H-2 and H-6, J¼8.0 Hz);
(enol, CDCl3): 5.52 (s, 1Ha), 14.2 (s, 1H, OH). 13C NMR
(keto, CDCl3): 21.6 (Me), 44.9 (Ca), 128.6 and 129.5 (C-2,
C-3, C-5 and C-6), 133.5 (C-1), 145.1 (C-4), 168.5 (amide
C), 195.1 (Cb); (enol, CDCl3): 86.6 (Ca). MS, m/z: 200.1
(MþNa)þ. Anal. (%) calcd for C10H11NO2 (177.20): C,
67.78; H, 6.26; N, 7.90; found: C, 67.52; H, 6.26; N, 7.70.
3.4.11. 3-(3,4-Dichlorophenyl)-3-oxopropanamide (2k).
From 74 mg (0.35 mmol) of 1k/0.74 mL EtOH, 57 mg
(0.25 mmol, 71%) of 2k were obtained. Mp 114–115.5 8C.
1H NMR (keto, CDCl3): 3.94 (s, 2Ha), 5.83 (broad s, 1NH),
6.93 (broad s, 1NH), 7.59 (d,1H, H-5, J¼8.6 Hz), 7.83 (dd,
1H, H-6, J¼2.3, 8.6 Hz), 8.09 (d, 1H, H-2, J¼2.3 Hz);
(enol, CDCl3): 5.53 (s, 1Ha), 14.2 (broad s, 1H, OH). 13C
NMR (keto, CDCl3): 45.2 (Ca), 127.6–138.9 (six aromatic
signals), 167.2 (amide C), 193.2 (Cb); (enol, CDCl3): 88.0
(Ca), 168.9 (amide C), 173.7 (Cb). MS, m/z (%): 272
(Mþ2þK)þ, 270 (MþK)þ, 256 (Mþ2þNa)þ, 254
(MþNa)þ, 232 (MþH)þ. Anal. (%) calcd for
C9H7Cl2NO2 (232.06): C, 46.58; H, 3.04; N, 6.04; found:
C, 46.80; H, 2.89; N, 5.91.
3.4.7. 3-(4-Methoxyphenyl)-3-oxopropanamide (2g).
From 112 mg (0.64 mmol) of 1g/1.0 mL EtOH, 116 mg
(0.60 mmol, 94%) of 2g were obtained. Mp 145–146.5 8C.
1H NMR (keto, CDCl3): 3.89 (s, 3H, Me), 3.92 (s, 2Ha),
5.56 (broad s, 1NH), 6.96 (d, 2H, H-3 and H-5, J¼9.2 Hz),
7.18 (broad s, 1NH), 7.98 (d, 2H, H-2 and H-6, J¼9.2 Hz);
(enol, CD3OD): 5.67 (s, 1Ha). 13C NMR (keto, CDCl3):
44.6 (Ca), 55.5 (Me), 114.0 (C-3 and C-5), 129.0 (C-1),
131.0 (C-2 and C-6), 164.3 (C-4), 168.3 (amide C), 194.0
(Cb). MS, m/z (%): 193 (Mþ, 18), 149 (22), 135 (100), 97
(33), 91 (30), 83 (38), 77 (41), 71 (60), 69 (55), 57 (98), 55
(56). Anal. (%) calcd for C10H11NO3 (193.20): C, 62.17; H,
5.74; N, 7.25; found: C, 62.34; H, 5.66; N, 7.14.
3.4.12. 3-(2-Furyl)-3-oxopropanamide (2m). From 103 mg
(0.76 mmol) of 1m/1.0 mL EtOH, 102 mg (0.66 mmol,
87%) of 2m were obtained. Mp 142.5–144 8C (lit.,51

616
V. Gotor et al. / Tetrahedron 60 (2004) 607–618
159 8C). 1H NMR (keto, CDCl3): 3.85 (s, 2Ha), 5.60 (broad
s, 1NH), 7.10 (broad s, 1NH), 6.61 (dd, 1H, H-4, J¼1.6,
3.7 Hz), 7.35 (d, 1H, H-3, J¼3.7 Hz), 7.67 (d, 1H, H-5,
J¼1.6 Hz); (enol, CDCl3): 5.52 (s, 1Ha). 13C NMR (keto,
CDCl3): 44.5 (Ca), 112.9 (C-4), 119.6 (C-3), 147.9 (C-5),
151.8 (C-2), 167.4 (amide C), 183.7 (Cb). MS, m/z (%): 153
(Mþ, 67), 110 (28), 97 (19), 95 (100), 71 (23), 57 (31).
