Kinetic resolution of chiral hydroperoxides: hydrogen-peroxide-mediated screening of peroxidase-active soil bacteria

Full text of DOI:10.1021/jo9818327

7598
J . Org. Chem. 1998, 63, 7598-7599
been developed. Since approximately 108-109 bacteria are
present in 1 g of soil,11 this covers a wide genetic diversity
suitable for initial screening. Hydrogen peroxide was chosen
as selecting agent, because it is a potent inductor of
peroxidase transcription; however, it may also select some
strains with catalase activity. In contrast to current screen-
ing procedures, in which the substrate is administered as
the only carbon, sulfur, or nitrogen source at an early stage
of the growth cycle, in our protocol, the substrate is added
as a selecting agent in addition to all the other nutrients
necessary for the normal growth of the bacteria.
Our growth medium was either a full (Plate Count Agar,
Difco Laboratories) or a minimal (Dworkin et al.12) medium,
which was supplemented with the necessary trace elements.
The rationale behind this strategy was that most of the
substrates will not be accepted as a sole nutrient source by
bacteria, but such substances are still toxic enough to induce
the detoxification in sufficient bacteria, which have the
necessary genes to express the required peroxidase. More-
over, the material may be added at any stage of the bacterial
growth cycle to allow for more diversity in the control of the
biotransformation under examination.
The present screening procedure with H2O2 resulted in
four distinct species, which have been subjected to charac-
terization by the sequencing of the first 300 base pairs of
their small subunit ribosomal RNA gene (16s/18s) with the
Taq-cycle-DyeDeoxy-terminator technique13 (ABI division of
Perkin-Elmer, Weiterstadt). The results show that three
Bacillus spp. have been isolated that were not identical.
Morphological and 18s-rRNA sequence data show that the
fourth strain isolated is most probably a Paecilomyces sp.
The Bacillus strain, which gave the best results in regard
to the kinetic resolution of the organic hydroperoxides, was
more rigorously characterized and determined to be a
Bacillus subtilis strain.
Kin etic Resolu tion of Ch ir a l Hyd r op er oxid es:
Hyd r ogen -P er oxid e-Med ia ted Scr een in g of
P er oxid a se-Active Soil Ba cter ia
Waldemar Adam,† Barbara Boss,‡ Dag Harmsen,§
Zoltan Lukacs,*,†,‡ Chantu R. Saha-Mo¨ller,† and
Peter Schreier‡
Institutes of Organic Chemistry and of Pharmacy and
Food Chemistry, University of Wu¨rzburg, Am Hubland,
D-97074 Wu¨rzburg, and Institute of Hygiene and
Microbiology, University of Wu¨rzburg, J osef-Schneider-Str. 2,
D-97080 Wu¨rzburg, Germany
Received September 9, 1998
Hydroperoxides have been used frequently in asymmetric
oxidation;1,2 in fact, recently, enantiomerically pure hydro-
peroxides have become important as oxidants in the syn-
thesis of optically active compounds.3a,b Since by conven-
tional chemical means it has been difficult to obtain
enantiomerically pure products, nowadays the popularly
used enzymes in preparative organic chemistry have been
successfully employed in view of their high degree of
enantioselectivity and catalytic activity.4 By this means,
optically active hydroperoxides were obtained with lipases,5
lipoxygenase,6 and horseradish peroxidase.7 Most recently,
a chemically modified subtilisin enzyme, namely selenosub-
tilisin, was successfully applied to the kinetic resolution of
racemic hydroperoxides.8 Nevertheless, large quantities of
pure enzyme, which are necessary for preparative applica-
tions, are difficult to come by and expensive.
These disadvantages are to be contrasted with whole cell
systems (bacteria, fungi, and plant/animal cells), which
circumvent such problems since they are potentially avail-
able in large quantities through self-replication. No work
along these interdisciplinary lines is currently available on
peroxidase systems in bacteria and fungi, certainly not for
the kinetic resolution of racemic hydroperoxides.
All of the bacterial species mentioned above were incuba-
ted with 0.07 mmol of the particular racemic hydroperoxide
for the kinetic resolution of the hydroperoxides (Scheme 1):
Sch em e 1. Kin etic Resolu tion of Hyd r op er oxid es 1
by Soil Ba cter ia or F u n gi
As a matter of fact, most of the focus in recent years has
been set on the elucidation of the genetic expression of
peroxidases during the response to stress factors.9,10 Here,
we communicate our results of the first kinetic resolution
of organic hydroperoxides by soil bacteria and fungi.
First, an adequate screening procedure to select suitable
soil bacteria for the biotransformation of hydroperoxides had
* To whom correspondence should be addressed. Fax: +49-931-8885484.
E-mail: zoltan@pzlc.uni-wuerzburg.de.
For these experiments, the (1-phenyl)ethyl (1a ) and 1-(1-
phenyl)propyl (1b) hydroperoxides were used as model
substrates. The results are summarized in Table 1.
