semi-covalent surface molecular imprinting of polymers by one-stage mini-emulsion polymerization:...

Full Paper

596

Semi-Covalent Surface Molecular Imprinting ofPolymers by One-Stage Mini-emulsionPolymerization: Glucopyranoside as a Model Analytea

Pasquale Curcio, Christelle Zandanel, Alain Wagner, Charles Mioskowski,Rachid Baati*

Dedicated to Charles Mioskowski, who passed away on June 2nd 2007

P. Curcio, C. Zandanel, A. Wagner, R. Baati, C. MioskowskiLaboratoire de Chimie des Systemes Fonctionnels, UMR 7199,Faculte de Pharmacie, Universite de Strasbourg, 74 Route du Rhin,67401 Illkirch, FranceFax: þ33 3-90-24-43-06; E-mail: [email protected]

a : Supporting information for this article is available at the bottom ofthe article’s abstract page, which can be accessed from the journal’shomepage at http://www.mbs-journal.de, or from the author.

This paper describes a new type of surface imprinting technique that combines the advantages ofboth the semi-covalent approach and one-stage miniemulsion polymerization. This process hasbeen successfully applied for the preparation of glucose surface-imprinted nanoparticles. Theselective artificial receptors for glucopyranoside were fully characterized by IR, TEM and BETanalyses, and their molecular recognition abilities by binding experiments carried out in batchprocesses. The molecular affinity and selectivity of the glucose molecularly imprinted polymerswere accurately quantified. These characteristics are essential for verification of the efficiency ofthe developed surface imprinting process. The imprinting effect was clearly demonstrated usingthe batch rebindingmethod.We have found that the glucose imprinted polymers produced usingthe optimized one-stage mini-emulsion exhibited quite fast kinetics of binding and equilibrationwith glucopyranoside templates, compared to polymers prepared by bulk polymerization tech-nique, as well as extremely low levels of unspecific bindings. We also demonstrated that glucosemolecular imprinted polymer (MIP) exhibited very good selectivity for its original templatecompared to other glycopyranoside derivatives, such as galactose. Finally, the extraction of thebinding properties from isotherms of binding by fitting to the bi-Langmuir and Freundlichmodelsallowed the determination of the affinity constant distribution of the binding sites. Thisimprinting protocol allowed thedetermination of an affinity constant(KD), involving exclusively H-bondinginteractions, for the glucoseMIP (P2C)with the best template 1, in CH3CN asthe solvent system.

Macromol. Biosci. 2009, 9, 596–604

� 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Introduction

Molecular imprinting is a technique that enables the

preparation of synthetic receptors in a highly cross-linked

polymer matrix.[1] The imprinted solid material ideally

contains cavities that have a shape and functional groups

DOI: 10.1002/mabi.200900056

Semi-Covalent Surface Molecular Imprinting of Polymers by One-Stage . . .

complementary to the imprinted template molecule. This

technology has found numerous potential applications in

synthesis, drug screening and delivery, catalysis, solid

phaseextraction (SPE), asartificial enzymesandasantibody

mimics.[2–5] Generally, two different approaches, based on

covalent[6–7] and non-covalent imprinting,[8] pioneered by

Wulff and Mosbach, respectively, are used for the prepara-

tionof artificial receptors,mainlybybulkpolymerizationof

vinylic monomers mixtures by free radical initiation. This

technique of polymerization provides, after grinding and

sieving, molecular imprinted polymers (MIPs) with large

size particles distribution,[9,10] slow binding kinetics,[11]

high non-specific binding interactions,[12] and a wide

distribution of affinity constants.[13–15]

To overcome these limitations, many efforts have been

devoted towards the development of other polymerization

techniques, as well as the development of new imprinting

approaches. As a consequence, several polymerization

methods, such as precipitation and emulsion polymeriza-

tion, have been optimized for molecular imprinting

purposes and have enabled the production of materials

with more homogeneous and monodispersed micro- and

nano-sphere particles.[16,17] According to recent reports by

Whitcombe and Tong,[18,19] who studied the non-covalent

imprinting of, respectively, cholesterol and proteins, mini-

Figure 1. Schematic representation of the surface semi-covalent imprinting approach bymini-emulsion polymerization.

emulsion polymerization has emerged as

a promising technique for the surface

imprinting of molecules of biological

interests. Besides this improvement, a

semi-covalent approach that combines

the advantage of both covalent and non-

covalent strategies has also appeared for

controlling the microenvironment of

imprinted polymers.[20] Indeed, the

semi-covalent strategy relies on the

polymerization of a template that is

covalently bound to a functional mono-

mer by a cleavable bond. After copoly-

merization, the template removal leaves

an imprint bearing functional groups,

which are capable of interactingwith the

template in a non-covalent sense in the

rebinding step.[20b] This alternative, using

carboxylic esters as first introduced by

Sellergren, appears to be a promising tool

for the controlled tailoring of homoge-

neous binding sites, decreasing the non

specific binding interactions.[20a] The

post-modification of an existing material

has also been studied in order to improve

the binding properties of MIPs.[14d]

