Binding assays with molecularly imprinted polymers—why do they work?

doi:10.1016/j.jchromb.2003.12.007

Journal of Chromatography B

Volume 804, Issue 1, 5 May 2004, Pages 167-172

https://doi.org/10.1016/j.jchromb.2003.12.007 Get rights and content

Abstract

The design of homologous displacement ligand binding assays based on molecularly imprinted polymers (MIP) is discussed in terms of the MIP adsorption isotherm. It is shown that only MIPs having a binding isotherm with varying slope are suitable for the assay, but there is no need to interpret the isotherm in terms of site affinity and population. One can calculate the calibration plot of the binding assay from the isotherm and vice versa.

Introduction

Molecularly imprinted polymers (MIP) are a novel class of selective sorbents [1]. They are usually custom made for every particular substrate that should be selectively bound. Their unique selectivity is due to the manufacturing procedure. A rigid, usually heavily crosslinked polymer is produced in the presence of a relatively large quantity of the substrate. When the process has been finished the substrate is removed by thorough washing of the polymer with a suitable strong solvent. When exposed next to a solution of the substrate in a weaker solvent, the polymer will adsorb (“rebind”) the substrate with notable selectivity over other, even closely related, substances. This remarkable memory effect has been attributed to interactions between the substrate and functional groups of the polymer during production. These interactions appear to arrange the polymer into a structure, which is a chemical and steric imprint of the substrate. As the substrate had been used in the process as a molecular template, it is alternatively called the template. There are two main types of imprinting: covalent [2], [3] and noncovalent. In this paper we consider only noncovalent imprints.

In an ideal case the MIP binding sites should all be chemically equal, i.e. they should bind the substrate in exactly the same way. If this were so one might expect that the adsorption equilibrium with solutions of the substrate in a particular solvent can be described with a Langmuir isotherm characterized by a single equilibrium constant. With some covalently imprinted polymers this is approximately true but with most noncovalent imprints the situation is more complex. Many investigators have established the binding isotherms of their novel MIPs. These binding isotherms are generally quite featureless, monotonously increasing curves with gradually decreasing slope. Based on the visual appearance of the respective Scatchard plots, several investigators fitted a bi-Langmuir isotherm [4], [5], [6], [7], [8] to the curves with reasonable success. This is equivalent to the assumption that there are two distinct types of binding sites on the polymer.

In recent years, several investigators expressed doubts concerning this two-site model. Umpleby et al. [9] assumed a continuous distribution of binding site strength (equilibrium constant) and obtained good fit to measured isotherms. Guiochon and coworkers [10] concluded from a large series of accurate isotherm measurements that the Freundlich isotherm and the bi-Langmuir isotherm gave equally good fits, while the simple Langmuir isotherm did not fit well. Umpleby et al. also noticed that the Freundlich isotherm gave good fits to their data as well. What these studies show from the point of view of the present discussion is, that binding isotherms do not reliably support any specific idea about the site distribution of MIPs.

These newer results raise some doubts about the concepts underlying a particularly fascinating application of MIPs, i.e. their use as artificial antibodies. The selective and strong binding of various substrates to their MIPs resembles the similar features of natural antibodies. Indeed, Haupt and Mosbach [12], and Andersson [11], [13] have shown in many papers that MIPs can be used in essentially the same assay formats as antibodies, with strikingly similar results. In the course of years better and better MIPs (“plastic antibodies”, “artificial antibodies”) have been developed for such assays (molecularly imprinted sorbent assay, MIA).

MIP binding assays have been generally based on homologous competition, i.e. competition between the substrate and its radiolabeled version, for the limited number of binding sites. This is the point where questions are raised by the difficulties of isotherm interpretation. If we cannot be sure that there is only one type of site present, which is limited in its quantity, why should there be any competition then? One answer lies immediately at hand and has been used by researchers quite often. In classical immunoassays polyclonal antibodies were used. These are mixtures of antibodies of varying binding strength (equilibrium constant). This is obvious from the calibration curves of the respective immunoassays as the curve extends over a considerably larger concentration range than with a monoclonal antibody population. Similar extended calibration curves have been observed in MIP binding assays. This has led to the conclusion that site heterogeneity is not a problem in understanding MIP binding assays.

