A Specific Artificial Antibody toward Mycophenolic Acid Prepared by Molecular Imprinting

Molecular imprinting is the preparation of polymeric materials with specific binding sites for a molecule. A monomer is allowed to interact with the molecule of interest (template) to create low-energy interactions. Polymerization is then induced with a bridging agent and heat or ultraviolet irradiation. During the process of polymerization, the molecule of interest is entrapped within the polymer, which finally can be crushed, sieved, and washed with highly polar solvents to remove the template molecule. The imprint of the template is maintained in the rigid polymer, which is usually made of plastic materials such as acrylics and styrenes. The molecular imprint contains many small crypts with shapes complementary to the interest molecule and which are stabilized by chemical groups oriented during the polymerization process in the presence of the substrate. The imprinted polymer can bind the original molecule with high specificity similar to the specificities observed in enzyme-substrate or antigen-antibody interactions. The high specificity of the binding led to the concept of artificial antibodies (1). The use of artificial antibodies in competitive assays is the most promising application of molecular imprints in clinical chemistry. It has been shown that they can replace monoclonal antibodies from animal origin in certain immunologic assays such as those for cortisol, theophylline, diazepam, morphine, s-propanolol, and methylglucoside (1). Artificial antibodies show numerous advantages compared with natural antibodies: Ease of production Artificial antibodies can be prepared without laboratory animals, and the preparation protocol is relatively simple. Chemical stability Natural antibodies are labile, whereas artificial antibodies are very robust. They can be used with strong acids, bases, and organic solvents, and are temperature- and pressure-resistant. They can be preserved for many years without loss of their specificity. Reusable The natural antibody-antigen bond is stable and almost irreversible. The binding to an artificial antibody is reversible, and the molecular imprint can be used hundred of times, especially in HPLC and competitive assays. Organic solvent compatibility Natural antibodies are denatured in the presence of organic solvents. Artificial antibodies can be used in the presence of solvents such as chloroform, acetonitrile, and toluene. Mycophenolate mofetil (MMF) is an ester derivative of mycophenolic acid (MPA) and is approved as an immunosuppressant drug in renal transplant patients. The prodrug MMF is rapidly transformed in vivo to the active immunosuppressant MPA (2), which inhibits inosine monophosphate dehydrogenase 2. It thus suppresses the de novo synthesis of guanosine nucleotides especially in T and B lymphocytes and stops their proliferation (3). Clinical trials in renal transplant patients have shown that the addition of MMF to steroids and cyclosporine immunosuppressive regimens can reduce rejection episodes observed early after transplantation as well as the incidence of acute rejection. Although treatment with MMF has been shown to have a good safety profile, it may be useful to monitor MPA concentrations in serum (4)(5)(6). HPLC methods specific for MPA have been described (7). An immunoenzymatic assay (Emit) is also commercially available for the determination of MPA in serum (8). It is more automated and requires less technical work than the HPLC methods. The Emit assay yields results that are higher than those obtained with the HPLC methods. This is believed attributable to the cross-reactivity of MPA metabolites with the monoclonal antibodies. MPA is metabolized mainly to a glucuronide derivative (MPAG), which is believed inactive; however, the MPAG concentrations in sera of patients treated with MMF are an order of magnitude higher than those of MPA, and thus its potential for interference in the various analytical methods is high. Recently, two other metabolites of MPA have been identified (9). One of these metabolites, the acyl glucuronide of MPA, is believed to be responsible for the higher results observed with the Emit method. The Emit method has relatively high reagent costs, and the stability of reagents is limited. With the increasing popularity of the serum determination of MPA in transplant patients (10), there is a need for a fast, economical, and specific assay for the determination of MPA in serum. In this study, we prepared a molecular imprint of MPA by polymerization of methacrylic acid (MAA) in the presence of MPA. The polymer is easily produced, inexpensive, and highly specific. We set up the analytical conditions for binding MPA in an aqueous medium. The preparation and characterization of the binding of MPA to this polymer are described in this report. For this study, MPA was obtained from Sigma. Acetonitrile, methanol, and toluene were obtained from Fisher Scientific. MAA, ethylene glycol dimethacrylate, and 2,2′-azobis(isobutyronitrile) were obtained from Aldrich. MPAG was isolated in our laboratory from the urine of transplant patients. The details of the polymer synthesis have been described by Muldoon and Stanker (11). Briefly, the molar ratios of print molecule (MPA), functional monomer (MAA), and cross-linker (ethylene glycol dimethacrylate) were 1:4:20. To initiate the polymerization, 4.125 mg of 2,2′-azobis(isobutyronitrile) was added. The mixture was deoxygenated continuously with a stream of nitrogen during polymerization at 60 °C for 2 h. The polymer was then crushed and ground repeatedly. The particles