Revisiting and re-engineering the classical zinc finger peptide: Consensus peptide-1 (CP-1)

Article abstract of DOI:10.1039/c5mb00796h

Zinc plays key structural and catalytic roles in biology. Structural zinc sites are often referred to as zinc finger (ZF) sites, and the classical ZF contains a Cys₂His₂ motif that is involved in coordinating Zn(II). An optimized Cys

Full text of DOI:10.1039/c5mb00796h

Molecular
BioSystems
View Article Online
View Journal
PAPER
Revisiting and re-engineering the classical zinc
finger peptide: consensus peptide-1 (CP-1)
Cite this:DOI: 10.1039/c5mb00796h
a
b
b
a
Angelique N. Besold, Leland R. Widger, Frances Namuswe, Jamie L. Michalek,
a
b
Sarah L. J. Michel* and David P. Goldberg*
Zinc plays key structural and catalytic roles in biology. Structural zinc sites are often referred to as zinc
finger (ZF) sites, and the classical ZF contains a Cys His motif that is involved in coordinating Zn(II).
An optimized Cys His ZF, named consensus peptide 1 (CP-1), was identified more than 20 years ago
2
2
2
2
using a limited set of sequenced proteins. We have reexamined the CP-1 sequence, using our current,
much larger database of sequenced proteins that have been identified from high-throughput sequencing
methods, and found the sequence to be largely unchanged. The CCHH ligand set of CP-1 was then
2
altered to a CAHH motif to impart hydrolytic activity. This ligand set mimics the His Cys ligand set of
peptide deformylase (PDF), a hydrolytically active M(II)-centered (M = Zn or Fe) protein. The resultant
peptide [CP-1(CAHH)] was evaluated for its ability to coordinate Zn(II) and Co(II) ions, adopt secondary
structure, and promote hydrolysis. CP-1(CAHH) was found to coordinate Co(II) and Zn(II) and a
pentacoordinate geometry for Co(II)–CP-1(CAHH) was implicated from UV-vis data. This suggests a
Received 19th November 2015,
Accepted 20th February 2016
His Cys(H O)
2 2 2
environment at the metal center. The Zn(II)-bound CP-1(CAHH) was shown to adopt
1
partial secondary structure by 1-D H NMR spectroscopy. Both Zn(II)–CP-1(CAHH) and Co(II)–CP-1(CAHH)
show good hydrolytic activity toward the test substrate 4-nitrophenyl acetate, exhibiting faster rates than
most active synthetic Zn(II) complexes.
DOI: 10.1039/c5mb00796h
www.rsc.org/molecularbiosystems
that coordinate zinc and the structure that the peptide adopts
Introduction
1
1,13–21
upon zinc(II) coordination.
In all ZF proteins, the zinc binding domains are well defined
It is estimated that one-third of all proteins are bound to metal
1
–4
11,12,22,24–26
ions. One key metal ion for proteins is zinc, which can play and make up smaller domains of the larger protein.
5
–10
either a structural or a catalytic role.
The group of proteins As a result, these ZF domains can be isolated independently of
1
1,12,26
for which zinc plays a structural role are collectively called zinc the rest of the protein, and still bind zinc ions and fold.
1
1,12
finger (ZF) proteins.
e.g. 3–10% of the human genome encodes for ZF proteins) and efforts to mimic ZFs by preparing short peptides that match the
are identified in genomes based upon the presence of cysteine sequence of a specific zinc(II) binding domain of a particular
These proteins are highly abundant This modular nature of ZF domains has led to many successful
(
1
1,12
19,27–31
and histidine rich domains.
ZF proteins utilize these protein.
In addition, shortly after the identification of
3
2
cysteine and/or histidine ligands to coordinate zinc(II) into a the classical ZF by Berg and co-workers in 1987, an optimal or
tetrahedral geometry and fold into a three-dimensional struc- consensus sequence was defined for this class of ZF proteins by
ture. Once folded, the ZF can function – i.e. bind to DNA, aligning the sequences of the few known classical ZFs and
1
1,13–21
RNA or another protein.
the ‘classical’ ZFs. These ZFs bind zinc(II) using a Cys
The best-studied class of ZFs are identifying a sequence based upon the most prevalent amino
His acid within each position of the sequence alignment. This
CCHH) motif and adopt a bba fold upon zinc coordination sequence, named consensus peptide-1 (or CP-1), was deter-
2
2
3
3
(
(
6
,11–13,22,23
Fig. 1).
Since the discovery of classical ZFs in the late mined by manually aligning all of the sequences of classical
1
980s, at least fourteen additional classes of (non-classical) ZFs zinc finger proteins and identifying which residue occurs most
1
1,12
have been identified.
Each class is delineated by the ligands frequently at each position. CP-1 was subsequently prepared
and characterized and shown to exhibit maximal zinc(II) binding
3
3–36
affinity and optimal protein folding.
a
Department of Pharmaceutical Sciences, School of Pharmacy, University of
Maryland, Baltimore, MD 21201, USA. E-mail: smichel@rx.umaryland.edu;
Tel: +1 410-706-7038
CP-1 was defined in 1991, before any major eukaryotic genomes
(e.g. Homo sapiens) had been published and before modern
b
methods of sequence analysis had become readily available,
yet continues to be the sequence utilized as the model classical
Department of Chemistry, Johns Hopkins University, Baltimore, MD 21218, USA.
E-mail: dpg@jhu.edu; Tel: +1 410-516-6658
This journal is ©The Royal Society of Chemistry 2016
Mol. BioSyst.

View Article Online
Paper
Molecular BioSystems
Fig. 1 Consensus peptide-1 (CP-1) constructs (a) ExPASy output of all known CCHH zinc sites. (b) The alignment of the amino acid sequences of CP-1
from 1991, CP-1 from 2015, and modified CP-1 [named CP-1(CAHH)]. (c) A cartoon figure showing the positions of the amino acids of CP-1(CCHH)-2015.
The zinc coordinating residues are colored pink and the highly conserved amino acids are colored teal. (d) The NMR structure of ZFP268, which is an
example of a classical CCHH ZF protein. The zinc coordinating residues are colored pink and the highly conserved amino acids are colored teal. This
figure was made in PyMol (v. 1.1), PDB ID: 2EMW.
ZF peptide. Here, we sought to determine whether this consensus coordinate to this peptide, it was not catalytically active toward
sequence remained the same in 2015, given the proliferation of a chromogenic substrate, 4-nitrophenyl acetate (4-NA), suggest-
sequenced genomes since 1991. Remarkably, we found that there ing that simple modification of a ZF was not enough to impart
4
2
are minimal changes to the CP-1 sequence, despite the explosion activity. However, more recently, Sugiura and co-workers took
of available protein sequence data. a different approach to convert a structural ZF peptide into a
We then sought to determine if CP-1 could be modified to hydrolytically active peptide. Their approach involved mutating
promote hydrolysis. There has long been an interest in utilizing specific amino acids within a singular ZF domain of Sp1, rather
ZF peptides in protein design because their small size, ease of than truncating the C-terminus. Several of these modified Sp1
synthesis, modular nature and metal-binding properties make domains exhibited hydrolytic activity toward carboxylic ester
2
5,37
them attractive building blocks.
