Dual-Oriented Solid-Phase Molecular Imprinting: Toward Selective Artificial Receptors for Recognition of Nucleotides in Water

Article abstract of DOI:10.1021/acs.macromol.7b01782

We describe the synthesis of water-soluble molecularly imprinted polymer nanoparticles (MIP-NPs) as a new artificial host receptor for the recognition of adenosine monophosphate (AMP), used herein, as a model nucleotide. MIP-NPs were prepared by solid-pha

Full text of DOI:10.1021/acs.macromol.7b01782

Article
pubs.acs.org/Macromolecules
Dual-Oriented Solid-Phase Molecular Imprinting: Toward Selective
Artificial Receptors for Recognition of Nucleotides in Water
Cecília A. Mourao,†,‡ Frank Bokeloh,† Jingjing Xu,† Elise Prost,† Luminita Duma,† Franck Merlier,†
̃
Sonia M. A. Bueno,‡ Karsten Haupt,*,† and Bernadette Tse Sum Bui*,†
̂
†Sorbonne Universites
Couttolenc, CS 60319, 60203 Cedex Compieg
‡School of Chemical Engineering, University of Campinas, Rua Albert Einstein, 500, Campinas, Sao Paulo, Brazil
́
, Universite
́
̀
de Technologie de Compiegne, CNRS Enzyme and Cell Engineering Laboratory, Rue Roger
̀
ne, France
̃
S
* Supporting Information
ABSTRACT: We describe the synthesis of water-soluble
molecularly imprinted polymer nanoparticles (MIP-NPs) as a
new artificial host receptor for the recognition of adenosine
monophosphate (AMP), used herein, as a model nucleotide.
MIP-NPs were prepared by solid-phase synthesis on glass
beads (GB) using, for the first time, immobilized Fe(III)-
chelate as an affinity ligand to orientate the AMP via its
phosphate group. A polymerizable thymine monomer which
can induce complementary base-pairing with the adenine
moiety of the nucleotide was synthesized and incorporated in
the polymerization mixture to constrain the AMP in a dual-
orientated configuration. The MIP-NPs were remarkably selective toward AMP as they did not bind other nucleotides, GMP,
UMP, and CMP. This strategy of using the phosphate group of AMP as a hinge enables unhindered pairing of the nucleobase
with its corresponding complementary base monomer and can be extended to the preparation of specific MIP receptors for other
key nucleotides in aqueous conditions.
However, nucleotides are very hydrophilic molecules due to
the presence of a ribose and a phosphate functionality, which
render them very soluble in water. This property is not really
compatible with traditional imprinting method which prefer-
entially takes place in apolar and aprotic organic solvents.6
Hence, imprinting of nucleosides (nucleotides without the
phosphate group), often used as template for the preparation of
MIP for nucleotides’ recognition, requires their prior
conversion to their acetylated or acylated forms.7−9 Sometimes,
a functional monomer, 2,6-bis(acrylamido)pyridine (BAApy),
is added to make multiple hydrogen bonds with the
nucleobase.10 Though this strategy results in MIPs with specific
binding in these solvents, no binding was observed in water.8,10
Attempts to obtain water-compatible MIPs for targeting the
nucleotide adenosine 5′-monophosphate (AMP), for instance,
have also been reported. Commonly used functional monomers
were based on boronic acid, associated with acrylamide to
target the cis-diol function of ribose and the nucleobase,
respectively.7,11 Others used cationic monomers like (3-
acrylamidopropyl)trimethylammonium chloride12 and 2-
(dimethylamino)ethyl methacrylate13 to target the phosphate
group by making electrostatic interactions. Silica-based MIPs
have also been reported.14 Most of these MIPs showed cross-
1. INTRODUCTION
Nucleotides are ubiquitously present in biological systems and
are highly important biomolecules as they are the building
blocks of nucleic acids (RNA and DNA). Additionally, they
play essential roles in cell signaling and metabolism and
participate in numerous enzymatic reactions. Consequently,
nucleotides are one of the most targeted anionic compounds,
and many different artificial host systems that can work in
aqueous conditions have been developed.1,2 The majority of
nucleotide receptors are based on ion-pairing of the phosphate
moiety with polycationic polyamines and polyguanidines.
Calixarenes and cyclodextrins have also been used as they
present hydrophobic pockets for the nucleobase and/or the
sugar.1,2 Another very promising type of artificial receptors is
molecularly imprinted polymers (MIPs). MIPs, often referred
as synthetic antibodies, are obtained by the copolymerization of
functional and cross-linking monomers around a template
molecule.3−5 The template molecule can be the target analyte
or a derivative thereof. Removal of the template leaves cavities
complementary in shape, size, and functional groups
orientation in the polymer network, which are able to rebind
the template molecule with high affinities and selectivities
comparable to those of natural antibodies. MIPs have
considerable advantages over antibodies as they possess a
higher chemical, physical, and thermal stability, ease of
obtention, and low cost.
Received: August 17, 2017
Revised: September 8, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acs.macromol.7b01782
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Figure 1. Immobilization of AMP on GBs using Fe3+ chelate. The phosphate group interacts with Fe3+ to form a four-membered ring (adapted from
ref 23).
selectivities with other nucleo(s/t)ides and were mostly
synthesized as films or in bulk which generally give rise to
heterogeneous polymers with a diverse population of binding
sites ranging from high to low affinity, a common feature of
MIPs prepared by the self-assembly approach. Moreover, the
presence of residual template even after extensive washing
remains a problem.3,6
To overcome these problems and also to respond to the
growing interest15,16 for nanosized, “monoclonal-type” MIPs
which are soluble in water, we17−19 and others20 have
developed a solid-phase synthesis approach where the template
is immobilized via an affinity ligand on glass beads as solid
support. This configuration allows an oriented immobilization
of the template upon which MIP-NPs are synthesized. The GBs
play the role of both a reactor and a separation column since
the MIP-NPs are synthesized and purified in situ. After
synthesis and washing of unreacted reagents and nanoparticles
that polymerized distantly from the immobilized template, only
uniform MIP-NPs with homogeneous (since all have the same
orientation) binding sites are eluted.
