A synthetic polypeptide conjugate from a 42-residue polypeptide and salicylhydroxamic acid binds human myeloperoxidase with high affinity

Article abstract of DOI:10.1002/psc.2459

Myeloperoxidase (MPO) is a 150kD tetrameric heme protein consisting of two heavy chains and two light chains, which is present in neutrophils, white blood cells, at concentrations between 2% and 5% and plays an important role in the innate immune system. The MPO concentration in serum or plasma has been shown to be linked to the risk for cardiovascular diseases, and MPO is considered to be a high potential diagnostic biomarker. To develop a molecule that binds MPO, salicylhydroxamic acid (SHA), a substrate analog inhibitor of MPO with a K_D=2μM, was conjugated to a designed set of 42-residue polypeptide scaffolds via 9- and 11-carbon atom aliphatic spacers to form 20 different protein binder candidates, and their interactions with MPO were evaluated by surface plasmon resonance analysis. The polypeptide conjugate 4C37L34C11SHA was found to bind to MPO with an affinity that could be estimated to have a dissociation constant of around 400pM, nearly four orders of magnitude higher than that of SHA. Inhibition of binding to MPO by free SHA was observed in competition experiments demonstrating that the binding of the polypeptide conjugate is dominated by the interactions of SHA with the heme cavity. Although still in the future, the discovery of these new synthetic binders for MPO suggests a route to clinical diagnostic tests in vivo or in vitro, independent of antibodies.

Full text of DOI:10.1002/psc.2459

Research Article
Received: 12 June 2012
Revised: 12 September 2012
Accepted: 21 September 2012
Published online in Wiley Online Library: 19 October 2012
(wileyonlinelibrary.com) DOI 10.1002/psc.2459
A synthetic polypeptide conjugate
from a 42-residue polypeptide and
salicylhydroxamic acid binds human
myeloperoxidase with high affinity
Xiaojiao Sun, Jie Yang, Thomas Norberg* and Lars Baltzer
Myeloperoxidase (MPO) is a 150 kD tetrameric heme protein consisting of two heavy chains and two light chains, which is
present in neutrophils, white blood cells, at concentrations between 2% and 5% and plays an important role in the innate
immune system. The MPO concentration in serum or plasma has been shown to be linked to the risk for cardiovascular
diseases, and MPO is considered to be a high potential diagnostic biomarker. To develop a molecule that binds MPO,
salicylhydroxamic acid (SHA), a substrate analog inhibitor of MPO with a KD = 2 mM, was conjugated to a designed set of
42-residue polypeptide scaffolds via 9- and 11-carbon atom aliphatic spacers to form 20 different protein binder candidates,
and their interactions with MPO were evaluated by surface plasmon resonance analysis. The polypeptide conjugate
4C37L34C11SHA was found to bind to MPO with an affinity that could be estimated to have a dissociation constant of around
400 pM, nearly four orders of magnitude higher than that of SHA. Inhibition of binding to MPO by free SHA was observed in
competition experiments demonstrating that the binding of the polypeptide conjugate is dominated by the interactions of
SHA with the heme cavity. Although still in the future, the discovery of these new synthetic binders for MPO suggests a
route to clinical diagnostic tests in vivo or in vitro, independent of antibodies. Copyright © 2012 European Peptide Society
and John Wiley & Sons, Ltd.
Supporting information can be found in the online version of this article.
Keywords: polypeptide; conjugate; salicylhydroxamic acid; myeloperoxidase
as a diagnostic biomarker for the early detection of CVD. In a
healthy individual, the MPO level is low, with a few to a hundred
Introduction
Cardiovascular diseases (CVD) remain the leading cause of death
globally, and more than 17 million people died from CVD in 2008.
By 2030, more than 23 million people are expected to die from
CVD, mainly from heart disease and stroke [1]. Diagnostic tools
are in great demand that enable early stage discovery of CVD
for therapeutic intervention and monitoring of disease as a
function of treatment.
micrograms per liter (20–600 pM) in serum or plasma, whereas in
patients with increased risks of a heart infarction, the level of
MPO is higher, and a cut-off value of 350 mg/l (2.3nM) has been
suggested [6–8]. It has also been suggested that the measurement
of MPO in combination with the C-reactive protein (CRP) could
provide improved precision in diagnosis [5].
Diagnostic protein biomarkers for CVD are routinely used in
the clinic, and antibodies remain the golden standard in bioana-
lytical applications including in vitro diagnostics. Currently, the
measurement of MPO is typically carried out by immunological
methods using anti-MPO antibodies as recognition elements
with a detection limit of 1.6 mg/l (10 pM). Drawbacks of antibody
technologies in comparison with molecules prepared by chemi-
cal synthesis include poor stability, batch-to-batch variation,
non-specific interactions with native proteins, and difficulties
encountered in site-specific modification. Small organic synthetic
binder molecules that recognize and bind proteins with sufficient
Myeloperoxidase (MPO) is a tetrameric, glycosylated, basic
heme protein with two heavy and two light chains and a pI of
9.2. MPO has a molecular weight of around 150 kD, depending
on isoform, and is abundant [2,3] in white blood cells (leuco-
cytes). Three isoforms are known, which only differ by the size
of the two heavy chains. MPO catalyzes the formation of
reactive species such as hypochlorous acid, which kills invading
bacteria and other pathogens as part of the innate immune
defense. During the acute inflammatory condition, MPO is
released in the extracellular medium and produces large amounts
of oxidants to kill the invading organisms. However, an excess of
MPO-generated oxidants do not only kill invading organisms but
also contribute to tissue damage during inflammation and also
to potentially proatherogenic biological activities throughout
the development of CVD. Correlations have been demonstrated
between elevated levels of MPO and CVD, e.g. atherosclerosis,
acute coronary syndrome, vascular endothelial dysfunction, and
multiple sclerosis [4,5], and MPO has therefore been suggested
*
Correspondence to: Thomas Norberg, Department of Chemistry-BMC,
Uppsala University, PO Box 576, SE-751 23 Uppsala, Sweden. E-mail:
thomas.norberg@kemi.uu.se
Department of Chemistry-BMC, Uppsala University, PO Box 576, SE-751 23
Uppsala, Sweden
J. Pept. Sci. 2012; 18: 731–739
Copyright © 2012 European Peptide Society and John Wiley & Sons, Ltd.

SUN ET AL.
affinities and selectivities for diagnostic applications have so far
proven to be cumbersome to develop, and the efforts required
to develop small organic molecules that bind proteins efficiently
seem only to be justified in pharmaceutical applications. New
strategies for protein binder development that address these
problems are of considerable interest.
Materials and Methods
General Methods
All solvents used, including DMF, N-methyl-2-pyrrolidone (NMP),
DCM, THF, ethyl acetate (EA), pentane (PE), chloroform, and
diethyl ether, were commercially available. DCM was distilled
after drying over CaH2. THF was distilled after refluxing/drying
over sodium/benzophenone. Reagents were used as obtained
from commercial sources unless otherwise stated. Concentra-
tions were performed at reduced pressure and a temperature
below 40 ꢀC. Measurement of pH was carried out using pH paper
(pH range 1–14). Flash chromatography was performed using
silica gel 60 (0.040–0.063 mm) as the stationary phase. TLC was
performed on aluminum plates pre-coated with silica gel (Merck
silica gel 60 F254), and spots were visualized by either UV light,
iodine vapor, 8% ethanolic phosphomolybdic acid (dipping/heating),
Recently, we proposed a general binder concept that uses
conjugates from small organic molecules and 42-residue poly-
peptides to form hybrid molecules that bind proteins with affini-
ties that are enhanced by three to four orders of magnitude in
comparison with those of the small molecules and selectivities
that are enhanced by one to two orders of magnitude [9–11].
