A combinatorial approach to minimal peptide models of a metalloprotein active site

Article abstract of DOI:10.1039/b601407k

Screening of a "one-bead-one-compound" peptide library containing biomimetic His/Cys ligands has led to the discovery of sequences that hydrolyze ester substrates in combination with Zn²⁺. The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b601407k

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
A combinatorial approach to minimal peptide models of a
metalloprotein active site{
Frances Namuswe and David P. Goldberg*
Received (in Austin, TX, USA) 31st January 2006, Accepted 18th March 2006
First published as an Advance Article on the web 11th April 2006
DOI: 10.1039/b601407k
Zn2+-induced hydrolysis of phosphate and carboxylic ester
Screening of a ‘‘one-bead-one-compound’’ peptide library
containing biomimetic His/Cys ligands has led to the discovery
of sequences that hydrolyze ester substrates in combination
with Zn2+.
substrates.
An earlier report by Berkessel and coworkers22 described the
synthesis of a library of short peptides (11 AAs) with 625 possible
unique sequences that was screened on-bead in the presence of
Zr4+ for the hydrolytic cleavage of 5-bromo-4-chloro-3-indolyl
phosphate (BCIP), a well known chromogenic probe.3 These
efforts resulted in the identification of three different peptide–Zr4+
constructs capable of hydrolyzing phosphoryl ester substrates. This
type of on-bead screening for catalyst discovery has been used
previously.23,24 In light of these results, we chose to make a library
of the same size (11 AAs in length, 625 members), with a
generalized sequence that is intended to mimic the first coordina-
tion sphere of PDF. The architecture of our bead-bound library is
shown in Fig. 1. TentaGel-NH2 resin was used as solid support
because it allows for deprotection of the sidechains while leaving
the peptide attached to the resin, and because of its good swelling
properties in water. As seen in Fig. 1, each member of the library
contains one Cys and two His metal-binding ligands fixed in
positions (AA2, AA6, AA10). The His residues were incorporated
as part of the sequence HEXXH, which is a highly conserved
motif found in PDF18 as well as in many zinc enzymes, and
invariably forms a chelate ring with the His donors bound to the
metal center. A second conserved motif found in PDF is GCLS,25
which contains the metal-binding Cys ligand, and thus Leu was
included next to the Cys. It has been shown previously that specific
metal binding properties can be incorporated into synthetic
peptides by correctly positioning the ligands in minimally
conserved motifs.26 The diversity of the library was achieved by
varying the four X positions over five amino acids, Ser, Arg, Gly,
Asp, or Ala, resulting in the desired maximum of 625 (54) unique
sequences. These amino acids were selected to impart a wide range
of properties to the library, including hydrophobic, hydrophilic,
ionic and H-bonding characteristics. A phenylalanine at the
C-terminus was included for subsequent UV-vis absorption assays
in the solution phase.
The screening of combinatorial peptide libraries has been used
previously to discover both short peptides as well as peptide–metal
complexes that function as catalysts for various organic trans-
formations.1–5 Similar combinatorial strategies have led to the
development of well defined peptide–metal complexes as biological
probes.6,7 However, there are still relatively few examples of
combinatorial approaches being used to obtain catalysts that have
both naturally occurring ligands (e.g. His, Cys) and biologically
relevant metal ions (e.g. Zn2+).8 We envision using this approach
to discover biomimetic catalysts that can reproduce key structural
and functional aspects of metalloproteins.9 Herein we demonstrate
the use of this methodology to select for sequences that contain
only proteinogenic metal-binding ligands and catalyze hydrolysis
in the presence of Zn2+ both on solid support and in solution.
Previous efforts in our lab have been aimed at preparing
synthetic analogues of the metal center of peptide deformylase
(PDF).10–12 The bacterial PDF enzyme is an unusual metallohy-
drolase that contains a mononuclear, four-coordinate Fe2+
cofactor bound to the protein by one Cys and two His residues,
and a catalytically active water molecule occupying the fourth
coordination site.13,14 Although the Zn2+ form of bacterial PDF is
nearly structurally identical to the iron(II), cobalt(II), and nickel(II)
forms, it is dramatically less active.15–17 This result is surprising
given that PDF carries a conserved sequence motif (HEXXH)
found in zinc enzymes, and has an active site structure and
hydrolytic function that is normally seen in the mononuclear zinc
enzyme family.18 Interestingly, a recent exception was found in
PDF isolated from plants, which is apparently highly active with
Zn2+ 19,20
also been reported to use Zn2+ 21
.
A single bacterial PDF from Leptospira interrogans has
.
