An effective way to imprint protein with the preservation of template structure by using a macromolecule as the functional monomer

Abstract

In this study, an assumption that a micromolecular monomer could easily permeate into the inside of a protein and alter its conformation, while an inflexible macromolecular monomer may interact with the surface of the protein and thus maintain the integrity of the template protein's structure was proposed for the first time and confirmed by using circular dichroism and synchronous fluorescence spectroscopy. The protein imprinted hydrogels composed of macromolecular monomers or their equivalent micromolecular monomers were characterized and carried out in the competitive adsorption and adsorption isotherm experiments. The adsorption isotherm behaviours described by the Langmuir model revealed that a higher binding affinity was observed between the template protein and imprinted hydrogels made by a macromolecular monomer. The competitive adsorption results also demonstrated the imprinted hydrogels prepared by the macromolecular monomer exhibited much better specific recognition ability to the template protein. Therefore, the strategy of using a macromolecule to imprint could effectively overcome the mutability of protein during the preparation of imprinted polymers, and consequently would promote the development of imprinting technology.

---

Introduction

Recently, molecular imprinting technology has drawn worldwide attention in sensing, catalysis, separation and drug delivery.^1–6 However, compared to the rapid development of micromolecule imprinting technology, the imprinting of macromolecules is much more difficult.^7–12 One of the most important challenges lies in the complex and flexible structures of these biomacromolecules,^13–16 especially for proteins, which are widely applied as templates in imprinting technology. In research reported by Kryscio,^17,18 the secondary structures of two common protein templates were significantly altered in the presence of micromolecular monomers or crosslinkers, whose concentrations were far below those normally used in the preparation of molecularly imprinted polymers (MIPs). When the modified protein structure was surrounded by the MIPs, the binding cavities formed were specific to this alternate conformation. As a consequence, a relatively low selectivity of protein MIPs can be expected when compared to small molecule MIPs. Therefore, the essential prerequisite of generating a protein imprint should be to ensure the native conformation of the template in the preparation of MIPs.

In a typical protein imprinting process, structural instability of the template molecule may be caused by relatively small functional monomers and crosslinkers, whose flexibility could easily allow permeation inside the protein and destroy the hydrogen bonds that maintain its native conformation, as shown in Scheme 1. Based on this assumption, we propose a versatile and novel strategy to obtain MIPs with greater template specificity by using an inflexible macromolecular monomer and crosslinker to imprint protein.

Scheme 1 The influence of macromolecular chain and micromolecular monomer on the conformation of template protein.

The imprinting of protein by utilizing macromolecular monomers or crosslinkers has in fact been previously reported. The natural polymers, like chitosan and alginates, which are water-soluble linear polysaccharides containing plenty of functional groups, are usually chosen as the matrix used to imprint the protein.^19–21 The obtained MIPs possessed excellent features including hydrophilicity, non-toxicity, biocompatibility and biodegradation. Recently, an artificial polymer chain was synthesized as an additional element of monomer to create effective recognition sites in protein MIPs.^22,23 The resulting MIPs exhibited good adsorption capacity of the template protein. Although excellent progress has been made with these researches, the use of macromolecular functional monomers focuses on increasing the biocompatibility of the matrix with the template protein or on having an increased number of binding cavities available.

In this work, we designed a linear macromolecular chain as both the functional monomer and the crosslinker in order to improve structural stability of the template protein during the preparation of MIPs. Circular dichroism (CD) and synchronous fluorescence spectroscopy was carried out to investigate the influence of macromolecular chain and its micromolecular monomeric components on the structural stability of ovalbumin and lysozyme. In order to study the effect of structural stability of template on the selectivity of MIPs, the imprinted hydrogels (MIH) made by macromolecular chain or its equivalent micromolecular monomers were prepared with the ovalbumin and lysozyme as the template respectively. The resultant MIHs were characterized and subjected to the competitive adsorption and isotherm experiments to investigate their selectivity and specific recognition ability.

Materials and methods

Materials

Hydroxyethyl acrylate (HEA), 1-vinylimidazole (VIM) and allyl chloroacetate (ACa) were purchased from Alfa Aesar. 2,2-Azobisisobutyronitrile (AIBN), ammonium persulfate (APS), N,N,N′,N′-tetramethylethylenediamine (TEMED) were obtained from Sigma. Lysozyme (Lyz, MW 14 kDa, pI 10.7), cytochrome C (Cyc, MW 12.4 kDa, pI 10.8), ovalbumin (OVA; MW 45 kDa, pI 4.7) and bovine serum albumin (BSA; MW 66.4 kDa, pI 4.8) were purchased from Sigma-Aldrich. Molecular low weight protein markers were obtained from Shanghai Jinsui Bio-Technology Co. Ltd. Deionized water was produced by a Millipore water system. All the chemicals used were at least of analytical grade.

