OsO4?streptavidin: A tunable hybrid catalyst for the enantioselective cis-dihydroxylation of olefins

Article abstract of DOI:10.1002/anie.201103632

Taking control: Selective catalysts for olefin dihydroxylation have been generated by the combination of apo-streptavidin and OsO₄. Site-directed mutagenesis allows improvement of enantioselectivity and even inversion of enantiopreference in certain cases. Notably allyl phenyl sulfide and cis-β-methylstyrene were converted with unprecedented enantiomeric excess.

Full text of DOI:10.1002/anie.201103632

DOI: 10.1002/anie.201103632
Artificial Metalloenzymes
OsO ·Streptavidin: A Tunable Hybrid Catalyst for the Enantioselective
4
cis-Dihydroxylation of Olefins**
Valentin Kçhler, Jincheng Mao, Tillmann Heinisch, Anca Pordea, Alessia Sardo,
Yvonne M. Wilson, Livia Knçrr, Marc Creus, Jean-Christophe Prost, Tilman Schirmer,* and
Thomas R. Ward*
Enzymatic and homogeneous catalysis have evolved inde-
pendently to address the challenges in the synthesis of
enantiopure products. With the aim of complementing these
fields, artificial metalloenzymes, which combine the structural
diversity of biocatalysts with the wealth of metal-catalyzed
[
1]
reactions, have attracted increasing attention. In homoge-
neous catalysis the cis-selective, OsO -dependent asymmetric
4
dihydroxylation (AD) of olefins ranks among the most
powerful methods for the synthesis of vicinal diols. Ligands
for homogeneous catalysis have been largely developed by
Sharpless and co-workers, and are, with few exceptions,
almost exclusively based on quinidine or quinine deriva-
tives. Although most classes of prochiral olefins are
dihydroxylated with good activity and selectivity, the cis-
substituted olefins are problematic. Nature relies on non-
heme iron dioxygenases such as naphthalene dioxygenase
Figure 1. Postulated transition-state structure for the dihydroxylation of
olefins: a) for the naphthalene dioxygenase; b) for the osmium-cata-
lyzed AD of prochiral olefins; and c) for an artificial cis-dihydroxylase
[2]
resulting from anchoring of OsO within a host protein.
4
resulting dihydroxylases could be optimized by genetic
means.
(
NDO) to perform a related reaction. These enzymes display
[3]
broad substrate scope. It is believed that both the OsO - and
Five proteins were evaluated as hosts for the AD of a-
methylstyrene: Wild-type streptavidin (SAV) clearly per-
formed best. In contrast, bovine serum albumin (BSA)
yielded the opposite enantiomer, albeit with a low turnover
number (Table 1). In view of the size of the proteins (66 kDa
for BSA and 16 kDa for the SAV monomer), the difficult
4
NDO-catalyzed dihydroxylations proceed by an outer sphere
[
3+2] mechanism in which the substrate is not bound to the
[2a,f,3c]
metal in the transition state (Figure 1).
Considering a biomimetic approach, we hypothesized that
VIII
anchoring a catalytically competent Os
center within a
[4]
protein might afford an artificial metalloenzyme for the AD
recombinant production of BSA, and the number of
[
2d]
of olefins. Encouraged by a report by Kokubo et al., we set
out to screen various proteins and to test whether the
Table 1: Identification of a suitable host for the AD of a-methylstyrene in
[
a]
the presence of OsO4.
[
+]
[+]
[+]
[
*] Dr. V. Kçhler, Dr. J. Mao, T. Heinisch, Dr. A. Pordea,
Dr. A. Sardo, Dr. Y. M. Wilson, L. Knçrr, Dr. M. Creus, J.-C. Prost,
Prof. Dr. T. R. Ward
Departement Chemie, Universitꢀt Basel
Spitalstrasse 51, 4056 Basel (Switzerland)
E-mail: thomas.ward@unibas.ch
[
b]
[c]
Entry
Protein
ee [%]
TON
[
+]
[
d]
T. Heinisch, Prof. Dr. T. Schirmer
Biozentrum, Universitꢀt Basel
1
2
3
4
5
avidin
BSA
lysozyme
human carbonic anhydrase II
SAV
2 (R)
77 (S)
5 (S)
25 (S)
95 (R)
16
4
6
<1
27
[
[
e]
f]
Klingelbergstrasse 50/70, 4056 Basel (Switzerland)
E-mail: tilman.schirmer@unibas.ch
+
[g]
[
] These authors contributed equally to this work.