Despite these improvements, the pre-

paration of highly selective MIP based

receptors with homogeneous binding



sites, exhibiting good affinity and selectivity for a template

molecule is still a long standing challenge.[14c] In the course

of our investigations on the development of molecular

imprinted polymers, we hypothesized that a novel

combination of the semi-covalent imprinting approach

using carboxylic esters, associated with a one-stage mini-

emulsion polymerization, would afford surface imprinted

nanoparticles with improved binding affinity and selectiv-

ity, compared to MIPs obtained by bulk polymerization.

Herein, we disclose this novel semi-covalent surface

imprinting approach, that allows the preparation of small

narrow sized imprinted nanospheres, by using a one-stage

mini-emulsion polymerization with a polymerizable-sur-

factant template T (Figure 1).

In this process, the distinguishing characteristic of this

type of imprinting is that the template T, in addition to its

conventional function, is linked covalently to the polymer

surface through ester linkage (after copolymerization),

avoiding the instability of the prepolymerization complex

encountered in the non-covalent approach.[8] The surfac-

tantnatureof the templateT forces it tobepositionedat the

interface between the organic and the aqueousphase of the

‘‘oil-in-water’’mini-emulsion. This feature therefore allows

the imprint to be realized on the surface of the polymer

particles, after cleavage and removal of the template

www.mbs-journal.de 597

P. Curcio, C. Zandanel, A. Wagner, C. Mioskowski, R. Baati

598

(Figure 1). Since the selective recognition of carbohydrates

still stands as one of the biggest challenges in chemical

biology, we studied the imprinting of glucopyranoside as a

model analyte. Moreover, to date monosaccharides and

polysaccharides have only been imprinted by bulk poly-

merization techniques[6,7,21–28] or by polymer hydrogels[29]

providing, therefore, an appropriate systems for probing

the efficiency of our mini-emulsion/semi-covalent

approach.

Experimental Part

Materials

Surfactant polymerizable templates T1, monomer T2 (Figure 2) and

analytes 1–4 (Figure 5) were first synthesized (see the Supporting

Information). 1H NMR and 13C NMR spectra of new compounds

were recorded on Bruker DPX200 and DPX300 instruments.

Mass spectrawere recorded on aMS/MS high-resolutionMicromass

ZABSpecTOFspectrometer. IRspectraofsamplesweretakenusinga

Perkin-Elmer 2000 FT-IR spectrometer. All reactions were carried

out under argon. Organic solvents were distilled under an argon

atmosphere using standard drying agents or stored overmolecular

sieves (4 A).All startingmaterialwerepurchasedcommerciallyand

usedwithout furtherpurification.Hexadecaneandsodiumdodecyl

sulfate (SDS) were purchased from Acros Organics. Styrene,

divinylbenzene (DVB), were distilled before use, and 2,20-azo (2-

methylbutyronitrile) (AMBN), were purchased from Aldrich. All

other chemicals used were analytical or HPLC reagent grade.

Preparation of Imprinted Polymers and Control

Polymers

The general procedure was as follows. The template T1 (52mg,

0.072mmol), styrene (1.4ml, 12.4mmol), divinylbenzene (0.35ml,

2.5mmol), hexadecane (0.1ml, 0.34mmol), radical initiator 2,20-

azo(2-methylbutyronitrile) (AMBN, 16mg, 0.083mmol), sodium

dodecyl sulfate (SDS, 63mg, 0.26mmol) and water (8mL) were

Figure 2. Structure of the polymerizable surfactant template (T1)and surfactant monomer (T2).



stirred vigorously for 15min at room temperature under an argon

atmosphere. Miniemulsification was then obtained by sonication

using a Bioblok Scientific Vibra-cell, Model 72412 at 40% power for

2min at room temperature. The polymerizationwas carried out by

heating the reaction mixture at 80 8C for 20h. At the end of the

reaction, the latex was diluted in acetone and the solid was

separated by ultracentrifugation, at 27 000 tr �min�1, washed

intensively with a mixture of water/acetone (1:1, v/v) and dried

in vacuum overnight to afford polymer P2.

The control polymer P4 was prepared by the same method by

replacing the template T1 by four equivalents of the surfactant

monomer T2 in order to have the same theoretical quantity of

carboxylic acids at the surface of the polymer P4C. The control

polymerP3waspreparedusing the samecompositionandprotocol

as for P2 but without the template T1.

