Clickable 5′-γ-Ferrocenyl Adenosine Triphosphate Bioconjugates in Kinase-Catalyzed Phosphorylations

Article abstract of DOI:10.1002/chem.201405510

Clickable co-substrate: A tri-functional 5′-γ-ferrocenyl adenosine triphosphate (Fc-ATP) derivative containing a clickable site was synthesized (see figure). This compound is an effective co-substrate in kinase-catalyzed phosphorylation reactions, which can be detected by both electrochemical and immunoassay detection methods. The clickable reaction site makes direct modification possible, which greatly expands its application.

Full text of DOI:10.1002/chem.201405510

DOI: 10.1002/chem.201405510
Full Paper
&
Bioconjugates |Hot Paper|
Clickable 5’-g-Ferrocenyl Adenosine Triphosphate Bioconjugates
in Kinase-Catalyzed Phosphorylations
[a, e]
[a]
[a, b]
[c]
[d]
Nan Wang,
Zhe She, Yen-Chun Lin,
Sanela Marti c´ , David J. Mann, and
[a, b]
Heinz-Bernhard Kraatz*
A 5’-g-ferrocenyl adenosine triphosphate (Fc-ATP) conjugate,
Fc-CO-Lys-ATP, containing a clickable alkyne moiety was syn-
thesized and compared to a non-alkynyl Fc-ATP analogue, Fc-
CO-C6-ATP, as co-substrate in kinase-catalyzed phosphorylation
reactions. Sarcoma-related kinase (Src), cyclin-dependent
kinase (CDK2), and casein kinase II (CK2a) were chosen for this
study. Surface electrochemical studies indicate that Fc-CO-Lys-
ATP is an effective co-substrate due to the generated Fc cur-
rent density signals, which is in line with the XPS measure-
ment. Molecular modeling studies were also carried out to
help understanding the high efficiency of Fc-CO-Lys-ATP as co-
substrate in CDK2-catalyzed phosphorylation reactions. The
steric hindrance due to Fc-CO-Lys-ATP binding to the Src
active pocket, but not CDK2 kinase, negatively affects substrate
interaction and reduces Fc-phosphorylation. The compatibility
of Fc-CO-Lys-ATP with antiferrocene antibodies and a fluores-
cently labeled secondary antibody was also demonstrated in
a biochemical format. Additionally, scanning electrochemical
microscopy (SECM) was successfully used to probe the phos-
phorylation reaction for the first time. The clickable nature of
Fc-CO-Lys-ATP was demonstrated on gold surfaces. The pres-
ent study illustrates that Fc-CO-Lys-ATP is an efficient co-sub-
strate for kinase-catalyzed phosphorylation reactions and
adaptable to multiplex arrays.
Introduction
Protein kinases regulate the majority of cellular pathways
through chemically adding phosphate groups, for example
from an adenosine triphosphate (ATP) molecule, to other pro-
teins, which accordingly results in modulation of downstream
[1]
functions. It has been shown that the over/under expression
of certain protein kinases has been linked to human disease in-
[2]
cluding many forms of cancer. Consequently, assessing the
activities and substrates of protein kinases is of great interest
from a basic biological perspective and in terms of therapeutic
[1]
targets.
The traditionally used method for studying protein kinase
32
[4]
activity is the [g- P]ATP radiometric assay. However, due to
the well-known drawbacks of radioactive materials, environ-
mental friendly methods are desired to replace this traditional
approach. Thus, numerous techniques have emerged including
[
5]
[6]
[7]
electrochemical, optical, mass spectrometry, and surface-
[8]
plasmon resonance-based systems. Among them, g-phos-
phate-modified ATP analogues have been synthesized and
demonstrated to be effective for these uses, such as biotin-,
[9]
dansyl-, azide-, and ferrocene (Fc)-labeled ATP compounds.
[
a] Dr. N. Wang, Dr. Z. She, Y.-C. Lin, Prof. H.-B. Kraatz
Department of Physical and Environmental Science
University of Toronto Scarborough
Additionally, a systematic study of the compatibility between
these ATP analogues and specific kinases (i.e., PKA, CK2, and
[10]
1
265 Military Trail, Toronto, Ontario M1C 1A4 (Canada)
Abl) has been reported recently. Among the electrochemical
systems, ferrocene-modified bioconjugates are one of the
most widely used redox reporters due to their ideal electro-
E-mail: bernie.kraatz@utoronto.ca
[b] Y.-C. Lin, Prof. H.-B. Kraatz
Department of Chemistry, University of Toronto
[11]
chemical properties and their stability in aqueous solutions.
Numerous Fc–amino acid and Fc–peptide conjugates have
8
0 St. George Street, Toronto, M5S 3H6 (Canada)
[
c] Prof. S. Martic´
[12]
Department of Chemistry, Oakland University
been reported. Their applications include but are not limit to
[
13]
[14]
2
200 North Squirrel Road
peptide foldamers,
electrochemical detection,
and bio-
Rochester, Michigan 48309 (USA)
[15]
medical application. In 2008, we reported the synthesis of
[d] Dr. D. J. Mann
a novel Fc-modified ATP conjugate, namely, 5’-g-ferrocenyl-ATP
Department of Life Sciences, Imperial College London
London SW7 2AZ (UK)
(
Fc-CO-C6-ATP), and the effectiveness of this compound as
a co-substrate to detect the kinase-catalyzed phosphorylation
[
e] Dr. N. Wang
Current address:
[9d]
reaction of surface-bound peptides (as shown in Scheme 1).
Beijing Key Laboratory of
Based on this discovery, a series of Fc-ATP derivative com-
pounds have been synthesized and used for kinase activity
Photoelectronic/Electrophotonic Conversion Materials
Department of Chemistry, Beijing Institute of Technology
Beijing 100081 (P.R. China)
[16]
and inhibitor screening studies. These works provided valua-
ble information for further understanding the exact nature of
the interaction between the ATP co-substrate with the pep-
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201405510.
Chem. Eur. J. 2015, 21, 4988 – 4999
4988
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
methodology between an acti-
vated ester and Boc-protected
alkyl amines in the presence of
hydroxybenzotriazole, 1-ethyl-3-
(
3-dimethylaminopropyl)carbo-
diimide, and N,N-diisopropyl-
[12j]
ethylamine. Fc-CO-Lys-Boc (2)
was obtained with good yield
(
75%) and fully characterized by
using NMR spectroscopy and
mass spectrometry. The conver-
sion of ATP into the g-substitut-
ed derivate was achieved by fol-
lowing the literature proce-
[20]
dure. The Boc-protected con-
jugate Fc-CO-Lys-Boc was first
deprotected in the presence of
an excess of trifluoroacetic acid
to generate the free amine 3.
Then the in-situ-formed adeno-
sine metaphosphate was react-
ed with the free amine under
Scheme 1. Schematic illustration of the strategy and principle for the detection of kinase-catalyzed phosphoryla-
tion by using Fc-CO-C6-ATP and Fc-Ab /Ab . The peptide/protein (blue curve) is immobilized on the N-hydroxysuc-
cinimide lipoic acid active ester (LPA) (green curve) modified Au surface. A protein kinase transfers the g-phos-
1
2
phate-Fc group to the peptide, which is then further detected through electrochemical techniques. In an immu-
noassay approach, primary polyclonal antiferrocene antibodies (Fc-Ab
duction of the electrochemical signals. This is followed by indirect immunodetection through fluorescently labeled
goat anti-rabbit antibodies (Ab ).
1
) bind to the Fc-phosphates resulting in re-
2
tide/proteins, as well as guidance for designing other effective
Fc-ATP co-substrates.
In our recent study, we have also found that a robust poly-
clonal rabbit antiferrocene antibody (Fc-Ab ) can selectively
1
bind to Fc-modified bioconjugates in both solution and on sur-
[
17]
faces. By combining this bioanalytical tool with Fc-CO-C6-
ATP, a powerful approach has been established, which can be
applied to multiplex arrays and effectively monitor kinase ac-
tivity by using both electrochemical and immunoassay detec-
tion methods as shown in Scheme 1.
In this work, we have designed and synthesized a novel Fc-
ATP derivative, namely, Fc-CO-Lys-ATP, as shown in Scheme 2.