The concept is robust and can be used to prepare selective,
high-affinity binders for, in principle, any protein in short time
by organic synthesis, provided that a small molecule or peptide
is available, which binds the protein with at least modest affinity
and selectivity. We have previously reported on a polypeptide
conjugate that binds CRP, and to develop a set of binder mole-
cules that could in principle be used to address CVD in a fully
synthetic measurement system, we have now developed a poly-
peptide conjugate that binds MPO. We therefore wish to report
on a synthetic polypeptide conjugate binder for MPO based on
salicylhydroxamic acid (SHA), a substrate analog inhibitor of
MPO with a KD = 2 mM, which was conjugated to a set of designed
polypeptides via 9 and 11 carbon aliphatic spacers. The best
binder molecule was selected from the resulting set of binder
candidates, and a powerful binder for MPO was identified using
surface plasmon resonance (SPR) analysis. The binder concept il-
lustrated for a binder for MPO is shown in Figure 1.
or 5% ethanolic sulfuric acid (dipping/heating). H NMR and 13C
1
NMR spectra for CDCl3 solutions were recorded at 25 ꢀC on a Varian
Unity Inova 400 spectrometer. Chemical shifts are reported in parts
per million using either CHCl3/TMS as reference (d = 7.26/0.00 for 1H
NMR) or CDCl3 (d = 77.0 for 13C NMR). Mass spectra of small mole-
cules (MW < 1000) were recorded on a SCIEX API 150-EX mass
spectrometer by direct injection of diluted samples in methanol.
Conjugation reactions were monitored by analytical RP-HPLC on
an analytical C18 column (Grace Vydac, pore size: 300Å; particle
size: 5 mM; length: 15 cm; diameter: 4.6 mm) using a 30-min gradient
of 5–95% of 90% acetonitrile in water containing 0.1% TFA at a flow
rate of 1 ml/min. The crude peptides were purified by RP-HPLC on a
preparative Hypersil Gold C18 column (Thermo Scientific, pore
Figure 1. Illustration of the principle of myeloperoxidase (MPO) binder molecule assembly. Salicylhydroxamic acid (SHA) is known to bind to the two
heme cavities with a dissociation constant, KD, of 2 mM. The three oxygen atoms of SHA are involved in hydrogen bond formation, and the aromatic ring
is important for hydrophobic interactions in the active site [16]. SHA was covalently linked to a lysine residue in each of the twenty 42-residue polypeptides
by 9- or 11-carbon aliphatic spacers attached to the aromatic ring. In the sequences, no other free lysines were accessible during conjugation to ascertain
that only one SHA residue was introduced. The structure of MPO was visualized by PyMOL 1.1 from X-ray structure (PDB code: 1MHL).
wileyonlinelibrary.com/journal/jpepsci Copyright © 2012 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2012; 18: 731–739

POLYPEPTIDE-SALICYLHYDROXAMIC ACID CONJUGATE BINDS MPO
size: 175 Å; particle size: 5 Å; length: 25 cm; diameter: 21.2 mm)
eluted with a 40-min gradient of 25–75% of 90% acetonitrile in
water containing 0.1% TFA at a flow rate of 10 ml/min. Peptides
and peptide conjugates were identified by MALDI-TOF MS
(Voyager-DE Pro, Applied Biosystems, Carlsbad, CA, USA), using
a-cyano-4-hydroxycinnamic acid as matrix.
dropwise to make a 2% HCl/ethanol solution. After 1 h of stirring
at room temperature, the solvent was evaporated, and the
residual yellow oil was purified by column chromatography
(PE/EA 9 : 1 to 2 : 1) to afford pure 3a or 3b.
Ethyl 5-(90-ethoxycarbonylnonyl)-2-hydroxybenzoate (3a). Yield:
0.83 g, 80%,1H NMR (CDCl3, 400 MHz): d 10.68 (s, 1H, OH), 7.62
(d, J = 2.3 Hz, 1H, ArH), 7.27 (dd, J = 8.4, 2.3 Hz, 1H, ArH), 6.90
(d, J = 8.4 Hz, 1H, ArH), 4.41 (q, J = 7.1 Hz, 2H, CH2), 4.12 (q,
J = 7.2 Hz, 2H, CH2), 2.53 (t, J = 8.0 Hz, 2H, CH2Ph), 2.28 (t,
J = 7.7 Hz, 2H, CH2COOEt), 1.58 (m, 4H, 2ꢂCH2), 1.42 (t, J = 7.1 Hz,
3H, CH3), 1.25 (t, J = 7.2 Hz, 3H, CH3), 1.27–1.33 (bm, 10H,
5ꢂCH2); 13C NMR (CDCl3, 100.6 MHz): 173.9, 170.2, 159.7, 135.9,
133.4, 128.9, 117.3, 112.2, 61.3, 60.1, 34.4, 31.6, 29.4, 29.3, 29.2,
29.1, 29.0, 24.9 14.2. ESI-MS m/z calcd for C21H32O5 364.2, found
m/z 387.2 [M + Na]+, 751.6 [2M + Na]+.
Small Molecule Synthesis
Preparation of 2a and 2b, general
Anhydrous aluminum chloride (7.2 g, 54.2 mmol) was added in
portions to a stirred and cooled (ꢁ15 ꢀC or below) solution of
ethyl salicylate (1; 5 g, 30.1 mmol) in dichloroethane (9.3 ml), then
sebacoyl chloride (6.4 mL, 29.9 mmol) or dodecanedioyl dichlor-
ide (8.0 g, 29.8 mmol) was added dropwise, and the mixture was
kept at ꢁ15 to ꢁ20 ꢀC for 3–4 h. The mixture was then allowed
to reach room temperature and was further stirred overnight.
Solvents were removed under vacuum, and the remaining or-
ange solid was transferred into a mixture of water (200 ml) and
DCM (60 ml). The aqueous solution was acidified to pH 3 with
concentrated hydrochloric acid. A white precipitate was formed
and filtered off. The aqueous solution was extracted with DCM
(2 ꢂ 30 ml), and the organic layer was dried over anhydrous
MgSO4. The solvent was evaporated, and the residue was purified
by flash column chromatography (PE/EA 10 : 1 to 1 : 1) to afford
2a or 2b as a colorless solid.
Ethyl 5-(110-ethoxycarbonylundecyl)-2-hydroxybenzoate (3b). Yield:
1
0.90 g, 76%, H NMR (CDCl3, 400 MHz): d 10.66 (s, 1H, OH), 7.62
(d, J = 2.3 Hz, 1H, ArH), 7.26 (dd, J = 8.5, 2.3Hz, 1H, ArH), 6.88 (d,
J = 8.5Hz, 1H, ArH), 4.40 (q, J = 7.1 Hz, 2H, CH2), 4.12 (q, J = 7.1 Hz,
2H, CH2), 2.53 (t, J = 8.0 Hz, 2H, CH2), 2.28 (t, J = 7.9Hz, 2H, CH2),
1.58 (m, 4H, 2ꢂCH2), 1.42 (t, J = 7.1Hz, 3H, CH3), 1.17–1.34 (bm,
17H, 7ꢂCH2 and CH3); 13C NMR (CDCl3, 100.6 MHz): d 173.9,
170.3, 159.7, 135.9, 133.4, 129.0, 117.3, 112.3, 61.3, 60.1, 34.9,34.3,
31.6,29.5, 29.4, 29.4, 29.2, 29.2, 29.1, 24.9, 14.2. ESI-MS m/z calcd
for C23H36O5 392.2, found m/z 415.2 [M + Na]+.