The conversion of a struc- and circular DNA substrates when zinc was bound, providing
tural ZF site to a catalytic Zn site is one area of protein design of compelling evidence that a structural ZF domain can be con-
3
8,43–50
particular interest for ZFs, because zinc can have a catalytic role in verted to a catalytic domain.
biology. Proteins that utilize zinc sites for a catalytic role typically The intriguing catalytic activity reported by Sugiura and
coordinate zinc(II) using three amino acid ligands, usually histidine, co-workers for the re-engineered Sp1 zinc finger peptide moti-
aspartate, glutamate and/or cysteine. Importantly, these catalytic vated us to re-investigate CP-1 for possible conversion to a
zinc centers usually contain one or more open coordination sites hydrolytically active peptide, since such a result would demon-
that are absent in structural ZFs, and which allow for binding of strate that the most basic and general consensus sequence for a
substrates and small molecule (e.g. H
2
O) co-factors to promote structural ZF protein domain is capable of sustaining catalytic
3
7–41
reactivity.
Thus, one simple design approach to convert a activity. Instead of simply deleting a His ligand as done pre-
4
2
ZF domain into a catalytic zinc site involves the modification of viously, we sought to impart hydrolytic activity to CP-1 by
one of the four zinc coordinating ligands to allow for an open modifying one of the zinc coordinating ligands. We chose
coordination site for substrate binding. This approach was to modify the second cysteine of CP-1 to an alanine (named
previously attempted for CP-1, by Berg and co-workers, who CP-1(CAHH)), as this would result in a CP-1 variant that contains
truncated the last four amino acids of CP-1. This truncation a ligand set that matches that of a known, hydrolytically active
4
2
4
1,51
deleted one of the metal coordinating histidine ligands, which was enzyme peptide deformylase (CHH).
We demonstrate that
predicted to make the peptide hydrolytically active by providing an this mutant, CP-1(CAHH), coordinates Zn(II) ion with high
open site for water coordination. Although zinc(II) was shown to affinity, and also coordinates cobalt(II) ion. The binding of
Mol. BioSyst.
This journal is ©The Royal Society of Chemistry 2016

View Article Online
Molecular BioSystems
Paper
these divalent metal ions is found to induce partial folding of consensus sequence, CP-1(CCHH)-1991, shows a high degree of
the peptide. It is shown that M(II)-CP-1(CAHH) (M = Zn, Co) sequence conservation (Fig. 1b). Despite the small number of
exhibits good hydrolytic activity toward 4-NA, and mechanistic ZF sequences (less than 1% of those known today) analyzed by
hypotheses for the observed hydrolytic reactivity are evaluated. Berg and co-workers over twenty years ago, their original CP-1
sequence remains an accurate representation of a classical ZF
domain. Thus the use of CP-1 was validated as a scaffold for the
work described herein.
Results and discussion
Revisiting the consensus peptide
Modifying CP-1 to promote hydrolytic activity
The ‘classical’ ZF proteins, which are the best studied, contain Previous work from Berg attempted to make a mutation at the
domains with two cysteine and two histidine ligands. The overall metal binding site of CP-1 to convert the structural zinc(II) site
4
2
form is (Tyr,Phe)-X-Cys-X2,4-Cys-X -Phe-X -Leu-X -His-X3,4-His, into a site capable of mediating hydrolysis. CP-1 was truncated
3
5
2
3
3
with X equaling any amino acid. These proteins were first such that the last four amino acids were removed, including one
identified in the late-1980s, before any comprehensive genome of the histidine ligands, with the goal of opening up a site on
5
2–54
sequencing of whole organisms had been achieved.
Berg and co-workers aligned all of the sequences of known ZFs, activation of H
31 sequences from 18 proteins, and identified a consensus the pK of the bound H
peptide sequence (CP-1) with the sequence ProTyrLysCysProGlu their function and provides access to a nucleophilic Zn-OH
In 1991, the metal center for the binding and activation of H
O at Zn(II) in hydrolytic zinc enzymes, in which
O molecule is lowered, is a key part of
2
O. The
2
1
a
2
3
9,40
CysGlyLysSerPheSerGlnLysSerAspLeuValLysHisGlnArgThrHisThrGly species.
No hydrolytic activity was observed in this earlier
(
Fig. 1a–d). This sequence was shown to bind cobalt(II) and study, despite the successful formation of the CCH (NS ) zinc
2
zinc(II) more tightly (lower K ) than any native ‘classical’ ZF. center and the generation of an open coordination site at the
d
3
3
Thus, CP-1 is considered the optimized ‘classical’ ZF. CP-1 zinc(II) ion in the CP-1(CCH) mutant.
has subsequently been utilized for a variety of studies aimed at
Some of us have previously constructed hydrolytically active,
understanding topics ranging from the thermodynamics and synthetic (small-molecule) analogs of the active site of peptide
5
9–61
kinetics of ZF protein folding, to the design of new ZF-based deformylase (PDF).
complexes.
This enzyme catalyzes the hydrolytic
3
1,33–36,55–57
A recent example is seen in the work of deformylation of the N-terminus of newly synthesized poly-
S ´e n `e que and Latour, where CP-1 and a series of mutants were peptides. PDF is activated by the coordination of a divalent metal
utilized to systematically measure binding affinities for Zn(II) ion [Zn(II), Fe(II), or Co(II)] to one cysteine and two histidine
4
1,51,62
and Co(II) ions by using competitive chelators, and to deter- residues (N S donor set) in the active site.
The activation of
2
3
6
mine the kinetics of metal ion exchange and speciation. Thus, H O occurs at the metal center, ultimately leading to the hydro-
2
the original CP-1, determined from a small subset of sequenced lysis of formylated substrates. We speculated that mutation of
proteins, is a canonical ZF sequence and remains relevant to one of the Cys ligands in CP-1 to a non-coordinating residue
II
current work on ZF biochemistry.
would convert the metal site to an N
2
S–M center and provide a
An additional 13 465 CCHH ZF sequences have been identi- mimic of the PDF active site that would have a higher Lewis
fied since the initial report of CP-1, as a result of the proliferation acidity than the early CP-1(CCH) mutant. The Lewis acidities
of sequenced genomes from high-throughput proteomics efforts. will directly influence the extent of water activation and the
The original CP-1 sequence was investigated here to deter- generation of the hydrolytically active zinc(II)-hydroxide species
mine if the peptide that was designed in 1991 from only 131 for CP-1(CCH) and CP-1(CAHH) according to eqn (1) and (2),
sequences is still the most conserved sequence. Using the respectively.
