A modular assembly strategy for improving the substrate specificity of small catalytic peptides

Article abstract of DOI:10.1021/ja0171815

In contrast to large proteins, small peptide catalysts typically display limited specificity for small molecule substrates. This is presumably a result of the limited opportunities small peptides have to fold in a manner that provides for the formation of an isolated reaction vessel that effectively binds and sequesters substrates from bulk solvent while at the same time catalyzing their transformation. For the preparation of small peptide catalysts that possess improved substrate specificity, we have developed a modular assembly strategy that involves appending phage display-derived substrate binding-domain modules to catalytically active peptide domains. We demonstrate the potential of this strategy with the construction of a small 35-amino acid residue aldolase peptide with improved substrate specificity. The advantages of this approach are that it reduces the demand on the functionalization of the catalytic site and it is modular, therefore making its adaptation to a variety of specificities rapid. The modular assembly strategy studied here may present advantages over exhaustive searches of large random-sequence peptide libraries for peptides with singular function. Copyright

Full text of DOI:10.1021/ja0171815

Published on Web 03/15/2002
A Modular Assembly Strategy for Improving the Substrate Specificity of Small
Catalytic Peptides
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
Received September 27, 2001
Substrate specificity is one of the hallmarks of enzymes. Substrate
specificity allows an enzyme to catalyze a reaction involving
substrate molecules found within a complex mixture of compounds
possessing the same functional groups. Generation of designer
protein catalysts that possess substrate specificity has been dem-
onstrated by modification of Nature’s enzymes1 and by the
preparation of catalytic antibodies.2 In contrast to large proteins,
small peptide catalysts have demonstrated limited specificity for
small-molecule substrates.3 This is presumably a result of the limited
opportunities small peptides have to fold in a manner that provides
for the formation of an isolated reaction vessel that effectively binds
and sequesters substrates from bulk solvent while at the same time
catalyzing their transformation. To explore routes for the preparation
of small peptide enzymes that possess improved substrate specific-
ity, we have examined a modular assembly strategy. Here we
demonstrate the potential of this strategy with the construction of
a small 35-amino acid residue aldolase peptide with improved
substrate specificity.
peptides, YPNEFDWWDYYY (FluS303 in their report) binds to
fluorescein. This peptide was selected for binding to fluorescein
derivative 2 from a combinatorial library of 12-mer peptides
displayed on phage.7 We chose this peptide as our substrate-binding
module and attached it to the N-terminus of FT-YLK3. Since FT-
YLK3 has a tyrosine residue at the N-terminus and FluS303 has
three tyrosine residues at the C-terminus, one tyrosine was removed
at the junction. Peptide FluS303-FTYLK3 was chemically synthe-
sized (C-terminal amide) and characterized.
Our design strategy attempts to recruit a substrate-specific module
for providing substrate specificity to an otherwise promiscuous
catalyst. Covalent combination of binding- and catalytic-domain
modules might improve the substrate specificity of the catalyst.
When the binding site is in close proximity to the catalytic site,
the catalytic site would receive the benefit of a higher local substrate
concentration provided by sequestering the substrate in close
proximity to the catalytic site. The potential advantages of this
approach are that it reduces the demand on the functionalization
of the catalytic site, which is limited in small peptides, and it is
modular, therefore making its adaptation to a variety of specificities
rapid.
Recently, we reported a 24-amino acid residue aldolase peptide
FT-YLK3 that catalyzes retro-aldol reactions via an enamine
mechanism.4 This peptide was the product of covalent mechanism-
based â-diketone selection of a phage-displayed peptide library
where six randomized residues were appended to an 18-residue
amphipathic R-helix5 (YKLLKELLAKLKWLLRKL) containing
five lysine residues. The ꢀ-amino groups of the lysyl residues are
essential for the catalytic reaction. Peptide FT-YLK3 is highly
R-helical and catalyzes the retro-aldol reaction of 1 with a kcat/
kuncat of 1900, albeit with a Km of 1.8 mM indicative of relatively
poor substrate recognition.
