Neocarzinostatin-based hybrid biocatalysts with a RNase like activity

Article abstract of DOI:10.1016/j.bmc.2014.05.063

A new zinc(II)-cofactor coupled to a testosterone anchor, zinc(II)-N,N-bis(2-pyridylmethyl)-1,3-diamino-propa-2-ol-N′(17′-succinimidyltestosterone) (Zn-Testo-BisPyPol) 1-Zn has been synthesized and fully characterized. It has been further associated with a neocarzinostatin variant, NCS-3.24, to generate a new artificial metalloenzyme following the so-called 'Trojan horse' strategy. This new 1-Zn-NCS-3.24 biocatalyst showed an interesting catalytic activity as it was found able to catalyze the hydrolysis of the RNA model HPNP with a good catalytic efficiency (k_cat/K_M = 13.6 M^-1 s^-1 at pH 7) that places it among the best artificial catalysts for this reaction. Molecular modeling studies showed that a synergy between the binding of the steroid moiety and that of the BisPyPol into the protein binding site can explain the experimental results, indicating a better affinity of 1-Zn for the NCS-3.24 variant than testosterone and testosterone-hemisuccinate themselves. They also show that the artificial cofactor entirely fills the cavity, the testosterone part of 1-Zn being bound to one the two subdomains of the protein providing with good complementarities whereas its metal ion remains widely exposed to the solvent which made it a valuable tool for the catalysis of hydrolysis reactions, such as that of HPNP. Some possible improvements in the 'Trojan horse' strategy for obtaining better catalysts of selective reactions will be further studied.

Full text of DOI:10.1016/j.bmc.2014.05.063

Bioorganic & Medicinal Chemistry 22 (2014) 5678–5686
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Neocarzinostatin-based hybrid biocatalysts with a RNase like activity
Agathe Urvoas b, Wadih Ghattas a, Jean-Didier Maréchal c, Frédéric Avenier a, Felix Bellande a, Wei Mao a,
a,
Rémy Ricoux a, , Jean-Pierre Mahy
⇑
⇑
a Institut de Chimie Moléculaire et des Matériaux d’Orsay, UMR 8182 CNRS, Laboratoire de Chimie Bioorganique et Bioinorganique, Bât. 420, Université Paris-sud,
91405 Orsay Cedex, France
b Institut de Biochimie et de Biophysique Moléculaire et Cellulaire, UMR 8619 CNRS, Laboratoire de Modélisation et d’Ingénierie des Protéines, Bât. 430, Université Paris XI,
91405 Orsay Cedex, France
c Departament de Química, Universitat Autònoma de Barcelona, Edifici C.n., Cerdonyola del Vallès, 08193 Barcelona, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
A new zinc(II)-cofactor coupled to a testosterone anchor, zinc(II)-N,N-bis(2-pyridylmethyl)-1,3-diamino-
propa-2-ol-N0(170-succinimidyltestosterone) (Zn-Testo-BisPyPol) 1-Zn has been synthesized and fully
characterized. It has been further associated with a neocarzinostatin variant, NCS-3.24, to generate a
new artificial metalloenzyme following the so-called ‘Trojan horse’ strategy. This new 1-Zn-NCS-3.24 bio-
catalyst showed an interesting catalytic activity as it was found able to catalyze the hydrolysis of the RNA
model HPNP with a good catalytic efficiency (kcat/KM = 13.6 Mꢀ1 sꢀ1 at pH 7) that places it among the best
artificial catalysts for this reaction. Molecular modeling studies showed that a synergy between the
binding of the steroid moiety and that of the BisPyPol into the protein binding site can explain the exper-
imental results, indicating a better affinity of 1-Zn for the NCS-3.24 variant than testosterone and testos-
terone-hemisuccinate themselves. They also show that the artificial cofactor entirely fills the cavity, the
testosterone part of 1-Zn being bound to one the two subdomains of the protein providing with good
complementarities whereas its metal ion remains widely exposed to the solvent which made it a valuable
tool for the catalysis of hydrolysis reactions, such as that of HPNP. Some possible improvements in the
‘Trojan horse’ strategy for obtaining better catalysts of selective reactions will be further studied.
Ó 2014 Elsevier Ltd. All rights reserved.
Received 22 April 2014
Revised 24 May 2014
Accepted 28 May 2014
Available online 14 June 2014
Keywords:
Artificial metalloenzyme
Trojan horse
Zinc complex
Phosphodiester hydrolysis
Molecular modeling
1. Introduction
systems able to catalyze various reactions such as selective bis-
hydroxylation and hydroformylation of alkenes, sulfoxidation of
A promising new kind of eco-compatible hybrid biocatalysts,
that combine the advantages of both biocatalysis and homogeneous
catalysis, has recently been developed: artificial metalloenzymes or
« Artzymes ».1 Such catalysts associate a synthetic metal-cofactor
that is responsible for the catalytic activity with a protein that pro-
vides the metal cofactor with a chiral environment and protects it
against degradation. They are not only able, like enzymes, to work
under mild conditions with a high stereoselectivity, but they can
also be used in a wide scope of reactions like synthetic catalysts.
Additionally, their efficiency and selectivity can be optimized both
on the protein side, by directed evolution, and on the metal cofactor
side by chemical modification of the ligand.
sulphides, epoxidation of styrene derivatives, and Diels–Alder
reactions.2–7
Second, other metalloenzymes have been obtained upon the
covalent attachment of a metal cofactor into a protein, most often
by a covalent bond with a cysteine residue, and have been used to
catalyze a wide range of enantioselective transformations such as
hydrolysis of esters, sulfoxidation, epoxidation, hydrogenation
and Diels Alder reaction.8–13 Finally, artzymes can also be obtained
after the supramolecular anchoring of metal cofactors into pro-
teins, using either the ‘host–guest strategy’ or the ‘Trojan-horse
strategy’. The former was used to generate artzymes that catalyzed
either the stereoselective sulfoxidation of thioanisole,14,15 stereo-
selective Diels–Alder reactions16 or dihydrogen production17 or
that displayed peroxidase activity18 and O2 binding properties.19
We ourselves prepared by non-covalent insertion of iron- and
manganese- anionic tetraarylporphyrins in xylanase A artificial
hemoproteins that were found able to catalyze the stereoselective
Different strategies have been described to obtain artzymes.
First, the direct insertion of a metal ion into proteins has led to
⇑
Corresponding authors. Tel.: +33 1 69157421; fax: +33 1 69154723 (R.R.); tel.:
20
+33 1 69157421; fax: +33 1 69157281 (J.-P.M.).
E-mail addresses: remy.ricoux@u-psud.fr (R. Ricoux), jean-pierre.mahy@u-psud.
fr (J.-P. Mahy).
sulfoxidation of thioanisole by H2O2 and epoxidation of styrene
derivatives by oxone.21
http://dx.doi.org/10.1016/j.bmc.2014.05.063
0968-0896/Ó 2014 Elsevier Ltd. All rights reserved.

A. Urvoas et al. / Bioorg. Med. Chem. 22 (2014) 5678–5686
5679
The ‘Trojan-horse strategy’ was historically introduced by
Whitesides and coworkers but mainly developed and optimized
by Ward and coll. taking advantage of the strong affinity between
streptavidin and biotin, biotinylated transition metal complexes
were inserted into streptavidin and its variants to build new
artificial metalloenzymes that were able to catalyze the stereose-
lective reduction of acetamido-acrylic acid by H2.22,23 We also pre-
viously used the ‘Trojan horse strategy’ to build new artificial
hemoproteins. Taking advantage of the strong affinity of a monoclo-
nal antibody for its anti-estradiol antigen (KD = 9.5 ꢁ 10ꢀ10 M), the
artificial hemoproteins were built by insertion of Fe- or Mn-porphy-
rin–estradiol conjugates into the anti-oestradiol antibody.24–26
These biocatalysts catalyze selective oxidations such as the
enantioselective sulfoxidation of thioanisole by H2O2 with a 10%
enantiomeric excess and the chemoselective epoxidation of styrene
by KHSO5.
cavity of NCS-3.24. Second, a lowering of the pKa value of the
Zn–H2O/Zn–OH couple due to the hydrophobic environment in
the protein binding pocket should promote the formation of metal
hydroxo species responsible for the catalytic activity.