ExPasy program, all of the CCHH type ZF domains that have
been sequenced and deposited in protein databases were
extracted and aligned, and a Prosite Sequence Logo was then
generated (Fig. 1a). A Sequence Logo is a visualization of the
sequence conservation between homologous proteins. The out-
put of a sequence logo shows a protein sequence with each
amino acid represented at a specific height, which corresponds
to units of ‘bits’ on the y-axis. The height of the amino acid
indicates the frequency with which it occurs in that specific
position throughout all of the sequences analyzed. When the
amino acid is conserved at the 100% level, the height of the
amino acid will be at its maximal level, which is 4 bits. When
the amino acid is not conserved 100% of the time, several
(
1)
(2)
CP-1(CAHH) and coordination of Co(II)
amino acids will be placed in a given column in the order CP-1 was modified to mimic the CHH metal binding motif
of their percent conservation, with the most conserved amino of the active site of PDF. The native protein, referred to as
5
8
acid appearing at the top. Remarkably, a comparison of the CP-1(CCHH), was altered such that the second coordinating
new consensus sequence, CP-1(CCHH)-2015, with the original cysteine residue was mutated to a non-coordinating alanine
This journal is ©The Royal Society of Chemistry 2016
Mol. BioSyst.

View Article Online
Paper
Molecular BioSystems
residue [CP-1(CAHH)] (Fig. 1b). Initial attempts to synthesize titrations in further experiments, contains a tryptophan residue
this peptide independently by using a Symphony Quartet peptide in place of a glutamine as well as two additional amino acids at
synthesizer proved problematic. A significant portion of the the C terminus (Fig. 2a). The Co(II)–CP-1(CCHH)QDW UV-visible
3
3
peptide was not synthesized to completion, and separation of spectrum is identical to that of Co(II)–CP-1(CCHH)-1991. The
the incompletely synthesized peptide from the fully intact pep- d–d bands for Co(II)–CP-1(CCHH)QDW and Co(II)–CP-1(CCHH)-
tide via HPLC proved prohibitive due to low yields. The incom- 1991 are red shifted with maxima at 570 nm and 645 nm
plete synthesis observed may be due to incomplete coupling of compared to CP-1(CAHH) (Fig. 2b and c). The band at 320 nm
each amino acid or aggregation of the shorter peptides during remains the same. The position of the d–d transitions in relation
synthesis. The peptide was then purchased from Bio-synthesis to the coordination environment of the cobalt(II) ion has been
(Lewisville, TX) in the crude form and purified further in our well studied. It is known that cobalt(II) in a 4-coordinate,
laboratory by HPLC. This peptide was not soluble in the presence tetrahedral coordination geometry exhibits absorption maxima
of metal ions. Pure, stable peptide was obtained from Biomatik at 625 ꢀ 50 nm, while cobalt(II) in a 6-coordinate, octahedral
(
Wilmington, DE). The purity of the peptide was independently geometry gives maxima at 525 ꢀ 50 nm. Cobalt(II) in a penta-
63
measured in our laboratory via analytical HPLC and found to be coordinate geometry has maxima between these values. While
95% pure. This peptide was utilized without further purifica- both the CCHH and CAHH CP-1 constructs exhibit absorbance
4
tion. We therefore note that the preparation of mutant CP-1 is not maxima that best fit for tetrahedral coordination at cobalt(II),
straightforward. Interestingly, S ´e n `e que and Latour report similar the absorbance bands for CP-1(CAHH) are blue shifted when
36
difficulties in the preparation of CP-1 in their work.
compared to CP-1(CCHH), suggesting that the geometry may be
Initial experiments focused on determining the influence of distorted from an ideal tetrahedral environment to a 5-coordinate
the Cys-to-Ala mutation on the metal-binding properties of the geometry. In addition, the extinction coefficient (e) at the absorp-
peptide. The ability of CP-1(CAHH) to coordinate Zn(II) was tion maxima provides information about coordination number.
determined indirectly by using Co(II) as a spectroscopic probe Typically, tetrahedral complexes have an e greater than
ꢁ1
ꢁ1
for Zn(II). This approach takes advantage of the rich spectro- 300 M cm , while octahedral complexes have an e below
ꢁ1
ꢁ1
scopic properties of Co(II) in a tetrahedral coordination environ- 30 M cm , with the e values for pentacoordinate complexes
7
ment with sulfur and nitrogen ligands. The Co(II) (d ) ion again falling somewhere in between 4- and 6-coordinate com-
ꢁ
1
ꢁ1 43,63
exhibits distinct d–d transitions between 550–750 nm that can plexes (50 r e r 250 M cm ).
At its maxima, CP-1(CCHH)
ꢁ1
ꢁ1
be correlated with 4-coordinate, 5-coordinate, and 6-coordinate has an extinction coefficient of 510 M cm , suggestive of
ligand environments around the metal. In addition, when sulfur, tetrahedral coordination. CP-1(CAHH) has a significantly lower
and to a lesser extent nitrogen, serve as ligands, ligand-to-metal extinction coefficient of 60 M cm , which is more consistent
charge transfer (LMCT) bands are often observed in the near-UV with a pentacoordinate (or even an octahedral) environment.
ꢁ
1
ꢁ1
3
2,63,64
region.
Addition of CoCl
Thus, in addition to the cysteine and two histidine residues,
to CP-1(CAHH) at pH 7.5 leads to a Co(II) [and, by inference, Zn(II)], is likely coordinated by one or
2
spectrum that exhibits bands with maxima at 320, 550, and two water molecules.
30 nm (Fig. 2b). For comparison, we measured the UV-visible There is precedence for water coordination to zinc ions in
spectrum of Co(II) bound to a modified version of CP-1(CCHH) re-engineered structural zinc sites. For example, when the second
Fig. 2a). This modified version, CP-1(CCHH)QDW, which was coordinating cysteine from a de novo four-helix bundle peptide
6
(
created to investigate metal ion binding using fluorescence called Za that contained four pre-organized zinc binding ligands
4
Fig. 2 (a) A sequence alignment of CP-1(CCHH) with CP-1(CAHH). Differences between the peptides shown in purple (b) UV-visible spectra of
Co(II)–CP-1(CCHH) QDW. (c) UV-visible spectra of Co(II)–CP-1(CAHH).
Mol. BioSyst.
This journal is ©The Royal Society of Chemistry 2016

View Article Online
Molecular BioSystems
Paper
(Cys
2 2
His ) was mutated to an alanine, the Co(II) UV-visible on the affinity for metal ions (Fig. 3a). In this approach, the
spectrum was altered: the lmax for the cobalt(II) d–d bands was apo-peptide was titrated with cobalt(II) dichloride at pH 7.5 in
blue-shifted to 592 nm from 617 nm observed for unmodified Za
and the extinction coefficient was decreased to e = 213 M cm
100 mM HEPES (50 mM NaCl) buffer until saturation was
reached, as monitored by the appearance of the d–d transitions
4
ꢁ1
ꢁ
1
ꢁ
1
ꢁ1 65
from e = 390 M
-coordinate geometry at the metal site, with two exogenous [Co(II)], Fig. 3b) was easily fit to a 1 :1 binding equilibrium,
water molecules serving as ligands, in addition to the cysteine consistent with the formation of a 1: 1 cobalt/peptide complex.
cm
.