Herein, the amino acid Lys(Boc)-OMe (Boc=tert-butoxycarbon-
yl) was used as a linker to connect the Fc moiety with the ATP
functional group. The performance of this compound as a co-
substrate in a kinase-catalyzed phosphorylation reaction was
investigated with direct comparison to Fc-CO-C6-ATP as g-
phosphate donor. The extra alkyne moiety incorporated into
[
18]
Fc-CO-Lys-ATP allows further modification, which has been
demonstrated by successful attachment of an azide-labeled
fluorescent dye through click chemistry. In addition, immuno-
detection and scanning electrochemical microscopy (SECM)
were also used to detect a phosphotransfer involving this
[19]
novel ATP derivative. This is the first time that SECM has
been used for monitoring phosphorylation process.
Results and Discussion
Synthesis and characterization of Fc-CO-Lys-ATP
Scheme 2. Synthesis of Fc-CO-Lys-ATP and click reaction on Fc-CO-Lys-Boc.
EDC=N’-(3-dimethylaminopropyl)-N-ethylcarbodiimide, HOBt=1-hydroxy-
benzotriazole, DIPEA=diisopropylethylamine, TFA=trifluoroacetic acid,
TEAB=tetraethylammonium bromide, DCC=dicyclohexylcarbodiimide,
TEA=triethylamine.
The novel Fc-modified ATP compound was synthesized accord-
ing to Scheme 2. The synthesis of the intermediate Fc-CO-Lys-
Boc (2) was based on the employment of standard coupling
Chem. Eur. J. 2015, 21, 4988 – 4999
www.chemeurj.org
4989
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
rocene compounds. Due to the solvent change (from
3
1
[a]
Table 1. P NMR chemical shifts and electrochemical data for the Fc conjugates.
[16]
CH CN to water) Fc-CO-Lys-ATP and Fc-CO-C6-ATP
3
3
1
[b]
[c]
Fc compound
P NMR [ppm]
–
E
1/2 [mV]
DE [mV]
(64Æ5)
i
pa/ipc
D
show half-wave potential at approximately 430 mV.
The diffusion coefficients (Ds) for all three com-
pounds were also calculated by using the Randles–
Fc-CO-Lys-Boc
Fc-CO-Lys-ATP
(574Æ5)
(434Æ3)
1.1
1.1
(22.1Æ0.23)
(3.06Æ0.37)
À0.96
[21]
À11.51
À23.01
À0.78
(64Æ8)
Sevcic equation, which demonstrates the effect of
different molecule sizes on their mobility and redox
activity in solution. As expected, the ATP-free com-
pound Fc-CO-Lys-Boc has a larger D value than the
ATP-modified compounds.
Fc-CO-C6-ATP
À11.48
À22.98
(431Æ3)
(59Æ8)
1.1
(4.54Æ0.11)
[
a] Electrochemical data for compound Fc-CO-Lys-Boc was obtained at 0.1 mm in
CH CN in the presence of 0.1m TBAP. Data for compounds Fc-CO-Lys-ATP and Fc-CO-
C6-ATP were measured at the concentration of 0.1 mm in the presence of 0.1m
NaClO as the supporting electrolyte. Electrochemical data were measured by cyclic
3
Electrochemical, fluorescent, and XPS detections
of protein kinase-catalyzed Fc phosphorylations
4
voltammetry with Ag/AgCl/3m KCl as the reference electrode, a Pt wire as the auxili-
ary electrode, a glassy carbon surface as the working electrode. The scan rate was
À1
1
00 mVs . The reported error was obtained from the triplicate measurements. E1/2
=
To investigate the general application of Fc-CO-Lys-
ATP as a co-substrate in protein kinase-catalyzed
phosphorylation reactions, three protein kinases, that
is, Src, CDK2 in combination with cyclin A, and CK2a,
were chosen as representative kinases that act on
Tyr, Ser, or Thr residues for the following electro-
chemical studies. Src is a protein Tyr kinase that is over ex-
half-wave potential; DE=peak separation; ipa/ipc =peak current ratio. [b] D
2
O, 298 K,
H
3
PO
4
85% as an external NMR reference. [c] Diffusion coefficient is given in [ꢁ
cm s ] and was determined electrochemically.
À5
2
À1
10
basic condition in the presence of DCC. After adding water,
the byproduct was removed by precipitation and the solution
was loaded onto a diethylaminoethyl (DEAE) cellulose column
for further purification. The desired ATP coupling fractions
were collected with 0.3–0.4m TEAB (pH 7.5) and subjected to
HPLC purification to give Fc-CO-Lys-ATP in 8% yield. Fc-CO-
Lys-ATP was characterized by means of NMR spectroscopy and
matrix-assisted laser desorption/ionization time-of-flight mass
spectrometry (MALDI-TOF MS) (as shown in Figure S1 in the
[22]
pressed in some forms of cancer. CDK2 is mainly involved in
[23]
the cell cycle and division and acts on Ser and Thr residues.
CK2a is primarily involved in the cellular metabolism and dif-
ferentiation and acts on Ser/Thr residues in key regulatory pro-
[
24]
teins. Fc-CO-C6-ATP, a demonstrated effective co-substrate,
was employed as a control.
Cyclic voltammetry and square-wave voltammetry (SWV)
were used for electrochemical studies of peptides bound to
31
Supporting Information). In the P NMR spectrum (as shown in
Table 1), compound Fc-CO-Lys-ATP exhibited three signals
coming from the a-, b-, and g-P atoms, confirming the success-
ful coupling of the ATP moiety. More importantly, the charac-
teristic doublet peak for the g-P atom at d=À0.96 ppm (vs.
H PO ) indicates the Fc-g-phosphoamide formation.
a
gold surfaces in the presence of ferri/ferrocyanide
3À/4À
([Fe(CN)6]
) as a redox probe. As shown in Figure 1, the cur-
rent density observed is reduced with each step of modifica-
tion until the loss of the redox probe signal and the peak po-
3
4
In order to confirm that the alkyne moiety in this Fc deriva-
tive is clickable, a click reaction was carried out. Due to limiting
amounts of Fc-CO-Lys-ATP, Fc-CO-Lys-Boc was used in a model
reaction. Under typical copper(I)-catalyzed click reaction condi-
tion, Fc-CO-Lys-Boc reacted with benzyl azide successfully. The
expected triazole derivative product was obtained with 82%
yield. The synthesis and characterization details are given in
the Supporting Information.
Electrochemistry of the Fc compounds in solution
The solution electrochemical properties of Fc-CO-Lys-Boc and
Fc-CO-Lys-ATP were investigated by cyclic voltammetry (CV) in
CH CN solution containing tetrabutylammonium perchlorate
3
(
TBAP) as electrolyte and in NaClO solution, respectively. The
4
data were summarized and are shown in Table 1. The data for
Fc-CO-C6-ATP (the structure is shown in Scheme 1) was also in-
cluded as a reference. All the Fc-modified compounds show
a fully reversible one-electron wave with a peak separation
Figure 1. A) Cyclic voltammograms and B) square-wave voltammograms for
peptide-modified gold electrodes in a 1m NaClO
4
solution containing 5 mm
3
À/4À
of the [Fe(CN) ]
6
redox probe: a) bare gold electrode, b) after modifica-
tion with LPA, c) after immobilization of the peptide substrate, and d) after
blocking with 100 mm ethanolamine and following with 10 mm dodecane-
(
DE) in the range 50–80 mV versus Ag/AgCl. Fc-CO-Lys-Boc
shows a half-wave potential (E ) at approximately 574 mV and
1/2
thiol. Ag/AgCl/3m KCl mixture was used as reference electrode, a Pt was in-
À1
a peak current ratio (i /i ) of 1.1, which is similar to other fer-
cluded as auxiliary electrode. The scan rate was 100 mVs
.
pa pc
Chem. Eur. J. 2015, 21, 4988 – 4999
www.chemeurj.org
4990
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
tential in the SWV (Figure 1B) shifts from approximately 0.26 V
to about 0.58 V. The results indicate a lower electron-transfer
ability through the surface due to blockage of growing films
with the corresponding peptides and blocking reagent.
the magnitude of the redox signal of Fc is directly related to
the efficiency of the co-substrate. Consequently, Fc-CO-C6-ATP
is a better co-substrate than Fc-CO-Lys-ATP, due to the higher
redox signals observed in SWV as shown in Figure 2A.