Three substrates were used to assay the hydrolytic activity of
the Zn-complex and the corresponding NCS-3.24-Zn-complex
artzymes: p-nitrophenyl-phosphate ester pNPP, a phosphomono-
ester substrate also used in enzyme assays, its phosphodiester
counterpart bis-p-nitrophenyl-phosphate ester, BNPP, which had
already been used for model complex studies, and p-Nitrophenyl
2-Hydroxypropyl Phosphate, HPNP, a phosphodiester mimic of
RNA that has been used in model studies for RNAase models.28
In this work we describe the synthesis and characterization of a
new zinc(II)-cofactor coupled to a testosterone anchor, zinc(II)-
N,N-bis(2-pyridylmethyl)-1,3-diamino-propa-2-ol-N0(170-succin-
imidyltestosterone) (1-Zn) which can specifically bind to the
More recently, a water-soluble anionic iron(III)-porphyrin–
testosterone conjugate, Testo-porph-Fe, was synthesized and fur-
ther associated to a neocarzinostatin variant, NCS-3.24, to generate
a new artificial metalloenzyme (Artzyme) following the so-called
‘Trojan horse’ strategy.27 This new Testo-porph-Fe-NCS-3.24 bio-
catalyst showed an interesting catalytic activity as it was found
able to catalyze the chemoselective and slightly enantioselective
(ee = 13%) sulfoxidation of thioanisole by H2O2. Molecular model-
ing studies showed that a synergy between the binding of the ste-
roid moiety and that of the porphyrin macrocycle into the protein
binding site could explain the experimental results, indicating a
better affinity of Testo-porph-Fe for the NCS-3.24 variant than tes-
tosterone and testosterone-hemisuccinate themselves. They also
showed that the Fe–porphyrin complex was sandwiched between
the two subdomains of the protein providing with good comple-
mentarities. However, owing to the large size of this artificial
cofactor, it entirely filled the cavity and its metal ion remained
widely exposed to the solvent, which explained the moderate
enantioselectivity observed.27
Starting from these preliminary observations, and in order to
improve the Trojan horse strategy involving the NCS-3.24 mutant,
we decided to associate this protein with a testosterone moiety
coupled with a much smaller bis-pyridyl-metal complex that
would have a better chance to be more deeply inserted in the
enzyme cavity. We also decided to explore other reactions involv-
ing this time Lewis acid type catalysis that are fully compatible
with artificial metalloproteins catalysis in aqueous medium. Our
choice fell on the hydrolysis of phosphate diesters, a reaction with
a high economic potential.
Enzymes and ribozymes that catalyze the hydrolysis of phos-
phate esters frequently use metal ions (Zn2+, Mg2+ etc. . .) cofactors
that play a key role in the catalysis. They provide rate acceleration
by offering a metal-bound hydroxide as nucleophile, stabilization
of the negatively charged transition state and/or leaving group,
and Lewis acid activation by binding to one or more oxygen of
the substrate phosphoryl. Numerous bio-inspired complexes also
present a very good hydrolytic activity,28 such as the mononuclear
ZnII complex reported by the Williams team in 2005, that effi-
ciently catalyzes the hydrolysis of p-Nitrophenyl-2-Hydroxypropyl
Phosphate (HPNP).29 We thus decided, in order to generate new
artificial metalloenzymes for phosphate ester hydrolysis, to syn-
thetize a zinc(II)-cofactor coupled to a testosterone anchor, that
would act as a Trojan horse to drive it into the binding cavity of
a testosterone binding protein, the neocarzinostatin NCS-3.24
variant.
NCS-3.24 protein with a KD of 4.0 0.4 lM and form a new artifi-
cial enzyme. This artzyme catalyses the hydrolysis of the RNA
model HPNP with a good catalytic activity (kcat = 3 ꢁ 10ꢀ4 sꢀ1
)
and a tight binding (KM = 22 lM) of the substrate. Overall, with a
catalytic efficiency kcat/KM = 13.6 Mꢀ1sꢀ1 at pH 7, this artzyme
stands in the upper part of artificial catalysts for this reaction.
2. Materials and methods
2.1. Physical measurements
All commercially available reagents and solvents were
purchased and used without further purification.
UV–visible spectroscopy studies were performed on double
beam UVIKON 860XL and CARY 300BIO VARIAN spectro-
photometers.
1H NMR and 13C NMR spectra were recorded on BRUKER AM
250, 300, 360 spectrometers. Chemical shifts d are given in ppm
with respect to TMS, using either CHCl3 or DMSO as an internal
reference (d 7.27 and 2.50 ppm/TMS respectively). Electrospray
ionization mass spectrometry experiments were carried on a
MicrOTOF-Q Bruker. Microwave assisted reactions were performed
on a CEM Discover apparatus.
The artificial metalloenzymes were characterized by quenching
of fluorescence performed on a CARY Eclipse spectrofluorimeter
and circular dichroism analyses were performed on a Jasco J-710
spectropolarimeter (Jasco, Japan).
2.2. Synthesis of zinc(II)-N,N-bis(2-pyridylmethyl)-1,3-diamino-
propa-2-ol-N0(170-succinimidyltestosterone)
ZnII(TestoBisPyPol) 1-Zn
The Zn(II)-N,N-bis(2-pyridylmethyl)-1,3-diaminopropa-2-ol-N0
(170-succinimidyltestosterone) ZnII(TestoBisPyPol) 1-Zn complex
was synthetized in 3 steps (Fig. 1). First, N0,N0-bis(2-pyridyl-
methyl)-1,3-diamino-propan-2-ol,
BisPyPol,
was
prepared
according to literature,30 whereas the synthesis of 170-succinim-
idyltestosterone (TestoCOOH) was inspired from results already
published by Chen et al. in 2008.31 Second, the N,N-Bis(2-pyridyl-
methyl)-1,3-diaminopropa-2-ol-N0(170-succinimidyltestosterone)
Testo-BisPyPol ligand 1 was synthetized by peptide coupling of
TestoCOOH with BisPyPol, and, finally, the reaction of 1 with zinc
chloride in acetonitrile at room temperature afforded the 1-Zn
complex.
In this study, two main effects that would help to increase the
catalytic efficiency kcat/KM of the artzyme for the hydrolysis of phos-
phate esters were expected from the host protein. First, a decrease
the value KM of the reaction was expected thanks to the favored
binding of the phosphate ester substrates into the hydrophobic
2.2.1. Synthesis of 170-succinimidyltestosterone (TestoCOOH)
A suspension of testosterone (1.16 g, 4.02 mmol) and succinic
anhydride (1.0 g, 10 mmol) in pyridine (0.6 mL) was placed in a
microwave vessel equipped with stirring. The vessel was sealed

5680
A. Urvoas et al. / Bioorg. Med. Chem. 22 (2014) 5678–5686
reduced pressure and the obtained brown oil was dissolved ace-
tone. After treatment with 5% NaHCO3 a brown precipitate was
obtained. This precipitate was dissolved in CH2Cl2 and washed
with 3 ꢁ 30 mL water. After drying and removal of the solvent
under reduced pressure, 1 was obtained as a brown oil with a
quantitative yield.
