Enantioselective artificial metalloenzymes based on a bovine pancreatic polypeptide scaffold

Article abstract of DOI:10.1002/anie.200901134

Site creation: Enantioselective artificial metalloenzymes have been created by grafting a new active site onto bovine pancreatic polypeptide through the introduction of an amino acid capable of coordinating a copper(II) ion. This hybrid catalyst gave good

Full text of DOI:10.1002/anie.200901134

Angewandte
Chemie
DOI: 10.1002/anie.200901134
Enzyme Design
Enantioselective Artificial Metalloenzymes Based on a Bovine
Pancreatic Polypeptide Scaffold**
David Coquiꢀre, Jeffrey Bos, Joris Beld, and Gerard Roelfes*
The catalytic efficiency and high selectivities achieved by
natural metalloenzymes are a continuing source of inspiration
for the design of novel bioinspired catalysts. An emerging
approach for creating artificial metalloenzymes involves the
incorporation of a synthetic transition-metal catalyst into a
protein binding site. By using covalent, dative, or noncovalent
anchoring strategies, a variety of highly enantioselective
metalloenzymes have been created.[1] To date, however, the
choice of protein scaffold has been limited; only proteins such
as avidin, streptavidin, bovine serum albumin (BSA), and
apo-myoglobin, which have a sufficiently large pocket to bind
the catalyst and still leave space for the substrates, have been
applied successfully in enantioselective catalysis.[2] Yet, our
work on DNA-based catalysis has demonstrated that biomo-
lecular scaffolds that do not contain a pre-existing active site
can, in principle, also supply the chiral environment for a
catalyzed reaction.[3] Herein, we introduce a novel strategy for
artificial metalloenzymes that involves grafting of a new
active site onto a small natural peptide scaffold by introducing
nonproteinogenic amino acids capable of binding a Cu2+
ion.[4,5] The resulting metalloenzymes catalyze Diels–Alder
and Michael addition reactions in water with high enantio-
selectivities.
Bovine pancreatic polypeptide (bPP), a member of the
pancreatic polypeptide family of peptide hormones,[6,7] was
selected as the protein scaffold. bPP is 36 amino acids in
length and adopts robust and closely packed tertiary and
quaternary structures. It comprises a polyproline type II helix
(residues 1–8), a turn (9–12), and an a helix (13–31), with the
polyproline helix backfolded on the a helix (Figure 1a).[7c]
Hydrophobic interactions drive dimerization in solution to
give an antiparallel homodimer quaternary structure in
solution (dissociation constant Kd = 3.24 ꢀ 10À7 m, pH 5.0).[7,8]
For synthetic convenience, the peptide was truncated to 31
residues (bPP1–31); residues 32–36 are disordered in the
NMR structure and not involved in either tertiary- or
Figure 1. a) Molmol representation of monomeric bPP based on the
NMR structure.[7c] Residues modified in the present study (Tyr7,
Asp10, and Leu24) are labeled with the one-letter amino acid code.
b) Amino acid sequence alignment of wild-type (wt) bPP and bPPa–e
,
represented in the one-letter amino acid code. c) Schematic represen-
tation of the monomer/dimer equilibrium of bPPx. d) Nonproteino-
genic amino acids used: X: l-3-pyridylalanine; Z: l-4-pyridylalanine.
quaternary-structure formation.[7c,8] Based on a calculated
model for the dimer structure,[7a] Tyr7 was selected for
replacement by a variety of amino acids containing a
heteroaromatic side chain that are capable of binding a
transition-metal ion, namely histidine (bPPa) and 3- (bPPc–e
)
and 4-pyridylalanine (bPPb, Figure 1b and d). Residue 7 is
located at the dimer interface and the distance to its
counterpart on the other subunit would be in the right
range to achieve bidentate coordination of a metal ion. As a
result, the hydrophobic interface between the two subunits
could potentially provide a chiral pocket for binding of
reactive substrates.