Next, the kinase-catalyzed Fc phosphorylation was carried
out by using these peptide-modified gold electrodes, Fc-ATP
analogues, and their corresponding kinases in the presence of
kinase assay buffers. CVs and SWVs were recorded after the re-
action. The representative CVs and SWVs for Src-catalyzed reac-
tions are given in Figure 2. The CV data show that Fc-CO-C6-
and Fc-CO-Lys-phosphorylated peptide films display very simi-
In order to confirm that the current signal was generated
from the surface-attached Fc moiety rather than diffusive con-
tributions, CV measurements with different scan rates were
also applied. As shown in the Supporting Information (Fig-
ure S7), the linear dependence of the anodic and cathodic cur-
rent density versus the scan rate clearly illustrates the exis-
tence of surface chemically attached Fc groups.
o
lar formal redox potential, E , located at approximately 490 mV
Besides Src kinase, similar CV and SWV results were also ob-
tained by using CDK2 and CK2a, however, Fc-CO-C6-ATP did
not always perform better than Fc-CO-Lys-ATP as co-substrate.
The electrochemical results given in Figure 2C indicate that Fc-
CO-C6-ATP displays better performance in the case of Src and
CK2a, because greater current densities were observed, where-
as in the case of CDK2 nearly equal current densities were ob-
tained by using these two co-substrates. In general, the differ-
ent efficiencies between the Fc-ATP analogues in the kinase-
catalyzed phosphorylation reactions might be explained by
due to the similar structure of the Fc-CO-C6- and Fc-CO-Lys
moiety, and also indicate the presence of redox-active Fc
groups on the surface. Because the protein kinases transfer the
Fc phosphate groups provided by the Fc-ATP analogues in this
phosphorylation process to the peptide substrates, the Fc
labels on the surface can be electrochemically evaluated and
[16]
steric hindrance and electronic factors. Our previous work
has demonstrated that a long alkyl spacer with more than six
carbon atoms can effectively reduce the steric hindrance and
[
16a]
hence help to maintain the kinase activity with Src.
More-
over, in order to be a good co-substrate, additional hydrogen-
[16b]
bonding and electrostatic interactions should be avoided.
In this work, comparing the linker in Fc-CO-Lys-ATP with the
one in Fc-CO-C6-ATP, the lysine moiety is only one carbon
atom short, but the existence of an amide-linked alkyne
moiety possibly introduces extra conformational constraints
with respect to the active sites of the protein kinase. Further-
more, even though the alkyne moiety in Fc-CO-Lys-ATP has
[25]
a low polarity as other hydrocarbon functional groups, the
amide group that linked the ferrocenyl and alkyne groups to-
gether still brings in additional hydrogen-bonding and electro-
static interactions with the catalytic pocket of the kinase or the
incoming peptide substrate. As a result, the extra amide group
negatively affects the phosphoryl transfer process.
In our previous work, we found that with Src and CK2a the
current density generated by the resulting phosphorylated Fc
peptides significantly decreased with an increasing length and
polarity of the linker, whereas with CDK2 the current density
[16b]
increased.
Based on this observation it is likely that the in-
fluence of electronic factors is less pronounced with CDK2
than with Src and CK2a.
To further confirm the formation of the Fc-CO-Lys-phospho-
peptide through a kinase-catalyzed phosphorylation reaction
by using Fc-CO-Lys-ATP as co-substrate, XPS was used to prove
the existence of Fc on the surface. The XPS survey scans of the
peptide film after phosphorylation in the presence of CDK2 is
provided in the Supporting Information (Figure S8); these
reveal the presence of S, C, and N elements in the films. The
XPS spectrum obtained with fifty scans given in Figure S9 in
the Supporting Information shows that a peak at 711.9 eV was
observed only when kinase was used. The binding energy of
this peak indicates the presence of Fe 2p3/2 on the surface,
Figure 2. A) Square-wave voltammograms and B) cyclic voltammograms
showing the current densities of modified surfaces following the Src kinase-
catalyzed phosphorylation reaction in the presence of compounds Fc-CO-
Lys-ATP (dashed lines) and Fc-CO-C6-ATP (solid lines). C) Plot of the current
densities of surfaces following the phosphorylation reactions with CDK2,
CK2a, and Src in the presence of Fc-CO-Lys-ATP (black bars) and Fc-CO-C6-
ATP (white bars). Data points are the average of at least triplicate measure-
4
ments taken from SWVs. Measurements were taken in 1m NaClO aqueous
solution with Ag/AgCl/3m KCl as the reference electrode, a Pt wire as the
À1
auxiliary electrode, and a scan rate of 100 mVs
Chem. Eur. J. 2015, 21, 4988 – 4999
.
www.chemeurj.org
4991
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
which resulted from a Fc-labeled phosphoryl transfer and this
reactions using Fc-CO-C6-ATP and Fc-CO-Lys-ATP generated
visible fluorescence under irradiation, confirming the effective-
ness of Fc-CO-Lys-ATP in the phosphorylation reaction and the
compatibility of the Fc-CO-Lys fragment with the immunode-
tection system. As expected, the antiferrocene antibodies did
not bind to either the peptide or the phosphopeptide films
(treated with unmodified ATP), these samples demonstrating
background fluorescence only (Figures 3Ba) and d)).
[26]
observation is in good agreement with previous works. In
the control experiment, when Fc-CO-Lys-ATP was present and
CDK2 kinase was absent, no corresponding signals related to
the Fc group were observed as shown in Figure S9 in the Sup-
porting Information (dashed line).
We also tested the compatibility of Fc-CO-Lys-ATP with the
antiferrocene antibody and fluorescently labeled secondary an-
tibody for monitoring protein kinase activity on a surface. Ini-
tially, the electrochemical method was applied to detect the
Molecular modeling studies
performance of the Fc-Ab in the CDK2-catalyzed phosphoryla-
1
tion reaction. As shown in Figure S10 in the Supporting Infor-
To rationalize the different efficiency between Fc-CO-Lys-ATP
and Fc-CO-C6-ATP as co-substrate in kinase-catalyzed phos-
phorylation reactions, we modeled both ATP derivatives into
the active site of Src and CDK2/cyclin A. With the Src kinase
mation and Figure 3A, after incubation with Fc-Ab at 1:500 di-
1
(
PDB 1y57), the Fc groups in both Fc-ATP compounds were
well outside the ATP binding pocket and exposed to the sol-
vent as shown in Figure 4A, making them theoretically tolera-
ble. However, the additional alkyne group of Fc-CO-Lys-ATP in-
troduced potential steric hindrance between the acetylene
proton and the surrounding Src amino acids (Figure 4C). This
observation may explain the higher catalytic activity for Src
with un-functional Fc-CO-C6-ATP than with Fc-CO-Lys-ATP be-
cause the latter substrate induced steric hindrance at the bind-
ing pocket of the kinase.
Modeling of both Fc-ATP analogous into the CDK2/cyclin A
ATP binding site (PDB 4eoo) resulted in favorable overall bind-
ing modes. As shown in Figure 5A both compounds may pro-
trude into the cyclin A binding interface to the CDK2 protein,
but may not negatively affect the CDK2/cyclin A complexation.
The distance between the acetylene proton to the surrounding
amino acids is around 2 ꢂ as shown in Figure 5C, indicating
the more significant steric hindrance between Fc-CO-Lys-ATP
and the active site. However, with CDK2, both Fc-ATP com-
pounds display similar efficiency. To explain this, the peptide
substrate binding should also be taken into account. Consider-
Figure 3. A) Immuno-electrochemical detection of Fc phosphorylation of im-
mobilized peptides. Background-subtracted cyclic voltammograms of the Au
surfaces containing the Fc-CO-Lys-ATP-phosphorylated CDK2 substrate pep-
1
tide films before (dash lines) and after (solid lines) addition of Fc-Ab (1:500).
B) Fluorescence images of a) peptide-modified, b) Fc-CO-Lys-phosphopep-
tide-modified, c) Fc-CO-C6-phosphopeptide-modified, and d) phosphopep-
[27]
tide-modified Au surfaces following the addition of Fc-Ab
rescent Ab (1:100). Filter set 17 was used to generate the images. Colored
images are given in the Supporting Information.