These data are suggestive of in the visible region. The resulting titration curve (Abs630nm versus
5
6
5
and histidine residues from Za ligand. Modification of the The best fit of the data led to the determination of an upper limit
first coordinating cysteine of Za to an alanine also altered the for the dissociation constant (K ) of 170 mM for cobalt(II) ion
absorption spectrum, with a measured extinction coefficient binding to CP-1(CAHH). In comparison, a K of 50 nM has been
for the Co(II) d–d bands of e = 60 M cm (the lmax was reported for CP-1(CCHH) when measured in the same manner
4
4
d
d
ꢁ
1
ꢁ1
ꢁ
15
not reported). Similarly, Nomura and Sugiura found that when while a Kapparent of 10
M is reported when a more rigorous,
3
6
they systematically mutated metal coordinating cysteine and competitive titration is performed. Thus, the single Ala muta-
histidine ligands of the second ZF domain of Sp1 to non- tion in CP-1(CAHH) leads to a significant weakening of the
coordinating ligands (alanine or glycine), the resultant metal binding affinity for Co(II) ions, but this result is not surprising
site gave Co(II) UV-vis spectra consistent with a pentacoordinate given that one of the peptide-derived metal-binding ligands has
geometry. The ligand sets of the Sp1 peptides were CCHG, been removed. Despite the deletion of one of the metal-binding
CCAH, and CCGH respectively, and extinction coefficients (e) Cys donors, the K
d
for CP-1(CAHH) is still within the regime of
ꢁ
1
ꢁ1
ꢁ1
ꢁ1
ꢁ1
ꢁ1
of 191 M cm , 165 M cm , and 135 M cm at 655, dissociation constants that have been reported for cobalt(II)
4
3,44
36,67,68
620 and 625 nm for Co(II) coordination were reported.
coordination to ZF proteins.
There are also examples of these types of metal centers in
naturally occurring zinc enzymes. For example, when Co(II) is
Zn(II) coordination of CP-1(CAHH)
bound to the zinc enzyme farnesyltransferase (FTase), the The cobalt(II) ion was next used as a spectroscopic probe for
UV-vis spectrum exhibits a lmax at 560 nm, with an extinc- measuring the binding of zinc(II) to CP-1(CAHH), which is
ꢁ
1
ꢁ1
tion coefficient of 140 M
cm , which is indicative of a spectroscopically silent. This strategy is well established for
19,24,28,64,68–71
5
-coordinate metal center with at least one water molecule bound measuring zinc(II) binding affinities for ZF proteins.
66
to the metal.
A solution of fully loaded Co(II)–CP-1(CAHH) was prepared by
In addition to creating an open coordination site to which combining the peptide with excess Co(II) ions (20 equiv. of
water can bind, it is not surprising that mutations that alter the CoCl ) at pH 7.5. Spectroscopic titrations were performed by
metal binding site in ZFs have also been shown to affect the the addition of successive aliquots of ZnCl , leading to the
ZF’s affinity for the metal ion. For instance, Klemba and Regan progressive loss of cobalt(II) d–d transitions at 630 and 550 nm.
demonstrated that mutagenesis of the peptide Za to favor A plot of A630nm versus [ZnCl ] is shown in Fig. 4a. The sharp
2
2
4
2
pentacoordinate geometry at the metal site resulted in a 33-fold decrease in the absorbance for the Co(II) marker band at
decrease in Co(II) ion binding affinity when compared to the 630 nm is indicative of displacement of Co(II) by Zn(II), con-
6
5
Co(II) binding to the wild type Za peptide.
firming the coordination of Zn(II) by CP-1(CAHH). The lack
We thus performed spectroscopic titrations of CP-1(CAHH) of curvature in this plot indicates that the Zn(II) ion is binding
with cobalt(II) ions to determine the effect of the alanine mutation very tightly in this concentration range, and the K for Zn(II)
4
d
Fig. 3 (a) Plot of the change in the absorption spectrum between 250–800 nm as 600 mM CP-1(CAHH) is titrated with CoCl
experiments were performed in 100 mM HEPES, 50 mM NaCl at pH 7.5. (b) Plot of absorbance at 630 nm versus the concentration of Co(II)Cl
2
. All spectrophotometric
added to
2
CP-1(CAHH). The data were fit to a 1 : 1 binding equilibrium to yield an upper limit dissociation constant of 170 mM. The solid line represents a nonlinear
least-squares fit to the 1 : 1 binding model.
This journal is ©The Royal Society of Chemistry 2016
Mol. BioSyst.

View Article Online
Paper
Molecular BioSystems
is too small to be measured with these data. These results resonances was observed between 7.2–8.2 ppm upon lowering
for Zn(II) ion coordination to CP-1(CAHH) as compared to the pH to 6.0 and adding a slight excess of Zn(II) ions (4 equiv.),
the Co(II) ion fall in line with other ZF proteins, in which as seen by comparison of the spectra for apo-CP-1(CAHH) and
the binding of Zn(II) is thermodynamically favored over that Zn–CP-1(CAHH) at pH 6.0 (Fig. 4d and e). Similar experiments
of Co(II) and typically has K s in the nanomolar to picomolar have been performed with CP-1(CCHH) and the peaks corres-
d
1
9,28,33,34,36,56,63,67–69,72–77
regime.
ponding to the amide N–H protons were dispersed over a wider
range (6.5–9.4 ppm) upon metal ion coordination at pH 6.5.
3
6,64
Secondary structure of CP-1(CAHH) upon metal ion
coordination
The NMR data indicate that CP-1(CAHH) clearly coordinates
zinc(II) ions at pH 6.0 and exhibits some secondary structure
formation upon metal ion binding, but does not fold to the
same extent as CP-1(CCHH).
The binding of zinc(II) ions to ZFs can be expected to induce a
structural change in the protein. ZF domains in the apo form are
usually unstructured, only adopting secondary structure upon the
3
2,78
Hydrolysis of 4-nitrophenyl acetate (4-NA)
coordination of a metal ion [e.g. Zn(II), Co(II)].
The chemical
shift dispersion of amide N–H resonances can be correlated with The ability of Zn(II)–CP-1(CAHH) and Co(II)–CP-1(CAHH) to mediate
the presence of secondary structural elements in small peptides hydrolysis was examined with the test substrate 4-nitrophenyl
7
9,80
and proteins,
and the folding of CP-1(CAHH) was examined acetate (4-NA). The 4-NA substrate has been commonly used
1
by 1-D ( H) NMR spectroscopy in the presence and absence of to measure the hydrolytic efficiency of both zinc complexes,
zinc(II) ions. The 1-D ( H) NMR spectrum for apo-CP-1(CAHH) is zinc peptides and proteins as well as other types of hydrolase
shown in Fig. 4b. The amide N–H peaks are found in the region mimics,
.5–8.2 ppm and are not well dispersed, as expected for an tivity of CP-1(CAHH) against other hydrolytically active zinc
unstructured peptide. The addition of one equivalent of ZnCl complexes. The M–CP-1(CAHH) [M = Zn(II) or Co(II)] peptide was
1
4
9,59,82–93
allowing for a direct comparison of the reac-
6
2
to apo-CP-1(CAHH) at pH 7.5 resulted in shifts in the aromatic incubated with 4-NA at pH 7.5 in 100 mM HEPES (1% acetonitrile,
region, which are likely due to Zn(II) binding to the histidine 50 mM NaCl) buffer and the hydrolysis reaction was monitored
ligands and some amide proton dispersion (Fig. 4c). It is well by UV-visible spectroscopy. All reactions were performed at
known that a slight lowering of the pH is often necessary to 100 mM M–CP-1(CAHH). The desired reaction gives the expected
induce good dispersion in the amide N–H region associated product, 4-nitrophenolate (4-NP), which exhibits a characteristic
8
1
ꢁ1
ꢁ1
with protein folding. Additional dispersion of the amide N–H absorbance at lmax = 400 nm (e = 12 800 M cm at pH 7.5),
Fig. 4 (A) Plot of absorbance at 630 nm versus the concentration of ZnCl
2
added to a saturated solution of Co(II)–CP-1(CAHH) [20 : 1
1
Co(II) : CP-1(CAHH)]. All spectrophotometric experiments were performed in 100 mM HEPES, 50 mM NaCl at pH 7.5. (B) H-NMR spectrum of apo-
1
1
CP-1(CAHH) at pH 7.5, (C) H-NMR spectrum of apo-CP-1(CAHH) with 1 equivalent of ZnCl
2
at pH 7.5, (D) H-NMR spectrum of apo-CP-1(CAHH) at
1
pH 6, (E) H-NMR spectrum of apo-CP-1(CAHH) with 4 equivalents of ZnCl
2
at pH 6. The peptides for all NMR experiments were at a concentration of
3
2
50 mM in 600 mL of 25 mM deuterated Tris with 5% D O.