TEM Microscopy

The morphological characterization of the functionalized nano-

spheres was carried out with transmission electron microscopy

(TEM) observation (Philips CM-120 microscope operating at

100 kV). The samples were deposited as a water suspension on a

copper grid. After drying by removal of the aqueous phase, the

examinationwas performed directlywithout prior any coloration.

Template Extraction from the Polymers

The general procedure was as follows. For the extraction of the

template T1 and monomer T2, the imprinted nanospheres P1, P2

and P4 were treated with a 1:1 (v/v) mixture of MeOH/NaOH 5M

under reflux for 3 days. The nanospheres were then recovered by

centrifugation and washed intensively with a mixture of water/

MeOH (4:1 v/v). The polymer was then recuperated by centrifuga-

tion, treated with a 2M aqueous solution of HCl, washed several

timewithwater, driedovernight invacuumandfinally lyophilized,

to give the imprinted polymer P1C, P2C or P4C.

Procedure for the Esterification of P2C

20mg of P2C was suspended in a 1:1 (v/v) mixture of toluene/

methanol. Then, at 0 8C, trimethylsilyldiazomethane (110mL, 2 eq/

COOH) was slowly added dropwise. The reaction mixture was

stirredat 0 8C for1 hand slowlyheated to roomtemperature.Acetic

acid (300mL) was added and the mixture was centrifuged. After

several washings and centrifugation with a 1:1 mixture of water/

acetonitrile, the polymer was recovered and dried under vacuum.

5mg of this polymer was then suspended in CH3CN in the

presence of excess template 1 and the binding was measured as

described previously, in the batch process after 24h of incubation.

BET Measurements

The specific surface area of the nanospheres was determined by

gravimetric nitrogen gas adsorptionmeasurements (BETmethod),

using a Carlo Erba Sorpty 1750 instrument. For the BET measure-

DOI: 10.1002/mabi.200900056


ments, the samples were first dried and degassed at 50 8C, untilconstant mass prior analysis.

Light Scattering Measurements

For the measurements of the polydispersity or the particle size

distribution, 1–2mgofpolymerswas suspended in chloroformand

analyzed by Malvern instrument.

Figure 3. TEM pictures of polymers P2 and P2C.
Binding Experiments Using the Batch Process and
HPLC Analysis

The percentage of bindingwasmeasured byHPLC (Shimadzu LC 10

ADVP) using a Thermo Betasil Column (4.6mm5u BS diol 100). For

themeasurements of the binding ability of the polymers, 10mg of

imprinted (P1C and P2C) and control polymers (P3 and P4C) were

dispersed in acetonitrile containing different concentration of

glucopyranoside 1 or analogues (2, 3 or 4) of the original template 1

(Figure 5), at 20–22 8C for the desired incubation time. The particles

were separated by centrifugation and the carbohydrate concen-

tration before and after particle suspension was measured and

quantified by HPLC. The concentration of missing template in the

solution after suspension and removing of the polymer particles

was assigned as that portion being absorbed by the different

polymers (Table 1).

Results and Discussion

The polymerizable surfactant template T1 (Figure 2)

bearing four acrylic esters tethered via a lipophilic chain

to the polar O-sulfonated group was first efficiently

synthesized using afive-step procedure (see the Supporting

Information).

T1 was then dissolved in styrene and divinyl benzene

(DVB), as the functional monomer and the cross-linker,

Table 1. Polymer formulations, physical characterisation and binding

Entry Polymer Hexadecane T1 or

T2/SDS

Template

1 P1 No 25/75 T1

2 P1C – 25/75 T1

3 P2 Yes 25/75 T1

4 P2C – 25/75 T1

5 P3 Yes – None

6 P4 – – T2

7 P4C Yes – T2

a)Measurements made on 1.3–1.5mg of dry polymer dispersed in 1.0

populations of particle size distibution (nm and mm) were observed.



respectively. The soluble radical initiator azomethylbutyr-

onitrile (AMBN) was then added to the resulting homo-

geneous organic phase. An aqueous solution of sodium

dodecylsulphate (SDS) was used as the surfactant system

for the preparation of the mini-emulsion, which was

obtained by vigorous stirring followed by ultrasonication

for 2min at room temperature. The copolymerization

performedwithouthexadecane, as reportedbyWhitcombe,

failed[18] in providing narrow sized particles, and a polymer

(P1)with aggregated and largeparticles size distribution, as

determined by light scatteringmeasurements, was formed

(Table 1, Entry 1).

