Selectivity-enhancement in enantioselective hydrolysis of sec-alkyl sulfates by an alkylsulfatase from Rhodococcus ruber DSM 44541

Article abstract of DOI:10.1016/S0957-4166(02)00363-4

The Enantioselectivity of the biohydrolysis of sec-alkyl sulfate esters using a bacterial alkylsulfatase from Rhodococcus ruber DSM 44541 was dramatically enhanced in presence of additives ('enhancers') such as carbohydrates, polyethylene glycol, detergents, metal ions and through enzyme immobilization. In presence of iron, the E value for the kinetic resolution of (±)-3- and (±)-4-octyl sulfate was improved from E=3.9 to ≥200 and E=1.1 to 10, respectively.

Full text of DOI:10.1016/S0957-4166(02)00363-4

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 13 (2002) 1443–1447
Selectivity-enhancement in enantioselective hydrolysis of
sec-alkyl sulfates by an alkylsulfatase from Rhodococcus ruber
DSM 44541
a
b
b
a,
Mateja Pogorevc, Ulrike T. Strauss, Thomas Riermeier and Kurt Faber *
a
Department of Chemistry, Organic & Bioorganic Chemistry, University of Graz, Heinrichstrasse 28, A-8010 Graz, Austria
b
Degussa AG, Projekthaus Katalyse, Industriepark H o¨ chst, D-65926, Frankfurt/M, Germany
Received 1 July 2002; accepted 3 July 2002
Abstract—The Enantioselectivity of the biohydrolysis of sec-alkyl sulfate esters using a bacterial alkylsulfatase from Rhodococcus
ruber DSM 44541 was dramatically enhanced in presence of additives (‘enhancers’) such as carbohydrates, polyethylene glycol,
detergents, metal ions and through enzyme immobilization. In presence of iron, the E value for the kinetic resolution of (±)-3- and
(
±)-4-octyl sulfate was improved from E=3.9 to ]200 and E=1.1 to 10, respectively. © 2002 Elsevier Science Ltd. All rights
reserved.
7
1. Introduction
being attacked. Thus, from a rac-sec-alkyl sulfate,
R)-enantiomers were preferentially hydrolyzed with
(
In contrast to the majority of hydrolytic enzymes, such
as esterases, lipases and proteases, which do not affect
inversion of configuration to furnish the corresponding
(S)-sec-alcohol and remaining non-converted (S)-alkyl
sulfate. As a result, a homochiral product mixture was
1
any stereogenic center located within the substrate dur-
2
8
ing catalysis, alkylsulfatases [EC 3.1.6.X] are flexible
obtained.
concerning the stereochemistry during catalysis
(
Scheme 1). Depending on the nature of the enzyme
The first enantioselective hydrolysis of sec-alkyl sulfates
by an alkylsulfatase was discovered by the chance
observation that several rac-alkyl sulfates were rapidly
hydrolyzed until 50% conversion, whereas the remain-
der took place at a very much reduced reaction rate or
and its mechanism of action, hydrolysis of a sec-alkyl
sulfate ester may proceed either by cleavage of the SꢀO
bond or the CꢀO bond, which results in retention or
3,4
inversion of configuration at the stereogenic carbon.
9,10
This stereochemical flexibility makes them particularly
attractive for the design of so-called deracemization
processes, which allow the transformation of a race-
mate into a single stereoisomeric product in 100%
not at all.
This observation—together with addi-
tional experiments performed with enantiopure sub-
strates—led to the conclusion that alkylsulfatase CS1
from Comamonas terrigena (NCIMB 8193) and S2 from
Pseudomonas C12B (NCIMB 11753=ATCC 43648)
exhibited a certain enantioselectivity for (R)-configured
5
,6
theoretical yield. Most recently, we have isolated an
alkylsulfatase from Rhodococcus ruber DSM 44541
10
(
termed RS2), which acts through absolute stereoinver-
sec-alkyl sulfates. An opposite enantiopreference was
sion of the stereogenic sec-carbon atom of the substrate
found for the CS2- and S1-enzyme from the above-
Scheme 1. Retention or inversion during enzymatic sulfate ester hydrolysis.
*
Corresponding author. E-mail: kurt.faber@uni-graz.at
0
957-4166/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(02)00363-4

1
444
M. Pogore6c et al. / Tetrahedron: Asymmetry 13 (2002) 1443–1447
mentioned strains. In spite of these biochemical data,
which indicate the possibility of kinetic resolution, no
data on the stereo- and enantioselectivities of alkylsul-
metric (bio)catalysis and thus it is the prime factor in
13
chiral discrimination, in particular for substrates pos-
sessing low polarity. This phenomenon was a suitable
explanation for the fact that the enantioselectivity of
enzyme RS2—expressed as the enantiomeric ratio (E
7
,8
fatases were reported to date.