Retro-aldol substrate 3 contains a structural moiety from
fluorescein and was designed and synthesized as a specific substrate
for FluS303-FTYLK3.8 Peptide-catalyzed retro-aldol reactions of
3 were examined using FluS303-FTYLK3 and FT-YLK3. Reactions
were performed using 50 µM peptide in 10% CH3CN-42.5 mM
Na phosphate (pH 7.5), and the product aldehyde 4 was monitored
by HPLC. Both peptides FluS303-FTYLK3 and FT-YLK3 dis-
played saturation kinetics. Peptide-catalyzed retro-aldol reactions
of 1 were also studied.9 Peptide FluS303-FTYLK3 catalyzed the
retro-aldol reaction of 3 with a Km of 8 µM and a kcat of 2.3 ×
10-4 min-1 (Table 1).10 FluS303-FTYLK3 catalyzed the retro-aldol
reaction of 1 with a 1.1 mM Km.11 The Km value of substrate 3 was
140-fold lower than that of 1 in the FluS303-FTYLK3-catalyzed
reaction and the kcat/Km value of 3 was 43-fold higher than that of
1. Thus FluS303-FTYLK3 successfully discriminates between
substrates 3 and 1. As calculated from the Km values, the differential
binding energy (∆∆G) in favor of substrate 3 over 1 in the FluS303-
FTYLK3-catalyzed reactions is 2.9 kcal/mol.12 The catalytic module
itself, FT-YLK3, catalyzed the reaction of 3 with a Km of 130 µM.
Fortuitously, the Km value of substrate 3 was 14-fold lower than
that of substrate 1 in the FT-YLK3-catalyzed reactions. Thus, for
FT-YLK3-catalyzed reactions, differential binding of substrate 3
over 1 is favored by 1.6 kcal/mol.12 Note that the Km value of
FluS303-FTYLK3 with substrate 3 was 16-fold lower than that of
FT-YLK3 with 3, while the Km values of 1 were similar in both
the FluS303-FTYLK3- and FT-YLK3-catalyzed reactions. These
results indicate that the enhanced substrate specificity of FluS303-
FTYLK3 for 3 is the result of two effects. First, binding of 3 to
To endow this catalyst with improved substrate specificity, we
sought to covalently connect this peptide to another peptide that
binds a substrate molecule with high affinity, anticipating that the
local concentration of substrate proximal to the catalytic peptide
would be increased. Recently, Rozinov and Nolan reported small
peptides that bind to small-molecule fluorophores.6 One of these
* Corresponding authors. E-mail: carlos@scripps.edu and ftanaka@
scripps.edu.
9
3510
J. AM. CHEM. SOC. 2002, 124, 3510-3511
10.1021/ja0171815 CCC: $22.00 © 2002 American Chemical Society

C O M M U N I C A T I O N S
Table 1. Kinetic Parameters of Peptide-Catalyzed Retro-Aldol Reactions
substrate 3a
substrate 1b
kcat (min-1
peptide
Km (µM)
kcat (min-1
)
kcat/Km (M-1 min-1
29
1.5
)
Km (mM)
)
kcat/Km (M-1min-1
)
FluS303-FTYLK3
FT-YLK3
8
130
2.3 × 10-4
1.1
7.4 × 10-4 c
0.67
2.0 × 10-4
1.8d
5.6 × 10-4 d
0.31d
a Reaction conditions: [Peptide] 50 µM in 10% CH3CN-42.5 mM Na phosphate (pH 7.5) at 25 °C. See ref 11. b [Peptide] 50 µM in 5% CH3CN-42.5
mM Na phosphate (pH 7.5) at 25 °C. The first-order kinetic constant of the background reaction (kuncat) was 3.9 × 10-7 min-1
.
c kcat/kuncat 1900. d Data were
taken from ref 4. [Peptide] 100 µM.