Mol. BioSyst.
This journal is ©The Royal Society of Chemistry 2016

View Article Online
Molecular BioSystems
Paper
Fig. 7 Plot of n
-NA with Co(II)–CP-1(CAHH) (blue line) and Zn(II)–CP-1(CAHH) (pink line).
All data were collected at pH 7.5, 100 mM HEPES, 50 mM NaCl, 25 1C.
i
/[M–CP-1(CAHH)] as a function of [4-NA] for the reaction of
4
00
Table 1 Second order rate constants (k ) of the hydrolysis of 4-NA by
various complexes and peptides
Fig. 5 A plot of the increase in the absorbance at 400 nm as 100 mM
Zn(II)–CP-1(CAHH) reacts with 100 mM 4-NA to yield the chromogenic
product 4-NP.
00
ꢁ1
ꢁ1
k
(M
s
)
Ref.
Complex
(
(
(
(
PATH)ZnOH
[12]aneN )ZnOH
Cyclen)ZnOH
CH cyclen)ZnOH
0.089 ꢀ 0003
0.036 ꢀ 0.003
0.1 ꢀ 0.01
58
83
84
85
81
86
86
86
3
appearing over time (Fig. 5). An initial rate method was employed
to monitor the kinetics of hydrolysis of 4-NA. Initial rates (ni)
3
0.047 ꢀ 0.001
0.6 ꢀ 0.06
([15]aneN O )ZnOH
3
2
were obtained from the best-fit lines of [4-NP] versus time plots ZnL1
0.3908 ꢀ 0.1
0.2791 ꢀ 0.02
0.3863 ꢀ 0.02
ZnL3
ZnL8
(
Fig. 6). An initial rate method was selected because of the relatively
slow rates observed for these reactions. The kinetics were measured
over a range of substrate concentrations (0.5–2 equiv. vs. Protein/peptide
(
His94Cys)CAII
117 ꢀ 20
82
43
43
43
43
43
43
i 0
[M–CP-1(CAHH)]), and plots of n /[M–CP-1(CAHH)] versus
Zn(II)–Sp1(CCAH)
Zn(II)–Sp1(CCHA)
0.232 ꢀ 0.0051
0.568 ꢀ 0.0228
0.458 ꢀ 0.0021
0.478 ꢀ 0.0057
0.370 ꢀ 0.0289
0.966 ꢀ 0.0492
0.482 ꢀ 0.034
0.556 ꢀ 0.072
[4-NA] for both Zn(II)- and Co(II)–CP-1(CAHH) are shown in
Fig. 7. The best-fit lines in Fig. 7 yield the second-order rate Zn(II)–Sp1(CGHH)
0
0
ꢁ1 ꢁ1
ꢁ1 ꢁ1
Zn(II)–Sp1(AHHH)
Zn(II)–Sp1(HHAH)
Zn(II)–Sp1(HHHH)
constants k = 0.556 ꢀ 0.072 M
s
and 0.482 ꢀ 0.034 M
s
for the Co(II) and Zn(II) peptides, respectively (eqn (7)). These
rate constants are on the high end of those reported for many Zn(II)–CP-1(CAHH)
This work
This work
ꢁ
1
ꢁ1
Co(II)–CP-1(CAHH)
zinc complexes, which mostly range from 0.036–0.6 M
s
5
9,82,84–87
(Table 1).
The rates are comparable to those seen
for the hydrolytic activity of mutant peptides based on the ZF
4
4
domain of Sp1 reported by Nomura and Sugiura, in which the
metal-binding ligands were systematically varied (Table 1) and
the hydrolytic cleavage of 4-NA was examined. Taken together,
these data show that CP-1 can be re-engineered into a hydro-
lytically active peptide through ligand substitution at the
metal center.
Conclusions
We have shown that the consensus peptide for a classical ZF
domain, CP-1, can be modified by rational design to convert the
structural zinc site into a zinc site capable of efficiently mediating
the hydrolysis of an external substrate. The single mutation of
Cys-to-Ala at the metal-binding site was made to give a CHH
binding motif (CP-1(CAHH)), based on the known ligand set for
Fig. 6 A plot of 4-NP concentration (A400/(eꢂl)) as a function of time
indicating the production of 4-NP in the reaction of 100 mM 4-NA with
Co(II)–CP-1(CAHH) (in blue), or Zn(II)–CP-1(CAHH) (in pink), all at 100 mM.
All data were collected at pH 7.5 in 100 mM HEPES, 50 mM NaCl at 25 1C. the hydrolytically active Zn(II)/Fe(II) enzyme peptide deformylase.
This journal is ©The Royal Society of Chemistry 2016
Mol. BioSyst.

View Article Online
Paper
Molecular BioSystems
Despite the removal of one of the key metal binding ligands, stored and manipulated under anaerobic conditions at all
CP-1(CAHH) is still capable of binding Co(II) and Zn(II) with times (Coy Anaerobic Box, (95% N
2 2
, 5% H )).
good affinity, and shows some secondary structure formation
upon binding metal ion, albeit less structurally ordered than
UV-visible metal binding titrations
the native CP-1(CCHH) peptide. The hydrolytic cleavage of the Metal binding titrations were performed on a PerkinElmer
test substrate 4-NA mediated by M(II)–CP-1(CAHH) (M = Co, Zn) Lambda 25 UV-visible spectrometer. In a typical experiment,
is faster than many of the known zinc model complexes that CoCl was titrated into a solution of apo-CP-1(CAHH) (600 mM
2
have been designed as catalytic mimics for zinc enzymes and in 1 mL) dissolved in HEPES (4-(2-hydroxyethyl)-1-piperazineethane-
tracks with the rates reported by Sugiura and co-workers for sulfonic acid) (100 mM) with NaCl (50 mM) at pH 7.5. The titration
a modified Sp1 ZF peptide. We have thus shown that the CP-1 included additions of 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0,
peptide represents a potential scaffold from which to create 2.0, 5.0, 10, 15, and 20 molar equivalents of CoCl . The shape of
2
efficient, catalytically active hydrolysis enzymes that utilize the spectrum did not change during the course of the experi-
biomimetic divalent metal ions. For high catalytic activity to ment, suggesting that Co(II) binds to the same site throughout the
be achieved, the non-coordinating amino acids may need to be titration. A plot of A630 versus CoCl
was fit to a 1 : 1 binding model
2
optimized to promote hydrolysis at the zinc site as well as (eqn (5)) using non-linear least squares analysis (KaleidaGraph,
provide the proper structure to the catalytic domain. A system- Synergy Software). An upper limit dissociation constant (K ) for
d
atic tuning of the metal active site and peptide structure by cobalt(II) was determined.
varying residues through a combinatorial and/or rational
approach may lead to such optimization. We have shown that
CP-1 may be an ideal system for this approach, given its high
stability and its ability to tolerate mutations without loss of
structural integrity.