There are, however, some arguments against this opinion. Biochemists can—at least in principle—separate mixtures of antibodies and determine the quantities and binding strengths of individual protein fractions. This is not possible at the moment with MIPs. Also, biochemists have developed refined immunization strategies that lead to practically useful antibody mixtures. Similarly, MIPs have been optimized for better performance in MIP based binding assays. These results have solved the practical problem, but the question still remains: Are all site distributions suitable for competitive binding assays? Or, to ask a more practical question: can we judge the usefulness of a MIP for competitive assays by measuring its adsorption isotherm but without attempting to interpret it in terms of site distribution? The goal of this paper is to answer these questions. We shall show that knowledge of the isotherm is sufficient to determine the MIP binding assay calibration curve and thus also the expectable detection limit and useful concentration range of the measurement. If the isotherm is linear, the assay is not possible. Our results will hopefully also contribute to improved MIP designs.

Section snippets

Assay format

Immunoassays exist in many different formats. Some of these are making use of antibody properties, which are unlikely to be imitated by MIPs. In a sandwich immunoassay, e.g. the solubility of the antibody in the assay solvent is important. MIPs are by their nature insoluble. For such reasons the practically demonstrated MIP binding assays have been limited to competitive assays. Competitive binding assays exist in two main types: homologous and non-homologous. In both cases two species compete

Quantitative analysis

Let us consider a typical assay experiment. The total molar amount of radiolabeled analyte (n^∗) distributes into adsorbed (“bound”) amount ( $n_{b}^{∗}$ ) and dissolved (“free”) amount ( $n_{f}^{∗}$ ). The molar amount of unlabeled analyte in the whole system is n, which distributes as n_b and n_f between the two phases. The concentrations in the solution phase shall be denoted by c and expressed in M, whereas in the adsorbed phase by q and expressed in mol/kg. To avoid complications we shall use c and q,

The role of isotherm shape

The MIP isotherm has been sometimes found to start with a linear section at low concentrations [17], [18]or it was described by the Langmuir or bi-Langmuir equations [19], [20] which become also linear at low concentrations. Occasionally, an approximate linearity may extend to high substance concentrations. This is actually a goal in developing MIPs for HPLC stationary phases. We show now that in this linear range of the isotherm the displacement assay does not work.

The measured quantity in the

Conclusion

Among the analytical applications of MIPs chromatographic type applications (HPLC, SPE, CEC, etc.) and binding assay type applications are probably the most important. Researchers working in these two fields of applications have been using the traditional terminologies of chromatography and immunoanalysis, respectively. In particular, chromatographers look at MIPs as solid sorbents, which can be characterized by their adsorption isotherms, kinetics and morphology. In immunoanalysis the MIP is

Acknowledgements

This work was supported by 3/043/2001 National Research and Development Programme (NKFP), the Pro Progressio Foundation and the Varga József Foundation.

References (23)

M.A. Khasawneh et al.
J. Chromatogr. A
(2001)
J. Zhou et al.
Anal. Chim. Acta
(1999)
Q.-Z. Zhu et al.
Anal. Chim. Acta
(2002)
A. Bereczki et al.
J. Chromatogr. A
(2001)
P. Sajonz et al.
J. Chromatogr. A
(1998)
L.I. Andersson
J. Chromatogr. B
(2000)
K. Haupt et al.
TIBTECH
(1998)
J.G. Karlsson et al.
Anal. Chim. Acta
(2001)
V.P. Joshi et al.
Chem. Eng. Sci.
(2000)
H.S. Andersson et al.
J. Chromatogr. A
(1999)

C. Baggiani et al.

J. Chromatogr. A

(1997)

Cited by (24)