1H NMR (300 MHz, CDCl3): d = 8.47 (d, J = 4.8 Hz, 2Hpyridyl), 7.52
(td, J = 7.5, 1.7 Hz, 2Hpyridyl), 7.28–7.16 (m, 2Hpyridyl), 7.10–7.06 (m,
2Hpyridyl), 6.34 (t, J = 5.6 Hz, 1HNH), 5.66 (s, 1Htesto), 4.52 (t,
J = 8.3 Hz, 1Htesto), 3.93–3.76 (m, 5H, CHOH, 2CH2methylpyridyl),
3.47–3.39 (m, 1HNHCH2), 3.06–2.98 (m, 1HNHCH2), 2.77–2.72 (m,
1HNCH2CH), 2.58–0.85 (m, 28H, 3CHtesto
, 8CH2testo, 1CH3testo,
2CH2succinic, 1HNCH2CH, OH), 0.75 (s, 3H, CH3testo) ppm; 13C NMR
(300 MHz, CDCl3): d = 199.4 COtesto, 172.7 COOsuccinic, 171.5 CONsuc-
cinic, 170.9 Ctesto, 159.0 2Cpyridyl, 148.9 2CHpyridyl, 136.5 2CHpyridyl
123.8 CHtesto, 123.1 2CHpyridyl, 122.1 2CHpyridyl, 82.6 Ctesto, 67.9
CHOH, 60.3 2CH2methylpyridyl, 58.7 CNCH2CH, 53.6 CHtesto, 50.2 CHtesto
43.1 CNHCH2, 42.5 Ctesto, 38.5 Ctesto, 36.6 CH2testo, 35.7 CH2testo, 35.4
CHtesto, 33.9 CH2testo, 32.7 CH2testo, 31.5 CH2testo, 30.9 CH2succinic
,
,
,
29.7 CH2succinic, 27.4 CH2testo, 23.4 CH2testo, 20.5 CH2testo, 17.3
CH3testo, 12.0 CH3testo ppm; ESI-HRMS: [M+H+]+, found 643.3855.
C38H51N4O5 requires 643.3859.
2.2.4. Synthesis of 1-Zn
Testo-BisPyPol 1 (43 mg, 0.158 mmol) was dissolved in 1 mL
acetonitrile at room temperature. 1 molar equiv ZnCl2 was then
added to obtain an orange solution that was used without further
purification. HRMS: M+ found m/z 741.2720 calculated for (C38H50-
N4O5ClZn)+ 741.2761
2.3. Synthesis of the barium p-Nitrophenyl-2-Hydroxypropyl
Phosphate (HPNP) substrate 2
Figure 1. Synthesis of the zinc(II)-N,N-bis(2-pyridylmethyl)-1,3-diamino-propa-2-
ol-N0(170-succinimidyltestosterone) ZnII(TestoBisPyPol) 1-Zn complex.
Barium p-Nitrophenyl-2-Hydroxypropyl Phosphate (HPNP) was
synthesized according to literature.32
and irradiated in the microwave oven for 20 min at 230 °C. The
obtained dark material was dried under vacuum and the residue
redissolved in ether (150 mL). The product was extracted in 1 M
aqueous Na2CO3 (3 ꢁ 150 mL). The aqueous phases were combined
and acidified to about pH = 2 using HCl (37%). The product was
extracted in ether (3 ꢁ 150 mL) dried over magnesium sulfate
and volatiles removed by evaporation yielding title compound
(96%, 1.50 g, 3.86 mmol) as a white solid.
2.4. Expression and purification of the neocarzinostatin variant
(NCS-3.24)
The NCS protein variant NCS-3.24 was prepared essentially as
previously reported by of Minard et al.33,34 Escherichia coli
BLR(DE3) pLysS strain (Novagen) were used as expression host.
Cells transformed with the phagemid containing the gene coding
for the protein were grown in 500 ml of 2YT medium in a 2 l flask
1H NMR (250 MHz, CDCl3): d = 5.71 (s, CH), 4.60 (t, J = 8.4 Hz,
1H, CH), 2.69–2.57 (m, 4H, 2 CH2), 2.41–0.86 (m, 22H, 3 CH, 8
CH2, CH3), 0.81 (s, 3H, CH3) ppm; 13C NMR (350 MHz, DMSO-d6):
for 70 h at 30 °C, 200 rpm, in the presence of 100
lg/ml ampicillin
and 30 g/ml chloramphenicol. Under these conditions the protein
l
is efficiently secreted. Bacterial pellets were removed by 30 min
centrifugation at 9000 g and the soluble secreted proteins were
precipitated with ammonium sulfate (650 g/l). Precipitated pro-
teins were solubilized in TBS, and desalted by overnight dialysis
at 4 °C against TBS. The dialysed extract was cleared from any
insoluble material by centrifugation for 45 min at 12000g. His-
tagged variant protein was purified using a nickel affinity chroma-
tography (Ni-NTA-agarose Qiagen). The protein, eluted in 0.3 M
imidazole, was concentrated and injected in a size-exclusion chro-
matography column (Hiload™ 16/60 Superdex™ 75) previously
equilibrated with 20 mM Hepes, 150 mM NaCl buffer, pH 7.5. The
purity of the final sample was checked with an overloaded
SDS–PAGE gel, which had to show one well-resolved band with
no visible contamination. The functionality of the purified proteins
was checked by ELISA as described,35 with 50 nM protein. A Max-
isorp ELISA plate (Nunc) was coated with streptavidin–biotinylated
testosterone complex. Bound proteins were detected with an anti-
His6-tag horseradish peroxidase conjugated monoclonal antibody
(Roche Diagnostics). TBS, streptavidin and BSA coated wells were
used as negative controls. Proteins were also incubated in the
d = 197.8 COtesto, 173.3 COOHsuccinic, 171.7 COOsuccinic, 170.7 Ctesto
123.2 CHtesto, 81.7 Ctesto, 53.1 CHtesto, 49.5 CHtesto, 42.2 Ctesto, 38.1
Ctesto, 36.1 CH2testo, 35.1 CH2testo, 34.7 CHtesto, 33.5 CH2testo, 31.8
,
CH2testo, 31.1 CH2testo, 30.0 CH2succinic, 29.8 CH2succinic, 27.0 CH2testo
,
23.0 CH2testo, 20.1 CH2testo, 16.9 CH3testo, 11.8 CH3testo ppm; ESI-
HRMS: [M–H+]ꢀ, found 387.2169. C23H31O5 requires 387.2171.
2.2.2. Synthesis of the N,N-bis(2-pyridylmethyl)-1,3-diamino-
propa-2-ol ligand (BisPyPol)
The N,N-bis(2-pyridylmethyl)-1,3-diamino-propa-2-ol ligand
was synthesized according to literature.30
2.2.3. Synthesis of TestoBisPyPol 1
A mixture of 147 mg. (0.38 mmol.) 17-succinimidyltestoster-
one, 1.2 equiv N,N-diisopropylethylamine (DIEA) and 1.2 equiv
(benzotriazol-1-yloxyl)-tris-(dimethylamino)-phosphonium hexa-
fluorophosphate (BOP) in CH2Cl2 is stirred for 4 h. at room temper-
ature. After addition of 1 equiv BisPyPol the mixture was stirred
again for 2 days in the dark. The solvent was removed under

A. Urvoas et al. / Bioorg. Med. Chem. 22 (2014) 5678–5686
5681
presence of a free competitor (100 mM and 1 mM testosterone
hemisuccinate (THS)) to check the specificity of the testosterone
binding.
a strategy reported to reproduce accurate enough results for orga-
nometallic compounds binding to protein.42 All the other degrees
of freedoms of the system were allowed to be explored during the
runs. Calculations were performed on the unique structure of the
3TES24 NCS species available in the Protein Data Bank (pdb code
2cbo). In this structure, the testosterone hemisuccinate used as a
template for the Trojan horse strategy here reported is bound at
two different sites at the flanks of the NCS cavity creating an
interdominial solvent exposed region able to bind chemical com-
pounds. Calculations have been undertaken using the GoldScore
scoring function which revealed the best reliability as previously
benchmarked.27
2.5. Circular dichroism studies of the structure of NCS in the
absence and in the presence of ZnII(TestoBisPyPol) 1-Zn
To evaluate the influence of the insertion of ZnII(TestoBisPyPol)
1-Zn on the folding of the NCS-3.24 variant, the molar circular
dichroïsm
in the presence and in the absence of metal complex as follows.
