Phage display selection of peptides possessing aldolase activity

Article abstract of DOI:10.1039/b100479b

Peptides catalyzing retro-aldol reactions were selected from a phage display peptide library using 1,3-diketones designed for the covalent selection of an enamine-based reaction mechanism.

Full text of DOI:10.1039/b100479b

View Article Online / Journal Homepage / Table of Contents for this issue
Phage display selection of peptides possessing aldolase activity
Fujie Tanaka and Carlos F. Barbas III*
The Skaggs Institute for Chemical Biology and the Department of Molecular Biology, The Scripps Research Institute,
10550 North Torrey Pines Road, La Jolla, California 92037, USA. E-mail: carlos@scripps.edu;
Fax: +1-858-784-2583; Tel: +1-858-784-9098
Received (in Corvallis, OR, USA) 10th January 2001, Accepted 1st March 2001
First published as an Advance Article on the web 3rd April 2001
Peptides catalyzing retro-aldol reactions were selected from
a phage display peptide library using 1,3-diketones designed
for the covalent selection of an enamine-based reaction
mechanism.
natural 20 amino acids) was selected by three rounds of reaction
with 1,3-diketones, 1-BSA4a (bovine serum albumin) and
The folded states of enzymes are key to their catalytic activity
since structure can be used productively to modify the chemical
reactivity of amino acid side chains, for example with
electrostatic or hydrophobic effects, to build isolated reaction
vessels, and to stabilize transition states of reactions. Small
peptides are limited in their potential to catalyze reactions since
they typically do not adopt well-defined structures. While the
design of folded peptides has advanced tremendously in recent
years, design at the level of catalysis is a challenging task.
Rationally designed small catalytic peptides (14–42 amino acid
residues peptides) have been reported that catalyze the dec-
arboxylation of oxaloacetic acid,1 ester hydrolysis,2 and
transesterifications.2 Rational design, however, is limited in
throughput.1b The combination of rational design with library
based methods may be an attractive approach to catalyst
development. Phage display allows for libraries of greater than
109 peptides to be probed for function, but this approach has not
been used to select for catalytic peptides.3 Here we report the
applicability of a mechanism-based covalent selection of
catalytic peptides in vitro. Peptides possessing aldolase activity
(aldolase peptides) that use an enamine-type mechanism were
selected by reaction of peptide phage with 1,3-diketones that
have been employed for the induction of aldolase antibodies4
(Scheme 1).
Amine catalyzed decarboxylation and aldol reactions share
imine and enamine intermediates on their reaction coordinates.
Accordingly, enzyme and antibody catalysts of these reactions
are generally bifunctional catalysts capable of accelerating both
decarboxylation and aldol reactions.5 In order to select aldolase
peptides that use an imine–enamine reaction mechanism, we
sought to exploit mechanistic homology by using as our starting
point an 18-residues peptide (YLK-18-opt)1b that catalyzes the
decarboxylation of oxaloacetic acid. This peptide contains
multiple lysyl residues that are key to its activity. To evolve the
catalytic activity of this peptide, a 6-amino acid residue library
was appended to the C-terminus of the peptide and the library
was displayed on phage.6 The phage display peptide library
2-BSA.4c Screening of selected phage by ELISA identified
clones that bound to 1- and 2-BSA. The two most reactive
peptides as judged by ELISA, FT-YLK-3 and FT-YLK-4, were
chemically synthesized (C-terminal amide). Both peptides
catalyzed the retro-aldol reactions of fluorogenic substrates
(±)-34b and (±)-anti-44c in 5% CH3CN–42.5 mM Na phosphate
(pH 7.5). Since the solubility of the peptide FT-YLK-4 was low
in buffer (around neutral pH) and prevented exact kinetic
analyses, FT-YLK-3 was characterized in detail.
YLK-18-opt
FT-YLK-3
FT-YLK-4
YKLLKELLAKLKWLLRKL-NH2
YKLLKELLAKLKWLLRKLLGPTCL-NH2
YKLLKELLAKLKWLLRKLSDHLCL-NH2
FT-YLK-3-R5 YRLLRELLARLRWLLRRLLGPTCL-NH2
FT-YLK-3-23S YKLLKELLAKLKWLLRKLLGPTSL-NH2
The catalyzed reaction displayed saturation kinetics as
described by the Michaelis–Menten equation. The parameters
of FT-YLK-3 and of the template peptide YLK-18-opt are
shown in Table 1.7 The Km values were in mM range. The rate
acceleration above the buffer catalyzed background reaction
(kcat/kuncat) of the selected peptide FT-YLK-3 was 1400 and 500
for (±)-3 and (±)-anti-4, respectively, and was superior to the
template peptide YLK-18-opt. FT-YLK-3 demonstrated moder-
ate enantioselectivity in the retro-aldol reaction of 3. Studies
performed using FT-YLK-3 (100 mM) and each of the
enantiomers of 3 (1 mM) in 10% CH3CN–40 mM Na phosphate
(pH 7.5), revealed that the ratio of the initial velocities of FT-
YLK-3-catalyzed reactions of (R)-3+(S)-3 was 1.8+1.0.8
YKLLKELLAKLKWLLRKLXXXXXX (X
= any of the
In order to probe the catalytic mechanism of the peptide,
reactivity and mutational studies were performed. The catalyzed
reaction was inhibited by the addition of diketone 5. When FT-
YLK-3 (100 mM) was mixed with diketone 5 (2 mM) at pH 7.5
for 2.5 h at rt and the retro-aldol activity assessed with (±)-3 (1.2
mM) in 5% CH3CN–42.5 mM Na phosphate (pH 7.5), the
velocity of the retro-aldol reaction was < 15% that observed in
the absence of 5. Further, FT-YLK-3 formed a UV-observable
enaminone with pentane-2,4-dione (6) when FT-YLK-3 (100
mM) was mixed with 6 (5 mM) as indicated by the appearance
of a new absorption at 318 nm.4 The pH profile of the velocity
of the enaminone formation indicates that FT-YLK-3 has a pKa
Scheme 1 (a) Enaminone formation of peptide catalyst (Cat) with
1,3-diketone. (b) An enamine mechanism for the retro-aldol reaction.
DOI: 10.1039/b100479b
Chem. Commun., 2001, 769–770
This journal is © The Royal Society of Chemistry 2001
769

