close

Вход

Забыли?

вход по аккаунту

?

Chiral Compounds Synthesized by Biocatalytic Reductions [New Synthetic Methods (51)].

код для вставкиСкачать
Chiral Compounds Synthesized by Biocatalytic Reductions
New Synthetic
Methods (51)
By Helmut Simon,* Johann Bader, Helmut Gunther, Stefan Neumann, and
Jordanes Thanos
It has been known for many decades that chiral compounds can be obtained by stereospecific biocatalytic reduction. Further significant methodological developments in this field
have, however, only been made during the past ten years; they include the application of
previously unused microorganisms and electron donors, the discovery of additional substrates for the known reductases, the development of methods for regenerating reduced pyridine nucleotides, and the discovery of new reductases which were sought for specific preparative purposes. Many chiral compounds can now be synthesized by microbial hydrogenation using H2 and hydrogenase-containing microorganisms as well as by electromicrobial
or electroenzymatic reduction. In the two latter methods, anaerobic or aerobic organisms
are supplied with electrons from electrochemically reduced, artificial mediators, e.g., methyl viologen. Reductases that d o not require pyridine nucleotides and can accept electrons
directly from reduced viologens are especially useful. Two examples of this type of enzyme
are described which are of preparative interest. Many cells contain methyl viologen-dependent NAD(P) reductases, a large number of which have still not been characterized. A
productivity number is proposed which allows different methods of bioconversion with microorganisms to be compared. The productivity numbers of compounds synthesized by the
methods described in this review are often 10- to 100-fold higher than those of substances
obtained by conventional techniques.
1. Introduction
A large number of chiral products can be obtained by
stereospecific reduction of appropriately substituted unsaturated compounds according to Reaction (a) or (b):
K
R
In Reaction (a), the biocatalyst only has to differentiate
between the Re and Si faces of a JI system. Two alternatives are possible in Reaction (b). If the originally sp2-hybridized C atom does not become chiral after reduction,
a
I
then only the pro-R or p r o 3 group (-C = X) has to be differentiated. If, however, substituent a is such that a chiral
group is formed from the
system after hydrogenation,
this raises the question of diastereomeric purity, since the
Re and Si faces must be differentiated in addition to the
pro-R and pro-S groups.
Reaction (a) is more important than (b). The reduction
of carbonyl groups has been described much more frequently than that of CC double bonds. The reduction of
C N double bonds has rarely been described but plays an
important part in the reductive amination of 0x0 compounds (e.g., in the synthesis of amino acids). A review of
('I
the microbial enantioselective reduction of ketones has recently been published"] (for reviews and literature surveys
see ref. [2-81).
The reduction of phenylglyoxylic acid to mandelic acid
by fermenting yeast was reported more than 70 years ago.[']
The reduction of ketones to chiral alcohols was described
in about 160 publications in the mid-1950s. Microbial reductions of systems of the type -CH=C< were the subject of only about 20 studies.[*]Later, in addition to microorganisms, enzymes of various purities were used for the
preparative reduction of unsaturated compounds.[4.lo. ' 'I
The reductases that have so far been employed, however,
require reduced pyridine nucleotides (nicotinamide adenine dinucleotide = NAD or nicotinamide adenine dinucleotide phosphate = NADP), which cost several thousand
marks per mole. The nucleotides must therefore be regenerated if they are to be used in catalytic quantities.
Six different methods of biocatalytic reduction can be
differentiated according to the nature of the donor and the
route of electron transport from the donor to the. substrate:
-Up to now carbohydrates have usually been employed
as electron donors for the microbial reduction of unsaturated substrates. NADH and an electron acceptor such
It
carbohydrate
I
Prof. Dr. H. Simon, Dr. J. Bader, Dr. H. Giinther,
Dr. S. Neumann, Dr. J. Thanos
Organisch-chemisches lnstitut der Technischen Universitat Miinchen
Lichtenbergstrasse 4, D-8046 Garching (FRG)
Angew. Chem. I n / . Ed. Enyl. 24 (1985) 539-553
0 VCH Vedag~gesellschafimhH, D-6940 Weinhelm. l Y B j
r ) ? ~ o - f ~ , ~ 3 . ~ / # 5 / / ~ 7 0 7 - f BI S02.501
I~Y 0
539
as acetaldehyde or pyruvate are formed from NAD and
the carbohydrate in a series of enzyme-catalyzed steps
(Route I). The desired reaction catalyzed by the “final”
reductase Ef is usually a side reaction.
--It is better to have a system in which the electron donor
Doredconverts NAD into NADH in a single, practically
irreversible step, the NADH then being consumed in the
reaction catalyzed by Ef (Route 11). Two enzymes isolated from different sources can be combined for this
route.
/”””\/“
R
H+XH
R‘
-Systems can also be very advantageous in which a natural electron mediator Mred (e.g., reduced ferredoxin) is
formed from the primary electron donor (e.g., Hzor formate); Mredthen converts NAD into NADH in a reaction catalyzed by E2 (Route 111.1). In this case, electron
transport can often be accelerated if, instead of a natural
mediator, a n artificial mediator such as viologen can be
used (Route 111.2).
Dored\
,,/”””“\/
results, methyl viologen-dependent NAD reductases
seem to be widely distributed.
-Further studies performed by our group have shown that
preparatively useful reductases are available which can
accept electrons directly from reduced viologens. Routes
V and VI thus become accessible. In Route V, enzymatically reduced methyl viologen is produced with, for example, H2 or formate as an electron donor. In Route VI,
the shortest and simplest electron transport system, the
enzyme Ef is supplied with electrons electrochemically
(i.e., using electricity) via the mediator MVO’.
R
V
d‘
R
Rpx
R
111. 1,
VI
-Route 1V has proved to be of value in many cases, because viologens (e.g., methyl viologen = M V @ @ 1,l’-di,
methyl-4,4’-bipyridinium dication) can be electrochemically reduced in a defined manner and, according to our
These six pathways are briefly described in Table 1 and
will be discussed from a practical point of view in Section
2.2.
Table I . Possible pathways of biological reductions.
Description
Electron donoi
Catalyst(s) for NAD(P)H regeneratiou o r reduction of the final
reductase
Scheme of
Comments
electron flow
NAD(P)H is produced after a
series of enzyme-catalyzed steps
and the formation of competing
electron acceptors
Glucose o r other carbohydrates
Whole cells, e.g., yeasts or bacteria
I
Used most often; productivity
numbers are usually low
Direct NADH formation
HCOOH, Hz, CH,CH>OH
Whole cells, e.g., Pseudomonas
oualis, Arthrobacrer spec., o r
Clostridium kluyueri
11
Seldom used; cells are employed
for NADH regeneration
Reduction of NAD(P) via natural or artificial mediators
H2, HCOOH
Whole cells, e.g., clostridia or
Proteus vulgaris: with or without
mediators such as viologens
111.1: 111.2
Numerous a$-unsaturated carhoxylates and 2-oxocarhoxylates
are reduced to chiral compounds
Two-enzyme system
HCOOH or other substrates of
an NAD-dependent dehydrogenase, e.g., isopropanol or cyclohexanol
Dehydrogenase and final reductase
I1
In some cases successful on a
pilot scale with NAD of increased molecular weight in
membrane reactors
Electromicrobial
Cathode of a n electrochemical
cell
IV
Examples are to he found in this
review
Electroenzymatic or electromicrobial
Cathode of an electrochemical
cell
Cells with methyl viologen-dependent NAD(P)H reductase
and pyridine nucleotide-dependent final reductase
Cells with final reductase that
can accept electrons directly
from mediators such as viologens or isolated final reductdse
VI
Examples known so far are
enoate reductase and 2-oxocarboxylate reductase. Both react
with a wide range of substrates
540
or V
Angew. Chem. Int. Ed. Engl. 24 11985) 539-553
As shown in this review, the methodology of biocatalytic
reduction can be considerably extended by employing microorganisms that have not so far been generally applied
together with the new types of enzymes found in them; a
second approach is to utilize coenzyme regeneration methods which until now have been used either very rarely or
not at all.
2. What Has Been and Can Be Improved by New
Methods?
2.1. Criteria for Evaluating the Conversion of Substances
Using Microorganisms
The efficiency of the conversion of substances using microorganisms can be characterized by a so-called productivity number (PN):[l2.l3I
PN
=
amount of product [mrnol]
dry weight of catalyst [kg] x time [h]
High productivity numbers mean better volume-time
yields and usually easier isolation of the product. If, for
example, 500 mmol of a product is to be synthesized in
24 h, approximately 1OOOg of catalyst is required for a
productivity number of 20 (which is frequently the case),
whereas only 10 g is needed if the productivity number is
2000. Moreover, it is much simpler to separate 100 g of a
product from 10 g of a catalyst (i.e., microorganisms) than
from 1000 g. If a microorganism consists of 5oy0 protein,
the productivity numbers and the enzyme units (U) customarily used in biochemistry are related as follows: if a
crude extract of a n organism contains the enzyme that is
rate-limiting for the conversion with a specific activity of
1 U/mg protein (1 pmol substrate transformed per mg protein per min), a productivity number of 30000 can be attained. This assumes, however, that other factors, such as
transport processes, concentration limits, and inhibition by
the substrate and/or product, d o not have any effect on the
reaction. (So far, activities have seldom been expressed in
SI units, according to which a quantity of a n enzyme has
an activity of 1 katal if it transforms 1 mol of substrate per
second. 1 U a 16.7 x
katal.)
The following factors are also important for biocatalytic
conversions in addition t o satisfactory productivity numbers :
-Availability of the cells. Are the cells or is the material
for inoculation commercially available from a microorganism collection? Is cell culture simple and reproducible?
-Reaction specificity. Is the substrate only converted into
the desired product or d o the substrate and/or product
react further?
-High stereospecificity of the enzyme (or the same high
stereospecificity of each of the enzymes catalyzing the
reaction). Many enzymes display a high degree of stereospecificity but there are a large number of except i o n ~ . ’ Furthermore,
’~~
a microorganism may possess two
or more enzymes that react with the substrate in stereoAngew. Chem. Int. Ed. Engl. 24 1198s) 539-553
chemically different ways. For a discussion of this topic
see ref. [l].
-Michaelis-Menten constant ( K , ) for the substrate and
inhibition constants (KJ for the substrate and product of
the catalyzing enzyme. Low K , values allow rapid conversion even of the last 5-10% of the substrate, i.e., the
reaction does not end “gradually.” Low K , values for the
substrate may mean that the substrate has to be added
continuously, which is usually not difficult if the reaction rate and thus the actual substrate concentration is
known. Low K , values for the product require selective
separation of the product, which is in most cases a complicated task.