P + Co $ PCo
(3)
(4)
½
Pꢃ½Coꢃ
Kd ¼
½
PCoꢃ
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
À
Á
2
½
Pꢃtotal þ ½Coꢃtotal þ Kd ꢁ
½Pꢃ
þ ½Coꢃ
þ Kd ꢁ 4½Pꢃ ½Coꢃ
total
total
total
total
(5)
½PCoꢃ ¼
2½Coꢃ
total
Materials and methods
Identification of consensus sequence for CCHH ZFs
where P = Peptide, Co = Cobalt. The ability of Zn(II) to displace
Co(II) was determined by titrating Zn(II) into a solution of
Co(II)–peptide (20 : 1 Co(II) : peptide stoichiometry). As Zn(II)
was titrated, the Co(II) d–d bands disappeared, indicating that
Zn(II) was replacing Co(II) at the coordination site.
(
CP-1 update)
To determine an updated consensus classical ZF sequence, a
search of ‘‘zinc finger’’ using PROSITE (prosite.expasy.org), which
is part of the ExPASy suite of programs, was performed.
9
4,95
From
Nuclear magnetic resonance (NMR) spectroscopy
this search, a document ‘PDOC00028,’ which contains informa-
tion about all CCHH-type ZF domains for which sequence
information is available, was selected. The weblogo for the
updated consensus sequence, named CP-1(CCHH)-2015 was
generated via the ‘retrieve the sequence logo from the align-
All NMR experiments were recorded on a 600 MHz Bruker spectro-
meter with the temperature maintained at 25 1C. Apo-CP-
1
2
(CAHH) and Zn(II)–CP-1(CAHH) (350 mM) were prepared in
5 mM deuterated tris (tris(hydroxymethyl)aminomethane)
containing 5% D O at pH 7.5 and at pH 6.0. The Zn(II)-(CAHH)
5
8,96
2
ment’ link.
true positive hits from the UniProtKB/Swiss-Prot databank and
This sequence logo was constructed from 13 465
samples included 1 and 4 equivalents of Zn(II) at pH values of
7.5 and 6.0, respectively. All of the NMR spectral data were
9
7
is shown in Fig. 1a.
processed using Spinworks software (http://www.umanitoba.ca./
chemistry/nmr/spinworks).
Peptide preparation
The CP-1(CAHH) peptide (PYKCPEAGKSFSQKSDLVKHQRTHTG)
Hydrolysis of 4-nitrophenyl acetate (4-NA)
was purchased from Biomatik (Wilmington, DE) and stored in The hydrolysis of 4-NA by the peptides was monitored using a
the solid form at ꢁ20 1C until use. A small amount of the solid PerkinElmer Lambda 25 UV-visible spectrometer by following
peptide was resuspended in buffer prior to use and assayed for the production of the reaction product, 4-nitrophenolate (4-NP),
purity. A single peak was observed at 20% acetonitrile : 80% at 400 nm. Each reaction was performed in 100 mM HEPES,
water via High Performance Liquid Chromatography (HPLC). 50 mM NaCl, pH 7.5 with 1% acetonitrile. 100 mM M–CP-
The identity of the isolated peak was confirmed by Matrix 1(CAHH) [M = Zn(II) or Co(II)] (prepared at a 1 : 1 stoichiometry
Assisted Laser Desorption Ionization-Time of Flight Mass Spectro- of M(II) : peptide) was incubated with various concentrations of
+
metry (MALDI-TOF MS); [observed: 2931.23 (M ); expected 4-NA (dissolved in acetonitrile) and the absorbance at 400 nm
2931.45]. The reduction state of the peptide was determined by was recorded at 10 second intervals. The initial rate of hydro-
measuring the ratio of total peptide to Co(II)–peptide. Typically, lysis was monitored up to a 5% yield of the product, 4-NP.
7
1,98
495% of the peptide was reduced.
The peptide was Background hydrolysis was measured by incubation of 4-NA
Mol. BioSyst.
This journal is ©The Royal Society of Chemistry 2016

View Article Online
Molecular BioSystems
Paper
with either 100 mM ZnCl
2
or 100 mM CoCl
2
in the above buffer to 14 J. M. Matthews and M. Sunde, IUBMB Life, 2002, 54,
determine if the metal ions alone could promote hydrolysis. In
both instances, a low level of hydrolysis was recorded, and the 15 J. S. Hanas, D. J. Hazuda, D. F. Bogenhagen, F. Y. Wu and
appropriate spectrum for this hydrolysis was subtracted from
C. W. Wu, J. Biol. Chem., 1983, 258, 14120–14125.
the spectrum of the corresponding M–CP-1(CAHH) reaction 16 J. Miller, A. D. McLachlan and A. Klug, EMBO J., 1985, 4,
with 4-NA. The method of initial rates was employed to monitor
1609–1614.
the reaction kinetics, and was then analyzed with the kinetic 17 J. M. Berg, Science, 1986, 232, 485–487.
model shown in eqn (6) and (7):
18 J. M. Berg, Proc. Natl. Acad. Sci. U. S. A., 1988, 85,
9–102.
351–355.
9
0
0
n
i
= k [M–CP-1(CAHH)]
0
[4-NA]
0
(6)
1
2
2
9 A. M. Rich, E. Bombarda, A. D. Schenk, P. E. Lee, E. H. Cox,
A. M. Spuches, L. D. Hudson, B. Kieffer and D. E. Wilcox,
J. Am. Chem. Soc., 2012, 134, 10405–10418.
0 K. L. Chan, I. Bakman, A. R. Marts, Y. Batir, T. L. Dowd,
D. L. Tierney and B. R. Gibney, Inorg. Chem., 2014, 53,
ni
0
0
00
¼
MꢁCP-1ðCAHHÞꢃ0
k ½4 -NAꢃ
(7)
0
½
where n = the initial rate of the reaction up to 5% conversion;
M–CP-1(CAHH)]0 = initial concentration of metal peptide
i
[
6
309–6320.
