Screening for Amidases: Isolation and Characterization of a Novel D-Amidase from Variovorax paradoxus

Article abstract of DOI:10.1002/1615-4169(200210)344:9<965::AID-ADSC965>3.0.CO;2-Z

Using racemic tert-leucine amide as sole nitrogen source in minimal medium, 162 strains were isolated by enrichment techniques and shown to contain amidase activity. Among these isolates three D-amidase producers were found and identified as Variovorax pa

Full text of DOI:10.1002/1615-4169(200210)344:9<965::AID-ADSC965>3.0.CO;2-Z

FULL PAPERS
Screening for Amidases: Isolation and Characterization of a
Novel d-Amidase from Variovorax paradoxus
Lutz Krieg,a Marion B. Ansorge-Schumacher,b Maria-Regina Kulaa,*
a
Institute of Enzyme Technology, Heinrich-Heine University D¸sseldorf, 52426 J¸lich, Germany
Fax: ( ‡ 49)-2461-612490, e-mail: M.-R.Kula@fz-juelich.de
b
Department of Biotechnology, Aachen University of Technology (RWTH), Worringerweg 1, 52074 Aachen, Germany
Fax: ( ‡ 49)-241-8022387, e-mail: M.Ansorge@biotec.rwth-aachen.de
Received: April 29, 2002; Accepted: July 28, 2002
Dedicated to Professor Dr. Roger A. Sheldon on the occasion of his 60th birthday.
Abstract: Using racemic tert-leucine amide as sole drolyzed amino acid amides as well as carboxamides
nitrogen source in minimal medium, 162 strains were and 2-hydroxy acid amides. The stereoselectivity of
isolated by enrichment techniques and shown to the reaction was variable, however. Hydrolyzing dl-
contain amidase activity. Among these isolates three Tle-amide the enantiomeric ratio E was >200
d-amidase producers were found and identified as resulting in d-Tle with an ee of >99% and up to
Variovorax paradoxus (two strains) and Klebsiella 47% conversion. Similar results were obtained with
spec. The d-amidase from Variovorax paradoxus was dl-Leu-amide and dl-Val-amide while dl-Phe-
purified to homogeneity by three chromatographic amide was hydrolyzed with an enantiomeric ratio E
À
steps. With dl-Tle-amide as substrate Michaelis
Menten kinetics were observed with a KM of 0.74 mM,
a KI of 640 mM and a Vmax of 1.4 U/mg. The amidase
of only 5.
has a broad pH-optimum between 7 and 9.5 and a Keywords: d-amidase; d-amino acids; enzyme screen-
temperature optimum at 47 49 8C. The amidase hy- ing; tert-leucine; Variovorax paradoxus
Introduction
The latter reaction is used for the industrial produc-
tion of l-tert-leucine [l-Tle, (S)-2-amino-3,3-dimethyl-
The production of chiral 2-amino acids was an important butanoic acid],[7] utilizing leucine dehydrogenase
driving force in the development of enzyme-catalyzed (EC 1.4.1.9) from Bacillus cereus as catalyst and
processes. l-2-Amino acids are the building blocks of all regenerating NADH in-situ by formate dehydrogenase
proteins and peptides and therefore an essential (EC 1.2.1.2), see Figure 1.[8] Several other l-amino acids
component of human and animal nutrition. Special with sterically demanding side chains can be produced
amino acids often with sterically demanding side by the same route.[9]
chains are important intermediates in the synthesis of
Only l-enantiomers are accessible by reductive
enzyme inhibitors or receptor blockers.[1] d-Amino amination since d-selective amino acid dehydrogen-
acids are widely used to synthesize derivatives of ases have not been described until now. There is a high
penicillin or cephalosporin with improved pharmaceut- interest to make also d-Tle and related d-amino acids
ical properties.[2] Furthermore, chiral amino acids or available by a feasible approach.
their derivatives are employed as catalysts[3] and chiral
Recently, an asymmetric catalytic variant of the
auxiliaries in chemical asymmetric synthesis[4] or as Strecker reaction has been described yielding d-Tle in
ligands in the development of new organometallic high enantiomeric purity (ee >99%).[10] Whether larger
catalysts.[5] Chiral 2-amino acids are produced by
exploiting different enzymes and precursors, e.g.:
enantioselective hydrolysis of racemates starting
with N-acetylamino acids, amino acid amides or
hydantoins,
stereoselective addition of ammonia to double bonds,
modification of chiral precursors,
transamination or reductive amination of 2-oxo
acids.[6]
Figure 1. Synthesis of l-tert-leucine via reductive amination.
Adv. Synth. Catal. 2002, 344, No. 9
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 1615-4150/02/34409-965 973 $ 17.50+.50/0
965

Lutz Krieg et al.
FULL PAPERS
amidases require a free amino function for substrate
recognition.[21] When dl-N-acetyl-Tle-amide was added
to medium (M1) as nitrogen source, a few strains were
obtained after the enrichment procedure growing
poorly. However, none of the isolated strains exhibited
amidase activity towards dl-N-acetyl-Tle-amide. The
negative result could be a consequence of the repressor
activity due to the glucose employed as the carbon
source. Attempts to employ dl-N-acetyl-Tle-amide as
single carbon and nitrogen source (medium M2) were
hampered by the unexpectedly low solubility of the
screening substrate (<0.5%) and no amidase producer
was obtained from enrichment cultures.