-Stability of the biocatalyst. Here, a distinction has to be
made between whether resting or multiplying cells are
used. In the former case, the activities of the required
enzymes or the concentrations of the cosubstrates, such
as NAD(P) or NAD(P)H, can decrease sharply (for example, within 24 h) before the substrate has been transformed. If growing cells are used, product isolation is
usually impeded by the large quantities of biocatalyst,
and, depending on the microorganism, the substrate
and/or product may be converted further.
-Suitability for immobilization. In order to facilitate
product isolation and to stabilize (and eventually reuse)
the biocatalyst (intact cells, crude extracts, or enzymes),
it may be advantageous to immobilize it. The suitabilities of different biocatalysts for immobilization vary.
-Storage stability. It is advantageous if the enzyme activities of a prepared biocatalyst remain stable for a long
period of time (e.g., when frozen).
2.2. Evaluation of Older and More Recent Procedures
Bakers’ yeast (Saccharomyces cereuisiae) is the biocatalyst most commonly used for reductions based on Route I
(see Table 1); according to the criteria described above,
however, it is not a very suitable catalyst even though it is
readily accessible and stable. High productivity numbers
cannot be achieved via Route I for fundamental reasons.
Carbohydrates as electron donors must undergo five or six
consecutive, enzyme-catalyzed reactions before NADH is
formed. Also, the uptake of carbohydrate may be rate-limiting.”’] Furthermore, acetaldehyde is formed in alcoholic
fermentation and competes for the reduction equivalents;
formation of ethanol is usually the main reaction; it has
been reported that for every mole of ketone produced
200-2000 moles of ethanol are formed.“6. ”I
The experimental data in many published studies d o not
allow the calculation or estimation of productivity numbers. Examples that permit c a l c ~ l a t i o n [ ’ ~show
- ~ ’ ~ that the
majority of reactions performed with microorganisms and
carbohydrates have productivity numbers of 5-50, values
of about 200 being very rare. To the best of our knowledge,
the highest value, approximately 1000, has been reported
for the reduction of 2,2,6-trimethyl-5-cyclohexene1,4dione to (6R)-2,2,6-trimethyl- 1,4-cyclohexanedione with
The special problems associated with the use of
non-standardized bakers’ yeast or non-standardized methodology are illustrated by the studies published during the
54 1
past 50 years on the synthesis of chiral 3-hydroxybutyric
acid esters.[z2.2s.2(71 Even if only studies reported during the
last seven years are considered, ee (enantiomeric excess)
values of between 80 and 97% have been reported for the
synthesis of ethyl (S)-3-hydroxybutyrate with bakers'
y e a ~ t . [ ~ ' .Productivity
~~I
numbers of about 100 were found.
The ee values of 95-97% for the (S)enantiomer can, however, only be obtained if the concentration of the 3-0x0 ester does not exceed 1 g/L; if the concentration is 5 g/L,
the ee values only amount to 72%. Reductions with Mucor
Ja~~anicus["~
and Geotrichum candidum[2z1yield the ( R ) enantiomers. The investigations of Sih et aI.['-2',z71
on the synthesis of esters of 4-chloro-3-hydroxybutyricacid, which is
of interest as a precursor of carnitine, have helped to clarify the various factors involved in reductions with bakers'
yeast (see Section 3.2).
Seldom and only recently have electron donors that are
not carbohydrates (e.g., ethanol) been used for whole cells.
Ethanol would also perhaps be a better electron donor
than glucose for yeast. Furthermore, it has been suggested
that ketones can be reduced by yeast without adding carbohydrates or another electron donor by the use of appropriately larger quantities of cell material.[l71
The different results obtained in a large number of investigations demonstrate that the biochemical fundamentals of reductions are not always well thought out and that
the investigators are sometimes not aware of the fact that
the bakers' yeast used is not always the same. Even if the
same strain is used, the behavior of the yeast in reductions
may depend strongly on the culture conditions (culture
medium, temperature, stage of growth). An example is
given by the reduction of the 5-0x0 group of the monothioacetal of methyl 5,6-dioxocaproate. Takaishi et al.1291
found specific rotation values of between +5.4 and
- 2 1.1 for the products obtained with different strains of
Saccharomyces cerevisiae. This applies of course to more
or less all microorganisms.
According to our findings,['2.'3.30-341Hz is an especially
good electron donor (Routes 111.1, 111.2, or V) for reductions with organisms containing hydrogenases. In these
cases, Dored is HZ;this applies particularly if Route V is
possible. Formate could presumably also be of advantage
in cases in which the cells contain an NAD-dependent formate dehydrogenase. Allais et al.[351found such an enzyme
with a specific activity of 0.16 U/mg protein in a crude extract of Pichia pastoris. Schiitte et al.[361isolated a formate
dehydrogenase from the yeast Candida boidinii which
proved to be highly suitable for NADH regeneration in the
enzyme membrane reactor of the groups led by Kula and
Wandrey (see below). Zzumi et al.[371employed the NADdependent formate dehydrogenase from methanol-metabolizing bacteria to regenerate NADH with formate,
whereby productivity numbers of u p to about 6000 were
obtained for regeneration.
Examples of the use of ethanol (Table 1) are given by
the reductions of (+)- and (-)-cawone with Pseudomonas
ovalis performed by Noma et al.'381We have reduced various ally1 alcohol derivatives with Clostridium kluyveri and
ethanol to chiral alcohols such as (R)-2-methyl-l-butanol,
(R)-3-methyl-l-pentanol, and (2R,3S)-2-methyl-3-pheny1[2,3-2H]-1-propanol.[391
The advantages of ethanol over
542
Hz for this special case will be discussed in Section 3.1 (see
also ref. [40]).
In the case of NAD(P)H-dependent reductases, if enzymes are utilized on a preparative scale instead of cells,
NAD(P)H must be regenerated enzymatically (Table I ,
Route 11). It has not so far been possible to reduce
NAD(P) chemically or electrochemically with practically
useful recyclization numbers (for more recent experiments
see Wienkamp and S t e ~ k h a n [ ~ ' The
. ~ ~ groups
]).
of Kula and
Wandrey reported that NAD bound to polyethylene glycol
in an enzyme membrane reactor could be used for the preparation of amino acids with the aid of formate dehydrog e n a ~ e , and
~ ~ ~could
'
be recycled up to 80000 times. Numerous reports have appeared during the past few years on
other possible ways of enzymatically regenerating
NAD(P)H144-471
(see also Section 4.2).
NADH regeneration, especially with formate dehydrogenase or one of the more recently described methods, is
now a routine operation on a laboratory scale. Many chiral
compounds have already become accessible by the use of
this technique in combination with a small number of reductases. In the course of the past few years Jones et
a1.[11.481
and Nakazaki et al.[491have, for example, found a
large number of additional substrates for liver alcohol dehydrogenase, an enzyme that has been known for several
decades. Lamed et al.[501have drawn attention to some interesting applications of an alcohol/aldehyde/ketone oxidoreductase from the therrnophilic bacterium Thermoanaerobium brokii. May and P ~ d g e t t e [ have
~ ' ] reported on the
possible applications of oxidoreductases. It is likely that
further preparatively or even industrially useful oxidoreductases will be discovered since an intensive search for
these enzymes with a view to their possible applications
only began a few years ago. (S)-Phenylalanine and (2S)-2hydroxy-4-methylpentanoate dehydrogenase are recently
discovered examples of such enzyme^.[^^^^^]
Nonpolar compounds can also be enzymatically reduced in conjunction with NADH regenerating systems.
Hilhorst et al.[s41regenerated NADH in reverse micelles
with hydrogenase, methyl viologen, and lipoamide dehydrogenase. 20B-Hydroxysteroid dehydrogenase was used
to reduce progesterone dissolved in hexanoVoctane to
20B-hydroxy-4-pregnen-3-oneat the boundary to the aqueous phase. Routes IV and VI can be utilized for the continuous reduction of catalytic quantities of artificial electron
carriers by employing electrochemical cells ; NAD(P)H can
thus be regenerated with the NAD(P) reductases frequently encountered in microorganisms (Route 1V) or electrons can be directly transferred to the final reductase Ef.
3. Hydrogenation with Microorganisms
3.1. Reduction of the CC Double Bond
of a$-Unsaturated Aldehydes and Carboxylic Acids
According to our studies, Hz can be employed as an
electron donor with some microorganisms."2~13~30-34.551
This has the following advantages: I) The equilibrium of
Reaction (c) lies strongly to the right. 2) In resting cells the
Angew. Chem. Ini. Ed. Engl. 24 11985) 539-553
(c)
electron donor can be used solely for the desired reaction.
Side reactions, which can occur, for example, with acetaldehyde or other products that are formed from carbohydrates, do not take place. 3) The reaction can be monitored
continuously by following the H2 consumption. 4) The
productivity numbers are often 10- 100 times higher than
those found for reductions with yeasts, which considerably
simplifies isolation of the products. According to our findings, productivity numbers of 1000-3000 are often
achieved, particularly if catalytic quantities (1 -4 mM) of
an artificial electron donor such as methyl viologen are added. The electron flow in reductions with clostridia can be
described by Route 111.1 or 111.2.
Hydrogenation with Clostridium spec. La1 (DSM 1460),
Clostridium kluyveri (DSM 555), Proteus mirabilis (DSM
30 115), and Proteus vulgaris (DSM 30 118) will now be described. Many a$-unsaturated carboxylic acids or aldehydes (Tables 2 and 3) can be hydrogenated with C . La1
Table 2. Enoates of the type R'R'C=CR'CO? that can be hydrogenated with
Closrridium La1 and H, [a]. The last five compounds are allene derivatives.