[*] Dr. D. Coquiꢀre, J. Bos, Dr. G. Roelfes
Stratingh Institute for Chemistry, University of Groningen
Nijenborgh 4, 9747 AG Groningen (The Netherlands)
Fax: (+31)50-363-4296
E-mail: j.g.roelfes@rug.nl
Homepage: http://roelfes.fmns.rug.nl
J. Beld
The bPP variants bPPa–e were prepared by solid-phase
peptide synthesis on a Rink amide resin by using standard
9-fluorenylmethoxycarbonyl (Fmoc) chemistry, purified by
preparative C18 reversed-phase (RP) HPLC, and character-
ized by electrospray and MALDI-TOF mass spectrometry. In
all cases, analytical RP-HPLC of the purified peptides showed
a single peak and the experimental and calculated molecular
Laboratorium fꢁr Organische Chemie, ETH Zꢁrich
Hꢂnggerbergg HCI, 8093 Zꢁrich (Switzerland)
[**] We wish to thank Dr. W. R. Browne and A. J. Boersma for useful
suggestions. Financial support from the NRSC-Catalysis is grate-
fully acknowledged.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200901134.
Angew. Chem. Int. Ed. 2009, 48, 5159 –5162
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5159

Communications
masses were in excellent agreement.[9] In the case of bPPc, a
small amount ( ꢀ 8%) of a side product was formed, which
had a mass that was 18 amu lower than expected. The latter
was attributed to aspartimide formation at position 10.[10]
A
D10E mutation was introduced in bPPa,b,d,e to circumvent this
problem. Indeed, the side product was not observed after the
synthesis of bPPa,b,d,e. An additional L24A mutation was
included in bPPe to increase the size of the pocket around the
pyridyl-bound CuII ion.
The UV/Vis spectra of wt bPP and bPPa are similar
(lmax value at 278 nm), whereas the spectra of bPPb–e have a
strong absorption between 259 and 268 nm, which is charac-
teristic of pyridylalanine residues.[9] The circular dichroism
spectra of bPPa–e at room temperature are highly similar to
that of wild-type bPP, with minima at (210 Æ 2) nm and (222 Æ
1) nm; these values are characteristic for a polypeptide with a
significant degree of a-helix conformation. These data show
that the modifications do not cause a significant disruption of
the secondary structure.[8a] How-
Scheme 1. Asymmetric Diels–Alder and 1,4-addition reactions cata-
lyzed by bPPx–Cu(NO3)2 complexes in water.
bPPb, no significant enantioselectivity was obtained (Table 1,
entry 1). In contrast, bPPc and bPPd, gave the (À) enantiomer
of 2a in good yield and with ee values of 83 and 80%,
ever,
sedimentation-equilibrium
Table 1: Results of Diels–Alder reactions catalyzed by Cu–bPPx (Scheme 1).[a]
analytical ultracentrifugation anal-
ysis at catalytically relevant concen-
trations indicates that bPPd exists in
Entry Substrate bPPa
ee [%][b]
bPPb
bPPc
bPPd
bPPe
ee [%][b]
(conv. [%])
ee [%][b]
(conv. [%])
ee [%][b]
(conv. [%])
ee [%][b]
(conv. [%])
(conv. [%])
a
monomer/dimer equilibrium
(association constant Ka = 1.5 ꢀ
102 mÀ1 Figure 1c).[9] Truncated
1
1a
1a
1b
1c
1d
1e
1 f
<3 (96)
6 (65)[d]
83 (73)
80 (full)
74 (full)
14 (58)
52 (9)
23 (10)
8 (<3)
8 (<3)
<3 (96)
4 (94)
2[c]
3
;
<3 (32)
65 (full)
62 (3)
39 (8)
20 (3)
13 (<3)
<3 (full)
<3 (full)
3 (7)
4 (4)
3 (21)
9 (<3)
bPP (bPP1–31) has been reported
to be dimeric,[8a] so the decrease in
dimerization affinity has to be due
to the Y7X mutation. In the pres-
ence of CuII ions, the dimerization
affinity increases only slightly (Ka =
3.9 ꢀ 102 mÀ1). A titration of Cu-
(H2O)6(NO3)2 into a 1 mm solution
of bPPd, monitored spectrophoto-
metrically at 720 nm, showed
approximately stoichiometric bind-
ing of Cu2+ ions to bPPd (see
4
5
6
7
8
1g
[a] Typical conditions: 200 mm bPPx and 95 mm Cu(H2O)6(NO3)2 in 20 mm MOPS buffer (pH 6.5) for
3 days at 58C, unless noted otherwise. The ee values are the average from 2 experiments and are
reproducible within 2%. In all cases, the (À) enantiomer was found in excess. [b] For the endo isomer,
endo/exo ꢁ95:5. [c] 5 mol% Cu2+ and 11 mol% bPPd.