MIP-based immunoassays: A critical review
2023, Analytica Chimica Acta
Molecularly imprinted polymers, MIPs, are man-made receptors mimicking the thermodynamic and kinetic binding behaviour of natural antibodies. Therefore, it is not surprising that many researchers have thought about MIPs as artificial receptors in immunoassay-like analytical applications, where the general machinery of the assay is maintained, but the molecular recognition is no longer assured by an antibody but by an artificial receptor. However, the number of papers devoted explicitly to applications of MIPs in the immunoassay field is quite limited if compared to the huge number of papers covering the multifaceted molecular imprinting technology. For this reason, this critical review wants to give a general view of MIP-based immunoassays, trying to highlight the critical points that have so far prevented a wider application of molecular imprinting technology in the immunoassay field and, possibly, try to suggest strategies to overcome them.
Review on molecularly imprinted polymers with a focus on their application to the analysis of protein biomarkers
2021, TrAC - Trends in Analytical Chemistry
Citation Excerpt :
This can be a very slow process, since molecules need to cross the porous solid polymeric network to access the binding sites. In the case of surface imprinting, binding is usually faster, since binding sites are on the surface and readily accessible [135]. The best way to calculate binding kinetics is to carry out a batch rebinding assay for both MIPs and NIPs.
Molecularly imprinted polymers (MIPs) are a type of artificial polymer, which have complementary cavities that are designed to bind a specific target molecule with a high degree of selectivity. Due to their effectiveness and stability, MIPs have found their way into many applications in medicine, chemistry, analysis and sensing fields. One of the most important modern uses of MIPs is the recognition of biological molecules of medical significance, which are called “biomarkers”. The use of MIPs enables easy and rapid extraction and detection of these biomarkers from different biological matrices. There are multiple techniques that arose for synthesis of MIPs each with their own set of advantages and drawbacks. In this review, we discuss MIPs in detail including their different types, methods of synthesis, characterisation methods, common challenges, in addition to their applications in different fields with a focus on their use in the analysis of protein biomarkers.
Dynamic electromembrane extraction: Automated movement of donor and acceptor phases to improve extraction efficiency
2015, Journal of Chromatography A
Citation Excerpt :
In comparison with molecular imprinted polymers (MIPs), although MIP systems have theoretically promising properties for selective recognition of various organic and inorganic analytes, such as pharmaceuticals and others, but the practical realization of these systems is still a challenge. This fact can be attributed to the poor efficiency in binding kinetics of some compounds in the bulk of polymer networks [36], poor recognition ability in some cases [37], laborious synthesis procedure, relatively long extraction time and the memory effect resulted from incomplete removal of templates from the polymer bulk. Magnetic solid phase extraction methods (MSPE) have many advantages including ease of operation, rapidity and providing high preconcentration factors however, relatively low cleanup, need to sample treatment of complicated matrices before extraction procedure as well as a time consuming and rather costly synthetic procedure are the challenges can be considered for MSPE.
In the present research, dynamic electromembrane extraction (DEME) was introduced for the first time for extraction and determination of ionizable species from different biological matrices. The setup proposed for DEME provides an efficient, stable, and reproducible method to increase extraction efficiency. This setup consists of a piece of hollow fiber mounted inside a glass flow cell by means of two plastics connector tubes. In this dynamic system, an organic solvent is impregnated into the pores of hollow fiber as supported liquid membrane (SLM); an aqueous acceptor solution is repeatedly pumped into the lumen of hollow fiber by a syringe pump whereas a peristaltic pump is used to move sample solution around the mounted hollow fiber into the flow cell. Two platinum electrodes connected to a power supply are used during extractions which are located into the lumen of the hollow fiber and glass flow cell, respectively. The method was applied for extraction of amitriptyline (AMI) and nortriptyline (NOR) as model analytes from biological fluids. Effective parameters on DEME of the model analytes were investigated and optimized. Under optimized conditions, the calibration curves were linear in the range of 2.0–100 μg L⁻¹ with coefficient of determination (r²) more than 0.9902 for both of the analytes. The relative standard deviations (RSD %) were less than 8.4% based on four replicate measurements. LODs less than 1.0 μg L⁻¹ were obtained for both AMI and NOR. The preconcentration factors higher than 83-fold were obtained for the extraction of AMI and NOR in various biological samples.
Determination of L-phenylalanine on-line based on molecularly imprinted polymeric microspheres and flow injection chemiluminescence
2012, Spectrochimica Acta - Part A: Molecular and Biomolecular Spectroscopy
Citation Excerpt :