The CD spectrum of a 18 M solution of NCS-3.24 dialyzed in
50 mM phosphate buffer, pH 7, containing no NaCl, was first
recorded in a 300 L quartz cells with a 1 mm path length. CD spec-
De
(mdeg Mꢀ1 cmꢀ1) of the protein35,36 was measured
l
l
2.8. Kinetic studies of phosphate hydrolysis catalyzed by
ZnII(TestoBisPyPol) and its complex with NCS 3.24
tra were then recorded after addition to this solution of increasing
amounts of Zn-Testo-BisPyPol in acetonitrile (1–10 molar equiv).
Finally, increasing amounts of acetonitrile were added to a 18
l
M
The hydrolysis of the Barium p-Nitrophenyl-2-Hydroxypropyl-
Phosphate (HPNP) substrate was studied at various pH (7, 7.5, and
8) in the presence of ZnII(TestoBisPyPol) 1-Zn and its complex with
NCS-3.24 as catalysts. For this, the kinetics of the catalytic hydroly-
sis of HPNP was studied at 20 °C using 20 mM HEPES buffer with
20% acetonitrile. Reactions were performed in 96-Well Microtiter
plates and the kinetics were followed by monitoring the formation
of the product p-nitrophenolate through the measurement of the
absorbance at 405 nm every 10 min. The catalytic assays with the
solution of NCS-3.24 in 50 mM phosphate buffer, pH 7, and CD spec-
tra were recorded in order to study the influence of acetonitrile on
the structure of the protein.
2.6. Determination of the dissociation constant of the
ZnII(TestoBisPyPol)-NCS 3.24 complex by quenching of
fluorescence
The interaction between the protein NCS-3.24 and the metal
complex was studied by intrinsic NCS-3.24 fluorescence quenching
upon successive additions of the complex. Fluorescence measure-
ments were performed with a CARY Eclipse spectrofluorimeter,
using a 1 cm-path length quartz cell. The spectra were recorded
with a bandwidth of 5 nm for both excitation and emission beams
at a scan rate of 250 nm minꢀ1. Emission spectra of the protein
NCS-3.24 were recorded between 300 and 450 nm with an excita-
tion wavelength of 290 nm. Increasing amounts of a 0.5 mM solu-
artzyme NCS-3.24-1-Zn contained 30
and 4 mM HPNP. Assays with 20 M 1-Zn and 4 mM HPNP as well
as assays with 30 M NCS-3.24 and 4 mM HPNP where also per-
formed for comparison. Finally, the hydrolysis of 4 mM HPNP alone
was similarly monitored under the same conditions. KM was mea-
sured at pH 7 using the same concentration of artzyme as above
lM NCS-3.24, 20 lM 1-Zn
l
l
and varying the substrate concentration from 4 mM to 40 lm.
3. Results and discussion
tion of ZnII(TestoBisPyPol) 1 in acetonitrile were added to a 5
lM
solution of NCS-3.24 in 20 mM HEPES buffer pH 7.5 in a final vol-
ume of 1 mL. The total added volume was inferior to 10% of the
total volume. The emitted fluorescence signal at 345 nm of the
solution mixture (F) was measured after each addition of complex
(1). F0 represents the fluorescence signal of the protein in the
absence of ligand and Fmin the fluorescence of the protein saturated
3.1. Synthesis and characterization of zinc(II)-N,N-bis(2-
pyridylmethyl)-1,3-diaminopropa-2-ol-N0(170-succinimidyl-
testosterone) ZnII(TestoBisPyPol) 1-Zn
The N,N-bis(2-pyridylmethyl)-1,3-diaminopropa-2-ol-N0(170-
succinimidyltestosterone) ligand 1 was first synthetized by pep-
tidic coupling of N,N-Bis(2-pyridylmethyl)-1,3-diaminopropa-2-ol
(BisPyPol) with 170-succinimidyl-testosterone (TestoCOOH) as
shown in Figure 1.
with the ligand. Fluorescence values (
D
F = F0 ꢀ F) were plotted as a
function of the total ligand concentration ([L]tot) in the solution
and fitted with the following equation using the OriginÒ program:
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
The BisPyPol ligand was prepared according to literature30 and
the synthesis of 170-succinimidyl-testosterone (TestoCOOH) was
inspired from results already published by Chen et al. in 2008.31
In particular, succinic anhydride was opened by reaction of the
170-OH function of testosterone in the presence of pyridine in a
microwave oven for 20 min at 230 °C. Subsequent treatment by a
saturated aqueous NaHCO3 solution and washing with diethyl-
ether, followed by acidification of the aqueous phase with 4 M
HCl and extraction by diethyl-ether as described in Materials and
Methods, allowed to recover the Testo-COOH product as white
solid in a 96% yield after removal of the solvent under reduced
pressure. The product was then characterized by 1H and 13C NMR
spectroscopy and ESI-HR Mass spectrometry.
2
½ðn½Pꢂ0 þ ½LꢂD þ KDÞ ꢀ ðn½Pꢂ0 þ ½Lꢂ0 þ KDÞ ꢀ 4n½Pꢂ0½LꢂDꢂ
D
F ¼
DFmax
2n½Pꢂ0
where [P]0 and [L]0 are respectively the total protein and ligand
concentrations; n is the stoichiometry (number of ligand binding
sites per protein);
Fmax = F0 ꢀ Fmin represents the asymptotic
value for F; Fmax and KD and n are the calculated values from
D
D
D
the fitting equation.
2.7. Molecular modeling
We carried out a series of protein–ligand docking calculations
using the same approach as recently reported.27 The organometallic
Trojan horse compounds have been prepared under the UCSF
Chimera package37 with the Gaff force field38 for the linker and tes-
tosterone groups and a DFT minimized model for the homogenous
catalyst regions. For the latest, the B3LYP functional39 with the
6–31 g basis set40 was used. The docking has been performed with
the CCDC program GOLD41 and the first coordination sphere of the
metal has been fixed during the docking using a covalent bonding;
Finally the peptide coupling reaction between the carboxylic
acid function of the modified steroid and the terminal primary
amine function of the BisPyPol ligand is realized in the presence of
DIEA and of BOP as a coupling agent. Stirring of the reaction mixture
for 24 h at room temperature was enough to fully convert Testo-
COOH into the desired N,N-Bis(2-pyridylmethyl)-1,3-diaminoprop-
a-2-ol-N0(170-succinimidyl-testosterone) Testo-BisPyPol ligand 1.
The product was fully characterized by Mass spectrometry

5682
A. Urvoas et al. / Bioorg. Med. Chem. 22 (2014) 5678–5686
ESI-HRMS showing a peak at m/z 643.3855 ([M+H+]) and by 1H and
13C NMR spectroscopies that were in agreement with the expected
structure.
purified protein was used for CD characterization and fluorescent
titrations with 1-Zn.
Zinc(II) was inserted into Testo-BisPyPol 1 simply by adding one
molar equivalent of ZnCl2 to a solution of 1 in 1 mL acetonitrile at
room temperature. An orange solution of 1-Zn was then obtained.
The ESI–HRMS analysis was in agreement with a (1-ZnIICl)+, Cl-
structure for the complex as shown by the spectrum in Figure 2,
with a peak at m/z 741,2720 the isotopic pattern of which was per-
fectly fitted with a theoretical (C38H50N4O5ClZn)+ formula. This
suggested that the Zn(II) cation was tetra-coordinated, with the 2
pyridine nitrogen atoms, the amino nitrogen atom and a chloride
ion as ligands, whereas the hydroxyl oxygen atom of the di-
amino-propanol moiety did not participate to its coordination
(Fig. 1), contrary to what was usually observed in di-iron com-
plexes involving similar ligand.