14
During our first study on the stereo- and enantioselec-
value) —was optimal for substrate rac-2-octyl sulfate
7
1
tive hydrolysis of sec-alkyl sulfates, it was found that
(1a) (E=21) where the relative size of substituents R
2
alkylsulfatase RS2 exhibited absolute stereoselectivity
with respect to inversion of configuration during cataly-
sis. In contrast, the corresponding enantioselectivities
were less than satisfactory as they ranged from poor to
acceptable and an E value of 21 for (±)-2-octyl sulfate
was obtained at most. In order to broaden the appli-
cability of this useful enzyme for the deracemization of
various sec-alcohols through stereo- and enantioselec-
tive hydrolysis of their corresponding rac-sulfate esters,
we studied the possibility to enhance enantioselectivities
by addition of various components which are known to
act as mediators for the chiral recognition process of
and R differs significantly (Scheme 2). In contrast,
enantioselectivities markedly dropped when the stereo-
genic sulfate ester group was moved towards the center
1
2
of the molecule, with R and R becoming similar in
size to give 3- (2a, E=3.9) and 4-octyl sulfate (3a,
E=1.1). This fact prompted us to focus our study on
(±)-2a and (±)-3a which were transformed with insuffi-
cient enantioselectivities. Furthermore, whereas several
methods for the preparation of nonracemic 2-octanol
15
are available, t6he synthesis of 3- and 4-octanol in high
1
e.e. is difficult.
1
1,12
enzymes.
2.1. Selectivity-enhancement of 3-octyl sulfate
rac-3-Octyl sulfate 2a was resolved with low enantiose-
lectivity in the absence of additives (E=3.9). The addi-
tion of carbohydrate derivatives and polyethylene
glycol (PEG) showed some positive, but overall still
insufficient effects on the enantioselectivity, with
DEAE–dextran being best (E=9.5) (Table 1). These
2
. Results and discussion
It is a commonly observed phenomenon that the rela-
tive size of substituents attached to a stereogenic center
dominates the spatial recognition process during asym-
Scheme 2. Stereo- and enantioselective hydrolysis of sec-alkyl sulfates by alkylsulfatase RS2.
Table 1. Enhancement of enantioselectivity for (9)-2a in presence of carbohydrates, PEG, enzyme immobilization and
detergents (see Scheme 2).
Additive
Nonea
Concentration (w/v)
Conversion (%)
41
E.e.p (%)
48
Enantioselectivity E
–
3.9
Trehalose
PEG 6000
Dextran MW 41 kDa
DEAE–Dextran
5
5
5
5
38
41
34
33
50
60
64
74
4.0
6.0
6.3
9.5
b
c
DEAE–Sephadex A-25
–
–
–
32
26
26
42
66
86
3.0
6.1
17.8
b
c
TEAE–Cellulose
b
Ecteola–Cellulose
Sodium dodecyl sulfate (SDS)
-Octanesulfonic acid
Triton X-100
0.2
0.2
2.0
0.2
B1
36
36
–
–
1
56
60
90
4.8
5.5
25.5
Cetyl trimethylammonium bromide
25
a
Under standard conditions using native enzyme RS2 in the absence of additive.
Enzyme RS2 immobilized onto this carrier.
Some of the enzyme activity was not bound.
b
c

M. Pogore6c et al. / Tetrahedron: Asymmetry 13 (2002) 1443–1447
1445
additives are believed to exert their selectivity-enhanc-
ing effects through hydrogen bonding on the enzyme
surface. Positive effects of PEG on the catalytic perfor-
Due to the fact that the addition of EDTA did not
19
abolish enzymatic activity, we concluded that enzyme
RS2 was not dependent on any metal ions as cofactors
for its mechanism of action. This property and the fact
that the substrate represents a negatively charged entity
at neutral pH, led us to the assumption that metal ions
might exert certain selectivity-enhancing effects (Table
12
mance of lipases were reported recently. Furthermore,
DEAE–dextran combines the properties of a polysac-
charide with those of an immobilization matrix due to
its positively charged DEAE residues, which can bind
proteins. It is assumed that this combination of proper-
ties makes it effective in comparison to ‘normal’ dex-
trans (MW 188 kDa) or DEAE–cellulose, both of
which alone had no effect on the enantioselectivity.
2+
+
2+
2, Scheme 2). Several metals, such as Ca , Li , Mg ,
are known to improve activity and/or selectivity of
12,20
lipases.