0
0
complex; [4-NA]
0
= initial concentration of substrate; k = the
was obtained from
linear fits of concentration vs. time plots for 4-nitrophenolate
1 A. R. Reddi, T. R. Guzman, R. M. Breece, D. L. Tierney and
B. R. Gibney, J. Am. Chem. Soc., 2007, 129, 12815–12827.
2 J. M. Berg, J. Biol. Chem., 1990, 265, 6513–6516.
3 J. M. Berg and Y. Shi, Science, 1996, 271, 1081–1085.
4 J. M. Berg and H. A. Godwin, Annu. Rev. Biophys. Biomol.
Struct., 1997, 26, 357–371.
5 D. Jantz, B. T. Amann, G. J. Gatto, Jr. and J. M. Berg, Chem.
Rev., 2004, 104, 789–799.
6 A. N. Besold and S. L. Michel, Biochemistry, 2015, 54,
4443–4452.
7 J. M. Berg and H. A. Godwin, Annu. Rev. Biophys. Biomol.
Struct., 1997, 26, 357–371.
8 J. C. Payne, B. W. Rous, A. L. Tenderholt and H. A. Godwin,
Biochemistry, 2003, 42, 14214–14224.
9 O. Seneque, E. Bonnet, F. L. Joumas and J. M. Latour,
Chemistry, 2009, 15, 4798–4810.
0 S. M. Quintal, Q. A. dePaula and N. P. Farrell, Metallomics,
2
1 R. T. Doku, G. Park, K. E. Wheeler and K. E. Splan, JBIC,
J. Biol. Inorg. Chem., 2013, 18, 669–678.
32 A. D. Frankel, J. M. Berg and C. O. Pabo, Proc. Natl. Acad. Sci.
U. S. A., 1987, 84, 4841–4845.
second-order rate constant. The initial rate n
i
2
2
2
(
(
4-NP), where the concentration is given by [4-NP] = A400/eꢂl
ꢁ1 ꢁ1
e = 12 800 M cm ; l = 1 cm path length). Second-order rate
00
constants (k values) were determined according to eqn (7) by
taking the slopes of the best-fit lines of n /[M–CP-1(CAHH)] vs.
i
0
2
2
2
2
2
3
3
5
9,60,99
[4-NA].
All experiments were performed in triplicate.
Acknowledgements
SLJM would like to thank the NSF (CHE-1306208), ANB would like
to thank the NIH (F31NS074768) and DPG would like to thank
the NIH (GM062309, GM101153) for support of this research.
We also acknowledge the Biomolecular NMR Center at JHU and
Dr. Ananya Majumdar for assistance with NMR experiments.
011, 3, 121–139.
References
1
W. Shi and M. R. Chance, Curr. Opin. Chem. Biol., 2011, 15,
44–148.
S. M. Yannone, S. Hartung, A. L. Menon, M. W.
1
2
Adams and J. A. Tainer, Curr. Opin. Biotechnol., 2012, 23, 33 B. A. Krizek, B. T. Amann, V. J. Kilfoil, D. L. Merkle and
9–95.
J. M. Berg, J. Am. Chem. Soc., 1991, 113, 4518–4523.
R. H. Holm, P. Kennepohl and E. I. Solomon, Chem. Rev., 34 B. A. Krizek, D. L. Merkle and J. M. Berg, Inorg. Chem., 1993,
8
3
1996, 96, 2239–2314.
32, 937–940.
4
5
6
7
8
9
A. C. Rosenzweig, Chem. Biol., 2002, 9, 673–677.
W. Maret, J. Inorg. Biochem., 2012, 111, 110–116.
W. Maret and Y. Li, Chem. Rev., 2009, 109, 4682–4707.
W. Maret, Met. Ions Life Sci., 2013, 12, 479–501.
W. Maret, Adv. Nutr., 2013, 4, 82–91.
35 B. A. Krizek, L. E. Zawadzke and J. M. Berg, Protein Sci.,
1993, 2, 1313–1319.
36 O. Seneque and J. M. Latour, J. Am. Chem. Soc., 2010, 132,
17760–17774.
37 F. Yu, V. M. Cangelosi, M. L. Zastrow, M. Tegoni, J. S.
Plegaria, A. G. Tebo, C. S. Mocny, L. Ruckthong, H. Qayyum
and V. L. Pecoraro, Chem. Rev., 2014, 114, 3495–3578.
C. Andreini, I. Bertini and G. Cavallaro, PLoS One, 2011,
6, e26325.
1
1
1
1
0 I. Bertini, L. Decaria and A. Rosato, J. Biol. Inorg. Chem., 38 M. L. Zastrow and V. L. Pecoraro, Biochemistry, 2014, 53,
010, 15, 1071–1078.
957–978.
1 S. J. Lee and S. L. Michel, Acc. Chem. Res., 2014, 47, 39 B. L. Vallee and D. S. Auld, Proc. Natl. Acad. Sci. U. S. A.,
643–2650. 1990, 87, 220–224.
2 J. L. Michalek, A. N. Besold and S. L. Michel, Dalton Trans., 40 K. A. McCall, C. Huang and C. A. Fierke, J. Nutr., 2000, 130,
011, 40, 12619–12632. 1437S–1446S.
3 J. H. Laity, B. M. Lee and P. E. Wright, Curr. Opin. Struct. 41 K. T. Nguyen, J. C. Wu, J. A. Boylan, F. C. Gherardini and
Biol., 2001, 11, 39–46. D. Pei, Arch. Biochem. Biophys., 2007, 468, 217–225.
2
2
2
This journal is ©The Royal Society of Chemistry 2016
Mol. BioSyst.

View Article Online
Paper
Molecular BioSystems
4
4
4
4
4
4
2 D. L. Merkle, M. H. Schmidt and J. M. Berg, J. Am. Chem. 69 A. L. Guerrerio and J. M. Berg, Biochemistry, 2004, 43,
Soc., 1991, 113, 5450–5451. 5437–5444.
3 A. Nomura and Y. Sugiura, Inorg. Chem., 2002, 41, 70 M. T. Worthington, B. T. Amann, D. Nathans and J. M. Berg,
693–3698. Proc. Natl. Acad. Sci. U. S. A., 1996, 93, 13754–13759.
4 A. Nomura and Y. Sugiura, Inorg. Chem., 2004, 43, 71 S. J. Lee, J. L. Michalek, A. N. Besold, S. E. Rokita and
708–1713. S. L. Michel, Inorg. Chem., 2011, 50, 5442–5450.
5 A. Nomura and Y. Sugiura, J. Am. Chem. Soc., 2004, 126, 72 S. L. Michel, A. L. Guerrerio and J. M. Berg, Biochemistry,
5374–15375. 2003, 42, 4626–4630.
6 S. Negi, Y. Umeda, S. Masuyama, K. Kano and Y. Sugiura, 73 A. N. Besold, A. A. Oluyadi and S. L. Michel, Inorg. Chem.,
Bioorg. Med. Chem. Lett., 2009, 19, 2789–2791. 2013, 52, 4721–4728.
7 S. Negi, M. Yoshioka, H. Mima, M. Mastumoto, M. Suzuki, 74 R. C. diTargiani, S. J. Lee, S. Wassink and S. L. Michel,
3
1
1
M. Yokoyama, K. Kano and Y. Sugiura, Bioorg. Med. Chem.
Lett., 2015, 25, 4074–4077.