3.2.2. Study of structure of the NCS-3.24 variant by circular
dichroism (CD) in absence and in the presence of 1-Zn
As the 1-Zn complex was poorly soluble in water, a minimal
presence of acetonitrile co-solvent was needed to solubilize it. In
order to appreciate the maximum amount of acetonitrile that
could be used to avoid denaturation of the protein, CD spectra of
a 18 lM solution of the NCS-3.24 variant previously dialyzed in
50 mM phosphate buffer, pH 7 were recorded in the presence of
increasing concentrations of acetonitrile (Fig. 3A).
A slight change was observed in the signal of the NCS 3.24 var-
iant in the presence of either 0%, 10% or 20% of acetonitrile, which
indicated that under those conditions the protein kept its second-
ary b-sheet structure, with possibly a slight change of conforma-
tion that could be caused by up to 20% acetonitrile. Accordingly,
the CD signal of NCS-3.24 was almost not modified after the incu-
3.2. Preparation and characterization of the
ZnII(TestoBisPyPol)-NCS-3.24 artificial metalloenzyme
bation of a 18 lM sample of NCS-3.24 in 50 mM phosphate buffer,
pH 7 for 24 h at 4 °C in the presence of 20% acetonitrile. As a con-
clusion, the NCS-3.24 variant appeared to be able to support the
presence of 20% of acetonitrile without an important change in
the protein structure and for further studies, such a concentration
of co-solvent could be used to incorporate the metal complexes
into the NCS-3.24 binding pocket.
Neocarzinostatin (NCS) is a 113 amino-acids protein that con-
stitutes the most studied member of a family known as chromo-
proteins, secreted by Streptomyces. Each member of this family
tightly and specifically binds a nine membered enediyne ‘chromo-
phore’ that is responsible for the powerful cytotoxic and antibiotic
activities of the chromoprotein complexes. It has thus been inten-
sively studied for its antitumor properties and has also been engi-
neered as a potential drug-carrying scaffold.33–35 In a previous
study, the ligand-binding pocket of NCS was remodeled and direc-
ted evolution techniques were used to create a new binding site for
a target unrelated to its natural ligand. In particular, the NCS-3.24
variant, that contains 7 mutations in the binding site : D33W,
G35A, C37W, W39R, L45W, C47Y and F52N, with respect to wild
type NCS, was found to bind testosterone with a good affinity
(KD = 13 ꢁ 10ꢀ6 M) and its X-ray structure was solved in the apo
and complexed forms.33 We thus decided to associate this NCS var-
iant with a zincii-N,N-bis(2-pyridylmethyl)-1,3-diaminopropa-2-
ol-N0(170-succinimidyl-testosterone) conjugate 1-Zn.
Thus, in order to check the influence of the binding of the 1-Zn
complex on the structure of the NCS-3.24 variant, its spectrum CD
3.2.1. Production and purification of NCS-3.24
The NCS protein variant NCS-3.24 was prepared as previously
reported by Minard et al.33,34 E. coli BLR(DE3) pLysS strain (Nova-
gen) were used as expression host. The His-tagged variant protein
was purified by nickel affinity chromatography (Ni-NTA-agarose
Qiagen) followed by a size-exclusion chromatography column
(Hiload™ 16/60 Superdex™ 75). The purity of the final sample
was checked with an overloaded SDS-PAGE. and the freshly
Figure 3. Circular dichroïsm spectra of the NCS 3.24 variant 18 lM in 50 mM
phosphate buffer, pH 7: (A) in the presence of increasing amounts of acetonitrile,
(B) in the presence of 10 equiv of the Zn-Testo-BisPyPol 1-Zn complex.
Figure 2. ESI-HRMS spectrum of the 1-ZnIICl complex.

A. Urvoas et al. / Bioorg. Med. Chem. 22 (2014) 5678–5686
5683
where [P]0 and [L]0 are respectively the total protein and ligand
concentrations, [NCS-3.24]0 and [1-Zn]0, n is the stoichiometry
was recorded in the distant UV (180–240 nm) in the presence or
not of 1-Zn (Fig. 3B). The spectrum of 18 M NCS-3.24 in 50 mM
phosphate buffer, pH 7 is characterized by a peak at 223 nm that
is typical of b-sheet secondary structure for the protein
l
(number of ligand binding sites per protein);
resents the asymptotic optimal value for F.
D
Fmax = F0 ꢀ Fmin rep-
D
a
As can be seen from Figure 4, the continuous curve corresponds
to the optimal fit with a single site model resulting in KD = 4.0
(Fig. 3B). Such a result is an agreement with the previously
reported studies that described a positive contribution towards
223 nm that was characteristic of the NCS-3.24 structure.33,34 The
binding of 10 equiv of ZnII(TestoBisPyPol) 1-Zn complex on the
NCS-3.24 variant (Fig. 3A) caused a slight change in its CD spec-
trum, with a decrease of the peak at 223 nm and an increase of
the trough at 210 nm. These variations indicated that insertion of
the 1-Zn complex in the binding pocket of NCS-3.24 could cause
a slight local modification of its structure.
0.4 lM, n = 1.2 0.1 and DFmax = 233 3. This clearly showed
that only one molecule of ZnII-TestoBisPyPol 1-Zn was bound in
the binding of the NCS-3.24 variant with a good affinity.
Indeed, it is noteworthy that the observed KD value was lower
than those described by Drevelle et al.33 for the binding by
NCS-3.24 of testosterone hemisuccinate (THS), KD = 40.15
lM,
and testosterone, KD = 13 M, which showed that the attachment
l
of the Zn-bispypol moiety to testosterone not only did not alter
its recognition by the NCS-3.24 variant but even improved it. Such
a phenomenon was also previously observed for the binding of a
metalloporphyrin-testosterone conjugate into NCS-3;24.27
3.2.3. Study of the affinity of the NCS-3.24 variant for the ZnII-
Testo-BisPyPol metal complex by quenching of fluorescence
The affinity of the NCS-3.24 variant for the ZnII-Testo-BisPyPol
1-Zn complex was studied by quenching of the fluorescence of
the tryptophanes caused by the interaction of 1-Zn with the pro-
tein. The excitation wavelength was set at kexc = 290 nm, which
corresponds to the wavelength at which photons are absorbed by
the tryptophanes, and the emitted fluorescence was analyzed at
an emission wavelength kem = 345 nm. An increase in the concen-
tration of metal complex bound to the protein causes an increase
in the absorption of the photons and thus a decrease of the fluores-
cence of the protein at 345 nm.
In addition, Drevelle et al.33 reported that NCS-3.24 possessed
two binding sites for THS. Indeed, using equilibrium dialysis exper-
iments, they verified the existence of one high-affinity binding site
that could be saturated for THS concentrations lower than 100 lM,
and another binding site of lower affinity that could not be satu-
rated even at high THS concentrations. The existence of the two
binding sites was further proven by crystallographic studies on
the structure of the NCS-3.24 variant that showed the binding of
two THS in two different binding sites located in the same crevice
of the protein.33 The afore mentioned results thus show that, con-
trary to what was observed for THS, when testosterone is coupled
to the ZnII-BisPyPol complex through a succinate arm, it only binds
in a single site of NCS-3.24, presumably that of higher affinity.
The fluorescence of a 5 lM solution of the NCS-3.24 variant in
20 mM HEPES buffer pH 7.5, was thus measured in the presence
of increasing amounts of a 0.5 mM solution of 1-Zn in acetonitrile,
taking into account, as shown by the CD studies that the final pro-
portion of acetonitrile should not exceed 20% (Fig. 4). To determine
the dissociation constant KD of the NCS-3.24–1-Zn complex, the
3.2.4. Molecular Studies of the structure of the NCS-3.24-1-Zn
artzyme
decrease in fluorescence at 345 nm (
D
F = F0 ꢀ F) was plotted as a
To better understand the complementarity between the Trojan
horse catalyst Zn-testoBisPyPol 1-Zn with the NCS-3.24 variant,
molecular modeling was performed. We carried out a series of
protein–ligand docking calculations using the same approach as
recently reported.27 Calculations were performed on the 1-Zn
complex into the NCS-3.24 scaffold and twenty solutions were
generated in each case. Of the resulting hypothetical complexes,
one could observe that the entire ensemble of predicted binding
orientations presents a large geometrical variability consistent
with the solvent exposure of the binding region of the protein.