Figure S3d in the Supporting Information), which indicates
that the catalyst is predominantly monomeric. The binding of
a Cu2+ ion to a single pyridine is relatively weak,[11] so this
suggests that, in addition to the 3-pyridylalanine, other
residues in the peptide are also involved in the binding of
the Cu2+ ion. In view of the ultracentrifugation results, small
but significant amounts of the copper-bound dimeric bPPd are
also likely to be present under catalytic conditions, that is,
0.2 mm bPPx and 95 mm Cu2+ ions.
The catalytic potential of the peptides was evaluated by
using the Diels–Alder reaction of azachalcone 1a with
cyclopentadiene as a model system (Scheme 1). The reactions
were performed with 15 mol% of Cu(H2O)6(NO3)2 and
33 mol% of ligand bPPx (2 ligands per metal center) in
3-(N-morpholine)propanesulfonic acid (MOPS) buffer solu-
tion (20 mm, pH 6.5). After reaction for 3 days at 58C, the
Diels–Alder product 2a was obtained as the only detectable
product.
respectively. The reaction proved to be ligand accelerated;
the apparent second-order rate constants, kapp, with 0.1 mm
Cu(H2O)6(NO3)2 in the absence and presence of bPPd were
(0.048 Æ 0.0038) and (0.17 Æ 0.013) mÀ1 sÀ1, respectively.[9] This
corresponds to a 3.5-fold rate acceleration as a result of
ligation of Cu2+ ions by bPPd. Thus, the presence of any
unbound Cu2+ ions will have only a minor effect on the
ee value. Combined, these results demonstrate that a pyridine
moiety is required and, furthermore, that the enantioselec-
tivity critically depends on the structure of this group. Only in
certain cases, as with the 3-pyridyl side chain, is an active site
created that is capable of asymmetric induction.
A control experiment with the tripeptide EXP, that is,
3-pyridylalanine flanked by the same amino acids as in bPPc–e
,
gave no enantioselectivity.[9] This demonstrates the impor-
tance of the second coordination sphere supplied by the
peptide scaffold in asymmetric induction. A bPPd/Cu2+ ion
ratio of 2 or higher gave the best ee values; a ratio of less than
2 induced a progressive reduction in the ee value down to
49% for a ratio of 0.5 (see Figure S4 in the Supporting
Information). A decrease in the copper loading to 5 mol%
No conversion was found in reactions performed with wt
bPP in the absence of copper.[9] With copper, the observed
reaction was not enantioselective. Similarly, with bPPa or
5160
www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 5159 –5162

Angewandte
Chemie
and in the bPPd loading to 11mol% still gave full conversion
but with a slightly decreased ee value (Table 1, entry 2).
The scope of the reaction was investigated by using a
variety of a,b-unsaturated 2-acyl imidazole substrates,[12]
which are excellent substrates for Lewis acid catalysis in
water (Scheme 1).[3c,d] When R’ = phenyl (1b; Table 1,
entry 3), good conversion and enantioselectivity were
obtained with Cu–bPPc. Substitution of the phenyl ring with
an electron-withdrawing or -donating group led to a dramatic
reduction in the yield (Table 1, entries 4–6), although signifi-
cant ee values were still obtained with 2c and 2d, that is, 62
and 39%, respectively. Replacement of the phenyl ring with a
2-furanyl group had a detrimental effect on both conversion
and enantioselectivity (Table 1, entry 7). By contrast, excel-
lent conversion to the Diels–Alder product was observed with
R = methyl, albeit without enantioselectivity (Table 1,
entry 8). Finally, with chalcone as the substrate, < 3%
conversion was observed, which suggests that bidentate
binding of the substrate to the Cu2+ ion is required for
activity, as was observed before.[2g,3d,9]
5 mol% led to slightly lower ee value of 75% (Table 2,
entry 3). Interestingly, with bPPe, which was nonselective in
the Diels–Alder reaction, a good ee value was obtained.