Preparation of 4a and 4b, general
Ethyl 5-(90-carboxy-10-oxononyl)-2-hydroxybenzoate (2a). Yield:
4.4 g, 50%, 1H NMR (CDCl3, 400 MHz): d 11.32 (s, 1H, OH), 8.49
(d, J = 2.3 Hz, 1H, ArH), 8.07 (dd, J = 8.8, 2.3 Hz, 1H, ArH), 7.02
(d, J = 8.8 Hz, 1H, ArH), 4.46 (q, J = 7.2 Hz, 2H, CH2), 2.91
(t, J = 7.4 Hz, 2H, CH2COPh), 2.35 (t, J = 7.6 Hz, 2H, CH2COOH),
1.72 (m, 2H, CH2), 1.63 (m, 2H, CH2), 1.45 (t, J = 7.2 Hz, 3H,
CH3), 1.27–1.4 (bm, 8H, 4ꢂCH2); 13C NMR (CDCl3, 100.6 MHz):
198.3, 180.0, 169.7, 165.1, 135.1, 130.9, 128.6, 117.7, 112.2, 61.9,
38.0, 33.9, 29.1, 29.0, 28.9, 24.5, 24.2, 14.1 (one aliphatic carbon
was overlapped). ESI-MS calcd for C19H26O6 350.1, found m/z
373.2 [M + Na]+, 349.4 [M ꢁ H]ꢁ.
Ethyl 5-(110-carboxy-10-oxoundecyl)-2-hydroxybenzoate (2b). Yield:
4.6g, 41%, 1H NMR (CDCl3, 400MHz): d 11.32 (s, 1H, OH), 8.50
(d, J= 2.3 Hz, 1H, ArH), 8.08 (dd, J= 8.8, 2.3Hz, 1H, ArH), 7.03
(d, J =8.8 Hz, 1H, ArH), 4.46 (q, J = 7.2Hz, 2H, CH2), 2.91
(t, J = 7.6Hz, 2H, CH2), 2.35 (t, J = 7.5Hz, 2H, CH2), 1.72 (m, 2H,
CH2), 1.63 (m, 2H, CH2), 1.45 (t, J = 7.2Hz, 3H, CH3), 1.23–1.39
(m, 12H, 6ꢂCH2). 13C NMR (CDCl3, 100.6MHz) d 198.3, 179.9,
170.0, 165.0, 135.3, 131.0, 128.8, 117.9, 111.9, 52.6, 38.1, 4.0, 29.4,
29.4, 29.3, 29.3, 24.6, 14.1. ESI-MS calcd for C21H30O6 , 378.2, found
m/z 401.2 [M + Na]+, 377.0 [M ꢁ H]ꢁ.
To a dried 100 ml round bottom flask was added 3a (0.58 g,
1.6 mmol) or 3b (0.63g, 1.60 mmol), sodium hydride (60% dispersion
in mineral oil, 0.16 g, 4.0mmol), followed by dry THF (24ml) in a
nitrogen atomosphere. The suspension was stirred and cooled in
ice for 30 min. Methoxyethoxymethyl chloride (0.45 ml, 4.0 mmol)
was added dropwise. After stirring for 2 h at 0ꢀC, the reaction
mixture was warmed to room temperature and stirred overnight.
The mixture was concentrated, and then a saturated aqueous solu-
tion of sodium bicarbonate was added. The resulting mixture was
extracted with DCM (3ꢂ 10 ml). The combined organic layers were
washed with H2O and brine, then dried with MgSO4, and concen-
trated. The residue was purified by flash chromatography (PE/EA
2 : 1) to afford 4a or 4b as a faint yellow oil.
Ethyl 5-(90-ethoxycarbonylnonyl)-2-[(200-methoxy)-ethoxymethyloxy]
benzoate (4a). Yield: 0.54 g, 76%. H NMR (CDCl3, 400 MHz): d
1
7.54 (d, J = 2.2 Hz, 1H, ArH), 7.21 (dd, J = 8.4, 2.2 Hz, 1H, ArH),
7.12 (d, J = 8.4 Hz, 1H, ArH), 5.29 (s, 2H, OCH2O), 4.33 (q, 2H,
CH2CH3), 4.10 (q, 2H, CH2CH3), 3.86 (m, 2H, OCH2CH2O), 3.54
(m, 2H, OCH2CH2O), 3.35 (s, 3H, CH3), 2.54 (t, J= 7.1 Hz, 2H, CH2Ph),
2.26 (t, J= 7.2Hz, 2H, CH2COOEt), 1.57 (m, 4H, 2ꢂCH2), 1.36
(t, J = 7.1 Hz, 3H, CH3), 1.23 (t, J = 7.2Hz, 3H, CH3), 1.23–1.32 (bm,
10H, 5ꢂCH2); 13C NMR (CDCl3, 100.6 MHz): d 173.7, 166.2, 154.5,
136.0, 133.0, 129.7, 121.4, 116.6, 94.2, 71.4, 67.7, 60.7, 60.0, 58.8,
34.8, 34.2, 31.3, 29.3, 29.2, 29.1, 29.0, 28.9, 24.8, 14.2, 14.1. ESI-MS
m/z calcd for C25H40O7 452.2, found m/z 475.0 [M + Na]+.
Preparation of 3a and 3b, general
To a 100 ml flask was added zinc dust (4 g, 61.5 mmol) and HgCl2
(77.6 mg, 0.29 mmol) in water (4 ml). The suspension was agitated
for 15 min, the liquid was then decanted, and the amalgamated
zinc was washed with water. Compound 2a (1.0 g, 2.9 mmol) or
2b (1.1 g, 3.0 mmol) was added, followed by concentrated HCl
(6 ml) and EtOH (6 ml). The mixture was refluxed with stirring
for 3–5 h, then cooled to room temperature, and DCM (40 ml)
was added. The aqueous layer was extracted with DCM
(2 ꢂ 40 ml), washed with water and brine (3ꢂ 20 ml), and dried with
MgSO4. The solvent was removed, and the residue was dissolved in
ethanol (99.5%, 20 ml). Then, thionyl chloride (0.8ml) was added
Ethyl 5-(110-ethoxycarbonylundecyl)-2-[(200-methoxy)-ethoxymethyloxy]
benzoate (4b). Yield: 0.54 g, 71%. Crude 1H NMR (CDCl3,
400 MHz): d 7.55 (d, J = 2.4 Hz, 1H), 7.23 (dd, J = 8.5, 2.4 Hz, 1H),
7.14 (d, J = 8.5 Hz, 1H), 5.30 (s, 2H), 4.34 (q, J = 7.1 Hz, 2H), 4.12
(q, J = 7.1 Hz, 2H), 3.90–3.85 (m, 2H), 3.58–3.53 (m, 2H), 3.37
(s, 3H), 2.58 (t, 2H), 2.24 (t, 2H), 1.66–1.50 (m, 4H), 1.38 (t, J=7.1Hz,
3H), 1.30–1.21 (m, 17H). ESI-MS m/z calcd for C27H44O7 480.3,
found m/z 503.2 [M + Na]+.
J. Pept. Sci. 2012; 18: 731–739 Copyright © 2012 European Peptide Society and John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jpepsci

SUN ET AL.