Figure 2. (A) Synthesis of the functional monomer 1-(vinylbenzyl)-
thymine (VBT). (B) Pairing between adenine of AMP and thymine of
VBT.
Herein, we describe the solid-phase synthesis of MIP-NPs
targeting AMP. Metal-chelate (Figure 1) and p-amino-
benzamidine (PAB) (Figure S1) were used as affinity ligands
to immobilize AMP on the glass beads. For the metal-chelate
setup, we exploited the strong coordinate bonding between the
phosphate of AMP and Fe3+, itself complexed with iminodi-
acetic acid (IDA).21 In order to confer high selectivity to the
MIP, a functional monomer bearing a complementary thymine
base, 1-(vinylbenzyl)thymine (VBT)22 (Figure 2), so as to form
hydrogen bonds with the adenine base on the AMP, was
synthesized and included as a comonomer in the polymer-
ization mixture.
A high amount of N-isopropylacrylamide (NIPAm) was
included in the MIP mixture to impart thermoresponsiveness
and allows its facile liberation from the immobilized AMP. The
resulting MIP-NPs were very specific for AMP as no binding
was observed with the NIP. No binding was observed with
guanosine 5′-monophosphate (GMP), uridine 5′-monophos-
phate (UMP), and cytidine 5′-monophosphate (CMP)
(structures in Figure 3), indicating the high selectivity of the
MIP-NPs for its template.
Interestingly, the MIP-AMP could bind adenosine to the
same extent as AMP. Recently, a solid-phase approach was
employed for synthesizing MIPs for nucleoside targeting.24 2′-
Deoxyadenosine (dA) was used as template, and a polymer-
izable 2′-deoxyuridine was incorporated in the MIP so as to
induce complementary base-pairing. Unfortunately, dA was
linked to the glass beads by its amine group, hence rendering
the base-pairing face partially inaccessible to the normal
hydrogen-bonding geometry. In this work, we used instead a
nucleotide as template to enable its immobilization via its
phosphate moiety, hence leaving the nucleobase sterically
nonhindered for recognition with the complementary base pair.
This configuration indeed leads to more selective MIPs, and at
B
DOI: 10.1021/acs.macromol.7b01782
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
precipitate. The precipitate was filtered and vacuum-dried overnight. It
was then solubilized in 180 mL of dichloromethane by stirring for 3 h.
The mixture was filtered in order to remove undissolved components
and the dichloromethane concentrated under vacuum to 50 mL. 150
mL of hexane was added, causing the precipitation of 1-(vinylbenzyl)-
thymine. The final product was stirred for 1 h, filtered, and vacuum-
1
dried. Yield was 47%. H NMR (400 MHz, DMSO-d6): 7.63 (broad
qd, 1H), 7.47−7.26 (d, 4H), 6.72 (dd, 1H), 5.82 (dd, 1H), 5.25 (dd,
1
1H), 4.82 (s, 2H), 1.75 (broad d, 1H). The H NMR spectrum is
shown in Figure S2.
Determination of the Lower Critical Solution Temperature
(LCST) of the Polymerization Mixture. The polymerization
mixture was prepared in a glass vial by mixing NIPAm (41.19 mg,
91 mol %) and VBT (3.88 mg in 1 mL of methanol, 4 mol %) as
functional monomers and N,N-ethylenebis(acrylamide) (EbAm) (3.36
mg, 5 mol %) as cross-linker in a total volume of 9.64 mL of 25 mM 2-
(N-morpholino)ethanesulfonic acid (MES) buffer (pH 5.5) so that the
total monomer concentration is 0.5% (w/w). The MES buffer will
herafter be referred as buffer A. 3.4 mg of potassium persulfate (KPS)/
0.3 μL of N,N,N′,N′-tetramethylethylenediamine (TEMED) (7.5/1
molar ratio; the amount of KPS was 3% mol/mol with respect to
polymerizable double bonds) as initiation system was then added. The
mixture was purged with nitrogen for 10 min. The polymerization
mixture was left to polymerize overnight in a water bath at 37 °C. The
LCST of the polymer was determined via turbidity measurement
(cloud point determination)29 using a UV−vis spectrophotometer
fitted with a circulation bath and single cell Peltier accessory. The
transmittance of the polymer solution (diluted 5 times with buffer A),
contained in a 1 cm path glass cuvette at a fixed wavelength of 720 nm
(due to light scattering), was recorded as a function of temperature
(heating rate: 1 °C/min). The polymer solution was heated from 25 to
40 °C, and the LCST was taken as the temperature at which the
solution transmittance reached 50%, which corresponds to ∼32 °C
(Figure 4).
Figure 3. Chemical structures of molecules employed in this study.
the same time the MIP can advantageously recognize both the
nucleotide and its associated nucleoside.
In contrast, MIP-NPs obtained with p-aminobenzamidine as
affinity ligand (Figure S1) were not as selective as those
obtained with the metal chelate (Figure 1). This difference will
be discussed in terms of 31P NMR studies of the interaction of
AMP with a benzamidine-based monomer.