Instrumentation

The absorption spectra were measured using a UV spectrophotometer (Varian, Cary-1E). The CD experiments were carried out using an Applied Photophysics Chirascan instrument. The ¹H nuclear magnetic resonance (¹H NMR) spectra were recorded using an Avance 300 MHz spectrometer. X-ray photoelectron spectroscopy (XPS) was carried out using Kratos Axis Ultra DLD to determine chemical composition. Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) was performed using a DYY-6C electrophoresis system (Beijing Liuyi Instrument Factory). Fluorescence measurements were performed on a spectrofluorimeter Model F-4500 (Hitachi, Japan). Scanning electron microscope (SEM) was employed using an Inca Oxford instrument to investigate the morphology of the hydrogels. The element analysis was carried out by using Elementar Vario EL (Germany) to determine chemical composition. Gel permeation chromatography (GPC) was carried out using Waters e2695 GPC system with Wat011525 water column.

Synthesis of the macromolecular chains

The macromolecular chains were prepared using a two-step polymerization method. The precursor for the macromolecular chains was synthesized by copolymerization of HEA and VIM. In a typical polymerization, a 25 mL flask was equipped with a magnetic stirrer. HEA (1.4 mL, 10.9 mmol) and VIM (0.6 mL, 6.1 mmol) were dissolved in 10 mL ethanol, and then AIBN (20 mg) was added to the solution. Subsequently, the mixed solution was deoxygenated by purging with nitrogen for 15 min. The copolymerization was carried out at 70 °C for 8 h. After the reaction, the resultant precursor of macromolecular chains was precipitated to a white viscous solid by using 20 mL diethyl ether. The precursor was subsequently dissolved in 8 mL DMF, followed by the slow addition of ACa (0.6 mL, 4.0 mmol) using constant pressure funnel at 50 °C. The reaction mixture was stirred for 18 h and the macromolecular chain was precipitated with diethyl ether to form a canary yellow viscous solid. After washing with 40 mL diethyl ether to remove the unreacted monomer, the macromolecular chain was then dissolved in 5 mL water and dialyzed by using dialysis tubing with a molecular cutoff of 20 kDa for 3 days. The aqueous solution containing the macromolecular chain was lyophilized and 1.72 g of the desired product was obtained.

One of micromolecular monomer, 1-(allyl acetate)-3-vinyl-imidazolium chloride, here we called [AVIM]Cl was also synthesized and details of its formation and characterization can be found in the ESI.†

Characterization of the influence of macromolecular chain or its equivalent micromolecular monomer on the structural stability of OVA and Lyz respectively

The CD experiments were carried out to estimate the effect of the macromolecular chains or its component micromolecular monomers on the secondary structure of OVA and Lyz respectively. The spectrophotometer was purged with sufficient nitrogen before starting the instrument. The CD calibration was performed using (1S)-(+)-10-camphorsulfonic acid (Aldrich), which exhibits a 34.5 M cm⁻¹ molar extinction coefficient at 285 nm, and 2.36 M cm⁻¹ molar ellipticity (θ) at 295 nm. The sample was pre-equilibrated at 25 °C for 20 min and scanned using 1 cm path length quartz cuvette cell in the range of 200–250 nm with 0.5 nm resolution. The concentration of protein was fixed at 1.1 × 10⁻⁶ M and the mass ratio of macromolecular chains or its component micromolecular monomers to the protein varied from 0 : 1 to 20 : 1. All the sample spectrums were obtained by subtracting appropriate blank media without protein from the experimental spectrum.

Synchronous fluorescence spectroscopy was carried out to investigate the effect of the macromolecular chains or its components micromolecular monomers on the internal microenvironment of OVA and Lyz respectively. Fluorescence measurements were performed using a 1 cm quartz cuvette at 25 °C. Synchronous fluorescence spectra were obtained by setting the excitation and emission wavelength interval (Δλ) at 15 and 60 nm. The concentration of protein was fixed at about 1.1 × 10⁻⁶ M and the mass ratios of macromolecular chains or micromolecular monomers to protein were varied as 0 : 1, 1 : 1, 2 : 1, 5 : 1, 10 : 1, 15 : 1 and 20 : 1.