In contrast, imprinted polymer P2, prepared in the

presence of hexadecane was obtained as colloid with a

rather regular particle size, as shown by TEM pictures

(Figure 3), and with a good and improved polydispersity

measured by light scattering (Table 1, Entry 3).

To determine whether the template molecule T1 was

successfully incorporatedwithin the nanospheres P2, FT-IR

spectroscopyanalyseswereperformed.The IRabsorptionof

the ester functionswith a characteristic band at 1 734 cm�1

is present for P2 and can be easily identified (see

properties.

Particle

diametera)

Polydispersitya) Uptake of Analyte

1 after 24 h

nm mmol � g�1

500–700b) 0.414 –

770b) 0.346 7.5� 0.9

144 0.142 –

154 0.173 17.0� 1

141 0.253 0.7� 0.2

138 0.101 –

141 0.180 1.75� 0.5

mL of chloroform by dynamic light scattering (Malvern); b)Two



Figure 4. Binding experiment in the batch process.

600

the Supporting Information). The concomitant disappear-

ance of the absorption band of C¼C double bond (acrylate

functions) located at 1 638 cm�1 provides evidence for

the incorporation of T1 into the polymer and, therefore, the

copolymerisation of T1 with styrene and DVB. Two control

polymers P3 and P4 were also prepared with exactly the

same procedure as P2, except T1 was not added for both

polymers (Table1, Entrie5and6), andP4wassynthesized in

the presence of the polymerizable surfactant monomer T2

(Figure 2). It should be noted that control polymers P3 and

P4 are required in order to quantify the unspecific binding,

as well as to assess unambiguously the surface imprinting

effect of imprinted polymers. The cleavage of the template

moleculeT1 andmonomerT2 from thepolymersP1 andP2,

andP4, respectively,were thenenvisioned inorder toafford

the ‘‘glucose’’ surface imprinted polymer P1C, P2C and the

control polymer P4C. It is known that the cleavage of ester

bonds in a highly cross-linked polymer matrix obtained by

bulk polymerization is not easy to achieve,[11,20] and

generally drastic reductive conditions are used when basic

treatments are ineffective.[30,31] In our case, the creation of

the hydrophilic carboxylic acid recognition binding sites at

the surface of the nanospheres, was performed by using

usual alkaline hydrolysis.[20a] The effectiveness of the

cleavage/extraction protocol of template T1 from the

corresponding polymers P2, was followed by FT-IR analysis

(see the Supporting Information). The intensity of the ester

absorption band at 1 734 cm�1 decreased for the polymer

P2C, whereas the characteristic carboxylic acid absorption

bands appeared concomitantly at 3 400 cm�1 (stretching of

O�Hbond) and 1 630–1700 cm�1 (stretching of C¼Obond).

At theendof thecleavageprocedure, theabsorptionbandat

1 734 cm�1 of the esters functions was still detectable to

some extent, suggesting that the template could not be

totally removed, as reported for other highly cross-linked

polymers.[11,20]

It is important to note that the distribution particle size

(Table 1, Entry 2, 4 and 7), as well as the morphological

aspect of the imprintednanospheresP1C,P2CandP4Cwere

not affected by the cleavage and extraction procedure, as

can be seen in Figure 3.Nitrogen adsorptionmeasurements

(BET) of the nanospheres revealed a specific area of

55m2 � g�1 for the control polymerP3, and adrastic increase

to about 133m2 � g�1 for P2C, accounting for the modifica-

tion of the surface of thenanoparticles. The efficiency of the

semi-covalent surface imprinting was finally examined

andquantifiedbymeasurements of thepolymers (P1C,P2C,

P3 and P4C) binding properties, by equilibration experi-

ments using the batch rebinding procedure. The results of

these experiments would also assess the imprinting effect

of the polymer P2C compared to control polymers P3 and

P4C.

The binding sites of the glucose imprinted polymer P2C,

theoretically exhibiting four carboxylic acids as specific



recognition elements, are complementary in interaction

and in shape with glucopyranoside. Molecular recognition

of sugars can be assured via non-covalent reversible

interactions such as hydrogen bonds, according to the

semi-covalent approach (Figure 4).[20,31]

We first studied the molecular recognition abilities for

P1C and P2C with the original glucose template 1 by

plotting the concentration of the analyte bound to the

polymers, versus the concentration of the analyte, for 24h

of incubationand for a constantmassofpolymer, at 22 8C inCH3CN. Under these conditions, the maximum uptake of

1 reached a value of 17.0� 1mmol � g�1 of P2C at a

concentration of 6.67mmol (1), while P1C reached a lower

capacity (7.5� 0.9mmol � g�1). The low uptake for P1C is

ascribed to the high level of aggregation observed

previously, precluding a good accessibility and diffusion

of pyranoside 1 inside the binding cavity. We were

encouraged by the fact that extremely low levels of

template 1 uptake was observed by the control polymers

P3 and P4C (0.70� 0.2mmol � g�1 and 1.75� 0.5mmol � g�1,

respectively), suggesting low contributions of non-specific

adsorption, such as hydrophobic and ionic interactions, in

the binding event between carbohydrate 1 with P2C.