Figure 2. Possible enzymatic routes for kinetic resolutions of
amino acid precursors with d-specific hydrolases.
In the next enrichment procedure, we used dl-Tle-
amide as nitrogen source (medium M3), which proved
amounts can be produced this way remains to be much more productive. Out of 81 samples collected
established.[11]
from soil/water/natural habitats, 224 single colonies
Enzymatic approaches for the production of d-Tle were isolated and tested for amidase activity. Among
have also been investigated. The commercially available those, 62 strains were discarded, because of their poor
acylases and amidases do not accept the corresponding growth in liquid media. The remaining 162 strains (72%)
Tle derivatives as substrates, however. Using whole cell revealed amidase activity when tested for ammonia
biocatalysts activity towards l-Tle-NH2 was observed in liberation from dl-Tle-amide. Using a chiral thin-layer
some cases.[12,13] With a d-hydantoinase (EC 3.5.2) it was chromatography for amino acids the enantioselectivity
possible to stereoselectively open the ring of racemic 5- was determined as high d- or l-selective (ee >80%) or
tert-butylhydantoin. The resulting N-carbamoyl-d-Tle non-selective amidase activity. Analysis of the positive
can be converted to d-Tle using a d-carbamoylase strains revealed 3 strains that produced the desired d-
(EC 3.5.1.77) as a second enzyme or by a chemical amidase, while all others produced predominantly an l-
reaction with nitrite.[7] 5-tert-Butylhydantoin is a very amidase. The screening results are summarized in
poor substrate of the available hydantoinase, therefore, Table 1. Apparently, transaminases do not tolerate
long reaction times are required, which make this route substrates highly branched at C-3 as all isolated
less attractive for a larger scale. To provide a better strains, which grew sufficiently in liquid minimal
alternative, we decided to screenfor amidases (EC 3.5.1.4) medium, exhibited amidase activity. Since l-amino
or acylases (EC 3.5.1.14) able to accommodate structures acids are predominantly found in nature, it is not a
in their reactive sites, which are highly branched in the C-3 surprising result that the majority of the isolates had an
position, like Tle (Figure 2).[14]
l-amidase. Among the previously described amidases
Since such enzymes appear quite rare we employed an only a few are d-selective enzymes and the majority of
enrichment technique to find enzyme producers able to these were obtained by screening on the corresponding
utilize racemic precursors of Tle as carbon and/or d-enantiomer.[16 18]
nitrogen source. A similar approach had proven
The three isolated d-amidase producers were identi-
successful before, e.g., in the isolation of carnitine fied by the DSMZ (Deutsche Sammlung von Mikroor-
amidase producers[15] or d-amidase producers.[16 18] In ganismen und Zellkulturen, Braunschweig, Germany).
this paper we describe the screening for Tle amidase The results as well as some activity data are shown in
producers, as well as the isolation and characterization Table 2. The two Variovorax strains originated from
of a new d-amidase.
different environmental samples. Since isolate 19-3 had
a significantly higher specific activity in the crude
extract, it was selected for further studies.
Results and Discussion
Table 1. Results of screening 224 strains isolated from
enrichment cultures of 81 samples from soil/water using dl-
Tle-amide as nitrogen source.
Screening
We decided to use dl-N-acetyl-Tle-amide as screening
substrate first to avoid nitrogen mobilization by trans-
amination and thus the generation of false positive
colonies. A microbial peptide amidase[19] has been
described hydrolyzing N-acetylamino acid amides[20]
stereoselectively, though most known amino acid
Strains
Number
[%]
Isolated
Amidase activity
l-amidase
224
162
159
3
100
72
71
d-amidase
1.3
966
Adv. Synth. Catal. 2002, 344, 965 973

Isolation and Characterization of a d-Amidase from Variovorax paradoxus
FULL PAPERS
Table 2. Identity of the isolated d-amidase producers. Activ-
ity was determined in the crude extract after growth of the
strains in medium M3.
could be minimized by using Omega membranes
instead.
After the HIC-step on butyl-Sepharose, the d-
amidase was about 50% pure and contaminated by
only one other protein as shown in Figure 3. The
contaminating protein could be completely removed
by gel filtration yielding the homogeneous enzyme.
From the gel filtration, the native molecular weight of
the d-amidase was determined as 171 kDa, while the
molecular weight of the denatured protein chain was
determined as 50 kDa by SDS-PAGE (Figure 3). How-
ever, additional information is required to establish the
quaternary structure of the d-amidase.
Strain No. Identity
Activity
Spec. activity
[mU/mL] [mU/mg]
19-3
34-4
47-4
Variovorax paradoxus 64 33
Variovorax paradoxus 38
Klebsiella sp. <5
19
Cultivation of Variovorax paradoxus
Growth of the Variovorax strain on medium M3 was
rather slow, after about 90 h the stationary phase was
reached at an optical density at 660 nm of about 2. The Characterization of the d-Amidase from Variovorax
specific activity of the d-amidase was almost constant paradoxus
throughout the growth period. Furthermore, in this mini-
malmedium thestrainproducedmuch slime, presumably Influence of Metals and Inhibitors on the Amidase
because of the unfavorable C/N ratio.[22] Growth on Activity
mediarecommendedbythe DSMZcatalogue (medium1
and 81) for Variovorax strains was faster but resulted in Mg2‡, Ca2‡, Zn2‡, Mn2‡, Co2‡ and Ni2‡ were tested for
about the same optical density (2.5) and no d-amidase possible enzyme activation but showed no effects. Also,
could be detected in the absence of dl-Tle-amide as
inducer. Additionofyeastextractintheminimalmedium
M3 accelerated the growth rate but gave less d-amidase
activity. Attempts to induce the amidase production with
other commercially available amides, such as isobutyr-
amide, instead of dl-Tle-amide also failed, as the cell
yield increased about 100%butthespecificactivityofthe
crude extract dropped to <3 mU/mg. Therefore, 30 L
minimal medium M3 was employed in a bioreactor to
produce enough biomass for the purification and charac-
terization of the d-amidase. On this scale an average cell
yield of approximately 5 g (wet weight)/L culture volume
and a specific activity of 20 mU/mg was reached. The
activity yield was about 1 U/L culture volume.
