Antibiotics constitute one of the emerging categories of persistent organic pollutants, characterised by their expansion of resistant pathogens. Antibiotic pollutants create a major public health challenge, with already identifiable detrimental effects on human and animal health. A fundamental aspect of controlling and preventing the spread of pollutants is the continuous screening and monitoring of environmental samples. Molecular imprinting is a state-of-the-art technique for designing robust biomimetic receptors called molecularly imprinted polymers (MIPs), which mimic natural biomolecules in target-selective recognition. When integrated with an appropriate sensor transducer, MIP demonstrates a potential for the needed environmental monitoring, thus justifying the observed rise in interest in this field of research.
1. Rational Design of Antibiotic Molecularly Imprinted Polymers (MIPs)
The success of molecular imprinting relies on the choice of template, functional and cross-linking monomers, as well as a suitable porogenic solvent. A detailed analysis of the significance of each component was previously reported
[1]. In summary, a template could be any molecule, ion, macromolecule, compound or whole cell that has functional moiety that can be harnessed for chemical interactions. Mostly, the target analyte serves as the template during MIP preparation. The polymerisation solvent plays important roles, including solubilisation of all the monomers in the pre-polymerisation mixture, stabilisation of template-monomer complexes and acting as a porogen that helps to control the porosity of the resulting polymer. Following polymerisation, functional groups of the monomers are held in position by cross-linking the polymeric structure, thereby cementing their orientation within the polymer after template extraction.
Indeed, MIP recognition of a target is greatly influenced by the strength of the interaction between the template and functional monomer during the pre-polymerisation complex formation stage that then dictates the binding affinity, usually represented by the dissociation constant (i.e., KD) of the MIP for the targets at rebinding. A lower KD would generally suggest that the recognition layer possesses a greater fraction of high-affinity binding sites, due to a strong interaction (high binding energy) between the template and the monomer that was formed during the pre-polymerisation stage. Moreover, a strong interaction between the template and functional monomer is critical, especially in the analysis of small targets, such as antibiotics, where a high binding affinity (low KD) is required to analyse low concentrations of analytes.
Depending on the interaction type in the template–monomer complex, molecular imprinting approaches are generally categorised as covalent and noncovalent. The covalent approach employs reversible covalent bonds between the functional monomer and template, such as boronate ester, ketal/acetal, and Schiff’s base formation. This strategy leads to the generation of a higher yield of specific and more homogeneous binding sites, while the applicability is limited because of the small number of compounds bearing required functionalities (alcohols (diols), aldehydes, ketones, amines, and carboxylic acids). Another disadvantage of covalent imprinting is the complicated template removal and slow binding kinetics of the resulting MIP. On the contrary, the most frequently employed noncovalent approach offers a wide variety of functional monomers, with high flexibility and rapid rebinding kinetics. This approach is based on noncovalent interactions, such as hydrogen bonding, electrostatic interactions and van der Waals forces in the template–functional monomer complex. In addition, molecular imprinting based on metal-ion coordination was introduced
[2][3]. In this approach, the functional monomer and template are bridged through coordination binding with various metal ions, such as Cu
2+, Ni
2+, Cd
2+ or Zn
2+. Compared to noncovalent interactions, metal-ion coordination bonds are stronger, leading to better stability of the MIP in aqueous media. Hence, the selection of an appropriate functional monomer capable of forming a stable complex with a target analyte via the reversible covalent, noncovalent or metal-ion coordination is of great importance for a successful imprinting process. Notwithstanding, not all monomers fulfilling this obligation are suitable candidates. This is because the appropriate monomer must also be compatible with the desired polymerization approach. Therefore, a potential functional monomer should possess chemical groups for polymerisation and formation of strong interactions with template molecules. Although such a demand may suggest a limitation in the number of suitable functional monomers for molecular imprinting, new functional monomers tailor-made to accommodate these challenges are being synthesised
[4][5][6][7][8].
Moreover, cross-linking monomers are essential for controlling the morphology of the MIP matrix during polymerisation by fixing functional monomers around template molecules. Typically, an insufficient amount of crosslinker reduces the structural stability of the polymer, whereas an excess reduces the number of MIP binding sites. Reportedly, most commercial crosslinkers, such as ethylene glycol dimethacrylate (EGDMA), are compatible with molecular imprinting
[9]. However, the selection of crosslinkers should be based on their solubility in the synthesis medium and the strength of the interaction with the template. It was reported that the crosslinker that displays lower binding of the template should be preferential because it generates MIPs with lower non-specific binding and a higher imprinting factor, and therefore specificity
[10]. If the level of non-specific binding background on the MIP is high, a similar polymeric material, i.e., non-imprinted polymer (NIP) generated in the absence of the template, might serve as a good reference, allowing one to compensate the background and accurately analyse label-free responses originating from a MIP modified sensor.