1
(1:500) and fluo-
ing the CDK2 residues that bind the peptide substrate, only
50Arg is close to the Fc group but not to the alkyne moiety as
shown in Figure 5B. Due to peptide substrate binding away
from the Fc side, the substitution of Fc-ATP would not appear
to significantly influence the final phosphoryl transfer out-
come. In contrast, with Src the residues that bind the peptide
2
lution, the current signals decreased from approximately 2 and
À2
À2
0
.5 mAcm down to 0.4 and 0.1 mAcm for Fc-CO-C6-phos-
[28]
phopeptide and Fc-CO-Lys-phosphopeptide films, respectively,
substrate are 343–519. Among these, residues 384–402 are
critical. As shown in Figure 4B all residues surrounding the
alkyne moiety belong to this list of amino acids. As a result,
the introduction of this functional group may influence the
phosphorylation reaction more substantially. It is also notewor-
thy that all the above-described molecular modeling studies
are based on the native state, whereas our electrochemical
study is carried out on a surface with the environment of the
substrate known to influence the phosphoryl transfer pro-
due to the binding of Fc-Ab to the Fc groups on the surface.
1
This dramatic electrochemical signal-off response indicates that
compared to the C linker, the modified lysine linker does not
6
greatly interfere with the recognition between the Fc group
and the antiferrocene antibody, confirming the compatibility
and usefulness of Fc-CO-Lys-ATP in this detection strategy.
An indirect fluorescence imaging study was also carried out
by using antiferrocene antibodies to visualize the phosphory-
lated peptide films by using different ATP substrates. In this
study, a CDK2-catalyzed phosphorylation reaction was per-
formed on a peptide-modified Au surface in the presence of
Fc-CO-Lys-ATP, Fc-CO-C6-ATP, and unmodified ATP co-sub-
strates. The unphosphorylated peptide film was included as
a control. The fluorescence images (Figure 3B) indicated that
[10]
cess.
Scanning electrochemical microscopy (SECM)
As a complimentary method to the above-described electro-
chemical and optical investigations, SECM was used to investi-
Chem. Eur. J. 2015, 21, 4988 – 4999
www.chemeurj.org
4992
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 4. A) Molecular modeling of Fc-CO-Lys-ATP and Fc-CO-C6-ATP into
the active site of Src. B) Molecular modeling of Fc-CO-Lys-ATP into the active
site of Src showing key amino acids. C) Proximity of the acetylene proton to
surrounding amino acids of Src. Colored images are given in the Supporting
Information.
Figure 5. A) Molecular modeling of Fc-CO-Lys-ATP and Fc-CO-C6-ATP into
the active site of CDK2. B) Molecular modeling of Fc-CO-Lys-ATP into the
active site of CDK2 showing the key amino acids. C) Proximity of the acety-
lene proton to surrounding amino acids of CDK2. Colored images are given
in the Supporting Information.
gate the peptide films before and after the CDK2 kinase-cata-
lyzed phosphorylation reaction. To the best of our knowledge,
this is the first time that SECM has been used to monitor phos-
phorylation reactions. Instead of using a gold or gold-modified
electrode with a large surface area (several square millimeters),
a Pt microelectrode was employed for evaluating the electro-
chemical properties of the surfaces at smaller scale of approxi-
mately hundreds of square micrometers. The working principle
creased during the approach of the tip to the substrate. This is
expected, as the SAMs on the gold surface will be more insu-
lating and the layer minimizes the electron charge transfer be-
tween the gold substrate and the redox probe, thus the regen-
2
+
3+
eration of Fe from Fe by the substrate is compromised
2
+
and obstruction of the Fe
diffusion dominates the pro-
[
19d,29]
[29a,b]
of SECM has been well discussed before.
cess.
The electron charge-transfer process was further re-
The electrochemical property of the surface was dramatically
changed when the bare gold surface was modified with LPA
self-assembled monolayers (SAMs). A negative feedback was
observed for LPA SAMs, as shown in Figure 6. The current de-
duced by immobilization of the CDK2 peptide onto LPA. The
SECM data in Figure 6 show that the steady current observed
during the approach was even less for the surface after CDK2
peptide binding than before. The additional insulating layer of
Chem. Eur. J. 2015, 21, 4988 – 4999
www.chemeurj.org
4993
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ATP as co-substrate. In this case, Fc groups were at-
tached onto the peptide surface through phosphoryl
[9d]
transfer reactions.
The Fc layer seemed to have
a similar conductive ability to metal surfaces and
could efficiently distribute the charges, as shown in
Figure 6D, and promote the regeneration process,
which is consistent with other similar Fc monolayer
[30]
systems.
However, negative feedback was ob-
served for the Fc-CO-Lys-phosphopeptide as shown
in Figure 6C. In this case, although the charge-trans-
fer property of the Fc-CO-Lys-phosphopeptide film
was less efficient than of the Fc-CO-C6-phosphopep-
tide film (positive feedback), the charge-transfer
property of this film was still improved by the intro-
duced Fc groups compared to the unmodified pep-
tide film.
The antiferrocene antibody was also used in the
SECM studies to probe the existence of Fc on the
sample surfaces. The unmodified ATP-treated pep-
tide surface showed no changes before and after an-
tibody treatment indicating no non-specific effects
Figure 6. Examples of SECM approach curves obtained for three sets of samples using
different co-substrates: A) ATP, B) Fc-CO-C6-ATP, and C) Fc-CO-Lys-ATP. Approach curves
(
Figure 6A). When Fc-Ab1 was applied to surfaces
treated with Fc-CO-C6-ATP and Fc-CO-Lys-ATP, re-
spectively, the approach curves shown in Figures 6B
and C were seen to have a reduced steady-state cur-
rent, which is consistent with antibody immobiliza-
tion. In order to evaluate the result more quantita-
tively, simulation of approach curves with a known
for the following surfaces are presented in the figure: bare gold ( ); gold modified with
&
LPA (~), immobilized CDK2 peptides after blocking by 100 mm ethanolamine and follow-
ing with 10 mm dodecanethiol. After phosphorylation by using ATP in the presence of
("). After phosphorylation by using
Fc-CO-C6-ATP in the presence of CDK2 kinase (*) and followed by exposure to Fc-Ab
CDK2 kinase (^) and followed by exposure to Fc-Ab
1
1
(
~). After phosphorylation by using Fc-CO-Lys-ATP in the presence of CDK2 kinase (!)
and followed by exposure to Fc-Ab ). The continuous and dashed lines are the ap-
proach curves calculated by COMSOL Multiphysics simulation
(
&
1
[
19b,c]
by using known values
surface regeneration rate was conducted following
of dimensionless rate constants (L). The tip steady-state current was normalized against
the current determined at an infinite distance from the substrate. The normalized dis-
tance (L) is the ratio of tip/substrate separation (d) to the tip radius (a). D) Illustration of
SECM measurements: positive feedback generated from catalytic oxidation of [Fe(CN)
over ferrocenyl peptide.
[19c,31]
a
published procedure
and by using the
[
19b,c]
COMSOL Multiphysics software.
The experimen-
]4À
tal curves were compared with the simulated ones
to estimate the regeneration kinetics.
6
The rate constants plot in Figure 8 shows that the
unmodified gold surface had the highest regenera-
3
+
2+
À3
À1
peptide increases the film thickness, which makes the electron
transfer channels even longer. Three sets of phosphorylation
experiments were carried out on the peptide-immobilized sur-
faces by using unmodified ATP, Fc-CO-C6-ATP, and Fc-CO-Lys-
ATP as co-substrates under identical conditions and their corre-
sponding approach curves have been obtained as shown in
Figures 6A–C. By using unmodified ATP as co-substrate, a posi-
tive feedback was observed as shown in Figure 6A. It is likely
due to the reason that the CDK2 peptide, HHASPRK, has multi-
ple free amine groups (His, Arg, and Lys) that can react with
the LPA SAMs, resulting in a more closely packed neutral pep-
tide film. After introducing the negatively charged phosphate
groups into the Ser residue through kinase-catalyzed phos-
phorylation, defects in the film were created due to ionic re-
pulsion interaction. Thus, there would be more channels for
electron charge transfer and the current would increase as the
tip approaches the substrate. This hypothesis is in agreement
with the CV and impedance measurements. As shown in Figur-
es 7A and B, after phosphorylation, the current signal is recov-
ered and the film resistance decreased.