Purification of the d-Amidase from Variovorax
paradoxus
Figure 3. Purification of d-amidase from Variovorax para-
doxus after HIC, analyzed by SDS-PAGE: IEC-pool (lane 1;
5 mg protein), flow through HIC (lane 2; 5 mg protein),
After some initial optimization, the d-amidase was
fraction No. 6: 0.07 U/mg (lane 3, 2.5 mg protein), fraction No.
purified by three chromatographic steps as summarized
in Table 3. The low yield after the last step resulted from
losses during ultrafiltration using YM 10 membranes to
concentrate the pool. Later, it was found that losses
9: 0.2 U/mg (lane 4, 2.5 mg protein), marker proteins (lane 5),
fraction No. 13: 0.8 U/mg (lane 6, 2.5 mg protein), fraction No.
17: 1.4 U/mg (lane 7, 2.5 mg protein). d-Amidase is indicated
with an arrow. The gel was silver stained.
Table 3. Purification of d-amidase from 80 g wet cells from Variovorax paradoxus.
Purification step
Spec. activity
[U/mg]
Protein
[mg]
Total activity
[U]
Yield
[%]
Purification
factor
Crude extract
0.02
0.16
0.68
1.4
800
94
20
16
15
14
4
100
94
87
Q-Sepharose FF
Butyl-Sepharose FF
Superdex 200 PG
8
7
2
2.9
25
Adv. Synth. Catal. 2002, 344, 965 973
967

Lutz Krieg et al.
FULL PAPERS
Table 4. Effects of various compounds on the d-amidase from Variovorax paradoxus. Enzyme preparations after two
chromatographic steps were used with a specific activity of 0.56 U/mg.
Additive
Relative activity [%][a]
Additive (1 mM)
Additive (0.1 mM)
none (control)
100
0
1.3
69
100
104
0
0.7
79
100
0
88
92
97
100
6.9
38
‡
Hg2‡, Ag
Cu2‡
Fe2‡
Fe3‡
Pb2‡
phenylmethanesulfonyl fluoride
p-chloromercuribenzoate
DTNB[b]
90
iodoacetamide
hydroxylamine
EDTA
93
101
100
n. d.
n. d.
100
104
99
o-phenanthroline
[a]
After preincubation of the enzyme for 30 min with the compound to be tested, enzyme activity was measured by HPLC
under the standard conditions described in the experimental section.
5,5'-Dithiobis-(2-nitrobenzoic acid), DTNB.
[b]
À
Figure 4. Michaelis Menten plot for the conversion of dl-
Tle-amide by the d-amidase from Variovorax paradoxus.
Enzyme preparations after one chromatographic step were
used with a specific activity of 0.16 U/mg.
Figure 5. Optimum pH for the d-amidase activity. Levels of
activity were determined by using the standard assay and the
following buffers (each at a final concentration of 0.1 M):
&
*
sodium citrate ( ), sodium acetate ( ), 2-(N-morpholi-
^
no)ethanesulfonic acid (MES) ( ), potassium phosphate
~
*
( ), Tris-(hydroxymethyl)aminomethane ( ), and sodium
the d-amidase was not inactivated by chelators such as
EDTA and o-phenanthroline (Table 4). Both results
demonstrate that the amidase does not require metal
ions as a cofactor. The other results summarized in
Table 4, indicate that thed-amidase may have a serine in
the catalytic center as indicated by the inhibition with
^
carbonate ( ). The activity for sodium carbonate buffer,
pH 9.2, with 0.61 U/mg was taken as 100%.
Kinetics
phenylmethanesulfonyl fluoride known to interact with With partially purified d-amidase (after ion exchange
serine in the active center of proteases.[23] Cysteine(s) chromatography) the initial rate of hydrolysis was
appears to play an important role also for activity and/or determined. The results are shown in Figure 4. The
À
structural integrity of the enzyme. The latter is referred kinetics can be described well by the Michaelis Menten
from the strong inhibition by the sulfhydryl reagent p- equation taking into account a slight substrate inhib-
‡
chloromercuribenzoate and the metal ions Hg2‡, Ag
ition. Thus for dl-Tle-amide a KM of 0.74 mM and a KI
of 640 mM were calculated. Pure amidase reached a
Vmax of 1.4 U/mg.
and Cu2‡ at 1 mM concentration.
The pH profile of the d-amidase is presented in
Figure 5. It shows a broad optimum from pH 7 to 9.5.