2. EXPERIMENTAL SECTION
Materials and Instruments. HPLC solvents were purchased from
Biosolve Chimie (Dieuze, France). All other solvents and chemicals
were of analytical grade and purchased from VWR International
(Fontenay sous Bois, France) or Sigma-Aldrich (St. Quentin Fallavier,
France), unless otherwise stated. Adenosine 5′-monophosphate
disodium salt (AMP), guanosine 5′-monophosphate disodium salt
(GMP), uridine 5′-monophosphate disodium salt (UMP), cytidine 5′-
monophosphate disodium salt (CMP), adenosine 5′-triphosphate
disodium salt hydrate (ATP), adenosine 3′,5′-cyclic monophosphate
sodium salt monohydrate (cAMP), adenosine, and FeCl3·6H2O were
purchased from Sigma-Aldrich. Glass beads (GBs) of diameter 0.1 mm
were obtained from Roth Sochiel E.U.R.L (Lauterbourg, France).
Deuterated NMR solvents and NMR tubes were purchased from
Euriso-top (France) and Norell (USA), respectively. Buffers were
prepared with Milli-Q water, purified using a Milli-Q system
(Millipore, Molsheim, France). Absorbance measurements were
done on a CARY60 UV−vis spectrophotometer (Agilent Technolo-
1
gies). H and 31P NMR experiments were performed on a 400 MHz
Bruker spectrometer at 25 °C.
Analysis of the Interaction of AB with AMP Using 31P NMR
Spectroscopy. AB monomer (Figure 3), the polymerizable form of
p-aminobenzamidine, was synthesized as its acetate salt by reacting
PAB with acryloyl chloride, as previously described (details in the
Supporting Information).25 The stoichiometry and strength of the
AB−AMP complex were determined by chemical shift analysis using
the methods of continuous variation (Job plot)26 and of titration,27,28
respectively. For the Job plot, solutions of AB (M) and AMP (T) in a
mixture of CD3OD/D2O (4/1), at a constant total concentration of 2
mM and with a template molar fraction, χT = [T]/([T] + [M]),
varying from 0 to 1 by steps of 0.1, were prepared. The total volume of
each sample was 0.6 mL. Titration experiments were done to
determine the association constant. A stock solution of 40 mM AB was
prepared and added (from 0 to 10 equiv) to a fixed amount of 4 mM
concentration of AMP.
Synthesis of 1-(Vinylbenzyl)thymine. The synthesis of 1-
(vinylbenzyl)thymine (VBT) (Figure 2A) was performed as previously
described.22 Briefly, 5 g (39.6 mmol) of thymine was dissolved in 50
mL of NaOH solution (2.42 M). The mixture was heated to 65 °C
under magnetic stirring. 25.7 mg (0.12 mmol) of the inhibitor 2,6-di-
tert-butyl-4-methylphenol and 50 mL of ethanol were added and left
under stirring for a further 15 min, followed by the addition of 5.6 mL
(39.7 mmol) of 4-vinylbenzyl chloride. The mixture was refluxed at 85
°C for 18 h and then left to cool to room temperature, and the ethanol
was evaporated under vacuum. The suspension was diluted with 150
mL of water, and the pH was adjusted to 7.0 so as to obtain a
Figure 4. Transmittance−temperature curve of NIP-NPs in 25 mM
MES buffer, pH 5.5.
Immobilization of IDA-Fe3+ Metal-Chelate on Glass Beads.
All GBs were first activated by boiling in 4 M NaOH (typically 100 g
of GBs in 100 mL of NaOH) for 10 min. After activation, they were
washed with water and acetone and then dried in an oven at 50 °C.
The GBs were then functionalized with (3-glycidyloxypropyl)-
trimethoxysilane (Glymo) so as to introduce iminodiacateic acid
(IDA), as previously described.18 Briefly, 60 g of activated GBs was
added to a mixture of 100 mL of anhydrous toluene and 100 mL of
Glymo in a round-bottom flask and refluxed at 110 °C for 17 h. The
Glymo-coupled glass beads (Glymo-GBs) were collected and washed
with toluene and acetone, three times each solvent. After drying, the
Glymo-GBs were immersed into 500 mL of a 0.75 M IDA solution
(containing 0.34 M NaCl and 2.0 M Na2CO3, pH 11.0) and stirred at
70 °C for 17 h. The final GBs were obtained after washing five times
with water.
Solid-Phase Synthesis of MIP-NPs for AMP. The solid-phase
synthesis of MIP-NPs was carried out in a glass column equipped with
a thermostated jacket (XK 26/40, GE Healthcare, Fontenay sous Bois,
France), connected to a circulation thermostated water bath (Bioblock
Scientific polystat 5, Fischer Scientific, France). All solvents were
C
DOI: 10.1021/acs.macromol.7b01782
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
pumped through the column using a peristaltic pump at a flow rate of
2 mL/min. The column was packed with 20 g of functionalized IDA-
GBs and washed with 10 volume column (VC) (200 mL) of water.
Then 10 VC of iron(III) chloride (50 mM), prepared in water, was
passed through the column. Unbound Fe3+ was removed by passing 10
VC water, and the column was equilibrated with 200 mL of buffer
A.21,30 Afterward, 20 mL of a solution of AMP (4 mg/mL buffer A)
was loaded on the column. The flow-through was collected and passed
through the column again for 1 h. Excess AMP was removed by
passing 200 mL of buffer A, until the absorbance at 259 nm was close
to zero.
The polymerization solution was prepared by mixing NIPAm
(123.57 mg, 91 mol %), VBT (11.63 mg, 4 mol %), and EbAm (10.09
mg, 5 mol %) in 29.1 mL of buffer A. The solution was purged with
nitrogen for 30 min. KPS (10.2 mg in 200 μL of buffer A)/TEMED
(0.8 μL) was added, and this mixture was percolated through the
column. The temperature was set to 39 °C, i.e., >LCST, for overnight.