The preparation of molecular imprinted hydrogel

A typical synthesis procedure of imprinted hydrogels by using macromolecular chain is as follows: the template protein (45 mg for OVA and 14 mg for Lyz respectively) was dissolved in 0.8 mL phosphate buffer (0.01 M, pH 7.0) in a 25 mm × 40 mm (diameter × height) weighing bottle. The macromolecular chains (132.8 mg) and APS (2 mg) were mixed into solution at 25 °C for interacting with template protein. After incubation of 20 min, the solution was deoxygenated by purging with nitrogen for 15 min. Before the weighing bottle was sealed, a volume of 2 μL TEMED was promptly added into the solution and the polymerization was carried out at 25 °C for 3 h. After the reaction, the resultant hydrogels were cut into 6 mm diameter disks with a thickness of 1.5 mm.

The washing procedure was accomplished by dipping the hydrogel disks in 100 mL of NaCl (0.5 M) solution at 25 °C. The complete removal of template protein from imprinted hydrogels was confirmed by using a UV spectrophotometer and XPS, which showed no characteristic peaks of template protein in its eluate and exhibited no elemental sulphur in the hydrogels respectively. After protein removal, 1 L of deionized water was used to remove any traces of NaCl. After washing, the disks were dried by lyophilization. The obtained molecular imprinted hydrogels were defined as MIH-C-OVA and MIH-C-Lyz, respectively. The corresponding non-imprinted hydrogel was prepared in the same way, but in the absence of the template protein and was defined as NIH-C.

The imprinted hydrogel made by micromolecular monomers was also prepared and the detailed methods were presented in ESI.† The resultant MIHs and NIHs were defined as MIH-M-OVA, MIH-M-Lyz and NIH-M respectively.

Characterization of molecular imprinted hydrogel

To assess the completion degree of gelation, gel fraction yield was calculated by dividing the final dry mass of hydrogel (m_dry) by the total mass of monomers used in the preparation of hydrogels as given in the following equation:²⁴

Gel fraction yield = m_dry/m_{total monomer} × 100

All the samples were tested in duplicate.

The gravimetric method was employed to study the hydrogel swelling ratio. After equilibrium at 25 °C, the hydrogels were removed from the phosphate buffer (0.01 M, pH 7.0) and blotted with filter paper for the removal of excess water on the surface. They were then weighed, and the swelling ratio was calculated from the following formula:²⁵

Swelling ratio = (W_s − W_d)/W_d × 100

where W_s and W_d are the weights of the swollen and dry hydrogels, respectively. Each sample was measured in duplicate.

The surface morphologies of hydrogels were investigated using SEM. The dried hydrogels were mounted on metal stubs and at a low vacuum degree (∼10⁻³ atm).

Protein adsorption experiments

The adsorption dynamic studies were conducted to investigate the adsorption equilibrium time of hydrogels. In this experiment, about 50 mg of dry hydrogels was first equilibrium in phosphate buffer (0.01 M, pH 7.0) at 25 °C. After swelling equilibrium, the solution was replaced by 10 mL protein solution with an initial concentration of 1.0 mg mL⁻¹. The concentration of residual protein solution adsorbed by hydrogel was measured by UV at different incubation times.

The adsorption isotherm was employed to investigate the bind affinity of MIH-C and MIH-M. In this experiment, about 50 mg of dry hydrogels was first equilibrium in phosphate buffer (0.01 M, pH 7.0) at 25 °C. After swelling equilibrium, the solution was replaced by 10 mL phosphate buffer (0.01 M, pH 7.0) containing various concentrations of protein (initial concentration range: 0–1.2 mg mL⁻¹). The solution was incubated at 25 °C for 3 h. After the incubation, the protein solution was collected, and the concentration was determined using a UV spectrophotometer. The amount of the protein adsorbed on hydrogel was calculated according to the following equation.

Q = (C₀ − C_e)V/m

where C₀ (mg mL⁻¹) and C_e (mg mL⁻¹) are the initial and final concentrations of the protein solution, respectively. V (mL) is the volume of the protein solution, whereas m (g) is the mass of the hydrogel.

The imprinting factor (IF) was used to evaluate the specificity of MIH toward the template or template analogues and was calculated from following equation:

IF = Q_MIH/Q_NIH

where Q_MIH and Q_NIH are the adsorption capacity of the template or template analogues by MIH and NIH respectively.

In the selective adsorption experiments, about 50 mg of dry hydrogels were first equilibrium in phosphate buffer (0.01 M, pH 7.0) at 25 °C. After swelling equilibrium, the solution was replaced by a 10 mL protein solution with an initial concentration of 1.0 mg mL⁻¹. The hydrogel was incubated in the protein solution at 25 °C for 3 h. After the incubation, the protein solution was collected, and the concentration was determined using a UV spectrophotometer. All the samples were tested in duplicate.

The selectivity factor (β) is defined as the following equation:

β = IF_temp/IF_ana

where IF_temp and IF_ana are the imprinting factors for the template molecule and for the analogues respectively.

