2.1 Materials

GSF, fluconazole, ketoconazole, clotrimazole, dexamethasone purchased from Aladdin Co. (Beijing, China). The 3-aminopropyltriethoxysilane (APTES), and naproxen were supplied from J&K Chemical Co. (Beijing, China). Dehydrogriseofulvin was purchased from Toronto Research Chemicals Inc. (Toronto, Canada). Silica gel (diameter: 50 μm, average pore radius: 43.54 nm) was supplied by Tokyo Chemical Co. (Tokyo, Japan). Methacrylic acid (MAA) was purchased from Tianjin Chemical Reagent Plant (Tianjin, China). Ethylene glycol dimethacrylate (EDMA) was obtained from Sigma-Aldrich (New Jersey, USA). 2,2′-Azobisisobutyronitrile (AIBN) was purchased from Shanghai No. 4 Reagent Factory (Shanghai, China). SPE cartridges (3 mL) were purchased by Biocomma limited (Shenzhen, China). Commercial C-18 SPE cartridge (100 mg, 3 mL) was purchased from Sepax Technologies, Inc (Newark, USA). Ultra-pure water was purified with Molement 1805b (Shanghai, China). Methanol and acetonitrile were of HPLC grade. All other chemicals used in this study were supplied from local suppliers and were of analytical grade.

2.2 Instruments

The HPLC analysis was performed with a Shimadzu LC-2010AHT series HPLC (Kyoto, Japan) and a Lab-Solution workstation. FT-IR spectra (4000–500 cm⁻¹) were recorded in a Thermo Nicolet Nexus 330 FT-IR spectrometer (Madison, USA). The SEM images were obtained by a TM-1000 Scanning Microscope (Hitachi, Japan). The shaking was performed by a SHA-C Vapour-bathing Constant Temperature Vibrator (Changzhou, China).

2.3 Preparation of surface molecularly and non-molecularly imprinted polymers

The silica particles were first activated and then activated silica particles were modified as amino modified silica particles (denoted as NH2-SiO2) by using APTES. The silica particles activation and modification were performed according to a previous method (Ge et al., 2017). The GSF SMIPs preparation method is as follows. Firstly, GSF (0.5 mM) as a template drug was dissolved in 20 mL of toluene. After that, the functional monomer (MAA, 2 mM) was added in the above mixture. Further, 0.5 g amino modified silica was added in the above solution. The above mixture was pre-polymerized with constant stirring at 30 °C for 2 hrs. After that the cross-linker (EDMA, 10 mM) and an initiator (AIBN, 16.4 mg) were added. The solvents were mixed on a sonication apparatus for 10 min, and then nitrogen was purged for 15 min to remove oxygen. The polymerization flask was sealed for 12 h. under a nitrogen atmosphere and the temperature was maintained at 60 °C. After polymerization, the polymers were subsequently washed and extracted by a mixture of methanol-acetic acid (4:1, v/v) to remove unreacted reagents and template molecules. At the same time, the preparation method of surface non-molecularly imprinted polymers (SNIPs) was the same as that of the imprinted polymers except that no template molecule was added. Graphical abstract Fig. 1 illustrates the scheme for the synthesis of SMIPs.

2.4 Adsorption and selectivity experiment

The adsorption quantities of GSF polymers were assessed by adsorption equilibrium and adsorption kinetics experiments. The analysis was performed by using following HPLC-UV conditions; the column was C₁₈ (Promosil 250 mm × 4.6 mm i.d., 5 µm); mobile phase was methanol-water (70:30, v/v), the flow rate was 1.0 mL/min, the detection wavelength was 291 nm, and the column oven temperature was 30 °C.

For the adsorption equilibrium experiment, 10 mg of SMIPs and SNIPs were added in different test flasks. Each test flask contained 10 mL of drug solution of concentrations (10, 50, 100, 200, 300, 400, 500, 600 µg/mL). The reaction flasks were constantly stirred on the mechanical shaker for 2 h., at 150 revolutions per minute (rpm). The adsorption quantities Qe (µg/mg) of SMIPs and SNIPs and imprinting factor (IF) were estimated by the following equations (Ge et al., 2017).

Furthermore, IF was calculated by the following equation.

The kinetic adsorption tests were performed by using 10 mg each of SMIPs and SNIPs polymers. The concentration of GSF drug solution was kept 400 µg/mL. The mixtures were continuously vibrated on a mechanical shaker at 150 rpm at 25 °C. The concentration of the drug was measured at different time intervals (0, 5, 10, 15, 30, 45, and 60 min). The adsorption amounts and IF were calculated by Equations (1) and (2).