968
Adv. Synth. Catal. 2002, 344, 965 973

Isolation and Characterization of a d-Amidase from Variovorax paradoxus
FULL PAPERS
Table 5. Substrate specificity of partially purified d-amidase Table 6. Substrate specificity of partially purified d-amidase
for amino acid amides and 2-hydroxy acid amides. Enzyme for carboxamides. Enzyme preparations after two chromato-
preparations after two chromatographic steps were used with graphic steps were used with a specific activity of 0.56 U/mg.
a specific activity of 0.56 U/mg.
Substrate
Relative activity[a]
Substrate
Relative activity[a, b]
dl-Tle-amide
1
dl-Tle-amide
d-Val-amide
l-Val-amide
d-Leu-amide
l-Leu-amide
l-Ile-amide
d-Ala-amide
l-Ala-amide
d-Phe-amide
l-Phe-amide
l-Tyr-amide
dl-Trp-amide
l-Trp-amide
dl-Met-amide
l-Met-amide
d-Pro-amide
l-Pro-amide
d-Lactic acid amide
l-Lactic acid amide
1
13
Formamide
Acetamide
Propionamide
n-Butyramide
Isobutyramide
Valeric acid amide
Acetoacetic acid amide
Malonic acid diamide
Succinic acid diamide
Adipic acid diamide
Acrylamide
0.13
4.0
26
37
33
1.2
400
26
0.35
44
36
130
2.9
2.1
14
7.7
5.2
19
15
890
190
100
160
62
230
86
24
56
59
66
150
Benzamide
Nicotinamide
l-Asparagine
d-Asparagine
0.15
<0.05
[a]
The activity was determined from the amount of liberated
ammonia, and the value for dl-Tle-amide (0.56 U/mg) was
taken as 1.
[c]
dl-HMB-amide
[a]
The activity was determined by HPLC at a substrate
concentration of 20 mM. The value for dl-Tle-amide
(0.56 U/mg) was taken as 1.
The activity for proline amides and the hydroxy acid
amides were determined from the amount of liberated
ammonia.
amino acid amide substrates has a strong influence on
the activity. In general, the d-enantiomers are the
preferred substrates. With proline the l-enantiomer is
hydrolyzed faster, however, and the l- and d-enantiom-
ers of lactic acid amide are hydrolyzed with nearly equal
rates.
The d-amidase from Variovorax paradoxus also
readily hydrolyzes carboxamides. The highest activity
was observed with acetoacetic acid amide. Whether the
amides of the dicarboxylic acids were hydrolyzed
[b]
[c]
dl-HMB-amide ˆ dl-2-hydroxy-4-methylthiobutanoic
acid amide.
The temperature optimum for incubation times of stepwise cannot be decided with the assay used. The
15 min was found at 47 49 8C (data not shown).
corresponding amino acid, asparagine, containing an
amide function in the g-carboxyl group, was a very poor
substrate; compared to dl-Tle-amide it was converted
with 15% activity only.
Substrate Range
As expected, N-acetylated amino acid amides were
Besides various amino acid amides, 2-hydroxy acid very poor substrates, if hydrolyzed at all (data not
amides, carboxamides and N-acylated amino acid shown).
amides were investigated as substrates for the d-
amidase using a partially purified enzyme (HIC pool).
The results are summarized in Table 5 and 6.
Enantioselectivity
If available, both enantiomers of the amino acid
amides were used in the study. With the exception of The optical purity of d-Tle obtained during hydrolysis of
proline, HPLC wasusedto quantify theamino acids. The the racemic substrate by the d-amidase was measured as
hydrolysis of proline amides and the hydroxy acid a function of conversion. The results are summarized in
amides was determined from the liberated ammonia. Table 7. It can be seen that up to 47% conversion the ee
Not unexpected was the observation that the activity of value of d-Tle is >99. At the theoretical maximum yield
the d-amidase was higher towards other substrates than of 50% the ee value of the product was still 97.1% and
dl-Tle-amide. With 500 U/mg the highest activity was dropped further upon prolonged incubation. The
found towards d-Phe-amide, which is a 890-fold enantiomeric ratio E[24] well above 200 demonstrated
increase compared to dl-Tle-amide. From the series the high preference of the d-amidase from Variovorax
Leu > Val > Tle it can be seen that branching at C-3 of for d-Tle-amide as substrate. With the enzyme pool
Adv. Synth. Catal. 2002, 344, 965 973
969

Lutz Krieg et al.
FULL PAPERS
Table 7. Enantioselectivity of the d-amidase as a function of
conversion c of dl-Tle-amide. Enzyme preparations after
two chromatographic steps were used with a specific activity
of 0.56 U/mg.
amino acid amidase from Ochrobactrum did not hydro-
lyze aliphatic carboxamides, like acetamide and pro-
pionamide. Whether these results described here for the
d-amidase from Variovorax paradoxus are due to the
second protein in the enzyme preparation remains to be
confirmed. The (R)-ketoprofen-amidase hydrolyzed
Phe-amide, Leu-amide and Gly-amide among 5 amino
acid amides tested. The d-alanine acid amidase from
Arthrobacter sp. NJ-26[26] and the d-aminopeptidase
from Ochrobactrum anthropi C1-38[17] are enzymes
specific for d-Ala-amide and d-Ala-peptides and
amino acid amides with larger side chains were not
hydrolyzed. The physiological substrate for the enzyme
from Variovorax paradoxus remains to be elucidated.