However, a clear trend appears from the calculations: while the
testosterone of the Trojan horse occupies one of the two native
binding sites in the protein, the homogenous catalyst region
occupies the other one (Fig. 5). As the lowest energy mode of 1-
Zn NCS-3.24 is one of the solutions of the most populated clusters
of binding orientations, the following part of the study has been
focused on this particular geometry.
function of the total ligand concentration ([1-Zn]tot) in the solution
(Fig. 4) and fitted with the following equation using the OriginÒ
program as described in Section 2:
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2
½ðn½Pꢂ0 þ ½LꢂD þ KDÞ ꢀ ðn½Pꢂ0 þ ½Lꢂ0 þ KDÞ ꢀ 4n½Pꢂ0½LꢂDꢂ
D
F ¼
DFmax
2n½Pꢂ0
In the lowest energy docking solution of 1-Zn-NCS-3.24, the tes-
tosterone binds in the site 2 of the NCS mutant and the homoge-
nous catalyst in the site 1 (Fig. 5). Unlike the previous artificial
hemoproteins we reported,27 in which testosterone was covalently
attached to bulky metalloporphyrin derivatives, the steroid is not
displaced entirely towards the solvent but remains inside its origi-
nal binding site with protein-ligand complementarities similar to
the original NCS-3.24 structure. Only a gentle twist around an axis
perpendicular to the macrocyclic part of the steroid and crossing at
its center of mass occurs with respect to the X-ray orientation. This
conformational change can easily be rationalized by the low num-
ber of polar residues and large number of hydrophobic ones in the
NCS binding site leading to low directionality for ligand binding. In
fact, only the protonated nitrogen of the Trp33 appears to partici-
pate in hydrogen bonding with the terminal carbonyl of the testos-
terone with system 1-Zn (Fig. 5). This result clearly indicates that
Figure 4. Fluorescence titration of ZnII(TestoBisPyPol) 1 binding on the NCS-3.24
protein: Intrinsic tryptophane fluorescence emission spectra of NCS-3.24 (5 M in
20 mM HEPES buffer pH 7.5, kex = 290 nm), in the presence of increasing concen-
trations of 1-Zn (0 to 40 M, from top to bottom). Insert: corresponding emission
signal at 345 nm fitted with the equation described in material and methods.
l
l

5684
A. Urvoas et al. / Bioorg. Med. Chem. 22 (2014) 5678–5686
buffer with 20% acetonitrile. The reactions were performed in 96-
Well Microtiter plates, and to follow the kinetics, the absorbance
at 405 nm due to the p-nitro-phenolate product formed was mon-
itored every 10 min. Blank reactions involving the hydrolysis of
4 mM HPNP alone or in the presence of 30
lM NCS-3.24 were sim-
ilarly monitored under the same conditions.
Whatever the catalyst, no reaction was observed with either
pNPP or BNPP as substrates. On the contrary, the hydrolysis of
p-nitrophenyl-2-Hydroxypropyl-Phosphate, HPNP, appeared to be
catalyzed by both 1-Zn and 1-Zn-NCS-3.24 complexes.
As shown in Figure 6, at pH 7 the ZnII-Testo-BiPyPol complex
1-Zn, as well as the NCS-3.24 protein, are very poor catalysts for
the HPNP transesterification reaction (in the same range as the
background reaction for HPNP in solution).
On the contrary, the artzyme NCS-3.24-1-Zn presents a good
catalytic activity during the first hour of reaction, characterized
by an optimal rate Vmax = 3.0 ꢁ 10ꢀ4 mol. p-nitrophenol (pNP)
formed/mol cat./s, before reaching a plateau. In this case, the initial
rate of the artzyme is more than 800 fold faster than 1-Zn, showing
a large impact of the protein environment on catalysis (Table 1).
Interestingly, when raising the pH to 7.5 and then to 8, the cat-
alytic activity of NCS-3.24-1-Zn decreases and the rate of hydroly-
sis reaches Vmax = 5.7 ꢁ 10ꢀ5 mol. pNP formed/mol cat./s at pH 8.
On the contrary the catalytic activity of 1-Zn increases and the rate
of hydrolysis reaches Vmax = 1.5 ꢁ 10ꢀ4 mol. pNP formed/mol cat./s
at pH 8. Consequently, the ratio Vmax (NCS-3.24-1-Zn)/Vmax (1-Zn)
drops to 3.6 at pH 7.5 and 0.37 at pH 8.2 (Table 1). This difference is
mainly due to the increase of activity of the zinc complex 1-Zn
which is very sensitive to the concentration of general base in
solution.
Based on this large effect of the protein environment at neutral
pH, catalysis was then performed with an increasing amount of
Figure 5. Representation of the lowest energy solutions of the docking of 1-Zn (up
view, in cyan) into the 3TES24 NCS scaffold (pdb code 2cbo). The protein is
represented in gold ribbons. Ligands and important residues of the binding site are
represented in stick, the metal and its bound halogen substituents in sphere.
substrate from 40 lM to 4 mM, and a KM of 22 lM together with
a kcat of 3 ꢁ 10ꢀ4 sꢀ1 could be calculated, showing a very good
the reduction of the size of the homogenous catalyst with regard to
the previously reported metallo-porphyrin derivatives27 has
allowed the testosterone to maintain most of its original binding
properties. This novel system appears therefore as a far more inter-
esting scaffold for rational enzyme design based on a Trojan horse
strategy into NCS.
Regarding the catalytic center, a good association is also
observed for 1-Zn into the NCS-3.24 variant although not reaching
the entire packing of the porphyrinic ring observed in our initial
Trojan horse intent.27 The entire bispypol ligand is bound to the
site 2 performing this way a good complementarity and reaching
a satisfactory score of about 70 Goldscore units. However, despite
this good insertion of 1-Zn into NCS-3.24, the lowest energy mode
shows that the metal and its bound chloride ion are mainly soaking
into the solvent media. The entire catalytic metal environment
remains therefore accessible to the solvent with no residue of the
protein even appearing at 5 A from the metal. Under these condi-
tions, the NCS-3.24-1–Zn appeared to be perfectly suited for being
able to catalyze hydrolysis reactions, such as that of PNPH.
3.3. Hydrolysis of phosphate esters catalyzed by 1-Zn and its
complex with NCS-3.24
To essay the catalytic activity of the 1-Zn and 1-Zn-NCS-3.24
complexes, the hydrolysis of 4 mM p-nitrophenyl-phosphate ester
(pNPP), or of 4 mM bis-p-nitrophenyl-phosphate ester (BNPP), or of
4 mM p-nitrophenyl-2-Hydroxypropyl-Phosphate (HPNP), was
Figure 6. Absorbance at 405 nm measured every 10 min at pH 7 and 20 °C for the
hydrolysis of 4 mM HPNP Catalyzed by: 20
lM NCS-3.24-1-Zn (
), 20 lM 1-Zn
performed in the presence of either 20
1-Zn in the presence of 30 M NCS-3.24 as catalysts, at 20 °C under
three different conditions of pH (7, 7.5 and 8) using 20 mM Hepes
lM 1-Zn alone or 20 lM
(
), 30 M NCS-3.24 ( ) and 4 mM HPNP alone (
l
). The black line is the
l
linear regression fit for the initial rate of the hydrolysis reaction in presence of NCS-
3.24-1-Zn.