Preparation of 5a and 5b, general
3.32 (s, 3H, CH3), 2.59 (t, J= 7.6 Hz, 2H, CH2), 2.53 (t, J= 7.7 Hz, 2H,
CH2), 1.75 (m, 2H, CH2), 1.55 (m, 2H, CH2), 1.44–1.18 (bm, 10H,
5ꢂCH2); 13C NMR (CDCl3, 100.6 MHz): d 171.3, 163.0, 152.6, 142.2,
137.0, 132.8, 131.5, 128.9, 128.0, 127.7, 125.2, 122.4, 120.1, 114.4,
93.4, 71.3, 68.0, 59.0, 34.8, 34.3, 31.3, 29.3 (2 carbons), 29.1 (2
carbons), 29.0, 24.7. (3 quaternary carbons were not found). ESI-MS
calcd for C46H50N2O9 774.3, found m/z 797.4 [M + Na]+.
To a stirred solution of 4a (289 mg, 0.6 mmol) or 4b (306 mg,
0.6 mmol) in MeOH (2.6 ml) was added aqueous 1 M NaOH
(2.6 ml, 2.6 mmol). The resulting mixture was kept at 60 ꢀC for
4 h. After cooling, the organic solvent was removed under
reduced pressure, and the aqueous layer was washed with EA
(3ꢂ 20ml). The resulting solution was then acidified with aqueous
10% citric acid to pH 4.5 and then extracted with EA (3 ꢂ 5 ml).
The combined organic extracts were washed with H2O
(3 ꢂ 5 ml), dried over anhydrous MgSO4, and concentrated to
afford 5a or 5b as a colorless liquid.
5-[110-(4000-Nitrophenyloxycarbonyl)-undecyl]-2-[(200-methoxy)-
ethoxymethyloxy]-N-trityloxybenzamide (7b). Yield: 56 mg, 15%
1
(total yield of final two steps). H NMR (CDCl3, 400 MHz): d 9.70
(s, 1H, NH), 8.26 (m, 2H, p-NO2ArH), 7.87 (d, J= 2.4 Hz, 1H, ArH),
7.56 (bm, 5H, Trityl and p-NO2ArH), 7.38–7.22 (m, 12H, Trityl), 7.14
(dd, J= 8.6, 2.4 Hz, 1H, ArH), 7.03 (d, J= 8.6 Hz, 1H, ArH), 4.89
(s, 2H, OCH2O), 3.44 (multiplet, 2H, OCH2CH2O), 3.42 (multiplet,
2H, OCH2CH2O), 3.33 (s, 3H, CH3), 2.60 (t, J= 7.8 Hz, 2H, CH2),
2.53 (t, J= 7.6 Hz, 2H, CH2), 1.76 (m, 2H, CH2), 1.55 (m, 2H, CH2),
1.46–1.18 (bm, 14H, 7ꢂCH2); 13C NMR (CDCl3, 100.6MHz): d 171.3,
163.0, 155.5, 152.6, 142.2, 137.0, 132.6, 131.5, 128.9, 128.0, 127.6,
125.2, 122.4, 120.1, 114.3, 93.4, 92.6, (95.4), 71.3, 68.0, 59.0,
34.8, 34.3, 31.3, 29.5, 29.4, 29.3, 29.1, 29.0, 24.7. ESI-MS calcd for
C48H54N2O9 802.3, found m/z 825.4 [M + Na]+.
5-(90-Carboxynonyl)-2-[(200-methoxy)-ethoxymethyloxy]benzoic acid
(5a). Yield: 0.23 g, 89%. 1H NMR (CDCl3, 400 MHz): d 7.94
(d, J = 2.1 Hz, 1H, ArH), 7.33 (dd, J = 8.4, 2.1 Hz, 1H, ArH), 7.19
(d, J = 8.4Hz, 1H, ArH), 5.46 (s, 2H, OCH2O), 3.89 (m, 2H, OCH2CH2O),
3.57 (m, 2H, OCH2CH2O), 3.36 (s, 3H, CH3), 2.58 (t, J= 7.9 Hz, 2H,
CH2), 2.34 (t, J = 7.5Hz, 2H, CH2),1.60 (m, 4H, 2ꢂCH2), 1.21–1.37
(bm, 10H, 5ꢂCH2); 13C NMR (CDCl3, 100.6 MHz): d 179.5, 166.5,
154.2, 137.5, 134.9, 132.9, 118.0, 115.0, 94.7, 71.4, 68.9, 59.0, 34.8,
34.0, 31.3, 29.3, 29.2, 29.1, 29.0, 28.9, 24.6. ESI-MS m/z calcd for
C21H32O7 396.2, found m/z 419.2 [M + Na]+.
5-(110-Carboxyundecyl)-2-[(200-methoxy)-ethoxymethyloxy]benzoic
acid (5b). This product was not purified but was directly used
in the next step. Crude H NMR (CDCl3, 400 MHz): d 10.42 (bs,
Polypeptide Synthesis
1
The polypeptides were synthesized on a 433A peptide synthe-
sizer (Applied Biosystems, Carlsbad, CA, USA) using standard
Fmoc chemistry with HBTU (Iris Biotech GmbH, Marktredwitz,
Germany) and DIPEA (Iris Biotech GmbH) as activating agents.
Fmoc deprotection of the amino terminal was performed using
20% piperidine in NMP. The synthesis was performed on a
0.1 mmol scale using Fmoc-PAL-polyethyleneglycol-polystyrene
resin (Fmoc-PAL-PEG-PS, Applied Biosystems) with a substitution
level of 0.19 mmol/g. A 10-fold excess of amino acid was used in
each coupling, and activation was with a mixture of HBTU (0.5M in
DMF) and DIPEA (2 M in NMP). The side chains of the amino acids (Iris
Biotech GmbH) were protected by base-stable groups: tert-butyl es-
ter (Asp and Glu), Trt (His, Asn, and Gln), 2,2,4,6,7-pentamethyldi-
hydrobenzofuran-5-sulfonyl (Arg), and acetamidomethyl (Cys).
For the site-selective incorporation of a fluorescent probe and
ligands, the side chains of lysine residues were N-protected by an
Alloc group, which can be selectively deprotected by treatment
with tetrakis(triphenylphosphine) palladium(0) (Pd(PPh3)4), a
Boc group, which can be deprotected by TFA, or an Mtt group,
which can be deprotected by a mixture of TFA, triisopropylsilane
(TIS), and DCM (v/v/v 1 : 1 : 500). The N-terminal of the polypep-
tide was capped manually with a mixture of acetic anhydride,
DIPEA, and NMP (v/v/v 10 : 5 : 85) for 20 min. After complete
machine synthesis, the resin was rinsed with DMF and DCM
and dried in a desiccator under vacuum.
Before the cleavage of the peptide from the resin, the Alloc-
protected lysine was deprotected by three equivalents of
tetrakis(triphenylphosphine)palladium(0) in a mixture of acetic
acid, N-methylmorpholine, and chloroform (v/v/v 2 : 1 : 17; 10 ml/g
of resin) at room temperature under N2 for 2.5 h. The resin was
then sequentially washed with 30 mM DIPEA in DMF, 20 mM
diethyldithiocarbamic acid in DMF, DMF, and DCM and then des-
iccated. For the chromophore coupling, the resin was mixed with
4 equiv of 7-methoxycoumarin-3-carboxylic acid, 4 equiv of
HBTU, and 6 equiv of DIPEA in DMF (4ml/g of resin) and was
gently shaken for 4 h at room temperature in a mini-rocker. Alter-
natively, the resin was mixed with 4 equiv of dansyl chloride and
1H, COOH), 7.91 (d, J = 2.3 Hz, 1H, ArH), 7.31 (dd, J = 8.5, 2.3 Hz,
1H, ArH), 7.18 (d, J = 8.5 Hz, 1H, ArH), 5.44 (s, 2H, OCH2O), 3.88
(ddd, J = 2.7 Hz, 1.8 Hz, 4.5 Hz, 2H, OCH2CH2O), 3.56 (ddd,
J = 2.7 Hz, 1.8 Hz, 4.5 Hz, 2H, OCH2CH2O), 3.35 (s, 3H, CH3), 2.57
(t, J = 7.8 Hz, 2H, CH2), 2.34 (t, J = 7.6 Hz, 2H, CH2),1.59 (m, 4H,
2ꢂCH2), 1.15–1.36 (bm, 14H, 7ꢂCH2). ESI-MS m/z calcd for
C
23H36O7 424.2, found m/z 447.2 [M + Na]+.