R'
R'
R'
Relative rates
H
H
Me
Me
H
Me
H
H
H
H
Me
H
Me
H
Et
Pr
H
Me
H
Me
Me
H
H
H
320
170
150
30
20
300
100
90
H
H
100
H
H
130
H
H
100
H
H
20
OEt [c]
H
H
Br
20
40
90
90
CN
Me
OMe
Br
Me
Et
Et
H
Et
H
C~HS
p-CIC6Ha
C,H,
EtCH=
MeCH=
PhCH=
Ph(Et)C=
Ph(Et)C=
['??a]
[b]
R'
R'
Me
H
H
H
H
H
H
H
H
H
H
H
H
NHCHO
H
F
CI
Br
Me
Me
Et
Me
iPr
COOMe
CH,CH=CH2
CsH,
p-CIC6H4
.t-O2NCe,H,
p-MeOC,H,
p-Me2NChH4
o-HOC6H4
C6HS
o.m-(Me0)2C6H1
Me
H
Me
H
C~HS
CoHS
p-CIC,H,
H
HOCH2CH2
HOCH2CH2
Me
Me
HO
(CH,)?C=CHCH=CHCH2
(CH,)~C=CH(CHZ)~
Me
(CH3)2C=CH(CH&
PhCH=
1
F
Br
CI
CI
Me
H
H
H
H
H
H
H
H
Me
H
H
Me
Et
H
H
H
H
H
H
H
H
H
H
H
H
Me
H
Me
100
280
II
H
11
130
9
18
60
88
30
44
29
3s
21
8
150
180
150 Is1
180 [h]
30
16
20
0.03
I20
H
H
p-CIC6H4
H
10
Me
HOCH2CH2
9
5
MeOCH2CH2
Me
6
H
7
Me
19
( C H ~ ) ~ C = C H ( C H Z )1~. 1
Me
60 [el
20
[a] If not otherwise stated, X = C O O Q . [b] Relative rate with respect to (E)-2methylbutenoate (100%). Purified enoate reductase displays a specific activity of 20 U/mg for this substrate with reduced methyl viologen at 25°C. [cl
K , values in mM. [dl Only the a$ double bond is hydrogenated. [el
X =CHO. [fJ90% E a n d 10% Z isomer. [g] Marked substrate inhibition above
40 mM. [h] Marked substrate inhibition above 10 m M .
found to apply: R' should not be too large; if R2 is a phenyl ring, NHCOCH3 and OC2H5residues are not accepted
in place of NHCHO and OCH3, respectively. The halogens
+
NADH
+
HQ
+
NAD@
(d)
-
80
-
[a] All the compounds listed in Table 3 can also be hydrogenated since they
are substrates of enoate reductase. [b] A relative rate of 100 signifies that 100
mM solutions are hydrogenated at a rate of 150 p n o l / h by 400 mg of wet cell
sediment ( 2SO-mg dry weight) at 35" and pH 7.0. This corresponds to a productivity number of about 1900. [c] E / Z ratio unknown.
(Route V, Table 1). Doredis H2, E l a hydrogenase, and Ef
the enoate reductase that we have discovered (EC
1.3.1.31)."3.56-601
Th e reactions catalyzed by this enzyme indude Reactions (d) and (e)."', '3.60.611
The reaction with MVQo proceeds 1.5 times faster than
that with NADH. Enoate reductase can react with a very
wide range of substrates. The following rules have been
Angew. Cheni. I n t . Ed. Engl. 24 (1985) 539-553
Table 3. Substrates (R'R'C=CR'X) of enoate reductase that are reduced according to Equations (d) or (e) [a] but can also be hydrogenated with H2 and
C. La1 cells
H
I
X = COOo, C=O; R', R2,R3 s e e Tables 2 a n d 3
F, C1, and Br are tolerated in the a position but are reductively eliminated in the 0 p ~ s i t i o n . [ ~The
~ , ~choice
~'
of R2 is
subject to the least number of restrictions. Branching in the
p position leads to reduced activities. If R2 and R3 are interchanged (i.e., if the E and Z isomers of substrates of the
type R3R2C=CHX are used), different enantiomers are
produced. Therefore, E / Z mixtures of substrates in which
the p-C atom becomes chiral on reduction cannot be employed. Complications can arise, however, if whole cells
are used, as we observed in the hydrogenation of ( E ) - and
(a-geraniate ( l a and 2a). As shown in Figure 1, l a yields
543
pure (R)-citronellate l b , but 2a is largely isomerized to the
thermodynamically more stable ( E ) isomer l a by clostridia
(presumably enzymatically) and thus also converted into
the ( R ) enantiomer. The ee value is lower because some of
2a is also reduced directly. Isolated enoate reductase must
be used in this case (see Section 3.6 and Fig. 1).
same rates, although, according to Table 3, 3a is reduced
approximately twelve times more rapidly by enoate reductase than 4a. Thus, the reduction of methyl viologen must
be rate-limiting for the hydrogenation of 3a. If 3a is hy-
COOQ
€
b)
I
0
4
+
Hz
+
~~200
as 'l a c t o n e )
Me
I
3b
3a
Me
H
O
COO'
S
HO
+
H,
__z
)-
COOQ
(isolated
as lactone)
H Me
lc
RR
/
SR
lb
Fig. I Bioreduction of ( E ) - and (a-geraniate ( l a and Za, respectively) to
( R ) - and (S)-citronellate ( l b and Zb, respectively). a) la is hydrogenated to
enantiomerically pure l b by C. La1 cells with a productivity number of approximately 3000. The (Z)compound Za is reduced much more slowly and
most of it is isomerized to l a prior to hydrogenation so that I b is formed
with ee values between 60 and 85%. Pure 2b can, however, be obtained if isolated enoate reductase is used for reduction. b) l b obtained after hydrogenation of l a with C. La1 cells was derivatized to l c and then separated o n 7 - 1
Lichrosorb columns by HPLC. (J. P. Lecomte et al., unpublished).
The high productivity numbers allow reactions to be carried out in small volumes. Products that are only chiral due
to the stereospecific replacement of ' H by 2 H o n one or
two methylene groups can therefore be obtained relatively
easily by means of
The cells are
freeze-dried and then used in 'H20 buffer. H2 can be used
as an electron donor since lo2-lo3 times more 'H is present in the buffer than ' H in the amount of hydrogen gas required. A preparative example has been described.[641
Some of the hydrogenations that have so far been carried out are summarized in Tables 2 and 3. All compounds
given in Table 3 had already been hydrogenated. It should
be noted that nitro groups are reduced by whole cells but
not by enoate reductase and NADH.'651Figure 2 shows the
time course of hydrogenation. Electron flow in the hydrogenation of a$-unsaturated carboxylates and other carbonyl compounds with C. La1 proceeds according to Route
V (Table 1). (Hydrqgenations without methyl viologen are
possible but usually proceed slower.) There are two alternatives. If the rate at which the unsaturated substrate
reacts with enoate reductase is greater than that at which
the hydrogenase ( E l ) can reduce methyl viologen, the hydrogenation mixture is not initially blue. The solution only
becomes blue due to reduced methyl viologen toward the
end of the hydrogenation when the substrate concentration
decreases and the enoate reductase is no longer saturated.
If, however, the unsaturated substrate reacts relatively
slowly so that the hydrogenase can reduce the methyl viologen rapidly enough, the solution remains blue throughout the entire reaction. Figure 2 shows that the hydrogenations of 3a and 4a with C. La1 proceeded at about the
544
4a
4b
drogenated with a mixture of C. La1 and C. kluyueri (the
latter displays a considerably higher hydrogenase activity),
the hydrogenation rate of 3a can be increased by a factor
of approximately 5 . In the case of 4a, this measure reduces
the rate of hydrogenation because, although C. kluyueri has
a higher hydrogenase activity than C. L a l , its enoate reductase activity is lower. As will be -demonstrated in Section 4.4, the conditions in an electrochemical cell are significantly different if the reduction of methyl viologen is
not rate-limiting.
1
2
3
k
tlhl
5
6
7
Fig. 2. Hydrogenation of a favorable (3a) and an unla\ordhlr ( 4 a ) aub\trate
of enoate reductase. 100 mM substrate was shaken at about 100 cycles per
minute with 400 mg of cells (wet sediment) in 3.0 mL of phosphate buffer under 1 atm H 2 at 35°C in a Warburg flask with a capacity of approximately 25
mL (manometer filled with Hg). 0 - 0 Hydrogenation of 3a with 250 mg of
C. La1 and 150 mg of C. kluyueri: 0 - 0 3a with only C. L a ] ; A - A 4a
with C. La!: 0 - U 4a with 250 mg of C. La! and 150 mg of C. kluyueri;
0-0 4a with only C. kluyueri. (The measurements were performed with C.
La1 cells that had been stored for 15 months and whose hydrogenase activity
was about half of the normal value).
Although the p H maximum of Reaction (e) occurs at
5.2,1'3Ja pH of between 6.5 and 7.0 has proved to be optimal for the overall hydrogenation reaction. The hydrogenation of the diketone 5a can also be carried out on a preparative scale-however, without methyl viologen (cf. Section 4.3).
5a
5b
Angew. Chem. Inr. Ed. Engl. 24 (1985) 539-553
In addition to a,P-unsaturated carboxylates and other
carbonyl compounds, the CC double bonds of ally1 alcohol derivatives can also be hydrogenated with C. La1 or C.
kluyveri cells. Enoate reductase does not, however, reduce
ally1 alcohol derivatives and the cells d o not possess any
other enzymes capable of catalyzing this reduction. It has
been shown that the ally1 alcohol derivatives are initially
dehydrogenated to the corresponding aldehydes, and the
CC double bonds of these compounds are then quickly reduced by enoate reductase. The aldehyde group is subsequently reduced back to an alcohol group. This is the reason why methyl viologen inhibits reduction in such cases;
the entire NAD is rapidly reduced to N A D H by Hz and
methyl viologen, and NAD is therefore not available for
dehydrogenation of the ally1 alcohol derivative. Ethanol is
a more suitable electron donor in these cases.139,401
3.2. Hydrogenation of Ketones
Ketones can also be reduced to chiral secondary alcohols with clostridia such as C. La1 or C. kluyveri [Reaction
(01.Examples are given in Table 4. Excellent ee values can
be obtained in a number of cases but low values are also
encountered.
Sih et al.1'.2'.271
have found that in the reduction of 4chloro-3-oxobutyrate esters with bakers' yeast the ee value
is a function of the size of the ester group. The methyl ester
is hydrogenated to the (S) enantiomer with a n ee value of
65%, whereas the octyl ester is converted to the ( R ) enantiomer with an ee value of 97%. For other alkyl residues of
the type (CH,),H (n=2-7), the ee values lie between
these two values. C. kluyveri hydrogenates the ethyl ester
of 4-chloro-3-oxobutyrate to the (R)enantiomer with an ee
value of >99% and an approximately 50-fold higher productivity number ( H . Simon et al. and B . Koppenhiifer et
al., unpublished). The enantiomeric purities of the phenylethyl and phenylpropyl alcohols obtained with C. kluyveri
d o not seem to have been reached yet by other means. In
other cases (e.g., phenylglycol) the ee values are poor, but
this has so far been seldom observed in clostridia. The very
low productivity numbers also seem to be more of an exception.
3.3. Hydrogenation of 2-Oxocarboxylates
A surprisingly wide range of 2-oxocarboxylate substrates can be reduced to (2R)-hydroxycarboxylates by Proteus mirabilis or Proteus vulgaris in the presence of methyl
or benzyl viologen with very high enantiomeric purities.[47.66.671
F7
R-C-R'
+
2 [HI
+
RCHOHR'
(f)
Table 4. Reduction of ketones with clostridia. If not otherwise stated, H L was
used as an electron donor.