The selectivity experiment was performed by using therapeutic and structural analogues such as fluconazole, clotrimazole, ketoconazole, dehydrogriseofulvin, naproxen, and dexamethasone. For this purpose, 5 mg of polymers were added into the 5 mL of each drug solution. The concentration of each drug solution was 100 μg/mL, and the incubation time was 60 min. at room temperature. Finally the adsorption amounts measured and IF of these drugs were determined.

2.5 Drug standards, spiked plasma, and real plasma sample preparations

A stock solution (1000 µg/mL) of GSF was prepared in methanol. Further the working standard solutions of 0.1, 1, 10, 25, 50, 75, 100, 200, 500 µg/mL, were prepared by serially diluting with methanol-water (70:30, v/v).

Blood samples were collected form untreated rats with heparin. The blood collected from untreated rats was subjected to centrifugation at 12000 rpm for 3 min. The lower layer was removed and clear plasma was collected and stored at −20 °C until used. Plasma samples were spiked with GSF working standard solutions to yield calibration standards in plasma at concentration of 0.1, 1, 10, 25, 50, 75, 100, 200 µg/500 µL. For 0.1 (10 µL of 10 µg/mL of working standard solution plus 490 µL of blank plasma), 1 (10 µL of 100 µg/mL of working standard solution and 490 µL of blank plasma), 10 (100 µL of 100 µg/mL of working standard solution plus 400 µL of blank plasma), 25 (250 µLof 100µg/mL of working standard solution plus 250 µL of blank plasma), 50 (250 µL of 200 µg/mL of working standard solutions and 250 µL of blank plasma), 75 (150 µL of 500 µg/mL of working standard solution and 350 µL of blank plasma), 100 (200 µL of 500 µg/mL of working standard solution and 300 µL of blank plasma) and for 200 (400 µL of 500 µg/mL working standard solution and 100 µL of blank plasma). After spiking a calibration curve was constructed. The spiked plasma samples stored at −20 °C until used.

For real plasma samples, a dose of 50 mg/kg as aqueous suspension was given to two adult male Sprague-Dawley rats. The blood samples were collected after 5 hrs., from the jugular vein, centrifuged at 12000 rpm for 3 min (Wei et al., 2008). The plasma was separated and stored at −20 °C until further analysis. The animal experiments were performed according to the International guidelines of experimental animal care and approval by the ethical committee of Xi’an Jiaotong University (certificate no. 22-9601018). All frozen plasma (blank, spiked and real) samples were thawed at room temperature before use.

2.6 SPE procedure and plasma samples analysis

The SPE cartridge (3 mL) was packed with SMIPs. After that, plasma sample was loaded, and then washing step was performed to wash and remove the impurities. After washing elution solvent was employed. The eluents were collected and evaporated to complete dryness under nitrogen stream at room temperature. After evaporation 1 mL methanol was added in the residual volume and an aliquot of 10 μL was analyzed under HPLC conditions.

For the analysis of plasma samples, a C₁₈ column (Promosil, 250 mm × 4.6 mm, i.d., 5 µm) was used. The mobile phase was a mixture of 20 mM aqueous solution of potassium dihydrogen phosphate and acetonitrile (55:45, v/v). The final pH of the mobile phase was 3.5, which was adjusted with phosphoric acid. The flow rate was 1.0 mL/min, the detection wavelength was 291 nm, and the column oven temperature was set at 30 °C.

3.1 Optimization of surface molecularly imprinted polymers

The amount of template, functional monomer and cross-linking agent commonly affect the morphology, structure, adsorption behavior, and selectivity performance of the polymers. The amount of functional monomer is linked with the adsorption performance of polymer. If, the amount of functional monomer is less, it cannot offer enough functional groups, which ultimately result in less adsorption and if the concentration of functional monomer is more, then non-specific adsorption sites are formed which result in decreased adsorption amount (Chen et al., 2016). In our investigation we used 1:4 M ratios of the template and functional monomer. At this optimized molar ratio, the adsorption amount was higher as compared with other molar ratios (Table S1). In the meantime, the molar quantity of cross-linking agent also disrupted the three-dimensional assembly of the imprinted surface. Due to an increase in the amount of cross linker, a rigid polymer structure was formed, which obstruct the template molecule to reach its recognition site. The mass transfer became difficult and resulted in a decreased adsorption amount. So, a molar quantity of 10 mM of EDMA provided enough sites for binding of the drug (Table S1).