tion rate of Fe to Fe at (1.55Æ0.29)ꢁ10 cms . The rate
À5 À1
dropped dramatically to (7.63Æ2.41)ꢁ10 cms when the
gold was modified with LPA. A further decrease was seen after
immobilization of the CDK2 peptide and the regeneration rate
À5
À1
was decreased to (5.47Æ2.94)ꢁ10 cms . Phosphorylation of
the immobilized peptide with unmodified ATP, Fc-CO-C6-ATP,
and Fc-CO-Lys-ATP by CDK2 kinase, respectively, increased the
À4
À3
regeneration rates to (9.95Æ1.62)ꢁ10 , (1.39Æ0.28)ꢁ10 ,
À4
À1
and (1.83Æ0.06)ꢁ10 cms , respectively. The regeneration
rates for the Fc-CO-C6-phosphopeptide surfaces were the fast-
est of all the treated samples and were approximately 10%
slower than the bare gold. As discussed in the description for
Figure 6, phosphorylation with ATP possibly creates defects for
electron charge transfer, Fc groups transferred to the peptide
surfaces might also be able to promote the regeneration pro-
cess by reduction and oxidation of their Fe centers. The Fc-CO-
Lys-phosphopeptide surfaces had the lowest regeneration
rates. This is possibly due to the steric hindrance produced by
having a lysine group in the chain, thus the regeneration rates
for Fc-CO-Lys-phosphopeptide surfaces would rely mostly on
the Fc group only. The regeneration rates for phosphate-pep-
tide surfaces were not changed by the treatment with the anti-
Figure 6B illustrates that the positive feedback was also ob-
served for Fc-phosphorylated peptide film by using Fc-CO-C6-
Chem. Eur. J. 2015, 21, 4988 – 4999
www.chemeurj.org
4994
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
0
Figure 8. Rate constant, k , plots for the surfaces indicated in the figure.
These rate constants were measured by using the L values estimated by
the comparison of the experimental approach curves against the calculated
approach curves from COMSOL Multiphysics simulations, both examples of
which are shown in the Figure 6. No. 1: bare gold, 2: LPA, 3: peptide film,
4
: phosphate-peptide, 5: Fc-CO-C6-phosphopeptide, 6: Fc-CO-Lys-phospho-
peptide, 7: phosphate-peptide+antibody, 8: Fc-CO-C6-phosphopeptide+an-
tibody, 9: Fc-CO-Lys-phosphopeptide+antibody.
Click reaction on Fc-CO-Lys-phosphopeptide films
Figure 7. Surface characterization of the CDK2 peptide film in a 1m NaClO
4
3
À 4À
solution containing 5 mm [Fe(CN)
6
]
/
as redox probe. A) Cyclic voltammo-
The extra alkyne moiety incorporated into the Fc-CO-Lys-ATP
compounds allows further modification. To further determine
the utility of this functionalized ATP compound, a surface-
based click reaction was designed as shown in Figure 9. In this
reaction, CK2a peptide was first immobilized onto the Au sur-
face and then phosphorylated by using CK2a kinase with Fc-
CO-Lys-ATP as co-substrate. The phosphorylated film further re-
acted with Azide-fluor 488 through a click reaction catalyzed
grams and B) Nyquist plots before (dashed lines or ꢁ) and after (solid lines
or *) phosphorylation reaction by using CDK2 kinase with Fc-CO-C6-ATP.
Ag/AgCl was used as the reference electrode, a Pt wire was used as auxiliary
electrode, a scan rate of 100 mVs was used. The impedance spectra were
recorded in the frequency range of 100 kHz to 0.1 Hz with a AC amplitude
of 5 mV. Experimental data are shown as points and calculated results corre-
À1
spond to solid lines. The inset shows the equivalent circuit model: R
solution resistance, Q is the constant phase elements, RCT is the charge-trans-
fer resistance, and Z is the finite length Warburg impedance.
S
is the
W
I
by Cu. In this study, Fc-CO-C6-ATP, the un-functionalized Fc-
ATP compound, was also included as a control. The fluores-
cence images were collected as shown in Figure 9. The image
in entry 1 clearly shows that the film formed by using Fc-CO-
ferrocene antibody, as shown in Figure 8, No. 4 and 7. This is
consistent with the examples discussed earlier for the ap-
proach curves shown in the
Figure 6. The regeneration rates
for Fc-CO-C6-phosphopeptide
and Fc-CO-Lys-phosphopeptide
surfaces decreased to (1.13Æ
À3
0
.09)ꢁ10
and (3.32Æ0.94)ꢁ
À5
À1
1
0
cms , respectively, after
the binding by the antibody,
because the electron transfer
between Fc and the redox
probe is blocked by the attach-
ment of non-conductive Fc-Ab1.
SECM investigation not only
provides
a
complementary
method to electrochemical
characterizations, but it is also
a method to monitor the elec-
trochemical property at more
localized sample areas.
Figure 9. Fluorescence images of 1) Fc-CO-Lys-phosphopeptide-modified and 2) Fc-CO-C6-phosphopeptide-modi-
I
fied Au surfaces after a Cu -catalyzed click reaction with Azide-fluor488. Colored images are given in the Support-
ing Information.
Chem. Eur. J. 2015, 21, 4988 – 4999
www.chemeurj.org
4995
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Lys-ATP generated visible fluorescence under irradiation, con-
firming the attachment of the fluorescent fluor488 fragment.
Due to the lack of an alkyne moiety, the Fc-CO-C6-phospho-
peptide displayed only weak background fluorescence signals
as shown in entry 2.
without further purification. Goat anti-rabbit IgG (H+L) secondary
antibody, DyLight 488 conjugate was purchased from Thermo Sci-
entific, Canada.
Synthesis
Preparation of Fc-CO-Lys-Boc (2): Compound 1 (400 mg) was dis-
solved in MeOH/H O/THF (3 mL:3 mL:1 mL). Then LiOH (6 equiv)
2
was added to this solution. The reaction mixture was stirred in the
dark at RT for two days. Then THF was removed in vacuum and
water (15 mL) was added. The aqueous phase was washed with di-
chloromethane (3ꢁ10 mL) and then treated with 2m HCl till pH
ꢀ2 was reached. Dichloromethane (15 mL) was added to dissolve
the precipitate, which was further washed with brine (3ꢁ10 mL)
Conclusion
A new Fc-ATP compound containing a functional alkyne
moiety has been synthesized and characterized. A comparative
electrochemical study between this compound and related Fc-
CO-C6-ATP was carried out by using surface-bounded peptides
phosphorylated by Src, CDK2, and CK2a kinases. Obvious elec-
trochemical signals were detected, indicating the effectiveness
of Fc-CO-Lys-ATP as co-substrate in the reaction. These studies
also show that this compound can be applied to the Fc-Ab1/
and dried over MgSO . After solvent removal, the yellow solid (Fc-
4
CO-Lys(OH)-NHBoc) was used in the next step without further pu-
rification. EDC·HCl (1.38 mmol, 266 mg, 2.5 equiv) and HOBt
(
1.38 mmol, 213 mg, 2.5 equiv) were added to an anhydrous solu-
tion of Fc-CO-Lys(OH)-NHBoc (250 mg, 1 equiv) in dichloromethane
15 mL) on an ice bath. Then DIPEA (0.14 mL, 2.5 equiv) and prop-
(
Ab system, providing both electrochemical results and immu-
2
argylamine (1.1 mmol, 61 mg, 2 equiv) were added to this solution.
The reaction mixture was continuously stirred overnight under N2
atmosphere. After the reaction was completed, the solution was
nodetection. An azide-attached organic dye was successfully
linked to this Fc-CO-Lys-phosphate peptide film through click
chemistry. Thus, this compound has significant potential for
further application in the study of enzymatic post-translations
and we will evaluate the feasibility of “clickable” substrates to
replace the well-established immunoassay.
washed with saturated aqueous NaHCO solution (3ꢁ50 mL), 10%
3
critic acid (3ꢁ50 mL), and brine (3ꢁ50 mL). The organic layers
were dried over MgSO and the solvent was evaporated. The resi-
4
dues were purified by column with dichloromethane/MeOH (97:3)
1
to isolate the product as yellow solids (yield 75%). H NMR (DMSO,
2
98 K): d=8.32 (t, J=5.0 Hz, 1H), 7.68 (d, J=8.3 Hz, 1H), 6.78 (t,
J=5.5 Hz, 1H), 4.88 (s, 2H), 4.35 (s, 2H), 4.35 (t, J=2.3 Hz, 1H), 4.18
s, 5H), 3.87 (q, J=2.3 Hz, 2H), 3.09 (t, J=2.3 Hz, 1H), 2.94–2.86 (m,
Experimental Section
(
2
H), 1.71–1.63 (m, 2H), 1.42–1.22 (m, 4H), 1.35 ppm (s, 9H);
General methods
1
3
C NMR (CDCl , 298 K): d=172.08, 170.95, 155.95, 79.40, 78.85,
3
All synthesis reactions were carried out under air unless indicated
otherwise. N,N-Diisopropylethylamine, propargylamine, diethylami-
noethyl cellulose, adenosine 5’-triphosphate disodium salt, Azide-
fluor488, and N,N’-dicyclohexylcarbodiimide were obtained from
Sigma Aldrich and used as received. Dowex AG 50W-X8 was pur-
chased from Bio-Rad Laboratories (Ontario, Canada). H-Lys(Boc)-
OMe·HCl, hydroxybenzotriazole, and 1-ethyl-3-(3-dimethylamino-
propyl)carbodiimide were bought from AAPPTec LLC (KY, USA). Dry
dimethylformamide was purchased from Fisher Scientific. Dichloro-
74.87, 71.37, 70.60, 69.65, 68.41, 68.22, 53.34, 52.62, 40.02, 31.86,
29.37, 28.93, 28.31, 22.70 ppm; MS (ESI+): m/z calcd for
+
C
H
25
34FeN O
3
: 496.1899 [M+H ]; found: 496.1879.