† Institute of Organic Chemistry.
‡ Institute of Food Chemistry.
§ Institute of Hygiene and Microbiology.
(1) J ohnson, R. A., Sharpless, B., Ojima, I., Ed. Catalytic Asymmetric
Synthesis; VCH: Weinheim, 1993; pp 103-158.
(2) Adam, W.; Richter, M. J . Acc. Chem. Res. 1994, 27, 57-62.
(3) (a) Shum, W. P. S.; Saxton, R. J .; Zajacek, J . G. US Patent 5663384,
1997. (b) Adam, W.; Korb, M. N.; Roschmann, K. J .; Saha-Mo¨ller, C. R. J .
Org. Chem. 1998, 63, 3423-3428.
(4) Adam, W.; Lazarus, M.; Saha-Mo¨ller, C. R.; Weichold, O.; Hoch, U.;
Ha¨ring, D.; Schreier, P. Biotransformations with Peroxidases. In Advances
in Biochemical Engineering/ Bio-technology/ Biotransformations; Faber, K.,
Ed.; Springer-Verlag: Heidelberg, in press.
Two of the isolated bacterial species showed slow conver-
sion of the hydroperoxides and poor enantiomeric excess for
both the alcohol and the hydroperoxide (see entry 6 for an
example). Thus, these two strains were not considered for
further experimentation. The Paecilomyces sp. (entry 4)
showed much faster conversions, so that after only 80 min
as much as 92% of the hydroperoxide 1a was converted and
an enantiomeric excess (ee) of 79% of the (R)-hydroperoxide
1a was achieved (entry 4). The best results were obtained
with B. subtilis (entry 1), which converted 64% of hydro-
(5) Baba, N.; Mimura, M.; Hiratake, J .; Uchida, K.; Oda, J . Agric. Biol.
Chem. 1988, 52, 2658-2687.
(6) Datechva, V. K.; Kiss, K.; Solomon, L.; Kyler, K. S. J . Am. Chem.
Soc. 1991, 113, 270-274.
(7) Adam, W.; Hoch, U.; Lazarus, M.; Saha-Mo¨ller, C. R.; Schreier, P. J .
Am. Chem. Soc. 1995, 117, 11898-11901.
(11) Blaine Metting, F., J r. Soil Microbiological Ecology; Marcel Dekker
Inc.: New York, 1993.
(12) Dworkin, M.; Foster, J . W. J . Bacteriol. 1958, 75, 592-603.
(13) Harmsen, D.; Heesemann, J .; Brabletz, T.; Kirchner, T.; Mu¨ller-
Hermelink, H. K. Lancet 1994, 343, 1288.
(8) Ha¨ring, D.; Herderich, M.; Schu¨ler, E.; Withopf, B.; Schreier, P.
Tetrahedron: Asymmetry 1997, 8, 853-856.
(9) Storz, G.; Tartaglia, L. A.; Ames, B. N. Science 1990, 248, 189-194.
(10) Demple, B. Annu. Rev. Genet. 1991, 25, 315-337.
10.1021/jo9818327 CCC: $15.00 © 1998 American Chemical Society
Published on Web 10/16/1998

Communications
J . Org. Chem., Vol. 63, No. 22, 1998 7599
Ta ble 1. Biotr a n sfor m a tion of Hyd r op er oxid es by Diver se Micr oor ga n ism s
precultivationa
(days)
incubation
timeb (min)
convnc
(%)
ROHd
ee (%)
ROOHe
ee (%)
entry
ROOH
microorganism
1
2
3
4
5
6
7
8
9
1a
1a
1b
1a
1b
1a
1a
1a
1b
1a
1a
B. subtilisf
3
1
1
11
11
3
8
8
8
8
240
30g
120g
80g
120
240
15
30
45
15
60
64
94
47
92
36
19
26
67
55
57
23
30 (S)
20 (S)
36 (S)
8 (S)
3 (S)
0
24 (R)
25 (R)
9 (R)
7 (R)
23 (R)
88 (R)
>99 (R)
64 (R)
79 (R)
3 (R)
B. subtilisf
B. subtilisf
Paecilomyces sp.f
Paecilomyces sp.f
Bacillus sp.f
A. nigerh
6 (R)
10 (S)
37 (S)
7 (S)
24 (S)
9 (S)
A. nigerh
A. nigerh
10
11
Botrytis cinereah
Penicillium v.h,i
8
a
b
Time at which the culture has been pregrown. Actual time for the incubation of the substrate with the pregrown culture. c Conversion
d
determined by HPLC analysis (error limit ( 5%). Determined by multidimensional gas chromatography on a 2,6-dimethyl-3-O-pentyl-
â-cyclodextrin column; 1a : (a) DB-Wax 100-10 °C/min to 240 °C, (b) cyclodextrin column, 100 °C (15 min isocratic) to 2 °C/min to 200
°C; 1b: (a) DB-Wax 80-10 °C/min to 240 °C, (b) cyclodextrin column, 60 °C (15 min isocratic) to 2 °C/min to 200 °C) (error limit (1%).