Preparation of 7a and 7b, general
Dry DCM (33 mL) was added to 5a (189 mg, 0.47 mmol) or
5b (203 mg, 0.47 mmol), 4-nitrophenol (67 mg, 0.47 mmol),
N,N-dicyclohexylcarbodiimide (99 mg, 0.47 mmol), and 4-N,N-
dimethylaminopyridine (5.8 mg, 0.096 mmol) under a nitrogen
atmosphere. The reaction mixture was stirred overnight, washed
with brine and water, dried with MgSO4, and concentrated. The res-
idue was purified by column chromatography (DCM/methanol
30 : 1) to afford a mixture of 4-nitrophenyl monoesters (aliphatic
and aromatic), which was used in the next step without further pu-
rification. To an ice-cooled solution of this mixture in DCM (15 ml)
under an atmosphere of nitrogen were added, consecutively,
hydroxylbenzotriazole (68 mg, 0.503 mmol), diisopropylcarbodii-
mide (DIPCI; 77.9 ml, 0.503 mmol), O-tritylhydroxylamine (115mg,
0.419mmol), and TEA (23.4ml, 0.168 mmol). The resulting mixture
was then stirred for 15 h at room temperature, diluted with
DCM (15 ml), washed with aqueous 1% citric acid solution
(3 ꢂ 15 ml), saturated aqueous NaHCO3 (3 ꢂ 15 ml), and water
(3 ꢂ 15 mL), dried over anhydrous MgSO4, and concentrated.
The residue was purified by flash chromatography (DCM/MeOH
35 : 1) to afford 7a or 7b as a colorless liquid.
5-[90-(4000-Nitrophenyloxycarbonyl)-nonyl]-2-[(200-methoxy)-
ethoxymethyloxy]-N-trityloxybenzamide (7a). Yield: 102 mg, 28%
1
(total yield of final two steps). H NMR (CDCl3, 400MHz): d 9.71 (s,
1H, NH), 8.26 (dm, J= 9.2 Hz, 2H, p-NO2ArH), 7.87 (d, J= 2.0 Hz, 1H,
ArH), 7.56 (m, 6H, Trityl), 7.39–7.19 (bm, 11H, Trityl and p-NO2ArH),
7.14 (dd, J= 8.6, 2.0 Hz, 1H, ArH), 7.03 (d, J= 8.6 Hz, 1H, ArH), 4.89
(s, 2H, OCH2O), 3.43 (m, 2H, OCH2CH2O), 3.42 (m, 2H, OCH2CH2O),
wileyonlinelibrary.com/journal/jpepsci Copyright © 2012 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2012; 18: 731–739

POLYPEPTIDE-SALICYLHYDROXAMIC ACID CONJUGATE BINDS MPO
8 equiv of DIPEA in DMF (4 ml/g of resin) and gently shaken for
2 h. The resin was then washed with DMF and DCM and desic-
cated. The peptide was cleaved from the resin using a mixture
of TFA, TIS, and water (v/v/v 95 : 2.5 : 2.5; 10 ml/g of resin) for 2 h
at room temperature. The mixture was filtered, and TFA was then
evaporated from the filtrate using a stream of nitrogen. The
polypeptide was precipitated by the addition of cold ether,
centrifuged, redissolved in water, and lyophilized. The crude
peptides were purified by RP-HPLC and characterized by
MALDI-TOF MS.
for 4-series 11-carbon linked conjugates, and 0.5–32 mM for 1-, 2-, 3-
series 9-carbon linked conjugates using 3 min contact time. For min-
imum sample dispersion, the samples were injected at a flow rate of
50 ml/min. After 10 min dissociation, the surface was regenerated by
30-s pulse injection of 10 mM glycine, pH 3.0. All injections were
serial and first passed over the deactivated dextran surface and
then over the immobilized surfaces. Blank injections were in-
cluded for each measurement series.
Competition Experiments
Human myeloperoxidase isoform C was immobilized to the sensor
chip in a final immobilization level of 15 491 RU. The detailed proce-
dure was as described previously. A total of 0.5, 5, and 50 mM of
4C37L34C11SHA in running buffer were prepared from 0.731 mM
4C37L34C11SHA DMSO solution. We used 6.52 mM SHA in running
buffer for the preparation of the different concentrations. In these
binding assays, to reach the saturation of binding, the injection
time was 8 min in a flow rate of 20 ml/min, and after 10min dissoci-
ation, the surface was regenerated by dissociation in running buffer
for 20 min. A series of concentration of SHA (0, 0.125, 0.25, 0.5, 1, 2,
and 4 mM) and 4C37L34C11SHA (0, 5, 10, 20, 40, 80, 160, and
320 nM) were flowed over the immobilized MPO C surface as posi-
tive controls. The mixtures of 100 nM 4C37L34C11SHA with a series
of concentration of SHA were injected. The sensograms were
obtained by subtraction of individual SHA concentration responses
and blank injections. The experiments were carried out in duplicate.
Conjugation Procedure
The ligands were conjugated to the free lysyl amino group of the
42-residue polypeptides by the addition of 2.5 equiv of 7a or 7b
in DMSO stock solution (50 mM) to a solution of the polypeptide
in DMSO (2 mM) in the presence of 30 equiv of DIPEA or TEA.
The reaction was monitored by analytical HPLC, and the fractions
were analyzed by MALDI-TOF MS. The functionalized polypeptide
was precipitated by the addition of cold diethyl ether and centri-
fuged. A mixture of TFA/H2O/TIS (95 : 2.5 : 2.5) was slowly added
to the precipitated polypeptide to make a 2 mM polypeptide solu-
tion, which was kept at room temperature for 2–3 h, after which
the TFA was evaporated by a stream of nitrogen. The depro-
tected polypeptide was precipitated by the addition of cold ether
and centrifuged. The pellet was purified by semi-preparative
HPLC, and the appropriate fractions were lyophilized to afford
the pure functionalized polypeptides, which were characterized by
MALDI-TOF MS (see supporting information) in yields of 10–65%.