Accordingly, the ratio of the binding capacities for

imprinted polymer P2C and the control polymer P4C

reached a high value of 9.7. This result might be explained

by the fact that, for P4C, the carboxylic acid functions are

theoretically randomly distributed at the surface of the

nanospheres and offer only low affinity and multiple non-

cooperative one-point binding site interactions, compared

to the well-designed binding cavity of P2C bearing four

carboxyl functions. To further demonstrate the efficiencyof

thesurface imprintingofP2C, thecarboxylicacids functions

of the ‘‘glucose binding site’’ were esterified as methyl

esters by treatmentwith excess diazomethane inDCM. The

resulting polymer, reengaged in the batch rebinding

conditions with analyte 1, completely lost the original

binding proprieties of P2C, suggesting that the post-

esterification protocol precludes effective H-bonding and

therefore inhibit the uptake of 1.

The binding specificity of the imprinted polymer P2C for

1 was then studied by comparing the binding properties

DOI: 10.1002/mabi.200900056


Figure 5. Structure of pyranosides 1–4 used in the binding exper-iments.

with other pyranosides analogues. Nanoparticles of

imprinted polymer P2C were incubated at a given

concentration (6.67� 10�3M) with carbohydrates 2–4

(Figure 5), with the same procedure used for the binding

of 1.

The uptake of the a-anomer 2 turned out to be

12.8� 2.2mmol � g�1 of imprinted polymer P2C, meaning

that theabsolute stereochemistryat theanomeric carbonof

the sugar does not seem to be a strong discriminating

element, although 3.9mmol � g�1 of difference between

bothanomers1and2wasobserved (Figure6). Similarly, the

maximumbinding of carbohydrates 3 and 4 wasmeasured

and found to be 3.0� 1.7 and 2.1� 0.8mmol � g�1 of

imprinted polymer P2C, respectively. Accordingly, the

selectivity factors for 1 relative to 3 and 4 were determined,

in anhydrous acetonitrile, and found to be a1,3¼ 6.5 and

a1,4¼ 9.6, respectively.[27]

Figure 6. Binding capabilities of polymers P2C, P3 and P4C withcarbohydrates 1–4.



To date, considerable attempts have been devoted to the

preparation of highly selective MIPs for monosaccharides.

These MIPs have been prepared exclusively by bulk

polymerization, and have used the covalent approach, by

forming reversible boronate esters between 1,2-cis

diol,[6,7,21] the metal coordination approach[27,28] and the

non-covalent strategy based on weaker reversible hydro-

gen bonds.[22–26] However, the selectivities reported so far

for MIPs prepared by bulk polymerisation techniques still

need to be improved, especially for the discrimination

between glucose and galactose. We were gratified to

observe that the glucose/galactose selectivity factor

(a1,3¼ 6.5) calculated for P2C is the highest selectivity

factor reported, compared to the data available for a non-

covalent Zn2þ porphyrin-based imprinted polymer.[22] The

fact thatP2Cexhibited ahigh selectivity for1 compared to3

and 4, strongly suggest that the dominating factor for the

epimeric recognition depends mainly on the orientation of

the carboxylic groups inside thebinding siteof thepolymer,

in sharp contrast to MIPs prepared by bulk polymerisation,

for which the stereochemistry of the anomeric position of

the monosaccharide was a crucial parameter for a good

selectivity.[23,24] Our result accounts, probably, for a better

sensitivity and stereochemical ‘‘fit’’ of the surface imprint-

ing technique, associated with the well-defined four-point

interacting binding sites.

Having examined the imprinting effect and the selectiv-

ities for P2C, to further characterize the surface glucose

imprinted polymers,we decided to determine andquantify

the binding properties of this new synthetic receptor for

carbohydrate 1. We performed a study of the binding

kinetics by plotting the concentration of 1 bound to P2C

versus the time of incubation for a constant mass of

polymer (10mg), at the concentration of 6.67� 10�3M of 1

in CH3CN (Figure 7).

Interestingly, the equilibrium of the complexation was

reached after 5 h at 22 8C, accounting for quite effective

diffusion of the templatemolecule 1 to the specific binding

cavity of P2C exposed at the surface of the nanoparticles.

Figure 7. Binding kinetic for P2C with analyte 1.