Microorganism
Product
PN
Enantiomer
ee
[%I
H OH
,l.Jooet
C. kluyveri
1500
80
C. kluyveri
C. kluyveri
(2R,3S)
H OH
CIJ.,JOOEt
H
C. La1
C. La?
(S)
This reaction is also based on an unknown type of enzyme which we have partially characterized (see Section
3.6). Instead of HZ,formate can be used as an electron donor in combination with a viologen. Since viologens (Section 4) can also be easily regenerated electrochemically,
(2R)-hydroxycarboxylic acids may be obtained in three
ways (Scheme l), corresponding to Route V with
OH
x
x
x
x
CgH5
m-HOC,H,
C. La1
p-IIOC,H,
C. kluyveri
H
YII
H
OH
H
OH
c6%
H
C. kluyveri
C. kluvveri and
Candida utilis
OH
x/
CsH5
H
LOII
HO
94
C. kluyueri and
Enterobarter agglomerans
[a] Determined by enzymatic analysis. [b] A racemic mixture of ethyl 2-methyl-3-oxobutyrate was used. The reaction mixture consisted of 81% ethyl
(2R.3S)-3-hydroxy-2-methylbutyrate,
13% of the (2R.3R) form, 4% of the
(2S.3R) form, and 2% of the (2S.3S) form. [c] Analysis: B. Koppenhofer et
at., unpublished. Id] 10% ethanol was used as an electron donor.
Angew. Chem. Inr. Ed. Engl. 24 (1985) 539-553
Scheme 1. Proreus w l y u r r \ can be u\ed bith three dliiereni electron donors
to synthesize (2R)-hydroxycarboxylic acids. ViolaG =oxidized methyl o r
benzyl viologen; ViolGa =reduced viologen. The mediator can be reduced
either electrochemically o r by HZgas and a hydrogenase present in the microorganism or by formate and a formate dehydrogenase also found in the
microorganism.
DO,,,, = Hz o r HCOO@ and Route VI. Examples are the
reductions of 6a - 8a to 6b - 8b.
545
6a
6b
Table 5 provides a list of the substrates that can be reduced by Proteus mirabilis or Proteus vulgaris according to
Scheme 1. Glyoxylic acid is so far the only 2-oxocarboxylate that does not react with this enzyme.
+
c02
(iso1att.d
as lactonel
3.4. Hydrogenation with Two Microorganisms
7a
As described in more detail in Section 4.2, crude extracts
of C. kluyueri exhibit high activities for Reaction (h) in the
presence of 1-2 mM methyl viologen. Reaction (h) corresponds to part of Route 111.2 (Table 1). Yeasts, in particu-
7b
HZ+ NAD(P)'
8b
8a
Figure 3 shows the time course of the reduction of 4-methyl-2-oxopentanoate 8a to the corresponding (2R)-hydroxy acid 8b with formate and H2. As described in Section 4.4, productivity numbers of > 100000 can be obtained by determining and exploiting the kinetic parameters of the electrochemical reduction.
t [rnin]
-
t i p . 3. Kcduction 0 1 4-nicrli~l2-oxtipcnianoatc with P. i d g a r i s and H2 or
formate as an electron donor. Decrease in 4-methyl-2-oxopentanoateon reduction with H2 0 -- 0 -~0 and with formate 0 -- 0 -~0 and formation of
and with formate H - H - H.
the product with H2 0 0
~
~
Table 5. Relative rates for the reduction of 2-oxocarboxylates by a reductase
found in P. miruhilis and P. uulgaris.
Substrate
phenylpyruvate [a]
pyruvate [a]
2-oxobutanoate
indolylpyruvate
5-bemy lox yindolylpyruvate
3-fluoropyruvate
4-methyl-2-oxopentanoate
(S)-3-methyl-2-oxopentanoate
(R.S)-3-methyl-2-oxopentanoate
4-hydroxy-3,3-dimethyl-2-oxopentanoate
[a]
phenylglyoxylate [a]
2-oxononanoate
2-oxododecanoate
2-oxotridecdnoate
oxaloacetate [a]
2-oxoglutarate
2-oxoadipate
4-hydroxymethylphosphinyl(2-oxo)butanoate
Relative rate [ O h ]
P. mirabilis P. vulgaris
100
92
40
35
30
22
81
28
63
7
16
83
21
23
73
78
70
24
100
85
65
61
-
20
83
25
46
4
5
93
46
26
50
62
13
6
[a] The enantiomeric purity of the products of these substrates was investi.
gated by various methods. ( S ) enantiomers were not detected.
546
-
NAD(P)H
+ He
(h)
lar, demonstrate high reductase activities with certain carbony1 compounds, but the availability of NADH can be
limiting. It therefore seemed reasonable to combine lysates
from two species of cells; cell type 1 rapidly produces
NADH using H r (or another electron donor): the NADH
is then used by cell type 2 to reduce the unsaturated substrate. An example of this principle will be described in
Section 4.4. 2(R)-Propanediol can, for example, be obtained in this way with a productivity number of u p to
5700.[4"13-Hydroxy-3-methyl-2-butanone
can be hydrogenated to (S)-2,3-dihydroxy-2-methylbutanewith a productivity number of 3300 (Table 4) by combining C. kluyueri
and Enterobacter agglomerans. Aerobically cultured Enterobacter agglomerans cannot metabolize hydrogen, and C.
khyueri alone does not react with the ketone.
This method has the previously described advantages of
high productivity numbers and simple control of the reaction (manometrically). In addition to C. kluyueri, Alcatigenes eutrophus can, for example, also be used for producing NADH according to Reaction (h). We found that crude
extracts of cells that had been cultured according to Schlegel et a1.1681and then stored for 5 years at - 15°C still had
a specific activity for NADH formation of 1 U/mg protein.
Substrates for which the hydrogenase is rate-limiting (see
Section 3.1) can also be more rapidly hydrogenated by employing a combination of C. kluyueri and C. L a l .
3.5. Hydrogenations with Immobilized Cells
According to our experience,[691clostridia can readily be
immobilized with acrylic acid derivatives. These so-called
photopolymerizable prepolymers introduced by Tanaka et
al.[701do not noticeably damage the cells. The observed decrease in activity can be accounted for solely by inhibition
of diffusion. We have performed hydrogenations at room
temperature for 63 d with immobilized C. Lal cells in the
presence of 2 mM methyl viologen. After 12 d the rate of
Hz uptake was still 50% of the starting rate. These immobilized preparations are highly suitable for storage. C. kluyueri stored in tris buffer with 50% glycerol did not show
any decrease in activity after 150 d ; under the same conditions C. La1 cells lost 40% of their activity.[691Purified
enoate reductase can also be immobilized with photopolymerizable prepolymers. These substances are not, however,
suitable for the immobilization of Proteus mirabilis, Proteus
vulgaris, or the 2-oxocarboxylate reductases obtained from
Angew. Chem. Int. Ed. Engl. 24 (1985)539-553
these microorganisms. Ionotropic gels, such as alginates
cross-linked with calcium ions, can be employed for these
bio~atalysts.'~''Since gels composed of photopolymerizable prepolymers are not very suitable for electrochemical
cells in which rapid stirring is necessary, we have coated
filter paper with alginate films containing immobilized enzymes. These systems have proved to be of value both for
cells and for more or less purified enzymes. We have been
able to increase the half-life of enoate reductase from approximately 6 h to 150 h in rapidly stirred electrochemical
cells by using this approach.'611
3.6. Kinetic Data for Enoate and 2-Oxocarboxylate
Reductases
The pH optimum of enoate reductase with reduced methyl viologen is 5.2;[13]at p H 4.5 and 6.2 the rate is twothirds that of the maximal value. A p H value of less than
5.2 should be avoided in an electrochemical cell. On the
other hand, p H values of u p to 8.0 d o not seem to have a
detrimental effect o n the stability of the enzyme. A p H
value of between 6.0 and 6.2 is recommended as a compromise between the stability of methyl viologen and the reaction rate of enoate reductase. A number of V,,, and K ,
values for substrates of enoate reductase are given in Table
3 . The K , values for reduced methyl viologen and NADH
are 0.4 mM and 0.012 mM, respectively. A concentration of
approximately 2 mM MVeo thus seems appropriate for a n
optimal rate of Reaction (e). The K , values for reduced and
oxidized methyl viologen are so high that the rate is not affected when the total methyl viologen concentration is 34 mM. The K , values for enoates vary over a range of two
orders of magnitude. The smaller the K , value, the later
will the reaction rate be limited by the falling substrate
concentration. The K , values for aliphatic enoates and the
reaction products are high (e.g., ca. 500 mM); hence, they
are suitable for preparative purposes. The K , values of aromatic products are, however, lower; the value for 3-phenylpropionate (15 mM) is the lowest that we have encountered so far. We have already reported in detail o n the
mechanisms and kinetics of the reductions of enoates with
NADH.I5'] The question arises, however, whether some of
the values reported therein are different in the case of reduction with MV@O
' , as is found, for example, for the p H
dependency of the reaction rate.
The 2-oxocarboxylate reductase from Proteus vulgaris
displays a broad p H optimum with reduced methyl viologen. The maximum activity occurs at pH 7.0; the activity is
still 80% of the maximal value at pH 6.0 and 8.0. Some typical K , and K , values are compiled in Table 6. The dependence of the reduction rate of 2-oxocarboxylates on the
concentration of methyl viologen does not exhibit Michaelis-Menten behavior. The rate of reduction of a 2-oxocarboxylate increases with decreasing concentration of reduced methyl viologen until a concentration of < 0.05 mM
is reached! The behavior at lower concentrations cannot be
followed due to experimental difficulties. MichaelisMenten behavior was observed in the determination of the
K , values of various 2-oxocarboxylates. The K , values
vary greatly for different substrates but are suitable for
preparative purposes. The K , values of aliphatic substrates
and products are favorable. The value for phenylpyruvate
is relatively low but that of the product is again suitable. In
practice this is of little importance. For instance, 100 mM
solutions of phenylpyruvate can be reacted in the electrochemical cell with productivity numbers of > 100000. The
surprising behavior of 2-oxocarboxylate reductase toward
oxidized methyl viologen plays an important part here. As
shown in Table 7, oxidized methyl viologen is a positive effector-i.e., depending on the concentration of the reduced methyl viologen, the enzyme activity can be augmented by a factor of three to six by adding oxidized methyl viologen. The MV@''/MV@@ratio can easily be adjusted in an electrochemical cell by means of the cathode
potential, thus allowing very high enzyme activities to be
achieved.
Table 7. Influence of oxidized methyl viologen on the rate of reduction of a
2-oxocarboxylate.
0. I4 mM
MV"
with increasing
concentrations
of MVea
[mMl
Ratio
MVQo/
MV'"
Relative
rate
0.02
0.07
0.14
0.7
7
2
1
0.2
18.1
41.9
46.0
62.0
1.6 mM
MVeo
with increasing
concentrations
of MV"*
[mM1
0.02
0.8
1.6
8.0
17.8
Ratio
MVe"/
Relative rate
MV"'
80
2
1
0.2
0.09
10.2 [a]
29.3
33.7
53. I
61.5
[a] The different rates obtained with the same MV"/MV""
ratio are presumably due to the inhibitory effect of high MVao concentrations.