[
**] We thank Prof. C. R. Cantor for the SAV gene, Chiral Technologies
Europe for help with the method development for chiral-phase
HPLC analysis, Beat Amrein for preliminary experiments, and
Maurus Schmid for his contribution to the artwork. This research
was funded by the Swiss National Science Foundation (Grant
[
a] Reactions were carried out with 6 mmol substrate in a total reaction
volume of 0.2 mL. [b] ee value determined by HPLC on a chiral stationary
phase; absolute configuration assigned by comparison with literature
data. [c] Determined by reverse-phase HPLC with internal standard.
TON=mol product per mol K [OsO (OH) ]. [d] Some precipitation was
observed at the end of the run. [e] Immediate precipitation occurred
upon addition of K CO /K [Fe(CN) ]/K [OsO (OH) ] stock solution to
2
2
4
200020-126366), the Cantons of Basel, and the Marie Curie Training
Network (FP7-ITN-238434).
2
3
3
6
2
2
4
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201103632.
the protein stock solution. [f] Protein was not fully soluble. [g] Average of
two independent runs.
Angew. Chem. Int. Ed. 2011, 50, 10863 –10866
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
10863

Communications
[
a]
histidine residues in BSA (17 His for BSA vs. 2 His per SAV
Table 2: Genetic optimization of the performance of artificial AD.
monomer), SAV was selected as the host for further studies.
Next, a range of substrates was subjected to AD in the
presence of SAV. With the exception of indene and precocene
I, all the substrates produced the same enantiomer as with the
[
2a,h–l]
quinidine scaffold in the Sharpless AD (i.e. AD-mix-b).
[
c]
[d]
Entry Olefin
SAV mutant
ee [%]
TON Ref.
ee [%]
Ref.]
Although some substrates were converted with a low to
moderate ee value, both cis-b-methylstyrene and a-methyl-
styrene were dihydroxylated with good selectivity (Table 2).
The expected syn specificity of the reaction was demonstrated
unambiguously (i.e. 1% or less anti product detected) by
comparison of the dihydroxylation products of cis-b-methyl-
styrene and trans-b-methylstyrene (see Supporting Informa-
tion).
To investigate the fate of SAV under the oxidizing
conditions, the reaction mixture was subjected to SDS-
PAGE after catalysis, and a remaining biotin-binding capa-
bility was revealed by biotin-4-fluorescein staining of the gel.
As the ee value remained constant over the course of the
reaction, we concluded that any decomposition pathway of
the catalyst leads to species with strongly reduced activity (see
Supporting Information).
[b]
(mol% Os)
[
1
2
3
4
WT (2.5)
40 (S)
82 (S)
77 (R)
0
13
14
21
9
S112Y (5.0)
D128A (2.5)
D128E (2.5)
88
[2h]
5
6
7
8
WT (2.5)
2 (R)
4
S112Y (5.0)
D128A (2.5)
D128E (2.5)
71 (S)
71 (R)
12 (R)
7
61
10
5
[2i]
9
10
11
WT (2.5)
S112Y (5.0)
S112M (5.0)
30 (1R,2S) 13
7 (1S,2R) 12
41 (1R,2S) 12
56
[2j]
12
13
14
15
WT (2.5)
90 (1R,2S) 26
45 (1R,2S) 11
92 (1R,2S) 16
91 (1R,2S) ꢀ16
S112Y (5.0)
S112T (2.5)
S112T (2.5)
72
[2j]
Addition of 1.05 equivalents of biotin per SAV monomer
afforded a nearly racemic product for the AD of a-methyl-
styrene. This finding suggests that the catalytically active Os
moiety is located in the vicinity of the biotin-binding pocket
Figure 2b and 2c). As the carboxylate D128 ··· HNurea
hydrogen bond of biotin is a key interaction for high
[e]
16
17
WT (2.5)
S112Y (5.0)
D128A (2.5)
5 (1S,2R) 12
16 (1R,2S) 12
45 (1R,2S) 13
53
(
[2k]
18
[8]
biotin·SAV affinity, SAV D128A was tested in the AD. a-
Methylstyrene was converted by the OsO ·SAV D128A
19
20
21
WT (2.5)
S112Y (5.0)
S112T (2.5)
62 (3R,4R)
26 (3R,4R)
68 (3R,4R)
6
5
6
67
2l]
4
[
9]
[
catalyst with significantly diminished selectivity. The AD
of allyl phenyl ether in the presence of SAV D128A afforded
the opposite enantiomer than with SAV and allyl phenyl
sulfide was converted with remarkable selectivity. In contrast,
the conservative mutation SAV D128E, which extends the
side chain of residue 128 by a methylene group, led to less
dramatic changes in the enantioselectivity (Table 2).