Given the broad substrate scope of the analogous DNA-
based catalytic reactions,[3] the high substrate selectivity of the
bPP-based catalysts is notable. Substituents on the enone
moiety of the substrate (R’) that are too large, such as a
substituted phenyl group, lead to loss of activity and
significant reduction of enantioselectivity. Activity is restored
with smaller substituents, such as R’ = CH3, but enantiose-
lectivity is then only seen for the Michael addition. The
enolate of dimethylmalonate, which is used as the Michael
donor, is much larger than cyclopentadiene, which is used in
the Diels–Alder reaction. Hence, a tentative conclusion is
that the active site provided by the hybrid catalyst is
compatible only with certain substrate/reactant combinations,
such as 1a/1b with cyclopentadiene or 1g with the enolate of
dimethylmalonate. Incompatibility with the structure of the
active site leads to a loss of activity and/or enantioselectivity.
In this sense, the present system resembles true enzymatic
catalysts, which generally also have an active site optimized
for one reaction. Current efforts are directed towards the
elucidation of the oligomerization state and the active-site
structure of the most efficient bPP-based artificial enzymes,
which will provide more insight into the origin of the observed
enantioselectivity.
Cu–bPPd, which has the D10E mutation, displays the
same trends with the a,b-unsaturated 2-acyl imidazole sub-
strates. This was expected because the D10E mutation is
conservative and is not expected to significantly affect the
structure of the peptide. A notable exception, however, is
substrate 1b, which gave rise to a significantly decreased
ee value than that obtained with Cu–bPPc (Table 1, entry 3).
At present, the origin of this decrease is still not understood.
bPPe, which contains an additional L24A mutation, gave
similar conversions but, surprisingly, displayed a complete
loss of enantioselectivity; this result suggests a role for Leu24
in determining the enantioselectivity of the Diels–Alder
reaction.
In conclusion, we have developed a novel strategy towards
the design of artificial metalloenzymes, which involves graft-
ing of an active site onto an existing small natural protein
scaffold by incorporation of a nonproteinogenic amino acid
that is capable of binding a transition-metal ion. A key
strength of the present approach is that an existing binding
pocket in the protein is not required; this greatly expands the
choice of protein scaffolds for artificial-metalloenzyme
design. By using bPP-based catalysts, good enantioselectiv-
ities were obtained in the Cu2+-catalyzed Diels–Alder and
Michael addition reactions in water, that is, up to 83 and
86% ee, respectively. A particularly interesting feature of the
present system is the high substrate selectivity, which is
reminiscent of natural enzymes.
Encouraged by these results, we decided to explore the
potential of these hybrid enzymes in the catalytic asymmetric
Michael addition reaction in water (Scheme 1). The 1,4-
addition reaction was performed by using dimethylmalonate
as the nucleophile. In marked contrast to the Diels–Alder
reaction, a modest conversion and no enantioselectivity were
observed with substrate 1b (Table 2, entry 1) with Cu–bPPc–e
.
However, good conversion and high ee values were obtained
with 1g, up to 86% ee for product 3g in the case of Cu–bPPd
(Table 2, entry 2). A decrease in the catalyst loading to
Experimental Section
Representative procedure for bPPx–Cu2+-catalyzed reactions: An
aqueous solution of Cu(H2O)6(NO3)2 (24 mL, 1 mm) was added to
bPPd (200 mm, 250 mL) in 20 mm MOPS buffer (pH 6.5) at 08C. A
fresh stock solution (5 mL) of substrate in CH3CN was added. After
addition of freshly distilled cyclopentadiene (1 mL) or dimethylmal-
onate (3 mL) at 08C, the reaction was mixed for 3 days by continuous
inversion at 58C. The product was isolated by extraction with Et2O
(2 ꢀ 1.5 mL). The organic phases were dried (Na2SO4) and evaporated
under reduced pressure to give the product. The conversion and
ee value were determined by RP-HPLC.