Table 6. Some typical K , and K , values for 2-oxocarboxylate reductase.
Substrate
K , [ m ~ ] Substrate/products
reduced methyl viologen
reduced benzyl viologen
phenylpyruvate
4-methyl-2-oxopentanoate
4-hydroxy-3,3-dimethyl-2
oxobutanoate
2-oxoglutarate
<0.05 [a]
0.1
0.15
2.1
7.5
2.5
reduced benzyl viologen
benzyl viologen
(R)-phenyllactate
(R)-Z-hydroxy-Cmethylpentanoate
4-hydroxy-3,3-dimethyl-2-oxobutanoate
phenylpyruvate
4-methyl-2-oxopentanoate
K , [mM]
4. Electromicrobial Reductions
0.7 [b]
10.0
100
100
50
12
70
[a] The hydrogenase of Proteus uulgaris has a K , value of approximately 0.3
mM for reduced methyl viologen. &] Concentration at which the reaction rate
is half that found with 0.1 mM reduced viologen.
Angew. Chem Int. Ed. Engl. 24 (1985) 539-553
4.1. Principles
The enoate reductase of C. La1 and 2-oxocarboxylate
reductase are enzymes that d o not have to or cannot accept
electrons from N A D H or NADPH, but can accept them
from reduced viologens. Since viologens can be reduced in
electrochemical cells, unsaturated substrates can be stereospecifically reduced electrochemically with the aid of
microorganisms containing enzymes such as enoate reductase o r 2-oxocarboxylate reductase. Only one enzyme is
547
then required. The electron flow is described by Route
VI .
We have also discovered that a large number of microorganisms catalyze Reactions (i) and (j) (see Section
4.2) :144.721
-
2MVS0+He+NADe
2 MV@'
-2MVQe+NADH
+ He + NADP'
2 MVGe
(i)
+ NADPH
ti)
Reductions can thus be performed electrochemically
with a variety of different microorganisms. Figure 4 shows
a current-time curve for such a reaction. The reactions that
occur are described by Equations (k), (i), and (l), which
can be added together to give Equation (m):1721
2 M V e e+ 2 e e
Ha
electrochemical
t
2MVG0
+ NADH + CH,COCH20H
NADe
CH,COCH,OH
+ 2H0 + 2eQ
--t
+ CH3CHOHCHzOH
(I)
(4
CH,CHOHCHZOH
Table 8. Specific activities of methyl viologen-dependent NAD(P) reductases
in crude extracts of various microorganisms (room temperature if not otherwise stated).
Microorganism
Source [a]
gmol
min x mg protein
NAD NADP
DSM
555
DSM 1460
DSM
663
ATCC 25755
DSM 1731
DSM
525
ATCC25522
ATCC 638
ATCC 9689
ATCC 25757
ATCC 25772
ATCC25761
ATCC 25647
ATCC 25784
ATCC 9714
ATCC 3584
DSM
519
ATCC 14940
ATCC 14963
DSM 20357
ATCC 25085
DSM 20402
DSM 2 133 [b]
DSM 2 161 [c]
DSM 2078 [c]
DSM 2476 [c]
16.6
3.3
2.4
0.31
3.3
0.55
1.2
0.79
0.77 0.81
0.17 0.05
1.02 0.04
1.56 1.1
0.13 0.13
2.7
0.46
1.2
0.25
0.69 0.23
1.03 0.38
0.55 0.60
1.96 1.5
4.62 0.61
0.29 0.79
0.47 0.045
0.54 3.2
0.29 0.06
1.3
0.24
1.14 1.03
0.33 0.25
0.86 23.5
0.46 1.0
0.33 4.6
DSM
DSM
DSM
DSM
0.18
0.066
0.040
0.10
0.018
0.16
0.022
0.016
0.1
0.80
0.13
0.013
0.04
0.12
0.09
0.095
0.086
0.12
0.008
0.011
0.001
0.04
0.027
Anaerobic organisms
Clostridiurn kluyueri
Clostridium La I
Clostridrum tyrobut.vricum
Clortridiurn fyroburyricurn
Clostridium acetoburylicum
Clostridium pasteuriarium
Closrridium propionicum
Clomidiurn hi/ermenrans
Clostrrdiurn difficile
Closrridium ghoni
Closfridium hastiforme
Clostridium mangenoti
Clostridium oceanicum
Closfrrdiumputrificurn
Clostridium sordellii
Clostridium sporogenes
Clostridiurn strcklandi
Cloc.tridiurn rym biosum
Peptococcus aerogenes
Peptostreptococcus anaerohius
Acidaminococcus fermentans
Eubacterium limosum
Methanobacterium thermoautotrophicum
Desulfurococcus mobilis
Thermoproteus tenax
Thermococcus ceier
Aerobic organisms
Candida ufilis
Candida boidinii
Candida ualida
Rhodotomla glutinis
Bakers' yeas!
Georrichum candidum
Bacillus cereus
Bacillus macerans
Bacillus spec.
Alcaligenes eutrophus H 16
Aeromonas hydrophila
Enterobacter agglomerans
Escherichia coli
Klebsiella aerogenes
Proteus mirahilis
Pseudomonas fluorescens
Pseudornonas aeruginosa
Micrococcus luteus
Staphylococcus aureus
Staphylococcus carnosus
Lactobacillus casei
Propionihacterium penrosaceum
Sulfolobus solfataricus
L
1
2
3
4
5
6
7
8
9
t[hl
Fig. 4. Hydroxyacetone (40 mmol), methyl viologen (1.2 mrnolj, and NAD
(0.1 mmol) were reduced in 200 mL of potassium phosphate buffer solution
(0.1 M , pH 7.0). Candida urilis (175 mg, dry weight) was added at B and C .
kluyueri (7.6 mg, dry weight) at C at - 792 mV vs. the saturated calomel electrode. A current of 93 mA was obtained after adding a further 350 mg of Candida utilis, 3.8 mg of C . kluyueri, and 0.1 mmol of NAD. After a total of
140 h, 33 mmol of propanediol had formed and 2 mmol of hydroxyacetone
was still present.
70167
70026
70178
70389
DSM 1240
DSM
31
DSM
24
DSM
406
DSM
428
ATCC 13 137
NCTC 9381
ATCC 10536
DSM 30102
DSM 30115
DSM 50090
DSM 50071
DSM 20030
DSM 20231
DSM 20501
ATCC 7469
DSM 20272
DSM 1616
0.023
0.013
0.011
0.007
0.012
0.10
0.059
0.076
~
0.026
0.071
0.016
0.01
0.07
0.03
0.025
0.010
0.010
0.011
0.015
0.001
0.008
0.21
(2 R)-l,2-Propanediol can be obtained by this meth~ d [ ~ with
~ , a~ 20~ to
. ~50-fold
~ l higher productivity number
than that found for a procedure published in Organic Synthesis.[741
[a]
DSM = Deutsche
Sammlung
Mikroorganismen,
Gottingen.
ATCC = American Type Culture Collection, Rockville, MD, USA.
NCTC= National Collection of Type Cultures, London, England. [b] Determined manometrically by measuring the H2 consumption. [c] Test at 80°C.
4.2. Types, Sources, and Applications of
Methyl Viologen-Dependent NAD and NADP Reductases
for NAD(P)H Regeneration
Four groups of microorganisms can be distinguished in
Table 8 :
1. Anaerobic eubacteria, in particular clostridia: These
organisms contain NAD(P) reductases with high activities.
If they contain a hydrogenase, the formation of hydrogen
[Reaction (n)] may compete with Reactions (i) and/or Q).
Table 8 shows that many microorganisms catalyze Reactions (i) and (j).(44.471
These activities are of interest for the
regeneration of NADH and NADPH, although their values vary by several orders of magnitude in different organisms. In order to evaluate these numbers it should be remembered that 0.1 U/mg protein corresponds to a productivity number of 3000 (see Section 2.1).
548
C. kluyueri is particularly advantageous at room temperature. This organism exhibits very high NAD(P) reductase
Angew. Chem. I n t . Ed. Engl. 24 (19881 839-583
activities and the hydrogenase activity only amounts to
about 10% of the value for the NAD reductases.
2. Archaebacteria: Methanogenic bacteria oxidize reduced methyl viologen mainly with production of H2. The
three thermophilic, sulfur-reducing archaebacteria that
have so far been tested contain surprisingly high levels of
N A D P reductase activity. In two cases these activities are
one order of magnitude higher than that for NADH formation.
3 . Yeasts: They display lower NAD(P) reductase activities but d o not contain any hydrogenases, which is of advantage for electromicrobial reductions.
4. Aerobic eubacteria: The majority of the aerobic eubacteria tested contain lower activities than the anaerobic
bacteria.
The nature of the methyl viologen-dependent NAD(P)
reductases is not always clear. A number of NAD(P)H-dependent flavoproteins catalyze these reactions. Diaphorase
has already been used in practice.[541In addition, Reactions (i) and (j) are known to take place with the hydrogenase from Alcaligenes eutrophu~1~~I
and plant ferredoxinNADP oxidoreducta~e,['~~
respectively. Reactions (i) and
(j) do not, however, seem to have been studied quantitatively so far. The measured activities of a series of enzymes
for Reactions (i) and G) are given in Table 9 and compared
with the physiological activities of these enzymes. The following enzymes, which might have been expected to catalyze Reactions (i) and 6 ) o n account of their structure, display low activities and are of little interest for practical
purposes : glutathione reductase from yeast, NAD( P)H
F M N reductase from Photobacterium fisheri and N A D H
dehydrogenase from the thermophilic bacterium Bacillus
spec DSM 406. A very stable, methyl viologen-dependent
N A D reductase can, however, be obtained from the latter
microorganism; its activity as a n NADH dehydrogenase
with 2,6-dichlorophenolindophenolas an acceptor is low.
This clearly shows that NADH dehydrogenase activity
does not necessarily parallel the activity as a methyl viologen-dependent N A D reductase. The activity of the methyl viologen-dependent N A D reductase from Bacillus
spec. DSM 406 remains practically constant over 50 d at
35°C in a dilute solution.
In all entries in Table 9 in which the stereochemistry of
[4-'H]NAD(P)H formation is not accompanied by a literature reference, we prepared NADH o r NADPH in an electrochemical cell using 2 H Z 0buffer and determined which
enantiomer was produced by means of N M R spectroscopy.[441The method used to determine the stereochemistry
has already been described."81
4.3. Construction of an Electrochemical Cell
Cathodes with a high overvoltage for hydrogen formation must be used for electrochemical reductions in water.
Only the mediator and not the substrate must be electrochemically reduced; this can be checked by means of cyclovoltammetry.'8'1 Moreover, the mediator must not react
spontaneously with the substrate. So far we have observed
spontaneous reactions between reduced methyl viologen
and the following substrates: phenylglyoxylate, ethyl 4chloro-3-oxobutyrate, and 5a. Phenylglyoxylate can be
reacted in the presence of benzyl viologen; the two other
compounds are reacted with H2 and C. kluyveri without a
mediator.
