Synthesis and evaluation of phosphopeptides containing iminodiacetate groups as binding ligands of the Src SH2 domain

Article abstract of DOI:10.1016/j.bioorg.2009.05.003

Phosphopeptide pTyr-Glu-Glu-Ile (pYEEI) has been introduced as an optimal Src SH2 domain ligand. Peptides, Ac-K(IDA)pYEEIEK(IDA) (1), Ac-KpYEEIEK (2), Ac-K(IDA)pYEEIEK (3), and Ac-KpYEEIEK(IDA) (4), containing 0-2 iminodiacetate (IDA) groups at the N- and

Full text of DOI:10.1016/j.bioorg.2009.05.003

Bioorganic Chemistry 37 (2009) 133–142
Contents lists available at ScienceDirect
Bioorganic Chemistry
journal homepage: www.elsevier.com/locate/bioorg
Synthesis and evaluation of phosphopeptides containing iminodiacetate groups
as binding ligands of the Src SH2 domain
Guofeng Ye a, Aaron D. Schuler b, Yousef Ahmadibeni a, Joel R. Morgan b, Absar Faruqui b, Kezhen Huang c,
Gongqin Sun c, John A. Zebala b,d, Keykavous Parang a,
*
a Department of Biomedical and Pharmaceutical Sciences, University of Rhode Island, Kingston, RI 02881, USA
b Syntrix Biosystems, Inc., Auburn, WA 98001, USA
c Department of Cell and Molecular Biology, University of Rhode Island, Kingston, RI 02881, USA
d Department of Laboratory Medicine, University of Washington, Seattle, WA 98105, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 3 April 2009
Available online 21 May 2009
Phosphopeptide pTyr-Glu-Glu-Ile (pYEEI) has been introduced as an optimal Src SH2 domain ligand. Pep-
tides, Ac-K(IDA)pYEEIEK(IDA) (1), Ac-KpYEEIEK (2), Ac-K(IDA)pYEEIEK (3), and Ac-KpYEEIEK(IDA) (4),
containing 0–2 iminodiacetate (IDA) groups at the N- and C-terminal lysine residues were synthesized
and evaluated as the Src SH2 domain binding ligands. Fluorescence polarization assays showed that pep-
Keywords:
tide 1 had a higher binding affinity (Kd = 0.6
(Kd = 1.7 M), an optimal Src SH2 domain ligand, and peptides 2–4 (Kd = 2.9–52.7
ity of peptide 1 to the SH2 domain was reduced by more than 2-fold (Kd = 1.6
M) upon addition of Ni2+
(300
M), possibly due to modest structural effect of Ni2+ on the protein as shown by circular dichroism
experimental results. The binding affinity of 1 was restored in the presence of EDTA (300 M)
(Kd = 0.79 M). These studies suggest that peptides containing IDA groups may be used for designing
novel SH2 domain binding ligands.
lM) to the Src SH2 domain when compared with Ac-pYEEI
Phosphopeptides
Metal chelation
Iminodiacetate group
SH2 domain
UV titration
Fluorescence polarization
Circular dichroism
l
l
M). The binding affin-
l
l
l
l
Ó 2009 Elsevier Inc. All rights reserved.
1. Introduction
8]. The Src SH2 domain preferentially binds peptides with the se-
quence pTyr-Glu-Glu-Ile (pYEEI) with high affinity [9].
Src, one of the first studied non-receptor tyrosine kinases, has
been implicated in the genesis and progression of multiple types
of human diseases including colon, breast, lung cancers, osteoporo-
sis, and inflammation-mediated bone loss [1,2]. Src tyrosine kinase
contains three major domains, in the order from N- to C-terminal,
two regulatory Src homology (SH) domains (SH3 and SH2 do-
mains), and a kinase domain. The Src SH2 domain is a relatively
small protein module of approximately 100 amino acids and has
extraordinary ability to recognize specifically sequences contain-
ing phosphotyrosine residue (pTyr), thereby facilitating phosphor-
ylation-dependent protein–protein interactions that result in
signal transduction process [3]. Short peptide inhibitors capable
of disrupting these interactions have been designed as useful tools
in the mechanistic studies of the Src SH2 domain interactions and
as potential inhibitors for further pharmaceutical development [4–
The use of conformationally constrained peptides has been
shown to be an effective strategy in developing therapeutic pep-
tidic and peptidomimetic agents [10,11]. Constrained peptides
are usually more stable towards enzymatic degradations. We pre-
viously reported the synthesis and evaluation of a series of cova-
lently bonded conformationally constrained peptide analogues
based on the optimal sequence pYEEI as inhibitors of the Src SH2
domain [8]. A number of conformationally constrained peptides
showed at least 100-fold increase in the binding affinities to the
Src SH2 domain relative to pYEEI and the corresponding linear pep-
tides, respectively. The enhancement in binding affinities was
accomplished solely through covalent bonding of N-terminal and
side chain functional groups, formation of a favorable conforma-
tion, reducing unfavorable entropic effects, and/or creating novel
bonding interaction between the peptides and the protein. How-
ever, the conformational changes were irreversible; the synthesis
was cumbersome and challenging; and the preparation of the cor-
responding linear peptides was necessary for comparative studies.
To minimize one or more of the problems associated with
covalent bonding described above, we attempted to introduce
the conformational constraints through non-covalent forces by
incorporating groups capable of forming of intramolecular hydro-
gen bindings or chelating to metals.
Abbreviations: IDA, iminodiacetate; F-GpYEEI, fluorescein-labeled GpYEEI; FP,
fluorescence polarization; CD, circular dichroism.
* Corresponding author. Address: Department of Biomedical and Pharmaceutical
Sciences, College of Pharmacy, University of Rhode Island, 41 Lower College Road,
Kingston, RI 02881, USA. Fax: +1 401 874 5787.
E-mail address: kparang@uri.edu (K. Parang).