Table 2: Results of Michael addition reactions catalyzed by Cu–bPPx
(Scheme 1).[a]
Entry
Substrate
bPPc
bPPd
bPPe
ee [%][b]
(conv. [%])
ee [%][b]
(conv. [%])
ee [%][b]
(conv. [%])
1
2
1b
1g
1g
1h
<3 (25)
66 (70)
<3 (59)
86 (85)
75 (70)
6 (54)
<3 (15)
65 (90)
3[c]
4
<3 (10)
Received: February 27, 2009
Revised: April 10, 2000
[a] Typical conditions: 200 mm bPPx and 95 mm Cu(H2O)6(NO3)2
(15 mol%; bPPx to CuII ratio: 2.1) in 20 mm MOPS buffer (pH 6.5) for
3 days at 58C, unless noted otherwise. [b] The ee values are the average
of 2 experiments and are reproducible within 2%. [c] 5 mol% Cu2+ and
11 mol% bPPd.
Keywords: artificial metalloenzymes · asymmetric catalysis ·
.
copper · Diels–Alder reactions · Michael addition
Angew. Chem. Int. Ed. 2009, 48, 5159 –5162
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5161

Communications
B. Breit, Chem. Eur. J. 2008, 14, 4488 – 4502; b) C. A. Christen-
sen, M. Meldal, J. Comb. Chem. 2007, 9, 79 – 85; c) A. Agarkov,
S. Greenfield, D. Xie, R. Pawlick, G. Starkey, S. R. Gilbertson,
Biopolymers 2005, 84, 48 – 73; d) W. F. DeGrado, C. M. Summa,
V. Pavone, F. Nastri, A. Lombardi, Annu. Rev. Biochem. 1999,
68, 779 – 819.
[1] a) J. Steinreiber, T. R. Ward, Coord. Chem. Rev. 2008, 252, 751 –
766; b) M. D. Mihovilovic, J. Chem. Technol. Biotechnol. 2007,
82, 1067 – 1071; c) M. T. Reetz, Proc. Natl. Acad. Sci. USA 2004,
101, 5716 – 5722.
[2] a) M. E. Wilson, G. M. Whitesides, J. Am. Chem. Soc. 1978, 100,
306 – 307; b) T. Kokubo, T. Sugimoto, T. Uchida, S. Tanimoto, M.
Okano, J. Chem. Soc. Chem. Commun. 1983, 769 – 770; c) J.
Collot, J. Gradinaru, N. Humbert, M. Skander, A. Zocchi, T. R.
Ward, J. Am. Chem. Soc. 2003, 125, 9030 – 9031; d) M. Ohashi, T.
Koshiyama, T. Ueno, M. Yanase, H. Fujii, Y. Watanabe, Angew.
Chem. 2003, 115, 1035 – 1038; Angew. Chem. Int. Ed. 2003, 42,
1005 – 1008; e) J. R. Carey, S. K. Ma, T. D. Pfister, D. K. Garner,
H. K. Kim, J. A. Abramite, Z. L. Wang, Z. J. Guo, Y. Lu, J. Am.
Chem. Soc. 2004, 126, 10812 – 10813; f) A. Mahammed, Z. Gross,
J. Am. Chem. Soc. 2005, 127, 2883 – 2887; g) M. T. Reetz, N. Jiao,
Angew. Chem. 2006, 118, 2476 – 2479; Angew. Chem. Int. Ed.
2006, 45, 2416 – 2419; h) T. Hayashi, D. Murata, M. Makino, H.
Sugimoto, T. Matsuo, H. Sato, Y. Shiro, Y. Hisaeda, Inorg. Chem.
2006, 45, 10530 – 10536; i) M. T. Reetz, M. Rentzsch, A. Pletsch,
M. Maywald, P. Maiwald, J. J. P. Peyralans, A. Maichele, Y. Fu, N.
Jiao, F. Hollmann, R. Mondiꢁre, A. Taglieber, Tetrahedron 2007,
63, 6404 – 6414; j) J. Pierron, C. Malan, M. Creus, J. Gradinaru, I.