We are interested in determining the fundamental role of the
metal ion in the mechanism of PDF by constructing and studying
small-molecule models of the metal active site. In this paper we
have taken a combinatorial approach to this problem by preparing
a library of short peptides (11 AAs) that are designed to bind metal
ions (e.g. Zn2+, Co2+, Fe2+) through a His2Cys motif. We have
successfully screened this library to identify sequences that exhibit
The peptide library was manually synthesized by standard
Fmoc procedures and split/pool methodology. These methods
result in a ‘‘one-bead-one-compound’’ library,3 where each
bead contains an undecapeptide of only one sequence, with
Department of Chemistry, Johns Hopkins University, 3400 N. Charles
Street, Baltimore, MD 21218, USA. E-mail: dpg@jhu.edu;
Fax: +1 410.516.8420; Tel: +1 410.516.6658
{ Electronic supplementary information (ESI) available: Peptide syntheses
and reactivity studies. See DOI: 10.1039/b601407k
Fig. 1 Combinatorial library of 625 (54) peptides synthesized by the
split-and-pool method with the X positions varied over the five AAs given
in parentheses.
2326 | Chem. Commun., 2006, 2326–2328
This journal is ß The Royal Society of Chemistry 2006

View Article Online
Fig. 4 All solutions contain HEPES (20 mM) at pH 6. From left to
right: (1) Sequence A + Zn(OAc)2 + BCIP (molar ratio 1 : 1 : 1), (2) A +
BCIP, (3) Zn(OAc)2 + BCIP and (4) BCIP alone. Photograph taken
after 72 h.
Fig. 2 Microscopic image of a portion of the TentaGel-peptide library
incubated with Zn(OAc)2 and BCIP at pH 6, showing a single active bead
(dark blue) among inactive beads (colorless).
formation. BCIP was then added and the solution exposed to air
to allow for the colored dye to develop. As seen in Fig. 4, the
solution containing the active peptide A + Zn2+ ions developed a
turquoise blue color over 48 h, while only a faint blue color due to
background hydrolysis is observed in the control experiments
where peptide A alone, Zn(OAc)2 alone, or buffer are mixed with
BCIP. Similar results were obtained with the active peptide B. In
contrast, no turquoise color developed over 48 h in the presence of
inactive peptide C and Zn2+ ions. These results clearly show that it
is only the A–Zn2+ or B–Zn2+ complexes that lead to hydrolysis of
the substrate.
We then tested A–Zn2+ and B–Zn2+ to determine if they were
capable of hydrolyzing the carboxylic ester p-nitrophenylacetate
(4-NA), which would further demonstrate their hydrolytic activity
in solution, and allow for comparative kinetic measurements.
Hydrolysis of 4-NA in the presence of sequences A or B and Zn2+
was examined by methods previously established in our labora-
tory.11 The reaction was monitored by observing the appearance
of the nitrophenolate product (4-NP) over time in a UV-vis
spectrophotometer. The initial rates for A–Zn2+ and B–Zn2+ are
compared with the background rate of hydrolysis (buffer alone) in
Fig. 5. There is a significant rate enhancement for both active
sequences, with A giving an approximately ten-fold increase in
rate, while B proceeds approximately 2 times faster than the
background reaction. Both A and B appear to function as catalysts
for the cleavage of 4-NA in the presence of Zn2+. Control
experiments with Zn(OAc)2 revealed almost no rate enhancement
under the same conditions. However, the metal-free peptides A
and B were found to be hydrolytically active toward 4-NA, and
exhibit 10 and 100 times faster initial rates than their respective Zn
complexes. These results are not surprising because of the well
80–100 pmol of peptide per bead. To screen for hydrolytic activity
in the presence of Zn2+, a portion of the library (5 mg,
y2500 beads) was incubated with a 1 mM solution of
Zn(OAc)2 in HEPES buffer at pH 6 for 15 min. The substrate
BCIP was then added and the library was visually inspected every
6–12 h under a standard stereoscopic microscope. Hydrolysis of
the phosphoester group in BCIP leads to rapid oxidative
dimerization to give the highly colored, insoluble blue dye shown
in Fig. 2. Thus, any beads that actively hydrolyze BCIP are
expected to become stained with a dark blue color. Approximately
1–3 dark blue beads had developed after 48 h, and microscopic
identification of one of the active beads is shown in Fig. 2. This
bead was isolated by microsyringe and its peptide sequence
determined by automated Edman degradation. The sequences of
two different dark blue beads isolated from separate screening
experiments are shown in Fig. 3, along with an inactive sequence
obtained from a non-colored bead. Clean single sequences were
obtained from each bead, indicating that the coupling steps in the
library synthesis were complete. It is interesting to note that the
two active sequences are rich in Arg residues in the variable
positions. One obvious consequence of Arg selection is that both A
and B are positively charged, which may help to attract the
negatively charged BCIP. In contrast, the inactive sequence C has
an overall negative charge.