Results and Discussion
Screening of Polypeptide Conjugates with Immobilized MPO
Binder Design
Human myeloperoxidase (MW = 145 kDa) or human myeloperox-
idase isoform C (MW = 146 kDa) was obtained commercially
(Biodesign Inc., Täby, Sweden, and Lee Biosolution, Inc. St Louis,
MO, USA, respectively). The 42-residue polypeptide-SHA conju-
gates were synthesized as described. Human myeloperoxidase
(1 mg/ml) in 50 mM sodium acetate, 0.1 M NaCl (pH 6.0) and mye-
loperoxidase isoform C (1.9 mg/ml) in sodium acetate buffer
(pH 6.0) were diluted in de-ionized water or 10 mM sodium ace-
tate (pH 6.0) to give a concentration of about 50 mg/ml. The
MPO was covalently immobilized to the sensor chip surface by
amine coupling with PBS (10 mM phosphate, 2.7 mM KCl, 137 mM
NaCl, and 0.005% P20, pH 7.4) (GE Healthcare Bio-Sciences,
Uppsala, Sweden) as a running buffer. The CM5 sensor chip sur-
face was activated for 7 min by injecting a solution of EDC/N-
hydroxysuccinimide (200 mM EDC/50 mM N-hydroxysuccinimide)
(GE Healthcare Bio-Sciences). The MPO was injected for 5–7 min
at a flow rate of 3–5 ml/min over the activated surface and fol-
lowed by 7-min pulse of 1 M ethanolamine at pH 8.5 to deactivate
the remaining active ester groups. The final immobilization levels
were between 11 000 and 28 000 RU. The interaction experiments
between MPO and binders and the competitive experiments
were carried out using a Biacore 2000 instrument (GE Healthcare
Bio-Sciences), equilibrated at 25 ꢀC. The CM5 sensor chip (Research
grade, Biacore) and reagents were from GE Healthcare Bio-
Sciences. For direct interaction studies between immobilized
MPO and polypeptide conjugates, PBS containing 1–3% DMSO
and 0.005% P20 was used as a running buffer. The polypeptide
conjugates were diluted from a 50 uM stock solution in the running
buffer and injected over the immobilized enzyme in concentration
series of 5–640 nM for 4-series 9-carbon linked conjugates, 5–320 nM
The binder concept has been described in detail previously [9–11].
In short, it is based on the idea that if a small molecule that is
known to bind to a protein, albeit with poor affinity and selectivity,
is conjugated to a polypeptide with hydrophobic as well as
charged residues, a dramatic increase in affinity and selectivity is
achieved although the polypeptide itself binds to the protein with
poor affinity. Because the entropic penalty has been ‘paid’ by the
small molecule, the polypeptide has to contribute very little binding
energy to increase the overall affinity by, say, four orders of mag-
nitude. The interactions between two leucine side chains and hy-
drophobic patches on the protein surface would be enough to
provide the required amount of binding energy. Alternatively,
charge–charge interactions and hydrogen bonding could contribute
similar amounts. The polypeptides have been shown to fold into
helix-loop-helix motifs that dimerize to form four-helix bundles at
micromolar concentrations but exist as unordered monomers at
low nanomolar concentrations. The polypeptides are therefore
capable of adopting a large number of conformations capable of
interacting with the protein and according to the design hypothe-
sis bind by adapting to the protein surface to form the binding
interaction. Because only a few contacts between polypeptide
and protein are needed, the number of possible binding modes
is very large; thus, the small set of polypeptides can interact
with a very large number of proteins, possibly all. The concept is
unlike that of binder molecules prepared by molecular biology
technologies that form pre-organized structures that are shape
complimentary to the target protein. The set of 16 polypeptide
sequences was described in detail previously (Figure 2) [9–11].
The sequences were designed to have some propensity for forming
helix-loop-helix motifs and were varied with regard to the number
J. Pept. Sci. 2012; 18: 731–739 Copyright © 2012 European Peptide Society and John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jpepsci

SUN ET AL.
Seq. name N-Terminus
C-terminus Charge
1C15L8
AcNEADLEAKIRHLAEKLEARGPEDAEQLAEQLARAFEAFARAG-OH
AcNAADLEAAIKHLAEALKERGPEDCEQLAEQLARAFEAFARAG-OH
-7
1C10L17
1C25L22
1C37L34
2C15L8
2C10L17
2C25L22
2C37L34
3C15L8
3C10L17
3C25L22
3C37L34
4C15L8
4C10L17
4C25L22
4C37L34
-7
-6
-5
-4
-4
-4
-3
-1
-1
-1
-1
+2
+2
+2
+2
AcNEADLEAAIRHLAEALEARGPKDAKQLAEQLARAFEAFERAG-OH
AcNEADLEAAIRHLAERLEARGPADAAQLAEQLAAKFEKFARAG-OH
AcNEADLEAKIRHLAEKLAARGPVDCAQLAEQLARAFEAFARAG-OH
AcNAADLEAAIKHLAEALKARGPVDCAQLAEQLARAFEAFARAG-OH
AcNEADLEAAIRHLAEALAARGPKDCKQLAEQLARAFEAFARAG-OH
AcNAADLEAAIRHLAERLAARGPVDCAQLAEQLAAKFEKFARAG-OH
AcNAADJEAKIRHLAEKJAARGPVDCAQJAEQLARRFEAFARAG-NH2
AcNAADJEARIKHLAERJKARGPVDCAQJAEQLARAFEAFARAG-NH2
AcNAADJEAAIRHLAERJAARGPKDCKQJAEQLARAFEAFARAG-NH2
AcNAADJEAAIRHLAERJAARGPVDCAQJAEQLARKFEKFARAG-NH2
AcNAADJEAKIRHLREKJAARGPRDCAQJAEQLARRFERFARAG-NH2
AcNAADJEARIKHLRERJKARGPRDCAQJAEQLARAFERFARAG-NH2
AcNAADJEARIRHLRERJAARGPKDCKQJAEQLARAFERFARAG-NH2
AcNAADJEARIRHLRERJAARGPRDCAQJAEQLARKFEKFARAG-NH2
Figure 2. The 42-residue polypeptides used for binder development in the one-letter code (J is norleucine, Nle). The total charge, the site of incorpo-
ration of SHA, and the chromophore site are the main variables. The lysine residue K shown in green is the attachment site of the chromophore and that
shown in red is the intended site of SHA conjugation. All N-terminals were acetylated, and all C residues were Acm-protected. The C-terminals were
either carboxylates (1- and 2-charge types) or amides (3- and 4-charge types). In the sequence names, the first number indicates the charge type of
the peptide, e.g. all peptides beginning with ‘4’ carry a total charge of +2 (at pH 7 and when amidated at both lysines). The second and third numbers
indicate, respectively, the lysine carrying the chromophore and that carrying a free side chain amino group, e.g. 4C37L34 has a 7-methoxycoumarinyl
chromophore group at the side chain of lysine-37 and a free amino group at the side chain of lysine-34 (available for conjugation). Sometimes, for future
purposes, the alanine-41 residue was replaced with an e-N-trifluoroacetylated lysine residue (see Supporting Information for details and characterizations of
peptides and conjugates).
Synthesis
and position of charged residues and also with regard to the site of
incorporation of the small molecules. The hydrophobic residues
Leu, Ile, and Phe remain the same throughout the sequences.
In the crystal structure of MPO, the heme group is located in a
crevice, about 15 Å in depth, with access to the solvent via an open
channel, approximately 10 Å in diameter, and the hydroxyl oxygen
and NH nitrogen of SHA are 3.1 and 3.5 Å away from the iron,
respectively [16]. SHA is a known substrate mimic inhibitor of ΜPO
with a KD of 2 mM in sodium phosphate buffer at pH6.0 [12]. It was
therefore selected for conjugation to the polypeptide. On the basis
of the SHA location and the depth of the heme cavity, 9 and 11
carbon aliphatic spacers were selected for MPO (Figure 1).
The strategy used for the synthesis of the polypeptide conjugates
was to first synthesize the protected ligands 7a and 7b, then to con-
jugate them in protected form to the side chain amino groups of
polypeptides (obtained by Fmoc-type solid-phase synthesis), and fi-
nally to subject the conjugates to acidic deprotection conditions.