The competitive adsorption was carried out to estimate the recognition ability of MIH-C and MIH-M. A mass of 50 mg dry hydrogel was used in this experiment. After the swelling equilibrium, the hydrogel was put into a 10 mL phosphate buffer (0.01 M, pH 7.0) containing the protein mixture of BSA, OVA and Lyz with each concentration of 1.0 mg mL⁻¹. The hydrogel was incubated in above solution at 25 °C for 3 h. After incubation, the hydrogel was taken out from the solution and treated with phosphate buffer (0.01 M, pH = 7.0) containing 5 mM NaCl to remove the weakly adsorbed proteins and then with 0.5 M NaCl solution to elute the strongly adsorbed proteins. The elution process was continued until no characteristic peaks of these proteins showed in the ultraviolet spectrum. The fractions of strongly adsorbed proteins were combined and desalted by using dialysis tubing with a molecular cutoff of 500 Da, and then freeze-dried using lyophilization. The obtained product dissolved in 3 mL phosphate buffer (0.01 M, pH = 7.0). 10 μL of product solution was measured by SDS-PAGE using a 12.5% polyacrylamide separating gel and 5% polyacrylamide stacking gel.

Results and discussion

Synthesis of the macromolecular chains

Macromolecular chain should be designed to adhere to two basic conditions. First, it must exhibit good solubility in water for interacting with the template protein. Secondly the resultant macromolecular chain should be able to be sequentially polymerized to form MIPs. Therefore, we copolymerized hydroxyethyl acrylate and N-vinyl imidazole to obtain precursor for the macromolecular chain, with the final product prepared by alkylation reaction with ACa. Thus the resultant macromolecular chain possessed hydrogen bond and electrostatic affinity to proteins, which is in favour of the imprinting and recognition process.

The ¹H NMR spectra and chemical structure of macromolecular chains and their precursors are shown in Fig. 1. The spectrum for the precursor in Fig. 1(a) has peaks at 7.6, 7.2 and 6.9 ppm, which are characteristic of the hydrogen atoms in imidazole ring. Additional peaks at 4.9, 4.0 and 3.5 ppm were attributed to the hydrogens in hydroxyethyl. The macromolecular chains were composed of three monomeric parts, HEA, VIM and [AVIM]Cl. Therefore, besides having the same peaks as in the precursor, the peaks at 9.5 and 7.9 ppm were attributed to the ionized imidazole ring, and the peaks at 6.1 and 5.4 ppm are characteristic of double bonds. Consequently, the spectrum in Fig. 1(b) illustrated the successful alkylation reaction between imidazole and Aca, with the resultant chain being our target product. The chemical structure of macromolecular chain and its precursor was also examined using FTIR, and the results were consistent with the findings in ¹H NMR spectra (the detailed information was presented in ESI and Fig. S2†).

Fig. 1 ¹H NMR spectra of the macromolecular chains (b) and their precursor (a).

Fig. S3† illustrates the XPS survey spectra of the macromolecular chain and its precursor. It shows that the precursor contains C, O, and N elements, whereas the macromolecular chain is composed of C, O, N, and Cl elements indicating the successful alkylation between the imidazole group in the precursor and ACa. The atomic composition of C, O, and N for the precursor is 65.32, 24.31 and 10.37% respectively, and the atomic composition of C, O, N and Cl for the macromolecular chain is 68.31, 23.71, 6.14 and 1.84%, respectively. Therefore, the mass ratio of HEA to VIM in the precursor was calculated as 65.2 : 34.8, and the mass ratio of HEA : VIM : [AVIM]Cl in the macromolecular chain was 65.4 : 7.5 : 27.1, respectively. Furthermore, the molecular weight of macromolecular chains and their precursors were analysed (Fig. S4†) and the results revealed their M_n values of 28 615 and 27 761 with the PDI value of 1.06 and 1.05 respectively.

The influence of macromolecular chain or its equivalent micro-molecular monomer on the structural stability of the protein