After polymerization, the column was washed with 150 mL (3 × 50
mL) of buffer A at 39 °C and then left to cool down to room
temperature. The MIP-NPs were eluted with 20 mL (4 × 5 mL) of
buffer A at room temperature.
Control nonimprinted nanoparticles (NIP-NPs) and MIP-NPs
devoid of thymine group were synthesized using the same protocol as
described above, except that AMP was not immobilized on the GBs
and VBT was absent in the polymerization mixture, respectively. All
eluted fractions were subjected to absorbance measurements at 259
nm for verifying eventual leakage of AMP by using its calibration curve
(Figure S3A). The absence of Fe3+ (see protocol in the next section)
was verified by using the calibration curve of the iron(III)−EDTA−
H2O2−NH3 complex at 513 nm (Figure S4).
Particle Size Determination. The hydrodynamic size of the MIP-
NPs was measured directly on the eluate fractions from the column by
dynamic light scattering (DLS) at 25 °C, using a Zeta-sizer NanoZS
(Malvern Instruments Ltd., Worcestershire, UK). DLS analysis was
performed on all eluted fractions, and MIPs having similar sizes and
dispersities (first three fractions, total 15 mL) were pooled. For
transmission electron microscopy (TEM) imaging, an aliquot of the
pooled fraction was dialyzed overnight against Milli-Q water. TEM
images were captured using a JEM-2100F (JEOL, Japan). The TEM
grid was a 300 mesh carbon-coated copper grid from AGAR Scientific
(Stansted, UK).
particles (0.02−0.4 mg/mL) were pipetted in separate 2 mL
polypropylene microcentrifuge tubes, and the analytes (final
concentration 50 μM) were added. The final volume was adjusted
to 1 mL with buffer A, and the mixture was incubated overnight on a
tube rotator in a thermostated chamber at 40 °C. The samples were
centrifuged at 41415g for 1 h, and the supernatant was placed in a
quartz cuvette for recording of their absorbance. The amount of
analyte bound to the polymers was calculated by subtracting the
amount of unbound analyte from the initial amount of analyte added
(λmax = 259 nm for AMP, adenosine, ATP, and cAMP, λmax = 252 nm
for GMP, λmax = 271 nm for CMP, and λmax = 262 nm for UMP). The
calibration curves of the different analytes are shown in Figure S3A−G.
3. RESULTS AND DISCUSSION
Initial Strategy Using Amidinium Moiety To Orientate
Phosphate Groups. Because of their ubiquitous nature,
nucleotides have relevant roles in biological processes and are
among the most targeted anionic molecules in supramolecular
chemistry for developing biomimetic artificial receptors.1,2
Natural receptors like proteins selectively recognize the
phosphate moiety of nucleotides via ammonium (lysine
amino acid) and guanidinium (arginine amino acid). Therefore,
polyammonium-based and more recently guanidinium-based
host receptors targeting AMP, ADP, and ATP with high
binding constants of up to 107 M−1 in aqueous media have been
reported.1,2 Like guanidine,32 amidine groups are superbases
with pKa of ∼12 in water33 and can form strong stoichiometric
electrostatic interactions with carboxylates34 and phosphates.35
Recently, we have synthesized a benzamidine-based monomer
((4-acrylamidophenyl)(amino)methaniminium acetate (AB))
(Figure 3), which forms a 1:1 AB-carboxylate template complex
with a high association constant (Ka > 900 M−1),34 in CD3OD/
25,36
1
D2O (4/1), as determined by H NMR spectroscopy.
The
resulting MIPs displayed high specificity and selectivity for their
target molecules in aqueous conditions.
Encouraged by the results obtained with carboxylate targets,
we hypothesized that AB would also form strong interaction
with phosphate moieties. Thus, in our initial work, we
synthesized MIPs for AMP by precipitation polymerization
using AB, VBT, and ethylene glycol dimethacrylate as
functional and cross-linking monomers in CH3OH/H2O (4/
1) as porogen (see details in Supporting Information). Thermal
polymerization, instead of UV light, was employed to avoid
photodimerization of the thymine moieties.37 Prior to syn-
thesis, the strength and stoichiometry of the AB−AMP complex
were studied in CD3OD−D2O (4/1), using 31P NMR
spectroscopy (see details in Supporting Information). The
Job plot and titration experiments predict the formation of a
1:2 AMP:AB complex (Figure S5B) with an overall binding
constant, β12 = Ka1Ka2, of 300 M−2 (Figure S5C), where Ka1 and
Ka2 are the association constants of the interaction of AB with
the first and second interaction site of AMP, respectively.
Unexpectedly, the association constant was relatively low (Ka
must be ≥900 M−1 to establish 1:1 stoichiometric noncovalent
interactions)34 and suggests that the conformation adopted by
AMP during interaction with AB may not completely favor
binding of the phosphate with the amidinium moiety but might
form weaker Π−Π interaction between the aromatic ring of AB
with the nucleobase of AMP. VBT contains a complementary
thymine residue and was included in the polymerization
mixture to mimic the base pair in DNA and hence favor pairing
with the adenine of AMP by hydrogen bonding. It was
synthesized from thymine and 4-vinylbenzyl chloride in one
step, with a yield of 47%, using a procedure adapted from a
Determination of Fe3+ in the Washing and Eluted Fractions.
The determination of iron(III) was adapted from Poeder and co-
workers.31 100 μL of each fraction was added to 790 μL of a 100 mM
ethylenediaminetetraacetic acid (EDTA) solution, followed by 60 μL
of H2O2 (35%) and 50 μL of ammonia solution (28%). The
absorbance of the iron(III)−EDTA−H2O2−NH3 complex which has a
deep violet coloration was measured at 513 nm. A calibration curve
prepared with 0.1−1 mM FeCl3 in water in a final volume of 1 mL was
used for quantification (Figure S4).