were free from MPA after soaking in 100 mL of methanol for 12 h. After a final drying at 60 °C, the polymer was ready for use. The polymerization produced a rigid solid that was ground into small particles averaging 25 μm in diameter. An important characteristic of the polymer is its robustness, which gives it the ability to retain its specific binding sites even in harsh conditions. This property also gives the polymer the advantage of being stored at ambient temperatures for many months without loss of its recognition capabilities (12). We determined that by passing a constant stream of nitrogen through the mixture during the polymerization, we reduced the time of polymerization to 2 h. Although it has been shown that the number of binding sites and the affinity at which the polymer binds to the print molecule can be affected by the conditions of polymerization (13), the deoxygenation process did not affect the affinity of binding or the number of binding sites of our polymer. Typically, the binding of MPA to the polymer is performed in a 0.01 mol/L phosphate buffer, pH 2.5, containing 5 mL/L methanol. This concentration of methanol was shown to be optimal. Most binding studies with molecular imprinted polymers have been done in organic solvents (1). The advantage of working in an aqueous binding solvent is that it is closer to the natural environment of human plasma. This characteristic could be useful in the development of an heterogeneous binding assay using the molecularly imprinted polymer. The binding of MPA to the polymer is affected by the pH of the binding solvent (2). As shown in Fig. 1A , the imprinted polymer gains affinity for MPA with a decrease of pH. In an acidic solvent, the carboxyl group of MPA should predominately be in its anionic form. Thus, we can make the hypothesis that the binding of MPA occurs by interaction of its carboxyl group with the polymer through hydrophobic interactions. To study the binding specificity of the polymer, a comparison of the binding to imprinted and nonimprinted polymers was performed (Fig. 1B ). We defined specific binding as the difference between binding to imprinted and nonimprinted polymers. It was determined that the artificial antibody could bind MPA with high specificity up to an average concentration of 125 mg/L. A Scatchard analysis was used to estimate the strength of binding of the print molecule (Fig. 1C ). This analysis clearly gave a nonlinear plot, which is typical of imprinted polymers and reflects the heterogeneity of the binding sites present. A two-site model was used to calculate the binding strength (KD) and the site populations (Bmax) of the polymer. The molecularly imprinted polymer gave KDs of 0.88 and 100 μmol/L and Bmax values of 15.0 and 233 μmol/g, respectively, which is comparable to other imprinted polymers (1). These results show the presence of binding sites with high affinity and sites with low affinity. The low-affinity sites are present in much greater numbers, even in the polymer prepared without MPA, and represent nonspecific binding to the polymer. Interference studies were performed with several drugs that may be co-administered with MPA. The drugs tested included acetylsalicylic acid, butobarbital, caffeine, cyclobarbital, diazepam, phenobarbital, phenytoin, salicylic acid, theophylline, and thiobarbital. The metabolite MPAG was also tested at a final concentration of 1500 μmol/L. All of the compounds tested did not interfere with the binding of MPA to the polymer. It was determined in our laboratory (data not shown) that plasma samples from renal transplant patients treated with MPA show a MPAG:MPA molar ratio of 20.4 ± 21.2 (n = 53). In our experiment, the glucuronide was tested to a ratio of 100:1 (MPAG:MPA) and was shown not to interfere. In summary, we prepared a molecularly imprinted polymer that can specifically bind MPA in an acidic aqueous medium. The polymer contains two distinct binding sites. The high affinity sites are responsible for the specific binding of MPA. The imprinted polymer has a hydrophobic interface, which is believed to interact with the protonated carboxyl group of MPA in an acidic medium. Current work is being directed toward the development of a heterogeneous binding assay for MPA in plasma with the molecularly imprinted polymers. Analysis of MPA binding by an anti-MPA polymer. Binding to the anti-MPA polymer was determined in a 0.01 mol/L phosphate buffer containing 5 mL/L methanol. Polymer (125 mg/L) was added to the solution and incubated for 30 min at room temperature with agitation (Eberbach). The polymer was centrifuged, and MPA was measured in the supernatant by HPLC. The HPLC procedure is based on the method of Shipkova et al. (7). The equipment consisted of a Hewlett Packard series 1100 pump, an automatic injector, an ultraviolet detector, a Supelco LC-318 column (25 cm × 4.6 mm; 5 μm particles), and a Hewlett Packard Kayak XA, Pentium II interface. The mobile phase consisted of 0.01 mol/L phosphoric acid buffer, pH 5.0 (adjusted with NaOH), containing 375 mL/L acetonitrile. The flow rate was 1.0 mL/min, and MPA was detected at 215 nm with a retention time of 7 min. The temperature was kept at 40 °C. (A), pH of the binding medium was adjusted from 2.5 to 7.5 with NaOH. (B), 50 to 1000 mg/L of polymer was added to the binding medium, which contained 15 μmol/L MPA. The dashed line and arrow indicate the region of specific binding. (C), Scatchard analysis. Dissociation constants (KD) were determined as 0.88 and 100 μmol/L, corresponding to high- and low-affinity sites, respectively. Site populations were 15 and 233 μmol/g polymer, respectively.