The adsorption behavior and structure of polymers are also reliant on the nature of solvents. The solvents such as toluene, acetonitrile, and chloroform are applied for the synthesis of non-covalent based polymers. Toluene is a less polar solvent and provided good imprinting factor, high adsorption capacity and uniform structure to the polymers (Chen et al., 2016). So, different solvents were used in this experiment. Toluene showed high adsorption amount and IF in comparison to other solvents (Fig. S.2).

3.2 Characterization

3.2.1

3.2.1 SEM analysis

The SEM images of activated silica particles, SNIPs and SMIPs are shown in (Fig. 2a,b,c). The activated silica particles presented a smoother surface as compared to the surfaces of SMIPs and SNIPs. The SMIPs surface was comparatively rougher than the surface of SNIPs and this was the only remarkable difference between these particles. These distinctive features in morphology may ensure the formation of MIP film on the surface of modified silica particles.

Close

Fig. 2

Characterization of silica particles, SMIPs, SNIPs. (a): SEM of Activated silica; (b): SEM of SNIPs; (c): SEM of SMIPs; (d): FT-IR spectra.

Export to PPT

3.3 Evaluation of the adsorption and selectivity characteristics

3.3.1

3.3.1 Adsorption isotherm experiment

The adsorption of the drug by SMIPs and SNIPs gradually increased as the concentration of drug in solution increased until it reached equilibrium concentration level. The equilibrium concentration level was 400 µg/mL (Fig. 3a). At this concentration level, the adsorption amount for SMIPs and SNIPs were 119.1 and 48.8 µg/mg respectively. At this point, the recognition surfaces for the drug molecule were saturated and no more drugs adsorbed after this equilibrium level. The adsorption amount of SNIPs was lower than the adsorption amount of SMIPs. The SMIPs have both the specific binding sites and the non-specific binding sites. However, the SNIPs just have non-specific binding sites. To further elucidate this process, two adsorption models were used to evaluate the adsorption process, Langmuir model and Freundlich model.

(3)

$$\text{Langmuir equation}:1/{\text{Q}}_{\text{e}}=1/\left({\text{K}}_{\text{L}}{\text{C}}_{\text{e}}{\text{Q}}_{\text{m}}\right)+1/{\text{Q}}_{\text{m}}$$

(4)

$$\text{Freundlich equation}:\text{log}{\text{Q}}_{\text{e}}=\left(\text{log}{\text{C}}_{\text{e}}\right)/\text{n}+\text{log}{\text{K}}_{\text{F}}$$ Q_m (mg/g) is the maximum adsorption capacity, K_L (mL/mg) is the Langmuir constant, n and K_F are Freundlich constants (Peng et al., 2016).

Close

Fig. 3

Adsorption experiments (a): adsorption isotherm curves of SMIPs and SNIPs; (b): adsorption kinetics curves of SMIPs and SNIPs; (c): adsorption selectivity of SMIPs and SNIPs.

Export to PPT

According to the data, the linear correlation coefficient (R²) value of Langmuir model (0.9902) was higher than Freundlich model (0.9534). This result indicated that the adsorption process belonged to the Langmuir adsorption process and was a monolayer adsorption process.

3.4 SPE optimization

A 25 μg/mL of spiked plasma solution was used in this experiment to optimize the SMISPE process. The amount of adsorbent material is a major contributing factor responsible for good recoveries. A high amount of adsorbent material assures more adsorption sites are available for target analyte. Firstly, we explored different amounts of the polymers such as 10, 50, 100 and 150 mg of polymer with 0.5 mL volume of the sample. Fig. 4a shows that 100 mg of the polymer resulted in a maximum recovery about 81.9%. While the lower or higher amount of polymers did not achieve maximum recovery. So, 100 mg SMIPs was packed in SPE cartridge. In washing step, solvents such as water, water: methanol (7:3, v/v), water: methanol (5:5, v/v) and methanol were optimized. The volume of each washing solvent used in washing step was 2 mL and a recovery of 85.4% was achieved by water: methanol (7:3, v/v) (Fig. 4b). Washing step was carried out to reduce the interference of substances and to achieve a good clean recovery. Subsequently, different types of elution solvents were applied such as methanol-acetic acid (4:1, v/v), methanol-acetic acid (9:1, v/v), methanol-1 M hydrochloric acid (4:1, v/v), methanol-0.1 M hydrochloric acid (4:1, v/v), and ethanol-1 M hydrochloric acid (9:1, v/v). The hydrophobic GSF absorbed on the SMIPs was supposed to be desorbed in the organic solvents. So we used less toxic organic solvents like methanol and ethanol as elution solvents in this step. The maximum recovery 89.4% was achieved by use of methanol and acetic acid (4:1, v/v) (Fig. 4c) In the last step, volume of elution solvent was also optimized. The results demonstrated that by using 2 mL volume of the elution solvent a maximum recovery was achieved (Fig. 4d).