Further insight on the location of the catalytically active
Os species was gathered by mutation of further residues
located close to the biotin-binding site. Inspection of the SAV
structure revealed that the l-7,8 loops (residues 113–121,
Figure 2b) of two neighboring monomers line the biotin-
22
23
WT (2.5)
95 (R)
24 (R)
16 (R)
89 (R)
75 (R)
97 (R)
53 (R)
80 (R)
9 (R)
27
8
K121N (5.0)
L124G (5.0)
L124K (2.5)
S112Y (5.0)
S112M (5.0)
D128A (2.5)
D128E (2.5)
WT (2.5) +
biotin
2
2
2
2
2
4
5
6
7
8
11
18
15
16
22
21
13
99
[2n]
29
30
[f]
[
a] Results are the average of two independent runs, see Table 1 and
[10]
binding pocket.
Single point mutants at position S112,
K121, and L124 were combined with various amounts of
Supporting Information for experimental details. [b] The ideal osmate
loading was determined for each mutant (see Supporting Information
and Ref. [7]). [c] ee determined by HPLC on a chiral stationary phase;
absolute configuration assigned by comparison with literature data.
d] Determined by reverse-phase HPLC with internal standard.
TON=mol product/mol K [OsO (OH) ]. [e] Carried out on a 120 mmol
[
11]
K [OsO (OH) ] and tested in AD.
Mutation at position
2
2
4
SAV S112 proved most effective for fine-tuning purposes.
Both SAV S112Y and SAV S112M led to an improvement of
selectivity in many cases. a-Methylstyrene was converted with
[
2
2
4
scale; TON is based on yield of isolated product. [f] 1.05 equivalents of
7
5% ee (R) with SAV S112Y and 97% ee (R) with SAV
d-biotin were added relative to protein monomer.
S112M. Allyl phenyl sulfide, which gave nearly racemic
product with SAV, was converted with 71% ee (S) with SAV
S112Y. The best result for the AD of 1,2-dihydronaphthalene
was obtained with SAV S112M (41% ee, 1R,2S). The
conservative mutation SAV S112T, which introduces an
additional methyl group to the amino acid side chain of
residue 112, proved best for the dihydroxylation of precocene
The system performed well on a range of challenging
substrates. Noteworthy results were obtained for allyl phenyl
sulfide (Table 2, entries 6, and 7) and cis-b-methylstyrene
(Table 2, entry 14) and are the highest ee values ever reported
for these substrates to the best of our knowledge.
(
68% ee, 3R,4R) and of cis-b-methylstyrene (92% ee 1R,2S),
For other demanding cis-substituted substrates, such as
1,2-dihydronaphtalene and indene, the enantioselectivities
could be significantly improved by site-directed mutagenesis
Table 2).
1
0864 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 10863 –10866

(
Table 2, entries 11 and 18). For the AD of precocene, the
site (e.g. H87, H127 and K80) were subjected to site-directed
mutagenesis. Catalysis in the presence of single mutants SAV
H87A, SAV H127A, and SAV K80G gave comparable results
to SAV in the AD of a-methylstyrene, (lowest ee 89% (R)
ee value obtained with SAV S112T matched the best result
obtained with quinidine-derived systems.
Importantly, exchange of the lysine side chain at posi-
tion 121 resulted in low levels of stereoselectivity: the highest
ee value observed for any SAV K121 mutant/substrate
combination was 24% (R) (a-methylstyrene; Table 2,
entry 23). Mutating the leucine L124 had a less pronounced
deleterious (or neutral in one case) effect on selectivity.
With the aim of gaining structural insight on the location
of the Os center, SAV crystals were grown (at pH 4.0 or 7.3)
and subsequently soaked in a solution of K [OsO (OH) ].
[6]
with H87A; see Table S1 in the Supporting Information).
Although the structural data did not allow identification of
the actual catalytic site, it suggests that not all of the osmium
is catalytically active, which accounts for the modest TONs
[
2m]
(up to 27 turnovers per added osmium).