Acid-labile methoxyethoxymethyl (MEM)- and Trt-protecting groups
were used for protection of the ligands. The synthesis of the pro-
tected ligands 7a and 7b from ethyl salicylate 1 (14% and 3.5%
overall yields, respectively) is shown in Scheme 1. Friedel–Crafts
acylation with sebacoyl or didodecanedioyl dichloride afforded
compounds 2a and 2b in 50% and 41% yields, respectively [13–15].
OH
OH
OH
a
b
HO
O
EtO
n
O
O
Et
O
n
O
O
Et
O
O
O
3a (n = 8)
3b (n=10)
2a (n = 8)
2b (n=10)
1
O
O
O
MEM
OH
MEM
c
d
HO
O
O
n
n
O
O
O
5a (n = 8)
5b (n=10)
4a (n = 8)
4b (n=10)
MEM
O
MEM
O
e
f
H
N
O
O
O
OH
Trt
n
O
n
O
O
O
O2N
O2N
7a (n = 8)
7b (n=10)
6a (n = 8)
6b (n=10)
Scheme 1. Synthesis of compounds 7a and 7b. Conditions: (a) AlCl3, ClCO(CH2)8 or 10COCl, dichloroethane, ꢁ20 to ꢁ15 ꢀC to rt; (b) i: Zn/Hg, 6N HCl,
ethanol, reflux, ii: HCl/abs. ethanol; (c) NaH, THF, MEMCl, 0 ꢀC; (d) 1N NaOH, H2O/MeOH (1 : 1), 60 ꢀC; (e) 4-nitrophenol, DCC, DMAP, dry DCM, rt, o.v.
(f) HOBt, DIPCI, NH2OTrt, NEt3, dry DCM, r.t., o.v.
wileyonlinelibrary.com/journal/jpepsci Copyright © 2012 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2012; 18: 731–739

POLYPEPTIDE-SALICYLHYDROXAMIC ACID CONJUGATE BINDS MPO
Clemmensen reduction [13] and subsequent esterification of
2a and 2b gave compounds 3a and 3b (80% and 76%, respec-
tively). The phenolic hydroxyl group was protected by a MEM
group to give derivatives 4a and 4b (76% and 71%), followed
by saponification to give diacids 5a and 5b. DCC-promoted
partial esterification of 5a and 5b gave predominantly the ali-
phatic monoesters 6a and 6b. The site of esterification in 6a
was determined by comparing the H-NMR spectrum of 6a to that
of the starting material 5a (see the Supporting Information). The
CH2COO methylene proton triplet was significantly shifted
downfield when going from the dicarboxylic acid 5a (2.33 ppm)
to 6a (2.59 ppm), indicating that the ester group was on the
aliphatic chain. Analogous shift changes were observed for 6b.
Finally, treatment of the monoesters with N-tritylhydroxylamine,
DIPCI, and HOBt together with TEA, afforded compounds 7a
and 7b (28% and 15%). That Trt-hydroxylamidation had indeed
taken place at the aromatic carboxyl group was again evident
from the H-NMR spectra of 7a and 7b (see the Supporting
Information), which showed significant chemical shift difference
for the MEM methylene protons between 6a and 7a, indicating
that trityloxyamidation had occurred at the aromatic carboxylic
acid. Finally, 7a and 7b were successfully conjugated to the
specified (Figure 2) free lysine side chain amino group of the
various peptides in DMSO solution to afford peptide conjugates.
The MEM- and Trt-protecting groups of the conjugates were
deprotected simultaneously by a cocktail mixture of TFA, TIS,
and water in a ratio of 95 : 2.5 : 2.5 to afford the final peptide
conjugates, which were purified by HPLC and characterized by
MALDI-TOF MS. A total of 20 peptide conjugates were prepared
in a yield of 10–65% (for structures and characterization, see
Figure 2 and the Supporting Information).
Evaluation of Binder Performance
The interactions between the 20 polypeptide-SHA conjugates
and MPO were investigated by SPR biosensor analysis (Figures 3
and 4). MPO was immobilized on a CM-5 chip (Biacore) using
standard protocols, and each one of the binder candidates were
allowed to flow over the immobilized protein for evaluation of
binding. Within the series of binder candidates equipped with
the shorter 9-carbon atom spacer, the highest uptake was
observed for the four polypeptide sequences that were highly
positively charged (four series), and the same phenomenon was
found for the series of binder candidates with the longer 11-carbon
Figure 3. Panel of sensograms showing screening of MPO conjugates carrying 9-carbon atom aliphatic spacer for binding to immobilized MPO in PBS
buffer containing 0–1% DMSO and 0.005% P20 at pH 7.4. The concentrations of polypeptide conjugates in the top 12 panels were 0.5, 1, 2, 4, 8, 16, and
32 uM, and the concentrations used for the bottom four panels were 5, 10, 20, 40, 80, 160, 320, and 640 nM. Axes are drawn to scale for direct comparison
of uptake between binders. The injection time was 3 min, and the clearance time 10 min. This figure is available in colour online at wileyonlinelibrary.
com/journal/jpepsci
J. Pept. Sci. 2012; 18: 731–739 Copyright © 2012 European Peptide Society and John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jpepsci

SUN ET AL.
Figure 4. Panel of sensogram for the conjugates with 11-carbon atom aliphatic spacer that bind the strongest to immobilized MPO isoform C in PBS
buffer containing 3% DMSO and 0.005% P20 at pH 7.4. Axes are drawn to scale for direct comparison of uptake between binders. The injection time was
3 min, and the clearance time 10 min. This figure is available in colour online at wileyonlinelibrary.com/journal/jpepsci
atom spacer. The weaker binders were not found to bind signifi-
cantly at low micromolar concentrations, whereas the strongest
binders showed strong uptake even at low nanomolar concentra-
tions. The polypeptide conjugate 4D37L34C9SHA was found quali-
tatively to bind the most strongly to immobilized MPO among
the 9-carbon spacer conjugates (Figure 3), and the analogous
4C37L34C11SHA conjugate bound strongest among the 11-carbon
spacer ones (Figure 4). The dissociation constant KD of SHA was
reported previously to be 2mM in sodium phosphate buffer
at pH 6.0 [12]. The sensogram panel of four series binders with
11-carbon spacers is shown in Figure 4. Strong uptake at low nano-
molar concentrations suggest higher affinity than that of SHA, but
as has been reported previously, binding does not follow single
exponential kinetics, and determination of dissociation constants
by SPR is associated with large errors. We therefore do not report
dissociation constants from simple SPR analysis.
To demonstrate that the polypeptide conjugate binds MPO by
anchoring the small molecule SHA to its binding site, a competi-
tion experiment was carried out using SPR analysis where a
mixture of 100 nM of the strong binder 4C37L34C11SHA and
SHA over a range of concentrations from 0 to 4 mM was used to
probe uptake by freshly immobilized MPO on a sensor chip
(Figure 5; see also Figure S2 in the Supporting Information).