Conversion c[a] [%]
eed-Tle [%]
[b]
18
29
40
43
45
47
48
50
98.8
99.0
99.0
99.2
99.4
99.1
98.6
97.1
[a]
Conversion c ˆ (d-Tle ‡ l-Tle)/(dl-Tle-amidetˆ0)  100.
eed-Tle ˆ (d-Tle l-Tle)/(d-Tle ‡ l-Tle)  100. ee, enan-
tiomeric excess.
[b]
Conclusion
In a screening with dl-Tle-amide as sole nitrogen source
a strain of Variovorax paradoxus was isolated producing
an inducible d-amidase, which is a promising enzyme for
the production of d-Tle by stereoselective resolution of
the racemate with an E value of >200. The enzyme had
the highest stereoselectivity towards aliphatic amino
acid amides with a chain length of about 5 carbon atoms.
At present only small amounts of the d-amidase can be
produced with the wild type organism Variovorax
paradoxus. Work is in progress, however, to clone and
express the gene in a recombinant host. The d-amidase
should be available soon in sufficient quantities for
further process development.
Figure 6. Enantioselectivity of the d-amidase as a function of
^
*
conversion c of dl-Val-amide ( ), dl-Leu-amide ( ) and dl-
Phe-amide ( ).
Experimental Section
&
Materials and General Methods
obtained after the ion exchange chromatography the
enantiomeric ratio E was slightly lower, at 120 180
yielding product at 50% conversion with 96.5% ee (data
not shown). The results suggest that partially purified
enzyme preparations can be used successfully for the
resolution of the racemate.
The enantioselectivity of the amidase is also high for
Leu and Val, which was demonstrated by measuring the
ee values as a function of conversion (Figure 6). d-Phe-
amide, the substrate with the highest hydrolysis rate
(500 U/mg) had an E value of only 5, resulting in an ee of
50.9 at 51% conversion.
Regarding the enantioselectivity for d-amino acid
amides and the described wide substrate range, the d-
amidase from Variovorax paradoxus appears to be
different from other well-known amidases.[14] Similar
to the d-amino acid amidase from Ochrobactrum
anthropi SV3[16] and the (R)-ketoprofen-amidase from
Comamonas acidovorans KPO-2771-4,[25] the new
dl-Tle-amide, l-Tle-amide, N-acetylamino acid amides, N-
acetylamino acids, l-Tle, dl-2-hydroxy-4-methylthiobutanoic
acid amide (dl-HMB-amide) were kindly supplied by Degussa
AG (Hanau-Wolfgang, Germany). Other amino acid deriva-
tives were purchased from Novabiochem (Bad Soden, Ger-
many) or Bachem Biochemica (Heidelberg, Germany).
NADH, 2-oxoglutarate and adenosine diphosphate were
products from Sigma-Aldrich (Deisenhofen, Germany), and
the glutamate dehydrogenase was obtained from Fluka
(Deisenhofen, Germany). All other chemicals unless noted
otherwise were commercial products of the highest purity
available. Media components were obtained from Merck
(Darmstadt, Germany) or Difco (Becton Dickinson, Sparks,
USA)
Culture Media
Media sterilization was carried out at 121 8C at 1.2 bar for
20 min. Glucose, MgSO4 ¥ 7 H2O, CaCl2 ¥ 2 H2O, Tle-deriva-
tives, vitamin solution and trace element solution were added
enzyme has high activity for d-Phe-amide. The d- separately after cooling of the medium by sterile filtration
970
Adv. Synth. Catal. 2002, 344, 965 973

Isolation and Characterization of a d-Amidase from Variovorax paradoxus
FULL PAPERS
using 0.2 mm membranes Minisart NML (Sartorius, Gˆttingen,
Germany). For solid media 1.5% agar (w/v) was added. The
minimal medium M1 comprised 1.95 g KH2PO4, 2.5 g K2HPO4,
1.0 g NaCl, 0.3 g MgSO4 ¥ 7 H2O, 0.05 g CaCl ¥ 2 H2O, 2.0 g
glucose, 2.0 g dl-N-acetyl-Tle-amide, 0.8 mL trace element
solution and 2.5 mL vitamin solution in 1000 mL. In minimal
medium M2 glucose was omitted, for the rest it had the same
composition as given for M1. Minimal medium M3 contained
3.3 g KH2PO4, 0.8 g K2HPO4, 1.0 g NaCl, 0.3 g MgSO4 ¥ 7 H2O,
0.05 g CaCl2 ¥ 2 H2O, 4.5 g glucose, 3.3 g dl-Tle-amide, 0.8 mL
trace element solution and 2.5 ml vitamin solution in 1000 mL.
Thetrace elementsolutionwas prepared asdescribed byJoeres
and Kula,[15] the composition of the vitamin solution was taken
from Schlegel.[27]
Analytical Methods
Amidase Assay
For the amidase assay, incubation was carried out at 30 8C in a
total volume of 500 mL containing 400 mL 0.1 M KPi, pH 7.5, 50
mL substrate (0.2 M dl-Tle-amide) and 50 mL crude extract.
After 10 60 min incubation the reaction was terminated by
heating the mixture for 3 min at 95 8C. Precipitated protein was
centrifuged off at 13000 rpm for 5 min. Two control reactions
were included, in which the hydrolysis reaction was performed
in the absence of substrate or enzyme extract, respectively.