8 M. Imanishi, S. Negi and Y. Sugiura, Methods Mol. Biol.,
Biochemistry, 2006, 45, 13641–13649.
75 M. C. Posewitz and D. E. Wilcox, Chem. Res. Toxicol., 1995, 8,
1020–1028.
4
4
5
5
2
010, 649, 337–349.
76 M. J. Lachenmann, J. E. Ladbury, J. Dong, K. Huang,
P. Carey and M. A. Weiss, Biochemistry, 2004, 43,
13910–13925.
9 M. L. Zastrow, A. F. Peacock, J. A. Stuckey and V. L. Pecoraro,
Nat. Chem., 2012, 4, 118–123.
0 M. L. Zastrow and V. L. Pecoraro, J. Am. Chem. Soc., 2013, 77 A. N. Besold, D. L. Amick and S. L. Michel, Mol. BioSyst.,
35, 5895–5903.
2014, 10, 1753–1756.
1 T. Meinnel, C. Lazennec and S. Blanquet, J. Mol. Biol., 1995, 78 M. S. Lee, G. P. Gippert, K. V. Soman, D. A. Case and
1
2
54, 175–183.
P. E. Wright, Science, 1989, 245, 635–637.
79 K. Wuthrich, Angew. Chem., Int. Ed. Engl., 2003, 42,
3340–3363.
5
5
5
2 F. Sanger, Nat. Med., 2001, 7, 267–268.
3 A. J. Simpson, Nat. Rev. Genet., 2001, 2, 979–983.
4 R. S. Brown, C. Sander and P. Argos, FEBS Lett., 1985, 186, 80 H. J. Dyson and P. E. Wright, Annu. Rev. Phys. Chem., 1996,
71–274.
47, 369–395.
5 C. A. Blasie and J. M. Berg, Biochemistry, 2004, 43, 81 K. W u¨ thrich, NMR of Proteins and Nucleic Acids, Wiley,
0600–10604. 1991.
6 C. A. Blasie and J. M. Berg, Biochemistry, 2002, 41, 82 C. Bazzicalupi, A. Bencini, A. Bianchi, F. Corana, V. Fusi,
2
5
5
5
5
5
1
1
5068–15073.
7 C. A. Blasie and J. M. Berg, Biochemistry, 1997, 36,
218–6222.
8 T. D. Schneider and R. M. Stephens, Nucleic Acids Res., 1990,
8, 6097–6100.
C. Giorgi, P. Paoli, P. Paoletti, B. Valtancoli and C. Zanchini,
Inorg. Chem., 1996, 35, 5540–5548.
83 L. L. Kiefer and C. A. Fierke, Biochemistry, 1994, 33,
15233–15240.
84 E. Kimura, I. Nakamura, T. Koike, M. Shionoya, Y. Kodama,
T. Ikeda and M. Shiro, J. Am. Chem. Soc., 1994, 116,
4764–4771.
6
1
9 R. C. diTargiani, S. Chang, M. H. Salter, Jr., R. D.
Hancock and D. P. Goldberg, Inorg. Chem., 2003, 42,
5
825–5836.
85 T. Koike, M. Takamura and E. Kimura, J. Am. Chem. Soc.,
1994, 116, 8443–8449.
6
0 D. P. Goldberg, R. C. diTargiani, F. Namuswe, E. C.
Minnihan, S. Chang, L. N. Zakharov and A. L. Rheingold, 86 T. Koike, S. Kajitani, I. Nakamura, E. Kimura and M. Shiro,
Inorg. Chem., 2005, 44, 7559–7569. J. Am. Chem. Soc., 1995, 117, 1210–1219.
1 V. V. Karambelkar, C. Xiao, Y. Zhang, A. A. Sarjeant and 87 M. Subat, K. Woinaroschy, S. Anthofer, B. Malterer and
D. P. Goldberg, Inorg. Chem., 2006, 45, 1409–1411. B. Konig, Inorg. Chem., 2007, 46, 4336–4356.
2 P. T. Rajagopalan, S. Grimme and D. Pei, Biochemistry, 2000, 88 K. L. Duncan and R. V. Ulijn, Biocatalysis, 2015, 1, 67–81.
6
6
6
6
6
6
6
6
3
9, 779–790.
3 I. Bertini and C. Luchinat, Adv. Inorg. Biochem., 1984, 6,
1–111.
4 J. M. Berg and D. L. Merkle, J. Am. Chem. Soc., 1989, 111,
759–3761.
5 M. Klemba and L. Regan, Biochemistry, 1995, 34,
0094–10100.
89 N. Singh, M. P. Conte, R. V. Ulijn, J. F. Miravet and
B. Escuder, Chem. Commun., 2015, 51, 13213–13216.
90 B. Goyal, K. Patel, K. R. Srivastava and S. Durani, RSC Adv.,
2015, 5, 105400.
91 C. M. Rufo, Y. S. Moroz, O. V. Moroz, J. Stohr, T. A. Smith,
X. Hu, W. F. DeGrado and I. V. Korendovych, Nat. Chem.,
2014, 6, 303–309.
7
3
1
6 C. C. Huang, P. J. Casey and C. A. Fierke, J. Biol. Chem., 1997, 92 Y. Maeda, N. Javid, K. Duncan, L. Birchall, K. F. Gibson,
2
72, 20–23.
7 J. S. Magyar and H. A. Godwin, Anal. Biochem., 2003, 320,
9–54.
D. Cannon, Y. Kanetsuki, C. Knapp, T. Tuttle, R. V.
Ulijn and H. Matsui, J. Am. Chem. Soc., 2014, 136,
15893–15896.
3
8 A. B. Ghering, J. E. Shokes, R. A. Scott, J. G. 93 C. Zhang, X. Xue, Q. Luo, Y. Li, K. Yang, X. Zhuang, Y. Jiang,
Omichinski and H. A. Godwin, Biochemistry, 2004, 43,
346–8355.
J. Zhang, J. Liu, G. Zou and X. J. Liang, ACS Nano, 2014, 8,
11715–11723.
8
Mol. BioSyst.
This journal is ©The Royal Society of Chemistry 2016

View Article Online
Molecular BioSystems
Paper
94 C. J. Sigrist, L. Cerutti, E. de Castro, P. S. Langendijk- 96 G. E. Crooks, G. Hon, J. M. Chandonia and S. E. Brenner,
Genevaux, V. Bulliard, A. Bairoch and N. Hulo, Nucleic Acids
Res., 2010, 38, D161–166.
Genome Res., 2004, 14, 1188–1190.
97 T. U. Consortium, Nucleic Acids Res., 2012, 40, D71–D75.
95 C. J. Sigrist, E. de Castro, L. Cerutti, B. A. Cuche, N. Hulo, 98 S. J. Lee and S. L. Michel, Inorg. Chem., 2010, 49, 1211–1219.
A. Bridge, L. Bougueleret and I. Xenarios, Nucleic Acids Res., 99 S. A. Li, J. Xia, D. X. Yang, Y. Xu, D. F. Li, M. F. Wu and
2013, 41, D344–D347.
W. X. Tang, Inorg. Chem., 2002, 41, 1807–1815.
This journal is ©The Royal Society of Chemistry 2016
Mol. BioSyst.