3.4.16. 3-(1-Methylindol-3-yl)-3-oxopropanamide (2r).
From 50 mg (0.25 mmol) of 1r/1.0 mL EtOH, 41 mg
(0.19 mmol, 75%) of 2r were obtained. Mp 179.5–
1
181.5 8C. H NMR (keto, CDCl3): 3.81 (s, 2Ha), 3.88 (s,
1H, N–Me), 5.48 (broad s, 1NH), 7.2–7.5 (m, 3HArþ1NH),
7.87 (s, 1H, H-2), 8.3–8.45 (m, 1H, H-4). 13C NMR (keto,
CDCl3): 33.9 (N–Me), 46.7 (Ca), 110.0 (C-7), 116.1 (C-3),
122.5, 123.3 and 124.0 (C-4, C-5 and C-6), 126.3 (C-3a),
137.2 (C-2), 126.3 (C-7a), 169.1 (amide C), 189.4 (Cb). MS,
m/z (%): 216 (Mþ, 26), 158 (100), 130 (12). Anal. (%) calcd
for C12H12N2O2 (216.24): C, 66.65; H, 5.59; N, 12.96;
found: C, 66.39; H, 5.55; N, 12.79.
3.4.13. 3-(3-Furyl)-3-oxopropanamide (2n). From 108 mg
(0.80 mmol) of 1n/1.0 mL EtOH, 106 mg (0.70 mmol,
1
87%) of 2n were obtained. Mp 105.5–107 8C. H NMR
(keto, CDCl3): 3.75 (s, 2Ha), 5.61 (broad s, 1NH), 6.80 (dd,
1H, H-4, J¼0.8, 2.0 Hz), 7.10 (broad s, 1NH), 7.48 (t, 1H,
H-5, J¼1.6 Hz), 8.15 (broad s, 1H, H-2); [enol, (CD3)2CO]:
5.54 (s, 1Ha), 6.65 (d, 1H, H-4, J¼1.0 Hz), 7.61 (t, 1H, H-5,
J¼1.6 Hz), 7.96 (broad s, 1H, H-2). 13C NMR (keto,
CD3OD): 109.3 (C-4), 128.8 (C-3), 146.1 (C-5), 150.8
(C-2), 171.7 (amide C), 190.5 (Cb); Ca signal is overlapped
with the solvent signal. MS, m/z (%): 153 (Mþ, 14), 111
(18), 97 (14), 95 (100), 71 (14), 57 (17). Anal. (%) calcd for
C7H7NO3 (153.14): C, 54.90; H, 4.61; N, 9.15; found: C,
54.64; H, 4.78; N, 8.97.
3.4.17. 3-Oxo-2-phenylbutanamide (2s). From 101 mg
(0.64 mmol) of 1s/1.0 mL EtOH, 89 mg (0.50 mmol, 79%)
of 2s were obtained. Mp 122.5–124.5 8C (lit.,10h 124–
1
126 8C). H NMR (keto, CD3OD):52 2.22 (s, 3H, Cg), 4.81
(s, 1Ha), 7.3–7.5 (m, 5HAr); (enol, CDCl3): 1.78 (s, 3H, Cg),
5.07 (broad s, 1NH), 5.30 (broad s, 1NH), 7.2–7.5 (m,
5HAr), 14.6 (s, 1H, OH). 13C NMR (keto, CDCl3): 29.8 (Cg),
65.0 (Ca), 170.6 (amide C), 204.8 (Cb); aromatic carbon
nuclei masked by the aromatic signals due to the major
(79%) enol tautomer; (enol, CDCl3): 19.7 (Cg), 104.0 (Ca),
129.1, 131.2, 133.5 and 135.5 (aromatic carbon atoms),
172.4 (amide C), 174.6 (Cb). MS, m/z (%): 177 (Mþ, 22),
160 (56), 135 (38), 118 (100), 105 (18), 90 (39), 77 (24).