To produce molecularly imprinted polymers, two different approaches have been developed: the set-up of covalent interactions between template molecules and monomers [13] and the development of reversible non-covalent interactions (mainly hydrogen bonding) between them [14,15]. The molecularly imprinted materials have been applied into affinity separation [16], antibody binding [17], biomimetic chloroperoxidase [18] and biomimetic sensors [19], which were excellently examined in recent developments and perspective area evolutions such as chromatographic stationary phase [20], solid phase extraction matrices [21,22], artificial receptors [23], antibody mimics [24], enzyme mimics [25], catalytic application [26], recognition elements in sensors [27] and CL sensing systems [28]. In this paper, a novel MIP-CL sensor for Phe determination using MIP as recognition element is developed.
A novel molecular imprinting-chemiluminescence (MIP-CL) sensor for the determination of l-phenylalanine (Phe) using molecularly imprinted polymer (MIP) as recognition element is reported. The Phe-MIP was synthesized using acrylamide (AM) as functional monomer and ethylene glycol dimethacrylate (EGDMA) as cross-linker, 2,2-azobisisobutyronitrile (AIBN) as initiator and the polymers’ properties were characterized. Then the synthesized MIP was employed as recognition element by packing into flow cell to establish a novel flow injection CL sensor. The CL intensity responded linearly to the concentration of Phe in the range 1.3 × 10⁻⁶ to 5.44 × 10⁻⁴ mol/L with a detection limit of 6.23 × 10⁻⁷ mol/L (3σ), which is lower than that of conventional methods. The sensor is reusable and has a great improvement in sensitivity and selectivity for CL analysis. As a result, the new MIP-CL sensor had been successfully applied to the determination of Phe in samples.
Optimizing the formulation of a myoglobin molecularly imprinted thin-film polymer-formed using a micro-contact imprinting method
2007, Biosensors and Bioelectronics
Thin-film myoglobin molecularly imprinted polymers have been fabricated using a micro-contact approach. By initially selecting the cross-linker on the basis of it having a minimal recognition for the template and using this as a starting point for functional monomer selection, we have produced myoglobin imprinted polymers with exceptionally high selectivities.
The affinity of the polymers, for myoglobin, when prepared with a variety of different cross-linkers and no functional monomer was evaluated. Of these, tetraethylene glycol dimethacrylate (TEGDMA) exhibited the lowest affinity for the template species. Methyl methacrylate (MMA) was chosen as the functional monomer as when it was used in conjunction with TEGDMA, it exhibited maximum selectivity for the template compared to polymers made with other functional monomers.
With a MMA to TEGDMA ratio of 1 to 3, the myoglobin molecularly imprinted polymer adsorbed 15.03 ± 0.89 × 10⁻¹¹ mole/cm² of template from a 5.68 × 10⁻⁷ M myoglobin solution, compared to 2.58 ± 0.02 × 10⁻¹¹ mole/cm² for a polymer of similar composition, but formed in the absence of a template. Various washing conditions, using alkaline media to remove the template, were investigated. An extraction solvent comprising 2 wt.% SDS and 0.6 wt.% NaOH used at 80 °C for 30 min was shown to give the highest imprinting factor i.e. 5.83 with 72.82% myoglobin removal.
The saturation kinetics of template binding to the thin-film MIP were examined and found to display a simple two-phase profile typical of non-cooperative binding. A Scatchard binding plot showed the dissociation constant (K_d) for the specific binding phase to be 3.4 × 10⁻⁷ M and the binding site capacity to be 7.24 × 10⁻¹¹ mole/cm². For the non-specific binding phase, K_d was found to be 1.355 × 10⁻⁵ M and the binding site capacity was determined as 9.62 × 10⁻¹⁰ mole/cm².
Selectivity experiments were carried out in both single protein and binary protein systems all using a total protein concentration of 5.68 × 10⁻⁷ M. The molar ratio of adsorbed myoglobin to IgG, HSA and hemoglobin was found to 115.5, 230.9 and 2.5, respectively. While, in binary competition systems, myoglobin selectivity to IgG, HSA and hemoglobin was, respectively, 94.18, 98.21 and 61.09%. Rebinding in natural biological matrices, i.e. human serum or urine, showed the imprinted films to have significantly greater uptake than non-imprinted films. Re-binding in undiluted urine was found to be a facile process, with the imprinting factor, i.e. the ratio of MIP to NIP binding, being determined as 37.4.
Characterization of binding sites in molecularly imprinted polymers
2007, Sensors and Actuators, B: Chemical
In recent years interest has grown in a deeper understanding of the heterogeneity nature of binding sites in molecularly imprinted polymers, purposively designed artificial receptor materials. This article reviews the progress of the understanding of the molecular recognition binding sites beyond the usual description of adsorption isotherms. Analytical and numerical methods used for calculating the adsorption isotherms and the adsorption energy distribution, as a quantitative measure of the imprinted polymer heterogeneity, are included. Advantages and limitations of the different approaches are also discussed.

View all citing articles on Scopus

View full text

Binding assays with molecularly imprinted polymers—why do they work?

Abstract

Introduction

Section snippets

Assay format

Quantitative analysis

The role of isotherm shape

Conclusion

Acknowledgements

J. Chromatogr. A

Anal. Chim. Acta

Anal. Chim. Acta

J. Chromatogr. A

J. Chromatogr. A

J. Chromatogr. B

TIBTECH

Anal. Chim. Acta

Chem. Eng. Sci.

J. Chromatogr. A

J. Chromatogr. A