Gold, silver, carbon, and particularly mercury are suitable for use as cathode materials when working in the neu-
Table 9. Enzymes that have been tested for the reduction of NAD(P) with reduced methyl viologen. Initial rates for Reactions (i) and fi) and comparison of this activity with the physiological reduction or dehydrogenation of NAD(P)H in the presence of a n artificial acceptor. All specific activities are for two electron transfers
(pmol/mg protein x min). If not otherwise stated the tests were performed at pH 7.0 and room temperature in 0.1 M tris acetate buffer. The NAD(P) concentration
was 2.5 mM, that of MVao was 1.8 mM. In order to determine the stereochemistry, NAD(P) was reacted with the relevant enzyme in a n electrochemical cell (Fig.
5a) in ' H 2 0 buffer [44].
Enzyme
(EC number)
Physiol. reaction or NAD(P)H
dehydrogenase reaction
Specific
activity
lipoamide reductase (diaphorase) (1.6.4.3)
NADH/lipoamide
62
diaphorase from microorganisms (Boehringer
Mannheim)
NADH/p-iodonitrotetrarolium
chloride
NADPH/GSSG
NADPH/cytochrome c
glutathione reductase from yeast (1.6.4.2)
NADPH-cytochrome P450 reductase from rat
liver (1.6.2.4) [a]
0.6
98
6.5
hydrogenase rrom Alcoligenes eutrophus
( I . 12. I .2)
H2/NAD
1.5 [b]
NADH-cytochrome c reductase from porcine
heart (1.6.99.3)
NADH/cytochrome c
0.16
NAD(P)H-FMN reductase from Photobacteriurnfirheri ( I 6.8.1)
NADH/FMN
10
xanthine oxidase (123.2)
enoate reductase (1.3.1.31)
NADH/hypoxanthine
NADH/(E)-2-methylbutenoate
NADH dehydrogenase from Bacillus spec.
DSM 406 (membrane-bound) [c]
NADH/dichlorophenolindophenol
NADH dehydrogenase from Bacillus spec.
DSM 406 (soluble) [c]
NADH/dichlorophenolindophenol
MV@'/NAD
46
1.1
0.02
1.2
70
MV"/NADP
Stereochemistry of
[4-2H]NAD(P)H [ref.]
0.15
0.03
s (791
0.06
-
-
15.2
R
-
S [Sol
0.22
0.001
S
0.1
0. I
-
0.04
7
0.02
-
12
0.01
S
61
1.5
0.16
8.8
44
~
S
[a] We are grateful to Prof. V. Ulrich. Constance, for this preparation. [b] The preparation had been stored for about 1 year. [c] Enzymes were isolated by S . Nagata
in our laboratory (unpublished).
Anqew. Chrm Int. Ed. En&. 24 (1985) 539-553
549
tral pH range and with potentials of approximately
-450 mV versus the saturated calomel electrode (all potentials in this review are given versus this electrode). In
the case of more negative potentials (e.g., -640 V), lead,
bismuth, antimony, and tin cathodes are suitable. The potential is fixed in a standard three-electrode arrangement
(Fig. 5). If cathode surface areas of <250 cm2 are used,
200-mA potentiostats with a maximal terminal voltage of
15 volts are suitable.
H
4.4. Practical Aspects of Performing Electromicrobial
and Electroenzymatic Reductions
The optimal concentrations of oxidized and reduced methyl viologen, as well as the substrate concentration, depend on the kinetic parameters of the enzymes. For instance, enoate reductase and 2-oxocarboxylate reductase
differ considerably in this respect (see Section 3.6). This
also applies if cells or cell extracts are used that contain
these enzymes. In the case of enoate reductase, 2-3 mM
reduced methyl viologen should be present; the concentration of oxidized methyl viologen does not affect this enzyme.
- 40
c
Fig. 5. Different e!ectrochemicdl cells. a) 50-mL three-necked flask; b) 200mL five-necked flask (two necks not shown); c) 70-mL cell with multiring
carbon electrodes. A, platinum electrode; B, Luggin-Haher capillary for the
reference electrode (saturated calomel electrode); C , mercury cathode in a)
and b); D, diaphragm; E, stirring rod; F, cathode contact: G, glass vessel, in
c) it can he thermostated; H, opening for taking samples and gassing with
N,: I, device for holding the cathode; J, device for preventing I from rotating; K, PVC lid and anode chamber.
Nafion membranes with a thickness of 0.5 mm have
proved to be suitable for use as diaphragms in cells of over
50 mL. Vycor tips are also suitable for smaller cells (Fig.
5a). The diaphragm must satisfy the following requirements: it must allow proton transfer from the anolyte to
the catholyte in which the protons are consumed (e.g., in
Reaction (m)); it must be sufficiently conductive and prevent oxygen from passing from the anolyte to the catholyte. Oxygen reacts with reduced methyl viologen at a diffusion-limited rate. The superoxide radical anion
f o ~ m e d [ ~and
* . ~its~subsequent
~
products can react with the
enzymes, substrates, products, and mediator. Furthermore,
the presence of oxygen results in an apparent rate of enzyme reaction that is too high. Potential differences of
> f 3 0 mV should be avoided by means of proper design
of the symmetry of the cathode.
Cells with mercury cathodes are especially easy to handle in the laboratory. A current of approximately 1 mA can
be produced per cm2 mercury surface area in solutions
which contain 3 mM oxidized methyl viologen when the
cell is in the stationary operational state. A current of
3.2 mA reduces 1 ymol of substrate per minute with two
reduction equivalents.
Since it is easy to accomodate 102-104 U of enzyme activity in 100 mL of catholyte, especially if the enzyme is enriched, the preparative capacity of the operating cells is
usually limited by the reduction of methyl viologen. The
cells should therefore be operated under diffusion-controlled conditions. The diffusion layer on the cathode
should be kept as thin as possible by, for example, efficient
stirring (magnetic stirrer) or rotation of the cathode; this
also results in thorough mixing of the reaction solution.
550
‘B
24
12
LB
t [hl
-10
96
Fig. 6. Time courie 01 the electromicrobial reduction of 30 mlri (L)-2-lluorocinnarnate and 30 mM 4a by 122 mg of C. L a ) and 184 mg of C . L a ] , respectively, in the cell shown in Figure 5c. A - A , Decrease in fluorocinnamate;
0 - 0, formation of (2R)-2-fluoro-3-phenyIpropionate;
A, the observed current-time course in this reaction. A A , decrease in 4a : 0 0 , formation
of 4b; B, current-time course in this reaction. After 50 h, a further 70 mg of
C . La1 was added. The methyl viologen concentration was 3 mM, T=28”C,
36 mL of 100 mbi phosphate buffer (pH 6.3), cell dry weights.
~
~
Figure 6 shows the reduction of (Z)-2-fluorocinnamate
and of 4a in an electrochemical cell (Fig. 5c) with C. Lal.
(Z)-2-Fluorocinnamate has a higher V,,,,,, value than 4a
and a much smaller K , value (Table 3). This has the following consequences : (Z)-2-fluorocinnamate is reduced
considerably faster than 4a and the current yield is appreciably higher; this applies particularly to the reaction of
the last third of the substrate and is mainly a result of the
approximately 100-fold more favorable K , value of
fluorocinnamate. H2 is produced from reduced methyl viologen (Reaction (n)) by the hydrogenase present and this
reaction competes with the reduction of both substrates.
Fluorocinnamate is reduced with a current yield of approximately 50%, whereas the current yield for the reduction of 4a is ca. 10% during the first 5 h, ca. 7% after 24 h,
and merely 4.5% after 72 h. The current is therefore mainly
used for H2 formation.
The current-time course during the reduction of phenylpyruvate with Proteus vulgaris is shown in Figure 7. The
reaction is started by adding oxidized methyl viologen.
The potential is adjusted in such a way that the optimal ratios given in Table 7 are obtained. A solution containing
<0.1 mM reduced methyl viologen is sky blue. The subsequent current-time course can be qualitatively interpreted
from the kinetic data given in Table 6. Phenylpyruvate
causes substrate inhibition and has a low K , value. As the
Anyew. Chem. In,. Ed. Engl. 24 f798S) 539-883
501
25
I
014
1
L
. . . .
,
3
5
7
. .
.
-
9
f[hl
1
.
.
.
.
1
Fig. 7. Current-time course during the electromicrobial reduction of 7a to
7b. For further details see Experimental Procedure.
total concentration decreases, the enzyme activity and thus
the current increase. Because of the K , value, the reaction
rate only begins to fall when substrate conversion is
>98%, but then decreases very sharply. The solution becomes dark blue within one minute. The reason for this is
that the steady state of the consecutive reaction sequence-the electrochemical reduction of methyl viologen
and its reoxidation by the enzyme-collapses due to lack
of substrate. The current does not fall to zero because Proteus vulgaris possesses hydrogenase. This reaction is of no
importance until the phenylpyruvate is exhausted because
the hydrogenase has the relatively large K , value of
0.3 mM for reduced methyl viologen and u p to this point
the concentration of reduced methyl viologen is
<0.05 mM. The product of current and time between the
current increase and decrease shown in Figure 7 corresponds to the quantity of phenyllactate formed.
Typical Experimental Procedures
The cultivation of C. L a l , C. kluyueri, P. mirabilis. and P. vulgaris has
been described 157, 66, 671. After the stationary growth phase has been
reached, the cells are centrifuged off at ca. 5000 g and stored at approximately - 15°C under exclusion of air without being washed. The dry weight
of the sediment amounts to about 20% of the wet weight. Hydrogenations are
performed in 0.1 M potassium phosphate buffer (pH 7.0) under normal pressure in an Hz atmosphere with shaking (approximately 200 cycles per min) at
35°C and carried out in the presence of 0.4 mg tetracycline/lO m L buffer to
prevent growth of other microorganisms. All experiments are performed under strict exclusion of oxygen. The H 2 consumption is read from a Warhurg
manometer that is filled with mercury a n d connected to the hydrogenation
vessel. Quantities of cells are expressed as dry weights.
I. Hydrogenation of l a to I b with C. L a l : The substrate (0.70 mmol) was
completely hydrogenated in 4.5 h in 26 mL of buffer containing the cells
(150 mg) and 1 mM methyl viologen (see Fig. I).
2. Hydrogenation of 3a, 4a, and 5a: C. La1 cells (1.6 g), methyl viologen dihydrochloride (12 mg), and 3a (6.0 mmol as the sodium salt) in 40 m L of buffer were hydrogenated for ca. 3-4 h in a shaking flask with a capacity of approximately 200 mL which was connected to a 2.5-L H2 reservoir. 3b was isolated in the form of a lactone: no contaminants could he detected by gas
chromatography after distillation. 4b was prepared from 4a in an analogous
fashion. The enantioselectivity of both reactions was >96%. 5a (6 mmol) was
hydrogenated with C. khyueri cells (1.6 g) in 52 m L of buffer without methyl
viologen in 3.5 h (er>96%). (We are grateful to Dr. H . G. Leuenberger and
Dr. M . Schmrd. Hoffmann-La Roche, Basel, for the analyses.)