Figure 9. Bi-Langmuir model applied for the binding of P2C withanalyte 1.

Figure 8. Binding isotherm for the uptake of 1 by glucoseimprinted polymer P2C obtained.

602

Compared to literature reports on the binding equilibration

experiments involving MIP prepared by bulk polymeriza-

tion, this time of equilibration can be considered as quite

fast, since 24 to 48h are usually required.[7,27] Under these

conditions, themaximumuptake of 1 reached amaximum

value of 16.0mmol � g�1 for P2C, in accordance with the

results obtained after 24h of incubation. With these data,

the binding isotherm was reconstructed by plotting the

concentration of bound ligand 1 versus the initial

concentration for a constant mass of polymer (10mg)

and for 5 h of incubation. The resulting curve fitting shown

in Figure 8, using Prism Software, gave a dissociation

constant (KD) of 950� 10�6M associated with a site

population of 21mmol � g�1 of polymer. The best fit is

Figure 10. a) Freundlich binding isotherm for P2C with carbohydrate 1. The correspond-ing affinity distribution: b) semi-log plot (N vs. logK); and, c) log plot (logN vs. logK).

obtained for a one-binding-site model

suggesting, eventually, a homogeneous

binding constant and the presence of

homogeneous binding sites. However,

data fitting with Prism software is

usually applicable for the study of

structurally homogeneous biological sys-

tems, such as enzymes or antibodies. The

application of more specific models and

assumptions about the distribution of

binding sites of polymer P2C was then of

crucial importance. Accordingly, as

reported in the literature,[15,32] we used

Langmuir models that allow more accu-

rate determination of whether an MIP

system exhibit homogeneous or hetero-

geneous binding sites. Non-linear regres-

sion of the Scatchard plot, constructed

from the replotting of the corresponding

binding isotherm, revealed, as shown in

Figure 9, that the best accommodated fit

is the heterogeneous binding model,

known as the Bi-Langmuir model. The

two straight lines drawn through the

curve in Figure 9 yielded two sets of



binding parameters corresponding to the low

(K1¼ 5 780� 10�6M) and high-affinity (K2¼ 439� 10�6

M)

binding sites of P2C. Furthermore, analysis of the data by

fitting to the Freundlich equations, after having verified the

appropriatenessof themodel for our rangeof concentration

for 1 (Figure 10a),[15,32] provided the affinity distribution

(Figure10bandc), andtheconcentrationof thebindingsites

for P2C: 14.5mmol � g�1 of polymer, in accordance with our

experimental data determined for 5h of incubation

(16.6mmol � g�1).

Therefore, altogether, we can assume that the imprinted

polymerP2Cexhibits twoormoredifferent typesofbinding

sites, eachwithabinding constant (Ki), as observed formost

of theMIPs. Unlike enzymes ormonoclonal anti-bodies, the

DOI: 10.1002/mabi.200900056


binding sites of imprinted polymers vary widely in size,

shape and rigidity. This tendency is also observed for P2C

prepared by the semi-covalent approach using mini-

emulsion techniques.

To thebest of our knowledge, fewMIPbinding properties

for sugars have been characterized using such mathema-

ticalmodels.[27]AlthoughKDmeasurementsare sensitive to

themethodandconditionsused, thevaluesobtained forour

polymers P2C are very useful for comparative purposes,

especially for artificial receptors involving exclusively

multiple hydrogen bonds as recognition elements.[10]

Conclusion

A novel, simple and straightforward semi-covalent surface

imprintingmethodwas developed and used in a one-stage

mini-emulsion polymerization. This technique, which

avoids the grinding and the sieving of the polymers, was

applied successfully for the preparation of artificial affinity

receptors for glucopyranosides, with interesting binding

properties and good selectivities, as compared to MIPs

prepared by bulk polymerization that use the covalent or

the non-covalent approaches. Due to the surface imprint-

ing, quite fast kinetics of binding and equilibration with

template 1 were observed. The selectivity factor of P2C

found between glucopyranoside and galactopyranoside is

the highest reported so far. The fitting of the binding

isotherm data with the Bi-Langmuir model, which better

characterizes the binding properties of our polymers

compared to Prism software, revealed the presence of

two populations of binding sites (low- and high-affinity

sites), affording a range of affinity constant involving

exclusively H-bonding interaction, for template 1 with

imprinted polymer P2C. Moreover, almost negligible non-

specific adsorption of the template molecule was observed

for such MIP in comparison to MIPs obtained by the non-

covalent approach. It is anticipated that this one-stage

mini-emulsion surface imprinting technique using a

polymerizable surfactant template could be applied to a

wide range of small molecules and biomolecules.