Equilibrium Binding Studies and Selectivity Test. A cleanup
procedure of the MIP-NPs was essential before doing the binding
experiments. The pooled MIP-NPs fractions eluted from the column
(15 mL) were subjected to repeated (4 times) concentration and
dilution with buffer A, using an Amicon Ultra-15 centrifugal filter unit
with MWCO 30 kDa (Millipore, Molsheim, France) in order to
remove tiny particles, which interfere with the absorbance values of the
nucleotides, during read-out on a spectrophotometer. The final
retentate was then diluted to 15 mL with buffer A, and this constituted
our stock of working MIP-NPs. Its concentration (mg/mL) was
determined by taking 2 mL and centrifuging at 40000g for 1 h at 40 °C
to precipitate the polymer NPs. The centrifuge machine was left for 30
min to equilibrate to 40 °C prior to use. After discarding the
supernatant, the precipitate was resuspended in 2 mL of water and
lyophilized. The dry MIP-NPs were weighed with a precision balance;
this allows to determine the concentration of the MIP-NPs (mg/mL)
and the yield of polymerization, calculated as mg of MIP-NPs per g of
GBs.
1 mM stock solutions of AMP, GMP, CMP, UMP, ATP, adenosine,
and cAMP were prepared in buffer A and kept at 4 °C. The polymer
D
DOI: 10.1021/acs.macromol.7b01782
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
previous report22 (Figure 2A). The resulting MIPs were very
specific for AMP (negligible NIP binding) but were not at all
selective since CMP and GMP were bound to the same extent
as AMP in water (Figure S6A). This nonselectivity indicates
that pairing was not or poorly achieved. The reason, as
suggested above, might be due to nonspecific interactions like
Π−Π binding, which occurs between AB and the two
nucleobases. By using a solid-phase strategy (Figure S1)
where the immobilized amidinium moiety was first constrained
to present an orientation favorable for phosphate interaction
with AMP, then followed by subsequent interaction of VBT to
react with the nucleobase of AMP, improved the selectivity to
some extent; the MIP did not bind CMP at all, but bound
GMP and AMP, though AMP to a higher extent, indicating that
the MIP could not distinguish between the two purine
nucleotides (Figure S6B).
No leakage of Fe3+ and AMP was detected in the MIP-NPs
(limit of quantification (LOQ): 100 μM for Fe3+ (Figure S4)
and 5 μM for AMP (Figure S3A)). Control NIP-NPs and MIP-
NPs without VBT were synthesized following the same
protocol as described above but in the absence of AMP and
in the presence of AMP but without VBT in the polymerization
mixture, respectively.
Physicochemical Characterization of MIP-NPs. MIP-
NPs were obtained as a transparent solution, and no
aggregation was observed for a period of six months when
stored in a refrigerator. The hydrodynamic size and
morphology of the MIP-NPs were analyzed by dynamic light
scattering (DLS) and transmission electron microscopy (TEM)
(Figure S7), respectively. The diameters of MIP-NPs, NIP-NPs,
and MIP-NPs without VBT were 55.9 2.6, 57.5 0.8, and
69.8 4.7 nm (n = 3), with a polydispersity index (PDI) of
∼0.3, respectively, as measured by DLS at 25 °C. The MIP-NPs
appear as spherical particles in the TEM image.
We then switched to another affinity ligand, devoid of
aromatic groups. The new ligand was based on immobilized
Fe(III) ion which enabled the coupling of the nucleotide via its
phosphate group.21
Evaluation of the Binding Characteristics of MIP-NPs.
The recognition behavior of MIP-NPs for AMP (50 μM) was
determined by equilibrium binding studies in free solution.
Figure 5A shows the high specificity of the MIP as no binding
Solid-Phase Synthesis of MIP-NPs Using IDA-Fe3+ as
Affinity Ligand. Figure 1 shows the steps used for the
immobilization of AMP on GBs. Briefly, GBs were activated by
boiling in NaOH, followed by silanization with Glymo and
coupling with IDA, as previously described.18,38 IDA-GBs were
then packed in a column, equipped with two adapters for
regulation of the bed volume and a thermostated jacket. A
solution of 50 mM FeCl3 in water was then pumped through.
After washing the excess Fe3+ with water and equilibration with
buffer A, AMP was loaded on the column. AMP adsorbs to
immobilized Fe3+ at pH 5.5, through its free terminal phosphate
group.21 The strong affinity of Fe3+ for phosphorylated
biomolecules (peptides, proteins, amino acids) was already
reported.23,30 At this stage, it is important to point out that
phosphate buffer, which was precedently employed as
equilibrium and working buffer,17−19,38 was substituted here
by MES buffer to avoid interference of the phosphate groups of
the buffer with that of AMP for binding with Fe3+.
A polymerization mixture, composed of NIPAm, VBT, and
EbAm (91/4/5, molar ratio) and the initiator system (KPS,
TEMED) in buffer A, was then passed on the column. NIPAm
is a functional monomer,39 capable of hydrogen bond
interaction, due to the presence of oxygen and nitrogen
atoms40 but at the same time was used herein as the major
component in this polymer recipe so as to obtain
thermoresponsive MIP-NPs. The lower critical solution
temperature of the polymer had been previously determined
via turbidity measurement (cloud point determination) by
measuring the transmittance versus temperature using a UV−
vis spectrophotometer and found to be ∼32 °C (Figure 4). The
imprinted polymer was synthesized at 39 °C, and the growing
polymeric nanoparticles, which are in the collapsed state at this
temperature, encapsulate the immobilized AMP during
polymerization.