View Article Online
Table 1 Kinetic parameters of the retro-aldol reactionsa
(±)-3
(±)-anti-4
Km, mM
Peptide
Km, mM
kcat, min21
kcat/kuncat
kcat, min21
kcat/kuncat
YLK-18-opt
FT-YLK-3
1.8
1.8
2.1 3 1024
5.6 3 1024
540
1400
0.9
1.1
4.1 3 1024
1.2 3 1023
170
500
a Reaction conditions: [peptide] 100 mM, 5% CH3CN–42.5 mM Na phosphate (pH 7.5) for (±)-3 and, 5% DMSO–42.5 mM Na phosphate (pH 7.5) for
(±)-anti-4, 25 °C. The first-order kinetic constant of the background reaction (kuncat) was 3.9 3 1027 for (±)-3 and 2.4 3 1026 min21 for (±)-anti-4. The
reaction was followed by monitoring the increase in fluorescence (lex 330 nm, lem 452 nm).
lower than that typical of a lysine e-amino group or a N-terminal
amino group. The lowest pKa is approximately 5.5. These
results are consistent with peptide catalysis of the retro-aldol
reaction using an enamine mechanism and chemically reactive
amino group(s).
by a disulfide linkage did not preferentially form. The role of the
Cys residue was assessed by synthesis of the Ser analog FT-
YLK-3-23S. FT-YLK-3-23S catalyzed the retro-aldol reactions
described above and the velocity of the catalyzed reaction of
(±)-3 (2 mM) was 75% that of FT-YLK-3. The ratio of the mean
residue ellipticity of FT-YLK-3-23S at 208 and 222 nm were
0.8 and this ratio is the same as that observed in FT-YLK-3. The
a-helix content of FT-YLK-3-23S was ~ 50%. The reduction in
a-helical of FT-YLK-3-23S was correlated with a reduction in
retro-aldolase activity. These results suggest that improved
catalytic activity is intimately linked with stabilization of the
conformation of the peptide.
In this study, we demonstrate that phage display and a
compound designed to covalently trap catalysts that operate by
a predefined mechanism can be used to directly select for small
structured peptides that catalyze the aldol reaction. The
coupling of phage display with this type of selection strategy
provides a direct selection for peptide folding. For the aldolase
peptide reported here, folding and catalytic activity were
intimately linked.
Since the low pKa nucleophilic amino group of LysH93 of
aldolase antibodies 38C2 and 33F12 can be specifically
covalently modified with lactam 7,9 FT-YLK-3 was treated
with 7 to further study amino group reactivity. FT-YLK-3 (100
mM) was mixed with 7 (1 mM) and analyzed by MALDI-mass
at 30 min, 2 h, and 6 h. Analysis indicated a time-dependent
modification at multiple sites. At 30 min, 0–4 additions of 7 per
molecule of FT-YLK-3 were observed while 1–5 and 1–7
modifications were observed at 2 and 6 h, respectively. In
addition to lysine and a free amino terminus, FT-YLK-3
contains tyrosine, threonine, and cysteine residues that can react
with 7. Even with consideration of labeling at these 3 additional
potential modification sites, the results suggest that more than
one reactive amino group is labeled.10
To examine the effect of the lysyl residues, all five lysyl
residues in FT-YLK-3 were replaced by arginine residues.
Peptide FT-YLK-3-R5 displayed less than 5% of the catalytic
activity of FT-YLK-3 with (±)-3. This result is also consistent
with catalytic role for the e-amino groups of the lysine residues
of FT-YLK-3. In addition, 1 mM of lysine, arginine, tyrosine,
proline, and mixtures of these amino acids, or lysyl-lysine did
not catalyze the retro-aldol reactions described above as assayed
in 5% CH3CN-42 mM Na phosphate (pH 7.5).
Since no additional amino functionalities were selected in
FT-YLK-3, improvement of the catalytic activity of the peptide