e Determined by HPLC analysis on a chiral Chiracel OD-H column (isohexane:2-propanol 9:1, 220 nm, 0.5 mL/min) (error limit (5%).
f Bacteria isolated from topsoil by a selective screening procedure. Yeast extract was added to the growth medium. Fungi obtained
g
h
i
from Deutsche Sammlung von Mikroorgansimen und Zellkulturen (DSM). Penicillium verrucosum var. verrucosum.
peroxide 1a within 240 min with an enantiomeric excess of
88% for the (R) hydroperoxide 1a and 30% for the (S)-alcohol
2a .
When the sterically more demanding 1-(1-phenyl)propyl
hydroperoxide (1b) was applied to the Paecilomyces sp. and
B. subtilis, the former showed a sharp decline in the
conversion rate and the enantiomeric excess (entry 5). Thus,
the ee values were only 3% for the (R)-peroxide 1b and 3%
for the (S)-alcohol 2b at a conversion of 36%. The conversion
was only approximately one-third of that for the hydroper-
oxide 1a (entry 4), despite the longer incubation time for
the propyl derivative 1b. In contrast, B. subtilis still gave
47% conversion after 120 min to afford an enantiomeric
F igu r e 1. Comparison of the enantioselectivities (ee values) for
excess of 64% for the (R)-hydroperoxide 1b and 36% for the
(S)-alcohol 2b (entry 3). The results show a moderate
decrease in enantioselectivitiy, which has also been observed
with horseradish peroxidase,7 because the propyl derivative
1b is sterically more demanding and presumably does not
fit into the enzyme cavity as well as the ethyl homologue
1a . A decrease in conversion rate was observed, but the drop
is smaller than for horseradish peroxidase.7
For comparison, arbitrarily chosen commercially available
fungal systems were subjected to the hydroperoxides 1a ,b,
administered in the same amount as for the bacterial
incubations. The fungal cultures showed an even faster
conversion, i.e., more than 50% after 15-30 min incubation
(Table 1, entries 8 and 10). Unfortunately, due to the much
higher biomass present in fungal systems, these results
cannot be readily compared to the bacterial ones. Neverthe-
less, the enantiomeric excess obtained here was significantly
lower than that achieved with the cultures isolated from
topsoil. The best fungal system was Aspergillus niger, which
gave an ee value of 37% for the (S)-hydroperoxide 1a and
25% for the corresponding (R)-alcohol 2a (entry 8). The
biotransformation of hydroperoxide 1b with A. niger (entry
9) showed again a decrease in the conversion rate and a
substantial diminution of the enantiomeric excess (entry 9)
compared to what was achieved with (1-phenyl)ethyl hydro-
peroxide (1a ).
the kinetic resolution of racemic 1-phenylethyl (1a ) and 1-phenyl-
1-propyl hydroperoxides (1b) by A. niger (cf. Table 1, entry 8), B.
subtilis (cf. Table 1, entry 1), and horseradish peroxidase.7
dase. Thus, while the fungal systems, with the exception
of the Paecilomyces sp., and horseradish peroxidase yield
the S enantiomer of the peroxide and the R enantiomer of
the alcohol, the bacterial strains afford the opposite enan-
tiomers, namely R for the peroxide and S for the correspond-
ing alcohol. Therefore, there appear to exist structurally
distinct peroxidase enzymes in nature that differ funda-
mentally in their substrate recognition to provide such
differential enantioselectivities.
For the first time, the kinetic resolution of racemic
hydroperoxides has been successfully performed with whole
cell systems, in the present case, bacteria. B. subtilis,
especially, affords high enantioselectivities, allows for easy
handling, is environmentally acceptable, and is a readily
accessible peroxidase-producing system with little demand
on the growth medium. Indeed, our preliminary efforts have
already shown novel and encouraging results, which should
initiate the further use of whole-cell biocatalysis in the
asymmetric synthesis of cumbersome to prepare organic
substances such as optically active peroxides.
Ack n ow led gm en t. We express our gratitude to Dr. U.
Rdest (Institute of Microbiology, University of Wu¨rzburg,
Germany) for her valuable help and advice. We also thank
the DFG (SFB 347) and the Bayerische Forschungsstiftung
(FORKAT) for their generous financial support.
The results described herein for the microorganism A.
niger and B. subtilis are compared with those of the isolated
enzyme horseradish peroxidase reported earlier7 (Figure 1).
Clearly, B. subtilis (Table 1, entries 1-3) not only provides
a good alternative to horseradish peroxidase, but the ad-
ditional advantage is the inversion of the sense of the
enantioselectivity vs that observed for horseradish peroxi-
Su p p or tin g In for m a tion Ava ila ble: Experimental details (5
pages).
J O9818327