After polymerization, the column reactor was washed with
buffer A at 39 °C to remove unreacted reagents and particles
that polymerized distantly from the immobilized template. The
reactor was then washed with the same buffer at room
temperature (25 °C), allowing the MIP-NPs to swell and be
eluted. The yield of polymerization of the MIP-NPs was 0.65
0.17 mg per g of GBs (n = 3), comparable to values previously
obtained with MIP-NPs synthesized by the solid-phase method,
using Glymo coupling and metal-chelate as affinity ligand.18,38
Figure 5. Equilibrium binding isotherms in buffer A of 50 μM AMP
(circle), GMP (diamond), CMP (triangle), and UMP (cross) on MIP-
AMP (filled) and NIP-AMP (open). Polymers were obtained in the
presence (A) and absence (B) of VBT; (C) adenosine (star) and ATP
(square). Data are means from three independent experiments with
three different batches of polymers. The error bars represent standard
deviations.
E
DOI: 10.1021/acs.macromol.7b01782
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
of the NIP toward AMP was observed. The MIP was also very
selective toward AMP as it did not bind GMP, CMP, and UMP.
This selectivity is remarkable as the two purine nucleobases,
AMP and GMP, are quite similar in size, shape, and
hydrophobicity. It is important to note that the selectivity
was partially lost when VBT was not present in the
polymerization mixture (Figure 5B). The MIP could not
differentiate between AMP and GMP, but it did not bind CMP
or UMP, indicating that the thymine moiety in the polymer
plays an essential role in the recognition procedure. The
binding capacity of the MIP for AMP was determined by
plotting a graph of B (bound) versus F (free) (Figure S8). The
nonlinear fitting of the data to a single-site Langmuir binding
isotherm showed a maximum binding capacity (Bmax) of 1350
nmol/mg of MIP and a dissociation constant (Kd) of 74.5 μM,
indicating that high binding capacity and affinity were achieved
by using the dual-oriented solid-phase imprinting strategy.
Interestingly, the MIP could recognize adenosine (Figure
5C) to the same extent as AMP. Adenosine plays an important
role in neuroprotection and brain functions activity and is a
biomarker of cancer.8,24 On the other hand, negligible and no
binding were observed with ATP (Figure 5C) and cAMP,
respectively. The bent conformation adopted by ATP due to its
three phosphate groups and the additional phospholinkage in
cAMP, probably prevent the fitting of their adenine base with
the thymine group within the MIP. Furthermore, the two −OH
groups of the ribose moiety are very apt to form hydrogen
bonding with NIPAm40 which gave additional structuration to
the binding cavity, explaining why cAMP cannot go in.
tion, detection, and delivery of nucleotides and nucleosides in
aqueous media.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.macro-
mol.7b01782.
Immobilization of PAB on GBs, solid-phase synthesis of
MIP-NPs for AMP, on PAB-immobilized GBs and
equilibrium binding studies of these MIP-NPs with
AMP and other nucleotides, 31P NMR studies of AB−
1
AMP complex, H NMR spectrum of VBT, calibration
curves of nucleotides and FeCl3, determination of size,
morphology, and binding capacity of MIP-NPs obtained
on IDA-Fe3+ (PDF)
AUTHOR INFORMATION
Corresponding Authors
*E-mail jeanne.tse-sum-bui@utc.fr (B.T.S.B.).
*E-mail karsten.haupt@utc.fr (K.H.).
■
ORCID
Bernadette Tse Sum Bui: 0000-0002-4170-2303
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
C. Mourao thanks the Brazilian National Council for Scientific
̃
and Technological Development and J. Xu thanks the china
scholarship council (CSC) for scholarships. We thank Frederic
Nadaud for the TEM images. This work was finally supported
by the European Regional Development Fund ERDF and the
Region of Picardy (CPER 2007-2013), the European
Commission (FP7Marie Curie Actions, project SAMOSS,
PITN-2013-607590).
4. CONCLUSIONS
Template-free water-soluble MIP-NPs targeting AMP were
prepared by solid-phase synthesis using for the first time
immobilized Fe(III)-chelate as affinity ligand to attract the
AMP by its phosphate group. A polymerizable thymine
monomer, which is complementary to the adenine moiety of
the nucleotide, was synthesized and incorporated in the MIP
polymerization mixture so as to constrain the AMP in a dual-
orientated configuration. The resulting MIP was very selective
toward AMP as it did not bind other nucleotides, GMP, CMP,
and UMP in buffer. This remarkable selectivity of the MIP for
its target molecule in the presence of surrounding water
molecules and counterions which compete with the donor and
acceptor sites indicates the creation of high fidelity binding sites
within the polymer. Noteworthy is the fact that the choice of
the affinity ligand is of prime importance as immobilized p-
aminobenzamidine which contains an aromatic moiety did not
lead to selective MIPs, probably due to Π−Π stacking with the
nucleobase of AMP.
REFERENCES
■
(1) Tamaru, S. I., Hamachi, I. In Structure and Bonding - Recognition of
Anions; Vilar, R., Ed.; Springer-Verlag: Berlin, 2008; Vol. 129, p 95.
(2) Kuchelmeister, H. Y.; Schmuck, C. In Designing Receptors for the
Next Generation of Biosensors, 1st ed.; Piletsky, S. A., Whitcombe, M. J.,
Eds.; Springer-Verlag: Berlin, 2013; Vol. 12, p 53.
(3) Haupt, K.; Linares, A. V.; Bompart, M.; Tse Sum Bui, B.
Molecularly Imprinted Polymers. Top. Curr. Chem. 2011, 325, 1−28.