4
Preparation of Fc-CO-Lys-NH2 (3): Neat TFA (4 mL) was used to
dissolve Fc-CO-Lys-Boc (720 mg, 1.44 mmol). After stirring the solu-
tion for one hour, dichloromethane (20 mL) was added to dilute
the solution before remove the solvent in vacuum. Three portions
of dichloromethane were added and evaporated to get rid of the
excess TFA. The residue was dissolved in dichloromethane (10 mL)
and TEA (5 mL) was added to convert the TFA salt to free amine
completely. After solvent removal, the mixture was used in the
next step without further purification.
methane was distilled by using CaH before use. Methanol was dis-
2
tilled from magnesium turnings with the presence of iodine. Ferro-
[
32]
[33]
cenecarboxylic acid, Fc-CO-Lys(OMe)-NHBoc (1), and adenosine
5
[9d]
’-[g-ferrocene] triphosphate (Fc-CO-C6-ATP) were prepared ac-
Preparation of Fc-CO-Lys-ATP (4): Adenosine 5’-triphosphate diso-
dium salt (100 mg, 0.18 mmol) was dissolved in 0.1m TEAB buffer
cording to the literature procedures assigned to each compound.
1
31
H and P NMR experiments were performed on a Bruker Advance
(
pH 7.5) (10 mL) and loaded on a column packed with a cation-ex-
change resin (AG 50W-X8), which has been pre-equilibrated with
.1m TEAB buffer. The desired fraction (monitored by UV light) was
5
00 MHz spectrometer. The initial geometry of Fc-CO-C6-ATP and
[34]
Fc-CO-Lys-ATP conjugates was energy-optimized by Gaussian.
0
The optimized co-substrate structures were superimposed onto
the binding site of the Src kinase (PBD 1y57) and CDK2/Cyclin A
kinase (PBD4eoo) and modeled by ArgusLab 4.0.1 (Mark A. Thomp-
son, Planaria Software LLC, Seattle, WA, USA, http://www.arguslab.-
com).
collected and evaporated in vacuum. The residue was co-evaporat-
ed with dry methanol (10 mL) for three times and dissolved in dry
DMF (1.8 mL) under argon. This ATP solution was further dried by
pre-activated molecular sieves over night. DCC (123 mg) was
added into the ATP solution and the mixture was stirred under Ar
for three hours at room temperature to form adenosine-5’-trimeta-
phosphate (ATMP). The ATMP solution was added to a mixture of
compound 3 (8 equiv) in dry MeOH (10 mL) and TEA (0.25 mL, pre-
dried by molecular sieves) under Ar. The mixture was stirred for
The Src peptide substrate, EGIYDVP, was obtained from BioBasic
Inc (Markham, Ontario). The CDK2 peptide substrate (HHASPRK),
the CDK2/cyclin A complex, and the Src kinase were purchased
from Cell Signaling (New England Biolabs Ltd., Pickering, Ontario).
The CK2a and the peptide substrate (RRRDDDSDDD) were pre-
pared in the laboratory of D. W. Litchfield (University of Western
Ontario).
one day, and poured into H O (20 mL). The solution was loaded on
2
a DEAE-cellulose column and washed with distilled H O to remove
2
excess ferrocene-amine. Then, a linear gradient of TEAB buffer
(0.1–0.5m) was carried out to give the desired fraction (yellow
band ꢀ0.3–0.4m TEAB), which was lyophilized to give a light
yellow powder. The product was further purified by HPLC (Varian
Serum enriched with polyclonal rabbit antiferrocene antibodies
(
Fc-Ab ) was produced at the YenZym Antibodies, LLC (CA, U.S.)
1
against the monosubstituted ferroceneamide-alkylamine and used
Chem. Eur. J. 2015, 21, 4988 – 4999
www.chemeurj.org 4996
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Modular reverse-phase Dynamax Macro-HPLC System C18
the kinase assay buffer and then Milli-Q water thoroughly in prior
to measurement.
À1
2
1.4 mmꢁ25 cm). The flow rate was 20 mLmin , the UV detector
was set to l=250 nm, and a 15 min linear gradient was composed
of A=TEAB (pH 7.5) and B=CH CN from 100 to 95% solvent A for
3
Click reaction on the phosphorylated peptide Au surface
2
min, from 95 to 80% solvent A for 2 min, from 80 to 75% solven-
t A for 1 min, at 75% solvent A for 5 min, from 75 to 100% solven-
t A for 2 min, and at 100% solvent A for 3 min. The product was
collected and dried as triethylammonium salt (yield 8%). HPLC:
The phosphorylated CK2a peptide-immobilized Au surfaces were
prepared according to the same procedure mentioned above. A
click-reaction solution containing 0.01% Azide-fluor488, 0.1% CuI,
and 1% DIPEA in DMF was applied to the Au surface overnight.
Then the Au surface was rinsed thoroughly with DMF, ethanol, and
1
t =8.3 min; H NMR (D O, 298 K): d=8.42 (s, 1H), 8.09 (s, 1H), 5.99
R
2
(
2
d, J=6.0 Hz, 1H), 4.70–4.63 (m, 3H), 4.44 (t, J=2.5 Hz, 1H), 4.37 (s,
H), 4.27 (s, 1H), 4.18–4.09 (m, 9H), 3.87 (d, J=4.7 Hz, 2H), 2.78 (q,
water, and blown-dried with a stream of N in prior to the mea-
2
31
J=7.2 Hz, 2H), 1.67–1.57 (m, 2H), 1.39–1.01 ppm (m, 4H); P NMR
surement.
(
D O, 298 K): À0.96 (d, J=21.4 Hz, g-P), À11.51 (d, J=18.3 Hz, a-P),
2
À23.01 ppm (t, J=21.4 Hz, b-P); MS (MALDI-TOF): m/z calcd for
Electrochemical experiments
+
C H FeN O P :885.1221 [M+H] ; found: 885.2879.
3
0
40
8
14 3
All electrochemical measurements were carried out by using a CH
Instrument potentiostat 660B (Austin, TX). An electrochemical cell
with a three-electrode configuration was used with a peptide-
modified gold electrode as the working electrode, a platinum wire
as the counter electrode, and Ag/AgCl in 3m KCl as the reference
electrode, which was connected by using a salt bridge filled with
Substrate cleaning and preparation
All electrochemical experiments were performed by using Au disk
electrodes (2 mm diameter, CH Instruments, Inc.). The fluorescence
experiments and SECM measurements were performed by using
Au silicon sputtering wafers, which were prepared by electron-
beam deposition of 200 nm thickness of Au on Si wafer with
Agar/3m KNO . The cyclic voltammetry measurements for charac-
3
terizing the peptide-modified electrodes were performed in 5 mm/
5 mm K [Fe(CN) ]/K [Fe(CN) ] solution with 1m NaClO as the sup-
2
0 nm Ti as the adhesion layer (Western University’s Nanofabrica-
3
6
4
6
4
tion Facility). The cleaning of the Au wafers was achieved by etch-
ing in piranha solution twenty seconds. After rinsing with Milli-Q
water, the Au wafers were further sonicated in Milli-Q water and
porting electrolyte. Electrochemical impedance spectroscopy (EIS)
experiments were conducted in the frequency range of 100 kHz to
0.1 Hz with an AC amplitude of 5 mV. The ZSimpWin 2.0 software
(Princeton Applied Research electrochemical software, http://
www.princetonappliedresearch.com/Our-Products/Electrochemical-
Software/ZSimpWin.aspx) was used to evaluate the EIS data by fit-
ting modeling curves. All impedance spectra are represented as
ethanol for 10 min each, and blown-dried by N . The Au electrodes
2
were also cleaned by piranha solution for 2 min. Then 0.3 and 0.05
micron aluminum powder were used for polishing. After rinsing
with Milli-Q water and sonication with Milli-Q water and ethanol
(
10 min each), the electrodes were electrochemically cleaned in
Nyquist plots with real impedance (Z ) and the imaginary impe-
re
0
.5m KOH by scanning between À2.0 and 0.1 V (vs. Ag/AgCl) at
dance (Z ) as the x axis and y axis, respectively. Square wave vol-
im
À1
a scan rate of 0.1 Vs and followed by scanning in 0.5m H SO in
the range of 0 to 1.5 V until a stable gold oxidation peak at 1.1 V
was obtained.