Aliquots of the supernatant were either analyzed for ammonia
generated by the amidase activity or for Tle by an HPLC
method. One unit amidase activity hydrolyzed 1 mmole Tle-
amide per minute under the specified conditions.
Ammonia was determined by the glutamate dehydrogenase
assay as described by Bergmeyer and Beutler[29]. The method is
fast but does not provide information on the stereoselectivity.
Screening Procedure
30 mL of a minimal medium containing either dl-N-acetyl-
Tle-amide as nitrogen source (M1) or as carbon and nitrogen
source (M2) or dl-Tle-amide as nitrogen source (M3) was
inoculated with a sample from natural water sources or a
suspension of soil from various habitats in sterile saline. The
shake flasks with the cultures were placed on a rotary shaker
HPLC for the Determination of d- and l-Amino Acids
To separate amino acid enantiomers by RP-HPLC and enable
their detection by fluorescence, they were converted into
diastereomers using N-isobutyryl-l-cysteine (IBLC) or N-
isobutyryl-d-cysteine (IBDC) as chiral reagents together with
o-phthaldialdehyde (OPA) according to Br¸ckner et al.[30] The
derivatization reagent was prepared in 100 mM sodium borate
buffer pH 10.4 containing 170 mM OPA and 260 mM IBLC or
IBDC. 20 mL reagent were mixed with 180 mL samplediluted in
100 mM sodium borate buffer pH 10.4. An aliquot of 20 mL was
used for analysis. As stationary phase, a Kromasil column (250
 4 mm, 5 mm, 100 ä) was employed in HPLC equipment
(Dionex, Germering, Germany). For separation, a two
component buffer was applied consisting of A: 23 mM sodium
acetate, titrated with acetic acid to pH 6.0 and B: acetonitrile
mixed with water 10:1.5 (v/v).[31] The Tle, Leu, Ile, Val, Phe, Tyr,
Trp und Met enantiomers were separated using a flow rate of
1 mL/min and the following gradient: time t ˆ 0 min 22%
buffer B, 25 min 29% B, 27 min 100% B, 30 min 100% B,
32 min 0% B, 42 min 0% B. For Gln, His and Ala the gradient
program had to be changed starting with 12% buffer B (Gln
and His) and 15% B (Ala) for sufficient separation of the
enantiomers. The corresponding amides were also modified
and eluted well separated from the free amino acid. A
fluorescence detector was used with the excitation wavelength
set at 340 nm and emission at 440 nm. The peak area was
calibrated using the pure enantiomers whenever available.
Optical purity was expressed as: enantiomeric excess (% ee) ˆ
(major enantiomer minor enantiomer)/(sum of enantiomers)
 100. The enantiomeric ratio E was calculated as: Eˆ ln[1
c(1 ‡ eed-Tle)]/ln[1 c(1 eed-Tle)].[21]
operated at 150 rpm and 30 8C. An aliquot was taken after 3
7
days depending on the developing turbidity indicating micro-
bial growth, and transferred into 30 mL of fresh medium. After
4 transfers the cultures were diluted and plated on agar plates
prepared with the same medium and incubated at 30 8C. The
development of colonies was observed by visual inspection
over several days. Up to four single colonies from each plate
were picked and purified by further plating until pure cultures
were obtained. Pure cultures were maintained on agar plates
with minimal medium at 4 8C and grown in 30 or 100 mL liquid
media in shake flasks for further evaluation.
Cultivation of Variovovax paradoxus
To obtain sufficient biomass for enzyme isolation the d-
amidase producer was cultivated in 30 L medium M3 in a
bioreactor (Techfors, 40 L reactor volume, Infors, Bottmingen,
Switzerland) for 92 h. Cells were harvested using a chamber
centrifuge KA 02 (Westfalia Separator, Oelde, Germany)
À
operated at 9600 rpm. Cells were frozen and stored at 20 8C
until further use.
Cell Disintegration
Wet cells were suspended in 0.1 M potassium phosphate (KPi
buffer) pH 7.5 to 20% (w/v) and disintegrated by wet milling
using 0.3 mm glass beads and a mixer mill (Retsch, Haan,
Germany) for small volumes[28] or the disintegrator S (IMA,
Frankfurt, Germany) operated at 3500 rpm, for 20 min. Glass
beads and cell suspension were mixed in a ratio 2:1. Thereafter,
cell debris and glass beads were sedimented by centrifugation
and the supernatant was used as the enzyme source.
Chiral Thin-Layer Chromatography (TLC)
For qualitative evaluation of enzyme selectivity during screen-
ing chiral thin-layer chromatography (CHIRAL-PLATE silica
gel RP modification coated with Cu2‡ and chiral reagent;
Macherey and Nagel, D¸ren, Germany) was employed and
methanol/water/acetonitrile 1:1:4 (v/v/v) was used as the
mobile phase. Plates were activated for 15 min at 100 8C prior
Adv. Synth. Catal. 2002, 344, 965 973
971

Lutz Krieg et al.
FULL PAPERS
to use. After development plates were dried and spots
visualized by reaction with ninhydrin (0.3% (w/v) ninhydrin
in 2-methylpropanol).
Acknowledgements
We thank Degussa AG for the generous supply with substrates,
Dr. K. G¸nther (Degussa) for his advice to separate amino acid
enantiomers by HPLC and Dr. A. Bommarius (Degussa) for
discussions. The work described was in part supported by a
fellowship from the Hermann-Schlosser-Stiftung of Degussa.
Protein Determination
Total protein was determined according to Bradford.[32] Bovine
serum albumin (BSA) was used for calibration.
References
Purification of d-Amidase from Variovorax
paradoxus
[1] a) A. S. Bommarius, M. Schwarm, K. Drauz, J. Mol.
Catal. B: Enzym. 1998, 5, 1 11; b) M. Kottenhahn, K.
Drauz, in Chemical Specialties USA, Chicago, 1994.
[2] A. Bruggink, E. C. Roos, E. de Vroom, Org. Process Res.
Dev. 1998, 2, 128 133.
[3] For reviews, see: a) E. R. Jarvo, S. J. Miller, Tetrahedron
2002, 58, 2481 2495; b) M. J. Porter, S. M. Roberts, J.
Skidmore, Bioorg. Med. Chem. 1999, 7, 2145 2156.
[4] For reviews, see: a) A. Job, C. F. Janeck, W. Bettray, R.
Peters, D. Enders, Tetrahedron 2002, 58, 2253 2329;
b) M. D. Groaning, A. I. Meyers, Tetrahedron 2000, 56,
9843 9873; c) R. O. Duthaler, Tetrahedron 1994, 50,
1539 1650; d) D. J. Ager, I. Prakash, D. R. Schaad,
Chem. Rev. 1996, 96, 835 875.
Step 1. Ion Exchange Chromatography (IEC): Q-Sepharose
FF (Amersham Biosciences, Freiburg, Germany) was em-
ployed for the first step during purification. The column (5 Â
21 cm, 410 mL volume) was equilibrated with 20 mM KPi
buffer, pH 6.5 (buffer A), and operated with a flow rate of
10 mL/min. A sample of 185 mL crude extract was applied to
the column, and after washing elution was carried out in a
linear gradient changing to buffer B (20 mM KPi buffer,
pH 6.5, 150 mM sodium sulfate) within 4 column volumes.
Fractions of 12 mL were collected and analyzed for activity.
The appropriate fractions were pooled and adjusted to a
sodium sulfate concentration of 150 mM measuring the
conductivity (25 mS/cm).
[5] For reviews, see: a) G. Helmchen, A. Pfaltz, Acc. Chem.
Res. 2000, 33, 336 345; b) A. Pfaltz, in Comprehensive
Asymmetric Catalysis, (Eds.: E. N. Jacobsen, A. Pfaltz, H.
Yamamoto), Spinger, Berlin 1999, Vol. 2, pp. 513 538;
c) A. Pfaltz, M. Lautens, ibid., 833 886; d) J. S. Johnson,
D. A. Evans, Acc. Chem. Res. 2000, 33, 325 335; e) D. A.
Evans, J. S. Johnson, in Comprehensive Asymmetric
Catalysis, (Eds.: E. N. Jacobsen, A. Pfaltz, H. Yamamo-
to), Spinger, Berlin 1999, Vol. 3, pp. 1177 1236.
[6] A. S. Bommarius, M. Schwarm, K. Drauz, Chimia 2001,
55, 50 59.
[7] A. S. Bommarius, M. Schwarm, K. Stingl, M. Kotten-
hahn, K. Huthmacher, K. Drauz, Tetrahedron: Asymme-
try 1995, 6, 2851 2888.
[8] R. Wichmann, C. Wandrey, A. F. Buckmann, M. R. Kula,
Biotechnol. Bioeng. 1981, 23, 2789 2802.
[9] G. Krix, A. S. Bommarius, K. Drauz, M. Kottenhahn, M.
Schwarm, M. R. Kula, J. Biotechnol. 1997, 53, 29 39.
[10] M. S. Sigman, P. Vachal, E. N. Jacobsen, Angew. Chem.
Int. Ed. 2000, 39, 1279 1281.
[11] L. Yet, Angew. Chem. Int. Ed. 2001, 40, 875 877.
[12] W. J. J. van den Tweel, T. J. G. M. van Dooren, P. H.
de Jonge, B. Kaptein, A. L. L. Duchateau, J. Kamphuis,
Appl. Microbiol. Biotechnol. 1993, 39, 296 300.
[13] Y. Fujio, N. Tetsuji, (Mitsubishi Rayon), EU Patent
1 174 499, 2002.
Step 2. Hydrophobic Interaction Chromatography (HIC): A
column with 6.5 mL Butyl-Sepharose FF (Amersham Bio-
sciences, Freiburg, Germany) was equilibrated with 20 mM
KPi buffer, pH 6.5, 150 mM sodium sulfate before the pool of
the anion exchange column was applied at a flow rate of 1 mL/
min. Elution was performed in a linear gradient decreasing the
sodium sulfate concentration to zero within 6 column volumes.
Fractions of 4 mL were collected. Fractions containing d-
amidase activity were pooled and concentrated by ultrafiltra-
tion.
Step 3. Gel Filtration (GPC): Superdex 200 PG (Amersham
Biosciences, Freiburg, Germany) was packed in a column (1.6
 62 cm, 124 mL volume) in 20 mM KPi buffer, 180 mM
NaCl. The column was operated with a flow rate of 1 mL/min in
isocratic mode. The effluent was analyzed at 280 nm and 1 mL
fractions collected. Aliquots (1.0 mL) of the pool from the HIC
column were applied to achieve the final purification of the d-
amidase. The column was calibrated using the proteins from
the low and high molecular weight Kit (Amersham Bioscien-
ces, Freiburg, Germany).
Ultrafiltration: Diluted enzyme fractions were concentrated
by ultrafiltration using YM 10 membranes (Millipore, Es-
chborn, Germany) or Omega membranes (Pall Gelman
Laboratory, Dreieich, Germany), both membranes have a cut
off value of 10 kDa. Concentrated enzyme samples were mixed
À
with equal volumes of glycerol and stored at 20 8C.
[14] For reviews, see: a) L. Martinkova, V. Kren, Biocatal.
Biotrans. 2002, 20, 73 93; b) M. Yagasaki, A. Ozaki, J.
Mol. Catal. B: Enzym. 1998, 4, 1 11.
SDS-PAGE
[15] U. Joeres, M. R. Kula, Appl. Microbiol. Biotechnol. 1994,
40, 599 605.
[16] a) H. Komeda, Y. Asano, Eur. J. Biochem. 2000, 267,
2028 2035; b) Y. Asano, T. L. L¸bbeh¸sen, J. Biosci.
Bioeng. 2000, 89, 295 306; c) Y. Asano, T. Mori, S.
SDS-PAGE was carried out following the protocol described
by Laemmli[33] using 12.5% separation gels. The premixed
protein standard ™low range∫ (Roche Diagnostics, Mannheim,
Germany) was employed as a marker. Silver staining was
carried out according to Blum et al.[34]
972
Adv. Synth. Catal. 2002, 344, 965 973

Isolation and Characterization of a d-Amidase from Variovorax paradoxus
FULL PAPERS
Hanamoto, Y. Kato, A. Nakazawa, Biochem. Biophys.
Res. Commun. 1989, 162, 470 474.
[24] C.-S. Chen, Y. Fujimoto, G. Girdaukas, C. J. Sih, J. Am.
Chem. Soc. 1982, 104, 7294 7299.
[25] a) T. Hayashi, K. Yamamoto, A. Matsuo, K. Otsubo, S.
Muramatsu, A. Matsuda, K. Komatsu, J. Ferment.
Bioeng. 1997, 83, 139 145; b) K. Yamamoto, K. Otsu-
bo, A. Matsuo, T. Hayashi, I. Fujimatsu, K. Komatsu,
Appl. Environ. Microbiol. 1996, 62, 152 155.
[26] a) A. Ozaki, H. Kawasaki, M. Yagasaki, Y. Hashimoto,
Biosci. Biotechnol. Biochem. 1993, 57, 520 521; b) A.
Ozaki, H. Kawasaki, M. Yagasaki, Y. Hashimoto, Biosci.
Biotechnol. Biochem. 1992, 56, 1980 1984.
[27] H. G. Schlegel, Allgemeine Mikrobiologie, 6th edn.,
Thieme Verlag, Stuttgart, 1985, p. 174.
[28] W. Hummel, M. R. Kula, J. Microbiol. Methods 1989, 9,
201 209.
[17] a) Y. Asano, T. L. L¸bbeh¸sen, J. Biosci. Bioeng. 2000,
89, 295 306; b) Y. Asano, Y. Kato, A. Yamada, K.
Kondo, Biochemistry 1992, 31, 2316 2328; c) Y. Asano,
A. Nakazawa, Y. Kato, K. Kondo, J. Biol. Chem. 1989,
264, 14233 14239.
[18] S. E. Godtfredsen, C. Kim, I. Kjeld, H. F. Hermes, J. A.
van Balken, E. M. Meijer, (Novo Industri A/S, Stam-
icarbon B. V.), EU Patent 0 307 023, 1988.
[19] U. Stelkes-Ritter, K. Wyzgol, M. R. Kula, Appl. Micro-
biol. Biotechnol. 1995, 44, 393 398.
[20] U. Stelkes-Ritter, G. Beckers, A. Bommarius, K. Drauz,
K. Gunther, M. Kottenhahn, M. Schwarm, M. R. Kula,
Biocatal. Biotransform. 1997, 15, 205 219.
[29] H. U. Bergmeyer, H.-O. Beutler, in Methods of Enzy-
matic Analysis, (Ed.: H. U. Bergmeyer), VCH-Verlag,
Weinheim, 1985, pp. 454 461.
[30] H. Br¸ckner, R. Wittner, H. Godel, Chromatographia
1991, 32, 383 388.
[21] J. Kamphuis, W. H. Boesten, Q. B. Broxterman, H. F.
Hermes, J. A. van Balken, E. M. Meijer, H. E. Schoe-
maker, Adv. Biochem. Eng. Biotechnol. 1990, 42, 133
186.
[22] M. T. Madigan, J. M. Martinko, J. Parker, Brock Biology
of Microorganisms, 9th edn., Prentice-Hall, Inc., Upper
Saddle River, New Jersey, 2000, pp. 86 87.
[23] D. E. Fahrney, A. M. Gold, J. Am. Chem. Soc. 1963, 85,
997 1000.
[31] K. G¸nther, personal communications.
[32] M. M. Bradford, Anal. Biochem. 1976, 72, 248 254.
[33] U. K. Laemmli, Nature 1970, 227, 680 685.
[34] H. Blum, H. Beier, H. J. Gross, Electrophoresis 1987, 8,
93 99.
Adv. Synth. Catal. 2002, 344, 965 973
973