The far-UV CD spectrum was obtained in order to reveal any changes in proteins' secondary structure. In the far-UV region, the CD spectrum is an effective probe for the study of the protein secondary structure such as α-helix, β-sheet and β-turn. In the case of OVA and Lyz, the regions at 208 and 222 nm provide information on α-helix content. As shown in Fig. 2, when the equal mass of micromolecular monomers was mixed with protein solution, two negative peaks of OVA and Lyz altered, especially at 208 nm. With the increasing concentrations of micromolecular monomers, their destructive influences on the secondary structure of both proteins were increasingly significant. Compared to the micromolecular monomers, an appropriate content of macromolecular chain could maintain the stability of protein structure. When mass ratio between OVA, Lyz and the macromolecular chains were 1 : 5 and 1 : 15 respectively, the native secondary structure of protein could be reserved. However, the further increases of macromolecular chains destroy the secondary structure of protein. This might be due to the fact that the excessive content of macromolecular chain increases the possibility of accessing the inside of protein by their segments, which can result in the unfolding of protein secondary structure. Furthermore, it is also found that the stabilized effect of macromolecular chains varied with different proteins. It seems like the smaller the protein is, the better the stability of structure, indicating that a larger protein is more likely to be permeated by the segments of macromolecular chain. Consequently, the results demonstrated that an appropriate content of macromolecular chain used as functional monomer results in the preparation of protein imprinted polymers with the preservation of protein secondary structure.

Fig. 2 Effect of micro-molecular monomers or macromolecular chains on the secondary structure of OVA and Lyz, M and C are represented micromolecular monomers and macromolecular chain respectively.

In order to verify the stabilizing mechanism of macromolecular chain, synchronous fluorescence spectra were recorded to study the changes in the micro-environment in the vicinity of the chromophore.²⁶ When the difference between excitation wavelength and emission wavelength (Δλ) were 15 and 60 nm, the spectra provided characteristic peaks for Tyr and Trp residues respectively.²⁷ The possible shift of the emission peak is related to the change of the micro-environment in the vicinity of the chromophore. A red shift implies the polarity of the cavity increases; while a blue shift indicates there is now a more hydrophobic environment around the Tyr or Trp residues. As shown in Fig. 3, when the micromolecular monomers were incubated with OVA and Lyz respectively, their fluorescence intensity of Tyr residues decreased with the red shift of their emission peak indicating that the polarity around the chromophores increased. A similar phenomenon was observed in the synchronous fluorescence spectrum of Trp by incubating with micromolecular monomers. However, when the protein solution was incubated with macromolecular chains, only fluorescence quenching of Tyr and Trp was found in synchronous fluorescence spectrum. The results indicated that the affinity of macromolecular chain to OVA and Lyz did not alter protein's internal micro-environment. Thus, the possible mechanism for the destructive and stabilizing effects on OVA and Lyz involves micromolecular monomer easily permeating the protein surface and subsequently destroying the hydrogen bond which maintain the structural stability of the protein, whereas macromolecule chain tends to interact with the protein surface, where it is difficult to diffuse into the proteins' internal environment and disturb their internal structure.

Fig. 3 Synchronous fluorescence spectrums of OVA and Lyz influenced by micromolecular monomers or macromolecular chains, (a), (b), (c) and (d) Δλ = 15 nm; while (A), (B), (C) and (D) Δλ = 60 nm, M and C are represented micro-molecular monomers and macromolecular chain respectively.

Consequently, in order to ensure the structural stability of the template protein, the mass ratio of 1 : 2.95 and 1 : 9.48 between OVA, Lyz and the macromolecular chain was employed in the preparation of MIH-C-OVA and MIH-C-Lyz.

Synthesis and characterization of molecular imprinted hydrogel

The molecular imprinted hydrogels for OVA and Lyz were prepared either using macromolecular chains or its equivalent micromolecular monomers. As can be noticed, allyl group which has relatively low reactivity compared to vinyl are employed in the preparation of hydrogels. In this work, the choice of using allyl could make two double bonds of crosslinker ([AVIM]Cl) with different polymerization reaction activity. Therefore, during the preparation of imprinted hydrogel made by micromolecular monomers, the relative high reactivity of vinyl in HEA, VIM and [AVIM]Cl were preferentially polymerized to form the polymer chain, and the chain was crosslinked by the group of allyl in the latter part of the polymerization. This polymerized process is much alike as the preparation of imprinted hydrogel with macromolecular chain. Therefore, the use of allyl makes the hydrogels either composed of micromolecular monomers or macromolecular chain more comparable.

The FT-IR spectroscopy was carried out to characterize the chemical composition of MIH-C and MIH-M, the detailed analysis was represented in ESI and Fig. S2.† The results demonstrated that both MIH-C and MIH-M had been successful polymerized, and MIH-M had the same chemical composition as MIH-C.