(4) Chen, L.; Wang, X.; Lu, W.; Wu, X.; Li, J. Molecular Imprinting:
Perspectives and Applications. Chem. Soc. Rev. 2016, 45, 2137−2211.
(5) Wackerlig, J.; Schirhagl, R. Applications of Molecularly Imprinted
Polymer Nanoparticles and Their Advances toward Industrial Use: A
Review. Anal. Chem. 2016, 88, 250−261.
Interestingly, the MIP could bind adenosine, AMP without
the phosphate group, to the same extent as AMP. Using the
phosphate group of the AMP as a hinge enables free pairing of
the nucleobase with a complementary base-pair residue
monomer. This strategy can be easily extended to the
preparation of selective water-soluble MIPs for other
nucleotides, nucleosides, and potentially oligonucleotides,
provided that there is no limitation of the synthetic availability
of the complementary polymerizable base monomer. We
believe that the development of cost-effective and chemically
and physically stable artificial receptors like MIPs will be of
great interest to different research areas of medicine,
biochemistry, and separation science for recognition, sequestra-
(6) Tse Sum Bui, B.; Haupt, K. Molecularly Imprinted Polymers:
Synthetic Receptors in Bioanalysis. Anal. Bioanal. Chem. 2010, 398,
2481−2492.
́ ́ ́
(7) Breton, F.; Delepee, R.; Jegourel, D.; Deville-Bonne, D.;
Agrofoglio, L. A. Selective Adenosine-5′-Monophosphate Uptake by
Water-Compatible Molecularly Imprinted Polymer. Anal. Chim. Acta
2008, 616, 222−229.
́ ́
(8) Krstulja, A.; Lettieri, S.; Hall, A. J.; Delepee, R.; Favetta, P.;
Agrofoglio, L. A. Evaluation of Molecularly Imprinted Polymers Using
2′,3′,5′-Tri-O-Acyluridines as Templates for Pyrimidine Nucleoside
Recognition. Anal. Bioanal. Chem. 2014, 406, 6275−6284.
(9) Spivak, D. A.; Shea, K. J. Investigation into the Scope and
Limitations of Molecular Imprinting with DNA Molecules. Anal. Chim.
Acta 2001, 435, 65−74.
F
DOI: 10.1021/acs.macromol.7b01782
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
(10) Yano, K.; Tanabe, K.; Takeuchi, T.; Matsui, J.; Ikebukuro, K.;
Karube, I. Molecularly Imprinted Polymers which Mimic Multiple
Hydrogen Bonds between Nucleotide Bases. Anal. Chim. Acta 1998,
363, 111−117.
(11) Sallacan, N.; Zayats, M.; Bourenko, T.; Kharitonov, A. B.;
Willner, I. Imprinting of Nucleotide and Monosaccharide Recognition
Sites in Acrylamidephenylboronic Acid-Acrylamide Copolymer
Membranes Associated with Electronic Transducers. Anal. Chem.
2002, 74, 702−712.
(12) Kanekiyo, Y.; Sano, M.; Iguchi, R.; Shinkai, S. Novel Nucleotide-
Responsive Hydrogels Designed from Copolymers of Boronic Acid
and Cationic Units and their Applications as a QCM Resonator
System to Nucleotide Sensing. J. Polym. Sci., Part A: Polym. Chem.
2000, 38, 1302−1310.
́ ́
(13) Breton, F.; Delepee, R.; Agrofoglio, L. A. Molecular Imprinting
of AMP by an Ionic-Noncovalent Dual Approach. J. Sep. Sci. 2009, 32,
3285−3291.
(14) Chen, Y.; Li, X.; Yin, D.; Li, D.; Bie, Z.; Liu, Z. Dual-Template
Docking Oriented Molecular Imprinting: a Facile Strategy for Highly
Efficient Imprinting within Mesoporous Materials. Chem. Commun.
2015, 51, 10929−10932.
(15) Canfarotta, F.; Poma, A.; Guerreiro, A.; Piletsky, S. Solid-Phase
Synthesis of Molecularly Imprinted Nanoparticles. Nat. Protoc. 2016,
11, 443−455.
(16) Chen, L.; Muhammad, T.; Yakup, B.; Piletsky, S. A. New
Immobilisation Protocol for the Template Used in Solid-Phase
Synthesis of MIP Nanoparticles. Appl. Surf. Sci. 2017, 406, 115−121.
(17) Ambrosini, S.; Beyazit, S.; Haupt, K.; Tse Sum Bui, B. Solid-
Phase Synthesis of Molecularly Imprinted Nanoparticles for Protein
Recognition. Chem. Commun. 2013, 49, 6746−6748.
(18) Xu, J.; Ambrosini, S.; Tamahkar, E.; Rossi, C.; Haupt, K.; Tse
Sum Bui, B. Toward a Universal Method for Preparing Molecularly
Imprinted Polymer Nanoparticles with Antibody-Like Affinity for
Proteins. Biomacromolecules 2016, 17, 345−353.
(29) Saeed, A.; Georget, D. M. R.; Mayes, A. G. Synthesis,
Characterisation and Solution Thermal Behaviour of a Family of
Poly(N-Isopropyl Acrylamide-Co-N-Hydroxymethyl Acrylamide) Co-
polymers. React. Funct. Polym. 2010, 70, 230−237.
(30) Muszynska, G.; Dobrowolska, G.; Medin, A.; Ekman, P.; Porath,
J. O. Model Studies on Iron(III) Ion Affinity Chromatography. II.
Interaction of Immobilized Iron(III) Ions with Phosphorylated Amino
Acids, Peptides and Proteins. J. Chromatogr. 1992, 604, 19−28.