tammetry (SWV) was performed at a pulse amplitude of 25 mV
under the same solution condition to EIS. CV and SWV measure-
ments for the phosphorylated peptide-modified electrodes were
2
4
To immobilize the peptide substrates on the Au wafer and elec-
trode, the Au surface/electrode was first incubated in 2 mm N-hy-
droxysuccinimide lipoic acid active ester (LPA) ethanolic solution at
performed in 1m NaClO as the supporting electrolyte in the po-
4
tential range of 0.2 to 0.6 V. At least five measurements were per-
formed for all experiments.
4
8C for 48 h. Then, the surface was washed with ethanol and incu-
bated with 0.1 mm peptide solution for 24 h at 48C. The Au surface
was removed from incubation, washed with Milli-Q water thor-
oughly and further blocked with incubation in 100 mm ethanolic
ethanolamine solution for one hour, which was followed with im-
mersion in 10 mm dodecanethiol solution 20 min to back-fill the
exposed spots of the surface.
Immunoarray fluorescence imaging
After the surface phosphorylation reaction, the Au wafer was im-
mersed in 10% bovine serum albumin (BSA) solution for one hour
at room temperature to further eliminate the nonspecific binding
of the antibodies. Then the substrates were rinsed in Tris buffer
saline Tween20 (TBST) and incubated with a primary Fc-Ab solu-
1
tion (1:500 dilution in 0.1% BSA/TBST) for one hour at room tem-
perature. The Au substrate was then rinsed with TBST and subse-
quently incubated with secondary antibody (1:100 dilution in 0.1%
BSA/TBST) for 45 min at room temperature. The substrates were
then rinsed twice with TBST and twice with Tris buffer saline (TBS)
and imaged by using a Zeiss AxioPlan2 imaging fluorescent micro-
scope (Zeiss) with filter set 17.
Phosphorylation reaction on the Au surface
The kinase-catalyzed phosphorylation reactions were performed in
the kinase assay buffer. The Src kinase assay buffer consisted of
5
mm 3-morpholinopropanesulfonic acid (MOPS) (pH 7.5), 2.5 mm
b-glycerophosphate, 1 mm ethylene glycol tetraacetic acid (EGTA),
0
4
6
.4 mm ethylenediaminetetraacetic acid (EDTA), 2.5 mm MnCl , and
2
mm MgCl . The CDK2/cyclin A complex assays were performed in
2
0 mm 4-(2-hydroxyethyl)-1-piperazineethansulfonic acid (HEPES)
X-ray photoelectron spectroscopy measurement (XPS)
À1
(
pH 7.5), 3 mm MnCl , 3 mm MgCl , 0.05 mgmL
polyethylene
2
2
glycol (PEG) 20000 and 3 mm sodium orthovanadate. The CK2a
assay buffer consisted of 50 mm Tris HCl (pH 7.5), 10 mm MgCl2,
The Au substrates were cleaned and the CDK2 peptide substrate
(HHASPRK) was immobilized onto the Au surface by using the
same procedure as for the preparation of the modified gold elec-
trode described above. The substrate was subsequently Fc-phos-
phorylated in the presence of the CDK2/cyclin A complex and Fc-
CO-Lys-ATP. Prior to analysis, the substrate was thoroughly rinsed
with kinase assay buffer and water, and dried with a stream of N2.
1
50 mm NaCl. The peptide-modified surfaces were incubated in
À1
the presence of protein kinase (160 ngmL
5
for CK2a, and
À1
mgmL for Src and CDK2) and Fc-CO-C6-ATP or Fc-CO-Lys-ATP or
ATP (350 mm). After 8 h of incubation at 378C in a heating block
VWR Scientific, USA), the gold substrates were washed by using
(
Chem. Eur. J. 2015, 21, 4988 – 4999
www.chemeurj.org
4997
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
The XPS analyses were performed with an AlKa X-ray source
[6] a) Z. Wang, J. Lee, A. R. Cossins, M. Brust, Anal. Chem. 2005, 77, 5770–
À10
5774; b) Z. Wang, R. Lꢃvy, D. G. Fernig, M. Brust, J. Am. Chem. Soc. 2006,
(
15 mA, 14 kV). The operating pressure was 1ꢁ10 Torr. The XPS
1
28, 2214–2215.
spectra were obtained with a pass energy of 200 eV for the survey
scan and 100 eV for the Fe 2p region. The binding energies were
calibrated by using Au 4f7/2 at 84.0 eV.
[
[
[
7] E. T. Lund, R. McKenna, D. B. Evans, S. K. Sharma, W. R. Mathews, J. Neu-
rochem. 2001, 76, 1221–1232.
8] Y.-P. Kim, E. Oh, Y.-H. Oh, D. W. Moon, T. G. Lee, H.-S. Kim, Angew. Chem.
Int. Ed. 2007, 46, 6816–6819; Angew. Chem. 2007, 119, 6940–6943.
9] a) S. Suwal, M. K. H. Pflum, Angew. Chem. Int. Ed. 2010, 49, 1627–1630;
Angew. Chem. 2010, 122, 1671–1674; b) K. D. Green, M. K. H. Pflum,
ChemBioChem 2009, 10, 234–237; c) K. D. Green, M. K. H. Pflum, J. Am.
Chem. Soc. 2007, 129, 10–11; d) H. Song, K. Kerman, H.-B. Kraatz, Chem.
Commun. 2008, 502–504.
Scanning electrochemical microscopy (SECM) measurements
SECM experiments were carried out with a CHI-900b (CH Instru-
ments, Austin, TX) at room temperature in an electrochemical cell
by using a three-electrode configuration. A Pt wire, a Ag/AgCl/
[
[
[
10] S. Suwal, C. Senevirathne, S. Garre, M. K. H. Pflum, Bioconjugate Chem.
012, 23, 2386–2391.
11] S. Marti c´ , M. Labib, P. O. Shipman, H.-B. Kraatz, Dalton Trans. 2011, 40,
264–7290.
3
.0m KCl electrode, and a Pt SECM tip were fitted in as the counter
2
electrode, reference electrode, and working electrode, respectively.
Modified Au/Si substrates were mounted in the cell and studied
during the experiment without bias on the substrates. The home-
made Pt tip was fabricated by sealing a Pt wire with a diameter of
7
12] a) S. Iijima, F. Mizutani, S. Yabuki, Y. Tanaka, M. Asai, T. Katsura, S.
Hosaka, M. Ibonai, Anal. Chim. Acta 1993, 281, 483–487; b) I. Suzuki, K.
Murakami, J.-i. Anzai, Mater. Sci. Eng., C 2001, 17, 149–154; c) K. A. Mah-
moud, H.-B. Kraatz, J. Inorg. Organomet. Polym. Mater. 2008, 18, 69–80;
d) G.-Y. Qing, T.-L. Sun, F. Wang, Y.-B. He, X. Yang, Eur. J. Org. Chem.
2
5 mm into a glass capillary and then the microelectrode was pol-
ished to an RG (ratio of total tip radius to electrode radius) of ap-
proximately five. The tip was electrochemically cleaned prior to
each experiment by performing cyclic voltammetry scans in 0.5m
H SO between 0 and 1.4 V for hundred cycles at a scan rate of
2
009, 841–849; e) Y. Xiao, M. Petryk, P. M. Diakowski, H.-B. Kraatz, Proc.