Hafner, A. Ivanova, A. Sardo, T. R. Ward, Angew. Chem. 2008,
120, 713 – 717; Angew. Chem. Int. Ed. 2008, 47, 701 – 705; k) A.
Pordea, M. Creus, J. Panek, C. Duboc, D. Mathis, M. Novic, T. R.
Ward, J. Am. Chem. Soc. 2008, 130, 8085 – 8088.
[5] For an example of a DNA-cleaving protein containing a (2,2’-
bipyridin-5-yl)alanine–copper complex, see: H. S. Lee, P. G.
Schultz, J. Am. Chem. Soc. 2008, 130, 13194 – 13195.
[6] a) A. Bettio, A. G. Beck-Sickinger, Biopolymers 2001, 60, 420 –
437; b) A. G. Beck-Sickinger, H. A. Wieland, H. Wittneben,
K. D. Willim, K. Rudolf, G. Jung, Eur. J. Biochem. 1994, 225,
947 – 958; c) I. D. Glover, D. J. Barlow, E. J. Pitts, S. P. Wood, I. J.
Tickle, T. L. Blundell, K. Tatemoto, J. R. Kimmel, A. Wollmer,
W. Strassburger, Eur. J. Biochem. 1984, 142, 379 – 385.
[7] a) M. Lerch, V. Gafner, R. Bader, B. Christen, G. Folkers, O.
Zerbe, J. Mol. Biol. 2002, 322, 1117 – 1133; b) N. J. Zondlo, A.
Schepartz, J. Am. Chem. Soc. 1999, 121, 6938 – 6939; c) X. Li,
M. J. Sutcliffe, T. W. Schwartz, C. M. Dobson, Biochemistry
1992, 31, 1245 – 1253; d) T. L. Blundell, J. E. Pitts, I. J. Tickle,
S. P. Wood, C. W. Wu, Proc. Natl. Acad. Sci. USA 1981, 78, 4175 –
4179.
[8] a) A. J. Nicoll, R. K. Allemann, Org. Biomol. Chem. 2004, 2,
2175 – 2180; b) S. E. Taylor, T. J. Rutherford, R. K. Allemann, J.
Chem. Soc. Perkin Trans. 2 2002, 751 – 755; c) S. E. Taylor, T. J.
Rutherford, R. K. Allemann, Bioorg. Med. Chem. Lett. 2001, 11,
2631 – 2635.
[3] a) N. Sancho Oltra, G. Roelfes, Chem. Commun. 2008, 6039 –
6041; b) A. J. Boersma, J. E. Klijn, B. L. Feringa, G. Roelfes, J.
Am. Chem. Soc. 2008, 130, 11783 – 11790; c) D. Coquiꢁre, B. L.
Feringa, G. Roelfes, Angew. Chem. 2007, 119, 9468 – 9471;
Angew. Chem. Int. Ed. 2007, 46, 9308 – 9311; d) A. J. Boersma,
B. L. Feringa, G. Roelfes, Org. Lett. 2007, 9, 3647 – 3650; e) G.
Roelfes, B. L. Feringa, Angew. Chem. 2005, 117, 3294 – 3296;
Angew. Chem. Int. Ed. 2005, 44, 3230 – 3232.
[9] See the Supporting Information.
[10] G. Barany, R. B. Merrifield, The Peptides, Vol. 2 (Eds.: E. Gross,
J. Meienhofer), Academic Press, New York, 1979, pp. 1 – 234.
[11] J. Bjerrum, Chem. Rev. 1950, 46, 381 – 401.
[12] a) D. A. Evans, K. R. Fandrick, H. J. Song, J. Am. Chem. Soc.
2005, 127, 8942 – 8943; b) D. A. Evans, H. J. Song, K. R. Fan-
drick, Org. Lett. 2006, 8, 3351 – 3354; c) M. C. Myers, A. R.
Bharadwaj, B. C. Milgram, K. A. Scheidt, J. Am. Chem. Soc.
2005, 127, 14675 – 14680.
[4] For other approaches to peptide-based ligands for transition-
metal catalysis, see: a) A. C. Laungani, J. M. Slattery, I. Krossing,
5162
www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 5159 –5162