Acknowledgements: This research has been supported by theEuropean Commission and a Marie Curie Action RTN IntegratedBio-mimetic Approach to Asymmetric Catalysis: IBAAC (contractnumber MCRTN-CT-2003-505020), grant to P. C.. The authors aregrateful to Pascal Marie and Joseph Selb from the Institut CharlesSadron (ICS, Strasbourg), and Celine Cakir-Kiefer for interestingscientific discussions. Patrick Schultz and Christine Ruhlmann(IGBMC, Strasbourg) are also acknowledged for fruitful discussionsand TEM pictures.

Received: February 6, 2009; Accepted: March 19, 2009; DOI:10.1002/mabi.200900056



Keywords: miniemulsion polymerization; molecular imprinting;nanoparticules; semi-covalent; surfactants

[1] [1a] G. Wulff, Angew. Chem., Int. Ed. 1995, 34, 1812; [1b] B.Sellergren, Angew. Chem., Int. Ed. 2000, 39, 1031; [1c] A. G.Mayes, K. Mosbach, Trends Anal. Chem. 1997, 16, 321;[1d]‘‘Molecular and Ionic Recognition with Imprinted Poly-mers’’, Vol. 178, R. A. Bartsch, M. Maeda, Eds., AmericanChemical Society, Washington, DC 1997, 2799.

[2] For the use MIP in solid phase extraction (SPE) see: [2a] F.Lanza, B. Sellergren, Adv. Chromatogr. 2001, 41, 137; [2b] L. I.Andersson, L. Schweitz, Handb. Anal. Sep. 2003, 4, 45.

[3] For the use MIP in synthesis see: [3a] H. Zhang, T. Piacham,M. Drew, M. Patek, K. Mosbach, L. Ye, J. Am. Chem. Soc. 2006,128, 4178; [3b] K. Mosbach, Y. Yu, L. Andersch, J. Am. Chem.Soc. 2001, 123, 12420.

[4] For the use of MIP as catalyst, artificial enzyme and antibodysee: [4a] C. Alexander, L. Davidson, W. Hayes, Tetrahedron2003, 59, 2025; [4b] M. Emgenbroich, G. Wulff, Chem. Eur. J.2003, 9, 4106; [4c] K. Haupt, K.Mosbach, Chem. Rev. 2000, 100,2495; [4d] S. Al-Kindy, R. Badia, J. L. Suarez-Rodriguez, M. E.Diaz-Garcia, Crit. Rev. Anal. Chem. 2000, 30, 291.

[5] For the use of MIP in drug screening and delivery see: [5a] L.Ye, Y. Yu, K. Mosbach, Analyst 2001, 126, 760; [5b] P. T.Vallano, V. T. Remcho, J. Chromatogr. 2000, 888, 23; [5c] H.Hiratani, A. Fujiwara, Y. Tamiya, Y. Mizutani, C. Alvarez-Lorenzo, Biomaterials 2005, 26, 1293.

[6] G. Wulff,, R. Grobe-Einsler, A. Sarhan,Makromol. Chem. 1997,178, 2799.

[7] G. Wulff, S. Schauhoff, J. Org. Chem. 1991, 56, 395.[8] R. Arshady, K. Mosbach, Macromol. Chem. Phys. 1981, 182,

687.[9] O. Brueggemann, K. Haupt, L. Ye, E. Yilmaz, K. Mosbach,

J. Chromatogr. A 2000, 15, 889.[10] G. Wulff, Chem. Rev. 2002, 102, 1.[11] C. Alexander, H. S. Andersson, L. I. Andersson, R. J. Ansell,

N. Kirsch, I. A. Nicholls, J. O’Mahony, M. J. Whitcombe, J. Mol.Recognit. 2006, 19, 106.

[12] M. J. Whitcombe, E. N. Vulfson, Adv. Mat. 2001, 13, 467.[13] A. Katz, M. E. Davis, Macromolecules 1999, 32, 4113.[14] [14a] B. Sellergren, K. Shea, J. Chromatogr. A 1995, 690, 29;

[14b] P. Sanjonz, M. Kele, G. Zong, B. Sellergren, G. J. Guiochon,J. Chromatogr. 1998, 810, 1; [14c] B. Sellergren, Macromol-ecules 2006, 39, 6306; [14d] B. Sellergren, ‘‘MolecularImprinted Polymers: Man Made Mimics of Antibodies andtheir Applications in Analytical Chemistry’’, Vol. 23, ElsevierScience B.V., Amsterdam 2001, pp. 113–184.