3. Reduction of 6a to 6b with Proteus vulgaris and formate: a mixture (90
mL) containing 85 mM phosphate buffer, 33.3 mM 6a (as the sodium salt),
65 mM sodium formate, Proteus vulgaris (600 mg), and 1.0 m M henzyl viologen was shaken for 21 h at 35°C under N'. The resulting solution contained 32 mM 6b; no 6a could he detected. 6b was isolated a s the pantolactone. [a]$ - 5 I . I ( H 2 0 , c = 30 mg/mL). ee>99.5% (determined by gas chromatography). (We are grateful to BASF AG for this analysis).
4. Electromicrobial reduction of 7a to 7b: 0.1 M potassium phosphate buffer
(200 mL, pH 7.0) containing 100 mM 7a was flushed free of oxygen with N Z
in the cathode compartment or an electrochemical cell (Fig. 5b) and mixed
with a suspension of Profeus uulgarb (15 mg dry weight). The cathode potential was set to -900 mV with respect to a saturated calomel electrode. A current of 70 mA flowed after addition of sufficient methyl viologen to give a 6
Angew Chrm Int. Ed Engl. 24 (1985) 539-553
mM solution. The maximum current (100 rnA) was obtained shortly before
the end of the reaction, and the subsequent drop in current showed that the
substrate was totally consumed (Fig. 7). The quantity of charge taken up at
this point corresponded to a conversion of 99%. No remaining substrate
could be detected by HPLC. The productivity number was 117000.
5. Electromicrobial reduction of 8b to 8a : The cathode compartment of the
electrochemical cell contained 0.1 M potassium phosphate buffer (41 mL, pH
7.0) and 0.6 M 8a. The diaphragm consisted of a 3.6-cm' Nafion cation exchange membrane; 10 m L of the above phosphate buffer served as an anolyte. Oxygen dissolved in the catholyte was removed by flushing with N2.
Proteus vulgaris (15.1 mg) was then added and the cathode potential set to
- 760 mV. After addition of sufficient methyl viologen to give a 6 m v solution and stirring at a rate of 400-500 rpm, a current of 120 mA was obtained. According to HPLC, 96% of the expected quantity of 8b was formed
after 12 h. [a]?; 11.1 (H'O, c=20.7 mg/mL). If the concentration dependence is taken into consideration, this corresponds to the reported values 184,
851. The productivity number was approximately I25 000.
+
5. Outlook
The methodology of biocatalytic, enantioselective reductions of unsaturated compounds has undergone considerable development over the past few years. A number of efficient procedures for NADH regeneration are now available. The groups led by white side^[^^] and
have synthesized a variety of compounds on a laboratory scale using NADH regeneration. Amino acids can even be prepared on a commercial scale with kg prices of approximately 50 DM as a result of the studies performed by the
groups of Kula. Wandrey, and L e ~ c h t e n b e r g e r . ~ ~This
~.'~'
new battery of methods is, however, still used relatively
seldom in preparative organic chemistry.
It can be assumed that further microorganisms and reductases will be found for reductions. These methods
must, however, compete with other developments: namely,
synthetic chiral hydrogenation catalysts and enzymes other
than reductases. The latter include esterases that have recently often been employed to selectively cleave the esters
of secondary alcohol^;^^^^"^ also noteworthy are the syntheses of chiral secondary alcohols by biocatalytic condensation reactions.['"' Syntheses in which a center of chirality
obtained by biocatalysis is utilized in subsequent sterically
controlled reactions will probably also be developed further. For instance, Reetd"] recently demonstrated the possibilities that are opened u p by addition reactions with chiral a- and p-alkoxycarbonyl compounds.
We assume that many applications will also be found for
the microbial hydrogenation and electromicrobial and
electroenzymatic reductions that we have investigated during the past few years. It would also be interesting if other
reductases such as enoate and 2-oxocarboxylate reductase
could be discovered that are highly stereoselective, can
react with a wide spectrum of substrates, and can also accept electrons from artificial mediators which can be electrochemically regenerated. We believe electrochemical regeneration, e.g., of reduced methyl viologen, to be superior
in many respects to that obtained with the system light/
[Ru(bpy),]/ethylenediaminetetraacetic acid.[921
The results described in this report were obtained with the
help of a large number of people and institutions. In addition
to the persons mentioned in the text and references, we are
especially grateful to Dr. P. Rauschenbach f o r providing appropriate HPLC methods, to the technicians R . Feicht, H.
Leichmann, L. Riesinger, C. Stuber, and F. Wendling for
55 1
pare the manuscript. We would also like to thank BASF, the
Fonds der Chemischen Industrie, and especially the
Deutsche Forschungsgerneinschaft ("Sonderforschungsbereich 145, Biokonversion'7 for theirjinancial support.
Received: December 10, 1984 [A 538 IE]
German version: Angew. Chem. 97 (1985) 541
Translated by Dr. Gail Schulz, Seeheim-Jugenheim
[ I ] C. J. Sih, C.-S. Chen, Angew. Chem. 96 (1984) 556; Angew. Chem. fnt.
Ed. Engl. 23 (1984) 570.
121 Type Reactions in Fermentation Chemistry. U S Department of Agriculture, Fermentation Laboratory Northern Utilization Research and Development Division, Peoria, IL, USA 1959.
[3] C . J. Sih, J. P. Rosazza in J. B. Jones, C. J. Sih, D. Perlman: Applications
of Biochemical Systems in Organic Chemistry, Part I , Wiley, New York
1976, chap. 3.
[4] J. B. Jones, J. F. Beck in J. B. Jones, C. J. Sih, D. Perlman: Applications
of Biochemical Systems in Organic Chemrsrry, Part I . Wiley, New York
1976, chap. 4.
151 K. Kieslich: Microbial Transformations of Nan-Steroid Cyclrc Compounds, Thieme, Stuttgart 1976.
[6] H. lizuka, A. Naito: Microbial Conuersion of Steroids and Alkaloids,
Springer, Tokyo 1981.
[7] H. G. W. Leuenherger in H. J. Rehm, G. Reed: Biotechnology, Vol. 6a
(Microbial Tranformation), Verlag Chemie, Weinheim 1984, p. 5.
[S] J. Mulzer, Nachr. Chem. Tech. Lab. 32 (1984) 520.
[9] 0. Neuhauer, K. Fromherz, Hoppe Seyleri Z . Physiol. Chem. 70 (191 1)
326.
[lo] W. Hummel, H. Schiitte, M.-R. Kula, Ann. N . Y. Acad. Sci.. in press.
[ I l l J. B. Jones in: Enzymes in Organic Synthesis (Ciba Found. Symp. 111).
Pitman, London 1985, p. 3.
1121 H. Simon, H. Giinther in 2. 1. Yoshida, N. Ise: Biomimetic Chemistry.
Kodansha, Tokyo/Elsevier, Amsterdam 1983, p. 207.
1131 H. Simon, J. Bader, H. Giinther, S. Neumann, J. Thanos, Ann. N. Y.
Acad. Sci. 434 (1984) 171.
1141 R. Bentley: Molecular Asymmetry in Biology Vol. I and 2. Academic
Press, New York 1970/71.
[IS] R. Lagunas, C. Dominguez, A. Busturia, M. J. Saez, J . Bacteriol. I52
(1982) 19.
1161 R. MacLeod, H. Prosser, L. Fikentscher, J. Lanyi, H. S. Mosher, Biochemistry 3 (1964) 838.
1171 M. Bucciarelli, A. Forni, I. Moretti, G. Torre, Synthesis 1983, 897.
[IS] B. S. Deol, D. D. Ridley, G. W. Simpson, Aust. J . Chem. 29 (1976)
2459.
1191 D. W. Brooks, P. G. Grothaus, W. L. Irvin, J . Org. Chem. 47 (1982)
2820.
1201 T. Kitazume, N. Ishikawa, Chem. Letr. 1983. 237.
1211 B. Zhou, A. S. Gopalan, F. Van Middlesworth, W. R. Shieh, C. J. Sih, J .
A m . Chern. Soc. I05 (1983) 5925.
(221 B. Wipf, E. Kupfer, R. Bertazzi, H. G. W. Leuenberger, Helv. Chim. Acta
66 (1983) 485.
1231 H. G. W. Leuenherger, W. Boguth, R. Barner, M. Schmid, R. Zell, Helu.
Chim. Acta 62 (1979) 455.
(241 H. G. W. Leuenberger, W. Boguth, E. Widmer, R. Zell, Helu. Chim. Acra
59 (1976) 1832.
1251 K. Mori, Tetrahedron 37 (1981) 1341; and references cited therein.
1261 D. Seehach, M. F. Zuger, F. Giovannini, B. Sonnleitner, A. Fiechter, Angew. Chem. 96 (1984) 155: Angew. Chem. Int. Ed. Engl. 23 (1984) 151.
(271 C. J. Sih, B. N. Zhou, A. S. Gopalan, W. R. Shieh, F. Van Middlesworth
in W. Bartmann, B. M. Trost: Selectivity - a Goalfor Synthetic Efficiency
(Workshop Conferences Hoechst, Vol. 14). Verlag Chemie, Weinheim
1983, p. 251.
(281 E. Hochuli, K. E. Taylor, H. Dutler, Eur. J . Biochem. 75 (1977) 433.
1291 Y. Takaishi, Y. L. Yang, D. DiTullio, C. J. Sih, Tetrahedron Lett. 23
( 1982) 5489.
1301 W. Tischer, W. Tiemeyer, H. Simon, Biochimie 62 (1980) 331.
[3 I ] B. Rambeck, H. Simon, Angew. Chem. 86 (1974) 675; Angew. Chem. Int.
Ed. Engl. I3 (1974) 609.
[32] H. Simon, J. Bader, B. Rambeck, E. Krezdorn, W. Tischer in H. Dellweg: 4 . Symp. Tech. Mikrobiol.. Versuchs- und Lehranstalt fur Spiritusfabrikation und Fermentationstechnologie im lnstitut fur Garungsgewerbe und Biotechnologie, Berlin 1979, p. 325.
1331 H. Gunther, J. Bader, S. Neumann, H. Simon in: 3. Eur. Congr. Biotechno/.. Ed. 1 . Verlag Chemie, Weinheim 1984, p. 261.
552
(1976) 151.
1371 Y. Izumi, S. K. Mishra, B. S. Gosh, Y. Tani, H. Yamada, J . Ferment.
Technol. 61 (1983) 135.