The hydrolytically active sequences A and B, and the inactive
sequence C, were then resynthesized on a larger scale for solution
phase studies. For this purpose, these peptides were synthesized on
Wang resin, which allows for cleavage from the solid support.
After synthesis and purification by HPLC, these peptides were
tested for their ability to hydrolyze BCIP with Zn2+ in solution. A
solution of the peptide (1 mM) was first incubated with Zn(OAc)2
(1 mM) in HEPES buffer at pH 6 for 1 h. Incubation with Zn2+
was performed under anaerobic conditions to avoid disulfide
Fig. 3 Sequences A and B were determined from microsequencing of
two active beads selected from separate BCIP assays. Sequence C was
determined from an inactive (colorless) bead selected at random.
Fig. 5 Initial rates for the hydrolysis of 4-NA. All reactions done in
HEPES at pH 7.5 with 1 mM 4-NA added to initiate the reaction. A +
Zn(OAc)2 (.); B + Zn(OAc)2 ($); background (&).
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 2326–2328 | 2327

View Article Online
Although the explicit structure of the catalytically active peptide–
metal complex has not been determined, these peptides bind Co2+
,
and by analogy Zn2+, through natural metal-binding ligands
specifically incorporated to mimic the first coordination sphere of
the metallohydrolase PDF. Combinatorial peptide libraries may be
useful for obtaining models of other types of metalloenzymes,
provided a proper reactivity screen can be developed.
We thank Dr. Sarah L. J. Michel at the University of Maryland,
Pharmaceutical Sciences and her research group for assistance with
peptide synthesis and HPLC. This work was supported by the
NIH (GM62309).
Fig. 6 UV-vis spectrum of B + CoCl2 in HEPES (20 mM) at pH 7.5.
Inset: expanded region showing the d–d transitions.
Notes and references
known catalytic activity of His residues toward the hydrolysis of
carboxylic esters, and similar activity has been seen for apo zinc
finger peptides vs the zinc-loaded peptide.27,28 The lower activity in
the presence of Zn2+ suggests that the Zn2+ ion is in fact
coordinated to the His residues of A and B, blocking them from
direct participation in the hydrolysis of 4-NA.
1 A. Berkessel, Curr. Opin. Chem. Biol., 2003, 7, 409–419.
2 K. W. Kuntz, M. L. Snapper and A. H. Hoveyda, Curr. Opin. Chem.
Biol., 1999, 3, 313–319.
3 K. S. Lam, M. Lebl and V. Krchna´k, Chem. Rev., 1997, 97, 411–448.
4 M. B. Francis, T. F. Jamison and E. N. Jacobsen, Curr. Opin. Chem.
Biol., 1998, 2, 422–428.
The library design and screening techniques presented here have
clearly resulted in the discovery of short peptides with proteino-
genic metal-binding ligands that, in combination with the most
common biological metal ion used for hydrolysis, Zn2+, function
as hydrolytic catalysts both on solid support and in solution. To
elucidate the structure of the catalytic species, Co2+ was added in
place of Zn2+ as a spectroscopic probe. This type of substitution
has been used extensively to obtain structural information on zinc
metalloproteins, and relies on the fact that Zn2+ and Co2+ typically
exhibit similar coordination geometries for a given ligand set.
Addition of CoCl2 to peptide A or B resulted in the appearance of
an intense blue color, indicative of the formation of a four- or five-
coordinate Co2+ complex. The UV-vis spectrum of B–Co2+ is
shown in Fig. 6, and reveals peaks characteristic of Co2+ d–d
transitions at 570, 625, and 660(sh) nm. This spectrum matches
quite well with that obtained for tetrahedral zinc proteins,
including zinc finger proteins with four-coordinate, CCHH
ligation at the metal center.29 In addition, there are strong
absorbances at 283, 317(sh) and 360(sh) nm which do not appear
in the spectrum of the free peptide or that of CoCl2 alone. These
latter peaks can be assigned to Cys-to-Co2+ LMCT bands. Similar
peaks are observed for Cys-to-Co2+ LMCT bands in cobalt-
substituted zinc fingers.29
5 S. J. Miller, Acc. Chem. Res., 2004, 37, 601–610.
6 L. J. Martin, B. R. Sculimbrene, M. Nitz and B. Imperiali, QSAR
Comb. Sci., 2005, 24, 1149–1157.