(31) Poeder, B. C.; den Boff, G.; Franswa, C. E. M. Selective
Spectrophotometric Determination of Iron(III) with EDTA. Anal.
Chim. Acta 1962, 27, 339−344.
́
(32) Blondeau, P.; Segura, M.; Perez-Fernandez, R.; de Mendoza, J.
Molecular Recognition of Oxoanions Based on Guanidinium
Receptors. Chem. Soc. Rev. 2007, 36, 198−210.
(33) Margetic, D. In Superbases for Organic Synthesis: Guanidines,
Amidines, Phosphazenes and Related Organocatalysts, 1st ed.; Ishikawa,
T., Ed.; John Wiley & Sons, Ltd.: 2009; p 9.
(34) Wulff, G.; Knorr, K. Stoichiometric Noncovalent Interaction in
Molecular Imprinting. Bioseparation 2001, 10, 257−276.
(35) Sebo, L.; Diederich, F.; Gramlich, V. Tetrakis-
(phenylamidinium)-Substituted Resorcin[4]arene Receptors for the
Complexation of Dicarboxylates and Phosphates in Protic Solvents.
Helv. Chim. Acta 2000, 83, 93−113.
(36) Panagiotopoulou, M.; Salinas, Y.; Beyazit, S.; Kunath, S.; Mayes,
A. G.; Duma, L.; Prost, E.; Resmini, M.; Tse Sum Bui, B.; Haupt, K.
Molecularly Imprinted Polymer-Coated Quantum Dots for Multi-
plexed Cell Targeting and Imaging. Angew. Chem., Int. Ed. 2016, 55,
8244−8248.
(37) Mikulska, A.; Inoue, M.; Kuroda, K.; Iwanowska, A.; Yusa, S. I.;
Nowakowska, M.; Szczubiałka, K. Polymeric/Silicagel Hybrid Molec-
ularly Photoimprinted Adsorbents for Adenosine and its Derivatives.
Eur. Polym. J. 2014, 59, 230−238.
(38) Xu, J.; Medina-Rangel, P. X.; Haupt, K.; Tse Sum Bui, B. Guide
to the preparation of molecularly imprinted polymer nanoparticles for
protein recognition, by solid-phase synthesis. Methods Enzymol. 2017,
590, 115−141.
(19) Xu, J.; Haupt, K.; Tse Sum Bui, B. Core−Shell Molecularly
Imprinted Polymer Nanoparticles as Synthetic Antibodies in a
Sandwich Fluoroimmunoassay for Trypsin Determination in Human
Serum. ACS Appl. Mater. Interfaces 2017, 9, 24476−24483.
(20) Guerreiro, A.; Poma, A.; Karim, K.; Moczko, E.; Takarada, J.; de
Vargas-Sansalvador, I. P.; Turner, N.; Piletska, E.; Schmidt de
(39) Hien Nguyen, T.; Ansell, R. J. N-isopropylacrylamide as a
functional monomer for noncovalent molecular imprinting. J. Mol.
Recognit. 2012, 25, 1−10.
(40) Pelton, R. Poly(N-isopropylacrylamide) (PNIPAM) is Never
Hydrophobic. J. Colloid Interface Sci. 2010, 348, 673−674.
Magalhaes, C.; Glazova, N.; Serkova, A.; Omelianova, A.; Piletsky, S.
̃
Influence of Surface-Imprinted Nanoparticles on Trypsin Activity. Adv.
Healthcare Mater. 2014, 3, 1426−1429.
(21) Dobrowolska, G.; Muszynska, G.; Porath, J. Model Studies on
Iron(III) Ion Affinity Chromatography Interaction of Immobilized
Metal Ions with Nucleotides. J. Chromatogr. 1991, 541, 333−339.
(22) Cheng, C. M.; Egbe, M. I.; Grasshoff, J. M.; Guarrera, D. J.; Pai,
R. P.; Warner, J. C.; Taylor, L. D. Synthesis of 1-(vinylbenzyl)thymine,
a novel, versatile multi-functional monomer. J. Polym. Sci., Part A:
Polym. Chem. 1995, 33, 2515−2519.
(23) Holmes, L. D.; Schiller, M. R. Affinity Chromatography for the
Separation of Phosphorylated Macromolecules: Ligands and Applica-
tions. J. Liq. Chromatogr. Relat. Technol. 1997, 20 (1), 123−142.
(24) Poma, A.; Brahmbhatt, H.; Watts, J. K.; Turner, N. W.
Nucleoside-Tailored Molecularly Imprinted Polymeric Nanoparticles
(MIP NPs). Macromolecules 2014, 47, 6322−6330.
(25) Nestora, S.; Merlier, F.; Beyazit, S.; Prost, E.; Duma, L.; Baril, B.;
Greaves, A.; Haupt, K.; Tse Sum Bui, B. Plastic Antibodies for
Cosmetics: Molecularly Imprinted Polymers Scavenge Precursors of
Malodors. Angew. Chem., Int. Ed. 2016, 55, 6252−6256.
(26) Renny, J. S.; Tomasevich, L. L.; Tallmadge, E. H.; Collum, D. B.
Method of Continuous Variations: Applications of Job Plots to the
Study of Molecular Associations in Organometallic Chemistry. Angew.
Chem., Int. Ed. 2013, 52, 11998−12013.
(27) Connors, K. A. In Binding Constants − The Measurement of
Complex Stability; John Wiley & Sons: 1987.
(28) Thordarson, P. Determining Association Constants from
Titration Experiments in Supramolecular Chemistry. Chem. Soc. Rev.
2011, 40, 1305−1323.
G
DOI: 10.1021/acs.macromol.7b01782
Macromolecules XXXX, XXX, XXX−XXX