Artificial metalloenzymes for enantioselective catalysis based on biotin-avidin

Article abstract of DOI:10.1021/ja035545i

Homogeneous and enzymatic catalysis offer complementary means to generate enantiomerically pure compounds. Incorporation of achiral biotinylated rhodium-diphosphine complexes into (strept)avidin yields artificial metalloenzymes for the hydrogenation of N-protected dehydroamino acids. A chemogenetic optimization procedure allows one to produce (R)-acetamidoalanine with 96% enantioselectivity. These hybrid catalysts display features reminiscent both of enzymatic and of homogeneous systems. Copyright

Full text of DOI:10.1021/ja035545i

Published on Web 07/03/2003
Artificial Metalloenzymes for Enantioselective Catalysis Based on
Biotin-Avidin
J e´ r oˆ me Collot, Julieta Gradinaru, Nicolas Humbert, Myriem Skander, Andrea Zocchi, and
Thomas R. Ward*
Institute of Chemistry, UniVersity of Neuch aˆ tel, AV. BelleVaux 51, CP 2, CH-2007 Neuch aˆ tel, Switzerland
Received April 9, 2003; E-mail: thomas.ward@unine.ch
Enantioselective catalysis is one of the most efficient ways to
synthesize high-added-value enantiomerically pure organic com-
pounds. As the subtle details which govern enantioselection cannot
Scheme 1
1
be reliably predicted or computed, catalysis relies more and more
2
on a combinatorial approach. Biocatalysis offers an attractive and
often complementary alternative for the synthesis of enantiopure
products.3 From a combinatorial perspective, the potential of
directed evolution techniques in optimizing an enzyme’s selectivity
4
is unrivaled. The present Communication outlines our efforts to
create artificial metalloenzymes by incorporating an achiral rhodium-
diphosphine moiety into a protein environment to yield efficient
enantioselective hydrogenation catalysts.
5
Inspired by the seminal work of E. T. Kaiser, several groups
have recently developed methods to covalently modify proteins by
incorporating transition-metal catalysts to yield hybrid catalytic
systems with promising properties.4
d,6,7
Table 1. Enantioselective Hydrogenation of the Acetamidoacrylic
Acid Using Biotinylated Complexes in (Strept)avidina
The approach delineated herein relies on noncovalent interactions
between the metal catalyst and the protein. As no chemical coupling
step is required upon addition of the catalyst precursor to the protein,
entry
ligand
host protein
buffer
ee (%)
b
1
2
3
4
5
6
7
8
9
Biot-1
avidin
avidin
neutravidin
neutravidin
streptavidin
streptavidin
avidin
phosphate
MOPS
phosphate
MOPS
acetate
acetate
MOPS
MOPS
acetate
acetate
MOPS
acetate
acetate
acetate
37 (S)
0
7 (S)
10 (S)
8
the integrity of organometallic species is warranted. As proposed
c
Biot-2
Biot-1
Biot-2
Biot-1
Biot-2
9
d
d
by Whitesides, the biotin-avidin system offers an attractive
scaffold to perform such experiments.10
92 (R)
Initial enantioselective hydrogenation of acetamidoacrylic acid
was performed with biotinylated amino-diphosphine ligands Biot-1
and Biot-2 using avidin as the host protein (Scheme 1).11 The
e
20 (S)
60 (S)
66 (S)
24 (S)
42 (R)
69 (S)
92 (R)
96 (R)
94 (R)
1
Biot-3 -1
1
Biot-3 -2
avidin
1
Biot-3 -1
streptavidin
streptavidin
avidin
streptavidin
strept. S122G
strept. S122G
modest levels of enantioselection obtained with both ligand scaffolds
1
1
1
0
1
Biot-3 -2
0a
(
Table 1, entries 11 and 2) led us to question the affinity of the
2
Biot-3 -1
+
+
catalyst precursors [Rh(COD)(Biot-1)] and [Rh(COD)(Biot-2)]
12
13
Biot-4ortho-1
for avidin. Indeed, the Coulomb repulsion caused by the cationic
character of avidin at neutral pH (pI ) 10.4) may dramatically
decrease the affinity of the biotinylated cationic catalysts for avidin.