To our disappointment, a range of metals
often tested for their effects on enzyme activity, such as
2+
2+
2+
2+
Mg , Ca , Mn and Co showed no effect within the
limits of detectability. However, the use of iron had a
dramatic effect, which was strongly concentration-
Due to the fact that DEAE–dextran showed the most
promising results, changes in enantioselectivity upon
enzyme immobilization was investigated. Successful
binding onto macroscopic materials was only feasible
onto anionic carriers. In contrast, adsorptive binding
onto various types of hydrophobic materials of
Accurel-type, as well as covalent attachment though
epoxy-groups (VA-Epoxy-Biosynth) or activated Sep-
harose (thiopropyl-activated Sepharose 6B and acti-
2
+
dependent: Addition of increasing amounts of Fe led
to a decrease of reaction rate going in hand with an
improvement of the E value (up to Eꢀ80). The same,
but even more pronounced effect was observed with
3+
Fe . In this case, virtually absolute enantioselectivity
(E]200) was achieved at the price of a reduced reac-
tion rate. The limits of this method are reached at
3+
vated
CH-Sepharose
6B)
failed.
Improved
elevated Fe -concentration, which led to enzyme deac-
tivation. Whereas the exact mechanism of action of
iron on the alkylsulfatase remains unclear, any redox-
enantioselectivities were only detected when the enzyme
was bound on Ecteola-Cellulose, a carrier possessing
epichlorohydrin–triethanolamin moieties. In this case,
the E value was close to preparative utility (E=17.8).
2
+
phenomena due to spontaneous oxidation of Fe “
3+
Fe
can be excluded since the addition of
mercaptoethanol and dithiothreitol had no effect on
19
The most pronounced effects of detergents on the cata-
lytic performance of enzymes were reported for lipases
and the effects were largely attributed to interfacial
activity and selectivity of the enzyme. It appears clear,
however, that the ionic strength of the metal ion seems
to be an important factor.
1
7
activation. Depending on the nature of the detergent,
its interaction with the enzyme is predominantly based
2.2. Selectivity-enhancement of 4-octyl sulfate
1
8
on ionic and/or hydrophobic parameters. Recalling
the detergent-like structure of alkyl sulfates, strong
effects were anticipated. Most interestingly, substrate-
like anionic detergents, such as sodium dodecyl sulfate
and di-n-octylsulfosuccinate resulted in a complete loss
of activity even at low concentration (0.2% w/v),
whereas 1-octanesulfonic acid had no effect. Similar
results were obtained with non-ionic detergents, such as
Triton X-100 and Tween 80. In contrast, a remarkable
improvement was obtained with a cationic detergent—
cetyl trimethylammonium bromide—which resulted in
an E value of preparative utility (E >25).
Since both alkyl chains in rac-4-octyl sulfate differ only
by a single methylene group, it is a quasi-symmetric
substrate, which makes chiral recognition a very
16d
difficult task.
tioselectivity was observed under standard conditions in
the absence of enhancers (E=1.1). Since
As a consequence, no notable enan-
(poly)saccharides, PEG and detergents had little or
limited effects on the enantioselectivity of substrate
(±)-2a, they were excluded from further studies and our
attention was focused on the effects of iron (Table 3,
Scheme 2).
Table 2. Enhancement of enantioselectivity for (9)-2a in presence of metal ions (see Scheme 2)
Additive
Nonea
Concentration (mM)
Conversion (%)
41
E.e.p (%)
48
Enantioselectivity E
—
3.9
MgCl2
CaCl2
MnCl2
CoCl2
10
10
10
10
41
41
39
37
40
46
38
50
3.0
3.7
2.8
4.0
FeCl2
FeCl2
FeCl2
2
5
10
37
35
18
66
88
97
7.1
25
81
FeCl3
FeCl3
FeCl3
2
5
10
27
9
B1
50
99
–
3.6
]200
–
a
Under standard conditions using native enzyme RS2 in the absence of additive.

1
446
M. Pogore6c et al. / Tetrahedron: Asymmetry 13 (2002) 1443–1447
Table 3. Enhancement of enantioselectivity for (9)-3a in presence of iron (see Scheme 2)
Additive
Concentration (mM)
Conversion (%)
E.e.p (%)
Enantioselectivity E
Nonea
FeCl2
FeCl2
FeCl2
FeCl3
FeCl3
FeCl3
–
54
50
32
17
48
4
14
62
80
10
78
—
1.1
1.5
5.6
10
1.3
9.7
—
2
5
10
2
5
19
B1
10
a
Under standard conditions using native enzyme RS2 in the absence of additive.