may originate from structural changes to the peptide and/or the
addition of factors such as acid–base catalysis and transition
state stabilization. The CD specta of FT-YLK-3 and YLK-
18-opt (100 mM) in 45 mM Na phosphate buffer (pH 7.5) at
25 °C are shown in Fig. 1. For FT-YLK-3 the mean residue
ellipticity at 208 and 222 nm were 23.04 3 104 and 22.47 3
104 deg cm2 dmol21, respectively.11 The CD spectra indicate
that the peptide adopts an a-helical structure with an a-helical
content of ~ 70%.12 The template peptide YLK-18-opt showed
a much reduced a-helical content under the same conditions,
~ 20%.13 These results suggest that selection of the 6 amino
acid residue C-terminal extension in FT-YLK-3 stabilizes the a-
helical conformation of the peptide. MALDI mass analysis of
FT-YLK-3 revealed the peptide forms a covalent dimer though
a monomer was the main species detected after storage in the
buffer used for kinetic studies. The dimer of FT-YLK-3 formed
This study was supported in part by the NIH (CA27489) and
The Skaggs Institute for Chemical Biology.
Notes and references
1 (a) K. Johnsson, R. F. Allemann, H. Widmer and S. Benner, Nature,
1993, 365, 530; (b) E. Perez-Paya, R. A. Houghten and S. E. Blondell,
J. Biol. Chem., 1996, 271, 4120; (c) M. Allert, M. Kjellstrand, K. Broo,
A. Nilsson and L. Baltzer, J. Chem. Soc., Perkin Trans. 2, 1998,
2271.
2 K. S. Broo, H. Nilsson, J. Nilsson and L. Baltzer, J. Am. Chem Soc.,
1998, 120, 10 287 and refs. therein.
3 Complex schemes for phage display selection based on catalytic activity
have recently been proposed: H. Pedersen, S. Holder, D. P. Sutherlin, U.
Schwitter, D. S. King and P. G. Schultz, Proc. Natl. Acad. Sci. USA,
1998, 95, 10 523; S. Demartis, A. Huber, F. Viti, L. Lozzi, L.
Giovannoni, P. Neri, G. Winter and D. Neri, J. Mol. Biol., 1999, 286,
617.
4 (a) J. Wagner, R. A. Lerner and C. F. Barbas III, Science, 1995, 270,
1797; (b) G. Zhong, R. A. Lerner and C. F. Barbas III, Angew. Chem.,
Int. Ed., 1999, 38, 3738; (c) F. Tanaka, R. A. Lerner and C. F. Barbas
III, J. Am. Chem. Soc., 2000, 122, 4835.
5 R. D. Kobes and E. E. Dekker, Biochem. Biophys. Res. Commun., 1967,
27, 607; H. Nishihara and E. E. Dekker, J. Biol. Chem., 1972, 247, 5079;
C. J. Vlahos and E. E. Dekker, J. Biol. Chem., 1986, 261, 11 049; R.
Björnestedt, G. Zhong, R. A. Lerner and C. F. Barbas III, J. Am. Chem.
Soc., 1996, 118, 11 720.
6 NNK codon mixture was used to encode 20 amino acids in the library.
The phage display library was selected using pComb3 system: Phage
Display: A Laboratory Manual, ed. C. F. Barbas III, D. R. Burton, J. K.
Scott and G. J. Silverman, Cold Spring Harbor Laboratory Press, Cold
Spring Harbor, New York, 2001.
7 The concentration of the peptide was determined by amino acid
analysis.
8 The purity of the enantiomeric substrates that were used in the reactions
is as follows: (R)-3, 99.5% ee; (S)-3, 98% ee.
9 F. Tanaka, R. A. Lerner and C. F. Barbas III, Chem. Commun., 1999,
1383.
10 The order of the reactivity of the 5 lysyl residues of the peptide was not
determined.
11 The mean residue ellipticity of FT-YLK-3 was calculated as the
monomer.
12 P. Korsgren, P. Ahlberg and L. Baltzer, J. Chem. Soc., Perkin Trans 2,
2000, 643.
Fig. 1 CD spectra of peptides FT-YLK-3 (solid square) and YLK-18-opt
(open circle). The CD spectra were recorded in 45 mM Na phosphate buffer
(pH 7.5) at 25 °C at a peptide concentration of 100 mM.
13 The a-helical content of YLK-18-opt is concentration dependent. See
ref. 1b.
770
Chem. Commun., 2001, 769–770