Close

Fig. 4

Optimization of SMISPE procedure. (a): Optimization of Mass of SMIPs; (b): optimization of washing solvents; (c): elution solvents (d): Volume of elution solvent.

Export to PPT

So, 0.5 mL loading sample, 100 mg polymers, 2 mL water: methanol (7:3, v/v) as washing solvent to reduce interferences and 2 mL of methanol-acetic acid (4:1, v/v) used as an elution solvent were chosen as the optimized quantities for the procedure.

3.5 Method validation

The practicality of SMISPE method coupled with HPLC was assessed at the optimum conditions. A linear calibration curve of GSF spiked samples were achieved for the concentration range of 0.1–50 μg/mL. The correlation coefficient of linear calibration curve was (R² = 0.997). The method’s limit of detection and limit of quantification were 0.02 μg/mL and 0.05 μg/mL respectively when estimated as three to ten times signal to noise ratio (S/N). The experimental accuracy (intra-day and inter-day) were determined in terms of percentage recoveries by three repeated experiments at four concentration levels (0.1, 1, 10, 25 μg/mL) of spiked plasma samples. Finally, the precision at concentration levels (0.1, 1, 10, 25 μg/mL) was calculated as intra-day and inter-day mean percentage of relative standard deviations (RSD). The RSD % of intra-day and inter-day was less than 4.5%. The SMISPE-HPLC had good recoveries (90.7–97.7%) at the four concentration levels (Table 1). The Fig. 5 shows the recovery chromatogram of the drug standard (10 μg/mL), spiked plasma sample after SMISPE procedure (10 μg/mL) and blank plasma sample. After SMISPE method, the interference of impurities was reduced in plasma samples chromatogram, relative content of the drug was improved and a recovery of 97.7% was achieved.

Close

Fig. 5

Recovery from spiked plasma. (a): Blank plasma; (b): Drug standard 10 µg/mL; (c): Recovery from spiked plasma (10 µg/mL) after SMISPE.

Export to PPT

3.6 Detection of GSF in real plasma samples

The above established method was applied to real plasma samples in order to detect and separate the GSF. The blood plasma concentration was 0.8 μg/mL (after 5 hrs. drug dosing interval). The chromatogram of real plasma samples (Fig. 6) showed that after the SMISPE method the GSF was extracted from real plasma samples. The interference of impurities decreased and the quantity of GSF increased as compared to the chromatogram obtained by using SNIPs and C-18 SPE conventional SPE cartridge. The results demonstrated that the established method can meet the analytical requirements of GSF in plasma samples and can be used as a promising sample treatment technique for the determination and estimation of GSF in real plasma samples.

Close

Fig. 6

Detection of GSF in real plasma samples. (a): Griseofulvin standard (10 µg/mL); (b): plasma sample collected at 5 hrs., after a 50 mg/kg oral dose of griseofulvin, treated with SMISPE; (c): plasma sample collected at 5 hrs., after a 50 mg/kg oral dose of griseofulvin, treated with SNISPE; (d): plasma sample collected at 5 hrs., after a 50 mg/kg oral dose of griseofulvin, treated with C-18, SPE; (e): blank plasma.

Export to PPT

3.7 Reusability of SMIPs

The % recovery was determined in order to evaluate the performance of SMISPE after 8 repeated experiments. The results of reusability experiment (Fig. S.4) demonstrated that SMIPs showed good recoveries and stability even after it was recycled eight times. The recoveries after eight cycles were ≥70.1%, which was still significant for reuse. The results assured that the prepared polymer has good durability and can be reused 8 times with relatively low performance loss.

3.8 Comparison with different extraction methods

The SMISPE-HPLC method was compared in terms of extraction efficiency and sensitivity with previously reported methods. The results (Table 2) demonstrated that the developed SMISPE method has better recoveries as compared with previously reported SPE methods. Additionally, the extraction process is highly selective, accurate, specific, simple, and reproducible. However, some limitations are also associated with this method such as low yield of specific binding sites and high non-specific binding sites, tedious optimization method and slow evaporation of eluents.

Considering the advantages like cost, easy availability of instrument, selectivity, good recovery, high precision, less organic solvent consumption, less amount of sample and sorbent applied and reusability of sorbent material for multiple analysis best suited our SMISPE-HPLC method for pre-clinical and pharmacokinetic evaluation in plasma samples.