To assess the degree of polymerization for MIH-C and MIH-M, the gel fraction yield was carried out in the characterization of hydrogels. As seen in Table S1,† the gel fraction yields of 68–82% were obtained for all the synthesized hydrogels, which confirmed a modest progress of polymerization reaction as compared to the reported values of 70–80%.²⁸ In order to investigate the crosslinking degree of hydrogel, swell ratio was also employed in the experiment. As shown in Table S1,† all the NIHs possessed the lower swell ratio compared to MIHs, indicating there is a higher crosslinking degree in NIH. This might be due to hindrance effect of protein with the large size on crosslinking reaction and resulted in a looser structure of MIH. It could also be seen that the swell ratio of MIH-C was a little larger than MIH-M, suggesting a slightly lower crosslinking degree for MIH-C, which might be due to the imbedding of reaction sites in macromolecular chain during the preparation of MIH-C. SEM was carried out to investigate the morphology of MIH-M and MIH-C. The images of SEM in Fig. 4 showed that both MIH and NIH exhibited a characteristic cellular structure with the interpenetrating porous network, which benefited the adsorption and transmission of protein.^29,30 Furthermore, the hydrogels made by macromolecular chain possessed similar morphology compared to those prepared by micromolecular monomers.

Fig. 4 SEM of MIH-C-OVA, MIH-M-OVA, MIH-C-Lyz, MIH-M-Lyz, NIH-C and NIH-M.

Adsorption properties of MIH-C and MIH-M

Adsorption dynamic studies were carried out for MIH-C-template and MIH-M-template to investigate their rate of adsorption, and their binding capacities were determined as a function of time. As shown in Fig. 5, the dynamic curve showed that the binding rate of the imprint hydrogels was faster than the non-imprinted ones in 1.0 mg mL⁻¹ protein solution within 120 min. This indicates that the interactions between MIH and template protein were stronger than NIH, probably because of the formation of template-sized cavities in MIH. In addition, it can be seen that the binding rate of both MIH-C-OVA and MIH-C-Lyz were higher than MIH-M-OVA and MIH-M-Lyz indicating more powerful binding site in MIH-C-template, which might be due to more integral structure of template protein during the preparation of MIH-C-template.

Fig. 5 Adsorption dynamic curve of MIH-C, NIH-C, MIH-M and NIH-M for OVA and Lyz respectively.

In characterizing the adsorption behaviours of the MIH-C-template and MIH-M-template, they were subjected to the isothermal experiments. As seen in Fig. 6, the amounts of protein adsorbed on MIH or NIH increased with an increase in protein concentration. Both MIH-C-template and MIH-M-template had higher binding capacities than non-imprinted hydrogel, indicating the presence of molecular recognition sites in MIH.^31,32 In addition, the binding capacities and IF value of MIH-M-template were significantly lower than MIH-C-template. It might be caused by the distorted template protein during the preparation of MIH-M-template.

Fig. 6 Adsorption isotherm of MIH-C, NIH-C, MIH-M and NIH-M for OVA and Lyz respectively.

In order to further investigate the affinity of MIH-C-template and MIH-M-template to their template protein, their adsorption behaviours were described by the Langmuir adsorption equation as Q = Q_mK_mC_e/(1 + K_mC_e), where Q_m and K_m are the Langmuir constants related to the theoretical maximum adsorption capacity and median binding affinity.^33,34 As shown in Table 1, the hydrogels have a lower Q_m for Lyz compared to OVA, which might be due to the electrostatic repulsion between Lyz and ionic imidazole ring in the neutral pH, while a higher K_m value for Lyz might be caused by their relative small molecular size which makes them lower mass transfer resistance in the micropore of hydrogels during the binding process. The K_m and Q_m value of both MIH-C-template and MIH-M-template were higher than NIH-C and NIH-M respectively indicating the existence of imprinted cavity in MIH. Additionally, compared to MIH-M-template, MIH-C-template possessed bigger K_m and Q_m values indicating much stronger affinities between imprinted cavities and the template protein in MIH-C-template. Since the similar preparation method was used in both MIH-C-template and MIH-M-template, and their hydrogels had the same chemical compositions and similar physical structures, the stronger affinities and more accurate imprinted cavities formed in MIH-C-template might be due to the more “natural” protein structures during the preparation of imprinted hydrogels.

Table 1 Theoretical maximum capacity (Q_m) and Langmuir adsorption equilibrium constant (K_m) from the Langmuir model

	MIH-C-OVA	NIH-C	MIH-M-OVA	NIH-M
Km (mL mg^−1)	2.74 ± 0.18	0.91 ± 0.07	1.79 ± 0.12	1.04 ± 0.09
Qm (mg g^−1)	104.2 ± 6.4	65.4 ± 4.3	83.3 ± 5.2	66.2 ± 4.6
R^2	0.9916	0.9849	0.9852	0.9926

---

	MIH-C-Lyz	NIH-C	MIH-M-Lyz	NIH-M
Km (mL mg^−1)	3.76 ± 0.23	1.37 ± 0.11	2.78 ± 0.17	1.57 ± 0.14
Qm (mg g^−1)	64.9 ± 4.1	41.2 ± 3.8	61.0 ± 4.7	48.8 ± 3.3
R^2	0.9915	0.9969	0.9988	0.9902