[15] [15a] R. J. Umpleby, II, M. Bode, K. D. Shimizu, The Analyst2000, 125, 1261; [15b] R. J. Umpleby, II, S. C. Baxter, Y. Chen,R. N. Shah, K. D. Shimizu, Anal. Chem. 2001, 73, 4584.

[16] [16a] L. Ye, P. A. G. Cormack, K.Mosbach,Anal. Commun. 1999,36, 35; [16b] L. Ye, R.Weiss, K.Mosbach,Macromolecules 2000,33, 8239; [16c] T. de Boer, R. A. de Zeeuw, G. J. de Jong, D. C.Sherrington, P. A. G. Cormack, K. Ensing, Electrophoresis 2002,23, 1296.

[17] D. Vaihinger, K. Landfester, I. Krauter, H. Brunner, E. M. Tovar,Macrom. Chem. Phys. 2002, 203, 1965.

[18] [18a] M. J. Whitcombe, M. E. Rodrigez, P. Villar, E. N. Vulfson,J. Am. Chem. Soc. 1995, 117, 7105; [18b] N. Perez, M. J.Whitcombe, E. N. Vulfson, Macromolecules 2001, 34, 830.

[19] C. J. Tan, S.Wangrangsimakul, R. Bai, Y.W. Tong, Chem.Mater.2008, 20, 118.



604

[20] [20a] B. Sellergren, L. Andersson, J. Org. Chem. 1990, 55, 3381;[20b] N. Kirsch, M. J. Whitcombe, ‘‘The Semi-CovalentApproach’’, in: Molecularly Imprinted Materials, M. Yan, O.Ramstrom, Eds., Marcel Dekker, New York 2005, p. 93 [20c]S. C. Zimmerman, M. S. Wendland, N. A. Rakow, I. Zharov, K. S.Suslick, Nature 2002, 418, 399.

[21] [21a] W. Wang,, S. Gao, B. Wang, Org. Lett. 1999, 8, 1209; [21b]R. Rajkumar, A. Warsinke, H. Mohwald, F. W. Scheller, M.Katterle, Biosens. Bioelectron. 2007, 22, 3318.

[22] J. D. Lee, N. T. Greene, G. T. Rushton, K. D. Shimizu, J. I. Hong,Org. Lett. 2005, 7, 963.

[23] A. G. Mayes, L. I. Anderson, K. Mosbach, Anal. Biochem. 1994,222, 483.

[24] K. G. I. Nilsson, K. Sakaguchi, P. Gemierner, K. Mosbach,J. Chromatogr. A 1995, 707, 199.

[25] [25a] D. Maury, F. Couderc, J.-C. Garrigues, V. Poinsot, Talanta2007, 73, 340; [25b] F. Sineriz, Y. Ikeda, E. Petit, L. Bultel, K.Haupt, J. Kovensky, D. Papy-Garcia, Tetrahedron 2007, 63,1857; [25c] Y. Ikeda, F. Sineriz, L. Bultel, E. Grand, J. Kovensky,D. Papy-Garcia, Carbohydr. Res. 2008, 343, 587.



[26] C. Malitesta, I. Losito, P. G. Zambonin, Anal. Chem. 1999, 71,1366.

[27] [27a] S. Striegler, Tetrahedron 2001, 57, 2349; [27b] S. Strieg-ler, Bioseparation 2002, 10, 307; [27c] S. Striegler, M. Dittel,Anal. Chim. Acta 2003, 53; [27d] S. Striegler, Macromol 2003,36, 1310.

[28] G. Chen, Z. Guan, C.-T. Chen, L. Fu, V. Sundaresan, F. H. Arnold,Nat. Biotech. 1997, 15, 354.

[29] [29a] W. J. Wizeman, P. Kofinas, Biomaterials 2001, 22, 1485;[29b] P. Parmpi, P. Kofinas, Biomaterials 2004, 25, 1969.

[30] S. E. Bystrom, A. Borje, B. Akermark, J. Am. Chem. Soc. 1993,115, 2081.

[31] A. G. Mayes, M. J. Whitcombe, Adv. Drug Delivery Rev. 2005,222, 1742.

[32] [32a] M. Rampey, R. J. Umpleby, G. T. Rushton, J. C. Iseman,R. N. Shah, K. D. Shimizu, Anal. Chem. 2004, 76, 1123; [32b]R. J. Umbleby, II, S. C. Baxter, M. Bode, J. K. Berch, R. N. Shah,K. D. Shimizu, Anal. Chim. Acta 2001, 435, 35; [32c] A. K.Thakur, P. J. Hunston, D. Rodbard, Anal. Biochem. 1980, 103,240.

DOI: 10.1002/mabi.200900056

semi-covalent surface molecular imprinting of polymers by one-stage mini-emulsion polymerization:...

Documents