Various functional monomers exist for building polymer matrices for molecular imprinting.
Figure 1 shows the structure of monomers commonly employed in antibiotic imprinting. Among these, methacrylic acid (MAA) is the most utilised. This is partly because hydrogen bonding constitutes one of the dominant interactions employed in MIP research and MAA can serve as a hydrogen bond acceptor or donor. Moreover, it was revealed that the dimerisation of MAA can improve, to an extent, the success of molecular imprinting
[11]. However, the growth in scientific efforts has resulted in an increasing utilisation of other monomers, thereby removing the monopoly in MAA usage.
Figure 1. Structures of commonly used functional monomers for molecular imprinting.
The understanding of the noncovalent interactions between a template and functional monomer could be obtained through computational software that assists in studying these interactions at the molecular level, thereby allowing the optimisation of factors that affect the performance of the MIP-based system. The adaptation of computational approaches for rational design of MIPs requires calculating binding energies between a template and functional monomer(s) and designing the more specific and selective recognition sites in the polymers
[12]. The computational approach demonstrates superior advantages over experimental trial and error due to its time efficiency, low costs, and the possibility to circumvent the use of expensive and toxic chemicals.
In molecular imprinting research, commonly employed computational approaches for estimating noncovalent interactions between a template molecule and a monomer in pre-polymerisation solution include quantum chemical calculations (QCC), molecular docking (MD), molecular mechanics (MM), and molecular dynamics. For small molecular weight analytes, QCC has become the most promising approach for calculating hydrogen binding energies between the template and monomer
[13]. To perform a QCC calculation, the geometry of the monomer-template complexes and their individual compounds are optimised by semi-empirical (SE) parameterised method 3 (PM3). This is then followed by the estimation of the binding energy by either density functional theory (DFT) or Hartree–Fock methods. Based on the DFT method that is commonly used to estimate hydrogen binding energy between small molecular-weight templates (e.g antibiotics) and monomers, a hydrogen bond interaction is usually classified as strong (>63 kJ mol
−1), moderate (16.8–63 kJ mol
−1), and weak (<16.8 kJ mol−1)
[14]. A brief account of antibiotic MIP-based sensing research preceded by a painstaking but rewarding computationally assisted selection of the functional monomer is provided here, while a more comprehensive list is shown in
Table 1.
Tadi et al.
[15] achieved a rational selection of pyrrole, as the functional monomer for the imprinting of sulfanilamide among many monomers, with the help of QCC and DFT methods. The high binding energy obtained between the template and monomer was traceable to the hydrogen bond formed between S=O and -NH2 groups of the template and the -N and -CH groups of pyrrole. After electrodeposition on a pencil graphite electrode and parameter optimisation, the MIP demonstrated a discriminatory recognition for the target and the assay indicated a significant LOD of 20 nM. Similarly, to select an appropriate monomer for the preparation of amoxicillin MIP on QCM, QCC was employed to estimate the binding energies between the target and several electropolymerisable monomers including pyrrole, meta-phenylenediamine (mPD) and 3,4-ethylenedioxythiophene (EDOT)
[16]. The study indicates the highest binding energy for mPD (273 kJ/mol), as compared to 63 and 8 kJ/mol for pyrrole and EDOT, respectively. The significance of this study was reflected in the sensor’s superior binding of the target (about 7 times higher adsorption capacity) over its reference NIP-based sensor.
Although MD and molecular dynamics are conventional in the rational monomer selection for imprinting of macromolecules, e.g., protein, their implementation is extended to small molecular weight templates
[17][18]. Thus, in two separate reports
[19][20], MD was applied for the rational selection of a monomer for the imprinting of norfloxacin using CDOCKER and SYBYL software. In both cases, MAA gave the highest binding energy among the tested monomers, including acrylamide, methacrylamide, polyvinylpyrrolidone, and MAA. Although the binding energies in both reports were slightly different (100.33 vs. 87.45 kJ/mol), a pronounced difference in the affinity (0.004 vs. 2.06 μM) and relative adsorption capacity (4.3 vs. 2.4) was observed. This further indicates the importance of accounting for molecular interaction between a template and a functional monomer and that a small change in the effectiveness of the interaction might lead to a significant effect on the performance of the MIP-based sensor.
This entry is adapted from the peer-reviewed paper 10.3390/bios12070441