A Site-Selective Dual Anchoring Strategy for Artificial Metalloprotein Design
James R. Carey,†Steven K. Ma,†Thomas D. Pfister,†Dewain K. Garner,†Hyeon K. Kim,†
Joseph A. Abramite,†Zhilin Wang,‡Zijian Guo,‡and Yi Lu*,†
Department of Chemistry, UniVersity of Illinois at Urbana-Champaign, Urbana, Illinois 61801, and Coordination
Chemistry Institute, Nanjing UniVersity, Nanjing 210093, P.R. China
Received May 25, 2004; E-mail: firstname.lastname@example.org
Introducing nonnative metal ions or metal-containing prosthetic
groups into a protein is a branch of metalloprotein design with
considerable impact, as it can dramatically expand the repertoire
of protein functionalities and thus their range of applications.1-9
Since protein scaffolds have not evolved to tightly bind artificial
metal complexes in a single conformation, one of the most
challenging aspects of this field is control of chemo- and/or
enantioselectivity.8-11To meet the challenges, two approaches,
noncovalent12-14and single-point covalent attachments,8,9have been
successfully demonstrated, either by altering the metal complex and
protein to promote binding in the noncovalent case12-14,15or by
carefully selecting a protein host for covalent attachment of
nonnative metal complexes.16Here we report a novel site-selective
two-point covalent attachment strategy to introduce an achiral
manganese salen complex (Mn(salen)), into apo sperm whale
myoglobin (Mb) (Figure 1) and to demonstrate the effectiveness
of the dual anchoring approach to markedly improve the enantio-
selectivity of a semi-synthetic enzyme with minimal structural
modification to either the metal complex or the protein.
In an attempt to expand the functionality of heme proteins, we12,13
and Watanabe and co-workers14,15have reported the replacement
of heme with noncovalently attached achiral Mn(salen) and
Cr(salen) complexes in the chiral pocket of cytochrome c peroxidase
(CcP) or Mb, respectively. Metallo-salens are known for their high
catalytic activity and versatility,17and their planar nature makes
them an ideal choice for introduction into heme proteins. Despite
their similarity, replacement of heme with metallo-salens has
proven challenging. Noncovalent placement of a Cr(salen) inside
the heme binding pocket requires the use of modified salens to
increase the binding affinity, which results in a low yield of the
Cr(salen) protein complex (0.3-15%) and low ee (0.3-13% ee).14
Further, the method that successfully incorporated Cr(salen) into
Mb is reported to be difficult to apply to Mn(salen).13,14
Covalent linkage is an alternative approach for site-specific
attachment of an artificial metallo complex to a protein with high
yield (∼100%) and with minimal structural information. Distefano
and co-workers have demonstrated that covalent attachment of a
Cu(II)/1,10-phenanthroline complex to a single cysteine in an
adipocyte lipid binding protein results in a catalyst that promotes
highly enantioselective hydrolysis.8,16We employed such a strategy
to incorporate Mn(salen) into apo-Mb(Y103C) by methane thio-
sulfonate groups (Figure 1B) with high selectivity and reactivity
toward cysteines.18This catalyst shows sulfoxidation activity (Table
1, entry 7); however, the ee remains low (12%) (Table 1).14
Similarly low ee (<10%) is reported by Reetz and co-workers for
a Mn(salen) complex attached to papain via a single maleimide
linker.19The low ee of the single attachment again suggests that
multiple orientations of the metallo complex may be possible. From
these results we hypothesized that an improved catalyst could be
engineered using a dual anchoring strategy, which allows for precise
control of the placement of the artificial metallo complex with
specific orientation and limited rotational freedom.
Apo-Mb was chosen as a chiral scaffold because it has been
determined to be essentially a folded globular protein20and has
been used as a template for engineering numerous artificial
proteins.4,7,14,21,22To covalently attach Mn(salen) to Mb, we again
utilized methane thiosulfonate groups (Figure 1).18By modeling
ethanediamino-manganese(III) bromide (1) molecule (Figure 1B)
into Mb using the InsightII (Accelrys) program and searching for
its best fit in the heme pocket, we identified two mutations (L72C
and Y103C) that would selectively anchor the compound (Figure
†Department of Chemistry, University of Illinois - Urbana-Champaign.
‡Coordination Chemistry Institute, Nanjing University.
Figure 1. (A) Computer model of Mb(L72C/Y103C) with 1 covalently
attached overlayed with heme. (B) Complex 1.