---

In order to investigate the selectivity of MIH-C-template and MIH-M-template, the hydrogels were subjected to the selective adsorption experiments. Due to the similar pI and molecular weight as OVA and Lyz, BSA and cytochrome C (Cyc) were chosen as the reference protein respectively. As shown in Table S2,† MIH-M-OVA had relatively good selectivity to Lyz and Cyc with the corresponding β value of 1.77 and 1.76 respectively. However, MIH-M-OVA exhibited poor selectivity to BSA with the β value of 1.14, indicating inaccurate imprinted cavities in the hydrogels. This might be due to the distorted structure of protein during the preparation of MIH-M-OVA. In contrast with MIH-M-OVA, MIH-C-OVA showed excellent recognition ability to BSA, and their β value could reach 2.23, demonstrating structural integrity of template protein during the preparation of MIH-C-OVA contributed to better selectivity. Similar phenomenon and result was also observed in the selective adsorption experiments for MIH-C-Lyz and MIH-M-Lyz.

The specific recognition ability of MIH-C-template and MIH-M-template was established by the competitive adsorption experiments in a ternary mixture solution. As shown in SDS-PAGE analysis (Fig. 7a), three transverse lines in lane 2 at 66.2, 43.0 and 14.4 kDa represented BSA, OVA and Lyz respectively with each concentration of 1.0 mg mL⁻¹. The lane 3 and 5 showed that the intensity of the OVA band eluted from MIH-C-OVA and MIH-M-OVA was significantly higher compared to non-imprinted ones (lane 4 and lane 6) indicating that the imprinted cavities were successfully formed in these hydrogels. Moreover, MIH-M-OVA also extracted other proteins more in contrast to MIH-C-OVA, indicating the better specific recognition ability of MIH-C-OVA to the template protein. Similar phenomenon was also observed in SDS-PAGE analysis of MIH-C-Lyz and MIH-M-Lyz (Fig. 7b). The above results demonstrated that more integral template protein could contribute to better selectivity of MIH, thus exhibiting the advantage of using macromolecule to imprint protein in imprinting technology.

Fig. 7 SDS-PAGE analysis of competitive adsorption of MIH-C-template and MIH-M-template, and their template is OVA (a) and Lyz (b) respectively. Lane 1, protein molecular weight marker; Lane 2, the mixture solution of BSA, OVA and Lyz; Lane 3, the protein eluted from MIH-C-template; Lane 4, the protein eluted from NIH-C; Lane 5, the protein eluted from MIH-M-template; Lane 6, the protein eluted from NIH-M.

Conclusions

In conclusion, the preliminary study confirmed the assumption that the micromolecule can easily permeate into the inside of protein disrupting its internal structure, while an inflexible macromolecule may interact with the surface of protein and thus maintain the integrity of the template protein's structure. By comparing the imprinted hydrogels either composed of the macromolecular chains or its micromolecular component monomers, it was found that MIH-C-template exhibited greater template recognition and affinities thus highlighting the marked advantage of using macromolecules to imprint protein. Furthermore, the architecture and composition of macromolecule used as functional monomer could be designed according to the biological template in contrast with the monotonous micromolecularly functional monomers. Consequently, we believe this strategy could significantly enhance the selectivity and specific recognition ability of MIPs, thus would promote the development of biomacromolecule imprinting technology.

Acknowledgements

This work was supported by the National Nature Science Foundation of China (Grant no. 21174111 and 51433008).