(4) circular dichroism studies, (5) multiturnover experiment, (6) catalysis
of the aldol reaction, and (7) effect of peptide concentration on
enaminone formation (PDF). This material is available free of charge
via the Internet at http://pubs.acs.org.
the FT-YLK3 domain as compared to substrate 1 is enhanced by
1.6 kcal/mol. While the exact mechanism for this increase is not
clear at present and was not part of the design strategy, heteroatom
presentation within 3 may play a role (see Supporting Information
for discussion and model). Second, the addition of the FluS303
sequence to FTYLK3 further increases the affinity of the peptide
for substrate 3 (∆∆G ) 1.7 kcal/mol)13 while having a minimal
effect on the binding of substrate 1 (∆∆G ) 0.3 kcal/mol). We
assign this increase in substrate specificity to the effect of the
modular assembly strategy.14
To further probe the role of the substrate-binding domain FluS303
in the peptide enzyme FluS303-FTYLK3, binding to fluorescein
and the effect of fluorescein on the catalyzed reaction was examined.
The dissociation constant (Kd) of fluorescein for FluS303-FTYLK3
was determined to be 3.6 µM by fluorescence-quenching, indicating
that the binding domain retains its original function in the fusion
peptide.15 Further, when the FluS303-FTYLK3-catalyzed reaction
of 3 was performed in the presence of fluorescein (500 µM), the
kinetic parameters were Km 60 µM, kcat 2.5 × 10-4 min-1, and
kcat/Km 4.2 M-1 min-1. Thus, the Km value substantially increased
while the kcat remained unchanged (within experimental deviation).
This result indicates that when fluorescein occupies the substrate-
binding domain, in competition with 3, the specificity constant is
greatly reduced. With respect to the catalytic domain, it is likely
that the highly R-helical structure of FT-YLK3 is maintained in
the fusion peptide since in all of our studies the kcat of the catalytic
domain FT-YLK3 is not altered as compared to that of the fusion
peptide. Our earlier study of FT-YLK3 found a strong correlation
between R-helical content and kcat.4
References
(1) Petrounia, I. P.; Arnold, F. H. Curr. Opin. Biotechnol. 2000, 11, 325.
Cedrone, F.; Menez, A.; Quemeneur, E. Curr. Opin. Struct. Biol. 2000,
10, 405.
(2) Reymond, J.-L. Top. Curr. Chem. 1999, 200, 59. Barbas, C. F., III; Rader,
C.; Segal, D. J.; List, B.; Turner, J. M. AdV. Protein Chem. 2000, 55,
317. Schultz, P. G.; Lerner, R. A. Science 1995, 269, 1835. Tanaka, F.;
Lerner, R. A.; Barbas, C. F., III. J. Am. Chem. Soc. 2000, 122, 4835.
(3) Broo, K. S.; Nilsson, H.; Nilsson, J.; Baltzer, L. J. Am. Chem. Soc. 1998,
120, 10287. Broo, K. S.; Brive, L.; Ahlberg, P.; Baltzer, L. J. Am. Chem.
Soc. 1997, 119, 11362. Johnsson, K.; Allemann, R. K.; Widmer, H.;
Benner, S. A. Nature 1993, 365, 530. Peptide ligases operate via template-
assisted catalysis and demonstrate good substrate specificity: Kennan,
A. K.; Haridas, V.; Severin, K.; Lee, D. H.; Ghadiri, M. R. J. Am. Chem.
Soc. 2001, 123, 1797. Yao, S.; Ghosh, I.; Zutshi, R.; Chmielewski, J.
Nature 1998, 396, 447.
(4) Tanaka, F.; Barbas, C. F., III. Chem. Commun. 2001, 769.
(5) Perez-Paya, E.; Houghton, R. A.; Blondelle, S. E. J. Biol. Chem. 1996,
271, 4120.
(6) Rozinov, M. N.; Nolan, G. P. Chem. Biol. 1998, 5, 713.
(7) The affinity of FluS303 displayed pentavalently on phage for 2 was
reported to be Kd 2.7 nM in ref 6. Lacking in avidity, the affinity of
monovalent peptide FluS303 for 2 is lower.