Table 1. Enantioselective Sulfoxidation of Thioanisole
apo-WTMb + 1
1 (0.5), S
1 (1.5), S
3 (2), R
3 (2), S
0 (0.5), S
12 (1), S
51 (1), S
aIn 50 mM NH4OAc pH 5.1 at 4 °C, 130 µM catalyst, thioanisole (5
mM), H2O2(6.5 mM) (for details see Supporting Information).bThe unit
of the rate is 10-3turnover min-1, std dev reported in parentheses.cReaction
rates and ee were determined by GC analysis using an ASTEC G-TA
cyclodextrin column, and acetophenone was added as an internal standard.25
dAttachment method (n ) noncovalent, s ) single, d ) double).eCr-2 )
Cr-5,5′-tBu2-salophen (from ref 14).
Published on Web 08/17/2004
10812 9 J. AM. CHEM. SOC. 2004, 126, 10812-10813
10.1021/ja046908x CCC: $27.50 © 2004 American Chemical Society
Construction, expression, and purification of Mb mutant proteins
were performed as described previously.21,23Addition of 1 to this
mutant results in the catalyst called Mn‚1‚apo-Mb(L72C/Y103C).
The UV-vis spectrum of this catalyst shows new absorption peaks
at 284 and 292 nm, suggesting the formation of a protein-salen
adduct (Figure 2A). Further confirmation of the adduct formation
comes from electrospray mass spectroscopy (ESI-MS), which shows
a single peak corresponding to the apo enzyme plus 1 minus two
methyl thiosulfonate groups displaced by covalent attachment
(measured: 17731.68 ( 2.92 Da; calcd: 17731.12 Da) (Figure 2
B). The mass spectrum also suggests a complete conversion from
apo enzyme to Mn(salen)-modified enzyme. These results are
consistent with elemental analysis data and the absence of free
cysteines by the Ellman’s test.24
Comparison of the dual anchor Mn‚1‚apo-Mb(L72C/Y103C)
catalyst with previously reported noncovalent and single-point
attached protein metallo-salen complexes, as well as various
controls with apo protein, 1 only, and noncovalent and single-point
attached catalysts reported herein demonstrate the advantages of
the dual anchor strategy. Control experiments (Table 1, entry 1-5)
show that 1 in water and heme-free apoproteins employed in this
study all displayed slow reaction rates and low ee (0-3%). This is
not surprising since, in water and without a chiral center (in the
case of 1) or without a cofactor (for the apoproteins), they have
previously been shown to be ineffective asymmetric catalysts.12-14
When 1 was added to apo-WTMb without covalent attachment,
the reaction rate and ee are also low (Table 1, entry 5); consistent
with observations made by Watanabe and co-workers.14In their
experiments, increased yield and ee were obtained by modifying
both the Cr(salen) complex (tert-butyl substitution) and the Mb
substrate binding pocket (H64D/A71G mutations). When we
incorporated 1 into apo-Mb(Y103C) through a single-point attach-
ment with no further modification of the protein, metal, or salen,
we observed an increased rate (0.051 min-1) and higher ee (12.3%).
These results are comparable to the rate (0.078 min-1) and ee (13%)
reported for the noncovalent strategy used by Watanabe and co-
workers (Table 1, entry 6),14indicating that covalent attachment is
a comparable alternative to the noncovalent attachment strategy
for controlling selectivity. More significantly, when 1 was incor-
porated into Mb through the site-selective dual anchoring strategy
using apo-Mb(L72C/Y103C) (Figure 1A), a significant increase in
rate (0.390 min-1) and ee (51%) were observed (Table 1, entry 8).
These results strongly suggest that the dual covalent anchoring
strategy limits the number of conformational states available to the
metallo complex inside the protein and by so doing improves both
rate of reaction and ee.
In conclusion, we have demonstrated for the first time the
effectiveness of a site-selective dual anchoring strategy for attach-
ment of artificial metal complexes to proteins. Dual covalent
anchoring of the metal complex to the protein affects catalytic
sulfoxidation of thioanisole with much higher ee and rate than either
noncovalent or single-point covalent attachment strategies. Since
these results were obtained without mutation or selectively modified
salen groups, the dual anchoring method provides an excellent
platform from which to obtain even higher rate and ee. The method
can be generally applied to protein incorporation of other metal
complexes with minimal structural information and is a promising
approach for generating new artificial enzymes.