We were pleased to observe the same effects, i.e. a
concentration-dependent enhancement of enantioselec-
tivity in general, and a positive influence of the ionic
described in the preceding paper (±)-1a–3a. Additives
were obtained from Sigma (immobilization carriers,
detergents), Merck and Fluka (salts). For the determi-
2
+
3+
strength of Fe versus Fe . Again, complete deactiva-
nation of conversion, e.e. as well as the elucidation of
p
3+
tion took place at elevated concentrations of Fe .
absolute configuration of products see the experimental
section of the preceding paper.
2+
Overall, the addition of 10 mM Fe improved the
initial selectivity from nil to a remarkable value of
E=10.
3.2. Biocatalytic procedures
Rhodococcus ruber DSM 44541 was grown under aero-
bic conditions in baffled Erlenmeyer flasks at 30°C and
The use of various components, such as carbohydrates,
PEG, detergents, and metal ions has been proven to be
a largely empirical, but powerful method for the tuning
of enzyme (enantio)selectivity for several types of bio-
catalyzed reactions. In general, allosteric effects
denoted as ‘enantioselective inhibition’ have been
quoted. However, since their mode of interaction with
the protein is still poorly understood on a molecular
basis, rationales for the explanation of these effects are
still highly speculative.
1
30 rpm using the following medium: 10 g/l glucose, 10
g/l peptone, 10 g/l yeast extract, 2 g/l NaCl, 1.5 g
MgSO ·7 H O, 1.3 g/l NaH PO and K HPO . Cell
4
2
2
4
2
4
growth was monitored by measurement of the optical
density via the absorption at 546 nm. Biotransforma-
tions were carried out using a partially purified enzyme.
2
1
A crude enzyme preparation was obtained as follows:
Cells of R. ruber DSM 44541 from the late exponential
growth phase were disintegrated using a bead mill
The data from this study allows the following points to
be addressed: It has been proposed that the docking of
the alkyl sulfate within the active site of the enzyme
predominantly involves (strong) ionic binding via the
(
Vibrogen Zellm u¨ hle, bead diameter 0.35 mm). After
centrifugation of cell debris (38 000×g, 2 h), the super-
natant was subjected to hydrophobic chromatography
using a Phenyl Sepharose column (Pharmacia). Active
fractions were collected, dialyzed against 10 mM Tris/
HCl buffer pH 7.5 and finally lyophilized. The activity
of the crude enzyme preparation was 1.2 U/ml; 1 U of
(
negatively charged) sulfate ester group, whereas chiral
discrimination occurs by (weak) Van der Waal’s inter-
actions involving the positioning of the alkyl groups
into two hydrophobic pockets. As a consequence, chiral
recognition largely depends on discrimination by
2−
−1
sulfatase liberates 1 mmol SO4 min .
1
0,22
size.
It is therefore unsurprising that (among several
Immobilization of alkylsulfatase: An aliquot of par-
tially purified alkylsulfatase preparation (1 ml, 11 mg
protein/ml) in Tris/HCl buffer (0.1 M, pH 7.5) was
added to the immobilization carrier (100 mg), which
was pre-washed with the same buffer. The mixture was
gently mixed for 5 min and allowed to stand for
another 60 min at 4°C. After centrifugation and separa-
tion of the supernatant, the carrier was repeatedly
washed with the same buffer and tested for alkylsulfa-
tase activity. The relative activities in relation to the
initial activity (set to 100%) were as follows: TEAE–
Cellulose, ꢀ100% bound and <1% nonbound activity;
Ecteola–cellulose, 22% bound and 34% nonbound
activity.
types of additives) metal ions had the most pronounced
effect on the enantioselectivity, because they are most
likely to interfere with the ionic substrate binding
through formation of ion pairs with the charged
substrate.
3. Experimental
3.1. General remarks
All GC analyses were carried out on a HP 6890 gas
chromatograph, equipped with FID and a HP-1 capil-
lary column (30m×0.53 mm×0.5 mm, N ). Enantiomeric
2
purities were analyzed on a chiral column Chrompack
CP7500 (25 m×0.25 mm×25 mm) using H as carrier
2
Acknowledgements
gas.
Financial support by Degussa AG (Frankfurt) is grate-
fully acknowledged and H. Trauthwein is cordially
thanked for his valuable contributions.
Alcohols were purchased from Aldrich (1b) and Fluka
(
2b,3b); sulfation of rac-1b–3b was completed as

M. Pogore6c et al. / Tetrahedron: Asymmetry 13 (2002) 1443–1447
1447
References
15. (a) Morgan, B.; Oehlschlager, A. C.; Stokes, T. M.