(8) Aldehyde 4 was synthesized from resorufin, sodium salt, and 4-fluoro-
benzaldehyde according to the procedure reported by Yeager et al. Yeager,
G. Y.; Schissel, D. N. Synthesis 1991, 63. Subsequently, the DL-proline-
catalyzed aldol reaction with acetone according to the procedure reported
by Sakthivel et al. gave aldol 3. Sakthivel, K.; Notz, W.; Bui, T.; Barbas,
C. F., III. J. Am. Chem. Soc. 2001, 123, 5260.
Circular dichroism studies indicate that FluS303-FTYLK3 retains
the R-helical structure of the parental peptide FTYLK3.16 FluS303-
FTYLK3 also formed a UV-observable enaminone upon addition
of 2,4-pentanedione.4 Denaturation of the R-helical structure (4.5
M guanidine hydrochloride) as observed in the CD spectra was
accompanied by complete loss of catalytic activity.
(9) The reaction of 1 was followed by monitoring the increase in fluorescence
(λex 330 nm, λem 452 nm). See ref 4 and List, B.; Barbas, C. F., III; Lerner,
R. A. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 15351.
(10) Since near-equal peptide-substrate concentrations were employed in the
assay, the kcat and Km were determined according to the procedure reported
by Smith et al. Smith, G. D.; Eisenthal, R.; Harrison, R. Anal. Biochem.
1977, 79, 643.
(11) The velocity of the catalyzed reaction was identical in 10 and 5% CH3-
CN. To increase the solubility of 3, 10% CH3CN was used. The first-
order kinetic constant of the background reaction (kuncat) of 3 was
significantly lower than that of 1.
In summary, we have shown that a modular assembly strategy
can be used to create small peptide enzymes with good specificity
for small molecule substrates. Since our approach accepts substrate-
binding domains derived from phage display selections, a wide
variety of substrate binding specificities might be readily selected
and recruited to rapidly create peptide enzymes with a wide variety
of substrate specificities. This is significant since peptide domains
that bind small molecules are not currently accessible by rational
design. We anticipate that this approach will be applicable to other
small peptide catalytic domains as well, allowing for the combi-
natorial preparation of substrate specific peptide enzymes. The
modular assembly strategy studied here may therefore present
advantages over exhaustive searches of large random-sequence
peptide libraries17 for peptides with singular function.
(12) ∆∆G ) -RT ln{(Km of 3)/(Km of 1)} at 25 °C.
(13) ∆∆G ) -RT ln{(Km of FluS303-FTYLK3-catalyzed reaction of 3)/(Km
of FT-YLK3-catalyzed reaction of 3)} at 25 °C.
(14) At this point, two potential mechanisms for providing enhanced substrate
discrimination are: (1) the binding domain provides a higher local substrate
concentration to the catalytic domain with the two domains acting
independently, and (2) the binding and catalytic domains work coopera-
tively to form a more specific active site.
(15) Quenching study was performed using fluorescein (5 µM) and FluS303-
FTYLK3 (1.5-50 µM) in 10% CH3CN-42.5 mM Na phosphate (pH 7.5)
by monitoring the fluorescence of fluorescein (λex 472 nm, λem 518 nm).
See ref 6. The Kd was calculated according to the procedure described in
ref 10.
(16) The CD spectra of FluS303-FTYLK3 (50 µM) in 45 mM Na phosphate
(pH 7.5) at 25 °C indicates that the peptide contains R-helical and â-sheet
structures. Studies using the program k2d indicate the R-helical content
is 27% (max error 22.7%). Since the R-helical content of FT-YLK3 (100
µM) was estimated to be ∼70%,4 the ideal content of FluS303-FTYLK3
is calculated to be 48% (24 residue x 0.7/35 residue). For the k2d program
see, Andrate, M. A.; Chacon, P.; Merelo, J. J.; Moran, F. Protein Eng.
1993, 6, 383.
Acknowledgment. We thank G. Nolan for helpful discussions.
This study was supported in part by the NIH (CA27489) and The
Skaggs Institute for Chemical Biology.
(17) Keefe, A. D.: Szostak, J. W. Nature 2001, 410, 715.
Supporting Information Available: (1) Data for compounds 3 and
4, (2) assays, (3) additional discussion concerning the binding of 3,
JA0171815
9
J. AM. CHEM. SOC. VOL. 124, NO. 14, 2002 3511