7 M. Nitz, K. J. Franz, R. L. Maglathlin and B. Imperiali,
ChemBioChem, 2003, 4, 272–276.
8 A. Berkessel and R. Riedl, J. Comb. Chem., 2000, 2, 215–219.
9 For an earlier report of this approach, see: N. Shibata, J. E. Baldwin
and M. E. Wood, Bioorg. Med. Chem. Lett., 1997, 7, 413–416.
10 S. Chang, V. V. Karambelkar, R. D. Sommer, A. L. Rheingold and
D. P. Goldberg, Inorg. Chem., 2002, 41, 239–248.
11 R. C. diTargiani, S. Chang, M. H. Salter, R. D. Hancock, Jr. and
D. P. Goldberg, Inorg. Chem., 2003, 42, 5825–5836.
12 D. P. Goldberg, R. C. diTargiani, F. Namuswe, E. C. Minnihan,
S. Chang, L. N. Zakharov and A. L. Rheingold, Inorg. Chem., 2005, 44,
7559–7569.
13 A. Becker, I. Schlichting, W. Kabsch, D. Groche, S. Schultz and
A. F. V. Wagner, Nat. Struct. Biol., 1998, 5, 1053–1058.
14 R. Jain, B. Hao, R. Liu and M. K. Chan, J. Am. Chem. Soc., 2005, 127,
4558–4559.
15 S. Ragusa, S. Blanquet and T. Meinnel, J. Mol. Biol., 1998, 280,
515–523.
16 P. T. R. Rajagopalan, S. Grimme and D. Pei, Biochemistry, 2000, 39,
779–790.
17 P. T. R. Rajagopalan, X. C. Yu and D. Pei, J. Am. Chem. Soc., 1997,
119, 12418–12419.
18 S. Ragusa, P. Mouchet, C. Lazennec, V. Dive and T. Meinnel, J. Mol.
Biol., 1999, 289, 1445–1457.
19 A. Serero, C. Giglione and T. Meinnel, J. Mol. Biol., 2001, 314,
695–708.
The Co2+ data provide good spectroscopic evidence that both
peptides A and B coordinate to Co2+ through the designed
His2Cys motif. However, we cannot rule out other coordination
modes based on these data alone. Titration experiments with Co2+
were attempted in order to shed further light on the stoichiometry
and strength of metal binding, but interpretation of these data
were hampered by the fact that the Co2+ complex slowly decays
over time to unidentified cobalt species. Although these studies
were done anaerobically, the instability of the peptide–Co2+
complexes is likely due to oxidative degradation from trace
amounts of dioxygen. We have previously observed extreme air-
sensitivity in (N2S(thiolate)CoII model complexes.10
20 S. Fieulaine, C. Juillan-Binard, A. Serero, F. Dardel, C. Giglione,
T. Meinnel and J.-L. Ferrer, J. Biol. Chem., 2005, 280, 42315–42324.
21 Y. K. Li, Z. F. Chen and W. M. Gong, Biochem. Biophys. Res.
Commun., 2002, 295, 884–889.
22 A. Berkessel and D. A. He´rault, Angew. Chem., Int. Ed., 1999, 38,
102–105.
23 G. T. Copeland and S. J. Miller, J. Am. Chem. Soc., 2001, 123,
6496–6502.
24 K. H. Shaughnessy, P. Kim and J. F. Hartwig, J. Am. Chem. Soc., 1999,
121, 2123–2132.
25 T. Meinnel, C. Lazennec, S. Villoing and S. Blanquet, J. Mol. Biol.,
1997, 267, 749–761.
26 G. Xing and V. J. DeRose, Curr. Opin. Chem. Biol., 2001, 5, 196–200.
27 A. Nomura and Y. Sugiura, Inorg. Chem., 2004, 43, 1708–1713.
28 M. Dhanasekaran, S. Negi and Y. Sugiura, Acc. Chem. Res., 2006, 39,
45–52.
In summary, on-bead screening of a combinatorial peptide
library has led to the discovery of short peptides that utilize Zn2+
to hydrolyze ester substrates both on solid support and in solution.
29 S. F. Michael, V. J. Kilfoil, M. H. Schmidt, B. T. Amann and J. M. Berg,
Proc. Natl. Acad. Sci. U. S. A., 1992, 89, 4796–4800.
2328 | Chem. Commun., 2006, 2326–2328
This journal is ß The Royal Society of Chemistry 2006