Notes and references

1. S. A. Piletsky and A. P. F. Turner, Electroanalysis, 2002, 14, 317–323.
2. A. Nematollahzadeh, W. Sun, C. S. A. Aureliano, D. Lütkemeyer, J. Stute, M. J. Abdekhodaie, A. Shojaei and B. Sellergren, Angew. Chem., Int. Ed., 2011, 50, 495–498.
3. U. Latif, A. Mujahid, A. Afzal, R. Sikorski, P. Lieberzeit and F. L. Dickert, Anal. Bioanal. Chem., 2011, 400, 2507–2515.
4. T. Renkecz, G. Mistlberger, M. Pawlak, V. Horvath and E. Bakker, ACS Appl. Mater. Interfaces, 2013, 5, 8537–8545.
5. S. Liu, H. Yan, M. Wang and L. Wang, J. Agric. Food Chem., 2013, 61, 11974–11980.
6. M. Jenik, A. Seifner, P. Lieberzeit and F. L. Dickert, Anal. Bioanal. Chem., 2009, 394, 523–528.
7. E. Verheyen, J. P. Schillemans, M. V. Wijk, M. A. Demeniex, W. E. Hennink and C. F. V. Nostrum, Biomaterials, 2011, 32, 3008–3020.
8. J. K. Awino and Y. Zhao, J. Am. Chem. Soc., 2013, 135, 12552–12555.
9. V. Venkatachalam, D. Brinks, D. Maclaurin, D. Hochbaum, J. Kralj and A. E. Cohen, J. Am. Chem. Soc., 2014, 136, 2529–2537.
10. M. Jenik, R. Schirhagl, C. Schirk, O. Hayden, P. Lieberzeit, D. Blaas, G. Paul and F. L. Dickert, Anal. Chem., 2009, 81, 5320–5326.
11. M. Mahmo-udi, S. Bonakdar, M. A. Shkrgozar, H. Aghaverdi, R. Hartmann, A. Pick, G. Witte and W. J. Parak, ACS Nano, 2013, 7, 8379–8384.
12. J. Liu, Q. Deng, D. Tao, K. Yang, L. Zhang, Z. Liang and Y. Zhang, Sci. Rep., 2014, 4, 5487–5492.
13. L. Chen, S. Xu and J. Li, Chem. Soc. Rev., 2011, 40, 2922–2942.
14. N. W. Turner, C. W. Jeans, K. R. Brain, C. J. Allender, V. Hlady and D. W. Britt, Biotechnol. Prog., 2006, 22, 1474–1489.
15. T. Takeuchi and T. Hishiya, Org. Biomol. Chem., 2008, 6, 2459–2467.
16. S. Li, S. Cao, M. J. Whitcombe and S. A. Piletsky, Prog. Polym. Sci., 2014, 39, 145–163.
17. D. R. Kryscio, M. Q. Fleming and N. A. Peppas, Macromol. Biosci., 2012, 12, 1137–1144.
18. D. R. Kryscio, M. Q. Fleming and N. A. Peppas, Biomed. Microdevices, 2012, 14, 679–687.
19. R. Dan, Y. Z. Wang, L. Du, S. H. Du, M. D. Huang, S. Yang and M. Zhang, Analyst, 2013, 138, 3433–3443.
20. S. Yang, Y. Z. Wang, M. L. Xu, M. Z. He, M. Zhang, D. Ran and X. P. Jia, Anal. Methods, 2013, 5, 5471–5477.
21. E. P. Herrero, E. M. Martín Del Valle and N. A. Peppas, Ind. Eng. Chem. Res., 2010, 49, 9811–9814.
22. M. J. Guo, Z. Zhao, Y. G. Fan, C. H. Wang, L. Q. Shi, J. J. Xia, Y. Long and H. F. Mi, Biomaterials, 2006, 27, 4381–4387.
23. D. Liu, Q. Yang, S. Jin, Y. Song, J. Gao, Y. Wang and H. F. Mi, Acta Biomater., 2014, 10, 769–775.
24. A. Derazshamshir, G. Baydemir, M. Andac, R. Say, I. Y. Galaev and A. Denizli, Macromol. Chem. Phys., 2010, 211(6), 657–668.
25. G. Demirel, G. Ozcetin, E. Turan and T. Caykara, Macromol. Biosci., 2005, 5, 1032–1037.
26. Z. Hu and C. L. Tong, Anal. Chim. Acta, 2007, 587, 187–193.
27. X. R. Pan, R. T. Liu, P. F. Qin, L. Wang and X. C. Zhao, J. Lumin., 2010, 130, 611–617.
28. F. M. Plieva, J. Anderssin, I. Y. Galaev and B. Mattlasson, J. Sep. Sci., 2004, 27, 828–836.
29. L. Qin, X. W. He, W. Zhang, W. Y. Li and Y. K. Zhang, Anal. Chem., 2009, 81, 7206–7216.
30. D. M. Hawkins, D. Stevenson and S. M. Reddy, Anal. Chim. Acta, 2005, 542, 61–65.
31. A. Bossi, F. Bonini, A. P. F. Turner and S. A. Piletsky, Biosens. Bioelectron., 2007, 22, 1131–1137.
32. H. Hiratani, Y. Mizutani and C. Alvarez-Lorenzo, Macromol. Biosci., 2005, 5, 728–733.
33. H. F. EL-Sharif, D. M. Hawkins, D. Stevenson and S. M. Reddy, Phys. Chem. Chem. Phys., 2014, 16, 15483–15489.
34. S. K. Tsermentseli, P. Manesiotis, A. N. Assimopoulou and V. P. Papageorgiou, J. Chromatogr. A, 2013, 1315, 15–20.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra08246c

---