We thus tested neutravidin (pI ) 6.3) with catalysts [Rh(COD)-
Biot-1
14
Biot-4ortho-1
a
Unless otherwise stated, conversions (conv.) were quantitative after
5 h. b 90% conv. c 94% conv. d 87% conv. e 55% conv. The ee and
1
+
+
(
Biot-1)] and [Rh(COD)(Biot-2)] (7% ee (S) and 10% ee (S),
conversion were determined by GC (after 15 h at room temperature, 5 atm
H2, buffer).11 Hydrogenation with all biotinylated catalysts in the absence
of a host protein produced essentially racemic material (ee < 8%,
quantitative).
respectively, entries 3 and 4).
As compared to avidin, streptavidin (pI ) 6.2) possesses a similar
a
affinity for biotin (K M ) but is endowed with a deeper
ca. 1014
-1
binding pocket.12 In the spirit of E. Fischer’s lock-and-key principle,
streptavidin may thus offer a more propitious environment for an
n
favorable chiral environment. Achiral alkylamino acid spacers 3
q
(n ) 1-5) and arylamino acid spacers 4 (q ) ortho-, meta-, para-)
enantiodiscrimination event. To our delight, catalyst precursor
were inserted between the biotin anchor and the amino-diphosphine
+
n
n
[
Rh(COD)(Biot-1)] displayed good enantioselectivity, in favor of
moieties 1 and 2 to afford ligands Biot-3 -1 and Biot-3 -2 as
q
q
11
the (R)-enantiomer (92% ee, entry 5). The catalyst derived from
well as Biot-4 -1 and Biot-4 -2 (Scheme 1).
+
[
Rh(COD)(Biot-2)] performed sluggishly to afford acetamidoala-
Using avidin as the host protein, introduction of a glycine spacer
nine in 20% ee (S) with only 55% conversion after 15 h (entry 6).
To optimize the enantioselectivity of the reduction, two comple-
mentary strategies can be pursued: chemical modification of the
first coordination sphere of the biotinylated catalyst or genetic
modification of the gene coding the protein. So far, our efforts have
focused primarily on the introduction of an amino acid spacer
between biotin and the aminodiphosphine. This allows one to probe
the topography of the host protein’s binding pocket in search of a
has a beneficial effect both on the activity and on the selectivity of
catalysts [Rh(COD)(Biot-3 -1)] and [Rh(COD)(Biot-3 -2)]
(60% ee (S) and 66% ee (S), respectively, entries 7 and 8). In
streptavidin, the sense of enantioselection for catalysts [Rh(COD)-
1
+
1
+
1
+
1
+
(Biot-3 -1)] and [Rh(COD)(Biot-3 -2)] is inverted as compared
to the catalysts devoid of spacer (24% ee (S) and 42% ee (R),
respectively, entries 9 and 10). The best result for avidin as the
2
+
host protein is obtained with [Rh(COD)(Biot-3 -1)] (69% ee (S),
9030
9
J. AM. CHEM. SOC. 2003, 125, 9030-9031
10.1021/ja035545i CCC: $25.00 © 2003 American Chemical Society

C O MMU N I C A T I O N S
entry 11). For streptavidin, [Rh(COD)(Biot-4ortho-1)]+ catalyzes
the reduction of acetamidoacrylic acid to (R)-acetamidoalanine in
sphere to produce catalysts with exquisite activity and selectivity.
The artificial metalloenzymes presented herein offer an attractive
way to exploit the second coordination sphere provided by a host
protein to produce versatile enantioselective catalysts with features
reminiscent both of enzymatic and of organometallic catalysts. We
have demonstrated that the enantioselectivity may be optimized
either chemically or genetically (i.e., chemogenetic), thus offering
an ideal scaffold for high-throughput optimization of enantiose-
lective catalysts.
Finally, these hybrid enantioselective catalysts offer interesting
perspectives toward “greener” organometallic catalysis as they
operate in water and should prove easy to recycle either by
immobilization or by size-selective filtration.