Acknowledgment. We thank Ms. Brook DeMoisy and Mr. Evan
Brower for help in preparing proteins, and the U.S. National
Institutes of Health and National Science Foundation of China for
Supporting Information Available:
sulfoxidation and the synthesis of 1. This material is available free of
charge via the Internet at http://pubs.acs.org.
Experimental details for
(1) DeGrado, W. F.; Summa, C. M.; Pavone, V.; Nastri, F.; Lombardi, A.
Annu. ReV. Biochem. 1999, 68, 779-819.
(2) Hellinga, H. W. Curr. Opin. Biotechnol. 1996, 7, 437-441.
(3) Kennedy, M. L.; Gibney, B. R. Curr. Opin. Struct. Biol. 2001, 11, 485-
(4) Lu, Y.; Berry, S. M.; Pfister, T. D. Chem. ReV. 2001, 101, 3047-3080.
(5) Farrer, B. T.; Pecoraro, V. L. Curr. Opin. Drug DiscoVery DeV. 2002, 5,
(6) Hoffman, B. M. In The Porphyrins; Dolphin, D., Ed.; Academic Press:
New York, 1979; Vol. VII, pp 403-444.
(7) Hayashi, T.; Hisaeda, Y. Acc. Chem. Res. 2002, 35, 35-43.
(8) Qi, D.; Tann, C.-M.; Haring, D.; Distefano, M. D. Chem. ReV. 2001, 101,
(9) Reetz, M. T. Tetrahedron 2002, 58, 6595-6602.
(10) Miller, V. P.; DePillis, G. D.; Ferrer, J. C.; Mauk, A. G.; Ortiz de
Montellano, P. R. J. Biol. Chem. 1992, 267, 8936-8942.
(11) Jin, S.; Makris, T. M.; Bryson, T. A.; Sligar, S. G.; Dawson, J. H. J. Am.
Chem. Soc. 2003, 125, 3406-3407.
(12) Ma, S. K.; Lu, Y. J. Inorg. Biochem. 1999, 74, 217.
(13) Ma, S. K. Ph.D. Thesis, University of Illinois - Urbana-Champaign:
(14) (a) Ohashi, M.; Koshiyama, T.; Ueno, T.; Yanase, M.; Fujii, H.; Watanabe,
Y. Angew. Chem., Int. Ed. 2003, 42, 1005-1008. (b) Wilson, M. E.;
Whitesides, G. M. J. Am. Chem. Soc. 1978, 100, 306-307. (c) Collot, J.;
Gradinaru, J.; Humbert, N.; Skander, M.; Zocchi, A.; Ward, T. R. J. Am.
Chem. Soc. 2003, 125, 9030-9031.
(15) Ueno, T.; Ohashi, M.; Kono, M.; Kondo, K.; Suzuki, A.; Yamane, T.;
Watanabe, Y. Inorg. Chem. 2004, 43, 2852-2858.
(16) Davies, R. R.; Distefano, M. D. J. Am. Chem. Soc. 1997, 119, 11643-
(17) Srinivasan, K.; Michaud, P.; Kochi, J. K. J. Am. Chem. Soc. 1986, 108,
(18) Smith, D. J.; Miggio, E. T.; Kenyon, G. L. Biochemistry 1975, 14, 766-
(19) Reetz, M. T.; Rentzsch, M.; Pletsch, A.; Maywald, M. Chimia 2002, 56,
(20) Eliezer, D.; Wright, P. E. J. Mol. Biol. 1996, 263, 531-538.
(21) Springer, B. A.; Sligar, S. G. Proc. Natl. Acad. Sci. U.S.A. 1987, 84, 8961-
(22) Watanabe, Y. Curr. Opin. Chem. Biol. 2002, 6, 208-216.
(23) Sigman, J. A.; Kwok, B. C.; Lu, Y. J. Am. Chem. Soc. 2000, 122, 8192-
(24) Ellman, G. L. Arch. Biochem. Biophys. 1959, 82, 70-77.
(25) Ozaki, S.-I.; Ortiz de Montellano, P. R. J. Am. Chem. Soc. 1995, 117,
Figure 2. (A) UV-vis spectra of apo-Mb(L72C/Y103C) (red dashed line)
and Mn‚1‚apo-Mb(L72C/Y103C) (blue solid line) in 50 mM ammonium
acetate buffer pH 5.1. (B) ESI-MS of apo-Mb(L72C/Y103C) and Mn‚1‚
C O MMU NI C A T I O NS
J. AM. CHEM. SOC. 9 VOL. 126, NO. 35, 2004 10813