0045-2068/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved.
doi:10.1016/j.bioorg.2009.05.003

134
G. Ye et al. / Bioorganic Chemistry 37 (2009) 133–142
Metal-chelating groups have been used in the areas of protein
was structured to incorporate two iminodiacetate groups (IDA), a
common metal-chelating ligand, at the N- and C-terminal lysine
residues, flanking the pYEEI motif. The peptide was expected to
achieve a constrained conformation through potential intramolec-
ular hydrogen bindings or chelating of the IDA groups with metal
ions. The binding affinity of the ligand toward the Src SH2 domain
was determined in the absence or presence of Ni2+ and/or a metal
scavenger. The free amino groups in lysine residues are also strong
hydrogen bond donors and metal-chelating groups. Three other
peptides (Fig. 1) containing 0 or 1 IDA group at one of the N- or
C-terminal lysine residues, Ac-KpYEEIEK (2), Ac K(IDA)pYEEIEK
(3), and Ac-KpYEEIEK(IDA) (4), were also synthesized and evalu-
ated for their binding affinities toward the Src SH2 domains.
purifications [12,13], peptide/protein engineering [14,15], and
designing of nucleic acid probes targeting specific DNA molecules
[16].
Hochuli and coworkers reported a novel nitrilotriacetic acid
quadridentate chelate resin for metal chelate affinity chromatogra-
phy. It was found that when charged with Ni2+, the resin was able
to bind peptides and proteins containing neighboring histidine res-
idues [12].
Hopkins group reported that transition metal could induce a
conformational change in the
a-helical peptides bearing IDA
groups at the i and i + 3 or 4 positions [17]. The cooperative binding
of the two IDA groups to one transition metal could cross-link the
peptide intramolecularly and stabilize the secondary structure. The
non-natural amino acid Fmoc-Lys(IDA)-OH was used in the con-
ventional Fmoc peptide synthesis and introduced into protein–
2. Materials and methods
peptide interaction studies of the S-peptide fragment (
of ribonuclease S complex [18].
a-helix rich)
2.1. Peptide synthesis
In a recent report from Syntrix Biosystems, Inc., probes where
the N- and C-termini of a PNA oligomer are modified with a pair
of metal-chelating ligands (IDA or nitrilotriacetic acid, NTA) were
described, and it was found that the coordination can force the
PNA oligomer into a macrocyclic configuration [16]. The intramo-
lecular chelate dissociates only when the PNA hybridizes to its per-
fectly matched DNA target.
As part of our ongoing studies of interactions of conformation-
ally constrained phosphopeptides with the Src SH2 domain [8],
herein we report the interactions of phosphopeptides containing
metal-chelating groups with the Src SH2 domain in the presence
and absence of metals. Peptide Ac-K(IDA)pYEEIEK(IDA) (1, Fig. 1)
All reactions were carried out in Bio-Rad polypropylene col-
umns by shaking and mixing, using a VWR rotator, at room tem-
perature. All reagents were purchased from Novabiochem (San
Diego, CA), unless otherwise noted.
The peptides (Ac-K(IDA)pYEEIEK(IDA), Ac-KpYEEIEK, Ac-K(IDA)-
pYEEIEK, Ac-KpYEEIEK(IDA), and Fluorescein-K(IDA)pYEEIEK(IDA))
were synthesized on a Rink amide NovaGel resin employing N-(9-
fluorenyl)methoxycarbonyl (Fmoc)-based chemistry and Fmoc-L-
amino acid building blocks. The synthesis of Fmoc-Lys(IDA)-OH
has been previously described [19]. The preparation of peptides
containing the IDA moiety was carried out according to the
Fig. 1. Chemical structures of peptides 1–4 containing 0–2 IDA groups and fluorescein-labelled peptide derivative 5.

G. Ye et al. / Bioorganic Chemistry 37 (2009) 133–142
135
previously described procedure [15] using Fmoc-Lys(IDA(OtBu))-
2.2. UV titration assay
OH as the building block. HBTU and DIPEA (TCI America, Portland,
OR) in N,N-dimethylformamide (DMF) (VWR, West Chester, PA)
were used as coupling and activating reagents, respectively. Fmoc
deprotection at each step was carried out using piperidine (EMD,
La Jolla, CA) in DMF (20%). After the final coupling, the peptides
were N-terminally acetylated with acetic anhydride (Sigma–Al-
drich, Milwaukee, WI), unless otherwise noted. Assembled pep-
tides were cleaved from the resin with TFA/water/TIS (Sigma–
Aldrich) (95:2.5:2.5) for 3 h, hydrolyzed for 19 h after addition of
10% water as described previously (Novabiochem Catalogue
2006/2007, 3.32), dried in vacuum, dissolved in water, and lyoph-
ilized on a Labconco FreeZone 6 (Kansas City, MO). The crude pep-
tides were purified by HPLC (Waters Delta Prep 4000, Millipore,
All spectra were collected on a Beckman Coulter DU-800 UV–
Vis spectrophotometer with a scan rate of 120 nm/s, a slit width
of 2 mm, and using 1 cm UV–Cuvette semi micro cells at pH = 8.5
from 210 to 310 nm. Difference titration experiments were per-
formed by subtracting the signal of a cuvette containing only buf-
fer solution from the signal of a similar solution containing 10 lM
NiCl2 after adding equivalent aliquots of the peptide solution to
both cuvettes. The extinction coefficients of the complexes were
calculated from the initial slopes of the titration curves and were
based on the total metal concentration. In another set of experi-
ments, the Kd values were determined by monitoring the changes
at 225 nm absorbance of solutions of 10 lM peptide upon addition
Billerica, MA) on a Phenonemex Gemini 10
lm C18 column (Tor-
of aliquots of NiCl2 up to a final concentration of 10 lM.
rance, CA) by elution at 20 mL/min using a gradient of 5–95% ace-
tonitrile (0.1% formic acid) and water (0.1% formic acid) over
30 min. The peptides had a purity of >95% determined by HPLC
as shown by analytical HPLC using a similar gradient system and
a flow rate of 1 mL/min. The collected fractions were analyzed by
LC–MS (Shimadzu LC-10; Waters Micromass Quattro II), and only
fractions containing the pure peptide were pooled and lyophilized.