9
2% ee (entry 12). All other ligand-spacer-protein combinations
11
yielded acetamidoalanine with an ee < 40%.
The catalytic experiments reported herein reveal several note-
worthy features:
(i) As the host protein, streptavidin is generally a better chiral
inducer than avidin. Good levels of enantioselection, that match or
exceed those obtained with catalysts devoid of spacers, can be
1
2
achieved with either a glycine 3 or a â-alanine 3 spacer for avidin
and an anthranilic acid 4
ortho
spacer in streptavidin. Although both
enantiomers can be obtained with both proteins, streptavidin
produces preferentially the (R)-enantiomer, and avidin produces
preferentially the (S)-enantiomer.
Acknowledgment. This work was funded by the Swiss National
Science Foundation. We acknowledge the preliminary experiments
performed by E. Joseph and D. Berdat. We thank Belovo Egg
Science and Technology for a generous gift of egg white avidin.
We thank C. R. Cantor for the streptavidin gene, J.-M. Neuhaus,
P. Sch u¨ rmann (University Neuch aˆ tel), and P. Arosio (University
Milano) for their help in setting up the streptavidin production, as
well as Z. Lei (University Berne) for help with the mutagenesis
experiments.
(
ii) The flexible ligand skeleton 1 generally catalyzes the
reduction with a higher enantioselectivity than the more rigid ligand
skeleton 2.13 Along the lines of the quadrant rule,1b we suggest
that the conformation of the biotinylated chelating ligand (which
exists as a racemic mixture in solution, but may be biased in favor
of one enantiomer, λ or δ, upon incorporation in the host protein)
plays a determinant role in the enantioselection event.
(iii) Buffer-screening experiments reveal that the fastest reduction
rates and slightly higher enantioselectivities are obtained at neutral
pH for avidin (0.1 M MOPS, pH 7.0 or 0.07 M phosphate buffer,
pH 7.1) and acidic pH for streptavidin (0.1 M acetate buffer, pH
Supporting Information Available: Materials and methods, comple-
mentary data tables and figures (PDF). This material is available free
of charge via the Internet at http://pubs.acs.org.
4
.0 or 0.1 M MES, pH 6.0). This suggests that the Coulomb
repulsion between the biotinylated catalyst and the cationic protein
is not responsible for the modest performance with avidin as a host
protein.
References
(
1) (a) Jacobsen, E. N., Pfaltz, A., Yamamoto, H., Eds. ComprehensiVe
Asymmetric Catalysis; Springer: Berlin, 1999; Vols. I-III. (b) Knowles,
W. S. Angew. Chem., Int. Ed. 2002, 41, 1998. (c) Sharpless, K. B. Angew.
Chem., Int. Ed. 2002, 41, 2024. (d) Noyori, R. Angew. Chem., Int. Ed.
(iv) To test the substrate specificity of these hybrid catalysts,
the reduction of acetamidocinnamic acid was carried out using
2
002, 41, 2008. (e) Blaser, H.-U.; Malan, C.; Pugin, P.; Spindler, F.;
+
ortho
[
1
Rh(COD)(Biot-1)] ⊂streptavidin and [Rh(COD)(Biot-4
-
Steiner, H.; Studer, M. AdV. Synth. Catal. 2003, 345, 103.
+
)] ⊂streptavidin to yield (R)-acetamidophenylalanine with 86%
(2) (a) Jandeleit, B.; Schaefer, D. J.; Powers, T. S.; Turner, H. W.; Weinberg,
W. H. Angew. Chem., Int. Ed. 1999, 38, 2494. (b) Josephsohn, N. S.;
Kuntz, K. W.; Snapper, M. L.; Hoveyda, A. H. J. Am. Chem. Soc. 2001,
123, 11594. (c) Wenzel, A. G.; Jacobsen, E. N. J. Am. Chem. Soc. 2002,
ee (pH 6.0, MES buffer) and 83% ee (pH 6.0, MES buffer),
respectively (as compared to 92% for (R)-acetamidoalanine, Table
124, 12964.
1, entries 5 and 12). The substrate tolerance displayed by these
(
3) (a) Faber, K. Biotransformations in Organic Chemistry, 3rd ed.;
Springer: Berlin, 1997. (b) Drauz, K., Waldmann, H., Eds. Enzyme
Catalysis in Organic Synthesis: A ComprehensiVe Handbook; VCH:
Weinheim, 1995; Vols. I-II.
hybrid catalysts is thus more typical of homogeneous than of
enzymatic catalysis.1e Because the host protein was by no means
designed to stabilize the transition state of a hydrogenation reaction,
it is more tolerant toward substrates of varying steric requirements.
(4) (a) Fong, S.; Machajewski, T. D.; Mak, C. C.; Wong, C. H. Chem. Biol.
2
000, 7, 873. (b) Bornscheuer, U. T.; Altenbuchner, J.; Meyer, H. H.
Biotechnol. Bioeng. 1998, 58, 554. (c) May, O.; Nguyen, P. T.; Arnold,
F. H. Nat. Biotechnol. 2000, 18, 317. (d) Reetz, M. T. Tetrahedron 2002,
(v) These artificial metalloenzymes are amenable to a chemo-
5
8, 6595 and references therein.
5) Kaiser, E. T.; Lawrence, D. S. Science 1984, 226, 505.
(6) Qi, D.; Tann, C.-M.; Haring, D.; Distefano, M. D. Chem. ReV. 2001, 101,
081.
7) Reetz, M. T.; Rentzsch, M.; Pletsch, A.; Maywald, M. Chimia 2002, 56,
genetic optimization procedure. Having identified by chemical
modification the most promising organometallic fragments (entries
(
3
5
and 12), we subjected the streptavidin to site-directed mutagenesis.
(
Preliminary experiments were performed by substituting single
amino acids by a glycine4d in the flexible regions of streptavidin
721.
(8) Ohashi, M.; Koshiyama, T.; Ueno, T.; Yanase, M.; Fujii, H.; Watanabe,
Y. Angew. Chem., Int. Ed. 2003, 42, 1005.
close to the biotin-binding site. Substitution of serine 112 by a
glycine residue in the L7,8 loop of streptavidin (S112G) yields an
improved host protein for the reduction of acetamidoacrylic acid
(9) Wilchek, M., Bayer, E. A., Eds. Methods in Enzymology: AVidin-Biotin
Technology; Academic Press: San Diego, 1990; Vol. 184.
(
10) (a) Wilson, M. E.; Whitesides, G. M. J. Am. Chem. Soc. 1978, 100, 306.
(
b) Lin, C.-C.; Lin, C.-W.; Chan, A. S. C. Tetrahedron: Asymmetry 1999,
+
both with [Rh(COD)(Biot-1)] ⊂streptavidin S112G and with
10, 1887.
ortho
+
(11) Materials and Methods are available free of charge as Supporting
Information.
[
9
Rh(COD) (Biot-4 -1)] ⊂streptavidin S112G (96% ee (R) and
4% ee (R), entries 13 and 14, respectively).
(12) (a) Pazy, Y.; Kulik, T.; Bayer, E. A.; Wilchek, M.; Livnah, O. J. Biol.
Chem. 2002, 277, 30892. (b) Stayton, P. S.; Freitag, S.; Klumb, L. A.;
Chilkoti, A.; Chu, V.; Penzotti, J. E.; To, R.; Hyre, D.; Le Trong, I.;
Lybrand, T. P.; Stenkamp, R. E. Biomol. Eng. 1999, 16, 39.
In the field of homogeneous catalysis, the steric and the electronic
control of a catalytic moiety is mostly limited to the first
coordination sphere of the metal. This contrasts with enzymes which
take advantage both of the first and of the second coordination
(13) Yue, T.-Y.; Nugent, W. A. J. Am. Chem. Soc. 2002, 124, 13692.
JA035545I
J. AM. CHEM. SOC.
9
VOL. 125, NO. 30, 2003 9031