3.4.14. 3-Oxo-3-(2-thienyl)propanamide (2p). From
100 mg (0.66 mmol) of 1p/1.0 mL EtOH, 100 mg
(0.59 mmol, 89%) of 2p were obtained. Mp 125.5–
127 8C. 1H NMR (keto, CDCl3): 3.92 (s, 2Ha), 5.57
(broad s, 1NH), 7.18 (dd, 1H, H-4, J¼3.9, 5.1 Hz), 7.20
(broad s, 1NH), 7.76 (dd, 1H, H-5 or H-3, J¼1.2, 5.1 Hz),
7.83 (dd, 1H, H-3 or H-5, J¼1.2, 3.9 Hz); [enol, (CD3)2CO]:
5.72 (s, 1Ha), 7.13 (dd, 1H, H-4, J¼3.9, 4.9 Hz), 7.55 (m,
1H, H-5 or H-3), 7.63 (broad d, 1H, H-3 or H-5, J¼4.9 Hz).
13C NMR [keto, (CD3)2CO]: 48.1 (Ca), 129.6 (C-4), 135.0
and 135.9 (C-5 and C-3), 145.3 (C-2), 168.7 (amide C),
188.4 (Cb); [enol, (CD3)2CO]: 88.0 (Ca). MS, m/z (%): 169
(Mþ, 27), 111 (100). Anal. (%) calcd for C7H7NO2S
(169.20): C, 49.69; H, 4.17; N, 8.28; found: C, 49.78; H,
4.12; N, 8.30.
3.4.18. 2-Benzoylbutanamide (2t). From 116 mg
(0.67 mmol) of 1t/1.0 mL EtOH, 99 mg (0.52 mmol, 77%)
of 2t were obtained. Mp 154–156 8C (lit.,53 153–153.5 8C).
1H NMR (keto, CDCl3): 1.00 (t, 3Hd, J¼7.4 Hz), 1.9–2.04
(m, 2Hg), 4.29 (t, 1Ha, J¼7.2 Hz), 5.46 (broad s, 1NH), 6.59
(broad s, 1NH), 7.4–7.7 (m, 3H, H-3, H-4 and H-5), 8.02 (d,
2H, H-2 and H-6, J¼7.0 Hz). 13C NMR (keto, CD3OD):
13.4 (Cd), 25.2 (Cg), 58.8 (Ca), 130.5 and 130.8 (C-2, C-3,
C-5 and C-6), 135.6 (C-4), 139.0 (C-1), 175.7 (amide C),
199.0 (Cb). MS, m/z (%): 191 (Mþ, 2), 163 (5), 122 (13),
105 (100), 77 (48).
3.4.15. 3-Oxo-3-(3-pyridyl)propanamide (2q). Due to the
special acid–base properties of this product, the work-up
was slightly different from the standard protocol. After
the bioconversion [103 mg (0.71 mmol) of 1p/1.0 mL
EtOH], the supernatant liquid (ca. 100 mL) was concen-
trated in vacuo until ca. 10 mL, and then extracted with
ethyl acetate–methanol 10:1 (5£10 mL), without wash-
ing. Usual drying and elimination of solvents led to a
85:15 crude mixture of 2q: 3-acetylpyridine (4q), finally
purified by flash column chromatography (eluent, chloro-
form–hexane–diethyl ether–methanol 3:1:1:0.7), to yield
69 mg (0.42 mmol, 60%) of 2q. Mp 110.5–112 8C (lit.,34a
3.5. Determination of keto–enol ratio in b-ketoamides 2
For each compound and solvent, comparison of integral
values (1H NMR) was made between those cleaner, non-
overlapped signals corresponding to both tautomers. Thus,
the most useful signals to this end with samples in CDCl3
were those of Ha protons (2a–e,h,k,m,p), followed by H-2
and H-6 protons (ortho to keto or enol group: 2d,f,g,j) and
H-2 protons (2i,n); for 2s, the signals due to Hg protons were
selected, whereas an average of integral values of H-2, H-4
and H-6 protons was taken for 2q.