In summary, we have demonstrated that high selectivity
for challenging substrates can be obtained with an artificial
AD based on OsO ·SAV. Importantly, it was shown that the
2
2
4
4
The resulting crystals were subjected to X-ray analysis (see
the Supporting Information).
hybrid catalyst offers vast opportunities for genetic optimi-
zation: Single point mutations can lead to a significant
increase in enantioselectivity and even to an inversion of
enantiopreference. As for several natural mono- and dioxy-
genases, this artificial metalloenzyme suffers from modest
TONs combined with a high molecular weight. Current
efforts are thus aimed at increasing the TON by identifying
the location and the kinetic profile of the catalytically
competent osmium species and testing other challenging
substrates, including tri- and tetrasubstituted olefins. Finally,
from a biomimetic perspective, combination with a flavin-
coupled NMO-recycling system with H O as the stoichio-
Soaking crystals, obtained at pH 7.3, resulted in anom-
alous scattering density (modeled with an Os atom), which
z
e
was detected in the proximity of four residues: N of K80, N
e
z
of H87, N of H127 and N of K132 (Figure 2a). The intrinsic
packing disorder, however, did not allow full refinement but
yielded precious information on potential Os-binding sites. In
contrast, the X-ray data of the crystals obtained at pH 4.0,
yielded a well-behaved structure with a single Os binding site
e
close to N of the H127 side chain (see the Supporting
Information).
2
2
[
2n,o]
The three osmium binding sites identified by X-ray
crystallography located in the proximity of the biotin binding
metric oxidant would be of great interest.
Figure 2. Crystal structure of SAV soaked in K [OsO (OH) ] at pH 7.3: a) ribbon representation of SAV tetramer highlighting positions of strong
2
2
4
anomalous difference density (in red, contoured at 4s); b) close-up view of the biotin-binding pocket (from PDB 1stp) highlighting the residues
and biotin (stick representation) which strongly influence catalysis; c) surface representation of close-up view.
Angew. Chem. Int. Ed. 2011, 50, 10863 –10866
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 10865

Communications
Received: May 27, 2011
Revised: August 29, 2011
Published online: September 21, 2011
Sharpless, J. Am. Chem. Soc. 1994, 116, 1278 – 1291; n) S. Y.
Jonsson, K. Fꢀrnegꢁrdh, J.-E. Bꢀckvall, J. Am. Chem. Soc. 2001,
123, 1365 – 1371; o) K. Bergstad, S. Y. Jonsson, J.-E. Bꢀckvall, J.
Am. Chem. Soc. 1999, 121, 10424 – 10425.
[
3] a) D. R. Boyd, G. N. Sheldrake, Nat. Prod. Rep. 1998, 15, 309 –
Keywords: dihydroxylation · metalloenzymes · mutagenesis ·
osmium
.
324; b) T. Hudlicky, D. Gonzalez, D. T. Gibson, Aldrichimica
Acta 1999, 32, 35 – 62; c) A. Karlsson, J. V. Parales, R. E. Parales,
D. T. Gibson, H. Eklund, S. Ramaswamy, Science 2003, 299,
1039 – 1042; d) T. Hudlicky, J. W. Reed, Synlett 2009, 685 – 703.
[
1] a) Y. Lu, N. Yeung, N. Sieracki, N. M. Marshall, Nature 2009, 460,
[
[
[
4] P. A. Colussi, T. C. Evans, C. H. Taron (New England Biolabs,
Inc.), WO2007123689, 2007.
5] N. Humbert, A. Zocchi, T. R. Ward, Electrophoresis 2005, 26,
8
9
55 – 862; b) F. Rosati, G. Roelfes, ChemCatChem 2010, 2, 916 –
27; c) M. T. Reetz, Top. Organomet. Chem. 2009, 25, 63 – 92;
d) S. Abe, T. Ueno, Y. Watanabe, Top. Organomet. Chem. 2009,
47 – 52.
25, 25 – 43; e) Q. Jing, R. J. Kazlauskas, ChemCatChem 2010, 2,
6] Although the effect of mutating K132 has not yet been
investigated, the results obtained with mutations at other sites
953 – 957; f) P. J. Deuss, G. Popa, C. H. Botting, W. Laan, P. C.
Kamer, Angew. Chem. 2010, 122, 5443 – 5445; Angew. Chem. Int.