1381 Y. Noma, S. Nonomura, H. Ueda, C. Tatsumi, Agric. Biol. Chem. 38
(1974) 735.
1391 L. Angermaier, J. Bader, H. Simon, Hoppe SeylerP Z . Physiol. Chem. 362
(1981) 33.
1401 J. Bader, M.-A. Kim, H. Simon, Hoppe Seyler's Z . Physiol. Chem. 362
(I98 I ) 809.
1411 R. Wienkamp, E. Steckhan, Angew. Chem. 94 (1982) 786; Angew. Chem.
Int. Ed. Engl. 21 (1982) 782; Angew. Chem. Suppl. 1982. 1739.
1421 R. Wienkamp, E. Steckhan, Angew. Chem. 95 (1983) 508; Angew. Chem.
f n t . Ed. Engl. 22 (1983) 497.
1431 C. Wandrey, R. Wichmann, W. Berke, M. Morr, M.-R. Kula in: 3. Eur.
Congr. Biotechnol., Ed. 1 . Verlag Chemie, Weinheim 1984, p. 239.
144) J. Bader, H. Giinther, S. Nagata, H. J. Schuetz, M. L. Link, H. Simon, J .
Biotechnol. 1 (1984) 95.
1451 C. H. Wong, G. M. Whitesides, J. Am. Chem. Soc. 105 (1983) 5012.
1461 J. P. Vandecasteele, Appl. Enuironm. Microbiol. 39 (1980) 327.
(471 H. Simon, H. Giinther, J. Bader, S. Neumann in: Enzymes in Organic
Synthesis (Ciba Found. Symp. 1 1 1 ) . Pitman, London 1985, p. 97.
(481 J. B. Jones in P. Dunnill, A. Wisemann, N. Blakeborough: Enzyme and
Non Enzymatic Catalysis, Wiley, New York 1980, p . 54.
1491 M. Nakazaki, H. Chikamatsu, T. Fujii, Y. Sasaki, S. Ao, J . Org. Chem.
48 (1983) 4337.
(501 R. J. Lamed, E. Kleinau, J. G. Zeikus, Enzyme Microb. Technol. 3 (1981)
144.
[Sl] S . W. May, S. R. Padgette, Bio/Technology 1983. 677.
1521 W. Hummel, N. Weiss, M.-R. Kula, Arch. Microbiol. 137 (1984) 47.
1531 H. Schiitte, W. Hummel, M.-R. Kula, Appl. Microbiol. Biorechnol. 19
(1984) 167.
1541 R. Hilhorst, R. Spruijt, C. Laane, C. Veeger, Eur. J . Biochem. 144 (1984)
459.
(551 J . Bader, H. Giinther, B. Rdmheck, H. Simon, Hoppe Seyler's Z . Physiol.
Chem. 359 (1978) 19.
1561 W. Tischer, J. Bader, H. Simon, Eur. J . Biochem. 97 (1979) 103.
[57] J. Bader, H. Simon, Arch. Mirrobiol. 127 (1980) 279.
(581 M. Biihler, H. Simon, Hoppe Seylerk Z . Physiol. Chem. 363 (1982) 609.
[59] H. Giesel, H. Simon, Arch. Microbiol. 135 (1983) 51.
(601 H. Simon, H. Giinther, J. Bader, W. Tischer, Angew. Chem. 93 (1981)
897; Angew. Chem. fnr. Ed. Engl. 20 (1981) 861.
1611 J. Thanos, J. Bader, H. Giinther, S. Neumann, H. Simon, Methods Enzymol.. in press.
[62] H. Hashimoto, H. Simon, Angew. Chem. 87 (1975) 1 1 I ; Angew. Chem.
Inr. Ed. Engl. 14 (1975) 106.
(631 H. Sedlmaier, W. Tischer, P. Rauschenbach, H. Simon, FEES Lett. 100
(1979) 129.
(641 K. Bartl, C. Cavalar, T. Krehs, E. Ripp, J. Retey, W. E. Hull, H. Giinther, H. Simon, Eur. J . Biochem. 72 (1977) 247.
(651 L. Angermaier, H. Simon, Hoppe Seyleri Z . Pltysiol. Chem. 364 (1982)
961, 1653.
(661 S. Neumann, H. Simon, FEES Left. 167 (1984) 29.
1671 S. Neumann, H. Giinther, H. Simon in: 3. Eur. Congr. Biorechnol.. Ed. 1 ,
Verlag Chemie, Weinheim 1984, p. 239.
[68] H. G. Schlegel, H. Kaltwasser, G. Gottschalk, Arch. Microbiol. 38 (1961)
209.
1691 P. Egerer, H. Simon, Biorechnol. Lett. 4 (1982) 501.
1701 A. Tanaka, S . Fukui, Adu. Biochem. Engin./BiotechnoL 29 (1984) 1.
(71) J. Klein, F. Wagner in L. B. Wingard, I. Chibata: Appl. Biochem. Bioeng.
Immobilized Cells 4 (1983) 11.
(721 J. Bader, H. Gunther, S. Nagata, H. Simon in: 3. Eur. Congr. Biorechnd,
Ed. I , Verlag Chemie, Weinheim 1984, p. 371.
(731 H. Gunther, C . Frank, H.-J. Schuetz, J. Bader, H. Simon, Angew. Chem.
95 (1983) 3 2 5 ; Angew. Chem. f n l . Ed. Engl. 22 (1983) 322; Angew. Chem.
Suppl. 1983, 463.
1741 P. A. Levene, A. Walti, Org. Synfh. 10 (1930) 84.
1751 C.-H. Wong, L. Daniels, W. H. Orme-Johnson, G. M. Whitesides, J. Am.
Chem. Soc. 103 (1981) 6227.
(761 P. Egerer, H. Simon, Biochim. Biophys. Acta 703 (1982) 158.
1771 M. Shin in A. San Pietro: Methods Enzymol. 23 (1971) 440.
[78] L. J . Arnold, K. You, W. S. Allison, N. 0. Kaplan, Biochemistry 15
(1976) 4844.
1791 L. Ernster, H. D. Hoberman, R. L. Howard, T. E. King, C:P. Lee, B.
Mackler, G. Scottocasa, Nature London 207 (1965) 940.
[SO] E. Schleicher, H. Simon, FEES Leu. 74 (1977) 269.
[Sl] H. R. Thirsk, J. A. Harrison: A Guide to the Study of Electrode Kinefrcs.
Academic Press, New York 1972.
(821 J. A. Farrington, M. Ehert, E. J. Land, K. Fletcher, Biochim. Bioph.v.7.
Acra 314 (1973) 372.
Angew. Chem. f n r . Ed. Engl 24 (1985) 539-553
1x31 C. C. Winterhourn, FEES Letf. 128 (1981) 339.
(841 M. Arai, Biochem. Z. 122 (1921) 254.
[85] 0. Lutr, B. Jirgensons, Chem. Ber. 6S (1932) 788.
(86) W. Leuchtenherger, M . Karrenhauer, U. Plocker, Ann. N . Y. Acad. Sci.
434 (1984) 78.
(87) B. D. Camhou, A. M. Klihanov, J . Am. Chem. SOC. 106 (1984) 2687, Ann
N . Y . Acad. Sci. 434 (1984) 219.
(881 M. Schneider, N . Engel, P. Honicke, G. Heinemann, H. Gorisch, Angew.
Chem. 96 (1984) 5 5 ; Angew. Chem. Int. Ed. Engl. 23 (1984) 67.
Syntheses with Radicals-C-C
Organomercury Compounds
[89] C:S. Chen, Y . Fujimoto, G. Girdaukas, C. J. Sih, J. Am. Chem. Soc. 104
(1982) 7294.
1901 C. Fuganti in: Enzymes in Organic Synthe.sis (Ciba Found. Symp. l / I J .
Pitman, London 1985, p. 112.
1911 M. T. Reetz, Angew. Chem. 96 (1984) 542; Angew. Chem. Int. Ed. Engl.
23 (1984) 556.
[92] D. Mandler, 1. Willner, J. Am. Chem. SOC. 106 (1984) 5352.
1
Bond Formation via Organotin and
New Synthetic
Methods (52)
By Bernd Giese*
Dedicated to Professor Rolf Huisgen on the occasion of his 65th birthday
C-C bond formation is one of the most important synthetic steps in the construction of organic molecules. In the last few years it has been increasingly achieved by radical addition
to alkenes. In such reactions the adduct radicals have to b e trapped by an donor subsequent
to the C-C bond formation in order to prevent polymerization. This task can be accomplished with organotin and organomercury hydrides, the use of which has led to new synthetic methods. The occurrence of radical chain reactions in which reactions take place between radicals and nonradicals is decisive for the success of the synthesis. In these cases
small amounts of radical initiators suffice and numerous functional groups may be used in
the C-C bond-forming reactions. The yields and selectivities of these radical reactions are
often very high.
1. Introduction
hydride, alkyl halides) with the formation of new radicals.
Until a few years ago radicals were still a domain for
mechanistically oriented research; a wealth of experimental data has led to a deeper understanding of their chemistry. Nowadays radical reactions are increasingly employed
in synthesis, wherein the product is formed through reactions of radicals either with other radicals or with molecules whose electron spins are paired. In radical-radical
reactions, the radicals must be continuously generated, so
that at least equivalent amounts of radical initiators are required. Examples are the coupling reactions of electrochemically generated radicals”] or the dimerization of capto-dative stabilized radicals developed by Viehe et aI.[’] in
recent years. Methods in which the products are produced
in reactions between radicals and nonradicals are fundamentally different. These syntheses, occurring via radical
chain reactions, require only small amounts of radical initiators, as shown, e.g., by the radical addition of alkyl halides to alkenes in the presence of tributyltin hydride.I3’ The
radicals 1 - 3 react with nonradicals (alkenes, tributyltin
[*I
Prof. Dr. B. Giese
lnstitut fur Organische Chemie und Biochemie
der Technischen Hochschule
Petersenstrasse 22, D-6100 Darmstadt (FRG)
Anyew. Chem. Int. Ed. Engl. 24 11985) 553-565
5.
k-X
Radical chain reactions such as the SRN1 reactions,’4’ the
additions of molecules with activated ~ a r b o n - h a l o g e n ~or~ ’
carbon-hydrogen bonds,[61 the Minisci alkylation~,”~and
the Meerwein arylations[’I have already been described in
detail. The subject of this review is the C-C bond-forming
reactions developed in recent years using organotin or organomercury compounds.
0 V C H Verlaysgesellschaft m b H , 0-6940 Weinheim. 1985
0570-0833/85/0707-0553 $ 02.50/0
553
Документ
Категория
Без категории
Просмотров
1
Размер файла
1 480 Кб
Теги
synthetic, chiral, compounds, biocatalytic, reduction, method, synthesizers, new
1/--страниц
Пожаловаться на содержимое документа