1
102–104 8C). H NMR (keto, CDCl3): 4.00 (s, 2Ha), 5.64
Using CD3OD as solvent, Ha signals disappeared as a result
of H/D interchange, in such a way that the determination
was impossible in several cases (for instance, 2d,h).
However, keto–enol percentages were easily deduced
from all aromatic or heterocyclic protons (2f,j,n,p), from
tert-butyl protons (2b), from Hg protons (2s) and from H-2,
H-4 and H-6 protons (2q).
(broad s, 1NH), 6.94 (broad s, 1NH), 7.45 (dd, 1H, H-5,
J¼4.7, 8.2 Hz), 8.29 (dt, 1H, H-4, J¼8.2, 2.0 Hz), 8.84
(dd, 1H, H-6, J¼1.6, 4.7 Hz), 9.21 (d, 1H, H-2,
J¼2.3 Hz); (enol, CDCl3): 5.60 (s, 1Ha), 7.37 (dd, 1H,
H-5, J¼4.7, 8.0 Hz), 8.06 (dt, 1H, H-4, J¼7.8, 2.0 Hz),
8.66 (dd, 1H, H-6, J¼1.6, 4.7 Hz), 8.96 (d, 1H, H-2,
J¼2.0 Hz), 14.2 (s, 1H, OH). 13C NMR (keto, CD3OD):
125.4 (C-5), 133.6 (C-3), 137.9 (C-4), 150.4 (C-6), 154.2
(C-2), 171.7 (amide C), 194.8 (Cb); Ca signal is
overlapped with the solvent signal. MS, m/z (%): 164
(Mþ, 52), 147 (91), 106 (100), 78 (66), 51 (20).
b-Ketoamides 2 showed a very slow Ha/D interchange
when dissolved in (CD3)2CO (except in the case of 2f). For
such a reason, Ha protons could be selected for a freshly
prepared sample of 2b, since the enol percentages remain

V. Gotor et al. / Tetrahedron 60 (2004) 607–618
617
G.-Q. Tetrahedron Lett. 2000, 41, 6501–6505. (c) Masutomo,
S.-i.; Inoue, A.; Kumagai, K.; Murai, R.; Mitsuda, S. Biosci.
Biotechnol. Biochem. 1995, 59, 720–722. (d) Gilligan, T.;
Yamada, H.; Nagasawa, T. Appl. Microbiol. Biotechnol. 1993,
39, 720–725. (e) Kakeya, H.; Sakai, N.; Sugai, T.; Ohta, H.
Tetrahedron Lett. 1991, 32, 1343–1346. (f) Bianchi, D.;
Bosetti, A.; Cesti, P.; Franzosi, G.; Spezia, S. Biotechnol. Lett.
1991, 13, 241–244.
almost invariable with time in this solvent. H-2 and H-6
protons were useful for measurements in 2d–f,j. The signals
selected for the remaining four compounds were as follows:
H-2 (2n); H-3 and H-5 (2p); H-2, H-4 and H-6 (2q); Hg (2s).
Acknowledgements
14. Meth-Cohn, O.; Wang, M.-X. J. Chem. Soc., Perkin Trans. 1
1997, 1099–1104.
Financial supports of this work by Principado de Asturias
(Project GE-EXP01-03) and MCYT (Project PPQ-2001-
2683) are gratefully acknowledged.
15. Yokoyama, M.; Nakatsuka, Y.; Sugai, T.; Ohta, H. Biosci.
Biotechnol. Biochem. 1996, 60, 1540–1542.
16. Sugai, T.; Yokoyama, M.; Takahiro, Y.; Ohta, H. Chem. Lett.
1997, 797–798.