All peptides were characterized by LC–MS or SELDI-TOF mass
spectrometer.
2.3. Expression and purification of the Src SH2 domain
Bacterial DH5a cells containing recombinant plasmid pGEX-
Src-SH2 domain were used to inoculate 500 ml of Luria-Bertani
(LB) culture medium. After 10–12 h culture at 37 °C, the bacterial
culture (OD600nm around 1.0) was cooled down to 25 °C and diluted
by another 500 mL of LB medium. Then 0.4 mM isopropyl b-D-isog-
alactopyranoside (IPTG) was added to induce the production of the
fusion protein for 7 h at 25 °C. After induction, the bacterial cells
were harvested by centrifugation at 5000 rpm for 5 min at 4 °C.
Then bacterial pellets were resuspended in 50 mL 1X PBS and son-
icated for 15 s. Resupension and sonication were repeated 5 times.
All cell debris, including insoluble proteins, were removed by cen-
trifugation at 20,000 rpm for 30 min. The supernatant was mixed
and incubated with 0.1 g of glutathione–agarose resin for 1 h and
loaded into purification column. When all liquid portions flowed
through the column, 1 or more column volumes of PBS was used
to wash the glutathione-agarose resin, which had already bound
to the fusion protein GST-tagged Src SH2 domain, until no proteins
can be detected in the flow-through by Bradford protein reagent. In
order to adjust the pH of the resin from 7.4 to 8.0, another 1 vol-
ume of 50 mM tris buffer (pH 7.4) and 1 volume of 50 mM tris buf-
2.1.1. Ac-Lys(IDA)-pTyr-Glu-Glu-Ile-Glu-Lys(IDA) (1)
The peptide was prepared as described above at 0.3 mmol scale
and purified by HPLC in three separate batches. The final overall
yield was 141 mg (36%). LC–MS (ESI-TOF) (m/z) Anal. Calcd for
C52H78N9O27P: 1291.5. Found: 1291.7 [M]+.
2.1.2. Ac-Lys-pTyr-Glu-Glu-Ile-Glu-Lys (2)
The peptide was prepared as described above at 0.095 mmol
scale and purified by HPLC. The final overall yield was 38 mg
(38%). LC–MS (ESI-TOF) (m/z) Anal. Calcd for C44H70N9O19P:
1059.5. Found: 1059.6 [M]+.
2.1.3. Ac-Lys(IDA)-pTyr-Glu-Glu-Ile-Glu-Lys (3)
fer (pH 8.0) were used to wash the beads. Reduced L-glutathione
The peptide was prepared as described above at 0.085 mmol
scale and purified by HPLC. The final overall yield was 12 mg
(12%). LC–MS (ESI-TOF) (m/z) Anal. Calcd for C48H74N9O23P:
1175.5. Found: 1176.0 [M + H]+.
(10 mM) in 50 mM tris (pH 8.0) buffer was used to elute the fusion
protein. All purification steps were carried out at 4 °C. Protein con-
centrations were determined by Bradford protein assay using bo-
vine serum albumin as a standard.
2.1.4. Ac-Lys-pTyr-Glu-Glu-Ile-Glu-Lys(IDA) (4)
2.4. Fluorescence polarization binding assay
The peptide was prepared as described above at 0.085 mmol
scale and purified by HPLC. The final overall yield was 25 mg
(25%). LC–MS (ESI-TOF) (m/z) Anal. Calcd for C48H74N9O23P:
1175.5. Found: 1175.9 [M + H]+.
2.4.1. Competitive fluorescence polarization binding assay in presence
of the Src SH2 domain
All peptides were tested as competitors of the fluorescent probe
F-GpYEEI for binding affinity to the Src SH2 domain using fluores-
cent polarization (FP) competitive binding assay as described be-
low. FP intensities were measured at 25 °C in a disposable glass
2.1.5. Fluorescein-Lys(IDA)-pTyr-Glu-Glu-Ile-Glu-Lys(IDA) (5)
The peptide was prepared as described above at 0.063 mmol
scale. After the final Lys(IDA) coupling and Fmoc deprotection,
5,6-carboxyfluorescein succinimidyl ester (33 mg, 0.070 mmol)
and DIPEA (0.066 mL, 0.25 mmol) were added to the swelled resin
in DMF (1 mL) and the mixture was shaken overnight in light-pro-
tected column. After the solution was drained, the resin was
washed with DMF (3 Â 20 mL), and the fluorescein coupling was
repeated overnight. After the solution was drained, the resin was
washed with DMF (4 Â 20 mL). The peptide was cleaved and puri-
fied by HPLC as described above. All the fractions were protected
from light. The final overall yield was 11 mg (11%). LC–MS (ESI-
TOF) (m/z) Anal. Calcd for C71H86N9O32P: for 1607.5. Found, 1609.2
[M + H]+; MS (SELDI-TOF) (m/z) Anal. Calcd for C71H86N9O32P:
1607.5. Found: 1606.3 [M À 1]+, 1629.2 [M + Na À 1]+, and 1645.7
[M + K À 1]+.
tube (volume of 600 lL) using a Perkin-Elmer LS 55 luminescence
spectrometer equipped with an FP apparatus. The excitation and
emission wavelengths were set at 485 and 535 nm, respectively.
For the competition assay, final concentrations of 80 nM fluores-
cent probe F-GpYEEI, 750 nM Src SH2 domain, phosphate buffer
(20 mM, pH 7.3, 100 mM NaCl, 2 mM DTT, 0.1% BSA), water, and
various concentrations (0–100
were used. The order of addition to each glass tube (600
l
M) of each competitor peptide
L) was
l
(i) buffer, (ii) water, (iii) fluorescent probe, (iv) Src SH2 domain,
and (v) competitor peptide. A blank control (with the Src SH2 do-
main but without a peptide) and a background control (without
both the Src SH2 domain and the peptide) were used. The inhibi-
tion percentage of fluorescence probe binding to the Src SH2 do-
main by the sample was calculated by the following equation:

136
G. Ye et al. / Bioorganic Chemistry 37 (2009) 133–142
ꢀ
ꢁ
amount of EDTA added to each concentration point of Ni2+ de-
scribed above. The Src SH2 domain had a fixed concentration of
0.75 M in all experiments.
FPblk À FPs
1 À
 100
FPblk À FPbgd
l
where FPblk is the fluorescent polarization value of the blank con-
trol. FPs is the fluorescent polarization value of the sample (pep-
tide), and FPbgd is the fluorescent polarization value of the
background control. The inhibition percentages of the various con-
centrations of the assayed peptides were plotted, and the Kd value
(the concentration of peptides at which the binding site on Src
SH2 domain is half occupied) was calculated using LAB Fit Curve Fit-
ting Software 7.2. The reported Kd values are the mean of three sep-
arate determinations with a standard deviation of less than 5%.
2.5. Circular dichroism (CD) spectroscopy
All CD spectra were recorded in 1 cm path length cuvette on a
nitrogen-flushed JASCO J-810 spectropolarimeter interfaced with
a 25 °C water bath by averaging six consecutive scans. All spectra
were recorded with a 4 ms response and a bandwidth of 1 nm. CD
spectra were measured with the spectropolarimeter using
a
50 nm minÀ1 scan speed. All spectra were corrected for back-
ground by subtraction of appropriate blanks. The data are repre-
sented in the 185–265 nm spectral range.
2.4.2. Competitive fluorescence polarization binding assay in presence
of the Src SH2 domain and different metal ions
The assays were performed as described above with the excep-
tion that the metal ions with different final concentrations [NiCl2
3. Results and discussion
(2, 5, 10, 50, 100, 200, 300 and 500
lM), CoCl2, CdCl2, CuCl2, and
3.1. Chemistry
ZnSO4 (2 and 300 M)], were added after addition of the competi-
l
tor peptides. In another set of assay, EDTA (300 lM) was also used
after the addition of the same concentration of Ni2+ and incubated
for 10 min.
Peptides containing 0–2 IDA groups (Fig. 1) were synthesized by
solid-phase Fmoc-based chemistry employing Rink amide NovaGel
resin as the solid support and Fmoc-L-amino acids as building
blocks. The synthesis of Fmoc-Lys(IDA(OtBu))-OH has been previ-
ously reported [19]. The synthesis of peptide 1 is shown as a repre-
sentative example in Scheme 1 using Fmoc-Lys(IDA(OtBu))-OH,
2.4.3. Saturation binding assay using fluorescein-labeled peptide 5
Fluorescein-labeled peptide 5 (80 nM) was used in the presence
of Ni2+ alone at different concentrations (0.08, 0.16, 0.32, 0.64, 1, 3,
Fmoc-Glu(OtBu)-OH,
Fmoc-Ile-OH,
Fmoc-Glu(OtBu)-OH,
10, 30, 100, 200, 300, 400, and 500 lM). In another experiment, F-
Fmoc-Glu(OtBu)-OH, Fmoc-Tyr(H2PO3)-OH, and Fmoc-Lys(IDA-
(OtBu))-OH, respectively, in coupling reactions with the resin.
GpYEEI (80 nM) was used in the place of peptide 5 as a control pep-
tide. The same assay was carried out in the presence of a similar
Scheme 1. Solid-phase synthesis of peptide 1.

G. Ye et al. / Bioorganic Chemistry 37 (2009) 133–142
137
2-(1H-Benzotriazole-1-yl)-1,1,3,3-tetramethyluronium
hexa-
compared in the presence of the reference peptide probe, F-
GpYEEI, using a fluorescence polarization competitive assay. The
results expressed as Kd values are shown in Table 2.
fluorophosphate (HBTU) and N,N-diisopropylethylamine (DIPEA)
in N,N-dimethylformamide (DMF) were used as coupling and acti-
vating reagents, respectively. Fmoc deprotection at each step was
carried out using piperidine in DMF (20%). After the final deprotec-
tion, the peptides were capped by N-terminal acetylation with ace-
tic anhydride. Then a mixture of trifluoroacetic acid (TFA):water:
triisopropylsilane (TIS) (95:2.5:2.5) was used for side chain depro-
tection of amino acids and cleavage of the synthesized peptide from
the resin. Finally, the resin was hydrolyzed for 19 h after addition of
10% water, dried in vacuum, and lyophilized. The crude peptides
were purified by preparative reverse-phase HPLC and analyzed by
LC–MS using a gradient of 5–95% acetonitrile (0.1% formic acid)
and water (0.1% formic acid) at 20 mL/min over 30 min. The purity
of final products was confirmed by analytical HPLC (>95%) using a
similar gradient system at 1 mL/min. A similar protocol was also
used for the synthesis of peptides 2–5 using appropriate Fmoc-ami-
no acid building blocks or 5,6-carboxyfluorescein succinimidyl
ester.
Peptide 1 with two IDA groups at the N- and C-terminal lysine
residues showed 2.7-fold higher binding affinity toward the SH2
domain when compared with the optimal sequence Ac-GpYEEI.
These data suggest that the carboxylic groups on the IDA groups
may have contributed to the binding to the Src SH2 domain bind-
ing pockets, possibly through additional hydrogen bonding or elec-
trostatic interactions. Furthermore, the introduction of the two
negatively charged IDA groups may have generated a favorable
conformational change within modified pYEEI derivative for
appropriate binding to the SH2 domain.
The crystal structure of the Src SH2 domain complexed with
peptide pYEEI showed two major binding pockets within the do-
main substrate binding site, the hydrophilic pTyr binding pocket
(P site) and the hydrophobic pocket (P + 3 site) [3]. The glutamic
acid residues (EE) in the P + 1 and P + 2 positions have minor inter-
actions with amino acids in the SH2 domain. The two lysine resi-
dues in peptide 1 are located at the P À 1 and P + 5 positions, and
may have minor interactions with other small binding grooves or
other important amino acids outside the P and P + 3 sites. Further
structural studies are required to identify the key amino acids out-
side these two major binding pockets that are involved in interac-
tions with IDA groups. That information may provide insights for
designing more potent Src SH2 domain binding ligands.
3.2. UV titration studies
UV Titration studies were carried out to determine whether Ni2+
binds to the peptides. Different concentrations of peptide 1 (1–
100 lM) were incubated with a fixed concentration of 10 lM
Ni2+ (Fig. 2a). In a lower range of peptide concentration (<10
lM)
and less than one equivalent of Ni2+ concentration, the monitored
absorbance was concentration-dependent and increased consis-
tently. However, at higher concentrations of the peptide
Peptide 2, which lacks the two IDA groups but has the two ly-
sine residues at both terminals, had a Kd value of 52.70
lM show-
ing significant decrease in the binding affinity when compared
with peptide 1 and Ac-pYEEI. The reduction of binding affinity in
2 suggests that the presence of positively-charged amino acid res-
idues at both N- and C-termini is detrimental to interactions with
the SH2 domain. Incorporation of one IDA group at the N- or C-ter-
minal in peptides 3 and 4 rescued partially the binding affinity
when compared with peptide 2. However, peptides 3 and 4 showed
approximately 5-fold lower binding affinity toward the SH2 do-
main when compared with that of peptide 1 containing two IDA
groups. These data suggest that the presence of two positively-
charged lysine residues at either the N- or C-terminals is deleteri-
ous in interactions with the SH2 domain. On the other hand, the
presence of one or two IDA groups can provide additional interac-
tions with amino acid residues existing on both sides of the bind-
ing pocket regions of the Src SH2 domain leading to enhanced
binding affinity of the peptides toward to the SH2 domain.
(>15
Ni2+ ions was already saturated with peptides.
When the was plotted with different equivalents (from 0 to
lM), the absorbance reached its steady state, suggesting that
D
e
10) of peptide 1, the curve was consistent with the formation of a
1:1 metal–peptide complex (Fig. 2b). This indicates that the two
IDA groups on the N- and C-termini can possibly interact with
one Ni2+ ion and form a cyclic peptide–metal chelation complex.
However, other peptide–metal complex species, such as those
resulting from intermolecular interactions between the peptide
and a Ni2+ ion in 1:2 metal–peptide complex, may also be present
at higher concentrations of the peptide. The conformation and the
binding profile of this cyclic complex were further investigated by
using CD and fluorescence polarization studies.
The difference absorption spectra of free and metal-bound cyc-
lic peptide 1 demonstrated that the absorbance at ꢀ225 nm can be
used to monitor the formation of peptide–metal complex (Fig. 2c).
The Kd was determined to be 234.7 nM using the Igor Pro algorithm
by plotting the absorbance change at specific wavelength versus
concentration of peptide upon addition of aliquots of Ni2+ from
an aqueous stock.
Peptide 2 containing two unsubstituted lysine residues exhib-
ited the highest binding affinity toward Ni2+ (Kd = 97.4 nM) (Table
1) when compared with other peptides. Peptides 3 and 4 also
exhibited higher binding affinity toward Ni2+ when compared with
peptide 1, possibly due to the presence of at least one free amino
group that is a stronger chelating group than a carboxylate. Control
peptide Ac-pYEEI does not have a strong chelating group at the N-
or C-termini. Thus, Ac-pYEEI did not exhibit a high binding affinity
toward Ni2+, suggesting that the high affinity binding between
peptide 1 and Ni2+ is not due to the interactions with negatively-
charged phosphate group.
3.3.2. Competitive binding assay in presence of the Src SH2 domain
and different metal ions
Table 3 shows that by incubating peptide 1 in the presence of
different concentrations of Ni2+ (0–500
l
M for peptide 1), the bind-
ing affinities toward the Src SH2 domain consistently decrease. The
binding affinity of peptide 1 (Kd = 0.6 M) was reduced by more
than 2-fold (Kd = 1.6 M). This
M) upon addition of Ni2+ (300
l
l
l
binding affinity is similar to the Kd value of Ac-pYEEI, suggesting
that Ni2+ possibly causes some structural effects on the binding
pocket of the protein or blocks the interactions of the carboxylate
groups with the SH2 domain through the formation of a complex
between the two IDA groups and Ni2+
.
These data were consistent with the results of UV titration stud-
ies showing the formation of a 1:1 complex between peptide 1 and
Ni2+ with a Kd value of 234.7 nM. In the experimental conditions of
the fluorescence polarization assay while free peptide and pep-
tide–metal complex coexist, the binding affinity of peptide 1 is
3.3. Fluorescence polarization (FP) binding assay
approximately 7-fold lower toward the SH2 domain than Ni2+
.
3.3.1. Competitive binding assay in presence of the Src SH2 domain
The binding affinities of the synthesized peptides 1–4 along
with Ac-pYEEI against the Src SH2 domain were determined and
On the other hand, even at the highest concentration of Ni2+
(500 M), the peptide retained a high binding affinity for the Src
SH2 domain. It is not clear from these experiments whether the
l

138
G. Ye et al. / Bioorganic Chemistry 37 (2009) 133–142
Fig. 2. (a) UV titration of peptide 1 (1–100
l
M) with a solution of Ni2+ (10
l
M). The data are plotted as
D
e
versus wavelength. (b) The
D
e
versus equivalents of peptide 1 (0–10
M) (black);
Ni2+-bound peptide: peptide 1 (10
lM) with NiCl2 (10 lM) (red). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version
equiv). One equivalent of Ni2+ interacts with one equivalent of peptide 1. (c) Difference absorption spectra of free and metal-bound peptide 1. Free peptide 1 (10
l
of this article.)

G. Ye et al. / Bioorganic Chemistry 37 (2009) 133–142
139
Table 1
3, upon adding EDTA to the system, the binding affinity of peptide
1 was almost restored to the original level, changing the Kd value
from 1.55 to 0.79 lM. EDTA was a very effective competitive che-
The binding affinity of peptides 1–4 to Ni2+
.
Peptides
Kd (nM)a
lating agent with a much higher affinity than peptide 1 for binding
to Ni2+. It is possible that EDTA reversed the structural effect of Ni2+
on the protein. Another possibility is that most of the peptide in
the peptide–Ni2+ complex was replaced by EDTA, forcing the pep-
tide 1 to return to the unbounded free state for optimal interaction
with the SH2 domain.
Ac-pYEEI
>600
234.7
97.4
153.2
161.9
1
2
3
4
a
Kd: the concentration of peptides (Ac-pYEEI or 1–4) at which half of them bind
to Ni2+. Values are means of three experiments with a standard deviation of less
than 5%.
Similarly the addition of Ni2+ ions (0, 2 and 300
lM) to peptides
2–4 slightly affected the Kd values. These data suggest that the addi-
tion of Ni2+ did not significantly change the pattern of interactions
for peptides 2–4 with the Src SH2 domain and the changes were be-
lieved to be caused by Ni2+ addition only. The Ni2+ ion itself may
have caused some structural effects on the Src SH2 domain that
may have eventually affected the binding with the peptides. This
phenomenon was shown in the CD experiments described later.
To test the hypothesis, other types of transition metals were se-
lected, including Co2+, Cd2+, Cu2+, and Zn2+, and used in similar
experimental conditions (Table 4). Interestingly, only Co2+ showed
Table 2
Binding affinity of peptides containing IDA group and Ac-pYEEI to the Src SH2
domain.
Peptides
Kd (l
M)a
Ac-pYEEI
1.70
0.63
52.70
2.96
3.35
1
2
3
4
similar binding patterns when compared to Ni2+ at 2 and 300
All other metals, Cd2+, Cu2+, and Zn2+, showed deleterious effects in
the assay at both 2 and 300 M conditions. Even at 2
M of Cd2+
lM.
a
Kd: the concentration of peptides (Ac-pYEEI or 1–4) at which the binding site at
the Src SH2 domain is half occupied. Values are means of three experiments with a
standard deviation of less than 5%.
l
l
,
Cu2+, and Zn2+, the FP values revealed that there were no binding
interactions of F-GpYEEI to the Src SH2 domain, possibly due to
the partial or complete denaturation/structural changes of the
Src SH2 domain and/or changing the conformation of the protein
binding sites in the presence of these metals. It is not clear why this
type of protein structural modifications do not happen in the pres-
ence of Ni2+ and Co2+, but could be related to the physicochemical
properties of the metals and their differences in interactions with
the SH2 domain.
Table 3
Binding affinity of the peptides containing 0–2 IDA groups (1–4) toward the Src SH2
domain in the presence of Ni2+
.
Compounds
Ni2+
(l
M)
Kd (l
M)a
1
1
1
1
1
1
1
1
1
0
2
5
0.63
0.69
1.02
1.08
1.18
1.15
1.07
1.55
1.45
0.79
52.70
55.68
39.30
2.96
2.98
3.70
3.35
4.13
4.01
10
50
100
200
300
500
300
0
2
300
0
2
300
0
3.3.3. Saturation binding assay using fluorescein-labeled peptide 5
Fluorescein-labeled peptide 5 (Fig. 1) was synthesized and used
as a probe instead of F-GpYEEI to study the interactions of peptide
containing two IDA groups with Ni2+ in the presence of EDTA or the
Src SH2 domain. Several different FP studies were performed, such
as assays of peptide 5 with Ni2+, peptide 5 with Ni2+ in the presence
or absence of same concentration of EDTA, and peptide 5 with Ni2+
in the presence of a fixed concentration of the Src SH2 domain. F-
GpYEEI with Ni2+ was used as a control.
1 (300 lM EDTA)
2
2
2
3
3
3
4
4
4
As shown in Fig. 3, the control peptide F-GpYEEI did not show
any dramatic change in FP values upon addition of Ni2+ even at
2
300
high concentration of 500 lM. This result was expected since Ac-
pYEEI had very weak affinity toward Ni2+ as described above. How-
ever, peptide 5 showed a significant increase in FP values in the
presence of Ni2+ when compared with F-GpYEEI, suggesting the
formation of peptide–metal complex in the assay condition. To
explain this observation, the same concentration of EDTA was
added to the assay system. In the presence of the metal scavenger,
the FP values showed no significant changes at different concentra-
a
Kd: the concentration of peptides (Ac-pYEEI or 1–4) at which the binding site at
the Src SH2 domain is half occupied. Values are means of three experiments with a
standard deviation of less than 5%.
peptide–metal complex binds to the SH2 domain in a similar man-
ner to peptide 1. One direct method to determine whether peptide
1 and the peptide–metal complex bind to the Src SH2 domain in
the cyclized form is to synthesize a covalently cyclized peptide
by linking between the two IDA groups or two lysine residues
and compare the binding affinity with the peptide 1 in the pres-
ence of Ni2+. Unfortunately, the synthesis of the covalently cyclized
peptide was not successful by using standard methods. Further
studies, such as synthesizing a new class of cyclized peptides with
different ring sizes, are required to understand effect of the ring
size and the role of the lysine residues in the binding to the Src
SH2 domain.
Table 4
Binding affinity of peptide 1 toward the Src SH2 domain in the presence of different
concentrations of metals (Co2+, Cd2+, Cu2+, and Zn2+).
Compounds
1 (2
M Co2+
1 (with 300
1 (with 2
Kd (l
M)a
l
)
l
1.09
2.55
NAb
NAb
M Co2+
)
l
M Cd2+, Cu2+, or Zn2+
)
1 (with 300 l )
M Cd2+, Cu2+, or Zn2+
a
Kd: the concentration of peptides (Ac-pYEEI or 1–4) at which the binding site at
the Src SH2 domain is half occupied. Values are means of three experiments with a
standard deviation of less than 5%.
To further prove the effect of the peptide–metal complex for-
mation on the binding affinity of peptide 1 to the Src SH2 domain,
FP assay was conducted by adding the same concentration of EDTA
b
NA = no binding of F-GpYEEI with the Src SH2 domain was observed after metal
(300 lM) to the system containing 300 l
M Ni2+. As shown in Table
addition.

140
G. Ye et al. / Bioorganic Chemistry 37 (2009) 133–142
Fig. 3. The fluorescence polarization values in different saturation assay conditions.
tions of Ni2+, meaning that EDTA disrupted the complex formation
of peptide 5 through chelating with Ni2+
showed more folding in the presence of Ni2+, but compared to pep-
tide 1, elipticity values were lower even at the highest concentra-
tions of Ni2+. This was mainly because the amino groups of peptide
.
As described above, the interaction between peptide 1 and the
Src SH2 domain was slightly disrupted upon the addition of high
2 can also chelate to Ni2+
.
concentrations of Ni2+ (Kd changed from 0.63 to 1.55
lM). Peptide
Peptides 3 and 4 showed similar pattern to 1 with a minor in-
crease of elipticity in the presence of Ni2+. However, their elipticity
values were higher than peptide 2. In general, it appears that pep-
tides 1, 3 and 4 have secondary structures with higher elipticities
than that of peptide 2 and their secondary structures are not af-
5 showed a similar binding disruption pattern in the presence of
the Src SH2 domain as shown by the reduced fluorescence polari-
zation values with increasing Ni2+ concentrations (Fig. 3). Similarly,
fluorescent-labeled peptide 5 was still able to bind to the Src SH2
domain even at 500
l
M of Ni2+. The presence of the fluorescein
fected significantly in the presence of Ni2+
.
group at the N-terminal of peptide 5 may have partially changed
the conformation of the optimal interaction with the SH2 domain.
3.4.2. CD studies of the Src SH2 domain (fixed concentration) in the
presence of different concentrations of Ni2+
Slight enhancement in elipticity of the Src SH2 domain was ob-
served in the presence of Ni2+, suggesting that this metal affects the
secondary structure of Src SH2 domain (Fig. 5). Because of there are
several types of secondary structures in the Src SH2 domain,
3.4. Circular dichroism (CD) spectroscopy study
3.4.1. CD studies of peptides with different concentration of Ni2+ in the
absence of the Src SH2 domain
including
a-helices and b-sheets, the peaks are broader from
Peptide 1 showed similar folded conformations at different con-
200–230 nm with a minimum at 210 nm.
centrations of Ni2+ (0–500
lM) (Fig. 4). The peptide showed a signif-
icantly folded structure in the absence of the metal based on the
elipticity values and retained its secondary structure in the presence
of Ni2+. This result was not surprising since the binding affinity of
peptide 1 to the SH2 domain was slightly decreased in the presence
of Ni2+. In other words, Ni2+ did not disrupt the conformation of pep-
tide 1 significantly for interaction with the SH2 domain. This may
explain why peptide 1 itself was able to bind to the Src SH2 domain
with a higher binding affinity than Ac-pYEEI. The addition of two
IDA groups may have assisted the peptide to adopt a favorable con-
formation for the interaction with the Src SH2 domain. The addition
of Ni2+ partially disrupted the interaction of IDA groups with the SH2
domain, but it did not change the conformation of the peptide since
the folding of the peptide was not disturbed. Peptide 1 appears to
have an appropriate folded structure in the absence of Ni2+ and addi-
tion of the metal did not change the secondary structure during the
interaction with the SH2 domain.
3.4.3. CD studies of fixed concentration of peptides and the Src SH2
domain with different concentrations of Ni2+
Similar equivalents of the peptides and SH2 domain were used
with different concentrations of Ni2+ in similar ratios as the other
experiments described above (Fig. 6). In general, the changes in
the secondary structures appear to be from the effect of Ni2+ on
the Src SH2 domain and not because of the interaction of the pep-
tide with the Src SH2 domain. Comparing the spectra with those of
the Src SH2 domain alone in the presence of Ni2+ (Fig. 5) shows that
the differences are not significant. These data were consistent with
the binding affinity data that showed lower binding affinity in the
presence of Ni2+, possibly because of the changes in the secondary
structure of the Src SH2 domain at high concentrations of the me-
tal. Peptides 1, 3, and 4 showed some elipticity enhancement after
1–1.5
significant change for the peak with the minima of 210 nm. Peptide
l
M of Ni2+. At higher concentrations of Ni2+, there was no
2 exhibited larger changes up to 1.5 l .
M of Ni2+
Peptide 2 exhibited a much lower binding affinity to the SH2
domain because of the importance of the IDA groups versus the
positively-charged amino groups. Peptide 2 was folded much less
compared to 1 in the absence of the metal based on the elipticity
4. Conclusions
values. Addition of 5
As shown in Table 3, the binding affinity of 2 increased in the pres-
ence of 300 M), suggesting the folding of
M of Ni2+ (Kd = 39.3
l
M Ni2+ to 2 enhanced the elipticity values.
Herein, we introduced novel peptides containing 0–2 IDA
groups at the side chains of lysine residues flanking the optimal
sequence of the Src SH2 domain ligand pYEEI to introduce confor-
mational constraints through non-covalent forces, such as intra-
molecular hydrogen bindings or metal chelation. UV titration
l
l
this peptide in the presence of the high concentration of the metal
may have assisted the interaction with the SH2 domain. Peptide 2

G. Ye et al. / Bioorganic Chemistry 37 (2009) 133–142
141
Fig. 4. The CD spectra of peptides 1–4 (100 l
M) in the presence of different concentrations of Ni2+ in the absence of the Src SH2 domain.
Fig. 5. The CD spectra of different concentration of Ni2+ and fixed concentration of Src SH2 domain.
studies demonstrated that all four peptides were capable of chelat-
ing with one Ni2+ ion possibly by formation of a cyclic complex.
However, only peptide 1 showed 2.7-fold higher binding affinity
toward the SH2 domain when compared with the optimal se-
quence Ac-GpYEEI. Although CD experiments did not show signif-
icant changes in the structure of the peptide in the absence or
presence of Ni2+, the increased affinity when compared to the ref-
erence peptide Ac-pYEEI indicated that the addition of IDA groups
at the two terminals may have already caused the formation of a
favorable conformation for peptide 1 for binding to the Src SH2 do-
main. The non-covalent interactions between the two IDA groups
within peptide 1 could explain the lack of dramatic effect of Ni2+
ion on the ligand binding affinity toward SH2 domain. Ni2+ could
serve as a bridge to bring the two IDA groups together, achieving
similar conformational effect as that via direct inter-IDA group
interactions in the absence of Ni2+. Furthermore, based on the CD
experimental data, Ni2+ caused the structural disruption on the
protein and thus modest effect on the peptide-SH2 domain binding
interaction. FP assays showed a 2-fold decrease of the binding
affinity for the Src SH2 domain for peptide 1 when Ni2+ was added

142
G. Ye et al. / Bioorganic Chemistry 37 (2009) 133–142
Fig. 6. The CD spectra of fixed concentration of peptides 1–4 and the Src SH2 domain with different concentrations of Ni2+
.
Schravendijk, S. Adams, S. Violette, J. Smith, W. Guan, C. Bartlett, J. Herson, J.
Iuliucci, M. Weigele, T. Sawyer, Proc. Natl. Acad. Sci. USA 97 (2000)
9373–9378.
to the system at high concentration of 300–500 lM. Nonetheless,
the addition of EDTA dropped the Kd value back to the almost ori-
ginal level, suggesting the disruption is reversible. In future stud-
ies, optimized substrate sequences from other protein families
may be used to construct IDA containing substrate peptides. With
further optimization, this class of peptides may serve as potential
SH2 domain binding ligands.
[5] T.R. Burke Jr, Z.J. Yao, D.G. Liu, J. Voigt, Y. Gao, Biopolymers 60 (2001) 32–44.
[6] N. Kawahata, M.G. Yang, G.P. Luke, W.C. Shakespeare, R. Sundaramoorthi, Y.
Wang, D. Johnson, T. Merry, S. Violette, W. Guan, C. Bartlett, J. Smith, M.
Hatada, X. Lu, D.C. Dalgarno, C.J. Eyermann, R.S. Bohacek, T.K. Sawyer, Bioorg.
Med. Chem. Lett. 11 (2001) 2319–2323.
[7] J.P. Davidson, O. Lubman, T. Rose, G. Waksman, S.F. Martin, J. Am. Chem. Soc.
124 (2002) 205–215.
[8] N.H. Nam, G. Ye, G. Sun, K. Parang, J. Med. Chem. 47 (2004) 3131–3141.
[9] Z. Songyang, S.E. Shoelson, M. Chaudhuri, G. Gish, T. Pawson, W.G. Haser, F.
King, T. Roberts, S. Ratnofsky, R.J. Lechleider, B.J. Neel, R.B. Birge, J.E. Fajardo,
M.M. Chou, H. Hanafusa, B. Schaffhausen, L.C. Cantley, Cell 72 (1993) 767–778.
[10] J.M. Humphrey, A.R. Chamberlin, Chem. Rev. 97 (1997) 2243–2266.
[11] A.R. Khan, J.C. Parrish, M.E. Fraser, W.W. Smith, P.A. Bartlett, M.N. James,
Biochemistry 37 (1998) 16839–16845.
[12] E. Hochuli, H. Dobeli, A. Schacher, J. Chromatogr. 411 (1987) 177–184.
[13] M.C. Smith, T.C. Furman, T.D. Ingolia, C. Pidgeon, J. Biol. Chem. 263 (1988)
7211–7215.
Acknowledgments
We acknowledge the financial support from NIH Grant #
R44CA099126, American Cancer Society Grant # RSG-07-290-01-
CDD, and National Science Foundation Grant # CHE 0748555. We
acknowledge National Center for Research Resources, NIH, Grant
# 1 P20 RR16457 for sponsoring the core facility.
[14] F.H. Arnold, B.L. Haymore, Engineered metal-binding proteins: purification to
protein folding, Science 252 (1991) 1796–1797.
References
[15] I. Hamachi, J.I. Watanabe, R. Eboshi, T. Hiraoka, S. Shinkai, Biopolymers 55
(2000) 459–468.
[16] J.R. Morgan, R.P. Lyon, D.Y. Maeda, J.A. Zebala, Nucleic Acids Res. 36 (2008)
3522–3530.
[1] J.S. Biscardi, D.A. Tice, S.J. Parsons, Adv. Cancer Res. 76 (1999) 61–119.
[2] C.A. Metcalf 3rd, M.R. van Schravendijk, D.C. Dalgarno, T.K. Sawyer, Curr.
Pharm. Des. 8 (2002) 2049–2075.
[17] F. Ruan, Y. Chen, K. Itoh, T. Sasaki, P.B. Hopkins, J. Org. Chem. 56(1991)4347–4354.
[18] I. Hamachi, Y. Yamada, T. Matsugi, S. Shinkai, Chem. Eur. J. 5 (1999) 1503–1511.
[19] S. Hutschenreiter, L. Neumann, U. Radler, L. Schmitt, R. Tampe, Chem. Biochem.
4 (2003) 1340–1344.
[3] T.K. Sawyer, Biopolymers 47 (1998) 243–261.
[4] W. Shakespeare, M. Yang, R. Bohacek, F. Cerasoli, K. Stebbins, R.
Sundaramoorthi, M. Azimioara, C. Vu, S. Pradeepan, C. Metcalf 3rd, C.
Haraldson, T. Merry, D. Dalgarno, S. Narula, M. Hatada, X. Lu, M.R. van