A. Urvoas et al. / Bioorg. Med. Chem. 22 (2014) 5678–5686
5685
Table 1
Ratios of Vmax at different pH
pH
Vmax (NCS-3.24-1-Zn) (mol. pNP/mol Cat/s)a
Vmax (1-Zn) (mol. pNP/mol Cat/s)a
Vmax (NCS-3.24-1-Zn)/Vmax (1-Zn)
7
7.5
8.2
3.0 ꢁ 10ꢀ4
3.5 ꢁ 10ꢀ7
2.2 ꢁ 10ꢀ5
1.5 ꢁ 10ꢀ4
845
3.63
0.37
8.0 ꢁ 10ꢀ5
5.7 ꢁ 10ꢀ5
a
VmaxNCS-3.24-1-Zn is in presence of the artzyme and Vmax1-Zn is in presence of 1-Zn without NCS-3.24.
affinity of the substrate for the artzyme, and a kcat/KM of 13.6 Mꢀ1
sꢀ1 at pH 7. Overall, the improvement of the catalytic activity of the
artzyme compared to the zinc complex probably reflects the incor-
poration of the zinc in a more hydrophobic environment brought
out by the close proximity of the protein. It is very likely that this
locally hydrophobic micro-environment lowers the pKa of the
metal-bound water molecule and allows catalysis at neutral pH.
The main weakness of the catalyst would be his life time since it
appears to be stable only for one hour before losing its activity.
However, this might be explained by the fact that the NCS protein
is known to be stable only for a few hours in solution.
The 1-Zn-NCS-3.24 artzyme thus showed a good catalytic activ-
ity characterized by a kcat value of 3 ꢁ 10ꢀ4 sꢀ1 and an almost 3
logs rate enhancement with respect to 1-Zn alone at pH 7. It also
shows a tight binding (KM = 22 lM) of the substrate which leads
to a catalytic efficiency kcat/KM of 13.6 Mꢀ1 sꢀ1 at pH 7, which
placed this artzyme among the best artificial catalysts for this reac-
tion. In addition, since its activity was optimal at physiological pH,
whereas that of 1-Zn was optimal at pH values above 8 it appeared
as a really valuable mimic for natural RNases.
This novel 1-Zn-NCS-3.24 system appears as a far more inter-
esting scaffold for rational enzyme design based on a Trojan horse
strategy into NCS. However, despite showing a clear improvement
regarding our previous metallo-porphyrinic systems, the relative
position of the homogenous catalysts in 1-Zn-NCS-3.24 does prob-
ably not show sufficient in depth penetration into the protein scaf-
fold to afford the asymmetric environment that would be required
for highly enantioselective catalysis of reactions such as oxidation
reactions. Such variables need to be further optimized in following
steps of our studies.
In terms of kcat, this artificial enzyme is about two orders of
magnitude slower than the best artificial catalyst known to date.43
Concerning the KM the value of 22 lM is comparable to the best
values obtained with zinc-calixarene systems44 and provides the
system with a kcat/KM of 13.6 Mꢀ1 sꢀ1 at pH 7. Which is a very
interesting value bringing the artzyme close to the best catalysts
described so far.28
4. Conclusion
Acknowledgments
The aforementioned results describe the synthesis and charac-
terization a new zinc(II)-cofactor coupled to a testosterone anchor,
zinc(II)-N,N-bis(2-pyridylmethyl)-1,3-diamino-propa-2-ol-N0(170-
succinimidyltestosterone) (Zn-Testo-BisPyPol, 1-Zn). In order to
generate new artificial metalloenzymes following the so-called
‘Trojan horse strategy’, this molecule was further associated with
a neocarzinostatin variant, NCS-3.24, that was previously engi-
neered using a directed evolution approach for the binding of
testosterone. CD studies showed that the addition of 1-Zn to the
NCS-3.24 protein only caused a slight local modification of its
structure, whereas quenching of fluorescence studies were clearly
in favor of an insertion of a single 1-Zn cofactor into its binding
site. These studies also allowed the determination of a KD value
The authors would like to thank Tania Inceoglu for her excellent
technical assistance in mass spectrometry. They are grateful to the
‘Agence Nationale de la Recherche’ for its financial support (ANR
REBAR, ANR CATHYMETOXY).
References and notes
1. Raffy, Q.; Ricoux, R.; Mahy, J.-P. In Focus on Catalysis Research: New
Developments; Ghang, M., Ramel, B., Eds.; Nova Science Publishers: New York,
2011; p 31.
2. Ward, T. R. Acc. Chem. Res. 2011, 44, 47.
3. Kokubo, T.; Sugimoto, T.; Uchida, T.; Tanimoto, S.; Okano, M. J. Chem. Soc., Chem.
Commun. 1983, 769.
4. Bertucci, C.; Botteghi, C.; Giunta, D.; Marchetti, M.; Paganelli, S. Adv. Synth.
Catal. 2002, 344, 556; Marchetti, M.; Mangano, G.; Paganelli, S.; Botteghi, C.
Tetrahedron Lett. 2000, 41, 3717.
of 4.7 lM for the 1-Zn-NCS-3.24 variant complex, which appeared
to be better than those reported for the NCS-3.24-testosterone and
-testosterone-hemisuccinate (THS) complexes.33 This made it clear
that a synergetic effect between the binding of the testosterone
moiety and that of the BisPyPol ligand led to better association of
1-Zn with NCS-3.24 than that of both testosterone and THS alone,
confirming what was previously observed for the binding of iron–
porphyrin–testosterone conjugates into the NCS-3.24 variant.27
Protein–ligand docking experiments clearly indicated that the
reduction of the size of the homogenous catalyst with regard to
our first porphyrin derivatives allowed the testosterone to main-
tain most of its original binding properties and therefore fulfilled
the general concepts of Trojan horse artificial metalloenzymes
design. In our models, the Zn catalytic center occupies the other
site of the NCS-3.24 variant, which apparently provides a good pro-
tection by the protein to enhance catalytic activity. They also show
an exposed metal center and suggest that the binding of the sub-
strate in a catalytic consistent conformation would mostly occur
in a solvent environment. Consequently, the requirements to use
the 1-Zn-NCS-3.24 artzyme as a biocatalyst for hydrolysis reac-
tions appeared to have been fulfilled, and it was then tested as a
biocatalyst for the hydrolysis of the RNA model HPNP.
5. van de Velde, F.; Könemann, L.; van Rantwijk, F.; Sheldon, R. A. Chem. Commun.
1998, 1891; Esmieu, C.; Cherrier, M. V.; Amara, P.; Girgenti, E.; Marchi-
Delapierre, C.; Oddon, F.; Iannello, M.; Jorge-Robin, A.; Cavazza, C.; Ménage, S.
Angew. Chem., Int. Ed. 2013, 52, 3922.
6. Okrasa, K.; Kazlauskas, R. J. Chem. Eur. J. 2006, 6, 1587; Fernández-Gacio, A.;
Codina, A.; Fastrez, J.; Riant, O.; Soumillion, P. ChemBioChem 2006, 7, 1013.
7. Coquière, D.; Bos, J.; Beld, J.; Roelfes, G. Angew. Chem., Int. Ed. 2009, 48, 5159.
8. Davies, R. R.; Distefano, M. D. J. Am. Chem. Soc. 1997, 119, 11643.
9. Carey, J. R.; Ma, S. K.; Pfister, T. D.; Garner, D. K.; Kim, H. K.; Abramite, J. A.;
Wang, Z.; Guo, Z.; Lu, Y. J. Am. Chem. Soc. 2004, 126, 10812.
10. Reetz, M. T. Tetrahedron 2002, 58, 6595.
11. Reetz, M. T.; Rentzsch, M.; Pletsch, A.; Maywald, M.; Maiwald, P.; Peyralans, J.
P.; Maichele, A.; Fu, Y.; Jiao, N.; Hollmann, F.; Mondiere, R.; Taglieber, A.
Tetrahedron 2007, 63, 6404.
12. Panella, L.; Broos, J.; Jin, J.; Fraaije, M. W.; Janssen, D. B.; Jeronimus-Stratingh,
M.; Feringa, B. L.; Minnaard, A. J.; de Vries, J. G. Chem. Commun. 2005, 5656.
13. Talbi, B.; Haquette, P.; Martel, A.; de Montigny, F.; Fosse, C.; Cordier, S.; Roisnel,
T.; Jaouen, G.; Salmain, M. Dalton Trans. 2010, 39, 5605; Cherrier, M. V.;
Engilberge, S.; Amara, P.; Chevalley, A.; Salmain, M.; Fontecilla-Camps, J. C. Eur.