Tetrahedron 1991, 47, 1611–1620; (b) Orrenius, C.;
1
. Bornscheuer, U. T.; Kazlauskas, R. J. Hydrolases in
Organic Synthesis; Verlag Chemie: Weinheim, 1999.
. Dodgson, K. S.; White, G. F.; Fitzgerald, J. W. Sulfa-
tases of Microbial Origin, 2 vols., CRC Press: Boca
Raton, 1982.
Ohrner, N.; Rotticci, D.; Mattson, A.; Hult, K.; Norin,
8
T. Tetrahedron: Asymmetry 1995, 6, 1217–1220; (c) Liese,
A.; Zelinski, T.; Kula, M.-R.; Kierkels, H.; Karutz, M.;
Kragl, U.; Wandrey, C. J. Mol. Catal. B: Enzymatic
2
1
998, 4, 91–99.
6. (a) Oritani, T.; Yamashita, K. Agric. Biol. Chem. 1980,
4, 2407–2411; (b) Kamezawa, M.; Raku, T.; Tachibana,
3
4
. Spencer, B. Biochem. J. 1958, 69, 155–159.
. Bartholomew, B.; Dodgson, K. S.; Matcham, G. W. J.;
Shaw, D. J.; White, G. F. Biochem. J. 1977, 165, 575–580.
. Strauss, U. T.; Felfer, U.; Faber, K. Tetrahedron: Asym-
metry 1999, 10, 107–117.
1
4
H.; Ohtani, T.; Naoshima, Y. Biosci. Biotechnol.
Biochem. 1994, 58, 598–599; (c) Nakamura, K.; Fujii, M.;
Ida, Y. J. Chem. Soc., Perkin Trans. 1 2000, 19, 3205–
5
6
7
. Faber, K. Chem. Eur. J. 2001, 7, 5004–5010.
. Pogorevc, M.; Faber, K., see the preceding paper in this
issue.
3
211; (d) Adam, W.; Lukacs, Z.; Saha-M o¨ ller, C. R.;
Weckerle, B.; Schreier, P. Eur. J. Org. Chem. 2000,
923–2926.
2
8
. For a preliminary communication see: Pogorevc, M.;
Kroutil, W.; Wallner, S. R.; Faber, K. Angew. Chem., Int.
Ed. Engl. 2002, 41, in press.
1
1
1
7. Ueji, S.-I.; Nishimura, M.; Kudo, R.; Matsumi, R.;
Watanabe, K.; Ebara, Y. Chem. Lett. 2001, 912–913.
8. Calvo, M. V.; Plou, F. J.; Ballesteros, A. Biocatalysis
9
. Matcham, G. W. J.; Bartholomew, B.; Dodgson, K. S.;
Fitzgerald, J. W.; Payne, W. J. FEMS Microbiol. Lett.
1
1996, 13, 271–285.
9. Pogorevc, M.; Faber, K., isolation and full biochemical
characterization of sulfatase RS2 will be published in due
course.
977, 1, 197–200.
0. Shaw, D. J.; Dodgson, K. S.; White, G. F. Biochem. J.
980, 187, 181–196.
1. Faber, K.; Ottolina, G.; Riva, S. Biocatalysis 1993, 8,
1–132.
2. Theil, F. Tetrahedron 2000, 56, 2905–2919.
3. (a) Hanson, K. R.; Rose, I. A. Acc. Chem. Res. 1975, 8,
1
1
1
2
0. (a) Okamoto, T.; Ueji, S. Biotechnol. Lett. 2000, 22,
1169–1171; (b) Okamoto, T.; Ueji, S. Chem. Commun.
9
1999, 939–940.
1
1
2
1. (a) Guo, Z.-W.; Sih, C. J. J. Am. Chem. Soc. 1989, 111,
836–6841; (b) Itoh, T.; Ohira, E.; Takagi, Y.;
Nishiyama, S.; Nakamura, K. Bull. Chem. Soc. Jpn. 1991,
4, 624–627; (c) Bamann, E.; Laeverenz, P. Ber. Dtsch.
6
1
–10; (b) Kazlauskas, R. J.; Weissfloch, A. N. E.; Rappa-
port, A. T.; Cuccia, L. A. J. Org. Chem. 1991, 56,
656–2665.
6
2
Chem. Ges. 1930, 63, 394–404.
1
4. Calculated from e.e.p and e.e. ; for a review see:
c
2
2. Matcham, G. W. J.; Dodgson, K. S.; Fitzgerald, J. W.
Straathof, A. J. J.; Jongejan, J. A. Enzyme Microb.
Technol. 1997, 21, 559–571.
Biochem. J. 1977, 167, 723–729.