17. (a) Bauer, R.; Knackmuss, H. J.; Stolz, A. Appl. Microbiol.
Biotechnol. 1998, 49, 89–95. (b) Bauer, R.; Hirrlinger, B.;
Layh, N.; Stolz, A.; Knackmuss, H. J. Appl. Microbiol.
Biotechnol. 1994, 42, 1–7.
References and notes
1. (a) Iida, T.; Hori, K.; Nomura, K.; Yoshii, E. Heterocycles
1994, 38, 1839–1844. (b) DeShong, P.; Lowmaster, N. E.;
Baralt, O. J. Org. Chem. 1983, 48, 1149–1150. (c) Boeckman,
R. K., Jr.; Thomas, A. J. J. Org. Chem. 1982, 47, 2823–2824.
2. Boeckman, R. K., Jr.; Starrett, J. E., Jr.; Nickell, D. G.; Sum,
P.-E. J. Am. Chem. Soc. 1986, 108, 5549–5559.
3. Amaresh, R. R.; Perumal, P. T. Tetrahedron 1999, 55,
8083–8094.
´
18. (a) Martinkova, L.; Stolz, A.; Knackmuss, H. J. Biotechnol.
´
¨
Lett. 1996, 18, 1073–1076. (b) Layh, N.; Stolz, A.; Bohme, J.;
Effenberger, F.; Knackmuss, H.-J. J. Biotechnol. 1994, 33,
175–182.
19. Eliel, E. L.; Freeman, J. P. Organic Syntheses; Wiley: New
York, 1963; Collect. Vol. IV. pp 58–62.
20. Basu, M. K.; Luo, F.-T. Tetrahedron Lett. 1998, 39,
3005–3006.
4. Cossy, J.; Belotti, D.; Leblanc, C. J. Org. Chem. 1993, 58,
2351–2354.
21. Cacchi, S.; Misiti, D.; La Torre, F. Synthesis 1980, 243–244.
22. Hall, J. H.; Gisler, M. J. Org. Chem. 1976, 41, 3769–3770.
23. Merchant, K. J. Tetrahedron Lett. 2000, 12, 3747–3749.
24. For example: (a) Linardi, V. R.; Dias, J. C. T.; Rosa, C. A.
FEMS Microbiol. Lett. 1996, 144, 67–71. (b) Harper, D. B.
Biochem. J. 1977, 165, 309–319.
5. (a) Cossy, J.; Belotti, D.; Cuong, N. K.; Chassagnard, C.
Tetrahedron 1993, 49, 7691–7700. (b) Cossy, J.; Leblanc, C.
Tetrahedron Lett. 1989, 30, 4531–4534.
6. Sakaki, J.-i.; Kobayashi, S.; Sato, M.; Kaneko, C. Chem.
Pharm. Bull. 1989, 37, 2952–2960.
7. Matzen, L.; Engesgaard, B. E.; Didriksen, M.; Frølund, B.;
Krogsgaard-Larsen, P.; Jarszewski, J. W. J. Med. Chem. 1997,
40, 520–527.
25. This strain was previously classified as Brevibacterium sp.
See: Briand, D.; Dubreucq, E.; Perrier, V.; Grimaud, J.; Galzy,
P. Microbios 1994, 78, 205–214.
8. Hilgenkamp, R.; Zercher, C. K. Tetrahedron 2001, 57,
8793–8800.
´
9. (a) Quiros, M.; Rebolledo, F.; Gotor, V. Tetrahedron:
26. (a) Osprian, I.; Jarret, C.; Strauss, U.; Kroutil, W.; Orru, R. V.
A.; Felfer, U.; Willets, A. J.; Faber, K. J. Mol. Catal. B:
Enzym. 1999, 6, 555–560. (b) Kerridge, A.; Parrat, J. S.;
Roberts, S. M.; Theil, F.; Turner, N. J.; Willets, A. J. Bioorg.
Med. Chem. 1994, 2, 447–455.
´
Asymmetry 1999, 10, 473–486. (b) Quiros, M.; Rebolledo,
F.; Liz, R.; Gotor, V. Tetrahedron: Asymmetry 1997, 8,
3035–3038.
27. To, F.; Mizunashi, W.; Endo, R. Jpn. Patent 04,365,491, 1992;
Chem. Abstr. 1993, 118, 146302b..
10. Preparation of b-ketoamides: (a) Kashima, C.; Fukuchi, I.;
Takahashi, K.; Hosomi, A. Tetrahedron 1996, 52,
10335–10346. (b) Cossy, J.; Belotti, D.; Bouzide, A.;
Thellend, A. Bull. Soc. Chim. Fr. 1994, 131, 723–729, and
references therein. (c) Pak, C. S.; Yang, H. C.; Choi, E. B.
Synthesis 1992, 1213–1214. (d) Cossy, J.; Thellend, A.
Synthesis 1989, 753–755. (e) Cossy, J.; Belotti, D.; Thellend,
A.; Pete, J. P. Synthesis 1988, 720–721, and references
therein. (f) Sato, M.; Ogasawara, H.; Komatsu, S.; Kato, T.
Chem. Pharm. Bull. 1984, 32, 3848–3856. (g) Colau, R.; Viel,
C. Bull. Soc. Chim. Fr. 1979, Vol. II, 362–365. (h) Hauser,
C. R.; Eby, C. J. J. Am. Chem. Soc. 1957, 79, 725–727.
11. Yamada, H.; Kobayashi, M. Biosci. Biotechnol. Biochem.
1996, 60, 1391–1400.
28. (a) Matoishi, K.; Sano, A.; Imai, N.; Yamazaki, T.;
Yokoyama, M.; Sugai, T.; Ohta, H. Tetrahedron: Asymmetry
1998, 9, 1097–1102. (b) Ohta, H. Chimia 1996, 50, 434–436.
(c) Yokoyama, M.; Imai, N.; Sugai, T.; Ohta, H. J. Mol. Catal.
B: Enzym. 1996, 1, 135–141. (d) Sugai, T.; Katoh, O.; Ohta,
H. Tetrahedron 1995, 51, 11987–11998.
29. Tail-end spots of lower Rf, especially when R. rhodochrous
was grown in the absence of DEPA. In no TLC plate was
ketoacid 3 observed before ketoamide 2 appeared, which is a
confirmation of the lack of nitrilase activity of this bacterium.
30. Although biotransformations were carried out at pH 8, it is
known that b-ketocarboxylic acids experience decarboxyla-
tion under slightly basic conditions: March, J. Advanced
Organic Chemistry. 4th ed. Wiley: New York, 1992, p. 629.
31. Ketone 4r was isolated from the supernatant after extracting
2r at pH 8 and further continuous extraction (40 8C, 8 h) at pH
6.
12. Hann, E. C.; Eisenberg, A.; Fager, S. K.; Perkins, N. E.;
Gallagher, F. G.; Cooper, S. M.; Gavagan, J. E.; Stieglitz, B.;
Hennessey, S. M.; DiCosimo, R. Bioorg. Med. Chem. 1999, 7,
2239–2245.
13. It is usual, however, to adjust the biotransformation time of
some racemic 2-substituted nitriles (conversion values of ca.
0.5) in order to obtain both amides and acids with high ee. See,
for example: (a) Wang, M.-X.; Zhao, S.-M. Tetrahedron:
Asymmetry 2002, 13, 1695–1702. (b) Wang, M.-X.; Feng,
32. When grown in this new culture medium N, R. rhodochrous
.
has shown to develop amidase activity28c
33. We assume that the remaining 16% includes the formed
b-ketoacid 3d and small losses during the work-up.