SPIE 2009, 7304, 73040R; f) S. Mondal, S. Ghosh, S. Verma, Tetrahedron
Lett. 2010, 51, 856–859; g) M. C. Gimeno, H. Goitia, A. Laguna, M. E.
Luque, M. D. Villacampa, C. Sepulveda, M. Meireles, J. Inorg. Biochem.
2
4
À1
0
.5 Vs . The SECM experiments were conducted in the electrolyte,
which contains 2 mm K [Fe(CN) ] aqueous solution as the redox
4
6
2
011, 105, 1373–1382; h) M. J. Sheehy, J. F. Gallagher, M. Yamashita, Y.
probe and 50 mm NaClO as the supporting electrolyte. A steady-
4
Ida, J. White Colangelo, J. Johnson, R. Orlando, P. T. M. Kenny, J. Organo-
met. Chem. 2004, 689, 1511–1520; i) A. J. Corry, A. Goel, P. T. M. Kenny,
Inorg. Chim. Acta 2012, 384, 293–301; j) R. Afrasiabi, H.-B. Kraatz, Chem.
Eur. J. 2013, 19, 15862–15871; k) A. G. Harry, W. E. Butler, J. C. Manton,
M. T. Pryce, N. O’Donovan, J. Crown, D. K. Rai, P. T. M. Kenny, J. Organo-
met. Chem. 2014, 757, 28–35.
state current was obtained for each measurement by holding the
Pt tip, which was immersed in the electrolyte, at a constant height
above the substrate for 300 seconds at a potential of 0.5 V before
making the approach. The same potential was applied for making
all approach curves measurements. The experimental approach
curves were normalized to the steady-state current at an infinite
distance before fitting against the theoretical curves simulated by
[
13] a) T. Hirao, J. Organomet. Chem. 2009, 694, 806–811; b) T. Moriuchi, T.
Nagai, T. Hirao, Org. Lett. 2005, 7, 5265–5268.
[14] D. R. van Staveren, N. Metzler-Nolte, Chem. Rev. 2004, 104, 5931–5985.
15] G. Jaouen, N. Metzler-Nolte, Topics in Organomet. Chem., Vol. 32, Spring-
er, Berlin, 2010, p. 32.
16] a) S. Marti c´ , M. Labib, D. Freeman, H.-B. Kraatz, Chem. Eur. J. 2011, 17,
[19b,c]
using the COMSOL Multiphysics software.
The surface regener-
[
[
ate kinetics was estimated for each set of samples.
6
744–6752; b) S. Marti c´ , M. K. Rains, D. Freeman, H.-B. Kraatz, Bioconju-
gate Chem. 2011, 22, 1663–1672.
17] S. Marti c´ , M. Gabriel, J. P. Turowec, D. W. Litchfield, H.-B. Kraatz, J. Am.
Chem. Soc. 2012, 134, 17036–17045.
18] H. C. Kolb, M. G. Finn, K. B. Sharpless, Angew. Chem. Int. Ed. 2001, 40,
2004–2021; Angew. Chem. 2001, 113, 2056–2075.
[19] a) S. Bergner, P. Vatsyayan, F.-M. Matysik, Anal. Chim. Acta 2013, 775, 1–
Acknowledgements
[
[
We are grateful to Rouzbeh Afrasiabi for his help with synthe-
sis. In the initial stage, he helped to synthesize the Fc-CO-Lys-
Boc compound. We appreciate funding from NSERCand from
the University of Toronto Scarborough.
1
4
3; b) M. N. Alam, M. H. Shamsi, H.-B. Kraatz, Analyst 2012, 137, 4220–
225; c) P. M. Diakowski, H.-B. Kraatz, Chem. Commun. 2009, 1189–1191;
d) J. Kwak, A. J. Bard, Anal. Chem. 1989, 61, 1221–1227.
20] R. K. Suto, M. A. Whalen, B. R. Bender, R. G. Finke, Nucleosides Nucleo-
tides 1998, 17, 1453–1471.
[
Keywords: click chemistry
phosphorylation · protein kinases
·
ferrocenes
·
peptides
·
[
[
[
[
[
21] E. Laviron, J. Electroanal. Chem. 1979, 101, 19–28.
22] J. Slack-Davis, J. O. DaSilva, S. J. Parsons, Cancer Cell 2010, 17, 527–529.
23] D. O. Morgan, Nature 1995, 374, 131–134.
24] D. W. Litchfield, Biochem. J. 2003, 369, 1–15.
25] P. Y. Bruice, Organic Chemistry, 5th ed., Pearson Education, Upper Saddle
River, New Jersey, 2007.
[1] a) G. Burnett, E. P. Kennedy, J. Biol. Chem. 1954, 211, 969–980; b) G.
Manning, D. B. Whyte, R. Martinez, T. Hunter, S. Sudarsanam, Science
2002, 298, 1912–1934; c) for the use of a Fc-labeled nucleobases by
terminal deoxynucleotidyl transferase-mediated 3’-base extension, see:
A. Anne, C. Bonnaudat, C. Demaille, K. Wang, J. Am. Chem. Soc. 2007,
[
26] a) Y. Li, R. Afrasiabi, F. Fathi, N. Wang, C. Xiang, R. Love, Z. She, H.-B.
Kraatz, Biosens. Bioelectron. 2014, 58, 193–199; b) S. Marti c´ , M. Labib,
H.-B. Kraatz, Analyst 2011, 136, 107–112.
129, 2734–2735.
[
[
2] a) P. Cohen, Nat. Rev. Drug Discovery 2002, 1, 309–315; b) M. Flajolet, G.
He, M. Heiman, A. Lin, A. C. Nairn, P. Greengard, Proc. Natl. Acad. Sci.
USA 2007, 104, 4159–4164.
3] a) Z. A. Knight, H. Lin, K. M. Shokat, Nat. Rev. Cancer 2010, 10, 130–137;
b) J. Zhang, P. L. Yang, N. S. Gray, Nat. Rev. Cancer 2009, 9, 28–39;
c) M. E. Anderson, L. S. Higgins, H. Schulman, Nat. Clin. Pract. Cardiovasc.
Med. 2006, 3, 437–445.
[
27] A. A. Russo, P. D. Jeffrey, A. K. Patten, J. Massague, N. P. Pavletich, Nature
1996, 382, 325–331.
[28] F. Sicheri, I. Moarefi, J. Kuriyan, Nature 1997, 385, 602–609.
[29] a) P. Sun, F. O. Laforge, M. V. Mirkin, Phys. Chem. Chem. Phys. 2007, 9,
802–823; b) J. L. Amphlett, G. Denuault, J. Phys. Chem. B 1998, 102,
9946–9951; c) M. V. Mirkin, F.-R. F. Fan, A. J. Bard, J. Electroanal. Chem.
1992, 328, 47–62.
[30] F. Hauquier, J. Ghilane, B. Fabre, P. Hapiot, J. Am. Chem. Soc. 2008, 130,
2748–2749.
[31] P. M. Diakowski, H.-B. Kraatz, Chem. Commun. 2011, 47, 1431–1433.
[32] P. C. Reeves, Organic Syntheses, Wiley, New York, 2003.
[
[
4] D. M. Olive, Expert Rev. Proteomics 2004, 1, 327–341.
5] a) S. Balakrishnan, N. J. Zondlo, J. Am. Chem. Soc. 2006, 128, 5590–
5
2
5
591; b) K. Kerman, M. Chikae, S. Yamamura, E. Tamiya, Anal. Chim. Acta
007, 588, 26–33; c) K. Kerman, H.-B. Kraatz, Chem. Commun. 2007,
019–5021.
Chem. Eur. J. 2015, 21, 4988 – 4999
www.chemeurj.org
4998
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[
[
33] P. M. Diakowski, Y. Xiao, M. W. P. Petryk, H.-B. Kraatz, Anal. Chem. 2010,
2, 3191–3197.
Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O.
Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V.
Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu,
A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M.
A. Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill,
B. Johnson, W. Chen, M. W. Wong, C. Gonzalez, and J. A. Pople, Gaussi-
an, Inc., Wallingford CT, 2004.
8
34] R. C. Gaussian 03, Revision C. 02 ed., M. J. Frisch, G. W. Trucks, H. B.
Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomery,
Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J.
Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Pe-
tersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasega-
wa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J.
E. Knox, H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R.
Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli,
J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J.
Received: October 2, 2014
Published online on February 9, 2015
Chem. Eur. J. 2015, 21, 4988 – 4999
www.chemeurj.org
4999
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim