Investigation of polyacrylamide hydrogel‐based molecularly imprinted polymers using protein gel electrophoresis

In conjunction with polyacrylamide gel electrophoresis (PAGE), molecular imprinting methods have been applied to produce a multilayer mini‐slab in order to evaluate how selectively and specifically a hydrogel‐based molecularly imprinted polymer (MIP) binds bovine haemoglobin (BHb, ~64.5 kDa). A three‐layer mini‐slab comprising an upper and lower layer and a MIP, or a non‐imprinted control polymer dispersion middle layer has been investigated. The discriminating MIP layer, also based on polyacrylamide, was able to specifically bind BHb molecules in preference to a protein similar in molecular weight such as bovine serum albumin (BSA, ~66 kDa). Protein staining allowed us to visualise the protein retention strength of the MIP layer under the influence of an electric field. This method could be applied to other proteins with implications in effective protein capture, disease diagnostics, and protein analysis.


| INTRODUCTION
Molecular imprinting is a means of introducing sites of specific molecular arrangement into an otherwise uniform polymeric matrix. [1][2][3] These techniques have found applications in biomedical engineering, as chiral stationary phase in high-performance liquid chromatography, as chiral selector in capillary electrophoresis methods and in the design of new drug delivery systems. [4][5][6] Compared to traditional molecularly imprinted polymers (MIPs) made in organic solvents, aqueous media synthesis of chemically and mechanically stable MIPs has received much traction over the past 20 years and is an interesting challenge in chemistry. 7 This is due to their capability of recognising higher molecular weight molecules despite the significant reduction in integral binding strength of non-covalent template/monomer interactions. More recently, the molecular imprinting of large biomolecules, such as nucleic acids, 8,9 viruses, 10,11 and proteins, [12][13][14] has become increasingly topical, especially with the aim of developing MIP-based sensors for the detection of disease markers. MIP-based biosensors have been reported for the determination of a number of protein biomarkers including bovine (and human) serum albumin, [15][16][17] haemoglobin, 18 myoglobin (Mb), 19 and prostate-specific antigen. 20,21 The approach with biomolecular imprinting has been to use an aqueous solvent system, 22 which allows the biomolecular template to remain structurally stable during and after the imprinting process. The use of water-soluble monomers and cross-linkers in the synthesis of MIPs for biomacromolecular targets is now common place. The resulting hydrogel materials are hydrophilic and highly crosslinked.
Due to their high-water compatibility, hydrogel-based MIPs have been shown to retain protein stability and provide a robust means for recognition of target analytes over long periods. 7,23,24 For hydrogel protein imprinting, water-soluble monomers such as acrylamide and functionalised acrylamides have been used alongside the cross-linker N,N 0 -methylenebisacrylamide (MBAm) to produce polyacrylamidebased MIPs. 4,12,13 Polyacrylamide (PAM) is biocompatible and inert to almost any nonspecific interactions with proteins. The PAM gel monolith is processed through a 100-mesh net to produce micron-sized particles.
The constraining factor of this imprinting technology is the difficulty of the template removal. Despite that, Hawkins et al in 2005 demonstrated the efficiency of the cooperation of a strong anionic surfactant like sodium dodecyl sulphate (SDS) and acetic acid in the template removal strategy. 12 Whereas this widely adopted method removes the surface exposed protein to leave protein-selective binding sites, the method is limited to exposed surfaces; any protein retained within the bulk of the microparticles remains entrapped, even after such stringent washing.
PAM materials also form the backbone of gel electrophoresis.
Researchers routinely use electrophoresis to study the properties of proteins. Separation relies on charged biomolecules having different electrophoretic mobility through the PAM gel matrix because of the application of an electric field. Under constant electric field, the difference in mobility through a matrix depends on the charge and molecular weights of the molecules. Generally, the sample is run in a support matrix such as agarose or polyacrylamide gel. Agarose is mainly used to separate larger macromolecules such as nucleic acids, whereas polyacrylamide gel is widely employed to separate proteins. Slab gels, 0.5 to 1.5 mm thick, have replaced cylindrical rod gels in glass tubes because it allows direct comparison of the band pattern of different samples under identical conditions in the same matrix gel. In gel electrophoretic methods, among several other detection methods (organic dyes, fluorescent staining, and negative staining), silver staining is considered the most sensitive at low protein concentrations. All silver methods, that is, diamine or ammoniacal stains, non-diamine silver nitrate stains, silver stains based on photo-development, depend on the reduction of the ionic silver to the metallic form.
Several chromatographic applications of MIPs have been developed and applied in proteins and nucleic acids separations. Ogiso et al 25,26 applied MIP technology to gel electrophoresis to develop a simple and inexpensive DNA detection method. However, the MIP was prepared in situ in a glass tube using a specific double-standard DNA (dsDNA) target sequence as a printer molecule. What we present is the first report of hydrogel-based MIPs applied to slab electrophoresis, involving a multilayer resolving system. Illustrated is the application of hydrogel-based MIPs in mini-slab gel electrophoresis whereby a MIP dispersion is layered between two control layers. Using the threelayered mini-slab, a non-imprinted polymer (NIP) dispersion instead of MIP as a control system was also investigated. In this article, we explore how, during electrophoresis, a protein imprinted polymer can retain its template protein vs a protein, which is analogous in molecular weight.  (Guildford, UK).  were not observed to contain any protein. Therefore, we are confident that any remaining template protein within the MIPs did not continue to leach out during subsequent studies.

| MIP characterisation and selectivity
The subsequent rebinding effect and selectivity of the conditioned and equilibrated BHb-MIPs and NIPs were characterised using spec-  As soon as the solution was poured, the gel was layered with a few drops of 2-propanol to flatten the layer. After 10 minutes, the 2-propanol drops were drained away and the second layer was applied. This comprised 0.1 g of the preconditioned and equilibrated MIP (or NIP for the control system) dispersed in 1.5 mL of an identical solution to that of the first (bottom) layer. Thus, 1.2 mL of that dispersion was poured as a second layer into the space in between the two plates, and the same procedure as the first layer was followed. The third and final layer was filled using an identical solution as the first layer. The three-layered gel polymerising was then left for at least 4 to 5 hours before injecting the samples.

| Native polyacrylamide gel electrophoresis
BHb or BSA protein samples (500-4000 ng) were prepared in sample buffer solution to give a final volume of 50 μL, and then, the solutions were loaded directly onto the sample wells. After loading the samples, the gel was run at 150 V at different run times. The gels were then carefully removed using a blade and the European Molecular Biology Laboratory (EMBL) silver-staining protocol was followed in staining the gels. 27 3 | RESULTS AND DISCUSSION Table 1 illustrates the molecular imprinting effects of a BHb-MIP in recognising its original template BHb and non-cognate BSA in relation to a NIP control. These have been characterised by calculating the rebinding capacity (Q, mg/g) of proteins to the gel polymer using:

| MIP characterisation
where C i and C r are the initial protein and the recovered protein concentrations (mg/mL), respectively (which specifies the specific protein bound within the gel), V is the volume of the initial solution (mL), and g is the mass of the gel polymers ( g).The imprinting factor (IF) servers as a standard and is expressed by comparing the latter calculated binding capacities (Q) for MIP and NIP control (Equation 2): The selectivity of the BHb-imprinted MIPs for cognate proteins was quantified using relative imprinting factors (k; Equation 3): where IF template is the imprinting factor for the original template, and IF analogue is the imprinting factor of the analogue proteins. For the template BHb, k = 1, and for the non-cognate proteins that are less specific for the BHb-MIP, k < 1. It is evident from our data that the BHb-imprinted MIP has a higher binding capacity for its original BHb template in comparison to BSA (Q = 4.79 and 3.95 mg/g, respectively). BSA has also been expressed as having a 0.52 k to a BHb-MIP, meaning that more BHb is specifically bound by our MIP. Thus, our BHb-MIP has more recognition for BHb, in terms of selectivity and affinity, than non-cognate BSA when analysed spectrophotometrically for bulk gel imprinting prior to PAGE application.

| Native protein polyacrylamide gel electrophoresis
In order to study how selectively and specifically a MIP binds a certain molecule during electrophoretic procedures, we set the experiments using a multilayer system in mini-slab gel electrophoresis where a dispersed MIP layer with binding sites available was in between other two "non-imprinted" gel layers. We used an imprinted polymer, based on polyacrylamide hydrogels for the selective imprinting of bovine haemoglobin (BHb, MW 65 kDa) as a discriminating layer able to specifically bind BHb molecules instead of other proteins similar in molecular weight, namely, albumin from bovine serum (BSA, MW 66 kDa). A three-layer mini-slab with a NIP dispersion instead of a MIP in the middle layer was considered as a control system. Resulting stains are shown in Figures 1 to 3. In the first set of experiments, we explored the critical amount of haemoglobin that was detectable in the multilayer system configuration. Figure 1 illustrates that the layer containing the MIP exhibits significant staining throughout the gel.
This is due to the fact that the MIP particles used have residual haemoglobin locked within the gel particles, which has therefore also been stained. Tracks a-c in the electrophoresis experiment in The three-layer system for MIP-and NIP-loaded middle layers was investigated further at 4000 ng BHb injection, and a comparison was made between electrophoresing at 120 and at 150 minutes (see  All experiments in our study were conducted at 150 V, which is typically the upper voltage limit used for conventional protein gel electrophoresis (typically 100-150 V). It is likely that higher applied

| CONCLUSIONS
The binding of a BHb-imprinted MIP to its native template and non-cognate BSA protein has been assessed using gel electrophoresis. Both spectrophotometry and PAGE methods demonstrate imprinting and selectivity towards template BHb over cognate BSA.
The results show how an imprinted polymer retains its specific pro-