J. Inorg. Chem. 2013, 21, 3596.
14. Mahammed, A.; Gross, Z. J. Am. Chem. Soc. 2005, 127, 2883.
15. Ohashi, M.; Koshiyama, T.; Ueno, T.; Yanase, M.; Fujii, H.; Watanabe, Y. Angew.
Chem., Int. Ed. 2003, 9, 1005; Ueno, T.; Koshiyama, T.; Ohashi, M.; Kondo, K.;
Kono, M.; Suzuki, A.; Yamane, T.; Watanabe, Y. J. Am. Chem. Soc. 2005, 127,
6556; Ueno, T.; Koshiyama, T.; Abe, S.; Yokoi, N.; Ohashi, M.; Nakajima, H.;
Watanabe, Y. J. Organomet. Chem. 2007, 692, 142.
16. Reetz, M. T.; Jiao, N. Angew. Chem., Int. Ed. 2006, 45, 2416.

5686
A. Urvoas et al. / Bioorg. Med. Chem. 22 (2014) 5678–5686
17. Matsuo, T.; Asano, A.; Ando, T.; Hisaeda, Y.; Hayashi, T. Chem. Commun. 2008,
3684.
29. Feng, G.; Mareque-Rivas, J. C.; Torres Martín de Rosales, R.; Williams, N. H. J.
Am. Chem. Soc. 2005, 127, 13470.
18. Hayashi, T.; Hisaeda, Y. Acc. Chem. Res. 2002, 35, 35; Sato, H.; Hayashi, T.; Ando,
T.; Hisaeda, Y.; Ueno, T.; Watanabe, Y. J. Am. Chem. Soc. 2004, 126, 436; Hayashi,
T.; Murata, D.; Makino, M.; Sugimoto, H.; Matsuo, T.; Sato, H.; Shiro, Y.;
Hisaeda, Y. Inorg. Chem. 2006, 45, 10530; Matsuo, T.; Murata, D.; Hisaeda, Y.;
Hori, H.; Hayashi, T. J. Am. Chem. Soc. 2007, 129, 12906; Matsuo, T.; Hayashi, A.;
Abe, M.; Matsuda, T.; Hisaeda, Y.; Hayashi, T. J. Am. Chem. Soc. 2009, 131, 15124.
19. Matsuo, T.; Tsuruta, T.; Maehara, K.; Sato, H.; Hisaeda, Y.; Hayashi, T. Inorg.
Chem. 2005, 44, 9391; Kano, K.; Kitagishi, H.; Dagallier, C.; Kodera, M.; Matsuo,
T.; Hayashi, T.; Hisaeda, Y.; Hirota, S. Inorg. Chem. 2006, 45, 4448; Matsuo, T.;
Ikegami, T.; Sato, H.; Hisaeda, Y.; Hayashi, T. J. Inorg. Biochem. 2006, 100, 1265;
Harada, K.; Makino, M.; Sugimoto, H.; Hirota, S.; Matsuo, T.; Shiro, Y.; Hisaeda,
Y.; Hayashi, T. Biochemistry 2007, 46, 9406; Matsuo, T.; Ito, K.; Nakashima, Y.;
Hisaeda, Y.; Hayashi, T. J. Inorg. Biochem. 2008, 102, 166.
30. Hayashi, Y.; Kayatani, T.; Sugimoto, H.; Suzuki, M.; Inomata, K.; Uehara, A.;
Mizutani, Y.; Kitagawa, T.; Maeda, Y. J. Am. Chem. Soc. 1995, 117, 11220.
31. Chen, H. X.; Zhang, X. X. Electrophoresis 2008, 29, 3406.
32. Brown, D. M.; Usher, D. A. J. Chem. Soc. 1965, 6558.
33. Drevelle, A.; Graille, M.; Heyd, B.; Sorel, I.; Ulryck, N.; Pecorari, F.; Desmadril,
M.; van Tilbeurgh, H.; Minard, P. J. Mol. Biol. 2006, 358, 455.
34. Drevelle, A.; Urvoas, A.; Ben Hamida-Rebaï, M.; Van Vooren, G.; Nicaise, M.;
Valerio-Lepiniec, M.; Desmadril, M.; Robert, Ch. H.; Minard, P. ChemBioChem
2009, 10, 1349.
35. Heyd, B.; Pecorari, F.; Collinet, B.; Adjadj, E.; Desmadril, M.; Minard, P.
Biochemistry 2003, 42, 5674.
36. Nicaise, M.; Valerio-Lepiniec, M.; Minard, P.; Desmadril, M. Protein Sci. 2004,
1882.
20. Ricoux, R.; Allard, M.; Dubuc, R.; Dupont, C.; Marechal, J.-D.; Mahy, J.-P. Org.
Biomol. Chem. 2009, 7, 3208.
37. Pettersen, E. F.; Goddard, T. D.; Huang, C. C.; Couch, G. S.; Greenblatt, D. M.;
Meng, E. C.; Ferrin, T. E. J. Comput. Chem. 2004, 25, 1605.
38. Wang, J.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.; Case, D. A. J. Comput. Chem.
2004, 25, 1157.
39. (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648; (b) Lee, C.; Yang, W.; Parr, R. G.
Phys. Rev. B 1988, 37, 785.
40. (a) Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. J. Chem. Phys. 1980, 72,
5639; (b) Chandrasekhar, J.; Spitznagel, G. W.; Schleyer, P. V. R.; Clark, T. J.
Comput. Chem. 1983, 4, 294.
21. Allard, M.; Dupont, C.; Munoz-Robles, V.; Doucet, N.; Lledos, A.; Marechal, J.-D.;
Urvoas, A.; Mahy, J.-P.; Ricoux, R. ChemBioChem 2012, 13, 240.
22. Collot, J.; Gradinaru, J.; Humbert, N.; Skander, M.; Zocchi, A.; Ward, T. R. J. Am.
Chem. Soc. 2003, 125, 9030; Letondor, C.; Ward, T. R. ChemBioChem 2006, 7,
1845.
23. Pordea, A.; Creus, M.; Panek, J.; Duboc, C.; Mathis, D.; Novic, M.; Ward, T. R. J.
Am. Chem. Soc. 2008, 130, 8085.
24. Raffy, Q.; Ricoux, R.; Mahy, J.-P. Tetrahedron Lett. 2008, 49, 1865.
25. Raffy, Q.; Ricoux, R.; Sansiaume, E.; Pethe, S.; Mahy, J.-P. J. Mol. Catal. A: Chem.
2010, 317, 19.
26. Sansiaume, E.; Ricoux, R.; Gori, D.; Mahy, J.-P. Tetrahedron: Asymmetry 2010, 21,
1593.
41. Verdonk, M. L.; Cole, J. C.; Hartshorn, M. J.; Murray, C. W.; Taylor, R. D. Proteins
2003, 52, 609.
42. Ortega-Carrasco, E.; Lledos, A.; Marechal, J.-D. J. Comput. Chem. 2014, 35, 192.
43. Feng, G.; Natale, D.; Prabaharan, R.; Mareque-Rivas, J. C.; Williams, N. H. Angew.
Chem. 2006, 7056.
27. Sansiaume-Dagousset, E.; Urvoas, A.; Chelly, K.; Ghattas, W.; Maréchal, J.-D.;
Mahy, J.-P.; Ricoux, R. Dalton Trans. 2014. http://dx.doi.org/10.1039/
c4dt00151f.
44. Molenveld, P.; Stikvoort, W. M. G.; Kooijman, H.; Spek, A. L.; Engbersen, J. F. J.;
Reinhoudt, D. N. J. Org. Chem. 1999, 64, 3896.
28. Desbouis, D.; Troitsky, I. P.; Belousoff, M. J.; Spiccia, L.; Graham, B. Coord. Chem.
Rev. 2013, 256, 897.