Ed. 2010, 49, 5315 – 5317; g) J. Podtetenieff, A. Taglieber, E. Bill,
E. J. Reijerse, M. T. Reetz, Angew. Chem. 2010, 122, 5277 – 5281;
Angew. Chem. Int. Ed. 2010, 49, 5151 – 5155; h) T. Heinisch, T. R.
Ward, Curr. Opin. Chem. Biol. 2010, 14, 184 – 199; i) P. Fournier,
R. Fiammengo, A. Jꢀschke, Angew. Chem. 2009, 121, 4490 –
(namely S112, K121, L124, and D128) render the involvement of
this amino acid side chain unlikely.
[
[
7] The performance of mutants/batches used in this study was first
evaluated with various ratios of osmate to protein to determine
the loading that led to the highest HPLC yield without erosion of
enantioselectivity; see the Supporting Information for details.
8] a) P. C. Weber, D. H. Ohlendorf, J. J. Wendoloski, F. R. Sale-
mme, Science 1989, 243, 85 – 88; b) P. C. Weber, J. J. Wendoloski,
M. W. Pantoliano, F. R. Salemme, J. Am. Chem. Soc. 1992, 114,
4493; Angew. Chem. Int. Ed. 2009, 48, 4426 – 4429.
[
2] a) H. C. Kolb, M. S. VanNieuwenhze, K. B. Sharpless, Chem.
Rev. 1994, 94, 2483 – 2587; b) T. Oishi, M. Hirama, Tetrahedron
Lett. 1992, 33, 639 – 642; c) Y. Imada, T. Saito, T. Kawakami, S.-I.
Murahashi, Tetrahedron Lett. 1992, 33, 5081 – 5084; d) T.
Kokubo, T. Sugimoto, T. Uchida, S. Tanimoto, M. Okano, J.
Chem. Soc. Chem. Commun. 1983, 769 – 770; e) M. A. Anders-
son, R. Epple, V. V. Fokin, K. B. Sharpless, Angew. Chem. 2002,
3197 – 3200; c) J. DeChancie, K. N. Houk, J. Am. Chem. Soc.
2007, 129, 5419 – 5429; d) S. Freitag, V. Chu, J. E. Penzotti, L. A.
Klumb, R. To, D. Hyre, I. Le Trong, T. P. Lybrand, R. E.
Stenkamp, P. S. Stayton, Proc. Natl. Acad. Sci. USA 1999, 96,
8384 – 8389.
114, 490 – 493; Angew. Chem. Int. Ed. 2002, 41, 472 – 475; f) A. B.
[
9] A. Pordea, M. Creus, J. Panek, C. Duboc, D. Mathis, M. Novic,
Zaitsev, H. Adolfsson, Synthesis 2006, 1725 – 1756; g) K. B.
Sharpless, Angew. Chem. 2002, 114, 2126 – 2135; Angew. Chem.
Int. Ed. 2002, 41, 2024 – 2032; h) Z.-M. Wang, X.-L. Zhang, K. B.
Sharpless, Tetrahedron Lett. 1993, 34, 2267 – 2270; i) P. J. Walsh,
P. T. Ho, S. B. King, K. B. Sharpless, Tetrahedron Lett. 1994, 35,
T. R. Ward, J. Am. Chem. Soc. 2008, 130, 8085 – 8088.
[
[
10] The residue K121 is significantly closer to residue D128 of the
adjacent monomer in the tetrameric structure of SAV than to the
D128 on the same oligopeptide chain (12.67 vs. 20.70 ꢂ for C ).
a
11] After an initial screening of a wider range of SAV mutants with
styrene, five SAV S112X mutants (S112N, S112A, S112T, S112M,
S112Y), four SAV K121X (K121A, K121F, K121N, K121H), and
four SAV L124X mutants (L124G, L124K, L124N, L124Y) were
selected for further study.
5129 – 5132; j) H. Becker, S. B. King, M. Taniguchi, K. P. M.
Vanhessche, K. B. Sharpless, J. Org. Chem. 1995, 60, 3940 – 3941;
k) L. Wang, K. B. Sharpless, J. Am. Chem. Soc. 1992, 114, 7568 –
7570; l) Z.-M. Wang, K. Kakiuchi, K. B. Sharpless, J. Org. Chem.
1994, 59, 6895 – 6897; m) H. C. Kolb, P. G. Andersson, K. B.
1
0866 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 10863 –10866