618
V. Gotor et al. / Tetrahedron 60 (2004) 607–618
34. See for example: (a) Bunting, J. W.; Kanter, J. P. J. Am. Chem.
Soc. 1993, 115, 11705–11715. (b) Yamasaki, K.; Kajimoto,
O. Chem. Phys. Lett. 1990, 172, 271–274. (c) Kallury, K. R.;
Krull, U. J.; Thompson, M. J. Org. Chem. 1988, 53,
1320–1322. (d) Zheglova, D. Kh.; Aleksiev, A. A.; Kmetska,
G. G.; Petrov, G. N.; Kol’tsov, A. I. J. Org. Chem. USSR 1988,
24, 1641–1643. (e) Mills, S. G.; Beak, P. J. Org. Chem. 1985,
50, 1216–1224. (f) Billman, J. H.; Sojka, S.; Taylor, P. R.
J. Chem. Soc., Perkin Trans. 2 1972, 2034–2035. (g) Allen,
G.; Dwek, R. A. J. Chem. Soc. (B) 1966, 161–163. (h) Burdett,
J. L.; Rogers, M. T. J. Am. Chem. Soc. 1964, 86, 2105–2109.
35. Reichardt, C. Solvents and Solvents Effects in Organic
Chemistry. 2nd ed. VCH Verlagsgesellschaft: Weinheim,
1988; pp 91–97.
40. Although 2c is formally alkyl substituted, the pyrrole ring can
substantially alter its electronic effects.
41. Two b-ketoamides analogous to 2, bearing strong electron-
withdrawing substituents [R1¼F3C, R2¼H; and R1¼H(CF2)4,
R2¼H], have been described to exist only as enol forms in
36d
CDCl3
42. When 2d and 2t bear an additional N-phenyl substituent,
.
concordant enol percentages of 20 and 0%, respectively, were
measured in chloroform using UV techniques36c
.
43. Okuro, K.; Furuune, M.; Miura, M.; Nomura, M. J. Org.
Chem. 1993, 58, 7606–7607.
44. See Ref. 35, pp 360 and 365.
45. See Ref. 35, p 113.
46. See Ref. 35, p 402.
47. Data from Ref. 34e and own [(CD3)2CO].
48. Eby, C. J.; Hauser, C. R. J. Am. Chem. Soc. 1957, 79,
723–725.
36. (a) Hynes, M. J.; Clarke, E. M. J. Chem. Soc., Perkin Trans. 1
1994, 901–904. (b) Wengenroth, H.; Meier, H. Chem. Ber.
1990, 123, 1403–1409. (c) Naoum, M. M.; Saad, G. R.
J. Solution Chem. 1988, 17, 67–76. (d) Pashkevich, K. I.;
Saloutin, V. I.; Fomin, A. N.; Rudaya, M. N.; Egorova, L. C.
Izv. Akad. Nauk. SSSR, Ser. Khim. 1986, 1586–1592. (e)
Dudek, G. O.; Volpp, G. P. J. Org. Chem. 1965, 30, 50–54.
37. Apart from papers with tautomeric orientation,36 see for
instance Ref. 10d.
38. A clear example is that of b-ketoamide 2s. Our 1H NMR
spectrum (CDCl3) shows that it is a 21:79 keto–enol mixture;
however, it has been described as only keto form from 1H
NMR (CDCl3).10g
49. Sato, M.; Ogasawara, H.; Komatsu, S.; Kato, T. Chem. Pharm.
Bull. 1984, 32, 3848–3856.
50. Beilsteins Handbuch der Organischen Chemie, Vierte Auflage,
Band 3, p 244.
ˆ
51. Mironescu, A.; Ioanid, G. Bull. Soc. Chim. Romania 1935, 17,
107–129, Chem. Abstr. 1936, 30, 1053.
52. When freshly dissolved in CD3OD, the keto–enol ratio is 95:5,
and virtually no Ha/D interchange is observed. In CDCl3
(keto): 2.24 (s, Hg), 4.66 (s, Ha), 5.75 (broad s, 1NH), 6.90
(broad s, 1NH); HAr masked by the aromatic signals due to the
major (79%) enol tautomer.
39. Hansen, P. E.; Duus, F.; Bolvig, S.; Jagodzinski, T. S. J. Mol.
Struct. 1996, 378, 45–59.
53. Bu¨chi, J.; Schneeberger, P.; Lieberherr, R. Helv. Chim. Acta
1953, 36, 1402–1410.