Although lower and higher SHA concentrations would have been
needed to produce a complete S-shaped inhibition curve (the
solubility of SHA was the limiting factor at the high end), the
diagram clearly shows that increasing concentrations of SHA
reduced the binding of 4C37L34C11SHA to the chip, with an
approximate 50% inhibition of binding at around 0.25–0.5 mM
concentration. The results show that the interaction between
the polypeptide conjugate and MPO is driven by the conjugated
SHA and that the polypeptide provides further interactions to
boost the affinity of the binder molecule. A rough estimate of
the dissociation constant KD for the MPO-binder complex can
be obtained from the competition experiments. At 50% inhibi-
tion of binding, the concentrations of the binder–MPO complex
and the SHA–MPO complex are approximately equal, and the
ratio of concentrations is directly related to the ratio of dissocia-
tion constants. We observe 50% inhibition at around 0.25–0.5 mM
of SHA and 100 nM of 4C37L34C11SHA, indicating a 2500- to
5000-fold higher concentration of SHA in comparison with
4C37L34C11SHA and therefore a 2500- to 5000-fold higher
affinity of the polypeptide conjugate compared with that of
SHA. The KD of SHA has been reported [12] to be 2 mM, and an
estimated dissociation constant of 4C37L34C11SHA is therefore
400–800 pM. Although this estimate is a rough one, it neverthe-
less shows that conjugation to the polypeptide provides a
binder molecule that binds to the site of the small molecule
but with dramatically enhanced affinity.
Mode of Binding
According to the crystal structure of SHA complexed with MPO, the
aromatic ring of SHA binds to a hydrophobic region at the entrance
to the distal heme pocket, and the hydroxamic acid moiety is
hydrogen bonded to both the distal histidine and the adjacent
Glu or Arg but is not coordinated to heme iron [16]. The structural
data suggested that the spacer should be introduced on the
aromatic ring in the para position relative to the phenolic hydroxy
group. Although the crystal structure suggested where to introduce
the spacer, the required spacer length could not be estimated. Two
spacer lengths were therefore evaluated, and the longer one was
found to provide slightly higher affinity (data not shown).
Myeloperoxidase is a cationic basic protein, with an isoelectric
point of over 9.2. The total charge of MPO is positive. The results
from the SPR analysis showed that positively charged polypep-
tide conjugates bound strongly to MPO, whereas negatively
charged ones did not. These facts seem to be contradictory but
instead illustrate that the overall charge of the protein is not
critical for strong interactions between polypeptide and protein.
Instead, the interactions are local and most likely depend on
Figure 5. Inhibition of 4C37L34C11SHA binding to immobilized MPO
isoform C by 0.0625, 0.125, 0.25, 0.5, 1, 2, and 4 mM concentrations of
salicylhydroxamic acid (SHA). The maximum solubility of SHA in the buffer
was around 6.5 mM, therefore higher concentrations than 4 mM were not
used. The data were obtained from the right sensogram in Figure S2
(supplementary information).
wileyonlinelibrary.com/journal/jpepsci Copyright © 2012 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2012; 18: 731–739

POLYPEPTIDE-SALICYLHYDROXAMIC ACID CONJUGATE BINDS MPO
5 Heslop CL, Frohlich JJ, Hill JS. Myeloperoxidase and C-reactive pro-
tein have combined utility for long-term prediction of cardiovascular
mortality after coronary angiography. J. Am. Coll. Cardiol. 2010; 55:
1102–1109.
hydrophobic patches on the surface or hydrophobic patches that
are exposed as a result of conformational changes.
6 Hoy A, Trégouët D, Leininger-Muller B, Poirier O, Maurice M, Sass C,
Siest G, Tiret L, Visvikis S. Serum myeloperoxidase concentration in a
healthy population: biological variations, familial resemblance and
new genetic polymorphisms. Eur. J. Hum. Genet. 2001; 10: 780–6.
7 Vita JA, Brennan M-L, Gokce N, Mann SA, Goormastic M, Shishehbor
MH, Penn MS, Keaney JF, Hazen SL. Serum myeloperoxidase levels
independently predict endothelial dysfunction in humans. Circulation
2004; 110: 1134–1139.
8 Brennan M-L, Penn MS, Van Lente F, Nambi V, Shishehbor MH, Aviles
RJ, Marlene Goormastic M, Pepoy ML, McErlean ES, Topol EJ, Nissen
SE, Hazen SL. Prognostic value of myeloperoxidase in patients with
chest pain. N. Engl. J. Med. 2003; 349: 1595–1604.
9 Tegler LT, Nonglaton G, Büttner F, Caldwell K, Christopeit T,
Danielson UH, Fromell K, Gossas T, Larsson A, Longati P, Norberg T,
Ramapanicker R, Rydberg J, Baltzer L. Powerful protein binders from
designed polypeptides and small organic molecules-A general con-
cept for protein recognition. Angew. Chem. Int. Ed Engl. 2011; 50:
1823–1827.
10 Baltzer L. Crossing borders to bind proteins-a new concept in protein
recognition based on the conjugation of small organic molecules or
short peptides to polypeptides from a designed set. Anal. Bioanal.
Chem. 2011; 400: 1653–1664.
11 Slosarczyk AT, Baltzer L. The molecular recognition of phosphorylated
proteins by designed polypeptides conjugated to a small molecule
that binds phosphate. Org. Biomol. Chem. 2011; 9: 7697–7704.
12 Ikeda-Saito M, Shelley DA, Linda Lu L, Booth KS, Caughey WS, Kimura
S. Salicylhydroxamic acid inhibits myeloperoxidase activity. J. Biol.
Chem. 1991; 266: 3611–3616.
13 Aultz DE, Mcfadden AR. Dibenz[b,e]oxepinalkanoic acids as nonsteroidal
antiinflammatory agents. 3. w-(6,ll-Dihydro-1 1-oxodibenz[b,e]oxepin-
2-yl)alkanoic acids. J. Med. Chem. 1977; 20: 1499–1501.
Conclusion
The active esters of protected SHA ligands were successfully syn-
thesized, and the ligands successfully conjugated to polypeptides
to afford 20 conjugates for MPO recognition and binding. The inter-
actions between polypeptide-SHA conjugates and MPO were eval-
uated by SPR, and competition experiments indicated that the SHA
group of the strong binder 4C37L34C11SHA bound to the active
site of MPO. The polypeptide provided further contacts with the
protein to form a binder molecule with an affinity for MPO that
was more than four orders of magnitude higher than that of SHA.
The reported binder molecule offers a starting point for the devel-
opment of a robust point of care test for a cardiovascular risk that
does not rely on antibodies. The combination with a previously de-
veloped synthetic binder for CRP provides an opportunity for an
improved clinical diagnostic test designed to address simulta-
neously two biomarkers that have been suggested to provide an
improved level of precision in predicting risk for heart infarction.
Acknowledgement
We would like to thank the Swedish Research Council Formas for
the financial support to Prof. Lars Baltzer.
References
1 Fact sheet Nꢀ317, World Health Organization 2011.
14 Papa D, Schwenk E, Hankin H. Friedel-Crafts synthesis of aroyl aliphatic
acids. J. Am. Chem. Soc. 1947; 69: 3018–3022.
2 Nauseef WM, Malech HL. Analysis of the peptide subunits of human
neutrophil myeloperoxidase. Blood 1986; 67: 1504–1507.
3 Arnhold J. Properties, functions, and secretion of human myeloperox-
idase. Biochemistry (Moscow) 2004; 69: 4–9.
4 Nicholls SJ, Hazen SL. Myeloperoxidase and cardiovascular disease.
Arterioscler. Thromb. Vasc. Biol. 2005; 25: 1102–1111.
15 Cox EH. The mechanism and application of the Fries reaction. J. Am.
Chem. Soc. 1930; 52: 352–358.
16 Davey CA, Fenna RE. 2.3 A resolution X-ray crystal structure of the
bisubstrate analogue inhibitor salicylhydroxamic acid bound to
human myeloperoxidase: a model for a prereaction complex with
hydrogen peroxide. Biochemistry 1996; 35: 10967–10973.
J. Pept. Sci. 2012; 18: 731–739 Copyright © 2012 European Peptide Society and John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jpepsci